DNV-RP-F202 Composite Risers October 2010

RECOMMENDED PRACTICE

DET NORSKE VERITAS

DNV-RP-F202

COMPOSITE RISERS

OCTOBER 2010

FOREWORDDET NORSKE VERITAS (DNV) is an autonomous and independent foundation with the objectives of safeguarding life,property and the environment, at sea and onshore. DNV undertakes classification, certification, and other verification andconsultancy services relating to quality of ships, offshore units and installations, and onshore industries worldwide, and carriesout research in relation to these functions.

DNV service documents consist of amongst other the following types of documents:— Service Specifications. Procedual requirements.— Standards. Technical requirements.— Recommended Practices. Guidance.

The Standards and Recommended Practices are offered within the following areas:A) Qualification, Quality and Safety MethodologyB) Materials TechnologyC) StructuresD) SystemsE) Special FacilitiesF) Pipelines and RisersG) Asset OperationH) Marine OperationsJ) Cleaner EnergyO) Subsea Systems

The electronic pdf version of this document found through http://www.dnv.com is the officially binding version© Det Norske Veritas

Any comments may be sent by e-mail to [email protected] subscription orders or information about subscription terms, please use [email protected] Typesetting (Adobe Frame Maker) by Det Norske Veritas

If any person suffers loss or damage which is proved to have been caused by any negligent act or omission of Det Norske Veritas, then Det Norske Veritas shall pay compensation to such personfor his proved direct loss or damage. However, the compensation shall not exceed an amount equal to ten times the fee charged for the service in question, provided that the maximum compen-sation shall never exceed USD 2 million.In this provision "Det Norske Veritas" shall mean the Foundation Det Norske Veritas as well as all its subsidiaries, directors, officers, employees, agents and any other acting on behalf of DetNorske Veritas.

Recommended Practice DNV-RP-F202, October 2010Changes – Page 3

MOTIVESNo design code for Fibre Reinforced Plastic, often called com-posite structures, exists today except for some special applica-tions like FRP pipes, pressure vessels and ships.The realisation of even simple designs of FRP structures tendsto become a major undertaking due to the lack of applicabledesign standards. It is DNV’s impression that the lack of agood FRP guideline is one of the major obstacles to utilise FRPstructurally in a reliable and economical way.For this reason DNV started a JIP to develop a guideline forcomposite risers directly linked to the newly developed Off-shore Standard for Dynamic (metal) Risers, in response torequest by the industry to develop a specific standard for thisimportant application.Upon termination of the JIP, the members participating i.e.ABB, Conoco, FMC Kongsberg Subsea, Gurit Suprem,Kværner Oilfield Products, Norsk Hydro, Statoil, Timet agreedthat DNV shall transform the resulting project report into aDNV Recommended Practice.The new DNV Recommended Practice is indexed: DNV-RP-F202 Composite Risers, and has a contents layout as shownoverleaf.

CHANGES• GeneralAs of October 2010 all DNV service documents are primarilypublished electronically.In order to ensure a practical transition from the “print” schemeto the “electronic” scheme, all documents having incorporatedamendments and corrections more recent than the date of thelatest printed issue, have been given the date October 2010.An overview of DNV service documents, their update statusand historical “amendments and corrections” may be foundthrough http://www.dnv.com/resources/rules_standards/.• Main changesSince the previous edition (May 2003), this document has beenamended, most recently in April 2009. All changes have beenincorporated and a new date (October 2010) has been given asexplained under “General”.

DET NORSKE VERITAS

Recommended Practice DNV-RP-F202, October 2010 Page 4 – Changes

DET NORSKE VERITAS

Recommended Practice DNV-RP-F202, October 2010 Contents – Page 5

CONTENTS

Sec. 1 General................................................................... 7

A. General....................................................................................7A 100 Introduction....................................................................... 7A 200 Objectives ......................................................................... 7A 300 Scope and application ....................................................... 7A 400 Other codes ....................................................................... 9A 500 Structure of the RP ........................................................... 9

B. Normative References ..........................................................10B 100 Offshore Service Specifications...................................... 10B 200 Offshore Standards ......................................................... 10B 300 Recommended Practices ................................................. 10B 400 DNV Rules...................................................................... 10B 500 DNV Standards for Certification

and Classification notes .................................................. 10B 600 Other (external) references ............................................. 11

C. General Definitions (see DNV-OS-F201) ............................11C 100 Definitions ...................................................................... 11C 200 Verbal forms used........................................................... 11

D. General Abbreviations and Symbols (see DNV-OS-F201).............................................................11

D 100 Abbreviations and symbols............................................. 11

E. Definitions for Composite Risers ......................................... 11E 100 Definitions ...................................................................... 11

F. Abbreviations and Symbols for Composite Risers ..................................................................12

F 100 Symbols and abbreviations ............................................. 12F 200 Ply and laminate co-ordinate systems............................. 13

Sec. 2 Design Philosophy and Design Principles......... 14

A. General..................................................................................14A 100 Objective......................................................................... 14A 200 Applicability ................................................................... 14

B. General Safety Philosophy ...................................................14B 100 General............................................................................ 14

C. Design Format ......................................................................14C 100 General............................................................................ 14C 200 Failure types.................................................................... 14C 300 Reliability based design .................................................. 14C 400 Design by testing combined with analysis...................... 14

Sec. 3 Design Input - Loads .......................................... 15

A. Introduction ..........................................................................15A 100 Introduction..................................................................... 15

B. Product Specifications ..........................................................15B 100 General function or main purpose of the riser ................ 15

C. Division of the Product or Structure into Components, Parts and Details ...................................................................15

C 100 Levels of division............................................................ 15

D. Phases ................................................................................... 15D 100 Phases.............................................................................. 15

E. Safety and Service Classes ...................................................15E 100 Safety classes .................................................................. 15E 200 Service classes ................................................................ 16

F. Loads ....................................................................................16F 100 General............................................................................ 16F 200 The sustained load effect ................................................ 16F 300 The fatigue load effects................................................... 17

G. Environment .........................................................................18G 100 General............................................................................ 18G 200 Effects of the environment on the material properties ... 18

Sec. 4 Analysis Methodology ........................................ 19

A. General..................................................................................19A 100 Objective......................................................................... 19

B. Combination of Load Effects and Environment .........................................................................19

B 100 General............................................................................ 19B 200 Fundamentals .................................................................. 19B 300 Load effect and environmental conditions

for ultimate limit state..................................................... 19B 400 Load effect and environmental conditions

for time-dependent material properties........................... 19B 500 Load effect and environmental conditions

for fatigue analysis.......................................................... 20B 600 Direct combination of loads and moments ..................... 20

C. Analysis Procedure for Composite Risers ............................20C 100 General............................................................................ 20C 200 'Global - Local' procedure............................................... 20C 300 Global procedure with response surface......................... 20C 400 Fatigue and long term analysis for composite risers ...... 22

D. Local Analysis ......................................................................22D 100 General............................................................................ 22D 200 Input data ........................................................................ 22D 300 Analysis types................................................................. 22D 400 Local linear analysis with degraded properties .............. 22D 500 Local progressive analysis .............................................. 23

E. Analytical Methods...............................................................23E 100 General............................................................................ 23E 200 Assumptions and Limitations ......................................... 23E 300 Link to Numerical Methods............................................ 23

F. Local Finite Element Analysis..............................................23F 100 General............................................................................ 23F 200 Modelling of structures – general ................................... 23F 300 Software requirements .................................................... 24F 400 Execution of analysis ...................................................... 25F 500 Evaluation of results ....................................................... 25F 600 Validation and Verification ............................................ 25

G. Local Dynamic Response Analysis ......................................25G 100 General............................................................................ 25

H. Impact Response...................................................................25H 100 General............................................................................ 25

I. Thermal Stresses...................................................................25I 100 General............................................................................ 25

J. Swelling Effects....................................................................25J 100 General............................................................................ 25

K. Buckling................................................................................25K 100 General............................................................................ 25K 200 Buckling analysis of isolated components...................... 26K 300 Buckling analysis of more complex elements

or entire structures .......................................................... 26

L. Partial Load-Model Factor....................................................26L 100 General............................................................................ 26L 200 Connection between partial load-model factor

and analytical analysis .................................................... 26L 300 Connection between partial load-model factor

and finite element analysis.............................................. 27L 400 Connection between partial load-model factor

and dynamic response analysis ....................................... 27

Sec. 5 Design Criteria for Riser Pipes ......................... 28

A. General..................................................................................28A 100 Objective......................................................................... 28A 200 Application ..................................................................... 28A 300 Pressure testing ............................................................... 28A 400 Limit states...................................................................... 28

DET NORSKE VERITAS

Recommended Practice DNV-RP-F202, October 2010 Page 6 – Contents

B. Load Effects.......................................................................... 29B 100 Design load effects..........................................................29B 200 Load effect factors ..........................................................29B 300 Load model factors..........................................................29

C. Resistance ............................................................................. 29C 100 Resistance factors............................................................29C 200 Geometrical parameters ..................................................30C 300 Material strength .............................................................30C 400 Resistance model factors.................................................30C 500 System effect factor ........................................................30

D. Ultimate Limit State ............................................................. 30D 100 General ............................................................................30D 200 Bursting ..........................................................................31D 300 Liquid tightness - leakage ...............................................31D 400 Buckling .........................................................................31D 500 Propagating buckling ......................................................33D 600 Wear and tear ..................................................................33D 700 Explosive decompression................................................33D 800 Chemical decomposition - corrosion ..............................33D 900 Displacement controlled conditions................................33

E. Fatigue Limit State ............................................................... 33E 100 General ............................................................................33E 200 Cyclic fatigue ..................................................................33E 300 Stress rupture...................................................................33E 400 Factors for static and dynamic fatigue analysis ..............33

F. Accidental Limit State .......................................................... 34F 100 General ............................................................................34F 200 Resistance against fire.....................................................34F 300 Resistance against dropped objects - impact...................34F 400 Impact testing ..................................................................34F 500 Evaluation after impact testing .......................................34

G. Serviceability Limit State ..................................................... 34G 100 General ............................................................................34

H. Special Considerations ......................................................... 34H 100 Interference .....................................................................34H 200 Unstable Fracture and Gross Plastic Deformation ..........34

Sec. 6 Connectors and Liners....................................... 35

A. General.................................................................................. 35A 100 Objective .........................................................................35A 200 Definition of joint............................................................35

B. Connector Designs................................................................ 35B 100 Functional requirements..................................................35B 200 Design and qualification considerations .........................35

C. Composite - Metal Connector Interface ............................... 35C 100 General ............................................................................35C 200 Limit states ......................................................................35

D. Inner Liner ............................................................................ 36D 100 General ............................................................................36D 200 Mechanical performance.................................................36D 300 Autofretage......................................................................36D 400 Liner buckling .................................................................36D 500 Liner composite interface................................................37D 600 Liner to end connector interface .....................................37D 700 Wear and tear .................................................................37

E. Outer Liner............................................................................37E 100 General ............................................................................37E 200 Mechanical performance.................................................38E 300 Blow out of outer liner ....................................................38

F. Joints of Materials or Components - general aspects ...........38F 100 Analysis and testing ........................................................38F 200 Qualification of analysis method for other load

conditions or for scaled joints .........................................38F 300 Multiple failure modes ....................................................39F 400 Evaluation of in-service experience................................39F 500 Laminated joints..............................................................39F 600 Adhesive joints................................................................39F 700 Mechanical joints ............................................................39

G. Test Requirements ................................................................39G 100 General ............................................................................39G 200 Axial/ pressure test of riser with composite metal

interface...........................................................................40G 300 Cyclic fatigue testing for end fittings and composite

metal interface.................................................................40G 400 Stress rupture testing for end fittings and composite

metal interface.................................................................40G 500 Inner liner test requirements ...........................................41G 600 Specimen geometry - Scaled specimen...........................41

Sec. 7 Materials ............................................................ 42

A. General..................................................................................42A 100 Objective .........................................................................42A 200 Material Description .......................................................42

B. Fabrication ............................................................................42B 100 Objective .........................................................................42B 200 Material Description .......................................................42

Sec. 8 Documentation and Verification....................... 43

A. General..................................................................................43A 100 Documentation and verification .....................................43

Sec. 9 Operation, Maintenance, Reassessment, Repair.................................................................. 44

A. General..................................................................................44A 100 Objective .........................................................................44

B. In-service Inspection, Replacement and Monitoring............44B 100 General ............................................................................44B 200 Inspection methods .........................................................44

C. Reassessment ........................................................................44C 100 General ............................................................................44

D. Repair....................................................................................44D 100 General ............................................................................44D 200 Repair procedure .............................................................44D 300 Requirements for a repair................................................44D 400 Qualification of a repair ..................................................44

E. Maintenance..........................................................................45E 100 General ............................................................................45

F. Retirement.............................................................................45F 100 General ............................................................................45

DET NORSKE VERITAS

Recommended Practice DNV-RP-F202, October 2010 Sec.1 – Page 7

SECTION 1GENERAL

A. General

A 100 Introduction101 This Recommended Practice (RP) document gives crite-ria, requirements and guidance on structural design and analy-sis of riser systems made of composite materials exposed tostatic and dynamic loading for use in the offshore petroleumand natural gas industries.102 The major benefits in using this RP comprise:

— provision of riser solutions with consistent safety levelbased on flexible limit state design principles

— application of safety class methodology linking accept-ance criteria to consequence of failure

— provision of state-of-the-art limit state functions in a Loadand Resistance Factor Design (LRFD) format with relia-bility-based calibration of partial safety factors

— guidance and requirements for efficient global and localanalyses and introduction of a consistent link betweendesign checks (failure modes), load conditions and loadeffect assessment in the course of the global and localanalyses

— allowance for the use of innovative techniques and proce-dures, such as reliability-based design methods.

103 The basic design principles and functional requirementscomply with state-of-the-art industry practice.

A 200 Objectives201 The main objectives of this RP are to:

— provide an international RP of safety for composite risersutilised for drilling, completion/ workover, production/injection, or transportation of hydrocarbons (import/export) in the petroleum and gas industries

— serve as a technical reference document in contractualmatters, and

— reflect the state-of-the-art and consensus on acceptedindustry practice and serve as a RP for riser design andanalysis.

A 300 Scope and application301 This RP provides the design philosophy, loads and glo-bal analysis aspects valid for risers made of composite materi-als. The RP applies to all new built riser systems and may beapplied to modification, operation and upgrading of existingrisers.302 The risers covered in the RP can be jointed or continu-ous. Bonded rubber risers and risers with un-bonded load bear-ing structures are not included. Applications are production,drilling and injection risers, as well as choke and kill lines.303 Composites are fibre reinforced plastics. The fibresshould have a higher modulus than the surrounding polymericmatrix material. The matrix may be thermoset or thermoplas-tic.304 Composite risers have typically internal and external lin-ers around the main pipe section. Any material may be chosen

for the liners, as long as long term performance of the linerscan be demonstrated. Standards related to chosen liner materialshall be used to document liner performance. Additionalrequirements to liners and interfaces are given in Sec.6.305 Composite risers have typically metal end flanges. Anymaterial may be chosen for the flanges, as long as long termperformance of the flanges can be demonstrated. Standardsrelated to chosen flange material shall be used to documentperformance of the flanges. Additional requirements to endflanges are given in Sec.6 (composite metal interface).306 The scope covers design, materials, fabrication, testing,operation, maintenance and re-assessment of riser systems.Aspects relating to documentation, verification and qualitycontrol are also addressed. The main purpose is to cover designand analysis of top tensioned and compliant composite risersystems operated from floaters and fixed platforms. The RPapplies for permanent operation (e.g. production and export/import of hydrocarbons and injection of fluids), as well as fortemporary operation (e.g. drilling and completion/ workoveractivities).307 This RP is applicable to structural design of all pressurecontaining components that comprise the riser system. Othercomposite components can be designed according to DNV-OS-C501.

Guidance note:Most composite risers of today consist of metallic or polymericliners within the composite pipes. The purpose of the liners is toprevent leakage of the riser, while the composite pipes are theload carrying part of the riser system. This RP covers risers with(and without) liners as well as riser connectors and other risercomponents such as tension joints and stress joints.

---e-n-d---of---G-u-i-d-a-n-c-e---n-o-t-e---

308 There are, in principle, no limitations regarding floatertype, water depth, riser application and configuration. How-ever, for novel applications where experience is limited, spe-cial attention shall be given to identify possible new failuremechanisms, validity/adequacy of analysis methodology andnew loads and load combinations.

Guidance note:Composite risers are novel applications and it shall be docu-mented that the global load effects can be predicted with sameprecision as for conventional riser systems. This may typicallyinvolve validation of computational methodology by physicaltesting.As an alternative, an appropriate conservatism in design shouldbe documented.Procedures of DNV-RP-A203 “Qualification of new technol-ogy” should be considered.


309 Examples of typical floater and riser configurations areshown schematically in Fig. 1. Examples of some typical com-ponents/important areas included in typical riser systems areillustrated in Fig. 2.

DET NORSKE VERITAS

Recommended Practice DNV-RP-F202, October 2010 Page 8 – Sec.1

Figure 1 Examples of typical riser configurations and floaters

Figure 2 Examples of riser components

DET NORSKE VERITAS


A 400 Other codes401 This RP shall be used in combination with the standardsfor dynamic risers and submarine pipeline systems denotedDNV-OS-F201 and DNV-OS-F101, respectively. This RPshall not be used as a stand-alone document. The RP is alsorelated to the offshore standard for composite componentsdenoted DNV-OS-C501. The limit state design checks for thisRP and DNV-OS-F201 and DNV-OS-F101 are similar, butdue to differences in the governing failure modes and prevail-ing uncertainties, some differences in safety factors exist.402 Where reference is made to codes other than DNV doc-uments, the valid revision shall be taken as the revision thatwas current at the date of issue of this RP unless otherwisenoted, see list under B600.403 The framework within DNV riser standards and RPs isillustrated in Fig. 3.

Figure 3 Framework for DNV riser standards and RPs

404 This RP provides specific aspects related to compositerisers, including material description, local analysis and designcriteria. General design philosophy, loads and global analysisaspects valid for all riser materials are covered by the DNV-

OS-F201. The present RP document subscribes, for consist-ency, to the safety philosophy and analyses methodology setforward by this standard.

A 500 Structure of the RP 501 This RP is organised as follows:Section 1 contains the objectives and scope of the RP. It fur-ther introduces essential concepts, definitions and abbrevia-tions.Section 2 contains additions to the fundamental design philos-ophy and design principles in DNV-OS-F201.Section 3 in DNV-OS-F201 contains a classification of loadsinto pressure loads, functional loads and environmental loads.Important internal pressure definitions are given. This RP con-tains additional aspects that should be considered for compos-ite risers. In particular the description of long term loads andenvironments.Section 4 in DNV-OS-F201 contains the framework for globalanalysis methodology. This RP provides some additions to thecombination of long term loads and concentrates mainly on thelocal analysis of composite risers.Section 5 contains acceptance criteria for the riser pipe forULS, SLS, ALS and FLS. This includes a definition of resist-ance and load effects and safety factors for explicit limit states.It provides links to DNV-OS-C501 for specific composite fail-ure criteria.Section 6 contains the fundamental functional requirementsfor connectors and liners. It also provides test requirements forthese components.Section 7 contains requirements for materials. They are iden-tical to the requirements in DNV-OS-C501.Section 8 contains requirements for documentation and verifi-cation of the riser system. They are identical to the require-ments in DNV-OS-F201.Section 9 contains basic requirements for operation and in-service operations in addition to DNV-OS-F201.502 The close relationship between this RP and DNV-OS-F201 and DNV-OS-C501 is shown in Fig. 4.

DNV-RP-F202

DNV-OS-F201

Deep water risers

STEEL

Design Philosophy

Loads

Analyses

DNV-OS-F101

Material

Testing

Installation

DNV-RP-F201

TITANIUM

Material

Testing

Design Criteria

COMPOSITE

Materials

Local Analysis

Testing

Design Criteria

DNV Rules

FLEXIBLES

Other

RP’s

CN

Guidelines

Rules

DNV-OS-C501

Composite

Components

DNV-RP-F202

DNV-OS-F201

Deep water risers

STEEL

Design Philosophy

Loads

Analyses

DNV-OS-F101

Material

Testing

Installation

DNV-RP-F201

TITANIUM

Material

Testing

Design Criteria

COMPOSITE

Materials

Local Analysis

Testing

Design Criteria

DNV Rules

FLEXIBLES

Other

RP’s

CN

Guidelines

Rules

DNV-OS-C501

Composite

Components

DNV-OS-F201

Deep water risers

STEEL

Design Philosophy

Loads

Analyses

DNV-OS-F101

Material

Testing

Installation

DNV-OS-F101

Material

Testing

Installation

DNV-RP-F201

TITANIUM

Material

Testing

Design Criteria

COMPOSITE

Materials

Local Analysis

Testing

Design Criteria

DNV Rules

FLEXIBLES

Other

RP’s

CN

Guidelines

Rules

DNV-OS-C501

Composite

Components

DNV-OS-C501

Composite

Components

DET NORSKE VERITAS


Figure 4 Relationship between this RP, DNV-OS-F201 and DNV-OS-C501

B. Normative ReferencesB 100 Offshore Service Specifications101 The following Offshore Service Specifications shall beused:

— DNV-OSS-301 Certification and Verification of Pipe-lines.

B 200 Offshore Standards201 The following Offshore Standards shall be used:

— DNV-OS-F101 Submarine Pipeline Systems— DNV-OS-F201 Dynamic Risers— DNV-OS-C105 Structural Design of TLPs by the LRFD

Method— DNV-OS-C106 Structural Design of Deep Draught Float-

ing Units — DNV-OS-C501 Composite Components.

B 300 Recommended Practices301 The following Recommended Practices shall be used:

— DNV-RP-B401 Cathodic Protection Design

— DNV-RP-C203 Fatigue Strength Analysis of OffshoreSteel Structures

— DNV-RP-C205 Environmental Conditions and Environ-mental Loads

— DNV-RP-F101 Corroded Pipelines— DNV-RP-F104 Mechanical Pipeline Couplings— DNV-RP-F105 Free Spanning Pipelines — DNV-RP-F106 Factory applied Pipeline Coatings for Cor-

rosion Control— DNV-RP-F201 Design of Titanium Risers— DNV-RP-O501 Erosive Wear in Piping Systems

B 400 DNV Rules401 The following Rules shall be used:

— Rules for Certification of Flexible Risers and Pipes— Rules for Planning and Execution of Marine operations.

B 500 DNV Standards for Certification and Classifica-tion notes501 The following Standards for Certification and Classifi-cation notes shall be used:

— No. 1.2 Conformity Certification Services, Type Approval

1 General

2 DesignPhilosophy

9 Operation,Repair, Maint.,Reassessment

5 Design Criteriafor Riser Pipes

4 AnalysisMethodology

3 Design InputLoads

8 DocumentationVerification

7 Materials

6 ConnectorsLiners Joints

DNV-RP-F202Composite

Risers

6 Failure Mechanisms and

Criteria

4 Materials

10 Component

testing

1 General

2 Design Philosophy

9 Operation, Repair, Maint ., Reassessment

5 Design Criteria for Riser Pipes

4 Analysis Methodology

3 Loads

8 Documentation Verification

7 Materials

6 Connectors and Components

DNV-OS-F201 Dynamic Risers

DNV-OS-C501 Composite

Components

Content of DNV-RP-F202 Composite Risersand links to other DNV standards

...

...

DET NORSKE VERITAS


— No. 7 Ultrasonic Inspection of Weld Connections— No. 30.4 Foundations— No. 30.6 Structural Reliability Analysis of Marine Struc-

tures

B 600 Other (external) references601 The following other references shall be used:

— BS 7910 Guide on methods for assessing the acceptabilityof flaws in fusion welded structures

— API RP1111 Design, Construction, Operation, and Main-tenance of Offshore Hydrocarbon Pipelines (Limit StateDesign)

— API RP2RD Design of Risers for Floating Production Sys-tems (FPSs) and Tension-Leg Platforms (TLPs)

— EUROCODE 3 Design of steel structures - Part 1.1: Gen-eral rules and rules for building

— ISO/FDIS 2394 General Principles on Reliability forStructures

— ISO/CD 13628-7 Petroleum and natural gas industries -Design and operation of sub-sea production systems - Part7: Completion/ workover riser systems

Guidance note:The latest revision of the referenced documents applies. The lat-est revision of the DNV documents may be found in the publica-tion list at the DNV website www.dnv.com.


C. General Definitions (see DNV-OS-F201)C 100 Definitions101 The general definitions are identical to and as found inDNV-OS-F201.

C 200 Verbal forms used201 “shall”= indicate requirements strictly to be followed inorder to conform to this RP and from which no deviation is per-mitted. 202 “should”= indicate that among several possibilities oneis recommended as particularly suitable, without mentioningor excluding others, or that a certain course of action is pre-ferred but not necessarily required as other possibilities may beapplied subject to agreement.203 “may”= indicate a course of action permissible withinthe limits of the RP.204 "agreement" and or "by agreement" = agreed in writingbetween the manufacturer or contractor, and the purchaser(unless otherwise indicated).

D. General Abbreviations and Symbols (see DNV-OS-F201)

D 100 Abbreviations and symbols101 The general abbreviations and symbols are identical toand as found in DNV-OS-F201.

E. Definitions for Composite RisersE 100 Definitions101 Angle-ply laminate: symmetric laminate, possessingequal plies with positive and negative angles.102 Anisotropy: material properties varying with the orienta-

tion or direction of the reference co-ordinate.103 Characteristic load: reference value of a load to be usedin the determination of the load effects. The characteristic loadis normally based upon a defined fractile in the upper end ofthe distribution function load.104 Characteristic resistance: the nominal value of thestructural strength to be used in the determination of the designstrength. The characteristic resistance is normally based upona defined fractile in the lower end of the distribution functionfor resistance.105 Constituent: in general, an element of a larger grouping.In advanced composites, the principal constituents are thefibres and the matrix.106 Cross-ply laminate: special laminate that contains only0 and 90 degree plies.107 De-lamination: separation or loss of bonds of plies (the2-D layers) of material in a laminate.108 Environmental conditions: environmental exposure thatmay harm or degrade the material constituents.109 Environmental loads: loads due to the environment,such as waves, current, wind, ice, earthquakes.110 Fabric: planar, woven material constructed by interlac-ing yarns, fibres or filaments.111 Failure criterion: criterion to define or identify whenfailure has occurred, usually expressed as an inequality in thegoverning variables, e.g. load greater than resistance.112 Failure mechanism: a mechanism of failure is the under-lying phenomenon at the material level that determines themode of failure. Depending on its level of severity a mecha-nism of failure can lead to various failures. Failure mecha-nisms are specific to material type.113 Failure mode: state of inability to perform a normalfunction, or an event causing an undesirable or adverse condi-tion, e.g. violation of functional requirement, loss of compo-nent or system function, or deterioration of functionalcapability to such an extent that the safety of the unit, person-nel or environment is significantly reduced.114 Failure type: failure types are based on safety margin,intrinsic to a given failure mechanism. A distinction is madebetween catastrophic and progressive failures, and betweenfailures with or without reserve capacity during failure.115 Fibre Reinforced Plastic (FRP): a general term poly-meric composite reinforced by fibres.116 Fibre: single filament, rolled or formed in one direction,and used as the principal constituent of woven or non-wovencomposite materials. 117 Filament: the smallest unit of a fibrous material. Thebasic units formed during drawing and spinning, which aregathered into strands of fibre. It is a continuous discrete fibrewith an effective diameter in the range of few micrometersdepending on the source.118 Functional requirement: a functional requirement isdefined as a requirement that the global structure has to fulfil.119 Glass Fibre Reinforced Plastic (GRP): general term pol-ymeric composite reinforced by glass fibres.120 Homogeneous: descriptive term for a material of uni-form composition throughout. A medium that has no internalphysical boundaries.121 Inspection: activities, such as, measuring, examination,testing, gauging one or more characteristic of a product or aservice, and comparing the results with specified requirementsto determine conformity.122 Interface: boundary or transition zone between constitu-ent materials, such as the fibre/matrix interface, or the bound-

DET NORSKE VERITAS


ary between plies of a laminate or layers of a sandwichstructure. Boundary between different materials in a joint. Aninterface can also be the area where two components or partstouch each other.123 Lamina: same as ply.124 Laminae: plural of lamina125 Laminate: layers of a plies bonded together to form asingle structure. Also the process to build a laminate.126 Laminate ply: same as ply.127 Layer: a single layer of reinforcement (see also defini-tion for ply)128 Liner: the thin wall/pipe (usually made of metal) that isapplied within the composite pipe of most composite risers.The purpose of the liner is to avoid leakage of the riser.129 Local analysis: detailed analysis of parts of the riser sys-tem, e.g. critical cross-sections, connectors and joints. Thelocal analysis should provide stresses and strains on the plylevel. 130 Matrix: the cured resin or polymer material in which thefibre system is imbedded in a ply or laminate.131 MCI: metal composite interface132 Monolithic structure: laminate consisting uniquely ofcomposites materials except core materials; also called single-skin structure.133 Off-axis: not coincident with the symmetry axis; alsocalled off-angle.134 On-axis: coincident with the symmetry axis; also calledon-angle.135 Orthotropic: having three mutually perpendicularplanes of material symmetry.136 Ply: basic building block of a laminate with orthotropicproperties. Reinforcement surrounded by a matrix. Severallayers of reinforcement may form a ply. Several plies form alaminate.137 Reinforcement: a strong material embedded into amatrix to improve strength, stiffness or impact resistance.138 Roving: a number of strands, tows, or ends collected intoa parallel bundle with little or no twist.139 Strand: normally an untwisted bundle or assembly ofcontinuous filaments used as a unit, including slivers. Tows,ends, yarn and so forth, sometimes a single filament are calleda strand.140 Stacking sequence: a description of the orientation ofplies in a laminate.

Guidance note:The term stacking sequence is also often used to describe theorder riser joints are mounted to make up an entire riser. It shouldbe clear from the context which definition is valid.


141 Warp: the direction along which yarn is orientated lon-gitudinally in a fabric and perpendicularly to the fill yarn.142 Weft: the transversal threads of fibres in a woven fabricrunning perpendicular to the warp.

F. Abbreviations and Symbols for Composite Risers

F 100 Symbols and abbreviations101 The symbols, abbreviation subscripts etc. given in

Table F1 to Table F5 are used.

Table F1 Definitions of symbols for variablesVariables1,2,3 ply, laminate, or core local co-ordinate system , 1 being

the main directiona half crack lengthai scalarAi,j matrix A components[A] extensional stiffness matrixb widthC swelling agent concentration coefficientCOV coefficient of variation, i.e., standard deviation over

meanD flexural rigiditye core width{e} expensional strain fieldE modulus of elasticityei general expensional strainf correction factor – scalarG shear modulusG strain energy release rateh height of boxed beamH anisotropy factorI 2nd moment of areak scalarK stress intensity factorl lengthm surface massM momentN in-plane loadQi,j matrix Q components[Q] stiffness matrixR Resistance or Radius of pipeS shear stiffness, local or global structure responseSCF stress concentration factorSi,j matrix S components[S] transformed compliance matrixt thicknessT transverse load, temperatureU strain energyu, v, w displacement in (x, y, z)V volume fractionx, y, z global co-ordinate systemΦ failure criteria functionΨ ratio between quantiles in the marginal distributions

and extreme-value distributionsα thermal expansion coefficientα loading mode factorβ thermal swelling coefficient, or boundary conditions

factorε direct strain, i.e. ε1 in the main direction

strain to failure

{ε} strain fieldγ shear strainγF partial load factors γFM partial load and resistance factorγM partial resistance factorsγRd partial model factor, resistance componentγSd partial model factors, load componentμ mean valueν Poissonís ratio, i.e. major ν12, minor ν21

ε̂

DET NORSKE VERITAS


F 200 Ply and laminate co-ordinate systems201 Local co-ordinate system and symmetry planes in anorthotropic bi-directional ply is shown in Fig. 5.

Figure 5 Local co-ordinate system and symmetry planes in an orthotropicbi-directional ply

θ ply angleρ densityσ direct stress, i.e. σ1 in the main direction, or standard

deviationstrength, or stress to failure

{σ} stress fieldτ shear stress, i.e. τ12 (or σ12 sometimes)ω angular velocity

Table F2 Definitions of subscriptsSubscriptsb bending effectsben bendingc compressioncore corecorrected value corrected by using a correction factorcr criticald designDelam delaminationE(n) time curveface faceFiber fiberi effects due to in-plane size of sandwich beamip effects due to in-plane size of sandwich panelk characteristic valueMatrix matrixmax maximummeas measured valuemin minimumnom nominalply plyref mean of the measured valuesShear shearsl shear-loadedSLS serviceability limit statet tensiontc core thickness effectstyp typical valueULS ultimate limit state

σ̂

Table F3 Definitions of superscriptsSuper-scripts ¯ maximum direct or shear stress in the structure/com-

ponent^ direct or shear stress of material at failure* elastic or shear modulus of damaged face or core nl non-linearlin linear0 initial1 finaltop top facebottom bottom face

Table F4 Definitions of sub-subscriptsSub-subscriptslin linear limit

y, v z, w

x, u O

23

1’

2’ 3

1

DET NORSKE VERITAS


SECTION 2DESIGN PHILOSOPHY AND DESIGN PRINCIPLES

A. GeneralA 100 Objective101 The purpose of this section is to present the safety phi-losophy and corresponding limit state design format applied inthis RP.102 The design philosophy and design principles are thesame as stated in the DNV-OS-F201. This RP refers to thisstandard and addresses additional issues that are relevant forcomposite risers.

A 200 Applicability201 This section applies to all risers that are to be built inaccordance with this RP.

B. General Safety PhilosophyB 100 General101 The general safety philosophy as described in the DNV-OS-F201, is also applicable for composite risers.102 The following issues are addressed in DNV-OS-F201:

— safety objective— systematic review— fundamental requirements— operational considerations— design principles— quality assurance and quality system.

C. Design FormatC 100 General101 The design objective is to keep the failure probability(i.e. probability of exceeding a limit state) below a certainvalue. All aspects described in the DNV-OS-F201 are alsoapplicable for composite risers.102 The following issues are addressed in DNV-OS-F201:

— safety class methodology— design by LRFD-method— reliability based design— design by testing.

103 Additional requirements specific for composite risersare given below.

C 200 Failure types201 Composite materials can fail in different ways than met-als. The safety factors given in this RP are linked to failuretypes that are modelled by the design criterion. Failure typesare based on the degree of pre-warning intrinsic to a given fail-ure mechanism. A distinction is made between catastrophicand progressive failures, and between failures with or withoutreserve capacity during failure. The failure types for each fail-

ure mechanism described in this RP are specified for eachdesign criterion. The specification is based on the following definitions:

— failure type 'ductile' = corresponds to ductile failure mech-anisms with reserve strength capacity. In a wider sense, itcorresponds to progressive non-linear failure mechanismswith reserve capacity during failure. The design criteriondescribes the onset of the failure process, e.g. it is based onthe yield point and not the ultimate strength, even thoughit is used to describe total failure

— failure type 'brittle' = corresponds to brittle failure mecha-nisms. In a wider sense, it corresponds to non-stable fail-ure mechanisms.

202 The different failure types should be used under the fol-lowing conditions for materials that show a yield point.

The failure type 'ductile' may be used if the design criterion isapplied to the yield point , and: σult > 1.2 σyield and εult > 2εyield where σult is the ultimate strength at a strain εult , andσyield is the yield strength at a strain εyield .

The failure type 'ductile' may be used if onset of damage ismodelled, but extensive damage is needed to cause failure, e.g.for the onset of matrix cracking, when failure is related to leak-age and only a substantial number of cracks causes leakage.

In all other cases, the failure type 'brittle' shall be used.

C 300 Reliability based design301 As an alternative to design according to the formatsspecified and used in this RP, a recognised structural reliabilityanalysis (SRA) design method may be used. All requirementsgiven in DNV-OS-F201 shall be followed.302 As far as possible, target reliability levels shall be cali-brated against existing riser designs that are known to haveadequate safety. If this is not feasible, the target safety levelshall be as given in Table C1. The values are nominal valuesreflecting structural failure due to normal variability in loadand resistance but excluding gross error.

C 400 Design by testing combined with analysis401 Testing may be performed as described in DNV-OS-F201. Additional guidance and requirements are given inSec.4, Sec.5 and Sec.6.

Table C1 Target annual failure probabilities PFT for ULS, FLS and ALS

Failure consequenceFailure type Low

safety class

Normalsafety class

Highsafety class

Ductile failure type (e.g. as for steel)

PF = 10-3 PF = 10-4 PF = 10-5

Brittle failure type (base case for composite)

PF = 10-4 PF = 10-5 PF = 10-6

DET NORSKE VERITAS


SECTION 3DESIGN INPUT - LOADS

A. IntroductionA 100 Introduction101 The offshore standard DNV-OS-F201 Sec.3 contains aclassification of loads into - pressure loads, functional loadsand environmental loads. Important internal pressure defini-tions are given. All these are also relevant for composite risers.102 This RP contains additional aspects that should be con-sidered for composite risers. In particular the description oflong term loads and environments.

B. Product SpecificationsB 100 General function or main purpose of the riser101 The general function or the main purpose of the riser andits main interactions with other components and the environ-ment shall be specified in the product specifications.102 The design life in service should be specified in theproduct specifications.

Guidance note:E.g., the riser will work as a production riser for a deep waterfield of 1500 m for 25 years.


C. Division of the Product or Structure into Components, Parts and Details

C 100 Levels of division101 The following levels of division of the riser (product orstructure) are used in this RP:

— riser (structure / product)— sub-structure / sub-product— components— parts— details.

102 The riser can be divided into sub-products or sub-struc-tures, each of which may belong to different safety classes.103 The riser can be divided into components correspondingto the same safety class but may be subject to different func-tional requirements.104 Each component can be divided into parts and each partinto details.

Guidance note:Structure = riserSub-structure = The riser can be divided into to sub-structurescorresponding to different safety classes, e.g. parts of the riserunderneath the platform and parts far away from the platform.Components = the riser could be constituted of an inner liner, anouter shell and the connectors (flanges). The liner’s function is tokeep the riser tight, whereas the shell’s function is to carry thepressure loads. The two components have different functionalrequirements. The connector caries all loads and transfers theloads into the main body of the riser.Parts and details = Different design approaches and design solu-tions may be used for the different parts and details.


105 A structure or substructure is an independent part forwhich a safety class can be defined. Components, parts anddetails are part of a structure or substructure. Failure of any ofthese components, parts or details shall be seen in combinationwith each other.106 The interfaces between parts, components or structuresshould be considered carefully. Interfaces shall be analysed asa part itself if they belong to a continuous structure. If the inter-faces are physical interfaces, the requirements of Sec.6 D andF shall be considered.

D. Phases

D 100 Phases101 The design life of the riser shall be divided into phases,i.e. well-defined periods within the life span of the product.102 All phases that could have an influence on the design ofthe riser shall be considered.103 As a minimum, the construction phase and the operationphase shall be considered. However, it may be convenient tosplit the design life into more detailed phases as shown inTable D1.

104 Spooling should be considered as part of the construc-tion phase if relevant for the riser solution.105 A decommissioning phase may be specified in somecases.106 The duration of each phase should be specified. Espe-cially, the lifetime in service shall be specified.

E. Safety and Service Classes

E 100 Safety classes101 The riser can be divided into sub-structures, each ofwhich may belong to different safety classes.102 For each sub-structure the safety classes, as described inDNV OS F201 Sec.2 C, shall be specified and documented. 103 The safety class of a riser or its sub-structures maychange from one phase to another during the life of the riser.

Table D1 Design phasesManufacturing ConstructionFabrication / AssemblyTransportHandlingStorageInstallationTestingCommissioningOperation OperationMaintenanceBOP handlingRepairRetrieval / re-circulation Post-operation

DET NORSKE VERITAS


E 200 Service classes201 The riser may be divided into sub-structures, each ofwhich may belong to different service classes.

Guidance note:Service classes may be used to discriminate between parts of ariser system with different maintenance requirements. For exam-ple, some parts of a riser system, which are less accessible, couldbe designed for a lower maintenance frequency.


F. LoadsF 100 General101 Loads for composite risers are as specified in DNV-OS-F201. Loads and deformations are categorised into fourgroups:

— pressure (P) loads— functional (F) loads— environmental (E) loads— accidental (A) loads

102 All the load cases shall be described separately for eachphase during the design life of the structure.103 Long term loads need special considerations. The effectof permanent loads like top tension shall be considered andfatigue loads shall be known in terms of mean loads and ampli-tude. More details are given below.

F 200 The sustained load effect201 The sustained load effect value should be used for thedetermination of time-dependent material properties asdescribed in DNV-OS-C501 Sec.4.

Guidance note:In general, it would be very conservative to determine the timedependent degradation of material properties under long-termloads by using the characteristic load effect value (i.e. extremeload effect value). The sustained value is defined in this RP as anaverage load effect value over the lifetime of the product.


202 Sustained load values are defined over an observationperiod, which can correspond to the entire design life of theriser or to a part of that design life. This observation periodshould be divided into several time intervals. Time intervalsshould not be chosen shorter than 1 hour. The maximum lengthof a time interval depends on the load variations. Variations inmagnitude of the load within a time interval shall not be largerthan half the absolute load amplitude during the total observa-tion period.203 Load effects are divided, according to their variationwith time, into:

— permanent load effects; effects likely to act or be sustained

throughout the design life and for which variations in mag-nitude with time are negligible relative to their mean val-ues; or load effects which are monotonically in - ordecreasing until they attain some limiting values

— variable load effects; effects which are unlikely to actthroughout the specified design life or whose variations inmagnitude with time are random rather than monotonicand not negligible relative to their mean values.

204 The sustained value of permanent load effects shall cor-respond to their characteristic value, the 99% quantile in thedistribution of the annual extreme value.205 The sustained value of variable load effects is defined asthe mean value of the effects over the time interval. The sus-tained value Ss during the time interval to is determined suchthat the corresponding total duration above Ss is a portionμ = 0.5 of the exposure period ts. See Fig. 1:

Figure 1 Sustained value of a variable load effect

206 The sustained value of the stress or strain fluctuations(load effect fluctuations) shall be specified within each obser-vation period for each time intervals. A 'table' of the following form should be established.

207 The sustained value of a load effect over an observationperiod may conservatively be chosen as the maximum value ofthat load effect during the observation period.208 The sustained conditions should be considered for fail-ure mechanisms or material property changes governed orinfluenced by long-term load effects.

Guidance note:For example, the sustained load effect value shall be used for thecalculation of creep and for stress rupture.


Exposure time (duration) Sustained valuets Ss

∑ ≤i

si tt μ

tim e

L o ad e ffect S

S u sta in edv alu e S s

ex p osu re p eriod ts

t1 t2 t3

DET NORSKE VERITAS


Figure 2 Division into time intervals and definition of sustained values Ssi for different load effect cases

F 300 The fatigue load effects301 All load effect fluctuations, e.g. stress or strain fluctua-tions, imposed during the entire design life, shall be taken intoaccount when determining the long-term distribution of stressor strain ranges. All loads as given in F100 and all phases shallbe included and both low-cycle fatigue and high-cycle fatigueshall be considered.302 Fatigue may be analysed for load effects in terms ofeither stress or strain. Strain is preferred for composite lami-nates.303 The characteristic distribution of load effect amplitudesshould be taken as the expected distribution of amplitudesdetermined from available data representative for all relevantloads. This is a long-term distribution with a total number ofstress/strain cycles equal to the expected number of stress/strain cycles over a reference period such as the design life ofthe structure. 304 For fatigue analysis, the mean and amplitude of thestress or strain fluctuations shall be specified. A 'table' of thefollowing form should be established.

As an alternative to the presentation in table format, the fatigueloads can be presented on matrix form with one row for eachmean strain, one column for each strain amplitude, and numberof cycles as the entry of each matrix element; i.e.

Matrix representation of rain-flow counted strain amplitudedistribution.

Guidance note:The history of mean and amplitude of stress should be estab-lished on discrete form by a rain-flow analysis.A minimum resolution of the discrete stresses has to be definedbefore the stress history is established.Note that for the fatigue analysis the history of mean stress/ strainand amplitude is needed. In a non-linear analysis, the mean mayshift relative to the amplitude during the transfer from appliedload to load response.If the time duration of some cycles is long or if the mean value isapplied over a long time, these loads may have to be consideredfor sustained load cases (stress rupture) as well.Degradation is a non-linear, history-dependent process. If differ-ent load and environmental conditions can cause different degra-dation histories, all relevant load combinations shall beconsidered.


305 Based on the material properties, in particular the char-acteristic S-N curve and the magnitude of its slope parameter,it shall be assessed whether the bulk of the fatigue damage willbe caused by several thousand or more stress cycles from thecharacteristic stress distribution, or if it will be caused by onlyone or a very few extreme stress amplitudes from this distribu-tion. In the former case, the natural variability in the individualstress amplitudes can be disregarded as its effect on the cumu-lative damage will average out, and the partial load factor canbe set equal to 1.0. In the latter case, the natural variability inthe few governing extreme stress amplitudes cannot be disre-garded and needs to be accounted for by a partial load factorgreater than 1.0. If no detailed analysis of the load factor canbe made, the same factors as those given for static loads shallbe used.

S

t

S s

S

t

S s 1

S s 2

S

t

S s

S

t

S s 1

S s 2

t 1 t 2

S

t

S s 1

S s 2

t 1 t 2

Number of cycles Mean load AmplitudeN S A

DET NORSKE VERITAS


G. Environment

G 100 General101 The term environment designates in this RP the sur-roundings that impose no direct load on the product.

Guidance note:Environment can be chemicals, temperature. The environmentshould not be confused with environmental loads as defined infollowing interactions should be considered:


102 The environment may impose indirect loads on thestructure, e.g. thermal stresses or swelling due to moistureuptake. This should be considered as a load effect and shouldbe calculated according to the relevant parts of Section 4.How-ever, the environment is generally considered for itseffect on the degradation of material strength or change of elas-tic properties.103 The following aspects should be considered when eval-uating the effect of the environment on local volume elementsin a structure:

— direct exposure— possible exposure if protective system fails— exposure after time— exposure after diffusion through a protective layer— exposure after accident— exposure after degradation of a barrier material, or any

material.

Guidance note:The most common environments to be considered are given inTable G1.

104 The time history of all quantities that characterise envi-ronmental conditions (e.g. temperature, humidity) should bedocumented for each phase during the design life of the struc-ture.105 The time history of all environments should be docu-mented for the entire life of the product. Time histories andcharacteristic values should be established according to thesame principles as described for loads in Sec.4 B.

106 Different environmental values are defined in this RP:

— the characteristic value — the sustained value

Guidance note:The definition of the different load values is summarisedin Table G2, for further details the definitions presented in rele-vant chapters shall be used.

For example: when considering temperature as an environment,the following values can be defined:

- sustained environmental value corresponding to the averagetemperature

- extreme environmental value corresponding to the maximumtemperature

- accidental environmental value corresponding to a fire situa-tion

- fatigue environmental values corresponding temperature fluc-tuations imposing thermal stress fluctuations in the material.


107 The notion of fatigue value for the environment is notconsidered in this chapter. If the environment imposes indirectfatigue loads on the structure the loads and their resultingstresses should be considered according to Sec.4 B, e.g. cyclicthermal stresses, stresses from waves and currents etc.108 Different types of loads and environment shall be com-bined. Depending on which load and environment values arecombined, different load and environmental conditions aredefined. These different load and environmental conditionsdefine the different design cases to be considered. Thesedesign cases are described in Sec.4 B.

G 200 Effects of the environment on the material prop-erties 201 All possible changes of material properties due to theeffect of the environment should be considered.

Guidance note:The following interactions should be considered:

- temperature: variation of the mechanical properties (stiffness,strength…)

- exposure to water (salinity / corrosion, marine fouling…)- exposure to humidity- exposure to chemicals- exposure to UV- exposure to other radiation- erosion.


202 The degradation of material properties caused by envi-ronmental conditions is described in DNV-OS-C501 Sec.4.

Table G1 Common environmentsNatural Temperature internal and external

Temperature variationsTemperature gradientsUV radiation (if above the water line)MoistureSea waterAnimals (e.g. shark bites

Functional Transported or contained fluids and chemicalsTemperature internal and externalPressure internal and externalOil spillCleaning materialsPaint solventsAccidental chemicalsFireProcess gas leaksService induced shocksAccidental high pressure steam

Table G2 Load ValuesDesignation Definition To be used forCharacteris-tic value

Extreme value with return period of 100 years

Check of Ultimate Limit States

Sustained value

Average value over a long period

Long-term degra-dation of material properties

Fatigue value

Only for loads

Accidental value

See DNV-OS-F201

DET NORSKE VERITAS


SECTION 4ANALYSIS METHODOLOGY

A. GeneralA 100 Objective101 The purpose of this section is to provide an overview ofthe analysis methodology for composite risers.102 Global analysis shall be performed as described inDNV-OS-F201.103 All phases identified in Sec.3 D shall be analysed.

B. Combination of Load Effects and Environment

B 100 General101 The fundamental approach to combine load effects isdescribed in DNV-OS-F201.102 Combined loading in DNV-OS-F201 is described foracceptance criteria that can be used directly with respect toapplied forces and moments. If such acceptance criteria can befound (see C300), the same methods as in DNV-OS-F201 canbe used. Otherwise, the procedures described in C200 shall beused.103 If the local load effect is linearly proportional to theactual load, loads may be combined directly instead of com-bining load effects. See also DNV-OS-F201 Appendix C onhow to combine loads for non-linear systems.

B 200 Fundamentals201 The combination and severity of load effects and/orenvironmental conditions should be determined taking intoaccount the probability of their simultaneous occurrence.

Guidance note:For example, a severe wave climate producing a large wave loadis usually accompanied by a severe vessel offset producing largeaxial loads or bending moments.


202 Load effects and/or environmental conditions, which aremutually exclusive, should not enter together into a combina-tion, e.g. ice load effects and wave load effects in a riser envi-ronment.203 All directions of load effects are to be taken as equallyprobable, unless data clearly show that the probability ofoccurrence is different in different directions, or unless loadeffects in a particular direction is particularly critical.204 Permanent load effects and permanent environmentalconditions shall be taken into consideration in all combinationsof load effects and environmental conditions. When combinedwith other load effects or environmental conditions, their char-acteristic values shall be included in the combination.205 The following load effect and environmental conditionsare defined in this RP:

— load effects and environmental conditions for ultimatelimit state

— load effects and environmental conditions for time-dependent material properties

— load effects and environmental conditions for fatigue anal-ysis.

206 Table B1 summarises the load and environmental condi-tions that should be considered for the determination of the

time-dependent material properties and those that should beused for the design checks during all phases of the life of theproduct, e.g., installation, transport, operation, etc.

B 300 Load effect and environmental conditions for ultimate limit state301 At any time during the design life of the structure itshould be documented that the structure can fulfil its functionalrequirements for:

— all characteristic load effect values combined with all sus-tained environmental values

— all sustained load effect values combined with all charac-teristic environmental values.

302 When environment and load effect are fully-correlated,their characteristic values shall be combined.303 The combination of characteristic load effects and envi-ronment should be determined such that the combined charac-teristic effect has a return-period of 100 years.

Guidance note:A method to determine the 100-years combined effect of severalload effects and environments is described in this chapter. It isbased on the so-called Turkstra’s rule.


304 When several stochastic load effect and/or environmen-tal conditions occur simultaneously, the extreme combinedeffects of the associated stochastic processes are required fordesign against the ultimate limit state. Each process is charac-terised by a characteristic value. The characteristic values areto be factored and combined to produce a design effect. Forthis purpose, a (limited) number of possible load effect and/orenvironmental condition combinations are considered. Themost unfavourable combination among these shall be foundand will govern the design.305 The most unfavourable relevant combinations shall bedefined for every point in time during the design life.

Guidance note:In most cases, the most unfavourable relevant combinations arethe same over the entire design life. However, in some cases con-ditions may change with time, which may in turn cause changesin the relevant combinations.


B 400 Load effect and environmental conditions for time-dependent material properties401 The sustained load effect values or the fatigue loadeffect values (when relevant) and the sustained environmental

Table B1 Combinations of load and environmental conditions to be considered for the determination of material degradation and for design checks

LoadsCharacteristic

valueSustained

valueFatigue value

Environ-ment

Characteristic value

ULS checkFully correlated

onlySee B302

ULS checkNot fully correlatedSee B304

Sustained value

ULS checkNot fully corre-

latedSee B304

Material degrada-

tionSee B400

Fatigue analysis

See B500

DET NORSKE VERITAS


values should be used for the determination of time-dependentmaterial properties as specified in Sec.3 F200.

B 500 Load effect and environmental conditions for fatigue analysis501 The fatigue load effects should be combined with thesustained environmental values for the fatigue analysis asspecified in Sec.3 F300.

B 600 Direct combination of loads and moments601 The combination of load effects and environments asdescribed above should be used to obtain the load effects, i.e.,local stresses and strains.602 If transfer functions and structural analysis are linear,loads or moments can be combined by the procedures givenabove instead of the load effects.

C. Analysis Procedure for Composite RisersC 100 General101 The global analysis of the riser system shall be per-formed the same way as described in DNV-OS-F201. Detailedlocal analysis should be applied for connectors/ joints andother critical parts of the riser system.102 Risers made of composites possess a complex behaviourdue to the fact that the development of failure in compositematerials usually involves a sequence of failure mechanisms(e.g. matrix cracking, de-lamination and fibre failure), each ofwhich leads to local change of material properties.103 Due to the large number of failure mechanisms and thefact that local effects are crucial for most failure modes relatedto composite structures, it is extremely difficult to establishanalytical acceptance criteria on a global level for all failuremodes. Therefore, local analysis should be extensively used inthe evaluation of failures for composite risers. A method toobtain global acceptance criteria by numerical analysis isgiven in 300.104 The development of local failure mechanisms, with cor-responding local degradation of material properties, may resultin decreased values for the global stiffness parameters. Thismay affect the overall global behaviour (e.g. displacements,bending moments and effective tension) of the riser system.Thus, the parameters that serve as boundary conditions for thelocal analysis may be modified.105 In the following two analysis procedures for compositeriser, systems are recommended. The principal differencebetween the methods is the level on which the failure criteria(or limit states) are evaluated. Another obvious difference,which follows from the prior, is the order in which the globaland local analysis is conducted. Other analysis procedures maybe found in DNV-OS-C502 Sec.9.

C 200 'Global - Local' procedure201 In order to evaluate the limit states one first performsglobal analysis of the entire riser system. The resulting globalload effects (e.g. effective tension, bending moment, thermalloads and internal or external overpressure) serve as boundaryconditions for the forthcoming local analysis.202 Based on the load effects from the global analysis, localanalysis, which leads to local load effects (stresses and strains),is now conducted.203 The local load effects resulting from the local analysisare finally applied in the local acceptance criteria (or failurecriteria) in order to detect possible failure mechanisms of theriser components.204 If the local investigations are performed by progressive

failure analysis (E300), it is possible to detect a sequence of(acceptable) failure mechanisms that may happen prior to thefinal (unacceptable) failure mechanism (often fibre failure).Let us assume that the local analysis predicts the presence ofmatrix cracking somewhere in the riser (and that matrix crack-ing is accepted), which in turn leads to reduced riser stiffness.This local reduction of stiffness may influence the overallbehaviour of the riser system. Therefore, in certain cases it maybe necessary to repeat the global analysis (with degraded mate-rial properties where relevant). Then, the presence of addi-tional failure mechanisms should be investigated through anew local analysis. This iterative procedure (between globalanalysis (with degraded material properties) and detailed localfailure analysis) should be performed until no new failuremechanism is observed (acceptable design) or until a crucialfailure mechanism is predicted (unacceptable design).

Guidance note:The change of axial stiffness due to local degradation mecha-nisms is usually small and does not influence the global loads onthe riser system. In such cases, the global (static and dynamic)analysis does not need to be repeated although the local analysisdemonstrates that (acceptable) failure mechanisms occur. A con-servative approach should be chosen for the simplified analysis.


C 300 Global procedure with response surface301 As an alternative to the global – local procedure pre-sented in 200, a procedure may be used that requires extensivelocal analysis to be conducted prior to the global failure analy-sis. The local analysis is used to establish response surface thatcan be used in subsequent global analysis.302 A riser system is a relatively simple structure on a globalscale. Usually, the riser pipes contain a large number of iden-tical pieces of composite pipes that are all connected with thesame type of connectors or joints. In other applications contin-uous riser pipes, with constant properties along the pipe, maybe used. In all these situations, the following procedure forevaluation of failure may be advantageous.303 Prior to the global analysis of the riser system, globallimit states (on the form gmax = 1) are established by perform-ing local failure analysis (of the pipes as well as the connec-tors/joints) for a large number of combinations of global loadeffects (bending moments, effective tension and internal orexternal overpressure). The global limit states are representedas surfaces in a space/ co-ordinate system with bendingmoments, effective tension and internal/external overpressurealong the axes. The surfaces are obtained by interpolating acollection of points (load cases) from the local analysis thatsatisfies gmax = 1. Such global limit states may be establishedfor several kinds of (local) failure mechanisms.304 After these initial local investigations, the rest of theriser analysis may be performed on a global level.305 If the initial local investigations are conducted by pro-gressive failure analysis (D500) global limit states may beestablished for a wide range of (local) failure mechanisms. Inthis way, an iterative procedure may be adopted. In the firststep (after having established the limit states) global analysis isperformed with initial (non-degraded) stiffness properties. Letus assume that a limit state (corresponding to a non-crucialfailure mechanism) is exceeded in certain global elements.Then the stiffness properties in those elements should bereduced (according to the observed local failure mechanism)and the global analysis should be repeated. This iterationshould continue until no new limit state is exceeded (accepta-ble design) or until a crucial limit state is exceeded (unaccept-able design).

Guidance note:The change of axial stiffness due to local degradation mecha-nisms is usually small and does not influence the global loads onthe riser system. In such cases, the global (static and dynamic)

DET NORSKE VERITAS


analysis does not need to be repeated although the local analysisdemonstrates that (acceptable) failure mechanisms occur.


Guidance note:Example of a global failure criterion. The global failure criterionshould be established for a small section of the riser that repeatsitself along the length of the string. Typically such a sectioncould be a riser joint of about 15m length consisting of a pipe sec-tion with two end fittings. A joint could also be modelled byestablishing two separate response surface, one for the pipe sec-tion and one for the joint. For a long continuous riser a globalfailure criterion would typically only be established for the pipesection. The two joints would be investigated individually.The loads and a riser section are shown schematically in Fig. 1.Typically a section is analysed for the following loads:

— pressure = P— axial load = A— moment = M— torsion = T

Figure 1 General loading conditions for a riser pipe


Guidance note:The axial load can be defined as effective axial load, i.e., Theaxial load without the axial end cap load caused by the pressure,or it can be defined as the absolute axial load. Which choice ismade is a matter of convenience, but it is important to use a con-sistent approach. Torsion can often be neglected for metal risers.However, even small torsional loads may cause damage in acomposite riser, depending on the particular layout and jointgeometry.The selected section of the riser should now be analysed for allpossible combinations of: P, A, M and T.For each combination a stress analysis of the section is carriedout and all relevant failure criteria are checked at all places of thesection. The relevant failure criteria are at least fibre failure andbuckling, but other criteria like matrix cracking may have to beconsidered. Which criteria should be considered is described inSec. 5.Once all combinations of: P, A, M and T have been analysed afour dimensional failure envelope can be defined for that sectionof the riser.To make the example more specific, just a riser pipe section isdescribed in the following part. The same type of arguments canalso be used for joints or a combined pipe-joint analysis. Thelaminate of the pipe has a 0/90 orientation with the same numberof fibres running in the hoop direction as in the axial direction. A typical failure envelope for such a laminate is shown in Fig. 2.

Figure 2 Simple schematic of a failure envelope of a 0/90 laminate


Guidance note:If the pipe is put under internal pressure, the fibres in the hoopdirection see twice as much stress as the fibres in the axial direc-tion (since we have the same number of fibres running in bothdirections). The burst pressure will be related to the maximumstress the fibres can take in the hoop direction, provided the lam-inate is thin and we have the same stress in the hoop fibresthrough the thickness (a condition that is often not fulfilled forcomposite risers). The calculation gives point P1 in the globalfailure envelope on the pressure axis. This is shown in Fig. 3 for a two dimensional P versus A, failurecriterion.

Figure 3 Global failure envelope for pressure and axial force


Guidance note:If the riser is exposed to additional effective axial loads thestresses in the axial fibres will increase. The strength of the axialfibres has to be large enough to carry the applied axial load plusthe end cap load from the pressure. This gives points P2 and P3in the global failure criterion. P2 describes the maximum axialload under maximum pressure. P3 the maximum axial load with-out internal pressure. Ignoring Poisson’s effects and interactionsbetween the fibres, the failure envelope is given by lines betweenP1, P2 and P3.Under external pressure, collapse is defined by a buckling crite-rion. The collapse pressure is shown as P4. If we assume that thecollapse pressure is not effected by an axial load, P5 indicates themaximum external pressure and maximum axial load combina-tion.Many risers are not exposed to compressive axial loads and thefailure envelope is not expanded into that direction in this exam-ple.If the riser sees torsion, the fibres of the 0/90 laminate will not bestressed. Torsional load must be carried by the matrix. The tor-sional load is then proportional to the in-plane shear strength ofthe matrix. Fig. 4 shows this in the global P-A-T failure envelope.

PA

M T

σhoop

σaxial

τ

pressure

axial force

hoop fibre failure axial fibre failure

laminate buckling

P1

P2

P3

P4 P5

DET NORSKE VERITAS


Figure 4 Global failure envelope for pressure, axial force and torsion


Guidance note:The effect of a moment can be added in a similar way. Note thatthe moment puts the highest stresses into the upper fibres andpossibly compressive stresses into the lower fibres. In addition,in-plane shear stresses are highest in the middle. These shearstresses may cause unacceptable matrix shear failure.In reality, the failure envelopes tend to be more complicated,because fibre orientations are more complicated, three dimen-sional stresses should be considered, and in particular the com-posite metal interface may behave totally differently from thesimple pipe. The intention of this example was just to show theprinciple of the development of a global failure envelope.


C 400 Fatigue and long term analysis for composite ris-ers401 The effect of cyclic loads and permanent static loadsshould be evaluated for composite risers.402 The presence of creep, stress relaxation and stress rup-ture-stress relaxation in composite structures depends on thelevel of stresses and or strains and the condition of the constit-uent materials (intact, presence of cracks or other failures).Permanent static load effects should be analysed as describedin Sec.3 F200.403 Development of fatigue failure depends on the strainamplitudes and mean levels during each cycle, as well as thetotal number of cycles. Loads should be analysed as describedin Sec.3 F300.404 The effect of long term loads and environments on thematerial properties should be considered in the analysis.

D. Local AnalysisD 100 General101 In the following two local analysis methods are outlined.More details about the methods and other applicable proce-dures may be found in DNV-OS-C501.102 High pressure risers have generally thick shells and a 3-D analysis is required. The region at and near the joints alsorequires a 3-D analysis. If a 2-D analysis is used it shall beshown that through thickness stresses can be neglected.

D 200 Input data201 The boundary conditions should be selected carefully inorder to represent the nature of the problem in the best possibleway. It should be demonstrated that the chosen boundary con-ditions lead to a realistic or conservative analysis of the struc-ture.

202 Thermal stresses that result from production process orin service loading should be considered in all analysis.203 Stresses due to swelling from absorbed fluids should beincluded if relevant.204 The elastic properties of the materials constituting thestructure should be taken as described in DNV-OS-501 Sec.4.In particular, time-dependent stiffness properties based on theexpected degradation due to environmental and loading condi-tions should be considered. Local variations of these condi-tions should also be considered.205 Laminates should be analysed on the ply level. Each plyshould be described by 4 elastic constants (E1, E2, G12, ν12) forin-plane 2-D analysis and by 9 elastic constants (E1, E2, G12,ν12, E3, G13, G23, ν13, ν23) in 3-D analysis. A nomenclature forthe various elastic constants is defined in Sec.1.206 As an alternative to elastic constants, the stiffness matrixfor orthotropic plies may be used.207 It should be shown that the estimated stiffness gives con-servative results with respect to load effects. The choice ofstiffness values may be different in the cases of strength andstiffness limited design. More details are given in the sectionsbelow.208 If fibres are not oriented in the principle stress directionsthey want to rotate into these directions. This rotation is usu-ally prevented by the matrix. If the matrix cracks or yields, thefibres may be free to rotate slightly. This rotation is usually notmodelled. However one should check that ply stresses trans-verse to the fibres and ply shear stresses are low in a ply withdegraded matrix. Otherwise a reanalysis with rotated fibredirections may be required.

Guidance note:The rotation of fibres may, for example, be important in filamentwound pipe designed for carrying internal pressure. In this casethe fibre orientation is typically about +55°. If the pipe experi-ences a strong axial load in addition to pressure, the fibres wantto orient themselves more into the axial direction.


D 300 Analysis types301 Analytical and or numerical calculations may be used inthe structural analysis. The finite element [FE] method is pres-ently the most commonly used numerical method for structuralanalysis, but other methods, such as finite difference or finiteseries methods may also be applied.

Guidance note:While the FE-method is applicable for a wide range of problems,analytical solutions and the finite series approach often put toomany restrictions on laminate lay-up, geometry etc., and are thusinsufficient in the design of most real world composite struc-tures.


302 Laminate analysis is an additional type of analysis thatis applied to layered composites in order to derive the proper-ties of a laminate from the properties of its constituent plies.303 The structural analysis should be performed for allphases over the entire lifetime of the structure. Initial anddegraded material properties should be considered if relevant.

D 400 Local linear analysis with degraded properties401 In many riser applications (for example risers with lin-ers) several failure mechanisms (e.g. matrix cracking) may beaccepted, while fibre failure is the mechanism of interest. Thelocal analysis of such risers may be performed by this linearprocedure with degraded properties. In certain applicationspresence of matrix cracking in the riser pipe may be acceptable(e.g. for risers with a liner). Assume that fibre failure is theonly failure mechanism of interest. Then the riser may be ana-

pressure

axial force

hoop fibre failure

axial fibre failure

laminate buckling

P1

P2

P3

P4 P5

torsion

DET NORSKE VERITAS


lysed with the assumption that matrix cracking has alreadyoccurred throughout the riser. This means that the riser may bemodelled with reduced stiffness for all riser elements. In mostcases this is a conservative (with respect to local and/or globalload effects) simplification. However, in certain displacementcontrolled problems (for example if the displacement of the topof the riser is dictated by prescribed movement of the platform)the simplification may be non-conservative. This effect shouldbe carefully investigated when relevant.402 The method may be applied for both 2-D and 3-D prob-lems.403 Due to the assumption of matrix cracking in the compo-nent the material properties are degraded in the entire domainby setting certain elasticity parameters to zero (default value).That is, for in-plane 2-D analysis the stiffness in the fibre direc-tion (of each ply) is kept unaltered, while the rest of the prop-erties are assumed to be changed according to DNV-OS-501Sec.4 I. If 3-D analysis is required, the in-plane parameters aredealt with as in the 2-D analysis, while all through thicknessparameters are changed if through thickness stresses causematrix cracking or non-linear deformation of the matrix.404 This method should be mainly used for statically deter-mined problems, as is the case for a riser pipe. Otherwise thissimplified method, with degradation of material properties inthe entire domain, may offer considerably incorrect stress/strain distributions. If the error cannot be analysed andincluded into the model factor a more refined method shall beused.

D 500 Local progressive analysis501 Local progressive analysis, which is presented herein,provides more accurate results than obtained by the simplifiedmethod presented in D400. Instead of degrading almost allparameters in the entire domain, this method is based on a step-wise degradation of a limited number of parameters inbounded regions.502 All kinds of local failure mechanisms may be detectedby the method.503 The method may be applied for both 2-D and 3-D prob-lems.504 Initially, non-degraded ply properties shall be used inthe progressive failure analysis.505 The boundary conditions (load effects from the globalanalysis) for the component are imposed in a step-wise man-ner, as a first step a small portion e.g. 10 % of the load isapplied. Based on this load level, laminate and ply stresses andstrains are calculated and analysed by the relevant failure cri-teria (for each ply). If a failure is detected somewhere in a ply,certain material properties of that ply shall be locally degraded,which means that the parameters shall be reduced in locations(e.g. finite elements) where the failure is detected. Then, thelocal analysis shall be repeated with locally degraded parame-ters for the same load level. If no failure is observed, the loadis increased to e.g. 0.2 x load, and a similar failure analysis isperformed.506 When the analysis finds that the matrix is cracked, theproperties should be changed according to DNV-OS-501 Sec.4I.507 The step-wise increase in loads as indicated in 505 con-tinuous until a critical failure mechanism is observed (unac-ceptable design) or until the entire load is applied and nocritical failure mechanism detected (acceptable design).

E. Analytical MethodsE 100 General101 Analytical methods can be divided into two classes:Analytical solutions of (differential) equations or use of hand-book formulae.

E 200 Assumptions and Limitations201 Analytical methods shall not be used outside theirassumptions and limitations.

Guidance note:The main disadvantage of available analytical solutions is thatsimplifications often put too many restrictions on geometry, lam-inate build-up etc. and hence, are insufficient in the design ofmore complex composite structures.Handbook formulae are usually too simple to cover all the designissues and are also in general not sufficient.Simplified isotropic calculation methods should not be used,unless it can be demonstrated that these methods give validresults.


E 300 Link to Numerical Methods301 Analytical solutions or handbook formulae used withintheir assumptions and limitations may be used to validate finiteelement analysis results.

F. Local Finite Element AnalysisF 100 General101 Only recognised FE-programs should be used. Otherprograms shall be verified by comparison with analytical solu-tions of relevant problems, recognised FE-codes and or exper-imental testing.

F 200 Modelling of structures – general201 Element types shall be chosen based on the physics ofthe problem.202 The choice of the mesh should be based on a systematiciterative process, which includes mesh refinements in areaswith large stress/strain gradients.203 Problems of moderate or large complexity shall be ana-lysed in a stepwise way, starting with a simplified model.204 Model behaviour shall be checked against behaviour ofthe structure. The following modelling aspects shall be treatedcarefully:

— loads— boundary conditions— important and unimportant actions— static, quasi-static or dynamic problem— damping— possibility of buckling— isotropic or an-isotropic material— temperature or strain rate dependent material properties— plastic flow— non-linearities (due to geometrical and material proper-

ties)— membrane effects.

205 Stresses and strains may be evaluated in nodal points orGauss points. Gauss point evaluation is generally most accu-rate, in particular for layered composites, in which the distribu-tion of stresses is discontinuous, and should therefore beapplied whenever possible.

DET NORSKE VERITAS


Guidance note:The analyst shall beware that Gauss point results are calculatedin local (element or ply based) co-ordinates and must be trans-formed (which is automatically performed in most FE codes) inorder to represent global results. Thus, Gauss point evaluation ismore time-consuming than nodal point calculations.


206 Support conditions shall be treated with care. Appar-ently minor changes in support can substantially affect results.In FE-models, supports are typically idealised as completelyrigid, or as ideally hinged, whereas actual supports often liesomewhere in between. In-plane restraints shall also be care-fully treated.207 Joints shall be modelled carefully. Joints may have lessstiffness than inherited in a simple model, which may lead toincorrect predictions of global model stiffness. Individualmodelling of joints is usually not appropriate unless the jointitself is the object of the study. See also requirements for theanalysis of joints in DNV-OS-501.208 Element shapes shall be kept compact and regular to per-form optimally. Different element types have different sensi-tivities to shape distortion. Element compatibility shall be keptsatisfactory to avoid locally poor results, such as artificial dis-continuities. Mesh should be graded rather than piecewise uni-form, thereby avoiding great discrepancy in size betweenadjacent elements.209 Models shall be checked (ideally independently) beforeresults are computed.210 The following points shall be satisfied in order to avoidill-conditioning, locking and instability:

— a stiff element shall not be supported by a flexible element,but rigid-body constraints shall be imposed on the stiff ele-ment

— for plane strain and solid problems, the analyst shall not letthe Poisson’s ratio approach 0.5, unless a special formula-tion is used

— 3-D elements, Mindlin plate or shell elements shall not beallowed to be extremely thin

— the analyst shall not use reduced integration rule withoutbeing aware of possible mechanism (e.g. hourglassmodes).

Guidance note:Some of these difficulties can be detected by error tests in thecoding, such as a test for the condition number of the structurestiffness matrix or a test for diagonal decay during equation solv-ing. Such tests are usually a posteriori rather than a priori.


211 Need for mesh refinement is usually indicated by visualinspection of stress discontinuities in the stress bands. Analo-gous numerical indices are also coded. 212 For local analysis, a local mesh refinement shall be used.In such an analysis, the original mesh is stiffer than the refinedmesh. When the portion of the mesh that contains the refinedmesh is analysed separately, a correction shall be made so theboundary displacements to be imposed on the local mesh areconsistent with the mesh refinement.213 For non-linear problems, the following special consider-ations shall be taken into account:

— the analyst shall make several trial runs in order to dis-cover and remove any mistake

— solution strategy shall be guided by what is learned fromthe previous attempts

— the analyst shall start with a simple model, possibly thelinear form of the problem, and then add the non-lineari-ties one by one.

214 Computed results shall be checked for self-consistencyand compared with, for example, approximate analyticalresults, experimental data, text-book and handbook cases, pre-ceding numerical analysis of similar problems and results pre-dicted for the same problem by another program. Ifdisagreements appear, then the reason for the discrepancy shallbe sought, and the amount of disagreement adequately clari-fied.215 The analyst shall beware the following aspects:

— for vibrations, buckling or non-linear analysis, symmetricgeometry and loads shall be used with care since in suchproblems symmetric response is not guaranteed. Unlesssymmetry is known to prevail, it shall not be imposed bychoice of boundary conditions

— for crack analysis, a quarter point element can be too largeor too small, thereby possibly making results from meshrefinement worse

— the wrong choice of elements may display a dependenceon Poison’s ratio in problems that shall be independent ofPoisson’s ratio

— if plane elements are warped, so that the nodes of the ele-ments are not co-planar, results may be erratic and verysensitive to changes in mesh

— imperfections of load, geometry, supports and mesh maybe far more important in a buckling problem than in prob-lems involving only linear response.

216 In the context of finite element analysis (FEA) of lami-nate structures (one of) the following element types should beapplied:

— layered shell elements with orthotropic material propertiesfor each layer (for in-plane 2-D analysis

— solid elements with orthotropic material properties (for 3-D and through thickness 2-D analysis.

The decision to use 2-D or 3-D analysis methods should bemade depending on the level of significance of through thick-ness stresses and gradients of in-plane stresses through thethickness. A 3-D analysis is usually required for risers withthick walls.

Guidance note:There are two options for the solid elements: The modelling maybe performed with (at least) two solid elements through the thick-ness of each ply. Alternatively, one may apply layered solid ele-ments where the thickness of a single element includes two ormore plies.


F 300 Software requirements301 Selection of finite element software package shall bebased on the followings:

— software availability— availability of qualified personnel having experience with

the software and type of analysis to be carried out— necessary model size— analysis options required— validated software for intended analysis.

302 Useful options for the analysis of composite structuresinclude:

— layered solid elements with orthotropic and an-isotropicmaterial behaviour

— layered shell elements— solid elements with correct material models or appropriate

interface elements allowing for de bond (for analysis ofbonded and laminated joints)

— interface elements allowing for large aspect ratio (for anal-ysis of thin layer bonds)

DET NORSKE VERITAS


— the possibility to select different co-ordinate systems in aclear and unambiguous way.

303 Depending on the area of application, additional analy-sis options should be available, such as:

— appropriate solver with stable and reliable analysis proce-dures

— options characterising large displacements and largestrains (for geometrically non-linear analysis)

— material models describing the behaviour of, e.g., lami-nates beyond first failure (for materially non-linear analy-sis)

— robust incremental procedures (for non-linear analysis ingeneral)

— tools for frequency domain analysis and/or options such astime integration procedures (for dynamic analyses)

— appropriate post-processing functionality— database options— sub-structuring or sub-modelling.

F 400 Execution of analysis401 FEA tasks shall be carried out by qualified engineersunder the supervision of an experienced senior engineer.402 Analysis shall be performed according to a plan, whichhas been defined prior to the analysis.403 Extreme care shall be taken when working with differentrelevant co-ordinate systems, i.e. global, ply based, laminatebased, element based and stiffener based systems.404 The approach shall be documented.

F 500 Evaluation of results501 Analysis results shall be presented in a clear and conciseway using appropriate post-processing options. The use ofgraphics is highly recommended, i.e. contour plots, (amplified)displacement plots, time histories, stress and strain distribu-tions etc.502 The results shall be documented in a way to help thedesigner in assessing the adequacy of the structure, identifyingweaknesses and ways of correcting them and, where desired,optimising the structure.

F 600 Validation and Verification601 FE-programs shall be validated against analytical solu-tions, test results, or shall be benchmarked against a number offinite element programs.602 Analysis designer shall check whether the envisagedcombination of options has been validated by suppliers. If thisis not the case, the necessary validation analysis shall be per-formed.603 FEA-results shall be verified by comparing against rele-vant analytical results, experimental data and/or results fromprevious similar analysis.604 Analysis and model assumptions shall be verified.605 Results shall be checked against the objectives of theanalysis.606 Verification whether the many different relevant co-ordinate systems have been applied correctly shall be consid-ered.

G. Local Dynamic Response AnalysisG 100 General101 In case of accidental loads, such as explosions, dynamiceffects on material properties should be considered carefully. 102 The dependence of the material properties on strain rate

should be taken into account, see DNV-OS-501 Sec.4 C1000.Guidance note:Although static material properties may yield conservative pre-dictions of displacements, a strength assessment based on staticproperties is not necessarily conservative since both the materialstrength and the material stiffness may be enhanced at high strainrates. The higher stiffness may increase the induced stress so thatthe benefit of the increase in the material strength may be lost.Furthermore, ductile materials often become brittle at high rates.Thus, the extra margin provided by ductile behaviour may bedestroyed.There is a lack of sophisticated material models taking the ratedependent behaviour into consideration.


H. Impact ResponseH 100 General101 Impact should be evaluated by testing as described inSec.5 F300.

I. Thermal StressesI 100 General101 Changes in temperature from the environment resultingin dimensional changes of the body shall be taken in account.The general thermal strains, ei, can be expressed as:

αi is the thermal expansion coefficients, and temperature isdenoted by T.102 Residual strains shall be calculated against the referencetemperature for which αi was determined. It is usually the cur-ing temperature103 Accordingly, the stress-strain relations shall be modifiedto account for the stress free environmentally induced expan-sional strains as follows:

J. Swelling EffectsJ 100 General101 Changes in gas/fluid absorption from the environmentresulting in dimensional changes of the body shall be taken inaccount. The general swelling strains, ei, can be expressed as:

βi is the swelling expansion coefficients and C is swellingagent concentration inside the laminate.102 Accordingly, the stress-strain relations shall be modifiedto account for the stress free environmentally induced expen-sional strains as follows:

K. BucklingK 100 General101 The need for special buckling analysis shall be assessedcarefully in every case. In particular the following aspects shallbe considered in making this assessment:

Te ii Δ=α

{ } [ ]{ } { }eS += σε

Ce ii β=

{ } [ ]{ } { }eS += σε

DET NORSKE VERITAS


— presence of axial compressive stresses in the riser pipe— presence of circumferential compressive or shear stresses

in the riser pipe— presence of all compressive stresses in the joint area.

102 All parts of the riser, like pipe, liners and fittings shouldbe evaluated for buckling.103 Two alternative approaches may be used in analysingbuckling problems:

— analysis of isolated components of standard type, such astubular sections, beams, plates and shells of simple shape

— analysis of an entire structure (or of an entire, complexstructural component).

K 200 Buckling analysis of isolated components201 When a member or component that is a part of a largerstructure is analysed separately a global analysis of the struc-ture shall be first applied to establish:

— the effective loading applied to the member/component bythe adjoining structural parts

— the boundary conditions for the structural member, interms of translational and rotational stiffness componentsin all relevant directions.

202 For simple members or components standard formulaeor tables may be used to estimate elastic critical loads (Pe),critical stresses (σe) or critical strains (εe), and the correspond-ing elastic buckling mode shapes. Alternatively these quanti-ties may be calculated using analytical or numerical methods.It shall always be checked that the buckling mode shape is con-sistent with the boundary conditions.203 An assessment shall be made of the shape and size of ini-tial, geometrical imperfections that may influence the bucklingbehaviour of the member. Normally the most critical imperfec-tion shape for a given buckling mode has a similar form to thebuckling mode itself. However, any geometrical feature(including eccentricity of loading) that results in compressiveforces that are not coincident with the neutral axis of the mem-ber may require consideration. The assumed form and ampli-tude of the imperfection shall be decided on the basis of theproduction process used with due consideration of the relevantproduction tolerances, see DNV-OS-C501 Sec.6 H.204 In some cases a geometrically non-linear analysis maybe avoided as follows. The elastic critical load (without imper-fections) Pe is calculated. In addition an ultimate failure loadPf is estimated at which the entire cross-section would fail bycompressive fibre failure, in the absence of bending stresses atthe section in question. If Pe > Pf the further assessment maybe based on geometrically linear analysis provided geometricalimperfections are included and the partial load effect model-ling factor is increased by multiplying it by the factor:

205 In cases where it is possible to establish the bendingresponses (stresses, strains or displacements) associated withan in-plane loading separately from the in-plane (axial)responses, a first estimate of the influence of geometrical non-linearity combined with the imperfection may be obtained bymultiplying the relevant bending response parameter obtainedfrom a geometrically linear analysis by a factor:

and combining the modified bending responses with the(unmodified) in-plane responses.206 The above procedures (205 and 206) may be non-con-servative for some cases where the post-buckling behaviour is

unstable. Examples include cylindrical shells and cylindricalpanels under axial loading. Such cases shall be subject to spe-cial analysis and or tests.

K 300 Buckling analysis of more complex elements or entire structures301 Buckling analysis of more complex elements or entirestructures shall be carried out with the aid of verified finite ele-ment software or equivalent.302 Initially a natural frequency buckling analysis shall beperformed assuming initial (non-degraded) elastic propertiesfor the laminates. This shall be repeated with alternative, finermeshes, until the lowest natural frequency and correspondingmodes are not significantly affected by further refinement. Themain purposes of this analysis are to clarify the relevant buck-ling mode shapes and to establish the required mesh density forsubsequent analysis.303 Careful attention shall be paid to correct modelling ofboundary conditions.304 If the applied load exceeds, or is close to, the calculatedelastic critical load, the design should be modified to improvethe buckling strength before proceeding further.305 A step-by-step analysis shall be carried out. Geometricalnon-linearity shall be included in the model. The failure crite-ria shall be checked at each step. If failure such as matrixcracking or de-lamination is predicted, any analysis for higherloads shall be performed with properties reduced as describedin DNV-OS-C501 Sec.4 I. 306 Alternatively to the requirement in 305 a geometricallynon-linear analysis may be performed using entirely degradedproperties throughout the structure. This will normally provideconservative estimates of stresses and deformations. However,provided reinforcing fibres are present in sufficient directions,so that the largest range of un-reinforced directions does notexceed 60º, such an estimate will not normally be excessivelyconservative.307 The influence of geometric imperfections should beassessed, on the basis of the production method and productiontolerances. See DNV-OS-C501 Sec.6 H.

L. Partial Load-Model FactorL 100 General101 A deterministic factor shall be assigned to each struc-tural analysis method. It is designated in this RP as the partialload-model factor γSd.102 The load-model factor accounts for uncertainties of thestructural analysis method being used to accurately describeand quantify the response of the structure.103 Model factors for the main structural analysis methodsare given in the following sub-sections.104 In some cases a structure is only evaluated by testing,and such an approach evaluates only the particular conditionstested. A procedure for this approach is given in DNV-OS-C501 Sec.10.

L 200 Connection between partial load-model factor and analytical analysis201 When analytical methods are used within their assump-tions and limitations a model factor of 1.0 should be used.202 If analytical methods are used outside their assumptionsand limitations, it shall be documented that the magnitude ofthe model factor ensures that all predicted stresses and strainsare higher than in reality. If the choice of model factor cannotbe documented, the analytical method shall not be used.

ef P4P11

−

eee 11or

11

PP11

εεσσ −−−,

DET NORSKE VERITAS


L 300 Connection between partial load-model factor and finite element analysis301 The accuracy of FE-methods is generally very goodwhen the structure is properly modelled. The use of thesemethods with unsatisfactory models is much more uncertain.302 When FE-methods are used within their assumptionsand limitations (and according to F) a model factor of 1.0 maybe used.303 If FE-methods are used outside their assumptions andlimitations, it shall be documented that the magnitude of themodel factor ensures that all predicted stresses and strains arehigher than in reality. If the model factor cannot be docu-mented, the analysis method shall not be used.

304 If the boundary conditions do not exactly represent thereal conditions the effect on the load model factor shall beevaluated. As a minimum a factor of 1.1 shall be used.305 If the load-model factor cannot be determined for calcu-lations in a critical region, e.g. a critical joint or region of stressconcentrations, experimental qualification should be done (seeDNV-OS-C501 Sec.10).

L 400 Connection between partial load-model factor and dynamic response analysis401 The accuracy of the dynamic analysis shall be estimated.The load-model factor used, which is described in sec. L200and L300, should include all uncertainties due to dynamiceffects.

DET NORSKE VERITAS


SECTION 5DESIGN CRITERIA FOR RISER PIPES

A. GeneralA 100 Objective101 The section provides the general framework for designof riser systems including provisions for checking of limitstates for pipes in riser systems. Design of connectors and risercomponents are covered in Sec.6.

A 200 Application201 This standard provides design checks with emphasis onULS, FLS, SLS and ALS load controlled conditions. Designprinciples for displacement controlled conditions are discussedin D900.202 Requirements for materials, manufacture, fabricationand documentation of riser pipe, components, equipment andstructural items in the riser system are given in DNV-OS-501Sec.4.

A 300 Pressure testing301 All risers of safety class normal or high shall be pressuretested before going into service.302 A test pressure in compliance with DNV-OS-F101should be used unless such a pressure would introduce damageto the component that may reduce its lifetime. The maximumservice pressure shall be the minimum test pressure.303 If the riser contains non-composite parts that weredesigned according to a standard that requires a pressure testup to a certain test pressure - p -, the pressure test shall be car-ried out at that pressure - p - or the pressure required by 302,whatever is highest.304 A detailed test programme should be defined. The fol-lowing should be stated as a minimum:

— rates of pressure increase— holding times— time over which the pressure in the system shall not drop

without actively applying pressure, i.e. a leakage test.

305 The test schedule should be developed for each applica-tion. The testing should allow detecting as many possibledefects in the structure as possible. As a general guidance thefollowing schedules are recommended:

— the minimum time over which the maximum test pressure

in the system should not drop without actively applyingpressure should be at least 10 minutes for systems that donot creep. The pressure should stay constant within 5% ofthe value at the start of the test.

— if the test fluid could possibly migrate slowly throughcracks, materials or interfaces testing up to 24 hours maybe necessary to detect leaks.

— for systems that show creep the maximum test pressureshould be kept for 1 hour applying active pressure. Thepressure should be monitored for another hour withoutactively applying pressure. The pressure drop should bepredicted before the test and the test result should bewithin 10% of the prediction.

306 Risers of low safety class should be tested up to theirdesign pressure. Pressures should be applied for at least 10minutes.307 Most authorities give general test requirements for pres-sure vessels, these may also apply to pressurised risers. Therequirements of the authorities that govern the location of theapplication should be followed.

A 400 Limit states401 The limit states are grouped into the following four cat-egories:

— serviceability limit state (SLS) requires that the riser mustbe able to remain in service and operate properly. Thislimit state corresponds to criteria limiting or governing thenormal operation (functional use) of the riser

— ultimate limit state (ULS) requires that the riser mustremain intact and avoid rupture, but not necessary be ableto operate. For operating condition this limit state corre-sponds to the maximum resistance to applied loads with10-2 annual exceedence probability

— accidental limit state (ALS) is a ULS due to accidentalloads (i.e. infrequent loads)

— fatigue limit state (FLS) is an ultimate limit state fromaccumulated excessive fatigue crack growth or damageunder cyclic loading.

402 As a minimum requirement, the riser pipes and connec-tors shall be designed for (not limited to) the potential modesof failures as listed in Table A1 for all relevant conditionsexpected during the various phases of its life.

DET NORSKE VERITAS


B. Load EffectsB 100 Design load effects101 Design load effects are obtained by multiplying the loadeffect of each category by their corresponding load effect fac-tor. 102 A load model factor shall be determined to account forsystematic errors in calculating local load effects from global

loads or events as described in Sec.4 L.

B 200 Load effect factors201 The design load effect is used in the design checks. Sev-eral combinations may have to be checked when load effectsfrom several load categories enter one design check. The loadeffect factors shown in Table B1 shall be used wherever thedesign load effect is referred to for all limit states and safetyclass.

B 300 Load model factors301 Load model factors γSd account for inaccuracies, ideali-sations, and biases in the engineering model used for represen-tation of the real response of the structure. Effects of geometrictolerances shall also be included in the load model factor. Thefactor is treated here as a deterministic parameter.302 Details about the load model factor are given in Sec.4 L.

C. ResistanceC 100 Resistance factors101 The following resistance factors apply:

— material resistance factor γm to account for material andresistance uncertainties

— a resistance model factor to account for possible inaccura-cies in the failure criteria used

— a system factor

102 The resistance factors applicable to ultimate limit states(ULS) are specified in Table C1 and C2. The factors are linkedto the safety class to account for the consequence of failure.Failure types are described in Sec.2 C200 and specified inDNV-OS-C501 Sec.6 A200 for all failure criteria.

103 The resistance factors applicable to accidental limitstates (ALS) are identical to the factors for ULS, specified inthe tables in 102.104 The resistance factors applicable to serviceability limit

Table A1 Typical limit states for the riser systemLimit State Category

Limit State or Failure Mode

Failure definition or Comments

SLS

Clearance No contact between e.g. riser-riser, riser-mooring line, riser-hull, surface tree- floater deck, subsea tree-seabed, surface jumper- floater deck.

Excessive angular response

Large angular deflections that are beyond the specified operational lim-its, e.g. inclination of flex joint or ball joint.

Excessive top displacement

Large relative top displacements between riser and floater that are beyond the specified operational lim-its for top tensioned risers, e.g. stroke of telescope joint, slick joint and ten-sioner, coiled tubing, surface equip-ment and drill floor. Note that systems can be designed for exceed-ing displacement limits if the struc-tural integrity is maintained.

Mechanical function

Mechanical function of a connector during make-up/break-out.

ULS

Bursting Membrane rupture of the pipe wall caused by internal overpressure, pos-sibly in combination with axial ten-sion or bending moments

Liquid tightness Leakage in the riser system including pipe and components, caused by internal overpressure, possibly in combination with axial tension or bending moments

Buckling Buckling of the pipe cross section and/or local buckling of the pipe wall due to the combined effect of external overpressure, effective tension and bending moment.

Propagating buckling

Propagating hoop buckling initiated by hoop buckling.

Damage due to wear and tear

Damage to the inside or possibly to the outside of the pipe during opera-tion or installation, resulting into burst or leakage.

Explosive decompression

Rapid expansion of fluid inside a material or interface leading to dam-age that may cause leakage or burst.

Chemical decomposition Corrosion

Chemical decomposition or corro-sion of materials with time that leads to a reduction and strength, resulting into burst or leakage.

ALS

Same as ULS and SLS

Failure caused by accidental loads directly, or by normal loads after accidental events (damage condi-tions).

Impact Damage introduced by dropped objects, like drill bits etc.

Fire Resistance to fire, if parts are above water

FLS Fatigue failure Excessive Miner fatigue damage or fatigue crack growth mainly due to environmental cyclic loading, directly or indirectly.

Table B1 Load effect factorsLimit state F-load

effectE-load effect

A-load effect

γF 1) γE 2) γAULS 1.1 1.3 NAFLS 1.0 1.0 NASLS and ALS 1.0 1.0 1.0NOTES

1) If the functional load effect reduces the combined load effects, γF shall be taken as 1/ 1.1.

2) If the environmental load effect reduces the combined load effects, γE shall be taken as 1/ 1.3.

Table C1 Brittle failure type

Safety classCOV of the strength

COV < 10% 10%-12.5% 12.5%-15%Low 1.22 1.33 1.49Normal 1.34 1.53 1.83High 1.47 1.75 2.29

Table C2 Plastic or ductile failure type


COV < 10% 10%-12.5% 12.5%-15%Low 1.11 1.16 1.23Normal 1.22 1.33 1.49High 1.34 1.53 1.83

DET NORSKE VERITAS


states (SLS) are specified in Table C3. The factors are linkedto the safety class to account for the consequence of failure.

Guidance note:For SLS, the set of resistance factors can be defined by theowner, see G.For ALS, the set of safety factors depends on the frequency ofoccurrence and is to be defined from case to case, see F. In cases,where the inherent uncertainty related to the accidental load isnegligible and, where a conservative estimate is applied, thematerial resistance factor in Table C1, Table C2 and Table C3can be reduced by 10%.


C 200 Geometrical parameters201 Nominal dimensions shall be used for all calculationsrelated to FRP laminates or polymers.202 For metals, the dimensions as described in the relatedmetal standards like DNV-OS-F201 for dynamic risers shall beused.

C 300 Material strength301 The characteristic material strength as described inDNV-OS-501 Sec.4 shall be used for all calculations.302 Both characteristic short term properties and character-istic long term properties up to the design life shall be consid-ered. How to obtain long term properties is described in DNV-OS-501 Sec.4.

Guidance note:If all long term properties are lower than short term properties,one analysis with long term properties is usually sufficient. How-ever, in some instances stresses may be distributed differently atthe beginning of the design life than at the end. In that case twoanalysis may be required.


303 If the strength of the material is temperature dependentor dependent on the surrounding environment within the rangeof operational conditions, the analysis should consider therange of strength using the same principles as given in 302.

C 400 Resistance model factors401 Resistance model factors γRd account for differencesbetween true and predicted resistance values given by the fail-ure criterion.402 Model factors shall be used for each failure criteria. Thefactors are given in DNV-OS-C501 Sec.6. A summary is given in Table C4.

C 500 System effect factor501 The safety factors are given for the entire system.Depending on how the components are connected to form asystem, the target probability of failure for individual compo-nents may need to be lower than the target probability of fail-ure of the entire system.502 In order to take this system effect into account, a systemeffect factor γS shall be introduced. If the system effect is notrelevant, γS = 1.0. Otherwise a system factor shall be docu-mented. A value of γS = 1.10 can be used as a first approach.

Guidance note:E.g. In the case of a riser string, the failure of one section (i.e.plain pipe or end connector) is equivalent to the failure of theentire system. This is a chain effect in which any component ofthe string can contribute. As a consequence, the target safety ofindividual section should be higher than the target safety of theentire system, in order to achieve the overall target safety.


Guidance note:A continuos spoolable riser has only two end connectors (one ateach end). Failure of an end connector is also a system failure.However, since there are only two connectors it is not a chaineffect and γS = 1.0 can be used.


503 In some cases a system may consist of parallel compo-nents that support each other and provide redundancy, even ifone component fails. In that case a system factor smaller than1 may be used if it can be based on a thorough structural relia-bility analysis.

D. Ultimate Limit State

D 100 General101 The riser pipe shall be designed against relevant modesof failure listed in Table A1.102 This section provides design checks with emphasis onload controlled conditions. Design principles for displacementcontrolled conditions are discussed in 900.103 Loading conditions for the limit state checks areobtained as described in Sec.4. Risers are typically evaluatedfor internal/ external pressure, axial tension, bending and allpossible combinations of these loads. Composite risers may bevery sensitive to small torsional loads and small axial compres-sive loads, depending on the laminate and end-fitting design.All loads should be considered in the limit state analysis.104 An example of the sequence of steps in evaluating a limitstate is shown in Fig.1. Each limit state can be related to vari-ous failure mechanisms. For each failure mechanism a relevantmathematical description, i.e. a failure criterion, has to befound.105 Failure criteria are given directly in this section or seerelevant failure criteria in DNV-OS-C501.

Table C3 SLS


COV < 10% 10 %-12.5% 12.5%-15%Normal 1.11 1.16 1.23High 1.22 1.33 1.49

Table C4 Model factorsFailure criterion Model factors

γRdReference in DNV-OS-C501

Fibre failure 1.0 or γA Sec.6 C202Matrix cracking 1.0-1.15 Sec.6 D100 to

D400De-lamination 1.0-2.0 Sec.6 EYielding 1.0 Sec.6 FUltimate failure of orthotropic homogenous materials

1.25 Sec.6 G

Buckling Same range as all other criteria

Sec.6 H

Displacements 1.0 Sec.6 IStress rupture 0.1-1.0 Sec.6 J400Fatigue 0.1-1.0 Sec.6 K300

DET NORSKE VERITAS


*) usually risers are designed that they do not experience compressive loads.

Figure 1 Example to illustrate flow from limit state to design criterion.

D 200 Bursting 201 Bursting of the pipe may be caused by internal overpres-sure, possibly in combination with axial tension or bendingmoments.202 The global analysis of the riser shall provide the worstcombination of the above loads for local analysis. The localanalysis shall establish load effects (stresses or strains) on theply level.203 The analysis shall check all failure mechanisms listed inDNV-OS-C501 Sec.6 A501.204 Fibre failure shall always be analysed (DNV-OS-C501Sec.6 C). Fibre failure defined here as ply failure in the fibredirection, as described in DNV-OS-501 Sec.4. Fibre failure isnot acceptable.205 Matrix cracking of the laminate may be acceptable aslong as a fluid barrier remains intact. Typically matrix crack-ing is acceptable for riser pipes with a liner. See Sec.6 D forrequirements for the liner.206 If the riser does not have a liner or fluid barrier, fluidtightness in the presence of matrix cracks should be docu-mented. Usually a laminate leaks only after a certain numberor density of matrix cracks has developed. It is recommendedto determine the point of leakage experimentally by compo-nent testing (DNV-OS-C501 Sec.10).207 Matrix cracking may reduce the compressive strengthunder some conditions (DNV-OS-C501 Sec.6 C400). The pos-sible consequences of such a strength reduction should be con-sidered.208 Fluid tightness can also be documented by showing thatthe matrix does not crack (DNV-OS-C501 Sec.6 D).209 Delaminations in the laminate may be acceptable ifthrough thickness stresses must not be carried by the laminate.However, delaminations may reduce the buckling strength. 210 Yielding is not a failure mode for most fibre reinforcedlaminates. If yielding can happen two options may be used.The design does not allow yielding (DNV-OS-C501 Sec.6 F).Alternatively, a fully non-linear analysis may be done consid-ering the effects of yielding. See Sec.6 D for requirements forthe liner.211 Buckling may happen locally under bending of the riser.Buckling is not acceptable. Requirements for buckling aregiven in D400.212 Large displacements or deformations due to high loadsdo usually not cause burst. Large deformations may weakencomposite metal interfaces. 213 Materials shall be chosen in a way that they do not

decompose chemically over the lifetime. Such decompositionwould weaken the material and may cause burst (DNV-OS-C501 Sec.6 Q).

D 300 Liquid tightness - leakage301 Leakage of a riser is similar to burst, but a more gradualprocess. All considerations for burst also apply for leakage.302 Diffusion or permeability of the fluid shall be lowenough that no or minimal amounts of fluid get out of the sys-tem.303 If the riser has a liner it is sufficient to show that the lineritself can contain the fluid. Properties of the laminate shall bechecked if there is no liner or if the liner is not fluid tight.

D 400 Buckling 401 Buckling of the riser tube shall be considered as a possi-ble failure mechanism.402 Relevant load conditions that may induce buckling ofthe riser tube are:

— axial compression— bending of the riser tube as a beam (i.e. such that the axis

of the riser tube bends)— torsion of the tube about its own axis— external overpressure.

It may be relevant to consider the simultaneous presence ofaxial tension along with bending, torsion or external pressure,or of internal overpressure along with axial tension/compres-sion, bending or torsion.403 Buckling shall be evaluated as described in DNV-OS-C501 Sec.6 H taking due account of geometric imperfections.404 For liner buckling, see Sec.6 D400.405 If analytical formulae are used for estimating criticalbuckling loads, due account shall be taken of the an-isotropicproperties of the riser wall. 406 to 410 provide some formulaethat may be used to estimate the elastic critical loads for theload cases listed in 402. These should be used with knock-down factors to give the buckling strength allowing for geo-metric imperfections, as indicated. The criterion to be checkedis given in 411. Combined loads may be considered in accord-ance with 412. 406 A negative effective tension may cause a riser to bucklein compression. Buckling may take the form of global beam-column buckling or local buckling of the riser wall, or a com-bination of the two. For axial compression the critical values ofthe mean axial compressive stress for elastic buckling in theglobal and local modes, σcr global and σcr local, and the bucklingresistance, buckling, may be derived as follows:

In the above, L is the effective length of the riser tube for globalbuckling as a beam-column, R and t are the radius and thick-ness of the riser tube, the suffices x and θ refer to the axial andcircumferential directions, -E- and -ν- are modulus and Pois-son’s ratio, and K1 is the an-isotropy factor given by:

Limit statescategory

Limit state orfailure mode

Failuremechanism

DesigncriterionDNV-OS-C501

Burst

Buckling

Fibre failure

Matrix cracking (if critical)

Global buckiling

Local buckling

ULS

Maximum fibrestrain criterion

Matrix crackingcriterion

Global criterion(if relevant*)Finite elementanalysis

σ̂

localbuck globalbuck buckling ˆ1

ˆ1

ˆ1

σσσ+=

2ˆ xx

2globalA

2

2

globalcr globalA globalbuck

EkLRk

πσσ ==

[ ] 2/1θxxθ

1xxlocalA localcr localA localbuck )1(3

ˆνν

σσ−

==KEk

Rtk

DET NORSKE VERITAS


Exx, Eθθ and Gxθ are the laminate elastic engineering modulifor in-plane deformations. Note that the engineering constantsare only defined for symmetric laminates.The knock-down factors to account for geometric imperfec-tions, kA global and kA local , should be taken as 0.67 and 0.5respectively, unless higher values can be demonstrated.Global buckling under conditions of displacement-controlledloading may be permitted, provided it does not result in otherfailure modes such as local buckling, unacceptable displace-ments, or unacceptable cyclic effects. In such cases only thelocal buckling mode need be considered in the above formulae.The above formulae apply to the true wall compression for thecase when the external and internal pressures on the tube wallare equal. Cases of simultaneous axial compression andexternal overpressure shall be treated in accordance with 412.

Guidance note:It is essential that an appropriate tensioned-beam model is usedfor the analysis of global buckling. The consequence of a too-small positive effective tension is excessive curvature and bend-ing moment near the location of minimum effective tension.Note that members above the tension joint for top tensioned ris-ers may be subjected to compressive forces for some riser types.


407 For bending of the riser tube the critical bendingmoment for elastic buckling - Mcr, -, and the buckling resist-ance moment - buckling - may be derived from:

in which R and t are the radius and thickness of the riser tube,the suffices x and θ refer to the axial and circumferential direc-tions, E and ν are modulus and Poisson’s ratio, and K1 is ananisotropy factor as defined in 406. The knock-down factor kMto account for geometric imperfections should be taken as 0.5unless a higher value can be demonstrated. 408 Special care shall be given when a small decrease in toptension of a top-tensioned riser could cause excessive bendingmoment. In that case, the designer shall establish a minimumeffective tension that gives a margin above the tension that ispredicted to cause excessive bending moments.409 For the case of torsional loading about the riser’s longi-tudinal axis, the critical torsional moment for elastic buckling,MTcr, and the buckling resistance torsional moment, T buck-ling, may be estimated from:

in which R, L and t are the radius, length and thickness of theriser tube, Axx, Aθθ and Axθ are the laminate elastic constantsfor in-plane deformations, Dxx, Dθθ and Dxθ are the laminateelastic constants for bending deformation, and the suffices xand θ refer to the axial and circumferential directions, respec-tively. The knock-down factor kT to account for geometricimperfections should be taken as 0.67 unless a higher value canbe demonstrated. This formula is valid only when the couplingcoefficient Bθθ is small or zero (as in the case of a symmetric

laminate lay-up), and when:

410 For the case of external pressure loading the criticalpressure for elastic buckling, pcr, and the buckling resistancepressure, buckling, may be estimated from:

in which the notation used is as defined in 409. The knock-down factor kp to account for geometric imperfections shouldbe taken as 0.75 unless a higher value can be demonstrated.The above formula applies for long tubes. For shorter lengthsof tube the following formula should be used if this gives ahigher value:

in which L is the length of the tube. This formula assumes asymmetric lay-up and is valid only when:

Guidance note:pcr is the local minimum internal pressure taken as the most unfa-vourable internal pressure plus static head of the internal fluid.For installation pmin equals zero. For installation with water-filled pipe, pmin equals pe.


411 The failure criterion for buckling when the resistance isdetermined by use of the above formulae is as follows:

F = characteristic value of the induced stress orstress resultant (σ, M, T or p)

= characteristic value of the resistance obtainedfrom the tests

γF = partial load or load effect factorγSd = partial load or load effect model factor γMbuckle = partial resistance factorγRdbuckle = partial resistance-model factor.The partial resistance factor γMbuckle and γRdbuckle may betaken as 1.0, if the knock-down factors given in the above sec-tions are adopted.The load effect model factor γSd shall take account of the accu-racy of representation of geometrical imperfections andboundary conditions. The value shall be determined fromSec.4 L.412 For cases of combined loadings a conservative assess-ment may be performed by assuming a linear interaction rela-tionship:

σ, M, T and p are the characteristic values of the axial com-pressive stress, the bending moment, the torsional moment andthe external pressure when considered in combination.

Guidance note:Symbols of the formulas in this section are explained after theequation the first time they occur. The components of the A, B,

2/1

xx

xθ2/1

xx

θθ2/1

xx

θθxθ1 12

⎪⎭

⎪⎬⎫

⎪⎩

⎪⎨⎧

⎟⎟⎠

⎞⎜⎜⎝

⎛

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛+=

EG

EE

EE

K ν

M̂

[ ] 2/1θxxθ

1xx2

McrMbuckling )1(3

3.1ˆνν

π−

==KERtkMkM

M̂

( ) 8/3

θθθθ

22xθθθx

4/32/1

4/5

θθTcrTTbuckling T 7.21ˆ⎥⎥⎦

⎤

⎢⎢⎣

⎡ −==

DAtAAA

tLRDkMkM x

( )⎟⎟⎠

⎞⎜⎜⎝

⎛

⎥⎥⎦

⎤

⎢⎢⎣

⎡ −⎟⎟⎠

⎞⎜⎜⎝

⎛≤

tRDAtAAA

DD

Lx 500

121

2/1

xxθθ

22xθθθx

6/5

xx

θθ2

p̂

3θθ

2θθ

θθp

crpbuckling

3ˆ

R

AB

Dkpkp

⎟⎟⎠

⎞⎜⎜⎝

⎛−

==

( ) 4/1

θθθθ

22xθθθx

2/12/3θθp

crpbuckling

5.5ˆ

⎥⎥⎦

⎤

⎢⎢⎣

⎡ −==

DAtAAA

tRLDk

pkp x

( )500

12

22/1

xxθθ

22xθθθx

2/3

xx

θθ ≥⎟⎟⎠

⎞⎜⎜⎝

⎛

⎥⎥⎦

⎤

⎢⎢⎣

⎡ −⎟⎟⎠

⎞⎜⎜⎝

⎛tR

LDA

tAAADD x

RdbuckleMbuckle

bucklingSdF

FFγγ

γγ.

..∧

≤

bucklingF∧

RdbuckleMbuckleSdFpp

MM

MM

FF

γγγγσσ

...1

ˆˆˆˆˆbucklingbuckling T

T

bucklingbucklingbuckling

≤+++=

DET NORSKE VERITAS


D stiffness matrix of the composites can be obtained from stand-ard laminate theory calculations, explained in most textbooks.


D 500 Propagating buckling501 Propagation buckling is not considered for compositerisers, since local buckling is already not acceptable.

D 600 Wear and tear601 The riser pipe shall have sufficient resistance to wearand tear from the fluids or equipment running through the pipe.(See also DNV-OS-C501 Sec.6 M).602 Composite riser have often an internal liner. In this casewear resistance of the liner shall be demonstrated.603 If the composite riser does not have a liner wear resist-ance of the laminate shall be demonstrated.604 The riser shall not leak and the laminate shall keep itsload carrying capacity after degradation due to wear and tear.605 Drilling risers should have a smooth inner bore to pre-vent tools from catching the inner liner.

D 700 Explosive decompression701 Explosive decompression may happen if fluid isentrapped under pressure within the material or the interface. A sudden reduction of pressure in the system may cause thefluid to expand and cause severe damage.702 If the fluid can diffuse into any riser materials explosivedecompression shall be considered (DNV-OS-C501 Sec.6 O).703 Fluids in risers with a tight metal inner liner cannot dif-fuse through the metal and explosive decompression due to theinternal fluids does not have to be considered.

D 800 Chemical decomposition - corrosion801 Materials shall be chosen which do not decomposechemically in the design environment within the lifetime of theriser (DNV-OS-C501 Sec.6 Q).802 Glass, aramid and carbon reinforced laminates made ofpolyester, vinyl or epoxy have not shown chemical decompo-sition in marine environments. However, they do age asdescribed in DNV-OS-C501 Sec.4 C through E.803 Possible corrosion of the constituent materials should beconsidered. Carbon fibres in contact with metals may causegalvanic corrosion.

D 900 Displacement controlled conditions901 Displacement controlled conditions can be addressed asdescribed in the standard for dynamic risers DNV-OS-F201.902 Accumulated yielding is usually not relevant for com-posite laminates, but it should be considered for all other com-ponents, like metal or thermoplastic liners etc.

E. Fatigue Limit StateE 100 General101 Cyclic fatigue and static fatigue (or stress rupture)should be considered for composite risers.102 All failure modes that were evaluated in static analysisshall also be evaluated for possible fatigue failures. A fewexceptions are given in the failure criteria in DNV-OS-C501Sec.6 J and Sec.6 K.103 The effects of possible creep or stress relaxation on theriser system should be evaluated.104 All critical sites of each unique component along theriser shall be evaluated. These sites normally include details

that causes stress concentrations or load introduction points.105 Possible changes in stiffness with time shall be consid-ered for cyclic and static fatigue (see DNV-OS-C501 Sec.6J200 and K200).

E 200 Cyclic fatigue201 The riser system shall have adequate safety againstcyclic fatigue within the service life of the system. See DNV-OS-F201 Sec.4 and Appendix B for more details with respectto fatigue design and analysis. The safety factors are given inthis document in Section E 400.202 All cyclic loading imposed during the entire service life,which have magnitude and corresponding number of cycleslarge enough to cause fatigue damage effects, shall be takeninto account. Temporary phases like transportation, towing,installation, running and hang-off shall be considered. Loadsshall be described as stated in Sec.3 F300.203 Normally, the methods based on characteristic S-Ncurves and reduction of strength with time are used duringdesign for fatigue life assessment (DNV-OS-C501 Sec.6 K).S-N curves should be obtained as described in DNV-OS-501Sec.4 C.204 If representative fatigue resistance data are not availa-ble, direct fatigue testing of the actual components shall be per-formed with due regard of the chemical composition of theinternal and external environment (DNV-OS-C501 Sec.10).205 The stress to be considered for fatigue in a riser is thecyclic (i.e., time-dependent) stress. Mean value and amplitudeshould be obtained.206 This combined stress varies around the circumference ofthe riser pipe. For cases where the waves are incident from sev-eral different directions, the fatigue damage must hence be cal-culated at a number of regularly spaced points to identify themost critical location.

E 300 Stress rupture301 The riser system shall have adequate safety againststress rupture (or static fatigue) within the service life of thesystem.302 All long term permanent loads during the entire servicelife, which have magnitude large enough to cause stress rup-ture effects, shall be taken into account. Temporary phases liketransportation, towing, installation, reeling, running and hang-off shall be considered. Loads shall be described as stated inSec.3 F200.303 Normally, the methods based on long term stress rupturecurves and reduction of strength with time are used duringdesign for fatigue life assessment (DNV-OS-C501 Sec.6 K).Material data should be obtained as described in DNV-OS-501Sec.4 C.304 If representative stress rupture resistance data are notavailable, direct testing of the actual components shall be per-formed with due regard of the chemical composition of theinternal and external environment (DNV-OS-C501 Sec.10).305 The stress to be considered for stress rupture in a riser isthe long term static stress. 306 This stress may vary around the circumference of theriser pipe. For cases where the waves are incident from severaldifferent directions, the stress rupture analysis must hence becalculated at a number of regularly spaced points to identifythe most critical location.

E 400 Factors for static and dynamic fatigue analysis401 The factors γfat in Table E1 shall be used for the predic-tion of failure due to cyclic fatigue or due to long term staticloads. The factors shall be used with the failure criteria in

DET NORSKE VERITAS


DNV-OS-C501 Sec. 6 J and K (Miner Sum).

402 When using the factors in Table E1, no in-serviceinspection is required. See also Section 9.403 In-service inspection may be used if it can be combinedwith damage growth and failure predictions, to optimise afatigue analysis.

F. Accidental Limit State

F 100 General101 The same considerations apply to composite risers asgiven in the DNV-OS-F201.

F 200 Resistance against fire201 If the riser is used above the water line resistance againstfire shall be demonstrated. A full scale fire test with a repre-sentative fire is the recommended option to demonstrate fireperformance.202 It is recommended to use composite risers made of flam-mable materials only below the water line. Metal riser sectionscan be used above the water line. In that case it is not necessaryto demonstrate fire resistance of the composite riser.

F 300 Resistance against dropped objects - impact301 Demonstration of resistance against dropped objects, asdrill-bits may be required. The impact scenario should bedefined by the local authorities or the user.302 It shall be shown that the riser remains leak tight after animpact of the defined scenario.303 When considering the effects of impact, it should bedocumented that no unintended failure mechanisms will hap-pen due to impact.304 The resistance of the riser to impact should be testedexperimentally. This can be done in two ways.

— the material or a small section is exposed to a relevantimpact scenario. The strength of the material with theimpact damage should be determined. This strength can beused for further design of the riser

— the full riser is exposed to a relevant impact scenario. Theriser is tested afterwards to show that it can still tolerate thecritical loads.

305 Impact failure criteria may be used if experimental evi-dence shows that they are applicable for the application.

F 400 Impact testing401 If the testing option is chosen in 303, the riser should betested lying on the ground under normal operating pressure andaxial tension. A higher pressure than operating pressure maybe used to simulate the axial tension.402 The riser pipe should be impacted in the middle andclose to the joint.403 It should be evaluated whether the riser should be able towithstand more than one impact scenario. In that case the riser

should be exposed to the expected number of impact events.

F 500 Evaluation after impact testing501 The impact tests should demonstrate that no unaccepta-ble damage is introduced into the riser. Once the riser has beenexposed to impact it should be carefully inspected to ensurethat no unexpected failure mechanisms occurred that mayreduce the riser's performance, in particular long term perform-ance. If the riser will be taken out of service after an impact,long term considerations do not have to be made.502 It shall be shown further that the riser can carry all rele-vant loads after impact until it can be taken out of service forrepair or replacement. This can be done by demonstrating thatthe riser remains pressure tight for the time it takes to take theriser out of service. The riser should remain pressure tight forat least 1 hour.503 If the riser may be exposed to impact but can or shouldnot be repaired afterwards, it should be shown that the riser canwithstand all long-term loads with the damage induced by theimpact. This can be done by analysis taking the observedimpact damage into account, by testing, or a combination ofanalysis and testing. Testing should be done according toDNV-OS-C501 Sec.10.

G. Serviceability Limit State

G 100 General101 The same considerations apply to composite risers asgiven in the DNV-OS-F201.

H. Special Considerations

H 100 Interference101 The same considerations apply to composite risers asgiven in the DNV-OS-F201.

H 200 Unstable Fracture and Gross Plastic Deformation201 Metal parts should fulfil the same requirements as givenin DNV-OS-F201 Sec.5 H200.202 Composite laminates should have a certain minimumfracture toughness to prevent unstable fracture growth.203 Laminates with all of the properties listed below havesufficient fracture toughness:

— Laminates that have somewhere layers with fibre direc-tions that are at least 30 degrees apart.

— The thickness of layers with fibres in one direction is lessthan 0.6 mm.

Layers of interwoven fibres with fibres in two directions withare at least 30 degrees apart may have any thickness.204 If 203 does not apply the notch sensitivity of a laminateshall be at least 0.8 in all directions. Notch sensitivity isdefined here as the strength of a laminate with notches dividedby the ultimate strength. The ratio of crack length to specimenwidth shall be at least 0.375. The thickness of the notch shallbe not more than 0.5 mm. It is recommended to test notch sen-sitivity with double edge notched specimens, where notchesare cut into the specimen on each side.

Table E1 Factor for fatigue calculations γfatSafety class

Low Normal High15 30 50

DET NORSKE VERITAS


SECTION 6CONNECTORS AND LINERS

A. GeneralA 100 Objective101 This section gives requirements to connectors, compo-nents and liners. Metal parts shall fulfil all requirements givenin DNV-OS-F201 Dynamic risers, in particular Sec.6 of thatstandard. This section covers composite and plastic compo-nents and gives additional requirements for metal components.102 The aim of the design is to ensure that the riser with itsconnectors, liner and riser components has adequate structuralresistance, leak tightness and fatigue resistance for all relevantload cases. Resistance against accidental loads such as fire andimpact shall also be considered when applicable.103 Risers and connectors shall achieve the same or higherlevel of reliability as the structure of which they are part.104 All connectors, components and interfaces shall be eval-uated against the same limit states as described for the riserpipe in Sec.5.105 Connectors and components made of fibre reinforcedplastics can in principle be analysed and tested the same wayas a structure or component. However, some special consider-ations are described in the following sections.

A 200 Definition of joint201 The term "joint" is used in this section as a connectionbetween two parts, like a mechanical joint or and adhesivejoint. It is not used as a riser joint, describing a section of riserpipe with two end connectors.

B. Connector DesignsB 100 Functional requirements101 Riser connectors shall allow for multiple makeup andbreakout in a reliable manner. The connector may permit forinterchangeability between connector halves to allow riserjoints to be run in any sequence.102 The basic requirements given for the performance ofconnectors in DNV-OS-F201 apply also for composite risers.

B 200 Design and qualification considerations201 It is recommended to make the connectors of metallicmaterials. If the connector is made of composite great careshould be taken to analyse the complicated stress states andstress concentrations. Through thickness stresses away fromthe fibre directions tend to be critical and as a consequencethrough thickness material properties. Extensive testing ismost likely needed to qualify a composite connector. 202 If the connector is made of metal all requirements givenin DNV-OS-F201 apply. Special consideration should begiven for the composite metal interface between the connectorand the riser pipe, see under C.

C. Composite - Metal Connector InterfaceC 100 General101 The interface between the metal connector and the com-posite pipe is a critical part of the riser design. The interface isbasically a joint and all general requirements given under Fshould be considered.

102 The fibres run usually in a complicated 3-D pattern at thejoint. The analysis should model the fibre arrangement prop-erly. In an FE-analysis, the elements should follow the direc-tion of the fibres.103 The composite metal connector interface shall be strongenough to transfer all loads considered for the connector andthe pipe section.104 If the riser has a liner, the liner is usually also connectedto the metal connector in some way. This connection shall notinhibit the functions of the liner in any way and shall be as reli-able as the liner itself.

Guidance note:The composite metal connector interface is typically a mechani-cal joint. Adhesive joints may be used, but it is difficult to dem-onstrate reliable long-term performance of the adhesive joint.


105 The performance of the composite metal interface shallbe verified by testing. Minimum requirements are given in G.

C 200 Limit states201 The composite metal connector interface shall be at leastanalysed for the same limit states as the riser pipe.202 A local analysis should be carried out based on loads andboundary conditions from the global analysis of the riser sys-tem.203 A careful analysis of all possible failure modes shall bemade. It shall be shown for all failure modes that they eitherwill not occur or are not critical for the performance of the risersystem.204 The possible mismatch of thermal properties of materi-als shall be considered in the analysis.205 Internal or external pressure on the riser system may bebeneficial or detrimental to the performance of the joint. Thiseffect shall be considered in the analysis.206 Creep of any of the materials used in the joint mayreduce friction, open up potential paths for leakage or lead tocracks. Effects of creep shall be carefully considered.

Guidance note:It is highly recommended to design the joint in a way that it alsofunctions if the matrix of the composite laminate is completelydegraded. In that case the joint can perform as long as the fibresare intact and sufficient friction between fibres and the fibremetal interface exists. Such a joint does not rely on the usuallyuncertain long-term properties of the matrix.


207 Metal parts should be designed in a way that they do notyield to ensure no changes in the geometric arrangement of thejoint. If any yielding can occur a non-linear analysis shall bedone taking all relevant load histories and accumulated plasticdeformations into account.208 Possible effects of corrosion on metals and interfacesshall be evaluated.209 Possible galvanic corrosion between different materialsshall be considered. An insulating layer between the differentmaterials can often provide good protection against galvaniccorrosion.210 Leak tightness of the joint shall be carefully evaluated.In particular possible flow along interfaces should be analysed.

DET NORSKE VERITAS


D. Inner Liner

D 100 General101 Most composite risers have an inner liner as a fluid bar-rier. This inner liner is typically made of metal or polymericmaterials.102 It shall be shown that the inner liner remains fluid tightthroughout the design life, if it is used as a fluid barrier.

D 200 Mechanical performance201 The inner liner may contribute to the overall stiffnessand strength of the riser system depending on its stiffness andthickness.202 The inner liner usually follows the deformations of themain load bearing laminate. It shall be shown that the innerliner has sufficiently high strains to failure and yield strains tofollow all movements of the riser system.203 Mismatch in thermal properties between inner liner andlaminate should be considered. The mismatch may introducehigh stresses or strains.204 The inner liner should be operated in its linear range.Neither operational conditions nor test conditions should bringit to yield. An exception is the first pressure loading D300.205 In addition to the requirements given here, metal innerliners and their welds shall be evaluated according to DNV-OS-F201 fatigue life, and capability to follow system deforma-tions. If the metal liner is load bearing static strength shall alsobe evaluated according to DNV-OS-F201.206 Polymeric inner liners, like thermoplastic inner linersmay be evaluated against the yield criterion in DNV-OS-C501Sec.6 F.

D 300 Autofretage301 It is common practice to pressurise the riser pipe initiallyat the factory to such a high pressure that the inner liner yields.After removing the pressure the inner liner will be compressedby the outer laminate. This procedure ensures a tight fitbetween inner liner and laminate.302 The yielding of the inner liner also causes the welds toyield. This may reduce stress concentrations, but it can alsocause local thinning around the weld. Any thickness variationsin the inner liner may cause localised yielding. The weld zonemay have lower yield strength than the main part of the innerliner. Due to this the inner liner may yield locally close to thewelds. The strain in the localised yield region can be very high,possibly leading to instant rupture, lower fatigue performance,enhanced creep. The inner liner and its welds shall be analysedtaking all these effects into account.

Guidance note:A small thin area in the inner liner can be worse than a larger thinarea, because the inner liner may only deform by yielding in thethin section. In that case the small thin section will have muchhigher strains than the large section, if the total deformation is thesame.


303 If the inner liner material can creep, than creep will hap-pen especially in the thin highly strained regions. The effect ofcreep with respect to fatigue, stress rupture and bucklingshould be evaluated.304 If the inner liner is under compression, local yieldingmay create deformations resulting in local or global buckling.305 Inner liner specifications with respect to acceptablethickness variations, weld quality, and maximum misalign-ments should be consistent with the worst cases evaluated inthe analysis.

D 400 Liner buckling401 Buckling of the liner due to hoop compression shall beconsidered as a potential failure mechanism. The followingfour slightly different phenomena should be considered:

— buckling due to internal under pressure, i.e., vacuum,without external pressure should always be evaluated

— buckling of the liner due to external pressure as a conse-quence of compression of the main laminate due toexternal pressure. This effect should be always evaluated

— buckling of the liner due to external water pressure. Thisis only relevant if the pressure of the outside water canreach the outer surface of the inner liner. The laminates areusually not pressure tight, but the presence of an outerliner can make it pressure tight

— explosive decompression causes a pressure to build upsuddenly between the liner and the composite riser tube, atthe same time as the pressure inside the liner suddenlydrops. This effect can happen if gas or liquid can diffusethrough the inner liner and accumulate in the interfacebetween liner and laminate or inside the laminate. Thiseffect can be ignored for metal liners, since they are diffu-sion tight, provided no other diffusion path through sealsetc. exists in the system

— as a result of the sustained internal pressure, the lineryields plastically (or undergoes creep deformation) in ten-sion in the hoop direction. Decompression causes the com-posite riser tube to contract, compressing the liner andcausing it to buckle. This effect can be prevented by usinginitially the autofretage process (D300) and by keeping theliner below yield during operation.

402 Possible buckling of the liner as a result of other loadingconditions on the riser shall also be evaluated. In particularbuckling of the liner associated with bending of the riser tubeas a beam should be considered.403 Buckling may be evaluated by treating the inner liner asan independent tube. This is a very conservative approach,because the support of the laminate outside the liner is not con-sidered. It may be a convenient approach to document that theliner can withstand an internal vacuum.404 When considering either of the effects described in 401the tightness of the fit between the liner and the riser tube shallbe taken into account. A relevant parameter is the liner fitparameter:

Ro (liner) and Ri (riser tube), respectively are the outer radius ofthe liner and the inner radius of the riser tube in conditionswhere the two components are considered separately withoutmechanical loading and at the temperature prevailing at thetime of decompression. The case η > 0 is that of a tight fit such that the liner is in a stateof circumferential (hoop) compression when fitted in the risertube. If the liner has undergone plastic or creep deformation asa result of the prior internal pressure the effective value of ηmay be different from that at initial assembly of the riser. Thefit parameter is also dependent on the temperature prevailing atthe time of the decompression event.405 The following aspects may also have a fundamentalinfluence and shall be considered when evaluating liner buck-ling:

— the extent to which the liner is and remains bonded to theriser tube (see also 408). Its recommended to treat the lineras un-bonded, because it is difficult to demonstrate bond-ing over the lifetime of the riser, unless liner and laminateare made of the same material (407)

— initial geometric out-of roundness or other unevenness inthe liner (or in the inner surface of the riser tube). In par-

liner)(o

)tube(riseri(liner)o

RRR −

=η

DET NORSKE VERITAS


ticular, a longitudinal seam in a liner may have a largeeffect on the buckling resistance

— plastic as well as elastic deformation in the liner duringbuckling

— the temperature at which the failure event is considered tooccur. As well as influencing the liner fit parameter η, thetemperature in the liner material may significantly affectboth its elastic properties and its plastic yield properties

— the mismatch of thermal properties between liner and lam-inate

— possible degradation of the liner properties prior to theevent, e.g. due to hydrolysis.

406 The buckling phenomena shall be evaluated by means ofnon-linear analysis and/or testing in accordance with Sec.4 K.In the case of non-linear analysis, an appropriate finite elementmethod shall be applied using contact elements at the liner/tube interface.407 If the only difference between liner and structural lami-nate is that the one has reinforcement and the other does not,i.e., they were made of the same matrix material, there is basi-cally no interface between liner and matrix. In this case theliner and laminate can be treated as bonded provided:

— inner liner and laminate were not connected by a second-ary bond

— through thickness testing (at least 3 tests) of the laminatewith liner shows that the combined material does not failat the connection between liner and laminate for more than30% of the tests

— the combined material has the same strength as thethrough thickness strength of the laminate within 10%.

408 Special attention shall be paid to the possibility of a fluidpressure occurring between the liner and riser tube in a regionwith de-bonding between the two components, and a subse-quent rapid growth of de-bonding (Fig. 1).This is essentially aproblem involving bending of the liner combined with growthof a crack at the interface and is not a buckling problem assuch.

Figure 1 Possible fluid pressure build-up in the interface between liner andriser tube causing rapid growth of initial de-bonding

D 500 Liner composite interface501 The liner may be bonded to the composite laminate or itmay be un-bonded. 502 A different layer of material may also be placed betweenthe laminate and the liner.503 All possible failure modes of the interface and their con-sequence to the performance of the system shall be evaluated.504 If a bond is required between laminate and liner, forexample to obtain good buckling resistance of the liner, theperformance of the bond shall be tested G500.

505 If interfaces only touch each other friction and wearshould be considered (DNV-OS-C501 Sec.6 M).506 Fluids may accumulate between interfaces. They mayaccumulate in voids or de-bonded areas and or break the bondof the interface. The effect of such fluids should be analysed.Possible effects of rapid decompression of gases should beconsidered.507 If the laminate may have matrix cracks, but the linershall not crack (or vice versa), it shall be shown that crackscannot propagate from one substrate across the interface intothe other substrate. Possible de-bonding of the interface due tothe high stresses at the crack tip should also be considered.508 It is recommended to demonstrate by experiments thatcracks cannot propagate across the interface from one sub-strate to the other. It should be shown that by stretching orbending both substrates and their interface that no cracks formin the one substrate even if the other substrate has the maxi-mum expected crack density.

Guidance note:A weak bond between the substrates is beneficial to preventcrack growth across the interface. However, it means thatdebonding may happen easily.


509 Local deformation of the liner may create high stressesor strains in the composite laminate next to the liner. Thiseffect shall be evaluated if local deformations of the liner canbe present.

D 600 Liner to end connector interface601 All requirements of D apply.602 The interface will be exposed to pressure, axial, andbending loads.603 Fluid tightness of the liner connector interface shall becarefully insured.604 Special attention shall be given to welds or seals, ensur-ing that the different movements of end connector and liner canbe followed.605 Aspects related to joints in general, under F, should alsobe considered.

D 700 Wear and tear 701 It should be demonstrated that the liner material couldwithstand the operating conditions with respect to wear andtear.702 Wear and tear should never destroy the fluid tightness ofthe liner.703 It may be good practice to add extra thickness to the lineror to provide a separate layer for wear and tear protection.704 The inner liner should be strong enough to withstandpossible shear, scraping and torsional loads from equipmentrunning inside riser. This is particularly important for drillingrisers.

E. Outer LinerE 100 General101 An outer liner is usually applied for keeping out externalfluids, for protection from rough handling and the outer envi-ronment and for impact protection.102 If no outer liner is applied the outer layers of the lami-nate have to take the functions of the outer liner.103 The outer liner material shall be resistant to the externalenvironment, e.g., seawater temperature, UV etc.

Region of initial de-bonding

Liner

Riser tube

DET NORSKE VERITAS


104 If the outer liner is exposed to UV radiation in service orduring storage, it should be UV resistant.

E 200 Mechanical performance201 Outer liners are not exposed to autofretage. They shouldbe kept below yielding.202 Resistance of the outer liner to handling and the externalenvironment shall be considered. The outer liner may get somedamage from handling, but the structural layer underneathshould not be effected.203 If the outer liner shall ensure that no water comes intothe laminate and the inner liner does not have to be analysedfor collapse under external water pressure the outer linershould meet the following requirements:

— the outer liner must be watertight— it should be demonstrated that no path exist for the water

to flow into the laminate. The seals at the end of the outerliner, usually against the end connector should be carefullyevaluated for long and short term performance

— water tightness should be demonstrated even when someexternal damage from handling is present

— it is recommended to apply an extra layer for protectionagainst handling.

204 The performance requirements to the outer liner shouldnot be effected by a possible impact scenario.

E 300 Blow out of outer liner301 If fluids can diffuse through the inner liner into the loadbearing laminate the outer liner may suffer from blow out if theexternal pressure is lower than the pressure inside the laminate.302 Blow out can be prevented by a venting mechanism.303 Blow out will also not happen if it can be shown that thefluids will diffuse from the laminate through the outer linerinto the external environment more rapidly than from theinside of the tube through the inner liner into the laminate. Inaddition, the remaining fluid concentration should be lowenough that even under low external pressure the outer linercannot blow out.

F. Joints of Materials or Components - general aspects

F 100 Analysis and testing101 The same design rules as applied for the rest of the struc-ture shall be applied to joints, as relevant.102 Joints are usually difficult to evaluate, because they havecomplicated stress fields and the material properties at theinterfaces are difficult to determine.103 Joints may be designed according to three differentapproaches:

— an analytical approach, i.e. the stress/ strain levels at allrelevant parts of the joint including the interface are deter-mined by means of a stress analysis (e.g. a FEM-analysis)and compared with the relevant data on the mechanicalstrength

— design by qualification testing only, i.e. full scale or scaleddown samples of the joint are tested under relevant condi-tions such that the characteristic strength of the completejoint can be determined

— a combination of an analytical approach and testing, i.e.the same approach specified in DNV-OS-C501 Sec.10 Cfor updating in combination with full scale componenttesting.

104 The options marked in Table F1 may be used for the dif-

ferent types of joints:

105 The level of all stress (strain) components in all relevantareas of the joint, including stress concentrations, shall bedetermined according to the same procedures as specified forthe rest of the structure. Special emphasis shall be put on pos-sible stress concentrations in the joint. It shall be recognisedthat the stress concentrations in the real structure may be dif-ferent than determined through the analyses due to e.g. simpli-fications made, effects of FEM-meshing etc.106 An analytical analysis is sufficient, if the stress field canbe determined with sufficient accuracy, i.e., all stress concen-trations are well characterised and a load model factor γSd canbe clearly defined. In all other cases experimental testingaccording to DNV-OS-C501 Sec.10 shall be carried out toconfirm the analysis.107 If the material properties, especially of the interface can-not be determined with sufficient accuracy, experimental test-ing according to DNV-OS-C501 Sec.10 shall be carried out.Scaled testing may be possible, as described in F200.108 Long term performance of a joint may be determinedbased on long-term materials data, if a clear link between thematerial properties and joint performance can be established.The requirements of 102 and 103 also apply for long term per-formance.109 The load cases should be analysed with great care forjoints. Relatively small loads in unfavourable directions can dogreat harm to a jointed connection. Especially loads due tounintended handling, like bending, stepping on a joint etc.should not be forgotten.110 Joints may be analysed by testing alone as described inDNV-OS-C501 Sec.10 B.111 The most practical approach is likely to use a combina-tion of analysis and testing. Since a large conservative biasmay be necessary in the analysis to account for the manyuncertainties in a joint design it is recommended to use theupdating procedures of DNV-OS-C501 Sec.10 C400 to obtaina better utilisation of the joint. The purpose of this approach isto update the predicted resistance of the joint with the resultsfrom a limited number of tests in a manner consistent with thereliability approach of the RP.

F 200 Qualification of analysis method for other load conditions or for scaled joints201 If an analysis method predicts the tested response andstrength of a joint based on basic independently determinedmaterial properties according to DNV-OS-C501 Sec.10 C, theanalysis works well for the tested load conditions. The sameanalysis method may be used:

— for the same joint under different load conditions, if theother load conditions do not introduce new stress concen-trations in the analysis

— for a joint that is similar to an already qualified joint, if alllocal stress concentration points are similar to the alreadyqualified joint and all material properties are known inde-pendently.

202 Local stress concentrations are similar if the local geom-etry of the two joints and the resulting stress fields at theselocal points can be scaled by the same factor.

Table F1 Design approaches for different categories of jointsType of joint Analytical

approachQualifica-tion testing

Analyses com-bined with testing (updating)

Laminated joint x x xAdhesive joint x xMechanical joint x x

DET NORSKE VERITAS


203 An analysis method that predicts the test results properlybut not entirely based on independently obtained materialsdata can only be used for other load conditions or joint geom-etry if it can be demonstrated that the material values that werenot obtained by independent measurements can also be appliedfor the new conditions.

F 300 Multiple failure modes301 Most joint designs can fail by various failure modes. Allpossible failure modes shall be carefully identified and ana-lysed. See DNV-OS-C501 Sec.10 D.

F 400 Evaluation of in-service experience401 In service experience may be used as experimental evi-dence that a joint functions well.402 This evidence shall only be used if the load and environ-mental conditions of the in-service experience can be clearlydefined and if they match or are conservative for the new appli-cation.403 Material properties of the joints to be compared shouldbe similar. The analysis method should be able to address alldifferences between the joints according to B100 and 200.

F 500 Laminated joints501 Laminated joints rely on the strength of the interface forload transfer. The interface has resin dominated strength prop-erties. Defects in the interface tend to be more critical thandefects in the interface of plies of laminate, because the jointinterface is the only and critical load path.502 The strength of the joint may be different from thethrough thickness matrix properties of the laminate, becausethe joint may be a resin rich layer and the joint may be appliedto an already cured surface instead of a wet on wet connection.(see manufacturing). The strength of the joint should be docu-mented.503 Laminated joints are very sensitive to peel conditions.Peel stresses should be avoided.504 For the interface between the joining laminates thematrix design rules given in Sec.5 apply. The resistance of theinterface shall be determined with the same level of confidenceas specified in Sec.4 A600. It shall be recognised that theresistance of the interface between the laminates may not bethe same as the corresponding resistance parameter of the join-ing laminates. Resin rich layers may even have to be analysedby different failure criteria, e.g., the yield criterion in Sec.6 F.505 The laminates themselves, including possible over-lam-inations, shall be analysed like regular laminates.

F 600 Adhesive joints601 All issues related to laminated joints also apply to adhe-sive joints.602 Geometrical details should be clearly specified, espe-cially at points of stress concentrations like the edges of thejoints.603 The relationship between all elastic constants of bothsubstrates and the adhesive should be carefully considered.Mismatches may introduce stresses or strains that can causefailure of the joint.604 Thermal stresses should be considered.605 Long term performance of adhesive should be estab-lished with great care. The long-term performance is not onlyinfluenced by properties of the substrate, the adhesive and theinterface, but also by the surface preparation and applicationmethod.606 Relevant long-term data shall be established exactly forthe combination of materials, geometry, surface preparationand fabrication procedures used in the joint.

607 An adhesive joint may also introduce local throughthickness stresses in the composite laminate that can lead tofailure inside the laminate in the joint region.

F 700 Mechanical joints701 Mechanical joints are often very sensitive to geometricaltolerances.702 Creep of the materials shall be considered.703 The pretension of bolted connections shall be chosen byconsidering possible creep of the material under the bolt.704 It is preferred to design the joint in a way that its per-formance is independent of the matrix. This way matrix crack-ing or degradation of matrix properties are not important forthe performance of the joint.

G. Test RequirementsG 100 General101 Due to the uncertainties in designing connectors and lin-ers some testing is required to confirm predicted performance.102 Testing should be done whenever uncertainties in theanalysis cannot be resolved. These uncertainties may berelated to the structural analysis, boundary conditions, model-ling of local geometry, material properties, properties of inter-faces, etc. The procedure given in DNV-OS-C501 should befollowed for testing.103 The predicted performance of the Composite Metal con-nector Interface CMI and the resistance of the riser to externalpressure shall be confirmed by testing. Minimum test require-ments are given in table G1.

104 Testing may be done at room temperature and withwater as a pressure medium if the effect of temperaturechanges and fluid changes can be well described. If the effect

Table G1 Summary of test requirementsHigh safety class Normal safety class Reference

Design PhaseAxial test / pressure test

1 test to failure 1 test to failure G200

Axial or bending fatigue of CMI

2 tests to 50x actual number of cycles or sur-vival test to about 100000 cycles

1 test to 30x actual number of cycles or survival test to about 100000 cycles

G300

Stress rupture test of CMI if matrix prop-erties are crit-ical or fibres can creep

2 tests to 50x actual lifetime or survival test to about 1000 hours

1 test to 30x actual lifetime or survival test to about 1000 hours

G400

If the inner liner is bonded to the laminate

Test bond between liner and laminate

Test bond between liner and laminate

G500

External pressure test

Use specimen after fatigue test and expose to maximum external pressure

Use specimen after fatigue test and expose to maximum external pressure

G500

If impact requirement

Impact tests Impact test Sec.5 F300 to F500

After fabricationPressure test Test to a given

internal pressure for each riser component

Test to a given internal pressurefor each riser com-ponent

Sec.5 A300

DET NORSKE VERITAS


of changing the environmental conditions is uncertain, testingshould be carried out in the worst conditions.

G 200 Axial/ pressure test of riser with composite metal interface201 One axial tensile test to failure shall be carried out. Theaxial tensile test can be replaced by a pressure test if the axialload created by the pressure test exceeds the maximum serviceload of the riser.202 The failure load or pressure should be at least the pre-dicted μ-σ for high safety class and μ-2σ for normal safetyclass, where μ is the mean prediction and σ is one standarddeviation of the predicted load. If more than one test is donethe requirements are given in DNV-OS-C501 Sec.10 C200.203 If the performance of the CMI is very dependent on theinternal pressure it should be evaluated if axial testing with andwithout internal pressure is required to demonstrate the per-formance of the CMI. The evaluation should be based on howwell the strength can be predicted by modelling for the twoconditions.

G 300 Cyclic fatigue testing for end fittings and compos-ite metal interface301 Fatigue testing should be performed in axial tension orin bending. The most relevant test should be found by evaluat-ing the design analysis.302 For high safety class at least two survival tests shall becarried out. The specimen should not fail during the survivaltest and it should not show unexpected damage. The require-ments to the testing are:

— tests should be carried out up to five times the maximumnumber of design cycles with realistic amplitudes andmean loads that the component will experience. If constantamplitude testing is carried out tests should be carried outup to 50 times the maximum number of design cycles tocompensate for uncertainty in sequence effects.

— if the anticipated lifetime exceeds 105 cycles testing up to105 cycles may be sufficient. The load levels should bechosen such that testing of the two specimens is completedafter at least 104 and 105 cycles respectively. The loga-rithms of the two test results shall fall within μ-σ of thelogarithm of the anticipated number of cycles to failure,where μ is the mean of the logarithm of the predictednumber of cycles to failure and σ is one standard deviationof the logarithm of the predicted number of cycles to fail-ure, both interpreted from a log(stress)-log(lifetime) dia-gram for the anticipated number of cycles to failure. Ifmore tests are made the requirements are given in DNV-OS-C501 Sec.4 H806.

303 For normal safety class at least one survival test shall becarried out. The specimen should not fail during the survivaltest and it should not show unexpected damage. The require-ments to the testing are:

— tests should be carried out up to three times the maximumnumber of design cycles with realistic amplitudes andmean loads that the component will experience. If constantamplitude testing is carried out tests should be carried outup to 30 times the maximum number of design cycles tocompensate for uncertainty in sequence effects.

— if the anticipated lifetime exceeds 105 cycles testing up to105 cycles may be sufficient. The load levels should bechosen such that testing of the two specimens is completedafter at least 104 and 105 cycles respectively. The loga-rithms of the test results shall fall within μ-2σ of the loga-rithm of the anticipated number of cycles to failure, whereμ is the mean of the logarithm of the predicted number ofcycles to failure and σ is one standard deviation of the log-arithm of the predicted number of cycles to failure, both

interpreted from a log(stress)-log(lifetime) diagram for theanticipated number of cycles to failure. If more tests aremade the requirements are given in DNV-OS-C501 Sec.4H806.

304 For low safety class long term testing is not required.305 The sequence of the failure modes in the test shall be thesame as predicted in the design. If the sequence is different orif other failure modes are observed, the design shall be care-fully re-evaluated.306 Fatigue tests should be carried out with a typical loadsequence or with constant load amplitude. If a clearly definedload sequence exists, load sequence testing should be pre-ferred.307 In some cases high amplitude fatigue testing may intro-duce unrealistic failure modes in the structure. In other cases,the required number of test cycles may lead to unreasonablelong test times. In these cases an individual evaluation of thetest conditions should be made that fulfils the requirements of302 or 303 as closely as possible.308 Additional tests may be required if resistance to a failuremode cannot be shown by analysis with sufficient confidenceand if this failure mode is not tested by the tests describedabove.

G 400 Stress rupture testing for end fittings and com-posite metal interface401 Only if the performance of the metal composite interfacedepends on matrix properties or adhesives, or if the fibres inthe laminate can creep, long term static testing should be per-formed. Two survival tests should be carried out for highsafety class applications and one survival test for normal safetyclass applications. If it can be shown that the CMI keeps itsstrength if the matrix in the laminate is cracked and degradedand the fibres do not creep, long term static testing is notrequired.402 For high safety class at least two survival tests shall becarried out. The specimen should not fail during the survivaltest and it should not show unexpected damage. The require-ments to the test results are:

— tests should be carried out up to five times the maximumdesign life with realistic mean loads that the componentwill experience. If constant load testing is carried out testsshould be carried out up to 50 times the design life to com-pensate for uncertainty in sequence effects.

— if the anticipated lifetime exceeds 1000 hours testing up to1000 hours may be sufficient. The load levels should bechosen such that testing is completed after 103 hours. Thelogarithms of the two test results shall fall within μ-σ ofthe logarithm of the anticipated lifetime, where μ is themean of the logarithm of the predicted lifetime and σ isone standard deviation of the logarithm of the predictedlifetime, both interpreted from a log(stress)-log(lifetime)diagram for the anticipated lifetime. If more tests are madethe requirements are given in DNV-OS-C501 Sec.4 H806.

403 For normal safety class at least one survival test shall becarried out. The specimen should not fail during the survivaltest and it should not show unexpected damage. The require-ments to the test results are:

— tests should be carried out up to five times the maximumdesign life with realistic mean loads that the componentwill experience. If constant load testing is carried out testsshould be carried out up to 30 times the design life to com-pensate for uncertainty in sequence effects.

— if the anticipated lifetime exceeds 1000 hours testing up to1000 hours may be sufficient. The load levels should bechosen such that testing is completed after 103 hours. Thelogarithms of the two test results shall fall within μ-2σ of

DET NORSKE VERITAS


the logarithm of the anticipated lifetime, where μ is themean of the logarithm of the predicted lifetime and σ isone standard deviation of the logarithm of the predictedlifetime, both interpreted from a log(stress)-log(lifetime)diagram for the anticipated lifetime. If more tests are madethe requirements are given inDNV-OS-C501 Sec.4 H806.

404 For low safety class long term testing is not required.405 The sequence of the failure modes in the test shall be thesame as predicted in the design. If the sequence is different orif other failure modes are observed, the design shall be care-fully re-evaluated.406 Stress rupture tests should be carried out with a typicalload sequence or with a constant load. If a clearly defined loadsequence exists, load sequence testing should be preferred.

G 500 Inner liner test requirements501 If the design relies on a bond between liner and compos-ite laminate, the quality of the bond shall be tested. Tests canbe done on the pipe or representative smaller specimens. If thelaminate may have cracks, but the liner not, the requirementsin D507 should be considered.502 If the riser may be exposed to external pressures itsresistance to buckling should be tested. The test shall be car-ried out by applying maximum external pressure to the riser.The riser and liner shall be produced with controlled and rep-resentative tolerances. Testing shall be carried out according tothe requirements in DNV-OS-C501 Sec.10. External pressuretesting shall be carried out on test specimens that have previ-ously been exposed to high loads and have developed repre-sentative degradation of material properties. The fatigue testsspecified in G200 can most likely also be used to introducerepresentative damage.503 Testing of a bent riser under external pressure should beconsidered if:

— if the riser can be bent in service and this bending couldreduce the resistance to internal pressure (e.g. due to oval-isation)

— if the effect of ovalisation on the buckling resistance cannotbe predicted by calculations with sufficient confidence.

G 600 Specimen geometry - Scaled specimen601 The specimen geometry for testing may be chosen to bedifferent from the actual under certain conditions.602 Specimens may be shorter than in reality. The freelength of the riser pipe between end-fittings should be at least6 x diameter.603 Most test specimens have a relevant CMI at both ends ofthe riser pipe. In this case testing one riser pipe with two CMIscan be used to fulfil the requirement of two survival tests forthe CMI, provided both CMIs are exposed to the same loadingconditions.604 Scaled specimens may be used if analytical calculationscan demonstrate that:

— all critical stress states and local stress concentrations inthe joint of the scaled specimen and the actual riser aresimilar, i.e., all stresses are scaled by the same factorbetween actual riser and test specimen

— the behaviour and failure of the specimen and the actualriser can be calculated based on independently obtainedmaterial parameters. This means no parameters in theanalysis should be based on adjustments to make largescale data fit

— the sequence of predicted failure modes is the same for thescaled specimen and the actual riser over the entire life-time of the riser

— an analysis method that predicts the test results properlybut not entirely based on independently obtained materialsdata, may be used for other joint geometry. In that case itshould be demonstrated that the material values that werenot obtained by independent measurements can also beapplied for the new conditions.

605 Tests on previous risers may be used as testing evidenceif the scaling requirement in 604 is fulfilled. Materials and pro-duction process should also be identical or similar. Similarityshould be evaluated based on the requirements in DNV-OS-C501 Sec.4.

DET NORSKE VERITAS


SECTION 7MATERIALS

A. GeneralA 100 Objective101 This section specifies the requirements for materials ofriser pipe, components, equipment and structural items in theriser system, with regard to the characteristic properties ofmaterials. The requirements are relevant both for pressure con-taining and for load carrying parts.

A 200 Material Description201 The materials selected shall be suitable for the intendeduse during the entire service life. The materials for use in theriser system shall have the dimensions and mechanical proper-ties, such as strength, ductility, toughness, corrosion and wearresistance, necessary to comply with the assumptions made inthe design.202 The materials selected shall be suitable for the intendeduse during the entire service life. The materials for use in theriser system shall have the dimensions and mechanical proper-ties, such as strength, ductility, toughness, corrosion and wearresistance, necessary to comply with the assumptions made inthe design. 203 Composite material properties shall be described andtested as given in DNV-OS-C501.204 Metal properties shall be described and tested as givenin DNV-OS-F201.

205 Titanium parts should be described and tested as givenin DNV-RP-F201.

B. Fabrication

B 100 Objective101 This section specifies the requirements for fabrication ofriser pipe, components, equipment and structural items in theriser system, The requirements are relevant both for pressurecontaining and for load carrying parts.

B 200 Material Description201 The fabrication process shall be well controlled toensure that the material properties and tolerances assumed inthe design are achieved.202 Fabrication of composite parts shall be evaluated asgiven in DNV Offshore standard for composite componentsDNV-OS-C501 Section 11. 203 Fabrication of metal parts shall be evaluated as given inDNV Offshore standard for dynamic risers DNV-OS-F201.204 Fabrication of titanium parts shall be evaluated as givenin DNV-RP-F201 "Titanium Risers".

DET NORSKE VERITAS


SECTION 8DOCUMENTATION AND VERIFICATION

A. GeneralA 100 Documentation and verification 101 All requirements given in DNV-OS-F201 apply also tocomposite risers.

DET NORSKE VERITAS


SECTION 9OPERATION, MAINTENANCE, REASSESSMENT, REPAIR

A. General

A 100 Objective101 The objective of this section is to provide requirementsfor operation and in-service inspections. This section also pro-vides general guidance on structural integrity assessment ofrisers to demonstrate fitness for purpose in case deviationsfrom design appear during operation.

B. In-service Inspection, Replacement and Monitoring

B 100 General101 The requirements for composite risers are the same asgiven in DNV-OS-F201.102 Some special considerations apply for inspection meth-ods.

B 200 Inspection methods201 If the riser is designed according to this document and ifthe predicted time to failure divided by the fatigue safety factorfrom Sec.5 is longer than the intended service life, inspectionis not expected to be necessary and need not be included in thein the operation and maintenance documents. However, if theintended service life is exceeded or if load or environmentalconditions were worse than planned, then the componentshould be inspected and refurbished if necessary or replaced.202 The reliability and functionality of all inspection meth-ods should be documented.203 Available inspection methods can often not detect allcritical failure mechanisms. However, the methods may detectpreceding failure mechanisms. A link between detectable fail-ure mechanisms and critical failure mechanisms shall be estab-lished.204 In many cases, a complete inspection programme cannotbe developed due to the limited capabilities of available NDTequipment. In that case the following alternatives may be used:205 Inspection of components during or right after manufac-turing may be replaced by well documented production con-trol.206 Inspection to detect damage due to accidental loads oroverloads may be compensated for by monitoring the loadsand comparing them to the design loads. If this method is usedthe component must be replaced after all overloads or otherevents exceeding the design requirements. This approach shallbe agreed upon with the customer.207 Inspection frequencies and acceptance criteria should bedetermined for each project.

C. Reassessment

C 100 General101 The requirements for composite risers are the same as

given in DNV-OS-F201. The only exception is references tocorrosion allowance. These are not relevant for compositematerials.

D. Repair

D 100 General101 This section applies to repairs of defects that influencethe structural integrity or a functional requirement, e.g. tight-ness.102 Cosmetic, non-structural and non-functional repairs donot need to be qualified.

D 200 Repair procedure201 A repair procedure shall be given for each component.202 A repair shall restore the same level of safety and func-tionality as the original structure, unless changes are acceptedby all parties in the project.203 An acceptable repair solution is to replace the entirecomponent if it is damaged. This approach requires that thecomponent can be taken out of the system.204 It may also be acceptable to keep a component in servicewith a certain amount of damage without repairing it. The sizeand kind of acceptable damage shall be defined and it must bepossible to inspect the damage. The possible damage shall beconsidered in the design of the structure.205 If local damage may happen to the structure detailedprocedures to repair such anticipated damage shall be given.206 If the damage is due to an unknown loading condition oraccident, an analysis of the damage situation shall be carriedout. The analysis shall identify whether the damage was due toa design mistake or an unexpected load condition. If the unex-pected load may reoccur, a design change may be required.

D 300 Requirements for a repair301 A repair should restore the stiffness and strength of theoriginal part. If the stiffness and/or strength cannot be restored,the performance of the component and the total system underthe new conditions shall be evaluated. 302 It shall be documented that local reduction in strengthmay not be critical for the total performance of the structure.

D 400 Qualification of a repair401 A repair is basically a joint introduced into the structure.The repair shall be qualified in the same way as a joint (seeSec.6 B).402 The repair procedure used to qualify the joint shall alsobe applicable for each particular repair situation.403 Suitable conditions for repair work shall be arranged andmaintained during the repair. This is mandatory, irrespectiveof whether the repair is carried out on site or elsewhere. If suit-able conditions cannot be arranged and maintained on site, thecomponent should be moved to a more suitable site.

DET NORSKE VERITAS


E. MaintenanceE 100 General101 A maintenance procedure shall be given for each com-ponent. All aspects related to maintenance should be covered.

Guidance note:Appropriate cleaning agents should be described. If the compo-nent is painted suitable paints should be identified and methodsfor removal and application of the paint should be given if rele-vant.


F. RetirementF 100 General101 A method for retirement of all components shall be doc-umented.

DET NORSKE VERITAS


DET NORSKE VERITAS

DNV-RP-F202 Composite Risers October 2010

Documents

det norske

fibre reinforced

gross plastic

actively applying

regularly

serviceability

ultimate limit

axial tensile