Beruflich Dokumente
Kultur Dokumente
1. Introduction
Dry bulk terminals are essential nodes in the supply chains for
coal and iron ore. These bulk materials are used for the worldwide production of energy and steel. To facilitate the expected
growing cargo ows, new dry bulk terminals will be built or
existing ones will be expanded. In the supply chains for these
materials, bulk ships and cargo trains are generally used for
transport. The terminal operation is complex when both ships and
trains have to be served at the same time to meet predened
agreements (Robinson, 2007). This research focuses on the
transport network at the terminals. Such networks have to
facilitate all required transportation needs linking several sources
and destinations and consist of belt conveyors and transfer points.
In a transfer point, the material ow is transferred between
different belt conveyors.
In this paper simulation is applied to determine the parameters
that affect the design of belt conveyor networks and to assess
such designs. In section 2, a literature review is presented about
dry bulk terminal design and in particular the network design.
The simulation model developed is introduced in Section 3.
In Section 4, the impact of terminal parameters on the network
design (like the network connectivity, the storage policy and the
redundancy of stockyard machines) is investigated. Section 5
demonstrates the integration of simulation for a belt conveyor
network design. Finally, conclusions are presented in Section 6.
2. Literature review
In Section 2.1, a literature review is presented for the design of
dry bulk terminals and in particular for belt conveyor networks.
*Correspondence: T van Vianen, Delft University of Technology, Department
of Marine and Transport Technology, Mekelweg 2, 2628 CD, Delft,
The Netherlands.
E-mail: T.A.vanVianen@tudelft.nl
Journal of Simulation
Table 1 Review of references that applied simulation-integrated design of dry bulk terminals
Author(s)
Year Design
Application
Baunach et al
El Sheikh et al
Park and Noh
Kondratowicz
King et al
Weiss et al
Dahal et al
Sanchez et al
Ottjes et al
Lodewijks et al
Boschert and Hellmuth
Cassettari et al
1985
1987
1987
1990
1993
1999
2003
2005
2007
2009
2010
2011
while taking the stochastic arrival processes, equipment breakdown behaviour and material ows into account. Furthermore,
dedicated transports have to be performed at the same time using
a limited number of belt conveyors. A simulation model will be
developed to consider the stochastic processes mentioned.
By varying characteristics in belt conveyor networks and by
registering the corresponding performances, relevant insight will
be acquired to design such networks.
3. Simulation model
This section introduces the simulation model that was developed
for the design of belt conveyor networks. The advantage of this
model is that not only the stochastic processes are considered but
also specic terminal operational procedures like the storage
policy and particular network characteristics are taken into
account. The approach followed is mentioned in Section 3.1.
Specic details of the simulation model are presented in Sections
3.2 and 3.3; the verication of this model is discussed.
Ship
generator
Train
generator
Interarrival time
distribution
L1
Shipload
distribution
L2
Pile
Storage time
distribution
SR1
SR2
L3
SR3
L4
Stackerreclaimer
Belt conveyor
Cargo train
Bulk ship
(un)loader
Transfer station
Pile
Stockyard
lane
Control signals
3.3. Verication
Verication of the simulation model is required to check the
correct translation of the conceptual model into computer code
and to determine if the simulation model performs as intended.
Simulation results for a simplied network (as shown in
Figure 2a) were compared with analytical results using queuing
theory. For the incoming (Qin) as well as the outgoing material
Journal of Simulation
Qin
Qin+ Qout
A
A
SR1
SR2
SR1
B
M/D/1
C
E
Qin+ Qout
D
Qout
SR2
M/D/1
Figure 2 Verication of the simulation model for a simplied network layout (a), which can be represented by two individual
M/D/1-queuing systems (b).
1.6
1.4
1.2
Wt [1/]
1.0
0.8
0.6
0.4
M/D/1
0.2
Simulation Results
0.0
0.3
0.4
0.5
0.6
0.7
0.8
0.9
sr [-]
Figure 5 shows the sum of the average ship and train port times
versus the annual throughput (Q) for both layouts as shown in
Figure 4. Another layout with a fully equipped network ( = 1)
was assessed. This layout is not displayed separately in this paper
but can easily be derived by combining the network conguration
from Figure 6b with the stockyard layout of Figure 4a. From
Figure 5 it can be concluded that stacker-reclaimer redundancy
realizes a larger reduction of the sum of the average ship and train
port times than a fully equipped network.
For the assessment of the ID-storage policy, the network
congurations as shown in Figure 1 and 6 are used. The network
connectivity varies from = 7/12 (Figure 1) until = 1 (Figure 6b).
150
120
Wship + Wtrain [h]
D
E
0
10
12
16
18
20
Figure 5 The sum of the average ship and train port times versus
the annual throughput for the CAM-storage policy.
SR3
SR1
SR2
SR2
SR3
(CAM, =7/12)
14
Q [Mt/y]
60
30
SR1
B
C
90
D
F
B
Figure 4 CAM-storage policy with different grade allocation procedures; (a) each grade is assigned to a single stacker-reclaimer and
(b) each grade is assigned to two stacker-reclaimers.
Value
Parameter
NED
Historical data*
NED
0.97
SR-capacities [kt/h]
Average piles storage time [h]
Average shipload [kt]
Trainload
Value
2.5
500
101
4
Journal of Simulation
SR1
SR1
SR2
SR2
SR3
SR3
(ID, =)
(ID, =1)
150
150
120
120
Wship + Wtrain [h]
Figure 6 ID-storage policy applied at two layouts each with different values for the network connectivity ().
90
60
Fig.1, ID, =7/12
30
90
60
30
Fig.6A, ID, =
Fig.6B, ID, =1
0
10
12
14
16
18
20
Q [Mt/y]
10
12
14
16
18
20
Q [Mt/y]
Figure 7 The sum of the average ship and train port time versus
the annual throughput for different network layouts.
SR6
L7
L6
SR5
L5
SR4
L4
SR3
L3
SR2
Figure 10
L2
SR1
L1
QCV1
QCV2 QCV3
Figure 9 The investigated terminal layout with the redesign object (shown with the hatch lled rectangle).
(100, 110, 120 and 130) while one or two of them are active with
reclaiming. Using dedicated belt conveyors reduced the total
number of belt conveyors in the terminal from 51 to 45. Two
extra transfer points are needed to realize all connections resulting
in an increase of the network connectivity () from 0.9 to 0.94.
Advantages of this design are the expected decrease of the
disturbance time, thanks to the reduction of belt conveyors, and
the decrease of the transportation power because the material
does not need to be fed up as frequent as in the existing layout.
For the formulation of the second design (as shown in
Figure 10c) it was allowed that some routes, which are hardly
used simultaneously based on historical operational data, cannot
be performed at the same time anymore. The transport of
materials to the second barge loader (belt conveyor 520 in
Figure 10c) cannot be performed at the same time anymore when
material is transported to the iron ore railcar loader (135) or to the
blending silos (114). This concession was justied by the fact
that at the terminal three barge loaders are installed and loading of
three barges at the same time did hardly happen. Moreover, a
relatively small amount of material (11% of coal) is fed to the
blending silos so the probability is limited that this conicting
situation will occur. Design 2 applies further the fundamentals of
the rst design; all quay conveyors need to be connected to the
cross-conveyors and dedicated belt conveyors are proposed for
the transport to the loading machines. In Design 2 less transfer
points are then needed resulting in a decrease of the network
connectivity () to 0.83.
6. Conclusions
Belt conveyor networks have to facilitate transport activities
to meet the contractual agreements for the terminals seaside
Journal of Simulation
340
220
144
154
120
510
100
111
120
210
110
110
100
210
130
120
712
510
510
220
410
712
131
210
130
111
130
121
111
132
410
712
111
131
230
230
220
30
20
10
135
330
330
230
122
134
1000
135
330
121
340
134
135
133
123
134
240
520
340
113
1001
240
520
520
114
420
240
410
154
420
144
131
114
420
154
136
144
b
114
30
30
20
20
10
10
Belt conveyor
Design 1 (: 0.94)
Design 2 (: 0.83)
Figure 10 Different network congurations for the redesign object; existing layout in 2011 (a) and two designs (bc).
120
105
90
75
Layout 2011 (=0.9)
Design 1 (=0.94)
Design 2 (=0.83)
60
25
30
35
40
Q [Mt/y]
Figure 11 The sum of the average port times for ships and
landside jobs (trains, barges, etc) for the existing layout and new
designs.
References
Altiok T (2000). Tandem queues in bulk port operations. Operations
Research 93(2000): 114.
Andr J et al (2013). Design and dimensioning of hydrogen transmission
pipeline networks. European Journal of Operational Research
229(1): 239251.
Baunach GR, Wibberley ES and Wood BR (1985). Simulation of a coal
transshipment terminal: Batam Island, Indonesia. Mathematics and
Computers in Simulation 27(1985): 115120.
van Beek A (2009). Advanced Engineering Design. Lifetime Performance and Reliability. VSSD: Delft.
Boschert S and Hellmuth T (2010). Simulation in bulk material handling.
Proceedings of the BulkSolids Europe Conference, Scotland.
Bugaric US and Petrovic DB (2007). Increasing the capacity of terminal
for bulk cargo unloading. Simulation and Modelling Practice and
Theory 15(2007): 13661381.
Cassettari L, Mosca R, Revetria R and Rolando F (2011). Sizing of a
3,000,000t bulk cargo port through discrete and stochastic simulation
integrated with response surface methodology techniques. Proceedings of the 11th WSEAS International Conference on Signal Processing, Computational Geometry and Articial Vision, Italy.
Dahal KP, Galloway SJ, Burt GM, McDonald JR and Hopkins I (2003).
A port system simulation facility with an optimization capability.
International Journal of Computational Intelligence and Applications 3(4): 395410.
El Sheikh AAR, Paul RJ, Harding AS and Balmer DW (1987).
A microcomputer-based simulation study of a port. Journal of the
Operational Research Society 38: 673681.
Fishmann GS (2001). Discrete Event SimulationModeling, Programming and Analysis. Springer-Verlag: New York.
Jagerman D and Altiok T (2003). Vessel arrival process and queuing in
marine ports handling bulk materials. Queuing Systems 45(3): 223243.
King DH, Radomske BA and Manocha GS (1993). Recent advances
in simulation models for bulk terminal design. Bulk Solids Handling
13(1): 2327.