Proceedings of the 2008 Winter Simulation Conference

S. J. Mason, R. R. Hill, L. Mnch, O. Rose, T. Jefferson, J. W. Fowler eds.

A SIMULATION APPROACH TO THE EVALUATION OF OPERATIONAL COSTS AND PERFORMANCE

IN LINER SHIPPING OPERATIONS

Aldo A. McLean
William E. Biles

Dept. of Industrial Engineering

University of Louisville
Louisville, KY 40292, USA

ABSTRACT fixed weekly schedules. These schedules are kept in place

for a few months or for a year (Shintani et al. 2007). How-
This paper presents a simulation model of the operation of ever, a unique service route cannot fully accomplish the
a liner shipping network that considers multiple service objective of shipping cargo to a large number of destina-
routes and schedules. The objective is to evaluate the op- tions. A group of service routes forms a service network.
erational costs and performance associated with liner ship- In liner shipping, container cargos move through the net-
ping, as well as the impact of individual service schedules work from their origin and a port of destination. Virtually
on the overall system. The approach proposes a discrete- any combination of origin-destination pairs is possible in
event simulation model where shipping activities, con- the network when transshipment of cargo is allowed.
tainer ship operations, and intermodal container move- Therefore, the configuration of the network is an important
ments are considered. The model allows for direct and aspect to the performance and quality of the shipping ac-
transshipment operations of container cargo, and the tivities. Evaluating the cost and performance of a service
evaluation of fuel consumption and other logistics metrics. network and their individual components, where conditions
The model is used to evaluate a liner shipping network are similar to the real operation, suggests opportunities for
consisting of four service routes, up to 64 container ships, improvements and savings in operations.
and up to 20 ports with diverse physical characteristics and Higher canal transit tolls, port fees, growing demand
cost components. The results show the contribution of ser- for liner services, and an expanding number of destinations
vice routes, ports, container ships, and containers to the suggest major challenges in modern liner operations. Per-
cost and performance of the system. haps one of the most important characteristics of the liner
shipping industry is its high fixed cost (Haralambides and
1 INTRODUCTION Veesnstra 2000). Despite improvement in the design and
propulsion of container ships, the instability of prices for
Simulation modeling is an analysis method where com- marine fuel oil is a major concern in the industry. Current
puters are used to evaluate a model numerically in order to prices oscillate between $550.00 and $600.00 per metric
estimate the desired true characteristics of the model (Law ton (pmt) based upon port locations. Cooperation and
and Kelton 2000). The nature of liner shipping processes global alliance, the increasing use of transshipment opera-
is characterized by its dynamic operations and diversity of tions and the size of container ships have helped the indus-
cost components. The dynamic behavior in liner shipping try to be flexible and resilient. Determining the effects of
is characterized by the movement of containers. Contain- these aspects in the performance of the shipping process is
ers are transported in container ships from port to port paramount to the design of new services and decision to-
throughout one or more predefined service routes. The ward current ones.
high costs in each activity and the large investments in as- This study describes the characteristics of a simulation
sets are characteristics of these operations. However, the model designed to mimic the operation of a liner shipping
operational performance determines the quality of shipping network. The Arena simulation package (Kelton et al.
service and the profitability of a service route. 2007) was used to code and run the model. An application
A service route is a sequence of container ports with a case was proposed and solved using the model. The re-
deployed order of container ships that complies with strict sults show the expected operational cost and performance

978-1-4244-2708-6/08/$25.00 2008 IEEE 2577

 McLean and Biles

of the system in various aspects of shipping, including fuel tion of the vessel at lower speeds results in fuel savings be-
consumption and logistics. cause of the reduced water resistance is modeled by the
This paper is organized in six sections. Section 2 pre- cube rule (Stopford 1997). This relationship is used in Lo
sents a previous work in liner shipping and simulation. and McCord (1991) to suggest fuel oil savings using ocean
Sections 3 and 4 present details of the simulation and logic current information.
model used represent and evaluate the network. A case
scenario and results are presented in Sections 5 and 6, re- 2 DESCRIPTION OF THE SIMULATION MODEL
spectively.
Rockwells Arena software (Kelton et al. 2007) was used
1 LITERATURE REVIEW to develop the simulation model. The model evaluates the
operation of liners, movement of intermodal containers,
Simulation modeling has been used effectively to evaluate operation in ports, and the contribution of each service
determine the design and evaluate the operation of con- route and schedule to the operation of the network.
tainer ports in liner shipping. The operation of liners and The model mimics the operation of container ships
their interaction with shipping process and service sched- sailing and transporting containers between ports. Once a
ules has been the objective of recent simulation studies. A container ship arrives at a port, arrivals container cargo are
simulation approach that considered container ships and offloaded and departing container are uploaded on the sto-
intermodal containers was presented in Van Rensburg, He, wage area of the vessel. The rate at which this process
and Kleywegth (2005). They proposed a discrete-event takes place is set by the simulated turnaround rate of the
simulation to evaluate events in liner shipping including port and the number of container assigned to the operation.
voyage costs, fuel cost, and capital cost built in a modular The ship waits until the service period is completed to sail
fashion using C++ functions. Goldsman, Pernet, and Kang slowly out of the port. Once at sea, the container ship
(2002) used Arena to build a simulation model that in- adopts the predefined service speed for the route segments
cluded container ship, RORO, and break bulk operations to in transit. A ship in a given route will adopt the predefined
estimate the time required to move materials by a given segment service speeds of the route until it reaches the an-
number of ship. Lai, Lam, and Chan (1995) developed a chorage area of the following port in the sequence. Sig-
model for shipping operations in the Europe-Middle East- nificant container traffic is generated upon arrival of con-
Far East routes to evaluate several policies for allocation of tainers at origin ports. The process of a container ends
containers. A simulation analysis of the advantage of fast once it reaches it destination port and exits the system. A
ship services in liner shipping was presented in Ryan container ship operates continuously until the end of the
(1998). Bendall and Stent (2001) presented a simulation planning period. Multiple service schedules can operate
analysis applied to fleet design and deployment in liner together, and more than one ship of a different route can
shipping under hub and spoke networks. Simulation has call at the same port at the same time.
been used to study inland waterway shipping processes in The model was validated using round trip service pro-
Swedish (1998) and Bush, Biles, and DePuy (2003). formas obtained from contributing liner operators. The re-
Other studies have integrated service routes analysis sults of the time performance in the simulation successfully
and the allocation of container cargo. Song, Zhang, Carter, matched the sailing, service, buffering, and docking time
and Field (2005) measure the cost-efficiency of trade of voyage summary in proformas. A complete walk-
routes based upon the assignment of container cargo. Ben- though of the model, using the debug feature in Arena, and
dall and Stent (1999) evaluate trade routes for container a visual analysis of models Animation features comple-
ships to avoid reducing load factors. Baird (2002) evalu- mented the models validation and verification of the cost
ates the feasibility of transshipment operations in northern and operation of ports, containers, and container ships.
European routes. The model was divided into seven main sub-modules
Fuel consumption is an important aspect in shipping. and three operational logics. The operation of each sub-
Different approaches are used to model and evaluate the modules are described next.
consumption and cost of containership operations. The
approach in Eljard (2006) includes a set of equations that 2.1 Container Port Operations
account for the specification of a container ship and cargo
load conditions to forecast the actual fuel consumption in The main component of the model is its Generic Container
the vessel at a given sailing speed. Cullinane and Khanna Port Sub-module. This module has the objective of resem-
(1998) present a set of equations to determine the fuel oil bling activities and events that take place in a container
consumption (FO) base upon the average specific fuel oil port, such as berthing, service, and storage. Port Modules
consumption and power utilization. Shintani, Imai, Nishi- are easily configured to represent the basics characteristics
mura, and Papadimitris (2007) evaluate fuel and lubricant of real container ports. Configurable parameters include
consumption using the specific consumption. That opera- some physical characteristics of ports. Length of access

2578
 McLean and Biles

channel, storage capacity, and number of quay cranes are lows large flexibility to the model to represent different
some of these parameters. In term of costs, container stor- sizes, capacity, and speeds of container ships. Although
age, stevedore, and container ship docking charges are in- any combination of container ships can be deployed, the
cluded. Evaluation of time for shipping events, logistic re- current model offers the following container ship specifica-
sults, and costs in the simulated liner process are evaluated tions:
in the generic module structures.
Table 1: Container Ships Simulated Parameters

Capacity Type LOA Service Speed

(TEU) (m) (knot)
9,600 Post-Panamax 295 24.7
8,200 Post-Panamax 335 25.2
4,895 Panamax 295 24.5
4,738 Panamax 260 24.5

The discrete events associated to the operation of con-

tainer ships in the model are listed as follows:

1. Sailing: movement of container ship from a de-

parting port to the next port in the sequence.
Figure 1: Port Module Structure 2. Anchorage: waiting for berthing.
3. Pilot in: sailing into port from anchorage to berth.
4. Service: moving containers off and on container
Figure 1 shows the basic structure of the model. ships.
Twenty such modules are available in the simulation. Port 5. Pilot out: sailing out of the port.
modules become\unique after defining a name and arrival
rate of container through the definition of the Port Assign- Container ships experience delays during sailing op-
ment Table. The arrival of containers at port was modeled eration between consecutive ports. The delay distance was
using exponential distribution with a choice of four levels modeled using uniform distributed random variates in the
of interarrival rates to simulating volumes. Distance be- range between 0 and 1% of the total segment distance.
tween ports, including canals, are contained in the Distance
Matrix. This distances are given in nautical miles (nm). 2.3 Service Route and Schedule Modules
Discrete events associated to port processes are the follow-
ing: The model works under fixed weekly schedules and sup-
ports end-to-end routes exclusively. Service routes are
1. Downloading: containers are offloaded from a controlled by the Service Route Sequence table to config-
container ship. uring up to five service routes and schedules with 20 ports
2. Storage: containers are stored in container port sequence each. Service schedules are controlled by tables
yards. containing the Estimated Time for Arrival (ETA) and Es-
3. Uploading: containers are loaded to a container timated Time for Service (ETS). The ETA define the time
ship. for berthing and the expected length for service at port re-
spectively.
2.2 Container Ship Operations In the model, links between two consecutives points
(e.g. ports and canal points) in a service route are seg-
Container ships are modeled as entities. A Container Ship ments. It is assumed that container ships sailing a segment
Initialization Table sets up the initial physical and service will adopt a predefined service speed. A Service Speed ta-
parameters of container ships. These parameters include ble is used to synchronize service schedules by defined
length overall (LOA), stowage capacity, consumption rate, service speeds for each segment of each service route in
maximum service speed, number of crew, minimum num- the network.
ber of quay cranes required during service, and deployed
service route assignments. A container ships departure
time and location at the beginning of the simulation are al-
so defined in this table. The Ship Initialization table al-

2579
 McLean and Biles

2.4 Intermodal Container Operations tainers are collected without considering storage or pickup
procedures at ports.
Container traffic is generated through a Demand Matrix
table. This table stores a percentage of shipments required 2.5 Canal Transit
to be transported from an specific origin to each destina-
tion port in the network. The destination port is defined for Canal transits are simulated through the Canal Points
each container after their arrival at originating ports. The Structure. Canal Points manage scheduled transits through
Routing Table details the service route or path required to the Panama and Suez Canals and measure distance trav-
move container to their destination ports. Upon the arrival eled, costs per transit, and time spent waiting before tran-
of containers at origin ports, destination port and transit siting canals. Distance between Canal Points and ports are
paths are assigned to containers following the details in the defined by the Distance Matrix. Canal transit time is de-
Demand Matrix and Routing Tables. The discrete events termined by the transit speed and specific channel distance.
associated to direct shipments operations of intermodal In the simulation transit speed for Panama and Suez are 5
container operation are the following: and 6 knots while the length of the channels are 48 and 87
nm respectively. Figure 2 shows the structure of Canal
1. Arrival: containers arrive to their originating Points as seen in the simulation model:
ports.
2. Storage: containers are stored before their depar-
ture.
3. Transport: containers are transported on container
ships to their destination.
4. Exit: containers arrive to their destination.

2.4.1 Transshipment Process

Regularly, container transits require intermediate port stops

before arriving to their final destinations. These interme-
diate ports are considered transshipment ports, and the re-
quired container storage during stops are transshipment
storage (CTT1).
The simulation allows transshipment operations when
one or two transshipment ports are required. The trans-
shipment sequence is controlled by a two reference tables:
Transshipment Routing and Transshipment Port tables.
The first table controls the sequence of service routes to
define a path for transshipment transports. The second Figure 2: Canal Routing Structure (Panama Canal)
stores a group of transshipment ports required to link any
transit origin-destination pair. The discrete-events associ-
ated to transshipments operations are:
3 SUPPORTING LOGICS
1. Arrival: arrival of containers to originating ports.
2. Storage: containers are stored before their first The following logics supporting the evaluation of parame-
departure. ter of interest in the simulation model. The supporting lo-
3. First Transport: containers are transported on con- gics evaluate consumption, logistic, and cost results.
tainer ship to their first transshipment port.
4. Transshipment Arrival: containers arrive to trans- 3.1 Consumption Model
shipment port.
5. Transshipment Storage: containers are stored dur- The approximation used to the consumption of fuel oil, di-
ing transshipment operations. (CTT1) esel, and lubricants is applied on every container ship in
6. Second Transport: containers are transported on the model. Shipbuilders define and supply the consump-
container ship to their final destination. tion rate of their ships measured in metric tons per day
7. Exit: container arrive to their destination. (mtd). The values of the consumption rate are assigned to
each container ship through the Container Ship Initializa-
Upon arrival at the destination ports, containers imme- tion Table. The following list shows the consumption rate
diately exit the shipping system. Statistics on exiting con- for container ships represented in the model:

2580
 McLean and Biles

9,600 TEU = 248.2 mtd 3.3 Cost Model

8,200 TEU = 248.2 mtd
4,895 TEU = 156.6 mtd The simulation model works under a cost model defined to
4,795 TEU = 133.4 mtd account for main activities in liner shipping operations.
The cost model proposed is divided into four main catego-
Fuel oil, diesel, and lubricant prices are specified for ries:
individual ports. Fuel oil and lubricant consumption are
measured when container ships are at sea. For simplicity, Container ship related costs
their consumption is reported together. Diesel consump- port related charges
tion is measured during anchorage and while these ships canal transit charges
undergoes service at port with the assumption that auxil- container related costs.
iary generators are used. The relationship between con-
sumption and speed is modeled through the cube rule Container ship costs include all costs related to opera-
(Stopford, 1998). Equation 1 shows the general form of tion and maintaining the fleet. Fuel and diesel costs are de-
the cube rule. The exponent a is set to be 3. termined for individual container ship and is determined
from its consumption and the contracted price during op-
a erations. In container ships, cost and other charges are spe-
S
F =F *
*
(1) cific to each vessel. The cost for insurance, capital, crew,
S auxiliary, and maintenance are considered Fixed Costs.
All ship related costs are defined individually per container
Where, F current fuel consumption rate ship regarding their physical characteristics.
F* designed fuel consumption rate Port related charges group stevedore, dockage, and
S current service speed container storage charges. Penalty costs are charged on
S* designed service speed container ships for overstaying outside the scheduled ser-
vice time at ports. Storage charges vary based upon con-
3.2 Logistic Model tainer ports and whether containers handled are full or
empty, twenty or forty feet type. Three categories of the
The logistic model describes a set of metrics designed to storage charges had been defined as storage before first
report various aspects of the operation of container ship, departure, during first transshipment, and during second
containers, and ports. The most representative metric in transshipment stop.
the model is the Network Operation Cost per TEU Trans- Canal transit costs are related to the charges per tran-
ported and the Network Operation Cost per Hour. Other sits through the Panama and Suez Canals. Transit costs are
important indicators of the operation of container ship, evaluated per container ship visit and charges are estimated
port, and container are listed below: in a per TEU units base. Container costs include the lease
cost for inter-modal container in the network in dollars per
Total transit time day of use.
Transit time, direct shipments All charges and fees are loaded into the model through
Transit time, transshipments global variable matrices and available throughout the
Storage time at ports evaluation run.
Time at sea
Miles traveled per ship 4 CASE STUDY
Total vehicle miles traveled
The current application describes a network of four service
Number of moves at ports
routes, 18 container ports, and 24 container ships. In addi-
Container yard levels, TEU
tion, four transshipment routes, 29 port calls and four canal
TEU transshipped per port transits a week are considered. The service route and port
Ships stowage utilization call sequences for the case application are shown in Figure
Departure Storage Time 3 and Table 2 respectively.
Transshipment Storage Time
Ship stowage capacity utilization
Total TEU delivered

2581
 McLean and Biles

CTT1 flat rate = $0.50 per TEU-hour

Canal Transit Charges

Panama Transit = $54.00 per registered TEU
Suez Transit Sub-PMX = $80.00 per register TEU
Suez Transit Panamax = $56.00 per register TEU
Suez Transit Post-PMX = $50.00 per registered TEU

Transshipments
Felixstowe, UK, to Hong Kong, CN, via Norfolk,
USA, in both directions
Algeciras, SP, to Charleston, USA, via Bremerhaven,
Figure 3: Service Routes DN, in both directions

Table 2 shows the sequence of ports of call for each

Four types of container ships are required to serve these service route evaluated in the application. These routes
schedules: were obtained form published schedules submitted by sup-
porting shipping operators. The application did not include
Route R1 - 4 container ships of 4,895 TEU evaluations of container related costs. Operation cost and
Route R2 - 5 container ships of 9,600 TEU other logistic performance metrics are reported for the
Route R3 - 8 container ships of 8,200 TEU overall network and individual service routes.
Route R4 - 8 container ships of 4,738 TEU
Table 2: Service Routes and Sequences
This application case assumes all four service routes R1 R2 R3 R4
and schedules will not change during the evaluation period Rotterdam Yantian Bremerhaven Ningbo
and that ship deployments are fixed and container ships are Bremerhaven Hong Kong Suez W Shanghai
not diverted at sea. It is assumed that distances between Le Havre Los Angeles Ningbo Yantian
ports, price for fuel, diesel, and lubricants remain constant Felixstowe Oakland Xiamen Hong Kong
during the planning period. The application case assumes New York Pusan Hong Kong Panama S
the following price structure: Norfolk Ningbo Yantian New York
Charleston Xiamen Tanjung P. Norfolk
Container ship cost - - Suez E Savannah
Fuel oil cost = $350.00 pmt - - Algeciras Panama N
Diesel cost = $560.00 pmt
- - Rotterdam -
Lubricant oil cost = $1,000.00 pmt
Capital cost = $12.00 per 14-ton slot per day (pd)
Insurance cost = 10% capital cost pd
5 SIMULATION RESULTS
Crew cost = $500.00 per crew member pd
Auxiliary cost = $50.00 per crew member pd
Maintenance and repair cost = 5% of capital pd A total of twenty replications of three months each was
sufficient to produce good statistical precision of the re-
Stevedore Charges sults in the liner shipping network. The model run satis-
20-foot loaded = $55.00 fied a 95% confidence on the mean value for the Network
20-foot empty = $33.00 Operation Cost per TEU Transported and the Network Op-
40-foot loaded = $82.50 eration Cost per Hour of Operation measures. Results for
40-foot empty = $49.50 the overall network and individual service routes are re-
ported in the following lines.
Dockage Charges
First 150m LOA = $150.00 per meter hours (pmh) 5.1 Assessment on the Network Operation
Additional LOA = $4.00 pmh
Overstaying = $4.00 pm/15 min units The network operation summary shows an average Opera-
tion Cost of $419.581 million in three months of opera-
Container Storage Charges tions. The largest contributor to the operation cost is the
CPT first 7 days = $0.00 per TEU-hour Fixed Cost, $258.243 million, representing a 62% of the
CPT thereafter = $1.00 per TEU-hour Operation Cost of the network. The second largest is Fuel
Oil Cost at $78.661 million or 19% of the Operation Cost.

2582
 McLean and Biles

A key indicator is the Network Operation cost per of throughput and 25% of the total operation cost. The
TEU was $977.69 dollars per TEU, obtained after deliver- 29% of delivered TEUs in service route R3 is offset by its
ing an average 429,156.40 TEU of cargo during the as- share in operation cost of 39%. The lowest rate of deliver-
sessment period. The Network Operation Cost per Hour of ies is found in R4 with 16% and operation cost of 26%.
Operation showed $7,530.52 per hour in the network.
The summary also reveals 10% for Stevedore Cost, Table 3: Service Routes and Schedule Contribution
handling containers off/on container ships. Canal transits Operation TEU
Service Vehicle-miles
represent 5% of the operation cost at $20.073 million. Cost Delivered
Storage cost was 4%, $15.442 million, driven by the cost R1 11% 13% 22%
of retaining containers before their first departure or CPT R2 25% 20% 33%
at $15.111 million and storage during transshipment pro- R3 38% 32% 29%
cedures or CTT1 at $0.330 million. R4 26% 36% 16%
Regarding the operation time, the most notable result
is the sailing time of 74.5% of the networks time. Service In terms of vehicle-miles traveled, R4 and R3 are the
of container ships at ports accounts for 11.7% of the net- service routes with the highest performance. The vehicle-
work time while transits through Panama and Suez Canals miles traveled measure the total distance traveled during
represent 2.8% of the time. the evaluation period. Therefore, the low performance in
Fuel consumption results reveal a consumption of R1 and R2 is caused mainly by the number of container
224,746 mt of fuel oil during the quarter. A total of 4,295 ships assigned to these routes, four and five respectively.
mt of diesel oil was consumed in the operation of auxiliary
engines at buffers and berths by the fleet. 6 CONCLUSION
Regarding the utilization of the stowage capacity of
container ships, the highest average utilization in the fleet The results for a three-month analysis of the operation of
was experienced on CC 17 and CC 18. These vessels be- the liner shipping network reveals operation costs, transit
long to service route R1, having a registered capacity of time, and container ship utilization consistent with quar-
4,895 TEU. Their average utilization reached over 85% terly reports in liner shipping industry. The results show
with a maximum observed occupancy of 100%. The low- details on the expected performance of a liner shipping net-
est average utilization was 50.1% observed in CC9 and work and service schedule can be analyzed through simula-
CC4. Both container ships belong to service route R3 with tion modeling. Simulation modeling can be used to ana-
capacity of 8,200 TEU. Despite of their low utilization, lyze the effects and contribution of each party when a
CC4 reached 93.2% of its total capacity utilization while global alliance and cooperation between firms is proposed.
CC9 reached 89.5% of its total capacity at some point dur- Likewise, it can be used to determine areas were the per-
ing the evaluation period. formance is above and below the expected level of opera-
Results on the logistic performance also show an aver- tions in shipping.
age container transit time of 23.38 days. This transit time The next step in this research would include optimiza-
includes both direct and transshipment time of 23.16 days tion of the overall network performance by changes in in-
and 51.28 days, respectively. The longest time observed dividual service routes and support for refueling plan anal-
during a direct shipment was 67.50 days. Likewise, the ysis for container ship in service.
longest time observed for transshipment was 81.50 days. This research contributes to the knowledge base in lin-
Containers spent an average of 5.06 days in container yards er shipping by proposing a modeling approach where the
at ports. During the evaluation, 43.89 days was observed dynamic operation of service routes, deployment, and
for some containers. On average, containers spent 12.07 scheduling decision in liner shipping are explored under a
days for transit at sea. A maximum high of 25.24 days at realistic container transit scenario.
sea was observed in this period.
The port of Bremerhaven accounts for 4,743.00 TEU, 7 ACKNOWLEGMENTS
and the port of Norfolk handled 2,145 TEU, both as trans-
shipment ports.
The authors thank the Department of Industrial Engineer-
ing and the Department of Geography of the University of
5.2 Assessment of Individual Service Routes Louisville for their logistic support as well as the liner
shipping firms interviewed during the research process.
Table 3 summarizes details of the results and overall op-
eration of service routes (R1, R2, R3, R4) in three aspects:
cost, productivity, and performance. Service schedule R1
handled 22% of cargo throughput consuming only 11% of
the cost in the system. Service schedule R2 is yield to 33%

2583
 McLean and Biles

REFERENCES Medeiros, E. F. Watson, J. S. Carson, and M. S.

Manivannan, 1219-1221.
Baird, A. 2002. The Economics of Container Transship- Van Rensburg, J. J., J. He, and A. J. Kleywegt. 2005. A
ment in Northern Europe. International Journal of computer simulation model of container movement
Maritime Economics, 4, 249-280. by Sea. Proceedings of the 2005 Winter Simulation
Bendall, H. B., and A. F. Stent. 1999. Longhaul feeder Conference, ed. M. E. Kuhl, N. M. Steiger, F. B.
services in an era of changing technology: An Asia- Armstrong, and J. A. Joines, 1559-1566.
Pacific perspective. Maritime Policy & Manage-
ment, 26(2), 145-159. AUTHOR BIOGRAPHIES
Bendall, H. B., and A. F. Stent. 2001. A Scheduling
model for a High Speed Container Service: A Hub ALDO McLEAN is an Instructor in the Department of In-
and Spoke Short-Sea Application. International dustrial Engineering at the University of Louisville. He
Journal of Maritime Economics, 3(3), 262-277. spent eight years working in consumer electronics and tel-
Bush, A., W. E. Biles, and G. W. DePuy. 2003. Interative ecommunication industries. He has experience related to
Optimization and Simulation of Barge Traffic on an engineering and management for global firms include LG
Inland Waterway. In Proceedings of the 2003 Win- Electronics, Siemens, NEC, and Panasonic. He developed
ter Simulation Conference, 1751-1756. expertise in logistics, quality improvement, and engineer-
Cullinane, K., and M. Khanna. 1998. Economies of Scale ing management. His current research interests include
in Large Container Ships. Journal of Transporta- transportation logistics, simulation modeling, and opera-
tion Economics and Policy, 33(2), 185-208. tion management. He earned is Ph.D. in Industrial Engi-
Eljardt, G. 2006. Development of a Fuel Oil Consumption neering in 2008. His e-mail address is
Monitoring System. Ph.D. thesis, Technische Uni- [email protected].
vesitt Hamburg-Harburg.
Goldsman, D., S. Pernet, and K. Kang. 2002. Simulation WILLIAM E. BILES is the Clark Chair of Computer
of Transportation Logistics. In Proceedings of the Aided Engineering in the Department of Industrial Engi-
2002 Winter Simulation Conference, ed. E. Yuce- neering of the University of Louisville. He spent two years
san, C.H. Chen, J. L. Snowdown, and J. M. Charnes, in the US Army and seven years in industrial R&D before
901-904. undertaking an academic career. He has held academic
Haralambides, H. E., and A. W. Veenstra. 2000. Modeling appointments at the University of Notre Dame, Penn State,
performance in liner shipping. In K.J. Button and Louisiana State University, and the University of Louis-
D.A. Hensher: Handbook of Transport modeling. ville. Dr. Biles received the Distinguished Engineering
Pergamon-Elsevier Science. Alumnus award from the University of Alabama in Hunts-
Kelton, W. D., R. P. Sadowski, and D. T. Sturrock, 2007 ville in 2004. His most recent publication is Environmen-
Simulation with Arena, 4th Edition, McGraw-Hill, tally Benign Manufacturing Chapter 1 in the 2007 John
New York Wiley handbook for Environmentally Conscious Manufac-
Lai, K. K., K. Lam, and W. K. Chan. 1995. Shipping con- turing. His e-mail address is
tainers logistics and allocation. Journal of the Op- [email protected].
erational Research Society, 46, 687-697.
Law, A. M., and W. D. Kelton. 2000. Simulation Modeling
and Analysis. Third Edition. McGraw-Hill.
Shintani, K., A. Imai, E. Nishimura, and S. Papadimitriou.
2007. The Container Shipping Network Design
Problem with Empty Container Repositioning.
Transportation Research: Part E, 43(1).
Song, D., J. Zhang, J. Carter, T. Field, J. Marshall, J. Po-
lak, K. Schumecher, P. Shina-Ray, and J. Woods.
2005. On Cost-Efficiency of the global container
shipping network. Maritime Policy and Manage-
ment, 32(1), 15-30.
Stopford, M. 1997. Maritime Economics. Chesham, Buck-
inghamshire: Pointing-Green Publishing Services.
Swedish, J. A. 1998. Simulation of an Inland Waterway
Barge Fleet Distribution Network. In Proceedings
of the 1998 Winter Simulation Conference, ed. D. J.

2584