Urban Freight Intermodal Transport: An Analogue Theory Using Electrical Circuits
Paul Charles Beavis
A thesis submitted in fulfillment of the requirement for the degree of
Doctor of Philosophy
School of Civil and Environmental Engineering
University of New South Wales
Sydney, Australia
Urban Freight Intermodal Transport:
An Analogue Theory Using Electrical Circuits
Abstract
Container Seaports co-located with populations in Australian cities generate and attract inexorable, increasing discrete- diffuse flows which lead to landside access constraints and conflicts. Network models in freight transport are deficient in addressing demand management initiatives to ecologically re-structure the landside container freight task. Conventional freight planning stages road building and port expansion to facilitate mobility and container storage, and a vicious, inefficient cycle ensues. The miss-specification of modelling and planning frameworks is due to a conceptualisation of node impedance as a constraint phenomenon rather than as a necessary measure of precision.
This thesis pioneers an activity-based approach to theoretically support investigations towards the adoption of urban freight intermodalism. A necessary signature of true transhipment- consolidation network formulations is the capture of service generation and inter-temporal storage mechanisms. Intermodal terminals offer a means to retrofit existing road-rail networks so that modes interface and deliver consolidation and accessibility outcomes. The essences captured are the changing value proposition and precision relations necessary to harmonise terminal complex activity with rail operating forms.
Analytical relations are proposed using electrical circuit and process control theories to represent impedance relations based on new flux variables and sparse system parameters. This transhipment calculus allows for precision specifications to be measured according to novel “Figures of Merit” which assess storage-handling and throughput trade-offs. To show potential model applications case studies are presented of alternate system formats in Waste transportation and the management of Seaport-Hinterland conflicts. Attributes of Hinterland Absorptive Capability illustrated include rail sidings activation and bi-modal overflow to support terminal space management. Resulting design criteria indicate node impedance requirements to deliver dispatch rail payloads.
The contribution made in this thesis is a new measure of impedance which allows for the investigation of utilisation of the terminal and its network. The thesis outlines an authentic freight intermodal science to be developed for sketch-planning satellite intermodal functions. The model framework demonstrates complex bundling initiatives and acts as a load following mechanism for coordination of the freight task. This leads to future research to address the multi-modal, multi-commodity, flow problem. Overall, the thesis contributes to measuring the effectiveness of transport infrastructure stock by modelling terminal retrofit leverage options.
Originality Statement
‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’
Signed ……………………………………………......
Date ……………………………………………......
Our diagnosis of the vital systems of cities … remains primitive; therefore the basis of most forecasts is questionable (Ausubel and Herman, 1988)
One of the major objectives of science is to elucidate the dynamics of change
(Nicolis and Prigogine, 1977),
To Ruby and Max, Natalie and Lucinda,
Acknowledgements
I would like to firstly thank Drs. Graham Turner and James Lennox of the CSIRO who introduced me to the systems “model”, Australian Stocks and Flows Framework where I did my early analysis on transportation and infrastructure stocks. Through this work I became conversant with the seemingly imponderable obstacles in achieving de-materialisation of the freight task.
I would like to thank Francesca Porta and Francesca Machetti of Politecnico di Milano who worked as research students with me at UNSW over the Australian Spring of 2006 on the Waste Case Study and who helped construct the Powersim simulation. This helped me think about what was lacking in system dynamics and discrete event modelling for decision-support in intermodal planning. I am also grateful to Vioelia (Collex) Ltd who allowed me to visit their Matraville and Clyde Transfer Stations several times and bombard their managers with questions, which they answered comprehensively and cheerfully.
I am grateful for several academics who guided my research, in discussions or through their literature, into understanding the essence of intermodal operations: to Pete Tapio, of Finland Futures Institute, who helped me conceptualise de-materialisation efforts for transportation; to Johan Woxenius, formerly of Chalmers University of Technology, Gothenburg, who has written widely on the need to interface terminal operations with rail operating forms and who pricked my interest in researching a time composition of freight; to Chris Skinner of University of Sydney who suggested I consider electrical circuit theory in my search to better define freight impedance relationships; to Jean-Paul Rodrigue of Hofstra University for his innovative work on terminals in transport geography. These interfaces were epochal moments in the thesis development.
For the seaport-hinterland case study, I wish to thank Michael Pfeiffer of the Victorian Department of Transport for his work in developing feasible shuttle intermodal schedules for Dandenong and Altona and advising me on the implications of different intermodal sidings configurations. Likewise I am indebted to Messers Adrian Polton and Hugh Gaynor of the Department of Transport for sharing their wisdom on freight train operations. These discussions added invaluably to the credibility of the intermodal operations and train service dispatch scenarios discussed.
I wish to thank Kim Hassall of University of Melbourne for his assistance in stressing the need for model tractability on the basis of the governance and business rules it could test.
To my supervisors, Stephen Moore, Iain MacGill, and John Black, thank you for prompting and prodding me to find and then prototype the essence of my thesis. Thanks also for your moral support over at times long distances.
To my father, Charles, who helped me understand the facts and some of the art in electrical circuit theory, I give you my love and admiration.
To my partner Thomas and the rest of my family- your support has been tremendous.
Paul Charles Beavis
Melbourne, April 2010
Publications
This thesis is underpinned by a number of publications:
Chapter 1 scope was in part based on a workshop held in Sydney on 28th May, 2006 and the resulting report:
Beavis, P., E. Blakely, J.A Black (2006) Critical Transportation Infrastructure in a Global Warming Future: Protecting NSW Seaports and their Hinterland. Report on Workshop held 25th May, 2006 at University of Sydney by the Botany Bay Studies Unit, UNSW and Planning Research Centre, University of Sydney: 16pp.
funded by the Commonwealth Scientific, Industry and Research Organisation (CSIRO) Education Endowment Fund.
Chapter 2 has been presented in part in:
Beavis, P., J.A. Black, J. Lennox, G. Turner, S. Moore (2009) Industrial Ecology Futures Scenarios: The Design Approach in Transportation. In Dynamics of Industrial Eco-Systems. Ruth and Davidsdottir (eds.) Edward Elgar. Pp. 179-200.
Chapter 3 has been presented in:
Beavis, P., J.A. Black, R.Golzar (2005) Functional Specification of Strategic Urban Freight Models: Modelling Attributes for the Port and Landside Freight Task in Sydney. Journal of The East Asia Society of Transportation Studies (EASTS) Vol. 6, 2005 16pp. which was based on an industry research report commissioned by the Australian Toll Roads operator, Transurban Infrastructure Developments.
Chapters 5, 6, and 8 have been outlined in:
Beavis, P. J.A.Black, I.MacGill, J.Woxenius, S.Moore (2007) Distributed Function Hinterland: Functional Design of Container Intermodal Terminal Systems for Sydney. World Conference on Transport Research (WCTRS), University of California, Berkeley, 24th -28th June 2007
Chapter 7 has been presented in part in:
Beavis, P., S. Bartoli, F. Michetti, F. Porta, S. Moore (2006) Intermodal Transportation Design for Waste Recovery 28th Conference of Australian Institutes of Transport Research (CAITR) 6-8th December, University of New South Wales
Glossary
The definitions in this Glossary specific to this thesis are designated with the prefix [L]. Where these terms have alternate meanings, such as in electrical power circuit theory, this will be recognised by [E]. The origin of the definitions used in this thesis indicate the work in progress in applying the analogue. Authors referred to are those who have coined and applied these terms.
Terms Meaning
Accessibility [L] A measure of effective transport infrastructure under alternate flow- load regimes.
Active Arrival Duration [L] The period of distribution of arrival carriers to map to a specific dispatch service. The length of this period affects the area required for stack reservation, consequently consuming space resources and affecting throughput metrics of TEU/ ha. (Taleb-Ibrahimi et al., 1993)
This time domain metric maps to the timing domain by the synchronisation band measure.
Activity Based [L] Modelling approach where intermediate terminal nodes affect Approach the level of freight activity over the network through consolidation and distribution. The approach represents an activity pattern arises from the scheduling of activities. This pattern is based on rules (McNally, 2000).
Agility The performance of flexible transformation capacity. Ability of terminal operations to respond to changing specifications in input and output flows. Freight terminals must possess attributes of flexible capacity before they may exhibit agile performance (Marlow and Paixao, 2004)
Analogue A comparison with an alternate science theory with the objective of gaining insights into a system which is pre-paradigmatic or less developed.
Anthroposphere Also known as the technosphere, this is a complex system of energy, material and information flows in space driven by human activity. It is a term for the collection of industrial eco-systems. (Brunner and Rechberger, 2004)
Automatic Generation [E] Mechanism in power distribution theory used for load Control following control to activate generative unit processes for
i
ancillary services between economic dispatch time steps. [L] Mechanism in deployment of logistical unit processes to form the impedance of the itinerary. This is discussed in Appendix F.
B-double truck Prime mover with two load carrying articulated trailers; nominally up to 25m in length and up to about 62 tonnes gross vehicle mass
Back-loading [L] The proportion of return trips with a full load. This affects the payload of the total return trip.
Bandwidth [E]Width of the passband.
[L] Representative of active arrival duration permitted for a specific carrier pair and anticipates the reservation area required for a stack.
Batch Consolidation of consignment units for storage or shipment. Also known as bundling.
Bimodal When intermodal terminals can improve the productivity of both road and rail systems –these are bimodal systems: i.e. they involve both transmodal (intra-modal) and intermodal transhipment operations.
In North America, Bimodal operations refers to the use of truck trailer as a container wagon. This technology is not addressed directly in this thesis.
Bode plot [E] A graphical technique which indicates the frequency response of a signal over a circuit and yields voltage and current information. There are both magnitude and phase angle Bode plots. The phase angle indicates the reactive impedance caused by the signal- circuit interaction.
[L] A figure of merit of value density or consignment flux to carrier synchronisation range which assesses the effect of capacity and handling controls. The relationship between LUP or terminal productivity and freight flow profile.
Bulk Queue [L]The business rule mechanism of passing flow to the next Logistical Unit Process (LUP) on the basis of sufficient accumulation or specific elapsed time to meet a scheduled service.
Business Rules Rules in operating a terminal system- may be over short- medium- to long term horizons and cover processes in freight terminals as well as terminal networks. The ability of business rules to achieve desired outcomes will depend on the terminal infrastructure and layout where they are applied. They test aspects
ii
of controlling complex systems.
Capability The measure of endogenous changes in the state of supply to variations in source flows and load flows. Saturation constraints may be variable. A measure of the flexible capacity of a number of linked activity units in a terminal node.
Capacity The measure of a terminal unit’s ability to store or pass through consignments based on fixed saturation constraints.
Capacity Risk [E] The measure of the likelihood that a proportion of the system will be available to function at a certain load.
[L] The risk that system infrastructure and resources will not be available under increasing throughput rates. Alternate configuration of terminals and rail operating forms may reduce the capacity-risk relationship by linking storage processes and providing temporal accessibility.
Carrier Pair / Carrier A transport relation between source flow and dispatch flow Transfer Pair tracked through a gateway terminal to observe the impact of conversion on value and precision relations.
A carrier pair from an export perspective may represent a number of source inflows mapped to one specific train dispatch service.
Circulation Rates [L] The productivity of truck fleets and rail rolling stock in completing a tour (return trip with intermediate stops). This measure depends on the fleet’s average payload based on back- loading rates. This measure is different from circulation time which tries to minimise the time taken on the network through measures which increase speed.
Combined Transport Use a combination of transport modes to increase the capability of the system with the minimum physical augmentation of the road network.
Commitment [logistical] is the step prior to dispatch allocation to assess the availability of resources (i.e. storage slots) for supply deployment.(Wood and Wollenberg, 1996)
Compensation Control [E] Application of control elements to achieve a sufficient response; [L] Application of operational business rules to achieve a sufficient response in storage and /or payload characteristics. They are ancillary services to tie the infrastructure response with the behaviour of actors.
Complex Bundling Coordinating multiple Beginning sources with multiple End loads. For the purposes of this thesis includes stuffing and unstuffing functions at the terminal
iii
Complex conjugate [E] The matching of complex impedance between signal and terminal response to minimise impedance [L] The design response of an LUP for source and load schedules which minimizes the value lost. Used as a space management design technique to maximize efficiency of terminals.
Complex Impedance [E] Reactive and resistive impedances
[L] The work requirements to manage the throughput of a certain consignment profile to match rail dispatch and arrival schedules. It is a measure of terminal inertia and cost.
Consignment The packaged commodity in transit. This has information attributes of the carrier relationship between origin and destination. In this thesis the consignment represents the combined carrier, container and mass being transferred to simplify the exposition.
Consistency An attribute of precision. Considers interoperability of flows
Constitutional laws [E]: Laws governing the relationships between driving forces and driven forces in a circuit; [L] Laws governing impedance relationships in a stencil forming the integro-differential equations to represent Logistical Unit Process responses to flows.
Controllability Ability to achieve a sufficient response under a range of flows
Control Strategy Application of control elements to achieve a sufficient response. This may or may not be an optimal control strategy.
Coordination The allocation of flows and resources over a gateway of networks. A more useful approach under conditions of asymmetrical flows and congested networks. Salient mechanisms are inter-temporal storage and generative scaling
Cross-docking Rapid direct transhipment of containers at intermodal terminals. There is little or no storage.
Cumulative Curves Depict the steady-state relation between inflow and outflow over a period
Damping Factor [E] a parameter of the characteristic equation which relates bandwidth to centre frequency and thus represents the Quality factor.
[L] A design ratio to determine handling- storage impedance to achieve batching and throughput objectives.
Demand Management Application of frictional controls (pricing, physical constraints, alternate mode pathways) to reduce transport activity
iv
Demand Matching Use of predictive models to address projected demand with increases in physical supply, in order to minimise increase in equilibrium travel time.
Dematerialisation A concept which may be considered relative and absolute. Relative dematerialisation is the reduction in inputs per unit output or per capita for a particular activity. Absolute dematerialisation is the reduction in total material requirements for a nation (across industrial eco-systems) or globally (those nations with which it trades). Absolute dematerialisation also includes the multiple of population size. De-linking human welfare from economic activity is another term for dematerialisation.
Dependent Variables Performance output. Set load specifications.
Design Value Rating [L] The dispatch payload desired. Equivalent to: [E] the power rating.
Discrete- diffuse The flow rhythm of consignments. Discrete-diffuse sequences sequences represent patterns of consolidation and distribution (i.e.: a network without intermodal terminals will have more diffuse truck operation). Diffuse flows are those small consignments of many carriers; discrete flows are consignments on scheduled carriers such as train or vessels. Discrete flows can often be measured as pulse signals.
Dispatch [L] Equivalent to the allocation step in transportation Four-Step Modelling. In power distribution theory this step may be combined with the test of availability, called unit commitment.
Dispositive elasticity (L) Opportunistic use of an infrastructure network characterized by high trip frequencies: this consumes temporal and spatial resources, contributing to network congestion.
Dissipative Structures Systems with few closed loops. The landside container logistics structure may be seen as dissipative as actor behaviour results in low levels of truck unit loading and low back-loading. This is due to poor coordination in depositing and collecting containers.
Distributed Resource Systems where resources are not centralized. This leads to Systems opportunities to reduce dissipative trips (less than full load) on links between nodes
Distributed Storage An aspect of distributed resource systems where storages functions can be located so that bundling opportunities for consolidated flows may occur
Dry Port An intermodal terminal where there is a distribution of functions from the Seaport. It is not simply an autonomous node with rail/ v
road connections to the seaport. There is a planning of functions which demonstrably complement the seaport.
Dwell time Duration a consignment (container) resides in a storage area. Duration will affect the throughput capacity of the terminal
Ecological Re- The re-organisation of processes among inter-connecting Structuring industrial eco-systems to achieve de-linking of human welfare from economic activity which consumes energy and materials. It represents transitions management in industrial eco-systems.
Effective Interfaces Interfaces wherein there is a minimum delay time and cost for the function required.
Evolution of System [L] The recognition that increasing demand requires supply Deficiencies constraints to be overcome by increasing the injection of infrastructure capacity.(Kanafani, 1982)
Feasibility The net cost-benefit of operating a system or terminal. Also extends to assessing modelling terminal operations as physically realisable i.e. have causation and can be controlled.
Fuel Scheduling [E] Mechanism of energy generative allocation over a long term horizon guiding infrastructure investment decisions in Power Distribution Theory. May involve alternate generative sources, such as hydro-electric, photovoltaic, and gas turbine and coal- fueled power plants (Wood and Wollenberg, 1996).
[L] Method in network coordination of a freight system as a distributed resource system to test long term infrastructure investment decisions, thus tying together the three tiers of freight modeling – strategic, tactical and operational.
Figures of Merit Trade-off relations between Performance Indicators used toward discerning an improved design outcome (Mischke, 1980)
Fine meshed systems [L] Transport networks where there are dense logistical links between a terminal and its hinterland. This encourages complex bundling. Characteristic of these systems are short line hauls and multiple stops between final origin and destination.
FEU Forty Foot Equivalent Unit size container
Fidelity Accuracy of model to represent behavior in reality
Figures of Merit Trade-Off relationships which indicate the relative performance of alternate system formats
vi
Filters [L] Modelling technique to establish precision relations between carrier pairs.
Flexible capacity Considers the ability to handle a variability of inflows whilst still meeting Level of Service requirements.
Fourier Series [E] Used to decompose the fundamental frequency signal into a series of harmonics.
[L] Used to decompose the train dispatch schedule into contributing carrier inputs according to their magnitude and timing.
Freight Exposure Measure coined by Hassall (2008) to express the comprehensive impacts of trucking in metropolitan areas. This measure drives analysis and design of network- node alternatives to consolidate loads and reduce truck trips
Freight Intensity Measure coined by Roth (2000) to express the dissipative work of the logistical structure. Units are in kilometres/ tonne.
Frequency [Electrical]; The reoccurrence interval of a signal
[Logistical] is a variable of timing. It denotes the pattern of time utilisation on space resources (such as train schedules on terminal sidings and rail mainline capacity.
Functional Discovery Measure coined by Gifford and Garrison (1993) to express a reduction in Material Input per Unit Service due to improvements in service. It is related to the Factor 4 debate in de-materialisation.
Gateway A pivotal node that links networks. A means to evaluate consolidation networks by collapsing the network depiction to critical connector nodes.
Generative capacity [L] The ability of a terminal to achieve a certain generative scaling called for.
Generative Scaling/ [E]: The engagement of generator units according to relative Service Generation speed-droop characteristics with changes in frequency. [L]: Control Engagement of terminals and their unit processes forming particular itineraries according to availability and assignment criteria.
Goods Are commodities that are carried as consignments.
Harmonisation The balancing of operational activities with logistical / commercial relationships. This is demonstrated by the precision
vii
of system resources
Headway The inter-arrival time of the means of transport; inverse of service frequency
High Productivity Trucks which can carry multiple containers such as B-doubles Vehicle (HPV) which can carry 2 TEUs or Super- B-Doubles which can carry 2 TEUs plus one FEU.
Hinterland A transport geography term referring to the landside trade catchment of the seaport. Can also refer to the logistical catchment of the intermodal terminal.
Hub A terminal node which coordinates flows from alternate origins to alternate destinations
Impedance [E] Measured in terms of a vector relationship of resistance (real) and reactance (imaginary); [L] Measure of work (depicted as saturation inertia) and precision requirements to meet the load required.
Impedance Attributes [L] Consists of transportation delay, resource productivity constraints; load matching for precision; and Value flux
Impedance Matching [E]See complex conjugates [L] Ensures Logistical Unit Processes are linked according to physical flow and precision criteria.
Independent Variables Performance variables
Intermodal Balance Synergy by connecting modal networks rather than dissipation in modes operating singly without complementary activities in the transportation task (Vuchic, 1999). Intermodal Balance is made operative in this thesis by proposing Figures of Merit in terminal tactical design that may characterise Hinterland Absorptive Capability- covering utilisation, productivity and effectiveness measures.
Intermodal Production The system boundary involving terminals stand-alone or Systems connected to other terminals, rail mainline and seaport facilities. Pre- and Post Haulage is not directly considered. It is implied.
Intermodalism The logistical science of facilitating consignment transfer between two or more modes i.e. road- rail. For this thesis its salient characteristics are the fostering of complementary modes and building accessibility opportunities.
Interoperability The degree of smoothness in the interface of consignment exchange between carrier pairs. Logistical friction is the measure of poor interoperability.
viii
Inter-temporal Storage Storage that allows the positioning of consignments to transship to alternate scheduled headways of their destination carrier.
Itinerary A pathway of commitment (availability) and dispatch (allocation).
Investment Optimisation Considers the most effective return of infrastructure deployment under scenarios of alternate flow rating and transhipment opportunities.
Landside The terminals and networks that support the seaport terminal.
Landside Absorptive Ability of infrastructure node elements to manage freight flows Capability under alternate load conditions.
Laplace transfer A mathematical representation (transform) of Ordinary functions Differential Equations. Can be used to represent Unit process activity in converting input signals to output signals.
Load The flux specification for output of a Logistical Unit Process. This drives design of each terminal Logistical Unit Process. Capacitive load devices are used to store additional consignment flows (indirect transshipment). Inductive load devices represent transshipment handling resources for direct transshipment activity.
Load Following / Load [E] The search process or compensation mechanism whereby Matching Voltage is generated from one or a number of sources according to optima criteria. The intention is to minimise the energy supply effort (cost) for the power load demanded.
[L] Load-based design where Logistical Unit Processes and intermodal and other freight terminals are designed to meet dispatch requirements according to consignment characteristics of quantity, type, destination, and timing. This enables an activity based approach for dynamic allocation of a freight itinerary rather than on the basis of a constraint only approach.
Load Regulation Process control of Logistical Unit Processes to maintain output volume. This could include modification of duty cycle of a train dispatch schedule to ensure sufficient source flow is carried through to minimum consignment load to be dispatched.
Load Shunting [E] The addition of reactive devices to control a voltage drop.
[L] Can be used to provide storage resources which delay dispatch but allow complex bundling.
Logistical Friction The physical and logistical impedances to rapid throughput of consignments. (Hesse and Rodrigue, 2004)
ix
Logistical Unit Unit process within a terminal where impedance may be designed Processes (LUP) and analysed. This thesis applies an electrical circuit analogue to act as a stencil for LUPs in order to characterise their flux relationships
Logistics Study of freight movements and actor relations. This may include supply chain relations and/or post supply chain relations in consolidation and distribution.
Metabolism An industrial ecology metaphor for the operation of industrial eco-systems and a means to describe their patterns of resource use.
Multi-commodity Flow The question of allocating multiple carrier pair relationships over Problem a network and over modes with transhipment activity. Also known as the Multi-Commodity, Multimodal flow problem (Crainic, 1986). Conventional modelling allocates upon the basis of substitution / competition, rather than complimentarity, and is thus not well defined for intermodal collaborative networks.
Multimodalism The policy approach which (whether intentionally or otherwise) fosters multiple means of transport between the same origin and destination. This can lead to the preference for a more convenient means of transport over another which may be less sustainable. For instance, a motorway is introduced on a parallel route with a rail corridor leads to growth in road trips and a reduction in rail patronage.
Multipath Coupling The available itinerary of paths which can be taken by a consignment through a terminal during the production cycle. The degree of multipath coupling refers to the intensity of links between terminal unit processes.
Multiple Input Multiple Process and terminals which transform several inputs into several Output output.
Network [L] A graphical space of nodes and links over which links between attractors and generators are allocated and transport flows are assigned routes.
Node [L] A fixed location in a network where intermediate activities may take place, such as consolidation, storage i.e. an intermodal terminal.
[E] A position on a circuit stencil where branches of the circuit meet. Used to calculate energy balances according to constitutional laws
Ontology A set of particular assumptions about the nature of being and
x
reality (Allenby, 2006).
Open Access Regime An equitable and transparent cost structure for all actors in the freight task (ARTC, 2002) ; the governing means whereby the operational roles of actors harmonise with the prevailing logistical structure thus reducing conflicts (SFCNSW, 2007).
Operational Amplifiers [E] Active electrical circuits [L] Unit of activity used to form a stencil.
Origin or destination [L] Beginning or End location of the container trip. Ultimate orgins and destinations refer also to less than container load consignments.
Parseval’s theorem [E] used to analyse the energy content of a signal.
[L] used to analyse the value intensity of a modal flow profile for the consignment type and value the profile has to a carrier pair mode to which the consignments will be transhipped.
Payload [L] Utilisation value of the means of transport. Defined by magnitude and timing.
Peak Shaving Demand management technique to reduce the peak of truck freight
Percent Overshoot [E]: An indicator of stability in transient analysis
Phase Angle See: Storage lag coefficient
Phase Plot [E]: The relationship between reactive elements and frequency for an electrical stencil
[L] The phase plot represent evolved storage lag indicating the degree of impedance with a certain consignment flow profile.
Phasor Diagram [E]: Resolution of resistive and reactive impedances acting in a stencil for input and output system variables of voltage and current; [L] Resolution of impedances for stencil design for desired payload attributes of value density and consignment rate
Physically Realisable [E]: Criteria of process control where a system is physically realisable if 1) it is causal (that a flux relationship around a process can be characterised) and 2) if it is controllable (degree of stability according to alternate flow regimes) (Kuo, 1966)
Polar notation An algebraic representation of integro-differential equations involving real and imaginary parts.
Port Capacity as The operational Capacity of a berth or terminal is the maximum Practical Operating cargo throughput which can be achieved at an acceptable level of
xi
Capacity service. Capacity is expressed in terms of tonnes throughput per annum or TEUs per annum
Power Distribution The theory of power load (demand) and generation (supply) over Theory a distribution network. This can involve the time horizon mechanisms of load following, unit commitment, economic dispatch, and fuel scheduling.
Power Rating [E] Elements in a circuit that have a specific load draw power from that circuit
[L] Design value rating of an activity designated in the Logistical Unit Process (LUP), such as dispatch of a train service from a siding. Design value rating is measured by the consignment flux, the value density held and the impedance load of the LUP.
Pre- and Post Haulage The freight task by trucks in collection and delivery of (PPH) consignments to and from an intermodal terminal.
Precision A measure of concurrent systems – involves synchronisation, the matching of input and output flows of different mode carrier classes based on a schedule; and consistency, the matching of such flows based on common destination or other criteria.
Process Control The control of variable performance due to disturbance to ensure output and level of service boundaries are met.
Production Cycle The transformation of a consignment through terminal activities stuffing/ unstuffing, storage, batching, and transhipment prior to dispatch or distribution.
Productivity Performs to the effectiveness of outcome of terminal production, such as costs per unit throughput. Efficiency refers to sub- measures, such as handling rate of containers. Note this is may be a sub-optimal measure with terminal utilisation when considering performance in terms of hinterland absorptive capability.
Pulse Correspondence [L] Synchronisation of payloads between carrier services using load matching storages. This allows diffuse truck inflows to be mapped and consolidated onto train services. The concept also pertains to train to train transshipment for like consignment destinations (Bostel and Dejax, 1998).
Quality Factor [E] Ratio of centre frequency to passband.
[L] A composite Figure of Merit assessing the productivity of Logistical Unit Processes and Intermodal Terminals and their flux characteristics in catering for certain consignment flow transformations. It assesses the value stored against the value dissipated in operating, demurrage, and delay costs.
xii
Rail Operating Forms Train formation with specific attributes of physical dimensions and itinerary to match their purpose. The identification of train types anticipates the functional requirements of the terminal.
Reactance [E] imaginary part of impedance; [L]: Level of storage or resource consumption
Resistance [E] Real part of impedance; [L] variable costs with consignment flow rate
Resonance [E] When a circuit is driven at the resonance frequency- the frequency of the forcing function is the same as the natural frequency of the circuit. No net imaginary impedance 9reactance) is recorded.
[L] Optimum storage utilisation. Storages are not saturated and handling resources are not limited given the flow regime profile.
Resources Are the means of transport, handling equipment and terminal infrastructure where utilisation and saturation may be measured.
Retrofit The practice of symbiotic investment decisions where combined transportation outcomes are fostered and thus the evolution of system deficiencies due to single mode networks is reduced.
The physical and logical restructure of existing infrastructure. It is the physical action of transitions management.
Schedule A periodic timetable of services. Source schedule is the expected inflow profile; Load schedule is the expected outflow profile. Schedule is characterized by value (mass and commodity type), number of services, and synchronisation band. Precision of carrier pairs matching schedules can be gauged through analysis of how well their synchronisation bands coincide.
Sequence A path of unit processes in series or parallel.
Sidings Rail tracks off a mainline used to stable or hold train sets.
Sketch Planning Strategic thinking to achieve better outcomes in infrastructure deployment
Source The input flow to a terminal
Specification Performance requirements of outputs as well as Level of Service ranges for an LUP or terminal.
Spectrum Analysis [E] Analysis of the energy content of a signal of specific frequency band
[L] Analysis of the value intensity of consignment carrier flows
xiii
(value over a specific timing period for a carrier service).
Stability [E]: Measures of responses in Transient analysis which include percent overshoot and settling time. Steady state stability is also measured.
Steady State Analysis when natural response of terminal has settled from forced response.
Stencil A diagram representing constitutional laws of driving and driven forces used to formulate differential equations to calculate value and volume fluxes and changes in precision relations .An electrical circuit analogue is used in this thesis.
Storage Claim A system capability measure which records the number of load units exchanged via the storage area per hub or rail operating form (Terminet, 1998, p.24). This relates the storages at a terminal with the system format: the rail schedules it serves and the operations of other linked terminals.
Storage Lag Coefficient which determines the relationship between through and across variables, i.e.: demands placed upon resource per unit output, such as storage requirement for output.
Stuffing/unstuffing The un/bundling of pallets from containers
Super B- double Prime mover with two 40 foot long container carrying trailers
Synchronisation An attribute of precision relations involving impedance matching, value amount and timing. The Synchronisation band is the value spectrum of consignments flows that corresponds to a desirable time interval
System Dynamics The study of feedback loops in human and natural systems.
System Format The physical, engineering, and economic aspects of an industrial eco-system (Beavis et al., 2009).
Tactical Modelling Refers to the medium term ‘time horizon’. Can involve physical network design (including physical supply aspects such as transhipment options) and service network design (including services scheduling via itineraries). (Crainic, 1987)
Technological Envelope [E]The range of technically feasible behavior of an industry (i.e. Electrical Industry) given the organization of physical elements. (MacGill, 2006)
[L] The manner in which an intermodal system format is deployed through use of operational business rules. This will impact on the hinterland absorptive capability of a system.
xiv
Tense flux Driving force in supply chains whereby terminal components must respond in a synchronized and consistent fashion. (Rodrigue, 1999)
TEU Twenty Foot Equivalent Unit container.
Tie line model control [E]: Links neighbouring systems in a network where purchase and sale of power among systems is profitable and frequency can be restored if there is generating disruption. [L]: Regions of freight activity can share resources to optimize the freight task (minimize road freight intensity) in their region such as sharing empty containers and/or access slots on train shuttles.
Time slots Time windows where consignment carriers can access resources. For instance, trucks carrying export containers destined for the seaport have booked time slots of 15 minute windows beyond which the truck must rebook and/or pay a penalty.
Time Window A period of infrastructure availability for a service to run or for a resource to be accessed. Related to time slots.
Timing Synchronisation. It is the aspect of flow information designating the carrier pair schedule to which the consignment belongs or is scheduled for. It does not necessarily entail that the consignment flow will meet that output schedule.
Timing Profile Carrier arrival / distribution and dispatch; the profiles represent, for instance, the multi-commodity flow problem of different truck origin and different vessel destination intentions, and how an intermediate node sorts these into a desired manifest mapping.
Terminal Complex The configuration of terminal functions which allows alternative Activity itineraries to be explored according desired transformations and dispatch outcomes.
Tractability Assessment criteria on model effectiveness to grant valuable insights into phenomena investigated.
Transformations Physical and information (timing) changes that occur to consignments in a terminal.
Transhipment Intensity The degree to which a terminal caters for flux of direct or indirect transhipment, where direct transhipment involves minimum storage.
Transhipment Calculus The mathematical theory of interface relations between intermodal terminal and the flow profiles of the means of transport (modes). The theory also extends to representing the interconnectivity between terminals.
xv
Transhipment Problem [L] The study of allocating flows over a network with intermediate points between origin and destination.
Transient Response over a particular time period: used to assess response to sudden changes in input and to compensate for these.
Transitions Management Manage trajectories of performance of large tech-system(s) over long time periods. The motivation is exploratory and explanatory.
Transmodal Intra-modal operations, such as train-train interchange. May / may not involve consolidation.
Transportation Cycle Cycle involving the means of transportation. This is affected by:
(i) The need to exchange goods. (ii) The logistical structure of operational requirements and actor relations. Urban Leverage The process of utilizing the existing stock of infrastructure to reduce the material intensity per unit service. Retrofitting can be seen as urban leverage.
Value Density The inverse of freight intensity: (t / km)
This acts as the driving force in devising impedance relations.
Value Flux The net value of a batched assignment according to the particular itinerary it takes through a terminal.
Value Intensity The value content of a particular carrier pair as a proportion of the total dispatch payload.
Value of Travel Time Conventional transport network models allocate routes taken Savings according to gravity (and other) mechanisms, which ration the exchange path between attractors and generators according to the least cost path. Cost is equated to time delay.
Value Proposition The net value flux after terminal impedance compared with alternate road based routes.
xvi
TABLE OF CONTENTS
GLOSSARY………………………………………………………………………………………………………………………………………I
1 INTRODUCTION ...... 1
1.1 PRESENTATION OF PROBLEM ...... 1 1.2 LEVEL OF TECHNICAL KNOWLEDGE AND GAPS ...... 5 1.3 OBJECTIVES AND THE RESEARCH PERSPECTIVE ...... 8 1.4 METHOD ...... 11 1.5 THESIS STRUCTURE ...... 16 1.6 FINDINGS ...... 19 1.7 CONTRIBUTION TO KNOWLEDGE ...... 19 1.7.1 Terminal Analytical Mechanisms ...... 19 1.7.2 Novel Indicators and Figures of Merit ...... 21 1.7.3 Anticipating Future Research ...... 23 2 RESEARCH DESIGN FOR SUSTAINABLE FREIGHT SYSTEM FORMATS ...... 24
2.1 INTRODUCTION ...... 24 2.1.1 Chapter Guide ...... 24 2.1.2 Freight Task as a Large Technological System ...... 25 2.2 THE AMBITION OF COMBINED TRANSPORTATION ...... 27 2.2.1 Ecologically Restructuring the Freight task ...... 27 2.2.2 The Nature of Combined Transportation ...... 27 2.2.3 Desirable Frame of Reference for Combined Transportation Modelling Framework ...... 28 2.3 MATERIALISATION OF THE FREIGHT TASK ...... 31 2.3.1 Predicament and Measurement ...... 31 2.3.2 Recasting Measurement to Drive Improvements ...... 32 2.4 THE SKETCH PLANNING ENDEAVOUR AS TRANSITIONS MANAGEMENT ...... 33 2.4.1 Transitions Management Technique for Sustainable Futures ...... 33 2.4.2 Freight Futures Simulation Example ...... 35 2.5 FREIGHT METABOLISM ...... 37 2.5.1 Structure of the Urban Freight Task ...... 37 2.5.2 Materialising Pressures of Land Use Devolution ...... 40 2.5.3 Modelling Paradigms and Claims ...... 41 2.5.3.1 The Perceived Need for Speed ...... 41 2.5.3.2 City Logistics and Reverse Logistics ...... 43 2.6 CONDITIONS IN FREIGHT INDUSTRIAL ECO-SYSTEMS ...... 45 2.6.1 Urban Landside International Sea Container Task ...... 45 2.6.1.1 Overview...... 45 2.6.1.2 Sydney Condition ...... 45 2.6.1.3 The Lack of Harmonisation ...... 47 2.6.2 Integrated Waste Management ...... 48 2.6.2.1 Internationally ...... 48 2.6.2.2 Sydney ...... 49 2.7 ROLE OF INTERMODAL TERMINALS ...... 49 2.7.1 New Concept of Impedance ...... 49 2.7.2 Accessibility as a Measure of Effective Infrastructure ...... 50 2.7.3 From Capacity to Capability: A New Concept in Sustainable Infrastructure Supply ...... 50 2.7.4 Controlling the Evolution of System Deficiencies ...... 51
2.7.5 Urban Leverage by Inter-temporal Generation and Storages ...... 54 2.7.6 Facilitating the Cost-Value Jump ...... 55 2.8 REQUIREMENTS OF INTERMODAL TERMINALS ...... 56 2.8.1 Physical Essence in a Freight Transportation System ...... 56 2.8.2 The Re-formulation of Logistical Friction ...... 57 2.8.3 Conditions for Financial Feasibility of Intermodal terminals ...... 58 2.9 PERFORMANCE INDICATOR FRAMEWORKS ...... 58 2.9.1 Temporal Performance Analysis of Intermodal Retrofits ...... 58 2.9.2 Hierarchy Indicators of Intermodal Capability...... 62 2.10 CONCLUDING COMMENTS FOR CONCEPTUAL DEVELOPMENT ...... 66 3 GAP ANALYSIS IN FREIGHT MODELLING FOR INTERMODAL OUTCOMES ...... 67
3.1 INTRODUCTION ...... 67 3.1.1 Chapter Guide...... 67 3.1.2 Research Design Drivers Which Limit the Study ...... 68 3.1.3 Addressing the Information Needs of Infrastructure Suppliers ...... 69 3.2 CONCEPTUAL MODEL ...... 69 3.3 SYSTEM DEFICIENCIES AND POLICY PLANNING RESPONSES ...... 72 3.3.1 Strategic Planning and Investment for Urban Landside Container Freight Task ...... 72 3.3.2 Land Use Planning for Seaport Functions...... 74 3.4 INTERFACE ATTRIBUTES SPECIFIC TO URBAN INTERMODALISM ...... 76 3.4.1 Hinterland Bundling and Consolidation: Land Use Analysis ...... 76 3.4.2 Benefits of Short Haul ...... 76 3.4.3 Performance Indicators ...... 77 3.5 NETWORK FREIGHT MODELLING TAXONOMY TOOLBOX ...... 79 3.5.1 Network Model Archetypes ...... 79 3.5.2 Freight Modelling Methods in the Transhipment Problem ...... 83 3.5.2.1 The Four Step Network Model ...... 83 3.5.2.2 Incorporating Transhipment ...... 85 3.5.2.3 Multi-Modal, Multi-Commodity Approaches ...... 85 3.5.2.4 Intermodal Service Design ...... 86 3.6 FACILITIES MANAGEMENT MODELS ...... 88 3.6.1 Frontiers in Design and Analysis of Container Terminals ...... 88 3.6.2 Models of Inventory and Distribution Theory ...... 90 3.6.3 Lessons from Terminal Design Methods ...... 91 3.7 MODEL HIERARCHY GAPS ...... 92 3.7.1 Critique of Network Methods to Address Bundling and Transhipment Phenomena...... 92 3.7.2 Gaps and Linkage Requirements across Model Hierarchy ...... 94 3.7.3 Terminal and Hub Activity and the Time Composition of Freight ...... 95 3.7.4 Advancing Activity Based Approaches ...... 96 3.7.5 New Rail Operating Forms and Handling Technologies ...... 98 3.7.6 Network Analysis Needs for Physical and Service Design ...... 99 3.8 ASSESSING CAPACITY MEASURES ...... 101 3.8.1 Network Capacity Measures ...... 101 3.8.2 Capacity Assessment of Terminals ...... 101 3.8.3 Misspecification of Capacity in Assessing Terminal Performance...... 104 3.8.4 Opportunities with a Process Control Approach ...... 104 3.9 ANTICIPATING THE CONNECTION OF TERMINAL DESIGN TO COORDINATION APPROACHES ...... 105 3.10 INTERMODAL MODELLING METHODS AND GAPS ...... 107 3.10.1 Straddling Horizons in Space and Time ...... 107 3.10.2 The Perils of Neglecting Urban Intermodalism...... 108 3.10.3 Outstanding Research Agenda for Urban Intermodalism ...... 109 3.11 SYNTHESIS MAPPING INTERMODAL INFORMATION NEEDS TO MATHEMATICAL MODELS ...... 110
3.11.1 Devising a Functional Specification ...... 110 3.11.2 The Need for Design Figures of Merit ...... 111 3.12 CONCLUSIONS FOR A NEW PHENOMOLOGICAL APPROACH ...... 112 4 FUNCTIONAL REQUIREMENTS FOR SKETCH PLANNING URBAN INTERMODAL PRODUCTION SYSTEMS ...... 115
4.1 INTRODUCTION ...... 115 4.1.1 Chapter Guide ...... 115 4.1.2 Two-Node System Scope...... 117 4.1.3 Need for an Activity Based Theory of Node Analysis ...... 120 4.1.3.1 Relating Terminal Complex Activity to Rail Operating Forms ...... 120 4.1.3.2 Addressing Pulse Flow Phenomenon ...... 121 4.1.3.3 The Role of a Load Following Response Mechanism ...... 121 4.2 DESIRABLE SYSTEM FORMAT OF COMBINED TRANSPORTATION ...... 122 4.3 BUSINESS RULES AND THEIR APPLICATION ...... 123 4.4 POSITIONING ELEMENTS OF THE DESIGN CYCLE FOR MODEL DEVELOPMENT ...... 125 4.4.1 Overview ...... 125 4.4.2 Attributes of Continuous Approximation Approaches ...... 125 4.4.3 Probing Logistical Relationships through Analogy ...... 126 4.5 CONCEPTUALISATION OF INTERMODAL PRODUCTION SYSTEMS ...... 127 4.5.1 Introduction ...... 127 4.5.2 Gateway ...... 127 4.5.3 Inter-nodal...... 129 4.5.4 Terminal ...... 129 4.6 NODE: CRITICAL TERMINAL OPERATIONS AND DESIRABLE CHARACTERISATION ...... 131 4.6.1 Areas of Performance Measurement and Design of Terminal Interfaces ...... 131 4.6.1.1 Introduction ...... 131 4.6.1.2 Logistics Performance ...... 131 4.6.1.3 Storage and Handling...... 132 4.6.1.4 Stack and Sidings ...... 133 4.6.2 Model Identification Requirements ...... 134 4.6.2.1 Logistics Cost Structure and Value Stream Mapping ...... 134 4.6.2.2 Storage and Space Management ...... 135 4.6.2.3 The Bulk Queue...... 136 4.6.2.4 Multi-path Coupling ...... 137 4.6.2.5 Multi-commodity Flows ...... 137 4.7 GATEWAY: CRITICAL LINK AND NODE OPERATIONS AND DESIRABLE CHARACTERISATION ...... 138 4.7.1 Link Time Windows ...... 138 4.7.2 Prioritisation and Inter-Node Stack Availability ...... 138 4.7.3 Security Functional Requirements ...... 138 4.8 IMPEDANCE ATTRIBUTES AND SPECIFICATIONS FOR A URBAN INTERMODAL TERMINAL ...... 139 4.9 RELATING OBJECTIVES AND CASE STUDIES TESTING ...... 140 4.9.1 Objectives ...... 140 4.9.2 The Container Hinterland Case Study ...... 141 4.9.3 Integrated Waste Management: Storage and Streaming ...... 142 4.10 CHAPTER CONCLUSIONS ...... 143 5 MODEL FORMULATION OF INTERMODAL PRODUCTION SYSTEMS ...... 144
5.1 INTRODUCTION ...... 144 5.1.1 Chapter Guide ...... 144 5.1.2 Addressing the Functional Requirements ...... 145 5.1.3 Challenges and Limits of the Analogy ...... 149 5.2 TERMINAL OPERATIONS DESIGN MAPPED TO ELECTRICAL CIRCUIT PHYSICS ...... 149
5.2.1 Overview ...... 149 5.2.1.1 Circuit Stencil as Logistical Unit Process ...... 149 5.2.1.2 Simplifying Assumptions ...... 151 5.2.2 Variables and Parameters and Balancing Laws ...... 152 5. 2.2.1 Variables...... 152 5.2.2.2 Parameters ...... 154 5.2.2.3 Constitutional laws ...... 155 5.2.3 Load Based Design...... 160 5.2.4 Physical Realisability of the Analogue to Terminal Operations ...... 162 5.2.5 Impedances ...... 163 5.2.5.1 General Concept ...... 163 5.2.5.2 Value Accounting...... 164 5.2.5.3 Precision ...... 165 5.2.6 Synchronising Storage and Handling Operations for Temporal Accessibility ...... 166 5.2.7 Response to Business Rule Levers ...... 173 5.2.8 An Itinerary Mechanism ...... 174 5.3 FLOWS AS SIGNALS ...... 177 5.3.1 Introduction ...... 177 5.3.2 Excitation Response ...... 177 5.3.3 Stability Control ...... 179 5.4 TRANSHIPMENT CALCULUS ...... 181 5.4.1 Logistical Unit Process Mathematical Tools ...... 181 5.4.1.1 Characteristic Impedance in Time Domain using Stencil Operational Amplifiers ...... 181 5.4.1.2 Stencil Op Amp Canonical Forms ...... 183 5.4.1.3 Cost and Consumption: Complex Impedance and Polar Accounting System ...... 184 5.4.2 Multi-path Coupling ...... 189 5.4.2.1 Multiple Input Multiple Output Systems ...... 189 5.4.2.2 Filters as Itinerary Control ...... 191 5.4.3 Characterising Complex Bundling ...... 193 5.4.3.1 Defining Complex Bundling ...... 193 5.4.3.2 Use of Fourier Frequency Spectrum Analysis ...... 195 5.4.4 Rail Sidings and Flexible Capacity ...... 195 5.4.4.1 Carrier Pairs and Sidings Occupation ...... 195 5.4.5 Bulk Queue Control ...... 198 5.5 SPECIFICATIONS ...... 200 5.5.1 Flow Arrival and Dispatch Signals ...... 200 5.5.2 Design according to Cost-Quality Ratios under Tense Flux Conditions ...... 201 5.5.3 Analysis Using Bode Plots ...... 203 5.5.4 Figure of Merit ...... 205 5.6 CHAPTER CONCLUSIONS ...... 209 6 SATELLITE TERMINALS TO A SEAPORT ...... 210
6.1 INTRODUCTION ...... 210 6.1.1 Chapter Guide...... 210 6.1.2 Novel Aspects to Illustrate Modelling Prototype ...... 211 6.1.3 Role of Intermodal Terminal Satellite ...... 212 6.2 DEVELOPMENT OF CONTAINER SYSTEM IN SYDNEY AND MELBOURNE ...... 213 6.2.1 Container Flows and Forecasts ...... 213 6.2.2 Misalignment of Functional Relations ...... 214 6.2.3 Container System Capacity in Melbourne ...... 215 6.2.4 Landside Container System Cost Structures and Value Accounting Through Terminals ...... 218 6.3 THE ALTONA CLUSTER FOR STACK MANAGEMENT AT TERMINAL AND SEAPORT ...... 220 6.3.1 System and Terminal Descriptions ...... 220 6.3.2 Proposed Schedules and Dispatch Signals ...... 223
6.3.3 Terminal Transhipment Calculus: Design of Storage Utilisation ...... 223 6.3.4 Extension: Prioritisation and Stack Management at Seaport ...... 227 6.3.5 Stack Management Figures of Merit ...... 230 6.4 DANDENONG SIDINGS ACTIVATION AND BIMODAL OVERFLOW ...... 230 6.4.1 System and Terminal Description ...... 230 6.4.1.1 Original Concept Design ...... 230 6.4.1.2 Change Envisioned ...... 231 6.4.2 Proposed Schedules and Dispatch Signals ...... 233 6.4.3 Sidings Configurations and Impedance...... 233 6.4.4 Extension: Bimodal Overflow Control ...... 236 6.5 ENFIELD HUB FOR COMPLEX BUNDLING ...... 240 6.5.1 System and Terminal Description ...... 240 6.5.2 Terminal Transhipment Calculus ...... 241 6.5.3 Extension: Complex Bundling Phenomena ...... 242 6.5.4 Bundling Figures of Merit ...... 248 6.6 CHAPTER CONCLUSIONS ...... 248 7 WASTE TRANSPORT AND RESOURCE RECOVERY ...... 249
7.1 INTRODUCTION ...... 249 7.1.1 Chapter Objectives ...... 249 7.1.2 Delimiting the Study ...... 251 7.2 CONCEPT OF A SUSTAINABLE WASTE LOGISTICS SYSTEM ...... 251 7.2.1 Transportation Leverage for Waste Recovery ...... 251 7.3 SYSTEM FORMAT IN SYDNEY ...... 254 7.3.1 Waste fractions of interest ...... 254 7.3.2 The Role of Clyde...... 255 7.4 TRANSFER STATION FORMULATION FROM SYSTEM DYNAMICS ...... 257 7.4.1 Time Series Analysis of Source Flows ...... 257 7.4.2 Terminal Description ...... 257 7.4.3 Feedback loop of resources matching ...... 260 7.4.4 Storage-Handling Limits ...... 262 7.4.5 Throughput Capacity Results ...... 262 7.4.6 Modelling Gaps ...... 263 7.5 FORMULATING TERMINAL COMPLEX ACTIVITY ...... 265 7.5.1 Complex Impedance Representation of Terminal Activity ...... 265 7.5.2 Stack Interface ...... 267 7.5.3 Dumping Floor Inertia and Evolution of Work ...... 272 7.5.3 Itinerary Control and Streaming Wastes ...... 279 7.5.4 Sidings Capability ...... 280 7.5.5 Feedback Loops Between and Within LUPs ...... 281 7.6 INTER-TERMINAL SYSTEM FORMULATION ...... 284 7.6.1 Protocol Description ...... 284 7.6.2 Signal Pulse forms ...... 285 7.6.3 Specification and Method ...... 286 7.7 SYSTEM FORMATS - DESCRIPTION AND SPECIFICATION ...... 286 7.7.1 Base Case ...... 286 7.7.2 Clyde direct improved compaction ...... 287 7.7.3 Satellite to Clyde: Road-Rail ...... 288 7.7.4 Matraville Satellite to Clyde: Direct Rail to Rail ...... 288 7.7.5 Clyde to Goulburn via Campbelltown: Liner Train ...... 289 7.8 ANALYTICAL METHOD AND FUNCTIONAL REQUIREMENTS ...... 290 7.8.1 Analytical Method ...... 290
7.8.2 Figures of Merit Supporting Functional Requirements ...... 291 7.9 CHAPTER CONCLUSIONS ...... 292 8 CONCLUSIONS ...... 293
8.1 RE-STATEMENT OF PROBLEM ...... 293 8.1.1 Transitions Management Ambition ...... 293 8.1.2 System Context in Sydney and Melbourne ...... 294 8.1.3 Investigations for an Intermodal Production System ...... 295 8.1.4 System Formats Assessed ...... 298 8.2 CONTRIBUTIONS TO KNOWLEDGE ...... 299 8.2.1 New Phenomenological Approach ...... 299 8.2.2 Terminal Dynamics and Terminal Complex Activity ...... 300 8.2.3 Performance Metrics of System...... 301 8.2.4 Terminal Multipath Coupling ...... 302 8.2.5 Addressing Business Rules ...... 303 8.3 AREAS FOR FUTURE RESEARCH ...... 303 8.3.1 Building the Analogy and Parameter Identification ...... 303 8.3.2 Compensation Control ...... 304 8.3.3 Building Coordination Schemes ...... 305 8.3.3.1 Load Following Basis ...... 305 8.3.3.2 Augmenting Multi-Commodity Multimodal Flow Problem ...... 306 8.3.4 New Transportation Science Fields for Application ...... 306 8.4 FINAL CONCLUSIONS ...... 307 BIBLIOGRAPHY ...... 309 APPENDIX A- STRENGTHS AND WEAKNESSES OF ANALOGUE ...... 320
A.1 DIFFERENCES AND SIMILARITIES IN REPRESENTING FREIGHT TRANSPORTATION AS AN ELECTRICAL ANALOGUE...... 320 A.2 COMPARATIVE STRENGTHS...... 321 A.3 COMPARATIVE WEAKNESSES ...... 325 A.4 ANTICIPATING VERIFICATION PROCEDURES ...... 325 APPENDIX B- TERMINAL ACTIVITY MAPPED TO CUMULATIVE STORAGE CURVES . 326
B.1 TERMINAL ACTIVITY: HEADWAY AND STORAGE CURVES ...... 326 B.2 STORAGE CURVES AND MULTI-COMMODITY FLOWS ...... 327 APPENDIX C: THE LAPLACE TRANSFORM APPLICATION TO FREIGHT TERMINAL LOGISTICS ...... 330
C.1 INTRODUCTION ...... 330 C.2 FILTERS ...... 331 C.3 COMPENSATION CONTROL ...... 333 C.4 SCOPE ...... 335 C.5 SPECIFICATIONS ...... 336 APPENDIX D: DANDENONG CASE STUDY ...... 338
D.1 DANDENONG CONFIGURATIONS ...... 338 D.2: DANDENONG FREIGHTER SCHEDULE ...... 340 APPENDIX E: WASTE STUDY - BALANCE EQUATIONS FROM SYSTEM DYNAMICS ..... 341 APPENDIX F: THEORY SYNTHESIS AND EXTENSION TO NETWORK COORDINATION343
F.1 INTRODUCTION ...... 343 F.1.1 Chapter Guide ...... 343
F.1.2 Drivers and Means to Introduce Intermodal Forms in Urban Container Networks ...... 344 F.1.3 The Connection Between Transhipment Calculus and Coordination Theory ...... 345 F.2 DISTRIBUTED RESOURCE SYSTEMS ...... 348 F. 2.1 Distributed Energy Resource Systems ...... 348 F.2.1.1 Description...... 348 F.2.1.2 Conceptual Modelling Mechanisms...... 349 F.2.2 Distributed Function Systems in Container Freight Task ...... 351 F.2.2.1 Functional Relief of Seaport...... 351 F.2.2.2 Actor Governance Relations ...... 351 F.2.3 Analysis of Gateway Relations Leading to Measurement of Hinterland Absorptive Capability ...... 352 F.3 DESCRIPTION OF COORDINATION APPLICATIONS...... 353 F.3.1 Melbourne Intermodal Tri-Lobal Network ...... 353 F.3.2 Dynamic Routing ...... 357 F.3.3 Empty Container Movement and Repatriation ...... 359 F.4 COORDINATION MECHANISMS ...... 359 F.4.1 Introduction ...... 359 F.4.2 Control of Generation ...... 361 F.4.3 Economic Dispatch...... 362 F.4.4 Economic-Security Mechanisms of Commitment and Dispatch ...... 362 F.4.4.1 Introduction to Security Requirements...... 362 F.4.4.2 Operating Ranges ...... 363 F.4.4.3 Reserves...... 363 F.4.4.4 Distribution Limits ...... 365 F.4.4.5 Investment Optimisation ...... 365 F.4.4.6 Minimising Freight Exposure ...... 366 F.4.5 Hydro-thermal Coordination ...... 366 F.4.6 Network Multi-Area Exchange ...... 369 F.4.7 Figures of Merit ...... 371 F.5 CONCLUSIONS ...... 372
TABLE OF FIGURES
Figure 1: Research Design Perspective for Thesis on Urban Intermodal Freight Transport (Dotted boxes represent future work which is scoped in Appendix F) ...... 10 Figure 2: Scope and Role of Transhipment Calculus Theory (Dotted Elements of Coordination are Outside the Scope of this Thesis and are Outlined in a Paper in Appendix F) ...... 15 Figure 3: The Kaya Identity (Azar C. and Schneider, 2002) ...... 26 Figure 4: Container Freight Task Using Rail-Road Intermodal Forms (GAO, 2007 p.5) ...... 28 Figure 5: Conceptual Model of Designing and Managing Intermodal Retrofit (Outhred, 2002) 29 Figure 6: Process of Modeler Intervention of Parameter Relationships Midstream to Track Feasible Transitions (De la Barra, 1989) ...... 34 Figure 7: Urban Freight Hydrocarbon Direct Energy Requirement with Different Strategies - ASFF Output (Beavis et al., 2009) ...... 36 Figure 8: Goods Demand and Freight Traffic Interaction (Priemus and Konings., 2001) ...... 38 Figure 9: The Complementary Sub-systems of Logistics and Transportation (Sjöstedt, 1994) .. 39 Figure 10: Metropolitan Road and Rail Links Serving Port Botany, Sydney (Triangles are intermodal terminals; Circles are import and export container activity centres) (SPC, 2000) .... 46 Figure 11: Approaches to Transport and Logistics System Complexity (Waidringer, 2001) ..... 53 Figure 12: Integration of System Elements Relating Infrastructure Capability with Load Operations ...... 61 Figure 13: Relating Engineering Impedance to Performance Indicators ...... 61 Figure 14: Hierarchy Indicator Diagram for Intermodal Balance Capability...... 65 Figure 15: Freight Conceptual Model (Rimmer and Black, 1981, Fig.1, p.16)...... 71 Figure 16: Freight Network Taxonomy (Beavis et al., 2005) ...... 82 Figure 17: Alternate Rail Operating Forms Interfacing Terminals (Woxenius et al., 2004)...... 98 Figure 18: Information Needs for Intermodal Design (Beavis, Black et al., Fig. 4, p. 2986, 2005) ...... 100 Figure 19: Terminal Time-Volume Curves with Varying Arrival Headways (Morlok, Figure 7.8, p.263, 1978) ...... 102 Figure 20: Hinterland Absorptive Capability and its impact on Seaport Capacity: Gateway Node Access and Storage Capacity ...... 118 Figure 21: Demonstrating Accessibility Improvements and the Feasibility of Intermodal Terminals Requires a Gateway Analysis to Link Networks (Styhre, 2005, Fig. 7) ...... 128 Figure 22: A Transit Terminal for Inter-Intra-City Transhipment (Morlok, Figure 7-4, p.254, 1978) ...... 130 Figure 23: Marginal Cost with Throughput (after Ballis and Golias, 2004) ...... 135 Figure 24 Stencil Derivation for Transfer Impedance for Operations at a Terminal ...... 157 Figure 25: Stencil for Value Density and Consignment Flux Change at Waste Transfer Station ...... 158 Figure 26: Impedance Z with Synchronisation Curve ...... 169 Figure 27: Storage lag (Phase Angle) with Transhipment Intensity (Frequency) (Generated from RLC Stencil in PSPICETM model) ...... 171 Figure 28: Development of Complex Impedance Due to a Synchronisation Response (Frequency) which does not meet the desired synchronisation rating of the unit process (F- resonance) ...... 172 Figure 29: Variation of Consignment Flux as a Function of Synchronisation (Frequency Spectrum) ...... 172
Figure 30: Terminal Impedance Response as a Resolution of Work from a Pulse Input ...... 178 Figure 31: Non- Inverting Operational Amplifier Stencil Sketch ...... 182 Figure 32: Terminal Space Management Analysis through Complex Impedance Phasor Diagram ...... 186 Figure 33: Attributes of Impedance and Value in a Logistics Unit Process for Indirect Transhipment ...... 188 Figure 34: Attributes of Impedance and Value in a Logistics Unit Process for Direct Transhipment ...... 188 Figure 35 A Multipath Coupling Diagram for Direct and Indirect Export Transhipment...... 189 Figure 36: Coupling Matrix (C) Relating Input Flows (P) from Different Hub Connections and carriers (α..ξ) from different paths (i…n) to Outputs to Other Hubs (L) ( after Geidl and Andersson, 2005a) ...... 190 Figure 37: The Relationship between Time and Synchronisation Domains: Carrier Pairs and Rail Sidings Occupation and Train Dispatch ...... 196 Figure 38: Pulse Train Signal in Time Domain (Bobrow, Figure 11.3. p.506, 1987) ...... 197 Figure 39: Determining Design Parameters from Stability Specification ...... 204 Figure 40: Heavy Truck Flows at Critical Junctions and Port Gate Bottlenecks in weekday morning peak around Port of Melbourne (April 2002) (DoT, 2007) ...... 217 Figure 41: Typical Container Supply Chain Costs for Sydney (SFCNSW, 2004b) ...... 219 Figure 42: Altona Intermodal Cluster of Three Intermodal Terminals West of Port of Melbourne (DoT, 2009) ...... 222 Figure 43: DC Mesh Circuit Analogue for Logistical Unit Process of Container Storage Buffer at Intermodal Terminal, inclusive Capacitive Load C1 ...... 224 Figure 44: Phasor Diagram Depicting Impact of High Throughput Direct Transhipment Intensity -Inductive Load (No Capacitive Load) (after Nilsson and Riedel (2005)) ...... 225 Figure 45: Phasor Diagram Depicting Impact of More Indirect Transhipment Intensity- Combined Inductive and Capacitive Load Operations (after Nilsson and Riedel, (2005)) ...... 226 Figure 46: Stencil Bandreject Filter for Sidings Configuration Constraint ...... 234 Figure 47: Sidings Range of Inactivation (PSPICE output) ...... 235 Figure 48: Intermodal Terminal Design Schematic (Railhead on west side of terminal) ...... 237 Figure 49: Bimodal Overflow Control Filter Logic ...... 239 Figure 50: Proposed Flow Balances for Enfield Intermodal Terminal (SKM, 2005) ...... 240 Figure 51: Enfield Intermodal Terminal Process Flowchart...... 242 Figure 52: Complex Bundling Filter Logic ...... 247 Figure 53: Train Path from Sydney to Woodlawn Bioreactor ...... 255 Figure 54: Distribution Analysis of Truck Arrivals at Clyde (Porta and Marchetti, 2007) ...... 257 Figure 55: Aerial View of Clyde Terminal (Collex, 2006) ...... 258 Figure 56: Process Flowchart of Waste Intermodal Terminal at Clyde ...... 260 Figure 57: Modelling Map of Logistical Feedback Loops at Intermodal Waste Terminal (Porta and Marchetti, 2007) ...... 261 Figure 58: Stencil for Waste Container Stack Management ...... 269 Figure 59: Bode Plots for Waste Stack Management Scenarios ...... 271 Figure 60: Dumping Floor Movements at Waste Transhipment Facility and growth of the waste zone...... 273 Figure 61: Transformer Circuit Stencil for Dumping -Compaction Operation ...... 276 Figure 62: Compaction Handling Bode Value Magnitude and Storage Lag (based on PSPICE output) ...... 277
Figure 63: Physical and Logistical Flowchart of Waste Transhipment Terminal ...... 283 Figure 64: Clyde Base Case- Clyde Compaction to 27t and then to 18t ...... 287 Figure 65: Road Rail Direct Improved Compaction- Clyde Process Streaming ...... 287 Figure 66: Satellite Hub- Clyde Cross Docking ...... 288 Figure 67: Rail Trunk and Feed- Clyde Pulse Correspondence ...... 288 Figure 68: Liner Operating Form- Clyde Streaming, Compaction and Ordering ...... 290 Figure 69: Research Perspective for Urban Intermodal Freight Transport Thesis ...... 297 Figure 70: Performance Indicator Hierarchy of Harmonisation Attributes for Freight Intermodal Balance...... 301 Figure 71: Cumulative Curves for Container Arrivals A(t), Departures D(t), Static R(t) and Dynamic R’(t) Assignment from Stack to Vessel (after Taleb- Ibrahimi, et al., Fig. 3 p. 18,1993) ...... 327 Figure 72: High Quality Bandpass Filter ...... 333 Figure 73: Design Modification of Bode Plot for Terminal Handling- Flux Goals (after Bobrow, Fig 10.16 p.456) ...... 337 Figure 74: Dandenong Intermodal terminal Design Single Sidings Access (DoT, 2008e) ...... 338 Figure 75: Dandenong Intermodal Terminal Design Central Sidings Access (DoT, 2008e) .... 339 Figure 76: Dandenong Freighter Form Distance Time Chart (2008e) ...... 340 Figure 77: Scope and Role of Transhipment Calculus...... 347 Figure 78: Time Mechanism Components of System Coordination (Beavis et al., 2007) ...... 350 Figure 79: Melbourne Tri-lobal Intermodal Network ...... 354 Figure 80: Network of an Area with Key Spanning Arcs ...... 356 Figure 81: All Road Haulage to the Seaport Precinct for Stuffing/UnStuffing (Elliptical arrows represent empty container returns) ...... 357 Figure 82: Intermodal Terminals in the Stuffing/Unstuffing Transhipment Task allowing rail leg (routes to the intermodal terminals are dynamic) ...... 357 Figure 83: Process Control of Generation of Response to Loads (Wood and Wollenberg, 1996, Fig, 9.25, p.354) ...... 361
TABLE OF TABLES
Table 1: Value Ledger of Road Only Haulage vs. Road-Rail Intermodal Service (Dandenong to Port of Melbourne Example) ...... 33 Table 2: Relationship of Logistics Drivers to Operational Freight Indicators (SULOGTRA, 2002) ...... 39 Table 3: Impact on Transportation Cycle ...... 59 Table 4: Impact on Production Cycle ...... 59 Table 5: Criteria and Attributes to Address in Intermodal Sketch Planning Functional Requirements ...... 146 Table 6: Mapping Logistical Unit Process Attributes from Functional Requirements to Mechanisms and Case Study Illustrations ...... 147 Table 7: Stencil Components Mapped to Freight Terminal Parameters...... 154 Table 8: Value Density Calculation through an Intermodal Waste Transfer Station ...... 159 Table 9: Transhipment Intensity and Impedance Condition ...... 167 Table 10: Canonical Op Amp Stencils for Representing Logistical Unit Processes ...... 184 Table 11: Mapping Freight Terminal Performance to Circuit Model Specifications and Performance Figures of Merit ...... 206 Table 12: Seaport-Hinterland Attributes to Investigate ...... 212 Table 13: Container Freight Forecast by Activity Centre and Mode Opportunity in Melbourne (in 000’s TEU/yr and percentage of total flows) (DoT, 2008b) ...... 214 Table 14: Proposed Principal market Areas of Enfield Intermodal (SPC, 2005) ...... 243 Table 15: Composition of Complex Bundling Problem in an Intermodal Terminal and Attributes Measured ...... 245 Table 16: Projections for Waste for Disposal Quantities Intermediated Through Clyde in 2015 (Uncapacitated) ...... 256 Table 17: Parameters for Base Case Used in System Dynamics Simulation...... 266 Table 18: Operating Times of Mobile Resources (Porta and Marchetti, 2007) ...... 267 Table 19: Scenarios for Stack Management ...... 268 Table 20: Parameters and Results for Waste Stack Management Scenarios ...... 270 Table 21: Variable Changes around Waste Dumping Floor and Compaction LUP...... 275 Table 22: Dumping Floor- Compaction Logistical Unit Process Results ...... 278 Table 23: Itinerary Control in Waste Transhipment Terminal Activities ...... 280 Table 24: System Formats to be Studied According to Transhipment Protocol ...... 284 Table 25: Payload and Synchronisation Band Change with Waste Transport System Format . 285 Table 26: System Format Relationships, Specifications and Figures of Merit ...... 291 Table 27: Transhipment Calculus Attributes Described Through System Format Case Studies ...... 298 Table 28: Positioning the Electrical Stencil Tool in Freight Terminal Analysis for Sketch Planning ...... 323
Chapter 1: Introduction
1 Introduction
1.1 Presentation of Problem
In the Australian cities of Sydney and Melbourne, over eighty per cent of container traffic has an origin or destination within the greater urban area (within a 40 km radius) to or from container seaports. The container flows are high value and generate significant pre- and post haulage road traffic. The logistical structure underpinning container freight flows is highly dissipative: a high proportion of wasteful trips of low load utilisation, high road use intensity, and poor connectivity leading to capacity constraints on access to gateway nodes, like seaports. There is significant residential disamenity, called Freight Exposure, and this conflicts with access on passenger carriageways both on road and rail modes. Both Port Botany in Sydney and the Port of Melbourne are undertaking significant seaside development to augment capacity with the expected increase and size of container vessels projected. Planning productivity improvements by overcoming landside logistical constraints still lag behind in theory and implementation.
The freight task, in both intensity (tonne/kilometres) and absolute terms (tonne-kilometres), has been increasing at a rate greater than the physical flow of goods in Australia, Europe and North America. Freight activity control, as one measure of a nation’s de-materialisation, and thus a vector of sustainability, becomes difficult to achieve. This growth has been ascribed to various logistics drivers, such as Time Compression Principles (Just-In-Time production), and cheap transport, fostering the geographic extension of the supply chain (SULOGTRA, 2002) .The demand matching response has been either an injection of new physical infrastructure, or otherwise planning inertia. Both responses have allowed urban transportation networks to become increasingly inequitable, congested and polluting.
Additionally, the planning jurisdictions of both Melbourne and Sydney have suffered from haphazard intermodal system planning practice and functional deployment1. Penetrating questions are only recently being asked about intermodal feasibility and the tools in design and implementation for an informed decision and performance measurement are found to be
1 This is evidenced in Sydney by the inability to achieve container freight flows on rail above 20% partly due to inefficient, poorly retrofitted terminals; in Melbourne, by there being no current metropolitan share of container rail traffic with a severe State Government policy disconnect, which has stated for years the ambition to achieve 30% of container freight flows on rail. Functional deployment considers the functions an intermodal terminal should undertake and thus this thesis is not concerned with the purely strategic operational problem. Page | 1
Chapter 1: Introduction
wanting. This thesis makes concrete steps toward a transportation science of urban freight intermodalism. It does this directly by proposing novel analytical tools to be used in intermodal sketch planning using the analogue of electrical circuit theory. Indirectly, it is identified that the analytical approach can be used to build modelling mechanisms of network coordination.
The aim of this thesis is to extend the capability of sketch planning in the design of urban freight intermodal terminal functions to improve understanding of intermodal feasibility and thus foster their adoption2. Urban rail-road freight intermodalism has the potential to achieve three outcomes: harness the use of less carbon-emitting transport modes; foster better space utilisation of unit loads; and consolidate the existing network by opening gateway access opportunities. This third aspect minimises physical growth of the network through techniques such as peak shaving and making intermodal connections with existing and new infrastructure. Such consolidation activity can lead to a lower road freight intensity (kilometres/tonne) and maximize system capability. Integrated transport through intermodal production systems has the potential to provide major capability improvements to the system through coordination mechanisms which increase capacity and flexibility. In order to deliver these outcomes, a heuristic approach must be developed which provides calculation mechanisms to assess intermodalism in differing network arrangements. Currently, implementing the right system formats (infrastructure configuration, governance arrangements, flow patterns, and logistics relationships) to achieve a physically realisable3 freight system remains uncertain, undocumented and uncoordinated.
Freight intermodalism involves the logistical arrangements and physical infrastructures at intermediate nodes which utilise at least two modes of transport for the transhipment of freight. Combined transport is another term for intermodalism and can also extend to intra-modalism or bimodalism, which is also covered in this thesis. Intermodalism may also refer to urban passenger transit but this meaning is outside the scope of this thesis. Two salient features of intermodalism for freight are the synchronisation of vehicle and train resources and the storage arrangements at the terminal. Unlike multimodalism, modes are considered as complementary and not competitive. This thesis focuses on road-rail intermodalism for freight interacting with the urban space. The critical mechanism to demonstrate is the cause and effect relationship between space management with matching train dispatch and arrival schedules.
2 This thesis recognises and reviews the designation of the classic three tiers of freight planning: strategic dealing with generation type land use planning and location issues; tactical dealing with medium term issues such as terminal design; and operational issues dealing with routing and scheduling of truck and consignment flows. These tiers are found to be inadequate in delivering integrated–cross network and cross infrastructure ownership solutions. To develop a transportation science of freight intermodal systems, the essences of all three tiers need to be combined. Intermodalism in this thesis also pertains to bimodal and transmodal arrangements. 3 Physical realisability infers the transportation system meets precision and controllability specifications. Page | 2
Chapter 1: Introduction
As well as facilitating an efficient interface between road and rail modes, surface intermodalism has the potential to provide integration of freight activities over a network. Integrated transport may be considered a significant sustainability objective of urban societies. Integrated transport is a means of reducing the resource intensity of mobility demands for accessibility to places necessary for the translocation of goods and people. A demand steering response would consider demand management initiatives to control the growth of the freight task. An intriguing proposition is whether logistical drivers and the growth congestion they generate (including additional trips) can be ameliorated by a physical and logical retrofit of existing network infrastructure. Such demand management objectives would redefine what is meant by effective transport infrastructure.
This thesis outlines a modelling framework for Intermodal balance4. Intermodal balance is a term to describe the technical envelope requirements of integrated transport. It is seen as a method in facilitating integrated transportation (Vuchic, 1999). The objective of integrated transport is to increase the capability of the system with the minimum physical augmentation of the road network. This capability allows the handling of varying fluxes as well as providing resilience; the quicker recovery of reliability under conditions of disturbance. Intermodal balance as a manifestation of integrated transport has not been mathematically formulated to date for either passenger or freight distribution networks. In order to formulate a mathematical basis for this technical envelope it is necessary to explore inter-temporal storage and service generation opportunities5 (MacGill, 2006). For this thesis, the intermodal balance concept is explored through various measures to characterise Hinterland Absorptive Capability. This term links landside storage and accessibility opportunities and constraints to seaport operations in order to define the productivity of container seaports. The introduction of intermodal operations in existing networks may provide significant relief to the use of road-based transport. However, it leads to more complexity in consignment transformation and transhipment. This complexity must be managed. Decision support is needed in understanding the design trade-offs between complexity in intermediation and controllability (Rodrigue, 1999, Waidringer, 2001, Sjöstedt, 1994). Precision relations become critical to define in demonstrating a physically realisable system of intermodal operations which improve the space utilisation of the system.
4 Intermodal balance is also referred to in this thesis as harmonisation. The novel modelling endeavour presented in this thesis addresses the attributes of harmonisation in order to guide the container freight task toward a new vector of sustainability. Harmonisation is discussed in Chapter 2. Intermodal Balance is used rather than harmonisation as the term harmonisation has been associated in the policy and planning literature with addressing the symptoms of landside container freight logistics problems rather than treating the underlying causes. 5 Also known as Generative Scaling. Page | 3
Chapter 1: Introduction
Terminal functions which deliver desirable precision and controllability outcomes may be related to network service forms such as hub and spoke formats, trunk and feeder lines or corridors with intermediate stops. These network forms can also be specific to modes, for instance another term is train operating forms. There is however little analytical means to investigate the desirable service relationships between such service flow forms and terminal functions under different coordination or trading regimes. This research design problem has been articulated by Woxenius (1998); Kreutzberger (2008); and Caris et al., (2008). Tactical design requirements extend to attributing functions to terminals acting as satellite dry ports and assessing how fine-meshed systems (a dense network of interdependent nodes and links) can accommodate the uptake of small freight consignments6. Freight transhipment operations have costs and also add value through such activities as consolidation, and storage which can provide higher productivity flows. These activities require a model abstraction of impedance which is richer than prevailing freight network modelling. Devising a modelling framework for such decision support would assist planners in positioning intermodal production systems, particularly urban intermodal systems, into existing all-road networks. The essence of the underpinning impedance relationship needs to consider terminal activity and its interface with service flows. This is the core methodological and technical problem addressed in this thesis.
The implementation of intermodal production systems, particularly in an urban context, is constrained by a number of engineering-economic factors. An effective modelling framework needs to incorporate both these aspects to ascertain design feasibility. Designs to be evaluated involve a different hub network with particular traffic and consignment transfer and storage relationships at nodes. Such novel network options face obstacles due to the existing infrastructure, ownership of resources, land availability, interoperability of interface exchanges, and the preference for large terminals based on their perceived superior cost performance based on economies of scale. Intermodal production systems may provide significant value in the improvement of unit space utilisation and minimising dwell time at critical node gateways, such as at the container seaport. However, the conventional wisdom in achieving a cost-value jump for the justification of intermodal operations generally necessitates significant line haul distances. Urban intermodal networks are rare and when they are proposed as in Sydney (FIAB, 2005), network arrangements are of the more basic direct connect type with large originating terminal nodes. There are also ownership issues in influencing the openness of the system and the types of functions attributed to these terminals. Ownership arrangements may detract from improvements in system coordination and lead to system sub-optimisation. Additionally, these large terminals face considerable opposition given scarce availability of urban land. Retrofitting the network as an intermodal production system requires careful design to ensure increased
6 This is also known as complex bundling networks (Kreutzenberger, 2008). Page | 4
Chapter 1: Introduction
system capability. Consequently, defining and testing functional relationships within and between terminals is just as significant as locational attributes.
1.2 Level of Technical Knowledge and Gaps
The methods developed to test novel intermodal network designs in metropolitan areas investigated in this thesis are based on gaps identified in prevailing freight transport models. There are two research contexts to draw together: how terminal resource productivity is measured and how to coordinate transhipment activity. The response to identified gaps in these research contexts is organised into two parts respectively: 1) the development of an activity- based approach concerning freight terminals; and 2) the development of a transitions approach for the deployment of intermodal infrastructure as a system retrofit. This thesis focuses the first response with mechanism that may yield useful analytical solutions. Research scoping is provided for the second response where the analytical mechanisms may be developed further for network applications.
There are eight gaps in conventional transportation science which need to be bridged before an authentic freight metropolitan intermodal science may be discerned:
1. Terminal node and network productivity are only superficially related in transportation science7. Critical coupling within terminals needs to be described and it must be identified how these couplings affect precision relations in networks; 2. An activity-based approach is required to accentuate the role of terminal nodes in a system. That is, terminal multi-path coupling needs to transform the Multi-Modal Multi-Commodity flow problem; 3. A systems control-unit process approach is required to characterise terminal nodes as interdependent entities which depend on and enable upstream, downstream and sidestream fluxes. This is a pre-condition for network coordination approaches; 4. Physical network design and service network design are not well linked. A more seamless modelling relationship would aid in control of the injection of lumpy supply that is often an inevitable outcome of conventional network predictive models; 5. Capability rather than capacity should be considered a measure of effective supply provision. Capability considers the flexible capacity of infrastructure under conditions
7 For instance, the process of defining truck types and thus the freight intensity is only introduced after third, allocation step in the classic 4-step model. Terminals as hub consolidation nodes is not easily introduced. Page | 5
Chapter 1: Introduction
of varying flow whereas capacity is predominantly defined as a constraint phenomenon. A means of measuring hinterland absorptive capability is required to relieve intractable hinterland- seaport conflicts; 6. Conventional characterisation of the itinerary of freight consignments does not consider that the itinerary ought to be a path of commitment (availability) and dispatch (control). This would anticipate a coordinating rather than allocative system mechanism; 7. The time composition of freight has neither been considered in detail nor in an analytical fashion. A transhipment calculus would provide a helpful tool for sketch planning and support novel designs in urban intermodalism; and 8. An activity approach requires new validation procedures in order for it to be considered a transportation science.
The strategic-location and tactical-scheduling aspects of decision support have generally been addressed by the transhipment problem of Operations Research. The physical design and service design aspects of freight planning and modelling are clearly delineated in freight infrastructure and logistical structure planning (Crainic and Laporte, 1997). However, there are problems with keeping both analyses separate when there are many necessary connections that would improve the decision support toward the symbiotic deployment of distribution infrastructure (Beavis et al., 2005). The injection of physical supply of infrastructure is a crucial problem in urban freight networks that have to compete for road and rail space with passenger transport. In a recent container freight capability report for Sydney, it was noted that there was a need to investigate alternatives to the physical augmentation of infrastructure as a supply response to growing the freight task (SFCNSW, 2005). This is a general predicament for delivering sustainable, effective infrastructure in Australia (AUSCID, 2002). Wilson (1979) has observed that the current network based modelling processes, with predictive intent, identify the need to augment physical nodes and links. This brings lumpiness to the supply problem. Such lumpiness can be undesirable due to large implementation and maintenance costs. Analytical approaches that test retrofit opportunities in network infrastructure for improved integration may offer incremental changes. These changes would involve both physical and logical re- configurations that increase the capacity of the system with less physical infrastructure growth.
In a node-centric modelling framework, where the decision support objective is demand management, a new means of framework verification is required along with the new identification structure. As this method is explanatory of terminal interactions, rather than predictive of flow growth on the network, there is less focus on network calibration. The
Page | 6
Chapter 1: Introduction
validation process in transport network planning conventionally involves structuring the land- use–transport system, empirical calibration and ensuring a goodness of fit (Black, 1981). In freight transport science dealing with intermodalism, we are interested in specific itinerary corridors and their gateway connections to other resources and not necessarily whole networks. The verification procedures for a physically realisable system now lie with the concepts of causality and stability. Model correctness must be satisfied according to flow continuity equations as well as controllability concerns. This underpins the need for a control systems approach. The lack of a control systems approach to analysing effectiveness of freight distribution infrastructure is a fundamental research gap in the city logistics literature. Transportation network theory does not easily allow for endogenous changes in the state of supply to be generated with alternate flow patterns. Consequently this theory is largely silent on network capability.
The complexity facing planners in securing a role for freight intermodal operations in the urban environment can be summarised as follows:
In order to reduce the impact of the urban container freight task, how can the freight system deliver the payload necessary to support freight container train shuttle services and thus justify metropolitan intermodal operations?
The means to improve the cost-quality ratio for critical short haul rail schedules is the issue this thesis is dedicated to addressing at the level of sketch planning. It is at the terminal level where consolidation activity is generated and attracted so the terminal becomes the critical focus of the system design. In order to achieve a sustained payload, both a terminal’s value proposition and physical capabilities need to be assessed. The essences of Open Access8 Regime requirements need to be captured and modeled to observe terminal feasibility. The scope of this regime, which characterises the system format of the terminal operation, extends to direct and enabling infrastructure, supporting stable operational interfaces such as train services, and the practice of other system and terminal based business rules that will engender coordination of resources and thus fewer dissipative outcomes.
Planners then have interests in applying what if scenario analysis of alternate terminal-network forms at the tactical level of scheduling to ascertain the financial and operational feasibility of intermodal networks as well as to ascertain the efficacy of so-called harmonisation objectives. As government support extends to both infrastructure investment (direct infrastructure and
8 Open Access Regimes are those where applicants to accessing the system are given fair conditions for access by the owners of the infrastructure and the operators of the means of transportation (ARTC 2002. Access Undertaking. In: AUSTRALIAN RAIL TRACK CORPORATION LTD (ARTC) FOR AUSTRALIAN COMPETITION AND CONSUMER COMMISSION (ACCC) (ed.). As an extension to this definition it is also seen as the means whereby the operational roles of actors harmonise with the prevailing logistical structure (SFC NSW, 2007). Page | 7
Chapter 1: Introduction
enabling infrastructure9) and a congenial regulatory framework, the application of different business rules are a quintessential measure in assessing the feasibility of alternate intermodal forms. A tool is required to allow exploration of the interface among flexibility, capacity and controllability objectives. It is this interface that designates the capability of the intermodal production system.
In order to obtain desirable guidance, a tool needs to provide helpful Figures of Merit on intermodal operations that can validate both the value proposition of terminal operations and their physical capabilities. These Figures of Merit would present key trade-offs in cost-value and capability-controllability as the new verification tools in intermodal transportation science.
1.3 Objectives and the Research Perspective
This thesis defines an activity-based approach for the urban container freight task which is not tied to the prevailing network in model formulation nor constrained by reductive network model identification relationships. The approach allows us to evaluate novel hub forms which support the role of rail in addressing growing avalanche pulse flows from a gateway node like a seaport. Thus, the objective of this thesis is to develop a novel modelling framework towards the design of intermodal balance. Specifically, the objective is to develop a conceptual and theoretical framework to support the modelling of analytical solutions for aspects of urban freight intermodal planning. This is a significant step towards developing a method to assess the multi- modal, interface, and infrastructure process control of the seaport-hinterland, which is identified as future research. Both the analytical approach and the aspiration it feeds for system coordination intend to support decision making in establishing physically feasible intermodal production systems in urban environments.
This main objective is complemented by four sub-objectives:
1. Propose a rigorous analytical procedure for the representation of urban intermodal terminal operations using the principles of activity based approaches. This would allow for assessment of terminal flexible capacity; 2. Provide an accounting system that takes into account impedance of logistical unit processes;
9 Such as terminal facilities and passing loops for access to mainline rail respectively. Page | 8
Chapter 1: Introduction
3. Develop, and demonstrate, a hierarchy of indicators for the planning of effective freight distribution infrastructure. This would link to design Figures of Merit to assist in trade- off analysis in sketch planning; and 4. Propose a method for analysis of itinerary routing through terminals of multimodal carrier, multi-commodity flows which is responsive to defined business rules.
The outcomes of this research are to gain a fair analytical solution with minimum data requirements in order to gauge the benefits and trade-offs of intermodal system retrofits for service alteration. This is the sketch planning nature of the methods developed in this thesis which are illustrated by case studies in waste transportation and seaport-landside relations in international container transportation. This will lead to an analytical approach with an explicit ecological restructuring aim that consolidates freight flow activity to a) direct the desirable evolution of the container hinterland supporting a gateway seaport and b) introduce rail based- hubs in urban waste management, for instance.
The investment logic map (ILM) is a tool used by the Victorian Government to discern the value of expenditures. It ties together, drivers, objectives, benefits, changing activities and enabling assets (investment inputs) to highlight that the investment has clear linkages to the problem definition (drivers) and will deliver desirable outcomes (benefits). A significant discipline is that the drivers and objectives do not specify the solution. They stand alone and the proposed solution method can be fully justified in meeting the component drivers and objectives. ILMs can cover all magnitudes of investments and can also be nested in a hierarchy of maps if the investment has discrete but inter-linking phases.
This thesis is an investment in a new method to deliver the task of urban freight coordination. Therefore an ILM is presented for this thesis to present the research perspective. It is a tool to demonstrate the linkages among drivers of knowledge gaps/problem definition, thesis objectives, the benefits (contributions to knowledge) and the tools and activities applied to deliver these contributions. The ILM of Urban Freight Intermodal Transport Systems also details Key Performance Indicators (KPIs) and Figures of Merit (FoM) or design performance indicators, which demonstrate the areas of contribution.
The ILM (Figure 1) allows the reader to orientate themselves throughout the thesis to confirm what contributions are being made and how they link to addressing the problem definition and critical knowledge gaps. The ILM leads to the method diagram depicting the scope and role of the transhipment calculus theory (Figure 2).
Page | 9
Chapter 1: Introduction
Figure 1: Research Design Perspective for Thesis on Urban Intermodal Freight Transport (Dotted boxes represent future work which is scoped in Appendix F)
Page | 10
Chapter 1: Introduction
1.4 Method
A comprehensive literature review of freight models was undertaken to understand approaches in modelling the placement and functioning of intermodal production systems. From this, specific research gaps were identified as well as critical processes that needed to be captured in any model formulation of terminals. In the development of this thesis a number of science theories were reviewed. The graphical network theory and the transhipment problem of classical transportation science were investigated for its gaps in yielding critical insights regarding desirable hub typologies in satellite intermodal-seaport systems.
A number of disciplines were reviewed for an appropriate modelling structure: queuing theory; system dynamics; graphical network theory; computer control systems; electrical circuit theory; and power distribution theory. The last three disciplines offered a means to assess the flexible capacity of a selected freight distribution network. Power distribution theory was selected as it is a node-centric approach and it is amenable for making cross network connections. This is ideal for incorporating the nature of freight terminals gateways that cross uncontainerised and containerised flow networks, for instance.
In order to integrate terminal operations with network forms, this thesis takes a node- centric approach. A node-centric approach to transport design requires a sparse representation of terminal activities to coordinate freight flows according to minimum costs and maximum space unit utilisation. The nodes need to become the generating unit of an activity-based approach, which has emerged in transport science as a significant theoretical concept in the demand management of network activity (McNally, 2000). A richer description of the time composition of freight can then be made which can be calculated with typical line impedances. Nodes as activity units then are recognised as having both constraints and compensation mechanisms. Node compensation, using feedback mechanisms, is currently not prevalent in allocation solutions provided by graphical network theory.
Freight transport may be seen as an industrial eco-system: it provides a particular set of services and utilises natural resources. This corresponds with the consideration of an intermodal system as a production system (Woxenius, 1998). As such, analysis of intermodal facilities is amenable to a unit process approach. The means of organisation of the industrial eco-system can be characterised by a specific system format involving land –use generators and attractors in space and time, technology, physical infrastructure, information and commercial relations. The intermodal production system is made up of these aspects and is the unit of investigation in this
Page | 11
Chapter 1: Introduction
thesis, rather than the intermodal terminal itself. Strategic location theory generally considers nodes that operate autonomously and do not interact.
The modelling framework is thus based on electrical circuit theory and computer controlled systems (Ånström, 1997) rather than combinatorial graph theory with its resolution through whole of network equal travel time or cost equilibrium. Computer controlled systems provide a means for decentralised control of freight terminal activity to maximise system benefits. This is achieved by firstly developing a transhipment calculus which calculates the time composition of freight. This provides a basis to calculate the characteristic impedance of Intermodal Production Systems. Secondly, this sparse representation of terminal activity is the foundation for inter- temporal coordination of terminal resources at time intervals from hours to include long term investment periods. Thus, strategic and tactical decision support issues are combined. This transitions approach marks this modelling framework as novel.
To model movement of consignments and information around a network as well as transformations within nodes, a novel representation of steady state dynamics is required. This thesis proposes the use of an electrical circuit analog to sparsely represent terminal activity. This method allows for both causality and stability to be verified. It associates the driving forces of consignment arrivals with transformations at terminals using transient storage states. Compensation mechanisms are introduced with control circuitry using Operational Amplifier circuits.
These analogs provide the stencils to design distribution terminals known as characteristic impedance. The stencils enabled freight logistics phenomena to be illustrated analytically with analogy to electrical circuit theory. Mechanisms were devised to address the following phenomena which were identified in the literature as necessary in order to characterize and manage urban intermodal operations under tense flux conditions:
1. Stack and handling interaction 2. Sidings capability 3. Bulk queue control 4. Bimodal overflow control 5. Order of consignments in a shipment 6. Pulse correspondence between modes
The characteristic impedance of each terminal logistical unit process acts to guide the itinerary of flows to dispatch. The mechanisms in guiding these flows are service generation and inter-
Page | 12
Chapter 1: Introduction
temporal storage. The itinerary mechanism allows multi-attribute information regarding the flow to be included in the terminal operations. This is known as multipath coupling.
Multipath coupling, based on terminal characteristic impedance, acts as a measure of terminal capability. This can be used later to generate an equilibrating mechanism for network coordination. Coordination schemes using the terminal impedance functions developed in this thesis are scoped in Appendix F and are flagged for future research.
A number of models were used to explore the research perspective of this thesis. System dynamics theory was reviewed and applied in a macro setting using the Australian Stocks and Flows Framework (ASFF) of the CSIRO in order to demonstrate the concept of system formats in transport science and transitions management. System dynamics using PowersimTM was also used at a terminal level in the development of an intermodal terminal operation at Clyde for the transhipment of container waste and to develop key terminal parameter relationships used to validate the circuit analog. PSPICE, an electrical circuit simulator, was used to develop phenomena and useful relationships through electrical circuit stencils.
For the waste freight case study, data was collected from two operating terminals in Sydney to provide verification of a simulation model. The results of the simulation model were used to enumerate parameters of the transhipment calculus. For the hinterland container case study, data and system characterisation were taken from a number of capacity audit reports on the container networks in Sydney and Melbourne. The Enfield intermodal terminal was modeled from data available from the Enfield EIS (SKM, 2005). Descriptions and data for the proposed Altona intermodal hub and Dandenong terminal were gathered during work undertaken at the Victorian Department of Transport. However, the case studies themselves are purely hypothetical. Constraints on the available urban freight rail network were discerned from engineering drawings, current timetables and discussions with professionals.
The scope and role of the transhipment calculus theory developed is presented in Figure 2. This diagram tracks the ambition which is largely tactical functional deployment of infrastructure and scheduling with the mechanisms developed, key variables and parameters, the logistical relationships captured and the design Figures of Merit. The transhipment calculus stands on its own in providing a tool for analysis of terminal impedance requirements with rail service schedules. This acts as a ready-reckoner for assessing the steady state capabilities of terminals to respond to different train schedules and what areas may need reconfiguration and/or technological injection for meeting these schedules. The role of the transhipment calculus goes beyond the analysis of particular terminal process relationships. It may act as the kernel for a systems coordination approach to harmonise critical container and ancillary flows. This
Page | 13
Chapter 1: Introduction
coordination approach covers the range of decision support from precision to infrastructure investment optimisation. The transhipment calculus theory is synthesised with the coordination theory in Appendix F.
Page | 14
Chapter 1: Introduction
Figure 2: Scope and Role of Transhipment Calculus Theory (Dotted Elements of Coordination are Outside the Scope of this Thesis and are Outlined in a Paper in Appendix F)
Page | 15
Chapter 1: Introduction
1.5 Thesis Structure
The thesis is structured as follows: Chapter Two identifies the need and problem of ecologically restructuring the freight task and presents a transitions management approach to move transport system formats along trajectories of more sustainable outcomes. Some high level simulation results from the Australian Stocks and Flows Framework are presented. The paradigm is further developed for urban freight intermodalism. The need to control logistical friction in a systems sense is discussed. Chapter Two also acts as a critical review of general logistics theory and practice in order to scope the problem definition. The potential of urban intermodal terminals is discussed.
The potential of urban intermodal terminals to facilitate integrated transport is supported by a hierarchy of indicators for the more sustainable practice of logistics, which is termed Intermodal Balance. This approach may be seen as a basis for investment decisions which may be seen as more symbiotic. That is, investment in effective infrastructure is required that better controls the dynamics of the logistical structure.
This indicator hierarchy allows policy directions in the sustainable evolution of infrastructure deployment to be connected to operations. It also leads to key Figures of Merit which guide design toward a crucial measure of Intermodal Balance: Hinterland Absorptive Capability.
A technical literature review of freight network and facilities models is presented to highlight the research gaps and anticipate appropriate methods dealing integrated transportation systems and the effectiveness of transport infrastructure using intermodal terminals. The problem context for freight movement in Sydney and Melbourne is explained. It highlights the critical constraints associated with developing new network relationships and anticipates the need to investigate attributes of node terminal tactical planning that could support wider network optimisation, especially for distributed resource systems. Explicit outline is made of the container freight task and the waste freight task as two industrial eco-systems that are candidates for restructuring demand-supply relations. This third chapter also reviews freight models and the transport science heritage that are used in freight analysis. There is a critical gap in characterising and planning for intermodal production systems. There have been few attempts to develop a node-centric approach to infrastructure development which can demonstrate how changing capabilities of distributed resources extends system capacity. Investigating the feasibility of urban intermodal terminals is stymied in two ways. Analytical solutions for storage and handling phenomena at urban freight terminals do not have a sufficiently comprehensive and robust form. Secondly, they do not consider wider system
Page | 16
Chapter 1: Introduction
implications. The tense flux conditions, with which intermodal terminals must contend, require modelling mechanisms to address at least six areas: sidings capability; container stack management; pulse correspondence between diffuse and discrete flows; bulk queue control; bimodal overflow control, and the condition of consignment ordering. To have meaning to the freight planner, these mechanisms must show how they impact on value stream mapping and carry multi-attribute information that can be used for multi-path coupling at the terminal and feed into the multimodal, multi-commodity transhipment problem of operations research. Chapter conclusions recommend the development of an activity-based approach for freight which addresses these phenomena. Such a modelling framework would demonstrate the flexibility of intermodal production systems and would guide appropriate retrofits for improved sustainability outcomes.
In the fourth chapter a functional requirements guideline is developed from the objectives of sustainable freight mentioned in Chapter 2 and the modelling gaps identified in Chapter 3. Its focus is on the broad requirements that need to be included in a model formulation for analytical solutions to the urban freight intermodal problem. The phenomena of stack and sidings management, pulse flow mapping, and business rule controls are further discussed under the rubric of multi-path coupling which is the terminal logic that needs to inform the system multi-modal, multi-commodity transhipment flow problem. This chapter outlines functional requirements, rather than being a functional specification. It does not prescribe the modelling method. Therefore it is left to future chapters to substantiate the value in the analogy to electrical circuit theory on the basis of these functional requirements.
In the fifth chapter, the model formulation is presented based on the freight logistics phenomena to be captured, as outlined in the functional requirements. This covers the variables, system states and figures of merit which can be represented in an electrical circuit analogue stencil. Freight terminal distribution systems are approximated as Linear Time Invariant systems and a stencil based on an electrical circuit analog is utilised to derive interface dynamics (integro- differential equations). Control design at the terminal level is made with utilisation of an Operational Amplifier circuit analog. A theoretical application of compensation feedback control is made to intermodal freight terminals. Analytical mechanisms for several of the phenomena are explained. The following two case studies further scope possible applications of the theoretical modelling framework developed in Chapter 5.
Chapter 6 contextualises the problem definition and model formulation for the Australian seaport-landside condition. Intermodal systems are introduced to effect peak shaving of the container road task and relieve seaport-landside constraints. Critical phenomena are explained in terms of the model formulation: space management techniques of bulk queue control and Page | 17
Chapter 1: Introduction
stack handling are illustrated with reference to the proposed Altona intermodal cluster. Sidings activation and bimodal overflow mechanisms are illustrated for the Dandenong terminal. The complex bundling mechanism is also outlined with reference to Enfield Terminal. Complex bundling is proposed as a means to incorporate multipath coupling mechanisms of the terminal with the multi-modal, multi-commodity transhipment problem. Filter logic flow diagrams are constructed for the bimodal overflow and complex bundling mechanisms as a means to illustrate the use of the electrical circuit analog to discern useful Figures of Merit. These diagrams also set the scene for future work.
The waste transportation case study is presented in Chapter 7. A systems dynamics model is applied to understand the critical constraints on the Clyde terminal with current and future flows. This demonstrated strengths and weaknesses in the classic system dynamics approach in analysing urban freight intermodalism. This review pointed to the significant attributes which the transhipment calculus needed to consider for tractability. It is theorized that the location of compaction and streaming activities fundamentally affects the transhipment and transport solution form. Novel modelling solutions are required to assess this interaction. The compaction unit process is modeled using a circuit analogue as a demonstration. Further mechanisms are outlined for itinerary control. A protocol of modelling scenarios in waste transportation management are described to foresee how the circuit analogue could assist planners to reconcile alternate train operating forms with intermodal impedance of configuration and operation.. These various satellite hub arrangements are illustrated as examples of a freight production system retrofit.
A key area of future research from the above work is outlined in Appendix F. Here the role of the transhipment calculus in supporting network coordination objectives is explored. The comparison with power distribution systems is made and the steps for a coordination mechanism are presented. This load following approach with coordination schemes offers implementation of models using engineering-security approaches to address issues in the urban freight intermodal system where they are considered as distributed resource networks. The theoretical case study of empty container repositioning is presented as a basis for future research focus given its current impact on Sydney and Melbourne.
The conclusion summarises the contributions made by this thesis and anticipates future work, particularly in the field of distributed resource networks and network reliability assessments.
Page | 18
Chapter 1: Introduction
1.6 Findings
The electrical circuit analogue provided mathematical stencils which were used to illustrate mechanisms of pertinent freight phenomena in urban intermodal terminals. The investigation yielded the following observations on the usefulness of the analogue in tactical design:
1. A value – consignment flux balancing around Logistical Unit Processes; 2. Tentative insights into value stream mapping; 3. Demonstration of the evolution of the impedance dual of resource saturation and cost; 4. Space management opportunities using filter control mechanisms of Operational Amplifier stencils; 5. Complex bundling to actualise multi-path coupling through terminals using such filters and to foreshadow further work into dynamic routing mechanisms for system coordination; 6. Design Figures of Merit could be defined from the above areas.
The case studies provided the means to illustrate certain terminal phenomena. These findings may be generalised to certain contributions to knowledge
1.7 Contribution to Knowledge
1.7.1 Terminal Analytical Mechanisms
Prevailing freight models, both facilities design models and network strategic- tactical models poorly differentiate between the concepts of time and timing. Associated with this is the research into solving freight problems that tries to reconcile the transport task with the often dissipative logistical structure. A model formulation of timing would allow essential qualitative aspects regarding synchronisation of freight flows to be incorporated in the transhipment problem of Operations Research. This allows precision analysis of freight fluxes and their transformations at terminals, with dispatch schedules. Freight models are inadequate in this measure of flexible capacity. Analytical models for sketch planning are just as deficient. Space management implications with changing flow profiles are not well captured. Gauging these relationships is vital when assessing urban freight intermodal investments. This thesis proposes
Page | 19
Chapter 1: Introduction
a means to address this multi-attribute concept and build in valuable terminal design aspects in sketch planning.
The contribution of this thesis to the transport science of urban freight intermodalism lies in the re-conceptualization of system impedance of transport-logistical relations using electrical circuit analogues. This defines an activity-based approach in landside container freight modelling centered on the capabilities of intermodal terminals. The analytical mechanisms developed in this thesis yield both performance indicators and sketch design Figures of Merit that cover the following attributes incorporating the concepts of time and timing:
1. Testing the value proposition of an urban intermodal production system by tracking value density through a terminal according to: a. Consignment flow changes b. Terminal configurations. This combines the costs of the transhipment activities as well as the batching value accumulated. The planner obtains an understanding of terminal value stream mapping which allows them to assess the contribution or otherwise of specific terminal functions.
2. Developing a characteristic impedance for each terminal stage which assesses input flexibility of the terminal: how it best fits in the role it has in the hub network type and how it can best respond to pulse flows. This allows an analysis of space management requirements and opportunities at the terminal that are necessary to support the transhipment and transformation of consignment flows. 3. Developing the mathematics of a transhipment calculus to combine precision and controllability criteria lacking for intermodal analysis in transport science. This allows critical business rules to be incorporated into terminal planning to assist in the assessment of flexible capacity. 4. Mapping the requirements of train operating forms with terminal activity through summarising a dynamical system of multipath coupling through the terminal. This better conjoins the costs of taking an itinerary through the terminal with service cycle frequency. 5. Extending the multi-commodity, multi-modal representation and solution of metropolitan container activity by applying impedance transfer functions for each itinerary taken by a carrier. Analysis of multi-commodity flows contributing to train service loads from the intermodal satellite to the seaport can underpin complex bundling approaches essential for the viability of urban intermodal system formats.
Page | 20
Chapter 1: Introduction
This is further demonstrated by using synchronisation information of different flows to coordinate itineraries in an intermodal terminal leading to better matching of the vessel pulse loads which will grow in frequency and magnitude at the seaport.
In essence, the contribution made through this novel modelling device is that the planner can identify the precision requirements and value proposition of terminal operations with different train operating forms. Furthermore, these system formats can be tested to assess how amenable they are to the application of novel business rules. Thus the planner can gauge their degree of flexible capacity. This offers an elegant and useful method to incorporate dynamic assignment of consignment flow in a terminal.
These mechanisms provide Figures of Merit which also lead to an appreciation of Hinterland Absorptive Capability, the flexible capacity of landside infrastructure to support a logistical structure that maximises vehicle load utilisation. Such insights at the sketch planning level can grant planners the skills to develop more robust intermodal policy and infrastructure investment proposals and direct more advanced modelling to this end. Indeed, infrastructure investments in intermodal facilities can be considered in their contribution as effective infrastructure, not simply investment that fuels efficiency based on underlying dissipative relations.
1.7.2 Novel Indicators and Figures of Merit
Although the above contributions to knowledge are framed in the context of transport science, there are broader contributions made to the debate about sustainable cities and infrastructure. The design method provides a hierarchy of indicators and figures of merit, which are considered significant in making the transition to more sustainable transport systems (Black et al., 2002, Doust and Black, 2009).
This hierarchy considers the following interconnected layers:
1. Repositioning the Freight Task; 2. Terminal Interface Productivity; 3. Flexible capacity; and 4. Storage stability.
Page | 21
Chapter 1: Introduction
This thesis draws a connection amongst the four levels of the hierarchy. A particular focus is on the steady-state relationships amongst these attributes. This hierarchy also sheds insights into the major trade-offs of system complexity and controllability which beset a transhipment network. Analysing such a hierarchy in intermodal system design allows strategists to check the vector of sustainability in making retrofits in physical and logical systems which complement and control the logistical structure. Whilst this thesis conceptually covers ground on all four layers, the mathematical focus in on the tactical second and third layers.
This thesis makes the connection between harmonisation objectives (Repositioning the Freight Task) with Terminal Interface Productivity as follows:
1. Translate harmonisation attributes to desirable terminal performance as network levers by defining load following requirements at the terminal level, 2. Load following requirements become impedance specifications, 3. Use of Figures of Merit to test the model to see whether specifications have been met, 4. Direct the user to alter parameters governing impedance relationships, 5. By this mechanism, the user has a means to relate the terminal operations with the rail operating form requirements that interface the terminal.
The Figures of Merit relate storage productivity with throughput productivity. In this way, two critical attributes of freight transport are captured that covers both time and space utilisation: Zeitnutzung, utilisation of time, and Auslastung, the utilisation of node and network space and other physical resources (Tarski, 1986). These attributes act as a vector for the performance indicators of Logistical Unit Processes (LUPs). The Quality Factor measure is a compilation of these Figures of Merit. The value stored is a function of the indicators value density stored, the rate of accumulation, and the saturation inertia. The value dissipated represents the value delivered to the load and is a function of the batch gain, relative and fixed costs and the throughput rate of the Logistical Unit Process.
Page | 22
Chapter 1: Introduction
1.7.3 Anticipating Future Research
The analogue has provided limited insights into tactical design of intermodal facilities. Further work is required to build on this by 1) confirming broad flow and flux relations against cumulative flow curves; and 2) enumerating parameters specific to terminal layouts and handling technologies and correlating different canonical stencils with known productivity measures.
Additionally, the transhipment calculus mechanisms outlined above provides an activity-based approach for freight intermodal design where the load following mechanisms elucidated above can be used as a centre-point for network coordination mechanisms using techniques from power systems theory such as hydrothermal coordination so that distributed resource system initiatives can be trailed to respond to the urban freight transportation task.
A coordination theory would also contribute to network productivity by identifying storage utilisation arrangements which will lead to peak shaving of the road network task around critical access points, such as a seaport gateway.
The node-centric approach of service generation developed in this thesis provides a means of system coordination rather than constraint based scheduling. It can be extended to assess the flexible capacities of different system formats by polling inter-temporal storage availability and applying economic- engineering solution techniques to maximize unit space utilisation, known as Industry Benefits of Trade. This is vital for assessing the feasibility of intermodal production systems, especially in urban environments where they are currently deemed infeasible due to a poor understanding of how existing networks can be leveraged.
This future work is outlined in Appendix F.
Page | 23
Chapter 2: Research Design for Sustainable Freight System Formats
2 Research Design for Sustainable Freight System Formats
2.1 Introduction
2.1.1 Chapter Guide
In this conceptual chapter, the research agenda for the thesis is presented. The chapter explains the motivation for and, benefits and requirements of investigating the feasibility of urban combined freight systems, particularly those utilising intermodal sequences. The investigation centres around developing a systems engineering approach to determining the feasibility of intermodal production systems. Feasibility involves physical, engineering and sustainability criteria. There remains a clear case for researching sustainability criteria in the interaction between logistics and transportation infrastructure and this is demonstrated by establishing the physical and logical feasibility of intermodal production systems.
The chapter commences with a discussion of the container freight task as an industrial eco- system having the traits of a large technological system. Different system formats of the freight task involving infrastructure, logistical structure of actor relations and goods movement attractors and generators designate a particular ecological imprint of the freight eco-system. A salient trait is the degree of integration of such a system in other industrial eco-systems that make up society’s anthroposphere10. Means to manoeuvre to improved vectors of sustainability can be impeded by a system’s degree of embeddedness with other industrial-ecosystems. Thus while policymakers can strive for efficiency gains, they may well be undermining longer term effectiveness in environmental outcomes.
This chapter sets the scene by including some modelling results with research undertaken at CSIRO, Canberra, by the author using the Australian Stocks and Flows Framework Model. The purpose was to understand the nature of the passenger and freight transportation task as embedded into the Australian economy; impediments to dematerialisation and trajectories of more sustainable futures. The results strongly suggested that there was no one technological optima for a sustainable trajectory and that consolidated land use and novel logistical hub structures were required to effect mode shift and reduce the freight road task.
10 Total industrial systems that define human activity Page | 24
Chapter 2: Research Design for Sustainable Freight System Formats
The container freight task is then analysed with reference made to the landside international container task and the waste transportation task. Both eco-systems are identified for vectors of unsustainability and conceptual applications of intermodal road-rail forms are introduced for their benefits and requirements.
This chapter is connected to the others in the following ways: it underpins and structures the research drivers that guide the literature review in Chapter 3; it anticipates the theoretical analysis in Chapter 5 of how the feasibility of intermodal production systems is to be demonstrated, and also outlines the role of the case studies in Chapter 6 and 7 respectively in demonstrating the concepts devised here. It particularly articulates the need for a systemic and transitions approach in intermodal retrofits which anticipates the coordination theory outlined as further research in the concluding chapter and in Appendix F.
2.1.2 Freight Task as a Large Technological System
Transportation systems imprint their operations as particular system formats. These system formats consist of a particular mix of stocks and arising flows. Such infrastructure is likely to remain located in the same place because of land use constraints. Thus the term infrastructure skeleton is apposite. Stocks such as roads, vehicles and intermediate storage nodes cause flows across the network. Transportation system formats can cover a variety of new and older technologies. These system formats are often mono-modal. The technology of car stocks and road infrastructure lock-in traffic behaviours with an impact on the anthroposphere.
System formats “are like putty in these early days and turn to hardened clay as they age” (Gifford and Garrison, 1993 p.115). The policy implications of lock-in are profound. Repercussions of paths taken can be tracked for their lock-in effects (though often assessed only in hindsight). Inappropriate action or lack of action is also significant. Starkie (1987) recognised the attempts of urban and transport policy makers to backtrack over mistaken paths or to remediate at a later date, and are an example of catastrophe theory. That is, it requires many more resources later to achieve improvements at the necessary rate if wrong decisions are made. Conventional transportation systems require huge resources to maintain mobility requirements. This can be observed through the widespread mismatch between policy assertion and practice where governments claim sustainable transportation infrastructure budgets yet the great majority of investment remains in road building and upkeep (Tapio and Hietanen, 2002). A supply response without the explicit objective of mode integration leads to, transportation planning
Page | 25
Chapter 2: Research Design for Sustainable Freight System Formats
fated to address a supply cycle culminating in the evolution of “system deficiencies” (Kanafani, 1982 p.67).
It seems that quite often, the functional attributes of servicing infrastructure are confused with its fixed location. This can frustrate the introduction of functional change among a set of infrastructure and thus deny the benefits of retrofits achieving urban leverage. A means to achieve a breakthrough in a locked-in transportation system format is by a transitions management approach with vintage modelling. This is demonstrated in the urban freight case study presented in this Chapter.
De-materialisation, also known as de-coupling, remains an elusive phenomenon for control in industrial eco-systems because they are a large technological system embedded with other systems (Fischer-Kowalski and C.Amann, 2001). The use of the Kaya identity (Figure 3) conceptually develops de-materialisation into a larger framework of components contributing to environmental impact (Azar C. and Schneider, 2002). We may consider various initiatives in reducing the environmental impact of the freight task identified by Pastowski (1998), according to the Kaya identity. For example, trans-materialisation is achieved through technological advances (fuel efficiency and material use in vehicles). De-materialisation is pursued through Transport Demand Management initiatives such as: load factor increases to the unit vehicle; reduction in the metabolism of materials; and substitute information exchange for the transport of goods. Structural change is undertaken, for example, in rearranging the spatial locations of generators and attractors of freight transport (manufacturers and retailers); and developing novel network relationships that may lead to mode shift, reduced fleet size requirements and a better utilisation of network infrastructure.
The Kaya Identity (Azar et al, 2002)
I= I [I/kg] * m [kg/utility]* u[utility/capita]*P[capita]
Trans- De- Structural Change materialisation materialisation • Efficiency of materials • Change in Production • Recycling • Change in Consumption • Substitute M aterials • Substitute Services • End of Pipe Technology
Figure 3: The Kaya Identity (Azar C. and Schneider, 2002)
Page | 26
Chapter 2: Research Design for Sustainable Freight System Formats
2.2 The Ambition of Combined Transportation
2.2.1 Ecologically Restructuring the Freight task
Ecological restructuring embeds de-materialisation by redesigning the stocks (infrastructure and vehicles) and their relations in the anthroposphere such that flows are transformed and integrated. These are seemingly elusive pursuits which have been identified by peer industry bodies in infrastructure development (AUSCID, 2002) as well as research institutes (Foran and Poldy, 2002) to address the crisis in infrastructure spending, and more fundamentally, the materialising metabolism effects created by the given infrastructure skeleton and land use patterns. We can nominally define the freight task (in tonne-kilometres) as the translocation of goods, generated by production–consumption relations in time and space and mediated by the transport infrastructure network. An assessment of how a freight transportation system utilises spatial and temporal resources is central to the understanding of how synergistic the set of logistics practices and infrastructure supply interaction is. This degree of synergy has implications for the system’s sustainability. That is, the logistics–infrastructure interaction represents an effective use of spatial and temporal resources.
2.2.2 The Nature of Combined Transportation
Combined or integrated transportation remains a loose concept with a range of possible outcomes (Potter and Skinner, 2000). The compelling objective is to increase the capability of the system with the minimum physical augmentation of the road network. Benefits for both the passenger and freight task include obtaining both scale economies and reducing inventory risk and environmental impacts (Bukold, 1996) and network benefits of improving accessibility with frequency based intermodal connections (Mees, 2000). Combined transportation is a concerted strategy to improve transportation benefits with fewer resources. It often can benefit from leveraging off existing infrastructures. For the freight task, achieving combined transportation outcomes is a major undertaking as it requires better harmonisation of system resources and inevitably affects the existing logistical structure of power relations between actors. The term intermodal balance was coined by Vuchic (1999) to describe the technical envelope requirements of integrated transportation. The interplay of activities affecting intermodal balance can be seen from the specific freight task of international container freight with its seaport- landside interfaces Figure 4. Page | 27
Chapter 2: Research Design for Sustainable Freight System Formats
Figure 4: Container Freight Task Using Rail-Road Intermodal Forms (GAO, 2007 p.5)
Figure 4 only broadly outlines the intermodal activities involving the means of transportation for this freight task. Storages distributed through the logistical chain are also critical. Intermodal interfaces must also deal with the multi-commodity flow problem where consignments of full containers or less than full containers of various origins and destinations must be logically and physically mapped to alternate carriers. The behaviour of actors affects the logistical structure and this adds to the complexity of the operational relationships anticipated by Figure 4.
2.2.3 Desirable Frame of Reference for Combined Transportation Modelling Framework
In seeking the desirable design retrofit of a transport infrastructure system, we are confronted by where to start the analysis. This is particularly perplexing when we wish to consider initiatives that embed demand management activities. When we are considering the transition to a more sustainable set of arrangements for the given infrastructure, it is necessary to start from a basis of existing physical assets. Before a retrofit can be implemented, we need to consider the available engineering opportunities and constraints that arise from this physical arrangement as well as how such a retrofit matches and also steers prevailing levels of demand (or load). Outhred’s four layers of transitions assessment (Figure 5), developed from the electrical power supply field, is helpful in commencing an analysis in redesigning logistics service networks with physical networks which can incorporate demand management controls (MacGill, 2006). For instance, when intermodal flexibility is investigated, different arrangements between consolidation networks and exchange nodes which are the components of the intermodal production system need to be considered. Intermodal delivery can have many logistical forms. This requires the engineering system to deliver a set of compatible temporal characteristics which may require further modification of the physical system. Ancillary services such as information on availability of resources to trade and amenable governance link the available infrastructure with actor behaviour.
Page | 28
Chapter 2: Research Design for Sustainable Freight System Formats
The freight logistical form has been devised by considering demand management and supply control alternatives. In this sense the concept leads to an understanding of the opportunities and requirements for retrofitting infrastructure. Consequently, this conceptual model may grant planners the capacity to make more symbiotic investment decisions. That is, to support investment in infrastructure which goes beyond capacity injection to consider demand management opportunities such as improving truck and carriageway resource utilisation and long-term accessibility to key transhipment and beginning and end nodes.
Figure 5: Conceptual Model of Designing and Managing Intermodal Retrofit (Outhred, 2002)
It may be surmised that making the intermodal system “attractive” or achieving intermodal breakthrough is more than correct pricing (the commercial phase) – it needs to consider existing infrastructure, the engineering feasibility for change, economic power relations as well as commercial prospects. It is truly a retrofit approach. The analysis of intermodal flexibility, through this method, seeks to develop network retrofits which embed a relationship between the logistical and infrastructure systems which work for the containment of peak demand through decentralisation and differentiation of functions.
Economic power relations are critical in supporting the feasibility of intermodal terminal production systems (SFCNSW, 2007) For instance, the operations of a terminal must complement the bundling network in which it operates with regard to operating times and functional support. It must also be sensitive to the dictates of mainline rail access and access to
Page | 29
Chapter 2: Research Design for Sustainable Freight System Formats
the destination gateway such as a seaport. The terminal must exhibit aspects of an open access regime where there is an equitable and transparent cost structure for all actors. Applying the modelling framework suggested by Outred allows consideration to be given to the nature of terminal interfaces. These interface include modifications to the logistical structure such as appropriate governance regimes and business rules which will interact with the technological system and are critical to evaluate.
The procedure outlined by Figure 5 may also resolve the predicament of regulatory abstention. In the case of urban freight distribution centres11 this is due to the tension between considering terminals in transhipment networks as either user attracting systems or more prescribed direct delivery systems (O'Kelly, 1998). This has considerable implications for model design and scenario application. It is often perceived that existing commercial relationships must be the basis for seeking an optimal solution and any realistic rendition in a model of the system under study. This can stymie sustainability design. For instance, despite public distribution centres having identifiable gains in reducing general freight traffic, their implementation in Kassel (Kohler, 1999), Los Angeles (Reagan and Golob, 2005), and Ottawa (Transport Canada, 1979), have not succeeded due to approaches that accept the belief that this infrastructure must be a user attracting system. That is, covenant agreements are in place among users of the facilities. In some cases, these initiatives do not come to fruition as preliminary surveys suggest that there would be little uptake of these facilities. The analytical basis for locating public distribution centres in Japan is more advanced mathematically, considering these terminals as delivery systems that can be optimised according to a single goal, i.e. minimisation of cost-delay impedance (Castro et al., 1999, Taniguchi et al., 1999a). Consequently, devising a governance structure to cover port and hinterland relationships is critical for correct intermodal design and implementation. Key Performance Indicators are required to support these initiatives.
11 Urban Freight Distribution Centres can involve both intramodal (bimodal) and intermodal forms of transformation and transhipment. They usually involve solely intramodal operations . They are a component of an extended container freight network that may involve intermodal terminals. They are not directly assessed in this thesis. Page | 30
Chapter 2: Research Design for Sustainable Freight System Formats
2.3 Materialisation of the Freight Task
2.3.1 Predicament and Measurement
The Total Material Requirement of Nations (TMR) is increasing (Adriannse et al., 1997), particularly for Australia (Picton and Daniels, 1999), and the move to service industries has increased service intensity per unit population (u) greater than reductions achieved in reducing end pipe emissions (i) and the physical quantity of goods (m). In freight terms, the novel logistics developed in recent decades (see Table 2) represents a major growth in the service component of freight movement. Ogden (1992) has noted the service complexity of freight: timeliness, arrival in space, quantity required and quality desired. Additionally, there is growing concern about the expanded transport infrastructure required to carry increased freight in terms of available finance required, land resources and the impact on the existing network. This has led to the suggestion of non-physical means to solve missing infrastructure links, such as improvements in technology and communications that will increase system capability (Maggi, 1994). The challenge is to investigate how these characteristics can be maintained yet reduce the growing impact of the freight task.
The service component of the freight task has risen disproportionally to the changes in commodity flows. Sydney, as a major Australian conurbation and global hub, has experienced tremendous growth in the intra-urban and transit freight task. The diffusion in spatial structure of origin and destination relations is a noticeable burden on the urban freight task (Rimmer and Black, 1982). In Sydney, for instance, articulated road freight in tonne-kilometres has increased by a factor of 12.1 compared to only a 3.5 factor increase in total tonnages over the period 1974- 2001 (Rimmer, 1978, ABS, 2003). This also compares with the experience of Sweden where Roth (2000) has commented that freight tonnages have reduced by 29% between 1975 and 1997 but transport work (tonne-kilometres) has increased by 64%. Tapio (2005) has identified that, for Europe, freight transport has moved from a condition of weak decoupling (a small decline in the rate of growth with GDP growth rates) to, more recently, expansive negative decoupling, or re-linking (growth in the freight transport task greater than the growth of GDP).
The metabolism of the freight transportation system is characterised by a network of links and nodes utilised in time and space by an underlying land use. An understanding of the drivers to the utilisation of the network allows us to investigate the means to control the growth of the urban freight task. Roth (2000) and Pastowski (1998) have argued that the growth of the freight task (in tonne-kilometres or Vehicle Kilometres Travelled) should not be complacently accepted as a necessary condition of rising Gross Domestic Product. Page | 31
Chapter 2: Research Design for Sustainable Freight System Formats
There are many pan-European initiatives (thematic networks) to work on reducing the impacts of freight and understanding the impediments to combined freight transportation forms (BESTUFS, 2003, SULOGTRA, 2002, Themis, 1999, TERMINET, 1998). These coordinated initiatives have arisen due to the pressing need to investigate means for a sustained de- materialisation of the freight task between member states.
2.3.2 Recasting Measurement to Drive Improvements
It has been argued that the measure of tonne-kilometres does not lead to transparent initiatives to modify the underpinning logistical structure towards more sustainable outcomes (Roth, 2000, Hassall, 2008). A measure of freight intensity in kilometres/tonne (km/t) is more instructive. This measure is particularly revealing of the concentration of loads in a fleet to reduce trips. Applying this metric, Roth (2000) observed that shippers and logistics companies with so-called green credentials had actually increased the freight intensity of their operations. This was due to increasingly long supply and logistics chains with new actors12 and Just-In-Time stock ordering imperatives.
The measure of freight intensity (km/t) allows the planner to consider the work which the operations are undertaking using the distribution network, and how dissipative this may be. An instance is the number of intermediate node visits (visitation calls) each load undergoes between original consignor and consignee. The measure is also insightful in discerning the benefits of consolidation consignments onto railheads for trunk line-haul. Mckinnon and Woodburn (1996) have hypothesised that a related notion of value density (value of consignment/ distance hauled) may also support initiatives to consolidate freight and reduce the absolute number of trips. A calculation is presented based on container transportation of single trips between Dandenong and the Port of Melbourne, a distance of approximately 25km (Table 1). Both measures of freight intensity and value density are illustrated. No additional penalties, such as low backloading rates or empty container costs, are considered. Nominally, the tonne-kilometres (tkm) between the two methods are the same (75 TEUs at 30 tonnes each, transported 25km being 56,250 tkm). If taking into account the added Pick up and Delivery (PUD) leg of the intermodal transport option, this option has an addition tkm of 11,200 tkm. These results are radically reversed when considering the freight intensity. Due to consolidation, the intermodal option has a factor four lower freight intensity than the all-road option (0.2 km/t compared with
12 These actors consist of 3rd and 4th party logistics providers. 3rd party providers are those that manage the component supply of the production chain; 4th party providers are those that manage the logistics task such as freight forwarding activities of warehousing, storage and distribution. Page | 32
Chapter 2: Research Design for Sustainable Freight System Formats
0.83 km/t).13 These new metrics allow for a re-casting of the container freight task value ledger and also raise the need for value stream mapping of intermodal terminal functions to record how they affect the load rates of mobile resources.
Table 1: Value Ledger of Road Only Haulage vs. Road-Rail Intermodal Service (Dandenong to Port of Melbourne Example) Summary Road Intermodal Units Freight Intensity 0.83 0.2 km/t 30 2250 t/batch Value Density 1200 5696 $/km 30,000 2,250,000 $/batched trip 1.2 5.7 t/km
Calculation Road Intermodal Note Intermodal: Inclusive extra $40/TEU for lift to road for last mile delivery Handling and to wharf. Assume rail Transport Costs 385 425 $/TEU access cost neutral Value of freight 30,000 30,000 $/TEU 30 30 t/TEU 1000 1000 $/tonne 1 75 TEU/Batch per trip Account for 5.3km road trunk for each TEU to be Distance travelled 25 395 km railed Freight intensity 0.83 0.2 km/t 25 5.3 km/TEU 1.2 5.7 t/km 1200 5696 $/km Value
2.4 The Sketch Planning Endeavour as Transitions Management
2.4.1 Transitions Management Technique for Sustainable Futures
Transitions management is a strategic thinking approach which has the ambition to manage trajectories of performance of large technological systems. It is a tool of futures studies which consider how to structure the uncertainty of evolving parameter relationships over long durations (typically greater than 50 years). Because parameter relationships are uncertain in an extended future, the motivation is exploratory and explanatory in nature rather than predictive. In transportation, such relationships to be “reviewed” can cover the parameters of GDP and freight activity, kilometres of built road and kilometres travelled. The motivation is explanatory and exploratory rather than predictive. This is to identify vectors of evolving unsustainability and given the foresight that is made available to the modeler today, take corrective paths which
13 Value net of costs has not been included in this calculation as costs due to transhipment impedance have not been considered. Page | 33
Chapter 2: Research Design for Sustainable Freight System Formats
are not yet locked out by constraints of built infrastructure. Consequently, the linking of scenario narratives can be an even more powerful technique in futures strategic thinking. (Hoejer and Mattsson, 2000) observe that scenarios can have an ambition additional to devising some alternate future: scenarios provide a means of searching for new paths along which development could take place. This is particularly appropriate when conventional paths do not solve a problem, or only suggest marginal improvements. In this sense, the model is used to explore how a desirable state might be reached, which is essentially the nature of a transitions approach. Transition management is a method of devising trajectories of scenarios (Rotmans et al., 2001). Such transition management requires the ability to develop complex scenarios and assess their biophysical implications.
To gain guidance on desirable frameworks for human–machine interaction that track transitions, real-time control theory should be considered, although it is impossible to condense the future and manage it in real time. A possible proxy is the use of scenario management that maps a feasible transition pathway into this future and notes the limits on deviations from certain paths. Rarely are paths and their disturbances known. Therefore, a transitional management approach is most necessary for futures studies. The user should be able to intervene at different points in time according to the system response (Figure 6).
I1 I2 I 3 I4 I5
P5 P5
P3
P4 P1 P3
P4 P2 P2 P1
Figure 6: Process of Modeler Intervention of Parameter Relationships Midstream to Track Feasible Transitions (De la Barra, 1989)
Page | 34
Chapter 2: Research Design for Sustainable Freight System Formats
2.4.2 Freight Futures Simulation Example
When a number of interacting policies are required to produce transition, feasible pathways need to be assessed that capture cross-cutting issues (Rotmans et al., 2001). For instance, the activity of urban goods freight is rising due to logistics drivers and control of operations from a systems point of view seem elusive (Rodrigue et al., 2001) Light goods vehicles represent an increasing traffic load in the urban transport network representing 30% of the urban road traffic and 33% of the urban transport fuel consumption (BTRE, 2003 Figures ES.1, 2.6, 3.4). The total energy intensity of freight is 48 mega joules/tonne-kilometre (MJ/tkm) (Lenzen, 1999) which is surprisingly high compared to other freight types due, primarily, to the low load factors. Applying different scenarios and strategies, it was first considered how to control the energy intensity of urban freight. The investigation was made of the effects of introducing electric vehicles into the fleet at different diffusion rates to gauge the prospects of reducing hydrocarbon energy demands (Low Elect and High Elect) and then hydrogen fueled vehicles (Elect & Hydro). Accelerating retirement of the existing stock as well was then considered (High and Retire). A new scenario approach was considered when it was realised the gains were delayed by stock inertia. This scenario was based on network restructuring by improving logistics hub arrangements so that pick up and put down delivery consignments are consolidated (Reduce Intensity). Finally, a combination of these innovations was made (Combination). From an analysis in Australian Stocks and Flows Framework (ASFF), the peak control of energy and emissions from urban goods freight requires a combination of aggressive technological improvements, network and land-use restructuring around public distribution hubs, receiver/sender cooperative agreements and renewable energy sources and suitable infrastructure refueling/recharging depots if Australia is to reduce fossil fuel consumption in this sector to 1990 levels by 2030 (Figure 7).
Page | 35
Chapter 2: Research Design for Sustainable Freight System Formats
3000
2500 BAU
Low Elect 2000 High Elect
1500 High & Retire
Reduce Intensity PJ per 5 year step 1000 Elect&Hydro
500 Combination
0 1950 2000 2050 2100
Figure 7: Urban Freight Hydrocarbon Direct Energy Requirement with Different Strategies -ASFF Output (Beavis et al., 2009)
Applying a vintaging model to scenario narratives reveals that there is no single optimum transport technology for long term sustainability (Mees, 2000). Speculative, and then exploratory, narratives need to be applied to structure a scenario query system which facilitates this necessary interdisciplinary response. Significantly, the engagement with the framework enabled the user to be educated in the interactive nature of the driving forces such as population, physical network typology and logistical service drivers affecting the urban freight task.
Thus we may consider whether the transportation system format provides functional discovery (urban leverage) over simple technical refinements (Gifford and Garrison, 1993). Under technical refinements, technology fuels growth in undesirable elements of our network system and locks-in inefficiencies rather than driving innovations. Under conditions of functional discovery, the initial infrastructure outlay leads to synergies in transportation and other activities. Intermodal freight infrastructure and logistics which foster consolidation networks may allow efficiency and productivity benefits. The concept of functional discovery may also be applied to the use of transportation in integrated waste management. The ability to bring waste together at a transfer station may deliver density and scale economies which enable streaming operations to occur and resource value to be recovered.
Page | 36
Chapter 2: Research Design for Sustainable Freight System Formats
2.5 Freight Metabolism
2.5.1 Structure of the Urban Freight Task
An understanding of the components of the metabolism of freight will lead to recognition of the impediments to de-materialisation. Freight transportation systems should be seen as driven by an integrated demand rather than simply a derived demand for goods (Hesse and Rodrigue, 2004). Integrated demand recognises that there are two flows in the freight task: the underlying material flow and the load unit (vehicle fleet) flow (Priemus and Konings, 2001). Load unit volumes are mediated by spatial relations, transport infrastructure, and the traffic system. Demand for commodities seems directly linked to economic growth. However, the spatial and temporal arrangement of freight in the urban environment is more significant that the actual quantity shipped itself (Wigan, 1978). These arrangements become the starting point of an investigation into the analysis of design for strong sustainability of the freight task.
We can begin to organise the element of the freight task for an urban environment as a means to build an analytical framework. Freight may be seen in the provision of the following functions:
• Coordination of the production supply chain; and • Distribution of goods. These functions may be characterised by further service characteristics (Ogden, 1992):
• Timeliness of delivery; • Arrival in space; • Quantity required; and • Quality desired. Within the supply chain and distribution functions, there are four key sub-functions:
• Inventory control (at the warehouse and throughout the distribution system); • Transhipment (break-bulk functions at a node transient); • Trip Chaining (particularly for urban goods movement); and • Empty movement management. Finally, the freight structure is built on three interaction levels (Priemus and Konings, 2001), namely: commodity (the material flow, in tonnes); unit load (vehicle flow); and infrastructure (capacity provision) (Figure 8).
Page | 37
Chapter 2: Research Design for Sustainable Freight System Formats
Figure 8: Goods Demand and Freight Traffic Interaction (Priemus and Konings., 2001)
McKinnon and Woodburn (1996) identify 4 levels in the hierarchy that affect the level of freight transport: logistical structure, the pattern of trading links, scheduling production flows and the management of transport resources. The growing transport intensity of goods (measured in kilometres/tonne with reference to road traffic) is due to a pre-occupation with the management and optimisation of transport resources (Roth, 2000). Fundamental changes need to be made to the logistical structure to drive down materialising freight activity.
A pressing research task is the understanding of what generates unit load growth and measures to contain it. One means to understand the effects of Supply Chain Logistics (SCL) on the freight task is to see how different SCL initiatives affect key indicators of freight performance such as load factors, Average haul length, Empty Runs, Handling time and Lead time (Zografos and Giannouli, 2001). SCL trends include Spatial Concentration of Inventories, Spatial Concentration of Production, Time Delivery Systems and Hub Satellite Systems. The relationships between SCL initiatives and the structure and utilisation of the freight transport system has been developed as a broad methodology but remains to be empirically tested or
Page | 38
Chapter 2: Research Design for Sustainable Freight System Formats
simulated (Zografos and Giannouli, 2001). The project, called SULOGTRA, links supply chain logistics trends with changes in freight system indicators (SULOGTRA, 2002).
Table 2: Relationship of Logistics Drivers to Operational Freight Indicators (SULOGTRA, 2002)
Operational Transport Indicators Rail Road Load Factor DBBTS, DHSS SCI, SCP, DBBTS, DHSS, TCP Length of Haul WGSD, SCI, DHSS, DBBTS TCP, DHSS, SCI, SCP Handling Factor DBBTS, DHSS DHSS, DBBTS, TCP Lead Time WGSD, SCI, DBBTS, DHSS TCP, DBBTS, DHSS Empty Runs DHSS, DBBTS TCP, DHSS, DBBTS, NDD Mode Share DBBTS, DHSS, WGSD DBBTS, DHSS, WGSD
The logistics drivers are summarised as follows: DBBTS- Development of Break-Bulk Transhipment System; DHSS-Development of Hub Satellite System; SCI-Spatial Concentration of Inventory; SCP-Spatial Concentration of Production; TCP-Time Compression Principles; WGSD- Wider Geographical Sourcing of Supplies and Distribution of Goods; NDD- Growth of Nominate Day Deliveries and Time Delivery System.
It is the interplay of these factors that describe the logistical structure and can lead to policy interventions to improve the urban freight task. From Figure 9 (Sjöstedt et al., 1994) it may be surmised that the logistics (service) and physical networks are closely interrelated. However the operations of the logistics network can place strains on the physical network. We need to seek mechanisms so that they operate more harmoniously. Combined transportation initiatives require a greater understanding between the operations of the two networks.
Figure 9: The Complementary Sub-systems of Logistics and Transportation (Sjöstedt, 1994)
Page | 39
Chapter 2: Research Design for Sustainable Freight System Formats
A number of controllable elements are open to authorities: socio-economic organisation (such as bans on truck sizes in urban areas); technical aids (such as the promotion of new vehicles); manipulation of the locational and activity structure for freight infrastructure (such as new consolidation terminals) or promoting the co-location of shippers and receivers (Rimmer and Hicks, 1978 p.547). The development of intermodal production systems can be seen as a comprehensive application of the third element.
The literature reviewed on infrastructure supply performance measurement has shown it to be sketchy both in terms of network size and in particular the level of activity which utilises resources that require the infrastructure. Uncertainty remains in decision making regarding the acceptable marginal costs of supply for a given freight effort. The concept of value density (McKinnon and Woodburn, 1996) has the potential to identify how planners may measure and mitigate the wanton growth in infrastructure supply and activity. This concept has not been developed further into design tools.
2.5.2 Materialising Pressures of Land Use Devolution
As freight throughput increases with economic growth, and logistical drivers, the governing belief has been for increasing the velocity of the system. In terms of fostering intermodal linkages, the increasing number of transhipments requires a further synchronisation of system resources. With the need to support the economic feasibility of intermodal terminals, the terminal requires a certain amount of volumetric flow. This is considered the economics of density. To further reduce the cost per tonne of translocation of consignments, large handling and mobile resources are considered more feasible. This is termed economies of scale. Transhipment technology requires unitisation of a variety of consignments: there are diverse commodities translocated, to different regions, with different service requirements. This is termed economies of scope. Often these phenomena result in the development of hubs as the physical manifestation of these processes (O’Kelly, 1998). Hub type arrangements facilitate connectivity between connecting places (O’Kelly, 1998). Hubs lead to certain efficiencies. For instance, in airport systems, hubs have been able to cater for the growth in passenger traffic with a lower growth in aircraft resources. They have also however facilitated this growth. Satellite hubs located in the seaport hinterland are one clear development in the integration of the seaport with its hinterland (Beuthe and Kreutzberger, 2001). Slack (1999) has noted that such hubs have recognised local spillover effects including congestion. Additionally the hub paradigm can be such as to reduce terminal reach. Consignments have to travel over increasingly greater
Page | 40
Chapter 2: Research Design for Sustainable Freight System Formats
distances to be transshipped. From an urban and peri-urban perspective, sparse intermodal resources could mean a reduction in the usage of rail to the port gateway.
For apparent commercial viability, intermodal terminals have needed to include a host of functions in their operations. A growth physical-logistical arrangement has been the peri-urban logistics centre. At such centres, terminals provide a composite of functions around freight forwarding activities. These functions include transformation of the consignment through break- bulk operations as well as storage and transhipment of the unit load. This dynamic has however led to cluster operations in the hinterland. Hesse and Rodrigue (2004) have noted that the search for commercial viability, including cheap land as well as a scale economy mentality, has encouraged the dispersal of facilities and the growth of subsidiary logistics industries in a process called “sub-harbourisation”.
The location and size of terminals is often dictated by available land area and economies of scale. Consequently terminals (intermodal and distribution terminals, for instance) can be large. These hubs create extensive traffic which can meet planning approval resistance. A recent report to the Sea Freight Council of NSW (SFCNSW, 2005) has suggested that more than 200 ha of land is required for future intermodal terminals alone in order to cater for the goal of shifting 40% of container volumes by rail in Sydney. There needs to be research in how we can control the size of terminals and their traffic levels in urban areas.
A minimum criterion is required to position intermodal terminal systems as retrofits to the existing transportation system which draws together efficiency and effectiveness in environmental outcomes. Woxenius (1998) and Slack (1999) have observed that this is a latent risk when introducing intermodal terminal networks where more localised desirable hub forms may be ignored for larger inter-city solutions.
2.5.3 Modelling Paradigms and Claims
2.5.3.1 The Perceived Need for Speed
It is a research and planning concern that short-term efficiency gains of building and augmenting physical transportation networks can undermine long term effectiveness and thus overall sustainability (Beavis, Black, et al., 2009). The vicious cycle of logistics, is a compelling example of this dilemma (Boerkamps and Van Binsbergen, 1999) . The cycle is denoted by infrastructure investment inducing further activity to yet another plateau of Page | 41
Chapter 2: Research Design for Sustainable Freight System Formats
congestion. The dilemma associated with this cycle is a major theme developed through the thesis that has received insightful but only piecemeal conceptual development in transportation planning (Starkie, 1987, Rimmer and Hicks, 1978 , Knoflacher, 1996, and Mees, 2000).
The economic justification for the augmentation of new infrastructure (most often roads) is often based on the concept of the Value of Travel Time Savings (VTTS). However, there are problems with using this measure.
1. The application of VTTS as a general welfare measure may be incorrect. The freight industry may receive great benefits from additional time-savings but local residents may lose amenity through new construction and localised concentration of flows created. 2. Time Savings are transitory. New infrastructure often creates induced traffic and leads to saturation levels that curtail the level of service estimates upon which the time savings are based. 3. An addition to the existing network to ease congestion may increase time savings through allowance of higher speeds. However, it may re-route a large proportion of freight traffic to travel a greater distance, thereby generating higher overall energy consumption and greenhouse gases. This was one of the findings of the CSIRO- BTRE study which investigated changing freight loads with the completion of the Sydney orbital (BTRE, 2004). 4. The provision of additional physical infrastructure may engender more logistics activity, for example, the relocation of warehouses to the outer west of Sydney. 5. A single economic measure based on forecasts of network flows does not consider how infrastructure can transform flows to increase load factors and to harmonise discrete- diffuse sequences.
The third salient point has been discussed by Knoflacher (1996). In this he identifies that the pursuit of time savings as a justification for physical infrastructure augmentation is illusory. The demands on the network will grow with the removal of system friction. That is, businesses will translocate to cheap land which is more attractive because it has become accessible, predicated of course on improving the mobility of the automobile, which is seen as the predominant means to achieve accessibility. Along similar lines, Mees (2000) has questioned the use of congestion pricing for road infrastructure investment outlays.
An increase in the velocity throughput of the intermodal system may heighten its attractiveness compared with unimodal options. However this is also a symptom of further materialising pressures. Rodrigue (2005) conceptualises that the segregation of manufacturing and pressure to
Page | 42
Chapter 2: Research Design for Sustainable Freight System Formats
reduce inventories have caused entropic pressures, which can only be countered by major inputs of energy into the system. These energetic inputs include facilitating increased inventory mobility through temporary storage and swift translocation with a larger vehicle fleet. Schliephake (2000) berates an unguided consumption driven demand for transportation and freight services which supports the growth of infrastructure to facilitate a turntable of mobility. Roson (2000) has also cautioned on the consequences of improving the velocity throughput of intermodal services. The materializing or resource consuming aspects of this feature of the logistics-infrastructure relationship appear embedded.
Consequently there is a fundamental problem in planning for the interdependency of temporal factors in transportation (Tarski, 1986). The demand for frequency of services, for instance, can lead to the growth in transportation resources and physical capacity demands. The forces of economies of scale generally lead to a large number of road trips and larger terminals located outside the urban area. If new forms of freight translocation through physical configuration systems are to be developed, they need to consider not simply the increase in mobility that they provide for consignments but how consignments are transshipped in ways that reduce intensity and overall transport load.
2.5.3.2 City Logistics and Reverse Logistics
There are unsubstantiated claims that logistics initiatives and the evolving structure of the freight network lead to greater efficiencies and environmental savings by reduced inventory levels and streamlining network arrangements. They also synchronise domestic with global freight flows - a necessity for economic productivity (DOTARS, 2002). Yet, paradoxically, so- called green logistics can also lead to a number of materialising trends that contribute to a vicious cycle of dispersed activities, reduced truck loading rates, increases in tonne-kilometres, urban congestion, reduced accessibility and fleet growth (Boerkamps and Van Binsbergen, 1999, Rodrigue et al., 2001, Slack, 2001). The growth in the freight task due to supply chain and distribution logistics and the congestion effects this creates has been coined logistical friction (Hesse and Rodrigue, 2004). The logistics agenda has been generally mapped by private and government organisations also with an ostensible focus on sustainability (DETR, 1999) but they do not demonstrate empirically, or by physical laws, how the supply chain and logistics industry be made to have environmentally beneficial consequences for the system as a whole. In this context, corridor-like paradigms for the swift hinterland translocation of container traffic,
Page | 43
Chapter 2: Research Design for Sustainable Freight System Formats
for instance the rail freight Bostwash corridor between Boston and Washington (Rodrigue, 2004) may lock-out more sustainable local freight transport and land-use interactions.
City logistics considers “the process for totally optimising the logistics and transport activities by private companies in urban areas while considering the traffic environment, traffic congestion and energy consumption within the framework of a free market economy” (Taniguchi et al., 1999b). Review of logistics output highlights its limitations in taking a systemic perspective. Similarly, reverse logistics is the growing theory of incorporating the recovery and recycling loops in the distribution chain. It is a supply chain approach often for one product system. It takes the infrastructure distribution system as given. These approaches with the implicit objective of a “sustainable supply chain” represent a problem of truncated system boundaries. Gudmundsson (2004) recognises that sustainability is a system level concept, therefore any claims of sustainable activity must take into account the full context of interactions. This intellectual integrity demands that model outputs be relevant in themselves and directly linked to a web of sustainability performance.
Logistics drivers place tensions on the transport system. The increasing frequency of flows has arisen in order to reduce accumulation costs in warehousing facilities. This does have a one-off, de-materialising benefit with respect to production as large inventories do not need to be held to offset long lead times. However, the inventories have been made more mobile and have thus increased transport intensity (km/tonne per good). As utilisation of the network is driven by a just-in-time imperative, where all components must respond in a synchronised and consistent fashion, tense fluxes are developed (Rodrigue, 1999). For a global-local freight network, where the intensity in use of transhipment infrastructure has increased, the new operating paradigm has created this notion of tense flux, where time-savings are paramount rather than achieving economies of scale. With a system of declining redundancy in time and space, maintaining reliability then becomes crucial. Of interest is how the flexibility and reliability of the infrastructure system may be measured as part of its overall operating performance.
Different sectors, discretely optimising their supply chains, can amplify burdens on the system by pursuing dispositive elasticity (Tarski, 1986) . Dispositive elasticity is the activity of operators using the network to their advantage but consuming temporal and spatial resources which are non-optimising for the whole system. Congestion along key link roads and node accesses is one result of the dispositive activity of logistics operators. The challenge is to provide suitable infrastructure to fit the requirements of the modern production system without fueling freight mobility that poorly utilises the temporal and spatial resources of the physical network. It is these interactions and trade-offs that this thesis wishes to investigate in order to establish new sustainability criteria for transportation infrastructure. Page | 44
Chapter 2: Research Design for Sustainable Freight System Formats
The interest lies in modifying the effects on the system of supply chain and distribution logistics activity through the judicious deposition of distribution infrastructure. For integrated waste management, we are interested in what effects recovery initiatives of a part of the waste stream have on the transport and management of the larger waste stream.
2.6 Conditions in Freight Industrial Eco-Systems
2.6.1 Urban Landside International Sea Container Task
2.6.1.1 Overview
The port as a gateway node siphons transit and urban originating and destination freight. Seaport capacity then must also be measured in terms of its hinterland relationships and interacting infrastructure. The relationship between port related and local flows remains poorly understood in planning and design practice, yet port productivity is now recognised to encompass landside constraints (Marlow and Paixao, 2004).
Global trade exerts pressures on the national freight landside condition. For instance, servicing infrastructure and links in China are not sufficient to handle the variety of container throughput (Wang, 2002). Empty container movement is a significant problem in urban environments for all supply chain members, landside and seaside with large asymmetrical flows leading to storage and transfer capacity problems, for instance around the Ports of Los Angeles and Long Beach (Hanh, 2003, 2007) . This will necessitate greater interaction between freight infrastructure nodes. Australian seaport-landside linkages also need further integration and this must be negotiated through existing heavily built up urban areas.
2.6.1.2 Sydney Condition
A diagram of the Sydney freight network is portrayed in Figure 10. The majority of container freight is shipped through Botany Bay (grey circles depict container activity centres). Whilst there is a network of intermodal terminals in greater Sydney (triangles), the use of most of them is constrained by congested rail space shared with passenger traffic. Only the planned terminals
Page | 45
Chapter 2: Research Design for Sustainable Freight System Formats
of Chullora and Enfield have a dedicated freight line with Botany terminal. Through (transit) freight traffic has recently been provided for with the ring road circling Sydney (the M7 orbital). There is currently no dedicated road freight corridor in Sydney. Eighty-five per cent of container traffic has an origin or destination within the greater metropolitan area (within a 40km radius) to or from Botany terminal (Mack, 2000). Sixty percent of this traffic is south of Sydney Harbour and the Parramatta River. Shaded areas represent origin and destination concentrations of containers. While not more than ten percent of total road freight movements (Brooker, 2005) , they are of high value, they have significant localised diurnal loads and also create secondary and unconsolidated flows that connect with domestic intra-urban movements.
Figure 10: Metropolitan Road and Rail Links Serving Port Botany, Sydney (Triangles are intermodal terminals; Circles are import and export container activity centres) (SPC, 2000)
Rail share has since declined with container freight consolidation at Port Botany. Recently, as part of the planned expansion of Port Botany, Sydney Ports has committed itself to an increased modal shift from short-haul road to rail from under 25% to 40% of tonnages. This modal shift is necessary to control the growth of freight road traffic to approximately 1.4 million TEU Containers/yr (Port Botany will have a planned throughput of 3 million TEU/yr by 2016). New intermodal terminals and the expansion of existing terminals are key concerns in the metropolitan strategy being developed in 2005 (see, www.metrostrategy.dipnr.nsw.gov.au ). Page | 46
Chapter 2: Research Design for Sustainable Freight System Formats
Increasing the rail share to 40% of container tonnages should reduce growing road traffic by 20 million Vehicle Kilometres Travelled VKT/yr by 2021 (Brooker, 2005).
Intermodalism is seen as relieving the landside constraints to port throughput growth. These landside constraints to transit flows have already led to shippers and receivers in the hinterland eschewing the Sydney port for the Ports of Brisbane in the north and Melbourne to the south (SFCNSW, 2004a). Introducing intermodalism on an existing congested network may be a means of relieving road transportation pressures. However, the land-use requirements and local impacts can be prohibitive. The scale of the proposed Enfield terminal is being reconsidered since the Morris Inquiry rejected the proposal for a capacity of 500,000 TEU/yr due to perceived concentration in secondary empty container and other less than container load truck flows causing road congestion around the terminal.
Intermodal transportation has not reached its potential in Europe, nor in North America. This is often due to a perception of large impedance cost. This impedance cost involves delays due to clearance, handling, storage and transhipment. The composition of this impedance cost is however poorly understood and therefore poorly managed. If the composition of terminal impedance is better understood according to a variety of interacting factors, the means to ameliorate this impedance, thus making intermodal operations more attractive, may be possible. Intermodal forms of freight transportation represent the opportunity to leverage off the existing network and increase system capability. It is perceived that the container industry in Australia will need to assess more transhipment options in the future due to the stuffing and un-stuffing of larger containers (40 ft and Hi-cube). A general methodology is needed to assess the nature of terminal-node visitations for given purposes to ensure lowest system cost and improved environmental outcomes.
2.6.1.3 The Lack of Harmonisation
Harmonisation of resources has been coined by the container logistics industry in Sydney and Melbourne to express an aspiration where logistical frictional elements are reduced so that costs are reduced. Logistical friction in the seaport-hinterland business includes:
1. A disparity in the ownership and management of resources such as vehicles and containers leading to lack of synchronisation in pick-ups of imports and delivery of exports (poor back-loading from the seaport);
Page | 47
Chapter 2: Research Design for Sustainable Freight System Formats
2. A mismatch in operating hours between the seaport and other facilities leading to intermediate storages of containers and idleness of vehicles; 3. An asymmetry in export and import container flows leading to poor coordination in the reuse of empty containers; 4. Growing landside congestion on the road network and bottlenecks in gateway access to the seaport leading to local disamenity in residential areas and delays in deliveries and collections of containers. These conflicts are characterised by a diffuse truck flow profile which under-utilises the transport network. Problems achieving harmonization and the planning- modelling gap are discussed in section 3.3.
2.6.2 Integrated Waste Management
2.6.2.1 Internationally
There is a need for a more comprehensive and general analytical approach to waste logistics, which considers the operation of terminal nodes. In Europe, including the UK, intermodal terminal configurations are seen as providing a means to trans-locate increasing volumes of wastes to be treated at a variety of facilities. Existing uni-modal networks (often road based) have increasing capacity constraints. The resource recovery EU directives mean that there will be fewer direct landfill options in favour of a range of treatments. There will need to be transformation of the wastes as well: either in storage, streaming and compaction. To ensure the necessary accessibility and transformation of consignments, a new network of waste treatment facilities will be based on a smart configuration or retrofit of transport infrastructure, which uses inter- and multi-modal forms. The UK STRAW project is an example of such initiatives, to combine waste mass trans-locations with multi-modal solutions (Curry, 2006). Increasing fossil fuel costs may also encourage the development of intermodal transport, particularly rail. Diverting wastes from landfill requires the development of more sophisticated methods of waste logistics. Supporting the right form of logistics is crucial as overall costs may delimit whether the material is considered a waste for disposal or a resource for recovery. This leads to research questions on the pivotal role of transfer station nodes.
Page | 48
Chapter 2: Research Design for Sustainable Freight System Formats
2.6.2.2 Sydney
Sydney transfer stations generally have a compaction role with some separation of recyclables. Their predominant function has been to act as the gateway to landfilling operations. Currently, there is only one intermodal operation for the consolidation and transhipment of solid wastes. This is the Collex facility at Clyde. Here putrescible wastes and commercial wastes are taken from a number of councils and businesses and compacted into containers where they are then loaded onto rail cars and transported to Goulburn to the Woodlawn landfill. These transfer stations have limited storage area and therefore there are tight indirect transhipment arrangements, especially when there is a rail-road interchange.
The lack of streaming activities reduces the ability to cascade the waste fraction. Transfer stations as streaming stations represent a significant transformation in the transport- waste management process. Transfer stations also act as consolidation hubs where economies of density can be obtained. Logistics systems currently do not support cascades. The logistics and waste material flow system then need to work synergistically: recovery of resources from Municipal Solid Waste requires intermediate streaming and consolidation facilities. For significant waste recovery, the transportation network needs to become much more productive. To avoid large increases in capacity supply, integrated network forms using intermodal networks need to be pursued.
2.7 Role of Intermodal Terminals
2.7.1 New Concept of Impedance
Freight should be considered as an integrated demand (Hesse and Rodrigue, 2004) -not only a derived demand as logistics and distribution affect the total demand for freight conveyancing. As Drewes Nielsen, Jespersen et al (2003) have observed the distance traveled depends on logistical structures as well as the pattern of trading links (p.303). Therefore the calculation of network capacity with growing use can be better assessed by considering whole transport geography, the core dimensions being: flows, nodes/locations and networks. Woxenius (2004) investigates the role transhipment terminals play in an urban network by assessing the linkage arrangements of five types.
Page | 49
Chapter 2: Research Design for Sustainable Freight System Formats
Logistical friction is the term to describe the factors which impede system utilisation that extend beyond the link line haul distance travelled (Hesse and Rodrigue, 2004). Logistical friction then becomes a multi-dimensional measure of impedance- it represents the nature of space-time convergence in a network with terminal intermediation. This complex definition of impedance becomes an insightful measure of areas of impediments leading to supply chain inefficiencies (physical and logical).
2.7.2 Accessibility as a Measure of Effective Infrastructure
Accessibility based on system capability to translocate goods with the minimum activity is a more sustainable design objective than maintaining mobility measures. The latter objective focuses on maintaining the speed of vehicles and says nothing about increasing the load density. Roads are constructed to maintain mobility even if this means vehicles need to travel further to reach their destinations. Accessibility to nodes such as seaport terminals is a significant measure of system capacity. Thus alternative connections can assist in improving accessibility. True accessibility measures focus on fostering intermodal relations based on precision and controllability objectives. Mobility measures can easily lead to undesirable multimodalism without developing the integrated dynamics of consolidation networks. Both greenfields and brownfields conditions (requiring retrofits) have the danger of denigrating into multi-modal rather than intermodal solutions.
The role of freight intermodalism as gateways in conjoining modal networks is further developed in section 4.5.2.
2.7.3 From Capacity to Capability: A New Concept in Sustainable Infrastructure Supply
Intermodal balance is a term to describe the technical envelope requirements of integrated transportation. It is seen as a method in facilitating integrated transportation (Vuchic, 1999). The objective of integrated transportation is to increase the capability of the system with the minimum physical augmentation of the road network. This capability allows the handling of varying fluxes as well as providing resilience; the quicker recovery of reliability under conditions of disturbance. Intermodal balance as a manifestation of integrated transportation has not been mathematically formulated to date for either passenger or freight distribution networks.
Page | 50
Chapter 2: Research Design for Sustainable Freight System Formats
A means to overcome logistical friction in the container freight network is the concept of physical infrastructure and logistical harmonisation. Whilst impediments to harmonisation are well documented (SFCNSW, 2004a, VFLC, 2005), jurisdictions in both Sydney and Melbourne have grappled, with only partial success, with the attributes and pathways of harmonisation. Intermodal balance is a means of addressing business activity harmonisation in container freight.
The interest of this thesis is in developing tools to observe how a network infrastructure, configured in space and the possible use of that space, performs according to temporal criteria. If we are able to observe the interaction of these criteria we can assess how the infrastructure mediates the throughput. Alternate terminal configurations can be assessed for how they relieve or accentuate impedance and how they may facilitate different consolidation sequences. Temporal criteria are discussed in section 2.9.1.
2.7.4 Controlling the Evolution of System Deficiencies
Many transportation systems have or will have capacity constraints. The life cycle of a component of the network is limited by the growth in mobility. In transportation planning, there is conceptually a facility for the anticipation of the evolution of these system deficiencies and this conventionally feeds back in to the strategic planning process to calculate critical corridors and to assess alternative transport systems (Kanafani, 1982). The typical tool is the augmentation of physical transportation infrastructure. Increasingly, this is perceived as an ineffective response to capacity constraints from an environmentally sustainable perspective for several reasons (Nijkamp et al., 1994):
• Supply often induces its own demand; • If in the public domain, there is a limit in investment capacity; • Improving transportation infrastructure (such as roads) may reduce the mode share of alternatives; and • Increasing physical supply consumes scarce land.
In sum, we may consider such a response as a traditional supply response devoid of system coordination and demand management aspects. The physical augmentation of infrastructure is often not a sustained solution to capacity limits as it panders to a chaotic deposition behaviour of vehicle operators, called depositive elasticity (Tarski, 1986). Trip activity grows with an
Page | 51
Chapter 2: Research Design for Sustainable Freight System Formats
expanded physical network. What is required is a capacity management approach alongside a demand management approach. Here, capacity is assessed according to the constituents of capacity constraints such as technology, environment, organisational arrangements, economic instruments, and information resources (Maggi, 1991) as well as opportunities to respond to alternate logistics arrangements. Thus, rather than measuring capacity of a network infrastructure by the speed or minimum Levels of Service it permits on road links, cross-modal network capabilities under different operating regimes need to be considered for how they improve access to intermediate and final freight nodes. This is a multi-dimensional view of capacity that is a novel way in considering how planners might leverage existing transportation infrastructure for more harmonious transport outcomes. Capability analysis provides an analytical platform for establishing the feasibility of intermodal production systems and is another way to respond to the concern of missing transportation infrastructure links for a vibrant economy.
The freight industrial eco-system is a complex system. There are various means of addressing complexity in the freight transportation task. There are two discerning pathways (Figure 11). One is the reductionist where infrastructure is either injected to meet growing demand, or there is some modulation of service availability through tools such as congestion charging. The other pathway is integrative whereby information is used to connect tasks and reduce unnecessary activity. The policy tools that may be selected from either path or a combination of pathways must be such that considers the imperative of finding a means to reducing materialising effects of dissipative activity.
Page | 52
Chapter 2: Research Design for Sustainable Freight System Formats
Figure 11: Approaches to Transport and Logistics System Complexity (Waidringer, 2001)
Waidringer (2001) developed metrics on assessing the complexity of freight supply chains based on their number and degree of intermediation. Broad ratios of network, process, and stakeholder complexity are defined. The usefulness of these measures is to benchmark the operations of different shippers and to see how supply chains of specific commodities may be simplified.
The complexity problem can be more usefully re-cast: complexity adds to the evolution of system deficiencies when it represents increasing dissipative activity of discrete- diffuse freight flows. The management of complexity can be achieved through synchronised consolidation and distribution activity at intermodal terminals. Complexity has space management and synchronisation implications at the terminal. Complexity may be necessary, especially in small systems characterized as fine- meshed: a many-to one relationship requiring numerous shippers to support minimum train shuttle payloads to justify the intermodal operation. Consequently a more relevant and intriguing line of research in urban intermodal design may be that of Sjöstedt et al., (1994) who consider the tradeoffs between complexity (openness versus specialization) and control (terminal flexibility).
Page | 53
Chapter 2: Research Design for Sustainable Freight System Formats
2.7.5 Urban Leverage by Inter-temporal Generation and Storages
Consider an existing transportation network as containing pockets of capacity reserve. This represents the ability to generate freight services and provide storages for consignments which will require later freight services. The uilisation of system capacity reserves will depend on the infrastructure function, degree of connectivity and how that network operates. Capacity reserves then represent both the ability for novel activities to occur within the transit task and for necessary redundancy in the system. There is regular asymmetry in utilisation of reserves during peak demand and perceived limitations on demand management tools often force the planning response as an augmentation of the physical infrastructure (such as arterial tunnels). Squeezing temporal reserves by increasing the velocity of the system and achieving more direct transhipments requires a high degree of synchronisation which may lead as Rodrigue (1999) anticipates, to greater system instability. From a regular performance point of view, simply removing frictional constraints may lead to higher levels of just-in-time sequences, reduced unit truck loads and more truck trip activity (Roson and Soriani, 2000).
In this context, intermodal operations may be seen as providing significant flexibility and capability to the infrastructure skeleton. There may be direct substitution of road link utilisation in the line haul component. There can also be significant complementarity. That is, the intermodal terminal allows a combination, for instance, of rail and road haulage, which may reduce long haulage truck distances.
An aspect of temporal accessibility, also known as precision, is necessary for the effective operation of intermodal production systems. In turn, precision is a measure of the degree of synchronisation of transport carriers. The environmental gains come however if we refer back to the concept of temporal reserves and the nature of integrated demand for freight. Freight distribution systems, inclusive of intermodal terminals are both generators and consumers of freight services. This is particularly so in the bilateral nature of export and import flows. In order to look inside the freight task, we need to go beyond the outlook of traditional origin and destination nodes to how the gateways involved in freight translocation interact.
For instance, intermodal operations can provide inter-temporal storage capacity. This has the potential to buffer the lumpy supply arriving from the port and also to share container resources. This capacity can be manifested on several levels and is demonstrated by break-bulk activities. For instance, arriving import containers can be unstuffed, their contents stored and later re- consolidated in different forms for dispatch, whilst the container is prepared for export service on site and then stuffed.
Page | 54
Chapter 2: Research Design for Sustainable Freight System Formats
This can be seen as analogous to electrical power distribution whereby the distributed generation and storages operate close to loads such that total distributed generation remains less than the total load in its locality (Outhred et al., 2002) and thus allows peak loads (demands) to be met without endangering the capacity of the central generation facility. A facility that allows inter-temporal storage and has the ability to ramp up service generation according to demand (a combined generation-load function) can control the repositioning of supporting resources. By offering system deferment of consignments the facility may also control the amount of traffic on roads for a preference for rail. This is a form of dynamic resource assignment in the terminal.
2.7.6 Facilitating the Cost-Value Jump
Several authors reviewed, for instance Daganzo (1995), have explained how the production cost structure has mobilised inventories and created the growth of logistics activities, where transportation costs remain relatively marginal. Additional costs involved with transhipment activities often reduce the attractiveness and uptake of this mode of operation.
Terminet (1998) and Kreutzberger (1999) have defined the essential relationship in facilitating intermodal relations as the cost–value jump. This involves the arrangement of consolidation networks alongside transhipment networks. This has been the basis for further flow bundling models. Flow consolidation provides economies of density whereby complex bundling can provide either higher utilisation rates and or higher transport frequencies. Economies of scale can occur when a network allows larger transporting units (such as Hi-cubes or 40 ft containers) with a larger utilisation of exchange facilities. This reduces the cost per tonne of the consignment. Graphical analysis suggests that intermodal forms of transport only become cost- effective over distances greater than 350 km. Juxtaposed to this is Rodrigue’s argument that only a time compression will allow intermodal operations to be competitive against direct road haulage over any distance (Rodrigue, 1999). Woxenius and Sjostedt (2004) have recognised that a competence leap is required for the extension of intermodal services, and this relies on appropriate technology and information exchange.
It needs to be investigated whether this cost–value jump can be achieved in sequential exchanges rather than simultaneous exchanges. That is, how deferment in the system can reduce the scale requirements of the system by relying on improvements in system density. A means to reconcile the cost-value jump is to consider users’ objectives in terms of precision rather than velocity. Precision may be said to incorporate timeliness (synchronisation) as well as reliability and thus linking services need to be considered. Page | 55
Chapter 2: Research Design for Sustainable Freight System Formats
Transportation nodes can assist in the value recovery of wastes. In providing a necessary storage, streaming and densification function, transfer terminals can provide for the increase in resource recovery due to physical- logistics arrangements.
A key indicator here is the value density termed by McKinnon and Woodburn (1996). Operations that maximise this value density might be considered more sustainable. In freight container terms, operations that minimise the deployment of freight resources but otherwise increase the load matching ability of the network (through higher load factors, greater time utilisation of fleet) may be seen to build value density. An accounting system of terminal interactions that tracks value stream mapping (Marlow and Paixao, 2004) would aid in analyzing the benefits of certain functional deployment amongst terminals, a tactical planning decision.
2.8 Requirements of Intermodal Terminals
2.8.1 Physical Essence in a Freight Transportation System
The essential sustainability features that combined freight transportation can provide to an urban freight system need to be considered as well as economic feasibility. The broad intention, in the container freight task, is to better coordinate lumpy supply with diffuse demand. The effect is potential peaks which propagate through the system causing large infrastructure and resource demands, thus stretching capacity. This condition will grow with the forecast container flows through the seaport. A combined transportation system could provide two essential features to counteract this: the configuration of terminals of distributed function could provide deferment and accumulation functions which could foster better discrete-diffuse operations; secondly, these configurations may reduce the re-positioning of resources required. Mechanisms that test these features under different configurations could be utilised for more symbiotic infrastructure investment decisions, a concern that has already been raised by the Sea Freight Council of NSW (SFCNSW, 2005). That is, this infrastructure deployment may provide higher capabilities to the system as a whole.
Locational issues also become more complex when considering how to best embed intermodal operations in a seaport hinterland. UNCTAD (2004) identify that terminals may have a hierarchy of functions within the unit and between units. As such, the location decision of terminals as logistics centres becomes a condition of functional attributes as well as spatial
Page | 56
Chapter 2: Research Design for Sustainable Freight System Formats
features. Thus inland terminals, including intermodal terminals, can act as articulation points. It is these inter-functional relationships that may allow for greater synergies in operations and lead to lower system impacts for the given translocation effort.
Intermodal production systems are essentially retrofits of transportation systems. That is, they are constructed and developed from usually existing transportation infrastructure – rail and road links - and possibly existing terminals of varying functions. Compensating mechanisms need to be developed and applied to cater for this interface dynamic. Consequently there are a new range of performance measures to assess intermodal feasibility and how these system formats contribute to sustainability.
2.8.2 The Re-formulation of Logistical Friction
Logistical friction is a concept which describes the complex impedance characterising integrated freight demand i.e. the demand between consignors and consignees and the intermediate demand for freight resources that arise with freight forwarding third and fourth party logistics. Urban rail/road freight intermodalism is seen as a means to at least partially relieve the growth in container road freight traffic. There are many impediments to its full development in Sydney. There is ignorance in how intermodal terminals might best be configured (size, location, operations) for a systems-wide optimisation and what ancillary infrastructure is required. There is no specification about what the performance criteria might be. Intermodalism represents one distribution infrastructure type. An investigation into sea freight container intermodalism can provide an analytical basis into initiatives that reduce or exacerbate logistical friction and thus materialising tendencies of the freight industry. An investigation into the effects of retrofitting the existing network to achieve synergies and enlarged capability without large physical augmentation is warranted. Logistical friction involves a consideration of node impedance which is poorly covered in existing transport models. However this thesis claims that logistical friction is only a partial concept to address the obstacles to implementing an effective, sustainable intermodal production system. This is so that relieving the friction, through increasing the velocity of transhipment, may simply fuel integrated demand and freight intensity (kilometres/tonne).
Thus the research interest lies in also testing the complementarity of time and synchronisation elements which may enhance the capability of transportation infrastructure to match the loading ability of the distribution system. At the level of terminal design and interaction, impedance matching becomes more significant than reducing the perceived friction. The desired research Page | 57
Chapter 2: Research Design for Sustainable Freight System Formats
task is to ascertain what are the required impedance relations at the terminal to deliver the dispatch loads which allow for terminal and system feasibility.
2.8.3 Conditions for Financial Feasibility of Intermodal terminals
The information needs to develop intermodalism may be summarised into three sections. These are: the physical and logical connections required; the means to engender desirable behaviour; and the assessment of distributional impacts on existing networks and land-use activity. The Sea Freight Council of NSW (SFCNSW, 2004c p.6) further defines the financial sustainability criteria of these connections on the basis of considering terminals as places, entities and elements in chains. In essence, terminals must have proximity to substantial and suitable volumes and have access to road and rail infrastructure. Terminals must be fit for the purpose of allowing frequent service by rail and vehicle interoperability (physical access to different vehicle classifications). To be competitive, terminals must exist within efficient supply chains. Whereas measuring these mechanical connections is demanding enough, embedding intermodal terminals within existing supply chain networks, so that such terminals may divert flows, is even more challenging.
2.9 Performance Indicator Frameworks
2.9.1 Temporal Performance Analysis of Intermodal Retrofits
Tarski (1986) evaluates rich definitions of time and information pertaining to precision in freight transport. These are vital to discern as their interaction represents the performance, constraints and adaptivity of the freight infrastructure (physical and logical) given material flow and unit load demands. In order for our system objectives to be achieved, we need to analyse the complementarity of the parameters for the intermodal production system.
These indicators may be defined as measures of flow patterns as an interface with distribution infrastructure (impacts on the transportation cycle) and measures of performance of infrastructure resulting from a flow pattern (impacts on the production cycle). Temporal components of these cycles are explained in Table 3 and
Page | 58
Chapter 2: Research Design for Sustainable Freight System Formats
Table 4.
Table 3: Impact on Transportation Cycle Component Description and Relationship Speed Increase velocity of cycle, increase carrying capacity of system. Rhythm/ Continuity Degree of evenly distributed connections. Consolidation patterns to increase utilisation of unit in time and increase use of its capacity. Frequency Reduces storage with goods dispatched in smaller quantities. Implications for terminal work. Synchronisation Timeliness of flows with available resources – meeting particular dispatch services. Consistency Flows that are compatible in time are also complements in function – such as empty container flows to serve consolidation operations. Symmetry Flow balance. There is asymmetric flow between exports and imports affecting carrier loading rates. Complexity Multicommodity source requirements to meet dispatch services.
Table 4: Impact on Production Cycle Component Description and Relationship Speed Increase velocity of cycle, increase carrying capacity of system; injection of work required. Capability Depends on manner loads are exchanged in the transportation cycle. Transport units are utilised in time and with respect to their total capacity. Flow rates also impact on accessibility of storages. Interoperability How vehicle types fit with terminal infrastructure and technology with respect to transhipment. Reliability A function of frequency, rhythm, punctuality, regularity. Pertains to predictability of temporal accessibility and the ability to adhere to a given schedule. Is also dependent on the redundancy of the system. Resilience System flexibility under conditions of disturbance. Assess the terminal’s ability to respond to business rules Multipurpose nature Terminal multi-path coupling typology due to the types of flows of services required at the terminal
Tarksi presents a number of interacting factors that pertain to temporal aspects of the freight task. He highlights the concept of temporal accessibility, which may be considered the moment when transport is demanded and available. This is the degree of synchronisation. Tarksi notes that it is here where the temporal reserves of the transport network may be released. The role of distributed storage buffers are critical in effecting temporal accessibility Vehicles party to transhipment need to arrive not only in a timely fashion but with complementary functions. This Page | 59
Chapter 2: Research Design for Sustainable Freight System Formats
is significant when characterising the heterogeneity of freight dynamically when trucks may be constrained by the type of container consignment they can collect or deliver based on physical capacity, destination, and institutional arrangement. Precision involves synchronisation against a schedule.
Tarksi outlines the tension between regularity of service and dispositive elasticity (that is, movements of consignments according to shipper-carrier whim). Regularity is an umbrella term for reliability, frequency, rhythm and punctuality. It represents conditions fixed in time and space. Tarski (1986) has noted that “transport users only prefer regular connections if their aspects shorten storage time and offer higher profits than they would have derived from the greater dispositive elasticity inherent in irregular connections” (p.44).14 Thus, the introduction of a more physically regulated logistics-infrastructure system needs to meet a number of criteria in order for uptake.
A powerful description of time in distribution supply operations allows us a language to negotiate the paradoxes of green logistics (Rodrigue and Hesse, 2004). There is a vital need to inquire about the potential for logistics-infrastructure synergies that can contain logistical friction pressures, such as those pressures from undampened logistical drivers. This fuller description of time elements also equips the planner to define and measure infrastructure capability and the synchronisation of consolidation activities. The time elements described span and tie together the operational scale with longer-term scales. Thus the load matching ability of the distribution system may be seen as dependent on a development of infrastructure capability. This connection is portrayed in
Figure 12.
14 Dispostive elasticity of the road network may be considered perceived slack capacity to be exploited. Page | 60
Chapter 2: Research Design for Sustainable Freight System Formats
Figure 12: Integration of System Elements Relating Infrastructure Capability with Load Operations
There is a need to coordinate these parameters within the concept of impedance measurement. Figure 13 links the elements of specification of desired payload dispatch, inflow pattern and terminal productivity level with mechanisms of work and control and outputs. Further development of the impedance mechanism will include Figures of Merit to assess terminal unit process single and combined performance.
Impedance
Inflow pattern Payload Dispatch Specification Productivity level Stability Work Control -Consolidation -Routing through -Access inertia terminal Mechanisms and handling -Process storage -Flow matching utilisation
Delay Costs/Value Outputs Utilisation of resources Flexibility
Figure 13: Relating Engineering Impedance to Performance Indicators
Page | 61
Chapter 2: Research Design for Sustainable Freight System Formats
2.9.2 Hierarchy Indicators of Intermodal Capability
Developing Sustainability criteria is vital as it allows us to apply strategic and anticipatory research which links design and operations of the network under formulation (Nijkamp, Vleugel et al., 1993). Our central challenge is, to re-configure network infrastructure (retrofitting) in order to reconcile operating efficiency with longer term effectiveness. This is a significant research area in sustainability of interdependent systems, such as transportation (Beavis et al., 2009). The aim is to utilise the benefits of the technological envelope that the system format represents. The nature and level of integrated demand for freight services is moulded by the technological envelope, being the available network infrastructure, its transformation by technology, land use intensity and mix and logistics practices (such as inventory control).
The design method provides a hierarchy of indicators and figures of merit, which are considered significant in making the transition to more sustainable transport systems (Black et al., 2002, Doust and Black, 2009).
For this thesis the hierarchy considers the following interconnected layers:
1. Repositioning the Freight Task; 2. Terminal Interface Productivity; 3. Flexible capacity; and 4. Storage stability.
Repositioning the Freight Task represents a set of indicators that consider the system benefits of intermediation. These pertain to reducing the freight intensity of road operations (km/tonne). This ambition also addresses business activity harmonisation objectives to reduce the logistical friction among operations over a network. This would involve network coordination solutions15. Terminal Interface Productivity considers the key performance indicators and Figures of Merit to assess whether terminals are operating according to precision and controllability objectives. Impedance functions are developed which represent the work a terminal process must undertake to yield the desired dispatch. Flexible capacity assesses how well the terminal and its interface can meet dispatch schedules using the existing configuration and changes to it. This involves a consideration of the application of business rules through compensation control mechanisms to assess available transformations with the given terminal configuration. Finally, Storage stability
15 Coordination measures as future work are outlined in Appendix F: Theory Synthesis and Extension to Network Coordination. Page | 62
Chapter 2: Research Design for Sustainable Freight System Formats
provides design guidance at the unit process level to ensure that the terminal processes are stable for the given level of flows and interface technologies.
This thesis focuses on mechanisms for measuring terminal interface productivity and flexible capacity. Demonstration of the first and fourth elements in the hierarchy is left for future research.
Many scholars have contributed to defining a systems perspective for freight transportation and its effective utilisation. Wigan (1978) observed that the spatial and temporal arrangement of freight in the urban environment is more significant in terms of consumption of resources than the actual quantity shipped itself. Morlok and Riddle (1999) have attempted to address the measurement of the physical capacity of freight infrastructure given the heterogeneity of freight services. This includes a consideration of adaptivity of the network in transitory events (such as road closure). D’Este (1996) has been explicit for the need to carry multi-attribute information through any freight model (such as commodity type, vehicle type, time window priority) and the need to associate both discrete event and continuous dynamic processes in a network of transhipment nodes (terminals of various function are sources of discontinuity). Other authors have identified the physical and operational components critical in effective services delivery (Ballis and Golias, 2004, Kreutzberger, 2008). It is evident that in order to address the challenge, analytical tools in capability assessment need to be developed. This incorporates both physical capacity as well as effects on temporal and spatial qualities. An engineering a solution to the logistics-infrastructure antagonism may be to retrofit key freight nodes, such as intermodal terminals to existing links. Such initiatives do not currently have a basis in performance analysis.
Four categories are devised for the assessment of capability for intermodal production systems: Accessibility; Financial Feasibility; and Freight Intensity and Functional Discovery. These form the objectives in a hierarchy of attributes and indicators towards measuring intermodal balance capability (Figure 14).
Accessibility pertains to the attributes of precision and flexible capacity of terminals. Financial feasibility is the testing of the value proposition of a terminal in a specific network relationship. This involves the complexity of the flows necessary to support dispatch loads and justify operating viability. Freight Intensity considers the value flux metrics through the terminal. Functional discovery considers the opportunities for urban leverage that a distributed resource system such as satellite terminals brings. Urban leverage can minimise the evolution of system deficiencies (the infrastructure to support transportation flows) where intermediate terminals can
Page | 63
Chapter 2: Research Design for Sustainable Freight System Formats
siphon diffuse flows and control their complexity in bundleness (full container load, or less than full), destination, and dispatch timing.
Page | 64
Chapter 2: Research Design for Sustainable Freight System Formats
Figure 14: Hierarchy Indicator Diagram for Intermodal Balance Capability
Page | 65
Chapter 2: Research Design for Sustainable Freight System Formats
2.10 Concluding Comments for Conceptual Development
This chapter has explained the motivations and the attributes of the conceptual model for assessing the feasibility of intermodal production systems.
The analysis of the freight task as a large technological system has shown that a concerted effort is required with cross-cutting policy initiatives in order to achieve future trajectories of sustainability. Intermodal road-rail hubs have the potential to yield may benefits to address emerging gaps in capacity with a demand management response of consignment consolidation.
Whilst there are obvious opportunities in developing theoretical models of intermodal systems, there are also noted dangers in contributing to materialising freight task by simply accelerating the throughput of the network by reducing frictions or by multimodal duplications. The guiding elements in modelling the types of intermodal production systems identified in this chapter are the performance indicators for sustainability for a combined freight container transportation system. System performance is linked to the performance of terminal operations by value flux assessment which monitors the changing freight intensity with the introduction of buffer storages and transhipment opportunities.
Page | 66
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
3 Gap Analysis in Freight Modelling for Intermodal Outcomes
3.1 Introduction
3.1.1 Chapter Guide
This chapter assesses the tools available to address the research design requirement. Research design drivers are outlined as objectives that define the area of study. A conceptual model of the urban freight container task is first developed. This acts as a reference to position where and how models formulate the freight metabolism and gives insight into their conceptual validity. A review of intermodal policy is made to highlight tensions and opportunities in making a transition to a new logistics structure. Research ambitions in linking terminal operations with rail operating forms are explained. Network methods are then reviewed for their applicability to intermodal terminal physical and service design. The insights from facilities management models are then reviewed. Gaps are identified in prevailing freight models that address tactical level problems. A critique of network models is made when considering the dynamics of consolidation networks. The benefits of a terminal design model formulation leading to coordination methods are anticipated. A synthesis of the gaps and opportunities are made in discussing worthwhile paths of investigation in mapping intermodal information needs to mathematical models. This leads to chapter 4 where a functional specification is presented.
This chapter reviews the full spectrum of freight models in order to define the essence of the transhipment problem and how it has been addressed by the modelling hierarchy of strategic, tactical and operational levels. The research drivers define the essence of the transhipment problem according to this thesis which extends the definition to also consider consignment transformations, such as bundling, and terminal functions such as storage. The initial focus is on network models with a critique of their demand steering objectives. Facilities dynamics models are reviewed for their insights in terminal design. Intermodal models are reviewed including their archetypes. Whilst they are instructive for specific purposes for the time horizon and stakeholder to whom they are directed, there remains no suitable intermodal transportation science that combines physical network and service network forms. This review of the freight modelling literature allows the reader to see the outstanding gaps in the Multi-Modal Multi-Criteria problem. The solution to this must take a network perspective whilst performing an authentic, albeit sparse terminal characterisation. Page | 67
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
3.1.2 Research Design Drivers Which Limit the Study
To achieve combined transportation networks, integration mechanisms between network flows and terminal operations are required. A critical mechanism is precision. The need to devise precision parameters of the terminal interface is based on five attributes which characterise the current urban freight task and transition management efforts to improve its productivity: 1) The current system is one of excess capacity, creating a practice based on dispositive elasticity (Tarski, 1986) where the network is treated largely as a free good and users engage in activity which is sub-optimal in a systems perspective causing acute transient bottlenecks. Injecting more capacity in the system may only fuel this dispositive elasticity. Rodrigue et al., (2001) have documented the spurious claims of green logistics once this underlying dynamic is understood. One measure of the success of controlling this spendthrift activity is measuring the freight task as kilometre/tonne (km/t) rather than the ubiquitous tonne-kilometres (tkm) (Roth, 2000). Control of freight complexity through terminal operations may be known as minimising the evolution of system deficiencies; 2) The limits on capacity at a terminal are factored by the headway frequency between carrier arrivals(Morlok, 1978a). There is a need to cope with even more pronounced peakiness to address increasing peaky shipments through the seaport. This will affect practical capacity at satellite terminals. It is at these satellite terminals where the consolidation, transhipment and block movement activity required to facilitate train shuttles between the terminal and port takes place. This driver may be known as developing and measuring hinterland absorptive capacity; 3) There needs to be a measure of logistical friction (impedance) which occurs with novel freight system formats as a comparison to achieve more optimal forms. This driver is known as interface precision; 4) Engendering the utilisation of intermodal facilities to consolidate and tranship, particularly over short haul distances (less than 40km), will require timely and synchronised processes between modes, intermediated by storage buffers. This is a condition of increasingly tense flux (Rodrigue, 1999), where flows are dependent on effective connections. This may lead to node and system instability with shock loads. This driver is known as controllability; and 5) For an intermodal production system to be feasible it must demonstrate a value proposition. The impact on value and costs with changing functions in intermodal terminals must be assessed to discern appropriate terminal layout and function and to communicate preferred options to stakeholders.
Page | 68
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
Without adequate theory and systemic design tools, intermodal planning is uncoordinated and may be currently seen as pre-paradigmatic (Bontekoning et al., 2004). One outcome of this situation is piecemeal modelling of system components and planning inertia. The following review considers limits and opportunities with current modelling paradigms in both network and facilities modelling.
3.1.3 Addressing the Information Needs of Infrastructure Suppliers
There is a dearth of analytical strategic frameworks for freight transport planning from which infrastructure providers may assess information needs from a suite of modelling methods. Exceptionally, some methods have been outlined briefly for a port authority in landside logistical analysis (Cartwright et al., 2003). Techniques in forecasting passenger transport needs are well established, but in government practice, freight planning uses the more traditional models in the transport planning toolbox, such as a synthetic approach with forecasting based on extrapolations of cross-sectional data. Nijkamp and Reichman (1987 p.314) have observed that these techniques have two limitations; ignorance of behavioural drivers, and an orientation towards past experience and not towards future mobility. These deterministic approaches can over-parameterise models leading to a model identification which builds in major theoretical errors and limits their fidelity and tractability for novel network forms.
Kanafani and Sperling (1982, Fig.1, p,6) developed the blueprint for a national transportation plan. This showed the necessary interaction of components supporting Financial, Supply and Demand Analysis. The interacting components include the identification of operational and capital improvements, calculating zonal activity growth, future modal options and anticipating the evolution of transportation deficiencies. The decision support context must be discerned for the appropriate deployment of models.
3.2 Conceptual Model
The value of a conceptual model is that it lies above any metabolism that depicts the freight task at hand. It assists in scrutinising the conceptual validity of modelling practice. Establishing this conceptual validity is of vital significance. Modelling outcomes will have much greater differences in guiding decision making in achievement of the delivery task depending on whether their focus is Page | 69
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
on maximising swift goods movement (mobility) or on goods transformation (consolidating to improve accessibility).
A conceptual model abstraction then is essential to understand the nature of the freight task and the changing driving forces that should be incorporated into urban freight modelling studies for either predictive or explanatory intent. Changing driving forces and the complexity of their composition make freight much less predictable than passenger traffic. Driving forces can be attributed to the nature of the multi-stakeholders involved in freight and logistics, the intertwining involvement of public and private sectors, and the parallel processes of physical movement of goods and the accompanying information flows in the logistical chain. A generalised freight conceptual model was proposed by Rimmer and Black (1981) , but this now requires considerable supplementation given the great changes in supply chain management and physical distribution systems. A conceptual model also allows planners to understand the scope in space, time, stakeholders, and nature of the task that any one model addresses or neglects.
Figure 15 represents the supply of transport infrastructure and services as a network of links and nodes A variety of freight flows underpin the conceptual model and invite an analysis of possible interactions amongst these flows: external flows to the urban system – through-transit flows with either a domestic or a global destination (through the ports without urban transhipment) or direct flows with urban destinations (with or without transhipment) or transit flows with urban transhipment – and internal or intra-urban flows with linked trip patterns or single destinations. The trip made on the network can be described in terms of a tour which consists of two arcs (directional links). Additionally, freight traffic has distinct diurnal peaks, in parts of the network often controlled by a formal or informal curfew in order to provide service to passenger transport.
Page | 70
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
Figure 15: Freight Conceptual Model (Rimmer and Black, 1981, Fig.1, p.16)
Land-use nodes represent freight generators or attractors. However, freight infrastructure may also become a component with repercussions on freight demand because consignment translocation must navigate through the network infrastructure. Roson (2000) has noted that intermodal terminals could invoke Just-in-Time logistics practices by reducing transport impedance. Additionally, transhipment (via terminals) may lead to increasing consolidation of flows (discretion), later to be distributed (diffusion). The freight trip is measured not only by the links taken but also by the nature of node visitations it makes: how the consignment characteristics change at each node.
Logistics trends are only faintly outlined in this conceptual model. A shipper has a number of options and conditions for delivery. As a manufacturer involved in process and assembly, the shipper can also reconfigure his/her supply chain. Supply chain logistics trends can affect the attributes of the freight task. For instance, Just-In-Time or Time Compression Principles make inventory holdings more mobile and increase freight traffic (Zografos and Giannouli, 2001). Logistics initiatives have reduced the loading factor of freight vehicles and increased the average distance travelled. Logistics dynamics contribute to freight growth more than the economic (linear) requirement for the physical movement of goods. The introduction of intelligent transportation systems (ITS) is possibly seen as a means to radically improve system performance - for example,
Page | 71
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
interoperability (Giannopoulos, 2002), and obviate the need for new physical infrastructure. Technology that provides close to perfect information, may induce widespread trip chaining and reduce the large proportion of freight vehicular traffic carrying empty loads.
3.3 System Deficiencies and Policy Planning Responses
3.3.1 Strategic Planning and Investment for Urban Landside Container Freight Task
In the United States the introduction of the Intermodal Surface Transportation Efficiency Act (1991) heralded new opportunities in connecting modes for improving mobility of freight and passengers (TCRP, 1993). Reviews since this introduction have cited restraining factors in successful implementation (GAO, 2007): 1. Limited Federal funding for infrastructure works; 2. Limited Collaboration between government institutions as governance enabling bodies; and 3. Limited ability to evaluate the benefits of intermodal projects
These barriers have led to planning inertia for intermodal networks and continued entrenched modal separation planning. In some large American cities, such as Chicago with the program CREATE, there have been concerted infrastructure programs to improve passenger and freight rail efficiencies through grade separations. Neither the literature generally nor such examples discuss governance issues in harmonising freight activity in the existing logistics structure. Many of the American intermodal initiatives are for inter-state and inter-city transit. European intermodal research initiatives were surveyed in Chapter 2. They are predominantly based on the operating concept of long line-haul distances and grant limited insight into the tensions that need to be reconciled for urban intermodal forms. Institutions in both American and European jurisdictions are thus intent in driving intermodal solutions to improve mobility and do not seem to identify any conflicts with this and reducing truck trips.
Page | 72
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
In Australia peer industry bodies and government have been diligent and comprehensive in identifying the so-called causes of degrading seaport-hinterland productivity and the impediments to harmonisation of the landside container task (SFCNSW, 2005, SFCNSW, 2004a, VFLC, 2005). These include the mis-match of operating hours, the conflicts of freight truck movements over the passenger vehicle network, and the mis-match in empty container movements.
The problem with these diagnoses is they are symptomatic rather than causal. There are essential (re)design requirements that arise from these harmonisation objectives. For instance, rarely is the lack of terminal consolidation and transhipment facilities given as a cause nor is the need highlighted to address pernicious logistical structures such as the dispersal of warehousing functions from the seaport precinct and the increasing length and number of truck trips. The drive to reduce logistical frictions such as road congestion leads to a focus on mobility and speed. This “removes all freedom to do anything but commute” (Hägerstrand, 1987 p.18). The dissipative logistical structure of seaport-hinterland relations are highlighted by transient storages of containers in the truck network overnight due to inaccessibility of the seaport. The symptom of combined poor TEU slot utilisation at the seaport with poor truck utilisation rates leads to low accessibility and will improve only marginally with the introduction of larger load bearing trucks (VFLC, 2008). The suggestion to address lack of harmonisation by 24/7 operations may relieve peak capacity constraints such as access to the seaport gate, however they do not necessarily encourage high load factors and more sustainable sequences of discrete/ diffuse flows.
Recently policy and strategic frameworks for intermodal networks in Melbourne and Sydney have developed considerably in identifying the guiding principles for strategic implementation to support the urban landside freight task (FIAB, 2005, SFCNSW, 2007, DoT, 2008b, DoT, 2008a). These guiding principles include: 1. A logistics systems approach that crosses modes; 2. An alignment of institutional and commercial relationships with physical activities; 3. Concerted action to consolidate base load demand and offset investment costs; 4. Proper governance to deliver inter-organisational and inter-commercial arrangements.
The strategic implementation of these guiding principles will involve (DoT, 2009): 1. Fostering the intermodal value proposition: a cost accounting of container transportation costs to stakeholders and the system; 2. Managing investment with incremental development;
Page | 73
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
3. Providing minimum infrastructure of dedicated linked terminal pairs to ensure availability of service route. Given freight moves over the passenger mainline, there must be sufficient mainline capacity to deliver sufficient paths; 4. Ensuring suitable terminal arrangements which are conducive to an efficient and effective logistics supply chain. This involves the terminal arrangements so that road- rail chains match; 5. Providing on-going network investment.
Strategy 2 involves the building up of base loads on rail with concurrent truck services. Thus the integrated nature of the recent Victorian Freight Futures strategy (DoT, 2008a) identifies the need to foster container road hubs in the vicinity of rail intermodal terminals. Ideally these road hubs will be co-located with the road-rail intermodal terminal becoming bimodal activity centres. It is perceived that intermodal start-up will be enhanced by the road based activity. Private investors would also see an opportunity to cross-subsidise the operating costs of rail services with road services. Whilst intuitively attractive, there remains the concern that this may lead to a multi-modal solution rather than an intermodal or bi-modal solution. That is, truck hubs are preferred over road- rail intermodal services.
Whilst, much of the strategic implementation is sound for the jurisdictions of Melbourne and Sydney, the implementation plans remain untested. These would cover the governance and investment interactions anticipated in Chapter 2, Figure 5. A necessary, though only partly sufficient, question to be addressed from the strategic principles can be phrased as follows:
In order to reduce the impact of the landside container freight task, how can the freight system of physical infrastructure and logistical structure deliver the payloads necessary to support freight container train shuttle services and thus justify urban road-rail intermodal operations?
3.3.2 Land Use Planning for Seaport Functions
Port authorities in both Sydney and Melbourne have until recently paid scant attention to hinterland constraints on seaport productivity. Port development plans (PoM, 2006, SPC, 2002) cite land-use pressures in their precinct for container storage and vessel visitations yet propose vast engineering works to dredge the bay (PoM, 2008), augment wharf capacity to increase the number of vessel Page | 74
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
berths, and recommend further acquisition of adjacent land for container related functions in order to maintain and increase seaport productivity (SPC, 2002) . There are significant landside congestion pressures with large peak bottlenecks on road access. For instance in Melbourne, only 51% of available TEU slots are utilised using truck carriers (VFLC, 2008). Preliminary modelling studies undertaken for Melbourne identify that with the threefold growth in container traffic expected to the Port of Melbourne, Melbourne’s East-West road routes accessing the Port precinct will become extremely constrained throughout an increasing peak period (DoT, 2008c). Alternate ports development to augment the Port of Melbourne, such as the Port of Hastings, are being staged to occur as an overflow from Melbourne once it reaches its “seaport capacity” of 8 million TEU/year projected by 2025. This capacity does not include the capability of the hinterland.
The practice of intermodalism may be seen as redefining port nodes and landside networks and thus affecting the equilibrium of transportation. The port as a gateway node siphons transit and urban originating and destination freight. Seaport capacity then must also be measured in terms of its hinterland relationships and interacting infrastructure. The relationship between port related and local flows remains poorly understood in planning and design practice, yet port productivity is now recognised to encompass landside constraints (Marlow and Paixao, 2004, Kia et al., 2002). Notteboom (2008) observes that ‘containerisation and intermodalism have strengthened the symbiotic relationship between foreland and hinterland in the sense that a true foreland-hinterland continuum has come into existence.
Global trade exerts pressures on the national freight landside condition. Servicing infrastructure and links in China are not sufficient to handle the variety of container throughput (Wang, 2002). Empty container flow management is a significant portside and landside problem at the ports of Los Angeles and Long Beach in Southern California (Hanh, 2003, 2007). To alleviate the pressure on truck access routes, greater interaction between freight infrastructure nodes is required. Australian seaport-landside linkages also need further integration and this must be negotiated through existing heavily built up urban areas.
The seaport dilemma in Australian cities is analogous to airport planning in cities. Jatmika (2004) calculated that there was insufficient land in the immediate hinterland for ancillary services including freight warehousing to support the planned growth of Sydney airport. Slack (1999) identified the benefits that a network of satellite terminals could deliver to a seaport as a gateway node included extending seaport holding capacity and relieving access constraints. Rimmer and Tsipouras (1978) identified the conflicts inherent in Australian ports and urban areas and the roles Page | 75
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
that different actors had. There has been very little policy and strategy work before 2005 on embedding hinterland intermodal operations into seaport productivity. As strategies are now being released ((DoT, 2008a, FIAB, 2005)) it is clear that much work remains in conjoining strategy with investment, governance and practice in delivering intermodal freight solutions in an urban environment.
3.4 Interface Attributes Specific to Urban Intermodalism
3.4.1 Hinterland Bundling and Consolidation: Land Use Analysis
A useful approach to transhipment analysis within the context of a feasible intermodal production system is one that considers the prospects for gateway interaction as a focus for consolidation networks. Intermodalism has a number of physical, engineering, and logistical requirements. For instance it is dependent on the availability of sufficient loads to develop consolidation networks. Bundlepoint optimisation studies, (Beuthe and Kreutzberger, 2001, Kreutzberger, 1999, Southworth and Peterson, 2000, Schmeddling et al., 2004, Jourquin et al., 1999) in part identify where these loads can be developed based on the current land use intensities. These in turn suggest where transhipment terminals can be located and how they should be interfaced i.e. their role in a hub arrangement. For instance, Notteboom (2008) observes that there are three highly interrelated decisions required by service planners: the service frequency, the loading capacity, and the itinerary taken in visiting intermediate terminals prior to the final destination. This is especially necessary when considering advanced forms of intermodal hub configurations and liner operating forms where there are several intermediate stops at intermodal terminals between beginning and end locations. The volatility of demand is a crucial aspect to be estimated and even on assured routes, can be highly seasonal.
3.4.2 Benefits of Short Haul
There are benefits for short haul urban intermodalism. Howard (1983) was an early proponent in Great Britain who questioned the efficacy of large economies of scale and noted that small intermodal terminals had scale benefits and had the additional benefit of capturing local
Page | 76
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
consignments and fostering dense mesh bundling networks. With simple, suitable transhipment technology, terminals can facilitate considerable throughput. Novel train operating forms prototyped such as the Swedish Cargo Sprinter, can prove economic with networks where consolidation loads are disparate and small so long as terminal operating costs are low and the train operating form such as a liner train. This can be assisted with mobile transhipment equipment (Woxenius, 1998, 2007). Roso et al., (2009) have argued that intermodalism should be seen in terms of close, medium and far location. This echoes Priemus and Konings (2001) who posit the development of a seaport hinterland using intermodal satellite terminals. Where more that 80% of container beginning and end trips take place within 40km of the seaport, as in the cities of Melbourne and Sydney, the feasibility of short-haul intermodal transport becomes a pressing research and public policy issue.
3.4.3 Performance Indicators
A number of terminal resource productivity measures have been developed to assess terminal operating performance. These include:
• Unit costs of container production ($/container throughput); • Throughput productivity (containers/day); • Area productivity (container throughput/area); • Equipment throughput (containers/crane/hour); • Lifting ratios (number of lifts required/ crane);
The ratios of terminal operations described above do not tie into network productivity or measures of network space utilisation. In facilities design and management (Thompkins, 1996), there is little extension of the physical and service design of the terminal with the complementary operations of other terminals in the network of interest. Problems with the separation of terminal design from network design include the ad hoc deployment of distribution infrastructure in size, time and function leading to high freight road intensity as these terminals are not design explicitly for a complementary logistical structure (Slack, 1999). Analytical expressions of terminal dynamics and relevant network forms have some precedence in Continuous Approximation Models as expounded by Daganzo (1995) and Langevin, Mbaraga et al., (1996). A persistent research gap is to summarise
Page | 77
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
the dynamics of design parameters into steady state analysis to lead to effective mechanisms for coordination of these terminals in networks.
Indicators of freight logistics may be seen in terms of utilisation, productivity, and effectiveness. Examples of the first may be vehicle fill, empty running, and vehicle time utilisation; of the second: fuel consumption; of the third: deviation from schedule. The deviation indicator is significant as any deviation often reduces back-loading opportunities and thus undermines capacity utilisation (Mckinnon, 2007). There are few indicator hierarchies that extend port performance to include intermodal satellite hinterland connectivity. Styhre has posited 5 areas of indicator research required in order to lead towards improved port performance in intermodal transportation (Styhre, 2005):
1. Transit time; 2. Frequency; 3. Reliability (punctuality and precision); 4. Information management; and 5. Agility
Styhre (2005) and Rodrigue (1999) point to the need to reconcile effectiveness goals with efficiency goals in intermodal transportation. Effectiveness goals can lead to overcapacity, whilst efficiency goals can lead to a lean system which is vulnerable to shock loads and has reduced agility (ability to adapt).
Precision of service is a central issue in positioning intermodal production systems as timing between system resources affects reliability and the deviation from schedules (Woxenius, 2006, Tarski, 1986 ). The time composition of freight in distribution networks has been evaluated generally in Continuous Approximation Models by generalising the synchronisation of headway arrivals and storage patterns (Daganzo, 1990, 1995, De Castilho and Daganzo, 1993). However detailed composition calculations that consider the transfer technology, terminal functions and layout patterns assessed by analytical means remain unresearched. Facilities design and management for intermodal terminals as separate entities has been a recent focus of research using discrete dynamics (Kozan, 2006). Previous work in queuing theory has also focused on limited analytical expressions for terminal operations such as estimating delay time in train yards due to shunting operations (Crainic, 1987, Petersen, 1977a, 1977b, Turnquist and Daskin, 1982) or Page | 78
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
computing the mechanism of the bulk service queue to control vehicle dispatch (Powell and Humblet, 1986). Freight tactical planning research needs closed form solutions of terminal operations which cater for different situations with greater precision (Crainic, 1987). Critically, timing information in itinerary design to route flows through terminal operations is also not well articulated.
Specific benefits of an analytical approach to the time composition of freight include:
1. Testing the feasibility of introducing a fine- meshed rail-road intermodal system of short distance, small consignment freight into an existing all road operation; and 2. Assessing the feasibility of intermodal production systems under different flow regimes.
3.5 Network Freight Modelling Taxonomy Toolbox
3.5.1 Network Model Archetypes
The freight model taxonomy presented here has the initial purpose of classifying freight models. Kanafani and Sperling (1982) also noted that along with the first order tasks in a transportation planning framework, there is considerable feedback among the tasks. This taxonomy anticipates that information needs for strategic urban freight planning will cross several model types. This interlinked taxonomy becomes the basis for constructing a systematic modelling approach.
Research that informed this taxonomy was organised into four areas. Conceptual transportation models and the theory of logistics acted as a foundation for the research. Papers summarising specific modelling fields formed the research area Model Archetypes and were significant in composing the model taxonomy. Strategic Freight Studies were reviewed with an emphasis on the experience for Sydney. Freight Models and Modelling Methods were reviewed with a focus on urban models (or how interregional models grant insights into urban activities). This approach is equally applicable to other cities. In all, more than 150 freight modelling, modelling method papers, and strategic studies contributed to this taxonomy. Freight Models have been categorised as Forecasting, Travel Demand Management, Integrated Urban, and Logistical (Figure 16).
Page | 79
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
Modelling approaches could also be broadly categorised as strategic freight network models (Friesz, 1985) or urban goods movement models (D'Este, 2001), dealing with supply chain and distribution logistics (SCL), although both approaches are closely related with regard to their interest in forecasting and analysis of supply capacities. The taxonomy combines both these interests. The starting point might be forecasting approaches. Freight forecasting models can be generated by a number of methods. These include aggregate approaches of general equilibrium (Friesz, 1985, Oppenheim, 1993), time series (Garrido and Mahmassani, 1998), commodity based (Rockcliffe et al., 1999), zonal truck trip (Black, 1977), or extrapolations (Cambridge Systematics, 1996).
Supply models, central to this thesis, can be considered at several scales. As a constraint on forecasting growth on network links, simple supply issues can be level of service (LOS) generation as inputs to network assignment models (Black, 1981), Physical network design models consider new and augmented feasible networks and terminal locations (Crainic, 2000). System capacity, performance and infrastructure quality models use optimisation techniques to assess likely system bottlenecks (Morlok and Riddle, 1999) or the requirements of system maintenance (Kalaitzidakis and Kalyvitis, 2002).
Disaggregate models consider the behaviour of shipper – carriers in affecting decisions in the four step forecasting sequence (Roberts and Kullman, 1979). Agent based models, such as neural networks, can run autonomously of this structure (Nijkamp et al., 2004). Activity models consider the generation of freight demand from activity nodes and are better placed to consider integrated demand from different distribution system configurations (Boerkamps and Van Binsbergen, 1999). Disaggregate approaches have also been developed for acquisition models (urban trip-chaining) for light goods vehicle movements (Russo and Comi, 2004). It is envisaged that infrastructure suppliers and operators need to consider the output of such models to understand changing vehicle flows and classifications.
Travel Demand Management models consider rationing demand across the network. These can be vehicle routing and scheduling models (Thompson and Van Duin, 1999) infrastructure pricing models (Madsen and Jensen-Butler, 2004), or more sketch modelling types, which may also include strategic decision support simulation (Tavasszy et al., 1998). Vehicle Routing and Scheduling models consider the deployment of a vehicle fleet. Infrastructure pricing can be used to control congestion over parts of the network. Pricing studies are particularly apposite in Australia with the increase in toll roads. Page | 80
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
Integrated urban modelling involves the integration of transport and land use, forecasting and logistics, transport and the economy, and passenger travel and freight. Models may consider that transportation drives land use or land use drives transportation growth. Noticeable developments in transport and land-use have been in micro-simulation (Parsons Brinckerhoff, 2001). Traditional models generally consider the economy as exogenous – the economy drives the transportation task (traffic models, activity based models). In transport and the economy, transportation influences the economy (location models and general equilibrium models) and transportation of goods interacts with the economy (structural input-output models) (Van der Vooren, 2004). Infrastructure providers and operators need to consider how the evolution of certain land-use activities and intensities drive changes in freight growth.
Logistics models can be considered optimisation and/or simulation approaches. Areas of modelling include Vehicle Routing and Scheduling, Service network design (terminal function and simulation (Crainic, 2000), Distribution models (Langevin and Mbaraga, 1996), mobility demand and tour simulation (Russo and Cateni, 2004). Models which attempt to abstract the reality of supply chain and distribution logistics are of great interest to infrastructure providers and operators as they depict assigned freight flows based on the demand generated by the transportation system, and not simply the demand for the commodity.
From this review of urban freight modelling, a number of gaps can be identified. Primarily, there are fewer modelling approaches for urban environments; they are often directed toward regional plans. System conflicts with passenger and freight, multi-modal (transhipment), multi-commodity models and land use activity generation models currently appear only theoretically and are not fully operationalised into models. Consequently, the taxonomy will not satisfy all information needs of planners.
The taxonomy presented in Figure 16 may be used by freight transport planners selectively. Planners can see the spectrum of model types available in the context of their purpose. It requires further definition of planning needs, and, inevitably, knowledge of data availability, to select the model and modelling methods of value. Each set of information needs will assist in guiding the planner in tracing what modelling-by-purpose category is of value, and, within the category, the likely methods. It is envisaged that planners may combine a number of modelling approaches toward an overall tailored functional specification. For instance, from general commodity forecasts in the region (forecasting), planners may wish to investigate the additional freight movements due to different distribution network infrastructure (logistics) and land-use activities (integrated urban).
Page | 81
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
FORECASTING (static node, unimodal, single actor) Regional/ Urban Network Scale Travel Demand LOGISTICAL (intelligent node, multimodal, multi-actor) Network/Sub-network, macro/ microscopic scale Management AGGREGATE NETWORK ACTIVITY DISAGGREGATE Regional Urban (Micro- Simulation) Control/ Decision VRS Support/ Logit models for Sketch Simulation Optimisation Modal Split & SUPPLY Assignment General DEMAND Continuous Event Equilibrium Infrastructure Nested Logit Operations Extrapolation Models for Supply Pricing Vehicle Routing Physical Network Chain & Network Activity Agent Systems Design Distribution (VRS) Ti me Structural Economics Commodity Series (Input-output or make Level of Service Neural Network Distribution (Cts use tabulations) Design INTEGRATED URBAN Approx. Models) Disaggregate Multimodal Network/ Whole of Network Truc k Tri p Tour Simulation Service Network Design Integrated Performance Simulation Land Use Terminal Management/ Direct Estimation Augmented 4-Step Spatial Geometry Cluster Terminal Location & Acquisition Methods Supply Techniques Capacity Generation/Attraction Estimation of Light Goods Distribution Vehicles (LGV) Cooperation in Combined Logistics/ distribution/ SCM Modal Split Forecasting Freight Generator/ VEHICLE LOADING Transport & Economy Environmental Attractor Location Iteration Impact Assignment Simulation Travel & Freight Network Forecast
Figure 16: Freight Network Taxonomy (Beavis et al., 2005)
Page | 82
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
The freight network model taxonomy in Figure 16 represents the breadth of the freight modelling toolbox available to transport professionals. The indicators of interest discussed in s.2.9, dealing with port-hinterland productivity based on combined transportation forms, cannot be easily supported or measured by these methods. A key reason for this deficiency is that there is little research in activity based methods which could lead to demand management initiatives or improved coordination of the use of supply. The bulk of methods are concerned with either forecasting network growth or partial system optimisation. They are thus highly empirical. Analytical relationships are not a focus of development and thus exploratory ways of learning are not so readily supported. The next two sections deal with existing approaches to the network transhipment problem and the facilities management problem. From this it may be identified what are the knowledge gaps which need to be bridged to assist the concept of urban freight intermodalism evolving from a conceptual paradigm to a discipline through both analytical procedures and network coordination procedures.
3.5.2 Freight Modelling Methods in the Transhipment Problem
3.5.2.1 The Four Step Network Model
Approaches in network theory in addressing freight have the following frames of reference: • Demand matching is the goal; • Flow unit is vehicle flux; • Minimise the time impedance of the system o Impedance is seen as a cost to minimise; o Link impedances are the most significant; • Allocate over the network, based on equal travel time; • Forecasting begets engineering response of augmentation of infrastructure; • Location of facilities usually based on equal travel time approaches.
Blunden (1971) describes the equations of state that govern the Hitchcockian formulation for a two route, two zone land use transport interaction mechanism using a gravity model. Land Use,
Page | 83
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
has a potential, L, to attract traffic; separated by travel time impedance, t. This generates the
number of trips between origins and destinations Tod. LL = od Tkod n …………………..(3.1) tod
1. Transport Impedance Equations
Land Use –Transport Incidence Matrix A(l,m,n)
= tAlmntod ∑ (, , )l ……………..(3.2) (n=tod) l
11− ()− jy tdt= 0 ll…………..(3.3) lll − ()1 yl where: j= level of service factor y= saturation level d= length of link t0 = free speed travel time
= 1 yl ∑∑ Almnqmn(, , ) ( , )……..(3.4) sl nm
sl = saturation flow of link l
q(m,n) = component of the desire line (od= n) which takes the mth path between o and d.
2. Transport Assignment Equation over the Network (equilibrium travel time ( minimise total costs)) = Tqmnn ∑ (,)…………………….(3.5) m == ∑ Al(,,1212 nt )l ∑ Al (, , nt )l .... n , …..(3.6) l l
Page | 84
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
These equations distill to the 4-step model: • Generation; • Distribution (equation set 1 and 2); • (Modal Split);and • Assignment (equation set 3).
3.5.2.2 Incorporating Transhipment
The classic transhipment problem (Blunden and Black.J.A., 1984) follows from the 4-step model and describes shipments that can be made via intermediate “Origins” and “Destinations”. This can consider capacitated and uncapacitated nodes.
• Combines desire line- distribution and assignment phases of classic modelling approach, • Variables are the traffic flows on links (not the total demand), • Volumetric balance around each node.
N N
∑∑ tqij ij ……… (3.7) i j== 11
3.5.2.3 Multi-Modal, Multi-Commodity Approaches
At the strategic level of design, the objective of prediction of flows is paramount. Shipper and carrier actors are not differentiated at this level. However it allows for detailed representation of the transportation infrastructure, facilities and services. Crainic and Laporte (1997) identify that this level can capture the competition of products for the service capacity available. We may link products, link flows and transfer flows according to the following objective cost function (Crainic and Laporte, 1997):
=+ρ ρρρ FSvvSvv∑ (()∑ a at∑ ())t …………… (3.8) ρ∈∈P aA∈ tT
where: Page | 85
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
ρρ Svvaa() is the total cost of the flow on arc a from product p
This covers flows on links, a, and flows over transfers, t. A path is a feasible arc for that mode. The constraint represents the total flow on all paths h, of itineraries, k equals the origin- destination mode-product matrix:
= mp() ∑ hgK od ………………………………… (3.9) ∈ mp() kKod
Loureiro and Ralston (1994) developed an approach to consider optimisation of investments along each arc with the location of nodes (including intermodal – bi and tri modal nodes). The demand for transportation services is assumed to be fixed and exogenous to the model. The source and destination commodity shipment matrices do not vary as a result of selected network improvements (p.40).
3.5.2.4 Intermodal Service Design
Service design objectives are generally addressed in tactical level models. These often demonstrate trade-offs in minimising costs and maximising service standards. These models can include decomposition of the optimisation function to ascertain the trade-offs (such as between costs of improving operations and costs of achieving promised service quality). This is known as post- optimisation analyses (Crainic and Laporte, 1997).
The function consists of the elements of traffic flow due to the specific traffic class, m in itinerary k, and the frequency, F of specific services, s (Crainic and Laporte, 1997):
ϕ m [,]XFk s ………………………………………. (3.10)
This is subject to the continuity constraint of flows equally total demand for each traffic class: Page | 86
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
m = m ∑ Xdk ……………………………………..(3.11) k
A problem with this model depiction is that the service characteristics and the transformations which depict the itinerary interact – this is especially so when peak (discrete) flows hit the terminal and the terminal response - depends on its ability to compensate for this pulse disturbance. A feedback control problem occurs, for instance with import containers if the service characteristics are set, the terminal must cope with the disturbance according to the output load profile required.
The Deferred Item Vehicle Routing Problem (DIVRP) is an extension of tactical transportation planning using multi-commodity constituents. For instance, Smilowitz et al. (2003) consider deferring less urgent items from the road network to excess capacity available on aeroplanes. As such, they consider an integrated, multi-modal approach. The additional complexity of this problem formulation type involves routing deferred and through items “consistent with delivery time windows and availability of excess capacity” (p. 307). The time-space network is defined as a set of nodes and arcs. The node represents both a physical location and a discrete instant in time; the types are consolidation, break-bulk, and air hub nodes. The arcs, linking the nodes, represent inventory (holding time), ground movement, and express air movement. The function to minimise by linear programming methods, is the sum of vehicle and item transportation costs and investment holding costs of commodities, k sent over all arcs, a:
k + i min ∑ bfa a ∑ cxa a …………………….. (3.12) kK∈ aGA∈ aA∈ iV∈
There are a number of equality and inequality constraints that make this a complex procedure. The commodity flow based equation is:
⎧ −=dnkk, Ο ⎪ k −=k ⎨ kk= ∑∑ffmn, nm, dn, D (.)mn∈∈ A (,nm ) A ⎪ ………………… (3.13) ⎩ 0,otherwise
Page | 87
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
The results suggest savings to be made by integration of modes. If there is great variability in excess capacity of the one mode, shifting deferred items through integration can have further benefits in improved system utilisation.
Similarly, deferment can be considered as an attribute of intermodal flexibility. Rather than simply specifying a priori the quantum of the breaking and dispatch activity, the rail road mix fosters the engagement in break-bulk activities depending on its functions, available area and catchment needs. These activities require some deferment in forward arrivals so that the containers may be broken down to pallets, mixed and re-consolidated. Potentially this distributes the diffusion process to occur closer to the end load. For instance in Sydney, the large proportion of export containers (31%) are dispatched by road from the Port Botany area and this is largely due to re-consolidation activity from forwarders receiving exports across town by road and re-consolidating exports for the multiple destinations required overseas (Witjayaratna, 2005). Daganzo (1995) has demonstrated that the economics of such break-bulk activity through a node is greatest with a large number of diffuse destinations. It is of great research interest to analyse how the interface opportunities of intermodal terminals can modify this activity in location, function and intensity.
3.6 Facilities Management Models
3.6.1 Frontiers in Design and Analysis of Container Terminals
The task of dimensioning port facilities extends to assessing the configuration and resource of the sections of the seaport such as the vessel berth-crane, container stack, and vehicles and rail access and loading/unloading interfaces (Crainic and Kim, 2005). The motivation has been typically to increase the throughput productivity of container seaports. Both system dynamics and discrete event simulation models have been used to dimension aspects of the port. Alessandri et al., (2004) has used feedback control to solve the optimal control problem of resource utilisation at intermodal terminals that evolves. They used discrete event theory and did not directly incorporate wider network coordination issues. Critical bottlenecks evolve with the stack interface and with berth sequencing of vessel arrivals at the seaport (Fararoui, 1988). Simulations of stack management with novel technologies at seaports have also been investigated and is a thriving field of research (Günther and Kim, 2005). Network modelling of seaport terminals by Jansen (2001) using queuing
Page | 88
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
theory around nodes within combinatorial graph theory is one means of incorporating the changing nature of terminal unit processes and configurations with changing freight flows.
There are emerging challenges in handling containers in seaports with increased numbers of vessel visitations and increased capacity of vessels (DoT, 2008a). Whilst Australia is unlikely to get the new generation of vessels of 12,000 containers or more, it may receive vessels of up to a carrying capacity of 6,000 containers. Currently international vessel visitations representing container capacities of 3,000 are the norm.. Such an avalanche of pulse flows will have major implications for seaport clearances and the configuration and operation of hinterland facilities to relieve anticipated seaport congestion remains poorly researched.
From a simulation perspective, the Discrete Event System (DES) approach can cater for the nature of the freight task in a more appropriate fashion than continuous system dynamics (although both have a potential role to play). The DES approach can be a more accurate abstraction of reality in that it can simulate break-bulk movements as well as multi-faceted impedances at nodes, a condition which is unique for freight. These consignment movements are determined by several attributes for each unit flow. D’Este (1996) has suggested decisions be made on the following attributes which may represent a specific state of the consignment:
• Product type; • Location; • Unit of consignment; • Mode (also truck type and size); and • Status (Commercial).
The common mechanism of progression in DES is a next-event technique – that is an action in the system is caused by a previous event: the movement in a queue is due to the completion of service of a forward unit. This may be distinguished from a continuous approach whereby time progresses the actions (Pidd, 1997). A significant difference with a DES approach is that events can be ordered in sequence according to some logic. This is of particular value in analysing a freight system that is non-continuous. For instance, the flow of import and export containers in system dynamics occurs simultaneously and it is difficult to build in interruptions such as export flows having to wait for the arrival of empty import containers before they can be stuffed and then dispatched.
Page | 89
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
3.6.2 Models of Inventory and Distribution Theory
Continuous Approximation models have been applied to ascertain the synchronisation issues associated with production and transportation, particularly in determining the optimum lot size of production (Blumenfeld et al., 1991, 1985) . These approaches are much more effective in assisting understanding when they depict the implications of changing parameters in generalised cost equations. Here, cumulative arrival and dispatch curves are used to effectively depict the impact of different business rules on inventory size and meeting dispatch. Taleb-Ibrahimi et al., (1993) have also used cumulative arrival and dispatch curves to represent the effect of business rules at a container seaport which compare changing storage requirements with handling (container shuffling work). Daganzo (2003)) uses these in developing a theory of supply chains in order to assess stability and control policies. Cumulative curves are an effective analytical means to demonstrate the broad relationships between terminal operations under alternate business rules and freight consignment fluxes.
Much has been made of the requirement for intermodal feasibility on the basis that they are elements in logistics and supply chains and must be synchronised in operation with other shipper and receiver functions (SFCNSW, 2005; VFLC, 2005). Research has been undertaken on the manifestation in European jurisdictions of large freight villages (Intermodal Logistics Centres) with warehousing and other functions to support consolidation/distribution bimodal and intermodal terminals (Höltgen, 1995). These refer to greenfields opportunities and are increasingly inappropriate for Australia’s major cities16. There remains no modelling or planning framework in either Sydney or Melbourne jurisdictions for the arrangement of necessary functional elements at intermodal terminals. The research needs to beyond spatial location theory in order to determine robust functional configurations. This gap in research persists despite observations of the devolution of warehouse activity to outer Sydney (Rimmer and Black, 1982) and a call since the 1970’s for a better analysis of the co-location of freight distribution elements so the planner can assess novel configurations (Meyburg and Lavery, 1974).
16 Opportunities for establishing freight villages which may include intermodal terminals are possible in regions such as Goulbourn Valley around Shepparton in Victoria where there is a growing export supply of fruits and other perishable goods. Page | 90
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
3.6.3 Lessons from Terminal Design Methods
Model development stems from the process flow diagram. De Neufville (1976), Morlok (1978), and Hart (1985) have cited the benefits to analysis in developing process flow diagrams to understand terminal relations. In identifying the sequence of operations, the process flowchart can be used to calculate the transfer time for arrivals and departures (Worrall and Bruggeman, 1969, Fig 5.3.15 – 5.3.18), assess the arrangement of functions, construct engineering drawings and may be further used to analyse the effects of changes in events (Hart, 1985, p.17).
Cumulative activity curves can be used to identify the necessary storages required in a transfer process. This can involve a number of curves, for a range of interacting activities, which can be identified by the process flowchart. The vertical distance between the arrivals and dispatches suggest the storages required and the horizontal distance between the curves suggests the average waiting time. Bruggeman and Worrall (1969) applied this device in airport terminals to suggest the size of passenger waiting lounges. These probability curves can also be applied to assessing multiple interacting flows, such as baggage arrivals and passenger de-planning, to calculate the size of a baggage collection carousel.
Intermodal terminals, in interaction with networks have dynamic constraints which may involve evolving bottlenecks within and between the container production cycle, the storage cycle and the train dispatch cycle. The process flowchart provides insight into the necessary coupling that can occur. How to represent this coupling is a real challenge. Coupling refers to the interacting processes inside the terminal. For instance, the saturation of a downstream storage may propagate backflows to an upstream storage. Here, there is an interaction of storage resources. A central role of flow schemes may also be to minimise the extent of coupling or interaction. This can otherwise lead to bottlenecks in the system. For instance, transit passengers are kept separate from passengers waiting for a direct flight in airport terminal planning. Internal access roads and service bays are designated separately for commercial and household disposals given the different servicing times required.
We wish to develop a connection between a flowchart which specifies the physical transitions required (flow and function relationships) and the mathematical schematic. The development of an electrical circuit analog has the potential for converting the physical transformations required into a mathematically reproducible format. However it should not be built literally around the mesh circuit. The flowchart may be considered the starting point whereby intermediate transformations
Page | 91
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
are delineated. At the end of the process when functions are incorporated, there is likely to be a nested circuit from which a transfer function, or a series of transfer functions is developed through mesh analysis etc. Differing input streams and their functional transformation needs can be assessed for output delays according to the established characteristic impedance. When there are a number of terminals of distributed function to be planned, it is valuable to assess how changes in the functional distribution change the effect of the characteristic impedance on the lead time and also the total traffic flow.
3.7 Model Hierarchy Gaps
3.7.1 Critique of Network Methods to Address Bundling and Transhipment Phenomena
A key challenge in modelling freight systems is guiding the discrete-diffuse nature of freight flows. This particular phenomenon has been documented since Rimmer and Hicks (1978) and is particularly the case when considering transformations involving the containerisation of freight.. Conventional allocation mechanisms which may be based on least impedance route and system equilibrium of equal travel time need to be restructured so as to consider the attracting force of consolidation corridors between hubs (O'Kelly, 1998). In the hub location and service frequency model applied by Racunica and Wynter (2000), a discount factor on inter-hub links is applied to denote the cost reductions obtained by consolidation at hub nodes. Marginal costs should also decrease with increasing flows. To further account for economies of scale due to differing train services such as shuttles, non-linear discount functions are applied.
The traditional transhipment problem of Operations Research (OR) and developments from it has worked on prediction of multi-commodity flows over a multimodal network (Crainic and Laporte, 1997). Equilibrium outcomes are based on a substitution and competition approach to modes. In Operations Research techniques, we are adept at incorporating constraints for capacitated links and nodes but this technique does not lead itself to simulation and design of compensating mechanisms in the system.
Page | 92
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
The physical abstraction in OR models generally ignores the density and rhythm of flows – significant for the design of discrete – diffuse sequences and also in the control of the intensity of freight. Rather “all possible physical movements on the infrastructure” are denoted by the elements of node, link and modes (Crainic and Laporte, 1997, p. 1997). Load factors in the four-step model are only applied exogenously after allocation and distribution as an accounting task rather than considering load factor evolution as integral to the freight distribution activity. The designation of model types among strategic, tactical and operational does not adequately handle the dynamic design issue of freight intermodal production systems.
Service quality is a notion that underpins the allocative nature of multi-modal solutions. Tarksi (1986) identified that there were clear quality differences in the carrier modes of shipping, trucks, rail and aircraft in terms of the niche in a network and transport functional set. This description unintentionally reifies a notion of modal service quality and is responsible for model allocative functions based on competitive, least cost approaches. Thus substitution effects of different modes are often the driving force for assessing investment decisions (Loureiro and Ralston, 1994) , flexible capacity of a network (Morlok and Chang, 2004), or solutions to routing problems (Smilowitz et al., 2003). If the landside container freight task is considered the manifestation of a Large Technical System, involving numerous technologies and infrastructure types, it is of insight to consider how modes synergistically can improve the trajectory toward more sustainable formats, rather than follow the competition-substitution imperative. Bukold (1996) has noted that each mode has disbenefits called Reverse Salients which work against opportunities to improve the freight task in terms of integration with production systems and sustainability outcomes. These reverse salients include political and economic drivers which enforce silo activities. Combined transportation in the freight task allows a re-casting of the transportation technological envelope - improving cost structures, and driving more sustainable outcomes. A truer intermodal modelling ethos would involve collaborative frameworks, whereby the predictive part of the container freight task is captured (Groothedde et al., 2005), or further coordination frameworks whereby modes interface at terminals dynamically. Novel terminal technology and alternate rail operating forms can considerably alter the service quality of a particular mode and its success depends on the effective interface across modes. There are numerous plant location studies which identify the optimum location of facilities to minimise the total distribution task. What is lacking is an understanding of the tactical level criteria that supports the technical and economic feasibility of such transportation networks based on concepts of precision and controllability.
Page | 93
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
Transportation planning has ‘almost exclusively paid attention to the art of covering distances’, whereby the concept of mobility should also encompass the need to change location. This second neglected aspect focuses on the trajectories that come together in bundles remain for a duration and then depart (Hagerstrand, 1987, p.26). The salient measure of this aspect is load factors and is a means to manage both the supply response and demand in networks. The need for improved system coordination requires an understanding and manipulation of the system’s distributive, accumulative, and communicative aspects (Jonsson, 2004, p.213).
3.7.2 Gaps and Linkage Requirements across Model Hierarchy
The model formulation expounded by Crainic and Laporte (1997) trades off service level with the itinerary chosen and resultant cost.
Ψ m t += t m + k s ),( ∑ s s ∑ ,km XCFCFX k s ,km s m −− σ m 2 + ∑C ,km (min{ m k ()(,0 TnTES k )}) ,km P α − 2 ∑Cs (min{ ,0 XF sss }) s ……………..(3.14)
This objective function represents the total cost of operating a service network at a given service
m level Fs for a certain amount of freight moved along certain selected itineraries X k . The second
m line represents the cost of failing service targets Sm due to transportation delays Tk . The third line is a trade-off between the cost of increasing the level of service and the extra costs of insufficient capacity. This is a capacity-penalty delay.
A problem with this model formulation is that it does not acknowledge that the service characteristics and the transformations which depict the itinerary interact – this is especially so when peak (discrete) flows hit the terminal and the terminal response depends on its ability to compensate for this pulse disturbance. A feedback control problem occurs, for instance with import containers, if the service characteristics are set the terminal must cope with the disturbance according to the output load profile required.
Page | 94
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
3.7.3 Terminal and Hub Activity and the Time Composition of Freight
The feasibility of intermodal production systems requires both a measure of performance of utilisation of time availability of infrastructure and resources (Zeitnutzung) and utilisation physical resource capacity (Auslastung) (Tarski, p.72, 1986). Often the Auslastung concept is not fully considered in a network setting. System formats may be considered, with respect to the logistical structure, in the way that they manage the transformation of lumpy supply and discrete demand in the freight translocation effort. An example is the nature of the break-bulk activity. When we are considering how a system format conducts this transformation through analysis of the distribution of activities and the cluster intensity of how consignments move through the network, we are expressly considering the Auslastung concept. In this case, we go significantly beyond the traditional network representation of transhipment nodes as a buffer state, constraint or simple dwell time. Terminal and network performance is expressed in terms of physical productivity.
The impedance calculations include the Zeitnutzung aspects of the mesh schedule: particularly synchronisation of mobile resources. Network models usually only consider Auslastung effects by super-imposing load factors between the modal split and assignment phases. This is also after the generation and distribution phases. The Auslastung measure however should be seen as more directly associated with the generative and distributive effect. This is the basis of the criticism of the 4-step model by the activity approach. In the Distributed Function Hinterland case study, we consider the nature and relative location of this activity, such that compensating mechanisms best coordinate the freight intensity. In this way we attempt to match the load system with the capacity of the (sub)network with low traffic activity. Intermodal systems can facilitate this by their flexibility.
Consequently, the pressing issue of transhipment simulation is the assessment of lumpy supply with finely divisible demand. This has direct relevance for measuring system capacity, capacity flexibility and impedance. For instance, at the intermodal interface, containers arrive in batches by trains in a scheduled system (discrete). Trucks to collect the containers and/or collect part of the container contents, if there is an unstuffing function, may have characteristics of more continuous arrivals. In transport planning there is usually no dynamic quantification made of the interface between train arrival/ departures and truck arrivals and departures. This is particularly so if there is any break-bulk activity. Representing the heterogeneous requirements of unit loads is essential to capture time (impedance) related issues of transportability (goods- vehicle interaction) , interoperability (vehicle-terminal interaction) and handling. In sum, we are endeavouring to find a Page | 95
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
competent mechanism of transit activity to more truly reflect the nature of the visitation call. Flows through a terminal need to actualise a driving function which represents utilisation in time and physical resources (load).
In Chapter 5, the possibilities in developing a more comprehensive and insightful time composition of freight using classical electrical circuit theory is considered. In this theory, interface design occurs incrementally. The feasibility of design is assessed mathematically by impedance matching. A source and a load need to be joined. The interface designed ensures that both resistance needs are met. In the process flowchart approach, the work is done graphically to assess the relationships between flows and function. The feasibility of a design is largely undertaken by hand in the process flowchart approach.
Of interest in this thesis is how the flowchart approach can be abstracted for mathematical analysis. This would allow an analysis of intra-terminal interactions as well as assessing the connectivity match with other terminal nodes. In facilities design and management (Tompkins, 1996), there is little extension of the physical and service design of the terminal with the complementary operations of other terminals in the network of interest. Therefore the development of schematics that include the concept of impedance matching can provide us an insight into the feasibility of transportation infrastructure of different distributed functions.
3.7.4 Advancing Activity Based Approaches
The space- time arrangements of freight movement and the physical realisability of intermodal production systems remain poorly explored in transportation science. The gap is most pronounced when applying graphical network theory. Its allocation mechanism of traffic along arcs and through nodes is based on the calculation of equal travel time or costs. The pre-occupation with the use of time rather than the use of unit load capacity means that the phenomena of diffuse-discrete flow sequences are generally neglected. These sequences rely on the transformation activity at hub nodes which are generally considered as buffer states in the transhipment problem of Operations Research. To represent fidelity and achieve tractability in freight modelling these bulk sequence rhythms must be captured (Rimmer and Hicks, 1978; D’Este, 1996). These rhythms are generated by intermediate transhipment nodes and define the performance of an intermodal production system. This abstraction underpins a model identification which has significant implications on system tactical
Page | 96
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
design particularly on issues of efficient interfaces and tactical demand management, rather than to predict and provide more road space.
Rhythms of flows may be represented as pulses. Simulating pulse activity is most insightful for assessing transport performance involving transfers at interfaces. Taleb-Ibrahimi et al.,(1993) has calculated by continuous approximation of Newell cumulative curves that the storage requirements of a seaport would reduce if there were block arrivals that were compatible with ship destinations. The intermodal production system, which generates these blocks, can then influence the productivity of the seaport-hinterland system. Specifically, insights may be gained into the compatibility of train production cycles with container production cycles.
The approach described here may be considered an activity based approach. The salient characteristic of an activity based approach is that an activity pattern arises from the scheduling of activities. The behaviour underpinning this pattern is based on a set of rules and procedures which are stable in time (McNally, 2000). In turn, the freight planning profession needs a design approach of transport networks that focuses on the generative capacity of nodes. To represent desirable transformations in the unit load, not only do the distributive aspects of a transportation network need to be considered, but also the cumulative and communicative aspects.
For instance, flexible storage capacity in a terminal or sequence of terminals may be a key issue in the earlier consolidation of consignment flows. Geidl and Andersson (2005a, 2004, 2005b), in the Electrical Power Distribution field, identify various transformation opportunities in terminal nodes as representing different node typologies. These typologies reflect different paths through a terminal requiring alternate couplings of processes. Assessing how functions may be coupled in a terminal allows planners to consider how nodes transform flows. Daganzo (1990), using a Continuous Approximation Model, identified how the functional pathways of bulk and container loading can be coupled at a seaport, allowing the utilisation of same berths through available slots. Rosebrock (1992) assessed the opportunities to extend the capacity and flexibility of Deutsche Bahn to move freight and improve intermodal service by decentralising rolling stock storage and train formation. He formulated network movement following the theory of computer information packets. This formulation facilitated an activities based approach. Typical network optimisation would only consider a static path through a terminal. Characterising the functions in a terminal facilitates an analysis of flexible supply.
Another characteristic of intermodal terminals as activity systems is that they act as gateways (Woxenius, et al., 2004a). They are interfaces with flows from other networks. When considering Page | 97
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
pre- and post- container haulage we may discern that the functions of the intermodal terminal are related to other terminal interfaces. Consequently, when considering the feasibility of intermodal production systems, the line haul impedance becomes less significant than the impedances due to interactions between sub-systems.
3.7.5 New Rail Operating Forms and Handling Technologies
The feasibility of an intermodal system lies with its interface with novel rail operating forms. This connection is deemed vital to address hinterland productivity needs to relieve seaport congestion. Recently transport researchers have identified novel rail operating forms which differ greatly from the block train arrangement of earlier rail forms (Woxenius, 1998 , Woxenius et al., 2004, Ballis and Golias, 2004, Kreutzberger, 1999, 2008). These rail operating forms develop alternate discrete- diffuse container movements and require appropriate terminal functions and configurations in bundling/unbundling, storage, and transhipment. Figure 17 depicts five possible types. There are many sub-types. These novel rail operating forms lead to significant research questions in measuring and designing the flexible capacity of these terminals to respond to the different pulse and bundled loads represented by these rail forms.
Figure 17: Alternate Rail Operating Forms Interfacing Terminals (Woxenius et al., 2004)17
17 Corridor operating forms may also be known as trunk and feed. For a common rail service, corridor forms may also acts as liner systems (multiple intermediate stops). The hub and spoke form can represent a number Page | 98
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
These new rail network itineraries will only be feasible with alternate rolling stock. Rail shuttles have been considered for both Melbourne and Sydney and currently run in Sydney in a direct connect between a designated satellite terminal and the seaport precinct. Rail shuttles with operating characteristics similar to passenger trains have been flagged as appropriate in order to foster intermodal operations over the Melbourne broad gauge network where freight rail services must share rail capacity with passenger services (DOI, 2006).
3.7.6 Network Analysis Needs for Physical and Service Design
Intermodal systems can be implemented in built-up regional and urban areas. Both the physical design network and the service design network need to be considered in planning. There are a number of information needs that must be addressed (Figure 18).
of typologies. See O'KELLY, M. E. & MILLER, H. J. 1994. The Hub Network Design Problem: A Review and Synthesis. Journal of Transport Geography, 2, 31-40.). Often the intermodal literature refers to hub and spoke iin the context of single mode rail forms. This thesis extends the form to consider road-rail over shorter distances. Allocated routes represent more dynamic routing opportunities depending on the availability of terminals and levels of road and gateway congestion. Page | 99
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
Figure 18: Information Needs for Intermodal Design (Beavis, Black et al., Fig. 4, p. 2986, 2005)
Issues to investigate include: appropriateness of location given existing network and hinterland land use; use of the intermodal facility given impedance and service characteristics; necessary capacity and technology inter-operability; supporting link/node infrastructures; and fostering of necessary flow characteristics – that is, load unit densification road-side and frequency of rail service. In the Sydney context, like many densely populated environments, there is a need to investigate the possible trade-off in the conflicting objectives of concentrating flows to sustain intermodalism, terminal optimum capacity, and minimising flow concentrations that might otherwise lead to localised spatial disutility. Sydney has an unusually highly concentrated container hinterland, therefore the congestion effects of new transhipment activity may be pronounced, particularly through large hubs.
Page | 100
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
3.8 Assessing Capacity Measures
3.8.1 Network Capacity Measures
Capacity is conventionally defined in the transhipment problem as a constraint phenomenon. It is becoming evident in recent facilities research associated with intermodal terminals (Ballis and Gollias, 2004; Hulten, 1997) that capacity depends on unit capabilities and on interoperability of system resources. Flexibility, as well as static physical capacity, is identified as a key characteristic of impedance. The sparse representation of terminal operations then needs to incorporate opportunities for changes in the rate of interface transactions between system resources (and how this alters transient storages). The use of different unit processes and paths within a terminal define a specific terminal activity. This representation needs to capture sufficient terminal activity to characterise capability measures. This would aid in both terminal and larger system optimisation through coordination schemes. Current network modelling is insufficient in capturing the flexible capacity of freight distribution systems.
A means to understand the gaps in network theory and practice toward intermodal system integration (Woxenius, 2007) is to discern the nature of the itinerary. For this thesis, an itinerary defines an available intermodal production system. Conventionally, the itinerary is selected as the unit in tactical planning involving a combination of nodes and links along a network which have the least impedance cost (Crainic and Laporte, 1997). The allocation mechanism selects the itineraries18.
3.8.2 Capacity Assessment of Terminals
Generally seaports have two measures of capacity: physical storage and throughput capacity. These measures are supported by efficiency figures of merit for stack area (TEU/ha/year) and crane handling rates. What is not evident are Figures of Merit that assess how the productivity based on these efficiency indicators would change with changing profiles of vessel arrivals (in size, and
18 Crainic and Laporte present an optimisation tactical model which includes aspects of itinerary cost and service frequency by iterative means.
Page | 101
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
frequency) and how trains taking consolidated loads could expedite throughput at the terminal. Morlok (1978) has illustrated that the practical operating capacity of a terminal critically depends on the arrival flow profile (Figure 19). That is, the more variable flows are, the slower the fluxes are through the terminal due to increased impedance.
t1
C1 C2 C3
Figure 19: Terminal Time-Volume Curves with Varying Arrival Headways (Morlok, Figure 7.8, p.263, 1978)
In essence the impact of pulse flow changes are not easily considered with this indicator set. In addition, the capability of the hinterland to store and prepare containers is not considered. Currently, in Melbourne and Sydney, much transient staging of containers and Less than Container Loads (LCL) occurs due to mismatch in operating hours and limited seaport accessibility during peaks. In Melbourne, it is estimated that some 50,000 transient slots are used throughout the network as there are severe road access constraints to the seaport.(VFLC,2008). Consolidating this activity would reduce the landside container freight task in trips (and thus reduce the kilometres/tonne).
A potential capacity constraint in intermodal terminal production systems is the availability of sidings and necessary shunting operations. Ballis and Golias (2004) have noted that sidings are a major opportunity and constraint to rail throughput between terminals. With limited frontage and siding capacity a terminal is constrained by train length. Without sidings, for instance, the train can only be as long as the terminal frontage. Train lengths can be increased through the stagewise production and loading onto available wagons and then shunting them to an adjacent terminal. The
Page | 102
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
train can then be “made up” by hooking the latest production wagons to the stored wagons at the sidings. This however adds to a significant impedance whereby the dispatch only occurs once shunting and hooking up has been completed. It can however reduce the storage size area requirements of the terminal by providing pre-dispatch loading of containers onto wagons. In this case the terminal may be able to handle a greater proportion of direct transhipment despite having the same dispatch sequence. Utilisation of sidings can also allow the terminal have multiple purposes (Daganzo, 1990). That is, provide additional throughput capacity for other flows to be dispatched on other services when the siding- mainline link is free. This can support the feasibility of an intermodal terminal when its baseload traffic is reduced.
There has been little recent work in characterising the capacity and impedance arising from sidings configuration. Petersen (1977a, 1977b) developed a typology and mathematical method using queuing theory for each sidings yard typology. His focus was in calculating the average delay in train dispatch based on the need to re-classify train carriages. Turnquist (1982) extended this analysis to consider a model formulation for carriage dispatch based on a bulk queue formulation. Crainic and Laporte (1997) in a review of freight terminal modelling have noted that the limited mathematical development and model formulation of rail sidings yards was a significant gap in rail freight transportation science. A significant research field is the capacity consumption of sidings and their flexibility under alternate train service scenarios.
In the Sydney container Landside Capability Study (SFCNSW, 2005) the authors identify that Intermodal Terminal Capacity is currently at 160,000 TEUs/yr - and it needs to rise to 1.2 million TEUs/yr in order to capture 40% of TEU flows. The seaport-rail interface is a major operating constraint due to marshalling/shunting operations required and thus train path timetable not always met. Their policy suggestions are for a new dedicated freight rail route east-west, and arrangements for empty container storage adjacent to intermodal terminals for stuffing/unstuffing activities. This report still considers capacity rather than capabilities. That is, it does not consider the interconnectivities among landside terminal infrastructure. Furthermore, there is no ready assessment tool for how rated landside capacity could be flexible to changing conditions of demand and supply. Flows through a terminal may not have a fixed itinerary as it depends on their physical and information attributes . Thus the impedance flux will alter. A case in point is that one perceived benefit of intermodal terminals in relieving seaport access and storage congestion is the ability to improve synchronisation of landside flows to arriving and departing vessels. Intermodal terminals offer a sequencing prioritization function, particularly valuable in close proximity to the seaport (Ponton, 2006). Work on sequencing consignments and vessels has like that of Fararoui (1988) Page | 103
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
focused on the seaport. There has been scant consideration in how intermodal terminals landside could manufacture such sequences to further facilitate swift seaport throughput.
What is needed is a mechanism that can assess the hinterland absorptive capacity of changing operations and flow profiles at the seaport.
3.8.3 Misspecification of Capacity in Assessing Terminal Performance
Capacity at intermediate terminals, particularly those involved in intermodal networks, is not a straight-forward physical constraint as it is usually portrayed in graphical network theory. The ability of the node to process throughput will alter according to the heterogeneous profile of the freight inflow (such as the rhythm of its intensity), the level of static and mobile resources to process this throughput, the scheduling of storage capacity (i.e. the operation of shunting) and the resolution of supply push and demand pull features. Terminals are not isolated entities and their performance should be considered within a larger sub-system of inter-acting terminals in cascade series and parallel. All these interactions involve consideration of the activity level of the terminal plant. Thus the impedance functions at the terminal are complex, and non-linear including regulating feedback loops.
3.8.4 Opportunities with a Process Control Approach
In the classical transhipment problem, the constraints are defined on the links and nodes with respect to saturation capacities and/or switches (i.e. a mode will take a certain commodity along a path or not). These non-linearities allow some dynamic response with a freight flow profile which may be heterogeneous. However these constraints, in being defined up-front (a priori) in the optimisation problem are in an aggregate fashion and do not account for the flexible capacity of the node infrastructure. It was reviewed in section 3.6.2 that the impact of broad business rule control on consignment fluxes through a terminal can be assessed analytically using cumulative flow curves. This approach does not easily lead to design for specific unit processes in the terminal.
The process control approach (Dorsey, 2002) allows constraints to operate according to process specific activities. The flexible capacity of the terminal and its system may be assessed by how well constraints are partially relaxed with bringing on-line additional equipment provision, server
Page | 104
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
scheduling, storages (through feedback operations). Here, complex terminal activity is incorporated in system capacity assessment. Thus we can get a better understanding of how the terminal production system can manage constraints. Performance is thus defined on capability (physical and non-linear coordination of services) rather than capacity (physical constraints in linear framework).
Usually, terminal design is divorced from system configuration. [Intra-unit flow does not generally form part of the detail of network tools]. The flow and function flowchart gives us a basis in forming system equations based on storage relationships (integro-differential equations). Process control can be easily applied to LTI systems and this allows us to depict complex terminal activity.
The flowchart in this approach is the starting and endpoint of design. We can approach design differently if we can tie in the demand load profiles with changes in the terminal operation. We consider at this level an interaction among strategic, tactical and operational concerns.
There is also a fundamental limitation with using a network approach for design solutions (ecological re-structuring). The distribution infrastructure, with constraints set a priori, is passive to the mutations of the logistical (flow) system. The network approach can be said to facilitate a notion of dispositive elasticity, which is most obviously manifested in localised congestion on key links. By considering the system as an interactive production system, we can design demand management procedures in terminal functions.
3.9 Anticipating the Connection of Terminal Design to Coordination Approaches
Coordination is a more tractable solution method to resource routing under conditions of flow imbalances than a conventional allocation approach. The increasing asymmetrical nature of container flows in the Sydney basin place a larger burden on the re-positioning of freight resources (vehicles and empty containers). A dominating unilateral flow also drives the requirement for the injection of infrastructure supply, making commercial arrangements potentially fraught with difficulty. Port Botany is predominantly an importing seaport and is destined to grow rapidly. The land-use development of generators and attractors and hinterland deployment of freight infrastructure may further stretch the capabilities of the system to feasibly operate (SFCNSW, 2005). Harmonising cycles then become the critical endeavour in ensuring resources (infrastructure
Page | 105
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
availability, ancillary resources) are utilised in a way that minimises any imbalance (Hulten, 1993). These cycles may involve a role for transient storages in intermediate terminals.
From a container freight perspective, the decentralisation of break-bulk activity from the seaport to a hinterland terminal represents distributed generation of the current freight service. Distributed generation systems establish couplings of transformation activities at nodes (Geidl and Andersson, 2004). Indirect and direct transhipment characterises composite functions at terminals. In incorporating this node complex activity, the network becomes potentially flexible in supply of freight services. Facilitating dynamic routes is a significant potential benefit of developing intermodal production systems in the hinterland. Improved communications may lead to better coordination between available container space on rail and ready consignments. This facilitation of the Industry Benefits of Trade may improve system utilisation. A node-centric approach to networks allows us to gauge the benefits and feasibility of distributed generation systems.
Electrical power distribution theory has developed a number of algorithms to analyse decentralised node generators because a major research endeavour is to assess the cost and energy savings and improved reliability of smaller generating plants. Specific applications include introducing renewable energy sources to the existing power network (Korpaas et al., 2005) and assessing the benefits of transient storages over augmenting transmission lines (Koeppel et al., 2004). The two mechanisms at play are generative scaling and inter-temporal storage which combined can reduce the total supply effort required for the load demanded. Inter-temporal storage represents a dynamic cumulative function of a network. Generative scaling considers the feasibility of smaller terminals to service the same overall throughput. These coordinating concepts also have validity in the design of novel freight networks. Consignments arriving at different rates can be qualitatively transformed such that the next dispatch in the sequence of flows has higher load factors and characteristics which mean less handling complexity upstream.
In a network of open access regimes, where multiple actors may utilise freight terminal resources, dynamic routing can improve the utilisation of infrastructure which requires a critical mass of consignments in a timely fashion in which to operate. Dynamic routing has been canvassed as a significant means to develop feasible intermodal networks by Ballis and Golias (2004) and Woxenius (2007), yet solution methods are cumbersome and incomplete. Demonstrating dynamic route utilisation necessitates model frameworks with scheduling mechanisms which poll the complex activity of terminals and their itineraries at various temporal resolutions. Thus what is distinctive in coordination mechanisms as opposed to allocation schemes is that the itinerary is an
Page | 106
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
inter-temporal path of commitment and dispatch. The terminal nodes generate the service according to their availability. Nodes, through their compensation mechanisms (such as bringing additional handling and storage capacity online), provide a feedback mechanism to central coordination. In this way, unit capability is explored to optimise system capacity. This coordination mechanism with node resolution of control mechanisms leads to a transitions approach for transport infrastructure deployment. In this way, strategic and tactical decision support needs are conjoined. A gap which needs addressing is how to test and design different functional attributes of a distributed resource system in order to ascertain lowest cost, and lowest road freight intensity outcomes. This thesis anticipates further research into testing the feasibility of a distributed resource system approach to the freight transhipment problem. It does this by outlining possible mechanisms at terminal nodes that may be used as a basis for system-wide coordination algorithms.. This extension is further investigated in Appendix F.
3.10 Intermodal Modelling Methods and Gaps
3.10.1 Straddling Horizons in Space and Time
The scene of intermodal planning and modelling has been surveyed (Caris et al., 2008, Crainic and Kim, 2005, Bontekoning et al., 2004). Bontekoning et al., (2004) noted that intermodal planning is without adequate theory and systemic design tools and should therefore be considered as pre- paradigmatic.
Caris et al., (2008) have surveyed the literature and classify planning problems according to four decision makers: drayage operators, terminal operators, network operators and intermodal operators. Across the three time horizons of strategic, tactical and operational this leads to 12 categories of intermodal operations problems. This segregation, though convenient, does not consider critical interfaces among the actors nor overarching governance issues. There are very few modelling approaches that straddle time horizons of strategic and tactical or all three horizons.
Sustainability gains through urban leverage require an analysis of how we can inter-connect these networks which do not act autonomously. Also required are model abstractions that reflect this i.e. pre- and post haulage from the container network. A fixation on the costs of the intermodal operator solely will not yield a feasible intermodal system. The shipper has considerable costs beyond the
Page | 107
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
transhipment operation; these include the management of empty containers. Where the intermodal terminal can hold, distribute and repatriate empty containers will add significantly to reducing whole of cycle container costs to the shipper/ receiver (SKM.2003). These model approaches are rarely considered in strategic freight container studies and models.
3.10.2 The Perils of Neglecting Urban Intermodalism
Assumed wisdom in freight intermodal theory coming from Europe and North America is that intermodalism is only feasible with an extensive line haul i.e. at least 150 km (Terminet, 1998). Thus it is best suited for inter-city container freight task including landbridging between ports via a hinterland. Intermodalism was until recently seen as being measured on the economies of scale, scope, and density (Slack, 2001). That is, benefits of volume for high throughput, catchment reach, and role in the hub network to consolidate and distribute diffuse consignments.
Additionally site footprints for metropolitan intermodal terminals are limited. In freight intermodal literature this, along with the fact of short line-haul increasing the relative transhipment cost, is cited as reasons why urban intermodal terminals are infeasible. Available land for intermodal terminals in the metropolitan region is typically between 10-20 ha.
Whilst there remains a need for large regional intermodal terminals, the undesirable effects of large intermodal terminals for the urban environment are being increasingly questioned. These problems include inflexibility and high carrying capacity risks, large areas required and large local disamenity due to high levels of truck movements. Additionally their colonisaton of network form in receiving and dispatching only long haul traffic means that local bundling networks are neglected (Rodrigue, 2004, Woxenius 1998) This undermines the premises that previously dominated intermodal theory for the urban environment.
A number of authors have lamented the lack of uptake of intermodal opportunities, particularly urban forms (Kreutzberger, 2008, Woxenius, 1998). This has led to research into understanding terminal requirements to meet alternate train operating forms such as shuttle trains, trunk and feed, liner and hub and spoke networks (Ballis and Golias, 2002, 2004; Kreutzberger, 2008). The most salient impact on intermodal terminals with these novel forms is the need to respond to increased frequency of service. Rodrigue (1999) has suggested a new modelling paradigm is required that accounts for the tense flux of precision relations that urban intermodalism will require.
Page | 108
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
3.10.3 Outstanding Research Agenda for Urban Intermodalism
The outstanding research agenda for intermodalism includes (Caris et al., 2008): 1. Drayage operations around an intermodal terminal; 2. Design of intermodal service networks using bundling concepts; 3. Developing solution methods further using Operations Research techniques; 4. Methods of coordination of intermodal networks including testing the degree of openness of access regimes.
The review in this thesis adds that there are critical outstanding interfaces such as understanding the interaction of physical design on service characteristics. Further, the Multi Commodity, Multi- Modal Allocation Problem needs to incorporate terminal Multi-Path Coupling so that novel terminal networks and configurations can be tested for alternate operating forms. The literature is heavily weighted toward the feasibility of terminals on the basis of truck vehicle access and there is less consideration given to optimum configurations for the variety of train operating forms necessary to make intermodalism, especially urban intermodalism, viable.
For the focus of this thesis, there are some pertinent gaps in the terminal container literature in addressing closed form analytical solutions for storage and handling phenomena. These are useful starting points to pursue in order to understand the character of urban intermodal systems. These research areas have been summarised by Crainic (1987) and include:
1. Measuring and controlling sidings capability; 2. Container stack control management; 3. Pulse correspondence (in the case of transmodal transhipment operations between train services, for instance).
Additional pertinent phenomena in tactical modelling of intermodal terminals, arising from the technical literature and the geographical problem context of this thesis, include: 4. Bulk queue control; 5. Sequence ordering; and 6. Bundling small freight consignments through the terminal unit processes.
Page | 109
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
The observation of Bontekoning et al, 2004 remains potent. The problems in intermodal transport are complex and require new knowledge to solve them. Research into terminals is needed in order to “obtain a fundamental understanding of the impact of the different arrival and departure dynamics of trucks and trains, terminal layout and operations strategy on terminal performance”, (Bontekoning, 2004, p.24).
Consequently, the orientation of this thesis is to develop analytical concepts which consider how intermodal terminals respond under alternate freight profiles according to a number of mechanisms. This focus feeds the larger network research agenda cited above as future research.
3.11 Synthesis Mapping Intermodal Information Needs to Mathematical Models
3.11.1 Devising a Functional Specification
In the transhipment problem of Operations Research, nodes are considered buffer states which add impedance to the line haul trip (Dantzig, 1959) rather than activity nodes which may facilitate consignment transformation. This transformation is critical in understanding the role of intermodal terminals and waste transfer stations in consolidation networks. Also precision aspects need to be incorporated which relate to synchronising production at the node and the train service.
Devising a functional specification for an intermodal system can be an extensive undertaking. The decision support requirement depends on the stakeholder client - for an infrastructure supplier and operator (government and or private) the requirements can be multi-faceted. Beavis, et al, (2005) developed an approach to devising a functional specification for urban freight planning. The modelling needs are identified from the conceptual model Figure 15. These relationships are further developed and suggested modelling methods are applied. The stakeholder links information (Figure 18), and modelling methods are sourced from the freight modelling taxonomy in Figure 16. The stakeholder may then construct their own taxonomy of model methods to address the problem at hand: location and investment, fostering flow bundling logistical networks, facilities design, rail scheduling. Significantly, stakeholders’ needs can span strategic, tactical and operational concerns.
Page | 110
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
There remains the challenge of integrating these disparate modelling methods to assure that the intermodal solution delivered is the most appropriate and is feasible.
3.11.2 The Need for Design Figures of Merit
Whilst there exist performance indicators for intermodal terminals, the mathematical language has not yet developed to synthesise a design perspective. Figures of Merit on Combined Transportation Systems are disparate and largely conceptual in characterising system performance with terminal operations. Ballis and Golias (2002; 2004) have constructed marginal cost versus throughput curves for alternate terminals configurations and train services. The capacity-risk, the propensity of a certain proportion of the system being available under different conditions with increasing throughput, has been conceptually discussed by (Kleinrock, 1974, Bukold, 1996, Rodrigue, 1999). All authors note that hub systems can potentially allow high throughput and manage capacity-risk to varying degrees. The level of complexity versus system controllability has been mooted as an insightful figure of merit in how varying system mobile resource technologies (trucks, transhipment equipment, train body types) should be (Sjöstedt, Woxenius, et al., 1994). This is a concern with technology interoperability. Interoperability also extends to managing consignment Origin- Destination information diversity and an intermodal architecture has been outlined (Duerr and Giannopoulos, 2003).The Simple Intermodal Tracing and Tracking Solutions (SITS) protocol offers a platform for information exchange over complex logistical chains. A measure of complexity in logistics chains has been developed by Waidringer (2001). This complexity index allows planners to identify how links can be made less complex and thus less vulnerable. Tarski (1986) has identified that the measure, degree of direct transhipment intensity, allows planners to assess the effect of flow information on terminal function, configuration and throughput.
These trade-off relationships are largely not in a form where performance assessment of alternate terminal functions and configurations can be readily applied. Additionally there is not a means to assess by closed form analytical solution the effect of alternate unit process arrangements that lead to different itinerary opportunities in the terminal on the availability, controllability criteria or on the impact on storage- throughput capacity of added functions such as complex bundling. The financial significance to the terminal of bundling functions has been identified for urban (SKM,
Page | 111
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
2003) and regional terminals (SFCNSW, 2004c) yet this activity does not have a model formulation within a larger formulation of terminal multipath coupling.
3.12 Conclusions for a new Phenomological Approach
This chapter has reviewed the literature with the intent of evaluating to what extent the essence of transhipment networks are captured for novel bundling design. The motivation was to assess what tools were available to address research design drivers of: 1. Minimising the evolution of system deficiencies by controlling dissipative flows and complexity; 2. Developing and measuring attributes torwards hinterland absorptive capacity; 3. Measuring Interface precision requirements based on ability to achieve synchronised consignment carrier pairing; 4. Designing terminal interface controllability between unit processes to manage storage saturation; and 5. Demonstrating the intermodal value proposition.
Freight network models have a predictive intent and are used to fuel infrastructure supply growth, predominantly in Australia for roads. At the other end of the spectrum, logistical vehicle routing and scheduling models aim to optimise specified fleets for specified tasks, and whilst a contribution, assume an underlying basis in infrastructure use which is dissipative. Facilities models cover a wide spectrum of analysis. Many also have an explanatory and design intent which is valuable. There is a gap in the linking of freight terminal dynamical models with network planning and modelling. This linking would assist the further development of an activities based approach and open opportunities for coordination schemes more suited to integrating the freight task over multiple modes. Novel model formulations can give rise to new opportunities to link terminal operations with wider network coordination. Augmenting the Multi-Modal Multi-Commodity problem with a Multi-path Coupling mechanism around terminal activities will offer opportunities in investigating the flexibility of itineraries and lead to outcomes with a coordination focus. The science theory which drives much network modelling has a focus on mobility improvements and anticipating the need of injections of physical supply to support this. Demand management and flexible supply through accessibility improvements and consolidation opportunities are less
Page | 112
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
developed. Consequently the tools to assess the opportunities for urban intermodal forms are scant. Specifically,
1. The landside terminal location decision cannot be effectively made in isolation from the functional decision - yet the model stratification of research between strategic and tactical does not easily facilitate this. This stratification, though mathematically convenient, does not address new decision support needs to articulating the functional relations between terminals for instance the relations between seaport and landside intermodal terminal. This can leads to ad hoc deployment of infrastructure which embeds logistical structures which in turn lead to dissipative and congested activities. 2. Stuffing mechanisms for bundling containers at terminals are not presented analytically in the literature despite this function being identified as a significant one to justify the financial feasibility of intermodal terminals and support the capture of local freight flows destined for container activity. This limits analysis and design into functional capability of intermodal terminals. 3. The activity based approach in freight modelling is not well developed operationally. A number of analytical relationships found in Continuous Approximation Modelling could be used in analytical solutions of terminal activities. 4. There is a limited concept of impedance. It neglects effects of work and resource utilisation at the terminal. 5. Performance measures of the effectiveness of infrastructure deployment to achieve combined transport outcomes have been conceptual articulated but not operationalised as design Figures of Merit. Thus there is reduced ability to compare alternate intermodal system formats to be deployed. Measures such as flexible capacity have only been addressed in network theory and modelling by redundancy aspects rather than adaptability- capability aspects. Consequently new dynamics in tense flux relations and measuring the support the landside logistical structure can provide to facilitate lean productivity seaports have not been mathematically formulated let alone developed as a modelling device. 6. Combined cost curves have been constructed for the system formats of terminal handling and rail operating form of marginal cost for throughput. It depicts how terminal limiting capacity can be extended or constrained with line access capacity. These are specific for European line haul distances (usually greater than 150km) and depict a trend for marginal cost performance with higher throughput based on larger terminals. This trend may be misleading for the urban intermodal case whereby terminals act as gateways for both the
Page | 113
Chapter 3: Gap Analysis in Freight Modelling for Intermodal Outcomes
road and rail network in improving overall productivity and utilisation. Pre- and Post Haulage costs are not considered in so-called intermodal cost curves. Alternate terminal configurations of small size can support high frequency rail services and lead to a throughput which equals or exceeds larger terminals. Thus terminals may not need to be located outside cities on very large tracts of land.
It was found that, for the planner, there are a number of gaps in systemic design tools to address their information needs for intermodal investment and governance. Following the research drivers and assessing the partial success of policy responses to intermodal deployment, the following abilities are required: 1. To link terminal function complexity with proposed rail operating form and reveal precision and value proposition information.; 2. To incorporate flow pulses of consignment blocks into a storage-handling scheme; 3. To consider retrofit of urban networks with rail-road intermodal exchanges as an engineering-economic endeavour. That is, the itinerary can be considered a pathway of commitment and dispatch; 4. To include governance and business rules; 5. To model and measure flexible capacity; 6. To lead to mechanisms of improved coordination in control diffuse and discrete freight flow sequences and thus better match the load system with the capacity of the network; and 7. To equip planners with skills to comment on harmonisation initiatives for the seaport- container hinterland system.
This chapter assists the reader in understanding the essence of the transhipment problem specific to urban intermodal system formats. The means to trial such formats are only partially available in network and facilities models. A model formulation of urban intermodal systems requires a new platform of conceptual validity of the freight task. The following chapter outlines a functional specification for intermodal systems which will support investigations towards a new model formulation then presented in chapter 5 and illustrated in the case studies of Chapters 6 and 7.
Page | 114
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
4 Functional Requirements for Sketch Planning Urban Intermodal Production Systems
4.1 Introduction
4.1.1 Chapter Guide
In order to justify the modelling approach presented in the rest of this thesis, it is necessary to provide a functional requirements outline for a tool in sketch planning urban intermodal terminal operations. This specification scopes the investigations of the essence of the transhipment problem with which this thesis is concerned. It can then be identified how well the proposed modelling mechanism fits in testing different system formats for fidelity and tractability. The tool which meets these specifications can then be used to identify opportunities in developing intermodal networks and assessing changes to existing networks.
The following objectives are considered in the sketch planning endeavour:
1. Discern how terminal configurations and business rule operations can deliver on certain train service schedules and thus support critical payloads; 2. Express the relationships between flow and function at a freight terminal mathematically so that we may assess how node complex activity and freight heterogeneity affects larger system performance; 3. Assess what functional configurations of distribution infrastructure can mediate freight road traffic in Sydney and Melbourne through increasing the load factors of freight and deferring delivery from peaks; 4. Consider how terminals which siphon more complex bundling19 can lead to further improvements in driving down the road freight task and improving seaport access clearance by reducing container dwell time at the seaport. This is enabled by the more efficient ordering of flows and smaller storage requirements.
19 Complex bundling is the notion that consignments with multiple Beginning and End points are consolidated and transshipped at an intermodal node. Often this will require more storage buffer requirements to meet specific outgoing mode services. Concept coined by Kreutzenberger (2008). Page | 115
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
To achieve these objectives, a tool is required that can:
1. Coordinate flows of goods, resources, and information; 2. Analyse node impedance dynamically in a partial network; 3. Investigate and characterise transit dynamics (inter-temporal storages) and approximate this continuously in the steady state; 4. Incorporate physical processes and business rules; 5. Account for multi-path coupling in a terminal. That is, track itineraries of consignments through unit processes and how they combine and separate; and 6. Provide a mechanism to track multi-commodity flows. The terminal should be treated as a multiple input – multiple output system.
The system boundary of the critical seaport-hinterland is first depicted with a two node linkage. This raises intriguing questions on intra-terminal and inter-terminal dynamics. An activity based theory of node analysis in transportation science, especially for the urban container freight task,20 is then described. This is compared against the traditional approach in node terminal design. Criteria for assessing the fidelity and tractability of models of intermodal systems are restated from the gap analysis of Chapter 3. Critical terminal operations are described and they provide the salient phenomena for model identification. Additionally, levers of control include application of different business rules for supply provision and demand management. This model identification requirement leads to new performance specifications which consist of five key variables of: freight value density, consignment flux, transhipment intensity, value intensity and dwell time.
The terminal node proposed has a rich impedance function that captures six critical measures of operations and alters according to the terminal complex activity and its use. The hierarchy diagram of indicators describing harmonisation goals in Chapter 2 is further developed. These goals are taken further in this chapter by developing specifications that link to impedance attributes. Impedance attributes measure load following requirements of the terminal in its interaction with consignment flows. A combination of impedance attributes, represented by transfer functions, characterise certain precision relationships between carrier pairs necessary to meet the load following requirement. This can lead to an analysis of how well flows between terminal nodes are
20 The Urban Container Freight Task is that where containers originate from, are destined to, or pass through the urban network. Often there is a large generator/ attractor gateway node such as a seaport or urban terminal hub that coordinates flows. Page | 116
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems facilitated. The impedance specifications act as performance objectives. The outputs of the modelling yield key Figures of Merit which allow us to assess how well we have met these objectives by comparing performance variables. In this way Figures of Merit can also show planners the need to trade-off variables to achieve the objectives. The specification concludes with a re-statement of the thesis objectives.
This requirements chapter positions the further definition and the development of the science theory of electrical circuits in freight transportation science in chapter 5.
4.1.2 Two-Node System Scope
The system scope of investigations is present in Figure 20. This anticipates the ambition of the functional requirements. Both intra-terminal and inter-terminal dynamics are considered. Certain observations can be made regarding the interface dynamics between nodes. The availability of the rail sidings at either Beginning or End point is dependent on the availability of the mainline and the availability of the receiving sidings. This must consider some transportation lag.
Page | 117
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
ty mp E
= [ −− ] )( ∑ j ijl actHBtR j )( ijl
Figure 20: Hinterland Absorptive Capability and its impact on Seaport Capacity: Gateway Node Access and Storage Capacity
Page | 118
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
The use of sidings to meet payload requirements depends on the operations at terminals. The interface between operations (the production cycle) and train services (the transportation cycle) may be buffered by a stack storing container for shorter and longer durations. This is known as indirect transhipment (1). These cycles can also be directly interfaced when the made up container arrives at the node and is timed to load a train service or to collect from a train service (2). Control of stack overflow is critical and this can be facilitated by movement of consignments to sidings or moved by a road vehicle with higher loading capacity (bi-modal exchange (3)). Intermodal terminals can also provide transmodal activities by acting as hubs for other shipments by rail (4). There are dynamic paths for goods to take within terminals according to required functions (stuffing, storage, direct loading) and available resources (stack, sidings).
The means to synchronise these operations requires the flow to carry information about timing. Alternate configurations of the terminals need to develop itineraries through the terminal which enable consignments to meet this timing imperative.
Intermodal terminals can improve the access clearance at a gateway node (such as a seaport) by synchronising train service dispatches with vessel departure dates and thus minimising the storage required at a seaport. Consignments are indexed by the cut-off date their designated vessel is due to depart (Ct-1, Ct-2,….) and are loaded onto train services close to a departure date. This can further reduce the dwell time at the seaport.
The control of the stack becomes more critical when carrier pairs are linked to vessel services of alternate arrivals and destinations (Taleb-Ibrahimi et al., 1993). This can be seen with the aid of cumulative arrival –dispatch curves. Taleb-Ibrahimi et al., (1993) develop the function
= [ −− ] )( ∑ j ijl actHBtR j )( ijl ………(4.1)
This represents the release- assignment storage curve based on a step function of vessel or train dispatch with arrival period of feed consignments at the terminal for that dispatch. The cumulative graph mathematics relevant to this problem is described in Appendix B.
Similar to the two-node, two–zone problem presented by Blunden (1981), Figure 1 anticipates that the phenomological approach required for an intermodal transportation science needs to capture driving and driven forces which must consider intra-terminal and inter-terminal dynamics. The imperatives are control of the stack with overflow options, sidings activation, and support of critical payloads on scheduled train services. When multiple commodities (multiple destinations) are Page | 119
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems included, this two-node problem can demonstrate how complex bundling supports train payload requirements in their mass utilisation and schedule.
Complex bundling is seen as a critical phenomenon for the development of urban intermodal systems (Kreutzberger, 2008, Notteboom, 2008). The impedance mechanism is not simply an elapsed time attribute: timing information is included in the mechanism to facilitate the analysis of carrier pair itineraries. This augmented mechanism allows the planner to pose questions on the capability of the hinterland to adapt to changing flows emanating from a gateway node such as a seaport and in complementary fashion explore opportunities for the hinterland to relieve seaport congestion.
4.1.3 Need for an Activity Based Theory of Node Analysis
4.1.3.1 Relating Terminal Complex Activity to Rail Operating Forms
Techniques are required to model options for the relief of seaport-hinterland interactions. These interactions represent considerable land-use dilemmas for city areas. Efficiency increases at port with swifter handling equipment are offset by growing demands to handle pulse flows with larger container vessels. Therefore overall productivity at a seaport will improve only marginally, meaning that practical operating capacity of seaports with fixed land resources may possibly saturate sooner. This depends on the growth rate of containers. Rail intermodal linkages offer capability improvements. For instance, a sketch modelling approach which considers the intermodal and transmodal opportunities of rail nodes allows us to evaluate novel hub forms which support the role of rail in addressing growing avalanche pulse flows from a gateway node like a seaport by adding flexible capacity, or capability, to the system. This leads to the necessity to determine precision requirements among and between modal flows according to timing and storage requirements.
There are two broad areas of decision support at the tactical level. The introduction of intermodal terminals in service design must be considered for their value proposition. That is, what value do terminals add in batching activities against the cost of handling and storage. Secondly, precision relations need to be established between carrier transfer pairs. This analysis involves balancing handling and storage resources for itineraries of different batch intensities to support carrier cycles
Page | 120
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems of certain headways necessary for system throughput goals. This addresses the central decision support requirement in intermodal terminals of the storage and handling trade-off.
4.1.3.2 Addressing Pulse Flow Phenomenon
Transfer terminals and transhipment nodes do not generally respond well to large variability in flows or high peaks. Discrete flows of train shipments represent large peaks. With sufficient handling equipment they can be accommodated at terminals better than truck peak flows with their requisite individual manoeuvring. As described in the introductory chapter and the case study on seaport-satellite terminals in chapter 8, gateway bottlenecks at the seaport are impacting landside capacity for growth in both Sydney and Melbourne. As shipments increase in frequency and lot size, an efficient means of moving container blocks in and out of the terminal are required.
Train flows between terminals may be characterised as pulses (also known in signal theory as pulse trains). These services have attributes of batching magnitude and timing (which may or may not be an adherence to schedule). In connecting terminal satellites with seaport gateways, for instance, these pulse services can have considerable productivity benefits to the seaport in increasing the consignment thoughput/slot reservation at the seaport. Container dwell time is reduced by more timely arrivals and train pulses can provide alternate access to the seaport thus hub bottlenecks via road access can be relieved.
The intermodal terminal must be responsive to the schedule of the train pulse. A modelling method is required that can characterise such flows and assess the capability of the terminal to meet such schedules through their handling, consolidation and storage processes for the inflows anticipated.
4.1.3.3 The Role of a Load Following Response Mechanism
Discerning the role of the load following response mechanism leads to an understanding of what the impedance functions should represent. For an activity based formulation of freight task, development of a load following mechanism is proposed which is based on electrical circuit theory and power distribution systems. The load following response represents a control of generation. It is
Page | 121
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems an ancillary service to support dispatch and scheduling of generator units (LUPs) in the intermodal production system. This control must provide terminal and system flexibility given that:
1. Loads change throughout the day (though a proportion of this can also be anticipated); 2. These changes must be allocated to the generating units; and 3. There needs to be a coordinated response.
The load following response represents a controller-observer mechanism which registers changes in load and according to the effect on its timing band rating, picks up this load.
The impedance functions we develop here are intended for further incorporation into a system operated by Automatic Generation Control in order to achieve system optimal control. This is conceptually and theoretically discussed in Appendix F.
4.2 Desirable System Format of Combined Transportation
Key characteristics of intermodal operations which require integration into a model formulation were identified in Chapter 3 and cover:
1. An activity based approach for design of terminal node impedance; 2. Stability and controllability; 3. Terminal flexible capacity.
The time composition of freight can be considered to affect the decision to route flows through certain transhipment nodes. This node-centered activity has precision requirements, not normally measured in graphical network models. Flows must be rationed according to timing behaviour. Capturing the time composition of freight is the new fidelity and tractability criteria to assess the performance of intermodal production systems.
To make the analysis of urban intermodal production systems meaningful, the planning tool must consider the interface among performance attributes of flexibility, throughput capacity and stability.
Page | 122
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
The kernel of the intermodal production system format is the Logistical Unit Process (LUP). This is a combination of physical processes and business rules. LUPs can also be considered as signal processing units as they use information about the flow: the consolidated nature of the goods, the consignment rate, the timing and information about the load they must meet in order to define performance such as cost, resource consumption, and level of dispatch utilisation. The itinerary taken by a consignment carrier pair through a terminal will traverse multiple LUPs.
4.3 Business Rules and Their Application
The success of an urban intermodal terminal cannot only be denoted by its utilisation. Performance also depends on how terminal operations are connected into a broader system. System utilisation then extends to cover the interactive processes between terminals and gateway networks. This expanded concept of utilisation has implications for how a sequence of flows (usually hard-wired flow paths), can become an itinerary. Modelling system utilisation requires the incorporation of open access regimes which support a terminal’s:
1. Logistical link in the supply chain; 2. Operation as an efficient terminal; and 3. Physical connection to various means of transport.
A representative model needs to incorporate different business rules in the operation of the terminal which affect its performance. These business rules are operational activities that change the response of the terminal system to different flow patterns. These business rules are transient and longer term. A useful extension of this response analysis would be considering back-end loads i.e. how the output requirements guide design configuration and operation.
These business rules may include:
1. Bulk queue control laws, 2. The management of storages, 3. The operations of rail sidings,
Page | 123
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
The management of empty containers pertains to the quantity held on site; when and to what extent they are repatriated either to shippers in the hinterland or back to the gateway node such as a seaport.
Gateway network business rules impact on the operation of intermodal terminals and can include:
1. Track access windows to the seaport21; 2. Operating hours of seaport terminals; 3. The configuration and operation of receiving and dispatch sidings; 4. Marginal handling costs at the seaport compared to road based transport; and 5. The requirement to repatriate a certain proportion of empty containers periodically.
It is critical that a model capture business rules of a terminal and the wider gateway network in which it operates. The implementation of business rules demonstrates the flexible capacity of a terminal and thus link the economic with engineering decision underpinned or constrained by the existing asset. These business rules extend to both terminal operations (the production) and dispatch decisions (the transportation cycle).
The development of business rules and their control mechanisms provide a means of incorporating the stochastic nature of freight logistics. Daganzo notes (1995) that “ as production is driven by consumption, the destination will request deliveries so that its inventory level can sustain at all times the demand that is anticipated”. In an exporting situation, the seaport acts as a drawing force to goods and consignment logistics production in the hinterland. Daganzo also notes that as there are stochastic variations in inventory control due to uncertain travel times and unpredictable demand, control triggers are required. These control triggers can allow stochastic effects to be incorporated into deterministic systems. Such a deterministic model formulation needs to consider the activation of control triggers.
21 Track access windows can be defined as clear mainline access as well as port-precinct docking access and availability. Page | 124
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
4.4 Positioning Elements of the Design Cycle for Model Development
4.4.1 Overview
Once the dimensions of intermodal production system assessment have been scoped, the task of developing a functional requirements outline for sketch planning can be given. The sketch planning endeavour can now be linked to the design cycle. For control systems, Dorsey (2002) denotes six stages: 1) Modelling; 2) Identification; 3) Analysis; 4) Design; 5) Simulation; and 6) Implementation. Modelling is concerned with finding the structure of the mathematical formulation; parameters are obtained from the identification process. The remaining stages deal with designing a control strategy, from preliminary testing to implementation. The remainder of this functional requirements chapter deals with describing both the essences of terminal operations that are necessary to capture, as well as the gateway representation of the network with which these operations must interact. Model identification requirements are recognised through this process. It is recognised that in exploring a new science theory, much work needs to be expended on structuring a prototype model formulation.
4.4.2 Attributes of Continuous Approximation Approaches
In modelling intermodal production systems, consideration must be given to many discrete operations that act to form a non-continuous system. We wish to formulate a continuous approximation of these operations as steady state phenomena. The challenge is to reconcile the interface between network flows with complexity transhipment points which are arguably discrete event systems. Continuous Approximation Models have been used in freight logistics (notably Daganzo, 1995; 2003) to characterise the interaction between the information logistics realm and the physical transportation network.
Their salient attributes are: 1. A drive to understand relationships towards better operating logistical networks rather than simply forecast future flows over an existing network;
Page | 125
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
2. Simple relationships using canonical forms such as geometry to model crucial system phenomena; and 3. Application to transportation fields where data is difficult to acquire and parameters difficult to calibrate Their key benefits are: 1. Minimum use of data input 2. Relatively small number of parameters required to ease verification 3. Yield insights into physical network and logistical (service network) relations for strategic planning
Consequently these methods provide an intriguing framework to apply to sketch planning intermodal production systems. What is needed is a robust science theory to support development and application of canonical forms of terminal impact or impedance with different flows.
4.4.3 Probing Logistical Relationships through Analogy
The novel relationships proposed to be investigated under an activity based approach and terminal impedance functions, require a new way at looking at how freight modelling and planning considers knowledge. The need to assess terminal operations against flow profiles in terms of value and precision require the application of new science theories to develop the discipline. Analogy is a powerful tool in such inductive reasoning (Polya, 1954). Four phases may be discerned in building a respectable analogy:
1. Making suggestive contacts by establishing conjectures that address the problem definition; 2. Testing the conjectures with supportive points of contact by mapping to specific relations; 3. Sketching parameters and decision variables ; 4. Establishing a model formulation through patterns of plausible inference (Polya, 1954)
This represents the preliminary phases to the model formulation stage Dorsey (2002) recognises. Analogy to electrical circuit theory is proposed and pursued in the later chapters as a feasible method in sketching intermodal impedance relations.
Page | 126
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
4.5 Conceptualisation of Intermodal Production Systems
4.5.1 Introduction
To understand the operation of freight terminals in the hinterland, we define them as interlinked industrial production systems. This comes from the conceptual model for intermodal terminal systems developed by Chalmers University, Gothenburg (Woxenius, 1998). The definition of freight terminals as production systems invites a process approach for the analysis of their operations. This process approach can be characterised as a Multiple Input Multiple Output System (MIMO) which is affected by a heterogeneous feedstock. Furthermore, intermodal terminals and their production systems should be considered as open systems (Muskin, 1983). They operate with many inter-dependencies. In this sense they are complex. Analysis of their supporting infrastructure helps the planner to determine what compensating mechanisms are required for bridging the gap in frequency, capacity and time (Styhre, 2005). Model reduction of this openness is valid to the extent that new insights in flexible capacity measurement may be obtained that drive new configurations and practices.
4.5.2 Gateway
A gateway is a pivotal node that links networks. In this sense a single transport assignment of carrier pairs between beginning and end point can be made. The focus is now on the transformation impedances at nodes as the consignment crosses modal networks. The concept becomes crucial in the assessment of freight production systems and the relations among terminal nodes that may involve primary and secondary container flows, pre-containerised, consolidated flows and ancillary flows. Most studies and models neglect the effects on pre- and post- haulage. The gateway approach enables a consideration of this activity, based on the functional arrangements of the gateway terminal, interfacing with other networks (Figure 21).
Page | 127
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
Figure 21: Demonstrating Accessibility Improvements and the Feasibility of Intermodal Terminals Requires a Gateway Analysis to Link Networks (Styhre, 2005, Fig. 7)
The gateway approach allows the effects of terminal activity to be more closely linked with upstream and downstream terminal connections. This allows the planner to assess the functional role of a terminal in a variety of hub forms. The application of this concept to a model is that the mathematical description of inter-terminal relations can be defined in terms of loads and this has direct requirements in intra-terminal operations. Link impedances between terminals can still be included with use of transportation lags. It is the transhipment impedance of consignment carrier pair transformations (un/stuffing, storage, batching, and transhipment) which now becomes more transparent.
Page | 128
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
4.5.3 Inter-nodal
This level refers to the directly connected vertical and horizontal production systems: that is up and downstream processes as well as side-stream states that channel the re-positioning of mobile resources. It is at this level, for instance, that staging posts, for transient use are included, when there is a mismatch of hours among terminal nodes. An indicator at this layer is the circulation time of the re-positioning resources.
Aspects of how terminal production systems harmonise:
• Bulk queuing –dispatch and arrivals depending on a minimum load available (part of queuing discipline), • Precision of synchronised resources – the possibility of delivery trucks making pick-ups on the return trip from the terminal, to improve backloading for instance, • Re-positioning resources – the circulation time of empty containers, vehicles, and train sets.
4.5.4 Terminal
The process flow chart for an urban freight terminal (consolidation/distribution) or intercity terminal (gateway node) is very complex (Figure 22). There are a number of characteristics: • The freight profile entering can be heterogeneous- i.e. frequency, rhythm (cluster arrivals); the container type (reefer, 20 ft, 40ft) or the consignment may not be unitised; the destination; and the commodity type; • There are parallel processes at work which may interact and be inter-dependant; • External resources are required to interface with the terminal in order to facilitate the transhipment and translocation process. In this way, the prime flow, calls other resources to be positioned; • The inter-operability of this system is dependent upon the level of equipment technology, and information transfer.
Page | 129
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
Figure 22: A Transit Terminal for Inter-Intra-City Transhipment (Morlok, Figure 7-4, p.254, 1978)
The focus of this thesis is on the interaction between storage and necessary function in the freight production system in order to control the impact of peak demand for freight services. The demand for freight services is seen as the demand for containers.
The measure of the feasibility of intermodal production systems in this thesis is based not only on the locational issues of sufficient flow volumes (such as the bundle type studies cited above) but on how the productions systems match the storage arrangements which control the demand for freight services. The engineering measure is terminal impedance matching and this links the functional requirements of the terminal with the logistical sequences arising from the distributed storage procedure. In this way the planner may ensure that intermodal production systems act to reduce freight flows.
A means to explore management of storage arrangements at a terminal is through reference to cumulative curve theory alluded to in section 4.1.2 and further explained in Appendix B. A critical control mechanism is the duration of the active arrival period in which consignments are expected for a particular discrete dispatch service (such as a train). This corresponds to the average dwell
Page | 130
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems time measure of containers at a seaport. Control of this mechanism represents dynamic assignment and can be used in stack and siding management with multi-commodity flows.
4.6 Node: Critical Terminal Operations and Desirable Characterisation
4.6.1 Areas of Performance Measurement and Design of Terminal Interfaces
4.6.1.1 Introduction
In this section critical state variables and operating parameters are established to characterise an intermodal freight terminal as a nodal gateway in a freight network. The following section deals with gateways (inter-node). This will lead to an understanding of what the model needs to identify.
Transhipment intensity is a measure of the time sensitivity of freight and certain business rules in operating a terminal which support this.
For urban intermodal terminals under conditions of increasing tense flux, precision and stability are critical figures of merit. Transhipment intensity information in the flow signal affects how the terminal responds. The performance of the sub-systems which make up the freight terminal can be rated at different transhipment intensities. The terminal can be configured such that it performs best at a certain transhipment intensity band (range of synchronisation values).
4.6.1.2 Logistics Performance
Logistics performance considers the interface between the terminal and other connecting nodes, such as other inland terminals, or gateway hubs such as seaports. There is a need to measure this performance at the sketch planning level in order to assess viable networks. In essence, viable networks of freight modal interchange are those that meet desirable specifications on the basis of
Page | 131
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems precision and stability22. An interface typology (or hub typology), can be used to relate terminal operations with rail operating forms. This means that the mode service frequencies and other modal flow dimensions (of rail and of truck) match with the complex activity of the terminal operation. That is, the path of impedances through the terminal. By this characterisation, we can get an integrated view of the itinerary through a terminal as a path of availability and dispatch.
The varying quantum of four variables characterise the flow-terminal interface:
1. Freight value density, 2. Consignment flux, 3. Transhipment intensity, 4. Value intensity.
Freight value density tracks the mass to volume ratio of the batch as i) individual consignments and as ii) batches of consignments. Consignment flux tracks the flow of consignments through each unit process with the changes in value density. Transhipment intensity tracks the timing of freight consignments and is used to direct flows through unit process paths to match timing requirements of loads. Value intensity is a measure of the value being stored according to each timing band. It is used to assess the contribution of that discrete flow to the combined dispatching load.
4.6.1.3 Storage and Handling
Metrics regarding container terminal capacity and productivity are often based on queuing model formulations of freight terminals. Ferreira and Sigut (1993) summarise key performance ratios including: number of lifts per container, and time taken per container lift to designate the handling productivity. Seaport productivity is often assessed on the storage capacity based on throughput/ storage area (often number of slots). The storage area in metres squared ( m2) can be used as a base measure rather than the measure of number of container slots. By this area measure the throughput (flux) productivity can also recognise increased accessibility problems with stack height. Stack
22 There has been little work in assessment of viable hub networks. Benefits have been characterised by Daganzo (1984) Hub networks have peculiar viability criteria. Page | 132
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems accessibility can be assessed by the additional shuffling of containers required to access desired containers.
Characterising the storage-handling buffer is the essential interface in a model of a freight terminal. Storages act as buffers in balancing inflows with outflows. The buffer storage can be defined in terms of slot occupations (average quantity and duration). This corresponds to measures in differences between cumulative arrival and dispatch curves. The rate of change of the buffer storage depends on the handling rate of the transhipment equipment.
Storages are characterised by:
1. The slots that are reserved for a carrier pair for a specific dispatch; and 2. The slots that are utilised for a carrier pair for a specific dispatch. The difference under alternate flow regimes, terminal configurations and strategies will affect the storage size and throughput productivity of unit processes.
4.6.1.4 Stack and Sidings
Sidings represent a critical interface between the road and rail network. How they are configured in the terminal and the method they are managed by terminal operators will affect the ability to service train schedules within the tight windows of the urban network. Rail shuttles, for instance require rapid turn-around and potentially require the capacity to load one train whilst un-loading another. This necessitates a hardstand with multiple active siding faces and has implications for transhipment resources i.e. a gantry or portal crane rather than a reach stacker or forklift. Here the classification task is not required as the shuttles are in fixed units. In Melbourne and Sydney, intermodal terminals are being retrofitted on old industrial sites with track leads to the main line. These often have multiple sidings which can act as valuable storage of railcars and loaded consignments. In these cases classification activity may occur, particularly if the intermodal terminal acts as a gateway to other rail networks in the hinterland.
There is no direct mathematical formulation of stack–sidings interaction to assess impedance and the precision capabilities of different configurations and uses. This interface needs to be incorporated into a model formulation of urban intermodal terminals that considers sidings Page | 133
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems availability based on current utilisation and future assignment. Rail freight schedules need to be incorporated into sidings occupation calculations.
4.6.2 Model Identification Requirements
4.6.2.1 Logistics Cost Structure and Value Stream Mapping
The logistics cost structure has been outlined by Daganzo (1995). These costs are grouped as motion and holding costs. Motion costs include handling and transportation costs. Holding costs include rent and waiting costs. Deterministic relationships can be made with headway, shipment size, holding and handling costs, and travel times. Cumulative cost curves of a “production” node and a “consumption” node demonstrate mathematically these relationships in a two zone – two node problem. The distribution of system item-hours can be gauged and altering this distribution affects the cost structure. Motion and Holding costs have both fixed and variable components. Shipment generally needs to balance inventory costs with transportation costs. Storage levels and associated inventory costs which control headways is considered a significant cost activity to be managed.
It can be seen that any useful model of intermodal activity must be able to formulate such a cost structure. Whilst handling costs may be seen as relatively small, the wait time effect may be considerable. Additionally, when considering the management of diffuse- discrete flow sequences (such as trucks from many origins to single destination dispatch trains) the need to synchronise shipments is essential. This is the key criteria in the successful performance of urban intermodal terminals. Space management aspects that foster or hinder this synchronisation need to be considered in any model formulation.
The performance of an intermodal terminal must be seen in relation to the seaport or other critically linked nodes that it serves. Value Stream mapping seeks to measure the marginal benefit the addition of a unit process function makes to freight system costs. Marlow and Paixao (2004) identify this process is essential in assessing the combined productivity of a seaport and its landside hinterland. Activities and tactical scenarios that improve system space management on both mobile and infrastructure resources with a view to improving the load capability of the network need to be measured.
Page | 134
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
Ballis and Golias (2004) present an insightful Figure of Merit in a context of a terminal linked to a specific rail operating form. This Figure of Merit assesses the trade-off between the variable cost/TEU to TEU throughput of different intermodal terminal systems at different capacities. These terminal systems involve terminal size and layout, available infrastructure including rail sidings, and stack and rail handling equipment. They show that different intermodal terminal systems have an optimum operating range for certain levels of annual throughput (Figure 23).
Figure 23: Marginal Cost with Throughput (after Ballis and Golias, 2004)
In reality, not one intermodal terminal type can accommodate all throughput levels so each terminal represents a partial band on this spectrum (dashed lines). This Figure of Merit can be used to identify what combination of terminal configuration can best address different throughput levels, at the least marginal cost. In the context of urban intermodal terminals a valuable extension in investigating this relationship would be the flexibility of the terminal to respond to different time window- synchronisation information of consignments as well as physical fluxes.
4.6.2.2 Storage and Space Management
There are two critical measures of space: space utilised and space reserved. These measures are constituted as follows:
• Space utilised= storage lag x Value flux (can be in total value stored or total containers), Page | 135
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
• Space reserved= timing band x Value intensity. These measures of space are significant in space management as they assist the planner in assessing how business rules can affect the space required to meet scheduled services and act as a means to improve precision and stability. The value of any sketch planning tool should be on how it can address space management issues.
This model formulation uses two domains to assess space management performance: the time domain and the frequency domain. The time domain expresses time continuity information and the frequency domain expresses resource consumption.
4.6.2.3 The Bulk Queue
The bulk queue is an extension of queuing theory whereby flows accumulate and are released to the next unit process by a control law which triggers the change. Its constituents are the flow arrival pattern and the server activity. This control law is either time based or quantity based. The purpose of modelling the bulk queue is to represent authentic terminal dynamics according to flow types, resource constraints and the operation of business rules to achieve precision of dispatching services and maximise the dispatching service capacity. Significantly there may be cost trade-offs in utilising the dispatch service to its full extent and delaying dispatch until there is sufficient accumulation.
The bulk queue is a critical phenomenon in characterising terminal dynamics. It is modelled by continuous approximation approaches or by discrete event mathematics. The continuous approximation is based on a modification of Markov Queuing chains. This method however has limited closed form solutions in its application to freight terminals (Crainic, 1987).
The model formulation of this thesis considers an alternate approach to the bulk queue problem using a frequency domain solution. Here the flow is passed once a certain value intensity is reached. This involves a number of source flows accumulating within a certain timing band, thus raising the intensity level which triggers the control law. By this method, the planner may consider any periodic flow for a closed form solution.
Page | 136
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
4.6.2.4 Multi-path Coupling
An itinerary path through a terminal is based on both precision and controllability. These attributes dimension the capability of a terminal. The capability of a terminal goes beyond the constraint notion of capacity measurement and is measured by the effectiveness of applied compensation control mechanisms permissible under certain terminal layouts.
In capturing the complex dynamics of terminal operations we wish to include flow couplings through unit processes within the terminal. Terminals are Multiple Input – Multiple Output hybrid facilities. There are potentially several paths that can be taken by consignments through the terminal and this affects the costs of transhipment as well as the value added by consolidation. Designing coupling activity can lead to undesirable bottlenecks within the terminal. It can also lead to improvements in flexible capacity and ability to consider the terminal as a multipurpose facility (Daganzo, 1990). Capturing the workings of terminal complex activity is a critical mechanism in the model formulation of this thesis. A prerequisite for gaining benefits from modelling multipath coupling is the inclusion of multi-commodity flows.
4.6.2.5 Multi-commodity Flows
The multi-commodity flow problem is an extension of the transportation science modelling solution over a network to consider flows with a range of origins and destinations. The extension to incorporate multi-commodity flows in the model formulation allows complex bundling forms serving train service loads to be considered. Complex bundling is the consolidation of different consignments into train load batches. Dealing with complex bundling offers opportunities for urban terminals to generate sufficient flows to meet train service frequency requirements and consequently operate feasibly. The feasibility of an urban intermodal terminal relies on capturing a proportion of consignment flows to the port which would otherwise go by the road network. For sufficient flows, several generators and attractors need to be sourced as it is unlikely in the urban area that there are a small number of very large shippers who would generate sufficient flows (SFC NSW, 2007). Additionally, storage savings from prioritisation will increase if consignments destined for different vessels are included (Taleb-Ibrahimi et al., 1993)).
Page | 137
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
4.7 Gateway: Critical Link and Node Operations and Desirable Characterisation
4.7.1 Link Time Windows
Urban rail freight is acutely constrained by access to mainline links to the seaport and other destinations. Terminal operations must support the train schedules devised to fit time access windows to the mainline. This condition adds to the tense flux of terminal operations. Precision assessment is necessary to determine whether the operations are feasible.
4.7.2 Prioritisation and Inter-Node Stack Availability
In linking land based consignment flows to container vessel dispatches, there is an active arrival period, whereby containers destined for a specific vessel accumulate in a reserved area at the seaport. With a long active arrival period, there is a pronounced excess of storage capacity which remains underutilised. Containers on rail may be prioritized by only those containers with impending vessel departures loaded. This may lead to a reduction in the size of the storage area at the seaport. Of greater significance is a test of whether throughput can increase with the same storage size. This prioritisation may also benefit from the pulse nature of freight movements by rail. This is so because dedicated stacks and stack segregation strategies may then be used. These strategies can reduce the shuffling time handling equipment to find desired containers.
4.7.3 Security Functional Requirements
The sequence of flows through an intermodal terminal should be modelled as an itinerary which is guided by a number of criteria including functional requirements of the consignment and timing. Because we are dealing with the node as an activity centre for other link and node connections, operating limits and boundaries are considered within the scope of system infrastructure availability. Unit Commitment is a filtering stage prior to scheduling which assesses the availability
Page | 138
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems of units at each period. In hydrothermal coordination, unit commitment constraints can include operating minimum and maximum levels, shut down and start up time and reserve requirements. Unit commitments are best included early in the algorithm in order to ensure operations are secured. That is, there is less risk of overloading unit capacity and critical transmission lines. The itinerary formulated becomes a path of commitment (availability) and dispatch (utilisation). This addresses the gap identified in Chapter 3 of harmonising service characteristics with itineraries and such tactical modelling responses in service network design which are iterative.
4.8 Impedance Attributes and Specifications for a Urban Intermodal Terminal
In chapter 2 a hierarchy diagram was presented, describing the constituents of harmonisation of the container freight task. In this section we go into detail regarding Figures of Merit and specifications. For design purposes, these harmonisation goals can be translated to desirable terminal performance through defining the load following requirement of the terminal service through a number of impedance specifications. These specifications define the load following outcome desired. The impedance specifications map to a desirable coordination of excitation- response relationships: carrier pair flows through unit processes. These relationships are designed, measured and controlled using impedance attributes. The characteristic impedance of each carrier pair through each unit process is conveniently captured in a transfer function. In order to test whether this system has met the load following requirement, working Figures of Merit are applied to observe the results and if necessary, to adjust results to meet the load following outcome. By this mechanism we have a means to relate the terminal operations with the rail operating form requirements that interface the terminal.
Specifications for a “close” urban intermodal and transmodal terminal need to include the following requirements:
1. Batching, 2. Storage (inertia from space consumption), 3. Cost-Value Ratio of operations, 4. Timing matching, or precision, 5. Terminal productivity assessing throughput-storage capability i.e. mass/m2/day, Page | 139
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
6. Flexible Capacity.
In this section both the novel specifications and Figures of Merit have been outlined in the context of identified terminal operations which can capture the phenomena of interest. These performance indicators act as the guiding points for a model formulation which proceeds next. As part of the functional requirements there must be the ability to scale the analysis according to time frame – from impacts of short term fluctuations to impacts over the steady state.
4.9 Relating Objectives and Case Studies Testing
4.9.1 Objectives
This functional requirements outline has further scoped the objectives arising from the problem drivers presented in chapter 1, Figure 1. The overarching objective is to position intermodal operations to control a dissipative logistical structure and to “minimise the evolution of system deficiencies”. These supporting objectives were to assess the feasibility of developing:
1. An activity based approach using terminal node characteristic impedance to assess precision and controllability specifications; 2. A measure of the value proposition of consignments taking itineraries through terminals; 3. A means to assess hinterland absorptive capability and thus relieve seaport scale capacity pressures; and 4. A means to test effects of governance and operating business rule levers on terminal and system capability.
The functional requirements have conceptualised urban intermodal systems for conceptual validity of activity measurement, flexible capacity and controllability. Significant state variables assist in enumerating terminal responses. Specific attributes of bulk queue control, stack–handling management, sidings control, multi-path coupling and addressing the multi-commodity flow problem have further been outlined.
Page | 140
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
The case studies may be seen to illustrate that the functional requirements have scoped the key phenomena of interest that need to be enumerated and explained.
4.9.2 The Container Hinterland Case Study
An analytical approach to the assessment and design of distributed interfaces with a case study in Sydney, Australia is presented. The focus of our investigation is the implications of dispersing seaport precinct functions, with the port remaining the dominant gateway node. Specifically, we consider the configurations of direct shipment, centralised intermodalism, both adjacent to the port and within the inner city, smaller terminals of distributed function, and hinterland terminals of functional clusters. Each of these configurations alters the nature of freight visitation calls and has their own requirements for ancillary infrastructure. Functions such as consignment storage, consolidation and distribution, stuffing/unstuffing, intra-modal and intermodal transhipment, empty container repositioning, and resource vehicle storage are evaluated with different hinterland terminal configurations. We assess the implications for distributed functions in terminals in order to consider the opportunities for a more coordinated nodal freight transport infrastructure. This provides insight into the challenge to locate terminals on the basis of functional as well as spatial issues.
The task is to assess the effect of dispersing functions from the seaport precinct in order to understand how the seaport gateway and hinterland distribution infrastructure may be further integrated. In turn the system is assessed for the changing levels of impedance that arise with the distribution of functions in the hinterland. The wider production systems supporting intermodal operations are investigated to assess hinterland capability and give direction on desirable retrofits.
The objectives are:
a) Investigate the requirements to operate a terminal configuration of distributed functions to support the international container distribution task. Consequently we detail the concept that the location decision of terminals becomes a condition of functional attributes as well as spatial features; b) Assess how the terminal hubbing effect may be controlled with development of terminals of distributed function;
Page | 141
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
c) Assess how freight infrastructure can be used to develop bundling networks of small freight flows; d) Developing indices that can be used to assess and communicate planning decisions involving the location and function of terminals in the container hinterland.
4.9.3 Integrated Waste Management: Storage and Streaming
The intention of the case study is to illustrate how an intermodal waste management system could create resource recovery opportunities. This considers that the re-configuration of the transportation system can act as a precursor to more sustainable forms of human organisation.
Initially the densification opportunities in solid waste management are considered. Containerisation of wastes can reduce the volume processed considerably. Secondly, opportunities to increase resource recovery by a combination of collection, storage, streaming and containerisation in the waste network are investigated. This is facilitated by the intermodal network. Any transformation activity needs to be considered in the value that it potentially recovers. This is another approach to assessing the cost-value jump of intermodal operations.
The recovery of wastes represents a new function in the anthroposphere. We investigate how the intermodal production system can lead to better resource recovery. We take a physical and logical perspective here to investigate what are the performance requirements of an intermodal system with respect to the transformation processes in the transit process: storage, streaming.
The objectives are:
a) Consider the results of waste terminal operations using system dynamics to understand broad critical path issues and limitations in the system dynamics application; b) Describe the application of the transhipment calculus to different aspects of unit processes at a waste intermodal terminal; and c) Describe system formats involving combined production and rail transportation cycles and outline how their distinctness may be modeled by the transhipment calculus;
Page | 142
Chapter 4: Functional Requirements for Sketch Planning of Urban Intermodal Production Systems
4.10 Chapter Conclusions
This chapter outlines the requirements of an analytical procedure of freight transportation networks to assess and minimise terminal node impedance and thus foster networks with feasible urban intermodal operations (i.e. rail-road) in the container distribution industry. This involves considering the complementary nature of linked terminal production cycles (i.e. for consolidation terminals and intermodal transhipment terminals). In turn the complementarity of temporal characteristics needs to be analysed. The location-size decision is based on relative functions and it is these functions that need to be assessed for design of more integrated systems. Therefore the portability of a functional analysis is desired. It is hoped that this will reduce the ad-hoc nature of distribution terminal infrastructure location and function and thus contain traffic pressures. As anticipated, the productivity of seaport will increasingly depend on the performance of the hinterland. A functional-process approach provides a research basis for the multi-modal process control of the seaport-hinterland (Marlow, and Paixao, 2004). The contribution to knowledge of this thesis is an analytical framework to address value proposition and precision relations criteria in sketch planning design of urban container intermodal terminals within their role in the hub transportation system.
In any new model formulation for intermodal terminal relations there is a need to make explicit the time and space composition of freight through a terminal under different flow and handling regimes. In the following Chapter the science theory of electrical circuits is presented for this task. This science theory offers a metaphor of conceptual validity and an analogue of constitutional laws which support a novel measurement of impedance relations. This model formulation is discussed in the next chapter.
Page | 143
Chapter 5: Model Formulation of Intermodal Production Systems
5 Model Formulation of Intermodal Production Systems
5.1 Introduction
5.1.1 Chapter Guide
Chapter 4 characterised intermodal operations and detailed requirements to be met in devising a tool for sketch planning. This chapter explains and justifies the model identification approach to meet these functional requirements. The analytical relationships of terminal impedance are developed in this chapter using the analogy approach. Electrical circuit theory relationships are introduced and are mapped to logistical processes.
A new physics of freight terminal operations called transhipment calculus is proposed. Electrical circuits can act as stencils of Logistical Unit Processes (LUPs) to characterise terminal impedance- precision measures and several relevant phenomena are described. Whilst based on established mechanisms in both electrical circuit and power theory and process control engineering, the application to transportation systems is novel.
Alongside precision, application of business rules to assess flexible capacity is a key attribute in the new transhipment calculus. Design aspects leading to compensation control mechanisms are outlined. Compensation control act as the ancillary services presented in the conceptual model of retrofitting systems (Chapter 2, Figure 5). This mechanism offers a means to consider the flexible capacity of the terminal (also known as capability). The control mechanism incorporates precision- impedance dynamics characterising terminal functions. Stability also allows us to assert the applicable boundaries of operation of terminals.
As the physical basis to the model formation proposed in this thesis is novel, the method needs to be carefully elucidated.
The tool development process consists of the following seven steps:
1. The concept and analogy is established; 2. The phenomena, design variables and Figures of Merit are formulated;
Page | 144
Chapter 5: Model Formulation of Intermodal Production Systems
3. The relevant mathematics from electrical circuit theory and process control theory are explained; 4. The mathematics is demonstrated; 5. A rudimentary simulation of the mathematical form is made; 6. The mathematics is refined for variability in parameter settings and feasible operating ranges; 7. The sketch planning capability for this mechanism of the tool is formalised.
This thesis deals directly with the first five steps. Chapter 4 presented the suggestive points of contact with the analogy. Chapter 5 covers steps 2 and 3 in the tool development process. Case studies in Chapters 6 and 7 illustrate steps 4 and 5 by prototyping individual mechanisms. Further work is scoped in the concluding chapter for investigating plausible patterns of inference using the circuit analog: determining the operable regions of parameter settings (step 6) and verification procedures to evaluate the physical realisability of the analogue model to intermodal logistics planning (step 7).
To make the discussion straightforward, the examples of terminal operation and response in this chapter are presented from the perspective of exporting consignments arriving as trucks, undergoing transformation at the terminal and then being dispatched on a train service. Complexity associated with multiple direction flows is sketched as well.
5.1.2 Addressing the Functional Requirements
The functional requirements outlined in Chapter 4 may be summarized according to a number of criteria and corresponding attributes that need to be characterised to achieve the sketch planning objective (Table 5).
Page | 145
Chapter 5: Model Formulation of Intermodal Production Systems
Table 5: Criteria and Attributes to Address in Intermodal Sketch Planning Functional Requirements
Functional Requirements Criteria Attributes developed through the electrical circuit analogue
Pulse flows Flow rate; Value Density; Characterise the time composition of freight for transient and steady state cases
Space Management Complex impedance; storage and handling; costs; resource consumption
Linking Terminal Operations with Impedance Matching of Logistical Unit Transportation Scheduling Processes for precision; Bulk Queue
Multi-commodity flow problems Signal carriers; Value intensity
Multi-path coupling Characterise impedance according to terminal complex activity; transhipment intensity
Complex Bundling Multi-path Coupling with Multi-commodity
Flexible Capacity Compensation control
Operating Business Rules Filters of itinerary control; bulk queue control using synchronisation band to proxy the active arrival duration
Figures of Merit Cost-value ratio at terminal; Capacity-Risk
It is the role of chapter 5 to identify and explain the specific mechanisms to address these criteria.
Table 6 is presented for rapid crossing referencing of the particular terminal logistics phenomena captured and modelled. This table maps the attributes of Logistical Unit Processes (LUPs) from the functional requirements defined in Chapter 4, with their circuit analogues described in Chapter 5 and with the prototype examples demonstrated in Chapters 6 and 7.
Page | 146
Chapter 5: Model Formulation of Intermodal Production Systems
Table 6: Mapping Logistical Unit Process Attributes from Functional Requirements to Mechanisms and Case Study Illustrations Section LUP Attribute Section Analogue Mechanism Section Illustration
4.1.3.2 Flows and Information 5.3 Flow Signal Characteristics 7.6.2 Pulse train forms described Clyde Attributes 5.2.2 Sources and Loads 5.2.5
4.2 Impedances at terminal in 5.2.4.2 Laplace Transform response to nature of flows and terminal configuration Impedances and Value Accounting System 4.6.2.1 Logistics Cost Structure
4.6.1.2/ Storage and Handling: Batching 5.4.2.3 Complex Impedance and Polar 7.5.2 Dumping floor – compaction dynamics- 4.6.2.1 Operations, Cost and Resource Accounting Clyde Consumption through Impedance Storage-Handling Interface as 5.2.5 Resonance
4.1.3.3 Precision and Impedance 5.2.5 Sources and Loads Matching Resonance as Equilibrium Operating Capacity
4.1.2.2/ Precision and Transhipment 5.2.4.3 Filter mathematics and the 7.5.3 Itinerary control and Multi-path 4.3.2.6 Intensity Frequency Domain Coupling- Clyde
4.6.1.3 Stack and Sidings and Flexible Feedback control 6.4 Bimodal overflow, Sidings Capability - Capacity Dandenong 5.4.5 Clyde description 7.5.4
Page | 147
Chapter 5: Model Formulation of Intermodal Production Systems
4.3 The Bulk Queue and Business 5.4.2.8 Transient switching Rules 4.6.2.2 5.4.6 Filters
Multi-Path Coupling: Routing 5.4.3.2 Filters as Itinerary Mechanism 6.5 Enfield Description Flows through an Itinerary 4.6.2.3
4.6.2.4 Complex Bundling and Multi- 5.4.4 Filters & Pulse Correspondence 7.7 Clyde system formats described Commodity Flows Fourier Analysis
4.7.2 Prioritisation 6.3.4 Altona Description
4.6.1.2 Technology Interface and Appendix Transient Analysis 6 Container Terminal Technology Applicable boundaries of B.3 Interfaces operation Stability: Design Specifications
4.2.4 Stability/ Flexible Capacity 5.6/5.6.3 Compensation Control 7.5.3 Itinerary Control- Clyde
4.7.3 System Security in Coordination Appendix Coordination Theory and Empty F Container Flow Problem
Page | 148
Chapter 5: Model Formulation of Intermodal Production Systems
5.1.3 Challenges and Limits of the Analogy
Some preliminary comments on the applicability of electric circuit and power distribution theory to urban freight intermodal system model formulation are warranted. Chapter 4 made some suggestive points of contact. There remain areas that are not so similar and may pose limitations in the application of the theory. For instance, electrical circuit theory does not map easily to the multi- modal, multi-commodity flow problem, electricity being fungible. Addressing this problem is vital when assessing opportunities and requirements for urban intermodal operations. Flow complexity, system capacity, and controllability under alternate terminal configurations and operating business rules are critical Figure of Merits. For the contribution of the electricity analogue to be of value such limitations must be addressed. In considering specific mechanisms in the transhipment calculus, this chapter makes some supportive points of contact of the circuit theory to freight transportation science. The purpose of this chapter is to further scope where the analogy may be justified. The key variables of value intensity and consignment rate are presented in their load implications on terminal unit processes. In Appendix A the strengths and weaknesses of the analogy are summarised. Verification procedures are highlighted for future work.
5.2 Terminal Operations Design Mapped to Electrical Circuit Physics
5.2.1 Overview
5.2.1.1 Circuit Stencil as Logistical Unit Process
This section outlines how freight terminal operations can be physically represented by close reference to electrical circuit theory. Key variables and activity parameters are introduced here for further development in the chapter. Following from the broad permissible assumptions granted in the functional requirements of chapter 4, assumptions are introduced which realize the benefits of an electrical circuit mapping. The section then expands to consider other attributes of electrical circuit physics which allow the modeler to theorize an activity based approach to urban intermodal terminal operations. These attributes are:
Page | 149
Chapter 5: Model Formulation of Intermodal Production Systems
1. Load Based Design 2. Physical Realisability criteria stemming from circuit physics to determine feasibility 3. Impedance attributes- covering value mapping, precision and stability/ control 4. Rating operating capacity of a unit process including storage utilisation 5. Business Rule levers 6. A mechanism to track the itinerary of consignment flows through a terminal
The circuit acts as a closed loop stencil which balances driving and driven forces over the specific logistical activity is represents. This activity can be transhipment, storage, bundling to a larger or smaller consignment, queuing, other transformations or loading onto a siding. Activity is a measure of value stream change and can thus represent a financial charge such as gate access charges, customs duties, demurrage, or payment of container rental, for instance. These costs can be fixed or variable.
The driving force, is considered the value density, denoted by V, and is in units of $/consignment.km and this is related to the freight trip intensity of tonne/kilometre. The driven force is the consignment flow rate, denoted by I and is in units of consignments flow/ unit time. Consignments are ostensibly container boxes, either TEUs or FEU23 but may also be standardised pallets.
For a certain value density flow, a consignment flow is induced with the impedance elements in the circuit. Load is the forward impedance signal for the relationship between value density inflow and consignment rate inflow. Their product is the value flux. In electrical circuits, the analogous measure of maximum power throughput is achieved with a conjugate. A conjugate is a mathematical expression of this load expressed as the driving and driven forces (Value density, V and consignment flux, I) which are carried and modified through each electrical circuit element (Nilsson, 2005). For the purposes of this analogy, each terminal unit process can be represented by one or a number of electrical circuits. This conjugate reflects the inflow signal. Conjugates are used in this mathematical formulation of terminal operations to match the terminal storage and handling response with inflow requirements. Complex conjugates are used to represent the design response of terminal elements to inflows which carry information on terminal handling requirements. This
23 Standardised container measures lengthwise: Twenty-Foot Equivalent Units (TEUs), Forty-Foot Equivalent Units (FEUs). Page | 150
Chapter 5: Model Formulation of Intermodal Production Systems
facilitates load based design introduced in section 4.1.3.3 and further developed as a method in section 5.2.3
The stencils provide a topological description of a Logistical Unit Process (LUP) which maps to a particular mathematical description. This supports the modularity and ease of use of the method as it allows canonical forms to be described and applied when creating one or a number of Logistical Unit Processes (LUPs).
5.2.1.2 Simplifying Assumptions
The electrical analogue allows a proxy relationship between system work and impedance for changing freight flow profiles. This provides the basis for a continuous approximation where it is not necessary to count the processing of each individual consignment, like in Discrete Event Simulations or actor-based simulations. This reduction means the modeler loses the ability to observe how a certain terminal layout can deal with asynchronous events. Where real time activity, including step changes can be surmised as periodic general insights can still be obtained.
At the terminal level, a major simplification is the use of lumped parameters. Lumped parameters are treated irrespective of their individual location or separation in a terminal. In the electrical analogue, lumped parameters allow the calculation of equivalent resistors in series and in parallel. An impedance component, which is the internal distance translocation, may be incorporated in the resistance coefficient. We can also, through the electrical circuit stencil (described below), identify what elements can be lumped, and what must be kept separate for our terminal analysis. Those elements lumped or kept separate may also interact i.e. load the system. Consequently, when terminal design is discussed, it pertains to broad relationships among transformation activity, storage size, flow profile and impedance and does not extend to the siting of particular elements within a terminal or the potential bottlenecks of intra-unit flows. This is a permissible modelling simplifying assumption. We are concerned with the physical feasibility of the terminal and how it processes inputs and disturbances and its broader role as a necessary transformation and impedance node in a network configuration.
Where there are multiple sources, linearity and the theory of superposition are assumed whereby the network can be analysed by allowing each source to act one at a time and then superimposing the results. Page | 151
Chapter 5: Model Formulation of Intermodal Production Systems
The assumption is also made that freight flows and modification of flows in an intermodal terminal can be represented by signals and that all the signal and signal processing of interest can be represented in terms of exponential functions. This assumption has a fair degree of modelling heritage: as Kulathinal (1988, p.14) notes, “Most of the input/output signals encountered in physical systems can be converted to exponential signals”.
5.2.2 Variables and Parameters and Balancing Laws
5. 2.2.1 Variables
True physical systems are a combination of driving and driven functions. The urban container freight task is characterised by the intermediation of numerous actors and although there is some coordination (time slots for truck arrivals at terminals), the system may be truly described as one of chaotic deposition (Tarski, 1986). Driving and driven functions are less transparent in such a system. However, once the essence of freight service activity is defined in terms of physical forces, meaningful abstractions may be made to grant modelling elements for more sustainable network and terminal design outcomes. The electrical circuit analogue provides a basis for formulating and assessing flow intensity and impedance variables in the freight network. Voltage acts as a value density (driving force) - it is the energy or work required to move a charge or consignment. From a freight perspective, the driving force of the bulk arrivals needs to be managed through the terminal, according to the output desired. The bulk arrival can be characterised by a pulse function. Current is the rate of transfer of the consignments. The size of our consignment, from a pallet to many containers in train, varies through a terminal. The designation of the consignment will determine the route and the required driving force. For instance, whether the consignment is to be further aggregated or disaggregated.
The value of the freight task is often difficult to quantify as there are many qualitative requirements including size of consignment, timeliness and desirable place. This might be summarised by the notion of precision. Power, a mass (value) flux, represents a value of delivery/ dispatch quality which can quantify precision aspects through an attribute pair of quantity and timing. Consequently we may characterise the cluster intensity of arrivals and how this affects the impedance-storage- throughput relationship at a terminal or series of terminals. This may be represented by an interaction of bulk value density (voltage) and transfer rate (current) at a prevailing timeliness Page | 152
Chapter 5: Model Formulation of Intermodal Production Systems
requirement. Power precision has three key aspects – the global frequency, local voltage and noise. The four step model has a similar rationing device based on both local impedances and equilibrating the system. System equilibration in classic transportation science is generally for continuous flows and has no consideration of the productive capability of the intermediate terminals.
Value Density is the value of the consignment over the line haul distance to be travelled. It is the payload. It is the direct inverse of the freight intensity (km/t) and therefore when other modes are compared, value density acts as a measure of the trips saved by batching operations (compaction and consolidation). Freight intensity has been identified in the literature as a critical KPI (or Figure of Merit) in tracking the energy intensity of freight logistics and its dissipative structures. The numerical example in section 2.3.2 demonstrated the significance of tracking the value density profile of a consignment.
Value density can be represented by a combination of:
1. Payload; 2. Order and coordination (priority shipping; storage and handling activities which increase the order of a system for upstream productivity gains will add to the value density); 3. Quality and contamination (composition of product shipped and its degree of value for its final use: i.e. the waste matrix). Activities which increase the end use value such as source separation for a compost product will add to the value density); 4. Relative emissions of competing carrier class modes; 5. A combination of above for each carrier class conversion pair24 that the terminal intermediates. The focus of the analysis in this thesis is in using the payload component.
The value flux through the system for different carrier pair conversions k to j can be calculated in equation (5.1).
= ValueFlux ∑{[ ]kj .[ dIV kjkj }.] …………. (5.1) kj
Where:
[V]kj is the value density in $/consignment.km,
[I]kj is the consignment flow rate (per operating period),
24 Carrier class pairs are those mode services between which the consignment is transhiped. Page | 153
Chapter 5: Model Formulation of Intermodal Production Systems
dkj is the link distance between each beginning and end carrier pair,
k is the incoming carrier traffic class (such as truck rated at a certain tonnage),
j is the outgoing carrier traffic class.
This initial analysis consists of a simple bundling concept of one Beginning-End pair. There is no change in link distances between incoming flow signals and there is only one traffic class conversion. When this thesis considers different hub configurations and up to a many-many mapping (Multiple Input- Multiple Outputs), different value densities per consignment class can be incorporated.
5.2.2.2 Parameters
The stencils can be devised by mesh circuits with components in parallel and in series. Impedance relations between incoming and dispatch value densities can be formed using the constitutional laws of Kirchhoffs Voltage Law and Kirchhoffs Current Law.
Key circuit components mapped to the freight terminal are shown in Table 7.
Table 7: Stencil Components Mapped to Freight Terminal Parameters
Component Parameter Mapped Use Value
Voltage V Value density fixed cost/consignment
Resistor R Variable cost of handling and queuing $/consignment
Capacitor C Storage resource capability
Inductor L Handling resource capability
Transformer L ,L , M Batching operation- either 1 2 compaction or consolidation
Page | 154
Chapter 5: Model Formulation of Intermodal Production Systems
A further development, described in section 5.4, uses canonical forms of operational amplifiers to represent circuits with compensation feedback to improve storage utilisation and control.
5.2.2.3 Constitutional laws
Flows of consignments through a terminal of a certain itinerary of LUPs can be characterised by the value flux. What decides this itinerary, the work required by the terminal, the costs arising, the precision and control requirements need to be characterised at a layer of dynamics below the value flux. This is provided by electrical circuit constitutional laws. These constitutional laws allow changes in value density and consignment flux to be calculated according to the dynamics of storage and handling. The easy calculation between the value flux and these constitutional variables allows terminal complex activity (complex interactions between LUPs involving multipath coupling) to be represented.
Following these constitutional laws ensures that the circuit stencil is correctly defined for the terminal dynamics required i.e. that there is a transparent, reproducible, and scalable mapping in the signal representation of flow and in signal processing representation of terminal conversions of consignments.
The relationship between driving and driven forces, Value density and consignment flux can be characterised in the circuit analogue by Ohm’s Law. This law characterises the ability of a path of devices to impede the flow of current (our consignment flux). In other words, there is a voltage drop (value density drop) according to the consignment flux that is being forced through that part of the terminal itinerary. Resistance acts as a variable cost, relative to the flow or build up of consignments. Fixed costs as value density drops can also be catered for in a circuit with reversed polarity voltage sources. The relationship is demonstrated in (5.2 and 5.3): v = iR ……………….. (5.2)
The relationship with resistance and changes to value flow is demonstrated in (2):
)( === 2RiiiRvip ……………….. (5.3)
Where p is the value flux, i is the consignment flow, and R is the variable costed component of the impedance.
Page | 155
Chapter 5: Model Formulation of Intermodal Production Systems
The efficiency of the truck-terminal interface can be measured in turns of Ohm’s law. For a fixed tonnage load to be moved, larger trucks will entail fewer trips. Truck productivity may be viewed as maximising its payload. That is, moving the most tonnages over the shortest time for the least effort. A theoretical maximum efficiency may be calculated based on the amount the truck must carry over the deployment cost of a fleet of such capacity. In this case I= V/R, V being the tonnages and R being the variable cost for the trips necessary based on loading capacity. Actual efficiency of this rated truck fleet would depend on the motion costs and the additional costs of deployment due to the transit time. Where the network of space and time can be collapsed through transhipment terminals actual efficiencies may increase. Transit of consignment flows via transhipment terminals reduces the line-haul transit of smaller capacity trucks and can provide higher backloading rates. Terminal activities which hold a truck such as queuing, loading and unloading (called resistance) can add to the deployment cost of the fleet. It can be calculated the backloading opportunities provided by terminal operations reduce the total calculated trip cost of the fleet. The operation of dispatch, receival activities of consolidated loads by rail or larger truck to end destinations can similarly be calculated. Efficiency ratios of total costs by original fleet to that of the transhipment interface can be calculated to consider the net efficiency benefit.
Circuit analogue stencils are formed by balances of driving forces (Voltages through Kirchoff’s Current Law) and balances of driven forces (Currents through Kirchoff’s Voltage law). These balances act as mathematical constraints to form continuous integro-differential equations depicting the relationship through and across a Logistical Unit Processes for specific carrier pairs25.
For the purposes of this thesis, Kirchoff’s Voltage Law (KVL) can be interpreted as stating that the algebraic sum of all value density values around any closed path in a stencil including value density drops equals zero.
Similarly, Kirchoff’s Current Law (KCL) is the algebraic sum of all consignment fluxes at any node in a stencil equalling zero. In forming equations using Kirchoff’s Current Law off a stencil, nodes need to be established where two or more circuit elements meet. The convention is followed that flows departing a node are positive and flows entering a node are negative. This grants the analyst n-1 independent consignment flux equations (Nilsson, 2005).
The combination of KVL and KCL allows the analyst to solve for n unknown variables in n equations.
25 The operations of circuit stencils allow the observation of the response to through variables (the consignment flux) and across variables (value density) at each Logistical Unit Process. Page | 156
Chapter 5: Model Formulation of Intermodal Production Systems
These laws can be used to solve for single closed loop circuits and complex mesh circuits, with and without storage devices. Mesh circuits can be simplified to a single stencil using Thevenin’s Theorem. This takes into account the loading effects of one circuit onto another. A stencil derivation of a unit process using a passive circuit with a storage term is depicted in Figure 24.
dv 11 +=v V dt RC RC 0
V 1 c = V 1 0 s + RC
Figure 24 Stencil Derivation for Transfer Impedance for Operations at a Terminal
The mathematical language applied is one of algebraic calculus through the Laplace transform, which is used for dynamical systems representation and their compensation process control (Dorsey, 2002). The stencil defines differential equations which capture certain values and interactions for parameters of resistance, R, and capacitance, C (storage). The stencil provides the canonical relationship between output and input.
The solution of equations presented in Figure 24 by inverse transform technique is:
−t −t RC == τ 0 0 tuevevv )( …………………. (5.4)
Page | 157
Chapter 5: Model Formulation of Intermodal Production Systems
Which denotes the Value density v, through this unit process element given the flow input forcing function u(t) after an elapsed time t and terminal processes of time constant ı .
Laplace transforms are useful in solving unit process operations using the analog of electrical circuits in three ways: understanding the response of multiple connected circuits which act as unit processes through a convenient means to solve sets of linear differential equations; understanding transient response due to alternate signal inputs; and analysing steady state responses when considering frequency (synchronisation) requirements (Nilsson, 2005).
A simple example indicates the role of value density and how it may be modified in a terminal process as represented by a circuit stencil and constitutional laws. Consider a simple flow through a waste transfer terminal. Consignments of approximately 7 tonnes arrive to be compacted to 18tonnes containers and then loaded onto a train for dispatch to a landfill bioreactor. The waste load to be shifted is Vi. This value is reduced according to the variable and fixed costs of waiting and handling. Variable costs are represented by circuit resistances and fixed costs are represented by voltage drops. These costs are based on the consignment flow rate which in this case is the same as the number of trucks and truck trips. Costs, most noticeably waiting costs, would reduce with a larger truck size and thus a smaller number of truck trips. Figure 25 demonstrates a simple circuit without storage elements depicting the transformation of value density with consignment flow rate.
Table 8 tracks the value changes.
Gate Wait Cost Handling Cost Train Network Line Reservation
Source truck Flow (Ii) 6% of consignment flow $20/truck consignment $93/container 180 truck consignments/day Train Operating Cost $20/container $160 $411 Source Waste Load (Vi) Batch Multiple $210/Consignment.km 18/7 Dispatch Load (Vo) $299/consignment.km Value Flux In = $ 37,800 Value Flux Out = $ 20,919
Figure 25: Stencil for Value Density and Consignment Flux Change at Waste Transfer Station
Page | 158
Chapter 5: Model Formulation of Intermodal Production Systems
Table 8: Value Density Calculation through an Intermodal Waste Transfer Station26
Parameter Values Units
Truck size 7 tonnes waste value $30 per tonne
Value Density $210 per consignment.km
Flow rate 180 Trucks/day
Ideal Value Flux $37,800
per truck (average 180 Variable Wait cost $30 trucks/day)
Static handling cost $20 per truck
Value net Value Drop in Transfer Container Production System $160 per consignment.km
Value Multiplier for consignment batching 2.6
Container Value Density $411 per consignment.km
Container Flow rate 70
Train network fixed cost- line reservation $5,000 per 54 car consist
$92.59 per container car
Train network fixed operating cost $20 per container
Value net Value Drop in Train Circulation System $299 per consignment.km
Value Flux at Destination $20,919
System Operating Cost of Transfer Station and Train System $16,881 per day
26 These cost and transport values are hypothetical. True costs of plant operation are not made available due to commercial confidentiality. Page | 159
Chapter 5: Model Formulation of Intermodal Production Systems
When assessing complex impedance from stencils with storage devices in later sections, KVL and KCL still hold. The relationships extend to reactance as well as resistance. Net reactance values are considered either net storage or resource consumption which exceed the static operating capacity of the terminal and which contribute to further delays. Polar notation and phasor diagrams are used to depict the relationship between resistance and reactance. These are described in section 5.4.1.3.
The load demand may be characterised by a total value. This in turn can be defined by the quantity batch, flow rate and timeliness information of arrivals. These parameters represent the nature of the driving force, necessary to deliver the load demand. The quantity and frequency also define the nature of the vehicle-load relationship. That is, the value can be broken down into the value- kg/consignment and the consignments/hour. Value can be associated for a particular timeliness band (value intensity-value/timeliness band). These connections are fundamental to our initiative to tie distribution infrastructure activity to freight services demand and thus arrive at a logistical structure of lower traffic intensity. The sub-system connected also allows the whole model to be calibrated on the basis of supply profiles from the seaport and the distribution of demand loads in the hinterland. The impedance relationships at the transhipment nodes are feasible only if they can cater for the time profile of both container supply and demand.
5.2.3 Load Based Design
The electrical circuit analogue approach to terminal design allows us to design from the back-end. That is, given our carrier departure requirements: train timing, length, utilisation, and other physical and logistical impediments associated with the load to be met, the terminal is so designed to address this load. In addressing this load, the terminal impedance design prescribes a source inflow profile. It does reveal the necessary precision and stability relations required to achieve the load. This is load based design.
This approach does not prohibit the application of different source inflow profiles. Where there is obstruction due to changing inflow profiles, this is revealed according to the range of impedances acceptable under the configuration. The modeller then is prompted to re-configure the terminal with available physical and logistical parameters to ensure a new set of precision relations are established. The acceptable impedance range is therefore a vital Figure of Merit.
Page | 160
Chapter 5: Model Formulation of Intermodal Production Systems
It is problematic to present a continuous basis for batching. For instance, the physical nature of the flow system is the throughput or mass flux. To address this, additional information is required regarding the intensity of the flow process: what is the bulk nature of the arrivals (quantity of pallets, container or train-load) and with what headway the consignments arrive (frequency). Furthermore this headway needs to be matched with the service to which the consignment is to be transhipped. This may be known as synchronisation.
One means to represent batching operations is to proxy the intensity of the batch arrival with the work required to process that batch. This will involve relating the nature of the arrival with the response of the terminal node: the nature of the arrival being characterised by freight bundledness27 and headway of arrivals (frequency). This approach is essential if the planner wishes to differentiate terminals on the basis of their functions and recognise that the system may be defined by the interruptions and transformations of this intensity. The planner then needs to observe how consignment flux changes through terminals.
Expressing these flow constituents aids in balancing and routing the flows of the system as well as facilitating some transformations at the terminal. The network procedures cited sofar however have generally detailed constituents in a linear fashion.
The mechanism of how the sub-network responds to the logistics load develops as follows:
1. Each hinterland region has a load demand profile to either receive consignments of a certain quality or to dispatch consignments of a certain quality. These quality attributes may be addressed in an engineering sense by system precision; 2. This load profile can be served by a specific itinerary; 3. This itinerary represents the necessary transformation path through an intermodal network. Thus it is a spatial-temporal path which incorporates the functional specificity of the terminal system. The itineraries are the paths of services to use; 4. The feasible configuration of an intermodal production system to these itinerary sets is assessed by an impedance design which is represented by an inter-connected set of impedance matrices; 5. These impedance functions represent terminal operations and incorporate process control mechanisms for considering how terminal complex activity can provide for intermodal flexibility to increase capability with different load demands;
27Bundledness pertains to level of consolidation of consignment (ie. whether a full container load) and commodity composition Page | 161
Chapter 5: Model Formulation of Intermodal Production Systems
6. The total hinterland load profile must match the impedance profile of the production system with the generator/ load profile of the seaport; 7. There is potentially some facility for the process control of terminal production generation based on changing demand load profiles.
Unlike earlier interpretations of the transhipment problem, which insufficiently considers the transformation operations at the terminal nodes, this mechanism is firmly centered around the impedance matching capability of terminal nodes, in isolation, and in synchronisation. This analysis is facilitated by making the terminal system the foci of generation.
5.2.4 Physical Realisability of the Analogue to Terminal Operations
Modelling freight intermodal terminals in an activity based approach must conform to criteria of fidelity and tractability28 which differs from current transhipment network optimisation methods. For freight terminal operations to have such feasibility, according to the circuit analogue stencil, circuit model criteria must also apply. This feasibility, or physical realisability, between electrical circuit stencils is based on determining their causality and stability (Kuo, 1966). Causality in this sense pertains to flow and value balancing around a Logistical Unit Process (a circuit stencil) and precision relations. Precision is the measure of how well services synchronise with flow load requirements. It will be shown in this chapter that a key insight that electrical circuit analogues yield to freight terminal logistics is how precision of terminal resources to flows affects space management. It follows that assessments of flexible capacity are based on seeking opportunities for space management responsiveness to changes in flow precision requirements.
The condition of Linear Time Invariant (LTI) systems implies that parameters are lumped and that flows are not sequential. Thus, the realisability between processes cannot be verified by this method alone. A means to consider the precision between processes may be obtained through applying the concept of impedance matching derived from electrical circuit design (Kuo, 1996). Impedance matching parameters, expressed as complex conjugates, specify the relationship between input and output signals and thus define the transformation to be designed in the process including the acceptable limits of operation. This allows both enumeration of the flexibility Figures of Merit
28 Model fidelity and tractability in this instance means accuracy to reality and meaningfulness. Page | 162
Chapter 5: Model Formulation of Intermodal Production Systems
outlined by Hulten (1997) described in section 4.2., and that the individual unit operations are made compatible with top-down design specifications.
Control of stability for the purposes of this thesis extends to transient stability, steady state accuracy, and flexible capacity. This thesis focuses on control mechanisms, called business rules, to ensure steady state accuracy and flexible capacity. Transient stability is outlined in Appendix B.2. It is flagged for further work beyond this thesis.
5.2.5 Impedances
5.2.5.1 General Concept
Impedance is the work function to achieve a certain precision relation. It includes handling work as well as delay time and consumption of storage resources (inventory cost). It is a variable cost. Storage requirements are due to a mismatch in the batching of arrivals with departures. Therefore the impedance characterises the necessary storage resources required to achieve a certain load carrier cycling. This can be represented by average dwell times of consignment carrier classes. We can also consider the characteristic of specific time-sensitive consignments. For instance, with perishable goods there is a decay in value (increasing impedance) with storage time. The impedance function is activated by the inflow source signal and thus the response depends on the inflow.
Mathematically, impedance of a value density gain can be represented by resolving the circuit stencil as a Laplace transfer function. The impedance of a circuit stencil, representing one or several unit processes can be calculated in this manner and thus its steady state response can be estimated from an input flow signal.
This chapter presents a new way of considering the storage-handling interface. Consider storage utilisation as a build-up of pressure at a point; the greater the pressure the greater the need for handling resources to provide throughput relief. Thus the pressure measurement (imaginary Value density) represents increasing inertia of throughput and is related inversely with the critical transfer metric at seaports of consignment throughput/unit area storage. This metric is influenced by a number of factors including:
1. Dwell time of consignment carrier pair types, 2. The number of handling resources, Page | 163
Chapter 5: Model Formulation of Intermodal Production Systems
3. The handling rate per resource, 4. The number of times a consignment is handled (including internal reshuffling), 5. The availability of upstream storage space (bottleneck loading effects).
Impedance is a steady-state dynamic measure of terminal response to achieve a transformation between two flow states which covers several attributes:
1. A value density net gain or loss; 2. A system cost; 3. A storage lag; 4. A transportation delay; 5. Load matching (cf. port parameter matching); 6. Resource productivity constraints; 7. The effect of filter business rules for multipath coupling; and 8. Compensation mechanisms for testing capability (impedance as an equilibrating mechanism to ensure stability).
Impedance is a complex measure. Impedance involves both handling and storage resources. Impedance here is a vector whose argument (phase angle) is formed by the measure of resource capacity consumption and variable cost. Normal variable handling costs increase due to the storage interface and represents a premium due to increasing inaccessibility of the storage resource (thus requiring additional lifts) or delays due to insufficient handling resources. This premium is represented by the phase angle. A reference phase angle can be carried through with the inflow and specify handling storage relations for the itinerary of that pair through the terminal. Alternatively, it can represent a deviation from optimum storage utilisation and thus lead the way for compensation activities (injection of additional storage and handling resources where possible by adding capacitors or inductors in series or in parallel).
5.2.5.2 Value Accounting
There are three value systems which we track through terminal unit process stencils:
1. The changes in value density across unit processes, Page | 164
Chapter 5: Model Formulation of Intermodal Production Systems
2. The changes in value throughput, 3. The impedance arising (variable costs).
Value density drop occurs in three instances: due to a fixed charge in supply chain activity such as a gate fee (negative voltage); due to a proportional charge, such as truck waiting costs due to a lengthening queue (resistance); and due to batching and other storage activities (a reflected impedance).
Storage elements can increase the value density held across unit processes and in doing so increase the impedance. Only when storage elements are balanced at an optimal synchronisation rate (resonant frequency) does maximum flux efficiency occur. At this point terminal space management matches with the incoming transhipment intensity required. Here terminal operations are most productive and fully utilised. Impedance may also be characterised as the ratio of value held across the unit process to consignment flux through the unit process. Impedance grows when there is an accumulation of consignment value. The rated operating capacity of the terminal, its resonance, will be set at a certain impedance level. Efficient terminals will be able to achieve consignment flux at a lower impedance ratio.
Impedance through increased reactances adds to the value density held across unit processes (adds to the value density stored), without increasing the flux. This represents a storage of consignment value which may be in designated stack or in queued resources (such as trucks). This will occur to a certain extent at the static operating capacity. When there is an imbalance in storage or handling resources at excess operating capacity, the impedance and value density stored rises. This reflects reduced accessibility of the stack, for instance and real variable costs increase. Space management methods may include injecting storage and handling resources to drive the performance to more optimum storage and handling resource utilisation. This will also entail an increase in variable costs (resistive impedance) due to the more handling resources employed, for instance.
5.2.5.3 Precision
Another way to look at design characteristic impedance is that loads that make up the train schedule departure have a design value rating (power rating) that is made up of the value density held across the stencil, the consignment flow rate and the load impedance. Impedance mismatch of load versus
Page | 165
Chapter 5: Model Formulation of Intermodal Production Systems
source will mean that the value rating of the carrier dispatch will not be met and that the train cycles, for instance, will be less than fully utilised. Either the volume slots are not being fully utilised (low consignments per cycle period) and/or the payload per consignment is less than desired.
To obtain impedance matching, the unit process, resolves a source impedance with a load impedance. A complex conjugate match allows for maximum value throughput. Any value dissipation is shared between the source and load impedances. Impedance matching is designed according to the impedance as seen by the source. That is, for maximum value throughput the stencil Logistical Unit Process must be so devised that it explicitly takes into account variations in the source flow. This enables the modeller to construct a mathematical approach to Hulten’s (1993) concept of transformation flexibility (described in Chapter 3), essential in measuring the capability of the terminal as in a hub.
5.2.6 Synchronising Storage and Handling Operations for Temporal Accessibility
Temporal Accessibility was identified in Chapter 2 as a key Figure of Merit in ensuring transport requirements are met effectively and in assessing the capacity-risk relationship of intermodal system formats. That is, there is some synchronisation between consignment flows (source loads) and the management of them at intermodal terminal. It is the effective storage utilisation at the unit process interface. This matching, called impedance matching, intends to control dissipative forces of poorly loaded truck trips, which load the network and reduce accessibility29. Maintaining temporal accessibility is seen as a vital measure in ensuring a feasible network of intermodal operations where intermodal points create more interface requirements. As an assessment of the capacity- risk relationship it is a measure of the effectiveness of retrofitted infrastructure.
Temporal Accessibility is not simple to measure. It can be measured partially by proxy attributes. One attribute is resource consumption. This assesses the level of imbalance between storage and handling resources given the degree of transhipment intensity of the incoming flow load. Initially it is valuable to design storage/ handling relations which complement a certain regularity of service. A storage utilisation measure will reflect this. An optimal storage utilisation reduces the holding cost of large headways associated with large accumulations whilst balancing transport costs of small
29 This is Tarski’s concept of dispositive elasticity discussed in Chapter 2 (Tarski, 1986). Page | 166
Chapter 5: Model Formulation of Intermodal Production Systems
shipments. The degree of transhipment intensity is a value which defines the synchronisation that must occur between carriers and unit operations.
In electrical circuit theory, the balance between resource components is called resonance. For the logistics analogy, the resonance frequency proxies the synchronisation rating to which the storage/ handling interface has been designed. The passive circuit stencil exhibits resonance presenting a pronounced peak in its transfer function. The performance of resonance is denoted by the quality factor of value flux to value lost which characterises the sharpness of the peak. The performance of resonance also relates to stability.
Resonance is also related to transhipment intensity (timing), specifically, how the unit processes in the terminal perform at the rated transhipment intensity (timing).
Optimum Storage Utilisation is when transhipment resources are in balance. This is represented by the resonant frequency. There is minimum phase angle. The frequency information attached to the incoming flow is resonant if it matches the circuit impedance caused by the L, C, parameter values in the circuit (equation 5.5). Table 9 depicts some simple relations on the level of transhipment intensity.
1 f = ………… (5.5) resonance 2Π LC
Table 9: Transhipment Intensity and Impedance Condition
Transhipment Storage Utilisation (Impedance) intensity
(frequency)
Good Poor
Direct Min. storage Excess land avail
Indirect Lifts offset storage Large no. lifts requirements Inaccessible Stack
Page | 167
Chapter 5: Model Formulation of Intermodal Production Systems
Synchronisation resonance can occur through a case of no storage requirements, perfect cross docking and no accessibility impedance. It may also occur when impedance elements are matched. This is called complex conjugates. Complex conjugates represent a case where flow information match terminal operating design. The complex conjugate is presented in detail in s.5.4 on complex impedance.
In reality, the storage facility will have a combination of direct and indirect transhipment requirements. This can be addressed by handling and storage resources proxied by inductive and capacitive loads respectively:
= Π XfLL 2 ……………………………. (5.6)
1 X = c Π 2 fC …………………………… (5.7)
Where XL is inductive impedance and Xc is capacitive impedance, f is the synchronisation rating, L is the handling coefficent, C is the storage coefficient. These components are termed reactive impedance, and with resistive impedance, is a measure of complex impedance which references terminal work effort for a specified flow throughput.
These storage and handling impedance elements arise from alternate configurations and operating business rules for terminal Logistical Unit Processes. The terminal designer can plan according to anticipated source flows or specify source flows which can be managed according to terminal technology and layout.
The interaction between direct and indirect transhipment can be depicted through an RLC series circuit stencil for illustrative purposes. At resonant synchronisation, there is no reactance component as: XL=XC. The resolution of resistive and reactive impedance is depicted in equation 5.8. How the impedance alters for a certain synchronisation rating is depicted in Figure 26.
ZRXX=−−22() LC……………………………………. (5.8) Page | 168
Chapter 5: Model Formulation of Intermodal Production Systems
Note an inductive load is proportional to synchronisation rating (frequency) and a capacitive load is hyperbolic.
Figure 26: Impedance Z with Synchronisation Curve
Figure 26 leads to the understanding that, with higher source synchronisation rates, there is a larger proportion of direct transhipment, and inductive load management predominates.
This analysis extends the trade- off relationship of marginal costs to throughput to identify the flexibility of the terminal type to respond to transhipment intensity needs. This was identified as a functional requirement in section 4.6.2.1. The parabolic curve of impedance to synchronisation (Figure 26) has a similar shape to that developed empirically by Ballis and Golias (2004) depicted in section 4.6.2.1.
The relationship between handling and storage effort, inductive and capacitive loads can also be rendered as a phasor diagram. The reactance component of impedance is made up of a vector resolving the impedance magnitude and a complex impedance argument (phase angle).The complex component of impedance represents the lift/storage access complexity and the argument proxies the transhipment storage lag (the impact on storage/handling access and capability that the specified dwell time has).
Transhipment Intensity is the information contained in the signal flow and is the degree of directness of transhipment. This is related to the average dwell time of consignments. Transhipment
Page | 169
Chapter 5: Model Formulation of Intermodal Production Systems
intensity characterises the handling storage relations required to meet carrier outflow cycles. It is represented by signal frequency. The frequency domain allows reactive elements to be incorporated into the transhipment calculus. These reactive elements denote the resource consumption component of work in the new impedance formulation and thus advise the planner on capacity limits and opportunities.
The magnitude of the signal frequency of the value density source corresponds to the time sensitivity of the freight consignment and thus to the direct transhipment intensity which must be managed. For the freight analogue this frequency band is termed the synchronisation rate. Terminals may be characterised by the band of synchronisation consignment flow they can address - either statically depicted by a passive circuit or flexibly, depicted by a compensation control circuit, such as an operational amplifier. This is further discussed in section 5.4.
The degree of directness of transhipment is a prime parameter in assessment of the precision interface between intermodal terminal resources (Tarski, 1986). This thesis uses synchronisation as a major parameter in assessment of terminal performance. For urban intermodal terminals, facilitating synchronisation of consignments is as significant as processing mass flows in assuring precision and stability relations in the terminal and between terminals in the network. The inclusion of the synchronisation attribute extends the concept of capacity to capability, where required functions are maintained at a rated flux under variable flow conditions. In achieving a resonance synchronisation of storage and handling activities, resources have been adapted such that the required transhipment intensity is met. At fresonance there is no additional impedance due to transportation or storage lags (phase angle shifts). In reality there is always some excess capacity (underutilised resources at parts of the day).
If a sigmoidal curve of phase angle (the argument of the complex impedance) against frequency increases is considered, it may be noted that the phase angle reduces for a certain frequency with increasing resistance. This may be preliminarily interpreted for the freight analogy that the storage lag is less sensitive to increasing transhipment intensity as more handling resources are deployed (Figure 27).
Page | 170
Chapter 5: Model Formulation of Intermodal Production Systems
Figure 27: Storage lag (Phase Angle) with Transhipment Intensity (Frequency) (Generated from RLC Stencil in PSPICETM model)
The combination of L, C can move the fresonance in either direction. They can also change the sharpness of the inflection. An increase in these parameter values increases the sharpness of the phase angle response, thus making phase angle increases more sensitive to frequency changes for the given circuit stencil.
The relevance of the fresonance is that the terminal is designed to operate at this point. Any change in the time sensitivity (transhipment mix) of the inflow will then lead to the development of a complex impedance where the overall impedance is affected by the interplay of storage and handling. A phase angle will develop which will add to the overall impedance cost ($/TEU throughput) as a resultant of the vector of resistances and reactances (Figure 28). Otherwise the terminal will need to provide compensation devices (see note on Op Amps below).
Page | 171
Chapter 5: Model Formulation of Intermodal Production Systems
Figure 28: Development of Complex Impedance Due to a Synchronisation Response (Frequency) which does not meet the desired synchronisation rating of the unit process (F-resonance)
Inductive and capacitive loads act as characteristics of the storage stack- i.e. more temporary storage to longer dwell time.
Through changes in C, L storage parameters, we can show at what frequency level and band, consignment flux, I, is greatest. For an RLC circuit stencil, increasing the inductive load (L) coefficient narrows the band and shifts its apex to the left. Reducing L, widens the spread and shifts the apex of throughput to the right. The consignment flow can be denominated over the resulting value density change to depict the cost of allowing this flux at a certain transhipment intensity (frequency) (Figure 29).
Figure 29: Variation of Consignment Flux as a Function of Synchronisation (Frequency Spectrum)
With an increase in consignments stored (V), the I/V metric reduces denoting a larger dwell time, f1.
At f2 , a higher flux is achieved however this is at a cost of a higher storage. Figure 29 is also known
Page | 172
Chapter 5: Model Formulation of Intermodal Production Systems
as a magnitude Bode plot and is useful in design of terminal functions to provide the required flux response, such as batching and throughput rate, under certain flow profiles.
The response to the space management problem can be addressed by the typology of the circuit stencil proposed. For instance, an equivalent capacitance in parallel adds to capacity (capacitance) whilst its reciprocal in series adds in series. In parallel, the value density transferred is constant. This can lead to a drop in consignment flow. In a stencil series arrangement the consignment flow is constant. An equivalent inductance in series is added directly, whilst in parallel its reciprocal adds.
Load impedance (the departing carrier) corresponds to source impedance if the departing carrier’s cycles, required to address the transhipment intensity profile, are provided.
The synchronisation band information plays five roles:
1. It designates the space requirements of carrier flows for a certain train load service; 2. It designates the transhipment intensity and what will be the critical limiting resources – storage and/or handling equipment. This has implications for the work effort at the terminal and relative cost; 3. It can be used to apply business rules through acting as a compensation variable in space management (ensuring stability of storage utilisation); 4. It provides precision correspondence of diffuse truck carrier flows with the train load service; and 5. It provides for multi-path coupling (selection of itinerary path within terminal). Timing information activates filters which form one part of the critical attributes of terminal impedance. These attributes are:
1. Gain/ loss of value density, 2. Control of consignment flow, 3. Storage lag, 4. Filter business rules.
5.2.7 Response to Business Rule Levers
The nature and role of business rules in sketch planning design of terminal operations was outlined in section 4.3.4. It is critical for the fidelity and tractability of any model formulation that a
Page | 173
Chapter 5: Model Formulation of Intermodal Production Systems
mathematical basis is developed that can yield responses to changes in business rule instruction both at the local terminal level and the system-wide level. In section 5.4 we develop the transhipment calculus and the mechanism of filter impedance functions which act as terminal itinerary control due to terminal specific parameters and system specific parameters.
The application of filter functions to business rules can represent:
1. The bulk queue, 2. Pulse correspondence, 3. Splitting dispatch, 4. Specifying the (storage) saturation range.
5.2.8 An Itinerary Mechanism
Modelling an itinerary through a freight terminal must consider impedance due to the critical path of services. This impedance must consider intentional dwell time, delays due to processing, evolving storage lag inertia due to capacity saturation, and other resource dependencies to facilitate critical path throughput. In using an electrical circuit analogue, four challenges in characterising these issues are:
1. Conservation of mass; 2. Instantaneous nature of flows; 3. Frequency dependent circuits with evolving impedances; and 4. The fungibility of electrical flow
The conservation of mass is dealt with by the constitutional law of current flows (consignment flows) which have a mass imputed to each consignment flow type. Consignment flux alters according to impedance relations in terminal logistical unit processes.
For tracking flows which can be stored throughout the terminal in order to meet a certain output sequence, the mass flow relationship and the instantaneous nature of electrical flow will not easily map. A further state is required.
Timing information (transhipment intensity) is used by filter devices to select itineraries amid multiple path options. “Filters are networks that process signals in a frequency-dependent manner. Page | 174
Chapter 5: Model Formulation of Intermodal Production Systems
The basic concept of a filter can be explained by examining the frequency-dependent nature of the impedance of capacitors and inductors”. Zumbahlen (2008, p.583). Filter stencils of LUPs, with their other functions, establish precision relations between carrier pairs.
Electricity is fungible. It is the sam unit of value whereever it may be in the system. Unlike freight, it cannot be easily differentiated on the basis of commodity type, destination or some other criteria. This poses challenges in apply electrical circuit analogue to freight terminal logistical processes.
Two Figures of Merit used to identify boundaries of feasible terminal operation:
1. Transhipment intensity against impedance; 2. Transhipment intensity against I (a quality factor measurement).
The mechanism of the itinerary is the synchronisation information embedded in the load flow which specifies an acceptable source flow. This information is the transhipment intensity level, a range within this spectra, a carrier pair, is designated to a specific dispatch load such as a train set. Thus the value contribution of the flow to the end train dispatch payload can be calculated.
When multi-commodity, multi-modal flow analysis is to be considered as part of the transhipment problem of operations research, there is multi-faceted information that must be mapped. For instance, in an exporting operation, a commodity from a particular shipper is carried by a particular truck type with a particular time window and this must map to a particular dispatch service. The itinerary mechanism must track the multi-path coupling to occur through the terminal to match these flow characteristics.
This thesis applies time domain techniques and frequency domain techniques. The frequency domain is termed the synchronisation domain for the freight application in this thesis. Time domain techniques are used to observe and manage physical flows through the terminal according to elapsed time. Time is the rationing device. Synchronisation domain techniques are used to observe and manage information regarding these physical flows.
In the time domain interest is in both transient and steady state flows. In the transient we may observe and control the stability of buffer storages and the interface with handling equipment. In the steady state the effect of average value density magnitudes, consignment dwell times and consignment fluxes may be observed when transformations to the batch and flow rate are made. The
Page | 175
Chapter 5: Model Formulation of Intermodal Production Systems
time domain analysis constructs delay functions to achieve a required output. Time domain techniques are used to assess relative and fixed costs.
The time domain has been classically used in the transhipment problem of transport in networks. It is used to calculate flux balances with storage and handling requirements demonstrated by cumulative graph theory and delays estimated by queuing theory.
Synchronisation Analysis as a dual of the time domain offers four further insights:
1. Examination of further encoded information dealing with driving forces of value density, consignment flux, value intensity and dwell time, 2. Finding terminal impedance response, involving resource consumption as well as system costs, 3. Applying more elaborate signal processing techniques which assist with design and analysis of terminal complex activity by continuous approximation. These analytical techniques include filters to control the itinerary of flows, and compensation control to assessing the flexible capacity of terminals, 4. A means to address the complex bundling problem in metropolitan intermodal terminals, a sub-set of the multi-commodity flow problem (see s.5.4.3).
Analysis of the synchronisation domain allows the inclusion of time sensitivity as well as carrier pairing information attached to the flow and allows an assessment of the consumption of resources in and through a terminal. It provides both a means of dynamic assignment and itinerary control (see Appendix F).
Information flows pertain to:
a. Belongingness of the source inflow to a specific dispatch carrier load; b. Degree of direct transhipment intensity; and c. Dwell time requirements.
This information allows the planner to develop an itinerary mechanism of multipath coupling through the terminal as well as attribute work effort. The synchronisation domain can be used to associate particular carrier pairs through the terminal. This is a prerequisite for formulating and assessing multi-commodity flow problems.
Page | 176
Chapter 5: Model Formulation of Intermodal Production Systems
Precision is achieved through synchronising the timing of the flows. Not only is the flow translated in time but the correct transformations and pair matching are made. Precision then becomes more than incorporating the delay function of a set of terminal operations.
5.3 Flows as Signals
5.3.1 Introduction
We commence the development of an activity-based approach for freight by proposing pertinent model elements from the theory of electrical engineering according to a fundamental activity: transport phenomena of attraction and impedance. These relationships foreshadow a means to address the allocation phenomena across networks (discussed in chapter 8). This section describes the transhipment calculus which enumerates the characteristic impedance of the intermodal production system.
5.3.2 Excitation Response
The passenger generator–attractor flux phenomena (Blunden, 1971) can also be seen in generating freight of different consignment types. The generation of freight occurs throughout nodes in the distribution network and affects flow rhythms. These freight types need to be defined in how much transformation is required at intermediate nodes for specified delivery to the seaport, in the case of export consignments, for instance. Consequently, the signal flux is a burst which represents frequency and bundledness complexity (from the point of view of further transformation). The relationship between flow and function may be considered one of excitation and response. Therefore, the signal approximates the flow with significant characteristics. We can define these characteristics as flow “burstiness” (Kleinrock, 1974) where “burstiness” describes the rhythm of the flow – that is the bundledness complexity of containers as well as the frequency. The pulse bundledness function, u(t), with magnitude, V, can be described in part by its duty factor: the ratio of pulse width to period or headway, T (Kuo, 1966). A greater width, denoting a longer train block, for instance, would require a longer time for transfer.
Page | 177
Chapter 5: Model Formulation of Intermodal Production Systems
This signal contains information on the work that needs to be enacted through the terminal node. How this workload is resolved depends on the resistance interface characteristics of the terminal node which yields a time impedance. The derivative of this work represents the rate of interface transfer of the consignment (consignment/time), including any conversion process (such as break- bulk). The mass flux is the product of these two relations (mass/time). We can attribute a certain value to the mass in order to get a value flux. The total mass stored in the unit process or in the terminal can be derived through integration. This yields the significant Figure of Merit of storage utilisation (consignment/unit area for a given period). Capacity then becomes dynamic: it is associated with fluctuating demand and the capability of the terminal’s handling and storage resources.
A typical response from a pulse burst of an arrival sequence may be represented by the shift equation and exponential response curve of Figure 30. This response is in two identifiable parts: an impedance exponential and negative exponential curve. This response can depict the cumulative transfer of a consignment between operating processes, whether pallet, container or train block. The cumulative load transferred has a changing rate of processing (time/ITU) and this could be a function of the resistance of the handling equipment and the growing complexity in accessing storage (descriptive of stacking strategies). The degraded signal can represent the storage on a departure unit, such as a train block, prior to departure. In this way, the transient storage charts are more detailed than Newell cumulative curves of arrivals and departures as resistance in handling is included as well as temporary storage on mobile units. Pipeline delay may also be included between the transfer stages. In its simplest form, the transient storage response gives a performance measure of the proportion of the load transferred between interfaces and the elapsed time for this sequence or group of sequences.
= − − vVututTs 0 [() ( )]
=−− TRC/ vV0 ()1 e = −−()/tT RC vVeT
Figure 30: Terminal Impedance Response as a Resolution of Work from a Pulse Input Page | 178
Chapter 5: Model Formulation of Intermodal Production Systems
The dynamical response of the processes within the terminal is a combination of forcing and natural responses. The forcing response is due to the forcing function (in this instance, a pulse burst). The natural response represents relations, independent of the forcing function. These relations may be defined by parameters that describe the resistance interface between system resources: the handling technologies, layout arrangements and storage capacities. These parameters are discussed in the next sub-section where we formulate a characteristic impedance.
The signals tracked through the control system represent mathematical relations rather than physical flows. This allows the modeller to include detailed information regarding the driving force of flows (their bundledness) and the response at the terminal in transforming these flows. This has direct implications for how the algebraic calculus is devised from the integro-differential equations. The electric analogue approach is selected over a physical flow approach as we can define more clearly the response of technical elements using the electric analogue.
5.3.3 Stability Control
Rodrigue (1999) has noted that in a system of increasing synchronisation, there is the increasing prospect of instability under disturbance events. This is a case of managing complexity under increasing tense flux. This is not approached in equilibrium network modelling. Of particular value in the process approach is the location and control of stability.
It is necessary to differential between transient and steady state analysis in evaluating precision and stability performance. The forced response is the terminal response due to shock loads; the natural response is the longer term response of the terminal based on its internal mechanisms. In the spectrum of considering the forced to natural response of a terminal unit to flows, transient analysis focuses on the forced response and steady state analysis focuses on the natural response. The steady state analysis allows further consideration of synchronisation requirements (in the frequency domain) and allows models to be built which connect other unit processes in the terminal.
This thesis is predominantly concerned with analysing steady state relationships. Identification and design of resistance interface parameter based on specific vehicle and handling technologies and the transient stability analysis arising from this are in the domain of transient analysis and are thus
Page | 179
Chapter 5: Model Formulation of Intermodal Production Systems
beyond the scope of this thesis. They remain highly pertinent to the establishment of a transhipment calculus based on the physics of electrical circuit theory. Transient analysis is discussed at in Appendix C and is flagged as further research in the field of stability.
Excitation and response can be represented by exponentials. These exponentials have a trigonometric representation. Notation represents the excitation – response signal of apparent values. Apparent values in consignment flow and value density represent the magnitude of a vector of a) a real flow which is carried between circuit components and b) an imaginary flow component which marks some storage (accumulation). This is very useful. It helps the planner identify and design the storage-handling interface to achieve the output flows required.
Transient time analysis is one aspect of stability analysis in intermodal freight production systems. Another level of stability analysis is undertaken in the frequency domain when timing aspects are considered. The transient storage is formed on the basis of the service rate of handling equipment, the number of times a consignment is handled (the lifting ratio) and the storage characteristics. The second order transfer function, formed by the analogue stencil with two or more memory (storage) elements, is critically defined by the parameters ω, the natural frequency, and ξ, the damping ratio. They represent the typical handling – storage relationship. Thus, we can use these parameters to represent the resistance interface between system resources. For instance, the more transfer equipment engaged the lower the impedance (and thus activity time) in unloading trains. Figure 73 in Appendix C depicts how the damping factor can be used to moderate the unit process response at different synchronisation (frequency) rates.
To analyse processes within a terminal it may be assumed that the traffic – transfer relationship can be represented by transient storages:
• Arrival flows are an excitation which may be represented by pulses (trucks and trains). • Transient storages from input pulses represent the unloading, sorting, re-loading resistance of equipment processes. • Transient storages can be analysed for synchronisation between coupled processes. • The critical time line of activities can be summarised as the network function and this is the characteristic impedance. • Impedance matching parameters between the processes provide the decomposition analysis to ensure that the characteristic impedance is met.
Page | 180
Chapter 5: Model Formulation of Intermodal Production Systems
5.4 Transhipment Calculus
5.4.1 Logistical Unit Process Mathematical Tools
5.4.1.1 Characteristic Impedance in Time Domain using Stencil Operational Amplifiers
The itinerary of an intermodal production system is a pathway which may involve a number of processes within the terminal and among terminals that are part of the production system. Each itinerary has a characteristic impedance which can be used to both calculate dwell time through the system and represent the physical realisability of operations, based on causality and controllability. The characteristic impedance is formulated from steady-state and transient dynamics of the system. In order to develop dynamic equations for freight processing at node distribution terminals, an electrical circuit analogy is proposed with the potential for converting the physical transformations required into a mathematically reproducible format. It is proposed to use electrical Operational Amplifier (Op Amp) circuits as a stencil to represent processing of traffic-transfer relationships in an intermodal terminal for each unit process30. Op Amps are active circuits, involving a combination of passive circuits, which can include feedback mechanisms. A number of Op Amps can be used to represent one unit process. Op Amps have two salient characteristics which are valuable for the investigation of positioning terminals of certain functions in an intermodal production system:
• Performing mathematical operations on storage devices; and
• Providing physical realisation of Business Rule devices such as controllers and compensators.
The demonstration of Op Amps supports the assessment of intermodal production systems as physically realisable: that there is continuity of flow (causation) and that they are stable or controllable within the itinerary of terminals that constitutes the system. Op Amps represent the steady state or natural response of terminal unit processes to flow profiles. Changes to how the handling operations interact with buffer storages by re-arrangement of its elements (R and C elements) can result in a new set of capabilities.
30 Op Amps are generally configured to perform mathematical operations on Voltage variables (value density) only. Consignment flux information can be evaluated using the technique of an R-shunt stencil. Page | 181
Chapter 5: Model Formulation of Intermodal Production Systems
The design of the functions of freight terminals using Op Amps can then lead to an understanding of:
1. The trade-off between handling work and storage space (capability and capacity), as well as storage interactions; and
2. The effects of coupling (queue interaction) with multi-function arrangements.
The Op Amp analogue thus permits the modeller the design flexibility to manipulate node impedance so to facilitate different combinations of the time composition of freight. In turn we can form itineraries through a decentralised control system which facilitates a load following response.
A non-inverting Op Amp is depicted in Figure 31. The Op Amp can connect with passive circuits in a variety of ways (Z1, Z2). Op Amp stencils can be designed for Logistical Unit Processes of an intermodal or seaport terminal to modify independent flow variables by enacting storage, batching, transhipment and other processes. The stencil allows circuit elements to be conveniently represented mathematically (Z1, Z2) where Z1 forms the source inflow. For this model formulation non-inverting Op Amps may be used for modifying export consignment flows.
Figure 31: Non- Inverting Operational Amplifier Stencil Sketch
Page | 182
Chapter 5: Model Formulation of Intermodal Production Systems
5.4.1.2 Stencil Op Amp Canonical Forms
The modelling endeavour is greatly assisted if canonical forms of stencils are identified and placed for an application of operation (Table 10). For this thesis, Operational Amplifiers are used with and without connection to passive circuit stencils. Op Amps provide clear advantages as analogues over passive circuits only:
1. They are non-loading. That is, upstream and downstream processes are treated independently, 2. They provide for consolidation processes leading to value density gains, 3. They act as filters for itinerary pathing within the terminal, 4. They act as filters to provide for control actions.
Page | 183
Chapter 5: Model Formulation of Intermodal Production Systems
Table 10: Canonical Op Amp Stencils for Representing Logistical Unit Processes
Canonical Form What it Does Application to Logistical Unit Processes and Relationships
High Q filter Pass only high Q signals Container stack management
Broad bandpass Passes signals within Analysis of Synchronisation filter bandpass profile
Gain amplifier Batching operation to a Compaction, batching multiple of the input flow value density
All-pass filter Alters the phase angle to Analysis of dwell times in meet adjacent LUP and storage dispatch requirements
Summing Combining signals of Combing source load flow differing amplitude, phase signals; combining dispatch angles and frequencies flow signals
Difference Subtract signals from each Used to calculate residual other flows after filter controls
R-shunts Converts I information to V Calculate the consignment information flow from the value density
Load shunts Adds a storage buffer to Cater for changes in degree reduce drop in dispatch of transhipment intensity payload
Bandreject filter Rejects band of frequencies Sidings occupation
5.4.1.3 Cost and Consumption: Complex Impedance and Polar Accounting System
Complex impedance allows the planner to capture the costs of using resources as well as the consumption of their capacity (and the residual capability) in an accounting framework. In this sense the accounting framework is a space management tool which can incorporate physical capacity of the available resources and the use of the configuration according to some specific business rules which the modeller chooses.
Page | 184
Chapter 5: Model Formulation of Intermodal Production Systems
The interaction between changes in value and space management requirements in freight terminal logistics, the complex impedance, is represented by using a novel application of the imaginary number system. The polar accounting system uses a convenient notation of the real and imaginary number system, called phasor notation. The significance of the evolution of the phase angle as the argument of this vector relationship under resonant and non-resonant terminal conditions was discussed in section 5.2.6 on transhipment intensity of the consignment. In this section we develop the mathematical formulation which has wide ranging applications in terminal logistics wherever there is storage-handling interaction. In particular, evolving storage-handling inertia can be depicted which designates capacity consumption.
The polar accounting system allows us to represent complex impedance. Complex impedance is a resolution of two different driven forces: inertia (space utilisation) and resistance (a variable cost of handling due to restricted access, and inventory costs31). It was established in the discussion of excitation- response relationships, that flow is a sinusoidal driving force. That is, it acts as a periodic signal of time and timing32. The response will also be sinusoidal. With such signals we can then observe their complex response through a unit process stencil: cost-value and space-handling structures according to the specifications we desire.
In electrical circuit theory, storage impedance (reactance) and variable costs (resistance causing voltage drop) can be represented on a phasor diagram, where there is a deviation from operating resonance at a prevailing synchronisation band (frequency). This relationship is applied to the accounting and space management system of the transhipment calculus (Figure 32).
31 Although past work with linear cost functions on staging freight transhipment (see Black and Rimmer, 1983) have used the inventory component as a trade-off with production and transport costs, this has been most relevant for transhipment nodes with long line-haul. In “close” metropolitan intermodal terminals, real inventory costs would normally be considered low as there is usually rapid turnover of consignments. Also terminal operators are unwilling to charge demurrage for longer dwell times given the competition with the all road network. In this thesis the relative inventory cost is an ideal cost that may or may not be passed on to the shipper; it rather represents the relative resource consumption cost as storage reaches saturation. 32 In the frequency domain. Page | 185
Chapter 5: Model Formulation of Intermodal Production Systems
Figure 32: Terminal Space Management Analysis through Complex Impedance Phasor Diagram
A driving force as a sinusoidal steady state function can be represented as an exponential signal:
-jθ 0 (θ …………………(5.10ے e = cosθ +/- jsinθ = Vm
Where the middle terms represent the rectangular form of real and imaginary values and the right hand term represents the polar form: both forms convey magnitude (Vm) and phase angle (storage lag) information (θ) at the prevailing synchronisation rate. This expression can be simplified to the cosine for sinusoidal signals at the prevailing synchronisation rate (ω):
jωt jθ V= Vmcos(ωt+ θ) = VmR(e e )……………….(5.11)
The phasor form is then:
jθ V= Vm e = P[Vmcos(ωt+ θ)]…………………..(5.12)
And to extract the real impedance, the inverse phasor is:
-1 jωt jθ P {Vmcos(ωt +θ)} = VmR(e e ) …………… (5.13)
The impedance magnitude Vm is an apparent magnitude. That is it has no practical meaning other than a convenient representation, with the phase angle, of cost and resource consumption quantities. The phasor notation considers the state variables of consignment flow, value density and impedance in both their real, throughput forms and their storage forms: The notation that relates these two are the apparent magnitude and the phase angle.
Page | 186
Chapter 5: Model Formulation of Intermodal Production Systems
Phasor diagrams are useful in depicting the dual nature of impedance which is the essence of the terminal accounting system we develop. Its usefulness is demonstrated in Chapter 6 where a running model is illustrated showing the effects of additional storage configuration on the value flux (a process known as load shunting).
Complex impedance represents the evolution of the storage lag through terminal operations under varying conditions. Complex impedance can be tracked and depicted by phasor notation for a given transhipment intensity (frequency). This shows how consignment flows and value density levels interact with their phase angles.
The concept of value flux was presented in section 5.2.2.1. It is a product of the driving and driven forces, of value density and consignment flow rate respectively. This conveniently maps to the state variables in the electrical circuit analogue of Voltage and Current. Vector representations of each of these state variables with impedance over a unit process grant the modeller insight into the cost- value trade-off of certain terminal activity. There are four attributes and eight indicators that assist in design according to specifications. These relationships are illustrated using two broad examples.
The impedance diagram shows the relative cost of storage consumption A storage buffer leading to saturation will develop increasing inertia (due to increased slot occupation and complex slot occupation). This will have an increasing variable cost. Saturation inertia can also be due to the parameter of storage lag as represented by the phase angle33. The corresponding value density vector diagram will depict a high value density stored over the element for a relatively low batch gain value density passed. The corresponding consignment flux vector diagram will depict a small consignment rate throughput given the value density stored. Finally, the composite Value Flux diagram demonstrates the relative magnitudes of value delivered to the load and value stored. The more value delivered to the load, the less is dissipated through a Value Density Drop (shown in the net batch gain and the relative impedance cost). These relationships are depicted in Figure 33.
33 The storage lag is a function of dwell time, stack-handling strategy, and loading- retrieval relations. The inertia impact on relative handling costs of a large average dwell time can be managed by business rules of configuring and handling the stack. Page | 187
Chapter 5: Model Formulation of Intermodal Production Systems
Figure 33: Attributes of Impedance and Value in a Logistics Unit Process for Indirect Transhipment
A storage buffer that is experiencing little constraint will be operating closer to resonance (Figure 34). For a direct docking operation, for instance, it may require greater handling activity and thus a larger real impedance or cost. The consignment flux may be high and the value density stored low. Storage-handling impedance, and capacity consumption, is low34.
Figure 34: Attributes of Impedance and Value in a Logistics Unit Process for Direct Transhipment
The cost of a higher direct transhipment intensity may reduce the value delivered to the dispatch load or it may be acceptable.
These Figures of Merit can also combine with frequency-domain information to give further performance indicators. For instance, the container throughput/unit area productivity measure can
34 All these relationships may be expressed utilising 2 state equations of driving and driven forces. Page | 188
Chapter 5: Model Formulation of Intermodal Production Systems
be represented by a similar measure in the synchronisation domain considering Consignment throughput per Value of slots utilized in the stack (the Iout/(value intensity x bandwidth reserved).
At resonance the vector relationship is not evident. This is because reactive impedances are matched. The productivity of different unit process configurations (through alternate stencil forms) may still be considered using the Quality-factor (Q) measure. This is described in section 5.5.2.
The generation of a phasor diagram is a sketch planning method in observing the unit process response at a prevailing synchronisation rate. Complex impedance can also be represented in Laplace transfer functions where there are complex roots. This leads to a sparse representation of complex terminal activity which can be further transformed into a state space representation for matrix operations when treating the terminal as a Multiple Input –Multiple Output System.
5.4.2 Multi-path Coupling
5.4.2.1 Multiple Input Multiple Output Systems
To further define terminal characteristic impedance, a two port terminal can be drawn, which represents a multiple input-multiple output (MIMO) terminal system of multiple-path couplings. This terminal flowchart may include: input-output relations, conversions, and storages. It may represent a number of Logistical Unit Processes which have been connected together by their own flux variables (impedance matching complex conjugates). For instance, Figure 35 represents a subset of processes that are planned to occur at the Enfield Intermodal Logistics Centre in Sydney (SKM, 2005).
Figure 35 A Multipath Coupling Diagram for Direct and Indirect Export Transhipment
Page | 189
Chapter 5: Model Formulation of Intermodal Production Systems
The configuration of the intermodal terminal may be defined as including break-bulk activity (that is distribution and consolidation of consignments F1, F2) or a direct transhipment from road to rail (D) where the container is consolidated upstream. System delays through link separation are reduced if we have these functions at the one terminal and may affect how empty container resources (E) are accessed. This may lead to instability and would require enlarged storages to buffer the activities. In this case, the flow of empty containers is coupled to the operations of break- bulking and break-bulking production is coupled to the production of the train block. The time taken to make empty containers available can be compared with the time taken to collect partial flows, sufficient for container production to a like destination. In this way, we can compare how synchronised flows P2 and P3 are in order to produce the train block output frequency L1. Assessing transient storages between coupled processes with different input flows can also guide control strategies for bulk queues. Synchronising transient storages is a potent technique in engineering design.
Flow conversions within the terminal can be represented by network functions (transfer functions) which describe the delay and activity time for processing according to the coupling units engaged in the itinerary. These represent the response to the excitation signal of the input disturbance, formed according the circuit analogue stencil, and whose interaction is depicted in the coupling diagram (Figure 36). These network functions are summarised in a coupling matrix C which represents the interaction steps between each carrier and each process.
…………………………….. (5.14)
Figure 36: Coupling Matrix (C) Relating Input Flows (P) from Different Hub Connections and carriers (α..ξ) from different paths (i…n) to Outputs to Other Hubs (L) ( after Geidl and Andersson, 2005a)
These coupling relations represent impedance operations. The diagonal elements represent the single transformation path of direct transhipments. Off-diagonal elements represent coupling between carrier flows along different paths. These are the indirect transhipments. Geidl and Andersson (2004, 2005b) demonstrated how terminal operations could be modeled as matrices according to Power flow laws. Relating these to impedance constitutional laws must take account of the superposition theory referred to in section 5.2.1.2 on Simplifying Assumptions. Additionally, Page | 190
Chapter 5: Model Formulation of Intermodal Production Systems
the value density information must be decoupled from the total value flow information to track the value streaming processes through each Logistical Unit Process.
Transmission delay effects to facilitate the connection amongst terminals may be incorporated by a preceding connectivity function matrix and a matrix of carrier flows along specific paths i.e. 40ft container, 20 ft container, empty container, partial loads. These carrier types interact through the coupling of the terminals. These flows may cause different impedances at the terminal this can be accommodated by calling another coupling matrix.
5.4.2.2 Filters as Itinerary Control
For the model formulation of continuous approximation to be accurate, s.4.5 identified that intermodal activity had to negotiate an open, complex system through some type of reduction which allowed for closed loop control of certain variables. As part of load flow based design, stencil mechanisms, called filters, control an itinerary through the terminal. This enables multi-path coupling of carrier pairs through the LUPs of a terminal.
Filters are a specific form of characterising the impedance of a Logistical Unit Process (LUP). They act on the synchronisation information of a flow rather than the consignment flow rate or value density to enact the itinerary. That is, filters direct the path of the flow through different unit processes of the terminal whilst enacting other operations like batching, storage and handling. This thesis uses filtering techniques to effect multi-path coupling by routing the itinerary of carrier pairs through a terminal. With filters this thesis incorporates all the phenomena described earlier such as resonance and impedance.
Active circuits such as operational amplifiers (Op Amps) are utilised as building blocks in this thesis for filter impedance. An additional role of Op Amps is to provide feedback lead or lag compensation such that storages and handling are re-oriented to facilitate the changing needs of inflows and load outflows. In electrical circuit control the compensation allows the resonance frequency to be achieved when changing flow signals occur. Similarly, modelling Logistical Unit Processes as compensation control devices allows terminal storage and handling to be altered when mass flow and time-window information of the flow profile alter. The running models in Chapter 6 include the application of compensation mechanisms.
Page | 191
Chapter 5: Model Formulation of Intermodal Production Systems
Design criteria of a terminal anticipates that the source load will correspond to the synchronisation rating (fundamental frequency) of the dispatch train load. This will define the time window which must be met to serve the train load. The train load will also be defined by its value rating (its value absorbed). This constitutes the consignment flow at a certain value density. The flow through the terminal will induce value density drops and storage lags (shown by an alteration of the phase angle).
Complex impedance also has a role in routing freight flow itineraries of a certain transhipment intensity through the intermodal terminal. This uses complex impedance forms in filter design. Filter design using complex impedance can deliver a tighter bandpass or bandreject performance, to meet certain cut-off time windows for instance. This may impact on stability.
The bandpass filter demonstrates the new mathematics of the terminal logistics. It incorporates the response of the unit process. This response may include a range of control activities:
1. Batching operations to increase or decrease the value density; 2. The synchronisation band to pass consignment value to the dispatch payload; 3. Accommodating dwell time change (minimizing evolved storage lag); 4. Control of the consignment throughput rate; 5. The degree to which the value rating of the load is met; and 6. Compensation control to deliver flexible capacity and stability outcomes.
Value intensity is defined as the value density within a specific synchronisation band. The value dissipated at any impedance component due to a synchronisation band may be calculated. This calculation is undertaken using Parseval’s Theorem. This is explained in the following section on addressing the bulk queue.
The payload of a train shipment has attributes of mass value and synchronisation. That is, the payload is effective under a schedule of operation. This payload then may be expressed as a signal. A pertinent analogue is the Fourier power signal expression. The intention is to linking the Fourier power signature of the load with impedance changes enacted through the terminal according to its circuit stencil. These impedance changes include:
• Value change, • Consignment flow change,
Page | 192
Chapter 5: Model Formulation of Intermodal Production Systems
• Storage lag, • Filter business rules.
Facilities design and management can be represented by filter design.
The block diagram of a terminal depicts the multiple coupling of unit processes, linked in the frequency (timing) domain by their synchronisation information.
A Multiple Input- Multiple Output (MIMO) system can be represented at a most simple level by carriers that track through an itinerary in the terminal to meet a train service.
This train service is defined by its:
1. Time window; and 2. Value rating: a. Batched density, b. Consignment flow rate. Each arrival- train service linkage is known as a carrier pair.
The train service with a particular time window has a fundamental frequency or timing which must be met. It may be conjectured that this service has a Fourier signature which defines how inflow signals should contribute to its characteristics.
5.4.3 Characterising Complex Bundling
5.4.3.1 Defining Complex Bundling
The complex bundling problem is an issue that particularly faces metropolitan intermodal terminals: attracting certain base load customers that will support the desired payload of a pulse dispatch by rail. This payload needs to be measured by its value density, consignment accumulation rate and scheduled timing band. The essence of addressing the complex bundling problem is devising a mechanism to measure an individual source flow’s contribution to the payload. What is especially novel in the model formulation of this thesis is the inclusion of impedance which a) affects the individual flow value through the terminal to the dispatch payload and measures the work effort
Page | 193
Chapter 5: Model Formulation of Intermodal Production Systems
(terminal resource consumption) and b) provides an assessment of the flexible capacity of Logistical Unit Processes to meet complex bundling requirements.
The multi-commodity problem can be partially addressed by representing flows with different shippers and carriers (analogous to power sources and power carriers). This allows aggregation and segregation according to the relative efficiency of conversion of each carrier within a processing terminal. Storage effects, terminal work, delay and precision requirements are not designed at this flow level. These impedance attributes are considered at the flux level of decision variables, V, value density of consignments and I, the consignment flow rate. Critical control mechanisms to assess flexible capacity such as bulk queue control, sidings operations and storage-handling balances are assessed using the synchronisation rating (analogous to the electrical frequency domain). This method triggers impedances and needs to be able to combine up to the power flow analysis for ease of intra-terminal and inter-terminal matrix representations.
Origin-destination flows from a number of regions around the terminal may be characterised by key signals with a specific period. Each of these signals may have for instance 3-4 key flow types (frequencies), based on commodity type, and or geographical location. Signals can map directly to a dispatch carrier service or can be broken down to individual commodity flows which are re- summed according to the service for which they are destined.
This process is known as Fourier series analysis. Fourier analysis allows the planner to understand the synchronisation band of matching freight flows. In electrical circuit parlance, this corresponds to the frequency spectrum characteristics of periodic waveforms (Nilsson, 2005). In Fourier series analysis, a signal can be segmented into harmonic signals which are integers of the fundamental frequency or period of the signal.
In a Fourier series analysis, the amplitudes of the frequency sub-signals will be a proportion of the full signal amplitude. Value density analysis is appropriate using a Fourier series when the flows are less than container load consignments to be consolidated at the terminal. Consignment flux analysis using the Fourier series approach is appropriate when sub-flows are container load value and they represent a fraction of the consignment flow of the signal. Signals can also have overall differing value densities according to the commodity they carry.
This analysis can be used for a back-engineering approach: once the time and space dimensions of sub-signals are defined to meet a specific payload an input signal can be developed to meet this payload. This may be one of several input signal forms. This in turn can define the Page | 194
Chapter 5: Model Formulation of Intermodal Production Systems
catchment/storage requirements for terminals to meet a train (or other such pulse carrier) dispatch schedule.
5.4.3.2 Use of Fourier Frequency Spectrum Analysis
The mechanism of Fourier series and Fourier transforms may be applied to analyse the following activities and implications of certain business rules on terminal tactical planning and operations:
1. The proportion of value that an inflow delivers to a dispatch payload and thus address complex bundling requirements, 2. The timing correspondence of inflows and outflows, which can both be represented as pulses, 3. Multipurpose opportunities to service different inflow and dispatch types (the bi-modal overflow problem, for instance) 4. The analysis of circulation rules for empty containers35.
5.4.4 Rail Sidings and Flexible Capacity
5.4.4.1 Carrier Pairs and Sidings Occupation
The relationship between time and synchronisation can be understood through analysis of terminal rail sidings activation and use.
A pulse train flow represents a schedule of desired services as well as a nominal availability of the sidings and mainline. It represents the value density magnitude and periodicity of the service in the time domain. The duration of a train service on the sidings in unloading and loading consignments, its occupation, can be assessed.
This pulse train can be decomposed in the synchronisation domain to specific inflows of services for instance from specific regions Figure 37. The synchronisation domain allows the planner to assess the contribution of different carrier sources to specific train dispatch services. This method,
35 The empty container problem is scoped in Appendix F dealing with possible system coordination extensions of this thesis. Page | 195
Chapter 5: Model Formulation of Intermodal Production Systems
appropriated from electrical circuit and signal theory, is considered appropriate in addressing the multi-commodity flow problem in the freight task and its extension for tracking complex bundling through intermodal terminals.
Where a sidings configuration, stacking strategy and/or operating business rule excludes the full provision of the train schedule, some carrier-pair relations are not carried through. Additionally if the distribution of arrivals for that train service are too dispersed, the train service will either go under-utilised or have need of a top up from available empty containers requiring repatriation to the seaport.
Figure 37: The Relationship between Time and Synchronisation Domains: Carrier Pairs and Rail Sidings Occupation and Train Dispatch
Page | 196
Chapter 5: Model Formulation of Intermodal Production Systems
The pulse train signal in the time domain can be represent as a Laplace transform (Bobrow, 1987) :
Figure 38: Pulse Train Signal in Time Domain (Bobrow, Figure 11.3. p.506, 1987)
The Laplace transform is:
e-1 -sa F(s) = -sT (Bobrow, p.506, 1987)…………………………………. (5.15) s(1 )e-
The relationship between a train service and carriers that feed it can be represented in the timing domain by a Fourier Series of harmonics (Ziemer, 1993) :
N 2πn )( += ∑ CCtf cos( t + φ ) -…………………………………… (5.16) 0 n=1 n T n
(Ziemer, Tranter, et al., Table 3-1, p.117, 1993)
Filters are non-causal mechanisms in that they are control elements in a terminal set by the operating business rules. Their control is denominated in the frequency domain. Filters, either bandpass or bandreject proscribe a band of frequencies that are passed through or excluded from the LUP or service. Design of these filters, using bode plots, gives a fair indication of the carrier pair relation that will be passed with the resulting terminal level of service. The Terminal level of service is defined on the basis of the payload: value density gain and service headway met, as well as net value density losses.
Page | 197
Chapter 5: Model Formulation of Intermodal Production Systems
5.4.5 Bulk Queue Control
Pulse correspondence was a term coined by Dejax and Bostel (1998) to synchronise consignment transhipment between carrier pairs of train services and are characterised by pulse flows. In this thesis, pulse flows between carriers can be matched according to their contribution to the load Fourier signature where subsidiary pulse flows are harmonics from the fundamental load timing (frequency). A measure of pulses, value intensity, can be used to assess when there are sufficient containers to be batched onto a train load. An alternate time based trigger may take precedence as a business rule as metropolitan trains have departure schedules to which they must adhere. Where the value intensity for that service time band is insufficient, this would herald a less than full train utilised.
The crux of the bulk queue problem is assessing the trade-off between the accumulation costs of awaiting additional consignment inflows verses the utilisation of the dispatched load. The control law open to terminal operators is either triggered by time or accumulation amount. For metropolitan intermodal terminals, a scheduled based trigger is more appropriate as train services must operate within strict time windows over scarce mainline paths. The interplay of bulk queue forces are conventionally depicted as a step change between consignment accumulation and dispatch.
The use of the frequency domain allows the critical dynamics to be observed based on a synchronisation rather than elapsed time trigger. Namely, a build-up of value over the timing band of operation may be observed and it may be compared to whether this meets the value load criteria for dispatch. The synchronisation band can map to the active arrival duration for a specific train carrier over which consignments are stored. This timing band and can consist of contributions from numerous source carriers (incoming trucks, for instance).
There are two metrics which may be gained from the analysis. The first evaluates the value passed in the active arrival period against the total value that could be dispatched: this is the percent utilisation of the dispatch load. The second, called the value intensity, considers value per timing unit for that band. This allows the planner to assess the distribution of the arrivals during the active period and whether there are supply side (more storage infrastructure, more regular train services) or demand- management initiatives (such as re-organisation of arrivals) required. These two metrics can be observed from the three-dimensional conceptual sketch of Figure 37.
Page | 198
Chapter 5: Model Formulation of Intermodal Production Systems
The method to assess consignment carrier pair mapping dynamics through the terminal is proposed as follows:
1. A signal entering a stencil is defined in exponential form, -jnωt f(t)= ΣCne ……………………………………………(5.17)
2. Parseval’s Theorem is applied to obtain the value intensity distribution of the flow signal, 1 = ∫ 2 |)(| dFW ωω Π ……………………………. (5.18)
3. The value of the load requirement is calculated, 4. The function representing the stencil impedance is calculated at the passband defined to get the passed value, 5. The passed value is compared to the load value to get a percentage utilisation of dispatch, 6. Application of Parseval’s Theorem allows the planner to consider the possible value that can be captured by the train dispatch if the synchronisation band within active arrival period is altered, 7. In a multi-commodity approach we can calculate how different carrier flows contribute to a specific train dispatch. We note that each carrier pair is a harmonic from the fundamental frequency (timing) of the train dispatch. Only these timing pairs interact to produce the value for that specific train service. The total value of dispatch output from an intermodal terminal can be considered as the sum of average powers associated with each frequency:
P= VdcIdc+ Σ((VnIn)/2)cos(θvn- θin)………………………(5.19)
The harmonic relationship of the source carrier flow to fundamental timing rate (frequency) of the dispatch service is captured in the Fourier Series expression:
F(t)=av+ΣAncos(nω0t-θn)...... (5.20)
Page | 199
Chapter 5: Model Formulation of Intermodal Production Systems
5.5 Specifications
5.5.1 Flow Arrival and Dispatch Signals
In formulating the Fourier transform of these flow signals, the key criteria are:
1. Value density amplitude (batch density of each container), 2. The container handling rate, 3. Dwell time duration of a number of source signals until dispatch.
A pulse train signal is a periodic sequence of asymmetric pulses. The Fourier exponential series of this signal can be represented thus:
Aτ Π− −1 = τ 2 nftj 00 = X n sin cnf 0 ; Tfe 00 (5.21) T0 ……………………………
Where n is the number of terms in the trigonometric series describing the pulse36.
Equation 5.27 allows conversion of time domain information into synchronisation domain coefficients for the dispatch flow. Following Figure 38, A represents the magnitude of the flow
(value density); ı is the period the train set occupies the terminal siding; and ı ı is the interval between train set arrivals. This transform contains information regarding the duty factor of the pulse: period of signal re-occurrence to width of signal. The desired payload dispatch signal accounts for siding utilisation by this duty factor. Sidings utilisation is the critical limiting capacity supply element in the dispatch of loads. That is, the timing of dispatch payloads must align with the time window granted by the schedule which includes sidings and mainline access37.
36 This will be used to describe both the carrier inflows that make up a specific dispatch carrier service and the intra-modal pulse correspondence. 37 The concept of the schedule is extended to include the time windows available at the sidings. Normally private operators must buy access to the mainline (through take or pay or other arrangements) and make their own internal arrangements regarding sidings availability. This formulation allows the planner to consider the time window to the mainline issue are integrated with sidings configuration and use. Page | 200
Chapter 5: Model Formulation of Intermodal Production Systems
In the synchronisation (frequency) domain we can assess how a number of these signals make up the final payload of a specific train service. This is the bandwidth of that train service. A long bandwidth due to a wider distribution of loads to be dispatched will utilise the rail siding and crowd out opportunities for other services. Synchronisation domain analysis captures the multi-commodity flow problem, provides easy identification of response generated and allows for compensation control to improve space management to cater for flow precision requirements (tense flux).
5.5.2 Design according to Cost-Quality Ratios under Tense Flux Conditions
The cost-quality ratio, or Q-factor, is a critical Figure of Merit in the operation of metropolitan intermodal terminals. It allows the designer to define the manner in which load is delivered to a dispatch carrier. The Q-factor is a ratio of value stored to value dissipated. For a high ratio, the load that needs to be stored is not being significantly dissipated in value due to handling costs. Additionally, a high Q factor specifies that the active arrival area that must be reserved for each carrier dispatch is lower. In order to supply sufficient and correct flows for dispatch schedules, there may be a need for terminal storage. A high Q factor designates high storage productivity. For lower throughput requirements, a lower Q-factor may suffice. Consequently, the Q- factor may be seen as a figure of merit in defining the design performance of a logistical unit process under certain tense flux conditions of the flow source load.
The Q factor is a compilation of the Figures of Merit presented in polar form in s.5.4.1.3. The value stored is a function of the value density stored, the rate of accumulation, and the saturation inertia. The value dissipated, or its obverse, the value delivered to the load, is a function of the batch gain, relative and fixed costs and the throughput rate of the Logistical Unit Process.
The Quality-factor (Q factor) can specify the design configuration of each Logistical Unit Process. Q-factors can be designed according to the function of the stencil and become part of the composite transfer function that represents the stencil. The flow response of these stencils can be represented in Bode plots. The Bode plots show the magnitude and phase angle of either state variable against the synchronisation band.
The Bode plots can be derived from the composite transfer function as follows:
Page | 201
Chapter 5: Model Formulation of Intermodal Production Systems
The roots of the characteristic equation of the transfer function, the poles, can be represented as:
2 2ξω ss ω 2 =++ 0 nn …………………………………………. (5.22)
Where 2ξ is the damping ratio. As the damping ratio becomes a higher proportion of Q, the magnitude gain declines .
This coefficient has a critical role in determining the signal response (transient behaviour). In relation to ω:
2 < ξω If 0 2 the roots are real and distinct and the response is overdamped
2 > ξω If 0 2 the roots are complex conjugates and the response is underdamped
2 = ξω If 0 2 the roots are real and equal and the response is critically damped
If the response is incorrectly damped, instability could arise.
In forming the Bode plot, the denominator and numerator of the transfer function can be factorised into poles and zeros respectively:
Π m + Kszi=1 ()i Gs()= sspl Π nl− ()+ i=1 i …………………………………………….(5.23)
This can be further factorised:
⎡ s ⎤ Π m + ⎢ = ()1 ⎥ KzΠ m i 1 z = i=1 i ⎢ i ⎥ Gs() nl− Π p ⎢ − s ⎥ i=1 i sl Π nl(1 + ⎣⎢ i=1 p ⎦⎥ i ………………………….(5.24)
The Bode plot can be formed from the logs of each pole and zero:
m jωωm j ω =+ +−+ 20log10 Gj ( )20 log10 K ∑∑20 log10 1 20log10 1 ==z p i 1 i i 1 i …………………………..(5.25) Page | 202
Chapter 5: Model Formulation of Intermodal Production Systems
The bode plot alters in magnitude gain according to the strength of the Q factor.
Similarly, the phase plot is formed by deducting the sum of poles from the sum of zeros:
ω − ω A broad bandpass filter, having a low Q factor can be made of a combination of low, high pass and gain filters in cascade. A high Q filter has an altered stencil morphology, with a storage element (capacitor) in both the forward and feedback loop. This makes for complex conjugate poles which allows for narrowband active arrival areas for each dispatch carrier. Design synthesis using process control techniques of Bode plots is flagged in the concluding chapter for further research. 5.5.3 Analysis Using Bode Plots Output is a function of input through constitutional laws of conservation of value and flow. An impedance function can relate outflows with inflows and changes in value density over a unit process in the terminal. This impedance function can be formed by the equations of state from the constitutional laws. These are in the form of Laplace transfer functions. Factorisation of these functions leads to identification of poles and zeros which can be used to assess: 1. The shape of the response at the given transhipment intensity; 2. The impedance – gain relationship as a fraction of the transhipment intensity at the static operating point; 3. The evolution of the storage (resource consumption) lags (the phase response). 4. The stability of the terminal response with source profiles; 5. Design opportunities in improving performance through parameter alteration, stencil re- drafting and compensation control. The use of poles and zeros to predict a Logistical Unit Process’s (LUP’s) response is outlined in Appendix C and the effect on value Density Magnitude- Synchronisation rating is depicted in Figure 73. Desirable outcomes depicted by Bode plots can also be used for filter synthesis. Thus Page | 203 Chapter 5: Model Formulation of Intermodal Production Systems terminal activities (configuration and their parameters) can be designed from preferred outcome dispatch profiles38. The number and location of poles critically affects the stability of the process-response and can be used in design of Logistical Unit Processes (LUPs). The pole effect can be depicted in a bode magnitude and phase plot. Phase shifts become critical for stability and the phase margin (the amount above -180 degrees) is a measure of oscillation of the impedance process. Each pole adds a break point in the magnitude plot which is a gradient of 20db/decade and a phase shift accumulation of 45 degrees. One metric of stability mentioned in the section on transient analysis was percent overshoot, a significant measure of storage capacity limits. We noted that damping ratio was a critical parameter in LUP design, related to bandwidth and Q-factor. In plotting curves of percent overshoot and phase margin against damping ratio, we can find the required damping ratio and phase margin required for a specified percent overshoot Figure 39. This in turn determines the impedance parameters of the stencil and the physical dimensions and business rules of that LUP. Figure 39: Determining Design Parameters from Stability Specification The damping ratio is derived from the characteristic equation from the impedance function: 2 2 2ζω ss ++ ω with =Φ −1 ξ NN M tan )2( ………………………(5.27) 38 Kuo (1996) observes the difficulty in filter synthesis with complex poles. Resolution of this challenge is beyond the scope of this thesis and is flagged for further work. Page | 204 Chapter 5: Model Formulation of Intermodal Production Systems 5.5.4 Figure of Merit Certain figures of merit assist with improving design of the freight terminal: 1. Assessing the flexible capacity of our unit process design with changing transhipment intensities (this is depicted in a phase-angle, frequency chart); 2. Use of phasor diagram to depict opportunities to reduce impedance and the value density drop with injection of capacitive and inductive loads (in parallel); 3. Trade off in handling and storage resource costs with changing transhipment intensities. A key Figure of Merit in freight terminals is the number of lifts required per container. With consignments of increasing dwell time (lower transhipment intensity), the storage demands increase and there are more reshuffling moves required to access certain containers. An associated Figure of Merit is the Quality factor which relates the value stored with the value dissipated. In electrical circuit theory this is shown in equation 5.28. Q= Pstored/ Pdissipated……………. (5.28) The Q-factor may be seen as an indicator of the cost- quality jump. Tense flux adds to costs but it is also can add considerably to quality in timely translocation of consignments. The quality factor is then a ratio of the degree of transhipment intensity to the synchronisation profile permitted. The synchronisation profile is demonstrated in Bode plots and may be mapped to the active arrival duration profile of source load consignments. This productivity factor can be used to represent storage utilisation: consignment throughput rate/land area which is another key Figure of Merit at container terminals. Table 11 depicts how the output variables of this model using the two domains of time and synchronisation (frequency) map to Figures of Merit of terminal logistics information and physical performance. These also act as measures of the inter-temporal storage and service generation opportunities across the terminal network (see Appendix F). Page | 205 Chapter 5: Model Formulation of Intermodal Production Systems Table 11: Mapping Freight Terminal Performance to Circuit Model Specifications and Performance Figures of Merit Freight Terminal Model Specification Logistics Meaning Logistical Design Design Figures of Merit Performance Figures of Merit (stencil) and Trade-Offs Value Stream Mapping Fixed and variable costs Value stored to value Cost/ slot utilised Q-factor and Bode Plot design dissipated Quality-Cost Ratio Q-factor Value Stored/payload Batching Condition V0/Vi Value Density Gain Gain margin Flow condition I0/Ii Consignment flux, Throughput/ slot reserved I/Bandwidth Accumulation rate, consignment flow System Efficiency ratios V, I, R Truck and system Actual productivities efficiency/theoretical efficiency for truck capacity type Terminal Efficiencies C,L,R, Storage and Handling Cost and Space Resistive and reactive efficiencies Management trade-offs impedances Work and Effectiveness Throughput capability Load impedance Resource consumption and Throughput/ slot reserved Resonance matching through maximum efficiency complex conjugates Meeting scheduled service Power load rating- Dispatch payload Synchronisation band and V,I, frequency band and requirements Fourier signature prescribed carrier pair Fourier spectrum analysis synchronisation rating Timing band of operations Bandwidth filter Synchronisation period mapped to active arrival Page | 206 Chapter 5: Model Formulation of Intermodal Production Systems period Storage buffer and V reactive impedance Storage reserved Bandwidth x Value Intensity; capability Gain Bandwidth Product Storage Saturation V reactive impedance Increasing stack Slots utilised Phasor Storage lag x Value Density inaccessibility Diagram Value Intensity Power transmitted to Payload; Measure of Value Bulk queue control Q-factor and bandwidth power absorbed in required to activate bandwidth transhipment Control Stability39 Lag and Lead Storage and handling Storage stability and Percent Overshoot; Damping compensation control resource injection flexible capacity factor and phase margin Multi-commodity flow Power carrier pairs Level of complex bundling Catchment/Storage Value intensity of source routing requirements signals Multi-path coupling Port parameters; Flux variables of Logistical Determining impedance Combined impedance transfer complex conjugates Unit Processes according to critical path functions of terminal unit processes Storage and Siding Power transmitted to Value intensity Terminal capability given Q- factor design and Capability power absorbed in headway constraints Bandwidth control bandwidth Bulk Queue Control Power intensity, Value over the active Percent Utilisation of the Bandwidth arrival duration (time dispatch load window) and trigger Value stream control Q-factor Level of Service control Terminal productivity Complex bundling and controllability 39 Transient stability measures are discussed in Appendix C. They are beyond the scope of this thesis and are flagged, with steady state measures, for future research in determining design parameters for terminal LUPs. Page | 207 Chapter 5: Model Formulation of Intermodal Production Systems The essence of making close intermodal terminals attractive is to enable a quality-cost jump in production. That is, to add value in consolidation opportunities and other value adding such as consignment ordering greater than any marginal increase in cost. To achieve a certain throughput production for a certain timing attribute, storage is required to buffer disparate flow rhythms and achieve specific consolidation gain requirements. The quality-cost ratio depicts the relationship between storage and its cost. A higher relative cost for storage activities leads to a reduction in gain magnitude and an increase in value bandwidth. This value bandwidth specifies the synchronisation band to pass the consignment to the scheduled dispatch load. If this gets broader, the gain is lower, referencing a poorer quality ratio and cost rises for same or lower storage capability (less efficient storage or more storage reserved relative to utilisation). Consignments with a generally higher dwell time will require a larger storage area for the same throughput requirement (Iout). Throughput increases are limited in such cases (value density gain is limited and/or consignment flow is limited and there will be an increase in resistance for the same reference storage quantity). For instance, earlier estimates for the Dandenong area intermodal terminal had large average dwell times for containers and thus large storage area requirements for not more than two, 60 container load train services per day. More recent estimates of up to 300,000 container movements by rail per annum to achieve the quota of 30% of international containers by rail, will require 8 or more return services per day by 2030. This will require a radical change in roadside logistics and much swifter terminal throughput i.e. dwell times of hours rather than of days. The Q-factor may be seen as a productivity of terminal operations in resonance. The Q-factor maps to the space productivity measure of throughput/unit area. A high Q corresponds to a greater storage efficiency and/or swifter throughput. High Q factors would be generally associated with direct transhipment intensity and dry docking operations where there is minimum dwell time/ occupation of storages. Page | 208 Chapter 5: Model Formulation of Intermodal Production Systems 5.6 Chapter Conclusions This chapter has presented a theory of logistical impedance to address handling-storage phenomena in an intermodal terminal as well as consider the interface with the gateway network. This logistical impedance mechanism is mathematically formulated as a transhipment calculus considering value and precision relations within and among Logistical Unit Processes (LUPs) that join the container production system with the transportation system. This sketch mathematical formulation will further be applied in the next two chapters where case studies of the landside international seaport container task and the waste transportation task are considered. These case studies illustrate the attributes amenable to this model formulation. Some parameter identification is also made. The following case studies also demonstrate the development of Figures of Merit which guide the prototyping of control laws. This chapter has proposed a number of novel efficiency and effectiveness Key Performance Indicators. These become design criteria through drawing up novel Figures of Merit. These are summarised in Table 11. These design indicators consider how space management control of logistical unit processes can address input and output flow profile requirements. They represent aspects of transformation flexibility that relate the terminal with its system. The electrical circuit analogue is applied to devise these Key Performance Indicators and Figures of Merit. Thus it is demonstrated through these suggestive points of contact that the analogue has value in characterising several freight intermodal terminal operations. Page | 209 Chapter 6: Satellite Terminals to a Seaport 6 Satellite Terminals to a Seaport 6.1 Introduction 6.1.1 Chapter Guide The objective of this chapter is to contextualise aspects of the model formulation of the transhipment calculus presented in Chapter 5 for the Australian conditions of seaport-landside interaction. This chapter outlines how the novel model formulation of the transhipment calculus can lead to the development of figures of merit dealing with terminal space management. These figures of merit may point to design trade-offs in achieving functional requirements relating to an improved value proposition, precision and controllability in urban intermodal operations. Urban intermodal terminals offer significant productivity benefits in seaport container throughput. The beneficial functions terminals provide include satellite buffer storage, consolidation, and improving back-loading rates through routing truck trips. Seaports which impose asymmetrical flows on their landside networks such as Sydney (net importing) or Melbourne (net exporting) have problems in balancing flows and there is significant road traffic activity due to transient storage of containers destined for the seaport and the repositioning of empty containers. For any intermodal system to take up a significant share of the urban landside container freight task, it must be able to cater for frequent train services. This has implications on terminal configuration and the practice of operational business rules. This dilemma is unique for urban intermodal activity. Three case studies are described of intermodal terminals currently planned in Melbourne and Sydney. Their system formats represent a combination of terminals and rail operating form specific to the Australian urban condition. These formats reflect line-haul distances to the seaport, shipper locations, available land area, a tendency towards saturation in the road network at critical junctions and the current logistics structure of container freight. The business rules and terminal configuration combinations presented are not intended as optimum solutions. They are used to demonstrate the spectrum of intermodal urban retrofit opportunities and their requirements and to suggest how the transhipment calculus tool can support decision-making by identifying feasible solutions. Page | 210 Chapter 6: Satellite Terminals to a Seaport This chapter demonstrates the capability improvements at terminals of applying alternate business rules. It is demonstrated that applying sidings activation and bi-modal overflow options can increase the productivity of the terminal as these mechanisms provide both sufficient transformations and throughput for more intensive train service schedules whilst controlling stack overflows. System productivity improvements with truck fleets interfacing the terminal can also be achieved. Pertinent phenomena to illustrate in seaport-hinterland interaction include: 1. The interaction of sidings capability with train operating forms; 2. Stack management at terminal level through bi-modal diversion or use of prioritisation to control stack at seaport; 3. Complex bundling opportunities and requirements. 6.1.2 Novel Aspects to Illustrate Modelling Prototype Attributes of various system formats are investigated for information needs and model application in sites that are planned for development in Australia. The Altona cluster of terminals is modelled for assessment of multiple access to the mainline and seaport precinct interaction (direct connect shuttle services in series). Dandenong terminal is modelled for capability with altered sidings configurations (alternate rail operating forms with bi-modal operations). Enfield terminal arrangements are outlined to anticipate the tool effectiveness in assessing complex bundling phenomena (urban terminal-hinterland interaction). All of the system formats presented here as based on Container on Flat Car (COFC) or wagon interface technology. The investigations using the transhipment calculus concept on real planning situations are summarised in Table 12. Page | 211 Chapter 6: Satellite Terminals to a Seaport Table 12: Seaport-Hinterland Attributes to Investigate Investigation System Format Altona- Feeder-Trunk and liner Impedance of terminal clusters linked by a corridor system Prioritisation and relieving storage requirements at Seaport Altona- Feeder-Trunk Stack-siding interface; Siding Capability and alternate rail Dandenong Direct connect with operating forms alternate rolling stock Bimodal diversion to alternate carriers in order to manage Dandenong Direct connect and the stack allocated routes Introducing multipurpose terminals (multi-path coupling Enfield- Collection-Distribution with multi-commodity flow) Hub Enfield- Collection-Distribution Complex bundling of multi-commodity flows Hub 6.1.3 Role of Intermodal Terminal Satellite Conventionally at a seaport there is a different dynamic associated with temporary storage of importing and exporting containers. Shipping companies are keen to have importing containers removed as swiftly as possible from their vessels. They are not interested in how they are retrieved by importers once unloaded. Average dwell time is unknown. Thus there can be stack formation where later retrieval may involve multiple shuffles depending on stack height and the number of ships arrivals in an interval. Strategies to control the handling effort can range from non-segregation to segregation strategies according to age structure of consignments (De Castilho and Daganzo, 1993). Page | 212 Chapter 6: Satellite Terminals to a Seaport Export containers can be more ordered at the seaport according to the specific vessel they are to be loaded onto. The reserved space can depend on the active arrival duration of containers destined for a specific vessel. Strategies to control the duration, and thus area, can include use of a mixed rough pile for transient storage prior to dispatch to delay use of a designated reserved area to reduce the storage area. This increases the additional handling moves (Taleb-Ibrahimi et al., 1993). The use of a satellite intermodal terminal can further modify these dynamics. For imports arriving by vessel at a seaport, containers can be loaded either directly to a train or to a stack that has a designated rail destination. Transhipment to an allocated stack would reduce the handling moves per container in retrieval (also called shuffles). This represents an improved segregation strategy. The importing terminal dynamics then occur at the satellite terminal. Additionally improvements can be made with a higher proportion of direct transhipment occurring at the satellite (rail head to truck). Where a temporary stack is required, the limited size may not allow a segregation strategy and a conventional mixed stack may be necessary. When a trunk and feed arrangement is in operation for distributing import containers, as along the Altona corridor, the seaport has an additional role in ordering consignments for a particular service on that particular corridor. The delay due to calling into each terminal must be limited. Exports at the terminal can be managed by decentralising the seaport storage-handling management function. Consignments are then shipped by rail in an ordered form that minimises their dwell at the seaport prior to being shipped. Care must be taken that this does not increase the demurrage at intermodal interminals. 6.2 Development of Container System in Sydney and Melbourne 6.2.1 Container Flows and Forecasts In both Melbourne and Sydney, more than 80% of international containers have an origin or destination within 40 km of the seaport40. Table 13 depicts a possible freight forecast of container flows by activity centre and mode for Melbourne with the introduction of significant intermodal 40 This is strictly a beginning or end trip. Ultimate origins and destination of consignments which are less than container load may be outside this radius. Page | 213 Chapter 6: Satellite Terminals to a Seaport operations to capture up to 30% of container flows onto rail (35% for Altona and 28% average overall by 2035). Table 13: Container Freight Forecast by Activity Centre and Mode Opportunity in Melbourne (in 000’s TEU/yr and percentage of total flows) (DoT, 2008b) 2010 Rail 2015 Rail 2025 Rail 2035 Rail (000’s TEUs/yr) % (000’s TEUs/yr) % (000’s TEUs/yr) % (000’s TEUs/yr) % Total Rail Total Rail Total Rail Total Rail North 283 0 0% 379.3 36 9% 616 116 19% 974 197 20% (Hume)` West 651 17 3% 873.7 153 18% 1517 528 35% 2565 867 34% (Altona) Southeast 579 0 0% 714.3 96 13% 1136 313 28% 1812 483 27% (Dandenong) Other 95 0 0% 135 0 0% 215 0 0% 153 0 0% Metro Total 1608 17 1% 2102 285 14% 3484 957 27% 5504 1547 28% 6.2.2 Misalignment of Functional Relations In order to show the usefulness of the transhipment calculus theory described here, the current situation and the change is described for Melbourne. Land use and transport conflicts between the seaport and landside interface have been documented since the 1970’s (Rimmer and Tsipouras, 1978). There are system disequilibria in the flows of import and export containers and empty containers. There are different containers used for imports and exports. Because of this, as well as the uneven spatial and temporal dispersion of these shipments in cities, empty containers availability is mismatched across the network. Up to 50% of road container transportation under current arrangements involves the re-positioning of empties. There is also a load demand for empties to return to the seaport and be repatriated to those countries, such as China, from whom Australia imports. A large proportion of empties are stored in the port precinct. This has led to a tension between improvements in the hinterland logistical structure and fulfillment of shipping companies load demands for empty containers. Page | 214 Chapter 6: Satellite Terminals to a Seaport The system can also be characterised as a profound mismatch between commercial and operating relationships. At the port there is a large proportion of partially loaded and empty truck movements. That is, trucks make deliveries without a return load or make pickups without a delivery. This is due to poor interaction of importing and exporting container booking systems. More significantly the commercial relationship between carrier and forwarder (shipper’s agent) dominates over the operational relationship between carrier and stevedore. This leads to an unwillingness for actors to optimise outside their own environment rather than across the supply chain. Due to the mismatch in operating hours between collection at the port and delivery at the importers’ warehouse, there is considerable overnight staging (transient storage) at transport depots. The capacity is estimated at 29,000 full container slots and 41,000 empty container slots (VFLC, 2008). This prevailing condition under the all road system without rail-road intermodal terminals represents a dissipative, inefficient use of resources. These are depot intermediate locations and represent further inefficient use of transport resources (vehicles and containers) and reduce the ability to achieve high back-loading rates. Space resources are consumed and additional consignment waiting time is endured due to mismatch in headways between landside and port vessel dispatch. These costs represent further system inefficiencies. It is useful to identify the causes and impacts of this current predicament to show how the mechanisms afforded by urban intermodal terminal can alleviate inefficient container staging. Intermodal terminals can improve headway matching between landside and vessel flows using pulse flow correspondence and consignment ordering leading to improved consolidation as well as improve accessibility to the seaport through providing alternate rail shuttle pathways. 6.2.3 Container System Capacity in Melbourne There is a growing impact on the road network due to freight movements in the inner west corridor. There is a cost to maintaining roads, health and amenity impact with container trucks, and greenhouse emissions. There will be growing congestion on the road network with growing passenger and freight task as the road network nears saturation. This is particularly the case in the inner west road corridor. International container traffic is of high value. Fostering an alternative rail mode on an available rail network would provide alternate capacity to the East-West road network for this high value freight. Page | 215 Chapter 6: Satellite Terminals to a Seaport The performance of freight along the western road network will deteriorate with growing congestion. Results from the strategic Melbourne Freight Model (DoT, 2008c) indicate that at key links, such as parts of the Princes and Westgate Freeways, the Westgate Bridge, and Footscray Road, access to the seaport precinct will become saturated by 2021 in peaks of increasing duration. Without linkages to an alternative mode developed, the costs of road haulage of freight to and from western Melbourne will increase significantly. The operational capacity of the Port of Melbourne is increasingly constrained. Assessment of port productivity should be considered as a system; the port and its landside networks. Landside accessibility of the seaport is a critical component of operational capacity and seaport productivity. Daily truck flows at critical junctions into the port precinct are depicted in Figure 40 (DoT, 2007). Loaded and empty containers as well as non-container loads are depicted with critical congestion points across the Maribrynong River, entering Swanson Dock and entering Appleton Dock across Footscray Road from Dynon terminal. There are also bottlenecks at the West and East ends of the West Gate Bridge. Significantly empty container flows are nearly as heavy as loaded flows. The seaport-road access capacity is estimated to be close to operating capacity and this will worsen, adding considerably to access waiting times and jeopardise truck ability to meet their scheduled access time windows. This is affected by the current peaking nature of the road based logistics system. The roadside access to the seaport is regulated by the Vehicle Booking System. This currently experiences severe peaks between 9 am – 3 pm. While trucks occupy most of the access timeslots, they are of varying container capacities and less than 40% carry return consignments. Thus only 51% of possible container truck slots are utilized due to less than full loading of containers per truck. This may be seen as symptomatic of not only high peak flows but also poor loading rates41. This is due in part to the mismatch of operating hours between freight forwarders and shippers. In turn, this reduces the access capacity of the port-landside road system. There are some initiatives in communications and pricing to achieve a demand flow smoothing of trucks entering and departing from the port precinct. A larger problem is the dissipative container logistical structure that does not maximize the loading of trucks across the network. The poor alignment of shippers and stevedore needs leads to very low back loading rates of trucks arriving and departing the seaport. Urban intermodal road-rail connections offer several opportunities to address these problems: 1) they will allow an alternate access to the seaport, not subject to gate 41 Average truck loading capacity is 2.2 TEUs. Average truck loads are 1.17 TEUs (VFLC 2008. Truck Optimisation Plan: Options Paper. Melbourne: Victorian Freight and Logistics Council. Page | 216 Chapter 6: Satellite Terminals to a Seaport bottlenecks; 2) they offer consolidation activities; 3) they offer a hub for higher loading of regionally specific container truck traffic. Figure 40: Heavy Truck Flows at Critical Junctions and Port Gate Bottlenecks in weekday morning peak around Port of Melbourne (April 2002) (DoT, 2007) In addition to landside accessibility, a second aspect of seaport operational capacity is container storage capacity of empty and full containers. Land use for empty and full container storage at the seaport is at a premium. The average dwell time of containers at the seaport is 3.5 days. Improved productivity at the seaport is expected to increase with crane rates increasing from 29 to 32 lifts/hr and berth productivity to increase from 850 TEU/metre to 1500 TEU/metre in the next 10 years (PoM, 2006). Overall seaport productivity is expected to increase from 19,000 TEU/hectare in 2000, to 32,000 TEU/hectare by 2015. Large growth projections in the number of TEUs through the Port will still place constraints on seaport precinct storage capacity. The Port of Melbourne have estimated that they will have a long term need for 160 hectares (up from the current 77 hectares). Currently there are no direct container rail services to the Port berths and no metropolitan rail origin- destination services. There are container shipments by rail to the port precinct. These have regional or inter-state origin/destinations and still need to be transported the last mile to the stack or dock by road vehicle. A satellite rail hub connection would allow the use of additional storage capacity outside the port and can assist in reducing dwell times for those containers it handles. Page | 217 Chapter 6: Satellite Terminals to a Seaport 6.2.4 Landside Container System Cost Structures and Value Accounting Through Terminals Figure 41 identifies that the most significant costs of shipping containers are associated with 1) the line haul transportation to and from the seaport; and 2) empty container movements. The line haul costs are based on a return trip. For both Sydney and Melbourne the back-loading is less than 60% and this inflates the cost of transporting containers. This refers to those single trips that have a load average less than 1.5 TEUs per trip. If loading rates could increase, the cost per trip would reduce. Empty container movements involve positioning containers to exporters and returning empty import containers to container parks or to the seaport. Transhipment costs (overnight storage, lifting containers) are much less significant in comparison. A study investigating the financial feasibility of the Dandenong intermodal terminal in Melbourne found significant improvements in hinterland attraction and Net Present Value Returns for a terminal operator when additional functions of empty container storage as well as consignment bundling prior to consolidation on train services were added (DOI, 2003). Whilst the function of stuffing Less than Container Load (LCL) cargo to container can be considered a cost born equally by any freight system format, such a batching function can be a considerable enabler to encourage shippers to utilise an intermodal terminal where warehousing functions can be added to this service42. 42 On this basis, the MIST intermodal terminal at Minto 40km south of Port Botany in Sydney provides the electrical importer Sunbeam with unstuffing and warehousing facilities. Also 1/3 of import container flows at Enfield are planned to be unstuffed and distributed from the terminal. Page | 218 Chapter 6: Satellite Terminals to a Seaport Figure 41: Typical Container Supply Chain Costs for Sydney (SFCNSW, 2004b)43 Significant value stream mapping benefits lie in the batching and consolidation of consignment flows which reduce the freight intensity (km/t). Tracking changes in the inverse relation, the value density, through the terminal, allows a direct comparison with the relative cost of transporting containers by road with associated gateway congestion costs with the cost of delay and 43 The cost structure for Melbourne will slightly differ as road haulage costs are believed to be up to $100/TEU cheaper than in Sydney due to less road congestion (DOT 2009. Altona Intermodal Access Enhancement Business Case. Melbourne: Victorian Department of Transport.). Page | 219 Chapter 6: Satellite Terminals to a Seaport transportation by rail via an intermodal terminal. In the export case the container production cycle of a terminal must match train dispatch cycles. In the import case, train arrival cycles must match dispatch distribution cycles. The matching of these two cycles will involve buffering at the terminal through storages as well as transhipment. The bundling of less than container flows may also assist in supporting high utilisation rates and thus high payloads of train services. This function of complex bundling is a critical aspect of urban intermodal terminal activity. The intermodal terminal can further increase the productivity of the road based container system. Back-loading rates can increase significantly with the development of an intermodal system which fosters new circulation trips of larger capacity vehicles between the seaport and its hinterland. This involves the movement of empty containers through the intermodal terminal. The use of the terminal for bi-modal operations assists in moving containers between the terminal, seaport and shipper. This can close loops in transport flows and minimise trips without containers or full containers and minimise unnecessary empty trips. 6.3 The Altona Cluster for Stack Management at Terminal and Seaport 6.3.1 System and Terminal Descriptions The potential Altona cluster of three intermodal terminals is a trunk-feed network in the sense that the cluster activity supports the intermodal service trunk headway on a common corridor (Figure 42). Intermodal shuttle services must be booked on the corridor in advance. A number of intermodal terminal generator and attractors located on the same main line trunk allow for a certain service frequency to be reliably booked. An extension is that each rail service may stop at more than one terminal towards their final destination (liner rail operating form). This will increase the impedance cost of that service but allows more certainty in utilising the payload opportunity of each service. The Altona case considers the complexity of utilising a common corridor with complementary or same rail services. Three terminals are distributed in the Altona cluster, approximately 3 km apart. The epicentre of these terminals is approximately 17 km from the Melbourne seaport precinct. The Altona cluster is an infrastructure artefact. It might be more efficient to have one highly efficient intermodal terminal at this or at a greater distance away from the seaport. Potential intermodal terminals are however Page | 220 Chapter 6: Satellite Terminals to a Seaport constrained by the available land use. It is also a State and Federal Government intention to foster container freight onto rail in an equitable and competitive manner. These terminal operators also run significant road container operations. It is known that each of these terminal operators have potential base-loads which ensures that they can support some rail service start ups. Although their catchments in the Altona/ Laverton/ Maribrynong areas of western Melbourne generate and attract some 30% of Melbourne’s container freight with peak movements occurring from the morning until 3pm, it is understood that a more regulated timing with rail transportation would be acceptable if imports were available by 11am in the morning and exports were dispatched to arrive at the seaport by the evening of the day they are lodged at the intermodal terminal. These terminals are typically 5-10ha in area. One of the terminals is constrained by its depth from the corridor access and trains of greater than 300 metres in length must be made up by use of a locomotive running around a spare siding to join rakes44 from two sidings. The corridor is a standard gauge line which carries interstate freight traffic of trains 1500m in length as well as some country and interstate passenger services Figure 42. Terminal services will access the corridor via a passing loop which will ensure that services can join the corridor at speed and not delay through running services. Due to signal headway arrangement and sidings length restrictions at the seaport precinct, the maximum length of shuttle train is expected to be 600m which can carry up to 75 TEUs (Twenty Foot Equivalent Unit containers). 44 Rakes are a length of train in a siding that are to be connected to a locomotive. Page | 221 Chapter 6: Satellite Terminals to a Seaport Figure 42: Altona Intermodal Cluster of Three Intermodal Terminals West of Port of Melbourne (DoT, 2009) Page | 222 Chapter 6: Satellite Terminals to a Seaport 6.3.2 Proposed Schedules and Dispatch Signals Rail capacity is defined by the terminals, the corridor mainline, and the siding capacity at the seaport precinct. In order to capture 30% of current container freight movements onto rail, some 12 return services must occur daily. It is estimated that this would necessitate major duplication along parts of the mainline (i.e. between Newport and Footscray). It is also not feasible that terminal operators could switch sufficient base load to rail to this extent immediately, rather it would take a ramp up of some years. A daily return flow of 6 services to be shared amongst the three terminal operators is deemed feasible in the next five years. There are some tight implications with this: train shuttles can make the return trip in 2+3hrs= 5hours. Shuttles have approximately 3 hours to turn-around once they dock at an intermodal siding .Thus each shuttle can provide 2 services for each terminal in a 16 hour period (8 hours for curfew and peaks).This means that containers must be unloaded and loaded during this time. With 50 containers and transhipment with traditional forklift technology taking 3mins/container, two forklifts will be required. This leaves approximately 30 mins for the train to be checked prior to dispatch. 6.3.3 Terminal Transhipment Calculus: Design of Storage Utilisation Terminals are a critical part of the trunk-feed rail-line capacity. The stack-siding interface configuration, the degree of transhipment intensity of consignment flows, and the dispatch load that is required to be met, affect the capability of a terminal to meet performance specification and to respond appropriately to alternate operating business rules. Routing flows through alternate unit process paths, called the itinerary, is a significant capability of terminals. Indeed, the destination within a terminal for containers of a certain group is not static and depends on their and the dispatch carrier’s arrival/departure time profile. Such dynamic assignment procedures can be considered analytically such as in cumulative flow curves (Taleb-Ibrahimi et al., 1993). Dynamic Assignment is a critical consideration in any modelling of an intermodal terminal connected to a network. Mechanisms in terminal space management which track storage and handling trade-offs according to the nature of consignment inflows are critical to this. Page | 223 Chapter 6: Satellite Terminals to a Seaport Each intermodal terminal in the Altona cluster needs to manage its storage operations according to the incoming flow rate, the degree of direct transhipment intensity in the incoming flow (dwell time) and the train dispatch schedule. Consequently the design of effective storage utilisation to meet the carrier services is required. In the electrical circuit analogue, this can be accomplished through the combination of parallel inductive and capacitive loads which the design circuit stencil as the Logistical Unit Process contains. For instance, at the satellite terminal the planner needs to provide sufficient storage such that the batch throughput rate can increase (I), without a reduction in the train dispatch carrier load utilisation (V, value density) or an increase in the source value density (tonnages per truck) which may be limited by physical dimensions, interoperability limits and/or available handling resources. The aim is to design a Logistical Unit Process such that it can handle an increase in storage of consignment value density for the desirable flux. For a certain value density dispatch and a desired batch rate, the stencil unit process (storage buffer) has an inductive-capacitive (net reactive) position which can be represented by a vector between handling and storage flows45. Inductive loads may be considered cross docking operations (direct transhipment). They Lag VL. Capacitive loads are internal storage movements (indirect transhipment). They Lead VL. A simple stencil of this relationship for storage buffer- handling is depicted in Figure 43. An increase in carrier services required (increase in train dispatch frequencies) will require a larger throughput over the storage buffer. This in turn will require swifter handling as well as an increase in the storage pool- this combination of capacity (static) and utilisation (dynamic) may be considered as the capability or storage utilisation of the terminal. R1 L1 1 2 1 V1 + R2 L2 C1 Ia Ib Ic VL 2 Figure 43: DC Mesh Circuit Analogue for Logistical Unit Process of Container Storage Buffer at 46 Intermodal Terminal, inclusive Capacitive Load C1 45 The following discussion assumes a constant synchronisation ration between incoming flows and terminal operations in each scenario. That is, transhipment intensities of flows do not vary within each scenario. 46 The output is VL. Additional unit process elements are added in parallel for this stencil. They act as a load on the output. The influx value density of the average consignment is V1 or Vs. Page | 224 Chapter 6: Satellite Terminals to a Seaport From this condition of high direct transhipment intensity, it may be required to sketch plan a terminal for a change in this tense flux condition to one of more indirect transhipment Intensity that produces less train services . The planner then wishes to retain the dispatch payload of a train 47 service (being the net value density outgoing VL ) and contain the loss in value density and loss in overall value due to any added storage and handling. The interaction of consignment flux and value density changes can be depicted on a phasor diagram. In Figure 44 only the RL circuit is shown (without the effects of the capacitive load of Figure 43). These diagrams are an overlay of the attributes of the cost-value trade off presented in s.5.4.1.3. Figure 44: Phasor Diagram Depicting Impact of High Throughput Direct Transhipment Intensity - Inductive Load (No Capacitive Load) (after Nilsson and Riedel (2005)) Under fully inductive loads there is a higher consignment flux passing over the storage area and out of the terminal (I) which requires a more rapid response in cross-docking type operations. This requires more injection of handling work and thus greater loss in value (Vs-VL). The higher net value density in (Vs) comes from pre-dispatch loading of wagons on available sidings (multiple containers per wagon) in order to meet the higher consignment dispatch rate. The vector quantity jwLI represents the value density change due to handling work. An alternative way to manage the terminal for the same consignment flux would require a much greater source inflow rate. 47 This net value density is the resolution of the batching value of the consignment with fixed operating costs. Page | 225 Chapter 6: Satellite Terminals to a Seaport Introduction of a capacitive load means that more consignments in are now streamed to storage (indirect transhipment). This postponement of dispatch increases the dwell time profile of consignments (and thus a greater storage lag). Addition of capacitive loads (Ic) can reduce the consignment flow out and can reduce the costs (drop in value density) over the storage buffer (Figure 45). Figure 45: Phasor Diagram Depicting Impact of More Indirect Transhipment Intensity- Combined Inductive and Capacitive Load Operations (after Nilsson and Riedel, (2005)) The inductive flows interact with the capacitive flows for more internal movements rather than direct cross-docking operations. The flux I reduces to Inew due to the introduction of a capacitive reactance vector. The reduction in consignment flux through the terminal is marked by greater rate of consignment accumulation. A combination of design inductive and capacitive loads can respond to business rules regarding the space management at terminals. They may act as a means to demonstrate the relative significance of direct and indirect transhipment. For instance, these tools can highlight the value to the system of indirect transhipment in capturing smaller Beginning and End flows where maintaining a constant stream of containers for dispatch by direct transhipment is not possible with the shipper hinterland of the intermodal terminal. This more complex bundling of smaller and diffuse flows would not normally make economic sense to go through intermodal shipment. In a dense urban network, this Page | 226 Chapter 6: Satellite Terminals to a Seaport method does however add to the economies of scale in operation and supply the minimum number of train carrier cycles necessary for commercial viability48. For instance, if the terminal is to capture more Beginning and End links it needs to apply some order to them for later dispatch at a certain value density and precision requirement. There will be a requirement for some dwell time of containers in allowing this consolidation. 6.3.4 Extension: Prioritisation and Stack Management at Seaport Any analysis of terminal dynamics must consider how terminal effectiveness maps to system effectiveness. For instance, space management capability involving the balancing of seaport and terminal resources needs to be measured. An analytical mechanism that relates precision aspects of the train dispatch to the truck arrivals and the vessel dispatch services to the train arrival needs to be devised. Seaports receive container arrivals for each vessel during an active period. This period ties up storage resources which can only be discharged once the vessel is loaded. Therefore the storage required is much more than the distance between the cumulative step departure curve and the cumulative arrival curve. This is depicted in Appendix B. The storage size will grow larger the larger the step size of the departing (and arriving) vessels (from 3,000 to 8,000 TEU, for instance for Melbourne (DoT, 2008a)). This requirement for a larger reservation of storage will accelerate capacity constraints at a container seaport. Techniques in stack management for export containers can include use of a transient, “roughpile storage” ((Taleb-Ibrahimi et al., 1993) A roughpile storage can postpone assignment to the permanent container stack until shortly before the vessel is to depart. Taleb- Ibrahimi et al., (1993) estimate this may reduce overall storage area requirements by up to 30%. There are trade-offs in additional handling requirements. The road vehicle booking system has gone some way in relieving congestion at access gates to the port. It however does not explicitly time the distribution of container arrivals relative to ship departures. Thus there remains a large active arrival time of containers for each vessel class which must be catered for in reserve storage. Additionally, import and export booking systems are not synchronised to improve back-loading of individual trucks. 48 This tool can also enumerate the dissipative effort of focusing on a sole objective of cross-docking speed at the expense of complex bundling (trade-off between Precision and Flexibility). Page | 227 Chapter 6: Satellite Terminals to a Seaport Making the intermodal terminal a satellite distributed storage of the seaport, may improve the rate of dynamic assignment at the seaport. This is through the use of ordering consignments to be dispatched as a consolidated, discrete flow. Containers are dispatched to the seaport when their vessel cut off time approaches. For instance, a train may order its consignments - 60% are to depart within 2 days of vessel departure49, 30% within 3 days, and 10% within 4 days. This reduces the dwell time at the seaport of containers which would normally require a larger active arrival interval for a particular vessel service. There is also the possibility of direct transhipment between rail and vessel. Consequently the satellite storage can relieve both storage size and handling requirement of the system. Such arrangements offer a more precise source response to the departing vessel loading requirement. The necessary business rules at the intermodal terminal require servicing a scheduled cycle of periodic trains. For a certain batch configuration (load utilisation and train length in the Single Input Single Output (SISO) model) there will be a certain batch headway rate (carrier services) that needs to be made-up and dispatched to meet schedule (and to meet the payload). At times the schedule may dictate additional carrier services. The active arrival duration in the time domain can be specified. This is the duration (start-stop) of inflows which are part of a dispatch consolidation service. The service flux characterising the number of services and their occupation of terminal, sidings and track is defined in the time domain. The synchronisation domain (analogous to the electrical frequency domain) is a transform which can be applied to grant information on the timing requirement of these specific flows and thus matching the combination of them which may be passed to the service. The active arrival duration profile is now referenced by the bandwidth. The bandwidth gives information on the inflow classes of service patterns which map to the dispatch service. The frequency domain allows for analysis of timing precision of different consignment flows. These consignment flows have a duty factor (width to period relationship). This can depict the duration a carrier class occupies part of the terminal, such as a siding and therefore terminal capacity, but it does not directly give insight on precision requirements of fitting in other services based on their time sensitivity. The frequency domain analysis allows evaluation of whether additional flows could meet the precision dispatch requirement. This covers both the degree to which both the payload and the planned schedule are met. The bandwidth can be altered on the basis of its value intensity measure. A bandwidth with a 49 Cut off times to receive containers for a vessel service may be a day or more before vessel departure. Page | 228 Chapter 6: Satellite Terminals to a Seaport low value intensity reflects an arrival distribution which is sparse and of low value density. This arrival duration may need to be prolonged to order to capture the minimum payload for the dispatch service. Alternatively we can design an arrival profile which reduces the requirement for a prolonged active arrival duration and thus reduce the storage requirement. With a simple illustration, consider an inflow of one consignment every 3 minutes and another flow profile of one consignment every 6 minutes. The more frequent inflow equates to a train formation every 2.7 hours compared to 5.4 hours for the less frequent flow (considering no other bottleneck processing in the terminal). If the inflow disturbance switched from the second flow pattern to the first, the bandwidth (active arrival duration) could be truncated, releasing storage reserves. The bandwidth control would respond to a more frequent inflow with a reduction in reserved space required and allow for storage of other consignment carrier pairs. Another measure of the benefits of ordering through consolidation of container flows by train services is the reduced work required in managing a less ordered flow. A seaport productivity measure is the average dwell time of the import and export containers it holds. In Chapter 5, the flow as a signal was discussed having attributes of Value density, consignment flow, transhipment intensity, and dwell time. It was presented that the dwell time was analogous to phase angle in electrical circuit theory. Where the dwell time of the source flow does not match the design dwell time of the Logistical Unit Process, a storage lag was said to evolve. This storage lag led to an increase in impedance and cost. This can be depicted in a frequency – phase angle diagram (Figure 27). The mechanism is described as follows. The ideal ordering of the consolidated flow from the perspective of the profile of vessel dispatch services at the seaport will be marked at a certain synchronisation rating (centre frequency). This can be modeled as a bandpass filter stencil. The train dispatch signal to the seaport must negotiate this filter. The deviation of the train service’s synchronisation rating from the ideal seaport rating will be resolved as a phase angle shift (storage lag). This storage lag can be viewed as additional work in accessing the stack for matching consignments and thus an increase in impedance. Minimising dwell time through ordering of consolidated consignment flows will lead to a reduction in dwell times and impedance. Page | 229 Chapter 6: Satellite Terminals to a Seaport 6.3.5 Stack Management Figures of Merit The feasibility of space management techniques can be assessed through the generation of Figures of Merit which depict the trade-offs in certain performance criteria. These criteria can pertain to either prescribed Levels of Service within a terminal or to the objective of matching output load requirements. Levels of service may be represented by the Quality factor. This factor assesses what value can be passed through a Logistical Unit Process with the minimum band of carrier inputs (bandwidth). Control of the active arrival duration allows terminal storage arrangements to sufficiently buffer the carrier inflow profile with dispatch load services according to their headways. Another Figure of Merit is the evolved storage lag. This depicts the degree to which the flow- terminal interface is operating below maximum throughput efficiency and thus storage-handling costs increase. In generating these relationships the transhipment calculus offers a tool in assessing the sufficiency of terminal configuration and operation with alternate train operating forms. 6.4 Dandenong Sidings Activation and Bimodal Overflow 6.4.1 System and Terminal Description 6.4.1.1 Original Concept Design The Dandenong intermodal terminal was originally scoped to serve up to three services per day (default one return service daily) and approximately 70,000 TEUs per year. The analysis suggested that at this throughput level there would be a Present Value return over 20 years of $4.8m for simple load/unload operations, $11.5m with an additional container park facility, and $15.4m with an additional warehousing, stuffing/unstuffing facility. As a minimum, an empty container park holding function was deemed necessary in order to break even or perform better financially against the logistics cost structure of importers and exporters who had to otherwise pay for the empty container to be held offsite until repatriated to the seaport. Page | 230 Chapter 6: Satellite Terminals to a Seaport From a survey of four different user groups of such an intermodal facility, it was found that all would benefit in cost savings per TEU. Importers would be penalised with a delay of 2-4 hours due to a single train service arriving per day before 6am. With more frequent services by rail, this penalty would be markedly reduced. The greatest consumption of land area at the intermodal terminal would be the empty container holding area. For a dwell time of up to 30 days, this would mean a land consumption of more than 11 hectares. 6.4.1.2 Change Envisioned In order for the Dandenong region to capture up to 30% of container traffic on rail, a significant number of train services would need to operate - up to 30 trips per day. Green’s Road terminal as one of two possible terminals in the region, would need to support 8 return services per day (16 trips). As these services must schedule within passenger services on the same broad gauge network, shuttle type rail operating forms are envisioned of short length (not more than 600m) and with acceleration/deceleration characteristics similar to passenger trains so that they may be comfortably schedule in the inter-peak and off peak periods (DOI, 2006)). Further mainline capacity for such service frequency would be required such as at least a partial duplication of the feeder Cranbourne line to provide terminal site access, and a triplication of the trunk Dandenong line (another track). Container storage would necessarily increase to hold approximately 3,600 TEU at any one time (based on a daily throughput of 2,250 TEUs with 80% held for 1 day and 20% held for 4 days). This would require an increase in the originally estimated container handling space from 1.4ha to approximately 5.5ha. Empty container storage would then have to be managed downwards from an average dwell time of 30 days to 5 days and thus reduce the space requirement to less than 6 ha. The interaction between node configuration and network form (type and length of train, for instance) affects system capacity. Port shuttle operating forms can be of a number of types. Each form will affect the optimal design of the terminal and the terminal throughput capacity and efficiency. This has been well characterised by Woxenius (2004) and Ballis and Golias (2002, 2004). Page | 231 Chapter 6: Satellite Terminals to a Seaport Two examples are provided and discussion is made to show how this may affect our model parameterisation. The first is a push-pull shuttle of 600m length. The cycle time for a return trip will be typically 6 hours. Unloading and loading of the shuttle will take 2 hours at either end. For a time window of 18 hours each train set can do 3 return trips. For a planned 12 return trips per day, 4 train sets (and 8 locomotives) will be required. Implications for the terminal interface include: 1. Double track will be required to allow unhindered arrival and swift departure from the terminal. There will be cases when a train arrives shortly before another at the same terminal is dispatched; 2. How will terminal equipment access farthest track for unloading? Will this require another hardstand to access this track? The second example will involve a shuttle system where locomotives are shared between train sets. A shuttle arrives at the destination. The locomotive is unhooked and this locomotive runs around to pick another loaded wagon which is then taken to its destination and the locomotive picks up another loaded set. The run around would take less than ½ an hour. In this way the loading and unloading function is taken off-line from the locomotive cycle time. The locomotive cycle time is then 3 hours instead of 6 and each locomotive can do 6 return trips. For a planned 12 return trips per day, 2 locomotives are required. Wagon set cycling time remains essentially the same. Thus the same number of wagons will be required (25 per set and 4 sets in total). This train operating form reduces the system cost as it reduces the idle time of locomotives. The throughput becomes independent of the number of locomotives. The transhipment calculus is applied for alternate system formats of the intermodal terminal system in the South East. One which caters for a traditional container production arrangement (with a maximum of two return shuttle trips per day) and the other which envisions a more rapid process of up to 8 return shuttles per day. The system format implications for each dispatch profile, based on minimum container storage requirements, track sidings layout, transhipment equipment and consumption of mainline resources are documented in SKM (2003) and further developed by the Victorian Department of Transport. This case study outlines how the transhipment calculus may demonstrate each of these system formats. Page | 232 Chapter 6: Satellite Terminals to a Seaport 6.4.2 Proposed Schedules and Dispatch Signals A proposed schedule for 8 return trips per day using the freighter technology was constructed by Department of Transport staff and is depicted in Appendix D. This avoids movements on the mainline during the passenger morning and evening peaks (7:30-9:30 and 16:00-18:00). These schedules can be further transformed into dispatch signals containing information on value flow, and synchronisation requirements. This is scoped for future work. The schedules are presented to depict how integrated the terminal and rail line capacity assessment needs to be. 6.4.3 Sidings Configurations and Impedance Sidings play a crucial role in the throughput capacity of a terminal. For urban freight networks such as Melbourne and Sydney, it has been identified that the interaction between train length, sidings length and shunting activity can impede optimum rail paths that must share access on the mainline with passenger services (Beavis et al., 2006b p.13) The level of train dispatch activity is affected by the availability of sidings to accept train occupations. This availability will depend on the current and assigned occupation on sidings of scheduled services and the shunting activity which blocks out sidings use. Utilisation is based on available slots that can be scheduled on the sidings and the payload of consignments produced by the terminal. Sidings throughput capacity can be modelled according to the continuous approximation berth utilisation techniques of Daganzo (1990). Utilisation depends on sidings being both available and unassigned50. The use of sidings can be considered an overflow of the stack. This mechanism can observe strategies that trigger overflow to sidings according to alternate inflow sequences and dispatch schedules. Sidings capability is represented by a) the nature of the train operating form, and b) the sidings configuration. The pulse signal represents the payload and periodicity and sidings occupation requirement of the proposed train operating form. Mainline access windows are also considered here. The bandreject filter represents the sidings configuration (Figure 46). This operation takes into account non-available periods and ranges of carrier pairs. Non-available periods can be due to shunting requirements, and availability of wagon resources for loading (COFC) arrangements, and 50 For additional flows to be incorporated in the terminal’s capacity, sidings must be available, that is, not occupied by the currents schedule of services and unassigned to any additional service. Business rules apply different assignment procedures. Page | 233 Chapter 6: Satellite Terminals to a Seaport occupation of sidings due to laden wagons awaiting classification. Significantly, a larger duty factor will demand more sidings utilisation (an extended bandwidth) and less opportunity for new services to be included within the available bandwidth. The bandreject filter also has a role with assessing pulse correspondence under transmodal operations. Carrier pairs which are filtered out will not be directly transhipped to end destination but will be diverted for temporary storage. The sidings configuration can limit the number of services that a terminal can turn around each day. For instance parallel active sidings limit the turn around time and access to each siding thus limiting the through-put.51 Loaded trains cannot be dispatched until they are physically checked after a rake is joined to a locomotive. This takes approximately 30 minutes. Meanwhile the train that has arrived adjacent to the impending train to be dispatched cannot be unloaded until the locomotive run around, coupling and checking has occurred. This potentially adds 40 minutes to a train occupying a siding and prevents access to the adjacent sidings. This reduces the throughput capacity of the terminal. This extends the bandwidth of sidings inactivation and narrows the bandpass for payload activity. C2 C1 U1 2 - 3 OUT 1 + Vo R2 R1 IdealOpAmp R4 V1 Vi R3 C3 U1 2 0 - 3 OUT 1 + R5 IdealOpAmp 0 Figure 46: Stencil Bandreject Filter for Sidings Configuration Constraint Different bandreject filters may be used as tools to operate as sidings constraints to a desired train dispatch schedule. The centre frequency, w0 presented here acts as the transition time and the Bandwidth acts as the transition duration during which no effective interaction with the sidings can take place. Figure 47 depicts the working of the bandreject filter stencil to represent a range of sidings inactivation when the sidings are either occupied or assigned. The range of inactivation 51 Active parallel sidings were originally designed in the concept for Dandenong Intermodal Terminal (SKM,2003), See Chapter 8 for further discussion. Page | 234 Chapter 6: Satellite Terminals to a Seaport pertains to the synchronisation rating values of carrier pair consignments that will not be passed (loaded) onto a train service. DB(V(VO)) 0.0 -10.0 -20.0 100.0 128.5 165.1 212.2 272.7 350.4 450.3 578.6 743.5 955.4 1227.7 1577.7 2027.4 2605.2 3347.7 4301.9 5528.1 7103.7 9128.4 -30.0 DB(V(VO)) -40.0 -50.0 -60.0 -70.0 e Figure 47: Sidings Range of Inactivation (PSPICE output)52 The bandreject filter acts as a control element in the throughput of container train services. This element modifies the train pulse dispatch signal which represents a schedule of trains. As an illustration, two configurations are presented in Appendix D for Dandenong Intermodal terminal. The first has three parallel tracks with one abutting a transhipment hardstand area. Whilst this was appropriate for a small number of train services per day, it was estimated that bottlenecks would arise with 6 or more return services due to sidings clearance constraints. The single sided sidings arrangement impeded swift throughput of services due to delays in train preparation which meant that the other sidings were blocked for access. The second siding configuration was designed in order to improve throughput of train services. In this system format the transhipment hardstands can access two active sidings which permits increased train shuttle service frequencies. 52 This bandreject mechanism indicates the synchronisation range for which the sidings is inactive, a sub-zero value density capture. Consignment carriers that fall within this range will be filtered to another storage zone. The synchronisation rating is in the units of Herz. Their correlation with transhipment intensity and time sensitivity of consignments is left for future research. Page | 235 Chapter 6: Satellite Terminals to a Seaport 6.4.4 Extension: Bimodal Overflow Control Signal flows aggregate to a certain value and source impedance. Increasing reactive impedance designating greater consignment value inflow and work requirements will lead to an increasing storage lag and reduce the consignment flux through the storage. This is a sign that the stack is reaching saturation. Sidings are inactive for loading as there is no available train capacity (train is fully committed or there are no train wagons at the sidings). Therefore sidings cannot relieve the stack. In such cases bi-modal solutions can be introduced whereby a B-double trailer or higher capacity truck fleet relieves the stack of the overflow swiftly. This is the case when train services, and/or loading capacity on sidings are not available. In this case it may be possible to minimise the stack size as well as meeting the necessary train dispatch service. Bi-modal operation rules cater for stack overflow. The triggers are a saturated stack and an inactive sidings. From Daganzo (1990) it is anticipated that such an overflow mechanism will experience lulls interrupted by flurries. B-double and Super B-double trucks are sufficiently mobile to respond quickly to relieve stack overload constraints. This concept of running intermodal terminals has been anticipated by State Government to relieve the freight exposure impact on city roads (DoT, 2008a). The mechanisms have not yet been developed to analytically assess such operations. The ultimate terminal design schematic for Dandenong as a bi-modal terminal is depicted in Figure 48. It can be seen that the terminal footprint is 15.9 hectares. The triple rail track provides for a wagon set to be loading whilst another loaded set is dropped off, the locomotive detaches and then runs around the siding (track 3). The central stack is accessed separately for loaded imports, exports and empties. These facilities are a combination of road-rail, intermodal and road consolidation/distribution facilities. Their operation may be feasible with the use of High Productivity Vehicles (HPVs). When there are time-sensitive trades and container slots on trains are not available, HPVs can shuttle containers from the intermodal terminal stack to the port and back. Container loading and unloading from the train is done directly and indirectly (via the stacks). The dwell profile of empty containers at a terminal may be considerably reduced. A significant proportion of containers may also be repatriated to the seaport by rail. Page | 236 Chapter 6: Satellite Terminals to a Seaport Figure 48: Intermodal Terminal Design Schematic (Railhead on west side of terminal) The HPVs provide a means of reducing the transit of containers. For instance, HPVs may take two FEUs from the stack and deliver to the customer. From the customer they collect 2 empty FEUs and deliver to the intermodal stack for railing. There they collect 2 loaded FEUs and drop off two to the port, where they may collect two loaded or empty FEUs for customers and the cycle commences again. The use of HPVs at the intermodal terminal also regulates the intermodal terminal stack within a smaller boundary of operation. Were single articulated trucks carrying a single TEU or FEU to be used, a much larger transient stack would need to be maintained in order to feed the block flows of trains. There would also need to be considerable truck movements. The transient storage of empty containers at the intermodal terminal also provides the ability to cycle empty containers within areas without having to return empty containers to an empty Page | 237 Chapter 6: Satellite Terminals to a Seaport container park at the port precinct or outside the area. This coordination leads to a very high back- loading rate and reduces the dwell time of empty containers in the system. The filter logic for bi-modal overflow control using electrical circuit stencils is devised and presented in Figure 49: Bimodal Overflow Control Filter Logic. The source flow impedance and the stack characteristic impedance leads to an evolved storage lag denoting stack saturation. This is the first trigger point. Sidings inactivation designated by a bandreject filter of a certain synchronisation spectrum acts as the second trigger. Higher Productivity Vehicles are then called upon to relieve the stack swiftly. They act to inject handling resources which counteract the reactive impedance of the stack. The storage lag reduces. Distribution analysis of the synchronisation band of sidings inactivation according to signal inflow can identify the value intensity of carrier pairs that require an alternate itinerary via the bi-modal solution. This value intensity will direct the number of vehicles of a certain capacity that need to be activated. This will guide further development of bi- modal circulation rules. This control logic using analytical methods provides decision support on areas such as: acceptable stack size, bimodal fleet requirements, and train schedules which can be serviced. Page | 238 Chapter 6: Satellite Terminals to a Seaport Figure 49: Bimodal Overflow Control Filter Logic Page | 239 Chapter 6: Satellite Terminals to a Seaport 6.5 Enfield Hub for Complex Bundling 6.5.1 System and Terminal Description Enfield hub is a proposed road–rail collection and distribution hub located approximately 17km from Port Botany. It is planned to accommodate throughput of 300,000 containers per annum. Atypical of urban intermodal terminals, it has a large land area of some 50ha. It also has many sidings. The proposed container flow balances for Enfield intermodal terminal are depicted in Figure 50. It can be seen that the terminal has a considerable empty container handling function. It also has a significant container stuffing function where 1/3 of the export container volumes arrive as less than container load. In this way, Enfield will operate as a complex bundling terminal. This relieves the requirement for Less than Container Loads to deliver to a site at or adjacent to the seaport for warehousing and bundling. Figure 50: Proposed Flow Balances for Enfield Intermodal Terminal (SKM, 2005) Page | 240 Chapter 6: Satellite Terminals to a Seaport Key: Ce : Empty containers Cx: Export containers Cm: Import containers Px: Export container equivalent loads which are less than container load Pm: Import container equivalent loads which are less than container load 6.5.2 Terminal Transhipment Calculus In the Enfield model a Multiple Input Multiple Output terminal framework is presented. This lays the basis for the transhipment calculus using the electrical analogue to carry multi-commodity flow information. In this case the flows cover the movement of import and export containers, and LCL consignments which are consolidated/deconsolidated (stuffed and unstuffed) on site. The transhipment calculus may represent both work impedance and control impedance. The flow relationships depicted in Figure 50 can be represented by a terminal flowchart (Figure 51). This represents the Multiple Input Multiple Output nature of the terminal node and the multi-path coupling arrangements of itineraries. Functions include direct and indirect transhipment of export and import containers, empty container storage, and bundling/unbundling of less than container loads into either full containers for dispatch or de-consolidated consignments for truck haulage. The block algebra of these relations can be represented in a matrix format. Page | 241 Chapter 6: Satellite Terminals to a Seaport Export Direct Cx Cx Transhipment Indirect Transhipment Store Import Direct Cm Cm Transhipment Ce Empties Direct Transhipment to Seaport Empties Direct Transhipment to Ce Hinterland Empties Store for Use Stuff to Full Export Pm Px Containers Figure 51: Enfield Intermodal Terminal Process Flowchart 6.5.3 Extension: Complex Bundling Phenomena As was described in chapter 5, this thesis proposes that complex bundling needs to be incorporated into the multi-commodity flow problem in transportation science in order for strategic planning to address urban intermodal freight issues of consolidation and transhipment. The thesis considers the application of terminal multi-path coupling in formulating impedance requirements for flows of certain origin, which may be less than full container load, to be converted to dispatch payloads. The profile of origin and destination consignments to and from Enfield by principle market area is presented. Using synchronisation domain (frequency spectrum) analysis, this information can be built into the model formulation. It can then be analysed to see how changes in the source of flows affects the train dispatch payload. That is, an understanding of catchment-payload relations can be Page | 242 Chapter 6: Satellite Terminals to a Seaport gauged with commensurate terminal impedance requirements. This is critical for assessing intermodal terminal feasibility. The proposed hinterland distribution of these flows for Port Botany is depicted in Table 14. This represents 90% of consignment origins and destinations sourced within a radius of 15km53. Table 14: Proposed Principal market Areas of Enfield Intermodal (SPC, 2005) Local Government Area Exports Imports Auburn 9% 9% Bankstown 38% 40% Baulkham Hills 0% 0% Blacktown 1% 4% Fairfield 6% 10% Holroyd 4% 8% Liverpool 3% 6% Parramatta 30% 19% Ryde 0% 2% Strathfield 9% 2% Total 100% 100% In order to capture a significant number of containers on rail, a minimum train service frequency is required. In order to support these services with respect to volume and frequency, the catchment must be more extensive and diverse so that smaller generators and attractors of freight can be captured. Through the transhipment calculus of terminal impedance, catchment size can be 53 The radius includes consolidation activity conducted in the hinterland and made up into containers and therefore the real radius of consignments may be much larger. Page | 243 Chapter 6: Satellite Terminals to a Seaport associated with active arrival duration (storage areas) at the terminal and with load dispatch frequency. Inevitably, more complex bundling may involve a greater dwell time due to variations in diverse traffic classes of diffuse truck flows(in catchment space and time). A greater dwell time then leads to increases in active arrival duration , more storage reserved and a lower throughput/container storage area ratio. The mathematical formulation is as follows. The consignment is a mass which has a container volume equivalence. We define the notation which describes this relationship and provides a means to track multiple commodities through the system. We define the number of consignments from carrier j, along path i, within period t Csij, Each carrier is synonymous with a carrying class which defines the type of container. Each consignment has a commodity type coCs Each commodity type has a density ρ Co Each consignment has a mass according to each carrier type MCs, j The volume of each consignment from carrier j is then = ρ VCscoMi,, j i j... Cs Co Cs, j These indices allow us to assess mass throughput balances, volume storage balances, and carrier balances. This expression can also be mapped to a transform and expressed in frequency spectrum analysis. Page | 244 Chapter 6: Satellite Terminals to a Seaport The composition of the complex bundling problem is presented in Table 15. The complex bundling problem is a specific subset of the multi-commodity flow problem, involving multi-modal transport forms and the use of node infrastructure by the mechanism of multi-path coupling. It uses a form of the steady state bulk queue solution presented in Chapter 5. Table 15: Composition of Complex Bundling Problem in an Intermodal Terminal and Attributes Measured Multi-Commodity Multi-Modal Multi-path Coupling Signal of multi-attribute flow Carrier pair- harmonics Terminal work information Transhipment intensity, value Desired Load value Itinerary magnitude, synchronisation rating Value Intensity Active Arrival Period and Storage control synchronisation band Siding Activation Dispatch control Passed Value % Utilisation Dispatch The filter logic of the complex bundling formulation is depicted in Figure 52. Complex bundling can be described in multi-commodity flow terms as an itinerary through the terminal. Containers or goods arrive on a specific carrier, k. These containers are to be transhipped to train j. Each of these carrier flows has a timing sequence which represents the active arrival period of these flows before the dispatch cut-off of train j. A synchronisation band will be specified which represents the number of container slots in storage available for this train service. Carrier flows can occupy whole or part of this timing band. This will reflect the efficiency of the storage management and to what extent storages are underutilised. Where there are peak inflows or Page | 245 Chapter 6: Satellite Terminals to a Seaport departure schedules, some excess storage area may be necessary. Each source carrier will have a synchronisation rating, part of which may overlap with another source carrier. The storage slot requirements for each train service can be added for a total storage requirement. A check for excess storage capacity can be made with calculation of value intensity of the storage band (value/frequency range). Outside this timing band, the source flow will not be able to make this train service. For each load service there will be a quantum amount that is stored - this can be expressed as both number of container slots and total value stored. The container or value flux can be multiplied by the storage lag (average stored time) to obtain the amount stored. For precision, the nominated storage lag in the source must be met in the storage resources available. The sum of storage slots required by load services of each type will represent the total storage. Multiple signals of source flows can be accepted for this load service so long as they fall within the timing band nominated for this scheduled service. Certain carrier sources will have a higher flux than others and carrier sources can overlap in timing bands – this will be reflected by a higher value density (value/ timing unit).The concept of resonance comes into play with filter devices. The source and load are at resonance when the maximum efficiency is achieved for the Level of Service required. Each train service, j, will have associated with it an arrival interval of carrier flows that are to be paired to it. This arrival interval can be represented by the bandwidth and is the space reservation duration required for each train service. The magnitude is the value flux (power) of each carrier flow. The duration multiplied by the value flux yields the space reservation. A number of carrier flows can meet one train service. They must arrive within the synchronisation band (bandwidth). Any flows outside this bandwidth are allocated to another train service. Page | 246 Chapter 6: Satellite Terminals to a Seaport Figure 52: Complex Bundling Filter Logic Page | 247 Chapter 6: Satellite Terminals to a Seaport 6.5.4 Bundling Figures of Merit The Figures of Merit associated with complex bundling at a terminal are associated with obtaining sufficient diverse sources of flows to support dispatch train services. Consequently, the catchment of source flows can be associated with the storage required and the dispatch service frequency. This is a significant Figure of Merit in establishing the location and functional attributes of an intermodal terminal. The distribution of source flows on the throughput of the terminal can be transparent. As the level of complex bundling a terminal can accept is also a measure of flexibility, another significant Figure of Merit on terminal performance is the precision – flexibility relationship i.e. what is the acceptable length of the period of arrival distributions in order to achieve consignment bundling sufficient for a specific train dispatch class payload. 6.6 Chapter Conclusions This chapter has presented the specific context for the implementation of intermodal terminals linked to seaports in Australian cities. Establishing mechanisms of functional relations and their distribution among the shipper hinterland, the intermodal terminal and the seaport are critical in assessing the feasibility of intermodal terminals. Distributing critical functions to intermodal terminals requires an assessment of their capability to incorporate these functions according to precision relations and the consequent value proposition. The chapter discussed design problems specific to attributing functions to landside intermodal terminals and then providing tools for designing their configuration. Qualitative case studies were described for real intermodal planning situations in Sydney and Melbourne. These case studies supported the need to investigate mechanisms of stack management, sidings activation, bimodal overflow and complex bundling. The transhipment calculus scopes and offers opportunities to generate the Figures of Merit necessary to understand the functions required at the terminals. Each of these mechanisms needs to be further investigated for enumeration of parameters to test the range of usefulness of the Figures of Merit. Page | 248 Chapter 7: Waste Transport and Resource Recovery Case Study 7 Waste Transport and Resource Recovery 7.1 Introduction 7.1.1 Chapter Objectives This chapter considers how the analytical framework of transhipment calculus may be applied to aspects of intermodal services in the waste transportation industry. The case study deals with an existing waste intermodal terminal at Clyde in Sydney and data was collected there as well as at another waste terminal at Matraville. Based on these observations a systems dynamics model was established under a team the candidate supervised to understand the relationships amongst the entities and plant parameters and to estimate impacts with changes in waste fluxes. The system dynamics (semi-discrete) model formed the basis for a gap analysis in current modelling of terminals for the requirements of the new intermodal modelling framework. It was identified that a representation of process dynamics is required that captures the essence of unit process capabilities, namely how throughput capacity is affected by flux variables,. It would be especially helpful if this could be developed as an analytical method for sketch planning purposes. The gaps highlighted the need to assess retrofit opportunities for waste intermodal terminals subject to alternate rail operating forms. The system dynamics model also was used to define some boundaries of terminal capacity and the critical path of operations. This set the scene to apply the transhipment calculus to sketch plan notable phenomena in waste intermodal production systems. The chapter illustrates the following mechanisms pertinent to the waste intermodal terminal: 1. Active Arrival Area and Stack Management, 2. Dumping- Compaction Activity, 3. Itinerary control, These mechanisms further illustrate attributes of terminal impedance that is described by the electrical circuit analogue. A protocol of 5 system formats are described with novel train operating forms that anticipates various degrees of terminal operating complexity. These system formats are seen as necessary if higher quantities of waste to disposal are to be diverted for beneficial reuse with the aid of an Page | 249 Chapter 7: Waste Transport and Resource Recovery Case Study intermodal terminal gateway. These system formats and likely Figures of Merit are described according to the new model formulation which is anticipated for future work. From this case study the planner can gain some understanding of the precision requirements and value proposition with different operating train forms. Some of these system formats require a major alteration in function and configuration. The system dynamics model demonstrated severe constraints on the existing facility to cope with a doubling of inflows. These results generated investigations into novel terminal business rules, new configurations and technologies which could manage higher throughput and alternative itinerary (and transformation) requirements. Critical phenomena of dumping floor inertia, sidings capability, feedback loops between LUPs in the terminal and system distribution of compaction activity can be represented by stencil dynamics. The waste case study describes the mathematics and block algebra to cater for feedback loops between LUPs. Such closed loops are a critical dynamic in consolidation freight terminals relying on flows of empty containers for production of new containers. Sidings capacity flexibility is also considered as this is a critical part of terminal throughput capacity, particularly where there is relatively short line-haul (i.e. less than 25km). The case study demonstrates the usefulness of signal processing using the circuit stencil. Increasing the train throughput to two or more return trips per day has fundamental implications to the operation and desirable configuration of the waste terminal. Through the transhipment calculus, alternate system formats of waste transportation systems can be sketch-planned for their accessibility and financial feasibility with the prevailing infrastructure. The planner can identify the precision requirements and value proposition with different train operating forms. Furthermore, these system formats can be tested to assess how amenable they are to the application of novel business rules and thus the planner can gauge their degree of flexible capacity. This planning tool can also anticipate what retrofits are required to achieve the desired train dispatch payloads. The information on current operations was collected with the generous assistance of Veolia Transport. The case studies are however purely hypothetical. Research students assisted the candidate with building and generating the system dynamics model. This contributed towards their Honours thesis (Michetti and Porta, 2007). This case study was scoped in a paper presented by Beavis et al, (2006a). Page | 250 Chapter 7: Waste Transport and Resource Recovery Case Study 7.1.2 Delimiting the Study The focus of this study is on the transportation of fractions of Municipal Solid Wastes (MSW) and Commercial and Industrial (C&I) waste from the point of delivery to a consolidation- transhipment (transfer) facility to dispatch and receipt at a waste processing site. The pre-consolidation network of collection is not assessed explicitly. Reverse logistics and industrial symbiosis, whereby certain materials within the waste fraction are tracked to their reuse to the same or complementary industry, is not considered. The study does not consider explicitly resolving the environmental–cost dilemma of source separation versus bulk transport and separation and processing at a waste recovery plant. The model however has the potential to enumerate the operational impedance to achieve consolidation, storage and transhipment tasks according to the composition of the demand-supply waste stream. This could contribute to the on-going policy and commercial discussion in supporting evolving resource recovery markets of current waste to disposal fractions. 7.2 Concept of a Sustainable Waste Logistics System 7.2.1 Transportation Leverage for Waste Recovery In this study we explore the tradeoff between the sustainable issue elucidated by the proximity principle and the logistics supply chain theory of the postponement principle. The proximity principle requires waste to be disposed as close as possible to its point of generation, in order to reduce transportation effort in re-processing and disposal. The postponement theory, used in logistic systems and distribution channels, outlines the motivation in taking decisions regarding in what manner the good should be produced and distributed. The principle of postponement, developed in the logistics field by Alderson and Bucklin in the fifties, considers, to what degree, activities in a value supply chain should be postponed in order to increase efficiency from reducing inventories (Jahre, 1995). For instance, fourth party supply logistics considers how to coordinate production of component parts of a good in dispersed Page | 251 Chapter 7: Waste Transport and Resource Recovery Case Study localities. Often this process coincides with Just in Time production, or time compression principles. This may have a trade-off in reducing the economies of scale of operations as smaller batches are produced. The final assembling plant would then postpone the manufacture of the good through the supply chain until it is actually ordered. The longer the chain the more involved the synchronisation requirements. Cheap and reliable transportation has facilitated these operating principles, often with a large increase in the road transportation effort and lower unit loading. Speculation considers how detailed activities should be along this supply and distribution chain and speculative inventories of finished or semi-finished goods could be located along this chain in warehouses to avoid stock-outs. In waste management theory, Jahre (1995) has proposed the concept of speculation- postponement to designate the degree of source separation that might occur in an MSW collection system or be postponed until sorted at a Material Recovery Facility (MRF). Cost-benefits of any strategy will depend on the number of waste fractions to be separated and the population densities of collection. This finding coincides with the observation of Daganzo that transhipment terminals are economically feasible in a many-to-many hub form, particularly when the destination densities are sparser that the collection densities (Daganzo, 1995). In the present project the postponement principle will be considered as the compaction activity in the waste value chain54. Here alternate locations are evaluated where compaction might be performed under conditions of combined transportation. Under conventional arrangements, MSW goes through three compaction stages. Waste faces firstly a low density compaction on trucks once collected from households, then it is uncompacted in transfer stations and compacted again (with an higher density) in trailers attached to B-double trucks and dispatched to the landfill (Bates, 2005). There, it is once again dumped (uncompacted) and disposed under final compaction in landfills. With the application of unitised consignments, such as 20ft or 40ft containers, there are new opportunities in managing waste and increasing load factors. Speculative inventories become significant in achieving flow densities to support rail operations. A better coordination of waste transport would therefore include investigation into the nature and the location of compaction itself. This has implications for the network and the role of transfer 54 Postponement of consignment dispatch generally is seen as a benefit of intermodal terminals to achieve greater consignment dispatch densities (such as pulse train loads). Postponement is the practice of bulk queue mechanisms. Page | 252 Chapter 7: Waste Transport and Resource Recovery Case Study stations in a combined transportation physical and service network. Capturing sufficient mass of waste, in a timely fashion, is crucial in order to build block trains. Postponing compaction means to collect waste on the road network and to concentrate it at the unique road-rail facility where it will be packed in at the very end and sent by rail. Under advanced network operating forms, speculative inventories might be facilitated with the development of satellite transfer stations (in addition to the only existing road-rail station in Clyde), where compaction takes place in standardised containers. A further anticipation in the application of the principle would consider compaction directly on trucks which collect the residual fraction of MSW. These trucks could go directly to the intermodal terminal or go via a local satellite terminal where the containers are transshipped to a B-double and dispatched to the intermodal terminal. In this way, collection is separated from the line-haul task with potential savings in truck numbers and trips. The proximity principle was established in the EC Framework Directive on waste (75/442/EEC, amended by Council Directive 91/156/EEC) (EEA, 2006) and declares that waste should generally be managed as near as possible to its place of production because transporting waste itself has environmental impact and costs. Investigations since the 1970’s of possible rail haul of containerised wastes have shown its technical and economic feasibility (US EPA, 1973)., Railing wastes is particularly feasible over long line haul. This assessment was based on single services with conventional handling technologies. The motivation of the investigation was to access landfill and re-processing sites which were far from waste generators. This drivers has also been the basis for more recent strategies in the intermodal transport of waste proposed in the STRAW (Sustainable Transport of Resource and Waste) project (Curry, 2006). This project recognises that combined transport networks should be developed to 1) alleviate the demands on the road network by moving wastes, and 2) to facilitate the development of waste cascades in resource recovery. It may be observed that Great Britain, as elsewhere has considerable missing links in its transportation network (Maggi, 1994) which means that access to potential re-processing sites is not evident. The project proposes a mass balance multi-modal approach to developing waste cascades where wastes can be separated, handled, stored and transshipped for effective recovery at low environmental impacts. In Sydney and the state of New South Wales context it has been identified that there is a distribution mismatch between generators of green waste and demand for compost with rich nutrients, including carbon for farmland (WMAA, 2005). Pricing mechanisms need to be built to Page | 253 Chapter 7: Waste Transport and Resource Recovery Case Study make the transfer attractive. An essential means to the industry Benefits of Trade (IBOT) is to develop the transportation infrastructure and logistics networks55. Consequently, smart allocation of compaction and a clever setting on intermodal hubs demonstrates how a sequence of intermodal transhipment of waste could be sustainable even if the processing point is at a distance from generation. The appraisal of the best compaction allocation can be examined through the investigation of different scenarios which contemplate network optimisation and in some cases new infrastructure to be constructed or resources to be reallocated. Our investigations involve different compaction activities prior to transhipment and how these activities are compatible with different rail operating forms, designed to reach different recovery facilities of different destinations. We may discern how these systems add to the flexible capacity of a network with different service designs including compaction arrangements, transient storages and rail operating forms. 7.3 System Format in Sydney 7.3.1 Waste fractions of interest In this case study the waste fractions of interest in MSW include the residual fraction, green waste as well as C&I waste. Currently the rail terminal at Clyde deals solely with C&I wastes. In the state of New South Wales, approximately 1.8 tonnes/waste/person is generated each year over the waste streams of MSW, C&I and C&D. The State represents a third of the population of Australia and generates slightly more than a third of its wastes. Significantly the Sydney Metropolitan Region disposes of 68% of wastes. If the Extended Regulated Area is considered, some 84% of wastes to disposal are generated in urban and peri- urban areas. 55 Extensions to network coordination methods and Industry Benefits of Trade mechanisms are scoped as future work in Appendix F: Theory Synthesis and Extension to Network Coordination Page | 254 Chapter 7: Waste Transport and Resource Recovery Case Study 7.3.2 The Role of Clyde Clyde is the only transfer station in Australia where waste is compacted in containers then sent by rail to landfill-bioreactor/ resource recovery complex. The receiving railhead is located at Crisps Creek Intermodal Facility where it is then transshipped and trucked another 10km to Woodlawn Figure 53: Train Path from Sydney to Woodlawn Bioreactor. Currently the station at Clyde receives approximately 400,000 tonnes of waste to dispose per year. This is mainly C&I waste. The Woodlawn landfill is some 250km Southwest of Sydney and has an available volume of 25 million m3. Assuming a constant inflow of up to 500,000 tonnes of waste/year the bioreactor is estimated to have sufficient capacity for the next 40 years. Figure 53: Train Path from Sydney to Woodlawn Bioreactor As well as producing EfW (Energy from Waste) biogas, the site also has composting facilities. There are also plans to build a mechanical-biological plant at Woodlawn for the recovery of another 120,000t each of MSW and C&I waste. Conventional landfills around Sydney are reaching capacity and there is also a strong move to minimise the amount of waste to disposal. It is estimated that there is a potential for Clyde to take a Page | 255 Chapter 7: Waste Transport and Resource Recovery Case Study significant amount of waste to disposal that can no longer be handled by the landfills around Sydney. The estimated amounts that Clyde could theoretically attract are depicted in Table 16 based on waste generation growth in its vicinity. This does not consider the constraints on Clyde to handle such threefold growth in flows. Table 16: Projections for Waste for Disposal Quantities Intermediated Through Clyde in 2015 (Uncapacitated)56 2015 Clyde with Truck Arrivals/ Hour Diversion Range Monday T-F Saturday Sunday Low 8.63 7.23 4.56 1.43 Normal 26.46 29.04 16.78 7.61 Peak 55.18 48.87 39 13.27 Mean 28.60 26.56 17.93 7.26 Tot arrivals/day 686 637 430 174 Mass/truck [t] 7.16 6.58 4.89 6.61 Total mass/day 4915.6 4197.4 2105.3 1153 Total mass/week[t] 24,963 Total mass/year [t] 1,248,185 56 Uncapacitated means that these flow projects have not considered the capacity limits at Clyde with the current terminal/sidings configuration. Page | 256 Chapter 7: Waste Transport and Resource Recovery Case Study 7.4 Transfer Station Formulation from System Dynamics 7.4.1 Time Series Analysis of Source Flows A time series was analysed from weekly flow data and a distribution was generated and used for the System Dynamics simulation (Figure 54). Figure 54: Distribution Analysis of Truck Arrivals at Clyde (Porta and Marchetti, 2007) 7.4.2 Terminal Description Clyde terminal is a single ended terminal with two active sidings adjacent to hardstand and 3 storage sidings to the north of the site (Figure 55). Trucks access the terminal from the North off Paramatta Road and enter the terminal through the gateway that bisects the active sidings to the North. Trucks queue to enter the dumping floor to the West. A turning circle is large enough for trucks to return from the same direction via a dual carriageway. Compaction of forty-foot containers (FEUs) occurs from either of two compactors. A forklift places full containers to the stack and collects empty containers for placement. The same forklift also removes empty containers on arriving trains and places full containers on empty train wagons. Train sets shunt, arrive and depart on the East side. From Figure 55 it can be surmised that there is limited stack storage for empty and Page | 257 Chapter 7: Waste Transport and Resource Recovery Case Study full containers. Operations accommodate this by pre-dispatch loading of containers on wagons which are then shunted to inactive sidings. Figure 55: Aerial View of Clyde Terminal (Collex, 2006) In this case study, interactions within intermodal transfer stations were analysed using the system dynamics software Powersim™. The model formulation is known as Continuous Variable Dynamic Systems (CVDS) which is based on continuous flows, however allows for triggering events to enact other flows and activities (Cassandras and Lafortune, 1999). Thus the model formulation can be considered a partially event based approach. System dynamics was selected over queuing software as system dynamics with explicit triggering allows incorporation of resource constraints and business rules as feedback loops. Models can be set-up and scenarios generated on limited data. This is valuable when specific fluxes over unit processes within the terminal are not easily measureable. Conventionally, waste transfer station design has been based on queuing theory at each unit process node to identify capacity constraints which are presumed to focus on the interaction between vehicle arrivals and process servicing, creating transient queues which may accumulate as storages. However in terminals with a transformation and storage function, space management also needs to Page | 258 Chapter 7: Waste Transport and Resource Recovery Case Study be related to quality factors such as synchronisation and other markers of consignment pairing between arrival and dispatch. Additionally, as noted in Chapter 3, the model abstraction of the bulk queue does not cater adequately for the variety of diffuse – discrete pulse flow sequences that might need to be tested for a road- rail intermodal operation. This is particularly the case with rail-road intermodal terminals where there is a potential mismatch between container production and train production. These cycles need to be synchronised. Intermodal terminals have at least two production cycles. One is the container production cycle, involving the compaction of waste, and the second is the train production cycle involving train and wagon shunting, formation of wagon blocks, line haul to destination, transfer, loading of empty containers and return. It is possible to identify four main areas concerning capacity issues: 1) Compaction service, 2) Transhipment sidings, 3) Train Service, 4) Storage. The compaction constraints include the number of trucks that may enter at any one time, and the time taken for compaction. The capacity limitations of the sidings/transhipment track sub-system are based on site layout and length of transhipment tracks. Train service represent a constraint in two ways: the train length: a maximum 54 wagons (800 m length) and frequency which is a function of distance and availability of the train path. Container storage represents a constraint in relation to two parameters: the number of containers (buffer of 165 containers) and the available area of placement. These parameters could represent a constraint if full container production increases which is not in synchronisation with the train service. If this happens, more buffer containers will be needed as well as a larger storage area. These constraints are dynamically related. The container storage represents a significant buffer between the container production cycle and the train production cycle. If the frequency of train service were to increase, the storage requirements may be less. An evolving constraint in this case would be the rate of handling. In simulating the operations of intermodal terminal, we obtain a time composition for freight production which may be termed the characteristic impedance. This characteristic impedance may be extended to describe the interval time to produce freight services under different intermodal production systems arrangements. Page | 259 Chapter 7: Waste Transport and Resource Recovery Case Study Figure 56: Process Flowchart of Waste Intermodal Terminal at Clyde 7.4.3 Feedback loop of resources matching In the system dynamics model developed for this terminal, flow resource attributes are tracked by applying dynamical operations to each flow attribute and connecting each dependant dynamical unit process. A number of feedback loops were created to simulate the interaction between the transportation cycle and the production cycle (Figure 57). There are four areas of interface affecting capability: 1) Truck arrivals and compaction service; 2) Storage; 3) Transhipment sidings (Train Service- arrivals, dispatch); and 4) Train dispatch services and mainline operations. Page | 260 Chapter 7: Waste Transport and Resource Recovery Case Study Figure 57: Modelling Map of Logistical Feedback Loops at Intermodal Waste Terminal (Porta and Marchetti, 2007) Page | 261 Chapter 7: Waste Transport and Resource Recovery Case Study Figure 57 depicts the critical role of information flows to operate a system of feedback loops. Solid lines represent physical flows, dotted lines are information flows such as relating one entity with another, and dashed lines are information triggers on resource use, and limits. The movement and availability of entities: trains, wagons, containers and their condition: empty, full, to be freighted, trucks, and incoming wastes are linked and their status triggers other activity. The utilisation of plant (compactors, handling equipment, the dumping floor and the stack) is triggered by these fluxes. Train arrivals contain empty containers. These are removed to the stack and are used to compact waste incoming. Containers are filled and then stored or directly transshipped to wagons on sidings. Trains are dispatched after wagons have been filled. In the sparse representation of terminal activity using transhipment calculus, Logistical Unit Processes (LUPs) need to contain feedback loops within themselves and between other dependent LUPs. 7.4.4 Storage-Handling Limits The simulation results presented identified that, with continued indirect transhipment operations, there were two significant bottlenecks: 1. Compaction operations: truck access to the dumping floor became increasingly constrained in the higher peaking of arrivals; 2. Stack capacity limitations in the storage of newly produced full containers and returned empty containers. Sidings limitations were not investigated explicitly in the simulation. It is perceived that the current sidings configuration and use is limited to two return trips per day. 7.4.5 Throughput Capacity Results The model was tested under a number of flows culminating in the flow level specified in Table 16. Page | 262 Chapter 7: Waste Transport and Resource Recovery Case Study Severe constraints arose in the dumping-compaction unit process. It was also uncertain how the terminal would deal with the interaction between storage stack and sidings to meet more frequent train dispatch requirements. The current infrastructure artefact with which the logistical structure must operate only provides for classification and shunting operations using the existing sidings configuration. That is, there is limited direct frontage of sidings for handling equipment to access. Alternate handling of alternate waste streams was not considered in the systems dynamics model. 7.4.6 Modelling Gaps The modelling approach using CVSD incorporated many attributes that supported the fidelity of the modelling endeavour. Variability in the flow profile can be considered by generating an inflow of a certain distribution or analysis of data to approximate a statistical distribution. Feedback resource triggers are used liberally to act as bulk queue control to the process flow. A good understanding of the storage-handling adequacies and deficiencies can be gained around each unit process. Whilst continuous, a time step approach allows for accumulations in storage buffers. Parameter limits can force a backlog of flow upstream. Advice on terminal design has recommended design be based on back-end approaches57 (Everet and Appelgate, 1994, Thompson, 2006), and the system dynamics model is not readily amenable to this. Intricate feedback loops can be devised to accommodate this to a certain extent. Analytical relationships cannot be easily derived from such construction and general rules cannot be surmised. As such, it is not a true load following approach whereby up-stream unit processes are explicitly designed for desirable output service profiles. This is a critical gap where the key driving force of urban intermodal model formulation should be adequate terminal performance due to alternate rail dispatch profiles. Secondly, feedback is exogenously connected. Besides limit constraints, there is little consideration of the endogenous response of each unit process to changes in flow profile and information. Therefore there is no consideration of compensation control such as bringing on more resources, storage capacity and changing operational rules to assess the flexible capacity of the unit processes. 57 Back end design approaches require the planner to consider the nature of the consolidation dispatch in the complex topology design of terminals – the itinerary based on linked unit processes and sizing and function of processes in order to achieve the desirable dispatch. When dealing with management of two way flows it may become a more successful approach to consider discrete design based on the dominating pulse (of trains and vessels), rather than use of system dynamics. Page | 263 Chapter 7: Waste Transport and Resource Recovery Case Study Thirdly, the System Dynamics approach is limited in the information each flow can carry. It is not a signal which can contain multiple attributes. This limits the use of system dynamics approaches in addressing the multi-modal, multi-commodity flow problem in intermodal tactical planning of complex bundling. For instance, the modeller cannot observe the impedance mapping effect of one carrier inflow onto a specific carrier outflow service. Fourthly, there is no discrimination between time and timing or synchronisation. System dynamics can give the planner an understanding of delays but not how flow constituents can specifically map to the schedule of a pulse output service. Associated with this is the lack of concurrent systems information. Relationships between terminals are not specifically reflected in these dynamics and thus there is no understanding of real time consistency in steady-state relationships. Therefore implementation of the full concept of precision is lacking. The results do not give the planner an understanding of how this terminal would act under a range of system conditions or how it could provide further system capacity. Finally, the system dynamics model considers capacity as a constraint phenomenon. There is no obvious mechanism to route flows through the terminal on the basis of commodity, timing or other information. In reciprocal fashion, there is no means to register changes in system slot capacity with node activity. Discrete event models can address a number of these issues, however, they are data and parameter intensive. Consequently the system dynamics model does not offer tractable solutions to the decision support needs at the level of sketch planning terminals of distributed function that have precision and controllability requirements. Page | 264 Chapter 7: Waste Transport and Resource Recovery Case Study 7.5 Formulating Terminal Complex Activity This section explains and mathematically tests the key operating phenomena specific to waste intermodal transfer terminals, which was outlined generally in Chapters 5 and 6. 7.5.1 Complex Impedance Representation of Terminal Activity The activity cycles of the terminal can be reduced to a critical path sequence (Morlok, 1978a). Additionally, these critical path itineraries of consignment flows can alter according to their transformation, synchronisation and destination needs. Thus incorporating flow multi-attribute information is critical in terminal itinerary decision making. For the transhipment calculus approach the terminal can be distilled to three unit processes involving indirect transhipment: 1. Dumping floor – compaction, 2. Stack-handling, 3. Sidings with pre-dispatch overflow. The operation of Logistical Unit Processes (LUPs) depends on the cycling of flows of the material and carriers. The availability of these entities trigger terminal production phases and transportation production phases. The relationship between parameters and performance specifications for each LUP at the waste terminal can be ascertained through the working tool of the Bode plot. The broad operating parameters of the terminal under existing conditions of a throughput of 400,000 t of C&I waste per year and a conventional operating form of single evening train dispatch are presented in Table 17. Page | 265 Chapter 7: Waste Transport and Resource Recovery Case Study Table 17: Parameters for Base Case Used in System Dynamics Simulation Parameter Value Unit Dumping Floor Capacity 800 tonnes Dumping Floor off-load capacity 3 Trucks Arriving trucks average mass 7.2 tonnes Compaction batch in FEUs 27.2 tonnes Number of bulldozers 1 Number of compactors 1 Number of forklifts 1 Handling rate 3 Mins/FEU Number of containers per wagon 1 Train production cycle 1 Train/day each way Container storage capacity onto hardstand 200 Containers (3 high) The operating times of carrier resources and mobile handling equipment is depicted in Table 18. Page | 266 Chapter 7: Waste Transport and Resource Recovery Case Study Table 18: Operating Times of Mobile Resources (Porta and Marchetti, 2007) Operative Times Mean Variance Lower Limit Upper Limit [min] Truck Arrivals 6 3 5 15 Bulldozer Mixing 3 1 2 10 Compaction 7 2 6 15 Container handling by 5 3 4 10 forklift Manoeuvring of wagons in sidings has not been considered in the system dynamics simulation. This was not considered on the critical itinerary path when a conventional train dispatch schedule of one return trip per day was considered. A greater inflow, supported by advanced rail operating forms with more frequency of service, will require consideration of the sidings interface with the terminal and with the mainline. The mechanism is addressed in section 7.5.4 Sidings Capability. 7.5.2 Stack Interface Taking the stack-handling interface, uncompensated transfer functions are presented and analysed using Bode plots for alternate Q-factors. These are under the conditions of a) all container production is stored in the stack and only called for when train wagon sets are available prior to dispatch for a single destination; a1) As per case A but stack is periodically relieved by transhipment onto wagons; b) As per case a. and there are headway dispatches of 2 destinations; b1) as per b) and there is stack management whereby containers for alternate destinations and schedules are loaded onto wagon sets pre-dispatch and shunted (classified). These scenarios and their implications are depicted in Table 19. . Page | 267 Chapter 7: Waste Transport and Resource Recovery Case Study Table 19: Scenarios for Stack Management Narrative: Description of Stack Condition and Scenario Handling Implications Specification Bandwidth receivals at Q-factor stack Gain All indirect transhipment; positioned in stack; loaded onto train No sidings Constraints; without wagon No internal stack classification; long shuffles; broad wide -but needs headways of single bandwidth of arrivals to be a constant, A destination permitted med even distribution High Include sidings work- allows wagon set rotation of active face Max Stack gain so that both empty required to load train Containers are loaded return containers can Bandwidth can can be lower as can onto wagon sets be off-loaded and be narrowed be loaded onto adjacent sidings and newly compacted without wagons in then positioned onto containers can be operational incremental blocks A1 inactive sidings loaded higher than A impedance (2 or more) As per A but there are Bandwidth for at least Gain needs to be two destination one carrier pair reduced higher than A as headway dispatches per - may need to segregate Bandwidth dispatch from stack B period (2 trains) or shuffle the stack low narrowed in half the time As per B and Bandwidth can Containers are (within be narrowed Max stack gain schedule) loaded onto without required can be wagon sets for pre- Sidings act as a bypass operational lowered -similar to B1 dispatch overflow mechanism low-med impedance A1 Page | 268 Chapter 7: Waste Transport and Resource Recovery Case Study This analysis considers the impact on the stack58. The stack is the critical buffer node that needs to be maintained for the terminal to remain in operation. C2 R2 1uF 20k R1 C1 U1 1 + 0.185k 1uF 3 OUT Out 2 - V1 0.54Vac R3 IdealOpAmp 0.168k 0 0 0 Figure 58: Stencil for Waste Container Stack Management Parameter set-up and results are depicted in Table 20. 58 The stack can have several sections pertaining to long and short term storage. In these illustrations the stack generally has a within day storage duration and is mostly accessibly with the active hardstand adjacent to the active sidings face. Page | 269 Chapter 7: Waste Transport and Resource Recovery Case Study Table 20: Parameters and Results for Waste Stack Management Scenarios59 Parameter A A1 B B1 Q-dissipative work 10 30 10 5.3 K-accumulation required in the stack60 54 26 108 54 C 1 uF 1 uF 1uF 1uF Bandwidth 16.2 11.2 48.7 71 R1 0.185 1.154 0.093 0.098 R2 0.068 0.017 0.109 2.431 R3 20 60 20 10.6 The transfer function derived from the stencil of Figure 58 has the following generic form: Ksβ Hs()= ss2 ++βω2 0 (6.1) The value losses are calculated on the basis of a difference in ideal gain and actual gain for the train service described in each scenario narrative during the headway period. This is the value density drop through the LUP. Generally handling costs increase when there is more shuffling of containers at the stack to meet a specific service i.e. there are alternate destinations for the waste inflows. 59 Results generated for value loss (costs) for loading each train set varied and suggest further work needs to be undertaken to correlate value drops in the model with terminal operating costs. 60 This is the gain multiple required for a certain train load and generally before a train can be loaded. A sidings activation strategy allows the accumulation in the stack to be less than the full train load required as the sidings acts as an overflow storage. Page | 270 Chapter 7: Waste Transport and Resource Recovery Case Study The Bode plot results for these scenarios are depicted in Figure 59. 40.0 B 30.0 A 20.0 B1 10.0 A A1 0.0 B -10.0 10.0 12.6 15.8 19.8 24.9 31.3 39.3 49.3 62.0 77.8 97.7 B1 122.8 154.2 193.7 243.3 305.6 383.8 482.1 605.6 760.7 955.4 -20.0 A1 -30.0 -40.0 Figure 59: Bode Plots for Waste Stack Management Scenarios61 Condition A1 allows better stack control with a reduction in peak amount that must be stored. It also anticipates a tense flux condition with a narrower active arrival area. This means that the maximum headway is reduced, more regular services can be supported and holding costs reduce though handling costs increase. It also means that consignments have a narrower pass-band in which to reach a specific service headway dispatch. Condition B is without a sidings overflow strategy and it is seen that there must be a larger gain and broader active arrival duration to accommodate two dispatch services per day. If we assume that shippers are indifferent as to which service their consignments dispatch, the tense flux condition is dissipated as denoted by wide active arrival duration. In condition B1, the 2 services per day are addressed by a sidings strategy as per A1. The stack accumulation can be reduced. The tense flux condition parameter has not been altered. This leads to 61 Bode plot results should be scanned for magnitude and bandwidth response only for relevance to the freight terminal logistics problem. Page | 271 Chapter 7: Waste Transport and Resource Recovery Case Study a wide active arrival duration at the stack. There are additional handling costs associated with moves between stack and siding. The quality factor has reduced due to the wide active arrival duration permitted and a standard transhipment intensity. If this condition were tightened handling costs would increase with an increase in value loss. These results also map to a cumulative flow diagram where the situation with a wider bandwidth entails a longer active arrival duration and means that the stack is reserved for that particular carrier pair shipment. Strategy A1 has the lowest stack area demand, whilst B has the greatest stack area demand. The wider arrival duration of Conditions B and B1 is reflective of a broader inflow profile of loaded trucks. These are a more disparate source of consignments to feed the necessary services. The Bode plot of Value Density gain (in decibels) to synchronisation rating can be used to tighten the active arrival duration measure. Where consignment arrival profiles can be more synchronized the active arrival duration of the stack can be reduced further. Care must be taken to ensure other interface elements, such as the truck gateway maintain minimum levels of service of access. Where the technology and configuration is not appropriate to the throughput profile required, and/or the containers have an additional dwell time requirement, there will be an evolution of storage lag. This will be evidenced by a lower value density accumulation and a phase angle62. This example has demonstrated the use of the circuit analogue to investigate relationships between tense flux requirements, stack management and temporal accessibility, dispatch service levels and costs. 7.5.3 Dumping Floor Inertia and Evolution of Work Combined waste transfer and transhipment terminals act as consolidation nodes, and the waste batching process is a central activity in the densification of waste and reduction in freight intensity. Trucks deliver collected MSW throughout the day (averaging 7.2tonnes) and this waste is supposed to be dumped into a pit where it is further compacted and sent to landfill in larger B-double trucks (averaging 25 tonnes). This represents the non-putrescible fraction of MSW. In reality most large transfer stations do not have or do not utilise the facility for trucks to dump directly into a compaction pit. Terminal operators sweat the storage capacity and the handling equipment by allowing build up of waste consignments on the dumping floor. This storage allows the operator to 62 According to a phasor diagram this will resolve as an additional imaginery Value Density value- representing storage accumulation. Page | 272 Chapter 7: Waste Transport and Resource Recovery Case Study buffer the mismatch between diffuse local arrival peaks and more discrete dispatch services to landfills63. They can also conserve bulldozer and compactor equipment and staffing. This system can be proxied by a sequential queue where the tipping face of 2 berths (2 servers) is followed by one compaction server. The system is increasingly impeded as bulldozers have to mix and position waste for compaction over a larger and deeper waste zone. The growth of the waste zone can reduce the access slots for trucks to dump and clear in a timely fashion64. The terminals visited benefit from a night leap in production. These terminals receive a steady flow of waste in from 5am with a lull after 4pm. The intermodal terminal at Clyde has the whole day to produce unitised waste for 54 containers prior to train dispatch in the evening (typically around 11pm). Under these conditions, the delay at the dumping floor is manageable. If there is to be more frequent rail services, with a commensurate increase in truck arrivals, this dumping floor approach may become infeasible as a buffer and a production line system will be required. Figure 60 depicts the truck- waste interface at the Clyde facility dumping floor. Figure 60: Dumping Floor Movements at Waste Transhipment Facility and growth of the waste zone. 63 Landfills in Melbourne and Sydney are located on the outskirts of the cities. Congestion is such that deliveries to landfill are scheduled outside peak traffic periods. Even so, the increasing congestion in Sydney means that 4 B-double trucks instead of three are required at the Matraville terminal in East Sydney as they can only do 2 return trips per day instead of 3 return trips per day they could achieve 5 years ago. 64 Despite several visits to both Clyde and Matraville facilities, the candidate was unable to correlate increased dumping and mixing time with the growth of the waste zone. This was also a product of differing truck types in the fleet and the quantities of wastes they carried. What was evident was that with only one marked peak per day in inflows and the practice of a night leap in production, there was a lot of slack capacity in the terminal plant that meant that the dumping floor could be reduced in the early afternoon. Page | 273 Chapter 7: Waste Transport and Resource Recovery Case Study Storage and Handling capability can be assessed through the phasor approach outlined in section 5.4.2. For a large mass carried in the dumping floor, the value density held will be greater for the same per unit batch gain. For a greater flow of trucks, the accumulation rate will increase. As the dwell time of wastes increases, this leads to a large saturation limit with large relative costs of dumping and handling (there is longer queuing to access the dumping floor; the bulldozer/s must be activated for longer to plough the waste from a greater distance). This will erode the value density batch gain leading to a reduced value delivered to the load. The Q-factor on storage productivity will be reduced. It is known that the dumping floor has a storage capacity of 800 tonnes. With the compactor taking 12 minutes to complete its cycle of 27.2 tonnes, this would take close to 6 hours to clear. The access constraints become more pronounced when there is a large increase in arrivals to meet dispatch requirements of 2 train services/ day or more frequent. The relationship between work and inertia can be depicted as a phasor diagram. In this scenario, the evolution of a storage lag can be observed. This acts to give us information on both the average time delay, the consignment amount stored and the utilisation impedance (increasing limits on capability) of storage and handling resources. In this sense, the phenomenon captured is analogous to that depicted in cumulative storage curves used in terminal design and related to queuing theory. In this case, the planner may wish to deliver considerably more consignments to the train yard during the operating period in order to serve a load of two trains per day instead of the current one. This will require an approximate flow from the compaction units of 11.5 consignments/hour65. The main constraints are gate and dumping floor access limitations (no more than 2 trucks at a time) which mean that a maximum peak of 25.6 consignments/hour is permitted. The batching multiple requirement remains at 3.8 (from 7.2t to 27.2t). 65 In this case we do not consider the stack-siding dynamics which will require a further storage lag based on the access of train sets at the sidings and the timing of train departures. This will be considered in a further extension. Page | 274 Chapter 7: Waste Transport and Resource Recovery Case Study Table 21: Variable Changes around Waste Dumping Floor and Compaction LUP State Changes Values Units V 210 $/consignment 1 V2 Batched multiple of $/consignment 3.8 less value density drop I1 Max 25.8 Consignments/ consignments/hour hour I2 Approx 11.5 Consignments/ hour The transhipment calculus allows the planner to assess how the operating capacity at the dumping floor-compaction LUP is affected with increasing saturation of the dumping floor. The dumping floor-compaction interaction can be represented by an electrical transformer stencil66. Either an ideal or linear transformer stencil can be applied. A linear transformer allows the modeller to consider dwell time of the dumping floor (Figure 61). 66 Dumping floor- compaction LUP may also be represented by one or a series of canonical Operational Amplifiers. A transformer stencil here is demonstrated to further explain electrical circuit mechanisms with currency for logistics operations. Page | 275 Chapter 7: Waste Transport and Resource Recovery Case Study Floor dumping Coupling efficiency- Access related and mixing Compaction impedance degree to which we Impedance impedance directly drop at/into the pit Zs jwM R1 V2 R3 R4 I1 Output flow TX1 characteristic, V1 1 2 L 100V jwL jwL Z further modified according to business 0 Zab 0 rules using Zr RLC circuit Batching step change Figure 61: Transformer Circuit Stencil for Dumping -Compaction Operation Page | 276 Chapter 7: Waste Transport and Resource Recovery Case Study A signal is generated through this stencil based on value density, consignment flow rate, and level of transhipment intensity (time sensitivity). The resulting Bode and phase angle plots of Value Density based on desired batching and limits on consignment flux, I, allow the planner to read the off synchronisation rating values (in degrees) which match the desired Value Density batching gain and controlled outflow rate. The evolved storage lag in degrees is reflective of the work (Figure 62) Figure 62: Compaction Handling Bode Value Magnitude and Storage Lag (based on PSPICE output) The results are summarised in Table 22. Page | 277 Chapter 7: Waste Transport and Resource Recovery Case Study Table 22: Dumping Floor- Compaction Logistical Unit Process Results Rated Throughput V2 dB 63.9 dB 57.5 0 θ2 126 I1 46.5 25 I2 23 12.6 ωr 107.8Hz ωc 109.5 Hz The Logistical Unit Process has a certain synchronisation rating, ωc, based on planned transhipment intensity. Where flows and transformation work are outside this rating, ωr, a storage lag evolves, θ2. Here V2 is the Value Density in the compacted batch (in dB) after handling and compaction costs; I1 and I2 are consignment flux in and out of the process as truck consignments and compacted containers respectively (per peak hour). This element of terminal activity is analogous to the container stuffing and unstuffing functions in satellite intermodal terminal connected to a seaport. This represents a technique in addressing the urban complex bundling problem. The problem description is presented in Chapter 6 with the Enfield Hub problem. Page | 278 Chapter 7: Waste Transport and Resource Recovery Case Study 7.5.3 Itinerary Control and Streaming Wastes The impedance depicted in each Logistical Unit Process (LUP) represents both work and control functions from the application of business rules. The precision and control settings for each LUP interface are enumerated by the poles and zeros of the impedance function. These are established through the Bode plot analysis for that LUP. These functions represent changing costs and resource consumption as well as itinerary control. The business rules to operate an LUP deal with terminal resource availability and overflow itinerary routing. Itinerary control is a mechanism to allow for the model to represent complex terminal topologies. These are the process flow multi-path couplings that can occur through a terminal for instance, with flows subject to direct and indirect transhipment. In this thesis, itinerary is the path taken of commitment and dispatch. LUPs must be checked for their availability such as storage space and resources, and flows must be filtered for required train services on the basis of payload and timing. In this way the itinerary chosen harmonises with the service pattern dispatch desired. With increasing tense flux rail services and the pursuit of network coordination opportunities, itinerary control becomes a valuable mechanism as inflows represent different waste types and/or different destinations, requiring different routing through the terminal, different storage locations, different loading positions on the train, and even different train services. It is foreseen that itinerary control could be used at the waste transhipment terminal to control a number of conditions (Table 23). The compensation control could be addressed with a lag- compensator. It can be surmised that application of itinerary control assists in evaluating the effects of space allocation strategies for it recognises that the destination within a terminal for containers of a group is not static, rather it depends on their arrival times (Taleb-Ibrahimi et al., 1993). In this respect, bandwidth is the parameter used to control the active arrival duration of the source inflow according to train service dispatch requirements. Page | 279 Chapter 7: Waste Transport and Resource Recovery Case Study Table 23: Itinerary Control in Waste Transhipment Terminal Activities Condition Parameter Modification in Control Alter acceptable active arrival duration band Maintain Gain, K, and widen Bandwidth for flows to map to specific train service schedules Storage overflow routing- divert to direct Narrow Bandwidth, introduce new pathway to transhipment or bimodal diversion sidings loading or to truck dispatch Account for more regular inflow of smaller Increase batch gain, narrow bandwidth carrier size Account for less regular inflow of larger Reduce batch gain, widen bandwidth consignment sizes The compensation control elements are an extension of the transhipment calculus, the objectives and mathematics of which are described in the Appendix. 7.5.4 Sidings Capability The terminal is currently configured with single ended access to the west. The production and storage facilities stand in-between two active sidings of approximately 400m length each. There are three additional sidings to the north which are used to store, empty wagons, load wagons pre- dispatched, and arrival wagons with empty containers to be unloaded. With the current single return dispatch per day. A train with empty containers arrives at approximately 9pm. Half of the train length (400m) is shunted onto an active siding; the other half is shunted onto a holding siding. The locomotive then collects one half of the new train production from the southern most active siding and then collects the other half of the new train production from another holding siding which was stored there mid afternoon after production of new container consignments. The train is then due to depart shortly after 11pm. In the Clyde case, the three holding sidings allow train rakes to be cycled Page | 280 Chapter 7: Waste Transport and Resource Recovery Case Study to and from the active sidings allowing the unloading of empty containers and loading of production containers and building a train length of 800m. The current cycling arrangement of train rakes leaves a 5 hour window at one of the active sidings. Additionally, the current limiting factor is the single train over a long line-haul (230km each way). Currently production rakes are sitting either at the active sidings or in a hold siding for long periods of time and this occupies the sidings capacity. The sidings would have capacity to accommodate additional services and novel rail operating forms if more train services were scheduled. 7.5.5 Feedback Loops Between and Within LUPs Unlike the CVSD approach, resource attributes are not explicit with the transhipment calculus approach. All necessary information regarding the consignment density and its carrier are contained in the signal. The dependency on availability of carriers: trucks, containers, wagons and trains to move consignments are not explicit. A limitation with the transhipment calculus then is that is does not directly enumerate and predict the scarcity of these supporting resource flows. Resource needs can be proxied. For instance the arriving truck is considered as part of the consignment and the access and handling impedance it experiences is also experienced by the consignment. For the sketch planning purpose of the model this is acceptable as the focus of the analysis and design is matching work and itinerary impedance through the terminal with the desired output flow profile. Mobile resource availability is considered an operations problem rather than the strategic- tactical concerns of the transhipment calculus. Critical feedback cycles are: 1. Return trains delivering containers for production and wagons for transhipment loading, 2. Triggers of truck arrivals, container filling, and wagon filling. The availability of empty containers is critical for compaction of waste. These empty containers are released on train arrivals either immediately or after some delay depending on shunting impedance. They arrive as a block. Servicing flows such as empty containers act as a filter signal to control the flow of consignments that are to be compacted. In this way they act as a source impedance to other flows to be transformed. Page | 281 Chapter 7: Waste Transport and Resource Recovery Case Study Filled containers also provide a means of trigger bulk queue control of train dispatch. In a simplified form three electrical filter stencils can act as information control triggers for guiding the itinerary of source inflows. The first filter acts as the source impedance of the empty container returns. This filters the unconsolidated waste flows for production limited to feedstock containers. The utilisation of the dispatch service is also limited to the ability to load sufficient containers against the scheduled requirements of the sidings/mainline system (the train dispatch control law) and the availability of wagons. This is formulated as additional filters. Constraints include stack quantity and wagon availability at sidings. Where there are “logistical” losses between the outflow services and the cumulative inflows, the system is miss-specified for that rail operating system. Flows not passing filters will overflow to a storage bay. The simplified uni-directional flow is depicted in Figure 63. Page | 282 Chapter 7: Waste Transport and Resource Recovery Case Study Figure 63: Physical and Logistical Flowchart of Waste Transhipment Terminal Page | 283 Chapter 7: Waste Transport and Resource Recovery Case Study 7.6 Inter-Terminal System Formulation 7.6.1 Protocol Description The location of compaction alters the consolidation-transhipment sequence of various intermodal production systems. In this section we describe five service design options which combine consolidation principles, relations between satellite transfer stations and intermodal terminals, and rail operating forms. These offer different prospects for the utilisation of intermediate storages and rail frequencies. These alternatives involve great complexity in terminal operation and link arrangements and protocols are developed after O’Kelly et al. (1994) and Ballis et al. (2004), in order for us to analyse this complexity. Such a protocol allows us to understand the distinguishing features of alternate intermodal production systems such as the routing of wastes and interaction between container and train production cycles. This has implications for the transfer technology to be applied, storage requirements and terminal sizing, for instance. The protocol is presented in Table 24 and describes the system formats. Table 24: System Formats to be Studied According to Transhipment Protocol System Strategy Network Location and Frequency of Transhipment Format Description Operating type of service type form compaction 1a Base case Direct truck Clyde, FEU 1/day Indirect to rail 1b Clyde Dense Direct truck Clyde, TEU 2-3/day Direct/Indirect to rail 2 Clyde Dense Direct truck Satellite, or 2-3/day Direct/Indirect to rail Clyde TEU 3 Satellite to Road-Rail hub Satellite, TEU 2/day Direct Clyde (B-doubles) 4 Matraville Rail-Rail Satellite, TEU 2/day Direct Satellite to Intramodal Clyde Trunk and Feed 5 Clyde to Liner Mixed, TEU 2/day Direct/Indirect Goulburn via Campbelltown Page | 284 Chapter 7: Waste Transport and Resource Recovery Case Study 7.6.2 Signal Pulse forms System format scenarios vary in terms of their network operating form and terminal configuration: whether they are road-rail, or rail-rail transhipment terminals; the hub nature of the terminal; and the transformation requirements within the terminal. Each strategy represents a change in operations based on the existing interface arrangement and configuration, such as batching densities required, and changes in source flow distribution. Business rules being the deployment of the system format are included as strategies. Where strategies do not perform well with the system format applied, a new system format must be devised. This may or may not involve a retrofit of the physical infrastructure. The performance of each system format and related deployment scenario can be characterised by how it meets specific dispatch criteria which is described by its own pulse signal. Each system format has a value density, consignment dispatch rate and a timing designation. The timing designation will determine the carrier pair match and the sidings utilisation. The dispatch flow payload captures these three attributes. It can be represented by sinusoidal signals of consignment flux and value density and thus has Fourier transforms. As was discussed in chapter 5, Fourier techniques of representing flows allow analysis of payload and precision. Table depicts broad relationship changes with alternate system formats. Table 25: Payload and Synchronisation Band Change with Waste Transport System Format System Mathematical parameter description in accordance with pulse train form Format compared to base case 1a Basecase 1b Higher payload 2 Higher payload 3 Smaller dwell time with direct transhipment, greater frequency 4 Pulse correspondence of bandwidths for adjacent sidings 5 Sidings capability for multiple bandwidths Page | 285 Chapter 7: Waste Transport and Resource Recovery Case Study 7.6.3 Specification and Method Each system format will have a specific relationship between its production cycle and transportation cycle. Each system format scenario and strategy has a set of specifications against which its performance may be measured. This will include load attributes of the output dispatch pulse (timing, value density). There will also be specifications according to each Logistical Unit Process (LUP). These LUP specifications cover: the Quality factor (storage productivity) and Gain Bandpass Product (the total mass-value permissible through the LUP for a given period). There may also be transient stability requirements such as the allowable percent overflow of a storage buffer67 In the following section, 5 system formats of different rail operating forms are described. These will have different impedance fluxes on the waste terminal. Section 7.8 outlines the method of analysis and the Figures of Merit that need to be tracked to assess the impact on each terminal and lead to design improvements. The methods highlighted cover those illustrated earlier in this chapter as well as techniques outlined in chapter 5. The analysis is left for future work. 7.7 System Formats - Description and Specification 7.7.1 Base Case The first scenario proposed (S1) depicted in Figure 64 represents two variations on the same system format. The first involves compaction to 27.2t into FEUs at Clyde terminal. For the second the network scheme remains the same with the same waste feed profile and conventional train form, but a reduction in the dimension of containers is taken into consideration. The FEU containers are replaced by TEU containers that can fit 18 tonnes of residual waste each. Because of their smaller length, two containers can be placed onto one wagon, increasing payload capacity in each dispatch and leaving other conditions and constraints of the system being the same. This change will help in decreasing the waste that needs to be stored over the night in Clyde as it allows a greater mass to be dispatched daily per train service. Concurrently a larger storage area will be required in the stack. 67 As measured by the phase margin on a Bode plot. See Figure 39: Determining Design Parameters from Stability Specification. Page | 286 Chapter 7: Waste Transport and Resource Recovery Case Study Figure 64: Clyde Base Case- Clyde Compaction to 27t and then to 18t 7.7.2 Clyde direct improved compaction In Figure 65 satellite transfer stations are introduced, interfacing the transport of waste from households to Clyde. The collection will still begin from kerbside with high density compaction enabled by a new technology for trucks will be introduced with 9 tonne containers. This container will continue to the final destination. In this way there will be a consistent reduction in the compaction-uncompaction activity and an overall acceleration of the whole transhipment design. Instead of being driven directly to Clyde, waste will go to intermediate satellite terminals. Here transhipment to B-double trucks takes place. Consequently containers are dispatched to Clyde. The existing infrastructure will permit the road route to be utilised. Two containers can be carried on B-double trucks to Clyde. No connection between satellite stations themselves is required. Compaction does not need to be undertaken in Clyde. Up to 4 containers per wagon can be loaded. The repatriation of empty containers by trucks to the satellite stations must occur. These containers are then connected to collection truck chassis. A compaction stream continues for waste shipments where the source catchment is nearer to Clyde than a waste transfer station. Figure 65: Road Rail Direct Improved Compaction- Clyde Process Streaming Page | 287 Chapter 7: Waste Transport and Resource Recovery Case Study 7.7.3 Satellite to Clyde: Road-Rail In case 3, Low density trucks are used for collection and are transported to satellite terminals where the compaction process to TEUs takes place. These are then transshipped to B-double trucks. In this scenario the sole task at Clyde is transhipment of containers from trucks to wagons (two container per each wagon) and management of dispatch and arrival of train services. In this case, the B-doubles are used to repatriate the empty containers to the satellite terminal. Figure 66: Satellite Hub- Clyde Cross Docking 7.7.4 Matraville Satellite to Clyde: Direct Rail to Rail Scenario 4 (Figure 67) represents a feeder trains plus trunk train operating form, where each satellite station will be connected to Clyde through a rail connection. Compaction will be done in each satellite terminal, where a couple of containers will be loaded on a single wagon. Different train services will arrive at Clyde, where again a unique 54-wagons vehicle will be assembled and dispatched. Again the service design from Clyde to Woodlawn will proceed as shuttles. Empty Containers will be repatriated to the satellite terminals by rail. Figure 67: Rail Trunk and Feed- Clyde Pulse Correspondence Page | 288 Chapter 7: Waste Transport and Resource Recovery Case Study If portal gantries are used to transship containers between train services the classification of wagons can be avoided. That is, load units rather than wagons are exchanged. Classification procedures will be increasingly onerous with an increase in the number of origins and destinations. 7.7.5 Clyde to Goulburn via Campbelltown: Liner Train The current waste intermodal terminal at Clyde acts as a regional intermodal terminal and has a conventional rail operating form. That is, it benefits from the day and night leap in production and has the flexibility in train block formation through sidings (Woxenius, et al., 2005). The container and train production cycles are separated in time sufficiently that there is much slack capacity in the container production cycle. If we were to consider closer located recovery plants, more frequent train services, under the liner rail operating form may be more compatible. The dual operating cycles now have an imperative to synchronise under an increasing tense flux condition. Tense flux describes the operating role of intermodal terminals, particularly in a dense network (Rodrigue, 2001). The translocation of residual (and possibly garden) wastes to regional NSW may be further facilitated by a liner rail operating form (Figure 68). In this service design, we consider the possibility of shipping wastes to Woodlawn via Campbelltown. There, residual waste containers are off-loaded for feed to a proposed mechanical-biological and Energy from Waste facility near Jack’s Gully landfill. Commensurately, the train can take on garden waste consignments that can be further shipped to Woodlawn for composting. On the return journey, the train can delivery empty containers for reuse. These arrangements facilitate more network options for the recovery cascade of wastes. The rail service changes composition along its route according to the intermediate stop requirements. This allows different waste types to be delivered to different destinations. Page | 289 Chapter 7: Waste Transport and Resource Recovery Case Study Figure 68: Liner Operating Form- Clyde Streaming, Compaction and Ordering The role of Clyde here is possibly enhanced. Either wastes are pre-streamed and delivered separately or Clyde acts as a mechanical-biological plant as well as a container management and gateway transhipment terminal. In this analysis, it is assumed that Clyde receives pre-stream wastes and the focus of the analysis is on the impedance arrangements in enhanced space management. Shipments arrive both containerised and uncompacted. A separate compaction facility for green waste is required. 7.8 Analytical Method and Functional Requirements 7.8.1 Analytical Method For an analysis of each system format, the following steps are recommended: 1. A system schematic is depicted; 2. A terminal flowchart is devised; 3. Specific Logistical Unit Process (LUP) transfer functions are enumerated; 4. Each LUP is connected to other LUPs according to a specific itinerary and this is validated by complex conjugates to ensure acceptable precision and flow; 5. Block algebra reduction is performed; 6. A process control model is formed which formulates the desired output as a reference signal and the input as a disturbance; 7. A test is made that terminal configuration is able to deliver the desired output signal; 8. The flowchart is transformed to an Multiple Input Multiple Output (MIMO) system; 9. Cumulative flow diagrams are presented as a means to validate the results. Page | 290 Chapter 7: Waste Transport and Resource Recovery Case Study 7.8.2 Figures of Merit Supporting Functional Requirements The design requirements and feasibility of each of these system formats can be assessed according to specifications measured by Figures of Merit. Key waste terminal operations, their specifications, and Figures of Merit are linked in Table 26. Table 26: System Format Relationships, Specifications and Figures of Merit Terminal Operating Specification Figure of Merit Relationship Consignment arrival – Flux; Saturation limits of Value Density batching, Dumping floor –Compaction dumping floor Consignment flux, storage lag congestion Container Storage Capability: Space- Handling relationship Bode Plot: Q-factor for space access and handling acceptable synchronisation Storage Claim band. Itinerary control for storage Triggers for sidings activation Synchronisation band filter overflow, sidings access and (pre-dispatch loading), matching specific train bimodal overflow services Cost- Value implication of Logistical Unit process and Impedance, Synchronisation system format terminal cost for capacity at a band curves to depict certain transhipment intensity operating “resonance” point From an application of the pulse train for each system format, it is evident that the sidings configuration and operation will have problems with addressing more than 2 train services/day. The current plan to remove storage sidings in order to make way for more stack area may be miss-specified if the longer term aim is to significantly increase container waste throughput. Compaction can be arranged at satellite transfer terminals and this would allow increased direct transhipment intensity at Clyde, relieving the dumping-compaction-storage bottleneck. Retaining sidings for pre-dispatch loading may yield more improved flexible capacity outcomes. Page | 291 Chapter 7: Waste Transport and Resource Recovery Case Study 7.9 Chapter Conclusions The system dynamics model demonstrated severe constraints at the Clyde terminal with a doubling of waste inflows under the current infrastructure configuration and rail operating form. This assessment identified that in order to improve the throughput of this facility, a re- configuration of the production and transportation system was required. These new system formats considered initiatives that included: 1. Earlier compaction in the waste management chain that led to direct transhipments at the intermodal facility which bypassed the dumping floor; 2. Further early consolidation and batching of flows towards a state of pulse correspondence at the intermodal facility (train transhipment to trains). 3. Inclusion of additional waste streams and transhipments punctuating the rail service with intermediate stops (liner system) to deliver to alternate reprocessing plants. This chapter illustrated elements of terminal operations by applying the new impedance methods developed in Chapter 5. These activities covered: 1. Active Arrival Area and Stack Management; 2. Itinerary control; 3. Dumping-Compaction Activity; and 4. Sidings activation. The transhipment calculus outlined in these mechanisms demonstrated the insights of alternate terminal complex activity fulfilling specific service designs. Essentially the sketch-planning technique demonstrated here can identify what distributed functions through the satellite- terminal hinterland should occur (i.e. location of compaction activity) in order to assist the capability of the intermodal terminal to meet alternate rail operating forms. Aspects of terminal impedance- work and control rules are enumerated. Future work would 1) assess the impedance flux impact terminals under different rail operating forms given the dispatch signals they generate; and 2) incorporate timing in flow variables so that impacts of greater dispatch service frequency can be assessed and such conditions of tense flux can be designed for in the terminal. Alternate aspects of terminal complex activity required could then be further identified in the provision of waste streaming and recovery of multiple waste fractions such as C&I and green waste. Page | 292 Chapter 8: Conclusions 8 Conclusions 8.1 Re-statement of Problem 8.1.1 Transitions Management Ambition The urban landside container freight task exhibits traits of a complex industrial eco-system which resists easy solutions in addressing energy and resource dissipation and social disamenity. The landside container freight task is perhaps the task in transportation where the evolution of system deficiencies of infrastructure supply in node and links is most noticeable in societies dependent on export income and import consumables. Road transportation activity has become increasingly dissipative: networks are used opportunistically and techniques in optimising truck fleets may increase freight intensity (km/t), a measure of increasing ineffective use of infrastructure. The generation of futures scenarios in the freight task depict trajectories of unsustainability. These pressures can only be addressed by concerted efforts to re-engineer the logistics structure. One means to achieve this is to implement combined transportation system formats. Rail-road Intermodal terminals and hub networks are a possible solution to drive down freight intensity of the landside container freight task. Combined transportation offers demand management solutions by fostering consolidation networks. The act of combining separate modal networks is a retrofitting task of transitions management. Here intermodal terminals become articulation point gateways helping to conjoin mono-modal networks and improve effectiveness in time and space (s.4.5.2). Demand management for transportation services is akin to electricity demand management which controls the timing and magnitude of demand i.e. importers waiting for consignments can have a tolerance for later in the day delivery and this allows container arrivals by rail to local intermodal terminals instead of dissipative activity of single truck trips. Empty container services generated from intermodal terminals means that individual trips do not need to be made to the seaport for collection from or delivery to central storage. This opportunity is analogous to the benefits of distributed power storage and generation in minimising losses in a distribution network from centralised generators. In order to present a transportation science of intermodalism the nature of change dynamics must be characterised. This dynamics may be termed, after Rodrigue (2005), tense flux. This dynamics recognises that flows of consignments are based not solely on origin and destination generators and attractors but also induced demand behaviours of the logistical structure of actor Page | 293 Chapter 8: Conclusions relations, operational requirements and the available infrastructure stock. The performance of a terminal depends on how well it synchronises carrier flow relationships. Modelling flexible storages and service generation control at intermediate terminals is critical in assessing novel transhipment opportunities. Networks are not just to be seen as distributive mechanisms but also accumulative in order to foster consignment bundling activities. Networks of alternate means of transport need to be accessed by gateways nodes such as rail-road intermodal terminals. The means of transport become complementary rather than substitutes. Models of urban intermodal relations require new criteria of physical realisability or feasibility. New measures of precision between flows are required. Secondly, techniques of controllability at the terminal need to be designed. This allows terminals to be considered for their capability rather than strict practical operational capacity. 8.1.2 System Context in Sydney and Melbourne Container freight activity in Australia’s largest cities is particularly intensive: container origins and destinations within the city radius of 40km represent some 80% of container volumes for both Sydney and Melbourne. Recent activities to increase the sea-side capacity of container ports in Sydney and Melbourne will achieve only limited objectives if landside capacity constraints are not considered. Despite mobility initiatives to inject further capacity into the road physical infrastructure, there is increasing access congestion at seaport gates in Sydney and Melbourne. The logistical structure does not align with operational relationships and thus trucks have poor load rates and access demand is especially peaked leading to slot access utilisation of only 51% throughout the week. Urban intermodal operations occur in Sydney and contribute to some 20% of urban sourced international containers going by rail. It has not been possible to increase this share. These terminal arrangements are largely conventional direct connect and advanced rail operating forms, and terminal hubs have yet to be developed. Melbourne has no current urban intermodal terminals in operation. Growing congestion levels have more recently placed pressure on developing alternate modes for seaport access in Melbourne. Plans are being made to facilitate intermodal connections in the west and south-east of the city. Both cities face infrastructure and logistical challenges in facilitating 30% or greater of container volumes on rail yet this has been a mission of governments for several years. Page | 294 Chapter 8: Conclusions To achieve significant container volumes on rail, planners need to develop feasible intermodal production systems which include the limits and opportunities of existing infrastructure and an understanding of the logistical structure this will require. Urban intermodal relations will be characterised by tense flux environments where consignments need to be consolidated and transhipped onto frequent rail services. These tense fluxes will add to the vulnerability (capacity-risk) of operations unless there are significant mechanisms for stack overflow such as bimodal diversion and sidings activation. Both precision and control design techniques are required to assess the physical realisability of urban intermodal operations. Planners have pressing needs to understand what functions intermodal terminals should adopt in order to foster their uptake. They also need to assess what combinations of infrastructure retrofit and application of broad governance strategies to support operating business rules are required to manage a transition of the prevailing landside container logistical structure. Planners also face a hazard of miss-specifying the problem leading to more dissipative solutions in attempts to optimise”business activity harmonisation” objectives. The problem is an economic-engineering one, requiring a consideration of leverage opportunities with existing physical infrastructure to generate closed loops and higher productivity in the logistical structure. 8.1.3 Investigations for an Intermodal Production System This thesis has investigated the root of the harmonisation question in container seaport- hinterland relations. This is a sub-set of the wider discussion in analysis and design of combined transportation system formats to address the landside container freight task. Waste transportation of fostering dense flows through a gateway intermodal node is another sub-set of this investigation. Through an analysis of the literature - theoretical and public policy, stakeholder interviews, and national dynamical systems analysis using the ASFF vintage model, it was found that there is no obvious technological optima: many logistical, infrastructure as well as technological solutions are required to lead to a major densification of carrier loads, carrier switching to more benign environmental transportation modes and new vehicle technologies. These initiatives could set a downward trajectory in energy use and decouple the freight task from economic activity. These broad parameters can only be achieved with novel combined transportation arrangements. Freight intermodal terminals offer significant productivity leverage of the existing transportation infrastructure. Page | 295 Chapter 8: Conclusions The harmonisation question is at root a problem of analysis and design of the absorptive capability of the hinterland which by fostering the means of accumulation provides a leap in accessibility opportunities. Improving accessibility is the true hallmark of intermodalism. Network modelling tools are predominantly predictive rather than explanatory in observing the effect of new network and node gateway arrangements. This thesis develops a model formulation for more effective design and analysis of how terminal functions can interface with the rail operating forms. These rail operating forms are necessary for retrofitting intermodal solutions into urban transportation network infrastructure. Analysis and design tools to assess the complexity of different terminal topologies are developed using electrical circuit theory and process control theory. It is anticipated that these analytical impedance functions will also act as load following devices for the coordination of freight container flows between nodes. This is seen as particularly pertinent to investigating harmonisation initiatives, including control of empty container movements. This future work is flagged in Appendix F. An investment logic map was devised to illustrate the research perspective this thesis took to address significant materialising pressures on the container freight task and to offer techniques in analysis and design of intermodal operations which improve seaport-hinterland productivity. This map is re-presented to aid the concluding comments (Figure 69). Drivers for the investigation included a broad concern with the current dissipative nature of freight activity and the poor formulation in many transportation models of an activity-based approach where demand management initiatives could be generated. Page | 296 Chapter 8: Conclusions Figure 69: Research Perspective for Urban Intermodal Freight Transport Thesis Page | 297 Chapter 8: Conclusions 8.1.4 System Formats Assessed In chapters 6 and 7, the currency of the transhipment calculus is tested with a number of system formats assessed. System format scenarios vary in terms of their network operating form and terminal configuration. They lead to a different relationship between production and transportation cycles. These system formats were formulated based on known baseline operations. As these system formats represent major logistical and configuration planning changes, where there is no operating data, parameter relationships can only be partially validated. As this thesis is a practical application of the sketch planning endeavour, the motivation for mathematical models is exploratory/explanatory rather than tightly predictive. Table 27 summarises the intermodal system formats illustrated in the case studies with mechanisms investigated of certain attributes and relationships broadly sketched for further work. Table 27: Transhipment Calculus Attributes Described Through System Format Case Studies System Format Attributes Relationships Altona cluster on mainline— Storage dynamics between Seaport precinct-terminal trunk and feed terminals interaction shared stack capacity Sidings configurations Sidings capacity Sidings- dispatch capability Enfield - Collection and MIMO, Complex bundling Catchment-bundling metric Distribution Clyde Waste Terminal Stack Management Bode plot design Dumping floor dynamics System formats of alternate rail operating forms described Response to alternate rail implications on stack and operating forms sidings capacity Tri-lobal network-hub and Empties repositioning Use of load following devices spoke (Outlined in Appendix for coordination F) Page | 298 Chapter 8: Conclusions 8.2 Contributions to Knowledge 8.2.1 New Phenomenological Approach The opportunities for urban intermodalism scoped in this thesis demands a new way of looking at the transhipment problem of Operations Research. As the field needs to develop from a conceptual paradigm to a discipline, it requires new metaphors or sign-posts to enable the planner to break the constraints of rationality and practice that inform transportation science. Electrical impedance offers a powerful metaphor. Analogy to other science theories like electrical power circuit theory enable “development of routines for the new field to become a discipline” (Ehrenfeld, 2004). This analogy supports original research in reclassifying the essence of phenomena and problems and providing improved decision support. The analogy establishes a vital conceptual validity to the problem. This thesis has used the analogy of electrical circuit relations to elegantly articulate conceptually the relationship between time and timing and consequently terminal space management and synchronizing freight flow profiles. The analogy is a conjecture that remains to be proven, The analogy taken to the extent in this thesis remains at a point of contact which is suggestive and supportive rather than substantiating a comprehensive model formulation of specialized and general rules (Polya, 1954). This is fertile ground for future research. Freight optimisation schemes are cost minimisation based. They have problems with including economies of scale, of scope and of density in their objective function. They do not consider precision relations except as a penalty function for certain saturation levels. These model formulations do not explicitly seek bundling opportunities to reduce truck trip numbers. By making precision relations explicit we can design terminal components to achieve precision objectives using the measure of characteristic impedance. The new conceptualisation of impedance allows the planner to design terminal unit processes as the basis of necessary production work rather than simplifying impedance phenomena to a cost that must be minimised. New dynamics of change which are specific to urban intermodal system formats can be modelled and guidance can be given on opportunities and tensions. The transhipment calculus allows a fresh look to be taken at defining problems associated with the logistical structure of urban container transportation. Driving forces are captured, impedance mathematics is defined for precision and flexible capacity relations and the dispatch requirement information in terms of timing, frequency, volume, and scale is incorporated. This tool, applied at the sketch planning level, works towards reducing the tensions and dissipative activity in a logistics framework by providing a basis for a robust governance framework. This framework Page | 299 Chapter 8: Conclusions assesses the use of the physical infrastructure by precision-impedance relations with the application of business rules through compensation control. The planner can now assess initiatives on how they will affect gateway harmonisation objectives. 8.2.2 Terminal Dynamics and Terminal Complex Activity This thesis formalises a means to consider terminal complex activity into the multi-modal, multi-commodity problem which can be summarised succinctly in cumulative flow-time charts. Salient contributions to capturing multi-path coupling include: 1. Ability to analyse the impedance relationship between carrier pairs and a number of source carriers and combined carrier load; 2. Enhancing capability in developing analytical closed form solutions for the bulk queue phenomena and sidings capability; 3. Representing itinerary control through the terminal as a path of availability and dispatch; 4. Including dwell time as well as active arrival duration in assessing impact on terminal stack storage; and 5. Combining expressions for desirable payload and sidings availability which correspond to specific train operating forms to be tested. 6. Providing a means of value stream mapping through terminal Logistical Unit Processes (LUPs) The work and value changes within the terminal remain to be correlated with specific terminal activities, handling technologies, terminal layouts and flow profiles. Suggestive points of contact are made regarding these relationships which act as a starting point for further investigation. Page | 300 Chapter 8: Conclusions 8.2.3 Performance Metrics of System This thesis found that there were no credible indicator schemes for assessing the performance of integrated, combined transport implementation or for advising on their optimum infrastructure investment deployment in an existing network. The concept of intermodal balance became a starting point to formulate a hierarchy of performance indicators, supported by Figures of Merit which could assess the pertinent trade-offs in implementing combined transport in the freight task and achieving harmonisation objectives firmly rooted in combined transportation formats (Figure 70) . Figure 70: Performance Indicator Hierarchy of Harmonisation Attributes for Freight Intermodal Balance These harmonisation attributes can be translated into impedance specifications for terminal performance. Figures of Merit were defined. These Figures of Merit dimension the physical realisability and feasibility of the urban intermodal production system by testing whether specifications can be met with the system format at play. The Figures of Merit then allow the planner to re-design the terminal configuration and/or apply modified business rules. The Figures of Merit defined and illustrated in this thesis extend to: 1. Marginal costs and throughput level, 2. Throughput and capacity-risk, Page | 301 Chapter 8: Conclusions 3. Transhipment intensity and impedance level, 4. Complex bundling and controllability, 5. Resource costs/container and transhipment intensity (summary of terminal functions), 6. Precision and Flexibility (Cross-docking speed vs. complex bundling opportunities, for instance), 7. Business circulation rules, stack control for stability and systems utilisation. Bode Plots were discussed in Chapter 5 and presented in Chapters 6 and 7 as a means to graphically depict Figures of Merit information. They allow use of timing concerns in the synchronisation domain to assess multi-commodity relations and the effect of business controls at terminals. 8.2.4 Terminal Multipath Coupling For any new freight methodology to have tractability it must be able to accommodate issues arising from the multi-commodity, multi-modal (MMMC) flow problem associated with freight transhipment. That is, multi-attribute data needs to be incorporated involving consignment information of origin-destination pairing and synchronisation needs. This is particularly the case when the functions of urban intermodal terminals are considered and the need arises to assess and foster opportunities for complex bundling. The complex bundling concept and draft formulation presented in this thesis may be seen as a critical phenomenon to capture in urban intermodal feasibility. Critically, the capacity issue needs to be reformulated as a capability issue. The productivity of the seaport critically relies on its immediate landside and hinterland satellite infrastructure, resources and logistical arrangements and their flexibility under different flow profiles. The transhipment calculus developed in this thesis offers tools to measure and design attributes of hinterland absorptive capability. The mechanisms that it characterises include: 1. Complex bundling, 2. Assessing necessary storage reserves with storages required, 3. Sidings activation, 4. Bi-modal overflow opportunities, 5. Empty container movement control. Page | 302 Chapter 8: Conclusions These mechanisms illustrate the usefulness of the transhipment calculus in granting planners the means to incorporate changes in the state of supply and thus a measure of capability is possible. 8.2.5 Addressing Business Rules A representative model needs to incorporate different business rules in the operation of the terminal which affect its performance. These business rules are operational activities that change the response of the terminal system to different flow patterns. The implementation of business rules demonstrate the flexible capacity of a terminal and thus link the economic with engineering decision underpinned or constrained by the existing asset. The transhipment calculus and the mechanism of filter impedance functions act as terminal itinerary control and track the effect of certain operational business rules. In this thesis, business rules were scoped and illustrated with the transhipment calculus covering: 1. Bulk queue control laws, 2. The management of storages, 3. The operations of rail sidings, 4. The management of empty containers, and 5. Complex bundling initiatives. The inclusion of governance structures and business rules grant planners the ability to define and assess trade-offs in openness, complexity and controllability so that the feasibility of the intermodal system format can be assessed. 8.3 Areas for Future Research 8.3.1 Building the Analogy and Parameter Identification This thesis established suggestive points of contact towards developing similarities between intermodal phenomena and electrical circuit theory in the pursuit of a more activity based approach for urban freight intermodal tactical planning. Associated with this, broad parameters were outlined. These parameters represented critical storage-handling relations at Logistical Unit Processes within a terminal. These parameters supported the conversion of driving forces into driven forces through impedance flux and allowed the representation of a number of freight Page | 303 Chapter 8: Conclusions terminal phenomena. These parameters included the storage saturation coefficient, the handling saturation coefficient, variable and fixed costs, dwell time, and evolution of the storage lag. A crucial next step in model development of the transhipment calculus is building supportive points of contact with the analogue. This would involve validating broad flow and flux relationship according to the accepted tool of cumulative storage curves outlined in detail in Appendix B2. Secondly, parameter identification would occur through experimentation of different circuit stencils in representing Logistical Unit Processes. That is, confirming the applicable range of values for these parameters under different canonical stencils, representing alternate Logistical Unit Process (LUP) configurations under a variety of source and load flows. Developing parameters for the novel technologies and terminal configurations to support short distance, small consignment freight flows that characterise the complex bundling arrangements of urban intermodal transport would be a valuable focus. 8.3.2 Compensation Control The thesis established that controllability with precision were significant attributes to represent for model physical realisability. Thus the thesis also scoped issues in compensation control for LUPs. Control of the stack by modifying active arrival durations and filters to control sidings availability were examples illustrated in this thesis of compensation control. Further work could be undertaken in designing compensation control mechanisms for terminal LUPs to gain a more detailed insight into terminal flexible capacity (capability) under load flow triggers. Typically terminals and facilities are designed using simulation techniques. The essence of the innovation of this thesis is the development of some analytical tools, taken from electrical circuit and control theory which may lead to “back end” design known as impedance synthesis. This addresses precision and controllability imperatives which have been largely lacking in terminal simulation techniques and are critical in evaluating the physical realisability of different terminal system formats to deliver a feasible service. Bode plots, as specifications of impedance transfer functions, can be used to alter the impedance necessary for the desired output. Bode plots give a pole-zero mapping which can be attributed to a stencil of a canonical form which represents the LUP in question. Specific parameter values can be calculated for handling resistance and storage coefficients for each LUP. The Bode plot can represent the specifications for one LUP or the number of LUPs in the itinerary a flow will make through the terminal. Feedback for controllability purposes can also be introduced through the Bode plot. The full exposition of this is flagged as future research. Page | 304 Chapter 8: Conclusions 8.3.3 Building Coordination Schemes 8.3.3.1 Load Following Basis The harmonisation attributes in the urban-international container freight task can be addressed in a more superior fashion through power systems theory. This science theory allows us to consider the phenomena of transient storages and has an extensive body of economic-security algorithms. That is, power systems theory considers both price signals and access availability due to variable operating constraints. In this way demand management objectives can be tested explicitly. The role of the transhipment calculus is to provide a load following mechanism for the coordination task. The transhipment calculus represents the attraction-impedance characteristic step that drives network transport models. The transhipment calculus captures impedance behaviour in a novel fashion in that 1) the node is considered the critical cause of impedance over the links; 2) impedance is more than a delay measure; 3) impedance is not only a capacity constraint phenomena but also a measure of precision and work necessary for required consignment transformations. The consideration of system coordination extends the scope of measurement of hinterland absorptive capability to include for instance the effect of truck and train circulation rules on terminal capability and system throughput capacity. This thesis has set up the mechanism for design of the topological complexity of a terminal as a means for use in systems design for coordination schemes. Appendix F scopes the role of transhipment calculus into a coordination mechanism to improve system productivity. The empty container problem is scoped here. Techniques provided by Power Distribution Theory offer opportunities to investigate means of presenting the freight task as a distribution resource system. This is a rich area of research leading from the fields studied in this thesis. Page | 305 Chapter 8: Conclusions 8.3.3.2 Augmenting Multi-Commodity Multimodal Flow Problem The thesis developed the concept of multi-path coupling whereby information pertaining to carrier timing requirements is carried through in the synchronisation domain to filter flows. Multi-path coupling can be incorporated into the Multi-Commodity, Multi-modal flow problem so that throughput capacity of various itineraries can be assessed as a path of commitment (availability) and dispatch (utilisation) as part of network coordination schemes. This then allows capacity to be considered as a compensation and coordination phenomenon rather than simply as a constraint basis. Liberation from such a reductionist approach may enable a better understanding and management of the effective deployment of infrastructure to harmonise the logistical flows. 8.3.4 New Transportation Science Fields for Application This thesis provides tools in the analysis and design of terminal functions in the framework of conceptualising networks as distributed storage and resource systems. System retrofit opportunities described in chapter 6 and 7, and in Appendix F, in particular the liner operation, are possible in other freight fields. The NSW and Victorian grain freight transportation networks are undergoing rationalization. These commodities have undergone considerable fluctuation in supply due to drought conditions which have prevailed for the last 10 years. This has led to uncertain payloads and train operators unwilling to commit to unfinancial rail lines. In the case of the Victorian country broad gauge network, the private owner neglected to maintain many lines with the result that restricted speeds of less than 50km/hr prevail on much of the network. The rationalisation strategy has included discussion of closing unproductive lines, upgrading degraded track on essential lines, rationalising silo storages and utilising new bimodal technologies with trailers shipped by rail then coupled to trucks for cross country transhipment to another rail line to meet another rail service. The benefits are improved access to markets and transport hub gateways like seaports without the expense of running long block trains which are poorly utilised. This strategy has its own precision-impedance requirements which can be assessed with the transhipment calculus presented in this thesis. The Melbourne rail passenger network has undergone considerable decentralisation of train stabling-sidings yards since the late 1980’s from its central location at Jolimont. This has increased the availability of trains to service morning and evening peak demand. Currently Page | 306 Chapter 8: Conclusions modelling by the Department of Transport does not consider sidings capability directly in service availability. The tool developed in this thesis can consider possible sidings configurations that improve the storage and availability of train types. 8.4 Final Conclusions The performance protocol presented in this thesis is titled Intermodal Balance. It provides an indicator hierarchy linked to Figures of Merit which allow terminal and system specifications to be tested. The mechanisms of Hinterland Absorptive Capability present the means to better equate the load system with network capacity. The working tool of intermodal balance is transhipment calculus. The motivation of this thesis lies in elucidating mechanisms which may controlling the logistical structure which otherwise drives the inexorable growth in freight intensity. Reducing the evolution of system deficiencies and large infrastructure investment means retrofitting modal networks to interface and , through consolidation and transhipment activity, create novel, more consolidated diffuse- discrete flow patterns. Intermodal networks offer significant reductions in freight intensity, a measure of the materialising pressures of freight task activity. In this sense the drivers which feed trends and logistical friction can be controlled. Terminal impedances should now be characterised and designed for improved accessibility rather than speed mobility. These impedance measurements give a precision and controllability specification which allows an assessment of terminal capability to coordinate novel carrier flow forms. Overall benefit of new model mechanism: 1) Provides a ready reckoner for assessing terminal implications with changing rail operating forms; 2) Allows the planner to assess throughput capability opportunities and requirements by attributing alternate functions and configurations to the terminal. There is much further work to be done on trialling the applicability of electrical circuit theory and power distribution theory to the landside container freight task. The objective of the analogue may be limited in supplanting network models and may have deficiencies in acting as a comprehensive mathematical platform for an activity approach. Indeed, it was not the purpose Page | 307 Chapter 8: Conclusions of this thesis to claim electrical circuit theory could be readily grafted onto intermodal issues or the transhipment problem of Operations Research. The value and validity of the analogue has been demonstrated to a limited degree in this thesis. It potentially offers a rich definition and detailed accounting approach to terminal impedance which has an explicit design objective. It offers insights into a number of figures of merit such as precision and controllability, complex bundling and throughput, marginal costs and throughput which assist in characterising opportunities for urban intermodal activity for the new dynamics of tense flux. Governance initiatives to foster system openness and operating business rules were also incorporated into the model formulation. The means to address persistent impacts of the container freight task and to improve system capability requires appropriate economic-engineering mechanisms to support decision making. The transhipment calculus further equips the planner with toolbox instruments to test alternative urban intermodal solutions. The planner may then address emerging supply constraints in the container freight task on the basis of combined transport solutions. The pursuit of the problems besetting urban intermodal forms using the electrical circuit analogue remains worthwhile. Page | 308 Bibliography BIBLIOGRAPHY ABS 2003. Freight Movements by Statistical Subdivision – Companion Data. Canberra: Australian Bureau of Statistics. ADRIANNSE, A., BRINGEZU, S., HAMMOND, A., MORIGUCHI, Y. & RODENBURG, E. 1997. Resource flows—The material basis of industrial economies. Washington D.C.: World Resource Institute. ALDRICH, J. F., HAPP, H. H. & LEUER, J. F. 1971. Multi-Area Dispatch. PICA Conference. Boston, Mass. ALLENBY, B. 2006. The Ontologies of Industrial Ecology? Progress in Industrial Ecology- An International Journal, 3, 28-40. ÅNSTRÖM 1997. Computer Controlled Systems: Theory and Design, Upper Saddle River, New Jersey, Prentice Hall. ARTC 2002. Access Undertaking. In: AUSTRALIAN RAIL TRACK CORPORATION LTD (ARTC) FOR AUSTRALIAN COMPETITION AND CONSUMER COMMISSION (ACCC) (ed.). AUSCID 2002. Transport infrastructure: a perspective and prospective analysis of its role in Australia’s economic growth. A report for the Australian Council for Infrastructure Development Limited (AusCID) and the Association of Australian Ports and Marine Authorities (AAMPA) Prepared by National Economics. AUSUBEL, J. H. & HERMAN, R. 1988. Cities and Their Vital Systems. Infrastructure: Past, Present and Future, Washington, D.C., National Academy of Engineering. AZAR C. & SCHNEIDER, S. H. 2002. Are the Economic Costs of Stabilising the Atmosphere Prohibitive? Ecological Economics, 42, 73-80. BAINBRIDGE, E. S., MCNAMEE, J. M. & ROBINSON, D. J. 1966. Hydrothermal Dispatch with Pumped Storage. IEEE Transactions on Power Apparatus and Systems, 85, 472- 485. BALLIS, A. & GOLIAS, J. 2002. Comparative Evaluation of Existing and Innovative Rail- Road Freight Transport Terminals. Transportation Research A, 36, 593-611. BALLIS, A. & GOLIAS, J. 2004. Towards the improvement of a combined transport chain performance. European Journal of Operational Research, 152, 420-436. BATES, M. 2005. Can we contain waste? Now might be a good time to look for multi-modal solutions. Waste Management World, 21-29. BEAVIS, P., BARTOLI, S., MICHETTI, F., PORTA, F. & MOORE, S. 2006a. Intermodal Transportation Design for Waste Recovery. 28th Conference of Australian Institutes of Transport Research (CAITR). The University of New South Wales (UNSW). BEAVIS, P., BLACK, J. A., LENNOX, J., TURNER, G. & MOORE, S. 2009. Industrial Ecology Futures Scenarios: The Design Approach in Transportation. In: DAVIDSDOTTIR, R. A. (ed.) Dynamics of Industrial Eco-Systems. Edward Elgar. BEAVIS, P., BLACK, J. A., MACGILL, I. F. & MOORE, S. 2007. Distributed Functional Hinterland: Functional Design of Container Inter-modal Terminal Systems for Sydney. In: WCTRS (ed.) 11th World Conference on Transport Research (WCTR). University of California, Berkeley. BEAVIS, P., BLAKELY, E. & BLACK, J. A. 2006b. Critical Transportation Infrastructure in a Global Warming Future: Protecting NSW Seaports and their Hinterland. Sydney: Botany Bay Studies Unit, University of New South Wales, and Planning Research Centre, University of Sydney. BEAVIS, P., J.A.BLACK & GOLZAR, R. 2005. Functional Specification of Strategic Urban Freight Models: Modeling Attributes for the Port and Landside Freight Task in Sydney. Journal of The East Asia Society of Transportation Studies (EASTS) 6, 16. Page | 309 Bibliography BESTUFS 2003. Best Urban Freight Solutions: Final Clustering Report. Advanced Railway Research Centre. BEUTHE, M. E. & KREUTZBERGER, I. E. 2001. Consolidation and Transhipment. In: BREWER, BUTTON & HENSCHER (eds.) Handbook of Logistics and Supply Chain Management. Elsevier. BLACK, J. A. Year. Micro- and Macro-planning Techniques and the Study of Urban Goods Movement. In: FISHER, G. P., ed. Engineering Foundation Conference, 4-9 December 1977 1977 Sea Island, Georgia. Washington, D.C.: Department of Transportation, 413- 437. BLACK, J. A. 1981. Urban Transport Planning, London, Croom Helm. BLACK, J. A., PAEZ, A. & SUTHANAYA, P. A. 2002. Sustainable Urban Transportation: Performance Indicators and Some Analytical Approaches. Journal of Urban Planning and Development, 184-209. BLUMENFELD, D. E., BURNS, L. D. & DAGANZO, C. 1985. Analyzing Trade-off Between Transportation, Inventory and Production Costs on Freight Networks. Transportation Research B, 19B, 361-380. BLUMENFELD, D. E., BURNS, L. D. & DAGANZO, C. 1991. Synchronisating Production and Transportation Schedules. Transportation Research B, 25B, 23-37. BLUNDEN, W. R. 1971. The Land Use/ Transport System, Sydney, Pergamon Press. BLUNDEN, W. R. & BLACK.J.A. 1984. The Land Use/ Transport System, Sydney, Pergamon Press. BOBROW, L. S. 1987. Elementary Linear Circuit Analysis, Saunders College Publishing (Harcourt Brace Jovanovich). BOERKAMPS, J. & VAN BINSBERGEN, A. 1999. Goodtrip - A New Approach for Modelling and Evaluating Urban Goods Distribution. In: TANIGUCHI, E. & THOMPSON, R. (eds.) City Logistics 1. Cairns, Australia: Institute for City Logistics. BONTEKONING, Y. M., MACHARIS, C. & TRIP, J. J. 2004. Is a new applied transportation research field emerging? - A review of intermodal rail- truck freight transport literature. Transportation Research A, 2004, 1-34. BOSTEL, N. & DEJAX, P. 1998. Models and Algorthims for Container Allocation Problems on Trains in a Rapid Transhipment Shunting Yard. Transportation Science, 34, 370-379. BROOKER, T. D. 2005. RE: Container Freight Movements in Sydney. Type to BEAVIS, P. BRUNNER, P. H. & RECHBERGER, H. 2004. Practical Handbook of Material Flow Analysis. Advanced Methods in Resource and Waste Management, Boca Raton, Lewis CRC. BTRE 2003. Urban Pollutant Emissions from Motor Vehicles: Australian Trends to 2020. Canberra: Australian Government Department of Transport and Regional Services. Bureau of Transport and Regional Economics. BTRE 2004. Modelling Responses of Urban Freight Patterns to Greenhouse Gas Abatement Scenarios. Canberra: Bureau of Transport and Regional Economics (BTRE). BUKOLD, S. 1996. Kombinerter Verkehr Schiene/ Strasse in Europa: Ein vergleichende Studie zur Transformation von Guetertransportsystemen, Frankfurt am Main, Peter Lang. CAMBRIDGE SYSTEMATICS 1996. Quick Response Freight Manual. Federal Highway Administration, Office of Planning and Environment. CARIS, A., MACHARIS, C. & JANSSENS, G. K. 2008. Planning Problems in Intermodal Freight Transport: Accomplishments and Prospects. Transportation Planning and Technology 31, 277-302. CARTWRIGHT, K., L. , COTTRILL, H., HAMICK, M. & LEUE, L. N. 2003. Ports of Long Beach and Los Angeles Transportation Study. Transportation Research Record 26-35. CASSANDRAS, C. G. & LAFORTUNE, S. 1999. Introduction to Discrete Event Systems, Boston, Kluwer. CASTRO, J., KUSE, H. & KUBO, M. Year. Study on location planning of distribution facilities. In: THOMPSON, T. A., ed. The First International Conference on City Logistics, 1999 Cairns, Australia. Institute for Systems Science. Page | 310 Bibliography COHEN, A. & SHERKAT, V. 1987. Optimisation based methods for operations scheduling. Proceedings of the IEEE, 75. COLLEX. 2006. http://www.collex.com.au/services/waste/waste-management_home.cfm [Online]. [Accessed November 2006]. CRAINIC, T. G. 1987. Rail Tactical Planning: Issues, Models and Tools. In: BELLA, L. B. A. A. L. (ed.) Freight Transport Planning and Logistics. 317 ed. Berlin: Springer Verlag. CRAINIC, T. G. 2000. Service Network Design in Freight Transportation. European Journal of Operational Research, 122, 272-288. CRAINIC, T. G. & KIM, K. H. 2005. Intermodal Transportation. University of Quebec, Pusan National University. CRAINIC, T. G. & LAPORTE, G. 1997. Planning models for freight transportation. European Journal of Operational Research, 97, 409-438. CRAINIC, T. G. A. J.-M. R. 1986. Multi-Commodity, Multimode Freight Transportation: A General Modeling and Algorithmic Framework for the Service Network Design Problem. Transportation Research B, 20B, 225-242. CURRY, R. 2006. Intermodal Transport of Wastes and Recyclables in England and Wales- the "STRAW" Project. 3rd International Conference of the International Society for Industrial Ecology. Royal Institute of Technology, Stockholm, Sweden. D'ESTE, G. 1996. An Event based approach to modelling intermodal freight systems. International Journal of Physical Distribution and Logistics Management, 26, 4-15. D'ESTE, G. 2001. Freight Models. In: BREWER, B. A. H. (ed.) Handbook of Logistics and Supply Chain Management. London: Elsevier Science. DAGANZO, C. 1990. The Productivity of Multipurpose Seaport Terminals. Transportation Science, 24, 205-270. DAGANZO, C. 1995. Logistics Systems Analysis, Berkeley, Institute of Transportation Studies, University of California, Berkeley. DAGANZO, C. 2003. A Theory of Supply Chains, Berlin, Springer. DANTZIG, G. 1959. Linear Programming and extensions, Santa Monica, Rand Corporation. DE CASTILHO, B. & DAGANZO, C. 1993. Handling Strategies for Import Containers at Marine Terminals. Transportation Research B, 27B, 151-166. DE LA BARRA, T. 1989. Integrated Land Use and Transport Modeling: Decision Chains and Hierarchies. , New York, Cambridge University Press. DE NEUFVILLE, R. 1976. Airport Systems Planning: A Critical Look at the Method and Experience, Cambridge, Mass., MIT Press. DETR 1999. Sustainable Distribution: A Strategy. London: Department of Environment Transport and Regions. DOI 2003. Greens Road Intermodal Freight Terminal - Feasibility Study. Melbourne: Report by Sinclair Knight Merz (SKM) Consulting to the Victorian Department of Infrastructure (now Department of Transport) DOI 2006. Improving Rail Mode Share at the Port of Melbourne. Melbourne: Report by Booz Allen Hamilton to Victorian Department of Infrastructure. DORSEY, J. 2002. Continuous and Discrete Control Systems: Modeling, Identification, Design and Implementation, Boston, McGraw-Hill. DOT 2007. Melbourne Urban Rail Corridor Strategy. Victorian Department of Transport. DOT 2008a. Freight Futures: Victorian Freight Network Strategy. Melbourne: Victorian Department of Transport. DOT 2008b. Melbourne Intermodal System Study Report Melbourne: Report by Strategic Design and Development for Victorian Department of Transport DOT 2008c. Modelling Results from Strategic Freight Model. Melbourne: Department of Melbourne. DOT 2009. Altona Intermodal Access Enhancement Business Case. Melbourne: Victorian Department of Transport. Page | 311 Bibliography DOTARS 2002. Freight Logistics in Australia: An Agenda for Action. Canberra: Report prepared for Department of Transport and Regional Services by Industry Steering Committee of Freight Transport Logistics Action Agenda. DOUST, K. & BLACK, J. A. 2009. A Holistic Assessment Framework for Urban Development and Transportation with Innovative Triple Bottom Line Sustainability Metrics. MIT Journal of Planning, 9, 10-27. DUERR, E. & GIANNOPOULOS, G. A. 2003. SITS: a system for uniform intermodal freight transport information exchange. International Journal of Transport Management, 1, 175-186. EEA 2006. EC Framework Directive on Waste. In: HTTP://GLOSSARY.EEA.EUROPA.EU/EEAGLOSSARY/P/PRINCIPLE_OF_PROX IMITY (ed.). EHRENFELD 2004. Industrial Ecology: A new field or only a metaphor? Journal of Cleaner Production, 12, 825-831. EVERET, J. W. & APPELGATE, D. 1994. Solid Waste Transfer Station Design Journal of Environmental Engineering, 121, 96-106. FARAROUI, F. 1988. Priority Berthing in Congested Ports: The Application of Multiattribute Decision Making Methods. University of Melbourne. FERRERA, L. & SIGUT 1993. Measuring Performance of Intermodal Freight Terminals. Transportation Planning and Technology, 17, 268-279. FIAB 2005. Railing Port Botany's Containers: Proposals to Ease Pressure on Sydney's Roads. Sydney: Freight Infrastructure Advisory Board, NSW Department of Premier and Cabinet. FISCHER-KOWALSKI, M. & C.AMANN 2001. Beyond IPAT and Kurznet Curves: Globalisation as a Vital Factor in Analysing the Environmental Impact of Socio- Economic Metabolism. Population and Environment 23, 7-41. FORAN, B. & POLDY, F. 2002. Future Dilemmas: Options to 2050 for Australia’s Population, Technology, Resources, and Environment. . CSRIO Sustainable Eco-systems. FRIESZ, T. 1985. Transportation Network Equilibrium Design and Aggregation: Key Developments and Research Opportunities. Transportation Research A, 19A, 413-427. GAO 2007. Intermodal Transportation: DoT Could Take Further Action to Address Intermodal Barriers. Washington: United States Government Accountability Office. GARRIDO, R. & MAHMASSANI, H. 1998. Forecasting Short-term Freight Transportation Demand. Transport Research Record, 1685, 8-16. GEIDL, M. & ANDERSSON, G. 2004. A Modelling and Optimisation Approach for Multiple Energy Carrier Power Flow. Zurich: ETH Swiss Federal Institute of Technology. GEIDL, M. & ANDERSSON, G. 2005a. A Modelling and Optimisation Approach for Multiple Energy Carrier Power Flow. Zurich: ETH Swiss Federal Institute of Technology. GEIDL, M. & ANDERSSON, G. 2005b. Operational and Topological Optimisation of Multi- Carrier Energy Systems. Zurich: ETH Swiss Federal Institute of Technology. GIANNOPOULOS, G. A. 2002. Integrating Freight Transportation with Intelligent Transportation Systems: Some European Issues and Priorities. Transportation Research Record, 29-34. GIFFORD, J. & GARRISON, W. A. 1993. Airports and the Air Transportation System: Functional Refinements and Functional Discovery. Technological Forecasting and Social Change, 43, 103-123. GROOTHEDDE, B., RUIJGROK, C. & TAVASSZY, L. 2005. Towards collaborative, intermodal hub networks. A case study in the fast moving consumer goods market. Transportation Research E, 41, 567-583. GUDMUNDSSON, H. 2004. Sustainable Transport and Performance Indicators. In: HARRISON, R. M. & HESTER, R. E. (eds.) Transport and the Environment. Cambridge: Royal Society of Chemistry. GÜNTHER, H.-O. & KIM, K. H. 2005. Logistics control issues fo container terminals and automated transportation systems, Berlin, Springer Verlag. Page | 312 Bibliography HÄGERSTRAND, T. 1987. Human Interaction and Spatial Mobility: Retrospect and Prospect. In: NIJKAMP, P. & REICHMAN, S. (eds.) Transportation Planning in a Changing World. Aldershot: Gower in association with European Science Foundation. HANH, L., GRIFFIN 2003. The Logistics of Empty Containers in Southern California Region. METRANS Research Centre http://www.metrans.org. HANH, L., GRIFFIN. Year. Managing Empty Container Flows through Short Sea Shipping and Regional Port Systems. In: WCTRS, ed. World Conference on Transportation Research (WCTR), 2007 Berkeley, California. HART, W. 1985. The Airport Passenger Terminal, New York, Wiley. HASSALL, K. 2008. Freight Exposure and How to Measure it. Melbourne: University of Melbourne. HESSE, M. & RODRIGUE, J.-P. 2004. The Transport Geography of Logistics and Freight Distribution. Journal of Transport Geography. HOEJER, M. & MATTSSON, L.-G. 2000. Determinism and backcasting in future studies. Futures, 32, pp.613-634. HÖLTGEN, D. 1995. Terminals, Intermodal Logistics Centres and European Infrastructure Policy. European Centre for Infrastructure Studies (ECIS). HOWARD, S. G. 1983. Large or Small Terminals in Intermodal Transport: What is the Optimum Size? Transportation Research Record, 903, 14-20. HULTEN, L. 1993. Logistics Models for Circulation and Utilisation of Cargo Carrying Equipment. Licentiate, Chalmers University of Technology. JAHRE, M. 1995. Logistics Systems for Recycling- Efficient Collection of Household Waste. Doctorate, Chalmers University of Technology. JANSEN, K. 2001. Network Modelling of Port Terminals. Doctorate, Chalmers University of Technology. JATMIKA, H. 2004. Economic, Social and Spatial Impacts of Major Commercial Airports and Their Management; A Case Study of Sydney (Kingsford Smith) Airport. The University of New South Wales. JONSSON, D. 2004. Combining Urban Infrastructure of Movement: A Vision of Sustainability. In: HANLEY, R. (ed.) Moving People Goods and Information in the 21st Century: The Cutting Edge Infrastructures of Networked Cities. New York: Routledge. JOURQUIN, B., BEUTHE, M. & DEMILIE, C. L. 1999. Freight Bundling Network Models: Methodology and Application. Transportation Planning and Technology, 23, 157-177. KALAITZIDAKIS, P. & KALYVITIS, S. 2002. On the Macroeconomic Implications of Maintenance in Public Capital. Journal of Public Economics, 88, 695-712. KANAFANI, A., SPERLING, D. 1982. National Transportation Planning, The Hague, Martinus Nijhoff. KIA, M., SHAYAN, E. & GHOTB, F. 2002. Investigation of port capacity under a new approach by computer simulation. Computers and Industrial Engineering, 42, 533-540. KLEINROCK, L. 1974. Queueing Systems Volume 1: Theory, New York, John Wiley and Sons. KNOFLACHER, H. 1996. Zur Harmonie von Stadt und Verkehr: Freiheit vom Zwang zum Autofahren, Vienna, Boehlau. KOEPPEL, G., GEIDL, M. & ANDERSSON, G. 2004. Value of Storage Devices in Congestion Constrained Distribution Networks. International Conference on Power System Technology. Singapore. KOHLER, U. Year. City Logistics in Kassel. In: THOMPSON, T. A., ed. The First International Conference on City Logistics, 1999 Cairns, Australia. Institute for Systems Science. KORPAAS, M., GREINER, C. J. & HOLEN, A. T. 2005. A Logistic Model for Assessment of Wind Power Combined with Electrolytic Hydrogen Production in Weak Grids’. 15th PSCC. Liege. KOZAN, E. 2006. Optimum Capacity for Intermodal Container Terminals. Transportation Planning and Technology, 29, 471-482. Page | 313 Bibliography KREUTZBERGER, I. E. 1999. Innovative Networks and New Generation Terminals for Intermodal Transport: Improving the Cost-quality Ratio by Bundling Flows. Trail Research School, Delft. KREUTZBERGER, I. E. 2008. The Impacts of Innovative Technical Concepts for Load Unit Exchange on the Design of Intermodal Freight Bundling Networks. The Future of Freight Intermodal Transport. Edward Elgar. KULATHINAL, J. 1988. Transform Analysis and Electronic Networks with Applications, Columbus, Merrill Publishing. KUO, F. F. 1966. Network Analysis and Synthesis, New York, John Wiley and Sons. LANGEVIN, A. & MBARAGA, P. 1996. Continuous Approximation Models in Freight Distribution: An Overview. Transportation Research B, 30, 163-188. LENZEN, M. 1999. Total Requirements of Energy and Greenhouse Gases for Australian Transport. Transportation Research D 4, 265-290. LOUREIRO, C. F. & RALSTON, B. 1994. Investment Selection Model for Multicommodity Multimodal Transportation Networks. Transportation Research Record. MACGILL, I. F. Semester 2 2006. RE: Lecture Notes on Power Distribution Systems School of Electrical Engineering University of New South Wales. MACK, S. A. 2000. Enfield Intermodal Terminal Development: Report on Traffic and Transport Issues. Sydney: Sydney Ports Corporation. MADSEN, B. & JENSEN-BUTLER, S. Year. A Systems Approach to Modelling the Regional Economic Effects of Road Pricing. In: World Conference on Transportation Research, July 2-4 2004 Istanbul, Turkey. WCTRS, 44. MAGGI, R. 1994. Environmental Implications of Missing Transport Networks in Europe. Transportation Research A, 28, 343-350. MARLOW, P. B. & PAIXAO, A. 2004. Measuring lean ports performance. International Journal of Transport Management. MCKINNON, A. 2007. Synchronised Auditing of Truck Utilisation and Energy Efficiency: A Review of the British Government’s Transport KPI Programme. In: WCTRS (ed.) World Conference on Transport Research. Berkeley, California: WCTRS. MCKINNON, A. & WOODBURN, A. 1996. Logistical Restructuring and Road Traffic Growth. Transportation, 23, 141-161. MCNALLY, M. 2000. The Activity Based Approach. In: HENSCHER D.A. , K. J. B. (ed.) Handbook of Transport Modelling. Oxford: Elsevier Science. MEES, P. 2000. A Very Public Solution: Transport in the Dispersed City, Melbourne, Melbourne University Press. MEYBURG, A. & LAVERY, L. 1974. Toward a Unified Distribution System- Warehouse Location Theory. Transportation Research Record. MICHETTI, F. & PORTA, F. 2007. Modelling and Simulation of a Waste Intermodal Terminal: A Case Study in Sydney Australia. Politecnico Di Milano. MISCHKE, C. 1980. Mathematical Model Building: An Introduction to Engineering, Iowa State University Press. MORLOK, E. 1978a. Introduction to Transportation Engineering and Planning, Tokyo, McGraw- Hill Kogakusha. MORLOK, E. & RIDDLE, S. 1999. Estimating the Capacity of Freight Transportation Systems: A Model and Its Applications in Transport Planning and Logistics. Transportation Research Record, 1-8. MORLOK, E. K. 1978b. Introduction to Transportation Engineering and Planning, Tokyo, McGraw-Hill. MORLOK, E. K. & CHANG, D. J. 2004. Measuring Capacity Flexibility of a Transportation System. Transportation Research A. MUSKIN, J. B. 1983. Intermodal Freight Terminal- An Open System: The Infrastructure Perspective. Transportation Research Record, 21-26. NICOLIS, G. & PRIGOGINE, I. 1977. Self-organisation in non-equilibrium systems: from dissipative structure to order through fluctuations, New York, Wiley. Page | 314 Bibliography NIJKAMP, P., REGGIANI, A. & TSANG, W. F. 2004. Comparative Modelling of Interregional Transport Flows: Applications to Multimodal European Freight Transport. European Journal of Operational Research, 155, 584-602. NIJKAMP, P. & REICHMAN, S. 1987. Mobility and Transportation Planning Revisited. In: NIJKAMP, P. & REICHMAN, S. (eds.) Transportation Planning in a Changing World. Aldershot: Gower in Association with the European Science Foundation. NIJKAMP, P., VLEUGEL, MAGGI, R. & MASSER, I. 1994. Missing Transport Networks in Europe, Aldershot, U.K., Avebury. NILSSON, J. W., RIEDEL, S.A. 2005. Electric Circuits, Upper Saddle River, N.J., Prentice Hall. NILSSON, O. & SJELVEGREN, D. 1997. Variable Splitting Applied to Modelling of Start-Up Costs in Short Term Hydro Generation Scheduling. IEEE Transactions on Power Systems, 12, 770-774. NOTTEBOOM, T. 2008. Bundling of Freight Flows and Hinterland Network Developments. The Future of Freight Intermodal Transport. Edward Elgar. O'KELLY, M. E. 1998. A Geographer's Analysis of Hub-and-Spoke Networks. Journal of Transport Geography, 6, 171-186. O'KELLY, M. E. & MILLER, H. J. 1994. The Hub Network Design Problem: A Review and Synthesis. Journal of Transport Geography, 2, 31-40. OGDEN, K. 1992. Urban Goods Movement: A Guide to Policy and Planning. OPPENHEIM, N. 1993. A combined equilibrium model of urban personal travel and goods movement. Transportation Science, 27, 161-173. OUTHRED, H. 2002. Fundamentals of Electricity Re-Structuring. Australia's Re-Structuerd Electricity Industry. Lecture Notes. Australian CRC for Renewable Energy LTD and University of New South Wales. OUTHRED, H., WATT, M., MACGILL, I. F. & NULLES, K. 2002. Submission on IPART Discussion Paper on Distributed Generation. Sydney: Australian CRC for Renewable Energy and University of New South Wales. PARSONS BRINCKERHOFF 2001. Transportation and Land Use Model Integration Program: Overview of the First Generation Models. Portland, Oregon: Oregon Department of Transport. PASTOWSKI, A. 1998. Decoupling Economic Development and Freight for Reducing its Negative Impacts. Wuppertal Instituet fuer Klima, Umwelt und Energie. PETERSEN, E. R. 1977a. Railyard Modeling: Part 1, Prediction of Put-Through Time. Transportation Science, 11, 37-49. PETERSEN, E. R. 1977b. Railyard Modeling: Part II: The Effect of Yard Facilities on Congestion. Transportation Science, 11, 50-73. PICTON, T. & DANIELS, P. L. 1999. Ecological restructuring for sustainable development: evidence from the Australian economy. Ecological Economics, 29, 405-425. PIDD, M. 1997. Computer Simulation in Management Science, Chichester, John Wiley and Sons. POLYA, G. 1954. Induction and Analogy in Mathematics, Princeton, New Jersey, Princeton University Press. POM 2006. Port Development Plan 2006-2035 Consultation Draft. Melbourne: Port of Melbourne Authority. POM 2008. Port Phillip Bay Channel Deepening: Supplementary Environmental Effects Statement. Melbourne: Port of Melbourne Corporation. PONTON, A. May-August 2006. RE: Discussion at Victorian Department of Transport on Melbourne intermodal terminals: Dynon, Austrack, CRT. Type to BEAVIS, P. POTTER, S. & SKINNER, M. J. 2000. On Transport Integration: A Contribution to Better Understanding. Futures, 32, 275-287. POWELL, W. B. & HUMBLET, P. 1986. The Bulk Service Queue with a General Control Strategy: Theoretical Analysis and a New Computational Procedure. Operations Research, 34, 267-275. Page | 315 Bibliography PRIEMUS, H. & KONINGS, R. 2001. Dynamics and Spatial Patterns of Intermodal Freight Transport Networks. In: BREWER, B. A. H. (ed.) Handbook of Logistics and Supply Chain Management. London: Elsevier Science. RACUNICA, I. & WYNTER, L. 2000. Optimal Location of Intermodal Freight Hubs. Rocquencourt, France: INRIA. Institut National De Recherche en Informatique et en Automatique. Adopt Project. REAGAN & GOLOB 2005. Trucking Industry Perceptions of Congestion Problems and Potential Solutions in Maritime Intermodal Operations in California. Transportation Research A: Policy and Planning, 34, 587-605. RIMMER, P., J. & BLACK, J. A. 1982. Land-use transport changes and global restructuring in Sydney since the 1970's: the container issue. In: RICHARD V. CARDEW, J. V. L., DAVID C. RICH (ed.) Why Cities Change: Urban Development and Economic Change in Sydney. Sydney: George Allen and Unwin. RIMMER, P., J. & TSIPOURAS, A. 1978. Ports and Urban Systems: Framework and Research Needs in Resolution of Port Generated Conflicts. Australian Transport Research Forum. RIMMER, P. J. 1978. Urban Goods Movement in Sydney. Canberra: Bureau of Transport Economics. RIMMER, P. J. & BLACK, J. A. 1981. Urban Goods and Commercial Vehicle Movements in Sydney: A Research Framework. Australian Road Research 11, 15-29. RIMMER, P. J. & HICKS, S. K. 1978. Urban Goods Movement: Process, Planning Approach and Policy. In: HENSCHER, D. A. & STOPHER, P. R. (eds.) Behavioural Travel Modelling. London: Croom Helm. ROBERTS, P. & KULLMAN, B. 1979. Urban Goods Movement: Behavioural Demand- Forecasting Procedures. In: STOPHER, H. A. (ed.) Behavioural Travel Modelling. London: Croom Helm. ROCKCLIFFE, N., WIGAN, M. & QUINLAN, H. 1999. Developing Database of Nationwide Freight Flows for Australia. Transportation Research Record, 1625, 147-155. RODRIGUE, J.-P. 1999. Globalization and the synchronization of transport terminals. Journal of Transport Geography, 7, 255-261. RODRIGUE, J.-P. 2004. Freight, Gateways and Mega-Urban Regions: The Logistical Integration of the BOSTWASH Corridor. Tijschrift voor Economische en Sociale Geografie, 95, 147-161. RODRIGUE, J.-P. 2005. Transportation and Global Production Networks: The Challenges of Geographical and Functional Integration. New York: Department of Economics and Geography, Hofstra University RODRIGUE, J.-P., SLACK, B. & COMTOIS, C. 2001. Green Logistics. In: BREWER, B. A. H. (ed.) Handbook of Logistics and Supply Chain Management. London: Elsevier Science. ROSEBROCK, M. 1992. Automatisierung und Dezentralisierung des Güterverkehrs der Bahn, Frankfurt am Main, Peter Lang. ROSO, V., WOXENIUS, J. & LUMSDEN, K. 2009. The Dry Port Concept- Connectiing Container Seaports with the Hinterland. Journal of Transport Geography, 17, 338-345. ROSON, R. & SORIANI, S. 2000. Intermodality and the Changing Role of Nodes in Transport Networks. Fondazione Eni Enrico Mattei. ROTH, A. 2000. Assessing long-term environmental goals in goods transportation. Doctorate, Chalmers University of Technology. ROTMANS, J., KEMP, R., VAN ASSELT, M., GEELS, F., VERBONG, G., MOLENDIJK, K. & VAN NOTTEN, P. 2001. Transition and Transition Mangement: The case for a low emission energy supply. Maastricht: International Centre for Integrative Studies. RUSSO, F. & CATENI, A. Year. An Integrated System of Models to Plan City Logistics. In: World Conference on Transportation Research, 4-8th July, 2004 2004 Istanbul, Turkey. 17. Page | 316 Bibliography RUSSO, F. & COMI, A. 2004. Acquisition Models for the Analysis of Freight Transportation in Urban Areas. World Conference on Transportation Research. Istanbul, Turkey: WCTRS. SCHLIEPHAKE, K. 2000. Verkehrsdrehscheibe Deutschland- Mobilitätszuwachs und Infrastrukturausbau im vereinigten Staat. Petermanns Geographische Mitteilungen, 144, 58-69. SCHMEDDLING, D., TIETZE-STÖCKINGER, I., LIEDTKE, G. & OTT, A. 2004. BundlePoints - An option for reducing economic and ecological loads in trade areas. 10th World Conference on Transport Research (WCTR). Istanbul, Turkey: WCTRS. SFCNSW 2004a. Freight Supply Chain - Coordination of Working Arrangements (Mismatch of Hours). Sydney: Report by Meyrick and Associates to Sea Freight Council of New South Wales. SFCNSW 2004b. New South Wales Import Export Container Mapping Study. Sydney: Report to Sea Freight Council of NSW by Jays Corporate Services. SFCNSW 2004c. Regional Intermodal Terminals- Indicators for Sustainability. Syndey: Report for the Sea Freight Council of NSW by Strategic Design and Development (Sd&D). SFCNSW 2005. NSW Landside Infrastructure Capability: International Containers. Sydney: Sea Freight Council of NSW. SFCNSW 2007. Sydney's Intermodal System: Development of a business plan for an open access intermodal terminal in Sydney. Sydney: Sea Freight Council of New South Wales. SHAW, J. J. 1995. A Direct Method for Security- Constraiined Unit Committment. IEEE Transactions on Power Systems, 10, 1329-1342. SJÖSTEDT, L., J. WOXENIUS & HULTEN, L. 1994. Flexibility Versus Specialisation – On the Controllability of Combined Transport Systems. The 7th IFAC Symposium of Transport Systems. Tianjin, China. SJÖSTEDT, L., J. WOXENIUS, AND L. HULTEN 1994. Flexibility Versus Specialisation – On the Controllability of Combined Transport Systems. The 7th IFAC Symposium of Transport Systems. Tianjin, China. SKM 2005. Intermodal Logistics Centre at Enfield Environmental Assessment. Sydney: Report to Sydney Ports Corporation by Sinclair Knight Merz SLACK, B. 1999. Satellite Terminals: a local solution to hub congestion? Journal of Transport Geography, 7, 241-246. SLACK, B. 2001. Intermodal Transportation. In: BREWER, B., HENSCHER (ed.) Handbook of Logistics and Supply Chain Management. Oxford: Elsevier Science. SMILOWITZ, K. R., ATAMTÜRK, A. & DAGANZO, C. F. 2003. Deferred Item and Vehicle Routing within Integrated Networks. Transportation Research E, 39, 305-323. SOUTHWORTH, F. & PETERSON, B. 2000. Intermodal and International Freight Network Modeling. Transportation Research C, 8, 147-166. SPC 2000. Metropolitan Sydney International Container Origin/ Destination Analysis. In: CORPORATION, S. P. (ed.). Sydney. SPC 2002. Port Botany Expansion: Environmental Impact Statement. Sydney: Report by URS Consulting for Sydney Ports Corporation. SPC 2005. Enfield Intermodal Logistics Centre. Environmental Impact Statement. Sydney: Report by Sinclair Knight Merz (SKM Consulting) for Sydney Ports Corporation (SPC). STARKIE, D. 1987. Configuring Change: Reflections on Transport Policy Processes. In: NIJKAMP, P. & REICHMAN, S. (eds.). Aldershot: Gower in association with European Science Foundation. STYHRE, L. 2005. Toward Improved Port Performance In Intermodal Transportation. Licentiate, Chalmers University of Technology. SULOGTRA 2002. SULOGTRA -Effects on Transport of Trends in Logistics and Supply Chain Management. Berlin: Technical University of Berlin, Logistics Department, Germany. Page | 317 Bibliography TALEB-IBRAHIMI, M., DE CASTILHO, B. & DAGANZO, C. 1993. Storage Space vs Handling Work in Container Terminals. Transportation Research B, 27B, 13-22. TANIGUCHI, E., NORITAKE, M., YAMADA, T. & IZUMITANI, T. 1999a. Optimal Size and Locatoon of Public Logsitics Terminals. Transportation Research E, 35, 207-222. TANIGUCHI, E., THOMPSON, R. & YAMADA, T. 1999b. Modelling City Logistics. In: TANIGUCHI, E. & THOMPSON, R. (eds.) First International Conference on City Logistics. Cairns, Australia: Institute for City Logistics. TAPIO, P. 2005. Towards a Theory of Decoupling: Degrees of Decoupling in the EU and the Case of Road Traffic in Finland Between 1970 and 2001. Transport Policy, 137-151. TAPIO, P. & HIETANEN, O. 2002. Epistimology and Public Policy: Using a New Typology to Analyse the Paradigm Shift in Finnish Transportation Futures Studies. Futures 34, 597- 620. TARSKI, I. 1986. The Time Factor in Transportation Processes, Amsterdam, Elsevier TAVASSZY, L., SMEENK, B. & RUIJGROK, C. 1998. A DSS For Modelling Logistic Chains in Freight Transport Policy Analysis. International Transactions in Operational Research, 5, 447-459. TCRP 1993. Institutional Barriers to Intermodal Transportation Policies and Planning in Metropolitan Areas. Washington: Transit Cooperative Research Program. TERMINET 1998. TOWARDS A NEW GENERATION OF NETWORKS AND TERMINALS FOR MULTIMODAL FREIGHT TRANSPORT. TECHNISCHE UNIVERSITEIT DELFT. THEMIS 1999. Thematic Network in Optimising the Management of Intermodal Transport Services. European Commission’s 5th Framework Programme. THOMPSON, G. 2006. Transfer Station Planning and Design 1.01. WMAA Transfer Stations Conference. Sunshine Coast, Australia, August 9-11: Waste Management Association of Australia (WMAA). THOMPSON, R. G. & VAN DUIN, R. Year. Vehicle Routing and Scheduling. In: TANIGUCHI, T. & THOMPSON, R., eds. The First International Conference on City Logistics, 1999 Cairns, Australia. Institute of Systems Research. TOMPKINS, J. A. 1996. Facilities Planning, New York, Wiley. TRANSPORT CANADA 1979. Urban Goods Movement Report Series. Montreal, Quebec: Urban Transportation Research Branch of Canadian Surface Transportation, Transport Canada. TURNQUIST, M. A. & DASKIN, M. S. 1982. Queuing Models of Classification and Connection Delay in Railyards. Transportation Science, 16, 207-230. UNCTAD 2004. Assessment of A Seaport Land Interface: An Analytical Framework. UNCTAD Secretariat. US EPA 1973. Rail Transport of Solid Wastes. Report prepared by the American Public Works Association for the United State Environmental Protection Authority Report no. SW- 22d. VAN DER VOOREN, F. W. C. J. 2004. Modelling Transport in Interaction with the Economy. Transportation Research E, 40, 417-437. VFLC 2005. Business Activity Harmonisation Study. Melbourne: Victorian Freight and Logistics Council. VFLC 2007. A Toolkit for the Development of Intermodal Hubs in Victoria. Melbourne: Victorian Freight Logistics Council (VFLC). VFLC 2008. Truck Optimisation Plan: Options Paper. Melbourne: Victorian Freight and Logistics Council. VUCHIC, V. R. 1999. Transportation for Livable Cities, New Brunswick, N.J., Centre for Urban Policy Research. WAIDRINGER, J. 2001. Complexity in Transportation and Logistics Systems. Doctorate, Chalmers University of Technology. Page | 318 Bibliography WANG, J. 2002. Dealing with Barriers in Maritime and Multimodal Services. Hong Kong: Department of Geography, University of Hong Kong. WIGAN, M. 1978. Indicators for Urban Commodity Movements. In: HENSCHER, D. A. & STOPHER, P. R. (eds.) Behavioural Travel Modelling. London: Croom Helm. WITJAYARATNA, O. 2005. RE: Discussion on freight activity in Port Botany precinct. Type to BEAVIS, P. WMAA 2005. New South Wales Compost Roadmap. Sydney: Waste Management Association of Australia. WOOD, A. J. & WOLLENBERG, B. F. 1996. Power Generation, Operation and Control, New York, John Wiley and Sons. WORRALL, R. D. & BRUGGEMAN, J. M. 1969. Analysis of the Locations and Functions of the Terminal Interface System. Report prepared by Peat Marwick, Livingston and Co. for the US Department of Transportation. WOXENIUS, J. 1998. Development of Small-Scale Intermodal Freight Transportation in a Systems Context. Chalmers University of Technology. WOXENIUS, J. 2006. Temporal Elements in Spatial Extension of Production Networks. Growth and Change, 37, 526-549. WOXENIUS, J. 2007. Intermodal freight transport network designs and their implications for transhipment technologies. European Transport - International Journal of Transport Economics, Engineering and Law. WOXENIUS, J., ANDERSSON, E., BAERTHEL, F., SOMMAR, R. & TROUVE, J. 2005. A Swedish Intermodal Transport Service Based on Line-Trains Servicing Freight Forwarders. Department of Logistics, Chalmers University of Technology. WOXENIUS, J., SJOESTEDT, L. & HELLGREN, J. 2004. Five Traffic Designs in Combined Transport Systems- A Theoretical Discussion. Logistyka- Engineering Machine Problems, 2, 111-127. ZIEMER, R. E., TRANTER, W.H., FANNIN, D.RONALD 1993. Signals and Systems: Continuous and Discrete, New York, MacMillan. ZOGRAFOS, K. & GIANNOULI, I. 2001. Emerging Trends in Logistics and Their Impact on Freight Transportation Systems. Transportation Research Record, 36-44. ZUMBAHLEN, H. 2008. Linear Circuit Design Handbook, Oxford, U.K., Newnes, Elsevier. Page | 319 Appendix A: Strengths and Weaknesses of Analogue APPENDIX A- Strengths and Weaknesses of Analogue A.1 Differences and Similarities in Representing Freight Transportation as an Electrical Analogue If the freight system is to be considered as a electrical circuit- power gateway analogue, combining both storage generation plants, and direct generation plants (road based), both with availability constraints, there remain some differences which must be resolved for the analogue to be adapted. There are seven areas of difference: 1. Intermodal operations must be considered competitively against the all road mode-it is not a free good: the intermodal gateway must be price competitive and offer timely and precision services between terminal and port origin/ destination. 2. Plant cost can be correlated with required output rates (within operating boundaries). The relationship between storage and throughput is more complex than for hydroplants. For nodes involving some transient storages, the cost needs to reflect the duration of consignments through the plant as well as the penalty for not matching the connecting service. This increases the duration of the consignment at the plant and cost might be passed on as demurrage. The duration through the plant depends on a number of parameters. These are: consignment arrival profiles (arriving as single containers by truck or as a 75 container block by train); the dwell profile of the consignment type; the consolidation/ distribution work required; the configuration of the plant and the equipment used which will affect number of lifts required and lift productivity. 3. Whilst freight is also a just-in-time operation, it is neither instantaneous nor fungible like electrical power. Elapsed time through the system is a crucial cost component and this must be recorded in any optimisation either directly as part of the production function or as a side constraint. 4. Freight terminal throughput capacity involves more than simple storage capacity. It involves both terminal configuration and train operating form. The transhipment equipment and their access to storage facilities will be different depending on the turn around time required of the train, its length and frequency and the destination of its containers. See section 4.2.4. Page | 320 Appendix A: Strengths and Weaknesses of Analogue 5. The international container system is defined by the coordination of three sets of flows: imports, exports, and empty container re-positioning. These flows mean that the plants must contend with multi-directional flows. Plants become both generating and load demanding nodes. In an open market system power nodes are becoming both generators and users of powers at different intervals. That is, they alternately sell and buy energy. For the intermodal production systems there is a need to account for the simultaneous production of containers (export destined) and the “purchase” of containers (import attracting). For modelling purposes, these distinct flows may be considered separate operations. 6. The international container system is a multi-commodity system. Hydro systems may be considered multi-commodity in the sense they may have a network of diffuse reservoir sources of flow. 7. The ownership of resources: containers, consignments, vehicles, intermodal terminals, port infrastructure, and rail paths is diverse in the urban-international container freight task, even under a centralised operations through the seaport. Often the commercial relationships between actors do not correspond with best operating practice for the system. A.2 Comparative Strengths Network models with transhipment nodes are open loop control systems whereby the control action to achieve the output from the input is calibrated. The activity based model approach utilises a closed loop system, whereby the control action is dependent on the output. Open loop systems are calibrated on time elapsed processes such as flow rates and their resulting impacts on capacity. Capacity is then constraint based and determined by time. For instance, penalty functions based on not meeting modal schedules are used in conventional network optimisation procedures. Closed loop control systems utilise the mechanism of feedback whereby the controlled variable such as the output is compared with the input to the system so that there may be an appropriate control action as some function of the output and the input. A closed sequence of cause and effect relations exists between system variables. Page | 321 Appendix A: Strengths and Weaknesses of Analogue This approach has undoubted strengths when we are considering systems where it is not easy to calibrate performance. This is due to: 1. The systems and infrastructure not currently operating 2. The dynamics considered is too complex to calibrate 3. The systems and infrastructure are retrofits over current systems In this instance, we may verify the relationship of different parameters. By utilising the frequency domain- impedance can be measured which characterises both capacity consumption and cost – due to time use (duration) and due to timing (precision). Meeting the desired output is also classified according to physical attributes of flow as well as timing information. Benefits beyond Queuing Theory: 1. Include control theory 2. Predict response from periodic sources 3. Detailed mathematical expressions for impedance 4. Consider storage inertia Table 28 compares and contrasts the electrical circuit analogue tool with other terminal and network methods prevailing in freight terminal analysis. Page | 322 Appendix A: Strengths and Weaknesses of Analogue Table 28: Positioning the Electrical Stencil Tool in Freight Terminal Analysis for Sketch Planning Criteria Electrical Stencil Network Methods Queuing Method System Dynamics Insights from Focus is the node Focus is the network of links Focus is unit process Derived from mass accounting system Related to energy balances. Shows activity between flows balance. changes in value density and and servers consignment flow. Impedance accounting allows measures of precision and stability. Limits of a lumped parameter system Application of Apply filters in the timing domain to Generally must be set apriori Can be applied as Can be applied as business rules unit processes within terminals as a constraint to objective independent variables independent variables functions Handling arrival/ Can handle all manner of flow Flow profiles are generally Can handle many flow Can handle all manner departure patterns profiles not pulsed profiles but limited in of flow profiles closed form responses Terminal Design to Time and Work Impedance basis. Time impedance based only. Design is input flow Design is input flow incorporate complex Filter and switching control Itinerary within terminal is set based based terminal activity techniques to allow for itinerary apriori. control. Design is load based (i.e. back-end) Stability Can detail transient analysis Does not detail transient Is based on steady state Uses cumulative response dynamics only unless storage relationships to Discrete event systems identify saturation. Can assess stability if time step is finely spliced. Closed form solutions Can resolve ODE of varying Limited or no dynamics Based on Markov chain Often ODEs cannot be complexity with variety of flow captured at the terminal level dynamics. More formed from the profiles. Most situations can reliably variable phenomena dynamics as tool reduce to second order not amenable to closed complex. solutions Page | 323 Appendix A: Strengths and Weaknesses of Analogue Insights into handling Requires extension to power Has well defined methods in Can excel in this if Poor unless they are multi-commodity distribution theory with power addressing the multi- extend to discrete event set as separate flow flows carrier proxy. commodity flow problem- the systems variables assignment step remains problematic and can lead to competitive allocation rather than complements and thus poor multimodal solutions Ability to assess Filter controls as well as switches Not amenable to process flexible capacity can be applied to assess changes in control of node terminals. load- amenable to process control Flexible capacity can generally y be only assessed in terms of redundancy. Page | 324 Appendix A: Strengths and Weaknesses of Analogue A.3 Comparative Weaknesses The use of circuit stencils to formulate the dynamical equations which underpin terminal impedance are unlike closed form queuing analysis where loading effects, flows requiring work on terminal resources leading to impedance, evolve endogenously from the underlying Markov chain mathematics. Loading effects can be included through use of coupled passive circuits, reflected impedance, impedance matching and careful formulation in active circuits. In order for the loading implications to be accounted for, they need to be recognised and explicitly included in the model formulation. The benefit of this is that we can develop more complex closed form solutions to transport terminal problems but the model formulation needs to be more explicitly defined. Queuing theory itself has limited closed form solutions to flow profiles, especially of a pulse nature. Event modelling better accommodates unplanned loading effects. This thesis employs continuous time control systems. These are averaged response systems. Unless loading effects are pre-defined they will not evolve from the operation of such systems. In discrete time control systems and the use of time based triggers, loading effects and delays can evolve due to asynchronous events (Pidd, 1997). The fidelity of our system approach is reduced. For sketch planning purposes this is acceptable. Detailed simulation would be advisable prior to full terminal design. A.4 Anticipating Verification Procedures Once the parameters and relationships of transient and steady state analysis can be broadly calibrated, design can be carried out according to compensation control techniques. Taleb- Ibrahimi et al., (1993) have demonstrated that the effects of business rules applied to terminals can be sketched out using cumulative arrival and departure curves. These cumulative curves can be used to verify the effects of compensation control in the steady state. Page | 325 Appendix B: Terminal Activity Mapped to Cumulative Storage Curves Appendix B- Terminal Activity Mapped to Cumulative Storage Curves B.1 Terminal Activity: Headway and Storage Curves Cumulative storage curves have been used widely to depict the operational needs of terminals according to arrival and dispatch profiles (Blumenfeld et al., 1985, 1991, Daganzo, 1995, De Castilho and Daganzo, 1993, Morlok, 1978b, Taleb-Ibrahimi et al., 1993). They can also depict the effects of business rules on the capability of the terminal to deliver on service requirements. In any model which captures the topological complexity of the terminal and relates the itinerary of flows with the service level, a mass flux balance must be assessable according to net arrival and dispatch profiles for fidelity. Changes in design and operation can also be depicted. Most commonly, cumulative flow-time charts can estimate the average waiting time and the average time in the server as well as the average number in the system (Morlok, 1978): (B.1) − 1 N t = ∑ − AD )( (B.2) N i=1 ii (B.3) When dealing with handling of import containers, De Castilho and Daganzo (1993) have suggested that under certain assumptions, the number of container moves can be predicted as a function of their accumulation: (B.4) Where S is the number of stacks for import containers A is the arriving containers per vessel headway, D is the dispatch rate of containers per vessel headway. Page | 326 Appendix B: Terminal Activity Mapped to Cumulative Storage Curves Business rules affecting storage size and handling requirements can also be depicted. B.2 Storage Curves and Multi-Commodity Flows For a desirable model specification, multi-modal, multi-commodity and multi-path coupling aspects in tactical terminal design need to be incorporated into a representable form such as the cumulative flow chart. The analysis by multi-commodity, multi-modal flows using the aggregate representation of a cumulative flow chart has been demonstrated by Taleb-Ibrahimi and Daganzo, (1993). The transhipment calculus to characterise impedance can translate multi- path coupling of alternate complex terminal topologies onto cumulative flow charts by understanding that the terminal activity cycles can be reduced to a critical flow sequence. Alternative itinerary paths can show how relationships between arrivals and dispatch sequences necessarily alter. The cumulative arrival and departures curve can be used in space management of the storage and handling buffer. Taleb- Ibrahimi et al. (1993) depict the role of static and dynamic assignment in affecting the reservation storage area required. This will be in addition to the distance between the arrival A(t) and the departure rate D(t) (Figure 71). Cumulative container count container Cumulative Figure 71: Cumulative Curves for Container Arrivals A(t), Departures D(t), Static R(t) and Dynamic R’(t) Assignment from Stack to Vessel (after Taleb- Ibrahimi, et al., Fig. 3 p. 18,1993) Page | 327 Appendix B: Terminal Activity Mapped to Cumulative Storage Curves They further extend this to the multiple input case whereby arrivals at any one time can be measured by: = − ∑ ij j ijl tcFBtA )()( ijl …………………………….(B.5) Where B j is the number of consignments for each carrier class i per train of carrier class j and cijl is the cut-off time (the time after which no more consignments will be allowed for that train service). The departure curve is depicted in summary form: −= ∑ ij stHBtD ijl )()( ijl …………………………(B.6) th Where H is the unit step function and sijl is the l departure of train carrier j of consignment carrier ij. The space, or slots reserved at any time, depends on the initial arrival time of the consignments − of that carrier pair ( ijl ac j ). The release-accumulation storage assignment curve is represented by: [ −−= ] )( ∑ j ijl actHBtR j )( ijl (B.7) And the amount of space reserved is: S = max {R(t) – D(t)} (B.8) This notation can be applied in developing a stencil response of multi-commodity flows operating through Logistics Unit Processes. The active arrival period, between train departure and cut-off time for next train departure, is proxied by the synchronisation range of a dispatching service (the bandwidth). A narrower bandwidth for the same average dwell time, corresponding to a tighter active arrival duration for the discrete train dispatch service will allow fewer slots reserved in storage to meet the utilisation requirements of that train service. Business Rules to manage the stack may achieve a dynamic assignment curve, R’(t), closer to departures. The time-averaged accumulation will be: ~ ijVB j = ∑∑ wQ j j P j J …………………………………………..(B.9) Page | 328 Appendix B: Terminal Activity Mapped to Cumulative Storage Curves Where: wj is the dwell time of consignments, Vj is the number of train services for that train class (to meet a specific vessel dispatch, for instance) and Pj is the period between visits for ~ each train service (of that class). For our purposes Q j represents the value accumulated and thus we have control over settings of both consignment flow and value density (the batching activity). ~ Where Q is much less that Q, maximum accumulation allowed, and space is underutilised. Additionally when the slots reserved are much greater that the slots utilised, space is underutilised. Space management options to reduce Q include changing departure schedules and or changing container arrival patterns. This difference may be inevitable if there are peak services. Q represents an ideal amount accumulated, separate from the space management strategy. Page | 329 Appendix C: The Laplace Transform Activity to Freight Terminal Logistics APPENDIX C: The Laplace Transform Application to Freight Terminal Logistics C.1 Introduction If it is established that the response of unit processes in a freight terminal can be represented by exponential curves, this corresponds to a sinusoidal response in the steady state. A powerful means to generate and analyse this response is application of the Laplace transform. This enables us to relate time-domain behaviour to so-called frequency-domain behaviour. For the purposes of this thesis, frequency domain behaviour is the timing behaviour we wish to characterise as it impacts on space-management capabilities and the business rules applied for the freight terminal. The integro-differential equations developed from analogue circuit stencils can be converted to Laplace transform and acts as an impedance function providing a deeper understanding of the stencil response. A significant aspect of this mathematical representation is that it allows easier mathematical transformations. In circuit stencil analysis we can use two forms of the Laplace transform: the functional transform which can be used to transform specific functions we may use for source inputs and operational transforms which we use to transform the integro-differential equations developed from the stencil. Other valuable transforms include translations in the time domain (to account for time lags); translations in the frequency domain and scale changing. From a single or set of integro-differential equations, the Laplace transform can be formed. Algebraic manipulation of this leads to a transfer function representing the impedance ratio of achieving an output response based on a source input flow. The output mathematical expression can be represented in the time domain by taking an inverse transform, often solved through partial fraction expansion of the impedance ratio. Partial fraction expansion is based on the nature of the denominator roots. The transform and inverse transform of complex roots allows inclusion of phasor notation and is helpful in applying space management techniques developed in this Chapter. Other benefits from the Laplace transform approach include: 1. Superior signal process: accommodating a wider range of source driving functions; 2. Superposition of multiple sources; 3. Convolution; 4. Inclusion of initial conditions; and Page | 330 Appendix C: The Laplace Transform Activity to Freight Terminal Logistics 5. Inclusion of a number of impedance attributes to characterise the response to various flow signals. The steady state response of an excitation source through a unit process, proxied by a circuit stencil, can be computed using a transfer function (Nilsson, p.618): Yss(t)= A|H(jω)|cos[ωt+φ+θ(ω)] That is, we combine the excitation source with the transfer function to get the response. C.2 Filters The filter mathematics is based on the stencil of Operational Amplifiers. Usually capacitive loads are used in the stencil configuration rather than inductive loads as capacitive loads are considered more stable. There are four filtering types: 1. Low pass 2. High pass 3. Bandpass 4. Bandreject Value density gain is a further filtering type. Types 3 and 4 are based on combinations of the first two types Type 3 is formed by types 1 and 2 in cascade; type 4 is formed by types 1 and 2 in parallel. These combinations only lead to broadband i.e. low quality factor forms. This is due to the existence of discrete poles in the transfer function. For narrowband filters complex poles formulation is required. Typologies can be formed on the basis of cascade, series and or parallel impedance blocks. A bandpass filter takes the following form: Page | 331 Appendix C: The Laplace Transform Activity to Freight Terminal Logistics Ksβ Hs()= ss2 ++βω2 0 (C.1) where: H(s) is the transfer function ( in the frequency domain) of impedance relating the work required for the batching multiple. K is the value density gain (based on a batching multiple of the consignment) β is the bandwidth (active arrival rate for that train service load). ω 0 is the centre of the timing band. For a simple bandpass filter, a block function can be formed from 3 components: a low bandpass, a high band pass and a gain function. This cascade bandpass can be represented by the following transfer function: ω Ksc Hs()= 2 ss22++ωωω ccc212 (C.2) where: ω is the transhipment intensity rate designated by the low pass band, calculated by the c2 parameters: 1 ω = c2 RC LL (C.3) ω is the transhipment intensity rate designated by the high pass band, calculated by the c1 parameters: 1 ω = c1 RCHH (C.4) Page | 332 Appendix C: The Laplace Transform Activity to Freight Terminal Logistics For this bandpass filter to be valid, the timing rate ratio must adhere to the following: ω c2 ≥ ω 2 c1 (C.5) A high-Q bandpass filter requires complex poles and has the following form: C2 R3 R1 C1 U1 3 + 6 OUT V1 Vi R2 2 - Vo Ideal Op Amp Figure 72: High Quality Bandpass Filter C.3 Compensation Control Compensation control offers a more tractable model formulation in addressing the particular design ambition posed by metropolitan intermodal terminals than do other facilities and network modelling methods. The design ambition is to deliver precision services whilst meeting internal performance criteria of capacity consumption and this requires a mechanism of control. From design of this mechanism and observation through figures of merit, the planner gains an understanding of the flexible capacity of the terminal under different rail operating forms. Other modelling methods that capture the unit process operations of terminals under different flow regimes: queuing models, discrete event models; continuous variable dynamics, focus on assessment of delay and/or evolving bottlenecks in the terminal and do not give direct insights into flexible capacity. Network optimisation models assume capacity at nodes as solely constraint based and do not consider the control- capacity dynamic. The model formulation proposed in this thesis uses control mechanisms to apply a richer description of impedance. Flexible capacity is defined as the combination of attributes of: terminal functions, layout configuration, physical resources, flow routing, and business rules which manage the performance of the terminal to meet alternate services according to different inflow patterns. It is a combination of constraint and response dynamics.. It will be shown that with the correct Page | 333 Appendix C: The Laplace Transform Activity to Freight Terminal Logistics variables and the appropriate impedance parameters selected, flexible capacity dynamics can be demonstrated around sparse functions. This representation and application of topological complexity allows consideration of multipath coupling and multi-commodity flows. Its role in decision support is explanatory rather than predictive. Second order circuits, those with irreducible storage devices such as RLC circuits, have some interesting properties and their s roots can be evaluated for system stability. For instance, we wish for a critically damped response as a transient response. That is, minimum oscillation around the end point. Coupled storages at the terminal can lead, however, to an underdamped response. This is due to the strength of the middle term in the denominator polynomial (Kulathinal, 1989, p.102). This is the coefficient of the first derivative in the natural response. This is the damping factor. If negative, it is considered unstable. If less than one, underdamped (decaying oscillations) and zero, the response is oscillating, marginally stable. In the electric analogue, the resistor acts as a damper. In the oscillating case we do not store efficiently or in the more direct manner, thus the storage either takes longer or we temporarily exceed our storages – our aim should be the critically damped response with the middle term close to or just over one. The analysis of the underdamped response can be made in quite a detailed fashion. This includes calculation of : • Overshoot, • Decay Ratio, • Rise Time, • Response Time, • Period of oscillation, • Natural period of oscillation. This has valuable implications for the operation of an intermodal production system. Interacting storage and other activity elements can represent instability (bottlenecks propagating downstream). Thus this can be a sign that the functional mix at a terminal is undesirable. There may be some need for compensation control here in the form of time lags between activities. The control ambition occurs when the purpose is design of LUPs and terminals rather than simply analysis. Control consists of three aspects- stability, flexible capacity and accuracy. Stability consists of a number of figures of merit: Page | 334 Appendix C: The Laplace Transform Activity to Freight Terminal Logistics 1. Time to peak 2. Rise time 3. Percent Overshoot 4. Settling Time Steady state accuracy pertains to assessing how close the system is to matching the desired output (can be considered as a reference input). Flexible capacity is the ability of the terminal to respond to changes in input and load requirements by bringing on line other resources (such as sidings, handling equipment; stack buffers).Circuit control techniques used here include: load regulation; transportation lags; lag-lead compensators. Compensators are the pole and zero factors of the impedance transfer function. It needs to be established what these compensators represent with respect to physical resources, terminal layout configuration, flow routing and business rules. C.4 Scope The ability to perform control actions enables the achievement of some flexible capacity at the terminal. In this case control is not only the response to mitigate imposed disturbance but also to implement business rules and demonstrate flexible capacity. Control techniques are the application of business rules to achieve terminal operating feasibility under conditions of varying inflows and dispatch schedules and tactical planning decisions. The implication of these business rules may extend to injection of additional capacity such as providing more stack or sidings storage or routing the carrier itinerary different according to changing conditions. In the seaport- hinterland case controls are required to: 1. Take different itineraries through the terminal under different carrier pair arrangements (degree of direct transhipment); 2. Provide for bimodal activity (consignments routed for dispatch to larger trucks when train services are not available); 3. Preferentially ship containers of different types (i.e. ship empty containers when payloads on train shuttles are low); 4. Inject additional storage and handling resources 5. Reconfigure the stack for more efficient accessibility as the flow rate or value density to be batched increases 6. Utilise non-active sidings to pre-load wagons for dispatch Page | 335 Appendix C: The Laplace Transform Activity to Freight Terminal Logistics These actions are motivated by a) the need to support the payloads of train shuttles and further b) to preserve the performance of the stack i.e. that it does not saturate. This allows additional consolidation benefits to accrue to the network. Terminal performance needs to remain satisfactory. Controls have an additional role in compensation so that as well as performing the business rule according to some trigger, they allow the system to recognise the flexible capacity of the action. Some systems will have less flexible capacity than others- for instance terminals without additional sidings will not be able to store pre-loaded consignments and there would be no feasible compensation to mediate the disjoint in pulse arrivals and dispatch. The existing storage arrangement would go into saturation leading to increasingly inaccessible storages and queues bottlenecking the system. C.5 Specifications The specifications nominated are centered on frequency domain requirements: 1. Value Density Gain (batching) 2. Quality Factor 3. Centre frequency (or end frequencies if a band) 4. Bandwidth The quality factor can be considered as the measure of controlling the consignment flux. A low quality factor implies that the consignment flux is throttled (there is storage accumulation) with a reduced flux output. The temporary accumulation corresponds to a time lag. A low quality factor can also mean unproductive storage as there is disproportionate value lost in the transformation to the payload due to access inertia, handling costs. The timing rating (centre frequency or resonant frequency) is defined for the value of the terminal components. For a low pass filter the timing ratings is w = 1/R2C. For a high Q band pass filter w = sqrt(1/ReqR3C). Bandwidth may be considered as the size of the active arrival time. The active arrival time is the duration for which a storage of a certain area must be kept aside for a particular vessel dispatch. It is part of a many to one problem-solution. Akin to the logistics situation, the duration is defined as the time distance between the first and the final freight arrival. A longer duration for the same quantity means that more storage area is required and the average time in the stack will be longer. From the characteristic equation the damping factor can be altered to affect alternate responses of changing value density magnitude, consignment flux, and carrier timing profile passed (Figure 73). Page | 336 Appendix C: The Laplace Transform Activity to Freight Terminal Logistics Figure 73: Design Modification of Bode Plot for Terminal Handling- Flux Goals (after Bobrow, Fig 10.16 p.456) Page | 337 Appendix D: Dandenong Case Study APPENDIX D: DANDENONG CASE STUDY D.1 Dandenong Configurations Frankston - Dandenong Up Direction Down Directio n To Dandenong To Pakenham DNG719 DNG749 DN G 7 2 9 DNG759 Road Converted to Home Signal N A A Dandenong 450 metres Southern Bypass Main Line Separation Down Directio n Safety Fence Greens At Grade A DNG770 DNG776 To Cranbourne DN G 7 7 1 DNG778 DNG772 Loco Release Track A DNG773 Road Bombardier Site HARD STAND LOADING/UNLOADING AREA - by Others 450 metres 90 0 m e tr es Existing Inf r astr uc ture New Inf rastructure Removed Infras truc tur e V ic Tra c k B ou n dar y Figure 74: Dandenong Intermodal terminal Design Single Sidings Access (DoT, 2008e) Page | 338 Appendix D: Dandenong Case Study Figure 75: Dandenong Intermodal Terminal Design Central Sidings Access (DoT, 2008e) Dandenong Frankston - To Dande non g To Pakenham Road A A Dandenong Southern Bypass M a in Line Sep aratio n S afet y Fenc e Greens At Grade A To Cranbo urn e A Road 06 m HARD STAND LOADING/UNLOADING AREA Bombardier te er Site s HARD STAND LOADING/UNLOADING AREA 06 m te er 9 00 met re s s VicTrack Boundary Page | 339 Appendix D: Dandenong Case Study D.2: DANDENONG FREIGHTER SCHEDULE Figure 76: Dandenong Freighter Form Distance Time Chart (2008e) 22:00 23:00 00:00 01:00 02:00 03:00 04:00 05:00 06:00 07 : 00 08 : 00 09:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22: R1 1 2 1 2 1 2 1 2 1 Loading R2 1 Loading 2 Runaround 3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 DOCKS R3 1 AM Pe ak PM Peak 2 1 2 1 2 1 2 1 2 R3 GR EE NS RD Loading 1 Loading 2 Runaround 3 2 3 1 DOCKS Maintenance Locomotives Page | 340 Appendix E: Waste Study – Balanace Equations from System Dynamics APPENDIX E: Waste Study - Balance Equations from System Dynamics (Porta and Michetti, 2007) G t+1 = G t + TINt - SRTt DFT t+1= DFT t + SRTt – TOUTt DFWt+1 = DFWt + WIN t – SRWt CWt+1 = CWt + SRWt – WCt FCt+1= FCt + ФFCt – PFC t SFC t+1 = SFC t + PFC t – DFFC t - IFFC t Wt+1 = Wt + FWt - WOUTt SECt+1 = SECt + ФECt - ФFCt EW t+1= EW t + WINt - FWt where : G t+1 = number of trucks in queue at the Gate at t TINt = number of Trucks arriving at t SRTt = Trucks Service Rate, that is the number of trucks being served at t DFT t+1= number of Trucks in the Dumping Floor at t TOUTt = number of Trucks going Out at t DFWt+1 = mass of Waste on the Dumping Floor at t WIN t = mass of Waste being unloaded onto the dumping floor at t SRWt = Waste Service Rate, that is the mass of waste being processed at t CWt+1 = Waste Compaction at t WCt = Compacted Waste at t FCt+1= Full Containers at t ФFCt+1 = Flow of Full Containers at t PFC t+1 = Placement of Full Container onto wagon or in the storage at t Page | 341 Appendix E: Waste Study – Balanace Equations from System Dynamics SFC t+1 = Storage of Full Containers at t DFFC t+1 = Direct Flow of Full Containers IFFC t+1 = Indirect Flow of Full Containers Wt+1 = Wagons at t FWt = Freighted Wagons at t WOUTt = Wagons Out at t SECt+1 = Storage of Empty Containers at t ФECt+1 = Flow of Empty Containers at t EW t+1= Empty Wagons at t, available to be freighted WINt = Wagons In at t Page | 342 Appendix F: Theory Synthesis and Extension to Network Coordination Appendix F: Theory Synthesis and Extension to Network Coordination F.1 Introduction F.1.1 Chapter Guide The intention of this chapter is to take stock of the theoretical ground covered in devising the transhipment calculus to address the impedance phenomena of rail-road container terminals. This language of terminal relations grants insights into logistical relationships and key figures of merit of satellite terminal interaction with other nodes, such as a gateway seaport. It considers analysis and design on the basis of the terminal back-end driving the impedance relationship. Various paths through a terminal, called itineraries, can be mapped and controlled. This yields an understanding of terminal work. Control also leads to the evolution of impedance, a key means of assessing flexible capacity of the terminal operation. The mechanisms offer insights into the capability of the terminal configuration to meet alternate train operating forms with alternate dispatch schedules. The mathematical formulation may stand alone. This transhipment calculus also forms the basis for considering terminals as load following devices for operation of a coordination framework of gateway nodes. The relations between terminals and with the seaport can be seen in terms of a distributed resource system. In such systems terminals generate and attract consolidated flows, in this case by the rail link. They can also trade consignments between each other and share resources such as empty containers. Terminals offer buffer storages which are critical in consolidating loads and controlling port related road trips. Cycles for road and rail operations interface at the terminal and can be harmonised. The total supply of resources and infrastructure can be reduced for the load demanded. The transhipment calculus is the kernel of a coordination theory of transportation supply for the container freight task. This chapter describes this relationship and illustrates it with reference to the empty container task in a proposed Melbourne intermodal network. Page | 343 Appendix F: Theory Synthesis and Extension to Network Coordination This chapter also discusses the steps in freight network coordination using available algorithms in Power Distribution Theory. This sheds new light on a range of direct and ancillary decision support issues relating to freight container activity. A focus of this chapter is a response to the empty container resource coordination problem. F.1.2 Drivers and Means to Introduce Intermodal Forms in Urban Container Networks In Sydney, for example, poor harmonisation of system resources (including storages, train path access and positioning of empty containers) under conditions of an increasing flow avalanche from the seaport is adding to hinterland congestion (SFC, 2004). However, capability assessment techniques of the infrastructure supply remain rudimentary (SFC, 2005) and give little guidance on how system retrofit with the introduction of intermodal operations can be designed. A design for five intermodal terminals has been proposed (FIAB, 2005) with dedicated freight lines to the seaport and shuttle trains of 600m length as a partial solution to constrained hinterland capability at Port Botany. A nominal levy on container road transport accessing the seaport (to be refundable if travel is after dark) has also been mooted. This proposal has much engineering merit in bypassing current governing impedances, such as the sharing of passenger rail with its consequent restricted time windows and providing frequent pulse rail services. Such options involving intermodal interfaces require analytical techniques at a terminal and system level in order to gauge their feasibility. Outstanding engineering –economic issues centre on integrating intermodal terminal systems in logistical networks. There are no analytical means to direct the desirable evolution of the container hinterland. For instance, development of fine meshed systems of dense terminal coverage may have a number of benefits. It would minimise the size area of the terminal, reduce collection and delivery line-haul by trucks and would capture smaller and fragmented flows (Howard, 1978), (Woxenius et al., 2005). Introducing flexibility in destination choices would require further differentiation of terminal functions to include Page | 344 Appendix F: Theory Synthesis and Extension to Network Coordination facilities for sorting and stuffing/unstuffing containers. These novel network options face obstacles due to the existing infrastructure, ownership of resources, land availability, interoperability of interface exchanges, and the perceived economies of scale of large terminals. A starting point to test novel intermodal designs in a freight production system is to develop techniques in understanding impedances at a node with differentiated functions. The transhipment calculus acts as a basis for an activity-based approach in designing the container freight task using intermodal infrastructure, particularly in the urban transit leg. This leads to sketch planning models which are able to observe and control the traffic- transfer relationship under different flow conditions. We can consequently ascertain i) how intermodal production systems might operate as distributed systems and their resulting synchronisation requirements, and ii) how such systems can be integrated with the freight logistical structure and their coordination requirements. F.1.3 The Connection Between Transhipment Calculus and Coordination Theory The scope and role of transhipment calculus theory is again depicted in Figure 77. The role of the transhipment calculus is to provide a load following mechanism for the coordination task which for this thesis is viewed as both the economic dispatch/ unit commitment task and the fuel scheduling tasks. The transhipment calculus represents the attraction- impedance characteristic step that drives network transport models. The transhipment calculus captures impedance behavior in a novel fashion in that 1) the node is considered the critical cause of impedance over the links; 2) impedance is more than a delay measure; and 3) impedance is not only a capacity constraint phenomena but also a measure of precision and work necessary for required consignment transformations. Page | 345 Appendix F: Theory Synthesis and Extension to Network Coordination The consideration of system coordination extends the scope of measurement of hinterland absorptive capability to include for instance the effect of truck and train circulation rules on terminal capability and system throughput capacity. Page | 346 Appendix F: Theory Synthesis and Extension to Network Coordination Figure 77: Scope and Role of Transhipment Calculus Page | 347 Appendix F: Theory Synthesis and Extension to Network Coordination F.2 Distributed Resource Systems F. 2.1 Distributed Energy Resource Systems F.2.1.1 Description A further analogy is drawn with power distribution systems and the freight task. Distributed Energy Resource Systems have developed in the last 20 years as a means to conserve system resources, minimise distribution losses, and offer opportunities for locally supplied fuel including renewable sources. A characteristic of distributed energy resource systems is that consumers may send surplus energy back into the grid. In this way attractors of activity (loads) can also become generators (sources). . Trade between suppliers can lead to further injections of supply and ensure supplies are reliable. Critical supply ranges are monitored in the USA by the North American Electricity Reliability Council (NERC) and in Australia by the National Electricity Market Management Company Ltd (NEMMCO). These bodies oversee a market which fosters power matching through Industry Benefits of Trade (IBOT). A significant area of research and practice is ensuring that such distributed systems provide similar or greater security than a centralised generation and distribution system. Power network theory and practice have developed metrics in assessing the capacity- risk relationship. The Economic- Security problem deals with the likelihood that a certain amount of supply slots (generative power) will be available under different demand-load conditions. Electrical power distribution theory has developed a number of algorithms to analyse decentralised node generators because a major research endeavour is to assess the cost and energy savings and improved reliability of smaller generating plants. Specific applications include introducing renewable energy sources to the existing power network (Korpaas, et al., 2005) and assessing the benefits of transient storages over augmenting transmission lines (Koeppel, et al., 2004). The two mechanisms at play are generative scaling and inter-temporal storage, which combined, can reduce the total supply effort required for the load demanded. Inter-temporal storage represents a dynamic cumulative function of a network. Generative scaling considers the feasibility Page | 348 Appendix F: Theory Synthesis and Extension to Network Coordination of smaller terminals to service the same overall throughput. These coordinating concepts also have validity in the design of novel freight networks. Consignments arriving at different rates can be qualitatively transformed such that the next dispatch in the sequence of flows has higher load factors and characteristics which mean less handling complexity upstream. F.2.1.2 Conceptual Modelling Mechanisms Using the power systems analogue it can be depicted how the attraction-impedance characteristic is related to the phenomena of route allocation in transportation science. The enumeration of a characteristic impedance for an intermodal production system is a sparse representation of terminal generation of services and provides the basis for generative control for coordination of logistical services. This is the transhipment calculus presented in this thesis. The capacity of the terminal is linked to respond to changing flow patterns. The ability of the network to switch flows enables the planner to investigate how different system formats coordinate freight flows most effectively. Significantly, when considering retrofitting terminals of distributed functions to an existing network, the time capacity, or temporal reserves of the system may alter significantly. Power Distribution Theory has developed methods in the coordination of the various components of time which connect the precision of operations with the infrastructure investment deployment (Wood and Wollenberg, 1996) . The components of assessing available temporal reserves, the level of temporal accessibility, are presented in Figure 78. This is a coordination approach over a conventional allocation approach. In the coordination approach, the impedance mechanism is not set a priori but rather responds to the loads required. Page | 349 Appendix F: Theory Synthesis and Extension to Network Coordination Figure 78: Time Mechanism Components of System Coordination (Beavis et al., 2007) In addition to considering capacity as a constraint-based phenomena, we consider how an allocation-search procedure engages the compensation mechanisms of an itinerary through the Logistical Unit Processes (LUP) of a terminal. These compensations may involve different combinations (or couplings) of processes and technologies within terminals (Load Following). The characteristic impedance function represents the value flux for the required transformation and this generates an available cost-volume curve. Using the least cost points across the curves generated, a trajectory is found within the available technical envelope which involves both static and dynamic physical capacity for the lowest system cost (Economic Dispatch and Unit Commitment). Dynamic Scheduling requires polling of node itineraries of the intermodal production system and the competing all-road system (Fuel Scheduling). This allows us to assess the benefits to accrue to a system with the retrofit of an intermodal sequence. The utilisation of intermodal production systems may be improved with the facilitation of dynamic routes. Additionally, regions of deficit and excess resources (including storage, and train wagon slots) can be balanced (Multi-Area Exchange). The load following response of the terminal itinerary is the key in assessing flexible capacity for a nominated system. This is based on the sparse representation of node topological complexity, the transhipment calculus. The overall measure of temporal reserves is the measure of hinterland absorptive capability in the intermodal network study. The components of assessing temporal reserves are contiguous. That is, though designed and enacted over different timeframes, they work closely together. Page | 350 Appendix F: Theory Synthesis and Extension to Network Coordination F.2.2 Distributed Function Systems in Container Freight Task F.2.2.1 Functional Relief of Seaport The current system is characterised by low efficiencies with extended closed loops. For instance empty container returns take up to 30 days to cycle back to the seaport for repatriation. Utilisation of vehicles is often poor with low backloading rates in their tours. Slot availability by road is reduced due to pronounced peaks. These events are symptomatic of the dispositive structure of the urban container network. Coordination is a more tractable solution method to resource routing under conditions of flow imbalances than a conventional allocation approach. The increasing asymmetrical nature of container flows in both Melbourne and Sydney place an inflated burden on the re-positioning of freight resources (vehicles and empty containers). A central aspect of a coordination technique is the harmonisation of system resources through the flow cycle and storage interaction of infrastructure and logistical forms. This will lead to increasing system productivity. Distributed resource systems offer a means to extend the capacity and flexibility of a network by offering storage opportunities which can accumulate flows and facilitate mode shift and increase carrier productivity on road. By reducing the total number of trips, distributed resource systems also have a demand management objective. This network form, and the manner of operation it entails, offers a response to the growing avalanche of flows anticipated with larger vessels berthing at the seaport. Our hypothesis is that distributed resource systems relieve critical congestion points by increasing access opportunities to the port via the hinterland and thus increase overall productivity, flexible capacity, and harmonisation of system. F.2.2.2 Actor Governance Relations Seaport-landside relations are in transition, analogous to the actor relations evolving in Power Generation and Distribution Networks. There is a growing realisation that the productivity of the seaport must be extended to consider landside constraints. Page | 351 Appendix F: Theory Synthesis and Extension to Network Coordination Centralising many container freight activities within the seaport precinct reduces the accessibility and throughput capacity of the seaport whilst acting as both an attractor and generator of dissipative truck activity. Dispersing and distributing functions traditionally centralised in the seaport precinct may aid in increased productivity. Satellite terminals may also assist in shaving peak demand of truck trips. The sketch planning techniques described in this chapter are to build decision support for government to assess investment options that have strategic and tactical traction. Provision of seed infrastructure (including rail head) around terminals to facilitate certain functions is one means of government to facilitate port shuttle activity. In order to guarantee track path access and utilisation, government might also subsidise operations such that a take-or-pay facility with the standard gauge rail track landlord, ARTC, could be arranged. A minimum number of rail services are guaranteed. This builds up trust with shippers. The Victorian Intermodal Toolkit (VFLC, 2007) has suggested a number of operational subsidies that the State Government could provide to foster intermodal start-ups. F.2.3 Analysis of Gateway Relations Leading to Measurement of Hinterland Absorptive Capability Using a network of gateway nodal relations allows the planner to focus on critical sections of the freight network amid the combined passenger and freight network. The interface between nodes and links can be more readily analysed as well as the functions attributed to the node and the interface between nodes. Decision-making support here is very different from the motivations using conventional freight network models. The gateway model is utilised to analyse storage dynamics which can reduce the link flows; the network model builds in as inexorable and growing. The coordination of a gateway network involves resolution of a number of interlocking constraints. These are necessary filters to apply in an algorithm before optimisation schemes (which are usually over minimising costs). These unit commitment constraints can be unit based or interacting (coupled) and ensure that capacity limits, etc. are considered. Page | 352 Appendix F: Theory Synthesis and Extension to Network Coordination F.3 Description of Coordination Applications F.3.1 Melbourne Intermodal Tri-Lobal Network From the highly inefficient, centralised and dissipative state of seaport-landside relations, we wish to attain a system retrofit of a new configuration of common user access intermodal terminals. These terminals act as nuclei for further distribution and consolidation activities at road hubs using B-double trucks. For Melbourne there are three areas that are to be developed with this configuration: Western, Hume, and Dandenong. Each area is to have at least one common user access terminal as well as a private terminal (currently planned). Intermodal terminals will generate train shuttles to the seaport. In order to operate within their storage limits and to interface successfully with the road and rail links, terminals must exercise precision in their operations. Figure 79 is a conceptual layout of this tri-lobal structure. Additional opportunities for interaction between the areas include the interchange of empty containers as one area develops surpluses to meet another’s deficit. These infrastructure and the logistical structures they facilitate, represent a distributed resource system alternative. This system is for international container trade, import and export. Flows are then destined for or originate at the seaport for our system. Once the terminal network areas are developed it is possible that cross city shipment for general freight (intra-urban) could be facilitated. This gateway network may extend to including direct connection between intermodal terminals within the same area. This is a feasible addition if train operating forms, such as liner trains, are used. These trains unload and load containers throughout their journey. Parts of the urban general freight task could be unitised and captured. Page | 353 Appendix F: Theory Synthesis and Extension to Network Coordination Figure 79: Melbourne Tri-lobal Intermodal Network A perceived benefit of dispersing seaport functions to dry ports include the ability for shuttle runs of empty containers to intermediate nodes by dry port rather than single returns to the hinterland by road. This provides transient storage as an alternative to re- positioning empty containers by single truck runs. Empty containers can provide a local resource in matching the importer’s de-hiring of empty containers with the exporter’s need for empty containers. This concept is called triangulation68. Empty containers may also be repatriated by a combination of road and rail to the seaport. Additionally, an extended means to handle empty container movements is to include them in a multi-area network exchange algorithm. More efficient and effective re-positioning of empty containers becomes a demonstration of improvements towards harmonisation. Figure 79 offers an alteration on the traffic-node configuration typologies offered by (Woxenius et al., 2004) and Kreutzberger (2008). The tri-lobal network is ostensibly a line network by the rail mode. It also incorporates pre- and post-rail haulage by the road network. The road network is seen as integral in supporting an overall higher 68 There are limitations on the full substitution of import containers for export consignments as imports are generally FEU and exports are generally TEUs, especially if they are agricultural produce such as wheat with high densities.. Page | 354 Appendix F: Theory Synthesis and Extension to Network Coordination productivity urban freight task by both road and rail using intermediate storage and bundling. This rail network skeleton form also facilitates road and rail based trunk collection and distribution in the local and multi-area circulation of empty containers. The network of one of the three areas from the tri-lobal network of Figure 79 is depicted in Figure 80. All freight to be containerised with an origin or destination at the seaport can potentially be coordinated through either the intermodal road-rail gateway, a road-road consolidation/distribution gateway or an all road link. The intermodal road-rail gateway acts as both road consolidation and rail intermodal. The prime objective is to dispatch and receive consignments by rail. Container stacks provide the source of these containers. These stacks are fed and maintained by the vehicle system. Where the stack is near overflow, there should be a facility to allow vehicles to reduce the stack by taking containers to the port directly. The operation of distributed local storages allows for improved productivity of container movements for both rail and vehicles. The vehicles can improve the productivity of their tour by taking containers to the seaport, collecting containers from the seaport to a local consumer and taking consumers’ empty containers to the stack at the intermodal terminal. On average the carriers on the intermodal gateway will be rail shuttles of 75 TEU each; through the road-road gateway these will be B-doubles able to carry 2.5 TEUs each; an all road link will be an articulated truck with one TEU each. There are critical node junctions through this network. These junctions designate when flows on links are combining. These links then may have more pronounced distribution capacity side constraints. One example is the sharing of the rail link to the seaport with interstate trains of 1500m in length which have a schedule with a level of unreliability which requires additional slack time. Both the rail and road networks contain transient storage gateways. These are coupled with consolidation and distribution nodes and represent the hydroplants in our analogue. Direct road shipment represents the thermal (non-fuel constrained) part of our analogue. Page | 355 Appendix F: Theory Synthesis and Extension to Network Coordination Figure 80: Network of an Area with Key Spanning Arcs The ultimate terminal design schematic was depicted in Chapter 6 for the Dandenong case study. The triple rail track provides for a wagon set to be loading whilst another loaded set is dropped off, the locomotive detaches and then runs-around the siding (track 3). The central stack is accessed separately for loaded imports, exports and empties. These facilities are a combination of road-rail intermodal and road consolidation/distribution facilities. Their operation may be feasible with the use of High Productivity Vehicles (HPVs). When there are time sensitive trades and train slots are not available, HPVs can shuttle containers from the IMT stack to the port and back. Container loading and unloading from the train is done directly and indirectly (via the stacks). The dwell profile of empty containers is considerably reduced. A significant proportion is repatriated to the seaport by rail. Page | 356 Appendix F: Theory Synthesis and Extension to Network Coordination F.3.2 Dynamic Routing The need for an activities-based approach to freight modelling may be illustrated with the break bulk function in container freight services69. Currently some 31% of export containers are sourced within the Sydney seaport precinct (considered here as Botany Bay Council boundary). Here, full container and less than full container loads undergo break-bulk transformation to form full containers of same destination for shipping. Break-bulk operations offer an economic justification for intermediate terminals in a network, particularly under a many-to-many hub form (Daganzo, 1995). Break-bulk operations represent indirect transhipment, and it is this activity we investigate in a decentralised form at intermodal terminals. Alternative sequences are depicted (Figure 82). The particular novel features of the decentralised break-bulk approach are that a rail leg is introduced as well as the opportunity for dynamic routing through intermediate terminals. The choices available are rail-road or all road. These urban intermodal terminals would be located within 40km of Port Botany as 85% of import and export containers are produced within this radius. <20km <20km Figure 81: All Road Haulage to the Seaport Precinct for Stuffing/UnStuffing (Elliptical arrows represent empty container returns) Figure 82: Intermodal Terminals in the Stuffing/Unstuffing Transhipment Task allowing rail leg (routes to the intermodal terminals are dynamic) There are some benefits and disadvantages with dispersing this seaport precinct function. Some apparent benefits may include: • Reducing the line-haul task by trucks and improved discrete-diffuse sequence – increasing the value density of freight; 69 This is also known as stuffing and unstuffing activity. Page | 357 Appendix F: Theory Synthesis and Extension to Network Coordination • Minimising the pre-haul fleet; • Reducing congestion effects from a large hub. Improving the access at a seaport for swift clearances; • With the possibility of decision-making flexibility becoming more decentralised, and separating collection from line haul, the control exercised at this level may reduce the peak level of road traffic and also improve the sequencing block of flows to facilitate more direct port clearance; • Increased flow attraction to these terminals based on dynamic routes of changing cost and availability; • Shuttle runs of empty containers via train to intermediate nodes rather than single returns to the hinterland by road. This provides transient storage as an alternative to re-positioning empty containers. Some challenges include: • Additional system storage required in order to bundle up smaller freight flows; • Intermodal precision required and thus greater decentralised control; • Increased delivery of empty containers to the hinterland. From a container freight perspective, the decentralisation of break-bulk activity from the seaport to a hinterland terminal represents distributed generation of the current freight service. Distributed generation systems establish couplings of transformation activities at nodes (Geidl and Andersson, 2004). Indirect and direct transhipment characterise composite functions at terminals. In incorporating this node topological complexity, the network becomes potentially flexible in supply of freight services. Facilitating dynamic routes is a significant potential benefit of developing intermodal production systems in the hinterland. Improved communications may lead to better coordination between available container space on rail and ready consignments. This facilitation of the Industry Benefits of Trade may improve system utilisation. A node-centric approach to networks allows us to gauge the benefits and feasibility of distributed generation systems. Page | 358 Appendix F: Theory Synthesis and Extension to Network Coordination F.3.3 Empty Container Movement and Repatriation The management of empty containers becomes a critical requirement to support the objectives of intermodal terminal local and networked areas. Transient storage in the local area allows empties to be held for future use in the local and neighbouring areas instead of all having to transit through a central portal. Current centralisation practices through the seaport lead to large transit times, high freight intensity, and low efficiency of truck movements. This centralising requirement s reduces the opportunity for novel consolidation and distribution networks in the hinterland. Repatriation of empties to the port will need to continue given the inherent mismatch in import and export consignments (based on absolute number, and container type). The freight intensity by road can be minimised by allowing a large proportion of empty containers to travel to the port on a rail block or as part of a harmonised sequence which includes smart vehicle back-loading between intermodal storage node, seaport and consumer. F.4 Coordination Mechanisms F.4.1 Introduction The harmonisation attributes in the urban-international container freight task can be addressed in a more superior fashion through power systems theory. This science theory allows us to consider the phenomena of transient storages and has an extensive body of economic-security algorithms. That is, power systems theory considers both price signals and access availability due to variable operating constraints. In this way demand management objectives can be tested explicitly. Power systems theory has four features that can be adapted to address harmonisation requirements in urban freight systems: 1. Coordinate the characteristics of different generators; 2. Inter-temporality and storage transients; 3. Security controls; Page | 359 Appendix F: Theory Synthesis and Extension to Network Coordination 4. Network depictions that support the characterisation of bundling and distribution areas. Hydro-thermal coordination is a sub-set of power systems theory that considers the deployment of distributed generative resources with their own operating constraints and performance. Often there are a small number of hydroplants that need to be coordinated with a large number of thermal units. The usual objective is to minimise the operating cost of the thermal units (assuming the hydroplants have a minimum cost generation). The state variable is often the changing level of hydroplant reservoir storage. The analogue addresses the problem of scheduling operation of storage resources across a time horizon of security and availability constraints and repeated economic dispatches. These generating units are fuel constrained. That is, water release schedules must be coordinated with inflows to the reservoir to maintain as high a head on turbines and limitations of the forebay operations will permit “(Wood and Wollenberg, 1996, p.213). Similarly with intermodal production systems, the production of port shuttles is constrained by sufficient inflow of containers and make up of containers from less than truck loads. They are also constrained by timely arrival of port shuttles and the unloading of import containers. Thus, storage constraints can be recognised. Hydrothermal coordination includes the tasks of scheduling, unit commitment and economic dispatch with each having pertinence to the rail-road intermodal network. That is, the task of scheduling an available network which is most efficient. The pursuit to improve the performance of the international container freight system at the seaport/landside interface has an analogue with developments in power network theory and practice. The seaport gateway is a generator of significant transportation conflicts when interfacing the urban system (Rimmer and Tsipouras, 1978). Convention power systems utilise large generating nodes (i.e. coal fired power stations). Alternative means of supplying load demands are investigated through the concept of distributed generating resources which may respond more appropriately to load demands. This has led to the disaggregation of utility functions and the introduction of power markets . Similarly, the development of intermodal hubs has the potential to relieve access constraints on the seaport. There are institutional and commercial arrangements which must be developed to support these new transportation supply structures. Page | 360 Appendix F: Theory Synthesis and Extension to Network Coordination F.4.2 Control of Generation Control of generation precedes allocation mechanisms of dispatch and scheduling. It is concerned with the control of generator units such as particular itineraries through intermodal terminals, or alternate modal itineraries (such as bi-modal overflow paths discussed in Chapter 6). A means is required to allocate load changes to the generators (Wood and Wollenberg, 1996). The control of generation problem in power distribution theory refers to instantaneous changes in mechanical and electrical torque speeds in response to changing loads and must be balanced. Speed-drop characteristics determine the sharing of load between multiple generators (Wood and Wollenberg, 1996. P. 337). The control optimisation mechanism is depicted in Figure 83. This is known as the Automatic Generation Control Logic (AGC). Figure 83: Process Control of Generation of Response to Loads (Wood and Wollenberg, 1996, Fig, 9.25, p.354) Transfer functions of characteristic impedance represent the relationship of deployment of terminal processes to changes in loads. The transhipment calculus demonstrated in this thesis has focused on precision and controllability response and requirements to changes in train pulse loads. The transhipment calculus could be extended for this control mechanism to allocate system loads. Page | 361 Appendix F: Theory Synthesis and Extension to Network Coordination F.4.3 Economic Dispatch The costs in transporting consignments over a road corridor may likely change over a day due to a number of costs including: congestion costs on the link, time window costs to access the seaport gateway at a certain time and resource costs to utilise available trucks. The rate of production of the intermodal service (in this instance, block export trains to the seaport) will change according to the rate of export destined consignment arrivals. If we compare these two sets of dynamics, we arrive at a ratio which indicates the relative benefits of dispatch via an intermodal terminal compared with direct connect dispatch to the seaport precinct. This shadow price can be mathematically represented as such (Wood and Wollenberg, 1996): T γ = dF/dPi/ dq /dPT ------(F.1) where: dFi/dPi is the incremental cost of production by road service with changing load T dq /dPT is the incremental cost of production by intermodal service From this we can gain a dispatch trajectory which minimises systems costs. When this ratio is highest, the intermodal service should be utilised. It can also depict the changing storage levels in our intermodal terminals – a proxy for size requirements. There is continuity in dispatch. That is, over a period throughput must be serviced. F.4.4 Economic-Security Mechanisms of Commitment and Dispatch F.4.4.1 Introduction to Security Requirements Unit Commitment is a filtering stage prior to scheduling which assesses the availability of units at each period. In hydrothermal coordination, unit commitment constraints can include operating minimum and maximum levels, shut down and start up time and reserve requirements. Unit commitments are best included early in the algorithm in order to ensure operations are secured. That is, there is less risk of overloading unit capacity and critical transmission lines. Unit level and system level constraints are included in the objective function using dual programming Lagrange relaxation. The Page | 362 Appendix F: Theory Synthesis and Extension to Network Coordination system level constraints become penalty functions which are adjusted to obtain unit schedules which more closely match system unit constraints (Shaw, 1995). F.4.4.2 Operating Ranges Ballis and Golias (2004) identify that intermodal throughput capacity works within a specific operating range. They observed productivity according to differently sized terminals and configuration/equipment types. This is also based on the rail interface form. The trade-off is cost/TEU versus TEU annual throughput. Their analysis for terminal configurations and rail operating forms is depicted in Appendix 1. We build on this for the synthesis of terminal production functions and observe that intermodal production systems capability can be clustered according to productivity curves (marginal cost and throughput). Productivity is a function of a number of elements: 1. Loading and sidings tracks, 2. Handling equipment, 3. Site Storage. Each element will have its own attributes such as number and dimensions. The topological configuration, in how these elements relate to each other spatially and logically, will also affect peak productivity boundaries of the terminal node. The configuration will be in part due to the service complexity at the terminal (serving different container sizes and product groups). The controllability under different load scenarios, the load following response, will be dependent on the layout flexibility of the terminal configuration. The upper and lower limits of feasible operation will be contained in the unit commitment constraints. These limits can be specified by best efficiency points whereby the terminal unit is operated at a point equal to or greater than optimal efficiency (Nilsson and Sjelvegren, 1997). This ensures that sub-efficiencies are not engaged. F.4.4.3 Reserves Reserves are a critical measure of flexible storage and resource capacity of a system format. Alternate distribution functions of gateway nodes may be trialled which can be Page | 363 Appendix F: Theory Synthesis and Extension to Network Coordination assessed for their flexible throughput capacity. The risk-capacity relationship is a significant trade off when considering the stability of distributed resource systems. Reserve results at each time-step, as measured by storage utilisation in the system, may indicate the performance of the system format in terms of flexible storage capacity. At least two notions of capacity reserves may be considered: spinning reserves and supplementary reserves. Spinning reserves are on-line reserves which specify the amount of on-line generating capacity in addition to what is required to meet loads at each period. Spinning reserves can be used to meet random disturbances such as loss of node or line access or unforeseen changes in load. The following equations are taken from (Cohen and Sherkat, 1987) −≥ ∑ min(GSPi Gii (), t SP ) SPRES () t ------(F.2) unitsi Where GSPi is the capability limit for that unit which can be supplied to spinning reserves, Gti ()is the generated amount to meet load demands, SPi and is the maximum contribution to spinning reserves for that unit. SPRES(t) represents the spinning reserve limit for all on-line plants. Supplementary reserves may also be considered. For these reserves, the system has a longer time in which to provide capacity. Units which are offline but are brought in at times of peak load can be also included. GSU i represents the capability limit of the unit which can contribute to supplementary reserves. −+ ≥ + ∑ min(GSUi Gii ( t )∑ GSU ) SPRES ( t ) SURES ( t ) ------(F.3) unitsi offlinei For the freight system under study, the evaluation of available reserves can be a measure of flexible capability as well as reliability/stability of the system. These reserve parameters are generally known before the optimisation. An output schedule can include spinning and supplementary reserves available and utilised for each unit. Page | 364 Appendix F: Theory Synthesis and Extension to Network Coordination F.4.4.4 Distribution Limits Distribution requirements of the reserve can also be considered on the basis that capacity limits on transmission links are not exceeded. I − −≤+ =ΤΤ+ −t ≤ FFll∑∑ li,, Pilkk D Fl ------(F.4) i=1 The flow over each transmission line, l, is a summation of flows from each generating node, j, less the demand over that transmission line, D required by k load sources. This feature can be used to assess the flexible capacity of alternate intermodal system configurations given the demand load profile envisaged. F.4.4.5 Investment Optimisation It is the intention of the sketch planning method to guarantee that the way the system is operated reflects the investment made. Investment optimisation in operations can be ψ t made through the Unit Commitment stage where a system opportunity cost []xi is φ t applied to determine whether units should be activated or not. ii()P represents the return for generating power during this period. ⎡ L ⎤ φμα()min()PCPRPt =−+−⎢ tt ()t ∑ (t ξt )Γ Pt ⎥ ------(F.5) ii t ii ii l l li, i Pi ⎣ l=1 ⎦ The opportunity value is calculated both from perspectives of units that are already on and already off. For those units already activated: ψφt =+−−−+tt++11[] ttt []xSxJi ii (,)01ii() PJx[ i 1]------(F.6) The decision to switch a unit on which is already off can be made by resolving the following: ψφtt=+++11t − −t −tt − [] []xJxi [ i 111] ii( PSxJ )ii (,) ------(F.7) t ψ t The best decision from state xi is to stay on, or turn on, if []xi is positive, and to turn off, or stay off, otherwise. Page | 365 Appendix F: Theory Synthesis and Extension to Network Coordination For intermodal production systems the use of the rail gateway can be predicated on the utilisation rates achieved within the scheduled departure time. Business rules according to minimum utilisation rates can mark a switch in the use of rail or truck. This investment optimisation can also be made at the interchange stage where a take or pay contract can be made in the utilisation of link paths. F.4.4.6 Minimising Freight Exposure Hassall (2007) has developed a metric to assess pollution and amenity effects of metropolitan freight trips. It reflects the productivity of freight carriers which can be affected by the business rules of logistical structures such as intermodal terminals. One proxy of this measure could be freight intensity (Roth 2000) which assesses the efficient movement of freight in kilometres/tonne (km/t). Smaller, more numerous consignments will have a greater freight intensity. Shaw(1995) suggests that such environmental measures can be included in a security-unit commitment protocol. If we associate the kilometres travelled by different carriers with a typical set of emissions (mg/km) we can associate the freight intensity with pollution levels. This can be controlled in our set of constraints according to each transmission control area (set of links) and time of day. F.4.5 Hydro-thermal Coordination A means to model this and depict the critical role in intermodal facilities in distributing resources is to use the analogue of pumped hydroplants. Pumping from a downstream storage, allows a hydroplant to generate at times of high marginal thermal unit cost. Pumped hydroplants are then used to shave peak loads from more costly thermal plants. The hydroplant generates power when the margin cost of the thermal plant production is high (high load demand). The hydroplant pumps from a downstream reservoir to its upstream reservoir at a time of cheap thermal generated power. This pumping continues to operate until the added pumping cost exceeds savings in thermal costs. Page | 366 Appendix F: Theory Synthesis and Extension to Network Coordination The pumped hydroplant model represents an inter-temporal volume flow loop and is used to match a load cycle. For the freight analogue, empty container returns to an intermodal facility as dehired from the importer allow both 1) local re-allocation opportunities for exporters; and 2) timely and efficient repatriation of empties to the seaport by a rail block. The movement of empty containers in and out of the intermodal facility can also be timed according to off-peak export and import load periods on rail and road. The local empty container system takes the supply of containers made available by imported containers and if they are of an appropriate type for the exports, matches these. Import containers arrive at the terminal by rail and are either 1) delivered directly, 2) unstuffed on site, or 3) sent to a local distribution centre. If we assume that case 3 reigns, the empty container is then carted to the local container park, available for local use or to be repatriated back to the port by a combination of the rail and road system at low peak loads. Where there is a surplus of supply for local needs, the empty containers are available for interchange with another area. Another layer of complexity has empty containers imported in order to meet export needs. In the importing case, the arrival of imports at the intermodal facility is further distributed and the container is lodged at the adjacent container park. At times of low demand for shuttle services, the empties are “pumped” to the intermodal terminal and either directly transshipped onto rail or stuffed and transshipped. Mechanism of pump storage plants: Developing a schedule according to the gradient method is based on considering the change in costs to the thermal plants of generating from the pumped-storage plant and also a change in thermal plant costs due to pumping an amount to return volume storage to original levels. Pumping requires some utilisation of thermal generated power (in our case, road-based work). There are two intervals at work: a loaded container generating interval {k} and an empty pumping interval {i} . The following equations are taken from (Bainbridge et al., 1966). Two equations determine whether the hydro-pumping mechanism will proceed: Page | 367 Appendix F: Theory Synthesis and Extension to Network Coordination dF ΔΔ=− k Fk PHk (F.8) (Change in cost to the thermal plant by hydroplant generation; dPsk in the freight case this represents the peak shaving capability of the intermodal terminal to supply empty containers for local as well as port needs) dF ⎛ Δ P ⎞ Δ F = i ⎜ Hk ⎟ i ⎝ η ⎠ (F.9) Change in cost to the thermal plant by hydroplant pumping; dPsi this describes the cost of setting up for the peak shaving period. If the total cost change is negative, then hydroplant generation will occur in k and storage replacement by pumping will occur in interval i. Otherwise these intervals will be mis-specified. The constraint to achieve the same volume at the intermodal facility as at the start of the process, can be specified for any future time period. Similarly, the logic of the empty container movements–repatriation by rail to the port, are based on the relative marginal costs of running road trips in peak load. If road is expensive, freight (including empties) will be railed. If road is cheap, loaded containers will be encouraged to travel by road and there will be less demand on rail. Consequently, empties will be “pumped” by rail to and from the intermodal facility. In this way, empty repatriation by rail over road is encouraged. This can be compared to base case costs whereby empty containers are repatriated to empty container parks adjacent to the seaport and then may be called out for stuffing in the hinterland or repatriated back overseas. There are two intervals at work: a loaded container generating interval {C} and an empty pumping interval {E} . We can find an optimum schedule for delivery and loading of empties versus loading of full containers onto trains. Further complexity can be added to allow a mix of empty and full containers to be sent in the same rail shuttle. Page | 368 Appendix F: Theory Synthesis and Extension to Network Coordination F.4.6 Network Multi-Area Exchange The maintenance of the stack at intermodal facilities is essential for buffering the interface between road and rail and maintaining critical consolidation and vehicle utilisation rates. Where there are insufficient bookings for rail, delivery to the port can switch to road carriers. The localised and multi-area exchange of empty containers is a crucial mechanism to support the operation of intermodal terminals and achieve objectives of reducing freight intensity on road. Once the empty stack falls below a certain threshold, empties from importers need to be called to the terminal for repatriation to the seaport or for exchange with local importers. If there is excess in the stack to these requirements and road pricing is cheap, there are opportunities for multi-area exchange of empty containers. The network is defined as coupled nodes in multi-areas (see Figure 79). Each area can be designated as a bundling/distribution network (number of pickup and delivery centers). The critical resource which they need to share is empty containers. This approach may indicate a preferable location of empty container repositories, their size and relation to the network and other nodes. Multi-area mechanisms can show how the distribution of freight functions improves system harmonisation objectives. Intermodal facilities and bundling networks will have a ready supply of empty containers to call upon on site or in their local area. Given the unsynchronised nature of flows, the exchange of empty containers can be between areas directly and also indirectly (passing through an area not requiring the containers). These indirect relations enlarge the scope of the market for such interchange transactions. They however utilise link capacity and should incur a coordinating charge called “wheeling”. Enabling the interaction of multi-areas represents pooling. Pooling is an extension beyond area allocation and acts to improve the efficiency and reliability of the system. The method of diakoptics has been developed to address allocation of flows in interconnected systems. Diakoptics segregates the system into discrete areas which are Page | 369 Appendix F: Theory Synthesis and Extension to Network Coordination solved separately, then ties these torn areas together back to the system (Aldrich et al., 1971). The multi-area formulation involves minimising fuel costs of generators in all areas subject to flow balances. The following equation set has been proposed by Aldrich, et al., ((1971) An inter-area matrix is solved which yields all tie and line flows: P Gε Tβ =+TβεG Tβ PCDGε ek F ------(F.10) Pek Area Lambdas are calculated by the Lambda ratio matrix below. The Lambda ratio matrix (H) acts to tie the torn pieces of the system together after they have been solved separately and their net interchange needs have been resolved. In this instance we have three interchange areas (A,B,N). We relate the partial differentiation of their area losses with the net interchange from the area. ⎡ ∂P ⎤ − LN ⎡ λ ⎤ ⎢1 ∂ ⎥ − P ⎢ A ⎥ = H 1 ⎢ eA ⎥ λ (F.11) where: λ ⎢ ∂P ⎥ N ⎣ B ⎦ − LN ⎢1 ∂ ⎥ ⎣ PeB ⎦ ⎡ ∂P ∂P ⎤ ⎢1 + LN LB ⎥ ∂P ∂P = ⎢ eA eA ⎥ H ∂P ∂P (F.12) ⎢ LA + LB ⎥ ⎢ ∂ 1 ∂ ⎥ ⎣ PeB PeB ⎦ Iteration is made between the area lambda ratios and the inter-area matrix until pool constraints are satisfied. This is a method to test the synchronisation resources over transient storages in the network. Empty containers can be traded according to areas of excess generation and demand. Achieving the objectives of the intermodal production system will require some centralised guidance on encouraging zoning of areas to use specific intermodal gateways to minimise road trips through the city. This justifies a model design whereby Page | 370 Appendix F: Theory Synthesis and Extension to Network Coordination resources can be exchanged between areas without use of the centralised port. The techniques of diakoptics described in this section are then appropriate. F.4.7 Figures of Merit It was observed in Chapter 3 that the performance metrics from the European Terminet project (Terminet, 1998) on achieving a quality jump in intermodal transport were of great interest but needed to be modified for urban intermodal operations in Australian cities. The measurements were based on terminals with generally large hinterlands and extensive line-haul connections. The system boundaries were rail leg focused and did not consider the productivity implications of pre- and post haulage by trucks. The results were consequently biased towards large terminals which supported large throughput levels based on perceived economies of scale. Smaller terminals thus faired less well with respect to cumulative storage capacity and land use intensity ratios. The coordination mechanism discussed in this chapter allows for smaller intermodal terminals to leverage the connection between rail and road modes. In these cases, the storage claim70 can be much larger than the nominal train operating form which services the terminal . The stack is managed by both rail and bimodal (B-double or super B- double) road operations. The stack is no longer the constraint to throughput. Sidings can also be opportunistically activated to relieve the stack and manage multiple train services. Therefore the cumulative storage capacity and land use ratios can be much higher even for conventional direct connect satellite terminals. Their capacity-risk operation can also be lower. For the urban container freight task the interest must lie in increasing the productivity of the whole logistics chain across road and rail modes. This grants improved accessibility and clearance throughput to the seaport. The operation of the intermodal terminal can thus improve load factors for both road and rail and reduce the freight intensity of the container task. Circulation indices for both road and rail trips interfacing the intermodal terminal are then of great significance. 70 The number of load units exchanged via the storage area per micro-constellation or rail operating form (Terminet, 1998, p.24) Page | 371 Appendix F: Theory Synthesis and Extension to Network Coordination F.5 Conclusions Distributed resource systems can be an innovative way to reconfigure transportation networks. They can be used to manage congested networks through transient storages which increase system productivity. Such systems have an explicit retrofitting role. The innovation may involve the structures infrastructure, transport technologies and logistical arrangements. To ensure the system’s stability and controllability we need to integrate the measure of precision and available reserves in an optimisation framework. The model formulation and solution method described here adapts theories from hydro- thermal power distribution systems. This science theory is attractive as it involves the dynamics of storages at nodes and can include economic and engineering mechanisms for testing the complements of different generating nodes and thus sketch an intermodal system which is both coordinating and demand steering. The model can act as a tool to support increased carrier productivity and enact peak shaving. The transhipment calculus developed in this thesis operates within the phenomena of attraction-generation and represents the impedance relations to freight service loads. It forms the nucleus of a coordination mechanism in harmonising cycles associated with container flows in the urban environment. It may be extended to consider the allocation task. The task is more responsive to the activity based approach whereby terminals generate itineraries which are paths of commitment (availability) and dispatch (allocation). This chapter outlines how the transhipment calculus can be incorporated into a larger coordination role. This involves unit commitment, fuel scheduling and multi-area allocation algorithms. The coordination mechanism responds to the need to harmonise container flow cycles and measure the absorptive capability of the seaport hinterland. It does this through the Figures of Merit: Capacity-risk and throughput and effect of circulation rules for both truck and train resources on system slot capacity. The economic-security mechanisms allow measurement of resource availability prior to allocation and dispatch. An understanding of system availability under alternate system formats of terminal configuration and function can now be assessed. Page | 372