420 Int. J. Logistics Systems and Management, Vol. 10, No.

4, 2011

A system dynamics approach to operational

and strategic planning of a container terminal

Jack Kie Cheng*

Division of Physical Sciences,
College of Arts and Sciences,
Universiti Utara Malaysia,
06010 Sintok, Kedah, Malaysia
E-mail: jackkiecheng@yahoo.com
*Corresponding author

Razman Mat Tahar

Department of Technology Management,
Universiti Malaysia Pahang,
Lebuhraya Tun Razak,
26300 Kuantan, Pahang, Malaysia
E-mail: razman779@ump.edu.my

Chooi-Leng Ang
Division of Physical Sciences,
College of Arts and Sciences,
Universiti Utara Malaysia,
06010 Sintok, Kedah, Malaysia
E-mail: ang@uum.edu.my

Abstract: Modern container terminal faces the pressure of providing both

efficient services and adequate facilities parallel with the demanding needs
from the customers. This paper presents the application of system dynamics
simulation in addressing both the issues at the operational and strategic level.
The operational level model is used to aid terminal operators on daily planning
by understanding the relationship and interdependency between the berth and
yard. The strategic level model on the other hand is used for capacity planning.
This study bridged the gap between the literatures by integrating both
operational and strategic level issues in a system dynamics model.

Keywords: container terminal; system dynamics; simulation; Microworlds.

Reference to this paper should be made as follows: Cheng, J.K., Tahar, R.M.
and Ang, C-L. (2011) ‘A system dynamics approach to operational and
strategic planning of a container terminal’, Int. J. Logistics Systems and
Management, Vol. 10, No. 4, pp.420–436.

Biographical notes: Jack Kie Cheng is a PhD Student at the College of Arts
and Sciences, Universiti Utara Malaysia, Malaysia. Her area of research
includes container terminal, logistics, simulation and system dynamics.

Copyright © 2011 Inderscience Enterprises Ltd.

 A system dynamics approach 421

Razman Mat Tahar is a Professor at the Department of Technology

Management, Universiti Malaysia Pahang, Malaysia. Besides serving as the
Deputy Dean of the Department of Technology Management, he has 13 years
of teaching and training experiences, including industrial exposure as well as
consultancy with government agencies and NGOs. His research interests
include the application of simulation and system dynamics modelling in
numerous industries such as logistics and supply chain management.

Chooi-Leng Ang is an Associate Professor at the College of Arts and Sciences,

Universiti Utara Malaysia, Malaysia. Currently, she is the Programme
Coordinator at the Executive Development Centre. She is an Active
Researcher, having authored numerous academic papers for journals and
conference proceedings. Her research interests include Information and
Communication Technology (ICT) applications in business, decision-support
science, quantitative analysis and modelling.

1 Introduction

With the positive growth of containerised shipping, the competition between container
terminals especially the geographically near ones is getting more intense. To attract
liner operators, container terminals compete to deliver the best services, providing
state-of-the-art facilities. Liner operators, on the other hand, accelerate the introduction of
larger vessels. This is partly due to the ability of larger vessels in carrying more
containers and also to achieve economy of scale (Parola and Sciomanchen, 2009).
The terminal operator faces several challenges to attract and serve these
large vessels (Notteboom and Rodrigue, 2009). First of all, there is the need for more
facilities. Larger vessels require deeper draft, more equipment to transfer the containers
and larger container yard for temporary container storage. However, constructing new
infrastructures are costly and there is the risk of overcapacity if done without proper
planning and understanding on future demand. Moreover, these large vessels do not
necessarily guarantee a visit even if the container terminal is well equipped with the best
facilities.
Besides that, there is the challenge to serve these mega vessels within the shortest
time period as liner operators demand for quick vessel turnaround time to save on the
terminal costs (Imai et al., 2007). The internal operation of the container terminal is
complex (Alessandria et al., 2009; Monaco et al., 2009) as the terminal consists
of many individual subsystems such as the berth and the yard (Zhang and Jiang, 2008).
The efficiency of the overall container terminal performances depends on the efficiency
of the berth and yard operation. However, due to complexity, decision-making and
planning of the berth and yard are executed separately. As the individual subsystems
organised within its own operating constraints, terminal operators fail to see the impact of
the decision made in one subsystem to the other subsystems as well as to the whole
container terminal performance.
To serve these mega vessels and remain competitive, the container terminal needs to
constantly ensure that the terminal is able to deliver efficient services (Sciomachen et al.,
2009) and at the same time providing necessary facilities parallel (Kozan and Casey,
2007) with current and future demand. Container terminals that fail to fulfil these two
requirements will be eliminated from the shipping business and to the point of extinction.
422 J.K. Cheng et al.

Therefore, there is a need for a tool that can aid the terminal operator in enhancing
decision-making and planning at both operational and strategic levels.
This study presents the strength of system dynamics modelling in capturing both
the complexity of the container terminal operation along with the dynamics of its external
environment. The modelling aspect of this study aims at enhancing the understanding
of the container terminal operation, thus guiding the terminal operator towards a better
operational and strategic-level planning and decision-making. The first objective of this
study is to model the flow of containers across the berth side and the yard side to
identify potential bottleneck and to measure performance indicators. Four types of
performance indicators are measured; they are the vessels performance indicators,
berth performance indicators, container handling performance indicators and container
yard performance indicators. By measuring these performance indicators, the
performance of the container terminal can be analysed, thus potential bottleneck can be
identified.
The second objective of this study is to evaluate the relationship and interdependency
between the berth and yard to determine how it affects the terminal’s overall efficiency.
By understanding the relationship and interdependency between subsystems, the cause
of the bottleneck can be identified, thus relevant solution can be implemented. Next, to
ensure that the container terminal has sufficient capacity to sustain demand and yet avoid
overcapacity, the third objective of this study is to develop a capacity planning. Finally, a
Microworlds is developed to serve as a learning laboratory. By navigating through the
Microworlds, terminal operator can improve their understanding on the container
terminal performance as well as experimenting different strategies and observed the
impact to the terminal before implementing it into the real world.
The system dynamics model developed in this study consists of an
operational-level model, which is to achieve the first two objectives of this study, and a
strategic-level model, which is for capacity planning (third objective). These two models,
however, are interrelated and the terminal operator can observe the impact of the decision
made at the operational level to the strategic level. This study is conducted at one
of the container terminals in Malaysia. Data and information were gathered during a
two-week visit to the case study container terminal, from 21 January 2008 to 1 February
2008.
The remaining of this paper presents the literature reviews related to this study,
followed by the overview of the container terminal operation. Next, the development
of system dynamics model and the model validation procedure are presented.
The analysis of outputs generated from both operational- and strategic-level models is
discussed subsequently. This paper ends with a short conclusion.

2 Literature review

There are many studies conducted at container terminals, but many of these studies focus
either only on the berth side (e.g., Bierwirth and Meusel, 2009; Dragovic et al., 2005;
Imai et al., 2007; Monaco and Sammarra, 2007; Moorthy and Teo, 2006; Razman and
Khalid, 2000) or only on the yard side (e.g., Chen, 1999; Chen et al., 2004; Han et al.,
2008; Kim and Kim, 2002; Kim and Park, 2003; Kim et al., 2008; Legato et al., 2009;
Ma and Hadjiconstantinou, 2008; Zhang et al., 2003). The complexity and dynamic
nature of the container terminal operations resulted in difficulties in using analytical
 A system dynamics approach 423

tools as the method of investigation. Analytical method offers no means to capture

interdependency (Warren and Langley, 1999) and does not allow for a dynamic approach
to strategic and operational decisions as they may lack flexibility in formulation or
the algorithms may be time consuming for application (Grasman, 2006). Most of
the conventional tools intended to optimise individual subsystems fail to anticipate the
performance of the overall system, and these tools are used in such a way that one tool’s
results feeds into another tool’s assumptions (Homer et al., 1999).
As such, Discrete Event Simulation (DES) is widely used to model the whole
container terminal (Ma and Hadjiconstantinou, 2008). Simulation models are more
flexible and comprehensive than optimisation techniques, and also possess the capacity
to handle temporal dimensions and probabilistic elements of logistics problems
(Moynihan and Padmanabhan, 2006). A simulation model can also represent all the
features of a real container terminal and may be very useful to test different
scenarios or to compare alternative solutions to be applied to the terminal (Alessandri
et al., 2007).
There are many studies that integrate both berth and yard operation using DES
(Alattar et al., 2006; Kozan and Casey, 2007; Ramani, 1996; Veeke and Ottjes, 2002;
Yun and Choi, 1999). DES works well for issues concerned with transaction, processing
and the flow of individual entities through a system (Koelling and Schwandt,
2005). However, DES is not capable in capturing the relationship and interdependency
between the berth and yard operations to investigate how these subsystems affect
each other as well as the overall container terminal performance. System dynamics, on
the other hand, is capable of capturing the relationship and feedbacks between
variables as well as providing platform for testing the impact of policy and strategy
decision-making.
System dynamics is a methodology for analysing complex systems and problems
with the aid of computer modelling and simulation software. System dynamics approach
originated from the research of Professor Jay W. Forrester at Massachusetts Institute of
Technology in the late 1950s. System dynamics has long been used to help management
team to formulate strategy and improve individual, team and organisation learning
(Warren and Langley, 1999). The aim of system dynamics is to provide an understanding
of the modes of behaviour (basic trends). Thus, system dynamics models tend to be more
holistic, with aggregate flows (Taylor and Dangerfield, 2005).
Another major advantage of system dynamics is through the development of
microworld or management flight simulator. The study by Moynihan et al. (2006)
developed a decision-support system, which allows managers to analyse different
scenarios relevant to the changing environment using the ‘what–if’ engine. Similarly, the
Microworlds developed in this study also provides a feasible experimental environment,
as exploring the effect of policy changes and experimenting with alternative policy
formulation is not feasible in the real world. Microworld offers the opportunity for role
playing and role switching to explore differences in critical assumptions. It also helps
players to cope with uncertainty, complexity and ambiguity.
Nevertheless, the application of system dynamics simulation modelling is uncommon
in container terminal study as opposed to other similarly complex industries such as
healthcare (Brailsford, 2008; Brailsford et al., 2004; Gonzalez-Busto and Garcia, 1999;
Lane et al., 2000; Taylor and Dangerfield, 2005), transportation management (Cassettari
et al., 2010; Homer et al., 1999; Liehr et al., 2001; Mayo et al., 2001) and supply chain
(Briano et al., 2010; Janamanchi and Burns, 2007). Up to date, there are only two related
424 J.K. Cheng et al.

works (Choi et al., 2007; Soemodihardjo, 2000) reported on the application of system
dynamics in solving issues of the container terminal. The difference between these two
studies is that Choi et al. (2007) focused on solving issues at the operational level while
Soemodihardjo (2000), on the other hand, focused on issues at the strategic level. These
two studies also did not incorporate the interdependency and relationship between
various subsystems at the container terminal.
This study contributes at bridging the gap between the literatures in two ways.
The first one is by developing a system dynamics model that is capable of capturing the
relationship and interdependency between the berth and yard operations as well as how
the operations of these two subsystems impact the overall container terminal. The second
contribution is that the model of this study focused on solving issues at both the
operational level and the strategic level. The operational-level model or the micro-level
model provides the understanding on the current terminal operation through the analysis
of the relationship between the berth and yard operations while the strategic-level model
or the macro-level model provides insight into the dynamics behaviour of the container
terminal in long run.

3 Overview of container terminal operation

A container terminal is the place where vessels dock on a berth, and containers are loaded
and unloaded (Lee et al., 2006). A container terminal can ideally be divided into two
areas, the berth side for berthing vessels and the yard side for storing containers
(Lee et al., 2006; Vacca et al., 2007). The berth side consists of several berths for vessel
to moor. Typically, the berth is linear in shape. On the other hand, the yard side consists
of several yard blocks where it serves as the storage area for import, export and
transshipment containers.
The transshipment container activity is slightly different from import and export
container activities. According to Vacca et al. (2007), recently, the container transport
tends to develop towards single-mode transportation, which is called transshipment,
where containers are exchanged between vessels commonly referred as mother vessels
and feeders.
When vessels arrive at a container terminal, it will be assigned to an empty berth,
waiting for the quay cranes to unload or discharge the import containers from the vessel
and transfer it to the terminal. Under these quay cranes, prime movers are waiting to
receive the containers from the quay cranes. These prime movers will then transfer these
containers to the container yard for temporary storage and eventually being loaded onto
other vessels. Therefore, most of the transshipment containers activities revolve around
the berth area as well as the yard area and seldom make it to the gate area. This study
focuses on the transshipment container activity, thus only operations at the berth
subsystem and the yard subsystem are modelled.

4 Model development

The system dynamics model in this study is developed using iThink® simulation
package. There are two types of models being developed. The first model aids the
operational-level planning whereas the second aids the strategic-level planning. However,
 A system dynamics approach 425

both models are interrelated and dependent on each other. Because of the complexity
of the model, only the simplified stock and flow diagram for the berth allocation model
(for both operational-level and strategic-level models) is presented in this paper
(the actual model also consists of the stock and flow diagrams for container unloading
and loading process as well as container yard operations). Figure 1 shows the simplified
stock and flow diagram of the operational-level berth allocation model whereas Figure 2
demonstrates the simplified stock and flow diagram of strategic-level berth allocation
model.

Figure 1 Operational level of berth allocation model (see online version for colours)

Figure 2 Strategic-level model of berth allocation model (see online version for colours)
426 J.K. Cheng et al.

There are two types of vessels arriving at the container terminal; the first type is
mainlines whereas the second is feeders. Figure 1 captures the arrival of these two types
of vessels. The mainline is given a higher priority to berth when compared with the
feeder if both arrived within the same time frame. Upon arrival, the vessel’s length will
be checked against the available berth space. If there are no vacant berths, vessels have
to queue for available berths. As the vessel enters the berth, containers unloading and
loading processes begin. Containers that are being unloaded from the vessel via quay
crane are transferred to the yard blocks via prime movers. Upon completion of container
unloading process, the container loading process starts. Again, these prime movers are
used to transport containers from the yard block to the berth, to be loaded onto the vessel
via quay cranes. After completing the container unloading and loading process, the vessel
will depart from the berth. The rate of the vessel departing from the container terminal is
dependent on the duration spend to service the vessel.
However, in the strategic-level model (Figure 2), mainlines and feeders are combined
when compared with the operational-level model, which separates these two types of
vessels. The second thing is that the strategic-level model runs by years as opposed to the
operational-level model, which runs by day. Apart from these two differences, berth
allocation model for strategic-level model runs with the same logic as the berth allocation
model for operational-level model.

5 Model validation

For model validation purposes, the outputs generated by the operational-level model were
compared with the actual data. Five data sets were compared; they are the number of
vessels arrived, the number of containers unloaded at each berth, the number of
containers loaded at each berth, the berth occupancy rate and the yard occupancy rate.
The paired sample t-test of the five independent samples showed that these five pairs
of data sets are statistically not different at 95% significance level.
The strategic-level model, on the other hand, is validated by comparing its output
with the output of the operational-level model. The operational-level model is run for
one year (representing year 2008) and the total container throughput for the year 2008 is
generated. The total container throughput for the year 2008 is then compared with the
data input (container throughput) of the strategic-level model (retrieved from the website
of the Ministry of Transportation, Malaysia) of the same year. The paired sample t-test
of these two independent samples shows they are not statistically different at 95%
significance level.
Therefore, by validating both the developed operational- and strategic-level models,
the system dynamics model is deemed fit to represent the actual system and can be used
for further analysis.

6 Analysis and results

The operational-level model is used to generate performance indicators. These

performance indicators are analysed as they revealed a lot regarding the current
performance of the container terminal. The performance indicators are analysed to
identify bottleneck and to understand the relationship and interdependency between the
 A system dynamics approach 427

berth and yard operations. The strategic-level model, on the other hand, is used for
capacity planning. It also served as a guideline for determining when additional facilities
should be introduced, parallel with the future demand.

6.1 Generating performance indicators

Four categories of performance indicators are generated from the model; they are vessels
performance indicators, berth performance indicators, container handling performance
indicators and container yard performance indicators. Table 1 shows the output of the
performance indicators generated from the operational-level system dynamics model.
The average vessel turnaround time is 14.97 h while both the average vessel service time
and average vessel waiting time are 12.77 h and 2.20 h, respectively. The total containers
handled are 5,293,560 TEUs while the berth and yard occupancy rate are 70% and
24.9%, respectively.

Table 1 Performance indicators generated from the system dynamics model

Performance indicators Output

Vessels performance indicators
1 Average turnaround time 14.97 h per vessel
2 Average service time 12.77 h per vessel
3 Average waiting time 2.20 h per vessel
Berth performance indicators
1 Berth occupancy rate 70%
Container handling performance indicators
1 Total containers handled 5, 293, 560 TEUs
Container yard performance indicators
1 Container yard occupancy 24.9%

6.2 Discussion on berth occupancy rate

High value for berth occupancy shows that either the terminal is congested or there is
inefficient utilisation of the terminal equipment, whereas a low berth occupancy rate
indicates that only few vessels have arrived and there is underutilisation of terminal
resources (Ramani, 1996). The maximum internationally acceptable standard for berth
occupancy rate according to Ray and Blankfeld (2002) is 40%. Therefore, the berth
occupancy rate of 70%, generated from the operational-level model, appeared to be high.
The yard occupancy rate is not discussed as the value of 24.9% appears to be low and it is
also the strategy of the case study container terminal to maintain a large yard capacity for
future container growth.
To determine how to lower the berth occupancy rate, the relationship of berth
occupancy rate with the other performance indicators such as vessel turnaround time,
vessel service time and vessel waiting time was investigated.
The berth occupancy rate is influenced by the vessel turnaround time, which in turn is
influenced by the vessel service time. Vessel waiting time is influenced by the berth
occupancy rate as the higher the berth occupancy rate, the longer the vessels have to wait
428 J.K. Cheng et al.

to be berthed. Vessel service time, on the other hand, is dependent on the speed of
container unloading and loading process while this process is influenced by quay cranes
moves and prime mover travelling distances. Therefore, quay crane moves and prime
mover travelling distances have impact on the berth occupancy rate. Figure 3 exhibits the
discussed relationships. The subsequent section focuses on how adjusting quay crane
moves and analysing the prime mover travelling distances can help in lowering the berth
occupancy rate of 70%.

Figure 3 Relationship and interdependency of berth occupancy rate with other elements in the
case study container terminal

6.3 Adjustment in quay crane moves

Quay cranes are used to unload containers from a vessel and load containers onto a
vessel. The quay cranes at the case study container terminal performed an average of
33 container moves per hour, which means in one hour, the quay cranes transferred
33 containers. Thirty-three container moves per hour appear to be quite low as Chou et al.
(2003) state that both Hong Kong Port and Kaohsiung Port, ranked top 10 in the world
leading ports, manage 40 container moves per hour and 35 container moves per hour,
respectively.
Therefore, by increasing the quay cranes productivity of 33 container moves per hour
to 40 container moves per hour, there is a sharp decrease in vessel turnaround time and
berth occupancy rate. Table 2 shows the adjustment of quay cranes moves and its impact
on the berth occupancy rate and vessel turnaround time. Notice that even with the
maximum 40 container moves per hours, the berth occupancy rate is 60% (higher than
the maximum international standard of 40%). This shows that additional capacity or new
berths need to be introduced as even the maximum quay crane moves fail to lower the
berth occupancy rate to below 40%.
 A system dynamics approach 429

Table 2 The impact of the adjustment in quay crane moves on average vessel turnaround time
and berth occupancy rate

Quay cranes moves per hour Average vessel turnaround time Berth occupancy rate (%)
33 14.98 h 70
34 14.62 h 68
35 14.28 h 67
36 13.96 h 65
37 13.66 h 64
38 13.37 h 62
39 13.09 h 61
40 13.83 h 60

6.4 Analysis on prime mover travelling distance

The function of prime mover is to transfer containers between the berth and blocks.
As revealed earlier, the travelling distances of the prime movers do play a role in
influencing vessel service time. Therefore, to minimise the travelling distance
of the prime movers between the yard blocks and the berth, it is better to store containers
at yard blocks, which are situated near to its loading point at the berth. There are a total
of 12 yard blocks at the case study container terminal. Using the operational-level model,
the travelling distance of prime movers was analysed.
Basically, from the analysis, two types of position are determined. The first type is
the favourable positions where most of the containers stored in these blocks situated near
to their loading position at berth, therefore the travelling distance of prime movers is
minimised. The second type is the unfavourable position where containers stored in these
blocks travel more to their loading position at the berth due to the further distance
between the yard block and the loading point at berths. Table 3 demonstrates the
percentage of containers exiting all 12 yard blocks to favourable loading point at berth
and to unfavourable loading point at berth.

Table 3 Percentage of containers at favourable position and unfavourable position for loading
point at berth

Percentage of containers exiting to Percentage of containers exiting to

Yard block favourable loading point at berth (%) unfavourable loading point at berth (%)
Yard Block 1 75.75 24.25
Yard Block 2 83.53 16.47
Yard Block 3 84.2 15.8
Yard Block 4 94.21 5.79
Yard Block 5 91.03 8.97
Yard Block 6 70.66 29.34
Yard Block 7 83.38 16.62
Yard Block 8 92.32 7.68
Yard Block 9 87.05 12.95
430 J.K. Cheng et al.

Table 3 Percentage of containers at favourable position and unfavourable position for loading
point at berth (continued)

Percentage of containers exiting to Percentage of containers exiting to

Yard block favourable loading point at berth (%) unfavourable loading point at berth (%)
Yard Block 10 90.67 9.32
Yard Block 11 95.03 4.98
Yard Block 12 88.45 11.55

Although majority of the containers stored at each of the 12 yard blocks are loaded at its
favourable loading points, there are containers stored in these blocks being loaded at
unfavourable loading point. Especially for yard block 1 and yard block 6 where the
percentage of containers being loaded to further berths are the highest among all 12 yard
blocks with 24.25% and 29.34%, respectively. This increases the prime mover travelling
distances and its time spent to move containers from yard to berth thus increases the
vessel service time and berth occupancy rate.

6.5 Analysis on capacity planning

With the trend of introducing larger vessels, the case study container terminal needs to be
prepared with adequate berth capacity and efficient services. This is to ensure that
the terminal has sufficient capacity to sustain demand and yet avoid overcapacity.
The strategic-level model is used to observe the impact of introducing additional
berths (initial: 8 berths) on both berth occupancy and capacity rate until the year 2012.
The benchmarks that are being used in this study to indicate when the container terminal
needs additional berths are 60% for berth occupancy rate and 90% for berth capacity rate.
Table 4 demonstrates the comparison of berth occupancy rate for the initial 8, 10 and
12 berths for the years 2007–2012.

Table 4 The comparison of berth occupancy rate for 8, 10 and 12 berths

Berth occupancy rate Berth occupancy rate Berth occupancy rate

Year with 8 berths (%) with 10 berths (%) with 12 berths (%)
2007 57.11 45.69 38.07
2008 58.93 47.14 38.29
2009 62.91 50.32 41.94
2010 68.48 54.78 45.65
2011 72.89 58.31 48.59
2012 77.30 61.84 51.53

With initial eight berths, the berth occupancy rate for the year 2009 is 62.91%.
By adding two berths, the berth occupancy rate decreased to 50.32%. With 10 berths, the
berth occupancy rate reaches 61.84% in the year 2012, signifying the need for extra
berths. By adding two new berths in the year 2012, the berth occupancy rate is reduced to
51.53%.
Table 5, on the other hand, exhibits the comparison of berth capacity rate for the
initial 8 and 10 berths for the years 2007–2012. The total length of the initial eight berths
 A system dynamics approach 431

is 2880 m (one berth: 360 m). In the year 2009, the berth capacity rate is 90.15%
(benchmark: 90%). With additional two berths, the berth capacity rate is reduced to
72.12%. These 10 berths are able to sustain until the year 2012, where the berth capacity
rate of 72.94% is still lower than 90%. Take note that the discussion on the berth capacity
rate so far is based on the average vessel length of 251.662 m. As the average vessel
length varies each year due to the quick growth in vessel length, the discussion so far
is an estimation based on the average vessel length that arrives at the case study
container terminal in the year 2007. The strategic-level model, however, allows users to
increase the average vessel size at any chosen year to see its effect on the overall berth
capacity rate.

Table 5 The comparison of berth capacity rate for 8 and 10 berths

Year Berth capacity rate with 8 berths (%) Berth capacity rate with 10 berths (%)
2007 89.46 71.57
2008 89.77 71.82
2009 90.15 72.12
2010 90.49 72.39
2011 90.82 72.66
2012 91.18 72.94

7 The development of Microworlds

System dynamics simulation also supports the development of Microworlds.

Microworlds or management flight simulator is a virtual world representing business
environment, where in the case of this study, representing the container terminal
operations. The developed Microworlds in this study allows terminal operator to enhance
the understanding on the berth and yard operation. By experimenting different
combination of strategies, the terminal operator can see how the decision made at the one
subsystem can influence the other subsystems as well as the whole container terminal
operation. The Microworlds provides an opportunity for the terminal operator to learn as
they explore, thus justifying the role of Microworlds in organisation learning.
Microworlds developed in this study integrate both the operational-level planning
and the strategic-level planning. The Microworlds developed for the operational-level
model aids users on daily planning and decision-making whereas the strategic-level
model provides users with the ability to understand the future performance of the case
study container terminal. Besides helping the terminal operator to understand the
relationship between subsystems at the terminal, the Microworlds also allow port
managers to see both the immediate and the future impact of their decision on the
container terminal performance.
Figure 4 shows the user interface for the operational planning of the case study
container terminal. The users can make numerous changes to the terminal operational
settings and see the impact on the overall terminal performances. Users may increase
or decrease both the number of vessel arrival and the number of containers unloaded
and loaded to see their impact on both the berth occupancy rate and the yard occupancy
rate. Besides that, users can experiment and learn how increasing or decreasing the
432 J.K. Cheng et al.

number of berths, number of quay cranes assignment and quay cranes moves affect the
terminal operations.

Figure 4 User interface for container terminal operational planning (see online version
for colours)

While the user interface for operational planning allows terminal operator to test how
changes in their daily operation affect the overall terminal operations, the user interface
for strategic planning assists terminal operator in making decisions at the strategic level
(Figure 5). The impact of the trend of vessel arrival and container throughput on berth
capacity and berth occupancy can be observed. Users can experiment with four types of
intervention (intervention in vessel size, number of vessel arrival, number of berths and
total container throughput) and observed their impact on the annual container terminal
performance.

Figure 5 User interface for container terminal strategic planning (see online version for colours)
 A system dynamics approach 433

8 Conclusion

In order for the container terminal to remain competitive in the shipping industry, it is
important to provide both efficient services and adequate facilities. To provide these two
criteria, there is a need to understand the internal operation of the container terminal as
well as the current and future demand of containerised shipping. Therefore, there is a
need for a model that can simultaneously capture the complexity of the container terminal
operation, along with dynamics environment, which the terminal operates. The system
dynamics model developed in this study is capable of assisting terminal operator in
planning and decision-making at the operational and strategic levels.
To the best of our knowledge, there are only two related works (Choi et al., 2007;
Soemodihardjo, 2000) reported on the application of system dynamics in solving
issues of the container terminal where the former focused on solving issues at the
operational level while the latter focused on issues at the strategic level. These two
studies also did not incorporate the interdependency and relationship between various
subsystems at the container terminal. This study, therefore, contributes at developing a
system dynamics model that is capable of capturing the relationship and interdependency
between the berth and yard operations as well as solving issues at both the operational
and strategic levels.
The berth occupancy rate of 70% generated from the operational-level model appears
to be higher than the maximum international standard of 40%. The quay crane moves
and prime mover travelling distance were analysed as these two factors influence
the berth occupancy rate. The strategic-level model, on the other hand, provides capacity
planning through the analysis on both berth occupancy and capacity rate until the year
2012. This strategic-level model is capable in guiding the terminal operator on when
additional capacity will be needed parallel with the growth of container throughput and
vessel size. This study, therefore, provides integration between issues at the operational
level (daily basis) and at the strategic level (yearly basis).
Besides that, the Microworlds developed in this study also serves as a learning tool,
as terminal operator can improve their understanding on how each subsystem
operates and influences each other. Terminal operator can also test new strategies or
policies and observe the immediate impact on the terminal operations. The future
impact of intervention at the daily operational-level model can be observed at the
strategic-level model, thus guiding the terminal operator in making better planning and
decision-making.
The model developed in this study has demonstrated the usefulness of system
dynamics. Further investigation and model development can be conducted to make
the system dynamics model more realistic. In this study, the speed of quay crane
moves is represented by an average value. However, in reality, the speed of the
quay crane moves is actually dependent on the skill of the quay crane drivers. If the
quay cranes are handled by experienced drivers, the quay cranes are most likely
to perform more moves as opposed to the quay cranes handled by inexperienced
drivers. Therefore, further research can incorporate this qualitative aspect when
modelling the container terminal operations using system dynamics. By capturing both
the qualitative and the quantitative aspects, a more realistic container terminal model can
be developed.
434 J.K. Cheng et al.

Acknowledgements

The authors wish to express their gratitude on the constructive and helpful comments
from the anonymous reviewers on the earlier version of the manuscript, which helped
to improve the presentation of the paper considerably.

References
Alattar, M.A., Karkare, B. and Rajhans, N. (2006) ‘Simulation of container queues for port
investment decisions’, Proceedings of the Sixth International Symposium on Operations
Research and its Applications, Xinjiang, China, pp.155–167, Available from: http://www.
aporc.org/LNOR/6/ISORA2006F12.pdf [6 October 2010].
Alessandri, A., Sacone, S. and Siri, S. (2007) ‘Modelling and optimal receding-horizon control of
maritime container terminals’, Journal of Mathematics Model Algorithm, Vol. 6, No. 1,
pp.109–133.
Alessandria, A., Cervellera, C., Cuneo, M., Gaggero, M. and Soncin, G. (2009) ‘Management of
logistics operations in intermodal terminals by using dynamic modelling and nonlinear
programming’, Maritime Economics & Logistics, Vol. 11, No. 1, pp.58–76.
Bierwirth, C. and Meusel, F. (2009) ‘A fast heuristic for quay crane scheduling with interference
constraints’, Journal of Scheduling, Vol. 12, No. 4, pp.345–361.
Brailsford, S.C. (2008) ‘System dynamics: what’s in it for healthcare simulation modelers’,
Proceedings of the 2008 Winter Simulation Conference, pp.1478–1483, Available from ACM
Portal: ACM Digital Library [6 October 2010].
Brailsford, S.C., Lattimer, V.A., Tarnaras, P. and Turnbull, J.C. (2004) ‘Emergency and on-demand
health care: modelling a large complex system’, Journal of the Operational Research Society,
Vol. 55, No. 1, pp.34–42.
Briano, E., Caballini, C., Giribone, P. and Revetria, R. (2010) ‘Using a system dynamics approach
for designing and simulation of short life-cycle products supply chain’, Proceedings of 4th
WSEAS International Conference on Computer Engineering and Applications, pp.143–149.
Available from ACM Portal: ACM Digital Library [6 October 2010].
Cassettari, L., Dagnino, F., Mosca, M. and Revetria, R. (2010) ‘Analysis of a urban route traffic
flow exploiting the system dynamic model’, Proceedings of 4th WSEAS International
Conference on Computer Engineering and Applications, pp.164–169, Available from ACM
Portal: ACM Digital Library [6 October 2010].
Chen, P., Fu, Z., Lim, A. and Rodrigues, B. (2004) ‘Port yard storage optimization’,
IEEE Transactions on Automation Science and Engineering, Vol. 1, No. 1, pp.26–36.
Chen, T. (1999) ‘Yard operations in the container terminal: a study in the ‘unproductive moves’’,
Maritime Policy Management, Vol. 26, No. 1, pp.27–38.
Choi, H.R., Park, B.J., Yoo, D.H. and Kang, M.H. (2007) ‘Development of a model based on
system dynamics to strengthen the competitiveness of a container terminal’, WSEAE
Transactions on Information Science and Application, Vol. 4, No. 5, pp.988–996.
Chou, C.C., Liang, G.S. and Chu, C.W. (2003) ‘Competitiveness analysis of major port in Eastern
Asia’, Journal of the Eastern Asia Society for Transportation Studies, Vol. 5, pp.682–697.
Dragovic, B., Park, N.K., Radmilovic, K. and Maras, V. (2005) ‘Simulation modeling of ship-berth
link with priority service’, Maritime Economics & Logistics, Vol. 7, No. 4, pp.316–335.
Gonzalez-Busto, B. and Garcia, R. (1999) ‘Waiting lists in Spanish public hospitals: A system
dynamics approach’, System Dynamics Review, Vol. 15, No. 3, pp.201–224.
Grasman, S.E. (2006) ‘Dynamic approach to strategic and operational multimodal routing
decisions’, International Journal of Logistics Systems and Management, Vol. 2, No. 1,
pp.96–106.
 A system dynamics approach 435

Han, Y., Lee, L.H., Chew, E.P. and Tan, K.C. (2008) ‘A yard storage strategy for minimizing
traffic congestion in a marine container transshipment hub’, OR Spectrum, Vol. 30, No. 4,
pp.697–720.
Homer, J.B., Keane, T.E., Lukiantseva, N.O. and Bell, D.W. (1999) ‘Evaluating strategies to
improve railroad performance: A system dynamics approach’, Proceedings of the 1999 Winter
Simulation Conference, Vol. 2, pp.1186–1193, Available from IEEE Xplore Digital Library
[6 October 2010].
Imai, A., Zhang, J.T., Nishimura, E. and Papadimitriou, S. (2007) ‘The berth allocation problem
with service time and delay time objectives’, Maritime Economics & Logistics, Vol. 9, No. 4,
pp.269–290.
Janamanchi, B. and Burns, J.R. (2007) ‘Reducing bullwhip oscillation in a supply chain: a system
dynamics model-based study’, Proceedings of the International Journal of Information
Systems and Change Management, pp.350–371, Available from ACM Portal: ACM Digital
Library [6 October 2010].
Kim, K.H. and Kim, H.B. (2002) ‘The optimal sizing of the storage space and handling facilities
for import containers’, Transportation Research Part B, Vol. 36, No. 9, pp.821–835.
Kim, K.H. and Park, K.T. (2003) ‘A note on a dynamic space-allocation method for outbound
containers’, European Journal of Operational Research, Vol. 148, No. 1, pp.92–101.
Kim, K.H., Park, Y.M. and Jin, M.J. (2008) ‘An optimal layout of container yard’, OR Spectrum,
Vol. 30, No. 4, pp.675–695.
Koelling, P. and Schwandt, M. (2005) ‘Health system: a dynamic system-benefits from system
dynamics’, Proceedings of the 2005 Winter Simulation Conference, pp.1321–1327, Available
from IEEE Xplore Digital Library [6 October 2010].
Kozan, E. and Casey, B. (2007) ‘Alternative algorithms for the optimization of a simulation model
of a multimodal container terminal’, Journal of the Operational Research Society, Vol. 58,
No. 9, pp.1203–1213.
Lane, D.C., Monefeldt, C. and Rosenhead, J.V. (2000) ‘Looking in the wrong place for healthcare
improvements: a system dynamics study of an accident and emergency department’, Journal
of the Operational Research Society, Vol. 51, No. 5, pp.518–531.
Lee, L.H., Chew, E.P., Tan, K.C. and Han, Y. (2006) ‘An optimization model for storage yard
management in transhipment hubs’, OR Spectrum, Vol. 28, No. 4, pp.539–561.
Legato, P., Canonaco, P. and Mazza, M.R. (2009) ‘Yard crane management by simulation and
optimisation’, Maritime Economics & Logistics, Vol. 11, No. 1, pp.36–57.
Liehr, M., Grosler, A., Klein, M. and Milling, P.M. (2001) ‘Cycles in the sky: understanding and
managing business cycles in the airline market’, System Dynamics Review, Vol. 17, No. 4,
pp.311–332.
Ma, N.L. and Hadjiconstantinou, E. (2008) ‘Evaluation of operational plans in container terminal
yards using discrete-event simulation’, The OR Society, Vol. 21, No. 4, pp.10–18.
Mayo, D.A., Callaghan, M.J. and Dalton, W.J. (2001) ‘Aiming for restructuring success at London
underground’, System Dynamics Review, Vol. 17, No. 3, pp.261–289.
Monaco, M.F. and Sammarra, M. (2007) ‘The berth allocation problem: a strong formulation
solved by a Lagrangian approach’, Transport Science, Vol. 41, No. 2, pp.265–281.
Monaco, M.F., Moccia, L. and Sammarra, M. (2009) ‘Operations research for the management of a
transhipment container terminal: the Gioia Tauro case’, Maritime Economics and Logistics,
Vol. 11, No. 1, pp.7–35.
Moorthy, R. and Teo, C.P. (2006) ‘Berth management in container terminal: the template design
problem’, OR Spectrum, Vol. 28, No. 4, pp.495–518.
Moynihan, G.P. and Padmanabhan, N. (2006) ‘AISLE: analytical integrated software for a logistics
environment’, International Journal of Logistics Systems and Management, Vol. 2, No. 1,
pp.78–95.
436 J.K. Cheng et al.

Moynihan, G.P., Saxena, P. and Fonseca, D.J. (2006) ‘Development of a decision support system
for procurement operations’, International Journal of Logistics Systems and Management,
Vol. 2, No. 1, pp.1–18.
Notteboom, T. and Rodrigue, J.P. (2009) ‘The future of containerization: perspectives from
maritime and inland freight distribution’, Geojournal, Vol. 74, No. 1, pp.7–22.
Parola, F. and Sciomanchen, A. (2009) ‘Modal split evaluation of a maritime container terminal’,
Maritime Economics & Logistics, Vol. 11, No. 1, pp.77–97.
Ramani, K.V. (1996) ‘An interactive simulation model for the logistics planning of container
operations in seaports’, Simulation, Vol. 66, No. 5, pp.291–300.
Ray, D. and Blankfeld, R. (2002) Reforming Indonesia’s Ports, Technical Report. Robert R.
Nathan Associates, Inc. and Checchi and Co. Available from: http://dec.usaid.gov/index.
cfm?p=search.getSQLResults&CFID=21302394&CFTOKEN=98804034 [6 October 2010].
Razman, M.T. and Khalid, H. (2000) ‘Simulation and analysis for the Kelang container terminal
operations’, Logistics Information Management, Vol. 13, No. 1, pp.14–20.
Sciomachen, A., Acciaro, M. and Liu, M. (2009) ‘Operations research methods in maritime
transport and freight logistics’, Maritime Economics & Logistics, Vol. 11, No. 1, pp.1–6.
Soemodihardjo, S. (2000) ‘Study on the optimization and development of Tanjung Emas container
terminal in Semarang using powersim dynamic model simulation’, Proceedings of 8th
International Conference of the System Dynamics Society, Norway, Available from:
http://www.systemdynamics.org/conferences/2000/PDFs/soemodih.pdf [6 October 2010].
Taylor, K. and Dangerfield, B. (2005) ‘Modeling the feedback effects of reconfiguring health
services’, Journal of the Operational Research Society, Vol. 56, No. 6, pp.659–675.
Vacca, I., Bierlaire, M. and Salani, M. (2007) ‘Optimization at container terminals: status,
trends and perspectives’, Proceedings of the 7th Swiss Transport Research Conference,
Available from: http://infoscience.epfl.ch/record/117161 [6 October 2010].
Veeke, H.P.M. and Ottjes, J.A. (2002) ‘A generic simulation model for systems of container
terminals’, Proceedings of 16th European Simulation Multiconference, pp.581–587, Available
from ACM Portal: ACM Digital Library [6 October 2010].
Warren, K. and Langley, P. (1999) ‘The effective communication of system dynamics to improve
insight and learning in management education’, Journal of the Operational Research Society,
Vol. 50, No. 4, pp.396–404.
Yun, W.Y. and Choi, Y.S. (1999) ‘A simulation model for container-terminal operation analysis
using an object-oriented approach’, International Journal of Production Economic, Vol. 59,
No. 1, pp.221–230.
Zhang, C., Liu, J., Wan, Y.W., Murty, K.G. and Linn, R.J. (2003) ‘Storage space allocation in
container terminals’, Transportation Research Part B, Vol. 37, No. 10, pp.883–903.
Zhang, H.L. and Jiang, Z.B. (2008) ‘Simulation studies of heuristic approaches for dynamic
scheduling of container terminal operations’, International Journal of Modelling and
Simulation, Vol. 28, No. 4, pp.410–422.