Professional Documents
Culture Documents
SEAPORT Case Study
SEAPORT Case Study
SEAPORT Case Study
Cover picture: a) Elements of channel width. P. Siregar (1995), ‘A study on the conceptual design
rules for approach channels’. Graduate thesis, Delft University of Technology,
Department of Civil Engineering, Hydraulic Engineering Group.
b) Artistic impression of the future Taman Seaport. Witteven + Bos (2012), ‘Creation
of Taman Seaport dry cargo area’. Engineering services for RosTranModernizatsiya.
PREFACE
I would like to take the opportunity of this section to thank immensely to all those who have
accompanied me in one or another way, during this ‘Dutch’ experience.
First, thanks to all the members of my committee: Professor Vellinga, who was always open
to answer each of my questions, especially during those moments when the calendar was
becoming tighter and tighter; Peter Quist, who suggested me this research subject some
months ago; Winnie Daamen, who was always willing to check the simulation model and also
gave me many comments to improve the quality of this thesis; and Johan de Boer, from
Witteveen + Bos, who always had time to receive me, answer my e-mails and give me
feedback on my work.
Second, thanks to all the friends I met during my stay in The Netherlands. Many of them were
an important support, particularly during the cold winter season and some ‘changes of
addresses’. Regardless our cultural differences, we learned how to act as a large family,
helping each other in small and large things. I think many of these friendships will remain in
spite of the thousands of kilometers and some oceans that will separate us. To all of them, my
best wishes and success for the future!
Finally, my deepest gratitude and admiration to my wife Lucely and my sons Santiago and
Matías. They are the engine that moves me, and I am completely sure that without their
support and love, I could not have finished my studies in Delft. Also, I know the large sacrifice
that they have done to be with me during these long two years. To them, I dedicate my work
and my life, and I tell them that this experience has strengthened our family. I’m sure that our
future will shine back home… I love you with all my heart!
Sebastián Rayo
September 2013, Delft
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
SUMMARY
The correct operation of a port requires that all of its components need to work coordinated. In this context, the
design of the access channel is fundamental, allowing safe passage for a maximum number of ships.
Nevertheless, when the number of ships increases, congestion occurs. Then, congestion depends on the ships
arrival pattern and the vessels’ characteristics (especially the size and speed).
Two typical types of channels are identified on the PIANC rules (one and two-way channel), which provide a
detailed set of rules for dimensioning of this infrastructure. However, it does not provide an explicit definition
about when to use a one-way or a two-way channel.
In order to provide an assessment tool for the performance of an access channel (existing or in a design stage), a
new simulation model is programmed. This model permits the assessment of different type and widths of access
channel, and it incorporates several functionalities that allow the enhancement the number of ships using it with
an acceptable related congestion.
As an application of the new simulation model, the Taman Seaport’s access channel is assessed. This future
Russian port is in a design stage, and a one-way channel has been selected. Nevertheless, the information on the
projected delays of Taman during its different stages of development has not been calculated. Then, the
utilisation of the simulation model allows defining if the design is appropriate or if excessive delays due to
congestion will occur.
II
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Verbal Model
The verbal model definition is the description of all processes and establishing of the different assumptions to be
considered in the new simulation model.
The main processes related to vessels using an access channel are: obtaining authorisation to navigate, entrance
into the access channel, sail with cruise speed across the outer access channel, speed reduction close to the port
entrance, sailing with reduced speed on inner channel, tugs making fast and the final stop at the moment of
arrival to the turning circle. These processes are related to a ship sailing towards the port: if a ship is sailing
towards the sea, the processes occur in reverse.
The authorisation to enter into the channel depends on several requirements that need to be satisfactory
simultaneously. These requirements are: the access channel must be opened for navigation (tidal windows,
daylight operation, etc.), a sufficient quay length has to be available (only for ships sailing towards the port),
traffic conditions are suitable for a safe passage of a new ship, and the weather conditions are suitable for the
sailing of the new vessel.
Several assumptions are considered for the simulation model. The main assumptions are: the ship movement is
modelled in a one dimensional scheme, the channel is modelled as a straight channel (no bends), and the delay
of a ship is classified according to fourth reasons (port, berth, traffic and weather) in a hierarchical scheme.
Furthermore, some exclussions are considered too: the model does not consider the movement of ships within
the anchorage area and the inner basins, at it does not consider the dynamic status of cranes, storage yards,
stocking piles, etc.
Basic Simulation Model
The basic model allows the modelling of a channel considering all typical options available on other simulation
models (Harboursim), like one or two-way channel, a tidal / current window and a single entrance to the access
channel. The simulation model is programmed in Matlab® as a fast-time type, and it is divided in four stages: set-
up stage (user definition), data generation (model definition), model running (fast-time simulation status) and
output stage.
Several modules are identified on it: input modules (information about a certain port can be defined by the
user), data generation modules (projection of the environment conditions during the simulation time, based on
probabilistic functions and the user’s definition) and output modules (results regarding delay of ships, utilisation
of the anchorage and encountering on different sections of the access channel).
The logical structure of ships to receive authorisation to enter the access channel is based on five call functions:
queue in/out call, port call (port and access channel status), berth call (berth availability), traffic call (congestion)
and pilot call (weather conditions). These call functions represent the different clearances required by a ship to
enter the access channel.
Several ‘manager’ matrixes maintain dynamic information on different components of the simulation model:
anchorage manager (number of ships), queue manager (ships waiting to enter into the access channel), ships
manager (characteristics and relevant information on active ships) and nautical manager (current utilisation of
the access channel).
Because of the absence of real data information, the simulation model is not validated. Nevertheless, several
testing of its main functions (behaviour of ships for different situations) has been executed. Additionally, checks
of the model’s results compared with other accepted sources of information are presented. Comparison with
queuing theory data shows a good agreement between the results. The comparison between the model’s results
and the analysis of the Port of Beira (Mozambique) via Harboursim present large differences for some terminals.
III
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Hence, future stages of this research should put emphasis on validating the model (with real data and other
master plans analysed via a simulation model).
Extended Simulation Model
The extended simulation model corresponds to the implementation of several functionalities that attempt to
enhance the number of ships using the access channel. These measures are classified as traffic measures and
enlargement of the nautical infrastructure.
The traffic measures corresponds to the possibility to operate the access channel in a mixed scheme (one-way
for large carriers, two-way for small vessels), allowance for multiple entrances (ships with small draught can
enter into the access channel via an intermediate entrance) and the possibility to allow speed reductions (faster
ships can adapt their speed when are being preceded by a slow ship).
The enlargement of the nautical infrastructure measures are the mplementation of a passing section (two-way
section on a section of the access channel) and the implementation of an intermediate anchorage site (ships
waiting for a berth to be freed can navigate to a position closer to the port).
From this set of alternatives, the mixed operation and the multiple entrances are good measures to reduce
congestion. Nevertheless, the real effect of these measures depends on the fleet composition (larger number of
small vessels generates a larger reduction of the congestion). The speed reduction option does not make
difference on reducing delays due to traffic.
Analysis of Results
The last part of the development of the simulation model is an analysis of its results. The analysis is divided into
two parts: on a first part, a relationship between average and maximum waiting time and several input
parameters is provided; on a second part, the effect of the available optimisation measures is analysed via some
testing scenarios.
The average and maximum waiting time depends on several parameters: the type of channel, the expected
number of ships, the length of the channel, the capacity of the channel, and the availability of the channel. Four
relationships are defined for average and maximum waiting time of a port with a single terminal and one fleet.
These relations are verified with the use of the simulation model, and the results shows good agreement
between results modelled and obtained via these equations.
The effect on reducing the congestion for each optimisation measure is analysed. The effectiveness of a mixed
operation of the channel and multiple entrances to it depends on the fleet composition. For a larger number of
small vessels, these measures present larger reductions of the congestion. On the other hand, the speed
reduction measure results do not show a reduction of the congestion.
The increment of the bottom width and the implementation of a passing section present progressive reductions
of the waiting time: the larger the modification, the larger the reduction of congestion. Nevertheless, because of
the capital and maintenance costs involved on expanding the nautical infrastructure, a financial analysis is
recommended (comparison between capital and maintenance costs against the benefit of reducing congestion)
for each specific case. Finally, the implementation of an intermediate anchorage does not present a positive
effect on reducing the delay due to congestion.
IV
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
V
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
number of ships per year (maximum of 6,000 ships/year for a one-way channel, and 18,000 ships/year for a two-
way channel). Third, with the purpose to improve the model’s performance, an optimisation of the algorithm is
recommended.
Finally, the assessment of the access channel of the Taman Seaport indicates that for most of the terminals the
delays are acceptable (during the initial and final throughput stages). Additionally, because of the presence of a
large number of small vessels on the Taman’s fleet, the application of a mixed operation of the channel and
multiple entrances are recommended. With these measures, the average congestion is reduced from 20 to 2
minutes. The enlargement of the nautical infrastructure (passing section or bottom width enlargement) are
recommended only if the benefits per reduced hour of congestion are larger than 13,300 €/hr or 30,000 €/hr,
respectively.
VI
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
TABLE OF CONTENTS
SUMMARY II
Simulation Model Development II
Assessment of the Taman Seaport’s Access Channel V
LIST OF FIGURES IX
LIST OF TABLES XII
LIST OF ABBREVIATIONS XV
LIST OF SYMBOLS XV
LIST OF UNITS XVIII
1 INTRODUCTION……………………………………………………………………………………………………………………………………… 1
1.1 Generic problem 1
1.2 Study case – The Taman Seaport 2
1.3 Thesis objectives 5
1.4 Research approach 7
1.5 Report structure 8
2 LITERATURE REVIEW………………………………………………………………………………………………………………………………. 10
2.1 Access channels design 10
2.2 Simulation models of ports operations 18
2.3 Conclusions 22
3 VERBAL MODEL………………………………………………………………………………………………………………………………………. 24
3.1 System boundaries 24
3.2 Processes definition and description 24
3.3 Model’s description 34
3.4 Model’s equations 38
3.4 Conclusions 40
4 BASIC SIMULATION MODEL……………………………………………………………………………………………………………………. 41
4.1 Model’s features 41
4.2 Model’s architecture 42
4.3 Algorithms 54
4.4 Model checks 57
4.5 Model testing 64
4.6 Conclusions 68
5 EXTENDED SIMULATION MODEL…………………………………………………………………………………………………………… 69
5.1 Description of the new functionalities 69
5.2 Modifications of the programming code 71
5.3 Verification of the new functionalities 80
5.4 Conclusions 85
6 ANALYSIS OF RESULTS…………………………………………………………………………………………………………………………….. 87
6.1 Dependence of the delay on input parameter 87
6.2 Effect of the available measures on reducing the delay 99
6.3 Conclusions 111
7 TAMAN SEAPORT CASE – ASSESSMENT OF THE DESIGN…………………………………………………………………………. 114
7.1 Taman Seaport description 114
VII
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
VIII
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
LIST OF FIGURES
1.1 Location of Black Sea relative to the Russian territory 3
1.2 General location of the Taman Seaport 4
1.3 Preliminary access channel (straight) with respect to the Ukrainian border 5
1.5 General methodology and phases of study 8
2.1 Turning radius as function of rudder angle and water depth 11
2.2 Width of a swept track in a turn as function of rudder angle and water depth 16
2.3 Under keel clearance factors 17
2.4 Squat estimation chart for full-bodied ships 17
2.5 Steps in the simulation process 19
2.6 Configuration simulation model Harboursim_Pro 21
3.1 Processes of a ship using an access channel 25
3.2 Annual variations in day length at different locations in the North Hemisphere 27
3.3 Minimum distance between ships in a convoy 30
3.4 Overtaking manoeuvre 31
3.5 Scheme with classification of access channels 31
3.6 Encountering manoeuvre 31
3.7 Bow thrust as a function of tugboat velocity 33
3.8 Relationship between berthing speed, displaced volume and berthing conditions 33
3.9 Speed profile scheme of a vessel using an access channel 36
3.10 Effect of variations of the ship’s speed 38
4.1 Model’s architecture scheme 43
4.2 Algorithm of the main run module 54
4.3 Algorithm of queue in / out call functions 55
4.4 Algorithm of port call function 55
4.5 Algorithm of berth call function 56
4.6 Algorithm of traffic call function 56
4.7 Algorithm of pilot call function 56
4.8 Histogram of inter arrival results of 1,000 ships per year with NED distribution 57
4.9 Histogram of service time results of 1,000 ships per year with k-Erlang distributions 58
4.10 Single ship sailing towards berth; a) position, b) speed 59
4.11 Single ship sailing towards sea; a) position, b) speed 60
4.12 Two ship with opposite directions in a two-way channel 60
4.13 Two ship with opposite directions in a one-way channel 61
4.14 Four ships in a convoy (one-way channel) 62
4.15 Four ships in a convoy, one-slow (one-way channel) 62
4.16 Four ships in a convoy, one-slow (two-way channel) 63
4.17 Vessel’s path with different time steps 64
4.18 Comparison between theoretical and model’s results: a) 1 berth, b) 2 berths 66
4.19 Comparison between Van der Meer (2009) and the simulation model: a) daylight navigation, b) 24
hours navigation 67
5.1 Scheme of ‘mixed’ operability of an access channel 69
5.2 Scheme of multiple entrances points to an access channel 70
5.3 Scheme of speed reduction: a) without speed reduction; b) with speed reduction 70
IX
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
X
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
XI
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
LIST OF TABLES
1.1 Taman Seaport’s terminals 4
2.1 Basic manoeuvring lane 12
2.2 Additional widths Wi for straight channel sections 12
2.3 Additional width for passing distance in two-way traffic 14
2.4 Additional width for bank clearance 14
3.1 Principal semidiurnal constituents and its characteristics 27
3.2 Maximum speed for several ships 29
3.3 Admiralty constant 39
3.4 Typical values of coefficient As 39
3.5 Typical values of coefficient Bs 39
3.6 Typical values of coefficient Cs 40
4.1 Matrix structure for ships generation module 48
4.2 Vector structure for fog generation module 48
4.3 Vector structure for ice generation module 49
4.4 Vector structure for wind generation module 49
4.5 Vector structure for waves generation module 50
4.6 Vector structure for port downtime generation module 50
4.7 Matrix structure for ships manager module 51
4.8 Matrix structure for berth manager module 51
4.9 Matrix structure for nautical manager module 51
4.10 Matrix structure for queue manager module 52
4.11 Matrix structure for ships summary 53
4.12 Matrix structure for encountering summary 53
4.13 Matrix structure for anchorage summary 54
4.14 Inter arrival results of generation of 1,000 ships per year with NED distribution 57
4.15 Service time of 1,000 ships per year with k-Erlang functions of second and third degree 58
4.16 Ships characteristics, convoy of four ships sailing towards port, one-way channel 61
4.17 Characteristics of the scenario solved via queuing theory 65
4.18 Characteristics of the scenario solved via queuing theory (M/E 2/1) 65
4.19 Characteristics of the scenario solved via queuing theory (M/E 2/2) 65
4.20 Fleet characteristics of Port of Beira 67
4.21 Fleet service times of Port of Beira 67
5.1 Modification of matrix structure for ships manager module 72
5.2 Modification of matrix structure for ships manager module 77
5.3 Matrix structure for queue manager 77
5.4 Matrix structure for anchorage summary 79
6.1 Testing simulation characteristics 98
6.2 Scenarios for testing allowance of ‘mixed’ channels capacity 100
6.3 Vessels’ characteristics 100
6.4 Channel’s characteristics 100
6.5 Summary of encountering for different vessels 100
6.6 Entrance points characteristics 102
6.7 Vessels’ characteristics 104
XII
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
XIII
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
XIV
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
LIST OF ABBREVIATIONS
CD : Chart datum
CDR : Conceptual Design Rules (PIANC, 1997)
DWT : Dead weight tons
ICORELS : International Commission for the Reception of Large Ship
ID : Identification
IMO : International Maritime Organization
LNG : liquefied natural gas
NED : negative exponential distribution
NH : North Hemisphere
PIANC : Permanent International Association of Navigation Congresses
ROM : Recomendaciones de Obras Marítimas (Maritime Works Recommendations)
SH : South Hemisphere
TEU : twenty-foot equivalent unit
USACE : U.S. Army Corps of Engineers
LIST OF SYMBOLS
2
a : Acceleration (deceleration) rate for a vessel [m/s ]
d1MAX
in : Maximum waiting time arriving ships, one – way channel [min]
d1MAX
out : Maximum waiting time departing ships, one – way channel [min]
d 2MAX
in : Maximum waiting time arriving ships, two – way channel [min]
d 2MAX
out : Maximum waiting time departing ships, two – way channel [min]
d1in : Average waiting time arriving ships, one – way channel [min]
d1out : Average waiting time departing ships, one – way channel [min]
d 2in : Average waiting time arriving ships, two – way channel [min]
d 2out : Average waiting time departing ships, two – way channel [min]
XV
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
D : Ship draught[m]
D : Daylength [hr]
XVI
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
L : Latitude [°] (positive for North Hemisphere, negative for South Hemisphere)
XVII
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
: Volume of displacement [m ]
3
LIST OF UNITS
°C : degrees Celcius
kn : knot (1 kn = 0.514 m/s)
kW : kilowatt (1 kW = 1,000 W)
hr : hour
m : meter
3
m : cubic meter
nmi : nautical mile (1 nmi = 1,852 m)
s : second
t : ton (1 t = 1,000 kg)
kton : kilo ton (1 kton = 1,000 t)
XVIII
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
1 INTRODUCTION
The research consists of several topics. The first one is the definition of the generic problem (assessment of the
design of an access channel) that will be studied. The second one is the assessment of the design of a projected
access channel, which will be part of the future port of Taman, in Russia.
This chapter consists on a definition of the generic problem, a description of the Taman Seaport study case, the
definition of the thesis objectives, the research approach and the report structure.
1
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Subsequently, when a new design is being developed or when a change is considered in an existing port, the
main goal is to maintain the waiting time within the accepted margins.
2
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
dimensions, including their draughts. This deficit was increased at the moment of the Soviet Union splitting,
because Russia lost direct access to Odessa and Sevastopol ports, which are now part of Ukraine. These ports
continue exporting Russian cargo, but the related costs are higher (import duty) than in case of using a port
located in the Russian territory.
Almost half of the Russian ports are shallow (capacity of service of 10,000 dwt) and currently, only ten ports can
handle vessels up to 50,000 dwt (UK Trade & Investment, 2012). The situation is even more delicate considering
that there is only one port (Murmansk, in the northern part of the country) that allows modern vessels over
150,000 dwt. Because of these limitations, future expansions of the total Russian turnover are difficult to be
executed by the use of the existing ports. Hence, new terminals and ports will be built in Russia: Sabetta and
Mursmansk in the Artic Basin; Ust-Luga, Saint Petersburg and Bronka in the Baltic Basin; Novorossiysk, Sochi and
Taman in the Azov – Black Sea Basin; Olya in the Kaspian Basin; and Vostchny – Nakhodka in the Far East Basin
(UK Trade & Investment, 2012). The expected turnover will increase largely (46% with respect to 2010 turnover),
reaching 770 million tons.
In order to increase the annual export of dry cargo in the Russian Federation, the Russian authorities have
decided to build a new port in the Taman peninsula, facing the Black Sea. According to UK Trade & Investment
(2012), the Taman Dry Cargo Port will provide 100 million extra tons turnover by the year 2030, providing service
for vessels with a draught up to 14 metres. With an intended start of operations during the year 2016, “Taman is
the first in modern Russian dry cargo port project…” (Malisheva, 2012).
Figure 1.1. Location of Black Sea relative to the Russian territory (WorldAtlas.com)
3
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Total turnover
Number Name Type Operation
Mton/year
1 Container terminal 1
Container Import / export 10
2 Container terminal 2
5 Steel terminal Break bulk Import / export 4.4
6 Grain terminal Dry bulk Import / export 13.4
8 Fertilizer terminal 2 Dry bulk Import / export 8
9 Fertilizer terminal 1 Dry bulk Export 10
10 Iron ore terminal Dry bulk Export 15
11 Sulphur terminal Dry bulk Export 5
12 Coal terminal 1 Dry bulk Export 12
13 Coal terminal 2 Dry bulk Export 16
4
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Ukrainian border
Taman Port access channel
Figure 1.3. Preliminary access channel (straight) with respect to the Ukrainian border (Witteeven+Bos, 2012)
5
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
6
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Identification of the degree of dependence of congestion on several parameters related with the
operation of an access channel.
7
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
o Model set-up: corresponds to the identification and valorisation of the different parameters
required by the simulation model, incorporating it in the simulation model.
o Current design assessment: corresponds to the execution of the simulation model, the
obtaining of results and the assessment of the current design of Taman Port.
Taman optimised case
o Optimisation alternatives: corresponds to the definition and description of alternatives that
will incorporate some of the optimisation measures available on the generic simulation model.
o Evaluation of alternatives: corresponds to the application of the measures, the execution of
the generic simulation model and the obtaining of results.
o Final recommendation: based on the evaluation of alternatives, the best alternative will be
identified and proposed to the designers of the Taman Port.
Figure 1.4 presents a block diagram with the different phases of this study.
Problem definition
- analysis results Generic simulation model Taman base case
Evaluation
Generic
Checking
simulation model
Evaluation of
Analysis of results Output
alternatives
Recommendations Final
for designers recommendation
8
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Chapter 3 – Verbal Model: corresponds to the identification and description of the different processes
that are related with the operation of an access channel.
Chapter 4 – Basic Simulation Model: corresponds to the development of the first stage of the simulation
model.
Chapter 5 – Extended Simulation Model: corresponds to the implementation of different measures that
expand the capabilities of the model.
Chapter 6 – Analysis of Results: corresponds to the analysis of the results obtained via the extended
simulation model. The analysis is divided in two sections: the first one is the estimation of the effect of
different parameters on the average and maximum waiting time, due to congestion; the second one is
the quantification of the effects of different measures (modification of traffic rules, in reducing the
congestion.
Chapter 7 – Taman Seaport Case, Assessment of the Design: corresponds to the simulation and
assessment of the access channel of the future port of Taman.
Chapter 8 – Taman Seaport Case, Design Optimization: corresponds to the application of several
measures in order to optimise the operability of the Taman’s access channel and reduce congestion.
Chapter 9 – Final Conclusions and Recommendations: corresponds to the definition of the main
conclusions obtained within this research, as well as incorporating recommendations for future
researchers.
Chapter 10 – References.
9
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
2 LITERATURE REVIEW
The development of this thesis needs a review of references that are related with the design and optimisation of
access channels. In this context, two main themes are highlighted. The first is a description of the current
procedures and tools for designing an access channel. The second is a review of the state-of-art of simulation
models related with ports operations.
The objectives of the literature review are:
For the access channels design, to give a framework of the most accepted methodologies of design for
access channel. This framework will be used for the development of the simulation model.
Furthermore, an identification of those parameters that are not explicitly defined will be provided.
For the simulation model for port operations, to give a description of the available software for ports
operations (Harboursim), identifying its capabilities and its improvement opportunities. This
information will allow the definition of the desired capabilities for the simulation model to be
developed.
2.1.1 Alignment
PIANC (1997) gives several requirements that should be applied to design the alignment. These influencing
factors are useful when a conceptual design is under development, and are often related with environmental
conditions (waves, currents, winds) and the layout of the channel.
i) The shortest possible length.
ii) Conditions / basins at either end of the channel.
iii) Avoid obstacles or areas of accretion.
iv) Prevailing winds, currents and waves.
v) Avoid bends close to the port entrance.
vi) The edge of the channel should be such that ships passing along it do not cause disturbance or damage.
Thorensen (2010) indicates that the access channels should be located in areas where the natural depth is the
maximum. In this manner, it is possible to reduce the capital and maintenance costs of dredging.
For scenarios where a more accurate level of engineering (detailed engineering stage) is required, PIANC (1997)
provides more recommendations, most of them related with the design and verification of bends. The
verification is usually done with the use of ship simulators. The assessment of the bend layout is recommended
by the use of real and fast time simulations, over a large range of tidal and wind conditions. Ligteringen (2009)
indicates that fast time simulations are useful when comparing different alternatives, while real time simulations
are more suitable for final check of the alignment and also training of pilots under different weather conditions.
Ligteringen (2009) specifies that the radius of bends depends on the maneuverability of the design ship (length
between perpendiculars, water depth and rudder angle). Figure 2.1 presents the dependence of turning radius
with respect to the other parameters.
10
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 2.1. Turning radius as function of rudder angle and water depth (Ligteringen, 2009)
Mayer et al. (1999) did an extensive comparison of the design common practice in America and annotated
bibliography. They remark that the channel alignment is dictated by the natural course of the river or estuary. By
this, it is possible to minimize the capital dredging costs (length reduction and channel alignment in deeper
areas) with a criterion that is comparable to those defined by PIANC.
The USACE (2006) recommendations are in line with PIANC recommendations. Shortly, USACE recommends that
the alignment of the channel should avoid crosswinds, avoid areas of accretion, and maintain a small angle with
respect to predominant waves. If some of these criteria are in conflict, then several alternatives can be
evaluated via a multi-criteria analysis (manoeuvrability, capital and maintenance costs, etc.).
n
W 2WBM 2 W W
i 1
i Br WBg W P (two-way channel) (Eq. 2.2)
The basic manoeuvring lane WBM depends on ship manoeuvrability and the design ship beam. Related values are
presented in Table 2.1. The ship manoeuvrability is a qualitative definition. It depends on the ships
11
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
characteristics (like the shape of the hull) and the hydrodynamic aspects for different operations. The judgement
of the level of manoeuvrability is typically performed by experts.
Table 2.1. Basic manoeuvring lane (PIANC, 1997)
PIANC (1997) gives a complete description and definition for all additional widths W i, which are presented in
Table 2.2.
Table 2.2. Additional widths Wi for straight channel sections (PIANC, 1997)
12
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Table 2.2. Additional widths Wi for straight channel sections (cont.) (PIANC, 1997)
Thorensen (2010) mentions that when a ship is displaced from the centre of the channel, it experiences a bank
suction effect due to the asymmetrical flow. The result of this effect is a yawing movement. To counteract it, a
clearance lane should be considered. Thorensen indicates that this clearance lane should be the beam of the
largest ship, with a minimum value of 30 m.
PIANC (1997) also gives additional widths for passing distance in a two-way access channel (Table 2.3) and
additional width for bank clearance (Table 2.4).
Ligteringen (2009) indicates that only in cases of a very large tidal range, the PIANC rules can be modified into a
width equal to the total length of the design ship. This width may help a ship to avoid going aground with its
stern when it is heading to the channel’s bank.
13
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Table 2.3. Additional width for passing distance in two-way traffic (PIANC, 1997)
The PIANC method (CDR) does not give an explicit definition of when to design a channel as one-way or two-
way, which should be a definition based on an expected number of ships using the channel. Siregar (1995a)
analysed real port’s information and compared it with the CDR method, defining that the CDR rules calculations
for one-way channels are much closer to the reality than the results for two-ways channels. Sigerar formulates a
rule of thumb: when the channel width is lower than six times the design ship beam, the channel is one-way;
when its larger, then the channel is two-ways. Nevertheless, no references to the type of ship against the
number of ships are mentioned.
Sigerar (1995a) mentions that the CDR rules uses the design ship, which in reality is not necessarily the largest
ship. A channel may normally have a two-way traffic for ships whose dimensions fit the design ships. Only as a
matter of exception, when the ‘largest’ ship enters or leaves, the channel allows one-way traffic only.
In addition, PIANC’s (1997) detailed design recommendations for width incorporate ships manoeuvring
simulation models as the definitive design tool for designers. With the aid of this type of models (real – time
models), it is possible to include the behaviour and reaction of the human factors in the design process.
Although PIANC rules are widely used, there are other methods available. USACE (2006) methodology gives less
detailed guides than PIANC to estimate the width of a channel. In case of a one-way channel, the design of the
width can be in the range 2 to 6 times (even 7) the design ship beam. The same reference indicates that a final
range has been developed from 2.5 to 5 times the design ship beam has been developed. Nonetheless, the
method still requires the definition of the traffic pattern (one-way or two-way) as an input.
USACE (200s<sz6) indicates that, after several simulator studies, it was concluded that is possible to control ships
navigating in narrow channels, which means that the design rules and criteria are conservative. Demirbilek and
Sargent (1999) indicate that for straight channels and low currents, a width of 2 to 2.5 times the design ship
beam should be conservative. Nevertheless, it is difficult to provide a similar recommendation for a two-way
14
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
approach channel because of the lack of information from simulations. A rule of thumb is given for two-way
channels: the width has to be from 4.5 to 6 times the average beam of opposing ships.
Gray et al. (2002) concluded that the channel width is treated somewhat similarly by PIANC and USACE methods.
Nevertheless, USACE does not provide the level of detail and attention to the channel width. In many cases, the
channel width is reduced to decrease capital and maintenance costs. This reduction can go down to the point
where a ship can no longer sail through a channel, based solely on its width. Some other impacts are the limited
definition of one versus two way and reduced vessel speed. An important remark is that the current trend is to
expand ships beams whereas the width of channels does not.
Finally, a simple definition of the width for a two way channel that incorporate the effect of wind and currents is
given by Xingyan et al (2011), which correspond to the Chinese Standard. According with this reference, the
width of the channel is given by:
W 2 A B 2c (Eq. 2.3)
A nL sin B (Eq. 2.4)
In case that opposing ships have different dimensions, the width of the manoeuvring lane A is calculated
separately for each vessel.
2.1.4 Depth
Mayer et al. (2009) mention that the channel depth should accommodate the draught of the loaded design
vessel, with allowance for squat, sinkage, tidal variations, effects of waves, wind and currents. Additionally, they
indicate that a concept design should establish the initial estimates for the depth: the detailed design refines
alternatives and finalizes the design with more extensive information (like a more detailed bathymetry or a
wave’s propagation study for different sections of the access channel). Nevertheless, for a fast – time simulation
model, an initial estimation for the depth is sufficient.
15
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 2.2. Width of a swept tract in a turn as a function of rudder angle and water depth (PIANC, 1997)
USACE (2006) indicates some extra considerations to those mentioned by Mayer et al. The extra considerations
are the effect of fresh water and allowances for dredging tolerance and advance maintenance.
Ligteringen (2009) defines several factors that influence the depth of a channel:
Draught of the design ship.
Other ship-related factors like squat and trim, and the vertical response to waves.
Water level, mostly related with tidal levels.
Channel bottom factors.
The depth can be determined according to the following equation (Ligteringen, 2009):
d D T smax r m (Eq. 2.5)
In simple words, for estimation of the required depth of a channel it is necessary to consider the draught of the
design ship, the maximum sinkage of the ship, the vertical motions due to wave response and a safety margin.
Additionally, it allows the designer to introduce a tidal window, which represents an opportunity to reduce the
volume to be dredged and therefore the capital and maintenance costs. Lighteringen (2009) describes a tidal
window as a restriction to navigation, “at least for the biggest ships, to a limited period of the tide”. Restrictions
may apply as a current window (tidal currents) and even as a wave window.
Figure 2.3 presents a scheme of the cross section of a channel and the related under keel clearance factors.
According to PIANC (1997), the calculation of the sinkage of the ship (squat and trim) can be obtained by three
different options:
i) Huuska / Guliev expression (ICORELS)
ii) Graphical method
iii) Minimum value for on water depth/draught ratio
16
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
In addition to the Huuska / Guliev expression, other expressions suitable for prediction of the squat are available
(Barrass II and Eryuzlu et al. formulas). Nevertheless, the Huuska / Guliev expression gives quick results and it is
valid for unrestricted waterways and restricted channels. An important restriction of this method is that it
should not be used for Froude depth numbers larger than 0.7.
The graphical method for squat estimation is presented in Figure 2.4. The method returns an estimation of the
bow or stern sinkage, based on the ship speed, the water depth, the contour of the bow or stern appropriate to
the at – rest trim of the ship and the ship’s length.
Figure 2.4. Squat estimation chart for full-bodied ships (PIANC, 1997)
17
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Dand (2004) comments “the vertical motions due to waves are usually considered only at the detailed design
stage with the use of mathematical models. For concept design a depth/draught ratio, based on experience, is
used to allow for wave-induced vertical motions.” PIANC (1997) gives some quick and simple factors to predict
the squat, draught and sounding uncertainties, by setting a minimum value of the depth/draught ratio. For calm
waters, the depth/draught ratio usually may be 1.1, although values of 1.15 can be found. For channels with
waves, the coefficient can be 1.3 or even larger.
A last method is the inclusion of safety factors, expressed in terms of the ship’s draught. Witteeven + Bos (2012),
in its conceptual design of the Taman Port, define a simpler criteria basis for the gross underkeel clearance. For
outer channels, the minimum clearance is 20% the ship’s draught, while for inner channels the clearance is
reduced to 15% the ship’s draught.
Several methods are presented. For the application of a fast – time simulation model, it is advisable to use a
simple approach (like definition of Witteeven + Bos) because the results are typically used or assessing a design
during a conceptual engineering stage.
18
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
19
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The assessment of an access channel design can be performed via both methods. However, the real-time
simulations are mainly attempted to evaluate the manoeuvrability conditions of a ship sailing through a certain
channel. On the other hand, the fast-time simulation models can allow the simulation of several ships in a short
time, while the configuration of the model (port layout, fleet composition, number of ships, etc.) can be easily
varied. Then, when the design of an access channel is needed to be evaluated in terms of congestion, it is
preferable to use a fast-time simulation model.
20
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
21
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Ship processes: the ship passes through the port system. Several components (pilot, harbourmaster and
administrator 2) run this process.
Procedure occupation channels: the procedure determines the possibility of entering or leaving of a
ship, related to the occupation of the channel section, turning circles and berthing areas.
Harboursim is a model that has been used widely, because it allows the evaluation of a port like the sum of
several systems. It is easy to find examples of researchers who have used this model in the evaluation of master
plans for different ports (like Budiyono (1998), Van Driel (1993), Deville (2011), etc.).
Nevertheless, Van der Meer (2013) found some errors in the model, which were corrected. Van der Meer’s
conclusion was a recommendation of do not use the original Harboursim model until it has been adapted to the
found errors. Nevertheless, the functionality of Harboursim has been tested for different evaluations of master
plans and its results are widely accepted. Hence, even considering that the software can be improved, its use is
still positive, especially for the assessment of master plans during the conceptual engineering stages.
Other simulation models (like those presented by Solari et al (2010), Demirci (2003) or Hofseth (2007)) are in
different stages of development and/or use. Nevertheless, these models were not available for this literature
review, hence, an identification of its capabilities and limitations should be performed during a next stage of
research.
2.3 Conclusions
22
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
These simulation models are focused on the evaluation of a port as a complex system, including several
processes (anchorage, access channel, service at the berth, yards utilisation, etc.). Nevertheless, these
simulation models do not have a focus on assessing the access channel.
Harboursim executes a simulation run based on a process approach, which in simple words is a sequence of
processes for an arriving ship (arrival, sailing to berth, service time, sailing to sea, leave). The processes are
defined as time cycles: each time a cycle is finished, the next process begins. When a certain ship is willing to use
the access channel, the model checks if the current utilisation of the channel allows the entrance of a new one.
Nevertheless, the process is simplified: if the channel is two – way type then encounters will be admitted;
otherwise, the ship will have to wait.
If an assessment of the access channel is needed, a more detailed simulation of the processes related to it is
required. Then, a verbal model of the different actions that ships can perform on an access channel is required
(speed reduction to enter into the port, overtaking of slow ships, etc.). Additionally, a new simulation model
should have enough capabilities to model different configurations of access channels (intermediate anchorages,
presence of a passing section) and related traffic rules (intermediate entrances for smaller vessels, operation of
one – way channel for large carriers and two – way channel for small vessels).
Finally, the results of Harboursim indicate the average waiting time for different fleets, both for arriving and
departing ships. Although the characterisation of these results is correct, there is no classification for the reasons
behind this delay. The delay can occur from different sources (berth unavailability, congestion, weather
conditions, etc.): if the access channel is being assessed, it is crucial to know if the waiting time of the vessels is
due to congestion or because of a lack of free berths. This type of results can give more information to the
designers in order to know exactly what processes are being the bottle-necks and what could the more suitable
measures to diminish the delay.
23
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
3 VERBAL MODEL
The generation of a simulation model to assess an access channel requires a previous description and
schematisation of the reality. These actions need to be done within the system boundaries, which depend on the
specific problem that is being studied.
This chapter presents the system boundaries, the definition and description of the main processes involved, the
model’s description and the equations considered to translate reality into the model.
Arrival to port
Wait at anchorage
Navigation towards the berth (through access channel and inner basins)
Berthing
(Un)loading
Wait at berth
Navigation towards open sea (through inner basins and access channel)
Leave port
It is necessary to emphasize that some of the processes listed before (navigation through inner basins, berthing
and (un)loading) are outside of the defined boundaries; hence, they will not be modelled in detail. Nevertheless,
they cannot be neglected because they are part of the natural sequence of ships and the time to execute these
tasks must be considered in the complete cycle.
24
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The typical sequence of these processes is schematized in Figure 3.1. Green boxes represent simplified
processes, while blue boxes represent detailed processes.
Arri va l
(Un)l oading
NaEnter
vi gate
i nto
within
outer
the Yes Sa i l to sea No Wa i t a t berth
port (tug
cha nnel
a ssisted) pos sible?
Sa i l through outer
Exi t outer channel
cha nnel
Lea ve
25
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Working days: some ports offer service seven days per week, others may be restricted or closed during
weekends.
Channel opening: some approach channels may be accessible 24 hours per day while others may be
accessible only with daylight.
Other events: other irregular events can force the port to close for a certain period of time.
If a certain port has a channel that operates only with daylight, it is important to know the duration of the day.
This duration depends on the latitude and the date. Figure 3.2 presents a graph with the daylight variation due
to these factors.
26
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 3.2. Annual variations in day length at different locations in the North Hemisphere
(PhysicalGeography.net)
Table 3.1 introduces the characteristics of the constituents M2 and S2.
The definition of the current behaviour in a specific area is much more complicated to estimate than the tides.
Although currents in a coastal area may have its origin in the tides, they depend more strongly on the
bathymetry and the propagation of the tides. In addition, if the port is located near an estuary, residual currents
could be found. Hence, no practical rule can be defined. A current window can be defined only if specific
information on the currents pattern is available.
Due to the stochastic behaviour of waves, is difficult to develop precise prognosis of the wave field during a
certain time. No practical information is available in this subject.
27
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Wind: a wind-rose with the yearly distribution of the wind direction and forces (or velocities) in
percentage of time.
Waves: a waves-rose with the yearly distribution of the wave’s direction, amplitude and period.
Additional statistics of extreme events (significant wave height versus period of return) is also required.
Ice: estimation of downtime due to ice problems per season or per month. Ice problems means that the
ice layer is sufficiently thick to do not allow vessels to passage.
Fog: estimation of downtime due to visibility problems per year. The definition of visibility problems is
when it is lower than 1,000 m (Thorensen, 2010).
The wind and waves are input parameters considered by the PIANC rules. Then, there are maximum values that
allow safe navigation. If some of these variables exceed the maximum, then the navigation will not be safe
anymore and the ship will not be authorised to enter to the channel.
28
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Channel depth
Channel restrictions (maximum allowed speed)
Another feature regarding the admiralty coefficient is that it depends on the total weight of a ship. Therefore, if
a ship is sailing empty, it will be able to sail faster than a ship that is fully loaded.
Table 3.2 presents some maximum speeds for different types of ships, which can be considered as reference
values. Nevertheless, when evaluating a real project, the characteristics of expected ships (including their speed
for different loads) should be defined.
Table 3.2. Maximum speed for several ships ((1) Lighteringen, 2009); (2) Lingwood, 2007)
The second criterion refers to the maximum speed considering the channel depth. The hydrodynamic resistance
to motion of a ship on shallow waters is governed by the Froude Depth number (PIANC, 1997). The calculation of
this dimensionless number is presented in equation 3.2.
V
Fnh (Eq. 3.2)
g d
PIANC (1997) defines some maximum factors for this Froude number (0.6 for tankers, 0.7 for container ships). If
the velocity of the ship is traduced into a larger Froude number, the resistance to the ship’s motion reaches very
high values: all the extra power consumed by the ship begins to enlarge the energy contained in primary and
secondary waves (due to ship movement). In this sense, navigation with high speeds becomes extremely
inefficient.
As some reference values, if a channel has a nominal depth of 15 m, the maximum speed for tankers and
container ships will be 14.2 kn and 16.5 kn, respectively. If the depth of the same channel increases to 20 m, the
maximum velocities also increases (16.3 kn for tankers; 19.1 kn for container ships).
The third criterion depends on maximum allowed speed on different sections of the channel. These restrictions
are related with the design of the channel, which are well covered by the CDR (PIANC, 1997).
A ship will navigate with the maximum available speed, which corresponds to the minimum of these three
restrictions. When the maximum speed is set, ships will move with a real speed that does not fluctuate very
significantly. Hence, a narrow uniform distribution can represent the sailing time (and the related speed) of a
given ship with acceptable accuracy (PIANC, 1997).
29
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Although it could be expected that a ship would try to reach the berth as fast as it cans, there are other
parameters that might influence the real speed of the vessel. Among these possibilities it can be considered the
existence of a scheduled arrival time, or a reduction of the fuel consumption by reducing the speed.
3 -6 LC 3 -6 LB
LC LB LA
30
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
CDR rules are not respected, then overtaking is not permitted. For this type of channel, an allowance of
this type of operation is required. The operability of the channel must be evaluated by the harbour
master, depending on the channel’s dimensions and the vessels’ beams.
If overtaking is not possible, the faster ship will have to reduce its speed in order to maintain the minimum
distance with the previous ship. Another possibility is to wait at the anchorage until the position of the slower
ship is sufficient for the faster ship to sail with cruise speed without reducing the distance below the safety
criterion.
Figure 3.4 presents a scheme with different the overtaking manoeuvre. Figure 3.5 presents a scheme with the
classification of access channels.
VB>VA VA
Figure 3.5. Scheme with classification of access channels: a) One-way access channel; b) Two-way access
channel; c) One or two-way access channel
Figure 3.6 presents a scheme with the encountering manoeuvre.
VC
VA
31
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The maximum allowed speed in the protected area is lower than the maximum allowed speed in the
outer channel.
The ship is approaching a turning circle and tug assistance is required.
Large vessels have a large stopping distance and usually they lack of control during that manoeuvre. In practice,
when tugboats are not available, in order to maintain sufficient rudder control, the vessel is forced to maintain a
minimum speed of 3 to 4 kn (relative to the water) (Ligteringen, 2009). So, for lower speeds (a final stop), the
assistance of tugboats is required.
Because the entrance speed is reduced, the ship will have to start reducing its speed in the final part of the outer
channel. The reduction starts from the cruise speed that the vessel can maintain in the channel, and typically
finalizes with the entrance speed. Both speeds were described previously.
The deceleration rate of a certain vessel depends on several factors (IMO, 2002):
The resistance of the hull is proportional to the square of the ship speed.
The astern thrust is constant throughout the stopping manoeuvre and equal to the aster thrust
generated by the propeller when the ship eventually stops dead in the water.
The propeller is reverse as rapidly as possible after the astern order is given.
The speed reduction to the entrance speed is totally executed on the outer access channel. Hence, it will be
covered in detail by the simulation model.
32
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 3.7. Bow thrust as a function of tugboat velocity (Puertos del Estado, 2000)
Figure 3.8 presents a graph with the relationship between the berthing speed, the displaced volume and the
berthing conditions.
The mooring process is not part of the processes included within the access channel. Therefore, the required
time to execute this task will be considered according to a probabilistic distribution. Typical average values for
the mooring process are between 15 – 30 minutes, and it depends on the size of the vessel, the configuration of
the quay (wall or piers) and the properties of the fenders.
Figure 3.8. Relationship between berthing speed, displaced volume and berthing conditions (Quist, 2012)
3.2.6 (Un)loading
The loading and unloading of cargo are the main purpose of a ship that is visiting a port. The required service
time depends on several characteristics:
Cargo type (container, liquid bulk, dry bulk, general cargo, etc.)
33
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Nominal and maximum rate (type of cranes and cargo, number of gangs, etc.)
Loading and/or unloading
Total cargo (tonnage, TEUs)
Some typical service rates are highlighted (Ligteringen, 2009):
Container terminals: peak 40 – 50 moves per hour; average 20 – 30 moves per hour.
General cargo: conventional 8.5 – 12.5 t/hr; timber and timer products 12.5 – 25 t/hr; steel products 20
– 40 t/hr; containerised cargo 30 – 55 t/hr.
Liquid bulk: 10% per hour of deadweight tonnage.
Dry cargo: loading 2,000 – 20,000 t/hr (linear and radial loaders); unloading depends on equipment
(grabs, pneumatic systems, vertical conveyors, bucket elevators, slurry systems, self-discharging
vessels), but range is wide (200 – 2,500 t/hr).
The service time can be modelled with a distribution function, which depends on the average productivity. For a
container terminal is possible to use a third degree k-Erlang; for other terminals a second degree k-Erlang is
more suitable.
Like the navigation of vessels within the inner basins, the service processes are outside of the boundaries of the
model. Hence, a correct estimation of the total time is required by the model. The estimation will be obtained
with the described probability functions.
3.3.1 Assumptions
The following assumptions are considered for the simulation model:
The ships movement is modelled in a one dimension scheme (position along x axis versus time).
34
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The position of a ship is defined as a single point, which is considered at the centre of the ship
(longitudinal direction).
A ship entering the channel at the outer entrance will already sail with its maximum velocity,
independently if it is entering directly or if it has already waited at the anchorage area.
The deceleration of a vessel is influenced by the friction of the vessel and the water. When the velocity
of the vessel increases, the friction becomes larger and the deceleration rate will also increase.
Nevertheless, for simplicity reasons, the simulation model assumes that the deceleration rate is
constant between two given speeds. This assumption underestimates the deceleration rate for high
speeds and overestimates it for low speeds.
During stage 2 (speed reduction), the deceleration rate is obtained from the IMO relationship.
During stage 3 (final stop), the deceleration rate is obtained from the stopping distance, which is
assumed to be one ship’s length.
Ships entering a port will pass through two stages of deceleration. It is assumed that ships sailing
towards the sea will pass through two stages of acceleration. These acceleration rate will be the same
than the deceleration rate (for ships moving towards the port), but with opposite direction.
The final speed corresponds to the velocity of a vessel being assisted by tugs. It is assumed that this
speed will be 1.5 knots.
The channel modelling will be considered as a straight channel.
The traffic rules for vessels that will be modelled are: type of channel operation (one, two or mixed
operation), safety distance between ships in a convoy, maximum speed on outer and inner channel,
possibility of reducing vessel speed and single / multiple entrances option.
Each ship will enter into a port with a pilot. There will be no limitation on the number of pilots available.
Each ship (independently on its size) will enter into the port with tug assistance. Furthermore, there will
be unlimited number of tugs.
The delay cause for a ship will be assumed to be caused by a single source. When two or more causes of
delay will be active simultaneously, the delay source will be defined according to a ranking of delays
(first port / channel unavailability; secondly berth unavailability; thirdly traffic conditions; and fourth
the weather conditions).
In case of encountering or overtaking, the speed of the involved vessels is not modified.
The longitudinal movement of the ship is not affect by longitudinal currents nor longitudinal wind.
3.3.2 Exclusions
The following items are excluded for the simulation model:
Movement of the ships entering, anchoring and leaving the anchorage area.
Movement of the ships through inner basins and other nautical units located beyond the first turning
circle.
Position of berthing of a ship in its terminal of destination.
Status of cranes, storage yards, stocking piles, etc.
35
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
xi 1 xi Vi t (Eq. 3.3)
a t2
xt x0 V 0 t (Eq. 3.4)
2
36
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
It is important to mention that equation 1.4 is valid for stages 2 and 4 (deceleration of the ship). For stages 1 and
3 there is no acceleration, therefore the velocity of the ship can be considered as constant.
The reduction and final stop of a ship is typically defined in terms of the ship’s length. Taking the kinematic
equations, it is possible to define the acceleration rate in terms of the initial and final speed, and the ship’s
length. Equation 3.7 presents this relationship.
37
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
f t 0 if t 0 (Eq. 3.9)
Sometimes, the inter-arrival time follows a k-Erlang distribution. The definition of a k-Erlang distribution is
presented in equation 3.10.
f t
k xk t k 1 exp k x t if x 0 (Eq. 3.10)
k 1!
The coefficient k corresponds to the degree of the k-Erlang distribution. In all of these equations, t is considered
as the stochastic variable.
P sin 1 0.39795 cos 0.2163108 2 tan 10.96713896 tan0.0080 J 186 (Eq. 3.11)
0.2667 L
sin 180 sin 180 sinP
D 24 cos 1
24 (Eq. 3.12)
L
cos cos P
180
38
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
In case the speed is not available, this formula allows the estimation of the maximum speed of a ship, based on
some of its characteristics (dimensions, total weight and installed power).
39
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
3.5 Conclusions
Within the system physical boundaries (outer edge of the access channel and centre of the turning circle) there
are several processes related to a vessel sailing through the access channel, either with direction towards the
berth or the open sea. The processes related are the sailing with cruise speed through the outer channel, the
first speed reduction to reduce speed to enter into the port (through the breakwaters opening), the sailing with
the entrance speed and tug making fast, and the final stop at the turning circle.
There are several interactions between ships that are using the access channel (encounter, overtaking and
convoy’s movement). Each interaction depends on the current traffic rules (one, two-way channel or mixed
operation, safety distance between ships), which are particular for each port.
The authorisation for a vessel to enter into the access channel (from the outer edge of the channel or the
turning circle protected by the breakwaters) depends on several conditions. First, the access channel must be
opened for navigation. Secondly, a minimum free quay length on the terminal of destination is required (only for
vessels navigating to the port). Third, the traffic conditions must allow the safe movement of the vessel,
according to the current traffic rules and the characteristics of the nautical infrastructure. Finally, the weather
conditions (wind, waves, ice or fog) must be suitable for a ship to navigate safely.
Other processes related to the typical cycle of a ship (waiting at the anchorage, navigation through inner basins,
(un)loading), will be considered into the modelled but only as simplified functions. The simplification is made
because these processes are outside of the defined boundaries, and are no directly related with the processes
occurring within the access channel. Nevertheless, these processes cannot be discarded because the access
channel operation is for vessels sailing to and from the port. Then, for estimating the departing time of a vessel
it is necessary to have an estimation of the duration of the different processes since its arrival.
40
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
4.1.1 Generalities
4.1.1.1 Programming language
The simulation model is programmed with Matlab®. The post – processing calculations can be made with Excel®.
The software Matlab® was selected to create the simulation model based on a trade – off analysis. Between its
benefits are present the possibility to develop calculations for large matrixes, the creation and application of
new functions, sufficiently good tools to create graphs. Furthermore, the use of Matlab® is massive; hence, the
possibilities for a new user to implement an improvement to the model are simplified with the use of this
software. Finally, Matlab® incorporates commands to import and export data to and from Excel®, which is
indeed a massive software to analyse information contained on large matrixes (output files).
4.1.1.4 Terminals
The maximum number of terminals that can be modelled is 10 (number set by the user). Still, if it is necessary, it
is possible to modify the code to incorporate extra terminals.
41
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
4.1.1.5 Output
The software returns a summary of the waiting time for each ship, indicating the cause of delay. The record of
waiting time is available for both ships waiting to enter or leave a port.
A second output is a record of the number of encounters obtained within the access channel.
The last output is a record of the number of ships waiting at the anchorage site during different moments of the
simulation.
42
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Set-up stage
General Layout Priority Vessel
(user definition)
Wait at Navigate to
anchorage Queue in call berth
Anchorage Service
manager
Pilot call
Port call Wait at berth
Queue manager Traffic call
Queue out call
Ship manager
Leave Navigate to sea
Nautical manager
Anchorage
summary
43
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
44
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
45
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Route i [m] (distance by inner wet infrastructure to the terminal i), i from 1 to the number of terminals.
Necessary for calculation of the sailing time through the inner basins.
46
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Cargo hazard definition vectors (input parameter for PIANC rules). It does not depend on the type of
vessel, but it depends on the cargo (1 means low hazard, 2 means medium hazard, and 3 means high
hazard).
x ln1 R
1
(Eq. 4.1)
iii) Repeat process for next ship, until the arrival time of last ship exceeds the total simulation time.
Similarly, for a k-Erlang distribution, a sequence is defined by Groenveld (2001):
i) Generate k random numbers Ri (Matlab® function).
ii) Obtain a respective value for the inter-arrival time x, according to equation 4.2.
1 k
x ln Ri (Eq. 4.2)
k i 1
iii) Repeat process for next ship, until the arrival time of last ship exceeds the total simulation time.
Because the arrival time is described by these functions, no boundaries are considered to prevent very short and
very long inter-arrival times. Furthermore, this method does not guarantee that the yearly calls for a certain ship
will be exactly the theoretical: the number of generated ships can be different, but the average value for several
simulations is similar to the theoretical value.
47
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Both sequences are suitable for defininig the arrival time of a ship that depends on the arrival time of the
previous one. Because there is no information for the ships arrival before the initial simulation time, the arrival
time for the first ship of each fleet is calculated with an uniform distribution, independently on the distribution
function chosen for the rest of the ships. This allows the possibility that the first ship of each fleet can arrive just
after the beginning of the simulation with the same probability of arriving after the theoretical inter – arrival
time.
The sequence is:
i) Generate a random number R (Matlab® function).
ii) Obtain a respective value for the inter-arrival time x, according to equation 4.3.
R
x (Eq. 4.3)
The ships generation module creates a matrix that compreses the arrival information for all terminals and fleets.
The matrix structure is presented in Table 4.1
Table 4.1. Matrix structure for ships generation module
R1
x (Eq. 4.4)
T
i) Obtain the duration of the event F, according to equation 4.5.
48
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
49
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
50
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Ships manager
The ships manager module is a matrix that incorporates all relevant information regarding the active ships. Table
4.7 presents a summary of the information contained in this module.
Table 4.7. Matrix structure for ships manager module
Berth manager
The berth manager module is a matrix that incorporates all relevant information regarding the utilisation of quay
for the different terminals. Table 4.8 presents a summary of the information contained in this module.
Table 4.8. Matrix structure for berth manager module
51
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Queue manager
The queue manager module is a matrix that incorporates the ships that are present at the queue lists (in and
out). Table 4.10 presents a summary of the information contained in this module.
Table 4.10. Matrix structure for queue manager module
52
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
the length of a ship or a modification of the ships manager module). Appendix A presents a summary of the
auxiliary functions, the necessary input parameters and the type of result.
53
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
4.3 Algorithms
In this section, the algorithms of the most important modules of the simulation model code are presented:
1. Main run module
2. Queue in / out call
3. Port call
4. Berth call
5. Traffic call
6. Pilot call
No
Crea te s hip i n ships Queue in No Ta ke fist ship in queue
s ummary empty? i n l ist
Yes
Incorporate s hip i n
Port ca l l ?
No
queue manager
Yes
Cha nge s hip s tatus to Incorporate s hip i n
Berth ca ll?
No
'queue out' s hi ps manager
No Yes
Obta i n new ship Check s hips geneation
Pi l ot call?
No Wri te delay i n s hip
pos ition l i st s ummary
No Yes
Queue out Yes La s t s hip i n Shi p changes s tatus Wri te utilisation in
empty? queue in? to 's a iling to port' a nchorage s ummary
No
Ta ke fist ship in queue
outlist
Port ca l l ?
No
Yes
Tra ffi c call?
No
Yes
Pi l ot call?
No Wri te delay i n s hip
Go to next ti me s tep
s ummary
Yes
Shi p changes s tatus La s t s hip i n Dea ctivation of s hips
to 's a iling to s ea' queue out? No 'l eaving port'
54
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
No
Return ship's ID (high Obtai n list of high
pri ori ty) pri ori ty ships
No
Ra nk higher Count hi gh priority
tha n count? s hi ps
Yes
Obta i n simulation
ti me
No
Get s hip's draught
Yes Port
opeating?
No
Ans wer positive
55
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Obta i n current
Obtai n ship length
a va ilable quay l ength
Yes
Ans wer negative
No Di s tance Cha nnel 1- Yes Ans wer negative
s a fe? wa y?
Yes Yes No
Cha nnel 1- Yes Encounters Obta i n distance with
wa y? >0? overta ked s hip
No No
Encounters > No Ans wer positive
Di s tance No
Yes 1? Yes s a fe?
No
Yes Fog a ctive?
No
Yes Ice a ctive?
No Ans wer negative
56
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 4.8 presents a histogram with the inter arrival results and its comparison with the theoretical NED
function.
Figure 4.8. Histogram of inter arrival results of 1,000 ships per year with NED distribution
2
The correlation coefficient R between the simulated and theoretical results is equal to 0.997. Hence, the
modelling of the inter arrival time is correct.
57
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 4.9 presents a histogram with the service time results and its comparison with the k-Erlang functions.
Figure 4.9. Histogram of service time results of 1,000 ships per year with k-Erlang distributions
2
The correlation coefficients R between the simulated and theoretical results for a second and a third degree k-
Erlang distribution are equal to 0.966 and 0.951, respectively. Therefore, the modelling of the service time with
both k-Erlang functions is correct.
58
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
59
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure
4.11. Single ship sailing towards sea, a) position, b) speed
The ships motion of a ship sailing towards the sea is similar to the motion of a ship entering a port, but in
opposite direction. The ship passes through acceleration stages (first acceleration to achieve the entering speed,
second acceleration to reach the maximum speed at the access channel). Because of this, the required time to
sail the entire access channel is almost the same for both situations.
60
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Instead of it, after one ship has finished its path, the second will be allowed to start sailing. Figure 4.13 presents
the trajectory of both ships.
Figure 4.14 presents the path of each ship for this scenario.
According to the graph, ships B and D are forced to wait at the anchorage site in order to maintain the minimum
distance with the previous ship. Specifically in the case of ship B, its length (280 m) requires a safe distance of
1,400 m with respect to ship A. At the outer channel the distance is larger (about 4,800 m), but because of the
speed reduction of ship A during stages 2, 3 and 4 (around time 50 min), the distance between both ships
reduces to a value of around 1,400 m. If the ship B does not wait at the anchorage, the distance with respect to
ship A will become too short around time 50 min.
61
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
62
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
63
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
64
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The simulation procedure is considering a time step of 4 minutes, a total simulation time of 1 year and ten
simulation runs. The results are obtained for two conditions (one and two berths), and are presented in Tables
4.18 and 4.19.
Table 4.18. Characteristics of the scenario solved via queuing theory, one terminal (M/E2/1)
Theoretical Obtained
Theoretical Yearly Average inter
normalized normalized Accuracy
utilisation calls arrival time
waiting time waiting time
[-] [-] [ships/yr] [min] [-] [-]
0.1 0.08 35 15,017 0.07 0.04
0.2 0.19 70 7,509 0.16 0.09
0.3 0.32 105 5,006 0.34 0.09
0.4 0.50 140 3,754 0.61 0.38
0.5 0.75 175 3,003 0.70 0.25
0.6 1.13 210 2,503 1.12 0.27
0.7 1.75 245 2,145 1.77 0.63
0.8 3.00 280 1,877 3.01 1.74
0.9 6.75 315 1,669 5.55 1.73
Table 4.19. Characteristics of the scenario solved via queuing theory, two terminals (M/E2/2)
Theoretical Obtained
Theoretical Yearly Average inter
normalized normalized Accuracy
utilisation calls arrival time
waiting time waiting time
[-] [-] [ships/yr] [min] [-] [-]
0.1 0.01 70 7,509 0.01 0.01
0.2 0.03 140 3,754 0.03 0.02
0.3 0.08 210 2,503 0.07 0.02
0.4 0.15 280 1,877 0.17 0.05
0.5 0.26 350 1,502 0.19 0.04
0.6 0.43 420 1,251 0.52 0.15
0.7 0.73 491 1,070 0.76 0.17
0.8 1.34 561 937 1.19 0.28
0.9 3.14 631 833 3.16 1.05
65
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 4.18 presents the comparison between the theoretical and the obtained results, including the upper and
lower boundaries (accuracy of the results).
Figure 4.18. Comparison between theoretical and model’s results: a) 1 berth, b) 2 berths
2
The correlation coefficient R between the theoretical and model’s results for 1 and 2 berths are 0.990 and
0.996, respectively. Hence, the performance of the simulation model is correct.
66
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The fleet characteristics for each terminal are presented in Table 4.20.
Table 4.20. Fleet characteristics of Port of Beira
Inter arrival
Berth Dead weight Ship’s
Yearly calls distribution
Fleet Terminal length tonnage length
[ships/yr] function
[m] [ton – TEUs] [m]
[-]
Container Container 261 NED 646 350 TEUs 154
General cargo General cargo 387 NED 321 + 336 2,500 113
Coal Vale 1 Coal Vale 4 x 21 k – Erlang (10) 230 27,000 194
Coal Vale 2 Coal Vale 3 x 21 K – Erlang (10) 230 27,000 194
Coal Rio Tinto Coal Tio Tinto 53 k – Erlang (10) 230 30,000 185
Fuel Fuel 90 NED 215 10,000 130
The service times for the different ships are presented in Table 4.21.
Table 4.21. Fleet service times of Port of Beira
A complete description of the port and its related master plan can be found in the text prepared by Van der
Meer (2009).
A comparison between the results obtained by Van der Meer (2009) and the simulation model is presented in
Figure 4.19.
Figure 4.19. Comparison of results between Van der Meer (2009) and the simulation model: a) daylight
navigation, b) 24 hours navigation
2
The correlation coefficient R for a daylight and 24 hours navigation are 0.457 and 0.593, respectively. These
coefficients are low because the results of two terminals are different (general cargo and Rio Tinto coal
67
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
terminals), obtaining a difference of the waiting time higher than 5 hours (for daylight navigation) and 4 hours
(for 24 hours navigation). Nevertheless, for the rest of the terminals (3 terminals for daylight operation; 4
terminals for 24 operation) the time difference is lower than 1 hour.
Independently on the previous results, it seems that the port considered to test the results of the new
simulation model can be a complicated example. The motivation to use the Port of Beira results obeys to the
intention to compare results with a real (and analysed) problem, as well as to take one of the latest documented
uses of Harboursim. Additionally, almost all the characteristics of the Beira port modelling with Harboursim can
be reproduced with the proposed simulation model (excepting the navigation from the access channel to an
offshore terminal, which was modelled as a navigation through inner basin with a speed and distance that gives
a sailing time similar to the cycle defined in Harboursim). However, the utility of using this case is questionable,
based on the complexity of Beira and the obtained results. Hence, for a future development stage of this model,
it is highly recommended to simulate more (and simpler) situations, in order to obtain a better testing /
validation of the model.
4.6 Conclusions
The modelling of a port via the proposed simulation model has four stages (input stage, data generation stage,
model running stage and output stage). While for the input stage, the set up of the model is required by the
user, the rest of the stages are automatically executed by the simulation model.
The output of the model contains detailed information of the delay of the modelled ships, the utilisation of the
anchorage area and a summary with the encounters within different sections of the access channel. Particularly,
the information of the delay of the ships is relevant. The delay (for arriving and departing ships) is classified
according to the cause of the delay (port, berth, traffic or weather conditions). This type of output represents an
improvement in the type of results available for similar simulation models, which typically only present the delay
as a single value. The available information allows to:
Identify the main reason(s) behind a large delay, allowing a designer or port operator to execute proper
measures according to the cause of the delay. For example, if the main reason of the delay is due to
congestion, a traffic rule modification or an enlargement of the access channel could be the best
options. But, if the main reason is due to berth unavailability, an improvement of the service rate
(update and/or increase of the number of cranes, enlargement of the quay length, etc.) could be the
best choices.
To construct histograms of the delay (total and classified). A further quantification of the benefit of
reducing the delay of the ships requires knowing how many vessels have a large delay and how many
do not have it. The estimation of the benefit of reducing the waiting time is complicated and exclusive
for each port; hence, the more information about delay, the better the accuracy of the benefit will be
obtained.
The simulation model check has probed that the simulated behaviour of the ships follows the expected
behaviour, based on the traffic rules modelled and the characteristics of the access channel. Nevertheless, an
important stage on the development of a simulation model is not present in this research, which is the validation
(and possible calibration) of the model with real data. Then, an improvement opportunity is the validation of the
simulation model, including a possible stage of calibration of some of its parameters (ships reducing speed
behaviour, mooring and sailing though inner basins times, etc.) with the feedback obtained from the validation
itself.
68
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
69
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Outer Intermediate
entrance entrance
Water surface
Figure 5.3. Scheme of speed reduction: a) without speed reduction; b) with speed reduction
It is important to notice that the first measure is suitable for fleets with a widely varied beam between the
different vessels. If the vessels’ beams are uniform (or close to it), then the possibilities of a one-way channel to
act as a two-way channel are small, hence, this measure could be not very effective in reducing the waiting time.
In other words, the effectiveness of this measure depends on the fleet composition.
Similarly, the second measure is suitable for fleets that have a widely varied draughts, combined with a
bathymetry that allows the navigation outside of the channel in a safe way. But, if the vessels’ draughts are
uniform or similar to the maximum, then the entrance position will probably be single, then, this measure will
have no effect on reducing the waiting time.
Finally, the speed reduction measure may help to reduce the waiting time of ships at the anchorage site.
Nevertheless, the navigation of a ship with a lower speed means that it will be delayed in comparison with the
same ship sailing at its maximum speed. Hence, the effective reduction of the waiting time can occur but only for
very specific situations (like a convoy slow – fast – slow vessel, see Figure 5.3).
70
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Inclusion of an intermediate passing section (or ‘buffer’ anchorage). With this measure, an extra
dredging is required but only in a certain section of the channel. The width will be such as a two-way
channel for all vessels, while the rest of the channel is maintained as a one-way channel (or with mixed
operation). Hence, more encounters will be permitted on the passing section.
Passing section
71
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 5.6 presents a scheme of the logical structure modification of the traffic call function. The
modifications are highlighted as coloured blocks, while the original algorithm is maintained with white
blocks.
Yes
Ans wer negative
No Di s tance Cha nnel 1- Yes Ans wer negative
s a fe? wa y?
Yes Yes No
Cha nnel 1- Yes Encounters Cha nnel 2-
wa y? >0? wa y?
No No Yes
Cha nnel 2-
Ans wer positive
Obta i n distance with No
wa y? overta ked s hip
Yes
Encounters > Di s tance Obta i n PIANC widths
Yes 1? No Yes s a fe? coefficients
Yes No
Obtai n minimum Wi dth > No
cha nnel wi dth requi red?
72
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
If at least one of the ships is found to be sailing in a position located at more distance than its entrance
point, the interaction does not exists.
Figure 5.7 presents a scheme of the logical structure modification of the traffic call function. The
modifications are highlighted as coloured blocks, while the original algorithm is maintained with white
blocks.
Defi ne ships to be Project position of Get s hip direction and Obta i n ships sailing
overtaked s hi ps pos ition towa rds the sea
Yes
Ans wer negative
No Di s tance Cha nnel 1- Yes Ans wer negative
s a fe? wa y?
Yes Yes No
Cha nnel 1- Yes Encounters Cha nnel 2-
wa y? >0? wa y?
No No Yes
Cha nnel 2-
Ans wer positive
Obta i n distance with No
wa y? overta ked s hip
Yes
Encounters > Di s tance Obtai n PIANC widths
Yes 1? No Yes s a fe? coefficients
Yes No
Obtai n minimum Wi dth > No
cha nnel wi dth requi red?
Figure 5.7. Modifications to the traffic call function – multiple entrances to the access channel
73
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Because of the reduction of the speed, a delay indicator (traffic reasons) is included in the ships
summary matrix. This delay has the same category than the delay related to ships waiting at the
anchorage because of congestion reasons.
Figure 5.8 presents a scheme of the modification to the logical structure of the main run module. The
modifications are highlighted as coloured blocks, while the original algorithm is maintained with white
blocks.
No
Crea te s hip i n ships Queue in No Ta ke fist ship in queue
s ummary empty? i n l ist
Yes
Incorporate s hip i n
Port ca l l ?
No
queue manager
Yes
Cha nge s hip s tatus to Incorporate s hip i n
Berth ca ll?
No
'queue out' s hi ps manager
No Yes
Obta i n new ship Check s hips geneation
Pi l ot call?
No Wri te delay i n s hip
pos ition l i st s ummary
No Yes
Queue out Yes La s t s hip i n Speed Yes Wri te utilisation in
empty? queue in? reducti on? a nchorage s ummary
No No
Ta ke fist ship in queue Shi p changes s tatus
outlist to 's a iling to port'
Port ca l l ?
No
Yes
Speed La s t s hip i n
Go to next s tep
reducti on? queue out? No
No
Shi p changes s tatus Dea ctivation of s hips
to 's a iling to s ea' 'l eaving port'
74
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 5.9 presents a scheme of the logical structure modification of the traffic call function. The
modifications are highlighted as coloured blocks, while the original algorithm is maintained with white
blocks.
Defi ne ships to be Project position of Get s hip direction and Obta i n ships sailing
overtaked s hi ps pos ition towa rds the sea
Yes
Ans wer negative
No Di s tance Cha nnel 1- Yes Ans wer negative
s a fe? wa y?
Yes Yes No
Cha nnel 1- Yes Encounters Cha nnel 2-
wa y? >0? wa y?
No No Yes
Cha nnel 2-
Ans wer positive
Obtai n distance with No
wa y? overta ked s hip
Yes
Encounters > Di s tance Obtai n PIANC widths
Yes 1? No Yes s a fe? coefficients
Yes No
Obta i n minimum Wi dth > No
cha nnel wi dth requi red?
75
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
4. Incorporation of variable ‘Wpp’ (value indicates the width of the passing section, in metres).
Defi ne ships to be Project position of Get s hip direction and Obtai n ships sailing
overta ked s hi ps pos ition towa rds the sea
Yes
Ans wer negative
No Di s tance Cha nnel 1- Yes Ans wer negative
s a fe? wa y?
Yes Yes No
Cha nnel 1- Yes Encounters Cha nnel 2-
wa y? >0? wa y?
No No Yes
Cha nnel 2-
Ans wer positive
Obta i n distance with No
wa y? overtaked s hip
Yes
Encounters > Di s tance Obta i n PIANC widths
Yes 1? No Yes s a fe? coefficients
Yes No
Obta i n minimum Wi dth > No Obta i n width at
cha nnel wi dth requi red? overta king location
76
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
5.2.5.1 Assumptions
The assumptions considered to model an intermediate anchorage area are:
Because of simplicity reasons on programming the code, a ship entering the intermediate anchorage
area will enter to it with its maximum speed.
The movement of the ship in the anchorage area is not modelled. Instead of it, the required time of a
vessel to move within the anchorage and do the manoeuvres to anchor and leave is 15 min.
All the delays related to a vessel which is using the intermediate anchorage exist because of the berth
unavailability. Therefore, these delays are classified as ‘berth causes’.
Because of the position of the intermediate anchorage (close to the port entrance), those ships that
leave the intermediate anchorage and sail to the port will have no sufficient length to overtake other
ships.
A simplification similar to the outer anchorage is considered: a ship leaving the intermediate anchorage
will enter the channel with its maximum speed.
The intermediate anchorage capacity is assumed to be unlimited.
77
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
No
Crea te s hip i n ships Queue in No Ta ke fist ship in queue
s ummary empty? i n l ist
Yes
Incorporate s hip i n
Port ca l l ?
No
queue manager
Yes
Cha nge s hip s tatus to Incorporate s hip i n
Berth ca ll?
No
'queue out' s hi ps manager
No Yes
Obta i n new ship Check s hips geneation
Pi l ot call?
No Wri te delay i n s hip
pos ition l i st s ummary
No Yes
Queue out Yes Queue in 2 Yes La s t s hip i n Speed Yes Wri te utilisation in
empty? empty? No queue in? reducti on? a nchorage s ummary
No No
Ta ke fist ship in queue Shi p changes s tatus
outl ist to 's a iling to port'
Port ca l l ?
No
Yes
Speed La s t s hip i n
Go to next s tep Tra ffi c call?
reducti on? queue out? No No
No Yes
Shi p changes s tatus Dea ctivation of s hips
Pi l ot call?
to 's a iling to s ea' 'l eaving port' No
Yes
Shi p changes s tatus to Wri te delay i n s hip Speed Yes
's a iling to port' s ummary reduction?
No
Modi fy s hip's speed
No Port ca l l ?
Wri te delay i n s hip Shi p changes s tatus
s ummary to 's a iling to i nt a nch'
No Yes
Tra ffi c call 2?
Yes Berth ca ll?
Shi p changes s tatus
Modi fy s hip's speed
to 's a iling to i nt a nch'
Yes No
Pi l ot call?
Yes Speed No Shi p changes s tatus Wri te delay i n s hip
reducti on? to 's a iling to port' s ummary
No Yes
La s t s hip i n Shi p changes s tatus
Modi fy s hip's speed
Yes queue? No to 's a iling to port'
78
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Defi ne ships to be Project position of Get s hip direction and Obtai n ships sailing
overta ked s hi ps pos ition towa rds the sea
Yes
Ans wer negative
No Di s tance
Ans wer negative
s a fe?
Yes Yes
Cha nnel 1- Yes Encounters Encounter >
Encounter
wa y? >0? entr? No
No No Yes
Cha nnel 2-
Ans wer positive No encounter
wa y?
Yes
Encounters >
Yes 1? No
Yes
Obta i n minimum Wi dth > No
cha nnel wi dth requi red?
Obta i n width at
encounter l ocation
79
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The ‘queue in 2’ function has the same structure than the ‘queue in’ and ‘queue out’ functions, which
were presented previously.
Figure 5.13. Simulated path of ships within testing scenario of a ‘mixed’ operation of a channel, a) with ‘mixed’
operation; b) one-way operation
According to this graph, ship A encounters with ship B (position 7,200 m approximately) and ship D (position
5,300 approximately). Nevertheless, the encounter with ship C is not allowed, hence, ship C is forced to wait at
the anchorage site during more than 30 minutes before obtaining authorisation to navigate towards the berth. If
the channel was operated as a one-way channel, ships B and D should wait at the anchorage area, increasing the
average waiting time.
80
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 5.14. Simulated path of ships within testing scenario: a) multiple entrances; b) single entrance
According to these graphs, the following commentaries are extracted:
For a situation where a multiple entrance scheme is accepted, some ships are using the channel only
partially (ships A, B and C). For these ships, no encounters are found: when the ships are in the same
position and at the same time, one or both ships are outside of the channel. This allows an earlier
81
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
authorisation to sail towards the port for ships C and D. The delay of ship D is because of the criteria of
minimum safe distance in a convoy.
In case that the channel is one way with a single entrance (outer edge of the channel), no encounters
can occur. Because all the ships will be forced to sail through the entire channel, the absence of
encounters can be obtained only if the ships that are using the channel are sailing in the same direction
(no encounters or overtaking). Only when the ships A and B have left the channel the conditions
become suitable for ships C and D to obtain a sequential authorisation.
Figure 5.15. Simulated path of ships within testing scenario: a) speed reduction available; b) speed reduction non
available
82
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
83
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
84
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 5.17. Simulated path of ships – intermediate anchorage: a) with intermediate anchorage; b) without
intermediate anchorage
Several commentaries are obtained from these figures:
For the situation with an intermediate anchorage:
o Ships B, C and D arrive to the outer channel when the berth is being occupied. When the traffic
conditions are suitable, those ships sail towards the intermediate anchorage.
o When these ships arrive to the intermediate anchorage, they maintain its position at that
location until the berth is free. That waiting time is because of berth reasons.
o The ship B stays at the berth more time than the service time, because when it finish its service
time, there are two ships (C and D) sailing towards the intermediate anchorage and the
encounters are not allowed for these beams. It can start to sail to the sea only when the traffic
conditions are suitable for it.
For the situation without an intermediate anchorage
o Ships B, C and D arrive to the outer channel when the berth is being occupied. Because of the
inexistence of an intermediate anchorage site, these ships are forced to wait at the outer
anchorage.
o Additionally, when the berth is freed, a vessel waiting at the anchorage cannot sail
immediately because of traffic reasons: it is authorised to sail only when the channel is empty,
which occurs only when the vessel that was using the berth leaves the access channel.
The entire scheme of waiting at the anchorage site generates a larger delay of the ships, because they
are allowed to enter into the channel after a complete cycle of sailing – service – sailing has been
finished. On the other hand, when an intermediate anchorage is available, ships can get closer to the
port when they are waiting for a berth. With this infrastructure, it may be allowed to sail when the
leaving ship have passed in front of the intermediate anchorage site: the final authorisation depends on
the traffic, weather and port conditions.
In this testing case, the difference in time for the last ship (D) to leave the port is more than 100 min.
5.4 Conclusions
The traffic rules measures of extension (mixed operability of the channel, multiple entrances and speed
reduction) are typical rules that can be applied by different port authorities. Hence, its implementation makes
the simulation model a more realistic tool, when it is compared to the basic model. For example, the idea of
85
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
forcing a river vessel or barge to sail across the entire access channel when it can enter directly into the port is
simply not realistic.
The enlargement of the nautical infrastructure is an improvement of the simulation tool, which allows a designer
to check the effect of different options of a conceptual design. The selected alternatives (width enlargement,
passing section and intermediate anchorage) are the most typical options of reducing the delay of ships using a
port. The implementation of these measures enlarges the capabilities of the simulation model, allowing the
assessment of extra configurations of the channel (not just a single width related to a one or two-way channel).
During the checking stage of this chapter, it was possible to observe that all proposed measures improve the
traffic conditions of the vessel by a reduction of the related delay. In many situations, ships that were not
allowed to sail under a typical configuration of one-way channel are now allowed, without the need of
implementing a two-way channel (the most expensive option because of the large amount of capital and
maintenance dredging works).
86
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
6 ANALYSIS OF RESULTS
The final stage in the development of the simulation model is the analysis of its results, which depend on
different parameters that are characteristics for each port and need to be set up. Hence, several simulations of
some fictitious ports were developed in order to estimate the real influence of these parameters in the delay on
an access channel due to traffic conditions.
Additionally, another topic that is important to analyse is the quantification of the real effect of the proposed
measures of optimisation. Hence, comparisons with and without the utilisation of the different available
measures are presented below.
This chapter is divided in the following sections: definition of the degree of dependence of the traffic delay on
some input parameters and quantification of the real effect of the available measures on reducing the traffic
delay.
87
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Capacity of the channel: if the capacity of a channel (defined as the maximum number of ships that sail on the
channel during a small period of time) decreases, then the required time to ‘evacuate’ a certain queue (arriving
or departing lists) will be higher, this will finally lead to a higher delay of the ships on the queues.
Availability of the channel: if the channel is not operating during a certain period of time, all ships that will arrive
to some of its edges will have to wait until the channel will be re-opened. Hence, the higher the channel will be
not operating, the higher the delay that will be obtained.
88
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
berths within the port). Figures 6.1 and 6.2 present the results of the average and maximum waiting time,
respectively.
Figure 6.1. Average traffic delay for a varying number of ships per year, one – way channel
Figure 6.2. Maximum traffic delay for a varying number of ships per year, one – way channel
The proposed fitting curves are power type.
2
The coefficients R for each adjustment are the following: 0.998 and 0.982 (average waiting time, arriving and
departing queues); and 0.959 and 0.892 (maximum waiting time, arriving and departing queues). These
coefficients are high, especially those related with the average waiting time, hence the proposed relationships
are valid within the tested range (500 to 7,000 ships per year).
The observation of these figures returns a notorious difference on the waiting time (average and maximum)
between the arriving and the departing ships. The difference can be explained by the probabilistic distribution of
each process. While the arriving ships follow a NED distribution, the departing ships follow a different
distribution, which is obtained from the sum of a NED distribution (arriving), two uniform distributions (sailing to
and from the port) and a k-Erlang distribution (service at the port). Especially when the k-Erlang function is
89
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
considered, it will tend to reduce the number of ships with a short inter-arrival time: this reduce congestion and
diminishes the waiting time.
Figure 6.3. Average traffic delay for a varying length of the outer channel, one – way channel
Figure 6.4. Maximum traffic delay for a varying length of the outer channel, one – way channel
The proposed fitting curves are power type. Relations for the average and maximum waiting time of the arriving
ships are presented for three different numbers of ships (1,000, 3,000 and 5,000 ships/year). Some remarks can
be obtained from it:
For the average waiting time, the three fitting curves are power type with coefficients in the range 0.92
to 1.22. Then, the coefficients are similar and a unique coefficient is proposed, which is considered as
90
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
the average of the three values, being equal to 1.04. Hence, when the extent of the channel increases,
the average waiting time increases almost proportionally.
For the maximum waiting time, the three fitting curves are power type with coefficients in the range
0.63 to 1.06. The coefficients have a larger dispersion than in the case of the average waiting time,
probably because of the nature of the maximum waiting time. Nevertheless, a similar adjustment is
proposed, obtaining a unique coefficient which is the average of the coefficients of each curve, being
equal to 0.84.
Figure 6.5. Average traffic delay for a varying capacity of the access channel, one – way channel
The proposed fitting curves are power type. Relations for the average and maximum waiting time of the arriving
ships are presented for three different numbers of ships (1,000, 3,000 and 5,000 ships/year). Some remarks can
be obtained from it:
For the average waiting time, the three fitting curves are power type with coefficients that decreases
when the number of ships increases (from -0.34 to -1.55). Then, the representative coefficient for
estimating the traffic delay with a varying capacity depends on the number of ships. A power
adjustment is proposed and presented in equation 1.7.
91
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
For the maximum waiting time, the three fitting curves are power type with coefficients in the range
0.14 to 0.31. The coefficients have a larger dispersion than in the case of the average waiting time,
probably because of the nature of the maximum waiting time. Nevertheless, a similar adjustment is
proposed, obtaining a unique coefficient which is the average of the coefficients of each curve, being
equal to 0.25.
Figure 6.6. Maximum traffic delay for a varying capacity of the access channel, one – way channel
The proposed adjustment coefficients for estimating the effect of changes on the channel’s capacity were
estimated on a short number of simulations. Hence, it is recommended to execute further research in order to
define if the proposed relationship can be confirmed or modified.
Figure 6.7. Average traffic delay for a varying availability of the access channel, one – way channel
92
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 6.8. Maximum traffic delay for a varying availability of the access channel, one – way channel
The proposed fitting curves are power type. Relations for the average and maximum waiting time of the arriving
ships are presented for two different numbers of ships (1,000 and 3,000 ships/year). Some remarks can be
obtained from it:
For the average waiting time, the two fitting curves are power type with coefficients in the range -0.89
to -1.01. Then, the coefficients are similar and a unique coefficient is proposed, which is considered as
the average of the two values, being equal to -0.95.
For the maximum waiting time, the two fitting curves are power type with coefficients in the range 0.42
to 0.76. Then, the coefficients are similar and a unique coefficient is proposed, which is considered as
the average of the two values, being equal to 0.59.
The proposed adjustment coefficients for estimating the effect of changes on the channel’s availability were
estimated on a small number of simulations. Hence, it is recommended to execute further research in order to
define if the proposed relationship can be confirmed or modified.
Considering all the fitting curves for the different parameters defined previously, two relationships (average and
maximum waiting time) are obtained for a one – way channel.
93
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
leaving their berths within the port). Figures 6.9 and 6.10 present the results of the average and maximum
waiting time, respectively.
Figure 6.9. Average traffic delay for a varying number of ships per year, two – way channel
Figure 6.10. Maximum traffic delay for a varying number of ships per year, two – way channel
The proposed fitting curves are potential.
2
The coefficients R for each adjustment are the following: 0.977 and 0.970 (average waiting time, arriving and
departing queues); and 0.914 and 0.918 (maximum waiting time, arriving and departing queues). These
coefficients are high, especially those related with the average waiting time, hence the proposed relationships
are valid within the tested range (4,000 to 22,000 ships per year).
Similarly to the situation of a one – way channel, it is possible to observe that there is a difference between the
waiting time for arriving and departing ships. The reasons of this difference are the same than the case of a one
– way channel (differences on the distribution function for each process).
94
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 6.11. Average traffic delay for a varying length of the outer channel, two – way channel
Figure 6.12. Maximum traffic delay for a varying length of the outer channel, one – way channel
The proposed fitting curves are power type. Relations for the average and maximum waiting time of the arriving
ships are presented for three different numbers of ships (6,000, 12,000 and 18,000 ships/year). Some remarks
can be obtained from it:
For the average waiting time, the three fitting curves are power type with coefficients in the range 0.45
to 0.58. Then, the coefficients are similar and a unique coefficient is proposed, which is considered as
the average of the three values, being equal to 0.49. Hence, when the extent of the channel increases,
the average waiting time increases with a power law close to ½.
95
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
For the maximum waiting time, the three fitting curves are power type with coefficients in the range
0.85 to 0.86. The coefficients are almost equal; hence, a unique coefficient of 0.85 is selected.
For the set of data of 18,000 ships/year, the maximum delay related to a channel extension of 5 km is not
considered for estimating the fitting curve.
Figure 6.13. Average traffic delay for a varying capacity of the access channel, two – way channel
Figure 6.14. Maximum traffic delay for a varying capacity of the access channel, two – way channel
The proposed fitting curves are power type. Relations for the average and maximum waiting time of the arriving
ships are presented for three different numbers of ships (6,000, 12,000 and 18,000 ships/year). Some remarks
can be obtained from it:
96
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
For the average waiting time, the three fitting curves are power type with coefficients that decreases
when the number of ships increases (from -0.71 to -2). Then, the representative coefficient for
estimating the traffic delay with a varying capacity depends on the number of ships. A power
adjustment is proposed and presented in equation 6.17.
For the maximum waiting time, the three fitting curves are power type with coefficients in the range -
0.87 to -2.09. The coefficients have a larger dispersion than in the case of the average waiting time,
probably because of the nature of the maximum waiting time. Nevertheless, a similar adjustment is
proposed, obtaining a unique coefficient which is the average of the coefficients of each curve, being
equal to -1.37.
Similarly to the case of a one – way channel, the proposed adjustment coefficients for estimating the effect of
changes on the channel’s capacity were estimated on a short number of simulations. Hence, it is recommended
to execute further research in order to define if the proposed relationship can be confirmed or modified.
Figure 6.15. Average traffic delay for a varying availability of the access channel, one – way channel
The proposed fitting curves are power type. Relations for the average and maximum waiting time of the arriving
ships are presented for two different numbers of ships (6,000 and 12,000 ships/year). Some remarks can be
obtained from it:
For the average waiting time, the two fitting curves are power type with coefficients contained in the
range -6 to -6.2. Then, the coefficients are similar and a unique coefficient is proposed, which is
considered as the average of the two values, being equal to -6.15.
For the maximum waiting time, the two fitting curves are power type with both coefficients equal to -
4.3.
97
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 6.16. Maximum traffic delay for a varying availability of the access channel, one – way channel
The proposed adjustment coefficients for estimating the effect of changes on the channel’s availability were
estimated on a short number of simulations. Hence, it is recommended to execute further research in order to
define if the proposed relationship can be confirmed or modified.
Considering all the fitting curves for the different parameters defined previously, two relationships (average and
maximum waiting time) are obtained for a one – way channel.
98
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
In order to verify the results obtained by the proposed relationships, a comparison between the delay (average
and maximum) obtained by the simulator and the formulas is done. The comparison is presented in Figure 6.17.
Figure 6.17. Comparison between delay time obtained by simulation model and relationships
In these figures, a dashed line is incorporated: it represents a perfect fitting of the delay obtained with the
proposed relationships. The better the relationship, the closer the points has to be in relation with the dashed
line. For the average delay, it is possible to see that all the points are close to the line, with a maximum
difference (vertical distance) lower than 5 min. For the maximum delay, it is possible to see a larger dispersion of
the data. The maximum difference (vertical distance) is lower than 50 min.
2
The coefficients R for the sets of data related with the average and maximum delay are 0.986 and 0.905,
respectively. These numbers (especially the coefficient for the average delay) are high; hence, the proposed
relationships are valid. It must be emphasized that the testing of the formulas has been executed for a small
number of data and only within the range of the parameters for which the simulations were performed. Then,
further research is recommended in order to confirm the validity of these relations, and also to check if the
results can be extrapolated.
Independently on the good correlations obtained for both formulas, it is important to remark that due to the
methodology used to obtain the different results (only one simulation for each condition), the reliability of these
formulas can be defined as limited, then, and further research is strongly advised.
99
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
All the verifications are made for an original one – way channel, because the scope of these measures is to
reduce the traffic delay without needing to expand the channel to a two – way channel.
It has to be emphasized that all of these verifications are made for some particular scenarios, which does not
cover all the possibilities. Hence, the results must be considered as quick and rough recommendations of which
measures are more suitable for a given scenario. Nevertheless, it is always advisable to execute simulations for a
particular case of study, in order to obtain more accurate results.
The characteristics of the vessels of both fleets are presented in Table 6.3.
Table 6.3. Vessels’ characteristics
Ship’s length Ship’s beam Ship’s draught Ship speed
Fleet
[m] [m] [m] [kn]
A 304 40 14.2 12
B 182 23 9.5 12
The quay length for both vessels is considered unlimited, hence, there is no waiting time related to a berth
unavailability.
With a mixed operation the channel can operate as a one or two – way channel, depending on the vessels’
characteristics and the channel’s dimensions. Within this scenario, encounters may or may be not allowed for
different vessels. Table 6.5 presents a summary with the allowance of encountering for the different vessels.
Table 6.5. Summary of encountering for different vessels
Entering ship / departing ship Fleet A Fleet B
Fleet A No No
Fleet B No Yes
100
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figures 6.18 and 6.19 present the average waiting time due to traffic reasons for the different scenarios, with
and without the allowance for ‘mixed’ operability of the channel.
Figure 6.18. Average waiting time with and without ‘mixed’ operability of an access channel – Fleet A
Figure 6.19. Average waiting time with and without ‘mixed’ operability of an access channel – Fleet B
From the results presented in these figures, the following commentaries are extracted from:
The average waiting time for fleet A does not present a significant reduction of its value when the
measure is applied. Only for the third scenario (lowest number of ships on fleet A) there is a reduction
close to a 25% of the original waiting time, but for the other scenarios the results are maintained.
The average waiting time for fleet B presents important reductions of the waiting time, which are larger
depending on the number of ships present on fleet B. Scenario 1 (1,000 ships) presents a reduction of
about 10%, scenario 2 (2,000 ships) increases the reduction to 55% while scenario 3 (3,000 ships)
returns a reduction of the waiting time of 75%.
The differences on the observed waiting time for both fleets are explained in terms of the encounter allowance
(ships and channel characteristics) and what is the percentage of encounters that are allowed. For the vessels of
the fleet A, they cannot encounter neither with ships from fleet A or B. Hence, the presence of a ‘mixed’
101
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
operation of the channel has no effect on these ships, because the channel operates as a one – way channel for
all situations.
On the other hand, the ships of fleet B can encounter with other ships from fleet B. When the number of ships of
the fleet B increases, the number of allowed encounters also increases. For the scenario 1, it is estimated that
only an 8% of the encounters will be allowed; this number rises to a 33% for scenario 2 and a 69% for scenario 3.
Hence, if for a certain port a mixed operation will lead to a high number of allowed encounters, the reduction of
the average waiting time will be more significant, but only for those vessels that can interact with other vessels
as a two – way channel. Then, the effectiveness of this measure is better for smaller ships.
Then, ships from fleet A will enter the channel through the entrance point 1 while ships from fleet B will enter
the channel through the entrance point 2. This means that ships from fleet A will have to sail the entire length of
the outer channel but ships from fleet B will only sail half of it.
Figures 6.20 and 6.21 presents the average waiting time due to traffic reasons for the different scenarios, with
and without the allowance multiple entrance points.
From the results presented in these figures, the following commentaries are extracted from:
When multiple entrance points are available, the average waiting time for fleet A present important
reductions of its value (31% to 61%, when the number of ships entering at a middle position is 25% to
75%, respectively). The reduction is higher when the number of ships that can enter through an
intermediate entrance points increases.
When the multiple entrance points are available, the average waiting time for fleet B present important
reductions of its value, which are around 80%. The observed reduction factors are similar for all the
scenarios; hence, it does not depend on the number of ships.
102
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 6.20. Average waiting time with / without multiple entrance points to an access channel – Fleet A
Figure 6.21. Average waiting time with / without multiple entrance points to an access channel – Fleet B
Both fleets present reductions of their average waiting time. Nevertheless, while the fleet A shows a progressive
reduction (dependent on the number of ships that can enter through an intermediate position of the channel);
the fleet B presents an almost constant (and higher) reduction.
The reason of the progressive increment on the reduction of the waiting time for fleet A is related with the
existence of ships sailing on the opposite direction and the possibility of occurrence of an encounter within the
access channel. If the number of ships entering to the channel on a middle position increases, the possibilities of
having encounter reduce and the ships from fleet A have less restrictions to sail. In simple words: the utilisation
of the channel is reduced and the delay decrease.
In the case of ships from fleet B the reduction of the waiting time is practically constant. All these ships sail
through the channel only on a part of it; hence, the possibilities of having encounters are limited. But, because
the channel is a one – way channel, the encounters are not allowed with ships from fleet A nor B. Because the
part of the channel that ships from fleet B uses is occupied by ships from both fleets, then independently on the
103
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
distribution of the total number of ships, encounters must be avoided. Then, for this fleet, the reduction of the
waiting time does not depend on the number of ships of this fleet.
The results present a reduction of the delay for both fleets (large and small vessels). For small vessels the
reduction is particularly large, while for large vessels it depends on the fleet composition. But the
implementation of this measure present reduction of the delay for all fleets, hence, it is highly recommended to
execute it when there are large differences between the largest and smallest ship of a fleet.
The quay length for both vessels is considered unlimited, hence, there is no waiting time related to a berth
unavailability.
Figures 6.22 and 6.23 present the average waiting time due to traffic reasons for the different scenarios, with
and without the speed reduction on the channel.
Figure 6.22. Average waiting time with and without speed reduction on the access channel – Fleet A
104
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 6.23. Average waiting time with and without speed reduction on the access channel – Fleet B
From the results presented in these figures, the following commentary is extracted from:
Both fleets present similar average waiting times for the different scenarios. Hence, no significant
reductions of the traffic delay are observed.
The most possible explanation for the absence of reduction of the waiting time on both fleets (fast and slow
ships) is related with the delay effect of a ship which is sailing with a slower speed than its maximum. When a
fast ship is being preceded by a slow ship, the speed reduction measure indicates that the fast ship will reduce
its speed in order to do not reduce the distance with the slow ship. But, the speed reduction is traduced into a
later arrival time to the assigned berth. That time is comparable to the time that the ship should wait at the
anchorage before receiving authorisation to sail with its normal speed.
One advantage of this measure is the lower utilisation of the anchorage site (for ships waiting due to traffic
conditions): a ship sailing with a slower speed will not enter the anchorage. Nevertheless, the lower utilisation of
the anchorage site refers only for ships waiting due to traffic conditions. In case that a berth is not available, or
there are storm conditions that block the access to the port, then the anchorage site will be used by those ships.
Number of ships Ship’s length Ship’s beam Ship’s draught Ship dwt
Fleet
[ships/yr] [m] [m] [m] [kton]
A 1,000 280 45 17.5 161
B 1,000 230 36 12.7 77.4
C 1,000 170 27 10 30
D 1,000 91 12 4.1 2.2
105
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The channel’s characteristics (excepting the width) are the same presented in Table 6.4.
Table 6.9 presents the width of the access channel for the different scenarios.
Table 6.9. Channel’s width for different scenarios
Channel width
Scenario
[m]
1 150
2 175
3 190
4 210
5 250
The quay length for both vessels is considered unlimited, hence, there is no waiting time related to a berth
unavailability.
Tables 6.10 to 6.14 present the encounter allowance between ships of the different fleets, considering a varying
width of the channel.
Table 6.10. Summary of encountering allowance for different vessels – Scenario 1
106
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The scenario 5 corresponds to an operation with a two – way channel (all encounters are permitted).
Figure 6.24 presents the average waiting time due to traffic reasons for the different scenarios.
Figure 6.24. Average waiting time of different fleets under different scenarios
From the results presented in this figure, the following information is highlighted:
Fleet A: there is an important decrease of the waiting time, from 14.8 min (scenario 1) to 0.8 min
(scenario 5), totalizing a delay reduction of 91%. Local reductions are observed between scenarios 1 – 2,
3 – 4 and 4 – 5.
Fleet B: there is an important decrease of the waiting time, from 9.4 min (scenario 1) to 0.6 min
(scenario 5), totalizing a delay reduction of 94%. Local reductions are observed between all scenarios.
Fleet C: there is an important decrease of the waiting time, from 7.6 min (scenario 1) to 0.2 min
(scenario 5), totalizing a delay reduction of 97%. Local reductions are observed between scenarios 1 – 2
and 3 – 4.
Fleet D: there is an important decrease of the waiting time, from 2.9 min (scenario 1) to 0.4 min
(scenario 5), totalizing a delay reduction of 88%. The most important local reduction is observed
between scenarios 1 – 2.
When a comparison is made between the results (Figure 6.24) and the encounter allowance (Tables 6.10 to
6.14), it is possible to notice that each local reduction is obtained when the width of the channel is increased to
a value that allows a new encounter. If the new width is not sufficient (or if encounters are allowed with all the
fleets), then there is no effective reduction of the waiting time.
107
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
In conclusion, the implementation of an increase of the channel depends strongly on the fleet composition,
specifically on the vessel’s dimensions. Then, in order to generate a positive effect on a specific fleet (reduction
of the delay due to traffic conditions), one or various new encounters must be allowed.
Outer edge
Width Length
Scenario position
[m] [m]
[km]
1 250 - -
2 250 2,500 3.75
3 250 5,000 2.5
4 250 7,500 1.25
5 250 10,000 -
The first scenario corresponds to a channel with a uniform width of 150 m, with no passing section. The fifth
scenario corresponds to a channel with a uniform width of 250 m, which is equivalent to a two – way channel.
The quay length for both vessels is considered unlimited, hence, there is no waiting time related to a berth
unavailability.
Figure 6.25 presents the average waiting time due to traffic reasons for the different scenarios.
Figure 6.25. Average waiting time of different fleets under different scenarios
108
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
From the results presented in this figure, the following information is highlighted:
Fleet A: reductions of the waiting time due to a passing section are observed. The reductions are in the
range from 71% (2,500 m length) to 93% (7,500 m length).
Fleet B: reductions of the waiting time due to a passing section are observed. The reductions are in the
range from 66% (2,500 m length) to 87% (7,500 m length).
Fleet C: reductions of the waiting time due to a passing section are observed. The reductions are in the
range from 67% (2,500 m length) to 91% (7,500 m length).
Fleet D: reductions of the waiting time due to a passing section are observed. The reductions are in the
range from 65% (2,500 m length) to 80% (7,500 m length).
It is possible to observe that in general the delay reduction is large for all the fleets, even considering the
simulation of the smallest length of the passing section. Nevertheless, if the length of the passing section is
increased, the observed effect on the delay becomes smaller. Hence, it seems to be not very effective to enlarge
the passing section excessively, because the reduction of the delay becomes less and less significant.
Another feature that is observed is a slight difference on the delay reduction between fleets with larger and
smaller vessels. While the dimensions of the vessels decrease, the delay reduction also decreases. This effect is
related to the allowance of encounters for the smaller vessels with other ships, even on the normal section of
the channel. For these ships, the presence of a passing section has a smaller impact than on large ships, which
only can encounter with other ships on the passing section.
Number of ships Ship’s length Ship’s beam Ship’s draught Ship dwt
Fleet
[ships/yr] [m] [m] [m] [kton]
A 500 280 45 17.5 161
B 500 230 36 12.7 77.4
C 500 170 27 10 30
D 500 91 12 4.1 2.2
The service time characteristics for each fleet are presented in Table 6.17.
Table 6.17. Service time characteristics per fleet
109
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The quay length and number of berths for each fleet are presented in Table 6.18.
Table 6.18. Service time characteristics per fleet
Figure 6.26 presents the average waiting time due to traffic reasons for the different scenarios.
Figure 6.26. Average waiting time of different fleets under different scenarios
Figure 6.27 presents the utilisation of both anchorage sites with and without the presence of an intermediate
anchorage site.
From the results presented in these figures, the following information is highlighted:
The delay reduction does not present a clear trend. For the largest ships (fleet A), an important increase
of the waiting time due to traffic conditions is observed, but for fleet C a reduction close of 50% of the
delay is observed.
The utilisation of the outer anchorage site is higher when there is no intermediate anchorage. But,
when the intermediate anchorage is available, almost a 100% of the time the outer anchorage site is
used.
110
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
6.3 Conclusions
From this chapter, the following conclusions are obtained:
111
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
proportional to the square root of the length increase. The difference on both coefficients is explained in terms
of the processes of obtaining delay for both channels:
For a one – way channel, the delay of ships (due to traffic conditions) can occur from two sources. First, the ships
that are waiting in a queue to enter the channel must wait until all ships sailing in the opposite direction have
left it (because encounters are not allowed on it). If the length of the channel increases, the waiting time at the
queue will increase too. Secondly, when a ship is sailing through the channel being preceded by a slower ship, it
will be forced to reduce its speed or wait at the queue until the conditions on the channel allows navigation
without reducing excessively the distance between the ships. Independently on which is the condition
(dependant on the traffic rules of the channel), the second ship will sail slowly, increasing the delay.
For a two – way channel, encounters are allowed and then there is no delay because the channel is being used
by ships sailing in the opposite direction. Nevertheless, when a slower ship is sailing in front of another one and
the speed difference is not significant, the possibilities of having an overtaking are reduced because of the
required distance to do it. Then, it is possible that the faster ship will have to reduce its speed, adding delay.
Because for a one – way channel the delay occur due to more sources, the dependence of the delay on the
channel’s length is more significant than in the case of a two – way channel.
The average waiting time depends on the capacity of the access channel, which is related to the minimum safety
distance between ships, the ships dimensions (length) and the speed through the inner channel. The proposed
adjustments show the inverse dependence of the traffic on the channel’s capacity: the higher the channel’s
capacity, the smaller the delay. Nevertheless, there is a dependence also on the number of ships sailing: when
the number is higher, the effect of the capacity becomes more important and vice versa.
This can be explained by the hypothesis that the capacity is a parameter related with the ‘ability’ of a channel to
‘evacuate’ queues: if capacity is higher, the queue will be emptied rapidly.
In order to confirm or discard this hypothesis, further research should be developed. This research should
consider a larger number of simulations and its comparison with real life data. Additionally, focus should be paid
to the behaviour of the queues within the simulation time.
112
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
intermediate entrance point for a specific draught is moved towards the outer edge of the channel. If that is the
case, the smaller vessels will have to sail through a larger distance, reducing the effectiveness of this measure.
The speed reduction measure results do not show a reduction of the average waiting time. The only advantage
of this measure is a lower utilisation of the anchorage site (because ships that have impediments due to traffic
conditions can enter into the channel with a slower speed). Nevertheless, the lower utilisation is only due to
traffic conditions: if there is berth unavailability or storm conditions, then the anchorage site will be used by
ships waiting for an authorisation. Hence, the real effect of reducing the anchorage utilisation may not be
significant.
The implementation of a passing section has a positive effect of reducing the delay due to traffic reasons on all
the fleets (slightly larger for fleets with larger vessels). According to the results, for an increasing length of the
passing section, the effect on the delay reduction becomes more inefficient. But, the definition of a passing
section as a two – way channel type means that it generates a benefit for all ships, independently on the fleet
composition. Then, for a uniform or quasi-uniform fleet, is more likely to implement a passing section instead of
an increase of the channel’s length. Like an increase of the channel’s width, the evaluation of this measure must
consider the effect on the cost reduction due to a lower waiting time versus the increase of the capital and
maintenance costs of dredging.
The implementation of an intermediate anchorage site does not seem to have a positive effect on reducing the
delay due to traffic conditions. Instead of it, the intermediate anchorage site implementation obeys to the
requirements of more complex ports (like Rotterdam), where a several areas of anchorage are available.
Another possibility of implementing an intermediate anchorage site is a restriction of the maximum surface of
the outer anchorage site: the presence of another site allows the distribution of ships between these areas,
reducing the occupation of the outer anchorage.
113
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Kerch Strait
Taman Seaport
Black Sea
Figure 7.1. Projected location of the Taman Seaport (Google Earth)
114
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Coal (export)
Grain (import / export)
Iron ore (export)
Sulphur (export)
Mineral fertilizer (import / export)
Steel products (export)
Table 7.1 presents a summary with the port requirements projected for the year 2025 (final throughput).
Table 7.1. Summary of port requirements (Witteveen + Bos, 2012)
In order to handle the different cargo presented previously, several terminals are considered. Figure 7.2
presents a layout of the Taman Seaport, presenting the location of each terminal.
Container Container
terminal 2 terminal 1
Steel terminal
Grain terminal
115
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Quay length
Terminal
[m]
Container terminal 1 835
Container terminal 2 685
Steel terminal 885
Grain terminal 1,175
Coal terminal 1 570
Coal terminal 2 655
Iron ore terminal 835
Mineral fertilizer terminal 1 835
Mineral fertilizer terminal 2 590
Sulphur terminal 590
Projections of the total number of vessels per year indicates that during the initial throughput (2017) the
expected number of yearly calls will be 2,689 ships/year, while for the final throughput (2025) the total yearly
calls will be 3,503 ships/year. The details of the expected number of ships per terminal and per fleet are
presented in Appendix D.
116
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Service rate
Terminal Type Operation
[t/hr] or [moves/hr]
Container 1 Container Import / export
720 moves/day/crane
Container 2 Container Import / export
Coal 1 Dry bulk Export
3,000 tons/hr
Coal 2 Dry bulk Export
960 tons/hr (import)
Steel General cargo Import / export
960 tons/hr (export)
1,500 tons/hr (import)
Grain Dry bulk Import / export
3,000 tons/hr (export)
Mineral fertilizer 1 Dry bulk Export 3,000 tons/hr (export)
1,500 tons/hr (import)
Mineral fertilizer 2 Dry bulk Import / export
3,000 tons/hr (export)
Iron Dry bulk Export 5,000 tons/hr
1,500 tons/hr (import)
Sulphur Dry bulk Import / export
3,000 tons/hr (export)
For the container terminals, the service rate is expressed in a number of moves (import or export) per hour per
crane working on a single ship. Hence, the number of cranes defines the final service rate of these terminals.
Depending on the ships’ capacity (expressed in TEUs), a different number of cranes will operate on it. It is
assumed that for a ship with capacity larger than 3,500 TEUs, 3 cranes will operate on it. If the capacity is
between 1,000 TEUs and 3,500 TEUs, 2 cranes will operate on it. Finally, if the ship is small (less than 1,000
TEUs,), only one crane will operate on it.
7.1.2.3 Currents
Currents are generally small (less than 0.3 m/s near the project location).
117
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
7.1.2.4 Waves
Table 7.5 presents a histogram with the significant wave height against the mean wave direction for the area
surrounding the project location.
Table 7.5. Significant wave against wave direction (Witteveen + Bos, 2012)
7.1.2.5 Tides
Due to the location of the project (Black Sea), the tidal variations are very small and can be neglected.
118
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figures 7.4 and 7.5 present a plan view and three sections of the access channel.
Figure 7.4. Plan view of the Taman’s access channel (Witteveen + Bos, 2012)
Figure 7.5. Cross sections of the Taman’s access channel (Witteveen + Bos, 2012)
The estimation of the channel width for the different sections depends on the definition of several parameters
(PIANC’s rules).
Manoeuvrability: moderate (for bulk carriers sailing through the access channel).
Vessel speed: moderate (8 – 12 kn).
Prevailing cross wind: moderate (>15 – 33 kn, Beaufort 4 – 7).
Prevailing cross current: moderate (0.5 – 1.5 kn).
Prevailing longitudinal current: moderate (1.5 – 3 kn).
Significant wave height and length: between 1 to 3 m, and wave length similar to the length of the ship.
119
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Aids to navigation: good (navigation marks which are well visible by naked eye or radar).
Bottom surface: depth smaller than 1.5 ship’s draught, smooth or sloping and hard.
Depth of the waterway: smaller than 1.25 the ship’s draught.
Cargo hazard level: low (dry bulk).
Additional width (bank clearances): sloping channel edges and shoals.
The average service time and the normalized average time for each terminal are presented in Table 7.8.
120
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Table 7.8. Average service time and normalized average waiting time – initial throughput
Figure 7.6 presents the distribution of the average waiting time, which is classified depending on the reason of
the delay (port unavailability, berth unavailability, congestion or weather conditions).
Figure 7.6. Average waiting time classification per terminal – initial throughput (units in minutes)
In any of these figures is possible to see that the delay is because of port unavailability. The reason is that the
channel operates 24 hours, there is no tidal window, the currents are negligible and the probability of downtime
of the port due to other reasons is very small (0.01%).
In some of the terminals, the berth unavailability appears to be the major cause of the delay (container
terminals 1 and 2, coal terminals 1 and 2). On some other terminals, when the berth availability is high, the
congestion becomes the main cause of delay. Nevertheless, it must be emphasized that the values of congestion
are contained in the range 9 to 16 minutes (with average of 13 minutes), which suggests that the waiting time
due to congestion do no depend strongly on the berth availability or the characteristics of the vessels, but it
does on the total number of ships per year that are actually using the access channel.
121
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The typical accepted values for the normalized average waiting time are between 5 – 10% (for container
terminals), and 15 – 20% (for dry bulk and general cargo terminals). The results for the initial throughput show
that all the terminals have an acceptable waiting time, then, no modifications for this scenario are required.
The average service time and the normalized average time divided by the average service time for each terminal
are presented in Table 7.10.
Table 7.10. Average service time and normalized average waiting time – final throughput
Figure 7.7 presents the distribution of the average waiting time, which is classified depending on the reason of
the delay (port unavailability, berth unavailability, congestion or weather conditions).
122
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 7.7. Average waiting time classification per terminal – final throughput (units in minutes)
Similarly to the scenario with the initial throughput, in any of these figures is possible to see that the delay is
because of port unavailability.
The terminals where it is possible to observe that the main cause of delay is due to berth unavailability are the
container terminals, the coal terminals, the iron terminal and the mineral fertilizer 2 terminal. On the rest of the
terminals, the berth availability stills high; hence, the traffic is the major cause of delay.
The congestion values are higher than during the initial throughput. For this scenario, the values of congestion
are container in the range 15 to 33 minutes (with average of 20 minutes). The increase of congestion is
explained because of the larger number of ships using the access channel during one year.
The typical accepted values for the normalized average waiting time are between 5 – 10% (for container
terminals), and 15 – 20% (for dry bulk and general cargo terminals). From Table 7.13 it is possible to observe that
the container terminal 1 has a related waiting time which is close to the maximum accepted value. Furthermore,
the coal terminal 1 has a waiting time that is almost within the accepted range. In view of these results, a deeper
analysis of the causes of delay on these two terminals is required.
For the rest of the terminals the normalized average waiting time present acceptable values.
123
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The service rate for this scenario is a 50% higher than the original assumption for Taman (see Table 7.3).
Figure 7.8 presents a comparison for the scenario with normal and high service rate, presenting the average
delay obtained for the different terminals (for berth and traffic reasons). The weather delay is not presented
because it does not depend on the berth or congestion conditions.
Figure 7.8. Comparison of delays between scenarios with different service rate, a) berth delay; b) traffic delay
From Figure 7.8 it is possible to observe that with a high service rate, the berth delay has diminished for all
terminals (from the situation with the normal service rate). Nevertheless, the traffic delay remains
approximately constant, independently on the faster service time induced by a high service rate. Then, it is
possible to conclude that the traffic delay is not directly affected by the utilisation of the berth. Then, for the
next steps in the Taman Seaport assessment (incorporation of optimisation measures), only the original service
rate will be considered.
124
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
410 (increment of 65%). For more details on the number and type of ships during both scenarios, see Appendix
D.
Figure 7.9. Comparison of causes of delay for initial and final throughput – Container Terminal 1
Finally, the design of the access channel can have an effect on the average waiting time due to congestion.
Hence, some measures can be applied in order to reduce it. Nevertheless, the reduction of the waiting time via
the optimisation of the access channel will have an effect only on the traffic delay, not on the berth availability.
Then, the possibilities of reducing the total waiting time (without modifying the quay length or the service rate)
are limited to a reduction of the traffic delay. Then, the maximum reduction of the waiting time is 23 minutes,
obtaining an ideal average waiting time of 119 minutes and a normalized average waiting time of 7.9%.
If the normalized waiting time is needed to be reduced to a value smaller than 5%, then a modification of the
master plan is required. Nevertheless, that modification is out of the scope of this research.
125
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
with a higher capacity (150 ktons). Then, the utilisation of the berth is higher, which is traduced into a higher
delay due to berth unavailability.
Figure 7.10. Comparison of causes of delay for initial and final throughput – Coal Terminal 1
The increase of the congestion is explained because of the larger number of ships per year that will use the
access channel. This number takes into account all fleets related with all the terminals of the port, not just the
coal terminal 1.
If a reduction of the waiting time via an optimisation of the access channel, the reduction will be effective on the
traffic delay. Then, for a total delay of 48 minutes, the maximum reduction of the waiting time is 27 minutes,
obtaining an ideal average waiting time of 27 minutes and a normalized average waiting time of 8.2%.
7.4 Conclusions
The available information of the Taman Seaport is sufficient to set up the simulation model and obtain results
regarding the waiting time and the performance of its access channel. Nevertheless, a better definition of the
service rate for the different terminals would be ideal. Independently on this, a change on the service rate will
modify the utilisation of the berths, having an effect on the waiting time but only due to berth reasons. Because
the scope of this work is to assess the access channel, a modification of these rates will not have a direct effect
on the congestion.
The results of the simulation for the Taman Seaport indicate that for the final stage of the port (year 2025), there
are two ports that have a waiting time close to the typically accepted values of delay: the container terminal 1
and the coal terminal 1. The reasons behind the observed delays and the opportunities to improve the
performance of these terminals are different:
For the container terminal 1, the large average waiting time (9.5% of the service time) is explained
because of the berth unavailability, which is an 83% of the total waiting time. The congestion
contributes to this delay, but only in a small part. Then, an optimisation of the access channel will have
a small effect on reducing this delay: a modification of the master plan (or an increment of the active
cranes per ship) will have a more significant result on reducing the delay.
For the coal terminal 1, the large average waiting time (14.4% of the service time) is explained because
of the sum of the berth unavailability and congestion on the access channel. Congestion reaches a value
126
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
of 43% of the total waiting time, then, an optimisation of the access channel will allow a significant
reduction of the waiting time for this terminal.
In order to improve the performance of the access channel of the Taman Seaport, two types of measures are
available (traffic measures and enlargement of the nautical infrastructure measures). Because of its reduced
capital and maintenance costs, the traffic measures are recommended in a first stage. If the effect of these
measures is not sufficient, the possibilities of enlarging the access channel (increase of its width or
implementation of a passing section) should be evaluated, in terms of the reduction of the delay and the related
costs.
127
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
128
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Table 8.1. Summary with allowed and non-allowed encounters for a mixed operation of the access channel
CKH-750, go500
CH-25, 23, 20
CH-5.4, 5, 4.5
CH-120, 100
CKH-6500
CH-200
CH-150
CH-8, 7
CH-3.5
CH-4.2
CH-70
CH-30
CH-12
CH-15
CH-3
CH-2
Vessel
CH-200 N N N N N N N N N N N N N N N N N
CH-150 N N N N N N N N N N N N Y Y Y Y Y
CH-120, 100 N N N N N N N N N N Y Y Y Y Y Y Y
CKH-6500 N N N N N N N N Y Y Y Y Y Y Y Y Y
CH-70 N N N N N Y Y Y Y Y Y Y Y Y Y Y Y
CH-80, 60, 52, 48;
N N N N Y Y Y Y Y Y Y Y Y Y Y Y Y
CKH-5000, 2500
CH-30 N N N N Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-12 N N N N Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-25, 23, 20 N N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-15 N N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CKH-750, go500 N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-8, 7 N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-5.4, 5, 4.5 N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-3.5 N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-4.2 N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-3 N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-2 N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
Table 8.2. Average and maximum waiting time for incoming vessels per terminal – mixed operability of the
channel
Figure 8.1 presents the distribution of the average waiting time, which is classified depending on the reason of
the delay (port unavailability, berth unavailability, congestion or weather conditions).
For most of the terminals, a reduction of the traffic delay is observed, with average values smaller than 10
minutes. An exception of this is the Iron Terminal, which still has a traffic delay of about 22 minutes. This is
because of all its ships are large carriers, with beams larger than 40 m. Hence, for all of these ships the
encounter allowance with the rest of the ships is not allowed.
129
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Table 8.3. Average service time and normalized average waiting time – mixed operability of the channel
Figure 8.1. Average waiting time classification per terminal – mixed operability of the channel (units in minutes)
When the results of the Container Terminal 1 are observed, it is possible to see that the traffic delay has reduced
from 23 to 5 minutes, while the berth and pilot delays remain practically constant (117 to 110 minutes for the
berth delay; 2 to 3 minutes for the pilot delay). Then, the application of this traffic measure returns a reduction
of the congestion close to an 80%.
When the results for the Coal Terminal 1 are observed, it is possible to see that the traffic delay has reduced
from 21 to 3 minutes, while the berth and pilot delays remain practically constant (24 to 22 minutes for the
berth delay; constant value of 3 minutes for the pilot delay). Then, the application of this traffic measure returns
a reduction of the congestion close to an 85%.
The reductions of the average waiting time were from 9.5% to 7.1% (Container Terminal 1) and from 14.4% to
8.4% (Coal Terminal 1). Then, for the Coal Terminal 1 the application of this measure returns a reduction of this
indicator to a value that is below the typically accepted values. On the other hand, the reduced waiting time of
130
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
the Container Terminal 1 is still in the range where it could or could not be tolerated. Then, an assessment of the
operation with the application of the multiple entrances traffic rule is recommended.
Entrance 2
Entrance 3
Entrance 1
Figure 8.2. Longitudinal section of the access channel and proposed entrance points
The position of the multiple entrances and its maximum admitted draught are indicated in Table 8.4.
Table 8.4. Multiple entrances position and maximum admitted draught
Table 8.5 presents a summary with the entrance point for each vessel type, which depends on its draught.
Table 8.5. Entrance per vessel type
131
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
According to the multiple entrance defined for each vessel type and the total number of it (details in Appendix
D), the projected utilisation of each entrance will be a 22% of the ships using the entrance 1; a 32% of the ships
using the entrance 2; and a 46% of the ships using the entrance 3. The large percentage of ships entering to the
port via the entrance 3 (closest to the port) is explained because of the large percentage of vessels with small
draughts that will enter into the Taman Seaport. Considering that there is an important part of the fleet that will
enter via the intermediate entrances, it seems to be an interesting opportunity to implement this rule.
The average and maximum waiting time (applying multiple entrances to the access channel) for the vessels
entering into the Taman Seaport is presented in Table 8.6.
Table 8.6. Average and maximum waiting time for incoming vessels per terminal – multiple entrances to the
channel
The average service time and the normalized average time divided by the average service time for each terminal
are presented in Table 8.7.
Table 8.7. Average service time and normalized average waiting time – multiple entrances to the channel
Figure 8.3 presents the distribution of the average waiting time, which is classified depending on the reason of
the delay (port unavailability, berth unavailability, congestion or weather conditions).
For most of the terminals, a reduction of the traffic delay is observed, with average values smaller than 8
minutes. An exception of this is the Iron Terminal, which still has a traffic delay of about 19 minutes. This is
132
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
because of all its ships are large carriers, with beams larger than 12 m. For all of these ships the entrance point
will be the outer edge of the access channel; hence, these vessels have to sail the entire channel.
Figure 8.3. Average waiting time classification per terminal – multiple entrances to the channel (units in minutes)
When the results of the Container Terminal 1 are observed, it is possible to see that the traffic delay has reduced
from 23 to 4 minutes, while the berth and pilot delays remain practically constant (117 to 109 minutes for the
berth delay; constant value of 2 minutes for the pilot delay). Then, the application of this traffic measure returns
a reduction of the congestion close to an 83%.
When the results for the Coal Terminal 1 are observed, it is possible to see that the traffic delay has reduced
from 21 to 4 minutes, while the berth and pilot delays remain practically constant (24 to 27 minutes for the
berth delay; constant value of 3 minutes for the pilot delay). Then, the application of this traffic measure returns
a reduction of the congestion close to an 83%.
The reductions of the average waiting time were from 9.5% to 7.3% (Container Terminal 1) and from 14.4% to
10.4% (Coal Terminal 1). Then, the implementation of multiple entrance points to the access channel returns
reductions of the average delay that are similar to the obtained with the implementation of a mixed operability.
Container Terminal 1 stills with an average waiting time that is within the range where it could or could not be
tolerated. On the other hand, for the Coal Terminal 1, the reduction of the waiting time is sufficient to obtain an
accepted value.
Because both measures (mixed operability and multiple entrances) were implemented separately, an
assessment of the effect of implementing both measures simultaneously is developed.
133
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The average and maximum waiting time (applying a ‘mixed’ operability of the access channel) for the vessels
entering into the Taman Seaport is presented in Table 8.8.
Table 8.8. Average and maximum waiting time for incoming vessels per terminal – mixed operability and
multiple entrances to the access channel
The average service time and the normalized average time divided by the average service time for each terminal
are presented in Table 8.9.
Figure 8.4 presents the distribution of the average waiting time, which is classified depending on the reason of
the delay (port unavailability, berth unavailability, congestion or weather conditions).
Table 8.9. Average service time and normalized average waiting time – mixed operability and multiple entrances
to the channel
For all the terminals, the average traffic delay has been reduced to a value lower than 10 minutes. The Iron
Terminal (which had the largest traffic delay due to the large dimensions of its related carriers) has reduced its
traffic delay to 8 minutes. Then, the simultaneous application of a mixed operation of the access channel and a
multiple entrances allowance have a positive impact on reducing the congestion to average values lower than 10
minutes.
When the results of the Container Terminal 1 are observed, it is possible to see that the traffic delay has reduced
from 23 to 4 minutes, which is a similar value than those observed with the separate application of a mixed
operation and multiple entrances to the channel. Then, there are no extra differences with the utilisation of both
134
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
measures. Nevertheless, there are no negative effects (increase of congestion) and the application of these
measures has no costs.
When the results for the Coal Terminal 1 are observed, it is possible to see that the traffic delay has reduced
from 21 to 1 minutes, while the berth and pilot delays remain practically constant (24 to 27 minutes for the
berth delay; constant value of 3 minutes for the pilot delay). Then, the application of this traffic measure returns
a reduction of the congestion close to a 95%. Then, the application of both measures has a positive impact on
reducing the congestion of the Coal Terminal 1, to a value that indicates that congestion practically does not
exists for this terminal.
Figure 8.4. Average waiting time classification per terminal – mixed operability and multiple entrances to the
channel (units in minutes)
135
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The implementation each traffic measure separately returns a reduction of the average traffic delay of 75%. If
both measures are implemented simultaneously, the reduction of the average traffic delay grows up to 88%.
Figures 8.5 and 8.6 presents the average and maximum traffic delay for the different situations, respectively.
Figure 8.5. Comparison of average traffic delay for different traffic measures
Figure 8.6. Comparison of maximum traffic delay for different traffic measures
From these figures, it is possible to observe the following behaviour of the traffic delay:
The application of any traffic measure (mixed operation or multiple entrances) has a positive effect on
reducing the traffic delay (average and maximum) on all terminals. A (partial) exception is the Iron
Terminal, where a reduction of the congestion is observed but it is lower than for the rest of the
terminals. The good effect of both measures separately is explained because of the diversity of vessel
type for the different terminals (excepting the Iron Terminal): there are large differences on dimensions
from the large carriers to the small river – sea vessels. The mixed operation success depends on a large
number of vessels with small beams, while the multiple entrances success depends on a large number
of vessels with small draughts: both characteristics are found on the Taman’s projected fleet.
136
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The simultaneous application of both traffic measures has a larger reduction of the congestion than the
application of each measure separately. Nevertheless, the effective reduction from a situation with any
of the traffic measures is small. This is because the congestion has been reduced to small values for
almost terminals (excepting the Iron Terminal): the available range of reducing the congestion becomes
very small.
The final values of average congestion are different for each terminal. Those terminals with a large
percentage of small vessels have a lower congestion than those terminals with a high percentage of
large vessels.
The same effect is observed for the values of maximum traffic delay (higher for fleets with large vessels,
smaller for fleets with small vessels). In general, the maximum waiting times are below three hours for
all terminals. For the dry bulk and general cargo, this situation might be not very complicated (because
of the higher allowance of waiting related to this type of cargo). Nevertheless, the Container Terminal 1
presents a maximum traffic delay of 140 minutes: a verification of the typically maximum waiting time
accepted by the Port Authority of Taman is recommended.
The application of extra measures (modification of the nautical infrastructure) appears to be difficult
because of the small values of traffic delay obtained with a mixed operation and multiple entrances.
Nevertheless, the effects of an increase of the width and an implementation of a passing section are
presented in the next section.
Because the terminal with a normalized waiting time closer to the typically accepted values is the Container
Terminal 1, the enlargement of the bottom width to a value of 190 m (increase of 20 m) increases the allowed
encounters up to 98%. For this modification, the mixed operation of the channel and the multiple entrances
measures are active. Table 8.12 presents a summary with the allowed and non-allowed encounters for the new
bottom width of the outer channel, which is dependent on the vessel type.
The average and maximum waiting time (applying multiple entrances to the access channel) for the vessels
entering into the Taman Seaport is presented in Table 8.13.
137
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Table 8.12. Summary with allowed and non-allowed encounters for a mixed operation of the access channel
CKH-750, go500
CH-25, 23, 20
CH-5.4, 5, 4.5
CH-120, 100
CKH-6500
CH-200
CH-150
CH-8, 7
CH-3.5
CH-4.2
CH-70
CH-30
CH-12
CH-15
CH-3
CH-2
Vessel
CH-200 N N N N N N N N N Y Y Y Y Y Y Y Y
CH-150 N N N N N N Y Y Y Y Y Y Y Y Y Y Y
CH-120, 100 N N N N N N Y Y Y Y Y Y Y Y Y Y Y
CKH-6500 N N N N Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-70 N N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-80, 60, 52, 48;
N N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CKH-5000, 2500
CH-30 N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-12 N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-25, 23, 20 N Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-15 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CKH-750, go500 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-8, 7 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-5.4, 5, 4.5 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-3.5 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-4.2 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-3 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
CH-2 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
Table 8.13. Average and maximum waiting time for incoming vessels per terminal – width of the outer channel
increased to 190 m
The average service time and the normalized average time divided by the average service time for each terminal
are presented in Table 8.14.
Figure 8.7 presents the distribution of the average waiting time, which is classified depending on the reason of
the delay (port unavailability, berth unavailability, congestion or weather conditions).
138
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Table 8.14. Average service time and normalized average waiting time – width of the outer channel increased to
190 m
Figure 8.7. Average waiting time classification per terminal – width of the outer channel increased to 190 m
(units in minutes)
Clearly, for all terminals the berth delay has become the main cause of delay (with the exception of the Grain
Terminal and Sulphur Terminal: the first one does not present delay because of berth unavailability while the
second has a small average delay due to berth of 1.8 minutes).
The traffic delay is in the range 0 to 11 minutes. The highest congestion is observed for the Iron Terminal’s
vessels. The reason is that the bottom width enlargement to 190 m does not allow the interaction between all
the vessels of the Iron Terminal’s fleet, although many encounters with other smaller vessels from other
terminals are now permitted.
When the results of the Container Terminal 1 are observed, it is possible to see that the traffic delay has reduced
a 50% from the original bottom width (4 minutes to 2 minutes). The rest of the causes of delay (port, berth and
pilot) remain approximately constant.
139
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
When the results of the Coal Terminal 1 are observed, it is possible to see that the traffic delay remains constant,
with a value of 1 minute. The other causes of delay maintain approximately constant too.
The increase of the bottom width presents an improvement of the congestion conditions on the access channel.
Nevertheless, this progress is smaller than the observed for the implementation of the traffic measures.
Additionally, there are capital and maintenance costs related with the dredging of a larger access channel;
hence, a financial analysis is necessary in order to confirm or discard the effectiveness of this measure.
The average and maximum waiting time (applying multiple entrances to the access channel) for the vessels
entering into the Taman Seaport is presented in Table 8.16.
The average service time and the normalized average time divided by the average service time for each terminal
are presented in Table 8.17.
Figure 8.8 presents the distribution of the average waiting time, which is classified depending on the reason of
the delay (port unavailability, berth unavailability, congestion or weather conditions).
Table 8.16. Average and maximum waiting time for incoming vessels per terminal – with passing section
Original waiting time
Average waiting time Maximum waiting time
Terminal (no measures)
[min] [min]
[min]
Container terminal 1 141.9 ± 84.4 130.0 ± 62.9 6,656
Container terminal 2 120.3 ± 75.1 70.2 ± 63.7 1,881
Steel terminal 24.6 ± 5.8 9.0 ± 5.6 711
Grain terminal 17.4 ± 2.0 4.5 ± 1.4 460
Coal terminal 1 47.8 ± 12.1 39.0 ± 22.5 1,837
Iron terminal 159.4 ± 76.1 636 ± 27.1 1,846
Coal terminal 2 66.0 ± 27.3 45.4 ± 17.8 1,679
Mineral fertilizer terminal 1 23.1 ± 3.4 6.4 ± 1.6 456
Sulphur terminal 20.4 ± 4.1 6.3 ± 1.9 466
Mineral fertilizer terminal 2 43.8 ± 18.4 28.2 ± 9.0 1,399
140
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Table 8.17. Average service time and normalized average waiting time – with passing section
Figure 8.8. Average waiting time classification per terminal – with passing section (units in minutes)
Like for the bottom width enlargement, the presence of a passing section generates that for most of the
terminals the berth delay is the main cause of delay, excepting the Grain, Mineral Fertilizer 1 and Sulphur
terminals.
The traffic delay is in the range 1 to 4 minutes. The highest congestion is observed for the Iron Terminal’s
vessels, but the average delay is closer to the delay observed on the rest of the terminals. The passing section
acts as a two – way channel on a limited part of the access channel, allowing the encounters of all vessels
(including large carriers that composite the fleet of the Iron Terminal).
When the results of the Container Terminal 1 are observed, it is possible to see that the traffic delay has reduced
a 75% from the original bottom width (4 minutes to 1 minutes). The rest of the causes of delay (port, berth and
pilot) remain approximately constant.
When the results of the Coal Terminal 1 are observed, it is possible to see that the traffic delay remains constant,
with a value of 1 minute. The other causes of delay maintain approximately constant too.
141
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The implementation of a passing section presents an improvement of the congestion conditions on the access
channel. Nevertheless, this progress is smaller than the observed for the implementation of the traffic measures.
Additionally, there are capital and maintenance costs related with the dredging of a larger access channel;
hence, a financial analysis is necessary in order to confirm or discard the effectiveness of this measure.
The enlargement of the bottom width of the outer channel generates a reduction of the congestion of 19%. The
implementation of a passing section returns a reduction of the congestion of 50%.
Figures 8.9 and 8.10 presents the average and maximum traffic delay for the different situations, respectively.
Figure 8.9. Comparison of average traffic delay for different modifications of the access channel
142
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure 8.10. Comparison of maximum traffic delay for different modifications of the access channel
From these figures, it is possible to observe the following behaviour of the traffic delay:
The modification of the access channel via an enlargement of the bottom width has a positive effect on
reducing the average traffic delay on almost all terminals, excepting the Iron Terminal. The same
behaviour is observed with the maximum traffic delay. For this terminal, the enlargement of the bottom
width has no effect on its related vessels, because the dimensions (beams) of it do not allow encounters
with the new configuration of the access channel.
The implementation of a passing section has a positive effect on reducing the average and maximum
traffic delay on all terminals. The positive effect on all terminals is explained because the passing
section allows encounters for all vessels, including the large carriers related to the Iron Terminal.
While the enlargement of the bottom width of the channel has a low reduction of the congestion (19%),
the passing section increases the reduction up to 50%. Nevertheless, the absolute values of reducing
the congestion are small (in the order of minutes). The highest observed reduction (passing section)
generates a reduction of the average traffic delay of only 1 minute.
Considering this, the modification of the access channel seems to have a small financial benefit because of the
reduction of the waiting time. Nevertheless, in order to confirm or discard the financial feasibility of any of these
measures, an estimation of the capital and maintenance costs is required. This analysis is provided in the next
section.
143
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
The average depth to be dredged in the access channel is approximately 4.1 m. Then, the extra volume to be
3
dredged is approximately 1,050,000 m .
Passing section
The implementation of the passing section (bottom width enlargement to 240 m) is projected to be executed on
3 kilometres on a middle position of the outer channel. Similarly to the width enlargement, the extra volume to
be dredged depends on the original depth of the sea and the projected depth of the access channel. On the
passing section (kilometres 4.9 to 7.9), the original depth varies from -18.2 m to -15.7 m.
The average depth to be dredged in the access channel is approximately 4 m. Then, the extra volume related to
3
the passing section is 840,000 m . Nevertheless, the passing section is an increase of the bottom width: two
transitions are required to connect the original bottom width of the access channel. Puertos del Estado (2000)
recommends an horizontal slope of 1:20 for this type of transitions.
The total length of each transition is 700 metres (increases of 35 metres on each side of the channel with an
angle of 1:20). The average depth to be dredged on the outer and inner transitions is approximately 3.3 m and
3
5.8 m, respectively. The volumes of the outer and inner transitions are approximately 160,000 m and 280,000
3
m , respectively.
3
Then, the total volume related to a passing section is approximately 1,280,000 m .
Cost
Item [€]
Bottom width enlargement Passing section
Capital 4,725,000 5,760,000
Maintenance (5, 10 and 15 years) 576,000 693,000
Total present costs 5,442,613 6,623,379
144
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
8.2.4.3 Benefit
The financial benefit of implementing any of these measures is a reduction of the congestion (average waiting
time due to traffic conditions). Nevertheless, the conversion of this reduction of the average waiting time into a
financial benefit is always complicated and depends on parameters that are characteristics of each port.
Additionally, this type of information is typically confidential.
Because the specific information of the benefit due to reducing the waiting time is not available, an estimation
of the minimum benefit that should be obtained per ship per hour is defined.
The reduction of the average traffic delay for the width enlargement and the passing section are 0.4 and 1.1
minutes per ship, respectively. The expected number of ships per year varies from an initial throughput (year
2017; 2,689 ships/year) to a final throughput (year 2025; 3,503 ships/year). It is assumed that the variation is
linear, while after the year 2025 the number of ships will remain constant. This information leads to a definition
of a total number of waited hours per year due to congestion.
Finally, for the financial analysis it is considered a discount rate of 10%, while the lifetime of the project is
assumed to be 20 years.
Table 8.20 presents the estimation of the total number of waited hours due to congestion per year, per
alternative.
Table 8.20. Yearly gained congestion hours per alternative
Considering the total hours earned, the minimum benefit to achieve a financial feasibility for the bottom width
enlargement and the passing section alternatives are 30,112 €/hr and 13,325 €/hr, respectively.
8.3 Conclusions
The implementation of the two available traffic measures (mixed operation and multiple entrances to the
channel) has an important effect on reducing the congestion. The reduction of the congestion is about an 88%
(16.2 minutes) of the original delay due to traffic conditions, and it is explained because of the characteristics of
the Taman’s fleet. The large percentage of small vessels that can encounter within the current dimensions of the
access channel and can enter into it in a intermediate position is large.
Additionally, the implementation of these traffic measures has no impact on the finances of the Taman project.
The mixed operation requires a good system of traffic control (VTS), as well as good information on the
dimensions of the vessels entering and leaving the port. The multiple entrances require the implementation of
buoys to indicate the position of an intermediate entrance, and a proper definition of the bathymetry conditions
145
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
outside of the channel (survey). All of these actions have costs that are considerably small, and it can be included
on the final cost of the Taman project.
The modification of the access channel (width enlargement or implementation of a passing section) has a small
impact (0.4 to 1.1 minutes) on reducing the absolute waiting time due to traffic conditions. Then, the
implementation of any of these measures depends on the financial feasibility of it. For the width enlargement, it
is required a minimum benefit of about 30,000 €/hr while for the passing section the minimum benefit drops to
about 13,300 €/hr. Then, if the benefits on reducing the waiting times are beyond 13,300 €/hr, it is
recommended to implement a passing section.
As an opportunity to reduce the required benefit of reducing the congestion is to execute the dredging of the
passing section after the beginning of the port operation (for example, close to the year 2017). In this way, it
would be possible to delay the capital cost, and then reduce the present cost.
The benefit of reducing the waiting time is a complex problem. On this current approach, it has been considered
the effect of reducing the waiting time for all vessels. Nevertheless, during the final throughput stage it has been
detected that only some of the terminals have relatively large waiting times. Then, the absorption of the extra
costs due to an enlargement of the nautical infrastructure should be related only to these terminals (Container
Terminal 1 and Coal Terminal 1).
Additionally, when the delay due to berth unavailability is observed for different ships of a terminal (like the
Container Terminal 1), it is possible to observe that the largest ships has a larger delay due to berth compared
with smaller ships. Obviously, the required quay length is larger for the largest ships, but the algorithm of the
simulation model is currently not assigning a priority for larger vessels, hence, when a quay length is freed, if a
larger vessel does not fit on it, it will be assigned to a smaller ship.
The large waiting time for the Container Terminal is not because of congestion, but because of berth
unavailability. There are different measures that could be taken in order to reduce the delay due to berth
conditions (incorporate extra cranes to accelerate the service time, combine the operation of Container
Terminals 1 and 2, extend the quay length, etc.). Nevertheless, this problem is related to the design of the
master plan of Taman, but no with its access channel. Because this research is related with assessing the
operation of the access channel, this problem is out of the scope of study.
146
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
147
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
148
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
An important aspect in the development of any simulation model is its validation with real data. Unfortunately,
the validation of a model of these characteristics is difficult because the required information needed to perform
it is simply not available. During a future stage of research, the validation of the new simulation model is
recommended.
The assumptions and capabilities of the simulation model allow its application for many ports of small and
intermediate size. Nevertheless, during the testing and data analysis stages, some increasingly problems with
the number of ships were detected. Depending on the type of channel, the simulation model’s performance can
become very slow: limits of 6,000 ships/year (one-way channel) and 18,000 ships/year (two-way channel) are
recommended.
This practical limitation reduces the applicability of the simulation model to ports with smaller size. For example,
if the Port of Rotterdam is wanted to be simulated with this new model, in addition to the complexity of
modelling it (several anchorage areas, different inner access channels, and many vessels moving to and from the
Rhine River), the number of ships using this port is more than 30,000 ships/year. The expected behaviour of this
simulation model would be extremely slow, and its use is not recommended for it.
The low speed of the simulation model with a large number of ships can be explained because of the algorithm
used. During a future stage of research, it is recommended that the code of the model could be reviewed and
improved by a programming engineer. Hence, an optimised algorithm can be implemented, as well as evaluating
the incorporation of extra features to the model (such as the implementation of an easy set up screen for new
users or graphical animations of the simulated vessels).
Independently on the possible optimisation of the algorithm, some modifications to the verbal model can be
implemented as well. For example, the priority conditions of the ships are time independent: it does not matter
how much time a ship have been waiting at the anchorage site, its place in the queue in list depends on a FIFO
scheme and a priority rank. But, if a limit time is exceeded, then a solution could be the implementation of a
new priority rank, which could even apply a restriction for vessels entering the channel from the port until the
delayed ship has arrived to its terminal.
Another possibility is to extend the boundaries of the simulation model to incorporate the navigation of ships
through the different inner basins. Currently, this process is simplified by considering a tugged speed (variable)
and a distance from the turning circle to the terminal of destination (deterministic). But, if a similar approach to
the access channel is applied for the inner basins, a better insight of the traffic conditions on that part of the
nautical infrastructure can be obtained.
The features of the simulation model (written in a well-known software, definition of its algorithm, open source
code) allows for the implementation of these or other improvement possibilities.
149
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
assumptions were considered (like a maximum of three cranes operating on a single ship). If these assumptions
are confirmed (or modified), more reliable results on the berth delay could be obtained.
The application of the traffic rule modification generates an average congestion of 2 minutes. This congestion is
very low, and it leads to two main recommendations. First, it does not seem very attractive the idea to enlarge
the access channel to reduce the congestion. Secondly, the access channel can allow the passage of a larger
number of ships than the projected for the year 2025 without incrementing excessively the delay due to
congestion.
The related minimum benefits of 30,000 € per reduced hour of congestion (width enlarged) or 13,300 € per
reduced hour of congestion (passing section) appears to be high. Additionally, the volumes to be dredged
(capital and maintenance) are large, while the absolute reduction of congestion is about 1 minute. Only if the
number of ships would increase over the maximum expected, a new evaluation of the congestion would be
required and, probably, the required minimum benefits could be reduced.
If the access channel can allow the passage for extra ships without incrementing the congestion excessively, it
could lead to an optimisation of it. The design of the access channel depends on a design vessel. The largest
vessel has a draught of 18 m, and are large bulk carriers related to the Iron Terminal. A possibility that can be
evaluated (using this simulation model) is to replace the largest vessels into intermediate vessels, with smaller
capacity. If the total turnover is maintained, then a larger number of intermediate vessels will be required. It will
increase congestion but the reduction of the dimensions of the largest ship can lead to an access channel with
smaller depth, which is finally traduced into a reduction of the capital and maintenance costs.
150
‘Development of a Simulation Model to Assess Approach Channels – The Taman Seaport Case’
10 REFERENCES
[1] AGERSCHOU, H., DAND, I., ERNST, T., GHOOS, H., JENSEN, O., KORSGAARD, J., LAND, J., MC KAY, T.,
OUMERACI, H., PETERSEN, J., RUNGE-SCHMIDT, L., SVENDSEN, H. (2004), “Planning and design of ports
and marine terminals”. Thomas Telford Limited.
[2] BOSBOOM, J., STIVE, J. (2013), “Coastal Dynamics I”. Delft University of Technology, Faculty of Civil
Engineering and Geosciences, Section of Hydraulic Engineering. Version 0.4. Published by VSSD.
[3] BUDIYONO “A computer simulation modelling study for port planning of main public terminal, port of
Palembag, South Sumatera, Indonesia”. International Institute for Hydraulic and Environmental
Engineering, Delft, the Netherlands.
[4] DEMIRCI, E. (2003), “Simulation modelling and analysis of a port investment”. Simulation, vol. 79, issue
2, February.
[5] DERMIBILEK, Z., SARGENT, F. (1999), “Deep-draft coastal navigation entrance channel practice”. US
Army Corps of Engineers, Coastal Engineering Technical Note I-63, March.
[6] DEVILLE, S. (2011), “Port of Rotterdam anchorages study, an occupancy evaluation using simulation”.
Master of Science thesis report, Delft University of Technology, Faculty of Civil Engineering and
Geosciences, Ports and Waterways Chair. October.
[7] FORSYTHE, W., RYKIEL JR. E., STAHL, R., WU, H., SCHOOFIELD, R. (1995), “A model comparison for
daylength as a function of latitude and day of year”. Ecological Modelling 80, 87 – 95.
[8] GOOGLE EARTH, satellite images.
[9] GRAY, W., WATERS, J., BLUME, A., LANDSBURG, A. (2002), “Channel design and vessel manoeuvrability
– next steps”. Society of Naval Architects and Marine Engineers, New York Metropolitan Section.
January.
[10] GROENVELD, R. (1983), “Harboursim, a generally applicable harbour simulation model”. Delft University
of Technology, Department of Civil Engineering, Hydraulic Engineering Group.
[11] GROENVELD, R. (2001), “Service systems in ports and inland waterways, lecture notes CT 4330/5306”.
Delft University of Technology, Faculty of Civil Engineering and Geosciences, Section of Hydraulic
Engineering. Published by VSSD.
[12] HOFSETH, K., ROGERS, C., HEISEY, S., MALES, R. (2007), “Planning level simulation modelling of channels
improvements”. Presentation of the US Army Corps of Engineers and the Institute for Water Resources
th
– IWR, 25 September.
[13] IMO (2002), “Explanatory notes to the standards for ship manoeuvrability”. MSC/Circ. 1053. 16
December.
[14] LIGTERINGEN, H. (2009), “Ports and terminals, lecture notes CTwa 4330-5306”. Delft University of
Technology, Faculty of Civil Engineering and Geosciences, Section of Hydraulic Engineering. Published
by VSSD.
[15] LINGWOOD, J., MRINA (2007), “Significant ships of 2006”. Published by The Royal Institution of Naval
Architects.
[16] MALYSHEVA, N. (2012), ‘Transportation Ministry Presents Taman Port in Paris’. Information Agency RZD
th
Partner, 4 April, 2012. Available on link http://www.rzd-partner.com/press/2012 /04/04/375518.html
[17] MARIA, A. (1997), “Introduction to modelling and simulation”. Proceedings of the 1997 Winter
Simulation Conference,
[18] MAYER, R., WATERS, J., KRIEBEL, D. (1999), “Design & maintenance of deep-draft navigation channels,
an overview of current practice with an annotated bibliography”. Department of Naval Architecture,
Ocean and Maritime Engineering. Division of Engineering and Weapons. U.S. Naval Academy, Annapolis.
151
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
[19] PIANC (1997), “Approach channels, a guide for design”. Final report of the joint Working Group PIANC
and IAPH, in cooperation with IMPA and IALA. Supplement to Bulletin no 95. Brussels.
[20] PHYSICALGEOGRAPHY.NET, “Earth – Sun relationships and isolation”. Available on link http://
www.physicalgeography.net/fundamentals/6i.html
[21] PUERTOS DEL ESTADO (2000), “ROM 3.1 – 99, proyecto de la configuración marítima de los puertos;
canales de acceso y áreas de flotación”. Obras Marítimas Tecnología.
[22] QUIST, P. (2012), “Liquid bulk terminals”. Delft University of Technology, Department of Civil
th
Engineering. Lecture notes Ports and Waterways II. May 24 .
[23] SCHNEEKLUTH, H., BERTRAM, V. (1998), “Ship design for efficiency & economy”. Butterworth-
Heinemann.
[24] SIREGAR, P. (1995a), “A study on the conceptual design rules for approach channels”. Graduate thesis,
Delft University of Technology, Department of Civil Engineering, Hydraulic Engineering Group. October.
[25] SIREGAR, P. (1995b), “Towards fast time simulation-based probabilistic design of channel width”.
Graduate thesis, Delft University of Technology, Department of Civil Engineering, Hydraulic Engineering
Group. December.
[26] SOLARI, S., MOÑINO, A., BAQUERIZO, A., LOSADA, M. (2010), “Simulation model for harbor verification
and management”. Coastal Engineering 2010.
[27] THORENSEN, C. (2010), “Port designer’s handbook, second edition”. Thomas Telford Limited.
th
[28] UK TRADE & INVESTMENT (2012), ‘Ports sector in Russia’. 24 April, 2012. Available on link
http://www.ukti.gov.uk/es_es/export/countries/europe/easterneurope/russia/sectorbriefing/294920.h
tml?null
[29] US ARMY CORP OF ENGINEERS (2006), “Hydraulic Design of Deep-Draft Navigation Projects”. Engineer
Manual, EM 1110-2-1613.
[30] VAN DER MEER, J. (2013), “Port master plan for the port of Beira, Mozambique”. Master of Science
thesis report, Delft University of Technology, Department of Civil Engineering, Chair of Ports and
Waterways. January.
[31] VAN DER SCHRIEK, G. (2012), “Dredging Technology, Guest lecture notes CIE5300 issue 2012”. GLM Van
der Schriek BV Dredging Education & Research.
[32] VAN DRIEL, J. (1993), “Simulation model for ports operations; application for Pontianak”. Master of
Science thesis report. Delft University of Technology, Faculty of Civil Engineering, Chair of Ports and
Waterways. September.
[33] WITTEVEEN+BOS (2012), ‘Creation of Taman Seaport Dry Cargo Area’. Engineering services for
RosTranModernizatsiya, February 2012.
[34] WORLDATLAS.COM, ‘Russia’. Available on link http://www.worldatlas.com/webimage/countrys
/asia/rularge.htm.
[35] XINGYAN, Y., HUA, J., WEI, L., HONG, L. (2011), “Study on the navigation capacity of the approach
channel of Tianjin Port”. Port Technology International.
152
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
i
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
ii
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
iii
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
General input
% GENERAL INPUT FILE
% Fleet characteristics
service_rate=1.5*[0,0,720;0,0,720;960,960,0;1500,3000,0;0,3000,0;0,5000,0;0,3000,0;1500,3000,0;
1500,3000,0;0,3000,0];
service_dist=[3;3;2;2;2;2;2;2;2;2];
% Simulation characteristics
iv
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
% General characteristics
% Strike characteristics
% Weather characteristics
save
('input.mat','number_terminals','Ti','Tf','dt','tidal_window','port_operation','latitude','str_
prob','str_dur','ice_prob','ice_dur','fog_prob','fog_dur','wave_prob','wind_prob','cross_curr',
'long_curr','vessel_speed','vessel_speed2','nav_aids','bot_surf','basic_manouv','traff_dens','b
ank_clear','service_rate','service_dist')
clear all
Layout input
% LAYOUT INPUT FILE
v
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
% Intermediate anchorage
% Dimensioning of berth
save layout
clear all
Priority input
% PRIORITY INPUT FILE
% This file allows user to incorporate those terminals that will have a
% larger priority (traffic purposes).
priority=[1 2];
save priority
Route input
% ROUTE INPUT
clear all
save route
clear all
vi
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
clear all
clear all
% Type of channel
% Safety distance
safety_distance=5; % minimum distance with respect to previous ship, expressed in ship length
% Maximum speed
% Entrance speed
% Speed reduction
save traffic_rules
Vessel input
% VESSEL INPUT FILE
vii
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
clear all
% Deadweight definition
i = 1;
% Length definition
viii
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
i = 2;
% Beam definition
i = 3;
% Draught definition
i = 4;
ix
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
draught(8,:)=[v(18,i) v(16,i) v(14,i) v(13,i) v(11,i) v(9,i) v(18,i) v(16,i) v(14,i) v(13,i)
v(11,i) v(9,i) 0 0 0 0 0]; % Mineral terminal 1
draught(9,:)=[v(18,i) v(16,i) v(14,i) v(13,i) v(12,i) v(11,i) v(9,i) v(7,i) v(4,i) v(3,i)
v(6,i) 0 0 0 0 0 0]; % Sulphur terminal
draught(10,:)=[v(19,i) v(16,i) v(14,i) v(9,i) 0 0 0 0 0 0 0 0 0 0 0 0 0];
% Mineral terminal 2
% Operation definition
i = 5;
save vessel
clear all
x
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
xi
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Figure C.1. Longitudinal and cross sections of the Taman’s access channel
xii
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
xiii
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
xiv
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Number of ship
Ships displace-
entries (in one
Draft loaded
depth (m)
direction)
Moulded
Нгр (m)
(тыс,т)
Vessels
Beam
B (m)
L (m)
ment
DWT
LOA
Per Year Berthing
Quay Type of ships representatives
of the vessel
full
Stage 1 development speed angle
(2017) 2025 (m/s) (º)
CKH-6500 40 73 85.7 304.1 40 24.2 14
Container Quay 1 Containership CKH-750 100 163 8,0-9,5 121.4 18.2 8.3 6.7
CKH-go500 108 174 5.2 100.6 18.2 8.25 6.5
xv
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Number of ship
Ships displace-
entries (in one
Draft loaded
depth (m)
direction)
Moulded
Нгр (m)
(тыс,т)
Vessels
Beam
B (m)
L (m)
ment
DWT
LOA
Per Year Berthing
Quay Type of ships representatives
of the vessel
full
Stage 1 development speed angle
(2017) 2025 (m/s) (º)
Not in
СН-150 2017 14 161 280 45 - 17.5
СН-115 14 Not in full dev 119.1 255 43 - 14
СН-100 16 16 105.7 234.9 43 - 15.27
СН-80 18 18 82 250 32.2 - 14.5
СН-70 20 20 77.4 230 36 - 12.7
Not in
СН-200 2017 36 200 290 48 - 18
Iron ore quay bulkcarrier Not in
СН-150 2017 42 161 280 45 - 17.5
СН-120 167 52 120 255 43 - 15.27
xvi
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Number of ship
Ships displace-
entries (in one
Draft loaded
depth (m)
direction)
Moulded
Нгр (m)
(тыс,т)
Vessels
Beam
B (m)
L (m)
ment
DWT
LOA
Per Year Berthing
Quay Type of ships representatives
of the vessel
full
Stage 1 development speed angle
(2017) 2025 (m/s) (º)
Not in
СН – 150 2017 5 161 280 45 - 17.5
СН – 115 3 Not in full dev 115 255 43 - 15
Coal 2 quay bulkcarrier СН – 100 12 25 105.7 234.9 43 - 15.27
СН – 70 29 57 77.5 230 36 - 12.7
СН – 60 49 99 58.6 196 32.26 - 13
СН – 48 55 109 47.8 197.1 32.2 - 11.5
xvii
‘Development of a Simulation Model for the Assessment of Approach Channels – The Taman Seaport Case’
Number of ship
Ships displace-
entries (in one
Draft loaded
depth (m)
direction)
Moulded
Нгр (m)
(тыс,т)
Vessels
Beam
B (m)
L (m)
ment
DWT
LOA
Per Year Berthing
Quay Type of ships representatives
of the vessel
full
Stage 1 development speed angle
(2017) 2025 (m/s) (º)
СН – 70 2 5 77.45 230 36 - 12.7
СН – 52 12 31 51.5 197.9 32.2 - 12
СН – 30 21 56 30 170 27 - 10
СН – 25 24 64 23.75 181.5 22.9 - 9.5
СН – 20 2 4 18.35 165 23 - 10.5
bulkcarrier sea
Sulphur quay СН – 15 2 5 16 153.5 22.4 - 8.2
СН – 8 4 11 7.52 122.3 17.8 - 7
СН – 5,4 6 16 5.4 110 16.5 - 6
СН – 4,2 7 21 4.2 100 13.5 - 5.6
СН – 3 10 31 3 82 12.3 - 5.7
River-sea vessel Unloading (CH-5) 189 189 5.9 140 16.65 - 4.5
xviii