Proceedings of the 15th International Ship and Offshore Structures Congress: 3-volume set

P R O C E E D I N G S OF THE 15th INTERNATIONAL SHIP AND O F F S H O R E STRUCTURES CONGRESS

VOLUME 2

Elsevier Internet Homepage:

http://www.elsevier.com Consult the Elsevier homepage for full catalogue information on all books, journals and electronic products and services. Elsevier Titles of Related Interest Ocean Engineering Series

WATSON Practical Ship Design ISBN: 008-042999-8 YOUNG Wind Generated Ocean Waves 1SBN: 008-043317-0 BAI Pipelines and Risers ISBN: 008-043 712-5 JENSEN Local and Global Response of Ships ISBN: 008-043953-5 TUCKER & PITT Waves in Ocean Engineering ISBN: 008-043566-1

OWENS Steel in Construction (CD-ROM Proceedings with Printed Abstracts Volume, 268 papers) ISBN: 008-042997-1 GODOY Thin-Walled Structures with Structural Imperfections: Analysis and Behavior ISBN: 008-042266-7 FUKUMOTO Structural Stability Design ISBN." 008-042263-2 GUEDES-SOARES Advances in Safety and Reliability (3 Volume Set) ISBN: 008-042835-5

Other Titles

CHAN & TENG ICASS '02, Advances in Steel Structures (2 Volume Set) ISBN: 008-044017-7

MOAN & BERGE 13th Int Ship & Offshore Structures Congress (ISSC 1997) ISBN: 008-042829-0

OHTSUBO & SUMI 14th Int. Ship and Offshore Structures Congress (ISSC 2000) ISBN." 008-043602-1

Related Journals Free specimen copy gladly sent on request. Elsevier Ltd, The Boulevard, Lang/'ord Lane, Kidlington, Oxford, OX5 1GB, UK

Applied Ocean Research Advances in Engineering Software CAD Coastal Engineering Composite Structures Computers and Structures Construction and Building Materials Engineering Failure Analysis Engineering Fracture Mechanics

Engineering Structures Finite Elements in Analysis and Design International Journal of Solids and Structures Journal of Constructional Steel Research Marine Structures NDT & E International Ocean Engineering Structural Safety Thin-Walled Structures

To Contact the Publisher

Elsevier welcomes enquiries concerning publishing proposals: books, journal special issues, conference proceedings, etc. All formats and media can be considered. Should you have a publishing proposal you wish to discuss, please contact, without obligation, the publisher responsible for Elsevier's civil and structural engineering publishing programme: James Sullivan Publishing Editor Elsevier Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB, UK

Phone: Fax: E.mail:

+44 1865 843178 +44 1865 84392 [email protected]

General enquiries, including placing orders, should be directed to Elsevier's Regional Sales Offices - please access the Elsevier homepage for full contact details (homepage details at the top of this page).

P R O C E E D I N G S OF THE 15th I N T E R N A T I O N A L SHIP AND O F F S H O R E S T R U C T U R E S C O N G R E S S

VOLUME 2

Edited by A.E. MANSOUR

University of California, Berkeley, USA and R.C. ERTEKIN

University of Hawaii, Honolulu, USA

.qy~oBESTt~tH^

o~Y" ~

~,~'~'~

IEGO,

2003 ELSEVIER Amsterdam

- Boston - Heidelberg - London - New York - Oxford-

San Diego - San Francisco - Singapore - Sydney - Tokyo

Paris

ELSEVIER SCIENCE Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB, UK 9 2003 Elsevier Science Ltd. All rights reserved. This work is protected under copyright by Elsevier Science, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier's Science & Technology Rights Department in Oxford, UK: phone: (+44) 1865 843830, fax: (+44) 1865 853333, e-mail: [email protected]. You may also complete your request on-line via the Elsevier Science homepage (http://www.elsevier.com), by selecting 'Customer Support' and then 'Obtaining Permissions'. In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (+1) (978) 7508400, fax: (+1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (+44) 207 631 5555; fax: (+44) 207 631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of Elsevier Science is required for external resale or distribution of such material.Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher.Address permissions requests to: Elsevier's Science & Technology Rights Department, at the phone, fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. First edition 2003 Library of Congress Cataloging in Publication Data A catalog record from the Library of Congress has been applied for. British Library Cataloguing in Publication Data A catalogue record from the British Library has been applied for. ISBN:

0-08-044076-2

The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Printed in The Netherlands.

PREFACE

This volume contains the 6 Specialist Committee and 2 Special Task Committee reports that will be presented and discussed at the 15th International Ship and Offshore Structures Congress (ISSC 2003)in San Diego, USA, 11-15 August, 2003. Volume 1 contains the 8 Technical Committee reports. Volume 3 will include the discussions of the reports, the chairmen's reply, the text of the invited lecture and the congress report of ISSC 2003, and it will appear in 2004. The Standing Committee of the 15th International Ship and Offshore Structures Congress in San Diego is: Chairman:

Secretary:

Prof. A.E. Mansour Prof. J.L. Armand Prof. B. Boon Dr. M. Dogliani

USA France The Netherlands Italy

Prof. W. Fricke

Germany

Dr. P.A. Frieze

UK

Prof. Prof. Prof. Prof.

Korea Poland Denmark Norway

C.D. Jang T. Jastrzebki J.J. Jensen T. Moan

Prof. H. Ohtsubo Dr. N. Pegg Prof. Y.S. Wu Prof. R.C. Ertekin

Japan (ex officio) Canada China USA

On behalf of the Standing Committee and members of the ISSC, I would like to thank the American Bureau of Shipping and the Ship Structure Committee for their financial support of ISSC 2003. The support of the City of San Diego is also gratefully acknowledged. Berkeley, USA March 2003

Alaa E. Mansour Chairman

This Page Intentionally Left Blank

CONTENTS

Preface

. . . . . . .

,

. . . . .

. . . . . . . . . . . . . .

.

. . .

.

V

REPORT OF SPECIALIST COMMITTEE V1:

RISK ASSESSMENT

REPORT OF SPECIALIST COMMITTEE V2:

INSPECTION AND MONITORING

REPORT OF SPECIALIST COMMITTEE V3:

COLLISION AND GROUNDING

.

71

REPORT OF SPECIALIST COMMITTEE V4:

STRUCTURAL DESIGN OF HIGH SPEED VESSELS . , . . . .

109

REPORT OF SPECIALIST COMMITTEE VS:

FLOATING PRODUCTION SYSTEMS

149

REPORT OF SPECIALIST COMMITTEE V6:

FABRICATION TECHNOLOGIES

189

REPORT OF SPECIAL TASK COMMITTEE V1.l: FATIGUE LOADING

.

. .

.

1

37

.

235

REPORT OF SPECIAL TASK COMMITTEE V1.2: FATIGUE STRENGTH ASSESSMENT 285 Indexes

. . . . . . . .

.

. . . . . . . . . . . . . .

vii

.

. . . .

327

This Page Intentionally Left Blank

15th INTERNATIONAL SHIP AND OFFSHORE STRUCTURE CONGRESS 2003 11-15 AUGUST 2003 SAN DIEGO, USA VOLUME 2 /EGO,

SPECIALIST C O M M I T T E E V.1

RISK ASSESSMENT

MANDATE Concern for the development of rational procedures for qualitative and quantitative risk assessment of ships. This shall include assessment of probability and consequence of accidental situations as well as evaluation of measures to control and mitigate the risk. Particular attention shall be paid to fire and explosion, extreme environmental condition, human element, traffic and obstructions, and operational hazards.

MEMBERS Chairman:

Dr. William Moore Professor Y. Chert Mr. A. Dinovitzer Professor O. Litonov Dr. Marc Prevosto Dr. Angelo Tonelli Professor Y.S. Yang Mr. Koichi Yoshida

KEYWORDS Risk assessment, risk analysis, formal safety assessment, hazard, accident, consequence, frequency, probability, cost, benefit

This Page Intentionally Left Blank

CONTENTS

1 INTRODUCTION

.

5

2 R E V I E W OF R I S K A S S E S S M E N T ACTIVITIES IN THE M A R I T I M E INDUSTRY 2.1 Regulatory . . . . . . . . . 2.1.1 Formal Safety Assessment (FSA) for the Maritime Rule Making Process 2.1.2 Risk Acceptance Criteria for IMO . . . . . 2.1.3 Bulk Carriers . . . . . . . . 2.1.3.1 F S A Study on Bulk Carrier Safety by Japan . 2.1.3.2 F S A Study on Life-Saving Appliances by Norway . . 2.1.3.3 FSA Study on Bulk Carrier Safety by Internationally Collaborated Group . . . 2.1.3.4 Decision Making at IMO . . 2.1.3.5 Other F S A Studies for Bulk Carriers 2.1.4 Passenger Ships . . . . . . . . 2.1.5 Maritime Security . . 2.2 Industry . . . . . . . . . 2.2.1 International Association o f Classification Societies (IACS) F S A Training 2.2.2 Guidance Publications on FSA . . . 2.2.3 Incorporation of Safety Assessment into the Rule Making Process 2.2.4 Application o f Risk Assessment to Icebreakers . 2.2.5 Joint Research Team on FSA . . . 2.2.6 Alternative Design and Arrangements for Fire Safety . 2.2.7 Marine Insurance Industry: Risk Assessment and Risk Selection 2.3 Applications . . . . 2.3.1 Risk Based Fire Safety Design . . . . . . . . 2.3.2 Event and Fault Tree Application . . . . . . . . . . 2.3.3 Fuzzy Set Modelling and its Application to Maritime Safety 2.3.4 F S A for Safety o f Coastal Trading Ships in Japanese Waters 2.3.5 Safety o f Ships Carrying Irradiated Nuclear Fuel . . . . . . . 2.3.6 Alert Communication from Small Craft Using Cellular Phones . .

8 8 13 13 13 14 14 14 14 15 16 19 19 21 21 22 24 25 26 26

3 E L E M E N T S OF R I S K A S S E S S M E N T . . . . . . . . . . . 3.1 Uncertainty of Data . . . . . . . . . 3.2 Decision Making Process based on the Results o f FSA . . 3.3 Effect of Safety Measures that have not Appeared in Historical Casualty Data 3.4 H u m a n Element . . . . . . .

27 27 28 29 29

4

Specialist Committee V.1

4 CONCLUSIONS

.

.

.

5 RECOMMENDATIONS

. .

. .

. .

. .

. .

. .

. .

.

.

.

.

.

.

.

. .

. .

. .

. .

. .

. .

29 .

.

29

APPENDICES . . . . . . . . . . . . . . . . . . . . . A p p e n d i x 1: Indices for Cost Effectiveness Analysis ( C E A ) . . . . . . . . A p p e n d i x 2: C o m b i n a t i o n o f R C O s and the Effect . . . . . . . . . . .

30 30 31

REFERENCES

32

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

Risk Assessment

1

5

INTRODUCTION

The application of risk assessment has evolved over 20 years in the offshore industry and within the last 5 years in the marine industry albeit in different directions. The offshore industry has focused on the application of risk assessment to evaluate the safety of individual offshore constructions. The marine industry has primarily focused on the applications of risk assessment to further enhance and bring greater clarity to the international rule making process. ISSC established, at its conference in 1997 in Trondheim, Norway, Risk Assessment. At the time of the first report of this Committee (Yoshida et al, 2000) the International Maritime Organization (IMO) risk assessment methodologies and techniques for the first time. The the application of risk-based approaches in the offshore industry.

Specialist Committee V.1, at ISSC 2000 in Nagasaki was agreeing to the use of 2000 report also addressed

This report provides the status of the application of risk assessment with a specific focus on the marine industry and provides insight into the direction that the industry is following in the research, development and application.

2

REVIEW OF RISK ASSESSMENT ACTIVITIES IN THE MARITIME INDUSTRY

Yoshida et al (2000) provides a review of the fundamentals for the application of FSA to the IMO rule making process. Since the development of formal safety assessment (FSA) approaches at IMO, there has been a wide range of activities associated with applying these techniques. This chapter provides a brief summary of these activities. 2.1

Regulatory

2.1.1

Formal Safety Assessment (FSA) for the maritime rule making process

As mentioned in the Yoshida et al (2000), IMO established interim guidelines for the application of FSA in IMO MSC/Circ.829 and MEPC/Circ.335 (IMO, 2002a). To date, these interim guidelines have been used, as trial basis, in several risk assessment in conjunction with IMO rule making process. Then, IMO decided to improve the interim guidelines whilst taking into account the experiences obtained through trial application. The Maritime Safety Committee (MSC) of IMO established a correspondence group to revise the interim FSA Guidelines. The group agreed to further include the following into the Guidelines. integration of analysis for human element through human reliability analysis (HRA); and 9 risk evaluation criteria. With regard to the human element, the group agreed that the HRA guidance developed by the International Association of Classification Societies (IACS) should be incorporated into FSA Guidelines, as an appendix. With regard to risk evaluation criteria, the group did not reach any firm conclusion. However, this topic was discussed at 74th session of IMO MSC, and it was agreed that Gross Cost of Averting a Fatality (Gross CAF or G-CAF) and Net Cost Averting Fatality (Net CAF or N-CAF) were most relevant for cost benefit assessment and that G-CAF and N-CAF should be used for comparison among risk control options (RCOs) in relation to the safety of life, and were included in FSA guidelines. In addition, it was further agreed that other indices are necessary to consider RCOs for

6

Specialist Committee E1

reducing the affect on property and the environment. This issue are remained for future consideration. The record of the discussion in the correspondence group was presented to 74th session of the MSC (2001a, 2001b). The record of the discussion at MSC 74 is given by paper of IMO (2001c) and further summarised in Gard Services (2001). IMO has since agreed, in both MSC and Marine Environmental Protection Committee (MEPC) to a final set of FSA guidelines as provided in IMO (2002a). 2.1.2

Risk acceptance criteria for IMO

As part of the FSA initiatives, recent efforts have also addressed the issue of risk acceptance criteria. As noted in Skjong (2002) it is difficult to make risk-based decisions without using or disclosing risk criteria. Risk acceptance criteria is of particular importance to IMO and efforts are currently underway to provide 'explicit' acceptable risk criteria. Skjong and Eknes (2001, 2002) provide an outline from which societal risk acceptance criteria may be established based on similar activities within other industries with similar maritime comparisons made for various ship types. Risk acceptance criteria will continue to be on the forefront of IMO related activities in the coming years. 2.1.3

Bulk carriers

IMO, recognizing the importance of enhancing the safety of bulk carriers, had considered and developed provisions, which were adopted as Chapter XII of 1974 International Convention for the Safety of life at Sea (SOLAS 74), as amended, at a SOLAS Conference held in November 1997. The Conference also adopted several resolutions concerning the safety of bulk carriers. Taking the resolutions into account, IMO MSC, at its 69th session in May 1999, agreed that it should further consider safety of bulk carriers. At the 70th session of MSC in December 1999, the United Kingdom offered a plan of conducting an internationally collaborated FSA study regarding bulk carrier safety. At that session, Japan announced that it would also conduct an FSA study on bulk carrier safety by itself. 2.1.3.1 FSA study on bulk carrier safety by Japan Since January 1999, a research committee (RR74BC-WG) in the Shipbuilding Research Association of Japan has been established under the supervision of the Ministry of Land, Infrastructure and Transport (MLIT) in co-operation with participants of the representatives of ship-builders, ship owners and operators, ship masters, officers and crew, the Japanese Coast Guard, National Maritime Research Institute and Class NK, for the purpose of conducting the FSA study on bulk carrier safety. The research committee conducted the FSA study, according to the FSA Guidelines in IMO (2002a), on typical bulk carriers with have topside tanks and hopper side tanks in the cargo spaces. The size of the bulk carriers under study was categorized into 4 groups by deadweight tonnes, (i.e. cape size, panamax size, handy size and small handy size). The casualty data-base was provided by Lloyd's Maritime Incident Service and Class NK was used. The results of the FSA study including final recommendations have been reported to IMO (2002b, 2002c and 2002e). The final recommendations for decision-making from the study are as follows: .1 The risk level of whole bulk carriers in future would stay at a relatively upper part of the 'As low as reasonably practicable' (ALARP) region even after recently adopted RCOs of SOLAS

Risk Assessment

7

chapter XII are implemented. Moreover, it is higher than other types of ships such as tankers and container ships. Therefore, IMO should pursue further cost effective safety measures that could reduce the risk of bulk carriers to ALARP (See Figure 1). .2 The risk level of the bulk carriers less than 150m in length is higher than that of the other size of bulk carriers. RCOs for mitigating consequences after hold flooding as required in SOLAS Chapter XII are not appropriate for those ships because only one hold flooding is fatal for bulk carriers of less than 150 m in length if the number of cargo holds of current design practice for such smaller ships can not be changed. Therefore, measures to prevent flooding is more important for such smaller bulk carriers.

1.0E-02 . . ~

.

.

.

.

.

.

.

.

.

.

.

.

,-a ~,~sr~ -..~..- T an ker ---m--Ge

[ 1.0E-03

~ - ~

ill

!

[

ne

1.0 F, 02

ralC arg~

J [[1 ] mmnmBulkCarri er

i ~-~'~~..

fi

""

~ ,.o~o~

!

2 o

1.0 F, 04

Z

i '"l'.,i ~'

"~

i,,i).,,j?~Jll!

i' , ~ l ]

1

i ["f-?,,~lii ~ ] I, I"-~!! .0F,05

i 1.0E-06

....~,[i! S

i; 1

10

100

i~ .01g06 1000

Basic risk level to date

Risk level after implementation of SOLAS XII and Enhanced survey Figure 1 F-N Curves of each ship type (IMO, 2002b)

.3 SOLAS Chapter XII can be justified based on the comparison of the relative cost effectiveness versus other relevant RCOs such as a mandatory requirement of double side skin. Exemption of double side skin bulk carriers from SOLAS Chapter XII is also justified based on the same comparison and consideration on the magnitude of risk of accidents for double side skin bulk carriers. .4 For single side skin bulk carrier of 150m and over in length, it is expected that preventive measures against water ingress from a breach of the side shell structure would be effective to reduce the risk. According to the cost effectiveness assessment, it is recommended that corrosion control requirements such as an increase of corrosion margin and preventive coating should be considered, since it was found to be more cost-effective than double side skin. In summary, further investigation on following RCOs was recommended: (1) increased corrosion margin (Design Stage) and (2) corrosion control of single side skin for vessels in service.

2.1.3.2 FSA study on life-saving appliances by Norway An FSA project on life-saving appliances for bulk carriers was carried out in Norway by Det Norske Veritas in co-operation with participants from Norwegian Maritime Directorate, Norwegian Union of Marine Engineers, Umoe Schat-Harding, Norwegian Shipowners' Association and International Transport Workers' Federation (IMO 200 ld).

8

Specialist Committee V.1

The hazard identification (step 1 of FSA) was carried out for conventional lifeboats, throw overboard liferafts, davit/crane launched liferafts and free-fall lifeboats. The study was considered representative for all SOLAS bulk carriers, with the exception of bulk carriers of less than 85 meters in length. The risk reduction effects of introducing free-fall lifeboats as a mandatory requirement was quantified, whilst conventional lifeboats were considered the base case in Step 4 of the FSA. The considered RCOs were: 9 shelter mustering and lifeboat area; 9 remote control of the ship from the mustering area; 9 level alarms to monitor water ingress in all holds and forepeak; 9 individual immersion suits to all personnel; 9 free-fall lifeboats; 9 free-fall lifeboat with an additional free float mode; 9 marine evacuation systems for throw overboard liferafts; 9 enclosing open lifeboats for all existing ships with open lifeboats; 9 redundant trained personnel; and 9 improved pick-up function (crane). After carrying out an extensive review of historical data and completing all steps of the FSA, it was concluded that the following RCOs were providing considerable improved lifesaving capacity in a cost-effective manner: .1 Free-fall lifeboats with an additional free float mode. This solution was slightly better than the flee-fall lifeboat solution itself and considered cost effective (new ships). .2 Water level alarm with continuous water level indication in all holds and fore peak (new and existing ships) .3 Personal immersion suits to all personnel (new and existing ships) It was noted that the success rate in evacuation from existing ships remain rather low also after implementing the suggested RCOs. This might call for additional measures, in particular, focused on crew competence and training.

2.1.3.3 FSA study on bulk carrier safety by internationally collaborated group At the 71st session of the Maritime Safety Committee (May 1999) it was agreed that in light of the investigation into the loss of the DERBYSHIRE, the UK proposed leading an initiative to perform a holistic FSA analysis on bulk carrier safety. Many papers were submitted to IMO and within the collaborative group. The final results of their initiative can be found in IMO (2002e). At the 76th session of the MSC (December 2002), a final list of risk control options was prepared for decision to be made by the MSC on how to proceed.

2.1.3.4 Decision making at IMO MSC 76 did not consider the combination of risk control options and their prioritization in terms of risk reduction and cost benefit assessment. Based on the outcome of the International Collaborative Bulk Carrier FSA study coordinated by the United Kingdom the following decisions were made as shown in Table 1 (Gard Services, 2002b). It was agreed that Gross Cost of Averting a Fatality (GCAF) and Net Cost of Averting a Fatality (NCAF) would be used as primary selection critiera. Appendix 1 provides a brief explanation as to the benefits of using

TABLE 1 BULK C A R R I E R S A F E T Y INITIATIVES A G R E E D T O A T T H E 7bTHSESSION O F THE M A R I T I M E S A F E T Y C O M M I l T E E (Gard Services. 2002b)

Application Risk control option Double side-skin construction (hull envelope)

Improve coatlngs (hull envelope)

Steel repair standards (hull envelope)

Iiold frames (hull envelope)

Summary

New ships

Agreed that this will apply to new shlps to be constructed lssucs such as unifonn intemstionnl rtandards for double side s k ~ n construction and coatings, strength of inner s h n s are to be cons~dered. Controls andlor performance For new ships of double side standards lor protective coatings construcuon improved coatings In rclatinn to compatibility with will he requ~redfor dedicated cargoeb. seawater ballast tanks and v o ~ d spaces w~thlndouble hull spaces coatlng. Thls I S to be done In accordance with current SOLAS requirements for ballast spaces. Coating In cargo spaces 1r lcft to the discretion of owner and classtficatton soclety. T~ghtercontrols on grades of stccl No requirements. but a draft and weld~ngrods used for ~n c~rcularreminding owners and operators of their responsib~llt~es service repalr. under SOLAS regulation 11-113-1 concerning provisions that s h ~ p s shall be maintamed In accordance with rhe structural rcqulremcnts of recognised class socteltes and other related ohl~gationsunder the ISM Code. Various measures were cons~dered Not appl~cable. for hold frame5 including reduced d~minutlonallowances. strengthening to comply with IACS UR S12 ( N ~ for S side structures in single skin bulk carriers) and coatings. Concerns regarding thc ingress via loss of slde s k ~ nstructural lntegnty leading to many accidents.

Existing ships

Status

Not applicahlc.

To be further developed by the Design and Equipment (DE) subcommittee and completed by 2004.

No new requirements. Suffic~ent control measures are already in place through thc enhanced survey programme.

T o he funher developed by the De%igiland Equlpment (DE) subcommittee and completed by 2004.

Same as for new ships.

To be further developed by the Destgn and Equipment (DE) subco~r~rn~ttee and completed by 2003.

No requirements. but governments will be urged via an MSC Resolution to ensure that all nonIACS classificaion soc~eties comply with the lACS UR S31 (rencw:~lcrlteria for s ~ d eshell frames for vessels not bull1 in compliance wlth the revised IACS UR S12.

To be funher developed by the Design and Equ~pmcnt(DE) subcommittee and completed by 2003

TABLE 1 (cont.) BULK CARRIER SAFETY INITIATIVES AGREED TO AT THE 76THSESSION OF THE MARITIME SAFETY COMMIlTEE (Gard Services, 2002b)

Application Risk control option Forecastle fittings (hull envelope)

Fore deck fittlngs(hul1 envelope)

envelope)

(hull envelope)

covers)

covers (hatch covers)

I

Summary Forecastle requirements to reduce green sea loads on forward end of vessel and protect foredeck fittings. Strengthen stud pipes for alr and vent pipes to be sufficient to withstand horizontal forces of green sea loading. Closing devices and strength of small hatches to be sufficient to withstand vertical and horizontal green sea load~ngsin accordance with standards being developed by IACS.

detecting water ingress into cargo holds and dry spaces forward with visual and audible alarms in permanently manned spaces.

double side construction.

rmnimum bow height and reserve

1988 Load Line Protocol on hatch cover loads.

New ships lACS ongoing development of Unified Requirement (UR) S28 requiring fitting of forecastles on bulk caniers contracted on or after 1st January 2004. No requirements, since it is already being addressed via two lACS URs: S26 (strengthening and securing small hatches on exposed foredecks) and S27 (strengthening requirements for foredeck fittings and equipment). Governments will be urge ships flying thelr flag to comply via an MSC Resolution to ensure that all non-IACS classification societies comply with the lACS URS. SOLAS regulation X11112, Hold, ballast and dry space water ingress alarms will apply to bulk carriers regardless of their date of construction and will enter into force on 1st July 2004. Shim of 150111in lenah and upwards of double side construction should also comply with all structural strength provision of regulation XI15 requiring that the ship shall have sufficient strength to withstand flood~ngof any one cargo hold. Amendments were made to Regulation 39 to enhance the requirements for minimum bow height and reserve buoyancy. A revised simplified formula for design wave loads on hatch covers for bulk caniers (Regulation 16.1 of the 1998 LL Protocol).

I

Existing ships Not applicable.

. -

requirement and.w!ll be delivered

1

I

Same as for new shivs. Design and Equipment (DE) subcommittee and completed by 2003.

2I?. s -. 2

1

Same as for new shivs. ingress alarms and these will be adopted at the next session of the MSC in June 2003.

1

..

Not av~licable.

Not applicable.

Design and Equipment (DE) subcommittee and completed by 2003.

Action taken at MSC 76 and the 1988 LL Protocol amendments were adopted.

Not applicable. 1988 LL Protocol amendments were adopted.

I

TABLE I (cont.) BULK CARRIER SAFETY INITIATIVES AGREED TO AT THE 76T" SESSION OF THE MARITIME SAFETY COMMITTEE (Gard Services, 2002h) Application Risk control option Redes~enlreinforcementof hatch covers (hatch covers)

lmmerslon suits (evacuation)

Frce fall lrfeboats wrth float-free mode (evacuation)

Early abandonment (evacuation)

Summary

I Redeslen of hatch covers and

New ships Not amlicable.

hatch cover securing arrangements for e x ~ s t ~ nshins g

securlng mechanisms to withstand both ven~caland horizontal loads. Personal immersion suits for all personnel onboard shlp

Slngle free-fall survlval craft wlth float-free capability enabling rapid evacuation of crew from ship. Consideration of guidel~nesfor when and ~fbulk carrlers should be abandoned vessels at an earller stage in the event of flooding.

Terminal interface improvement (operational)

Improvement of ship-shore communications, training of stevedores and terminal operators and better control of loading capabilities.

Port State Control training (operational)

Prov~sionsof special~sedtraining for pon State control Inspectors in bulk carrier design and operation, with panlcular emphasls on areas of vulnerab~l~ty.

Existing ships

I Standards are to be developed tor

be provided.

free fall llfcboats wlth lloat-free capability. Not applicable.

No requirements, but Governments are to be urged via an MSC circular to apply the Bulk Load~ngCode of Practice for Safe Loading and Unloading of Bulk Carriers (BLU) Code. In addition, a manual for terminal representatives is to be developed. No requirements, but an MSC circulir would be developed on this recommending that the various Port State Control Memoranda of Understanding (e.g. Paris MoU, Tokyo MoU, 1 etc.) develop speclalised tra~ning for pinpointing vulnerability within structures panicularlv for

be prov~ded.

Status

1

To be further developed by. the . Design and Equipment (DE) subcommittee and com~letedby

develop draft amendments to SOLAS chapter 111 andfor the Llfe

develop draft amendments to SOLAS chapter I11 and/or the L ~ f e

( No requirements, but a circular urglng shipowners to issue guldance to ship's personnel on the possible need for early abandonment of a bulk carrier that has any single hold flooded and in particular for vessels which may not withstand flooding of any one cargo hold. Same as for new ships.

Same as for new s h ~ p s

I Saving Appl~ances(LSA) Code. I To be further developed by the Design and Equipment (DE) subcommittee and completed by 2003.

T o be funher developed by the Dangerous Goods, Solld Cargoes and Containers (DSC) subcommittee and completed by 2003. The manual will be developed In conjunct~onwith the Sh~p-PonInterface Woking Group. To be developed by the Flag State (FSI) ' ~m~~ernentatrbn subcommittee and completed in 2004.

TABLE 1 (cont.) BULK C A R R I E R S A F E T Y INITIATIVES AGREED T O A T T H E 76T11SESSION O F T H E MARITIME S A F E T Y C O M M I T T E E (Gard Services, 2002b)

Application Risk control option

Summary

New ships

Improved loading/stability ~nformation(operat~onal)

Provision of detail, comprehensive and user-friendly information covering stability and stress characteristics of the ship's hull.

Alternative hold loading (operational)

Alternative hold loading has been observed to ~ncreasethe longitudinal stresses on ships panicularly for ships over a defined age.

Guidelines are to be developed for the provision of detail, comprehensive and user-friendly information covering stability and longitudinal stress characteristics of the ships hull during loading and unloading. Not applicable.

Availability of pumping systems (operational)

Availab~lityof draining and pumping ballast tanks forward of the collision bulkhead, and bilges of dry spaces any part of which extends forward of the foremost cargo hold shall be capable of being brought into operation from a readily accessible enclosed space accessible from the navigational bridge or propulsion machinery control position without transversing exposed freeboard or superstructure decks.

Not applicable.

Existing ships

Status

Same as for new ships.

To be developed by the Stability and Load Line and F~shingVessel Safety (SLF) and DE subcommittees and completed in 2004.

Consideration is to be given as at what age alternative hold loading will be banned. This will be based on the types of cargoes and may include requirements based on a successful completion of a condition assessment. Bulk carriers constructed on or before 1st July 2004 shall comply with the requirements of this regulation no later than the date of the first intermediate or renewal survey of the ship to be carried out after I July 2004, but in any case, no later than 1st July 2007.

To be developed by the DSC subcommittee and will repon their findings and recommendations in 2004.

Agreed to by MSC 76.

Risk Assessment

13

GCAF and NCAF for the RCO selection process. In addition, criticisms were made regarding the lack of proper consideration of RCOs chosen in combination. Appendix 2 provides some description of how RCOs chosen in combination should be considered in the selection process.

2.1.3.5 Other FSA studies for bulk carriers Further papers addressing various FSA related activities regarding the IMO study on bulk carriers and other related ship reliability aspects include Bitner-Gregersen et al (2002), Lee, J.O. (2001), Lee, J.O. et al (2001), Skjong (2002), Skjong and Wentworth (2001), and Yeo et al (2000).

2.1.4

Passenger ships

FSA activities are also being considered for application to the safety of large passenger ships as discussed in Lee, J.K. (2002). The Netherlands addressed the issue of finding the root accident causes through systematic events analysis in IMO (2002d). The investigation identified operations and management as being the main causal category whilst navigational and watchkeeping figured as the main sub-category within the operations and management segment. The pattern for large passenger ships was found to be a clearer version of the general pattern. Particular efforts are being progressed in the area of evacuation studies for safety of large passenger ships as described in Boer and Skjong (2001), ICCL (2000), Park et al (2002). Due to the limitations of regulations, many research teams are developing new evacuation models based on microscopic simulation. Korea Research Institute of Ships and Ocean Engineering (KRISO) launched 2-year project to develop a new maritime evacuation model, IMEX (Intelligent Model for Extrication simulation). In the IMEX project an evacuation model was defined and briefly discusses some models and their limitations. Also, the report focuses on describing the configuration, feature, and status of IMEX that is designed to overcome those limitations. SOLAS Chapter II/2 (fire protection) was recently amended including the possibility, through regulation 17, to deviate from prescriptive fire protection requirements of SOLAS (Regulation II2/17 "Alternative design and arrangements"). Proposed alternative design and arrangements must achieve a fire safety level at least equivalent to the prescriptive design criteria in SOLAS. In order to provide this demonstration, a risk-based fire engineering analysis is to be carried out according to IMO's Guidelines as laid down in MSC Circ./1002.

2.1.5

Maritime security

In December 2002, the IMO adopted amendments to the SOLAS Convention to address maritime security in wake of the terrorist attacks in the United States in September 2001. These amendments require security assessments for both ships and port facilities using the requirements developed in the International Ship and Port Security Code. Although no specified criterion has been established on how these assessments are to be performed, the Code does specify criteria that should be considered in making these assessments. The USCG (2002) and Gard Services (2002, 2003) outline the current status of these activities and provide guidelines on outlining these criteria. Gard Services (2002, 2003) highlights that ports facility assessments and security plans with three proposed levels of security have been suggested: Level 1 (low level security measures required), Level 2 (additional security measures required) and Level 3 (high security measures required). Authorised Port Security personnel in accordance with general requirements for the appropriate security level will carry out assessments. These requirements will include a security assessment, a security plan, a designated port facility security officer and security training and drills. It is

14

Specialist Committee V.1

generally agreed that the assessment of an appropriate security level is a matter for national administrations and the ship and port facility security plans should allow for changes in security levels. Owners will be required to obtain an International Ship Security Certificate issued by the flag Administration for each ship indicating compliance with the mandatory sections of the ISPS Code. Compliance with the ISPS Code will require: 9 development of a ship security assessment and plan; 9 documenting training, incidents, breaches of security, maintenance and calibration of security equipment records, etc.; 9 a designated and properly trained ship security officer; 9 a designated and properly trained company security officer to co-ordinate the company security plan; and 9 training and drills for ships and companies to respond to terrorist threats. Companies will be expected to follow established procedures in keeping copies of all ships' documents, certificates and plans ashore as already required by the International Safety Management (ISM) Code for other types of documentation. The company or ship's personnel will conduct ship security assessments. The assessment addresses the security risk level to be levied for each ship or each class of ships as a prerequisite for the development of the ship security plan (SSP).

2.2

Industry

2.2.1

International Association of Classification Societies (IACS) FSA training

In 2001, IACS developed a standardised training course on FSA in order to establish a common understanding of FSA within the maritime community, provide a basis of information of the sequence of analysis steps and demonstrate the uses of various FSA techniques. This training course is provided in two levels: Level 1 provides a high-level overview of the FSA process and Level 2 (9 modules) provides a detailed overview that includes exercises and case studies. More information on the IACS FSA training course can be found at the following website http://www.i acs.org.uk/t'sw'wlp5/fsatrainin g.htm.

2.2.2

Guidance publications on FSA

The American Bureau of Shipping (2000) has developed the Guidance Notes on Risk Assessment Application for the Marine and Offshore Oil and Gas Industries. These Guidance Notes provide an overview for managers and technical professionals for application of risk assessment to the maritime and offshore industries. This guidance introduces the concept of risk and introduces risk assessment tools that can be used in risk determination.

2.2.3

Incorporation of safety assessment into the rule making process

In 2001, the Russian Register (RS) published the Rules for Classing, Building and Equipping Offshore Drilling Units and Sea Stationary Platforms. For the first time in the world practice RS introduced the chapter "Safety Assessment", incorporating the following factors: 9 risk identification; 9 the concept of analysing hazard situations; 9 methods for analysing accident situations; 9 methods for risk quantitative assessments;

Risk Assessment 9 9 9 9 9 9 9 9 9

15

statistic models of accident situations; assessment of individual and social risks; control of risks; selection of risk control options; cost evaluation associated with measures for reducing risks; platform sufficient safety criteria; recommendations for accepting solutions on reducing accident risks; principle of as low as reasonable practicable level (ALARP); and negligible and unacceptable risk levels.

In the future, RS intends to introduce a similar chapter in the Rules for Classing and Building Transport Ships.

2.2.4

Application of risk assessment to icebreakers

Appolonov et al (2001) inform that in the Rules in force of Russian Maritime Register of Shipping (2001) the following two basic principles using safety notions are applied to ice-going ships (lOS): 9 the ice category is considered as a ship safety guarantee in specified permissible ice service conditions (safety guarantee principle); and 9 within the ice category all ships independently of their main dimensions, hull lines and configuration should have equal permissible ice service conditions (unified safety standard principle). The analysis of the data of accident rate for the Russian ice fleet has permitted to disclose the following types of emergency conditions: ship wreck in ice; breach in outer shell; mass damage of ice strengthening structures. As a result of processing the ice damage statistics, a statistical model including the main service factors as parameters has been developed. The statistical model is based on the assumption that IGS service under impermissible ice conditions (IIC) is a main cause of mass ice damage. The probability of service under IIC was determined by the following relationship:

Pc,,c = P~,c "Pc

(1)

where P~,c=probability of ship getting in IIC and Pc =conventional probability of the situation that after ship getting in IIC a ship navigator decides to continue ship operation on the specified route. The following expression was obtained for function P1~c, approximated by two-parametric Weibull's law:

P1,c(K)=exp(ln(Pl,,,g,/3XK-Kp~r) 2' )

(2)

where K = index of the actual navigation type: K=I - e a s y navigation" K = 2 - middle navigation; K = 3 - heavy navigation; K > 4 - extreme navigation" Kpe r = K perO . ~ f " ~ f = function considering additional strength reserves in comparison with the ones required by the RS Rules" ~/i > 1" KperO = index of the permissible navigation type by Table 2; Plon~ = probability of long-range forecast for one type" Pt,mxl =Pitc (K,)"

K 1 -- Kper_

. .

16

Specialist Committee V.1 TABLE 2 VALUES OF PARAMETER KperO (Appolonov et al, 2001) Category LU4

Way of operation IO

Winter-spring navigation B 1

K 0

L 0

ES 0

2,5 1 0 0 2 1 0 0 3,5 1,5 1 1 3,5 1 1 1 IO LU6 4 2,5 1,5 1,5 4 2 1 1 IO LU7 4 4 3,5 3,5 4 4 2,5 2,5 IO LU8 4 4 4 4 4 4 4 4 IO LU9 4 4 4 4 B - The Barents Sea; K - The Kara Sea; L - The Laptev Sea; VS - The East-Siberian Sea; Ch - The Chuckchee Sea; IO - independent operation; PI - pilotage by icebreaker. LU5

IO

Ch 0

0,5 0 1,5 1 2,5 2 3,5 3,5 4 4 4

B

4 4 4 4 4 4 4 4 4 4 4 4

Summer-fall navigation K L ES 2 1 1 3,5 2 2 3 2 2 3,5 3,5 3,5 4 3 3 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4

Ch 2 2,5 2 3,5 3 4 4 4 4 4 4 4

A quantitative estimate of the ship's design risk and assessment of the effectiveness of measures for risk monitoring can be made and thus decrease the scope the following FSA steps.

2.2.5

Joint Research Team on FSA

In 1995, the China Classification Society (CCS), Nippon Kaiji Kyokai (Class NK), the Indian Register of Shipping (IRS) and the Korean Register (KR) have formed a Joint Research Team (JRT) to address matters related to FSA. The JRT/FSA tasks were assigned to focus on: 9 safety assessment of passenger ship fire; 9 study on bulk carrier safety due to structural defects of the hull structures in cargo area; 9 A New Idea Considering FSA on a new tool or methodology applying FSA; 9 Application of formal safety assessment to rule making process; 9 Application of formal safety assessment to electro-hydraulic steering gear; 9 Application of FSA methodology to ship structure (especially for bulk carriers); 9 Safety assessment of engine room's fire of ships; 9 Safety assessment of collision and stranding of ships; 9 Safety assessment of fire in cargo holds; and 9 Database related to FSA. The main objectives of the above tasks were: i) identification of risk to be considered; ii) procedure of probabilistic approach to ships; iii) analytical method of risk advancement to accident; iv) acceptable risk level for ships; v) data base for all risks to be considered; and vi) case study of assessment The above research achievements were introduced into the Guidance of Application of FSA drafted by Dasgupta et al (1998, 1999) on behalf of the four classification societies above and the Guidance was made public in 2002 in order to indicote application of FSA at maritime community in Asia. Some achievements were applied to many aspects in China, Japan, Korea and India, e.g.

Risk Assessment

17

the methodologies adopted in the submissions to IMO provided by these countries mentioned above were from the Guidance. The Guidance included the following aspects: 9 The Guidance for Application of FSA (Version 1) containing details on the process of actual application of FSA as supplement to IACS Training Modules. It also addresses the issues on Quality Assurance of the FSA process. 9 Considering that IMO has formally issued amended Guidelines for FSA application (MSC Circ. 1023/MEPC Circ. 392) incorporating the HRA issues. This aspect is to be incorporated in the Guidance. 9 IACS has also evolved a basic Glossary of Terms on FSA that has been put up to IMO (MSC 76/INF.3). 9 The IACS AHG/FSA unanimously agreed that the IACS and the Japanese studies in which the main methodologies come from are more effective RCOs for bulk carriers. 9 A few issues have raised some controversy about the acceptability of the RCOs after the cost-benefit assessment. There has been an opinionated discussion on the use of GCAF and NCAF as the criterion of acceptance. Similar issues and other anomalies observed in the study reports strongly suggest the need for the introduction of process oriented quality assurance (QA) approach during the FSA study. Along with HRA study and updating the Guidance document, the application of QA to FSA can form a direction of future JRP of classification societies in Asia. Chen et al (1996, 1997, 1998a, 1998b) has been carried out a series of research projects in fire protection area for passenger ships sailing in the water area in China including the fire protection mathematical model, fuzzy theory application in fuzzy database as well as combination application of fuzzy theory and neural network theory. In the papers, they discussed the acceptable criteria of fire risk for passenger ships by means of fuzzy methods. This method describes the fuzzy property of the objects considered the structure property, knowledge property and other properties of objects. For that, they also try to reconsider the problem on the basis of Neural Network Method, which can describe the properties of objects synthetically. Chen and Lin (2001) carried out study on bulk carrier safety due to structural defects of the hull structures in cargo ships using FSA. The main purpose is to develop a method to carry out hazard identification and a corresponding database in which the fuzzy features in the process of application of FSA will be not avoided, so it is necessary to seek a new method to deal with the fuzzy feature in the practicable collection of data. The new ideas are applied to the process of FSA. In the paper, a comprehensive fuzzy method in the application of hazard identification of FSA is described, in the method the ship system is regarded as a whole that covers: 9 9 9 9 9 9 9

ship set; crew set; environment set; relationship between relationship between relationship between relationship amongst

ship and environment; ship and crew; crew and environment; and ship, crew and environment.

So this is a common method, which can be used in assessment of many events in ship system, such as the strength assessment for the number 1 cargo hold of bulk carriers, fire protecting, HRA and other nature assessments.

18

Specialist Committee V.1

Dasgupta et al (1998) studied both prescriptive and performance based regulations. The difference between the two forms in terms of the flexibility and their application were studied. The current indications of developments of performance based regulations as noted in the new SOLAS Chapter III (Life Saving Appliances) and SOLAS Chapter II-2 (Fire Safety) were highlighted. The authors proposed use of "success tree" (a similar as the fault tree; except that the probability of the success of a regulation is used instead of the failure probability) in the identification of regulatory requirements. Two issues were investigated: collisions as per SOLAS Chapter II-1 and for fire as per SOLAS Chapter II-2. The procedure enables a qualitative estimation of the effectiveness of a set of regulations as a safeguard against a specific hazard both in terms of its preventive and the mitigating clauses. The procedure can also be used to quantify the regulatory effectiveness once the individual clauses can be quantified. The procedure can also possibly be used in the quantification of the regulatory impact in future. The paper also identified the various sources of databases useful in the application of the FSA. Dasgupta et al (1999) studied a genetic model that was established for the steering gear system comprising of the following subsystems: main and auxiliary steering gear, the steering gear control system, and the steering gear power. The steering gear was identified to interface with main and emergency electrical powers source and switchboard. Based on the published failure statistics the relative reliability of the steering gear components were chosen with respect to the operational life. Where marine data were lacking failure data of components was chosen from the 'NPRD-95' handbook of "Reliability Analysis Center". The human factor affecting operator function was not considered separately through HRA. Arima et al (1996) carried out a trial application of FSA to the safety assessment of engine room fire of cargo ships was carried out. Its main purpose is to examine the applicability of FSA methodology, especially risk assessment and risk control options (Step 2 and 3 of the IMO FSA Guidelines). The fuel oil and lube oil systems in the engine room of a specific typical bulk carrier was taken as an example in this study instead of making its generic model. In addition, the occurrence and effects of fire due to combustible oil leakage from potential leakage locations were assessed. This paper proposed a technique for assessing the relative frequency ratio at each probable location where a fire may occur, and for assessing the effects of the layout, and number and types of safety equipment such as fire detectors and fire extinguishers on the scale of the fire. This trial application demonstrated that the methodology of probabilistic risk assessment could be applied to ship safety issues. Arima et al (1996) studied application of FSA on collision and stranding. At the same time related database were investigated. Base on the survey of data and information, fault tree analysis (FTA) and event tree analysis (ETA) was performed. Based on the casualty data, the basic patterns of FTA and ETA were constructed selecting macroscopic factors affecting collision and stranding, taking normal ships as the object of investigation with the objective of creating a framework for analysis in the future. It is convenient to classify collision and stranding casualties into light, heavy and total loss for the purpose of analysis and assessment in line with the categorization used by classification societies. A light casualty does not necessitate dry-docking after the accident or require the assistance of salvage tugs. Depending on the stranding of ship, the ship can move under power either by

Risk Assessment

19

waiting for the high tide or by shifting cargo on board the ship. A heavy casualty is an accident that requires the assistance of tugs, and total loss refers to a disabled ship, which can no longer be used; total loss includes a sunk or ships lost at sea. The human element comes into play in all cases of collision and stranding casualties. Based on actual accident reports, ETA and FTA were developed including human elements. Concerning collision and stranding, management factors related to human elements are very important. In the study, it can be concluded that the quantitative risk assessment can be carried out including human element, but is difficult to find universal accident measures based on survey about a small number of actual accidents. In addition, it is effective to survey minor accident because there is a possibility that those events could result in catastrophic accidents so we can prepare counter measures before catastrophic accident occur. Arima et al (1998) performed a safety assessment for fires in cargo holds, as the result may depend on ship type and size, they focused on a typical Aframax double-hull tanker. In this research, ships during unloading, tank cleaning and gas freeing are considered because the IGS is active and not a few accidents have been reported in this period. First, five initial events and scenarios were selected through a literature survey and discussion within the research group including consultants outside NK and preliminary risk assessment of them were carried out. Yeo et al (2000) studied trial applications of FSA methodologies to flooding of the number 1 cargo hold of bulk carriers. In this trial application, it was intended to: 9 develop the structural safety assessment system by applying the FSA methodology; 9 apply the developed system to risk assessment of bulk carriers whose dead weight is greater than 50,000 tones; and 9 suggest possible risk control options (RCOs) aimed at reducing the potential loss of life (PLL) from the viewpoint of structural integrity by implementing the proposed RCOs.

2.2.6

Alternative design and arrangements for fire safety

The new SOLAS Chapter II-2 entered into force on July 1st 2002, accordingly Registro Italiano Navale (RINA) developed class requirements as well as an expanded version of the FSA Guidelines where worked examples of all the steps in the analysis are included (RINA, 2002). For example, a fire event tree is provided in Figure 2 (reproduced from RINA's Rules).

2.2.7

Marine Insurance Industry: risk assessment and risk selection

The marine insurance industry has for assessing risks financial and economic, technical and operational factors to properly select and price risks. Insurance protection against risks and perils are handled differently amongst the well-known insurance conditions for hull and machinery. Hull and machinery policies are written for named risks and perils (International Hull Clauses, 2002) as well as for all risks and perils (Norwegian Marine Insurance Plan, 1996). Hull and machinery insurance is primarily covered by commercial insurance companies with a few notable exceptions. Therefore, risk selection is very important prior to accepting the risk. Significant effort has been placed on determining risks based on a number of criteria that include: 9 recent claims records; 9 classification society and current class records; 9 flag State; 9 ship type;

Specialist Committee 1/:.1

20

YES I NOT

I

A

~

Bl B2 Ct C2 D1 02 D3 ,.~MC~EINCIDENTCLA,..~ LOCATED LOCALt~ED MAJOR MAJOR MAJOR

MAJOR

LOCAL~ED

LOCALISED LCICAL~EO LOC&USEO'

MAJOR

MAJOR

ED

E~_~_

R~REII'r LOCAUSEO LOCALtSED LOCAUSEO LOCALLSED

LOCAUSEO LOC,~U~EO' LOCAUSEO

MAJOR

MAJOR

t.OC~Lt~SED

LOCAL~EO LOCAUSEO LOCAL~EO MAJOR

MAJOR

LOCALL~EO LOCAL~$EO

MAJOR

MAJOR

L~JOR

LOCAL~ECI LOCAL~EO

MAJOR

M~IOA LOCAL~EO LOCAL~ED MAJOR

MAJOR

LOC&LtSED MAJOR

JM,JC~

LOCAL~EO.

MAJOR

LOCAL~SED LOCAL~EO

~ $ V O G ~ I

MAJOR ~,~r

MAJOR LOCAUSEO LOCAL~$ED

L~MAJC~

Code EI B1 B2 C1 C2 D1 D2 D3

Event Ignition Event Rapid self termination Manual detection Automatic detection Forced ventilation shutdown Natural ventilation prevented Local manual suppression Automatic suppression On board manual suppression

Description Fire self terminates in the first instants subsequent to ignition People awake in the space are able to detect fire Equipment is able to detect fire Forced ventilation shutdown Whether door and/or windows are open/closed People in the space are able to extinguish fire Fixed equipment is able to extinguish fire People external to the space (fire brigade on board) are able to extinguish fire

Figure 2: Figure: event tree for a fire in a cabin (RINA, 2002) 9 9 9 9 9 9

ship size; ship technologymparticularly machinery type and manufacturer; ship age; port State inspection record; nature of trade including types of cargoes; domicile country of seafarers; and

9

trading regions.

The majority of ships with P&I cover are entered with P&I mutual clubs. The concept behind mutuality is insurance 'at cost'. Therefore, risk pricing is primarily based upon payment for claims incurred by the individual club member and, in some cases, shared payment of claims of other club members. P&I insurance costs are primarily based on cost of claims, cost of reinsurance and

Risk Assessment

21

performance of the club's investment portfolio. Risk selection also includes those elements described above. For further information, Gold (2002) provides an excellent summary of property and casualty insurance (hull and machinery) and protection and indemnity insurance (P&I) as well as the North of England P&I Club et al (1998).

2.3

Applications

2.3.1

Risk based fire safety design

Lee, J.H. et al (2001) provide a fire safety assessment about ship's fire protection design and classification society rules, statistical information and modeling techniques for the fire safety engineering are investigated and probabilistic safety assessment methods in the structural reliability engineering. A fire safety evaluation module (FSEM) developed in this paper calculates the probability of fatality, which can be used as an index of fire safety. FSEM is used to calculate the probability of fatality of the evacuees in a small room installed according to the rules for fire protection. Sensitivity analysis is executed to investigate the FSEM's applicability to ships. From the results, the necessity of new criterion for ship's fire safety design, the need to study the human behaviour in the evacuation from fire, and the development of new fire progress model considering special situations in ships are acknowledged. Yang et al (2001) summarises an FSEM that quantitatively evaluates the risk of evacuees in case fire occurs in ship has developed based on the research works done at Lurid University. The developed FSEM is applicable to multi-room structures as well as one-room structures. The necessary input data for the FSEM are obtained from a fire model, CFAST, and an evacuation model, MonteDEM. The MonteDEM evacuation model is developed by combining the Monte Carlo method for the random simulation of evacuee behaviour with the distinct element method (DEM) for the deterministic prediction of evacuee's movement. Compared with other evacuation models, the advantage of the MonteDEM evacuation model is that it includes the effects of ship motions to handle transverse inclinations. To verify the extended MonteDEM evacuation model, some numerical examples are demonstrated using the improved FSEM. The effects due to rolling motion should be considered to correctly evaluate the safety of evacuees in fire evacuation program. Through the numerical example, the quantitative estimation method for the latent risk of evacuation program is verified to be applicable and effective. Perhaps the most significant and thorough application of risk based fire safety design for ships is being completed within the 3 years duration SAFETY FIRST R&D project (European Community DGXII - 1998-2002 Growth Programme "SAFETY FIRST: Design For S a f e t y - Ship Fire Engineering Analysis Toolkit" Contract G3RD - CT99 - 0031). The aim of SAFETY FIRST is to ensure that a tried and tested fire ship engineering analysis toolkit, enabling ship designers to comply with IMO's revised fire safety regulations in place by the 1st July 2002. Now, it is possible for shipyards and ship designers to use a new approach where alternative, performancebased fire safety design and arrangements are accepted in lieu of traditional prescriptive designs as allowed by SOLAS regulation II-17 (alternative design arrangements). The analysis toolkit developed within the project, according to IMO's alternative design guidelines, is based on three pilot case studies of practical engineering relevance involving the fire protection design of passenger and crew accommodations and technical spaces on cruise ships as well as vehicle decks on ro-ro passenger ferries. Achieving these goals involves a good degree of scientific and applied research and development activities being performed by a team of 9 partners from 4 European Countries. Due to the novelty

22

Specialist Committee V.1

and complexity of SAFETY FIRST objectives, experience from civil buildings, nuclear and transport industry (railway and aircraft applications) is being considered, since fire engineering science is more developed in other industrial fields rather than in shipbuilding. To find an acceptable way of overcoming the limitations imposed by current fire safety regulations, that are not able to keep the pace with technological developments and customers' demand, the SAFETY FIRST project is structured into two main parts. The first one is based on scientific research and risk assessment techniques, aimed at assessing the applicability on ship design of performance-based analyses of fire and smoke models derived from other industrial fields. The second part of the project is devoted to practical simulations - with fire consequence modelling, qualitative and quantitative analyses of the selected Case studies - involving therefore a higher degree of applied research. The two parts of the project are linked to ensure that the models and the tests are always focused on real design alternatives. The expected achievements of the SAFETY FIRST project are to: 9 assess the practical applicability of IMO guidelines on the alternative design and arrangements with significant case studies; 9 develop a library of risk models, products and materials to identify representative fire scenarios to be readily available for application on the alternative performance-based approach to fire safety design; 9 provide a comprehensive toolkit for designers for practical application, including a costbenefit analysis to assess whether the alternative design is economically competitive; and 9 allow EU shipbuilding industry and ship owners to take immediate advantage of the new regulations leading to both enhanced competitiveness and improved safety.

2.3.2

Event and fault tree application

Assessment of risks is made through statistic models including those based on the full probability formula, Bayes' theorem, Monte-Carlo's and Delphi's methods, etc. Event and fault trees play important part in developing statistic models. Event trees and fault trees are used actively in investigation of different hazardous situations. Risk evaluation performed within the investigation floating production storage and offioading (FPSO) systems is a good example of such usage. In Wolford et al (2001), various events and fault trees including process loss of containment, mooring, transient induced leak frequency are explained. The function of the mentioned trees in specific analysis is of interest. There is information that marine event scenarios were represented with 89 unique initiating events, 12 frontline systems event trees, one support tree and 141 marine fault trees of which 89 developed specific initiating events and 52 modeled system response function (see Table 3). Over 2 billion unique event sequences were evaluated. Fire initiating event frequencies were developed for 70 individual hazard zones combined with an assessment of initiator density. The modeling of structural failures also followed a broadly similar approach. Structural event scenarios were represented with 46 unique initiating events, 13 frontline system event trees and one support tree. In the paper of Karsan et al (2001) it is said that risks associated with a FPSO system differ from those for the existing systems such as the conventional steel jacket, compliant tower, TLP and Spar. In the first phase of the existing Joint Industry Project (JIP), focus was made on evaluation of risks associated with the production operations from a Gulf of Mexico (GOM) FPSO. The JIP objectives included demonstration of the acceptability of a FPSO in the GOM: identification of

Risk Assessment

23

accidental events and F P S O components with high environmental pollution, loss of life, and financial risks, and recommendation of reasonably practicable risk reducing measures. Ten (10) oil companies, 3 F P S O contractors, 4 certification agencies and M M S sponsored the JIP. The risk concept developed for the G O M by a major oil company. TABLE 3 DESCRIPTION OF MODEL PARTITIONS (Wolford et al 2001) System Category Process

Marine

Initiating Events

171 Loss of Containment 9 Fault Tree modeling each of the 171 initiating events (Parts Count) 9 57 Escalation Events 9 89 (70 fires by zone)

Event Trees

9

Structural

9

46 Structural Damage

Total

363 Initiating Events

9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 17

Comprised of 54 process section Each of the 2 phases in 3 separators is modeled as a separate initiating event 3 Hole sizes (small, medium, large) Model Functions 9 Ignition/Explosion 9 Isolation 9 Blowdown 9 Fire Suppression Failure of cargo management Ballast control failure Flooding from seawater system Flooding from cargo oil system Rupture of marine pressure vessels Energetic Release - turbine breakup Marine fire Crude oil spill Pump/engine room explosion Inadvertent discharge of oily waste due to bilge system failure Inadvertent discharge of oily waste due to surface runoff Single mooring line failure Corrosion holes or fatigue crack turret shell Vessel impact with turret Hull damage following vessel impact or helicopter crash Reduced weather vaning Turret superstructure or foundation damage Turret superstructure underdeck damage Process support damage Process support underdeck damage Transverse bulkhead damage Longitudinal bulkhead damage Ship hull damage during extreme weather Turret events following fire and explosion Billion individual event sequences

The use of quantitative risk analysis technique provided a tool to identify potential escalation scenarios and probabilities of terminating events, and consequences. Similar issues are discussed in the paper of Nesje et al (1999). Here the qualitative and quantitative methods to assess F P S O risks, including event trees and fault trees, are presented. The risks addressed cover those originating from subsea releases from risers and flowlines, leaks from process systems possibly causing fires and explosions, crude oil storage tanks and engine rooms, offloading to shuttle tankers, collisions with different types of vessels, and escape and evacuation operations. Another aspect associated with F P S O problems where the event trees are used is reflected in the work of M a c D o n a l d et al (2001). The question is about Collision Risks Associated with F P S O in

24

Specialist Committee V.1

Deep Water Gulf of Mexico. The paper focuses on three main collision scenarios: route based traffic, random traffic, area density traffic. Formulae for evaluation of the passing vessel collision risks and for calculating the drift vessel collision frequency using the COLLIDE methodology are cited. At that the input parameter include the following: number of vessels using the route; vessel type category on the route, e.g. merchant, tanker, supply, standby, etc; vessel size categories on the route (divided by vessel type and deadweight tonnage); Closest Point of Approach (CPA) of the route to the installation; beating of the route to the installation at its CPA; standard deviation of the route (related to route width). The paper of Xu et al (2001), in which the principles and strategies of in-service inspection programs for FPSO's are explained, could be joined to the mentioned papers by its content. The paper summarizes the technical basis for three levels of inspection strategies: 1) probability-based inspection method, 2) risk-based inspection method, and 3) "optimum" inspection method.

2.3.3

Fuzzy set modeling and its application to maritime safety (prepared by Dr J. Wang of Liverpool JMU, UK)

With the cost of construction, operation and maintenance estimates in the multi-millions of dollars, the marine industry is seeking ways of reducing both the time and money spent to provide the high-quality marine engineering systems. Decision-making based on conventional mathematics that combines qualitative and quantitative concepts always exhibit difficulty in modeling actual problems. The successful selection process for choosing a design/procurement proposal is based on a high degree of technical integrity, safety levels and low costs in construction, corrective measures, maintenance, operation, inspection as well as preventive measures. However, the objectives of maximizing the degree of technical performance, maximizing the safety levels and minimizing the costs incurred are usually in conflict, and the evaluation of the technical performance, safety and costs is always associated with the uncertainty, especially for a novel marine system at the initial concept design stage. Furthermore, the safety of a large marine system is affected by many factors involved in its design, manufacturing, installation, commissioning, operation and maintenance. Consequently, it may be extremely difficult, if not impossible, to construct an accurate and complete mathematical model for the system in order to assess the safety because of inadequate knowledge about the basic failure events. This leads inevitably to problems of uncertainty in representation. In probabilistic risk assessment, probability distributions are used to describe a set of states for a system and to deal with uncertainty in order to evaluate potential hazards and assessment system safety. In many cases, however, it may be difficult or even impossible to precisely determine the parameters of a probability distribution for a given event due to lack of evidence or due to the inability of the safety engineer/designer to make firm assessments. Therefore one may have to describe a given events in terms of vague and imprecise descriptors such as "likely" or "impossible", terms that are commonly used by safety analysts/designers. Such judgments are obviously fuzzy and non-probabilistic, and hence non-probabilistic methods such as fuzzy set modeling may be more appropriate to analyse the safety of systems with incomplete information of the kind described above. .1 Use of fuzzy set modeling in risk assessment of offshore support vessels and fishing vessels (Sii et al, 2001; Pillay et al, 2002). A rule based fuzzy set modeling method was developed to carry out risk based design/operation decision making for offshore support vessels. A fuzzy set modeling method was combined with the fault tree analysis to deal with fishing vessel safety.

Risk Assessment

25

.2 Use of approximate reasoning approach for the design of offshore engineering products (Sii et al, 2002; Sii and Wang, 2002). Three different modeling frameworks were developed for safety-based design evaluation and decision support. These include: (1) a framework for risk analysis of offshore engineering products using approximate reasoning and evidential reasoning methods, (2) a decision support framework for evaluation of design options/proposals using a fuzzy-logic-based composite structure methodology, and (3) a design-decision support framework for evaluation of design options using a composite structure methodology based on approximate reasoning approach and evidential reasoning method (Sii et al., 2002). The first framework is designed for risk analysis of an engineering system having a hierarchical structure involved in safety assessment. The other two frameworks are used for design-decision support, using a composite structure methodology grounded in approximate reasoning and evidential reasoning methods. Fuzzy set modeling can also be used together with multiple attribute decision-making (MADM) methods to assist decision makers in selecting the winning design/procurement proposal that best satisfies the requirement in hand. It can also be used together with Analytical Hierarchy Process (AHP) and the Delphi method in carrying out design support evaluation (Sii et al, 2002). In maritime risk assessment, the application of numerical risk criteria may not always be appropriate because of uncertainties in inputs as discussed by Wang and Kieran (2000). Accordingly, acceptance of a safety case or formal safety assessment is unlikely to be based solely on a numerical assessment of risk. Where there is a lack of safety data for analysis or the level of uncertainty in safety data is unacceptably high, maritime safety analysts to facilitate risk modeling and decision-making may effectively use fuzzy set modeling as a useful alternative approach. Application of fuzzy set theory to risk assessment can also be found in Zolotukhin and Gudmestad (2000), Chert and Lin (2001), Wang et al (1995a, 1995b) and Wang and Kieran (2000).

2.3.4 FSA for safety of coastal trading ships in Japanese waters The Shipbuilding Research Association of Japan has been conducting a series of wide range of FSA studies for safety of coastal trading ships in Japanese water. The main topics are collision, grounding and fire casualties. The studies have been conducted in principle with the IMO FSA Guidelines and have used other analytical techniques for determining probability of collisions, groundings and fires. Progress reports of the studies have been issued annual basis (Shipbuilding Research Association of Japan, 2002). The main contributors to the studies are National Maritime Research Institute (NMRI), Nippon Kaiji Kyokai (Class NK), Shipowners Association of Japan (JCS), representatives of shipbuilding companies, Japan Ship-masters Association, professors of naval architecture and Ministry of Land, Infrastructure and Transport (MLIT) of Japan. The studies have derived casualty data from the records of Judges taken at Maritime Casualty Coat in Japan. In addition, statistical data of rescue record of Japanese Coast Guard have been used. However, it has been found that such data do not necessarily contain information on ship operations leading to casualties. Therefore, investigations on operations in near miss cases were conducted by way of questionnaires and interviews with seafarers. Hydrographical data on main sea routes and density of traffics on the routes were also used. Based on these data sources, hazards have been identified and several main casualty scenarios have been developed. Then, several event trees have been developed based on the scenarios.

26

2.3.5

Specialist Committee E1 Safety of ships carrying irradiated nuclear fuel

Irradiated nuclear fuel, Plutonium and high-level radioactive wastes are categorized as "B" type irradiated cargo by IAEA safety standards, and carries in flasks in accordance with IMO's Code for the safe carriage of irradiated nuclear fuel, plutonium and high-level wastes in flask on board ships (INF Code). The requirement for the flask is "to withstand a fire of 800~ for 30 minutes", which has been developed based on risk assessments for land transport. Concern was expressed that the fire resistance level would not be sufficient for maritime transport. Therefore, IAEA evaluated the fire safety of such flask carried on board ships (IAEA, 2001). In this project, the Shipbuilding Research Association of Japan conducted a risk based assessment study for the flask for the carriage of high-level radioactive wastes and irradiated nuclear fuel (Shipbuilding Research Association of Japan, 2000). The assessment comprised (1) collection of ships' fire casualty data and real scale fire test data, (2) establishment of fire scenario based on the fire casualty data, (3) estimation of temperature and fire conditions in the cargo space during fire casualty scenario, and (4) risk evaluation of fire around the flask. Two fire scenarios, under which the cargo space for the flask would be affected, were derived as: (a) engine room fire, and (b) fire after collision with a tanker or gas carrier. An event tree analysis was carried out for engine room fires, and two major serious fire scenario were considered: (i) Oxygen rich fire where closure of engine room fails and fire spread rapidly and reach high temperature. The fire continues for few hours and decades. (ii) Oxygen controlled fire where closure of engine room succeeds, but fire continues for longer hours in relatively low temperature. The temperature history in the engine room was simulated based on various real scale fire tests. It was concluded that temperature in the aft-most cargo hold during such engine room fire does not reach the temperature of 800~ in any case. An event tree analysis was carried out for fire resulted from collision with tankers. A scenario of cargo oil fire on the surface of sea around the ship was derived as the major serious fire scenario. Temperature of and heat flux from the fire was estimated based on several real scale oil pool fire test data. Then, temperature in the cargo hold where the flask was stored was calculated. Because the flask carrier has thick double sided shell and the double side spaces are ballast water tanks capable of filling water, it was estimated that the inside temperature of the cargo hold did not reach 800~ in nay case.

2.3.6 Alert communication from small craft using cellular phones A risk based evaluation of FSA on the use of personal cellular phone in board small craft was conducted in National Maritime Research Institute of Japan by Mitomo et al (2002). As an example of an application of FSA for the maritime field, the event tree analysis was applied to assess the effectiveness of cellular phone for reducing number of fatalities or missing people in maritime accidents happened on smaller crafts. Casualty data on small crafts, statistics of such crafts and statistics of population of cellular phone available in Japan were used. Scenarios of casualties of small crafts were identified and an even tree was developed for the case in which a cellular phone was carried in such craft, and another case in which cellular phone was not carried. The conclusion of the report shows that cellular phone would be a valuable means of communication when the craft stays in up to about 6 nautical miles from the coast and can reduce the number of fatalities by about 60% which have no sufficient standard marine communication facilities installed, or in casualties which are unexpectedly rapid with little or no time or

27

Risk Assessment

opportunity to communication with installed communication facilities. This study provides a good example of application of FSA for smaller study items, in which FSA study is relatively easily conducted.

3

ELEMENTS OF RISK ASSESSMENT

3.1

Uncertainty of Data

A report of a study on treatment of uncertain casualty data during FSA was submitted by Japan to IMO (2002g). It became clear, during discussions within the international collaborating FSA study group and FSA team of Japan for bulk carrier safety (see section 2.1.3), that the existence of casualty cases where the causes of the accident are unknown results in a discrepancy between the calculated probabilities of hatch cover failure reported. Transparency and neutrality are of paramount importance for FSA. Therefore, judgment of cause of such casualties needed to be done in a transparent and neutral manner where expert judgment should be avoided as far as possible. So, this study provides a method of estimation of number of casualties and fatalities caused by hatch cover failures using Bayesian theorem as follows. Two separate causes of casualties about a group of ships each of which is the cause of a certain number of casualties were used. However, there were a certain number of casualties whose causes remain unknown and cannot be investigated or solved due to lack of relevant data. During the investigation of bulk carrier casualties by FSA the cause of casualties of bulk carriers was classified into 'hatch cover failure', 'side shell failure' and 'unknown'. However, even in such a case, it would be possible to estimate the true number of casualties fatalities of each category by a probabilistic approach based on Bayesian theorem as follows. Following symbols are used in following sections. nh is the number of casualties cause of which is obviously hatch cover failure; ns is the number of casualties cause of which is obviously side shell failure; nu is the number of casualties cause of which is unknown; nt is the total number of all casualties, i.e. nt = nh + ns + nu; th is the true number of casualties cause of which is hatch cover failure; ts is the true number of casualties cause of which is obviously side shell failure; then ts = n t - th.

Here, let nh = H, ns=S, nu=U, th -- X. Then, next equation is produced from Bayesian theorem.

P((th : X ) [ (n h = H )

From the theory of probability,

(n s : S )) : P((nh : H )

(n' : S ) [ (t h : X )) . P(t h = X ) = : S))

(3)

Specialist Committee V.1

28

P(,~ : x ) :

~ c~

~ c~

_

2n,

nt

Z nCi

9

i=0

H+U ZiCH P((M h = Y)(~(/~/s

"- S))--

i=H

~

Cs

n,

--

H+U Z i CH On,-i CS i=H n,

Z ,,, Ci i=0

P((,,~ : H)~ (n, : s)l (t~ : x)):

~ c.

.._~

c,

nt C x

Therefore,

P((,~ = x) I(.~ = H ) ~ (.~ = s)):

* c..~

H+U

ca

(4)

Z i CH Ont-i CS i=H

Equation 4 means the conditional probability of the number of hatch cover related casualties (th = X) when nh - H, ns=S, nu=U. The next step is to obtain the probability of the number of fatalities (fatalities). Let the number of fatalities for each casualty i (i=1, nu) cause of which is unknown denote as R(i), the number of fatalities of all unknown caused casualties denoted as Nu and the number of fatalities of all obviously hatch cover related casualties denote as NH. Then the probability of the number of fatalities of all hatch cover related casualties NffNH

A

Gaussian integration point

Figure 5. Example of derivation of hot spot stress.

Figure 6. Different hot spot positions, Fricke (2001).

Models with thin plate or shell elements or alternatively with solid elements are normally used. It should be noted that on the one hand the arrangement and type of elements have to allow for steep stress gradients as well as for the formation of plate bending, and on the other hand, only the linear stress distribution in the plate thickness direction needs to be evaluated with respect to the definition of structural stress. The following methods of modelling are recommended:

296

Special Task Committee 111.2

9 The simplest way of modelling is offered by thin plate and shell elements which have to be arranged in the mid-plane of the structural components, see also Figure 7.8-node elements are recommended particularly in case of steep stress gradients. Care should be given to possible stress underestimation especially at weld toes of type b) in connection with 4-node elements, which should contain improved in-plane bending modes. The welds are usually not modelled except for special cases where the results are affected by high local bending, e.g. due to an offset between plates or due to a small free plate length between adjacent welds such as at lug (or collar) plates. Here, the weld may be included by vertical or inclined plate elements having appropriate stiffness or by introducing constrained equations for coupled node displacements. 9 An alternative particularly for complex cases is offered by solid elements which need to have a displacement function allowing steep stress gradients as well as plate bending with linear stress distribution in the plate thickness direction. This is offered, e.g., by isoparametric 20-node elements (with mid-side nodes at the edges) which means that only one element in plate thickness direction is required. An easy evaluation of the membrane and bending stress components is then possible if a reduced integration order with only two integration points in the thickness direction is chosen. A finer mesh sub-division is necessary particularly if 8-node solid elements are selected. Here, at least four elements are recommended in thickness direction. Modelling of the welds is generally recommended and easily possible as shown in Figure 8. 9 For both types of modelling, the dimensions of the first two or three elements in front of the weld toe should be chosen as follows. The element length should correspond to the plate thickness. In the transverse direction, the plate thickness may be chosen again for the breadth of the plate elements. However, the breadth over the first two elements should not exceed the "attachment width", i.e. the thickness of the attached plate plus 2 times the weld leg length (in case of type c: the thickness of the web plate behind plus 2 times weld leg length). This attachment width may also be taken for the width of solid elements in front of the weld toe, see Figure 8. The structural stress components on the plate surface should be evaluated along the paths shown in Figure 7 and Figure 8 and extrapolated to the hot spot. The average stress components between adjacent elements are used for the extrapolation. Recommended stress evaluation points are located at distances 0.5t and 1.5t away from the hot spot modelled, where t is the plate thickness at the weld toe. If the weld is not modelled, the hot spot is the structural intersection point modelled. The principal stress at the hot spot is calculated from the extrapolated component values. If the element sizes mentioned above are chosen, the stresses may be evaluated as follows: 9 In case of plate or shell elements the surface stress may be evaluated at the corresponding mid-side points. 9 In case of solid elements the stress may be extrapolated linearly to the surface centre. Normally, a linear extrapolation of the stresses to the hot spot modelled is performed. Altematively, a simplified approach without stress extrapolation is reasonable where the stress is taken at the location 0.5t away from the hot spot modelled and assessed with a reduced design S-N curve as described below. Much effort has been made during the last years to define S-N design curves for ship details, consistent with the assessment of hot spot stress. S-N data for weld toe cracking of tanker- and FPSO-specific details were developed based on fatigue tests, Kim (1997, 2001). This work included small-scale testing of 5 different types of typical FPSO details (75 specimens of each geometry) and 5 full-scale specimens of FPSO longitudinal to transverse frame connections for verification of the proposed procedure. The approach of using one hot spot S-N curve was supported by full scale tests of side longitudinals in ships, Lotsberg et al. (200 lb).

Fatigue Strength Assessment

297

Figure 7. Stress extrapolation in a three-dimensional FE model with shell elements.

Figure 8. Stress extrapolation in a three-dimensional FE model with solid elements. A design hot spot S-N curve was recommended based on a literature survey, The Welding Institute (TWI) in-house database and fatigue test data, Maddox (2001). The recommended S-N curve is linked to the finite element modelling and the method used for derivation of hot spot stress. For the linear extrapolation methods based on the FE modelling described above it is recommended to link the derived hot spot stress to the FAT90 curve IIW (1996). For stress calculated at 0.5t it is recommended to link the derived hot spot stress to that of FAT80 curve IIW (1996). The FAT90 curve may also be used together with this hot spot stress assessment. Then the hot spot stress can be calculated as 1.10 times the stress from 0.5t outside the weld toe or the intersection line. The nodal stress at the mid side node along line A-B in Figure 5 may be used for this stress using 8-node shell elements of size t by t at the hot spot region. As an alternative to this, Niemi (2001) proposes FAT90 for load-carrying fillet welds, while FAT100 is recommended for all other cases. The described procedure is considered to be conservative for details with significant stress gradients at the hot spots such as at plate bending at a hopper corner detail. Then a reduced hot spot stress may be

Special Task Committee VI.2

298

calculated taking this stress gradient into account, Kang et al. (2002). Reference is also made to the comparative study of the analysed bilge knuckle.

3.3

S-N Curve Formulation

Fatigue design criteria for structural components are based on a statistical analysis of fatigue test data obtained from constant amplitude tests. A linear relationship between log AS and log N is assumed (Figure 9), where AS is the stress range and N is the number of cycles for the fatigue life: (AS) m" N = C

or, on logarithmic form: m. log (AS) + log N = log C The exponent m is the slope of the S-N curve, taken with reference to the vertical axis. The value of the exponent m is generally between 3 and 7, i.e. a smaller slope exponent for a notched or welded component in relation to a smooth component. The S-N curve is the basis of the national and international codes for the calculation of fatigue design life. Thus for standard welded joints m = 3.0, consistent with fracture mechanics theory. The definition of a fatigue limit differs among design codes, for example N = 5" 106 (IIW, 1996) or 1.107 (NORSOK, 1998). In the I1W (1996) recommendations fatigue design curves are catalogued according to the characteristic stress range value ASR at N = 2. 106 (FAT classes). Only little is said about the scatter band of the individual joint types in the codes. In general the design S-N curves are calculated from a regression analysis giving a mean life curve with a two-sided confidence band. Design curves are assessed on the basis of a notional probability of survival of at least 95%. In most codes a "mean life curve minus two standard deviations" is used for design. To comply with the 95% probability of survival criterion the analysis must be based on a sample with at least 50 test results. In constant amplitude loading of steel components there will in general be a fatigue limit below which the fatigue life is "infinite." In design, the fatigue limit or the concept of a non-growing crack may be used only when the following conditions are met: - No cycles in the load history are above the fatigue limit No deterioration mechanism that could lead to an increase in the size of the initial defect is acting (corrosion, wear, etc.) -

Fatigue Strength Assessment

"9- .

Scatterband ~';'~'"..

2 Stdv S-N desig n~ c u r v e ' ' ~

R = Sm,n = const.

Smax

299

Mean curve

""'.. ""

x~"x

""-. """

"~'x'x """ x """ .................. ~]

k Number of cycles N [log]

Figure 9. S-N curve formulation in the codes.

A common case in fatigue design is to have a design load history with some cycles above, and some cycles below the fatigue limit. During the load history the cycles above the fatigue limit will contribute to crack initiation and growth, and the fatigue limit will be gradually lowered. For ships and offshore structures a major contribution to the fatigue limit comes from the small cycles in the spectrum, and the lowering of the fatigue limit may have a significant effect on cumulative damage. The lowering of the fatigue limit may be taken into account by a fictitious extrapolation of the S-N curve with a slope m2, Haibach (1970): m2 = 2 m l - 1 For welded joints m~ = 3.0 and m2 = 5.0. An alternative method is to apply a linear extrapolation of the S-N curve with slope ml to a fictitious cut-off level at N = 2- 107, Gurney (1976). Both these models were developed on the basis of fracture mechanics analysis. The Haibach model has been implemented in most design codes.

3.4

Combination of Stresses and Link to S-N Curves

In some design codes it is normal practice to use the principal stress within + 45 ~ normal to the weld toe together with an S-N curve for the weld for fatigue assessment as shown in Figure 11. Then the principal stress is linked to the FAT90 following the IIW(1996) notation.

Special Task Committee VI.2

300

& r-

Co._.nst_ant..am.,plitude .fati_gue l i m i t

03

R =

s

rain = const. Smax

z I

I

{

/ @1

z

1

Variable amplitude ] cut off life, / e'g" N = 10'

Number of cycles N [log]

Figure 10.

Modification of S-N curve for cumulative damage calculations.

If the angle between the principal stress is larger than 45 ~ the fatigue life with a stress component parallel with the weld together with an appropriate S-N curve that depends on the welding process should be assessed, IIW (1996). Some actual fatigue cracks can only be explained if it assumed that it is a principal stress at an angle larger than 45 ~ that is initiating and driving the crack growth. Thus IIW has suggested increasing this angle to 60 ~ Niemi (2001). In fatigue assessment of ship structures it is in most cases appropriate to assume that the calculated principal stress acts either normal to or parallel with the weld. However, there are some cases at hatches and circumferential welds around penetrations in deck structures where the new definition of principal stress and link to S-N curve may be of importance. Designers should be aware that fatigue cracking may occur at several locations around a circumferential weld. What region is most critical depends on geometrical properties as indicated in Figure 12. Fatigue cracking around a circumferential weld may occur at several locations on reinforced rings in plates depending on geometry of ring and weld size, ref. Figure 12. Fatigue cracking transverse to the weld toe in a region with a large stress concentration giving large stress parallel to the weld (Flexible reinforcement). See Figure 12a. Fatigue cracking parallel to the weld toe (Stiff reinforcement with large weld size). See Figure 12b. Fatigue cracking from the weld root (Stiff reinforcement with small fillet weld size). See Figure 12c. All these potential regions for fatigue cracking should be assessed in a design with use of appropriate stress concentration factors for holes with reinforcement, DNV-RP-C203 (2001).

Fatigue Strength Assessment

301

(3"2

ltlllItllllllllllll

0"1

45~

lllllllllllllllllll 0"2 Figure 11. Definition of stress in relation to S-N curve.

Position of fatigue crack

Comment

Fillet weld Fatigue crack growing normal to the weld toe due to large stress concentration when insert tubular is thin.

a)

1

Fatigue crack initiating from the weld toe for thicker insert tubular. The principal stress o 1 is the crack driving stress.

b) (~n

Fatigue crack in the fillet weld (initiating from the weld root) at region with large normal stress and shear stress (Small fillet weld size in relation to thickness of insert tubular or stiffening ring).

Figure 12. Fatigue cracks at reinforced cut-outs.

3.5

Effect of Tolerances

3.5.1 Introduction During the last years the reliability of equations for stress concentration factors for butt welds in design rules for floating production vessels (FPSOs) have become an important issue for fatigue design. Block sections welded from one side only and buttwelds at plates going from a thin to a thicker plate are

302

S p e c i a l Task C o m m i t t e e VI.2

found to be critical areas in terms of calculated fatigue life. A representative stress distribution at butt welds is required in order to perform a reliable fatigue design of these areas and to establish a sound basis for planning in-service inspection. A stress concentration factor can be defined as a stress magnification at a detail due to the detail itself or due to a fabrication tolerance with the nominal stress as a reference value. The maximum stress is often referred to as the hot spot stress that is used together with S-N data for fatigue life calculation. This hot spot stress is thus derived as the stress concentration factor times the nominal stress. Stress concentration factors for butt welds in plates have e.g. been presented in the DNV rules (1977, 1987) and by Maddox (1985, 1997). For plates and tubular sections SCFs have been presented by Connoly and Zettlemoyer (1993) and for tubular sections taking into account also slope of transition by Lotsberg (1998). The effect of stiffeners on the stress concentration factor for the butt welds in plated structures has been investigated by Lotsberg and Rove (2000). 3.5.2

Eccentric butt welds in plates

The stress at a butt weld between two plates as shown in Figure 13 is considered. It is assumed that the plates are welded together with an eccentricity e (and without angular mismatch). The plates are subjected to a nominal axial loading. Due to the eccentricity there will be secondary bending Sb at the weld. e Sb =3tS,om

and the stress concentration frequently referred to at an unstiffened plate weld joint is obtained from the definition given in the introduction as S C F - S"~ + S~ = 1 + 3 e Snom t

3.5.3

Fabrication tolerances in plated structures

Geometric stress concentration factors for butt welds in stiffened plates were investigated by finite element analysis using 20-node isoparametric elements. The geometry included cope holes, and is typical for stiffened plates in floating production vessels and ships. The geometry is also relevant for semi-submersibles.

Fatigue Strength A s s e s s m e n t

303

6 ~~162162

A

l

e

Notch

Figure 12. Eccentricity of butt welds. The FE analysis results were compared with the following equation for eccentricity with shift in neutral axis due to difference in plates with thickness: 0.5

SCF=I+3(tl_t2

)

t2 t~5 + t21.5

This equation was proposed by Maddox (1985). It is also used by the International Institute of Welding (1996) and in British Standard 7910 (1999). The FE analysis results were also compared with the following equation for eccentricity with shift in neutral axis due to eccentricity of plates in addition to shift in neutral axis due to joining of plates with different thickness: 0.5

S C F = 1 + 6e

tz

t~ 5 + t 21.5

0.5

+ 3 (t~ - t 2 )

t2__~____ t~5 + t 21.5

It was observed from the analyses that the cope holes imply an increased stress at that region. It is likely that a reduced area in way of the cope hole will imply an increase in SCF. This increase can be expressed as SCF = 1 +

Asection Asection - mcope hole

where: Asectio n = Sectional area of the plate and longitudinals without cope hole Acopehole = Area of cope hole in a section normal to the force.

Further details are discussed by Lotsberg and Rove (2000).

Special Task Committee VI.2

304

Tolerances are important for calculation of stress concentration factors for butt welds and cruciform joints. It is not quite obvious what tolerances are accounted for in the S-N data, as eccentricities were not measured for most of the test data that are used as a basis for derivation of the design S-N curves that are used today. In IIW (1996) it is stated that an eccentricity of 0.10t is included in the test data for butt welds and 0.15t for cruciform joints. In testing of cruciform joints the transverse plate is free in terms of boundary conditions. In a real structure it is restrained with respect to rotation in a similar manner as the two other plates. Therefore this corresponds to a tolerance of 0.3t that is accounted for in the S-N data for cruciform joints. This is in the range that normally is being accepted as a fabrication tolerance for cruciform joints. For butt welds an effective eccentricity in the equations given above can be calculated as e

= e m a x fabrication code - - O .

It

The maximum value of eccentricity in the IACS construction standard is the minimum of 0.15t and 3 mm for butt welds and t/3 for cruciform joints. Thus when the IACS standard is fulfilled with respect to these tolerances mainly butt welds at connections with different plate thickness such as shown in Figure 14 need to be investigated with respect to additional stress resulting from eccentricity. [

tlIi

[

i

t2

Figure 14. Worst case combination of eccentricity and transition in thickness.

3.6

Effect of Corrosive Environment

The effect of marine environment on fatigue strength has been researched and debated extensively over the last three decades. A major problem has been the apparent inconsistency of the effects. The following section is based on a literature survey that was made as basis for development of a Norsok standard for fatigue design of offshore structures presented by Lotsberg and Larsen (2001c). This work also formed the basis for a revision of the Norwegian standard, NS3472 (2001) and DNV-RP-C203 (2001). Since 1987 it has been generally accepted that the fatigue life of joints with cathodic protection in seawater environment is not less than that for joints in air for N>107. This is documented in the HSE (1992) background document. GrCvlen (1987) can also be cited: "The overall effect of cathodic protection upon fatigue life will depend on geometrical and loading conditions. Under long life conditions cathodic protection may give longer life than air exposure. Under conditions where shorter fatigue life is to be expected (f. ex. large initial defects), cathodic protection may give even shorter fatigue life than free corrosion." Thus, it may be concluded that there is some reduction in fatigue strength for low values of N, see HSE background document from 1992 and Berge et al. (1987). A similar conclusion was made by Mohaupt et al. (1987): "A comparison of data obtained in air with that obtained in artificial seawater, with or without cathodic protection, indicates that the fatigue life is reduced by a factor of about 2.5 to 3.0, and the beneficial effects, if any of cathodic protection, are too small to be recognised in fatigue design guidelines or codes". It should, however, be noted that this conclusion is based on test data with N < 3" 105 cycles.

Fatigue Strength Assessment

305

Based on this evidence the S-N curves in the Norwegian design standards were modified in the same manner as adopted by HSE (1995), i.e. the curves are shifted to the left in seawater environment with cathodic protection compared to those for air conditions. This means that the fatigue life, in terms of number of cycles N to failure, is reduced by a factor of approximately 2.5 for N < 106 cycles, while the fatigue life is kept equal to that in air for N > 107 cycles. For 106