99
1 INTRODUCTION
Low freight rates and high competition in the
shipping and maritime industry has driven a strong
demand towards optimizing the supply chain and
establish better utilization of increasingly larger
vessels entering the world fleet.
Both trends are currently increasing pressures on
many ports and terminal operators to accommodate
tighter transit schedules, reduce delays and
accommodate larger and deeper drafted vessels. For
more than a decade, full mission bridge 3D ship
simulators such as SIMFLEX4 by FORCE
TECHNOLOGY have been an essential tool in
identifying and solving capacity constraints through
the shipping channel, turning basin and during
berthing. Capacity and operability assessment of
moored vessels at berth are typically not carried out
using simulators but often using a more simplistic
static force balance approach. These tools used for
strategic port capacity assessment are almost never
used in day-to-day operational support systems.
Instead strategic studies are usually limited to
Next Generation of Physics-based System for Port
Planning and Efficient Operation
S.B. Mortensen, A. Harkin & H. Kofoed-Hansen
DHI Group
W. Mlas
DHI Polska
ABSTRACT: The continuing surge in commercial vessel sizes is putting increasing pressures on the world’s
port authorities to adopt effective expansion strategies to ensure that their asset is able to meet growing
capacity demands. Among the key challenges is to assure that correct strategic planning and operational
measures are adopted to guarantee safe and efficient traffic not only through its shipping channel, but also at
the port berthing facilities. DHI and FORCE TECHNOLOGY have collaborated to develop a novel physics-
based vessel traffic management system named NCOS ONLINE. It is capable of taking into account of any
relevant vessel constraints such as under-keel clearance (UKC), maneuverability and berth configuration that
may constrict the movement of vessels through the channel or operability at berth, facilitating scenario
planning and capacity assessment of proven unparalleled accuracy. The system incorporates the accuracy of
high-end Full Mission Bridge Simulators with regards to vessel response under power and at berth. The
underlying computational engines uses a powerful 3D panel method for vessel response calculations in
combination with highly detailed environmental data such as wind, waves and hydrodynamics (water level
and currents) simulated by use of MIKE Powered by DHI’s recognized and scientific based computational
models. The modular and integrated framework-based system has already been adopted by numerous port
authorities, terminal operators and pilots worldwide for strategic port planning, design and 24/7 operational
vessel traffic management. The paper focus on presenting the underlying equational framework and validation
of the underlying physical response engines and provide a brief introduction to how they are integrated and
operated through a series of user-friendly web dashboards.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 13
Number 1
March 2019
DOI: 10.12716/1001.13.01.09
100
supporting operations through provision of static
guidelines which does not utilize the full dynamic
optimization potential of the port.
In this paper we will introduce a next generation
type physics-based port traffic management system
called NCOS ONLINE. The cloud-based system was
originally presented in Mortensen et al 2016
[5]
and
Mortensen et al 2018
[6]
and is capable of optimizing
port capacity and reduce delays from channel to berth
with the accuracy of 3D full bridge simulators and
using the same numerical framework for both
strategic planning and subsequent operational
decision support. This paper focus on presenting the
underlying equational framework and validation of
the underlying physical response engines and provide
a brief introduction to how they are integrated and
operated through a series of user-friendly web
dashboards.
2 NUMERICAL MODELLING FRAMEWORK
2.1 Vessel Frequency Response
For simulating wave vessel frequency response under
power, NCOS ONLINE uses the 2nd order 3D panel
method engine, S-OMEGA, which is used in the Full
Bridge Simulator SIMFLEX4 by FORCE
TECHNOLOY incorporating implicitly the effect of
vessel forward speed and varying water depths. For
simulating vessel wave frequency response at berth
the numerical solution can be made more
computationally efficient by neglecting forward
speed. For this purpose, NCOS utilizes DHI’s own 3D
frequency radiation-diffraction solver FRC, which is
DHI’s boundary element code used to solve the linear
boundary problem for the free surface flow around a
body to calculate vessel first order wave forces and
second order wave drift forces in the frequency-
domain. Because of its computational efficiency FRC
is capable of fast and accurately modelling a wide
range of typical berth type multibody scenarios such
as caused by reflective quays or tandem moored
vessels.
2.2 Vessel Motion Analysis
NCOS is directly integrated with the phase-averaged
3rd generation wave model MIKE 21 Spectral Wave
model (MIKE 21 SW) and phase resolving models like
MIKE 21 Boussinesq Wave model (MIKE 21 BW) or
the 3D model MIKE 3 Wave FM enabling the accurate
prediction of spatially and temporally varying wave
response through a channel and at berth.
For powered vessel response the frequency
domain wave response is evaluated in the form of 1st
order motion response amplitude operators (RAO’s)
and MIKE 21 SW model provides the wave conditions
along with 2nd order vertical motions (T
(2)
).
The spectral form of the 1st order motions of a
user-specified number of motion points (d) on the
vessel in a specific sea state is calculated from the
motion RAO for the specified point and the sea
spectrum as shown in (1).
2
.
dd
S RAO S
(1)
where
d
R
AO is the RAO calculated by S-OMEGA
translated to each motion point on the vessel
d
, S
is the wave spectra at each timestep and vessel
position
and
d
S is the resulting motion response
spectra at each motion point
d . Equational
framework for calculating dynamic heel and squat is
presented in Harkin A. et al 2018
[1]
.
The 2
nd
order set down,
2
T
, is calculated from
(2) to (4) where
2
F is the second order
force/moments extracted from S-OMEGA,
is the
wave direction and
C is the restoring force.
2
2
, , ,
F
SFS
(2)
2
2
,
F
Sd
x
C
(3)
222 2
,**
d heave roll y pitch x
Txxdxd
(4)
For moored vessel response NCOS uses its non-
linear dynamic vessel mooring analysis model MIKE
21 MA to calculate vessel motions in the time-domain.
The incident wave potential
I
and first order
dynamic pressure
I
P is evaluated across the 3D
vessel hull from a 2D/3D wave input from MIKE 21
BW/MIKE 3 Wave FM or a surface evaluation time
series synthesized from MIKE 21 SW. The first order
radiation velocity potential
j
is computed by FRC.
The wave exciting force
jD
Ft is then calculated
from the Haskind relations:
,.
,,
b
b
jD I j
S
jIn
S
Ft Pxtnxdx
x
txdxd
∬
∬
(5)
Subsequently the equation of motion for the
moored vessel is evaluated in the time domain using
Equation 2.
¨
0
1,2,...,
t
N
k
jk jk jk k jk k
k
jD jnl
MxtKtxdCxt
Ft Ft j N
(6)
Second order wave drift forces are calculated
using the far field approximation method presented
in Newman J. N. (1974)
[2]
. Wind and current forcing
are accounted for as either spatially uniform (0D) or
spatially varying (2D) data files using vessel specific
drag coefficients. Mooring line and fender forces are
calculated based on actual load-displacement curves.
Viscous damping is included as a combination of
constant friction damping plus linear, quadratic and
cubic damping.
101
2.3 Numerical Model Validation Examples
2.3.1 Vessel motion and underkeel clearance in
Navigational Channel
Serving as the bases for this validation are
measurements taken during vessel transits through
the Port of Brisbane, Australia. Differential Global
Positioning Systems (DGPS) were located at the bow
(Figure 2-1) and on both the port and starboard bridge
wings (Figure 2-2) of the vessels to measure trajectory
and the vertical position at each location. An XSens
Inertial Measurement Unit (IMU) was also used to
measure the vessel roll and pitch as contingency in
the event that any of the DGPS units did not work
correctly. From these measurements roll, pitch, heave
and total vertical excursion of the vessels throughout
the transits were calculated.
Figure 2-1. DGPS Setup on Bow
Figure 2-2. DGPS Setup on Starboard Bridge Wing
Table 1 describes the vessel transits included in the
validation.
Table 1. Description of Vessels included in the Measurement
Campaign
_______________________________________________
Vessel Vessel Class Draft (m) LOA (m) LPP (m)
Name
_______________________________________________
B2 Bulk Carrier 13.50 253.50 249.20
B3 Bulk Carrier 13.53 253.50 249.20
C1 Container 12.10 294.10 282.20
C2 Container 12.68 255.00 244.00
_______________________________________________
For this validation a MIKE 21 SW model was
produced for the Port of Brisbane. The MIKE 21 SW
model has been setup to generate a 2D unstructured
data file containing sea and swell integral wave
parameters to be used with wave response
implementation 1. For wave response implementation
No 0 the MIKE21 SW model was setup to produce 40
fully direction spectra timeseries along the Port of
Brisbane shipping channel. The locations of these 40
directional spectra are displayed in Figure 2-3.
Figure 2-3. Port of Brisbane Directional Spectra Locations
Figure 2-4 and Figure 2-5 show that the calculated
wave-induced significant roll of the bulk carriers was
well represented by both wave response
implementations with implementation No 2 being the
most accurate. This included the relatively calm inner
components of the transit, moving to the more
exposed regions in the latter half of the outbound
transits.
Figure 2-4. B2 Filtered Roll Validation (Bulk Carrier,
Outbound)
Figure 2-5. B3 Heave Validation
102
Figure 2-6. C1 Non-filtered Roll Validation incl. effects of
wind and turning (Container Vessel, Inbound)
Figure 2-7. C1 Heave Validation
Figure 2-8. B3 UKC Validation
Figure 2-9. C1 UKC Validation
Direct comparison between measured and
modelled UKC was excellent and demonstrated a
high level of accuracy in capturing various drivers
conservatively without being overly conservative.
Additional validation is presented in Harkin A. et al
2018
[1]
.
2.3.2 Moored vessel motion at sheltered LNG Berth
Offshore structures, such as gravity-based
structures (GBS), are used in the petroleum industry
as drilling, extraction and storage units for crude oil
or natural gas. The interaction between moored
vessels and such structures involves complex wave
diffraction/sheltering and radiation effects. In order to
accurately perform mooring analyses of these
scenarios, a coupling of mooring analysis and
complex wave modelling is required.
An LNG tanker was moored at a water depth of
15m using 14 synthetic rope lines with 11m tails and 4
SCN2000 E1.5 fenders. The vessel characteristics are
given in Table 2 and mooring system is illustrated in
Figure 2-10.
Table 2. Vessel characteristics.
_______________________________________________
Characteristic Value
_______________________________________________
LOA [m] 318.2
Beam [m] 50.6
Draft [m] 12
Displacement [m
3
] 133824
_______________________________________________
Figure 2-10. Mooring system definitions.
The physical model is displayed in Figure 2-11.
Figure 2-11. Sheltered berth physical model setup.
The wave conditions considered for this validation
are a significant wave height of 3m and a peak period
of 12s. The wave direction is perpendicular to the GBS
to maximise its sheltering effects. A MIKE 21 BW
model was again used to replicate these wave
conditions numerically to serve as the wave forcing
input for MIKE 21 MA.
When moored vessels are in close proximity to
fixed structures, such as a GBS, viscous damping
effects become complex and significant (especially in
shallow water) MIKE 21 MA can account for viscous
damping and its effect on moored vessel motions but
103
requires damping coefficients to be established a
priory. Based on physical modelling results a linear
roll induced damping coefficient of 4.3E+06kNms/rad
was established, which was applied to the numerical
model setup. Viscous damping in other modes were
considered too small to have practical significance.
Test results obtained with the numerical model
have been compared to the physical model results.
Comparison of motion spectra and peak-to-peak
motions are presented in Figure 2-12 and Figure 2-13
and comparisons of maximum line and fender forces
are presented in Figure 2-14 and Figure 2-15.
The result comparisons show that MIKE 21 MA
has very accurately reproduced the physical model
results. Again, a long period response of the moored
ship has been observed and is mostly due to the effect
of the natural resonance frequency of the mooring
system. Additional validation is presented in Harkin
A. et al 2017
[3]
.
Figure 2-12. Motion spectra validation.
Figure 2-13. Peak to peak motion validation.
Figure 2-14. Maximum line forces.
Figure 2-15. Maximum fender forces.
3 NCOS ONLINE INTEGRATION
NCOS ONLINE incorporates the powerful vessel
response engine S-OMEGA and MIKE 21 MA in
combination with the hydrodynamic and wave
modelling capabilities of MIKE computational engine
models to provide a powerful cloud-based physics-
based port traffic management system supporting
both strategic planning and 24/7 operational decision
support.
Figure 3-1. Cannel capacity windows for deep draft
Suezmax and Aframax tankers for port of Geelong
104
Through its sophisticated framework the system is
capable of implicitly taking into account a wide
variety of relevant vessel constraint such as UKC,
maneuverability and moored vessel motions that may
constrict the movement of vessels through the channel
or operability at berth.
As an example of strategic planning application is
its capability to provide a detailed assessment of
navigational channel capacity through detailed
simulation of future vessel traffic scenarios based on
typically 1 to 20 years of deterministically modelled
variations in water level, winds currents and waves
through the entire domain. Figure 3-1 show the tidal
dependent operability constraints of the navigational
channel into Port of Geelong (Australia) based on
approximately 2,700 simulations of inbound transits
with the larger Suezmax tanker with a draft of 12.0 m
and an Aframax tankers at 11.65 m as presented in
Mortensen et al 2017
[4]
. As observed operability for
these large vessels for this port are significantly
constrained to suitable tide and weather windows.
Figure 3-2. Operational Planning Interface
NCOS ONLINE provides the option to plan such
transits through its operational planning interface
illustrated in Figure 3-2. Using an intuitive
dashboard, the port traffic controller types in each
unique vessel calling the port and the system returns
its safe transit windows based on forecasted weather
and water levels for the next 7 days. The system
provides differentiated support to multiple user
groups each with their own unique access levels and
tailored dashboards such as for Shipping Operators,
Port Owners, Port Traffic Control, Terminal Operators
and Pilots.
Figure 3-3. As soon as a vessel transit is designated, the
Pilot has access to a range of detailed decision support
information.
As soon as a transit is locked by traffic control, the
scheduled transit becomes accessible to the vessel
Pilot, who gets notified of the roster update and is
then able to inspect the specifics of his transit such as
the UKC profile, his speed profile, vessel response
and weather conditions as illustrated in Figure 3-3.
Harbour Masters oversee the entire process through
their own dashboards and have exclusive access
rights to increase the conservatism of various safety
parameters in the system if unusual circumstances
dictate this type of action.
The berth planning itself is also operated through
similar dashboards that is used to plan the
appropriate mooring line configuration and berth
marker that fits the forecasted weather conditions and
scheduled berth utilization. The system provides a
graphical GIS overview of proposed mooring system
layouts and provides the user the option for planning
is mooring arrangement based either on forecasted
values or pre-defined environmental forces as
illustrated in Figure 3-4.
Figure 3-4. Effective GIS based mooring arrangement
planning through NCOS ONLINE.
Once a mooring plan has been locked in the
system The system provides a running forecast of
motions and mooring system forces while the vessel is
at berth and provide automatic alerts and
notifications if attention is required for mitigating
potentially hazardous conditions for the moored
vessel as illustrated in Figure 3-5.
Figure 3-5. Example of operational berth forecast
The system also provides a number of automated
operational mooring reports, which allows
communication to be easily shared with the vessel
105
Master, terminal operators and stevedores see Figure
3-6.
Figure 3-7 shows an example of calculated wave
fields (short and long waves) within a port used as
input for dynamic mooring analysis of a LNG carrier
at berth.
Figure 3-6. Example of Mooring System Forecast showing
max bollard forces, line forces and vessel surge. Should also
have the option to include max Fender force, sway, heave,
roll, pitch, yaw.
Figure 3-7. Example of vessel (LNG carrier) motion
calculation using MIKE 21 MA at berth caused by offshore
waves (short waves, the white foam indicates wave
breaking at the coast) and infragravity waves (long waves,
colored).
4 CONCLUSIONS
This paper presents the equational frame work,
numerical accuracy and practical application of
utilizing an online physics-based traffic management
system can significantly increase port capacity to
accommodate larger and deeper drafted vessels from
channel to berth. The system demonstrates a capacity
to incorporate the operational user group
requirements of multiple port operators and provides
a flexible platform for a more effective and user-
friendly management tool for optimizing constrained
port traffic flow both now and in the future.
REFERENCES
[1] Harkin A, Harkin J, Suhr J, Mortensen S.B, Tree M,
Hibberd, W (2018) Validation of a 3D underkeel
clearance model with full scale measurements, 34
th
PIANC World Congress, Panama City, Panama.
[2] Newman J. N. (1974). Second-order, Slowly-varying
Forces on Vessel in Irregular Waves. International
Symposium on Dynamics of Marine Vehicles and
Structures in waves, London.
[3] Harkin A, Mortensen S.B, Dixen M. (2017) Validation of
Moored Vessel Response Simulator with Physcial Model
Comparisons, Coast and Ports Conference, Cairns,
Australia.
[4] Mortensen S.B, Jensen B.T, Harkin A, Tree M, Nave R
(2017) An Improved Integrated Approach for
Optimizing Shipping Channel Capacity for Australian
Ports, Coast and Ports Conference, Cairns, Australia.
[5] Mortensen S.B, Jensen B.T, Nave R. (2016) A Nonlinear
Channel Optimization Simulation Framework for Port of
Brisbane Australia, PIANC COPEDEC Conference
Proceedings, Rio de Janeiro, Brazil.
[6] Mortensen S.B, Thomsen F, Harkin A,
Shanmugasundaram S.K, Simonsen C, Nave R (2018)
Web Based Operational System for Optimising Ship
Traffic in Depth Constrained Ports, 34
th
PIANC World
Congress, Panama City, Panama.
