DNV-OS-J101: Design of Offshore Wind Turbine Structureshuniv.hongik.ac.kr/~geotech/key...

OFFSHORE STANDARD

DET NORSKE VERITAS

DNV-OS-J101

DESIGN OF OFFSHORE WIND TURBINE STRUCTURES

OCTOBER 2007

Since issued in print (October 2007), this booklet has been amended, latest in December 2008. See the reference to “Amendments and Corrections” on the next page.

FOREWORDDET NORSKE VERITAS (DNV) is an autonomous and independent foundation with the objectives of safeguarding life, prop-erty and the environment, at sea and onshore. DNV undertakes classification, certification, and other verification and consultancyservices relating to quality of ships, offshore units and installations, and onshore industries worldwide, and carries out researchin relation to these functions.DNV Offshore Codes consist of a three level hierarchy of documents:— Offshore Service Specifications. Provide principles and procedures of DNV classification, certification, verification and con-

sultancy services.— Offshore Standards. Provide technical provisions and acceptance criteria for general use by the offshore industry as well as

the technical basis for DNV offshore services.— Recommended Practices. Provide proven technology and sound engineering practice as well as guidance for the higher level

Offshore Service Specifications and Offshore Standards.DNV Offshore Codes 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) Wind Turbines

Amendments and Corrections This document is valid until superseded by a new revision. Minor amendments and corrections will be published in a separatedocument normally updated twice per year (April and October). For a complete listing of the changes, see the “Amendments and Corrections” document located at: http://webshop.dnv.com/global/, under category “Offshore Codes”.
The electronic web-versions of the DNV Offshore Codes will be regularly updated to include these amendments and corrections.
Comments may be sent by e-mail to [email protected] subscription orders or information about subscription terms, please use [email protected] information about DNV services, research and publications can be found at http://www.dnv.com, or can be obtained from DNV, Veritasveien 1, NO-1322 Høvik, Norway; Tel +47 67 57 99 00, Fax +47 67 57 99 11.

© Det Norske Veritas. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, including photocopying and recording, without the prior written consent of Det Norske Veritas.

Computer Typesetting (FM+SGML) by Det Norske Veritas.Printed in Norway.

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.

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover Changes –

AcknowledgmentsThis Offshore Standard makes use of eight figures and onetable provided by Mærsk Olie og Gas AS. The eight figuresconsist of Figures 11 and 12 in Section 7, Figure 1 in AppendixA, Figure 1 in Appendix C and Figures 1 through 4 in Appen-dix D. The table consists of Table A1 in Appendix C. MærskOlie og Gas AS is gratefully acknowledged for granting DNVpermission to use this material.The standard also makes use of one figure provided by Prof.S.K. Chakrabarti. The figure appears as Figure 7 in Sec.3. Prof.Chakrabarti is gratefully acknowledged for granting DNV per-mission to use this figure.

GeneralThis document supersedes the June 2004 edition.

Main changes

— Reformulation of load cases in line with Committee Draftof IEC61400-3.

— Adjustment of safety factor requirements to reflect targetsafety level of IEC61400-1.

— Expanded guidance for wind and wave modelling, in par-ticular for waves and wave loading in shallow waters.

— Rewritten section on concrete design.— Removal of non-technical parts and non-design parts from

the standard, i.e. commercial service descriptions anddescriptions of manufacturing surveys and instructions tosurveyors are removed.

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – Changes see note on front cover

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover Contents –

CONTENTS

Sec. 1 Introduction........................................................... 9

A. General....................................................................................9A 100 General.............................................................................. 9A 200 Objectives ......................................................................... 9A 300 Scope and application ....................................................... 9A 400 Non-DNV codes ............................................................... 9

B. Normative References ..........................................................10B 100 General............................................................................ 10

C. Informative References.........................................................10C 100 General............................................................................ 10

D. Definitions ............................................................................11D 100 Verbal forms ................................................................... 11D 200 Terms .............................................................................. 11

E. Abbreviations and Symbols..................................................13E 100 Abbreviations.................................................................. 13E 200 Symbols .......................................................................... 14

F. Support Structure Concepts ..................................................15F 100 Introduction..................................................................... 15F 200 Gravity-based structures and gravity-pile structures ...... 16F 300 Jacket-monopile hybrids and tripods .............................. 16F 400 Monopiles ....................................................................... 16F 500 Supported monopiles and guyed towers ......................... 16F 600 Tripods with buckets....................................................... 16F 700 Suction buckets ............................................................... 16F 800 Lattice towers.................................................................. 16F 900 Low-roll floaters ............................................................. 17F 1000 Tension leg platforms ..................................................... 17

Sec. 2 Design Principles................................................. 18

A. Introduction ..........................................................................18A 100 General............................................................................ 18A 200 Aim of the design............................................................ 18

B. General Design Conditions...................................................18B 100 General............................................................................ 18

C. Safety Classes and Target Safety Level ...............................18C 100 Safety classes .................................................................. 18C 200 Target safety ................................................................... 18

D. Limit States...........................................................................19D 100 General............................................................................ 19

E. Design by the Partial Safety Factor Method.........................19E 100 General............................................................................ 19E 200 The partial safety factor format ...................................... 19E 300 Characteristic load effect ................................................ 21E 400 Characteristic resistance ................................................. 21E 500 Load and resistance factors ............................................ 21

F. Design by Direct Simulation of Combined Load Effect of Simultaneous Load Processes ......................21

F 100 General............................................................................ 21F 200 Design format ................................................................. 21F 300 Characteristic load effect ................................................ 21F 400 Characteristic resistance ................................................. 22

G. Design Assisted by Testing ..................................................22G 100 General............................................................................ 22G 200 Full-scale testing and observation

of performance of existing structures ............................. 22

H. Probability-based Design......................................................22H 100 Definition ........................................................................ 22H 200 General............................................................................ 22

Sec. 3 Site Conditions ................................................... 23

A. General..................................................................................23A 100 Definition ........................................................................ 23

B. Wind Climate........................................................................23B 100 Wind conditions.............................................................. 23B 200 Parameters for normal wind conditions.......................... 23B 300 Wind data........................................................................ 23B 400 Wind modelling .............................................................. 24B 500 Reference wind conditions

and reference wind speeds .............................................. 26

C. Wave Climate .......................................................................29C 100 Wave parameters ............................................................ 29C 200 Wave data ....................................................................... 29C 300 Wave modelling.............................................................. 30C 400 Reference sea states and reference wave heights ........... 32C 500 Wave theories and wave kinematics............................... 33C 600 Breaking waves............................................................... 35

D. Current ..................................................................................35D 100 Current parameters.......................................................... 35D 200 Current data .................................................................... 35D 300 Current modelling........................................................... 35

E. Water Level ..........................................................................36E 100 Water level parameters ................................................... 36E 200 Water level data .............................................................. 36E 300 Water level modelling..................................................... 36

F. Ice .........................................................................................36F 100 Sea ice............................................................................. 36F 200 Snow and ice accumulation ............................................ 36F 300 Ice modelling .................................................................. 36

G. Soil Investigations and Geotechnical Data ...........................37G 100 Soil investigations........................................................... 37

H. Other Site Conditions ...........................................................38H 100 Seismicity ....................................................................... 38H 200 Salinity............................................................................ 38H 300 Temperature.................................................................... 38H 400 Marine growth ................................................................ 38H 500 Air density ...................................................................... 38H 600 Ship traffic ...................................................................... 38H 700 Disposed matters............................................................. 39H 800 Pipelines and cables........................................................ 39

Sec. 4 Loads and Load Effects .................................... 40

A. Introduction...........................................................................40A 100 General............................................................................ 40

B. Basis for Selection of Characteristic Loads..........................40B 100 General............................................................................ 40

C. Permanent Loads (G)............................................................40C 100 General............................................................................ 40

D. Variable Functional Loads (Q) .............................................40D 100 General............................................................................ 40D 200 Variable functional loads on platform areas................... 41D 300 Ship impacts and collisions ............................................ 41D 400 Tank pressures ................................................................ 41D 500 Miscellaneous loads........................................................ 42

E. Environmental Loads (E)......................................................42E 100 General............................................................................ 42E 200 Wind turbine loads.......................................................... 42E 300 Determination of characteristic hydrodynamic loads ..... 48E 400 Wave loads...................................................................... 48E 500 Ice loads .......................................................................... 52E 600 Water level loads ............................................................ 54E 700 Earthquake loads............................................................. 54E 800 Marine growth ................................................................ 55E 900 Scour ............................................................................... 55E 1000 Transportation loads and installation loads .................... 55

F. Combination of Environmental Loads..................................55F 100 General............................................................................ 55F 200 Environmental states....................................................... 56F 300 Environmental contours.................................................. 56

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – Contents see note on front cover

F 400 Combined load and load effect due to wind load and wave load.............................................56

F 500 Linear combinations of wind load and wave load ..........56F 600 Combination of wind load and wave load

by simulation...................................................................57F 700 Basic load cases ..............................................................57F 800 Transient load cases .......................................................59F 900 Load combination for the fatigue limit state ...................59

G. Load Effect Analysis ............................................................ 60G 100 General ............................................................................60G 200 Global motion analysis....................................................60G 300 Load effects in structures and foundation soils...............60

H. Deformation Loads ............................................................... 60H 100 General ............................................................................60H 200 Temperature loads...........................................................61H 300 Settlements ......................................................................61

Sec. 5 Load and Resistance Factors ............................ 62

A. Load Factors ......................................................................... 62A 100 Load factors for the ULS ................................................62A 200 Load factor for the FLS...................................................62A 300 Load factor for the SLS...................................................62

B. Resistance Factors ................................................................ 62B 100 Resistance factors for the ULS........................................62B 200 Resistance factors for the FLS ........................................62B 300 Resistance factors for the SLS ........................................62

Sec. 6 Materials.............................................................. 63

A. Selection of Steel Materials and Inspection Principles ........ 63A 100 General ............................................................................63A 200 Design temperatures........................................................63A 300 Structural category ..........................................................63A 400 Structural steel.................................................................64

B. Selection of Concrete Materials .......................................... 66B 100 General ............................................................................66B 200 Material requirements .....................................................66B 300 Concrete ..........................................................................67B 400 Grout and mortar .............................................................67B 500 Reinforcement steel.........................................................67B 600 Prestressing steel .............................................................67

C. Grout Materials and Material Testing .................................. 67C 100 General ............................................................................67C 200 Experimental verification................................................68

Sec. 7 Design of Steel Structures.................................. 69

A. Ultimate Limit States – General ........................................... 69A 100 General ............................................................................69A 200 Structural analysis ...........................................................69A 300 Ductility ..........................................................................69A 400 Yield check .....................................................................69A 500 Buckling check................................................................69

B. Ultimate Limit States – Shell Structures .............................. 69B 100 General ............................................................................69

C. Ultimate Limit States – Tubular Members, Tubular Joints and Conical Transitions ................................ 70

C 100 General ............................................................................70

D. Ultimate Limit States – Non-Tubular Beams, Columns and Frames ............................................................ 70

D 100 General ............................................................................70

E. Ultimate Limit States – Special Provisions for Plating and Stiffeners ....................... 70

E 100 Scope...............................................................................70E 200 Minimum thickness.........................................................70E 300 Bending of plating...........................................................70E 400 Stiffeners .........................................................................71

F. Ultimate Limit States – Special Provisions for Girders and Girder Systems ............. 71

F 100 Scope...............................................................................71F 200 Minimum thickness.........................................................71

F 300 Bending and shear...........................................................71F 400 Effective flange ...............................................................71F 500 Effective web ..................................................................72F 600 Strength requirements for simple girders........................72F 700 Complex girder system ...................................................72

G. Ultimate Limit States – Slip-resistant Bolt Connections......73G 100 General ............................................................................73

H. Ultimate Limit States – Welded Connections ......................74H 100 General ............................................................................74H 200 Types of welded steel joints............................................74H 300 Weld size.........................................................................75

I. Serviceability Limit States....................................................78I 100 General ............................................................................78I 200 Deflection criteria ...........................................................78I 300 Out-of-plane deflection of local plates............................78

J. Fatigue Limit States..............................................................78J 100 Fatigue limit state............................................................78J 200 Characteristic S-N curves................................................78J 300 Characteristic stress range distribution ...........................79J 400 Characteristic and design cumulative damage ................81J 500 Design fatigue factors .....................................................81J 600 Material factors for fatigue .............................................82J 700 Design requirement .........................................................82J 800 Improved fatigue performance

of welded structures by grinding.....................................82

Sec. 8 Detailed Design of Offshore Concrete Structures............................................ 84

A. General..................................................................................84A 100 Introduction.....................................................................84A 200 Material ...........................................................................84A 300 Composite structures.......................................................84

B. Design Principles ..................................................................84B 100 Design material strength .................................................84

C. Basis for Design by Calculation ...........................................85C 100 Concrete grades and in-situ strength of concrete ............85

D. Bending Moment and Axial Force (ULS) ............................85D 100 General ............................................................................85

E. Fatigue Limit State ...............................................................85E 100 General ...........................................................................85

F. Accidental Limit State ..........................................................85F 100 General ............................................................................85

G. Serviceability Limit State .....................................................85G 100 Durability ........................................................................85G 200 Crack width calculation ..................................................85G 300 Other serviceability limit states.......................................86

H. Detailing of Reinforcement ..................................................86H 100 Positioning ......................................................................86

I. Corrosion Control and Electrical Earthing ...........................86I 100 Corrosion control ............................................................86I 200 Electrical earthing ...........................................................86

J. Construction..........................................................................86J 100 General ............................................................................86J 200 Inspection classes............................................................86

Sec. 9 Design and Construction of Grouted Connections ..................................... 87

A. Introduction...........................................................................87A 100 General ............................................................................87A 200 Design principles.............................................................87

B. Ultimate Limit States............................................................87B 100 Connections subjected to axial load and torque..............87B 200 Connections subjected to bending moment

and shear loading ............................................................89

C. Fatigue Limit States..............................................................90C 100 General ............................................................................90

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover Contents –

C 200 Connections subjected to axial load and torque.............. 90C 300 Connections subjected to bending moment

and shear loading ............................................................ 90

D. Requirements to Verification and Material Factors .............91D 100 Experimental verification ............................................... 91D 200 Material factors for grouted connections ........................ 91

E. Grouting Operations .............................................................92E 100 General............................................................................ 92E 200 Operations prior to grouting ........................................... 92E 300 Monitoring ...................................................................... 92

Sec. 10 Foundation Design .............................................. 93

A. General..................................................................................93A 100 Introduction..................................................................... 93A 200 Soil investigations........................................................... 93A 300 Characteristic properties of soil ...................................... 93A 400 Effects of cyclic loading ................................................. 94A 500 Soil-structure interaction................................................. 94

B. Stability of Seabed................................................................94B 100 Slope stability ................................................................. 94B 200 Hydraulic stability........................................................... 94B 300 Scour and scour prevention............................................. 94

C. Pile Foundations ...................................................................95C 100 General............................................................................ 95C 200 Design criteria for monopile foundations ....................... 95C 300 Design criteria for jacket pile foundations...................... 96C 400 Design of piles subject to scour ...................................... 97

D. Gravity Base Foundations .................................................... 97D 100 General............................................................................ 97D 200 Stability of foundations................................................... 97D 300 Settlements and displacements ....................................... 98D 400 Soil reactions on foundation structure ............................ 98D 500 Soil modelling for dynamic analysis .............................. 98D 600 Filling of voids................................................................ 98

Sec. 11 Corrosion Protection ........................................ 101

A. General................................................................................101A 100 General.......................................................................... 101

B. Acceptable Corrosion Protection........................................101B 100 Atmospheric zone ......................................................... 101B 200 Splash zone ................................................................... 101B 300 Submerged zone............................................................ 101B 400 Closed compartments.................................................... 102

C. Cathodic Protection ............................................................102C 100 General.......................................................................... 102

D. Coating................................................................................102D 100 General.......................................................................... 102

Sec. 12 Transport and Installation .............................. 103

A. Marine Operations ..............................................................103A 100 Warranty surveys .......................................................... 103A 200 Planning of operations .................................................. 103A 300 Design loads.................................................................. 104A 400 Structural design ........................................................... 104A 500 Load transfer operations ............................................... 104A 600 Towing .......................................................................... 104A 700 Offshore installation ..................................................... 104A 800 Lifting ........................................................................... 104A 900 Subsea operations ......................................................... 104

Sec. 13 In-Service Inspection, Maintenance and Monitoring........................... 105

A. General................................................................................105A 100 General.......................................................................... 105

B. Periodical Inspections.........................................................105B 100 General.......................................................................... 105B 200 Preparation for periodical inspections .......................... 105B 300 Interval between inspections......................................... 105

B 400 Inspection results .......................................................... 105B 500 Reporting ...................................................................... 105

C. Periodical Inspection of Wind Turbines ............................105C 100 Interval between inspections......................................... 105C 200 Scope for inspection ..................................................... 105

D. Periodical Inspection of Structural and Electrical Systems above Water ..................................105

D 100 Interval between inspections......................................... 105D 200 Scope for inspection ..................................................... 106

E. Periodical Inspection of Structures Below Water...............106E 100 Interval between inspections......................................... 106E 200 Scope for inspection ..................................................... 106

F. Periodical Inspection of Sea Cables ...................................107F 100 Interval between inspections......................................... 107F 200 Scope for inspection ..................................................... 107

G. Deviations ...........................................................................107G 100 General.......................................................................... 107

App. A Stress Concentration Factors for Tubular Joints............................................................ 108

A. Calculation of Stress Concentration Factors.......................108A 100 General.......................................................................... 108

App. B Local Joint Flexibilities for Tubular Joints............................................................ 115

A. Calculation of Local Joint Flexibilities...............................115A 100 General.......................................................................... 115

App. C Stress Concentration Factors for Girth Welds ................................................................ 117

A. Calculation of Stress Concentration Factors for Hot Spots.......................................................................117

A 100 General.......................................................................... 117

App. D Stress Extrapolation for Welds ....................... 118

A. Stress Extrapolation to Determine Hot Spot Stresses.........118A 100 General.......................................................................... 118

App. E Tubular Connections – Fracture Mechanics Analyses and Calculations ........... 120

A. Stress Concentrations at Tubular Joints..............................120A 100 General.......................................................................... 120

B. Stresses at Tubular Joints....................................................120B 100 General.......................................................................... 120

C. Stress Intensity Factor.........................................................121C 100 General.......................................................................... 121C 200 Correction factor for membrane stress component....... 121C 300 Correction factor for bending stress component........... 122C 400 Crack shape and initial crack size................................. 122C 500 Load Shedding .............................................................. 122C 600 Crack Growth ............................................................... 123

App. F Pile Resistance and Load-displacement Relationships ........................... 124

A. Axial Pile Resistance ..........................................................124A 100 General.......................................................................... 124A 200 Clay............................................................................... 124A 300 Sand .............................................................................. 124A 400 t-z curves....................................................................... 125

B. Laterally Loaded Piles ........................................................125B 100 General.......................................................................... 125B 200 Clay............................................................................... 126B 300 Sand .............................................................................. 126B 400 Application of p-y curves ............................................. 127

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – Contents see note on front cover

App. G Bearing Capacity Formulae for Gravity Base Foundations......................................... 128

A. Forces.................................................................................. 128A 100 General ..........................................................................128

B. Correction for Torque......................................................... 128B 100 General ..........................................................................128

C. Effective Foundation Area.................................................. 128C 100 General ..........................................................................128

D. Bearing Capacity ................................................................ 129D 100 General ..........................................................................129D 200 Bearing capacity formulae for drained conditions ........129D 300 Bearing capacity formulae

for undrained conditions, φ = 0 .....................................129

E. Extremely Eccentric Loading ............................................. 129E 100 General ..........................................................................129

F. Sliding Resistance............................................................... 130F 100 General ..........................................................................130

App. H Cross Section Types.......................................... 131

A. Cross Section Types ........................................................... 131A 100 General ..........................................................................131A 200 Cross section requirements for plastic analysis ............131A 300 Cross section requirements

when elastic global analysis is used ..............................131

App. I Extreme Wind Speed Events ........................... 133

App. J Scour at a Vertical Pile .................................... 134

A. Flow around a Vertical Pile ................................................ 134A 100 General ..........................................................................134

B. Bed Shear Stress ................................................................. 134B 100 General ..........................................................................134

C. Local Scour......................................................................... 134C 100 General ..........................................................................134C 200 Scour depth ...................................................................134C 300 Lateral extension of scour hole .....................................135C 400 Time scale of scour .......................................................135

App. K Calculations by Finite Element Method......... 136

A. Introduction.........................................................................136A 100 General ..........................................................................136

B. Types of Analysis ...............................................................136B 100 General ..........................................................................136B 200 Static analysis................................................................136B 300 Frequency analysis........................................................136B 400 Dynamic analysis ..........................................................136B 500 Stability/buckling analysis ............................................136B 600 Thermal analysis ...........................................................136B 700 Other types of analyses .................................................136

C. Modelling............................................................................136C 100 General ..........................................................................136C 200 Model ............................................................................136C 300 Coordinate systems .......................................................136C 400 Material properties ........................................................137C 500 Material models.............................................................137C 600 Elements........................................................................137C 700 Element types................................................................137C 800 Combinations ................................................................137C 900 Element size and distribution of elements ....................137C 1000 Element quality .............................................................138C 1100 Boundary conditions .....................................................138C 1200 Types of restraints.........................................................138C 1300 Symmetry/antimetry......................................................138C 1400 Loads.............................................................................139C 1500 Load application............................................................139

D. Documentation....................................................................139D 100 Model ............................................................................139D 200 Geometry control ..........................................................139D 300 Mass – volume – centre of gravity................................139D 400 Material .........................................................................139D 500 Element type .................................................................139D 600 Local coordinate system................................................139D 700 Loads and boundary conditions ....................................139D 800 Reactions.......................................................................139D 900 Mesh refinement ...........................................................139D 1000 Results...........................................................................139

App. L Ice Loads for Conical Structures.................... 141

A. Calculation of Ice Loads.....................................................141A 100 General ..........................................................................141

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover Sec.1 –

SECTION 1INTRODUCTION

A. General

A 100 General101 This offshore standard provides principles, technicalrequirements and guidance for design, construction and in-service inspection of offshore wind turbine structures.102 DNV-OS-J101 is the DNV standard for design of off-shore wind turbine structures. The standard covers design,construction, installation and inspection of offshore wind tur-bine structures. The design principles and overall requirementsare defined in this standard. The standard can be used as astand-alone document.103 The standard shall be used for design of support struc-tures and foundations for offshore wind turbines. The standardshall also be used for design of support structures and founda-tions for other structures in an offshore wind farm, such asmeteorological masts.The standard does not cover design of support structures andfoundations for transformer stations for wind farms. For designof support structures and foundations for transformer stationsDNV-OS-C101 applies.

Guidance note:DNV-OS-C101 offers the choice of designing unmanned struc-tures with a lower requirement to the load factor than that whichapplies to manned structures, hence reflecting the difference inconsequence of failure between unmanned and manned struc-tures. Transformer stations are usually unmanned, but the eco-nomical consequences of a failure may be very large. Whensupport structures and foundations for transformer stations aredesigned according to DNV-OS-C101, it should therefore beconsidered whether it will be necessary from an economicalpoint of view to carry out the design based on the load factorrequirement for manned structures, even if the transformer sta-tions are unmanned.

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

104 The standard does not cover design of wind turbine com-ponents such as nacelle, rotor, generator and gear box. Forstructural design of rotor blades DNV-OS-J102 applies. Forstructural design of wind turbine components for which noDNV standard exists, the IEC61400-1 standard applies. 105 The tower, which usually extends from somewhereabove the water level to just below the nacelle, is considered apart of the support structure. The structural design of the toweris therefore covered by this standard, regardless of whether atype approval of the tower exists and is to be applied.

Guidance note:For a type-approved tower, the stiffnesses of the tower form partof the basis for the approval. It is important to make sure not tochange the weight and stiffness distributions over the height ofthe tower relative to those assumed for the type approval.


106 The standard has been written for general world-wideapplication. National and governmental regulations mayinclude requirements in excess of the provisions given by thisstandard depending on the size, type, location and intendedservice of the wind turbine structure.

Guidance note:An attempt has been made to harmonise DNV-OS-J101 with thecoming IEC61400-3 standard, in particular with respect to the

specification of load cases. For further information, reference ismade to the Committee Draft of IEC61400-3.


107 DNV-OS-J101 is applied as part of the basis for carryingout a DNV project certification of an offshore wind farm.

A 200 Objectives201 The standard specifies general principles and guidelinesfor the structural design of offshore wind turbine structures. 202 The objectives of this standard are to:

— provide an internationally acceptable level of safety bydefining minimum requirements for structures and struc-tural components (in combination with referenced stand-ards, recommended practices, guidelines, etc.)

— serve as a contractual reference document between suppli-ers and purchasers related to design, construction, installa-tion and in-service inspection

— serve as a guideline for designers, suppliers, purchasersand regulators

— specify procedures and requirements for offshore struc-tures subject to DNV certification

— serve as a basis for verification of offshore wind turbinestructures for which DNV is contracted to perform the ver-ification.

A 300 Scope and application301 The standard is applicable to all types of support struc-tures and foundations for offshore wind turbines. 302 The standard is applicable to the design of completestructures, including substructures and foundations, butexcluding wind turbine components such as nacelles androtors. 303 This standard gives requirements for the following:

— design principles— selection of material and extent of inspection— design loads— load effect analyses— load combinations— structural design— foundation design— corrosion protection.

A 400 Non-DNV codes401 In case of conflict between the requirements of thisstandard and a reference document other than DNV docu-ments, the requirements of this standard shall prevail. 402 The provision for using non-DNV codes or standards isthat the same safety level as the one resulting for designsaccording to this standard is obtained.403 Where reference in this standard is made to codes otherthan DNV documents, the valid revision of these codes shall betaken as the revision which was current at the date of issue ofthis standard, unless otherwise noted.404 When code checks are performed according to othercodes than DNV codes, the resistance and material factors asgiven in the respective codes shall be used.405 National and governmental regulations may override therequirements of this standard as applicable.

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – Sec.1 see note on front cover

B. Normative ReferencesB 100 General101 The standards in Table B1 include provisions, whichthrough reference in this text constitute provisions of thisstandard.

C. Informative ReferencesC 100 General101 The documents in Tables C1, C2 and C3 include accept-able methods for fulfilling the requirements in the standards.See also current DNV List of Publications. Other recognisedcodes or standards may be applied provided it is shown thatthey meet or exceed the level of safety of the actual standard.

Table B1 DNV Offshore Standards, Rules and Standards for CertificationReference TitleDNV-OS-B101 Metallic Materials

DNV-OS-C101 Design of Offshore Steel Structures, General (LRFD Method)

DNV-OS-C201 Structural Design of Offshore Units (WSD method)

DNV-OS-C401 Fabrication and Testing of Offshore StructuresDNV-OS-C502 Offshore Concrete StructuresDNV-OS-J102 Design and Manufacture of Wind Turbine

BladesRules for Planning and Execution of Marine OperationsRules for Fixed Offshore InstallationsRules for Classification of ShipsDNV Standard for Certification No. 2.22 Lifting Appliances

Table C1 DNV Offshore Standards for structural designReference TitleDNV-OS-C501 Composite Components

Table C2 DNV Recommended Practices and Classification NotesReference TitleDNV/Risø Guidelines for Design of Wind TurbinesDNV-RP-B401 Cathodic Protection DesignDNV-RP-C201 Buckling Strength of Plated StructuresDNV-RP-C202 Buckling Strength of ShellsDNV-RP-C203 Fatigue Strength Analysis of Offshore Steel

StructuresDNV-RP-C205 Environmental Conditions and Environmental

LoadsDNV-RP-C207 Statistical Representation of Soil DataClassification Notes 30.1 Buckling Strength Analysis

Classification Notes 30.4 Foundations

Classification Notes 30.6

Structural Reliability Analysis of Marine Struc-tures

Classification Notes 30.7 Fatigue Assessments of Ship Structures

Table C3 Other referencesReference TitleAISC LRFD Manual of Steel ConstructionAPI RP 2A LRFD Planning, Designing, and Constructing Fixed

Offshore Platforms – Load and Resistance Factor Design

API RP 2N Recommended Practice for Planning, Designing, and Constructing Structures and Pipelines for Arctic Conditions

BS 7910 Guide on methods for assessing the acceptabil-ity of flaws in fusion welded structures

BSH 7004 Standard Baugrunderkundung. Mindestenan-forderungen für Gründungen von Offshore-Windenergieanlagen.

DIN 4020 Geotechnische Untersuchungen für bautechnische Zwecke

DIN 4021 Baugrund; Aufschluss durch Schürfe und Bohrungen sowie Entnahme von Proben

DIN 4131 Stählerne AntennentragwerkeDIN 4133 Schornsteine aus StahlEN10025-1 Hot rolled products of non-alloy structural

steelsEN10025-2 Hot rolled products of structural steels. Techni-

cal delivery conditions for non-alloy structural steels

EN10025-3 Hot rolled products of structural steels. Techni-cal delivery conditions for normalized/normal-ized rolled weldable fine grain structural steels

EN 10204 Metallic products – types of inspection docu-ments

EN10225 Weldable structural steels for fixed offshore structures – technical delivery conditions

EN 13670-1 Execution of Concrete Structures – Part 1: Common rules

EN 1991-1-4 Eurocode 1: Actions on structures – Part 1-4: General actions – wind actions

EN 1992-1-1 Eurocode 2: Design of Concrete Structures EN 1993-1-1 Eurocode 3: Design of Steel Structures, Part 1-

1: General Rules and Rules for BuildingsENV 1993-1-6 Eurocode 3: Design of Steel Structures, Part 1-

6: General Rules – Supplementary Rules for the Shell Structures

ENV 1090-1 Execution of steel structures – Part 1: General rules and rules for buildings

ENV 1090-5 Execution of steel structures – Part 5: Supple-mentary rules for bridges

prEN50308 Wind Turbines – Labour SafetyIEC61400-1 Wind Turbines – Part 1: Design RequirementsIEC61400-3 Wind Turbines – Part 3: Design requirements

for offshore wind turbines, committee draftISO6934 Steel for the prestressing of concreteISO6935 Steel for the reinforcement of concreteISO 12944 Paints and varnishes – Corrosion protection of

steel structures by protective paint systemsISO 14688 Geotechnical investigations and testing – identi-

fication and classification of soil – Part 1: Iden-tification and description

ISO/IEC 17020 General criteria for the operation of various types of bodies performing inspections

ISO/IEC 17025 General requirements for the competence of calibration and testing laboratories

ISO 19900:2002 Petroleum and natural gas industries – Offshore structures – General requirements for offshore structures

ISO 19901-2 Seismic design procedures and criteria

ISO 19902 Petroleum and Natural Gas Industries – Fixed Steel Offshore Structures

DET NORSKE VERITAS


D. DefinitionsD 100 Verbal forms101 Shall: Indicates a mandatory requirement to be followedfor fulfilment or compliance with the present standard. Devia-tions are not permitted unless formally and rigorously justified,and accepted by all relevant contracting parties.102 Should: Indicates a recommendation that a certaincourse of action is preferred or is particularly suitable. Alterna-tive courses of action are allowable under the standard whereagreed between contracting parties, but shall be justified anddocumented.103 May: Indicates a permission, or an option, which is per-mitted as part of conformance with the standard.104 Can: Requirements with can are conditional and indi-cate a possibility to the user of the standard.105 Agreement, or by agreement: Unless otherwise indi-cated, agreed in writing between contractor and purchaser.

D 200 Terms201 Abnormal load: Wind load resulting from one of anumber of severe fault situations for the wind turbine, whichresult in activation of system protection functions. Abnormalwind loads are in general less likely to occur than loads fromany of the normal wind load cases considered for the ULS.202 Accidental Limit States (ALS): Ensure that the structureresists accidental loads and maintain integrity and performanceof the structure due to local damage or flooding.203 ALARP: As low as reasonably practicable; notation usedfor risk.204 Atmospheric zone: The external region exposed toatmospheric conditions.205 Cathodic protection: A technique to prevent corrosionof a steel surface by making the surface to be the cathode of anelectrochemical cell.206 Characteristic load: The reference value of a load to beused in the determination of the design load. The characteristicload is normally based upon a defined quantile in the upper tailof the distribution function for load.207 Characteristic load effect: The reference value of a loadeffect to be used in the determination of the design load effect.The characteristic load effect is normally based upon a definedquantile in the upper tail of the distribution function for loadeffect.208 Characteristic resistance: The reference value of astructural strength to be used in the determination of the designresistance. The characteristic resistance is normally basedupon a 5% quantile in the lower tail of the distribution functionfor resistance.

209 Characteristic material strength: The nominal value ofa material strength to be used in the determination of the designstrength. The characteristic material strength is normally basedupon a 5% quantile in the lower tail of the distribution functionfor material strength.210 Characteristic value: A representative value of a loadvariable or a resistance variable. For a load variable, it is a highbut measurable value with a prescribed probability of not beingunfavourably exceeded during some reference period. For aresistance variable it is a low but measurable value with a pre-scribed probability of being favourably exceeded. 211 Classification Notes: The classification notes coverproven technology and solutions which are found to representgood practice by DNV, and which represent one alternative forsatisfying the requirements stipulated in the DNV Rules orother codes and standards cited by DNV. The classificationnotes will in the same manner be applicable for fulfilling therequirements in the DNV offshore standards.212 Coating: Metallic, inorganic or organic material appliedto steel surfaces for prevention of corrosion.213 Co-directional: Wind and waves acting in the samedirection.214 Contractor: A party contractually appointed by the pur-chaser to fulfil all, or any of, the activities associated with fab-rication and testing.215 Corrosion allowance: Extra steel thickness that may rustaway during design life time.216 Current: A flow of water past a fixed point and usuallyrepresented by a velocity and a direction.217 Cut-in wind speed: Lowest mean wind speed at hubheight at which a wind turbine produces power.218 Cut-out wind speed: Highest mean wind speed at hubheight at which a wind turbine is designed to produce power.219 Design brief: An agreed document where owners’requirements in excess of this standard should be given.220 Design temperature: The lowest daily mean temperaturethat the structure may be exposed to during installation andoperation.221 Design value: The value to be used in the deterministicdesign procedure, i.e. characteristic value modified by theresistance factor or the load factor, whichever is applicable.222 Driving voltage: The difference between closed circuitanode potential and protection potential.223 Environmental state: Short term condition of typically10 minutes, 1 hour or 3 hours duration during which the inten-sities of environmental processes such as wave and wind proc-esses can be assumed to be constant, i.e. the processesthemselves are stationary.224 Expected loads and response history: Expected load andresponse history for a specified time period, taking intoaccount the number of load cycles and the resulting load levelsand response for each cycle.225 Expected value: The mean value, e.g. the mean value ofa load during a specified time period.226 Fatigue: Degradation of the material caused by cyclicloading.227 Fatigue critical: Structure with predicted fatigue lifenear the design fatigue life.228 Fatigue Limit States (FLS): Related to the possibility offailure due to the cumulative damage effect of cyclic loading.229 Foundation: The foundation of a support structure for awind turbine is in this document reckoned as a structural orgeotechnical component, or both, extending from the seabeddownwards.

ISO 20340 Paints and varnishes – Performance require-ments for protective paint systems for offshore and related structures

NACE TPC Publication No. 3. The role of bacteria in corro-sion of oil field equipment

NORSOK M-501 Surface preparation and protective coat-ing

NORSOK N-003 Actions and Action EffectsNORSOK N-004 Design of Steel StructuresNORSOK G-001 Marine Soil InvestigationsNS 3473:2003 Prosjektering av betongkonstruksjoner. Beregn-

ings- og konstruksjonsregler.NVN 11400-0 Wind turbines – Part 0: Criteria for type certifi-

cation – Technical Criteria

DET NORSKE VERITAS


230 Guidance note: Information in the standards in order toincrease the understanding of the requirements.231 Gust: Sudden and brief increase of the wind speed overits mean value.232 Highest astronomical tide (HAT): Level of high tidewhen all harmonic components causing the tide are in phase.233 Hindcast: A method using registered meteorologicaldata to reproduce environmental parameters. Mostly used forreproduction of wave data and wave parameters.234 Hub height: Height of centre of swept area of wind tur-bine rotor, measured from mean sea level.235 Idling: Condition of a wind turbine, which is rotatingslowly and not producing power.236 Independent organisations: Accredited or nationallyapproved certification bodies.237 Inspection: Activities such as measuring, examination,testing, gauging one or more characteristics of an object orservice and comparing the results with specified requirementsto determine conformity.238 Limit State: A state beyond which the structure nolonger satisfies the requirements. The following categories oflimit states are of relevance for structures: ULS = ultimatelimit state; FLS = fatigue limit state; ALS = accidental limitstate; SLS = serviceability limit state.239 Load effect: Effect of a single design load or combina-tion of loads on the equipment or system, such as stress, strain,deformation, displacement, motion, etc.240 Lowest astronomical tide (LAT): Level of low tide whenall harmonic components causing the tide are in phase.241 Lowest mean daily temperature: The lowest value of theannual mean daily temperature curve for the area in question.For seasonally restricted service the lowest value within thetime of operation applies.242 Lowest waterline: Typical light ballast waterline forships, transit waterline or inspection waterline for other typesof units. 243 Mean: Statistical mean over observation period.244 Mean water level (MWL): Mean still water level,defined as mean level between highest astronomical tide andlowest astronomical tide. 245 Mean zero-upcrossing period: Average period betweentwo consecutive zero-upcrossings of ocean waves in a seastate.246 Metocean: Abbreviation of meteorological and oceano-graphic.247 Non-destructive testing (NDT): Structural tests andinspection of welds by visual inspection, radiographic testing,ultrasonic testing, magnetic particle testing, penetrant testingand other non-destructive methods for revealing defects andirregularities.248 Object Standard: The standards listed in Table C1.249 Offshore Standard: The DNV offshore standards aredocuments which presents the principles and technical require-ments for design of offshore structures. The standards areoffered as DNV’s interpretation of engineering practice forgeneral use by the offshore industry for achieving safe struc-tures.250 Offshore wind turbine structure: A structural systemconsisting of a support structure for an offshore wind turbineand a foundation for the support structure.251 Omni-directional: Wind or waves acting in all direc-tions.252 Operating conditions: Conditions wherein a unit is onlocation for purposes of drilling or other similar operations,

and combined environmental and operational loadings arewithin the appropriate design limits established for such oper-ations. The unit may be either afloat or supported by the seabed, as applicable.253 Parking: The condition to which a wind turbine returnsafter a normal shutdown. Depending on the construction of thewind turbine, parking refers to the turbine being either in astand-still or an idling condition.254 Partial Safety Factor Method: Method for design whereuncertainties in loads are represented by a load factor anduncertainties in strengths are represented by a material factor.255 Pile head: The position along a foundation pile in levelwith the seabed. This definition applies regardless of whetherthe pile extends above the seabed.256 Pile length: Length along a pile from pile head to piletip.257 Pile penetration: Vertical distance from the seabed tothe pile tip.258 Potential: The voltage between a submerged metal sur-face and a reference electrode.259 Purchaser: The owner or another party acting on hisbehalf, who is responsible for procuring materials, componentsor services intended for the design, construction or modifica-tion of a structure.260 Qualified welding procedure specification (WPS): Awelding procedure specification, which has been qualified byconforming to one or more qualified WPQRs.261 Rated power: Quantity of power assigned, generally bya manufacturer, for a specified operating condition of a com-ponent, device or equipment. For a wind turbine, the ratedpower is the maximum continuous electrical power outputwhich a wind turbine is designed to achieve under normaloperating conditions.262 Rated wind speed: Minimum wind speed at hub heightat which a wind turbine’s rated power is achieved in the caseof a steady wind without turbulence.263 Recommended Practice (RP): The recommended prac-tice publications cover proven technology and solutions whichhave been found by DNV to represent good practice, andwhich represent one alternative for satisfying the requirementsstipulated in the DNV offshore standards or other codes andstandards cited by DNV.264 Redundancy: The ability of a component or system tomaintain or restore its function when a failure of a member orconnection has occurred. Redundancy can be achieved forinstance by strengthening or introducing alternative load paths.265 Reference electrode: Electrode with stable open-circuitpotential used as reference for potential measurements.266 Refraction: Process by which wave energy is redistrib-uted as a result of changes in the wave propagation velocitycaused by variations in the water depth.267 Reliability: The ability of a component or a system toperform its required function without failure during a specifiedtime interval.268 Residual currents: All other components of a currentthan tidal current.269 Risk: The qualitative or quantitative likelihood of anaccidental or unplanned event occurring considered in con-junction with the potential consequences of such a failure. Inquantitative terms, risk is the quantified probability of adefined failure mode times its quantified consequence.270 Rotor-nacelle assembly: Part of wind turbine carried bythe support structure.271 Scour zone: The external region of the unit which islocated at the seabed and which is exposed to scour.

DET NORSKE VERITAS


272 Serviceability Limit States (SLS): Imply deformations inexcess of tolerance without exceeding the load-carrying capac-ity, i.e., they correspond to tolerance criteria applicable to nor-mal use.273 Shakedown: A linear elastic structural behaviour isestablished after yielding of the material has occurred.274 Slamming: Impact load on an approximately horizontalmember from a rising water surface as a wave passes. Thedirection of the impact load is mainly vertical.275 Specified Minimum Yield Strength (SMYS): The mini-mum yield strength prescribed by the specification or standardunder which the material is purchased.276 Specified value: Minimum or maximum value duringthe period considered. This value may take into account oper-ational requirements, limitations and measures taken such thatthe required safety level is obtained.277 Splash zone: The external region of the unit which ismost frequently exposed to wave action.278 Standstill: The condition of a wind turbine generatorsystem that is stopped.279 Submerged zone: The part of the installation which isbelow the splash zone, including buried parts.280 Support structure: The support structure for an offshorewind turbine is in this document reckoned as the structurebetween the seabed and the nacelle of the wind turbine that thestructure supports. The foundation of the support structure,reckoned from the seabed downwards, is not included in thisdefinition, but is treated as a separate component. 281 Survival condition: A condition during which a unit maybe subjected to the most severe environmental loadings forwhich the unit is designed. Operation of the unit may havebeen discontinued due to the severity of the environmentalloadings. The unit may be either afloat or supported by the seabed, as applicable.282 Target safety level: A nominal acceptable probability ofstructural failure.283 Temporary condition: An operational condition thatmay be a design condition, for example the mating, transit orinstallation phases.284 Tensile strength: Minimum stress level where strainhardening is at maximum or at rupture.285 Tidal range: Distance between highest and lowest astro-nomical tide.286 Tide: Regular and predictable movements of the seagenerated by astronomical forces.287 Tower: Structural component, which forms a part of thesupport structure for a wind turbine, usually extending fromsomewhere above the still water level to just below the nacelleof the wind turbine.288 Transit conditions: All unit movements from one geo-graphical location to another.289 Turbulence intensity: Ratio between the standard devia-tion of the wind speed and the 10-minute mean wind speed.290 Ultimate Limit States (ULS): Correspond to the limit ofthe load-carrying capacity, i.e., to the maximum load-carryingresistance.291 Unidirectional: Wind and/or waves acting in one singledirection.292 Utilisation factor: The fraction of anode material thatcan be utilised for design purposes.293 Verification: Examination to confirm that an activity, aproduct or a service is in accordance with specified requirements.294 Welding procedure: A specified course of action to befollowed in making a weld, including reference to materials,

welding consumables, preparation, preheating (if necessary),method and control of welding and post-weld heat treatment (ifrelevant), and necessary equipment to be used.295 Welding procedure specification: A document provid-ing in detail the required variables of the welding procedure toensure repeatability.296 Welding procedure test: The making and testing of astandardised test piece, as indicated in the WPS, in order toqualify a welding procedure specification.297 Wind shear: Variation of wind speed across a plane per-pendicular to the wind direction.298 Wind shear law: Wind profile; mathematical expressionfor wind speed variation with height above sea surface.299 Yawing: Rotation of the rotor axis of a wind turbineabout a vertical axis.

E. Abbreviations and SymbolsE 100 Abbreviations101 Abbreviations as shown in Table E1 are used in thisstandard.

Table E1 AbbreviationsAbbreviation In fullAISC American Institute of Steel ConstructionALARP As Low As Reasonably PracticableALS Accidental Limit StateAPI American Petroleum InstituteBS British Standard (issued by British Standard Insti-

tute)BSH Bundesamt für Seeschifffahrt und HydrographieCN Classification NotesCTOD Crack Tip Opening DisplacementDDF Deep Draught FloatersDFF Design Fatigue FactorDNV Det Norske VeritasEHS Extra High StrengthFLS Fatigue Limit StateHAT Highest Astronomical TideHISC Hydrogen Induced Stress CrackingHS High Strength IEC International Electrotechnical CommissionISO International Organization for StandardisationLAT Lowest Astronomical TideMWL Mean Water LevelNACE National Association of Corrosion EngineersNDT Non-Destructive Testing NS Normal StrengthRP Recommended PracticeRHS Rectangular Hollow SectionRNA Rotor-Nacelle AssemblySCE Saturated Calomel ElectrodeSCF Stress Concentration FactorSLS Serviceability Limit StateSMYS Specified Minimum Yield StressSRB Sulphate Reducing BacteriaSWL Still Water LevelTLP Tension Leg PlatformULS Ultimate Limit StateWPS Welding Procedure SpecificationWSD Working Stress Design

DET NORSKE VERITAS


E 200 Symbols201 Latin characters

a0 connection areab full breadth of plate flangebe effective plate flange widthc detail shape factorc wave celerityd bolt diameterd water depthf frequencyf load distribution factorfice frequency of ice loadfn natural frequency of structurefr strength ratiofu nominal lowest ultimate tensile strengthfub ultimate tensile strength of boltfw strength ratiofy specified minimum yield stressg acceleration of gravityh heighth water depthh0 reference depth for wind-generated currenthD dynamic pressure head due to flow through pipeshpc vertical distance from the load point to the position of

max filling heightht threshold for wave heightk wave numberka correction factor for aspect ratio of plate fieldkm bending moment factorkpp fixation parameter for platekps fixation parameter for stiffenerskr correction factor for curvature perpendicular to the

stiffenersks hole clearance factorkt shear force factorl stiffener spanlo distance between points of zero bending momentsn numberp pressurepd design pressurep0 valve opening pressurer root facerc radius of curvaturerf flexural strength of icerlocal local ice pressureru compressive strength of ices distance between stiffenerst ice thicknesst0 net thickness of platetk corrosion additiontw throat thicknessvave annual average wind speed at hub heightvin cut-in wind speedvout cut-out wind speedvr rated wind speedvtide0 tidal current at still water level

vwind0 wind-driven current at still water levelz vertical distance from still water level, positive

upwardsz0 terrain roughness parameterA scale parameter in logarithmic wind speed profileAC Charnock’s constantAs net area in the threaded part of a boltAW wave amplitudeC weld factorCD drag coefficientCM mass coefficientCS slamming coefficientCe factor for effective plate flangeD deformation loadE modulus of elasticityE environmental loadE[·] mean valueF cumulative distribution functionF force, loadFd design loadFk characteristic loadFpd design preloading force in boltG permanent loadH heightHmax maximum wave heightH0 wave height in deep watersHRMS root mean squared wave heightHS significant wave heightIT turbulence intensityIref expected turbulence intensity, reference turbulence

intensityK frost indexKC Keulegan-Carpenter numberL length of crack in iceM momentMp plastic moment resistanceMy elastic moment resistanceN fatigue life, i.e. number of cycles to failureNp number of supported stiffeners on the girder spanNs number of stiffeners between considered section and

nearest supportP loadPpd average design point load from stiffenersQ variable functional loadR radiusR resistanceRd design resistanceRk characteristic resistanceS girder span as if simply supportedS power spectral densitySA response spectral accelerationSD response spectral displacementSV response spectral velocitySd design load effectSk characteristic load effectSZl lower limit of the splash zone

DET NORSKE VERITAS


202 Greek characters

203 Subscripts

F. Support Structure ConceptsF 100 Introduction101 Bottom-mounted support structures for large offshorewind farm developments fall into a number of generic typeswhich can be categorised by their nature and configuration,their method of installation, their structural configuration andthe selection of their construction materials. The options foroffshore support structures basically consist of:

— piled structures— gravity-based structures— skirt and bucket structures— moored floating structures.

The structural configuration of support structures can be cate-gorised into five basic types:

— monopile structures— tripod structures— lattice structures— gravity structures— floating structures.

Hybrid support structure designs may be utilised combiningthe features of the categorised structures.Water depth limits proposed for the different types of supportstructures in the following subsections are meant to be treatedas guidance rather than limitations.102 Monopile structures provide the benefit of simplicity infabrication and installation. Tripod and lattice structures areusually piled. Piled foundations by far forms the most commonform of offshore foundation. Piled offshore structures havebeen installed since the late 1940’es and have been installed inwater depth in excess of 150 metres. The standard method ofoffshore and near-shore marine installation of piled structuresis to lift or float the structure into position and then drive thepiles into the seabed using either steam or hydraulic poweredhammers. The handling of piles and hammers generallyrequires the use of a crane with sufficient capacity, ideally afloating crane vessel (revolving or shear leg crane). However,other types of offshore installation units are sometimes usedsuch as drilling jack-ups, specially constructed installationvessels or flat top barges mounted with a land based crawlercrane.103 Gravity foundations, unlike piled foundations, aredesigned with the objective of avoiding tensile loads (lifting)between the bottom of the support structure and the seabed.This is achieved by providing sufficient dead loads such thatthe structure maintains its stability in all environmental condi-tions solely by means of its own gravity. Gravity structures areusually competitive when the environmental loads are rela-tively modest and the “natural” dead load is significant orwhen additional ballast can relatively easily be provided at amodest cost. The ballast can be pumped-in sand, concrete, rock

SZu upper limit of the splash zoneT wave periodTP peak periodTR return periodTS sea state durationTZ zero-upcrossing periodU wind speed, instantaneous wind speedUhub wind speed at hub heightU0 1-hour mean wind speed U10 10-minute mean wind speedU10,hub 10-minute mean wind speed at hub heightUice velocity of ice floeV wind speedW steel with improved weldabilityZ steel grade with proven through thickness properties

with respect to lamellar tearing.

α angle between the stiffener web plane and the plane perpendicular to the plating

α exponent in power-law model for wind speed profileα coefficient in representation of wave loads according to

diffraction theoryα slope angle of seabedβw correlation factorδ deflectionΔσ stress rangeφ resistance factorγ spectral peak enhancement factorγf load factorγM material factorγMw material factor for weldsη ratio of fatigue utilisation, cumulative fatigue damage

ratioκ Von Karman’s constantλ wave lengthλ reduced slendernessθ rotation angleμ friction coefficientν Poisson’s ratioν spectral width parameterρ factor in crack length model for iceρ densityσd design stressσe elastic buckling stressσf flexural strength of iceσfw characteristic yield stress of weld depositσjd equivalent design stress for global in-plane membrane

stressσpd1 design bending stressσpd2 design bending stressσU standard deviation of wind speedτd design shear stressω angular frequencyξ coefficient in representation of wave loads according to

diffraction theoryψ wake amplification factor

ψ load combination factor, load reduction factorψ load factor for permanent load and variable functional

loadΦ standard normal cumulative distribution function.

c characteristic valued design value k characteristic value p plasticy yield.

DET NORSKE VERITAS


or iron ore. The additional ballast can partly be installed in thefabrication yard and partly at the final position; all dependingon the capacity of the construction yard, the available draftduring sea transport and the availability of ballast materials.The gravity based structure is especially suited where theinstallation of the support structure cannot be performed by aheavy lift vessel or other special offshore installation vessels,either because of non-availability or prohibitive costs of mobi-lising the vessels to the site.104 Floating structures can by their very nature be floatingdirectly in a fully commissioned condition from the fabricationand out-fitting yard to the site. Floating structures are espe-cially competitive at large water depths where the depth makesthe conventional bottom-supported structures non-competi-tive.

F 200 Gravity-based structures and gravity-pile struc-tures201 The gravity type support structure is a concrete basedstructure which can be constructed with or without small steelor concrete skirts. The ballast required to obtain sufficientgravity consists of sand, iron ore or rock that is filled into thebase of the support structure. The base width can be adjustedto suit the actual soil conditions. The proposed design includesa central steel or concrete shaft for transition to the wind tur-bine tower. The structure requires a flat base and will for alllocations require some form for scour protection, the extent ofwhich is to be determined during the detailed design.202 The gravity-pile support structure is very much like thegravity support structure. The structure can be filled with ironore or rock as required. The base width can be adjusted to suitthe actual soil conditions. The structure is designed such thatthe variable loads are shared between gravity and pile actions.203 These types of structures are well suited for sites withfirm soils and water depth ranging from 0 to 25 metres.

F 300 Jacket-monopile hybrids and tripods301 The jacket-monopile hybrid structure is a three-leggedjacked structure in the lower section, connected to a monopilein the upper part of the water column, all made of cylindricalsteel tubes. The base width and the pile penetration depth canbe adjusted to suit the actual soil conditions. 302 The tripod is a standard three-leg structure made ofcylindrical steel tubes. The central steel shaft of the tripodmakes the transition to the wind turbine tower. The tripod canhave either vertical or inclined pile sleeves. Inclined pilesleeves are used when the structure is to be installed with ajack-up drilling rig. The base width and pile penetration depthcan be adjusted to suit the actual environmental and soil con-ditions.303 These types of structures are well suited for sites withwater depth ranging from 20 to 50 metres.

F 400 Monopiles401 The monopile support structure is a simple design bywhich the tower is supported by the monopile, either directlyor through a transition piece, which is a transitional sectionbetween the tower and the monopile. The monopile continuesdown into the soil. The structure is made of cylindrical steeltubes. 402 The pile penetration depth can be adjusted to suit theactual environmental and soil conditions. The monopile isadvantageous in areas with movable seabed and scour. A pos-sible disadvantage is a too high flexibility in deep waters. Thelimiting condition of this type of support structure is the over-all deflection and vibration.403 This type of structure is well suited for sites with waterdepth ranging from 0 to 25 metres.

F 500 Supported monopiles and guyed towers501 The supported monopile structure is a standard mono-pile supported by two beams piled into the soil at a distancefrom the monopile. The structure is made of cylindrical steeltubes. The pile penetration of the supporting piles can beadjusted to suit the actual environmental and soil conditions. 502 The guyed tower support structure is a monotower con-nected to a double hinge near the seabed and allowed to movefreely. The tower is supported in four directions by guy wiresextending from the tower (above water level) to anchors in theseabed. The support structure installation requires use of smallto relatively large offshore vessels. Anchors including mudmats are installed. Guy wires are installed and secured to float-ers. Seabed support is installed and the tower is landed. Guywires are connected to tensioning system. Scour protection isinstalled as required.503 These types of structures are well suited for sites withwater depth ranging from 20 to 40 metres.

F 600 Tripods with buckets601 The tripod with buckets is a tripod structure equippedwith suction bucket anchors instead of piles as for the conven-tional tripod structure. The wind turbine support structure canbe transported afloat to the site. During installation, eachbucket can be emptied in a controlled manner, thus avoidingthe use of heavy lift equipment. Further, the use of the suctionbuckets eliminates the need for pile driving of piles as requiredfor the conventional tripod support structure.602 The support structure shall be installed at locations,which allow for the suction anchor to penetrate the prevalentsoils (sand or clay) and which are not prone to significantscour.603 This type of structure is well suited for sites with waterdepth ranging from 20 to 50 metres.

F 700 Suction buckets701 The suction bucket steel structure consists of a centrecolumn connected to a steel bucket through flange-reinforcedshear panels, which distribute the loads from the centre columnto the edge of the bucket. The wind turbine tower is connectedto the centre tubular above mean sea level. The steel bucketconsists of vertical steel skirts extending down from a horizon-tal base resting on the soil surface. 702 The bucket is installed by means of suction and will inthe permanent case behave as a gravity foundation, relying onthe weight of the soil encompassed by the steel bucket with askirt length of approximately the same dimension as the widthof the bucket.703 The stability is ensured because there is not enough timefor the bucket to be pulled from the bottom during a wave period.When the bucket is pulled from the soil during the passing of awave, a cavity will tend to develop between the soil surface andthe top of the bucket at the heel. However, the development ofsuch a cavity depends on water to flow in and fill up the cavityand thereby allow the bucket to be pulled up, but the typicalwave periods are too short to allow this to happen. The conceptallows for a simple decommissioning procedure.704 This type of structure is well suited for sites with waterdepth ranging from 0 to 25 metres.

F 800 Lattice towers801 The three-legged lattice tower consists of three cornerpiles interconnected with bracings. At the seabed pile sleevesare mounted to the corner piles. The soil piles are driven insidethe pile sleeves to the desired depth to gain adequate stabilityof the structure. 802 This type of structure is well suited for sites with waterdepth ranging from 20 to 40 metres.

DET NORSKE VERITAS


F 900 Low-roll floaters901 The low-roll floater is basically a floater kept in positionby mooring chains and anchors. In addition to keeping thefloater in place, the chains have the advantage that they con-tribute to dampen the motions of the floater. At the bottom ofthe hull of the floater, a stabiliser is placed to further reduceroll.902 The installation is simple since the structure can betowed to the site and then be connected by the chains to theanchors. The anchors can be fluke anchors, drag-in plateanchors and other plate anchors, suction anchors or pileanchors, depending on the actual seabed conditions. When theanchors have been installed, the chains can be installed andtightened and hook-up cables can be installed.903 This structure is a feasible solution in large water depths.

F 1000 Tension leg platforms1001 The tension leg support platform is a floater submergedby means of tensioned vertical anchor legs. The base structurehelps dampen the motions of the structural system. The instal-lation is simple since the structure can be towed to the site andthen be connected to the anchors. When anchors such asanchor piles have been installed and steel legs have been put inplace, the hook-up cable can be installed. The platform is sub-sequently lowered by use of ballast tanks and/or tension sys-tems.1002 The entire structure can be disconnected from the ten-sion legs and floated to shore in case of major maintenance orrepair of the wind turbine.1003 This structure is a feasible solution in large waterdepths.

DET NORSKE VERITAS


SECTION 2DESIGN PRINCIPLES

A. IntroductionA 100 General101 This section describes design principles and designmethods for structural design, including:

— design by partial safety factor method with linear combi-nation of loads or load effects

— design by partial safety factor method with direct simula-tion of combined load effect of simultaneous load proc-esses

— design assisted by testing— probability-based design.

102 General design considerations regardless of designmethod are also given in B101. 103 This standard is based on the partial safety factormethod, which is based on separate assessment of the loadeffect in the structure due to each applied load process. Thestandard allows for design by direct simulation of the com-bined load effect of simultaneously applied load processes,which is useful in cases where it is not feasible to carry out sep-arate assessments of the different individual process-specificload effects.104 As an alternative or as a supplement to analytical meth-ods, determination of load effects or resistance may in somecases be based either on testing or on observation of structuralperformance of models or full-scale structures.105 Structural reliability analysis methods for direct proba-bility-based design are mainly considered as applicable to spe-cial case design problems, to calibrate the load and materialfactors to be used in the partial safety factor method, and todesign for conditions where limited experience exists.

A 200 Aim of the design201 Structures and structural elements shall be designed to:

— sustain loads liable to occur during all temporary, operat-ing and damaged conditions if required

— ensure acceptable safety of structure during the design lifeof the structure

— maintain acceptable safety for personnel and environment— have adequate durability against deterioration during the

design life of the structure.

B. General Design ConditionsB 100 General101 The design of a structural system, its components anddetails shall, as far as possible, satisfy the following require-ments:

— resistance against relevant mechanical, physical andchemical deterioration is achieved

— fabrication and construction comply with relevant, recog-nised techniques and practice

— inspection, maintenance and repair are possible.

102 Structures and structural components shall possess duc-tile resistance unless the specified purpose requires otherwise.103 Structural connections are, in general, to be designedwith the aim to minimise stress concentrations and reducecomplex stress flow patterns.

104 As far as possible, transmission of high tensile stressesthrough the thickness of plates during welding, block assemblyand operation shall be avoided. In cases where transmission ofhigh tensile stresses through the thickness occurs, structuralmaterial with proven through-thickness properties shall beused. Object standards may give examples where to use plateswith proven through thickness properties.105 Structural elements may be manufactured according tothe requirements given in DNV-OS-C401.

C. Safety Classes and Target Safety LevelC 100 Safety classes101 In this standard, structural safety is ensured by use of asafety class methodology. The structure to be designed is clas-sified into a safety class based on the failure consequences.The classification is normally determined by the purpose of thestructure. For each safety class, a target safety level is definedin terms of a nominal annual probability of failure. 102 For structures in offshore wind farms, three safetyclasses are considered. Low safety class is used for structures,whose failures imply low risk for personal injuries and pollu-tion, low risk for economical consequences and negligible riskto human life. Normal safety class is used for structures, whosefailures imply some risk for personal injuries, pollution orminor societal losses, or possibility of significant economicconsequences. High safety class is used for structures, whosefailures imply large possibilities for personal injuries or fatali-ties, for significant pollution or major societal losses, or verylarge economic consequences.

Guidance note:Support structures and foundations for wind turbines, which arenormally unmanned, are usually to be designed to the normalsafety class. Also support structures and foundations for meteor-ological measuring masts are usually to be designed to the nor-mal safety class. Note, however, that the possibility of designing these supportstructures and foundations to a different safety class than the nor-mal safety class should always be considered, based on econom-ical motivations and considerations about human safety.For example, the design of a meteorological measuring mast fora large wind farm may need to be carried out to the high safetyclass, because a loss of the mast may cause a delay in the com-pletion of the wind farm or it may imply overdesign of the tur-bines and support structures in the wind farm owing to theimplied incomplete knowledge of the wind. The costs associatedwith the loss of such a mast may well exceed the costs associatedwith the loss of a turbine and thereby call for design to the highsafety class.Also, in order to protect the investments in a wind farm, it maybe wise to design the support structures and foundations for thewind turbines to high safety class.


103 In this standard, the different safety classes applicablefor different types of structures are reflected in differentrequirements to load factors. The requirements to material fac-tors remain unchanged regardless of which safety class isapplicable for a structure in question.

C 200 Target safety201 The target safety level for structural design of supportstructures and foundations for wind turbines to the normalsafety class according to this standard is a nominal annual

DET NORSKE VERITAS


probability of failure of 10–4. This target safety is the levelaimed at for structures, whose failures are ductile, and whichhave some reserve capacity.

Guidance note:The target safety level of 10–4 represents DNV's interpretation ofthe safety level inherent in the normal safety class for wind tur-bines defined in IEC61400-1.The target safety level of 10–4 is compatible with the safety levelimplied by DNV-OS-C101 for unmanned structures.This reflects that wind turbines and wind turbine structuresdesigned to normal safety class according to this standard areunmanned structures. For wind turbines where personnel areplanned to be present during severe loading conditions, design tohigh safety class with a nominal annual probability of failure of10–5 is warranted. Structural components and details should be shaped such that thestructure as far as possible will behave in the presumed ductilemanner. Connections should be designed with smooth transitionsand proper alignment of elements. Stress concentrations shouldbe avoided as far as possible. A structure or a structural compo-nent may behave as brittle, even if it is made of ductile materials,for example when there are sudden changes in section properties.


202 The target safety level is the same, regardless of whichdesign philosophy is applied.

Guidance note:A design of a structural component which is based on an assump-tion of inspections and possible maintenance and repair through-out its design life may benefit from a reduced structuraldimension, e.g. a reduced cross-sectional area, compared to thatof a design without such an inspection and maintenance plan, inorder to achieve the same safety level for the two designs.This refers in particular to designs which are governed by theFLS or the SLS. It may be difficult to apply this to designs whichare governed by the ULS or the ALS.


D. Limit StatesD 100 General101 A limit state is a condition beyond which a structure orstructural component will no longer satisfy the design require-ments. 102 The following limit states are considered in this stand-ard:Ultimate limit states (ULS) correspond to the maximum load-carrying resistanceFatigue limit states (FLS) correspond to failure due to theeffect of cyclic loadingAccidental limit state (ALS) correspond to damage to compo-nents due to an accidental event or operational failureServiceability limit states (SLS) correspond to tolerance crite-ria applicable to normal use.103 Examples of limit states within each category:Ultimate limit states (ULS)

— loss of structural resistance (excessive yielding and buck-ling)

— failure of components due to brittle fracture— loss of static equilibrium of the structure, or of a part of the

structure, considered as a rigid body, e.g. overturning orcapsizing

— failure of critical components of the structure caused byexceeding the ultimate resistance (which in some cases isreduced due to repetitive loading) or the ultimate deforma-tion of the components

— transformation of the structure into a mechanism (collapseor excessive deformation).

Fatigue limit states (FLS)

— cumulative damage due to repeated loads.

Accidental limit states (ALS)

— accidental conditions such as structural damage caused byaccidental loads and resistance of damaged structures.

Serviceability limit states (SLS)

— deflections that may alter the effect of the acting forces— deformations that may change the distribution of loads

between supported rigid objects and the supporting struc-ture

— excessive vibrations producing discomfort or affectingnon-structural components

— motions that exceed the limitation of equipment— differential settlements of foundations soils causing intol-

erable tilt of the wind turbine— temperature-induced deformations.

E. Design by the Partial Safety Factor MethodE 100 General101 The partial safety factor method is a design method bywhich the target safety level is obtained as closely as possibleby applying load and resistance factors to characteristic valuesof the governing variables and subsequently fulfilling a speci-fied design criterion expressed in terms of these factors andthese characteristic values. The governing variables consist of

— loads acting on the structure or load effects in the structure— resistance of the structure or strength of the materials in

the structure.

102 The characteristic values of loads and resistance, or ofload effects and material strengths, are chosen as specificquantiles in their respective probability distributions. Therequirements to the load and resistance factors are set such thatpossible unfavourable realisations of loads and resistance, aswell as their possible simultaneous occurrences, are accountedfor to an extent which ensures that a satisfactory safety level isachieved.

E 200 The partial safety factor format201 The safety level of a structure or a structural componentis considered to be satisfactory when the design load effect Sddoes not exceed the design resistance Rd:

This is the design criterion. The design criterion is also knownas the design inequality. The corresponding equation Sd = Rdforms the design equation.202 There are two approaches to establish the design loadeffect Sdi associated with a particular load Fi:(1) The design load effect Sdi is obtained by multiplication ofthe characteristic load effect Ski by a specified load factor γfi

where the characteristic load effect Ski is determined in a struc-tural analysis for the characteristic load Fki.(2) The design load effect Sdi is obtained from a structuralanalysis for the design load Fdi, where the design load Fdi isobtained by multiplication of the characteristic load Fki by a

dd RS ≤

kifidi SS γ=

DET NORSKE VERITAS


specified load factor γfi

Approach (1) shall be used to determine the design load effectwhen a proper representation of the dynamic response is theprime concern, whereas Approach (2) shall be used if a properrepresentation of nonlinear material behaviour or geometricalnonlinearities or both are the prime concern. Approach (1) typ-ically applies to the determination of design load effects in thesupport structure, including the tower, from the wind loadingon the turbine, whereas Approach (2) typically applies to thedesign of the support structure and foundation with the loadeffects in the tower applied as a boundary condition.

Guidance note:For structural design of monopiles and other piled structures,Approach (2) can be used to properly account for the influencefrom the nonlinearities of the soil. In a typical design situation fora monopile, the main loads will be wind loads and wave loads inaddition to permanent loads. The design combined wind andwave load effects at an appropriate interface level, such as thetower flange, can be determined from an integrated structuralanalysis of the tower and support structure by Approach (1) andconsist of a shear force in combination with a bending moment.These design load effects can then be applied as external designloads at the chosen interface level, and the design load effects inthe monopile structure and foundation pile for these design loadscan then be determined from a structural analysis of the monopilestructure and foundation pile by Approach (2).


203 The design load effect Sd is the most unfavourable com-bined load effect resulting from the simultaneous occurrenceof n loads Fi, i = 1,...n. It may be expressed as

where f denotes a functional relationship.According to the partial safety factor format, the design com-bined load effect Sd resulting from the occurrence of n inde-pendent loads Fi, i = 1,...n, can be taken as

where Sdi(Fki) denotes the design load effect corresponding tothe characteristic load Fki.When there is a linear relationship between the load Fi actingon the structure and its associated load effect Si in the structure,the design combined load effect Sd resulting from the simulta-neous occurrence of n loads Fi, i = 1,…n, can be achieved as

Guidance note:As an example, the combined load effect could be the bendingstress in a vertical foundation pile, resulting from a wind load anda wave load that act concurrently on a structure supported by thepile.


When there is a linear relationship between the load Fi and itsload effect Si, the characteristic combined load effect Sk result-ing from the simultaneous occurrence of n loads Fi, i = 1,…n,can be achieved as

204 Characteristic load effect values Ski are obtained asspecific quantiles in the distributions of the respective loadeffects Si. In the same manner, characteristic load values Fkiare obtained as specific quantiles in the distributions of therespective loads Fi.

Guidance note:Which quantiles are specified as characteristic values maydepend on which limit state is considered. Which quantiles arespecified as characteristic values may also vary from one speci-fied combination of load effects to another. For further details seeSec.4F.


205 In this standard, design in the ULS is either based on acharacteristic combined load effect Sk defined as the 98%quantile in the distribution of the annual maximum combinedload effect, or on a characteristic load Fk defined as the 98%quantile in the distribution of the annual maximum of the com-bined load. The result is a combined load or a combined loadeffect whose return period is 50 years.

Guidance note:When n load processes occur simultaneously, the standard spec-ifies more than one set of characteristic load effects (Sk1,...Skn)to be considered in order for the characteristic combined loadeffect Sk to come out as close as possible to the 98% quantile. Foreach specified set (Sk1,...Skn), the corresponding design com-bined load effect is determined according to item 203. For use indesign, the design combined load effect Sd is selected as the mostunfavourable value among the design combined load effects thatresult for these specified sets of characteristic load effects.


206 When the structure is subjected to the simultaneousoccurrence of n load processes, and the structural behaviour,e.g. the damping, is influenced by the character of at least oneof these loads, then it may not always be feasible to determinethe design load effect Sd, resulting from the simultaneousoccurrence of the n loads, by a linear combination of separatelydetermined individual load effects as set forth in 203. Withinthe framework of the partial safety factor method, the designcombined load effect Sd, resulting from the simultaneousoccurrence of the n loads, may then be established as a charac-teristic combined load effect Sk multiplied by a common loadfactor γf. The characteristic combined load effect Sk will in thiscase need to be defined as a quantile in the upper tail of the dis-tribution of the combined load effect that results in the struc-ture from the simultaneous occurrence of the n loads. Inprinciple, the distribution of this combined load effect comesabout from a structural analysis in which the n respective loadprocesses are applied simultaneously.

Guidance note:The total damping of a wind turbine depends on the wind loadingand its direction relative to other loads, such that for example thewave load effect in the support structure becomes dependent onthe characteristics of the wind loading. Unless the wind loadcharacteristics can be properly accounted for to produce a correcttotal damping and a correct separate wave load effect in a struc-tural analysis for the wave load, then the structure may need to beanalysed for the sought-after combined load effect for a simulta-neous application of the wind load process and the wave loadprocess.


207 The resistance R against a particular load effect S is, ingeneral, a function of parameters such as geometry, materialproperties, environment, and load effects themselves, the latterthrough interaction effects such as degradation.208 There are two approaches to establish the design resist-ance Rd of the structure or structural component:

kifidi FF γ=

),...( 1 dndd FFfS =

∑=

=n

ikidid FSS

1)(

∑=

=n

ikifid SS

1γ

∑=

=n

ikik SS

1

DET NORSKE VERITAS


1) The design resistance Rd is obtained by dividing the char-acteristic resistance Rk by a specified material factor γm:

2) The design resistance Rd is obtained from the design mate-rial strength σd by a capacity analysis

in which R denotes the functional relationship betweenmaterial strength and resistance and in which the designmaterial strength σd is obtained by dividing the character-istic material strength σk by a material factor γm,

Which of the two approaches applies depends on the design sit-uation. In this standard, the approach to be applied is specifiedfrom case to case.209 The characteristic resistance Rk is obtained as a specificquantile in the distribution of the resistance. It may be obtainedby testing, or it may be calculated from the characteristic val-ues of the parameters that govern the resistance. In the lattercase, the functional relationship between the resistance and thegoverning parameters is applied. Likewise, the characteristicmaterial strength σk is obtained as a specific quantile in theprobability distribution of the material strength and may beobtained by testing.210 Load factors account for:

— possible unfavourable deviations of the loads from theircharacteristic values

— the limited probability that different loads exceed theirrespective characteristic values simultaneously

— uncertainties in the model and analysis used for determi-nation of load effects.

211 Material factors account for:

— possible unfavourable deviations in the resistance of mate-rials from the characteristic value

— uncertainties in the model and analysis used for determi-nation of resistance

— a possibly lower characteristic resistance of the materialsin the structure, as a whole, as compared with the charac-teristic values interpreted from test specimens.

E 300 Characteristic load effect301 For operational design conditions, the characteristicvalue Sk of the load effect resulting from an applied load com-bination is defined as follows, depending on the limit state:

— For load combinations relevant for design against theULS, the characteristic value of the resulting load effect isdefined as a load effect with an annual probability ofexceedance equal to or less than 0.02, i.e. a load effectwhose return period is at least 50 years.

— For load combinations relevant for design against the FLS,the characteristic load effect history is defined as theexpected load effect history.

— For load combinations relevant for design against the SLS,the characteristic load effect is a specified value, depend-ent on operational requirements.

Load combinations to arrive at the characteristic value Sk ofthe resulting load effect are given in Sec.4.302 For temporary design conditions, the characteristicvalue Sk of the load effect resulting from an applied load com-bination is a specified value, which shall be selected dependenton the measures taken to achieve the required safety level. Thevalue shall be specified with due attention to the actual loca-

tion, the season of the year, the duration of the temporary con-dition, the weather forecast, and the consequences of failure.

E 400 Characteristic resistance401 The characteristic resistance is defined as the 5% quan-tile in the distribution of the resistance.

E 500 Load and resistance factors 501 Load and resistance factors for the various limit statesare given in Sec.5.

F. Design by Direct Simulation of Combined Load Effect of Simultaneous Load Processes

F 100 General101 Design by direct simulation of the combined load effectof simultaneously acting load processes is similar to design bythe partial safety factor method, except that it is based on adirect simulation of the characteristic combined load effectfrom the simultaneously applied load processes in stead ofbeing based on a linear combination of individual characteris-tic load effects determined separately for each of the appliedload processes. 102 For design of wind turbine structures which are sub-jected to two or more simultaneously acting load processes,design by direct simulation of the combined load effect mayprove an attractive alternative to design by the linear load com-bination model of the partial safety factor method. The linearcombination model of the partial safety factor method may beinadequate in cases where the load effect associated with oneof the applied load processes depends on structural propertieswhich are sensitive to the characteristics of one or more of theother load processes.

Guidance note:The aerodynamic damping of a wind turbine depends on whetherthere is wind or not, whether the turbine is in power productionor at stand-still, and whether the wind is aligned or misalignedwith other loads such as wave loads. Unless correct assumptionscan be made about the aerodynamic damping of the wind turbinein accordance with the actual status of the wind loading regime,separate determination of the load effect due to wave load aloneto be used with the partial safety factor format may not be feasi-ble. In a structural time domain analysis of the turbine subjected con-currently to both wind and wave loading, the aerodynamic damp-ing of the turbine will come out right since the wind loading isincluded, and the resulting combined load effect, usuallyobtained by simulations in the time domain, form the basis forinterpretation of the characteristic combined load effect.


F 200 Design format201 For design of wind turbine structures which are sub-jected to two or more simultaneously acting load processes, thedesign inequality

applies. The design combined load effect Sd is obtained bymultiplication of the characteristic combined load effect Sk bya specified load factor γf,

F 300 Characteristic load effect301 The characteristic combined load effect Sk may beestablished directly from the distribution of the annual maxi-

m

kd

RR

γ=

)( dd RR σ=

m

kd γ

σσ =

dd RS ≤

kfd SS γ=

DET NORSKE VERITAS


mum combined load effect that results from a structural analy-sis, which is based on simultaneous application of the two ormore load processes. In the case of ULS design, the character-istic combined load effect Sk shall be taken as the 98% quantilein the distribution of the annual maximum combined loadeffect, i.e. the combined load effect whose return period is 50years.

Guidance note:There may be several ways in which the 98% quantile in the dis-tribution of the annual maximum combined load effect can bedetermined. Regardless of the approach, a global structural anal-ysis model must be established, e.g. in the form of a beam-ele-ment based frame model, to which loads from severalsimultaneously acting load processes can be applied.A structural analysis in the time domain is usually carried out fora specified environmental state of duration typically 10 minutesor one or 3 hours, during which period of time stationary condi-tions are assumed with constant intensities of the involved loadprocesses. The input consists of concurrent time series of therespective load processes, e.g. wind load and wave load, withspecified directions. The output consists of time series of loadeffects in specified points in the structure.In principle, determination of the 98% quantile in the distributionof the annual maximum load effect requires structural analyses tobe carried out for a large number of environmental states, viz. allthose states that contribute to the upper tail of the distribution ofthe annual maximum load effect. Once the upper tail of this dis-tribution has been determined by integration over the results forthe various environmental states, weighted according to their fre-quencies of occurrence, the 98% quantile in the distribution canbe interpreted.The computational efforts can be considerably reduced when itcan be assumed that the 98% quantile in the distribution of theannual maximum load effect can be estimated by the expectedvalue of the maximum load effect in the environmental statewhose return period is 50 years.Further guidance on how to determine the 98% quantile in thedistribution of the annual maximum load effect is provided inSec.4.


F 400 Characteristic resistance401 The characteristic resistance is to be calculated as for thepartial safety factor method.

G. Design Assisted by Testing

G 100 General101 Design by testing or observation of performance is ingeneral to be supported by analytical design methods.102 Load effects, structural resistance and resistance againstmaterial degradation may be established by means of testing orobservation of the actual performance of full-scale structures.103 To the extent that testing is used for design, the testingshall be verifiable.

G 200 Full-scale testing and observation of performance of existing structures201 Full-scale tests or monitoring on existing structures maybe used to give information on response and load effects to beutilised in calibration and updating of the safety level of thestructure.

H. Probability-based DesignH 100 Definition101 The structural reliability, or the structural safety, isdefined as the probability that failure will not occur or that aspecified failure criterion will not be met within a specifiedperiod of time.

H 200 General201 This section gives requirements for structural reliabilityanalysis undertaken in order to document compliance with theoffshore standards.202 Acceptable procedures for structural reliability analysesare documented in Classification Notes No. 30.6.203 Reliability analyses shall be based on Level 3 reliabilitymethods. These methods utilise probability of failure as ameasure of safety and require knowledge of the probabilitydistribution of all governing load and resistance variables.204 In this standard, Level 3 reliability methods are mainlyconsidered applicable to:

— calibration of a Level 1 method to account for improvedknowledge. (Level 1 methods are deterministic analysismethods that use only one characteristic value to describeeach uncertain variable, i.e. the partial safety factormethod applied in the standards)

— special case design problems— novel designs for which limited or no experience exists.

205 Reliability analysis may be updated by utilisation of newinformation. Where such updating indicates that the assump-tions upon which the original analysis was based are not valid,and the result of such non-validation is deemed to be essentialto safety, the subject approval may be revoked.206 Target reliabilities shall be commensurate with the con-sequence of failure. The method of establishing such targetreliabilities, and the values of the target reliabilities them-selves, should be agreed in each separate case. To the extentpossible, the minimum target reliabilities shall be based onestablished cases that are known to have adequate safety.207 Where well established cases do not exist, e.g. in thecase of novel and unique design solutions; the minimum targetreliability values shall be based upon one or a combination ofthe following considerations:

— transferable target reliabilities for similar existing designsolutions

— internationally recognised codes and standards— Classification Notes No. 30.6.

DET NORSKE VERITAS


SECTION 3SITE CONDITIONS

A. GeneralA 100 Definition101 Site conditions consist of all site-specific conditionswhich may influence the design of a wind turbine structure bygoverning its loading, its capacity or both. 102 Site conditions cover virtually all environmental condi-tions on the site, including but not limited to meteorologicalconditions, oceanographic conditions, soil conditions, seismic-ity, biology, and various human activities.

Guidance note:The meteorological and oceanographic conditions which mayinfluence the design of a wind turbine structure consist of phe-nomena such as wind, waves, current and water level. These phe-nomena may be mutually dependent and for the three first ofthem the respective directions are part of the conditions that maygovern the design.Micro-siting of the wind turbines within a wind farm requiresthat local wake effects from adjacent wind turbines be consideredpart of the site conditions at each individual wind turbine struc-ture in the farm.


B. Wind ClimateB 100 Wind conditions101 For representation of wind climate, a distinction is madebetween normal wind conditions and extreme wind conditions.The normal wind conditions generally concern recurrent struc-tural loading conditions, while the extreme wind conditionsrepresent rare external design conditions. Normal wind condi-tions are used as basis for determination of primarily fatigueloads, but also extreme loads from extrapolation of normaloperation loads. Extreme wind conditions are wind conditionsthat can lead to extreme loads in the components of the windturbine and in the support structure and the foundation. 102 The normal wind conditions are specified in terms of anair density, a long-term distribution of the 10-minute meanwind speed, a wind shear in terms of a gradient in the meanwind speed with respect to height above the sea surface, andturbulence.103 The extreme wind conditions are specified in terms of anair density in conjunction with prescribed wind events. Theextreme wind conditions include wind shear events, as well aspeak wind speeds due to storms, extreme turbulence, and rapidextreme changes in wind speed and direction. 104 The normal wind conditions and the extreme wind con-ditions shall be taken in accordance with IEC61400-1.

Guidance note:The normal wind conditions and the extreme wind conditions,specified in IEC61400-1 and used herein, may be insufficient forrepresentation of special conditions experienced in tropicalstorms such as hurricanes, cyclones and typhoons.


B 200 Parameters for normal wind conditions201 The wind climate is represented by the 10-minute meanwind speed U10 and the standard deviation σU of the windspeed. In the short term, i.e. over a 10-minute period, station-ary wind conditions with constant U10 and constant σU areassumed to prevail.

Guidance note:The 10-minute mean wind speed U10 is a measure of the intensityof the wind. The standard deviation σU is a measure of the vari-ability of the wind speed about the mean.


202 The arbitrary wind speed under stationary 10-minuteconditions in the short term follows a probability distributionwhose mean value is U10 and whose standard deviation is σU. 203 The turbulence intensity is defined as the ratio σU/U10.204 The short term 10-minute stationary wind climate maybe represented by a wind spectrum, i.e. the power spectral den-sity function of the wind speed process, S(f). S(f) is a functionof U10 and σU and expresses how the energy of the wind speedis distributed between various frequencies.

B 300 Wind data301 Wind speed statistics are to be used as a basis for repre-sentation of the long-term and short-term wind conditions.Empirical statistical data used as a basis for design must covera sufficiently long period of time.

Guidance note:Site-specific measured wind data over sufficiently long periodswith minimum or no gaps are to be sought.


302 Wind speed data are height-dependent. The mean windspeed at the hub height of the wind turbine shall be used as areference. When wind speed data for other heights than the ref-erence height are not available, the wind speeds in theseheights can be calculated from the wind speeds in the referenceheight in conjunction with a wind speed profile above the stillwater level. 303 The long-term distributions of U10 and σU should pref-erably be based on statistical data for the same averagingperiod for the wind speed as the averaging period which is usedfor the determination of loads. If a different averaging periodthan 10 minutes is used for the determination of loads, the winddata may be converted by application of appropriate gust fac-tors. The short-term distribution of the arbitrary wind speeditself is conditional on U10 and σU.

Guidance note:An appropriate gust factor to convert wind statistics from otheraveraging periods than 10 minutes depends on the frequencylocation of a spectral gap, when such a gap is present. Applica-tion of a fixed gust factor, which is independent of the frequencylocation of a spectral gap, can lead to erroneous results.The latest insights for wind profiles above water should be con-sidered for conversion of wind speed data between different ref-erence heights or different averaging periods.Unless data indicate otherwise, the following expression can beused for calculation of the mean wind speed U with averagingperiod T at height z above sea level as

where h = 10 m and T10 = 10 minutes, and where U10 is the 10-minute mean wind speed at height h. This expression convertsmean wind speeds between different averaging periods. WhenT < T10, the expression provides the most likely largest meanwind speed over the specified averaging period T, given the orig-inal 10-minute averaging period with stationary conditions andgiven the specified 10-minute mean wind speed U10. The conver-sion does not preserve the return period associated with U10. The

)ln047.0ln137.01(),(10

10 TT

hzUzTU −+⋅=

DET NORSKE VERITAS


expression originates from the NORSOK standard and is repre-sentative for North Sea conditions.For extreme mean wind speeds corresponding to specified returnperiods in excess of approximately 50 years, the followingexpression can be used for conversion of the one-hour mean windspeed U0 at height h above sea level to the mean wind speed Uwith averaging period T at height z above sea level

where h = 10 m, T0 = 1 hour and T < T0 and where

and

and where U will have the same return period as U0.This conversion expression is recognised as the Frøya wind pro-file. More details can be found in DNV-RP-C205.Both conversion expressions are based on data from North Seaand Norwegian Sea locations and may not necessarily lend them-selves for use at other offshore locations. The expressions shouldnot be extrapolated for use beyond the height range for whichthey are calibrated, i.e. they should not be used for heights aboveapproximately 100 m. Possible influences from geostrophicwinds down to about 100 m height emphasises the importance ofobserving this restriction.Both expressions are based on the application of a logarithmicwind profile. For locations where an exponential wind profile isused or prescribed, the expressions should be considered usedonly for conversions between different averaging periods at aheight z equal to the reference height h = 10 m.


304 Empirical statistical wind data used as a basis for designmust cover a sufficiently long period of time.

Guidance note:Wind speed data for the long-term determination of the 10-minute mean wind speed U10 are usually available for power out-put prediction. Turbulence data are usually more difficult toestablish, in particular because of wake effects from adjacentoperating wind turbines. The latest insights for wind profileswithin wind farms should be considered.


305 The wind velocity at the location of the structure shall beestablished on the basis of previous measurements at the actuallocation and adjacent locations, hindcast predictions as well astheoretical models and other meteorological information. If thewind velocity is of significant importance to the design andexisting wind data are scarce and uncertain, wind velocitymeasurements should be carried out at the location in question.306 Characteristic values of the wind velocity should bedetermined with due account of the inherent uncertainties.307 Characteristic values of the wind velocity shall be deter-mined with due account of wake effects owing to the presenceof other wind turbines upstream, such as in a wind farm.

Guidance note:A wind farm generates its own wind climate due to downstreamwake effects, and the wind climate in the centre of the wind farmmay therefore be very different from the ambient wind climate.The layout of the wind farm has an impact on the wind at the indi-vidual wind turbines. Wake effects in a wind farm will in generalimply a considerably increased turbulence, reflected in anincreased standard deviation σU of the wind speed. This effect

may be significant even when the spacing between the wind tur-bines in the wind farm is as large as 8 to 10 rotor diameters. Wakeeffects in a wind farm may also imply a reduction in the 10-minute mean wind speed U10 relative to that of the ambient windclimate.Wake effects in wind farms will often dominate the fatigue loadsin offshore wind turbine structures.Wake effects fade out more slowly and over longer distances off-shore than they do over land.For assessment of wake effects in wind farms, the effects ofchanged wind turbine positions within specified installation tol-erances for the wind turbines relative to their planned positionsshould be evaluated.


308 Wind speed data are usually specified for a specific ref-erence temperature. When wind speed data are used for struc-tural design, it is important to be aware of this referencetemperature, in particular with a view to the operation philos-ophy adopted for the wind turbine design and the temperatureassumptions made in this context.

Guidance note:The wind load on a wind turbine tower is induced by the windpressure which depends both on density and wind speed. Thewind load on the rotor does not depend on the wind pressurealone but also on stall characteristics of the blade profile andactive control of the blade pitch and the rotor speed. Design loadsin type certification normally refer to an air density of 1.225 kg/m3. Project specific design loads shall address the air densityobserved with the wind speed measurements in a rational man-ner. The air density can increase by up to 10% in arctic areas dur-ing the winter season.


B 400 Wind modelling401 The spectral density of the wind speed process expresseshow the energy of the wind turbulence is distributed betweendifferent frequencies. The spectral density of the wind speedprocess including wake effects from any upstream wind tur-bines is ultimately of interest.

Guidance note:The latest insights for wind spectrum modelling within windfarms should be considered when the spectral density of the windspeed process is to be established.


402 Site-specific spectral densities of the wind speed processcan be determined from available measured wind data. Whenmeasured wind data are insufficient to establish site-specificspectral densities, it is recommended to use a spectral densitymodel which fulfils that the spectral density SU(f) asymptoti-cally approaches the following form as the frequency f in theinertial subrange increases:

403 Unless data indicate otherwise, the spectral density ofthe wind speed process may be represented by the Kaimalspectrum,

in which f denotes frequency, and the integral scale parameter

⎭⎬⎫

⎩⎨⎧

⋅⋅−⋅⎭⎬⎫

⎩⎨⎧ ⋅+⋅=

00 ln)(41.01ln1),(

TTzI

hzCUzTU U

02 15.011073.5 UC +⋅= −

22.00 )()043.01(06.0 −⋅+⋅=

hzUIU

35

32

10

2202.0)(−−

⎟⎟⎠

⎞⎜⎜⎝

⎛= f

ULfS k

UU σ

3/5

10

102

)61(

4)(

UfLUL

fSk

k

UU+

= σ

DET NORSKE VERITAS


Lk is to be taken as

where z denotes the height above the seawater level. Thismodel spectrum fulfils the requirement in 402. Other modelspectra for wind speed processes than the Kaimal spectrum canbe found in DNV-RP-C205.

Guidance note:Caution should be exercised when model spectra such as the Kai-mal spectrum are used. In particular, it is important to bewarethat the true length scale may deviate significantly from thelength scale Lk of the model spectrum. The Kaimal spectrum and other model spectra can be used to rep-resent the upstream wind field in front of the wind turbine. How-ever, a rotational sampling turbulence due to the rotation of therotor blades will come in addition to the turbulence of theupstream wind field as represented by the model spectrum andwill increase the wind fluctuations that the rotor blades effec-tively will experience. For wind turbines located behind other wind turbines in a windfarm, the wind fluctuations represented by the model spectrumwill become superimposed by an additional turbulence due towake effects behind upstream wind turbines.


404 The long-term probability distributions for the wind cli-mate parameters U10 and σU that are interpreted from availabledata can be represented in terms of generic distributions or interms of scattergrams. A typical generic distribution represen-tation consists of a Weibull distribution for the 10-minutemean wind speed U10 in conjunction with a lognormal distri-bution of σU conditional on U10. A scattergram provides thefrequency of occurrence of given pairs (U10, σU) in a givendiscretisation of the (U10, σU) space.405 Unless data indicate otherwise, a Weibull distributioncan be assumed for the 10-minute mean wind speed U10 in agiven height H above the seawater level,

in which the scale parameter A and the shape parameter k aresite- and height-dependent.

Guidance note:In areas where hurricanes occur, the Weibull distribution asdetermined from available 10-minute wind speed records maynot provide an adequate representation of the upper tail of thetrue distribution of U10. In such areas, the upper tail of the distri-bution of U10 needs to be determined on the basis of hurricanedata.


406 The wind speed profile represents the variation of thewind speed with height above the seawater level.

Guidance note:A logarithmic wind speed profile may be assumed,

in which z is the height above the seawater level and z0 is aroughness parameter, which for offshore locations depends onthe wind speed, the upstream distance to land, the water depthand the wave field. The logarithmic wind speed profile impliesthat the scale parameter A(z) in height z can be expressed in

terms of the scale parameter A(H) in height H as follows

The roughness parameter z0 typically varies between 0.0001 m inopen sea without waves and 0.003 m in coastal areas withonshore wind. The roughness parameter may be solved implicitlyfrom the following equation:

where g is the acceleration of gravity, κ = 0.4 is von Karman’sconstant, and AC is Charnock’s constant. For open sea with fullydeveloped waves, AC = 0.011 to 0.014 is recommended. Fornear-coastal locations, AC is usually higher with values of 0.018or more. Whenever extrapolation of wind speeds to other heightsthan the height of the wind speed measurements is to be carriedout, conservative worst-case values of AC should be applied.As an alternative to the logarithmic wind profile, the power lawprofile may be assumed,

Offshore wind profiles can be governed more by atmosphericstability than by the roughness parameter z0. For stability correc-tions of wind profiles reference is made to DNV-RP-C205.


407 Let FU10(u) denote the long-term distribution of the 10-minute mean wind speed U10. In areas where hurricanes do notoccur, the distribution of the annual maximum 10-minutemean wind speed U10,max can be approximated by

where N = 52 560 is the number of stationary 10-minute peri-ods in one year.

Guidance note:The quoted power-law approximation to the distribution of theannual maximum 10-minute mean wind speed is a good approx-imation to the upper tail of this distribution. Usually only quan-tiles in the upper tail of the distribution are of interest, viz. the98% quantile which defines the 50-year mean wind speed. Theupper tail of the distribution can be well approximated by a Gum-bel distribution, whose expression is more operational than thequoted power-law expression.Since the quoted power-law approximation to the distribution ofthe annual maximum 10-minute mean wind speed is used to esti-mate the 50-year mean wind speed by extrapolation, cautionmust be exercised when the underlying distribution FU10 of thearbitrary 10-minute mean wind speed is established. This appliesin particular if FU10 is represented by the Weibull distribution ofU10 commonly used for prediction of the annual power produc-tion from the wind turbine, since this distribution is usually fittedto mid-range wind velocities and may not necessarily honourhigh-range wind speed data adequately.


In areas where hurricanes occur, the distribution of the annualmaximum 10-minute mean wind speed U10,max shall be basedon available hurricane data. This refers to hurricanes for whichthe 10-minute mean wind speed forms a sufficient representa-tion of the wind climate. 408 The 10-minute mean wind speed with return period TRin units of years is defined as the (1− 1/TR) quantile in the dis-tribution of the annual maximum 10-minute mean wind speed,i.e. it is the 10-minute mean wind speed whose probability ofexceedance in one year is 1/TR. It is denoted U10,TR and is

⎩⎨⎧

≥<

=m 60for m 2.340m 60for 67.5

zzz

Lk

))(exp(1)(10k

U AuuF −−=

0

ln)(zzzu ∝

0

0

ln

ln)()(

zHzz

HAzA =

2

0

100 )/ln( ⎟⎟

⎠

⎞⎜⎜⎝

⎛=

zzU

gAz C κ

α⎟⎠⎞

⎜⎝⎛=

HzHUzu )()( 10

NUU uFuF ))(()(

10year1max,,10=

DET NORSKE VERITAS


expressed as

in which TR > 1 year and FU10,max,1 year denotes the cumula-tive distribution function of the annual maximum of the 10-minute mean wind speed.The 10-minute mean wind speed with return period one year isdefined as the mode of the cumulative distribution function ofthe annual maximum of the 10-minute mean wind speed.

Guidance note:The 50-year 10-minute mean wind speed becomes U10,50 = FU10,max,1 year

–1(0.98) and the 100-year 10-minutemean wind speed becomes U10,100 = FU10,max,1 year

–1(0.99). Note that these values, calculated as specified, are to be consid-ered as central estimates of the respective 10-minute wind speedswhen the underlying distribution function FU10,max is determinedfrom limited data and is encumbered with statistical uncertainty.


409 The natural variability of the wind speed about the meanwind speed U10 in a 10-minute period is known as turbulenceand is characterised by the standard deviation σU. For givenvalue of U10, the standard deviation σU of the wind speedexhibits a natural variability from one 10-minute period toanother. Caution should be exercised when fitting a distribu-tion model to data for the standard deviation σU. Often, thelognormal distribution provides a good fit to data for σU con-ditioned on U10, but use of a normal distribution, a Weibulldistribution or a Frechet distribution is also seen. The choice ofthe distribution model may depend on the application, i.e.,whether a good fit to data is required to the entire distributionor only in the body or the upper tail of the distribution.

Guidance note:When the lognormal distribution is an adequate distributionmodel, the distribution of σU conditioned on U10 can beexpressed as

in which Φ() denotes the standard Gaussian cumulative distribu-tion function. The coefficients b0 and b1 are site-dependent coef-ficients dependent on U10. The coefficient b0 can be interpreted as the mean value of lnσU,and b1 as the standard deviation of lnσU. The following relation-ships can be used to calculate the mean value E[σU] and thestandard deviation D[σU] of σU from the values of b0 and b1,

E[σU] and D[σU] will, in addition to their dependency on U10,also depend on local conditions, first of all the surface roughnessz0.Caution should be exercised when the distribution of σU condi-tioned on U10 is interpreted from data. It is important to identifyand remove data, which belong to 10-minute series for which thestationarity assumption for U10 is not fulfilled. If this is not done,such data may confuse the determination of an appropriate distri-bution model for σU conditioned on U10. Techniques for“detrending” of data are available for application in the case thatthe mean wind speed follows a trend rather than stays stationaryduring a considered 10-minute period.


410 Let U10 and σU denote the 10-minute mean wind speedand the standard deviation of the wind speed, respectively, in aconsidered 10-minute period of stationary wind conditions.Unless data indicate otherwise, the short-term probability dis-

tribution for the instantaneous wind speed U at an arbitrarypoint in time during this period can be assumed to be a normaldistribution. The cumulative distribution function for U canthen be expressed as

in which Φ() denotes the standard Gaussian cumulative distri-bution function.

Guidance note:When data do not support the assumption of a normal distribu-tion of the wind speed U conditioned on U10 and σU, othergeneric distribution types may be tried out, and it may be neces-sary to introduce additional distribution parameters such as theskewness α3 of the wind speed in order to arrive at an adequaterepresentation of the wind speed distribution.


411 Practical information regarding wind modelling is givenin DNV-RP-C205, in IEC61400-1 and in DNV/Risø Guide-lines for Design of Wind Turbines.

B 500 Reference wind conditions and reference wind speeds501 For use in load combinations for design, a number of ref-erence wind conditions and reference wind speeds are defined.502 The Normal Wind Profile (NWP) represents the averagewind speed as a function of height above sea level.

Guidance note:For standard wind turbine classes according to IEC61400-1, thenormal wind profile is given by the power law model with expo-nent α = 0.2. For offshore locations it is recommended to applyan exponent α = 0.14.


503 The Normal Turbulence Model (NTM) represents turbu-lent wind speed in terms of a characteristic standard deviationof wind speed, σU,c. The characteristic standard deviation σU,cis defined as the 90% quantile in the probability distribution ofthe standard deviation σU of the wind speed conditioned on the10-minute mean wind speed at the hub height.

Guidance note:For standard wind turbine classes according to IEC61400-1, pre-scribed values for the characteristic standard deviation σU,c aregiven in IEC61400-1.


504 The Extreme Wind Speed Model (EWM) is used to rep-resent extreme wind conditions with a specified return period,usually either one year or 50 years. It shall be either a steadywind model or a turbulent wind model. In case of a steady windmodel, the extreme wind speed (UEWM) at the hub height witha return period of 50 years shall be calculated as

where U10,hub,50-yr denotes the 10-minute mean wind speed atthe hub height with a return period of 50 years. The extremewind speed (UEWM) at the hub height with a return period ofone year shall be calculated as

The quantities Uhub,50-yr and Uhub,1-yr refer to wind speedaveraged over three seconds. In the steady extreme windmodel, allowance for short-term deviations from the meanwind direction shall be made by assuming constant yaw mis-alignment in the range of ±15°.The turbulent extreme wind model makes use of the 10-minutemean wind speed at the hub height with a return period of 50

)11(1,10 year1max,,10

RUT T

FU R −= −

)ln

()(1

0| 10 b

bF UU

−Φ=

σσσ

[ ] )21exp( 2

10 bbE U +=σ

[ ] [ ] 1)exp( 21 −= bED UU σσ

)()( 10,| 10

UUU

UuuF

U σσ

−Φ=

yrhubyrhub UU −− ⋅= 50,,1050, 4.1

yrhubyrhub UU −− ⋅= 50,1, 8.0

DET NORSKE VERITAS


years, U10,hub,50-yr. The 10-minute mean wind speed at the hubheight with a return period of one year shall be calculated as

Further, for representation of turbulent wind speeds, the turbu-lent extreme wind model makes use of a characteristic standarddeviation of the wind speed. The characteristic standard devi-ation of the wind speed shall be calculated as

Guidance note:For calculation of wind speeds and 10-minute mean wind speedsat other heights than the hub height, IEC61400-1 prescribes awind profile given by the power law model with exponentα = 0.11.


505 The Extreme Operating Gust (EOG) at the hub heighthas a magnitude which shall be calculated as

in which

σU,c = characteristic standard deviation of wind speed,defined as 90% quantile in the probability distribu-tion of σU

Λ1 = longitudal turbulence scale parameter, is related tothe integral scale parameter Lk of the Kaimal spec-tral density through the relationship Lk=8.1Λ1.

D = rotor diameter.The wind speed V as a function of height z and time t shall bedefined as follows

whereT = 10.5 secandu(z) is defined by the Normal Wind Profile. An example of extreme operating gust at the hub height isgiven in Figure 1 for a case where the 10-minute mean windspeed is 25 m/sec.

Figure 1 Example of extreme operating gust

506 The Extreme Turbulence Model (ETM) combines thenormal wind profile model NPM with a turbulent wind speed

whose characteristic standard deviation is given by

in which

c = 2 m/sUhub = wind speed at hub heightUaverage = long-term average wind speed at hub heightIref = expected value of turbulence intensity at hub

height at U10,hub = 15 m/s507 The Extreme Direction Change (EDC) has a magnitudewhose value shall be calculated according to the followingexpression

where

σU,c = characteristic standard deviation of wind speed,defined according to the Normal Turbulence Modelas the 90% quantile in the probability distribution ofσU

Λ1 = longitudal turbulence scale parameter, is related tothe integral scale parameter Lk of the Kaimal spec-tral density through the relationshipLk = 8.1Λ1.U10,hub = 10-minute mean wind speed athub height

D = rotor diameter.θe is limited to the range ±180°.

The extreme direction change transient, θ(t), as a function oftime t shall be given by:

where T = 6 sec is the duration of the extreme direction change.The sign shall be chosen so that the worst transient loadingoccurs. At the end of the direction change transient, the direc-tion is assumed to remain unchanged. The wind speed isassumed to follow the normal wind profile model given in 502.As an example, the magnitude of the extreme direction changewith a return period of one year is shown in Figure 2 for vari-ous values of Vhub = U10,hub. The corresponding transient forVhub = U10,hub = 25 m/s is shown in Figure 3.

Figure 2 Example of extreme direction change magnitude

yrhubyrhub UU −− ⋅= 50,,101,,10 8.0

hubcU U ,10, 11.0 ⋅=σ

⎭⎬⎫

⎩⎨⎧

Λ+−= −

1

,,101, 1.01

3.3);(35.1min

DUUV cU

hubyrhubgust

σ

otherwiseTtfor

zuT

tT

tVzutzV gust≤≤

⎪⎩

⎪⎨⎧ ⋅

−⋅

−=0

)(

))2cos(1)(3sin(37.0)(),(ππ

2022242628

30323436

0 2 4 6 8 10

Time t in secs.

EO

G W

ind

spee

d in

hub

hei

ght

⎟⎟⎠

⎞⎜⎜⎝

⎛+⎟

⎠⎞

⎜⎝⎛ −⋅⎟⎟

⎠

⎞⎜⎜⎝

⎛+⋅⋅⋅= 1043072.0, c

Uc

UIc hubaverage

refcUσ

)1.01(arctan4

1,10

,

Λ+⋅±=

DU hub

cUe

σθ

TtforTtfor

tforTtt e

≥≤≤

<

⎪⎩

⎪⎨

⎧⋅−=

0

0

0))/cos(1(5.0

0)( πθθ

-200

-100

0

100

200

0 10 20 30 40

Wind speed, V hub (m/s)

ED

C c

hang

e, θ

e (

deg)

DET NORSKE VERITAS


Figure 3 Example of extreme direction change

508 The Extreme Coherent Gust with Direction Change(ECD) shall have a magnitude of Vcg = 15 m/sec.The wind speed V as a function of height z and time t shall bedefined as follows

where T = 10 sec is the rise time and u(z) is the wind speedgiven in 502. The extreme coherent gust is illustrated in Figure4 for Vhub = U10,hub = 25 m/s.

Figure 4 Example of extreme coherent gust amplitude

The rise in wind speed (described by the extreme coherentgust, see Figure 4) shall be assumed to occur simultaneouslywith the direction change θ from 0 degrees up to and includingθcg, where θcg is defined by:

The direction change θcg is shown in Figure 5 as a function ofthe 10-minute mean wind speed Vhub = U10,hub at hub height.The direction change which takes place simultaneously as thewind speed rises is given by

where T = 10 sec is the rise time. The normal wind profilemodel as specified in 502 shall be used. An example of thedirection change is shown in Figure 6 as a function of time forVhub = U10,hub = 25 m/s.

Figure 5 Direction change

Figure 6 Temporal evolution of direction change for Vhub = 25 m/s

509 The Extreme Wind Shear model (EWS) is used toaccount for extreme transient wind shear events. It consists ofa transient vertical wind shear and a transient horizontal windshear. The extreme transient positive and negative verticalshear shall be calculated as

The extreme transient horizontal shear shall be calculated as

0

10

20

30

40

-5 0 5 10

Time, t (s)

EDC

Win

d di

rect

ion

chan

ge,

θ(t

) (d

eg)

⎪⎩

⎪⎨

⎧

>+≤≤⋅−+

<=

TtforVzuTtforTtVzu

tforzutzV

cg

cg

)(0))/cos(1(5.0)(

0)(),( π

01020304050

-2 0 2 4 6 8 10 12 14Time, t (s)

Win

d sp

eed

V(z

,t)

(m/s

)

⎪⎩

⎪⎨

⎧

>

≤= smUfor

Usm

smUforU

hubhub

hub

hubcg / 4 /720/ 4 180

)(,10

,10

,10

,10o

o

θ

⎪⎩

⎪⎨

⎧

>±≤≤⋅−±

<=

TtforTtforTt

tfort

cg

cg

θπθθ 0))/cos(1(5.0

00)(

o

0

50

100

150

200

0 10 20 30 40

Wind speed, V hub (m/s)

Dire

ctio

n ch

ange

, θcg

(deg

)

0

5

10

15

20

25

30

-2 0 2 4 6 8 10 12

Time, t (s)

Dire

ctio

n ch

ange

, V(z

,t) (

deg)

⎪⎪⎩

⎪⎪⎨

⎧≤≤⎟

⎠⎞

⎜⎝⎛ ⋅

−⋅⎟⎟

⎠

⎞

⎜⎜

⎝

⎛⎟⎟⎠

⎞⎜⎜⎝

⎛Λ

+⋅−

±=

otherwise)(

0for )2cos(12.05.2)(),(

10

41

1,10

zU

TtT

tDDzzzUtzV cU

hub πβσ

⎪⎪⎩

⎪⎪⎨

⎧≤≤⎟

⎠⎞

⎜⎝⎛ ⋅

−⋅⎟⎟

⎠

⎞

⎜⎜

⎝

⎛⎟⎟⎠

⎞⎜⎜⎝

⎛Λ

+⋅±=

otherwise)(

0for )2cos(12.05.2)(),,(

10

41

1,10

zU

TtT

tDDyzUtzyV cU

πβσ

DET NORSKE VERITAS


Here, U10(z) denotes the wind shear profile according to theNormal Wind Profile model, z is the height above sea level,zhub is the hub height, y is the lateral cross-wind distance, Λ1 isthe longitudinal turbulence scale parameter, σU,c is the charac-teristic standard deviation of wind speed, defined according tothe Normal Turbulence Model as the 90% quantile in the prob-ability distribution of σU, D is the rotor diameter, β = 6.4 andT = 12 sec. The sign for the horizontal wind shear transient shall be chosenin such a manner that the most unfavourable transient loadingoccurs. The extreme transient horizontal shear and the extremetransient vertical shear shall not be applied simultaneously.

Guidance note:For standard wind turbine classes according to IEC61400-1, thenormal wind profile U10(z) is given by the power law model withexponent α = 0.2. For offshore locations it is recommended toapply an exponent α = 0.14.


510 The Reduced Wind Speed Model (RWM) defines a com-panion wind speed URWM to be used in combination with theextreme wave height (EWH) for definition of an extreme eventwith a specified return period. The reduced wind speed can beexpressed as a fraction of the extreme wind speed,URWM = ψ · UEWM, ψ < 1. The Reduced Wind Speed is usedfor definition of events with return periods of 50 years and 1year, and the corresponding reduced wind speeds are denotedURed,50-yr and URed,1-yr, respectively.

Guidance note:IEC61400-3/CD requires use of URed,50-yr = 1.1U10,50-yr, whichimplies ψ = 0.79. Other values for ψ can be applied, providedthey can be substantiated by site-specific data.


C. Wave ClimateC 100 Wave parameters101 The wave climate is represented by the significant waveheight HS and the spectral peak period TP. In the short term, i.e.over a 3-hour or 6-hour period, stationary wave conditionswith constant HS and constant TP are assumed to prevail.

Guidance note:The significant wave height HS is defined as four times the stand-ard deviation of the sea elevation process. The significant waveheight is a measure of the intensity of the wave climate as well asof the variability in the arbitrary wave heights. The peak periodTP is related to the mean zero-crossing period TZ of the sea ele-vation process.


102 The wave height H of a wave cycle is the differencebetween the highest crest and the deepest trough between twosuccessive zero-upcrossings of the sea elevation process. Thearbitrary wave height H under stationary 3- or 6-hour condi-tions in the short term follows a probability distribution whichis a function of the significant wave height HS. 103 The wave period is defined as the time between two suc-cessive zero-upcrossings of the sea elevation process. Thearbitrary wave period T under stationary 3- or 6-hour condi-tions in the short term follows a probability distribution, whichis a function of HS, TP and H.104 The wave crest height HC is the height of the highestcrest between two successive zero-upcrossings of the sea ele-vation process. The arbitrary wave crest height HC under sta-tionary 3- or 6-hour conditions in the short term follows aprobability distribution which is a function of the significant

wave height HS.105 The short term 3- or 6-hour sea state may be representedby a wave spectrum, i.e. the power spectral density function ofthe sea elevation process, S(f). S(f) is a function of HS and TPand expresses how the energy of the sea elevation is distributedbetween various frequencies.

C 200 Wave data201 Wave statistics are to be used as a basis for representa-tion of the long-term and short-term wave conditions. Empiri-cal statistical data used as a basis for design must cover asufficiently long period of time.

Guidance note:Wave data obtained on site are to be preferred over wave dataobserved at an adjacent location. Measured wave data are to bepreferred over visually observed wave data. Continuous recordsof data are to be preferred over records with gaps. Longer periodsof observation are to be preferred over shorter periods. When no site-specific wave data are available and data fromadjacent locations are to be capitalised on in stead, proper trans-formation of such other data shall be performed to account forpossible differences due to different water depths and differentseabed topographies. Such transformation shall take effects ofshoaling and refraction into account.Hindcast of wave data may be used to extend measured timeseries, or to interpolate to places where measured data have notbeen collected. If hindcast is used, the hindcast model shall becalibrated against measured data to ensure that the hindcastresults comply with available measured data.


202 The long-term distributions of HS and TP should prefer-ably be based on statistical data for the same reference periodfor the waves as the reference period which is used for thedetermination of loads. If a different reference period than 3 or6 hours is used for the determination of loads, the wave datamay be converted by application of appropriate adjustmentfactors.

Guidance note:When the long-term distribution of the arbitrary significant waveheight HS is given by a Weibull distribution,

the significant wave height HS,Ts for a reference period of dura-tion TS can be obtained from the significant wave height HS,Ts0for a reference period of duration TS0 according to the followingrelationship,

in which N0 is the number of sea states of duration TS0 in oneyear and TR is the specified return period of the significant waveheight, which is to be converted. N0 = 2920 when TS0 = 3 hours.TR must be given in units of years.


203 Wave climate and wind climate are correlated, becausewaves are usually wind-generated. The correlation betweenwave data and wind data shall be accounted for in design.

Guidance note:Simultaneous observations of wave and wind data in terms ofsimultaneous values of HS and U10 should be obtained. It is rec-ommended that directionality of wind and waves are recorded.Extreme waves may not always come from the same direction asextreme winds. This may in particular be so when the fetch in thedirection of the extreme winds is short.

))(exp(1)(0

βhhhF SH −−=

β1

0

0,, )ln(

)ln(1

0 ⎟⎟⎠

⎞⎜⎜⎝

⎛+⋅=

R

SSTSTS TN

TTHH

SS

DET NORSKE VERITAS


Within a period of stationary wind and wave climates, individualwind speeds and wave heights can be assumed independent anduncorrelated.


C 300 Wave modelling301 Site-specific spectral densities of the sea elevation proc-ess can be determined from available wave data.302 Unless data indicate otherwise, the spectral density ofthe sea elevation process may be represented by the JON-SWAP spectrum,

where

f = wave frequency, f = 1/TT = wave periodfp = spectral peak frequency, fp = 1/TpTp = peak periodg = acceleration of gravityα = generalised Phillips’ constant

= 5 · (HS2fp

4/g2) · (1− 0.287lnγ) · π 4

σ = spectral width parameter= 0.07 for f ≤ fp and σ = 0.09 for f > fp

γ = peak-enhancement factor.The zero-upcrossing period TZ depends on the peak period Tpthrough the following relationship,

The peak-enhancement factor is

where Tp is in seconds and HS is in metres.Guidance note:When γ = 1 the JONSWAP spectrum reduces to the Pierson-Moskowitz spectrum.


303 The long-term probability distributions for the wave cli-mate parameters HS and TP that are interpreted from availabledata can be represented in terms of generic distributions or interms of scattergrams. A typical generic distribution represen-tation consists of a Weibull distribution for the significantwave height HS in conjunction with a lognormal distribution ofTP conditional on HS. A scattergram gives the frequency ofoccurrence of given pairs (HS,TP) in a given discretisation ofthe (HS,TP) space.304 Unless data indicate otherwise, a Weibull distributioncan be assumed for the significant wave height,

305 When FHs(h) denotes the distribution of the significantwave height in an arbitrary t-hour sea state, the distribution ofthe annual maximum significant wave height HSmax can betaken as

where N is the number of t-hour sea states in one year. For t = 3hours, N = 2920.

Guidance note:The quoted power-law approximation to the distribution of theannual maximum significant wave height is a good approxima-tion to the upper tail of this distribution. Usually only quantilesin the upper tail of the distribution are of interest, in particular the98% quantile which defines the 50-year significant wave height.The upper tail of the distribution can be well approximated by aGumbel distribution, whose expression is more operational thanthe quoted power-law expression.


306 The significant wave height with return period TR inunits of years is defined as the (1− 1/TR) quantile in the distri-bution of the annual maximum significant wave height, i.e. itis the significant wave height whose probability of exceedancein one year is 1/TR. It is denoted HS,TR and is expressed as

in which TR > 1 year.The significant wave height with return period one year isdefined as the mode of the distribution function of the annualmaximum of the significant wave height.

Guidance note:The 50-year significant wave height becomes HS,50 = FHs,max,1 year

–1(0.98) and the 100-year significant waveheight becomes HS,100 = FHs,max,1 year

–1(0.99). Note that these values, calculated as specified, are to be consid-ered as central estimates of the respective significant waveheights when the underlying distribution function FHs,max isdetermined from limited data and is encumbered with statisticaluncertainty. In the southern and central parts of the North Sea, experienceshows that the ratio between the 100- and 50-year significantwave heights HS,100/HS,50 attains a value approximately equal to1.04 to 1.05. Unless data indicate otherwise, this value of theratio HS,100/HS,50 may be applied to achieve the 50-year signifi-cant wave height HS,50 in cases where only the 100-year valueHS,100 is available, provided the location in question is located inthe southern or central parts of the North Sea.


307 In deep waters, the short-term probability distribution ofthe arbitrary wave height H can be assumed to follow aRayleigh distribution when the significant wave height HS isgiven,

where FH|Hs denotes the cumulative distribution function andν is a spectral width parameter whose value is ν = 0.0 for a nar-row-banded sea elevation process. The maximum wave height Hmax in a 3-hour sea state charac-terised by a significant wave height HS can be calculated as aconstant factor times HS.

Guidance note:The maximum wave height in a sea state can be estimated by themean of the highest wave height in the record of waves that occurduring the sea state, or by the most probable highest wave heightin the record. The most probable highest wave height is also

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛

⎟⎟⎠

⎞⎜⎜⎝

⎛

⋅σ

−−−

−

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛

⎟⎟⎠

⎞⎜⎜⎝

⎛−

π=

2

5.0exp45

4

2

45exp

)2()(

p

pfff

pγ

fffgαfS

γγ

++

=115

pZ TT

⎪⎪⎪⎪

⎩

⎪⎪⎪⎪

⎨

⎧

<

≤<−

≤

=γ

S

pS

p

S

pS

p

H

Tfor

H

Tfor

H

TH

Tfor

51

56.3)15.175.5exp(

6.35

))(exp(1)( βαhhF

SH −−=

NHH hFhF

SyearS))(()(

1max,,=

)11(1, 1,max,

RHTS T

FH yearSR −= −

))1(

2exp(1)( 22

2

|S

HHH

hhF S ν−−−=

DET NORSKE VERITAS


known as the mode of the highest wave height. Both of these esti-mates for the maximum wave height in a sea state depend on thenumber of waves, N, in the record. N can be defined as the ratiobetween the duration TS of the sea state and the mean zero-upcrossing period TZ of the waves. For a narrow-banded sea ele-vation process, the appropriate expression for the mean of thehighest wave height Hmax reads

while the expression for the mode of the highest wave heightreads

For a sea state of duration TS = 3 hours and a mean zero-upcross-ing period TZ of about 10.8 sec, N = 1000 results. For this exam-ple, the mean of the highest wave height becomesHmax = 1.936HS ≈ 1.94HS, while the mode of the highest waveheight becomes Hmax = 1.858HS ≈ 1.86HS. For shorter meanzero-upcrossing periods than the assumed 10.8 sec, N becomeslarger, and so does the factor on HS. Table C1 gives the ratioHmax/HS for various values of N.

Other ratios than those quoted in Table C1 apply to waves inshallow waters and in cases where the sea elevation process is notnarrow-banded.It is common to base the estimation of Hmax on the results for themode rather than on the results for the mean.Table C1 is valid for HS/d < 0.2, where d denotes water depth.


308 In shallow waters, the wave heights will be limited bythe water depth. Unless data indicate otherwise, the maximumwave height can be taken as 78% of the water depth. TheRayleigh distribution of the wave heights will become dis-torted in the upper tail to approach this limit asymptotically.Use of the unmodified Rayleigh distribution for representationof the distribution of wave heights in shallow waters maytherefore be on the conservative side.309 In shallow waters with constant seabed slope, the Battjesand Groenendijk distribution can be used to represent the prob-ability distribution of the arbitrary wave height H conditionalon the significant wave height HS. It is a requirement for thisuse of the Battjes and Groenendijk distribution that it is vali-dated by measured site-specific wave data. The Battjes andGroenendijk distribution is a composite Weibull distribution

whose cumulative distribution function reads

in which the transition wave height hT is defined as

where α is the slope angle of the sea floor and d is the waterdepth. The parameters h1 and h2 are functions of the transitionwave height hT and of the root mean square HRMS of the waveheights. The root mean square HRMS is calculated from the sig-nificant wave height HS and the water depth d as

and the parameters h1 and h2 can be found from the followingapproximate expressions, valid for 0.05HRMS < hT < 3HRMS,

The Battjes and Groenendijk distribution is not defined forhT > 3HRMS.

Guidance note:The Battjes and Groenendijk distribution has the drawback thatit has an unphysical “knee” at the transition height hT. The Bat-tjes and Groenendijk distribution should therefore be used withcaution and only when supported by data.Other distribution models for wave heights in shallow waters existand can be used as alternatives to the Battjes and Groenendijk dis-tribution.


310 The long-term probability distribution of the arbitrarywave height H can be found by integration over all significantwave heights

where

in which fHsTp(hs,t) is the joint probability density of the sig-nificant wave height HS and the peak period TP and ν0(hs,t) isthe zero-upcrossing rate of the sea elevation process for givencombination of HS and TP. FH|HsTp(h) denotes the short-termcumulative distribution function for the wave height H condi-tioned on HS and TP.311 When FH(h) denotes the distribution of the arbitrarywave height H, the distribution of the annual maximum waveheight Hmax can be taken as

Table C1 Ratio for deep water waves in narrow-banded sea elevation process

No. of waves N = TS/TZ

Ratio Hmax/HSmode mean

500 1.763 1.8451000 1.858 1.9361500 1.912 1.9882000 1.949 2.0232500 1.978 2.0515000 2.064 2.134

smean HN

NH⎥⎥⎦

⎤

⎢⎢⎣

⎡+=

ln22886.0ln

21

max,

sHNH⎥⎥⎦

⎤

⎢⎢⎣

⎡= ln

21

modemax,

Nln21

NN

ln22886.0ln

21

+

⎪⎪⎩

⎪⎪⎨

⎧

>−−

≤−−=

T

T

HH

hhhh

hhhh

hFS

for ))(exp(1

for ))(exp(1)(

6.3

2

2

1|

dhT ⋅⋅+= )tan8.535.0( α

dH

Hh SSRMS

2

2025.06725.0 +=

⎟⎟⎠

⎞⎜⎜⎝

⎛+⎟⎟

⎠

⎞⎜⎜⎝

⎛−⎟⎟

⎠

⎞⎜⎜⎝

⎛=

RMS

T

RMS

T

RMS

TRMS

Hh

Hh

HhH

h

3339.1583.00835.0

123

1

43

2

2

01925.0083259.0

01532.006.1

⎟⎟⎠

⎞⎜⎜⎝

⎛−⎟⎟

⎠

⎞⎜⎜⎝

⎛+

⎟⎟⎠

⎞⎜⎜⎝

⎛−=

RMS

T

RMS

T

RMS

T

RMS

Hh

Hh

Hh

Hh

∫ ∫ ⋅⋅=S

PSPSh

SSTHTHHSt

H dtdhthfhFthhF ),()(),(1)( |00

νν

∫ ∫ ⋅=S

PSh t

SSTHS dtdhthfth ),(),(00 νν

Wyear

NHH hFhF ))(()(1,max =

DET NORSKE VERITAS


where NW is the number of wave heights in one year.312 Unless data indicate otherwise, the wave crest height HCcan be assumed to be 0.65 times the associated arbitrary waveheight H. 313 The wave height with return period TR in units of yearsis defined as the (1−1/TR) quantile in the distribution of theannual maximum wave height, i.e. it is the wave height whoseprobability of exceedance in one year is 1/TR. It is denotedHTR and is expressed as

in which TR > 1 year.The wave height with return period one year is defined as themode of the distribution function of the annual maximum waveheight.

Guidance note:The 50-year wave height becomes H50 = FHmax,1 year

–1(0.98) andthe 100-year wave height becomes H100 = FHs,max,1 year

–1(0.99).Note that these values, calculated as specified, are to be consid-ered as central estimates of the respective wave heights when theunderlying distribution function FHmax is determined from lim-ited data and is encumbered with statistical uncertainty. Note also that the 50-year wave height H50 is always greater thanthe maximum wave height Hmax in the 3-hour sea state whosereturn period is 50 years and whose significant wave height isdenoted HS,50. This implies that in deep waters H50 will take ona value greater than Hmax = 1.86HS,50. Values of H50 equal toabout 2.0 times HS,50 are not uncommon in deep waters.


314 Directionality of waves shall be considered for determi-nation of wave height distributions and wave heights withspecified return periods when such directionality has an impacton the design of a wind turbine structure.

C 400 Reference sea states and reference wave heights401 For use in load combinations for design, a number of ref-erence sea states and reference wave heights are defined.402 The Normal Sea State (NSS) is characterised by a signif-icant wave height, a peak period and a wave direction. It isassociated with a concurrent mean wind speed. The significantwave height HS,NSS of the normal sea state is defined as theexpected value of the significant wave height conditioned onthe concurrent 10-minute mean wind speed. The normal seastate is used for calculation of ultimate loads and fatigue loads.For fatigue load calculations a series of normal sea states haveto be considered, associated with different mean wind speeds.It must be ensured that the number and resolution of these nor-mal sea states are sufficient to predict the fatigue damage asso-ciated with the full long-term distribution of metoceanparameters. The range of peak periods TP appropriate to eachsignificant wave height shall be considered. Design calcula-tions shall be based on values of the peak period which resultin the highest loads or load effects in the structure. 403 The Normal Wave Height (NWH) HNWH is defined asthe expected value of the significant wave height conditionedon the concurrent 10-minute mean wind speed, i.e.HNWH = HS,NSS. The range of wave periods T appropriate tothe normal wave height shall be considered. Design calcula-tions shall be based on values of the wave period within thisrange that result in the highest loads or load effects in the struc-ture.

Guidance note:In deep waters, the wave periods T to be used with HNWH maybe assumed to be within the range given by


404 The Severe Sea State (SSS) is characterised by a signifi-cant wave height, a peak period and a wave direction. It is asso-ciated with a concurrent mean wind speed. The significant waveheight of the severe sea state HS,SSS is defined by extrapolationof appropriate site-specific metocean data such that the loadeffect from the combination of the significant wave heightHS,SSS and the 10-minute mean wind speed U10 has a returnperiod of 50 years. The SSS model is used in combination withnormal wind conditions for calculation of the ultimate loading ofan offshore wind turbine during power production. The SSSmodel is used to associate a severe sea state with each mean windspeed in the range corresponding to power production. For all10-minute mean wind speeds U10 during power production, theunconditional extreme significant wave height, HS,50-yr, with areturn period of 50 years may be used as a conservative estimatefor HS,SSS(U10). Further guidance regarding estimation ofHS,SSS is provided in 4F703. The range of peak periods TPappropriate to each significant wave height shall be considered.Design calculations shall be based on values of the peak periodwhich result in the highest loads or load effects in the structure.405 The Severe Wave Height (SWH) HSWH is associatedwith a concurrent mean wind speed and is defined by extrapo-lation of appropriate site-specific metocean data such that theload effect from the combination of the severe wave heightHSWH and the 10-minute mean wind speed U10 has a returnperiod of 50 years. The SWH model is used in combinationwith normal wind conditions for calculation of the ultimateloading of an offshore wind turbine during power production.The SWH model is used to associate a severe wave height witheach mean wind speed in the range corresponding to powerproduction. For all 10-minute mean wind speeds U10 duringpower production, the unconditional extreme wave height,H50-yr, with a return period of 50 years may be used as a con-servative estimate for HSWH(U10). The range of wave periodsT appropriate to the severe wave height shall be considered.Design calculations shall be based on values of the waveperiod within this range that result in the highest loads or loadeffects in the structure.

Guidance note:In deep waters, the wave periods T to be used with HSWH may beassumed to be within the range given by


406 The Extreme Sea State (ESS) is characterised by a sig-nificant wave height, a peak period and a wave direction. Thesignificant wave height HS,ESS is the unconditional significantwave height with a specified return period, determined fromthe distribution of the annual maximum significant waveheight as outlined in 306. The Extreme Sea State is used forreturn periods of 50 years and 1 year, and the correspondingsignificant wave heights are denoted HS,50-yr and HS,1-yr,respectively. The range of peak periods TP appropriate to eachof these significant wave heights shall be considered. Designcalculations shall be based on values of the peak period whichresult in the highest loads or load effects in the structure. 407 The Extreme Wave Height (EWH) HEWH is a waveheight with a specified return period. It can be determined fromthe distribution of the annual maximum wave height as out-lined in 313. In deep waters, it can be estimated based on thesignificant wave height HS,ESS with the relevant return periodas outlined in 307. The Extreme Wave Height is used for returnperiods of 50 years and 1 year, and the corresponding wave

)11(11,max

RHT T

FH yearR −= −

gUHTgUH NSSSNSSS )(3.14)(1.11 10,10, ≤≤

gUHTgUH SSSSSSSS )(3.14)(1.11 10,10, ≤≤

DET NORSKE VERITAS


heights are denoted H50-yr and H1-yr, respectively. The rangeof wave periods T appropriate to the severe wave height shallbe considered. Design calculations shall be based on values ofthe wave period within this range that result in the highestloads or load effects in the structure.

Guidance note:In deep waters, the wave periods T to be used with HEWH may beassumed to be within the range given by


408 The Reduced Wave Height (RWH) HRWH is a compan-ion wave height to be used in combination with the extremewind speed (EWS) for definition of an extreme event with aspecified return period. The reduced wave height can beexpressed as a fraction of the extreme wave height,HRWH = ψ · HEWH, ψ < 1. The Reduced Wave Height is usedfor definition of events with return periods of 50 years and 1year, and the corresponding reduced wave heights are denotedHRed,50-yr and HRed,1-yr, respectively.

Guidance note:It is practice for offshore structures to apply ψ = H5-yr/H50-yr, whereH5-yr and H50-yr denote the individual wave heights with 5- and 50-year return period, respectively. The shallower the water depth, thelarger is usually the value of ψ.


409 The wave period T associated with the wave heights in403, 405, 407 and 408 has a depth-dependent lower limitderived from wave breaking considerations,

where H is the wave height, d is the water depth and g is theacceleration of gravity.

C 500 Wave theories and wave kinematics501 The kinematics of regular waves may be represented by

analytical or numerical wave theories, which are listed below:

— linear wave theory (Airy theory) for small-amplitude deepwater waves; by this theory the wave profile is representedby a sine function

— Stokes wave theories for high waves— stream function theory, based on numerical methods and

accurately representing the wave kinematics over a broadrange of water depths

— Boussinesq higher-order theory for shallow water waves— solitary wave theory for waves in particularly shallow water.

502 Three wave parameters determine which wave theory toapply in a specific problem. These are the wave height H, thewave period T and the water depth d. These parameters areused to define three non-dimensional parameters that deter-mine ranges of validity of different wave theories,

— Wave steepness parameter:

— Shallow water parameter:

— Ursell parameter:

where λ 0 and κ 0 are the linear deepwater wavelength and wavenumber corresponding to wave period T. The ranges of applica-tion of the different wave theories are given in Table C2.

Figure 7 shows the ranges of validity for different wave theories.

Figure 7 Ranges of validity for wave theories

gUHTgUH ESSSESSS )(3.14)(1.11 10,10, ≤≤

)78.0

(tanh5.34 1

dH

gdT −>

Table C2 Ranges of application of regular wave theories

TheoryApplication

Depth Approximate rangeLinear (Airy) wave Deep and shallow S < 0.006; S/μ < 0.032nd order Stokes wave Deep water Ur < 0.65; S < 0.045th order Stokes wave Deep water Ur < 0.65; S < 0.14Cnoidal theory Shallow water Ur > 0.65; μ < 0.125

02

2λ

π HgTHS ==

02

2λ

πμ dgT

d==

3232 41

0μπS

dkHU r ==

DET NORSKE VERITAS


503 Linear wave theory is the simplest wave theory and isobtained by taking the wave height to be much smaller thanboth the wavelength and the water depth, or equivalently

This theory is referred to as small amplitude wave theory, lin-ear wave theory, sinusoidal wave theory or as Airy theory. For regular linear waves the wave crest height AC is equal tothe wave trough height AH and denoted the wave amplitude A,hence H = 2A.The surface elevation is given by

where and β is the direction of prop-agation, measured from the positive x-axis. The dispersion relationship gives the relationship betweenwave period T and wavelength λ. For waves in waters withfinite water depth d the dispersion relationship is given by thetranscendental equation

in which g denotes the acceleration of gravity. A good approx-imation to the wavelength λ as a function of the wave period Tis given by

where

and

α1 = 0.666, α2 = 0.445, α3 = –0.105, α4 = 0.272.504 Stokes wave theory implies the Stokes wave expansion,which is an expansion of the surface elevation in powers of thelinear wave height H. A Stokes wave expansion can be shownto be formally valid for

A first-order Stokes wave is identical to a linear wave, or Airywave. A second-order Stokes wave is a reasonably accurateapproximation when

The surface elevation profile for a regular second-order Stokeswave is given by

where . Second-order and higherorder Stokes waves are asymmetric with AC > AT. Crests aresteeper and troughs are wider than for Airy waves. The linear dispersion relation holds for second-order Stokeswaves, hence the phase velocity c and the wavelength λ remainindependent of wave height.To third order, the phase velocity depends on wave height

according to

The wave height is limited by breaking. The maximum steep-ness is

where λ is the wavelength corresponding to water depth d. For deep water the breaking wave limit is approximated bySmax = 1/7. Use of second order Stokes waves is limited by the steepnesscriterion

For regular steep waves S < Smax (and Ur < 0.65) Stokes fifthorder wave theory applies.Stokes wave theory is not applicable for very shallow waterwhere cnoidal wave theory or “Stream Function” wave theoryshould be used.505 Cnoidal wave theory defines a wave which is a periodicwave with sharp crests separated by wide troughs. The rangeof validity of cnoidal wave theory isμ < 0.125 and Ur > 0.65The surface profile of cnoidal waves of wave height H andperiod T in water depth d is given by

where K, E are the complete elliptic integrals of the first andsecond kind respectively, cn is the Jacobian elliptic functionand k is a parameter determined implicitly as a function of Hand T by the formulae

506 The “Stream Function” wave theory is a purely numeri-cal procedure for approximating a given wave profile and hasa broader range of validity than the wave theories in 503through 505.A stream function wave solution has the general form

where c is the wave celerity and N is the order of the wave the-ory. The required order, N, of the stream function theory, rang-ing from 1 to 10, is determined by the wave parameters S andμ. The closer to the breaking wave height, the more terms arerequired in order to give an accurate representation of thewave. Figure 8 shows the required order N of stream functionwave theory such that errors in maximum velocity and accel-eration are less than one percent.

1;1 <<<< rUS

Θ= cos2

),,( Htyxη

)sincos( ctyxk −+=Θ ββ

⎟⎠⎞

⎜⎝⎛=

λπ

πλ dgT 2tanh

2

2

2/1

2/1

)(1)()( ⎟⎟

⎠

⎞⎜⎜⎝

⎛+

=ϖϖ

ϖλf

fgdT

n

nnf ϖαϖ ∑

=

+=4

1

1)(

)/()4( 22 gTdπϖ =

1;1 <<<< rUS

65.0and04.0 << rUS

[ ] Θ++Θ= 2cos2cosh2sinhcosh

8cos

2 3

2

kdkdkdHH

λπη

)sincos( ctyxk −+=Θ ββ

⎪⎭

⎪⎬⎫

⎪⎩

⎪⎨⎧

⎥⎦

⎤⎢⎣

⎡ +−⎟⎠⎞

⎜⎝⎛+=

)(sinh8)(cosh8)(cosh89

21)tanh( 4

4222

kdkdkdkHkd

kgc

λπ

λdHS 2tanh142.0max ==

⎟⎟⎠

⎞⎜⎜⎝

⎛

+=

kd

kdkH3

3

cosh81sinh924.0

[ ]{ }

⎥⎦⎤

⎢⎣⎡ −+

−+−=

kTtxkKHcn

dHkEkKkKdtx

),)((2

1)()()(3

16),(

2

2

2

λ

λη

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛−+=

⎟⎟⎠

⎞⎜⎜⎝

⎛=

=

)()(

2111)()(

)(3

16)(

)()()(

22/1

2/13

kKkE

kdHgdkc

kkKHdk

kckkT

λ

λ

∑=

++=ΨN

n

nkxdznknXczzx1

cos)(sinh)(),(

DET NORSKE VERITAS


Figure 8 Required order N of stream function wave theory

C 600 Breaking waves601 Wave breaking may take place as a result of shoalingand limited water depth. Such breaking may take place eitherbefore the waves arrive at the site or when they have arrived atthe site. In both cases, the wave breaking implies that a depth-dependent limitation is imposed on the waves at the site. Thisdepth dependency shall be taken into account when waveheights for use in design are to be determined. For this deter-mination, the water depth corresponding to the maximumwater level on the site shall be assumed. The breaking criterionis identified in Figure 1. Breaking waves are irregular waves,for which the kinematics deviate from those implied by thewave theories referenced in 503 through 506. The kinematicsof breaking waves depends on the type of breaking. 602 There are three types of breaking waves depending onthe wave steepness and the slope of the seabed:

— surging breaker— plunging breaker— spilling breaker.

Figure 9 indicates which type of breaking wave can be expectedas a function of the slope of the seabed and as a function of thewave period T and the wave height H0 in deep waters.

Figure 9 Transitions between different types of breaking waves as a func-tion of seabed slope, wave height in deep waters and wave period

D. CurrentD 100 Current parameters101 The current consists of a wind-generated current and atidal current, and a density current when relevant. 102 The current is represented by the wind-generated currentvelocity vwind0 at the still water level and the tidal currentvelocity vtide0 at the still water level.103 Other current components than wind-generated currents,tidal currents and density currents may exist. Examples of suchcurrent components are

— subsurface currents generated by storm surge and atmos-pheric pressure variations

— near-shore, wave-induced surf currents running parallel tothe coast.

D 200 Current data201 Current statistics are to be used as a basis for representa-tion of the long-term and short-term current conditions. Empir-ical statistical data used as a basis for design must cover asufficiently long period of time.

Guidance note:Current data obtained on site are to be preferred over current dataobserved at an adjacent location. Measured current data are to bepreferred over visually observed current data. Continuousrecords of data are to be preferred over records with gaps. Longerperiods of observation are to be preferred over shorter periods.


202 The variation of the current with the water depth shall beconsidered when relevant.203 In regions where bottom material is likely to erode, spe-cial studies of current conditions near the sea bottom may berequired.

D 300 Current modelling301 When detailed field measurements are not available, thevariation in current velocity with depth may be taken as

where

for z ≤ 0and

for –h0 ≤ z ≤ 0in which

v(z) = total current velocity at level zz = distance from still water level, positive upwardsvtide0 = tidal current at still water levelvwind0 = wind-generated current at still water levelh = water depth from still water level (taken as positive)h0 = reference depth for wind-generated current;

h0 = 50 m.302 The variation in current profile with variation in waterdepth due to wave action shall be accounted for. In such cases,the current profile may be stretched or compressed vertically,such that the current velocity at any proportion of the instanta-neous depth is kept constant. By this approach, the surface cur-rent component remains constant, regardless of the sea

)()()( zvzvzv windtide +=

71

0)( ⎟⎠⎞

⎜⎝⎛ +

=h

zhvzv tidetide

⎟⎟⎠

⎞⎜⎜⎝

⎛ +⋅=

0

00)(

hzh

vzv windwind

DET NORSKE VERITAS


elevation during the wave action.303 Unless data indicate otherwise, the wind-generated cur-rent at still water level may be estimated as

whereU0 = 1-hour mean wind speed at 10 m height.

E. Water LevelE 100 Water level parameters101 The water level consists of a mean water level in con-junction with tidal water and a wind- and pressure-inducedstorm surge. The tidal range is defined as the range between thehighest astronomical tide (HAT) and the lowest astronomicaltide (LAT), see Figure 10.

Figure 10 Definition of water levels

E 200 Water level data201 Water level statistics are to be used as a basis for repre-sentation of the long-term and short-term water level condi-tions. Empirical statistical data used as a basis for design mustcover a sufficiently long period of time.

Guidance note:Water level data obtained on site are to be preferred over waterlevel data observed at an adjacent location. Measured water leveldata are to be preferred over visually observed water level data.Continuous records of data are to be preferred over records withgaps. Longer periods of observation are to be preferred overshorter periods.


202 Water level and wind are correlated, because the waterlevel has a wind-generated component. The correlationbetween water level data and wind data shall be accounted forin design.

Guidance note:Simultaneous observations of water level and wind data in termsof simultaneous values of water level and U10 should beobtained.


E 300 Water level modelling301 For determination of the water level for calculation ofloads and load effects, both tidal water and pressure- and wind-induced storm surge shall be taken into account.

Guidance note:Water level conditions are of particular importance for predictionof depth-limited wave heights.


F. IceF 100 Sea ice101 When the wind turbine structure is to be located in anarea where ice may develop or where ice may drift, ice condi-tions shall be properly considered.102 Relevant statistical data for the following sea ice condi-tions and properties shall be considered:

— geometry and nature of ice— concentration and distribution of ice— type of ice (ice floes, ice ridges, rafted ice etc.)— mechanical properties of ice (compressive strength ru,

bending strength rf)— velocity and direction of drifting ice— thickness of ice— probability of encountering icebergs.

F 200 Snow and ice accumulation201 Ice accretion from sea spray, snow and rain and airhumidity shall be considered wherever relevant.202 Snow and ice loads due to snow and ice accumulationmay be reduced or neglected if a snow and ice removal proce-dure is established.203 Possible increases of cross-sectional areas and changesin surface roughness caused by icing shall be considered wher-ever relevant, when wind loads and hydrodynamic loads are tobe determined.204 For buoyant structures, the possibility of uneven distri-bution of snow and ice accretion shall be considered.

F 300 Ice modelling301 The ice thickness forms an important parameter for cal-culation of ice loads. The ice thickness shall be based on localice data, e.g. as available in an ice atlas or as derived from frostindex data.302 As a basis for design against ice loads, the frost index Kmay be used. The frost index for a location is defined as theabsolute value of the sum of the daily mean temperature overall days whose mean temperature is less than 0° C in one year.The frost index K exhibits variability from year to year and canbe represented by its probability distribution.

Guidance note:Unless data indicate otherwise, the frost index may be repre-sented by a three-parameter Weibull distribution,


303 The frost index with return period TR in units of years isdefined as the (1−1/TR) quantile in the distribution of the frostindex, i.e. it is the frost index whose probability of exceedancein one year is 1/TR. It is denoted KTR and is expressed as

304 The ice thickness t at the end of a frost period can be esti-mated by

where t is in units of metres and K is the frost index in units ofdegree-days.305 In near-coastal waters and in sheltered waters, such as inlakes and archipelagos, the ice sheet is normally not movingafter having grown to some limiting thickness, tlimit. The lim-

00 01.0 Uvwind ⋅=

))(exp(1)( βa

bkkFK−

−−=

)11(1

RKT T

FK R −= −

509.0032.0 −= Kt

DET NORSKE VERITAS


iting thickness can therefore be used to define extreme thick-ness events for moving ice in such waters. Unless data indicateotherwise, the limiting thickness tlimit can be taken as the long-term mean value of the annual maximum ice thickness. Nosuch limiting thickness is associated with moving ice in opensea, for which larger thicknesses can therefore be expected inthe extreme thickness events.

Guidance note:The long-term mean value of the annual maximum ice thicknessmay be interpreted as a measure of the ice thickness associatedwith a "normal winter".


306 The compression strength ru, the bending strength rf andthe thickness of the ice may be expressed as functions of thefrost index or, alternatively, in terms of their respective proba-bility distributions. Other location-dependent parameterswhich may need to be considered are the floe size and the driftspeed of floes.307 Unless data indicate otherwise, the following generalvalues of ice parameters apply, regardless of location:

G. Soil Investigations and Geotechnical DataG 100 Soil investigations101 The soil investigations shall provide all necessary soildata for a detailed design. The soil investigations may bedivided into geological studies, geophysical surveys and geo-technical soil investigations.

Guidance note:A geological study, based on the geological history, can form abasis for selection of methods and extent of the geotechnical soilinvestigations. A geophysical survey, based on shallow seismic,can be combined with the results from a geotechnical soil inves-tigation to establish information about soil stratification and sea-bed topography for an extended area such as the area covered bya wind farm. A geotechnical soil investigation consists of in-situtesting of soil and of soil sampling for laboratory testing.


102 The extent of soil investigations and the choice of soilinvestigation methods shall take into account the type, size andimportance of the wind turbine structure, the complexity ofsoil and seabed conditions and the actual type of soil deposits.The area to be covered by soil investigations shall account forpositioning and installation tolerances.

Guidance note:The line spacing of the seismic survey at the selected locationshould be sufficiently small to detect all soil strata of significancefor the design and installation of the wind turbine structures. Spe-cial concern should be given to the possibility of buried erosionchannels with soft infill material.


103 For multiple foundations such as in a wind farm, the soilstratigraphy and range of soil strength properties shall beassessed within each group of foundations or per foundationlocation, as relevant.

Guidance note:Whether the soil stratigraphy and range of soil strength proper-ties shall be assessed within each group of foundations or perfoundation location is much a function of the degree to which thesoil deposit can be considered as homogeneous. Thus, when veryhomogeneous soil conditions prevail, the group of foundations tobe covered by such a common assessment may consist of all thefoundations within the entire area of a wind farm or it may con-sist of all the foundations within a sub-area of a wind farm. Suchsub-areas are typically defined when groups of wind turbineswithin the wind farm are separated by kilometre-wide straits ortraffic corridors. When complex or non-homogeneous soil con-ditions prevail, it may be necessary to limit common assessmentsof the soil stratigraphy and soil strength properties to cover onlya few close foundations, and in the ultimate case to carry out indi-vidual assessments for individual foundations.


104 Soil investigations shall provide relevant informationabout the soil to a depth below which possible existence ofweak formations will not influence the safety or performanceof the wind turbine and its support structure and foundation.

Guidance note:For design of pile foundations against lateral loads, a combina-tion of CPTs and soil borings with sampling should be carried outto sufficient depth. For slender and flexible piles in jacket typefoundations, a depth of about 10 pile diameters suffices. For lessflexible monopiles with larger diameters, a depth equal to the pilepenetration plus half a pile diameter suffices.For design of piles against axial loads, at least one CPT and onenearby boring should be carried out to the anticipated penetrationdepth of the pile plus a zone of influence. If potential end bearinglayers or other dense layers, which may create driving problems,are found this scope should be increased. For design of gravity base foundations, the soil investigationsshould extend at least to the depth of any critical shear surface.Further, all soil layers influenced by the wind turbine structurefrom a settlement point of view should be thoroughly investi-gated.In seismic areas, it may be necessary to obtain information aboutthe shear modulus of the soil to large depths.


105 Soil investigations are normally to comprise the follow-ing types of investigation:

— site geology survey— topography survey of the seabed— geophysical investigations for correlation with soil bor-

ings and in-situ testing— soil sampling with subsequent laboratory testing— in-situ tests, e.g. cone penetration tests (CPT).

Guidance note:The extent and contents of a soil investigation program are nostraight-forward issue and will depend on the foundation type.The guidance given in this guidance note therefore forms recom-mendations of a general nature which the designer, either on hisown initiative or in cooperation with the classification society,may elaborate further on.An experienced geotechnical engineer who is familiar with theconsidered foundation concepts and who represents the owner ordeveloper should be present during the soil investigations on thesite. Depending on the findings during the soil investigations,actions may then be taken, as relevant, to change the soil investi-gation program during its execution. This may include sugges-tions for increased depths of planned soil borings, suggestionsfor additional soil borings, and suggestions for changed positionsof soil borings. When non-homogeneous soil deposits are encountered or whendifficult or weak soils are identified locally, it may be necessaryto carry out more soil borings and CPTs than the tentative mini-mum recommended below.

Density 900 kg/m3

Unit weight 8.84 kN/m3

Modulus of elasticity 2 GPaPoisson’s ratio 0.33Ice-ice frictional coefficient 0.1Ice-concrete dynamic frictional coefficient 0.2Ice-steel dynamic frictional coefficient 0.1

DET NORSKE VERITAS


For solitary wind turbine structures, one soil boring to sufficientdepth for recovery of soil samples for laboratory testing is recom-mended as a minimum.For wind turbine structures in a wind farm, a tentative minimumsoil investigation program may contain one CPT per foundationin combination with one soil boring to sufficient depth in eachcorner of the area covered by the wind farm for recovery of soilsamples for laboratory testing. An additional soil boring in themiddle of the area will provide additional information about pos-sible non-homogeneities over the area. For cable routes, the soil investigations should be sufficientlydetailed to identify the soils of the surface deposits to the planneddepth of the cables along the routes.Seabed samples should be taken for evaluation of scour potential.


106 For further guidance and industry practice regardingrequirements to scope, execution and reporting of offshore soilinvestigations, and to equipment, reference is made to DNVClassification Notes No. 30.4, NORSOK N-004 (App. K) andNORSOK G-001. National and international standards may beconsidered from case to case, if relevant.107 The geotechnical investigation at the actual site com-prising a combination of sampling with subsequent laboratorytesting and in situ testing shall provide the following types ofgeotechnical data for all important layers:

— data for soil classification and description— shear strength and deformation properties, as required for

the type of analysis to be carried out— in-situ stress conditions.

The soil parameters provided shall cover the scope required fora detailed and complete foundation design, including the lat-eral extent of significant soil layers, and the lateral variation ofsoil properties in these layers.108 The laboratory test program for determination of soilstrength and deformation properties shall cover a set of differ-ent types of tests and a number of tests of each type, which willsuffice to carry out a detailed foundation design.

Guidance note:For mineral soils, such as sand and clay, direct simple shear testsand triaxial tests are relevant types of tests for determination ofstrength properties.For fibrous peats, neither direct simple shear tests nor triaxialtests are recommended for determination of strength properties.Shear strength properties of low-humified peat can be deter-mined by ring shear tests.


H. Other Site ConditionsH 100 Seismicity101 The level of seismic activity of the area where the windturbine structure is to be installed shall be assessed on the basisof previous record of earthquake activity as expressed in termsof frequency of occurrence and magnitude.102 For areas where detailed information on seismic activityis available, the seismicity of the area may be determined fromsuch information.103 For areas where detailed information on seismic activityis not available, the seismicity is to be determined on the basisof detailed investigations, including a study of the geologicalhistory and the seismic events of the region.104 If the area is determined to be seismically active and thewind turbine structure will be affected by an earthquake, anevaluation shall be made of the regional and local geology inorder to determine the location and alignment of faults, epicen-

tral and focal distances, the source mechanism for energyrelease and the source-to-site attenuation characteristics. Localsoil conditions shall be taken into account to the extent thatthey may affect the ground motion. The seismic design, includ-ing the development of the seismic design criteria for the site,shall be in accordance with recognised industry practice.105 The potential for earthquake-induced sea waves, alsoknown as tsunamis, shall be assessed as part of the seismicityassessment.106 For details of seismic design criteria, reference is madeto ISO 19901-2.

H 200 Salinity201 The salinity of the seawater shall be addressed with aview to its influence with respect to corrosion.

H 300 Temperature301 Extreme values of high and low temperatures are to beexpressed in terms of the most probable highest and lowest val-ues, respectively, with their corresponding return periods.302 Both air and seawater temperatures are to be consideredwhen describing the temperature environment.

H 400 Marine growth401 The plant, animal and bacteria life on the site causesmarine growth on structural components in the water and in thesplash zone. The potential for marine growth shall beaddressed. Marine growth adds weight to a structural compo-nent and influences the geometry and the surface texture of thecomponent. The marine growth may hence influence thehydrodynamic loads, the dynamic response, the accessibilityand the corrosion rate of the component.

Guidance note:Marine growth can broadly be divided into hard growth and softgrowth. Hard growth generally consists of animal growth such asmussels, barnacles and tubeworms, whereas soft growth consistsof organisms such as hydroids, sea anemones and corals. Marinegrowth may also appear in terms of seaweeds and kelps. Marineorganisms generally colonise a structure soon after installation,but the growth tapers off after a few years.The thickness of marine growth depend on the position of thestructural component relative to the sea level, the orientation ofthe component relative to the sea level and relative to the domi-nant current, the age of the component, and the maintenancestrategy for the component.Marine growth also depends on other site conditions such assalinity, oxygen content, pH value, current and temperature.The corrosive environment is normally modified by marinegrowth in the upper submerged zone and in the lower part of thesplash zone of the structural component. Depending on the typeof marine growth and on other local conditions, the net effectmay be either an enhancement or a retardation of the corrosionrate. Marine growth may also interfere with systems for corro-sion protection, such as coating and cathodic protection.


H 500 Air density501 Air density shall be addressed since it affects the struc-tural design through wind loading.

H 600 Ship traffic601 Risk associated with possible ship collisions shall beaddressed as part of the basis for design of support structuresfor offshore wind turbines.602 For service vessel collisions, the risk can be managed bydesigning the support structure against relevant service vesselimpacts. For this purpose the limit state shall be considered asa ULS. The service vessel designs and the impact velocities tobe considered are normally specified in the design basis forstructural design.

DET NORSKE VERITAS


H 700 Disposed matters701 The presence of obstacles and wrecks within the area ofinstallation shall be mapped.

H 800 Pipelines and cables801 The presence of pipelines and cables within the area ofinstallation shall be mapped.

DET NORSKE VERITAS


SECTION 4LOADS AND LOAD EFFECTS

A. IntroductionA 100 General101 The requirements in this section define and specify loadcomponents and load combinations to be considered in theoverall strength analysis as well as design pressures applicablein formulae for local design.102 It is a prerequisite that the wind turbine and supportstructure as a minimum meet the requirements to loads givenin IEC61400-1 for site-specific wind conditions.

B. Basis for Selection of Characteristic LoadsB 100 General101 Unless specific exceptions apply, as documented withinthis standard, the basis for selection of characteristic loads orcharacteristic load effects specified in 102 and 103 shall applyin the temporary and operational design conditions, respec-tively.

Guidance note:Temporary design conditions cover design conditions duringtransport, assembly, maintenance, repair and decommissioningof the wind turbine structure.Operational design conditions cover design conditions in the per-manent phase which includes steady conditions such as powerproduction, idling and stand-still as well as transient conditionsassociated with start-up, shutdown, yawing and faults.


102 For the temporary design conditions, the characteristicvalues shall be based on specified values, which shall beselected dependent on the measures taken to achieve therequired safety level. The values shall be specified with dueattention to the actual location, the season of the year, theweather forecast and the consequences of failure. For design

conditions during transport and installation, reference is madeto DNV Rules for Planning and Execution of Marine Opera-tions.103 For the operational design conditions, the basis forselection of characteristic loads and load effects specified inTable B1 refers to statistical terms whose definitions are givenin Table B2.

Guidance note:The environmental loading on support structures and foundationsfor wind turbines does – as far as wind loading is concerned – notalways remain the way it is produced by nature, because the con-trol system of the wind turbine interferes by introducing meas-ures to reduce the loads.


104 Characteristic values of environmental loads or loadeffects, which are specified as the 98% quantile in the distribu-tion of the annual maximum of the load or load effect, shall beestimated by their central estimates.

C. Permanent Loads (G)

C 100 General101 Permanent loads are loads that will not vary in magni-tude, position or direction during the period considered. Exam-ples are:

— mass of structure— mass of permanent ballast and equipment— external and internal hydrostatic pressure of a permanent

nature— reaction to the above, e.g. articulated tower base reaction.

102 The characteristic value of a permanent load is definedas the expected value based on accurate data of the unit, mass

of the material and the volume in question.

D. Variable Functional Loads (Q)

D 100 General101 Variable functional loads are loads which may vary inmagnitude, position and direction during the period under con-sideration, and which are related to operations and normal useof the installation. Examples are:

— personnel— crane operational loads— ship impacts

Table B1 Basis for selection of characteristic loads and load effects for operational design conditions

Limit states – operational design conditionsLoad category ULS FLS SLSPermanent (G) Expected valueVariable (Q) Specified valueEnvironmental (E)

98% quantile in distri-bution of annual maxi-

mum load or load effect (Load or load

effect with return period 50 years)

Expected load history or expected load effect

history

Specified value

Abnormal wind turbine loads

Specified value

Deformation (D) Expected extreme value

Table B2 Statistical terms used for specification of characteristic loads and load effects

Term Return period (years) Quantile in distribution of annual maximum

Probability of exceedance in distribution of annual maximum

100-year value 100 99% quantile 0.0150-year value 50 98% quantile 0.0210-year value 10 90% quantile 0.105-year value 5 80% quantile 0.201-year value 1 Most probable highest value in one year

DET NORSKE VERITAS


— loads from fendering— loads associated with installation operations— loads from variable ballast and equipment— stored materials, equipment, gas, fluids and fluid pressure— lifeboats.

102 For an offshore wind turbine structure, the variable func-tional loads usually consist of:

— actuation loads— loads on access platforms and internal structures such as

ladders and platforms— ship impacts from service vessels— crane operational loads.

103 Actuation loads result from the operation and control ofthe wind turbine. They are in several categories includingtorque control from a generator or inverter, yaw and pitch actu-ator loads and mechanical braking loads. In each case, it isimportant in the calculation of loading and response to con-sider the range of actuator forces available. In particular, formechanical brakes, the range of friction, spring force or pres-sure as influenced by temperature and ageing shall be takeninto account in checking the response and the loading duringany braking event.104 Actuation loads are usually represented as an integratedelement in the wind turbine loads that result from an analysisof the wind turbine subjected to wind loading. They are there-fore in this standard treated as environmental wind turbineloads and do therefore not appear as separate functional loadsin load combinations.105 Loads on access platforms and internal structures areused only for local design of these structures and do thereforeusually not appear in any load combination for design of pri-mary support structures and foundations.106 Loads and dynamic factors from maintenance and serv-ice cranes on structures are to be determined in accordancewith requirements given in DNV Standard for CertificationNo. 2.22 Lifting Appliances, latest edition.107 Ship impact loads are used for the design of primary sup-port structures and foundations and for design of some second-ary structures.108 The characteristic value of a variable functional load isthe maximum (or minimum) specified value, which producesthe most unfavourable load effects in the structure under con-sideration.109 Variable loads can contribute to fatigue. In this casecharacteristic load histories shall be developed based on spec-ified conditions for operation.

Guidance note:For a specified condition for operation, the characteristic loadhistory is often taken as the expected load history.


D 200 Variable functional loads on platform areas201 Variable functional loads on platform areas of the sup-port structure shall be based on Table D1 unless specified oth-erwise in the design basis or the design brief. For offshore windturbine structures, the platform area of most interest is theexternal platform, which shall be designed for ice loads, waveloads and ship impacts. The external platform area consists oflay down area and other platform areas. The intensity of thedistributed loads depends on local or global aspects as given inTable D1. The following notions are used:

D 300 Ship impacts and collisions301 Impacts from approaching ships shall be considered asvariable functional loads. Analyses of such impacts in designshall be carried out as ULS analyses. The impact analyses shallinclude associated environmental loads from wind, waves andcurrent. The added water mass contributes to the kineticenergy of the ship and has to be taken into account.302 For design against ship impact in the ULS, the load shallbe taken as the largest unintended impact load in normal serv-ice conditions. It is a requirement that the support structure andthe foundation do not suffer from damage. Secondary struc-tural parts such as boat landings and ladders shall not sufferfrom damage leading to loss of their respective functions.

Guidance note:A risk analysis forms the backbone of a ship impact analysis. Thelargest unintended impact load is part of the results from the riskanalysis.


D 400 Tank pressures401 Requirements to hydrostatic pressures in tanks are givenin DNV-OS-C101.

Local design: For example design of plates, stiffeners, beams and brackets

Primary design: For example design of girders and columns

Global design: For example design of support structure

Table D1 Variable functional loads on platform areas

Local design Primary design

Global design

Distributed load p

(kN/m2)

Point load,P(kN)

Apply factor to

distributed load

Apply factor to primary design load

Storage areas q 1.5 q 1.0 1.0Lay down areas q 1.5 q f fArea between equipment 5.0 5.0 f may be

ignoredWalkways, staircases and external platforms

4.0 4.0 f may be ignored

Walkways and staircases for inspection only


Internal platforms, e.g. in towers


Areas not exposed to other func-tional loads

2.5 2.5 1.0 –

Notes:

— Point loads are to be applied on an area 100 mm × 100 mm, and at the most severe position, but not added to wheel loads or dis-tributed loads.

— For internal platforms, point loads are to be applied on an area 200 mm × 200 mm

— q to be evaluated for each case. Lay down areas should not be designed for less than 15 kN/m2.

— f = min{1.0 ; (0.5 + 3/ )}, where A is the loaded area in m2.

— Global load cases shall be established based upon “worst case”, characteristic load combinations, complying with the limiting global criteria to the structure. For buoyant structures these cri-teria are established by requirements for the floating position in still water, and intact and damage stability requirements, as documented in the operational manual, considering variable load on the deck and in tanks.

A

DET NORSKE VERITAS


D 500 Miscellaneous loads501 Railing shall be designed for a horizontal line load equalto 1.5 kN/m, applied to the top of the railing.502 Ladders shall be designed for a concentrated load of 2.5kN.503 Requirements given in prEN50308 should be met whenrailing, ladders and other structures for use by personnel aredesigned.

E. Environmental Loads (E)E 100 General101 Environmental loads are loads which may vary in mag-nitude, position and direction during the period under consid-eration, and which are related to operations and normal use ofthe structure. Examples are:

— wind loads — hydrodynamic loads induced by waves and current,

including drag forces and inertia forces— earthquake loads— current-induced loads— tidal effects— marine growth— snow and ice loads.

102 Practical information regarding environmental loadsand environmental conditions is given in DNV-RP-C205.103 According to this standard, characteristic environmentalloads and load effects shall be determined as quantiles withspecified probabilities of exceedance. The statistical analysisof measured data or simulated data should make use of differ-ent statistical methods to evaluate the sensitivity of the result.The validation of distributions with respect to data should betested by means of recognised methods. The analysis of thedata shall be based on the longest possible time period for therelevant area. In the case of short time series, statistical uncer-tainty shall be accounted for when characteristic values aredetermined.

E 200 Wind turbine loads201 Wind-generated loads on the rotor and the tower shall beconsidered. Wind-generated loads on the rotor and the towerinclude wind loads produced directly by the inflowing wind aswell as indirect loads that result from the wind-generatedmotions of the wind turbine and the operation of the wind tur-bine. The direct wind-generated loads consist of

— aerodynamic blade loads (during operation, during park-ing and idling, during braking, and during start-up)

— aerodynamic drag forces on tower and nacelle.

The following loads, which only indirectly are produced bywind and which are a result of the operation of the wind tur-bine, shall be considered as wind loads in structural designaccording to this standard:

— gravity loads on the rotor blades, vary with time due torotation

— centrifugal forces and Coriolis forces due to rotation— gyroscopic forces due to yawing— braking forces due to braking of the wind turbine.

Guidance note:Aerodynamic wind loads on the rotor and the tower may bedetermined by means of aeroelastic load models.Gyroscopic loads on the rotor will occur regardless of the struc-tural flexibility whenever the turbine is yawing during operationand will lead to a yaw moment about the vertical axis and a tilt

moment about a horizontal axis in the rotor plane. For yawspeeds below 0.5°/s gyroscopic loads can be disregarded.


202 For determination of wind loads, the following factorsshall be considered:

— tower shadow, tower stemming and vortex shedding,which are disturbances of the wind flow owing to the pres-ence of the tower

— wake effects wherever the wind turbine is located behindother turbines such as in wind farms

— misaligned wind flow relative to the rotor axis, e.g. owingto a yaw error

— rotational sampling, i.e. low-frequent turbulence will betransferred to high-frequent loads due to blades cuttingthrough vortices

— aeroelastic effects, i.e., the interaction between the motionof the turbine on the one hand and the wind field on theother

— aerodynamic imbalance and rotor-mass imbalance due todifferences in blade pitch

— influence of the control system on the wind turbine, forexample by limiting loads through blade pitching

— turbulence and gusts— instabilities caused by stall-induced flapwise and edge-

wise vibrations must be avoided— damping— wind turbine controller.

Guidance note:The damping comes about as a combination of structural damp-ing and aerodynamic damping. The structural damping dependson the blade material and material in other components such asthe tower. The aerodynamic damping can be determined as theoutcome of an aeroelastic calculation in which correct propertiesfor the aerodynamics are used.The coherence of the wind and the turbulence spectrum of thewind are of significant importance for determination of towerloads such as the bending moment in the tower.


203 Wind turbine loads during power production andselected transient events shall be verified by load measure-ments that cover the intended operational range, i.e. windspeeds between cut-in and cut-out. Measurements shall be car-ried out by an accredited testing laboratory or the certifyingbody shall verify that the party conducting the testing as a min-imum complies with the criteria set forth in ISO/IEC 17020 orISO/IEC 17025, as applicable.204 For design of the support structure and the foundation, anumber of load cases for wind turbine loads due to wind loadon the rotor and on the tower shall be considered, correspond-ing to different design situations for the wind turbine. Differentdesign situations may govern the designs of different parts ofthe support structure and the foundation.The load cases shall be defined such that it is ensured that theycapture the 50-year load or load effect, as applicable, for eachstructural part to be designed in the ULS. Likewise, the loadcases shall be defined such that it is ensured that they captureall contributions to fatigue damage for design in the FLS.Finally, the load cases shall include load cases to adequatelycapture abnormal conditions associated with severe fault situ-ations for the wind turbine in the ULS.Because the wind turbine loads occur concurrently with otherenvironmental loads such as loads from waves, current andwater level, the load cases to be considered shall specify notonly the wind turbine load conditions, but also their compan-ion wave load conditions, current conditions and water levelconditions.Table E1 specifies a proposal for 31 load cases to consider for

DET NORSKE VERITAS


wind turbine load conditions and their companion wave loadconditions, current conditions and water level conditions inorder to fulfil the requirements in this item. The load cases inTable E1 refer to design in the ULS and in the FLS and includea number of abnormal load cases for the ULS.The load cases in Table E1 are defined in terms of wind condi-tions, which are characterised by wind speed. For most of theload cases, the wind speed is defined as a particular 10-minutemean wind speed plus a particular extreme coherent gust,which forms a perturbation on the mean wind speed. Extremecoherent gusts are specified in Sec.3 B505. Some load cases inTable E1 refer to the normal wind profile. The normal windprofile is given in Sec.3. For each specified load case in Table E1, simulations forsimultaneously acting wind and waves based on the wavesgiven in the 4th column of Table E1 can be waived when it can

be documented that it is not relevant to include a wave load orwave load effect for the design of a structural part in question.

Guidance note:The 31 proposed load cases in Table E1 corresponds to 31 loadcases defined in the committee draft of the coming standardIEC61400-3 on the basis of the load cases in IEC61400-1. The 31load cases defined in the committee draft of IEC61400-3 are sub-ject to discussion and may become subject to modifications.Wind load case 1.4 is usually only relevant for design of the topof the tower, and wave loading may only in rare cases have animpact on the design of this structural part.For analysis of the dynamic behaviour of the wind turbine and itssupport structure for concurrently acting wind and waves, it isimportant to carry out the analysis using time histories of bothwind and waves or relevant dynamic amplification factors shouldbe applied to a constant wind speed or individual wave height.


DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 – Sec.4

Tabl

e E

1 Pr

opos

ed lo

ad c

ases

com

bini

ng v

ario

us e

nvir

onm

enta

l con

ditio

ns

Des

ign

situ

atio

nLo

ad

case

Win

d co

nditi

on: W

ind

clim

ate

(U10

,hub

) or

win

d sp

eed

(Uhu

b)W

ave

cond

ition

: Sea

stat

e (H

S) o

r in

divi

dual

wav

e hei

ght (

H) t

o co

m-

bine

with

in si

mul

atio

ns fo

r sim

ul-

tane

ous w

ind

and

wav

es (7

)

Win

d an

d w

ave

dire

ctio

nalit

yC

urre

ntW

ater

leve

lO

ther

co

nditi

ons

Lim

it st

ate

Pow

er

prod

uctio

n1.

1N

TMv i

n < U

10,h

ub <

vou

tN

SSH

S =

E[H

S|U10

,hub

]C

odire

ctio

nal i

n on

e di

rect

ion

Win

d-ge

nera

ted

curr

ent

MW

LFo

r pre

dict

ion

of

extre

me

load

s on

RN

A a

nd in

ter-

face

to to

wer

ULS

1.2

NTM

v in <

U10

,hub

< v

out

NSS

HS

acco

rdin

g to

join

t pr

obab

ility

dis

tribu

tion

of H

S, T

P an

d U

10,h

ub

Cod

irect

iona

l in

one

dire

ctio

n (S

ee

F900

)

(5)

Ran

ge b

etw

een

uppe

r and

low

er

1-ye

ar w

ater

leve

l

FLS

1.3

ETM

v in <

U10

,hub

< v

out

NSS

HS

= E[

HS|U

10,h

ub]

Cod

irect

iona

l in

one

dire

ctio

nW

ind-

gene

rate

d cu

rren

tM

WL

ULS

1.4

ECD

U10

hub

= v r

− 2

m/s

, vr,

v r+2

m/s

N

SSH

S =

E[H

S|U10

,hub

] or

NW

HH

= E

[HS|U

10,h

ub] (

3)

Mis

alig

ned

Win

d-ge

nera

ted

curr

ent

MW

LU

LS

1.5

EWS

v in <

U10

,hub

< v

out

NSS

HS

= E[

HS|U

10,h

ub]

or N

WH

H =

E[H

S|U10

,hub

] (3)

Cod

irect

iona

l in

one

dire

ctio

nW

ind-

gene

rate

d cu

rren

tM

WL

ULS

1.6a

NTM

v in <

U10

,hub

< v

out

SSS

HS

= H

S,50

-yr

(See

item

F70

3)

Cod

irect

iona

l in

one

dire

ctio

nW

ind-

gene

rate

d cu

rren

t1-

year

wat

er le

vel

(4)

ULS

1.6b

NTM

v in <

U10

,hub

< v

out

SWH

H =

H50

-yr

(See

item

F70

3)

Cod

irect

iona

l in

one

dire

ctio

nW

ind-

gene

rate

d cu

rren

t1-

year

wat

er le

vel

(4)

ULS

Pow

er

prod

uctio

n pl

us

occu

rren

ce

of fa

ult

2.1

NTM

v in <

U10

,hub

< v

out

NSS

HS

= E[

HS|U

10,h

ub]

Cod

irect

iona

l in

one

dire

ctio

nW

ind-

gene

rate

d cu

rren

tM

WL

Con

trol s

yste

m

faul

t or l

oss o

f el

ectri

cal c

onne

c-tio

n

ULS

2.2

NTM

v in <

U10

,hub

< v

out

NSS

HS

= E[

HS|U

10,h

ub]

Cod

irect

iona

l in

one

dire

ctio

nW

ind-

gene

rate

d cu

rren

tM

WL

Prot

ectio

n sy

stem

fa

ult o

r pre

cedi

ng

inte

rnal

ele

ctric

al

faul

t

ULS

A

bnor

mal

2.3

EOG

U10

,hub

= v

out a

nd v

r ± 2

m/s

NSS

HS

= E[

HS|U

10,h

ub]

or N

WH

H =

E[H

S|U10

,hub

] (3)

(6)

Cod

irect

iona

l in

one

dire

ctio

nW

ind-

gene

rate

d cu

rren

tM

WL

Exte

rnal

or i

nter

-na

l ele

ctric

al fa

ult

incl

udin

g lo

ss o

f el

ectri

cal n

etw

ork

conn

ectio

n

ULS

A

bnor

mal

2.4

NTM

v in <

U10

,hub

< v

out

NSS

HS

= E[

HS|U

10,h

ub]

Cod

irect

iona

l in

one

dire

ctio

n (S

ee

F900

)

(5)

Ran

ge b

etw

een

uppe

r and

low

er

1-ye

ar w

ater

leve

l

Con

trol o

r pro

tec-

tion

syst

em fa

ult

incl

udin

g lo

ss o

f el

ectri

cal n

etw

ork

FLS

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Sec.4 –

Star

t up

3.1

NW

Pv i

n < U

10,h

ub <

vou

t+

norm

al w

ind

prof

ile to

find

ave

rage

ver

tical

w

ind

shea

r acr

oss s

wep

t are

a of

roto

r

NSS

HS

= E[

HS|U

10,h

ub]

or N

WH

H =

E[H

S|U10

,hub

] (3)

Cod

irect

iona

l in

one

dire

ctio

n (S

ee F

900)

(5)

Ran

ge b

etw

een

uppe

r and

low

er

1-ye

ar w

ater

leve

l

FLS

3.2

EOG

U10

,hub

= v

in, v

out a

nd v

r ± 2

m/s

NSS

HS

= E[

HS|U

10,h

ub]

or N

WH

H =

E[H

S|U10

,hub

] (3)

Cod

irect

iona

l in

one

dire

ctio

nW

ind-

gene

rate

d cu

rren

tM

WL

ULS

3.3

EDC

U10

,hub

= v

in, v

out a

nd v

r ±

2 m

/sN

SSH

S =

E[H

S|U10

,hub

] or

NW

HH

= E

[HS|U

10,h

ub] (

3)

Mis

alig

ned

Win

d-ge

nera

ted

curr

ent

MW

LU

LS

Nor

mal

sh

utdo

wn

4.1

NW

Pv i

n < U

10,h

ub <

vou

t+

norm

al w

ind

prof

ile to

find

ave

rage

ver

tical

w

ind

shea

r acr

oss s

wep

t are

a of

roto

r

NSS

HS

= E[

HS|U

10,h

ub]

or N

WH

H =

E[H

S|U10

,hub

] (3)

Cod

irect

iona

l in

one

dire

ctio

n (S

ee F

900)

(5)

Ran

ge b

etw

een

uppe

r and

low

er

1-ye

ar w

ater

leve

l

FLS

4.2

EOG

U10

,hub

= v

out a

nd v

r ± 2

m/s

NSS

HS

= E[

HS|U

10,h

ub]

or N

WH

H =

E[H

S|U10

,hub

] (3)

Cod

irect

iona

l in

one

dire

ctio

nW

ind-

gene

rate

d cu

rren

tM

WL

ULS

Emer

genc

y sh

utdo

wn

5.1

NTM

U10

,hub

= v

out a

nd v

r ± 2

m/s

NSS

HS

= E[

HS|U

10,h

ub]

Cod

irect

iona

l in

one

dire

ctio

nW

ind-

gene

rate

d cu

rren

tM

WL

ULS

Tabl

e E

1 Pr

opos

ed lo

ad c

ases

com

bini

ng v

ario

us e

nvir

onm

enta

l con

ditio

ns (C

ontin

ued)

D

esig

n si

tuat

ion

Load

ca

seW

ind

cond

ition

: Win

d cl

imat

e (U

10,h

ub) o

r w

ind

spee

d (U

hub)

Wav

e co

nditi

on: S

ea st

ate

(HS)

or

indi

vidu

al w

ave h

eigh

t (H

) to

com

-bi

ne w

ith in

sim

ulat

ions

for s

imul

-ta

neou

s win

d an

d w

aves

(7)

Win

d an

d w

ave

dire

ctio

nalit

yC

urre

ntW

ater

leve

lO

ther

co

nditi

ons

Lim

it st

ate

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 – Sec.4

Park

ed

(sta

ndin

g st

ill

or id

ling)

6.1a

EWM

Turb

ulen

t win

dU

10,h

ub =

U10

,50-

yr(c

hara

cter

istic

stan

dard

dev

iatio

n of

win

d sp

eed

σ U,c

= 0

.11

⋅ U10

hub)

ESS

HS

= H

S,50

-yr (

1)M

isal

igne

dM

ultip

le

dire

ctio

ns

50-y

ear c

urre

nt50

-yea

r wat

er

leve

lU

LS

6.1b

EWM

Stea

dy w

ind

Uhu

b =

1.4

⋅ U10

,50-

yr

RW

HH

= ψ

⋅H50

-yr (

2)M

isal

igne

dM

ultip

le

dire

ctio

ns

50-y

ear c

urre

nt50

-yea

r wat

er

leve

lU

LS

6.1c

RW

MSt

eady

win

dU

hub

= 1.

1 ⋅U

10,5

0-yr

EWH

H =

H50

-yr

Mis

alig

ned

Mul

tiple

di

rect

ions

50-y

ear c

urre

nt50

-yea

r wat

er

leve

lU

LS

6.2a

EWM

Turb

ulen

t win

dU

10,h

ub =

U10

,50-

yr(c

hara

cter

istic

stan

dard

dev

iatio

n of

win

d sp

eed

σ U,c

= 0

.11

⋅ U10

hub)

ESS

HS

= H

S,50

-yr (

1)M

isal

igne

dM

ultip

le

dire

ctio

ns

50-y

ear c

urre

nt50

-yea

r wat

er

leve

lLo

ss o

f ele

ctric

al

netw

ork

conn

ectio

n

ULS

A

bnor

mal

6.2b

EWM

Stea

dy w

ind

Uhu

b =

1.4

⋅ U10

,50-

yr

RW

HH

= ψ

⋅H50

-yr (

2)M

isal

igne

dM

ultip

le

dire

ctio

ns

50-y

ear c

urre

nt50

-yea

r wat

er

leve

lLo

ss o

f ele

ctric

al

netw

ork

conn

ectio

n

ULS

A

bnor

mal

6.3a

EWM

Turb

ulen

t win

dU

10,h

ub =

U10

,1-y

r(c

hara

cter

istic

stan

dard

dev

iatio

n of

win

d sp

eed

σ U,c

= 0

.11

⋅ U10

hub)

ESS

HS

= H

S,1-

yr (1

)M

isal

igne

dM

ultip

le

dire

ctio

ns

1-ye

ar c

urre

nt1-

year

wat

er le

vel

Extre

me

yaw

mis

-al

ignm

ent

ULS

6.3b

EWM

Stea

dy w

ind

Uhu

b =

1.4

⋅ U10

,1-y

r

RW

HH

= ψ

⋅H1-

yr (2

)M

isal

igne

dM

ultip

le

dire

ctio

ns

1-ye

ar c

urre

nt1-

year

wat

er le

vel

Extre

me

yaw

mis

-al

ignm

ent

ULS

6.4

NTM

U10

,hub

< 0

.7U

10,5

0-yr

NSS

HS

acco

rdin

g to

join

t pro

babi

lity

dist

ribut

ion

of H

S, T

P an

d U

10,h

ub

Cod

irect

iona

l in

mul

tiple

dire

ctio

n (S

ee F

900)

(5)

Ran

ge b

etw

een

uppe

r and

low

er

1-ye

ar w

ater

leve

l

FLS

Park

ed a

nd

faul

t co

nditi

ons

7.1a

EWM

Turb

ulen

t win

dU

10,h

ub =

U10

,1-y

r(c

hara

cter

istic

stan

dard

dev

iatio

n of

win

d sp

eed

σ U,c

= 0

.11

⋅ U10

hub)

ESS

HS

= H

S,1-

yr (1

)M

isal

igne

dM

ultip

le

dire

ctio

ns

1-ye

ar c

urre

nt1-

year

wat

er le

vel

ULS

A

bnor

mal

7.1b

EWM

Stea

dy w

ind

Uhu

b =

1.4

⋅ U10

,1-y

r

RW

HH

= ψ

⋅H1-

yr (2

)M

isal

igne

dM

ultip

le

dire

ctio

ns

1-ye

ar c

urre

nt1-

year

wat

er le

vel

ULS

A

bnor

mal

7.1c

RW

MSt

eady

win

dU

hub

= 0.

88 ⋅

U10

,50-

yr

EWH

H =

H1-

yr

Mis

alig

ned

Mul

tiple

di

rect

ions

1-ye

ar c

urre

nt1-

year

wat

er le

vel

ULS

A

bnor

mal

7.2

NTM

U10

,hub

< 0

.7U

10,5

0-yr

NSS

HS

acco

rdin

g to

join

t pro

babi

lity

dist

ribut

ion

of H

S, T

P an

d U

10,h

ub

Cod

irect

iona

l in

mul

tiple

dire

ctio

n (S

ee F

900)

(5)

Ran

ge b

etw

een

uppe

r and

low

er

1-ye

ar w

ater

leve

l

FLS

Tabl

e E

1 Pr

opos

ed lo

ad c

ases

com

bini

ng v

ario

us e

nvir

onm

enta

l con

ditio

ns (C

ontin

ued)

D

esig

n si

tuat

ion

Load

ca

seW

ind

cond

ition

: Win

d cl

imat

e (U

10,h

ub) o

r w

ind

spee

d (U

hub)

Wav

e co

nditi

on: S

ea st

ate

(HS)

or

indi

vidu

al w

ave h

eigh

t (H

) to

com

-bi

ne w

ith in

sim

ulat

ions

for s

imul

-ta

neou

s win

d an

d w

aves

(7)

Win

d an

d w

ave

dire

ctio

nalit

yC

urre

ntW

ater

leve

lO

ther

co

nditi

ons

Lim

it st

ate

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Sec.4 –

Tran

spor

t, as

sem

bly,

m

aint

enan

ce

and

repa

ir

8.2a

EWM

Stea

dy w

ind

Uhu

b =

1.4

⋅ U10

,1-y

r

RW

HH

= ψ

⋅ H

1-yr

(2)

Cod

irect

iona

l in

one

dire

ctio

n1-

year

cur

rent

1-ye

ar w

ater

leve

lU

LS

Abn

orm

al

8.2b

RW

MSt

eady

win

dU

hub

= 0.

88 ⋅

U10

,50-

yr

EWH

H =

H1-

yr

Cod

irect

iona

l in

one

dire

ctio

n1-

year

cur

rent

1-ye

ar w

ater

leve

lU

LS

Abn

orm

al

8.3

NTM

U10

,hub

< 0

.7U

10,5

0-yr

NSS

HS

acco

rdin

g to

join

t pro

babi

lity

dist

ribut

ion

of H

S, T

P an

d U

10,h

ub

Cod

irect

iona

l in

mul

tiple

dire

ctio

n (S

ee F

900)

(5)

Ran

ge b

etw

een

uppe

r and

low

er

1-ye

ar w

ater

leve

l

FLS

1)In

cas

es w

here

load

and

resp

onse

sim

ulat

ions

are

to b

e pe

rfor

med

and

the

sim

ulat

ion

perio

d is

shor

ter t

han

the

refe

renc

e pe

riod

for t

he si

gnifi

cant

wav

e he

ight

HS,

the

sign

ifica

nt w

ave

heig

ht n

eeds

to b

e co

nver

ted

to a

re

fere

nce

perio

d eq

ual t

o th

e si

mul

atio

n pe

riod,

see

3C20

2. M

oreo

ver,

an in

flatio

n fa

ctor

on

the

sign

ifica

nt w

ave

heig

ht n

eeds

to b

e ap

plie

d in

ord

er to

mak

e su

re th

at th

e sh

orte

r sim

ulat

ion

perio

d ca

ptur

es th

e m

axim

um

wav

e he

ight

whe

n th

e or

igin

al re

fere

nce

perio

d do

es. W

hen

the

refe

renc

e pe

riod

is 3

hou

rs a

nd th

e si

mul

atio

n pe

riod

is 1

hou

r, th

e co

mbi

ned

conv

ersi

on a

nd in

flatio

n fa

ctor

is 1

.09

prov

ided

the

wav

e he

ight

s are

Ray

leig

h-di

strib

uted

and

the

num

ber o

f wav

es in

3 h

ours

is 1

000.

Lik

ewis

e, if

the

sim

ulat

ion

perio

d is

long

er th

an th

e av

erag

ing

perio

d fo

r the

mea

n w

ind

spee

d, a

def

latio

n fa

ctor

on

U10

may

be

appl

ied.

Whe

n th

e si

mul

atio

n pe

riod

is 1

hou

r and

the

aver

agin

g pe

riod

is 1

0 m

inut

es, t

he d

efla

tion

fact

or m

ay b

e ta

ken

as 0

.95.

2)It

is p

ract

ice

for o

ffsh

ore

stru

ctur

es to

app

ly ψ

= H

5-yr

/H50

-yr,

whe

re H

5-yr

and

H50

-yr d

enot

e th

e in

divi

dual

wav

e he

ight

s with

5- a

nd 5

0-ye

ar re

turn

per

iod,

resp

ectiv

ely.

The

shal

low

er th

e w

ater

dep

th, t

he la

rger

is u

sual

ly

the

valu

e of

ψ.

3)Th

e lo

ad c

ase

is n

ot d

riven

by

wav

es a

nd it

is o

ptio

nal w

heth

er th

e w

ind

load

shal

l be

com

bine

d w

ith a

n in

divi

dual

wav

e he

ight

or w

ith a

sea

stat

e.

4)Th

e w

ater

leve

l sha

ll be

take

n as

the

uppe

r-ta

il 50

-yea

r wat

er le

vel i

n ca

ses w

here

the

extre

me

wav

e he

ight

will

bec

ome

limite

d by

the

wat

er d

epth

.5)

In p

rinci

ple,

cur

rent

act

ing

conc

urre

ntly

with

the

desi

gn si

tuat

ion

in q

uest

ion

need

s to

be in

clud

ed, b

ecau

se th

e cu

rren

t inf

luen

ces t

he h

ydro

dyna

mic

coe

ffic

ient

s and

ther

eby

the

fatig

ue lo

adin

g re

lativ

e to

the

case

with

out

curr

ent.

How

ever

, in

man

y ca

ses c

urre

nt w

ill b

e of

littl

e im

porta

nce

and

can

be ig

nore

d, e

.g. w

hen

the

wav

e lo

adin

g is

iner

tia-d

omin

ated

or w

hen

the

curr

ent s

peed

is sm

all.

6)In

the

case

that

the

extre

me

oper

atio

nal g

ust i

s com

bine

d w

ith a

n in

divi

dual

wav

e he

ight

rath

er th

an w

ith a

sea

stat

e, th

e re

sulti

ng lo

ad sh

all b

e ca

lcul

ated

for t

he m

ost u

nfav

oura

ble

loca

tion

of th

e pr

ofile

of t

he in

divi

dual

w

ave

rela

tive

to th

e te

mpo

ral p

rofil

e of

the

gust

.7)

Whe

neve

r the

wav

e lo

adin

g as

soci

ated

with

a sp

ecifi

c lo

ad c

ase

refe

rs to

a w

ave

train

or a

tim

e se

ries o

f wav

e lo

ads,

the

soug

ht-a

fter c

ombi

ned

load

eff

ect s

hall

be in

terp

rete

d as

the

max

imum

resu

lting

load

eff

ect f

rom

the

time

serie

s of l

oad

effe

cts w

hich

is p

rodu

ced

by th

e si

mul

atio

ns.

Tabl

e E

1 Pr

opos

ed lo

ad c

ases

com

bini

ng v

ario

us e

nvir

onm

enta

l con

ditio

ns (C

ontin

ued)

D

esig

n si

tuat

ion

Load

ca

seW

ind

cond

ition

: Win

d cl

imat

e (U

10,h

ub) o

r w

ind

spee

d (U

hub)

Wav

e co

nditi

on: S

ea st

ate

(HS)

or

indi

vidu

al w

ave h

eigh

t (H

) to

com

-bi

ne w

ith in

sim

ulat

ions

for s

imul

-ta

neou

s win

d an

d w

aves

(7)

Win

d an

d w

ave

dire

ctio

nalit

yC

urre

ntW

ater

leve

lO

ther

co

nditi

ons

Lim

it st

ate

DET NORSKE VERITAS


205 Analysis of the load cases in Table E1 shall be carriedout for assumptions of aligned wind and waves or misalignedwind and waves, or both, as relevant. Analysis of the load casesin Table E1 shall be carried out for assumptions of wind in onesingle direction or wind in multiple directions, as relevant. 206 9 of the 31 load cases specified in Table E1 defineabnormal load cases to be considered for loads and load effectsdue to wind loading on the rotor and the tower in the ULS.Abnormal load cases are wind load cases associated with anumber of severe fault situations for the wind turbine, whichresult in activation of system protection functions. Abnormalload cases are in general less likely to occur than the normalload cases considered for the ULS in Table E1.207 Computer codes which are used for prediction of windturbine loads shall be validated for the purpose. The validationshall be documented.208 Table E1 refers to two turbulence models, viz. the nor-mal turbulence model NTM and the extreme turbulence modelETM. By the NTM the characteristic value σU,C of the stand-ard deviation σU of the wind speed shall be taken as the 90%quantile in the probability distribution of σU conditional onU10hub. By the ETM the characteristic value σU,C of the stand-ard deviation σU of the wind speed shall be taken as the valueof σU which together with U10hub forms a combined (U10hub,σU) event with a return period of 50 years.

Guidance note:When available turbulence data are insufficient to establish thecharacteristic standard deviation σU of the wind speed, the fol-lowing expressions may be applied for this standard deviation forthe normal and extreme turbulence models, respectively:

in which Iref is a reference turbulence intensity defined as theexpected turbulence intensity at a 10-minute mean wind speed of15 m/s, Vave is the annual average wind speed at hub height,b = 5.6 m/s and c = 2 m/s.The expressions are based on probability distribution assump-tions which do not account for wake effects in wind farms. Theexpressions are therefore not valid for design of wind turbinestructures for locations whose extreme turbulences are governedby wake effects.


209 The wind turbine loads in items 201 through 208 do notapply to meteorological masts nor to other structures which donot support wind turbines. For such structures, wind loads,which have not been filtered through a wind turbine to formwind turbine loads, shall be considered. Wind loads on mete-orological masts may be calculated according to EN 1991-1-4.Load combinations where these wind loads are combined withother types of environmental loads can be taken according toDNV-OS-C101.

Guidance note:Detailed methods for calculation of wind loads on meteorologi-cal masts are given in DIN 4131 and DIN 4133.


E 300 Determination of characteristic hydrodynamic loads301 Hydrodynamic loads shall be determined by analysis.When theoretical predictions are subjected to significantuncertainties, theoretical calculations shall be supported bymodel tests or full scale measurements of existing structures orby a combination of such tests and full scale measurements.302 Hydrodynamic model tests should be carried out to:

— confirm that no important hydrodynamic feature has beenoverlooked by varying the wave parameters (for new typesof installations, environmental conditions, adjacent struc-ture, etc.)

— support theoretical calculations when available analyticalmethods are susceptible to large uncertainties

— verify theoretical methods on a general basis.

303 Models shall be sufficient to represent the actual instal-lation. The test set-up and registration system shall provide abasis for reliable, repeatable interpretation. 304 Full-scale measurements may be used to update theresponse prediction of the relevant structure and to validate theresponse analysis for future analysis. Such tests may especiallybe applied to reduce uncertainties associated with loads andload effects which are difficult to simulate in model scale.305 In full-scale measurements it is important to ensure suf-ficient instrumentation and logging of environmental condi-tions and responses to ensure reliable interpretation.

E 400 Wave loads401 For calculation of wave loads, a recognised wave theoryfor representation of the wave kinematics shall be applied. Thewave theory shall be selected with due consideration of thewater depth and of the range of validity of the theory.402 Methods for wave load prediction shall be applied thatproperly account for the size, shape and type of structure.403 For slender structures, such as jacket structure compo-nents and monopile structures, Morison’s equation can beapplied to calculate the wave loads.404 For large volume structures, for which the wave kine-matics are disturbed by the presence of the structure, wave dif-fraction analysis shall be performed to determine local(pressure force) and global wave loads. For floating structureswave radiation forces must be included. 405 Both viscous effects and potential flow effects may beimportant in determining the wave-induced loads on a windturbine support structure. Wave diffraction and radiation areincluded in the potential flow effects.

Guidance note:Figure 1 can be used as a guidance to establish when viscouseffects or potential flow effects are important. Figure 1 refers tohorizontal wave-induced forces on a vertical cylinder, whichstands on the seabed and penetrates the free water surface, andwhich is subject to incoming regular waves.


Figure 1 Relative importance of inertia, drag and diffraction wave forces

406 Wave forces on slender structural members, such as a

)75.0( 10,, bUI hubrefNTMCU +⋅=σ

)10)4()3(072.0( 10,, +−⋅+⋅⋅⋅=

cU

cV

Ic hubaverefETMCUσ

DET NORSKE VERITAS


cylinder submerged in water, can be predicted by Morison’sequation. By this equation, the horizontal force on a verticalelement dz of the structure at level z is expressed as:

where the first term is an inertia force and the second term is adrag force. Here, CD and CM are drag and inertia coefficients,respectively, D is the diameter of the cylinder, ρ is the densityof water, is the horizontal wave-induced velocity of water,and is the horizontal wave-induced acceleration of water.The level z is measured from still water level, and the z axispoints upwards. Thus, at seabed z = −d, when the water depthis d.

Guidance note:The drag and inertia coefficients are in general functions of theReynolds number, the Keulegan-Carpenter number and the rela-tive roughness. The coefficient also depends on the cross-sec-tional shape of the structure and of the orientation of the body.For a cylindrical structural member of diameter D, the Reynoldsnumber is defined as Re = umaxD/ν and the Keulegan-Carpenternumber as KC = umaxTi/D, where umax is the maximum horizon-tal particle velocity at still water level, ν is the kinematic viscos-

ity of seawater, and Ti is the intrinsic period of the waves. Re andKC, and in turn CD and CM, may attain different values for theextreme waves that govern the ULS and for the moderate wavesthat govern the FLS.The drag coefficient CDS for steady-state flow can be used as abasis for calculation of CD and CM. The drag coefficient CDS forsteady-state flow depends on the roughness of the surface of thestructural member and may be taken as

in which k is the surface roughness and D is the diameter of thestructural member. New uncoated steel and painted steel can beassumed to be smooth. For concrete and highly rusted steel,k = 0.003 m can be assumed. For marine growth, k = 0.005 to0.05 m can be assumed. The drag coefficient CD depends on CDS and on the KC numberand can be calculated as

in which the wake amplification factor ψ can be read off fromFigure 2. For intermediate roughnesses between smooth andrough, linear interpolation is allowed between the curves forsmooth and rough cylinder surfaces in Figure 2.

Figure 2 Wake amplification factor as function of KC number for smooth (solid line) and rough (dotted line)

For KC < 3, potential theory is valid with CM = 2.0. For KC > 3,the inertia coefficient CM can be taken as

where CDS depends on the surface roughness of the structuralmember as specified above.As an example, in 30 to 40 meters of water in the southern andcentral parts of the North Sea, CD = 0.8 and CM = 1.6 can beapplied for diameters less than 2.2 m for use in load calculationsfor fatigue limit states.For structures in shallow waters and near coastlines where thereis a significant current in addition to the waves, CM should not betaken less than 2.0. For long waves in shallow water, the depth variation of the waterparticle velocity is usually not large. Hence it is recommended touse force coefficients based on the maximum horizontal waterparticle velocity umax at the free surface.When waves are asymmetric, which may in particular be the casein shallow waters, the front of the wave has a different steepnessthan the rear of the wave. Since the wave force on a structuredepends on the steepness of the wave, caution must be exercisedto apply the asymmetric wave to the structure in such a manner

that the wave load impact is calculated from that of the two wavesteepnesses which will produce the largest force on the structure.


407 The resulting horizontal force F on the cylinder can befound by integration of Morison’s equation for values of z from–d to the wave crest, η(t).

Guidance note:For non-breaking waves, the resulting horizontal force becomes

The integration from –d to 0 ignores contributions to the forcefrom the wave crest above the still water level at z = 0. This is aminor problem when the inertia force FM is the dominating forcecomponent in F, since FM has its maximum when a nodal line at

dzxxDCdzxDC

dFdFdF

DM

DM

&&&&24

2ρρπ +=

+=

x·x··

⎪⎪⎩

⎪⎪⎨

⎧

>

<<+

<

=−

−

−

(rough) 10/for 05.1

10/10for 20

)/(log429(smooth) 10/for 65.0

2

24-10

4

Dk

DkDk

Dk

CDS

),( KCCCC DSDSD ψ⋅=

0.00.20.40.60.81.01.21.41.61.82.0

0 5 10 15 20 25 30 35 40 45 50 55 60KC/CDS

{ })65.0(6.1);3(044.00.2max −−−−= DSM CKCC

dzxxDC

dzxDC

FFF

t

dD

t

dM

DM

∫

∫

−

−

+

=

+=

)(

)( 2

2

4

η

η

ρ

ρπ

&&

&&

DET NORSKE VERITAS


the still water level passes the structure. The drag force FD has itsmaximum when the crest or trough passes the structure. If thisforce is the dominating force component in F, a significant errorcan be introduced by ignoring the contribution from the wavecrest.The relative magnitude between the inertia force component FMand the drag force component FD can be expressed by the ratiobetween their amplitudes, A = AM/AD. Figure 2 can be used toquickly establish whether the inertia force or the drag force is thedominating force, once the ratios H/D and d/λ have been calcu-lated. Structures which come out above the curve marked A = 1in Figure 2 experience drag-dominated loads, whereas structureswhich come out below this curve experience inertia-dominatedloads.Morison’s equation is only valid when the dimension of thestructure is small relative to the wave length, i.e. when D < 0.2λ.The integrated version of Morison’s equation given here is onlyvalid for non-breaking waves. However, Morison’s equation as

formulated for a vertical element dz is valid for calculation ofwave forces from both breaking and non-breaking waves as longas the element is fully submerged. In deep water, waves breakwhen H/λ exceeds about 0.14. In shallow water, waves breakwhen H/d exceeds about 0.78.Figure 2 is based on linear wave theory and should be used withcaution, since linear wave theory may not always be an adequatewave theory as a basis for prediction of wave forces in particu-larly shallow waters. 5th order stream function theory is usuallyconsidered the best wave theory for representation of wave kine-matics in shallow waters. For prediction of wave forces forfatigue assessment, higher order stream function theory can beapplied for water depths less than approximately 15 m, whereasStokes 5th order theory is recommended for water depths inexcess of approximately 30 m.


Figure 3 Relative magnitude of inertia and drag forces for cylinders with D/λ < 0.2

408 When the dimension of the structure is large comparedwith the wave length, typically when D > 0.2λ, Morison’sequation is not valid. The inertia force will be dominating andcan be predicted by diffraction theory.

Guidance note:For linear waves, the maximum horizontal force on a vertical cyl-inder of radius R = D/2 installed in water of depth d and subjectedto a wave of amplitude AW, can be calculated as

and its arm measured from the seabed is

The coefficients ξ and α are given in Table E2.

The diffraction solution for a vertical cylinder given above isreferred to as the MacCamy-Fuchs solution. The terms given rep-resents essentially a corrected inertia term which can be used inMorison’s equation together with the drag term.The formulae given in this guidance note are limited to verticalcircular cylinders with constant diameter D. For other geometriesof the support structure, such as when a conical component ispresent in the wave-splash zone to absorb or reduce ice loads, dif-fraction theory is still valid, but the resulting force and momentarm will come out different from the vertical cylinder solutionsgiven here.


409 For evaluation of load effects from wave loads, possibleringing effects shall be included in the considerations. When asteep, high wave encounters a monopile, high frequency non-linear wave load components can coincide with natural fre-quencies of the structure causing resonant transient response inthe global bending modes of the pile. Such ringing effects are

[ ][ ] ξαρkdAdk

kgAF WW

X tanh)sin(sinh4

2max,+

=

[ ] [ ][ ]kdkd

kdkdkddhF sinh1coshsinh +−

=

DET NORSKE VERITAS


only of significance in combination with extreme first orderwave frequency effects. Ringing should be evaluated in thetime domain with due consideration of higher order wave loadeffects. The magnitude of the first ringing cycles is governedby the magnitude of the wave impact load and its duration isrelated to the structural resonance period.

Guidance note:Ringing can occur if the lowest natural frequencies of the struc-ture do not exceed three to four times the typical wave frequency.In case the natural frequency exceeds about five to six times fp,where fp denotes the peak frequency, ringing can be ruled out.When a dynamic analysis is carried out, any ringing responsewill automatically appear as part of the results from the analysis,provided the wave forces are properly modelled and included inthe analysis.


Table E2 Coefficients ξ and α

410 When waves are likely to break on the site of the struc-ture or in its vicinity, wave loads from breaking waves shall beconsidered in the design of the structure. Wave loads frombreaking waves depend on the type of breaking waves. A dis-tinction is made between surging, plunging and spilling waves.The kinematics is different for these three types of breakingwaves.

Guidance note:For plunging waves, an impact model may be used to calculatethe wave forces on a structure. The impact force from a plungingwave can be expressed as

F = ½ ρ CS Au2

where u denotes the water particle velocity in the plunging wavecrest, ρ is the mass density of the fluid, A is the area on the struc-ture which is assumed exposed to the slamming force, and CS isthe slamming coefficient. For a smooth circular cylinder, theslamming coefficient should not be taken less than 3.0. The upperlimit for the slamming coefficient is 2π. Careful selection ofslamming coefficients for structures can be made according toDNV-RP-C205. The area A exposed to the slamming forcedepends on how far the plunging breaker has come relative to thestructure, i.e., how wide or pointed it is when it hits the structure.Plunging waves are rare in Danish and German waters.

For a plunging wave that breaks immediately in front of a verticalcylinder of diameter D, the duration T of the impact force on thecylinder is given by

For surging and spilling waves, an approach to calculate the asso-ciated wave forces on a vertical cylindrical structure of diameterD can be outlined as follows: The cylinder is divided into anumber of sections. As the breaking wave approaches the struc-ture, the instantaneous wave elevation close to the cylinderdefines the time instant when a section is hit by the wave andstarts to penetrate the sloping water surface. The instantaneousforce per vertical length unit on this section and on underlyingsections, which have not yet fully penetrated the sloping watersurface, can be calculated as

f = ½ ρCSDu2

where u denotes the horizontal water particle velocity, ρ is themass density of the fluid, and CS is the slamming coefficientwhose value can be taken as

for 0 < s < D.The penetration distance s for a section in question is the horizon-tal distance from the periphery on the wet side of the cylinder tothe sloping water surface, measured in the direction of the wavepropagation. For fully submerged sections of the cylinder, thewave forces can be determined from classical Morison theorywith mass and drag terms using constant mass and drag coeffi-cients,

The water particle velocity u is to be determined from the wavekinematics for the particular type of breaking wave in question.


411 Computer codes which are used for prediction of waveloads on wind turbine structures shall be validated for the pur-pose. The validation shall be documented.412 Characteristic extreme wave loads are in this standarddefined as wave load values with a 50-year return period.

Guidance note:In the southern and central parts of the North Sea, experienceshows that the ratio between the 100- and 50-year wave load val-ues Fwave,100/Fwave,50 attains a value approximately equal to1.10. Unless data indicate otherwise, this value of the ratioFwave,100/Fwave,50 may be applied to achieve the 50-year waveload Fwave,50 in cases where only the 100-year value Fwave,100 isavailable, provided the location in question is located in thesouthern or central parts of the North Sea.


413 Any walkways or platforms mounted on the supportstructure of an offshore wind turbine shall be located above thesplash zone.For determination of the deck elevation of access platformswhich are not designed to resist wave forces, a sufficient airgapbased on design water level and design wave crest height shallbe ensured, such that extreme wave crests up to the height of

cDT

6413

=

⎟⎠⎞

⎜⎝⎛ +

+=

Ds

sDDCS

107.019

15.5

22

21

4DuC

dtduDCf DM ρρπ +=

DET NORSKE VERITAS


the design wave crest are allowed to pass without risk of touch-ing the platform.

Guidance note:Sufficient airgap is necessary in order to avoid slamming forceson an access platform. When the airgap is calculated, it is recom-mended to consider an extra allowance to account for possiblelocal wave effects due to local seabed topography and shorelineorientation. The extra allowance should be at least 1.0 m. Forlarge-volume structures, airgap calculation should include awave diffraction analysis.It is also important to consider run-up, i.e. water pressed upwardsalong the surface of the structure or the structural members thatsupport the access platform, either by including such run-up inthe calculation of the necessary airgap or by designing the plat-form for the loads from such run-up.


414 For prediction of wave loading, the effect of disturbedwater particle kinematics due to secondary structures shall beaccounted for. Disturbed kinematics due to large volume struc-tures should be calculated by a wave diffraction analysis. Forassessment of shielding effects due to multiple slender struc-tures reference is made to DNV-RP-C205.

E 500 Ice loads501 Loads from laterally moving ice shall be based on rele-vant full scale measurements, on model experiments which canbe reliably scaled, or on recognised theoretical methods. Whendetermining the magnitude and direction of ice loads, consid-eration is to be given to the nature of the ice, the mechanicalproperties of the ice, the ice-structure contact area, the size andshape of the structure, and the direction of the ice movements.The oscillating nature of the ice loads, including build-up andfracture of moving ice, is to be considered.

Guidance note:Theoretical methods for calculation of ice loads should always beused with caution.In sheltered waters and in waters close to the coastline, a rigid icecover will usually not move once it has grown to exceed somelimiting thickness, see 3F305. In such land-locked waters, loadscaused by moving ice may be calculated on the basis of this lim-iting thickness only, while loads associated with thermal pres-sures, arch effects and vertical lift need to be calculated on thebasis of the actual characteristic ice thickness as required by thisstandard.In open sea, where moving ice can be expected regardless ofthickness, all ice loads shall be based on the actual characteristicice thickness as required by this standard.


502 Where relevant, ice loads other than those caused by lat-

erally moving ice are to be taken into account. Such ice loadsinclude, but are not limited to, the following:

— loads due to rigid covers of ice, including loads due to archeffects and water level fluctuations

— loads due to masses of ice frozen to the structure— pressures from pack ice and ice walls— thermal ice pressures associated with temperature fluctua-

tions in a rigid ice cover— possible impact loads during thaw of the ice, e.g. from fall-

ing blocks of ice— loads due to icing and ice accretion.

Guidance note:Owing to the very large forces associated with pack ice, it is notrecommended to install wind turbines in areas where pack icemay build up.


503 Table E3 specifies a proposal for 7 load cases to considerfor ice load conditions and their companion wind load condi-tions in order to fulfil the requirements in 501 and 502. Theload cases in Table E3 refer to design in the ULS and in theFLS. The load cases for design in the ULS are based on a char-acteristic ice thickness tC equal to the 50-year ice thickness t50or equal to the limiting thickness tlimit, depending on location.504 Wherever there is a risk that falling blocks of ice may hita structural member, a system to protect these members fromthe falling ice shall be arranged.505 Possible increases in volume due to icing are to be con-sidered when wind and wave loads acting on such volumes areto be determined.506 The structure shall be designed for horizontal and verti-cal static ice loads. Frictional coefficients between ice and var-ious structural materials are given in 3F307. Ice loads onvertical structures may be determined according to API RP2N.

Guidance note:Horizontal loads from moving ice should be considered to act inthe same direction as the concurrent wind loads.Unilateral thermal ice pressures due to thermal expansion andshrinkage can be assumed to act from land outwards toward theopen sea or from the centre of a wind farm radially outwards.Larger values of unilateral thermal ice pressures will apply tostand-alone structures and to the peripheral structures of a windfarm than to structures in the interior of a wind farm.The water level to be used in conjunction with calculation of iceloads shall be taken as the high water level or the low water levelwith the required return period, whichever is the most unfavour-able.


DET NORSKE VERITAS


507 Ice loads on inclined structural parts such as ice-loadreducing cones in the splash zone may be determined accord-ing to Ralston’s formulae. Ralston’s formulae are given inAppendix L.

Guidance note:To achieve an optimal ice cone design and avoid that ice loadgoverns the design of the support structure and foundation, it isrecommended to adjust the inclination angle of the cone such thatthe design ice load is just less than the design wave load.For ice-load reducing cones of the “inverted cone” type that willtend to force moving ice downwards, the bottom of ice-loadreducing cones is recommended to be located a distance of atleast one ice thickness below the water level.The flexural strength of ice governs the ice loads on inclinedstructures. Table E4 specifies values of the flexural strength forvarious return periods in different waters.


508 The characteristic local ice pressure for use in designagainst moving ice shall be taken as

where ru,C is the characteristic compressive strength of the ice,tC is the characteristic thickness of the ice, and Alocal is the area

over which the locale ice pressure is applied.Guidance note:The characteristic compressive strength of ice depends on localconditions such as the salinity. For load cases, which representrare events, the characteristic compressive strength is expressedin terms of a required return period. Table E5 specifies values ofthe compressive strength for various return periods in differentwaters.For load cases, which are based on special events during thaw,break-up and melting, lower values than those associated withrare events during extreme colds apply. 1.5 MPa applies to rigidice during spring at temperatures near the melting point. 1.0 MPaapplies to partly weakened, melting ice at temperatures near themelting point. Local values for the characteristic ice thickness tc shall beapplied.


509 The structure shall be designed for horizontal and verti-cal dynamic ice loads.

Guidance note:For structures located in areas, where current is prevailing, thedynamic ice load may govern the design when it is combinedwith the concurrent wind load. This may apply to the situationwhen the ice breaks in the spring.


Table E3 Proposed load cases combining ice loading and wind loading Design situation

Load case

Ice condition Wind condition: Wind climate (U10hub)

Water level Other conditions

Limit state

Power production

E1 Horizontal load due to tempera-ture fluctuations

vin < U10hub < vout+ NTM10-minute mean wind speed resulting in maximum thrust

1-year water level ULS

E2 Horizontal load due to water level fluctuations or arch effects

vin < U10hub < vout+ NTM10-minute mean wind speed resulting in maximum thrust


E3 Horizontal load from moving ice floeIce thickness:tC = t50 in open seatC = tlimit in land-locked waters

vin < U10hub < vout+ ETM 10-minute mean wind speed resulting in maximum thrust

50-year water level For prediction of extreme loads

ULS


vin < U10hub < vout 1-year water level FLS

E5 Vertical force from fast ice covers due to water level fluctuations

No wind load applied 1-year water level ULS

Parked (standing still or idling)

E6 Pressure from hummocked ice and ice ridges

Turbulent windU10hub = U10,50-yr+ characteristic standard deviation of wind speed σU,c = 0.11 ⋅ U10hub



U10hub < 0.7U10,50-yr+ NTM

1-year water level FLS

Table E4 Flexural strength of sea iceReturn period

(years)Flexural strength of ice, rf (MPa)

Southern North Sea, Skagerrak, Kattegat

Southwestern Baltic Sea

5 – 0.2510 – 0.3950 0.50 0.50

100 – 0.53

local

CCuClocal A

trr2

,, 51+=

Table E5 Compressive strength of sea ice

Return period (years)

Compressive strength of ice, ru (MPa)Southern North Sea, Skagerrak, Kattegat

Southwestern Baltic Sea

5 1.0 1.010 1.5 1.550 1.6 1.9100 1.7 2.1

DET NORSKE VERITAS


510 The level of application of horizontal ice load dependson the water level and the possible presence of ice-load reduc-ing cones. Usually a range of application levels needs to beconsidered.511 When ice breaks up, static and dynamic interactions willtake place between the structure and the ice. For structureswith vertical walls, the natural vibrations of the structure willaffect the break-up frequency of the ice, such that it becomestuned to the natural frequency of the structure. This phenome-non is known as lock-in and implies that the structure becomesexcited to vibrations in its natural mode shapes. The structureshall be designed to withstand the loads and load effects fromdynamic ice loading associated with lock-in when tuningoccurs. All contributions to damping in the structure shall beconsidered. Additional damping owing to pile-up of ice floesmay be accounted for when it can be documented.

Guidance note:The criterion for occurrence of tuning is

where Uice is the velocity of the ice floe, t is the thickness of theice, and fn is the natural frequency of the structure.The loading can be assumed to follow a serrated profile in thetime domain as shown in Figure 4. The maximum value of theload shall be set equal to the static horizontal ice load. Aftercrushing of the ice, the loading is reduced to 20% of the maxi-mum load. The load is applied with a frequency that correspondsto the natural frequency of the structure. All such frequencies thatfulfil the tuning criterion shall be considered.


Figure 4 Serrated load profile for dynamic ice load

512 For conical structures, the break-up frequency of the iceshall be assumed independent of the natural vibrations of thestructure. It shall be assured in the design that the frequency ofthe ice load is not close to the natural frequency of the struc-ture.

Guidance note:The frequency of the ice load can be determined as

where Uice is the velocity of the ice floe, and L is the crack lengthin the ice.The force can be applied according to the simplified model inFigure 4, even though the failure mechanism in the ice is differ-ent for conical structures than for vertical structures.

For prediction of the crack length L, the following two modelsare available:

1) L = ½ ρ D, where D is the diameter of the cone at the watertable and ρ is determined from Figure 5 as a function ofγ WD2/(σ ft), in which σf is the flexural strength of the ice,γ W is the unit weight of water and t is the thickness of theice.

2)

where E is Young’s modulus of the ice and ν is Poisson’s ra-tio of the ice.

Neither of these formulae for prediction of L reflects the depend-ency of L on the velocity of the ice floe, and the formulae musttherefore be used with caution. The prediction of L is in generalrather uncertain, and relative wide ranges for the frequency ficemust therefore be assumed in design to ensure that an adequatestructural safety is achieved.


Figure 5 Factor ρ for calculation of crack length in ice

E 600 Water level loads601 Tidal effects and storm surge effects shall be consideredin evaluation of responses of interest. Higher water levels tendto increase hydrostatic loads and current loads on the structure;however, situations may exist where lower water levels willimply the larger hydrodynamic loads. Higher mean water lev-els also imply a decrease in the available airgap to access plat-forms and other structural components which depend on someminimum clearance.

Guidance note:In general, both high water levels and low water levels shall beconsidered, whichever is most unfavourable, when water levelloads are predicted.For prediction of extreme responses, there are thus two 50-yearwater levels to consider, viz. a low 50-year water level and a high50-year water level. Situations may exist where a water levelbetween these two 50-year levels will produce the most unfa-vourable responses.


E 700 Earthquake loads701 When a wind turbine structure is to be designed forinstallation on a site which may be subject to an earthquake,

3.0>⋅ n

iceft

U

LUf ice

ice =

25.02

3

))1(12

21

(νγ −

=W

EtL

DET NORSKE VERITAS


the structure shall be designed to withstand the earthquakeloads. Response spectra in terms of so-called pseudo responsespectra may be used for this purpose.

Guidance note:Pseudo response spectra for a structure are defined for displace-ment, velocity and acceleration. For a given damping ratio γ andangular frequency ω, the pseudo response spectrum S gives themaximum value of the response in question over the duration ofthe response. This can be calculated from the ground accelerationhistory by Duhamel’s integral. The following pseudo responsespectra are considered:- SD, response spectral displacement- SV, response spectral velocity- SA, response spectral accelerationFor a lightly damped structure, the following approximate rela-tionships apply, SA≈ ω2SD and SV ≈ ω SD, such that it suffices toestablish the acceleration spectrum and then use this to establishthe other two spectra.It is important to analyse the wind turbine structure for the earth-quake-induced accelerations in one vertical and two horizontaldirections. It usually suffices to reduce the analysis in two hori-zontal directions to an analysis in one horizontal direction, due tothe symmetry of the dynamic system. The vertical accelerationmay lead to buckling in the tower. Since there is not expected tobe much dynamics involved with the vertical motion, the towermay be analysed with respect to buckling for the load induced bythe maximum vertical acceleration caused by the earthquake.However, normally the only apparent buckling is that associatedwith the ground motion in the two horizontal directions, and thebuckling analysis for the vertical motion may then not be rele-vant. For detailed buckling analysis for the tower, reference ismade to DNV-OS-C101 and NORSOK.For analysis of the horizontal motions and accelerations, thewind turbine can be represented by a concentrated mass on top ofa vertical rod, and the response spectra can be used directly todetermine the horizontal loads set up by the ground motions. Fora typical wind turbine, the concentrated mass can be taken as themass of the nacelle, including the rotor mass, plus ¼ of the towermass.


702 When a wind turbine structure is to be installed in areaswhich may be subject to tsunamis set up by earthquakes, theload effect of the tsunamis on the structure shall be considered.

Guidance note:Tsunamis are seismic sea waves. To account for load effects oftsunamis on wind turbine structures in shallow waters, an accept-able approach is to calculate the loads for the maximum sea wavethat can exist on the site for the given water depth.


E 800 Marine growth801 Marine growth shall be taken into account by increasingthe outer diameter of the structural member in question in thecalculations of hydrodynamic wave and current loads. Thethickness of the marine growth depends on the depth below sealevel and the orientation of the structural component. Thethickness shall be assessed based on relevant local experienceand existing measurements. Site-specific studies may be nec-essary in order to establish the likely thickness and depthdependence of the growth.

Guidance note:Unless data indicate otherwise, the following marine growth pro-file may be used for design in Norwegian and UK waters:

Somewhat higher values, up to 150 mm between sea level andLAT –10 m, may be seen in the Southern North Sea.Offshore central and southern California, marine growth thick-nesses of 200 mm are common.In the Gulf of Mexico, the marine growth thickness may be takenas 38 mm between LAT+3 m and 50 m depth, unless site-specificdata and studies indicate otherwise.Offshore West Africa, the marine growth thickness may be takenas 100 mm between LAT and 50 m depth and as 300 mm in thesplash zone above LAT, unless data indicate otherwise.The outer diameter of a structural member subject to marinegrowth shall be increased by twice the recommended thickness atthe location in question.The type of marine growth may have an impact on the values ofthe hydrodynamic coefficients that are used in the calculations ofhydrodynamic loads from waves and current.


802 Due to the uncertainties involved in the assumptionsregarding the marine growth on a structure, a strategy forinspection and possible removal of the marine growth shouldbe planned as part of the design of the structure. When such astrategy is planned, the inspection frequency, the inspectionmethod and the criteria for growth removal shall be based onthe impact of the marine growth on the safety of the structureand on the extent of experience with marine growth under thespecific conditions prevailing at the site.

E 900 Scour901 Scour is the result of erosion of soil particles at and neara foundation and is caused by waves and current. Scour is aload effect and may have an impact on the geotechnical capac-ity of a foundation and thereby on the structural response thatgoverns the ultimate and fatigue load effects in structural com-ponents.902 Means to prevent scour and requirements to such meansare given in Sec.10 B300.

E 1000 Transportation loads and installation loads1001 Criteria shall be defined for acceptable external condi-tions during transportation, installation and dismantling of off-shore wind turbine structures and their foundations. Thisincludes external conditions during installation, dismantlingand replacement of wind turbine rotors and nacelles as far asthe involved loads on the support structures and foundationsare concerned. Based on the applied working procedures, onthe vessels used and on the duration of the operation in ques-tion, acceptable limits for the following environmental proper-ties shall be specified:

— wind speed— wave height and wave crest— water level— current— ice.

1002 It shall be documented that lifting fittings mounted ona structure subject to lifting is shaped and handled in such amanner that the structure will not be damaged during liftingunder the specified external conditions.1003 DNV Rules for Planning and Execution of MarineOperations apply.

F. Combination of Environmental LoadsF 100 General101 This section gives requirements for combination of envi-ronmental loads in the operational condition.102 The requirements refer to characteristic wind turbine

Depth below MWL (m)

Marine growth thickness (mm)Central and

Northern North Sea(56° to 59° N)

Norwegian Sea(59° to 72° N)

–2 to 40 100 60> 40 50 30

DET NORSKE VERITAS


loads based on an investigation of the load cases specified inTables E1 and E3.103 For design against the ULS, the characteristic environ-mental load effect shall be taken as the 98% quantile in the dis-tribution of the annual maximum environmental load effect,i.e. it is the load effect whose return period is 50 years, andwhose associated probability of exceedance is 0.02. When theload effect is the result of the simultaneous occurrence of twoor more environmental load processes, these requirements tothe characteristic load effect apply to the combined load effectfrom the concurrently acting load processes. The subsequentitems specify how concurrently acting environmental loadscan be combined to arrive at the required characteristic com-bined load effect. 104 Environmental loads are loads exerted by the environ-ments that surround the structure. Typical environments arewind, waves, current, and ice, but other environments may alsobe thought of such as temperature and ship traffic. Each envi-ronment is usually characterized by an intensity parameter.Wind is usually characterized by the 10-minute mean windspeed, waves by the significant wave height, current by themean current, and ice by the ice thickness.

F 200 Environmental states201 Environmental states are defined as short-term environ-mental conditions of approximately constant intensity param-eters. The typical duration of an environmental state is 10minutes or one hour. The long-term variability of multipleintensity parameters representative of multiple, concurrentlyactive load environments can be represented by a scattergramor by a joint probability distribution function including infor-mation about load direction.

F 300 Environmental contours301 An environmental contour is a contour drawn through aset of environmental states on a scattergram or in a joint prob-ability density plot for the intensity parameters of concurrentlyactive environmental processes. The environmental statesdefined by the contour are states whose common quality is thatthe probability of a more rare environmental state is p = TS/TRwhere TS is the duration of the environmental state and TR is aspecified return period.

Guidance note:The idea of the environmental contour is that the environmentalstate whose return period is TR is located somewhere along theenvironmental contour defined based on TR. When only oneenvironmental process is active, the environmental contourreduces to a point on a one-dimensional probability density plotfor the intensity parameter of the process in question, and thevalue of the intensity in this point becomes equal to the valuewhose return period is TR. For an offshore wind turbine, the wind process and the waveprocess are two typical concurrent environmental processes. The10-minute mean wind speed U10 represents the intensity of thewind process, and the significant wave height HS represents theintensity of the wave process. The joint probability distributionof U10 and HS can be represented in terms of the cumulative dis-tribution function FU10 for U10 and the cumulative distributionfunction FHS|U10 for HS conditional on U10. A first-order approx-imation to the environmental contour for return period TR can beobtained as the infinite number of solutions (U10, HS) to the fol-lowing equation

valid for TS < TR.in which Φ –1 denotes the inverse of the standard normal cumu-lative distribution function.The environmental contour whose associated return period is 50years is useful for finding the 50-year load effect in the wind tur-bine structure when the assumption can be made that the 50-year

load effect occurs during the 50-year environmental state. Whenthis assumption can be made, the 50-year load effect can be esti-mated by the expected value of the maximum load effect that canbe found among the environmental states of duration TS alongthe 50-year environmental contour.The environmental state is characterised by a specific duration,e.g. one hour. Whenever data for U10 and HS refer to referenceperiods which are different from this duration, appropriate con-versions of these data to the specified environmental state dura-tion must be carried out.


F 400 Combined load and load effect due to wind load and wave load401 In a short-term period with a combination of waves andfluctuating wind, the individual variations of the two load proc-esses about their respective means can be assumed uncorrelated.This assumption can be made regardless of the intensities anddirections of the load processes, and regardless of possible cor-relations between intensities and between directions.402 Two methods for combination of wind load and waveload are given in this standard:

— Linear combination of wind load and wave load, or ofwind load effects and wave load effects, see F500.

— Combination of wind load and wave load by simulation,see F600.

403 The load combination methods presented in F500 andF600 and the load combinations specified in F700 areexpressed in terms of combinations of wind load effects, waveload effects and possible other load effects. This correspondsto design according to Approach (1) in Sec.2 E202. For designaccording to Approach (2) in Sec.2 E202, the term “loadeffect” in F500, F600 and F700 shall be interpreted as “load”such that “design loads” are produced by the prescribed com-bination procedures rather than “design load effects”. Follow-ing Approach (2), the design load effects then result fromstructural analyses for these design loads.

F 500 Linear combinations of wind load and wave load501 The combined load effect in the structure due to concur-rent wind and wave loads may be calculated by combining theseparately calculated wind load effect and the separately calcu-lated wave load effect by linear superposition. This methodmay be applied to concept evaluations and in some cases alsoto load calculations for final design, for example in shallowwater or when it can be demonstrated that there is no particulardynamic effect from waves, wind, ice or combinations thereof.According to the linear combination format presented in Sec.2,the design combined load effect is expressed as

in which Swind,k denotes the characteristic wind load effect andSwave,k denotes the characteristic wave load effect. It is a pre-requisite for using this approach to determine the design com-bined load effect that the separately calculated value of thecharacteristic wave load effect Swind,k is obtained for realisticassumptions about the equivalent damping that results fromthe structural damping and the aerodynamic damping. Theequivalent damping depends on the following conditionsrelated to the wind turbine and the wind load on the turbine:

— whether the wind turbine is exposed to wind or not— whether the wind turbine is in operation or is parked— whether the wind turbine is a pitch-regulated turbine or a

stall-regulated turbine— the direction of the wind loading relative to the direction

of the wave loading.

Correct assumptions for the wind turbine and the wind load

)1()))((()))((( 12|

1210

11010

R

SSUHU T

THFUF S −Φ=Φ+Φ −−−

kwavefkwindfd SSS ,2,1 γγ +=

DET NORSKE VERITAS


shall be made according to this list. The equivalent dampingshall be determined in correspondence with these assumptions.Structural analyses by an adequate structural analysis modeland based on this equivalent damping shall then be used todetermine the characteristic wave load effect Swave,k. Thedamping from the wind turbine should preferably be calculateddirectly in an integrated model.

Guidance note:When the characteristic load effect Swave,k is defined as the loadeffect whose return period is TR, the determination of Swave,k asa quantile in the distribution of the annual maximum load effectmay prove cumbersome and involve a large number of structuralanalyses to be carried out before contributions to this distributionfrom all important sea states have been included. When the assumption can be made that Swave,k occurs during theparticular sea state of duration TS whose significant wave heightHS has a return period equal to TR, then Swave,k may be estimatedby the expected value of the maximum load effect during this seastate, and the analytical efforts needed may become considerablyreduced. The assumption that Swave,k occurs in the sea statewhose return period is TR is often reasonable, unless sea statesexist for which the structure becomes more dynamically excitedthan by this particular sea state, for example sea states involvingwave trains whose periods are close to integer multiples of thenatural period of the structure.When the structural analysis involves executions of a number ofsimulations of the maximum load effect that occurs during thesea state whose significant wave height has a return period TR,then Swave,k shall be estimated by the mean of these simulatedmaximum load effects.The wind loads in the wind direction during idling and with theyaw system in function will be quite small and will consistmainly of drag on the tower and the nacelle cover. During thiscondition it is implied that the blades are pitched such that theblade profiles point in the direction up against the wind or in thewind direction. The largest wind loads in this condition will bethe blade loads that act perpendicular to the wind direction.


F 600 Combination of wind load and wave load by simulation601 The combined load effect in the structure due to concur-rent wind and wave loads may alternatively be calculated bydirect simulation. This approach is based on structural analysesin the time domain for simultaneously applied simulated timeseries of the wind load and the wave load. By this approach,simulated time series of the combined load effect results, fromwhich the characteristic combined load effect Sk is interpreted.

Guidance note:The approach requires that a global structural analysis model isestablished, e.g. in the form of a beam-element based framemodel, to which load series from several simultaneously actingload processes can be applied. Although this is here exemplifiedfor two concurrently acting load processes, viz. wind loads andwave loads, this can be generalised to include also other concur-rent load processes.When the characteristic load effect Sk is defined as the load effectwhose return period is TR, the determination of Sk as a quantilein the distribution of the annual maximum load effect may provecumbersome and involve a large number of structural analyses tobe carried out before contributions to this distribution from allimportant environmental states have been included.

When the assumption can be made that Sk occurs during an envi-ronmental state of duration TS associated with a return period TR,then Sk may be estimated by the expected value of the maximumload effect during such an environmental state, and the analyticalefforts needed may become considerably reduced. Under thisassumption, Sk can be estimated by the expected value of themaximum load effect that can be found among the environmentalstates on the environmental contour whose associated returnperiod is TR. To simulate one realisation of the maximum load effect along theenvironmental contour whose associated return period is TR, onestructural simulation analysis is carried out for each environmen-tal state along the environmental contour and one maximum loadeffect results for each one of these states. The same seed needs tobe applied for each environmental state investigated this way. Afollowing search along the contour will identify the sought-afterrealisation of the maximum load effect. In practice, it will sufficeto carry out structural simulation analyses only for a limitednumber of environmental states along a part of the environmentalcontour. The procedure is repeated for a number of differentseeds, and a corresponding number of maximum load effect real-isations are obtained. The sought-after characteristic load effectSk is estimated by the mean of these simulated maximum loadeffects.When dynamic simulations utilising a structural dynamics modelare used to calculate load effects, the total period of load effectdata simulated shall be long enough to ensure statistical reliabil-ity of the estimate of the sought-after maximum load effect. Atleast six ten-minute stochastic realisations (or a continuous 60-minute period) shall be required for each 10-minute mean, hub-height wind speed considered in the simulations. Since the initialconditions used for the dynamic simulations typically have aneffect on the load statistics during the beginning of the simulationperiod, the first 5 seconds of data (or longer if necessary) shall beeliminated from consideration in any analysis interval involvingturbulent wind input.The wind loads in the wind direction during idling and with theyaw system in function will be quite small and will consistmainly of drag on the tower and the nacelle cover. During thiscondition it is implied that the blades are pitched such that theblade profiles point in the direction up against the wind or in thewind direction. The largest wind loads in this condition will bethe blade loads that act perpendicular to the wind direction.The wave field must be simulated by applying a valid wave the-ory according to Sec.3. Simulation using linear wave theory(Airy theory) in shallow waters may significantly underestimatethe wave loads.


F 700 Basic load cases701 When information is not available to produce the requiredcharacteristic combined load effect directly, the required charac-teristic combined load effect can be obtained by combining theindividual characteristic load effects due to the respective indi-vidual environmental load types. Table F1 specifies a list of loadcases that shall be considered when this approach is followed,thereby to ensure that the required characteristic combined loadeffect, defined as the combined load effect with a return periodof 50 years, is obtained for the design. Each load case is definedas the combination of two or more environmental load types. Foreach load type in the load combination of a particular load case,the table specifies the characteristic value of the corresponding,separately determined load effect. The characteristic value isspecified in terms of the return period.

DET NORSKE VERITAS


Guidance note:Table F1 forms the basis for determination of the design com-bined load effect according to the linear combination format initem 501. Table F1 refers to a characteristic combined load effectwith a return period of 50 years and shall be used in conjunctionwith load factors specified in Sec.5.When it can be assumed that a load effect whose return period isTR occurs during the environmental state whose return period isTR, then the tabulated recurrence values in Table F1 can be usedas the return period for the load intensity parameter for the loadtype that causes the particular load effect in question. With thisinterpretation, Table F1 may be used as the basis for determina-tion of the characteristic combined load effect by linear combi-nation, in which case the analyses for the particular load cases ofTable F1 replace the more cumbersome searches for the charac-teristic load effect on environmental contours as described initem 301.When the direction of the loading is an important issue, it may beof particular relevance to maintain that the return periods ofTable F1 refer to load effects rather than to load intensities.For determination of the 50-year water level, two values shall beconsidered, viz. the high water level which is the 98% quantile inthe distribution of the annual maximum water level and the lowwater level which is the 2% quantile in the distribution of theannual minimum water level. For each load combination in TableF1, the most unfavourable value among the two shall be used forthe 50-year water level.


702 Every time a load combination is investigated, whichcontains a load effect contribution from wind load, the loadcombination shall be analysed for two different assumptionsabout the state of the wind turbine:

— wind turbine in operation (power production)— parked wind turbine (idling or standing still).

The largest load effect resulting from the corresponding twoanalyses shall be used for design.

Guidance note:It will usually not be clear beforehand which of the two assump-tions will produce the largest load effect, even if the blades of theparked turbine are put in the braking position to minimise thewind loads.In a ULS situation where the characteristic wind load effect is tobe taken as the 50-year wind load effect, the calculation for thewind turbine in operation will correspond to calculation of theload effect for a wind climate whose intensity is somewherebetween the cut-in wind speed and the cut-out wind speed. Forstall-regulated wind turbines, the cut-out wind speed dominatesthe extreme operational forces. For pitch-regulated wind tur-bines, the extreme operational forces occur for wind climateswhose intensities are near the 10-minute mean wind speed whereregulation starts, typically 13 to 14 m/s.For the parked wind turbine, the calculation in a ULS situationwill correspond to the calculation of the 50-year wind load effectas if the wind turbine would be in the parked condition during itsentire design life.


703 When it can be established as unlikely that the wind tur-bine will be in operation during the wave, ice, current and/or

water level conditions that form part of a load combinationunder investigation, the requirement of item 702 to analyse theload combination for the assumption of wind turbine in opera-tion may be too strict. When such an unlikely situation isencountered, the fulfilment of this requirement of item 702may be deviated from in the following manner: The load com-bination under investigation shall still be analysed for theassumption of wind turbine in operation; however, the require-ments to the return periods of the wave, ice, current and waterlevel conditions that the wind load effect is combined withmay be relaxed and set lower than the values specified in TableF1 for the particular load combination, as long as it can be doc-umented that the return period for the resulting combined loadeffect does not fall below 50 years.

Guidance note:When the fetch is limited and wind and waves have the samedirection in severe storms, then the wind climate intensity islikely to reach its extreme maximum at the same time as the waveclimate intensity reaches its extreme maximum, and it may beunlikely to see wind speeds below the cut-out wind speed duringthe presence of the 50-year wave climate. Likewise, it may beunlikely to see the 50-year wave climate during operation of thewind turbine.When the topography, e.g. in terms of a nearby coastline, forcesthe extreme maximum of the wind climate to take place at a dif-ferent time than the extreme maximum of the wave climate inten-sity, then it may be likely to see wind speeds below the cut-outwind speed during the presence of the 50-year wave climate.Likewise, it may be likely to see the 50-year wave climate duringoperation of the wind turbine.When a large fetch is present, there may be a phase differencebetween the occurrence of the extreme maximum of the wind cli-mate intensity and the extreme maximum of the wave climateintensity, and it may be likely to see wind speeds below the cut-out wind speed during the presence of the 50-year wave climate.Likewise, it may be likely to see the 50-year wave climate duringoperation of the wind turbine.


704 Every time a load case is investigated, which contains aload effect contribution from ice loads, loads from moving iceshall be considered as well as loads from fast-frozen ice andloads due to temperature fluctuations in the ice. 705 Load combination No. 5 in Table F1 is of relevance forstructures in waters which are covered by ice every year.Investigations for Load combination No. 5 in Table F1 can bewaived for structures in waters which are covered by ice lessfrequently than every year. 706 When a load case is investigated, which contains a loadeffect contribution from wave loads, loads from wave trains inless severe sea states than the sea state of the specified returnperiod shall be considered if these loads prove to produce alarger load effect than the sea state of the specified returnperiod.

Guidance note:Dynamic effects may cause less severe sea states than the seastate of the specified return period to produce more severe loadeffects, e.g. if these sea states imply wave trains arriving at thewind turbine structure with frequencies which coincide with afrequency of one of the natural vibration modes of the structure.

Table F1 Proposed load combinations for load calculations according to item 501Environmental load type and return period to define characteristic value of corresponding load effect

Limit state Load combination Wind Waves Current Ice Water level

ULS

1 50 years 5 years 5 years 50 years2 5 years 50 years 5 years 50 years3 5 years 5 years 50 years 50 years4 5 years 5 years 50 years Mean water level5 50 years 5 years 50 years Mean water level

DET NORSKE VERITAS


The possibility that waves break at the wind turbine structuremay play a role in this context and should be included in the con-siderations.


707 Co-directionality of wind and waves may be assumedfor calculation of the wave loads acting on the support struc-ture for all design cases except those corresponding to the windturbine in a parked (standstill or idling) design situation. Themisalignment of wind and wave directions in the parked situa-tion is to be accounted for.

Guidance note:Allowance for short term deviations from the mean wind direc-tion in the parked situation should be made by assuming a con-stant yaw misalignment. It is recommended to apply a yawmisalignment of ±15°.In areas where swell may be expected, special attention needs tobe given to swell, which has a low correlation with wind speedand wind direction.


708 The multi-directionality of the wind and the waves mayin some cases have an important influence on the loads actingon the support structure, depending primarily on whether thestructure is axisymmetric or not. For some design load casesthe load calculations may be undertaken by assuming that thewind and the waves are acting co-directionally from a single,worst case direction. 709 Characteristic extreme wind load effects are in thisstandard defined as wind load effects with a 50-year returnperiod. 5-year wind load effects form part of some load com-binations. When only wind load effects with a 100-year returnperiod are available, the 100-year wind load effects have to beconverted to 50-year values. This can be done by multiplica-tion by a conversion factor. Likewise, to the extent that 5-yearwind load effects are needed in load combinations and only 50-year values are available, the 50-year values have to be con-verted to 5-year values for use in these load combinations.

Guidance note:The ratio Fwind,100/Fwind,50 between the 100- and 50-year windload effects depends on the coefficient of variation in the distri-bution of the annual maximum wind load and can be used as aconversion factor to achieve the 50-year wind load effect Fwind,50in cases where only the 100-year value Fwind,100 is available.Unless data indicate otherwise, the ratio Fwind,100/Fwind,50 can betaken from Table F2. Table F2 also gives the ratio Fwind,5/Fwind,50 between the 5-year wind load effect Fwind,5 and the 50-year wind load effect Fwind,50. This is useful in some load com-binations that require the 5-year wind load effect. Table F2 isbased on an assumption of a Gumbel-distributed annual maxi-mum wind load effect.

The conversion factors are given as functions of the coefficientof variation of the annual maximum wind load effect. There is norequirement in this standard to document this coefficient of var-iation.Note that use of the conversion factor Fwind,5/Fwind,50 given inTable F2 to obtain the 5-year wind load effect from the 50-yearwind load effect will be nonconservative if the distribution of the

annual maximum wind load effect is not a Gumbel distributionand has a less heavy upper tail than the Gumbel distribution.Note also that for a particular wind turbine, the coefficient of var-iation of the annual maximum wind load effect may be differentdepending on whether the wind turbine is located on an offshorelocation or on an onshore location. For offshore wind turbines thecoefficient of variation is assumed to have a value of approxi-mately 20 to 30%.


F 800 Transient load cases 801 Actuation loads from the operation and control of thewind turbine produce transient wind loads on the wind turbinestructure. The following events produce transient loads andshall be considered:

— start up from stand-still or from idling— normal shutdown— emergency shutdown— normal fault events: faults in control system and loss of

electrical network connection— abnormal fault events: faults in protection system and

electrical systems— yawing.

802 The characteristic transient wind load effect shall be cal-culated as the maximum load effect during a 10-minute periodwhose wind intensity shall be taken as the most unfavourable10-minute mean wind speed in the range between the cut-inwind speed and the cut-out wind speed. In order to establish themost critical wind speed, i.e. the wind speed that produce themost severe load during the transient loading, gusts, turbu-lence, shift in wind direction, wind shear, timing of fault situ-ations, and grid loss in connection with deterministic gustsshall be considered.803 The characteristic transient wind load effect shall becombined with the 10-year wave load effect. The combinationmay be worked out according to the linear combination formatto produce the design load effect from the separately calculatedcharacteristic wind load effect and wave load effect. The com-bination may alternatively be worked out by direct simulationof the characteristic combined load effect in a structural anal-ysis in the time domain for simultaneously applied simulatedtime series of the wind load and the wave load.804 When transient wind loads are combined with waveloads, misalignment between wind and waves shall be consid-ered. For non-axisymmetric support structures, the most unfa-vourable wind load direction and wave load direction shall beassumed.

F 900 Load combination for the fatigue limit state901 For analyses of the fatigue limit state, a characteristicload effect distribution shall be applied which is the expectedload effect distribution over the design life. The expected loadeffect distribution is a distribution of stress ranges owing toload fluctuations and contains contributions from wind, waves,current, ice and water level as well as from possible othersources. The expected load effect distribution shall includecontributions from

— wind turbine in operation— parked wind turbine (idling and standing still)— start up— normal shutdown— control, protection and system faults, including loss of

electrical network connection— transport and assembly prior to putting the wind turbine to

service— maintenance and repair during the service life.

For fatigue analysis of a foundation pile, the characteristic loadeffect distribution shall include the history of stress ranges

Table F2 Conversion factors for wind load effectsCoefficient of

variation of annual maximum wind load

effect (%)

Ratio between 100- and 50-year wind

load effects, Fwind,100/Fwind,50

Ratio between 5- and 50-year wind

load effects, Fwind,5/Fwind,50

10 1.05 0.8515 1.06 0.8020 1.07 0.7525 1.08 0.7230 1.09 0.6835 1.10 0.64

DET NORSKE VERITAS


associated with the driving of the pile prior to installing thewind turbine and putting it to service.

Guidance note:The characteristic load effect distribution can be represented as ahistogram of stress ranges, i.e., the number of constant-rangestress cycles is given for each stress range in a sufficiently finediscretisation of the stress ranges. The individual contributions tothis load effect distribution from different sources can be repre-sented the same way.For contributions to the expected load effect distribution that areconsecutive in time or otherwise mutually exclusive, such as thecontribution from the transportation and installation phase andthe contribution from the in-service phase, the fatigue damagedue to each contribution can be calculated separately and addedtogether without introducing any particular prior combination ofthe contributions to the distribution. Alternatively, the differentcontributions can be combined to form the expected load effectdistribution prior to the fatigue damage calculation by addingtogether the number of stress cycles at each defined discretestress range from the respective underlying distributions.When the expected load effect distribution contains load effectswhich result from two or more concurrently acting load proc-esses, such as a wind load and a concurrent wave load, therespective underlying stress range distributions for separate windload effect and separate wave load effect need to be adequatelycombined prior to the calculation of the fatigue damage. Whenwind loads and wave loads act concurrently, it can be expectedthat their combined load effect distribution will contain some-what higher stress ranges than those of the underlying individualwave load effect and wind load effect distributions. The follow-ing idealised approach to combination of the two underlyingstress range distributions will usually be conservative: Thenumber of stress cycles of the combined load effect distributionis assumed equal to the number of stress cycles in that of theunderlying distributions (i.e. the distribution of wind stresscycles and the distribution of wave stress cycles) which containsthe highest number of cycles. Then the largest stress range in thewind load effect distribution is combined by the largest stressrange in the wave load effect distribution by simple superposi-tion, the second largest stress ranges are combined analogously,the third largest stress ranges the same, and so on. There may be some ambiguity involved with how concurrentwave load effects and wind load effects shall be combined toform the resulting load effect distribution for fatigue damage pre-diction. The proposed method of combination is idealised andimplies an assumption of colinear wind and waves. However,when combining wind load effects and wave load effects forfatigue, consideration of the distribution of the wind direction,the distribution of the wave direction and the distribution of themisalignment between wind and waves is important and may,relative to the situation with colinear wind and waves, oftenimply gains in terms of reduced fatigue damage that will morethan outweigh the possible effects of conservatism in the ideal-ised combination method. Caution should be exercised whencounting on such gains when wind and waves are not colinear,since situations may exist for which larger fatigue damage willaccumulate if the waves act perpendicular to the wind rather thancolinearly with the wind.


G. Load Effect AnalysisG 100 General101 Load effects, in terms of motions, displacements, andinternal forces and stresses in the wind turbine structure, shallbe determined with due regard for:

— their spatial and temporal nature including:

— possible non-linearities of the load— dynamic character of the response

— the relevant limit states for design checks— the desired accuracy in the relevant design phase.

102 Permanent loads, functional loads, deformation loads,and fire loads can generally be treated by static methods ofanalysis. Environmental loads (by wind, waves, current, iceand earthquake) and certain accidental loads (by impacts andexplosions) may require dynamic analysis. Inertia and damp-ing forces are important when the periods of steady-state loadsare close to natural periods or when transient loads occur.103 In general, three frequency bands need to be consideredfor offshore structures:

104 For fully restrained structures a static or dynamic wind-wave-structure-foundation analysis is required.105 Uncertainties in the analysis model are expected to betaken care of by the load and resistance factors. If uncertaintiesare particularly high, conservative assumptions shall be made.106 If analytical models are particularly uncertain, the sensi-tivity of the models and the parameters utilised in the modelsshall be examined. If geometric deviations or imperfectionshave a significant effect on load effects, conservative geomet-ric parameters shall be used in the calculation. 107 In the final design stage theoretical methods for predic-tion of important responses of any novel system should be ver-ified by appropriate model tests. Full scale tests may also beappropriate, in particular for large wind farms.108 Earthquake loads need only be considered for restrainedmodes of behaviour.

G 200 Global motion analysis201 The purpose of a motion analysis is to determine dis-placements, accelerations, velocities and hydrodynamic pres-sures relevant for the loading on the wind turbine supportstructure. Excitation by waves, current and wind should beconsidered.

G 300 Load effects in structures and foundation soils301 Displacements, forces and stresses in the structure andfoundation, shall be determined for relevant combinations ofloads by means of recognised methods, which take adequateaccount of the variation of loads in time and space, the motionsof the structure and the limit state which shall be verified.Characteristic values of the load effects shall be determined.302 Non-linear and dynamic effects associated with loadsand structural response, shall be accounted for whenever rele-vant.303 The stochastic nature of environmental loads shall beadequately accounted for.

H. Deformation LoadsH 100 General101 Deformation loads are loads caused by inflicted defor-mations such as:

— temperature loads— built-in deformations— settlement of foundations.

High frequency(HF)

Rigid body natural periods below the dominating wave periods, e.g. ringing and springing responses.

Wave frequency(WF)

Typically wave periods in the range 4 to 25 sec-onds. Applicable to all offshore structures located in the wave active zone.

Low frequency(LF)

This frequency band relates to slowly varying responses with natural periods beyond those of the dominating wave energy (typically slowly varying motions).

DET NORSKE VERITAS


H 200 Temperature loads201 Structures shall be designed for the most extreme tem-perature differences they may be exposed to. 202 The ambient sea or air temperature shall be calculated asthe extreme value whose return period is 50 years.203 Structures shall be designed for a solar radiation inten-

sity of 1 000 W/m2.

H 300 Settlements301 Settlement of the support structure and its foundationdue to vertical deformations of the supporting soils shall beconsidered. This includes consideration of differential settle-ments.

DET NORSKE VERITAS


SECTION 5LOAD AND RESISTANCE FACTORS

A. Load FactorsA 100 Load factors for the ULS101 Table A1 provides two sets of load factors to be usedwhen characteristic loads or load effects from different loadcategories are combined to form the design load or the designload effect for use in design. For analysis of the ULS, the setdenoted (a) shall be used when the characteristic environmen-tal load or load effect is established as the 98% quantile in the

distribution of the annual maximum load or load effect. Foranalyses of the ULS for abnormal wind load cases, the setdenoted (b) shall be users. The load factors apply in the operational condition as well asin the temporary condition. The load factors are generallyapplicable for all types of support structures and foundationsand they apply to design of support structures and foundationswhich qualify for design to the normal safety class.

Guidance note:Load factor set (a) is relevant for any design in the ULS exceptfor designs for abnormal load cases.


102 The characteristic environmental load effect (E), whichforms part of the load combinations of Table A1, is to be takenas the characteristic combined load effect, determined accord-ing to Sec.4, and representing the load effect that results fromtwo or more concurrently acting load processes.103 For permanent loads (G) and variable functional loads(Q), the load factor in the ULS shall normally be taken as ψ = 1.0 for load combinations (a) and (b).104 When a permanent load (G) or a variable functional load(Q) is a favourable load, then a load factor ψ = 0.9 shall beapplied for this load in combinations (a) and (b) of Table A1instead of the value of 1.0 otherwise required. The only excep-tion from this applies to favourable loads from foundation soilsin geotechnical engineering problems, for which a load factorψ =1.0 shall be applied. A load is a favourable load when areduced value of the load leads to an increased load effect inthe structure.

Guidance note:One example of a favourable load is the weight of a soil volumewhich has a stabilising effect in an overturning problem for afoundation.Another example is pretension and gravity loads that signifi-cantly relieve the total load response.


105 For design to high safety class, the requirements to theload factor γf specified for design to normal safety class inTable A1 for environmental loads shall be increased by 13%.

A 200 Load factor for the FLS201 The structure shall be able to resist expected fatigue

loads, which may occur during temporary and operationaldesign conditions. Whenever significant cyclic loads mayoccur in other phases, e.g. during manufacturing and transpor-tation, such cyclic loads shall be included in the fatigue loadestimates.202 The load factor γf in the FLS is 1.0 for all load catego-ries.

A 300 Load factor for the SLS301 For analysis of the SLS, the load factor γf is 1.0 for allload categories, both for temporary and operational designconditions.

B. Resistance Factors

B 100 Resistance factors for the ULS101 Resistance factors for the ULS are given in the relevantsections for design in the ULS. These resistance factors applyto design of support structures and foundations which qualifyfor design to normal safety class.102 For design of support structures and foundations to highsafety class, the same resistance factors as those required fordesign to normal safety class can be applied, provided the loadfactors for environmental loads are taken in accordance withA105.

B 200 Resistance factors for the FLS201 Resistance factors for the FLS are given in the relevantsections for design in the FLS.

B 300 Resistance factors for the SLS301 The material factor γm for the SLS shall be taken as 1.0.

Table A1 Load factors γf for the ULS

Load factor set Limit state

Load categories

G Q E D

(a) ULS ψ ψ 1.35 1.0(b) ULS for abnormal wind load cases ψ ψ 1.1 1.0

Load categories are:G = permanent loadQ = variable functional load, normally relevant only for design against ship impacts and for local design of platformsE = environmental loadD = deformation load.For description of load categories, see Sec.4.For values of ψ, see items 103 and 104.

DET NORSKE VERITAS


SECTION 6MATERIALS

A. Selection of Steel Materials and Inspection Principles

A 100 General101 This section describes the selection of steel materialsand inspection principles to be applied in design and construc-tion of offshore steel structures.

A 200 Design temperatures201 The design temperature is a reference temperature usedas a criterion for the selection of steel grades. The design tem-perature shall be based on lowest daily mean temperature. 202 In all cases where the service temperature is reduced bylocalised cryogenic storage or other cooling conditions, suchfactors shall be taken into account in establishing the minimumdesign temperatures. 203 The design temperature for floating units shall notexceed the lowest service temperature of the steel as definedfor various structural parts.204 External structures above the lowest waterline shall bedesigned with service temperatures equal to the lowest dailymean temperature for the area(s) where the unit is to operate. 205 Further details regarding design temperature for differ-ent structural elements are given in the object standards.206 External structures below the lowest waterline need notbe designed for service temperatures lower than 0°C. A higherservice temperature may be accepted if adequate supportingdata can be presented relative to the lowest average tempera-ture applicable to the relevant actual water depths.207 Internal structures in way of permanently heated roomsneed not be designed for service temperatures lower than 0°C.208 For fixed units, materials in structures above the lowestastronomical tide (LAT) shall be designed for service temper-atures down to the lowest daily mean temperature.209 Materials in structures below the lowest astronomicaltide (LAT) need not be designed for service temperatureslower than of 0°C. A higher service temperature may beaccepted if adequate supporting data can be presented relativeto the lowest daily mean temperature applicable for the rele-vant water depths.

A 300 Structural category301 The purpose of the structural categorisation is to assureadequate material and suitable inspection to avoid brittle frac-ture. The purpose of inspection is also to remove defects thatmay grow into fatigue cracks during service life.

Guidance note:Conditions that may result in brittle fracture are to be avoided.Brittle fracture may occur under a combination of:- presence of sharp defects such as cracks- high tensile stress in direction normal to planar defect(s)- material with low fracture toughness.Sharp cracks resulting from fabrication may be found by inspec-tion and repaired. Fatigue cracks may also be discovered duringservice life by inspection.High stresses in a component may occur due to welding. A com-plex connection is likely to provide more restraint and largerresidual stress than a simple one. This residual stress may bepartly removed by post weld heat treatment if necessary. Also acomplex connection shows a more three-dimensional stress statedue to external loading than simple connections. This stress statemay provide basis for a cleavage fracture.

The fracture toughness is dependent on temperature and materialthickness. These parameters are accounted for separately inselection of material. The resulting fracture toughness in theweld and the heat affected zone is also dependent on the fabrica-tion method.Thus, to avoid brittle fracture, first a material with suitable frac-ture toughness for the actual design temperature and thickness isselected. Then a proper fabrication method is used. In specialcases post weld heat treatment may be performed to reduce crackdriving stresses, see also DNV-OS-C401. A suitable amount ofinspection is carried out to remove planar defects larger thanthose considered acceptable. In this standard selection of mate-rial with appropriate fracture toughness and avoidance of unac-ceptable defects are achieved by linking different types ofconnections to different structural categories and inspection cat-egories.


302 Components are classified into structural categoriesaccording to the following criteria:

— significance of component in terms of consequence of fail-ure

— stress condition at the considered detail that together withpossible weld defects or fatigue cracks may provoke brittlefracture.

Guidance note:The consequence of failure may be quantified in terms of residualstrength of the structure when considering failure of the actualcomponent.


303 The structural category for selection of materials shallbe determined according to principles given in Table A1.

Guidance note:Monopile structures are categorised as “Primary”, because theyare non-redundant structures whose stress pattern is primarilyuniaxial and whose risk of brittle fracture is negligible. Likewise, towers are also categorised as “Primary”.Tubular joints are categorised as “Special” due to their biaxial ortriaxial stress patterns and risk of brittle fracture. This will influ-ence the thickness limitations as specified in Table A8.


304 Requirements and guidance for manufacturing of steelmaterials are given in DNV-OS-C401. For supplementaryguidance, reference is made to ENV 1090-1 and ENV 1090-5.Steel materials and products shall be delivered with inspectiondocuments as defined in EN 10204 or in an equivalent stand-

Table A1 Structural categories for selection of materialsStructural category

Principles for determination of structural category

Special Structural parts where failure will have substantial consequences and are subject to a stress condition that may increase the probability of a brittle fracture.1)

Primary Structural parts where failure will have substantial consequences.

Secondary Structural parts where failure will be without significant consequence.

1) In complex joints a triaxial or biaxial stress pattern will be present. This may give conditions for brittle fracture where ten-sile stresses are present in addition to presence of defects and material with low fracture toughness.

DET NORSKE VERITAS


ard. Unless otherwise specified, material certificates accordingto Table A2 shall be presented.

305 Requirements for type and extent of inspection of weldsare given in DNV-OS-C401 depending on assigned inspectioncategory for the welds. The requirements are based on the con-sideration of fatigue damage and assessment of general fabri-cation quality. 306 The inspection category is by default related to the struc-tural category according to Table A3.

307 The weld connection between two components shall beassigned an inspection category according to the highest cate-gory of the joined components. For stiffened plates, the weldconnection between stiffener and stringer and girder web to theplate may be inspected according to inspection Category III.308 If the fabrication quality is assessed by testing, or if it isof a well known quality based on previous experience, theextent of inspection required for elements within structural cat-egory primary may be reduced, but the extent must not be lessthan that for inspection Category III.309 Fatigue-critical details within structural category pri-mary and secondary shall be inspected according to require-ments given for Category I. This requirement applies tofatigue-critical details in the support structure and the founda-tion, but not in the tower.310 Welds in fatigue-critical areas not accessible for inspec-tion and repair during operation shall be inspected according torequirements in Category I during construction.311 For monopile type structures, the longitudinal welds inthe monopile and in the transition piece to the grouted connec-tion shall be inspected according to requirements given forCategory I.

A 400 Structural steel401 Wherever the subsequent requirements for steel gradesare dependent on plate thickness, these requirements are basedon the nominal thickness as built.402 The requirements in this subsection deal with the selec-tion of various structural steel grades in compliance with the

requirements given in DNV-OS-B101. Where other codes orstandards have been agreed on and utilised in the specificationof steels, the application of such steel grades within the struc-ture shall be specially considered.403 The steel grades selected for structural components shallbe related to calculated stresses and requirements to toughnessproperties. Requirements for toughness properties are in gen-eral based on the Charpy V-notch test and are dependent ondesign temperature, structural category and thickness of thecomponent in question. 404 The material toughness may also be evaluated by frac-ture mechanics testing in special cases.405 In structural cross-joints where high tensile stresses areacting perpendicular to the plane of the plate, the plate materialshall be tested to prove the ability to resist lamellar tearing, Z-quality, see 411.406 Requirements for forgings and castings are given inDNV-OS-B101.407 Material designations for steel are given in terms of astrength group and a specified minimum yield stress accordingto steel grade definitions given in DNV-OS-B101 Ch.2 Sec.1.The steel grades are referred to as NV grades. Structural steeldesignations for various strength groups are referred to asgiven in Table A4.

408 Each strength group consists of two parallel series ofsteel grades:

— steels of normal weldability— steels of improved weldability.

The two series are intended for the same applications. How-ever, the improved weldability grades have, in addition toleaner chemistry and better weldability, extra margins toaccount for reduced toughness after welding. These grades arealso limited to a specified minimum yield stress of 500 N/mm2.409 Conversions between NV grades as used in Table A4and steel grades used in the EN10025-2 standard are used forthe sole purpose of determining plate thicknesses and are givenin Table A5. The number of one-to-one conversions betweenNV grades and EN10025-2 grades given in Table A5 is lim-ited, because the E-qualities of the NV grades are not definedin EN10025-2 and because no qualities with specified mini-mum yield stress fy greater than 355 MPa are given inEN10025-2.

Table A2 Material certificatesCertification process Material

certificate (EN10204)

Structural category

Test certificateAs work certificate, inspection and tests witnessed and signed by an independent third party body

3.2 Special

Work certificateTest results of all specified tests from samples taken from the prod-ucts supplied. Inspection and tests witnessed and signed by QA department

3.1 Primary

Test reportConfirmation by the manufacturer that the supplied products fulfil the purchase specification, and test data from regular production, not necessarily from products supplied

2.2 Secondary

Table A3 Inspection categories Inspection category Structural category

I SpecialII PrimaryIII Secondary

Table A4 Material designations

Designation Strength group Specified minimum yield stress fy (N/mm2)1)

NV Normal strength steel (NS) 235

NV-27High strength

steel (HS)

265NV-32 315NV-36 355NV-40 390NV-420

Extra high strength steel

(EHS)

420NV-460 460NV-500 500NV-550 550NV-620 620NV-690 690

1) For steels of improved weldability the required specified minimum yield stress is reduced for increasing material thickness, see DNV-OS-B101.

DET NORSKE VERITAS


Guidance note:Important notes to the conversions between NV grades andEN10025-2 grades in Table A5:NV grades are, in general, better steel qualities than comparableEN10025-2 grades. For example, all NV grades except NV A andNV B, are fully killed and fine grain treated. This is the case onlyfor the J2G3 and K2G3 grades in EN10025-2.The delivery condition is specified as a function of thickness forall NV grades, while this is either optional or at the manufac-turer’s discretion in EN10025-2. The steel manufacturing process is also at the manufacturer’soption in EN10025-2, while only the electric process or one ofthe basic oxygen processes is generally allowed according to theDNV standard.For the grades NV A, NV B and NV D, an averaged impactenergy of minimum 27 Joule is specified for thicknesses up toand including 50 mm. For larger thicknesses, higher energyrequirements are specified. EN10025-2 requires an averagedimpact energy of minimum 27 Joule regardless of thickness.Concerning NV A36 and NV D36, minimum 34 Joule averagedimpact energy is required for thicknesses below 50 mm, 41 Joulefor thicknesses between 50 and 70 mm, and 50 Joule for thick-nesses above 70 mm. EN10025-2 specifies 27 Joule averagedimpact energy for the S355J0 and S355J2G3 grades and 40 Joulefor the S355K2G3 grade.In EN10025-2, minimum specified mechanical properties (yieldstress, tensile strength range and elongation) are thicknessdependent. The corresponding properties for NV grades are spec-ified independently of thickness.


410 Conversions between NV grades as used in Table A4and steel grades used in the EN10025-3 standard are used forthe sole purpose of determining plate thicknesses and are givenin Table A6.

Guidance note:Important notes to the conversions between NV grades andEN10025-3 grades in Table A6:The conversions are based on comparable requirements tostrength and toughness.Because EN10025-3 specifies requirements to fine grain treat-ment, the EN10025-3 grades are in general better grades thancorresponding grades listed in EN10025-2 and can be consideredequivalent with the corresponding NV grades.


411 Within each defined strength group, different steelgrades are given, depending on the required impact toughnessproperties. The grades are referred to as A, B, D, E, and F fornormal weldability grades and AW, BW, DW, and EW forimproved weldability grades as defined in Table A7.Additional symbol:

Z = steel grade of proven through-thickness properties.This symbol is omitted for steels of improved weldabil-ity although improved through-thickness properties arerequired.

412 The grade of steel to be used shall in general be selectedaccording to the design temperature and the thickness for theapplicable structural category as specified in Table A8. Thesteel grades in Table A8 are NV grade designations.

Table A5 Steel grade conversionsNV grade EN10025-2

NVANVBNVDNVE

S235JR+NS235J0

S235J2+N–

NV A27NV D27NV E27

S275J0S275J2+N

–NV A32NV D32NV E32

–––

NV A36NV D36NV E36

S355J0S355K2+N and S355J2+N

–NV A40NV D40NV E40

NV D420NV E420NV F420NV D460NV E460

––––––––

NV F460 –

Table A6 Steel grade conversionsNV grade EN10025-3 grade

NVANVBNVDNVE

––––

NV A27NV D27NV E27

–S275N

S275NLNV A32NV D32NV E32

–––

NV A36NV D36NV E36

–S355N

S355NLNV A40NV D40NV E40

NV D420NV E420NV F420NV D460NV E460

–––

S420NL––

S460NS460NL

NV F460 –

Table A7 Applicable steel grades

Strength group

GradeTest

temperature (ºC)Normal weld-ability

Improved weldability

NS

A – Not testedB 1) BW 0D DW –20E EW –40

HS

AH AHW 0DH DHW –20EH EHW –40FH – –60

EHS

AEH – 0DEH DEHW –20EEH EEHW –40FEH – –60

1) Charpy V-notch tests are required for thickness above 25 mm but is subject to agreement between the contracting parties for thickness of 25 mm or less.

DET NORSKE VERITAS


413 Selection of a better steel grade than minimum required indesign shall not lead to more stringent requirements in fabrication. 414 Grade of steel to be used for thickness less than 10 mmand/or design temperature above 0°C will be specially consid-ered in each case. For submerged structures, i.e. for structuresbelow LAT−1.5 m e.g. in the North Sea, the design tempera-ture will be somewhat above 0°C (typically 2°C) and specialconsiderations can be made in such cases. 415 Welded steel plates and sections of thickness exceedingthe upper limits for the actual steel grade as given in Table A8shall be evaluated in each individual case with respect to thefitness for purpose of the weldments. The evaluation should bebased on fracture mechanics testing and analysis, e.g. inaccordance with BS 7910.416 For regions subjected to compressive and/or low tensilestresses, consideration may be given to the use of lower steelgrades than stated in Table A8.417 The use of steels with specified minimum yield stressgreater than 550 N/mm2 (NV550) shall be subject to specialconsideration for applications where anaerobic environmentalconditions such as stagnant water, organically active mud(bacteria) and hydrogen sulphide may predominate.418 Predominantly anaerobic conditions can for this purposebe characterised by a concentration of sulphate reducing bac-teria, SRB, in the order of magnitude < 103 SRB/ml, deter-mined by method according to NACE TPC Publication No. 3. 419 The susceptibility of the steel to hydrogen-inducedstress cracking (HISC) shall be specially considered whenused for critical applications. See also Sec.11.420 The grade of steel to be used shall in general be selectedsuch that there will be no risk of pitting damage.

B. Selection of Concrete Materials B 100 General101 For selection of structural concrete materials, DNV-OS-C502 Sec.4, “Structural Concrete and Materials” shall apply.102 The present subsection, B, provides a short summary ofDNV-OS-C502 Sec.4, focusing on issues which typically per-tain to offshore concrete structures, but not necessarily tostandard concrete design. For all design purposes, the usershould always refer to the complete description in DNV-OS-C502 and the text in Subsection B shall be considered applica-tion text for the text of DNV-OS-C502 with respect to offshorewind turbine concrete structures.

B 200 Material requirements201 The materials selected for the load-bearing structuresshall be suitable for the purpose. The material properties andverification that these materials fulfil the requirements shall bedocumented.202 The materials, all structural components and the struc-ture itself shall be ensured to maintain the specified qualityduring all stages of construction and for the intended structurallife.203 Constituent materials for structural concrete are cement,aggregates and water. Structural concrete may also includeadmixtures and additions.204 Constituent materials shall be sound, durable, free fromdefects and suitable for making concrete that will attain andretain the required properties. Constituent materials shall notcontain harmful ingredients in quantities that can be detrimen-tal to the durability of the concrete or cause corrosion of thereinforcement and shall be suitable for the intended use.205 The following types of Portland cement are, in general,assumed to be suitable for use in structural concrete and/orgrout in a marine environment if unmixed with other cements:

— Portland cements— Portland composite cements— Blastfurnace cements, with high clinker content.

Provided suitability is demonstrated also the following typesof cement may be considered:

— Blastfurnace cements— Pozzolanic cements— Composite cements.

The above types of cement have characteristics specified ininternational and national standards. They can be specified ingrades based on the 28-day strength in mortar. Cements shallnormally be classified as normal hardening, rapid hardening orslowly hardening cements.

Guidance note:Low heat cement may be used where heat of hydration may havean adverse effect on the concrete during curing.


206 The required water content is to be determined by con-sidering the strength and durability of hardened concrete andthe workability of fresh concrete. The water-to-cement ratio byweight may be used as a measure. For requirements to thewater-to-cement ratio, see B305.207 Salt water, such as raw seawater, shall not be used asmixing or curing water for structural concrete.208 Normal weight aggregates shall, in general, be of naturalmineral substances. They shall be either crushed or uncrushedwith particle sizes, grading and shapes such that they are suit-able for the production of concrete. Relevant properties ofaggregate shall be defined, e.g. type of material, shape, surface

Table A8 Thickness limitations (mm) of structural steels for different structural categories and design temperatures (ºC)

Structural Category Grade ≥ 10 0 –10 –20

Secondary

AB/BWD/DWE/EW

AH/AHWDH/DHWEH/EHW

FHAEH

DEH/DEHWEEH/EEHW

FEH

3060

15015050

10015015060

150150150

30601501505010015015060150150150

2550

1001504080

15015050

100150150

20408015030601501504080150150

Primary

AB/BWD/DWE/EW

AH/AHWDH/DHWEH/EHW

FHAEH

DEH/DEHWEEH/EEHW

FEH

304060

1502550

1001503060

150150

20306015025501001503060150150

102550

100204080

1502550

100150

N.A.204080153060150204080150

Special

D/DWE/EW

AH/AHWDH/DHWEH/EHW

FHAEH

DEH/DEHWEEH/EEHW

FEH

3560102550

100153060

150

3060102550100153060150

2550

N.A.204080102550

100

2040

N.A.153060

N.A.204080

N.A. = no application

DET NORSKE VERITAS


texture, physical properties and chemical properties. Aggregates shall be free from harmful substances in quantitiesthat can affect the properties and the durability of the concreteadversely. Examples of harmful substances are claylike andsilty particles, organic materials and sulphates and other salts.209 Aggregates shall be evaluated for risk of Alkali SilicaReaction (ASR) in concrete according to internationally recog-nised test methods. Suspect aggregates shall not be used unlessspecifically tested and approved. The approval of aggregatesthat might combine with the hydration products of the cementto cause ASR shall state which cement the approval applies to.The aggregate for structural concrete shall have sufficientstrength and durability.210 An appropriate grading of the fine and coarse aggregatesfor use in concrete shall be established. The grading and shapecharacteristics of the aggregates shall be consistent throughoutthe concrete production.211 Maximum aggregate size is to be specified based onconsiderations concerning concrete properties, spacing of rein-forcement and cover to the reinforcement.212 Latent hydraulic or pozzolanic supplementary materialssuch as silica fume, pulverized fly ash and granulated blast fur-nace slag may be used as additions. The amount is dependenton requirements to workability of fresh concrete and requiredproperties of the hardened concrete. The content of silica fumeused as additions should normally not exceed 10% of theweight of Portland cement clinker. When fly ash, slag or otherpozzolana is used as additions, their content should normallynot exceed 35% of the total weight of cement and additions.When Portland cement is used in combination with onlyground granulated blast furnace slag, the slag content may beincreased. The clinker content shall, however, not be less than30% of the total weight of cement and slag.213 The composition and properties of repair materials shallbe such that the material fulfils its intended use. Only materialswith established suitability shall be used. Emphasis shall begiven to ensure that such materials are compatible with theadjacent material, particularly with regard to the elasticity andtemperature dependent properties.

B 300 Concrete301 Normal Strength Concrete is a concrete of grade C30 toC65.302 High Strength Concrete is a concrete of grade in excessof C65.303 The concrete composition and the constituent materialsshall be selected to satisfy the requirements of DNV-OS-C502and the project specifications for fresh and hardened concretesuch as consistency, density, strength, durability and protec-tion of embedded steel against corrosion. Due account shall betaken of the methods of execution to be applied. The require-ments of the fresh concrete shall ensure that the material isfully workable in all stages of its manufacture, transport, plac-ing and compaction.304 The required properties of fresh and hardened concreteshall be specified. These required properties shall be verifiedby the use of recognised testing methods, international stand-ards or recognised national standards. Recognised standard isrelevant ASTM, ACI and EN standard.305 Compressive strength shall always be specified. In addi-tion, tensile strength, modulus of elasticity (E-modulus) andfracture energy may be specified. Properties which can causecracking of structural concrete shall be accounted for, i.e.creep, shrinkage, heat of hydration, thermal expansion andsimilar effects. The durability of structural concrete is relatedto permeability, absorption, diffusion and resistance to physi-cal and chemical attacks in the given environment, a lowwater/cement-binder ratio is generally required in order to

obtain adequate durability. The concrete shall normally have awater/cement-binder ratio not greater than 0.45. In the splashzone, this ratio shall not be higher than 0.40.306 The demands given for cement content in DNV-OS-C502 Sec.4 D309 shall be considered demands as for cement/filler content calculated according to a recognised standard.The demands may be waived based on conditions such as lessstrict national requirements or track records for good perform-ance and durability in marine environments for similar struc-tures.307 The concrete grades are defined as specified in DNV-OS-C502 Sec.6.The properties of hardened concrete are generally related to theconcrete grade. For concrete exposed to seawater the minimumgrade is C40. For concrete which is not directly exposed to themarine environment, the grade shall not be less than C30.308 The concrete grades are defined in DNV-OS-C502Sec.6 Table C1 as a function of the Characteristic CompressiveCylinder strength of the concrete, fcck. However, the gradenumbers are related to the Characteristic Compressive Cubestrength of the concrete, fck (100 mm cube).

B 400 Grout and mortar401 The mix design of grout and mortar shall be specified forits designated purpose.402 The constituents of grout and mortar shall meet the sametype of requirements for their properties as those given for theconstituents of concrete.

B 500 Reinforcement steel501 Reinforcements shall be suitable for their intended serv-ice conditions and are to have adequate properties with respectto strength, ductility, toughness, weldability, bond properties(ribbed), corrosion resistance and chemical composition.These properties shall be specified by the supplier or deter-mined by a recognised test method.502 Reinforcement steel shall comply with ISO 6935, Parts2 and 3 or relevant national or international standards for rein-forcement steel.503 Consistency shall be ensured between material proper-ties assumed in the design and the requirements of the standardused. In general, hot-rolled, ribbed bars of weldable qualityand with high ductility shall be used. Where the use of seismicdetailing is required, the reinforcement provided shall meet theductility requirements of the reference standard used in thedesign.504 Fatigue properties and S-N curves shall be consistentwith the assumptions of design.

B 600 Prestressing steel601 Prestressing steel shall comply with ISO 6934 and/orrelevant national or international standards for prestressingsteel.

C. Grout Materials and Material TestingC 100 General101 The grout materials for grouted connections shall com-ply with relevant requirements given for both concrete andgrout in DNV-OS-C502 “Offshore Concrete Structures”,Sec.3, as well as with requirements given for concrete in thisstandard (DNV-OS-J101) Sec.6 B. 102 The materials shall have sufficient workability to ensurefilling of the annulus without establishing air pockets or waterpockets or both in the grout.103 Test specimens are to be made with varying mix propor-

DET NORSKE VERITAS


tions to simulate the batching tolerances under field condi-tions. Grout mixes shall as a minimum be tested for thefollowing properties:

— density— air content— workability— viscosity— stability (separation and bleeding)— setting time— compressive strength— shrinkage/expansion — effect of admixtures and compatibility of admixtures.

104 For some applications, other properties of the grout mixmay be required to be confirmed by testing. For example, ifhardening of the grout may introduce unacceptable thermalstrains in the structure, it shall be confirmed that the maximumtemperature-rise caused by the hardening process is withinacceptable limits. 105 Samples for testing of the grout quality shall preferablybe taken from the emerging, surplus grout. If this is not possi-ble, other means of monitoring the density of the return groutare to be provided.106 Tests on grout samples shall be carried out in order toverify the characteristic compressive strength of the grout. Thecharacteristic compressive strength is normally defined as thecompressive strength after setting 28 days at 20°C or theequivalent. If the grout is to be subjected to loading before thecharacteristic design strength has been achieved, for exampledue to installation of other structures or due to wave and wind

loading before 28 days have passed, the assumed allowablegrout strength at the time of the loading shall be verified. Cur-ing of the specimens shall take place under conditions whichare as similar to the curing conditions of the placed grout aspossible.107 The compressive strength is normally to be tested bymaking sets of 5 test specimens each. One such set of 5 speci-mens shall be used every time a test is to be carried out. Eachspecimen is to be taken from a single, random sample. Thetotal number of test sets required according to this specifica-tion shall be obtained for every consumed 200 m3 of grout,once per shift or once per grouted compartment/annulus,whichever gives the largest number of test specimens. The testspecimens are to be adequately marked and recorded foridentification.

Guidance note:The specified requirement to the number of test sets usuallyimplies that one set consisting of 5 test specimens is obtainedfrom each annulus. It is acceptable to calculate the grout strengthas the average strength over all obtained samples, i.e. over anumber of samples equal to five times the number of groutedstructures, provided that it is demonstrated that the compressivestrengths obtained from the tests on these samples belong statis-tically to the same population.


C 200 Experimental verification201 If no sufficient documentation of the material propertiesof the grout is available experimental verification of the prop-erties must be carried out.

DET NORSKE VERITAS


SECTION 7DESIGN OF STEEL STRUCTURES

A. Ultimate Limit States – General

A 100 General101 This subsection gives provisions for checking the ulti-mate limit states for typical structural elements used in off-shore steel structures. 102 The ultimate strength capacity of structural elements inyielding and buckling shall be assessed using a rational andjustifiable engineering approach.103 The structural capacity of all structural components shallbe checked. The capacity check shall consider both excessiveyielding and buckling.104 Simplified assumptions regarding stress distributionsmay be used provided that the assumptions are made in accord-ance with generally accepted practice, or in accordance withsufficiently comprehensive experience or tests.105 Prediction of structural capacity shall be carried out withdue consideration of capacity reductions which are implied bythe corrosion allowance specified in Sec.11.

Guidance note:The increase in wall thickness for a structural component, addedto allow for corrosion, shall not be included in the calculation ofthe structural capacity of the component.


A 200 Structural analysis201 The structural analysis may be carried out as linear elas-tic, simplified rigid-plastic, or elastic-plastic analyses. Bothfirst order or second order analyses may be applied. In allcases, the structural detailing with respect to strength and duc-tility requirement shall conform to the assumption made for theanalysis.202 When plastic or elastic-plastic analyses are used forstructures exposed to cyclic loading, i.e. wind turbine loadsand wave loads, checks shall be carried out to verify that thestructure will shake down without excessive plastic deforma-tions or fracture due to repeated yielding. A characteristic ordesign cyclic load history needs to be defined in such a waythat the structural reliability in case of cyclic loading, e.g.storm loading, is not less than the structural reliability in theULS for non-cyclic loads.203 In case of linear analysis combined with the resistanceformulations set down in this standard, shakedown can beassumed without further checks.204 If plastic or elastic-plastic structural analyses are usedfor determining the sectional stress resultants, limitations tothe width-to-thickness ratios apply. Relevant width-to-thick-ness ratios are found in the relevant codes used for capacitychecks.205 When plastic analysis and/or plastic capacity checks areused (cross section type I and II, according to Appendix H), themembers shall be capable of forming plastic hinges with suffi-cient rotation capacity to enable the required redistribution ofbending moments to develop. It shall also be checked that theload pattern will not be changed due to the deformations.206 Cross sections of beams are divided into different typesdependent on their ability to develop plastic hinges. A methodfor determination of cross sectional types is given inAppendix H.

A 300 Ductility301 It is a fundamental requirement that all failure modes aresufficiently ductile such that the structural behaviour will be inaccordance with the anticipated model used for determinationof the responses. In general all design procedures, regardless ofanalysis method, will not capture the true structural behaviour.Ductile failure modes will allow the structure to redistributeforces in accordance with the presupposed static model. Brittlefailure modes shall therefore be avoided, or they shall be veri-fied to have excess resistance compared to ductile modes andin this way protect the structure from brittle failure.302 The following sources for brittle structural behaviourmay need to be considered for a steel structure:

— unstable fracture caused by a combination of the followingfactors: brittle material, low temperature in the steel, adesign resulting in high local stresses and the possibilitiesfor weld defects

— structural details where ultimate resistance is reached withplastic deformations only in limited areas, making the glo-bal behaviour brittle

— shell buckling— buckling where interaction between local and global buck-

ling modes occurs.

A 400 Yield check401 Structural members for which excessive yielding is apossible mode of failure, are to be investigated for yielding.402 Local peak stresses from linear elastic analysis in areaswith pronounced geometrical changes, may exceed the yieldstress provided that the adjacent structural parts has capacityfor the redistributed stresses.403 Yield checks may be performed based on net sectionalproperties. For large volume hull structures gross scantlingsmay be applied.404 For yield check of welded connections, seeSubsection H regarding welded connections.

A 500 Buckling check501 Requirements for the elements of the cross section notfulfilling requirements to cross section type III need to bechecked for local buckling.502 Buckling analysis shall be based on the characteristicbuckling resistance for the most unfavourable buckling mode.503 The characteristic buckling strength shall be based onthe 5th percentile of test results. 504 Initial imperfections and residual stresses in structuralmembers shall be accounted for. 505 It shall be ensured that there is conformity between theinitial imperfections in the buckling resistance formulae andthe tolerances in the applied fabrication standard.

B. Ultimate Limit States – Shell StructuresB 100 General101 The buckling stability of shell structures may bechecked according to DNV-RP-C202 or Eurocode 3/EN 1993-1-1 and ENV 1993-1-6.102 For interaction between shell buckling and columnbuckling, DNV-RP-C202 may be used.

DET NORSKE VERITAS


103 If DNV-RP-C202 is applied, the material factor forshells shall be in accordance with Table B1.

Guidance note:Note that the slenderness is based on the buckling mode underconsideration.

λ = reduced slenderness parameter

=

fy = specified minimum yield stressσe = elastic buckling stress for the buckling mode under con-

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

104 For global buckling, the material factor γM shall be 1.2as a minimum, cf. IEC61400-1.

C. Ultimate Limit States – Tubular Members, Tubular Joints and Conical Transitions

C 100 General101 Tubular members shall be checked according to recog-nised standards. Standards for the strength of tubular memberstypically have limitations with respect to the D/t ratio and withrespect to the effect of hydrostatic pressure. The followingstandards are relevant for checking tubular member strength:Classification Notes 30.1 Sec.2 (Compact cross sections), APIRP2A-LRFD (D/t < 300), Eurocode 3/EN 1993-1-1 and ENV1993-1-6 or NORSOK N-004 (D/t < 120). For interactionbetween local shell buckling and column buckling and foreffect of external pressure, DNV-RP-C202 may be used.

Guidance note:Compact tubular cross section is in this context defined as whenthe diameter (D) to thickness (t) ratio satisfy the following crite-rion:

E = modulus of elasticity andfy = minimum yield strength


102 Tubular members with external pressure, tubular jointsand conical transitions may be checked according to API RP2A – LRFD or NORSOK N-004.103 The material factor γM for tubular structures is 1.10.104 For global buckling, the material factor γM shall be 1.2as a minimum, cf. IEC61400-1.

D. Ultimate Limit States – Non-Tubular Beams, Columns and Frames

D 100 General101 The design of members shall take into account the possiblelimits on the resistance of the cross section due to local buckling.

102 Buckling checks may be performed according to Classi-fication Notes 30.1.103 Capacity checks may be performed according to recog-nised standards such as EN 1993-1-1 or AISC LRFD Manualof Steel Construction.104 The material factors according to Table D1 shall be usedif EN 1993-1-1 is used for calculation of structural resistance.

E. Ultimate Limit States – Special Provisions for Plating and Stiffeners

E 100 Scope101 The requirements in E will normally give minimumscantlings to plate and stiffened panels with respect to yield.102 The buckling stability of plates may be checked accord-ing to DNV-RP-C201.

E 200 Minimum thickness201 The thickness of plates should not to be less than:

E 300 Bending of plating301 The thickness of plating subjected to lateral pressureshall not be less than:

Table B1 Material factors γM for bucklingType of structure λ ≤ 0.5 0.5 < λ < 1.0 λ ≥ 1.0Girder, beams stiffeners on shells 1.10 1.10 1.10

Shells of single curvature (cylindrical shells, coni-cal shells)

1.10 0.80 + 0.60 λ 1.40

fyσe-----

Dt---- 0.5 E

fy----≤

Table D1 Material factors used with EN 1993-1-1Type of calculation Material factor 1) ValueResistance of Class 1, 2 or 3 cross sections

γM0 1.10

Resistance of Class 4 cross sections

γM1 1.10

Resistance of members to buckling

γM1 1.10

1) Symbols according to EN 1993-1-1.

fyd = design yield strength fy/γM fy is the minimum yield stress (N/mm2) as given in Sec.6 Table A3

t0 = 7 mm for primary structural elements= 5 mm for secondary structural elements

γM = material factor for steel= 1.10.

ka = correction factor for aspect ratio of plate field= (1.1 − 0.25 s/l)2

= maximum 1.0 for s/l = 0.4= minimum 0.72 for s/l = 1.0

kr = correction factor for curvature perpendicular to the stiffeners

= (1 − 0.5 s/rc)rc = radius of curvature (m)s = stiffener spacing (m), measured along the plating

pd = design pressure (kN/m2) as given in Sec.4σpd1 = design bending stress

= 1.3 (fyd −σjd), but less than fyd = fy /γM

t14.3t0

fyd--------------- (mm)=

(mm) 8.15

1 pppd

dra

k

pskkt

σ=

DET NORSKE VERITAS


Guidance note:The design bending stress σpd1 is given as a bi-linear capacitycurve.


E 400 Stiffeners401 The section modulus for longitudinals, beams, framesand other stiffeners subjected to lateral pressure shall not beless than:

402 The formula given in 401 shall be regarded as therequirement about an axis parallel to the plating. As an approx-imation the requirement for standard section modulus for stiff-eners at an oblique angle with the plating may be obtained ifthe formula in 401 is multiplied by the factor:

403 Stiffeners with sniped ends may be accepted wheredynamic stresses are small and vibrations are considered to beof small importance, provided that the plate thickness sup-ported by the stiffener is not less than:

In such cases the section modulus of the stiffener calculated asindicated in 401 is normally to be based on the followingparameter values:

The stiffeners should normally be snipped with an angle ofmaximum 30º.

Guidance note:For typical sniped end detail as described above, a stress rangelower than 30 MPa can be considered as a small dynamic stress.


F. Ultimate Limit States – Special Provisions for Girders and Girder Systems

F 100 Scope101 The requirements in F give minimum scantlings to sim-ple girders with respect to yield. Further procedures for the cal-culations of complex girder systems are indicated.102 The buckling stability of girders may be checkedaccording to Classification Notes No. 30.1.

F 200 Minimum thickness201 The thickness of web and flange plating is not to be lessthan given in E200 and E300.

F 300 Bending and shear301 The requirements for section modulus and web area areapplicable to simple girders supporting stiffeners and to othergirders exposed to linearly distributed lateral pressures. It isassumed that the girder satisfies the basic assumptions of sim-ple beam theory and that the supported members are approxi-mately evenly spaced and has similar support conditions atboth ends. Other loads will have to be specially considered.302 When boundary conditions for individual girders are notpredictable due to dependence on adjacent structures, directcalculations according to the procedures given in F700 will berequired.303 The section modulus and web area of the girder shall betaken in accordance with particulars as given in F600 andF700. Structural modelling in connection with direct stressanalysis shall be based on the same particulars when applica-ble.

F 400 Effective flange401 The effective plate flange area is defined as the crosssectional area of plating within the effective flange width. Thecross section area of continuous stiffeners within the effectiveflange may be included. The effective flange width be is deter-mined by the following formula:

σjd = equivalent design stress for global in-plane mem-brane stress

kpp = fixation parameter for plate= 1.0 for clamped edges= 0.5 for simply supported edges.

l = stiffener span (m)km = bending moment factor, see Table G1σpd2 = design bending stress

= fyd − σjdkps = fixation parameter for stiffeners

= 1.0 if at least one end is clamped= 0.9 if both ends are simply supported.

α = angle between the stiffener web plane and the plane perpendicular to the plating.

km = 8kps = 0.9

)(mm 1015 minimum),(mm 10 3336

2

2⋅⋅=

pspdm

ds kk

splZ

σ

αcos1

t 16l 0.5s–( )spd

fyd-------------------------------- mm( )≥

Ce = as given in Figure 1 for various numbers of evenly spaced point loads (Np) on the span

b = full breadth of plate flange e.g. span of the stiffeners supported by the girder with effective flange be, see also 602.

l0 = distance between points of zero bending moments (m)= S for simply supported girders= 0.6 S for girders fixed at both ends

S = girder span as if simply supported, see also 602.

be Ceb=

DET NORSKE VERITAS


Figure 1 Graphs for the effective flange parameter C

F 500 Effective web501 Holes in girders will generally be accepted provided theshear stress level is acceptable and the buckling capacity andfatigue life are documented to be sufficient.

F 600 Strength requirements for simple girders601 Simple girders subjected to lateral pressure and whichare not taking part in the overall strength of the structure, shallcomply with the following minimum requirements:

— net section modulus according to 602— net web area according to 603.

602 Section modulus:

603 Net web area:

604 The km and kτ values referred to in 602 and 603 may becalculated according to general beam theory. In Table F1, kmand kτ values are given for some defined load and boundaryconditions. Note that the smallest km value shall be applied tosimple girders. For girders where brackets are fitted or theflange area has been partly increased due to large bendingmoment, a larger km value may be used outside the strength-ened region.

F 700 Complex girder system701 For girders that are parts of a complex 2- or 3-dimen-sional structural system, a complete structural analysis shall becarried out.702 Calculation methods or computer programs appliedshall take into account the effects of bending, shear, axial andtorsional deformation.703 The calculations shall reflect the structural response ofthe 2- or 3-dimensional structure considered, with due atten-tion to boundary conditions.704 For systems consisting of slender girders, calculationsbased on beam theory (frame work analysis) may be applied,with due attention to:

— shear area variation, e.g. cut-outs— moment of inertia variation

S = girder span (m). The web height of in-plane girders may be deducted. When brackets are fitted at the ends, the girder span S may be reduced by two thirds of the bracket arm length, provided the girder ends may be assumed clamped and provided the section modulus at the bracketed ends is satis-factory

b = breadth of load area (m) (plate flange) b may be determined as:

= 0.5 (l1 + l2) (m), l1 and l2 are the spans of the sup-ported stiffeners, or distance between girders

km = bending moment factor km–values in accordance with Table F1 may be applied

σpd2 = design bending stress= fyd − σjd

σjd = equivalent design stress for global in-plane mem-brane stress.

kτ = shear force factor kτ may be in accordance with 604

Ns = number of stiffeners between considered sec-tion and nearest supportThe Ns–value is in no case to be taken greater than (Np+1)/4

)(mm 10 36

2

2⋅=

pdm

dg k

bpSZ

σ

)(mm 10 23⋅−

=p

pdSdtW

PNSbpkA

τ

Np = number of supported stiffeners on the girder span

Ppd = average design point load (kN) from stiffeners between considered section and nearest support

τp = 0.5 fyd (N/mm2).

Table F1 Values of km and kτ

Load and boundary conditions Bending moment and shear force factors

Positions 1km1kτ1

2km2

–

3km3kτ3

1Support

2Field

3Support

12

0.5

24 12

0.5

0.38

14.2 8

0.63

0.5

8

0.5

15

0.3

23.3 10

0.7

0.2

16.8 7.5

0.8

0.33

7.8

0.67

DET NORSKE VERITAS


— effective flange— lateral buckling of girder flanges.

705 The most unfavourable of the loading conditions givenin Sec.4 shall be applied.706 For girders taking part in the overall strength of the unit,stresses due to the design pressures given in Sec.4 shall becombined with relevant overall stresses.

G. Ultimate Limit States – Slip-resistant Bolt Connections

G 100 General101 The requirements in G give the slip capacity of pre-ten-sioned bolt connections with high-strength bolts.102 A high-strength bolt is defined as a bolt that has an ulti-mate tensile strength larger than 800 N/mm2 and a yieldstrength which as a minimum is 80% of the ultimate tensilestrength.103 The bolt shall be pre-tensioned in accordance with inter-national recognised standards. Procedures for measurementand maintenance of the bolt tension shall be established.104 The design slip resistance Rd may be specified equal toor higher than the design loads Fd.

105 In addition, the slip resistant connection shall have thecapacity to withstand ULS and ALS loads as a bearing boltconnection. The capacity of a bolted connection may be deter-mined according to international recognised standards whichgive equivalent level of safety such as EN 1993-1-1 or AISCLRFD Manual of Steel Construction.106 The design slip resistance of a preloaded high-strengthbolt shall be taken as:

107 For high strength bolts, the controlled design pre-ten-sioning force in the bolts used in slip resistant connections are:

108 The design value of the friction coefficient μ is depend-ent on the specified class of surface treatment as given inDNV-OS-C401 Sec.7. The value of μ shall be taken accordingto Table G1.

109 The classification of any surface treatment shall bebased on tests or specimens representative of the surfaces usedin the structure using the procedure set out in DNV-OS-C401.110 Provided the contact surfaces have been treated in con-formity with DNV-OS-C401 Sec.7, the surface treatmentsgiven in Table G2 may be categorised without further testing.

111 Normal clearance for fitted bolts shall be assumed if nototherwise specified. The clearances are defined in Table G3.

112 Oversized holes in the outer ply of a slip resistant con-nection shall be covered by hardened washers.113 The nominal sizes of short slotted holes for slip resistantconnections shall not be greater than given in Table G4.

ks = hole clearance factor= 1.00 for standard clearances in the direction of

the force = 0.85 for oversized holes= 0.70 for long slotted holes in the direction of

the forcen = number of friction interfacesμ = friction coefficientγMs = 1.25 for standard clearances in the direction of

the force= 1.4 for oversize holes or long slotted holes in

the direction of the force= 1.1 for design shear forces with load factor

1.0.Fpd = design preloading force.

fub = ultimate tensile strength of the bolt

Rd Fd≥

pdMs

sd F

nkR

γμ

=

Fpd 0.7fubAs=

As = tensile stress area of the bolt (net area in the threaded part of the bolt).

Table G1 Friction coefficient μSurface category μ

A 0.5B 0.4C 0.3D 0.2

Table G2 Surface treatmentSurface category

Surface treatment

A

Surfaces blasted with shot or grit:

— with any loose rust removed, no pitting— spray metallised with aluminium— spray metallised with a zinc-based coating certi-

fied to prove a slip factor of not less than 0.5

BSurfaces blasted with shot or grit, and painted with an alkali-zinc silicate paint to produce a coating thickness of 50 to 80 μm

C Surfaces cleaned by wire brushing or flame cleaning, with any loose rust removed

D Surfaces not treated

Table G3 Clearances in bolt holes

Clearance type Clearancemm

Bolt diameter d (maximum)mm

Standard

1 12 and 142 16 to 243 27 to 364 42 to 485 566 64

Oversized

3 124 14 to 226 248 27

Table G4 Short slotted holesMaximum size

mmBolt diameter d (maximum)

mm(d + 1) by (d + 4) 12 and 14(d + 2) by (d + 6) 16 to 22(d + 2) by (d + 8) 24

(d + 3) by (d + 10) 27 and larger

DET NORSKE VERITAS


114 The nominal sizes of long slotted holes for slip resistantconnections shall not be greater than given in Table G5.

115 Long slots in an outer ply shall be covered by coverplates of appropriate dimensions and thickness. The holes inthe cover plate shall not be larger than standard holes.

H. Ultimate Limit States – Welded Connections

H 100 General101 The requirements in this subsection apply to types andsizes of welds.

H 200 Types of welded steel joints201 All types of butt joints should be welded from bothsides. Before welding is carried out from the second side,unsound weld metal shall be removed at the root by a suitablemethod.202 The connection of a plate abutting on another plate in atee joint or a cross joint may be made as indicated in Figure 2.203 The throat thickness of the weld is always to be measuredas the normal to the weld surface, as indicated in Figure 2d.204 The type of connection should be adopted as follows:

a) Full penetration weldImportant cross connections in structures exposed to highstress, especially dynamic, e.g. for special areas and fa-tigue utilised primary structure. All external welds in wayof opening to open sea e.g. pipes, sea-chests or tee-jointsas applicable.

b) Partial penetration weldConnections where the static stress level is high. Accepta-ble also for dynamically stressed connections, providedthe equivalent stress is acceptable, see 312.

c) Fillet weldConnections where stresses in the weld are mainly shear,or direct stresses are moderate and mainly static, or dy-namic stresses in the abutting plate are small.

205 Double continuous welds are required in the followingconnections, irrespective of the stress level:

— oil-tight and watertight connections— connections at supports and ends of girders, stiffeners and

pillars— connections in foundations and supporting structures for

machinery— connections in rudders, except where access difficulties

necessitate slot welds.

Figure 2 Tee and cross joints

206 Intermittent fillet welds may be used in the connectionof girder and stiffener webs to plate and girder flange plate,respectively, where the connection is moderately stressed.With reference to Figure 3, the various types of intermittentwelds are as follows:

— chain weld— staggered weld— scallop weld (closed).

207 Where intermittent welds are accepted, scallop weldsshall be used in tanks for water ballast or fresh water. Chainand staggered welds may be used in dry spaces and tanksarranged for fuel oil only.

Table G5 Long slotted holesMaximum size

mmBolt diameter d (maximum)

mm(d + 1) by 2.5 d 12 and 14(d + 2) by 2.5 d 16 to 24(d + 3) by 2.5 d 27 and larger

DET NORSKE VERITAS


Figure 3 Intermittent welds

208 Slot welds, see Figure 4, may be used for connection ofplating to internal webs, where access for welding is not prac-ticable, e.g. rudders. The length of slots and distance betweenslots shall be considered in view of the required size of weld-ing.

Figure 4 Slot welds

209 Lap joints as indicated in Figure 5 may be used in endconnections of stiffeners. Lap joints should be avoided in con-nections with dynamic stresses.

Figure 5 Lap joint

H 300 Weld size301 The material factors γMw for welded connections are

given in Table H1.

302 If the yield stress of the weld deposit is higher than thatof the base metal, the size of ordinary fillet weld connectionsmay be reduced as indicated in 304. The yield stress of the weld deposit is in no case to be less thangiven in DNV-OS-C401.303 Welding consumables used for welding of normal steeland some high strength steels are assumed to give weld depos-its with characteristic yield stress σfw as indicated in Table H2.If welding consumables with deposits of lower yield stressthan specified in Table H2 are used, the applied yield strengthshall be clearly informed on drawings and in design reports.304 The size of some weld connections may be reduced:

— corresponding to the strength of the weld metal, fw:

or— corresponding to the strength ratio value fr, base metal to

weld metal:

Ordinary values for fw and fr for normal strength and high-strength steels are given in Table H2. When deep penetratingwelding processes are applied, the required throat thicknessesmay be reduced by 15% provided that sufficient weld penetra-tion is demonstrated.305 Conversions between NV grades as used in Table H2and steel grades used in the EN10025-2 standard are given inSec.6.306 Where the connection of girder and stiffener webs andplate panel or girder flange plate, respectively, are mainly shearstressed, fillet welds as specified in 307 to 309 should be adopted.

Table H1 Material factors γMw for welded connections Limit state Material factor

ULS 1.25

minimum 0.75

fy = characteristic yield stress of base material, abut-ting plate (N/mm2)

σfw = characteristic yield stress of weld deposit (N/mm2)

75.0

235 ⎟⎟⎠

⎞⎜⎜⎝

⎛= fw

wfσ

75.0

⎟⎟⎠

⎞⎜⎜⎝

⎛=

fw

yw

ff

σ

Table H2 Strength ratios, fw and frBase metal Weld deposit Strength ratios

Strength group DesignationNV grade

Yield stressσfw

(N/mm2)

Weld metal Base metal/weld metal

Normal strength steels NV NS 355 1.36 0.75

High strength steelsNV27 NV32 NV36 NV40

375 375 375 390

1.42 1.42 1.42 1.46

0.75 0.88 0.96 1.00

75.0

235 ⎟⎟⎠

⎞⎜⎜⎝

⎛= fw

wfσ

75.075.0

≥⎟⎟⎠

⎞⎜⎜⎝

⎛=

fw

yw

ff

σ

DET NORSKE VERITAS


307 Unless otherwise established, the throat thickness ofdouble continuous fillet welds should not be less than:

308 The throat thickness of intermittent welds may be asrequired in 307 for double continuous welds provided thewelded length is not less than:

— 50% of total length for connections in tanks— 35% of total length for connections elsewhere.

Double continuous welds shall be adopted at stiffener endswhen necessary due to bracketed end connections.309 For intermittent welds, the throat thickness is not toexceed:

— for chain welds and scallop welds:

tw = 0.6 frt0 (mm)t0 = net thickness abutting plate:

— for staggered welds:

tw = 0.75 frt0 (mm)If the calculated throat thickness exceeds that given in one ofthe equations above, the considered weld length shall beincreased correspondingly.310 In structural parts where dynamic stresses or high statictensile stresses act through an intermediate plate, see Figure 2,penetration welds or increased fillet welds shall be used.311 When the abutting plate carries dynamic stresses, theconnection shall fulfil the requirements with respect to fatigue,see J.312 When the abutting plate carries tensile stresses higherthan 120 N/mm2, the throat thickness of a double continuousweld is not to be less than:

minimum 3 mm.

313 Stiffeners may be connected to the web plate of girdersin the following ways:

— welded directly to the web plate on one or both sides of thestiffener

— connected by single- or double-sided lugs— with stiffener or bracket welded on top of frame— a combination of the ways listed above.

In locations where large shear forces are transferred from thestiffener to the girder web plate, a double-sided connection or

stiffening should be required. A double-sided connection maybe taken into account when calculating the effective web area.314 Various standard types of connections between stiffen-ers and girders are shown in Figure 6.

Figure 6 Connections of stiffeners

315 Connection lugs should have a thickness not less than75% of the web plate thickness.316 The total connection area (parent material) at supports ofstiffeners should not to be less than:

tw = 0.43 fr t0 (mm), minimum 3 mmfr = strength ratio as defined in 304 t0 = net thickness (mm) of abutting plate.

For stiffeners and for girders within 60% of the middle of span, t0 should not be taken greater than 11 mm, however, in no case less than 0.5 times the net thickness of the web.

fw = strength ratio as defined in 304 σd = calculated maximum design tensile stress in

abutting plate (N/mm2)r = root face (mm), see Figure 2bt0 = net thickness (mm) of abutting plate.

)mm(25.0320

2.036.10

0t

tr

ft d

ww ⎥

⎦

⎤⎢⎣

⎡⎟⎠

⎞⎜⎝

⎛ −+=σ

c = detail shape factor as given in Table H3fyd = minimum yield design stress (N/mm2)l = span of stiffener (m)s = distance between stiffeners (m)pd = design pressure (kN/m2).

a0 3 cfyd-------103 l 0.5s–( )spd mm2( )=

DET NORSKE VERITAS


The total weld area a is not to be less than:

The throat thickness is not to exceed the maximum for scallopwelds given in 309. 317 The weld connection between stiffener end and bracketis in principle to be designed such that the design shear stressesof the connection correspond to the design resistance.318 The weld area of brackets to stiffeners which are carry-ing longitudinal stresses or which are taking part in thestrength of heavy girders etc., is not to be less than the sectionalarea of the longitudinal. 319 Brackets shall be connected to bulkhead by a doublecontinuous weld, for heavily stressed connections by a partlyor full penetration weld.320 The weld connection area of bracket to adjoining girdersor other structural parts shall be based on the calculated normaland shear stresses. Double continuous welding shall be used.Where large tensile stresses are expected, design according to310, 311, and 312 shall be applied.321 The end connections of simple girders shall satisfy therequirements for section modulus given for the girder in question.Where the shear design stresses in web plate exceed 90 N/mm2, double continuous boundary fillet welds should havethroat thickness not less than:

322 The distribution of forces in a welded connection may becalculated directly based on an assumption of either elastic orplastic behaviour.323 Residual stresses and stresses not participating in thetransfer of load need not be included when checking the resist-ance of a weld. This applies specifically to the normal stressparallel to the axis of a weld.324 Welded connections shall be designed to have adequatedeformation capacity.325 In joints where plastic hinges may form, the welds shallbe designed to provide at least the same design resistance asthe weakest of the connected parts.326 In other joints where deformation capacity for joint rota-tion is required due to the possibility of excessive straining, thewelds require sufficient strength not to rupture before generalyielding in the adjacent parent material.

Guidance note:In general this will be satisfied if the design resistance of the weldis not less than 80% of the design resistance of the weakest of theconnected parts.


327 The design resistance of fillet welds is adequate if, atevery point in its length, the resultant of all the forces per unitlength transmitted by the weld does not exceed its designresistance.328 The design resistance of the fillet weld will be sufficientif both the following conditions are satisfied:

and

Figure 7 Stresses in fillet weld

Table H3 Detail shape factor c

Type of connec-tion

(see Figure 6)

IWeb-to-web con-

nection only

IIStiffener or bracket on top of

stiffenerSingle-sided Double-sided

abc

1.000.900.80

1.251.151.00

1.000.900.80

fr = strength ratio as defined in 304 a0 = connection area (mm2) as given in 316.

τd = design shear stress in web plate (N/mm2)fw = strength ratio for weld as defined in 304t0 = net thickness (mm) of web plate.

a fra0 mm2( )=

(mm) 260 0tf

tw

dw

τ=

σ⊥d = normal design stress perpendicular to the throat (including load factors)

τ⊥d = shear design stress (in plane of the throat) perpen-dicular to the axis of the weld

τ ||d = shear design stress (in plane of the throat) parallel to the axis of the weld, see Figure 7

fu = nominal lowest ultimate tensile strength of the weaker part joined

βw = appropriate correlation factor, see Table H4 γMw = material factor for welds

Table H4 The correlation factor βw

Steel gradeLowest ultimate tensile

strengthfu

Correlation factorβw

NV NS 400 0.83NV 27 400 0.83NV 32 440 0.86NV 36 490 0.89NV 40 510 0.9

NV 420 530 1.0NV 460 570 1.0

Mww

udd

fγβ

ττσ ≤++ ⊥⊥ )(3 22||d

2

Mw

ud

fγ

σ ≤⊥

DET NORSKE VERITAS


I. Serviceability Limit States

I 100 General101 Serviceability limit states for offshore steel structuresare associated with:

— deflections which may prevent the intended operation ofequipment

— deflections which may be detrimental to finishes or non-structural elements

— vibrations which may cause discomfort to personnel— deformations and deflections which may spoil the aes-

thetic appearance of the structure.

I 200 Deflection criteria201 For calculations in the serviceability limit states γM = 1.0.202 Limiting values for vertical deflections should be givenin the design brief. In lieu of such deflection criteria limitingvalues given in Table I1 may be used.

203 The maximum vertical deflection is:

Figure 8 Definitions of vertical deflections

204 Shear lag effects need to be considered for beams withwide flanges.

I 300 Out-of-plane deflection of local plates301 Checks of serviceability limit states for slender platesrelated to out-of-plane deflection may be omitted if the small-est span of the plate is less than 150 times the plate thickness.

J. Fatigue Limit States

J 100 Fatigue limit state101 The aim of fatigue design is to ensure that the structurehas sufficient resistance against fatigue failure, i.e. that it hasan adequate fatigue life. Prediction of fatigue lives is used infatigue design to fulfil this aim. Prediction of fatigue lives canalso form the basis for definition of efficient inspection pro-grams, both during manufacturing and during the operationallife of the structure. 102 The resistance against fatigue is normally given in termsof an S-N curve. The S-N curve gives the number of cycles tofailure N versus the stress range S. The S-N curve is usuallybased on fatigue tests in the laboratory. For interpretation of S-N curves from fatigue tests, the fatigue failure is defined tohave occurred when a fatigue crack has grown through thethickness of the structure or structural component. 103 The characteristic S-N curve shall in general be taken asthe curve that corresponds to the 2.3% quantile of N for givenS, i.e. the S-N curve that provides 97.7% probability of sur-vival. 104 The design fatigue life for structural components shouldbe based on the specified service life of the structure. If a serv-ice life is not specified, 20 years should be used.105 To ensure that the structure will fulfil the intended func-tion, a fatigue assessment shall be carried out for each individ-ual member, which is subjected to fatigue loading. Whereappropriate, the fatigue assessment shall be supported by adetailed fatigue analysis.

Guidance note:Any element or member of the structure, every welded joint orattachment or other form of stress concentration is potentially asource of fatigue cracking and should be considered individually.


J 200 Characteristic S-N curves201 For structural steel, the characteristic S-N curve can betaken as

in which

N = fatigue life, i.e. number of stress cycles to failure atstress range Δσ

Δσ = stress range in MPam = negative slope of S-N curve on logN-logS plotloga = intercept of logN axistref = reference thickness, tref = 32 mm for tubular joints,

tref= 25 mm for welded connections other than tubularjoints, such as girth welds

t = thickness through which the potential fatigue crackwill grow; t = tref shall be used in expression when t < tref

k = thickness exponent, also known as scale exponent.

Table I1 Limiting values for vertical deflectionsCondition Limit for δmax Limit for δ2

Deck beams

Deck beams supporting plaster or other brittle finish or non- flexible partitions

L is the span of the beam. For cantilever beams L is twice the pro-jecting length of the cantilever.

δmax = the sagging in the final state relative to the straight line joining the supports

δ 0 = the pre-camberδ 1 = the variation of the deflection of the beam due to

the permanent loads immediately after loadingδ 2 = the variation of the deflection of the beam due to

the variable loading plus any time dependent deformations due to the permanent load.

L200--------- L

300---------

L250--------- L

350---------

021max δδδδ −+=

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛

⎟⎟⎠

⎞⎜⎜⎝

⎛Δ−=

k

refttmaN σ101010 logloglog

DET NORSKE VERITAS


Guidance note:In general, the classification of structural details and their corre-sponding S-N curves in air, in seawater with adequate corrosionprotection and in free corrosion conditions, can be taken fromDNV-RP-C203 “Fatigue Strength Analyses of Offshore SteelStructures”. The S-N curves for the most frequently used structural details insteel support structures for offshore wind turbines are given inTable J1.Curves specified for material in air are valid for details, which arelocated above the splash zone. The “in air” curves may also beutilised for the internal parts of air-filled members below waterand for pile driving fatigue analysis.The basis for the use of the S-N curves in Table J1 is that a highfabrication quality of the details is present, i.e. welding and NDTshall be in accordance with Inspection Category I and StructuralCategory ‘Special’ according to DNV-OS-C401 Chapter 2 Sec.3Tables C3, C4 and C5. For structural details in the tower, therequirement of NDT inspections in accordance with InspectionCategory I is waived, cf. 6A309.For S-N curves for plated structures, I-girders and other struc-tural details than those covered by Table J1, reference is made toDNV-RP-C203.


202 Calculation of the fatigue life may be based on fracturemechanics design analysis separately or as supplement to an S-N fatigue calculation, see DNV-RP-C203. Appendix E pro-vides a method for calculation of the fatigue life for tubularconnections (tubular joints and tubular girth welds) based onfracture mechanics. An alternative method for fracturemechanics calculations can be found in BS 7910.

J 300 Characteristic stress range distribution301 A characteristic long-term stress range distribution shallbe established for the structure or structural component. 302 All significant stress ranges, which contribute to fatiguedamage in the structure, shall be considered.

Guidance note:Stress ranges caused by wave loading shall be established fromsite-specific wave statistics. Discrete wave statistics can beapplied for this purpose and usually imply that the number ofwaves are specified from eight different compass directions inone-meter wave height intervals. For wave heights between 0and 1 m, a finer discretisation with 0.2 m wave height intervals,is recommended in order to enhance the accuracy of the fatiguedamage predictions for the loading arising from waves heights inthis range.The choice of wave theory to be applied for calculation of wavekinematics is to be made according to Sec.3. The wave theorydepends much on the water depth. For water depths less thanapproximately 15 m, higher order stream function theory is to beapplied. For water depths in excess of approximately 30 m,Stokes 5th order theory is to be applied.Stress ranges caused by wind loading shall be established fromsite-specific wind statistics. Stress ranges caused by wind load-ing shall be established under due consideration of the actualalignment of the rotor axis of the wind turbine relative to thedirection of the wind. Stress ranges arising during fault condi-tions where a yaw error is present need to be considered.Stress ranges caused by the operation and control of the wind tur-bine shall be included. They include stress ranges owing to drivetrain mechanical braking and transient loads caused by rotorstopping and starting, generator connection and disconnection,and yawing loads.


303 Whenever appropriate, all stress ranges of the long-termstress range distribution shall be multiplied by a stress concen-tration factor (SCF). The SCF depends on the structural geom-etry. SCFs can be calculated from parametric equations or byfinite element analysis.

Guidance note:In wind farms, where the same joint or structural detail isrepeated many times in many identical support structures,requirements to cost-effectiveness makes it particularly impor-

Table J1 S-N curves for most frequently used structural details

Structural detailEnvironment

In air In seawater with corrosion protection Free corrosionloga m Range of validity k loga m Range of validity k loga m k

Weld in tubular joint12.164 3 N < 107 0.25 11.764 3 N < 106 0.25 11.687 3 0.2515.606 5 N > 107 0.25 15.606 5 N > 106 0.25

Butt weld and tubular girth weld, weld toe (1)

12.164 3 N < 107 0.20 11.764 3 N < 106 0.20 11.687 3 0.2015.606 5 N > 107 0.20 15.606 5 N > 106 0.20

Butt weld and tubular girth weld, weld root (1) (2)

11.855 3 N < 107 0.25 11.455 3 N < 106 0.25 11.378 3 0.2515.091 5 N > 107 0.25 15.091 5 N > 106 0.25

Non-load carrying welded attachments of length L in main stress direction

12.010 3 N < 107L < 50 mm

0.20 11.610 3 N < 106L < 50 mm

0.20 11.533 3 0.20

15.350 5 N > 107L < 50 mm

0.20 15.350 5 N > 106L < 50 mm

0.20

11.855 3 N < 10750 mm < L < 120 mm

0.25 11.455 3 N < 10650 mm < L < 120 mm

0.25 11.378 3 0.25

15.091 5 N > 10750 mm < L < 120 mm

0.25 15.091 5 N > 10650 mm < L < 120 mm

0.25

11.699 3 N < 107120 mm < L < 300 mm

0.25 11.299 3 N < 106120 mm < L < 300 mm

0.25 11.222 3 0.25

14.832 5 N > 107120 mm < L < 300 mm

0.25 14.832 5 N > 106120 mm < L < 300 mm

0.25

11.546 3 N < 107L > 300 mm

0.25 11.146 3 N < 106L > 300 mm

0.25 11.068 3 0.25

14.576 5 N > 107L > 300 mm

0.25 14.576 5 N > 106L > 300 mm

0.25

1) For girth welds welded from both sides, the S-N curves for the weld toe apply at both sides. For girth welds welded from one side only, the S-N curves for the weld toe position apply to the side from which the weld has been welded up, and the S-N curves for the weld root apply to the other side.

2) Transverse butt weld on a temporary or permanent backing strip without fillet welds.

DET NORSKE VERITAS


tant to assess the SCFs accurately, and assessment by finite ele-ment analysis is recommended.When parametric equations are used to calculate SCFs for tubu-lar joints, the Efthymiou equations should be applied for T, Y,DT and X joints, as well as for K and KT joints. For details, seeAppendix A.When finite element methods based on conventional rigid-jointframe models of beam elements are used to calculate SCFs fortubular joints, it is important to include local joint flexibilities.Such local joint flexibilities exist, but are not reflected in the rigidbeam element connections of such frame models. For inclusionof local joint flexibilities, Buitrago’s parametric formulae shallbe used. Details are given in Appendix B.For multi-planar tubular joints for which the multi-planar effectsare not negligible, the SCFs may either be determined by adetailed FEM analysis of each joint or by selecting the largestSCF for each brace among the values resulting from consideringthe joint to be a Y, X and K joint.When conical stubs are used, the SCF may be determined byusing the cone cross section at the point where the centre line ofthe cone intersects the outer surface of the chord. For gappedjoints with conical stubs, the true gaps shall be applied.A minimum SCF equal to 1.5 should be adopted for tubular jointsif no other documentation is available.In tube-to-tube girth welds, geometrical stress increases areinduced by local bending moments in the tube wall, created bycentre line misalignment from tapering and fabrication toler-ances and by differences in hoop stiffness for tubes of differentthickness. Details for calculation of SCFs for tube-to-tube girthwelds are given in Appendix C. It is recommended that as strictfabrication tolerances as possible are required for tube-to-tubewelds as a means for minimising the stress concentration factor.


304 For fatigue analysis of regions in base material not sig-nificantly affected by residual stresses due to welding, thestress ranges may be reduced prior to the fatigue analysisdepending on whether the mean stress is a tensile stress or acompressive stress.

Guidance note:The reduction is meant to account for effects of partial or fullfatigue crack closure when the material is in compression. Anexample of application is cut-outs in the base material. The meanstress σm is the static notch stress including stress concentrationfactors. Let Δσ denote the stress range including stress concen-tration factors. Prior to execution of the fatigue analysis, in whichthe long-term stress range distribution is applied together withthe S-N curve for prediction of fatigue damage, the stress rangesmay be multiplied by a reduction factor fm which is in generalobtained from Figure 9:

Figure 9 Reduction factor fm

This implies in particular that fm is 1.0 when the material is intension during the entire stress cycle, 0.6 when it is in compres-sion during the entire stress cycle, and 0.8 when it is subject tozero-mean stress.


305 For fatigue analysis of regions in welded structuraldetails where the residual stresses due to welding are relativelylow or where the residual stresses due to welding becomereduced over time due to relaxation from application of high

tensile peak loads, the stress ranges may be reduced prior to thefatigue analysis depending on whether the mean stress is a ten-sile stress or a compressive stress.

Guidance note:An example of application of the reduction is welded structuraldetails in plate structures, such as longitudinal stiffeners weldedonto the pile wall. The mean stress σm is the static notch stressincluding stress concentration factors. Let Δσ denote the stressrange including stress concentration factors. Prior to execution ofthe fatigue analysis, in which the long-term stress range distribu-tion is applied together with the S-N curve for prediction offatigue damage, the stress ranges may be multiplied by a reduc-tion factor fm which is in general obtained from Figure 10:

Figure 10 Reduction factor fm

This implies in particular that fm is 1.0 when the material is intension during the entire stress cycle, 0.7 when it is in compres-sion during the entire stress cycle, and 0.85 when it is subject tozero-mean stress.It is emphasised that the reduction factor fm as implied by the fig-ure does not apply to tubular joints and large-scale tubular girthwelds owing to the presence of high stress concentration factorsand high, long-range residual stresses due to external constraints(which are not easily relaxed due to loading) in these connec-tions.


306 Dynamic effects, including dynamic amplification, shallbe duly accounted for when establishing the long-term stressrange distribution.

Guidance note:When the natural period of the wind turbine, support structureand foundation is less than or equal to 2.5 sec, a dynamic ampli-fication factor DAF may be applied to the wave load on the struc-ture, when the wind turbine, support structure and foundation aremodelled as a single-degree-of-freedom system

in which

ξ = damping ratio relative to critical dampingΩ = ratio between applied frequency and natural frequencyWhen the natural period of the wind turbine, support structureand foundation is greater than 2.5 sec, a time domain analysisshall be carried out to determine the dynamic amplification fac-tor.The damping ratio for jacket type support structures can gener-ally be chosen as 1% relative to critical damping. The vibrationmodes relevant for determination of dynamic amplification fac-tors are typically the global sway modes, which can be excited bywave loading.


307 The stress ranges in the stress range distribution must becompatible with the stress ranges of the S-N curve that the dis-tribution is to be used with for fatigue damage predictions. Atwelds, where stress singularities are present and extrapolationneeds to be applied to solve for the stress ranges, this implies

222 )2()1(

1

Ω+Ω−=

ξDAF

DET NORSKE VERITAS


that the same extrapolation procedure must be applied to estab-lish the stress ranges of the stress range distribution as the onethat was used to establish the stress range values of the S-Ncurve for the weld.

Guidance note:S-N curves are based on fatigue tests of representative steel spec-imens. During testing, stresses are measured by means of straingauges. Stresses in the notch zone at the weld root and the weldtoe cannot be measured directly, because strain gauges cannot befitted in this area due to the presence of the weld. In additioncomes that a stress singularity will be present in this area, i.e.stresses will approach infinity.The stress which is recorded in standard fatigue tests is the so-called hot spot stress which is an imaginary reference stress. Thehot spot stress at the weld root and the weld toe is established byextrapolation from stresses measured outside the notch zone.During testing for interpretation of the S-N curve, strain gaugesare located in specific positions on the test specimens, and the hotspot stress is established by processing the measurements. Toensure an unambiguous stress reference for welded structuraldetails, the strain gauge positions to be used for application of thestrain gauges and for subsequent stress extrapolation are pre-scribed for each type of structural detail. To fulfil the compatibility requirement, the stresses in the weldsfrom the applied loading must be established as hot spot stressesfor the weld in question, i.e. the stresses in the welds must beestablished by extrapolation from stresses in the extrapolationpoints which are prescribed for the actual structural detail. Thuswhen a finite element analysis is used to establish the stresses inthe welds from the applied loading, the stresses in the welds areto be found by extrapolation from the stresses that are calculatedby the analysis in the prescribed extrapolation points. Appendix D provides definitions of the stress extrapolationpoints to be used for various structural details.Reference is made to DNV-RP-C203 for more details.


J 400 Characteristic and design cumulative damage401 Predictions of fatigue life may be based on calculationsof cumulative fatigue damage under the assumption of linearlycumulative damage.

Guidance note:There are two approaches to the calculation of the design cumu-lative fatigue damage. The two approaches are denoted Method(1) and Method (2).Method (1):The characteristic cumulative damage DC is calculated byMiner’s sum as

in which

DC = characteristic cumulative damage

I = number of stress range blocks in a sufficiently fine,chosen discretisation of the stress range axis

nC,i = number of stress cycles in the ith stress block, inter-preted from the characteristic long-term distributionof stress ranges

NC,i = number of cycles to failure at stress range of the ithstress block, interpreted from the characteristic S-Ncurve

The design cumulative damage DD is then obtained by multiply-ing the characteristic cumulative damage DC by the designfatigue factor DFFDD = DFF⋅DCMethod (2):The design cumulative fatigue damage DD is calculated byMiner’s sum as

in which

DD = design cumulative fatigue damageI = number of stress range blocks in a sufficiently fine,

chosen discretisation of the stress range axisnC,i = number of stress cycles in the ith stress block, inter-

preted from the characteristic long-term distributionof stress ranges

ND,i = number of cycles to failure at the design stress rangeΔσd,i = γmΔσi of the ith stress block, interpreted fromthe characteristic S-N curve

γm = material factor for fatigue Δσi = stress range of the ith stress block in the characteristic

long-term distribution of stress ranges.---e-n-d---of---G-u-i-d-a-n-c-e---n-o-t-e---

J 500 Design fatigue factors501 The design fatigue factor DFF for use with Method (1)is a partial safety factor to be applied to the characteristiccumulative fatigue damage DC in order to obtain the designfatigue damage.

Guidance note:Because fatigue life is inversely proportional to fatigue damage,the design fatigue factor can be applied as a divisor on the char-acteristic fatigue life to obtain the design fatigue life.


502 The DFF depends on the significance of the structure orstructural component with respect to structural integrity andaccessibility for inspection and repair.503 The design fatigue factors in Table J2 are valid for struc-tures or structural components with low consequence of fail-ure. The design fatigue factors in Table J2 depend on thelocation of the structural detail, of the accessibility for inspec-tion and repair, and of the type of corrosion protection.

∑=

=I

i iC

iCC N

nD

1 ,

,

∑=

=I

i iD

iCD N

nD

1 ,

,

DET NORSKE VERITAS


J 600 Material factors for fatigue601 The material factor γm for use with Method (2) is a par-tial safety factor to be applied to all stress ranges before calcu-lating the corresponding numbers of cycles to failure that areused to obtain the design fatigue damage. 602 The material factor depends on the significance of thestructure or structural component with respect to structuralintegrity and accessibility for inspection and repair.603 The material factors in Table J3 are given as a functionof the corresponding design fatigue factor DFF from Table J2and are valid for structures or structural components, when theapplied number of load cycles during the design life is large,i.e. in excess of 107.

J 700 Design requirement701 The design criterion is

J 800 Improved fatigue performance of welded struc-tures by grinding801 The fatigue performance of welds in tubular joints canbe improved by grinding. If the critical hotspot is at the weldtoe, reduction of the local notch stresses by grinding the weldtoe to a circular profile will improve the fatigue performance,as the grinding removes defects and some of the notch stressesat the weld toe.If the grinding is performed in accordance with Figure 11, animprovement in fatigue life by a factor of 3.5 can be obtained.Further, the scale exponent, k, in the S-N curves may bereduced from 0.25 to 0.20.

Table J2 Requirements to design fatigue factors, DFF

LocationAccessibility for

inspection and repair of initial fatigue and coating damages (8)

Corrosion protection

Corrosion allowance

(10)S-N curve DFF

Atmospheric zone Yes Coating (1) No In air 1.0

Splash zone(5)

YesCoating (2) (3) Yes

(6)

Combination of curves marked “air” and “free corrosion” (9)

2.0

No 3.0

Submerged zoneYes

Cathodic protection and optional coating (2) (4)

No

In seawater

2.0No 3.0

Scour zone No Yes(7) 3.0

Below seabed No None No 3.0

Closed compartments with seawater

Yes Cathodic protection, coating near free sur-faces and above free surfaces (12)

Yes (11)2.0

No 3.0

1) Coating for structures above the splash zone shall be a high quality multilayer coating in accordance with corrosivity category C5M in ISO 12944.2) Coating for structures in the splash zone and below the splash zone shall be taken as for (1) and shall furthermore be qualified for compatibility with

cathodic protection systems. Selection and qualification of coating systems shall include consideration of all conditions relevant for necessary repair after installation.

3) Coating shall be selected with due consideration of loads from impacts from service vessels and floating ice.4) Below the splash zone coating is considered optional. Coating can provide a reliable corrosion protection and can be designed such as to reduce marine

growth. However, coating can be damaged during inspection and maintenance sessions where marine growth is removed. 5) Splash zone definition according to Sec.11 B200.6) The corrosion allowance in the splash zone shall be selected in accordance with the corrosion rate for the structural steel in seawater and in accordance

with the planned inspection and repair strategy. In the North Sea the corrosion allowance for coated primary steel structures without planned coating repair in a 20-year design life is 6 mm. A corrosion allowance of minimum 2 mm is recommended for replaceable secondary structures.

7) In the scour zone the cathodic protection might not be fully effective and anaerobic corrosion might occur. For typical North Sea conditions it is recom-mended to design with a corrosion allowance of 3 mm in the scour zone.

8) If the designer considers the steel surface accessible for inspection and repair of initial fatigue damage and coating this must be documented through qual-ified procedures for these activities. See also Sec.11 and Sec.13.

9) The basic S-N curve for unprotected steel in the splash zone is the curve marked “free corrosion”. The basic S-N curve for coated steel is the curve marked “in air”. It is acceptable to carry out fatigue life calculations in the splash zone based on accumulated damage for steel considering the probable coating conditions throughout the design life – intact, damaged and repaired. The coating conditions shall refer to an inspection and repair plan as specified in Sec.13. For coating systems with a specified coating life of 15 years and without any qualified coating repair procedure, it is acceptable to use the “in seawater” S-N curve as a representative fatigue curve throughout a service life of 20 years.

10) The corrosion allowance shall be considered in all limit state analyses. Fatigue calculations can be based on a steel wall thickness equal to the thickness that corresponds to half the allowance over the full service life.

11) The corrosion allowance for closed compartments with seawater shall be established from experience data on a case to case basis.12) Biocides and scavengers can reduce corrosion in closed compartments.

Table J3 Material factors, γm, to be applied to all stress ranges for calculation of design fatigue life

DFF γm1.0 1.02.0 1.153.0 1.25

1≤DD

DET NORSKE VERITAS


Figure 11 Weld toe grinding

802 The following conditions shall be fulfilled when weldsin tubular joints are ground:

— a ball type rotary burr shall be used for grinding— final grid marks should be kept small and always be nor-

mal to the weld toe— the diameter of the ball shall be between 8 and 10 mm. If

the brace thickness is less than 16 mm, the diameter of thegrinder may be reduced to 6 mm.

— the edges between the ground profile and the brace/chordshall be rounded, i.e. no sharp edges are allowed

— if the weld toe grinding shall not be performed on the com-plete circumference of the joint, a smooth transitionbetween the ground profile and the non-ground weld shall

be ensured— the ground surface shall be proven free of defects by an

approved NDT method, e.g. MPI— the depth of grinding shall be 0.5 mm below any visible

undercut. However, the grinding depth is not to exceed 2mm or 5% of wall thickness whichever is less.

If weld toe grinding is performed on “old” joints according tothe above specification, these joints can be considered as ‘new-born’ when their fatigue lives are to be predicted.803 The fatigue performance of girth welds can be improvedby grinding. Grinding of girth welds will increase the fatiguelife of the welded connection if performed according to theconditions specified in Figure 12.

Figure 12 Grinding of girth welds. A local grinding by small-scale rotary burr (left) may not be performed. The figure only shows weld profilegrinding (right) of the weld at the one side, but a grinding of the weld root may be performed following the same principle

If the grinding is performed as shown to the right in Figure 12and the below conditions are fulfilled, an improved S-N curvemay be applied for the weld toe. If the weld root is groundaccording to the same principles, an improved S-N curve mayalso be applied for the weld root. The SCF due to fabricationtolerances and geometry such as tapering shall still be applied,see also Appendix C.

— Final grid marks should be kept small and always be nor-mal to the weld toe.

— The largest radius possible considering the actual geome-try shall be selected.

— The edges between the ground profile and the brace/chordshall be rounded, i.e. no sharp edges are allowed.

— If the weld toe grinding shall not be performed on the com-plete circumference of the joint, a smooth transitionbetween the ground profile and the non-ground weld shallbe ensured.

— The ground surface shall be proven free of defects by anapproved NDT method, e.g. MPI.

— The depth of grinding shall be 0.5 mm below any visibleundercut. However, the grinding depth is not to exceed 2

mm or 5% of wall thickness whichever is less.

Guidance note:The following improved S-N curves can be applied for girthwelds if grinding is carried out according to the above specifica-tions:


Rotary burr

Edge to be rounded (Typ.)

Smooth transition at start/stop of grinding

Grinding as for tubular joints

Large radius grinding removing weld caps and weld toe undercut

For ground girth welds in air: loga = 12.592 and m = 3 for N < 107, k = 0.15

(Curve ‘C’/Curve 125)

loga = 16.320 and m = 5 for N > 107, k = 0.15

For girth welds in seawater with adequate cathodic protection:loga = 12.192 and m = 3 for N < 106, k = 0.15

(Curve ‘C’/Curve 125)

loga = 16.320 and m = 5 for N > 106, k = 0.15

DET NORSKE VERITAS


SECTION 8DETAILED DESIGN OF OFFSHORE CONCRETE STRUCTURES

A. GeneralA 100 Introduction101 For detailed design of offshore wind turbine concretestructures, DNV-OS-C502, “Offshore Concrete Structures”shall apply together with the provisions of this section. Alter-natively, other standards can be used as specified in Sec.1A400. It is the responsibility of the designer to document thatthe requirements in Sec.1 A400 are met. 102 The loads that govern the design of an offshore wind tur-bine concrete structure are specified in Sec.4 and Sec.5. SLSloads for offshore wind turbine concrete structures are definedin this section. Details regarding the process of determining theload effects within the concrete structure can be found inDNV-OS-C502. 103 Sec.8 in general provides requirements and guidancewhich are supplementary to the provisions of DNV-OS-C502.Hence, Sec.8 shall be considered application text for DNV-OS-C502 with respect to offshore wind turbine concrete structures.For all design purposes, the user should always refer to the com-plete description in DNV-OS-C502 together with this section.104 Sec.8 in particular provides requirements and guidancefor how to use EN standards as a supplement to DNV standardsfor design of offshore concrete structures. Such use of ENstandards as a supplement to DNV standards shall be carriedout according to the requirements in Sec.1 A400.

A 200 Material201 The requirements to materials given in DNV-OS-C502Sec.4 and in Sec.6 of this standard shall apply for structuresdesigned in accordance with this section.

A 300 Composite structures301 For design of composite structures such as pile-to-sleeveconnections and similar connections, the requirements given inDNV-OS-C502 Sec.5 A500 shall be supplemented by therequirements given in Sec.9.

B. Design Principles

B 100 Design material strength101 In design by calculation according to DNV-OS-C502together with this standard, the design material strength shallbe taken as a normalized value of the in-situ strength dividedby a material factor γm (ref. DNV-OS-C502 Sec.6 B600 andSec.8 B103 in this standard).

Guidance note:It is important to note that the partial safety factor γm for materialstrength of concrete shall be applied as a divisor on the normal-ized compressive strength fcn and not as a divisor on the charac-teristic compressive strength defined as the 5% quantile in theprobability distribution of the compressive strength of concrete.The normalized compressive strength and the characteristic com-pressive strength are not necessarily the same.


102 For wind turbine structures, Young’s Modulus for con-crete shall be taken equal to the characteristic value Eck, bothfor the serviceability limit state and for the fatigue limit state(ref. DNV-OS-C502 Sec.6 B605).103 The material factors, γm, for concrete and reinforcement foroffshore wind turbine concrete structures are given in Table B1.

Guidance note:It is noted that the requirements to the material factor for ULSdesign as specified in Table B1 are somewhat lower than the corre-sponding requirements in DNV-OS-C502. This difference merelyreflects that DNV-OS-C502 is meant for design to high safety class(manned structures with large consequence of failure) whereasDNV-OS-J101 aims at design to normal safety class (unmannedstructures, structures with small consequences of failure).


Table B1 Material factors for concrete and reinforcementLimit State ULS FLS SLSReinforced Concrete Concrete, γc 1.21 (1.35) 2 1.11 (1.20) 2 1.0

Reinforcement, γs 1.11 (1.2) 2 1.001 (1.10) 2 1.0

Plain Concrete γc 1.451 (1.7) 2 1.251 (1.50) 2 1.0

1) When the design is to be based on dimensional data that include specified tolerances at their most unfavourable limits, structural imperfections, placement tolerances as to positioning of reinforcement, then these material factors can be used. When these factors are used, then any geometric deviations from the “approved for construction” drawings must be evaluated and considered in relation to the tolerances used in the design calculations.

2) Design with these material factors allows for tolerances in accordance with DNV-OS-C502 Sec.6 C400 or, alternatively, tolerances for cross sectional dimensions and placing of reinforcements that do not reduce the calculated resistance by more than 10 percent. If the specified tolerances are in excess of those given in DNV-OS-C502 Sec.6 C400 or the specified tolerances lead to greater reductions in the calculated resistance than 10 percent, then the excess tolerance or the reduction in excess of 10 percent is to be accounted for in the resistance calculations. Alternatively, the material factors may be taken according to those given under 1).

DET NORSKE VERITAS


C. Basis for Design by CalculationC 100 Concrete grades and in-situ strength of concrete101 In DNV-OS-C502 Sec.6 C100 normal weight concretehas grades identified by the symbol C and lightweight aggre-gate concrete grades are identified by the symbol LC. Thegrades are defined in DNV-OS-C502 Sec.6 Table C1 as a func-tion of the Characteristic Compressive Cylinder strength of theconcrete, fcck. However, the grade numbers are related to theCharacteristic Compressive Cube strength of the concrete, fck(100 mm cube).

Guidance note:Care shall be taken when using the notations C and LC. Otherstandard systems (e.g. EN standards) use the notation C in rela-tion to characteristic compressive cylinder strength.


D. Bending Moment and Axial Force (ULS)D 100 General101 For design according to EN 1992-1-1:2004 in the Ulti-mate Limit State the strength definition can be used from theEN standard together with the general material factors (ref. EN1992-1-1: 2004, Table 2.1N).

E. Fatigue Limit StateE 100 General 101 For fatigue design according to EN standards, the cumu-lative fatigue damage in the Fatigue Limit State shall be deter-mined according to DNV-OS-C502, which takes into accountfatigue under wet conditions.

F. Accidental Limit StateF 100 General101 According to DNV-OS-C502, structures classified inSafety Classes 2 and 3 (see DNV-OS-C502 Sec.2 A300) shallbe designed in such a way that an accidental load will not causeextensive failure. Support structures and foundations for off-shore wind turbines are in this standard defined to belong toSafety Class 2.

G. Serviceability Limit StateG 100 Durability101 When the formula in DNV-OS-C502 Sec.6 O206 for thenominal crack width (wk = wck ⋅ (c1 / c2) > 0.7 ⋅ wck) is used,the value for c2 shall be taken as given below:

c2 = actual nominal concrete cover to the outermost reinforce-ment (e.g. stirrups)

102 For offshore wind turbine concrete structures, the loadfor crack width calculations is to be taken as the maximumcharacteristic load that can be defined among the wind andwave climate combinations used for the FLS load cases. Thewind and wave climate combinations used for the FLS loadcases are specified in Sec.4 Table E1. The characteristic loadfor a particular combination of wind climate and wave climateis defined as the 90% quantile in the distribution of the maxi-mum load in a 10-minute reference period with this particularclimate combination. Based on this, the following procedure

can be used to determine the load for crack width calculations:

1) For each considered applicable combination of wind cli-mate and wave climate, at least 6 10-minutes time series ofload (or load effect) in relevant cross sections shall be cal-culated by simulation with different seeds.

2) From each of the time series for a particular cross sectionand a particular combination of wind and wave climate,the maximum load or load effect shall be interpreted.

3) For each relevant cross section and particular combinationof wind and wave climate, the mean value and the standarddeviation of the interpreted six or more maxima (one fromeach simulated time series of load or load effect) shall becalculated.

4) For each relevant cross section and particular combinationof wind and wave climate, the characteristic load can becalculated as mean value + 1.28 × standard deviation.

5) For each relevant cross section considered, the load forcrack width calculation shall be taken as the maximumcharacteristic load over all applicable combinations ofwind and wave climate considered.

Guidance note:Usually it will be sufficient to consider the production and idlingload cases i.e. Load Cases 1.2 and 6.4 according to Sec.4, Table E1.


103 In order to fulfil the requirements of DNV-OS-C502Sec.6 O213, the strain in the reinforcement shall be calculatedfor an SLS load which shall be set equal to the characteristicextreme ULS load and it shall be substantiated that this straindoes not exceed the yield strain of the reinforcement.

G 200 Crack width calculation201 Crack widths shall be calculated in accordance with themethod described in DNV-OS-C502 Sec.6 O700 and DNV-OS-C502 Appendix F.Let εsm denote the mean principal tensile strain in the rein-forcement over the crack’s influence length at the outer layerof the reinforcement. Let εcm denote the mean stress-depend-ent tensile strain in the concrete at the same layer and over thesame length as εsm.For estimation of (εsm – εcm) the following expression shall beused:

where:

σs = the stress in reinforcement at the crack calculated forthe actual load.

σsr = the stress in reinforcement at the crack calculated forthe load for which the first crack is developed. The ten-sile strength of the concrete to be used in this calcula-tion is the normalised structural tensile strength, ftn,according to DNV-OS-C502 Sec.6 Table C1.

σsr ≤ σsβs = 0.4.202 For guidance on how to calculate the free shrinkagestrain of the concrete, εcs, reference is made to NS 3473:2003,Section A9.3.2.203 For design according to EN standards the crack widthformulae in EN 1992-1-1:2004 can be used with the followingprescribed coefficient values which will yield results similar toresults according to DNV-OS-C502:

a) hcef shall be defined according to DNV-OS-C502 Appendix Fb) k2 shall be defined according to DNV-OS-C502 Appendix F

)1(Es

)(s

srs

sk

scmsm σ

σβ−=ε−ε

DET NORSKE VERITAS


c) k3 shall as a minimum be taken as 1.36d) k4 shall be taken as 0.425

Guidance note:For crack width calculation according to EN 1992-1-1:2004 withthe prescribed coefficient values, the crack width criterion can betaken according to DNV-OS-C502.


G 300 Other serviceability limit states301 Limitations of stresses in the concrete (ref. DNV-OS-C502 Sec 6 O802) are also governing for concrete wind turbinesupport structures with normal reinforcement. The SLS load tobe considered is the load defined for the crack width calcula-tion in G201.

H. Detailing of Reinforcement

H 100 Positioning101 All shear reinforcements and stirrups shall be anchoredoutside the main reinforcement (i.e. they shall encircle thereinforcement).

I. Corrosion Control and Electrical Earthing

I 100 Corrosion control101 Requirements to corrosion protection arrangement andequipments are generally given in Section 11. Special evalua-tions relevant for offshore concrete structures are given inDNV-OS-C502 Sec.6 S100-S400 and in 102.102 Concrete rebars and prestressing tendons are adequatelyprotected by the concrete itself, i.e. provided adequate cover-age and adequate type and quality of the aggregate. However,rebar portions freely exposed to seawater in case of concretedefects and embedment plates, penetration sleeves and varioussupports (e.g. appurtenances) which are freely exposed to sea-water or to the marine atmosphere will normally require corro-sion protection.

Guidance note:It is recommended always to install cathodic protection for anoffshore wind turbine concrete structure. The corrosion protec-tion may be combined with the electrical earthing system for thewind turbine structure, See Section I200.


I 200 Electrical earthing201 All metallic components in an offshore support structureincluding appurtenances shall have equipotential bonding andelectrical earthing in order to protect against potential differ-ences, stray currents and lightning. Documentation for thisshall be included in the design documentation.

Guidance note:Often the transfer resistance for the reinforcement in an offshoreconcrete structure will be low and could then be used for earth-ing. If used for earthing the reinforcement should as a minimumbe tied with metallic wire at every second crossing and the verti-cal and horizontal connection shall be supplemented by separateelectrical connections clamped to the reinforcement at a suitabledistance. Care shall be taken to ensure that the corrosion protec-tion system and the electrical earthing are not in conflict.


J. ConstructionJ 100 General101 Construction shall be performed according to DNV-OS-C502 Sec.7, if necessary together with other relevant standardsas stated in DNV-OS-C502 Sec.7 A201. 102 For structures designed according to other standards sys-tems (e.g. EN standards) the construction standards in theactual system shall be also be applied.

J 200 Inspection classes201 In general, inspection class IC2, “Normal inspection”,(see DNV-OS-C502 Sec.7 D201) applies for offshore windturbine concrete structures.202 For construction according to EN 13670-1:2000, Inspec-tion Class 2 applies for offshore wind turbine concrete struc-tures.

DET NORSKE VERITAS


SECTION 9DESIGN AND CONSTRUCTION OF GROUTED CONNECTIONS

A. IntroductionA 100 General101 The requirements in this section apply to grouted tubularconnections in steel support structures for offshore wind tur-bines.102 Grouted tubular connections are structural connections,which consist of two concentric tubular sections where theannulus between the outer and the inner tubular has been filledwith grout. Typical grouted connections used in offshore windturbine support structures consist of pile-to-sleeve or pile-to-structure grouted connections, single- or double-skin groutedtubular joints, and grout-filled tubes.

Guidance note:In steel monopile support structures, grouted connections typi-cally consist of pile-to-sleeve connections. In tripod legs, pile-to-structure grouted connections are typically used.


103 Types of grouted connections not specifically coveredby this standard shall be specially considered.104 All relevant factors which may influence the strength ofa grouted connection are to be adequately considered andaccounted for in the design.

Guidance note:The strength of grouted connections may depend on factors suchas:- grout strength and modulus of elasticity- tubular and grout annulus geometries- application of mechanical shear keys - grouted length to pile-diameter ratio- surface conditions of tubular surfaces in contact with grout- grout shrinkage or expansion- load history (mean stress level, stress ranges).


105 Grout materials are to comply with the requirementsgiven in Sec.6 B “Selection of Concrete Materials” and Sec.6C “Grout Materials and Material Testing” as relevant.106 Grouted connections in wind turbine support structuresmust be designed for the ULS and the FLS load combinationsspecified in Sec.5 for the loads specified in Sec.4.

A 200 Design principles201 Design rules for grouted connections are given for axialloading combined with torque and for bending moment com-bined with shear loading, respectively.

Guidance note:Long experience with connections subjected to axial load incombination with torque exists, and parametric formulae havebeen established for design of connections subjected to this typeof loading. For connections subjected to bending moment andshear force, no parametric design formulae have yet been estab-lished. Therefore, detailed investigations must be carried out forsuch connections.


202 For design of grouted connections, it may be conserva-tive to assume that axial load and bending moment do notinteract. When it can be demonstrated for a grouted connectionthat it will be conservative to assume that axial load and bend-ing moment do not interact, the grouted connection shall sat-isfy two separate requirements. The first requirement to satisfyis the capacity requirement specified for the combined action

of axial load and torque under the assumption of no concur-rently acting bending moment and shear force. The secondrequirement to satisfy is the capacity requirement specified forthe combined action of bending moment and shear force underthe assumption of no concurrently acting axial force andtorque.203 When shear stresses in grouted connections of piles sub-jected to axial load are calculated, due account shall be takenof the distribution of global loads between the various piles ina group or cluster of piles. Analyses of the connections are totake account of the highest calculated load with due consider-ation of the possible range of in-situ soil stiffness.204 A grouted connection can be established with or withoutshear keys as shown in Figure 1.

Guidance note:Shear keys can reduce the fatigue strength of the tubular mem-bers and of the grout due to the stress concentrations around thekeys. If shear keys are used in a grouted connection subjected tobending, they should be placed at the mid level of the connectionin order to minimise the influence on the fatigue damage,because the maximum grout stresses from bending will developat the top and the bottom of the grout member.


205 The distance between the mean seawater level (MSL)and the connection has to be considered in the early designphase since it may have great influence on the behaviour of theconnection.

Guidance note:The location of the connection relative to MSL may influence theshrinkage of the grout, the size of the bending moment in the con-nection, the fatigue performance of the connection, and thegrouting operation.


206 A grouted connection in a monopile can be constructedwith the transition piece placed either inside or outside thefoundation pile.

Guidance note:Traditionally the transition piece is located outside the founda-tion pile for connections near MWL. This is mainly to be able tomount accessories like boat landing and to paint the structurebefore load-out. These issues must be paid special attention if thetransition piece is placed inside the foundation pile.


207 The steel tubes shall be checked according to Sec.7.208 Local buckling in the steel tubes shall be considered.

B. Ultimate Limit StatesB 100 Connections subjected to axial load and torque101 The characteristic ultimate capacity of axially loadedgrouted tubular connections is defined as the mean value of thedistribution of the ultimate capacity. The design ultimatecapacity is defined as the characteristic ultimate capacitydivided by a material factor γm.102 The characteristic ultimate capacity of axially loadedgrouted tubular connections may be calculated according tothe method given in DNV Rules for Fixed Offshore Installa-tions, January 1998. The method is reproduced in the guidancenote with torque included.

DET NORSKE VERITAS


Guidance note:The shear stress to be transferred in an axially loaded connectionis:

where:

τ sa= shear stress in axially loaded connectionP = axial force from factored load actionsRp = pile outer radius (see Figure 1)L = effective grouted connection length.The shear stress to be transferred in a connection subjected totorque is:

where:

τ st = shear stress in torsionally loaded connectionMT= torque from factored load actions.For grouted connections with mill rolled surface where the millscale has been removed completely by corrosion or mechanicalmeans, the following simplified design equations may be used.The ultimate strength is the lesser of the interface shear strengthand the grout matrix strength. The interface shear strength due to friction may be taken as:

The interface shear strength due to shear keys may be taken as:

where:

τ kf= characteristic interface shear strength due to frictionτ ks= characteristic interface shear strength due to shear keysμ = grout to steel interface coefficient of friction to be taken

as 0.4 to 0.6 for corroded or grit blasted steel surfaceswith the mill scale removed.

δ = height of surface irregularities to be taken as 0.00037 Rp for rolled steel surfaces

N = number of shear keysRs = sleeve outer radiusts = wall thickness of sleevetp = wall thickness of piletg = thickness of grouth = shear key outstands = shear key spacingE = modulus of elasticity for steelEg = modulus of elasticity for groutfck = characteristic compressive cube strength of the grout. (A

conversion from characteristic compressive cylinderstrength, fcck, to cube strength, fck, can be made accord-ing to DNV-OS-C502 Sec.6 Table C1. For a cylinder strength fcck > 94 MPa, the cube strengthcan be taken as fck = fcck + 11 MPa.).fck shall be given in units of MPa.

Where more precise information is not available, Eg may be

taken as equal to 150 fck MPa.The above equations have been proven valid within the followinglimits:

The upper limit for the ratio Rp/tp can be exceeded for low utili-zation of the axial capacity of the grouted connection. The allow-able upper limit for Rp/tp must be evaluated for the actualconnection and the actual utilization.It is to be noted that when the shear key spacing, s, approaches alimit of

no further significant increase in strength may be obtained bydecreasing the shear key spacing.The capacity of the grout matrix may be taken as:

where:

There is only modest test experience of early age cycling effectssuch as will be caused by relative movement between pile andsleeve due to wave action during setting of the grout. The aboveequations are for initial estimation only. It will be necessary toverify the performance of the specimens subject to early agecycling effects with ad hoc tests. The shear stress in an axially and torsionally loaded connectionwithout shear keys shall satisfy:

where:

τ k = characteristic shear strength of the connection, min (τ kf,τ kg)

γ m = the material factor according to D200.The shear stress in an axially and torsionally loaded connectionwith shear keys perpendicular to the circumference of the tubu-

LR2P

psa ⋅⋅⋅

=π

τ

LR2M

2p

Tst

⋅⋅⋅=

πτ

⎥⎦

⎤⎢⎣

⎡⋅

⋅=

Pkf RK

E δμτ

NLs

Rt

fs

hK

E

p

pckks ⋅⋅

⎥⎥⎦

⎤

⎢⎢⎣

⎡⋅⋅

⋅⋅

⋅= 4.0

21μτ

s

s

pg

g

p

p

tR

REtE

tR

factor stiffnessK +⋅

⋅+==

τ kg = characteristic shear strength of the groutκ = early age cycling reduction factor

=

for

= 1for

Δ = early age cycling movement.

30t

R5

p

p≤≤

70tR

9s

s ≤≤

1.0sh

<

pp tRs ⋅>

pp tR ⋅

⎟⎠⎞⎜

⎝⎛ −⋅⋅= p-2L/R0.7

ckkg e1fκτ

p/R31 Δ⋅−

3tR/s pp <⋅

3tR/s pp ≥⋅

m

k22sa γ

τττ ≤+ st

DET NORSKE VERITAS


lars shall satisfy the following three requirements:

If the torque can be considered negligible (τst ≈ 0), then the shearstress from the axial load shall satisfy the following requirement:

Figure 1 Grouted pile-to-sleeve connection


103 As an alternative to calculation of the characteristic ulti-mate capacity by the prescriptive method given in 102, thecharacteristic ultimate capacity may be estimated on the basisof full-scale capacity tests carried out on representativegrouted connections. The estimate of the characteristic ulti-mate capacity shall then be obtained as the estimated meanvalue of the capacities observed from the tests. The estimateshall be obtained with a confidence of 95%.

B 200 Connections subjected to bending moment and shear loading201 For grouted connection subjected to bending momentand shear loading, the grout will mainly be exposed to radialstresses given a sufficient length-to-pile-diameter ratio.

Guidance note:The length-to-pile-diameter ratio (L/D ratio) of the connectionshould typically be in the order of L/D ≈ 1.5 to ensure that thebending moment is safely transferred by radial stresses in thegrout.

Due to load transfer by radial stresses, no shear keys in pile-to-sleeve connections are necessary to transfer the moment. For a pile-to-sleeve connection, for example for a monopile sup-port structure, relatively high loads must be transferred in thegrouted connection. Due to this, it is most likely that such con-nections require the use of high strength grout (i.e. compressivestrength in excess of 65 MPa).


202 The ultimate strength capacity of grouted connectionsshall be documented. This documentation shall include a buck-ling check. The documentation of the ultimate strength capacitymay be carried out by the use of non-linear finite element (FE)analyses. However, both the connection modelling and the solu-tion methodology should be calibrated to experimental data incases where no prior knowledge or experimental data exists.

Guidance note:The FE analyses should as a minimum represent the interactionbetween the grout and the steel. Further the FE analyses couldinclude the buckling check for the steel tubes by including non-linear geometric effects. For FE analyses guidelines and recommendations stated by themanufacturer of the FE program applied, such as in user’s man-uals, should always be followed.FE analyses of the grouted connection shall be modelled withdouble contact interfaces between the grout and the steel tubes(both sides of the grout member). FE analyses shall be carried out both with and without contactfriction on surfaces without shear keys. Friction coefficientsshould be in the order of 0.4 to 0.6, if not documented by testing.The effect of slip should be included in the contact formulationwhen the friction is present.The mesh size on the contact surfaces shall account for the non-linear stress singularities at the surface edges. The mesh sizeshall therefore ensure that contact occurs on minimum 3 ele-ments in the slip direction. Further, the element edge aspect ratioon the contact surfaces should not exceed 1:5. The grout elements should as a minimum be linear 8-node solid ele-ments with 3 translation degrees of freedom. Through the thicknessof the grout member, a minimum of two first-order elements, oralternatively one second-order element, should be applied.The constitutive model for the grout should account for the non-linear behaviour of the grout. The non-linear properties to beregarded are e.g. the difference in compressive and tensilestrength, possible cracking due to tension and effects from con-finement. In general, cracking of the grout will not be a problemfor a grouted connection. Cracking, if any, will appear verticallyto the circumference of the connection due to hoop stresses in thegrout. Since the loads on the connection will be transferredthrough radial stresses in the grout, possible cracking will notchange the load transfer significantly. Possible cracking should,however, be included in the constitutive model for the grout togive the most precise representation of the material.The steel elements should as a minimum be modelled with first-order shell elements with 5 integration points through the thick-ness.The element choice for the steel tubes and the grout shall togetherprovide a consistent deformation field.If shrinkage can be expected this should be accounted for in themodel.The input and output for the FE model must be documented thor-oughly by relevant printouts and plots. The input shall as a min-imum be documented by input file and plots showing geometry,boundary conditions and loads. The output shall as a minimumbe documented by plots showing total stresses (von Misesstresses in steel and Tresca stresses in grout) together with plotsshowing principal stresses and strains.


203 In the ultimate limit state, the stresses in the grout,expressed as Tresca stresses, shall satisfy the followingrequirement:

m

kssa γ

ττ ≤

m

kfst γ

ττ ≤

( ) ( )m

kg2st

2sa γ

τττ ≤+

m

kfkssa γ

τττ

+≤

DET NORSKE VERITAS


where:

fs = Tresca stress in the grout, fs = σ1 – σ3σ1 = maximum principal stress in the considered point in the

groutσ3 = minimum principal stress in the considered point in the

groutfcck= characteristic compressive cylinder strength of the

groutγ m = the material factor according to D200.This approach will in general be conservative.204 Alternatively, the ultimate strength capacity can be doc-umented by calculations using the design grout strength andallowing for plastic distribution of stresses.

C. Fatigue Limit StatesC 100 General101 The fatigue strength of the grout in the grouted connec-tions subjected to bending moment shall be based on codes forgrout and concrete. The documentation of the fatigue strengthcapacity of grouted connections may be carried out by meansof non-linear finite element (FE) analyses. However, both theconnection model and the solution methodology should be cal-ibrated to experimental data in cases where no prior knowledgeor experimental data exists.

Guidance note:The guidance note in B202 applies.For determination of stresses in the fatigue limit state, the peakstresses can be averaged over a length of about 100 mm.


102 A characteristic long-term stress cycle history shall beestablished for the grouted connection. All significant stresscycles, which contribute to fatigue damage in the structure dur-ing its design lifetime, shall be considered. Each stress cycle ischaracterised by its mean stress and its stress range. The designlifetime shall be based on the specified service life of the struc-ture. If a service life is not specified, 20 years should be used.

C 200 Connections subjected to axial load and torque201 The fatigue strength of axially loaded grouted connec-tions is to be based on relevant test data or experience relevantfor the actual properties of the connection. Provided a groutedconnection, exposed to environmental loading as the only formof dynamic loading, is designed to comply with the ultimatestrength requirements of B101 no further check will be

required for fatigue strength of a grouted connection only sub-jected to axial load and torque.

C 300 Connections subjected to bending moment and shear loading301 The accumulated damage, D, for the long-term stresscycle history shall be calculated using the Palmgren-Minersummation and is required not to exceed 1.0:

where:

ni = number of stress cycles for the actual combination ofmean stress and stress range (applied number of stresscycles at ith stress combination over design life)

Ni = allowable number of cycles for the actual combinationof mean stress and stress range (number of cycles tofailure at ith stress combination)

j = total number of combinations of mean stress and stressrange in a suitable discretisation of the mean stress andstress range plane.

Guidance note:When it can be demonstrated that the compressive stresses in thefatigue-critical sections of a high-strength grout member are pre-dominantly unidirectional, the calculations of the accumulateddamage can be carried out according to FIB/CEB SR90/1, Bulle-tin d’Information No. 197, “High Strength Concrete”, 1990. First calculate an intermediate value NI for the number of cyclesto failure:

The number of cycles to failure N can then be calculated accord-ing to the following S-N curve:

An endurance limit is defined for stress ranges ΔS < 0.30 – 0.375Smin. For these stress ranges, an infinitenumber of cycles to failure applies and overrides the value of Nresulting from the above expressions.Definitions:

fcck,f = design fatigue strength, fcck/γmsmax,f = maximum compressive stress in cyclesmin,f = minimum compressive stress in cycleSmax = max. relative stress, i.e. smax,f/fcck,fSmin = min. relative stress, i.e. smin,f/fcck,fΔS = stress range, Smax – Sminγm = material factor for the FLS to be taken according to

D200.---e-n-d---of---G-u-i-d-a-n-c-e---n-o-t-e---

m

ccks

ffγ

≤

∑=

≤=j

i i

i

Nn

D1

1

)1()81612(log max2minmin10 SSSNI −⋅++=

⎩⎨⎧

>≤≤

−⋅+⋅=

6log6log0

))6(log2.01(loglog

log10

10

1010

1010

I

I

II

IN

NforNNforN

N

DET NORSKE VERITAS


Figure 2 S-N curves for fatigue of high-strength grout

D. Requirements to Verification and Material Factors

D 100 Experimental verification101 If no sufficient documentation of the behaviour of agrouted connection is available, experimental verification ofthe behaviour must be carried out.

D 200 Material factors for grouted connections201 To account for uncertainties in the strength of thegrouted connections, including but not limited to natural vari-ability and uncertainties due to the offshore grouting opera-tions, the material factor γm is in general to be taken as:

202 For the FLS, the material factor γm can be expressed asa product of four factors,γm = γ1 ⋅ γ2 ⋅ γ3 ⋅ γ4where the following definitions and requirements apply

γ1 = 1.25 factor to account for the possible deviationbetween in-situ strength and laboratory testspecimen strength due to inferior in-situ com-paction and curing.

γ2 = 1.18 factor to account for the combined effect of longterm duration loading and the use of a rectangu-lar, constant stress distribution in the calcula-tions

γ3 = 1.18 factor to account for an extra safety for higher-grade concrete due to possible less ductility ofhigher-strength concrete

γ4 = 1.5 is the material factor to account for the statisticalvariation in the compressive strength.

When these four factors are applied with their required values,the “overall material factor” for concrete fatigue designγm = 1.25 · 1.18 · 1.18 · 1.5 = 2.6 is obtained as required in 201. 203 The documentation and verification activities associatedwith the grouting operation consist of:

— verification of the grout procedure and test sample require-ments

— on site verification of the grouting operation and of theresults of the operation

— verification and survey of test samples, mechanical testsand test results.

Provided that the actual in-situ concrete compressive strengthand the grouting operation are documented and further verifiedon site, and that the stress distribution in the grout is particu-larly well-controlled, a lower material factor γm than 2.6required in 201 can be accepted for design in the FLS. Thereduced requirement to γm is expressed in terms of reducedrequirements to the factors γ1 and γ2.Provided that the actual in-situ concrete compressive strength,the grouting procedure and the grouting operation are docu-mented and further verified on site, the factor γ1 can be takenas 1.0. The following two conditions shall be fulfilled before therequirement to γ1 can be reduced to 1.0:

— The certifying body shall verify the grouting operation, thecompressive testing of grout samples and the documenta-tion for the operation and the tests. The verification shallbe carried out by surveys and documentation reviews.

— The compressive testing shall be carried out on grout sam-ples which are representative of the grout in situ and whichlead to compressive strengths representative of the com-pressive strength in situ.

In order to obtain compressive strengths representative of thecompressive strength in situ, when the reduced γ1 =1.0 isapplied and it is unfeasible or should be avoided to obtaindrilled samples of the grout in situ, it suffices to carry out thecompressive tests on samples obtained from the emerging, sur-plus grout.

Guidance note:The grouting procedures should always be verified and the com-pressive strength of the grout should always be tested, even whenthe unreduced γ1 = 1.25 is applied. When the unreduced γ1 = 1.25is applied, it suffices to obtain the grout samples for compressivetesting from the grout mixer.


As the fatigue loading is short term duration loading from thewind turbine and waves, the factor γ2 can be reduced to 1.0 ifthe stress check is based on Gauss stresses including local

Smin = 0.0

Smin = 0.2

Smin = 0.4

Smin = 0.6

Smin = 0.8

Log10N = 6

0.0

0.2

0.4

0.6

0.8

1.0

0 5 10 15 20 Log10N

Smax

Limit state ULS FLSγm 3.0 2.6

DET NORSKE VERITAS


stress concentrations derived from the Finite Element Analysisperformed. However, if the fatigue limit state peak stresseshave been averaged over a length of 100 mm, as recommendedin the guidance note in C101, the factor γ2 shall remain equalto 1.18.The factors γ3 and γ4 always have to be applied with the valuesspecified in 202.

E. Grouting OperationsE 100 General101 The grouting operations of connections are to complywith relevant requirements given in DNV-OS-C502 Sec.7together with the requirements given for concrete in Sec.8 ofthis standard. 102 It is to be ensured that the grouting system has sufficientventing capacity to enable air, water and surplus grout to beevacuated from the annuli and compartments required to begrout filled at a rate exceeding the filling rate of grout.103 Injection of grout shall be carried out from the bottom ofthe annulus. Complete filling of the annulus is to be confirmedby grout overfill at top of grout connection or at top outlet hole.104 Sufficient strength of formwork or similar (e.g. an inflat-able rubber seal) must be ensured.105 To avoid casting joints in the grout member, the grout-ing should be carried out in one process.106 Sufficient material of acceptable quality is to be availa-ble at the start of a grouting operation to enable fabrication ofgrout for the biggest compartment to be grouted. A reliablesystem for replenishment of accepted material according to theconsumption rate is to be established.107 Adequate back-up equipment for the grouting processmust be available before the process is initiated.108 The temperature of all surroundings (air, water, steelstructures etc.) must be between 5°C and 35°C during thegrouting operation. 109 In general, piling operations are not to be performedafter commencement of pile-grouting operations.

E 200 Operations prior to grouting201 Prior to commencement of grouting operations, theproperties of the proposed grout mix are to be determined by

appropriate qualification tests according to a recognised codeor standard, see also Sec.6 C. 202 All steel surfaces must be clean before grouting. Beforepositioning of the tubes, the surfaces must be checked forgrease, oil, paint, marine growth etc. and cleaned if necessary.

E 300 Monitoring301 Parameters considered as important for controlling thegrouting operation are to be monitored prior to and during thegrouting operation. Records are to be kept of all monitoredparameters. These parameters are normally to include:

— results from qualification tests for grout mix— results from grout tests during operation— records of grout density at mixer and of total volumes

pumped for each compartment or annulus— records from differential pressure measurements, if appli-

cable— observation records from evacuation points— records of grout density at evacuation points or density of

return grout— results from compressive strength testing.

302 Means are to be provided for observing the emergenceof grout from the evacuation point from the compartment/annulus being grouted. 303 During fabrication of grout, regular tests are to be car-ried out for confirming of the following properties:

— density— air content— viscosity — workability— bleeding— temperature of grout — compressive strength.

Guidance note:A Grouting Procedure including the Quality Control Scheme forthe grout operation is to be worked out.The Quality Control Scheme shall name the responsible compa-nies or personnel for each grouting operation activity.The density and air content are normally to be checked manuallyevery half hour. The viscosity, workability, bleeding and temper-ature are to be checked once every two hours or once per com-partment or annulus to be grouted if the grouting takes less thantwo hours.


DET NORSKE VERITAS


SECTION 10FOUNDATION DESIGN

A. GeneralA 100 Introduction101 The requirements in this section apply to pile founda-tions, gravity type foundations, and stability of sea bottom.102 Foundation types not specifically covered by this stand-ard shall be specially considered.103 Design of foundations shall be based on site-specificinformation, see Sec.3.104 The geotechnical design of foundations shall considerboth the strength and the deformations of the foundation struc-ture and of the foundation soils.This section states requirements for

— foundation soils— soil reactions upon the foundation structure— soil-structure interaction.

Requirements for the foundation structure itself are given inSec.7 to Sec.9 as relevant for a foundation structure con-structed from steel and/or concrete.105 A foundation failure mode is defined as the mode inwhich the foundation reaches any of its limit states. Examplesof such failure modes are

— bearing failure— sliding— overturning— pile pull-out— large settlements or displacements.

106 The definition of limit state categories as given in Sec.2is valid for foundation design with the exception that failuredue to effect of cyclic loading is treated as an ultimate limitstate (ULS), alternatively as an accidental limit state (ALS),using partial load and material factors as defined for these limitstate categories. The load factors are in this case to be appliedto all cyclic loads in the design load history. Lower load factorsthan prescribed in Sec.5 may be accepted if the total safetylevel can be demonstrated to be within acceptable limits.107 The load factors to be used for design related to the dif-ferent categories of limit states are given in Sec.5. 108 The material factors to be used are specified in the rele-vant subsection for design in this Section. The characteristicstrength of soil shall be assessed in accordance with item 300.109 Material factors shall be applied to soil shear strength asfollows:

— for effective stress analysis, the tangent to the characteris-tic friction angle shall be divided by the material factor γm

— for total stress analysis, the characteristic undrained shearstrength shall be divided by the material factor γm.

For soil resistance to axial pile load, material factors shall beapplied to the characteristic resistance as described in C107.For soil resistance to lateral pile load, material factors shall beapplied to the characteristic resistance as described in C106.110 Settlements caused by increased stresses in the soil dueto structural weight shall be considered for structures withgravity type foundations. The risk of uneven settlementsshould be considered in relation to the tolerable tilt of the windturbine support structure.111 Further elaborations on design principles and examples

of design solutions for foundation design are given in DNVClassification Notes 30.4.

A 200 Soil investigations201 Requirements to soil investigations as a basis for estab-lishing necessary soil data for a detailed design are given inSec.3.

A 300 Characteristic properties of soil301 The characteristic strength and deformation propertiesof soil shall be determined for all deposits of importance.302 The characteristic value of a soil property shall accountfor the variability in that property based on an assessment ofthe soil volume that governs the limit state in consideration.

Guidance note:Variability in a soil property is usually a variability of that prop-erty from point to point within a soil volume. When small soilvolumes are involved, it is necessary to base calculations on thelocal soil property with its full variability. When large soil vol-umes are involved, the effect of spatial averaging of the fluctua-tions in the soil property from point to point over the soil volumecomes into play. Calculations may then be based on the spatiallyaveraged soil property, which eventually becomes equal to themean of the soil property when the soil volume is large enough.


303 The results of both laboratory tests and in-situ tests shallbe evaluated and corrected as relevant on the basis of recog-nised practice and experience. Such evaluations and correc-tions shall be documented. In this process account shall begiven to possible differences between properties measured inthe tests and those soil properties that govern the behaviour ofthe in-situ soil for the limit state in question. Such differencesmay be due to:

— soil disturbance due to sampling and samples not reconsti-tuted to in-situ stress history

— presence of fissures— different loading rate between test and limit state in ques-

tion— simplified representation in laboratory tests of certain

complex load histories— soil anisotropy effects giving results which are dependent

on the type of test.

304 Possible effects of installation activities on the soil prop-erties should be considered.305 The characteristic value of a soil property shall be a cau-tious estimate of the value that affects the occurrence of thelimit state, selected such that the probability of a worse valueis low.306 A limit state may involve a large volume of soil and it isthen governed by the spatial average of the soil property withinthat volume. The choice of the characteristic value shall takedue account of the number and quality of tests within the soilvolume involved. Specific care should be made when the limitstate is governed by a narrow zone of soil.307 The characteristic value of a soil property shall beselected as a lower value, being less than the most probablevalue, or an upper value being greater, depending on which isworse for the limit state in question.

Guidance note:Relevant statistical methods should be used. When such methodsare used, the characteristic value of a local soil property should

DET NORSKE VERITAS


be derived such that the probability of a worse value governingthe occurrence of the limit state is not greater than 5%. For selection of characteristic values of soil properties by meansof statistical methods, reference is made to DNV-RP-C207.


A 400 Effects of cyclic loading401 The effects of cyclic loading on the soil properties shallbe considered in foundation design where relevant.402 Cyclic shear stresses may lead to a gradual increase inpore pressure. Such pore pressure build-up and the accompa-nying increase in cyclic and permanent shear strains mayreduce the shear strength of the soil. These effects shall beaccounted for in the assessment of the characteristic shearstrength for use in design within the applicable limit state cat-egories.403 In the SLS design condition the effects of cyclic loadingon the soil’s shear modulus shall be corrected for as relevantwhen dynamic motions, settlements and permanent (long-term) horizontal displacements shall be calculated. See alsoD500.404 The effects of wave- and wind-induced forces on the soilproperties shall be investigated for single storms and for sev-eral succeeding storms, where relevant.405 In seismically active areas, where the structure-founda-tion system may be subjected to earthquake forces, the deteri-orating effects of cyclic loading on the soil properties shall beevaluated for the site-specific conditions and considered in thedesign where relevant. See also 500.

A 500 Soil-structure interaction501 Evaluation of structural load effects shall be based on anintegrated analysis of the soil and structure system. The analy-sis shall be based on realistic assumptions regarding stiffnessand damping of both the soil and structural members.502 Due consideration shall be given to the effects of adja-cent structures, where relevant.503 For analysis of the structural response to earthquakevibrations, ground motion characteristics valid at the base ofthe structure shall be determined. This determination shall bebased on ground motion characteristics in free field and onlocal soil conditions using recognised methods for soil andstructure interaction analysis.

B. Stability of SeabedB 100 Slope stability101 The risk of slope failure shall be evaluated. Such evalu-ations shall cover:

— natural slopes— slopes developed during and after installation of the structure— future anticipated changes of existing slopes— effect of continuous mudflows— wave induced soil movements.

The effect of wave loads on the sea bottom shall be included inthe evaluation when such loads are unfavourable.102 When the structure is located in a seismically activeregion, the effects of earthquakes on the slope stability shall beincluded in the analyses.103 The safety against slope failure for ULS design shall beanalysed using material factors (γM):

B 200 Hydraulic stability201 The possibility of failure due to hydrodynamic instabil-ity shall be considered where soils susceptible to erosion orsoftening are present.202 An investigation of hydraulic stability shall assess therisk for:

— softening of the soil and consequent reduction of bearingcapacity due to hydraulic gradients and seepage forces

— formation of piping channels with accompanying internalerosion in the soil

— surface erosion in local areas under the foundation due tohydraulic pressure variations resulting from environmen-tal loads.

203 When erosion is likely to reduce the effective foundationarea, measures shall be taken to prevent, control and/or moni-tor such erosion, as relevant, see 300.

B 300 Scour and scour prevention301 The risk for scour around the foundation of a structureshall be taken into account unless it can be demonstrated thatthe foundation soils will not be subject to scour for theexpected range of water particle velocities.

Guidance note:When a structure is placed on the seabed, the water-particle flowassociated with steady currents and passing waves will undergosubstantial changes. The local change in the flow will generallycause an increase in the shear stress on the seabed, and the sedi-ment transport capacity of the flow will increase. In the case ofan erodible seabed, this may result in a local scour around thestructure. Such scour will be a threat to the stability of the struc-ture and its foundation.


302 The effect of scour, where relevant, shall be accountedfor according to at least one of the following methods:

a) Adequate means for scour protection is placed around thestructure as early as possible after installation.

b) The foundation is designed for a condition where all mate-rials, which are not scour-resistant, are assumed removed.

c) The seabed around the structure is kept under close sur-veillance and remedial works to prevent further scour arecarried out shortly after detection of significant scour.

303 In an analysis of scour, the effect of steady current,waves, or current and waves in combination shall be taken intoaccount as relevant.

Guidance note:The extent of a scour hole will depend on the dimensions of thestructure and on the soil properties. In cases where a scour pro-tection is in place, it will also depend on the design of the scourprotection.


304 Scour protection material shall be designed to provideboth external and internal stability, i.e. protection againstexcessive surface erosion of the scour protection material andprotection against transportation of soil particles from theunderlying natural soil.

Guidance note:When scour protection consists of an earth structure, such as asequence of artificially laid-out soil layers, it must be ensuredthat standard filter criteria are met when the particle sizes of theindividual layers of such an earth structure are selected.


305 In cases where a scour protection is in place at a founda-tion structure and consists of an earth structure, the effect ofsoil support from the scour protection can be taken into

γM = 1.15 for effective stress analysis= 1.25 for total stress analysis.

DET NORSKE VERITAS


account in the design of the foundation structure. For this pur-pose, a scour hole in the scour protection material shall beassumed with dimensions equal to those that are assumed inthe design of the scour protection for the relevant governingULS event.306 A methodology for prediction of scour around a verticalpile that penetrates the seabed is given in Appendix J.

C. Pile FoundationsC 100 General101 The load-carrying capacity of piles shall be based onstrength and deformation properties of the pile material as wellas on the ability of the soil to resist pile loads.102 In evaluation of soil resistance against pile loads, the fol-lowing factors shall be amongst those to be considered:

— shear strength characteristics— deformation properties and in-situ stress conditions of the

foundation soil— method of installation— geometry and dimensions of pile— type of loads.

103 The data bases of existing methods for calculation of soilresistance to axial and lateral pile loads are often not coveringall conditions of relevance for offshore piles. This in particularrelates to size of piles, soil shear strength and type of load.When determining the soil resistance to axial and lateral pileloads, extrapolations beyond the data base of a chosen methodshall be made with thorough evaluation of all relevant param-eters involved.104 It shall be demonstrated that the selected solution for thepile foundation is feasible with respect to installation of thepiles. For driven piles, this may be achieved by a driveabilitystudy or an equivalent analysis.105 Structures with piled foundations shall be assessed withrespect to stability for both operation and temporary designconditions, e.g. prior to and during installation of the piles.

Guidance note:For drilled piles, it is important to check the stability of thedrilled hole in the temporary phase before the pile is installed inthe hole.


106 Unless otherwise specified, the following material fac-tors γM shall be applied to the characteristic soil strengthparameters for determination of design soil resistance againstlateral loading of piles in the ULS and the SLS:

107 For determination of design pile resistance against axialpile loads in ULS design, a material factor γM = 1.25 shall beapplied to all characteristic values of pile resistance, i.e. tocharacteristic limit skin friction and characteristic tip resist-ance.

Guidance note:This material factor may be applied to pile foundations of multi-legged jacket or template structures. The design pile loads shallbe determined from structural analyses in which the pile founda-tion is modelled either with an adequate equivalent elastic stiff-ness or with non-linear models that reflect the true non-linear

stress-strain properties of the soil in conjunction with the charac-teristic soil strength.If the ultimate plastic resistance of the foundation system is ana-lysed by modelling the soil with its design strength and allowingfull plastic redistribution until a global foundation failure isreached, higher material factors should be used.For individual piles in a group lower material factors may beaccepted, as long as the pile group as a whole is designed with therequired material factor. A pile group in this context shall notinclude more piles that those supporting one specific leg.


108 For drilled piles, the assumptions made for the limit skinfriction in design shall be verified during the installation.

Guidance note:The drilling mud which is used during the drilling of the hole forthe pile influences the adhesion between the pile and the soil andthereby also the limit skin friction.


109 Laterally loaded piles may be analysed on the basis ofrealistic stress-strain curves for soil and pile. The pile deflec-tions induced by the combination of lateral and axial loadingmay be so large that inelastic behaviour of the soil takes place.110 The lateral resistance of a pile or a pile group may in theULS be based on the theory of plasticity provided that the char-acteristic resistance is in accordance with recognised plastictheorems so as to avoid nonconservative estimates of thesafety. The calculations are then to be based on the assumptionthat the lateral deformations of the pile are sufficiently large toplastify the soil completely. 111 When pile penetrations are governed by lateral pileresistance, the design resistance shall be checked with respectto the ULS. For the ULS, material factors as prescribed in 106shall be used.112 For analysis of pile stresses and lateral pile head dis-placements, the lateral pile resistance shall be modelled usingcharacteristic soil strength parameters, with the material factorfor soil strength equal to γm=1.0. Non-linear response of soilshall be accounted for, including the effects of cyclic loading.

C 200 Design criteria for monopile foundations201 For geotechnical design of monopile foundations, boththe ultimate limit state and the serviceability limit state shall beconsidered.202 For design in the ultimate limit state, design soil strengthvalues are to be used for the soil strength, defined as the char-acteristic soil strength values divided by the specified materi-als factor. Design loads are to be used for the loads, eachdesign load being defined as the characteristic load multipliedby the relevant specified load factor. The loads are to be repre-sentative of the extreme load conditions. Two cases are to beconsidered:

— axial loading— combined lateral loading and moment loading.

203 For axial loading in the ULS, sufficient axial pile capac-ity shall be ensured.

Guidance note:The pile head is defined to be the position along the pile in levelwith the seabed. Sufficient axial pile capacity can be ensured bychecking that the design axial load on the pile head does notexceed the design axial resistance, obtained as the design unitskin friction, integrated over the pile surface area, plus a possiblepile tip resistance.For clay, the unit skin friction is a function of the undrained shearstrength. For sand, the unit skin friction is a function of the rela-tive density. In both cases, the unit skin friction may be deter-mined as specified in the API RP2A and the DNV ClassificationNotes No. 30.4.

Type of geotechnical analysisLimit state

ULS SLSγM γM

Effective stress analysis 1.15 1.0Total stress analysis 1.25 1.0

DET NORSKE VERITAS


The effects of cyclic loading on the axial pile resistance shouldbe considered in design. The main objective is to determine theshear strength degradation, i.e. the degradation of the unit skinfriction, along the pile shaft for the appropriate prevailing load-ing intensities.The effects of cyclic loading are most significant for piles incohesive soils, in cemented calcareous soils and in fine-grainedcohesionless soils (silt), whereas these effects are much less sig-nificant in medium to coarsely grained cohesionless soils.


204 For combined lateral loading and moment loading in theULS, sufficient pile capacity against this loading shall beensured. The pile capacity is formed by lateral pile resistance.Verification of sufficient pile capacity implies that the follow-ing two requirements shall be fulfilled: (1) The theoretical design total lateral pile resistance, which isfound by vectorial integration of the design lateral resistanceover the length of the pile, shall not be less than the design lat-eral load applied at the pile head.(2) The lateral displacement at the pile head shall not exceedsome specified limit. The lateral displacement shall be calcu-lated for the design lateral load and moment in conjunctionwith characteristic values of the soil resistance and soil stiff-ness.Requirement (1) is the conventional design rule, which isbased on full plastification of the soil. Requirement (2) is anecessary additional requirement, because lateral soil resist-ance cannot be mobilised locally in zones near points along thepile where the direction of the lateral pile deflection isreversed, i.e. the soil in these zones will not be fully plastified,regardless of how much the pile head deflects laterally.

Guidance note:Sufficient pile capacity against combined lateral loading andmoment loading can be ensured by means of a so-called singlepile analysis in which the pile is discretised into a number ofstructural elements, interconnected by nodal points, and with soilsupport springs in terms of p-y and t-z curves attached at thesenodal points. Lateral forces and overturning moments are appliedto the pile head. Also axial forces acting at the pile head need tobe included, because they may contribute to the bending momentand the mobilization of lateral soil resistance owing to second-order effects.The p-y curves specified for cyclic loading conditions in Appen-dix F can be applied for representation of the lateral support inthis analysis.The p-y curve formulations in Appendix F automaticallyaccounts for the cyclic degradation effects in the lateral resist-ances.The acceptance criterion for sufficient lateral pile resistanceneeds to be a criterion on displacement, cf. Requirement (2). Acriterion on the lateral deflection of the pile head or a criterion onthe rotation of the pile head about a horizontal axis will be prac-tical. When particularly conservative assumptions have beenmade for the lateral soil resistance, Requirement (2) can bewaived.It will usually not suffice to ensure that the lateral design load atthe pile head does not exceed the design total lateral resistancethat is theoretically available and which can be obtained from thesingle-pile analysis. This is so because long before the total avail-able lateral resistance becomes mobilised by mobilisation of alllateral soil resistance along the pile, excessive (and unaccepta-ble) lateral pile displacements will take place at the pile head. When carrying out a single-pile analysis, it is recommended topay attention to the lateral pile head displacements that resultfrom the single-pile analysis and make sure that they do notbecome too large, e.g. by following the predicted pile head dis-placement as function of the pile length and making sure that thedesign is on the flat part of the corresponding displacemen-s.-length curve.

It is also recommended to make sure that the soil zones along thepile, which are plastified for the lateral ULS loads, are not tooextensive.


205 For design in the serviceability limit state, characteristicsoil strength values are to be used for the soil strength. Charac-teristic loads are to be used for the loads. The loading shall berepresentative of loads that will cause permanent deformationsof the soil in the long term, and which in turn will lead to per-manent deformations of the pile foundation, e.g. a permanentaccumulated tilt of the pile head. For this purpose, the behav-iour of the soil under cyclic loading needs to be represented insuch a manner that the permanent cumulative deformations inthe soil are appropriately calculated as a function of thenumber of cycles at each load amplitude in the applied historyof SLS loads.206 For design in the serviceability limit state, it shall beensured that deformation tolerances are not exceeded. Thedeformation tolerances refer to permanent deformations.

Guidance note:Deformation tolerances are usually given in the design basis andthey are often specified in terms of maximum allowable rotationsof the pile head in a vertical plane. The pile head is usuallydefined to be at the seabed. The deformation tolerances are typi-cally derived from visual requirements and requirements for theoperation of the wind turbine. The deformation tolerances shouldtherefore always be clarified with the wind turbine manufacturer.Usually, an installation tolerance is specified which is a require-ment to the maximum allowable rotation of the pile head at thecompletion of the installation of the monopile. In addition, another tolerance is usually specified which is anupper limit for the accumulated permanent rotation of the pilehead due to the history of SLS loads applied to the monopilethroughout the design life. The accumulated permanent rotationsubject to meeting this tolerance usually results from permanentaccumulated soil deformations caused by cyclic wave and windloads about a non-zero mean.In some cases, an installation tolerance is specified together witha tolerance for the total rotation owing to installation and perma-nent accumulated deformations. This is usually expressed as arequirement to the rotation or tilt of the pile at the pile head,where the pile head is defined as the position along the pile inlevel with the seabed. If, for example, the tolerance for the totalrotation at seabed is 0.5° and the installation tolerance at seabedis 0.25°, then the limit for the permanent accumulated rotationbecomes 0.25° at seabed.


C 300 Design criteria for jacket pile foundations301 Jacket piles are the piles that support a jacket or framestructure such as a tripod platform. For geotechnical design ofjacket piles, both the ultimate limit state and the serviceabilitylimit state shall be considered.302 For design in the ultimate limit state, design soil strengthvalues are to be used for the soil strength, defined as the char-acteristic soil strength values divided by the specified materi-als factor. Design loads are to be used for the loads, eachdesign load being defined as the characteristic load multipliedby the relevant specified load factor. The loads are to be repre-sentative of the extreme load conditions. Two cases are to beconsidered:

— axial loading— combined lateral loading and moment loading

303 For axial loading, sufficient axial pile capacity in theULS shall be ensured for each single pile. For combined lateralloading and moment loading, sufficient pile capacity againstthis loading in the ULS shall be ensured for each single pile.

DET NORSKE VERITAS


Guidance note:The verification of sufficient axial and lateral capacities of theindividual piles can be performed by means of an integrated anal-ysis of the entire support structure and its foundation piles, sub-ject to the relevant design loads.In such an analysis, the piles are discretised into a number ofstructural elements, interconnected by nodal points, and with soilsupport springs in terms of p-y and t-z curves attached at thesenodal points to represent lateral and axial load-displacement rela-tionships, respectively. The p-y curves can be generated according to procedures givenin Appendix F for cyclic loading conditions. p-y curves estab-lished according to these procedures will automatically accountfor cyclic degradation effects in the lateral resistances. The t-z curves depend on the unit skin friction. For clay, the unitskin friction is a function of the undrained shear strength. Forsand, the unit skin friction is a function of the relative density. Inboth cases, the unit skin friction may be determined as specifiedin Appendix F.It is important to consider the effects of the cyclic loading on theunit skin friction. The degradation of the unit skin friction shouldbe determined for the relevant prevailing load intensities beforethe t-z curves are generated.The effects of cyclic loading are most significant for piles incohesive soils, in cemented calcareous soils and in fine-grainedcohesionless soils (silt), whereas these effects are much less sig-nificant in medium to coarsely grained cohesionless soils.


304 Pile group effects shall be accounted for.Guidance note:When piles are closely spaced, the resistance of the piles as agroup may be less than the sum of the individual pile capacities,both laterally and axially, and the lateral and axial resistances ofthe p-y and t-z curves should be adjusted accordingly. When piles are closely spaced, the load transferred from eachpile to its surrounding soils leads to displacements of the soilsthat support the other piles, and the behaviour of the piles as agroup may be softer than if the piles were considered to have sup-ports which were not displaced by influence from the neighbour-ing piles. This effect may in principle be accounted for by elastichalf-space solutions for displacements in a soil volume due toapplied point loads.


305 For design in the serviceability limit state, characteristicsoil strength values are to be used for the soil strength. Charac-teristic loads are to be used for the loads. The loading shall berepresentative of loads that will cause permanent deformationsof the soil in the long term, and which in turn will lead to per-manent deformations of the pile foundation, e.g. a permanentaccumulated tilt of the support structure. For this purpose, thebehaviour of the soil under cyclic loading needs to be repre-sented in such a manner that the permanent cumulative defor-mations in the soil are appropriately calculated as a function ofthe number of cycles at each load amplitude in the applied his-tory of SLS loads. 306 For design in the serviceability limit state, it shall beensured that deformation tolerances are not exceeded.

Guidance note:Deformation tolerances are usually given in the design basis andthey are often specified in terms of maximum allowable rotationsof the support structure and maximum allowable horizontal dis-placements of the pile heads.Separate tolerances may be specified for the support structureand piles for the situation immediately after completion of theinstallation and for the permanent cumulative damages owing tothe history of SLS loads applied to the structure and foundationthroughout the design life.


C 400 Design of piles subject to scour401 Effects of scour shall be accounted for. Scour will leadto complete loss of lateral and axial resistance down to thedepth of scour below the original seabed. Both general scourand local scour shall be considered.

Guidance note:The p-y and t-z curves must be constructed with due considera-tion of the effects of scour. In the case of general scour, which is characterised by a generalerosion and removal of soil over a large area, all p-y and t-zcurves are to be generated on the basis of a modified seabed levelwhich is to be taken as the original seabed level lowered by aheight equal to the depth of the general scour.General scour reduces the effective overburden. This has animpact on the lateral and axial pile resistances in cohesionlesssoils. This also has an impact on the depth of transition betweenshallow and deep ultimate lateral resistances for piles in cohesivesoils.In the case of local scour, which is characterised by erosion andremoval of soil only locally around each pile, the p-y and t-zcurves should be generated with due account for the depth of thescour hole as well as for the lateral extent of the scour hole. Thescour-hole slope and the lateral extent of the scour hole can beestimated based on the soil type and the soil strength. Over thedepth of the scour hole below the original seabed level, no soilresistance and thus no p-y or t-z curves are to be applied.Unless data indicate otherwise, the depth of a current-inducedscour hole around a pile in sand can be assumed equal to a factor1.3 times the pile diameter. For large-diameter piles such asmonopiles, this emphasises the need for scour protection unlessthe piles are designed with additional lengths to counteract theeffects of the scour.


D. Gravity Base FoundationsD 100 General101 Failure modes within the categories of limit states ULSand ALS shall be considered as described in 200.102 Failure modes within the SLS, i.e. settlements and dis-placements, shall be considered as described in 300 usingmaterial coefficient γM = 1.0.

D 200 Stability of foundations201 The risk of shear failure below the base of the structureshall be investigated for all gravity type foundations. Suchinvestigations shall cover failure along any potential shear sur-face with special consideration given to the effect of soft layersand the effect of cyclic loading. The geometry of the founda-tion base shall be accounted for.

Guidance note:For gravity base structures equipped with skirts which penetratethe seabed, the theoretical foundation base shall be assumed to beat the skirt tip level. Bucket foundations, for which penetratingskirts are part of the foundation solution, and for which suctionis applied to facilitate the installation, shall be considered asgravity base structures for the condition after the installation iscompleted.


202 The analyses shall be carried out for fully drained, par-tially drained or undrained conditions, whatever representsmost accurately the actual conditions.203 For design within the applicable limit state categoriesULS and ALS, the foundation stability shall be evaluated byone of the following methods:

— effective stress stability analysis— total stress stability analysis.

DET NORSKE VERITAS


204 An effective stress stability analysis shall be based oneffective strength parameters of the soil and realistic estimatesof the pore water pressures in the soil.205 A total stress stability analysis shall be based on totalshear strength parameters determined from tests on represent-ative soil samples subjected to similar stress conditions as thecorresponding elements in the foundation soil.206 Both effective stress and total stress analysis methodsshall be based on laboratory shear strength with pore pressuremeasurements included. The test results should preferably beinterpreted by means of stress paths.207 Stability analyses by conventional bearing capacity for-mulae are only acceptable for uniform soil conditions.

Guidance note:Gravity base foundations of wind turbines usually have relativelysmall areas, such that bearing capacity formulae for idealisedconditions will normally suffice and be acceptable for design.


208 For structures where skirts, dowels or similar foundationmembers transfer loads to the foundation soil, the contribu-tions of these members to the bearing capacity and lateralresistance may be accounted for as relevant. The feasibility ofpenetrating the skirts shall be adequately documented.209 Foundation stability shall be analysed in the ULS byapplication of the following material factors to the characteris-tic soil shear strength parameters:

210 Effects of cyclic loading shall be included by applyingload factors in accordance with A106.211 In an effective stress analysis, evaluation of pore pres-sures shall include:

— initial pore pressure— build-up of pore pressures due to cyclic load history— transient pore pressures through each load cycle— effects of dissipation.

212 The safety against overturning shall be investigated inthe ULS and in the ALS.

D 300 Settlements and displacements301 For SLS design conditions, analyses of settlements anddisplacements are, in general, to include calculations of:

— initial consolidation and secondary settlements— differential settlements— permanent (long term) horizontal displacements— dynamic motions.

302 Displacements of the structure, as well as of its founda-tion soils, shall be evaluated to provide the basis for design ofconductors and other members connected to the structurewhich are penetrating the seabed or resting on the seabed.303 Analysis of differential settlements shall account for lat-eral variations in soil conditions within the foundation area,non-symmetrical weight distributions and possible predomi-nating directions of environmental loads. Differential settle-ments or tilt due to soil liquefaction shall be considered inseismically active areas.

D 400 Soil reactions on foundation structure401 The reactions from the foundation soils shall be

accounted for in the design of the supported structure for alldesign conditions.402 The distribution of soil reactions against structural mem-bers, seated on or penetrated into the sea floor, shall be esti-mated from conservatively assessed distributions of strengthand deformation properties of the foundation soil. Possiblespatial variation in soil conditions, including uneven seabedtopography, shall be considered. The stiffness of the structuralmembers shall be taken into account.403 The penetration resistance of dowels and skirts shall becalculated based on a realistic range of soil strength parame-ters. The structure shall be provided with sufficient capacity toovercome the maximum expected penetration resistance inorder to reach the required penetration depth.404 As the penetration resistance may vary across the foun-dation site, eccentric penetration forces may be necessary tokeep the platform inclination within specified limits.

D 500 Soil modelling for dynamic analysis501 Dynamic analyses of a gravity structure shall considerthe effects of soil-structure interaction. For homogeneous soilconditions, modelling of the foundation soils using the contin-uum approach may be used. For non-homogeneous conditions,modelling by finite element techniques or other recognisedmethods accounting for non-homogenous conditions shall beperformed.

Guidance note:When the soil conditions are fairly homogeneous and an equiva-lent shear modulus G can be determined, representative for theparticipating soil volume as well as for the prevailing strain levelin the soil, then the foundation stiffnesses may be determinedbased on formulae from elastic theory, see Table D1 and TableD2. Foundation springs based on these formulae will be repre-sentative for the dynamic foundation stiffnesses that are neededin structural analyses for wind and wave loading on the wind tur-bine and its support structure. In structural analyses for earth-quake loads, however, it may be necessary to apply frequency-dependent reductions of the stiffnesses from Table D1 and TableD2 to get appropriate dynamic stiffness values for the analyses.


502 Due account shall be taken of the strain dependency ofshear modulus and internal soil damping. Uncertainties in thechoice of soil properties shall be reflected in parametric studiesto find the influence on response. The parametric studiesshould include upper and lower boundaries on shear moduliand damping ratios of the soil. Both internal soil damping andradiation damping shall be considered.

D 600 Filling of voids601 In order to assure sufficient stability of the structure orto provide a uniform vertical reaction, filling of the voidsbetween the structure and the seabed, e.g. by underbase grout-ing, may be necessary.602 The foundation skirt system and the void-filling systemshall be designed so that filling pressures do not cause channel-ling from one skirt compartment to another or to the seabedoutside the periphery of the structure.603 The filling material used shall be capable of retainingsufficient strength during the lifetime of the structure consid-ering all relevant forms of deterioration such as:

— chemical— mechanical— placement problems such as incomplete mixing and dilu-

tion.

γM = 1.15 for effective stress analysis= 1.25 for total stress analysis.

DET NORSKE VERITAS


Table D1 Circular footing on stratum over bedrock or on stratum over half spaceOn stratum over bedrock On stratum over half space

Mode of motion Foundation stiffness Foundation stiffness

Vertical

Horizontal

Rocking

Torsion Not given

)28.11(14

HRGRKV +

−=

ν

2

11

1

28.11

28.11

14

GG

HR

HR

RGKV+

+

−=

ν; 1≤H/R≤5

)2

1(28

HRGRK H +

−=

ν

2

11

1

21

21

28

GG

HR

HR

RGKH+

+

−=

ν; 1≤H/R≤4

)6

1()1(3

8 3

HRGRK R +

−=

ν

2

11

31

61

61

)1(38

GG

HR

HR

RGKR+

+

−=

ν; 0.75≤ H/R ≤2

316 3GRKT =

DET NORSKE VERITAS


Table D2 Circular footing embedded in stratum over bedrock

Mode of motion

Foundation stiffness

Vertical

Horizontal

Rocking

Torsion

)/1

/)28.085.0(1)(2

1)(28.11(14

HDHD

RD

RD

HRGRK V −

−+++−

=ν

)451)(

321)(

21(

28

HD

RD

HRGRK H +++

−=

ν

)7.01)(21)(6

1()1(3

8 3

HD

RD

HRGRK R +++

−=

ν

)381(

316 3

RDGRKT +=

DET NORSKE VERITAS


SECTION 11CORROSION PROTECTION

A. GeneralA 100 General101 In this section, the requirements regarding corrosionprotection arrangement and equipment are given.102 Methods for corrosion protection include, but are notlimited to, corrosion allowance, cathodic protection and coat-ing. Biocides and scavengers can reduce corrosion in closedcompartments.103 When corrosion allowance is part of the required corro-sion protection, the corrosion allowance shall be considered inall limit state analyses. Fatigue calculations can be based on asteel wall thickness equal to the nominal thickness reduced byhalf the allowance over the full service life.

B. Acceptable Corrosion ProtectionB 100 Atmospheric zone101 Steel structure components in the atmospheric zone shallbe protected by coating.

B 200 Splash zone201 Steel structure components in the splash zone shall beprotected by coating and corrosion allowance. The splash zoneis the part of a support structure which is intermittentlyexposed to air and immersed in the sea. The zone has specialrequirements to fatigue. 202 The wave height to be used to determine the upper andlower limits of the splash zone shall be taken as one-third of the100-year wave height. 203 The upper limit of the splash zone SZU shall be calcu-lated asSZU = U1 + U2 + U3in whichU1 = 60% of the wave height defined in 202U2 = highest astronomical tide (HAT)U3 = foundation settlement, if applicable.SZU is measured from mean seawater level. U1, U2 and U3shall be applied as relevant to the structure in question with asign leading to the largest or larger value of SZU.For floating support structures, the upper limit of the splashzone should be calculated according to DNV-OS-C101.204 The lower limit of the splash zone SZL shall be calcu-lated asSZL = L1 + L2in whichL1 = 40% of the wave height defined in 202L2 = lowest astronomical tide (LAT).SZL is measured from mean seawater level. L1 and L2 shall beapplied as relevant to the structure in question with a sign lead-ing to the smallest or smaller value of SZL.For floating support structures, the lower limit of the splashzone should be calculated according to DNV-OS-C101.205 The corrosion protection systems shall be suitable forresisting the aggressive environment in the splash zone. Appli-cation of corrosion allowance may form the main system for

corrosion protection, i.e. the wall thicknesses of structuralcomponents are increased during design to allow for corrosionin operation. The particular corrosion allowance shall beassessed in each particular case. The corrosion allowance shallbe selected in accordance with the site-specific corrosion ratefor steel in the submerged zone and in the splash zone and inaccordance with the planned inspection and repair strategy.Advanced corrosion protection systems can reduce the corro-sion rate. A reduced corrosion rate can be utilised in design,provided inspection and repair are feasible and provided aplanned strategy for inspection and repair is in place.

Guidance note:Corrosion rates for steel in the submerged zone and in the splashzone depend on the chloride content of the seawater. The chlo-ride content of seawater is site-specific. In the North Sea, it can generally be assumed that the corrosionrate in the splash zone is in the range 0.3 to 0.5 mm per year. Acorrosion allowance of minimum 6 mm is recommended forcoated primary steel structures without planned coating repair ina 20-year design life. A corrosion allowance of minimum 2 mmis recommended for replaceable secondary structures.It is recommended to combine a protection system based on cor-rosion allowance with surface protection such as glass flake rein-forced epoxy coating. When such a combination is applied, thereducing effect of the surface protection on the corrosion rateshall not be taken into account. The beneficial effect of the sur-face protection on the fatigue life may be taken into accountthrough selection of the relevant S-N curve.


206 Corrosion allowance shall be taken into account bydecreasing the nominal wall thickness in the correspondinglimit state analyses.

Guidance note:Fatigue calculations can be based on a steel wall thickness equalto the nominal thickness reduced by half the corrosion allowanceover the full service life.For North Sea conditions, a reduced corrosion allowance of 3 to5 mm should be applied to all primary steel structures in thesplash zone for fatigue analyses for a 20-year lifetime. Forreplaceable secondary structures in the splash zone, a reducedcorrosion allowance of 2 mm can be applied.


B 300 Submerged zone301 Steel structure components in the submerged zone shallbe cathodically protected. Use of coating is optional.

Guidance note:The submerged zone consists of the region below the splashzone, including the scour zone and the zone of permanent embed-ment.In the scour zone, the cathodic protection might not be fullyeffective and anaerobic corrosion can occur. A corrosion allowance is advisable both internally and externallyon steel piles near the seabed, depending on the detailed design,for example for jacket piles where the arrangement of pilesleeves and mudmats complicates effective cathodic protection.For typical North Sea conditions and a 20-year lifetime, it is rec-ommended to design with a corrosion allowance of 2 mm in thescour zone.If the inside of piles such as monopiles is ensured to be airtight,i.e. there is no or very low content of oxygen, corrosion protec-tion inside of the piles is not required.


DET NORSKE VERITAS


B 400 Closed compartments401 Closed compartments with seawater shall be protectedby cathodic protection, by coating near the water line andabove the water line, and by corrosion allowance. The neces-sary corrosion allowance shall be established from experiencedata on a case to case basis.

C. Cathodic ProtectionC 100 General101 Requirements to cathodic protection are given in DNV-RP-B401. 102 The electrical potential for the cathodic protection shallbe verified after the cathodic protection has been installed.

Guidance note:The recommendations for corrosion allowance in the zone nearthe seabed, see B301, where the cathodic protection may not besufficiently effective, can be disregarded when a good electricalconnection is established for the cathodic protection system.


D. CoatingD 100 General101 Requirements to coating are given in DNV-OS-C101. Forapplication of coating, reference is made to DNV-OS-C401.

102 Structures above the splash zone shall be protected by ahigh quality multilayer coating system as specified for corro-sivity category C5M in ISO 12944.103 Coating systems for structures in the splash zone and inzones below the splash zone shall be designed as for structuresabove the splash zone, see 102. In addition, they shall be qual-ified for compatibility with cathodic protection systems. Selec-tion and qualification of coating systems shall address allconditions relevant for necessary repair after installation.

Guidance note:Coating systems for the splash zone should meet the require-ments of NORSOK M-501 and ISO 20340.


104 Coating systems for structures in the splash zone shall beselected with due consideration of loads from impacts fromservice vessels and floating ice.

Guidance note:Glass flakes can be used to reinforce epoxy-based coating sys-tems to improve their resistance against mechanical loads.


105 Below the splash zone coating is optional.Guidance note:Coating can provide a reliable corrosion protection and can bedesigned to reduce marine growth. However, coating canbecome damaged during inspection and maintenance sessionswhere marine growth is removed.


DET NORSKE VERITAS


SECTION 12TRANSPORT AND INSTALLATION

A. Marine Operations

A 100 Warranty surveys101 Warranty surveys are required for insurance of the seatransport project phase and the installation project phase.102 Warranty surveys are to be carried out in accordancewith an internationally recognised scheme. The DNV ‘Rulesfor Planning and Execution of Marine Operations’ is acceptedby the insurance, finance and marine industries. Marine oper-ations cover yard lift, load out, sea transportation, offshore liftand installation operations.103 DNV ‘Rules for Planning and Execution of MarineOperations’, Part 1, Chapter 1, describes in detail the princi-ples, the scope and the procedures for insurance warranty sur-veys.

A 200 Planning of operations201 The planning of the operations should cover planningprinciples, documentation and risk evaluation. The planningand design sequence is given in Figure 1.

Figure 1 Planning and design sequence

202 Operational prerequisites such as design criteria,weather forecast, organisation, marine operation manuals aswell as preparation and testing should be covered. 203 The stability of the installation vessels shall be evalu-ated. This evaluation includes evaluation of stability duringbarge transports and load-out operations and applies to all ves-sels used during the installation, including special vessels suchas floating cranes. Equipment including equipment used fortowing of vessels and for mooring systems is also subject toevaluation.204 Acceptable characteristics shall be documented for thehandled object and all equipment, temporary or permanentstructures, vessels etc. involved in the operation.

Guidance note:Note that all elements of the marine operation shall be docu-mented. This also includes onshore facilities such as quays, soil,pullers and foundations.


205 Properties for object, equipment, structures, vessels etc.may be documented with recognised certificates. The basis forthe certification shall then be clearly stated, i.e. acceptancestandard, basic assumptions, dynamics considered etc., andshall comply with the philosophy and intentions of DNV‘Rules for Planning and Execution of Marine Operations’.206 Design analysis should typically consist of various lev-els with a “global” analysis at top level, and with strength cal-culations for details as a lowest level. Different types ofanalysis methods and tools may apply for different levels.207 Operational aspects shall be documented in the form ofprocedure, operation manuals, certificates, calculations etc.Relevant qualifications of key personnel shall be documented.208 All relevant documentation shall be available on siteduring execution of the operation.209 The documentation shall demonstrate that philosophies,principles and requirements of DNV ‘Rules for Planning andExecution of Marine Operations’ are complied with.210 Documentation for marine operations shall be self con-tained or clearly refer to other relevant documents.211 The quality and details of the documentation shall besuch that it allows for independent reviews of plans, proce-dures and calculations for all parts of the operation.

Guidance note:A document plan describing the document hierarchy and scopefor each document is recommended for major marine operations.


212 Applicable input documentation such as:

— statutory requirements— rules— company specifications— standards and codes— concept descriptions— basic engineering results (drawings, calculations etc.)— relevant contracts or parts of contracts.

should be identified before any design work is performed.213 Necessary documentation shall be prepared to proveacceptable quality of the intended marine operation. Typically,output documentation consists of:

— planning documents including design briefs and designbasis, schedules, concept evaluations, general arrange-ment drawings and specifications

— design documentation including load analysis, globalstrength analysis, local design strength calculations, sta-bility and ballast calculations and structural drawings

— operational procedure including testing program and pro-cedure, operational plans and procedure, arrangementdrawings, safety requirement and administrative proce-dures

— certificates, test reports, survey reports, NDE documenta-tion, as built reports, etc.

214 Execution of marine operations shall be logged. Sam-ples of planned recording forms shall be included in the marine

Regulations, RulesSpecifications, Standards

Overall Planning

Design Brief &Design Basis

Engineering &Design Verification

Operational Procedure

DET NORSKE VERITAS


operations manual.215 Further requirements are given in DNV ‘Rules for Plan-ning and Execution of Marine Operations’, Part 1, Chapter 2.

A 300 Design loads301 The design loads include basic environmental condi-tions like wind, wave, current and tide. The design processinvolving characteristic conditions, characteristic loads anddesign loads is illustrated in Figure 2.

Figure 2 Design process

302 The load analysis should take into account dynamiceffects and non-linear effects. Permanent loads, live loads,deformation loads, environmental loads as well as accidentalloads should be considered.303 Further requirements are given in DNV ‘Rules for Plan-ning and Execution of Marine Operations’, Part 1, Chapter 3.

A 400 Structural design401 Prerequisites for structures involved in marine opera-tions shall include design principles, strength criteria for limitstate design, testing, material selection and fabrication.402 Requirements and guidelines are given in DNV ‘Rulesfor Planning and Execution of Marine Operations’, Part 1,Chapter 4.

A 500 Load transfer operations501 The load transfer operations cover load-out, float-out,lift-off and mating operations. 502 Requirements to load transfer operations are given inDNV ‘Rules for Planning and Execution of Marine Opera-tions’, Part 2, Chapter 1.

A 600 Towing601 Specific requirements and guidelines for single-vesseland barge-towing operations are given in DNV ‘Rules forPlanning and Execution of Marine Operations’, Part 2, Chapter2.

A 700 Offshore installation701 Specific requirements and recommendations for off-shore installation operations particularly applicable for fixedoffshore structures like piled or gravity based wind turbinesupport structures are given in DNV ‘Rules for Planning andExecution of Marine Operations’, Part 2, Chapter 4. Environ-mental loads and load cases to be considered are described aswell as on-bottom stability requirements and requirements tostructural strength. 702 Operational aspects for ballasting, pile installation andgrouting shall be considered.

A 800 Lifting801 Guidance and recommendations for well controlled lift-ing operations, onshore, inshore and offshore, of objects withweight exceeding 50 tonnes are given in DNV ‘Rules for Plan-ning and Execution of Marine Operations’, Part 2, Chapter 5.The chapter describes in detail the basic loads, dynamic loads,skew loads and load cases to be considered. Design of slings,grommets and shackles as well as design of the lifted objectitself are covered.802 In addition, operational aspects such as clearances, mon-itoring of lift and cutting of sea fastening are described.

A 900 Subsea operations901 Subsea operations are relevant for tie-in of, for example,electrical cables. Planning, design and operational aspects forsuch installations are described in DNV ‘Rules for Planningand Execution of Marine Operations’, Part 2, Chapter 6.

Characteristic Conditions

Characteristic Loads

Analysis &Calculations

LoadFactors

Design Loads& Load Cases

Design &Verification

DET NORSKE VERITAS


SECTION 13IN-SERVICE INSPECTION, MAINTENANCE AND MONITORING

A. GeneralA 100 General101 An offshore wind farm is typically planned for a 20-yeardesign lifetime. In order to sustain the harsh offshore environ-ment, adequate inspections and maintenance have to be carriedout. This applies to the entire wind farm including substationand power cables.102 This section provides the requirements to the mainte-nance and inspection system for the wind turbines, the supportstructures, the substation and the power cables.

B. Periodical InspectionsB 100 General101 The following periodical inspections shall be performedin order to evaluate the condition of the offshore wind farmduring its design lifetime:

— periodical inspection of wind turbines— periodical inspection of structural and electrical systems

above water— periodical inspection of structures below water — periodical inspection of sea cable.

The periodical inspection consists of three levels of inspection,viz. general visual inspection, close visual inspection and non-destructive examination. General visual inspections can becarried out using an ROV (Remote Operated Vehicle),whereas close visual inspections require inspections carriedout by a diver.

B 200 Preparation for periodical inspections201 A Long Term Inspection Program for the wind farmshall be prepared, in which all disciplines and systems arespecified. In this program, inspection coverage over a five-year period should be specified in order to ensure that allessential components, systems and installations in the offshorewind farm will be covered by annual inspections over the five-year period.202 The periodical inspections should be carried out with ascope of work necessary to provide evidence as to whether theinspected installation or parts thereof continue to comply withthe design assumptions as specified in the Certificate of Com-pliance.203 The scope of work for an inspection shall always containa sufficient number of elements and also highlight any findingsor deviations reported during previous inspections which havenot been reported or dealt with.

Guidance note:The inspection will typically consist of an onshore part and anoffshore part. The onshore part typically includes:- follow up on outstanding points from the previous inspection- revision of inspection procedures- revision of maintenance documentation- interview with discipline engineers, including presentation/

clarification of any comments deduced during review of pro-cedures

- review of maintenance history.- preparation of the offshore program, based on findings from

the onshore part and systems selected from the Long TermInspection Program.

The offshore inspection typically includes test and inspections onsite as well as an assessment of the findings in order to distin-guish between random failures and systematic failures.


B 300 Interval between inspections301 The interval between inspections of critical items shouldnot exceed one year. For less critical items longer intervals areacceptable. The entire wind farm should be inspected at leastonce during a five-year period. Inspection intervals for subse-quent inspections should be modified based on findings. Criti-cal items are assumed to be specified for the specific project inquestion.

B 400 Inspection results401 The results of the periodical inspections shall beassessed and remedial actions taken, if necessary. Inspectionresults and possible remedial actions shall be documented.

B 500 Reporting501 The inspection shall be reported. The inspection reportshall give reference to the basis for the inspection such asnational regulations, rules and inspection programs, instruc-tions to surveyors and procedures. It shall be objective, havesufficient content to justify its conclusions and should includegood quality sketches and/or photographs as considered appro-priate.

C. Periodical Inspection of Wind Turbines C 100 Interval between inspections101 The interval between inspections above water should notexceed one year. In addition the requirements in the wind tur-bine service manual shall be followed.

C 200 Scope for inspection201 The following items shall be covered by the inspection:

— blades— gear boxes— electrical systems— transformers and generators— lifting appliances— fatigue cracks— dents and deformation(s)— bolt pre-tension— status on outstanding issues from previous periodical

inspections of wind turbines.

202 Inspections as required in the wind turbine service man-ual come in addition to the inspection implied by 201.

D. Periodical Inspection of Structural and Electrical Systems above Water

D 100 Interval between inspections101 The interval between inspections above water should notexceed one year. In addition the requirements in the wind tur-bine service manual shall be followed.

DET NORSKE VERITAS


D 200 Scope for inspection201 The following items shall be covered by the inspection:

— electrical systems— transformers and generators— tower structures — lifting appliances— access platforms— upper part of J-tubes— upper part of ladders— upper part of fenders— heli-hoist platforms— corrosion protection systems — marine growth— fatigue cracks— dents — deformation(s)— bolt pre-tension— status on outstanding issues from previous periodical

inspections above water.

202 Inspection for fatigue cracks at least every year asrequired by the list in 101 may be waived depending on whichdesign philosophy has been used for the structural detail inquestion: When the fatigue design of the structural detail hasbeen carried out by use of safety factors corresponding to anassumption of no access for inspection according to Sec.7Table J2, then there is no need to inspect for fatigue cracks andinspection for fatigue cracks may be waived. When smallersafety factors have been used for the fatigue design, inspec-tions need to be carried out. The inspection interval depends onthe structural detail in question and the inspection method andmay be determined based on the magnitude of the safety factorapplied in design. In general, the smaller the safety factor, theshorter is the interval between consecutive inspections.

Guidance note:Provided a reliable inspection, such as an inspection by eddy cur-rent or a magnetic particle inspection, is carried out after a goodcleaning of the hot spot area, the interval between consecutiveinspections can be calculated from the safety level expressed interms of the material factor γm as follows:Inspection interval = Calculated fatigue life ⋅ γm

5/1.255.This implies the following requirements to inspection:

γm = 1.25 No check for fatigue cracks is needed, correspondingto an assumption of no access to the structural detail.

γm = 1.15 Checks for fatigue cracks needed every 13 years ifthe calculated fatigue life is 20 years. This will resultin the same safety level as that achieved for γm = 1.25without inspections.

γm = 1.0 Checks for fatigue cracks needed every 7 years if thecalculated fatigue life is 20 years. This will result inthe same safety level as that achieved for γm = 1.25without inspections.


203 Inspections as required in the wind turbine service man-ual come in addition to the inspection implied by 201.

E. Periodical Inspection of Structures Below Water

E 100 Interval between inspections101 The interval between inspections below water shouldnot exceed five years.

Guidance note:Five-year inspection intervals are common; however, more fre-quent inspections during the first few years after installation arerecommended.


E 200 Scope for inspection201 The following items shall be covered by the inspection:

— support structures— lower part of J-tubes— lower part of ladders— lower part of fenders— corrosion protection systems (anodes, coating etc.)— marine growth— fatigue cracks— scour and scour protection — damages and dents — deformations— debris— status on outstanding issues from previous periodical

inspections below water.

Visual inspections may be carried out by a remotely operatedvehicle (ROV).202 Inspection for fatigue cracks at least every five years asrequired by the list in 101 may be waived depending on whichdesign philosophy has been used for the structural detail inquestion: When the fatigue design of the structural detail hasbeen carried out by use of safety factors corresponding to anassumption of no access for inspection according to Sec.7Table J2, then there is no need to inspect for fatigue cracks andinspection for fatigue cracks may be waived. When smallersafety factors have been used for the fatigue design, inspec-tions need to be carried out. The inspection interval depends onthe structural detail in question and the inspection method andmay be determined based on the magnitude of the safety factorapplied in design. In general, the smaller the safety factor, theshorter is the interval between consecutive inspections.

Guidance note:Provided a reliable inspection, such as an inspection by eddy cur-rent or a magnetic particle inspection, is carried out after a goodcleaning of the hot spot area, the interval between consecutiveinspections can be calculated from the safety level expressed interms of the material factor γm as follows:Inspection interval = Calculated fatigue life ⋅ γm

5/1.255.This implies the following requirements to inspection:

γm = 1.25 No check for fatigue cracks is needed, correspondingto an assumption of no access to the structural detail.

γm = 1.15 Checks for fatigue cracks needed every 13 years ifthe calculated fatigue life is 20 years. This will resultin the same safety level as that achieved for γm = 1.25without inspections.

γm = 1.0 Checks for fatigue cracks needed every 7 years if thecalculated fatigue life is 20 years. This will result inthe same safety level as that achieved for γm = 1.25without inspections.


203 The anode potential shall be measured and fulfil mini-mum requirements. 204 If deemed critical, steel wall thickness shall be meas-ured.

DET NORSKE VERITAS


F. Periodical Inspection of Sea CablesF 100 Interval between inspections101 The interval between inspections of sea cables shouldnot exceed five years.

F 200 Scope for inspection201 Interconnecting power cables between the wind turbinesand the transformer station as well as power cables to the shoreshall be inspected, unless they are buried.202 To the extent that power cables are to be buried, it shallbe ensured that the cables are buried to design depth.

G. DeviationsG 100 General101 Deviations or non-conformances are findings made dur-ing an inspection that require special follow-up. Deviationsmay be assigned one of three different levels of concernaccording to their criticality:

1) Those impairing the overall safety, integrity and fitness ofthe installation or parts thereof and/or the persons onboard.

2) Those which are found to present a hazard for the personsonboard due to deterioration and/or damage, and thosewhere documents are missing for completing a matter.

3) Those which are found starting to deteriorate or thosewhich are found to have minor defects.

The deviations shall be handled and reported accordingly.

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.A see note on front cover

APPENDIX A STRESS CONCENTRATION FACTORS FOR TUBULAR JOINTS

A. Calculation of Stress Concentration FactorsA 100 General101 Calculation of stress concentration factors (SCFs) forsimple planar tubular joints can be carried out by application

of available closed form solutions. The Efthymiou equationsshould be applied for T, Y, DT, and X joints, as well as for Kand KT joints. These parametric equations are expressed interms of a number of geometric parameters whose definitionsare given in Figure 1. The ranges of these parameters for whichthe parametric equations are valid are given in item 103.

Figure 1 Non-dimensional tubular joint parameters

102 The parametric equations for calculation of SCFs fortubular joints are given on the following pages.103 In 1985, Efhtymiou and Durkin published a series ofparametric equations covering T/Y and gap/overlap K joints.Over 150 configurations were analysed with the PMBSHELLfinite element program using 3D thick shell elements for thetubular members and 3D brick elements for the welds with pro-files as per AWS (1994). The hot-spot SCFs were based onmaximum principal stresses linearly extrapolated to the mod-elled weld toe, in accordance with the HSE recommendations,with some consideration being given to boundary conditions(i.e. short cords and cord end fixity). In 1988, Efthymiou pub-lished a comprehensive set of parametric equations coveringT/Y, X, K and KT simple joint configurations. These equationswere designed using influence functions to describe K, KT andmultiplanar joints in terms of simple T braces with carry-overeffects from the additional loaded braces.With respect to the Efthymiou equations reproduced below,the following points should be noted:

— The Efthymiou equations give a comprehensive coverageof all parametric variations and were developed as meanfit equations. They tend to give less conservative SCFsthan the other SCF equations, with the exception of theLloyd’s Register mean equations.

— It has been shown by Efthymiou that the saddle SCF isreduced in joints with short chord lengths, due to therestriction in chord ovalisation caused by either the pres-ence of chord end diaphragms or by the rigidity of thechord end fixing onto the test rig. Therefore, the measuredsaddle SCFs on joints with short chords may be less thanfor the equivalent joint with a more realistic chord length,a factor considered first in the Efthymiou equations andlater adopted in the Lloyd’s Register SCF equations.

— The equations introduce SCF modifiers to account for theinfluence of chord end fixity on beam bending (C) and forthe reduction in chord wall deformations when the chordends are close to the intersection (α < 12) and arerestrained (F).

— For wide gap K joints under balanced axial load, a Y clas-sification is appropriate with chord length parameter α setat 12 to account for the limited beam bending.

The validity range for the Efthymiou equations are as follows:

Chord outer data: Dchord: Diameter Tchord: Thickness

Brace outer data: Dbrace: Diameter Tbrace: Thickness

Lchord

Lgap

2,1,

22

2

bracebrace

gap

chord

brace

chord

chord

chord

brace

chord

chord

DDL

TT

TD

DD

DL

+===== ζτγβα

0.2 ≤ β ≤ 1.00.2 ≤ τ ≤ 1.08 ≤ γ ≤ 324 ≤ α ≤ 4020° ≤ θ ≤ 90°

≤ ζ ≤ 1.0θβ

sin6.0

−

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.A –

Table A1 Stress Concentration Factors for Simple Tubular T/Y Joints

Load type and fixity conditions

SCF equations Eqn. No. Short chord correction

Axial load-Chord ends fixed

Chord saddle:(1) F1

Chord crown:(2) None

Brace saddle:(3) F1

Brace crown:(4) None

Axial load-General fixity conditions

Chord saddle:(5) F2

Chord crown:(6) None

Brace saddle:(Eqn.(3)) F2


In-plane bending Chord crown:(8) None


Out-of-plane bending Chord saddle:(10) F3

Brace saddle:(11) F3

Short chord correction factors (α < 12)

where exp(x) = ex

Chord-end fixity parameterC1 = 2(C-0.5)C2 = C/2C3 = C/5C = chord end fixity parameter0.5 ≤ C ≤ 1.0, Typically C = 0.7

( )( ) ( )1.621.1 θsin0.52β31.11τγ −−

( )( ) ( ) θsin3α0.25βτ0.65β52.65τγ 20.2 −+−+

( )( ) ( )( )0.01α-2.71.10.10.52 θsin0.96β1.25β0.187ατγ1.3 −−+

( )( ) ( )1.2α0.1τβ0.0450.011β4β0.12expγ3 21.2 −+−+−+

( ) ( ) ( )20.5221 sin2θβ1βτ6α0.8C −−+(Eqn.(1))

( )( ) ( ) sinθ3αCβτ0.65β52.65τγ 220.2 −+−+

( )( ) ( )1.2αCτβ0.0450.011ββ4exp0.12γ3 321.2 −+−+−+

( ) ( )0.70.68β10.85 θsinγτ1.45β −

( ) ( )( )1.16-0.06γ0.77β1.090.4 θsinγτ0.65β1 −+

( )( )1.63 θsin1.05β1.7βτγ −

( ) ⋅+−−− 40.050.54 β0.08β0.470.99γτ (Eqn.(10))

( ) ( )2.5-1.160.232 αγ0.21-expγ0.02β0.56-β0.83-1F1 −=

( ) ( )2.5-1.380.042 αγ0.71-expγ0.03β0.97-β1.43-1F2 −=

( )1.8-0.890.161.8 αγ0.49-expγ0.55 β-1F3 =

DET NORSKE VERITAS


Table A2 Stress Concentration Factors for Simple X Tubular Joints Load type and fixity conditions

SCF equation Eqn. no.

Axial load (balanced) Chord saddle:(12)

Chord crown:(13)

Brace saddle:(14)

Brace crown:(15)

In joints with short cords (α < 12) the saddle SCF can be reduced by the factor F1 (fixed chord ends) or F2 (pinned chord ends) where

In plane bending Chord crown:(Eqn.(8))

Brace crown:(Eqn. (9))

Out of plane bending (balanced)

Chord saddle:(16)

Brace saddle:(17)

In joints with short chords (α < 12) eqns. (16) and (17) can be reduced by the factor F3 where:

( ) ( )1.71.8 θsinβ1.10βτγ3.87 −

( )( ) θsinβτ30.65β52.65τγ 20.2 −−+

( ) ( )2.51.70.90.5 θsinβ1.09βτγ1.91 −+

( )( )0.0450.011β4β0.12expγ3 21.2 −+−+

( ) ( )2.5-1.160.232 αγ0.21-expγ0.02β0.56-β0.83-1F1 −=

( ) ( )2.5-1.380.042 αγ0.71-expγ0.03β0.97-β1.43-1F2 −=

( )( )1.64 θsin1.34β1.56βτγ −

( ) ⋅+−−− 40.050.54 β0.08β0.470.99γτ (Eqn.(16))

( )1.8-0.890.161.8 αγ0.49-expγβ0.55-1F3 =

DET NORSKE VERITAS


Table A3 Stress Concentration Factors for Simple Tubular K Joints and Overlap K JointsLoad type and fixity conditions SCF equation Eqn. no. Short chord

correctionBalanced axial load Chord:

(20) None

Brace:(21) None

Where:C = 0 for gap jointsC = 1 for the through braceC = 0.5 for the overlapping braceNote that τ, β, θ and the nominal stress relate to the brace under considerationATAN is arctangent evaluated in radians

Unbalanced in plane bending Chord crown:(Eqn. (8))(for overlaps exceeding 30% of contact length use 1.2 ⋅ (Eqn. (8))

Gap joint brace crown:(Eqn. (9))

Overlap joint brace crown:(Eqn. (9)) ⋅ (0.9+0.4β) (22)

Unbalanced out-of-plane bending

Chord saddle SCF adjacent to brace A:

where

(23) F4

Brace A saddle SCF(24) F4

(Eqn. (10))A is the chord SCF adjacent to brace A as estimated from eqn.(10).Note that the designation of braces A and B is not geometry dependent. It is nominated by the user.

( )

( )( )ζ8ATANβ0.291.64ββ

sinθsinθ

sinθβ1.16β0.67γτ

0.380.30

min

max

0.30

min

max20.50.9

−+⎟⎟⎠

⎞⎜⎜⎝

⎛

⋅⎟⎟⎠

⎞⎜⎜⎝

⎛+−

( ) ( ) ⋅−+ − 0.70.140.25 θsinτβ1.571.971 (Eqn. (20))+

( ) ( )( )1.220.51.5

minmax1.8

τγβC

β4.2ζ14ATAN0.0840.131θθsin−

⋅+−⋅+

( ) ( )( )+− x0.8-expγβ0.081 0.5B(Eqn. (10))A (Eqn. (10))B

( ) ( )( ) ( )( )x1.3-expβ2.05x0.8-expγβ0.081 0.5max

0.5A−

A

A

βsinθζ

1x +=

( )⋅+−−− 40.050.54 β0.08β0.470.99γτ (Eqn. (23))

( )2.4-1.061.88 αγ0.16-expβ1.07-1F4 =

DET NORSKE VERITAS


Table A4 Stress Concentration Factors for Simple Tubular K Joints and Overlap K JointsLoad type and fixity conditions

SCF equations Eqn. No.

Short chord correction

Axial load on one brace only Chord saddle:(Eqn. (5)) F1

Chord crown:(Eqn. (6)) –

Brace saddle:(Eqn.(3)) F1

Brace crown:(Eqn. (7)) –

Note that all geometric parameters and the resulting SCF’s relate to the loaded brace.

In-plane-bending on one brace only

Chord crown:(Eqn. (8))


Note that all geometric parameters and the resulting SCF’s relate to the loaded brace.

Out-of-plane bending on one brace only

Chord saddle:

where

(25) F3

Brace saddle:(26) F3

Short chord correction factors:

( ) ( )( )x0.8-expγβ0.081 0.5B−⋅(Eqn. (10))A

A

A

βsinθζ

1x +=

( )⋅+−−− 40.050.54 β0.08β0.470.99γτ (Eqn. (25))

( ) ( )2.5-1.160.232 αγ0.21-expγ0.02β0.56-β0.83-1F1 −=

( )1.8-0.890.161.8 αγ0.49-expγβ0.55-1F3 =

DET NORSKE VERITAS


Table A5 Stress Concentration Factors for Simple KT Tubular Joints and Overlap KT JointsLoad type SCF equation Eqn. no.Balanced axial load Chord:

(Eqn. (20))Brace:(Eqn. (21))For the diagonal braces A & C use ζ = ζAB + ζBC + βBFor the central brace, B, use ζ = maximum of ζAB, ζBC

In-plane bending Chord crown:(Eqn. (8))Brace crown:(Eqn. (9))

Unbalanced out-of-plane bending

Chord saddle SCF adjacent to diagonal brace A:(Eqn. (10))A

where

(27)

Chord saddle SCF adjacent to central brace B:

where

(28)

Out-of-plane bending brace SCFs

Out-of-plane bending brace SCFs are obtained directly from the adjacent chord SCFs using:

where SCFchord = (Eqn. (27)) or (Eqn. (28))

(29)

( ) ( )( ) ( ) ( )( )+−− AC0.5

CAB0.5

B x0.8-expγβ0.081x0.8-expγβ0.081

( ) ( )( ) ( )( )+−⋅ AB0.5maxAB

0.5A 1.3x-expβ2.050.8x-expγβ0.081(Eqn.(10)) B

( ) ( )( ) ( )( )AC0.5maxAC

0.5A 1.3x-expβ2.050.8x-expγβ0.081−⋅(Eqn.(10)) C

A

AABAB β

θsinζ1x +=

( )A

ABBCABAC β

θsinβζζ1x

+++=

( ) ( )( ) ⋅−⋅ 1PAB

0.5A x0.8-expγβ0.081(Eqn. (10))B

( ) ( )( ) +− 2PBC

0.5C x0.8-expγβ0.081

( ) ( )( ) ( )( )+−⋅ AB0.5maxAB

0.5B x1.3-expβ2.05x0.8-expγβ0.081(Eqn. (10))A

( ) ( )( ) ( )( )BC0.5maxBC

0.5B x1.3-expβ2.05x0.8-expγβ0.081−⋅(Eqn. (10)C

B

BABAB β

θsinζ1x +=

B

BBCBC β

θsinζ1x +=

2

B

A1 β

βP ⎟⎟

⎠

⎞⎜⎜⎝

⎛=

2

B

C2 β

βP ⎟⎟

⎠

⎞⎜⎜⎝

⎛=

( ) chord40.050.54 SCFβ0.08β0.470.99γτ ⋅+−−−

DET NORSKE VERITAS


Axial load on one brace only

Chord saddle:(Eqn. (5))

Chord crown:(Eqn. (6))

Brace saddle:(Eqn. (3))


Out-of-plane bending on one brace only

Chord SCF adjacent to diagonal brace A:

where

(30)

Chord SCF adjacent to central brace B:

where

(31)

Out-of-plane brace SCFs Out-of-plane brace SCFs are obtained directly from the adjacent chord SCFs using:(32)

Table A5 Stress Concentration Factors for Simple KT Tubular Joints and Overlap KT Joints (Continued)Load type SCF equation Eqn. no.

( ) ( )( ) ( ) ( )( )AC0.5

CAB0.5

B x0.8-expγβ0.081x0.8-expγβ0.081 −−⋅(Eqn. (10))A

A

AABAB β

θsinζ1x +=

( )A

ABBCABAC β

θsinβζζ1x

+++=

( ) ( )( ) ⋅−⋅ 1PAB

0.5A x0.8-expγβ0.081(Eqn. (10))B

( ) ( )( ) 2PBC

0.5C x0.8-expγβ0.081−

B

BABAB β

θsinζ1x +=

B

BBCBC β

θsinζ1x +=

2

B

A1 β

βP ⎟⎟

⎠

⎞⎜⎜⎝

⎛=

2

B

C2 β

βP ⎟⎟

⎠

⎞⎜⎜⎝

⎛=

( ) chord40.050.54 SCFβ0.08β0.470.99γτ ⋅+−−−

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.B –

APPENDIX B LOCAL JOINT FLEXIBILITIES FOR TUBULAR JOINTS

A. Calculation of Local Joint FlexibilitiesA 100 General101 Calculation of local joint flexibilities (LJFs) for simpleplanar tubular joints can be carried out by application of avail-able closed form solutions. Buitrago’s parametric expressionsfor LJFs should be used. These expressions give local jointflexibilities of brace ends for axial loading, for in-plane bend-ing and for out-of-plane bending. There are expressions forsingle-brace joints (Y joints), for cross joints (X joints), and forgapped K joints and overlapped K joints. The expressions are

given in terms of a number of geometric parameters whose def-initions are given in Figure 1. LJFs influence the global staticand dynamic structural response.102 In addition to direct flexibility terms between loadingand deformation of a particular brace end, there are cross termsbetween loading of one brace end and deformation of anotherbrace end in joints where more than one brace join in with thechord beam. Figure 1 provides information of degrees of free-dom for which cross terms of local joint flexibility existbetween different brace ends.

Figure 1 General joint geometry, loads, and degrees of freedom

103 The local joint flexibility LJF for a considered degree offreedom of a brace end is defined as the net local deformationof the brace-chord intersection (“footprint”) in the brace localcoordinates due to a unit load applied to the brace end. 104 The local joint flexibilities are expressed in terms ofnon-dimensional local joint flexibilities, f, which are alsoknown as non-dimensional influence factors, as follows

in which E denotes Young’s modulus of elasticity, D is theouter chord diameter, IPB denotes in-plane bending, and OPBdenotes out-of-plane bending. Expressions for faxial, fIPB andfOPB are given in the following for various types of joints.105 Implementation of LJFs in conventional frame analysismodels requires springs, whose spring stiffnesses are equal tothe inverse of the local joint flexibilities, to be includedbetween the brace end and the corresponding point on thechord surface. Alternatively, a short flexible beam element canbe included between the brace end and the chord at the chordsurface.106 LJFs are given separately for different joint types. How-ever, note that for multi-brace joints, such as X and K joints,the LJFs are dependent on the load pattern. This implies thatfor a given load case, the joint should be classified by the loads

or the load pattern, rather than by its actual geometry. This fur-ther implies that a multi-brace joint may be classified as a dif-ferent joint type than the one which is given by its geometry,or it may be classified as a combination of joint types. In theformer case, its LJFs shall be calculated according to the for-mulae given for the joint type to which the joint has becomeclassified. In the latter case, its LJFs shall be calculated as

in which the λ values are the fractions corresponding to thejoint type designated by the subscript when the joint is classi-fied by loads.107 It is important to include LJFs not only in joints whichare being analysed, but also in joints which influence the forcedistribution at the joints which are being analysed. 108 The expressions for LJFs are developed for planar joints.For fatigue assessments in a traditionally braced jacket structure,the expressions can be applied to multi-planar joints as well, aslong as these joints are un-stiffened and non-overlapping. 109 According to the above, the following steps should thusbe included in a global analysis of a wind turbine support struc-ture, based on a conventional frame analysis model of beamelements:

1) Classification of joints (T/Y/X/XT joints) by load pattern,i.e. not by geometry.

2) Implementation of local joint flexibility in all jointsaccording to classification and parametric expressions byBuitrago.

EDf

LJF axialaxial =

3ED

fLJF IPB

IPB =

3ED

fLJF OPB

OPB =

KKXXYY LJFLJFLJFLJF λλλ ++=

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.B see note on front cover

3) Calculation of sectional forces at the surface footprint ofthe brace-to-chord connection.

110 The parametric expressions for calculation of LJFs fortubular joints are given in the following.Table A1 Non-dimensional influence factor expressions forlocal joint flexibility of single-brace joints

Table A2 Non-dimensional influence factor expressions forlocal joint flexibility of X joints

Table A3 Non-dimensional influence factor expressions for local joint flexibility of K joints

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.C –

APPENDIX C STRESS CONCENTRATION FACTORS FOR GIRTH WELDS

A. Calculation of Stress Concentration Factors for Hot Spots

A 100 General101 Stress concentration factors (SCFs) for hot spot stressesin tube-to-tube girth welds can be calculated by means of oneof the equations given in Table A1.

Table A1 SCF expressions for girth welds102 Equation A is for the SCF between two plates of equalthickness and will always yield conservative results whenapplied to girth welds including girth welds with differences inwall thickness. Equation B is an extension of Equation A,

accounting for differences in wall thickness.103 Distinction is to be made between design misalignmentsδ (e.g. thickness step) and misalignments from manufacturingtolerances x (e.g. due to out-of-roundness).104 The SCF due to design misalignment δ is always to betaken into account. If the manufacturing misalignment x islarger than 10% of the smaller thickness, the fraction exceed-ing 10% of this thickness shall be included when the wall cen-tre line offset e is calculated. This implies that the wall centreline offset shall be calculated as

Misalignment from manufacturing tolerances x below 0.1T1 iscovered by detail categories; thus no further SCF is to be takeninto account.105 Manufacturing tolerances for the local wall centre linemisalignment are hence to be included in the determination ofthe SCF. If the location and magnitude of the fabrication mis-alignments are unknown, i.e. they are not measured; the toler-ances are to be applied in the direction that gives the highestSCF. The maximum fabrication tolerances given in Figure 1can in general be applied.106 However, it should be noted that if very strict fabricationtolerances are secured, the tolerances will be less than the tol-erances given in Figure 1.

Figure 1 Fabrication tolerances for tube-to-tube girth welds. T1 is the smallest wall thickness of the adjoining tubes

Low

High

Deg

ree

of C

onse

rvat

ism

Equation ID Equation Nomenclature

A

B

T: Member thickness

T1 ≤ T2

e: Wall centre line offset between Tube 1 and Tube 2

⎟⎟⎟⎟⎟

⎠

⎞

⎜⎜⎜⎜⎜

⎝

⎛

⎟⎟⎠

⎞⎜⎜⎝

⎛+

+= 5.1

1

211

161

TTT

eSCF

1

31TeSCF += )1.0( 1Txe −+= δ

Single Sided Full Penetration Welds Double Sided Full Penetration Welds

efab

efab

efab

efab

⎩⎨⎧

=12.0

3min

Tmm

ofe fab⎩⎨⎧

=12.0

6min

Tmm

ofe fab

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.D see note on front cover

APPENDIX D STRESS EXTRAPOLATION FOR WELDS

A. Stress Extrapolation to Determine Hot Spot Stresses

A 100 General101 Since stress singularities are present at weld roots andweld toes, stress extrapolation is required to determine hot spot

stresses at welds. Figure 1 illustrates how the stress distribu-tion over a plate or tube wall thickness varies between zones ofdifferent proximity to a weld. In the notch stress zone, thestress at the weld approaches infinity. The stresses in the geo-metric stress zone are used as a basis for extrapolation to findthe hot spot stress at the weld.

Figure 1 Definition of stresses in welded structures. The three lower drawings show how the distribution of stresses through the thickness of aplate or tube wall varies in different stress zones

102 For welds in tubular joints, the hot spot stress is foundby linear extrapolation as defined in Figure 2.

Figure 2 Definition of the geometric stress zone in tubular joints. The hot spot stress is calculated by a linear extrapolation of the stresses in thegeometric stress zone to the weld toe

Stress distribution along surface normal to weld

Stress Distributions through the thickness of the plate/tube wall Notch Stress Zone Geometric stress zone Nominal Stress Zone

Section A-A Section B-B Section C-C

Notch stress

Geometric stress Nominal stress

A B C

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.D –

103 For welds in plate structures and for girth welds in tubu-lar sections, the hot spot stress is found by linear extrapolationas defined in Figure 3.104 For determination of hot spot stresses by finite elementanalysis, the notch stress as resulting from the analysis shall beexcluded and the hot spot stress shall be calculated by extrap-olation from the geometric stresses. The stress concentrationfactor shall be calculated on the basis of the extrapolated geo-metric stresses. The definition of the hot spot location (weld

toe or weld root singularity) for stress extrapolation is given inFigure 4 for different modelling approaches in the finite ele-ment analysis.105 Stress extrapolations, which are based on finite elementanalyses, shall be based on surface stresses, i.e. not the midlinestress from shell models. The most correct stress to use is thenormal-to-weld stress. Unless otherwise agreed, the surfacestress that is used should be based on averaged nodal stresses.

Figure 3 Stress extrapolation positions for plate structures and girth welds Distances are measured from the notch, i.e. typically the weld toe or the weld root. The positions 0.4 T/1.0 T are recommended in IIW94, whilethe positions 0.5 T/1.5 T are recommended by NORSOK.

Figure 4 Location of weld singularity for hot spot stress extrapolation dependent on element types used in tubular joint FE models The grey arrows define the primary positions to be used as the location of the weld singularity when the stress extrapolation is to be carried out.The light grey arrow pointing at the imaginary surface intersection in shell models defines an alternative location, which may be adopted forshell models if it can be justified. The locations marked by the dark arrows, i.e. “imaginary weld toes” in FE models where the weld is notmodelled, may not be used as the location of the weld singularity when the stress extrapolation is to be carried out.

1.0T (1.5T)

0.4T (0.5T)

Normal to Weld Stress

SCF

T

1.0T (1.5T)

0.4T (1.5T)

SCF

Stress singularity at weld root

Stress singularity at weld toe

Normal to Weld Stress

Solid elements with weld profile modelled

Solid elements without weld profile

Shell elements (no weld profile included)

Extrapolation to: Extrapolation to: Extrapolation to: Weld toe Intersection of surfaces Midline intersection

Solid element model with weld profile

Solid element model without weld profile

Shell element model (no weld modelled)

”OK” ”OK””OK”

”Maybe”

”No” ”No”

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.E see note on front cover

APPENDIX E TUBULAR CONNECTIONS – FRACTURE MECHANICS ANALYSES AND CAL-

CULATIONS

A. Stress Concentrations at Tubular Joints

A 100 General101 High stress concentrations normally exist at the weld toeof tubular joints. The stresses may be divided into three typesas shown schematically in Figure 1:

1) The geometric stress which depends on the structuralgeometry of the joint

2) The notch stress, which depends on the local geometryconfiguration of the brace-weld-chord connection

3) The local stress at the weld toe due to the geometry of theweld bead

Figure 1 Definition of stresses at Tubular Joint

The geometric stress can be defined by a linear extrapolationof two stresses to the weld toe of the joint, see also AppendixD for definition of stress extrapolation points. Since the hotspot stress is defined by extrapolating the stresses at points Aand B in Figure 1, it is a rather arbitrary value and it will notrepresent the actual stress condition at the weld toes. However,the hot spot stress is a useful parameter and it is normally usedfor both fatigue design and for comparisons with test data fortubular joints.The notch stress can be defined as the locally raised stressbetween point B and the weld toe.The local stress at the weld toe depends on the local geometryof the weld bead, but it is independent of the joint geometry.The local stress at the weld toe quickly decays and may onlybe influential up to about 2 to 3 mm in depth. The local stress concentration due to the local geometry of theweld bead may be taken into account in fracture mechanics cal-culations using the geometry correction factor, FG, which isgiven in C200.

B. Stresses at Tubular Joints

B 100 General101 Figure 2 shows a schematic view of the stresses whichmay be expected to be present at a tubular joint.

Figure 2 Schematic view of:a) Stresses due to global bending moment at the joint.b) Nominal tensile or compressive stresses.c) Stresses due to local plate bending in chord member/wall.

In Figure 2a, the stresses due to the global bending moment atthe joint are shown. These stresses can be computed by apply-ing simple beam theory. The stresses may be assumed constantthrough the thickness of the chord wall, where the fatiguecrack penetrates.102 When a load is applied at the top of the brace, a part ofthe chord wall is pulled up or pushed down to accommodatethe deformation of the brace, see Figure 2b. It may be notedthat the centre of rotation of the brace is at the intersectionbetween the centre line of the brace and the line A-B, see Fig-ure 2a. The deformation of brace results in tensile or compres-sive membrane stresses in the chord wall. Tensile membranestresses arise at side A when the load acts in the direction indi-cated by –P in Figure 2a. 103 As illustrated in Figure 2c, the chord wall further

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.E –

deforms and local bending stresses arise in the chord wall.Typically a high percentage of the total stresses in the hot spotareas are due to this local plate bending. Hence, the degree-of-bending parameter, DoB, defined as the ratio between thebending stress and the total stress at the outer side of the chordwall, is typically 70 to 80% for tubular joints.

C. Stress Intensity FactorC 100 General101 The stress intensity factor for a semi-elliptical surfacecrack subjected to tensile membrane stress, Sm and bendingstress, Sb can be expressed by the following semi-empiricalequation,

K = (E.1) Sm = tensile membrane stress componentSb = outer-fibre bending stress componentc = crack depthF = correction factor depending on structural geometry,

crack size and shape, proximity of the crack tip to freesurfaces and the type of loading. Subscript “m” refersto membrane and the subscript “b” refers to bending.

It should be emphasised, that the expression in eqn. (E.1) wasderived for statically determinate flat plate configurations. Inthe case of tubular joints, which contain some degree of redun-dancy, the cracked section may transfer significantly lowerload as a consequence of the load shedding from the crackedsection to less stressed parts of the joint, see C500.

C 200 Correction factor for membrane stress compo-nent201 An approximate method for calculation of the stressintensity factor for a semi-elliptical crack in a welded struc-tural detail is outlined in the following. Reference is made toFigure 3.

Figure 3 Schematic of semi-elliptical surface crack growing from weld toe

202 Separating the stress intensity factor into a finite numberof dimensionless stress intensity factor corrections, the stressintensity factor K can be expressed as follows:K = FS · FE · FT · FG · S · (E.2)where FS is the (front) free surface correction factor, FE is theelliptical crack shape correction factor, FT is the finite plate thick-ness correction factor (or finite width correction factor), FG is thestress gradient or geometry correction factor, S is the external,remote applied stress and c is the physical crack length.

For a semi-elliptical crack emanating from the weld toe, seeFigure 3, the following correction factors can be applied toexpress the stress intensity factor corrections in eqn. (E.2).

FS = 1.12 – 0.12 · (E.3)andFT = (E.4)where t is the thickness of the specimen.

Figure 4 Semi-elliptical crack

The elliptical crack shape correction factor, FE is given by

FE = (E.5)

in which the symbols used are explained in Figure 4. The value of FE is largest where the minor axis intersects thecrack front (point A in Figs. 3 and 4). At this point ϕ = π/2 andeqn. (E.5) reduces to

FE = (E.6)

The value of EK in eqs. (X.5 and X.6) is the complete ellipticalintegral of the second kind. i.e.

EK = (E.7)

which depends only upon the semi-axis ratio, c/b.The value of the elliptical integral varies from EK = π/2 for thecircular crack, c/b = 1, to a value of EK = 1.0 for the tunnelcrack, as the semi-axis ratio, c/b, approaches zero.A good approximation to eqn. (E.6) is obtained through theexpression:

FE = (E.8)

which also pertains to point A in Figs. 3 and 4.The geometry correction factor, FG, can be calculated applyingthe following formula:

FG = (E.9)

where σ(x) is the stress distribution in the un-cracked body atthe line of potential crack growth due to a unit remote appliedstress, and c is the physical crack length. σ(x) may, for exam-ple, be determined by a finite element calculation.If only a finite number of stress values, σi (i = 1, 2, ….,n), areknown, the following equation may be used instead of eqn. (E.9)

( ) c ·S · F S · bbm π+mF

cπ

bc

( )tc 2/sec π

4/12

22 cossin1

2 ⎪⎭

⎪⎬⎫

⎪⎩

⎪⎨⎧

+ ϕϕbc

EK

KE1

( )∫ ⎥

⎥⎦

⎤

⎢⎢⎣

⎡ −− θθ

π

ο

db

cb2/1

22

222sin1

2/165.1

25945.41

−

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎠⎞

⎜⎝⎛+

bc

( ) dxxc

xc

22

2

−°∫

σπ

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.E see note on front cover

FG = (E.10)

where (ai+1 – ai) is the width of stress element i carrying thestress σi and j is the number of discrete stress elements from thecentre of the crack to the physical crack tip, see Figure 5.

Figure 5 Crack subjected to pairs of discrete stresses

Since the fatigue crack in a tubular connection or joint will ini-tiate at the weld toe as a semi-elliptical crack and finally prop-agate through the thickness of the chord wall, the same stressintensity factor as given in eqn. (E.1) can be applied:

(E.11)whereFS = 1.12 – 0.12 · (E.12)

FE = (E.13)

FT = (E.14)FGm = Geometry correction factor for the membrane stresscomponent to be calculated according to eqn. (E.9) or eqn.(E.10).In eqn. (E.14), t denotes the wall thickness.

C 300 Correction factor for bending stress componentAt the deepest point of the crack front of a semi-elliptical sur-face crack, the stress intensity factor correction for the bendingstress component can be determined as

Fb = (E.15)

whereH = 1 + G1 · (E.16)

G1 = – 1.22 – 0.12 · (E.17)

G2 = 0.55 – 1.05 (E.18)

for

In general, the geometry correction factor FGb for the bending

stress component is different from FGm for the membranestress component. FGb can be calculated from the results of afinite element analysis applying eqn. (E.9) or eqn. (E.10).In Figure 6, the parameter H (which is equal to the ratiobetween the stress intensities for bending and membrane stresscomponents, if FGb = FGm) is plotted against the relative crackdepth c/t.

Figure 6 Ratio between stress intensity factors for bending and membranestress components

It appears from Figure 6 that the reduction in H for increasingcrack depth is largest for the semi-circular surface crack (c/b = 1). In Figure 6 it may also be seen that for high values ofthe semi-axis ratio c/b and large relative crack depths, theparameter H becomes negative and thus the bending effectmay lead to a reduction in the total stress intensity and henceto lower crack growth rates.

C 400 Crack shape and initial crack size401 Fatigue cracks at the weld toe in tubular joints appear tobe very slender with semi-axis ratios, c/b less than ~ 0.2. Thelow aspect ratios for cracks in tubular joints are mainly due tocrack coalescence.Therefore, for a semi-axis ratio of c/b = 0.2the initial crack size can be chosen asci = 0.1 mm

C 500 Load Shedding501 The stress distribution through a tubular joint is stronglyaffected by the presence of a crack. As a crack is growingthrough the hot spot region, the load is redistributed to lessstressed parts of the joint – the load shedding effect.502 A simplified model can be applied to model load shed-ding. By a hinge analogy the membrane stress component inthe cracked section can be assumed to be unaffected by thecrack, whereas the bending stress component is allowed todecrease linearly with crack depth according to the expression:

Sb = S (E.19)

where Sbo is the bending stress component of the hot spot stress

at the outer side of the chord wall in the un-cracked state. It maybe noted that eqn. (E.19) has been implemented in a fracturemechanics code for crack growth analysis in weld geometries.

( )njca

ca ii

j

ii ,...2,1,arcsinarcsin2 1

1=⎥⎦

⎤⎢⎣⎡ −

+

=∑σ

π

FM FS FE FT FGm⋅ ⋅ ⋅=

bc

2/165.1

25945.41

−

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎠⎞

⎜⎝⎛+

bc

( )tc 2/sec π

mF · H ·m

b

G

G

FF

2

2 tc · ⎥⎦

⎤⎢⎣⎡+⎥⎦

⎤⎢⎣⎡ G

tc

bc

5.175.047.0 ⎥⎦

⎤⎢⎣⎡+⎥⎦

⎤⎢⎣⎡

bc

bc

.1≤bc

( ) front.crack ofpoint Deepest ·mb GG FF =

⎟⎠⎞

⎜⎝⎛ −

tc

b 1ο

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.E –

C 600 Crack Growth 601 The crack growth can be calculated using the followingrelation

= C (ΔK m eff – ΔK m

eff,th), for ΔKeff ≥ ΔK eff,th (E.20)

= 0, for ΔKeff < ΔK eff,th

For the fracture mechanics calculations the crack growth coef-ficients given in Table C1 can be applied:

ΔK eff,th = 79.1 MPa (valid in air as well as in seawaterwith/without corrosion protection)602 The fatigue life can then be calculated by applying themethod outlined above and using eqn. (E.20). For determinis-tic fatigue life calculations, the data tabulated for the mean + 2standard deviations of logC are to be applied. For probabilistic

fatigue life calculations, the data tabulated for the mean valueof logC are to be applied. The fatigue life is calculated basedon the through thickness crack criterion for the final crack sizecf, i.e. cf ~ t, where t is the wall thickness.603 Reference is made to BS 7910 for an alternative methodfor fracture mechanics analyses and calculations.

dndc

dndc

Table C1 Crack growth coefficients

m

log C

Mean value

(μlogC)

Mean value + 2 standard

deviations(μlogC+2σlogC)

Value corresponding to mean value of logC

( )

Value corresponding to mean + 2 st.dev. of logC

( )Welds in air and in seawater with adequate corrosion protec-tion

3.1 –12.96 –12.48 1.1 · 10–13 3.3 · 10–13

Welds subjected to seawater without corrosion protection 3.5 –13.47 –12.80 3.4 · 10–14 1.6 · 10–13

Here, μlogC denotes the mean value of logC, and σlogC denotes the standard deviation of logC.

( ) ⎥⎥

⎦

⎤

⎢⎢

⎣

⎡

mmmMPa

mmC

Clog10μ CC loglog 210 σμ +

mm

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.F see note on front cover

APPENDIX F PILE RESISTANCE AND LOAD-DISPLACEMENT RELATIONSHIPS

A. Axial Pile ResistanceA 100 General101 Axial pile resistance is composed of two parts

— accumulated skin resistance— tip resistance.

For a pile in a stratified soil deposit of N soil layers, the pileresistance R can be expressed as

where fSi is the average unit skin friction along the pile shaft inlayer i, ASi is the shaft area of the pile in layer i, qT is the unitend resistance, and AT is the gross tip area of the pile.

A 200 Clay201 For piles in mainly cohesive soils, the average unit skinfriction fS may be calculated according to (1) total stress methods, e.g. the α method, which yieldsfsi = α suin which

where su is the undrained shear strength of the soil and p0’ isthe effective overburden pressure at the point in question.(2) effective stress methods, e.g. the β method, which yields

in which β values in the range 0.10 to 0.25 are suggested forpile lengths exceeding 15 m.(3) semi-empirical λ method, by which the soil deposit is takenas one single layer, for which the average skin friction is cal-culated as

where p0m’ is the average effective overburden pressurebetween the pile head and the pile tip, sum is the average un-drained shear strength along the pile shaft, and λ is the dimen-sionless coefficient, which depends on the pile length as shownin Figure 1. Hence, by this method, the total shaft resistancebecomes RS = fSAS, where AS is the pile shaft area.For long flexible piles, failure between pile and soil may occurclose to the seabed even before the soil resistance near the piletip has been mobilized at all. This is a result of the flexibilityof the pile and the associated differences in relative pile-soildisplacement along the length of the pile. This is a lengtheffect, which for a strain-softening soil will imply that the

static capacity of the pile will be less than that of a rigid pile.

Figure 1 Coefficient λ vs. pile length

For deformation and stress analysis of an axially loaded flexi-ble pile, the pile can be modelled as a number of consecutivecolumn elements supported by nonlinear springs applied at thenodal points between the elements. The nonlinear springs aredenoted t-z curves and represent the axial load-displacementrelationship between the pile and the soil. The stress t is theaxial skin friction per unit area of pile surface and z is the rel-ative axial pile-soil displacement necessary to mobilize thisskin friction.

A 300 Sand301 For piles in mainly cohesionless soils (sand), the averageunit skin friction may be calculated according to

fS = Kp0’tanδ≤flin which K = 0.8 for open-ended piles and K = 1.0 for closed-ended piles, p0’ is the effective overburden pressure, δ is theangle of soil friction on the pile wall as given in Table A1, andfl is a limiting unit skin friction, see Table A1 for guidance.The unit tip resistance of plugged piles in cohesionless soilscan be calculated as

qp = Nqp0’≤qlin which the bearing factor Nq can be taken from Table A1 andql is a limiting tip resistance, see Table A1 for guidance.The unit tip resistance of piles in cohesive soils can be calcu-lated as

qp = Ncsuwhere Nc = 9 and su is the undrained shear strength of the soilat the pile tip.

∑=

+=+=N

iTTSiSiTS AqAfRRR

1

⎪⎪⎩

⎪⎪⎨

⎧

>

≤

=0.1

'for

'21

0.1'

for '2

1

04 0

00

ps

ps

ps

psu

u

u

uα

'0pfSi β=

)2'( 0 ummS spf += λ

0

0,05

0,1

0,15

0,2

0,25

0,3

0,35

0,4

0,45

0,5

0 10 20 30 40 50 60 70

Pile length (m)

λ

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.F –

A 400 t-z curves401 The t-z curves can be generated according to a methodby which a nonlinear relation applies between the origin andthe point where the maximum skin resistance tmax is reached,

in which R denotes the radius of the pile, G0 is the initial shearmodulus of the soil, zIF is a dimensionless zone of influence,defined as the radius of the zone of influence around the piledivided by R, and rf is a curve fitting factor. For displacementsz beyond the displacement where tmax is reached, the skinresistance t decreases in linear manner with z until a residualskin resistance tres is reached. For further displacementsbeyond this point, the skin resistance t stays constant. Anexample of t-z curves generated according to this method isgiven in Figure 2. The maximum skin resistance can be calcu-lated according to one of the methods for prediction of unitskin friction given above.

Figure 2 Example of t-z curves generated by model

For clays, the initial shear modulus of the soil to be used forgeneration of t-z curves can be taken as

G0 = 2600 cuHowever, Eide and Andersen (1984) suggest a somewhatsofter value according to the formula

where su is the undrained shear strength of the clay, and OCRis the overconsolidation ratio. For sands, the initial shear mod-ulus of the soil to be used for generation of t-z curves is to betaken as

in which σa = 100 kPa is a reference pressure and σv is the ver-tical effective stress, ν is the Poisson’s ratio of the soil, and φis the friction angle of the soil.

B. Laterally Loaded PilesB 100 General101 The most common method for analysis of laterallyloaded piles is based on the use of so-called p-y curves. The p-y curves give the relation between the integral value p of themobilized resistance from the surrounding soil when the piledeflects a distance y laterally. The pile is modelled as a numberof consecutive beam-column elements, supported by nonlinearsprings applied at the nodal points between the elements. Thenonlinear support springs are characterized by one p-y curve ateach nodal point, see Figure 3.The solution of pile displacements and pile stresses in anypoint along the pile for any applied load at the pile head resultsas the solution to the differential equation of the pile

with

where x denotes the position along the pile axis, y is the lateraldisplacement of the pile, EI is the flexural rigidity of the pile,QA is the axial force in the pile, QL is the lateral force in thepile, p(y) is the lateral soil reaction, q is a distributed load alongthe pile, and M is the bending moment in the pile, all at theposition x.

Table A1 Design parameters for axial resistance of driven piles in cohesionless silicious soil 1)

Density Soil description

δ(degrees)

f1(kPa)

Nq(—)

q1(MPa)

Very looseLooseMedium

SandSand-silt 2)Silt

15 48 8 1.9

LooseMediumDense

SandSand-silt 2)Silt

20 67 12 2.9

MediumDense

SandSand-silt 2) 25 81 20 4.8

DenseVery dense

SandSand-silt 2) 30 96 40 9.6

DenseVery dense

GravelSand 35 115 50 12.0

1) The parameters listed in this table are intended as guidelines only. Where detailed information such as in-situ cone penetrometer tests, strength tests on high quality soil samples, model tests or pile driving performance is available, other values may be justified.

2) Sand-silt includes those soils with significant fractions of both sand and silt. Strength values generally increase with increasing sand fractions and decrease with increasing silt fractions.

max

max

max

00for

1ln tt

ttr

ttrz

GRtz

f

fIF≤≤

−

−=

11706000 −−= OCRccG uu

φνσσ

1000tanwith )1(20 =

+= m

mG va

0)(2

2

4

4=+−+ qyp

dxydQ

dxydEI A

Mdx

ydEIQdxdyQ

dxydEI LA ==+ 2

2

3

3 and

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.F see note on front cover

Figure 3 p-y curves applied at nodal points in beam-column representationof pile

102 A finite difference method usually forms the most feasi-ble approach to achieve the sought-after solution of the differ-ential equation of the pile. A number of commercial computerprograms are available for this purpose. These programs usu-ally provide full solutions of pile stresses and displacementsfor a combination of axial force, lateral force and bendingmoment at the pile head, i.e., also the gradual transfer of axialload to the soil along the pile according to the t-z curveapproach presented above is included. Some of the availableprograms can be used to analyse not only single piles but alsopile groups, including possible pile-soil-pile interaction andallowing for proper representation of a superstructure attachedat the pile heads, either as a rigid cap or as a structure of finitestiffness.For construction of p-y curves, the type of soil, the type ofloading, the remoulding due to pile installation and the effectof scour should be considered. A recommended method forconstruction of p-y curves is presented in the following:The lateral resistance per unit length of pile for a lateral piledeflection y is denoted p. The static ultimate lateral resistanceper unit length is denoted pu. This is the maximum value thatp can take on when the pile is deflected laterally.

B 200 Clay201 For piles in cohesive soils, the static ultimate lateralresistance is recommended to be calculated as

where X is the depth below soil surface and XR is a transitiondepth, below which the value of (3su+γ’X)D+JsuX exceeds9suD. Further, D is the pile diameter, su is the undrained shearstrength of the soil, γ’ is the effective unit weight of soil, and Jis a dimensionless empirical constant whose value is in therange 0.25 to 0.50 with 0.50 recommended for soft normallyconsolidated clay.For static loading, the p-y curve can be generated according to

For cyclic loading and X > XR, the p-y curve can be generatedaccording to

For cyclic loading and X≤XR, the p-y curve can be generatedaccording to

Here, yc = 2.5εcD, in which D is the pile diameter and εc is thestrain which occurs at one-half the maximum stress in labora-tory undrained compression tests of undisturbed soil samples.For further details, reference is made to Classification NotesNo. 30.4.

B 300 Sand301 For piles in cohesionless soils, the static ultimate lateralresistance is recommended to be calculated as

where the coefficients C1, C2 and C3 depend on the frictionangle φ as shown in Figure 4, and where X is the depth belowsoil surface and XR is a transition depth, below which the valueof (C1X+C2D)γ’X exceeds C3Dγ’X. Further, D is the pilediameter, and γ’ is the submerged unit weight of soil.The p-y curve can be generated according to

Figure 4 Coefficients as functions of friction angle

⎩⎨⎧

>≤<++

=R

RXXfor 9

XX0for )'3(Ds

XJsDXsp

u

uuu

γ

⎪⎩

⎪⎨⎧

>

≤=

c

c3/1

8yyfor

8yyfor )(2u

c

u

pyyp

p

⎪⎩

⎪⎨⎧

>

≤=

c

c3/1

3yyfor 72.0

3yyfor )(2

uc

u

pyyp

p

⎪⎪⎪

⎩

⎪⎪⎪

⎨

⎧

>

≤<−

−−

≤

=

cR

u

cc

c

Ru

c

u

yXXp

yyy

yyXXp

yyp

p

15yfor 72.0

153yfor )12

3)1(1(72.0

3yyfor )(2

c

c3/1

⎩⎨⎧

>≤<+

=R3

R21XXfor '

XX0for ')(XDC

XDCXCpu γ

γ

)tanh( yApkXApp

uu=

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.F –

in which k is the initial modulus of subgrade reaction anddepends on the friction angle φ as given in Figure 5, and A is afactor to account for static or cyclic loading conditions as fol-lows

For further details, reference is made to Classification NotesNo. 30.4.

Figure 5 Initial modulus of subgrade reaction k as function of friction angle φ

B 400 Application of p-y curves401 The recommended nonlinear p-y curves are meant pri-marily for analysis of piles for evaluation of lateral pile capac-ity in the ULS. 402 Caution must be exercised when the recommended non-linear p-y curves are used in other contexts than for evaluationof lateral pile capacity in the ULS. Such contexts include, butare not limited to, SLS analysis of the pile, fatigue analysis ofthe pile, determination of equivalent spring stiffnesses to rep-resent the stiffness of the pile-soil system as boundary condi-tion in analyses of the structure that the pile-soil systemsupports, and in general all cases where the initial slope of thep-y curves may have an impact.

403 Caution must be exercised regardless of whether the rec-ommended nonlinear p-y curves are applied directly as theyare specified on closed form or whether piece-wise linearapproximations according to some discretisation of the curvesare applied.404 The p-y curves that are recommended for clay aredefined as 3rd order polynomials such that they have infiniteinitial slopes, i.e. the initial stiffnesses of the load-displace-ment relationships are infinite. This is unphysical; however,the curves are still valid for use for their primary purpose, viz.evaluation of lateral pile capacity in the ULS. However, theclosed-form p-y curves that are recommended for clay cannotbe used directly in cases where the initial stiffness matters,such as for determination of equivalent pile head stiffnesses. 405 When a p-y curve for clay is to be used in contexts wherethe initial slope of the curve matters, the curve need to be dis-cretised and approximated by a piece-wise linear curve drawnbetween the discretisation points. The discretisation must becarried out in such a manner that the first discretisation pointof the curve beyond the origin is localised such that a correctinitial slope results in the piece-wise linear representation ofthe p-y curve.406 Unless data indicate otherwise, the true initial slope of ap-y curve in clay may be calculated as

where ξ is an empirical coefficient and εc is the vertical strainat one-half the maximum principal stress difference in a staticundrained triaxial compression tests on an undisturbed soilsample. For normally consolidated clay ξ = 10 is recom-mended, and for over-consolidated clay ξ = 30 is recom-mended. 407 As an alternative to localise the first discretisation pointbeyond the origin such that a correct initial slope results in thepiece-wise linear approximation of the p-y curve for clay, thefirst discretisation point beyond the origin may be localised at therelative displacement y/yc = 0.1 with ordinate value p/pu = 0.23. 408 The recommended closed form p-y curves for sand havefinite initial slopes and thus final initial stiffnesses. Wheneverdiscretised approximations to these curves are needed in anal-yses with piece-wise linear curves drawn through the discreti-sation points, it is important to impose a sufficiently finediscretisation near the origin of the p-y curves in order to get acorrect representation of the initial slopes. 409 Whenever p-y curves are used to establish equivalentpile head stiffnesses to be applied as boundary conditions foranalysis of structures supported by a pile-soil system, it is rec-ommended that a sensitivity study be carried out to investigatethe effect of changes in or different assumptions for the initialslopes of the p-y curves.

⎪⎩

⎪⎨⎧

≥−= loading staticfor 9.0)8.03(

loading cyclicfor 9.0

DXA

25.0)( c

u

Dp

kε

ξ⋅

⋅=

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.G see note on front cover

APPENDIX G BEARING CAPACITY FORMULAE FOR GRAVITY BASE FOUNDATIONS

A. Forces

A 100 General101 All forces acting on the foundation, including forcestransferred from the wind turbine, are transferred to the foun-dation base and combined into resultant forces H and V in thehorizontal and vertical direction, respectively, at the founda-tion-soil interface.

Figure 1 Loading under idealised conditions

In the following, it is assumed that H and V are design forces,i.e., they are characteristic forces that have been multiplied bytheir relevant partial load factor γf. This is indicated by index din the bearing capacity formulae, hence Hd and Vd. The loadcentre, denoted LC, is the point where the resultant of H and Vintersects the foundation-soil interface, and implies an eccen-tricity e of the vertical force V relative to the centre line of thefoundation. Reference is made to Figure 1, and the eccentricityis calculated as

where Md denotes the resulting design overturning momentabout the foundation-soil interface.

B. Correction for Torque

B 100 General101 When a torque MZ is applied to the foundation in addi-tion to the forces H and V, the interaction between the torqueand these forces can be accounted for by replacing H and MZwith an equivalent horizontal force H’. The bearing capacity ofthe foundation is then to be evaluated for the force set (H’,V)instead of the force set (H,V). The equivalent horizontal forcecan be calculated as

in which leff is the length of the effective area as determined inC100.

C. Effective Foundation AreaC 100 General101 For use in bearing capacity analysis an effective founda-tion area Aeff is needed. The effective foundation area is con-structed such that its geometrical centre coincides with the loadcentre, and such that it follows as closely as possible the near-est contour of the true area of the foundation base. For a quad-ratic area of width b, the effective area Aeff can be defined as

in which the effective dimensions beff and leff depend on whichof two idealised loading scenarios leads to the most criticalbearing capacity for the actual foundation.

Figure 2 Quadratic footing with two approaches to how to make up the ef-fective foundation area

Scenario 1 corresponds to load eccentricity with respect to oneof the two symmetry axes of the foundation. By this scenario,the following effective dimensions are used:

Scenario 2 corresponds to load eccentricity with respect toboth symmetry axes of the foundation. By this scenario, thefollowing effective dimensions are used:

Reference is made to Figure 2. The effective area representa-tion that leads to the poorest or most critical result for the bear-ing capacity of the foundation is the effective arearepresentation to be chosen.

f [kN

/m2 ]

e [m]

V

rupture 2 rupture 1

LC

H

d

dVMe =

22 22' ⎟

⎟⎠

⎞⎜⎜⎝

⎛ ⋅++

⋅=

eff

z

eff

zlMH

lMH

effeffeff lbA ⋅=

Aeffe

l eff

beff

LC1

Aeff

el ef

f

beff

LC2

ebbeff ⋅−= 2 , bleff =

2eblb effeff −==

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.G –

For a circular foundation area with radius R, an elliptical effec-tive foundation area Aeff can be defined as

with major axes

and

Figure 3 Circular and octangular footings with effective foundation areamarked out

The effective foundation area Aeff can now be represented bya rectangle with the following dimensions

For an area shaped as a double symmetrical polygon (octago-nal or more), the above formulae for the circular foundationarea can be used provided that a radius equal to the radius ofthe inscribed circle of the polygon is used for the calculations.

D. Bearing CapacityD 100 General101 For fully drained conditions and failure according toRupture 1 as indicated in Figure 1, the following general for-mula can be applied for the bearing capacity of a foundationwith a horizontal base, resting on the soil surface:

For undrained conditions, which imply φ = 0, the followingformula for the bearing capacity applies:

The symbols used have the following explanations

qd design bearing capacity [kN/m2]γ' effective (submerged) unit weight of soil [kN/m3]p'0 effective overburden pressure at the level of the

foundation-soil interface [kN/m2]cd design cohesion or design undrained shear strength

assessed on the basis of the actual shear strengthprofile, load configuration and estimated depth ofpotential failure surface [kN/m2]

Nγ Nq Nc bearing capacity factors, dimensionlesssγ sq sc shape factors, dimensionlessiγ iq ic inclination factors, dimensionless102 In principle, the quoted formulae apply to foundations,which are not embedded. However, the formulae may also beapplied to embedded foundations, for which they will lead toresults, which will be on the conservative side. Alternatively,depth effects associated with embedded foundations can becalculated according to formulae given in DNV ClassificationNotes No. 30.4.The calculations are to be based on design shear strengthparameters:

The material factors γc an γφ must be those associated with theactual design code and the type of analysis, i.e. whetherdrained or undrained conditions apply. The dimensionless factors N, s and i can be determined bymeans of formulae given in the following.

D 200 Bearing capacity formulae for drained conditions201 Bearing capacity factors N:

When the bearing capacity formulae are used to predict soilreaction stresses on foundation structures for design of suchstructures, it is recommended that the factor Nγ is calculatedaccording to the following formula

Shape factors s:

Inclination factors i:

D 300 Bearing capacity formulae for undrained condi-tions, φ = 0

E. Extremely Eccentric LoadingE 100 General101 In the case of extremely eccentric loading, i.e., an eccen-tricity in excess of 0.3 times the foundation width, e > 0.3b, anadditional bearing capacity calculation needs to be carried out,

⎥⎦⎤

⎢⎣⎡ −−= 222 )arccos(2 eRe

ReRAeff

( )eRbe −= 2

le 2R 1 1be2R-------–⎝ ⎠

⎛ ⎞ 2–=

LC

Aeff

e

beff

l eff

be

l e

R

e

eeffeff b

lAl = and e

e

effeff b

ll

b =

cccdqqqeffd isNcisNpisNbq ++= '0'

21

γγγγ

0000 pisNcq cccudd +⋅⋅⋅=

cud

ccγ

= and ))tan(arctan(φγφφ =d

d

dq deN

φφφπ

sin1sin1tan

−+

⋅= ; dqc NN φcot)1( ⋅−= ; dqNN φγ tan)1(23

⋅−⋅=

dqNN φγ tan)1(2 ⋅+⋅=

eff

eff

lb

s ⋅−= 4.01γ ;eff

effcq l

bss ⋅+== 2.01

2

cot1 ⎟

⎟⎠

⎞⎜⎜⎝

⎛

⋅⋅+−==

ddeffd

dcq cAV

Hiiφ

; 2qii =γ

20 += πcN

cc ss =0

udeffc cA

Hi⋅

−⋅+= 15.05.00

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.G see note on front cover

corresponding to the possibility of a failure according to Rup-ture 2 in Figure 1. This failure mode involves failure of the soilalso under the unloaded part of the foundation area, i.e., underthe heel of the foundation. For this failure mode, the followingformula for the bearing capacity applies

with inclination factors

The bearing capacity is to be taken as the smallest of the valuesfor qd resulting from the calculations for Rupture 1 and Rup-ture 2.

F. Sliding ResistanceF 100 General101 Foundations subjected to horizontal loading must alsobe investigated for sufficient sliding resistance. The followingcriterion applies in the case of drained conditions:

For undrained conditions in clay, φ = 0, the following criterionapplies:

and it must in addition be verified that

)tan05.1(' 3 φγ γγγ ++= cccdeffd isNcisNbq

φcot1

⋅⋅++==

cAVHii

effcq ; 2

qii =γ ;udeff

c cAHi⋅

+⋅+= 15.05.00

φtan⋅+⋅< VcAH eff

udeff cAH ⋅<

4.0<VH

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.H –

APPENDIX H CROSS SECTION TYPES

A. Cross Section TypesA 100 General101 Cross sections of beams are divided into different typesdependent of their ability to develop plastic hinges as given inTable A1.

Figure 1 Relation between moment M and plastic moment resistance Mp,and rotation θ for cross sectional types. My is elastic moment re-sistance

102 The categorisation of cross sections depends on the pro-portions of each of its compression elements, see Table A3.103 Compression elements include every element of a crosssection which is either totally or partially in compression, due

to axial force or bending moment, under the load combinationconsidered.104 The various compression elements in a cross sectionsuch as web or flange, can be in different classes.105 The selection of cross sectional type is normally quotedby the highest or less favourable type of its compression ele-ments.

A 200 Cross section requirements for plastic analysis201 At plastic hinge locations, the cross section of the mem-ber which contains the plastic hinge shall have an axis of sym-metry in the plane of loading.202 At plastic hinge locations, the cross section of the mem-ber which contains the plastic hinge shall have a rotationcapacity not less than the required rotation at that plastic hingelocation.

A 300 Cross section requirements when elastic global analysis is used301 When elastic global analysis is used, the role of crosssection classification is to identify the extent to which theresistance of a cross section is limited by its local bucklingresistance.302 When all the compression elements of a cross section aretype III, its resistance may be based on an elastic distributionof stresses across the cross section, limited to the yield strengthat the extreme fibres.

Table A1 Cross sectional typesI Cross sections that can form a plastic hinge with the rotation

capacity required for plastic analysisII Cross sections that can develop their plastic moment resist-

ance, but have limited rotation capacityIII Cross sections where the calculated stress in the extreme

compression fibre of the steel member can reach its yield strength, but local buckling is liable to prevent development of the plastic moment resistance

IV Cross sections where it is necessary to make explicit allow-ances for the effects of local buckling when determining their moment resistance or compression resistance

Table A2 Coefficient related to relative strainNV Steel grade 1) ε 2)

NV-NS 1NV-27 0.94NV-32 0.86NV-36 0.81NV-40 0.78

NV-420 0.75NV-460 0.72NV-500 0.69NV-550 0.65NV-620 0.62NV-690 0.58

1) The table is not valid for steel with improved weldability. See Sec.6, Table A3, footnote 1).

2) e 235fy

--------- where fy is yield strength=

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.H see note on front cover

Table A3 Maximum width-to-thickness ratios for compression elementsCross section part Type I Type II Type III

d = h – 3 tw 3)

d / tw ≤ 33 ε 2) d / tw ≤ 38 ε d / tw≤ 42 ε

d / tw ≤ 72 ε d / tw ≤ 83 ε d / tw ≤ 124 ε

when α > 0.5:

when α ≤ 0.5:

when α > 0.5:

when α ≤ 0.5:

when ψ > –1:

when ψ ≤ –1:

Tip in compression Tip in compression Tip in compression

Tip in tension Tip in tension Tip in tension

d / tp ≤ 50 ε 2 d / tp ≤ 70 ε 2 d / tp ≤ 90 ε 2

1) Compression negative2) ε is defined in Table A23) Valid for rectangular hollow sections (RHS) where h is the height of the profile4) C is the buckling coefficient. See EN 1993-1-1 Table 5.3.3 (denoted kσ)5) Valid for axial and bending, not external pressure.

d tw⁄ 396ε13α 1–----------------≤

d tw⁄ 36εα

--------≤

d tw⁄ 456ε13α 1–-------------------≤

d tw⁄ 41.5εα

--------------≤

d tw⁄ 126ε2 ψ+-------------≤

d tw⁄ 62ε 1 ψ–( ) ψ≤

Rolled: c tf⁄ 10ε≤

Welded: c tf⁄ 9ε≤





Rolled: c tf⁄ 10ε α⁄≤

Welded: c tf⁄ 9ε α⁄≤Rolled: c tf⁄ 11ε α⁄≤

Welded: c tf⁄ 10ε α⁄≤

Rolled: c tf⁄ 23ε C 4 )

≤

Welded: c tf⁄ 21ε C≤

Rolled: c tf⁄ 10εα α------------≤

Welded: c tf⁄ 9εα α------------≤

Rolled: c tf⁄ 11εα α------------≤

Welded: c tf⁄ 10εα α------------≤

Rolled: c tf⁄ 23ε C≤

Welded: c tf⁄ 21ε C≤

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.I –

APPENDIX I EXTREME WIND SPEED EVENTS

Ref. Sec.3 B400.

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.J see note on front cover

APPENDIX J SCOUR AT A VERTICAL PILE

A. Flow around a Vertical PileA 100 General101 When a vertical pile is placed on a seabed, the water-par-ticle flow associated with currents and passing waves willundergo substantial changes, see Figure 1. First, a horseshoevortex will be formed at the base in front of the pile. Second, avortex flow pattern in the form of vortex shedding will beformed at the lee-side of the pile. Third, the streamlines willcontract at the side edges of the pile. This local change in theflow will increase the bed shear stress and the sediment trans-port capacity will increase accordingly. In the case of an erod-ible seabed, this may result in a local scour around the pile.Such scour is a threat to the stability of the pile.

Figure 1 Flow around the base of a vertical pile

B. Bed Shear StressB 100 General101 The increase in the bed shear stress can be expressed interms of the amplification factor α, which is defined by

(J.1)

in which τmax is the maximum value of the bed shear stress τwhen the pile structure is present and τmax,∞ is the maximumvalue of the bed shear stress τ∞ for the undisturbed flow. In thecase of a steady current, τmax and τmax,∞ are replaced by con-stant τ and τ∞, respectively, in the expression for α.102 In the case of a steady current, the amplification factorcan become as large as α = 7-11. This is due to the presence ofa very significant horseshoe vortex. For waves the amplifica-tion is smaller.

C. Local ScourC 100 General101 When local scour is analysed, it is important to distin-guish between clear-water scour and live-bed scour. This dis-

tinction is necessary because the development of a scour holewith time and the relationship between the scour depth and theapproach-flow velocity both depend on which of the two typesof scour is occurring.102 Under ‘clear water’ conditions, i.e. when the sedimentsfar from the pile are not in motion, a state of static equilibriumis reached when the scour hole has developed to an extent suchthat the flow no longer has the ability to resuspend sedimentand remove it from the scour hole. Under ‘live bed’ conditions,i.e. when the sediment transport prevails over the entire bed, astate of dynamic equilibrium is reached when the rate ofremoval of material from the scour hole is equal to the rate atwhich material is being deposited in the scour hole from ambi-ent suspended material and bed loads. 103 In the case of a steady current, the scour process ismainly caused by the presence of the horseshoe vortex com-bined with the effect of contraction of streamlines at the sideedges of the pile. The shape of the scour hole will virtually besymmetrical, see Figure 2.

Figure 2 Scour hole around a vertical pile

104 In the case of waves, the horseshoe vortex and the lee-wake vortex form the two processes that govern the scour.These two processes are primarily governed by the Keulegan-Carpenter number, KC, which is defined by

(J.2)

where T is the wave period, D is the cylinder diameter and umaxis the maximum value of the orbital velocity at the bed, givenby linear theory as:

(J.3)

Here H is the wave height, h is the water depth and k is thewave number which can be found by solving the dispersionequation:

(J.4)

where g denotes the acceleration of gravity, i.e. 9.81 m/s2.

C 200 Scour depth201 Unless data, e.g. from model tests, indicate otherwise,the following empirical expression for the equilibrium scourdepth S may be used:

x

y

z

U

δ Boundary layer

Lee-wake vortices

Contraction of streamlines

Boundary layer separation

Horseshoe vortex

∞=

max,

maxττ

α

Flow

Scour hole

D

Pile

DTuKC ⋅

= max

)sinh(max khTHu ⋅

=π

)tanh(2 2khkg

T⋅=⎟

⎠⎞

⎜⎝⎛ π

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.J –

(J.5)

This expression is valid for live-bed conditions, i.e. for θ >θcr,in which the Shields parameter θ is defined below togetherwith its critical threshold θcr. For steady current, which impliesKC→∞, it appears from this expression that S/D→1.3. Forwaves it appears that for KC < 6 no scour hole is formed. Thephysical explanation for this is that no horseshoe vortex devel-ops for KC < 6. The Shields parameter S is defined by:

(J.6)

where s is the specific gravity of the sediment, d is the graindiameter for the specific grain that will be eroded and Uf is thebed shear velocity. For practical purposes, d50 can be used ford, where d50 is defined as the median grain diameter in the par-ticle size distribution of the seabed material. The criticalShields parameter, θcr, is the value of θ at the initiation of sed-iment motion. The critical value θcr for the Shields parameteris about 0.05 to 0.06. Seabed erosion starts when the Shieldsparameter exceeds the critical value.For steady current the bed shear velocity, Uf, is given by theColebrook and White equation

(J.7)

where ν equal to 10–6 m2/s is the kinematic viscosity. Forwaves, the maximum value of the undisturbed bed shear veloc-ity is calculated by:

(J.8)

where fw is the frictional coefficient given by

(J.9)

Here, a is the free stream amplitude, defined by

(J.10)and kN is the bed roughness equal to 2.5·d50, where d50 denotesthe median grain diameter in the particle size distribution of theseabed material.

C 300 Lateral extension of scour hole301 The scour depth S is estimated by means of the empiricalexpression in eqn. (J.5), which is valid for live bed conditions.The lateral extension of the scour hole at the original level ofthe seabed can be estimated based on the friction angle ϕ of thesoil, and assuming that the slope of the scour hole equals thisfriction angle. By this approach, the radius of the scour hole,measured at the original level of the seabed from the centre ofa pile of diameter D, is estimated as

(J.11)

C 400 Time scale of scour401 The temporal evolution of the scour depth, S, can beexpressed as:

(J.12)in which t denotes the time, and T1 denotes the time scale ofthe scour process. The time scale T1 of the scour process canbe found from the non-dimensional time scale T* through thefollowing relationship

(J.13)

where T* is given by the empirical expressions:

for steady current (J.14)

for waves (J.15)

[ ]{ } 6)6(03.0exp13.1 ≥−−−⋅= KCKCDS

dsg

U f

)1(

2

−=θ

⎟⎟⎠

⎞⎜⎜⎝

⎛

⋅⋅

+⋅

⋅−=ff

c

Uhhd

UU ν7.45.2ln5.24.6

max2ufU w

f ⋅=

⎪⎩

⎪⎨⎧

<>

⋅⋅= −

−

100/100/

)/(4.0)/(04.0

75.0

25.0

N

N

N

Nw ka

ka

kakaf

π2max Tua ⋅

=

ϕtan2SDr +=

))/exp(1( 1TtSSt −−=

12

3* )1(

TD

dsgT

−=

2.2*2000

1 −= θDhT

36* 10 ⎟

⎠⎞

⎜⎝⎛= −

θKCT

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.K see note on front cover

APPENDIX K CALCULATIONS BY FINITE ELEMENT METHOD

A. IntroductionA 100 General101 If simple calculations cannot be performed to documentthe strength and stiffness of a structural component, a FiniteElement analysis should be carried out.102 The model to be included in the analysis and the type ofanalysis should be chosen with due consideration to the inter-action of the structural component with the rest of the struc-ture.103 Since a FEM analysis is normally used when simple cal-culations are insufficient or impossible, care must be taken toensure that the model and analysis reflect the physical reality.This must be done by means of carrying out an evaluation ofthe input to as well as the results from the analysis. Guidelinesfor such an evaluation are given below.

B. Types of AnalysisB 100 General101 Though different types of analyses can be performed bymeans of FEM analysis, most analyses take the form of staticanalyses for determination of the strength and stiffness ofstructures or structural components. FEM analyses are usuallycomputer-based analyses which make use of FEM computerprograms.

B 200 Static analysis201 In a static analysis, structural parts are commonlyexamined with respect to determining which extreme loadsgovern the extreme stress, strain and deflection responses. Asthe analysis is linear, unit loads can be applied, and theresponse caused by single loads can be calculated. The actualextreme load cases can subsequently be examined by means oflinear combinations – superposition.

B 300 Frequency analysis301 Frequency analysis is used to determine the eigenfre-quencies and normal modes of a structural part. 302 The FEM program will normally perform an analysis onthe basis of the lowest frequencies. However, by specifying ashift value, it is possible to obtain results also for a set of higherfrequencies around a user-defined frequency.

Guidance note:The normal modes resulting from a frequency analysis only rep-resent the shape of the deflection profiles, not the actual deflec-tions.


B 400 Dynamic analysis401 Dynamic FEM analysis can be used to determine thetime-dependent response of a structural part, e.g. as a transferfunction. The analysis is normally based on modal superposi-tion, as this type of analysis is much less time consuming thana ‘real’ time dependent analysis.

B 500 Stability/buckling analysis501 Stability/buckling analysis is relevant for slender struc-tural parts or sub-parts. This is due to the fact that the loadscausing local or global buckling may be lower than the loadscausing strength problems.

502 The analysis is normally performed by applying a set ofstatic loads. Hereafter, the factor by which this set of loads hasto be multiplied for stability problems to occur is determinedby the analysis program.

B 600 Thermal analysis601 By thermal analysis, the temperature distribution instructural parts is determined, based on the initial temperature,heat input/output, convection, etc. This is normally a time-dependent analysis; however, it is usually not very time-con-suming as only one degree of freedom is present at each mod-elled node.

Guidance note:A thermal analysis set-up as described can be used to analyseanalogous types of problems involving other time-dependentquantities than temperature. This applies to problems governedby the same differential equation as the one which governs heattransfer. An example of such an application can be found in foun-dation engineering for analysis of the temporal evolution of set-tlements in foundation soils.


B 700 Other types of analyses701 The analyses listed in B200 through B600 only encom-pass some of the types of analyses that can be performed byFEM analysis. Other types of analyses are: plastic analyses andanalyses including geometric non-linearities.702 Combinations of several analyses can be performed. Asexamples hereof, the results of an initial frequency analysiscan be used as a basis for subsequent dynamic analysis, and theresults of a thermal analysis may be used to form a load case ina subsequent static analysis.

C. ModellingC 100 General101 The results of a FEM analysis are normally documentedby plots and printouts of selected extreme response values.However, as the structural FEM model used can be very com-plex, it is important also to document the model itself. Evenminor deviations from the intention may give results that donot reflect reality properly.

C 200 Model201 The input for a FEM model must be documented thor-oughly by relevant printouts and plots. The printed data shouldpreferably be stored or supplied as files on a CD-ROM

C 300 Coordinate systems301 Different coordinate systems may be used to define themodel and the boundary conditions. Hence the coordinate sys-tem valid for the elements and boundary conditions should bechecked, e.g. by plots. This is particularly important for beamelements given that it is not always logical which axes are usedto define the sectional properties.302 In a similar manner, as a wrong coordinate system forsymmetry conditions may seriously corrupt the results, theboundary conditions should be checked.303 Insofar as regards laminate elements, the default coordi-nate system often constitutes an element coordinate system,which may have as a consequence that the fibre directions aredistributed randomly across a model.

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.K –

C 400 Material properties401 Several different material properties may be used acrossa model, and plots should be checked to verify that the materialis distributed correctly.402 Drawings are often made by means of using units of mmto obtain appropriate values. When the model is transferred tothe FEM program, the dimensions are maintained. In this casecare should be taken in setting the material properties (andloads) correctly, as kg-mm-N-s is not a consistent set of units.It is advisable to use SI-units (kg-m-N-s).

C 500 Material models501 The material model used is usually a model for isotropicmaterial, i.e. the same properties prevail in all directions. Note,however, that for composite materials an orthotropic materialmodel has to be used to reflect the different material propertiesin the different directions. For this model, material propertiesare defined for three orthogonal directions. By definition ofthis material, the choice of coordinate system for the elementshas to be made carefully.

C 600 Elements601 For a specific structural part, several different elementtypes and element distributions may be relevant depending onthe type of analysis to be carried out. Usually, one particularelement type is used for the creation of a FEM model. How-ever, different element types may be combined within thesame FEM model. For such a combination special considera-tions may be necessary.

C 700 Element types701 1D elements consist of beam elements. Models with beam elements are quite simple to create and pro-vide good results for framework structures.One difficulty may be that the sectional properties are not vis-ible. Hence, the input should be checked carefully for thedirection of the section and the numerical values of the sec-tional properties. Some FEM programs can generate 3D viewsshowing the dimensions of the sections. This facility should beused, if present.Naturally, the stresses in the connections cannot be calculatedaccurately by the use of beam elements only.702 2D elements consist of shell and plate elements. Shell and plate elements should be used for parts consisting ofplates or constant thickness sub-parts. As shell elements suita-ble for thick plates exist, the wall thickness does not need to bevery thin to obtain a good representation by such elements.These elements include the desired behaviour through thethickness of the plate. The same problems as for beam ele-ments are present for shell elements as the thickness of theplates is not shown. The thickness can, however, for most FEMprograms be shown by means of colour codes, and for someprograms the thickness can be shown by 3D views.703 The stresses at connections such as welds cannot befound directly by these elements either.704 3D elements consist of solid elements. 705 By the use of solid elements the correct geometry can bemodelled to the degree of detail wanted. However, this may

imply that the model will include a very large number of nodesand elements, and hence the solution time will be very long.Furthermore, as most solid element types only have threedegrees of freedom at each node, the mesh for a solid modelmay need to be denser than for a beam or shell element model.

C 800 Combinations801 Combination of the three types of elements is possible,however, as the elements may not have the same number ofdegrees of freedom (DOF) at each node, care should be takennot to create unintended hinges in the model.802 Beam elements have six degrees of freedom in eachnode – three translations and three rotations, while solid ele-ments normally only have three – the three translations. Shellelements normally have five degrees of freedom – the rotationaround the surface normal is missing. However, these elementsmay have six degrees of freedom, while the stiffness for thelast rotation is fictive.803 The connection of beam or shell elements to solid ele-ments in a point, respectively a line, introduces a hinge. Thisproblem may be solved by adding additional ‘dummy’ ele-ments to get the correct connection. Alternatively, constraintsmay be set up between the surrounding nodal displacementsand rotations. Some FEM programs can set up such constraintsautomatically.

C 900 Element size and distribution of elements901 The size, number and distribution of elements requiredin an actual FEM model depend on the type of analysis to beperformed and on the type of elements used.902 Generally, as beam and shell elements have five or sixdegrees of freedom in each node, good results can be obtainedwith a small number of elements. As solid elements only havethree degrees of freedom in each node, they tend to be morestiff. Hence, more elements are needed.903 The shape and order of the elements influence therequired number of elements. Triangular elements are morestiff than quadrilateral elements, and first-order elements aremore stiff than second-order elements.

Guidance note:The required number of elements and its dependency on the ele-ment shape are illustrated in an example, in which a cantilever ismodelled by beam, membrane, shell and solid elements, see Fig-ure 1.

Figure 1 Cantilever

Table C1 gives the required number of elements as a function ofthe element type applied, and the corresponding analysis resultsin terms of displacements and stresses are also given.


100 N

10 mm

100 mm

E = 2.1⋅105 N/mm2

DET NORSKE VERITAS


C 1000 Element quality1001 The results achieved by a certain type and number ofelements depend on the quality of the elements. Several meas-ures for the quality of elements can be used; however, the mostcommonly used are aspect ratio and element warping.1002 The aspect ratio is the ratio between the side lengths ofthe element. This should ideally be equal to 1, but aspect ratiosof up to 3 to 5 do usually not influence the results and are thusacceptable.1003 Element warping is the term used for non-flatness ortwist of the elements. Even a slight warping of the elementsmay influence the results significantly.1004 Most available FEM programs can perform checks ofthe element quality, and they may even try to improve the ele-ment quality by redistribution of the nodes.1005 The quality of the elements should always be checkedfor an automatically generated mesh, in particular, for theinternal nodes and elements. It is usually possible to generategood quality elements for a manually generated mesh.1006 With regard to automatically generated high-order ele-ments, care should be taken to check that the nodes on the ele-ment sides are placed on the surface of the model and not juston the linear connection between the corner nodes. This prob-lem often arises when linear elements are used in the initial cal-culations, and the elements are then changed into higher-orderelements for a final calculation.1007 Benchmark tests to check the element quality for dif-ferent element distributions and load cases are given byNAFEMS. These tests include beam, shell and solid elements,as well as static and dynamic loads.

C 1100 Boundary conditions1101 The boundary conditions applied to the model shouldbe as realistic as possible. This may require that the FEMmodel becomes extended to include element models of struc-tural parts other than the particular one to be investigated. Onesituation where this comes about is when the true supports of aconsidered structure have stiffness properties which cannot bewell-defined unless they are modelled by means of elementsthat are included in the FEM model. When such an extended FEM model is adopted, deviationsfrom the true stiffness at the boundary of the structural part inquestion may then become minor only. As a consequence ofthis, the non-realistic effects due to inadequately modelledboundary conditions become transferred further away to the

neighbouring structural parts or sub-parts, which are now rep-resented by elements in the extended FEM model.

C 1200 Types of restraints1201 The types of restraints normally used are constrained orfree displacements/rotations or supporting springs. Other typesof restraints may be a fixed non-zero displacement or rotationor a so-called contact, i.e. the displacement is restrained in onedirection but not in the opposite direction.1202 The way that a FEM program handles the fixed bound-ary condition may vary from one program to another. Oneapproach is to remove the actual degree of freedom from themodel; another is to apply a spring with a large stiffness at theactual degree of freedom. The latter approach may lead to sin-gularities if the stiffness of the spring is much larger than thestiffness of the element model. Evidently, the stiffness can alsobe too small, which may in turn result in singularities. An appropriate value for the stiffness of such a stiff spring maybe approximately 106 times the largest stiffness of the model.1203 As the program must first identify whether the dis-placement has to be constrained or free, the contact boundarycondition requires a non-linear calculation.

C 1300 Symmetry/antimetry1301 Other types of boundary conditions are symmetric andantimetric conditions, which may be applied if the model andthe loads possess some kind of symmetry. Taking such sym-metry into account may reduce the size of the FEM model sig-nificantly.1302 The two types of symmetry that are most frequentlyused are planar and rotational symmetries. The boundary con-ditions for these types of symmetry can normally be defined inan easy manner in most FEM programs by using appropriatecoordinate systems.1303 The loads for a symmetric model may be a combinationof a symmetric and an antimetric load. This can be consideredby calculating the response from the symmetric loads for amodel with symmetric boundary conditions, and adding theresponse from the antimetric loads for a model with antimetricboundary conditions.1304 If both model and loads have rotational symmetry, asectional model is sufficient for calculating the response.1305 Some FEM programs offer the possibility to calculatethe response of a model with rotational symmetry by a sec-tional model, even if the load is not rotational-symmetric, as

Table C1 Analysis of cantilever with different types of elements.Element type Description Number of

elementsuy

[mm]σx,node

[N/mm2]σx,element[N/mm2]

Analytical result – 1.9048 600 600BEAM2D Beam element, 2 nodes per element, 3 DOF

per node, ux, uy and θz10 1.9048 600 6001 1.9048 600 600

PLANE2D Membrane element, 4 nodes per element, 2 DOF per node, ux and uy

10 × 1 1.9124 570 0

TRIANG Membrane element, 3 nodes per element, 2 DOF per node, ux and uy

10 × 1 × 2 0.4402 141 14120 × 2 × 2 1.0316 333 33340 × 4 × 2 1.5750 510 510

SHELL3 Shell element, 3 nodes per element, 6 DOF per node

20 × 2 × 2 1.7658 578 405

SOLID Solid element, 8 nodes per element, 3 DOF per node ux, uy and uz

10 × 1 1.8980 570 570

TETRA4 Solid element, 4 nodes per element, 3 DOF per node ux, uy and uz

10 × 1 × 1 0.0792 26.7 26.720 × 2 × 1 0.6326 239 23940 × 4 × 1 1.6011 558 558

TETRA4R Solid element, 4 nodes per element, 6 DOF per node

20 × 2 × 1 1.7903 653 487

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.K –

the program can model the load in terms of Fourier series.

C 1400 Loads1401 The loads applied for the FEM calculation are usuallystructural loads, however, centrifugal loads and temperatureloads are also relevant.1402 Structural loads consist of nodal forces and momentsand of surface pressure. Nodal forces and moments are easilyapplied, but may result in unrealistic results locally. This is dueto the fact that no true loads act in a single point. Thus, appli-cation of loads as pressure loads will in most cases form themost realistic way of load application.

C 1500 Load application1501 The loading normally consists of several load compo-nents, and all of these components may be applied at the sametime. As a slightly different load combination in a new analysiswill require an entirely new calculation, this is, however, notvery rational.1502 To circumvent the problems involved with executionof an entirely new calculation when only a slightly differentload combination is considered, each of the load componentsshould be applied separately as a single load case, and theresults found from each of the corresponding analyses shouldthen be combined. In this way, a large range of load combina-tions can be considered. To facilitate this procedure, unit loadsshould be used in the single load cases, and the actual loadsshould then be used in the linear combinations.1503 As only one or more parts of the total structure is mod-elled, care should be taken to apply the loads as they are expe-rienced by the actual part. To facilitate such load application,‘dummy’ elements may be added, i.e. elements with a stiffnessrepresentative of the parts which are not modelled – these areoften beam elements. The loads can then be applied at the geo-metrically correct points and be transferred via the beam ele-ments to the structural part being considered.

D. DocumentationD 100 Model101 The results of a FEM analysis can be documented by alarge number of plots and printouts, which can make it an over-whelming task to find out what has actually been calculatedand how the calculations have been carried out.102 The documentation for the analysis should clearly docu-ment which model is considered, and the relevant resultsshould be documented by plots and printouts.103 The model aspects listed in D200 through D700 can andshould be checked prior to execution of the FEM analysis.

D 200 Geometry control201 A verification of the geometric model by a check of thedimensions is an important and often rather simple task. Thissimple check may reveal if numbers have unintentionally beenentered in an incorrect manner.

D 300 Mass – volume – centre of gravity301 The mass and volume of the model should always bechecked. Similarly, the centre of gravity should correspondwith the expected value.

D 400 Material401 Several different materials can be used in the same FEMmodel. Some of these may be fictitious. This should bechecked on the basis of plots showing which material isassigned to each element, and by listing the material proper-ties. Here, care should be taken to check that the material prop-

erties are given according to a consistent set of units.

D 500 Element type501 Several different element types can be used, and hereplots and listing of the element types should also be presented.

D 600 Local coordinate system601 With regard to beam and composite elements, the localcoordinate systems should be checked, preferably, by plottingthe element coordinate systems.

D 700 Loads and boundary conditions701 The loads and boundary conditions should be plotted tocheck the directions of these, and the actual numbers should bechecked from listings. To be able to check the correspondencebetween plots and listings, documentation of node/elementnumbers and coordinates may be required.

D 800 Reactions801 The reaction forces and moments are normally calcu-lated by the FEM programs and should be properly checked.As a minimum, it should be checked that the total reaction cor-responds with the applied loads. This is especially relevantwhen loads are applied to areas and volumes, and not merelyas discrete point loads. For some programs it is possible to plotthe nodal reactions, which can be very illustrative.802 A major reason for choosing a FEM analysis as the anal-ysis tool for a structure or structural part is that no simple cal-culation can be applied for the purpose. This implies that thereis no simple way to check the results. Instead checks can becarried out to make probable that the results from the FEManalysis are correct.

D 900 Mesh refinement901 The simplest way of establishing whether the presentmodel or mesh is dense enough is to remesh the model with amore dense mesh, and then calculate the differences betweenanalysis results from use of the two meshes. As several meshesmay have to be created and tried out, this procedure can, how-ever, be very time-consuming. Moreover, as modelling simpli-fication can induce unrealistic behaviour locally, thisprocedure may in some cases also result in too dense meshes.Instead, an indication of whether the model or mesh is suffi-cient would be preferable.

D 1000 Results1001 Initially, the results should be checked to see if theyappear to be realistic. A simple check is made on the basis ofan evaluation of the deflection of the component, whichshould, naturally, reflect the load and boundary conditionsapplied as well as the stiffness of the component. Also, thestresses on a free surface should be zero.1002 Most commercial FEM programs have some means forcalculation of error estimates. Such estimates can be defined inseveral ways. One of the most commonly used estimates is anestimate of the error in the stress. The estimated ‘correct’ stressis found by interpolating the stresses by the same interpolationfunctions as are used for displacements in defining the elementstiffness properties.Another way of getting an indication of stress errors is givenby means of comparison of the nodal stresses calculated at anode for each of the elements that are connected to that node.Large variations indicate that the mesh should be more dense.1003 If the results of the analysis are established as linearcombinations of the results from single load cases, the loadcombination factors used should be clearly stated. 1004 The global deflection of the structure should be plottedwith appropriately scaled deflections. For further evaluation,deflection components could be plotted as contour plots to see

DET NORSKE VERITAS


the absolute deflections.For models with rotational symmetry, a plot of the deflectionrelative to a polar coordinate system may be more relevant forevaluation of the results.1005 All components of the stresses are calculated, and itshould be possible to plot each component separately to eval-uate the calculated stress distribution.

1006 The principal stresses should be plotted with an indica-tion of the direction of the stress component, and these direc-tions should be evaluated in relation to the expecteddistribution.1007 As for the evaluation of the resulting stresses, also thecomponents of the resulting strains and the principal strainshould be plotted in an evaluation of the results from the anal-ysis.

DET NORSKE VERITAS

Amended December 2008 Offshore Standard DNV-OS-J101, October 2007see note on front cover App.L –

APPENDIX L ICE LOADS FOR CONICAL STRUCTURES

A. Calculation of Ice Loads

A 100 General101 Calculation of ice loads on conical structures such as icecones in the splash zone of monopiles and gravity base struc-tures can be carried out by application of Ralston’s formulae,which are based on plastic limit analysis. Ralston’s formulae distinguish between upward breakingcones and downward breaking cones, see Figure 1. For off-shore wind turbine structures, downward breaking cones aremost common.

Figure 1 Upward breaking cone (left) and downward breaking cone (right)

For upward breaking cones, the horizontal force on the cone is

The vertical force on the cone is

For downward breaking cones, also known as inverted cones,

the horizontal force on the cone is

The vertical force on the cone is

The following symbols are used in these expressions

σf = flexural strength of iceγw = specific weight of seawaterh = ice sheet thicknessb = cone diameter at the water linebT = cone diameter at top of cone

A1, A2, A3, A4, B1 and B2 are dimensionless coefficients,whose values are functions of the ice-to-cone friction coeffi-cient μ and of the inclination angle α of the cone with the hor-izontal. Graphs for determination of the coefficients are givenin Figure 2. The argument k is used for determination of the coefficients A1and A2 from Figure 2.For upward breaking cones,

shall be used. For downward breaking cones,

shall be used.The inclination angle α with the horizontal should not exceedapproximately 65° in order for the theories underlying the for-mulae to be valid.

b

bT

α

bT

b α

SWL SWL

422

32

22

1 ))(( AbbhAhbAhAR TwwfH −++= γγσ

)( 2221 TwHV bbhBRBR −+= γ

422

32

22

1 ))(91

91( AbbhAhbAhAR TwwfH −++= γγσ

)(91 22

21 TwHV bbhBRBR −+= γ

hb

kf

W

σγ 2

=

hb

kf

W

σγ9

2

=

DET NORSKE VERITAS

Offshore Standard DNV-OS-J101, October 2007 Amended December 2008 – App.L see note on front cover

Figure 2 Ice force coefficients for plastic limit analysis according to Ralston’s formulae.

DET NORSKE VERITAS

DNV-OS-J101: Design of Offshore Wind Turbine Structureshuniv.hongik.ac.kr/~geotech/key...

Documents

Transcript of DNV-OS-J101: Design of Offshore Wind Turbine Structureshuniv.hongik.ac.kr/~geotech/key...