- ICSOT 2006 - Design, Construction and Operation of Natural Gas Carriers and Offshore Systems-The Royal Institution of Naval Architects (2006)

Sung Kil Nam, Wha Soo, Kim, Byeong Jae, Noh Hyung Cheol, Shin and Ick Hung, Choe, Hyundai Heavy Industries, Korea Selected Hydrodynamic Issues in Design of Large LNG Carriers Mirela Zal

ICSOT 2006: DESIGN, CONSTRUCTION & OPERATION OF NATURAL GAS CARRIERS &

OFFSHORE SYSTEMS

14 – 15 September 2006, Busan, Korea

© 2006: The Royal Institution of Naval Architects

The Institution is not, as a body, responsible for the

opinions expressed by the individual authors or

Ship Structural Design and Construction of Large LNG Carriers (LNGC’s) at Samsung Heavy Industries (SHI) – Malaysia International Shipping Corporation (MISC) Representative Perspectives

Mohd Fauzi Yaakob, International Shipping Corporation, Malaysia

Gas Carrier Development for an Expanding Market

Sverre Valsgård, Tom Klungseth Østvold, Olav Rognebakke, Eirik Byklum and Hans

O Sele ,Det Norske Veritas, Norway

The Propulsion of a 250000m³ LNG Ship

John Carlton, Lloyds Register, UK

Gas Combination Units for Dual Fuel Diesel / Electric or Slow Speed Diesel LNG Carriers

Damien Féger, Snecma Moteurs, France

The New Generation of LNG Carrier Machinery

Barend Thijssen, Wärtsilä, Finland

Trimariner Corporation’s LNG SeaTrain©, The First Truly-Modular LNG Shipping System

Stephen Henderson and Mary Lou Harrold, Trimariner Corporation, USA

LNG LiteTM – The Real Alternative to LNG

Bruce Hall, SeaOne Maritime Corp, USA

Ian Robinson, SeaTec Engineering, UK

Optimization of a Composite CNG Tank System

Thomas Plonski, Galal Galal, Gerhard Würsig and John Holland,

Germanischer Lloyd, Germany.

Design and Construction of Bilobe Cargo Tanks

Ivo Senjanovi, Smiljko Rudan and Vedran Slapniþar, University of Zagreb, Croatia

A study on Fatigue Management System for LNG Carriers Using Fatigue

Damage Sensor.

Toshiro Koiwa, Norio Yamamoto and Hirotsugu Dobashi, Nippon Kaiji Kyokai, Japan Osamu Muragishi, Kawasaki Heavy Industries Ltd., Japan and Yukichi Takaoka, Kawasaki Shipbuilding Corporation, Japan

CSA-2 Analysis of a 216k LNGc Membrance Carrier

Torbjørn Lindemark, Håvard N Austefjord Hans O Sele and Hang Sub Urm, Det Norske Veritas, Norway Keon Jong Lee, Hyundai Heavy Industries Co, Ltd, Korea, and T M Ha, Samsung Heavy Industries Co., Ltd, Korea

Extreme Sloshing and Whipping-Induced Pressures and Structural Response in Membrane LNG Tanks

Mateusz Graczyk, Torgeir Moan and MingKang Wu, Norwegian University of Science

and Technology (NTNU), Norway

The Parametric Study on the Response of Membrane Tanks in a Mark III Type LNG Carrier Using Fully Coupled Hydro-elastic Model

Sung Kil Nam, Wha Soo, Kim, Byeong Jae, Noh Hyung Cheol, Shin and Ick Hung, Choe, Hyundai Heavy Industries, Korea

Selected Hydrodynamic Issues in Design of Large LNG Carriers

Mirela Zalar, Sime Malenica and Louis Diebold, Bureau Veritas, France

Veritification of Numerical Methods applied to Sloshing Studies in Membrane Tanks of LNG Ships

Nagaraja Reddy Devalapalli and Dejan Radosavljevic, Lloyds’s Register, UK

Strength Assessment of Box Type LNG Containment System

Bo Wang, Jang Whan Kim, and Yung Shin, American Bureau of Shipping, USA

Dynamic Strength Characteristics of Membrane Type LNG Cargo Containment System

Jae Myung Lee, Jeom Kee Paik and Myung Hyun Kim, Pusan National University, Korea and Wha Soo Kim, Byeong Jae Noh and Ick Heung Choe, Hyundai Heavy

Industries Co, Ltd Korea

Numerical Analysis of 3-D Sloshing in Tanks of Membrane-Type LNG Carriers

M Arai and H S Makiyama, Yokohama National University, Japan

L Y Cheng, University of Sao Paulo, Brazil

A Kumano, Nippon Kaiji Kyokai, Japan

T Ando, National Maritime Research Institute, Japan

A Imakita, Mitsui Engineering & Shipbuilding Co, Ltd, Japan.

Authors’ Contact Details

SHIP STRUCTURAL DESIGN AND CONSTRUCTION OF LARGE LNG CARRIERS (LNGCS) AT SAMSUNG HEAVY INDUSTRIES – MISC BERHAD REPRESENTATIVE PERSPECTIVES

M F Yaakob, MISC Berhad, Malaysia

* Large LNG Carriers is generally defined as cargo tank capacity bigger than 100,000cbm whilst Very Large LNG Carriers is defined as cargo tank bigger than 200,000cbm

1 INTRODUCTION

The trend of current LNG newbuildings is that the cargo

capacity keeps increasing every year The shipyard

proposal keeps adding the numbers of tank capacity until

there is no limit to the membrane LNG carriers Prior to

this phenomenon, the Large LNG Carriers standard

designs are limited to 130,000cbm to 138,000cbm In

2004, Qatar Gas selected two designs proposed by

Hyundai Heavy Industries (HHI)/Samsung Heavy

Industries (SHI) Consortium and Daewoo Shipbuilding

and Marine Engineering (DSME) to build larger than

200,000cbm capacity LNG Carriers The contract of the

Very Large LNGCs clearly showed that there is no

limitation of what is coming to the industry

The construction of the LNG newbuildings around the

world will increase until the Qatar Gas acquisitions of

LNG ships settled sometime in 2012 Previously in the

past, any LNG newbuildings will be based on a fixed

charter contract between the Charterer and the Owner

However, recent trend of the LNG newbuildings is now

moving towards the spot charter market and speculative

in nature making the newbuildings slots for LNG very

tight among the LNGC capable shipyards

In order to become pro-active player in the LNG

transportation market and promoting high quality

standards in LNGC newbuildings, MISC would like to

share the experience gained during supervision of

newbuildings of Large LNGC MISC experience in

LNGC newbuildings is further augment by the fact that

most of the newbuildings ordered in the recent years are

based the membrane-type insulation rather than other

type of insulation like the Moss-type or independent

tank-type In 2004 alone, the big three shipyards in

Korea won almost 90 percent of the LNG tanker

contracts awarded (membrane-based insulation) (Herald Tribune, 2004) All of the fleet under MISC operation is based on the membrane type insulation, NO96 and Mark III systems

MISC, as the front-runner of the LNG Carrier ship owner/operator in the world, is keeping pace with the development constantly Six new LNGC project in Japan with capacity of 137,000cbm were contracted in 1998, designed by Mitsubishi Heavy Industries (MHI) whilst the construction were shared between MHI and Mitsui Engineering and Shipbuilding (MES) In 2003, another five LNGC were signed with SHI with capacity of 145,000cbm to be delivered between 2005 until 2007 Recently in 2004, another five LNG were signed with MHI with capacity of 152,300cbm to be delivered between 2007 and 2009 The trend of the LNGC capacity growing as the time goes by By the 2009, MISC will have about 29 LNG ships for operation worldwide

Figure 1: MISC First LNGC at SHI, SERI ALAM,

during Gas Trial

2 INTRINSIC FACTORS

2.1 CLASSIFICATION

The six LNG Carriers built in MHI/MES was classed by

LR under the notation LR +100A1, Liquefied Gas

Tanker, Ship Type 2G, Shipright (SDA, FDA, CM,

HCM, SERS), PCWBT, +LMC, UMS, IWS (Maximum

Vapor Pressure 25 kPaG at sea, Minimum cargo

temperature –163oC, Maximum cargo density 500

kg/m3) while the LNG Carriers built in SHI is classed by

BV under the notation of BV I + Hull, + Mach, Liquefied

Gas Carrier, Shiptype 2G (Membrane Tank, Maximum

Pressure 25kPaG and Minimum Temperature –163oC,

Specific Gravity 500 kg/m3), Unrestricted Navigation, +

VeriSTAR-HULL, +AUT-UMS, InWaterSurvey,

+SysNeq-1, Mon-Shaft, Mon-Hull

Both Classification Societies have their own concept of

approval the structural design of the large LNGCs The

Societies requirements on the local scantling are the

same where a simple program is able to calculate the

minimum requirements of each Class Then, in order to

minimize and optimize the steel structure the Yards will

pursue the matter using the direct calculation method

where a finite element modeling is performed for the

cargo hold area Both shipyards only performed

minimum 3-cargo tanks structural modeling of FEA

(minimum requirements by the Classification Societies)

Full structural modeling of the ship structure integrity

were not performed both shipyards because there is no

requirement for the full modeling under the Building

Specification

However, due to importance of the connection between

the cargo holds and the engine room and the forepeak

tank, SHI performed the detail connection analysis of the

structure as required BV Detail discussion of the FEA

approach by SHI will be discussed in detail later

2.2 COST

Based on the experience of MISC over the past 20 years,

the cost of the Large LNGC is going down from the early

deliveries of Large LNGC from French shipyard to the

Japanese shipyards During the early stage of the Large

LNGC construction in Europe the prices may reach more

than USD200 million per ship The price offered by

MHI/MES consortium was lower than the French

shipyard when MISC decided to build the next batch of

Large LNGC in Japan The price was lower than the

Japanese consortium when MISC decided to build the

Large LNGC in Korea specifically in Samsung Heavy

Industries (SHI) However, due to sudden requirements

from Charterers or Ship Owners for LNG tonnage the

price increases for the LNG Carriers in the recent years

The situation worsens when the slot for the construction

of the LNG Carriers diminishing rapidly with the high

requirements for new start up projects like QatarGas and

RasGas, NLNG Train 7 In order to compete with the

Korean shipyards, the Japanese shipyards need to increase the capacity and other modifications / improvements to maintain or lower the price of Large LNGC in Japan

Due to high volume of orders from the Ship Owners, the Korean shipyards like DSME and SHI, are able to offer prices lower than their Japanese counterparts In 2004 alone, the big three Korean shipbuilders grabbed about 17.3 million compensated gross tons (CGT) last year, the world’s largest, and far higher than No.2 Japan’s 12.2 million tons (Yonhapnews, 2004) At the end of 2004, South Korean yards had a combined order backlog of 35.4 million gross tonnage, the first time it has passed the 30-million-gross tonnage level (Tradewinds, 2005) 2.3 DESIGN

The design of the 145000cbm LNGC at SHI is a development from its standard design of 138000cbm LNGC As Owners are pushing the capacity higher and higher, SHI keep coming up with various designs for Owner consideration The designs vary from higher capacities to various type of propulsion system But one thing for sure, the capacities of the LNGCs are increasing

to lower the cost of LNG transported per shipment The bigger the capacity with the similar power, the capital cost is obviously lower For example, the cargo tank capacity of MISC Large LNGC ordered and built increasing from the Tenaga ships (130,000cbm) built in France, Puteri Satu ships (137,000cbm) in Japan, Seri

‘A’ series (145,000cbm) in Korea to the latest LNGCs signed at Mitsubishi Heavy Industries of 152,3000cbm for steam propulsion and of 157,000cbm for dual-fuel diesel electric propulsion

The size of Large LNGC will increase when Charterer or Ship Owner is looking at attractive or lower economics from newbuilding to operation in order to lower the cost

of transported per MMBTU For example, the cost of transported on the 200,000cbm LNGC is cheaper compared with 137,000cbm LNGC The cost is even lower when the Very Large LNGC is using slow speed diesel where the propulsion system is much more efficient than the steam propulsion system

As the size of the ships keep increasing, the Designers in the shipyards will try to come up with various design possibilities to get the basic idea of lower cost of LNG transported per shipment For whatever possible combination for the new designs by the shipyards, the limitations of Very Large LNGCs are the propulsion system, visibility, sloshing and import and export terminals

Since all of the LNG fleet in MISC is all steam driven, the limitation for the size is always the propulsion system Currently, the biggest steam turbine delivered

by the top turbine Maker is limited to lower than 30,000

kW Therefore, new generation of MISC LNG Carriers

need to be changed from traditional steam propulsion to

dual-fuel diesel electric propulsion to cater for increasing

capacity of LNG Carriers as required by the market

forces Otherwise MISC will have a limited market for

LNG transportation when the size of steam ships is

limited to 150,000cbm The shipyards on the other hand

had proposed diesel-electric propulsion system to cater

for the increase in the efficiency by achieving higher

capacity of the LNG transported per the same size of

ships

2.4 STRUCTURAL ARRANGEMENT

The structural arrangement of the LNGC is an evolution

of previous standard size of 138,000cbm LNGC by SHI

The structural arrangement is to remain the same but

rearranged to suit the current 145,000cbm For MISC

ships at SHI, the connection of the fore and aft structure

is analyzed in detail through finite element analysis

Some reinforcement was done to increase the strength of

the structure globally (Sohn, C.H., August 2003)

2.5 LONGITUDINAL GIRDERS

ARRANGEMENT

Detail review of the structure showed that the forepeak

alignment of the longitudinal girders is easy to perform

because of no foundation alignment required at the fore

area However, there is a slight mis-alignment of the

main girders of the cargo tanks with the engine room

girders

The outer most main girder (cruciform joint lower

hopper arrangement) for the cargo tanks is arranged at

15120mm off centerline while the sea chest girder is

arranged 14960mm off the centerline All other girders

are aligned between the cargo holds and the engine room

SHI did not explain the reason for the mis-alignment

arrangement, but willing to perform finite element study

to verify the connection (Yoon, K.S., 2004)

Based on the study, through various loading conditions

and static and dynamic conditions for the machineries,

the misalignment connection is lower than the allowable

stress but the opening for the seawater cross connection

pipe showed that the stress is above the allowable stress

limits by the Society Further improvement was

suggested by the Hull Structure Design Team to improve

the conditions Thicker plate was arranged in way of the

opening to compensate the loss of the structure and

reduce the stress concentration due to large opening

(1800mm diameter) at the girder

2.6 OTHER AREAS ARRANGEMENT

Other specific studies, other than the standard studies

requested by the Society, were also performed by the

Hull Structure Design Team to look into areas of

concern The studies include;

a Engine Room FE analysis (to study the effect of static and dynamic acceleration of the machineries and the ship hogging and sagging)

b Manifold Deck Connection FE analysis (to study the effect of effective connection between the ship structure and the manifold deck

c Manifold Deck Saddle Strength FE Analysis (to study the strength of the manifold deck under stress)

d Engine Room Girder FE Analysis (to study the misalignment of the 14960mm off CL with 15120mm off CL)

Figure 2: Results from the Engine Room FE Analysis at

Reduction Gear and Turbine Foundation 2.7 CARGO CONTAINMENT STRUCTURAL

DESIGN The ship structural arrangement is designed around the inner hull geometry for carrying the LNG cargo The structural design of the ship is fairly typical of a tanker design except for the notable differences in the detail joints inside the cargo holds and the cargo hold connection with the fore and aft section of the ship structure The hopper connection of the cargo tanks and the cofferdam foot arrangement is very important for the LNGC due to the fact LNGC allows zero tolerance of any possibility of crack or failure on the joints Not like any other tankers, GTT Mark III membrane LNGC cargo tank is insulated by mastic glue with Reinforced Polyurethane Form (R-PUF) and covered by Triplex (Continuous Strand Matt and aluminum) as the secondary barrier and finally covered by a Primary Membrane of 1.2mm SUS 304 corrugated membrane There is a notable difference between the NO96 and Mark III system where the maximum allowable stress for the LNGC by SHI is 185 MPa (GTT, 2001) while the allowable stress for LNGC built by MHI is 120 MPa (GTT, 1982) The difference of the allowable limit is because of the type of membrane system used for each ship The Mark III system allows higher limit because of the inherent properties of the double layers of R-PUF that

is separated by the Triplex While the NO96 system

limitation is the box design of the insulation system

Since both membrane systems are having lower

allowable stress than the ship structural allowable stress

by Bureau Veritas (BV, 2003), the Designers are playing

with delicate structure in order to ensure 40 years of

fatigue life for the cargo tanks as required by MISC The

highest von Mises stress designed for the No96 system

by MHI is only 118MPa while the highest stress

designed for the Mark III system by SHI is 186MPa

(coarse mesh study) at the vertical web of the cofferdam

foot opening Even with that condition BV still requires

SHI to increase the thickness of the vertical web from the

proposed design

2.8 HOPPER/COFFERDAM CONNECTIONS

Through MISC experience, the typical failures in the

structure normally occur at the top and bottom knuckle

due to discontinuity of stress flow vertically between the

top and bottom structure Therefore, MISC highlighted

the possible problems early to SHI during the design

process The problems of hopper connection were

acknowledged and experienced by SHI Hull Design and

Welding Research Institute Thus through various

discussions, SHI agreed to design extra leg length at the

cruciform joint and smooth grind the cruciform to

increase the fatigue life of the connection Smooth

grinding of the cofferdam and the longitudinal

connection was also applied to increase the fatigue life

and facilitate better stress transfer The effect of the

effect of the smooth grinding increased the fatigue life of

the connection up to 70 years

MISC also insisted for SHI to simulate the proposed full

penetration welding at the hopper area and smooth

grinding in order to provide a proper sequence for the

welders Based on the simulation at the Welding

Laboratory, a welding sequence was written by SHI to be

used by the welders during construction of the MISC

LNGC at SHI

2.9 FATIGUE ANALYSIS

The Building Specification of the MISC LNGC also

specified 40 years fatigue life for the ship structure based

on the North Atlantic wave data as specified in IACS

Recommendation 34 The fatigue analysis is performed

by the Society based on the Society’s propriety FEA

software, BV VeriSTAR Hull, with the dynamic loads at

10-8 probability level The results of the ship showed that

improvements were made to achieve the required fatigue

life stated by MISC The fatigue analysis performed by

Society is based on the damage ratio calculation of the

Society’s software The fatigue analysis results made by

the Society showed that the lowest fatigue life were at

the fwd and aft cofferdam foot, 41 and 44 years

respectively (Sohn, C.H., August 2003)

3 FE ANALYSIS

The Hull Strength Analysis required under the Society for the approval of the ship structure was also performed

by the Society However, SHI only requested the Society

to perform the minimal analysis as required by the Society, mid cargo hold structure analysis (Sohn, C.H., August 2003) and the cargo hold analysis and fore and aft connection (Sohn, C.H., November 2003)

Other than the structural design technology at the hopper connections, LNGC design lies in the connection between the fore and aft part of the ship where the stress transfer is high If the connection details are not carefully managed and arranged, the connection will create hotspots for the stress transfer thus promoting possible failure of the ship connection between the strong cargo hold area and the forepeak and engine room area The main concentration of the structure analysis is the connection between the cargo holds and the fore and aft sections Some modifications are required to improve the connection between cargo holds and the fore and aft sections The strengthening of the connections includes the following

a Reinforcement of Cofferdam Vertical Bulkhead for Cargo Hold 1 (at CL)

b Reinforcement of Cofferdam Vertical Bulkhead for Cargo Hold 1 (2700 off CL)

c Reinforcement of Cofferdam Vertical Bulkhead for Cargo Hold 1 (5370 off CL)

d Reinforcement at Cofferdam Vertical Bulkhead for Cargo Hold 1 (10710 off CL)

e Reinforcement of Stringer 2 (145000 AB) inside the Cofferdam no.5 for Cargo Hold 4

f Reinforcement of Stringer 1 (22790 AB) inside Fwd Pump Room for Cargo Hold 1

SHI also perform other specific local FEA to cater for local strength analysis besides the global strength analysis as required by the Society The specific FEAs for the ship structural design like the Cargo Hose Handling Crane FEA, The Provision Crane FEA, Sunken Bit FEA and forward mooring/anchor windlass foundation beam analysis

4 CONSTRUCTION

The number of ships constructed in SHI keeps increasing

as the year goes by In 2003, the total ships delivered by SHI were only 43, while the total ships delivered in 2004 were 50 Before the recent spat of increasing trend of LNG newbuildings, the total number of LNGC ships normally built by SHI is around 4 per year However, in

2005 SHI is targeting to build around 9 LNGC to take advantage of the current drive of LNGC by Owners around the world The number of LNGC to be built in SHI will also keep increasing as SHI is planning to bring

in second floating dock for the docking of Very Large

LNGC signed under the QatarGas project

The construction of the LNG ships at SHI is simplified

through proper planning of the hull construction and

relieving the chock point of the hull construction –

drydocks SHI will design and plan the construction of

the ships so that a fixed and regime timeframe will be

observed at the drydocks So far, Dock 1 Hull Erection

team already conversed with the LNG newbuildings and

can easily build, erect, weld and launch a typical

145,000m3 LNG ships within 45 working days The

main target for the SHI Hull Erection Team is always

and has been the Keel Laying date and the Launching

date for every ship erected inside Dock 1

5 BLOCK DIVISION

The ship is roughly divided into 265 assembly blocks including the 411 sub-assembly blocks The ship is also divided into several main block location i.e B-block for the double bottom ballast,

S-block for the wing ballast, T-block for the cofferdam, F-block for the forward ballast and forepeak, DS-block for the trunk passageway, DC-block for the inner trunk deck, E-block for the engine room, A-block for the stern section, M-block for the superstructure and funnel casing SHI went one step further in freeing the dock time by making mega blocks (3000 – 3500tonnes) around the shipyards and Sub-Contractors

Table 1: Ship block erection history for Hn1502 in the Dock 1 Practically, the shipyard is able to erect complete ship

for launching/commissioning within 40 working days

For MISC Project Hn1502 and 1503, there were 4 mega

blocks consist of the Tank 3 double bottom to the 2nd

stringer, engine room double bottom up to the 2nd Flat,

the whole accommodation block and the whole funnel

casing For the MISC Project Hn1589, 1590 and 1591,

SHI will further increase the number of the mega blocks

to minimum of six with additional two mega blocks for

the engine room from the 2nd Flat up to the Main Deck

and Tank 2 double bottom to the 2nd stringer

Figure 3 Lifting of Tank3 into Dock 1

Figure 4: Lifting of Engine Room into Dock 1

The block division of the ship is common in the

shipbuilding industry nowadays after the ingenuity of the

Japanese shipbuilder However, SHI is moving one step

further by simplifying the system to facilitate dock

erection time Combined with the minimum of 6-8 Hull

Sub-Contractors around the shipyard and China, and

better arrangement of the block division, SHI could

simply shave off 4-6 months from the normal

shipbuilding process

The Hull Sub Contractors are located within

20-60minutes driving from the Yard complete with the

gantry cranes allow the block to be fabricated bigger and

bigger This will make the block erection in the shipyard

simpler For example, Anjung Hull Fabrication area that

is located about 60 minutes from SHI consists of four

Sub-Contractors namely, Sung Dong, Dong Yang,

Kastech and Gaya Sung Dong and Dong Yang already

expanded their facilities to go in hand with SHI corporate agenda to become a Global Leader by 2010

Figure 5: Hull erection method is even simpler

Figure 6: Simple block division through various

improvement The block division is designed so that all complicated sections and detail construction sections like the lower hopper joints will be done during the block stage This strategy will limit the Hull Erection Team to concentrate

on the straight joints A typical erection of the wing ballast tank is taking about 30 minutes where the actual erection time is about 5 minutes In fact, the erection team needs more time on bringing the block in place from the turning over area and tacking the block in place This concept is possible to be put in placed because of the well-thought out block division to suit the facilities and reduce the time in dock

Due to simpler erection at dock, the accuracy of the hull erection is very high The connection between a typical transverse bulkhead is about 5-8mm and the side shell connection is almost perfect match even when the blocks are constructed about 45% outside of SHI

Figure 7: High accuracy of erection

Figure 8: Gap between the transverse bulkhead

6 CONSTRUCTION MANAGEMENT

On average, SHI needed only 6 months to build the

blocks for the LNGC and another two months in the

Erection period inside the Dock SHI could practically

complete the hull construction of the LNGC within 8

months a feat comparably very efficient compared with

other Builders around the world like Japan or France

The main reason for the fast construction of the hull is

because SHI is good in managing the Sub-Contractors

(inside and outside the Yard) especially when the

construction of the ships largely dependent on the

performance of the Sub-Contractors The project

management works on the set target dates where it is the

normal practice of current shipbuilding practice Then,

the Planning department will work backward and

distribute the construction of the blocks within the

available contractors

For MISC Project, the construction of the blocks for the

outside contractors is about 45-47% Among the blocks

that were built inside Yard, 52 blocks built the

Sub-Contractors and 371 blocks built by the SHI workers

The quality of the construction of the blocks between the

inside and outside Sub Contractors is comparatively the

same However, the amount of re-inspection (Owner

Confirmation) of repair work due to defect noted (based

on Samsung Shipbuilding Quality Standard) is higher for outside fabrication than the inside fabrication In short, the quality the block inside the Yard is better than the quality of the block outside the Yard

1502

Manufacturer Blocks %

Outside SC 254 37.5% Total 677

1590

Manufacturer Blocks %

Outside SC 139 20.8% Total 669

1591

Manufacturer Blocks %

Outside SC 252 37.7% Total 668 Table 2: Comparison Table for hull construction inside SHI and outside SHI

for Hn1502, Hn1503, Hn1589, Hn1590 and Hn1591

7 SHARED LESSONS

The main success criterion for any project management

is planning MISC planned for the Site Team is set up in

December 2003 From the creation of the core Site

Team, further planning was made to mobilize the rest of

the member prior to the Steel Cutting date for the first

ship in February 2004 Similar to any new ship

construction project, the design and the construction of

the first ship were overlapped for a year with the design

started from May 2003 Therefore, the engineers were

collated from various disciplines mainly from the Project

Department from Technical Services and Fleet

Operations Departments

The construction of the Large LNGCs is the second

project by MISC in SHI, after 105,000 Aframax Crude

Oil Tanker and the project also marked the first LNG

construction in Korea So far all of the LNGCs operated

by MISC were built in France and Japan Currently,

based on the statistics, most of the LNGC’s Owners,

experienced or new, opted to build the ship in Korea

rather than elsewhere (Herald Tribune, 2005) This is

due to the competitive cost offered by the Korean Yards

compared anywhere around the world

Most Owners in SHI LNG series are all well-known and

established Owners like MISC, BG, AP Moeller and 3

Japanese Consortium lead by MOL (3J/4J)

Furthermore, the buyers normally would purchase series

of vessel rather than 1 or 2 sister vessels For example,

between MISC and BG, there are 10-12 LNGCs signed under the contract with SHI Starting from 2004, SHI expected to increase the productivity to produce the LNGCs from 4 per year to 9 per year Increase in productivity means that the project management team assigned to manage, appraise and monitor the contract in SHI will face daunting tasks Therefore, MISC implemented three-prong strategy to maximize the resources and still maintain the quality of the ships produced at SHI

8 PROJECT MANAGEMENT

IMPLEMENTATION AND EXECUTION

8.1 PROPER PLANNING The overlapping of design process and the construction

is further aggravated with the overlapped of construction

of series of ships For MISC LNG project in SHI, the construction of five ships is spread over four years from

2004 until 2007 If it is not planned properly, there will

be lack of resources during the peak load of block inspections and the Cargo Containment System inspection together with the Machinery System commissioning

Figure 9: Construction program for 5 MISC LNGC at SHI

MISC 145000m3 LNG NEWBUILDING PROJECT SCHEDULE

8.2 PROPER TOOLS

Various tools were used during the implementation of the

project to facilitate the design and construction of the

LNG Carriers at SHI The tools used by the Site Team

include the Plan Approval Database that keeps history

and status of the drawing approval process with SHI and

the Customer Remarks Database that keeps MISC

remarks on non-compliance or comment during

construction of the LNGC

The Plan Approval database is set up and maintained by

MISC Site Office keeps the history and status of the

drawings The database will give the record of previous

comments and status of overall design approval stage

The Customer Remarks database is set up to record

comments and non-compliance to the actual production

In order to streamline with the SHI QA Management

System, the Customer Remarks database is included in

the SHI FOCUS Web-based progress and monitoring

system online monitoring and solutions The online

system allows all parties including the Head Office in

Kuala Lumpur to monitor the progress, inspection and all

Customer Remarks realtime

8.3 PROJECT MANAGEMENT TEAM

As mentioned earlier, the construction of the ships at SHI

is roughly divided into two, inside and outside of the Yard Compounded to the problem of outside inspection

is the location of the Sub Contractors that are located within 20-60 minutes driving radius from SHI Therefore, in order to maximize and improve the inspection, the Senior Site Manager employs and encourages the patrol from Bureau Veritas Surveyors and GTT Surveyors besides the Site Team Surveyors This concept will certainly increase the level of awareness by the production and improve quality control at Site

Since MISC already involved in so many types of ships and projects all over the world especially Korea, Japan and France, the level of interpersonal skill is very high Interpersonal skill in the newbuildings project management is important because short construction period and details of changes to suit Owner’s specific requirements Since most of the requirements cannot be spelled out in the Building Specifications and because of the fast turnaround in commercial shipbuilding, comments cannot be addressed during short approval period Therefore, interpersonal skill is very important factor in accomplishing objectives without compromising quality and cost

Overall Quality Control Hierarchy

EM

CONCERNS OBJECTIVES SOLUTION

Structural Design

1 Cruciform Joint To design of the cruciform joint

including the offset of the median line of the hopper and the longitudinal girders

The median line opted for the cruciform joint is zero between the hopper and the longitudinal girders Additional pass is also designed to provide enough leg length for the smooth grinding application at the cruciform joint The application resulted in prolong fatigue life

2 Main Structural

Girder Alignment

To achieve better alignment of the main longitudinal girders between the cargo holds and the engine room structure for better stress transfer

SHI agreed to perform the misalignment of the main girders to confirm the adequacy of the stress as required by BV The results showed the stress within acceptable level by BV but the opening of the crossover seawater supply along the girder need to be strengthened due to loss of structure

SHI agreed with the proposed hold and witness points to achieve the production quality as per approved design by MISC

2 Cruciform Welding

Sequence

To check and confirm the welding sequence of the cruciform joint in order to achieve the full penetration welding as designed and approved

SHI agreed to perform the welding test piece for the cruciform joint to simulate the design of the cruciform joints The test piece showed that SHI welder could only achieve the full penetration with sequence different from the standard practice A new set of the welding sequence was distributed to the production for MISC Project

3 Quality Inspection

Tools

To provide the correct tools for the inspection of the cruciform joint during fit-up and final inspection

For the hold and witness point of the cruciform joint, templates were made and distributed to the SHI Production and Sub-Contractors The templates facilitate the hopper angle and the median line of the longitudinal girders with the hopper plates

Table 3 Summary of Lesson Learnt at SHI

9 DESIGN AND CONSTRUCTION

PERTINENT LESSONS AT SHI

After having a dedicated Project Management Team,

MISC is able to tackle and resolve the design and

construction issues directly with SHI in order to

achieve high quality standards required prior to

entering the esteemed fleet of MISC vessels The

efforts are collective among the Team Members in

order to deliver the vessels as per Specification agreed

between MISC and SHI The major hurdles in

delivering high quality products to the Fleet Operations

in MISC after battling through the design process with

SHI are the construction monitoring and system

commissioning – quality inspection Therefore, the

best and only option for the Project Management Team

is to follow through the design discussions and

solutions to the construction of the ships in SHI

diligently

Below are some of the extracts of notable lessons for the ship structural design and construction that MISC learned from the construction of the LNGC at SHI

10 CONCLUSIONS

The management of the design and construction of the LNGCs at SHI really opened the eye of any LNG Owners due to the fact that SHI is building the LNGCs faster than ever Therefore, Owners shall react positively by adopting the new style of shorter and faster LNGC construction program where rigorous design process and continuous construction patrols are recommended There are various ways that are possible to suit this new trend of LNGC newbuildings but the most important strategy lies on the specific interest of the company

Due to recent high interest in LNGC long-term charter,

a lot of new Owners try to jump into the bandwagon of

LNG business These new comers will mainly follow

the proposal from the shipyards with little input or

without any major requirements for the design and

construction of the LNGC Lack of control in design

and construction will jeopardize the maintainability and

availability of the LNGC during in-service operations

Off charter hire during operation other than scheduled

dry-docking will squeeze the margins for this type of

operation Only then, the survival of the fittest will

ensure continuous and profitable operations in the LNG

business

The interesting development in the LNGC

newbuildings will never last forever, as the gas markets

will eventually become saturated Even now, there are

few LNGCs already idle around the world waiting for

cargoes Therefore, careful planned strategy in terms of

newbuildings commitment and existing LNGC

maintenance and reliability play major roles to ensure

the success in the LNG business

11 ACKNOWLEDGEMENTS

The views, comments and contributions from the MISC

Bhd Management, MISC LNG Site Team at SHI,

Samsung Heavy Industries (SHI) Management and SHI

Project Management & Design Teams are gratefully

acknowledged

12 REFERENCES

1 Bureau Veritas, “Rules for Steel Ships”, 2003

2 GazTransport and Technigaz (GTT), “Hull

Design and Tank Dimension”, GTT External

Document 1187, 2001

3 GazTransport and Technigaz (GTT), “Cargo

Tanks Arrangement Dimensions and Filling

Ratios Hull Scantling Requirements”, GTT

External Document NO DG 33, 1982

4 Herald Tribune, Various Articles, 2005

5 Sohn, C.H., “MISC 145000m3 Fore and Aft

Cargo Hold VeriSTAR Report”, 25 November

2005

6 Sohn, C.H., “MISC 145000m3 VeriSTAR

Report”, Bureau Veritas, 25 August 2003

7 Tradewinds, Weekly News Internet

Publication, 4 Feb 2005

8 Yonhapnews, Various Articles, 2005

9 Yoon, K.S., “Structure Misalignment at 14960

Girder and Frame 72”, SHI, 2004

13 AUTHOR’S BIOGRAPHY

Ir Mohd Fauzi Yaakob is presently a Senior Naval

Architect at the Technical Services Department, MISC Berhad Currently attached at Newbuilding Program for LNGC at Samsung Heavy Industries (SHI) since

2003 A graduate, in Bachelor of Science Engineering

in Naval Architecture and Marine Engineering, from University of Michigan joined Grand Banks Yachts in

1994 as an Engineer and later joined Penang Shipbuilding and Construction as a Naval Architect Attached to Wavemaster International for several years

as a Naval Architect and concentrated in design and development of fast passenger and car ferries Headed the Platform Systems Section for the Project Management Team for the design and construction of the Royal Malaysian Navy Patrol Vessel Project at Blohm+Voss Hamburg Experienced in new hull shape, structure, material and general arrangement

GAS CARRIER DEVELOPMENT FOR AN EXPANDING MARKET

S Valsgård, T K Østvold, O Rognebakke, E Byklum, and H O Sele, Det Norske Veritas, Norway

SUMMARY

The paper describes the work carried out in DNV to meet the new challenges facing the Gas Carrier industry, emphasising sloshing loads and tank system strength for normal tank fillings as well as reduced tank filling operations of membrane type LNG carriers The current applicability of Computational Fluid Dynamics (CFD) computer codes in sloshing analysis is discussed Results from scaling of model tests results between different model scales are shown and

it is concluded that a full scale sloshing impact measurement campaign is necessary to better understand the model scaling issue Short-term expected extreme as estimate for lifetime expected extreme sloshing loads is discussed and some remarks are given on sloshing loads at low filling ballast operation as compared to high filling full load operation Due to the current shortcomings in the CFD analysis tools DNV has concluded that sloshing load determination has to

be based on model testing As uncertainties still exists in the determination of absolute values of sloshing impact loads, a comparative approach has been selected for the containment assessment procedure in the new DNV guideline on

“Sloshing Analyses of LNG Membrane Tanks”

The status of the emerging CNG shipping industry is outlined Work is in progress for establishing a common basis for steel tank system design Several CNG proponents are working with prototype testing of tank system designs The main results from a specific full scale prototype testing campaign is reviewed highlighting fatigue testing, burst testing and live gas cool-down testing of a particular steel tank concept

1 INTRODUCTION

The world use of natural gas is increasing For long

distance seaborne transportation of natural gas LNG

represents today the most efficient commercial

alternative, but economically competitive systems for

smaller volumes and shorter distance trades like

Compressed Natural Gas (CNG) are emerging

1.1 GAS CARRIER DEVELOPMENT

From the very beginning of the gas carrier industry great

care has been taken to include all relevant failure modes

in the design of the tank systems and fatigue, buckling

and sloshing loads have been important design

parameters

The seaborne transport of liquefied gases in bulk is older

than often realised Already in 1949 the first dedicated

liquefied gas carrier was delivered with DNV class This

was a vessel with fully pressurised cargo tanks for

transport of LPG/Ammonia The vessel, named Herøya,

had vertical cylindrical tanks and was built at the Horten

Navy Shipyard in Norway DNV, therefore, became

involved very early in the setting of safety standards, and

was in 1962 the first classification society to publish

comprehensive rules for gas carriers

A research team on LNG was established in DNV in

1959 A membrane tank system was developed and

tested successfully in 1962 The system used double

corrugated aluminium sheets as the primary barrier This

system was later taken over and further developed by

Technigaz in France

The Moss spherical tank design was developed by the Kvaerner Group in Norway during 1969-1972 Basic design criteria for type B tanks were formulated by DNV

in the 1972 rules In order to confirm compliance with the design criteria comprehensive R&D programmes were carried out in DNV, e.g sloshing loads from liquid movement inside the cargo tanks, crack propagation, fatigue characteristics and buckling strength

The idea of shipping gas on keel without a costly liquefaction process is equally old as the LNG industry, but have until recently been no success due to the heavy gas containment systems if the tanks (cargo cylinders) were to be designed according to conventional pressure vessel codes or the international Gas Code (IGC) This leads to heavy containments systems with virtually no lifting capacity left for cargo unless unreasonably large and costly ships are to be used

Most CNG concepts apply high pressure (130-250 bars)

in a semi-chilled or at ambient temperature condition in order to keep the gas in a gaseous state with basically no liquid hydrate fall-out CNG tanks are mostly based on the use of cylindrical bottles or pipes with diameters up

to 48 inch being designed according to modern limit state

pipeline or pressure vessel codes For such tanks fatigue becomes the driving design parameter, ref [8]

1.2 NEW OPERATIONAL CHALLENGES The consumption of natural gas is projected to increase

by nearly 70% between 2002 and 2025 [1] and the market for seaborne gas transport is increasing at an

unprecedented pace The latter is characterised by rapid

increase in the carrier fleet, spot trading, speculative

ordering, increased carrier size, a move away from the

traditional one propeller steam plant towards diesel

propulsion and two propellers, more cross Atlantic

trading, partial load trading (milk runs) and an emerging

market for cold climate (Artic) operation Fatigue

considerations and tank sloshing loads are becoming

more important design parameters

Offshore receiving/storage terminals and regasification

and discharge terminals will in some parts of the world

be the preferred future option due to safety

considerations and environmental concerns Floating

units for receiving, storage, regasification and export

(FSRUs) of natural gas as well as units for offshore

production (FPSOs) are emerging markets For these new

applications safe operation with partial tank fillings has

to be carefully studied on a case to case basis Sloshing

loads and tank system strength are therefore key issues in

the design and operation of such systems

Seaborne LNG transport has historically been a high

standard, low accident operation Damage statistics from

DNV in-house studies indicate an average accident rate

in the range 30-80% lower than for average shipping

operations It is a challenge for everyone involved to

maintain this favourable situation in order to further

develop the industry

The larger sizes of carriers and the new operational

profiles outlined above make relying on past experience

for structural performance of the vessel hulls and

containment systems rather uncertain Hence, the use of

state-of-the-art design for ultimate strength and fatigue

will be essential for safe and trouble free operation

2 SLOSHING LOADS AND STRENGTH

Sloshing can induce various types of loads Motions

and/or more rapidly varying motions, causing higher

accelerations, induce dynamic effects The pressure

fields inside the tanks can still be described by smoothly

varying pressure distribution functions and the structural

response can be calculated in a quasi-static manner In

case of more frequency content around sloshing

resonance the fluid behaviour becomes violent, causing

breaking waves and high velocities of the fluid surface

In this case the fluid can cause impact loads on the

containment system These loads can be characterised by

a high pressure load with short duration acting on a

limited area

Violent sloshing can be characterised by various fluid

flow phenomena illustrated in Figure 1 In the high

filling range >90%H (H denoting the tank height) the

impacts typically occur on the tank roof at the connection

with the transverse bulkheads Typically a ‘flat’ fluid

surface hits the roof at high velocity causing the impact

For fillings in the range of ~60% to ~80% the largest impacts occur in the corners and knuckles of the chamfer These impacts can be caused by run-ups against the longitudinal or transverse bulkheads or by a ‘flat’ fluid surface impact

For fillings in the range of ~20%H to ~40%H the largest impacts occur at the longitudinal and transverse bulkheads due to breaking waves, Figure 2

A characteristic phenomenon, which can occur at lower

fillings is the so-called hydraulic jump or bore This

wave phenomenon is characterised by a ‘jump’ in the free surface level, which travels at high speed, and can cause a large impact Sloshing model experiments are required in order to assess the violent sloshing causing impact loads

The sloshing loads vary in size, duration and load area

In addition, the containment system and hull structure have different failure modes Consequently, a careful analysis of the structural response and strength needs to

be conducted for the various loads to assess the structural integrity

Figure 1: Typical high-filling (>90%H) impact in

near head sea conditions

Figure 2: Schematic illustration of a hydraulic jump

or hydraulic bore

2.1 LNG CARRIER DEVELOPMENTS

Many of the current developments in the LNG shipping

industry affect ship classification R&D is a key element

to develop the required competence to adapt

classification rules and guidelines to new ship designs

and operations DNV therefore applies a significant

amount of resources to R&D and has defined a specific

R&D portfolio for gas carriers - LNG as well as CNG

Some of the key elements in these efforts related to gas

carriers are:

x LNG sloshing in membrane LNG tanks

x Alternative propulsion arrangements

x Vibrations

x Hull fatigue

x Operation in cold climate

The last three items are not only related to gas carriers,

but are of major importance to all types of ships The two

last items are organised in separate R&D programmes on

“Hull Loads and Strength” and “Cold Climate”

respectively

Three Class Notes are under development and are

expected to be issued in 2006 One is focussing on the

hull and tank support design of membrane tankers,

excluding the containment system, the second is focussed

on sloshing loads and strength of membrane tanks [3]

and the third one is concerned with the design and

analysis of the hull and tank system of spherical type

LNG carriers

Industry co-operations complement the pure internal

R&D work Joint Industry Projects with yards and ship

owners are important for DNV in order to improve

knowledge sharing and competence exchange

Most attention in the LNG R&D portfolio has been paid

to the first item in the list - LNG sloshing in membrane tanks This R&D work is divided into four projects:

The first two listed projects are primarily focused on competence development in order to support DNV classification All the knowledge and competence gained are used to develop a load and structural strength assessment scheme This forms the basis for a dedicated Class Note (Guideline) on sloshing in membrane LNG tanks [3]

Sloshing is a highly complex phenomenon and despite the huge R&D efforts some aspects are still under discussion or difficult to put down in a practical guideline Based on this and as a response to market requests a sloshing full-scale measurement campaign has been designed by DNV This measurement campaign is intended to provide a validation database for sloshing loads and structural responses

2.2 MODEL TESTS OR COMPUTER

SIMULATIONS?

Computational fluid dynamics (CFD) software has found many engineering applications enabling designers to simulate fluid flow, heat and mass transfer, and a host of related phenomena involving turbulent, reacting, and multiphase flow Hence, CFD has been considered a potential tool for sloshing impact analyses in LNG tanks and much effort has been made to adapt CFD tools for such applications

DNV is using the ComFLOW CFD software developed

by University of Groningen in Holland [4], and has evaluated the programme for simulation of sloshing phenomena in LNG tanks The present version is simulating one-phase flow only, but does have facilities for random 6 d.o.f motion time series input to the programme Presently work is underway in a Joint Industry Programme (JIP) aiming at implementing two-phase capabilities which will be essential for possible future applications to LNG sloshing phenomena In this connection large 1:10 scale sloshing tests of a transverse cross-sectional cut of an LNG membrane tank has been carried out in the DNV laboratories at Høvik using water and air at atmospheric pressure in order to provide a verification data base for further development, Figure 4 Figure 3: Integration of sloshing load and response

projects for membrane LNG carriers

Previous model tests [7] have shown that ullage gas

depressurisation according to the linear Froude scaling

law and tests with heavy gas aiming to have a correct gas

to liquid density ratio gives quite different sloshing test

results, the latter giving the lower measured pressures,

Figure 5 Hence, for approaching an absolute sloshing

impact load assessment with a CFD code a two-phase

simulation capability for simulating both the liquid and

the gas is essential

Another challenge is the correct mathematical modelling

of the gas/liquid interface In an LNG tank continuous

phase transition from liquid to gas (boil-off) takes place

due to the heat influx through the insulation system The

motion of the LNG creates gas bubbles and turbulence

effect on the gas/liquid interface Also, when the liquid

hits the sharp knuckles and corners of the tank wall

cushioning effects may occur caused by entrapped gas

and the increased flexibility of the fluid/bubble mixture

Being able to capture the local compressibility effect of

the fluid/bubble mixture which varies in time and space

is therefore important for determining cushioning effects

In order to simulate these effects very small meshes have

to be used (in the order of 1 mm) This is prohibitive for

CFD calculations

Harmonic model test have shown that sloshing impact

pressures is a highly stochastic process Consecutive tests

with the same harmonic input signal do not give the same

result and statistical treatment of the results can be done

in the same way as for tests with randomly generated

input signals CFD codes may not have implemented

such facilities in their solution schemes – the same input

gives the same results

For these reasons Det Norske Veritas has concluded that

for the time being the only viable and practical approach

to determining sloshing loads in LNG tank systems is to

perform model tests We are then faced with the pressure

scaling issue which has been a continuous question mark

in the LNG industry This will be discussed later in the

paper

However, liquid motions can be modelled in a CFD

programme using much coarser meshes than necessary

for impact pressures Hence, pipe tower loads, i.e drag forces from liquid motions, as well as inertia load effects can be modelled adequately with today’s CFD codes This also means that simulation of global sloshing loads for sloshing-ship motion coupling is possible

2.3 SCALING OF MODEL TEST RESULTS For the sloshing experiments there has been quite some discussion about the scaling of the impact pressure and the properties of the ullage gas The Froude scaling is a well known scaling law used in fluid mechanics This scaling law is valid for inertia dominated fluid behaviour But for compressibility, surface tension & viscous effects other scaling laws apply DNV has studied this further and have earlier (1970s/1980s) recommended a reduction

in the ullage gas pressure according to linear scaling The ullage gas pressure was reduced linearly with the same factor as the geometric scale factor This was based on experimental sloshing investigations with water and air which showed that the measured extreme pressures are highly sensitive to the variation of ullage pressure when depressurised to lower than approximately 100 mbar [5] This also had implications on the selection of model scale as the uncertainties increase with reduced model scale

The validity of this was checked by DNV in 2004 when impact pressure result for roof impact at 90% filling level were compared between two tanks at different scale A range of different ullage gas densities and pressures were tested One tank had a scale-ratio of 1/70 and the other a scale-ratio of 1/20

The following effects were observed:

x The impact pressure was dependent on the ullage gas density and gas pressure

x A reduction in the ullage gas density was needed in the small tank by a factor close to the scale-ratio to predict the large tank impact pressure

Figure 5: Froude scaling of sloshing impact

pressures Figure 4: Test rig with 1:10 scale model for the

ComFLOW 2-phase JIP

x The effect was important for single sensor loads and

not that pronounced for a larger area (average

pressure for a cluster of sensors)

The implications of these issues are that questions still

may be asked if single sensor measurements might be

non-conservative if they are scaled directly by Froude

scale without modifying the ullage gas density Figure 5

illustrates the observations The results from the large

(1/20 scale) tank and the small (1/70 scale) tank are

compared in the same figure Both the horizontal and

vertical axes are scaled linearly Then the graphs match

If the ullage pressure in the small tank is scaled down

according to the scale ratio, as recommended by DNV

from earlier work, the impact pressure follows

reasonably Froude scaling

An example:

The vertical dashed arrows show the points where the air

at 1 atmosphere (atm) is tested The dimensionless

impact pressure in the small tank with air at 1 atm is 3.0

If the same gas condition (air 1 atm) is tested in the large

tank, the result is 4.6 This leads to an under prediction of

the large tank impact pressure by the small tank impact

pressure tests by the ratio 4.6/3.0 § 1.5

However, the following remarks should be made:

x The amount of data is limited

x There is an uncertainty in the motion due to the large

rig and small tank (1/70 scale)

x A validation with full scale measurement should be

carried out

x For some of the points, both the ullage gas density

and pressure is changed Thus it is not only one

parameter which is changed

A final conclusion cannot be drawn as the quality of the

tests is not found sufficient However, the various test

cases indicate a trend/direction, which seems to confirm

the recommendation given previously by DNV, [5]

2.4 FULL SCALE MEASUREMENT CAMPAIGN

As concluded above a full scale measurement campaign

onboard a membrane LNG carrier need to be carried out

to be able to get a better background for understanding

the scaling issue

Hence, DNV has together with industry partners

designed a full-scale sloshing measurement program in

order to obtain a full-scale sloshing measurement

validation database

The objectives of a full-scale measurement program are:

x Development of a full scale LNG sloshing

measurement system

x Obtain full-scale validation data

x Validate sloshing assessment procedures

Of prime interest is of course the measurement of sloshing pressures Traditional pressure sensors cannot

be used inside the containment system; hence a new sensor has been developed and qualified The newly developed sensor is based on fibre-optic measurement technique and is mounted in the insulation system under the primary membrane

The sensor has been developed by a Norwegian supplier

of fibre optical hull monitoring systems with assistance from DNV A test program consisting of static and dynamic functional tests has been carried out in a pressure test tank with a steel membrane between the sensor and the pressure transmitting liquid The sensor is capable of measuring dynamic responses with rise time down to 0.5 ms and has a working pressure range up to

40 bars The sensor has been verified for use with both the No 96 system and the Mk III system

The sensor has been qualified and approved by DNV for mounting onboard LNG carriers and will be connected to

a commercial fibre optical hull monitoring system A typical installation lay-out will comprise at least a set of instrumented containment boxes/insulation panels inside tank no 2 at positions likely to encounter the highest sloshing loads

The main objective with a full-scale measurement program is to obtain validation data It is therefore of vital importance to integrate full-scale measurements of sloshing load and structural response assessments Hence, simultaneous measurements of environmental data, ship motions and strains in the supporting insulation system and the supporting hull structure will be carried out to complement the sloshing pressure measurements

A measurement campaign was initially agreed upon between DSME, DNV, Bergesen Worldwide Gas ASA, Golar Management and STASCO in August 2005 The initiative was well received in the market and several other players in the LNG industry have shown interest in participating

2.5 LONG TERM AND SHORT TERM LOADS

In standard wave load analysis a wave scatter diagram is used to describe the long-term wave environment Using linear response transfer functions a complete wave scatter-diagram can be assessed and the long-term response amplitude distribution can be calculated From this distribution, the lifetime expected extremes can be determined, e.g corresponding to 108 wave encounters

in North Atlantic environment as given in the IGC [10]

or to a certain return period and probability level

Another approach is to determine the most critical sea state from the scatter-diagram and assess only the short-term statistics for that sea state For linear responses these typically give short-term expected extremes some 10% to 20% lower than the long-term expected extreme

However, for nonlinear responses the response behaviour

may be characterized by response amplitude distributions

having more “flat tails” (i.e Weibull fitted function with

low values for the slope factor) Figure 6 shows some

example curves with varying slope parameters For linear

responses the amplitude distribution is given by the

Rayleigh distribution However, when the response is

characterised by more “flat tails” the difference in

determining a short-term or a long-term extreme is larger

than 10% to 20%

Figure 7 shows an example based on use of the IACS

North Atlantic scatter diagram (recommendation 34) and

a Weibull slope factor equal to 0.6 that can typically be

observed from sloshing tests For simplicity 3165

observations of duration 3 hours at a wave period of 8.5

seconds (Hs=5.5–12.5 m) is used for a 40 year period

This corresponds to a return period of 13 months

Consequently, by testing only the worst sea state the

short-term expected extreme is not representative for the

lifetime expected extreme but significantly lower

For completeness an additional example is shown, where

the Weibull function is modelled with a slope factor

E=2.0, which corresponds to the Rayleigh distribution as

usually applied for linear ship responses Figure 8 shows

results from which it can be seen that the ‘long-term’

extreme is only slightly larger than the short-term extreme (1.04)

In order to study this effect experimentally and mainly to

be able to determine a first estimate of the long-term distribution for sloshing-impact fatigue considerations DNV has conducted a sloshing experimental program by carrying out a number of tests for a range of sea state

combinations From this study a very crude estimate of

the long-term distribution indicated a difference factor between the short-term extreme and a lifetime extreme of 1.8 but with a large uncertainty Most importantly from this study it was seen that the slope parameters for lower significant wave heights remained similar

In principle LNG carriers sail fully loaded or in ballast Partially filled trading, i.e tank filling between 70%H and 90%H, may appear only occasionally It is therefore

an important question how to compare a high filling sloshing case, e.g ~95%H, versus a lower filling, e.g

~80%H, when both are analysed for the worst short-term

40 years sea state or how to treat the short-term loads in a long-term absolute load-strength assessment From the discussion above it is clear that the actual lifetime expected extreme for the high filling case is much larger than the short-term value whereas the lifetime expected extreme for the partially filled case (70-90%H) is presumable only slightly larger

Hence the analysis (both comparative and absolute) must

account for this difference Roughly speaking it might be stated that:

x > 90%H filling – representative for normal operation over the lifetime of the vessel – 3hr expected extreme value is not representative for the long term expected extreme

x < 90%H filling – rare or very rare occasions – 3 hour expected extreme for a sea state with 1 year return period may possibly be representative for the long term expected extreme

Figure 6: The effect of different Weibull slope factors

Figure 7: Short–term and “long-term” exceedance

probability curves (Weibull slope=0.6)

Figure 8: Short–term and “long-term” exceedance

probability curves (Weibull slope=2)

2.6 LOADS AT HIGH AND LOW FILLINGS

Many of today’s LNG receiving terminals may have a

marginal capacity for serving the growing fleet of LNG

carriers of sizes larger than the 138 000 m3 reference size

Also, in cases where the terminals have not been able to

make available sufficient storage capacity in time for the

arrival of the next unloading carrier, the ship may be

forced to carry more than the normal heel on the return

ballast voyage

The current maximum low filling height for all DNV

classed membrane carriers is <10%L up to 155 000 m3 in

size and <10%H for all larger carriers

Sloshing impact loads for high and low fillings are

sketched in princilpe in Figure 9 as functions of loaded

area The high filling load curve is associated with the

situation shown in Figure 1 whereas low fillings are

illustrated in Figure 2 The latter is often associated with

a hydraulic jump which generates a larger impact pulse

acting over a larger area than is the case for high fillings,

ref [6] and [7] Hence, the low filling impact pulse may

be more demanding on the strength of the insulation

system [6] The situation can be summarised as follows:

x At low tank fillings (<10%L) sloshing impact

footprints are larger than at high fillings (a 95%H)

due to “hydraulic jumps/breaking wave” effects

x The sloshing impact loads at large areas may in

general be higher at the low 10%L filling level than

the high 95%H filling pressures

If enhanced safety of low filling operations is an issue

basically two options are available:

x Operate with lower fillings at the ballast voyage, i.e

reduce maximum allowed filling height from 10%L

to 10%H

x Reinforce the insulation system to withstand 10%L

fillings; vertical sides above upper hopper knuckle

and transverse bulkheads up to the same level

Altenatively, apply a combination of the above measures

In general little attention has been paid to the low filling issue by the LNG industry Sloshing loads at low fillings therefore need to be more thoroughly investigated 2.7 TESTING AND ANALYSIS OF MEMBRANE

SYSTEM RESPONSE AND STRENGTH The new market demands have identified the need for a methodology for strength assessment of membrane insulation systems under the action of sloshing loads Experiments have shown that changes in design and operation not only affect the magnitude of the sloshing loads, but that large variations also can be observed for its spatial extent and the time history Since the time and spatial distribution of the sloshing impact has significant impact on the response of the containment systems, the

various sloshing events can only be compared in terms of structural response and strength of the systems In

addition, a rational strength assessment methodology is required to identify the necessary strengthening and design improvements to maintain the required safety margins during the new operation

A significant amount of work has been done over the years to study both the static and dynamic impact response and strength of the entire system under certain loading conditions However, in order to construct analysis models for structural response and strength also information on material properties and representative failure modes are needed

DNV has during the last years carried out R&D work with the aim of developing a methodology for capacity assessment of membrane containment systems The objective has been that the methodology should be sufficiently general to allow for assessment of strength changes caused by moderate structural modifications such as

x Change of plate thickness

x Modified distance between lateral supports

x Other minor modifications expected to be proposed

to add strength to the systems The scope of the development work includes:

x Identification of critical failure modes

x Experimental and analytical investigation of the identified failure modes

x Gathering, developing and/or selection of representative stiffness and strength properties of the materials used in the insulation systems

x Specification and development of requirements and procedures for structural response assessment

x Formulation of strength criteria including dynamic and low temperature effects

Figure 9: Principle sketch of sloshing loads vs

load area at high and low fillings

Examples from this work have been published in ref [6]

and [7] and will therefore not be repeated here Testing

has been carried on both single components and system

sub-assemblies Further, the experimental studies have

been complemented with the development of dynamic

non-linear FE response models that has been tested

against, and validated by, the experimental results Based

on this quasi-static and dynamic response models for use

in design have been established and implemented into the

new sloshing guideline [3] Here the dynamic response is

determined by a simplified method using dynamic

amplification factors specified as functions of the ratio

between pulse rise time and system natural period

Most of the work has been focused on ultimate strength

(ULS) behaviour of the systems However, the

international gas code (IGC) [10] and the Classification

Society rules [2] require also the fatigue endurance to be

evaluated Hence, some work has also been done on

fatigue and the conclusions have been incorporated in the

new guideline Figure 10 illustrates the findings

x The number of sloshing impact fatigue cycles is less

than the number of sea loads with a factor of 100

x The Weibull shape factor of the long term response

distribution curve is in the order of 0.6

x This indicates that high cycle low impact loads are

not important for the insulations system

x However, the damaging effect of a limited number

of repeated high impact loads may need to be

considered

x Only the 10-50 highest low frequency load cycles

contribute to the fatigue damage resulting in an

accumulated damage effect (Miner sum) < 0.1

2.8 THE SLOSHING CLASS NOTE

Due to uncertainties in the sloshing impact load

assessment a comparative approach is used for assessing

the strength of the containment system and the

supporting hull structure This is contrary to traditional

direct wave load and strength analysis of ships where an

absolute approach is used However, for the pump tower

structure an absolute approach may be used, [3]

In the comparative approach the sloshing load and

strength of a new LNG carrier design or a new operation

of an LNG carrier is compared with the sloshing load and

strength of the existing fleet of membrane type LNG

carriers that have traded in a safe and damage free

operation The former is referred to as the target vessel,

whereas the latter is referred to as the reference case

2.8 (a) Design Safety Format

The safety format used is a partial safety factor format

which allows uncertainties to be defined and associated

with the actual load response and strength effect rather

than combining everything into one common usage

factor

Load comparative approach:

The following acceptance criterion should be satisfied:

M

ref F tar

p p

JF is the partial safety load factor

JM is the partial safety resistance factor The criterion should be satisfied for the entire range of load areas relevant for the unit dimensions of the containment system

Strength comparative approach:

The format is defined as follows:

M

c F

R DAF

p S

DAF is the dynamic load factor

R c is the capacity in terms of the considered response parameter

JF is the partial safety load factor

JM is the partial safety resistance factor

Pump tower assessment

The strength assessment of the pump tower and supports may be carried out using either a comparative approach

Figure 10: Comparison of scaled long term response

distributions for 20 years and 40 years North Atlantic operation

The absolute approach will usually be most convenient,

since the load and strength need to be calculated for the

target case LNG carrier only The strength is satisfactory

which is a simplification of eq (2) above

Direct vs comparative strength assessment

In a direct strength assessment, the absolute magnitude of

the loads is of major importance, and a thorough

investigation of the loads is necessary Larger load

factors need to be used in the absolute approach than in

the comparative approach, in order to account for the

uncertainties related to the load level

If the comparative approach is followed, as

recommended for the containment system and the hull

strength, the load and strength of the pump tower in the

reference case are compared with the load and strength of

the pump tower in the target case The utilization for the

target case, multiplied with a safety factor, should be

lower than for the reference case The strength is

satisfactory if:

ref c compare

tar

S R

S is the structural response

R c is the capacity in terms of the considered

response parameter

Jcompare is a load factor that reflects the statistical

uncertainty in the comparative load assessment

In the comparative approach, the uncertainty related to

load level is reduced, since the main concern is the load

increase from the reference case to the target case, rather

than the absolute load level

2.8 (b) Strength Assessment Methodology

The methodology can be summarised as follows:

Reference case, Figure 11:

1) Establish a curve relating sloshing impact pressure and sloshing exposed area based on experimental results Load factors to be disregarded in this step 2) Establish a curve relating the impact load capacity of the insulation system and the sloshing exposed surface area of the structure Resistance factors to be disregarded in this step

3) Establish the ratio between the load and the capacity for the entire range of load areas, and identify the maximum ratio between load and capacity Denote this ratio by Dcomp

4) Scale the load uniformly for all load areas using the maximum identified ratio between the load and the capacity The scaled load will now for any load area size be lower than the ultimate capacity of the insulation panels This step is motivated by the damage free operational experience with the membrane type LNG carriers

The resulting load curve is now the basis for the strength assessment of the insulation system and its supporting hull structure

Target case, Figure 12:

1) Establish a curve relating sloshing impact pressure and sloshing exposed area based on experimental results

2) Scale the load using the maximum ratio, Dcomp,between load and response determined from the reference case

3) Carry out a strength assessment of the insulation 4) Carry out the necessary reinforcement of the insulation system so that the load for any load area is lower than the ultimate capacity of the insulation panels

5) Carry out a strength assessment of the supporting hull structure

Figure 12: Scaling target case with same load factor

as for reference case Strengthen containment system according to scaled load curve

Figure 11: Scaling of loads for the reference case to

the ULS capacity

The comparative strength assessment should be carried

out for all insulation structure elements that will

experience sloshing impact loads in the cargo tank In

practice this means the dedicated transverse and

longitudinal corner/knuckle structure and standard flat

wall structure adjacent to the corner/knuckles The

specific locations are determined by the applicable tank

filling limitations and the operation of the vessel

A single comparative load scaling factor, Dcomp,

representative for the weakest element of the considered

insulation structures should be applied in the assessment

of all relevant insulation structures of the target vessel

This means that potential strength margins determined

for the reference case can be utilised in the target case

2.8 (c) Application of the sloshing guideline

The main focus of the Classification Note 30.9 [3] is to

provide guidance to assess sloshing for:

x Increased size LNG carriers

x Offshore loading/unloading

x Partially filled LNG tanks on a particular trade route

However, the class note provides detailed information on

the specification, execution and analysis of sloshing

experiments Consequently, it may be used to assess

other applications than the specific applications listed

above

3 THE CNG ALTERNATIVE

The Compressed Natural Gas (CNG) technology offers

interesting possibilities for handling of associated gas

and for exploitation of marginal gas fields (stranded gas)

The system does not require a gas liquefaction plant and

LNG storage tanks, nor will LNG storage and

regasification at the discharge location be necessary A

fleet of CNG ships may serve as both storage and

transport vehicles and can discharge directly into the land

based gas grid via an on/offshore discharge terminal, an

offshore platform or offshore buoys

3.1 CNG SYSTEM DESIGNS

Methods for shipping gas on keel without a costly

liquefaction process have been studied for decades

without any apparent success Design of containment

systems using pressure vessel codes like the International

Gas carrier Code (IGC), leads to heavy containment

systems with virtually no lifting capacity left for cargo

unless unreasonably large and costly ships were to be

used

The key to the realization of the idea is to use modern

reliability calibrated design codes that offer the same

system safety, but with the use of smaller nominal safety

factors on the structural design A typical example is the

DNV Standard for Submarine Pipeline Systems,

OS-F101, ref [9] that for the X-80 standard pipeline steel allows for a 50% reduction in wall thickness of the containment cylinders as compared to the IGC This weight reduction is the essential door opener for realization of steel based CNG systems onboard ships

A majority of the CNG concepts being proposed or under development are based on using pipelines as the pressure vessels The steel based systems can be designed using the DNV Submarine Pipeline Standard which has become the “world Industry standard” within the pipeline industry Some selected examples are shown in Table 1 and Figure 13

Table 1 Example of CNG Systems

Design condition System CNG

o C]

Coselle (SeaNG)

Vertical Steel Pipes

Trans Ocean Gas

Vertical composite pipes

The CNG concepts apply high pressure in order to keep

the gas in a gaseous state with basically no liquid hydrate

fall-out Concepts with such high pressure (250 bars) are

far beyond the scope for pressure vessel type C tanks

defined in the IGC This gap has been filled by the new

DNV Class Rules for Compressed Natural Gas Carriers

[8] following an equivalent Formal Safety Assessment

(FSA) approach according to IMO MSC 72/16, [11] and

MSC 74/19, [12]

The Trans Ocean Gas design is based on use of 12 m

long gas bottles built in composite materials This is well

proven technology from the aerospace industry and has

several advantages;

x Good track record from the aerospace industry since

1960 and now also successfully being used in CNG

powered buses since 1995

x Better rupture characteristics than steel

x Corrosion resistant

x Lighter than steel (about 1/3 of the weight for

comparable configurations)

x Excellent low temperature characteristics

As for cost comparisons, both steel tank designers and

composite designers maintain that their system is the best

and the most cost effective However, both types of

systems have their advantages and drawbacks They are

all technically feasible, but only the future will be able to

judge on the cost effectiveness

3.2 CNG TRANSPORT ECONOMY

Case studies indicate that for distances from about 500

nautical miles and up to 2500 to 3000 nautical miles it

could be more interesting to use CNG rather than LNG

Figure 14 shows an example case worked out for the

Knutsen PNG® system The figures are based on

x Total cost of capital is 10% (Internal Rate of Return

- IRR)

x 20 year amortisation

x Costs included are operating & maintenance costs, fuel, loading/unloading facilities (jetties/buoys, compression and heating during discharge)

x Costs not included are gas production costs, entry fees market, possible port charges, and government tax

3.3 CNG RULE DEVELOPMENT Since the IGC never was intended to cover compressed natural gas cargo containment systems, no existing rules were previously available for such concepts However, according to IMO, Formal Safety Assessment (FSA) principles can be applied where existing rules do not cover new applications ref [12]

3.3 (a) The DNV rule development Rules for classification of ships, Part 5 Chapter 15 for Compressed Natural Gas Carriers were issued for the first time by DNV in January 2003, [8] This was the first time a complete set of rules has been issued for CNG Carriers The rules are to some extent generic and apply

to ships carrying gases in the superheated phase above the critical temperature The scope covered steel pipeline designs, including PNG®, but did not apply for all proposed CNG concepts In the January 2005 issue design requirements for composite pipes and composite wrapped steel pipes were introduced

Technical background documentation started in the summer of 2000 The rule development was initiated early December 2001, and the first draft was available on February 14th 2002 Internal and external hearings took place from July through October 2002 The new rules that were issued in January 2003 came formally into force by July 1rst 2003 During that time frame the rules had been presented to and discussed twice with the Norwegian Maritime authorities (NMD) and the US Coast Guard (USCG) Also the Norwegian Petroleum Directorate (NPD) had been informed about the development Valuable input and comments were received throughout this process

3.3 (b) Rule Harmonization

At the “2nd International Marine CNG Standards Forum”

at St John’s Newfoundland in August 2005 it was decided that the Classification Societies that had issued rule or guidelines (DNV and ABS) or were in the process

of doing so (Bureau Veritas) were to meet to work out a common set of basic design requirements applicable to steel CNG cylinders Secretary for the work was to be the Centre for Marine CNG at St John’s So far one meeting has been held at ABS premises in Houston and the work

PNG is competitive from about 500 nm and up to 2500-3000 nm

Figure 14: Competitive range for Knutsen PNG®

3.4 VERIFICATION TESTING

An essential part of the verification of the safety of the

containment cylinders is to carry out full scale prototype

tests in accordance with the testing requirements set

fourth in the DNV CNG rules, [8] For steel cylinders the

requirements are:

a Full scale fatigue tests of two end-capped pipes The

fatigue capacity to be at least 15 times the number of

design life pressure induced stress cycles

b Burst test of one full scale end-capped pipe after 2

times the number of design life pressure induced

stress cycles

c Crack tip cool-down during gas leaks through a

fatigue crack at the longitudinal weld seam of a

containment pipe

d Cool-down testing from gas leaks in cargo piping

impinging on the cargo containment cylinders

e Verification that the loading/unloading process

works as intended, by full scale process prototype

testing, small scale model testing or numerical

simulations

The factor of 15 times the design life rather than 10 times

is applied to account for system effects by testing only 2

randomly selected pipes out of more than the 1000-3000

pipes in an actual ship

The reason for the cool-down testing is to explore if, and

under which conditions, the pipe may exhibit brittle

behaviour due to the nozzle effect (Joule Thompson)

from cold gas escaping under high pressure

3.4 (a) Full scale prototype tests and fatigue tests

In order to document that the requirements to burst and

fatigue for the PNG® system were complied with tests

were carried out by Europipe GmbH at their

Mannesmann Research laboratory in Duisburg, Germany

A series of small scale fatigue tests of longitudinal welds

and circumferential welds were made together with full

scale burst and fatigue tests with end-capped pipes The

following tests were carried out:

x One full scale burst test where the requirement was

to maintain full burst capacity after fatigue cycling

to two times the design lifetime, 4 000 cycles

x Two full scale fatigue tests with a required safety factor of 15 to the design life which results in 30 000 cycles

x The creation of an individual S/N curve at mean value minus three standard deviation (m-3s) probability level for the special product and applied production methods

x Proof of statistical safety of the (m-3s) S/N curve between test results from small sample testing and the required limit The requirement for the statistical testing was a safety factor of ten leading to a minimum required number of cycles of 20 000

To establish the product related S/N curve, fatigue tests were performed with full scale samples as well as a higher number of smaller samples in order to create statistical back-up for the individual S/N curve

The smaller samples were cut out across the welds from the full scale fabricated pipes Hence, thickness and welding properties were correctly represented These tests were therefore representative for the actual production quality and production control standards at the steel mill

The tests demonstrated fulfilment of the CNG rule requirements with ample margins No trace of brittle behaviour could be seen - the material behaviour proved ductile, [13], [14]

3.4 (b) Live Gas Cool-down Testing Prior to the gas leak testing Europipe prepared the X-80 test cylinder with a semi elliptical through thickness fatigue crack with a crack length close to the estimated critical length (150 mm at the outside), [15] The crack was positioned in the base material as close as practically possible to the long seam and was made by notch grinding, spark erosion and hydraulic fatigue cycling

The gas leakage test was carried out in full scale with live gas at the Advantica Spadeadam test site in Cumbria

in Great Britain Figure 16 shows the arrangement at the test site The end-capped cargo test cylinder (vessel) was positioned horizontally and was supported and secured

Figure 15: Full scale end-capped test pipe at

Europipe’s test site

Figure 16: Test arrangement

on a concrete pad A second cylinder section from the

same pipe used for the vessel was placed horizontally in

front of the test vessel in line at a distance of 300 mm to

the vessel to represent the cylinder spacing onboard a

PNG® carrier

This pipe section had one temperature gauge on the inner

surface directly in line with the centre of the release from

the test vessel This instrument was aimed at providing

information on the cool-down effect of the adjacent

vessels in case of a direct gas impingement

To be able to pressurize the vessel without prematurely

cooling the pipe wall the crack was sealed by a rubber

padded steel bar pressed to the crack by a hydraulic

cylinder which could be released by remote control

Temperature gauges were fixed on the outside surface of

the cargo containment cylinder close to the crack at six

locations shown schematically in Figure 17

The strain on the vessel wall close to the crack was

measured at six locations in longitudinal and transverse

direction The pressure in the vessel was measured with a

gauge placed directly at the gas inlet into the vessel

The test vessel was pressurized with natural gas (96%

Methane/4% Ethane) to a pressure level of 250 barg

Some leakage was experienced from 140 to 250 bars at

the tips of the seal In order to place the thermocouples

as close to the crack tip as possible not enough rubber

material was present to seal the crack completely The

pump rate had therefore to be increased to reach the

target pressure of 250 barg Due to this mishap the gas

flow from the leakage passed thermocouple T3 and the

temperature in the gas jet could be measured

After reaching 250 bar the sealing mechanism was

released to allow the gas to escape while monitoring

temperature and strain response The pressure level of

250 bars was maintained by pumping for 15 minutes

during free gas flow through the crack Then the

pumping was stopped and the pressure in the vessel

dropped to a pressure of approximately 180 bars over a

period of about 30 minutes After this the vessel was

vented through the pipework During the test the crack

was stable and no indication of fatigue crack growth

could be seen.

Prior to the testing DNV had calculated the temperature

profile through the leaking crack [15] By comparing the

test results with the theoretical predictions the following

could be observed:

x The lowest gas temperature predicted at the crack exit was -70 oC and the lowest gas temperature measured was -68 oC

x At a distance of 20 mm from the crack an outside temperature of -24 oC was measured The predicted temperature at the same location was ~ -30 oC

x At a distance 120 mm from the crack an outside temperature of -0 oC was measured The predicted temperature at the same location was ~ -5 oC

x The temperature at the inside of the adjacent pipe was measured to -16oC The predicted temperature was ~ -10 oC

After testing the crack surfaces were examined in an electron microscope and no indication of crack growth could be seen This means that leak-before–failure had been demonstrated This is a very important result that enhances the safety of the system and simplifies the in-service monitoring arrangement The test also showed that gas impingement onto a neighbouring cylinder will not be critical

4 THE WAY AHEAD

As outlined previously the future will see increased demand for gas carriers able to operate under more severe environmental conditions - cross Atlantic trading and operation in cold climates Due to the trend towards offshore storage and discharge gas carriers able to operate with reduced tank filling levels is, and will be, in demand

In general most independent tank systems can be designed to operate at any tank filling for any fraction of their design lifetime, whereas membrane carriers normally will have to operate inside a carefully determined site specific operational envelope of significant wave heights and heading angles

In order to participate in the transportation of natural gas out of the vast gas fields in the Russian arctic the carriers have to be able to operate under extreme cold and darkness, with icing from sea spray and fog and under adverse ice load conditions This requires carriers specifically adapted to the intended trade and with hull constructions providing adequate protection of the containment systems, the people onboard as well as the environment

For membrane type carriers more work will be needed on part loaded and low filling operations, sloshing loads and strengthening of the containment system Full scale sloshing load and response measurements need to be carried out to better understand the load scaling issue in order to move into an absolute sloshing assessment methodology

Figure 17: Positioning of temperature and strain

measurements on the vessel

To meet these challenges, the future may see a renewed

competition between independent tank systems and

membrane tank systems

On the CNG carrier side the systems closest to

realization appear to be the steel tank systems and

composite wrapped steel tank design Pure composite

based CNG designs are under development and

depending of costs the future may see more CNG

composite solutions

5 CONCLUSIONS

The paper describes the work carried out in DNV to meet

the challenges from the expanding gas transport market,

emphasising sloshing loads and tank system strength for

normal tank fillings as well as reduced tank filling

operations of membrane LNG carriers

An overview of the LNG R&D efforts related to

membrane carriers is given outlining the work on

sloshing loads, system response and strength and full

scale measurements/verification These are important

milestones in developing a sloshing analysis guideline

The first version was issued in June 2006 [3]

Work on the further development of Computational Fluid

Dynamics (CFD) tool for liquid motion analysis and two

phase analysis capabilities is ongoing, both testing and

programme development However, due to the

shortcomings in the present CFD analysis tools DNV has

concluded that for the time being sloshing load

determination has to be based on model testing

This makes a proper understanding of model test scaling

laws vitally important, and a full scale test campaign is

set up aiming to get better insight into these matters

Model testing in different model scales (1:70 and 1:20)

indicates that quite different results are obtained

depending on test conditions, e.g ullage gas Froude

scaling gives considerably higher pressures than gas

density scaling

An example is shown highlighting that the 3 hour

expected extreme sloshing load which is usually

determined from sloshing model tests is not

representative for the long term expected extreme for

high filling, >90%H, but may possibly be so for lower

filling levels

Due to the hydraulic jump effects sloshing loads at low

filling ballast operation (10%L) may in some cases be

larger than the reference case high filling loads (95%H)

This may need special consideration either in terms of

reduced filling height and/or strengthening of the

containment system in the lower parts of the tanks

Work with the aim of developing a methodology for

capacity assessment of membrane containment systems

has been carried out and is implemented into the new sloshing guideline [3]

Some of the background for, and the basic principles behind, the sloshing guideline are described Due to uncertainties in the sloshing impact load assessment a

comparative approach is used for assessing the strength

of the containment system and the supporting hull structure However, for the pump tower structure an absolute approach may be used

The emerging Compressed Natural Gas (CNG) technology offers interesting possibilities for handling of associated gas and for exploitation of marginal gas fields (stranded gas) Examples of CNG designs are shown; both steel based and composite solutions Compared to pipeline and LNG transport, an example of transportation costs for the Knutsen PNG® system indicates that CNG can come in as an interesting supplement in the transport range from 500 – 2500/3000 nautical miles

In order to provide a common basis of design criteria for CNG, DNV, ABS and BV are working on rule harmonization together with the Centre for Marine CNG

at St John’s, Newfoundland

Full scale prototype verification tests of CNG containment cylinders have been done/are underway for several of the CNG systems For the PNG® design successful tests at ambient temperature have been reported in ref [13] and [14] Further, a full scale leakage test with live gas has been carried out at the Advantica Spadeadam test site in Cumbria in Great Britain [15] The outcome of the test was that even when exposed to the cooling effect of the leaking gas (down to -70 oC) the crack was stable and no indication of crack growth could

be seen The test also showed that gas impingement onto

a neighbouring cylinder was not critical

6 ACKNOWLEDGEMENTS

The work on LNG development presented in the present paper is the joint efforts of the DNV LNG R&D team Valuable input and comments have been given by Jens Bloch Helmers, Øyvind Lund-Johansen and Thor Hysing Thanks also go to our previous DNV collegues Trym Tveitnes and Wouter Pastoor

On the CNG part thanks go to Kim Mørk, Erling Fredriksen, Morten Lerø, Kjell Olav Halsen and Lars Even Torbergsen

7 REFERENCES

1 Energy Information Administration (EIA):

“International Energy Outlook 2005”

2 Det Norske Veritas: “Liquefied Gas Carriers “, Rules

for Classification of Ships Pt 5 Ch 5, July 2005

3 Det Norske Veritas: “Sloshing Analysis of LNG

Membrane Tanks”, Classification Note 30.9, June

2006

4 Kleefsman, K.M.T, Fekken, G., Veldman, A.E.P.,

Iwanowski, B and Buchner, B.: ”A

Volume-of-Fluid based simulation method for wave impact

problems”, J Comp Phys (2005) 206 363-393

5 Berg, A.: “Scaling Laws and Statistical Distributions

of Impact Pressures in Liquid Sloshing”, A.S Veritas

Research, Report no 87-2008, (1987)

6 Pastoor, W., Tveitnes, T., Valsgård, S and Sele, H

O.: “Sloshing in Partially filled LNG Tanks – an

Experimental Survey”, OTC 16581, May 2004

7 Pastoor, W., Østvold, T K., Byklum, E and

Valsgård, S.: “Sloshing Load and Response in LNG

Carriers for New Designs, New Operations and New

Trades”, GasTech 2005, Bilbao, Spain, March 14-17,

2005

8. Det Norske Veritas: “Compressed Natural Gas

Carriers”, Rules for Classification of ships Pt.5

Ch.15, January 2006

9 Det Norske Veritas: “Offshore Standard –

Submarine pipeline Systems”, DNV-OS- F101,

January 2000

10 International Maritime Organisation (IMO):

“International Code for the Construction and

Equipment of Ships Carrying Liquefied Gases in

Bulk - IGC Code”, 1993 Edition

11 International Maritime Organisation (IMO): “Formal

Safety Assessment Decision parameters including

Risk Acceptance Criteria”, Maritime Safety

Committee, MSC 72/16, 14 February 2000

(Submitted by Norway)

12 International Maritime Organization (IMO):

“Guidelines for Formal Safety Assessment (FSA) for

use in the Rule-Making Process”, MSC/Circ.1023,

MEPC/Circ.392, 5 April 2002

13 Valsgård, S., Reepmeyer, O., Lothe, P., Strøm, N.K

and Mørk, K.: “The Development of a Compressed

Natural Gas Carrier”, The 9 th International

Symposium on Practical Design of Ships and other

Floating Structures (PRADS 2004),

Lübeck-Travemünde, Germany, September 12-17, 2004

14 Valsgård, S., Mørk, K.J., Lothe, P and Strøm, N.K.:

“Compressed Natural Gas Carrier Development –

The Knutsen PNG Concept”, SNAME Annual Meeting, Washington DC, September 29th - October

2nd , 2004

15 Reepmeyer, O., Lothe, P, Valsgård, S., Peppler, M and Knauf, G.: “ Full Scale Gas leak Test at a Large Diameter X-80 DSAW Pipe”,

Erdelen-Proceedings of IPC 2006, 6 th International Pipeline Conference, Calgary, Canada, September 25-29,

2006

8 AUTHORS’ BIOGRAPHIES Sverre Valsgård, Ph.D., is a Senior Principal Engineer

and focus responsible for gas carrier development on the maritime side of Det Norske Veritas He has more than

30 years experience from DNV on R&D and consultancy

on ships including gas carriers and offshore structures as well as from management positions

Tom Klungseth Østvold, M.Sc., is a Senior engineer at

Det Norske Veritas He is mainly working on development projects to improve calculation procedures and acceptance criteria for ultimate strength assessment

of (ship) structures Currently he is working as task leader on strength assessment of LNG cargo containment systems and tank structures exposed to sloshing impact loads Other activities include strength assessment consultancy studies on behalf of clients and DNV Classification

Olav Rognebakke, Ph.D is a Senior Engineer in Det

Norske Veritas He has five years of working experience from Marintek and DNV with focus on R&D activities and commercial model tests Main competence areas are hydrodynamic theory, experimental techniques, analysis methods and seakeeping issues relevant for sloshing in membrane type containment systems for transportation

of Liquified Natural Gas (LNG)

Eirik Byklum, Ph.D., is a Senior engineer at Det Norske

Veritas He is mainly working with development of procedures and calculation tools for ultimate strength assessment of marine structures Currently he is working

on strength assessment of LNG cargo containment systems and tank structures exposed to sloshing impact loads He is also involved in strength assessment consultancy studies on behalf of clients and DNV Classification

Hans O Sele, M.Sc., is a Project Engineer in Det Norske

Veritas He has 5 years experience from DNV on hydrodynamics and structural analyses and R&D on LNG sloshing

ASPECTS OF THE PROPULSION OF A 250000 M LNG SHIP

J S Carlton, Lloyd’s Register, London

SUMMARY

Recognising the present market trends for the transport of LNG, it is considered that the construction of a 250000m3LNG ship may be probable in the not too distant future Based on the results from Lloyd’s Register’s continuing research programme into LNG ships this paper considers some aspects of the propulsion of such a ship and the challenges it presents in terms of hydrodynamic and machinery design The relative merits of different propulsion configurations are discussed in the context of their propulsion efficiency, cavitation characteristics, reliability and the ship’s vibration performance

1 INTRODUCTION

The current sizes of LNG ships in service have reached

around 150000 m3 while a number of ships of up to

210000 m3 and beyond in capacity are in the various

stages of contemplation and design Moreover, resulting

from a number of economic and commercial forecasts for

LNG transportation, it may be that these ships could

require additional capacity and speed capabilities from

those currently contemplated Indeed, within the

literature predictions are made of 250000 m3 capacity

and potential speed requirements up to 25 knots within

the foreseeable future These predictions represent a

considerable extrapolation of current practice which

demands that careful thought is exercised on how best to

achieve the engineering challenge set by the potential

commercial market demand Notwithstanding the trade

routes currently served from the Middle East to Japan

and the Far East as well as services to Europe, other trade

routes are now being considered which include the

export of LNG from Arctic waters and this implies an

additional set of design constraints

Recognising these developments Lloyd’s Register has

implemented an ongoing research programme devoted to

the problems associated with the future development of

LNG ships One facet of this initiative has been related

to the propulsion of large LNG ships over a range of

speeds from 19 to 25 knots This paper considers some

of the options associated with single and twin screw hull

forms in terms of various propulsor configurations to

satisfy the design problem

Although LNG ships have traditionally been the preserve

of steam turbine propulsion which, while relatively

inefficient as a prime mover by today’s standards, has the

advantage of being able to use boil-off from the cargo in

order to redress the propulsion economics Furthermore,

notwithstanding the efficiency implications of steam

turbine propulsion, there have also been occasional

difficulties in some cases with the reduction gearing

systems as well as the ever attendant problems of finding

suitably qualified personnel to operate and maintain

steam plant Today, however, a number of other

propulsion options present themselves as potential

alternatives; typically, dual-fuel diesel engines, gas turbines and electric propulsion These differing prime mover options introduce associated machinery design implications and some of these are explored within the paper

The current developments in Artic regions for the export

of LNG also pose a further set of design challenges In the context of ship propulsion these are given some preliminary consideration here both in the context of the LNG research programme as well as other Lloyd’s Register initiatives specifically dealing with operations in ice

2 SHIP SIZE PARAMETERS

A sample of 167 existing LNG ships, having cargo capacities ranging from 1100 m3 to 147600 m3, has been analysed in terms of their principal hull form parameters The cargo carrying capacity of this sample of ships, which were all built between 1965 and 2004, is shown in Figure 1

Figure 1: Ship Size with respect to Year of Build From the Figure it can be seen that there was step change

in capacities to around 120000 to 130000 m3 capacity in

1975 and, apart from relatively few exceptions, this remained the size of LNG ships until about 1995

0 20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000

Following this there was a tendency for some ships to

have a slight increase in capacity to just less than 140000

m3 and then again in recent years a further marginal

increase to close to 150000 m3

With regard to hull dimensions a consistent and growing

increase in ship length is apparent from Figure 2 in

which the rate of increase of length with respect to

capacity, while initially increasing rapidly, shows a

reducing trend as the ships become larger Similarly in

terms of beam, while the draught exhibits an asymptotic

tendency to a value of around 12m This latter trend is

due to the ships being centred on a trading pattern from

the Gulf where draught limitations exist at the loading

ports due to the geology of the sea bed Therefore, for

ships intended for trading in this geographical location

this requires that increased capacity has to be obtained

from either length or breadth increases However, these

variables need careful exploration in order to optimise

the hydrodynamic performance of the hull form

It will be noted from Figure 3 that there is a fairly wide

scatter of breadth in the larger sizes of ships above

120000m3 This, in turn introduces a correspondingly

wide range of L/B and B/T ratios for the ships For ships

above 120000 m3 in the data sample considered the range

of L/B ratio is seen to vary from 5.49 to 6.76 whereas in

the case of the B/T ratio, this varies from an extreme

value in one case of 2.65 up to 4.31 More

characteristically, however, for the data set a lower

bound for the B/T ratio is 3.16 In the case of block

coefficient it is seen that there is an inverse relationship

with L/B ratio, with values of 0.71 to around 0.77

corresponding approximately to L/B ratios in the range

6.3 to 6.0 respectively However, as might be expected,

there is some scatter of data points about this mean

tendency

To explore the propulsion options two basic ship forms

were derived which were consistent with the general

development of LNG ships in recent years: particularly

with the constraints applying to the larger ships These

basic ship forms were designated the parent single and

twin screw hulls and would be expected to comprise the

equivalent of a five membrane tank capacity; each tank

having a 50,000 m3 capacity Nevertheless, within this

overall configuration it was recognised that a variety of

tank configurations, in terms of type and geometry,

might be employed for various cargo transportation and

tank sloshing attenuation reasons In each case, however,

the basic restriction of a 12 m design draught was

initially retained after which this constraint was then

relaxed to explore the implications of the design space

Figure 2: Ship Length between Perpendiculars

Figure 3: Moulded Breadth and Maximum Draught

3 PARENT SINGLE SCREW HULL FORM

The basic hull form parameters of the parent single screw ship are shown in Table 1

Length Overall 344.0 m Length Between

Perpendiculars

333.0 m Breadth (Moulded) 56.0 m Design Draught 12.0 m Block Coefficient 0.756 Mid-ship Section

Coefficient

0.991 LCB Position 0.33% (fwd) Displacement 174081 tonnes

Table 1: Parent Single Screw Ship Principal

Dimensions

0 50 100 150 200 250 300 350

These dimensions yield L/B and B/T ratios of 5.95 and

4.67 respectively in the design loaded condition Then

by assuming a residual ballast voyage cargo of 2%

together with the normal boil-off rates, the ballast

condition for this ship is characterised by a draught of

9.65 m At this condition the block coefficient reduces to

0.737 and the longitudinal centre of buoyancy moves to a

location 0.7% forward

4 PARENT TWIN SCREW HULL FORM

For the corresponding case of the parent twin screw hull

form the principal hull dimensions are shown in Table 2

Length Overall 344.0 m

Length Between

Perpendiculars

333.0 m Breadth (Moulded) 56.0 m

Displacement 174839 tonnes

Table 2: Parent Twin Screw Ship Principal Dimensions

While having similar nominal L/B and B/T ratios to its

single screw counterpart this ship is a little fuller due to

the influence of the stern gondolas enclosing the shaft

lines In the ballast condition the ship the block

coefficient for this version of the ship reduces to 0.743

5 SHIP PROPULSION

The propulsion characteristics have been considered in

relation to three principal conditions: the trial contract

condition, the service loaded draught and the ballast

condition In the latter two conditions there is a

requirement to introduce a sea margin into the power

prediction The most appropriate sea margin to use is

dependent upon the service routing and overall voyage

speed requirements of the ship Some discussion of these

aspects, albeit in that case in relation to container ships,

is given in [1] For the purposes of this propulsion study

a sea margin of 15% has been adopted

5.1 SINGLE SCREW SHIP

The reference case has been taken as the power

absorption associated with a ship speed of 20 knots

Such a power absorption would lead to a mono-block,

fixed pitch nickel-aluminium bronze propeller of around

97 tonnes which is also well within the existing

capabilities of foundries The principal constraint with

this propulsion configuration, however, is considered to

be one of propeller cavitation together with the attendant

risk of high levels of ship vibration and propeller blade

and rudder erosion In the loaded condition, because of

the ship’s restricted draught, the immersion of the propeller blade’s 0.9R radial position at TDC is around 5.3 m, even when allowing for the existence of a relatively small dynamic sinkage and a representative stern wave at the propeller location of the ship In the ballast condition the resultant immersion reduces to 2.95

m This creates in both conditions, but more particularly

in the ballast condition, an onerous cavitation environment which leads to a number of probable cavitation problems for the propeller when working in the flow field generated by the hull and its appendages These unwelcome cavitation effects are that in addition

to the generation of a strong and potentially fluctuating tip vortex, there is likely to be an extensive sheet cavity generated over the suction surfaces of the blades This may have a tendency towards strong instability, particularly at the trailing edges of the cavities Moreover, the extent of the sheet cavity close to the blade tips is predicted from lifting surface based propeller computations to extend across the whole blade section and become supercavitating Lloyd’s Register’s full scale observation experience has suggested that when this occurs the sheet cavity will interact with a strong tip vortex to create the potential for broadband excitation within that part of the frequency spectrum embracing the first to fourth blade rate harmonic frequencies: as distinct from the more normally expected broadband excitation characteristics located further up the frequency spectrum

A typical example of this type of interaction is shown in Figure 4a In this image, which was recorded under constant course heading conditions, the tip vortex has swept up the supercavitating part of the sheet cavity which then, under the centrifugal action, initially decomposed into an expanding cloud before reconstituting itself some way down stream into the cavitating tip vortex core Reference [2] discusses this full scale phenomenon in more detail Furthermore, during turning manoeuvres of this LNG ship the complexity of this cavitating flow regime was significantly enhanced in that the size and dynamic behaviour of the vortex structure increased, Figure 4b Here, in addition to the introduction of a hub vortex, the already robust and complex tip vortex structure also included a system of ring vortices In this case, as well

as in a number of others relating to large ships by today’s standards, the cavitation dynamics contributed significantly to the ship’s internal vibration signature both during straight running and turning conditions

Experience with the smaller single screw LNG tankers has shown that considerable attention has to be paid to achieving an acceptable quality of the wake field This relates to a number of aspects: first, the underlying flow field generated by the hull; secondly, from any appendages located on the hull and finally from the influence of any outflows from the hull which pass into the propeller part of the ship’s wake field Indeed, considerable care is required in the citing and influence

of the outflows and the design of the model test programme needs to specifically address this issue

Figure 4(a) Cavitation on a Straight Course

Figure 4(b) Tip Vortex Behaviour during Turning

Figures 4 (a) and (b):Full Scale Cavitating Sheet and

Vortex Cavitation on an LNG Ship

A further attendant risk with relatively high block

coefficient ships, such as those under consideration here,

is that the propeller cannot draw sufficient water from

ahead, particularly in the upper parts of the propeller disc

The propeller, therefore, has to resort to drawing water

from astern of the propeller disc which induces a reversal

of the flow over the after part of the hull surface Indeed,

this tendency has been observed on some model tests

over a range of LNG ship sizes When this occurs

stagnation streamlines will be generated between the

propeller blades and the hull surface around which a

propulsor-hull vortex can be induced Typically these

cavitate intermittently and their periodic collapse

introduces an unwelcome aperiodic addition to the ship’s

vibration signature

Given the cavitation environment generated by the

250000 m3 ship with a 12 m draught limitation, it is considered that it will be extremely difficult to avoid serious shipboard vibration and propeller blade erosion if the ship is propelled by a single screw Moreover, these difficulties are seen to manifest themselves from ship speeds as low as 18 to 19 knots which then gives little scope for increasing the design speed if future commercial considerations require higher speeds

If the 250000 m3 ship were required to operate on a trading pattern which removed the need for the existing

12 m draught restriction then the ship’s draught could be increased To explore this effect the parent hull was transformed into a series of forms having design draughts

in the range 12 to 15 m together with the constraint of keeping the hull volume and length constant This tends

to have a beneficial effect on the hull effective power at the nominal design speed of 20 knots since the changed L/B and B/T ratios give a more favourable propulsion situation The increase in draught also creates a better cavitation environment, subject to controlling the propeller diameter and rotational speed appropriately and also assists in developing a more helpful wake field Consequently, if the draught constraint were relaxed there is some limited scope for balancing an enhancement in the quasi-propulsive coefficient against a reduction in propeller radiated hull pressure signature and propeller blade cavitation erosion potential

To explore the implications of varying the alternative parameter of length, the parent hull form was increased

in length in steps up to a maximum of 20% while maintaining the draught constant at 12m This permitted

a variation in the ship length to breadth ratio of 5.87 ”L/B” 8.45 and the breadth to draught ratio of 3.9 ” B/T

” 4.68 The maximum length increase and the consequent reduction in beam, within the range considered, gave an effective power benefit of around 5% based on the parent hull form The reduction in breadth also maintained the ship’s beam to similar proportions of some of the largest ships in service today

At a ship design speed of 20 knots the radiated propeller blade rate hull surface pressures are predicted to be of the order of 10 kPa over the range of length increase Since this is a relatively high value, it is unlikely that the potential for any significant increase in ship design speed will accrue from increasing the ship length under the restrictions defined above Furthermore, it is not anticipated that any significant advantages in the propeller cavitation environment will derive from these hull variations and, consequently, much of the earlier discussion remains valid for these variations

5.2 TWIN SCREW SHIPThe twin screw propulsion option, although having the potential for suffering from a quasi propulsion coefficient penalty, offers the benefit of an easier propeller design environment The propulsion efficiency penalty,

however, may not be too severe since by judicious

development of the ship’s hull lines it is considered

possible to minimise this effect, if not derive some

propulsion advantage, when compared to its single screw

counterpart

The parent twin screw hull form propulsion arrangement,

despite the restricted draught in both the loaded and

ballast conditions, establishes a significantly better

propeller design environment For such a propeller,

lifting surface computations suggest acceptable levels of

sheet cavitation in the loaded and ballast conditions,

Figure 5, for a ship scale effective wake field

Furthermore, the margin against face cavitation was

found to be of the order of 0.68 KT and, in addition, the

vortex cavitation characteristics are considered to be far

more benign in this case and are unlikely to lead to the

complexity suggested in the single screw alternative

Figure 5: Typical Twin Screw Sheet Cavitation

Characteristics

For this design option gondolas were used to lead the

shafts outboard from the hull The design of these

gondolas is particularly important if a good flow field is

to be created at the propeller station and the full

efficiency potential of the twin screw arrangement

realised In this context the use of flow visualisation

techniques, such as paint and tufts, on large scale ship

models is strongly recommended as forming an integral

part of the hull design process Notwithstanding the

design of the shaft gondolas, it is equally important to

construct those appendages such that the intended

geometry and relationship to the hull is reproduced to

within the desired tolerances so as not to unduly disturb

the design flow field

When discussing the single screw propulsion option it was concluded that there was a ceiling on ship speed above which the cavitation related problems of propeller blade erosion and vibration would become difficult to overcome In the case of the twin screw option there is considerably more flexibility to accommodate design speed increases beyond 20 knots without incurring significant design problems Furthermore, an additional benefit in the case of twin screw propulsion arrangements is that there is also redundancy in terms of the propulsion capability

The podded propulsor concept lends itself without significant extrapolation of present installed powers to the propulsion of a twin screw version of a 250000 m3LNG ship Currently, propulsors having power ratings

up to 23MW are operating satisfactorily and, therefore, would be compatible with the power requirements for a large LNG ship of the type under consideration Since podded propulsors would operate in a tractor mode the inflow, by virtue of their location on the hull, would be disturbed only by the hull boundary layer and, therefore, the cavitation development and collapse over the blades would be considerably less onerous than for the conventional twin screw arrangement Nevertheless, when selecting the propulsors’ position on the hull there

is sensitivity to location from a propulsive efficiency perspective Consequently, if a podded propulsion arrangement is contemplated optimisation of the location

of the units on the hull with the aid of model tests is recommended at an early design stage Moreover, if podded propulsors are contemplated the propeller generated loadings must be correctly estimated in both the free running and manoeuvring conditions since it has been found that the loadings, particularly those related to shaft bending and shear forces, are sensitive to factors such as sea conditions, stopping, turning and propulsor interaction effects [3,4]

The propeller blade cavitation characteristics of podded propulsors have shown themselves to be particularly good when used with cruise ship hull forms, this, however, has not always been the case for hull forms with strong buttock flow characteristics In these cases care needs to be exercised in the choice of blade loading distribution in order to prevent the unwelcome effects of sheet and tip vortex cavitation from occurring Furthermore, an additional cavitation related aspect which requires care is the occurrence of cavitation phenomenological behaviour which may give rise to broadband excitation characteristics While the broadband problem is currently not yet fully understood,

it is possible to design the blade radial and chordal loading within the constraints of the ship’s wake field so

as to minimise the probability of incurring significant broadband excitation

200

Predicted Cavitation Pattern at

deg (ITTC Angle)

6 OVERLAPPING PROPELLERS

Overlapping propellers were a concept originally

proposed by Pien and Strom-Tejsen [5] and subsequently

explored by Kerlen et al in the context of container

ships [6] and Restad et al for LNG ships [7] While

these studies related to earlier generations ships than

those contemplated today the relevance of the

overlapping propeller concept, Figure 6, may well benefit

from re-examination in the context of the next generation

of LNG ships In essence it was found that by deploying

two propellers, one just in front of the other and whose

shaft line spacing was less than their diameter, the

propulsion benefits of the single screw system could be

partially regained but, due to the power absorption being

split between two propellers, the cavitation and

consequent vibration characteristics were more amenable

to control In the case of overlapping propellers the pitch

ratio of the two propellers are slightly different in order

to account for the mean flow induction effects and the

propellers rotating in opposite directions

Figure 6: Outline of Overlapping Propeller System

With an overlapping arrangement, however, the

propellers have to be phased so as to inhibit the mutual

interaction of the individual propeller cavitation

characteristics Typically, these interactions are

characterised by the tip vortex from the leading propeller

interacting with the blade cavitation on the following

propeller as well as mutual tip vortex interaction

Furthermore, with an overlapping propeller arrangement

the requirement for long shaft gondolas is removed and

smaller, more energy efficient fairings can replace them

7 SHIPS OPERATING IN ICE

The exploitation of the arctic gas fields have led to the

development of a ship hull form with a double acting

ability, shown schematically in Figure 7 These ships,

powered by podded propulsors, are designed to travel in

the conventional bow first direction when operating in

ice free or lightly populated ice waters, but in the reverse

direction when operating in heavy ice conditions

Figure 7: Double-Acting Ship Earlier developments in ice breaker propulsion technology showed that if the hull was lubricated at the bow then the resistance when travelling though ice was considerably reduced This led to the development of a number of hull forms but particularly the spoon shaped, water lubricated bow form In the case of the double acting ships, typical of which are the Mastera and Tempera both classed by Lloyd’s Register, the forward part of the hull is optimised for open water conditions while the after part is designed to accommodate heavy ice interaction The water lubrication for minimum ice breaking resistance is provided by the podded propulsors which induce a flow of water over the after part of the hull surface when working astern It has been found by the Finnish designers that this type of arrangement gives

a significant power saving when operating stern first in ice conditions such that typically a third to a half of the installed power in the pods is required during the ice navigation Double acting hull forms of this type are rather different to those of conventional LNG ships, consequently, if these double acting ships will eventually

be required to operate in open ice free seaways, for example in the North Atlantic, they may exhibit different seakeeping capabilities with corresponding potential changes in the sloshing characteristics of the cargo in the tanks Such characteristics need to be carefully understood should trade routings of this type become a reality

The propeller design for arctic operation, because of its requirement to withstand ice milling and impact scenarios, will have thickened blade tips Figure 8 outlines a typical, but by no means extreme, full scale variation in loading measured by Lloyd’s Register on a ship when navigating in moderate ice conditions with a conventional propulsion system While thickened blades are necessary for the blade strength they are counter to the requirements for controlled cavitation development

and the choice of the propeller blades’ distribution of

loading will require careful selection if severe cavitation

influences are to be avoided: both in terms of unstable

sheet cavitation and also strong tip vorticity In the case

of podded propulsors full scale observations, again made

by Lloyd’s Register, of the cavitation development on

blades designed to operate in ice has underlined the

vibration and erosive potential of these types of propeller

blade if optimisation has not taken place at the design

stage Furthermore, the aggressiveness of the tip vortices

can be such that erosion and paint removal can extend to

the podded propulsor body Blade skew is one parameter

which can assist in the attenuation of these unwelcome forms of cavitation However, skew because of the blade geometry that it induces can also have an undesirable influence on the blade stress and loading distributions under ice interaction conditions In a joint programme between Transport Canada, Lloyd’s Register and the National Research Council Canada these influences were examined at model scale [8] in order to better understand these interactions

Figure 8: Characteristic Ice Interaction Sequence in an Ahead and Astern Manoeuvre

8 PROPULSION MACHINERY

Although the propulsion of LNG ships has been

dominated by the steam turbine there has been a quest for

more economic solutions in recent years Given the

desirability of adopting a twin screw based propulsion

system for a 250000 m3 capacity ship and the

requirement that an LNG ship must maintain a

propulsion power capability when loading and unloading

at terminals, the alternatives that present themselves for

consideration are:

x Slow speed diesel engines

x Slow speed engines with reliquifaction plants

x Dual-fuel, medium speed diesel-electric machinery

x Combined gas turbine–electric plants

When ranking the basic unit efficiencies associated with these options the planned ship operation together with the possible changes that might occur due to economic and market factors need also to be taken into the assessment process Implicit in this process is the provision of some flexibility of operation unless the commercial long term operation is unambiguously known In particular the required relationship between the availability of natural and forced boil-off LNG for propulsion purposes in relation to heavy fuel and marine diesel oil needs to be clearly understood before a meaningful comparison between propulsion options can

be made Moreover, within this overall assessment process the failure probabilities of the different propulsion system options form a further important variable

As one basis for comparison the reliability of marine

steam turbine plant over the last 10 years can be judged

from Figure 9 which shows the failure incidence per 10

years of service for the principal components comprising

the system

Figure 9: Failure Incidence of Steam Turbine

Propulsion System Components

Within this broad classification of failure the most

significant failure sub-categories were:

Boilers Tube failures 28.9%

Superheaters 25.1%

Drum/shell 11.6%

Turbines Rotor failures 40.8%

Condensers 13.1%

Shafting Outboard Stern Gland 46.6%

Inboard stern gland 24.0%

In an alternative case of slow speed, two stroke

propulsion two principal systems present themselves

These are the use of a slow speed diesel engine operating

on the normal grades of fuel with a re-liquefaction plant

to either preserve the boil-off from the cargo or,

alternatively, use the boil-off LNG as a partial fuel for

the engine in association with conventional fuel Clearly,

the determinant in this case is whether it is considered

economically appropriate to use either natural or forced

boil-off for propulsion purposes

In the former case the system principally reduces to a

normal slow speed marine propulsion plant which utilises

the high thermal efficiencies of the two stroke diesel

engine In such a case for the propulsion of a 250000 m3

capacity ship two slow speed engines, each serving one

shaft, and housed in separate engine rooms with the

re-liquefaction plant sited elsewhere for cargo recycling

would form the basis of an appropriate arrangement

Furthermore, while fixed pitch propellers would perform

satisfactorily under normal service conditions, if

controllable pitch propellers were fitted this would

provide a further degree of propulsion system control in

the event that one shaft line had to be stopped during

service for repairs Nevertheless, the economic advantage of making this provision in terms of initial cost would need to be carefully weighed against the operational cost penalty of the probability of the system failing For slow speed marine diesel propulsion a considerable body of reliability data exists and Figure 10 demonstrates the failure incidence associated with this type of system

Figure 10: Failure Incidence of Direct Drive, Slow

Speed Diesel Propulsion System Components

If it were admissible commercially to use the natural and forced boil-off gas from the cargo, then dual fuel slow speed engines are still a potential candidate for propulsion Since the fuel oil and gas are admitted to the cylinder via the injector and the combination of fuel oil and gas can be effectively varied according to the circumstances by the use of electronic control systems, the boil-off gas requires to be brought up to injection pressures Therefore, there is a need for compressors and this raises safety issues of having compressed gases within the engine room However, there are a number of engineering solutions for attenuating this danger: for example, by the use of double wall piping concepts within the engine room

For a slow speed, two stroke direct drive propulsion system comprising one prime mover per shaft the electrical power generation would be achieved though a system of medium speed diesel generator sets with a possible contribution from a shaft driven generator Moreover, there is also the possibility of using a shaft generation capability in the reverse sense of a shaft motor driven from either the generator sets or an exhaust gas thermal recovery system should additional propulsion power be required during a voyage Notwithstanding this, the generator sets might also be driven from dual fuel medium speed engines if so desired by the ship’s operating policy

An alternative concept of dual fuel diesel-electric propulsion would comprise a set of medium speed, four stroke dual fuel generator units driving electric motors

Ge ar g

Sh afti ng

Engine Turbocharger Shafting

coupled to the propulsion shafting This, with the

exception of the dual-fuel element, is analogous to the

well established propulsion practice for some other ship

types; most notably cruise ships The failure incidence

rates for diesel-electric propulsion systems in ships over

the last ten years are shown by Figure 11 This, as the

diagram suggests, relates to propulsion systems having a

conventional shafting system and propellers However,

if podded propulsors are included into the comparison,

then by comparing the failures associated with these

units to the combined motor, shaft-line and propeller

incidences of failure for the conventional propulsion

systems, the ratio of the incidence of failure of podded to

conventional systems is 4.6 to 1.4 per ten years of service

However, when making comparisons of this type it must

be noted that podded propulsors have been a maturing

propulsion concept during the review period and in such

circumstances it would be reasonable to anticipate a

higher incidence of failure

Figure 11: Failure Incidence of Diesel-Electric

Propulsion System Components

Diesel-electric propulsion systems are particularly

versatile in that they can accommodate a range of

propulsion, hotel, cargo and other loadings from the set

of diesel generators In the case of the subject 250000

m3 LNG ship the arrangement of dual fuel, four stroke

diesel generators would likely comprise some 6 units,

typically four, twelve cylinder and two, six cylinder

generator sets Moreover, in terms of maintenance,

multi-engine systems such as these give considerable

flexibility without unduly interrupting the ship trading

pattern

If podded propulsors were contemplated then experience

has shown that adequate control of the podded propulsors

with respect to azimuthing angles, ship speed and shaft

rotational speed will be necessary in order to limit the

loads imposed on the motor and shaft bearings

Additionally, because the shafting arrangements of

podded propulsors deploy rolling element bearings a

lubricating oil cleanliness strategy must be implemented

which introduces a higher level of control than that

normally used in conventional marine engineering

practice While the bearings are designed for a given L10

life which is defined as at least 65000 hrs in Lloyd’s Register’s Rules, unless some other agreement between the owner and shipbuilder has been reached and which is compatible with survey intervals, experience has shown that oil contamination levels of at least NAS 6 are required in order to give an acceptable probability of achieving the bearing life [9]

Gas turbine propulsion systems bear some similarity to those for dual fuel diesel-electric systems in that the gas turbines will most likely drive alternators which supply electrical power to a main electrical bus bar From this bus, power will then be drawn for propulsion and other purposes by means of suitable converter and transformation systems Gas turbines are able to run on boil-of gas as well their normal fuels by means of dual passage fuel injectors into the combustion chamber and are capable of achieving the power requirements for a ship of this size by means of two turbo-generators, perhaps employing a father and son arrangement While the reliability of aero-derivative engines is generally good the maintenance strategy is different with these types of prime mover Typically, on board inspections will normally concentrate on the visual inspection of safety items and the fluid levels while the units are changed-out at a specified time interval: for example, after 24000 hrs

9 CONCLUSIONS

The discussion has centred upon a number of aspects relating to the propulsion of a 250000 m3 LNG tanker From this analysis the following conclusions have been reached

9.1 A typical 250000m3 LNG ship will have a

length between perpendiculars, breadth and draught of 333m, 56 m and 12 m respectively However, once a particular trading situation is known then a propulsion advantage can be gained by optimising these dimensions with respect to each other In particular increases in length and reductions in breadth are beneficial if the draught of 12 m has to be preserved Clearly, any relaxation of the 12 m draught restriction is likely to be advantageous with regard to hydrodynamic propulsion efficiency 9.2 A twin screw propulsion arrangement needs to

be adopted in order to achieve an acceptable cavitation performance from the propulsor and minimise the probability of ship vibration problems that are difficult to solve

9.3 While the traditional expectation for a twin

screw propulsion arrangement is that it will lead

to inferior propulsion efficiency, this, however, through proper attention to the hull design need not be particularly severe for these large ships Indeed, some slight advantage may be accrued

9.4 It is recommended that in any model test

programme paint or tuft tests form part of the

procedure in order to perform flow visualisation

over the hull under propulsion conditions

9.5 The twin screw arrangement, unlike the single

screw alternative, offers scope for embracing

higher design ship speeds than those currently

required should these be commercially desirable

9.6 In the case of the twin screw propulsion

arrangement a number of propulsor options

present themselves Among the more promising

are the conventional fixed and controllable pitch

propellers and podded propulsors Additionally,

the overlapping propulsor concept may have

potential in this ship application

9.7 A number of options present themselves as

contenders for the propulsion machinery to

replace the traditional steam turbine propulsion

system The economic case for the choice is

dependent on the constraints of how much

boil-off gas, either natural or forced, can be used for

propulsion purposes Additionally, machinery

reliability forms a further variable for

consideration

9.8 Of the available machinery options analysis of

historical failure trends shows that the slow

speed diesel propulsion systems exhibit the least

incidence of failure

9.9 Slow speed and medium speed diesel engine

technology are well understood in the merchant

marine industry Notwithstanding this, gas

turbine propulsion systems have been used with

success in the navies but require a different

maintenance philosophy

10 ACKNOWLEDGEMENTS

The author is grateful to the Committee of Lloyd’s

Register for permission to publish this paper In addition,

many of Lloyd’s Register’s engineers and scientists have

been involved in various parts of the work and thanks are

due to them In particular Mr A Boorsma, Mr P.A

Fitzsimmons and Mr R McAllister deserve particular

mention

11 REFERENCES

1 CARLTON, J.S The Propulsion of a 12500 teu

Container Ship I.Mar.EST, London, Jan 2006

2 CARLTON, J.S AND FITZSIMMONS, P.A Full Scale Cavitation Observations Relating to Propellers

CAV 2006, Wageningen, The Netherlands, Sept

Manoeuvres In the course of publication.

5 PIEN P.C., AND STROM-TEJSEN, J A Proposed

New Stern Arrangement NSRDC Rep 2410 Washington , May 1967

6 KERLEN, H., ESVELDT, J AND WERELDSMA,

R Propulsion, Cavitation and Vibration Characteristics of Overlapping Propellers for a

Container Ship Jahrbuch der Schiffbautechnische Gesellschaft e.v Vol 64, pp 301-341, 1970

7 RESTAD, K., VOLCY, G.C., GARNIER, H AND MASSON, J.C Investigation on Free and Forced Vibrations of an LNG Tanker with Overlapping

Propeller Arrangement Trans SNAME, New York

Nov 1973

8 MOORES, C., VEITCH, B., BOSE, N., JONES, S

AND CARLTON, J.S Multi-Component Blade Load Measurements on a Propeller in Ice Trans

SNAME

9. CARLTON, J.S Podded Propulsors: Some Results

of Recent Research and Full Scale Experience

Trans LRTA Session 2005-2006 Lloyd’s Register, London

OPERATIONS, DESIGN REQUIREMENTS AND INNOVATIVE TECHNOLOGIES FOR GAS COMBUSTION UNITS FOR THE NEW GENERATION OF LNG CARRIERS

D Feger and N Martin, Snecma - Space Engine Division, France

D Julien, North American Stordy Combustion, France

SUMMARY

The new generation of LNG carriers use Dual Fuel Diesel Electric or Slow speed Diesel combined with an on board re liquefaction plant Compared to previous Steam turbines LNG carriers propulsions systems, where excess boil off gas coming from the cargo tanks could be burned in the boiler and the corresponding excess steam dumped in the condenser, these new types of propulsion systems require either in normal operation or as a back up, a capability to dispose of the excess boil off gas, which cannot be used as fuel or treated by the re liquefaction plant, in a safe and environmentally friendly way This is provided by specific equipment, the Gas Combustion Unit (GCU)

This paper presents a review of the corresponding requirement’s, the main design features and operational performances

of the GCUs proposed by Snecma based on North American Stordy burner technology and already on order for two vessels being built in Japanese shipyards

1 INTRODUCTION

1.1 PRESENTATION

Liquefied Natural Gas (LNG) carriers are part of the

LNG chain, which is based on three links:

x the liquefaction terminal, in the producing country,

which purifies, liquefies and stores (under ambient

pressure and cryogenic temperature) the natural gas

prior to its loading into the LNG carrier,

x the LNG carriers, which ship the LNG from the

loading terminal to the off-loading one,

x the regasification terminal, in the gas consuming

country, which stores, pressurises and regasifies the

LNG prior to injecting it into the gas pipe, which

distributes it to the gas consumers

In LNG carriers, the liquefied gas is stored in a boiling

state, at cryogenic temperature (- 160°C) slightly above

the atmospheric pressure in insulated tanks Due to the

heat leaks getting though this insulation into the liquefied

gas, a part of the cargo is boiling off the tanks (typically

0,1 to 0,3 % per day)

To avoid wastage of this boil off vapours, the thermal

performance of the insulation is usually optimised so that

the boil off vapours flow can be used to provide part of

the ship's propulsion needs when it is on its way For this

purpose the propulsion system is of a dual type,

compatible with the use as fuel of either the heavy oil

either, when available, the natural gas boil off vapours

coming from the cargo tanks

When the ship propulsion requirements are reduced,

during harbour manoeuvres or at anchor for example, the

boil off vapours exceeds the propulsion needs, although

the cargo tank pressure has to be kept within acceptable

limits To dispose of this excess boil off and avoid a

pressure rise in the cargo tanks, several strategies can be

considered:

x implement an on board re liquefaction plant which

re liquefies the vapours and send back to the cargo tanks the boil off vapours in a liquefied state

x dispose of this excess boil off by burning it in an

on board thermal oxidiser complying with safety and environmental regulations which do not allow direct release of natural gas into the atmosphere for both safety and environmental concerns (green house gas effect of methane which is very significantly higher than the one of carbon dioxide) The standard approach: steam turbine propulsion:

Up to now, most LNG carriers strategy has been to use for this reason a steam turbine propulsion system as it allows to use either heavy oil or boil off vapours for fuel, the steam boiler being equipped with heavy oil and natural gas burners This propulsion system had the further advantage that the excess boil vapours could be disposed of directly in the steam boiler, the corresponding excess steam being sent to the sea water cooled condenser rather than to the propulsion turbine, without requiring any specific equipment, other than a bypass valve towards the condenser to fulfil this additional boil off disposal function

Although it has been considered attractive for LNG carriers for decades, this steam propulsion approach is very much challenged nowadays as it has the following drawbacks:

x compared with other propulsion modes, such as Diesel or Gas turbine, it is bulkier, and therefore leads, for the same hull size, to a lower shipping capability,

x its fuel efficiency is 30 % instead of 45 %,