DNV-RP-F105.

qxd 03-05-02 13:00 Page 2

RECOMMENDED PRACTICE
DNV-RP-F105

FREE SPANNING PIPELINES

MARCH 2002

DET NORSKE VERITAS

FOREWORD
DET 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 consultancy
services relating to quality of ships, offshore units and installations, and onshore industries worldwide, and carries out research
in 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 Methodology
B) Materials Technology
C) Structures
D) Systems
E) Special Facilities
F) Pipelines and Risers
G) Asset Operation

Amendments October 2003

This Code has been amended, but not reprinted in October 2003. The changes are incorporated in the printable (pdf) version.
All changes affecting DNV Offshore Codes that have not been reprinted, are published separately in the current Amendments
and Corrections, issued as a printable (pdf) file.

Comments may be sent by e-mail to rules@dnv.com

For subscription orders or information about subscription terms, please use distribution@dnv.com
Comprehensive information about DNV services, research and publications can be found at http://www.dnv.com, or can be obtained from DNV,
Veritasveien 1, N-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 person
for 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 Det
Norske Veritas.

CONTENTS
1 General ...................................................................................4
1.1 Introduction............................................................. 4
1.2 Objective .................................................................. 4
1.3 Scope and Application........................................... 4
1.4 Safety philosophy................................................... 5
1.5 Free Span Response Classification...................... 6
1.6 Flow regimes ........................................................... 6
1.7 Relationship to other Rules ................................... 7
1.8 Definitions ............................................................... 7
1.9 Abbreviations.......................................................... 7
1.10 Symbols .................................................................... 7
2 Design Criteria.................................................................. 10
2.1 General ................................................................... 10
2.2 Temporal classification ....................................... 10
2.3 Screening Fatigue Criteria .................................. 10
2.4 Fatigue Criterion................................................... 11
2.5 ULS Criterion........................................................ 12
2.6 Safety Factors........................................................ 13
3 Environmental Condi tions ............................................ 15
3.1 General ................................................................... 15
3.2 Current conditions................................................ 15
3.3 Short-term wave conditions................................ 16
3.4 Reduction functions............................................. 18
3.5 Long-term environmental modelling ................ 18
3.6 Return Period Values ........................................... 19
DET NORSKE VERITAS
Recommended Practice DNV-RP-F105, March 2002
Page 3
4 Response Models ...............................................................20
4.1 General....................................................................20
4.2 Marginal Fatigue Life Capacity..........................20
4.3 In-line Response Model.......................................21
4.4 Cross-flow Response Model...............................22
5 Force Model........................................................................25
5.1 General....................................................................25
5.2 FD solution for In-line direction.........................25
5.3 Simplified Fatigue Assessment...........................26
5.4 Force Coefficients .................................................26
6 Structural Analysis...........................................................29
6.1 General....................................................................29
6.2 Morphological classification...............................29
6.3 Structural modelling .............................................29
6.4 Functional Loads...................................................30
6.5 Static analysis ........................................................30
6.6 Eigen-value analysis .............................................31
6.7 Added Mass............................................................31
6.8 Approximate response quantities .......................31
7 Pipe -soil interaction..........................................................34
7.1 General....................................................................34
7.2 Modelling of pipe-soil interaction......................34
7.3 Approximate Soil Stiffness..................................35
7.4 Artificial supports..................................................38
8 References ...........................................................................39
Recommended Practice DNV-RP-F105, March 2002
Page 4
1 General
1.1 Introduction
1.1.1 The present document considers free spanning
pipelines subjected to combined wave and current loading.
The premises for the document are based on technical de-
velopment within pipeline free span technology in recent
R&D projects, as well as design experience from recent
and ongoing projects, i.e.
• DNV Guideline 14, see Mørk & Fyrileiv (1998)
• The sections regarding Geotechnical Conditions and
part of the hydrodynamic model are based on the re-
search performed in the GUDESP project, see Tura et
al., (1994).
• The sections regarding Free Span Analysis and in-line
Vortex Induced Vibrations (VIV) fatigue analyses are
based on the published results from the MULTISPAN
project, see Mørk et al., (1997).
• Numerical study based on CFD simulations for vibra-
tions of a pipeline in the vicinity of a trench, per-
formed by Statoil, DHI & DNV, see Hansen et al,
2001.
• Further, recent R&D and design experience e.g. from
Åsgard Transport, ZEEPIPE, TOGI and TROLL OIL
pipeline projects are implemented, see Fyrileiv &
Mørk (1998).
The basic principles applied in this document are in
agreement with most recognised rules and reflect state-of-
the-art industry practice and latest research.
This document includes a brief introduction of the basic
hydrodynamic phenomena, principles and parameters. For
a thorough introduction see e.g. Sumer & Fredsøe, (1997)
and Blevins (1994).
1.2 Objective
1.2.1 The objective of this document is to provide ra-
tional design criteria and guidance for assessment of pipe-
line free spans subjected to combined wave and current
loading.
1.3 Scope and Application
1.3.1 Detailed design criteria are specified for Ultimate
Limit State (ULS) and Fatigue Limit State (FLS) due to in-
line and cross-flow Vortex Induced Vibrations (VIV) and
direct wave loading.
The following topics are considered:
• methodologies for free span analysis;
• requirements for structural modelling;
• geotechnical conditions ;
• environmental conditions & loads;
• requirements for fatigue analysis;
• response and direct wave force analysis models; and
• acceptance criteria.
1.3.2 Free spans can be caused by:
• seabed unevenness
• change of seabed topology (e.g. scouring, sand waves)
• artificial supports/rock berms etc.
1.3.3 The following environmental flow conditions are
described in this document:
• steady flow due to current;
• oscillatory flow due to waves; and
• combined flow due to current and waves.
The flow regimes are discussed in section 1.6.
1.3.4 There are no limitations to span length and span
gap with respect to application of this Recommended Prac -
tice.
The basic cross-flow VIV Response model is, however,
based on single mode response.
In case several potential vibration modes can become ac-
tive at a given flow velocity, the mode associated with the
largest contribution to the fatigue damage shall be applied.
Unless otherwise documented the damage contribution for
any modes should relate to the same critical (weld) loca-
tion.
1.3.5 The free span analysis may be based on approxi-
mate response expressions or a refined FE approach de-
pending on the free span classification, see section 6.2.
The following cases are considered:
• single spans
• spans interacting with adjacent/side spans.
The stress ranges and natural frequencies should normally
be obtained from an FE-approach. Requirements to the
structural modelling and free span analyses are given in
section 6.
1.3.6 The following models are considered:
• Response Models (RM)
• Force Models (FM)
An amplitude response model is applicable when the
vibration of the free span is dominated by vortex induced
resonance phenomena. A force model may be used when
the free span response can be found through application of
calibrated hydrodynamic loads. The selection of an
appropriate model may be based on the prevailing flow
regimes, see section 1.6.
1.3.7 The fatigue criterion is limited to stress cycles
within the elastic range. Low cycle fatigue due to elasto-
plastic behaviour is considered outside the scope of this
document.
 Recommended Practice DNV-RP-F105, March 2002
Page 5

1.3.8 Fatigue loads due to trawl interaction, cyclic 1.3.9 The main aspects of a free span assessment to-
loads during installation or pressure variations are not con- gether with key parameters and main results are illustrated
sidered herein but must be considered as a part of the inte- in the figure below.
grated fatigue damage assessment.
Components

Environmental Structural Response model Acceptance

Main

description response Force model criteria

ch. 3 ch. 6 & 7 ch. 4 & 5 ch. 2

Project data Pipe data VR Safety class

Key Parameters

Current statistics Seabed & pipe profile KC Safety factors

Current profile Soil data α SN curve
Wave statistics Lay tension Damping
Wave spectrum Operational conditions Free span parameters
Directionality
Turbulence

Screening

Wave & current

Return period Natural Frequency OK / not OK
Design Criteria & Main results

values (span length)

Fatigue

Wave & current Natural Frequency Stress ranges

long-term Stress amplitude No of cycles OK / not OK
description (fatigue life)

ULS

Wave & current Natural Frequency

Return period Stress ranges Extreme stresses OK / not OK
values Static Bending (local buckling)

Figure 1-1 Overview of main components in a Free Span Assessment

1.4 Safety philosophy

1.4.1 The safety philosophy adopted herein complies

with section 2 in DNV-OS-F101.

1.4.2 The reliability of the pipeline against fatigue fail-

ure is ensured by use of a Load and Resistance Factors
Design Format (LRFD).
• For the in-line VIV acceptance criterion, the set of
safety factors is calibrated to acceptable target reliabil-
ity levels using reliability-based methods.
• For all other acceptance criteria, the recommended
safety factors are based on engineering judgement in
order to obtain a safety level equivalent to modern in-
dustry practice.

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 6

1.5 Free Span Response Classification

1.5.1 An overview of typical free span characteristics is length. The ranges indicated for the normalised free span
given in the table below as a function of the free span length in terms of (L/D) are tentative and given for illus-
tration only.

L/D Response description

L/D<30 1) Very little dynamic amplification.
Normally not required to perform comprehensive fatigue design check. Insignificant dynamic re -
sponse from environmental loads expected and unlikely to experience VIV.
30<L/D<100 Response dominated by beam behaviour
Typical span length for operating conditions.
Natural frequencies sensitive to boundary conditions (and effective axial force).
100<L/D<200 Response dominated by combined beam and cable behaviour
Relevant for free spans at uneven seabed in temporary conditions. Natural frequencies sensitive to
boundary conditions, effective axial force (including initial deflection, geometric stiffness) and pipe
“feed in”.
L/D>200 Response dominated by cable behaviour
Relevant for small diameter pipes in temporary conditions. Natural frequencies governed by de-
flected shape and effective axial force.
1) For hot pipelines (response dominated by the effective axial force) or under extreme current conditions (Uc > 1.0-2.0 m/s)
this L/D limit may be misleading.

1.6 Flow regimes

is the significant wave induced velocity amplitude normal

1.6.1 The current flow velocity ratio, α=Uc/(Uc+Uw )
to the pipe, see section 4) may be applied to classify the
(where Uc is the current velocity normal to the pipe and Uw
flow regimes as follows:

α < 0.5 wave dominant - wave superimposed by current.

In-line direction: in-line loads may be described according to Morison’s formulae, see section 5. In-
line VIV due to vortex shedding is negligible.
Cross-flow direction: cross-flow loads are mainly due to asymmetric vortex shedding. A response
model, see section 4, is recommended.
0.5< α < 0.8 wave dominant – current superimposed by wave
In-line direction: in-line loads may be described according to Morison’s formulae, see section 5. In-
line VIV due to vortex shedding is mitigated due to the presence of waves.
Cross-flow direction: cross-flow loads are mainly due to asymmetric vortex shedding and resemble
the current dominated situation. A response model, see section 4, is recommended.
α > 0.8 current dominant
In-line direction: in-line loads comprises the following components :
• a steady drag dominated comp onent
• a oscillatory component due to regular vortex shedding
For fatigue analyses a response model applies, see section 4. In -line loads according to Morison’s
formulae are normally negligible.
Cross-flow direction: cross-flow loads are cyclic and due to vortex shedding and resembles the pure
current situation. A response model, see section 4, is recommended.

Note that α=0 correspond to pure oscillatory flow due to waves and α =1 corresponds to pure (steady) current flow.
The flow regimes are illustrated in Figure 1-2.
 Recommended Practice DNV-RP-F105, March 2002
Page 7

6
1.8.4 Marginal Fatigue Capacity is defined as the fa-
tigue capacity (life) with respect to one seastate defined by
5 its significant wave height, peak period and direction.
current dominated flow
4
Flow Velocity(Uc+U w)

wave dominated flow α=0.8 1.8.5 Response Model is a model where the structural
3 response due to VIV is determined by hydrodynamical
2 parameters.
α=0.5
1
1.8.6 Span Length is defined as the length where a con-
0 tinuous gap exists, i.e. as the visual span length.
α=0.0
-1
time
-2 1.9 Abbreviations

Figure 1-2 Flow regimes CSF concrete stiffness factor

FM force model
LRFD load and resistance factors des ign format
1.6.2 Oscillatory flow due to waves is stochastic in
OCR over-consolidation ratio (only clays)
nature, and a random sequence of wave heights and asso-
RM response model (VIV)
ciated wave periods generates a random sequence of near
RD response domain
seabed horizontal oscillations. For VIV analyses, the sig-
RPV return period values
nificant velocity amplitude, Uw , is assumed to represent a
TD time domain
single sea-state. This is likely to be a conservative ap-
ULS ulitmate limit state
proximation.
VIV vortex induced vibrations

1.7 Relationship to other Rules

1.10 Symbols
1.7.1 This document formally supports and complies
with the DNV Offshore Standard “Submarine Pipeline
1.10.1 Latin
Systems”, DNV-OS-F101, 2000 and is considered to be a
supplement to relevant National Rules and Regulations.
aκ parameter for rain-flow counting factor
a characteristic fatigue strength constant
1.7.2 This document is supported by other DNV off-
shore codes as follows: Ae external cross-section area
• Recommended Practice DNV-RP-C203 “Fatigue Ai internal cross-section (bore) area
Strength Analysis of Offshore Steel Structures” A IL inline unit amplitude stress (stress induced by
• Offshore Standard DNV-OS-F201 “Dynamic Risers” a pipe (vibration mode) deflection equal to an
• Classification Note No. 30.5 “Environmental Condi- outer diameter D)
A CF cross-flow unit amplitude stress
tions and Environmental Loads”, 2000.
As pipe steel cross section area
In case of discrepancies between the recommendations, (A Y /D) normalised in-line VIV amplitude
(A Z/D) normalised cross-flow VIV amplitude
this Document supersedes the Recommended Practice and
Classification Notes listed above. B chord corresponding to pipe embedment
equal to v or linearisation constant
bκ parameter for rainflow counting factor
1.8 Definitions
Ca added mass coefficient (CM-1)
CD drag coefficient
1.8.1 Effective Span Length is the length of an idealised CM the inertia coefficient
fixed-fixed span having the same structural response in CL coefficient for lateral soil stiffness
terms of natural frequencies as the real free span supported CV coefficient for vertical soil stiffness
on soil. CT constant for long-term wave period distribu-
tion
1.8.2 Force Model is in this document a model where C1-6 Boundary condition coefficients
the environmental load is based on Morison’s force ex- c(s) soil damping per unit length
pression. D pipe outer diameter (including any coating)
D fat deterministic fatigue damage
1.8.3 Gap is defined as the distance between the pipe Ds outer steel diameter
and the seabed. The gap used in design, as a single repre- E Young's modulus
sentative value, must be characteristic for the free span EI bending stiffness
The gap may be calculated as the average value over the E seabed gap
central third of the span. es void ratio

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 8

(e/D) seabed gap ratio Ntr true steel wall axial force
f0 in-line (f0,in) or cross-flow (f0,cr) natural fre - Nc soil bearing capacity parameter
quency (determined at no flow around the Nq soil bearing capacity parameter
pipe) Nγ soil bearing capacity parameter
fsn concrete construction strength pe external pressure
fs vortex shedding frequency (Strouhal pi internal pressure
U Pi probability of occurrence for i’th stress cycle
frequency) = S t q deflection load per unit length
D
FL lateral pipe-soil contact force
FV vertical pipe-soil contact force PE Euler load = (1+CSF)π2 EI/Leff 2
fv dominating vibration frequency Ra axial soil reaction
fw wave frequency Rc current reduction factor
F() distribution function RD reduction factor from wave direction and
g gravity spreading
gc correction function due to steady current Rv vertical soil reaction
gD drag force term RIθ reduction factor from turbulence and flow
gI inertia force term direction
G shear modulus of soil or incomplete comple - Rk reduction factor from damping
mentary Gamma function UD
Re Reynolds number D=
G(ω) frequency transfer function from wave eleva- ν
tion to flow velocity s spreading parameter
Heff effective lay tension S stress range, i.e. double stress amplitude
HS significant wave height Ssw stress at intersection between two SN-curves
h water depth, i.e. distance from the mean sea Seff effective axial force
level to the pipe Sηη wave spectral density
I moment of inertia SSS stress spectra
Ic turbulence intensity over 30 minutes SUU wave velocity spectra at pipe level
ip plasticity index, cohesive soils su undrained shear strength, cohesive soils
k Wave number St Strouhal number
kc soil parameter or empirical constant for con- t pipe wall thickness or time
crete stiffening Texposure load exposure time
kp peak factor Tlife fatigue design life capacity
ks soil coefficient Tp peak period
kw normalisation constant Tu mean zero up-crossing period of oscillating
K soil stiffness flow
KL lateral (horizontal) dynamic soil stiffness Tw wave period
KV vertical dynamic soil stiffness U current velocity
(k/D) pipe roughness Uc current velocity normal to the pipe
U Us significant wave velocity
KC Keulegan Carpenter number = w
fw D Uw significant wave induced flow velocity normal
4 πm e ζ T to the pipe, corrected for wave direction and
KS stability parameter = spreading
ρD 2 v vertical soil settlement (pipe embedment)
K0 coefficient of earth pressure at rest U + Uw
k1 soil stiffness VR reduced velocity = c
f0D
k2 soil stiffness
w wave energy spreading function
L free span length, (apparent, visual)
y lateral pipe displacement
La length of adjacent span
z height above seabed or in-line pipe displace-
Leff effective span length
ment
Ls span length with vortex shedding loads
zD height to the mid pipe
Lsh length of span shoulder
zm macro roughness parameter
me effective mass per unit length
z0 sea-bottom roughness
m fatigue exponent
zr reference (measurement) height
m(s) mass per unit length including structural
mass, added mass and mass of internal fluid
Mn spectral moments of order n
ni number of stress cycles for stress block i
N number of independent events in a return
period
Ni number of cycles to failure for stress block i
 Recommended Practice DNV-RP-F105, March 2002
Page 9

1.10.2 Greek ψtrench,onset reduction factor for onset cross-flow due to

the effect of a trench
α current flow velocity ratio, generalised Phil- ψα,in correction factor for onset cross-flow due to
lips’ constant or Weibull scale parameter seabed proximity
αe temperature expansion coefficient ρ density of water
β Weibull shape parameter and relative soil ρ s/ρ specific mass ratio between the pipe mass (not
stiffness parameter including added mass) and the displaced wa-
∆/D relative trench depth ter.
∆p i internal pressure difference relative to laying σ stress, spectral width parameter or standard
∆T temperature difference relative to laying or deviation
storm duration σc standard deviation of current velocity fluctua-
δ pipe deflection tions
ε band-width parameter σu standard deviation of wave induced flow ve-
Γ gamma function locity
γ peak-enhancement factor for JONSWAP σdyn dynamic stress
spectrum or Weibull location parameter σl longitudinal stress
γsoil total unit weight of soil σN static axial stress
γsoil ’ submerged unit weight of soil σstat static stress
γwater unit weight of water σs effective soil stress or standard deviation of
γs safety factor on stress amplitude wave induced stress amplitude
γf safety factor on natural frequency θrel relative angle between flow and pipeline di-
rection
γcr safety factor for cross-flow screening criterion
θ
¯ flow direction
γin safety factor for in-line screening criterion
ζΤ total modal damping ratio
γk safety factor on stability parameter
ζsoil soil modal damping ratio
γon safety factor on onset value for VR
ζstr structural modal damping ratio
κRFC rainflow counting factor
ζh hydrodynamic modal damping ratio
κ curvature
ω0 angular natural frequency
λ1 mode shape factor
ω angular wave frequency
λmax equivalent stress factor
ωp angular spectral peak wave frequency
η usage factor
τmax soil shear strength
µ mean value
µa axial friction coefficient
µL lateral friction coefficient
ν Poisson's ratio
1.10.3 Subscripts
or kinematic viscosity (≈1.5·10-6 [m2 /s]
φ mode shape IL in-line
Φ() cumulative normal distribution function CF cross-flow
ϕ() normal distribution function onset onset of VIV
ϕs angle of friction, cohesionless soils 100year 100 year return period value
ψk,CM correction factor for CM due to pipe rough- 1year 1 year return period value
ness
ψtrench,CM correction factor for CM due to effect of pipe
in trench
ψproxi,CM reduction factor for CM due to seabed prox-
imity
ψk,CD correction factor for CD due to pipe rough-
ness
ψtrench,CD correction factor for CD due to effect of pipe
in trench
ψA,CD amplification factor for CD due to cross-flow
vibrations
ψproxi,CD reduction factor for CD due to seabed prox-
imity
ψproxi,onset correction factor for onset cross-flow due to
seabed proximity
ψmass,onset correction factor for onset cross-flow due to
specific mass of the pipe
ψα,onset correction factor for onset cross-flow due to
waves

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 10

2.1.5 Figure 2-1 gives an overview of the required de-

2 Design Criteria sign checks for a free span.

2.2 Temporal classification

2.1 General
2.2.1 The temporal criterion categorises the free span
2.1.1 For all temporary and permanent free spans a free as being caused due to scour or seabed unevenness, i.e.
span assessment addressing the integrity with respect to
• Scour induced free spans are caused by seabed erosion
fatigue and local buckling (ULS) shall be performed.
or bed-form activities. The free span scenarios (span
length, gap ratio etc.) may change with time.
2.1.2 Vibrations due to vortex shedding and direct
• Unevenness induced free spans are caused by an ir-
wave loads are acceptable provided the fatigue and ULS
regular seabed profile. Normally the free span sce-
criteria specified herein is fulfilled.
nario is time invariant unless operational parameters
such as pressure and temperature change significantly.
2.1.3 In case several potential vibration mo des can b e-
come active at a given flow velocity, the mode associated
2.2.2 In the case of scour induced spans, where no de-
with the largest contribution to the fatigue damage shall be
tailed information is available on the maximum expected
applied. Unless otherwise documented the damage contri-
span length, gap ratio and exposure time, the following
bution for any modes should relate to the same critical
apply:
(weld) location.
• Where uniform conditions exist and no large-scale
mobile bed-forms are present the maximum span
2.1.4 The following functional requirements apply:
length may be taken as the length resulting in a stati-
• The aim of fatigue design is to ensure an adequate
cally mid span deflection equal to one external diame -
safety against fatigue failure within the design life of ter (including any coating).
the pipeline.
• The exposure time may be taken as the remaining
• The fatigue analysis should cover a period which is
operational lifetime or the time duration until possible
representative for the free span exp osure period.
intervention works will take place. All previous dam-
• All stress fluctuations imposed during the entire de- age accumulation must be included.
sign life of the pipeline capable of causing fatigue
damage shall be accounted for.
2.2.3 Additional information (e.g. free span length, gap
• The local fatigue design checks are to be performed at ratio, natural frequencies) from surveys combined with an
all free spanning pipe sections accounting for damage inspection strategy may be used to qualify scour induced
contributions from all potential vibration modes re- free spans. These aspects are not covered in this document.
lated to the actual and neighbouring spans. Guidance may be found in Mørk et al., (1999) and Fyrileiv
et al., (2000).
Start
2.3 Screening Fatigue Criteria

2.3.1 The screening criteria proposed herein apply to

Free Span Data & Span intervention fatigue caused by Vortex Induced Vibrations (VIV) and
Characteristics Detailed analysis direct wave loading in combined current and wave loading
conditions. The screening criteria have been calibrated
against full fatigue analyses to provide a fatigue life in
excess of 50 years. The criteria apply to spans with a re-
Screening Fatigue sponse dominated by the 1st symmetric mode (one half
not OK not OK
wave) and should preferably be applied for screening
OK OK
analyses only and, if violated, more detailed fatigue analy-
ses should be performed. The ULS criterion in 2.5 must
always be checked.
ULS Check
not OK
2.3.2 The screening criteria proposed herein are based
OK on the assumption that the current velocity may be repre -
sented by a 3-parameter Weibull distribution. If this is not
the case, e.g. for bi-modal current distributions, care must
Stop
be taken and the applicability of these screening criteria
checked by full fatigue calculations.
Figure 2-1 Flow chart over design checks for a free
span.
 Recommended Practice DNV-RP-F105, March 2002
Page 11

2.3.3 The in-line natural frequency f0,IL must fulfil: where η is the allowable fatigue damage ratio, Tlife the
fatigue design life capacity and Texposure the life or load
f 0, IL Uc ,100year  L / D  γ IL exposure time.
> ⋅ 1 − ⋅
γf VRIL,onset ⋅ D  250  α
2.4.2 The fatigue damage assessment is based on the
Where accumulation law by Palmgren-Miner:
n
γf Safety factor on the natural frequency, see Dfat = ∑ i
Ni
2.6
γIL Screening factor for inline, see 2.6 Where
Current flow ratio=
α  Uc ,100year  D fat Accumulated fatigue damage.
max  ;0.6 
 Uw ,1year + Uc ,100year  ni Total number of stress cycles corresponding to the
  (mid-wall) stress range Si
D Outer pipe diameter incl. coating N Number of cycles to failure at stress range Si
L Free span length Σ Implies summation over all stress fluctuations
Uc,100year 100 year return period value for the cur- in the design life
rent velocity at the pipe level, see 3
Uw,1year Significant 1 year return period value for 2.4.3 The number of cycles to failure at stress range S
the wave induced flow velocity at the pipe is defined by the SN curve of the form:
level corresponding to the annual signifi-
cant wave height Hs,1year, see 3 a ⋅ S − m1 S > S sw
VRIL, onset In-line onset value for the reduced veloc- N =  1 −m
ity, see 4. a 2 ⋅ S 2 S ≤ S sw

If the above criterion is violated, then a full in-line VIV Where

fatigue analysis is required.
m1 , m2 Fatigue exponents (the inverse slope of the bi-
2.3.4 The cross-flow natural frequency f0,CF must fulfil: linear S-N curve)
a 1 , a 2 Characteristic fatigue strength constant de-
f 0, CF Uc ,100year + Uw ,1year fined as the mean-minus-two-standard-
> ⋅ γ CF deviation curve
γf VRCF,onset ⋅ D Ssw Stress at intersection of the two SN-curves
given by:
Where  log a 1 − log N sw 
 
 m1 
γCF Screening factor for cross-flow, see 2.6 S sw = 10  

VRCF,onset Cross-flow onset value for the reduced

velocity, see 4 Where Nsw is the number of cycles for which change in
slope appear. LogNsw is typically 6 – 7.
If the above criterion is violated, then a full in-line and
cross-flow VIV fatigue analysis is required.
1000

2.3.5 Fatigue analysis due to direct wave action is not

(a 1;m1)
required provided:
SSW
U c,100year 2 100
Stress Range, S

>
U w ,1year + U c,100year 3 (a2; m2 )

and the above screening criteria for in-line VIV is ful- 10

filled. If this criterion is violated, then a full fatigue analy-

ses due to in-line VIV and direct wave action is required
NSW
1
2.4 Fatigue Criterion 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 1.E+08 1.E+09 1.E+10
No of cycles, N

2.4.1 The fatigue criterion can be formulated as: Figure 2-2 Typical two-slope SN curve.

η⋅ Tlife ≥ Texposure

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 12

2.4.4 The SN-curves may be determined from: Where PHS , TP , θ is the probability of occurrence of each
• Dedicated laboratory test data,
individual sea-state, e.g. the probability of occurrence re-
• Accepted fracture mechanics theory, or
flected by the cell in a scatter diagram. The in-line fatigue
• DNV-RP-C203 “Fatigue Strength Analysis of Off-
life capacity is conservatively taken as the minimum ca-
shore Steel Structures”.
pacity (i.e., maximum damage) from VIV (RM) or direct
The SN-curve must be applicable for the material,
wave loads (FM) in each sea state.
construction detail, location of the intial defect (crack
initiation point) and corrosive environment. The basic
The fatigue life is the minimum of the in -line and the
principles in DNV-RP-C203 apply.
cross-flow fatigue lives.
2.4.5 The fatigue life capacity, Tlife , can be formally
2.4.9 The following marginal fatigue life capacities are
expressed as:
evaluated for (all) sea states characterised by (Hs , Tp , θ)
1
Tlife = RM , IL
THs Marginal fatigue capacity against in-
 f v ⋅ S im ⋅ Pi  , Tp , θ line VIV and cross-flow induced in-
∑  
 line motion in a single sea-state (Hs , Tp ,
 a  θ) integrated over long term pdf for the
Where current, see section 4.2.2.
RM , CF Marginal fatigue capacity against
THs, Tp, θ
Pi Probability of occurrence for the “i”th stress cross-flow VIV in a sea-state (Hs , Tp , θ)
cycle integrated over long term pdf for the
current, see section 4.2.1.
2.4.6 The concept adopted for the fatigue analysis ap- FM , IL Marginal fatigue capacity against direct
THs, Tp, θ
plies to both response models and force models. The stress wave actions in a single sea-state char-
ranges to be used may be determined by: acterised by (Hs , Tp , θ) using mean
• a response model, see section 4 value of current, see section 5.2.2.
• a force model, see section 5.

2.4.7 The following approach is recommended:
• The fatigue damage is evaluated independently in
each sea-state, i.e., the fatigue damage in each cell of
a scatter diagram in terms of (Hs , Tp , θ) times the
probability of occurrence for the individual sea state.
• In each sea-state (Hs , Tp , θ) is transformed into (Uw ,
Tu , θw ) at the pipe level as described in section 3.3.
• The sea state is represented by a significant short-term
flow induced velocity amplitude Uw with mean zero
up-crossing period Tu , i.e. by a train of regular wave
induced flow velocities with amplitudes equal to Uw
and period Tu . The effect of irregularity will reduce
the number of large amplitudes. Irregularity may be
accounted for provided it is properly documented.
• Integration over the long-term current velocity distri-
bution for the combined wave and current flow is per-
formed in each sea-state.
2.4.10 Unless otherwise documented, the following
assumptions apply:
• The current and wave induced flow components at the
pipe level are statistically independent.
• The current and wave-induced flow are assumed co-
linear. This implies that the directional probability of
occurrence data for either waves or current (the most
conservative with respect to fatigue damage) must be
used for both waves and current.
2.5 ULS Criterion
2.5.1 For the local buckling check reference is made to
the combined loading – load controlled condition criterion
in DNV-OS-F101, section 5, D500. Static and dynamic
bending moment, axial force and pressure shall be ac-
counted for.

2.4.8 The total fatigue life capacity in the in -line and 2.5.2 For extreme wave conditions, which can be as-
cross-flow direction is established by integrating over all sumed to cause large deformations on the shoulders, de-
sea-states, i.e. tailed analyses of the soil stiffness at the shoulders may be
−1 required. In lieu of detailed documentation, the boundary
 PHs,T p,θ  conditions for the free span should be assumed as pinned-
=  ∑ ∑∑ 
( )
IL
Tlife
 θ H T min T RM ,IL ; T FM,IL  pinned (only valid for the Force Model calculations).
 S P Hs,T p,θ Hs,T p,θ 
−1 2.5.3 The maximum dynamic bending moment due to
 PHs,T p,θ  VIV and/or direct wave action may be found from the dy-
CF
Tlife 
= ∑ ∑∑ RM ,CF
θ H T T  namic stresses:
 S P Hs,T p,θ 
M E = σ dyn
2 ⋅I
Ds − t
 Recommended Practice DNV-RP-F105, March 2002
Page 13

Where Alternatively a regular wave approach using Hmax may be

applied.
σdyn Dynamic stress given below
I Moment of inertia
2.5.7 A simplified ULS screening may be performed in
Ds Outer diameter of steel pipe
terms of an equivalent stress check, ref. DNV-OS-F101,
t Wall thickness
section 12, F1200.
2.5.4 The dynamic stress, σdyn , is taken as:
2.5.8 The longitudinal stress is given by:
1  A 
in − line σdyn = max S IL ; 0.5 ⋅ SCF IL ; S FM , max 
2  A CF  {
σ l = σ N ± σ static + σ dyn }
1
cross − flow σdyn = S CF
2 Where
Where
σN Static axial stress from true wall axia l force,
Ntr given by: σN = Ntr/A s
SILIn-line stress range, see section 4
σstatic Static bending stress from static bending mo -
SCFCross-flow stress range, see section 4 ment, see 6.8.6. Applies to both vertical and
SFM,max Maximum stress range due to direct wave lateral directions.
loading (force model) , see section 5 σdyn Dynamic bending stress, see 2.5.4. Applies to
A IL In-line unit deflection stress amplitude due to both directions, in-line and cross-flow inde-
VIV, see section 6.8.4 pendently.
A CF Cross-flow unit deflection stress amplitude due
to VIV, see section 6.8.4 2.6 Safety Factors
The unit deflection stress amplitude A IL and A CF may be
evaluated using the following return period environmental
flow conditions: 2.6.1 The safety factors to be used with the screening
criteria are listed below.
U1year = U c,1year + U w ,1year
U10 year (
= max Uc,10 year + U w ,1year ; Uc,1year + U w ,10 year ) Table 2-1 Safety factors for screening criteria
(
U100 year = max Uc,100 year + U w,10 year ; Uc,10 year + U w ,100 year ) γIL
γCF
1.15
1.3

2.6.2 Pipeline reliability against fatigue uses the safety

2.5.5 For the cross-flow direction, the stress simply class concept, which takes account of the failure conse-
stems from the VIV induced amplitude. For the in -line quences, see DNV-OS-F101, Section 2.
direction, the dynamic stress range is taken as the maxi- The following safety factor format is used:
f v (γ s S( γ f , γ k , γ on .) ) P( •)
mum of: m

• The return period stress range, e.g. 100 year, for in- D fat = Texp osure ⋅ ∑ ≤η
line VIV, Sin , defined in section 4.3. a
• The stress from 50% of the cross-flow induced VIV γf, γon , γk and γs denote partial safety factors for the natural
motion. All parameters are defined in section 4.4. frequency, onset of VIV, stability parameter and stress
range respectively. The set of partial safety factor to be
2.5.6 The maximum dynamic stress σFM,max from direct applied are specified in the table below for the individual
wave loading may be calculated using a “design storm” safety classes.
approach using:
σFM,max= kp σs Table 2-2 Safety factors for fatigue
Safety Class
Safety Factor
where kp is the peak factor given by: Low Normal High
η 1.0 0.5 0.25
γs 1.051) (1.0)
k p = 2 ln (f v ∆T ) +
0.577
2 ln (f v ∆T ) γf 1.201) (1.15)
γk 1.30
∆T is the storm duration equal to 3 hour and fv is given in γon 1.10
section 5.2. kp may conservatively be taken equal to 4. 1) This safety factor is intended to be used in design when detailed
data about span length, gap etc is not known. If a span is assessed
σs is the standard deviation of the stress (amplitude) re - in-service with updated and measured span data, the safety factor in
sponse calculated from a time domain or frequency do- brackets may be used.
main analysis, see section 5.

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 14

Comments:
2.6.3 The reliability of the pipeline against local buck-
• η apply to both Response Model and Force Model
ling (ULS criterion) is ensured by use of the safety class
• γs is to be multiplied to the stress (S γS ) concept as implemented by use of safety factors according
• γf applies to the natural frequency (fo /γf) to DNV-OS-F101, section 5 D500 alternatively section 12.
• γon applies to onset values for in-line and cross-flow
VIV (VR,on/γon )
• γk applies to the stability parameter (KS/γk )
• For ULS, the calculation of load effects is to be per-
formed without safety factors (γS = γ f = γk = γon = 1.0),
see also section 2.6.3.

3 Environmental Conditions
3.1 General
3.1.1 The objective of the present section is to provide
guidance on:
• the long term current velocity distribution,
• short-term and long-term description of wave induced
flow velocity amplitude and period of oscillating flow
at the pipe level, and
• return period values.
3.1.2 The environmental data to be used in the assess-
ment of the long-term distributions shall be representative
for the particular geographical location of the pipeline free
span.
3.1.3 The flow conditions due to current and wave ac-
tion at the pipe level govern the response of free spanning
pipelines. The principles and methods as described in
Classification Note CN 30.5 may be used in addition to
this document as a basis when establishing the environ-
mental load conditions.
3.1.4 The environmental data must be collected from
periods that are representative for the long-term variation
of the wave and current climate, respectively. In case of
less reliable or limited number of wave and current data,
the statistical uncertainty should be assessed and, if sig-
nificant, included in the analysis.
3.1.5 Preferably, the environmental load conditions
should be established near the pipeline using measurement
data of acceptable quality and duration. The wave and cur-
rent characteristics must be transferred (extrapolated) to
the free span level and location using appropriate conser-
vative assumptions.
3.1.6 The following environmental description may be
applied:
• Directional information, i.e., flow characteristic versus
sector probability;
• Omnidirectional statistics may be used if the flow is
uniformly distributed.
If no such information is available, the flow should be
assumed to act perpendicular to the axis of the pipeline at
all times.
3.2 Current conditions
3.2.1 The steady current flow at the free span level may
be a compound of:
• tidal current;
• wind induced current;
• storm surge induced current, and
• density driven current.
DET NORSKE VERITAS
Recommended Practice DNV-RP-F105, March 2002
Page 15
When detailed field measurements are not available, the
tidal, wind and storm surge driven current velocity com-
ponents may be taken from Classification No. 30.5.
3.2.2 For water depths greater than 100 m, the ocean
currents can be characterised in terms of the driving and
steering agents:
• the driving agents are tidal forces, pressure gradients
due to surface elevation or density changes, wind and
storm surge forces.
• the steering agents are topography and the rotation of
the earth.
The modelling should account adequately for all agents.
3.2.3 The flow can be divided into two zones:
• an Outer Zone far from the seabed where the mean
current velocity and turbulence vary only slightly in
the horizontal direction.
• an Inner Zone where the mean current velocity and
turbulence show significant variations in the horizon-
tal direction and the current speed and direction is a
function of the local sea bed geometry.
3.2.4 The outer zone is located approximately one local
seabed form height above the seabed crest. In case of a flat
seabed, the outer zone is located approximately at height
(3600 z0 ) where z0 is the bottom roughness, see Table 3-1.
3.2.5 Current measurements (current meter) should be
made in the outer zone outside the boundary layer at a
level 1-2 seabed form heights above the crest. For large-
scale currents, such as wind driven and tidal currents, the
choice of measurement positions may be based on the
variations in the bottom topography assuming that the cur-
rent is geo-strophic, i.e., mainly running parallel to the
large-scale bottom contours.
Over smooth hills, flow separation occurs when the hill
slope exceeds about 20o . Current data from measurements
in the boundary layer over irregular bed forms are of little
practical value when extrapolating current values to other
locations.
3.2.6 In the inner zone the current velocity profile is
approximately logarithmic in areas where flow separation
does not occur:
(ln(z ) − ln(z 0 ) )
U ( z) = R c ⋅ U ( z r )
(ln (z r ) − ln(z 0 ) )
where:
Rc reduction factor, see 3.4.1.
z elevation above the seabed
zr reference measurement height (in the outer zone)
z0 bottom roughness parameter to be taken from
Table 3-1:
Recommended Practice DNV-RP-F105, March 2002
Page 16

Table 3-1 Seabed roughness 3.2.10 For ULS, 1 min average values should be applied.
The 1 minute average values may be established from 10
Seabed Roughness z0 (m) or 30 min average values as follows:
Silt ≈ 5 10-6  (1 + 1.9 ⋅ I c ) ⋅ U10min
U1min = 
fine sand ≈ 1 10-5 (1 + 2. 3 ⋅ I c ) ⋅ U30 min
Medium sand ≈ 4 10-5
where Ic is the turbulence intensity defined below.
coarse sand ≈ 1 10-4
Gravel ≈ 3 10-4 3.2.11 The turbulence intensity, Ic, is defined by:
Pebble ≈ 2 10 -3 σ
Ic = c
Cobble ≈ 1 10-2 Uc
Boulder ≈ 4 10-2 where σc is the standard deviation of the velocity fluctua-
tions and Uc is the 10 min or 30 min average (mean) veloc-
3.2.7 If no detailed analyses are performed, the mean ity (1 Hz sampling rate).
current values at the free span location may assume the
values at the nearest suitable measurement point. The flow
(and macro-roughness) is normally 3D and transformation 3.2.12 If no other information is available, the turbu-
lence intensity should be taken as 5%. Experience indi-
of current characteristics should account for the local bot-
cates that the turbulence intensity for macro -roughness
tom topography e.g. guided by numerical simulations.
areas is 20-40% higher than the intensity over a flat seabed
with the same small-scale seabed roughness. The turbu-
3.2.8 For conditions where the mean current is spread
lence intensities in a rough seabed area to be applied for
over a small sector (e.g. tide-dominated current) and the
in-line fatigue assessment may conservatively be taken as
flow condition can be assumed to be bi-dimensional, the
typical turbulence intensities over a flat bottom (at the
following model may be applied in transforming the mean
same height) with similar small-scale seabed roughness.
current locally. It is assumed that the current velocity U(zr )
in the outer zone is known, see Figure 3-1. The velocity
profile U(z*) at a location near the measuring point (with 3.2.13 Detailed turbulence measurements, if deemed
essential, should be made at 1 m and 3 m above the sea-
zr*>zr) may be approximated by:
bed. High frequency turbulence (with periods lower than 1
(ln (z *) − ln(z m ) ) minute) and low frequency turbulence must be distin-
U (z*) = U ( z r )
(ln (z )− ln(z ))
guished.
*
r m
3.3 Short-term wave conditions
The “macro-roughness” parameter zm is given by:

z∗r 3.3.1 The wave induced oscillatory flow condition at

ln(z m ) = ln(z ∗r ) −
(z
zm is to be taken less than 0.2.
∗
r )
− zr +
zr
(ln(z r ) − ln(z o ) )
the free span level may be calculated using numerical or
analytical wave theories. The wave theory shall be capable
of describing the conditions at the pipe location, including
effects due to shallow water, if applicable. For most prac-
tical cases, linear wave theory can be applied. Wave
boundary layer effects can normally be neglected.

3.3.2 The short-term, stationary, irregular sea states

Outer Zone U0 Measuring Point may be described by a wave spectrum Sηη (ω) i.e. the
Inner Zone Iso-line for horisontal power spectral density function of the sea surface eleva-
Zr z mean velocity tion. Wave spectra may be given in table form, as meas-
z*r ured spectra, or in an analytical form.
z*

Seabed profile 3.3.3 The JONSWAP or the Pierson-Moskowitz spec-

trum is often appropriate. The spectral density function is:
  ω− ω 
2 
 
exp  − 0 .5
p

 5 ω    σ ωp  
 
Figure 3-1 Definitions for 2D model S ηη (ω) = αg 2 ω−5 exp − ( ) −4  γ  
 4 ωp 
 
3.2.9 It is recommended to perform current measure- where
ments with 10 min or 30 min averages for use with FLS.

ω =2π/Tw is the angular wave frequency.

Tw Wave period.
 Recommended Practice DNV-RP-F105, March 2002
Page 17

Tp Peak period. • Mean zero up-crossing period of oscillating flow at

ωp =2π/Tp is the angular spectral peak frequency pipe level:
g Acceleration of gravity M0
Tu = 2π
The Generalised Phillips’ constant is given by M2

• The bandwidth parameter ε:

5 H s ωp
2 4
α= ⋅ ⋅ (1 − 0. 287 ln γ )
16 g2 M 22
ε = 1−
M 0M4
The spectral width parameter is given by

 0. 07 if ω ≤ ω p The velocity process is narrow-banded for ε → 0 and

σ =
 0.09 else broad banded for ε → 1 (in practice the process may be
considered broad-banded for ε larger than 0.6).
The peak-enhancement factor is given by:

 5 ϕ ≤ 3. 6 US, Tu and ε may be taken from Figure 3-2 to Figure as-

 Tp suming linear wave theory.
γ = exp(5 .75 − 1.15 ϕ) 3 .6 < ϕ < 5 ; ϕ =
 ϕ≥5
HS
 1
0.50
0.45
where Hs is to be given in metres and Tp in seconds.
0.40
The Pierson-Moskowitz spectrum appears for γ = 1.0. 0.35
0.30
UST n/H s

0.25
3.3.4 Both spectra are describing wind sea conditions 0.20
that are reasonable for the most severe seastates. However, 0.15 0.5
moderate and low sea states, not dominated by limited γ=5.0 T n=(h/g)
0.10
γ=1.0
fetch, are often composed of both wind-sea and swell. A 0.05
two peak (bi-modal) spectrum should be considered to 0.00
account for swell if considered important. Further details 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
may be found in Classification Note No. 30.5. Tn /Tp

Figure 3-2 Significant flow velocity amplitude at pipe

3.3.5 The wave induced velocity spectrum at the pipe
level, US
level SUU (ω) may be obtained through a spectral transfor-
mation of the waves at sea level using a first order wave
theory: 1.4
S UU ( ω) = G 2 (ω) ⋅ S ηη ( ω) 1.3
γ=5.0
G2 (ω) is a frequency transfer function from sea surface 1.2
γ=1.0
elevation to wave induced flow velocities at pipe level 1.1
Tu/Tp

given by:
1

ω ⋅ cosh( k ⋅ ( D + e) )
G( ω) = 0.9
sinh(k ⋅ h) 0.8 0.5
Tn =(h/g)
Where h is the water depth and k is the wave number es- 0.7
tablished by iteration from the transcendental equation: 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Tn /Tp

ω2 ⋅ h
kh = coth(k ⋅ h ) Figure 3-3 Mean zero up-crossing period of oscillating
g flow at pipe level, Tu

3.3.6 The spectral moments of order n is defined as:

∞
M n = ∫ ωn SUU ( ω)dω
0

The following spectrally derived parameters appear:

• Significant flow velocity amplitude at pipe level:
US = 2 M0

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 18

Γ is the gamma function; see 3.5.1 and s is a spreading

0.7 parameter, typically modelled as a function of the sea
state. Normally s is taken as a real number, between 2
0.6 γ=1.0 and 8; 2 ≤ s ≤ 8. If no information is available, the most
Bandwidth parameter,ε

γ=5.0
0.5 conservative value in the range 2-8 shall be selected.
0.4
1
0.3 0.9

0.2 0.8

Reduction factor RD
0.7
0.1 0.5
Tn =(h/g) 0.6
0 s=2
0.5 s=4
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Tn/Tp 0.4 s=6
s=1000
0.3
Figure 3-4 Bandwidth parameter for flow velocity at 0.2
sin( θrel)
pipe level, ε 0.1
0
0 20 40 60 80 100 120 140 160 180
Relative pipeline-wave direction, θrel

3.4 Reduction functions

Figure 3-5 Reduction factor due to wave spreading and
3.4.1 The mean current velocity over a pipe diameter directionality
(i.e. taken as current at e+D/2) apply. Introducing the ef-
fect of directionality, Rc becomes: 3.5 Long-term environmental modelling
R c = sin(θ rel )
3.5.1 A 3-parameter Weibull distribution is often ap-
where θrel is the relative direction between the pipeline propriate for modelling of the long-term statistics for the
direction and the current flow direction. current velocity Uc or significant wave height, Hs . The
Weibull distribution is given by:
3.4.2 In case of combined wave and current flow the
  x − γ β 
seabed roughness is increased from the non-linear interac- FX ( x ) = 1− exp  −   
tion between wave and current flow. The modified veloc-   α  
 
ity profile and hereby-introduced reduction factor may be
taken from DNV-RP-E305. where F(•) is the cumulative distribution function and α is
the scale, β is the shape and γ is the location parameter.
3.4.3 The effect of wave directionality and wave Note that the Rayleigh distribution is obtained for β=2 and
spreading is introduced in the form of a reduction factor on an Exponential distribution for β=1.
the significant flow velocity, i.e. projection onto the veloc-
ity normal to the pipe and effect of wave spreading. The Weibull distribution parameters are linked to the sta-
tistical moments (µ: mean value; σ: standard deviation; δ:
UW = US ⋅R D skewness) as follows:
 1
The reduction factor is given by; see Figure 3-5. µ = α Γ1 +  + γ
 β
π /2 2
 2  1 
RD = ∫ w (β) sin (θ rel − β )dβ
2
σ = α Γ 1+  − Γ 1+ 
−π / 2  β  β
 α    3   1  
3 3
where θrel is the relative direction between the pipeline  1   2
δ =   ⋅ Γ1 +  − 3Γ1 + Γ 1+  + 2Γ 1+ 
direction and wave direction  σ    β   β  β  β  
Γ is the Gamma function defined as
3.4.4 The wave energy spreading (directional) function
∞
given by a frequency independent cosine power function Γ ( x) = ∫ tx −1e− t dt
is: 0

 s
 π Γ1 +  3.5.2 The directional (i.e. versus θ) or omni-directional
 k coss (β) β < 1  2
w (β) =  w 2 ; kw = current data can be specified as follows:
 π 1 s 
0 else Γ +  • A Histogram in terms of (Uc, θ) versus probability of
2 2 occurrence
 Recommended Practice DNV-RP-F105, March 2002
Page 19

The fatigue analysis is based on the discrete events in

the histogram. The corresponding Return Period Va l- 3.6 Return Period Values
ues (RPV) are estimated from the corresponding ex-
ceedance probability in the histogram or from a fitted
3.6.1 Return period values are to be used for ULS con-
pdf, see section 3.6.
ditions. A Return Period Value (RPV) xc is defined as:
• A long term probability density function (pdf)
1
The corresponding Return Period Values for 1, 10 and F(x c ) = 1 −
N
100 year are established from section 3.6.
• Based on Return Period Values where N is the number of independent events in the return
period (e.g. 100 year). For discrete directions, N may be
Distribution parameters for an assumed distribution
taken as the total number of independent events times the
e.g. Weibull, are established using e.g. 3 equations
sector probability.
(for 1, 10 and 100 year) with 3 unknowns (α, β and γ).
This is, in principle, always feasible but engineering The time between independent events depends on the envi-
judgement applied in defining return period values can ronmental condition. For currents, this time is often taken
lead to an unphysical underlying long-term pdf. as 24 hours, whereas the time between independent sea-
states (described by Hs ) normally may be taken as 3-6
3.5.3 The wave climate at a given location may be hours.
characterised by a series of short-term sea-states. Each
short-term sea state may be characterised by Hs , Tp , and the 3.6.2 For a Weibull distributed variable the return pe-
main wave direction θ, measured relative to a given refe r- riod value is given by:
x c = α (ln (N ))1/ β + γ
ence direction
The directional (i.e. versus θ) or omni-directional signifi-
cant wave height may be specified as follows:
• A scatter diagram in terms of Hs , Tp , θ 3.6.3 In case the statistic is given in terms of a scatter
diagram, a long term Weibull distribution (α, β, γ) is es-
The fatigue analysis is based on the discrete sea-states tablished from 3.5.1 using statistical moments derived di-
reflected in the individual cells in the scatter diagram. rectly from the scatter diagram as follows:
• A histogram in terms of (Hs , θ) versus probability of µ = ∑ (H s ) ⋅ PHS
HS
occurrence
σ = ∑ (H s − µ) ⋅ PH S
2
The fatigue analysis is based on the discrete events for
Hs in the histogram. The corresponding peak period is HS

∑ (H s − µ) ⋅ PHS
3
assumed on the versatile form
HS
δ =
Tp = C T (H s )α T σ3
Where 6 ≤ CT ≤ 8 and 0.3 ≤ α T ≤ 0.5 are location spe-
cific Where PHs is the discrete occurrence probability. The same
• A long term probability density function (pdf) principle apply for current histograms.
The corresponding Return Period Values (RPV) for 1,
10 and 100 year are established from section 3.6. 3.6.4 The return period value for to be used for direc-
tional data is taken as the maximum projected flow veloc-
• Based on Return Period Values ity, i.e.
The corresponding Weibull distribution is established ( )
max x c, i ⋅ R D (θ rel , i , s ) / R D ( θ rel = 0, s )
i =1.. n
from 3.6.2 using 3 equations (xc for 1, 10 and 100
year) with 3 unknowns (α, β and γ). This is, in princi- where RD is a reduction factor defined by 3.4.3, θrel,i is
ple, always feasible but engineering judgement ap- the relative direction between the pipeline direction
plied in defining return period values can correspond and the flow direction for direction i. For current flow
to an unphysical Weibull pdf. s>8.0 may be applied.

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 20

4.1.5 The Keulegan-Carpenter number is defined as:

4 Response Models
Uw
KC =
4.1 General fwD

where fw is the wave frequency.

4.1.1 Amplitude response models are empirical models
providing the maximum steady state VIV amplitude re- 4.1.6 The current flow velocity ratio is defined by:
sponse as a function of the basic hydrodynamic and struc-
tural parameters. The response models provided herein Uc
have been derived based on available experimental labora- α=
Uc + Uw
tory test data and a limited amount of full-scale tests for
the following flow conditions:
• In-line VIV in steady current and current dominated 4.1.7 The stability parameter, KS, representing the
conditions; damping for a given modal shape is given by:
• Cross-flow VIV induced in-line motion;
• Cross-flow VIV in steady current and combined wave 4π me ζ T
and current conditions; KS =
ρ ⋅ D2
The response models are in agreement with the generally
accepted concept of VIV. where:
ρ Water density
4.1.2 In the response models, in-line and cross-flow ζT Total modal damping ratio
vibrations are considered separately. Damage contribu- me Effective mass, see 6.8.3
tions from both first and second in-line instability regions
in current dominated conditions are implicit in the in-line
model. Cross-flow induced additional in-line VIV result- 4.1.8 The total modal damping ratio, ζT, comprises
ing in possible increased fatigue damage is considered • structural damping, ζstr, see section 6.3.10
approximately. • soil damping, ζsoil , For screening purposes ζsoil = 0.01
may be assumed. For details, see 7.2.10.
4.1.3 The amplitude response depends on a set of hy- • hydrodynamic damping, ζh. For VIV within the lock-
drodynamic parameters constituting the link between the in region, the hydrodynamic modal damping ratio ζh is
environmental data and the Response Models: norma lly to be taken as zero, i.e. ζh = 0.00
• Reduced velocity, VR
• Keulegan-Carpenter number, KC
• Current flow velocity ratio, α
• Turbulence intensity, Ic, see 3.2.11. 4.2 Marginal Fatigue Life Capacity
• Flow angle, relative to the pipe, θrel
• Stability parameter, KS 4.2.1 For cross-flow VIV, the marginal fatigue capac-
Note that the Reynolds number, Re, is not explicit in the ity against VIV in a single sea-state characterised by (Hs ,
evaluation of response amplitudes. Tp , θ) is defined by, see section 2.4:

1
T Hs, T p, θ =
(f )dF
4.1.4 The reduced velocity, VR, is in the general case RM, CF
∞
⋅ S CF
m
with combined current and wave induced flow, defined as:
∫
v
Uc
0 a
U + Uw
VR = c where
f0D
SCF Cross-flow stress range defined in 4.4
where
fv Vibration frequency; see 4.2.3.
f0 Natural frequency for a given vibration mode a Fatigue constant, depending on the relevant
Uc Mean current velocity normal to the pipe; see stress range, see 2.4.3
section 3.4. m Fatigue exponent, depending on the relevant
Uw Significant wave induced flow velocity; see stress range, see 2.4.3
section 3.4.
D Outer pipe diameter
The integral ∫ (...) dFu c indicates integration over the long-
term dis tribution for the current velocity represented by a
Weibull distribution, or histogram.
 Recommended Practice DNV-RP-F105, March 2002
Page 21

4.2.2 For the in-line direction, the marginal fatigue 4.3.4 (A y /D) is defined as the maximum in-line VIV
capacity against VIV in a single sea-state characterised by response amplitude (normalised with D) as a function of
(Hs , Tp , θ) is taken as: VR and KS, see Figure 4-1 for illustration. The corre -
sponding standard deviation may be obtained as (A Y
1 /D)/√2.
, Tp ,θ =
RM , IL
THs m
 S A 
0.20
f ⋅ max  S IL; CF IL 
 2 ACF 
0.18
∞ v Riθ,1 = 1.0

Ksd =0.00

Inline VIV Amplitude (Ay/D)

0.16
dFUc Riθ,2 = 1.0
a 0.14 γon = 1.0
0
Ksd =0.25
0.12
where 0.10 Ksd =0.50

SIL In-line stress range defined in 4.3 0.08 Ksd =0.75

A IL Stress due to unit diameter in-line mode shape 0.06 Ksd =1.00

deflection; 0.04 Ksd =1.25

A CF Stress due to unit diameter cross-flow mode 0.02 Ksd =1.50

shape deflection 0.00

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Reduced Velocity Vrd (=Vr*γ f)
The in-line stress range is taken as the maximum of:
• The in-line VIV stress range Sin Figure 4 -1 Illustration of the in-line VIV Response
• The in-line stress range corresponding to a figure 8 or Ampl itude versus VRd and KSd.
half-moon motion, i.e., stress induced by 50% of the
cross-flow induced VIV amplitude. 4.3.5 In the evaluation of (A Y /D) the design values for
the reduced velocity and stability parameter shall be ap-
4.2.3 The dominating vibration frequency, fv , is to be plied:
taken as: VRd = VR γ f
• fv = f0,IL for in-line VIV K
• fv = f0,CF for cross-flow VIV K sd = s
γk
• fv = 2⋅f0,CF for cross-flow induced in-line motion
where f0,il and f0,CF are the in-line and cross-flow natural where γ f and γk are safety factors related to the natural fre-
vibration frequencies. quency and damping respectively. In addition, an onset
safety factor is needed, see section 2.6.
4.3 In-line Response Model
4.3.6 The Response Model can be constructed from the
4.3.1 The in-line response of a pipeline span in current co-ordinates in Figure 4-2:
dominated conditions is associated with either alternating   1.0 
or symmetric vortex shedding. Contributions from both the   
 for K sd < 0. 4
first in-line instability region and the second instability   γ on 
 0. 6 + K sd 
region are included in the model. VRIL, onset =   for 0. 4 < K sd < 1.6

 γ on 
4.3.2 The amplitude response depends mainly on the   2.2 
   for K sd > 1.6
  γ on 
reduced velocity, VR, the stability parameter, KS, the tur-
bulence intensity, Ic, and the flow angle, θrel relative to the
 A Y,1 
pipe. Further, mitigation effects from the seabed prox- VRIL,1 = 10 ⋅   + VRIL, onset

imity, (e/D) is conservatively not included.  D 
 A Y, 2 
4.3.3 The in-line VIV induced stress range SIL is calcu- VRIL, 2 = VRIL, end − 2 ⋅  

lated by the Response Model:  D 
4.5 − 0. 8 K sd for K sd < 1. 0
VRIL, end =
S IL = 2 ⋅ A IL ⋅ ( A Y / D ) ⋅ ψ α, IL ⋅ γ s  3.7 for K sd ≥ 1. 0
 A Y ,1    K sd   A Y, 2  
    
 D  = max  0.18 ⋅ 1 − 1. 2  ⋅ R Iθ,1 ;  D  
Where
A IL Unit stress amplitude (stress due to unit diame -       
ter in-line mode shape deflection);  A Y ,2   K sd 
 
ψα,IL Correction factor for current flow ratio α  D  = 0. 13 ⋅ 1 − 1. 8  ⋅ R Iθ,2
   
γs Safety factor to be multiplied on the stress
range

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 22

AY 
  4.4 Cross-flow Response Model
 D 
 IL  A Y ,1  
 VR 1, ;  
  D 
   4.4.1 Cross-flow VIV are affected by several parame-
 IL  A Y,2  
Inline VIV Amplitude

 V R ,2 ;  
  D  
   ters, such as the reduced velocity VR, the Keulegan-
Carpenter number, KC, the current flow velocity ratio, α,
the stability parameter, KS, the seabed gap ratio, (e/D), the
Strouhal number, St and the pipe roughness, (k/D), among
others. Note that Reynolds number, Re, is not explicit in
(V ) the model.
)
IL
;0
R ,onset
VRIL,end ;0

Reduced Velocity
4.4.2 For steady current dominated flow situations,
onset of cross-flow VIV of significant amplitude occurs
Figure 4 -2 Response Model generation principle. typically at a value of VR between 3.0 and 5.0, whereas
maximum vibration levels occurs at a value of VR between
5 and 7. For pipes with low specific mass, wave dominated
4.3.7 The reductions RIθ,1(Ic,θrel) and RIθ,2 (Ic) accounts flow situations or span scenarios with a low gap ratio,
for the effect of the turbulence intensity and angle of at- cross-flow vibration may be initiated for VR between 2 and
tack (in radians) for the flow, see Figure 4-3. 3.

π 
R Iθ,1 = 1 − π 2  − 2 ⋅θ rel ( I c − 0.03 )
4.4.3 The cross-flow VIV induced stress range SCF due
0 ≤ R Iθ,1 ≤ 1 to a combined current and wave flow is assessed using the
2 
(I − 0.03) following Response Model:
R Iθ, 2 = 1. 0 − c 0 ≤ R Iθ, 2 ≤ 1
0. 17
S CF = 2 ⋅ A CF ⋅ ( A Z / D) ⋅ R k ⋅ γ s
1
Where
0.9
Ri θ, 1 θrel=60ο
0.8 A CF
Unit stress amplitude (stress due to unit diame -
0.7 ter cross-flow mode shape deflection);
0.6 Rk Amplitude reduction factor due to damp-
0.5 ο
ing
Ri θ, 1 θrel=45
0.4
γs Safety factor to be multiplied on the stress
ο
range
0.3 Ri θ,1 θrel =30
The cross-flow VIV amplitude (A Z/D) in combined current
0.2
R iθ ,2 and wave flow conditions may be taken from Figure 4-4.
0.1 ο
R iθ ,1 θrel=0 all angles
The figure provides characteristic maximum values. The
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
corresponding standard deviation may be obtained as (A Z
Turbulence Intensity, Ic /D)/√2.
1.5
Figure 4-3 Reduction function wrt turbulence intensity 1.4 α > 0.8 ; all KC ψproxi,onset = 1.0
1.3 ψ trench,onset = 1.0
Cross-Flow VIVAmplitude (AZ/D)

and flow angle 1.2

1.1 ψ mass,onset = 1.0
1.0 α < 0.8 ; KC > 30
ψ a,onset = 1.0
0.9
4.3.8 ψα,IL is a reduction function to account for re - 0.8
0.7
α < 0.8 ; KC < 10 γ on = 1.0

duced in-line VIV in wave dominated conditions: 0.6

 0.0 for α < 0.5 0.5


0.4

ψ α, IL = ( α − 0.5) / 0.3 for 0.5 < α < 0.8 0.3

0.2 VRcr,onset
 for α > 0. 8 0.1
 1.0 0.0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Reduced Velocity Vrd (=Vr*γ f )
Hence, in case of α < 0.5 in-line VIV may be ignored.

Figure 4-4 Basic Cross-Flow Response Model

 Recommended Practice DNV-RP-F105, March 2002
Page 23

4.4.4 The amplitude response (A Z/D) as a function of α 4.4.8 ψα,onset is a correction factor accounting for the
and KC can be constructed from: current wave ratio:
3 ⋅ ψproxi,onset ⋅ ψmass,onset ⋅ ψα ,onset ⋅ ψtrench,onset  α
VRCF,onset = 1+ for α < 0.5
γon ψ α, onset = 3

VRCF,1 =5 1.167 else
 9  A 
VRCF,2 = VRCF, end −   ⋅  Z,1 
 1.3   D 
4.4.9 ψtrench,onset is a correction factor accounting for the
VRCF,end = 16 effect of a pipe located in/over a trench:
 1.3 α f 0.8 all KC

 A Z,1   0.7 KC p 10 ∆
  =  ψ trench,onset = 1 + 0.5
 D  0.7 + 0.01 ⋅ (KC − 10 ) α ≤ 0.8 10 ≤ KC ≤ 30 D
 0.9 KC f 30 where ∆/D denotes a relative trench depth given by:
 A Z, 2  A 
  =  Z,1  ∆ 1. 25d − e
 D   D  =
D D
∆
where 0 ≤ ≤ 1
D
 C F  A Z,1    C F  A Z ,2  
V ;   V ;  The trench depth d is to be taken at a width equal to 3
 R ,1  D    R , 2  D  
   
outer diameters. ∆/D = 0 corresponds to a flat seabed or a
Cross-Flow VIVAmplitude

pipe located in excess of D/4 above the trench, i.e. the pipe
is not affected by the presence of the trench, see Figure
4-6. The restriction ∆/D < 1.0 is applied in order to limit
the relative trench depth

3D
(V CF
R ,onset ;0.1 )
VRCF,end ;0 )
(2.0;0.0 ) D
Reduced Velocity

Figure 4-5 Response Model generation principle d

e
4.4.5 The reduced onset velocity for cross-flow VIV,
CF
VR ,onset depends on the seabed proximity, trench geometry Figure 4-6 Definition of Trench Factor

and current flow ratio α, whereas the maximum amplitude

is a function of α and KC. 4.4.10 The characteristic amplitude response for cross-
flow VIV may be reduced due to the effect of damping.
4.4.6 ψproxi,onset is a correction factor accounting for the The reduction factor, Rk is given by:
seabed proximity:
1 e e 1 − 0.15 K sd for K sd ≤ 4
 (3 + 1. 25 ) for < 0.8 Rk = −1 .5
ψ proxi , onset =  4 D D  3.2 K sd for K sd f 4
 1 else

4.4.11 The normalised amplitude curves in Figure 4-4 to

4.4.7 ψmass,onset is a correction factor accounting for the a large degree embody all available test result. In addition,
specific mass (gravity) of the pipe: the following comments apply:
• The response for small gap ratio (e/D<< 1) is associ-
1 1 ρ s ρs ated with one-sided vortex shedding and may not be
 + for < 1. 5 characterised by VIV parameters as VR and KC. How-
ψ mass ,onset =  2 3 ρ ρ
 ever, the indicated response curve is considered con-
1 else
servative in general.
where ρ s/ρ is the specific mass. • The response for low KC numbers in the cross-flow
response model are not in a narrow sense related to
the VIV phenomena but rather linked to wave induced
water particle motions. Typical maximum response at
VR between 2.5 and 3.0 occur at Tu /T0 ≈ 2. Tu is the

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 24

wave induced flow period at pipe level and T0 is the

natural period.

4.4.12 Potential vibrations at low KC numbers must be

accounted for and care should be observed in case:
KC
VR > and 3 < KC < 9
3⋅ (1− α )

This corresponds to rare cases where Tu < 3T0 . If violated,

the criticality should be evaluated using an appropriate
force model.
 Recommended Practice DNV-RP-F105, March 2002
Page 25

5.2.2 The short term fatigue capacity against direct

5 Force Model wave actions in a single sea-state characterised by (Hs , Tp ,
θ) is given in the following form:
5.1 General a1 ⋅ S− m1
THFM
S , TP , θ
= ×
f v ⋅ κ RFC(m1 )
5.1.1 In principle, force models may be used for both −1
  m   S  2   m   S  2 
vortex induced and direct wave and current dominated  G 1 1 + 1 ,  sw   + χ ⋅ G 2 1 + 2 ,  sw  
loads if appropriate formulations of force models exist and   2   S    2   S  
 
reliable and consistent data are available for calibration. κRFC (m 2 ) a 1 ( m 2 −m 1 )
For cross-flow VIV, generally applicable force models do χ = S
κ RFC (m1 ) a 2
not exist and empirical response models presented in 4.4 S = 2 2 σ s ⋅ γs
reflecting observed pipeline response in a variety of flow
conditions is at present superior. Where
σS Standard deviation of stress amplitude
5.1.2 A force model based on the well-known Mori- fv Vibration frequency
son’s equation for direct in-line loading is considered a1 , a 2 Fatigue constant; see 2.4.3
herein. Both time domain (TD) and frequency domain
(FD) solutions are allowed. A time domain solution may m1 , m2 Fatigue exponent; see 2.4.3
account for all significant non-linearities but is in general Ssw Stress range, for which change in slope oc-
very time consuming if a large number of sea-states are to curs; see 2.4.3
∞
be analysed. For fatigue analyses, a frequency domain
G 1 (ϕ, x ) = ∫ e− t t ϕ −1 dt is the
solution (if thoroughly verified) is more tractable since it x
facilitates analyses of a very large number of sea-states at Complementary incomplete Gamma function
a small fraction of the time required for a time domain x
− t ϕ −1
G 2 ( ϕ, x) = ∫ e t
solution. dt is the
0
5.1.3 In this document, a complete FD approach for Incomplete Gamma function
short-term fatigue analyses is presented. Recommended γs Safety factor on stress range, see 2.6
procedures for state-of-the-art Time Domain (TD) short-
term damage calculation may be found in DNV-OS-F201. 5.2.3 The standard deviation of the wave induced stress
A simplified assessment method is given in 5.3. amplitude σS is given by the square root of the spectral
moment of the 0th order defined by 5.2.6.
5.2 FD solution for In-line direction
σS = M0
5.2.1 The recommended FD solution for the short term-
fatigue damage due to combined current and direct wave
actions in a single sea-state is based on: 5.2.4 The characteristic vibration frequency of consid-
• Palmgren-Miner approach using SN-curves; ered pipe stress response, fv , is taken equal to the mean up-
• linearisation scheme for drag term in the Morison crossing frequency defined by:
equation based on conservation of damage;
• effect of co-linear mean current included in linearis a- 1 M2
tion term; fv ≈
2π M0
• narrow banded fatigue damage with semi -empirical
correction to account for wide-band characteristic;
The formulation presented in this document has been suc- M 0 and M2 is defined by 5.2.6.
cessfully verified against comprehensive time domain
simulations using Rain flow Counting techniques, see e.g. 5.2.5 The rain flow counting correction factor, κRFC,
Mørk &Fyrileiv, 1998. The formulation is based on the accounts for the “exact” wide-banded damage, i.e. correct-
following assumptions: ing the implicit narrow-banded Rayleigh assumption for
• the main damage contribution comes from lowest the stress amplitudes to provide results similar to those
natural mode, i.e. the excitation frequency is far from arising from a state-of-the-art rain flow counting tech-
the natural frequency for the higher order modes; nique. The Rain Flow Counting factor κRFC is given by:
• the effective mass, me , and standard deviation of the
flow velocity σU is invariant over the free span length, κRFC ( m ) = a κ + (1 − a κ )(1 − ε) b κ
i.e. for span length less than the dominant wavelength. where
a κ = 0. 926 − 0. 033m
b κ = 1.587 m − 2.323

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 26

The bandwidth parameter ε is defined by: In lieu of more detailed data, λmax may be taken as:
A
M 22 λ max = IL λ1
ε = 1− D
M 0M4
where A IL is given by 6.8.4.

The process (spectrum) is narrow-banded for ε → 0 and 5.2.8 The linearisation constant b is given by:
broad banded for ε → 1 (in practice the process may be
considered broad-banded for ε larger than 0.6). Uc
b = 2.11 ⋅ σ u ⋅ g c ( )
σu
5.2.6 The n th response spectral moment is given by:
∞ where Uc is the mean current and σu =Uw /2 is the standard
M n = ∫ ω n S SS ( ω) dω deviation of the wave induced flow velocity. g c(•) is a cor-
0 rection function accounting for the effect of a steady cur-
rent given by:
Where SSS(ω) is the one-sided stress response spectral
density function given by:   1
g c ( x) = 2π  ϕ(x ) + x ⋅  Φ(x )−  ;
( )
SSS (ω) = R 2D ⋅ b 2 g 2D + ω2 g 2I ⋅ G 2 (ω) ⋅S ηη (ω) ×

1
− x2
 2

ϕ(x ) =
1 2
λ2max 2π
e

m 2e (
(ω20 − ω ) + (2ζ T ω0 ω)
2 2 2
) Φ ( x) = ∫
x
−∞
ϕ( x) dx

ϕ(x) is the Gaussian probability density function and Φ(x)

Where is the corresponding distribution function.
RD Factor accounting for the wave spreading and
direction; see 3.4.3. 5.2.9 The (linearized) hydrodynamic damping ratio ζh
b Linearisation constant; see 5.2.8. is given by:
gD Drag force term; see 5.4.1. 1 σ u gD
ζh = g c  C  λ 1
U
gI Inertia force term; see 5.4.1.
2π m e f 0  σU 
G(ω) Frequency transfer function; see 3.3
Sηη Single -sided wave elevation spectrum; see 3.3
ω0 = 2πf0 /γ f is the angular natural frequency 5.3 Simplified Fatigue Assessment
ζT Total damping ratio fro m
• structural damping; see 6.3.10
5.3.1 In situations where quasi-static stress response
• soil damping; see 7.2.10. can be assumed (when the wave period is far larger than
• hydrodynamic damping; see 5.2.9. the natural vibration period of the span), a simplified fa-
me Effective mass per unit length incl. added mass; tigue assessment may be tractable rather than a complete
see 6.8.3 TD or FD approach.

5.2.7 λmax is an equivalent stress factor given by: In such cases, the short term fatigue capacity against direct
wave actions in a single sea-state characterised by (Hs , Tp ,
(D s − t )E  ∂ 2 φ1  θ) may be estimated as follows: (cf. 2.4.5)
λ max = (1 + CSF ) λ 1 max  
L  ∂x 
2 2
 
THFM
S , TP ,θ
= a ⋅ S− m Tu
φ1 (x)) 1st mode shape
E Young’s modulus where S is the quasi-static stress range response from a
CSF Concrete stiffness factor; see 6.3.5 direct regular wave load at pipe level using Morison’s
Ds Outer steel pipe diameter equation. Tu is the mean zero upcrossing period in 3.3.6.
t Pipe wall thickness
L Length of mode shape 5.4 Force Coefficients
λ1 Mode shape weighting factor given by:
L
5.4.1 The force P(x,t) per unit length of a pipe free span
∫ φ1 ( x)dx is represented by the Morison’s equation. Assuming that
λ1 = 0
L the velocity of the structure is not negligible compared
∫ φ1 ( x) dx
2
with the water particle velocity Morison’s equation reads:
0
π
λ1 is typically in the order of 1.3 P ( x, t) = g D (U − y& ) U − y& + g I U& − C a ρ D 2 &y&
4
 Recommended Practice DNV-RP-F105, March 2002
Page 27

where The drag load is often of small practical importance for

small KC values and CD,0 may be interpolated for com-
ρ Water density
pleteness for KC < 5, see Figure 5-1.
D Outer pipe diameter
U Instantaneous (time dependent) flow ve-
locity 5.4.6 ψproxi,CD is a correction factor accounting for the
y Pipe lateral displacement seabed proximity:
gD = 0.5ρDCD is the drag force term  0.5
π 0.9 + for e/D < 0.8
gI = ρD 2 C M is the inertia force term ψ proxi , CD =  (1 + 5 ⋅ (e / D) )
4
 1 else
5.4.2 The added mass term in the Morison equation
π ψ k, CD is a correction factor for the pipe rough-
Ca ρD 2&z& is assumed implicit in the effective mass me , see 5.4.7
4
6.7.1. ness.
k
ψ k, CD = 1. 25 + 0. 05 ⋅ ln  
5.4.3 The drag coefficient CD and inertia coefficient CM D
to be used in Morisons equation is a function of :
• the Keulegan Carpenter number, KC; In lieu of detailed documentation of the surface roughness
• the current flow ratio, α; the following values may be applied:
• the gap ratio, (e/D); Pipe surface k [metres]
• the trench depth, (∆/D);
Steel, painted 10-6
• Reynolds number, Re;
• the pipe roughness, (k/D); Steel, un-coated (not rusted) 10-5
In addition also the cross-flow vibration level, (A z/D) in - Concrete 1/300
fluences the drag coefficient. Note that the dependency of Marine growth 1/200 → 1/20
the Reynolds number is embedded in the cylinder rough-
ness effect.
5.4.8 ψtrench,CD is a correction factor accounting for the
effect of a pipe in a trench:
5.4.4 The drag coefficient CD is to be taken as
2∆ 
ψ trench, CD = 1 −  
C D = C D, 0 ⋅ ψ k, CD ⋅ ψ proxi ,CD ⋅ ψ trench,CD ⋅ ψ A,CD 3 D

∆/D is the relative trench depth given by 4.4.9.

5.4.5 CD,0 is the basic drag coefficient in oscillatory
flow for a free concrete coated pipe (KC>5):
5.4.9 ψA,CD is an amplification factor due to cross-flow
vibrations, i.e., proportional to increased projected area to

0.9 ⋅ (1− α ) + KC
5
α ≤ 0.5 mean flow.
C D, 0 =   Ay 
0.45 +
5
α f 0. 5 ψ A, CD = 1 + 2 ⋅  

 KC  D 

For pure current CD,0 = 0.45 (i.e., for KC→∝).

5.4.10 The inertia coefficient CM is to be taken as:

2.0
1.8
C M = C M , 0 ⋅ ψ k ,CM ⋅ ψ proxi ,CM ⋅ ψ trench,CM
1.6
Drag Coefficient CD

1.4 5.4.11 CM,0 is the basic inertia coefficient for a free con-
1.2
α crete coated pipe taken as, see Figure 5-2:
1.0 0.0

5 ⋅ (2 − f (α ) )
0.1
0.2
0.8
C M, 0 = f ( α) +
0.3
ψ k, CD = 1 .0
(KC + 5)
0.4
0.6
ψ proxi, CD = 1.0 0.5

0.4 ψ trench,CD = 1.0

1. 6 − 2 ⋅ α α ≤ 0.5
0.2 ψ A,CD = 1.0 f (α ) = 
0.0  0.6 α f 0. 5
0 5 10 15 20 25 30 35 40
KC

Figure 5-1 Drag coefficient CD versus KC and α

DET NORSKE VERITAS

Recommended Practice DNV- RP- F105, March 2002
Page 28

2.0
α 5.4.12 ψ k ,CM is a correction factor accounting for the
1.8
0.0
1.6 0.1 pipe roughness.
Inertia Coefficient CM

1.4 0.2 ψ k ,CM = 2 − ψ k ,CD

1.2 0.3
1.0 0.4
0.5
0.8 5.4.13 ψproxi,M is a correction factor accounting for the
0.6 ψ k, CM = 1.0 seabed proximity:
0.4 ψ proxi,CM = 1.0
ψ trench, CM = 1.0
0.2
 0.8
 0.84 + for e/D < 0.8
(1 + 5 ⋅ ( e / D) )
0.0
0 5 10 15 20 25 30 35 40 ψ proxi ,M =
KC  1 else
Figure 5-2 Inertia coefficient C M versus KC and α
5.4.14 ψtrench,M is a correction factor accounting for the
effect of a pipe in a trench:
1∆ 
ψ trench, M = 1 −  
3D

∆/D is the relative trench depth given by 4.4.9

 Recommended Practice DNV-RP-F105, March 2002
Page 29

The curves are, in strict terms, only valid for the vertical
6 Structural Analysis (cross-flow) dynamic response but may also be used for
assessment of the horizontal (in-line) response.

6.1 General 6.3 Structural modelling

6.1.1 The following tasks are normally required for
assessment of free spans:
• structural modelling
• modelling of pipe-soil interaction
• load modelling
• a static analysis to obtain the static configuration of
the pipeline
• an eigenvalue analysis which provides natural fre-
quencies and corresponding modal shapes for the in-
line and cross-flow vibrations of the free spans
• a response analysis using a response model or a force
model in order to obtain the stress ranges from envi-
ronmental actions.
6.2 Morphological classification
6.2.1 The objective of the morphological classification
is to define whether the free span is isolated or interacting.
The morphological classification determines the degree of
complexity required of the free span analysis:
• two or more consecutive free spans are considered to
be isolated (i.e. single span) if the static and dynamic
behaviour are unaffected by neighbouring spans;
• a sequence of free spans is interacting (i.e. multi-
spanning) if the static and dynamic behaviour is af-
fected by the presence of neighbouring spans. If the
free span is interacting, more than one span must be
included in the pipe/seabed model.
6.2.2 The morphological classification should in gen-
eral be determined based on detailed static and dynamic
analyses. The classification may be useful for evaluation
of scour induced free spans and in application of approxi-
mate response quantities.
In lieu of detailed data, Figure 6-1 may be used to classify
the spans into isolated or interacting dependent on the soil
types and span and support lengths.
1.1
Interacting
a/L
6.3.1 The structural behaviour of the pipeline shall be
evaluated by modelling the pipeline, seabed and relevant
artificial supports and performing static and dynamic
analyses. This section presents requirements for the struc-
tural modelling.
Soil-pipe interactions are treated in section 7.
6.3.2 A realistic characterisation of the cross-sectional
behaviour of a pipeline can be based on the following as-
sumptions:
• the pipe cross-sections remain circular and plane;
• the stresses may be assumed constant across the pipe-
wall thickness;
• load effect calculation is normally to be performed
using nominal un-corroded cross section values;
• the application of this document is limited to elastic
response, hence plasticity models and effects of two-
dimensional state of stress (axial and hoop) on bend-
ing stiffness need not be considered.
6.3.3 The effect of coating is generally limited to in-
creasing submerged weight, drag forces, added mass or
buoyancy. The positive effect on the stiffness and strength,
see 6.3.5, is normally to be disregarded. If the contribution
of the coating to the structural response is considered sig-
nificant, appropriate models shall be used.
6.3.4 Non-homogeneity of the bending stiffness along
the pipe, due to discontinuities of the coating across field
joints or other effects, may imply strain concentrations that
shall be taken into account.
6.3.5 The stiffening effect of concrete coating may be
accounted for by:
0. 75
 EI 
CSF = k c  conc 

 EI steel 
sand
where CSF denotes the stiffness of concrete coating rela-

1.0 tive to the steel pipe stiffness and (1+CSF) is the stress
Relative length of adjaent span, L

0.9 stiff clay

Isolated concentration factor due to the concrete coating and local-
0.8
very soft clay ised bending. The parameter kc is an empirical constant
0.7
soft/firm clay accounting for the deformation/slippage in the corrosion
0.6
0.5
coating and the cracking of the concrete coating. The
0.4
value of kc may be taken as 0.33 for asphalt and 0.25 for
0.3
PP/PE coating.
0.2 In case the increased stiffness effect is utilised, the in-
La L sh L
0.1
creased bending stresses due to field joints must also be
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
accounted for.
Relative span shoulder length, Lsh/L
The CSF given above is assumed valid for all relevant pipe
Figure 6-1 Classification of free spans diameters, D/t-ratios and concrete strengths, fcn, provided
that that the pipe joint length exceeds 12m, the field joint

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 30
length is 0.5-1.0m and the concrete coating thickness does
not exceed 150mm.
6.3.6 In lieu of detailed data, it is conservative to as-
sume that a girth weld is present in the most heavily
loaded cross-section. This is also a basis for the concrete
stiffening effect given above.
6.3.7 The cross-sectional bending stiffness of the con-
crete coating, EIconc, is the initial, uncracked stiffness.
Young’s modulus for concrete may be taken as:
E conc = 10000 ⋅ fcn 0.3
where fcn is the construction strength of the concrete. Both
Econc and fcn are to be in N/mm2 .
6.3.8 The boundary conditions applied at the ends of
the pipeline section modelled shall be able to simulate the
pipe-soil interaction and the continuity of the pipeline.
Sufficient lengths of the pipeline at both sides of the span
must be included in the model to account for the effects of
side spans, if relevant.
6.3.9 The element length to be used in a finite element
model is dictated by the accuracy required. If the stress
ranges are to be derived from the mode shapes, see 6.8.4,
the accuracy of the stress ranges becomes strongly affected
by the element length, especially at the span shoulders.
Ideally the maximum element length should be found by
reducing the length until the results (natural frequencies
and stresses) converge towards constant values. In practice
this may be difficult to perform and, as guidance, the ele-
ment length should be in the order of the outer diameter of
the pipeline (1D). However, higher order modes and/or
short spans (L/Ds < 30) may require shorter elements.
6.3.10 Structural damping is due to internal friction
forces of the pipe material and depends on the strain level
and associated deflections. If no information is available, a
structural modal damping ratio of
ζstr = 0.005
can be assumed. If concrete coating is present, the sliding
at the interface between concrete and corrosion coating
may further increase the damping to typically 0.01-0.02.
6.3.11 It is recommended to verify the finite element
modelling and the post-processing by comparing the re-
sults from the finite element analysis with the approximate
response quantities of section 6.8 for a single span with
zero effective axial force and L/Ds = 60. The in-line and
cross-flow natural frequencies and stress ranges shall show
similar values within ±5%.
6.3.12 ULS conditions may require a more refined pipe-
soil modelling than the linearized eigen-value analysis due
to sliding at the span supports.
6.4 Functional Loads
6.4.1 The functional loads which shall be considered
are:
• weight of the pipe and internal fluid;
• external and internal fluid pressure;
• thermal expansion and contraction, and
• residual installation forces.
6.4.2 Response calculations must account for the rele -
vant sequence of load application, if important.
6.5 Static analysis
6.5.1 The static configuration is to be determined for
the following conditions if relevant:
• as-laid condition;
• flooded condition;
• pressure test condition, and
• operating condition.
6.5.2 The static analysis should normally account for
non-linear effects such as:
• large displacements (geometric non-linearity);
• soil non-linear response, and
• loading sequence.
6.5.3 The stiffness of the pipeline consists of material
stiffness and geometrical stiffness. The geometrical stiff-
ness is governed by the effective axial force, Seff. This
force is equal to the true steel wall axial force, Ntr, with
corrections for the effect of external and internal pressures:
S eff = N tr − p i A i + p e A e
Where
Ntr “True” steel wall axial force
pi Internal pressure
pe External pressure
Ai Internal pipe (bore) cross section area
Ae External pipe (incl. Coating) cross section area
The effective axial force in a span is difficult to estimate
due to uncertainties in operational temperature and pres-
sure, residual lay tension and axial force relaxation by
sagging, axial sliding (feed-in), lateral buckling, multi-
spanning and significant seabed unevenness. All these
effects should be considered and taken into account if
relevant. Then the most reliable method to estimate the
effective axial force is use of non-linear FE analysis.
As boundary values, the effective axial force for a
completely unrestrained (axially) pipe becomes:
S eff = 0

while for a totally restrained pipe the following effective
axial force apply (if pipe considered thin-walled):
S eff = H eff − ∆ pi Ai (1 − 2ν) − As E∆Tα e
Where
Heff Effective lay tension
∆p i Internal pressure difference relative to laying,
see DNV-OS-F101
As Pipe steel cross section area
∆T Temperature difference relative to laying
αe Temperature expansion coefficient, may be
temperature dependent
6.5.4 In this document, the static environmental loads
are confined to those from near bottom current. If the load
is much smaller than the vertical functional loads, then it
may be disregarded in the analysis. However, for light
pipes or long span lengths it should be considered if rele-
vant.
6.5.5 Load history effects such as the lay tension and
submerged weight during installation will influence the
static deflection and stresses which are mainly determined
by the submerged weight and effective axial force in the
phase considered.
Furthermore, the span geometry such as inclination of the
span shoulders will have a significant influence on the
static stresses and deflection. For this reason, the static
response should be based on survey results (measured de-
flections) and/or FE analysis if considered as critical for
the span assessment.
6.5.6 In addition to the static penetration due to the
submerged weight of the pipeline, the penetration may
increase due to effects from laying, erosional processes
and self-burial.
6.6 Eigen-value analysis
6.6.1 The aim of eigen-value analyses is to calculate
the natural frequencies and corresponding stress due to
associated mode shapes. The analysis is normally complex
and depends on:
• the temporal classification (scour or uneveness in-
duced free span);
• the morphological classification (single or multispan);
• the pipeline condition (i.e. as-laid, water-filled, pres-
sure test and operation);
• the pipe and soil properties;
• the effective axial force and the initial deflected shape
after laying;
• the loading history and axial displacement (“feed-in”)
of the pipe.
In general, it is recommended that the response quantities
be assessed using non-linear FE-analyses conducted over
an appropriate stretch of the pipeline. However,
DET NORSKE VERITAS
Recommended Practice DNV-RP-F105, March 2002
Page 31
approximate response quantities may be applied under
certain limitations, see section 6.8.
6.6.2 Using an FE-approach, the following comments
apply:
• the eigenvalue analysis shall account for the static
equilibrium configuration;
• in the eigenvalue analysis, a consistent linearisation of
the problem must be made;
• the pipe-soil linearisation should be validated;
• the effect of geometric non-linearity on the dynamic
response should be assessed;
• the span support may be assumed invariant during
Vortex-Induced Vibrations (VIV) and motions from
direct wave loading.
6.6.3 For analysis of a pipeline stretch with several
spans and especially with interacting spans, special care
must be paid to the determination of the eigenvalues and
associated eigenvectors. This is due to the potential occur-
rence of very close eigenvalues, especially with respect to
the identification of correct eigenvectors.
6.7 Added Mass
6.7.1 The added mass may be considered as
( )
C a ( e / D) = C M α = 0; KC ↓ 0; e / D − 1
The ↓ symbolise KC approaching 0. Note that the effect of
pipe roughness and trench is not accounted for.
According to section 5.4.10, Ca becomes:
 1. 6
0.68 + for e/D < 0.8
Ca =  (1 + 5 ⋅ ( e / D) )
 1 for e/D ≥ 0.8
where e/D is the span gap ratio. This expression applies for
both smooth and rough pipe surfaces.
6.8 Approximate response quantities
6.8.1 The approximate response quantities specified in
this section may be applied for free span assessment pro-
vided:
• Conservative assumptions are applied with respect to
span lengths, soil stiffness and effective axial force.;
• The span is a single span on a relatively flat seabed,
i.e. the span shoulders are almost horizontal and at the
same level;
• The symmetrical mode shape dominates the dynamic
response (normally relevant for the vertical, cross-
flow response only). Here the following limits apply:
L/Ds < 140
δ/D < 2.5
Note that these are not absolute limits; the shift in
cross-flow response from the symmetrical to the un-
symmetrical mode will depend on the sagging and the
Recommended Practice DNV-RP-F105, March 2002
Page 32

levelling/inclination of the span shoulders. In cases −3 / 2

φ   ∂ φ  
κ( x) = ∂
2 2
where a shift in the cross-flow response is considered 1 +  
as likely, the structural response of the span should be ∂x 2   ∂x  
assessed by using FE analysis including all important
aspects;
• Bar buckling is not influencing on the response, i.e. 6.8.5 In lieu of detailed information, the (unit diameter)
C2 Seff/PE > -0.5; stress amplitude A IL/CF may be estimated by:
• A sensitivity study is performed in order to quantify
the criticality of the assumptions. D ⋅ ( Ds − t ) ⋅ E ∂ 2 φ
A IL / CF = (1 + CSF ) max
2 ∂x 2
Approximate response quantities are considered relevant D ⋅ ( Ds − t ) ⋅ E
in performing efficient screening of FE or survey results in = C 4 (1 + CSF )
order to identify critical spans to be assessed with methods L2eff
that are more accurate, see Fyrileiv & Mørk, (1998).
where t is the steel pipe wall thickness and C4 is a bound-
ary condition coefficient.
6.8.2 The fundamental natural frequency may be ap-
proximated by:
6.8.6 The static bending moment may be estimated by:
q ⋅ L2eff
EI  S δ 
2 M static = C5
f 0 ≈ C1 ⋅ 1 + CSF ⋅ 1 + C2 ⋅ eff + C3     S
1 + C2 ⋅ eff


m eL4eff  PE  D    
  PE 
where where q represents the loading, i.e. the submerged weight
C1 – C3 Boundary condition coefficients of the pipe in the vertical (cross-flow) direction and/or the
E Youngs modulus for steel drag loading in the horizontal (in-line) direction, see 5.4.1.
I Moment of inertia for steel
CSF Concrete stiffness enhancement factor Note that Leff shall be calculated using the static soil stiff-
Leff Effective span length, see 6.8.9 ness in the Leff/L calculation.
me Effective mass, see below
D Outer diameter of pipe Note that, due to historical effects and the local seabed
PE Euler buckling load = (1+CSF)π2 EI/Leff 2 geometry, there is a large uncertainty associated with this
(positive sign) simplified expression, see 6.5.5.
δ Static deflection, normally ignored for in-
line direction). Shall be limited to 4D. Note that the term Seff/PE becomes negative when the ef-
Seff Effective axial force (negative in compres- fective axial force is in compression since PE is defined as
sion) , see section 6.5 positive.

6.8.3 The effective mass, me, is defined by 6.8.7 In case the static deflection is not given by direct
measurement (survey) or estimated by accurate analytical
 ∫ m (s) φ2 (s )ds 
  tools, it may be estimated by:
me =  L 
 ∫ φ ( s) ds 
2
q ⋅ L4eff 1
 L  δ = C6
EI ⋅ (1 + CSF )  S 
where φ(s) is the assumed mode shape satisfying the 1 + C 2 ⋅ eff 
 
boundary conditions and m(s) is the mass per unit length  PE 
including structural mass, added mass and mass of internal where C6 is a boundary condition coefficient.
fluid.
Note that Leff shall be calculated using the static soil stiff-
6.8.4 The unit diameter stress amplitude (stress due to ness in the Leff/L calculation.
unit outer diameter mode shape deflection) may be calcu-
lated by: Note that, due to historical effects and the local seabed
A IL / CF = D
κEI D s = 1 D E D κ geometry, there is a large uncertainty associated with this
s
I 2 2 simplified expression, see 6.5.5.
where Ds is the steel pipe diameter and D is the outer pipe
6.8.8 The coefficients C1 to C6 are given in the table
diameter (including any coating). κ is the curvature of the below for different boundary conditions. For multi-
assumed mode shape satisfying the boundary conditions
spanning scenarios the choice of coefficient should be
mode shape:
supported by dedicated FE-analyses.
 Recommended Practice DNV-RP-F105, March 2002
Page 33

Table 6-1 Boundary conditions coefficients

Pinned- 6.8.9 The Leff/L term used above accounts for the effec-
Fixed- Single span on seabed
Pinned 2) Fixed 3) tive span length in order to consider the span as fully
fixed. This ratio decreases as the L/Ds ratio and soil stiff-
C1 1.57 3.56 3.56 ness increase.
C2 1.00 0.25 0.25
C3 0.8 1) 0.2 1) 0.4 1) The Leff/L term is given by (for reference see Hobbs,
C4 4.39 14.1 Shoulder: 14.1(L/Leff)2 1986):
Mid-span: 8.6
C5 1/8 1/12 1  4. 73
Shoulder:  for β ≥ 2. 7
 − 0. 066β + 1.02β + 0.63
2
18 (L eff / L) 2 − 6 L eff
=
Mid-span: 1/24 L  4. 73 for β < 2.7
C6 5/384 1/384 1/384 
 0.036 β + 0.61β + 1.0
2
1) Note that C3 =0 is normally assumed for in-line if the
steady current is not accounted for.
2) For pinned-pinned boundary condition Leff is to be
 K ⋅ L4 
replaced by L in the above expressions also for PE . β = log10  
 (1+ CSF ) EI 
3) For fixed-fixed boundary conditions, Leff/L=1 per  
definition.
4) C5 shall be calculated using the static soil stiffness in where K is the relevant soil stiffness (vertical or horizon-
the Leff/L calculation. tal, static or dynamic).

6.8.10 The boundary coefficients in Table 6-1 based on

the effective span length are found appropriate for fatigue
assessment (FLS) under the assumption of small displace-
ments. For the check of maximum bending moments
(ULS) due to direct wave loading, the pinned-pinned
boundary condition may be applied in combination with
the apparent span length (not the effective span length).

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 34

7 Pipe-soil interaction Table 7-2 Typical geotechnical parameters for clay

Soil type su γsoil ’ ν es
7.1 General [kN/m2 ] [kN/m3 ]
Very soft <12.5 4-7 0.45 1.0-3.0
7.1.1 The soil is to be classified as cohesive (clays) or Soft 12.5-25 5-8 0.45 0.8-2.5
cohesionless (sands). As basis for the evaluations of the Firm 25-50 6-11 0.45 0.5-2.0
pipe-soil interaction the following basic soil parameters
Stiff 50-100 7-12 0.45 0.4-1.7
are relevant:
• type of soil; Very stiff 100-200 10-13 0.45 0.3-0.9
• in-situ stress conditions; Hard >200 10-13 0.45 0.3-0.9
• shear strength parameters for drained or undrained
condition including remoulded shear strength for
clays; 7.1.4 Uncertainties in the soil data should be consid-
• soil moduli and damping coefficients as function of ered, e.g. by sensitivity analysis. These uncertainties may
cyclic shear strain; arise from variations in the soil conditions along the pipe-
• soil settlement parameters; line route and difficulties in determining reliable in-situ
• general soil data as submerged unit weight, void ratio, soil characteristics of the upper soil layer, which is the soil
water content and plasticity limits. of most importance for the pipeline. Soil data down to a
depth equal to about 0.5-1.0 times the pipe diameter are
7.1.2 If the approximate soil stiffness expressions in 7.3 most important to consider in this context.
are to be used, then the following specific parameters are
relevant: 7.2 Modelling of pipe -soil interaction
• submerged unit weight of soil, γsoil ’
• Poisson’s ratio, ν 7.2.1 The pipe-soil interaction is important in the
• void ratio, es evaluation of the static equilibrium configuration and the
• angle of friction, cohesionless soils, ϕs dynamic response of a free spanning pipeline. The follow-
• undrained shear strength, cohesive soils, su ing functional requirements apply for the modelling of soil
• over-consolidation ratio, OCR resistance:
• plasticity index, cohesive soils, ip • the seabed topography along the pipeline route must
be represented;
The plasticity index ip is a standard geotechnical parame- • the modelling of soil resistance must account for non-
ter whose value is usually specified in the soil reports for linear contact forces normal to the pipeline and lift
the pipeline route. The plasticity index influences the dy- off;
namic spring stiffness given in Section 7.3 only for clay • the modelling of soil resistance must account for slid-
and only when the over-consolidation ratio of the clay is ing in the axial direction. For force models this also
greater than one, OCR > 1. Note that OCR > 1 is rarely applies in the lateral direction;
assumed. • appropriate (different) short- and long-term character-
istics for stiffness and damping shall be applied, i.e.
static and dynamic stiffness and damping.
7.1.3 The parameters listed above should preferably be
obtained by means of geotechnical tests on undisturbed
soil samples and be representative for the particular geo- 7.2.2 The seabed topography may be defined by a ver-
graphical location of the pipeline. In lieu of detailed in - tical profile along the pipeline route. The spacing of the
formation, the values given in Table 7-1 and Table 7-2 data points characterising the profile should relate to the
may be used. actual roughness of the seabed.

7.2.3 The axial and lateral frictional coefficients be-

Table 7-1 Typical geotechnical parameters for sand
tween the pipe and the seabed shall reflect the actual sea-
Soil type ϕs γsoil ’ ν es bed condition, the roughness, the pipe, and the passive soil
3
[kN/m ] resistance.
Loose 28-30o 8.5-11.0 0.35 0.7-0.9
Medium o 7.2.4 The axial and lateral resistance is not always of a
30-36 9.0-12.5 0.35 0.5-0.8
pure frictional type. Rapid changes in vertical stresses are
Dense 36-41o 10.0-13.5 0.35 0.4-0.6 (in low-permeable soil) reacted by pore water and not by a
change in effective contact stresses between the soil and
the pipe. In addition, the lateral resistance will have a con-
tribution due to the penetration of the pipe into the soil,
which needs be accounted for.
 Recommended Practice DNV-RP-F105, March 2002
Page 35

7.2.5 For sands with low content of fines, the frictional Note that the sand type is identified by the value of the
component of the axial and lateral resistance is propor- friction angle ϕs (Table 7-1), and the clay type is identified
tional with the vertical force at any time. For clays, the by the value of the undrained shear strength s u (Table 7-2).
frictional component is proportional with the undrained
shear strength. For pipes supported by rock, values for the modal soil
damping ratios may be taken as for dense sand.
7.2.6 Where linear soil stiffness has to be defined for
the eigenvalue analysis, the soil stiffness should be s e- Table 7-4 Modal soil damping ratios (in %) for clay.
lected considering the actual soil resistance and the ampli-
Horizontal (in- Vertical (cross-
tude of the oscillations.
Clay type line) direction flow) direction
7.2.7 The soil stiffness for vertical loading should be L/D L/D
evaluated differently for static and dynamic analyses. The <40 100 >160 <40 100 >160
static soil response will be governed mainly by the maxi- Very soft - Soft 3.0 2.0 1.0 3.0 2.0 1.0
mum reaction, including some cyclic effects. Dynamic Firm – Stiff 2.0 1.4 0.8 1.2 1.0 0.8
stiffness will be characterised mainly by the unloading/re-
loading situation. Very stiff - Hard 1.4 1.0 0.6 0.7 0.6 0.5

7.2.8 The soil damping is generally dependent on the

dynamic loads acting on the soil. Two different types of 7.3 Approximate Soil Stiffness
soil damping mechanisms can be distinguished:
• Material damping associated with hysteresis taking 7.3.1 The following expressions may be used for the
place close to the yield zone in contact with the pipe. static vertical soil reaction per unit length as a function of
• Radiation damping associated with propagation of the penetration, v:
elastic waves through the yield zone.
Rv = γsoil ’b(Nq v+0.5Nγb) − sandy soils
7.2.9 The radiation damping may be evaluated from Rv = b(γsoil ’Nq v+Ncsu ) − clayey soils
available solutions for elastic soils using relevant soil Where
modulus reflecting the soil stress (or strain) levels. The
radiation damping depends highly on the frequency of the B Load distribution width:
oscillations, and is more important for high frequency os- 2 ( D − v) v for v ≤ 0.5D
cillations. =
 D for v > 0.5D
7.2.10 The modal soil damping ratio, ζsoil , due to thesoil- D Outer pipe diameter (including any coat-
pipe interaction may be determined by: ing)
γsoil ’ Submerged unit weight of soil.
 ∫ c( s) φ2 (s )ds 
1  L  su Undrained shear strength
ζ soil =  
4πf 0  ∫ m (s) φ ( s)ds 
2
  The expressions for R V are based on bearing capacity for-
L  mulas for ideal 2-D strip foundations. Note that if these
where the soil damping per unit length, c(s), may be de- formulas are used to predict the expected penetration v for
fined on the basis of an energy balance between the maxi- a given contact pressure RV, they may lead to underestima-
mum elastic energy stored by the soil during an oscillation tion of the true penetration due to effects of the pipelay
cycle and the energy dissipated by a viscous damper in the process and erosion as well as possible 3-D effects on the
same cycle. shoulder near the free span.

Alternatively, the modal soil damping ratio, ζsoil , may be 7.3.2 The bearing capacity factors Nc, Nq and Nγ versus
taken from Table 7-3 or Table 7-4, in which L denotes the
the internal friction angle ϕs may be calculated from the
length of the free span and D is the outer diameter of the following formulas:
pipeline. Interpolation is allowed.
ϕ
N q = exp( π tan ϕ s ) tan 2 ( 45 + s )
Table 7-3 Modal soil damping ratios (in %) for sand. 2
N c = ( N q − 1) cot ϕs
Horizontal (in-line) Vertical (cross-flow)
Sand direction direction N γ = 1.5( N q − 1) tan ϕs
type L/D L/D
<40 100 >160 <40 100 >160
For clayey soils the friction angle is set equal to 0°, i.e. Nq
Loose 3.0 2.0 1.0 2.0 1.4 0.8 = 1.0 and Nc = 5.14.
Medium 1.5 1.5 1.5 1.2 1.0 0.8
Dense 1.5 1.5 1.5 1.2 1.0 0.8

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 36

the associated incremental vertical displacement of the

100
pipe.

7.3.6 The lateral (horizontal) dynamic stiffness KL is

defined as KL =∆FL /∆δL , where ∆FL is the incremental
Nc horizontal force between pipe and soil per unit length of
pipe, and ∆δL is the associated incremental horizontal dis-
Nc, Nq and Nγ

placement of the pipe.

10
Nq
7.3.7 For a detailed determination of KV , the following
expression may be applied:
0.88 ⋅ G
Nγ Kv =
1− ν
which is based on elastic halfspace theory for a rectangular
1 foundation under assumption of a pipe length that equals
0 10 20 30 40 50 10 times the contact width between pipe and soil. Pois-
ϕ[degree] son’s ratio ν is given in Table 7-1 and Table 7-2.

Figure 7-1 Bearing capacity factors Nc, Nq and Nγ ver- 7.3.8 For a detailed determination of KL , the following
sus the internal friction angle ϕs expression may be applied:

K L = 0.76 ⋅ G ⋅ (1 + ν )
7.3.3 The maximum static, axial soil reaction per unit
length may be taken as: which is based on elastic halfspace theory for a rectangular
foundation under assumption of a pipe length that equals
Ra = Rv µa − sandy soils 10 times the contact width between pipe and soil. Pois-
Ra = min{Rv µa, bτmax} − clayey soils son’s ratio ν is given in Table 7-1 and Table 7-2.
Where
7.3.9 For conditions with small-amplitude deforma -
Rv Vertical static soil reaction given by 7.3.1 tions, the shear modulus for the soil can be taken as:
µa Axial friction coefficient
B Given in 7.3.1  2000 ⋅ (3 − es )2
σs
OCR  i   1. 3i p  
 1 + es
for sand
[kN/m2 ]
= 1 − p  +   ip in % G= 
2.61  200   200 

kc
1300 ⋅ (3 − e s ) σ (OCR )k s
2
for clay

s
Soil shear strength: 1 + es
2
 0. 5(1 − k c ) Rv  where
τmax = s u2 − 
 b  σs Effective mean stress (in units of kPa), see
 
7.3.10.
OCR Over-consolidation ratio
es Void ratio
7.3.4 The static vertical stiffness is a secant stiffness
ks Coefficient, taken from Figure 7-2
representative for penetration conditions such as during
installation and erosion and during development of free
spans.
0.5
The static vertical stiffness KV,S is defined as KV,S=RV/v,
where RV is the static vertical soil reaction per unit length 0.4
of pipe and v is the vertical penetration of the pipe re-
quired to mobilise this reaction. Unless effects of pipelay 0.3
ks

and erosion and 3-D shoulder effects are significant, the 2-

D approach outlined in 7.3.1 can be used to predict v. Oth- 0.2

erwise, or when no detailed information is available, the

0.1
static stiffness value may be taken according to Table 7-7
for sand and Table 7-8 for clay.
0
0 20 40 60 80 100 120 140
7.3.5 The vertical dynamic stiffness KV is defined as Plasticity index, i p

KV=∆FV/∆δV, where ∆FV is the incremental vertical force Figure 7-2 k s versus plasticity index, i p
between pipe and soil per unit length of pipe, and ∆δV is
 Recommended Practice DNV-RP-F105, March 2002
Page 37

This expression for G, which gives lower-bound values for
the initial shear modulus, has been calibrated to measure-
ments from free pipeline spans. It is based on an expres-
sion for the initial shear modulus, formulated by Hardin
and Drnevich (1970) and based on experimental results for
a broad range of soil types.
7.3.11 The procedure in Sections 7.3.7-7.3.10 leads to
values of the dynamic stiffness KV and KL , which can be
considered as lower-bound values for initial small-strain
stiffness at either end of the free span, but which are not
adjusted for possible non-linear soil behaviour at larger
strains.

For extreme conditions, which can be assumed to cause

If there are indications that the values for KV and KL
large deformations, considerably lower G values apply.
should be different from those produced by this procedure,
then the ratio between the assumed pipe length on the
7.3.10 The effective mean stress, σs , in the soil at the shoulder and the contact width may be adjusted from the
span supports may be calculated from the stress conditions adopted value of 10. Calculation of the effective mean
at a representative depth below the pipe. The representa- stress σs from the stress conditions at a different represen-
tive depth may be assumed equal to the contact width b, tative depth below the pipe than b may also be considered.
which is given in Section 7.3.1. The following formula Note that, in this context, it is acceptable to distinguish
may then be applied: between representative depths for calculation of KV and
for calculation of KL .
1 q L
σS = (1 + K0 ) bγ soil' + (1 + )
2 3b 2Lsh 7.3.12 When normal conditions prevail and when no
detailed analysis is carried out for determination of KV and
in which KL for small-strain conditions, the values of these stiff-
K0 Coefficient of earth pressure at rest, usu- nesses in units of kN/m/m may be calculated in simplified
ally K0 =0.5 manner as (D is in units of metre):
γsoil ’ Submerged unit weight soil (kN/m3 )
2 ρs 1
q Submerged weight of pipe per unit length K V = CV ( + ) D
of pipe (kN/m) 3 ρ 3
Lsh Span support length on one shoulder (for
2 ρs 1
transfer of one-half the weight of the free K L = CL ( + ) D
span) 3 ρ 3
L Span length
in which the coefficients CV and CL are taken according to
Table 7-7 and Table 7-8. The soil type, which is used as
Note that for pipes on clay, the clay will not be consoli- entry to these tables, is identified by the value of the fric-
dated for the weight of the pipe in the temporary phase
tion angle ϕs for sand (Table 7-1) and by the value of the
immediately after pipelay. For calculations for a pipe on
undrained shear strength su for clay (Table 7-2).
clay in this phase, the formula for σs reduces to
1
σS = (1 + K 0) bγ soil' Table 7-7 Dynamic stiffness factor and static stiffness
2
for pipe-soil interaction in sand.
When detailed information does not exist and the topog- Sand type CV CL KV,S
raphical conditions are not complex, the support length (kN/m5/2 ) (kN/m5/2 ) (kN/m/m)
ratio Lsh /L may be taken according to Table 7-5 for sand
Loose 16000 12000 250
and according to Table 7-6 for clay.
Medium 22000 16500 530
Dense 32000 24000 1350
Table 7-5 Support length ratio Lsh/L for sand
Sand type Lsh /L
Table 7-8 Dynamic stiffness factor and static stiffness
Loose 0.3 for pipe-soil interaction in clay with OCR=1.
Medium 0.2
Clay type CV CL KV,S
Dense 0.1
(kN/m5/2 ) (kN/m5/2 ) (kN/m/m)
Very soft 1200 800 50-100
Table 7-6 Support length ratio Lsh /L for clay Soft 2700 1800 160-260
Clay type Lsh /L Firm 6000 4000 500-800
Very soft 0.5 Stiff 9000 6000 1000-1600
Soft 0.4 Very stiff 21000 14000 2000-3000
Firm 0.3 Hard 24000 16000 2600-4200
Stiff 0.2
Very stiff 0.1
Hard 0.07 7.3.13 The simplified expressions are based on the pro-
cedure in Sections 7.3.7-7.3.10 for average properties of
the considered soil type in conjunction with an assumed

DET NORSKE VERITAS

Recommended Practice DNV-RP-F105, March 2002
Page 38

penetration of 20-40% of the diameter and Lsh /L as given Table 7-9 Equivalent stiffness, k 2 ,
in Table 7-7 and Table 7-8. for span supports on sand
v/D k2 (kN/m/m)
The simplified expressions for KV and KL , with values of 0.00 0
CV and CL given in Table 7-7 and Table 7-8, are valid for
0.25 19
dynamic conditions with small-amplitude deformations.
0.35 28
0.50 44
7.3.14 For free spans supported by sand, the lateral dy-
1.00 105
namic stiffness KL should be calculated under an assump-
tion of loose sand properties in order to properly account
for effects of erosion. For span supports on clay, the equivalent stiffness k2 can
be calculated as
−0 .4
 s 
1 .3
7.3.15 For extreme conditions, which can be assumed to v
cause large deformations on the shoulders, a smaller spring k 2 = 8.26 ⋅ s u ⋅  u  ⋅ 
stiffness than that associated with small-strain conditions  Dγ soil  D
apply. For vertical loading, this smaller spring stiffness is in which
typically of the same order of magnitude as the static ver- γsoil Total unit weight of soil (=γsoil ’+γwater)
tical stiffness KV,S. For lateral loading, reference is made γsoil ’ Submerged unit weight of water
to Section 7.3.16. Unit weight of water (=10 kN/m3 )
γwater
v Vertical pipe penetration at span shoulder
7.3.16 In case of lateral loads large enough to cause slid-
ing at the span supports, the following bilinear force-
For span supports on sand, the maximum lateral resistance
displacement curve, which covers both small-strain condi-
per unit length of pipe is
tions and extreme displacement conditions, may be applied 1. 25
at span supports on sand and clay:  v
FL,max = µ LFV + 5. 0 ⋅ γ soil'⋅D2 ⋅  
D
 k1 ⋅ y for FL < µ LFV
 µ F For span supports on clay, the maximum lateral resistance
FL = 
µ F + k2 ⋅ ( y − L V ) for µLFV ≤ FL < FL, max per unit length of pipe is
 L V k1 − 0. 4
 s 
1 .3
 v
FL,max = µ LFV + 4.13 ⋅ D ⋅ s u ⋅  u  ⋅ 
in which  Dγ soil  D
y Lateral displacement of pipe on shoulder
FL Lateral force per unit length of pipe at dis- 7.3.17 The axial dynamic soil stiffness is usually not
placement y important. However, when long free spans are considered,
FL,max Maximum lateral resistance per unit length it is important to include an axial soil-support model with
of pipe friction and stiffness. If no information is available about
FV Vertical contact force per unit length of the axial dynamic soil stiffness, it may be taken as equal to
pipe on shoulder the lateral dynamic soil stiffness KL as described above.
µL Lateral friction coefficient
k1 Initial stiffness up to mobilisation of full 7.4 Artificial supports
friction
k2 Equivalent stiffness for deformations past
mobilisation of full friction 7.4.1 Gravel sleepers can be modelled by modifying
the seabed profile, considering the rock dump support
Unless data indicate otherwise, µL =0.6 may be applied for shape and applying appropriate stiffness and damping
characteristics.
span supports on sand and µL =0.2 for span supports on
clay.
7.4.2 The purpose of mechanical supports is generally
For supports on sand, the initial stiffness k1 may be taken to impose locally a pipeline configuration in the vertical
and/or transverse directions. Such supports can be mo d-
as equal to the lateral dynamic stiffness KL for loose sand.
For supports on clay, the initial stiffness k1 may be taken elled by concentrated springs having a defined stiffness,
as equal to the lateral dynamic stiffness KL for the clay taking into account the soil deformation beneath the sup-
port and disregarding the damping effect.
type in question.

The equivalent stiffness k2 depends on the vertical penetra-

tion, v, of the pipe on the shoulder. For span supports on
sand, values for k2 are given in Table 7-9.

8 References
Blevins, R.D., “Flow-Induced Vibrations”, Krieger
Publishing Company, Florida, 1994
Bruschi, R. & Vitali, L., “Large-Amplitude Oscillations of
Geometrically Non-linear Elastic Beams Subjected to
Hydrodynamic Excitation”, JOMAE, Vol. 113, May,
1991.
DNV Offshore Standard OS-F101, “Submarine Pipeline
Systems”.
DNV Offshore Standard OS-F201, “Dynamic Risers”.
DNV Recommended Practice RP-C203, “Fatigue Strength
Analysis of Offshore Steel Structures”.
DNV Classification Note CN 30.5, ”Environmental
Conditions and Environmental Loads”.
DNV Recommended Practice RP-E305, “On-bottom
Stability Design of Submarine Pipelines”.
Fyrileiv, O. & Mørk, K.J., “Assessment of Free Spanning
Pipelines using the DNV Guideline”, ISOPE’98, Montreal,
Canada, May 24-29, 1998.
Fyrileiv, O., Mørk, K.J. & Rongved, K. “TOGI Pipeline –
Assessment of Non-stationary Free Spans”, OMAE’00,
New Orleans, USA, February 14-17, 2000.
Hansen, E.A., Bryndum, M., Mørk, K.J., Verley, R.,
Sortland, L. & Nes, H.,”Vibrations of Free Spanning
Pipeline Located in the Vicinity of a Trench", OMAE'01-
4016, Rio de Janeiro, Brazil, June 3-8, 2001.
DET NORSKE VERITAS
Recommended Practice DNV-RP-F105, March 2002
Page 39
Hardin, B.O., and V.P. Drnevich, “Shear Modulus and
Damping in Soils: I. Measurements and Parameters
Effects, II. Design Equations and Curves,” Technical
Reports UKY 27-70-CE 2 and 3, College of Engineering,
University of Kentucky, Lexington, Kentucky, July 1970.
Hobbs, R.E:, “Influence of Structural Boundary
Conditions on pipeline Free Span Dynamics”, OMAE’86,
1986.
Mørk, K.J., Vitali, L. & Verley, R., “The MULTISPAN
Project: Design Guideline for Free Spanning Pipelines”,
Proc. of OMAE’97 conf., Yokohama, Japan, April 13-17,
1997.
Mørk, K.J. & Fyrileiv, O. “Fatigue Design According to
the DNV Gu ideline for Free Spanning Pipelines”, OPT’98,
Oslo, Norway, February 23-24, 1998.
Mørk K.J., Fyrileiv, O., Verley, R., Bryndum, M. &
Bruschi, R. “Introduction to the DNV Guideline for Free
Spanning Pipelines”, OMAE’98, Lisboa, July 6-9, 1998.
Mørk, K.J., Fyrileiv, O., Nes, H. & Sortland, L. “A
Strategy for Assessment of Non-Stationary Free Spans”,
ISOPE-99, Brest, France, May 30-June 4, 1999.
Sumer B.M. & Fredsøe, J. “Hydrodynamics around
Cylindrical Structures”, Advanced Series on Ocean
Engineering – Volume 12, World Scientific, London,
1997.
Tura, F., Dumitrescu, A., Bryndum, M. B. & Smeed, P.F.
“Guidelines for Free Spanning Pipelines: The GUDESP
Project”, OMAE’94, Volume V, pp 247-256, Houston,
1994.