Proceedings of the 1st Workshop on "Development of Advanced Ship Support System Using Information Technology", Tokyo University of Marine Science and Technology, February 5, 2009, Tokyo, Japan

(1)

Proceedings of The 1st Workshop

on

"DEVELOPMENT OF ADVANCED SHIP

SUPPORT SYSTEM

USING INFORMATION TECHNOLOGY"

February 5, 2009

Etchujima Hall, Etchujima

_Campus,

Tokyo University of Marine

_{Science and Technology}

Tokyo, Japan

P2009-14

Delft University of Technology

Ship Hydromechanics laboratory

Library

Mekelweg 2 _{26282 CD Delft}

Phone: +31 (0)15 2786873

(2)

Closing 17:20-17:30

Program

Opening 13:00-13:10

Sessionl: Maritime Broadband Communication System 13:10-14:10

»- Development of Maritime Broadband Communication / Kohei Ohtsu and Ruri Shoji, Tokyo University of Marine Science and Technology

An Estimation Tool of Long Term Benefits of Auxiliar Wind Propulsion by Means of a Traction Kite Including the Effect of Route Optimization / Peter Naaijen, Wei Shi,

Delft University of Technology, and Jean-Gregoire Kherian, Gusto MSC

Session2: Electronic Navigation System 14:10-15:10

Development of an e-Navigation Strategy in IMO and Study on Navigational Intention Exchange Support System / Yasuyuki Niwa, National Maritime Research

Institute

A Scaled-down version of the Integrated Navigational InformationSystem on Seascape Image for the training ship SHIOJIMARU / Takahiko Fujisaka , Tokyo University of Marine Science and Technology

Coffee Break 15:10-15:20

Session3: Advanced Operation Control System 15:20-16:20

)0. Practical Education for Marine Control Engineering Using an Actual Training Ship /

Jun Kayano, Tadatsugi Okazaki, Hayato Kondo, Hitoi Tamaru and Kohei Ohtsu, Tokyo University of Marine Science and Technology

On the Development of Decision-support and Guidance System for Ships Operating in Close Proximity / [gil Pedersen, Norwegian University of Science and Technology

Session4: Advanced Management System for Marine Engineer 16:20-17:20

'fr Knowledge Bank System for Marine Engineering Operation / Sachiyo Horiki and

Masahiro Osakabe, Tokyo University of Marine Science and Technology

>. Recording System of Seamless Work-log With a Wearable device and its Application to an Education Support System / Katsunori Matsuoka, National Institute of

(3)

Foreword to the Textbook

The Information Technology (IT) has found its way into shipping to achieve safe and efficient marine transportation as is exemplified by the IMO e-Navigation Strategy. In order to lead the way in this new field, the Tokyo University of Marine Science and Technology (TUMSAT) launched a new research and development project in 2008 on advanced IT-based ship operation technologies such as Maritime Broadband Communication System, Electronic Navigation System, Advanced Ship Operation and Control System and Advanced Management System for Marine Engineering. The TUMSAT also intends to integrate the outcome of this project into its curriculum to provide young professionals with sufficient knowledge in advanced

marine-related IT technologies to the broad maritime community including the maritime

industry, maritime education and training institutions and maritime authorities.

The project is run by the following 4 groups:

The Maritime Broadband Communication System Group attempts to develop a maritime

broadband communication system, which enables ships at sea and land-based personnel to share ship operation information.

The Electronic Navigation System Group aims at developing a navigation assistance system following the E-Navigation Strategy proposed by the IMO.

The Advanced Ship Operation and Control System Group focuses on ship control

technologies such as ship to ship operation, tracking control, and automatic berthing.

The Advanced Management System for Marine Engineering Group endeavors to develop a knowledge bank system for marine engineering operation.

We hope that this workshop will become a milestone in the development of advanced ship support systems for safe and efficient ship operation.

Hideo Yabuki

Project manager Professor

(5)

MARITIME BROADBAND

(6)

Development of Maritime Broadband Communication

Kohei Ohtsu

Emeritus Prof TLTMSAT

Ruri Shoji

Associate Prof. .TITMSAT

Introduction

We might say that the first generation of navigation style had continued until 19

century when G.Marconi developed a wireless communication technique in 1987.

During the first generation, a ship was navigating in standing alone, in other word, not

depending on land support except for a light house and manual signal using hand flag

and so on. A ship could receive on board some information from land for the first time

Marconi developed a wireless communication technique. However, the information

which the ship could receive was extremely few as the signals were exchanged by Morse

signal code. Until this generation, however, the gap in remote communication technique

between sea and land were small. The gap has abruptly expanded since G.Bell

developed a remote telephone technique. And nowadays, the gap becomes decisive by

development of internet and satellite communication system. Inversely considering this,

the innovation of communication technique in navigation will yield a break-through in

order to reduce the gap between sea and land and realize safety eco navigation which is

the urgent problem for navigation, by co-operation between ship and land.

In this paper, we will review the real state of the communication technique at

sea

and introduce the actual experimental projects which has be carried out in TUMSAT

and discuss a new world in navigation which will been derived by realization of a high

speed and a large capacity broad band (Maritime Broad Band Network; MBB)

communication network at sea.

2

Sea Infrastructure of Communication Technique

2.1 Infrastructure of Communication System at Land

Fig1 shows transition of the communication speed in LAN

system. This figure

shows that the communication speed has become faster at the rate of 4.3db/year. The

middle line of Fig2 shows transition of the communication speed in cellular

_{phone. The}

upper line is the one of the LAN copied from Fig. 1. We can know that the incline of these

two lines are almost parallel. In any case, transition of the communication speed in land

(7)

g 80 ,11111 140 60 40 20 o 1G 100M: -10MbPs 1 00 o 100 10 1 0. Transmission Spee Mbps -1?m-f 4ADSL1.5M .57.6k 136N 128k 24delay 384kbps(FOMA) 4-- \--32kbps(PHS) 9.8kbps(PDC) 1990 1995 2000 2005

Fig.2 Speed Up of Cellular Phone

2.2 Infrastructure of Communication System at Sea

2.2.1 Gap of communication speed between land and sea

On the other hand, the communication speed at sea is surprisingly slow. Fig 3

shows

the gap between them (Japanese Ministry of Internal Affairs and

Communications).

Seeing this figure, we might perceive that overtaking at sea communication to

developing at land communication services is too late.

Ultra High Speed Wir AN

Phone

Avtate e

ration _{". Ku}

.

IrirrAfi SAT NStar----

Band

1995 2000 2005 2010

Fig.3 Communication Speed Gap between Land and Sea

Gap!!

2 4Mbp (EV-DO 64kbps(PHS)

1985 1990 1995 2000 2005

Fig.1 Speed Up of LAN

2 -

#.-t74') 140 120 c,s, 100 80 60 40 20 O

(8)

2.2.2 Coastal Navigation

In coastal navigation, most popular communication method in Japan is to utilize a

cellular phone. However, the cover area of cellular phone at sea is extremely narrow

because the antennas for cellular phone direct to inland. Navigating out of the range,

the satellite communication systems called by N-Star ( DoCoMo Mobile Coltd) are

popularly utilized. It covers about 200 NM with max 64kbps.

According to the questionnaire to crews of inland ship carried out by Ministry of Land,

Infrastructure, Transport and Tourism, large claims on the present communication at

sea, are that the communication speed is slow and the cover range, narrow, considering

high cost

2.2.3 Ocean Navigation

There is no method except for a satellite communication system from ship to land

because the radio operator has not got on board in many recent ships. The most typical

communication service in the satellite communication is the INMARSAT satellite

services operated by International Maritime Satellite Organization whose band is L

band (1.6/1.5GHz). Table 1 shows the recent services in the INMARSAT. A frequently

used one among them is the INMARSAT-F (the communication speed is 64kbps).We can

understand that the communication speed is extremely slow comparing with land

communication because this communication speed was equivalent to the one in the first

generation in land service.

Conclusively speaking, the most urgent problem in the communication services at sea

to be grappled with, is realization of higher (at least over 1Mbps in up link speed,Ku

Band) speed low cost services as much as possible.

Table 1 Feature of INMARSAT Services

Type Voice Telex Fax Data A Analog Yes 4.8kbps 4.8kbps B Digital Yes 9.6kbps 9.6kbps C Yes _600bps M Digital 2.4 kbps 2.4kbps miniM Digital 2.4 kbps 64kbps Aviate Digital 4.8 kbps 4.8kbps F _Digital _{9.6 kbps} _64kbps D _32kbps Bgan Digital 14.4kbps 492kbps Aviate Digital 2.4kbps 64kbps

(9)

2.3 The effects obtained by realization of High Speed Large Capacity Communication

Table 2 demonstrates some effects induced by realization of broadband communication

service between sea and land. Most expected field of the contents in the last

questionnaire is information support to ship from land.

Table 2 Applications of High Speed Large Capacity

Communication at Sea

3 Marine Broad Band Project

3.1 Introduction

TUMSAT began a projects called by Marine Broad Band Network in 2005

aiming realization of a new world of navigational technique which is derived by a high

speed and a large capacity communication network system. This project is supported by

co-working team between TUMSAT and 7 companies, NTT communication system , Sky Perfect JSAT and DoCoMo Mobile treat

with network, Ube Kaiun, Mitsui Zosen

,Mitsui Akishima, JRC..

This project aims that

Realization of communication services

between sea and land in coastal and

Field Item Form

Image, Data

TV phone,

Internet,

Mail,TV

Medical Remote MedicalService

Image, Data

Diagnosis to patient

Safety

Navigation

Outside

and

Inside

Lookout

Image,Voice Lookout by camera etc

Monitoring

Hull Monitoring

Engine Monitoring

Cargo Monitoring

Data

Monitoring of Hull stress,

ship's

motions, rpm, FO.

Thrust ,Cargo

Information

Support

to Ship

Environmental

Information

Data

Weather , Wave Forecasting

Weather Routing

Route Information

Guidance Way

Point

Tracking

and Control

Data

Guiidance and Control of

Ship

in Ocean and Port

Ship

Management

Damage Control Ship Performance

Data, Image

Damage Control,

Long Term

Performance

Management

Risk

Management

Hazard at sea

Data, Image

Real Time

Information

(10)

ocean navigation and development of useful contents for safety and eco and energy

saving navigation

In this chapter, we introduce the activity of the marine broad band network project.

3.2

Maritime Broad Band Network for Ocean Going Ship

The marine broad band network project began with a satellite communication

between land and sea using the JSAT's satellite and the service of NTT communications.

Fig.4 shows the feature of the satellite maritime broad band system developed by the

project. The communication speed of the network realized 1 Mbps in up link and 1.2

Mbps in

down link

between

the

training

ship

Shioji Maru in TUMSAT.

Witt - )1

a

i

aita3- 5,, 'MIMI JD'1-41#3

II

7,0 1 RF I Ammon K25--BER2D00 I mares1-1

sp

S/Gt

4'44*

IIIIIP 7-0 EM45111''5 [77(Ilg) I_7-0 --1 e ,.. F. A

"RIM=

11E71)-5- tRiil ciFtOft SW-1-11-16 GSM=

Fig.4 Feature of Maritime Broad Band Network

3.3 Coastal Marine Broad Band Network

On the other hand, as the coastal navigation project, a wireless LAN network system

using many light houses, in which many ones has already be operated without human

(11)

operations(Fig.5). In Japan, there are now 3315 lighthouses and 1251 buoys. All of them

are unmanned. Except for the light house of entrance of port and so on, they are located

at high places in tips of peninsula. Their light arriving range are 20-30 NM.

We began

it in the entrance of Tokyo Bay. Fig 6 shows one of the light house to be installed and

antenna and transmitter.

Light House and

Fig.5 Concept of Coastal Broad Band Project Utilizing Light Houses

Fig.7 shows one example of the wireless LAN bases in the entrance of Tokyo Bay.

According to Fig.7 we need at least over 5 light houses in order to cover the entrance of

Tokyo Bay when the range of wireless output is 3 nm whose value is maximum range

permitted by law.

In near future, this project will become a national project.

(12)

- '12. 4.

Tokyo_ Báy

5Miles

Fig.7 Light Houses Alignment at Entrance of Tokyo Bay (The Doted Circles are not permitted by Law)

3.4 On-board Local Area Network (LAN) System

In order to realize the broad band communication between land and

_{sea, we need real}

time observing system and data transmission technique through

on board LAN system

and finally, establishment of unified standard of signal communication.

Especially, an

establishment and standardization of on board LAN system is urgent problem.

Our

project has technically supported the project supported by Japan

_{Marine Equipment}

Association.

4 Applications

4.1 Example of Image Data Communication

In this chapter, we describe some examples carried by the project using the satellite

network

system. Most powerful

effect

gained by realization

of broad band

communication network is to be possible to exchange

_{many image data. The people}

living on land and the people on board

_{can receive or send the real and clear image data.}

Effective application of the image data exchange

_{between ship and land is damage}

(13)

time image using a head mounted camera in an engine room. The image data can are

sent to land through the established network. Not only seafarers on board but also the

staff on land can monitor the state of engine and find the place of damage.

Engine

Monitoring

MEM =NM EMI *LI MO-MN

Fig.8 Engine Room Monitoring using Images

Nowadays, important information on navigation, for example on prediction of weather,

wave, and tide are gained through

INTERNET. The crews on board can gain such a

comfortable internet environment by the broad band

communication system. Fig.9

shows one example of the navigation information view. In this example, the people on

land can monitor ship's motions in real time.

Fig.10 shows a lookout view from a flying

bridge through a Skype camera system.

Another feasible important application is a remote

medical services. Using a clear

picture through the network system, a patient on

board can receive effective medical

examination by doctor in a hospital on land. This project will be actually tested by

National Institute for Sea Training.

(14)

Navigation

Information

Views

.

8, -111,

1.1 II e ria ..1.,. El ...um, OPRAnnrail_a__a_

- rrniT2----114,,,,. imir Igor mi., 'Arms. loom

Fig.9 Navigation Information Views on Land

L.-INV a 111

M

r . 1 "B. ,i1V77.114.3,0 4-1R4 12011 ..tsakom ibwro .4.S. 4 LI EA (.00

Fig.10 View from Flying Bridge through

a Skype Camera

100 200 210 220 '0 2*

MliM=1 ;i00% I 8 ?;

(15)

4.2 Example of Ship's Monitoring System

Nowadays, concerns having the highest priority in earth level are saving energy and

the realization of eco-society. The navigation of ship is also not exceptionable. In ship's

navigation, most energy loss is given by strong ship's motions at rough sea, the

movement of main engine and other machineries. Therefore the monitoring of ship's

state is important for realization of safety, eco and energy saving navigation

However,

the number of crews are now reduced and resultantly, they have no room to monitor all

states of ship. Thus, it is important to support the crews from land.

Fig. 11 are heading, rudder, true wind direction, wind force, drift angle, rolling,

pitching and thrust of a large ocean going ship sampled at 1 sec ,their spectra so on. Fig.

12 shows the change of their spectra. By monitoring such a data by staff on land in

stead of busy crews on board, we can realize energy saving and eco

navigation.

I FF

-

---Data Inspection v,., E. A A r E.

Fig.11 Time Series Records of Ship's Motion Fig.12 Analyzing of Ship Motion

4.3 Example of

Ship's Long Span Management

Received data from ship are stored on land at each ship and then we can make own

data base of each ship. For examples, by analyzing such a data, we can gain information

on trend, seasonal trend

and prediction of sea margin in each ship. Fig.13 shows the

change of sea margin in long span. We can decompose such long span data of trend,

seasonal

element and aging effect. Based on this information, we can judge the timing

(16)

J

Fouling+Aging+Seasonal =Se SI Slargin Slodel

1982 Actual Nlargin ¡OW fi014101: Strike!

/

I 4#P10 : : n a II-.. Hal" AgingFffct

Sea Nlargin Model

Fig.13 Analysis of Sea Margin in one middle class container ship

4.4 Example of Radar Information Exchange

RADAR gives precious information to navigator in order to know own and other ship's

positions. By acknowledgement of other ship's positions, the ship can determine her

next course and speed in order to avoid collision with other ships. Moreover, recently

an AIS (automatic identification system) became compulsory in large ship. If other

ship's position not only near own ship but also out of range of own radar get, safety of

navigation would increase. Our project team is now creating radar information

exchange system between many ships. Fig.14 shows example of the radar information

to be sent from ship to other ship or to traffic control center through the broad band net

work system. 1981 1 e 1983 J n _n 4181 6410 Dap. in 'cr. ice 8110 Fouling Effect 200

(17)

i

Fig.14 RADAR Information of Tokyo Bay in TUMSAT

4.5 Example of Ship's Weather Routing

System

A ship builds navigation plan before leaving, considering change of weather during

sailing to destination port. However, the weather condition during sailing often

changes from the expected one. Thus, the ship need demand rerouting based on

updated weather conditions to land. The staff in land rebuild an energy saving or

minimum time navigation plan, considering the updated ship's states received through

satellite network. The marine broad band

project has carried out the actual

experiments of weather routing and re-routing

using inland sailing ship. In this

experiment, the way points are given to the ship, the ship tracks along the given route

(18)

Information_ 40_

O

38

Give an Optimal Route

From land

42 36 ... ... ... 34 139 140 141 142 143 144 145 LONGITUDE o o 5444 Wind 13044110n Y5.31 /1441 34444 V/ 144 5,445 07 1. 4.31

Optimal Route

Tracking

Fig.15

Optimal Weather Routing Navigation in an Inland Ship

5

_Conclusion

This paper described about the necessity of innovation in its speed and capacity

of

communication between ship and land and introduced the recent projects on the marine

broad band. in TUMSAT. We hope that the marine broad band network fixes

_{to and}

realize safety and energy saving echo navigation system through the network.

We thank for the contribution to our project to NTT Communication

Co.Ltd

Skyperfect JSAT Co.Ltd, Ube Shipping Co.Ltd, Mitsui Ship Building

_{Co.Ltd JRC Co}

Ltd ,Mitsui,

Akishima Co.Ltd and Toyo Engineering Co.Ltd

(19)

An Estimation Tool of Long Term Benefits of Auxiliar Wind

Propulsion by Means of a Traction Kite Including the Effect of

Route Optimization

Peter Naaijen, Wei Shi, Delft University of Technology, the Netherlands Jean-Gregoire Kherian, GustoMSC, the Netherlands

SUMMARY

This paper describes an estimation tool to assess long-term benefits of auxiliary wind propulsion means

of traction kites. Methods are described to estimate kite propulsion force, ship and resistance and

propulsion performance without using any specific experimental data of the considered ship and kite

system. A voyage optimization/simulation tool is presented that has been developed in order to

assess

the effect of route optimization. Results of the performance prediction for the 52560 tons displacement

bulk carrier Jin Hui is presented.

INTRODUCTION

Although the recent drop in oil price

dramatically

decreases the direct financial

benefit for ship owners of reducing fossil fuel

consumption, there is

still enough reason to

search for sustainable alternatives if the long

term is considered: sooner or later, the oil price

will increase again. Besides, there is an ongoing

process to regulate emissions in the UN-based

International Maritime Organization (IMO).

which is based on Art. 2.2 of the Kyoto Protocol.

EU has stated that international shipping to the

EU region may be regulated by EU by linking it

to EU ETS from 2012 unless a global regulation

is established [1]. Either way, polluting will

have to be paid for in the near future.

Shipping is the most energy-efficient mode of

transport which is probably why the sector has

got away with doing very

little

to reduce

emissions until now. The fact that shipping

(responsible for transporting more than 90 % of

world trade) accounts for approx. 3-5 % of

global CO2 emissions

is

more and more

considered to be a good reason to change this.

One way to achieve this is to reintroduce wind

power. Having some advantages compared to

conventional sails, a traction kite propulsion

system for commercial ships is an interesting

concept which is being worked on by various

parties in industry

This paper describes an accurate method to

assess the long term benefit of wind assisted

propulsion, applied on the specific concept of

a

high performance (high lift to drag ratio)

kite

for traction force generation.

Attached to a single tow line via a steering

gondola, these kites can be actively controlled in

order to create high flying speeds resulting in

high traction force. Figure 1 depicts such

a

(20)

Figure 1, Ship with kite (courtesy: Skysails)

Compared

to more conventional wind

propulsion by sails, there are some benefits

involved with applying kites:

a kite can be actively controlled in order

to create

its own flying speed thus

increasing its apparent wind speed and

the traction force: more traction power

can be created with less 'sail' area this

way.

due to the fact that a kite can fly at

higher altitudes it is exposed to higher

wind speed

due to the low attachment point of the

tow line the roll heeling moment is

considerably smaller

there are no masts taking deck space

A logical

first

step when quantifying the

benefits of such a system is to analyze its

performance for

a range of environmental

conditions. A detailed description of such an

analysis can be found in Naaijen et.al. [8].

When considering the area of operation of the

specific

ship of which the performance

is

known, average long-term wind and wave data

can be used to estimate the long term fuel

saving potential. This kind of analysis has been

presented by Jdger [2].

However, since the environmental conditions

play such an important role in the case of wind

propulsion, it can be expected that the efficiency

increase that can be obtained by applying

weather routing is even more pronounced than

is the case for conventional propulsed ships, and

therefore should be taken into account to come

towards a fair and accurate estimate of

long-term fuel saving.

In the following, the estimation

of

the

performance of a kite equipped ship, based on

some general input parameters

is briefly

described. Results for case study are presented.

Next, the

tool is

described that has been

developed to include the effect

of route

optimization of long-term fuel saving.

Performance analysis

Introduction

This section describes how the performance of

the ship

kite system is analyzed. First the kite

traction force calculation is explained followed

by a description of how kite, hull, propeller and

engine interact finally

resulting

in fuel

consumption and forward speed that can be

obtained in prescribed wind conditions.

Kite traction

A kite can be considered as a wing surface

which enables the application

of existing

aerodynamic concepts: the resulting force acting

on a kite is determined by

calculating lift and

drag on a 3D wing surface being governed by

relative wind speed and angle of attack. The

main assumption for the model is instantaneous

equilibrium between the direction of the tow

line and the direction

of

the

resultant

aerodynamic force on the kite. For a kite, this

equilibrium is depending on its position in space

which will be described using the so-called

flight envelope (FE). The apparent wind speed,

experienced by the kite is a combination of true

wind, ship speed and the kite's own flying

speed. Depending on the position, the kite

will

develop its flying speed in such a way that

the

resulting force is parallel to the tow line. It's

the

(21)

calculation to determine this equilibrium speed

and the resulting towing force.

Flight Envelope

The set of possible positions in space of a kite,

attached to a tow line with length r, is described

by a quarter sphere with radius r, which is called

the flight envelope (FE). See Figure 2 where the

direction of the true wind is indicated. True

wind speed is defined by W. Point F is the

attachment point of the tow line. The half circle

LUR is called the edge of the FE. P is the centre

of the so-called power zone. When assuming

uniform inflow over the altitude in the FE, P is

the point where the highest speed and traction of

the kite are obtained. The boundary layer for the

wind speed in which wind speed increases with

altitude will result in an maximum speed and

traction force occurring at a certain altitude

above the mentioned point P.

All half circles parallel to LUR are called

iso-power lines: kite speed and traction are constant

on these lines. All circle segments from P to the

edge are iso-gradient lines: the gradient of speed

and power has a constant maximum value on

these lines.

The position of the kite within the FE, indicated

by K is described by two angles (See Figure 2):

O is the inclination of the tow line FK with

respect to the line FP

(1)

is the inclination of the plane FKP with

respect to the horizontal plane

In order to describe the flying direction of the

kite, a kite reference system _{xic, YK, zK is defined}

having its origin at K. The xi( axis is tangential

to the iso-power line through K pointing from L

to R, while the yK axis is tangential to the

iso-gradient line through K pointing towards P. The

zK axis is parallel to the tow line, pointing

outwards the FE.

y is the angle between the flying direction of the

kite and the positive xK axis. See Figure 3

showing a top view (looking in negative

zK direction) on the kite.

Figure 2, Flight Envelope (FE)

YK

v,

XK

V.

Figure 3, Kite in projected plane of FE

Having defined how the kite's position and

flying direction are described, this definition is

used now to

formulate the apparent wind

experienced by the kite.

First, the apparent wind is split up into a part

tangential to the FE (vt) which is a combination

of tangential velocities vt_x and lit, in xK and yK

direction respectively, and a radial part parallel

(22)

The tangential

velocity

is caused by a

combination of true wind and kite speed. The

part due to the kite speed can be split up into a

part tangential to the iso-power lines in the

direction of xK (vx_k) and a part tangential to the iso-gradient lines in the direction of yK (vy_k). In

YK direction, there is a contribution by the true

wind (v) as well. With the above mentioned

definitions, the following formulae for these

velocities can be found:

= -rd + Wsin(0) = v y_k+v),_, (1)

i 43= vx, (2)

Combining these two contributions to the total

tangential velocity yields:

(3)

For the flying speed vk and direction y of the kite follows:

Vk=NIVy_k2 + V k_k2

,Vk_k

The radial velocity is caused by true wind only:

= Wcos(0) (6)

With the above defined velocities, a formula

for

the total relative velocity vrei and the angle

of

attack a experienced by the kite can be derived:

(See Figure 4 which depicts a cross section of

the kite and the involved angle of attack.)

Figure 4, Kite Angle of Attack

v a =

arctanH

v,

a, = a -;

k

Perpendicular to tow line

ChorNd line Re l edative wind' N

a

ow line Vt e %

v,

(8)

To find the effective angle of attack, ae, the

angle at which the kite is attached to the tow

line, ak, has to be subtracted from a. (Figure 4)

(9)

Lift and Drag

To determine the resulting force on the kite, first

lift and drag of the 2D airfoil are determined.

Corrections to this 2D lift and drag coefficients

are made in order to take into account

3D

induced drag and the curvature of the kite.

Furthermore some additional

contributions to

the drag are considered which take into account

line drag, inlet drag, and drag due to

irregularities and surface roughness.

2D lift and drag coefficients are calculated

by

the free available panel method program XFOIL

by Drela & Youngren [3]. The inviscid

flow

calculated by the program is constructed by a

superposition of three potential flows being the

free stream, a flow created by a vortex sheet on

the airfoil surface and a source sheet on

the

airfoil surface and wake. y = arctan v v_k

(23)

The viscous part of the solution resulting in

frictional resistance is described by boundary

layer shape parameter equations. For a detailed

description of XFOIL reference

is made to

Drela & Youngren [3] and Drela [4].

The obtained 2D lift and drag coefficients have

to be corrected in order to include 3D effects.

This is done based on Prandtl's lifting line

theory assuming elliptical lift distribution over

the wing span.

The fact that a kite has a certain span wise

curvature will also effect the lift of the whole

kite. Resolving the lift force perpendicular to the

curved kite into the ZK direction results in the

following formulation for the lift on the curved

kite according to Lingard [6]: (See Figure 5).

CL = C COS' (4) (10)

where:

CL= 3D Lift coefficient of straight wing CL, = 3D Lift coefficient of curved wing

= angle of curvature (See Figure 5.)

.-kite

4 controlling gondola

Figure 5, lines between gondola and kite

The lines between steering gondola and kite,

where the speed is assumed to equal the speed

of the

kite itself,

appear

to generate a

considerable amount

of

drag. The drag

coefficient of all these lines together can be

determined using the formulation of Prakash

[7]:

CD, =n Rdcos3(a,)

S (11)

where:

Cal = drag

coefficients

of lines

between gondola and kite

n = number of lines

R= length of lines between gondola and kite

d = diameter of individual lines

= angle of attack between relative speed and

kite line S = kite area

According to Prakash [7] the number of lines

depends on the kite aspect ratio as follows:

n=8+16AR

(12)

The diameter of the individual lines between

gondola and kite is chosen such that their total

cross sectional area equals that of the tow line

between gondola and ship.

For determining the line drag the part of the tow

line underneath the

gondola and the

part

between gondola

and kite are considered

separately. For the lower part it appeared that

even when taking into account a line drag

coefficient of 3, the drag of the tow line did not

exceed 1% of the total drag (partly because the

speed of the tow line itself is low). Therefore the

lower tow line drag is neglected. (It must be

noted however that an increase in line drag can

be expected due to vortex induced vibrations

which is not taken into account here.)

Other additional drag coefficients come from

the air

inlet, and surface irregularities and

surface roughness. The air inlet is an opening at

the nose of the kite enabling air flow into the

inflatable kite. Approximations

for

drag coefficients

of these components given by

Prakash [7] have been used.

As mentioned,

the key assumption of the

presented approach to calculate the kite traction

is the resultant force on the kite being parallel to

the tow line. When assuming that the Reynolds

number (based on which the 2D lift and drag

coefficients are calculated) is independent of the

instantaneous relative kite velocity v

I and as a

consequence independent of the position of the

kite in the FE, the Lift to Drag ratio (LID) is

also independent of the kite

position. This

means that the direction of the resultant kite

force depends on the angle of attack only. With

a known L/D, the required angle of attack for

which the resultant kite force is parallel to the

(24)

tow line is easily determined and also

independent of the kite position. Equation (8)

gives the relation between angle of attack and

radial and tangential relative velocity of the kite.

The radial velocity vr_, is a result of true wind

only (equation (6)) and depends on the position

of the kite on the FE. The tangential velocities

12,, and 1/1_,,

however are strongly dependent of

the kite's own speed in terms of (i5 and Ò. So for

a given flying direction y and a given

position

on the FE (defined by 0 and t9), the

required

angle of attack for which the resultant force on

the kite is indeed parallel to the tow line can be

obtained by tuning the kite's own speed in terms

of

and d:

By combining equations (1), (2) and (5) é can

be expressed as follows:

sin(9) 0 tan(y)

(13)

By substituting equation (13) and equations (1)

and (2) in the expression for the angle of attack

(8), a quadratic equation for 0 can be obtained:

2(1-2sin2(61).(1+ tan2(y))) (2rsin(19)tan(y))- Wsin(61)

=

'sin2(0)+(Wc°s(e)

pv

tan(a) (14)

Having solved q3 from equation (14), é follows

from equation (13).

Knowing the kite velocities the instantaneous

relative velocity v,/ of the kite is known which

enables the calculation of the resultant force on

the kite from the lift and drag force:

L=Y2pv,.2SCL (15)

D=Y2pv,2S-C (16)

As mentioned calculation of lift and drag

coefficients is based on a constant Reynolds

number (independent of location in the FE

and)

independent

of

the instantaneous

relative

velocity of the kite. Therefore the lift to drag

ratio is also independent of the location on the

FE and only depending on the angle of attack.

The angle ak between kite cord line and tow line

has been chosen such that the angle of attack for

which the resultant force is parallel to the tow

line is giving the maximum lift to drag ratio. For

the kite considered during the case study that

will be described in the last paragraph, this L to

D ratio amounts to 3.5.

As mentioned, one of the benefits of a kite is

that its relative velocity can be increased by

actively maneuvering it on a desired track on the

FE.

Such a track could be an orbit shaped as

depicted in Figure 6.

O

-300 -200 -100 0

Figure 6, possible kite track on FE

When a certain track is prescribed, total relative

kite velocity and traction force at a finite

number of points on the track can be

determined. The average traction force and its

direction can be calculated by a time integration

over the chosen orbit. In the present study an

orbital shape as depicted in Figure 6 is

considered. The mathematical expression of the

selected type of orbit is presented in Wellicome

[9].

In case of a kite towing a ship, the wind that

enters the FE (being called 'true'

wind until

now) is in fact a combination of true wind and

wind created by the ship's own speed. In

general the direction of these two will not

coincide. The FE is positioned on the ship

in

(25)

such a way that its edge is perpendicular to the

direction of combined true wind and ship speed

at the average flying altitude of the kite. As the

kite's flying altitude is supposed to be within the

so-called surface layer of the atmosphere, where

the occurring wind is dominated by pressure

differences and no geotropic winds occur, the

variation of wind speed with altitude can be

expressed by a logarithmic profile (Troen [5]):

W (z)= Cbg

H-

(17)

z,

where:

W(z) = Wind speed at altitude z above (sea) surface

Ctog where:

uf= known wind speed at reference level

zf= reference level (10 m)

zo = surface roughness (depending on wave height)

Figure 7 depicts a top view on ship and FE for

a

stern quartering wind condition. The hatched

area represents a likely area for positioning the

kite track.

App. wind direction at kite's altitude

Figure 7, FE on ship for bow quartering wind

For a given ship course and wind direction,

a

range

of tracks

having similar horizontal

amplitude as shown in Figure 6 are tested. The

track generating the highest force in the

ship

direction is selected. The horizontal and vertical

amplitudes of the kite track are kept

constant

and were chosen so as to represent a realistic

kite flying behavior.

Concerning the vertical position of the kite

orbit, an optimum can be found resulting in the

highest mean traction force in the ship's sailing

direction.

This optimum flying altitude is governed by:

the variation of traction force over the

FE

the variation

of wind velocity over

altitude

variation of horizontal

traction force

with inclination of tow line

To assess the effect of the flying altitude on the

kite force in the direction of the forward speed

Xkite, and to find the optimum flying altitude in

terms of maximum Xkite,

simulations of the kite,

flying along an orbital track, have been made

for various flying altitudes during a case study.

(Details about this case study are presented later

on.) For three different towing line lengths, 150

m, 350 m and 550 m and three different wind

speeds, the traction force in the direction of the

forward ship speed (time-averaged over one

revolution on the orbit) has been calculated for

various flying altitudes.

It appears that the altitude at which maximum

forward towing force occurs amounts to 27

to

33 % of the towing

line _{length, slightly} decreasing with increasing wind speed.

Above presented results are for a wind direction

of 0 degrees off the stern. Similar calculations

have been made for different wind directions

resulting in optimum flying altitudes in the

same

range.

The properties of the kite that has been applied

for the current case study are as follows: Kite area:

Lift coefficient: Drag coefficient: LID ratio:

Line length:

Average flying altitude:

S = 400 m2 CL=0.73 CD=0.21 L/D=3.5 r=300 m

100m

(26)

The lift and drag coefficients determined as

described above have been used. Wind tunnel

measurements reported by Gernez [10] resulted

in lift and drag coefficients of 0.70 and 0.18

showing that

the calculated values can be

considered to be realistic.

For a constant ship speed of 15 kn the resulting

kite force in the direction of the forward speed

of the ship is presented in Figure 8

o 5 10 16 a, o_ 4 20 25 a) f-'1 30 35

:700

600 500 400 300 200 100 o

paragraph, only the longitudinal drag is relevant

for the current study.

The input parameters that govern the drag force

coefficient

G given by Isherwood for the

longitudinal force are given in Table 1.

Coefficient

G

and corresponding force are

related as follows:

c =

X

pc,V2A,

2

I"

Where V, indicates the

experienced by the ship. The

dependent of the

apparent

relative to the ship.

(18)

apparent

wind

coefficient G is

wind direction

Table 1, parameters used for Bulk Carier to determine

C, wind coefficient

Calculation of Fuel Consumption

and

ship speed

In this section the method used to determine the

performance of the existing propulsion system.

As will be explained in the next section, the ship

operation scenario

to be considered

is the

constant propeller revolutions scenario: given a

certain fixed propeller speed, ship speed and

fuel costs engine are to be determined.

Since

ships hull resistance, provided thrust

by the

propeller and kite force are all

depending on

LOA, Length o.a. [m] 189.99

B, Breadth [m] 32.26

AT, Transverse projected wind

area[m2]

756

AL, Lateral projected wind area

[m2] 3570

S, Length of perimeter of lateral

projection [m]

(Excluding waterline and slender

bodies)

300.00

C, Distance from bow to centroid

of lateral projected area [m]

115.00

M, Number of distinct groups of

masts or king posts [-]

2

o 50 100 150

True wind direction [deg]

Figure 8, Kite force 1kNi] against true wind speed [kill and direction [deg] for ship forward speed of 15.0 kn

An operational limit in terms of a maximum

wind speed at which the kite can be applied has

been assumed. Both apparent wind speed

and

true wind speed are capped at 20 m/s for kite

application.

Wind drag

Apart from the wind driven force generated

by

the kite a second wind force has to be taken into

account that results from the air flow around the

hull

part

above the water line and the

superstructure. To approximate this force the

method of Isherwood [11] is applied, which is

based on regression analysis of

wind tunnel

measurements.

This method provides drag

force coefficients in longitudinal and transverse

direction and a yaw moment coefficient that can

be obtained

by

substituting some main

characteristics concerning the ship' s

(27)

forward speed, the determination of the forward

speed is an iterative process. Once the ship

speed has been solved, corresponding engine

power and fuel consumption can be determined.

A detailed

description

of

the calculation methods can also be found in Shi [12]

In this section the

following nomenclature is

used:

diameter of the propeller, m

advanced ratio

engine brake power, kW

effective towing power, kW thrust power, kW

transmission power, kW

torque of propeller, Nm

model-ship correlation resistance, kN resistance of appendages, N

additional pressure resistance of bulbous

bow near the water surface, kN

frictional resistance according to the

ITTC-1957 friction formula, 1(1\1

total ship resistance, kN

additional pressure resistance of immersed transom stern, kN

wave-making and wave-breaking resistance,

kN

thrust of propeller, N ship speed, m/s relative quality

form factor describing the viscous resistance

of the hull form in relation of R1 propeller speed, rad/s

thrust deduction factor wake factor

hull efficiency open water efficiency relative rotative efficiency transmission efficiency

seawater density, kg/m3

Power Transmission

From the propulsion system point of view, the

main engine generates amount of power, known

as the engine brake power, transferred by the

transmission system, (gearbox and shaft), the

power propels the propeller, generating amount

of thrust to overcome the ship resistance. See

Figure 9 for the power transmission through the

propulsion system and the consequent

efficiencies. b P, P th

P

Q prop RA R4 RB RE Rtotal RTR Rw TB, Vship 0* I +k prop tin lo

Figure 9, Propulsion chain

Ship Resistance

In general, the ship resistance is determined by

the sailing speed, the shape of the hull and the

wave conditions. In the ship design phase, there

are different kinds of methods to predict ship

resistance and other factors, such as the thrust

deduction factor (t) and the wake factor (w), on

the basis of ship dimensional parameters. The

Holtrop and Mennen method, presented

in

Holtrop [12] & [13], has been implemented to

calculate ship resistance.

According to the Holtrop and Mennen method,

the total ship resistance is divided into six parts,

as shown in (20):

R1010, = (1+ k,)R, + R4 +R + RB + RTR +R,

In Holtrop [12] & [13], a statistical method,

which resulted into a set of prediction formulas,

was presented for the determination of each

component of the resistance. This method was

based on regression analysis of random model

and full-scale test data. By running this part of

simulation program with the input of ship

design data and ship speed, an accurate value of

ship resistance, as well as the thrust deduction

factor (t) and the wake factor (w), are generated.

Thus, the effective towing power and the hull

efficiency are calculated:

D

Hull Hydromechanics

As was concluded by Naaijen et. al. [8], the drift angle of the hull and required rudder angles

resulting from the transverse force and

yaw

moment caused by applying a kite as auxiliary

propulsion are very small. The effects

on the

Pe fotalVship

(1 t)

H

_{1 w)}

(28)

4-overall efficiency appeared to be negligible. For

this reason, in the current study only the

longitudinal force balance is considered. No

additional resistance components due to drift and rudder are taken into account.

Propeller

In order to balance the moving force acted on

the ship body, an estimate of the performances

of the propeller is required. The Wageningen

B-Screw Series are used to predict the thrust, the

delivered torque and the open water efficiency

of the propeller.

Concerning with the steady-state shipping

operation conditions, the ship speed remains

constant, by means of the balance of the ship

resistance, the wind force and the propeller

thrust.

The thrust and torque performance of propeller

of

the

Wageningen B-Screw Series are

represented by the standard open water curves

using J, KT and KQ. On the basis of experimental

data, refer to Kuiper ,

the Maritime Research

Institute Netherlands (MARIN) developed KT ,

K Q versus J diagram to describe open water

diagram of propellers within Wageningen

B-Screw Series, see Eq. (21), (22) and (23) Also,

fig. 2 illustrates the KT, K Q versus J diagram of

the reference bulk carrier.

j

n D K =TP"P (21)

(22)

0.0 0.4 go; 02 0.1 OpenWater Dagrern

Figure 10. open water propeller diagram of the considered bulk carrier

When calculating VA from ship speed,

the

effective wake factor (w) according to Holtrop

[13] is used. The same source gives values for

the thrust deduction factor

(t)

and relative

rotative efficiency (17,) are used to correct the

open water

thrust

Tpro,

and torque Qpp.

Afterward, the demanded power in propelling

the propeller is:

p P

(77,77 0)

(8)

Transmission system & Main Diesel Engine

With the assumption of ?it, = 0.97, the demanded

engine brake power at specific ship speed is

achieved:

(9)

Concerning with the fuel

consumption,

a

simplified calculation method, based on engine

brake power and engine rotation speed,

is

implemented. As a shown in Brussen [15], the

algorithm can be expressed as:

FC* = (a + bl),* + c(Pb* )2)(d + eneng* + f (,I; )2) (25)

According to Eq. (26), there are 6 parameters

which determine the

trend

of engine

fuel

consumption. By combining typical operation

conditions, Table. 1 illustrate a f coefficients

of the main diesel engine of the reference bulk

carrier.

KQ=

P"/

(pnpr,2D5)

(23)

Since the open water diagram is based on

the open water test, during which the flow in

front of the propeller is uniform, the open water

propeller efficiency is introduced, as shown

in

(24),

as well as the

rio

versus J diagram

illustrated in Fig. 2: T

V /

_{=Kr. 1/} ₍₂₄₎ / (27z- Qpropn prv) (271- KQ)

Iop

0.1 02 0.3 0.4 07 0 &Man. RM. J

(29)

Table 2, typical engine operation point and coefficients

Table 3, main particulars, engine and propeller characteristics of considered bulk carrier

Overall propulsion performance Bulk

Carrier

The overall performance of the bulk carrier due

to kite application and wind drag is presented

here. For an appropriate range of wind

conditions the force balance of propeller thrust,

kite force and wind drag and hull resistance (all

speed dependent) is solved iteratively. The

resulting ship speed and fuel consumption

are

stored in two dimensional lookup tables in order

to avoid time consuming iterative calculations

during the route optimization / simulation

described in the next section. These lookup

tables are visualized in Figure 11 and Figure 12.

In both figures the nominal operation

performance (without wind influence) is shown

by the single contour line and indicated value of

fuel consumption and forward speed.

500

Figure 11, Fuel consumption for given true wind speed

and direction o 10 20 30 40

Fuel consumption

[ton/1;2470r--Effect kite -81

Effect kite

'4

50 100

150

50

50 100 150

:1

52 1.5 1.48 1.46 1.44 1.42 1.4 1.30 1 36 17.5 17 16.5 16 15.5 15 14.5 14 13.5

Figure 12, Ship speed at constant propeller rate for

given true wind speed and direction

Route optimization

As mentioned, the added value of the developed

tool above existing estimation methods is to

include the effect of route optimization which

might be of significant impact in

case the

performance strongly depends on environmental

conditions.

Route optimization Algorithm

The method used for the actual route

optimization is the so-called modified isochrone

method that was developed at Delft University

of Technology by Hagiwara [16].

Instead of a spatial grid, a time grid is applied in

this method. See Figure 13 : from the starting

Sulzer 6RTA48TB Op. point 1 2 3 4 5 Pi, (kW) 8203 o o 6152 4102 no,0(11m1) 118 118 35 107 94 FC (gis) 389.7 31.1 23.3 283.1 188.1 a 0.0908200 b 1.0532000 c -0.0045346 d 0.6783600 e -0.1477300

f

0.3463700 Bulk carrier General Name: Jin Hui Classification: ABS Al DWT, ton: 44579 Displacement, rn3: 52559 Design speed, knot: 14.8 Principal dimensions Length o.a., m: 189.99 Length p.p., m: 182 Beam mld, m: 32.26 Draught, m : 10.75 Depth, m : 16.69 Cb: 0.802 Cw: 0.84 Cp: 0.818 Engine Main engine: Sulzer Type: 6RTA48TB No: I Power, kW: 8203 Speed, rpm: 118 Fuel type: HFO Propeller No: 1 Diameter, m: 6.35 Type: fixed pitch P/D: 0.59 AE/AO: 0.55 No. of blades: 4 Speed, rpm: 118 D 10 20 co 30 40

(30)

point the ship is navigated for a preset time

interval At with a certain heading.

X_o

Departure point

Sub-sector angle AS2

tl

fi

tl

Sub-sector S2(k)

Great circle route from X0 to X

- -f

Great circles departing from X.0

Figure 13, Construction of isochrones (Source:

Hagiwara 11611)

This is done for a range of n headings. Doing so,

a number of positions is obtained that can be

reached by the ship at t. From each of these

reached positions, the procedure is repeated

resulting in possible arrival points after 2At.

Form these n arrival points a selection is made

by choosing in every sub-sector (as depicted

inFigure 13) that one having the largest great

circle distance from the starting point. This

selection

of

points defines the so-called

isochrone at t2. The points on the previous

isochrone from which the new found isochrone

points can be reached are memorised.

By repeating

this procedure,

a number of

isochrones and the possible tracks between them

are constructed between starting point and final

destination. As soon as the minimum distance

from the latest constructed isochrone is less than

the expected distance that can be covered

during

At hours, the ship is navigated from each of the

points on the final isochrone to the

destination

along a rhumb line. This method is particularly

suitable for minimum passage time

optimizations.

When it comes to minimum fuel optimization

there are

some restrictions

to the applied operational scenario:

For several reasons, an interesting scenario for

kite application is that of constant speed with

varying

thnist

(and

corresponding

varying propeller speed):

increasing the ship speed by auxiliary

kite propulsion will decrease the

operability and efficiency of the kite: a

'slow'

ship benefits more from kite

propulsion than a fast one due to the

more favourable apparent wind direction

for slower ships

varying speed may result

in earlier

arrival but since arrival time remains

uncertain due to uncertainty in predicted

wind conditions, the question is how

favourable this possible early arrival will

be

For these reasons, the constant speed scenario is

an attractive one which unfortunately cannot be

dealt

with by

the isochrone method: the isochrones

are known in

advance for

this

problem and the resulting possible tracks will

diverge from the great circle track until the final

isochrone and then converge to the destination.

The minimum time route is per definition the

shortest route between departure and destination

location. The minimum fuel route however may

not be found this way.

However, the scenario of constant propeller

speed is likely to benefit significantly from kite

propulsion as well. Both ship speed and fuel

consumption will be affected in a similar way

by kite application as was shown in the previous

section: the kite's driving force results in

both

increased speed and as a consequence earlier

arrival and less overall fuel consumption due to

less engine operation hours. Besides, the

fuel

consumption per hour decreases due to the kite

in this scenario as can be seen in Figure 11.

Environmental data

The environmental data to be used in

combination with the developed tool is provided

in so-called ww3-type grib files by the National

Oceanic and Atmospheric

Administration

(31)

of various environmental quantities such as

wind speed and wind direction, significant wave

height, period and direction. Records for 0 hours

forecast up to 180 hours forecast with 3 hours

intervals are provided in the grib files that are

updated with the latest forecasts every 6 hours.

Optimization vs. Simulation

The developed tool provides the possibility to

both optimize and simulate voyages in a

realistic way: when analyzing a certain voyage

the first step is to find the optimal route between

departure location and destination. This is done

as described above using only the latest grib file

available at the entered departure date and time.

Once the optimal route is found, simulation of

the journey for At hours (the time between the

isochrones) along the optimal course is carried

out, using only the 0 hours forecast values of

several grib files.

Considering the

location

reached after At hours as a new departure point,

the optimization process is repeated to find the

updated optimal route from the reached location

to the destination. This process can be repeated

until the destination is reached.

This way, the minimum At that can be applied

using the described grib files, is 6 hours.

An example of the graphical output of the first

optimization step

for an east bound North

Atlantic crossing from New York to the English

Channel is shown. Isochrones and

interconnecting tracks are depicted in black. The

wind field shown by the blue arrows

corresponds with the forecast for the last part of

the voyage.

The minimum

distance route

deviates from the great circle since the great

circle crosses some pieces of land (Nova Scotia

and New Foundland) in this example. As can be

seen the minimum fuel route hardly deviates

from the minimum distance route in this

case.

However this

will

_{strongly depend on the}

predicted wind conditions.

6 4

\\\\\A

\\\\\\\

.Nm,

\\\,,

\i"\\\\\,

\\''

\\\\\\% \\\,,

4""\"1

\\\ _//J.' _\\\4' 4j. 11

,

Figure 14, example of optimization of east bound North Atlantic crossing

Future work

Added resistance in waves has not been taken

into account in any of the presented results. It

will

_{be implemented by means of using}

quadratic RAO's for the added resistance in

waves resulting from calculations with existing

strip theory software.

Having developed the described tool, it will be

used to optimize and simulate

a sufficient

number of voyages for different ocean areas and

seasons, with and without kite propulsion and

compare those

to great circle (or shortest

distance) crossings in order to obtain statistics

on the long term benefits of applying kite

propulsion and the effect of route optimization

on it.

Acknowledgements

Ashish Jha and Anoop Mohan contributed to the

development of the presented simulation tool for

which they are gratefully acknowledged.

3

(32)

The authors thank Prof. Hideki Hagiwara for

providing parts of the original code of the

modified isochrone method.

REFERENCES

[li

Torvanger, A. What happens in sectors

outside of the EU ETS: Shipping and Aviation, CLIPORE seminar on

Critical Aspects of the Post 2012 EU

Climate Policy, Gothenburg 25th June

2008

Jager, D. A tool to estimate the economical effects of kite propulsion technology Bsc. Thesis work

commissioned by "Stichting de

Noordzee", the Netherlands 2008

Drela, M., Youngren, H., XFOIL 6.94

User Guide, MIT Dept. of

Aerodynamica and & Astrodynamics, Aerocraft, Inc, 2001

(http://raphael.mit.edu/xfoil/xfoil doc.t

xt)

Drela, M.,An analysis and design

system for low Reynold's number

airfoils, MIT Dept. of Aerodynamica

and & Astrodynamics,1989

Troen, I. and Petersen, E.L., European

Wind Atlas, Riso National Laboratory,

1989

Lingard, J. S., Ram-air Parachute

Design, Precision Aerial Delivery

Seminar, 13th AIAA Aerodynamic

Decelerator Systems Technology

Conference, Clearwater Beach,1995

Prakash, O., Aerodynamics and

Longitudinal Stability of

Parafoil/Payload System, Department

of Aerospace Engineering, Indian

Institute of Technology, Bombay, 2004

[8]

Naaijen P., Koster V., Dallinga R.P. On

the Power Savings by an Auxiliary

Kite

Propulsion System, ISP Volume 53

No.4 2006

[7]

Wellicome J.F., Wilkinson S., Ship

Propulsion Kites An Initial Study. Ship Science dept. report SSSU 19,

1984

Gernez E., Experimental and numerical

investigation of the Performance of a

Kite, MSc Report, Ship Science dept,

University of Southampton, 2006

Isherwood, R.M., Wind Resistance of

Merchant Ships, Trans. of the Royal

Institution of Naval Architects, 1973 W. Shi, D. Stapersma and H.

Grimmelius, "Simulation of the

influence of ship voyage profiles on

exhaust emissions", in Proc. IMECE

08, ASME Conference

J. Holtrop, G.G.J. Mennen, 1978, "A

statistical power prediction method",

International Shipbuilding Progress,

Vol. 25.

J. Holtrop, G.G.J. Mennen, 1982, "An

approximate power prediction

method", International Shipbuilding

Progress, Vol. 29.,

P. Brussen, 2006, "CO2-emissions of

various ship types, simulated in an

operational year profile", TNO report,

2006-D-R0262.

Hagiwara, Hideki: 1989, Weather

Routing of Sail Assisted Motor

Vessels, Phd. Thesis Delft University of

(33)

Proceedings of the 1st Workshop on "Development of Advanced Ship Support System Using Information Technology", Tokyo University of Marine Science and Technology, February 5, 2009, Tokyo, Japan