Ocean Engineering 97 (2015) 207–217

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

Motion responses of a moored barge in shallow water

Lingzhi Xiong a, Haining Lu a,n, Jianmin Yang a, Wenhua Zhao b
a
State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai, China
b
Faculty of Engineering, Computing and Mathematics, The University of Western Australia, 35 Stirling Highway, Crawley 6009, WA, Australia

art ic l e i nf o a b s t r a c t

Article history: Motion characteristics of a floating structure in shallow water are of great concern in ocean engineering.
Received 30 May 2014 Shallow water effects will significantly affect the hydrodynamic performance of a floating structure. In
Accepted 31 January 2015 this study, both numerical and experimental studies have been conducted to investigate the hydro-
Available online 25 February 2015
dynamic performance of a barge in shallow water. Numerical simulations are conducted in both
Keywords: frequency and time domains based on 3D potential theory. Second-order wave forces have been
Shallow water effects incorporated in the numerical model through the calculation of fully quadratic transfer function. The
QTF numerical results are validated through a series of physical model tests, including free decay test and
Numerical simulations irregular wave tests. Responses of the floating structure in different water depths and sea states are
Physical model tests
studied to clarify the safe conditions for a float-over installation in shallow water. Based on the
Moored barge
numerical and experimental results, some conclusions have been drawn. Furthermore, a critical depth
for shallow water effects is clarified.
& 2015 Elsevier Ltd. All rights reserved.

1. Introduction wave height in the Gulf of Bohai can reach up to 6 m in extreme

Traditionally, fixed offshore platforms have been widely applied in
shallow waters for the exploitation of oil and gas resources. However,
nowadays more floating structures have been proposed to be applied
in shallow waters. For example, the Floating Production Storage and
Offloading system (FPSO) is widely adopted in shallow waters due to
its low cost and large capacity of oil storage avoiding the usage of
subsea pipelines in a fixed platform program (Xiao and Yang, 2006).
More recently, a new concept Floating Storage and Regasification Unit
(FSRU) is also proposed as a promising alternative for onshore
Liquefied Natural Gas (LNG) terminals due to the environmental
and safety issues (Lee et al., 2010). Shallow water area that is not
far away from the coast line is the preferred site for FSRU considering
the technical efforts and financial risks (Kim et al., 2012). It is reported
that seven FPSOs and one floating LNG terminal have been in service
in the Gulf of Bohai in China, where the average water depth is only
18 m (Li et al., 2014). Additionally, a lot of large Offshore Support
Vessels (OSVs) also operate in this shallow water region such as the
derrick vessel for heavy lift operation and the float-over barges for
topside installation.
Although the floating structures show advantages in this area,
the shallow water depths introduce more technical challenges
such as the strong nonlinearities in the propagating waves and the
complex wave-structures interactions. In particular, the significant
n
Corresponding author.
E-mail address: haining@sjtu.edu.cn (H. Lu).
storms with the period of 10.5 s (Xiao et al., 2014). The large wave
height and the relatively shallow water depth enhance the non-
linearities and result in more challenges.
There have been research efforts in the estimation of the
motion response of a moored structure in shallow waters. Molin
and Fauveau (1984) carried out an analysis on the set-down wave
in deep and shallow water depths. It is observed that the loads
induced by the set-down in shallow water contribute to the slow-
drift motion of moored structures. Recently, Yang et al. (2002)
used an experimental method to investigate the motions of a
single-point moored (SPM) FPSO in different shallow water
depths. Li et al. (2003) further conducted numerical studies to
analyze the shallow water effects. It is observed that the wave
frequency motions decrease with the reduction of water depth,
while the low-frequency motions increase at the same time. Naciri
et al. (2004) studied the low-frequency motions of LNG carriers
moored in shallow water. It is concluded that the shallow water
effects play an important role in wave drift forces. When the water
depth decreases, the wave drift forces increase. Wim and Ivo
(2008) used a new method consisting of a Boussinesq-type wave
model with a linear-domain diffraction model to calculate the
wave forces on a moored ship in irregular wave in shallow water. It
is concluded that both time-domain and frequency-domain meth-
ods are capable of dealing with the non-linear incident waves. The
frequency-domain method is more efficient, while the time-
domain method is practical with good accuracy. Pinkster (2009)
studied an LNG carrier moored in shallow water. It is shown that
bound waves appear to affect the low-frequency forces more than

http://dx.doi.org/10.1016/j.oceaneng.2015.01.018
0029-8018/& 2015 Elsevier Ltd. All rights reserved.
208 L. Xiong et al. / Ocean Engineering 97 (2015) 207–217

the first-order waves, which is typically a shallow water effect. Yan Reinholdtsen and Falkenberg (2001)
et al. (2010) investigated the fully nonlinear interaction between a

gA cosh kðz þ HÞ
moored FPSO and the shallow water waves, together with the ϕI ¼ cos ωt  kx cos β ky sin β þ φζ ; ð3Þ
effects of water depth on those issues. The results show that the
ω cosh kH
wave induced force decreases but the nonlinear components and where A represents the wave amplitude, g the acceleration of
the surge motions become more significant as the water depth gravity, ω the wave frequency, β the direction of wave propagation
decreases. The low-frequency motion response plays a critical role and φζ the wave component phase angle. The wave number k
in the design of mooring systems due to the significant resonant satisfies the dispersion relation:
responses which might be induced by the second-order wave
ω2
forces. ¼ k tanh kH: ð4Þ
g
To estimate the second-order wave excitation forces, Newman
(1974) proposed an approximation to get the Quadratic Transfer The fluid velocity components can be computed from Eq. (4). It
Function (QTF) matrix, by using the mean drift force to interpolate suggests that trajectories of fluid particle will become elliptical in
the off-diagonal values. Naciri and Poldervaart (2004) studied the shallow water, which would induce the shallow water effects.
low-frequency motions and design aspects of LNG carriers moored Each part in Eq. (2) can be solved under their corresponding
in shallow water. The Newman approximation and full QTF boundary conditions, which is shown as follows:
approach are expected to yield comparable responses in the sea 8 ∂2 ϕ
state with a relatively short peak period. However, studies (Chen, >
> þ g ∂∂zϕ ¼ 0; at z ¼ 0
>
> ∂t 2
1994; Tahar and Kim, 2003; Pessoa and Fonseca, 2013) have >
> ! ! ∂ðϕI þ ϕD Þ
> ϕR
< ∂∂n ¼U  n; ¼ 0; at body surface
indicated that Newman approximation underestimates the low- ∂n
∂ϕ ; ð5Þ
frequency response of a floating vessel in shallow water. Newman >
> ∂n ¼ 0; at z ¼ H
>
>

>
> pffiffiffi
(2004) confirmed that the zero-order approximation is poor as the >
: lim R ∂∂Rϕ  ikϕ ¼ 0
water depth is less than 100 m. Furthermore, Grant and Holboke R-1

(2004) concluded that the Newman assumption has limited !

applicability in cases with shallow water and long wave periods,
and full QTF is recommended in shallow water. Xiao (2007)
conducted a study regarding the shallow water effects on a soft
yoke moored FPSO. Results based on Newman approximation
show good agreement with experimental data in general water
depth. However, the agreement will become less satisfactory as
the water depth decreases. Guillaume et al. (2012) reviewed the
approximation methods for the calculation of the second-order
wave loads and suggested that the full QTF should be calculated in
shallow water to obtain a good prediction.
In this study, a direct second-order pressure integral method is
adopted to calculate the full QTF of a moored barge at Gulf of Bohai
with water depth of 9.41 m. Coupled analysis has been conducted
in time domain to estimate the motion response of the barge and
loads on the mooring lines. Numerical simulations are completed
by the commercial software SESAM. Model tests are conducted in
the wave basin at Shanghai Jiao Tong University to validate the
numerical results. The numerical results agree well with the
experimental data. Based on the numerical and experimental
study, some conclusions have been drawn regarding the hydro-
dynamic performance of the referenced floating barge in shallow
waters.
2. Numerical modeling
2.1. Potential theory
With the assumption of irrotational motion and incompressible
fluid, the fluid motion can therefore be described by a velocity
potential ϕ, which satisfies the Laplace equations:
∇2 ϕðx; y; z; t Þ ¼ 0;
where n stands for the normal vector of the corresponding
!
surface, U the velocity vector of body surface and H the finite
water depth.
Once the velocity potential is known, it is easy to calculate the
hydrodynamic forces of the barge, including the wave exciting
force, wave radiation force and restoring force by the pressure
integration on the wet surface. At the same time, hydrodynamic
coefficients such as the added mass and the potential damping can
be obtained from the radiation potential (Bingham, 2000; Hong
et al., 2005).
2.2. Motion equation of a floating body
Motion equation of the barge can be expressed in time domain
as follows (DNV, 2011; Zhang et al., 2013; Zhao et al., 2011):
n o n o
 n o
   R t
½m þ að1Þ ξ€ þ D1 ξ_ þ D2 f ξ_ þ K ξ ξ þ 0 hðt  τÞ ξ_ dτ
¼ F wave þ F wind þ F current þ F ext ; ð6Þ
where m is the mass matrix for the barge  hull, að1Þ the added
mass matrix at the infinite frequency, ξ the matrix of vessel’s six
DOF motion, K the hydrostatic restoring stiffness matrix, and D1
and D2 the linear and quadratic damping matrix, respectively. The
forces excited by waves, wind and current are represented by
F wave ; F wind ; and F current respectively. The last item F ext denotes the
other possible force such as the specified forces, mooring force and
so on. The retardation function hðτÞ represents the memory effect
in the free surface. It can be obtained from the added mass matrix
a and potential damping matrix c:
Z þ1
1
hðτÞ ¼ ½cðωÞ þ iωaðωÞeiωt dω; ð7Þ
ð1Þ 2π  1

where x; y; z are the spatial coordinates with xy-plane coincides where ω is the frequency. It should be noted that the viscous
with the calm water and z-axis points upwards, t represents time. damping and the wave n o drift damping should also be included in
Rt
The total velocity potential can be divided into three parts: the term 0 hðt  τÞ ξ_ dτ in the form of critical damping, because
incident potential ϕI , radiation potential ϕR and diffraction poten- these variants are related to the motion velocity of the vessel
tial ϕD , namely (Zhao et al., 2014): (Zhao et al., 2014).
Using the fact that cðωÞ ¼ cð  ωÞ and aðωÞ ¼ að  ωÞ, it gives
ϕ ¼ ϕI þ ϕR þ ϕD : ð2Þ
Z Z
1 1 2 1
The incident potential ϕI for a regular wave at finite water hðτÞ ¼ ½cðωÞ cos ωτ  ωaðωÞ sin ωτdω ¼ cðωÞ cos ωτdω
π 0 π 0
depth is, according to Airy’s wave theory, expressed by
 L. Xiong et al. / Ocean Engineering 97 (2015) 207–217 209

Z 1
2 (1) Contribution due to products of first-order angular motions
¼ ωaðωÞ sin ωτdω: ð8Þ
π 0 and inertia forces,
(2) Pressure drop due to first-order velocity,
(3) Pressure due to product of gradient of first-order pressure and
There are three methods to calculate the low-frequency loads and
first-order motion,
motions, such as near-field method, middle-field method and far-
(4) Second-order potential,
field method. The near-field method uses the integral on the body
(5) Influence due to couple of the first-order motion,
surface, which is direct and easy to be understood. The middle-field
(6) Contribution due to first-order relative wave elevation.
method uses the integral on a control surface and consists of body
surface, free surface and the surface in the field. The derivation of
The second-order moment can be derived in the same way.
middle-field method starts with the most general near-field method
From Eq. (13), it is not difficult to get the following equation in
formulation. Then it obtains a near-field formulation by using two
random sea (ignoring the sum frequency items):
variants of Stokes theorems. The analysis in a finite volume limited
by the hull and a control surface surrounding the body yields new !ð2Þ XN XN
1 ð1Þ ð1Þ   
middle-field formulation (Chen, 2007). The far-field method uses the F ¼ η ηj P ij cos ωi  ωj t þ ðεi  εjÞ
i¼1j¼1
2 i
integral on the infinite control surface. It is simple and direct, but
only calculates the mean drift forces and cannot provide the drift X
N X
N
1   
þ ηi ð1Þ ηj ð1Þ Q ij sin ωi  ωj t þ ðεi  εjÞ : ð14Þ
loads which can be important in shallow water case. In this study, the i¼1j¼1
2
near-field method is used. The clearness physical meaning of each
item and its simplicity is the obvious advantage for the simulation. It is easy to know that P ij ; Q ij is the in-phase and quadrature
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

The main procedure of this method is to get the non-linear parts of the QTF T ¼ P 2 þ Q 2 , details can be found in Pinkster
pressure on the body surface, then integrate on the wet surface to
the second-order (Pinkster, 1975) (1975).

The spectral density of the low-frequency components of the
!   !ð0Þ !ð1Þ !ð2Þ !ð0Þ
F ¼ ∬ pð0Þ þ εpð1Þ þ ε2 pð2Þ n þ ε N þ ε2 N dS þ ∬ εpð1Þ n dS force in Eq. (14) can be written as follows:
S0 ΔS Z 1
   2
!ð0Þ !ð1Þ !ð2Þ   SF μ ¼ 8 Sζ ðω þ μÞ  Sζ ðωÞ  Tðω þ μ; ωÞ dω; ð15Þ
¼ F þ ε F þ ε2 F þ o ε3 ; ð9Þ 0
 
where SF μ is the spectral density of the force, Sζ ðωÞ is the
where S0 stands for the mean wet surface and ΔS the instant
spectral density of the wave.
means wet surface. The hydrostatic pressure pð0Þ , first-order
pressure pð1Þ , second-order pressure pð2Þ can be obtained from
the following equations: 2.3. Mooring dynamics

pð0Þ ¼  ρgx3 ð0Þ Coupled analysis has been conducted regarding a moored
ð1Þ
∂ϕ barge. In the coupled approach, the vessel motions and the
p ¼  ρgX 3  ρ
ð1Þ ð1Þ
;
∂t !
mooring dynamics are solved simultaneously at each time step.
1   ∂ϕ
ð2Þ
!ð1Þ ∂ϕ
ð1Þ Therefore, the full interaction between the barge and the mooring
ð1Þ 2
p ¼  ρgX 3  ρ∇ϕ   ρ
ð2Þ ð2Þ
ρ X  ∇ ð10Þ lines is accounted overcoming the limitations of an uncoupled
2 ∂t ∂t
approach, especially when non-linearity is dominant. The mooring
!
where the instantaneous position vector X in a global coordinate lines are treated as a finite element model. The spatial discretized
!
system of the point fixed on the body fixed coordinate system x is system dynamic equilibrium equation is governed by DNV (2005)
given by  
RI r; r€ ; t þ RD ðr; r_ ; t Þ þ RS ðr; t Þ ¼ RE ðr; r_ ; t Þ; ð16Þ
! ! ! ! ! !ð0Þ !ð1Þ !ð2Þ
X ¼ x þ x0 þ α  x þ D x ¼ x þ ε X þ ε2 X ; ð11Þ where RI ; RD ; RS and RE represent inertia force vector, the damping
!
where x0 stands for the translational modes, α ¼ ½α1 α2 α3  the force vector, the internal reaction force vector and the external
!ð0Þ !ð1Þ force vector respectively. The structural displacement, velocity and
rotational modes, x the zero-order position vector, X the acceleration vectors are denoted by r; r_ and r€ .
!ð2Þ The inertia force vector is expressed as
first-order position vector, and X the second-order position
 
vector. The transformation matrix D is given by Pinkster (1975) R r; r€ ; t ¼ MðrÞr€ ;
I
ð17Þ
0 1 2  1
 2 α2 þ α3 2 0 0 and the damping force vector is expressed as
B  2  C
D¼B α1 α2  2 α1 þ α3
1 2
0 C: ð12Þ
@   A RD ðr; r_ ; t Þ ¼ CðrÞr_ ; ð18Þ
α1 α3 α2 α3  12 α1 2 þ α2 2
where M is the system mass matrix that includes structural mass
!ð2Þ and hydrodynamic mass. The damping matrix C includes both the
Then, we can obtain the expression of the F as follows:
internal structural and hydrodynamic dampings.
!ð2Þ 1!ð1Þ
F ¼ α
ε
!
1   ð1Þ
!ð1Þ ð1Þ 2 !ð0Þ !ð1Þ ∂ϕ !ð0Þ 3. Description of the reference barge
 F ∬ ρ∇ϕ  n dS  ∬ ρ X  ∇ n
S0 2 S0 ∂t
ð2Þ
A T-shape barge used for floater-over installation is selected as
∂ϕ !ð0Þ !ð0Þ !ð2Þ
dS ∬ ρ n dS  ∬ ρgX 3 ð2Þ n þ ρgx3 ð0Þ N the reference in this study. During the installation, the barge is
S0 ∂t S0 moored by four mooring lines for positioning purpose. The main
Z
1 !ð0Þ particulars of the barge in prototype and model scale (scale ratio is
dS þ ρgηr  n dl:
ð1Þ2
ð13Þ
2 1:30) are listed in Table 1. Fig. 1 presents the panel model and the
WL
second-order free surface for the numerical simulations. The
The components represent: details of the mooring system are shown in Table 2. Material
210 L. Xiong et al. / Ocean Engineering 97 (2015) 207–217

properties of the mooring lines are shown in Table 3. Fig. 2 shows monitor the clearance between the barge bottom and the seabed
the barge moored by 4 mooring lines. As shown in Fig. 2, the in case of collision. If collision happens, the displacement sensor
mooring lines are designed totally different with each other, due to will record the signal. Waves were calibrated prior to the model
the fact that the sea states and seabed are quite complex in tests. The calibration results are shown in Fig. 4. For each case, the
this area. time duration is 33 min in model scale corresponding to 3 h in
The wave, wind and current are all in collinear direction. prototype. The sampling frequency is 40 Hz in model scale.
Constant wind and current are applied with a speed of 10.7 m/s
and 1.26 m/s, respectively. The JONSWAP spectrum with a peaked-
ness parameter of 1.8 is adopted to represent the random waves.
Table 2
Different water depths are set to get the characteristics of the ship Configurations of the mooring system.
motion in shallow water. The water depths are 9.41 m, 10.31 m,
11.39 m, 13.93 m and 16.88 m in prototype. In this study, all the Mooring line Segment 1 Segment 2
ID
parameters are presented in prototype if there is no special
Type Diameter Length Type Diameter Length
annotation. (mm) (m) (mm) (m)

SPM-FS Chain 76 150 Wire 76 585

SPM-AS Chain 76 150 Wire 76 712
4. Experimental set-ups SPM-FP Rope 76 135 / / /
SPM-AP Wire_2 76 733 / / /
Model test of the moored barge was conducted in the wave
basin, State Key Laboratory of Ocean Engineering, Shanghai Jiao
Tong University. The basin is 50 m in length, 30 m in width and Table 3
6 m in depth. The artificial bottom can be adjusted from 0 to 6 m. Material properties of the mooring lines.
The scale between model and prototype is 1:30. Fig. 3 is a
Property Weight in air (kg/m) Weight in water (MT/km) EA (N)
snapshot of the barge in irregular waves in the wave basin.
A non-contact optical motion tracking system was used to CHAIN 126 110 4.90E þ 08
capture the 6-DOF motions at the reference point (center of WIRE 24 21 2.60E þ 08
gravity). Force transducers and accelerometers were utilized to ROPE 4.5 0 1.30E þ 08
WIRE_2 24 21 1.1E þ08
measure the mooring line tensions and barge accelerations
respectively. Two wave gauges were set in front of the barge and
in the port of the barge to monitor the ambient wave height. Four
displacement sensors were arranged in the barge’s four corners to

Table 1
Main particulars of the Barge.

Parameters Unit Prototype Model

Length m 180.0 6.0

Breadth m 52.5/36.0 1.75/0.60
Depth m 12.75 0.425
Draft m 8.44 0.281
Displacement ton 67,855 2.452
X COG m 80 2.700
YCOG m 0.00 0.000
ZCOG m 6.5 0.217
Rxx m 15.45 0.515
Ryy m 52.32 1.744
Rzz m 54.12 1.804
Fig. 2. Mooring configuration of the barge.

Fig. 1. Panel model and second order surface model: (a) second order surface; (b) panel model.
 L. Xiong et al. / Ocean Engineering 97 (2015) 207–217 211

5. Results and discussion
5.1. Static offset test
To verify the mooring system in model test, static offset tests
were conducted to obtain the static stiffness curves. This is the
basic requirement for the dynamic analysis later. The comparison
between the measured data and the target value is shown in Fig. 5.
One can see in Fig. 5 that a good agreement has been achieved.
5.2. Free decay tests
Free decay tests in still water were conducted to get the natural
periods and damping levels of the vessel in different
  water depths.
The water depths are expressed with the ratio γ of water depth
to draft (8.44 m). γ equals 1.12, 1.22, 1.35, 1.65 and 2. Results for
heave, roll and pitch motions are presented as follows.
As shown in Fig. 6, the natural period is very sensitive to the
water depth in shallow water. It is apparent that with the decrease
of the ratio γ , the natural period of barge motion increases and
shifts away from the encountered wave period in the Gulf of Bohai.
Because the added mass increases with the decrease of the water
depth, especially when the ratio γ is less than 1.35, a slight
reduction of water depth leads to a large increase of the added
mass. As we know, the natural period of the barge is proportional
to the square root of the sum of the barge mass and the added
mass. Thus, the natural period increases significantly that will
avoid the encountered wave period in design environment condi-
tions as a result. In this way, the motion amplitude drops lower,
keeping the safe bottom clearance.
Meanwhile, the roll damping coefficient gets larger with the
decrease of γ as Fig. 7 shows. It may be related to that the potential
damping of the barge increase dominantly, which affects the roll
damping.
One can further observe in Figs. 6 and 7 that the shallow water
exhibits obvious influences on the natural period when γ varies
from 1.12 to 2.0, while it shows different effects on the damping.
For instance, when γ decreases from 1.22 to 1.12, the damping
keeps stable. The reason for this phenomenon is that the set-down
wave appears when the water depth decreases. The set-down
wave becomes dominant and results in the first-order wave-
exciting force in low-frequency which influences the natural
period while does not affect the damping. Xiao (2007) also gives
similar analysis to this phenomenon.

5.3. Response amplitude operators

Calculations of motion RAOs have been conducted in different

directions and water depths. Numerical results are compared with
Fig. 3. Snapshot of the barge in irregular waves in the wave basin. the experimental data. Fig. 8 presents the comparison of 6-DOF

0.15 0.08
Measured Measured
Target
0.12 Target
0.06
Spectrum(m ·s/rad)

Spectrum(m ·s/rad)

0.09
2

0.04
0.06

0.02
0.03

0.00 0.00
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Frequency (rad/s) Frequency (rad/s)

0.04 0.04
Measured Measured
Target Target
0.03 0.03
Spectrum(m ·s/rad)

Spectrum(m ·s/rad)
2

0.02 0.02

0.01 0.01

0.00 0.00
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5
Frequency (rad/s) Frequency (rad/s)
Fig. 4. Target and measured wave spectra: (a) Hs ¼1 m, Tp ¼ 5.8 s, (b) Hs ¼0.75 m, Tp ¼ 5.5 s, (c) Hs ¼0.5 m, Tp ¼ 5 s, (d) white noise, Hs ¼ 0.75 m, Tp ¼4–25 s.
212 L. Xiong et al. / Ocean Engineering 97 (2015) 207–217

5000 0
Tar,γ=1.12

Restoring Force (kN)

Restoring Force (kN)

4000 Tar,γ=1.65
Tar,γ=2.0 -1000
Mea,γ=1.12
3000
Mea,γ=1.65 Tar,γ=1.12
Mea,γ=2.0 -2000
Tar,γ=1.65
2000 Tar,γ=2.0
Mea,γ=1.12
-3000
1000 Mea,γ=1.65
Mea,γ=2.0
0 -4000
-4 -3 -2 -1 0 0 1 2 3 4 5 6
X Offset (m) Y Offset (m)
Fig. 5. Offset static stiffness curve of mooring system in Surge/Sway direction.

25 The water depth of 3,000 m that is regarded as the infinite water

depth is also calculated and presented.
Heave
Roll
For the surge, sway and yaw motions, obvious differences can
Natural period (s)

20
Pitch be found between shallow water and infinite water depth cases.
Even though the ratio γ reaches up to 4, the difference in low-
15 frequency waves is dominant, while the responses in wave
frequency are relatively similar to each other. Furthermore, when
the water depth gets larger, the RAO curves get closer to the
10 infinite water case. Therefore, it can be concluded that when the
ratio γ reaches up to 4, the shallow water depth still affects the
low-frequency motion.
5
1.0 1.2 1.4 1.6 1.8 2.0 2.2 In the heave motion, when γ reaches up to 3 or larger, the
shallow water curves are extremely close to the infinite case.
Water depth/draft (γ)
However, as for roll RAOs, good agreements have been shown in
Fig. 6. Natural period variations with the ratio ðγ Þ. wave frequency, while some differences have been shown in low-
frequency. Because the low-frequency is away from the resonant
0.13 frequency of roll motion and the roll response is limited, it exhibits
little influences on the barge’s roll motion. The shallow water
Roll
exhibits similar effects on pitch motion.
Damping coefficient

0.12 From the analysis above, it is concluded that the low-frequency

motions (surge, sway and yaw) are more sensitive to the shallow
water effects than the wave frequency motions (heave, roll and
0.11
pitch). Furthermore, when γ Z 3, the wave frequency motions are
very close to the infinite results; when γ o 3, the shallow water
0.10 effects cannot be ignored and should be treated with great care.

5.4. Barge motions in random sea states

0.09
1.0 1.2 1.4 1.6
Water depth/draft γ
Fig. 7. Roll damping variations with the ratio ðγ Þ.
motion RAOs between the model test and the numerical simula-
tion at γ ¼1.12. As shown in Fig. 8, RAOs in heave, roll, pitch and
yaw directions are in good agreement with the model test. While
the calculated surge and sway RAOs show a slight difference. This
is caused by the mooring system. The RAOs are calculated based
on the potential theory without considering any mooring system,
while the barge is moored in the wave basin to avoid drifting off.
The mooring system in the model tests has a certain influence on
the low-frequency motion. Besides, there is non-linearity in
shallow water which is difficult to predict. It also has certain
influences on the RAOs. Therefore, the RAOs show a slight
difference. In fact, the difference is acceptable.
To clarify the effects of water depths, numerical simulations of
the barge in different water depths are shown in Fig. 9. The water
depths of 13.93 m, 16.88 m, 25.32 m and 33.76 m, with a corre-
 
sponding ratio γ of 1.65, 2, 3, and 4, are considered in this study.
1.8 2.0 2.2
Barge motions in random sea states are great concerned in
shallow water. Moored barge exhibits resonant low-frequency
motions in the horizontal plane, especially surge motion in head
sea and sway motion in beam sea. The resonant low-frequency
responses can significantly affect the barge motions and thus
impact the performance of mooring lines and risers. Due to the
limitation of Newman approximation, full QTFs are adopted to
reveal the shallow water effects.
One can see in Fig. 10 that the maximum amplitude of the surge
QTF in shallow water is much larger than that in deep waters.
Large responses appear near the diagonal area and mainly on the
low-frequency area. This means when two low-frequency waves
meet, the difference frequency is very small. For example, if the
difference frequency of the waves is 0.04 rad/s which is close to
the natural frequency of surge, a very large force in surge motion
will be induced in shallow water, as shown in Fig. 10(a). It can be
seen in Fig. 10(b) that the responses in the diagonal area are
nearly flat.
In the Newman approximation method, the diagonal values are
used to interpolate the off-diagonal elements in QTF matrix.
However, this handling will induce a large difference when
 L. Xiong et al. / Ocean Engineering 97 (2015) 207–217 213

2.5 5

2.0 Cal,γ=1.12 4 Cal,γ=1.12

Test,γ=1.12 Test,γ=1.12
Surge (m/m)

Sway (m/m)
1.5 3

1.0 2

0.5 1

0.0 0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Frequency (rad/s) Frequency (rad/s)

1.4 10

1.2 Cal,γ=1.12
8 Cal,γ=1.12
Test,γ=1.12 Test,γ=1.12
1.0
Heave (m/m)

Roll (deg/m)
6
0.8

0.6 4
0.4
2
0.2

0.0 0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Frequency (rad/s) Frequency (rad/s)

2.5 2.5

Cal,γ=1.12 Cal,γ=1.12
2.0 2.0
Test,γ=1.12 Test,γ=1.12
Pitch(deg/m)

Yaw(deg/m)

1.5 1.5

1.0 1.0

0.5 0.5

0.0 0.0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Frequency (rad/s) Frequency (rad/s)
Fig. 8. Comparison between model test and calculation of barge RAOs in γ ¼ 1.12: (a) surge, (b) sway, (c) heave, (d) roll, (e) pitch, and (f) yaw.

predicting the full QTF in shallow water, leading to a large
deviation in predicting the barge motions.
Besides the surge motion, the heave motion is directly con-
cerned due to the possibility of bottom collision. A series of model
tests have been conducted in different water depths to validate the
numerical simulations. Some representative comparison of heave
motion between the model test and numerical simulation are
shown in Fig. 8 in the form of response spectrum. To provide a
quantity analysis, the statistical analysis is also conducted and
some representative results are listed in Table 4.
Table 4 shows the minimum and the standard deviation of
heave motions at position of Sensor 1 (shown in Fig. 2). The six
DOF motions are measured at the center of gravity of the vessel
during the model tests, and the heave motions at Sensor 1 are
calculated based on the measured data. Table 4 reveals that the full
QTF provides a good prediction of the heave motion, while the
Newman approximation underestimates it, especially when the
water depth is shallow. For example, the difference between
Newman approximation and the test is 31.19%, while the differ-
ence between full QTF and the test is only 5.95% in Case 4. In this
case, although the significant wave height is large (H s ¼1 m,
bottom clearance is 0.97 m), the barge will not collide with the
seabed, because the peak period is far away from the resonant
period of the barge in six directions. Most of all, the heave
response at 5.8 s is quite small, thus the barge will be safe in
this case.
6. Sensitivity studies
Sensitivity studies have been conducted in different
  conditions
such as significant wave height ðH s Þ, peak periods T p , and water
depths ðH Þ. Parameters are selected according to the sea states in
Gulf of Bohai, for example, the average significant wave height is
1 m and peak period 4.8 s. Therefore, H s is selected from 0.5 m to
1.5 m with a step of 0.25 m; T p has the following values: 5 s, 5.5 s,
214 L. Xiong et al. / Ocean Engineering 97 (2015) 207–217

2.5 3.5

Cal,γ=1.65 3.0 Cal,γ=1.65

2.0 Cal,γ=2 Cal,γ=2
Cal,γ=3 2.5 Cal,γ=3

Sway (m/m)
Cal,γ=4
Surge(m/m)

Cal,γ=4
1.5 2.0 Cal,γ=inf
Cal,γ=inf

1.0 1.5

1.0
0.5
0.5

0.0 0.0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Frequency (rad/s) Frequency (rad/s)

1.4 8

1.2 Cal,γ=1.65 7 Cal,γ=1.65

Cal,γ=2 Cal,γ=2
6
1.0 Cal,γ=3 Cal,γ=3
Heave (m/m)

Roll (deg/m)
Cal,γ=4 5 Cal,γ=4
0.8 Cal,γ=inf Cal,γ=inf
4
0.6
3
0.4
2
0.2 1

0.0 0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Frequency (rad/s) Frequency (rad/s)

1.6 2.0

1.4 Cal,γ=1.65 Cal,γ=1.65

Cal,γ=2 1.6 Cal,γ=2
1.2 Cal,γ=3 Cal,γ=3
Pitch (deg/m)

Cal,γ=4 Cal,γ=4
1.0
Yaw (deg)

Cal,γ=inf 1.2 Cal,γ=inf

0.8

0.6 0.8

0.4
0.4
0.2

0.0 0.0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Frequency (rad/s) Frequency (rad/s)
Fig. 9. RAOs in different water depths: (a) surge, (b) sway, (c) heave, (d) roll, (e) pitch and (f) yaw.

Fig. 10. QTF for surge (0–01) in different water depth: (a) γ ¼ 1:12 and (b) γ ¼ inf .
 L. Xiong et al. / Ocean Engineering 97 (2015) 207–217 215

Table 4
Comparison between model test and calculation in random sea.

Case no. γ Hs (m) Tp (s) Direction (deg) Heave (m) Heave Std.

Test Full QTF Newman Test Full QTF Newman

Case 1 1.12 1 5.8 180  0.040  0.037  0.018 0.010 0.008 0.004
Case 2 1.22 1 5.8 0  0.056  0.057  0.041 0.013 0.011 0.009
Case 3 1.35 1 5.8 180  0.050  0.045  0.043 0.014 0.011 0.010
Case 4 1.12 1 12 135  0.420  0.445  0.289 0.138 0.097 0.073
Case 5 1.12 0.75 5.5 90  0.060  0.066  0.047 0.015 0.018 0.015

Note: Newman refers to the calculation method using Newman approximation.

0.20 0.20
9.41m 9.41m
Tp=5.5s Hs=0.75m
11.39m 11.39m
0.15 13.93m 0.15 13.93m

Min Heave (m)

Min Heave (m)

0.10 0.10

0.05 0.05

0.00 0.00
0.4 0.6 0.8 1.0 1.2 1.4 1.6 4.5 5.0 5.5 6.0 6.5 7.0 7.5

Hs (m) Tp (s)

Fig. 11. Heave sensitivity: (a) fixed Tp� and (b) fixed Hs. Min heave means the minimum z of the rigid barge.

1.5 0.8
9.41m 9.41m
Tp=5.5s Hs=0.75m
1.2 11.39m 11.39m
13.93m 0.6 13.93m
Surge (m)

Surge (m)

0.9
0.4
0.6

0.2
0.3

0.0 0.0
0.4 0.6 0.8 1.0 1.2 1.4 1.6 4.5 5.0 5.5 6.0 6.5 7.0 7.5
Hs (m) Tp (s)
Fig. 12. Surge sensitivity: (a) constant Tp and (b) constant Hs.

6 s, 6.5 s and 7 s; and three water depths are adopted: 9.41 m,
11.39 m and 13.93 m.
6.1. Heave motion
In this section, the absolute values are selected as the repre-
sentative results.
Sensitivities to significant wave height ðH s Þ for heave motion
are shown in Fig. 11(a). It can be seen in Fig. 11(a) that when H s
increases from 0.5 m to 1.5 m with fixed T p ¼5.5 s and water
depth, the increased mean amplitude of the heave motion reaches
up to 41%
 at each step. Fig. 11(b) shows the sensitivities to peak
period T p for heave motion. As shown in Fig. 11(b), the heave
motion increases evidently when the T p increases from 5 s to 7 s.
The increased amplitude is 52% at most and 22% for the average at
each step. One can see in Fig. 11 that the heave motion of the
moored barge is less sensitive to T p than H s . This phenomenon can
be explained through the results in Figs. 8 and 9. When the wave
period is 7 s or less, it is far away from the resonant period of the
system in six degree of freedom, resulting in a small motion
response. Small variances of T p will have little influences on the
barge’s heave motion.
Influence of the water depths on the heave motion is shown in
Fig. 11. Generally, a reduction of the water depth leads to a
significant decrease of the heave motion. This phenomenon is
one aspect of shallow water effects. It means that heave motion in
shallow water will be mitigated, which benefits the bottom
clearance.
6.2. Surge motion
The motions of the moored barge in the horizontal plane can
also be significantly affected by shallow water effects. Taking surge
motion as an example, comparison results are shown in Fig. 12.
Sensitivities to H s for surge motion are shown in Fig. 12(a). The
surge motion increases with the growth of H s . At each increase
216 L. Xiong et al. / Ocean Engineering 97 (2015) 207–217
0.4
Surge
0.3
Surge (m)
0.2
0.1
0.0
-2 0 2 4 6
Damping (% Cr)
Fig. 13. Low-frequency damping sensitivity study on surge.
step of H s , the average growth of the surge motion is 63.16%, 58.4%
and 74.11%, corresponding to water depths of 9.41 m, 11.39 m and
13.93 m. Fig. 12(b) reveals that when the T p increases from 5 s to
7 s, the average increase of surge motion is 13.2%, 17.22% and
25.0%, corresponding to water depths of 9.41 m, 11.39 m and
13.93 m. Possible reasons for this phenomenon have been dis-
cussed in Section 6.1.
Influence of water depths on surge motion can also be found in
Fig. 12. It is observed in Fig. 12 that a decrease of the water depth will
result in a significant growth of the surge motion. This indicates that
the surge motion in shallow water will be intensified, which leads to
a larger mooring force. This phenomenon is possibly induced by the
following two reasons. First, the second-order wave force gets larger
as the water depth becomes shallower, and its frequency gets closer
to the resonant frequency of the mooring system. Second, the set-
down becomes obvious as the shallow water depth decreases. It will
stimulate large first-order wave excitation force at low-frequency
area, which increases the surge response of the barge.
6.3. Damping sensitivity
The resonance between the wave induced low-frequency drift
forces and the vessel’s mooring system will lead to large motions
in the horizontal plane, such as surge and sway. At the same time,
an increase of the low-frequency damping will reduce the
motions. In the scale model tests, damping is considered larger
than the full-scale damping due to the difference of Reynolds
number, which causes the experimental results to become non-
conservative. Thus, a sensitivity study on the damping effects for
the surge response is conducted in this section.
Additional dampings are adopted with damping ratios of 2.5%,
5.0% and 7.5% of the surge critical damping. As shown in Fig. 13, the
results indicate that with the increase of the damping ratio, the surge
motion is slightly reduced. It is demonstrated that the damping has
finite effects on the low-frequency motion in the present case. This is
caused by the fact that with the decrease of water depth, the
damping of the low-frequency motion will increase to a certain level,
at which the additional damping will be less effective in the
reduction of the low-frequency motion. Therefore, the damping
differences between the scaled model tests and the prototype will
not bring significant influences to the motion prediction.
7. Conclusions
In this study, both numerical simulations and physical model
tests have been conducted on a moored floating barge in shallow
water. Some conclusions have been drawn and summarized as
follows:
(1) The simulated results agree well with the experimental data,
validating the reliability of the numerical models.
(2) It is observed that the natural periods of the barge in heave,
roll and pitch become larger as water depths decrease,
together with the roll damping levels.
(3) As water depth becomes smaller, the heave motion will
reduce, while the surge motion will increase. This conclusion
can be applied to increase the vessel draft in shallow water.
(4) The low-frequency motions are more sensitive to the shallow
8 water effects than the wave frequency motions. For the low-
frequency motions of the moored floating structure, the
critical value is around γ ¼ 4 for the consideration of the
shallow water effects. While for the heave motions, the critical
value is around γ ¼ 3 for consideration of the shallow water
effects.
Acknowledgement
This work was financially supported by National Natural
Science Foundation of China (Project no. 51239007). These sources
of support are gratefully acknowledged.
References
Bingham, H.B., 2000. A hybrid Boussinesq-panel method for predicting the motion
of a moored ship. Coastal Eng. 40 (1), 21–38.
Chen, X., 1994. Approximation on the quadratic transfer function of low-frequency
loads. In: Proceedings of the 7th BOSS, Boston (USA), vol. 2, pp. 289–302.
Chen, X., 2007. Middle-field formulation for the computation of wave-drift loads.
J. Eng. Math. 59 (1), 61–82.
DNV, 2005. DeepC Theory Manual. Oslo, Norway.
DNV, 2011. Modelling and Analysis of Marine Operations. Tech. Rep. DNV-RP-H103.
Det Norske Veritas.
Grant R., Holboke M., 2004. Shallow water effects on low-frequency wave excita-
tion of moored ships. In: Offshore Technology Conference, Houston, USA, Paper
OTC-16718-MS.
Guillaume, d.H., Flávia, R., Olaf, W., Chen, X., 2012. Review of approximations to
evaluate second-order low-frequency load. In: ASME 2012 31st International
Conference on Ocean, Offshore and Arctic Engineering, Rio de Janeiro, Brazil,
pp. 363–371.
Hong, S., Kim, J., Cho, S., Choi, Y., Kim, Y., 2005. Numerical and experimental study
on hydrodynamic interaction of side-by-side moored multiple vessels. Ocean
Eng. 32 (7), 783–801.
Kim, Y.S., Sung, H.G., Kim, J.H., et al., 2012. An experimental study on the response
of FSRU in shallow water in comparison of mooring systems. In: The 22nd
International Offshore and Polar Engineering Conference, Rhodes, Greece,
pp. 875–880.
Lee, H.W., Lee, D.Y., Kim, B., et al., 2010 A motion analysis of two floaters in shallow
water using Boussinesq equations. In: Proceedings of the 20th International
Offshore and Polar Engineering Conference, Beijing, China, pp. 480–484.
Li, D., Yi, C., Xu, Z., Bai, X., 2014. Scenario research and design of FPSO in South
China Sea. In: Proceedings of the 24th International Ocean and Polar Engineer-
ing Conference, Busan, Korea, vol. 1, pp. 964–969.
Li, X., Yang, J., Xiao, L., 2003. Motion analysis on a large FPSO in shallow water. In:
Proceedings of the 13th International Offshore and Polar Engineering Con-
ference, pp. 235–239.
Molin, B., Fauveau, V., 1984. Effect of wave-directionality on second-order loads
induced by the set-down. Appl. Ocean Res. 6 (2), 66–72.
Naciri, M., Buchner, B., Bunnik, T., Huijsmans, R., Andrews, J., 2004. Low-frequency
motions of LNG carriers moored in shallow water. In: ASME 2004 23rd
International Conference on Offshore Mechanics and Arctic Engineering.
American Society of Mechanical Engineers, BC, Canada, vol. 3, pp. 995–1006.
Naciri, M., Poldervaart, L., 2004. Design aspects of SPM LNG terminals in shallow
water. In: Offshore Technology Conference, Houston, USA, Paper OTC-16715-
MS.
Newman, J.N., 1974. Second-order slowly-varying forces on vessels in irregular
waves. In: Proceedings of the Symposium on the Dynamics of Marine Vehicles
and Structures in Waves, London, pp. 182–186.
Newman, J.N., 2004. Progress in wave load computations on offshore structures. In:
Invited Lecture, 23rd International Conference on Offshore Mechanics and
Arctic Engineering, Vancouver, Canada.
 L. Xiong et al. / Ocean Engineering 97 (2015) 207–217
Pessoa, J., Fonseca, N., 2013. Investigation of depth effects on the wave exciting low-
frequency drift forces by different approximation methods. Appl. Ocean Res. 42
(0), 182–199.
Pinkster, J., 1975. Low-frequency phenomena associated with vessels moored at
sea. Soc. Pet. Eng. J. 15 (06), 487–494.
Pinkster, J., 2009. Wave drift forces in directional seas in shallow water. In: ASME
2009 28th International Conference on Ocean, Offshore and Arctic Engineering.
American Society of Mechanical Engineers, pp. 831–842.
Reinholdtsen, S., Falkenberg, E., 2001. SIMO—Theory/User Manual. MT51 F93-0184,
MARINTEK.
Tahar, A., Kim, M., 2003. Hull/mooring/riser coupled dynamic analysis and
sensitivity study of a tanker-based FPSO. Appl. Ocean Res. 25 (6), 367–382.
Wim, v.d.M., Ivo, W., 2008. Time-domain calculation of moored ship motions in
nonlinear waves. Coastal Eng. 55 (5), 409–422.
Xiao, L., 2007. Research on Shallow Water Waves and Shallow Water Effect on Soft
Yoke Moored FPSO. Ph.D. Thesis. Shanghai Jiao Tong University, Shanghai.
Xiao, L., Tao, L., Yang, J., Li, X., 2014. An experimental investigation on wave run-up
along the broadside of a single point moored FPSO exposed to oblique waves.
Ocean Eng. 88 (0), 81–90.
217
Xiao, L., Yang, J., 2006. Review of the research on FPSO hydrodynamics. Ocean Eng.
24 (4), 116–123.
Yan, S., Ma, Q., Lu, J., Chen, S., 2010. Fully nonlinear analysis on responses of a
moored FPSO to waves in shallow water. In: Proc. ISOE2010, Beijing.
Yang, J., Xiao, L., Peng, T., 2002. Experimental research on motion performance of
large FPSO in shallow water. In: Proc. of the third New S-Tech Conf. Kobe, Japan.
Zhang, L., Lu, H., Yang, J., Peng, T., Xiao, L., 2013. Low-frequency drift forces and
horizontal motions of a moored FPSO in bi-directional swell and wind-sea
offshore West Africa. Ships Offshore Struct. 8 (5), 425–440.
Zhao, W., Yang, J., Hu, Z., Tao, L., 2014. Prediction of hydrodynamic performance of
an FLNG system in side-by-side offloading operation. J. Fluids Struct. 46,
89–110.
Zhao, W., Yang, J., Hu, Z., Wei, Y., 2011. Recent developments on the hydrodynamics
of floating liquid natural gas (FLNG). Ocean Eng. 38 (14), 1555–1567.