425
1 INTRODUCTION
Ship drift refers to the involuntary displacement of a
vessel under the influence of currents, winds, or other
natural forces, without active control by the crew. In
other words, it is the passive movement of the ship
caused by external forces rather than planned
navigation or its own propulsion. When the vessel is in
transit, drift may occur due to propulsion or rudder
system failures or in areas with strong maritime
currents, among other factors [3].
When the ship is docked, drifting can occur due to
extreme environmental conditions or human errors
that result in the rupture of mooring lines. In this case,
what is known as unintentional unberthing occurs,
which usually constitutes an even more critical
situation, as the ship is near port structures such as
quays, piers, or other vessels, increasing the likelihood
of accidents and damage to the environment (leakage
of hazardous substances) and port assets (collisions
with other vessels or structures). Additionally, there
are economic impacts, including delays in cargo
delivery and costs associated with damage repairs. In
such cases, a rapid response from the crew and port
Simulation of the Unintentional Unberthing of Vessels
(Ship Drift) in a Physical Hydraulic Model
R. Esferra, C.C. Fernandes de Jesus, R. de Oliveira Bezerra & J.C. de Melo Bernardino
University of Sao Paulo, Sao Paulo, Brazil
ABSTRACT: Ship drift refers to the unintentional movement of a vessel caused by external forces—such as wind,
currents, and waves—acting on the hull without deliberate control by the crew. During navigation, drift may
result from failures in the propulsion system or rudder, which impair or prevent maneuverability, leading to
course deviations and increasing the risk of grounding or collisions. When the vessel is moored, drift may occur
due to extreme environmental conditions or human error that result in the breaking of mooring lines. In such
cases, the problem known as unintentional unberthing occurs, often representing an even more critical situation,
as the vessel is located near fixed structures such as quays, piers, or other vessels, thereby increasing the
probability of accidents and damage to the environment and port infrastructure. This paper presents a study of
an estuarine port area, carried out in a Froude-number based reduced-scale physical modeling, to assess the risks
associated with the unintentional unberthing of a VLOC-class vessel (400,000 DWT). The study involved the
analysis of the drifting trajectory of the vessel under various environmental conditions, positioning of the vessel
at the berths, and occupancy of adjacent berths. A digital camera tracking system was employed to monitor the
vessel's position at each moment in time, allowing for the assessment of collision risks with port structures or
other ships, as well as the potential for grounding in shallow areas. The results of the physical model simulations
identified the scenarios with the highest potential for damage, underscoring the importance of strict maintenance
of mooring systems and serving as a basis for the development of an emergency action plan to mitigate accident
risks in the port area.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 19
Number 2
June 2025
DOI: 10.12716/1001.19.02.11
426
authorities is essential to prevent harm and ensure the
safety of those involved [4] [6].
For these reasons, studying unintentional
unberthing is crucial. Understanding drift patterns
helps to prevent maritime accidents, optimize port and
navigation operations, and support the development of
effective maritime regulations and guidelines.
Froude-number-based reduced-scale model, also
called physical hydraulic model (PHM), is an effective
tool in this type of analysis. It allows a high-fidelity
representation of hydraulic conditions affecting
vessels, enabling direct assessment of the interaction
between the flow and the port structures. Furthermore,
this approach helps identify critical operational
scenarios, validate damage mitigation strategies, and
provide valuable insights for port operation planning
and safety [5].
The application of PHM’s is particularly
advantageous compared to computer simulations,
which often require greater simplifications of physical
phenomena involved and face limitations in accurately
representing complex flows around solid structures
(such as vessels, berths, etc.). In contrast, when
properly designed to avoid or mitigate major scale
effects, PHM provides a realistic representation of the
flow dynamics and their effect on vessels.
This paper presents a study of an estuarine port
area carried out on a 1:170 scale PHM to assess the risks
associated with the unintentional unberthing of a
VLOC (Very Large Ore Carrier) class vessel (400,000
DWT). The study analyzed the vessel’s drift trajectory
under different scenarios of environmental conditions,
VLOC positioning at berths, and presence of other
vessels at adjacent berths.
A high-resolution digital camera-based tracking
system was employed to accurately monitor the
vessel’s position. This tracking system also allows real-
time calculation of the ship’s drift velocity, in addition
to a detailed analysis of potential collisions with port
structures or other vessels and risks of grounding in
shallow areas.
The main objective of this paper is to present the
methodology for analyzing unintentional unberthing
and vessel drift using physical hydraulic modeling,
presenting the testing techniques and the results
obtained in a case study.
2 MATERIAL AND METHODS
PHM’s of estuarine and port regions allow the
simulation of various environmental scenarios,
including tides, currents, and waves, as well as
different ship types, loading conditions, and mooring
systems.
For this study, a non-distorted 1:170 geometric scale
PHM was used, covering an area of 1,700 m² at the
Laboratory of Hydraulics of the University of Sao
Paulo (USP), Brazil. The model represents a port area
located inside a bay, sheltered from wave action but
subject to strong tidal currents (the current speed can
reach up to 6 knots near the berths). Hydraulic
similarity was ensured by maintaining equality of the
dimensionless Froude number between the model and
the prototype, guaranteeing the correct relationship
between inertial and gravitational forces. Additionally,
the model was designed to ensure that the studied flow
conditions always remained in a rough turbulent
regime, avoiding scale effects associated with viscosity.
Beyond topographic and bathymetric features, as
well as the representation of port structures, the PHM
includes scaled-down vessels (Figure 1) constructed
based on real ship line plans and general
arrangements, preserving geometric similarity. Vessel
models calibration involved verifying the center of
gravity and radius of gyration to ensure an accurate
representation of the movements of the full-scale
vessels.
Figure 1. View of the real 400,000 DWT VLOC (left) and its
1:170 scale model (right).
The calibration of currents in the PHM (Figure 2)
was conducted using georeferenced measurements of
velocity, direction, and water level, based on field
surveys. Twenty-four (24) homologous points were
established to assess tidal current velocity and
direction, employing MicroADV sensors and
limnimetric probes. The PHM was validated to
accurately represent flood and ebb currents within the
local tidal range. While the PHM calibration process is
discussed in greater detail in [2], this paper focuses
only on the aspects most relevant to the present study.
Figure 2. General view of the case study PHM (Geometric
scale 1:170).
The position and elevation of the ship and berth’s
mooring elements (bollards, bitts, winches, chocks,
fairleads) are strictly respected, ensuring that the
angles formed by the mooring lines in relation to the
vertical and horizontal planes are equal to the real
values. The lines are positioned on the mooring
elements (of both pier and ship) according to the
predefined mooring arrangement.
Before the test begins, with the vessel fixed and
centered in the berth (Figure 3), pretensions
corresponding to 10% of the MBL (Minimum Breaking
427
Load) are applied to the lines. The MBL represents the
minimum nominal breaking load of a line. This
procedure follows the practice adopted in the mooring
of real ships.
Figure 3. Ship positioning process at the center of the berth.
The simulation system for unintentional unberthing
in the PHM consists of a set of mechanisms to which
the mooring lines are attached and positioned on the
ship's deck (Figure 4). These mechanisms allow each
line to be individually released by means of a remotely
sent signal. The release sequence of the mooring lines
can be carried out in different ways, either based on
hypothetical scenarios or real events observed in the
field.
After the release of the mooring lines, with the ship
drifting, a tracking system monitors the vessel's
position, allowing analysis of its movement, including
potential collisions with structures, displacement
direction and speed, as well as grounding risk. The
tracking system, known as ship tracking, operates
using images captured by a set of digital cameras
installed in a zenithal position and strategically
distributed along the PHM, ensuring full coverage of
the ship’s movement area during the tests (Figure 5).
The images generated are processed by a software
developed specifically for this purpose, which detects
and continuously tracks the position of markers
attached to the ship’s deck [1] (Figure 4).
Figure 4. Mechanisms (mooring elements in red) and
markers (triangle and rectangle in white) installed on the
ship's deck for unintentional unberthing simulation.
Figure 5. Digital camera system installed in a zenithal
position.
In addition, the software correlates the captured
images with a corresponding map of the study area,
enabling the georeferencing of the tests. To achieve
this, at least two reference points with known
coordinates are used in the image from each camera.
These points are marked on the floor of the PHM
(Figure 6).
Figure 6. Image of the vessel's position monitoring system,
highlighting reference points with known coordinates (in
meters).
Thus, at each moment in time, the relative position
between the deck markers and the reference points on
the model floor is calculated, determining the ship’s
location on the chart. Using this same system, it is also
possible to calculate the drift speed of the ship model.
428
For the case study presented in this paper, the
modeled port in the PHM has six berths (1 to 6),
numbered from North to South (Figure 7).
Figure 7. General layout of the case study port.
For the two simulation scenarios presented, one
VLOC (400,000 DWT), with the unintentional
unberthing system installed, was docked at berth 4,
chosen because it is one of the berths used by large
vessels and it is surrounded by adjacent berths, making
it a critical point in cases of ship drift. For the first
scenario, under flood tide conditions, another VLOC
was docked at berth 5. For the second scenario, under
ebb tide conditions, another VLOC was docked at the
berth 2.
Through tests to measure forces on the mooring
lines, previously carried out in the PHM, the most
stressed lines of the simulated mooring arrangement
were identified. The criterion adopted for the release
sequence of the lines followed the order from the most
stressed (highest measured forces) to the least stressed.
The release was carried out at regular intervals of 5
seconds (in scale model), which corresponds to
approximately 1 minute in full scale. To validate the
results, the simulation of each scenario was repeated 12
times. The adopted mooring arrangement included a
total of 20 HMPE (High Modulus PolyEthylene) lines
with a minimum breaking load (MBL) of 100 tf,
distributed as follows:
Table 1. Composition of the mooring arrangement.
Group
Nº of ropes
Rope material
(MBL)
After Breast 1
2
HMPE (100 tf)
After Breast 2
2
HMPE (100 tf)
After Breast 3
2
HMPE (100 tf)
After Spring 1
2
HMPE (100 tf)
After Spring 2
2
HMPE (100 tf)
Forward Spring 2
2
HMPE (100 tf)
Forward Spring 1
2
HMPE (100 tf)
Forward Breast 2
2
HMPE (100 tf)
Forward Breast 1
1
HMPE (100 tf)
Head 2
1
HMPE (100 tf)
Head 1
2
HMPE (100 tf)
Figure 8. Mooring arrangement of the ship, moored to
starboard.
The mooring lines release sequences for flood and
ebb tide conditions were:
Table 2. Mooring line release sequence for the VLOC docked
at the berth 4.
Vessel
Berth used in
the simulation
of ship drifting
Tide
condition
Release sequence
of mooring lines
Neighboring
berth with a
moored vessel
VLOC
(400,000
DWT)
4
Flood
FB1-FB2-FS2-H1-
AB3-AB2-AB1-
H2-FS1-AS1-AS2
5
Ebb
FB1-FS1-FS2-H1-
FB2-H2-AB3-AB1-
AB2-AS1-AS2
2
This method allows the assessment of damage risks
associated with the ship drift. The objective is to
provide technical support for defining risk mitigation
strategies in situations involving mooring line failures.
3 RESULTS AND DISCUSSION
The results of the two drifting test scenarios are
presented in Figures 9 and 10, which show the
trajectory of the twelve repetitions of the tests carried
out for each scenario: flood tide and ebb tide,
respectively. The analysis of these tests was based on
the following criteria:
− Allision: Checking for impacts of the unberthing
vessel with the berth structure or adjacent berths.
− Collision: Evaluating collisions between the
unberthing vessel and ships moored at neighboring
berths.
− Yawing: Identifying uncontrolled yawing
tendencies of the drifting vessel, which could pose
risks to tugboat approaches.
− Speed: Analyzing drift speed, considering that
speeds equal to or greater than 2 knots represent a
risk for safe tugboat operations, as indicated by the
port operation team and tugboat commanders.
In the flood tide scenario (Figure 9), the 400,000
DWT vessel was observed unberthing from the berth 4
and drifting. Allisions with the berth structure
occurred, as well as a collision with the vessel moored
429
at berth 5, and there was little tendency for yawing,
resulting in a drift toward an island located further
south, where the vessel grounded. The first line
released was the forward breast line, considered the
most stressful in the mooring arrangement. The
maximum observed speed of the vessel was 2.7 knots
(a value above the safety limit for tugboat approaches),
and the average distance traveled by the vessel was 1.3
km. Considering the vessel's behavior during the drift,
the scenario was classified as having a high potential
for damage to the port due to the allisions with the
berth structure, collision with the vessel at berth 5, and
the grounding tendency on the island. It is emphasized
that unintentional unberthing of ships at berth 4 could
lead to the suspension of navigation in the access
channel to berth 6.
Figure 9. Vessel movement after unintentional unberthing
during flood tide.
In the ebb tide scenario (Figure 10), the allision of
the vessel with the structure of berth 4 was observed,
along with a reduced yawing tendency and impact
with the access bridge to berths 1 and 2, resulting in the
vessel grounding at the back of berth 2, in a shallow
area. It is highlighted that the allision with the access
bridge to berths 1 and 2 could disrupt operations at
those berths.
The first line released was also the forward breast
line (highest observed forces). The maximum observed
speed was 3.7 knots (a value above the safety limit for
tugboat approaches), and the average distance traveled
was 0.8 km. This scenario was classified as having a
high potential for damage to the port, due to the
multiple allisions, particularly with the access bridge to
berths 1 and 2, and the high drift speed.
Figure 10. Vessel movement after unintentional unberthing
during ebb tide.
Table 3 presents a summary of the results obtained
in the tests, highlighting the main events observed:
allision, collision, grounding, and yawing, as well as
the maximum speed of the vessel. To facilitate data
interpretation, the table uses a color-coded system
indicating the probability of occurrence of each event:
− Red: High probability (event observed in more than
50% of the tests).
− Yellow: Moderate probability (event observed in
less than 50% of the tests).
− White: Low probability (event not observed in any
test).
In the case of maximum speed, the color scale
reflects the magnitude:
− Red: High speeds (above 3 knots).
− Yellow: Moderate speeds (between 2 and 3 knots).
− White: Low speeds (below 2 knots).
Since there was little tendency to yaw in both
scenarios, the "yawing" column of the table 3 was left
blank.
The last column of the table presents a qualitative
evaluation of the potential damage to the port,
calculated by the sum of points assigned to each event:
− Red: 2 points.
− Yellow: 1 point.
430
− White: 0 points.
Scenarios with a total of 5 or more points are
classified as high damage potential (red); between 1
and 4 points, as moderate damage potential (yellow);
and with no points, as low damage potential (white).
Table 3. Summary of drift test results conducted at berth 4.
Ship
(dwt)
Berth
Tide
Allision
Collision
Grounding
Yawning
Average
speed
(knots)
Potential
for
damage
400,000
4
Flood
2.7
7
Ebb
3.7
6
The tests demonstrated that the involuntary
unberthing of 400,000 DWT vessels at berth 4, followed
by free drift, represents a high-risk scenario for the
port, both under flood tide and ebb tide conditions.
Allisions with fixed structures of the port, collisions
with ships moored at adjacent berths, and groundings
were frequent events. Additionally, it is important to
highlight that drift speeds greater than 2 knots cause
difficulties for the operations of tugboats, increasing
the risk of accidents.
4 CONCLUSIONS
Physical hydraulic models are highly reliable tools for
analyzing risks associated with accidents at ports.
Mooring lines breakage can cause ship drifting. In this
case, it is essential to understand the possible scenarios
of the vessel's displacement in order to define
emergency action plans to mitigate damage to the
environment and port assets.
In the case study, for berth 4, both current
conditions (flood tide and ebb tide) result in scenarios
with high potential damage to the port. Allisions with
the own berth, collisions with ships moored at berth 5
(flood tide), and groundings were observed, either on
an island (flood tide) or on the back of berth 2 (ebb
tide). The collision with the access bridge to berths 1
and 2, in particular, represents a serious consequence,
potentially disrupting their operations.
This type of study highlights the importance of
rigorously following inspection and maintenance
procedures for mooring elements, ensuring that the
operation teams can rely on the established load limits.
Furthermore, when signs of breakage or overload of
the mooring lines are detected, it is crucial to
immediately call the tugboats, because most of the time
only quick action can avoid accidents.
Since the study did not address the possibility of
unberthing ships at adjacent berths due to collisions
(the "domino effect"), further tests can be conducted to
evaluate different scenarios of simultaneous
occupancy of adjacent berths and multiple
unberthings, which would provide a more
comprehensive assessment of the potential damage to
the port.
REFERENCES
[1] de Melo Bernardino, J.C. Abordagem experimental para
avaliação de manobras de navios em modelos físicos de
espaços náuticos. PhD thesis. 2015;
[2] Esferra, R., de Melo Bernardino, J.C., Alfredini, P. Physical
modeling applied in evaluation of the safety and
efficiency of vessel mooring plans. RBRH. 2018;
[3] Gao X, Makino H, Furusho M. Analysis of a ship drifting
in a narrow channel using Automatic Identification
System (AIS) data. WMU J Marit Affairs. 2016;
[4] Montewka J, Ehlers S, Goerlandt F, Kujala T. Risk
assessment of ship grounding and its environmental
consequences. Mar Pollut Bull. 2012;
[5] Uittenbogaard RE, Battjes JA, Stive MJF. Physical
modeling of estuarine hydraulic phenomena using scale
models. Coast Eng. 2005;
[6] Zhang S, Pedersen PT, Villavicencio R. Probability and
mechanics of ship collision and grounding. 1ª ed. Oxford:
Butterworth-Heinemann. 2019;