Recommended Practice for Flexible Pipe - My Committeesmycommittees.api.org/standards/isotc67sc4/NDocs/2003/n246annex.pdf · Recommended Practice for Flexible Pipe Upstream Segment

Recommended Practicefor Flexible Pipe

API RECOMMENDED PRACTICE 17BTHIRD EDITION, MARCH 2002

Recommended Practicefor Flexible Pipe

Upstream Segment

API RECOMMENDED PRACTICE 17BTHIRD EDITION, MARCH 2002

SPECIAL NOTES

API publications necessarily address problems of a general nature. With respect to partic-ular circumstances, local, state, and federal laws and regulations should be reviewed.

API is not undertaking to meet the duties of employers, manufacturers, or suppliers towarn and properly train and equip their employees, and others exposed, concerning healthand safety risks and precautions, nor undertaking their obligations under local, state, or fed-eral laws.

Information concerning safety and health risks and proper precautions with respect to par-ticular materials and conditions should be obtained from the employer, the manufacturer orsupplier of that material, or the material safety data sheet.

Nothing contained in any API publication is to be construed as granting any right, byimplication or otherwise, for the manufacture, sale, or use of any method, apparatus, or prod-uct covered by letters patent. Neither should anything contained in the publication be con-strued as insuring anyone against liability for infringement of letters patent.

Generally, API standards are reviewed and revised, reafÞrmed, or withdrawn at least everyÞve years. Sometimes a one-time extension of up to two years will be added to this reviewcycle. This publication will no longer be in effect Þve years after its publication date as anoperative API standard or, where an extension has been granted, upon republication. Statusof the publication can be ascertained from the API Upstream Segment [telephone (202) 682-8000]. A catalog of API publications and materials is published annually and updated quar-terly by API, 1220 L Street, N.W., Washington, D.C. 20005.

This document was produced under API standardization procedures that ensure appropri-ate notiÞcation and participation in the developmental process and is designated as an APIstandard. Questions concerning the interpretation of the content of this standard or com-ments and questions concerning the procedures under which this standard was developedshould be directed in writing to the standardization manager, American Petroleum Institute,1220 L Street, N.W., Washington, D.C. 20005. Requests for permission to reproduce ortranslate all or any part of the material published herein should also be addressed to the gen-eral manager.

API standards are published to facilitate the broad availability of proven, sound engineer-ing and operating practices. These standards are not intended to obviate the need for apply-ing sound engineering judgment regarding when and where these standards should beutilized. The formulation and publication of API standards is not intended in any way toinhibit anyone from using any other practices.

Any manufacturer marking equipment or materials in conformance with the markingrequirements of an API standard is solely responsible for complying with all the applicablerequirements of that standard. API does not represent, warrant, or guarantee that such prod-ucts do in fact conform to the applicable API standard.

All rights reserved. No part of this work may be reproduced, stored in a retrieval system, or transmitted by any means, electronic, mechanical, photocopying, recording, or otherwise,

without prior written permission from the publisher. Contact the Publisher, API Publishing Services, 1220 L Street, N.W., Washington, D.C. 20005.

Copyright © 2002 American Petroleum Institute

FOREWORD

This

Recommended Practice

(RP) for

Flexible Pipe

is under the jurisdiction of the APISubcommittee on Subsea Production Systems. This RP provides complementary informa-tion to API SpeciÞcations 17J and 17K and addresses ßexible pipe system issues and evolv-ing technologies.

API publications may be used by anyone desiring to do so. Every effort has been made bythe Institute to assure the accuracy and reliability of the data contained in them; however, theInstitute makes no representation, warranty, or guarantee in connection with this publicationand hereby expressly disclaims any liability or responsibility for loss or damage resultingfrom its use or for the violation of any federal, state, or municipal regulation with which thispublication may conßict.

Suggested revisions are invited and should be submitted to the standardization manager,American Petroleum Institute, 1220 L Street, N.W., Washington, D.C. 20005.

iii

CONTENTS

Page

1 GENERAL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.3 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.4 Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.5 Referenced Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

3 DEFINITIONS AND ABBREVIATIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33.1 DeÞnitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33.2 Symbols and Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

4 SYSTEM, PIPE, AND COMPONENT DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . 54.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54.2 Flexible Pipe Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54.3 Flexible Pipe Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84.4 Ancillary Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

5 PIPE DESIGN CONSIDERATIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255.2 Design Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255.3 Failure Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295.4 Design Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325.5 Load Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

6 MATERIALS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416.1 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416.2 MaterialsÑUnbonded Pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416.3 MaterialsÑBonded Pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446.4 Alternative Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466.5 Polymer/elastomer Test Procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486.6 Metallic Material Test Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

7 SYSTEM DESIGN CONSIDERATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527.1 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527.2 General System Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527.3 Flowline Design Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537.4 Riser Design Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567.5 Ancillary Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587.6 System Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

8 ANALYSIS CONSIDERATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618.2 Analysis Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618.3 Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658.4 Global Response Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

v

Page

9 PROTOTYPE TESTING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719.2 ClassiÞcation of Prototype Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719.3 Test Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719.4 Test Protocol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749.5 ProceduresÑStandard Prototype Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769.6 ProceduresÑSpecial Prototype Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

10 MANUFACTURING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9010.1 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9010.2 ManufacturingÑUnbonded Pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9010.3 ManufacturingÑBonded Pipe. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9210.4 Marking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9410.5 Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

11 HANDLING, TRANSPORTATION, AND INSTALLATION . . . . . . . . . . . . . . . . . 9811.1 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9811.2 Handling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9811.3 Transportation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9911.4 Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9911.5 Pre-commissioning/Commissioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

12 RETRIEVAL AND REUSE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11312.1 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11312.2 Retrieval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11312.3 Reuse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

13 INTEGRITY AND CONDITION MONITORING . . . . . . . . . . . . . . . . . . . . . . . . . . 11713.1 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11713.2 General Philosophy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11713.3 Failure Modes and Potential Pipe Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11813.4 Monitoring Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11813.5 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

APPENDIX A FLEXIBLE PIPE HIGH TEMPERATURE END FITTINGQUALIFICATION TEST PROTOCOL: VOLATILE CONTENTPOLYMERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

APPENDIX A1 PVDF COUPON CRUDE OIL EXPOSURE TESTPROCEDURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

APPENDIX B FLEXIBLE PIPE HIGH TEMPERATURE END FITTINGQUALIFICATION TEST PROTOCOL: LOW VOLATILECONTENT POLYMERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

APPENDIX B1 POLYMER COUPON CRUDE OIL EXPOSURE TEST PROCEDURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

Figures1 Flexible Pipe Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Examples of Static Applications for Flexible Pipe. . . . . . . . . . . . . . . . . . . . . . . . . . 73 Examples of Dynamic Applications for Flexible Pipe . . . . . . . . . . . . . . . . . . . . . . . 94 Examples of Flexible Riser ConÞgurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

vi

Page

5 Examples of Flexible Pipe Jumper Line Applications . . . . . . . . . . . . . . . . . . . . . . 116 Schematic of Typical Flexible Riser Cross-Sections . . . . . . . . . . . . . . . . . . . . . . . 127 Pressure Armor and Carcass Interlock ProÞles . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 Example of an Unbonded Flexible Pipe End Fitting . . . . . . . . . . . . . . . . . . . . . . . 169 Schematic Drawing of an Example ISU. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1610 Examples of Multibore Constructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1711 Bend Limiters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1912 Schematic of a Bend Restrictor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1913 Example of a Quick Disconnect (QDC) System . . . . . . . . . . . . . . . . . . . . . . . . . . 2014 Subsea Buoy/Arch Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2215 Example of a Buoyancy Module for Wave ConÞgurations . . . . . . . . . . . . . . . . . . 2316 Example of a Clamp for Piggybacked Flexible Risers. . . . . . . . . . . . . . . . . . . . . . 2317 Example of a Typical Riser Base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2418 Example of a Typical Riser Hang-Off Structure . . . . . . . . . . . . . . . . . . . . . . . . . . 2519 Static Application Design Flowchart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2720 Dynamic Application Design Flowchart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2821 PA-11 Service Life vs. Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4922 Parameters Used to DeÞne a Bellmouth Shape . . . . . . . . . . . . . . . . . . . . . . . . . . . 5923 Example of Haigh Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6424 Schematic of Set-Up for the Burst Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7725 Schematic of Set-Up for the Axial Tension Test. . . . . . . . . . . . . . . . . . . . . . . . . . . 7826 Schematic of Set-Up for the Collapse Test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8027 Dynamic Fatigue Test Program DeÞnition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8128 Typical Set-Up for a Dynamic Fatigue Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8229 Schematic of Set-Up for the Erosion Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8830 Schematic of Horizontal Lay Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10231 Schematic of Vertical Lay Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10332 Representative Flowline Installation Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . 10433 Schematic of J-Tube Pull-In Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10534 Representative Lazy-S Riser Installation Procedure . . . . . . . . . . . . . . . . . . . . . . 10635 Representative Steep-S Riser Installation Procedure . . . . . . . . . . . . . . . . . . . . . . 10736 Representative Lazy Wave Riser Installation Procedure . . . . . . . . . . . . . . . . . . . 10837 Representative Steep Wave Riser Installation Procedure . . . . . . . . . . . . . . . . . . . 10938 Representative Free-Hanging Catenary Installation Procedure . . . . . . . . . . . . . . 11039 Schematic of Possible Test Pipe Arrangements . . . . . . . . . . . . . . . . . . . . . . . . . . 12140 Schematic of Topside Di-Electric Sensing Layout and Instrumentation

for Thermoplastic Monitoring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121A-1 Monitoring Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134B-1 Monitoring Assembly (Case II Only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

Tables1 Description of Standard Flexible Pipe FamiliesÑUnbonded Pipe . . . . . . . . . . . . 152 Description of Standard Flexible Pipe FamiliesÑBonded Pipe . . . . . . . . . . . . . . 153 Check List of Failure Modes for Primary Structural Design of Unbonded

Flexible Pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304 Check List of Failure Modes for Primary Structural Design of Bonded

Flexible Pipe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315 Recommended Allowable Degradation for Unbonded Pipes. . . . . . . . . . . . . . . . . 356 Recommended Allowable Degradation for Bonded Pipe. . . . . . . . . . . . . . . . . . . . 377 Recommendations on Annual Probabilities for Installation, and Normal

and Abnormal Operation for a 20 Year Service Life . . . . . . . . . . . . . . . . . . . . . . . 37

vii

Page

8 Typical Static Global Analysis Load CasesÑOperating Conditions. . . . . . . . . . . 389 Example of Dynamic Load Cases for FPSO/FPS Applications. . . . . . . . . . . . . . . 3910 Example of a Dynamic Load Case MatrixÑNormal OperationÑ

Functional and Environmental Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3911 Example Global Analysis Load Cases for Installation Conditions . . . . . . . . . . . . 4012 Example Local Analysis Load Cases for Installation Condition . . . . . . . . . . . . . . 4013 Typical Polymer Materials for Flexible Pipe Applications . . . . . . . . . . . . . . . . . . 4114 Temperature Limits for Thermoplastic Polymers in Flexible Pipe

Internal Pressure Sheath Applications Based on 20-Year Service Life . . . . . . . . . 4215 Typical Fluid Compatibility and Blistering Characteristics for Flexible

Thermoplastic Pipe Polymer Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4316 Typical Elastomer Materials For Bonded Flexible Pipe Applications . . . . . . . . . . 4417 Temperature limits for Thermosetting Elastomers in a Bonded Flexible

Pipe Liner Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4518 Typical Soil Stiffness and Friction CoefÞcients for Flexible Pipes . . . . . . . . . . . . 6719 ClassiÞcation of Prototype Tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7220 Recommendations for Prototype TestsÑModiÞcations to Pipe Structure Design . 7321 Recommendations for Prototype TestingÑChanges in Pipe Application . . . . . . . 7322 Potential Flexible Pipe Failure Modes and Associated Critical Prototype Tests . . 7423 Recommendations for Class I Prototype Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7524 Recommendations for Class II Prototype Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . 7525 Sample Dynamic Fatigue Test Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8326 Critical Aspects in Selection of Unbonded Flexible Pipe Manufacturing

Tolerances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9527 Critical Aspects in Selection of Bonded Flexible Pipe Manufacturing Tolerances . 9628 Marking Recommendations for Flexible Pipe Products . . . . . . . . . . . . . . . . . . . . 9729 Potential Pipe Defects for Static Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . 12230 Potential Pipe Defects for Dynamic Applications . . . . . . . . . . . . . . . . . . . . . . . . 12631 Potential System Defects for Static and Dynamic Applications . . . . . . . . . . . . . 12832 Current Integrity and Condition Monitoring Methods . . . . . . . . . . . . . . . . . . . . . 131

viii

1

Recommended Practice for Flexible Pipe

1 General

1.1 SCOPE

This recommended practice provides guidelines for thedesign, analysis, manufacture, testing, installation, and opera-tion of ßexible pipes and ßexible pipe systems for onshore,subsea, and marine applications. This recommended practicesupplements API SpeciÞcation 17J [1] and API SpeciÞcation17K [2], which speciÞes minimum requirements for thedesign, material selection, manufacture, testing, marking, andpackaging of unbonded and bonded ßexible pipes respec-tively [3].

In general, ßexible pipe is a custom-built product that canbe designed and manufactured in a variety of methods. It isnot the intent of this document to discourage novel or newdevelopments in ßexible pipe. On the contrary, it is recog-nized that a variety of designs and methods of analysis arepossible. For this reason, some topics are presented in generalterms to provide guidance to the user while still leaving openthe possibility of using alternative approaches.

The reader should be aware that ßexible pipe technology(i.e., concepts, design and analysis methodologies and crite-ria, components manufacturing and testing, operational rolesand demands, maintenance and inspection, etc.) is in a stateof rapid and continuing evolution. Potential users thereforeneed to apply care in their application of the recommenda-tions within this document.

1.2 PRODUCTS

As with API SpeciÞcation 17J and 17K, this recommendedpractice applies to ßexible pipe assemblies, consisting of seg-ments of ßexible pipe body with end Þttings attached to bothends. Both bonded and unbonded pipe types are covered. Inaddition this recommended practice applies to ßexible pipesystems, including ancillary components.

This recommended practice does not cover umbilical andcontrol lines.

1.3 APPLICATIONS

The applications covered by this recommended practiceare sweet and sour service production, including export andinjection applications. Production products include oil, gas,water and injection chemicals, and combinations of these ser-vices. The recommended practice applies to both static anddynamic ßexible pipe systems, used as ßowlines, risers, andjumpers. The recommended practice does cover in generalterms, the use of ßexible pipes for offshore loading systems.Refer to [4] also for this application.

The recommended practice does not apply to ßexible pipesfor use in choke and kill line or umbilical applications. Refer-

ence API SpeciÞcation 16C for choke and kill line applica-tions and API SpeciÞcation 17E for umbilical applications.

1.4 UNITS

The Systeme Internationale (SI) units are used in this rec-ommended practice. Imperial units may be given in bracketsafter the SI units.

1.5 REFERENCED STANDARDS

See Section 2 of API SpeciÞcation 17J [1] and API SpeciÞ-cation 17K [2] for referenced standards. Standards referencedonly in this document are listed in Section 2, along with rele-vant technical papers and publications.

2 References

1. API SpeciÞcation 17J,

SpeciÞcation for Unbonded Flexi-ble Pipe

.

2. API SpeciÞcation 17K,

SpeciÞcation for Bonded FlexiblePipe

.

3. Grealish, F., Bliault, A., and Caveny, K., ÒNew Standardsin Flexible Pipe Technology Including API Spec 17J,Ó Pro-ceedings of Offshore Technology Conference, OTC PaperNo. 8181, Houston, Texas, May, 1995.

4. OCIMF,

Guide to Purchasing, Manufacturing and Testingof Loading and Discharge Hoses for Offshore Moorings

,Fourth Edition, 1991.

5. FPS 2000,

Handbook on Design and Operation of Flexi-ble Pipes

, February, 1992.

6. Mahoney T., and Bouvard M., ÒFlexible Production RiserSystem for Floating Production Application in the NorthSea,Ó OTC Paper No. 5163, 1986.

7. Colquhoun, R.S., Hill, R.T., and Nielsen, R.,

Design andMaterials Considerations for High Pressure Flexible Flow-lines

, Society for Underwater Technology, Aberdeen, May1990.

8. Rigaud, J., ÒNew Materials and New Designs for FlexiblePipes,Ó

Flexible Pipe Technology

, February 1992, Trondheim,Norway.

9. Bouvard, M., Mollard, M., and J. Rigaud, ÒSpecifying,Monitoring and Verifying Quality and Reliability of FlexiblePipe,Ó OTC Paper 6873, May, 1992.

10. Kalman, M. and Rosenow, M., ÒEmploying CompositeMaterials in Flexible Riser Design for Deepwater Applica-tions,Ó Deeptec Ô95, Aberdeen, February 1995.

2 API R

ECOMMENDED

P

RACTICE

17B

11. Moore, F., ÒMaterials for Flexible Riser Systems: Prob-lems and Solutions,Ó

Engineering Structures

, Vol. 11, Octo-ber, 1989.

12. Dawans, F.A., Jarrin, J., Lefevre, T.O. and Pelisson, M.,ÒImproved Thermoplastic Materials for Offshore FlexiblePipes,Ó Proceedings of Offshore Technology Conference,OTC Paper No. 5231, Houston, Texas, May, 1986.

13. European Federation of Corrosion, ÒGuidelines on Mate-rial Requirements for Carbon and Low Allow Steels for H

2

SContinuing Oil and Gas Field Service,Ó Document No: EFCO&G 93-1.

14. Kvernold, O., ÒErosionÑCorrosion of Internal SteelCarcass of Flexible Pipes,Ó Proceedings of Flexible PipeTechnology Seminar FPS 2000, Trondheim, Norway, Febru-ary, 1992.

15. ASTM E739,

Practice for Statistical Analysis of Linearor Linearized Stress-Life (S-N) and Strain-Life (e-N) FatigueData

.

16. ASTM E468,

Practice for Presentation of ConstantAmplitude Fatigue Test Results for Metallic Materials.

17. Veritec RP E305,

On-Bottom Stability Design of Subma-rine Pipelines

, October, 1988.

18. DnV,

Rules for Submarine Pipeline Systems

, 1981.

19. PRC/Pipeline Research Committee of American GasAssociation (A.G.A.), ÒSubmarine Pipeline On-bottom Sta-bility,Ó Vol. 1,

Analysis and Design Guidelines

, AGA ProjectPR-178-9333, Sept. 1993.

20. Det Norske Veritas,

Rules for CertiÞcation of FlexibleRisers and Pipes

, 1995.

21. Nielsen, N.J.A., ÒUpheaval Buckling Analysis of Flexi-ble Flowlines,Ó Proceedings of Flexible Pipe TechnologySeminar, February, 1992, Trondheim, Norway.

22. Pedersen, P.T. and Jensen, J.J., ÒUpheaval Creep of Bur-ied Heated Pipelines with Initial Imperfection,Ó

Journal ofMarine Structures, Design, Construction, and Safety

, Vol. 1,Elsevier Applied Science, 1988.

23. Pedersen, P.T. and Michelsen, J., ÒLarge DeßectionUpheaval Buckling of Marine Pipelines,Ó Proceedings ofBehaviour of Offshore Structures (BOSS), Trondheim, Nor-way, June, 1988.

24. Det Norske Veritas,

Rules for the Design, Constructionand Inspection of Offshore Structures

, 1977.

25. NPD,

Guidelines on Design and Analysis of Steel Struc-tures

, January, 1990.

26. API Recommended Practice, RP 2A-WSD,

Recom-mended Practice for Planning, Designing and ConstructingFixed Offshore PlatformsÑWorking Stress Design

, TwentiethEdition, July 1, 1993.

27. DnV,

Rules for CertiÞcation of Lifting Appliances

, 1989.

28. Frost, S.R. and Buchner, S., ÒA Permeation Model ToCalculate the Pressure Accumulation of Bore Gases in theAnnulus of Flexible Flowlines or Risers,Ó Proceedings of Oil-Þeld Engineering with Polymers, London, October, 1996.

29. McNamara, J.F., OÕBrien, P.J. and Gilroy, J.P., ÒNonlin-ear Analysis of Flexible Risers Using Hybrid Finite Ele-ments,Ó

Journal of Offshore Mechanics and ArcticEngineering

, ASME, Vol. 110, No. 3, August 1988, pp. 197Ð204.

30. OÕBrien, P.J. and McNamara J.F., ÒSigniÞcant Character-istics of Three-Dimensional Flexible Riser Analysis,Ó

Engi-neering Structures

, Vol. 11, October 1989, pp. 223Ð233.

31. McNamara, J.F., OÕBrien, P. and Grealish, F. ÒExtremeBending and Torsional Responses of Flexible Pipelines,Ó Pro-ceedings of the 11th International Conference on OffshoreMechanics and Arctic Engineering, eds. S.K. Chakrabarti etal., Vol. 1, Part A, pp. 319Ð324, New York.

32. Morison, J. R., OÕBrien, M. P., Johnson, J. W., Schaaf, S.A., (1950) ÒThe Force Exerted by Surface Waves on Piles,Ó

Petroleum Transaction

, American Institute of Mining Engi-neers, Vol. 189, pp. 149Ð154.

33. Kodaissi, E., Lemarchand, E. and Narzul, P., ÒState of theArt on Dynamic Programs for Flexible Riser Systems,Ó

Off-shore Mechanics and Arctic Engineering

, OMAE, Houston,1990.

34. Larsen, C.M., ÒFlexible Riser AnalysisÑComparison ofResults from Computer Programs,Ó

Marine Structures

, Vol. 5,pp. 103Ð119, 1992.

35. Feret, J.J., Bournazel, C.L., and Rigaud, J., ÒEvaluationof Flexible Pipes Life Expectancy Under Dynamic Condi-tion,Ó OTC Paper No. 5230, 1986.

36. Nielsen R., Colquhoun R.S. and McCone A., ÒTools forPredicting Service Life of Dynamic Flexible Risers,Ó Proc.European Offshore Mechanics Symp., NTH, Trondheim,1990.

37. Rodenbusch, G., Kallstrom, C., ÒForces on a Large Cyl-inder in Random Two Dimensional Flows,Ó Proc. of OffshoreTechnology Conference, Houston Texas, OTC Paper No.5096, 1986.

38. Huse, E., ÒHydrodynamic Forces on Risers with Buoy-ancy Elements,Ó Proceedings of 9th OMAE Conference,Houston, 1990.

R

ECOMMENDED

P

RACTICE

FOR

F

LEXIBLE

P

IPE

3

39. Pontazopoulus, M.S., ÒVortex Induced Vibration Parame-ters: Critical Review,Ó OMAE Conference, 1994.

40. OÕBrien, P.J., McNamara, J.F. and Dunne, F.P.E., ÒThree-Dimensional Nonlinear Motions of Risers and OffshoreLoading Towers,Ó

ASME J. OMAE

, August 1988, Vol. 110 pp.232Ð237.

41. Engseth, A., Bech, A., and Larsen C.M., ÒEfÞcientMethod for Analysis of Flexible Risers,Ó BOSS 88, 1988.

42. Leira, B.J., and Olufsen, A., ÒApplication of FrequencyDomain Procedures for Flexible Riser Analysis,Ó ISOPE1992, Vol. II, pp. 192Ð201.

43. McIver, D.B., ÒRiser Effective TensionÑNow You Seeit, Now You DonÕt,Ó 37th Mechanical Engineering Workshopand Conference, Petroleum Division ASME, September,1981.

44. Sparks, C., ÒThe Inßuence of Tension, Pressure andWeight on Pipe and Riser Deformations and Stresses,Ó

Trans-actions of the ASME

, Vol. 106, March, 1984.

45. Brekke, J.N., and Gardner, T.N., ÒAnalysis of Brief Ten-sion Loss in TLP Tethers,Ó OMAE Tokyo, 1986, pp. 134Ð140.

46. API RP 17B,

Recommended Practice for Flexible Pipe

,First Edition, June, 1988.

47. ASTM D413,

Test Methods for Rubber PropertyÑAdhe-sion to Flexible Substrate

.

48. ASTM D2143,

Test Method for Cyclic Pressure Strengthof Reinforced, Thermosetting Plastic Pipe

.

49. ASTM D2924,

Test Method for External Pressure Resis-tance of Reinforced Thermosetting Resin Pipe

.

50. Boef, W., MacKenzie, V. and OÕBrien, P., ÒEngineeringthe Dynamic Flexible Riser System for the Guillemot andTeal Project (Anasuria FPSO),Ó Proceedings of the Advancesin Riser Technologies Conference, Aberdeen, 1996.

51. MCS International & Robit AS,

Guidelines for IntegrityMonitoring of Unbonded Flexible Pipe

, Rev. 03, Nov. 1996.

52. Verdu, J. et.al.; ÒAgeing of polyamide 11 in acid solu-tions,Ó

Polymer

Vol.38, No. 8, pages 1911Ð1917; 1997;

53. Verdu, J. et.al.; ÒMolecular weight distribution and masschanges during polyamide hydrolysis,Ó

Polymer

Vol. 39,Numbers 6Ð7; 1998;

54. Jarrin, J. et.al.; ÒDurability of Polyamide 11 for OffshoreFlexible Pipe Applications;Ó 26th and 27th October, 1998;2nd International Conference on OilÞeld Engineering withPolymers; Westminster, London, UK.

3 Definitions and Abbreviations

3.1 DEFINITIONS

For the purpose of this standard, the deÞnitions in Section3.1 of the API SpeciÞcations 17J and 17K and the followingapply:

3.1.1 Arrhenius plot:

Used to plot service life against theinverse of temperature for some polymer materials by meansof a log-linear scale.

3.1.2 basket:

Used for storage and transport of ßexiblepipe (all pipes are laid freely into the basket).

3.1.3 birdcaging:

Buckling of the tensile armor wires,which results in signiÞcant radial deformation. It is usuallycaused by extreme axial compression.

3.1.4 buoyancy module:

A buoy used in signiÞcantnumbers at discrete points over a section of riser to achievewave shape riser conÞgurations (see 4.4.6). See also deÞni-tion for subsea buoy.

3.1.5 carousel:

Used for storage and transport of ßexiblepipe in very long lengths and rotates about a vertical axis.Pipe is wound under tension around the center hub.

3.1.6 chinese fingers:

A woven steel wire or fabricsleeve that can be installed over a ßexible pipe and drawntight to grip it for support or applying tension to the pipe.

3.1.7 chinese lantern:

Riser conÞguration used in shal-low water offshore loading systems to connect a PLEM to abuoy directly above it. The upper and lower connections arevertical and excess riser length is supported by distributedbuoyancy. See Figure 4.

3.1.8 flexible pipe system:

A ßuid conveyance systemfor which the ßexible pipe(s) is the primary component andincludes ancillary components attached directly or indirectlyto the pipe.

3.1.9 free-hanging catenary:

Riser conÞgurationÑseeFigure 4.

3.1.10 heat trace:

An element incorporated into pipestructure to provide heating.

3.1.11 integrated service umbilical (ISU™):

A struc-ture in which the inner core is a standard ßexible pipe con-struction. Umbilical components are wound around the corepipe and covered with a protective outer sheath (see 4.3.4).

Note: ISU is a trademark of Coßexip Stena Offshore.

3.1.12 J-S:

Riser conÞguration similar to a lazy-S (seeFigure 4), with the exception that the lower catenary passesback underneath the subsea buoy. Also called reverse-S.

3.1.13 lazy wave:

Riser conÞgurationÑsee Figure 4.

4 API R

ECOMMENDED

P

RACTICE

17B

3.1.14 lazy-S:


3.1.15 multibore:

Multiple ßexible pipes and/or umbili-cals are contained in a single construction. An outer sheath isextruded over the bundle (see 4.3.6).

3.1.16 multiple configuration:

A riser system whichhas more than one riser connected at a mid-depth location,such as at a subsea buoy/arch system.

3.1.17 ovalisation:

The out-of-roundness of the pipe,deÞned as the following:

where

D

max

and

D

min

are maximum and minimum pipediameter respectively.

3.1.18 rapid decompression:

Sudden depressurizationof a system. Gas in the pipe will expand rapidly and maycause blistering or collapse of the internal pressure sheath orother gas-saturated layers.

3.1.19 reel:

Large diameter structures used for storage ofßexible pipe in long lengths and rotates about a horizontalaxis.

3.1.20 riser base:

Seabed structure (gravity or piled) forsupporting subsea buoy/arch systems and/or riser/ßowlineconnections (see 4.4.8).

3.1.21 riser hang-off:

Structure for supporting riser atthe connection to a platform (jacket, semi-sub, tanker, etc.).

3.1.22 steep wave:


3.1.23 steep-S:


3.1.24 subsea buoy:

Concentrated buoyancy system,generally consisting of steel or syntactic foam tanks, as usedin S-type riser conÞgurations (see 4.4.5). See also buoyancymodule.

3.1.25 tensioner:

Mechanical device used to support orapply tension to a pipe during installation. Also called cater-pillars.

3.1.26 umbilical:

A bundle of helically or sinusoidallywound small diameter chemical, hydraulic, and electricalconductors for power and control systems.

3.2 SYMBOLS AND ABBREVIATIONS

The following symbols and abbreviations are used in thisdocument:

17J/17K API SpeciÞcations 17J and 17K

AISI American Iron and Steel Institute

ANSI American National Standards Institute

API American Petroleum Institute

ASME American Society of Mechanical Engineers

ASTM American Society for Testing and Materials

C

D

Hydrodynamic Drag CoefÞcient

C

m

Hydrodynamic Inertia CoefÞcient

DMA Deplasticizing Monitoring Assembly

DnV Det norske Veritas

DOF Degrees of Freedom

FAT Factory Acceptance Test

FEM Finite Element Method

FPS Floating Production System

FPSO Floating Production Storage and Ofßoading

GA General Arrangement

GRP GlassÞber Reinforced Plastic

HAZID Hazard IdentiÞcation Study

HAZOP Hazard and Operability Study

HDPE High Density Polyethylene

HIC Hydrogen-induced Cracking

ID Inside Diameter

ISO International Standards Organization

MBR Minimum Bend Radius

MWL Mean Water Level

NACE National Association of Corrosion Engineers

NDE Non-Destructive Examination

NPD Norwegian Petroleum Directorate

OCIMF Oil Companies International Marine Forum

OD Outer Diameter

PA Polyamide

PE Polyethylene

PP Polypropylene

PLEM Pipeline End Manifold

PU Polyurethane

PVC Polyvinyl Chloride

PVDF Polyvinylidene Fluoride

QCDC Quick Connect Disconnect

QDC Quick Disconnect

RAO Response Amplitude Operators

SSC SulÞde Stress Cracking

TAN Titrated Acid Number

TFL Through Flowline

UV Ultraviolet

VIV Vortex Induced Vibration

XLPE Cross-linked Polyethylene

σ

u

Material Ultimate Stress

σ

y

Material Yield Stress

Dmax DminÐDmax Dmin+---------------------------

R

ECOMMENDED

P

RACTICE

FOR

F

LEXIBLE

P

IPE

5

4 System, Pipe, and Component Description

4.1 INTRODUCTION

4.1.1 Scope

This section provides a general overview of ßexible pipesystems, pipe cross-section designs, and ancillary compo-nents. In addition, this section gives an overview of allaspects of ßexible pipe technology and identiÞes the sectionsof this recommended practice and API SpeciÞcations 17J/17K to be consulted for relevant issues.

4.1.2 Recommended Practice and Specification Overview

4.1.2.1

API standards facilitate the broad availability ofproven, sound engineering and operating practices. A recom-mended practice shows guidance on best practice in a particu-lar area of technology, while a speciÞcation deÞnes theminimum technical requirements for supply of a product(unbonded ßexible pipe in the case of API SpeciÞcation 17Jand bonded ßexible pipe in the case of API SpeciÞcation17K) that are designed and manufactured to uniform stan-dards and criteria.

4.1.2.2 This document provides the current best practicefor design and procurement of ßexible pipe systems and givesguidance on the implementation of the speciÞcation for stan-dard ßexible pipe products. In addition the recommendedpractice shows guidelines on the qualiÞcation of prototypeproducts.

4.1.2.3 All aspects of ßexible pipe technology, from func-tional deÞnition to installation, are addressed in either thisrecommended practice or API SpeciÞcations 17J/17K. Someissues are addressed in both documents. The various stages inthe procurement and use of ßexible pipes are deÞned in Fig-ure 1, which also speciÞes the sections of the recommendedpractice and API SpeciÞcations 17J/17K to be referenced foreach of the individual stages in the process.

4.2 FLEXIBLE PIPE SYSTEMS

4.2.1 Definition of System

4.2.1.1 The ßexible pipe system is an important part of theoverall Þeld development and may inßuence or be inßuencedby the design and speciÞcation of other components in thedevelopment. The deÞnition of the ßexible pipe systemshould therefore commence at the initiation of the overallproject as development strategies evolve. Aspects of thedevelopment strategy which may inßuence the ßexible pipesystem include Þeld layout (template versus satellite wells)and production vessel type (platform, tanker including turretlocation, semi-sub, etc.). Current limitations in ßexible pipe

technology, such as application range and manufacturingcapability, may also fundamentally inßuence potential overallÞeld development options.

4.2.1.2 Two aspects need to be addressed: namely, the ßex-ible pipe system and the ßexible pipe/or pipes within that sys-tem. The relevant parameters need to be considered as well asthe interactions between the pipe design and the systemdesign. Critical parameters that may affect the pipe designshould be identiÞed early in the process and could include thefollowing:

a. Severe internal conditions, such as high H2S content (sourservice).b. Extreme external environmental conditions.c. DifÞcult installation conditions (e.g., extreme environment).d. Frequent cyclic large amplitude pressure and temperatureßuctuations.e. Large vessel offsets.

4.2.1.3 To deÞne accurately all relevant parameters, inter-action between the purchaser and manufacturer is required atan early stage in the project. An important aspect of this is theidentiÞcation of critical system issues, such as interfaces.Section 7.6 lists potentially critical interfaces that should beconsidered at project commencement.

4.2.1.4 Appendix A of API SpeciÞcations 17J/17K givespurchasing guidelines, which may be used in the deÞnition ofthe ßexible pipe system and address all aspects from generaldesign parameters to detailed ßowline and riser speciÞcrequirements.

4.2.2 Applications

4.2.2.1 General

4.2.2.1.1 Flexible pipe for offshore and onshore applica-tions is grouped into either a static or dynamic category (seeFigures 2 and 3). It is used for a multitude of functions,including the following:

a. ProductionÑoil, gas, condensate, water.b. InjectionÑwater, gas, downhole chemicals.c. ExportÑsemi-processed oil and gas.d. ServicesÑwellhead chemicals, control ßuids.

4.2.2.1.2 The static and dynamic categories place differentphysical demands on the pipe. While both require long life,mechanical strength, internal and external damage resistance,and minimal maintenance, dynamic service pipes addition-ally require pliancy and high fatigue resistance.

4.2.2.2 Static Applications

4.2.2.2.1 The use of ßexible pipe for static applications isprimarily for ßowline and Þxed jacket riser service. In theseapplications, ßexible pipe is used to simplify design or

6 API RECOMMENDED PRACTICE 17B

Figure 1—Flexible Pipe Overview

Retrievaland reuse

Integrity andcondition monitoring

InstallationCommissioning

Handlingand transportation

Marking, packaging,and storage

Documentation

Factory acceptancetesting

Manufacturing

Prototype test(optional)

Component designand materials

System design

Analysisconsiderations

Pipedesign

Materialselection

Functionaldefinition

RP Section 13

RP Section 11

RP Section 12

RP Section 11

Specification Section 10RP Section 10

Specification Section 8

Specification Section 9


RP Section 9

Specification 17J App. BRP Section 7

RP Section 8

RP Section 7



Specification Section 4Purchasing Guidelines

(Specification Appendix A)

(Bend stiffeners)(Bend restrictors)

Function Activity RP/Specification Reference1

Design

Testing

Manufacturing

Installation

Operation

Note:�1RP refers to this document and Specification refers to API Specification 17J and API Specification 17K.

RECOMMENDED PRACTICE FOR FLEXIBLE PIPE 7

installation procedures, or for the inherent insulation or cor-rosion resistant properties. In addition, reduction of installa-tion and end connection loads and moments may beachieved using ßexible pipe. Examples of where the use ofßexible pipe results in simpliÞed ßowline design or installa-tion include the following (see also Figure 2):

a. Subsea ßowline end connections where expensive or difÞ-cult operations, such as exact orientation measurements forspool pieces or the use of large alignment equipment to re-position the ßowline, can be eliminated.

b. Situations involving gross movements and damage toßowlines because of mudslides can be reduced through theuse of slack sections of ßexible pipe.

c. Applications in which Þeld hardware and ßowline locationchange with the ÞeldÕs production characteristics, which maynecessitate the recovery and reuse of ßowlines.

d. Applications with uneven seabed to avoid seabedpreparation.

e. In deepwater or severe environment applications, whereßexible pipe installation is economically attractive relative torigid pipe installation. Instead of mobilizing an expensivepipelaying spread, it is often preferable to use ßexible pipeinstalled from a dynamically positioned vessel.

4.2.2.2.2 Flexible pipe ßowlines generally have internaldiameters in the range 0.05 to 0.5 meter (2 to 20 inches)although some low pressure bonded ßexible pipes such asoil suction and discharge hoses, have internal diameters upto 36 inches. Section lengths are limited by transport capa-bilities, and diameter is limited only by current manufactur-ing capability.

4.2.2.2.3 The functional requirements of a ßexible pipeßowline are generally the same as for a steel pipe ßowline.

Figure 2—Examples of Static Applications for Flexible Pipe

Early field production scheme Flowlines repositioned formature field production scheme

J-tube Flexible pipe

Rigid pipe

Manifold

Flexible pipespool piece

Rigid steelflowline


As signiÞcant dynamic loading or motions are generally notexperienced, the ßexibility properties of ßexible pipe sim-plify the project transport and installation phases.

4.2.2.3 Dynamic Applications

4.2.2.3.1 Dynamic applications use ßexible pipe betweensupply and delivery points where there is relative movementbetween these two points while in service. These types ofapplications usually involve an offshore ßoating productionfacility or terminal connected to another ßoating facility,Þxed structure, or Þxed base (see Figure 3). Examples ofdynamic applications include the following:

a. Flexible pipe risers for offshore loading systems.

b. Flexible pipe riser connections between ßoating produc-tion facilities and subsea equipment.

4.2.2.3.2 The riser conÞgurations typically used are shownschematically in Figure 4. Note in general the critical sectionsin the riser conÞgurations are at the top (or bottom), wherethere are high tensile forces (and large curvatures): at the sagbend, where there is large curvature (at low tension); and atthe hog of a wave buoyancy section, where there is large cur-vature (at low tension).

4.2.2.3.3 The present dynamic applications of ßexiblepipes have only been for the production phase. However, withthe advent of down hole motors, ßexibles may also be used asdrilling risers [5].

4.2.2.3.4 In addition to riser systems that use ßexible pipethroughout, systems that combine ßexible pipe and rigid pipein the ßow path have been used. Described as hybrid risersystems, they typically use a lower rigid riser section (such asa free standing riser) and an upper ßexible pipe section(jumper line).

4.2.2.4 Jumper Lines

4.2.2.4.1 In addition to ßowlines and risers, jumper lines, afurther category, may be used for either static or dynamicapplications. Examples of ßexible pipes used in jumper lineapplications include the following (see also Figure 5):

a. Static Application.

1. Intra-Þeld connection of wellheads and manifolds (typ-ically in lengths less than 100 meters).

2. Connection of topside wellheads and platform pipingon TLPs.

b. Dynamic Applications.

1. Connection of wellhead platforms and ßoating supportvessels.

2. Lines in FPSO turret motion transfer systems.

4.2.2.4.2 The functions of the dynamic jumper lines(excluding internal turret lines) are in many respects similarto riser systems. Their operation, however, is somewhat dif-ferent; the lines generally are more exposed to wave loading,and the conÞguration varies between the connected conditionand the stand-off condition, posing extra requirements on theend connectors and bend stiffeners. The performance of thesecomponents should be evaluated carefully for dynamicjumper line applications.

4.3 FLEXIBLE PIPE DESCRIPTION

4.3.1 General

4.3.1.1 This recommended practice does not apply toßexible pipes for use in choke and kill line or umbilicalapplications. See API SpeciÞcation 16C for choke and killline applications and API SpeciÞcation 17E for umbilicalapplications.

4.3.1.2 A ßexible pipe generally combines low bendingstiffness with high axial tensile stiffness, which is achieved bya composite pipe wall construction. This is more applicable tounbonded ßexible pipes rather than bonded ßexible pipes.The two basic components are helical armoring layers andpolymer sealing layers, which allow a much smaller radius ofcurvature than for a steel pipe with the same pressure capac-ity. Generally, a ßexible pipe is designed speciÞcally for eachapplication and is not an off-the-shelf product, although theymay be grouped according to speciÞc designs and henceapplications. This allows the pipe to be optimized for eachapplication.

4.3.2 Unbonded Flexible Pipe Construction

4.3.2.1 A typical cross-section of a ßexible pipe is shownin Figure 6. The main layers identiÞed are as follows:

a. Carcass: This is an interlocked metallic layer which pro-vides collapse resistance. An example of a carcass proÞle isshown in Figure 7.

b. Internal pressure sheath: This is an extruded polymer layerwhich provides internal ßuid integrity.

c. Pressure armor: This is an interlocked metallic layer whichsupports the internal pressure sheath and system internalpressure loads in the radial direction. Some example proÞlesfor the pressure armor wires are shown in Figure 7. A back-uppressure armor layer (generally not interlocked) also may beused for higher pressure applications.

d. Tensile armors: The tensile armor layers typically use ßat,round, or shaped metallic wires, in two or four layers cross-wound at an angle between 20 degrees and 60 degrees. Thelower angles are used for pipe constructions, which include apressure armor layer. Where no pressure armor layer is usedthe tensile armor layers are crosswound at an angle close to55 degrees to obtain a torsionally balanced pipe and to bal-ance hoop and axial loads.


Figure 3—Examples of Dynamic Applications for Flexible Pipe

Flexible riser

Floating production system (FPS)

Floating production system (FPS)

Flexibleriser

Rigid riser

Anchor chain

Flexible riser

Floating tanker/terminal mooring

Subseabuoy


Figure 4—Examples of Flexible Riser Configurations

Free hanging

Steep-S Lazy-S

Steep wave Lazy-wave


Figure 5—Examples of Flexible Pipe Jumper Line Applications

Supportvessel

Flexiblejumper

Wellheadplatform

Fixed end

Topsidespiping

End fittingXmas tree

Flexiblejumper

Moving end

Wellbay

Rigid riser

Grated deck

Tree deck

Manifold

Wellheads

Flexiblejumper


Figure 6—Schematic of Typical Flexible Riser Cross-Sections

Bonded Flexible Pipe

Unbonded Flexible Pipe

Outer sheath

Outer layer of tensile armor

Inner layer of tensile armor

Anti-wear layer

Anti-wear layer

Back-up pressure armor

Interlocked pressure armor

Internal pressure sheath

Carcass


Figure 7—Pressure Armor and Carcass Interlock Profiles

Pressure Armor Profiles

Carcass Profile

A) Z-shape

B) C-shape

C) T shape 1

D) T shape 2

Clip

T. wire T. wire


e. Outer sheath: This is an extruded polymer sheath, whichprovides external ßuid integrity.

4.3.3 Bonded Flexible Pipe Construction

A typical bonded ßexible pipe consists of several layers ofelastomer either wrapped or extruded individually and thenbonded together through the use of adhesives or by applyingheat and/or pressure to fuse the layers into a single construc-tion. An example of a bonded pipe construction is shown inFigure 6. The main layers identiÞed are as follows:

a. Carcass: This is an interlocked metallic layer which pro-vides collapse resistance. An example of a carcass proÞle isshown in Figure 7.

b. Liner: This is a wrapped or extruded elastomer layerwhich provides internal ßuid integrity.

c. Reinforcement layer: This layer is comprised typically ofhelically wound steel cables in an embedding elastomer com-pound used to sustain tensile and internal pressure load on thepipe. The steel cables are typically laid at an angle of 55degrees to obtain a torsionally balanced pipe in addition toequivalent hoop and longitudinal forces in the layer due topressure. However, this angle may increase or decreasedepending on the required strength characteristics of the pipe.For example, a higher angle may be used if increased strengthin the hoop direction is required at the expense of tensilecapacity and axial stiffness of the pipe.

d. Cover: This is a wrapped or extruded elastomer layer whichprovides external ßuid integrity and protection against externalenvironments, corrosion, abrasion, and mechanical damage.

Note that the concept of separate layers in a bonded pipeconstruction is notional as the Þnal pipe cross section is abonded composite construction.

4.3.4 Classification of Flexible Pipe

4.3.4.1 Currently, unbonded ßexible pipes can be generallyclassiÞed into three distinct families. These classiÞcations areidentiÞed in Table 1. The footnotes to Table 1 list the typicalvariations within these standard pipe design families. Thereare also distinctions within these families between pipes forstatic and dynamic applications, with the main distinctionbeing the use of anti-wear layers for dynamic applications ifthey are required to achieve service life criteria.

4.3.4.2 The classiÞcations for bonded ßexible pipe areidentiÞed in Table 2. Smooth bore ßexible pipes (ProductFamilies I, unbonded, and IV, bonded) are often used forwater injection or dead crude applications.

4.3.5 End fittings

4.3.5.1 The terminations in a ßexible pipe are described asend Þttings. A typical unbonded pipe end Þtting is illustrated

in Figure 8. End Þttings may be built in during pipe manufac-ture or installed in the Þeld. The purpose of a ßexible pipeend Þtting is twofold, namely:

a. To terminate all the strength members in the pipeÕs con-struction so that axial loads and bending moments can betransmitted into the end connector without adversely affect-ing the ßuid-containing layers.

b. To provide a pressure tight transition between the pipebody and the connector.

4.3.5.2 End connectors may be an integral part of orattached to the end Þtting. A variety of end connectors exist,such as bolted ßanges, clamp hubs, proprietary connectors,and welded joints (two end Þttings welded together to joinpipe segments into a longer segment). The selection of con-nector depends on operational and service requirements.

4.3.6 Integrated Service Umbilicals

4.3.6.1 A recent development in ßexible pipe technology isto combine the functionality of ßexible pipes with umbilicals,to form an integrated service umbilical (ISUª). A schematicof a typical ISU is shown in Figure 9. The inner core is a stan-dard ßexible pipe construction and provides the axial load-bearing capacity of the structure. The umbilical components(electrical, hydraulic, and control lines) are helically (or sinu-soidally) wound around the core pipe.

4.3.6.2 Spacers (Þllers) are included between the umbilicallines to increase the crushing load resistance of the ISU. Theassembly is covered by a protective outer sheath. In somecases, a layer of helical or sinusoidal armoring is appliedbetween the control lines and the outer sheath. This layerincreases the weight/diameter ratio of the ISU, which reducesthe dynamic motions, thereby minimizing the potential forinterference with adjacent risers. It also protects the controllines against external damage.

4.3.6.3 The end terminations of an ISU are complex con-structions. The core of the termination is the end Þtting of thecentral ßexible pipe, around which the terminations of thecontrol lines are grouped. This assembly is integrated in asteel housing or frame, which may also carry the bend stiff-ener and transfer bending loads. The detailed design of thetermination is to a large extent governed by the installationand tie-in strategy.

4.3.6.4 Stainless steel conduits may also be used in theISU. These overcome the problem of ßuid diffusion throughthe polymer hoses (in particular methanol) and reduceresponse time in control systems. However, stainless steelconduits may be sensitive to fatigue in dynamic applicationsand installation loads.


4.3.7 Multibores

4.3.7.1 The multibore concept involves combining multi-ple ßexible pipes and/or umbilical components into a singleconstruction, thus reducing the number of lines in a Þelddevelopment and thereby simplifying the Þeld layout andinstallation requirements. It may also reduce the number of Ior J-tubes required for some development options. Someexamples of multibore constructions are shown in Figure 10.The individual pipes are helically or sinusoidally wound andÞller/spacer materials are used to obtain a circular cross-sec-tion. External armoring may be applied outside the bundle. Apolymer sheath is extruded over the bundle and providesstructural integrity and protection.

4.3.7.2 The design of a multibore construction is muchmore complex than a single bore, and important consider-ations include the following:

a. The most desirable shape in a multibore structure is a cir-cular cross-section, since this results in optimal hydro-dynamic performance, efÞcient space utilization, and easyhandling during installation and retrieval.

b. Standard components (ßexible pipes and umbilicals)should be used as much as possible.

Table 1—Description of Standard Flexible Pipe Families—Unbonded Pipe

Layer No.

Layer Primary Function

Product Family I Product Family II Product Family III

Smooth Bore Pipe Rough Bore PipeRough Bore

Reinforced Pipe

1 Prevent collapse Carcass Carcass

2 Internal ßuid integrity Internal pressure sheath Internal pressure sheath Internal pressure sheath

3 Hoop stress resistance Pressure armor layer(s) Pressure armor layer(s)

4 External ßuid integrity Intermediate sheath

5 Tensile stress resistance Crosswound tensile armors Crosswound tensile armors Crosswound tensile armors

6 External ßuid integrity Outer sheath Outer sheath Outer sheath

Notes:1. All pipe constructions may include various nonstructural layers, such as anti-wear layers, tapes, manufacturing aid layers, etc.2. An external carcass may be added for protection purposes.3. The pressure layer may be subdivided into an interlocked layer(s) and back-up layer(s).4. The number of crosswound armor layers may vary, though generally is either two or four.5. Thermal insulation may be added to the pipe.6. The internal pressure and outer sheaths may consist of a number of sublayers.7. Product family III is generally used for higher pressure applications than II.8. The intermediate sheath for smooth bore pipes is optional when there is no external pressure or external pressure is less than the collapse pressure of the internal pressure sheath for the given application.

Table 2—Description of Standard Flexible Pipe Families—Bonded Pipe

Layer No.

LayerPrimary Function

Product Family IV

Product Family V

Smooth Bore Pipe

Rough Bore Pipe

1 Prevent collapse Carcass

2 Internal ßuid integrity

Liner Liner

3 Hoop and tensile load resistance

Reinforcement layer(s)

Reinforcement layer(s)

4 External ßuid integrity and protection

Cover Cover

Notes:

1. All pipe constructions may include various non structural layers, such as Þller layers and breaker fabrics.

2. An external carcass may be added for protection purposes.

3. The number of crosswound reinforcement plies may vary, though generally is either two, four, or six.


Figure 8—Example of an Unbonded Flexible Pipe End Fitting

Figure 9—Schematic Drawing of an Example ISU

Mounting flange

End fitting housing(inner casing)

End fitting housing(outer casing)

Tensile armors(embedded in epoxy)

Pressure armor layer

Outer sheath

Carcass

Seal ring

Carcass end ring

Internal pressure sheath(and sacrificial layers)

InsulatorEnd fittingneck

Electrical power cable

Outer sheath

Tape

Pipe outer sheath

Tape

Tensile armor layer

Internal tensile armor layer

Pressure armor layer

Internal pressure sheath

Carcass

Anti-friction tape

Hydraulic hose

Fiber optical cable

Filler material

Electrical signal cable


Figure 10—Examples of Multibore Constructions

101.6mm (4 inch)

25.4mm (1 inch)

Filler

101.6mm (4 inch)

63.5mm (2.5 inch) Umbilical

38.1mm (1.5 inch)

Filler

101.6mm (4 inch)

63.5mm (2.5 inch) Umbilical

38.1mm (1.5 inch)

Filler

ODWeight in air emptyMinimum bending radius

: 355.6mm (14 inch): 181 kg/m: 2.4m


: 381mm (15 inch): 201 kg/m: 2.5m


: 431.8mm (17 inch): 249 kg/m: 2.9m

Example 1

Example 2

Example 3

Note: Properties given for the three examples are typical.


c. Depending on the manufacturing process, the internalcomponents may or may not provide the axial load capacityof the structure. The axial load capacity or additional capacitymay be provided by armor layers. The structural stability (dif-fering elongations in the components) and torsional balanceof the multibore under various loading conditions (unequalpressure levels and bending) should be evaluated.

d. The crushing resistance of the multibore must be largeenough to allow for ßexibility in installation methods.

e. The maximum outer diameter is limited by the extrusioncapability of the manufacturer for the outer sheath.

f. Care should be taken during winding to minimize torsionloads induced in the individual components.

g. A symmetrical construction is recommended to ensureuniform mechanical properties and to prevent structural rear-rangement under dynamic loading.

4.3.7.3 The end termination for the multibore constructionwould typically use standard end Þttings, contained within abox type structure.

4.4 ANCILLARY COMPONENTS

4.4.1 General

Ancillary components commonly used in ßexible pipe sys-tems are described in the following sections.

4.4.2 Bend Limiters

4.4.2.1 Two types of bend limiters in common use are bendstiffeners and bellmouths, which are shown schematically inFigure 11. A third type is a bend restrictor as described in thefollowing section. Bend stiffeners and bellmouths are gener-ally used for dynamic applications; however, they may alsobe used in static applications. An example of the latter is theuse of bend stiffeners on ßowlines to prevent overbending atthe end Þtting during installation.

4.4.2.2 Bend limiters should be designed to give literallyno bending in the pipe for a length of approximately one ODfrom the end Þtting. Below this, the bending is allowed toincrease gradually, with a smooth variation of bendingmoment within MBR criteria limitations.

4.4.2.3 Bend limiters may be built into the pipe construc-tion in some bonded pipes. This is achieved by extruding orwrapping additional layers of elastomer and then curing thestructure to form an integral bend limiter and pipe.

4.4.3 Bend Restrictors

4.4.3.1 Bend restrictors are designed to mechanicallyrestrict the ßexible pipe from bending beyond its allowableMBR and are currently only used in static applications. Anexample of a bend restrictor is shown in Figure 12. Bend

restrictors are used to support a ßexible pipe over free spanswhere there is the possibility of damaging the pipe structurebecause of overbending. Typical applications are at well-head connections, J-tube exits, and rigid pipe crossovers.Restrictors also may be used to prevent overbending duringinstallation.

4.4.3.2 The restrictor normally consists of interlocking halfrings that fasten together around the pipe so that they do notaffect the pipe until a speciÞed bend radius is reached, atwhich stage they lock. Full rings may be used if the restrictoris mounted prior to the end Þtting. The locking of the restric-tor prevents further bending of the pipe and additional loadsare carried by the bend restrictor. Care should be taken thatthe locking of the rings does not damage the outer sheath ofthe pipe, i.e., there is smooth support with no sharp edges inthe restrictor design.

4.4.3.3 The bend restrictor elements may be manufacturedfrom metallic materials, creep resistant elastomers, or GRP.All materials should be selected for the speciÞed environmentand have sufÞcient corrosion resistance.

4.4.4 Connectors

4.4.4.1 The design of ßexible pipe end Þttings allows forthe use of a variety of end connectors, such as bolted ßanges,clamped hubs, and proprietary connectors. The connectorsare typically welded to the end Þtting prior to connecting tothe ßexible pipe, or they may be integrally machined from theend Þtting body.

4.4.4.2 The ßexible pipe and end Þtting may also be con-nected directly to a steel pipe, e.g., by welding. However,when the end Þtting is already connected to the ßexible pipe,welding close to the end Þtting (approximately 0.5 to 0.8meter) should not be performed, or overheating of the end Þt-ting may adversely affect the layer terminations or seals.

4.4.4.3 For dynamic riser applications, quick disconnect(QDC) and quick connect disconnect (QCDC) systems maybe used as connectors where emergency release is an opera-tional requirement. An example of a QDC system is shown inFigure 13. The main features of emergency release systemsare typically as follows:

a. Isolation ball valve in upper and lower halves of thestructure.

b. Ability to disconnect under full design loads and internalpressure.

c. Minimal size and weight for structure.

d. Full bore throughout to allow for pigging.

e. Pressure tight connection with face-to-face type primaryseals to avoid damage to seals during disconnect/reconnectand dynamic loading.


Figure 11—Bend Limiters

Figure 12—Schematic of a Bend Restrictor

I-tube

Bellmouth

Pontoon

Flexible riser

Bendstiffener

End fitting

Reaction collar

Bend restrictor

Subsea line

Section showing bend restrictor in“locked” position

Side elevation


Figure 13—Example of a Quick Disconnect (QDC) System


f. Ball valves to be interlocked with release mechanism toensure closure on disconnection (may not be required for allapplications).

g. SimpliÞed support structure (guide post funnels) to alloweasy and safe reconnection.

h. Capability to periodically test release mechanism withoutreleasing the riser or breaking primary seals (or if this is notfeasible an alternative test procedure is required whichincludes retesting of primary seals after reconnection).

4.4.4.4 Disconnect systems may have emergency shut-down valves on one or both sides of the interface. There alsomay be cases where no valve is required. Important consid-erations in this decision include: risk of disconnection,transported ßuid, environmental concerns, and topsidesvalving.

4.4.5 Subsea Buoys

4.4.5.1 Subsea buoy/arch systems are used to achieve S-shaped riser conÞgurations, including, lazy, steep, andreverse conÞgurations (note that in the reverse conÞgura-tion the lower catenary of the riser passes back under-neath the buoy). The systems typically consist of one ormore buoyancy tanks supported by a steel structure overwhich lies individual gutters (arches) for each riser. Twotypical systems are shown in Figure 14. The buoyancytanks may be constructed from either steel tanks or syn-tactic foam modules. The tanks may be positioned asshown in Figure 14.

4.4.5.2 As an alternative, the S conÞguration may beachieved by using a Þxed support instead of a ßoating buoy.An example of this is also shown in Figure 14. The main dis-advantage of this system is the reduction in compliancy of theriser system.

4.4.5.3 The subsea buoy/arch system is held in place by ariser base to which it is connected by tethers (lazy-S) or byßexible risers (steep-S). The subsea buoy/arch systems aredesigned to typically support two to six risers, though there isno theoretical limit on the number. The risers are held in placeon the arch.

4.4.6 Buoyancy Modules

4.4.6.1 Buoyancy modules are used to achieve the waveshape riser conÞgurations (lazy, steep, and pliant). A sche-matic of a typical module is shown in Figure 15. A number ofmodules (e.g., 30) are required to achieve the wave conÞgura-tion and are generally sized (both length and diameter) to beabout two to three times the pipe OD, though this depends onbuoyancy and installation requirements. The number of mod-ules is largely based on riser weight, water depth, offset

requirements, and manufacturing/commercial issues. As themodules are individually clamped to the riser, the designshould ensure that they do not slide along the pipe or damageit. Some bonded ßexible pipes have integral elastomer collarsat intervals along the pipe to facilitate the attachment of ancil-lary devices. These collars are generally built and cured withthe pipe.

4.4.6.2 The buoyancy modules are typically composed oftwo components: an internal clamp and an syntactic foambuoyancy element. A polymer (e.g., polyurethane) casingprovides impact and abrasion resistance. The internalclamp bolts directly onto the ßexible pipe, and the buoy-ancy element Þts around the clamp. The buoyancy elementis generally in two halves that are securely fastenedtogether. The density of the syntactic foam is selectedbased on the speciÞed water depth and service life. A typi-cal density is 350 kg/m3.

4.4.7 Clamping Devices

4.4.7.1 Clamping devices may be used in ßexible pipeapplications to connect ancillary components to the pipe,such as buoyancy modules, subsea arches, tethers, and bendrestrictors. In addition, bundle clamps may be used to joinseveral pipes together at discrete intervals, such as withpiggy back lines (see example in Figure 16). The main com-ponent of bundle and piggy back clamps is a spacer deviceor body, which may be in two half sections. The body is pro-vided with cylindrical recesses into which individual linesare Þtted. The assembly is joined together with bolts or a setof circumferential straps. Alternatively, band straps may beused for static piggy back assemblies where they are neededonly for installation.

4.4.7.2 Care should be taken that excessive contact pres-sure is not caused. If high contact pressure is required, sometype of protection shell should be Þtted so as to distribute theapplied load. The clamp design should also ensure that thereare no sharp edges that may cause local overbending of thepipe.

4.4.8 Riser and Tether Bases

4.4.8.1 Riser bases are used to connect ßexible risers toßowlines and may also be required to support subsea buoy/arch systems (e.g., steep-S conÞguration). Tether bases areused only to anchor subsea buoy/arch systems (e.g., lazy-SconÞguration).

4.4.8.2 The riser base may be either a gravity structure, apiled structure, or a suction/anchor pad. Selection of gravitybased or piled structure depends on applied loads and soil con-ditions. A typical riser base structure is shown in Figure 17. Asan alternative, the ßexible pipe may be connected directly to a


Figure 14—Subsea Buoy/Arch Systems

Buoys

Gutters

Risers

RisersBuoys

Gutters

Tethering point

Tethering point

Riser clamp

Option 1–Twin Buoys

Option 2–Single Buoy

Option 3–Fixed Arch

Flexible riser

Arch/gutters

Support structure

Note: The buoys may be steel tanks or syntactic foam structures.


Figure 15—Example of a Buoyancy Module for Wave Configurations

Figure 16—Example of a Clamp for Piggybacked Flexible Risers

Side Elevation Cross Section

Flexible pipe

Buoyancy module

Clamp

Riser clamp

Front Elevation Side Elevation


manifold or a PLEM, in which case the manifold or PLEMacts as the riser base.

4.4.9 Riser Hang-Off Structures

4.4.9.1 The top connection of a ßexible riser may be hung-off from the support structure (e.g., platform, tanker, semi-sub, etc.) either externally or internally. In an external con-nection, the riser, for example, would be connected to top-sides piping at pontoon level or hung-off at upper deck level,while in an internal connection the riser is typically pulledthrough an I-tube and hung-off at the top of the I-tube (seeFigure 18 for an example). The loading on the two hang-offstructures is very different, with the internal connection sub-ject only to axial loads, while the external connection experi-ences axial, bending, and shear loads.

4.4.9.2 Important considerations in the design of riserhang-off structures include the following:

a. The main constraints in the design of the hang-off struc-ture are load limitations, space limitations, and spool piecerequirements.

b. For internal connections, the design of the hang-off struc-ture should account for the weight of the riser within the I-tube.

c. For some hang-off structures, the critical loading willoccur during installation, when there may be a signiÞcantpull-in load (including friction effects).

d. Overbending of the riser at a base of an I-tube is preventedby use of a bend limiter (bend stiffener or bellmouth).

e. As the limiter is structurally supported by the I-tube, thiscan induce substantial loads on the I-tube, which should there-fore be designed for all relevant loads. Note that these loadsmay be signiÞcantly increased by the use of short spool pieces(e.g., between a bend stiffener and the base of the I-tube), andthis should be considered during design of the I-tubes.

f. In some cases, corrosion inhibitors are added to the seawa-ter inside the I-tubes, which requires the bottom of the I-tubeto be sealed to prevent loss of inhibitor. If relevant, the designof the riser installation/connection system should account forthe requirement for sealing of the I-tube. Compatibility ofcorrosion inhibitors in the I-tube with materials of the ßexiblepipe riser must be veriÞed.

Flexible riserconnections

Spool pieces to theflowlines

Concreteslab

Figure 17—Example of a Typical Riser Base [6]


5 Pipe Design Considerations

5.1 GENERAL

Section 5 of API SpeciÞcations 17J/17K speciÞes ßexiblepipe design requirements. The objective of this section is toelaborate and give guidance on ßexible pipe design consistentwith the requirements of API SpeciÞcations 17J/17K. Thissection addresses the following speciÞc issues:

a. Design process.

b. Pipe structural failure modes.

c. Design criteria.

d. Design load cases.

5.2 DESIGN OVERVIEW

The objective of this section is to give a general overviewon the typical design process for ßexible pipe applications.The design process, however, is a function of the pipe appli-cation, and a distinction is made between the process for thedesign of the following two generic ßexible pipe applications:

a. Static (applies to static riser, ßowline, and jumperapplications). b. Dynamic or loading line (applies to dynamic riser, loadingline, and jumper applications).

Design of the end Þtting is also discussed in this section.The end Þtting is considered an integral part of the pipe.

Figure 18—Example of a Typical Riser Hang-Off Structure

Carcass bleed line

Bonnet

Splitted plate

Skirt

3" Flex riser

I-tube

6" Flex riser

Top of deck

Test port


5.2.1 Static Application Design

The main design stages for static applications are repre-sented in ßowchart form in Figure 19 and are as follows:

a. Stage 1ÑMaterial selection.

b. Stage 2ÑCross-section conÞguration design.

c. Stage 3ÑSystem conÞguration design.

d. Stage 4ÑDetail and service life design.

e. Stage 5ÑInstallation design.

5.2.1.1 In Stage 1, the pipe material selection is madebased on internal environment (transported product), func-tional requirements, and material options. Materials compati-ble with the transported product are selected. See Section 6for guidelines on material selection.

5.2.1.2 In Stage 2, the cross-section conÞguration anddimensions are selected based on the pipeÕs functionalrequirements and experience in the selection of the layerstructure. Cross-section design calculations and checks typi-cally are carried out by the manufacturer using proprietarysoftware that has been validated with test data.

5.2.1.3 Stage 3 involves selection of the system conÞgura-tion. For a ßowline, this is generally a straightforward task,with the only complications typically being the design of theend sections and any requirements to accommodate the rela-tive movement envelope. However, thermal analysis,upheaval buckling, and stability analysis may dictate designrequirements in certain situations.

5.2.1.4 Stage 4 includes the detailed design of ancillarycomponents, as described in 4.4, and corrosion protection.Service life analysis is also performed at this stage as itapplies to the pipe and components.

5.2.1.5 Stage 5 completes the design process and involvesthe selection/design of the installation system, including ves-sel, equipment, methodology, and environment conditions.Stage 5 requires detailed global and local analyses to conÞrmthe feasibility of the selected installation system. For ßow-lines, this stage isÑin many casesÑcritical for the pipedesign, and it is therefore recommended that preliminaryinstallation analyses be performed at an early stage in thedesign process.

5.2.2 Dynamic Application Design

5.2.2.1 The main design stages for dynamic applicationsare represented in ßowchart form in Figure 20, and are asfollows:

a. Stage 1ÑMaterial selection.

b. Stage 2ÑCross-section conÞguration design.

c. Stage 3ÑSystem conÞguration design.

d. Stage 4ÑDynamic analysis and design.

e. Stage 5ÑDetail and service life design.

f. Stage 6ÑInstallation design.

5.2.2.2 In Stage 1, the pipe material selection is made, asfor a static ßowline, based on internal environment (trans-ported product), functional requirements, and materialoptions. In this case, materials compatible with both the trans-ported product and the dynamic service of the ßexible pipeare selected (see Section 6).

5.2.2.3 In Stage 2, the cross-section conÞguration anddimensions are selected, and design calculations and checksare carried out as for a static ßowline.

5.2.2.4 Stage 3 involves selection of the system conÞgura-tion. For a dynamic riser, this task involves selecting a pipeconÞguration from available options, some of which areshown in Figure 4. Some guidelines on the selection of riserconÞgurations are provided in 7.4.1. System conÞgurationdesign also requires the effect of ancillary components, suchas concentrated or distributed buoyancy, to be quantiÞed atthis stage.

5.2.2.5 Stage 4 involves the dynamic design of the riser orriser system. Typically, this considers the dynamic responseof the riser, subject to a series of imposed loading conditionsderived from the functional, environmental, and accidentalloads on the system. Other important issues to be addressedhere include possible interference with other system compo-nents, top tensions, departure angles, and curvatures. Suchanalysis is typically performed using Þnite element dynamicanalysis software (refer to Section 8.2.3.3).

5.2.2.6 Stage 5 includes the detailed design of ancillarycomponents, as described in 4.4, and corrosion protection.Service life analysis is also performed at this stage, as itapplies to the pipe and components. Section 7 shows guide-lines on design of the pipe system and ancillary components.

5.2.2.7 Stage 6, installation design, completes the designprocess and is largely similar to the equivalent stage in staticßowline design. For risers, however, the complexity of thesystem to be installed is generally signiÞcantly greater thanfor a ßowline.

5.2.3 End Fitting Design

The design of the end Þtting for ßexible pipes is critical.Section 4.3.5 describes the end Þttings used for ßexible pipes,while Figure 8 shows a schematic of a typical unbonded pipeend Þtting. As a minimum, the end Þtting design should meetthe requirements of API SpeciÞcations 17J/17K.


Figure 19—Static Application Design Flowchart

��Material options, functionalrequirements, internalenvironment

��

Structural options

��Guidelines, experience

��Functional requirements,monitoring requirements

��Aging, degradationrequirements, and designcriteria

��Installation data, guidelines,experience

��Installation load casesand criteria

��

Static load cases anddesign criteria

��Functional requirementsand design criteria

Select materials

Select cross-sectiondimensions

Pipe local analysis

Select pipe configurationparameters

Static global andlocal analysis

Configuration design

Design end fittings, bendrestrictors, corrosion protection

monitoring provisions, etc.

Check service life

Design/selection of installationsystem (vessel, equipment,

methodology, season)

Global and localinstallation analysis

Final static flowline design

Stage 1

Materialsselection

Stage 2

Cross-sectionconfiguration

design

Stage 3

Systemconfiguration

design

Stage 4

Detail andservice life

design

Stage 5

Installationdesign

Response not OK

Response OK

Response not OK

Response OK

Service life not OK

Service life OK

Unfeasible installation

Feasible installation


Figure 20—Dynamic Application Design Flowchart

��Material options, functionalrequirements, internalenvironment

��Structural options

��Guidelines, experience

��Functional requirements,monitoring requirements

��

Fatigue, wear, aging, degradation requirements, and design criteria, fatigueload cases

��Installation data, guidelines,experience

��Installation load casesand criteria

��

Static load cases anddesign criteria

��

Functional requirementsand design criteria

Select materials

Select cross-sectiondimensions

Pipe local analysis

Select pipe configurationand buoyancy parameters

Static global analysis

Static configuration design

Design end fittings, stiffeners,corrosion protection

monitoring provisions, etc.

Check service life

Design/selection of installationsystem (vessel, equipment,

methodology, season)

Global and localinstallation analysis

Final dynamic riser design

Stage 1

Materialsselection

Stage 2

Cross-sectionconfiguration

design

Stage 3

Systemconfiguration

design

Stage 4

Dynamicanalysis and

design

Stage 5

Detail andservice life

design

Stage 6

Installationdesign

Response not OK

Response OK

Response not OK

Response OK

��Dynamic load cases anddesign criteria

��Dynamic response loadsand design criteria

Dynamic global analysis

Dynamic design

Analyze cross-sectionfor final loads

Response not OK

Response OK

Response not OK

Response OK

Service life not OK

Service life OK

Unfeasible installation

Feasible installation


5.2.3.1 Unbonded Pipe

5.2.3.1.1 The end Þtting design for unbonded ßexiblepipes should consider the potential pipe defects identiÞed inSection 13.3. Of particular relevance are high pressure,deepwater, and the potential for pull-out of the internal pres-sure sheath from the inner seal. Critical issues include thefollowing:

a. Loss of plasticizer from internal pressure sheath.

b. Dimensional changes in sheath because of plasticizer lossand other phenomena.

c. Friction coefÞcient between seal and adjacent layers.

d. Creep and stress relaxation in sheath material.

e. Thermal coefÞcient of expansion for sheath material.

f. Variation of sheath material properties over service life.

g. Requirement for multiple layers in internal pressuresheath.

h. For vertical risers, potential support of internal carcass byinternal pressure sheath during periods when pipe is depres-surized (decompression results in no support from pressurearmor, as depressurization results in insigniÞcant frictionalforce between the sheath and supported pressure armor).

i. Number and range of temperature cycles.

j. Cool down rates during temperature cycles of end Þttingand main pipe body.

k. Variations in polymer material properties withtemperature.

l. Armor wire pull-out.

m. Epoxy degradation.

n. Corrosion.

o. Pressure and tension retaining capability.

p. Resistance to seawater ingress.

q. Resistance to external sheath pull-out during installation.

5.2.3.1.2 The design of the end Þtting internal crimping/sealing mechanism, for PVDF based pipes in particular, iscritical, for riser applications, the effectiveness of the seal canbe reduced by large temperature cycles, high thermal expan-sion coefÞcient, plasticizer loss, or use of a multiple layerconstruction for a PVDF internal pressure sheath. The end Þt-ting design should be veriÞed with high temperature cyclingtests (see Appendix A for guidelines). These tests should berepresentative of service conditions, including thermal anddynamic loading, and the effect of plasticizer loss as applica-ble. For new designs the prototype tests of Section 9 shouldalso be considered.

5.2.3.2 Bonded Pipe

5.2.3.2.1 The end Þtting design for bonded ßexible pipesshould consider the potential pipe defects identiÞed in Sec-tion 13.3. Issues of particular relevance include high pres-sure, deepwater, the potential for pull-out of reinforcingcables and loss of ßuid seal integrity. Critical issues includethe following:

a. Change of pipe body, particularly liner material propertiesover service life.

b. Dimensional changes in pipe body due to the highly elas-tic nature of pipe body elastomer material.

c. Bonding of liner material layers and bonding of liner toremainder of ßexible pipe body.

d. Reinforcing armor pull-out.

e. Epoxy degradation.

f. Corrosion.

g. Pressure and tension retaining capability.

h. Resistance to seawater ingress.

i. Integrated gasket integrity.

j. Crimping over-pressure applied.

k. Number and range of temperature and pressure cycles.

l. Incorporation of integrated bend stiffeners.

5.3 FAILURE MODES

It is important to design a ßexible pipe in the knowledge ofthe potential degradation and failure modes for the intendedapplication. This section notes those failure modes which areexplicitly considered in ßexible pipe structural design calcu-lations. It is important to note that other modes of pipe degra-dation and failure possibly may be implicitly provided for indesign (e.g., through materials selection, see Section 6) or areconsidered elsewhere (e.g., as part of manufacture, see Sec-tion 10; or handling, transportation and installation, see Sec-tion 11).

Tables 3 and 4 provide a list of pipe failure modes that areexplicitly provided for in typical unbonded and bonded pipedesign respectively, and identiÞes relevant failure mecha-nisms and appropriate design strategies/solutions. The designsolutions should, in all cases, meet the design criteria speci-Þed in Section 5.2 of API SpeciÞcations 17J/17K. A morecomplete, though not exhaustive, list of potential pipe defectsfor ßowline and riser applications is presented in Tables 29through 31 of Section 13.3. Furthermore, some of the modesidentiÞed in Tables 3, 4, 29, 30 and 31 are being addressed bycontinuing design improvements and may therefore not berelevant to future pipe designs.


Table 3—Check List of Failure Modes for Primary Structural Design of Unbonded Flexible Pipe [Detailed Listing of Failure Modes shown in Section 13]

Pipe Global Failure Mode to Design Against Potential Failure Mechanisms

SA or DA1

Design Solution/Variables [Ref. API Spec 17J Design Criteria]

Collapse 1. Collapse of carcass and/or pressure armor due to exces-sive tension.

SA, DA 1. Increase thickness of carcass strip, pressure armor or internal pressure sheath (smooth bore collapse).

2. Collapse of carcass and/or pressure armors due to excess external pressure.

SA, DA 2. Modify conÞguration or installation design to reduce loads.

3. Collapse of carcass and/or pressure armor due to installa-tion loads or ovalisation due to installation loads.

SA, DA 3. Add intermediate leak-proof sheath (smooth bore pipes).

4. Collapse of internal pressure sheath in smooth bore pipe. SA, DA 4. Increase the area moment of inertia of carcass or pressure armor.

Burst 1. Rupture of pressure armors because of excess internal pressure.

SA, DA 1. Modify design, e.g., change lay angle, wire shape, etc.

2. Rupture of tensile armors due to excess internal pressure. SA, DA 2. Increase wire thickness or select higher strength material if feasible.

3. Add additional pressure or tensile armor layers.

Tensile failure 1. Rupture of tensile armors due to excess tension. SA, DA 1. Increase wire thickness or select higher strength material if feasible.

2. Collapse of carcass and/or pressure armors and/or inter-nal pressure sheath due to excess tension.

SA, DA 2. Modify conÞguration designs to reduce loads.

3. Snagging by Þshing trawl board or anchor, causing over-bending or tensile failure.

SA, DA 3. Add two more armor layers.

4. Bury pipe.

Compressive failure 1. Birdcaging of tensile armor wires. SA, DA 1. Avoid riser conÞgurations that cause excessive pipe com-pression.

2. Compression leading to upheaval buckling and excess bending (see also Upheaval Buckling failure mode).

SA, DA 2. Provide additional support/restraint for tensile armors, such as tape and/or additional or thicker outer sheath.

Overbending 1. Collapse of carcass and/or pressure armor or internal pressure sheath.


2. Rupture of internal pressure sheath. SA, DA

3. Unlocking of interlocked pressure or tensile armor layer. SA, DA

4. Crack in outer sheath. SA, DA

Torsional failure 1. Failure of tensile armor wires. SA, DA 1. Modify system design to reduce torsional loads.

2. Collapse of carcass and/or internal pressure sheath. SA, DA 2. Modify cross-section design (e.g. change lay angle of wires, add extra layer outside armor wires, etc.) to increase torsional capacity.

3. Birdcaging of tensile armor wires. SA, DA

Fatigue failure 1. Tensile armor wire fatigue. DA 1. Increase wire thickness or select alternative material, so that fatigue stresses are compatible with service life requirements.

2. Pressure armor wire fatigue. DA 2. Modify design to reduce fatigue loads.

Erosion 1. Of internal carcass. SA, DA 1. Material selection.

2. Increase thickness of carcass.

3. Reduce sand content.

4. Increase MBR.

Corrosion 1. Of internal carcass. SA, DA 1. Material selection.

2. Of pressure or tensile armor exposed to seawater, if applicable.

SA, DA 2. Cathodic protection system design.

3. Of pressure or tensile armor exposed to diffused product. SA, DA 3. Increase layer thickness.

4. Add coatings or lubricants.

Notes:1. SA = static application, DA = dynamic application.2. Burst, tensile, overbending and torsional failure are not considered in isolation for Þnal design of the ßexible pipe.3. Refer to Tables 29 through 31 for defects important in end Þtting designs.


Table 4—Check List of Failure Modes for Primary Structural Design of Bonded Flexible Pipe [Detailed Listing of Failure Modes presented in Section 13]

Pipe Global Failure Mode to Design Against Potential Failure Mechanisms

SA or DA1

Design Solution/Variables [Ref. API Spec 17K Design Criteria]

Collapse 1. Collapse of carcass due to excessive tension. SA, DA 1. Increase thickness of carcass strip, or pipe body (smooth bore collapse).

2. Collapse of carcass due to excess external pressure. SA, DA 2. Modify conÞguration or installation design to reduce loads.

3. Collapse of carcass and due to installation loads or ovalisation due to installation loads.

SA, DA 3. Increase the area moment of inertia of car-cass.

4. Collapse of pipe in smooth bore pipe.

Burst 1. Rupture of reinforcing armors due to excess inter-nal pressure.

SA, DA 1. Modify design, e.g. change lay angle, cable type, etc.

2. Increase cable thickness or select higher strength material if feasible.

3. Add additional reinforcing armor layers.

Tensile Failure 1. Rupture of reinforcing armors due to excess ten-sion.

SA, DA 1. Increase cable thickness or select higher strength material if feasible.

2. Collapse of carcass and/pipe body sheath due to excess tension.


3. Snagging by Þshing trawl board or anchor, caus-ing overbending or tensile failure.

SA, DA 3. Add two more armor layers.

4. Bury pipe.

Compressive Failure 1. Compression leading to upheaval buckling and excess bending (refer to Upheaval Buckling fail-ure mode also).

SA, DA 1. Avoid riser conÞgurations which cause exces-sive pipe compression.

Overbending 1. Collapse of carcass or pipe body. SA, DA 1. Modify conÞguration designs to reduce loads.

2. Rupture of liner SA, DA

3. Crack/tear in outer sheath. SA, DA

Torsional failure 1. Failure of tensile armor wires. SA, DA 1. Modify system design to reduce torsional loads.

2. Collapse of carcass and/or liner. SA, DA 2. Modify cross-section design (e.g. change lay angle of wires, add extra layer outside armor wires, etc.) to increase torsional capacity.

3. Birdcaging of tensile armor wires. SA, DA

Fatigue failure 1. Tensile armor wire fatigue. DA 1. Increase wire thickness or select alternative material, so that fatigue stresses are compati-ble with service life requirements.

2. Pressure armor wire fatigue. DA 2. Modify design to reduce fatigue loads.

Erosion 1. Of internal carcass. or liner SA, DA 1. Material selection.

2. Increase thickness of carcass.

3. Reduce sand content.

4. Increase MBR.

Corrosion 1. Of internal carcass. SA, DA 1. Material selection.

2. Of pressure or tensile armor exposed to seawater, if applicable.

SA, DA 2. Cathodic protection system design.

3. Of pressure or tensile armor exposed to diffused product.

SA, DA 3. Increase layer thickness.

4. Add coatings or lubricants.

Notes:1. SA = Static Application, DA = Dynamic Application2. Burst, tensile, overbending and torsional failure are not considered in isolation for Þnal design of the ßexible pipe.3. See Tables 29 through 31 for defects important in end Þtting design.


5.4 DESIGN CRITERIA

5.4.1 Unbonded Flexible Pipe

5.4.1.1 Introduction

5.4.1.1.1 The design criteria for unbonded ßexible pipesare given in Section 5.3.1 of API SpeciÞcation 17J in terms ofthe following:

a. Strain (polymer sheath).b. Creep (internal pressure sheath).c. Stress (metallic layers and end Þtting).d. Hydrostatic collapse (buckling load).e. Mechanical collapse (stress induced from armor layers).f. Torsion.g. Crushing collapse and ovalisation (during installation).h. Compression (axial and effective).i. Service life factors.

5.4.1.1.2 These are discussed further in the following sec-tions, which give some guidance on their derivation. In addi-tion, criteria are also introduced which provide for designagainst failure additional to the criteria speciÞed in API Spec-iÞcation 17J.

5.4.1.1.3 The criteria speciÞed by API SpeciÞcation 17Japply to the materials currently used in ßexible pipe applica-tions. Where new materials are proposed or used, the designcriteria for the new materials should give at least the safetylevel speciÞed in this recommended practice and API SpeciÞ-cation 17J. The design criteria should consider all materialcharacteristics, such as susceptibility to creep, fatigue, exces-sive strain, cracking, etc.

5.4.1.1.4 SimpliÞed approaches exist for the approxima-tion of pipe characteristics (axial, bending and torsional stiff-ness, etc.) and for calculating loads in the individual layers.These simpliÞed methodologies may be used for preliminarycomparison of design loads with design criteria. For Þnaldesign calculations, however, a veriÞed (with prototype tests)methodology is to be used, as deÞned in Section 5.2.1 of APISpeciÞcation 17J.

5.4.1.2 Strain

5.4.1.2.1 A critical parameter in the design of the internalpressure and outer sheaths is the allowable strain. Table 6 inAPI SpeciÞcation 17J speciÞes allowable strain values for themost commonly used materials. For materials not explicitlyprovided for in Table 6 of API SpeciÞcation 17J, the allow-able strain is speciÞed by the manufacturer.

5.4.1.2.2 Allowable strains have been veriÞed by materialtests performed under relevant service and ageing conditions.A safety factor is typically applied to results of such tests toderive the allowable strain of the material over its service life,accounting for material ageing and degradation in the appro-priate environment.

5.4.1.2.3 Section 5.3.1 of API SpeciÞcation 17J also pro-vides for the calculation of minimum bend radius (MBR) toprevent locking of the interlocked pressure armor wires.

5.4.1.3 Creep

5.4.1.3.1 Under normal service conditions, the internalpressure sheath will creep into gaps in the pressure or ten-sile armor layer as a result of pressure and temperatureeffects. If the sheath is too thin or the gap too large, theinternal pressure sheath will creep until a failure (leakage)occurs. Creep of the sheath at the end Þtting seal is also animportant issue (see 5.2.3).

5.4.1.3.2 The design of the internal pressure sheath (wallthickness) should therefore account for creep. The main fac-tors to be accounted for are material properties, layer thick-ness, pressure or tensile armor geometry, temperature, andpressure. Two methodologies are currently used to determinethe wall thickness required to prevent creep failure:

a. Physical tests to determine the required wall thickness.

b. Finite element analyses, calibrated with gap span test data,to determine the required wall thickness.

5.4.1.3.3 The creep design criterion speciÞed in Table 6 ofAPI SpeciÞcation 17J is based on both of these methodolo-gies. This speciÞes the maximum allowable reduction in wallthickness below the minimum design value under all loadconditions.

5.4.1.4 Stress

5.4.1.4.1 The design stress criteria (utilization factors)given in Tables 6 and 7 of API SpeciÞcation 17J were derivedto give acceptable factors of safety against failure. These fac-tors prescribe the maximum nominal applied stress as a pro-portion of the structural capacity of steel materials (deÞnedby Section 5.3.1.4 of API SpeciÞcation 17J). The utilizationfactors make implicit allowance for the presence of residualwire stress.

Note: The published utilization factors relate to steel materials. Noinference may be made about allowable stress in new materialsbased on these values.

5.4.1.5 Hydrostatic Collapse

5.4.1.5.1 Utilization factors which relate to buckling of theinternal carcass under hydrostatic pressure are speciÞed inTable 6 of API SpeciÞcation 17J as a function of water depth,with a higher permissible utilization factor (smaller safetyfactor) allowed for deep water applications. This is so that thesafety factor (the reciprocal of the utilization factor) is relatedto the absolute, rather than relative, margin between collapseand design depth.

5.4.1.5.2 Hydrostatic collapse calculations should be per-formed for both an intact outer sheath and a breached outersheath (i.e., seawater penetrated into the annulus), with the


hydrostatic collapse resistance taken as the minimum of thetwo collapse pressure values. If analytical methods are usedfor calculating collapse resistance, these should be based onan assumed initial ovalisation. This ovalisation should beselected by the manufacturer, based on manufacturing toler-ance limits and residual ovalisation from the installation pro-cess. If no other data exists, a minimum ovality of 0.2 percentshould be used.

5.4.1.5.3 The collapse resistance for smooth bore pipesshould also be calculated based on the resistance of the inter-nal pressure sheath only, and standard analytical methodsmay be used. If the collapse to design ratio is below therequired value, then it should be speciÞed that sufÞcient inter-nal pressure be maintained to prevent collapse (such as byensuring line is full of liquid at hydrostatic pressure). Alterna-tively, an impermeable intermediate sheath should be pro-vided to ensure that the pressure armor provides the requiredcollapse resistance.

5.4.1.5.4 References [5] and [7] give recommended pro-cedures for calculating the hydrostatic buckling load (col-lapse pressure) of a carcass. However, these procedures arefor the carcass layer alone. In pipe designs which include apressure armor layer, this layer assists the carcass and signif-icantly increases the collapse strength of the pipe. Whenused, methodologies for calculating the collapse strength(design water depth) of a ßexible pipe with contributionfrom the pressure armor layer should be veriÞed by docu-mented prototype tests.

5.4.1.6 Mechanical Collapse

5.4.1.6.1 The utilization factors which relate to mechanicalcollapse of the internal carcass due to excessive tension arespeciÞed in Table 6 of API SpeciÞcation 17J and, from Note(a) of this table, are identical to the utilization factors for thetensile and pressure armors.

5.4.1.6.2 The contribution of all supporting steel layersmay be taken into account when designing against mechani-cal collapse.

5.4.1.7 Torsion

5.4.1.7.1 The ßexible pipe should have a torsional strengthsufÞcient to withstand torsional loads induced during installa-tion and service conditions without any structural damage.The torsional stiffness indicates the resistance of a ßexiblepipe to rotation around its axis under a torsional moment andis a performance characteristic of the pipe.

5.4.1.7.2 The maximum acceptable torsion derives fromthe following two scenarios, depending on the direction of theapplied torsion:

a. The outer tensile armor layer is turned inwards andpressed against the internal layer (in which case the allow-able tension causes overstressing of the tensile armor) by

inducing a stress corresponding to its structural capacity(deÞned by Section 5.3.1.4 of API SpeciÞcation 17J multi-plied by the utilization factor, as speciÞed in Table 6 of APISpeciÞcation 17J).

b. The outer tensile armor layer is turned outwards andpressed against the outer layers, leading to a gap betweenthe two tensile armor layers in which case, the damagingtorsion, induces a gap between tensile armor layers, (inwhich case, the damaging torsion, induces a gap betweentensile armor layers equal to half the thickness of the tensilearmor wire). The allowable torsion for this case should becalculated from the damaging torsion using a safety factornot less than 1.0.

5.4.1.8 Crushing Collapse and Ovalisation

5.4.1.8.1 During conventional laying operations, the ten-sion in the ßexible pipe is generally controlled with a ten-sioner or with a laying winch. The load applied to the ßexiblepipe, when tightening it in a tensioner or unreeling/reeling theßexible pipe under tension (possibly over a V-shaped sheave),has to be controlled to avoid sudden collapse (or signiÞcantovalisation) of the structure or overstressing of the metalliclayers. The tension loads and crushing effect on the structureduring installation should be accounted for in the design ofthe ßexible pipe.

5.4.1.8.2 The feasibility of installing the ßexible pipe withthe selected procedure should be evaluated, considering thefollowing effects:

a. Crushing of the ßexible pipe under radial compression in atensioner.

b. Crushing effect on a laying pulley or sheave.

c. Damaging pull of the ßexible pipe at the top of thecatenary.

5.4.1.8.3 The collapse load should be calculated based onthe resistance of the internal carcass and supporting pressurelayers (pressure armor and ßat steel spiral), as applicable.Two alternative approaches are recommended for the collapsecalculation: Þnite element analysis or analytical/empiricalformulae, which have been calibrated against full scale tests.

5.4.1.8.4 The following load cases should be investigated,as applicable:

a. Reeling/unreeling on a sheave of a ßexible pipe submittedto design maximum axial load.

b. Radial compression in a tensioner of a ßexible pipe sub-mitted to design maximum axial load.

5.4.1.8.5 The minimum of the following two limits shouldthen be taken as the design maximum allowable installationtension:

a. The axial tension or radial compression in the ßexible pipeshould remain less than that which induces a stress corre-


sponding to the structural capacity of the pressure or tensilearmors (deÞned by Section 5.3.1.4 of API SpeciÞcation 17Jmultiplied by the utilization factor for installation, as speci-Þed in Table 6 of API SpeciÞcation 17J).

b. The effective tension or radial compression in the ßexiblepipe should be less than that which induces mechanical col-lapse, multiplied by the utilization factor for installation, asspeciÞed in Table 6 of API SpeciÞcation 17J.

5.4.1.8.6 In addition, the maximum permanent ovalisationof the pipe for both installation methods should be less thanthe value of initial ovalisation used for hydrostatic collapsecalculations (see 5.4.1.5).

5.4.1.9 Compression

5.4.1.9.1 A ßexible pipe may be subject to two types ofcompression: namely, effective compression (negative effec-tive tension) and axial (or true wall) compression. Effectivecompression will cause increased deformations in the pipe,while axial compression may potentially cause birdcaging inthe tensile armor layer. The behaviour of ßexible pipe undercompressive load is based on the pipe temperature.

5.4.1.9.2 The potential for both types of compression tooccur should be checked in the design of the ßexible pipe sys-tem. If effective compression occurs, the following designcriteria should be veriÞed:

a. The effective compression should be less than that whichwould cause the MBR criteria to be violated (see Table 6,Section 5.3.1 of API SpeciÞcation 17J).

b. Bar buckling of the pipe should not occur.

5.4.1.9.3 The maximum axial compression for anunbonded ßexible pipe should be calculated as the valuewhich causes a gap between the tensile armor wires and theunderlying layer equal to half the thickness of the armor wire.The allowable axial compression for stress and stabilityshould be calculated from the maximum axial compressionusing a safety factor not less than 1.0, and any axial compres-sion experienced by the pipe should be less than the allow-able. Tensile wire buckling analysis should also be conducted.

5.4.1.10 Service Life Factors

Section 8.2.4 presents a more detailed discussion of servicelife analysis, including fatigue calculations. The criteria forfatigue calculations are speciÞed in Section 5 of API SpeciÞ-cation 17J. Furthermore, permissible levels of degradationshould be deÞned for the service life analysis. Recommenda-tions on these are given in Table 5.

5.4.2 Bonded Flexible Pipe


5.4.2.1.1 The design criteria for bonded ßexible pipes areshown in Section 5.3.1 of API SpeciÞcation 17K in terms ofthe following:

¥ Strain (elastomer layers).¥ Stress/load (reinforcement layers, carcass and end

Þtting).¥ Hydrostatic collapse (buckling load).¥ Mechanical collapse (stress induced from reinforce-

ment layers).¥ Crushing collapse and ovalisation (during installation)¥ Service life factors.

5.4.2.1.2 These are discussed further in the following sec-tions, which give some guidance on their derivation. In addi-tion, criteria are also introduced which provide for designagainst failure additional to the criteria speciÞed in API Spec-iÞcation 17K.

5.4.2.1.3 The criteria speciÞed by API SpeciÞcation 17Kapply to the materials currently used in bonded ßexible pipeapplications. Where new materials are proposed or used, thedesign criteria for the new materials should give at least thesafety level speciÞed in this recommended practice and APISpeciÞcation 17K. The design criteria should consider allmaterial characteristics, such as ageing, fatigue, excessivestrain, etc.

5.4.2.1.4 SimpliÞed approaches exist for the calculation ofpipe characteristics (axial, bending and torsional stiffnessetc.) and for calculating loads in the individual materials ofthe pipe (reinforcing cables, elastomer body etc.). These sim-pliÞed methodologies may be used for preliminary compari-son of design loads with design criteria. For Þnal designcalculations, however, a veriÞed (with prototype tests) meth-odology is to be used, as deÞned in Section 5.2.1 of APISpeciÞcation 17K.

5.4.2.1.5 Due to the composite nature of bonded ßexiblepipes, the veriÞed design methodology should account forinteraction between metallic and elastomer components, andfor load sharing between different layers and components inparticular at and adjacent to the end Þtting.

5.4.2.1.6 Two distinctly types of design methodology areused by bonded ßexible pipe manufacturers. Some manufac-turers use analytical or Þnite element methods to account forthe load sharing between the various components making upthe bonded pipe. Others use standard analytical methodsderived from geometrical considerations of the pipe in con-junction with empirical efÞciency factors. The efÞciency fac-tors are calculated based on prototype tests for example burstand tensile tests.

5.4.2.2 Strain

5.4.2.2.1 Table 7 of API SpeciÞcation 17K speciÞes allow-able strain values for elastomer layers as a maximum of 50%of design maximum strain for aged material. Due to the typi-cally large strain capacity of elastomer materials used in themanufacture of bonded ßexible pipes this design criterionmay not be as critical as it is for the thermoplastic materialsused in the manufacture of unbonded ßexible pipe.


Table 5—Recommended Allowable Degradation for Unbonded Pipes

Component Degradation Mode Recommendation

Carcass 1. Corrosion Limited corrosion acceptable provided structural capacity and functional requirements are maintained.

2. Erosion Same as for corrosion.

Internal Pressure Sheath 1. Creep Limited creep acceptable provided:

Ð Structural capacity to bridge gaps maintained.

Ð No cracks.

Ð No locking of carcass or pressure armor layers.

Ð No leakage.

Ð Sealing maintained at end Þttings.

2. Thermal/chemical degradation Capacity at design life to remain within speciÞed usage factors with maximum gaps between layers. No leakage allowed. Increased perme-ation allowed, if system has been designed for the increased level of per-meation. Important considerations are increased damage rates (corrosion, HIC, SSC) for armors and limits on gas venting system capacity. Strain capacity sufÞcient to meet the design requirements of Table 6, API Spec-iÞcation 17J.

3. Cracking No cracking because of dynamic service.

Pressure and Tensile Armors

1. Corrosion Only general corrosion accepted. No crack initiation acceptable.

2. Disorganization or locking of armoring wires

No disorganization of armoring wires when bending to minimum bend radius.

3. Fatigue and Wear See Section 8.2.4

Anti-Wear Layer 1. Wear No wear through the thickness of the layer over its service life.

Intermediate Sheath 1. Thermal degradation Functional requirements are maintained.

Thermal Insulation 1. Thermal degradation Insulation capacity to be maintained equal to or above minimum speci-Þed value.

Outer Sheath 1. General degradation Strain capacity sufÞcient to meet the design requirements of Table 6, API SpeciÞcation 17J.

2. Radial deformation (loosening) No loosening that will cause disorganization of armor wires or strain fail-ure of outer sheath material.

3. Breaching No breaching allowed unless pipe design under ßooded annulus condi-tions can be shown to meet the design requirements and remaining ser-vice life requirements.

End Fitting and Carcass/Sheath Interface

1. Corrosion No corrosion acceptable which results in reduction of capacity, possibil-ity for leakage, or damage to any sealing or locking mechanism.


5.4.2.2.2 Section 5.3.1 of API SpeciÞcation 17K also pro-vides for the calculation of minimum bend radius (MBR) toprevent damage to the interlocked inner or outer carcass ifpresent.

5.4.2.3 Stress/Load

5.4.2.3.1 The design stress and load criteria (utilizationfactors) given in Tables 7 and 8 of API SpeciÞcation 17Kwere derived to give acceptable factors of safety against fail-ure. These factors prescribe the maximum nominal appliedstress or load as a proportion of the structural capacity of steelmaterials (deÞned by Section 5.3.1 of API SpeciÞcation17K). The utilization factors make implicit allowance for thepresence of residual wire stress.

Note: The published utilization factors relate to steel materials. Noinference may be made about allowable stress in new materialsbased on these values.

5.4.2.4 Hydrostatic Collapse

5.4.2.4.1 Utilization factors which relate to buckling of theinternal carcass under hydrostatic pressure are speciÞed inTable 7 of API SpeciÞcation 17K as a function of waterdepth, with a higher permissible utilization factor (smallersafety factor) allowed for deep water applications. This is sothat the safety factor (the reciprocal of the utilization factor) isrelated to the absolute, rather than relative, margin betweencollapse and design depth.

5.4.2.4.2 If analytical methods are used for calculating col-lapse resistance, these should be based on an assumed initialovalisation. This ovalisation should be selected by the manu-facturer, based on manufacturing tolerance limits and residualovalisation from the installation process. If no other dataexists, a minimum ovality of 0.2 percent should be used.

5.4.2.4.3 The collapse resistance for smooth bore pipesshould be calculated based on the resistance of the pipe body,and standard analytical methods may be used. If the collapseto design ratio is below the required value and if the pipe is notdesigned to be collapsible then it should be speciÞed that sufÞ-cient internal pressure be maintained to prevent collapse (suchas by ensuring line is full of liquid at hydrostatic pressure.)

5.4.2.4.4 References [5] and [7] give recommended proce-dures for calculating the hydrostatic buckling load (collapsepressure) of a carcass.

5.4.2.5 Mechanical Collapse

Refer to Section 5.4.1.6.

5.4.2.6 Torsion


5.4.2.7 Crushing Collapse and Ovalisation


5.4.2.8 Compression

Refer to Sections 5.4.1.9.1 and 5.4.1.9.2.

5.4.2.9 Service Life Factors

Section 8.2.4 presents a more detailed discussion of servicelife analysis, including fatigue calculations. The criteria forfatigue calculations are speciÞed in Section 5 of API SpeciÞ-cations 17K. Furthermore, permissible levels of degradationshould be deÞned for the service life analysis. Recommenda-tions on these are given in Table 6.

5.5 LOAD CASES

5.5.1 General

5.5.1.1 The ßexible pipe is to be designed to satisfy itsfunctional requirements under loading conditions corre-sponding to the internal environment, external environment,system requirements, and service life deÞned by the pur-chaser of the pipe.

5.5.1.2 All potential load cases for the ßexible pipe system,including manufacture, storage, transportation, testing, instal-lation, operation, retrieval, and accidental events are to bedeÞned by the manufacturer in the design premise speciÞedby Section 8.2 of API SpeciÞcation 17J/17K. The designpremise should specify a load case matrix which deÞnes allnormal, abnormal, installation, and fatigue loading conditionsaccording to requirements speciÞed by the purchaser inAppendix A of API SpeciÞcations 17J/17K.

5.5.1.3 The recommended annual probabilities of occur-rence for installation, and normal and abnormal loads aregiven in Table 7 for a 20-year service life. These may bechanged for different service lives. When combining annualprobabilities of waves and currents for 100 year conditions,the following two load combinations should be consideredunless more speciÞc data is available:

a. 100-year wave combined with 10-year current

b. 10-year wave combined with 100-year current

5.5.1.4 The requirement to analyze load cases for acciden-tal events should be based on an assessment of the probabilityof the events occurring. The accidental events typically con-sidered for static applications include impact from trawlboards and dropped objects. For dynamic applications, acci-dental events typically considered include one or more moor-ing lines broken and partial loss of buoyancy. Furthermore,for dynamic applications, consideration should be given toperforming extreme event load cases (e.g., events with proba-bilities of occurrence equal to or less than 10Ð4) to assess therobustness of the design.

5.5.1.5 The load case matrix constitutes the full set of load-ing conditions examined as part of the structural analysis and


Table 6—Recommended Allowable Degradation for Bonded Pipe

Component Degradation Recommendation

Carcass 1. Corrosion Limited corrosion acceptable provided structural capacity and functional requirements are maintained.

2. Erosion Same as for corrosion.

Liner 1. Blistering, delamination No blistering, delamination or leakage paths because of gas rapid decom-pression. Damage due to dissection process should be ignored.

2. Thermal/chemical degradation No leakage allowed. Increases permeation allowed, if system has been designed for the increased level of permeation. Limited degradation accept-able provided sealing is maintained at end Þtting in addition to the above.

Reinforced Layers 1. Corrosion No corrosion acceptable which results in increase in utilization of cables in reinforcing layer to beyond allowable values shown in Table 7 of API Speci-Þcation 17K.

2. Fatigue and Wear See Section 8.2.4.

Cover 1. General degradation Strain capacity sufÞcient to meet the design requirements of Table 7, API SpeciÞcation 17K.

End Fitting 1. Corrosion No corrosion acceptable which results in reduction of capacity, possibilities for leakage, or damage to any sealing or locking mechanism.

Table 7—Recommendations on Annual Probabilities for Installation, and Normal and Abnormal Operation for a 20 Year Service Life

Type ofLoad

Service Condition

Installation Service

Normal Service Abnormal Service

Functional Expected, speciÞed or extreme value. Expected, speciÞed or extreme value.

Expected, speciÞed or extreme value.

External Environmental

Probability of exceedance according to season and duration of installation period.

Yearly probability of exceedance > 10Ð2.

Yearly probability of exceed-ance between 10Ð2 and 10Ð4.

If abandonment is possible, the maximum weather in a period 3 times the expected installation duration may be used.

If combined with an accidental load the environmental load may be reduced such that the yearly probability of joint occurrence is >10Ð2.

If combined with an accidental load the environmental load may be reduced such that the yearly probability of joint occurrence is >10Ð4.

If abandonment is impossible, a more con-servative approach shall be used or the duration of the operation reduced to a period where reliable weather forecast is available (typically hours).

Accidental As appropriate to installation method. As appropriate to normal operation conditions, i.e., annual probability >10Ð2

Individual considerations. Yearly probability between 10Ð2 and 10Ð4.

Notes:1. Yearly probabilities of 10Ð2 and 10Ð4 are equivalent to return periods of 100 years and 10,000 years respectively.2. See Section 5.1.3.2 of API SpeciÞcations 17J/17K for load combination requirements.


design process. SpeciÞc load cases form inputs to Þve stagesin the overall pipe design, as follows:

a. Cross-section conÞguration design (local analyses).b. System conÞguration design (static global and localanalyses).c. Dynamic analysis and design (global analyses for dynamicriser design only).d. Detail and service life design (Þnal local and service lifeanalyses).e. Installation design (global and local analyses).

5.5.1.6 These stages are illustrated in Figures 19 and 20 forthe static ßowline (or static riser) and dynamic riser (ordynamic jumper) design processes, respectively, and are dis-cussed further in the following sections.

All stages of the design process involve either global orlocal (cross-section) analyses of the ßexible pipe. The pri-mary objectives of the global analyses are to verify that themain design criteria are satisÞed (e.g., MBR, allowable ten-sion, and stability of dynamic motions) and to identify criti-cal load combinations. Local analysis is then performed toverify that these critical global load combinations do notexceed the criteria speciÞed in Section 5.3.1 of API SpeciÞ-cation 17J/17K.

5.5.2 Cross Section Configuration Design

The results of initial local analyses, to determine burstpressure, response to FAT pressure, MBR, collapse depth,damaging tension, thermal properties, apparent weight in sea-water, drag to apparent weight ratio, etc., provide information

which may be compared with design requirements (waterdepth, design pressure, etc.) and experience to arrive at a pre-liminary cross-section design. This initial cross-sectiondesign may be subsequently modiÞed based on the resultsfrom the remaining stages in the design process. For deepwater applications, in particular it may be necessary to con-sider installation loads at the start of the design process.

5.5.3 System Configuration Design

5.5.3.1 Input to this stage includes all static loads relatingto the system design. The pipe is analyzed under all func-tional, environmental, and accidental loading combinationsderiving from the internal environment (pressure, tempera-ture, ßuid composition), deÞned by Table 1 of API SpeciÞca-tion 17J and Table 2 of API SpeciÞcation 17K, and the staticcomponents of the external environment, deÞned by Table 2of API SpeciÞcation 17J and Table 2 of API SpeciÞcation17K. In this context, functional, environmental, and acciden-tal loads are deÞned by Table 5 of API SpeciÞcation 17J andTable 6 of API SpeciÞcation 17K.

5.5.3.2 Examples of the global static analysis load caseswhich form an input to this process include thermal analysis,upheaval buckling load cases (static ßowlines only), on-bot-tom stability load cases (static ßowlines only), and/or staticglobal conÞguration load cases. A typical example of the glo-bal static analysis load cases relating to this stage of design ispresented in Table 8. In this phase of the design, local analy-ses are generally only required for static applications. Fordynamic applications, the local analyses are performed inStage 4 of Figure 20. For static applications, local analysisload cases should include all relevant test, installation, andoperational load cases. A typical example of the local analysisload cases relating to this stage of design is as follows:

Case A Design pressure, mean tension, bending to maxi-mum expected curvature.

Case B No internal ßuid, external hydrostatic pressure atmaximum water depth, damaged outer sheath.

Case C Maximum axial compression.

5.5.4 Dynamic Analysis and Design

5.5.4.1 Load cases for this stage relate only to dynamicriser (or jumper) applications and includes all dynamic loadsfor the global system design. The pipe is again analyzedunder all functional, environmental, and accidental loadingcombinations deriving from the internal and external environ-ment. For static design, functional, environmental, and acci-dental loads are deÞned by Table 5 of API SpeciÞcation 17Jand Table 6 of API SpeciÞcation 17K.

5.5.4.2 All dynamic operational and accidental load cases,typically combining static internal with dynamic externalenvironmental conditions (e.g., wave, current, and riser topmotions) are considered as part of the dynamic analysis. Suf-Þcient load cases should be analyzed to cover the completeenvelope of response, in terms of motions and forces. Sensi-

Table 8—Typical Static Global Analysis Load Cases—Operating Conditions

Load Case Description Application1

A Global static analysis at design pressure, operating internal ßuid, mean vessel offset, no current.

DA

B Global static analysis at design pressure, operating internal ßuid, 100 year return inline near current, 100 year near vessel offset.

DA

C Global static analysis at design pressure, operating internal ßuid, 100 year return far current, 100 year far vessel offset.

DA

D Global static analysis at design pressure, operating internal ßuid, 100 year return cross current, 100 year cross vessel offset.

DA

E Thermal analysis. SA, DA

F On-bottom stability analysis. SA

G Upheaval buckling analysis. SA

Note:1. SA = Static Application, DA = Dynamic Application.


tivity studies should be performed to evaluate the effect ofvariations in critical parameters, including internal ßuid,marine growth, wave periods, VIV effects, etc. The load casematrix will depend largely on site speciÞc conditions.

5.5.4.3 A set of artiÞcial but representative tables, describedbelow, illustrate elements of the recommended approach.

5.5.4.3.1 An example sub-set of load cases for an FPSO/FPS application is illustrated in Table 9. Each of the deÞnedload cases would be analyzed for different combinations ofenvironment conditions. A typical example of a global

dynamic analysis load matrix is presented in Table 10 for aset of Òfunctional and environmentalÓ operational load cases.

5.5.4.3.2 Table 10 shows the use of regular wave analyses.Consideration may be given to also using irregular sea analy-ses for complete design or design veriÞcation. Generally, ves-sel offset data is given as maximum values. If signiÞcantvalues are available, then these may be used for regular waveanalyses. Maximum values should be used for irregular seaanalyses. See 8.4.1 for guidance on analysis types, i.e., designwave (regular wave) or design storm (irregular sea) loading.

Table 9—Example of Dynamic Load Cases for FPSO/FPS Applications

Load Case Load Condition Load Type Stress1 Criterion

MBR2 Criterion Description

A Normal Operation Functional and Environment 0.55Pressure Armor

0.67Tensile Armor

1.5 Operating internal ßuid conditions, intact mooring system and 100 year environmental conditions.

B Normal Operation Functional, Environment and Accidental

0.85 1.25 No internal ßuid, one mooring line broken and 100 year environmental conditions.

C Abnormal Operation Functional, Environment and Accidental

0.85 1.25 No internal ßuid, two mooring lines broken and 10 year environmental conditions.

1. The stress criterion is permissible utilization as a function of structural capacity.

2. The MBR criterion is a factor of safety on storage MBR.

Note: Regulatory or contractual requirements should deÞne actual ÒnormalÓ or ÒabnormalÓ operations

Table 10—Example of a Dynamic Load Case Matrix—Normal Operation—Functional and Environmental Loads

Parameter

Load Case Matrix

Near Near Far Far Cross Cross

Water Depth Min. MWL Min. MWL Max. MWL Max. MWL Max. MWL Max. MWL

Internal Pressure Operating Operating Operating Operating Operating Operating

Vessel Draft Loaded Loaded Ballasted Ballasted Ballasted Ballasted

Vessel Offset Near Intact Near Intact Far Intact Far Intact Cross Intact Cross Intact

Current Near10 Year

Near10 Year

Far10 Year

Far10 Year

Cross10 Year

Cross10 Year

Regular Wave Height

Near100 Year

Near100 Year

Far100 Year

Far100 Year

Cross100 Year

Cross100 Year

Ass. Regular Wave Period

Minimum Maximum Minimum Maximum Minimum Maximum

Notes:1. Vessel offset includes installation tolerances. Intact refers to the mooring system condition.2. Near case has the environment and offset orientated along the plane of the riser towards the riser seabed connection.3. Far case has the environment and offset orientated along the plane of the riser away from the riser seabed connection.4. Cross case has the environment and offset orientated perpendicular to the plane of the riser.5. Appropriate vessel motions to be included in the load cases.6. Similar matrices should also be prepared for the load cases B and C in Table 9.7. The minimum/maximum wave periods should represent one standard deviation about the mean value.


5.5.4.3.3 A set of load cases should be performed to evalu-ate potential interference between different system compo-nents. Guidance on the issue of interference is shown in 7.4.2and API RP2RD. The load cases should include normal oper-ation (1-year and 100-year conditions) with relevant acciden-tal loading conditions.

5.5.5 Detail and Service Life Design

5.5.5.1 For dynamic applications, the Þnal local analysisload cases are checked at this stage of the design using loadswhich have been derived from previous global dynamic anal-yses. Local analyses should be performed for all critical loca-tions in the pipe, considering loads calculated in the globalanalyses for all relevant conditions during the life of the pipe(i.e., FAT, installation, and normal and abnormal operation).A typical example of the local analysis load cases relating tothis stage of design are as follows:

Case A Design pressure, maximum top tension from 10-year design storm, pipe bent to operational MBR.

Case B No internal pressure, maximum top tension from100-year design wave, pipe bent to operationalMBR.

Case C Design minimum pressure, maximum axial com-pression.

5.5.5.2 Service life calculations to be performed relate tothe polymer degradation, to the corrosion of metallic layersand to fatigue analysis (see Sections 6 and 8). Unless thestresses in the pressure and tensile armors (unbonded) andreinforcing cables (bonded) are below the endurance limit forall load cases, a fatigue analysis will be required. For thefatigue analysis, the pipe is analyzed under all fatigue loadingcombinations speciÞed in the Design Premise. The combina-tions derived from the internal environment and the fatigue(typically seastate) components of the external environment.

5.5.5.3 The number of seastates analyzed should be shownto be conservative. The selected seastates should represent thewave scatter diagram for the location. The wave scatter dia-gram is generally divided into a minimum of Þve blocks, withthe maximum seastate from each block being used also it maybe necessary to perform the analyses for a number of direc-tions, e.g., near, far, and cross loading.

5.5.6 Installation Design

5.5.6.1 In this stage of the design process, the ßexible pipeis analyzed to check the feasibility of the proposed installa-tion method. The load cases should account for all relevantfunctional, environmental, and accidental loads as applicableto the installation method, vessel, season, test pressure, etc.Table 11 shows a typical set of installation load cases.

5.5.6.2 For riser systems, the load cases should cover allphases in the installation process, e.g., for a wave conÞgura-tion this could include analyses of the initial bare riser sec-

tion, after buoyancy modules paid out, and during Þnalconnection. The installation internal ßuid conditions shouldbe in agreement with the purchaser and deÞned in the DesignPremise. Consideration may be given to ßushing the lineswith seawater for normal or extreme environment installationconditions, if the material of the innermost layer is suitable.

5.5.6.3 Based on the results of the global analyses, a criti-cal set of local installation load cases should be selected.Table 12 shows an example set of local load cases. The resultsof these analyses should be compared with the design criteriaspeciÞed in Table 6 and Section 5.3.1 of API SpeciÞcation17J and Table 7 and Section 5.3.1 of API SpeciÞcation 17Kfor installation conditions. Additional criteria given in 5.4.1.8and 5.4.2 of this recommended practice for crushing collapseand ovalisation should also be checked.

Table 11—Example Global Analysis Load Cases for Installation Conditions

Load Cases Description

A Static analysis, Þeld hydrotest pressure.

B Static analysis, installation internal ßuid conditions, maximum installation current, equivalent vessel offset.

C Dynamic analysis, installation internal ßuid conditions, maximum installation current and design wave, equiva-lent vessel offset.

D Dynamic analysis, hydrotest pressure, maximum cur-rent and design wave at hydrotest conditions, equiva-lent vessel offset.

E Static analysis, post-installation plough operation.

Notes:1. Load cases B, C, and D would typically be performed for a num-ber of loading directions, such as 0¡, 45¡, 90¡, 135¡, and 180¡.

Table 12—Example Local Analysis Load Cases for Installation Condition

Load Cases Description

A Field hydrotest pressure, maximum top tension at hydrotest conditions.

B Installation internal ßuid conditions, maximum instal-lation top tension, installation MBR.

C Maximum top tension, maximum radial compression over chute or at tensioners.

D Maximum top tension, minimum radial compression from tensioners.

Notes:Load Cases C and D are used to check two critical load conditions for vertical installation with tensioners. Load Case C checks for potential collapse of the carcass and Case D checks for slippage of the pipe due to insufÞcient friction between the outer sheath and the outer tensile armor layer. (unbonded pipe only).


6 Materials

6.1 SCOPE

This section provides support for the material requirementsspeciÞed in Section 6 of API SpeciÞcations 17J/17K andgives general guidance on material selection for ßexible pipeapplications. Commonly used ßexible pipe materials areidentiÞed and their performance characteristics are given.Alternative materials, including composites, are discussed.Recommendations are given for ßuid compatibility and age-ing resistance testing of polymer/elastomer and metallicmaterials.

Due to the complexity of the applications for ßexible pipes,the guidelines in this section should only be used as a basisfor discussions between the purchaser and the manufacturerfor each speciÞc application. These discussions should alsobe based on the requirements in Section 6 of API SpeciÞca-tions 17J/17K, which deÞnes detailed requirements for thequaliÞcation and use of polymer/elastomer materials in ßexi-ble pipe applications. See API SpeciÞcation 17J, Tables 9 and10 and API SpeciÞcation 17K, Tables 11 and 12, which listminimum property requirements for the materials.

6.2 MATERIALS—UNBONDED PIPE

6.2.1 General

6.2.1.1 This section identiÞes the commonly used materi-als in the ßexible pipe industry and presents in general termsthe performance characteristics of these materials, such asallowable temperature ranges and ßuid compatibility.

6.2.1.2 For speciÞc applications, the characteristics identi-Þed for the various materials may not be appropriate, as thesuitability of a particular material is based several factorsincluding transported ßuid components, temperature, pres-sure, and parameter variations over the service life (see Sec-tion 4 and Appendix A (Purchasing Guidelines) of APISpeciÞcations 17J for a detailed listing of relevant parame-ters). The purchaser should therefore specify to the manufac-turer the design and operating values of all relevantparameters, including variations over the service life, withreference to API SpeciÞcation 17J requirements.

6.2.1.3 The materials and their properties should bereviewed against potential failure modes so as to identify thecritical requirements of the materials in each layer of the pipe.A detailed list of potential failure modes is given in Section13 of this recommended practice.

6.2.2 Polymer Materials

Table 13 lists the polymer materials typically used in ßexi-ble pipes. The properties for PA-12 are initially largely simi-lar to PA-11; though their ageing process, however, is verydifferent. For higher temperature or dynamic applications,

PA-11 may be more suitable than HDPE for the outer sheathbecause of better abrasion and fatigue characteristics.

XLPE is a special grade of PE, which is achieved by acrosslinking process so as to improve the base material char-acteristics. The crosslinking is generally achieved by circulat-ing hot water after the extrusion process.

The properties of PVDF partially depend on the polymer-ization process. The two processes used currently for themanufacture of PVDF for the ßexible pipe industry are theemulsion and suspension processes. A critical issue with theuse of PVDF is sealing of the layer in the end Þtting. See5.2.3 for guidelines on this issue.

Typical properties (operating temperature range, ßuid com-patibility and blistering characteristics) for the main polymersheath materials (HDPE, XLPE, PA-11, and PVDF) arefound in 6.2.2.1 through 6.2.2.3. Note that for many applica-tions the polymer material properties/characteristics are inter-dependent, e.g., the allowable temperature range may be afunction of the transported ßuid or the blistering characteris-tics may be a function of temperature and pressure.

6.2.2.1 Temperature

6.2.2.1.1 Table 14 shows guidelines for selection of poly-mers for ßexible pipe applications based on a 20-year servicelife. For detailed engineering, a validated ageing model isrequired to conÞrm the polymer service life requirements (seeSection 6.2.3.4 of API SpeciÞcation 17J and Section 6.5.2 ofthis recommended practice).

6.2.2.1.2 Note that Table 14 shows only general limitsand may not apply for speciÞc applications. The tempera-ture ranges for each of the materials also depend on thecomponents of the conveyed ßuids. For example, the maxi-mum temperature for PA-11 will be signiÞcantly lowerwith water cuts. Also, higher operating temperatures maybe feasible for many polymers when the required designlife is shorter than 20 years, because higher temperaturestypically accelerate ageing. This point is not valid for allpolymer materials, and the ageing characteristics should be

Table 13—Typical Polymer Materials for Flexible Pipe Applications

Layer Material Type

Internal pressure sheath HDPE, XLPE, PA-11, PA-12, PVDF

Intermediate sheaths HDPE, XLPE, PA-11, PA-12, PVDF

Outer sheath HDPE, PA-11, PA-12

Insulation PP, PVC, PU

Note:1. The insulation may be solid material, foam, or syntactic foam.2. MDPE may be used instead of HDPE.


based on test data. Temperature excursions above the maxi-mum stated values may also be acceptable for relativelyshort durations with supplier acceptance.

6.2.2.2 Fluid Compatibility

Table 15 lists typical ßuid compatibility characteristics forßexible pipe polymer materials. Note that ßuid compatibilityis highly dependent on temperature.

6.2.2.3 Gas Exposure

6.2.2.3.1 Gas in the transported ßuid is an importantconsideration in material selection for the polymer layers.The main issues relate to blistering resistance and perme-ability of the material of the internal pressure sheath; per-meability characteristics of the outer sheath, however, willalso be required. Table 15 lists typical blistering resistancecharacteristics for the internal pressure sheath polymermaterials.

6.2.2.3.2 The gas permeation rate depends on many factors(see 8.2.2). The main issues to be considered in relation to gaspermeation are the transported ßuid components to be evalu-ated (the main components being CH4, CO2, H2S, and watervapor), their effect on the steel layers in the annulus (see 6.6),and the gas venting system capacity.

6.2.3 Metallic Materials

Property requirements for metallic materials are listed inTables 10 and 12 of API SpeciÞcation 17J. These propertiesshould be compared with the requirements of each applica-tion, with reference to the critical failure modes identiÞed inSection 13.3.

6.2.3.1 Carcass

6.2.3.1.1 Materials typically used for the carcass layer areas follows:

a. Carbon steel.

b. Ferritic stainless steel (AISIs 409 and 430).

c. Austenitic stainless steel (AISIs 304, 304L, 316, 316L).

d. High-alloyed stainless steel (e.g., Duplex UNS S31803).

e. Nickel based alloys (e.g., N08825).

6.2.3.1.2 Material selection for the carcass is based on theinternal ßuid components and expected use of the ßexiblepipe. Important parameters that should be considered areidentiÞed in Section 4.4.4 of API SpeciÞcation 17J.

6.2.3.1.3 As the severity of the internal ßuid environmentincreases, the material selected for the carcass will move from(a) to (e), i.e., carbon steel will be used for non-corrosiveenvironments while high-alloyed stainless steels will be usedfor corrosive applications. The most commonly used materi-als are 304L and 316L austenitic stainless steel. A highmolybdenum content (2.7 to 3.0 percent) may be speciÞed forAISI 316L material to improve its corrosion resistance char-acteristics.

6.2.3.1.4 The main parameters to be considered in thematerial selection for the carcass are ßuid temperature, CO2,H2S, chloride, and oxygen content. Other parameters thatshould be considered include pH, water, free sulphur andmercury content of internal ßuid. In sour service environ-ments, the carcass material should be resistant to HIC andSSC with reference to NACE MR01-75, as applicable.

6.2.3.1.5 If the transported ßuid is oxygenated (aer-ated), e.g., seawater injection, and a carcass is required,

Table 14—Temperature Limits for Thermoplastic Polymers in Flexible Pipe Internal Pressure Sheath Applications Based on 20-Year Service Life

Polymer Material

Minimum Exposure

Temperature (¡C)

Maximum Continuous Operating

Temperature (¡C)

Water Cut Limits1 Comments

HDPE Ð50 +60 0Ð100 percent High tensile and impact resistance at low temperature.

XLPE Ð50 +90 0Ð100 percent May be used for high water cut applications [8]. Maximum temperature is a function of operating pressure, with a reduction in temperature for pressures above 13.8 MPa (2000 psi).

PA-11 Ð20Ð20

+90+65

0 percent0Ð100 percent

See Section 6.5.2 for further details on the effect of water cut on service life.

PVDF Ð20 +130 0Ð100 percent


consideration may be given to using non-metallic material(e.g., polymers, composites) for the carcass. However,this unproven technology would need to be validated bytesting.

6.2.3.1.6 It is important that the hydrotest ßuid is benign tothe carcass material. As a minimum for carbon steel car-casses, dissolved oxygen should be removed from thehydrotest water, even for potable waters. In addition, consid-eration may need to be given to the use of biocide and, forparticularly aggressive cases, corrosion inhibitor.

6.2.3.2 Pressure and Tensile Armor Layers

6.2.3.2.1 For the pressure and tensile armor layers, the typ-ical material used is carbon steel, with carbon content depen-dent on the design requirements. High carbon content steel isused where the design requires very high strength and where

the environment permits. Low or medium carbon contentsteels are used for sour service environments. Not all wires,however, meet NACE MR01-75 sour service requirements.For sour service environments, the steels may also be heattreated, e.g., quenched and tempered.

6.2.3.2.2 Chemical composition of the steel material forboth the pressure and tensile armors should be reviewed toconÞrm suitability for the speciÞed application. Other impor-tant issues are manufacturability, weldability, sour servicerequirements, conformance to speciÞed structural capacity,and compliance with API SpeciÞcation 17J requirements.Important components to be speciÞed and controlled includecarbon, manganese, phosphorus, sulphur, silicon, and copper.The manufacturerÕs material speciÞcations should deÞne con-tent limits for these components and distinguish betweensweet and sour service applications. For some applications,

Table 15—Typical Fluid Compatibility and Blistering Characteristics for Flexible Thermoplastic Pipe Polymer Materials

Polymer Material General Compatibility Characteristics Blistering Characteristics1

HDPE Good ageing behaviour and resistance to acids, seawater and oil. Good blistering resistance at low tempera-tures and pressures only.

Weak resistance to amines and sensitive to oxidation.

Susceptible to environmental stress cracking (environments include alcohols and liquid hydrocarbons).

XLPE Good ageing behaviour and resistance to seawater, weak acids (dependent on concentrations and dosage frequency) and production ßuid with high water cuts.

Better blistering resistance than HDPE, with positive results obtained in excess of 3000 psi.

Weak resistance to amines and strong acids (dependent on concentrations and dosage frequency) and sensitive to oxidation. Less susceptible to environmental stress cracking than HDPE (environments include alcohols and liquid hydrocar-bons).

PA-11 Good ageing behaviour and resistance to crude oil. Good blistering resistance up to 7500 psi and 100¡C.

Good resistance to environmental stress cracking.

Limited resistance to acids at high temperatures (recommend pH > 4.5 or TAN < 4.0). Limited resistance to bromides.

Weak resistance to high temperatures when any liquid water is present.

PVDF High resistance to ageing and environmental stress cracking. Good blistering resistance up to 7500 psi and 130¡C.

Compatible with most produced or injected well ßuids at high temperatures including alcohols, acids, chloride solvents, aliphatic and aromatic hydrocarbons and crude oil.

Weak resistance to strong amines, concentrated sulfuric and nitric acids and sodium hydroxide (recommend pH < 8.5)

Notes:1. Blistering characteristics are taken from [9]. Note that blistering characteristics will be a function of transported ßuid, pressure, depressurization rate, and temperature.2. The suitability of a material for a particular application should be veriÞed by the manufacturer.


consideration should also be given to minimizing the manga-nese content and performing calcium treatment of the melt.

6.2.3.2.3 Wire weldability should be veriÞed by conduct-ing tests with deÞned and documented acceptance criteria.For evaluation of material weldability, the maximum carbonequivalent (CE) content should be speciÞed when no postweld heat treatments (PWHT) are performed. CE may bedeÞned by formulas similar to the following:

(1)

6.2.4 End Fittings

6.2.4.1 The materials typically used for the primarymetallic end Þtting components are AISI 4130 steel oralloyed stainless steel (e.g., Duplex, 6Mo). The corrosionresistant coatings typically used for the end Þttings includethe following:

a. Electroless nickel plating, thickness at least 75 µm.

b. Inconel 625 inlay, thickness at least 3 mm.

c. Epoxy coating systems.

d. Fluoropolymer coatings.

6.2.4.2 The material and corrosion coating selection for theend Þtting is a function of the application, in particular, theinternal and external environmental conditions. End Þttingmaterials and coatings should meet the requirements of Sec-tions 6.1.4 and 6.2.5 of API SpeciÞcation 17J.

6.3 MATERIALS—BONDED PIPE

6.3.1 General

6.3.1.1 This section identiÞes the commonly used materi-als in the bonded ßexible pipe industry and presents, in gen-eral terms, the performance characteristics of these materials,such as allowable temperature range and ßuid permeability.Note that the elastomer materials are identiÞed by their pri-mary elastomeric component, for example, nitrile butadienerubber (NBR). While the primary component is given, therecipe or mix used is, in general, speciÞc to each companyand not usually released to second parties.

6.3.1.2 For speciÞc applications, the characteristics identi-Þed for the various materials may not be appropriate as thesuitability of a particular material is dependent on a largenumber of factors including transported ßuid components,temperature, pressure, compound mix and parameter varia-tions over the service life (refer to Section 4 and Appendix A[Purchasing Guidelines] of API SpeciÞcation 17K for adetailed listing of relevant parameters). The purchaser shouldtherefore specify to the manufacturer the design and operat-ing values of all relevant parameters including variations over

the service life, with reference to the requirements of APISpeciÞcation 17K.

6.3.1.3 The materials and their properties should bereviewed against potential failure modes so as to identify thecritical requirements of the materials in each layer of the pipe.A detailed list of failure modes is given in Section 13 of thisrecommended practice.

6.3.2 Elastomer Materials

Table 16 lists the elastomer materials typically used in ßex-ible pipes. These elastomer materials constitute approxi-mately 40Ð65 percent of the Þnal compound mix, with carbonblack, antioxidants, activators, plasticizers and curing agentsmaking up the remainder amongst other ingredients. The Þnalproperties of the rubber compound are dependent on the Þnalmix of all ingredients. For example, the higher the carbonblack content in a compound mix the lower the electricalresistance will be in addition to a generally higher tensilestrength (although the structure and size of the carbon blackparticles play a signiÞcant role also). NBR is extensively usedas a liner material because of its low permeability to gas, suchas N2 and O2. However, NBR is itself dependent on the per-centage of acrylonitrite in the elastomer. This is usually 17 to50%. The higher the acrylonitrite content in the NBR, thehigher the heat and oil resistance and the lower the elasticityof the material at low temperature.

CPE is a typical elastomer used for bonded pipe covers. Itscharacteristics make it suitable for a relatively high abrasiveenvironment where it can be exposed to both seawater andozone.

Typical properties (operating temperature range, ßuid com-patibility, and ßuid permeability) for the main elastomermaterials are found in 6.3.2.1 through 6.3.2.3 Note that aseach rubber compound material is made up of an elastomermaterial and several other materials the properties will there-fore vary with mix type. In addition, for most applications theelastomer material properties/characteristics are interdepen-dent. For example, the allowable maximum operating temper-ature may be a function of the transported ßuid.

An API Technical Bulletin, ÒEvaluation Standard for Inter-nal Pressure Sheath Polymers for High Temperature FlexiblePipesÓ has been developed by a Joint Industry Project. Thedocument describes development of test plans to evaluate the

CE CMn6

--------Cr Mo V+ +

5-------------------------------

Ni Cu+15

------------------- + + +=

Table 16—Typical Elastomer Materials For Bonded Flexible Pipe Applications

Application Material

Liner NBR, HNBR, CR, NR, EPDM

Cover CR, CPE

Filler Various

Insulation PVC, PE, closed cell foam, glass Þber


suitability of candidate polymers for high temperature ser-vice. Also deÞned are a set of evaluation criteria for materialqualiÞcation.

6.3.2.1 Temperature

6.3.2.1.1 Table 17 gives guidelines for selection of elas-tomers for bonded ßexible pipe applications. These guide-lines consider a relatively benign transported ßuid.

6.3.2.1.2 For detailed engineering a validated ageingmodel is required to conÞrm the elastomer service liferequirements, see Section 6.2.3.4 of API SpeciÞcation 17Kand Section 6.5.2 of this recommended practice.

6.3.2.1.3 Note that Table 17 shows only general limits andmay not apply for speciÞc applications. The temperatureranges for each of the materials also depend on the compo-nents of the conveyed ßuid. For example, the maximum oper-ating temperature for NBR may be reduced by as much as20¡C should the transported ßuid contain a relatively largepercentage of aromatics. Temperature excursions above themaximum stated values may also be acceptable for relativelyshort durations with supplier acceptance.

6.3.2.2 Fluid Compatibility

Table 17 lists typical ßuid compatibility characteristics forbonded ßexible pipe elastomer materials. Note that ßuid com-patibility is highly dependent on temperature.

6.3.2.3 Gas Exposure

6.3.2.3.1 If bonded ßexible pipe is to be used for service inwhich the transported ßuid contains gas then this will have tobe taken into consideration in material selection for the elas-tomer layers. The main issues relate to the blistering resistanceand continuing curing of both the pipe liner and the remainderof the pipe body. In general, elastomer materials are more sus-ceptible to blistering than thermoplastic materials used in

unbonded pipe applications. This is attributed to the relativelyhigher permeability to gas and lower tearing resistance of elas-tomer materials over thermoplastic materials. A mitigatingfactor used by industry is that the bonded ßexible pipe, madeup of elastomeric materials, is supported by an internal steelstripwound carcass and so the liner is not quite as susceptibleto blistering as small scale test results on elastomer alonewould suggest. Since sulfur is a cross-linking agent for manyelastomers, bonded pipe bodies exposed to H2S may experi-ence continuing curing in Þeld applications. This can result inreduced local ßexibility and increased global riser stiffness

6.3.2.3.2 The gas permeation rate through the elastomermaterial is dependent on many factors including internal andexternal pressure, surface area, liner thickness and permeabil-ity coefÞcient. The main issues to be considered in relation togas permeation are the propensity for blistering to occurunder rapid decompression, the likelihood of transported ßuidcomponents to permeate through the body of the pipe andtheir effect on the elastomer and steel reinforcing layers.

6.3.3 Metallic Materials

Property requirement for metallic materials are listed inTables 11 and 13 of API SpeciÞcation 17K. These propertiesshould be compared with the requirements of each applica-tion, with reference to the critical failure modes identiÞed inSection 13.3.

6.3.3.1 Carcass

6.3.3.1.1 Materials typically used for the carcass layer areas follows:

a. Carbon steel.

b. Ferritic stainless steel (AISI 409 and 430).

c. Austenitic stainless steel (AISI 304, 304L, 316, 316L).

d. High alloyed stainless steel (e.g., Duplex UNS S31803).

e. Nickel based alloys (e.g., N08825).

Table 17—Temperature limits for Thermosetting Elastomers in a Bonded Flexible Pipe Liner Application

Brittleness Temperature

(¡C)

Maximum continuous operating

temperature (¡C) Comments

NBR Ð20¡ to 40¡ 125¡ Properties dependent on acrylnitrite content. Excellent resistance to hydrocarbons. Very good tensile strength and dynamic properties. Good impermeability and heat resistance. Poor resistance to weather and ozone.

HNBR Ð40¡ to 50¡ 150¡ Good resistance to hydrocarbon. Very good tensile strength and dynamic properties. Good impermeability and very good resistance to weather and ozone.

CR Ð30¡ to 40¡ 100¡ Reasonable resistance to hydrocarbons. Good tensile strength and reasonable dynamic properties. Good impermeability and heat resistance. Very good resistance to weather and ozone.


6.3.3.1.2 The selection of the material for the carcass isdependent on the internal ßuid components and expecteduse of the ßexible pipe. Important parameters that should beconsidered are identiÞed in Section 4.4.4 of API SpeciÞca-tion 17K.

6.3.3.1.3 As the severity of the internal ßuid environmentincreases, the material selected for the carcass will move from(a) to (e), i.e. carbon steel will be used for non-corrosive envi-ronments while high alloyed stainless steels will be used forcorrosive applications. The most commonly used materialsare 304L and 316L austenitic stainless steel. A high molybde-num content (2.7%Ð3.0%) may be speciÞed for AISI 316Lmaterial to improve its corrosion resistance characteristics.

6.3.3.1.4 The main parameters to be considered in theselection of the carcass material are ßuid temperature, andCO2, H2S, chloride and oxygen content. Other parametersthat should be considered include pH, water, free sulphur, andmercury content of internal ßuid. In sour service environ-ments the carcass material should meet the requirements ofNACE MR 01-75.

6.3.3.1.5 If the transported ßuid is oxygenated (aerated),e.g., seawater injection, and a carcass is required, consider-ation may be given to using non-metallic material (e.g., poly-mers, composites) for the carcass. However, this is unproventechnology and would need to be validated by testing.

6.3.3.1.6 It is important that the hydrotest ßuid is benignto the carcass material. As a minimum for carbon steel car-casses, dissolved oxygen should be removed from thehydrotest water, even for potable waters. In addition, consid-eration may need to be given to the use of biocide and for par-ticularly aggressive cases, corrosion inhibitor.

6.3.3.2 Reinforcing Layers

6.3.3.2.1 For the cables of the primary reinforcing layers,the typical material used is carbon steel. High carbon contentsteel is used to give a high strength cable.

6.3.3.2.2 Chemical composition of the steel material forthe reinforcing layers should be reviewed to conÞrm suitabil-ity for the speciÞed environment. Other important issues aresour service requirements, conformance to speciÞed struc-tural capacity and compliance to API SpeciÞcation 17Krequirements. In determining suitability, the effect of theenclosing rubber should be considered.

6.3.3.2.3 Important components to be speciÞed and con-trolled included carbon, manganese, phosphorus, sulphur, sil-icon, and copper. The manufacturersÕ material speciÞcationsshould deÞne content limits for these components. For someapplications, consideration should also be given to minimiz-ing the manganese content and performing calcium treatmentof the melt.

6.3.4 End Fittings

6.3.4.1 The materials typically used for the primarymetallic end Þtting components are AISI 4130 steel oralloyed stainless steel (e.g., duplex, 6Mo). The corrosionresistant coatings typically used for the end Þttings includethe following:

a. Electroless nickel plating, thickness at least 75 µm.

b. Inconel 625 inlay, thickness at least 3 mm.

c. Epoxy coating system.

d. Fluoropolymer coatings.

e. Zinc Coating.

6.3.4.2 The material and corrosion coating selection for theend Þtting is a function of the application, in particular, theinternal and external environmental conditions. End Þttingmaterials and coatings should meet the requirements of Sec-tions 6.1.4 and 6.2.5 of API SpeciÞcation 17K.

6.4 ALTERNATIVE MATERIALS

6.4.1 Aluminum

6.4.1.1 Aluminum material may be used to replace steel inany of the structural layers of the ßexible pipe, including car-cass, pressure armor and tensile armor, layers. AluminumÕsmain advantage is that, compared to steel, it gives a weightsaving of between 30 and 60 percent for the same strengthcharacteristics.

6.4.1.2 Careful evaluation of aluminumÕs corrosionbehaviour is required prior to its use for ßexible pipe appli-cations. Other important issues to be addressed includeabrasion/wear resistance, SSC and HIC resistance, fatigueand welding.

6.4.2 Composite Materials

6.4.2.1 Composites are materials in which a reinforcingÞbre is combined in a resin matrix and cured. For ßexiblepipes, composite materials are currently only used for thereplacement of carbon steel in the tensile armor layers. Con-sequently, this section considers only this particular use ofcomposites in ßexible pipe applications.

6.4.2.2 The steel tensile armor wires used in unbondedßexibles are typically 3 mm to 6 mm thick and are mechani-cally preformed to a helical structure. The composite armorwires may be 1 mm to 2 mm thick and helically wound inseveral layers per equivalent steel layer. Alternatively, theymay be the same thickness as the equivalent steel armor layer(up to 8 mm).

6.4.2.3 For the tensile armor wires, composites offer arange of beneÞcial properties when compared to steel, includ-ing the following:


a. High strength-to-weight ratio.

b. Good fatigue resistance (not notch sensitive).

c. Good impact resistance and toughness (materialdependent).

d. Immunity to corrosion and degradation by most oil Þeldchemicals and seawater.

e. High stiffness or modulus (in one direction).

Note: These characteristics are highly dependent on the compositeresin and reinforcing Þbers.

6.4.2.4 The main potential for use of composite-based ten-sile armor ßexible pipes is in deep water applications, wherethe weight reduction can be signiÞcant compared to steelbased tensile armor pipes (density of composites approxi-mately 25 percent of steel). In addition, there is potential foruse of composites in high pressure, sour service applications.Service life determination is an evolving technology for com-posites and currently limits their application.

6.4.2.5 The reinforcing Þbers used in composites includeE-glass, carbon and aramide Þbers. The glass Þbre compositeis more economical than the carbon Þbre material. The car-bon Þbre material, however, has more favorable strengthproperties and characteristics. For both glass and carbon Þbrecomposites, the reinforcing Þbers are orientated parallel tothe wire longitudinal axis. The matrix materials used includeepoxy and vinyl ester resins, and thermoplastic polymers.

6.4.2.6 Some of the main considerations when using com-posites are as follows:

a. Potential wear problems between armor layers andbetween individual armor wires, which are subject to relativemotion and high contact pressure, should be addressed.

b. Inßuence of defects on composite wire performanceshould be assessed. Failure mechanisms need to be identiÞedand assessed.

c. Effective anchoring of the composites in the pipe end Þt-ting should be conÞrmed with suitable tests. Join-upprocedures for the individual composite wires should be care-fully evaluated.

d. Experiments should be performed to characterize theeffects of permeated ßuids upon Þbre-matrix interfaces in thecomposites. The susceptibility of glass Þbre composites tostress corrosion cracking in seawater should be investigated.The potential for galvanic corrosion in carbon Þbre compos-ites should be determined. The use of glass Þbre compositesin water at high temperatures is limited and should be veriÞedby testing [10].

e. The structure of the composite after being subject to rele-vant loads and environmental conditions should bedetermined by scanning electron microscopy (SEM), whichcan be used to determine microcracking and delamination.

f. Normally, composite wires are preformed during wire fab-rication rather than during winding on to the pipe. Thisprocess may induce reduction of performance properties(e.g., σy) compared to the non-formed wire properties, andshould be checked by testing. If the composite wire is not pre-formed, bending stresses are induced when the material iswound on to the pipe. Because of these additional bendingstresses, the reduction in performance should be evaluated byanalysis and testing.

6.4.2.7 Composite materials should be qualiÞed in the Þnalprocessed state, under test conditions representative of theactual operational conditions. The manufacturer and pur-chaser should agree on the test procedures, with reference toapplicable international standards. The following properties/characteristics should be determined for composite materialsin ßexible pipe applications:

a. Tensile strength/elongation.

b. Modulus of elasticity.

c. Density.

d. Fatigue properties, including endurance limit (tensile, ßex-ural, and fretting fatigue).

e. Creep characteristics.

f. Fracture resistance.

g. Ageing characteristics (reduction of material propertieswith time).

h. Microbial (bacterial) degradation.

i. PoissonÕs ratio.

j. Wear/abrasion resistance.

k. Chemical resistance (to corrosion inhibitors, etc.).

6.4.3 Aramide Fibers

6.4.3.1 A potential alternative material for ßexible pipes issynthetic Þbers, such as aramide. These Þbers could be usedto replace the steel armor layers, giving signiÞcant weightreduction and potentially improved performance in sour ser-vice applications. In addition, aramide Þbers have the follow-ing positive characteristics for ßexible pipe applications:

a. No corrosion.

b. Good chemical resistance to most production ßuids.

c. Good fatigue properties.

d. Good creep properties.

e. Low temperature sensitivity.

6.4.3.2 Areas of concern for the use of aramide Þbersinclude the following:

a. Time/temperature dependency of mechanical properties.

b. Termination in the end Þttings.

c. Ageing characteristics (UV sensitivity).


d. Non-isotropic behaviour.

e. Static and dynamic bending ßexibility requirements.

f. Notch sensitivity.

g. Environmental stress cracking resistance.

6.5 POLYMER/ELASTOMER TEST PROCEDURES

Section 6 of API SpeciÞcation 17J/17K speciÞes materialproperty requirements and test procedures. As standard pro-cedures are unavailable for polymer/elastomer elastomer ßuidcompatibility and ageing resistance tests, procedures are notgiven in API SpeciÞcation 17J/17K. This section gives guide-lines and recommendations for performing these tests.

6.5.1 Fluid Compatibility

6.5.1.1 Section 6.2.3.3 of API SpeciÞcations 17J/17Kgives general requirements for the performance of ßuid com-patibility tests and identiÞes critical parameters for evaluatingcompatibility. This section gives recommendations on the testprocedures.

6.5.1.2 Laboratory tests with extruded samples of the poly-mer or calenderised or extruded samples of the elastomer canbe used to determine gross incompatibility. Tests should bebased on the design conditions, subject to the following rec-ommendations:

a. Test ßuid - To contain components of design internal ßuidwhich possibly have adverse effects on the polymer, in partic-ular seawater, production ßuid, H2S, CO2 and injectionchemicals. Fluid pH level to be controlled to designconditions.

b. TemperatureÑMaximum operating temperature as aminimum.

c. PressureÑAmbient for liquids and design pressure orgreater for gases.

d. Stress ConditionsÑZero. If there is potential for stresscracking, also test at maximum design strain.

e. Exposure TimeÑMinimum 300 hours for acceleratedtests (increased temperature) or minimum 2,000 hours foroperating temperature.

f. SamplesÑSample thickness should be at least 3 mm.Sample length should be based on the test equipment. If testßuid is multi-phase, sample should be immersed in all phases.

g. ParametersÑCritical parameters and acceptance criteriashould be established based on the polymer/elastomer beingevaluated and the particular application. Tensile strength,elongation, visual appearance, and ßuid absorption (weightgain) and desorption (weight loss) parameters should be con-sidered for evaluation/ measurement.

6.5.1.3 Sulphur can be liberated from H2S reacting withsteel components or the elastomer compounds in the bonded

pipe to cause cross-linking and hardening. The effects ofreleased sulphur on either metallic or elastomeric compo-nents should be evaluated.

6.5.2 Ageing Test

6.5.2.1 Ageing of elastomer/polymer material is an irre-versible process, which occurs when the material is exposedto particular environmental conditions. Polymer/elastomerageing is dependent on the ßuid transported in ßexible pipes,temperature, pressure, and external conditions, such as UVradiation. The ageing process is characterized by change inproperties, such as reduction in strength or ductility, andembrittlement or softening. In addition, the physical proper-ties of the polymer/elastomer may be signiÞcantly altered bymigration of plasticizers.

6.5.2.2 Section 6.2.3.4 of API SpeciÞcations 17J/17Kgives general requirements for the performance of ageingtests and identiÞes critical parameters for the most commonlyused polymers. The objective in performing ageing tests is todevelop satisfactory ageing prediction and monitoring mod-els, which may include Arrhenius plots. This gives the mate-rial service life as a function of the inverse of temperature,plotted to a log-linear scale. Some materials (e.g., PA-11)have been found to be more amenable to the development ofArrhenius plots, than other materials (e.g., PVDF).

6.5.2.3 An Arrhenius plot deÞnes an exponential decaymechanism between temperature and exposure time, asfollows:

(2)

where tcrit is the critical exposure time at a given value oftemperature (T), and Ea and R are constants.

6.5.2.4 Prior to test start-up, the ageing criteria should beestablished for review by the purchaser. The ageing criteriashould be based on measurable performance properties at theend of the pipeÕs service life. Recommended properties at theend of service life for polymer materials, using uniaxial shorttime tensile tests to ASTM D638 at 20¡C, are as follows:

HDPE Ñ Tensile strength: Min. 15 MPa

XLPE Ñ Tensile strength: Min. 15 MPa

PA-11 Ñ Tensile strength: Min. 20 Mpa

Ñ Elongation at break: Min. 50 percent

PVDF Ñ Tensile strength: Min. 25 MPa

Ñ Elongation at yield Min. 7 percent

6.5.2.5 Aging of PA-11 or Rilsan in water, alcohols, hydro-carbons, acids, and combinations thereof is currently the sub-ject of study by a number of researchers (See Ref [52], [53],

tcrit A.eEaRT-------

=


and [54]). An API work group (The Rilsan Users Group)under Subcommittee 17 is coordinating activities related toaging of PA-11 in ßexible pipes. A plot of estimated PA-11lifetimes in hydrocarbon-water exposure is shown in Figure21, which is based on data similar to that presented in [12]and criteria given above and in [11]. The plot is based on datameasured at temperatures above 100¡C and extrapolateddown into the application temperature range. For PA-11 indynamic ßexible pipe applications the ßowing ßuid tempera-ture has been used for aging evaluations. In some static appli-cations where the pipe is not expected to be subjected tosigniÞcant alternating strains, the mean temperature of theinternal pressure sheath, based on the radial temperature dis-tribution has been used.

6.5.2.6 The aging process for PA-11 is strongly inßuencedby the water content and pH of the transported ßuid. Figure21 shows two lines for hydrocarbon saturated with water andunsaturated. Current experience indicates that aging in hydro-carbon/water exposure above saturation is very similar to theÒSaturatedÓ curve and therefore this curve can be used for allwater cut values that would saturate the hydrocarbon at theproduction conditions. Similarly, the ÒUnsaturatedÓ curvemay be used for all water cut values that do not reach satura-tion at the production conditions. It has been suggested thatthe transition may occur at 80% of saturation.

6.5.2.7 Where the transported ßuid temperature is constantover the service life of the pipe, then the design life can beread directly from Figure 21. For varying temperatures andwater cuts, the degradation over the total service life shouldbe calculated by an integration of the exposure periods at thedifferent temperatures and water cuts.

6.5.2.8 This Palmgren-Miner cumulative damage typeapproach is believed to give a conservative estimate of designlife.

6.5.3 Epoxy Shear Strength Test

6.5.3.1 The epoxy shear strength test is intended as analternative to the ASTM D695 compressive strength test indetermining the shear capacity of the epoxy resin used foranchoring the reinforcing cables in some bonded pipe endÞttings.

6.5.3.2 Section 6.2.5.3 of API SpeciÞcation 17K givesgeneral requirements for the performance of epoxy shearstrength tests. This section provides recommendations on thetest procedures.

6.5.3.3 The epoxy shear strength test involves testing curedepoxy samples by shearing the sample at different tempera-tures, thereby obtaining the temperature dependent shearcapacity of the material. Tests should be based on operatingconditions, subject to the following recommendations:

a. SampleÑsample size should be based on the test equip-ment. A minimum of three samples per temperature should betested.

b. TemperatureÑsample should be tested at both minimumand maximum operating temperature and at sufÞcient tem-perature intervals in between to satisfactorily deÞne the shearstrength/temperature relationship of the material.

c. Curing of samplesÑsamples should be molded and curedunder the same temperature and humidity condition as pre-vailing when Þlling the end-Þtting.

d. Sample preparationÑthe epoxy resin should be mixedaccording to the manufacturers speciÞcation and pouredslowly into the prepared mold to ensure no air bubbles areenclosed.

e. Quality controlÑif the shear strength test is required aspart of the pipe manufacture quality process then the epoxysamples should be taken from the batch used to Þll the endÞtting.

1000.0

100.0

10.0

1.040 50 60 70 80 90 100

Water saturated hydrocarbon

Unsaturated hydrocarbon

Estimated PA-11 Service Life inHydrocarbon-Water Exposure(extrapolated data at pH = 7)

Temperature (degrees C)

Tim

e (y

ears

)

Figure 21—PA-11 Service Life vs. Temperature [12]


6.6 METALLIC MATERIAL TEST REQUIREMENTS

6.6.1 General

6.6.1.1 This section discusses the qualiÞcation test require-ments for ßexible pipe metallic materials and gives recom-mendations on the performance of the tests and interpretationof results. For qualiÞcation of materials for the carcass, pres-sure armor (unbonded pipe), tensile armor layers (unbondedpipe) and reinforcing layers (bonded pipe), Table 12 in APISpeciÞcation 17J and Table 13 in API speciÞcation 17K speci-Þes test requirements. The following required tests do not havestandard (e.g., ASTM) test procedures for their performance:

a. SSC and HIC resistance.

b. Corrosion resistance.

c. Erosion resistance.

d. Fatigue resistance.

e. Hydrogen embrittlement resistance.

f. Chemical resistance.

6.6.1.2 These tests are discussed in the following sectionsin detail and also supplement API SpeciÞcations 17J/17Krequirements.

6.6.2 SSC and HIC Resistance

6.6.2.1 In wet H2S environments, hydrogen enters steelcomponents at the corroding surface. Depending on the type ofsteel, its microstructure and the inclusion distribution, thehydrogen may give rise to internal decohesion resulting in HICor brittle fracture, termed SSC. Section 6.2.4.2 of API SpeciÞ-cations 17J/17K speciÞes SSC and HIC test procedures forsteel wire cables materials used in ßexible pipe applications.

6.6.2.2 Two types of SSC tests are required by API SpeciÞ-cations 17J/17K:

a. Use of NACE TM 01-77 environment at constant pHbetween 3.5 and 3.8 to determine stress threshold levels forthe occurrence of SSC.

b. SSC test with actual service conditions, with the samplesstressed to 0.9 times the actual yield stress of the sample, asdeÞned in Section 6.2.4.2.3. of API SpeciÞcation 17J ordesign stress levels as deÞned in Section 6.2.4.2.3 of APISpeciÞcation 17K.

6.6.2.3 Results from both of these tests are used to deter-mine suitability of the steel material for the proposed applica-tion. Important considerations in the performance of thesetests include the following:

a. For both SSC tests described above, the recommended testprocedures are as follows:

1. For pressure armor wires of unbonded pipe (includinginterlock and backup ßat wires), ring tests should be usedwhere practical for pipe diameters less than 6-inches;

otherwise, four-point bend tests from ring samplesshould be used.

2. For tensile armor wires of unbonded pipe, dependingon the wire size, Method A of ASTM A370 or four-pointbend tests should be used.

3. For reinforcing layer cables a coating of embeddingcompound should be applied, the maximum thickness ofwhich should not be greater than the minimum designthickness of the embedding compound in the pipeconstruction.

b. SSC tests in the actual service conditions will probablyhighlight any susceptibility of the material to HIC and/orSOHIC (stress-oriented HIC), and therefore examination pro-cedures should check for both of these characteristics. Inaddition, API SpeciÞcations 17J/17K requires HIC to bechecked for in the NACE TM 01-77 SSC tests describedabove.

c. All samples should represent, as close as possible, the as-manufactured wires/cables and should be tested on a statisti-cal basis to verify resistance. Welded samples should betested to qualify welding procedures for wires used inunbonded pipe.

d. Test procedures should ensure that the important testparameters are kept largely constant, including stress/strainlevels, pH, temperature, and H2S partial pressure.

e. The material is considered to have failed the test if there isevidence of cracking from visual, microscopic, or magneticparticle inspection, other than surface blisters. See [13] forfurther guidance on acceptance criteria.

f. A 20¡C (±3¡C) test temperature is recommended, as this isconsidered the worst temperature for hydrogen effects.

g. Consideration should be given to using NACE TM 02-84test method to determine the HIC resistance of the steel wirematerials of unbonded ßexible pipe. This test is a muchshorter test than TM 01-77 and may be used as a quality con-trol test on the wire material.

6.6.2.4 The speciÞed tests apply to pipes for both static anddynamic applications. In addition, for dynamic applications,fatigue and corrosion fatigue tests will be required, as dis-cussed in Section 6.6.5.

6.6.3 Corrosion

6.6.3.1 This section addresses uniform or pitting corrosion.This is particularly relevant for unbonded pipe armor wirecorrosion. Corrosion problems in the carcass are generallyavoided by proper material selection, as discussed in 6.2.3.1.Though the pressure and tensile armors are not directly incontact with the transported ßuid, they will be exposed to per-meated ßuids, such as CO2 and H2S gas, and seawater if thereis a breach in the integrity of the outer sheath.


6.6.3.2 Uniform corrosion will be caused by CO2 in thepresence of deoxygenated seawater. This uniform corrosionshould be accounted for in the selection of the armor wirethickness. Corrosion from oxygenated water, in the immedi-ate vicinity of tears in the outer sheath, should be controlledby appropriate design of the cathodic protection system. Nopitting corrosion should occur under design environmentaland stress conditions that could cause utilization factors toexceed design criteria or affect the service life requirements.

6.6.4 Erosion

6.6.4.1 The production of reservoir sand may cause ero-sion in the carcass layer of ßexible pipes. In addition, the sandmay remove any protective Þlms on the carcass, therebyincreasing corrosion. Therefore, the erosion and erosion/cor-rosion rates should be calculated, with the calculations basedon test data (see 9.6.6 for guidelines on erosion tests). Calcu-lations should conÞrm the following:

a. The hydrostatic collapse resistance of the eroded/corrodedpipe is not lower than the design requirements for the speci-Þed service life.b. The tensile load capacity with the eroded/corroded pipe isnot less than the design requirements.

6.6.4.2 Erosion rates will be most severe at high curvatureareas. Important parameters that inßuence erosion ratesinclude ßuid velocity, amount and size of produced sand, car-cass geometry, and steel material. The partial pressure of CO2and the ßuid temperature have a signiÞcant effect on the ero-sion/corrosion characteristics of the carcass. Some test data ispresented on erosion tests in [14].

6.6.5 Fatigue Resistance

6.6.5.1 Adequate fatigue resistance of steel wire materialsfor dynamic applications is required. Fatigue analysis (see8.2.4) should show that all stresses are below the materialendurance limit. Otherwise, fatigue damage calculationsshould be performed, such as with MinerÕs method usingdesign S-N curves and accounting for damage due to cycleswith stresses below the endurance limit. The determinationof the S-N curves is critical for the fatigue analysis. Section6.2.4.5 of API SpeciÞcation 17J and Section 6.2.4.4 of APISpeciÞcation 17K speciÞes relevant test requirements,namely that S-N data are to be developed based on theactual annulus conditions and the design basis for the annu-lus, i.e., exposure to air, seawater, or design annulus envi-ronment for unbonded pipes, and based on rubberizedcables and pipe bore conditions for bonded pipe.

6.6.5.2 The initial objective of the S-N tests should be toidentify the endurance limit of the material, accounting forthe relevant environment. Data from previous testing in moresevere conditions may be used. Note that a reduction in theendurance limit is expected for sour service applications. Thefollowing recommendations are given for S-N testing:

a. Tests should consider variations in the material strength andhardness. Softer material will generally give a lower fatiguelimit in air but this may change for corrosive environments.

b. The standard S-N for wires of unbonded pipe tests arebased on un-notched specimens. When pitting, wear, corro-sion, or other sources of notches are likely to occur,consideration should also be given to performing tests withnotched specimens or to use the results of full scale tests forvalidation. This would give a lower-bound S-N curve for pit-ted or worn wires, or wires scratched during manufacture.

c. The recommended notch is a 60 degrees Vee, with a depthof 0.2 mm and a root radius of 0.025 mm. This represents typ-ical surface anomalies found in full-scale sour service testsbecause of corrosion and also represents the worst case forscratches, damage, and corrosion experienced during manu-facture and service. For round bar specimens, the notchshould be fully circumferential. For ßat wires, it should be asingle-sided notch.

d. The number of samples and stress levels for developmentof S-N data should be in accordance with ASTM E739 [15].Where appropriate strain gauges should generally be used forstress measurements. The cyclic load test frequency shouldrepresent the in service load frequency. A higher test fre-quency is allowed if the effect of the higher frequency isdocumented. A recommended maximum frequency is 0.5 Hz.

e. SufÞcient S-N data should be available to conÞdentlyextrapolate the S-N curve to stress levels below the endurancelimit. The S-N curve may have a reduced slope below theendurance limit. Results should be presented in accordancewith ASTM E468 [16].

f. The endurance limit should be the stress level at whichspecimens exceed 1x107 cycles with no evidence of fatiguecracks. The endurance limit stress is only relevant for fatiguelife analyses that do not include any cycles with stressesabove the endurance limit.

6.6.5.3 Unbonded ßexible risers generally will be designedon the basis that the outer sheath will never be breached, i.e.,no ßooding of the annulus with seawater. However, servicelife analysis for dynamic applications should calculate thelength of time to failure of tensile and pressure armors whenthe annulus is ßooded with seawater from a rupture of theouter sheath. This is deÞned as an accidental situation, withthe calculated service life determining the length of time dur-ing which the pipe can be replaced. The replacement timeshould be included in the operation manual.

6.6.6 Hydrogen Embrittlement

Cathodically protected, high tensile strength steels may besubject to hydrogen embrittlement. Section 6.2.4.6 of APISpeciÞcation 17J speciÞes required testing to conÞrm satis-factory performance of high strength wires of unbonded ßex-ible pipes subject to cathodic protection.


7 System Design Considerations

7.1 SCOPE

The scope of this section relates to the overall ßexible pipesystem and not speciÞcally to the ßexible pipe itself. The sec-tion gives recommendations on system related design issues,as follows:

a. General system design requirements.

b. Flowline design requirements.

c. Riser design requirements.

d. Floating pipes.

e. Ancillary component design.

f. System interfaces.

In addition, system issues signiÞcantly impacting theoverall project are identiÞed throughout this section.Detailed consideration of these issues at an early stage inthe project can result in signiÞcant cost savings and designsimpliÞcations.

7.2 GENERAL SYSTEM REQUIREMENTS

7.2.1 Introduction

This section covers requirements that are common to allßexible pipe systems.

7.2.2 Transported Fluid Considerations

7.2.2.1 The ßuid velocity is important, particularly whereabrasive materials such as sand in the produced ßuids mayresult in wear of the pipeÕs internal layer. Fluid velocities ofthe ßowline/riser system are based on system pressure dropand the internal friction parameter for the ßexible pipe. Thefriction parameter varies signiÞcantly between smooth andrough bore pipes because of the carcass construction in arough bore pipe. Typical values for absolute friction factor areas follows:

Rough bore pipe: ID (mm)/250

Smooth bore pipe: 5 µm

7.2.2.2 The above roughness values can generally be con-sidered to be conservative. For the rough bore pipe, the fric-tion is strongly inßuenced by the carcass characteristics, e.g.,ID and proÞle dimensions. If required, a more accurate fric-tion factor can be calculated from tests.

7.2.2.3 The design of ßexible pipe systems should considerthe effect of variations in internal ßuid density over the life ofthe project, particularly for riser systems, where a change inßuid density can change the shape of the riser conÞguration.In the case of two phase ßow, the effect of slug induced vibra-tion should be considered.

7.2.3 Corrosion Protection

7.2.3.1 The metallic components of the ßexible pipe sys-tem exposed to corrosive ßuids should be selected so as to becorrosion resistant or alternatively be protected from corro-sion. Corrosion protection can be achieved by one or more ofthe following methods:

a. Coating.

b. Application of corrosion inhibitors.

c. Application of special metallic materials or cladding.

d. SpeciÞcation of corrosion allowance.

e. Cathodic protection.

7.2.3.2 The implications for overall system design of pro-viding corrosion protection should be assessed. Reference ismade to Section 5.4.2 of API SpeciÞcations 17J/17K for cor-rosion protection requirements and DnV recommended prac-tice B401 for guidelines on the design of cathodic protectionsystems.

7.2.4 Thermal Insulation

7.2.4.1 If the ßuid temperature inside the system must bemaintained at a particular level, thermal insulating layers maybe added to the ßexible pipe cross-section to provide addedthermal insulation. It is important to ensure that the insulatingmaterial used is compatible with the annulus ßuids to which itis likely to be exposed. Typically, both pressure and tempera-ture limits apply to the use of these insulating materials andshould be considered in the selection process. API SpeciÞca-tions 17J/17K lay down minimum requirements for the use ofthermal insulating layers.

7.2.4.2 Design of a ßexible pipe to meet a speciÞed ther-mal insulation coefÞcient should include resistance from thesurrounding environment. Burial or trenching and backÞllwill provide signiÞcant thermal resistance and may minimizeor avoid a requirement for thermal insulation layers.

7.2.5 Gas Venting—Unbonded Pipe

7.2.5.1 The purpose of gas venting is to enable gas whichhas diffused through the internal pressure sheath of the ßexi-ble pipe to escape, and thus avoid build-up of gas pressure inthe annulus of the ßexible pipe system (see 8.2.2).

7.2.5.2 A gas venting system comprises small bore pipesconnecting the pipe annulus to gas relief valves in the pipeend Þttings. Burst disks may also be placed along the outersheath of the ßexible pipe for ßowline systems; API SpeciÞ-cation 17J speciÞes that burst disks are not to be used on ris-ers. The minimum requirements for the design of gas reliefvalves and burst disks are given in Section 5.4.4 of API Spec-iÞcation 17J. At the topside connection, the gas bleed off sys-tem from a ßexible riser should be connected to process vents


through a check valve, or vented locally to atmospherethrough a gas relief valve. The gas bleed off system shouldnever be capped during operation to avoid excessive pressurebuild-up in the annulus.

7.2.6 Pigging and TFL Requirements

7.2.6.1 The user should specify any pigs or tools to bepassed through the ßexible pipe. If pigging is required for theßexible pipe system, the following is recommended, as theymay have an important impact on the system layout. Designissues include whether to use loops (pipes in parallel) or sub-sea receivers.

7.2.6.2 For smooth bore pipes, foam or PU pigs may beused. For rough bore pipes, brush, foam or PU pigs may beused. Scraper pigs are not suitable for ßexible pipes.

7.2.6.3 Flexible pipe intended for use in TFL serviceshould be constructed with an innermost layer that will notimpede or suffer signiÞcant damage from the passage ofTFL tools. For TFL service, the pipe should conform to APIRecommended Practice 17C requirements in regards todesign, fabrication, and testing; and Annex A of the samerecommended practice in regards to internal diameter anddrift testing.

7.2.7 Fire Resistance

API SpeciÞcations 17J/17K lists the issues which shouldbe considered in assessing the resistance to Þre of the ßexiblepipe. Ultimately, Þre resistance tests may need to be per-formed. Additional resistance against Þre may be provided bythe application of an insulating protective cover on the outersheath of the pipe. Special consideration should be given tothe effect of Þre on the interface between pipe and end Þtting.

7.2.8 Piggy Back Lines

7.2.8.1 Piggy back is deÞned as the attachment of two par-allel and adjacent, independent pipes, rigid or ßexible, over asigniÞcant length. When a ßexible pipe is piggy backed to asteel pipeline or other steel structure, the ßexible pipe shouldbe sufÞciently protected against pipe/steel scufÞng and thepotential transfer of high temperatures from the steel to theßexible pipe.

7.2.8.2 Where an umbilical or smaller diameter line ispiggy backed to a ßexible pipe, the piggy back system shouldbe designed considering the following:

a. Hydrodynamic interaction, including shielding, solidiÞca-tion, hydroelastic vibrations, lift, marine growth, etc.

b. Relative motion between the lines.

c. Relative changes in length between the two lines (particu-larly due to different expansion coefÞcients between ßexibleand steel lines).

d. Clamp loads.

e. Loads and wear of the ßexible pipe.

f. Creep and long term degradation of pipe and clampmaterials.

g. Internal pressure, tension, external pressure, bending andtorsion induced change in cross-section geometry of the pipe.

7.2.8.3 For the case of a ßexible pipe riser, the method ofconnecting the piggy backed line at the vessel interfaceshould be carefully designed.

7.2.9 Connector Design

7.2.9.1 The materials from which connectors are to bemanufactured should be compatible with those within theßexible pipe and any interfacing topside piping or seabedpipeline.

7.2.9.2 If release functions are required in connectordesign, their abandonment philosophy should be clearly iden-tiÞed and detailed prior to manufacturing commencement.See 4.4.4 for a description of typical disconnection systems.

7.2.9.3 System design and fatigue loads should be clearlyidentiÞed prior to connector design commencement. Wherestrength or leak testing of a ßexible pipe is to be carried outthrough a connector, any exposed valvesÑeither open orclosedÑmust be capable of sustaining such pressures.

7.3 FLOWLINE DESIGN REQUIREMENTS

7.3.1 Seabed/Overland Routing

7.3.1.1 Routes should be selected with regard to the proba-bility and consequences of all forms of pipe damage. The fol-lowing factors should be taken into account:

a. Installation.

b. Seabed or overland route contour and conditions.

c. Trenching or rock dumping (if applicable).

d. Location of other installed equipment and pipelines.

e. Pipe expansion.

f. Accuracy of structure positions.

g. Accuracy of installation vessel positioning system.

h. As pulled-in conÞguration.

i. Ship trafÞc.

j. Fishing activities.

k. Offshore operations.

l. Corrosivity of the environment.

m. Launching of lifeboats.

n. Anchoring and mooring of other installations and vessels.


7.3.1.2 The pipe route should be selected to:

a. Minimize the need for seabed preparation.b. Minimize the vertical imperfections to be crossed.c. Ensure space for individual trenching, if required.d. Minimize pipe length.

7.3.1.3 The layout (i.e., location of wellheads, manifolds,mooring lines, PLEMs, etc.) will signiÞcantly inßuence theselection of ßowline layouts and riser conÞgurations andshould be considered early in the design.

7.3.2 Protection

7.3.2.1 Pipe protection against damage caused by objectssuch as Þshing gear, anchors, mooring lines, etc., should beconsidered, and requirements speciÞed in agreement by thepurchaser and manufacturer.

7.3.2.2 The impact energy and geometry of objects to beconsidered should be deÞned in the project Design Premise(see Section 8.2 of API SpeciÞcations 17J/17K). Impact loadsshould be quantiÞed for the intended service as normal orabnormal operations following the results of a safety analysis.The recommended requirements for pressure containmentequipment, such as the pipe structure, end Þttings, and con-nectors are as follows:

a. Normal operations: such equipment should not be perma-nently deformed.

b. Abnormal operations: such equipment should not leak.

7.3.2.3 Based upon this classiÞcation and the protectionmethod adopted, representative calculations should show thatthe pipe structure, end Þtting, and connector utilization com-ply with Section 5 of API SpeciÞcations 17J/17K.

7.3.2.4 In the evaluation of the optimum technical and eco-nomical protection method, the following should be takeninto account:

a. Seabed or ground conditions.

b. Pipe and protection facility installation.

c. Pipe expansion from temperature, pressure, etc.

d. Bending as a result of upheaval buckling.

e. Inspection and maintenance.

f. Pipe retrieval.

7.3.3 On-bottom Stability

7.3.3.1 General

7.3.3.1.1 The stability of a section of ßowline on the sea-bed or ground is directly related to its (submerged) weight,the environmental forces, and the resistance developed by thesoil. A stability analysis would demonstrate that the (sub-merged) weight of the unburied ßowline is sufÞcient to meet

the required stability criteria. Pipeline stability is to be con-sidered for both installation and operation conditions. Flota-tion and/or sinking of the pipe for the most critical internalßuid conditions should be checked. Issues to be consideredduring the stability analysis should include the following:

a. Lateral displacement from an installed position as a resultof expansion, settlement, or hydrodynamic effects.

b. Geometric limitations of surrounding system.

c. Distance from other pipes, structures, or obstacles.

d. Internal ßuid density and its variation during the servicelife.

e. Pipeline tension, curvature, and torsion.

f. Interaction with lateral buckling resulting from axialforces.

g. Fatigue damage.

h. Wear and deterioration of outer sheath.

i. Damage to sacriÞcial anodes.

j. Loading on end connections.

7.3.3.1.2 If the incorporation of mattresses is required toprovide stability, their suitability with respect to pipe coverabrasion and damage from protrusions should be conÞrmed.If rock dumping is provided, the general form and size ofrocks should be such that no damage is sustained to the pipeduring deployment.

7.3.3.2 Analysis Methods

7.3.3.2.1 The following stability analysis methods may beemployed:

a. Dynamic analysisÑinvolving a full dynamic simulation ofthe pipeline resting on the seabed, including modeling of soilresistance, hydrodynamic forces, boundary conditions anddynamic response.

b. Generalized stability analysisÑbased on a set of non-dimensional stability curves that have been derived from aseries of runs with a dynamic response model.

c. SimpliÞed stability analysisÑbased on a quasi-static bal-ance of forces acting on the pipe.

7.3.3.2.2 Further details on the above analysis are given inVeritec Recommended Practice 305 [17] and American GasAssociation Guidelines [18].

7.3.3.3 Stability Criteria

The pipe supplier/designer should specify and justify sta-bility criteria for the particular application, which may bebased on guidelines in Veritec Recommended Practice E305[17] and DnV Rules for Submarine Pipeline Systems [19]. Asa minimum, the design criteria speciÞed in API SpeciÞcation17J/17K should be satisÞed.


7.3.4 Upheaval Buckling


7.3.4.1.1 A ßexible pipe laid in a trench may be suscepti-ble to upheaval buckling stemming from longitudinal expan-sion of the ßowline caused by internal pressure andtemperature loadings. For ßexible pipe, internal pressure isthe dominating factor contributing to upheaval buckling ofthe pipe.

7.3.4.1.2 In addition, changing the lay angle of the pipecan produce longitudinal expansion of the pipe, with the opti-mal angle being approximately 55 degrees for a two-ply pipe.Additional pairs of plies may change the optimal angle signif-icantly.

7.3.4.1.3 The ßexible ßowline may be allowed to buckleprovided that the design criteria of 7.3.4.4 are not violated.The potential for upheaval buckling can be evaluated by ana-lytical and/or experimental methods. The parameters thatinßuence the upheaval behaviour of a ßexible ßowline andthat should be incorporated into any investigation of upheavalbuckling include the following [20]:

a. Operational pressure and temperature distributions alongthe ßowline, including hydrotest conditions.

b. Vertical imperfections in the ßowline foundation.

c. Variations in uplift resistance along the line, such as vary-ing soil cover height and conditions, soil longitudinal friction,soil rotational stiffness, and its contribution to bending resis-tance of the pipe.

d. Uplift resistance as a function of pipe uplift displacement.

e. Stiffness properties of the pipe cross-section as a functionof pressure and temperature; in particular, axial compressionstiffness and bending stiffness of the pipe.

f. Relaxation with time of the initial lay pretension stressesin the pipe.

7.3.4.2 Methods of Prevention

7.3.4.2.1 Measures to prevent or limit the extent ofupheaval buckling include the following:

a. Burying the pipe in a trench.

b. Rock dumping.

c. Wide and open trench to allow horizontal snaking.

d. Laying the pipe with internal pressure to provide initialpretension in the line prior to burial.

e. Optimize tensile armor lay angle.

7.3.4.2.2 A feasible way of pre-tensioning a ßexible pipe-line is to restrain the pipe (e.g., by rock dumping) while it issubjected to axial expansion due to internal pressure. When

evaluating the resulting effective pre-tension in the line, thefollowing should be considered:

a. Residual axial compression loads due to the frictionalresistance between the pipe and the seabed.b. Relaxation of pre-tension loads because of possiblestraightening of formed loops (lateral buckles).c. Creep of pipe materials with time.

7.3.4.3 Analysis Methods

7.3.4.3.1 A linear model may be used to determine ifupheaval buckling may occur. If it is a concern, then a non-linear model is required for analysis of upheaval buckling.The nonlinear model should account for varying soil coverfrom imperfection geometry, nonlinear pipe/soil interaction,and geometric nonlinearities because of large deßections ofthe ßowline. It may be assumed that the material propertiesexhibit a linear behaviour.

7.3.4.3.2 An initial imperfection in the installed ßowlineconÞguration is characterized by an imperfection amplitudeand a corresponding imperfection wavelength, assuming asymmetrical shape about the imperfection apex. In theunloaded condition, the pipe is assumed fully supported bythe soil. Subjecting the pipe to temperature and pressure loadsgenerates an axial compression force in the pipe, causing thepipe to buckle into a new equilibrium shape characterized bya buckling wavelength and a buckling amplitude, thereby cre-ating a resulting uplift amplitude at the apex of the imperfec-tion. See [21, 22, and 23] for more details on the analysismethodologies.

7.3.4.4 Design Criteria

7.3.4.4.1 The upheaval buckling design criteria should bebased on the following [20]:

a. The pipe is not anywhere bent below its minimum allow-able bend radius.

b. The pipe does not deviate beyond the trench or bermboundaries.

c. Movement restrictions imposed by the trench and inÞll donot result in pipe structure stresses or loads which violate thedesign criteria in Section 5 of API SpeciÞcation 17J/17K.

d. The upheaval buckling process does not subject the pipe toother failure modes that could cause leakage of the pipe, e.g.,expose the pipe to trawl board snagging.

e. Adequate safety margin against snap-through buckling.

7.3.4.4.2 To avoid an upheaval creep mechanism takingplace because of variations in temperature and pressure dur-ing the service life of the line, the uplift displacement is to belimited to a maximum of 0.75 dult, where dult is the burialdepth.


7.3.4.4.3 To ensure an adequate safety margin againstsnap-through buckling failure, the distance between the pre-and post-buckling equilibrium curves at speciÞed design con-ditions is not to be less than 0.1 meter when plotted in a tem-perature (or pressure) versus uplift displacement plane.

7.3.4.4.4 The uplift resistance of the protection cover is tobe documented. Consideration is to be given to possibledecrease in uplift resistance because of undrained cover/back-Þll or change in cover properties as a result of the installationmethod employed.

7.3.4.4.5 Following the installation of a ßexible pipewhich is susceptible to upheaval buckling, the design require-ments are to be veriÞed with regard to the following:

a. Vertical imperfections of installed line.

b. Burial depth and berm height/width.

7.3.4.4.6 For a trenched pipeline with natural backÞll, it isto be documented that the required cover is present prior totaking the line into service. When a pipeline is situated in anopen trench, the resulting pipe conÞguration should bechecked when the line is brought into service.

7.3.5 Pipeline Crossing

7.3.5.1 If a ßexible pipe crosses another ßexible, steel pipeor umbilical in service, suitable protection should be placedbetween the two pipes unless it can be shown that the MBRand other design criteria are not violated. Protection mayinclude sand bags, stabilization mattresses, structural bridges,or low friction matting. If multiple lines are installed in a sin-gle trench, the number of crossovers should be minimized bythe installation procedures.

7.3.5.2 If a crossover involves both liquid and gas-carryingpipes, the gas-carrying pipe should be placed above the liquidpipe, unless the liquid pipe is lighter than the gas pipe,accounting for content. Where crossed ßexibles are suscepti-ble to movement, any protection facility should take suchmovement into account.

7.3.5.3 Where a number of pipes come into contact underconstant or frequent movement, the ßexible pipes concernedshould be provided with abrasion sleeves constructed ofmetal or polymer. The sleeves should sufÞciently cover themaximum extent of relative movement and have enoughthickness to account for expected wear. The sleeve require-ments should be determined during the detail design of thepipe system.

7.4 RISER DESIGN REQUIREMENTS

7.4.1 Riser Configuration

7.4.1.1 A considerable part of ßexible riser system designis the determination of conÞguration parameters so that the

riser can safely sustain the extreme seastate loadings forwhich it is to be designed. A safe riser design nowhereexceeds maximum allowable tension or minimum allowablebend radius criteria, as per API SpeciÞcation 17J/17K, whensubjected to these extreme wave and current loadings. Awell designed riser conÞguration is safe and provides com-pliancy to vessel motions in a cost-effective manner. A riserthat is compliant to vessel motions minimizes the station-keeping requirements for the vessel and, in turn, reducesmooring costs.

7.4.1.2 Large riser rotations, combined with large tensionsnear the riser/vessel or riser/seabed termination points, arealso an undesirable riser response to seastate loading. In thiscase, large bend stiffeners would be required at the pipe endÞtting to avoid exceeding the minimum allowable bend crite-rion in the ßexible pipe at this location.

7.4.1.3 Flexible risers are commonly deployed in one ofÞve standard conÞgurations, as illustrated schematically inFigure 4.

a. Free hanging catenary.

b. Lazy-S.

c. Steep-S.

d. Lazy wave.

e. Steep wave.

7.4.1.4 Key points about these riser conÞgurations are asfollow:

7.4.1.4.1 Free Hanging Configuration

This is the simplest, and generally the least expensive, riserconÞguration. A key problem with this solution, however, isthat if there are any signiÞcant Þrst order wave motions at thevessel connection (particularly heave), the amplitude ofdynamic tension is transferred directly to the seabed and thisinevitably leads to compression at the riser touchdown point.Buckling and overbending of the pipe below its allowablelimit are consequences of this effect. Furthermore, the free-hanging riser is not very compliant to vessel motions: risertop tension increases rapidly with far vessel offset, and largevessel offset motions result in correspondingly large andundesirable motions of the riser/seabed touchdown point.

Because of its simplicity, however, the free hanging conÞg-uration is always worth considering as a potential solution,particularly for mild environment deep water applications. Indeep water applications, the hang-off loads on the vessel canbe large due to the suspended riser length.

7.4.1.4.2 Lazy-S and Steep-S Configurations

The introduction of a subsea buoy (see Figure 14) into theriser conÞguration has two main functions:


a. Provides a Þlter to stop the direct transfer of dynamic ten-sion amplitude to the seabed that occurs with the free-hanging conÞguration.

b. Supports part of the weight of the riser, thereby reducingstatic tension at the vessel connection.

The change in seabed touchdown point is controlled by thelateral motion of the subsea buoy. Increasing the size of thebuoy correspondingly also increases the lateral restoringforce through the buoy tethers, and this in turn tends to reducethe lateral motions of the buoy. However, a larger buoy is alsosusceptible to increased hydrodynamic loading. S conÞgura-tions enable ßexible lines to ascend to the ßoating vessel inbundles over a single buoy. The analysis of the hydrodynamicbehaviour of the buoy is an important consideration in thedesign of these systems. In general, the steep-S riser buoy ismore susceptible to torsional instability than is the lazy-Ssolution.

7.4.1.4.3 Wave Configurations

For wave conÞgurations, the buoyancy (see Figure 15) isapplied to the pipe in a distributed manner rather than as aconcentrated point load as for the S conÞgurations. Generally,the wave conÞgurations are more compliant to environmentalloading than the S conÞgurations and ascend to the ßoatingvessel as individual lines (or clamped bundles). While theincreased compliancy to vessel motions of the wave conÞgu-rations is a deÞnite advantage, the compliant nature of theriser conÞguration itself to environmental loading and partic-ularly to cross loading makes riser interference with adjacentrisers or structures an important design consideration.

In general, the steep wave riser is less compliant than thelazy wave. The shape of the lazy wave riser is particularlysusceptible to variations in internal ßuid density, thoughundesirably large motions can be avoided by designing a ßex-ible pipe cross-section with low drag to weight properties.

The pliant wave shown in Figure 4 is a modiÞcation to thesteep wave conÞguration. Close to the seabed touchdown, thetension in the riser is transferred via a riser clamp to ananchor line, which is tied to the seabed by a clump weight, ora suction anchor. The riser itself touches down on the seabedalmost like a lazy wave conÞguration, except in this instancethe touchdown point is well controlled by the near verticalriser anchor line and an optional horizontal anchor lineclamped between the seabed section of the riser and theclump weight or suction anchor.

7.4.2 Riser Interference

7.4.2.1 The riser system design should include evaluationor analysis of potential riser interference (including hydrody-namic interaction) with other risers and between risers andmooring legs, tendons, vessel hull, seabed or any otherobstruction. Interference should be considered during all

phases of the riser design life, including installation, in-place,disconnected, and unusual events. The accuracy and suitabil-ity of the selected analytical technique should be assessedwhen determining the probability and severity of contact.

7.4.2.2 Riser systems should be designed to control inter-ference because of potential damage to the risers or otherparts of the system if interference occurs. Hydrodynamicinteraction of multiple risers, including shielding, should beconsidered.

7.4.2.3 Either of two design approaches may be taken tocontrol riser interference. One approach requires that the risersystem has an acceptably low probability that the clearancebetween a riser and another object is less than a speciÞedminimum value. The other approach permits contact betweenthe riser and the other object but requires analysis and designfor the effects of contact.

7.4.2.4 Interference may occur between a riser and anyobject which has dynamic characteristics different from thoseof the riser and which is sufÞciently close to it. Objects mayinclude the vessel hull; a riser of different size; or a riser hav-ing different properties, such as different contents, extent ofmarine growth, top tension or tension distribution, or otherboundary conditions; or a riser in a different ßow Þeld causedby wake effects. Clearly, this type of interference is moresevere than between the risers with similar dynamic charac-teristics, and the size and direction of impact loads should bequantiÞed.

7.4.2.5 Interference between adjacent wave type risers atthe buoyancy section should not be allowed.

7.4.3 Load Bearing Structures

7.4.3.1 If load bearing structures are used to support ßexi-ble pipes, they should be designed such that the pipe is notsubjected to excessive wear, bending, or crushing. As such,steel materials should be provided with suitable cathodic orcoating protection, and all surfaces in contact with the ßexiblepipe should be provided with a surface radius greater than thepermissible MBR for the ßexible pipe.

7.4.3.2 Structures within a ßexible pipe system should bedesigned to accommodate ßexible pipe movements. Loadbearing steel components should be designed in accordancewith relevant steel standards for offshore structure design.

7.4.4 Pipe Attachments

Interactive forces between pipe and any attachment shouldbe determined along with resultant pipe deßections. Due con-sideration should be given to mid-water support buoys withrespect to their overall behaviour within the system in mini-mize dynamic effects imposed on the pipe.


7.4.5 Riser Bases

Riser bases should be located in relation to the overallsystem such that the pipe does not exceed design MBR inany load case and the maximum excursion capability of theßexible pipe top end is facilitated. Installation tolerance forthe riser base should be accounted for in the riser systemdesign.

7.4.6 Jumper and Spool Pieces

7.4.6.1 Each ßexible pipe jumper and spool piece shouldbe analyzed in accordance with the load cases deÞned in theDesign Premise, see Section 8.2 of API SpeciÞcation 17J/17K. All associated equipment should be subjected to a simi-lar level of analysis in order to establish suitability. The analy-sis shall take account of seabed conditions and pipe stability.

7.4.6.2 The conÞguration of a spool piece should be suchthat minimal loading is imposed on the ßexible pipe, withspecial emphasis being placed on the area immediatelyaround the end Þtting. Spool pieces and their systems shouldbe manufactured so that pipe lengths provide sufÞcient ßexi-bility during installation and operation.

7.5 ANCILLARY COMPONENTS

7.5.1 General

Design requirements for typical ancillary components inßexible pipe systems are presented in this section.

7.5.2 Connectors and Rigid Pipe Components

Connectors and rigid pipe components should be designedaccording to the same requirements as the ßexible pipe endÞtting as speciÞed in Section 5 of API SpeciÞcation 17J/17Kor should be a standard connector (such as API SpeciÞcation16A, API SpeciÞcation 17D, ANSI B16.5, etc.) rated for thedesign pressure and other imposed loads.

7.5.3 Bend Stiffener

Appendix B of API SpeciÞcation 17J17K gives recom-mended procedures for the design, material selection, manu-facture, testing, and marking of bend stiffeners.

7.5.4 Bend Restrictor

Appendix B of API SpeciÞcation 17J/K gives recom-mended procedures for the design, material selection, manu-facture, testing, and marking of bend restrictors.

7.5.5 Bellmouths

7.5.5.1 A bellmouth is one type of bend limiter for a ßexi-ble pipe and is used for dynamic applications where ßexible

risers are pulled through guide tubes to vessel deck level. Thelower end is ßared to avoid overbending, thus producing thebellmouth. The bellmouth design is based on the maximumoffset angle of the ßexible riser and its minimum allowablebend radius. In evaluating the fatigue life of ßexible pipe, theeffect of bellmouth contact pressure on the structural layersalternating stress should be considered.

7.5.5.2 The simplest shape of bellmouth has a constantradius along its length. This shape, however, does not providethe best protection against fatigue. Therefore, it is moreadvantageous to apply a large radius at the top section wherethe pipe is in regular contact with the bellmouth, and asmaller radius at the bottom section, where there is only inter-mittent contact in extreme conditions.

7.5.5.3 A class of bellmouths with a linear variation incurvature along the bellmouth can be described by fourparameters:

sb = length of bellmouth, measured along the curved wall,

φb = angle of bottom entry,

κb = curvature at bottom entry (equal yRb where Rb is the radius),

α = ratio between minimum (top) and maximum (bot-tom) curvature,

Rb = minimum allowable bend radius.

7.5.5.4 Figure 22 shows a schematic of the parameters inthe design of a bellmouth. The shape of the bellmouth can bedeÞned as a function of s as follows:

(3)

(4)

(5)

7.5.5.5 In general, both the required length and diameter ofa bellmouth are dependent on the entry angle, and the lengthis also dependent on the ratio of minimum (top) and maxi-mum (bottom) curvature.

7.5.5.6 The entry angle, φb, should be at least 5 degreesgreater than that calculated to be required from all design loadcases, accounting for all effects including vessel rotation.

φ s( )1 α2Ð( ).κ b

2

4φb

--------------------------.s2 α .κ b.s+=

κ s( )1 α2Ð( ).κ b

2

2.φb

--------------------------.s α .κ b+=

sb

2.φb

1 α+( ).κ b

-------------------------=


7.5.6 Clamping Devices

7.5.6.1 Permanent clamping devices should be designedaccording to the requirements for load bearing structures.If such clamps are applied, sufÞcient testing of similarclamps on samples of the proposed pipe in simulated con-ditions should be carried out and fully documented prior toinstallation.

7.5.6.2 Clamping should not impose local or preferentialloading on the pipe structure so that its pressure and structuralintegrity are compromised during its design life. Clampingshould not accentuate fatigue, abrasion or fretting in the pipestructure beyond the limits imposed by the appropriate usagefactors. The materials selected for the clamps should be creepresistant and suitable for long-term exposure in the speciÞedenvironment.

7.5.7 Buoyancy Devices

7.5.7.1 The analysis should identify the interactive forcesbetween pipe and buoyancy devices and resultant pipe deßec-tions. The design should show that sliding of the buoyancydevices along the pipe is prevented, i.e., the clamping forceshould be sufÞcient that the friction between the buoyancyclamp and pipe is greater that the maximum longitudinalloads on the buoyancy devices, including a safety factor of atleast 1.0.

7.5.7.2 When selecting either a steel or polymer material,the following should be considered:

a. Suitability for water depth.

b. Length of service at water depth.

c. Resultant size and dynamic loading effects on pipe.

Figure 22—Parameters Used to Define a Bellmouth Shape

Db

Do

X

L

Y

S

S=0, φ=0, κ=α.κb,R0=1/α.κb

S=Sb, φ=0b, κ=κb,Rb=1κb

φ


d. Durability.

e. Previous history under similar conditions.

f. Safety.

g. Handling characteristics.

7.5.7.3 The arch supporting structure should be designedin accordance with load bearing structures (see Section7.4.3). Attachment of buoyancy modules to a riser shouldtake account of hydrodynamic forces, self weight, inertialforces, slamming forces, and effect of pressure on moduleclamp contact pressure.

7.5.7.4 Buoyancy modules are to maintain sufÞcient buoy-ancy over their service life to fulÞl their function, i.e., longterm resistance to hydrostatic pressure is required. All materi-als in the structure should be selected based on the environ-mental requirements with sufÞcient corrosion resistance forthe speciÞed service life.

7.5.7.5 Damage to one single buoyancy element should notresult in unacceptable loss of buoyancy for the pipe system asa whole. This may require the installation of bulkheads insteel buoyancy tanks. After loss of 10 percent of distributedbuoyancy or one compartment in a subsea buoy/arch system,the riser conÞguration should still be Þt for purpose.

7.5.7.6 Materials, such as syntactic foam, for buoyancymodules should be qualiÞed by tests in order to conÞrm theirresistance to hydrostatic pressure for the speciÞed waterdepth. Water absorption over the speciÞed service life shouldbe included in the analysis of the performance of the materi-als. The loss of buoyancy from water absorption should bedocumented and the end of life value used for a design check.

7.5.8 Riser Base

7.5.8.1 The riser base, including pipework, structural sup-ports, foundation, etc., should be designed in accordance withindustry standards [24, 25, 26]. The pipe and J-tubes should(where applicable) be arranged such that no bendingmoments are imposed on the end Þtting of the static pipe. Thefollowing issues may be pertinent to the Þeld in question:

a. Gravity or piled structure.

b. Isolation/manifolding facility.

c. Emergency abandonment procedure.

d. Riser conÞguration.

All such details should be fully evaluated prior to designcommencement.

7.5.9 Temporary Lifting Appliances

Temporary appliances should be designed in accordancewith industry standards, such as the DnV Rules for CertiÞca-tion of Lifting Appliances [27]. As a general rule, the lifting

gear should be designed for dynamic loading duties. Thisrequirement should also apply for equipment, such as shack-les and forerunners with associated gear.

7.5.10 Tether Design

If a ßexible static or dynamic riser requires tethering (suchas S-type riser conÞguration), the strength of the tether shouldallow the pipe separate from the tether prior to failure of thepipe structure (unless a pipe failure or load-limiting joint isdesigned at the tether connection). The tether should bedesigned for all events with probability of occurrence greaterthan 10Ð4.

7.6 SYSTEM INTERFACES

Interface issues should be considered at an early stage of aproject, as they may have a serious impact on both the pipeand system design. Clear interface deÞnition will allow thedevelopment of an optimized overall solution for the system.Relevant issues include the following:

a. Connection locationÑconnection of the risers above orbelow the water line will have important implications fordesign, installation, and use (condition monitoring).

b. Bend limiter selectionÑselection of bellmouths or bendstiffeners should be addressed prior to the design of the top-sides interface. Note that bellmouths will require signiÞcantlymore space than stiffeners.

c. Location of bend limiterÑwhether the bend limiter islocated at the end Þtting or at the end of an I-tube spool pieceshould be considered.

d. Flowline installation conditionsÑfor ßowlines to betrenched and back-Þlled, consideration should be given toupheaval buckling requirements and to the possible need topressurize prior to burial.

e. Connection designÑconsideration should be given to thepossible future requirement for internal inspection tools to beused. This would require a pigging system to be designed toallow access for the launch of inspection tools. This alsoapplies where pigs may be launched from the top connectionto the ßexible.

f. ConnectorsÑaspects which should be speciÞed includeheight and location of ßanges, diverless or diver assisted tie-ins, and ßange and hub speciÞcation.

g. I and J-tubesÑuse of I and J-tubes will affect ßexible pipeinstallation options, and this should be considered during thedesign of the tubes. Any requirement for spool pieces at theend of I-tubes will signiÞcantly effect the loads on the I-tube.

h. Subsea connectionsÑuse of riser conÞgurations with hori-zontal connections (e.g., lazy-S) can simplify installation andsigniÞcantly reduce the complexity of the PLEM/riser basestructure.


8 Analysis Considerations

8.1 INTRODUCTION

The objective of this section is to give recommendations onßexible pipe analysis techniques, deÞne the loads typicallyexperienced in pipe applications, and provide guidelines onthe evaluation of the pipe or system response to these loads.

8.2 ANALYSIS TECHNIQUES

8.2.1 Local Analysis

8.2.1.1 Because of the composite layer structure of a ßexi-ble pipe, local cross-section analysis is a complex subject,particularly for combined loads. Local analysis is required torelate global loadings to stresses and strains in the pipe. Thecalculated stresses and strains are then compared to the speci-Þed design criteria (Tables 6 and 7 in API SpeciÞcation 17Jand 17K respectively lists relevant criteria) for the load casesidentiÞed in the project design premise (see 5.5 of this recom-mended practice for guidelines on selection of load cases).

8.2.1.2 The simpliÞed formulae given in [5] may be usedfor a preliminary check of loads on ßexible pipe. For detaileddesign, more reÞned analysis techniques that account for allrelevant effects are required. The required analysis can beperformed by a number of computer programs. Minimumrequirements for the cross-section analysis methodology areprovided in Section 5.2.1 of API SpeciÞcations 17J/17K.

8.2.1.3 Load effects in pipe wall sections may be docu-mented by prototype testing. Numerical analysis methodsalso may be used to predict local stresses. Under numericalanalysis, the analysis results may be validated by prototypetesting.

8.2.1.4 Design formulas are to be related to the speciÞctype of pipe design and may be validated for those speciÞcdesigns by strain gauge results from prototype tests. JustiÞca-tion for extrapolation of results is to be documented. Whenconsidering use of analytical methods, the actual load situa-tion in the pipe is to be considered, especially with regard tocombined loading.

8.2.2 Analysis of Pipe Wall Environment

8.2.2.1 The pipe wall for either bonded or unbonded con-struction is the space occupied by the primary reinforcementelements.

8.2.2.2 The analysis of the pipe wall environment of a ßexi-ble pipe is an important consideration, particularly for thedetermination of gas release requirements and metallic materialfailure modes. The following pipe wall environment character-istics should be considered for the design of the ßexible pipe:

a. Permeated gas and liquids.

b. External ßuid ingress (seawater).

8.2.2.3 The polymers used for the internal pressure sheathallow ßuids in the pipe to permeate into the pipe wall. Thispermeation rate (leakage) is negligible with regard to pipeperformance (ßow capacity). The pipe system design, how-ever, must allow for safe escape of the permeated gas. Gaspermeation from the conveyed ßuid into the pipe wall shouldbe calculated using a qualiÞed procedure. The permeationrate is a function of internal and external pressures, surfaceareas, sheath thickness, and permeability coefÞcient. Notethat the permeability coefÞcient depends on material, gascomponent, and temperature.

8.2.2.4 H2S gas permeation into the pipe wall environmentwill determine if a particular application is to be consideredas sweet or sour service. To make this determination, thepressure in the pipe wall and the concentration of H2S in thepipe wall must be calculated. In addition, CO2 permeationrates are required to determine the annulus pH level.

8.2.2.5 After a transient period, an equilibrium condition isreached in which the partial pressures in the pipe wall will belower or at a maximum equal to the partial pressures in thepipe bore, with the actual value dependent on pressure, tem-perature, polymer materials, etc.

8.2.2.6 As an initial approximation, the partial pressure ofH2S in the pipe wall can be assumed to be the same as in thepipe bore. This should be conservative, as the pipe wall pres-sure is limited to the gas escape pressure at the particularlocation accounting for external seawater pressure. Note alsothat because the different permeation rates of H2S and othercomponents, there may be differences between the ßuid com-position in the pipe bore and pipe wall.

8.2.2.7 Parameters that inßuence the actual partial pressureof H2S in the pipe wall are discussed in reference [28]. Thepartial pressure should then be used to check against NACErequirements. If testing is required, the partial pressure ofH2S used in testing should be greater than or equal to the cal-culated pressure.

8.2.2.8 For unbonded ßexible pipe for static service, thepipe wall of the ßexible pipe should be assumed ßooded withseawater. For dynamic service, the outer sheath should bequaliÞed as watertight. In addition, the service life with thepipe wall ßooded with seawater should be calculated andspeciÞed in the operation manual.

8.2.3 Global Analysis

8.2.3.1 General

8.2.3.1.1 Global analysis is performed to evaluate the glo-bal load effects on the pipe during all stages of installation,operation, and retrieval, as applicable. The static conÞgura-tion and extreme response of displacement, curvature, force,and moment from environmental effects should be evaluatedin the global analysis.


8.2.3.1.2 Global load effects generally are to be docu-mented by numerical analysis methods, such as the Þnite ele-ment method (FEM) [29]. The analysis should account forthree-dimensional dynamic response, stochastic response(irregular sea), and nonlinear effects [30, 31]. The computermodel and results should be fully documented.

8.2.3.1.3 Static and quasi-static analysis methods may beused for preliminary conÞguration design. In the detaileddesign stage of a project, however, all time-varying loads(such as waves) should be modeled with dynamic analyses toaccurately account for inertia effects.

8.2.3.1.4 Critical phases during installation and operationmay be analyzed by a stepwise time integration procedure.Very large changes in the riser conÞguration require a nonlin-ear solution procedure. For dynamically sensitive structures,nonlinear time domain simulations are required. The waveconditions to be considered may be described by determinis-tic or stochastic methods. Structural damping may be takeninto account with a proportional type damping used withoutthe inertia component.

8.2.3.1.5 Pipe characteristics and operational data are to beconsidered in the analysis. For some applications the bendingstiffness characteristics of the pipe will be critical, e.g., lightlines which are subject to severe dynamic motions or the sea-bed touchdown region in the lower catenary section of a lazy-S conÞguration. In such cases, the bending stiffness will needto be assessed accurately, to determine if buckling is occur-ring or if MBR design criteria are being violated. Parametersrelevant to the pipe bending stiffness include number, thick-ness (including tolerances) and material in polymer layers,mean temperature in the layers (pipe will be stiffer at lowertemperatures), non-linear material characteristics (aged andunaged material) and internal pressure. The effect of the ten-sile armor layers on the stiffness can generally be ignored, asthe armor wires will have slipped when the pipe is bent to ahigh curvature.

8.2.3.1.6 Hydrodynamic loads may be calculated bymeans of MorisonÕs Equation [32]. CoefÞcients in 8.3.1.4may be used. For ßexible pipes with buoyancy elements, tan-gential forces are also to be taken into account.

8.2.3.1.7 For riser conÞgurations with a part of the riserresting on the sea ßoor a model of riser/sea ßoor, interactionis required. Where the local behaviour close to the sea ßoor, isof particular interest, a complete nonlinear formulation is tobe used.

8.2.3.1.8 The minimum effective tension (described in8.4.5) should be examined in order to check for possiblebuckling of the pipe. The effective tension is normallyrequired to be positive. Any effective compression which mayoccur should be shown to be tolerable for the pipe (see 5.4.1.9for compression criteria).

8.2.3.2 Static Analysis

8.2.3.2.1 The aim of the static analysis (sometimes aidedby preliminary dynamic analysis) is to determine the initialstatic geometry of the pipe conÞguration. The design parame-ters to be selected in the static analysis are typically length(s),weight, buoyancy requirements, and location of seabedtouchdown point and subsea buoy(s). The loads considered inthe static analysis stage are generally gravity, buoyancy, inter-nal ßuid, vessel offsets and current loads.

8.2.3.2.2 For ßexible risers, at least three extreme casesshould be investigated, as follows:

a. Near position analysis.

b. Far position analysis.

c. Maximum out of plane excursion.

8.2.3.2.3 Note that the extreme positions may not neces-sarily be in the plane of the riser, particularly if environmentdirectionality effects are considered.

8.2.3.3 Dynamic Analysis

8.2.3.3.1 The next stage in the design procedure (dynamicapplications only) is to perform dynamic analyses of the sys-tem to assess the global dynamic response. A system layoutand vessel position is chosen from the static analysis and aseries of dynamic load cases are considered. These load casescombine different wave and current conditions, vessel posi-tions and motions, and riser content conditions to provide anoverall assessment of the riser suitability in operating andextreme environmental conditions. See 5.5 for recommenda-tions on load case selection.

8.2.3.3.2 In the dynamic analysis phase, the effect of ves-sel motions should be combined with wave and current forcesto obtain the response of the riser. The hydrodynamic forcescan be calculated based on MorisonÕs Equation [32]. The ves-sel motions can be obtained from model tests, computer sim-ulations, or from a knowledge of the vessel RAOs and thedesign wave data.

8.2.3.3.3 Because of the geometrical nonlinearities gener-ally associated with dynamic behaviour of ßexible risers,analysis in the frequency domain is generally inappropriate;consequently, ßexible riser analyses are usually performedwith time domain simulations.

8.2.3.3.4 Analyses for the static and dynamic analysisphases are often interrelated in the sense that a certain amountof iteration will be needed to achieve a preliminary sizing andlayout design. For preliminary dynamic analysis, a coarsermesh may often be adequate.

8.2.3.3.5 Any results from a dynamic analysis should bescrutinized for their accuracy and convergence prior toaccepting them for design. Particular attention should begiven to the adequacy of mesh selection and time stepping


used in the analysis. Sensitivity of the response to the waveapproach direction and wave period should be evaluated toproduce the most unfavorable load conditions.

8.2.3.3.6 For dynamic analysis, the seastate can be repre-sented by either regular waves (design wave) or irregular seas(design storm) as described in 8.4.1. It is recommended thatinitially the regular wave approach be used in parametricstudies. Irregular sea analysis can then be used on a preferredconÞguration for the Þnal design load cases. Irregular seas oran appropriate family of regular waves selected to representthe fatigue seastates may be used for the fatigue analysis ofthe Þnal riser design.

8.2.3.3.7 The signiÞcant response parameters that arerequired from a dynamic analysis may include the following:

a. Riser angles at top, and base (for steep conÞgurations).

b. Effective tension at top, and base (for steep conÞgurations).

c. Maximum and minimum effective tension distributionalong riser.

d. Buoy tether tensions.

e. Buoy movement.

f. Buoy riser departure angle (either side of arch).

g. Riser tension at support buoy.

h. Maximum curvature (MBR).

i. Clearances between risers for multiple risers.

j. Clearance from structure or seabed.

k. Movement and curvature of riser at touchdown point.

Note that the angles and tensions of the riser at the connec-tion points may be used to design bend limiters to preventoverbending of the riser at these locations. For vessel connec-tions the measured angles should account for the relative rota-tion (e.g., pitching) of the vessel.

8.2.3.4 Computer Programs

There are a number of proprietary computer programsavailable for riser analysis, based on both Þnite element andÞnite difference methods [33, 34]. Any program selected foruse must be capable of modeling the risers appropriately(including axial, bending, and torsional effects where rele-vant) and veriÞed as to its accuracy and dependability ofresults.

8.2.3.5 Modeling Considerations

8.2.3.5.1 The following modeling considerations are criti-cal for accuracy of results:

a. Mesh size in relation to radius of curvature obtained fromthe analysis.

b. Selection of Cd and Cm for wave load calculations (see8.3.1.4).

c. Selection of boundary conditions.

d. Selection of time step and duration for dynamic analysis.

e. Type of Þnite element.

f. Selection of damping model and coefÞcients.

8.2.3.5.2 In some cases, it may be desirable to run multipleanalyses to check the sensitivity of the results to these param-eters.

8.2.3.6 Analysis of multiple configurations

8.2.3.6.1 In many situations, risers used in a productionfacility are bundled together. Three types of bundles are asfollows:

a. Free bundle.

b. Integral bundle.

c. Multibore risers.

8.2.3.6.2 In a free bundle, the risers are free to move inde-pendently and are connected only at the termination pointsand a subsea buoy. In the analysis of a free bundle, all risersshould be included individually in a single model, i.e., singleriser models or equivalent models are not recommended fordetail design. The free bundle model should be sufÞcientlydetailed so that all motions and loads in the risers, subseabuoy, and tethers can be calculated. The hydrodynamic inter-action of the risers is minimal, provided they are separated bya distance greater than Þve times their individual diameters,unless there is a large enough riser array to create wake syn-chronization or other ßow disruptions.

8.2.3.6.3 In an integral bundle, the riser pipes are con-nected together at short intervals, (e.g., at intervals of about10 meters), so that they all move as one unit. The analysis ofsuch bundles can be carried out by suitably combining theindividual riser line properties and treating the bundle as anequivalent single pipe. It should be noted that in bundles withrisers with unequal properties the total tension will be distrib-uted based on the axial stiffness of the individual risers. Also,unsymmetric bundle arrangement will produce unsymmetri-cal hydrodynamic loads, which might lead to torsional rota-tion of the riser bundle. In modeling such bundles, thefollowing is recommended:

a. The overall motion of the bundle is compared to thatexpected from individual risers.

b. The relative motions of the individual risers in a bundle areassessed so that the possibilities of riser entanglement andexternal wear are minimized.

c. The distribution of the tension at the terminal points isevaluated (for preliminary design it can be conservativelyassumed that the largest pipe in the bundle takes all the load).

8.2.3.6.4 The global analysis requirements for multiborerisers are the same as for standard risers.


8.2.4 Service Life Analysis

8.2.4.1 General

8.2.4.1.1 Pipe design service life is to be speciÞed anddocumented. Design service life may be based on speciÞcproject or application duration or may be related to a replace-ment program. Consideration is to be given in the design ofßexible pipe to service life or replacement of components/ancillary equipment as part of an overall service life policy.

8.2.4.1.2 SpeciÞcation of pipe service life may also berelated to an in-service inspection program. The inspectionmethod and inspection interval are to be documented and jus-tiÞed with respect to suitability for the speciÞc application(see Section 13).

8.2.4.1.3 Evaluation of service life should address the fol-lowing as a minimum:

a. Metallic material corrosion and other failure modes (SSC,HIC, erosion, hydrogen embrittlement).

b. Wear of metallic material.

c. Fatigue of metallic material.

d. Polymer/elastomer material degradation.

e. Wear/abrasion of polymers/elastomers.

f. End Þtting design.

8.2.4.1.4 The wear and fatigue failure modes are generallyonly applicable to dynamic applications. The metallic materi-als may be selected so as not to corrode or alternatively thecorrosion rate is calculated based on the predicted annulusenvironment and accounted for in the pipe design. Corrosionfatigue tests may need to be performed for the armor wires.Other potential failure modes, including SSC, HIC, erosionand hydrogen embrittlement, should be accounted for bymaterial selection, with reference to the requirements of Sec-tion 6.2.4 of API SpeciÞcations 17J/17K.

8.2.4.1.5 Wear and fatigue in the metallic layers is dis-cussed in 8.2.4.2. Polymer/elastomer layer degradation andwear/abrasion of polymers/elastomer is accounted for bymaterial selection for the speciÞed application and by ageinganalysis/testing (see 5.4.1.10 and 5.4.2.9 for recommenda-tions on permissible levels of degradation and Section 6 forguidelines on material selection and ageing tests). The end Þt-ting should be designed to comply with the requirements ofAPI SpeciÞcation 17J/17K, with particular emphasis beingplaced on material selection and fatigue analysis.

8.2.4.2 Fatigue and Wear Analysis

8.2.4.2.1 Flexible pipes are complicated structures, partic-ularly from a fatigue and wear point of view. For each type ofpipe, there are several potential fatigue and wear mechanismsthat may be critical. Therefore, each application should be

carefully evaluated, particularly for riser applications. Fatiguecalculations for ßexible risers involve substantial uncertain-ties because of simpliÞcations in the long-term load data andmathematical models, and complexities in the wear andfatigue processes. An in-service condition and integrity moni-toring program should be implemented (see Section 13) ifappropriate.

8.2.4.2.2 For the tensile armor wires, potential failuremechanisms include the following:

a. Wear between layers and strands of individual cables.

b. Fatigue of armor wires.

c. Fretting fatigue of individual wiresÑunbonded pipe.

d. Wear or fretting between strands within cables.

8.2.4.2.3 In bending of an unbonded ßexible pipe, thearmor layers will slide over each other, with a resulting poten-tial for wear. The wear rate is a function of the contact pres-sure, wear coefÞcients, and degree of slippage (bendingrelated). Models have been developed using experimentallyderived data to simulate this failure mode [35, 36]. However,this problem has generally been overcome in current designsby the use of polymer/elastomer anti-wear layers between thearmor layers. The service life analysis should conÞrm thefunctional performance of this layer for the speciÞed designlife, particularly for high temperature applications. Similarsliding occurs between the individual strands in reinforcingcables.

8.2.4.2.4 The fatigue analysis should show that theextreme stresses in the tensile armors of unbonded and rein-forcement layers of bonded pipe are below the materialendurance limit (Goodman line in Figure 23), or else fatiguedamage calculations should be performed. The Haigh dia-gram should be based on relevant test data and shouldaccount for material properties, wire sizes and shapes, andservice environment. An example of a classical Haigh dia-gram is Figure 23, showing the fatigue and non-fatigueregions.

Figure 23—Example of Haigh Diagram

Fatigue region

Mean stress

Goodman line

Nonfatigue region

Alte

rnat

ing

stre

ss

σa

σm


8.2.4.2.5 Fatigue damage calculations may be based on alimited number of seastate classes, provided selection of suchclasses is based on conservative criteria. See 5.5.5 for guide-lines on selection of load cases for fatigue analysis. Fatiguelife may be calculated based on the S-N fatigue approachunder the assumption of linear cumulative damage. The S-Ndata should be derived based on the requirements of Section6.2.4.5 of API SpeciÞcation 17J and Section 6.2.4.4 of APISpeciÞcation 17K. Calculations should be performed for allcritical locations in the riser, such as at connection points andin the sag bend region based on combinations of mean andalternating stresses.

8.2.4.2.6 Conditions leading to fretting fatigue may causea large reduction in fatigue strength of individual armor wiresor cables, particularly in the low stress/long life region. In theHaigh diagram for armor wires, the Goodman line under fret-ting conditions would be considerably lowered. The potentialfor fretting fatigue should therefore be the subject of closescrutiny.

8.2.4.2.7 Fretting fatigue cracks are nucleated at the stick/slip interface, primarily by the oscillating tangential (friction)force transmitted in the stick region. Important parametersinclude surface reactions (oxidation and other environmentalinteractions), water ingress (a result of damage to the outersheath), and lubrication. When cracks reach a length of about1 mm, the crack driving force of the tangential stresses hasdecayed. In the absence of normal stresses in the wire, thecracks may become arrested at that point. With oscillatingnormal stresses, the cracks may continue to grow, and the netresult is a signiÞcant reduction in fatigue life, particularly inthe low stress/long life region. This emphasizes the require-ment for dynamic axial stresses in prototype fatigue tests.

8.2.4.2.8 Interlocked pressure armor may also fail fromfatigue, fretting fatigue, or wear, and therefore this potentialfailure mode should also be addressed in the service life anal-ysis. Note that a single fracture of the pressure armor wiremay be critical for the whole pipe. Theoretical models may beused to predict the service life of the interlocking proÞle.These models should be validated by experimental testresults. The primary loading parameters to be considered forthe pressure armor are as follows:

a. Static stress and contact pressure from internal pressureand axial tension.b. Dynamic stresses, sliding, and friction forces as a result ofbending.c. Combined effect of corrosion, wear, and fatigue.

8.2.4.2.9 A critical parameter in pressure armor fatiguecalculations is the residual stress in the wires after preform-ing. The residual stress should be accurately assessed, e.g., bylocal FE analysis. If fatigue in this layer is a problem, consid-eration may be given to taking account of the hydrotest effect

in changing the residual stress state of the wires, therebyimproving the fatigue performance of the layer. The test pres-sure should not cause stresses in the pipe above the criteriadeÞned in API SpeciÞcation 17J/17K. The manufacturershould have documented test results for the formed wires toverify the improvement in structural strength.

8.2.4.2.10 In addition to the armor layers, fatigue analysisof end Þttings and connectors should be performed where rel-evant. The analysis should be based on standard methodolo-gies and account for all relevant fatigue loads (the load casesfrom the fatigue analysis of the armor layers may be used).

8.2.5 Component Analysis

8.2.5.1 Where practical, all ancillary components of theßexible pipe system should be included explicitly in the glo-bal analysis at the detailed design stage. This includes buoy-ancy modules, subsea arch/buoy systems, tethers, bendstiffeners, etc. In addition, local analysis of the individualcomponents may need to be performed.

8.2.5.2 Components in a pipe system are to be designedwith regard to the same design parameters as the ßexiblepipe, including load cases (global loads and service condi-tions), and service life. Components should be designed inaccordance with recognized codes and standards, with refer-ence to the design guidelines in Section 7.5.

8.2.5.3 Component interference, which refers to the rub-bing together or impact of system components, is alsoincluded in component analysis. The interaction betweenpipes in a bundle system is one potential interference prob-lem. Possible impact between system components, such asbetween buoys or chains and risers, is another potential prob-lem. See 7.4.2 for guidelines on interference issues.

8.2.5.4 For certain ßexible riser conÞgurations, possibleÒweathervaningÓ of the subsea buoy/arch system is a criticalaspect. Generally, care must be used to ensure that unsym-metrical hydrodynamic loads do not cause the buoy/arch sys-tem to weathervane and twist the riser beyond acceptablelevels. The riser conÞguration and buoy/arch system shouldbe designed to avoid this problem.

8.3 LOADS

8.3.1 Hydrodynamic Loads

8.3.1.1 Wave Kinematics

8.3.1.1.1 In the derivation of hydrodynamic forces, it is Þrstnecessary to deÞne the wave-induced water particle velocitiesand accelerations, i.e., the wave kinematics. Common practiceis to model the wave using linear Airy wave theory. In somecases, particularly with shallow water, a nonlinear theory, suchas StokeÕs Þfth order wave theory, may apply.


8.3.1.1.2 Note that linear wave theory only calculates thekinematics for inÞnitesimal wave heights. Stretching tech-niques are available to extend the theory to Þnite waveheights. The riser response is generally not sensitive to thestretching theory, except possibly for shallow water. SigniÞ-cant ampliÞcation of the wave kinematics can occur adjacentto large structures, such as the columns of a semi-sub, andwhere relevant this may need to be considered in the riserdesign. One method for modeling this ampliÞcation is to useincreased hydrodynamic coefÞcients at the relevant location.

8.3.1.2 Morison’s Equation

8.3.1.2.1 The general practice for modeling of hydrody-namic forces on ßexible pipes is to use the Morison formula-tion, which is largely empirical based. The formula [32] wasoriginally derived for calculating the hydrodynamic forces onvertical, shallow water, Þxed piles with only wave loading. Ithas since been extended to apply to arbitrary orientation mov-ing structures such as risers, with both wave and current load-ing. The transverse Morison load per unit length [ ] due toßuid-structure interaction is typically written as follows:

(6)

where

= is the normal relative ßuid velocity, i.e., the rela-

tive ßuid/structure velocity in the transverse direction,

= is the normal water particle acceleration,

= is the normal structural acceleration,

D = is the effective drag diameter,

CD = is the drag coefÞcient,

Cm = is the inertia coefÞcient.

= is the seawater density.

8.3.1.2.2 This formulation represents the most commonlyused extension of the original MorisonÕs Equation. A numberof comments are appropriate here.

a. Firstly, the inertia component of the original MorisonÕsEquation is replaced by two terms, one proportional to thenormal water particle acceleration, the other to the normalstructure accelerationÑthis is because the inertia force on amoving cylinder in a wave Þeld comprises a hydrodynamicÒadded massÓ term representing the additional inertia or resis-

tance to motion due to the ßuid ÒentrainedÓ with the movingmember, in addition to the force on a stationary member in anaccelerating ßuid (the term in the original Morisonformulation).

b. Secondly, in the drag term the ßuid velocity is replaceddirectly by the relative ßuid structure velocity (including cur-rent). The validity of this is open to question, but thisapproach is in widespread use.

8.3.1.3 Limitations to Morison’s Equation

The following comments are relevant to the formulation ofEquation 6:

a. For the inertia force term, the acceleration of the ßuid ßowis evaluated at the centerline of the riser. Therefore, higherorder convective acceleration terms are neglected.

b. The inertia, added mass, and drag coefÞcients are timeinvariant. Time varying parameters may be used; generallysufÞcient data is not available.

c. The hydrodynamic forces are determined by the accelera-tion and velocity components normal to the riser centerline.The three-dimensional from the axial component of the inci-dent ßow can be accounted for by calculating a tangentialdrag force as a function of the tangential velocity squared.

d. The riser response is in-line with the incident ßow. The liftforce is omitted. The ßuctuating lift and drag forces as aresult of vortex shedding are generally neglected. For shortjumpers or ÒtautÓ conÞgurations, however, vortex sheddingresponse should be taken into account

e. The force on a member in close proximity to another isaffected by the wake Þeld due to interference and shieldingeffects. It is possible that the wake of the Þrst memberdynamically excites the member behind it. Conversely, it ispossible that an adjacent large member shields a smallermember and leads to a reduction in hydrodynamic force.These effects, which in general inßuence only the drag forcecomponent, are difÞcult to incorporate into MorisonÕsequation.

f. If several risers are close together, there is a tendency for aproportion of the mass of ßuid enclosed collectively by themto act as part of the structure. This leads to increased ÒaddedmassÓ forces, which may be modeled empirically by increas-ing Cm and also modiÞes the inertia forces, which should notbe changed.

8.3.1.4 Drag and Inertia Coefficients

8.3.1.4.1 The drag (CD) and inertia (Cm) coefÞcientsincorporated into MorisonÕs formulation are empirical coefÞ-cients which have been derived from a large body of reportedexperiments. These experiments have shown good agreementbetween measured forces and forces calculated from Mori-sonÕs Equation.

f m

÷

f m

÷

12--- ρw DCD V

÷RN V

÷RN ρw

πD2

4---------- Cm V

÷ú

WN+=

ρwπD2

4---------- Cm 1Ð( )V

÷ú

PNÐ

V÷

RN

V÷ú

WN

V÷ú

PN

ρw


8.3.1.4.2 In theory, the drag and inertia forces are a func-tion of ReynoldÕs number, the Keulegan-Carpenter number,structure geometry, and surface roughness, and strictly shouldbe considered as varying along a member and with time. Inpractice, this would render hydrodynamic force computationsimpractical, and a constant coefÞcient is invariably used inriser analysis. This introduces a considerable source of uncer-tainty in the accuracy of results.

References [5], [24], and [37] provide recommendationson the selection of drag and inertia coefÞcients for ßowlinesand risers. In ßexible pipe analyses, Cm is usually taken to be2.0, while CD varies between 0.7 and 1.2. It is recommendedthat sensitivity studies be performed to investigate the effecton global analysis results of the selected coefÞcients. Notealso that the selection of hydrodynamic coefÞcients for largesystem components, such as buoyancy tanks, can be criticaland should be carefully evaluated. Consideration should alsobe given to the potential effect of VIV and marine growth onhydrodynamic coefÞcients.

For wave type riser conÞgurations, which use distributedbuoyancy modules, the buoyancy section will be subject tosigniÞcant tangential as well as transverse hydrodynamicforces. Huse [38] gives some recommendations on the selec-tion of tangential hydrodynamic coefÞcients for buoyancymodule riser sections.

8.3.2 Gravity and Buoyancy Loads

The analysis should include the gravity and buoyancy loadsfrom all components of the system, including ßexible pipe,buoys, clump masses, etc. Consideration should also be givento loads resulting from marine growth and ice accumulations.

8.3.3 Internal Fluid Loading

The mass of the internal ßuid should be included in allanalyses. Variation in the density should be considered. Notethat changes in the internal ßuid density over the design lifemay signiÞcantly affect some riser conÞgurations, particu-larly the wave conÞgurations. For some applications, it alsomay be necessary to consider the effect of slugs (liquid andgas) on the system. The loads induced by slugs, which shouldbe accounted for in the analysis, are gravity, inertia, centrifu-gal forces, and Coriolis loads.

8.3.4 Seabed and Soil Interaction Loads

The effects of the seabed, including frictional loads,should be included where relevant. In particular, these willbe required for ßowline stability analyses and motion analy-sis for riser sections lying on the seabed (lazy conÞgura-tions). Reference [5] lists representative soil stiffness andfriction coefÞcients for ßexible pipes in contact with the sea-bed. The soil stiffness and friction coefÞcients are repro-duced in Table 18.

8.3.5 Temperature and Pressure Loads

Temperature and pressure induced elongation are generallyonly a concern in trenched ßowlines where there is a possibil-ity of upheaval buckling. In addition, short unbonded jumperßowlines may experience signiÞcant compression loads fromtemperature and pressure effects, in which case the pipe mayneed to be reinforced with additional polymer layers to pre-vent birdcaging.

8.3.6 Vortex Induced Loads

8.3.6.1 The sensitivity of ßexible risers to vortex sheddinghas been the subject of a number of experimental investiga-tions, which have shown that though VIV occurred in themodeled risers, the vibration amplitudes were insufÞcient tocause fatigue damage. This can be attributed to the following:

a. Relatively low vibration amplitudes, probably a result ofthe inherent structural damping.

b. The complexity of ßow incident to typical ßexible risersystems and difÞculty in obtaining coherence of vortices in aheaving inclined riser.

c. Hydrodynamic damping.

8.3.6.2 Many of the contributory factors to VIV are difÞ-cult to model accurately in small-scale tests. In full scale,especially with deep water risers, the effects of VIV maybecome more signiÞcant due to the following [39]:

a. Increased tension-reducing inßuence of structuraldamping.

b. Increase in hydrodynamic drag coefÞcients from VIV.

c. Strong currents present in some deep water regions.

8.3.6.3 As a result of the above, the effects of VIV on boththe structural strength of components and on riser global

Table 18—Typical Soil Stiffness and Friction Coefficients for Flexible Pipes [5]

SeabedType Direction

Stiffness(kN/m2)

FrictionCoefÞcient

Clay Axial 50Ð100 0.2

Lateral 20Ð401 0.2Ð0.43

Vertical 100Ð50001 Ñ

Sand Axial 100Ð200 0.6

Lateral 50Ð100 0.8

Vertical 200Ð100002 Ñ

Notes:1Value increases with increasing undrained soilshear strength.2Value increases with increasing soil density.3Value increases with decreasing soil shear strength.


behaviour, particularly with respect to the potential for inter-ference, should be reviewed on a case-by-case basis. Currentpractice is to conduct analysis with increased effective hydro-dynamic cross-section to account for vortex-induced loading.

8.4 GLOBAL RESPONSE EVALUATION

8.4.1 Design Wave and Design Storm

8.4.1.1 The objective in performing dynamic analyses is topredict the lifetime maximum or extreme response of theßexible pipe system. The two approaches commonly used forthis purpose are design wave and design storm analyses.

8.4.1.2 The design wave (or regular wave) approach isbased on a deterministic seastate description of the waveenvironment using a single wave height and period to modelthe seastate. These parameters are derived using wave statis-tics or simple physical considerations. The advantages of theapproach is that the response calculation is straight forward,periodic input generally giving periodic output with no fur-ther requirement for statistical post-processing. The methodis often reasonable; for ßexible risers, the design wave willrepresent the extreme seastate with reasonable accuracy.

8.4.1.3 In the design wave method, consideration shouldbe given to performing analyses for a number of wave periodsto identify the critical system responses for both short andlong wave periods. For example, the short period may givethe critical loads at the vessel connection, while the longperiod may give larger motions in subsea buoy systems.

8.4.1.4 The limitation of the design wave approach is thatits use is uncertain in systems whose response is stronglydependent on frequency, based on uncertainties in the choiceof the design wave. It is often impossible to determinewhether the result is conservative or unconservative, particu-larly in the case of ßexibles where conventional methods andsoftware for the estimation of eigenfrequencies contain sig-niÞcant uncertainties. In such situations, the use of the designstorm approach may be necessary.

8.4.1.5 The design storm or irregular sea approach is basedon a stochastic description of the wave environment. The sea-state is modeled as a wave spectrum with energy distributedover a range of frequencies. The most common spectra usedare the Pierson-Moskowitz (fully developed sea) and theJONSWAP (developing sea) spectra. The response in thiscase is also stochastic, and statistical post-processing is nec-essary to identify the design value of the response. A three-hour design storm duration should normally be considered.

8.4.1.6 If a full-three-hour simulation is not performed, theduration of the simulated wave record should not be less than30 minutes, provided the generated sea state is qualiÞed withrespect to theoretically known statistical properties of a Gaus-sian process. The extreme response for the design storm

should be found by using a recognized, most probable maxi-mum extrapolation technique.

8.4.2 Formulation of Equations of Motion

8.4.2.1 The formulation of the motion equations for solu-tion of global response analyses involves consideration of thefollowing main issues:

a. 2D versus 3D response.

b. 3D wave kinematics.

c. Use of small angle versus large angle theory.

d. Modeling of intermittent seabed contact and frictioneffects.

8.4.2.2 A simpliÞcation for some riser analyses is the useof planar (two-dimensional) analysis in which vessel motion,waves, current, and any initial displacement of the riser are allassumed to be in the same plane. For many cases, especiallyfor initially straight (vertical) risers, this is an adequateassumption that can signiÞcantly reduce the resourcesrequired for a single analysis. Planar analysis is therefore use-ful for preliminary design work.

8.4.2.3 Spread seas and noncollinear wave and currentloads cannot be solved directly with two-dimensional tech-niques. In some cases, reasonable approximations will stillpermit the use of two-dimensional formulations. However,certain problems are inherently three-dimensional and there-fore require a three-dimensional analysis. This is generallythe case for ßexible risers.

8.4.2.4 The Òsmall angleÓ assumption has been used forformulating some riser analysis methods, particularly for ver-tical rigid risers. Use of the small angle theory simpliÞes thesolution through approximation of the curvature term, whichlimits its use to cases where the maximum angle change isless than 10 degrees. A large angle formulation must be usedfor analyses where the maximum angle change is greater than10 degrees, which is typically the case for ßexible risers sub-jected to extreme loading conditions. A number of large angleformulations are described in the literature [40, 41].

8.4.2.5 Interaction of the seabed with ßexible pipes is animportant consideration in global analysis. The verticalrestraint of the seabed may be modeled as a rigid surface oran elastic foundation. Use of either method should be evalu-ated for the particular application; in general, the rigid surfacemodel is satisfactory. This is dependent on the coefÞcient ofelasticity of the seabed soil. If a riser is strongly impactingwith the seabed, the analysis should be able to accurately sim-ulate the nonlinear behaviour.

8.4.2.6 The axial and lateral resistance to movement of thepipe at the soil interface can be modeled by a constant frictionmodel or a hysteretic model. In a hysteresis model, the fric-


tion force is gradually built up as the pipe slides on the sea-bed, up to the maximum value depending upon the normalforce and the friction coefÞcient; if the movement is reversed,the build-up starts in the opposite direction. However, thehysteretic model is difÞcult to apply in practice because thehistory of deformations is required. For this reason, a constantfriction may be used if a proper hysteretic model is not avail-able. In this case, it is recommended that the accuracy of theresults be evaluated using sensitivity studies.

8.4.2.7 The equations of motion are differential equationsand therefore do not have a closed formed solution. Theselection of appropriate solution methods will therefore becritical for efÞcient analyses.

8.4.3 Solution of Equations of Motion

8.4.3.1 Spatial Solution

8.4.3.1.1 Spatial solution of the equations of motion maybe based on analytical techniques (generally not applicable toglobal analysis of ßexible pipes) or approximate numericalmethods. The numerical methods used may be either Þniteelement or Þnite difference based.

8.4.3.1.2 A numerical solution to the equilibrium equa-tions is typically obtained by assembling equations for eachelement comprising the riser into a system of equationsdescribing the force displacement relationships for all degreesof freedom (DOF). By combining all equations for elementsconnected to a particular node, in a manner consistent withrequirements for equilibrium at the node and compatibilitybetween elements, equations relating forces at all globalDOFs to displacement at each DOF at the node are obtained.Assembling all such equations for N global DOFs leads to asystem of N coupled algebraic equations. These equations canbe expressed in matrix form as follows:

(7)

where [M], [C] and [K] are respectively the mass, dampingand stiffness matrices, {R} is the load vector, and , ,

are the acceleration, velocity, and displacement vectors,respectively.

8.4.3.2 Temporal Solution

8.4.3.2.1 Frequency Domain

Frequency domain analysis can be used if there are no non-linearities that signiÞcantly affect the system response. Fre-quency domain may be used for fatigue analysis, as it allowsfor reasonable statistical estimates of forces in the pipe. Thelinear fatigue analyses should generally be combined withnonlinear static analyses for ßexible riser systems.

The principal advantage of frequency domain analysis is areduction in computational effort for linear systems, coupledwith very simple, unambiguous output. Analysis of linearsystems is well understood, and the application of frequencydomain results to design criteria for truly linear systems isstraightforward. The limitation of frequency domain analysisare the difÞculties and added complexities associated withmodeling nonlinear behaviour. This generally invalidates thetechnique for use in large displacement ßexible riser analyses.

There are several applications of the method to riser analy-sis in the literature, though most apply to rigid riser analysis.Reference [42] describes the application of the method toßexible riser analysis. Important considerations in frequencydomain analysis include proper linearization of the wave andcurrent drag forces, and careful selection of analysis frequen-cies. Frequencies used in the analysis should result in ade-quate deÞnition of the wave energy spectrum, vessel responsecharacteristics, and natural frequencies of the riser.

8.4.3.2.2 Time Domain

Time domain analysis is generally required for ßexibleriser design, where accurate representation of the nonlinearbehaviour is important. Nonlinear effects encountered in ßex-ible riser analyses, including large deformations, nonlinearloads, and seabed interaction, can be directly modeled in thetime domain. Time domain can also be used to assess the rel-ative accuracy of equivalent frequency domain analyses andcalibrate them for use in design.

Analysis in the time domain requires deÞnition of environ-ment and applied loading, such as vessel motions, as a func-tion of time, typically by simulating wave time histories.Time domain analysis essentially requires solution of equilib-rium position at discrete points in time, by considering iner-tia, damping, and applied loads.

The equilibrium equation may be solved by implicit orexplicit integration methods. Explicit methods solve forresponse at t + ∆t based on equilibrium conditions at time t.Implicit methods solve for response at t + ∆t based on equi-librium at time t + ∆t. This has implications on the numericaleffort required to perform the integration. Explicit methodstypically require fewer computations per time step but oftenrequire shorter time steps to achieve an accurate solution.Implicit methods often require substantial numerical effort ateach time step (like decomposition of the coefÞcient matrix)but can often utilize larger time steps and are more typicallyused for ßexible riser analysis.

All methods have some degree of integration error that isassociated with frequency and amplitude of the integratedresponse. In certain situations, slight errors in frequencyalone can accumulate and lead to numerical ÒbeatingÓ of theresponse. It is important to recognize and understand theseerrors when performing time domain analysis, particularly forthe purpose of simulating long time histories and developingstatistics for extremes.

M[ ] xúú{ } C[ ] xú{ } K[ ] x{ }+ + R{ }=

xúú{ } xú{ }x{ }


8.4.3.2.3 Modal Analysis

A modal analysis may be performed to determine the modeshapes and natural frequencies of the system, in particular forshort jumpers or taut conÞgurations. An important consider-ation in modal analyses is the modeling of nonlinearities,such as the effect of the seabed in lazy riser conÞgurations.

8.4.4 Modeling Considerations

8.4.4.1 Model Discretization

8.4.4.1.1 Finite element or Þnite difference techniques aretypically employed to reduce the differential equilibriumequations to a set of coupled algebraic equations that can besolved numerically. Discretization of the riser must be donecarefully to avoid numerical errors resulting from too coarse amesh while producing a model that can be analyzed with areasonable amount of computational effort.

8.4.4.1.2 The level of discretization that is ultimatelyacceptable depends on the numerical representation of ten-sion variation, the spatial variation in physical properties ofthe riser, the magnitude of applied load, the frequency contentof the applied load, and the accuracy of the desired results. Ingeneral, coarser meshes are acceptable for determiningapproximate displacement solutions to problems dominatedby vessel motions, while Þner meshes are essential for accu-rately determining stresses in the splash zone or at discontinu-ities, such as support points.

8.4.4.2 Frequency Content Selection

8.4.4.2.1 For irregular sea (design storm) analyses it isimportant that the frequency content of the input seastatespectrum is accurately represented. The following commentsapply:

a. Total spread of frequencies should cover all frequencieswith signiÞcant energy.

b. The discretization of the spectrum (i.e., number of fre-quencies used) should accurately represent the seastate. Thediscretization may be based on an equal area approach or anequal frequency increment approach. The equal areaapproach is recommended.

8.4.4.2.2 For time domain analyses, the seastate spectrumis synthesized into a wave time history and may be achievedby a number of methods, including Monte Carlo and digitalÞltering approaches. The realized spectrum (from the timehistory) should be compared to the input spectrum for accu-racy of the synthesis method.

8.4.4.3 Time Step Selection

8.4.4.3.1 The time step used for a time domain analysiswill depend on the solution methodology and software pro-

gram. All methods require that the time step be small enoughto accurately reßect important frequencies in the load orresponse. This is analogous to proper spatial discretization ofthe model and careful selection of frequencies in the fre-quency domain method. Large time steps may result in aquicker analysis that is accurate for the frequencies repre-sented but may miss important high frequency contributions.

8.4.4.3.2 The time stepping scheme used may be based onÞxed or variable steps. Fixed steps are recommended; vari-able time steps, however, can result in signiÞcantly less com-putational effort. Results from variable time step analysesshould be checked to ensure that changes in the time step donot induce numerically spurious values.

8.4.5 Effective Tension

Effective tension is an important parameter in riser analy-sis, though it is a subject of much debate. The equation foreffective tension (Te) is as follows:

Te = Ta + Po.Ao Ð Pi.Ai (8)

where

Ta = axial (true wall) force,

Pi, Po = internal and external pressures,

Ai, Ao = internal and external cross-sectional areas of pipe.

8.4.5.1 Effective tension has a real effect on the displace-ment of a tensioned beam and it is often convenient to treat Teas a physical quantity. Effective tension, however, is not aphysical tensile force, nor is it an internal force of any kind.Effective tension is a grouping of applied load terms withinthe equation of motion. Dynamic analysis results normallyreport the effective tension not the true wall tension.

8.4.5.2 It is important to understand this distinction whenformulating the analysis model as well as to avoid misinter-preting results of typical riser analyses. For example, lateralforce at any cross-section of a riser is equal to shear plus theeffective tension times the slope. This calculation is validonly because it is equivalent to integrating pressure aroundthe tube circumference and adding shear and the lateral com-ponent of tension. Detailed discussions on effective tensionare provided in [43] and [44].

8.4.5.3 Low-or even negative-effective tension over a por-tion of the riser does not imply the riser is unstable, nor doesit cause the riser to instantaneously experience Euler buck-ling. The direct consequence of low or negative effective ten-sion is low lateral stiffness, the result of which is adequatelyestimated by the standard global riser analysis if changes ineffective tension are accounted for [45]. Any effective com-pression which occurs should be shown to be tolerable for thepipe (see the design criteria in Section 5).


9 Prototype Testing

9.1 GENERAL

9.1.1 Scope

9.1.1.1 This section gives guidelines on the requirementsfor prototype tests and presents procedures for performingthese tests. Refer to API SpeciÞcation 17J/17K for factoryacceptance and material test requirements. A prototype test isdeÞned as a test carried out in order to establish or verify aprincipal performance characteristic for a particular pipedesign, which may be a new or established design.

Prototype test documentation is intended to be reviewed bythe independent veriÞcation agent as part of the pipe designmethodology veriÞcation (see Section 5.2 of API SpeciÞca-tion 17J/17K).

9.1.1.2 The requirements for prototype testing are subject toagreement between the manufacturer and the purchaser, andmay be based on the recommendations given in this section. Asan alternative to prototype testing, the manufacturer may pro-vide objective evidence that the product satisÞes the designrequirements. Objective evidence is deÞned as documentedÞeld experience, test data, technical publications, Þnite elementanalysis (FEA) or calculations that verify the performancerequirements, may be used if the envelope of applications foran established design is proposed to be marginally extended.

9.1.1.3 The number and range of prototype tests that canbe performed on ßexible pipe is extensive. Prototype tests aregenerally destructive and are therefore expensive to under-take. Cost and/or time implications make it impossible to per-form a full range of prototype tests for each pipe design.

9.1.1.4 For high temperature applications, the design of theend Þtting sealing mechanism for unbonded pipe is critical.The test procedures currently being used are given in Appen-dix A and B for both static and dynamic applications. Notethat these protocols may be superseded based on results offuture tests.

9.1.1.5 It should be noted that a selected group of tests forqualiÞcation of a prototype design will normally includematerial and FAT tests, as speciÞed in Sections 6 and 9respectively of API SpeciÞcations 17J/17K.

9.1.2 Design Programs

9.1.2.1 As a minimum, Class I prototype testing is recom-mended for new or unproven ßexible pipe designs. The objec-tives of prototype testing should be twofold, as follows:

a. Prove or validate new or unproven pipe designs.

b. Validate the manufacturersÕ design methodology for a newpipe design.

9.1.2.2 The second objective will increase the level of con-Þdence in the design methodology and thereby reduce the

requirements for prototype testing in the future. The require-ments for the manufacturersÕ design methodology are speci-Þed in Section 5.2.1 of API SpeciÞcation 17J. The designmethodology should provide a conservative estimate of thefailure load for the particular prototype test. A conÞdencelimit should be established by which the design methodologycan be shown to be conservative.

9.1.2.3 Fundamental to reducing prototype test require-ments is the necessity to increase conÞdence levels in thedesign methodology. All tests performed should therefore beused to validate the design methodology and so minimizefuture requirements for prototype testing. It is fully permissi-ble to use validated analytical approaches to perform extrapo-lations from relevant tests, taking parameter variations intoaccount, subject to the recommendations of this section.

9.2 CLASSIFICATION OF PROTOTYPE TESTS

Prototype tests are classiÞed into three classes as follows:

a. Class IÑStandard prototype tests, as most commonlyused.

b. Class IIÑSpecial prototype tests, used regularly to verifyspeciÞc aspects of performance such as installation or operat-ing conditions.

c. Class IIIÑTests used only for characterization of the pipeproperties.

Tests that come under these classiÞcations are listed inTable 19. The loading used in the dynamic fatigue test listedas a Class II test may be single or combined loading. Theselection will depend on the application; a combined bendingand axial test is recommended.

Procedures for Class I and II tests are given in Sections 9.5and 9.6 respectively. Procedures for Class III tests should beas per the speciÞcations of the purchaser or manufacturer.

9.3 TEST REQUIREMENTS

9.3.1 General

The requirements for prototype tests should considerwhether the pipe is a new design or new application, and whatare the critical failures modes and consequences. In addition,scaling limitations and applicable tests should be addressed.These are discussed in the following sections.

9.3.2 New Pipe Design or Application

9.3.2.1 A new pipe design is deÞned by a substantivechange or modiÞcation to one of the following:

a. Pipe manufacturing process (structural layers, internalpressure sheath or end Þtting).

b. Pipe structure.

c. Pipe application.


9.3.2.2 Critical issues related to pipe structure and applica-tion are identiÞed in Tables 20 and 21, respectively, togetherwith recommendations on prototype test requirements. Therequirements for prototype testing of a new design are verydependent on the application and this should be considered.For example, there is a large difference between a low pres-sure static ßowline and a high pressure riser application.

9.3.3 Failure Modes

The requirements for prototype tests should consider thecriticality and consequences of pipe failure. In particular,potential defects, the consequences of these defects, andcauses should be identiÞed. The major potential defects, inunbonded ßexible pipes are identiÞed in Section 13.3. Criti-cal prototype tests which may be used to verify the pipedesign for some of these potential defects and failure modesare identiÞed in Table 22. This table should be referred towhen determining prototype test requirements.

9.3.4 Scaling Limitations

9.3.4.1 Scaling of previous test results may be used to ver-ify the members of a product family in accordance with theguidelines of this section. Flexible pipe product families arelisted in Tables 1 and 2. For scaling purposes, the pipe designprinciples and functional operation should be similar. In addi-tion, the design stress levels in relation to material mechanicalproperties should be based on the same criteria, i.e., equiva-lency in utilization or accumulated fatigue damage. The fol-lowing scaling limitations are recommended:

a. PressureÑthe test pipe may be used to qualify pipes of thesame family having equal or lower pressure rating.

b. Internal DiameterÑtesting of one pipe of a product familyshould verify products two inches larger or smaller than thesize tested.

c. TemperatureÑthe temperature range and number of cyclesveriÞed by the test product should verify all temperatures that

Table 19—Classification of Prototype Tests

Class Type Description Test Condition/Comment

I Standard Prototype Tests a) Burst Pressure Test Typically in straight line.

b) Axial Tension Test At ambient pressure.

c) Collapse Test With outer sheath perforated or omitted.

II Special Prototype Tests a) Dynamic Fatigue Test Bending, tension, torsional, cyclic pressure, rotational bending or combined bending and tension fatigue tests.

b) Crush Strength Test Installation test.

c) Combined Bending and Tensile Test Installation test.

d) Sour Service Test To examine degradation of steel wires.

e) Fire Test

f) Erosion Test To examine degradation of carcass.

g) TFL Test Also includes pigging test.

h) Vacuum Test Bond strength in test for bonded pipes

i) Kerosene Test Detect permeation or leakage of highly seeking hydrocar-bon through liner of bonded pipe.

j) Adhesion Test Verify bond strength of bonded pipe

k) Full Scale Blistering Test Determine suitability of bonded pipe to gas service.

III Characterization and Other Prototype Tests

a) Bending Stiffness Test To MBR (non-destructive).

b) Torsional Stiffness Test To allowable torque (non-destructive).

c) Abrasion Test Test for external abrasion.

d) Rapid Decompression Test

e) Axial Compression Test Upheaval buckling and compression capacity.

f) Thermal Characteristics Test Dry and ßooded conditions.

g) Temperature Test High and low temperature cycling.

h) Arctic Test Low temperature test.

i) Weathering Test UV resistance.

j) Structural Damping Test Characterization test.


Table 20—Recommendations for Prototype Tests—Modifications to Pipe Structure Design

No. Design ModiÞcation Recommendation on Requirement for Prototype Tests

1. Internal/External Diameter Probably not required. However, it may be necessary for large variations from previously qualiÞed designs to be veriÞed by prototype testing. See Section 9.3.4.

2. Number and Order of Layers Required for substantive change to structural layers only.

3. Metallic Layer Construction Required if cross-sectional shape or material type is substantially changed.

Material qualiÞcation required.

4. Polymer/elastomer Layer Material qualiÞcation tests only required.

5. Spiraling Angle Only required for angle (θ) changes outside the following, where θ is measured relative to longitudinal axis:

Ð Carcass or pressure armor (unbonded) layers: θ < 80¡

Ð Tensile armor (unbonded) and reinforcement (bonded) layers: 20¡ <θ< 60¡

6. End Fitting Required for substantive change to the end Þtting design, in particular:

Ð Change in armor /reinforcing layer anchoring system.

Ð Change in epoxy material.

Ð Change in internal/external ßuid integrity systems (sheath /liner anchoring).

7. Lubricant (unbonded) Not required. Material qualiÞcation is required.

8. Materials Generally sufÞcient for materials testing to be performed.

Note: The above recommendations may vary for different applications, such as ßowlines and risers.

Table 21—Recommendations for Prototype Testing—Changes in Pipe Application

No. Change in Pipe Application Requirement for Prototype Testing

1. Transported Fluid Generally not required. Compatibility to transported ßuid can generally be determined by material testing. However, for unusual transported ßuid conditions, prototype testing may be required. In par-ticular the following will require consideration for prototype tests:

Ð Sour service and corrosive environments

Ð High temperature applications

Ð High pressure applications

2. Service Life Not required for static applications, as material testing is generally more relevant. Not required for dynamic applications if previous testing can be extrapolated to the required service life.

3. External Environment Dependent on the environmental conditions. Not required if interpolation from previous tests can be performed.


fall entirely within that range, for the particular test ßuidcomponent.

d. Test FluidÑthe test ßuid should verify all products withthe same materials as the tested pipe.

9.3.4.2 The scaling comparison may also be made basedon pressure by internal diameter (P x ID), with the test pipequalifying pipes with a lower P x ID value, subject to theinternal diameter limitations.

9.3.5 Applicable Prototype Tests

9.3.5.1 This section describes the prototype tests which areapplicable to the design modiÞcations and application changeslisted in Section 9.3.2. The requirements for Class I and Class

II prototype tests, as deÞned by Table 19, are presented inTables 23 and 24 respectively. These requirements are subjectto the recommendations of Sections 9.3.2 to 9.3.4 inclusive.

9.3.5.2 Changes to transported ßuid, service life or exter-nal environment do not require Class 1 prototype tests, butmay require materials testing as in Section 6 of API SpeciÞ-cation 17J/17K.

9.4 TEST PROTOCOL

9.4.1 Test Sample

9.4.1.1 Prototype testing should be conducted on full sizeproducts that represent the speciÞed dimensions for the rele-vant components of the end product being veriÞed. This does

Table 22—Potential Flexible Pipe Failure Modes and Associated Critical Prototype Tests

Pipe Component Failure Mode Prototype Test

Carcass Layer 1. Collapse Failure Modes:

Ð Due to external pressure Collapse Test.

Ð Due to armor layer pressure Tensile Test.

Ð Due to installation loads Combined Bending and Tensile Test, Crush Strength Test.

2. Wear Erosion Test.

3. Material Failure Material Tests.

Internal Pressure Sheath or Bonded Pipe Liner

1. Rupture due to Pressure Burst Test

2. Creep Extrusion Burst Test and Temperature Test.

3. Material Failure Material Tests.

4. Wear Erosion Test.

5. Fatigue Dynamic Fatigue Test.

Structural Layers 1. Structural Failure due to Loading:

Ð Tension Tensile Test

Ð Compression Axial Compression Test

Ð Pressure Burst Test

2. Wear and Fatigue Dynamic Fatigue Test

3. Birdcaging Axial Compression Test

4. Adhesion/Delamination for Elastomers Adhesion Test

5. Material Failure Material Tests

Insulation Layers 1. Loss of Insulation due to Flooding Thermal Characteristics Test

2. Installation Crushing Loads Crush Strength Test

End Fitting 1. Pressure Sheath/liner Pull-Out Temperature Test

2. Armor/reinforcing Layer Anchoring Dynamic Test, Tension Test

3. Epoxy Failure Dynamic Test, Temperature Test

Note: Detailed lists of potential pipe defects are given in Section 13.3.


not apply to the length of the ßexible pipe, excluding end Þt-tings. Unless speciÞed in the test procedures of Sections 9.5and 9.6, the minimum length excluding end Þttings should bethe greater of 3 m or ten times the internal diameter. The testsamples should have been subjected to FAT testing.

9.4.1.2 The actual dimensions of pipe subjected to proto-type testing should be within the allowable tolerance rangefor dimensions speciÞed for normal production pipe. Wherepractical, these actual dimensions should represent the worstcase conditions. The sample should include any weak pointswhich may occur in the Þnal product. These include welds,repaired or damaged sections and process variations.

9.4.1.3 Test samples should represent the actual productto be supplied, considering both the design and manufac-turing procedures. If samples are made up using semi-man-ual procedures (i.e., not from a production run),consideration should be given to potential differencesbetween sample and production pipe. It may be necessaryto consider reproducing some of the critical test results onproduction samples to verify the manufacturing equipmentand procedures.

9.4.1.4 All tests should be carried out with end Þttingsmounted which are identical to those to be used on the prod-uct to be qualiÞed, except where recommended by this rec-ommended practice.

9.4.2 Test Equipment

Test equipment should conform to internationally recog-nized standards. All test equipment and instrumentationshould be calibrated on a regular basis, at least once a year.Current certiÞcation/calibration certiÞcates for all test equip-ment should be included in the test report.

9.4.3 Test Procedures

9.4.3.1 If tests require variables, such as temperature orpressure, to be constant, then the particular variable should bestabilized prior to commencement of the test. Stabilization isdeÞned as follows for pressure and temperature parameters:

a. PressureÑPressure variation in one hour is within ±1% ofthe test pressure.

b. TemperatureÑTemperature variation in one hour is within±2.5¡C of the test temperature.

9.4.3.2 When structure accommodation (bedding-in) canaffect the results, the necessity for pressure cycling the sam-ple prior to test start-up should be evaluated by the manufac-turer. For example, in a burst test where deformationmeasurements are required, a minimum of three cycles (fromzero to test pressure) is generally sufÞcient, performed asfollows:

a. First cycle for structure accommodations (bedding-in).

b. Second cycle for accurate measurements.

c. Third cycle to verify measurements from second cycle.

9.4.3.3 The load application requirements are different foreach test type, and are discussed in the individual test descrip-tions. The load applications rate should be representative ofthe load application rate applied under factory and Þeldacceptance testing, installation, and service conditions. Themaximum loading rate should not exceed 5 percent of theexpected maximum load per minute.

Table 23—Recommendations for Class I Prototype Tests

Design ModiÞcation orChange in Application

Recommended Class I Prototype Tests

Burst Tension Collapse

Internal/External Diameter X X X

Number or Order of Layers X X

Internal Carcass X

Internal Pressure X X

Pressure Armor Layer X X

Tensile Armor Layer X

Spiralling Angle X X

End Fitting Design X X

Table 24—Recommendations for Class II Prototype Tests

Design ModiÞcation or Change in Application

Recommended Class II Prototype Tests

New design or more severe dynamic loading conditions.

Dynamic Fatigue Test.

New installation system or water depth.

Crush Strength Test.

Installation of new design or deeper water using horizontal laying spread.

Combined Bending and Tension Test.

Sour service conditions. Sour Service Test.

Critical Þre protection require-ments and untested design.

Fire Test or calculated Þre sur-vival time conservatively calcu-lated by a method validated by previous Þre test results.

Severe sand production and severe consequences of failure.

Erosion Test.


9.4.4 Post-test Examination

Pipe dissection should be performed whenever a samplefails. Failure evaluations and abnormalities should bereported. All relevant items should be photographed. Theexamination document should include a written statementdescribing any defects which were found in the test sampleand whether or not these defects resulted in design criteriabeing violated.

9.4.5 Documentation

9.4.5.1 Before testing, the manufacturer should issue adetailed test procedure to the purchaser which should includethe following items as a minimum:

a. Type of tests to be performed.

b. Schedule and duration of tests.

c. Test descriptions (including sketches and equipment setup).

d. Type and size of samples to be tested.

e. Equipment descriptions (including accuracy, calibration,and sensitivity).

f. Data forms to be Þlled during the tests.

g. Acceptance criteria.

h. Predicted results and failure modes, where applicable.

i. References to applicable quality control procedures,codes, standards, etc.

j. Documentation of as-built dimensions and materialstrength.

9.4.5.2 After testing, the manufacturer should submit adetailed test report to the purchaser for approval. This shouldcontain the following as a minimum:

a. Gathered data and Þnal results.

b. Report on post-test examination.

c. Comparisons between predicted and observed values.

d. Conclusions.

9.4.6 Availability of Results

Tests should as much as possible be carried out in a consis-tent manner, such that the results will be applicable to futuredesigns. All test results should be available for veriÞcation offuture designs. Where practical, tests should be conductedsuch that the results and records could be accepted in lieu ofrepeated testing for other applications.

9.4.7 Intermediate Results

Results of all tests, including results at intermediatestages, should be compared with analytical results from thedesign program of the manufacturer. Discrepancies should

be investigated and reported to the purchaser. Where possi-ble, intermediate results should also be used to deÞne pipeproperties, such as axial and bending stiffness.

9.4.8 Validity of Test Results

Test results are valid unless a substantial change to the pro-cess (test procedure, design, or manufacturing procedure)invalidates the results.

9.4.9 Accelerated Tests

9.4.9.1 Accelerated tests may be performed by increasingthe following, subject to the approval of the purchaser:

a. Cyclic frequency.

b. Internal pressure.

c. Magnitude of movement.

d. Temperature.

9.4.9.2 For accelerated tests, the manufacturer should pro-vide documented evidence that the variation in test parameterdoes not signiÞcantly affect the results or change the mode offailure, and that the test period is satisfactory.

9.4.10 Multiple Tests

Single samples may be subjected to multiple tests, withnon-destructive tests (such as bending, torsional stiffnesstests, and FAT tests) performed prior to a destructive test. Thetest sequence needs to be carefully evaluated to ensure thatearlier tests do not affect the results of subsequent tests.

9.4.11 Repeatability of Results

When a single sample is tested, the design parameters andmanufacturing tolerance parameters that affect the perfor-mance should deÞne the bounds for the qualiÞcation achievedand should be accounted for in the deÞnition of the accept-able application envelope. Application of the test results indesign and analysis should use the critical parameters in aconservative manner.

9.5 PROCEDURES—STANDARD PROTOTYPE TESTS

Procedures are given in this section for the standard Class Iprototype tests, namely burst, tensile, and collapse tests.

9.5.1 Burst Test

9.5.1.1 Description

The test set-up for the burst test is shown in Figure 24. Theburst test should be performed with the specimen in a straightconÞguration. The minimum length of the test sample,excluding end Þttings, should be either two times the pitch


Figure 24—Schematic of Set-Up for the Burst Test

Temperature recorder

Pressure recorder

Thermocouple(optional)

140°Fhot water(optional)

Ambienttemperature

water

Counter &controller(optional)

Circulationpump

(optional)



PumpNote: This pressure and/or optional temperature test is similar to ASTM D2143 [48].

Linearmeasurement

device

P

T

Specimen


length of the outer armor wires/reinforcing cables for astraight conÞguration or three times the pitch length of theouter wires for a bent pipe. The test ßuid is generally water.

9.5.1.2 Procedure

Prior to commencement of the burst test, the requirementfor pressure cycling as per Section 9.4.3 should be consid-ered. The Þrst 50 percent of the expected load shall be appliedat a maximum rate of 1 percent per second with no holdingperiod prior to applying the balance of the load at a maximumrate of 5 percent per minute without holds. Failure is deÞnedby a sudden loss in pressure. The burst pressure, mode andlocation of failure should be noted. Internal pressure, pipetwist and pipe elongation should be continuously monitoredduring the test.

9.5.1.3 Acceptance Criteria

The measured burst pressure should be greater than thedesign requirements speciÞed in Section 5.3 of API SpeciÞ-cation 17J/17K. Failure of the end Þtting itself or failure due

to armor wire reinforcing layer pull out from the end Þttingshould not occur.

9.5.1.4 Analytical Requirements

The effect of tension and bending on burst pressure shouldbe analyzed.

9.5.1.5 Alternatives

The burst test may be performed with the sample bent to itsdesign MBR.

9.5.2 Axial tension Test

9.5.2.1 Description

The test set-up for the axial tension test is shown in Figure25. The axial tension test should be performed with the speci-men empty and free to twist. The minimum length of the testsample, excluding end Þttings, should be two times the pitchlength of the outer armoring wires reinforcing cables. One ormore pigs may be used to check the reduction in the internaldiameter during the test.

Tensile gauge

Tensile gauge

Pulling jack

Pulling jack

Elongation sensors

Elongation sensors

Coupling bar

Translating guides

Swivel

Angular sensors

Tensile Test – Free in Torsion

Tensile Test – Fixed in Torsion

Notes:�1. Test can be conducted at ambient pressure, design pressure, or both.�2.� Strain gauges are optional. If used they will only indicate surface conditions or the conditions at�� the layer where they are applied. They should not be considered to be representative of the�� general stress state of the pipe.�3.� Since catastrophic failure is probable, care should be taken to protect the personnel conducting�� the test.

Figure 25—Schematic of Set-Up for the Axial Tension Test [5]


9.5.2.2 Procedure

One end of the sample is Þxed and an axial load applied tothe other end at the rate given in Section 9.4.3. Load applica-tion should sufÞciently show that dynamic ampliÞcation isnot introduced. As a guideline, load application should becompleted in approximately 5 minutes. The failure tension,mode and location of failure should be noted. In addition,applied load, elongation and twist of the sample should becontinuously recorded. Failure occurs if the tensile load dropsor sudden elongation occurs.


The measured failure tension should be greater than thedesign requirements speciÞed in Section 5.3 of API SpeciÞ-cations 17J/17K. Failure of the end Þtting itself or failure dueto armor wire reinforcing layer pull out from the end Þttingshould not occur.


The effect of internal pressure and Þxing the ends fromrotating on the failure tension should be analyzed.


The axial tension test may be performed with the pipe fullof water at design or a lower internal pressure. In this case,the internal pressure should be continuously monitored dur-ing the test with sudden pressure drop (indicating an internalsealing failure) or reduction in tensile load taken as failure ofthe sample. The test may also be performed with the ends ofthe pipe Þxed in rotation.

9.5.3 Collapse Test

9.5.3.1 Description

9.5.3.1.1 The test set-up for the collapse test is shown inFigure 26(a). The test set-up should be such that the end Þt-tings (or sealed simple end caps) are not exposed to externalpressure-or, if exposed, a rigid bar should be installedbetween the two ends to eliminate end cap loads. The rigidbar may be omitted if the manufacturer wishes to demonstratethat the pipe design is suitable for compression loads. The testshould be performed with the specimen in a straight conÞgu-ration. The minimum length of the sample, excluding end Þt-tings, should be Þve times the internal diameter.

9.5.3.1.2 Prior to the test, the outer sheath should beremoved or perforated such that water ingress into the annu-lus of the pipe occurs. The sample should be at ambient inter-nal pressure and may be empty or Þlled (partially orcompletely) with water. In general, water is used as the testßuid. Note that it is not necessary to include the tensile armorlayers or the outer sheath in the sample. If included in the

sample, intermediate sheaths should also be removed or per-forated, unless the pipe design is based on an imperviousintermediate sheath.

9.5.3.1.3 Procedure

The external pressure may be applied at a maximum rate of1,500 psi per minute until failure occurs in the pipe. Failure isdeÞned as a sudden variation of the volumetric measurementor, depending on the test equipment, a sudden pressure loss.The collapse pressure, mode, and location of failure shouldbe noted.


The measured collapse pressure should be greater than thedesign requirements speciÞed in Section 5.3 of API SpeciÞ-cations 17J/17K.


The effect of bending and/or axial tension, and includingthe outer sheath, on the collapse pressure should be analyzed.


The sample may include end Þttings. The test may be per-formed with a leakproof outer sheath or with support to pre-vent axial compression of the pipe. The test may also beperformed with an axial tension load applied.

9.6 PROCEDURES—SPECIAL PROTOTYPE TESTS

This section lists recommended procedures for Class IIprototype tests, namely dynamic fatigue, crush strength, com-bined bending and tensile, sour service, Þre, erosion. TFL,vacuum, kerosene, adhesion, and full-scale blistering tests.

9.6.1 Dynamic Fatigue Test

9.6.1.1 Description

9.6.1.1.1 A schematic showing the overall deÞnition of thedynamic test program, including riser and bend limiterdesign, is shown in Figure 27. A typical test set-up is shownin Figure 28. The sample is hung vertically or tensioned hori-zontally from a rocker arm which can apply cyclic rotations.A tension load is applied to the opposite end. There are twotypes of full scale dynamic tests, a service simulation and ser-vice life model validation. The objective of a service simula-tion test is to determine the structural integrity of the topsection of the ßexible pipe, including end Þtting and bendlimiter, under simulated operational conditions. The objectiveof a service life model validation test is to apply loadingwhich results in cumulative damage equal to 1.0 based on theservice life analysis for a structural layer, normally either thepressure armor or the tensile armor.


9.6.1.1.2 The minimum length of the test sample, exclud-ing end Þttings, should be as follows:

a. The length between the lower end Þtting and the bottom ofthe bend protection device should be at least three times thepitch of the outer armor wires/reinforcing cables.

b. The length between the top end Þtting and the top of thebend protection device should be at least one pitch of theouter armor wires, unless the end Þtting is attached to a bendstiffener.

9.6.1.1.3 The test sample should have end Þttingsattached at both ends, with a bend stiffener attached to thetop end Þtting. As an alternative, a pipe without a bend stiff-ener may be tested if the set-up includes a suitable bell-mouth. The sample should be subjected to maximumoperating internal pressure and a conservative tensile loadrelated to the dynamic environment.

9.6.1.2 Procedure

9.6.1.2.1 The cyclic loading of the riser top should bedivided into a number of blocks each with a different angleamplitude, frequency, and number of cycles. The frequencyfor each load case should be speciÞed by the manufacturer.Typically the frequency increases as the angle range isreduced. Note that a higher frequency may reduce the totaltest period but may generate an unacceptable temperatureincrease in the riser top because of friction between the lay-ers. Local test site conditions, including temperature, machin-ery, and cooling requirements, will inßuence the cycling rate.Thermal analysis is recommended to determine the cyclingrate. An example of a typical cycling program is shown inTable 25.

9.6.1.2.2 In a service simulation test, the total number ofcycles in all blocks should be approximately 2 to 4 million.

Figure 26—Schematic of Set-Up for the Collapse Test

Pressuregauge Volumetric

measurementdevice

Pressuresource

Pressuresource

(a)

(b)

Pressuregauge

Volumetricmeasurement

device

Notes:�1. Pressure vessel and pressure source are to be capable of operating up to pipe collapse pressure.�2. Pipe specimen is axially stiffened in (a).�3. Ref. ASTM D2924 [49].


Figure 27—Dynamic Fatigue Test Program Definition

Floater data Functionalrequirements

Environmentalconditions

Static configurationdesign

Static analysis�tool

Dynamicconfiguration design& fatigue analysis

data

Service lifeprediction

Final pipedesign

Prototype testfunctional

requirements

Prototype testconfiguration

design

Final stiffenerdesign

Bend stiffenerdesign tool

Preliminarystiffener design

Fatigueanalysis data

Dynamic analysis�tool

Selection data

Testprogram

Bend stiffenerfor prototype

Prototype servicelife prediction

Test acceleration�tool

Bend stiffener�design tool

Testprogram

Fieldapplication

Input data

Notes:�1.� The objective of the flowchart is to show the following:�� (a) Flexible riser and bend stiffener design�� (b) Definition of dynamic qualification program.�2.� The bend stiffener design may be modified for the prototype sample so as to change the stress levels in the pipe.��

Test program�definition tool

Preliminary pipedesign

Flexible pipe�design tool


Figure 28—Typical Set-Up for a Dynamic Fatigue Test


The number of cycles in each block depends on the applica-tion (e.g., ßoater motions, environmental conditions). Anexample of a relative distribution of cycles per block isincluded in Table 25. The load cases should be selected suchthat the structural layer most susceptible to fatigue (i.e., short-est life calculated in the service life analysis) experiencescumulative damage in the test greater than or equal to thatexperienced in service over the Þeld life. The difference inannulus environment between test and Þeld conditions shouldbe considered in comparing the test damage with the Þelddamage. The loading should be applied either randomly or ingroups of a speciÞed percentage of all of the load blocks.Non-destructive inspection may be conducted periodically tocheck for damage to the structural layers in the bending zone.In a service life model validation test approximately 400,000cycles are applied with a single angle range, tension andinternal pressure. This block is selected based on achieving acumulative damage of 1.0 in the layer most susceptible todamage. The test conditions may be adjusted to attempt toachieve 1.0 cumulative damage in any of the structural layersbased on the service life analysis.

9.6.1.2.3 In the service simulation test, the last block, withthe largest cycle amplitude, normally represents the extremeoperation conditions. Only a limited number of cycles arerequired to represent this condition, preferably at the end ofthe test program. Application of the largest amplitude block isheld until the end because it may artiÞcially improve thefatigue performance of the pipe by strain hardening the armorwires. If it can be shown that strain hardening does not occur,the largest amplitude blocks may be applied at both the begin-ning and end of the test to create a more conservative test.

9.6.1.2.4 The following variables should be continuouslyrecorded:

a. Number of cycles.

b. Internal temperature.

c. External ambient temperature.

d. Internal pressure.

e. Applied tension load.

f. Actual angles applied.

9.6.1.2.5 The end of the initial dynamic fatigue test isdeÞned as failure of the pipe (or bend stiffener) or, alterna-tively, successful completion of all cycles. If fatigue failure ofthe pipe does not occur, the sample should be subsequentlypressure tested at a minimum of 1.25 times the design pres-sure with the tensile load applied. If it is planned to conductadditional dynamic loading upon completion of the test, thepipe should be non-destructively inspected to verify the con-dition of the structural layers prior to proceeding. Proposedmethods of non-destructive inspection are presented in Table27. The dynamic fatigue test may be continued in a servicesimulation test if there is notable damage to one of the struc-tural layers which does not result in failure of a pipe layer. Itmay also be continued in a service life model validation test ifthe non-destructive inspection does not indicate any notabledamage. Notable damage is deÞned in Section 9.6.1.3.

9.6.1.2.6 A layer by layer dissection of the test sampleshould be conducted to record the condition and evidence ofdegradation of the pipe structure over an area including thelocation of highest curvature variation. Layers which showsigns of damage should be subjected to detailed examination.

9.6.1.3 Acceptance Criteria—Service Simulation

The pipe should have passed the test sequence withoutleakage or failure of the pipe structural layers as deÞnedbelow. If there is notable damage to any of the structural lay-ers, the test should be continued for an additional 25% cumu-lative damage to the layer which is notably damaged.Reference is made to Section 13.3 and Tables 24 through 26for other defects which may be considered to affect the integ-rity of the pipe structure. A test pipe which has been through aservice simulation test is expected to suffer some layer degra-dation from the as-built condition. The acceptance criteria for

Table 25—Sample Dynamic Fatigue Test Program

Block No.

Mean Angle(¡)

Cycle Amplitude(¡)

Minimum Angle(¡)

Maximum Angle(¡)

Relative No.of Cycles

1 5.0 1.25 3.75 6.25 1.000

2 5.0 2.50 2.50 7.50 0.550

3 5.0 3.75 1.25 8.75 0.250

4 5.0 5.00 0.00 10.00 0.075

5 5.0 7.50 -2.50 12.50 0.025

6 5.0 10.00 -5.00 15.00 0.010

7 5.0 15.00 -10.00 20.00 0.001

Note: No more than 25% of the cycles of any block containing more than 1% of the total cycles should be applied prior toswitching to another block.


each layer should be clearly agreed between the purchaser andmanufacturer prior to completion of the initial dynamic test.


The result of this test is a curvature histogram indicatingthe number of cycles per class without failure of the pipestructure, end Þtting, or bend stiffener and documentation ofthe dissection. A comparison of the predicted and actualresults based on the service life analysis should also be pro-vided. This information may be used to estimate the life timeof a particular riser design for the expected history of ßoatermotion and environmental conditions.


This particular test focuses on fatigue at a riser top connec-tion. Alternative test set-ups will be required if other sectionsof the riser are considered critical, e.g., riser sag bend or sea-bed touchdown region for catenary risers. In this particulartest conÞguration the following parameters may be altered:

a. Internal pressure.

b. Internal temperature.

c. Mean angle.

d. Cycle amplitude.

e. Number of cycles.

In addition, strain in the outer tensile wires/reinforcingcables near the bend stiffener may be recorded.

9.6.2 Crush Strength Test

9.6.2.1 Description

9.6.2.1.1 The crush strength test is performed to determinethe suitability of a particular design for installation with ten-sioners. The number of tensioner belts is typically three orfour.

9.6.2.1.2 The test set-up should represent the tensionersystem on the particular installation vessel. In particular, thenumber of belts and geometry of shoes should be compara-ble. The minimum length of the sample should be two timesthe pitch length of the outer armoring wire when tensile loadsare applied.

9.6.2.2 Procedure

The ßexible pipe sample should be positioned-empty,without internal pressure-on the test device. The crushingload is increased from zero up to 110% of the pipe designcompression capacity at a rate not greater than 1% of themaximum load per second (1%/s). The compression loadshould be kept constant (within ±2%) for a period of at leastone hour. In the loaded condition and after unloading com-pletely, the ovalization of the pipe is measured. Test loadsshould be based on the load expected during installation witha safety factor. The radial load is a function of pipe weight,depth, and other factors.


The permissible ovalization of the pipe in the loaded condi-tion is 3% and in the unloaded condition is 0.2% (the value forthe unloaded condition may be increased if the larger value isused in collapse calculationsÑrefer to Section 5.4.2.4).


The effect of tensile load on the crush strength of the ßexi-ble pipe should be analyzed.


The crush strength test may be performed with a tensileload applied. It is recommended that the tensile load be atleast the design installation tension and be applied prior to thecompression load at a rate not to exceed 1% of the load persecond. Also, the compression load could be increased insteps until the acceptance criteria are exceeded, so as to deter-mine the maximum compression load of the pipe.

Layer Failure DeÞnition Notable damage

Internal carcass Through wall crack or loss of interlock which would cause pipe col-lapse or damage to the pressure sheath if the pipe was bent to the SBR in any plane.

Deformation of proÞle, loss of cross section

Pressure armor Through wall crack or loss of interlock which would cause failure of the internal pressure sheath if the pipe was bent to the SBR in any plane

Variance from the pro-Þle shape which results in the service life (by analysis) being reduced below the Þeld life.

Non-through wall cracks in areas with the highest alternating stress.

Tensile armor Torsion imbalance greater than 1 degree per meter in the Þeld hydrotest (one end free to rotate)

Less than 5% of the armor wires broken in any layer

Axial stiffness of the pipe reduced by a factor of 20% from value at beginning of test

More than 5% of the armor wires broken in any layer


9.6.3 Combined Bending and Tension Test

9.6.3.1 Description

9.6.3.1.1 The combined bending and tension test is per-formed to verify the installation of a particular ßexible pipedesign with a horizontal installation spread. This test simu-lates the passage of the pipe over the sheave of an installa-tion vessel. It is not necessary in this test for the sample toinclude production type end Þttings. The terminations needonly be capable of transferring tensile load to the ßexiblepipe. Damage due to the dissection process should beignored.

9.6.3.1.2 The test sample should be positioned, empty, atambient internal pressure, on a special device that simulatesthe pipe laying sheave of the installation vessel, with anidentical bend radius and transverse proÞle. The sampleshould also be connected to a suitable tensile load machine.The straight section of pipe connected to the tensile loadmachine should be at least the length of the pipe bent overthe sheave.

9.6.3.2 Procedure

9.6.3.2.1 The axial load is applied at a rate not greater than1% of the design installation tension per second up to 110%of the design tension. The allowable variation in the designtension should be ±2%. This load is held for a minimumperiod of 1 hour.

9.6.3.2.2 The external diameter of the pipe is measured attwo locations 90¡ apart on the pipe circumference in thecurved section of the pipe, with one measurement locationbeing the contact face of the pipe. The tensile load is released,and the diameter measurements retaken.


The allowable variations in the external diameter are as fol-lows:

a. Loaded Condition: ±3%

b. Unloaded Condition: ±1%


The effect of different sheave bend radii and tensile loadson the pipe deformation should be analyzed.


After completion of the above test, the tensile load may beincreased in steps not greater than 1% of the design installa-tion tension per second until the acceptance criteria above areexceeded. This is deÞned as the failure installation tension.

9.6.4 Sour Service Test

9.6.4.1 Description

9.6.4.1.1 In addition to bench tests of the steel wire/cablematerials (refer to API SpeciÞcations 17J/17K, Section 6.4.2)to verify performance in sour service conditions, prototypetests on a full scale pipe may also be carried out. Tests of thiskind may be used to generate a realistic sour service environ-ment in the pipe annulus (unbonded) containing the steelwires and at the cable surface (bonded), and in addition simu-late wire loading conditions by ßexing the pipe.

9.6.4.1.2 The tests will normally be carried out while sim-ulating a wet annulus for unbonded pipe, either with saltwater to test the failure condition, or with fresh water to simu-late normal operating conditions assuming shutdowns havecaused condensation. Rubberized cables will normally beused for bonded pipe tests

9.6.4.1.3 Two approaches may be taken as follows:

a. Injection of a known concentration of H2S/CO2 into thewet annulus directly.

b. Injection of the known H2S/CO2 concentration into thepipe bore and allowing the annulus cable surface to reach anequilibrium state from permeation through the internal pres-sure sheath.

Note: Only approach b. is relevant to bonded ßexible

9.6.4.1.4 In either case in 9.6.4.1.3, it is necessary to carryout a prediction of the steady state annulus/cable surface con-ditions based upon a diffusion/corrosion model which isagreed with the ßexible pipe manufacturer.

9.6.4.1.5 Unless the concentration of H2S is high, it islikely that to achieve steady state in a reasonable period oftime (e.g., 2 to 3 months), an artiÞcially high concentrationwill be necessary for an initial period to accelerate stabiliza-tion. Prediction of the stabilization process should also bemade using a consistent diffusion/corrosion model agreedupon with the manufacturer.

9.6.4.1.6 The test ßuid characteristics may simulate ser-vice conditions for the pipe product, or be in accordance withNACE TM 01-77 if a general qualiÞcation is sought. The testshould be designed to obtain saturation of the steel compo-nents in the annulus of the pipe or at the surface of the cableto a level at least equal to the design partial pressure (in theannulus/pipe bore (bonded) of H2S and CO2. The internalßuid in the pipe should be at design pressure.

9.6.4.1.7 The ßuid temperature is recommended to beapproximately 25¡C, unless operational temperature isexpected to be considerably less, in which case the operatingtemperature should be used. The test sample should includeend Þttings identical to those proposed for the application.


9.6.4.1.8 Tests based on injection into the pipe bore arepreferred because the diffusion of H2S and CO2 correctlymodels pipes in service.

9.6.4.1.9 Tests for dynamic risers may be carried out intwo phasesÑÞrst, injection of H2S/CO2 while the pipe isstatic, and then once the desired equilibrium is reached, ßex-ure of the pipe, producing known alternating stresses. Thealternating stresses should be representative of the stressrange blocks modeled in a dynamic fatigue program (refer toSection 9.6.1), adjusted so as to generate a known level offatigue damage in the wires/cables.

9.6.4.1.10 Following completion of the full scale exposuredynamic ßexure test, the resulting fatigue damage may beassessed by completing in-air fatigue tests on samples of thewire to determine the Òremaining life.Ó The pipe to be testedhas to be sited in a facility suitable for large scale sour servicetesting. This normally comprises a concrete bunker or anenclosed space with extraction ventilation in accordance withlocal health and safety regulations.

9.6.4.2 Procedure

9.6.4.2.1 Exposure of the ßexible pipe armor wires/rein-forcing cables to H2S and CO2 is achieved by ßowing ßuid(water plus dissolved gas components through the annulus oroil plus gas components through the bore) through the pipesample at a predetermined rate.

9.6.4.2.2 Sampling of ßuid from the pipe outlet (annulus/bore) is required to determine the consumption of H2S andCO2. Where injection is into the bore, sampling of the annu-lus is also required.

9.6.4.2.3 The test solution will then be continuouslyinjected for a given period of time after equilibrium isreached to determine either the corrosion rate (static pipes) orfatigue performance (dynamic pipes).

9.6.4.2.4 At the end of the exposure test, the pipe shouldÞrst be pressure tested and then dissected.

9.6.4.2.5 A decision needs to be made at this point as towhether burst test data is required, which may be most appro-priate for static ßowlines, or if remaining fatigue life data isrequired. In the latter case, appropriate to dynamic risers, thepipe should be dissected and wire samples bench tested forremaining fatigue life compared to new unexposed formedwires.

9.6.4.2.6 A burst pressure test should be carried out instages, raising the pressure by 20% of design (or lesser stepsif desired) from the exposure test pressure, with a hold timeof at least 3 hours between each step. The ßuid in the pipeshould be clean of H2S, while precautions should still bemaintained for H2S due to release of the gas when burstoccurs.

9.6.4.2.7 Flexure of a pipe to simulate dynamic serviceconditions may be most conveniently achieved by installationin a horizontal ßexing frame. One or both pipe ends may needto have bend stiffeners installed to control curvature. The pipeßexure should be designed so as to induce appropriate tensileloads for fatigue in the tensile wires or reinforcing cables inan area of maximum curvature of the pipe, in addition to real-istic loadings in the pressure armor (unbonded).


9.6.4.3.1 It should be noted that full scale sour service testsare a very challenging task and should be considered as partof a product development program rather than as part of aproduct qualiÞcation for a speciÞc project. Test duration mayexceed a calendar year, and interpretation of the results maynot be straightforward.

9.6.4.3.2 Static pipe tests may be assessed on the basis ofdecay of the burst pressure over time because of corrosion,assumed to be linear with time after equilibrium is reached.This is with a proviso that the corrosion is generalized ratherthan local pitting. If the latter, then the average depth and rateof growth of pits may be used to predict expected service life.

9.6.4.3.3 Dynamic pipe tests are rather more difÞcult topredict, as the combination of the loading environment andthe corrosion phenomena are complex. At present, the bestadvice to give is that the manufacturer and user shouldtogether develop a model to predict service life which ismutually acceptable.


There should be an analytical model available for the cor-rosion rate, the loading conditions (including annulus envi-ronment and the service life assessment) which has beenaccepted by both manufacturer and user prior to the tests.

9.6.5 Fire test

9.6.5.1 Description

9.6.5.1.1 The objective of the Þre test is to determine thesurvival time for the ßexible pipe in a particular Þre situation.The Þre resistance may be designed into the pipe structure ormay be achieved by non-integral passive Þre protection.

9.6.5.1.2 The Þre test may be carried out to the conditionsdeÞned in the LloydÕs Register recommended test (Fire Test-ing Memorandum 00/OSG 1000/499 Rev. 1). These can besummarized as a Þre temperature of 700¡C and a Þre durationof 30 minutes.

9.6.5.1.3 The pipe should be tested at the design pressure.The pipe internal ßuid may be water or another agreed ßuid.The ßuid should be stationary to simulate worst case loadingconditions. The end Þtting design to be used in the applica-tion should be used in the test sample.


9.6.5.2 Procedure

9.6.5.2.1 The pipe is pressurized to the design pressure.The Þre test should commence once pressure stabilizationoccurs. Both the ßexible pipe body and end Þtting should besubjected to the required test conditions. Pressure in excess ofthe design pressure may be relieved.

9.6.5.2.2 When the pressure in the pipe drops below 90percent of the test pressure, pipe failure should be consideredto have occurred. The survival time is then deÞned by thetime from Þre start-up to pipe failure.


The survival time should exceed the design requirements.


There are no analytical requirements for this test.


Alternatively, the test set-up may be in accordance withDnV ClassiÞcation Note 6.1 Test (i.e., furnace or propaneburners). The ßame temperature should be based on the worstcase likely Þre loading condition. Typical ßame temperaturesfor a jet Þre are approximately 1100¡C and for a pool Þre areapproximately 1000¡C, speciÞcally for a pipe engulfed byßames. If the pipe is not engulfed, ßame temperatures of400¡C to 600¡C may be appropriate.

9.6.6 Erosion Test

9.6.6.1 Description

9.6.6.1.1 A typical test set-up for an erosion test is shownin Figure 29. The test sample should be Þxed at its minimumbend radius in a 90 degree angle. Erosion rates may be deter-mined by thickness reduction (localized erosion rate) or byweight loss (average erosion rate) in the internal carcass.

9.6.6.1.2 The internal ßuid composition should representdesign conditions or be conservative. Consideration should begiven to the following:

a. Flow rate.

b. Sand content.

c. Particle size.

d. Temperature.

e. Pressure.

f. Corrosive gas content.

9.6.6.2 Procedure

The test ßuid should be circulated through the ßexible pipefor a minimum of 7 days. After completion of the test, erosion

measurements should as a minimum be made at Þve pointsaround the bend (0¡, 15¡, 30¡, 45¡, and 90¡ measurementpoints are recommended).


The erosion rate should be such that the design require-ments for the pipe are not violated for the speciÞed servicelife.


The effect of variations in the test ßuid composition, ßowrate, and pipe bend radius should be analyzed.


The effect of corrosive ßuids on the erosion rate may betested, as an alternative. This would be to determine corrosionenhanced erosion rates.

9.6.7 TFL Test

9.6.7.1 The purposes of the TFL test are to verify that TFLpumpdown tools adequately drift through the ßexible pipeand to determine ßexible pipe wear rates because of repeatedtool travel. The test unit, shown in Figure 29, simulates a TFLpipe run using a ßexible pipe which is 150 feet long.

9.6.7.2 The pipe is attached to both ends of a pump andmanifold unit that provides measurable hydraulic ßuid powerand a means for reversing the ßuid direction inside the pipe.The ßexible pipe is laid out in two conÞgurations: a wide ÒUÓshape with a twelve-foot-bend radius, and a narrow ÒUÓshape with a Þve-foot-bend radius (measured to the center-line).

9.6.7.3 Prior to hook-up, a TFL pumpdown tool string isinserted in the pipe. The TFL tool string should consist offour ÒupÓ locomotives, four ÒdownÓ locomotives, and a run-ning tool. The running tool may be either a TFL drift mandrelor two Òsharp shoulderedÓ drift mandrels, in which the ÞrstdriftÕs spring loaded keys are oriented 90 degrees out frombehind the second driftÕs keys. Both running tools should berun through both test conÞgurations and cycled through thepipe several times.

9.6.7.4 In general, the TFL drift mandrel tool string shouldbe able to pass freely through the pipe in either direction (seeAPI Recommended Practice 17C for drift mandrel dimen-sions, forces, and pressures). After the tests are completed,the tool strings and the pipe interior should be inspected foradverse wear or damage.

9.6.7.5 If specialized running tools for an application areknown (such as parafÞn scraper, sand wash wand, or Òkick-overÓ tool), then it is recommended to run these tools in thetest loop as well.


9.6.8 Vacuum Test

9.6.8.1 Description

9.6.8.1.1 This prototype test is intended for bonded ßexi-ble pipes only. The objective of the vacuum test is to indicatethe adequacy of the bond strength of the liner to other pipelayers.

9.6.8.1.2 The vacuum test is not applicable to pipes inwhich an internal steel interlocked carcass is used. In addi-

tion, the vacuum test may not be practical for long length(>11 m) or small diameter pipes.

9.6.8.2 Procedure

9.6.8.2.1 The pipe should be vacuum tested to a pressure

of 0.85 bar gauge and held for a 10 minute period.

9.6.8.2.2 A clear plastic window should be Þtted at eitherend of the test sample so that visual inspection of the interior

Figure 29—Schematic of Set-Up for the Erosion Test [5]

T1 P1

PT

300

90°

45°

30°

15°

0°r =

100

0

50o/o/

o/ 54

Water•

Sand

5650

1000

61


can be made by an adequate light source in one end with itsbeam directed to the other.


9.6.8.3.1 Collapse of the pipe liner, failure of adhesionbetween layers within the pipe body, blisters, and other defor-mities should not occur.

9.6.8.3.2 The pipe should be examined outside as well asinside for any deformities that may occur.


There are no analytical requirements for this test.


This test may be carried out within 24 hours of the kero-sene test to determine the resistance of the pipe to permeationor migration of highly seeking ßuids or gases. The vacuumtest will ÒpullÓ the kerosene out of the pipe body if signiÞcantpermeation or migration has occurred.

9.6.9 Kerosene Test

9.6.9.1 Description

9.6.9.1.1 The prototype test is intended for bonded ßexiblepipes only. The objective of the kerosene test is to detect anypermeation or leakage of a highly seeking hydrocarbon liquidthrough the pipe liner.

9.6.9.1.2 This test may be followed immediately by a vac-uum test to further detect any residual kerosene which mayhave migrated into the pipe body.

9.6.9.1.3 This test is primarily for bonded ßexible pipewith no internal interlocked steel carcass.

9.6.9.2 Procedure

9.6.9.2.1 The pipe should be laid out straight and Þlledwith kerosene venting all air. The pipe should then be pressur-ized to the design pressure and maintained for 24 hours.

9.6.9.2.2 Consideration should be given to cycling thepressure, prior to initiating the test, to help stabilize the pres-sure over the 24 hour period.


After 24 hours the pipe should be depressurized, drained,dried, and observed for any blistering, leakage, or separationof the liner from the carcass or from the end Þtting.


An analytical model for the permeation of highly seekingßuids or gases should be available and accepted by both man-ufacturer and purchaser prior to the tests.


After completion of this test, a vacuum test should be per-formed to further detect the permeation or migration ofhighly seeking ßuids or gases into the pipe body.

9.6.10 Adhesion Test

9.6.10.1 Description

9.6.10.1.1 This prototype test is intended for bonded ßexi-ble pipe only. The adhesion test is used to verify the bondstrength of the manufactured pipe.

9.6.10.1.2 Adhesion tests should be performed on samplesmade from materials taken from current manufacture and onsamples representative of every tenth hose thereafter (in thecase of speciÞc lengths).

9.6.10.1.3 Samples should be built with the same cross-section make up as the production pipe and should be built atthe same time as the production pipe or as agreed on by thepurchaser and manufacturer. The sample piece may be builtwith the cables at the reinforcing layer wound in the radialdirection i.e. a lay angle of 90 degrees to pipe longitudinalaxis, to facilitate this test. Vulcanization should occur underthe same conditions as the production pipe.

9.6.10.2 Procedure

Adhesion tests should be carried out according to eitherASTM D413 Machine Method or BS 903 part A12 usingstrip pieces.

9.6.10.3 Acceptance

The measured adhesion strength should not be less than 6N/mm.

9.6.11 Full Scale Blistering Test

9.6.11.1 Description

9.6.11.1.1 The full scale blistering test is performed todetermine the suitability of a particular pipe design for ser-vice in a gas containing environment and hence qualify thematerials used for service.

9.6.11.1.2 The pressure, depressurization rate, temperatureand ßuid type should as a minimum be consistent with whatthe pipe is expected to be subjected to during a typical appli-cation. It is preferable to use an inert gas of similar molecularstructure as the gas expected to be conveyed and with a mini-mum CO2 content of 5%.

9.6.11.1.3 The test pipe should be at least 3 m long includ-ing end Þttings or sufÞciently long as to ensure that any bene-Þcial effects the end Þttings have on inßuencing the outcomeare eliminated.


9.6.11.2 Procedure

9.6.11.2.1 The manufacturer should have documentedprocedures to ensure that the test gas occupies 100% of theinternal pipe volume. Once the pipe is Þlled with the test gas,the pressure should be gradually increased at a rate notgreater than the manufacturerÕs test procedure, to the designpressure, and held for a period of at least 2 hours to allow forstabilization. If necessary the pressure shall be consideredstabilized when the pressure drop is less than 1% in a onehour period. The pressure should be cycled to this pressureuntil stabilization is achieved. The pipe should then be heldat this pressure to ensure saturation of the pipe body with gasfor a length of time not shorter than that of the manufac-turerÕs test procedure.

9.6.11.2.2 Once saturation of the pipe body is achieved,the pipe should be depressurized at a rate equal to theexpected depressurization rate or else a minimum of 70 barper minute.

9.6.11.2.3 The procedures set out in 9.6.11.2.1 and9.6.11.2.2 above should be repeated for the expected numberof cycles or a minimum of 60 cycles.


9.6.11.3.1 Once the test is complete, the end Þttingsshould be cut off the test pipe, the pipe body should be cut inhalf lengthwise and the half shells cut radially into threeapproximately equal lengths. The carcass layer should beremoved to expose the elastomer surface beneath it. When theeight sample pieces are inspected on all surfaces at 1 x mag-niÞcation, there should be no evidence of delamination, blis-tering, or voids in the elastomer layers.


The soak time should be computed based on the measuredpermeability of the elastomer to the gas under consideration.


As an alternative to this test, small scale blistering resis-tance tests which reßect the design requirements, relating inparticular to ßuid conditions, pressure, temperature, numberof depressurizations and depressurization rate may be per-formed, (see Section 6.2.3.2 of API SpeciÞcation 17K fordetails of these tests).

In addition, the full scale prototype test piece may be usedto measure adhesion of the elastomer to the end Þtting oncethe blistering test is completed. The full scale blistering testmay also be carried out on the pipe after it has been used in afull scale fatigue test program.

10 Manufacturing

10.1 SCOPE

10.1.1 Section 7 of API SpeciÞcation 17J/17K speciÞesmanufacturing requirements for unbonded and bonded ßexi-ble pipes. This section describes the processes involved in themanufacture of the pipe. In addition, guidelines on selectionof manufacturing tolerances are given. Guidelines on assem-bly of end Þttings are also included.

10.1.2 Furthermore, this section provides guidelines onmarking and storage of ßexible pipes. The marking guide-lines supplement the minimum requirements for markingshown in Section 10.1 of API SpeciÞcations 17J/17K.

10.2 MANUFACTURING—UNBONDED PIPE

The manufacturing of unbonded ßexible pipe is composedof two main stages, as follows:

a. Fabrication of the ßexible pipe body.

b. Assembly and mounting of the end Þttings.

These two stages in the process are described in the follow-ing sections.

10.2.1 Manufacturing Processes

The main processes in the fabrication of the ßexible pipebody are as follows:

a. Carcass forming.

b. Polymer extrusion.

c. Pressure armor winding.

d. Tensile armor winding.

e. Tape winding.

Depending on the pipe design, processes (a) and (c) maynot be required.

10.2.1.1 Carcass forming

In the carcass forming process, ßat metallic strips arepulled into a forming head in which they are shaped into aninterlocking helical tube (see Figure 7 for an example of thecarcass shape).

10.2.1.2 Polymer Extrusion

10.2.1.2.1 Extruded components in a ßexible pipe includepolymer sheaths (internal pressure, intermediate, or outersheath) and solid anti-wear layers. The stations/equipment inthe polymer extrusion line are typically as follows (for arough bore structure):

a. Payoff reel (or basket) with the inner carcass layer.

b. Caterpillar (pre-extrusion).


c. Extruder.

d. Quench tanks (hot and cold water).

e. Caterpillar (post-extrusion).

f. Take-up reel (or basket).

10.2.1.2.2 The control of the extrusion process is impor-tant for quality of Þnished product, and a feedback controlsystem is recommended. See Section 7.3 of API SpeciÞcation17J.

10.2.1.3 Pressure Armor Winding

10.2.1.3.1 Using shaped wires (see Figure 7 for someexamples), the pressure armor winding machine preforms,interlocks, and winds the wires circumferentially around theinternal pressure sheath. Payout/takeup reels or (baskets) andcaterpillars are used to control the feed of the pipe through thewinding machine.

10.2.1.3.2 The interlocking pressure armor is laid as oneor two wires at a lay angle of close to 90 degrees. A ßat back-up layer may also be wound on top of the interlocked layerusing the same process.

10.2.1.4 Tensile Armor Winding

10.2.1.4.1 The tensile armor winding machine takes ßat,round, or shaped wires and preforms and winds the wiresonto the surface of the pipe. The number of wires wound inone layer is typically between 30 and 80. The wires are gener-ally laid with an angle range between 20 and 60 degrees. Thewires are stored in individual drums connected to the windingmachine. The drums rotate with the winding machine whilefeeding it with wire.

10.2.1.4.2 Two machines in sequence or one machine usedtwice can be used to apply the double crosswound tensilearmor layers used in most applications. These machines maybe subject to regular stoppages for reloading of drums andwelding of new wires.

10.2.1.5 Tape Winding

Tape winding machines are used to apply anti-wear, manu-facturing aid, or insulation layers. These machines are typi-cally used in sequence with one of the other processes.

10.2.2 End Fittings

10.2.2.1 The end Þtting is a critical part of the ßexiblepipe. A well designed transition zone is required for all thepipe wall components to converge into one ßange or connec-tor piece that carries all the pipe wall forces.

10.2.2.2 The pressure and tensile armor layers are lockedto the end termination body so as to ensure reliable attach-ment in both radial and axial directions. The pressure integrity

of the external and internal sealing layers (polymer sheaths)are provided by a seal arrangement that also ensures radialand axial attachment. The zone near the end Þtting will nothave the same ßexibility as the rest of the pipe. This zone, cor-responding to the length of a couple of turns of the tensilearmor, therefore does not have the same curvature capacity(ßexibility) as the main pipe section.

10.2.2.3 A schematic of a typical unbonded pipe end Þt-ting is shown in Figure 8. Most of the components in the endÞtting are applied manually with special tools and Þxtures.Quality control of all processes in the fabrication of the endÞtting is therefore critical.

10.2.2.4 The main steps in the process are as follows:

a. Separate individual layers of pipe.

b. Mount inner seal assembly and main end Þtting body.

c. Clamp pressure armor layer.

d. Secure tensile armors around body.

e. Mount external jacket.

f. Mount outer locking assembly (sealing of outer sheath).

g. Fill voids in end Þtting with epoxy resin and allow to set.

10.2.2.5 When bend stiffeners are required at the end ofthe ßexible pipe, they are usually mounted on the pipe priorto the end Þtting and subsequently pulled up and attached tothe end Þtting once it is mounted.

10.2.3 Tolerances

10.2.3.1 This section provides guidelines on the selectionof manufacturing tolerances (see Section 7.8 of API SpeciÞ-cation 17J for the minimum requirements of the selected tol-erances). The tolerances speciÞed in this section are deÞnedin terms of percentage of nominal values.

10.2.3.2 For unbonded ßexible pipes the length tolerancefor lengths up to 100 meters should typically be Ð0 metersand +1 meters. For unbonded lengths greater than 100 metersthe length tolerance may be increased to Ð0 percent and +1percent. For bonded pipes the tolerance may typically be ±1percent. For certain projects there may be additional require-ments on the length tolerance to be considered, including thefollowing:

a. For certain applications, such as jumpers, the tolerancesmay need to be reduced.

b. Some applications may have problems if the length is toolong, e.g., for long ßowlines a maximum tolerance of +1 per-cent may be too large because of insufÞcient space at the endconnection to accommodate excess length. This may be morecritical for trenched pipe.

c. If two or more risers are clamped together (such as withumbilicals in some applications), consideration should be


given to possible problems caused by the individual risershaving different lengths.

d. The calculation of the required ßowline length shouldaccurately account for all parameters, including undulationsin the route, accuracy of end point locations, installation tol-erances, manufacturing tolerance, and orientation of theßowline to the component (e.g., the pipe may be laid in a looparound the component, such as at a wellhead, and connectedat a 90 degrees orientation to the main ßowline direction).

10.2.3.3 The recommended tolerance on the ßexible pipeoverall outer diameter is ±3 percent. For carcass layers thatare not manufactured on a mandrel, the tolerance on internaldiameter should be Ð0 percent and +2 percent. For internalpolymer sheaths that are not extruded on to an inner carcass,it is recommended that the tolerance on the internal diameterbe between Ð0 percent and +2 percent.

10.2.3.4 Tolerances should be established and controlledby the manufacturer for each layer of the pipe. Recommenda-tions on critical aspects of dimensional tolerances for the ßex-ible pipe layers are listed in Table 26.

10.2.3.5 The manufacturer should check pressure andarmor layer tolerances for the allowable gap between adjacentwires or the allowable average gap over a group of wiresagainst manufacturer speciÞcations.

10.3 MANUFACTURING—BONDED PIPE

The manufacture of bonded ßexible pipe is composed ofthree main stages, as follows:

a. Fabrication of the ßexible pipe body

b. Assembly and mounting of the end Þttings

c. Curing of ßexible pipe

Note: Stages b and c are interchangeable in sequence for some pipes.

10.3.1 Manufacturing Processes

The main processes in the fabrication of the ßexible pipebody are as follows:

a. Carcass forming.

b. Preparation of compound and calendering.

c. Elastomer winding.

d. Reinforcement armor winding.

Depending on the pipe design and application, process a.may not be required.

10.3.1.1 Carcass Forming

In the carcass forming process, ßat metallic strips arepulled into a forming head in which they are shaped into aninterlocked helical tube (see Figure 7 for an example of the

carcass shape). Some bonded ßexible pipe manufacturers donot carry out this task, preferring instead to obtain pre-manu-factured carcasses.

10.3.1.2 Preparation of Compound and Calendering

10.3.1.2.1 The process by which the compound is pre-pared involves accurately weighing out each ingredient of thecompound, mixing the ingredients in the speciÞed order andat speciÞed temperatures in a large ÔBanburyÕ type mixeruntil a homogenous consistent compound is formed.

10.3.1.2.2 The calendering process involves passing theprepared compound between rollers repeatedly until the com-pound takes on the form of a smooth, even sheet with noßaws or blisters. This sheet may be subdivided into smallerstrips and subsequently wound onto reels for storage or cutand stored as smaller ßat sheets. The friction caused by forc-ing the compound through the calendering rollers causes anincrease in the temperature. This temperature should be con-trolled so as to ensure that over-curing does not occur duringcalendering. The compound is generally passed through abath containing an anti-adhesion substance prior to storage.Alternatively, the compound material may be stored withplastic sheets between each layer.

10.3.1.2.3 The steel cables of the reinforcing layer may beincorporated into a sheet of compound during the calenderingprocess or by an extrusion process. This facilitates winding ofthe reinforcing layer onto the pipe, and speeds up the fabrica-tion stage. These sheets are generally stored on reels for easeof use.

10.3.1.3 Elastomer Winding

10.3.1.3.1 The production pipe is generally built up bywinding sheets of calendered elastomer onto a mandrel orinterlocked steel carcass. The winding process continues withdifferent compounds as per the cross-sectional speciÞcation,including calendered reinforcing cables, until the pipe is fullybuilt up.

10.3.1.3.2 The control of the winding process is importantfor the quality of the Þnished product as irregular overlapsand gaps in the winding process may cause unevenness in thepipe cross-section. See Section 7.4 of API SpeciÞcation 17K.

10.3.1.3.3 The elastomer may also be extruded to build upthe pipe cross section although winding is more common. See10.2.1.2 for guidelines on elastomer extrusion.

10.3.1.4 Reinforcement Armor Winding

10.3.1.4.1 The cables which make up the reinforcementmay be wound onto the pipe body in two formats. The Þrstformat is simply by an armor winding machine where the


cables are stored in individual drums connected to the wind-ing machine. The drums rotate with the winding machinewhile feeding it with cable as the pipe advances through themachine. In some cases it is the pipe which rotates while thewinding machine traverses horizontally. The second format isidentical to the way in which the elastomer sheets are woundon. The cables are pre-calenderised and stored on reels inlong narrow strips. These are then wound onto the pipe bodyby rotating the pipe body and advancing either the pipe orwinding machine at a pre-deÞned rate.

10.3.1.4.2 Two machines (or more) in sequence or onemachine (or more) used twice can be used to apply the doublecrosswound armor cables.

10.3.1.4.3 The control of the winding process is importantto maintain the quality of the Þnished product. See Section7.5 of API SpeciÞcation 17K.

10.3.2 End Fittings

10.3.2.1 The end Þtting is a critical part of the ßexiblepipe. A well-designed transition zone is required for all thepipe wall components to converge into a ßange or connectorpiece that carries all the pipe wall forces.

10.3.2.2 The cables of the reinforcement armor layer arelocked to the end termination body so as to ensure reliableattachment in both radial and axial directions. The pressureintegrity of the external and internal sealing layers (elastomercover and liner) is provided by curing the layer onto the end-Þtting which also ensures radial and axial attachment.

10.3.2.3 In some cases the end Þtting can be swagged ontothe pipe body. This involves an internal and external steelendÞtting piece which encapsulates the pipe body and, whenswaged, compresses the pipe body sufÞciently to ensure bothÞxity and sealing of the liner, cover and cables of the rein-forcement layer. The end Þtting face in contact with the pipebody may be smooth or toothed. The toothed end Þtting isdesigned to contact the cables of reinforcement layer and soprovide a stronger mechanical grip.

10.3.2.4 The zone near the endÞtting may not have thesame ßexibility as the rest of the pipe. This zone, correspond-ing to the length of a couple of turns of the reinforcing cables,therefore, does not have the same curvature capacity (ßexibil-ity) as the main pipe section.

10.3.2.5 When bend stiffeners are required at the end ofthe ßexible pipe, they are usually mounted onto the pipe priorto the end Þtting and subsequently pulled up and attached tothe end Þtting once it is mounted. Alternatively, inherent stiff-ness may be introduced into the pipe during the manufactur-ing process by winding on additional elastomer layers.

10.3.2.6 The ßexible pipe may be cured fully or partiallycured prior to mounting the endÞtting. Alternatively, the end

Þtting may be mounted prior to cure and cured with the pipe.The difference in procedures is partially due to the differingtemperature and time required to cure elastomer compoundand epoxy resin.

10.3.3 Curing Process

10.3.3.1 Curing of the elastomer of bonded ßexible pipes isgenerally accomplished by applying heat and pressure in thepresence of curing agents to the pipe. Heat may be applied viaa steam oven or via electrical inductance. Pressure is generallyapplied by wrapping the pipe tightly with nylon prior to cure.

10.3.3.2 During the curing process, the elastomer com-pound will change properties irreversibly and the elastomermaterial making up the pipe cross-section will initially ßowand subsequently form one composite cross-section.

10.3.3.3 A composite cross-section with minimal ßawswill be formed once proper manufacturing procedures areadhered to, followed by manufacturers documented curingprocedures. However, a sample piece, identical in construc-tion to the pipe should be constructed with the pipe, dissectedand inspected for voids in accordance with manufacturersprocedures. The acceptance criterion should be that no visiblevoids are observed.

10.3.4 Tolerances

10.3.4.1 This section provides guidelines on the selectionof manufacturing tolerances (see Section 7.10 of API SpeciÞ-cation 17K for the minimum requirements of the selected tol-erances). The tolerances speciÞed in this section are deÞnedin terms of percentage of nominal values.

10.3.4.2 The length tolerance for bonded ßexible pipeshould typically be Ð0 percent and +1 percent. For certainprojects, there may be additional requirements on the lengthtolerance to be considered, including the following:

a. For certain applications, such as jumpers, the tolerancesmay need to be reduced.

b. Some applications may have problems if the length is toolong, e.g., for long ßowlines a maximum tolerance of +1 per-cent may be too large because of insufÞcient space at the endconnection to accommodate excess length. This may be morecritical for trenched pipe.

c. If two or more risers are clamped together (such as withumbilical in some applications), consideration should hegiven to possible problems caused by the individual risershaving different lengths.

d. The calculation of the required ßowline length shouldaccurately account for all parameters, including undulationsin the route, accuracy of end point locations, installation tol-erances, manufacturing tolerance, and orientation of theßowline to the component (e.g., the pipe may be laid in a loop


around the component, such as at a wellhead, and connectedat a 90 degrees orientation to the main ßowline direction).

10.3.4.3 The recommended tolerance on the ßexible pipeoverall outer diameter is ±3 percent. For carcass layers thatare not manufactured on a mandrel, the tolerance on internaldiameter should be Ð0 percent and +2 percent. For liners thatare not built on an inner carcass, it is recommended that thetolerance on the internal diameter be between Ð0 percent and+2 percent

10.3.4.4 Tolerances should be established and controlledby the manufacturer for each pipe layer. Recommendationson critical aspects of dimensional tolerances for the ßexiblepipe layers are listed in Table 27.

10.3.4.5 The manufacturer should check reinforcementarmor layer tolerances for the allowable gap between adjacentwires or the allowable average gap over a group of wiresagainst manufacturer speciÞcations.

10.4 MARKING

10.4.1 General

Section 10.1 of API SpeciÞcations 17J/17K speciÞes mini-mum requirements for marking of ßexible pipes. The objec-tive of this section is to provide recommendations onadditional markings that may be applied to the pipe. Theseadditional markings will be useful for particular applicationsand can make the pipe and its intended use more identiÞableduring its service life.

The marking system should be sufÞcient to resist installa-tion and operational abrasions, with letters and numbers atleast 10 mm high. All markings should be sufÞciently clear tobe read and/or recognized, in situ, by a remotely operatedvehicle (ROV), and be suitable for the required service life inthe design environment. This does not apply to markings thatare only required for installation purposes (e.g., circumferen-tial bands for length measurement or for clamp or buoyancylocations) and therefore only need to be sufÞcient to resist theinstallation procedures.

10.4.2 Flexible Pipe

10.4.2.1 Nameplates (AISI 316 material is recommended)should be securely attached to both ends of the pipe. Thenameplate should not be covered by any ancillary component,such as bend stiffeners or bend restrictors. In addition to therequirements of API SpeciÞcations 17J/17K, it is recom-mended that consideration be given to including the markingslisted in Table 28.

10.4.2.2 To allow identiÞcation of the length of the pipe,length measurements, typically every 10 m, should be markedon the pipe, highlighted by a colored circumferential band allaround the outer sheath. The length markings should indicatethe direction of the length measurement.

10.4.2.3 For riser applications, the following markingrequirements may also be considered:

a. Unique and logical markings applied to identify differentrisers or the locations for the attachment of any ancillaryitems, such as clamps or buoyancy modules.

b. If applicable, the location of the seabed touchdown pointshould be marked.

10.4.3 End Fittings

In general, the nameplate with the pipe markings isattached to the end Þtting and applies to both pipe and endÞttings. Separate markings are therefore not generallyrequired for the pipe and end Þtting. If there is the possibil-ity of the end Þtting being replaced, consideration should begiven to the markings listed in Table 28 for the end Þtting.Special care should be taken to ensure that identiÞcationmarkings do not damage any surface anti-corrosion treat-ment on the end Þtting.

10.4.4 Connectors and Flanges

Marking requirements for connectors, ßanges, and associ-ated components should be as speciÞed in API SpeciÞcation6A.

10.5 STORAGE

10.5.1 General

10.5.1.1 Flexible pipe can be stored in a number of ways,with the most common being reels, baskets and crates, or pal-lets. Reels and baskets in particular should be marked suchthat the manufacturer, serial number, ßange and drum diame-ters, width, empty weight, and weight capacity are identiÞed.

10.5.1.2 The ßexible pipe should be stored under environ-mental conditions that do not affect its performance charac-teristics. In particular, the following is recommended:

a. The storage temperature should be within the acceptablelimits of the ßexible pipe structure and its end Þttings.

b. The end Þtting connections should be protected to preventdamage of the seal area, threads, and other areas susceptibleto damage. The strapping of the end Þtting should ensure thatit can not become loose and possibly damage the pipe. Secur-ing of the end Þtting should not damage the pipe byoverbending the section adjacent to the Þtting.

c. For materials sensitive to sunlight, the ßexible pipe shouldbe covered to prevent degradation by ultraviolet radiation.

d. End-cuts of ßexible pipe should be covered for long-termstorage.

e. If ßexible pipe is stored for a long period of time after hav-ing been pressure tested, the possible effect of the test ßuid onthe ßexible pipe materials should be taken into consideration.


Table 26—Critical Aspects in Selection of Unbonded Flexible Pipe Manufacturing Tolerances

Recommendations on

Layer Thickness Layer Diameter (Inner/Outer) Other Parameters

Internal Carcass

The minimum value should meet the design requirements of Table 7 in API Spec 17J, considering the potential for erosion/corrosion over the service life. The strip thickness should be controlled by the man-ufacturerÕs material speciÞcation.

The minimum ID should ensure clear pas-sage for equipment such as gauging pigs. The maximum OD should consider the effect on collapse resistance and tolerance buildup of the other layers.

The maximum ovality should be less than that used in the calculation of collapse resis-tance.

Internal Pressure Sheath

The minimum thickness should be deter-mined based on the requirements of API Spec 17J, Section 5.3.2.1.

The maximum OD should consider the effect on hoop strength of the pressure armor layer in accordance with API Spec 17J, Section 5.3.2.5.

Surface Þnish and texture to be controlled such that potential defects do not occur which could propagate through the layer thickness.

Pressure Armor Layer

Should be controlled by the manufacturerÕs material speciÞcation. The minimum thickness should consider the effect on hoop strength in accordance in with API Spec 17J, Section 5.3.2.5.

The maximum OD should consider the effect on hoop strength in accordance with API Spec 17J, Section 5.3.2.5. Variations in OD with length should consider the load sharing along the length in a tensioner installation.

The OD should be controlled such that gaps between the pressure armor layer and the internal pressure sheath do not affect the load sharing between the carcass and pressure armor layer under external radial compres-sion and hydrostatic loading. The maximum gap should assure utilization is as speciÞed in API Spec 17J, Section 5.3.2.5.

Intermediate Sheath/Anti-Wear Layers

In dynamic applications, the minimum thickness should ensure that the sheath does not wear through over the service life. Where the intermediate sheath is to bear hydrostatic loading, the minimum thick-ness should ensure that the layer is not breached (lose pressure integrity) over the service life.

The maximum value should consider the effect of tolerance build-up on subsequent layers.

Tensile Armor Layer

Should be controlled by the manufacturerÕs material speciÞcation. The minimum thickness should consider the effect on hoop and axial strength in accordance with API Spec 17J, Section 5.3.2.6.

The maximum diameter should consider the effect of tolerance build-up on subse-quent layers, and ensure that the tensile wires lay ßat against the pipe.

Variations in lay angle should assure that allowable utilization is in accordance with API Spec 17J, Section 5.3.2.5.The maximum gap between wires should be determined considering the effect of circum-ferential stress concentration in the pressure armor (local bending of the pressure armors within the gaps). Where no pressure armor is present, the maximum gap should be deter-mined based on the requirements of API Spec 17J, Section 5.3.2.1.

Insulation Layer

Should be controlled by the manufacturerÕs material speciÞcation. The minimum thickness should give an overall heat trans-fer coefÞcient for the pipe, smaller than the speciÞed maximum.

The maximum outer diameter should con-sider the effect of tolerance build-up on subsequent layers, and ensure that the insu-lation lays ßat against the pipe.

Outer Sheath

The minimum thickness should assure watertight integrity over the service life, including at the end Þttings. Shear transfer to the underlying layers during installation with a tensioner should also be considered. The variation in thickness along the length of a pipe should consider the effect of stress concentration and possible thinning during installation.

The maximum outer diameter should con-sider the effect on packaging, installation loading, hydrodynamic loading and attach-ment of ancillary equipment such as buoy-ancy clamps.

External Carcass

The minimum thickness should consider the requirement for abrasion and impact protection in the speciÞc application.

The maximum outer diameter should con-sider the effect on packaging, installation and hydrodynamic loading.


f. Long-term pipe storage may cause a permanent curvatureset of the pipe because of the polymer layers. This may needto be accounted for in installation planning.

10.5.1.3 Product handling while in storage should be keptto a minimum. A full and thorough inspection program forthe ßexible pipe while in storage should be performed.Inspection reports should be provided to the purchaser.

10.5.1.4 Repairs carried out while in storage should beperformed under permanent or temporary cover along withthe environmental control facilities normally provided dur-ing manufacture. Work carried out in the storage area shouldbe strictly controlled and performed in such a manner as tocause no damage or contamination to stored products. Thestorage area should be subject to purchaser acceptance andshould be in a location where the pipe will not be suscepti-ble to damage.

10.5.2 Reels

10.5.2.1 Reels rotated around a horizontal axis are the sup-port most commonly used for storage of ßexible pipe in longlengths. Reels, when driven by a winch system, can also beused to maintain the ßexible pipeÕs tension during installationand recovery. The tension applied to the pipe during reelingshould be sufÞcient to prevent the pipe being stored slack,which can damage the pipe during subsequent unreeling. Theparameters to be considered in selecting storage reels for ßex-ible pipe include the following:

a. The drum radius should meet or exceed the storage MBRrequirements of the ßexible pipe.

b. The size of the reel should accommodate the length ofßexible pipe, including end Þttings and accessories.

Table 27—Critical Aspects in Selection of Bonded Flexible Pipe Manufacturing Tolerances

Layer

Recommendations on

Thickness Layer Diameter (Inner/Outer) Other Parameters

Internal Carcass

The minimum value should meet the design requirements of Table 7 in API SpeciÞcation 17K, considering the potential for erosion/corro-sion over the service life. The strip thickness should be controlled by the manufacturerÕs material speciÞcation.

The minimum ID should ensure clear passage for equipment such as gauging pigs. The maximum OD should consider the effect on collapse resistance and tol-erance build-up of the other layers.

The maximum ovality should be less than that used in the calculation of collapse resistance.

Liner The minimum thickness should be determined based on the requirements of API SpeciÞcation 17K, Section 5.3.2.1.

The maximum OD should consider the effect of tolerance build-up on subse-quent layers.

Surface Þnish and texture to be controlled such that potential defects do not occur that could propagate through the pipe body.

Reinforcement Armor Layer

Should be controlled by the manufacturerÕs material speciÞcation. The minimum thickness should consider the effect on hoop and axial strength in accordance with API SpeciÞcation 17K, Section 5.3.2.4.

The maximum diameter should consider the effect of tolerance build-up on subse-quent layers.

Variations in lay angle should assure that allowable utilization is in accordance with API SpeciÞcation 17K, Section 5.3.2.4.

Insulation Layer

Should be controlled by the manufacturerÕs material speciÞcation. The minimum thickness should give an overall heat transfer coefÞcient for the pipe smaller than the speciÞed maximum.

The maximum outer diameter should consider the effect of tolerance build-up on subsequent layers, and ensure that the insulation lays ßat against the pipe.

Cover The minimum thickness should assure water-tight integrity over the service life, including at the end Þttings. Shear transfer to the underlying layers during installation with a tensioner should also be considered. The variation in thickness along the length of a pipe should con-sider the effect of stress concentration and possi-ble thinning during installation.

The maximum outer diameter should consider the effect on packaging, instal-lation loading, hydrodynamic loading, and attachment of ancillary equipment, such as buoyancy clamps.

External Carcass

The minimum thickness should consider the requirement for abrasion and impact protection in the speciÞc application.

The maximum outer diameter should consider the effect on packaging, instal-lation, and hydrodynamic loading.


c. The structure of the reel should be capable of safely sup-porting the weight of the ßexible pipe and its contents.

d. If the reel is to be used for offshore installation, its dimen-sions, structural design, and construction should account forthe loads induced by the vessel motions and the ßexible pipetension during installation and recovery.

10.5.2.2 In the fabrication of reels, particular attentionshould be given to ensuring that all surfaces in contact withthe ßexible pipe are free of any sharp edge, burr, or cut whichmight damage the ßexible pipe. This also applies to partitionswhen used to subdivide reels into separate sections.

10.5.3 Baskets

Baskets or carousels rotated around a vertical axis are fre-quently used for the storage of ßexible pipe in very long

lengths. Baskets are normally used only for storage and arenot capable of supporting any signiÞcant tension in the ßexi-ble pipe. Therefore, a tensioning system is generally requiredfor installation of ßexible pipe from a basket. Design parame-ters and fabrication requirements are otherwise similar tothose of reels.

10.5.4 Crates/Pallets

Crates or pallets are commonly used for storage of ßexiblepipes in short lengths, either straight or coiled. If stored incoil, the storage MBR criteria for the ßexible pipe should bemet. The ßexible pipe should be tightly secured to the crate orpallet to prevent damage due to abrasion. The crate or palletshould contain no sharp edge, burr, or cut which would dam-age the pipe.

Table 28—Marking Recommendations for Flexible Pipe Products

Mark Flexible Pipe End Fitting Comments

API SpeciÞcation 17J/17K Designation X X Required by API SpeciÞcations 17J/17K.

Serial Number X X Required by API SpeciÞcations 17J/17K. Should ensure full trace-ability of all materials, processes and tests during manufacture.

Manufacturer Name or Mark X X Required by API SpeciÞcations 17J/17K.

Date of Manufacture X X Required by API SpeciÞcations 17J/17K. Month and year.

API Licence Number X X Required by API SpeciÞcations 17J/17K. API licensees only.

API Monogram X X Required by API SpeciÞcations 17J/17K. API licensees only.

Design Pressure X X Required by API SpeciÞcations 17J/17K. In MPa units. Specify absolute or differential pressure.

Storage MBR X NA Required by API SpeciÞcation 17J/17K.

Sweet or Sour Service Applications X X Designated by letters SW (sweet) or SO (sour).

Static or Dynamic Application X X Designated by letters S (static ßowline, riser or jumper) or D (dynamic riser or jumper).

Internal Diameter X X In mm units.

External Diameter X NA In mm units.

Design Temperatures X X Min. and max. design temperatures in ¡C.

Length X X Length of ßexible including end Þttings.

End Fitting Condition NA X Designated by letters OEF (Original End Fitting) or REF (Replaced End Fitting)

Notes:1. Imperial units (inches, psi, and ¡F) may be given in brackets after the SI units.2. The marking for the pipe and end Þtting may be covered by a single template attached to the end Þtting.


11 Handling, Transportation, and Installation

11.1 SCOPE

This section provides guidelines and recommendations forhandling, transportation, and installation of ßexible pipe sys-tems. The installation section addresses general consider-ations and describes sample installation procedures and Þnalcommissioning.

11.2 HANDLING

11.2.1 General

11.2.1.1 Precautions should be taken during handling andtransportation of ßexible pipe to prevent damage, as follows:

a. When ßexible pipe is to be transferred from reel (or bas-ket) to reel (or basket), precautions should be taken to ensurethat it will not be damaged by dragging on the ßoor or againstsharp edges of handling equipment or by unacceptable tor-sional/bending loading as a result of improper procedures.

b. The ßexible pipe should be securely fastened to its sup-porting reel, basket, or crate. The end Þttings will usuallyrequire additional fastening by means of wire ropes, Þberslings, bands, adjustable lever hoists, or clamps, as well asprotection with a soft packaging material, in order to protectadjacent pipe layers and to take up any creep or subsequentmotion.

c. Handling and lifting appliances used for ßexible pipesboth onshore and offshore, whether temporary or permanent,include items such as the following:

a. Cranes and A-frames.

b. Reels, carousels, baskets, and strip-out pallets.

c. Lifting frames and cradles.

d. Caterpillars/tensioners.

e. Pulling heads.

f. Winches.

g. Load cells.

h. Chutes and bend limiters.

i. Spreader beams and bars.

j. Tirfors and Òcome-alongs.Ó

k. Lifting ropes, slings, and webbing straps.

l. Chinese Þngers.

m. Control lines.

n. Shackles.

o. Sheaves.

p. CaribinaÕs.

q. Lifting eyes.

All handling equipment should meet the following require-ments and additional best offshore working practices:

a. Used in accordance with the rules and regulations of rele-vant international or national standards. CertiÞcationrequirements may apply.

b. Protected from damage and deterioration while not in use.

c. Inspected for signs of damage and deterioration prior touse.

d. Designed and speciÞed for dynamic applications whenintended for offshore use.

11.2.2 Steel Pipe-Lay Tensioners and Equipment

11.2.2.1 If steel pipe-lay tensioning or other type of equip-ment which is not especially designed to handle ßexible pipeis to be used for the installation of ßexible pipes, it should bedocumented by detailed calculation that the crushing loads onthe pipe do not exceed the design requirements of API Speci-Þcation 17J/17K. The tensioner compression force shouldalso be shown to be sufÞcient to resist the tension in the pipe.

11.2.2.2 As a principle, the calculations should be veriÞedby trials of either the actual equipment or a shoe and loadingconÞguration which consistently simulates the actual equip-ment used and that the relevant installation loads are simu-lated or validated by representative test/use of the equipment.

11.2.3 Reels, Carousels, Baskets, and Strip-Out Pallets

If appropriate, support and drive frames, shoes, cradles,and bobbins forming a part of an assembly should bedesigned and certiÞed for offshore dynamic applications,including lifting both individually and as an assembly. Poten-tial damage or collapse of pipes on reels and carouselsbecause of excess overlying weight should be assessed whererelevant. The drive facility when used for installation reelsand carousels should be Þtted with the following facilities:

a. Fully controllable braking.

b. Manual override for automatic tensioning devices.

c. Back tensioning facility, e.g., for re-reeling.

11.2.4 Overboarding Chutes—Rotating and Fixed

11.2.4.1 Fixed or rotating bend limiters (such as archesand chutes) to be employed as installation or handling aidsshould be designed as recommended by the ßexible pipemanufacturer in accordance with relevant international ornational standards. All such equipment should be maintainedin good condition. Surfaces which will come into contactwith the ßexible pipe should not be corroded or abrasive andshould be free from sharp edges. Wetting of the chute may beused in some cases to reduce the friction with the pipe.


11.2.4.2 When tensions or other installation parameters aresuch that an overboarding chute may damage any structuralor component part of a ßexible pipe, a larger diameter rolleror conveyor, sheave, or other type of equipment should beused in its place. Alternatively, the vertical lay system couldbe employed. A stinger constituting a number of small rollersis generally not acceptable.

11.2.5 Chinese Fingers

If used, Chinese Þngers should be selected with due con-sideration for the ßexible pipe materials, and acceptance forthe selected design should be obtained from the ßexiblepipe manufacturer. Chinese Þngers should have a suitableÞnish to prevent pipe cover damage when used for ßexiblepipe installations.

11.3 TRANSPORTATION

11.3.1 General

11.3.1.1 This section includes any movement of a par-tially or fully manufactured product which is not a normalpart of the manufacturing procedure. The transportationfacility should be selected to minimize handling and oppor-tunity for damage. If use of craneage is required, it shouldbe fully certiÞed and rated in accordance with the liftrequirements.

11.3.1.2 The manufacturer and purchaser should satisfythemselves of the validity of travel authorization prior totransportation. If transportation involves international travel,due regard should be given to all rules and regulationsimposed by relevant countries en route.

11.3.2 Load-Out

11.3.2.1 Load-out covers the period from immediatelyprior to lifting or transferring ßexible pipes on-board a vesselup to and immediately after the vessel leaves the quay side.All ßexible pipes should be visually inspected prior to andduring load-out. Such inspection should be carried out by themanufacturer, purchaser, and installation or transport repre-sentatives, where employed. The inspection should be fullydocumented and signed off by the above parties.

11.3.2.2 All ßexible pipes should be packed and han-dled in accordance with the requirements of Section 10.2of API SpeciÞcations 17J/17K, and further protectedagainst deck activities where necessary. Such protectionand packaging should remain in place during load-out. Thetransportation vessel should not be permitted to leave thequay-side until the purchaser has issued a load-out accep-tance certiÞcate, unless otherwise agreed by the manufac-turer and purchaser.

11.3.3 Sea Fastenings

Sea fastenings should be designed for the Þnal transportedweight in a dynamic environment appropriate for the trans-portation vessel and the sailing route. All sea fasteningsshould be fully certiÞed in accordance with the appropriatedesign code prior to sail-away. All designs should beapproved by purchaser prior to load-out.

11.3.4 Reeled Flexible Pipe

11.3.4.1 Reeled ßexible pipe in this context covers ßexiblepipe which is on a reel, carrousel, or basket. Flexible pipesshould not be placed on a reel so that end Þttings or otherattachments induce unacceptable local loading in the pipestructure. End Þttings or attachments that are not wrappedand packed should not be over-wrapped with unprotectedpipe.

11.3.4.2 Weights should be accurately monitored andrecorded during lifting, either with load cells certiÞed inaccordance with established practice, or crane gauges, wheresuch gauges have been individually certiÞed. When liftingreels in a drive or support frame, the reel should be Þxed toprevent rotation prior to lifting. If relevant, the reel shouldclearly identify that the pipe is full of ßuid and the effect ofthe ßuid weight on the total weight.

11.3.5 Coiled Flexible Pipe

Coiled ßexible pipe covers all pipes loaded out and securedon deck in coiled condition, either packaged or unpackaged.The ßexible pipes should be coiled so that removal of storagestraps will not result in uncontrolled release. Coiled ßexiblepipes should be suitably sea fastened prior to sail away. Decklocation should be such that potential hazards are minimizedduring over-boarding.

11.3.6 Uncoiled Flexible Pipe

11.3.6.1 Uncoiled ßexible pipe covers all ßexible pipessecured on deck neither reeled nor coiled. The ßexible pipesshould be provided with suitable protection from draggingover dockside surfaces.

11.3.6.2 The ßexible pipes are to be located on deck withinreach of the deck crane, or lifting facility, so that draggingacross the deck and lifting around objects during subsequentinstallation is minimized. The ßexible pipes should be suit-ably sea fastened and provided with protection from normaldeck activities prior to sail away.

11.4 INSTALLATION

11.4.1 Installation Analysis

11.4.1.1 The installation analysis should take into accountcontingency scenarios. Dynamic installation analyses should


be used to deÞne the maximum seastate and current proÞlesuitable for deck and installation activities on the particularvessel. The loads applied in the analyses should be for themaximum deÞned seastate for the planned activities.

11.4.1.2 If tensioners are used, the installation load casesshould check that minimum and maximum tensioner loads donot violate the pipe design criteria. The maximum load (withpipe hang-off tension) should be checked for potential col-lapse of the pipe, while the minimum tensioner load shouldbe greater than the force required to prevent the pipe slipping(Fmin), deÞned as follows:

(9)

where

Fmin = minimum tensioner load required to hold the pipe,

T = maximum tension in pipe,

µ1 = friction coefÞcient between pipe outer sheath and tensioner pads,

µ2 = friction coefÞcient between pipe outer sheath and underlying armor layer.

11.4.2 Monitoring

The subsea activities should be constantly monitored usingdiver and/or ROV mounted cameras as approved by the clientand installation contractor. The monitor recordings should bestored for review of subsea activities after installation has Þn-ished. The recordings should identify all visible markings,conÞrm lay patterns and conÞgurations, and status of boltedßanges, connectors, bend restrictors, bend stiffeners, andbuoyancy modules. All recordings should be stored with a logand uniquely marked for storage and retrieval.

11.4.3 Installation of Reeled Flexible Pipes

Whenever possible, deployment reels should be placeddirectly in line with overboard chutes. The use of rollers, sin-gle point attachments, or sheaves should not induce unaccept-able loads on the ßexible pipe structure; pipe deßection unitsmay be used provided the MBR criterion is met. Single-pointcontacts should be minimized. Detailed calculations shouldbe carried out to ensure that no unacceptable loads areinduced at any contact point.

11.4.4 Installation of Carouselled Flexible Pipes

The recommendations in Section 11.4.3 also apply to ßexi-ble pipes on carousels.

11.4.5 Installation of Coiled Flexible Pipes

Storage straps should be replaced by temporary deploy-ment rigging prior to deployment of coils overboard unlessthe storage straps can be used for installation. When possible,the ßexible pipe should be coiled on a rotating pallet and thestrip out rigging should have a suitable swivel. The craneshould slowly raise the pipe to a vertical position, allowing itto release any inherent twist through the swivel. Diversshould not use sharp tools for removal of temporary deploy-ment rigging.

11.4.6 Installation of Uncoiled Flexible Pipes

Uncoiled ßexible pipes should be lifted overboard with acrane using a multiple point lift. If over boarding chutes andwinches are used, then care should be taken to ensure that nodamage is caused to the ßexible pipe and/or end Þttings. Thepipe may also be laid out straight on the deck and picked upby one end. In this case, the installation procedures shouldensure that the MBR criteria are not exceeded.

11.4.7 Deployment and Tie-In

11.4.7.1 Loads and deformations during deploymentshould be within allowable limits. Bend radii should be moni-tored during installation or the installation method and layingparameters deÞned to ensure the MBR criterion is notexceeded, e.g., by monitoring the seabed touchdown pointwith an ROV and using a transponder to maintain a minimumlayback distance, thereby ensuring the conÞguration does notexceed the MBR criteria. If feasible, pull-in wires (or weaklinks if used) should be such that they break before damage issustained to the ßexible pipe as a result of excessive tension.Flexible pipes should not be over tensioned during deploy-ment through a steel pipe or J-tube, while accounting for themaximum friction force from the pull-in. Back tension will berequired during these operations.

11.4.7.2 The tie-in sequence should be arranged such thatminimal inhibited ßuid is lost after blind ßanges are removed,unless ßooding with inhibited water is carried out immedi-ately after tie-in. In general, ßexible pipes should not be laidaround obstacles such that natural movement is restricted.This may be acceptable, however, if the procedures, equip-ment, and ßexible pipe is designed for the application. Theuse of scour mats should be considered in preference to phys-ical restriction if scour is considered a problem.

11.4.7.3 It is recommended that ßowlines be connected totheir termination point (e.g., wellhead, manifold) at rightangles to the main lay direction. This allows excess lengthsand expansions of the line to be absorbed in the Þnal loop atthe connection point. This Þnal loop may also be used if thereis an underestimation of the ßowline length.

Fmin maximum Tµ1

----- , Tµ2

----- =


11.4.8 Trenching and Burial

If an installed ßexible pipe is expected to become buriedin soft seabed conditions, a pipe tracking facility should beincorporated to facilitate route conÞrmation at a later date.If a ßexible pipe enters a trench in hard seabed conditions orpasses over a boulder within the trench, suitable sand bag-ging or some such method should be provided to support thepipe over sharp edges or corners in the event the MBR crite-ria could be violated or if the outer sheath cover could bedamaged.

11.4.9 Vessel and Equipment

11.4.9.1 The vessel and equipment should be in good con-dition and working order and be checked prior to vesselmobilization. All measurement equipment, particularly formeasuring load, should be calibrated. All lifting equipmentshould have suitable certiÞcation.

11.4.9.2 Where pipe tension is to be distributed betweentensioners, reel drives, and carousel drives, the installationprocedures and control systems should be sufÞcient to ensurecontrol of the tension in the pipe.

11.4.9.3 Typically, the vessel spread should include thefollowing equipment for monitoring the ßexible pipe duringinstallation:

a. ROV for conÞguration.

b. Tension measuring equipment for maximum top tension.

c. Departure angle measuring equipment.

d. Compression load measurement for caterpillar tensioners.

11.4.10 Installation Procedures

11.4.10.1 General

11.4.10.1.1 The installation procedure employed for eachßexible pipe is dependent on the system conÞguration and theparticularities of the system components. In the sample instal-lation procedures in this section, horizontal installation usingan overboarding chute is shown. Vertical installation may alsobe used. Schematics of both are shown in Figures 30 and 31respectively.

11.4.10.1.2 The ßexible pipes may be installed eitherßooded, free-ßooding, or empty. The manufacturer andinstallation contractor should determine the installation con-ditions. Some pipes may require to be installed ßooded orfree-ßooding to prevent collapse of the pipe, or to ensure thestability of the installed line. In this case, the suitability of thecarcass material (for rough bore structures) should be con-Þrmed with the manufacturer.

11.4.10.1.3 In determining the installation strategy to beused, some of the issues which need to be addressed and

which may inßuence schedule and risks include the fol-lowing:

a. Pre-installation of risers prior to hook-up.

b. Number and size of ancillary components, including buoy-ancy, to be installed.

c. Type of bases, if any, to be used and anchoring system(gravity, pile, or suction).

d. Tension in line.

e. Tie-in systems, such as riser/ßowline connections.

f. Maximum environmental conditions (installation window).

g. Interfaces with installation of other systems, such as moor-ing lines.

h. Diver assisted or diverless operations.

i. Installation vessel requirements, including number, size,and mobilization/demobilization costs.

j. Trenching and/or protection requirements.

k. Installation of bundles or multiple lines.

l. Subsea versus topside operations.

m. IdentiÞcation of components/equipment to be installedonshore to minimize offshore operations.

n. ROV operations.

11.4.10.2 Flowlines

A typical installation procedure for a ßexible ßowline ispresented in Figure 32. The ßowline is attached to a pile orclump weight in the vicinity of the start ßowline base and islaid out along the seabed towards the end ßowline base. TheÞnal portion of the ßowline is laid out in an overlengthshape. Inßatable buoyancy units may then be attached to theßowline ends, which are then winched into the ßowlinebases for connection. An example installation of a ßexibleßowline through a J-tube is shown in Figure 33. For a J-tubepull-in, a pre-installed sealing plug may be used to seal theJ-tube at the lower bellmouth so as to prevent loss of corro-sion inhibitors.

11.4.10.3 Riser Configurations

Typical installation procedures for ßexible riser conÞgura-tions are shown in Figures 34 to 38, respectively for lazy-S,steep-S, lazy wave, steep wave, and free hanging catenaryconÞgurations. These Þgures show the ßexible pipe beinginstalled with the Þrst end connected to the vessel. Thismethod may not suit all applications and can be reversed. Thevessel is represented schematically as a semi-submersible butthis is of no consequence with regard to the actual installa-tion. Many installers would prefer to handle ßexibles, buoys,and clump weights separately.


Figure 30—Schematic of Horizontal Lay Installation

Low tension (T<20t)

High tension (T>20t)

Flexible pipe

Installationreel

Laying chute Installationvessel

Installationvessel

Laying chute

MWL

Flexible pipe

Tensioners

Installationreel

Notes:�1. For low tension systems, holdback tension is provided by the installation reel or a winch.�2. For high tension systems, the pipe is kept slack behind the tensioners.

MWL


Figure 31—Schematic of Vertical Lay Installation [7]

Installationderrick

Caterpillartensioners

Buoyancy modules,anodes, etc. fitted atthis point

Flexible pipe

Vesselmoonpool


Figure 32—Representative Flowline Installation Procedure

MSL MSL

Flowlinebase (FB) FB

FBFB

FB FB

FB

FB

FB

Pull-in wire

Flexible pipe

Elevation View Elevation View

Plan View Plan View

Plan View Plan View

1. Overboard first flange to seabed 2. Layout pipe

3. Overboard end flange to seabed 4. Pull-in pipe ends

Pull-in wire

Pull-in wire


Figure 33—Schematic of J-Tube Pull-In Operation

Platform Platform

Platform

Platform

Platform

Installationvessel

Installationvessel

Installationvessel

Installationvessel

Installationvessel

J-tubeMessengerline

Pull-inwire

Pull-inwire

(a) (b)

(c) (d)

(e)

Flexiblepipe

Flexiblepipe

Flexiblepipe

End fitting

Notes:�(a)�Installation vessel moves up to the platform.�(b)Preinstalled messenger line followed by a�� pull-in wire is transferred to the vessel from�� the platform.�(c)�Pull-in wire attached to end of flexible pipe�� and pull-in operation begins.�(d)When end of flexible pipe reaches the top of�� the J-tube, the end fittings are attached to a �� hang-off structure.�(e)�The installation of the flexible pipe is then�� continued in a layaway operation.


Figure 34—Representative Lazy-S Riser Installation Procedure

DSV

DSV

DSV

Pull-in winch

Pull-in wire

Riser end fitting (REF)

Layout Wire

REF

Clump weight

SeabedSeabed

SeabedSeabed

1. Pull-in riser end fitting (REF) 2. Overboard subsea buoy/arch system

3. Overboard riser lower end flange 4. System as installed

FPSFPS

REFREF

Riser flange

Riser flange

Flexible riser

Notes:�1.� The above procedure is based on connecting to the FPS firstly and then laying away from the�� FPS. The procedure may also be reversed.�2.� The horizontal lay procedure may be replaced with a vertical lay procedure.�3.� Many installers would prefer to handle flexibles, buoys, and clump weights separately.

FPSFPS


Figure 35—Representative Steep-S Riser Installation Procedure

DSV

DSV

DSV

Pull-in winch

Pull-in wire


REF

SeabedSeabed

SeabedSeabed

1. Pull-in riser end fitting (REF) 2. Overboard subsea buoy/arch system


FPS

FPS

FPS

FPS

REFREF

Riser flange

Flexible riser

Notes:�1.� The above procedure is based on connecting to the FPS firstly and then laying away from the�� FPS. The procedure may also be reversed.�2.� The horizontal lay procedure may be replaced with a vertical lay procedure.�3.� Many installers would prefer to handle flexibles, buoys, and clump weights separately.

Riser flange

Pull-in wire


Figure 36—Representative Lazy Wave Riser Installation Procedure

DSV

DSV

DSV

Pull-in winch

Pull-in wire


REF

SeabedSeabed

SeabedSeabed

1. Pull-in riser end fitting (REF) 2. Overboard mid-water buoyancy modules


FPSFPS

FPSFPS

REFREF

Riser flange

Riser flange

Flexible riser

Notes:�1.� The above procedure is based on connecting to the FPS firstly and then laying away from the�� FPS. The procedure may also be reversed.�2.� The horizontal lay procedure may be replaced with a vertical lay procedure.


Figure 37—Representative Steep Wave Riser Installation Procedure

DSV

DSV

DSV

Pull-in winch

Pull-in wire


REF

SeabedSeabed

SeabedSeabed

1. Pull-in riser end fitting (REF) 2. Overboard mid-water buoyancy modules


FPSFPS

REFREF

Riser flange

Pull-in wire

Riser flange

Flexible riser


FPSFPS


Figure 38—Representative Free-Hanging Catenary Installation Procedure

DSV

DSV

DSV

Pull-in winch

Pull-in wire


REF

SeabedSeabed

SeabedSeabed

1. Pull-in riser end fitting (REF) 2. Lay out riser catenary


MSLMSL

REFREF

Riser flange

Riserflange

Touchdownpoint

Flexible riser


FPSFPS


11.4.11 Diverless and Diver Assisted Installation

The selection of diver assisted or diverless installation willdepend on a number of factors, including the following:

a. Safety aspects.

b. Water depth.

c. Regulatory requirements or guidelines.

d. Available space for tie-in operations, e.g., if a large num-ber of risers are to be connected to a turret there may beinsufÞcient space for divers.

e. Economic factors (diverless tie-in equipment may havesigniÞcant costs).

f. Environmental conditions.

g. Equipment reliability (technical risks).

h. Schedule requirements, e.g., diverless operations may bemuch quicker.

11.5 PRE-COMMISSIONING/COMMISSIONING

11.5.1 Introduction

11.5.1.1 This process involves the testing and monitoringof ßexible pipes after tie-in and completion of the full system,of which the ßexible riser and/or ßexible ßowlines are anintegral part. If the ßexible pipe incurs damage during thecommissioning period, the damage should be repaired andthe commissioning should be restarted. The decision onwhether the pipe is repairable should be taken in consultationwith the pipe manufacturer and the purchaser.

11.5.1.2 The purchaser should provide the test speciÞca-tion. The manufacturerÕs recommendations on testing shouldbe taken into account, and the testing should be carried outprior to any backÞlling.

11.5.2 Pigging

11.5.2.1 If commissioning requires pigging of the ßexiblepipe the guidelines in this section should be implemented.Metallic brushes should not be used in ßexible pipes withouta metallic carcass layer. Metallic brushes may be used wherethe internal liner comprises a steel carcass, provided thematerials are compatible and the brush does not damage thecarcass. Metallic scrapers should not be used.

11.5.2.2 Gauges may be used, provided the discs aredesigned such that any obstruction protruding within thegauged diameter will be indicated by a permanent deforma-tion. The gauge plate should be approved by the ßexible pipemanufacturer. See API SpeciÞcations 17J/17K, Section9.2.1.2.

11.5.2.3 Articulated pigs should only be used where thenatural weight of the pipe or installed imposed bend radiusis sufÞciently large to accommodate the segment lengths in

the pig assembly. Foam pigs should be used for pipes with-out a metallic carcass layer where possible, but other typesof pig may be used subject to acceptance by the ßexiblepipe manufacturer.

11.5.3 Hydrostatic Pressure Test

11.5.3.1 General

11.5.3.1.1 The hydrostatic test may be performed sepa-rately on the ßexible pipe or as a system test if the ßexiblepipe is part of the total system. The pipe system may includemanifolds, trees, valve assemblies, couplings, seals, etc. Allcomponents in the system should be veriÞed as being capableof withstanding the maximum test pressure. Where relevant,the installation test procedure should be in accordance withthe requirements of API SpeciÞcations 17J/17K, Section 9.3(hydrostatic pressure test).

11.5.3.1.2 The hydrostatic test should be in accordancewith the following recommendations:

a. If the ßexible pipe is installed without the occurrence ofany suspected damage, then it will only be necessary to per-form a leak test, as a structural integrity test will have alreadybeen performed (i.e., FAT hydrotest as per Section 9.3 of APISpeciÞcations 17J/17K). The recommended leak test pressureis 1.1 times the design pressure.

b. A structural integrity test may be required if the pipe hasbeen damaged, repaired, end Þttings replaced, retrieved, andre-installed without a FAT hydrotest, or other such occurrencewhich may be considered relevant. The recommended struc-tural integrity test pressure is 1.25 times the design pressure.

c. Unless otherwise recommended, the hold period for thetest should be 24 hours (see section 11.5.3.5).

d. Regulatory authority requirements may exceed the recom-mended test pressures in (a) and (b) above and should bechecked with the relevant authorities.

e. The ßexible pipe design should be checked against allow-able criteria for the pressure test load case, including loadsfrom maximum test pressure (which will be between 1.04 and1.1 times nominal, as per Section 11.5.3.3), functional loads(including weight and buoyancy of pipe, contents, and attach-ments), relevant environmental loads, and any appropriateaccidental loads.

11.5.3.1.3 The hydrostatic test procedure should identifythe following as and where applicable:

a. Pre-test pigging requirements.

b. Fill medium details.

c. Pressurization and depressurization rates.

d. Stabilization criteria.

e. Pressure isolation details.


f. Entrapped air assessment.

g. Permissible unidentiÞable pressure loss.

h. Pressure variation calculation method.

i. Visual inspection details.

j. Data recording details.

k. Third party inspection requirements.

l. Acceptance Criteria.

11.5.3.1.4 During the test, all annulus vents in unbondedpipes should be opened in end Þttings which are notimmersed in seawater. The hydrostatic pressure test com-prises the following main tasks:

a. Test of instrumentation and connections.

b. Pressurization of the line.

c. Stabilization period.

d. Hold period.

e. Depressurization.

11.5.3.1.5 Recommendations for these tasks, and accep-tance criteria, measuring equipment and test records are givenin the following sections.

11.5.3.2 Test of Instrumentation and Connections

A pressure test should be performed on the test equipmentand connections at a pressure not less than 104 percent of thenominal test pressure of the ßexible pipe. The duration of thistest is half an hour.

11.5.3.3 Pressurization

Pressurization of the pipe should be carried out at asteady and controlled rate to be speciÞed by the manufac-turer. Too high of a rate can lead to excess stabilization peri-ods. A typical maximum rate is 18 MPa/hour. The pressureshould be raised to a value no greater than 110 percent ofthe nominal test pressure. (Different manufacturers specifyfactors between 104 percent and 110 percent of the nominaltest pressure; any factor within this range is suitable, so longas it is documented and used consistently throughout designand test activities.) The air content should not exceed 0.5percent for smooth bore pipes and 1.0 percent for roughbore pipes. If the air content exceeds the above values, thenventing at the pipe ends should be performed and pressur-ization recommenced.

11.5.3.4 Stabilization

The stabilization period should last for 10 hours after theend of pressurization. This stabilization period may beextended if signiÞcant pressure drops are still occurring afterthe Þrst 10 hours because of the stabilization process or ther-mal stabilization in the ßexible pipe. The period may also be

reduced if the line is stabilized. Stabilization is deÞned as apressure change over one hour of less than 1 percent of thetest pressure. During stabilization, the pressure curve shouldbe recorded and a log of pressure, and subsea and test ßuidtemperatures should be maintained (every half hour for pres-sure readings and every two hours for temperature readings).

11.5.3.5 Hold Period

11.5.3.5.1 When the stabilization period is completed, the24 hour hold period may start. A log of pressure, and subseaand test ßuid temperature readings should be taken at halfhour intervals during the hold period. The pressure must begreater than or equal to the nominal test pressure for the holdperiod. There must be no unaccountable pressure drop duringthe test. The maximum pressure drop during the hold periodshould not exceed 4 percent of the nominal test pressure.

11.5.3.5.2 For a leak test, the hold period may be reducedto 6 hours if all of the ßexible pipe, including both end Þttingscan be visually inspected for leakage during the test.

11.5.3.5.3 Once the test has commenced, should the pres-sure fall below the test pressure, the line should be repressur-ized. In such a case, the hold period is considered asrecommencing from this point.

11.5.3.6 Depressurization

The depressurization of the pipe should be performed at asteady and controlled rate. The maximum depressurizationrate should be deÞned by the manufacturer. Pipe failure canbe caused by depressurization at too high a rate. A typicalmaximum rate is 108 MPa/hour.

11.5.3.7 Qualitative Acceptance Criteria

The following acceptance criteria are recommended as aminimum:

a. The test pressure is maintained for the period speciÞedabove.

b. The test pipe does not undergo unintended or majorchanges in shape or conÞguration under pressure.

c. The pipe does not leak.

11.5.3.7.1 If the pressure loss is excessive such that a leakis suspected, leaks through all components in the pipe systemshould be evaluated, as there is potential that the leak may befrom valves, seals, etc., rather than the pipe itself.

11.5.3.8 Measurement Equipment

Measurement equipment used for pressure testing shouldbe calibrated at least every 6 months. Equipment should bemaintained in good order and used only for the purpose forwhich each item has been designed and intended. Equipment


used should be listed with all relevant details in the test docu-mentation and should be calibrated to within the followinglevels of accuracy:

a. Hydrostatic pressure gauges: +0.0, Ð0.5 percent

b. Dead weight testers: +0.0, Ð0.1 percent

c. Pressure chart recorders: ±0.5 percent

d. All other measurement equipment: ±1.0 percent

11.5.3.9 Test Records

11.5.3.9.1 It is recommended that the following testrecords be maintained:

a. Date and time.

b. Location, condition and situation details.

c. Test and safety personnel.

d. Fill medium details.

e. All equipment and certiÞcation details.

f. Pressure recorder charts showing continuous recordings.

g. Periodic pressure readings, every 30 minutes as aminimum.

h. Periodic ambient temperature readings, every 30 minutesas a minimum.

i. Periodic Þll medium temperature readings, every 30 min-utes as a minimum.

j. Visual observations.

11.5.3.9.2 The test records should be signed by the appro-priate personnel and Þled for reference.

11.5.3.9.3 A post commissioning survey should be carriedout and recorded on video tape to verify that the ßexible pipesystem is installed as designed.

11.5.4 Drying of Pipe

11.5.4.1 In some cases, there may be stringent require-ments on the amount of water that may be left in a ßexiblepipe after the hydrostatic pressure test. An example of this isgas export ßexible risers tied-in to major export lines, whichhave stringent requirements on the dryness of the gas. Arough bore pipe will be required for the riser. With this con-struction the interlocking carcass layer forms a large trap forwater which, subsequent to a hydrotest, could violate the gasdryness requirements. Vacuum drying of the ßexible riser ispotentially a very costly and time consuming operation on thecritical path of a project.

11.5.4.2 A special valve skid could be developed for theseabed end to allow dry installation and tie-in [50]. In addi-tion, the factory hydrotest of the riser could be performed withglycol instead of water, and the riser pressurized with nitrogenduring transportation and installation, to ensure dryness.

12 Retrieval and Reuse

12.1 SCOPE

12.1.1 This section addresses the retrieval of ßexible pipeand reuse at an alternative location. Recommendations areprovided on the inspection and test requirements for the pipeprior to reuse. Note that the retrieval recommendations forwhen the pipe is to be reused also apply where the pipe is tobe retrieved and scrapped.

12.1.2 Consideration should also be given to the recom-mendations in this section for a pipe that is to be retrieved forrepair purposes and reinstalled after repair.

12.2 RETRIEVAL

12.2.1 General

The following sections apply to bonded and unbondedpipe.

12.2.1.1 A ßexible pipe may be retrieved because of thecessation of its usefulness at a particular location or becauseof damage to the pipe. The retrieval operation is essentiallythe reverse of installation. A pre-survey to assess the condi-tion of the pipe should be carried out to highlight any poten-tial problems, such as the following:

a. Pipe burialÑjetting may be necessary to unbury the pipe,so as to avoid kinking the pipe during recovery.

b. Pipe crossings and adjacent linesÑto ensure these are notdamaged by retrieval operations.

c. Hard marine growthÑthis can cut through the outersheath as the pipe comes in contact with layover arches,bending shoes, tensioners, etc.

12.2.1.2 A procedure for pipe retrieval should be preparedto preserve the pipe integrity during the operation. The sameconditions considered in the global and local analysis of theoriginal installation should be used for the pipe retrieval oper-ation (e.g., pipe ßooded or empty, restrictions because ofenvironmental conditions, equipment imposed loads and con-Þgurations considered), as applicable.

12.2.1.3 Local environmental laws and regulations shouldalso be considered. Special care regarding pipe ßuid spillagesshould be taken to avoid pollution. The potential for hazard-ous elements in the pipe, such as radioactive materials, mer-curic compounds, etc., should be evaluated and appropriatesafety procedures and equipment speciÞed. Pipe ßushingwith inhibited seawater and cleaning may be necessary priorto disconnection and retrieval.

12.2.1.4 Risks involving personnel are to be a subject ofspecial review. A HAZID/HAZOP type study should be per-formed for all operations. ParafÞn plugging is a major safetyand environmental hazard. If there is a possibility of parafÞnplugging occurring, it may not be safe to recover the pipe.


12.2.1.5 Procedures for pipe retrieval should foresee howthe pipe will be identiÞed. Proper visual identiÞcation(through ROV, for example) should be used for this purpose.In the case of buried pipe, special procedures will be requiredto avoid possible damage to the pipe or other subsea equip-ment from trawler equipment used for unburying the pipe.

12.2.1.6 All limitations of the pipe during installation andhandling (e.g., MBR, maximum allowable torsion, maximumcrushing load and tension, and winding/unwinding and stor-age recommendations) should be included in the retrieval pro-cedure to avoid damage or failure of the pipe. Considerationshould be given to the pipeÕs aged condition (i.e., reducedstructural capacity) when specifying retrieval criteria.

12.2.1.7 The tensions experienced by the pipe are greaterduring retrieval than installation because of friction on theoverboarding chute. Depending on the tension and riser conÞg-uration, it may be necessary to void the pipe prior to retrieval.

12.2.1.8 The recovery operation may be simulated usingsuitable software. The simulation should take into account rel-evant factors, such as seastate, current proÞle, vessel motions,and possible restrictions to recovery, including burial material(soil, clay, or rocks), protection mats, and structures.

12.2.1.9 Loads, deformations, and abrasions of the pipeshould be monitored at all times during pipe retrieval. As arule, the pipe should be inspected during recovery. Any dam-age should be clearly identiÞed on the pipe outer sheath bymeans of suitable markings. The manufacturer should be con-sulted for cleaning and storage procedures.

12.2.2 Unbonded Pipe

12.2.2.1 The potential for corrosive or toxic ßuids in thepipe annulus should be evaluated. If such ßuids are present,the vent ports or valves in the end Þttings should be immedi-ately plugged on retrieval of the pipe until these ßuids can besafely discharged. One possibility for discharging the ßuids isto pump air or nitrogen into one end Þtting and allow releaseat the other end Þtting.

12.2.2.2 Special care should be taken during retrieval toavoid bursting the outer sheath because of excess differentialpressure between the annulus and exterior of the pipe. Excessdifferential pressure can also cause loosening of the outersheath and may result in problems (including damage to thesheath) if the pipe is retrieved using tensioners or if ChineseÞngers are used (compression created may not be sufÞcient totake tension load through friction). The retrieval rate shouldbe controlled to allow such excess pressure to be bled off atthe end Þtting vent valve during retrieval. If the annulus con-tains toxic ßuids, the pressure release system should be con-trolled to ensure the safety of personnel.

12.2.2.3 The allowable retrieval rate should be calculatedbased on the condition of the gas relief system. Gas relief

valves that have not been operational for a substantial periodmay become stuck because of scale deposition, marinegrowth, corrosion, etc. If feasible, clogged valves should befreed prior to recovery of the pipe. As an alternative, consid-eration may be given to drilling burst discs in the outer sheathprior to recovery to safeguard the integrity of the outer sheath.

12.2.3 Bonded Pipe

12.2.3.1 For ßoat/sink type bonded pipes that ßoat whenempty (i.e., full of air), the pipeline should be retrieved byßoating it to the surface then heaving it onto a reel from thesurface of the water. The manufacturer should be consultedfor de-watering/pigging procedures and limitations.

12.2.3.2 For bonded pipelines that consist of multiplelengths, care should be taken during reeling to protect theadjacent pipe layers from damage due to contact with an endÞtting. The manufacturer should be consulted for dunnagerecommendations related to his particular end Þtting.

12.2.3.3 For smooth bore (i.e., collapsible) bonded pipes,care should be taken during retrieval to avoid capturing exces-sive twist on the reel. Consideration should be given to main-taining a nominal pressure in the pipe bore during retrieval tocontrol twist. If full twists are captured on the reel, theyshould be relieved by transpooling the pipe under internalpressure. Alternatively, if the pipe ßoats when full of air, thetwist can be relieved by pulling the pipe off the retrieval reelonto the water, pressurizing it with air, then heaving it backon to the reel off the surface of the water. The pipe should notbe stored or re-used in the twisted condition.

12.2.3.4 Smooth bore bonded pipe can exhibit high elon-gation due to tension, particularly when the pipe is unpressur-ized. Care should be taken during retrieval to minimize theamount of elongation that is captured on the reel. If excessiveelongation is captured on the reel, the pipe should betranspooled, under internal pressure, to relieve the elongationbefore the pipe is stored or re-used. Alternatively, if the pipeßoats when full of air, the pipe can be pulled off the retrievalreel onto the water, pressurized with air, then heaved back onto the reel off the surface of the water.

12.3 REUSE

12.3.1 General

12.3.1.1 To reuse a ßexible pipe in a new application, it isrecommended that as a minimum the following stages in theprocess be addressed:

a. Documentation.

b. Pipe evaluation.

c. Pipe retrieval.

d. Inspection and repair.

e. Test requirements.

f. Installation.


12.3.1.2 See Section 11 for guidelines on installation andSection 12.2 for guidelines on pipe retrieval. The remainingstages in the process are addressed in Sections 12.3.2 to12.3.5. Note that a retrieved pipe that is designed for staticapplications should not be reused for a dynamic application.Stages (a) and (b) should be performed prior to pipe retrievalto determine if it will be feasible to reuse the pipe.

12.3.2 Documentation

12.3.2.1 The user should maintain a detailed record of pre-vious use so that it will be possible to accurately evaluate thefeasibility of reusing the pipe. The record should specifywater depth, production ßuid characteristics, installation date,length in service, operating pressure and temperature, and anyunanticipated events that might affect the pipe function.

12.3.2.2 Any events that may have damaged the pipe andany previous repairs to the pipe should also be documentedand held as evidence of the pipeÕs service history. In addition,records of all previous inspections and monitoring operationsrelating to the pipe should be maintained.

12.3.3 Pipe Evaluation

12.3.3.1 General

12.3.3.1.1 When a pipe is under evaluation for reuse, thenew design conditions should be deÞned using the purchasingguidelines in Appendix A of API SpeciÞcation 17J/17K. Theßexible pipe to be reused should comply with the pipe struc-ture design criteria speciÞed in Table 6 of API SpeciÞcation17J and Table 7 of API SpeciÞcation 17K for the new designconditions.

12.3.3.1.2 Prior to the pipe reuse, a general review should becarried out considering the pipe design characteristics, the newconditions of use, the remaining pipe service life, and all previ-ous conditions that may have affected its characteristics. Theevaluation should also address any accidental damage foundfrom the pipe inspection after retrieval. The effect of corrosiveßuids on the structural layers of the pipe should be evaluated inthe calculation of the remaining service life. In addition, theaged state and remaining life of the liner or internal pressuresheath/line polymer/elastomer material should be evaluated.

12.3.3.1.3 Pipe veriÞcation and assessment for reuse areaddressed in the subsequent sections for the following reuseconditions:

a. Similar use.b. New conditions.c. Special cases.

12.3.3.2 Evaluation for Similar Use

12.3.3.2.1 In this case, the pipe is to be reused in condi-tions similar to the original application. It does not includesituations in which the pipe was subjected to abnormal occur-rences, damage, or other events that could have signiÞcantly

reduced the service life. The information to be determined forthe evaluation is as follows:

a. The new conditions of use (refer to Appendix A of the APISpeciÞcation 17J/17K), including identiÞcation of any majorchanges in the application (e.g., H2S or CO2 levels).b. The remaining service life.c. The original data speciÞed by the manufacturer, includingpipe capacity (e.g., data sheet and design report).

12.3.3.2.2 If the new conditions of use (including installa-tion/retrieval equipment and procedure, and environmentaland operational conditions) are easily identiÞed as equivalentor less critical than the original conditions or original designcriteria, and if the remaining service life is greater than thelife required for the new location, an inspection of the pipe fordamage should be sufÞcient to approve the pipe for reuse.

12.3.3.2.3 Attention should be given to the procedures andequipment used for installation and retrieval, particularly fordeep water applications where installation conditions can becritical. The installation loads should be conÞrmed to be lessthan the original installation, or alternatively a new analysisshould be performed to conÞrm that the pipe meets the designrequirements speciÞed in API SpeciÞcation 17J/17K and Sec-tion 5 of this recommended practice.

12.3.3.3 Evaluation for New Conditions of Use

12.3.3.3.1 If the new conditions of use are not similar tothe original ones, or if the evaluation carried out according to12.3.3.2 is inconclusive, it is necessary to assess the followingadditional information:

a. New global and cross-section analyses (considering newinstallation equipment, new operational conditions, newapplication, etc.).

b. The results of prototype tests, as available (short- andlong-term tests).

12.3.3.3.2 Liner or internal pressure sheath of ßexible pipeto be reused should be suitable for the new transported ßuidconditions, considering aspects such as chemical compatibil-ity, temperature, gas permeation, and aging. Where available,aging models and methods for determination of elastomer orpolymer residual life should be used in the analysis withappropriate safety margins.

12.3.3.3.3 If sour conditions are foreseen, the metallicmaterials should be qualiÞed for SSC and HIC resistance inthe new design conditions. Elastomer, polymer, and metalliclayer thickness reduction as a result of fretting/abrasion,which may have occurred during previous use, should beproperly evaluated.

12.3.3.4 Evaluation of Special Cases

12.3.3.4.1 Additional analysis may be necessary if thepipe was subjected to abnormal occurrences, damage, critical


stresses, or other events that could have signiÞcantly reducedthe service life of the pipe. In such situations, the followingmay be required:

a. Special local analyses.

b. New prototype tests.

c. Records of abnormal operation, i.e., occurrences wherethe pipe was submitted to conditions beyond those consideredby the original design (e.g., extreme loads or temperatures).

d. Records of defects or condition detected from inspectionduring operation or after retrieval (e.g., damage, corrosion,aging).

e. Records of former conditions of long-term pipe storage.

f. Tests for material qualiÞcation (e.g., aging tests, compati-bility tests, SSC/HIC NACE qualiÞcation tests).

12.3.3.4.2 Special local analyses may be useful for evalua-tion of damage, such as wire rupture, corrosion, wear, etc.New prototype tests may be performed to conÞrm some spe-ciÞc characteristic required for reuse of the pipe in new con-ditions (e.g., if new installation equipment applies high stressto the pipe).

12.3.3.4.3 Results of qualiÞcation tests on materials (referto Section 6.2 of API SpeciÞcation 17J/17K) may be usefulfor evaluation of their remaining life when exposed to opera-tional ßuid or to environmental conditions. New tests may benecessary if data is not available. For test procedures and cri-teria, refer to Tables 11 and 12 in API SpeciÞcation 17J/17K.

12.3.3.4.4 To carry out the global and local analysis,qualiÞed methods for the pipe and system design should beavailable. Operators can use their own methods or those of amanufacturer or a third party to carry out the pipe assess-ment. In all cases, the programs and methods used should bevalidated as required by Section 5.2.1 of API SpeciÞcation17J/17K.

12.3.3.4.5 Special attention should be given to calculatingthe pipe remaining life. Safety margins should be the same asspeciÞed in API SpeciÞcation 17J/17K. Information concern-ing materialsÕ long-term performance under the original useconditions is essential for taking any decision about pipereuse. Sources of data that can be useful for this purposeinclude operational experience with materials and pipes,results of long-term tests performed for material qualiÞcation,prototype testing (e.g., destructive testing of sample fromretrieved pipe), inspection of retrieved pipes, suitably quali-Þed NDE monitoring techniques, and calibrated models forcalculating service life, both theoretically and with tests.

12.3.4 Inspection and Repair

ManufacturerÕs technical personnel should be involved inany inspection and/or repair operation.

12.3.4.1 Unbonded Pipe

12.3.4.1.1 If the pipe outer sheath is damaged (caused, forinstance, during the pipe retrieval), rapid corrosion ofexposed pipe armors can occur when it is subjected to theatmosphere. It is therefore recommended that such areas beimmediately protected by using special anti-corrosion prod-ucts and by covering with tape or bandage if they cannot beimmediately repaired.

12.3.4.1.2 If there is damage to the outer sheath thatallows the ingress of water, then an inspection should assessthe degree of corrosion that has taken place and evaluate thecorrosion that may be present in areas with an intact outersheath. Corrosion may both reduce the armor load capacityand adversely affect its wear characteristics. Areas of the pipewhere burst disks occurred during the pipeÕs previous opera-tion are an example of a pipe section where signiÞcant corro-sive damage can occur. Acceptance tests (see 12.3.5) andlocal analysis should be performed to evaluate if the damageis critical.

12.3.4.1.3 If damage in a localized area turns out to becritical, it may be convenient to cut it out and install end Þt-tings on the extremities of remaining sections to make theirreuse feasible. Special attention should be given to the inter-face between the pipe and the bend stiffener/restrictor, wheredamage and corrosion are likely to appear.

12.3.4.1.4 For outer sheath repair, qualiÞed proceduresand personnel should be used. The procedures should guaran-tee the minimum required pipe performance properties. ThequaliÞcation of repair procedures should include tests whichconÞrm pipe characteristics. The long-term degradation ofthe repaired area should also be considered. As an alternativeto outer sheath repair, it may be more convenient to strip offthe whole layer and re-extrude a new outer sheath.

12.3.4.1.5 End Þttings should be subjected to detailedinspection. The corrosion protection system should be evalu-ated for all components (end Þtting body, bolts, nuts). Thegasket seat should be checked against the design standard forthe required surface Þnish. If the face does not meet therequirements, it should be decided whether regrooving bymachining will be feasible or whether the ßange should bereplaced. Replacing the ßange may require replacement ofthe end Þtting, as it may not be possible to weld on a newßange. Relief valves should be tested and recalibrated orreplaced.

12.3.4.1.6 The long-term degradation of plastic compo-nents of end Þttings should be evaluated. Service life of resinsand gaskets should be obtained from the pipe supplier.

12.3.4.1.7 If for some reason the end Þttings are removed,the new end Þttings should be assembled using a procedureapproved by the pipe supplier or other competent body.


12.3.4.2 Bonded Pipe

12.3.4.2.1 The exterior surface of the pipe should be thor-oughly cleaned and inspected during or after retrieval.

12.3.4.2.2 If the cover layer of a bonded ßexible pipe isdamaged (caused, for instance, during pipe retrieval) rapidcorrosion of exposed reinforcing plies can occur. It is there-fore recommended that the area be immediately protected byapplying anti-corrosion product(s) and covering with a tem-porary, impermeable layer.

12.3.4.2.3 All areas of cover damage should be inspectedfor corrosion. Corrosion can produce rapid degradation in theÞlament-wire cables typically used in bonded ßexible pipe.Acceptance tests and local analysis should be performed todetermine if the corrosion damage is critical.

12.3.4.2.4 If the damage in a localized area is determinedto be critical, it may be possible to cut out the damage andinstall new end Þttings at the cut ends of the remaining sec-tions. New end Þttings should be installed by qualiÞed per-sonnel using qualiÞed procedures. Bonded ßexible pipes withbuilt-in end Þttings may use temporary repair Þttings but, ingeneral, this type pipe cannot be permanently re-terminated.

12.3.4.2.5 QualiÞed procedures and personnel should beused for all cover repairs. The repair procedure qualiÞca-tion should include tests that conÞrm pipe characteristics.Long-term degradation of the repaired area should also beconsidered.

12.3.4.2.6 End Þttings should be subjected to detailedinspection. The corrosion protection system should be evalu-ated for all components (end Þtting body, bolts, nuts). Thegasket seat should be checked against the design standard forthe required surface Þnish. If the face does not meet therequirements, it should be decided whether regrooving bymachining will be feasible or whether the ßange should bereplaced. Replacing the ßange may require replacement of theend Þtting, as it may not be possible to weld on a new ßange.

12.3.4.2.7 The long-term degradation of plastic compo-nents of end Þttings should be evaluated. Service life of resinsand gaskets should be obtained from the pipe supplier.

12.3.4.2.8 For bonded pipes with built-in end Þttings, theinterface between the built-in nipple and the liner layershould be visually inspected using a mirror and/or borescope.Any evidence of delamination of the liner layer, linear move-ment (i.e., slippage) between the nipple and liner, or seepageof oil into the nipple-liner interface should be thoroughlyevaluated to determine if it is critical.

12.3.5 Test Requirements

12.3.5.1 After a pipe is prepared for reuse, it should be sub-jected to the factory tests speciÞed in API SpeciÞcation 17J/17K or as required by the user (e.g., hydrostatic test, gauge

test, electric continuity test). The hydrostatic test pressureshould be in accordance with FAT requirements in API Speci-Þcation 17J. If the test pressure is reduced, then the designpressure should also be reduced to 0.67 times the test pressure.

12.3.5.2 After the pressure test, pipe ßushing and corro-sion protection for storage may be necessary. Other tests orinspection methods (refer to Section 12) may be used tocheck for defects in the pipe, such as material loss by corro-sion or cracks/ßaws in the structural layers. If abnormalitiesare identiÞed, the pipe should be subjected to further analysis,as recommended in 12.3.3.4.

12.3.5.3 Re-installation and commissioning of the pipeshould be in accordance with the recommendations of 11.4and 11.5.

13 Integrity and Condition Monitoring

13.1 SCOPE

This section provides guidelines and recommendations onintegrity and condition monitoring, including potential pipedefects, for unbonded ßexible pipes. In general this sectiondoes not apply to bonded ßexible pipes.

13.2 GENERAL PHILOSOPHY

13.2.1 Inspection/Monitoring Philosophy

13.2.1.1 A detailed integrity and condition monitoring pro-gram should be established, based on an evaluation of thefailure modes to which ßexible pipe are exposed and the riskattributed to failure from each source [51].

13.2.1.2 It may be required to design a monitoring systemto operate throughout the Þeld design life, or for a reducedperiod, on one or more dynamic risers or ßowlines forresearch or operational use. These issues should be fullyresolved and a Þeld philosophy completed prior to designcommencement. The monitoring and inspection philosophyshould be identiÞed in the project design premise.

13.2.2 Scope

The inspection/monitoring program should typicallyinclude all applications of ßexible pipe and their ancillarycomponents.

13.2.3 Objectives

The objectives of an in-service integrity and conditionmonitoring program should include the following:

a. Detection of possible degradation at a sufÞciently earlystage to allow for remedial action and thereby:

1. Protect against accidents or loss of life.2. Protect against environmental pollution.


3. Avoid downtime.4. Minimize the risk of economic loss arising from pipesystem degradation or damage to Þeld equipment.

b. Demonstration of continued Þtness for purpose.

c. Compliance with all relevant statutory and regulatoryrequirements.

d. Provision of a record of service data which may berequired when considering future re-use.

13.2.4 Establishment of Inspection/Monitoring Program

13.2.4.1 Potential modes of failure should be identiÞed forthe speciÞc design and application of the ßexible pipe. Thepipe systemÕs functional and operational requirements shouldbe taken into account when assessing potential failure modes.

13.2.4.2 A risk analysis should seek to quantify the riskattributed to each failure mode, typically as a function of theprobability and consequence of failure. The establishment ofinspection/monitoring strategy should relate the degree ofrequired monitoring or inspection to the calculated level ofrisk.

13.2.4.3 Available direct or indirect methods to inspect/access the pipe should be evaluated for their suitability for theintended ßowline or riser application. Furthermore, adequateprovision for facilitating pipe monitoring should be made inthe design of the pipe system and associated topside and sub-sea facilities. In this respect, topside piping should bedesigned to allow access for internal inspection tools. Notethat this area of ßexible pipe technology is continually evolv-ing, and the pipe and pipe system design should consider thelikelihood that some developing methods will become stan-dard practice in the future.

13.2.4.4 The requirements for a baseline survey should beconsidered for each of the methods that are selected as part ofthe integrity and condition monitoring program. Provisionshould be made for any such baseline survey before the pipeis brought into service, and records should be held for the fulllife of the ßexible pipe system.

13.2.4.5 It is important that integrity monitoring begins atthe factory with thorough inspection, quality control, anddocumentation of the manufacturing process. Installationoperations need thorough planning to avoid damage causedby handling equipment. Special care should be taken with theÞrst baseline visual inspection after installation to documentminor anomalies or damage which may indicate undetectedproblems and the need for more frequent monitoring.

13.2.5 Inspection/Monitoring Program Review

The inspection/monitoring program should be subjected toregular documented review throughout the service life of the

ßexible pipe Þeld system. This review should reconsider themethods and frequency of review based on the results ofinspection or monitoring, experience of this or similar sys-tems or additional knowledge of ßexible pipe behavior. Doc-umented records of the review process should be retained forthe service life of the Þeld system, or the service life of eachßexible pipe in the Þeld system if any pipes are reused.

13.3 FAILURE MODES AND POTENTIAL PIPE DEFECTS

13.3.1 A ßexible pipe failure mode describes one possibleprocess by which a ßexible pipe could fail. A single failuremode typically represents a succession of pipe defects whichhave the potential to culminate in pipe failure. The identiÞca-tion of relevant failure modes should be based on a detailedknowledge of ßexible pipe behavior.

13.3.2 Tables 29 to 31 identify potential defects that applyto the integrity of ßexible pipe systems. Each defect is num-bered, and the likely cause and consequence of the defect hasbeen identiÞed.

13.3.3 Tables 29 and 30 relate to riser and ßowline applica-tions, respectively, individually classifying defects in eachlayer of pipe. Table 31 applies to defects associated with sys-tem components and pipe attachmentsÑdamage that mayaffect the condition or integrity of the ßexible pipe itself.

13.3.4 These tables should be reviewed during the selec-tion of the integrity and condition monitoring program. Thereview will allow identiÞcation of critical components in thepipe system and potentially critical defects, thereby facilitat-ing a better deÞnition of the requirement and relevancy ofavailable monitoring methods.

13.4 MONITORING METHODS

13.4.1 Current methods available for the monitoring ofßexible pipes in service are presented in Table 32. Visualinspection and periodic pressure testing have been, to date,the most common forms of in-service monitoring used for thedemonstration of continued Þtness for purpose.

13.4.2 Non-destructive testing of pipes in service includesdirect intrusive and non-intrusive techniques which have beenÞeld demonstrated and suitably qualiÞed as measurementmethods.

13.4.3 The ageing of non-metallic components and thecorrosion or erosion of metallic components can be moni-tored by installation in the ßow path of short test pipes or cou-pons placed in coupon sampling traps. The test material canbe retrieved and destructively or non-destructively tested atpre-deÞned intervals throughout the service life of the com-ponent. Figure 39 shows a removable rigid test pipe arrange-ment (in series or in parallel with the ßow), which uses a


mock-up of the internal layers of ßexible pipe; it allows gasventing through a pressure relief valve.

13.4.4 Di-electric sensing of the internal pressure sheathshould be used only if qualiÞed for the material and for thetemperature and pressure ranges applicable to the serviceconditions. A schematic representation of the measurementmethod applied to topside internal pressure sheath monitoringis shown in Figure 40.

13.4.5 Gas diffusion monitoring of a ßexible riser annulusmeasures the composition of gas sampled via a vent valve atthe pipe end Þtting, typically at the riser top. The objective isto relate the results to the potential for metallic layer corro-sion (including SSC and HIC) or the aged condition of theinternal pressure sheath, which may provide early warning ofsevere deterioration before the integrity of the pipe isaffected.

13.4.6 Load, deformation and environmental monitoringincludes methods that involve the measurement of the follow-ing:

a. Pipe tension.

b. Deßection.

c. Torsion.

d. Bending.

e. Internal product composition.

f. Internal pressure and temperature.

g. Vessel motions and environmental conditions.

13.5 RECOMMENDATIONS

13.5.1 Scope of Recommendations

Although the methods and frequency of required monitor-ing or inspection should be determined based on the results ofa documented risk analysis, some general comments are pro-vided on available inspection and monitoring measures.

13.5.2 General Recommendations

13.5.2.1 Subsea visual inspection can be performed usingdivers or remotely operated vehicles (ROVs). Subsea and top-side visual inspection should be used periodically to provideevidence of observable damage to ßexibles from accidents,degradation during service, or damage during installation.Where possible, risers and ßowlines should be inspected fol-lowing potentially damaging incidents. After repair work, thepipe system should be re-inspected to conÞrm that any pipeor component repairs or replacements have been properlyperformed. Visual inspection should also occur after re-con-nection following an emergency or routine pipe disconnec-

tion. Visual inspection should seek to identify the followingpotential problems:

a. Extent and type of marine growth.

b. Pipe general integrity and condition, including leaks.

c. Pipe outer sheath or external carcass integrity andcondition.

d. Noticeable debris.

e. Evidence of scour and estimated length of free spans.

f. Condition of end Þttings.

g. Condition of cathodic protection system.

h. Any identiÞable damage, distortion, or degradation.

i. Any identiÞable disarrangement of pipe and disarrange-ment or loss of pipe ancillary components.

j. Interference with other subsea hardware.

k. Loops and kinks.

13.5.2.2 Defects should be documented in terms of type,size, location (pipe identiÞcation and Northing and Eastingcoordinates), depth, and time of observation. The inßuence ofdefects on structural or pressure integrity should be assessed.An acoustic survey may also be performed to identify thelocation of buried pipes and depth of cover.

13.5.2.3 The outer surface of the pipe should be examinedfor cuts, gouges, abrasion, bulges, soft spots, loose outersheath, or any sign of separation of the ßexible pipe from endÞttings. Any tendency for a suspended line to form a loopshould also be noted, since these could form kinks under ten-sion. Exposed surfaces of end Þttings should be examined forcracks or excessive corrosion. Slow permeation of a chemicalor product through the internal pressure sheath may Þrstbecome evident when product migrates along the line and isdischarged through vent valves at the end Þtting.

13.5.2.4 Where direct measurement of a section of pipe ispossibleÑeither through external or internal non-destructivetest methodsÑthe pipe locations selected for measurementshould be chosen to reßect the severity of service conditionsin terms of loading, deformation, and internal or externalenvironmental conditions. For ßexible risers, areas of criticaldesign loading may include one or more of the following:

a. Top end connection for tension, bending, and deßectionmonitoring.

b. Pre- and post-mid water support buoy for bending and tor-sion measurement.

c. Riser base connection for temperature, bending, and pres-sure monitoring.

13.5.2.5 If practicable, test pipes should be exposed to thesame pressures, stresses, and diffusion transport conditions asthose existing in the ßexible pipe. For polymer monitoring ofproduction ßuid, the test pipe or sample trap should be


located at the end of the ßexible pipe nearest the wellhead orelse the temperature at the test pipe should be controlled to atemperature at least as high as that in the ßexible pipe sectionclosest to the wellhead. For erosion monitoring, the bendradius at the test pipe should be designed to be less than theminimum bend radius of the ßexible in service.

13.5.2.6 Annulus monitoring methods should be demon-strated to be practical, and quantitative criteria should bedeveloped for their implementation. If the requirement formonitoring of dynamic risers or ßowlines has been speciÞed,suitable methods of measurement should be proposed for therelevant areas.

13.5.3 Inspection Intervals

13.5.3.1 Inspection intervals should be devised from a con-sideration of pipe failure modes. The following factors shouldbe considered when determining inspection intervals:

a. Consequences of failure to human life, property, or theenvironment.

b. Operational criticality.

c. Degree of innovation or lack of service experience undersimilar conditions.

d. Pipe product and service conditions, e.g., sour service,high pressure.

13.5.3.2 Present condition, and inspection and service his-tory of the pipe.

13.5.3.3 The external visual inspection interval for a ßexi-ble pipe should be deÞned in the inspection plan and shouldbe carried out immediately after suspected damage, re-con-nection, or installation, and prior to any trench backÞlling orrock dumping.


Figure 39—Schematic of Possible Test Pipe Arrangements

Figure 40—Schematic of Topside Di-Electric Sensing Layout and Instrumentation for Thermoplastic Monitoring

Flow

direction

Flow

direction

Testpipe

Testpipe

(a) In Parallel (b) In Series

IBMcompatible

PC

LCR meter

AC supply

MultiplexerDistribution

boardJunction

box

Weather deckHazardous area

Sensor spoolpiece in product

flow

Physical arrangement of remote sensors and measuring system


Table 29—Potential Pipe Defects for Static Applications

Pipe LayerDefect Ref. Defect Consequence Possible Cause

CARCASS 1.1 Hole, crevice, pitting, or thinning

Reduced collapse resistance and reduced tension capacity

a. Sand erosionb. Crevice, pitting or uniform corrosion (and SSC/HIC)c. Excessively sour serviced. Pigging damage

1.2 Unlocking deformation Locally reduced collapse resis-tance and tension capacity.

a. Overbendingb. Excess tension with bendingc. Pigging damage

1.3 Collapse or ovalization Blocked or reduced bore a. Excess tension b. External over-pressure (possible hole in outer

sheath)c. High initial ovality (manufacturing defect)d. Excess loading or deformation during installatione. High radial gap between pressure armor and inter-

nal Pressure sheath (manufacturing defect)f. Side impact or point contact

INTERNALPRESSURE SHEATH

2.1 Crack or hole Leak of medium into annulus and/or rupture of outer sheath and/or pipe rupture/leakage

a. Hole, bubble or inclusion during fabricationb. Pressure armor rupturec. Pressure armor unlockingd. Ageing (embrittlement)e. Temperature above design levelsf. Carcass defectg. Pressure above design levelsh. Pigging damagei. ENVIRONMENT Assisted Cracking (EAC)j. Erosion (smooth bore pipes)k. Product composition outside design limits

2.2 Rupture Failure of pipe a. Pipe bending (tension side)b. Collapse (outer sheath leak, low internal pressure,

collapsed carcass) c. Ageing/embrittlementd. Failure of pressure armor

2.3 Collapse Recoverable, but plastic straining

a. Excessive reduction in product pressure or exces-sive external relative to internal pressure (no car-cass or collapsing carcass)

2.4 Ageing embrittlement Reduced elasticity and greater susceptibility to cracking

a. Material property changes (degradation) arising from exposure to ßuid

2.5 Excess creep (extrusion) of polymer into metallic layer

Possible hole or crack rupture a. Operation at pressures and/or temperatures outside limits

b. Inadequate material selectionc. Inadequate wall thickness

2.6 Blistering Possible hole or crack rupture a. Rapid decomposition due to operation at pressures and/or temperatures outside limits.

b. Rapid decomposition under inadequate material selection


PRESSURE ARMOR LAYER

3.1 Individual or multiple wire rupture

Reduced structural capacity or pipe rupture (burst) or extru-sion/leakage of internal pres-sure sheath

a. Corrosion b. SulÞde Stress Cracking (SSC)c. Hydrogen Induced Cracking, (HIC)d. Excess internal pressuree. Failure of tensile/backup pressure armor (excess

tension/pressure)f. Unlockingg. Manufacturing (welding) defect

3.2 Unlocking Reduced structural capacity or pipe rupture (burst) or extru-sion/leakage of internal pres-sure sheath

a. Overbendingb. Excess tensionc. Impactd. Failure of tensile or backup pressure armor e. Radial compression at installationf. Excess torsion during installation

3.3 Collapse or ovalization Reduced bore a. Side impactb. Point contactc. Excess tension (in service)d. Radial compression at installation

3.4 Corrosion Pressure armor tensile failure a. Sour service/corrosive annulusb. Ingress of seawater into annulus

BACKUP PRESSURE ARMOR LAYER

4.1 Rupture (single or all wires)

Reduced structural capacity or pipe rupture (burst)

a. Corrosionb. SulÞde Stress Cracking (SSC)c. Hydrogen Induced Cracking (HIC)d. Excess internal pressuree. Failure of tensile/pressure armorsf. Manufacturing (welding) defect

4.2 Ovality Reduced bore a. Side impactb. Point contactc. Excess tension

4.3 Clustering Uneven support of pressure armor layer, failure

a. Manufacturing defect

4.4 Corrosion Pressure armor tensile failure a. Sour service/corrosive annulusb. Ingress of seawater into annulus

TENSILE ARMOUR LAYERS

5.1 Multiple wire rupture Reduced structural capacity or pipe rupture (burst)

a. Corrosionb. SulÞde Stress Cracking (SSC)c. Hydrogen Induced Cracking (HIC)d. Excess tension or internal pressuree. Manufacturing (welding) defectf. Accidental impact

5.2 Birdcaging or clustering Reduced tension capacity a. Overtwistb. Compression

Table 29—Potential Pipe Defects for Static Applications (Continued)



TENSILE ARMOR LAYERS(Continued)

5.3 Kinking Reduced tension capacity a. Side impactb. Point contactc. Loop in line due to design, manufacturing defect

or installation error

5.4 Corrosion Tensile armor rupture a. Sour service/corrosive annulusb. Ingress of seawater into annulus

5.5 Individual wire rupture Reduced tension capacity a. Corrosionb. SulÞde Stress Cracking (SSC)c. Hydrogen Induced Cracking (HIC)d. Overstressed armors (excess tension or internal

pressure)e. Improper clamp design or Þtf. Manufacturing (welding) defectg. Accidental impact

INSULATION LAYER

7.1 Crushed layer Inadequate insulation a. Crushing during installationb. External overpressure

7.2 Flooded layer Inadequate insulation a. Hole in outer sheath or another leakproof layer between outer sheath and insulation layer

7.3 Pipe clogging Wax deposit a. Inappropriate design

OUTER SHEATH

8.1 Hole, tear, rupture or crack

Ingress of seawater (if through wall)

a. Manufacturing defectb. Tear during installation c. Point contact, impact or shearingd. Improper clamp design or Þte. Pressure buildup in annulusf. Blocked vent valveg. Internal pressure sheath leak/holeh. Overbending + existing defecti. Ageing, weathering (UV radiation)

8.2 Ingress of seawater Tensile or pressure armor wire corrosion (especially splash zone) or collapse (smooth bore) or ßooded insulation layer

a. Hole, tear, rupture, crack in outer sheath

END FITTING 9.1 Internal pressure sheath pull-out

Leak of medium into annulus, failure

a. Loss of friction (carcass deformation etc.)b. Tearc. Sheath shrinkage due to temperature cycling d. Creep

9.2 Tensile armor pull-out (all wires)

Failure, burst a. Wire break within end Þttingb. Epoxy failure (sour service)c. Epoxy failure (high temperature ageing)d. Loss of frictione. Excess tension




END FITTING(Continued)

9.3 Outer sheath pull-out Ingress of seawater (hydrostatic pressure)

a. Excess annulus pressureb. Creep

9.4 Vent valve blockage Outer sheath burst (if it occurs to all vent valves)

a. Debrisb. Marine growthc. Mechanism failure (corrosion etc.)d. Fabrication errors

9.5 Vent valve leakage Possible seawater ingress into annulus

a. Corrosionb. Failure of mechanism (seal failure etc.)

9.6 Individual tensile armor pull-out

Reduced structural capacity a. Wire break within end Þttingb. Epoxy failure (sour service)c. Epoxy failure (high temperature ageing)d. Loss of frictione. Excess tension

9.7 Failure of sealing system (sealing rings etc.)

Leak of medium into annulus, possible vent valve blockage, Possible outer sheath burst and pipe leakage (failure)

a. Fabrication errors - ineffective seal of internal pressure sheath.

b. Inadequate designc. Excess internal pressured. Excess tension or torsione. Inadequate installationf. Excessively low production temperature

9.8 Crack or rupture of pres-sure armor or backup pressure armor

Possible pipe burst or reduced pressure capacity

a. Corrosionb. SulÞde Stress Cracking (SSC)c. Hydrogen Induced Cracking (HIC)d. Excess internal pressuree. Failure of tensile armor layer (excess tension or

internal pressure)

9.9 Crack or rupture of ten-sile armor

Possible progressive pullout and pipe failure or reduced structural capacity

a. Corrosionb. SulÞde Stress Cracking (SSC)c. Hydrogen Induced Cracking (HIC)d. Excess internal pressuree. Failure of tensile armor layer (excess tension or

internal pressure)

9.10 Structural failure of end Þtting body or ßange

Pipe burst/catastrophic failure a. Excess internal pressureb. Inadequate designc. Excess tension or torsion loadsd. Hydrostatic collapsee. Corrosion/chemical degradationf. Brittle fractureg. Fatigue




Table 30—Potential Pipe Defects for Dynamic Applications


CARCASS 1.1Ð1.3 As Table 29 for static applications.

As Table 29 for static applica-tions.

As Table 29 for static applications

1.4 Circumferential cracking/wear

Reduced collapse resistance and reduced tension capacity or pres-sure sheath rupture

a. Fatigue + crevice, pitting or uniform corrosion

b. Carcass to carcass wear/friction

INTERNAL PRESSURE SHEATH

2.1Ð2.6 As Table 29 for static applications.


As Table 29 for static applications.

2.7 Rupture Failure of pipe a. Fatigue cracking

2.8 Wear/nibbling No adverse consequence or inter-nal pressure sheath crack or hole

a. Abrasion between internal pressure sheath and carcass

b. Abrasion between internal pressure sheath and pressure armor

PRESSURE ARMOR LAYER





Reduced structural capacity or pipe rupture (burst) or extrusion/leakage of internal pressure sheath

a. Wear at inter-wire contactb. Wear from contact with back-up pressure

layer.c. Cracking along wired. Fatigue failuree. Welding defect

3.6 Longitudinal wire crack Potential elongation to critical defect size

a. Inter-wire contact and local stress concentration

BACKUP PRESSURE ARMOR LAYER





Reduced structural capacity or pipe rupture (burst)

a. Wear from contact with pressure armor layer.b. Fatigue failure

TENSILE ARMOR LAYERS





TENSILE ARMORLAYERS(Continued)

5.6 Multiple wire rupture Reduced structural capacity or pipe rupture (burst)

a. Wear between armor layers (gap in anti-wear layer, loss of lubricating oil)

b. Fretting fatiguec. Notch or crack fatigue failured. Fatigue failure

5.7 Individual wire rupture Reduced structural capacity or pipe rupture (burst)

a. Wear between armor layers (gap in anti-wear layer, loss of lubricating oil)

b. Fretting fatiguec. Notch or crack fatigue failured. Fatigue failure

ANTI-WEAR LAYER

6.1 Wear, cracking Radial contact of armor layers, wear

a. Relative movement between layers b. Temperaturec. Manufacturing defect

6.2 Clustering Radial contact of armor layers, wear

a. Manufacturing defect

INSULATION LAYER


As Table 29 for static applications As Table 29 for static applications

OUTER SHEATH

8.1 As Table 29 for static applications.


8.2 Wear, tear Possible rupture due to annulus pressure or possible hole due to wear or accelerated corrosion of metallic armor layers

a. Abrasive contact with seabed, other lines or other surfaces.

END FITTING 9.1Ð9.10 As Table 29 for static applications


Table 30—Potential Pipe Defects for Dynamic Applications (Continued)



Table 31—Potential System Defects for Static and Dynamic Applications

Pipe Layer Defect Ref. Defect Consequence Possible Cause

BEND LIMITERS

(Stiffeners and Bellmouths)

10.1 Stiffener crack Possible pipe overbending a. Stiffener fatigueb. Excessive bending at stiffenerc. Material degradation

10.2 Stiffener rupture Possible pipe overbending or possible tear of outer sheath

a. Stiffener fatigueb. Excessive bending at stiffenerc. Abrasion or impact damaged. Material degradation

10.3 Stiffener support structure failure

Possible pipe overbending or possible tear of outer sheath

a. Excessive bending at stiffener and overloading of bindings or support

b. Impact damagec. Structural fatigue of bindings or support structure

10.4 Bellmouth deformation or inadequate size

Pipe overbending a. Bellmouth design or manufacturing faultb. Excessive pipe bending around bellmouthc. Impact damage to bellmouth.d. Òpig tailingÓ of pipe

10.5 Stiffener misperformance Pipe overbending a. Inadequate design/design uncertainty (stiffness vs. temp)

b. Inadequate manufacture (PU curing)

BEND RESTRICTORS

11.1 Unlocking disarrangement Possible pipe overbending a.Excessive bending in pipeb. Defective or damaged restrictor

11.2 Position disarrangement Possible pipe overbending a. Inadequate clamping of bend restrictor(s)b. Impact or abrasion

11.3 Loss of bend Restrictor(s) Possible pipe overbending a. Inadequate or damaged clamp (s)b. Abrasion or impact damage

BUOYANCY MODULES

12.1 Position disarrangement Possible pipe overbending or excess tension or tear of outer sheath.

a. Defective buoyancy modulesb. Abrasion or impact damagec. Inadequate or damaged clamp(s)

12.2 Loss or failure of buoyancy module(s)

Possible pipe overbending or excess tension or tear of outer sheath.

a. inadequate or damaged clamp (s)b. Abrasion or impact damage

12.3 Reduced buoyancy Possible pipe overbending or excess tension or seabed contact (abrasion, compres-sion, overbending or impact) in sag.

a. Defective buoyancy modulesb. Abrasion or impact damagec. Inadequate or damaged clamp (s)d. Hydrostatic compression, water absorption or

creep


SUBSEA BUOYS

13.1 Position disarrangement Possible pipe overbending or excess tension, failure of pressure or tensile armors.

a. Defective buoy b. Abrasion, dropped object, collision or trawl-

board impact damagec. Inadequate or damaged clamp(s)

13.2 Loss of buoy Likely pipe overbending or excess tension, failure of pressure or tensile armors

a. Underdesign of bindings/anchorsb. Dropped object, collision or trawl-board damage

to tethers or buoy c. Fatigue of tethers/bindingsd. Flooding of buoye. Degradation of buoy material

13.3 Reduced buoyancy Possible pipe overbending or excess tension, failure of pressure or tensile armors

a. Defective buoyb. Abrasion or impact damagec. Inadequate or damaged clampsd. Flooding of buoye. Degradation of buoy materialf. Hydrostatic compression, water absorption or

creep

CLAMPS 14.1 Rupture Loss of buoyancy module or bend restrictor

a. Defective clampb. Abrasion or impact damage

14.2 Damage Reduced clamping capacity a. Abrasion or impact damage

14.3 Degradation Possible rupture a. Ageing or creep of plastic or corrosion of metallic clamp

RISER BASES 15.1 Damage to riser connection Possible end-Þtting damage or leakage at connection

a. Dropped object, anchor drag or trawl-board impact damage

15.2 Displacement Possible pipe overbending or possible excess tension

a. Dropped object, anchor drag or trawl-board impact damage

RISER SUPPORTSTRUCTURES

16.1 Disarrangement of risers Possible pipe overbending or possible excess tension or tear of outer sheath

a. Inadequate or damaged clamp(s)

16.2 Structural failure or dis-placement of support structure itself

Possible pipe overbending and possible excess tension or tear of outer sheath

a. Loads in excess of design valuesb. Inadequate design or manufacturec. Dropped object, collision or trawl-board impact

damage

CATHODIC PROTECTION

17.1 Disarrangement Inoperative cathodic protec-tion with risk of excessive corrosion

a. Dropped object, collision or trawl-board impact damage

b. Clamping failure

Table 31—Potential System Defects for Static and Dynamic Applications (Continued)



CATHODICPROTECTION(Continued)

17.2 Electrical discontinuity Inoperative cathodic protec-tion with risk of excessive corrosion

a. Inadequate manufacturing QAb. Dropped object, collision or trawl-board impact

damage

17.3 anode exhaustion Inoperative cathodic protec-tion with risk of excessive corrosion

a. Anode depletion in excess of design assumptions

MATTRESSES OR SAND BAGS

18.1 Disarrangement Free spans or possible over-bending or interference or abrasion

a. Excessive uplift due to riser motionb. Excessive uplift or horizontal motion due to acci-

dental loading

DUMPED ROCK OR TRENCH BACKFILL

19.1 loss of cover Possible pipe free spans and overbending, exposure to trawl-board or other impact damage

a. Gradual upward motion of pipe towards surface

FLEXIBLE PIPE LAYOUT

20.1 Upheaval buckling or upheaval creep of buried pipe

Possible overbending and local unlocking of pressure armor, exposure to trawl-board or other impact damage

a. Axial compression (temperature and/or pressure induced elongation)

b. Inadequate installation for buried pipe.

20.2 pipe loop Possible overbending pipe excess torsion

a. Excess torsion during installationb. Excess pipe length at installation

20.3 pipe disarrangement [com-pared to designed or as-built layout]

Possible overbending or pos-sible excess tension or possi-ble ovalization or possible tear of outer sheath

a. Anchor draggingb. FPS or FPSO excursion outside design limitsc. Trawl board or other side impactd. Point contact

For tear to outer sheath, refer to Table 29, defect Ref. 8.2 andTable 30, defect Refs. 8.2 and 8.5

20.4 pipe free spans Possible overbending a. Routing over sharp seabed featureb. Loss of cover of trenched or rock-dumped pipe

20.5 riser interference Possible damage to buoy-ancy devices, clamps or bend restrictors or possible over-bending or possible impact damage or wear / abrasion of pipe outer sheath

a. Extreme environmental conditions in excess of design values

b. Inadequate design to provide required clearancec. Loss of buoyancy modules or clamping devices

maintaining pipe separationd. Anchor dragginge. Excessive vessel excursion

Table 31—Potential System Defects for Static and Dynamic Applications (Continued)



Table 32—Current Integrity and Condition Monitoring Methods

Method No. Monitoring Method Description Purpose

1. Visual inspection(1) External

By ROV or manually to assess leakage or vis-ible deformation or damage to pipe or outer sheath.

To establish the overall integrity of visible sections of the pipe and the general arrangement of the pipe system.

(ii) Internal By camera device inserted into the pipe bore. To check the condition of the internal carcass or internal pressure sheath.

2. Pressure test [hydrotest] Pressure applied to pipe and decay measured as a function of time. Leakages or anomalies identiÞed.

To establish the ability of the pipe to withstand pressure loads, typically in excess of max. allowable operating pressure, at a given time.

3. Destructive analysis of removed samples

Generally applied to coupon testing for age-ing of internal pressure sheath whereby age-ing coupons of polymer are exposed to ßow environment in a spool piece sample trap and removed periodically for destructive testing.

To predict the state of ageing or degradation of the inter-nal pressure sheath by extrapolation from tensile or other testing of thermoplastic material samples removed from actual ßow conditions.

4. Load, deformation and environment monitoring

Measured parameters include wind, wave or current environment, vessel motions, product temperature, pressure and composition, and structural (or ßexible pipe) loads and defor-mations.

Used for design veriÞcation or remaining life assessment.

Actual loads and environmental conditions may be com-pared with those predicted during design, thereby estab-lishing the degree of conservatism in the design.

Service life calculations may also predict remaining life based on measured environment or loads.

5. Non-destructive testing of pipes in service

These may include radiography or eddy cur-rent measurement of steel layers.

To establish the condition of steel tensile armor and pres-sure armor layers in service.

6. Gauging operations Gauging pigs to determine pipe ovality. To check for damage to the internal pipe proÞle.

7. Spool Piece/Test Pipes:(i) Di-electric sensing or ultrasonic condition monitoring

Options:Applied to on-line ageing analysis of internal pressure sheath coupon inserted into a rigid test spool which is designed to emulate ßow conditions. The test pipe is likely to be in series with the ßow.

To predict the state of ageing or degradation of the inter-nal pressure sheath by extrapolation from online mea-surement of a material sample exposed to actual ßow conditions.

(ii) Test Pipe Use of a ßexible (or rigid with mock-up inter-nal) test pipe in series or in parallel with the ßow which is periodically removed for destructive or non-destructive testing.

To examine the state of ageing or degradation of the internal carcass, internal pressure sheath and/or pressure and tensile armor layers of the ßexible pipe.

8. Annulus Monitoring(i) Gas Diffusion moni-toring

Measurement of annulus ßuid (pH, chemical composition, volume)

To predict degradation of the steel pressure armor or ten-sile armor layers or the aged condition of the internal pressure sheath or susceptibility of annulus environment to such degradation.

133

APPENDIX A—FLEXIBLE PIPE HIGH TEMPERATURE END FITTINGQUALIFICATION TEST PROTOCOL: VOLATILE CONTENT POLYMERS

This protocol is a synthesis of the various requirements andobjectives of many ßexible pipe operators and manufacturers.It is primarily intended to qualify end Þttings genericallyrather than for speciÞc project requirements. Section 6 pro-vides discussion of topics that may be appropriate to testsconducted for speciÞc projects and for interpreting the resultsof tests conducted under this protocol for speciÞc projects.The protocol may also be used together with the Crude OilExposure Test Procedure (Appendix A1) to evaluate end Þt-ting performance when subjected to speciÞc crude oil envi-ronments. In addition to the mechanical behavior tested bythis protocol, appropriate testing is required to qualify thechemical and physical suitability of the end Þtting and pres-sure sheath materials. The protocol does not qualify thestrength or stiffness of the end Þttings. See Section 6 for otherqualiÞcation topics.

Pairs of identical samples will be tested to identical condi-tions. Four end Þttings are required to meet the acceptancecriteria to achieve unrestricted qualiÞcation for the envelopeof service covered by the test conditions.

The protocol may be used to qualify static or dynamic endÞttings. The protocol is applicable for plasticized polymerßuid barriers. Its development is based on the behavior ofPVDF (poly-vinylidene ßuoride) plasticized with DBS (dibu-tyl sebacate). The protocol, however, is not restricted to thismaterial combination.

A.1 Test Objective

A.1.1 The protocol deÞned below provides an industry-acceptable methodology to qualify the mechanical perfor-mance of both existing and newly developed end Þttingdesigns for ßexible pipes made with high temperature poly-mer internal pressure sheaths for a representative service lifeof 20 years.

A.1.2 The protocol is applicable for plasticized polymerßuid barriers.

A.1.3 The protocol is applicable for ßexible pipes in oilservice, gas service, and water injection service.

A.2 Initial Data

Prior to the start of testing, the manufacturer is to specify:

a. The rated service temperature THI, TLO, for which the endÞtting design is being qualiÞed.b. The Òinitial movementÓ because of Òbedding-inÓ or Òcom-pliance take-upÓ that is predicted to occur in the early stagesof the testing.

c. An objective weight percent, W, that is equal to or greaterthan the plasticizer loss expected under the seal grip ring dur-ing 20 years of production at the upper test temperature. Themanufacturer shall specify the following deplasticizing times:

T1 = The time at the upper test temperature THI required to reduce by one third of W, the average weight per-cent of plasticizer below the seal grip ring.

T2 = The incremental time at the upper test temperature, beyond T1, required to reduce the average weight percent of plasticizer below the seal grip ring by an

additional 1/3 W for a total of two thirds of W.

T3 = The incremental time at the upper test temperature, beyond T2, required to reduce the average weight percent or plasticizer below the seal grip ring by an

additional 1/3 W for a total of W.

d. The full-scale or mid-scale test simulates Þeld perfor-mance of the end Þtting design for project speciÞc crude oilapplications if the percent volume change correlating with thepercent weight change achieved under the seal ring in thefull-scale test is greater than the equilibrium percent volumechange expected under the seal ring in the service conditionsduring 20 years or during a shorter service life based on testsas outlined in Appendix A1.

Comment: The Þnal weight percent plasticizer present in thepolymer is dependent on the service use of the pipe. Oil ser-vice may result in Þnal plasticizer content between 3.5 and 6percent, while high temperature gas service may result incomplete removal of plasticizer.

A.3 Test Samples

A.3.1 Two test samples are required. The test samples shallbe complete production ßexible pipes with all layers and fea-tures. All end Þttings shall be of the same design and assem-bled to the same procedure and dimensional tolerancespeciÞcation. Pipe length shall be 10 meters or more. Thepipe annulus should be vented. The pipe should be manufac-tured according to normal procedures; in particular, thehydrostatic test shall be at ambient temperature and shall notexceed 1.5 times the rated design pressure.

A.3.2 Alternatively, two mid-scale test samples can beused. A mid-scale test sample shall include the following:

a. All layers of the full-sized ßexible pipe through the pres-sure armor. Thermal mass or insulation external to the


pressure armor should be applied to represent the remainingßexible pipe layers.

b. Mid-scale end Þttings that include identical design,dimensions, and tolerancing attributes for the followingfunctions:

¥ Anchoring of the internal carcass

¥ Anchoring and sealing of the internal pressure sheathand any sacriÞcial or tape layers adjacent to the internalpressure sheath

¥ Thermal mass and insulation which is representative ofa full end Þtting

The length of a mid-scale sample is typically 1 to 2 meters.It shall be demonstrated that the tension force applied to apressure sheath seal during thermal cycling of a mid-scalesample is equal to or greater than that in the full-scale sample.This is accomplished by Þxed restraints applied to both endÞttings. The restraints prevent the mid-scale sample fromchanging length during thermal cycling.

A.3.3 The manufacturer shall have available, for review byany interested parties, detailed records of the as-built mate-rial, dimensions, Þts, and clearances of all pieces of the endÞtting and pipe body that may affect the performance of theend Þtting during testing. The records shall include thedimensioned and toleranced manufacturing drawings for thepipe and end Þttings and all manufacturing and procurementprocedures and standards. In addition, the records shallinclude the calculations associated with the initial data (initialmovement, W, T1, T2, T3, etc.)

A.3.4 Four monitoring assemblies shall be placed insideeach test pipe (see Figure A-1). Each assembly may consist ofa square pressure barrier material sample with edge dimen-sions at least twice the width of the seal grip ring. The barriermaterial shall be compressed between a rigid plate that islarger than the material sample and a rigid bar that is at leastas wide as the seal/grip ring and longer than the material sam-ple width. The percent compression of the material sampleshall be equal (±5 percent) to the compression achieved underthe seal/grip ring.

A.3.5 Alternative monitoring assembly conÞgurations maybe accepted, by agreement. The purpose is to identify plasti-cizer content and hydrocarbon uptake in the seal ring area.This is based on the assumption that a validated analytical orempirical model exists for the relationship between plasti-cizer in the main body of the pipe and the plasticizer condi-tion at the seal grip ring. Development and validation of thismodel is a necessary part of pre-qualiÞcation testing. Valida-tion will include survey of the barrier condition in the sealarea from a dissected end Þtting after a documented deplasti-Þcation process.

A.4 Test ProceduresA.4.1 TEST SET-UP

A.4.1.1 The test pieces shall be set up initially for statictemperature cycling, and subsequently in a dynamic testbench or alternative test structure to allow ßexing of the upperend of the test riser. The static phases (block 1 through block4, see below) may be carried out with the sample on a work-shop ßoor. The dynamic test blocks shall be carried out withthe test sample(s) mounted in a testing apparatus suitable toßex the riser upper end sufÞciently to ensure any effects ofinter-layer friction are removed from the temperature cycling.Dynamic ßexing is not required in the mid-scale test samplesas the axial stiffness of the hoop strength layer alone is negli-gible. Therefore, the inter-layer friction between the internalpressure sheath and the hoop strength layer will not affect theload on the internal pressure sheath anchoring.

A.4.1.2 Thermocouples shall be installed on the inside andoutside of each end Þtting approximately in the plane of theseal grip ring. Additional thermocouples may be applied fordata taking at the manufacturerÕs discretion.

A.4.1.3 The test pipes shall be Þlled with a nonhazardousoil that facilitates deplasticizing of the inner polymersheath(s).

A.4.1.4 Load cells shall be installed between the Þxedrestraints and the test samples so that axial loads generatedduring thermal cycling may be measured.

A.4.2 TEST TEMPERATURES AND PRESSURES

A.4.2.1 An upper and lower test temperature shall be spec-iÞed by the manufacturer (Thi and Tlo).

A.4.2.2 It is intended that this protocol may be used forqualiÞcation without the application of design margins. Themaximum service temperature for which the pipe becomes

Figure A-1—Monitoring Assembly

Materialsample

2 x Aor more

A = seal/grip ring width

A


qualiÞed shall be the average of Thi achieved during the testprogram. The minimum service temperature for which thepipe becomes qualiÞed shall be the average of Tlo.

Comment: An industry target for Thi is 130¡C. Target for Tlois Ð25¡C but no higher than 0¡C. An acceptable value for Tloexcluding blowdown may be Ð5 to Ð8¡C.

A.4.2.3 The internal pressure shall vary with the tempera-ture such that no less than atmospheric pressure is inducedat ambient temperature, and a maximum pressure of approx-imately 20 bars is induced at the top ßange at maximum testtemperature. Relief valves shall be provided so that theinternal pressure does not fall below ambient at any time (novacuum).

A.4.2.4 Cooling rates should be no slower than those pre-dicted for typical Þeld applications. Cooling shall be con-trolled so as to simulate these typical operating conditions.Heating at a slower rate than predicted for typical Þeld appli-cations is acceptable but will increase the time required tocomplete the temperature cycling process.

Comment: An industry basis for cooling rate has been agreedas a riser termination at the deck level of an FPSO turret or asemi-submersible, in air. See Section 6 for discussion ofÒHang-offÓ and ÒInsulationÓ effects.

A.4.3 THERMAL CYCLING PROCEDURE

A.4.3.1 Full Scale Test

Each thermal cycle shall consist of Þve steps:

¥ Step 1 The pipe internal temperature THI shall beraised to the test temperature.

¥ Step 2 After internal and external thermocouples onthe pipe reach a stable temperature, the testtemperature shall be maintained for an addi-tional 24 hours.

Comment: The soaking period is related to the creep andrelaxation behavior of the polymer that is considered. The 24-hour period is valid for PVDF; other polymers may requiredifferent values.

¥ Step 3 The test pipe shall be cooled until the internaland external thermocouples stabilize at ambi-ent temperature. Dynamic pipes shall be ßexedat least two times while at this step. Coolingshall be at a rate equivalent to natural convec-tion, with representative temperature gradientwithin the end Þttings.

¥ Step 4 The temperature shall be reduced to the lowertemperature by controlled cooling until theinternal and external thermocouples stabilize.

¥ Step 5 The temperature shall be maintained at thelower temperature for a minimum of one hour.

A.4.3.2 Mid Scale Test

Each thermal cycle shall consist of 5 steps:

¥ Step 1 At installation, the restraints shall be adjustedso that the axial force is within 500 N of zerowhile the sample is at ambient temperature.


¥ Step 3 The pipe internal temperature shall be main-tained at THI until 24 hours from the start ofthe heating cycle.

¥ Step 4 The pipe internal temperature shall be reducedto the lower temperature by controlled cooling,until the pipe internal temperature reachesTLO.

¥ Step 5 The pipe internal temperature shall be main-tained at TLO until 24 hours from the start ofthe cooling cycle. The cycle is repeated atStep 2.

A.4.4 TEST BLOCKS

A.4.4.1 Descriptions

A.4.4.1.1 Block 1

Block 1 consists of 10 cycles of static thermal cycling. Thebore of each end Þtting shall be inspected after 5 (±1) and 10(±1) cycles.

During Block 1 thermal cycling, the pipes should be essen-tially horizontal, and Þttings may be raised for convenience inÞlling, inspecting, etc., with the pipes free to expand and dis-tort as a result of heating and induced loads.

A.4.4.1.2 Block 2

Block 2 consists of deplasticizing at the test temperaturefor no less a period of time than T1. At the end of the block apressure test shall be conducted, one of the monitoringassemblies shall be removed from the test pipe, and thedegree of deplasticizing in the center of the material sampleshall be compared with the manufacturerÕs predictions. If thepredicted extent of deplasticizing has not been achieved, thedeplasticizing times for all blocks shall be recalculated toachieve the removal of the objective fractions of W and thecurrent block shall be continued to achieve the recalculatedtime. If the intended deplasticizing has been exceeded, thefuture deplasticizing times shall be recalculated and reducedaccordingly.

A.4.4.1.3 Block 3

Block 3 consists of a repeat of Block 2 for no less thanduration T2, including any necessary adjustment of T2 toachieve the intended level of deplasticizing.


A.4.4.1.4 Block 4

Block 4 consists of a repeat of Block 2 for no less thanduration T3 achievement of the objective weight percent ofdeplasticizing W in test samples is to be conÞrmed beforeproceeding to Block 5.

A.4.4.1.5 Block 5

A.4.4.1.5.1 Static Flexible Pipes

Block 5 consists of at least 40 cycles of thermal cycling.If any apparent movement is recorded by changes in

dimensions during the Þrst 40 cycles, the thermal cyclingshall be continued until 20 cycles without any dimensionalchanges are achieved, or until a steady rate of change isachieved.

Each end Þtting shall be inspected after 10 (±1), cycles andthereafter every 10 (±1) cycles if no changes occur, or every 5(±1), cycles if apparent movement occurs.

A.4.4.1.5.2 Dynamic Flexible Pipes

Block 5 consists of at least 40 cycles of thermal cyclingwhile ßexing the pipe through an angle. Flexing is notrequired on the mid-scale tests.

During Block 5, ßexing of at least one end of the test pipeshall be carried out by lifting, or ßexing in a hinged frame to,preferably, a radius of curvature equal to the design minimumfor the pipe structure. The natural radius of curvature result-ing from lifting up one end Þtting to a vertical position, thesecond one being kept at its horizontal position, is acceptablefor samples with a length up to approximately 20 meters. Theßexure shall be repeated at least two times in each tempera-ture cycle while the pipe is at ambient temperature.

If any apparent movement is recorded by changes indimensions during the Þrst 40 cycles, the thermal cyclingshall be continued until 20 cycles without any dimensionalchanges are achieved, or until a steady rate of change isachieved.

Each end Þtting shall be inspected after 10 (±1) cycles andthereafter every 10 (±1) cycles if no changes occur or every 5(±1) cycles if apparent movement occurs.

A.4.4.1.6 Block 6

Block 6 consists of dissecting the end Þttings and measur-ing the plasticizer content under the seal/grip ring and at 2tand 4t (t is the uncompressed sheath thickness) on either sideof the seal grip ring center to conÞrm that the acceptance cri-teria have been met. If the objective weight percent of deplas-ticizing W is not achieved under the seal/grip ring in the Þrstpipe end Þttings, the second pipe shall not be dissected until ithas been subjected to a T3 duration recalculated to achievethe objective.

A.4.4.2 General

The second test pipe shall not be subjected to Block 4 test-ing until the Þrst test pipe has completed Block 6 and thedeplasticizing time T2 has been conÞrmed or corrected.Thereafter, the second test pipes deplasticizing times (T2 andT3) shall be adjusted according to the test results for the Þrstpipe.

To facilitate testing, deplasticizing in Blocks 2, 3, and 4can be continued while the monitoring assemblies are evalu-ated and deplasticizing times (T1, T2, T3) are adjusted.

A.4.5 INSPECTION AND TEST ACTIVITIES

When test blocks include inspection or additional testing itshall be conducted as follows:

a. Inspection: The bore areas of each end Þtting shall beinspected for movement of the layers. The position of theßuid barrier and any sacriÞcial or metallic layers adjacent tothe ßuid barrier which are retained in the end Þtting by theseal/grip ring, relative to a Þxed reference location, shall bemeasured and recorded. Special ÒportsÓ or ÒwindowsÓ mayneed to be cut in the carcass or other layers, or through theend Þtting body, to facilitate such measurements.

b. Pressure Testing: Each pipe shall be subjected to a two-hour leak test at design pressure (or a value agreed by the par-ties) and room temperature at the end of each test block. Forthe mid-scale test samples, the test pressure should be sufÞ-ciently high to achieve axial extension equal to or greater thanthat which would be experienced in the full scale test.

A.5 Acceptance CriteriaThe acceptance criteria for the testing shall include fulÞll-

ment of all three following items:

a. The objective weight percent W of plasticizer shall havebeen removed under the seal/grip ring in at least two end Þt-tings and achieved within 0.5 weight percent in the others.

b. There shall be no leakage, cracking, or blistering.

c. There shall be no evidence of movement under the seal/grip ring; or the movement shall be steady, predictable, andprogressing at a rate that would not cause failure within 20years.

A.6 Technical Issues—Discussion of Parameters

The following paragraphs are a commentary. The protocolincludes these paragraphs as advice upon qualiÞcation, crite-ria, or interpretation of results from the testing. Although theprotocol is aimed to be material independent, the technicalissues discussed below are somewhat more speciÞc to PVDF,for historical reasons.


A.6.1 VOLUMETRIC STABILITY

A.6.1.1 Plasticizer content will decline to zero, followingthe laws of diffusion. If the transported medium is gas orwater, this will be the Þnal condition. If the transportedmedium is crude oil, absorption of some of the crude com-ponents will occur, dependent on the crude and the opera-tional temperature. For PVDF, the equilibrium is expectedsomewhere between 2 to 4 percent DBS by weight for tem-peratures between 110 and 130¡C, with higher levels orreplasticizing at lower temperatures. This has to be veriÞedby small-scale testing, to be reported to the industry steeringcommittee. The differences between crudes (condensate,light, heavy, aromatics or not) are being investigated, theother main parameter being temperature. Tests with a num-ber of crudes are required to build an empirical model ofbehavior.

A.6.1.2 QualiÞcation for crude oil service may be achievedby means of assessment against predicted equilibrium for thespeciÞc Þeld conditionsÑbasically, the qualiÞcation testinghas to show stability in conditions of greater effective deplas-tiÞcation than the predicted Þnal in service conditions.

A.6.1.3 An assessment for crude oil service may be carriedout as in 6.9 below following acceptance of crude oil replasti-Þcation test results determined according to the protocol inAppendix A1.

A.6.2 NUMBER OF TEMPERATURE CYCLES FOR QUALIFICATION

A.6.2.1 For PVDF, it is agreed that 10 (static) cycles aresufÞcient to Òpre-conditionÓ a test pipeÑi.e., generate thepredicted tensile load in the barrier when cooled to the lowesttest temperature and reduce the hysteresis in the response to astable level.

A.6.2.2 Based on the rate of decay to ÒfailureÓ of previousdesign end Þttings in service, and an empirical relationship of1:2 between cycles in the Þeld vs. cycles in test pipes, it isproposed that a further 40 cycles (static or dynamic, depend-ing on the pipe application) after completing the speciÞeddeplastiÞcation process is sufÞcient to demonstrate Þtness forpurpose. Alternatively, if temperature cycling is carried out instages during the deplastiÞcation, the Þnal temperaturecycling series may be reduced to 20 cycles, subject to theminimum total being 50 cycles.

A.6.2.3 Zero movement may be interpreted as permanentlystable. If steady movement is identiÞed, this may be projectedlinearly, based on the progression of the early test specimens(Þeld monitoring is recommended to conÞrm the projectionsare accurate).

A.6.2.4 Simulation of Þeld applications where the servicelife is 20 years and operations involve frequent temperature

cycles might require several additional years of continuouscycling. In practice, therefore, the most practical approachmay be to accept qualiÞcation for the service period simu-lated by the testing, introduce markers in the PVDF barrier,and set up a monitoring program to calibrate against the full-scale test data.

A.6.3 NUMBER AND NATURE OF DYNAMIC FLEXURES FOR QUALIFICATION OF DYNAMIC PIPE

A.6.3.1 It is necessary to ßex at least one end of the testpipe sufÞciently that any interlayer friction between thePVDF layers, and the carcass/PVDF/pressure armor arereleased. This will then ensure that the tension generated inthe critical PVDF layers will be delivered to the crimped seal.Flexing is not required on the mid-scale tests.

A.6.3.2 It is unnecessary to apply a program of ßexures asfor a riser mechanical fatigue test because the bend stiffenerwill reduce the loading at the end Þtting to varying tensionload, which is considerably smaller than the temperatureinduced loading.

A.6.4 DIAMETER SCALING

The key parameters to PVDF barrier behavior in the sealring area are percentage indentation and the related stresses inthe crimp zone. If test results are to be used for other diame-ters, then the indentation of the sheath in radial direction aspercentage of the barrier thickness should be constant. Thefollowing elements must be evaluated in calculating the per-centage indentation or crimp:

a. Crimp geometry (generally scaled to ensure similar stressdistribution).

b. Deßection of any underlying steel supporting inserts.

c. Manufacturing and assembly tolerances: these should beadjusted so that the designs being compared have the sameminimum barrier compression under the crimp ring.

A.6.5 NUMBER OF END FITTINGS AND ALTERNATIVE METHODS OF INTERPRETATION

While one pipe (two end Þttings) may be sufÞcient to iden-tify mechanisms and provide a preliminary basis for qualiÞ-cation, a second test (one pipe, two end Þttings) is required toverify repeatability of results and interpret variability of man-ufacturing tolerances.

It may be possible to use test pipes with end Þtting designswhich are sufÞciently similar, rather than identical. The crite-ria for acceptance of marginally different end Þttings are to beestablished (see below).


A.6.6 CARCASS WEIGHT

The inner layer of PVDF (for risers) intrudes into the spiralspaces in the carcass. The carcass weight is transferred to thisPVDF layer via these protrusions. If the PVDF is a singlelayer construction, it also protrudes into the spiral spaces ofthe pressure armor. By this means, for a static line, anyweight loads are distributed along the suspended pipe length.

Multiple layer risers have a smooth surface between thePVDF layers. Unless the internal pressure is able to transferthe weight loading (plus the temperature cycling induced ten-sile loads), the weight and temperature induced load (propor-tional to barrier thickness) is transferred directly to the upperend Þtting. Based on typical examples of in service condi-tions, it is likely that the barrier weight loading wouldincrease the total loading by 10 to 15 percent.

A.6.7 DIMENSIONAL TOLERANCES

The effect of dimensional tolerances on performance isspeciÞc to the manufacturerÕs end Þtting design. No generalguidance can be given with the exception that production endÞttings must be able to be veriÞed to have assembly toler-ances equal to or better than the tolerances achieved for thetest pipe end Þttings. Minimum barrier indentation percent-age shall be greater than or equal to the indentation percent-age of the qualiÞcation samples.

To provide this veriÞcation, the manufacturer-detaileddesign, design basis, and tolerances all need documenta-tionÑwith the tests as a benchmark.

A.6.8 PRE-DEPLASTIFICATION

Pre-deplastiÞcation of the PVDF sheath prior to assemblymay be used as a means to document a minimum service lifefor pipes that transport hot gas or condensate. The deplastiÞ-cation required is related to the status achieved by suitabletest pipes. As an example, consider the case that the pipequaliÞcation tests have successfully reached 5 percent plasti-cizer, from 12 percent (including temperature cycling toprove end Þtting stability), and the predicted equilibrium for acondensate line to be qualiÞed is 2 percent. In this case, aplasticizer loss of 7 percent has been proven and the end Þt-tings should be pre-deplastiÞed to less than 9 percent (7 per-cent +2 percent)Ñi.e., a pre-deplastiÞcation of more than 3percent to verify long-term stability.

A.6.9 ASSESSMENT OF SERVICE LIFE FOR PIPES IN CRUDE OIL SERVICE

To qualify for long-term service, the percent volumechange V corresponding to the weight percent change Wachieved in the test pipe shall be greater than that determinedby exposure testing as in Appendix A1 for the maximumoperating temperature of the pipe in the given crude, or equiv-alent. If there is evidence of movement of the barrier in the

end Þtting, the service life shall be determined by the creeprate based on temperature cycles over the service life. If thereis no evidence of movement of the barrier, the pipe shall beconsidered qualiÞed.

A.6.10 INTERIM ASSESSMENT OF SERVICE LIFE

If the testing has reached a given percent of DBS (n per-cent) at a point necessary to assess projected service life, thefollowing procedure may be used:

a. The percentage volume change (v percent) correspondingto the percentage weight change (n percent) achieved in thefull-scale test shall be determined.

b. The equilibrium percentage volume change (e percent) forthe production ßuid and maximum production temperatureshall be determined in accordance with Appendix A1.

c. The time (Tv) to reach v percent under the expected tem-perature exposure proÞle shall be determined based on thedecay curve with e percent as the asymptote.

d. The projected service life is the time Tv, subject to veriÞ-cation of the following:

1. Completion of 10 + 40 temperature cycles.

2. No evidence of barrier movement under the crimp seal.

Field monitoring is recommended to conÞrm that Tv isaccurate.

A.6.11 PROJECT SPECIFIC CONSIDERATIONS

Each project needs to assess what elements of the protocoltesting are or are not representative of the projectÕs conditionsand exposures. Some possible differences may occur in thefollowing areas:

a. Top end hang-offÑthe methods and mechanical details ofthe top hang-off of ßexible pipe end Þttings can affect theheating and cooling rates for the end Þtting and pressuresheath, depending on how the structural support may conductheat from the end Þtting or shroud it from wind or other con-vection or cooling effects. Bend stiffeners and other ancillarydevices can also signiÞcantly inßuence the local thermalconditions.

b. Immersion/insulationÑtwo elements of the design sur-rounding the end Þtting can affect both the temperatureextremes and rates of heating and cooling. In particular, someend Þttings are insulated to provide Þre protection while otherend Þttings are mounted subsea. The former are likely toexperience higher steady-state temperatures and slower cool-ing and faster heating rates. Submerged end Þttings are likelyto experience lower steady-state temperatures and faster cool-ing rates and slower heating rates.

c. System blowdownÑgas production system risers may besubject to rapid depressurization or blowdown during process


shut-downs or other emergency activities. Because of theJoule-Thomson effects of natural gas, such blowdowns cancause rapid cooling to low temperatures signiÞcantly belowambient. It may be important to consider the thermal capacityof the gas when assessing the cooling rates and minimumtemperature achieved in the pressure sheath duringblowdown.

A.6.12 OTHER TEST PROTOCOLS

In addition to this protocol, there may be other protocolsdeveloped by other groups. In particular, Sintef in Norwayhas conducted end Þtting tests using mid-scale end Þttingsimulators.

A.6.13 MATERIAL CONSIDERATIONS AND FAILURE MODES

This test protocol focuses on the effects of long contin-uous high temperature exposures with periodic cool-downcycles. These conditions may affect the volatile content ofthe pressure sheath polymer and the stresses that maydevelop in the sheath because of thermal expansion andcontraction. However, there may be other signiÞcant

material consideration and failure modes that could affectend Þtting performance. One example of possible materialconsiderations would be changes in the crystallinity of thepolymer and the associated free volume because of pro-longed high temperature exposures. Additional testing onmaterial samples or end Þttings may be required to fullyunderstand other effects.

A.6.14 NUMBER OF THERMAL TEST CYCLES

These protocols expose end Þttings to 20 initial and 20Þnal thermal cycles after the objective plasticized state isachieved. The number of cycles was chosen based on earlytesting experience and an expectation that the load andstrength conditions would be adequately tested in the Þnal 20cycles. However, it should be recognized that risers and otherßexible pipes may be exposed to signiÞcantly more thermalcycling due to process ÒtripsÓ and other shut-downs. It hasbeen estimated that typical North Sea gas plants may experi-ence 1,000 thermal cycles over a typical 20-year life. Projectsshould consider additional thermal cycling if there is reasonto believe that additional cycling would affect otherwise sta-ble end Þtting performance.

141

APPENDIX A1—PVDF COUPON CRUDE OIL EXPOSURE TEST PROCEDURE

A1.1 Test Protocol

The objective of this protocol is to measure the progress to,and the Þnal state of deplasticizing and replasticizing of,PVDF samples representative of ßexible pipe liners whenexposed to a speciÞc liquid hydrocarbon production ßuids.

Note: The procedure described herein includes the heating and han-dling of hot equipment and hydrocarbon products. It is the responsi-bility of any individuals or organizations using this procedure toassure that all appropriate safety procedures are implemented to pre-vent injuries to personnel or damage to equipment or facilities.

A1.1.1 REQUIRED MATERIALS

A1.1.1.1 PVDF Samples

Fourteen samples of PVDF are required, each approxi-mately 35 x 75 mm. The samples should be ßat and rectangu-lar with opposite sides parallel, adjacent sides perpendicular,and uniform thickness (preferably between 0.5 and 3 mm).The samples shall be of the same grade and have the sameinitial amount of plasticizer (0 to 2 percent) as is used in mak-ing ßexible pipe pressure sheaths, and be taken from exam-ples of extruded sheaths.

A1.1.1.2 Exposure Fluid

Approximately one liter of liquid hydrocarbon is requiredto test 12 samples as described above.

Note: Consideration should be given to using both the gaseous andliquid components (in appropriate ratios) of the production ßuid andusing an autoclave so that the production pressure can be maintainedduring the exposures. Although these effects have not been thor-oughly evaluated, there are indications that some hydrocarbon com-ponents are more effective than others in deplasticizing/replasticizing the PVDF, and the exposure pressure may also affectthe interaction.

A1.1.1.3 Exposure Bottles

Exposure bottles should be 0.5 liter or 1 liter inert auto-clave sample bottles suitable for use at temperatures of 130¡Cwith hydrocarbons.

A1.1.2 MEASUREMENT ACCURACY

Thickness shall be measured to 0.01 mm.

Weights shall be measured to ±0.0001 gram.

Temperatures shall be recorded continuously and shall bemeasured to ±3¡C.

Volumes (using ArchimedesÕ Law) shall be measured to5 mm3.

A1.1.3 PROCEDURE

1. Prepare 12 clean dry samples and uniquely mark eachsample by notching the edges or in some other way thatwill not be obliterated by the exposure.

Remove any loose edges or debris from the samples,wipe them with a dry cloth or paper towel, and place themin a desiccator for 24 hours prior to conducting the fol-lowing steps.

2. Measure and record the thickness and weight (W1) ofeach sample. The samples should not be touched withbare hands during the measurements. Calculate the vol-ume of each sample (V1) using ArchimedesÕ Law anddedicated balance or picnometer, and average samplethickness (tavg).

3. Place 12 samples and approximately one liter of expo-sure ßuid in a closed container that is suitable for heat-ing the ßuid to a temperature, T. (Two separatecontainers with 6 samples and approximately one-halfliter of oil each may be used as an alternative). Heat andmaintain the oil temperature at T. Place the two remain-ing samples in a ventilated oven at 220¡C for 24 hoursand measure the weight (W2) and volume (V2) and cal-culate the initial plasticizer weight percent and the maxi-mum volume change percent.

Note: Initial experiments may be conducted at T = 130¡C toobtain initial results quickly. It is also necessary to identify therelationship of plasticizer/crude equilibrium with different oper-ating temperatures. It is therefore recommended to completethese exposure tests for a range of temperatures to address thisissue.

4. Calculate the following heating times in hours:

T1 = 225 x (tavg)2

T2 = 400 x (tavg)2

The deplastiÞcation and crude absorption process isexpected to be brought to equilibrium at approximatelyT1. It should be noted that tavg is thickness in mm, T1, T2are times in hours.

5. When times T1/4, T1/2, (3*T1)/4, and T1 are achieved,remove two samples from the oil bath. Wash the sampleswith a mild soap and water solution, rinse the sampleswith clean water, and thoroughly dry the surface of thesamples by wiping with a clean dry paper or cloth towel.

Place the samples in a desiccator to cool for 24 hours.When the samples have cooled, measure and record thelength, width, thickness, and weight (W2) of each sampleand calculate the volumes (V2). If the samples deform,


making direct measurement difÞcult, weigh initially inair, and then suspended in water; determine volume byArchimedesÕ principle. Calculate percent weight and per-cent volume change.

6. When times (T2 + T1)/2, and T2 are achieved, remove twosamples from the oil bath and process and calculate as instep 5. Allow the oil bath to cool and dispose of the test oilafter removing the Þnal samples. The test oil should notbe used for further replasticization tests.

7. As an option, the samples tested in steps 5 and 6 may befurther processed immediately after measurement, as fol-lows:

7A. Obtain and uniquely mark six commercially availableinert metal sample cups suitable for weighing the sam-ples from 5 and 6. Place a sample in each cup andmeasure the total weight of each cup and sample.

Place the cups and samples in a vacuum or ventilatedoven at 220¡C.

7B. After heating for 24 hours, remove the samples fromthe oven and place them in a desiccator to cool for 24hours. When the samples have cooled, measure andrecord the weight of each sample.

8. Complete the calculation sheets attached to calculate thenet loss of volatiles weight (Net ∆ Weight), the net changein volume (Net ∆ Volume), and to conÞrm the total weightchange (Total ∆ Weight percent) to be consistent with theinitial plasticizer content.

The Total ∆ Weight percent for the T1, (T2 + T1)/2, andT2 measurements should be consistent (±0.1 percent)between samples; if they are not, the procedure should berepeated with additional samples, and/or considerationgiven to extended tests on at least two samples withlonger time T2.

Note: Calculations as above may then be used to plot the total content of plasticizer and crude components against time.

A1.1.4 DATA FORM

Sample Ident: ____________________________ 1 2 3 4 5 6 7 8 9 10 11 12

Data (ref item 2)

Exposure time (hours)

Thickness

Weight W2

Volume V2

Initial data

Weight W1

Volume V1

Calculations for exposure time

Weight loss ∆W = W1 Ð W2

Vol change ∆V = V1 Ð V2

Percent weight loss ∆W/W1 x 100 percent

Percent vol change ∆V/V1 x 100 percent

Final Data

Cup sample weight W7A

Cup sample weight W7B

Plasticizer remaining (W7A Ð W7B)

143

APPENDIX B—FLEXIBLE PIPE HIGH TEMPERATURE END FITTING QUALIFICATION TEST PROTOCOL: LOW VOLATILE CONTENT POLYMERS

This protocol is a synthesis of the various requirementsand objectives of many ßexible pipe operators and manufac-turers. This test is primarily intended to qualify end Þttingsgenerically rather than for speciÞc project requirements. Sec-tion 6 provides discussion of topics that may be appropriateto tests conducted for speciÞc projects and for interpretingthe results of tests conducted under this protocol for speciÞcprojects. The protocol may also be used together with theCrude Exposure Procedure (Appendix B1) to evaluate endÞtting performance when subjected to speciÞc crude oil envi-ronments. In addition to the mechanical behavior tested bythis protocol, appropriate testing is required to qualify thechemical and physical suitability of the end Þtting and pres-sure sheath materials. The protocol does not qualify thestrength or stiffness of the end Þttings. See Section 6 forother qualiÞcation topics.

Pairs of identical samples will be tested to identical condi-tions. Four end Þttings are required to meet the acceptancecriteria to achieve unrestricted qualiÞcation for the envelopeof service covered by the test conditions.

The protocol may be used to qualify static or dynamic endÞttings. This protocol is applicable for unplasticized poly-mers (those which have only about 2 percent by weight vola-tile content); a separate protocol has been developed forplasticized polymers.

B.1 Test Objective

B.1.1 The test protocol deÞned below provides an indus-try-acceptable methodology to qualify the mechanical perfor-mance of both existing and newly developed end Þttingdesigns for dynamic ßexible pipes made with high tempera-ture polymer internal pressure sheaths for a representativeservice life of 20 years.

B.1.2 The protocol is applicable for unplasticized polymerßuid barriers.

B.1.3 The protocol is applicable for ßexible pipes in oilservice, gas service, or water injection service.

B.1.4 This protocol is based on the concepts that the basepolymer will lose or absorb volatile components during expo-sure to the test media to achieve stable equilibrium states inthe free polymer and under the seal/grip ring. Further, it isassumed that the equilibrium state in either region can becharacterized by a) weighing small samples of material takenfrom the region, b) driving off the volatile content of the sam-ples by heating it to just above the melting temperature of thepolymer, and c) determining the change in weight of the sam-ple ∆W region.

B.1.5 It is assumed throughout this protocol that ∆V isapproximately proportional to ∆W and that both changesresult from the loss or gain of volatile materials with simi-lar densities. The assumption is made as a simpliÞcationthat allows the use of easily made weight change measure-ments to be representative of speciÞc volume changes thatmay take place under the seal ring where they cannot bemeasured directly. For some materials and exposures, itmay be necessary to establish more complex relationshipsbetween changes in volume and weight, based on addi-tional testing.

Note: This protocol may require a mixture of hydrocarbon or otherliquids or gases. Appropriate safety practices will be required to pro-tect test personnel, facilities, and the environment.

B.2 Initial Data

B.2.1 Prior to the start of testing, the manufacturer is tospecify:

a. The rated service temperature for which the end Þttingdesign is being qualiÞed (Thi).

b. The minimum service temperature for which the end Þt-ting design is being qualiÞed (Tlo).

c. The average linear thermal expansion coefÞcient of thematerial between the minimum service temperature and ratedservice temperature (a).

d. The ∆Vs and ∆Ws of the barrier material (expressed as apercentage of weight change ∆W or volume change ∆V asdetermined by tests carried out according to Appendix B1, forboth unconstrained and constrained regions) are those forwhich the test will qualify the end Þtting designs.

B.2.2 Next, for unconstrained regions of the polymer, [1 +a (Thi Ð Tlo)] is compared with (1 + ∆V/100)1/3, leading to twocases:

Case I:

If [1 + a (Thi Ð Tlo)] is equal to or larger than (1 + ∆V/

100)1/3 then:

The thermal expansion during initial thermal cycles willbe the dominating factor. This is based on the reasonableassumption that the time scale characterizing swelling isone order of magnitude longer than the one associatedwith a heating period (days versus hours). No special mea-sures, including monitoring assemblies, are required forqualiÞcation procedure. Blocks 2, 3, and 4 (see B.4.4) canbe omitted.


Case II:

If [1 + a (Thi Ð Tlo)] is smaller than (1 + ∆V/100)1/3 then:

Volume change is relatively large, and long-term integrityof the seal could be affected. This prompts a procedure wherethe manufacturer shall specify ∆Ws and T1 to T3 deÞned as:

∆Ws = Seventy percent of the total expected change in sample weight under the seal/grip ring, over twenty years, ∆W seal.

T1 = The time at the upper test temperature required to change volume by one third of ∆Ws.

T2 = The incremental time at the upper test tempera-ture, beyond T1, required to change the volume by an additional 1/3∆Ws for a total of two thirds of ∆Ws.

T3 = The incremental time at the upper test tempera-ture, beyond T2, required to change the volume by an additional 1/3∆Ws for a total of ∆Ws.

B.3 Test SamplesB.3.1 Two test samples are required. The test samples shallbe complete production ßexible pipes with all layers and fea-tures. All end Þttings shall be of the same design and assem-bled to the same procedure and dimensional tolerancespeciÞcation. Pipe length shall be 10 meters or more. Thepipe annulus should be vented. The pipe should be manufac-tured according to normal procedures, in particular, thehydrostatic test shall be at ambient temperature and shall notexceed 1.5 times the rated design pressure.

B.3.2 Alternatively, two mid-scale test samples can beused. A mid-scale test sample shall include the following:

a. All layers of the full-sized ßexible pipe through the pres-sure armor. Thermal mass or insulation external to thepressure armor should be applied to represent the remainingßexible pipe layers.

b. Mid-scale end Þttings that include identical design, dimen-sions, and tolerancing attributes for the following functions:

¥ Anchoring of the internal carcass.

¥ Anchoring and sealing of the internal pressure sheathand any sacriÞcial or tape layers adjacent to the internalpressure sheath.

¥ Thermal mass and insulation which is representative ofa full end Þtting.

The length of a mid-scale sample is typically 1 to 2 meters.It shall be demonstrated that the tension force applied to apressure sheath seal during thermal cycling of a mid-scale

sample is equal to or greater than that in the full-scale sample.This is accomplished by Þxed restraints applied to both endÞttings. The restraints prevent the mid-scale sample fromchanging length during thermal cycling.

B.3.3 The manufacturer shall have available, for review byany interested parties, detailed records of the as-built mate-rial, dimensions, Þts and clearances of all pieces of the endÞtting and pipe body that may affect the performance of theend Þtting during testing. The records shall include thedimensioned and toleranced manufacturing drawings for thepipe and end Þttings and all manufacturing and procurementprocedures and standards. In addition, the records shallinclude the calculations associated with the initial data (initialmovement, ∆W, T1, T2, T3, etc.)

B.3.4 Four monitoring assemblies (see Figure B-1) shall beplaced inside each test pipe (Case II only). Each assemblymay consist of a square pressure barrier material sample withedge dimensions at least twice the width of the seal grip ring.The barrier material shall be compressed between a rigidplate that is larger than the material sample and a rigid barthat is at least as wide as the seal/grip ring, and longer thanthe material sample width. The percent compression of thematerial sample shall be equal (±5 percent) to the compres-sion achieved under the seal/grip ring.

B.3.5 Alternative monitoring assembly conÞgurations maybe accepted, by agreement. The purpose is to quantifychanges in the volatile content of the seal ring region. It isassumed that a validated analytical or empirical model will bedeveloped by each manufacturer using this protocol for therelationship between volatile components in the content ofthe pipe and at the seal grip ring. Validation will include sur-vey of the barrier condition in the seal area from a dissectedend Þtting, after a documented exposure process.

B.4 Test Procedures

B.4.1 TEST SET-UP

B.4.1.1 The test samples shall be set up initially for statictemperature cycling, and subsequently in a dynamic testbench or alternative test structure, to allow ßexing of theupper end of the test riser. The static phases (Block 1, Block4, see below) may be carried out with the sample on a work-shop ßoor. The dynamic test blocks shall be carried out withthe test sample(s) mounted in a testing apparatus suitable toßex the riser upper end sufÞciently to ensure any effects ofinter-layer friction are removed from the temperature cycling.Dynamic ßexing is not required in the mid-scale test samplesas the axial stiffness of the hoop strength layer alone is negli-gible. Therefore, the inter-layer friction between the internalpressure sheath and the hoop strength layer will not affect theload on the internal pressure sheath anchoring.


B.4.1.2 Thermocouples shall be installed on the inside andoutside of each end Þtting approximately in the plane of theseal grip ring. Additional thermocouples may be applied fordata gathering, at the manufacturerÕs discretion.

B.4.1.3 The test samples shall be Þlled with an oil thatyields a representative amount of equilibrium volume changeof the polymer. Suitable safety precautions shall be taken forall testing.

B.4.1.4 Load cells shall be installed between the Þxedrestraints and the test samples so that axial loads generatedduring thermal cycling may be measured.

B.4.2 TEST TEMPERATURES AND PRESSURES

B.4.2.1 An upper (maximum) and lower (minimum) testtemperature shall be speciÞed by the manufacturer.

B.4.2.2 It is intended that this protocol may be used forqualiÞcation without the application of design margins. Themaximum service temperature for which the pipe becomesqualiÞed shall be the average upper test temperature. Theminimum service temperature for which the pipe becomesqualiÞed shall be the average lower test temperature.

Note: An industry objective upper service temperature is 130¡C. Anindustry objective lower service temperature is Ð25¡C but no higherthan 0¡C. An acceptable value for lower temperature for operationsexcluding blowdown may be Ð5 to Ð8¡C.

B.4.2.3 The internal pressure shall vary with the tempera-ture such that no less than atmospheric pressure is inducedat ambient temperature and a maximum pressure of approx-imately 20 bars is induced at the top ßange at maximum testtemperature. Relief valves shall be provided so that theinternal pressure does not fall below ambient at any time (novacuum).

B.4.2.4 Cooling rates should be no slower than those pre-dicted for a typical Þeld applications. Cooling shall be con-trolled so as to simulate these typical operating conditions.Heating at a slower rate than predicted for typical Þeld appli-cations is acceptable but will increase the time required tocomplete the temperature cycling process.

Note: An industry basis for cooling rate has been agreed as a risertermination at the deck level of an FPSO turret or a semi submers-ible in air. See Section 6 for discussion of ÒHang-offÓ and ÒInsula-tionÓ effects.

B.4.3 THERMAL CYCLING PROCEDURE

B.4.3.1 Full Scale Test

Each thermal cycle shall consist of Þve steps:

¥ Step 1 The pipe internal temperature shall be raised tothe test temperature.

¥ Step 2 After internal and external thermocouples onthe pipe reach a stable temperature, the testtemperature shall be maintained for an addi-tional 24 hours.

¥ Step 3 The test pipe shall be cooled until the internaland external thermocouples stabilize at ambi-ent temperature. Dynamic pipes shall be ßexedat least 2 times while at this step. Cooling shallbe at a rate equivalent to natural convection,with representative temperature gradientwithin the end Þttings.

¥ Step 4 The temperature shall be reduced to the lowertemperature by controlled cooling, until theinternal and external thermocouples stabilize.

¥ Step 5 The temperature shall be maintained at thelower temperature for a minimum of 1 hour.

Note: The soaking period is related to the creep and relaxationbehaviour of the polymer that is considered. The 24-hour period isvalid for PVDF, while other polymers may require different values.

B.4.3.2 Mid Scale Test

Each thermal cycle shall consist of 5 steps:

¥ Step 1 At installation, the restraints shall be adjustedso that the axial force is within 500 N of zerowhile the sample is at ambient temperature.


¥ Step 3 The pipe internal temperature shall be main-tained at THI until 24 hours from the start ofthe heating cycle.

Figure B-1—Monitoring Assembly (Case II Only)

Materialsample

2 x Aor more

A = seal/grip ring width

A


¥ Step 4 The pipe internal temperature shall be reducedto the lower temperature by controlled cooling,until the pipe internal temperature reachesTLO.

¥ Step 5 The pipe internal temperature shall be main-tained at TLO until 24 hours from the start ofthe cooling cycle. The cycle is repeated atStep 2.

B.4.4 TEST BLOCKS

B.4.4.1 Descriptions

B.4.4.1.1 Block 1

Block 1 consists of 10 cycles of static thermal cycling. Thebore of each end Þtting shall be inspected after 5 [±1 and 10(±1)] cycles.

During Block 1 thermal cycling, the pipes should be essen-tially horizontal, and Þttings may be raised for convenience inÞlling, inspecting, etc. with the pipes free to expand and dis-tort as a result of heating and induced loads.

B.4.4.1.2 Block 2

Block 2 (Case II only) consists of exposure at the test tem-perature for no less a period of time than T1. At the end of theBlock a pressure test shall be conducted, one of the monitor-ing assemblies shall be removed from the test pipe and thechange of weight in the center of the material sample shall becompared with the manufacturerÕs predictions. If the pre-dicted change has not been achieved, the exposure times forall Blocks shall be recalculated to achieve the change of theobjective fractions of ∆W and the current block shall be con-tinued to achieve the recalculated time. If the expectedchange has been exceeded, the future times shall be recalcu-lated and reduced accordingly.

B.4.4.1.3 Block 3

Block 3 (Case II only) consists of a repeat of Block 2, forno less than duration T2, including any necessary adjustmentof T2 to achieve the intended level of change.

B.4.4.1.4 Block 4

Block 4 (Case II only) consists of a repeat of block 2 for noless than duration T3. Achievement of the objective ∆W inmonitoring assemblies, is to be conÞrmed before proceedingto Block 5.

B.4.4.1.5 Block 5

B.4.4.1.5.1 Static Flexible Pipes

Block 5 consists of at least 40 cycles of thermal cycling.

If any apparent movement is recorded, by changes indimensions, during the Þrst 40 cycles, the thermal cyclingshall be continued until 20 cycles without any dimensionalchanges are achieved, or until a steady rate of change isachieved.

Each end Þtting shall be inspected after 10 (±1) cycles andthereafter every 10 (±1) cycles if no changes occur, or every 5(±1) cycles if apparent movement occurs.

B.4.4.1.5.2 Dynamic Flexible Pipes

Block 5 consists of at least 40 cycles of thermal cyclingwhile dynamically ßexing the pipe. Flexing is not required onthe mid-scale tests.

If any apparent movement is recorded, by changes indimensions, during the Þrst 40 cycles, the thermal cyclingshall be continued until 20 cycles without any dimensionalchanges are achieved, or until a steady rate of change isachieved.

Each end Þtting shall be inspected after 10 (±1) cycles andthereafter every 10 (±1) cycles if no changes occur, or every 5(±1) cycles if apparent movement occurs.

During Block 5, ßexing of at least one end of the test pipeshall be carried out, by lifting, or ßexing in a hinged frameto a radius of curvature equal to the design minimum for thepipe structure. The ßexure shall be repeated at least 2 timesin each temperature cycle, while the pipe is at ambient tem-perature.

B.4.4.1.6 Block 6

Block 6 consists of dissecting the end Þttings and measur-ing the content of volatile species in the polymer under theseal/grip ring and at 2t and 4t (t is the uncompressed sheaththickness) on either side of the seal grip ring center to conÞrmthat the acceptance criteria have been met. If the objectiveweight percent ∆W is not achieved under the seal/grip ring inthe Þrst pipe end Þttings, the second pipe shall not be dis-sected until it has been subjected to a T3 duration recalculatedto achieve the objective.

B.4.4.2 General

The second test pipe shall not be subjected to Block 4 test-ing until the Þrst test pipe has completed Block 6 and the totaldeplasticizing time (T1 + T2 + T3) has been conÞrmed or cor-rected. Thereafter, the second test pipes exposure times (T2and T3) shall be adjusted according to the test results for theÞrst pipe.

To facilitate testing, deplasticizing in Blocks 2, 3, and 4can be continued while monitoring assemblies are evaluatedand exposure times (T1, T2, T3) are adjusted.


B.4.5 INSPECTION AND TEST ACTIVITIES

When test blocks include Inspection or additional Testingit shall be conducted as follows:

a. InspectionÑthe bore areas of each end Þtting shall beinspected for movement of the layers. The position of theßuid barrier and any sacriÞcial or metallic layers adjacent tothe ßuid barrier which are retained in the end Þtting by theseal/grip ring, relative to a Þxed reference location, shall bemeasured and recorded. Special ÒportsÓ or ÒwindowsÓ mayneed to be cut in the carcass or other layers, or through theend Þtting body, to facilitate such measurements.

b. Pressure testingÑeach pipe shall be subjected to a two-hour leak test at design pressure and room temperature at theend of each test block. For the mid-scale test samples, the testpressure should be sufÞciently high to achieve axial extensionequal to or greater than that which would be experienced inthe full scale test.

B.5 Acceptance CriteriaThe acceptance criteria for the testing shall include compli-

ance with all three of the following items:

a. The objective weight percent change shall have occurredunder the seal/grip ring in at least two end Þttings andachieved within 0.5 weight percent in the others.

b. There shall be no leakage, cracking, splitting, blistering, orother degradation.

c. There shall be no evidence of movement under the seal/grip ring; or the movement shall be steady, predictable, andprogressing at a rate that would not cause failure within 20years.

B.6 Technical Issues—Discussion of Parameters

The following paragraphs are a commentary as adviceupon qualiÞcation, criteria or interpretation of results fromthe testing. Although the protocol is aimed to be materialindependent, the technical issues discussed below are some-what more speciÞc to PVDF, for historical reasons.

B.6.1 VOLUMETRIC STABILITY

Unplasticized materials will swell to some equilibrium,which is related to the exposure media.

The end Þtting, on assembly, may be Òover-squeezedÓ tosimulate the maximum expected swell condition and test car-ried out in a Ònon-swellingÓ oil, or the ßuid used for testingshould be veriÞed as causing swell greater than the opera-tional ßuid.

Small scale swell exposure tests, as in Appendix B1,should be carried out to calibrate barrier response prior to thequaliÞcation tests.

If the barrier material will relax at high temperature overtime, including response to swell, then in the long-term,shrinkage will be larger than (ßuid absorption induced) swell.Cycle time for temperature cycling should take account ofrelaxation time, which should be determined by small-scaletesting beforehand.

B.6.2 NUMBER OF TEMPERATURE CYCLES FOR QUALIFICATION

B.6.2.1 Based on tests with plasticized PVDF, it isaccepted that 10 (static) cycles are sufÞcient to Ôpre-condi-tionÕ a test pipeÑi.e., generate the predicted tensile load inthe barrier when cooled to the lowest test temperature, andreduce the hysteresis in the response to a stable level.

B.6.2.2 Based on the rate of decay to ÒfailureÓ of previousdesign end Þttings in service, and an empirical relationship of1:2 between cycles in the Þeld vs. cycles in test pipes, it isproposed that a further 40 cycles (static or dynamic depend-ing on the pipe application) after completing the speciÞeddeplastiÞcation process is sufÞcient to demonstrate Þtness forpurpose. Alternatively, if temperature cycling is carried out instages during the exposure, the Þnal temperature cyclingseries may be reduced to 20 cycles, subject to the minimumtotal being 50 cycles.

B.6.2.3 Zero movement may be interpreted as permanentlystable. If steady movement is identiÞed, this may be projectedlinearly, based on the progression of the early test specimens.

B.6.2.4 Simulation of Þeld applications where the servicelife is 20 years and operations involve frequent temperaturecycles would require several years of continuous cycling. Inpractice, therefore, the most practical approach may be toaccept qualiÞcation for the service period simulated by thetesting, introduce markers in the PVDF barrier, and set up amonitoring program, to calibrate against the full-scale testdata.

B.6.3 NUMBER AND NATURE OF DYNAMIC FLEXURES FOR QUALIFICATION OF DYNAMIC PIPE

B.6.3.1 It is necessary to ßex at least one end of the testpipe sufÞciently that any interlayer friction between thePVDF layers, and the carcass/PVDF/pressure armor arereleased. This will then ensure that the tension generated inthe critical PVDF layers will be delivered to the crimped seal.Flexing is not required on the mid-scale tests.

B.6.3.2 It is not necessary to apply a program of ßexures asfor a riser mechanical fatigue test, because the bend stiffenerwill reduce the loading at the end Þtting to varying tensionload, which is considerable smaller than the temperatureinduced loading.


B.6.4 DIAMETER SCALING

The key parameters to polymer barrier behavior in the sealring area are percentage indentation and the related stresses inthe crimp zone. If test results are to be used for other diame-ters, then the indentation of the sheath in radial direction aspercentage of the barrier thickness should be constant. Thefollowing elements must be evaluated in calculating the per-centage indentation or crimp:

a. Crimp geometry (generally scaled to ensure similar stressdistribution).

b. Deßection of any underlying steel supporting inserts.

c. Manufacturing and assembly tolerancesÑthese should beadjusted so that the designs being compared have the sameminimum barrier compression under the crimp ring.

B.6.5 NUMBER OF END FITTINGS AND ALTERNATIVE METHODS OF INTERPRETATION

While one pipe (two end Þttings) may be sufÞcient to iden-tify mechanisms and provide a preliminary basis for qualiÞ-cation, a second test (one pipe, two end Þttings) is required toverify repeatability of results, and interpret variability ofmanufacturing tolerances.

It may be possible to use test pipes with end Þtting designsthat are sufÞciently similar, rather than identical. The criteriafor acceptance of marginally different end Þttings is to beestablished (see below).

B.6.6 CARCASS WEIGHT

The inner layer of PVDF (for multilayer PVDF ßuid bar-rier risers) intrudes into the spiral spaces in the carcass. Theweight of the carcass is transferred to this PVDF layer viathese protrusions. If the PVDF is single layer construction, italso protrudes into the spiral spaces of the pressure armor. Bythis means, for a static line, any weight loads are distributedalong the suspended pipe length.

Multiple layer risers have a smooth surface between thePVDF layers, unless the internal pressure is able to transferthe weight loading (plus the temperature cycling induced ten-sile loads), the weight and temperature induced load (propor-tional to barrier thickness) is transferred direct to the upperend Þtting. Based on typical examples of in service condi-tions, it is likely that the barrier weight loading wouldincrease the total loading by 10 to 15 percent.

B.6.7 DIMENSIONAL TOLERANCES

The effect of dimensional tolerances on performance isspeciÞc to the manufacturers end Þtting design. No generalguidance can be given with the exception that production end

Þttings must be able to be veriÞed to have assembly toler-ances equal to or better than the tolerances achieved for thetest pipe end Þttings.

To provide this veriÞcation, the manufacturer detaileddesign, design basis, and tolerances all need documenta-tionÑwith the tests as a benchmark.

B.6.8 ASSESSMENT OF SERVICE LIFE FOR PIPES IN CRUDE OIL SERVICE

To qualify for long-term service, the percent change ∆Wachieved in the test pipe, shall be greater than 70 percent ofthat determined by exposure testing as in Appendix B1 for themaximum operating temperature of the pipe in the givencrude, or equivalent. If there is evidence of movement of thebarrier in the end Þtting, the service life shall be determinedby the creep rate based on temperature cycles over the servicelife. If there is no evidence of movement of the barrier, thepipe shall be considered qualiÞed.

B.6.9 PROJECT SPECIFIC CONSIDERATIONS

Each project needs to assess what elements of the protocoltesting are or are not representative of the projectÕs conditionsand exposures. Some possible differences may occur in thefollowing areas:

a. Top end hang-offÑthe methods and mechanical details ofthe top hang-off of ßexible pipe end Þttings can effect theheating and cooling rates for the end Þtting and pressuresheath depending on how the structural support may conductheat from the end Þtting, or shrouding it from wind or otherconvection or cooling effects. Bend stiffeners and other ancil-lary devices can also signiÞcantly inßuence the local thermalconditions.

b. Immersion/insulationÑtwo elements of the design sur-rounding the end Þtting can affect both the temperatureextremes and rates of heating and cooling. In particular, someend Þttings are insulated to provide Þre protection while otherend Þttings are mounted subsea. The former are likely toexperience higher steady-state temperatures and slower cool-ing and faster heating rates. Submerged end Þttings are likelyto experience lower steady-state temperatures and faster cool-ing rates and slower heating rates.

c. System blowdownÑgas production system risers may besubject to rapid depressurization or blowdown during processshut-downs or other emergency activities. Because of theJoule-Thomson effects of natural gas, such blowdowns cancause rapid cooling to low temperatures signiÞcantly belowambient. It may be important to consider the thermal capacityof the gas when assessing the cooling rates and minimumtemperature achieved in the pressure sheath duringblowdown.


B.6.10 OTHER TEST PROTOCOLS

In addition to this protocol, there may be other protocolsdeveloped by other groups. In particular, Sintef in Norwayhas conducted end Þtting tests using mid-scale end Þttingsimulators.

B.6.11 MATERIAL CONSIDERATIONS AND FAILURE MODES

This test protocol focuses on the effects of long continu-ous high temperature exposures with periodic cool-downcycles. These conditions may affect the volatile content ofthe pressure sheath polymer and the stresses that maydevelop in the sheath due to thermal expansion and contrac-tion. However, there may be other signiÞcant material con-sideration and failure modes that could affect end Þttingperformance. One example of possible material consider-ations would be changes in the crystallinity of the polymerand the associated free volume as a result of prolonged high

temperature exposures. Additional testing on material sam-ples or end Þttings may be required to fully understandother effects.

B.6.12 NUMBER OF THERMAL TEST CYCLES

These protocols expose end Þttings to 20 initial and 20Þnal thermal cycles after the objective plasticized state isachieved. The number of cycles was chosen based on earlytesting experience and an expectation that the load andstrength conditions would be adequately tested in the Þnal 20cycles. However, it should be recognized that risers and otherßexible pipes may be exposed to signiÞcantly more thermalcycling because of process ÒtripsÓ and other shut-downs. Ithas been estimated that typical North Sea gas plants mayexperience 1000 thermal cycles over a typical 20-year life.Projects should consider additional thermal cycling if there isreason to believe that additional cycling would affect other-wise stable end Þtting performance.

151

APPENDIX B1—POLYMER COUPON CRUDE OIL EXPOSURE TEST PROCEDURE

B1.1 Test ProtocolThe objective of this protocol is to measure the progress to,

and the Þnal state of deplasticizing and replasticizing of poly-mer samples representative of ßexible pipe liners, whenexposed to a speciÞc liquid hydrocarbon production ßuids.

Note: The procedure described herein includes the heating and han-dling of hot equipment and hydrocarbon products. It is the responsi-bility of any individuals or organizations using this procedure toassure that all appropriate safety procedures are implemented to pre-vent injuries to personnel or damage to equipment or facilities.

B1.1.1 REQUIRED MATERIALS

B1.1.1.1 Polymer Samples

Fourteen samples of polymer are required, each approxi-mately 35 x 75 mm. The samples should be ßat and rectangu-lar with opposite sides parallel, adjacent sides perpendicular,and uniform thickness (preferably between 0.5 and 3 mm).The samples shall be of the same grade and have the sameinitial amount of plasticizer (0 to 2 percent) as is used in mak-ing ßexible pipe pressure sheaths, and be taken from exam-ples of extruded sheaths.

B1.1.1.2 Exposure Fluid

Approximately one liter of liquid hydrocarbon is requiredto test 12 samples as described above.

B.1.1.3 Exposure Bottles

Exposure bottles should be 0.5 liter or 1 liter inert auto-clave sample bottles suitable for use at temperatures of 130¡Cwith hydrocarbons.

B1.1.2 MEASUREMENT ACCURACY

ThicknessÑshall be measured to 0.01 mm.WeightsÑshall be measured to ±0.0001 gram.Temperatures shall be recorded continuouslyÑshall be

measured to ±3¡C.Volumes (using Archimedes Law)Ñshall be measured to

5 mm3.

B1.1.3 PROCEDURE

1. Prepare 12 clean dry samples and uniquely mark eachsample by notching the edges or in some other way thatwill not be obliterated by the exposure.

Remove any loose edges or debris from the samples,wipe them with a dry cloth or paper towel, and placethem in a desiccator for 24 hours prior to conducting thefollowing steps.

2. Measure and record the thickness and weight (W1) ofeach sample. Do not touch the samples with bare handsduring the measurements. Calculate the volume of eachsample (V1) using Archimedes Law and dedicated bal-ance or picnometer, and average sample thickness (tavg).

3. Place 12 samples and approximately one liter of expo-sure ßuid in a closed container that is suitable for heat-ing the ßuid to a temperature, T. (Two separatecontainers with 6 samples and approximately one-halfliter of oil each may be used as an alternative). Heat andmaintain the oil temperature at T. Place the two remain-ing samples in a ventilated oven at 220¡C for 24 hoursand measure the weight (W2) and volume (V2) and cal-culate the Initial Plasticizer Weight Percent and theMaximum Volume Change Percent.

Note: Initial experiments may be conducted at T = 130¡C toobtain initial results quickly. It is also necessary to identify therelationship of plasticizer/crude equilibrium with different oper-ating temperatures. It is therefore recommended to completethese exposure tests for a range of temperatures to address thisissue.

4. Calculate the following heating times in hours:

T1 = 225 x (tavg)2

T2 = 400 x (tavg)2

The deplastiÞcation and crude absorption process isexpected to be brought to equilibrium at approximatelyT1. It should be noted that tavg is thickness in mm, T1,T2 are times in hours.

5. When times T1/4, T1/2, (3*T1)/4, and T1 are achieved,remove two samples from the oil bath. Wash the sampleswith a mild soap and water solution, rinse the sampleswith clean water, and thoroughly dry the surface of thesamples by wiping with a clean dry paper of cloth towel.

Place the samples in a desiccator to cool for 24 hours.When the samples have cooled, measure and record thelength, width, thickness, and weight (W2) of each sam-ple and calculate the volumes (V2). If the samplesdeform, making direct measurement difÞcult, weigh ini-tially in air, and then suspended in water; determine vol-ume by Archimedes principle. Calculate percent weightand percent volume change.

6. When times (T2 + T1)/2, and T2 are achieved, removetwo samples from the oil bath and process and calculateas in Step 5. Allow the oil bath to cool and dispose of thetest oil after removing the Þnal samples. The test oilshould not be used for further replasticization tests.


7. As an option, the samples tested in Steps 5 and 6 may befurther processed immediately after measurement, asfollows:

7A. Obtain and uniquely mark six commercially availableinert metal sample cups suitable for weighing thesamples from 5 and 6. Place a sample in each cup andmeasure the total weight of each cup and sample.Place the cups and samples in a vacuum or ventilatedoven at 220¡C.

7B. After heating for 24 hours, remove the samples fromthe oven and place them in a desiccator to cool for 24hours. When the samples have cooled, measure andrecord the weight of each sample.

8. Complete the calculation sheets attached to calculate thenet loss of volatiles weight (Net ∆ Weight), the netchange in volume (Net ∆ Volume) and to conÞrm thetotal weight change (Total ∆ Weight Percent) to be con-sistent with the initial plasticizer content.

The Total ∆ Weight Percent for the T1, (T2 + T1)/2,and T2 measurements should be consistent (±0.1 per-cent) between samples; if they are not, the procedureshould be repeated with additional samples, and/or con-sideration given to extended tests on at least two sampleswith longer time T2.

Note: Calculations as above may then be used to plot the total content of plasticizer and crude components against time.

B1.1.4 DATA FORM

Sample Ident: ___________________________ 1 2 3 4 5 6 7 8 9 10 11 12

Data (ref item 2)

Exposure time (hours)

Thickness

Weight W2

Volume V2

Initial data

Weight W1

Volume V1

Calculations for exposure time

Weight loss ∆W = W1 Ð W2

Vol change ∆V = V1 Ð V2

Percent weight loss ∆W/W1 x 100 percent

Percent vol change ∆V/V1 x 100 percent

Final Data

Cup sample weight W7A

Cup sample weight W7B

Plasticizer remaining (W7A Ð W7B)

There’s more where thiscame from.

The American Petroleum Institute provides additional resources and programs to the oil and natural gas industry which are based on APIStandards. For more information, contact:

• Monogram Licensing Program Phone: 202-962-4791Fax: 202-682-8070

• American Petroleum Institute Quality Registrar Phone: 202-962-4791(APIQR) Fax: 202-682-8070

• API Spec Q1 Registration Phone: 202-962-4791Fax: 202-682-8070

• Perforator System Registration Phone: 202-962-4791Fax: 202-682-8070

• Inspector Certification Programs Phone: 202-682-8161Fax: 202-962-4739

• Engine Oil Licensing and Certification System Phone: 202-682-8233(EOLCS) Fax: 202-962-4739

• Training/Workshops Phone: 202-682-8490Fax: 202-962-4797

Check out the API Publications, Programs, and Services Catalog online atwww.api.org.

APIAmerican Petroleum Institute Helping You Get The Job Done Right.SM01.01.02

Available through Global Engineering Documents.

Effective January 1, 2002.

Phone Orders: 1-800-854-7179 (Toll-free in the U.S. and Canada)303-397-7956 (Local and International)

Fax Orders: 303-397-2740Online Orders: www.global.ihs.com

Invoice To (❏ Check here if same as ÒShip ToÓ)

Name:

Title:

Company:

Department:

Address:

City: State/Province:

Zip/Postal Code: Country:

Telephone:

Fax:

E-Mail:

❏ Payment Enclosed ❏ P.O. No. (Enclose Copy)

❏ Charge My Global Account No.

❏ VISA ❏ MasterCard ❏ American Express ❏ Diners Club ❏ Discover

Credit Card No.:

Print Name (As It Appears on Card):

Expiration Date:

Signature:

Quantity Product Number Title Total

Subtotal

Applicable Sales Tax (see below)

Rush Shipping Fee (see below)

Shipping and Handling (see below)

Total (in U.S. Dollars)

★ To be placed on Standing Order for future editions of this publication,place a check mark in the SO column and sign here:

Pricing and availability subject to change without notice.

Date:

SO★ Unit Price

❏ API Member (Check if Yes)

Ship To (UPS will not deliver to a P.O. Box)

Name:

Title:

Company:

Department:

Address:

City: State/Province:

Zip/Postal Code: Country:

Telephone:

Fax:

E-Mail:

Mail Orders – Payment by check or money order in U.S. dollars is required except for established accounts. State and local taxes, $10 processing fee*, and 5% shipping must be added. Sendmail orders to: API Publications, Global Engineering Documents, 15 Inverness Way East, M/S C303B, Englewood, CO 80112-5776, USA.Purchase Orders – Purchase orders are accepted from established accounts. Invoice will include actual freight cost, a $10 processing fee*, plus state and local taxes.Telephone Orders – If ordering by telephone, a $10 processing fee* and actual freight costs will be added to the order.Sales Tax – All U.S. purchases must include applicable state and local sales tax. Customers claiming tax-exempt status must provide Global with a copy of their exemption certificate.Shipping (U.S. Orders) Ð Orders shipped within the U.S. are sent via traceable means. Most orders are shipped the same day. Subscription updates are sent by First-Class Mail. Other options,including next-day service, air service, and fax transmission are available at additional cost. Call 1-800-854-7179 for more information.Shipping (International Orders) – Standard international shipping is by air express courier service. Subscription updates are sent by World Mail. Normal delivery is 3-4 days from shipping date.Rush Shipping Fee – Next Day Delivery orders charge is $20 in addition to the carrier charges. Next Day Delivery orders must be placed by 2:00 p.m. MST to ensure overnight delivery.Returns – All returns must be pre-approved by calling GlobalÕs Customer Service Department at 1-800-624-3974 for information and assistance. There may be a 15% restocking fee. Special orderitems, electronic documents, and age-dated materials are non-returnable.*Minimum Order – There is a $50 minimum for all orders containing hardcopy documents. The $50 minimum applies to the order subtotal including the $10 processing fee, excluding anyapplicable taxes and freight charges. If the total cost of the documents on the order plus the $10 processing fee is less than $50, the processing fee will be increased to bring the order amountup to the $50 minimum. This processing fee will be applied before any applicable deposit account, quantity or member discounts have been applied. There is no minimum for orders containing onlyelectronically delivered documents.

APIAmerican Petroleum Institute

2002 Publications Order Form

G17A02 RP 17A, Design and Operation of Subsea Production Systems $102.00

Spec 17J, Unbonded Flexible Pipe

Spec 17K, Specification for Bonded Flexible Pipe

$ 75.00

$ 75.00

G17J01

G17K01

03/02

Additional copies are available through Global EngineeringDocuments at (800) 854-7179 or (303) 397-7956

Information about API Publications, Programs and Services isavailable on the World Wide Web at: http://www.api.org

Product No. G17B03