Journal of Marine Science and Technology (2024) 29:387–403

https://doi.org/10.1007/s00773-024-00993-6

ORIGINAL ARTICLE

Investigation of the efficiency of wind‑assisted systems using

model‑based design approach
Cem Guzelbulut1 · Tomohiro Sugimoto2 · Yasuaki Fujita3 · Katsuyuki Suzuki1

Received: 31 July 2023 / Accepted: 13 March 2024 / Published online: 5 April 2024
© The Author(s) 2024

Abstract
To achieve the goal of a 50% reduction of ­CO2 emission in the maritime industry by 2050, different systems and solutions
were proposed by researchers. Rigid wind sails, rotor sails, suction wings, and kites were developed to contribute to cleaner
and environment-friendly transportation by reducing total fuel and energy consumption. In the present study, a ship dynam-
ics model of KVLCC2 consisting of hull, rudder, propeller, and sailing system was built considering the effects of wind and
wave. Firstly, the amount of energy consumption reduction of both systems was examined under different wind directions and
wind speeds. It was found that a single sailing system can reduce total energy consumption by up to 10%. Then, the effects
of the ship speed, the position of the sailing system, and the number of sails on the reduction of energy consumption were
examined. It was found that the amount of overall energy reduction reaches around 23% and 16% when the number of sails
was increased to 10 rigid wind sails and 10 rotor sails, respectively. The effects of waves were also investigated, and it was
revealed that wave forces decrease the percent energy reduction more when environmental conditions become more severe,
starting from the Beaufort scale of 7.

Keywords Wind-assisted ship propulsion · Rigid wind sails · Rotor sails · MMG model

List of symbols cL , cD , cP Lift, drag and power coeffi-

AF Frontal projected area of ship cients of sailing systems
above sea level DP Diameter of propeller
AL Lateral projected area of ship Drigid Lift force of rigid wind sail
above sea level Drotor Lift force of rotor sail
AR Rudder area FN Normal force component of
aH Rudder force increase factor rudder force
C1 , C2 Coefficients to define fRigid Function to determine angle
wake coefficients during of attack of rigid wind sail
maneuvering fRotor Function to determine spin
CXA , CYA , CNA Aerodynamic force coef- ratio of rotor sail
ficients for hull-wind f𝛼 Rudder lift gradient
interaction coefficient
CXW , CYW , CNW Force coefficients for hull- H1∕3 Significant wave height
wave interaction IzG Mass moment of inertia
around gravitational center
JP Advance ratio of propeller
* Cem Guzelbulut Jz Added mass moment of
[email protected] inertia
1 k0 , k1 , k2 Coefficients of polynomial to
Department of Systems Innovation, School of Engineering,
The University of Tokyo, Tokyo, Japan describe the J − KT relation
2 KT Thrust characteristics of
Rule Development Department, Class NK, Tokyo, Japan
propeller
3
Design Division, Japan Marine United Corporation,
Yokohama, Japan

Vol.:(0123456789)
388 Journal of Marine Science and Technology (2024) 29:387–403

Lpp Length of ship between XHull−Wave , YHull−Wave , NHull−Wave Force components

perpendiculars related to interaction
Lrigid Lift force of rigid wind sail of hull and wave
Lrotor Lift force of rotor sail XP , YP , NP Force components related to
lR Effective longitudinal posi- propeller
tion of rudder force XR , YR , NR Force components related to
m Mass of the ship rudder
mx , my Added mass in x and y XWAD , YWAD , NWAD Force components related to
directions wind assisting device
nP( Propeller speed Xvv′
, Xvr′ ′
, Xrr′ , Xvvvv Hydrodynamic derivatives in
P ui , uwind,j , 𝜙wind,k Probability of scenario x-direction
)

PPropeller Power consumption of ′

Yv′ , Yr′ , Yvvv ′
, Yvvr ′
, Yvrr ′ Hydrodynamic derivatives in
, Yrrr
propeller y-direction
PWAD Power consumption of rotor Nv′ , Nr′ , Nvvv′ ′
, Nvvr ′
, Nvrr ′ Hydrodynamic derivatives in
, Nrrr
sail yaw direction
r Yaw rate x − y − z Moving coordinate frame at
R′0 Resistance coefficient during midship position
straight moving xH Effective position of hull
SR Spin ratio of rotor sail hydrodynamic forces in lon-
Tfinal Voyage time gitudinal direction
Tref Controller parameter to xG X-coordinate of center of
describe time difference gravity according to moving
between actual ship and coordinate frame of ship
reference ship xP Propeller position in longitu-
Tv Wave period dinal direction
tP Thrust deduction factor xR Rudder position in longitudi-
tR Steering resistance deduction nal direction
factor xref , yref Reference position to define
U Resultant speed heading and speed error
UR Resultant inflow velocity to xSail Sail position in longitudinal
rudder direction
u Forward ship speed x0 − y0 − z0 Global coordinate frame
ue Ship speed error 𝛼A Angle of attack of rigid wind
uR , vR Forward and lateral inflow sail
velocity components of 𝛼R Effective inflow angle to
rudder rudder
VA Apparent wind speed 𝛽 Drift angle
Vtrue True wind speed 𝛽P Geometric inflow angle to
v Lateral velocity at gravita- propeller
tional center 𝛾R Flow straightening
vm Lateral velocity at midship coefficient
position 𝛿 Rudder angle
wP Wake fraction 𝜖 Percent reduction of total
wP0 Wake fraction during straight energy consumption
moving E Ratio of wake fraction at pro-
X, Y, Nm Forces in x and y directions, peller and rudder positions
and moment around midship 𝜂 Ratio of propeller diameter
position to rudder span (= Dp ∕HR )
XH , YH , NH Force components related to 𝜃A Apparent wind direction
hull hydrodynamics 𝜃true Apparent wind direction
XHull−Wind , YHull−Wind , NHull−Wind Force components 𝜅 Experimental factor to define
related to interaction uR
of hull and wind 𝜌 Density of water
Journal of Marine Science and Technology (2024) 29:387–403 389

𝜌air Density of air thrust power generation capacity as a clean source of energy.
𝜓e Heading error Therefore, some ships installing wind-assisted devices such
𝜒0 Relative wave direction as rigid wind sails and rotor sails, as shown in Fig. 1, were
being developed.
Previous studies about sailing systems were mostly focused
1 Introduction on rigid wind sails and rotor sails. Tillig and Ringsberg con-
ducted a comprehensive study to understand how different
Increasing ­CO2 emissions over the years plays an important configurations of rotor sails affect the performance of a RoRo
role in climate change. Recently, each sector has started to and tanker by using a 4 degrees-of-freedom model under
take some actions to reduce ­CO2 emission levels. The ship- steady-state conditions [8]. Thies and Ringsberg investigated
ping industry, whose share is around 3% of global ­CO2 emis- the effects of the Flettner rotor and suction wing of different
sion [1], is one of the main industries looking for a solution sizes on a ferry considering steady-state characteristics [9].
to reduce CO2 emission. Since the International Maritime Fujiwara et al. implemented a sail system on the MMG model
Organization (IMO) declared that the C ­ O2 emission level and investigated the steady sailing performance of a bulk
should be reduced by 50% in 2050 compared to ­CO2 emis- carrier [10]. Vilenova et al. conducted a comparative study
sion levels in 2008 [2], studies related to the decarbonization between rotor sails, rigid wind sails, and DynaRigs based on
of the shipping industry gained momentum. steady-state solutions. It was concluded that rotor sails per-
The aim of reducing C ­ O2 emission level by 50% requires formed best in terms of power saving per unit area [11]. In
many efforts including the design of engines, the type of another study, Lu and Ringsberg compared the Flettner rotor,
fuels, the design of hull and propeller, the optimization of rigid wing sail, and DynaRig. Rotor sail contributed more
route, and introducing innovative technologies. Soft sails, than other sails and further investigations were conducted. It
fixed sails, rotor sails, kites, and wind turbines use wind was revealed that reducing speed is as effective as implement-
energy, which is a renewable and clean energy source, to cre- ing a sailing system and the performance of a sailing system
ate additional thrust forces [3]. In addition, there are some depends on ship type, size, route, and weather conditions
conceptual studies to replace diesel engines with electric [12]. Viola et al. considered implementing a 2-piece wing sail
power systems using solar and wind power [4]. Controlla- on KVLCC2 ship and the effect of the number, aspect ratio
ble pitch propellers are also promising to achieve a reduc- and area of sails, and ship speed on the power reduction was
tion in power consumption. Gypa et al. proposed using a examined [13]. Ouchi et al. conducted CFD (computational
controllable-pitch propeller together with a wind sail system fluid dynamics) and FEM (finite element method) analyses to
to achieve a larger amount of reduction in power consump- understand the aerodynamics of sails and evaluate the struc-
tion [5]. Among the diverse options, wind power is one of tural performance of rigid wind sails [14]. Wang et al. used
the promising solutions due to the availability of strong a grid system to find energy-efficient routes by determining
winds in the open sea, being free to use, and having a high wind sail angle, ship speed, and position in a collaborative way

Fig. 1  Wind assisted ships equipped with rigid wind sails [6], and rotor sails [7]
390 Journal of Marine Science and Technology (2024) 29:387–403

using particle swarm optimization [15]. Although many stud- based on the previous work, conducted by Yasukawa and
ies were conducted to predict the effect of wind sails on per- Yoshimura [16]. Two coordinate systems were defined:
formance, they mostly focused on steady-state solutions. In the global fixed coordinate system ( x0 − y0 − z0 ) and moving
present study, we proposed using a ship dynamics model with ship coordinate system ( x − y − z ) as shown in Fig. 2.
controllers of rudder angle and propeller speed to maintain ( x − y − z ) coordinate system is fixed at mid-ship position,
ship course and speed due to generated thrust and side forces x and y axes are through the bow and starboard of the ship.
by wind sails. Thus, the time responses of each component and The velocity components at the center of mass were denoted
their interactions with each other will be understood clearly. as u and v in forward and starboard directions. Since the
Although different wind sails were proposed and imple- ( x − y − z ) coordinate system is shifted from center of grav-
mented on ships, there is a requirement to understand how a ity to mid-ship ( xG , 0, 0), the velocity component at mid-ship
sailing system affects the ship dynamics and energy consump- position in y-direction, vm , was found as given in Eq. 1. The
tion depending on the design of the ship, the sailing system, resultant
√ speed, U , and drift angle, 𝛽 , were defined by
and environmental factors. The MMG model, which allows U = u2 + v2m and 𝛽 = atan(−vm ∕u).
us to examine ship dynamics by considering all these effects,
forms the basis of this study due to the suitability to investi- vm = v − xG r (1)
gate different ship types, wind, and wave. The MMG model The equation of motion is given in Eq. 2. m and IzG are
was used to investigate the maneuverability of different ships the mass and inertia of the ship around the center of gravity.
with very good accuracy [16]. Each subsystem, such as pro- xG is the position of the center of mass according to the mid-
peller, hull resistance, etc., is modeled and the equation of ship coordinates. mx , my and Jz are added mass and inertia
motion under different conditions can be tested. Yasukawa terms. X , Y , and Nm are the resultant forces and moment in
et al. investigated the maneuverability of KVLCC2 ship under x , y , and r directions.
adverse weather conditions by considering the effects of wind
and waves on the hull [17]. In the present study, the MMG
( ) ( ) 2
(m + mx )u̇ − m( + my v)m r − xG mr = X
model was extended by considering a wind sailing system and (2)
( m + m2y vṁ + )m + mx (ur + xG m)ṙ = Y
controllers for propeller speed, rudder angle, and operational IzG + xG m + Jz ṙ + xG m vṁ + ur = Nm
parameters of the sailing system. Apart from previous studies,
the angle of attack of wind sails, and the spin ratio of rotor sails The resultant forces acting on the ship were obtained by
were optimized at each time step to maximize thrust depending the summation of the hull, propeller, rudder forces and hull-
on environmental conditions. wind interaction, hull-wave interaction, and wind-assisted
In the present study, the performance of wind-assis- system generated force as shown in Eq. 3.
tance systems was investigated using a model-based design
approach. In the proposed method, the MMG model was used
to investigate ship dynamics under the environmental effects
of wind and waves. Each source of force; hull resistance in the
water, propeller, rudder, wind sail, hull-wind interaction, and
hull-wave interaction, were modeled. Controllers for rudder 0
angle and propeller speed were implemented to follow a track
in a given time. Then, propeller power with and without wind
assistive devices and power consumption by wind assistive
devices were obtained. The reduction of total energy consump-
tion under different conditions was investigated. Finally, the
effects of route optimization on the performance of windsails
were discussed.

2 Methods

2.1 Standard MMG model

0
0
The target ship chosen for the implementation of wind assis-
tive is KVLCC2 [18] due to extensive research and availabil-
ity of data about it. Ship dynamics of KVLCC2 were inves-
tigated using 3 degrees-of-freedom: surge, sway, and yaw Fig. 2  Definition of coordinate systems
Journal of Marine Science and Technology (2024) 29:387–403 391

X =XH + XP + XR + XHull−Wind + XHull−Wave + XWAD the rudder and given in Eq. 11. The resultant inflow speed
and angle were defined in Eqs. 12 and 13.
Y =YH + YP + YR + YHull−Wind + YHull−Wave + YWAD (3)
Nm =NH + NP + NR + NHull−Wind + NHull−Wave + NWAD
( )
XR = − 1 − tR FN sin 𝛿
(10)
( )
YR = − 1 + aH FN cos 𝛿
Hydrodynamic forces acting on the ship hull were cal- ( )
culated by using Eq. 4. Here, v′m and r′ are nondimensional NR = − xR + aH xH FN cos 𝛿
speeds and were defined by v�m = vm ∕U and r� = rLpp ∕U .
Nondimensional hydrodynamic resistance on the hull FN = (1∕2)𝜌AR UR2 f𝛼 sin𝛼R (11)
was expressed by using polynomial functions as given by
Eq. 5. Hydrodynamic coefficients for maneuvering simu- √
lation of KVLCC2 were obtained based on the previous UR = u2R + v2R (12)
work by Yasukawa and Yoshimura [16].

XH = (1∕2)𝜌Lpp dU 2 X � 𝛼R = 𝛿 − atan(vR ∕uR ) (13)

H (v�m ,r� )
YH = (1∕2)𝜌Lpp dU Y � � �
2
(4) Longitudinal and lateral velocity components, uR and vR ,
H (vm ,r )
NH = 2
(1∕2)𝜌Lpp dU 2 N � � � around the rudder are required to calculate rudder forces as
H (vm ,r )
shown in Eqs. 14 and 15.
( ′ )2 ( )2 ( ′ )4
√
XH′ = − R′0 + Xvv
′ ′ ′ ′
vm + Xvr vm r + Xrr′ r′ + Xvvvv
′
vm
√ { (√ )}2
(v′m ,r′ )
√
( )√ 8KT
( ′ )3 ( ′ )2 ′ ( )2 ( ′ )3 uR = 𝜀u 1 − wP √ 𝜂 1+𝜅 1+ −1 + (1 + 𝜂)
YH′ = Yv′ v′m + Yr′ r′ + Yvvv
′ ′
vm + Yvvr ′ ′
vm r + Yvrr vm r′ + Yrrr
′
r 𝜋JP2
(v′m ,r′ )
( ′ )3 ( ′ )2 ′ ( )2 ( ′ )3
NH′
(v′m ,r′ )
= Yv′ v′m + Yr′ r′ + Yvvv
′ ′
vm + Yvvr ′ ′
vm r + Yvrr vm r′ + Yrrr
′
r (14)
(5) �
vR = U𝛾R 𝛽 − lR r�
( )
(15)
Propeller generates thrust only in the x direction as
given in Eq. 6 where tP is the thrust deduction factor, nP is Environmental factors, like wind and waves, also affect
the rotational speed of propeller, DP is the diameter of the the motion of ships. Therefore, the interaction between the
propeller, and KT is propeller thrust which depends on the ship and the environment was also considered. The effect
advance ratio of the propeller, JP. of wind on the hull was defined according to Fig. 3 using
the density of air, 𝜌air , frontal and lateral area, AF and AL ,
XP = 1 − tP 𝜌n2P D4P KT(Jp ) apparent wind speed, VA , and force coefficients as given
( )

YP = 0 (6) Eq. 16. Wind force coefficients, CXA , CYA , and CNA , at dif-
ferent apparent wind direction, 𝜃A, were obtained based on
NP = 0
the previous study by Yasukawa et al. [17].
The relation between KT and JP was defined by a sec- XHull−Wind = (1∕2)𝜌air AF VA2 CXA(𝜃A )
ond-order polynomial as shown in Eq. 7. Advance ratio
of the propeller can be found by using Eq. 8. Wake coef- YHull−Wind = (1∕2)𝜌air AL VA2 CYA(𝜃A ) (16)
ficient,wP , depends on maneuvering motion and wake coef- NHull−Wind = (1∕2)𝜌air AL Lpp VA2 CNA(𝜃A )
ficient at straight moving,wP0 , as given in Eq. 9. Geometric
inflow angle,𝛽P , was defined as 𝛽P = 𝛽 − xP� r�.
Wave-induced steady forces were defined using water
KT = k0 + k1 JP + k2 JP2 (7) density, 𝜌, gravitational acceleration, g , significant wave
height, H1∕3 , length of the ship, Lpp , and resistance coeffi-
( )
u 1 − wP cients, CXW , CYW , and CNW as given in Eq. 17. Wave resist-
JP = (8)
nP DP ance coefficients, which depend on the resultant ship speed,
U , wave period, Tv , and relative wave direction, 𝜒0 , were
( ) ( ) { ( )}( ) obtained by previous work of Yasukawa et al. [17].
1 − wP ∕ 1 − wP0 = 1 + 1 − exp −C1 ||𝛽P || C2 − 1
(9) 2
XHull−Wave = 𝜌gH1∕3 Lpp CXW (U,Tv ,𝜒0 )
The definition of rudder forces was given in Eq. 10. The 2
YHull−Wave = 𝜌gH1∕3 Lpp CYW (Tv ,𝜒0 ) (17)
coefficients of tR , aH and xH are related to the interaction
between hull and rudder. FN is the normal force created by
2 2
NHull−Wave = 𝜌gH1∕3 Lpp CNW (Tv ,𝜒0 )
392 Journal of Marine Science and Technology (2024) 29:387–403

Table 1  Ship particular, hydrodynamic coefficients, propeller and

rudder characteristics
Parameter Value Parameter Value Parameter Value

Lpp (m) 320 ′

Xvv – 0.04 aH 0.312
B(m) 58 ′
Xvr 0.002 xR′ – 0.5
d(m) 20.8 Xrr′ 0.011 tR 0.387
312,600 ′
0.771 ′ – 0.464
( )
Δ m3 Xvvvv xH
CB 0.81 Yr′ 0.083 xP′ – 0.48
11.2 Yv′ – 0.315 112.5
( )
xG (m) AR m 2
m′x 0.022 ′
Yvvv – 1.607 HR (m) 15.8
m′y 0.223 ′
0.379 0.6241
( )
Yvvr 𝜂 DP ∕HR
′
Jzz 0.011 ′
Yvrr – 0.391 f𝛼 2.747
Dp (m) 9.86 ′
Yrrr 0.008 𝜅 0.5
tP 0.22 Nr′ – 0.049 E 1.09
k0 0.2931 Nv ′ – 0.137 𝛾R (𝛽R < 0) 0.395
k1 – 0.1856 ′
Nvvv – 0.03 𝛾R (𝛽R > 0) 0.64
Fig. 3  Definition of true and apparent wind speed directions k2 – 0.2045 ′
Nvvr – 0.294 lR′ – 0.71
0.35 ′
0.055 1100
( )
wP0 Nvrr AF m 2
Principal particulars, hydrodynamic coefficients, interac- C1 2 ′
Nrrr – 0.013 AL (m2 ) 3000
tion coefficients, propeller, and rudder characteristics used C2 (𝛽P < 0) 1.1
in the simulation were given in Table 1. C2 (𝛽P > 0) 1.6

2.2 Characteristics of wind sail systems obtained using Eq. 20 where xSail is the position of the
wind-assisted device along the x-axis of the ship coordinate
In the present study, a rigid wind sail and a rotor sail were system.
considered to reduce energy consumption. Rigid wind
sails produce lift and drag forces by adjusting the angle of Lrigid = (1∕2)𝜌air VA2 AcL (𝛼A )
attack to maximize thrust generated by sail systems. The Drigid = (1∕2)𝜌air VA2 AcD (𝛼A ) (18)
airfoil shape considered in this study is NACA0015. Lift
and drag characteristics of NACA0015 airfoil between 0
and 90 degree was used for Reynolds number of 10 million Lrotor = (1∕2)𝜌air VA2 AcL(SR)
Drotor = (1∕2)𝜌air VA2 AcD(SR) (19)
[19]. The height and width of the wind-assisted device were
assumed to be 50 m and 30 m.
Rotor sails are another type of wind assistance propulsion
( ) ( )
XWAD = D cos 𝜃A − L sin 𝜃A
system investigated in the present study. Unlike rigid wind
(20)
( ) ( )
YWAD = D sin 𝜃A + L cos 𝜃A
sails, rotor sails consume additional power through rotating.
Lift and drag characteristics of the rotor sails depend on the NWAD = YWAD × xSail
ratio of the rotational speed of the rotor sail and apparent
wind speed. The generated forces and power consumption
by rotor sails were obtained based on the results in [8]. Since 2.3 Controllers
coefficients were valid for rotor sails having an aspect ratio
of 6 and disc diameter twice the diameter of the rotor, the To investigate the performance of wind assistive systems at
dimensions of rotor sails were determined as a diameter of different sea and weather conditions, rudder angle, propeller
5 m and a height of 30 m. The lift and drag coefficients speed, angle of attack of the rigid wind sail, and spin ratio of
used in simulations for rigid wind sails and rotor sails were the rotor sail need to be controlled. In the present study, we
shown in Fig. 4. assumed that the planned ship trajectory is straight, and wind
Lift and drag forces generated by different sail systems and wave are subject to change. Thus, the same trajectory is
were calculated as given in Eqs. 18 and 19 where lift and followed under different conditions of wind, wave, and cruis-
drag coefficients are cL and cD , apparent wind speed and ing speed of the ship.
angle are VA and 𝜃A, angle of attack is 𝛼A, and the spin ratio Reference position and heading of the ship were defined
is SR . Then, the forces and moment acting on the ship were as the position and heading of the ship after Tref seconds later,
Journal of Marine Science and Technology (2024) 29:387–403 393

2
Rigid Wind Sail
15
Rotor Sail of attack was determined by maximizing the fRigid function,
which is given in Eq. 23, in each time step. When it comes to
1
10
rotor sails, they require external power input to rotate the sail
at a certain rotational speed. Therefore, the spin ratio of rotor
Value

Value
0

5
sails was determined between ± 5 SR to maximize average
-1 power reduction, fRotor , which is given in Eq. 24. Average
-2 0
power reduction was predicted as the difference between the
-100 -50 0 50 100 0 2 4 6 approximate reduction of power created by rotor sail and
Angle of Attack (deg) Spin Ratio
power consumption of rotor sail, PWAD . The approximate
Drag Coefficient Lift Coefficient
reduction of power was assumed to be the multiplication of
thrust force generated, XWAD, and surge speed, u. It should
Fig. 4  Lift and drag coefficients for rigid wind sails [19] and rotor
sails [8] be noted that functions to determine operational condition of
rigid wind sails and rotor sails, fRigid and fRotor, have different
dimensions, because rotor sails should be operated consider-
ing power required to rotate rotor sails and generated thrust
power, but rigid wind sails should be operated considering
thrust force and resultant force.

(23)
( 2 2
)
fRigid = XWAD × XWAD + YWAD

Fig. 5  The definition of heading and speed error

fRotor = XWAD × u − PWAD (24)

which is acquired based on the route information as shown 2.4 MATLAB/Simulink integration

in Fig. 5. The definition of ship heading error and ship speed
error was given in Eqs. 21 and 22. After calculating the head- Wind-assisted ship dynamics model was built using MAT-
ing error, a proportional controller was used to determine the LAB R2022b/Simulink as shown in Fig. 6. Each source of
rudder angle. force, propeller, rudder, hull resistance, added mass and
The propeller speed rate was also controlled to consider the inertia, wind sails, hull-wind interaction, and hull-wave
effect of different sails under different environments. When interaction, was considered separately. Control models for
external forces acting on a ship change due to wind and waves, propeller speed and rudder angle were considered autopilot.
propeller speed should be controlled to reach the destination Then, wind sail controllers were designed to determine the
point at the desired time. The reference ship speed was calcu- angle of attack and the spin ratio.
lated at each time step and PID (proportional-integral-deriva-
tive) controller was designed to determine the propeller speed 2.5 Performance investigation index
rate. Then, propeller speed was obtained by integrating the
propeller speed rate. Through control of the rudder and pro- The performance of different sailing systems was pre-
peller, the same conditions were obtained for ship with and dicted by considering many parameters including envi-
without wind-assistive systems. ronmental conditions during the route. To evaluate the
( ) effect of the sailing system, the target ship followed a
y −y
𝜓e = 𝜓actual − atan ref (21) straight line under different environmental conditions by
xref − x using a controller. A scenario database, consisting of dif-
ferent environmental conditions and the probability of
each, was created. The overall performance function was
√( )2 ( )2
yref − y + xref − x
ue = −u (22) evaluated based on the total energy consumption and the
Tref probabilities as given in Eq. 25 where P is the probability
which depends on ship speed, ui , true wind speed, Vtrue,j,
Wind assistance systems also require a controller to maxi-
and true wind direction, 𝜃true,k . To compare ship with and
mize the thrust under different external conditions. Since
without wind sailing system, energy consumptions of
rigid wind sails require power mainly to adjust angle of
both ships were calculated considering energy consump-
attack, operating steady power of rigid wind sails is neg-
tion of propeller and rotor sail between two points along
ligible. Angle of attack of rigid wind sails was determined
straight route considering same true wind speed, true wind
to maximize thrust between −90 and 90 degrees. The angle
394 Journal of Marine Science and Technology (2024) 29:387–403

direction, and ship speed. Then, 𝜖i,j,k is calculated as the

∑∑∑ ( )
fWAD = P ui , Vtrue,j , 𝜃true,k × 𝜖i,j,k (25)
percent reduction in the total energy consumption to main- i j k

tain the ship speed constant.

vm
Xadd
r
Added MassX
u
Propeller vm XR

u XP r
1-wp 1-wP
v JP JP YR
Summation Equation of Motion
PropPow

r Beta Beta
KT KT 1 u
u

Pos_X PropRot n U U NR 2 1 X
Pos_Y RudderAngle 3 vm

Rudder Angle 4
heading Rudder r
5
u 2 Y
Autopilot r
Yadd 6 pos_x
7
Added MassY 8 pos_y
U XH 9 3 N
10 pos_alpha

11
12
13
14
15
16
vm YH

r NH

Hull Resistance
u
Nadd
r
Added MassN

++ ++ ++
u

heading

X
wind dir dir WindDir

wind speed speed WindSpeed

N

Hull-Wind
heading

heading
wave dir dir dir Relative Wave Dir X
wave dir

Wave Height
Y
U

T N

Hull-Wave

WindSpeed XWAD
AoA
Wind Dir
dir YWAD
AoA
OnOff On/Off NWAD
Wind Sail
Sail Controller

Fig. 6  An integrative model of each subsystem using Simulink

Journal of Marine Science and Technology (2024) 29:387–403 395

2.6 Route optimization were given in Figs. 9 and 10. The initial propeller speed
was specified considering the case without wind, wave,
The performance of wind sails depends on environmental and sailing system. When the wind comes at 90 degrees,
conditions significantly and this makes which direction to the required thrust by the propeller decreases and propeller
follow to harvest the maximum amount of power from the speed reduces slightly. The existence of a rigid wind sail or
wind quite important. An optimization problem was defined rotor sail creates propulsive power and reduces the thrust
to understand how route optimization and simultaneous con- demand of the propeller. On the other hand, wind generates
trol of sailing system control variables (angle of attack or forces also along sway direction, and lateral forces are larger
spin ratio) affect power consumption. Route control points, if wind assistance systems were used. The rudder angle is
xi , were defined as deviations from the straight path from controlled to overcome lateral force. Thus, it is ensured that
the start point to the end point of the route as shown in the ship follows a straight course at a constant speed. During
Fig. 7. The length of the route was assumed to be ­105 m and the course of the ship, rigid wind sails and rotor sails adjust
the upper and lower bounds of design variables, xi , were the angle of attack and spin ratio to maximize the thrust.
assumed to be ± ­105 m. Wind and wave conditions were To understand how wind sails behave at different true
assumed to be constant during simulations. The formulation wind speeds and directions, the percent reduction of energy
of the optimization problem was given in Eq. 26 consumption was obtained for ship speed of 7 m/s and true
wind speeds of 5, 10, and 15 m/s between 0 and 360 degrees
Tfinal
min f(x) = ∫
( )
PPropeller(t,x) + PWAD(t,x) dt as given in Fig. 11. Whereas rigid wind sail gives maximum
𝐱 0 (26) reduction around true wind direction of 80 degrees, rotor
5 5
subject to − 10 ≤ xi ≤ 10 where i = 1, 2, 3 sail gives maximum reduction around true wind direction of
110 degrees. Although the rotor sail has one-tenth projected
area of the rigid wind sail, the amount of energy consump-
tion reduction was similar to that of rigid wind sails. On the
other hand, the rigid wind sail system provided a reduction
3 Results in energy consumption for a larger range of wind direction
compared to the rotor sail. Rigid wind sails used backwind
In the present study, the dynamics of wind-assisted ships as thrust power in a more efficient way than rotor sails due
were investigated by using a system modeling approach in to having a larger surface area.
MATLAB R2022b/Simulink. First, the controller parameter, The position of the wind-assisted systems on ships may
Tref , and coefficients of PID controllers were tuned to ensure also affect the reduction of propeller power consumption.
that the ship follows the given trajectory. Different Tref val- For this purpose, the performance of wind-assisted ships
ues were tested on the given route and trajectory following where rigid wind sails were mounted at front ( xSail �
= 0.4),
performances were investigated as given in Fig. 8. When Tref middle ( xSail = 0), and back ( xSail = −0.4) were investigated.
� �
is low, the ship leaves the route for some time and balances The percent reduction of energy consumption was calculated
itself again. On the other hand, high Tref caused the ship to for the true wind speed of 10 m/s and wind angles between
have a larger error around control points. It was found that 0 and 360 degrees as given in Fig. 12. It was that the posi-
tracking performance was the best at Tref = 100s. tion of sailing systems did not affect the amount of energy
Then, each component that generates force was consid- consumption significantly in any wind direction.
ered separately, and interactions between environmental The amount of force generated by sailing systems
effects and each component were examined. First, a constant depends on the apparent wind speed and apparent wind
wind condition, so that true wind speed is 15 m/s and true direction. Ship speed is therefore important as much as true
wind direction is 90 degrees, was created for the ship trave- wind speed and direction. The percent reduction of energy
ling along a straight line at 7 m/s. Responses of propeller consumption was obtained for ship speeds of 5 and 7 m/s as
speed, rudder angle, ship speed, and wind assistance systems given in Fig. 13. It can be clearly said that speed reduction

Fig. 7  Design variables of the

route optimization problem
396 Journal of Marine Science and Technology (2024) 29:387–403

Fig. 8  The effect of different

Tref values on the trajectory fol-
lowing performance

96 0 96 0
Propeller Speed (rpm)

Propeller Speed (rpm)

Rudder Angle (deg)

Rudder Angle (deg)

without WAD without WAD
with WAD with WAD
94 -0.5 94 -0.5

92 -1 92 -1

90 -1.5 90 -1.5
0 100 200 300 400 0 100 200 300 400 0 100 200 300 400 0 100 200 300 400
Time (s) Time (s) Time (s) Time (s)
7.01 23.5 7.01 3.988
Angle of Attack (deg)
Ship Speed (m/s)

Ship Speed (m/s)

23
7.005 7.005 3.986

Spin Ratio
22.5
7 7 3.984
22
6.995 6.995 3.982
21.5

6.99 21 6.99 3.98

0 100 200 300 400 0 100 200 300 400 0 100 200 300 400 0 100 200 300 400
Time (s) Time (s) Time (s) Time (s)

Fig. 9  Propeller, rudder, ship speed, and wind sail response at the Fig. 10  Propeller, rudder, ship speed, and wind sail response at the
true wind speed of 15 m/s and true wind direction of 90 degrees with true wind speed of 15 m/s and true wind direction of 90 degrees with
and without rigid wind sail and without rotor sail

significantly reduces total energy consumption. The effect

of speed reduction from 7 m/s to 5 m/s was almost the same be quite similar in both cases. However, the effect of wave
as adding one more sailing. increases at higher Beaufort scale values.
In addition to the wind, waves also have a significant The overall performance of wind-assisted ships was pre-
effect on the performance of ship propulsion. Therefore, the dicted based on probabilistic way. The scenario database
percent reduction in energy consumption including both contains various ship speeds, true wind speeds, and direc-
wave and wind effects was compared to performance evalu- tions as given in Table 3. The probability of true wind speed,
ation considering only the effect of wind. True wind speed, true wind direction, and ship speed for the ship route was
significant wave height, and wave period were determined assumed as given in Fig. 15.
based on the Beaufort scale as shown in Table 2. Wind and The overall reductions of energy consumption generated
wave direction was assumed to be the same, and the propel- by a single rigid wind sail and rotor sail were found as 3.07%
ler power was obtained between 0 and 360 degrees of true and 1.97% respectively, based on the probability given in
wind direction. The percent reduction in energy consump- Fig. 15. To achieve a higher reduction in energy consump-
tion considering both wind and wave effects and considering tion, the number of sails can be increased. For this purpose,
only wind effect were given in Fig. 14. According to Fig. 14, the overall reduction of energy consumption was obtained
the percent reduction of energy consumption was found to for different numbers of sails as shown in Fig. 16. It can
Journal of Marine Science and Technology (2024) 29:387–403 397

Fig. 11  The percent reduc- Percent Reduction of Energy Consumption Percent Reduction of Energy Consumption
tion of energy consumption at for Rigid Wind Sail for Rotor Sail
different true wind speeds and 0° 0°
directions 30° 10 330° 30° 10 330°
8 8
60° 6 300° 60° 6 300°
4 4
2 2
90° 0 270° 90° 0 270°

120° 240° 120° 240°

150° 210° 5 m/s 150° 210°

180° 10 m/s 180°
15 m/s

Rigid Wind Sail, Area=1500 m

2 be said that the reduction of energy consumption can reach
Absolute Wind Speed=10 m/s 20% on average when the number of sails were increased to
0° 10. It should be noted that the relation between the number
30° 5 330° of sails and percent reduction in energy consumption is not
4 linearly increasing.
60° 3 300° In addition to the investigation of sailing performance
2 during a straight path, it is also important to predict how
1
routing can improve the reduction of energy consumption
further. Although sail systems reduce energy consumption in
90° 0 270°
general, total energy consumption may increase depending
on the environmental conditions. For example, when a ship
encounters an upwind condition, sails cannot generate any
120° 240° thrust force and causes additional resistive forces. For this
purpose, a scenario having an upwind condition at 20 m/s
150° 210° was created to find optimal routing for rigid wind sail and
180° rotor sail considering the ship dynamics including resist-
x' =-0.4 x' =0 x' =0.4 ance during maneuvering. Ship speed was fixed at 6 m/s,
Sail Sail Sail
and 4 rotor sails (height of 30 m and diameter of 5 m) and
4 rigid wind sails (height of 50 m and width of 30 m) were
Fig. 12  The effect of positions of rigid wind sails on the performance

Fig. 13  The effect of ship speed Rigid Wind Sail, Area = 1500 m
2
Rotor Sail, Area = 150 m
2

on the percent reduction of Absolute Wind Speed = 10 m/s Absolute Wind Speed = 10 m/s
energy consumption 0° 0°
30° 10 330° 30° 10 330°
8 8

60° 6 300° 60° 6 300°

4 4

2 2

90° 0 270° 90° 0 270°

120° 240° 120° 240°

150° 210° 150° 210°

5 m/s
180° 7 m/s
180°
398 Journal of Marine Science and Technology (2024) 29:387–403

Table 2  Wind speed, wave height, and wave period at Beaufort scales modeling approach to predict the performance of different
of 4, 5, and 6 sail systems. Through a system model of ship dynamics
Beaufort scale Wind speed Significant wave Wave in 3 degrees-of-freedom, the interactions between propel-
(m/s) height (m) period ler, rudder, hull dynamics, and wind-assisted device were
(s) investigated.
5 9.8 2 5.5 Wind-assisted ship system was modeled based on model-
6 12.6 3 6.7 based design approach, as shown in Fig. 6. First, each com-
7 15.7 4 7.7 ponent of the system, such as propeller, sailing system, etc.
was created and combined to investigate the overall ship
system performance. Then, rudder and propeller speed rate
considered as wind assistance systems. Then, optimal routes controllers were designed to follow given route. In addition
were found for both sailing systems as shown in Fig. 17. to motion controllers, sailing system controllers were cre-
Ship which is equipped rigid sails deviated from straight ated to provide the maximum performance from sails. The
route more than the ship having rotor sails. Compared to a reason why model-based design approach was adopted is to
ship without any sailing system, rotor sails and rigid wind consider the interaction between components and obtain-
sails increased the total energy consumption by 6.8% and ing overall system response depending on the component
0.6% under upwind condition. When the route was opti- design parameters. In addition, the model-based design of
mized, the performances of both wind sails were improved. wind-assisted ship system forms a general design platform
The amount of increase in energy consumption was reduced for future studies to optimize the performance of subsystems
from 6.8% to 3.1% for a ship having rotor sails under upwind and overall ship system through combining high fidelity sub-
condition whereas the ship having rigid wing sails caused system models and overall ship system model.
a 15.9% decrease in energy consumption compared to the In the present study, the performance of wind-assisted
ship having no sails. Although a significant amount of power ships was investigated using time domain solutions unlike
reduction was achieved in the case of rigid wind sails, the previous studies which focus on solving equilibrium equa-
voyage time was increased significantly due to increasing tions [10, 20]. The main reason why time domain solutions
traveled distance. During the simulations, the angle of attack were chosen is that subsystems and controllers interact
and the spin ratio of sailing systems were optimized at each
time step to maximize the thrust generation. Table 3  The scenarios that consider different ship speeds, wind
speeds and wind directions
Parameter Range
4 Discussion
Ship speed (m/s) 5, 6, 7, 8
Wind sails are one of the promising solutions to reduce True wind speed (m/s) 0, 3, 6, …, 18, 21
the ­CO2 emission of ships. In the present study, the perfor- True wind direction (deg) 0, 10, 20, …, 350, 360
mances of rigid wind sails and rotor sails were investigated
under different environmental conditions. Each compo-
nent that generates forces was modeled based on a system
Beaufort-5 Scale Beaufort-6 Scale Beaufort-7 Scale
0° 0° 0°
30° 330° 30° 330° 30° 330°
10 10 10
60° 300° 60° 300° 60° 300°
5 5 5

90° 0 270° 90° 0 270° 90° 0 270°

120° 240° 120° 240° 120° 240°

150° 210° 150° 210° 150° 210°

180° 180° 180°

Wind Wind+Wave

Fig. 14  The effect of considering forces related to wave at different Beaufort scales
Journal of Marine Science and Technology (2024) 29:387–403 399

0.25 0.06
0.3
0.05
0.2
0.25
0.04
Probability

Probability

Probability
0.15 0.2
0.03
0.15
0.1
0.02 0.1
0.05
0.01 0.05

0 0 0
0 3 6 9 12 15 18 21 0 50 100 150 5 6 7 8
Wind Speed (m/s) Wind Direction (deg) Ship Speed (m/s)

Fig. 15  The probability of wind speed, wind direction, and ship speed

dynamically with each other under disturbance of wind and the reference ship and the actual ship. Whereas the rudder
wave. Time domain analysis of ship system provides a tran- angle controller adjusted the position of the ship to follow
sient response of each subsystem, allows to build controllers the path under the external effects of wind and waves acting
to follow given trajectory which is necessary to conduct a on the ship and sailing system, the propeller speed controller
fair comparison of different sailing systems, and gives more adjusted the thrust force required to keep ship speed con-
accurate results by including acceleration terms especially stant as shown in Figs. 9 and 10. Thus, the conditions with
during route optimization. Besides, the fluctuation of wind and without wind assistance were compared fairly and the
and wave, which will be considered in future studies, leads reduction in propeller power generated by wind assistance
to fluctuations in other subsystems and it makes time domain was obtained.
solutions necessary to evaluate the performance of ship for The performances of rigid wind sail and rotor sail showed
real case conditions. different characteristics depending on the true wind angle
To conduct a fair comparison among wind sail systems, according to Fig. 11. The maximum percent reduction in
controllers were designed to follow straight route under dif- energy consumption was achieved at around 80 degrees in
ferent environmental conditions. The value of Tref was used rigid sail and 100 degrees in rotor sail. The force genera-
to determine the reference position and it affected the per- tion capacity of sailing systems depends on apparent wind
formance of tracking significantly as shown in Fig. 8. Low speed and lift/drag coefficients. Although the apparent wind
Tref values make speed error more sensitive and cause insta- speed is lower around 100 degrees, rotor sails produce larger
bility. On the other hand, high Tref values resulted in large thrust forces due to having a larger range of lift coefficient
errors around the corners due to a larger distance between depending on the spin ratio. Thus, rotor sails can generate
maximum thrust force around 100 degrees by increasing the
25 lift coefficient significantly although apparent wind speed
drops. On the other hand, rigid wind sails have a smaller
Energy Consumption (%)

20 Rigid Wind Sail range of lift coefficient, and the maximum percent reduc-
Percent Reduction of

Rotor Sail
tion was obtained where the apparent wind speed is higher.
15 Another different behavior was observed when the wind is
coming from the back. Due to having a larger surface area,
10
rigid wind sails produce larger thrust forces. Moreover, the
disadvantage of having a larger surface area under opposite
wind condition can be reduced using a retractable design.
5
The position where the wind-assisted device was mounted
was found to have a negligible effect on the reduction of
0 energy consumption according to Fig. 12. When wind condi-
1 3 5 10
tions were considered the same over the ship, mounting sails
Number of Sails
away from the mid-ship causes an additional yaw moment
and rudder angle is adjusted to follow the route. Therefore,
Fig. 16  The effect of increasing the number of sails on the reduction
of energy consumption
400 Journal of Marine Science and Technology (2024) 29:387–403

104

0
Position (m)

-2

-4

0 1 2 3 4 5 6 7 8 9 10
Position (m) 104
Rigid Wind Sail Rotor Sail w/o Route Opt.

(a)
7 Rigid Wind Sail 7 Rotor Sail
10 10
15 7
w/o Route Opt. w/o Route Opt.
w/ Route Opt. w/ Route Opt.
6
10
Power (W)

Power (W) 5

5
4

0 3
0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2
Time (s) 4 Time (s) 4
10 10
(b) (c)

Rigid Wind Sail Rotor Sail

30

20 0.4
Angle of Attack (deg)

10 0.2
Spin Ratio

0 0

-10 -0.2

-20 -0.4

-30 -0.6

-40 -0.8
0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2
Time (s) 104 Time (s) 104
(d) (e)

Fig. 17  a Optimal routes and b, c power consumption of ship with rigid wind sail and rotor sail. d, e variation of the angle of attack and the spin
ratio through route
Journal of Marine Science and Technology (2024) 29:387–403 401

the resistive forces generated by the rudder change, and only of ship and sailing system should be considered together to
small variations in energy consumption occur. use wind power more efficiently.
In addition to sailing system parameters, ship speed The effect of increasing the number of sails was handled
also significantly affects the energy consumption of ships. by increasing the wing area without considering the interac-
According to Fig. 13, the amount of reduction in energy con- tion between sails. According to Fig. 16, when the number
sumption significantly increases when ship speed is reduced of sails increased, the percent reduction of energy consump-
from 7 m/s to 5 m/s. Because, when ship speed was reduced, tion also increased but at a lower rate than the number of
the amount of required thrust dropped, and propeller power sails due to propeller characteristics and nonlinear hydro-
diminished. Besides, the ratio of thrust obtained by wind dynamic resistance. It was found that the amount of energy
sails to the total required thrust increased. Thus, propeller consumed by propeller reduction reaches 22.8% and 16.03%
speed and energy consumption were reduced significantly. for 10 rigid wind sails and rotor sails, respectively. However,
The effect of waves on the performance of wind-assisted there are geometrical limits and larger interaction effects to
ships was investigated by comparing the percent reduction in build many rigid wind sails or rotor sails. It can be said that
energy consumption of propeller. It appeared that the effect there is a need to optimize the shape of wind sails or rotor
of considering forces related to waves was not significant sails to achieve a higher reduction in energy consumption,
for environmental conditions at lower Beaufort scales. How- especially for larger ships.
ever, it becomes clearer starting from the Beaufort-7 scale The average performances of wind sails were predicted
according to Fig. 14. If the existence of strong wind and by assuming different scenarios and the probability of each.
higher waves shows a correlation, it is necessary to consider The change in the route affects the probability of each sce-
the wave effect through the optimization process of route of nario and the performance of wind sails can be further
wind-assisted ships. Also, since the effect of waves increases improved. Unlike conventional ships, wind-assisted ships
significantly at higher Beaufort scales, strong winds are not should adjust their route where higher wind power can
desirable always and there will be an optimal wind speed be harvested. Therefore, it is necessary to consider route
that different sailing systems perform best. In the present optimization conducted for specific to wind-assisted ships.
study, the directions of wind and wave were assumed to be Route optimization improved the performance of sailing
the same, but there might be some differences in fact. systems under upwind conditions, according to Fig. 17.
According to Fig. 16, the reductions in energy con- The ship having rigid wind sails showed a large deviation
sumption of propeller of ships equipped with single rigid from a straight path to minimize the energy consumption
wind sail and rotor sail were found as 3.07% and 1.97%, and voyage time increased. On the other hand, ships having
respectively. Compared to the previous studies, the amount rotor sails showed less deviation. Considering the rotor sails
of reduction in energy consumption was lower due to the performance in head wind conditions, the result is seemingly
ratio of ship size and sail size. Ouchi et al. investigated the contradictory because ships having rotor sails should devi-
effect of using 9 rigid wind sails, each having a sail area of ate much larger than ships with rigid wind sails to reduce
1000 ­m2, on a bulk carrier with 180,000 DWT and reached power. However, the obtained result oppositely showed less
22% power reduction on Great Circle when ship speed is deviation through the integral evaluation of the voyage time
constant [14]. Kramer and Steen investigated different and the total energy consumption. In both sailing systems, a
design and control alternatives for wind-assisted ship hav- lot of effort is required to understand the potential of route
ing 5000 DWT and fuel saving was predicted between 17 optimization under different conditions. The scope of the
and 65% with varying speeds between 8 and 12 knots and study will be extended to route optimization to harvest the
varying wind sail area between 784 m ­ 2 and 3920 ­m2 [19]. wind power efficiently depending on sailing systems and
Viola et al. considered implementing a multi-element rigid different weather conditions.
wind sail on KVLCC2M ship and the effects of the num- In the present study, different aspects of rigid wind sails
ber of sails, ship speed, and wind speed were considered. and rotor sails were investigated. However, the performance
It was concluded that an approximately 10% reduction in of sailing systems also depends on the ship. Through updat-
propeller thrust was achieved at a slower speed and higher ing ship and sail parameters, the performance of different
sail area [13]. Tilling and Ringsberg found that six rotor sailing systems on different ships can be obtained to under-
sails can provide up to 30% reduction for a tanker having stand how much a sailing system can contribute to reduc-
50,600 DWT on the Pacific Ocean route [8]. It can be con- tion in energy consumption. Moreover, the performance of
cluded that the amount of reduction in energy consumption sailing systems can be further increased by optimizing the
depends on many factors based on the results obtained in the route of ship and the design of the sailing system. Besides,
present study and previous works. In addition to the design there are some limitations of the existing model. For exam-
and control of wind sails, the size and speed of ship affect ple, when a tower-like sailing system was mounted on the
the reduction of energy consumption. Therefore, the design ship, forces in surge and sway direction will generate a
402 Journal of Marine Science and Technology (2024) 29:387–403

considerable amount of moment in roll and pitch direction. the article’s Creative Commons licence and your intended use is not
In future studies, the existing model should be extended to permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a
six degrees-of-freedom model to consider roll, pitch, and copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
heave motions as well. Also, the inertial effects of sailing
systems and interaction between sails were not taken into
account.
References
1. International Maritime Organization (2009) Second IMO GHG
5 Conclusion study 2009. CPI books. [Online], https://​w ww.​i mo.​o rg/​e n/​
OurWo​rk/​Envir​onment/​Pages/​Second-​IMO-​GHG-​Study-​2009.​
aspx. Accessed 27 June 2023
The present study investigates the effect of wind sail systems 2. International Maritime Organization (2018) Initial imo strategy
on ship dynamics and propeller power consumption using on reduction of GHG emissions from ships. [Online], https://​
the model-based design approach. The interaction between www.​i mo.​o rg/​e n/​O urWo​r k/​E nvir​o nment/​Pages/​Vision-​a nd-​
each subsystem was considered, and controllers for propeller level-​of-​ambit​ion-​of-​t he-​Initi​al-​IMO-​Strat​egy.​aspx. Accessed
27 June 2023
speed and rudder angle were designed to compare wind- 3. International Renewable Energy Agency (2015) Renewable
assisted ships and conventional ships. We have conducted energy options for shipping. [Online] https://​www.​irena.​org/-/​
several simulations to understand the effect of wind speed, media/​Files/​IRENA/​Agency/​Publi​cation/​2015/​IRENA_​Tech_​
wind angle, wave, ship speed, and the type and position Brief_​RE_​for-​Shipp​ing_​2015.​pdf. Accessed 27 June 2023
4. Julià Lluis E, Tillig F, Ringsberg J (2020) Concept design and
of wind sails. It was found that rigid wind sails and rotor performance evaluation of a fossil-free operated cargo ship with
sails have different characteristics at different wind angles. unlimited range. Sustainability 12(16):1–23. https://​doi.​org/​10.​
Both sailing systems generated a higher reduction in energy 3390/​su121​66609
consumption of propeller when wind speed increased. On 5. Gypa I, Jansson M, Gustafsson R, Werner S, Bensow R (2023)
Controllable-pitch propeller design process for a wind-powered
the other hand, the existence of strong wind shows some car-carrier optimising for total energy consumption. Ocean Eng
correlation with higher waves, which increases energy con- 269:113426. https://​doi.​org/​10.​1016/j.​ocean​eng.​2022.​113426
sumption. Therefore, it is quite necessary to conduct route 6. Delivery of 1st Wind Challenger|Case Study|Mitsui O.S.K.
optimization specific to wind-assisted ships to determine Lines Service Site. [Online], https://​www.​mol-​servi​ce.​com/​en/​
case/​windc​halle​nger01. Accessed 1 Nov 2023
how the ship route changes depending on the conditions 7. Bulk carriers—Norsepower|Rotor Sails|Wind Propulsion.
of wind and waves. The overall performances of using a [Online], https://​www.​norse​power.​com/​bulk-​carri​er/. Accessed
single rigid sail and single rotor sail were found as 3.07% 1 Nov 2023
and 1.97% based on the different scenarios and their prob- 8. Tillig F, Ringsberg J (2020) Design, operation and analysis of
wind-assisted cargo ships. Ocean Eng 211(1):1–23. https://​doi.​
abilities. When the number of sails was increased to 10, the org/​10.​1016/j.​ocean​eng.​2020.​107603
reduction in energy consumption of propeller was 22.8% and 9. Thies F, Ringsberg JW (2023) Retrofitting WASP to a RoPax
16.03% for rigid wind sails and rotor sails, respectively. It vessel—design performance and uncertainties. Energies 16:673.
was concluded that the optimization of the shape of the sail https://​doi.​org/​10.​3390/​en160​20673
10. Fujiwara T, Hearn G, Kitamura F, Ueno M, Minami Y (2005)
system is necessary to achieve higher reduction of energy Steady sailing performance of a hybrid-sail assisted bulk car-
consumed by propeller. rier. J Mar Sci Technol 10:131–146. https://​doi.​org/​10.​1007/​
s00773-​004-​0189-3
Acknowledgements We would like to thank Maritime and Ocean Digi- 11. Reche-Vilanova M, Hansen H, Bingham HB (2021) Perfor-
tal Engineering (MODE) Laboratory members at the University of mance prediction program for wind-assisted cargo ships. J Sail-
Tokyo for their research advice. ing Technol 6(01):91–117. https://​doi.​org/​10.​5957/​jst/​2021.6.​1.​
91
Funding Open Access funding provided by The University of Tokyo. 12. Lu R, Ringsberg JW (2019) Ship energy performance study of
This research was supported by REDAS: Fundamental Research Devel- three wind-assisted ship propulsion technologies including a
oping Association for Shipbuilding and Offshore, the Shipbuilders’ parametric study of the Flettner rotor technology. Ships Off-
Association of Japan. shore Struct 15(3):249–258. https://​doi.​org/​10.​1080/​17445​302.​
2019.​16125​44
Data availability All data generated or analyzed during this study are 13. Viola IM, Sacher M, Xu J, Fei W (2015) A numerical method
included in this published article. for the design of ships with wind-assisted propulsion. Ocean
Eng 105:33–42. https://​doi.​org/​10.​1016/j.​ocean​eng.​2015.​06.​009
Open Access This article is licensed under a Creative Commons Attri- 14. Ouchi K, Uzawa K, Kanai A, Katori M (2013) “Wind Chal-
bution 4.0 International License, which permits use, sharing, adapta- lenger” the next generation hybrid sailing vessel. Third Inter-
tion, distribution and reproduction in any medium or format, as long national Symposium on Marine Propulsors]
as you give appropriate credit to the original author(s) and the source, 15. Wang K, Guo X, Zhao J, Ma R, Huang L, Tian F, Dong S,
provide a link to the Creative Commons licence, and indicate if changes Zhang P, Liu C, Wang Z (2022) An integrated collaborative
were made. The images or other third party material in this article are decision-making method for optimizing energy consumption
included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
Journal of Marine Science and Technology (2024) 29:387–403 403

of sail-assisted ships towards low-carbon shipping. Ocean Eng 19. Kramer JV, Steen S (2022) Sail-induced resistance on a wind-
266:112810. https://​doi.​org/​10.​1016/j.​ocean​eng.​2022.​112810 powered cargo ship. Ocean Eng 261:111688. https://​doi.​org/​10.​
16. Yasukawa H, Yoshimura Y (2015) Introduction of MMG stand- 1016/j.​ocean​eng.​2022.​111688
ard method for ship maneuvering predictions. J Mar Sci Technol 20. Sogihara N, Kume K (2023) Lifecycle assessment of fuel saving
20:37–52. https://​doi.​org/​10.​1007/​s00773-​014-​0293-y in bulk carrier with a flettner rotor, Proceedings of OMAE2023,
17. Yasukawa H, Zaky M, Yonemasu I, Miyake R (2017) Effect of OMAE2023-101078
engine output on maneuverability of a VLCC in still water and
adverse weather conditions. J Mar Sci Technol 22:574–586. Publisher's Note Springer Nature remains neutral with regard to
https://​doi.​org/​10.​1007/​s00773-​017-​0435-0 jurisdictional claims in published maps and institutional affiliations.
18. MOERI KVLCC2 geometry and conditions, SIMMAN 2008,
FORCE Technology. [Online] http://​w ww.​s imma​n 2008.​d k/​
KVLCC/​KVLCC2/​kvlcc2_​geome​try.​html. Accessed 27 June
2023