Faculty of Engineering

Department of Naval Architecture and Marine Engineering

On-Bottom Stability Design of Subsea

Pipelines in Shallow Water

A thesis submitted in partial fulfillment of the requirements for

the degree of Master of Science

In
Naval Architecture and Marine Engineering

Submitted by
Mohammed Amr Mohammed Abdelsalam

B.Sc. in Naval Architecture and Marine Engineering

Faculty of Engineering, Alexandria University, 2013

2022
 DEDICATION

I dedicate this thesis to

my late father

my mother

my wife, Sarah

my son, Kenan

iv
 ACKNOWLEDGMENTS

First of all, I would like to thank my Lord “Allah” (Glory be to Him), whose will, guidance
and inspiration were the driving factors to get the work of this thesis accomplished. I am
sincerely thankful and grateful to him, and I ask him for more help in this life and the
forthcoming.

I would like to express my gratitude to my thesis advisor Prof. Khaled Ali Hafez for
continuous support, motivation, and his help in revising this thesis and the research paper. My
sincere thanks to my co-advisor Dr. Ahmed Naguib Abdelhameed for his friendly and
encouraging attitude.

I would like to express my gratitude to Eng. Abdulhameed Yehya (Pipeline Technical

Authority at GUPCO) for his time, guidance and the information provided at the start of this
work.

I would like to thank John Wood Group PLC company for providing an educational license of
Flexcom software and their software support team for the continuous help and guidance. Also,
I would like to thank Sigma Technical Services company, the local agent of Technical
Toolboxes Inc. in Egypt, for providing a license of AGA/PRCI software.

Finally, I would like to express my very profound gratitude to my wife, who suffered a lot
during my research studies, for understanding all the time and providing me with unfailing
support and continuous encouragement. This accomplishment would not have been possible
without you. Thank you.

v
 DECLARATION

I hereby declare that no part of this dissertation has been submitted to Alexandria University
or to any other university in Egypt or abroad as part of the requirements to obtain another
degree.

Furthermore, Alexandria University code of ethics and rules of academic integrity have been
followed.

Name: Mohammed Amr Mohammed Abdelsalam

Signature:…………………..

vi
 ABSTRACT

Pipelines are the main means of transport of oil and gas products in the industry. Products are
being transported from well or reservoir to subsea manifolds, shore, or production facility
platforms for end-users. Arrival to optimum design for subsea pipeline system, many
complicated engineering studies are required. This starts with pipeline sizing, material
selection, pipeline routing, on-bottom stability design, etc.

On-bottom stability analysis is one of the most important and fundamental tasks during the
design process of subsea pipelines. On-bottom stability is necessary to ensure vertical and
lateral stability for the as-laid pipeline on the seabed against hydrodynamic loads from waves
and currents. The external concrete coating layer is mostly used to add weight to the pipe as a
cost-effective solution for pipeline stabilization.

In this thesis, the factors that affect the pipeline stability are discussed, the conventional and
advanced methods for the on-bottom stability design are expounded, and a case study is
conducted using three design approaches: static analysis approach, calibrated methods
approach, and dynamic analysis approach.

For on-bottom stability design using static analysis and calibrated methods approach, on-
bottom stability analysis software is developed using MATLAB programming language in
compliance with the recommended practice DNVGL-RP-F109. AGA/PRCI Level 2 program
is also used as another calibrated method. AGA/PRCI Level 2 was developed by the Pipeline
Research Council International (PRCI) that was a part of the American Gas Association
(AGA). Comparison between both calibrated methods from DNVGL and AGA/PRCI is
carried out as well as a sensitivity analysis for various water depths.

The dynamic on-bottom stability analysis is performed using a finite element-based advanced
offshore engineering simulation software called Flexcom to predict the pipeline response in a
time-domain simulation based on a given environmental condition. The resultant maximum
lateral displacements and the associated stresses are examined via different thicknesses of the
concrete weight coating. Pipeline response from the numerical simulation is investigated and
reasons behind the dynamic response are defined, and seven simulation runs using random
seed numbers are performed to confirm the on-bottom stability against different random sea
states. It is observed from the results the importance of several factors on pipeline stability
such as pipeline submerged weight using the simple friction model, hydrodynamic loads
induced by random sea states on the stresses and lateral displacements, and soil friction model
being used incorporating the passive soil resistance term.

Comparison between the three stability approaches is conducted to study the effect of
ignoring passive soil resistance and the importance of dynamic stability analysis to optimize
the on-bottom stability design. The comparison has emphasized the reduction in concrete
weight coating using dynamic stability analysis if the actual case is modeled accurately.

vii
 TABLE OF CONTENTS

DEDICATION ......................................................................................................................... iv
ACKNOWLEDGMENTS ........................................................................................................ v
DECLARATION ..................................................................................................................... vi
ABSTRACT ............................................................................................................................ vii
TABLE OF CONTENTS ......................................................................................................viii
LIST OF FIGURES ................................................................................................................ xii
LIST OF TABLES ................................................................................................................. xiv
NOMENCLATURE ............................................................................................................... xv
LIST OF ABBREVIATIONS ............................................................................................... xix
CHAPTER 1: INTRODUCTION ........................................................................................... 1
1.1. General............................................................................................................................. 1
1.2. On-bottom Stability Design of Pipeline .......................................................................... 1
1.3. Thesis Objectives ............................................................................................................. 2
1.4. Thesis Outlines ................................................................................................................ 3
CHAPTER 2: LITERATURE REVIEW ............................................................................... 5
2.1. Introduction ..................................................................................................................... 5
2.2. Pipeline Classification ..................................................................................................... 5
2.3. Design Process ................................................................................................................. 6
2.4. Pipeline Stabilization Methods ........................................................................................ 8
2.5. Limit State Design ........................................................................................................... 9
2.5.1. Serviceability Limit State (SLS) .............................................................................. 9
2.5.2. Ultimate Limit State (ULS) .................................................................................... 10
2.5.3. Fatigue Limit State (FLS) ....................................................................................... 10
2.5.4. Accidental Limit State (ALS) ................................................................................. 10
2.6. On-bottom Stability Design Aspects ............................................................................. 10
2.6.1. Hydrodynamic Forces ............................................................................................ 11
2.6.1.1. Drag Force ....................................................................................................... 12
2.6.1.2. Inertia Force ..................................................................................................... 12
2.6.1.3. Lift Force ......................................................................................................... 12
2.6.1.4. The Complete Morison’s Equation.................................................................. 13
2.6.2. Pipe-Soil Interaction ............................................................................................... 13
2.6.2.1. Seabed Soil Classification ............................................................................... 13
2.6.2.2. Pipeline Initial Penetration .............................................................................. 14
2.6.2.3. Soil Resistance................................................................................................. 15
2.6.2.3.1. Coulomb Friction Model .......................................................................... 15
2.6.2.3.2. Passive Soil Resistance ............................................................................. 16
2.7. DNV On-bottom Stability Design History .................................................................... 18
2.7.1. Force Balance Method ............................................................................................ 18

viii
 2.7.2. DNV-RP-E305 Design Methods ............................................................................ 18
2.7.2.1. Overview of the Design Method ..................................................................... 18
2.7.2.2. Dynamic Analysis ........................................................................................... 19
2.7.2.3. Generalized Stability Analysis ........................................................................ 19
2.7.2.4. Simplified Static Stability Analysis................................................................. 19
2.7.3. DNVGL-RP-F109 Design Methods ....................................................................... 20
2.7.3.1. Dynamic Lateral Stability Analysis (DLSA) .................................................. 20
2.7.3.2. Generalized Lateral Stability Method (GLS) .................................................. 22
2.7.3.3. Absolute Lateral Static Stability Method (ALSS) ........................................... 22
2.7.4. Limitations of DNV Design Methods .................................................................... 23
2.8. AGA/PRCI Methodology .............................................................................................. 23
2.8.1. Level 1 Analysis ..................................................................................................... 24
2.8.2. Level 2 Analysis ..................................................................................................... 24
2.8.3. Level 3 Analysis ..................................................................................................... 24
2.8.4. Factor-of-Safety and Risk Levels ........................................................................... 25
2.8.4.1. Level 1 and Level 2 ......................................................................................... 25
2.8.4.2. Level 3 ............................................................................................................. 26
2.8.5. Stability Design Procedure Assumptions ............................................................... 26
2.8.6. Selection of Appropriate Design Procedure ........................................................... 26
2.8.7. Limitation of AGA/PRCI Levels............................................................................ 28
2.9. Methodologies Comparison ........................................................................................... 28
2.9.1. Similarities .............................................................................................................. 28
2.9.2. Differences ............................................................................................................. 28
CHAPTER 3: ON-BOTTOM STABILITY DESIGN METHODS ................................... 29
3.1. Introduction ................................................................................................................... 29
3.2. DNVGL-RP-F109 Recommended Practice................................................................... 29
3.2.1. Load Combinations ................................................................................................ 29
3.2.2. Current Loadings .................................................................................................... 30
3.2.3. Wave Loadings ....................................................................................................... 32
3.2.4. Wave Directionality and Spreading........................................................................ 34
3.2.5. Vertical Stability ..................................................................................................... 35
3.2.5.1. Vertical stability in soil.................................................................................... 35
3.2.5.2. Vertical stability in water ................................................................................ 36
3.2.6. Lateral Stability Methods ....................................................................................... 36
3.2.6.1. Absolute Lateral Static Stability Method (ALSS) ........................................... 36
3.2.6.2. Generalized Lateral Stability Method (GLS) .................................................. 38
3.2.6.3. Dynamic Lateral Stability Analysis (DLSA) .................................................. 42
3.3. AGA/PRCI On-bottom Stability Simulator................................................................... 42
3.3.1. Level 2 Methodology ............................................................................................. 43
3.3.2. Input Form .............................................................................................................. 47
3.3.3. Plot Output.............................................................................................................. 47

ix
 3.4. Flexcom Software .......................................................................................................... 49
3.4.1. Finite Element Modeling ........................................................................................ 49
3.4.2. Pipe-Soil Interaction ............................................................................................... 50
3.4.3. Environmental Loads.............................................................................................. 51
CHAPTER 4: ON-BOTTOM STABILITY DEVELOPED SOFTWARE PACKAGE .. 52
4.1. Introduction ................................................................................................................... 52
4.2. Software Development .................................................................................................. 52
4.3. Functions of the Developed Tool .................................................................................. 52
4.4. Graphical User Interface (GUI) ..................................................................................... 53
4.5. Brief Operation Instructions .......................................................................................... 54
4.5.1. Warning and Error Messages ................................................................................. 54
4.5.2. Output Results ........................................................................................................ 54
CHAPTER 5: CASE STUDY ANALYSIS AND RESULTS .............................................. 57
5.1. Introduction ................................................................................................................... 57
5.2. Case Study Design Data ................................................................................................ 57
5.3. Static Analysis ............................................................................................................... 59
5.4. Calibrated Methods........................................................................................................ 60
5.4.1. Generalized Lateral Stability Method .................................................................... 60
5.4.1.1. 0.5D Displacement Criterion ........................................................................... 60
5.4.1.2. 10D Displacement Criterion ............................................................................ 61
5.4.2. AGA/PRCI Level 2 ................................................................................................ 62
5.4.3. A Brief Comparison Study ..................................................................................... 63
5.4.3.1. ALSS method versus GLS methods ................................................................ 63
5.4.3.2. AGA/PRCI Level 2 method versus GLS (0.5D displacement) method .......... 63
5.4.4. Sensitivity Analysis ................................................................................................ 65
5.5. Dynamic Analysis.......................................................................................................... 67
5.5.1. Case Study Data ..................................................................................................... 67
5.5.1.1. Pipeline Data ................................................................................................... 67
5.5.1.2. Environmental Data ......................................................................................... 67
5.5.1.3. Seabed and Soil Data ....................................................................................... 68
5.5.1.4. Boundary Conditions ....................................................................................... 68
5.5.1.5. Assumptions .................................................................................................... 68
5.5.2. Results .................................................................................................................... 68
5.5.2.1. The Combined Load of 1-year RPV Wave plus 10-year RPV Current........... 69
5.5.2.2. The Combined Load of 10-year RPV Wave plus 1-year RPV Current........... 69
5.5.2.3. Numerical Investigation of Pipeline Response ................................................ 74
5.5.2.4. Random Seed Numbers ................................................................................... 76
5.5.2.5. Comparison with conventional stability methods ........................................... 76
CHAPTER 6: CONCLUSIONS AND RECOMMENDATIONS ...................................... 79
6.1. Introduction ................................................................................................................... 79
6.2. Conclusions ................................................................................................................... 79

x
 6.2.1. General ................................................................................................................... 79
6.2.2. Static Analysis ........................................................................................................ 79
6.2.3. Calibrated Methods ................................................................................................ 80
6.2.4. Dynamic Analysis .................................................................................................. 80
6.3. Recommendations ......................................................................................................... 81
REFERENCES ....................................................................................................................... 82
APPENDIX A RESULTS OF ON-BOTTOM STABILITY ANALYSIS ......................... 87
A.1 Static Analysis Results .................................................................................................. 87
A.2 Calibrated Methods Results ........................................................................................... 87
A.2.1 Generalized Lateral Stability Method (0.5D Displacement) .................................. 87
A.2.2 Generalized Lateral Stability Method (10D Displacement) ................................... 87
A.2.3 AGA/PRCI Level 2 Software ................................................................................. 88
A.2.4 Sensitivity Analysis Results between AGA/PRCI Level 2 and GLS 0.5D ............ 88
APPENDIX B FLEXCOM INPUT FILES .......................................................................... 89
B.1 Initial Static Analysis ..................................................................................................... 89
B.2 Quasi-Static Analysis ..................................................................................................... 90
B.3 Static Analysis for Current Loadings ............................................................................. 91
B.4 Dynamic Analysis for Wave Loadings .......................................................................... 92

xi
 LIST OF FIGURES

Figure 1.1: Thesis scope diagram ............................................................................................... 2

Figure 1.2: Computer software packages used in this thesis for on-bottom stability design ..... 3
Figure 2.1: Classification of Subsea pipelines (Y. Bai, 2001) ................................................... 5
Figure 2.2 Typical flexible pipeline cross-section details (Youssef, 2011) ............................... 6
Figure 2.3: Typical rigid pipeline cross-section details (Youssef, 2011) ................................... 6
Figure 2.4: The pipeline design process (Y. Bai, 2001) ............................................................. 7
Figure 2.5: Pipeline trenching using a trenching machine (Allseas Company, 2021) ............... 8
Figure 2.6: Concrete mattresses installation (Allseas Company, 2021) ..................................... 8
Figure 2.7: Rock dumping installation (Seatools Company, 2021) ........................................... 9
Figure 2.8: Gravity anchor installation (Zeitoun et al., 2009) .................................................... 9
Figure 2.9: Forces acting on the pipeline resting on the seabed ............................................... 11
Figure 2.10: Pipeline penetration into the soil (Bruton et al., 2008) ........................................ 14
Figure 2.11: Verley Force-Displacement curve (Verley & Sotberg, 1994) ............................. 16
Figure 2.12: RP-E305 design method (DNV, 1988) ................................................................ 19
Figure 2.13: RP-F109 design method ....................................................................................... 21
Figure 2.14: Limitation of current stability design approach (McMaster et al., 2012) ............ 23
Figure 3.1: The influence of current velocity on a pipeline laid on the seabed (Q. Bai, 2013) 31
Figure 3.2: The Pierson-Moskowitz and JONSWAP spectra (Abankwa et al., 2015) ............. 32
Figure 3.3: Reduction factor due to wave spreading (𝑅𝐷) versus relative angle between wave
and pipeline (𝜃𝑟𝑒𝑙) (DNVGL, 2017a) ...................................................................................... 35
Figure 3.4: Calculation flowchart for wave and current velocities .......................................... 35
Figure 3.5: Bottom velocity amplitude content during 4-hour build-up (largest 200 velocity
amplitudes used to predict embedment) (PRCI, 2008)............................................................. 43
Figure 3.6: Bottom velocity amplitude content during the 3-hour design storm (largest 50
velocity amplitude used to predict embedment) (PRCI, 2008) ................................................ 44
Figure 3.7: Level 2 build-up sea state model employed to predict pipe embedment (PRCI,
2008) ......................................................................................................................................... 45
Figure 3.8: Level 2 software package logic (PRCI, 2008) ....................................................... 46
Figure 3.9: Level 2 pipe embedment logic (PRCI, 2008) ........................................................ 47
Figure 3.10: Level 2 input form (PRCI, 2008) ......................................................................... 48
Figure 3.11: Plot safety factor example (PRCI, 2008) ............................................................. 48
Figure 3.12: Coulomb friction model used in Flexcom (Wood Group, 2018) ......................... 50
Figure 4.1: Main window of the on-bottom stability software tool.......................................... 52
Figure 4.2: Main window for wave and current characteristics input ...................................... 53
Figure 4.3: Warning message window ..................................................................................... 54
Figure 4.4: Error message window ........................................................................................... 54
Figure 4.5: Results window of ALSS method .......................................................................... 55
Figure 4.6: Results window of the GLS method (a) 0.5D displacement criterion and (b) 10D
displacement criterion............................................................................................................... 55

xii
Figure 4.7: Graphical representation of pipeline weight requirement and CWC thickness
versus combined directions of waves and currents .................................................................. 56
Figure 5.1: Gamba Project Location (Zeetech Engineering B.V., 2013) ................................. 57
Figure 5.2: Required concrete coating thickness using ALSS method .................................... 60
Figure 5.3: Required concrete coating thickness using GLS (0.5D displacement) method ..... 61
Figure 5.4: Required concrete coating thickness using GLS (10D displacement) method ...... 62
Figure 5.5: Required concrete coating thickness using AGA/PRCI Level 2 method .............. 63
Figure 5.6: Required concrete coating thickness using ALSS and GLS methods during
installation phase ...................................................................................................................... 64
Figure 5.7: Required concrete coating thickness using ALSS and GLS methods during
operation phase ......................................................................................................................... 64
Figure 5.8: Required concrete coating thickness using GLS (0.5D displacement) and
AGA/PRCI Level 2 methods .................................................................................................... 65
Figure 5.9: Sensitivity analysis results under different loading conditions using GLS (0.5D
displacement) and AGA/PRCI Level 2 methods ...................................................................... 66
Figure 5.10: Envelopes of lateral displacements for concrete coating thickness range 135-145
mm (1-year RPV wave + 10-year RPV current) ...................................................................... 70
Figure 5.11: Envelopes of von Mises stresses for concrete coating thickness range 135-145
mm (1-year RPV wave + 10-year RPV current) ...................................................................... 70
Figure 5.12: Envelopes of lateral displacements for concrete coating thickness range 150-160
mm (1-year RPV wave + 10-year RPV current) ...................................................................... 71
Figure 5.13: Envelopes of von Mises stresses for concrete coating thickness range 150-160
mm (1-year RPV wave + 10-year RPV current) ...................................................................... 71
Figure 5.14: Envelopes of lateral displacements for concrete coating thickness range 155-165
mm (10-year RPV wave + 1-year RPV current) ...................................................................... 72
Figure 5.15: Envelopes of von Mises stresses for concrete coating thickness range 155-165
mm (10-year RPV wave + 1-year RPV current) ...................................................................... 72
Figure 5.16: Envelopes of lateral displacements for concrete coating thickness range 170-180
mm (10-year RPV wave + 1-year RPV current) ...................................................................... 73
Figure 5.17: Envelopes of von Mises stresses for concrete coating thickness range 170-180
mm (10-year RPV wave + 1-year RPV current) ...................................................................... 73
Figure 5.18: Time history of lateral displacement for the end node for 180mm concrete
coating thickness ...................................................................................................................... 74
Figure 5.19: Time history of vertical displacement for the end node for 180mm concrete
coating thickness ...................................................................................................................... 75
Figure 5.20: Time history of total water particle velocity at the pipeline end node for 180mm
concrete coating thickness ........................................................................................................ 75
Figure 5.21: Envelopes of pipeline lateral displacements for the seven simulations ............... 77
Figure 5.22: Envelopes of von Mises stresses for the seven simulations ................................. 77

xiii
 LIST OF TABLES

Table 2.1: Typical link between scenarios and limit states (DNVGL, 2017c) ......................... 11
Table 2.2: Sand soil properties (Q. Bai & Bai, 2014) .............................................................. 14
Table 2.3: Estimates of undrained shear strength of clays (Q. Bai & Bai, 2014) .................... 14
Table 2.4: Parameter limitation of generalized and simplified methods .................................. 20
Table 3.1: Seabed roughness (DNVGL, 2017a) ....................................................................... 30
Table 3.2: Peak horizontal load coefficient (DNVGL, 2017a)................................................. 38
Table 3.3: Peak vertical load coefficient (DNVGL, 2017a) ..................................................... 38
Table 3.4: Minimum weight for pipe laid on the sand, K ≥ 10 (DNVGL, 2017a) ................... 40
Table 3.5: Minimum weight for pipe laid on the sand, K ≤ 5 (DNVGL, 2017a) ..................... 41
Table 3.6: Minimum weight for pipe laid on sand (DNVGL, 2017a) ...................................... 41
Table 5.1: Pipeline data (Zeetech Engineering B.V., 2013) ..................................................... 58
Table 5.2: Seawater properties (Zeetech Engineering B.V., 2013) .......................................... 58
Table 5.3: Soil data (Zeetech Engineering B.V., 2013) ........................................................... 58
Table 5.4: Waves and currents design data (Zeetech Engineering B.V., 2013) ....................... 58
Table 5.5: Significant wave heights and wave hitting angles relative to the water depth
(Zeetech Engineering B.V., 2013) ............................................................................................ 59
Table 5.6: Design load conditions ............................................................................................ 59
Table 5.7 Wave and current design data for the sensitivity analysis ........................................ 65
Table 5.8: Wave and current design data for the dynamic analysis (Zeetech Engineering B.V.,
2013) ......................................................................................................................................... 67
Table 5.9: Results of comparison between DLSA, ALSS and GLS methods .......................... 78
Table A.1: Full results of ALSS method .................................................................................. 87
Table A.2: Full results of GLS method (0.5D displacement) ................................................... 87
Table A.3: Full results of GLS method (10D displacement) .................................................... 87
Table A.4: Full results of AGA/PRCI Level 2 ......................................................................... 88
Table A.5: Full results of GLS 0.5D for the sensitivity analysis ............................................. 88
Table A.6: Full results of AGA/PRCI Level 2 for the sensitivity analysis .............................. 88

xiv
 NOMENCLATURE

Latin Symbols

𝑏 Pipe buoyancy per unit length

𝐶𝑎 Added mass coefficient
𝐶𝐷 Drag force coefficient
𝐶𝐿 Lift force coefficient
𝐶𝑀 Inertia force coefficient
𝐶𝑌∗ Peak horizontal load coefficient
𝐶𝑍∗ Peak vertical load coefficient
𝑑 Water depth
𝑑50 Mean grain size
𝐷 Pipe outer diameter including all coating
𝐷𝑠 Steel pipe outer diameter
𝐸 Young’s modulus
𝐹𝐶 Vertical contact force between the soil and the pipe
𝐹𝐷 Drag force
𝐹𝐹 Seabed friction force
𝐹𝐻 History dependent soil resistance
𝐹𝐿 Lift force
𝐹𝑀 Inertia force
𝐹𝑅 Passive soil resistance
𝐹𝑤 Load factor
𝐹𝑌 Horizontal hydrodynamic (drag and inertia) load
𝐹𝑌∗ Peak horizontal load
𝐹𝑍 Vertical hydrodynamic (lift) load
𝐹𝑍∗ Peak vertical load
𝑔 Acceleration of gravity. Taken as 9.81 m/s2
𝐺 Transfer function
𝐺𝑐 Soil (clay) strength parameter
𝐺𝑠 Soil (sand) density parameter
𝐻𝑠 Significant wave height during a sea state
𝐻∗ Maximum wave height during a sea state
𝐾 Significant Keulegan-Carpenter number

xv
𝑘 Wave number
𝐾∗ Keulegan-Carpenter number for single design oscillation
𝐾𝑏 Equivalent sand roughness parameter
𝑘𝑇 Ratio between the period of single design oscillation and design spectrum
𝑘𝑈 Ratio between oscillatory velocity amplitude of single design oscillation and design
spectrum
𝑘𝑉 Ratio between steady velocity component applied with single design oscillation and
with design spectrum.
𝐿 Significant weight parameter
𝐿∗ Weight parameter related to single design oscillation
𝐿10 Required weight parameter for limiting allowable pipeline displacement up to 10-
diameter
𝐿𝑚 Mobilization length for the stiffness of the Coulomb friction model
𝐿𝑠𝑡𝑎𝑏𝑙𝑒 Required weight parameter for obtaining virtual stable pipeline
𝐿𝑌 Required weight for an intermediate displacement criterion
𝑀 Steady to oscillatory velocity ratio for design spectrum
𝑀∗ Steady to oscillatory velocity ratio for single design oscillation
𝑀𝑛 Spectral moment of order 𝑛
𝑁 Spectral acceleration factor
𝑅𝐷 Reduction factor due to spectral directionality and spreading
𝑟𝑝𝑒𝑛 Load reduction factor due to penetration
𝑟𝑝𝑒𝑟𝑚 Load reduction factor due to a permeable seabed
𝑟𝑡𝑜𝑡 Load reduction factor
𝑟𝑡𝑟 Load reduction factor due to trench.
𝑠 Spectral spreading exponent
𝑠𝑔 Pipe specific density
𝑠𝑢 Undrained clay shear strength
𝑆𝑈𝑈 Wave induced velocity spectrum
𝑆𝜂𝜂 Wave spectral density
𝑡 Computational time
𝑡𝑐𝑜𝑎𝑡 Corrosion coating thickness
𝑇𝑛 Reference period
𝑇𝑝 Peak period for design spectrum
𝑡𝑠𝑡𝑒𝑒𝑙 Steel pipe wall thickness
𝑇𝑢 Zero up-crossing period

xvi
𝑇∗ Period associated with single design oscillation
𝑈 Total water particle velocity from wave and current
𝑈(𝑧) Reduced velocity due to boundary layer effect at the pipeline level
𝑈(𝑧𝑟 ) Current velocity at the reference measurements height above the seabed
𝑈𝑐 Current velocity
𝑈𝑝 Pipeline velocity
𝑈𝑟 Relative velocity between the total water particle velocity and the pipeline
𝑈𝑠 Spectrally derived oscillatory velocity (significant amplitude) for design spectrum,
perpendicular to pipeline
𝑈𝑠𝜃 Spectrally derived oscillatory velocity (significant amplitude) for design spectrum, at
an angle 𝜃 to the pipeline
𝑈𝑤 Wave induced water particle velocity
𝑈∗ Oscillatory velocity amplitude for single design oscillation, perpendicular to pipeline
𝑉 Steady current velocity associated with design spectrum, perpendicular to pipeline
𝑉∗ Steady current velocity associated with design oscillation, perpendicular to pipeline
𝑊𝑠 Pipe submerged weight per unit length
𝑦 Lateral pipe displacement
𝑌 Non-dimensional lateral pipe displacement
𝑧 Pipeline elevation above the seabed
𝑧0 Bottom roughness parameter
𝑍𝑝 Penetration depth
𝑍𝑝𝑖 Initial pipe penetration
𝑧𝑟 Current reference measurement height above the seabed
𝑍𝑡𝑟 Trench depth

Greek Symbols

𝛼 Generalised Phillips’ constant

Γ Gamma function
𝛾𝑆𝐶 Safety factor
𝛾𝑤 Safety factor
𝛾 Peak-enhancement factor
𝛾𝑠 Dry unit soil weight
𝛾′𝑠 Submerged unit soil weight
𝜃𝑐 Angle between current direction and pipe
𝜃𝑤 Angle between wave heading and pipe

xvii
𝜇 Coefficient of friction
𝜌𝑐𝑜𝑎𝑡 Corrosion coating density
𝜌𝑐𝑜𝑛𝑐 Concrete coating density
𝜌𝑐𝑜𝑛𝑡 Minimum product density
𝜌𝑠𝑡𝑒𝑒𝑙 Pipe steel density
𝜌𝑤 Mass density of water
𝜎 Spectral width parameter
𝜏 Number of oscillations in the design bottom velocity spectrum
𝜙 angle of internal friction
𝜔 Wave frequency
𝜔𝑝 Peak wave frequency
𝑣 Poisson ratio

xviii
 LIST OF ABBREVIATIONS

2D Two-dimensional
AGA American Gas Association
ALS Accidental Limit State
ALSS Absolute Lateral Static Stability
API American Petroleum Institute
CWC Concrete Weight Coating
DLSA Dynamic Lateral Stability Analysis
DNV Det Norske Veritas
DNVGL Det Norske Veritas – Germanischer Lloyd
FE Finite Element
FEA Finite Element Analysis
FEM Finite Element Method
FLS Fatigue Limit State
GL Germanischer Lloyd
GLS Generalized Lateral Stability
GUI Graphical User Interface
JONSWAP Joint North Sea Wave Project
KM Kilometer
LRFD Load and Resistance Factor Design
OBS On-bottom Stability Simulator
PRCI Pipeline Research Council International
RP Recommended Practice
RPV Return Period Values
SLS Serviceability Limit State
SMYS Specified Minimum Yield Stress
ULS Ultimate Limit State
WT Wall Thickness

xix
 CHAPTER 1: INTRODUCTION

1.1. General
Pipelines are a major hub in the oil and gas industry and are one of the main means of
transporting many fluids such as oil, gas, chemicals, and water between two points during oil
and gas production. Their use has increased over time, as they are more economical than
competing means of transport. Although the subsea pipelines require huge initial investments,
they last up to 40 years and require relatively small maintenance. The subsea pipelines can be
sometimes country crossing and continental borders. Billions of dollars in investment are
spent on such large projects and can also be controversial (Kyriakides & Corona, 2007).

In subsea pipelines field, many research and development attempts have been put in place to
address the technical problems of this sector. They cover, for example, improving the quality
of the pipeline's steel; developing several methods for installing the pipelines and lay-vessels;
establishing the mechanical behavior and failure modes of pipes under various loads;
investigating new fluid-pipe and soil-pipe interactions; corrosion problems, welding, and
many others. (Kyriakides & Corona, 2007).

In 1947, the first oil-producing well was drilled. In 1952 the first pipeline barge was
commissioned and in 1954 the first pipeline was laid on the seabed. Because of the increased
use of subsea pipelines, pipeline manufacturing, design, and installation methods have
developed significantly. More recently, around 5,000 km of pipelines are installed annually
(Braestrup et al., 2009).

1.2. On-bottom Stability Design of Pipeline

During the design of subsea pipelines which run for hundreds of kilometers, there are several
challenges. These include pipeline stability in tough environments with strong currents and
large storms. The design of the subsea pipelines should ensure that the pipeline can be
constructed, monitored, and maintained to resist this tough marine environment. Since
pipelines are the primary hub for transporting the hydrocarbons, they must be intended to be
stable in order to prevent environmental or economical disasters. The precise design of the
pipeline must predict the loading conditions on the pipe, when and where the pipe may
displace, and how much resistance the pipe has. This design reduces the risk of pipeline
breakage which could cost the environment millions of dollars and lead to the leaking of oil
and gas.

After their installation on the seabed, pipelines are exposed to heavy hydrodynamic loads
from waves and currents. To ensure pipeline stability, the pipeline must be stable during the
worst loading conditions. The most followed methodologies in the industry for the on-bottom
stability design are those methods stated in the recommended practice (RP) of RP-F109
(DNVGL, 2017a) and PRCI On-Bottom Stability Simulator (OBS) (PRCI, 2008), the latter
was developed by the Pipeline Research Council International, Inc. (PRCI) that was formerly
a part of the American Gas Association (AGA).

DNVGL-RP-F109 has three design methods for on-bottom stability analysis:

• Absolute Lateral Static Stability Method (ALSS)

1
 • Generalized Lateral Stability Method (GLS)
• Dynamic Lateral Stability Analysis (DLSA)

AGA/PRCI OBS (PRCI, 2008) has three levels for on-bottom stability analysis:

• Level 1 concerning the basic stability check by comparing the hydrodynamic loads
and the soil resistance.
• Level 2 dealing with the stability check of the pipeline including the embedment of the
pipeline from storm builds up.
• Level 3 performs dynamic analysis by considering pipeline displacements on the
seabed.

The on-bottom stability design methods can be classified on the basis of hydrodynamic loads
and soil resistance into three approaches (Zeitoun et al., 2008):

• Static analysis approach that is based on the force-balance equation where no pipeline
movement is allowed under extreme sea states such as ALSS method and AGA/PRCI
Level 1 software.
• Calibrated or empirical design methods approach such as the AGA/PRCI Level 2
software (Hale et al., 1989; PRCI, 2008) and GLS method (Lambrakos et al., 1987).
• Dynamic analysis approach that is based on a time-domain simulation of dynamic
pipeline response, hydrodynamic loads from sea state condition, and pipe-soil
interaction using specialized finite element (FE) software packages.

1.3. Thesis Objectives

The objective of this thesis, as illustrated in its scope in Figure 1.1, is to study the on-bottom
stability design using different approaches under different loading combinations from waves
and currents during installation and operation conditions. The case study that is presented in
this thesis is for a pipeline project titled “Gamba Export Loading Line” which is located in
Gabon country. This pipeline has external diameter of 22 inches placed on sand soil and in
shallow water with total length of about 9.4 km offshore. This case study is selected to
perform on-bottom stability analysis using three design approaches that are: static analysis
approach, calibrated methods approach, and dynamic analysis approach.

Figure 1.1: Thesis scope diagram

2
This thesis also studies several factors affecting pipeline stability such as the pipe-soil
interaction model, sea-state, and water depth by performing several comparisons between
different methods and approaches.

Three computer software packages are used to perform the on-bottom stability design as
shown in Figure 1.2. Following are brief description of the capabilities of the three computer
software packages:

• The developed on-bottom stability software which is built and designed using
MATLAB programming language (The MathWorks, 2018) based on RP-F109
(DNVGL, 2017a). It is used to perform the on-bottom stability design using the ALSS
method (as static analysis approach) and GLS method (as calibrated method
approach).
• AGA/PRCI Level 2 Version 3.00 was developed by the PRCI of AGA. PRCI Level 2
program is another calibrated method.
• Flexcom software package was developed by John Wood Group PLC (Wood Group,
2018). It is finite-element-based advanced and highly versatile simulation software
package, which is commonly used in the structural analysis of conventional and
unconventional offshore structures. It is used to perform the dynamic on-bottom
stability analysis to calculate the lateral displacements of the pipeline and its induced
stresses under the action of hydrodynamic loads from two-dimensional irregular
waves and steady currents and restrained by soil resistance force.

Figure 1.2: Computer software packages used in this thesis for on-bottom stability
design

1.4. Thesis Outlines

Chapter 2 (Literature Review) contains an introduction about the subsea pipelines and their
classification, design process, stabilization methods, limit state design for on-bottom stability.
It also gives an introduction on the on-bottom stability design aspects such as the
hydrodynamic forces and pipe-soil interaction including the classification of soils and soil
resistance components, the history of DNV design codes for on-bottom stability design, and
AGA/PRCI levels for pipeline stability assessment.

Chapter 3 (On-bottom Stability Design Methods) describes the procedures and methods of
on-bottom stability analysis and the calculation of current and wave loadings based on the
RP-F109 (DNVGL, 2017a). Also, it introduces the methodologies developed by AGA/PRCI
(PRCI, 2008) with special attention and details on Level 2 program methodology.

Chapter 4 (Developed On-bottom Stability Software Package) describes the interface and
commands of that developed software package using MATLAB programming language (The
MathWorks, 2018), in addition to its capability to determine the stability of the pipeline in
compliance with RP-F109 (DNVGL, 2017a).

3
Chapter 5 (Case Study Analysis and Results) focuses on the analysis of on-bottom stability
using the three design approaches; static analysis approach using ALSS method, calibrated
methods approach using GLS and AGA/PRCI Level 2 methods, and dynamic analysis
approach based on DLSA guidelines presented in RP-F109 (DNVGL, 2017a). Also, it
discusses the results from each design method and introduces a comparison between different
approaches and methods.

Chapter 6 (Conclusions and Recommendations) summarizes the conclusions drawn from the
analysis based on each design approach and presents some directions for future work.

Appendix A (Results of On-bottom Stability Analysis) presents the full results of static and
calibrated methods during installation and operation phase conditions under different loading
combinations from waves and currents.

Appendix B (Flexcom Input Files) contains the script of each analysis file which used in
performing the dynamic stability analysis by means of Flexcom software package.

4
 CHAPTER 2: LITERATURE REVIEW

2.1. Introduction
On-bottom stability analysis should be carried out to ensure that the pipeline will be stable
during installation and operation phase conditions by determining the required stability
weight. The stability weight of the pipeline should be designed to ensure adequate seabed
embedment and soil resistance to resist the hydrodynamic loads without excessive movements
or stresses. The stability analysis is performed by considering forces acting at a pipeline cross-
section to ensure both vertical and lateral stability. For lateral stability, the meteorological and
oceanographic data for storm waves and currents are used to establish design values for water
particle velocity and acceleration at the seabed (PRCI, 2008).

This chapter gives a literature review on the subsea pipelines and their classification, on-
bottom stability design and methods of stabilization, and the aspects to be considered in the
stability analysis such as hydrodynamic forces and pipe-soil interaction. It also covers the
history of the methodologies used for on-bottom stability analysis and their scope of
applications.

2.2. Pipeline Classification

Subsea pipelines can be classified as suggested by (Y. Bai, 2001) and shown in Figure 2.1:

• Export (transportation) pipelines;

• Flowlines to transfer product from a platform to export lines;
• Water injection or chemical injection flowlines;
• Flowlines to transfer product between platforms, subsea manifolds, and satellite
subsea wells;
• Pipeline bundles.

Figure 2.1: Classification of Subsea pipelines (Y. Bai, 2001)

5
The subsea pipelines are widely divided into two main categories by cross-section: flexible
and rigid pipelines. Flexible pipe is consisting of wrapping of many intertwining layers of the
high-strength stainless steel carcass and special polymers (Guo et al., 2005), see Figure 2.2.
The rigid pipe consists mainly of a steel pipe with external concrete coating and
external/internal layers of insulation, see Figure 2.3. This type of pipe can be constructed with
large diameters and operated under high pressure. These features make the rigid pipes suitable
for export lines, flowlines, and infield flowlines. Due to the importance of rigid pipe type, it is
considered in this thesis for on-bottom stability analysis.

The total pipeline weight is the summation of all pipeline layers weight, which basically
includes the steel pipe wall and the concrete coating layer. However, additional weight
components, such as internal content, external/internal corrosion coating, insulation coating,
and marine growth, should be considered in the pipe weight calculation if they exist.

Figure 2.2 Typical flexible pipeline cross-section details (Youssef, 2011)

Figure 2.3: Typical rigid pipeline cross-section details (Youssef, 2011)

Based on their installation method, the subsea pipelines can also be classified into two types,
trenched and untrenched pipelines. The untrenched pipelines are more popular than trenched
ones due to its lower operational cost. In the early 1980s, extensive research showed that a
properly designed steel pipe with a concrete coating layer is strong enough to provide the
pipeline with on-bottom stability (Moshagen & Kjeldsen, 1980).

2.3. Design Process

To reach the optimum pipeline design, many design elements have to be well designed, as
shown in Figure 2.4, these elements include:

6
• Wall thickness selection;
• Material grade selection;
• Route selection;
• Flowline protection;
• Flowline stress analysis;
• Flowline installation analysis.

Figure 2.4: The pipeline design process (Y. Bai, 2001)

7
2.4. Pipeline Stabilization Methods
In most subsea pipeline projects, an important design activity is to ensure that the pipeline is
stable on the seabed under the action of hydrodynamic loads caused by waves and currents.
Pipeline designers achieve pipeline stability by using primary stabilization methods, i.e., by
increasing the pipeline weight either by adding an external concrete weight coating (CWC) or
increasing the steel pipe wall thickness (WT). The concrete coating layer can provide
mechanical protection to the corrosion resistance coating layer while increasing pipeline wall
thickness is a comparatively expensive option. Pipeline stability can also be achieved by using
secondary stabilization methods such as burying the pipeline into the seabed by trenching (see
Figure 2.5), installation of concrete mattresses (see Figure 2.6), rock dumping over the
pipeline (see Figure 2.7), or installation of gravity anchors (see Figure 2.8) (Palmer & King,
2006).

Most secondary pipeline stabilization methods are time-consuming and expensive. For rock
dumping, it is necessary to carry comparatively big quantities of quarry rock from domestic
quarries and place them on the seabed along or over the pipelines. The pipeline engineer faces
challenges, including a selection of alignment, the definition of the design basis, stability
analysis, operational needs, and definitions of the construction method. This also involves
maintaining a significant reduction in construction costs and the pipeline system's operational
reliability.

Figure 2.5: Pipeline trenching using a trenching machine (Allseas Company, 2021)

Figure 2.6: Concrete mattresses installation (Allseas Company, 2021)

8
 Figure 2.7: Rock dumping installation (Seatools Company, 2021)

Figure 2.8: Gravity anchor installation (Zeitoun et al., 2009)

2.5. Limit State Design

The design of subsea pipeline under DNVGL standard ST-F101 (DNVGL, 2017c) is based on
a Load Resistance Factor Design (LRFD) format, i.e., load factors are applied to the loads and
a resistance factor is applied to the material strength. The design according to ST-F101
(DNVGL, 2017c) requires satisfying the Limit State Design (LSD) approach for four different
limit states:

• Serviceability Limit State (SLS),

• Ultimate Limit State (ULS),
• Fatigue Limit State (FLS), and
• Accidental Limit State (ALS).

All four limit states should be considered in the assessment of the pipeline response under
hydrodynamic loads. The recommended practice RP-F109 (DNVGL, 2017a) for the on-
bottom stability analysis is also based on the LSD approach. The standard ST-F101 (DNVGL,
2017c) states that, for other stability methods than the ALSS method, the recommended safety
level should be determined based on engineering judgment to obtain a safety level equivalent
to modern industry practice.

2.5.1. Serviceability Limit State (SLS)

For subsea pipelines, it must be ensured that they will be safe for their planned function
during its installation and operation. The SLS relates to the load condition that may cause the

9
pipeline to be unsafe for its operation if it is exceeded. For all the related loading conditions
which can be formulated, the SLS is defined. The following conditions may be considered
(Ghebreghiorghis, 2014):

• Deformation and movements due to waves and currents (hydrodynamic stability);

• Longitudinal deformations due to temperature and pressure variations (pipeline
expansion);
• Lateral deformations due to unrestrained temperature and pressure expansion (lateral
and upheaval buckling, i.e., deformation in horizontal and vertical direction
respectively);
• Blockage of the pipeline, due to hydrate formation or wax deposition (flow assurance
study);
• Pipeline interference with other subsea structures.

In the case of permanent local damage or permanent unacceptable deformations, the SLS is
not the appropriate formulation and the ULS design factors shall be introduced.

2.5.2. Ultimate Limit State (ULS)

Since pipelines on the seabed can move laterally, very large bending moments can occur
(particularly at fixations) therefore local buckling limit state must be checked through
buckling design analysis (Zeitoun et al., 2008).

2.5.3. Fatigue Limit State (FLS)

The FLS involves fatigue damage caused and accumulated over its entire lifespan by cycling
loads. The structure is designed to meet or exceed the design life, taking into account fatigue
damage from all load sources. Fatigue is mainly produced from waves and currents such as
stresses induced by the cyclic movement that can be happened between two stabilization
gravity anchors (Zeitoun et al., 2008).

2.5.4. Accidental Limit State (ALS)

The ALS involves damage or failure due to accidental, unusual, or unplanned loading
conditions such as higher return period metocean conditions that should be taken into account
and confirmed not to risk the pipeline safety (Zeitoun et al., 2008).

Pipeline limit state checks are typically split into different scenarios. A scenario is not
identical to a limit state and may include different limit states. A typical link between
scenarios and limit states is given in Table 2.1.

2.6. On-bottom Stability Design Aspects

Although the principle of pipeline stability is a simple concept to be captured, the design of
pipeline stability involves the consideration of a multitude of complex design aspects, such as
hydrodynamic forces, pipe-soil-fluid interaction, pipe-soil interaction, and the selection of an
approach that ensures a robust final design result. Many of these design aspects are not yet
fully understood by the pipeline community. Oversimplification of these design aspects can
lead to conservative results like very heavy pipelines which are costly to construct or non-

10
conservative results, that are often linked to expensive secondary stabilization requirements
(Zeitoun et al., 2008).

Table 2.1: Typical link between scenarios and limit states (DNVGL, 2017c)

2.6.1. Hydrodynamic Forces

The hydrodynamic forces are generated from wave and current actions that acting on the
pipeline installed on the seabed, see Figure 2.9. These forces are generated from the combined
effect of flow separation from the pipe surface and flow velocity over the pipe. The following
sections describe the different force components and the analytical expressions that are
commonly used in calculation of these forces on the pipeline.

The Morison equation (Morison et al., 1953) is a well-established method used to estimate the
forces around a cylinder. When applied to a pipe on the seabed, the lift (𝐹𝐿 ), drag (𝐹𝐷 ) and
inertia (𝐹𝑀 ) force under combined wave and current can be calculated as follows:

Figure 2.9: Forces acting on the pipeline resting on the seabed

11
2.6.1.1. Drag Force

Drag force is associated with velocities from currents superposed by waves. The drag is
mainly caused by the pressure difference between that in front of the pipeline and that in the
wake region behind the pipeline. The drag force per unit length of the pipeline (𝐹𝐷 ) can be
determined based on Eq. (2.1).

1
𝐹𝐷 = 𝜌𝑤 𝐷𝐶𝐷 𝑈𝑟 |𝑈𝑟 | Eq. (2.1)
2

where,
𝐶𝐷 drag force coefficient
𝐷 pipeline external diameter including all coatings
𝑈 total water particle velocity from wave and current, 𝑈 = (𝑈𝑤 + 𝑈𝑐 )
𝑈𝑝 pipeline velocity
𝑈𝑟 the relative velocity between the fluid and the pipeline, 𝑈𝑟 = (𝑈 − 𝑈𝑝 )
𝜌𝑤 seawater density

2.6.1.2. Inertia Force

Waves produce cyclic loadings in the water column above the water particles. These loadings
accelerate and deaccelerate water particles in the vertical and horizontal directions. The
pipeline which is exposed to an accelerated flow is subject to a force that is proportional to the
acceleration called inertia force. The inertia force per unit length (𝐹𝑀 ) can be determined
based on Eq. (2.2).

1 𝜕𝑈 𝜕𝑈𝑝
𝐹𝑀 = 𝜌𝑤 𝜋𝐷 2 (𝐶𝑀 − 𝐶𝑎 ) Eq. (2.2)
4 𝜕𝑡 𝜕𝑡

where,
𝐶𝑎 added mass coefficient, 𝐶𝑎 = (𝐶𝑀 − 1)
𝐶𝑀 inertia force coefficient
𝑡 time
𝜕𝑈
water particle acceleration
𝜕𝑡

2.6.1.3. Lift Force

The flow of seawater over the pipeline produces lift force as the flow past an airplane. The
existence of the seabed leads to an asymmetry between the flow above the pipeline and the
flow below. The lifting force is caused by high pressures from the slower (no flow velocities)
under the pipeline and low pressures from high flow velocities over the pipeline. The pipeline
lifting force per unit length (𝐹𝐿 ) can be determined based on Eq. (2.3).

1 2
𝐹𝐿 = 𝜌𝑤 𝐷𝐶𝐿 (𝑈 − 𝑈𝑝 ) Eq. (2.3)
2

where 𝐶𝐿 is the lift force coefficient.

12
2.6.1.4. The Complete Morison’s Equation

The inline force per unit length of a pipe (𝐹𝑌 ) is determined using the complete Morison’s
equation (Morison et al., 1953) given by Eq. (2.4).

1 1 𝜕𝑈 𝜕𝑈𝑝
𝐹𝑌 = 𝜌𝑤 𝐷𝐶𝐷 𝑈𝑟 |𝑈𝑟 | + 𝜌𝑤 𝜋𝐷 2 (𝐶𝑀 − 𝐶𝑎 ) Eq. (2.4)
2 4 𝜕𝑡 𝜕𝑡

2.6.2. Pipe-Soil Interaction

A model of contact interaction between the pipeline and the seabed is usually referred to as a
model of pipe-soil interaction. The pipe-soil interaction consists of seabed stiffness for initial
penetration and friction resistance to pipeline movement. The friction is based mainly on the
Coulomb friction model, soil type (cohesive/non-cohesive or combination of both together),
the density of the soil, and the contact force between the soil and the pipeline. The contact
force, friction resistance, and soil stiffness should, therefore, be precisely predicted (Q. Bai &
Bai, 2014).

2.6.2.1. Seabed Soil Classification

To accurately model the pipe-soil interaction along the pipeline route, soil data is required.
Soil should be classified along the pipeline route to determine the soil properties. In pipe-soil
interaction, it is mainly the strength and friction properties that are of concern.

Soil classification is based on visual inspection and testing in the laboratory. The investigation
aims to classify the soil to determine the strength parameters. The soil can be classified
according to grain size, cohesion, and behavior as below (Q. Bai & Bai, 2014).

The grain size-based soil classification:

• Coarse-grained soil (larger than 0.05 mm)

• Fine-grained soil (finer than 0.05 mm)

The cohesion type-based soil classification:

• Cohesive soil (fine soil particles that are sticking together)

• Non-cohesive soil (soil particles are not sticking together)

The behavior-based soil classification depends on the rate of loading with respect to the
permeability of the soil:

• Drained soil (the loading rate is slower than that at which pore water can move in or
out of soil interparticle voids)
• Undrained soil (the loading rate is greater than that at which pore water can move in or
out of soil interparticle voids)

For example, sand can be classified as coarse-grained, non-cohesive, and drained soil, unlike
clay that is classified as fine-grained, cohesive, and undrained soil.

13
Table 2.2. shows the sand soil properties, in which the sand strength is given by the “internal
friction angle” denoted by symbol 𝜙, and the clay strength is given in terms of the “undrained
shear strength”, denoted by the symbol 𝑠𝑢 and measured in kilopascals, see Table 2.3.
However, if the clay is sheared at a very slow rate, such that enough time is allowed for the
pore water to move in or out of the soil interparticle voids, then it will not exhibit undrained
shear strength. Instead, it behaves more like sand with an applicable clay friction angle.
Similarly, if sand is sheared under load at a very fast rate, such that the pore water lacks
enough time to move in or out of the soil interparticle voids, then sand can exhibit undrained
behavior.

Table 2.2: Sand soil properties (Q. Bai & Bai, 2014)
Type of Sand Angle of Internal Friction, 𝜙 [°] Submerged Unit Weight, 𝛾′𝑠 [kN/m3]
Very loose <28 <10
Loose 28−30 8.8−10
Medium dense 30−36 9.4−10.5
Dense 36−41 10.5−12.5
Very dense >41 >10.5

Table 2.3: Estimates of undrained shear strength of clays (Q. Bai & Bai, 2014)
Description of Clay Undrained Shear Strength, 𝑠𝑢 (kN/m2)
Very soft 0−20
Soft 20−40
Firm 40−75
Stiff 75−150
Very stiff −
Hard −
Very hard 

2.6.2.2. Pipeline Initial Penetration

The measurement of the initial pipeline penetration into the seabed is an essential step in
subsequent assessments during the pipeline design process, it is affecting interactions between
the pipeline and soil horizontally and vertically. Figure 2.10 shows the definition of pipeline
initial penetration. It is possible to calculate the initial penetration of a statically loaded pipe
into the soil as a function of pipe diameter, vertical contact force, soil strength parameters for
clay (based on undrained shear strength and dry unit soil weight for clay), and submerged unit
weight for sands. Pressure/penetration is a non-linear relationship and can be used in a finite
element model (Y. Bai & Bai, 2005).

Figure 2.10: Pipeline penetration into the soil (Bruton et al., 2008)

14
Initial penetration for sand and clay soils can be calculated based on Eq. (2.5) and Eq. (2.6).
These equations were developed by (Verley & Sotberg, 1994) and (Verley & Lund, 1995) for
sand and clay soil respectively based on the curve fitting to data and other test data developed
by (Brennodden et al., 1989) and (Wagner et al., 1987) that accounts for vertical cyclic loads.

The dimensionless initial penetration for the pipeline on sand can be determined by Eq. (2.5):

𝑍𝑝𝑖
= 0.037 ⋅ 𝑘𝑠 −0.67
𝐷
Eq. (2.5)
𝛾′𝑠 ⋅ 𝐷 2
𝑘𝑠 =
𝐹𝐶

For pipeline on clay soil, the dimensionless initial penetration can be determined by Eq. (2.6):

3.2 0.7
𝑍𝑝𝑖 𝐺𝑐 0.3 𝐺𝑐 0.3
= 0.0071 ⋅ ( ) + 0.062 ⋅ ( )
𝐷 𝑘𝑐 𝑘𝑐
Eq. (2.6)
𝑠𝑢 ⋅ 𝐷
𝑘𝑐 =
𝐹𝐶

where,
𝐷 pipe outer diameter (including coating)
𝐹𝐶 the vertical contact force between the soil and pipe; 𝐹𝐶 = 𝑊𝑠 – 𝐹𝑍
𝐹𝑍 vertical hydrodynamic (lift) load
𝑠𝑢
𝐺𝑐 soil (clay) strength parameter; 𝐺𝑐 = 𝐷⋅𝛾
𝑠
𝑠𝑢 undrained clay shear strength
𝑍𝑝𝑖 initial pipe penetration
𝑊𝑠 submerged pipe weight per unit length
𝛾𝑠 dry unit soil weight
𝛾′𝑠 submerged unit soil weight

2.6.2.3. Soil Resistance

Soil resistance arises from two distinct sources, the friction between pipeline and seabed soil
(pure friction term), and resistance due to pipe embedment into the seabed soil (passive soil
resistance term). Neglecting the passive soil resistance results in decreasing its total soil
resistance (comparing to its actual resistance); especially when the pipeline is placed on soil
of soft-clay or loose sand, the interaction between both pipe and soil will not be modeled
accurately (Zeitoun et al., 2008), and the pipe may also largely displace in the lateral
direction. In this regard, a reliable on-bottom stability analysis software package must be able
to precisely model this interaction.

2.6.2.3.1. Coulomb Friction Model

The Coulomb friction model is the simplest method for modeling pipe-soil interaction, and it
can be used in static or dynamic analyses (Zeitoun et al., 2008). Coulomb friction presumes
pure steady plastic friction between the pipeline and the seabed and does not take into
consideration the embedment caused by cyclic loads or passive soil resistance.

15
To ensure the pipeline stability on the seabed in the presence of Coulomb friction model, the
relation based on Eq. (2.7) must be established.

𝐹𝐹 ≤ 𝜇 ⋅ 𝐹𝐶 Eq. (2.7)

where,
𝐹𝐹 seabed friction force
0.6 for sand and rock soils
𝜇 coefficient of friction, where 𝜇 = {
0.2 for clay soils

2.6.2.3.2. Passive Soil Resistance

Pipeline embedment is a significant parameter for the evaluation of soil resistance and
hydrodynamic loads. The embedment decreases hydrodynamic loads and also increases
passive soil resistance.

The pipeline embedment starts during installation and increases during the design life of the
pipeline. At the touchdown point, the pipeline experiences much higher forces that cause the
pipeline to embed further. During the hydro testing and throughout the operation phase, the
total weight of the pipe is increased which increases the pipeline load on the seabed with the
result that of increasing that pipeline embedment. In addition to the increased weight, the pipe
experiences cyclic loadings from waves and currents that increase the vertical loads on the
pipeline and hence on the seabed and thus increase the embedment.

Total penetration of the pipeline can be taken as the sum of:

• self-weight initial penetration

• piping penetration (internal erosion of soil particles)
• penetration due to dynamics during pipelaying
• penetration due to pipe movements under the action of hydrodynamic loads.

Passive soil resistance is typically used for pipeline dynamic stability analysis, and it is often
based on the soil resistance model developed by (Verley & Sotberg, 1994) as shown in Figure
2.11.

Figure 2.11: Verley Force-Displacement curve (Verley & Sotberg, 1994)

In the elastic region, no work is done, and penetration is constant and equal to the initial
penetration. In the region 𝑌1 < 𝑌 ≤ 𝑌2 , the pipe soil interaction creates work which again

16
increases the penetration and thus the passive resistance. Note that the value of the break-out
resistance 𝐹𝑅2 cannot be computed a priori, as it is dependent on the accumulated pipe
displacement in the region between 𝑌1 and 𝑌2 .

If the displacement exceeds 𝑌2 , the pipe is assumed to break out. The accumulated work is set
to zero and no work is done in this region. Penetration is reduced linearly from the breakout
value at 𝑌2 to half this value at 𝑌 = 𝑌3 and passive resistance is reduced accordingly.

For a displacement larger than 𝑌3 , penetration and passive resistance can be assumed to be
constant. i.e., no work is done (DNVGL, 2017a).

(Verley & Sotberg, 1994) and (Verley & Lund, 1995) developed an empirical formula to
calculate passive soil resistance for sand and clay soil as given in respectively. The equations
were slightly modified by (DNV, 2007) as given in Eq. (2.8) and Eq. (2.9) for sand soil and
Eq. (2.10) for clay soil. This modification was because of some confusion resulted from the
formulations of the empirical methods.

For Sand

If 𝑘𝑠 ≤ 26.7

𝐹𝑅 2 𝑍𝑝 1.25
= (5 ⋅ 𝑘𝑠 − 0.15 ⋅ 𝑘𝑠 ) ⋅ ( ) Eq. (2.8)
𝐹𝐶 𝐷

If 𝑘𝑠 > 26.7

𝐹𝑅 𝑍𝑝 1.25
= 𝑘𝑠 ⋅ ( ) Eq. (2.9)
𝐹𝐶 𝐷

For Clay

𝐹𝑅 4.1 ⋅ 𝑘𝑐 𝑍𝑝 1.31
= ⋅( ) Eq. (2.10)
𝐹𝐶 𝐺𝑐 0.39 𝐷
where,
𝐷 pipe outer diameter (including external coatings)
𝐹𝐶 vertical contact force between the soil and pipe
𝐹𝑅 passive soil resistance
𝑠𝑢
𝐺𝑐 soil (clay) strength parameter; 𝐺𝑐 = 𝐷⋅𝛾
𝑠
𝑘𝑠 is given by Eq. (2.5)
𝑘𝑐 is given by Eq. (2.6)
𝑠𝑢 undrained clay shear strength
𝑍𝑝 total pipeline penetration into the soil
𝛾𝑠 dry unit soil weight

17
2.7. DNV On-bottom Stability Design History
In the early 80s, due to the absence of advanced software packages and finite element analysis
(FEA), pipeline design was developed based on traditional methods in the early 80s, these
methods were based fundamentally on construction experience and simple equations from
field experiments or laboratory tests. At the end of the 80s, pipeline design started to be based
on advanced software and FEA which contributed to new recommended practices with
different design methods to give more space for reducing the construction cost and optimizing
pipeline design.

Pipeline on-bottom stability is one of the design aspects that affect the design and installation
procedure of subsea pipelines. With the new exploration of oil and gas wells, the need for new
pipelines in tougher seabed conditions is growing and therefore the precise and accurate
assessment of on-bottom pipeline stability plays an important role in the cost-effective design
and ensures pipeline integrity.

2.7.1. Force Balance Method

(DNV, 1981) presents the simple method of designing the static stability of the pipeline. This
method does not allow the pipeline to experience any horizontal movement under the
environmental conditions of a 100-year return period.

The static stability approach is based on a simple equation of the force balance from which
the necessary pipe weight is calculated based on equation Eq. (2.11):

𝛾𝑆𝐶 ⋅ 𝐹𝑌 = 𝜇 ⋅ 𝐹𝐶 Eq. (2.11)

where,
𝐹𝐶 vertical contact force between the soil and pipe
𝐹𝑌 inline force of a pipe given by Eq. (2.4)
𝛾𝑆𝐶 safety factor.
𝜇 coefficient of friction

In Eq. (2.11), Morison’s equation was recommended by (DNV, 1981) for calculating the
hydrodynamic loads (𝐹𝑌 ).

2.7.2. DNV-RP-E305 Design Methods

RP-E305 (DNV, 1988) presents three methods of pipeline analysis, namely dynamic analysis,
generalized stability analysis, and simplified static stability analysis. These methods are
ordered based on the complexity, starting from dynamic analysis to generalized stability
analysis and finally simplified static stability analysis. The generalized and simplified analysis
methods are based on the generalization of dynamic analysis simulation results.

2.7.2.1. Overview of the Design Method

The flow diagram presented in Figure 2.12 shows an overview of the design method of RP-
E309 (DNV, 1988).

18
 Figure 2.12: RP-E305 design method (DNV, 1988)

2.7.2.2. Dynamic Analysis

For special circumstances in which the detailed pipe response is essential, the dynamic
analysis method is recommended. In this analysis, it is necessary to consider a sufficient
pipeline length of (250 to 1000 m) under hydrodynamic loads of 3 hours in a full sea state. If
there is an expectation of a critical strain response to the pipeline, non-linear stress/strain
material behavior-based analysis of the pipe is recommended.

2.7.2.3. Generalized Stability Analysis

This method was developed by (Lambrakos et al., 1987). The generalized stability method is
based on the generalization of the results of a large number of dynamic finite element
simulations carried out following the recommendations for dynamic analysis. A set of non-
dimensional parameters is used to present groups of design curves for different allowable
displacements, ranging from 0 pipe external diameter (for Zone 2 i.e., less than 500 m away
from a platform) to 40 times pipe external diameter (for Zone 1 i.e., more than 500 m from a
platform). These calibrated design curves are simple and may be considered as conservative
design tools if used within their design scope.

2.7.2.4. Simplified Static Stability Analysis

This method is based on the equation of force equilibrium similar to the equation of force
balance Eq. (2.11) with one non-dimensional parameter. This method simplifies the

19
generalized method using design curves with a maximum lateral displacement of 20 meters.
The simplified static stability equation is given by Eq. (2.12):

𝐹𝑌 + 𝜇𝐹𝑍
𝑊𝑠 = ( ) 𝐹𝑤 Eq. (2.12)
𝜇
where,
𝐹𝑤 non-dimensional parameter with ranges from 1.0 to 1.6 depending on the Keulegan-
Carpenter number and the current to wave ratio as described in RP-E305 (DNV,
1988).
𝐹𝑌 inline force per unit length of a pipe given by Eq. (2.4)
𝐹𝑍 vertical hydrodynamic (lift) load
𝑊𝑠 submerged pipe weight
𝜇 coefficient of friction

The applicability of the generalized and simplified stability methods is limited to certain
parameter values as shown in Table 2.4.

Table 2.4: Parameter limitation of generalized and simplified methods

Parameter Value
The external diameter of the pipeline > 0.4 m
Current to wave ratio 0 ~ 0.8
Keulegan-Carpenter number 4 ~ 40
The relative weight of the sandy soil 0.7 ~ 1.0
𝑊
Clay soil with the ratio 𝑠 0.05 ~ 8
𝐷⋅𝑠𝑢

2.7.3. DNVGL-RP-F109 Design Methods

RP-F109 (DNVGL, 2017a) is the first edition after merger between Det Norske Veritas
(DNV) and Germanischer Lloyd (GL) with no changes from its previous edition RP-F109
(DNV, 2007), which replaced RP-E305 (DNV, 1988).

RP-F109 (DNVGL, 2017a) provides three different design methods; dynamic lateral stability
analysis, generalized lateral stability method, and absolute lateral static stability method.
Absolute static and generalized lateral stability methods are widely used in the industry.
Figure 2.13 shows the design method based on RP-F109 (DNVGL, 2017a).

2.7.3.1. Dynamic Lateral Stability Analysis (DLSA)

DLSA works towards calculating the lateral displacement of the pipeline under the action of
combined hydrodynamics loads from waves and currents against soil resistance forces in a
complete sea state using finite element analysis (FEA) software package. It is therefore
expected that the dynamic analysis uses realistic calculated hydrodynamic loads based on a
complete sea state. Modeling of the pipe-soil interaction is recommended to consider at least
the friction resistance and passive soil resistance leading to the penetration of the pipe.

The requirements of dynamic analysis for all versions of recommended practice are the same
and the following items should be modeled accurately:

• Three-dimensional wave model with energy spreading and mean direction

20
 • Current velocity near the seabed
• Pipe structural behavior
• Pipe-soil interaction
• Pipe-fluid interaction
• Boundary conditions (e.g., gravity anchors)

Figure 2.13: RP-F109 design method

The DLSA analysis method is not widely used because of the associated complexities over
the conventional stability methods. Such complexities result from building a model with
accurate simulation involving the fluid-pipe-soil interaction, the limited availability of the
software packages, and the incentive to replace the static and calibrated methods with
advanced dynamic ones is not existing (Tørnes et al., 2009). In this regard, there are only two
software packages recognized in the industry for the dynamic on-bottom stability analysis of
the subsea pipelines: PONDUS (Holthe et al., 1987) and AGA Level 3 (Allen et al., 1989;
Lammert et al., 1989; PRCI, 2008). It is possible to use the other general FE packages such as
ABAQUS (Dassault Systèmes, 2021) and ANSYS (ANSYS Inc., 2021) with integrated
modules for modeling the hydrodynamic loads and the pipe-soil interaction. (Ose et al., 1999)

21
introduced the first finite element model using ABAQUS (Hibbitt et al., 1988) for the
dynamic on-bottom stability analysis of the pipelines by using simplified modeling of the
hydrodynamic loads and the pipe-soil interaction; (Y. Bai & Yu, 2011) used ABAQUS to
compare the on-bottom stability analysis results from the finite element model against those
of the DNV recommended practice (DNV, 1988) to validate the numerical results. ABAQUS
has also been used to build (DNV, 1988) to ensure the validity of the numerical model.
ABAQUS has also been used to build and solve the numerical model of pipeline stability
problems using a finite element method as presented in (Yang & Wang, 2013); and (Y. Bai et
al., 2014). Based on different methodologies of calculating the hydrodynamic loads and the
pipe-soil interaction, other modules have been developed and integrated with the ABAQUS
FE solver for dynamic on-bottom stability analysis of the pipelines. These modules are
referred to by different names, such as SIMSTAB (Zeitoun et al., 2009), UWAINT (Tian et
al., 2015), and CORUS-3D (Atteris Pty Ltd, 2011). SIMSTAB has been utilized in several
case studies as presented in (Anderson et al., 2017; Kien et al., 2010; Robertson et al., 2015),
(Robertson et al., 2015). UWAINT has been developed based on several numerical models
developed by (Tian & Cassidy, 2008, 2010, 2011). CORUS-3D has been used to study the
dynamic response of pipelines as presented in (Jas et al., 2012; McMaster et al., 2012;
Youssef & O’Brien, 2017).

2.7.3.2. Generalized Lateral Stability Method (GLS)

The GLS method is based on design curves employing non-dimensional parameters that are
extracted from numerical simulations using PONDUS (Holthe et al., 1987) following the
recommendations of dynamic stability (DNV, 1988) on a flat seabed and ignoring
deformations from axial and bending forces of the pipeline. In (DNVGL, 2017a), the range of
parameters for which the GLS method is applicable has expanded significantly compared to
the range of applicability of the parameters defined in RP-E305 (DNV, 1988). The GLS is
calibrated by dynamic finite element analysis using half or ten diameters as allowable
displacement criterion under extreme storm conditions. The pipe-soil models from (Verley &
Sotberg, 1994) and (Verley & Lund, 1995) were used in the calibration. Although GLS
remains user-friendly, it is still empirical and has limited application flexibility.

(DNVGL, 2017a) recommends limiting the accumulation of the lateral displacement in the
installation condition and during operation to ten diameters. For larger displacements that
exceed ten diameters, a full dynamic analysis should be carried out and special considerations
concerning bending and fatigue should be taken.

2.7.3.3. Absolute Lateral Static Stability Method (ALSS)

ALSS method assures absolute stability with zero movements to ensure that the resistance
from the soil and pipe’s weight can resist the extreme hydrodynamic loads on the pipe. The
hydrodynamic loads are calculated from the current profile and an extreme single design wave
which often leads to the excessively heavy pipeline. It is also ignoring the fundamental of
fluid-pipe-soil interaction.

The final design weight in the ALSS method is subject to the assumed lateral resistance. A
single friction factor related to submerged weight, as described in Eq. (2.11), is the simplest
method. Friction factor values of 0.6 for sand and rock and 0.2 for clay are often assumed.
Additional resistance is often assumed by considering additional lateral passive soil resistance
from soil heaves on both sides of the pipe as shown in Figure 2.10 and the additional
resistance term is added to the right side of Eq. (2.11).

22
2.7.4. Limitations of DNV Design Methods
The conventional on-bottom stability design methods outlined in (DNV, 1988) and the
updated edition of (DNV, 2007) are not considering the nature of seabed sediments, seabed
instability, and the level of pipeline embedment (Jas et al., 2012).

(DNV, 1988) does not provide guidance on pipe-soil interaction forces for pipelines on
carbonate soils; does consider the effect of pipeline embedment on soil resistance and
hydrodynamic loads and does not consider the effects of seabed instability within the pipeline
response during storm build up.

The updated recommended practice (DNV, 2007) and its newer edition (DNVGL, 2017a) do
allow for some effects of pipeline embedment; however, the do not consider asymmetrical
embedment levels and also do not provide guidance for carbonate soils.

Both recommended practices (DNV, 2007) and (DNVGL, 2017a) focus on pipeline on-
bottom stability with relatively low embedment levels and are not suited to the assessment of
highly embedded pipeline sections.

In addition, none of the existing recommended practices considers the effects of the changes
in seabed bathymetry and characteristics during a storm condition (i.e., when subjected to the
wave and current induced hydrodynamic loads) on pipeline stability. Such changes can
include sediment scour and deposition, excess pore pressure build-up and dissipation, and,
sometimes, liquefaction. See Figure 2.14.

Figure 2.14: Limitation of current stability design approach (McMaster et al., 2012)

2.8. AGA/PRCI Methodology

PRCI offers one of the oldest pipeline stabilization methodologies. During the 1970s and
1980s, PRCI undertook a large multi-year attempt to establish the technical basis for
determining the stability of subsea pipelines under the wave and current actions. Several
research and laboratory experiments have been conducted by (Allen et al., 1989; Brennodden
et al., 1989; Hale et al., 1989; Jacobsen et al., 1989; Lammert et al., 1989) for the PRCI
pipeline on-bottom stability projects up to that time.

23
During the PRCI's on-bottom stability research, three levels of stability design have been
identified (Level 1, Level 2, and Level 3), and computer software has been developed for each
level of design. The three levels of design, the assumptions associated with each level, the
data requirements, and the design procedures are described below (PRCI, 2008).

These design procedures differ in level of accuracy, type of data required, degree of
complexity and the type of analysis associated with the procedure. The Level 1 (static) design
procedure is representative of the manner in which pipeline stability design has traditionally
been performed. Levels 2 and 3 incorporate the results of the PRCI's latest research into
hydrodynamic and pipe/soil interaction forces. As such, these two procedures represent an
improved approach to submarine pipeline stability design. The two new methods differ from
the traditional method in several important areas (PRCI, 2008).

1. In oscillatory flow conditions, the hydrodynamic forces predicted by the Level 2 and
Level 3 software are typically much larger than those calculated using the traditional,
Level 1 approach. The Level 2 and 3 software is accurate, and hydrodynamic forces
have typically been under predicted in the traditional approach.
2. Levels 2 and 3 account for the effect that past cyclic loading history has on pipe/soil
interaction and, as a result, soil resistance forces are typically larger than those
considered for a Level 1 analysis. Again, the Level 2 and 3 predictions more
accurately represent actual soil resistance.
3. Levels 2 and 3 allow a much better representation of the sea state than the "design
wave" approach used in Level 1.

2.8.1. Level 1 Analysis

Level 1 is the simplest of the three levels. The stability problem is considered static without
considering the dynamic effects. The pipeline must, therefore, stay stationary on the seabed
with no displacement. The results of the analysis give the required weight or thickness of the
coating to satisfy the no displacement criterion. Level 1 method uses a single wave with
Morison's force equation; hence attention should be given to the hydrodynamic coefficients
used in the analysis and the maximum wave characteristics to avoid unpredicted pipeline
displacement (PRCI, 2008).

In the following conditions, it is expected that Level 1 is used:

• To obtain a simple, quick, reasonable estimate of historical concrete coating

requirements.
• To check Level 2 or Level 3 analysis against a more conservative criterion, used in
earlier designs.

2.8.2. Level 2 Analysis

The Level 2 analysis is used to determine CWC requirements based on the latest pipeline
hydrodynamic and soil force formulations. It is a simplified method for estimating pipeline
embedment and the resulting soil resistance. This analysis should be used for all preliminary
designs and resulting concrete thicknesses should be sufficient for most final designs (PRCI,
2008).

2.8.3. Level 3 Analysis

24
In a Level 3 analysis, a two-dimensional dynamic pipeline model is used to calculate stresses
and deflections in the line during storm conditions. Input parameters are specified over a grid
taking account of lengthwise variations along the pipeline route. A time history of the on-
bottom current velocities and wave kinematics is simulated by the software to model the
pipeline behavior in irregular seas. The pipeline model considers the effect of end restraints
such as risers and anchor points by modeling them as a series of springs on the end of the
pipeline (PRCI, 2008).

It is expected that this type of analysis is used to check and/or calibrate designs produced
using a Level 2 analysis, particularly where the Level 2 assumptions do not accurately
represent the actual situation. This may occur close to end restraints or other structural
restraints, if axial tension could develop, or if excessive pipe movements are allowed.

Level 2 is recommended for most cases and is widely recognized in the industry. Level 3 can
be used to predict pipeline movements especially for dense sand or stiff clay where the pipe
embedment does not take a big role. However, Level 3 takes a long computer running time
and it is difficult to estimate how far the pipeline will move over the design life. Therefore,
Level 3 is not recommended unless small savings of concrete coating can significantly affect
the project cost.

2.8.4. Factor-of-Safety and Risk Levels

In performing any design there is a degree of uncertainty. The purpose of using safety factors
is to account for the level of uncertainty associated with the design data and analysis
technique. Also associated with any design is a level of risk that the design conditions may be
exceeded. This acceptable level of risk depends on individual company risk philosophy and/or
code requirements. Realistically, there is no way of quantifying the actual "risk of failure" of a
pipeline because of the numerous uncertainties associated with each design variable,
environmental input, and analysis technique (PRCI, 2008).

2.8.4.1. Level 1 and Level 2

In the Level 1 and 2 analyses, factor-of-safety against lateral movement is defined as the net
hydrodynamic in-line force divided by the available soil resistance at the instant of least
stability during passage of a wave. In theory, an appropriate factor-of-safety against lateral
movement should be selected based on the level of confidence in the design data and design
calculations. That is, estimates of the possible error in soil and hydrodynamic forces should be
made, and then a factor-of-safety selected to be consistent with the error estimate. In practice,
however, such evaluations are almost never made, instead more emphasis is placed on
precedence (what has been used before). This is reasonable and is an efficient way to select an
appropriate factor-of-safety (PRCI, 2008).

In traditional design, it has been common to use a factor-of-safety from 1.0 to 1.1. When
accurate assessments of both hydrographic as well as geotechnical information are available,
and the design sea state is known to a high degree of confidence, then a factor-of-safety of 1.0
seems appropriate for the Level 2 design. Another consideration is that during data collection,
the tendency is to err on the side of conservatism. That is, when collecting wave height data or
soil properties, if there is a question regarding which data to use; the data which produces the
most conservative design is usually selected (i.e., the higher wave height and period, worst
angle of attack, the larger force coefficients, the smaller boundary layer, the smaller soil
resistance value, the larger current, etc.). With conservatism built into every input, there is no

25
need to add still another. From this point of view, a factor-of-safety of 1.0 also seems
appropriate (PRCI, 2008).

2.8.4.2. Level 3

In the Level 3 analysis, there is no factor-of-safety which comes from the analysis. Instead,
the output indicates:

1. pipe movements, and

2. pipe stresses

Thus, definitions of "allowable movements" and "allowable stresses" must be selected which
include adequate safety against pipe failure.

2.8.5. Stability Design Procedure Assumptions

This subsection discusses the basic assumptions embodied in the design procedures. The basic
assumptions for the Level 1 static and the Level 2 quasi-static analyses are as follows:

1. The pipe is assumed stationary and resting on the seabed.

2. Only a short length of pipe is considered in the analysis and this length of pipe is
assumed to act independent of adjacent pipe.

The effect of pipe stiffness is not included in the analysis. Typically, this is a reasonable
assumption because similar hydrodynamic forces act along a significant portion of the
pipeline regardless of wave spreading. However, for cases where the characteristic length
(based on bending stiffness) of the pipe is greater than crest wavelength (or length of pipe
affected by similar hydrodynamic forces), this assumption may lead to overly conservative
designs. This is especially true in cases of large diameter pipes where wave spreading is large
or incident waves approach at angles close to the pipe axis (60 degrees or less) (PRCI, 2008).

The assumptions made for the Level 3 dynamic analysis include the following:

1. The irregular sea state governing design can be represented by superpositioning a

series of randomly phased sinusoidal waves with user specified height and period
distribution (or spectral properties).
2. Wave energy is directionally spread (independent of frequency) based on a normal
distribution with user specified standard deviation and mean direction.
3. Hydrodynamic forces are reduced based on the relative velocity between pipe and the
water near the pipe.
4. An initially smooth seabed is assumed for purposes of calculating soil resistance.
5. The pipeline is initially straight.
6. Lateral motions of the pipeline are small.
7. The mass per unit length, bending stiffness, and tension are constant along the pipeline
axis.

2.8.6. Selection of Appropriate Design Procedure

As previously stated, the Level 1 design procedure should only be used to compare results of
the new procedures (Levels 2 and 3) to the traditional method of design. It should not be
selected as the primary design tool. The use of Level 2, Level 3, or both during the design

26
process depends primarily on the nature of the bottom soil conditions, and the type of
information required from the analysis. However, in general, the design process should begin
at Level 2 since the Level 3 procedure is much more time consuming. The effect of numerous
weight coating selections can be quickly checked with Level 2, and these results can
subsequently be used to guide the Level 3 analysis if it becomes necessary (PRCI, 2008).

A Level 3 analysis will not typically be required in the weaker end of the spectrum of typical
marine sediments. In clays, Level 3 analysis will not necessarily be of interest if the Level 2
analysis indicates the pipe to be stable and the undrained shear strength of the seabed soil is
less than about 100 psf. In sands, the same is true for a relative density less than about 60% to
70% (PRCI, 2008).

In dense sands or stiff clays, a pipe will not embed a great deal regardless of its submerged
weight. Level 3 analysis can be used to predict the nature of pipe movement, and the level of
bending stress induced into the pipe by the movement (PRCI, 2008).

There has not been extensive use of the latest version of the pipe/soil interaction module of
the Level 3 analysis software. However, the following type of results are anticipated based on
limited Level 3 analyses and should be used as guidance for determining when a Level 3
analysis may benefit the design (PRCI, 2008).

1. In both sands and clays, a pipe design which has been predicted to be stable with the
Level 2 program will exhibit little or no movement when analyzed with the Level 3
software. A limited number of Level 3 runs have been made to confirm this.
2. If the submerged weight of a pipe is reduced below that required to meet the Level 2
stability criteria, the behavior of the pipe will strongly depend on the strength of the
bottom soils.
In loose sands, soft clays - some scatter will be observed if several Level 3 analyses
are conducted using different sea state realizations. The scatter in the Level 3 results is
expected since:
a. at the end of the build-up sea state, the pipe will likely have embedded in the
weak bottom materials to a point where it is stable in all but the largest waves
in the 3-hour design sea state;
b. during the 3-hour design sea state, the pipe is likely to "break-out" of the soil
when one of the larger waves is applied. If a "break-out" occurs, the available
soil resistance will change by a large degree since, in weak soils, a substantial
portion of the soil resistance is load history dependent. After a "break-out", the
history term is reset to zero;
c. with the large change in available soil resistance at break-out, the pipe will
begin to undergo large movements, and may not become re-embedded in the
soil; and,
d. the time at which the break-out occurs will change from one sea state
realization to another; and as a result, the net movement of the pipe after break-
out will likely differ from one realization to another by a large degree.

In Dense Sands, Hard Clays - very little scatter is expected for different sea state
realizations. This scatter is expected because the load history dependent soil
resistance term is likely to be small in comparison to the history dependent portion
of the soil resistance.

27
To summarize, pipe designed to be stable using Level 2 analysis will not move significantly
(more than 1 diameter) during a Level 3 analysis. Pipes which Level 2 analysis indicates are
unstable may move significantly during a Level 3 analysis. This is especially true for soft
clays and loose sands where much of the soil resistance stems from the pipe embedment into
the soil during the storm build-up. If movements cause the pipe to break-out, this loss of soil
resistance then makes the pipe very unstable and large movements can be expected. Thus,
pipes in soft clays and loose sands should be designed to be stable using Level 2; however,
pipes in stiff clays or dense sands may be designed with either Level 2 (no net movements) or
Level 3 allowing some pipe movement (PRCI, 2008).

2.8.7. Limitation of AGA/PRCI Levels

• Level 1 and Level 2 design methodology do not allow for pipeline displacement.
• The hydrodynamic model is not based on field measurement data, it’s based solely on
laboratory model tests.
• The software does not assume a design sea spectrum but allows the user to specify the
sea state representation.

2.9. Methodologies Comparison

RP-F109 and AGA/PRCI design methodologies give a very detailed design procedure to
which most pipeline engineers refer to. Therefore, a comparison between these two
methodologies has been briefed as follows:

2.9.1. Similarities
1. Both methods are based on very similar dynamic simulation tools which use a FE
pipeline model subject to irregular waves plus steady current loadings.
2. Both methods include history dependent pipe-soil interaction models, where soil
resistance and pipe embedment are based on recent pipe movement history.
3. Both assume that the duration of the design sea-state is 3 hours.
4. The hydrodynamic force formulations are quite different between the two design tools.
The resulting forces produced by both models are very similar though. This
demonstrates the accuracy of the hydrodynamic forces’ estimation.

2.9.2. Differences
1. The AGA/PRCI Level 2 and GLS method have been developed based on very
different approaches.
• The AGA/PRCI tool performs a simplified analysis by estimating wave
content of the design event. It then simulates the embedment process checks
static stability after some embedment, i.e., zero net pipe movement.
• The RP-F109 non-dimensionalises results of numerous dynamic analyses and
presents these results in general form. These results anticipate limited pipe
movements between zero to 10 pipe diameters.
2. DNVGL-RP-109 was developed based on field measurement data and laboratory tests.
The AGA/PRCI’s hydrodynamic model is based solely on laboratory model tests.
3. DNVGL-RP-109 design procedure assumes that the pipe has no initial embedment at
the time the design sea state occurs. The Level 2 design procedure uses a conservative
estimate of the pipe embedment expected during storm build up to the design sea.

28
 CHAPTER 3: ON-BOTTOM STABILITY DESIGN
METHODS

3.1. Introduction
On-bottom stability analysis is performed to ensure the stability of the pipeline when exposed
to wave and current loadings. The required target is that no lateral movements of the pipeline
at all are accepted, or that certain limited movements neither cause interference with adjacent
objects nor overstressing the pipe.

This chapter describes the procedure for assessment of on-bottom stability design based on
RP-F109 (DNVGL, 2017a) and AGA/PRCI methodology (PRCI, 2008) with focusing on
AGA/PRCI Level 2 that is used as a design tool in this thesis.

3.2. DNVGL-RP-F109 Recommended Practice

This recommended practice describes three different methods for on-bottom stability
assessment with different levels of complexity starting with the absolute lateral static stability
(ALSS) method as a simple design method, generalized lateral stability (GLS) method, and
dynamic lateral stability analysis (DLSA). The first method is based on two-dimensional (2D)
force balance equilibrium equations while the second method is based on extensive dynamic
modeling results of 2D pipeline models. These two methods are intended to provide
conservative on-bottom stability designs for rigid pipelines. The third method involves
detailed finite element modeling of the pipeline under a time-domain hydrodynamic loading
(Youssef & O’Brien, 2017).

3.2.1. Load Combinations

As stated in RP-F109 (DNVGL, 2017a), for permanent operation and temporary phases with a
duration of more than 12 months, a 100-year Return Period Value (RPV) applies. If no data is
available, the condition may be approximated by the most severe condition among the
following two combinations:

1. The 100-year RPV for waves combined with the 10-year RPV for current.
2. The 10-year RPV for waves combined with the 100-year RPV for current.

For a temporary phase with a duration of fewer than 12 months but more than three days, a
10-year return period condition applies. An approximation to this condition is to use the most
severe condition among the following two combinations:

1. The 10-year RPV for waves combined with 1-year RPV for current.
2. The 1-year RPV for waves combined with the 10-year RPV for current.

One should make sure that the season covered by the environmental data is enough to cover
uncertainties in the beginning and the end of the temporary condition, e.g., delays.

For a temporary phase with a duration of fewer than three days, an extreme load condition
may be specified based on reliable weather forecasts.

29
A pipeline near the seabed is exposed to hydrodynamic forces induced by waves and currents.
The following paragraphs present the procedures and equations for calculating the wave and
current velocities at the seabed level based on RP-F109 (DNVGL, 2017a).

The subsea environmental conditions play a prime role in the design of almost all subsea
pipelines. Environmental data relevant to subsea pipelines include:

1. Wave characteristics (wave heights, periods, and directions), as it generates

hydrodynamic forces on pipelines.
2. Wind characteristics (wind speeds and directions), as it decided the sea state.
3. Current characteristics (currents and directions), as it causes hydrodynamic loads on
pipelines.

These environmental data are used in the following analysis for subsea pipelines:

• Pipeline installation analysis

• Pipeline stability analysis
• Pipeline free span analysis

The wave loads are generated by a representative irregular wave, which can be described by
its height and its associated periods (Q. Bai & Bai, 2014).

3.2.2. Current Loadings

Based on RP-F109 (DNVGL, 2017a), the steady current flow at the pipeline level may have
components from:

• tidal current;
• wind-induced current;
• storm surge induced current;
• density-driven currents.

Different seabed roughness (𝑧0 ) figures based on soil type are presented in Table 3.1. Each
soil type affects current velocity in a specific way because of the induced friction beneath the
fluid particles (near the seabed) and the soil upper surface.

Table 3.1: Seabed roughness (DNVGL, 2017a)

Seabed Grain size, 𝑑50 [mm] Roughness, 𝑧0 [m]
Silt and clay* 0.0625 ≈ 5·10-6
Fine sand 0.25 ≈ 1·10-5
Medium sand 0.5 ≈ 4·10-5
Coarse sand 1 ≈ 1·10-4
Gravel 4 ≈ 3·10-4
Pebble 25 ≈ 2·10-3
Cobble 125 ≈ 1·10-2
Boulder 500 ≈ 4·10-2
*For a clay seabed, the seabed roughness parameter of silt should be used.

30
The pipeline location at the velocity boundary layer lowers the effective velocity at pipeline
height and can be used in the analysis to measure the mean current velocity across the
pipeline diameter as shown in Figure 3.1.

The reference current should be measured at a depth at which the mean current varies only
slightly in the horizontal direction. On a relatively flat seabed, this reference height should be
larger than 1m depending on the seabed roughness.

Figure 3.1: The influence of current velocity on a pipeline laid on the seabed (Q. Bai,
2013)

The current velocity (𝑈(𝑧)) may be reduced to consider the effect of the bottom boundary
layer and directionality by using Eq. (3.1).

ln(𝑧 + 𝑧0 ) − ln(𝑧0 )
𝑈(𝑧) = 𝑈(𝑧𝑟 ) ⋅ ⋅ sin(𝜃𝑐 ) Eq. (3.1)
ln(𝑧𝑟 + 𝑧0 ) − ln(𝑧0 )
where,
𝑧 is the pipeline elevation above the seabed
𝑧0 is the seabed roughness
𝑧𝑟 is current reference measurement height above the seabed
𝑈(𝑧) is the current velocity
𝑈(𝑧𝑟 ) is the current velocity at current reference height
𝜃𝑐 is the angle between the current velocity and the pipeline axis

The mean current velocity (𝑈𝑐 ) over a pipe diameter is given by Eq. (3.2).

𝑧 𝐷
(1 + 𝐷0 ) ⋅ ln (𝑧 + 1) − 1
0
𝑈𝑐 = 𝑈𝑐 (𝑧𝑟 ) ⋅ ( 𝑧𝑟 ) ⋅ sin(𝜃𝑐 ) Eq. (3.2)
ln (𝑧 + 1)
0

where, 𝐷 is the external pipe diameter.

The directionality of the current velocity is accounted for through 𝜃𝑐 . If the information on
current directionality is unavailable, it should be assumed to act perpendicular to the pipeline.

31
The non-linear interaction between wave and current flow may be accounted for by modifying
the steady current velocity profile.

3.2.3. Wave Loadings

The wave-induced oscillatory flow condition at the pipe level may be calculated using
numerical or analytical wave theories. The wave theory can describe the conditions at the pipe
location, including the shallow water effects if applicable.

In the field, it is well known that the most commonly used wave spectra are Pierson-
Moskowitz spectrum (Pierson & Moskowitz, 1964) and Joint North Sea Wave Project
(JONSWAP) spectrum (Hasselmann et al., 1973). The JONSWAP spectrum (Hasselmann et
al., 1973) is similar to Pierson-Moskowitz spectrum (Pierson & Moskowitz, 1964), except
that the waves continue to grow at a distance (or time) and the peak of the spectrum is
stronger. The wave peak is particularly important as it leads to an increase in the non-linear
interactions and a time-changing spectrum. JONSWAP spectrum (Hasselmann et al., 1973) is
considered in the analysis carried out in this thesis because it describes wind sea conditions
that are reasonable for the most severe sea states.

Figure 3.2: The Pierson-Moskowitz and JONSWAP spectra (Abankwa et al., 2015)

The spectral density function 𝑆𝜂𝜂 for the JONSWAP spectrum (Hasselmann et al., 1973) as
described in RP-F109 (DNVGL, 2017a) is given by Eq. (3.3).

2
𝜔−𝜔𝑝
5 𝜔 −4 (−0.5( ) )
(−4(𝜔 ) ) 𝜎⋅𝜔𝑝 Eq. (3.3)
𝑆𝜂𝜂 (𝜔) = 𝛼 ⋅ 𝑔2 ⋅ 𝜔−5 ⋅ 𝑒 𝑝
⋅ 𝛾𝑒
where,
𝑔 is the gravitational acceleration
𝛼 is the Generalized Phillips’ constant
𝛾 is a peak-enhancement factor

32
𝜔 is the wave frequency
2𝜋
𝜔𝑝 is the peak wave frequency = 𝑇 , where 𝑇𝑝 is the peak period for the design spectrum.
𝑝

The Generalized Phillips’ constant 𝛼 is given by Eq. (3.4).

5 𝐻𝑠2 ⋅ 𝜔𝑝4
𝛼= ⋅ ⋅ (1 − 0.287 ⋅ ln 𝛾) Eq. (3.4)
16 𝑔2

where 𝐻𝑠 is the significant wave height during the sea state.

The spectral width parameter 𝜎 is given by Eq. (3.5).

0.07, 𝑖𝑓 𝜔 < 𝜔𝑝
𝜎={ Eq. (3.5)
0.09, 𝑒𝑙𝑠𝑒

The peak-enhancement factor (𝛾) may be taken as given by Eq. (3.6).

5.0 𝜑 ≤ 3.6 𝑇𝑝
𝛾 = {e(5.75−1.15𝜑) 3.6 < 𝜑 < 5.0; 𝜑 = Eq. (3.6)
1.0 𝜑 ≥ 5.0 √𝐻𝑠

The wave-induced velocity spectrum at the seabed level 𝑆𝑈𝑈 (𝜔) may be obtained through a
spectral transformation of the waves at sea level using a first-order wave theory as given by
Eq. (3.7).

𝑆𝑈𝑈 (𝜔) = 𝐺 2 (𝜔) ⋅ 𝑆𝜂𝜂 (𝜔) Eq. (3.7)

Where the transfer function 𝐺 transforms sea surface elevation to wave-induced flow
velocities at the seabed and is given by Eq. (3.8).

𝜔
𝐺(𝜔) = Eq. (3.8)
sinh(𝑘 ⋅ 𝑑)
where,
𝑑 is the water depth
𝑘 is the wave-number established by iteration from the transcendental equation given in Eq.
(3.9).
𝜔2
= 𝑘 ⋅ tanh(𝑘 ⋅ 𝑑) Eq. (3.9)
𝑔

The nth spectral moment 𝑀𝑛 is defined as given by Eq. (3.10).

𝑀𝑛 = ∫ 𝜔𝑛 ⋅ 𝑆𝑈𝑈 (𝜔)𝑑𝜔 Eq. (3.10)

0

33
The significant flow velocity amplitude 𝑈𝑠 at the pipeline level is calculated by Eq. (3.11).

𝑈𝑠 = 2√𝑀0 Eq. (3.11)

Where 𝑀0 is the zero spectral moment.

The mean zero up-crossing period 𝑇𝑢 of oscillating flow at the pipeline level is calculated by
Eq. (3.12).

𝑀0
𝑇𝑢 = 2𝜋√ Eq. (3.12)
𝑀2

Where 𝑀2 is the second spectral moment.

3.2.4. Wave Directionality and Spreading

The effect of main wave directionality and wave spreading on the wave velocity (𝑈𝑤 ) is given
in the form of a reduction factor (𝑅𝐷 ) on the significant flow velocity (𝑈𝑠 ) as stated in Eq.
(3.13).

𝑈𝑤 = 𝑅𝐷 ⋅ 𝑈𝑠 Eq. (3.13)

The reduction factor 𝑅𝐷 is given by Eq. (3.14).

𝜋⁄2

𝑅𝐷 = √ ∫ 𝐷𝑤 (𝜃)𝑑𝜃 Eq. (3.14)

−𝜋⁄2

where the wave energy spreading directional function (𝐷𝑤 (𝜃)) is given by Eq. (3.15) of the
frequency-independent cosine power function.

1 Γ(1 + 𝑠⁄2) 𝑆 2 (𝜃
𝜋
|𝜃| <
𝐷𝑤 (𝜃) = {√𝜋 Γ(0.5 + 𝑠⁄2) ⋅ cos 𝜃 ⋅ sin 𝑤 − 𝜃)
⋅
2 Eq. (3.15)
0 𝑒𝑙𝑠𝑒
where,
𝑠 is a site-specific spreading parameter
Γ is the gamma function
𝜃𝑤 is the angle between the wave heading and pipeline

Normally 𝑠 is taken between 2 and 8. If no information is available, the most conservative

value in the range 2 to 8 is selected. A value in the range of 6 to 8 may generally be used in
the North Sea. Figure 3.3 shows the effect of wave spreading and directionality on the
reduction factor.

34
 Figure 3.3: Reduction factor due to wave spreading (𝑹𝑫 ) versus relative angle between
wave and pipeline (𝜽𝒓𝒆𝒍 ) (DNVGL, 2017a)

Figure 3.4 is the summary of calculatation procedure of the significate flow velocity
amplitude and mean zero up-crossing period for wave, and mean velocity over the pipe
diameter for current as an input for a selected stability method.

Figure 3.4: Calculation flowchart for wave and current velocities

3.2.5. Vertical Stability

3.2.5.1. Vertical stability in soil

Pipelines that are intended to be buried should be checked for possible sinking or floatation.
Sinking should be considered with maximum content density, e.g., water-filled, and floatation
should be considered with minimum content density, e.g., air-filled.

If the specific weight of the pipe is less than that of the soil (including water contents), no
further analysis is required to document the safety against sinking. For pipelines to be placed
on or in soils having low shear strength, a consideration of the soil stress may be necessary. If
the soil is, or is likely to be, liquefied, the depth of sinking should be limited to a satisfactory
value by consideration of the depth of liquefaction or the build-up of resistance during the
sinking.

35
If the specific gravity of the pipe is less than that of the soil, the shear strength of the soil
should be documented as being enough to prevent floatation. Consequently, in soils that are or
may be liquefied, the specific weight of the pipe should not be less than that of the soil if
burial is required.

Exposed pipelines resting directly on the seabed should be checked for possible sinking in the
same manner as explained above for buried pipes.

3.2.5.2. Vertical stability in water

In order to avoid floatation in water, the submerged weight of the pipeline should meet the
criterion given by Eq. (3.16).

𝑏
𝛾𝑤 ⋅ ≤ 1.0 Eq. (3.16)
𝑊𝑠 + 𝑏
where,
𝑏 is the pipe buoyancy force
𝑊𝑠 is the submerged pipeline weight
𝛾𝑤 is a safety factor.

If a sufficiently low probability of negative buoyancy is not documented, the safety factor
𝛾𝑤 = 1.1 can be applied.

3.2.6. Lateral Stability Methods

3.2.6.1. Absolute Lateral Static Stability Method (ALSS)

This is based on a static equilibrium of forces that ensures the resistance of the pipe against
the motion is enough to withstand maximum hydrodynamic loads during a sea state, i.e., the
pipe will experience no lateral displacement under the design extreme single wave-induced
oscillatory cycle in the sea state considered.

A pipeline can be considered to satisfy the absolute static stability requirement if both Eq.
(3.17) and Eq. (3.18) are satisfied.

𝐹𝑌∗ + 𝜇 ⋅ 𝐹𝑍∗
𝛾𝑆𝐶 ⋅ ≤ 1.0 Eq. (3.17)
𝜇 ⋅ 𝑊𝑠 + 𝐹𝑅

𝐹𝑍∗
𝛾𝑆𝐶 ⋅ ≤ 1.0 Eq. (3.18)
𝑊𝑠
Where,
𝐹𝑅 is the passive soil resistance
𝐹𝑌∗ is the peak horizontal load defined by Eq. (3.19)
𝐹𝑍∗ is the peak vertical load defined by Eq. (3.20)
𝛾𝑆𝐶 is the safety factor

𝐹𝑌∗ = 𝑟𝑡𝑜𝑡,𝑦 ⋅ 0.5 ⋅ 𝜌𝑤 ⋅ 𝐷 ⋅ 𝐶𝑌∗ ⋅ (𝑈 ∗ + 𝑉 ∗ )2 Eq. (3.19)

𝐹𝑍∗ = 𝑟𝑡𝑜𝑡,𝑧 ⋅ 0.5 ⋅ 𝜌𝑤 ⋅ 𝐷 ⋅ 𝐶𝑍∗ ⋅ (𝑈 ∗ + 𝑉 ∗ )2 Eq. (3.20)

36
where,
𝐷 is the pipeline diameter
𝑟𝑡𝑜𝑡,𝑦 is the load reduction factor in horizontal directions
𝑟𝑡𝑜𝑡,𝑧 is the load reduction factor in vertical directions
𝑟𝑡𝑜𝑡,𝑦 and 𝑟𝑡𝑜𝑡,𝑧 defined by the sum of several reduction factors due to soil penetration,
trenching, and permeable seabed as presented in Section 3.4.5 in RP-F109 (DNVGL, 2017a)
𝐶𝑌∗ is the peak horizontal load coefficient
𝐶𝑍∗ is peak vertical load coefficient
𝑈 ∗ is the design single oscillation velocity amplitude
𝑉 ∗ is the steady current velocity associated with design oscillation perpendicular to the
pipeline.
𝜌𝑤 is the density of seawater

The passive soil force 𝐹𝑅 is defined either for sand or clay soils based on equations presented
in Section 2.6.2.3.2.

The design single oscillation velocity amplitude 𝑈 ∗ can be obtained by Eq. (3.21).

𝑈∗ 1 0.5772
= ⋅ (√2 ⋅ ln 𝜏 + ) Eq. (3.21)
𝑈𝑠 2 √2 ⋅ ln 𝜏
where,
𝑈𝑠 is the significant flow velocity amplitude at the pipeline level
𝜏 is the number of oscillations = 𝑇𝑠𝑡𝑜𝑟𝑚 ⁄𝑇𝑢
𝑇𝑠𝑡𝑜𝑟𝑚 is the duration of the storm sea state

The design single oscillation velocity period 𝑇 ∗ can be taken as given in Eq. (3.22).

𝑇∗ 𝑘 − 5 ⋅ (𝑘𝑡 − 1) ⋅ 𝑇𝑛 ⁄𝑇𝑢 for 𝑇𝑛 ⁄𝑇𝑢 ≤ 0.2

={ 𝑡 Eq. (3.22)
𝑇𝑢 1 for 𝑇𝑛 ⁄𝑇𝑢 > 0.2
where,
𝑇𝑛 is the reference period
𝑑
𝑇𝑛 = √𝑔 , where 𝑑 is the water depth and 𝑔 is the gravity acceleration
𝑇𝑢 is the mean zero up-crossing period of oscillating flow at the pipeline level
𝑘𝑡 is decided according to Eq. (3.23)

1.25 for 𝛾 = 1.0

𝑘𝑡 = {1.21 for 𝛾 = 3.3 Eq. (3.23)
1.17 for 𝛾 = 5.0

Where, 𝛾 is the peak-enhancement factor given by Eq. (3.6).

The peak load horizontal coefficient 𝐶𝑌∗ and peak load vertical coefficient 𝐶𝑍∗ are taken from
Table 3.2 and Table 3.3 respectively, where 𝐾 ∗ is the Keulegan-Carpenter number for single
design oscillation given by Eq. (3.24), and 𝑀∗ is the steady to oscillatory velocity ratio for
single design oscillation given by Eq. (3.25).

37
 Table 3.2: Peak horizontal load coefficient (DNVGL, 2017a)
𝐾∗
𝐶𝑌∗
2.5 5 10 20 30 40 50 60 70 100 ≥140
0 13 6.8 4.55 3.33 2.72 2.4 2.15 1.95 1.8 1.52 1.3
0.1 10.7 5.76 3.72 2.72 2.2 1.9 1.71 1.58 1.49 1.33 1.22
0.2 9.02 5 3.15 2.3 1.85 1.58 1.42 1.33 1.27 1.18 1.14
0.3 7.64 4.32 2.79 2.01 1.63 1.44 1.33 1.26 1.21 1.14 1.09
0.4 6.63 3.8 2.51 1.78 1.46 1.32 1.25 1.19 1.16 1.1 1.05
𝑀∗ 0.6 5.07 3.3 2.27 1.71 1.43 1.34 1.29 1.24 1.18 1.08 1
0.8 4.01 2.7 2.01 1.57 1.44 1.37 1.31 1.24 1.17 1.05 1
1 3.25 2.3 1.75 1.49 1.4 1.34 1.27 1.2 1.13 1.01 1
2 1.52 1.5 1.45 1.39 1.34 1.2 1.08 1.03 1 1 1
5 1.11 1.1 1.07 1.06 1.04 1.01 1 1 1 1 1
10 1 1 1 1 1 1 1 1 1 1 1

𝑇𝑢
𝐾∗ = 𝑈∗ Eq. (3.24)
𝐷
𝑉∗
𝑀∗ = Eq. (3.25)
𝑈∗

where 𝑉 ∗ is the steady current velocity associated with design oscillation.

Table 3.3: Peak vertical load coefficient (DNVGL, 2017a)

𝐾∗
𝐶𝑍∗
≤2.5 5 10 20 30 40 50 60 70 100 ≥140
0 5 5 4.85 3.21 2.55 2.26 2.01 1.81 1.63 1.26 1.05
0.1 3.87 4.08 4.23 2.87 2.15 1.77 1.55 1.41 1.31 1.11 0.97
0.2 3.16 3.45 3.74 2.6 1.86 1.45 1.26 1.16 1.09 1 0.9
0.3 3.01 3.25 3.53 2.14 1.52 1.26 1.1 1.01 0.99 0.95 0.9
0.4 2.87 3.08 3.35 1.82 1.29 1.11 0.98 0.9 0.9 0.9 0.9
𝑀∗ 0.6 2.21 2.36 2.59 1.59 1.2 1.03 0.92 0.9 0.9 0.9 0.9
0.8 1.53 1.61 1.8 1.18 1.05 0.97 0.92 0.9 0.9 0.9 0.9
1 1.05 1.13 1.28 1.12 0.99 0.91 0.9 0.9 0.9 0.9 0.9
2 0.96 1.03 1.05 1 0.9 0.9 0.9 0.9 0.9 0.9 0.9
5 0.91 0.92 0.93 0.91 0.9 0.9 0.9 0.9 0.9 0.9 0.9
10 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9
∗
For 𝐾 ∗ < 2.5, the peak horizontal load coefficient can be taken as 𝐶𝑌,𝐾=2.5 ⋅ 2.5/𝐾 ∗ , where
∗
𝐶𝑌,𝐾=2.5 is the relevant value in under 𝐾 ∗ = 2.5.

3.2.6.2. Generalized Lateral Stability Method (GLS)

One can take advantage of a large reduction in weight requirement by allowing some
displacement which would be limited to a value that most pipelines can experience without
problems, e.g., large strains.

The dimensionless lateral pipe displacement 𝑌 is extensively governed by a set of non-

dimensional parameters as given by Eq. (3.26).

38
 𝑌 = 𝑓(𝐿, 𝐾, 𝑀, 𝑁, 𝜏, 𝐺𝑠 , 𝐺𝑐 ) Eq. (3.26)
Where,
𝐺𝑐 is the clay strength parameter given by Eq. (3.32)
𝐺𝑠 is the sand density parameter given by Eq. (3.31)
𝐾 is the significant Keulegan-Carpenter number given by Eq. (3.28)
𝐿 is the significant weight parameter given by Eq. (3.27)
𝑀 is the steady to oscillatory velocity ratio for design spectrum given by Eq. (3.29)
𝑁 is the spectral acceleration factor given by Eq. (3.30)
𝑌 is equal to 𝑦/𝐷 and 𝑦 is the lateral pipe displacement
𝜏 is the number of oscillations
𝑊𝑠
𝐿= Eq. (3.27)
0.5 ⋅ 𝜌𝑤 ⋅ 𝐷 ⋅ 𝑈𝑠2

𝑇𝑢
𝐾 = 𝑈𝑠 Eq. (3.28)
𝐷
𝑉
𝑀= Eq. (3.29)
𝑈𝑠

𝑈𝑠
𝑁= Eq. (3.30)
𝑔 ⋅ 𝑇𝑢

𝛾′𝑠
𝐺𝑠 = Eq. (3.31)
𝑔 ⋅ 𝜌𝑤
𝑠𝑢
𝐺𝑐 = Eq. (3.32)
𝐷 ⋅ 𝛾𝑠
where,
𝐷 is the pipeline outer diameter
𝑔 is the gravity acceleration
𝑠𝑢 is undrained clay shear strength
𝑇𝑢 is the mean zero up-crossing period of oscillating flow at the pipeline level
𝑈𝑠 is the significant flow velocity amplitude at the pipeline level
𝑉 is the steady current velocity associated with the design spectrum
𝑊𝑠 is the submerged pipe weight
𝛾𝑠 is the dry unit soil weight
𝛾′𝑠 is the submerged unit soil weight
𝜌𝑤 is the density of seawater

The following paragraphs, in this section, provide design curves, in form of tables, for on-
bottom stability of subsea pipelines with an allowed lateral displacement in the range from
less than half a pipe diameter, i.e., for a virtually stable pipe, up to a significant displacement
of 10 pipe diameters during the given sea state. The weight required for obtaining a virtually
stable pipe is here denoted 𝐿𝑠𝑡𝑎𝑏𝑙𝑒 whereas the weight required for obtaining a 10-diameter
displacement is denoted 𝐿10 .

The curves are obtained from many one-dimensional dynamic analyses, i.e., on the flat seabed
and neglecting bending and axial deformation of the pipe due to e.g., elevated temperature,
pressure, and restraints at the pipe ends.

39
All cases with high values of 𝑁, 𝐾 and 𝑀 do not necessarily represent realistic physical
conditions. The given values are not valid for extreme cases requiring a pipe specific weight
𝑠𝑔 > 3 and 𝑠𝑔 < 1.05 either. The specific weight of a pipe (𝑠𝑔 ) is given by Eq. (3.33).

2
𝑠𝑔 = 1 + ⋅𝑁⋅𝐾⋅𝐿 Eq. (3.33)
𝜋

In deep waters, 𝐾 may be very small whereas the presence of current gives a large value of
𝑀, and hence, it is recommended to apply the ALSS method in such cases.

The required weight for an intermediate displacement criterion 𝐿𝑌 can be calculated based on
Eq. (3.34).

log(𝐿𝑠𝑡𝑎𝑏𝑙𝑒 ⁄𝐿10 )
log 𝐿𝑌 = log 𝐿𝑠𝑡𝑎𝑏𝑙𝑒 + ⋅ log(𝑌⁄0.5) Eq. (3.34)
log(0.5⁄(0.01 ⋅ 𝜏))

This design approach is applicable to 𝑁 ≤ 0.024 for clay and 𝑁 ≤ 0.048 for sand.

The curves are valid for 𝐺𝑐 ≤ 2.78 only, but for higher values of 𝐺𝑐 it is recommended to
apply the ALSS method.

The minimum pipe weight required to obtain a virtually stable pipe can be found in Table 3.4
independent of 𝑁.

Table 3.4: Minimum weight for pipe laid on the sand, K ≥ 10 (DNVGL, 2017a)
𝐾
𝐿𝑠𝑡𝑎𝑏𝑙𝑒 ⁄(2 + 𝑀)2
10 15 20 30 40 ≥ 60
≤ 0.2 1.5 1.42 1.35 1.25 1.22 1.22
0.4 1.82 1.7 1.61 1.53 1.5 1.5
0.5 2.19 1.97 1.83 1.69 1.61 1.61
0.6 2.65 2.35 2.18 1.99 1.85 1.72
0.8 3.05 2.55 2.32 2.13 2.01 1.9
𝑀
1 3.05 2.55 2.4 2.2 2.06 1.95
1.5 2.65 2.45 2.36 2.24 2.11 2.09
2 2.5 2.4 2.35 2.27 2.22 2.19
4 2.45 2.4 2.39 2.37 2.37 2.37
≥ 10 2.5 2.5 2.5 2.5 2.5 2.5

For 𝐾 ≤ 5, the required weight is more dependent on 𝑁 and the minimum pipe weight
required to obtain a virtually stable pipe can be found in Table 3.5.

The minimum pipe weight required to limit the lateral displacement to 10 pipe diameters for
pipes laid on sand can be found in Table 3.6.

40
 Table 3.5: Minimum weight for pipe laid on the sand, K ≤ 5 (DNVGL, 2017a)
𝑁
𝐿𝑠𝑡𝑎𝑏𝑙𝑒 ⁄(2 + 𝑀)2
0.003 0.006 0.012 0.024 0.048
≤ 0.2 1.55 1.45 1.34 1.24 1.13
0.4 2 1.65 1.34 1.24 1.13
0.5 3.3 2.6 1.91 1.24 1.13
0.6 3.75 3.07 2.38 1.7 1.13
0.8 4 3.45 2.9 2.36 1.81
𝑀
1 3.9 3.5 3.1 2.71 2.31
1.5 3.25 3.13 3 2.88 2.75
2 2.75 2.75 2.75 2.75 2.75
4 2.6 2.6 2.6 2.6 2.6
≥ 10 2.5 2.5 2.5 2.5 2.5

The minimum pipe weight required to limit maximum relative displacement 𝑌 to less than
0.5 on a clay seabed can be calculated by Eq. (3.35).

𝐺𝑐
𝐿𝑠𝑡𝑎𝑏𝑙𝑒 = 90√ 0.67 ⋅ 𝑓(𝑀) Eq. (3.35)
𝑁 ⋅𝐾
𝑓(𝑀) = (0.58 ⋅ (log 𝑀)2 + 0.60 ⋅ (log 𝑀) + 0.47)1.1 ≤ 1.0

Table 3.6: Minimum weight for pipe laid on sand (DNVGL, 2017a)
𝐾
𝐿10 ⁄(2 + 𝑀)2
=5 10 15 20 30 40 60 ≥ 100
≤ 0.2 0.2 0.41 0.61 0.81 0.69 0.69 0.69 0.69
0.4 0.31 0.62 0.93 0.81 0.75 0.72 0.7 0.7
0.5 0.34 0.69 1.03 0.93 0.83 0.78 0.75 1
0.6 0.79 1.2 1.13 1.1 1.07 1.05 1.03 1.02
0.8 0.85 1.4 1.37 1.35 1.33 1.33 1.32 1.31
𝑀
1 1.6 1.5 1.47 1.45 1.43 1.43 1.42 1.41
1.5 1.8 1.7 1.67 1.65 1.63 1.63 1.62 1.61
2 1.9 1.8 1.77 1.75 1.73 1.73 1.72 1.71
4 2.1 2 1.97 1.95 1.93 1.93 1.92 1.91
≥ 10 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5

The formula of Eq. (3.35) may give large weights for high values of 𝐺𝑐 and if the criterion for
absolute stability gives a lighter weight, that criterion can be applied.

The minimum pipe weight required to limit maximum relative displacement 𝑌 to 10 ⋅

𝜏/1000 on clay, can be calculated by Eq. (3.36).

𝐶2
𝐶1 + 𝑓𝑜𝑟 𝐾 ≥ 𝐾𝑏
𝐿10 𝐾 𝐶3
= 𝐶2 Eq. (3.36)
(2 + 𝑀)2 𝐶1 + 𝐶3 𝑓𝑜𝑟 𝐾 < 𝐾𝑏
{ 𝐾𝑏

41
𝐶1 , 𝐶2 , 𝐶3 𝑎𝑛𝑑 𝐾𝑏 are taken from coefficients tabulated in Appendix A (DNVGL, 2017a).

3.2.6.3. Dynamic Lateral Stability Analysis (DLSA)

The DLSA is based on a time-domain simulation of the pipeline dynamic response, subjected
to hydrodynamic loads from an irregular sea state, soil resistance forces, and boundary
conditions. Usually, the dynamic analysis forms the basis for the validation of other simplified
methods such as the GLS method to assure stability. Full dynamic analysis can provide less
conservative optimized design. Thus, it is normally used for the detailed analysis of critical
areas along a pipeline, such as pipeline crossings, riser connections, etc., when the
uncertainties in the design parameters entails a detailed assessment.

The surface wave spectrum shall be transformed to a time series of the wave-induced particle
velocity at the pipe location on the seabed. Normally a constant current velocity is added to
the wave-induced velocity and the hydrodynamic loads are based on the relative velocity and
acceleration between the pipe and the total particle velocity. The resisting force from the soil
consists normally of two parts, a pure friction term, and a passive resistance term depending
on the pipe’s depth of penetration into the soil.

A complete sea state should be used because of the high nonlinearities involved in the
pipeline response. In the absence of full sea state data, it is recommended that the sea state be
three hours. Sea state build-up may be modeled by applying a linear ramp function on wave-
induced particle velocity and acceleration so that the load increases from zero to full load
during approximately the first 20 percent of the analysis duration. This will subject the pipe to
moderate waves with a small displacement that leads to increased penetration and accordingly
increased passive resistance.

The application of different phase shifts between the harmonic wave components gives rise to
different time series realizations with varying maximum wave height and sequence of waves
that both are important factors for the calculated maximum displacement. Hence, at least
seven analyses with randomly, or onerously, chosen seeds to the random number generator
should be performed.

The simulation time step should be small enough to capture the actual nonlinear behavior of
the pipe-soil interaction. In addition, the axial force (due to internal pressure and temperature)
and end effects should be appropriately taken into account.

The dynamic analysis approach is very time-consuming and requires much design data to be
available, i.e., in the detailed design stage where site-specific environmental data is available
and hence optimization of the pipeline weight may be possible (Amlashi, 2017).

3.3. AGA/PRCI On-bottom Stability Simulator

The simulator has three modules:

• Level 1 - Simple and quick static analysis using a linear wave theory and Morison
equations as presented above, without accounting for pipe movement or self-
embedment.
• Level 2 - Reliable quasi-static analysis using a non-linear wave theory and numerous
model test results considering pipe’s self-embedment.

42
 • Level 3 - Complicated dynamic time-domain analysis using a series of linear waves
and allowing some pipeline movements. It also compares the computed pipe stresses
and deflections with allowable limits.

Among the three levels, Level 2 is widely used in the industry and its used in preparing the
case study analysis in this thesis.

3.3.1. Level 2 Methodology

Level 2 is a quasi-static analysis that has been designed to take advantage of the results from
the PRCI's hydrodynamic and pipe/soil interaction tests. Whereas Level 3 dynamic analysis
requires careful development of the input, Level 2 is easy to use (PRCI, 2008).

Description of the analysis conducted by AGA/PRCI Level 2 is as follows:

1. Based on user inputs, the Level 2 simulator calculates the ordinates of the design wave
height spectral density function. The wave height spectral density function is then
transformed to the corresponding bottom velocity spectral density function. The area
under the bottom velocity spectrum is numerically integrated, and the significant
bottom velocity is calculated. The peak frequency of the bottom velocity spectrum is
determined.
2. Maximum and minimum in-line hydrodynamic forces for the largest 200 waves
contained in an assumed 4-hour long build-up sea state are calculated (the 4-hour long
build-up period is considered to start with a zero-wave height and to linearly increase
with time to the design sea state wave height). The 200 largest waves are characterized
by the five-wave heights illustrated in Figure 3.5.
Wave forces for each of the five-wave heights are calculated using the PRCI's new
hydrodynamic force calculation procedure and the associated database of force
coefficients.

Figure 3.5: Bottom velocity amplitude content during 4-hour build-up

(largest 200 velocity amplitudes used to predict embedment) (PRCI, 2008)

43
3. Maximum and minimum in-line forces for the largest 50 waves during a subsequent 3-
hour long design sea state are calculated as in Step 2 above. These 50 waves are
characterized by the four different wave heights illustrated in Figure 3.6.

Figure 3.6: Bottom velocity amplitude content during the 3-hour design storm
(largest 50 velocity amplitude used to predict embedment) (PRCI, 2008)

4. Based on the forces calculated in Step 2, a conservative estimate of pipe embedment at

the end of the 4-hour storm build-up period is calculated. This estimate is obtained by
subjecting the pipe to 200 small oscillations. The oscillations are limited in amplitude
to be no larger than that which the wave forces can produce, or 0.07 times the pipe
diameter, whichever is smaller. To simulate the build-up sea state, the smaller waves
shown in Figure 3.7 are considered first. Not all the 200 oscillations necessarily
produce pipe embedment. Only the waves which produce in-line forces sufficient to
overcome frictional resistance between the pipe and soil are considered to produce
embedment.
For each of the 200 waves, the in-line hydrodynamic force is reduced to account for
the pipe embedment just prior to its application. The estimated pipe embedment and
the available soil resistance force at the end of the build-up period are then saved for
further processing. Pipe embedment and history-dependent soil resistance are
calculated using the PRCI's pipe-soil interaction model.
5. Based on the forces calculated in Step 3 and the pipe embedment calculated in Step 4,
the amount of additional pipe embedment that can be produced by the 50 largest
waves in the design sea state, is calculated in a fashion similar to that described in Step
4 for the storm build-up period. This embedment and the associated soil resistance
force are saved for further processing.
6. Hydrodynamic forces for a complete wave cycle are calculated for four levels of
bottom velocity which are expected in a 3-hour long design sea state. The four bottom
velocities are given by Eq. (3.37).

44
 𝑈1⁄3 = 1.0 𝑈𝑠
𝑈1⁄10 = 1.27 𝑈𝑠
𝑈1⁄100 = 1.66 𝑈𝑠 Eq. (3.37)
𝑈1⁄1000 = 1.86 𝑈𝑠

Figure 3.7: Level 2 build-up sea state model employed to predict pipe embedment
(PRCI, 2008)

7. Using the soil resistance values obtained in Steps 4 and 5 in addition to the
hydrodynamic forces calculated in Step 6, the minimum factor of safety against lateral
sliding is calculated for the pipe embedment at the end of the 4-hour long build-up
period, and the end of the 3-hour long design sea state.
The minimum factor of safety is calculated as given by Eq. (3.38).

𝜇(𝑊𝑠 − 𝐹𝐿 (𝑡)) + 𝐹𝐻
Factor of Safety = Eq. (3.38)
𝐹𝐷 (𝑡) + 𝐹𝑀 (𝑡)

where 𝐹𝐿 (𝑡), 𝐹𝑀 (𝑡), 𝐹𝐷 (𝑡) are the hydrodynamic lift, inertia, and drag forces (as a
function of time) and 𝐹𝐻 is the history-dependent soil resistance. The factor of safety is
calculated at 1-degree intervals of wave passage for a complete 360-degrees.

The above procedure was adopted based on the results of the typical analysis using Level 3
dynamics to calibrate and confirm that the results for pipe embedment are reasonable and that
the results are conservative.

The pipe embedment developed by the "assumed recent wave history" in steps 2 through 4
above is computed using conservative assumptions which include the following:

a. No pipe embedment is considered to have occurred until just prior to the design storm,
b. A short, 4-hour storm build-up period is assumed to precede the design storm during
which some pipe embedment is allowed to occur,

45
 c. The significant wave height during the build-up period starts at a zero-wave height and
increases linearly with time to the significant wave height of the design storm (see
Figure 3.7),
d. The pipe is considered to undergo only very small oscillations and thus does not
embed as far as it might otherwise.

The Level 2 provides a better estimate of pipe embedment than static calculations (Level 1)
which do not consider the effect of pipe movement, but it does not overestimate the
embedment. It also provides realistic estimates of both hydrodynamic and soil resistance
forces during the design sea state. Figure 3.8 shows Level 2 software logic.

Other assumptions specific to the Level 2 analysis tool are as follows:

1. Wave induced near seabed water particle velocities are assumed to have a Rayleigh
distribution (i.e., similar to the wave height distribution).
2. Bottom velocity amplitudes are based on a 3-hour storm duration with input spectral
parameters.
3. Soil resistance is based on the PRCI's pipe/soil interaction model which includes a
frictional resistance (dependent on the normal force applied to the soil) and a passive
soil resistance (dependent upon pipe embedment and independent of instantaneous
pipe normal force).
4. Pipe embedment at the end of the storm build-up period is based on 200 small
amplitude cyclic oscillations. The amplitude of the oscillations is limited by the
hydrodynamic forces expected from a rapidly developing build-up sea state model.
5. Subsequent pipe embedment during the design storm is estimated using 50 small
amplitude cyclic oscillations of the pipe. The amplitude of these oscillations is also
limited by the hydrodynamic force contained in the storm.

Figure 3.8: Level 2 software package logic (PRCI, 2008)

46
These last two assumptions describe the basis for the soil resistance, and detail the
conservative estimate of both number and magnitude of oscillations expected to embed the
pipe just before the design sea state is encountered. Figure 3.9 shows the logic for determining
pipe embedment at the end of the build-up sea state.

Figure 3.9: Level 2 pipe embedment logic (PRCI, 2008)

3.3.2. Input Form

The input form of the user interface consists of three basic areas. The input area, the message
area and the data range area. Information from the input area is used when the Level 2
program is run. The message area provides the information of the input data status. As the
cursor moves to each new field the data range area display the description, minimum,
maximum and default values of the input field. The information within the pair of the square
brackets [ ] is the data item name used in previous (DOS) version.

The input area is subdivided into six (6) groups namely: Title, Soil Properties, Pipe Properties,
Hydro Properties, Wave, and Concrete thickness ranges. See Figure 3.10 Level 2 Input Form.

3.3.3. Plot Output

The Plot file can be viewed by clicking the Plot option under the Report Menu. This will open
a “Plot Safety Factors” window that allows viewing and printing of the plot of Factor of
Safety and Embedment as a function of concrete thickness. The Plot Safety Factors window
has button options to CLOSE or PRINT the selected project. The upper graph shows the
factor of safety, and the lower graph is the corresponding embedment as a function of
concrete thickness. Press either left or right mouse button inside the graph area to activate a
small yellow text window which shows the curve values (x, y) at the mouse location. See
Figure 3.11 Plot Safety Factors.

47
 Figure 3.10: Level 2 input form (PRCI, 2008)

Figure 3.11: Plot safety factor example (PRCI, 2008)

48
3.4. Flexcom Software
Flexcom is a premium structural analysis software package which has underpinned the
engineering design on some of the world’s most demanding offshore projects. Developed by
Wood's software solution specialists, the program is designed and maintained by engineers
who are immersed in the offshore energy industry. Flexcom’s design philosophy is based on
the provision of advanced computational techniques to provide confidence in the engineering
design, coupled with a user-friendly interface which facilitates optimum productivity. The
software uses an industry-proven finite element formulation, incorporating a hybrid beam-
column element with fully coupled axial, bending and torque forces. Flexcom is a highly
versatile software package, capable of simulating risers, mooring lines, umbilicals, CALM
buoys, offloading lines, pipelines, installation processes, offshore wind turbines and wave
energy devices.

3.4.1. Finite Element Modeling

The pipeline is modeled as a line geometry with automatic mesh generation depending on the
specified element length. The pipeline geometric properties (i.e., internal and external
diameters, wall thickness, Young’s modulus, shear modulus, mass density, etc.) are assigned
to the pipeline. The hydrodynamic loads properties such as drag, inertia, and lift force
coefficients are assigned to the pipeline geometry. External coatings such as corrosion and
concrete coatings are added by defining the thickness and its relevant density for each coating
layer. The sea state consists of two-dimensional irregular sea waves combined with steady
current, and both of them are considered perpendicular to the pipeline as the most
conservative and worst-case scenario. A complete sea state simulation duration of three hours
(10800 sec) is considered as recommended by RP-F109 (DNVGL, 2017a).

In order to obtain a full dynamic stability analysis, three types of analysis are performed and
combined into one dynamic analysis run. These three types of analyses include static analysis,
quasi-static analysis, and time-domain dynamic analysis.

The static analysis is to study the effect of time-invariant loadings such as the current and
temperature loadings. In the initial static analysis, the full model is designed including the
geometrical and material properties of the pipeline, surrounding environment properties,
hydrodynamic properties of the pipeline, and seabed soil properties.

In the quasi-static analysis, any vertical constraints are removed, and the pipe is allowed to
fall on the seabed under the combined effect of the gravity and buoyancy loads. The static and
quasi-static analyses merely facilitate contact initiation between the pipeline and the seabed
and also allow for self-weight penetration of the pipeline based on the seabed soil stiffness.

The dynamic analysis is performed as a time-domain analysis to capture the effect of time-
variant loadings from waves and non-linear pipeline response. For a given design condition,
in which the pipeline is required to withstand a combined static and dynamic loads, Flexcom
(Wood Group, 2018) builds a fully dynamic solution for the combined loads by integrating
static, quasi-static, and dynamic analyses into stages. All loads are combined through all
stages to obtain the pipeline response at the end of the dynamic simulation.

49
3.4.2. Pipe-Soil Interaction
At every iteration at each time step during the analysis, the program checks all the nodes in
the pipeline model for seabed contact. If any node has a seabed contact, based on friction
coefficients, the node is made attached by modeling the seabed as non-linear spring.

In an ideal, Coulomb friction model, each of these non-linear springs would have a force-
deflection relationship in which its stiffness (in the region corresponding to zero-deflection
point as shown in Figure 3.12) is infinite. Such infinite force-deflection relationship is
assumed instantaneously, refers to the longitudinal direction. If there is no longitudinal force
on the node, the node does not allow to move (corresponding to zero deflection in the infinite
force-deflection relationship). Indeed, the node should remain in the same location until the
total nodal force exceeds the limiting friction force (𝜇𝐹𝑐 ). At this instant, the node may move
with by a constant force equal to the limiting friction force.

Flexcom (Wood Group, 2018) uses the modified Coulomb friction model as shown in Figure
3.12, in which the infinite stiffness of the ideal Coulomb friction model (in the region
corresponding to zero-deflection point) is replaced by a very high (but not infinite) stiffness
(𝑘) around the zero-deflection point. Such modification is performed by employing a slightly
modified non-linear spring characteristic. The latter is performed to overcome the
convergence difficulty inside the iterative scheme of the software while searching for the
correct solution of the deflections (as is usually performed in most finite element software
packages). This modification is crucial to the operation of the seabed friction model. As the
displacement at any arbitrary node depends, principally, on the stiffness of this section of the
force-deflection curve, then the stiffness of the spring is given by Eq. (3.39).

Figure 3.12: Coulomb friction model used in Flexcom (Wood Group, 2018)
𝜇𝐹𝑐
𝑘= Eq. (3.39)
𝐿𝑚

where 𝐿𝑚 is the mobilization length which is taken as 5% of steel pipeline outer diameter
(𝐷𝑠 ).

50
3.4.3. Environmental Loads
The environmental loads on the subsea pipelines for a randomly selected sea state are
represented by a combination of two-dimensional irregular waves and current loadings. Such
environmental loads depend on the total water particle velocity at the pipeline level and the
pipeline velocity. The total water particle velocity is calculated as the superposition of wave-
induced particle velocity at the pipeline level generated from transforming the free surface
wave spectrum and the steady current velocity. JONSWAP wave spectrum model
(Hasselmann et al., 1973) as given by Eq. (3.40) is selected to represent the two-dimensional
irregular sea waves.

2
𝜔−𝜔𝑝
5 𝜔 −4 (−0.5( ) )
(−4(𝜔 ) ) 𝜎⋅𝜔𝑝 Eq. (3.40)
𝑆𝜂𝜂 (𝜔) = 𝛼 ⋅ 𝑔2 ⋅ (2𝜋)−4 ⋅ 𝜔−5 ⋅ 𝑒 𝑝
⋅ 𝛾𝑒

1
where 𝜔𝑝 is the peak wave frequency = 𝑇 .
𝑝

Based on the value of 𝜑 = (𝑇𝑝 ⁄√𝐻𝑠 ), the value of Phillips’ constant (𝛼) is calculated from
Eq. (3.41). The spectral width parameter 𝜎 and the peak-enhancement factor 𝛾 are calculated
based on Eq. (3.5) and Eq. (3.6) respectively.

2.73 𝐻𝑠2 ⁄𝑇𝑝4 𝜑 ≤ 3.6

𝛼 = {0.036 − 0.0056𝜑 3.6 < 𝜑 < 5.0 Eq. (3.41)
5.07 𝐻𝑠2 ⁄𝑇𝑝4 𝜑 ≥ 5.0

Different random seed numbers are used to get different time series of wave heights by
assigning different phase shift values to wave components while discretizing the wave
spectrum.

The current loading is represented by a uniform velocity distribution considering the effect of
the boundary layer at the pipeline level and seabed roughness by applying Eq. (3.2) according
to RP-F109 (DNVGL, 2017a).

The hydrodynamic forces induced from the combined actions of wave and current loadings on
the pipeline are calculated using Morison’s equation (Morison et al., 1953) as already
described in Section 2.6.1.

51
CHAPTER 4: ON-BOTTOM STABILITY DEVELOPED
SOFTWARE PACKAGE

4.1. Introduction
The limited availability and the expensive price (if it exist) of specialized software tools that
help deciding the required pipeline stabilizing weight, magnified the need to develop such a
tool. The latter is based on the recommended practice RP-F109 (DNVGL, 2017a). This tool
can calculate the necessary pipeline stabilizing weight and optimum concrete coating
thickness using two methods of stability: ALSS and GLS design methods. Also, it may help
comparing between the results of both ALSS and GLS design methods as well as GLS 0.5D
displacement criterion and AGA/PRCI Level 2 software package.

4.2. Software Development

The developed software tool is a standalone program code designed using MATLAB
programming language (The MathWorks, 2018) for the design and assessment of on-bottom
stability of subsea pipelines in compliance with the recommended practice (DNVGL, 2017a).
Figure 4.1 shows the graphical user interface (GUI) of the main window.

Figure 4.1: Main window of the on-bottom stability software tool

4.3. Functions of the Developed Tool

The developed tool has the following features:

• Calculating pipeline dimensions and total submerged weight based on user inputs.
• Two criteria for pipeline stability design:
- Absolute lateral static stability method (ALSS).
- Generalized lateral stability method (GLS) (less than 0.5D allowable displacement
and up to 10D allowable displacement).

52
 • Different soil models (Sand, Clay, and Rock).
• The irregular sea state condition is represented by JONSWAP spectrum.
• The wave seabed induced velocity is calculated in each direction from user input.
• The reduction effect of the directionality of waves and currents on their induced
velocities is considered.
• Additional seabed penetration is optional as a reduction factor of hydrodynamic
forces.
• Two pipeline phase conditions (installation and operation) are considered.
• The boundary layer effect from the seabed on the current velocity is dealt with.
• The optimal concrete coating thickness and necessary pipeline stabilizing weight
(which ensures lateral and vertical stability for each wave and current direction
combinations) with graphical presentation may be suggested.

4.4. Graphical User Interface (GUI)

The software tool has one main window as shown in Figure 4.1. The main window contains
all necessary data divided into small panels (pipe data, environmental parameters, soil
interaction, properties of the pipeline layers, stability criterion, return period values (RPV) for
wave and current, and pipe condition (empty or operation). All input and output parameters as
shown in the main window are introduced in SI units. The input required for each design
method is limited based on user selection for easy use.

For more convenience, return period values for waves and currents are to be assigned in a
separated window as shown in Figure 4.2 by clicking the “Enter RPV” button. Users can
add/delete directions and their relevant wave and current characteristics or even edit existing
characteristics by just double clicking the chosen cell.

Figure 4.2: Main window for wave and current characteristics input

53
4.5. Brief Operation Instructions
To limit the required user input data, the cells of required data are allowed to be enabled or
disabled based on the user selection. The user selects the stability method and criterion first,
then the soil type model.

The required input data should be assigned for all panel sections as shown in the main
window Figure 4.1. The selection between tabs “Empty Pipe” or “Pipe in Operation”, and its
design condition, is depending on the results that the user requires. Special consideration
should be taken for the input index of “Peak Enhancement Factor”, if the input index is “0”,
the Peak Enhancement Factor will be calculated automatically inside the program according
to Eq. (3.6).

To start the analysis, the “Calculate” button should be clicked and another window appears
(according to the selected stability method) showing the results for each direction
combinations considered from waves and currents. The window should be closed to see the
results of the next direction combinations.

4.5.1. Warning and Error Messages

All parameters should be within acceptable limits that avoids unacceptable or overestimates
results. If there is any parameter outside the acceptable limits, the tool will not accept this
input.

During the calculation, the tool checks the vertical stability. If the pipe is not stable vertically,
i.e., pipe floats, the required CWC thickness is calculated to ensure vertical stability and is
added automatically before checking the lateral stability as shown by the warning message
shown in Figure 4.3. Also, if in case any calculation parameter goes beyond the specified
range governed by RP-F109 (DNVGL, 2017a), an error message as shown in Figure 4.4
appears, indicating the error location, and the calculation is terminated.

Figure 4.3: Warning message window Figure 4.4: Error message window

4.5.2. Output Results

The numerical results of the calculated design parameters for each loading combination from
waves and currents appear according to the selected stability method. Figure 4.5 shows the
results of ALSS method, Figure 4.6 (a) shown the results for GLS (0.5D displacement)
method, and Figure 4.6 (b) shows the results for GLS (10D displacement) method. Each
window shows the condition of stability followed by the required stabilizing weight and its
CWC thickness in case of instability of the pipeline.

After completion of all calculations for all combinations of directions from waves and
currents, the “Results” button in the main window is clicked to show two graphical

54
representations for required stabilizing weight and CWC thickness versus each direction
combination as shown in Figure 4.7.

Figure 4.5: Results window of ALSS method

(a) 0.5D Displacement Criterion (b) 10D Displacement Criterion

Figure 4.6: Results window of the GLS method (a) 0.5D displacement criterion and (b)
10D displacement criterion

55
Figure 4.7: Graphical representation of pipeline weight requirement and CWC
thickness versus combined directions of waves and currents

56
CHAPTER 5: CASE STUDY ANALYSIS AND RESULTS

5.1. Introduction
Pipeline stability against hydrodynamic forces generated from wave and current loadings
requires enough resistance to its lateral displacement, increasing pipeline weight is the
conventional and cost-effective stability method used to increase the pipeline resistance
against environmental loads. Several approaches can be used to determine the submerged
weight required so that the lateral resistance can be large enough to keep the pipe safe against
hydrodynamic loads.

In this chapter, the analysis of a case study is performed to calculate the required concrete
thickness for on-bottom stability. Three design approaches are used in this analysis; static
analysis approach using ALSS method, calibrated design methods approach using GLS
method and AGA/PRCI Level 2 software, and finally dynamic analysis approach using finite
element software called Flexcom based on DLSA guidelines of RP-F109 (DNVGL, 2017a).

5.2. Case Study Design Data

A case study of 22 inches pipeline for a project titled “Gamba Export Loading Line” (Zeetech
Engineering B.V., 2013). This project was executed in Gabon with total offshore length of
about 9.4km. See Figure 5.1 for exact project location. The pipeline is placed on sand soil in
shallow water. This case study was selected due to the availability of its data and the difficulty
of providing a local case study due to confidentiality. Also, its available data was necessary to
compare some of actual output results stated in this case study with the developed MATLAB
software to confirm its accuracy.

Figure 5.1: Gamba Project Location (Zeetech Engineering B.V., 2013)

57
A pipeline of 22-inches with material grade API 5L X65 (API, 2018) and a steel wall
thickness of 9.5 mm, the pipe has material properties and coating layers properties presented
in Table 5.1.

Table 5.1: Pipeline data (Zeetech Engineering B.V., 2013)

Parameter Unit Value
Pipeline outside diameter, D mm 559
Pipe wall thickness, 𝑡𝑠𝑡𝑒𝑒𝑙 mm 9.5
Steel density, 𝜌𝑠𝑡𝑒𝑒𝑙 kg/m3 7850
Minimum product density, 𝜌𝑐𝑜𝑛𝑡 kg/m3 850
Specified min. yield stress, SMYS MPa 448
Young’s modulus, 𝐸 GPa 207
Poisson ratio, 𝑣 - 0.3
Corrosion coating thickness, 𝑡𝑐𝑜𝑎𝑡 mm 3.2
Corrosion coating density, 𝜌𝑐𝑜𝑎𝑡 kg/m3 950
Concrete coating density, 𝜌𝑐𝑜𝑛𝑐 kg/m3 3040

The pipe is installed on sand soil in shallow water. The seawater characteristics and soil
properties are presented in Table 5.2 and Table 5.3 respectively.

Table 5.2: Seawater properties (Zeetech Engineering B.V., 2013)

Parameter Unit Value
Seawater density, 𝜌𝑤 kg/m3 1030
Gravity acceleration, g m/s2 9.81

Wave and current design data are presented in Table 5.4. The current is assumed to be
perpendicular to the pipelines throughout the analysis. The water depth starts from 6.1 up to
25 m with different significant wave heights and wave hitting angles as presented in

Table 5.5. The starting depth of water (6.1m) is selected to avoid the limitation of AGA/PRCI
Level 2 and also to consider the combined effect of waves and currents.

Table 5.3: Soil data (Zeetech Engineering B.V., 2013)

Parameter Unit Value
Friction coefficient, 𝜇 - 0.6
Seabed roughness, 𝑧0 m 1E-5
Submerged unit weight, 𝛾𝑠′ N/m3 9000
Dry unit weight of soil, 𝛾𝑠 N/m3 19000
Soil stiffness kN/m/m 300

Table 5.4: Waves and currents design data (Zeetech Engineering B.V., 2013)
Return Period Value (RPV)
Seawater Characteristics Unit
1-year 10-year 100-year
Peak period, Tp sec 13 13.9 14.3
Current velocity, 𝑈𝑐 (𝑧𝑟 ) m/s 0.23 0.29 0.35
Current reference measurement height, 𝑧𝑟 m 1 1 1
Spreading parameter, 𝑠 - 8 8 8

58
 Table 5.5: Significant wave heights and wave hitting angles relative to the water depth
(Zeetech Engineering B.V., 2013)
Water Significant Wave Height, Hs (m) Wave Hitting
Depth (m) 1-year 10-year 100-year Angle (deg)
6.1 2.4 3.02 3.62 9
8 2.26 2.84 3.4 11
10 2.16 2.71 3.25 12.8
15 2.01 2.52 3.01 16.5
20 1.93 2.41 2.87 65
25 2.01 2.46 2.91 67

The two phases of installation (empty pipe condition) and operation (minimum product
density) are considered with environmental loads from waves and currents as given in Table
5.6.

Table 5.6: Design load conditions

Design condition
Design load
Installation Phase Operation Phase
Environmental load 10-year wave + 1-year current 100-year wave + 10-year current
return period 1-year wave + 10-year current 10-year wave + 100-year current
Internal product Empty Minimum product density

The presented design data is used for on-bottom stability analysis using static and calibrated
methods approaches. Sensitivity analysis is performed to investigate the effect of increasing
the water depth on the calculated concrete coating thickness using the GLS method (0.5D
allowable displacement) and AGA/PRCI Level 2 method as both are calibrated methods.

The numerical results of ALSS and GLS methods are obtained using the developed software
tool (Chapter 4). The combined loadings from waves and currents (i.e., different return period
values from waves and currents) during installation and operation phases provide different
CWC at each water depth. The maximum CWC that satisfies the stability among all water
depths should be selected as the optimum CWC that can resist the hydrodynamic loads from
wave and current loadings on the pipeline against the lateral displacement.

For dynamic analysis method, because of its complexity and its time-consumption, only one
case is selected at a water depth of 20 m under two different loading combination from waves
and currents during only installation phase. Such combined loadings are assumed to be
perpendicular to the pipeline.

5.3. Static Analysis

The ALSS method is used as a static analysis method. The design philosophy of the ALSS
method is that to ensure enough pipeline resistance against displacement to withstand the
maximum hydrodynamic loads resulting from waves and currents. This in turn ensures zero
lateral displacements under the design extreme single wave. This traditional method is still
used in on-bottom stability design of pipelines where no motion of the pipeline is permitted
(such as locations near risers). This method provides high CWC due to the conservatism of its
design restrictions.

59
The safety factor for the ALSS method is taken as unity to remove any uncertainty in this
approach. Two loading combinations are considered in this analysis for each phase. For the
installation phase, 1-year RPV current + 10-year RPV wave and 10-year RPV current + 1-
year RPV are considered. For operation phase, 10-year RPV current + 100-year RPV wave
and 100-year RPV current + 10-year RPV wave are considered.

Figure 5.2 shows the concrete coating thickness required at each water depth using the ALSS
method.

Figure 5.2: Required concrete coating thickness using ALSS method

The required concrete coating thickness for stability is 152 mm at water depth 6.1 m during
both installation and operation phases. The concrete thickness required for stability during
installation phase (empty pipe condition) is higher than the operation phase (filled with
minimum product density condition) because of the light pipeline weight.

5.4. Calibrated Methods

The GLS and AGA/PRCI Level 2 are the calibrated methods used in this case study. The
former method has two design criteria: virtually stable pipe up to 0.5D and allowable
displacement up to 10D. Both calibrated methods are widely used in the industry as they are
considered as the most fast and reliable design methods.

5.4.1. Generalized Lateral Stability Method

5.4.1.1. 0.5D Displacement Criterion

The allowable displacement limit of this stability criterion is up to 0.5D, this criterion is
decided to avoid breaking out from its cavity which makes the pipeline virtually stable. The

60
stability weight required under this criterion is less than that required by the ALSS method to
prevent the lateral displacement under extreme single design wave.

Figure 5.3 shows that the concrete coating thickness required for stability is 69 mm and it
occurs at water depth of 6.1 m during the installation phase. The calculated concrete coating
thicknesses at water depth 6.1 m and 20 m are very close, although the significant wave
height is smaller at water depth 20 m than at water depth 6.1 m. This is because of the wave
hitting angle is larger at water depth 20 m than at water depth 6.1 m, which leads to strong
hydrodynamic forces on the pipe and increases the stability weight requirements.

Figure 5.3: Required concrete coating thickness using GLS (0.5D displacement) method

5.4.1.2. 10D Displacement Criterion

The allowable displacement limit of this stability criterion is 10D which consequently
provides less CWC than the 0.5D displacement criterion.

The required concrete coating thickness for pipeline stability is 43 mm at water depth 6.1 m
and 20 m during installation phase as shown in Figure 5.4. This thickness is almost the
thickness required to ensure the vertical stability of the pipeline as 42 mm is the calculated
concrete coating thickness that is required to avoid floatation. During the operation phase, the
required concrete coating thickness is almost “zero”. This is due the fact that the pipeline is
filled with the minimum product density which provides a sufficient submergence weight for
the vertical stability of the pipeline and ensures allowable displacement within the limit of this
stability criterion.

61
Figure 5.4: Required concrete coating thickness using GLS (10D displacement) method

5.4.2. AGA/PRCI Level 2

To apply the same environmental loads as in the GLS method, the developed on-bottom
stability software presented in Chapter 4 is used to calculate the near seabed wave velocity
and zero-up crossing period. Such results are assigned as the input to the Level 2 software
package.

Conservative assumptions are made for inline and lift force reduction factors, reflecting both
the trench effect and embedment by assuming their values as unity in the input data.

The fundamental acceptance criteria for AGA/PRCI Level 2 stability assessment are as
follows:

• At the end of a 4-hour storm build-up, the pipeline should be stable in U1/100 condition,
• And, at the end of a 3-hour storm period, the pipeline should be stable in the U1/1000
condition.

A minimum factor of safety of 1.0 is used to indicate that the pipeline is stable at the end
of the 4-hour storm build-up in the U1/100 condition.

The required concrete coating thicknesses based on AGA/PRCI Level 2 are presented in
Figure 5.5. The selected concrete coating thickness which satisfies the stability in the
installation and operation phases is 86 mm at water depth of 6.1 m (highest value). A brief
comparison between AGA/PRCI Level 2 and 0.5D GLS methods is presented in the next
section.

62
 Figure 5.5: Required concrete coating thickness using AGA/PRCI Level 2 method

5.4.3. A Brief Comparison Study

5.4.3.1. ALSS method versus GLS methods

Figure 5.6 and Figure 5.7 show that the ALSS method provides higher concrete coating
thickness than GLS methods (0.5D and 10D displacement criteria) during installation and
operation phases. The reason behind this is the absence of lateral movements imposed by the
ALSS method against the maximum hydrodynamic loads, unlike the GLS method which
allows for some lateral displacement ranging from 0.5D up to 10 D.

It is noticeable from the results of both ALSS and GLS methods, that the highest concrete
coating thickness is at a water depth of 6.1 m, although the wave hitting angle is small at this
water depth. The reason for this increase is the large value of significant wave height at water
depth 6.1 m, which increases the wave-induced water particle velocity near the seabed and
consequently generates strong hydrodynamic forces that require high stabilizing weight.

5.4.3.2. AGA/PRCI Level 2 method versus GLS (0.5D displacement) method

As shown in Figure 5.8, the maximum concrete coating thickness calculated by AGA/PRCI
Level 2 method is 86 mm compared to 69 mm calculated by GLS method, noting that both
methods provide high concrete coating thickness during the installation phase because of the
condition of the empty pipe. At 20 m and 25 m water depths, the concrete coating thickness
increases because of the changes in the wave hitting angle that approaches 90°.

63
Figure 5.6: Required concrete coating thickness using ALSS and GLS methods during
installation phase

Figure 5.7: Required concrete coating thickness using ALSS and GLS methods during
operation phase

64
 Figure 5.8: Required concrete coating thickness using GLS (0.5D displacement) and
AGA/PRCI Level 2 methods

5.4.4. Sensitivity Analysis

Sensitivity analysis is conducted using both calibrated methods; GLS (0.5D displacement)
method and AGA/PRCI Level 2 method. The objective of this analysis is to investigate the
effect of water depth as well as effect of each calibrated method on the concrete coating
thickness selection under two different loading combinations from waves and currents (1-year
RPV wave + 10-year RPV current and vice versa). Two more loading combination are
considered to check the effect of wave and current characteristics on the concrete coating
thickness selection (1-year RPV + 1-year RPV and 10-year RPV + 10-year RPV). The wave
hitting angle and characteristics are fixed during this analysis as given in Table 5.7. The
current is assumed to be perpendicular to the pipeline.

Table 5.7 Wave and current design data for the sensitivity analysis
RPV
Seawater Characteristics Unit
1-year 10-year
Peak period, Tp sec 13 13.9
Significant wave height, Hs m 2.01 2.46
Current velocity, 𝑈𝑐 (𝑧𝑟 ) m/s 0.23 0.29
Current reference measurements height, 𝑧𝑟 m 1 1
Wave hitting angle deg 67 67

65
The seawater depths start from 6.1 m due to the limitation of AGA/PRCI Level 2 method and
up to 40 m depth, to consider different loading combinations from waves and currents. The
installation phase is considered the worst-case based on the highest CWC provided in
different loading combinations as shown in the above results.

The wave characteristics have a pronounced effect on the calculated concrete coating
thickness as shown from Figure 5.9 (a) and (c) or Figure 5.9 (b) and (d), while the current has
a slight effect on the calculated concrete coating thickness as shown from Figure 5.9 (a) and
(b) or Figure 5.9 (c) and (d) which appears using the GLS (0.5D displacement) method.

As shown in Figure 5.9 (a-d) under all different loading combinations, the calculated concrete
coating thickness becomes less as the water depth increases, this is as a result of the wave-
induced seabed velocity that is decreasing against increasing water depth under each loading
combination. It is also observed that by increasing the water depth, AGA/PRCI Level 2
method provides concrete coating thickness less than the GLS method under different loading
combinations from waves and currents. For shallow water depth less than 25 m, AGA/PRCI
Level 2 method provides higher concrete thickness than the GLS method.

It is also observed that the difference between the calculated concrete coating thickness using
both methods for water depth less than 20 m is higher than the difference between the
calculated concrete coating thicknesses for water depth greater than 20 m.

(a) 1-year wave + 1-year current (b) 1-year wave + 10-year current

(c) 10-year wave + 1-year current (d) 10-year wave + 10-year current

Figure 5.9: Sensitivity analysis results under different loading conditions using GLS (0.5D
displacement) and AGA/PRCI Level 2 methods

66
5.5. Dynamic Analysis
The dynamic on-bottom stability analysis of the subsea pipeline is performed by using finite-
element-based advanced simulation software, which is named Flexcom (Wood Group, 2018).
Flexcom is a highly versatile software package, which is commonly used in the structural
analysis of conventional and unconventional offshore structures.

The dynamic on-bottom stability analysis is conducted on the pipeline that has the same
properties as given in Table 5.1 in shallow water with a depth of 20 m on a sandy seabed. Sea
state condition consists of two-dimensional irregular sea waves represented by the JONSWAP
spectrum (Hasselmann et al., 1973) combined with a steady current. Installation phase with an
empty pipe condition is only considered in this analysis under the condition of combined
loads of 1-year RPV wave + 10-year RPV current and 10-year RPV wave + 1-year RPV
current.

The acceptance criteria of on-bottom stability design are depending on the design target and
the factors that may affect the integrity of the pipeline. Obviously, the acceptance criteria can
be satisfied either by the maximum allowable lateral displacement or by the stress limits for
the design of the pipeline. The results using both criteria are presented.

5.5.1. Case Study Data

The case study used in the dynamic analysis is that of the 22-inch diameter pipeline used
previously with a 250-meter pipe length subjected to the combined loads from two-
dimensional irregular waves and steady currents. The following sub-sections present the
design data for the dynamic stability analysis.

5.5.1.1. Pipeline Data

The pipeline has a geometrical property in the form of a rigid pipeline with linear elastic
structural properties. The pipeline data and the material properties are presented in Table 5.1.
The pipeline is modeled as a single continuous line restrained just above the seabed as an
initial position. Mesh is generated using a constant element length of 5-meter throughout the
pipe length.

5.5.1.2. Environmental Data

The pipeline is situated at a water depth of 20 m. The seawater characteristics are already
presented in Table 5.2. The characteristic data of the wave and current loadings
(corresponding to two return periods) are presented in Table 5.8. The directions of the applied
waves and currents are assumed to be perpendicular to the pipeline.

Table 5.8: Wave and current design data for the dynamic analysis (Zeetech Engineering
B.V., 2013)
RPV
Seawater Characteristics Unit
1-year 10-year
Peak period, Tp sec 13 13.9
Significant wave height, Hs m 2.4 3.02
Current velocity, 𝑈𝑐 (𝑧𝑟 ) m/s 0.23 0.29

67
5.5.1.3. Seabed and Soil Data

The seabed is considered elastic and flat surface without slope with longitudinal and
transverse friction coefficients. Soil assumed to be loose to medium sand, and elastic seabed
stiffness defined as a linear constant stiffness value. The longitudinal and transverse friction
coefficients for the Coulomb friction model, soil stiffness, and seabed roughness are already
presented in Table 5.3.

In the present case study, the pure friction term is only considered by using the Coulomb
friction model, while passive soil resistance is ignored due to the present capabilities of the
Flexcom software package (Wood Group, 2018).

5.5.1.4. Boundary Conditions

In the initial static analysis, both ends of the pipeline are assumed to be fixed in all
translational and rotational degrees of freedom. An additional boundary condition is applied at
regular intervals of 5-meter along the pipeline to position the pipeline slightly hanging above
the seabed.

In the quasi-static analysis, all vertical constraints along the length of the pipeline are
removed to allow for the pipeline to fall onto the seabed, under the combined influence of
gravity and buoyancy, and to allow for its penetration into the soil.

In the dynamic analysis, the pipeline is assumed to be fixed in all translational and rotational
degrees of freedom at the left end, whereas it is assumed to be fixed in all rotational degrees
of freedom only at the right end.

5.5.1.5. Assumptions

• There is no temperature variation along the pipeline, to avoid excessive lateral

displacements and bending moments.
• There is no temporal deterioration in the pipeline wall thickness due to corrosion.
• A two-dimensional irregular wave is considered satisfactory to represent the sea state.
• The coefficients of the hydrodynamic forces (i.e., drag, inertia and lift) for the pipeline
resting on the seabed are assumed to be constant during the analysis. As stated by
(Verley et al., 1987), the drag, inertia and lift coefficients are often taken according to
(DNV, 1981) rules as equal to 0.7, 3.29, and 0.9 respectively.
• All added coating layers increase the mass of the pipeline per unit length for the
pipeline and do not contribute to the pipe stiffness.
• During the first 100 seconds of the simulation, the wave loads increased from zero to
maximum by applying a linear ramp function according to RP-F109 (DNVGL,
2017a).
• The effect of the cyclic loads on the embedment of the pipeline into the soil is not
considered because of using linear soil stiffness.

5.5.2. Results
The DLSA method is conducted using Flexcom software package to determine the required
concrete coating thickness which satisfies the stability of the pipeline against the applied
hydrodynamic loads and to investigate the dynamic pipeline response. Several simulation runs
are performed under different loading combinations from waves and steady currents using

68
fixed and random seed numbers, for different concrete coating thicknesses. The initial guess
of the concrete coating thickness required for stabilizing the pipeline was determined using
the ALSS method. Also, such initial concrete coating thickness is used as the basis of
conducting the dynamic analysis aiming at minimizing the number of dynamic simulations.

The acceptance criteria, which are used in this case study, are to achieve a relatively stable
pipeline, i.e., to keep the lateral displacement of the pipeline less than half of its external
diameter (0.5D), and to keep the resultant induced stresses within their allowable limits. In
this case study, the allowable limit of the von Mises stress is taken as 430 MPa, i.e., 96% of
the SMYS. The resultant induced stresses are to be checked thoroughly to avoid the excessive
loading effects such as bending moments which may affect the integrity of the pipeline by
causing local buckling or collapsing of the pipeline wall.

5.5.2.1. The Combined Load of 1-year RPV Wave plus 10-year RPV Current

Under the action of the combined loading condition of 1-year RPV for wave + 10-year RPV
for current, and for a concrete coating thickness less than or equal to 145 mm, the pipeline
records an instability as shown in Figure 5.10, and the lateral displacements recorded are
beyond the allowable limit.

The associated von Mises stresses to these lateral displacements are presented in Figure 5.11.
The associated stresses do not exceed the allowable limit, except that recorded at a concrete
coating thickness of 140 mm which reached to 580 MPa. For this range of concrete coating
thicknesses (135-145mm), both design criteria (displacements and stresses) are not satisfied,
which necessitates increasing the stability weight.

By increasing the concrete coating thickness, the pipeline becomes relatively stable at a
concrete coating thickness greater than or equal to 155 mm as shown in Figure 5.12, and the
associated von Mises stresses do not also exceed the allowable limit of stresses as shown in
Figure 5.13. Therefore, the design criteria (displacements and stresses) are satisfied for
concrete coating thicknesses greater than or equal to 155 mm.

5.5.2.2. The Combined Load of 10-year RPV Wave plus 1-year RPV Current

Under the action of the combined loading condition of 10-year RPV wave + 1-year RPV
current, and for a concrete coating thickness less than or equal to 165 mm, the pipeline
records an instability and the lateral displacements recorded are beyond the allowable limit as
shown in Figure 5.14.

The associated von Mises stresses to these lateral displacements are presented in Figure 5.15.
It is shown that the associated stresses do not exceed the allowable limit, except the stresses
recorded at a concrete coating thickness of 160 mm which reached to 650 MPa. For this range
of concrete coating thicknesses, both design criteria (displacement and stress) are not
satisfied, which necessitates increasing the stability weight.

By increasing the concrete coating thickness, the pipeline becomes relatively stable as shown
in Figure 5.16, and the associated von Mises stresses are also varying within the specified
allowable limit as shown in Figure 5.17. Accordingly, both design criteria (displacements and
stresses) are satisfied for concrete coating thicknesses greater than 170 mm.

69
Figure 5.10: Envelopes of lateral displacements for concrete coating thickness range
135-145 mm (1-year RPV wave + 10-year RPV current)

Figure 5.11: Envelopes of von Mises stresses for concrete coating thickness range 135-
145 mm (1-year RPV wave + 10-year RPV current)

70
Figure 5.12: Envelopes of lateral displacements for concrete coating thickness range
150-160 mm (1-year RPV wave + 10-year RPV current)

Figure 5.13: Envelopes of von Mises stresses for concrete coating thickness range 150-
160 mm (1-year RPV wave + 10-year RPV current)

71
Figure 5.14: Envelopes of lateral displacements for concrete coating thickness range
155-165 mm (10-year RPV wave + 1-year RPV current)

Figure 5.15: Envelopes of von Mises stresses for concrete coating thickness range 155-
165 mm (10-year RPV wave + 1-year RPV current)

72
Figure 5.16: Envelopes of lateral displacements for concrete coating thickness range
170-180 mm (10-year RPV wave + 1-year RPV current)

Figure 5.17: Envelopes of von Mises stresses for concrete coating thickness range 170-
180 mm (10-year RPV wave + 1-year RPV current)

73
5.5.2.3. Numerical Investigation of Pipeline Response

In this numerical investigation, the first 3000-seconds only during the time-domain simulation
of the right end are considered to investigate the pipeline response. Based on the numerical
results of the envelope of lateral displacement for 180mm thickness which is a part of Figure
5.16, the time-history of the lateral and vertical displacements of the pipeline, and the total
water particles velocity are presented in Figure 5.18 to Figure 5.20 respectively. It is worth
mentioning that the load combination of 10-year RPV wave + 1-year RPV current represents
the worst loading condition, as is already emphasized by the high concrete coating thickness
required for the pipeline stabilization.

Figure 5.18 shows that the right end experiences highly oscillatory lateral behavior
displacements at certain time intervals (e.g., at 898, 1417, 2054, and 2452-second). Similar is
shown in Figure 5.19, in which the end node also experiences highly oscillatory vertical
displacements for the same time intervals as in the case of the lateral displacements.

The principal reason behind these lateral and vertical displacements is reverted to the drag,
inertia, and lift forces which are induced under the action of the total velocity of the water
particles at the pipeline level. As shown in Figure 5.20, the total velocity of the water particles
has few peak values at the same time intervals, which increases the hydrodynamic forces
affecting the pipeline, and results in these lateral and vertical displacements.

Figure 5.18: Time history of lateral displacement for the end node for 180mm concrete
coating thickness

74
Figure 5.19: Time history of vertical displacement for the end node for 180mm concrete
coating thickness

Figure 5.20: Time history of total water particle velocity at the pipeline end node for
180mm concrete coating thickness

75
The absence of passive soil resistance can be clearly noticed from the lateral displacement of
the right end. The right end displaces laterally steadily in proportion to the applied
hydrodynamic forces applied if the latter exceed the soil resistance force. The absence of the
passive soil resistance can also be noticed from the rapid change in the pipeline stability with
increasing its stabilizing weight.

5.5.2.4. Random Seed Numbers

The numerical results obtained in the previous sections are based on a single seed number,
i.e., the same phase shift is assigned to all wave components constituting the irregular wave,
for all simulation runs. However, to ensure the stability of the pipeline under the action of
various irregular sea-states, different seed numbers should be used. RP-F109 (DNVGL,
2017a) recommends performing seven simulations at least with randomly chosen seed
numbers, to consider the effect of the different irregular wave patterns on the pipeline
dynamic response.

Figure 5.21 shows the results for lateral displacements corresponding to different irregular
sea-states based on randomly selected seed numbers assigned. The maximum lateral
displacement of the pipeline has occurred at seed numbers 1 and 2, and both values exceed the
allowable limit, which emphasizes the importance of performing seven simulations using
randomly selected seed numbers. However, the von Mises stresses are satisfying the stress
limit criterion for the same different irregular sea-states as shown in Figure 5.22.

It is observed that the induced stresses are increased in the region nearby the left fixed end
because of the boundary condition at the left end of the pipeline. Therefore, the induced
stresses at a location of 75-meter from the left end of the pipeline satisfactorily represent the
von Mises stresses.

5.5.2.5. Comparison with conventional stability methods

The results obtained from the DLSA method ignore the passive soil resistance term, therefore,
a comparison is conducted with the other conventional stability methods provided in the
recommended practice of RP-F109 (DNVGL, 2017a) such as ALSS and GLS methods to
study the effect of ignoring the passive soil resistance term. It is worth highlighting that no
embedment is assumed while using the ALSS method, thus, the passive soil resistance is
equal to zero, while the GLS method considers the passive soil resistance based on the design
curves presented in RP-F109 (DNVGL, 2017a). These design curves are extracted from finite
element dynamic simulations following the recommendations of dynamic lateral stability on
the flat seabed.

Table 5.9 shows the concrete coating thickness required for stabilization of the pipeline based
on each method of stability, it is observed that both DLSA and ALSS methods result in
relatively high concrete coating thickness in comparison with the GLS method because of
ignoring the passive soil resistance term. It is observed also that DLSA method results in less
concrete coating thickness in comparison with ALSS method despite ignoring the passive soil
resistance, which proves that the DLSA method can optimize the on-bottom stability design
and results in less CWC if the actual case is modeled correctly.

76
Figure 5.21: Envelopes of pipeline lateral displacements for the seven simulations

Figure 5.22: Envelopes of von Mises stresses for the seven simulations

77
 Table 5.9: Results of comparison between DLSA, ALSS and GLS methods
Concrete Coating Thickness (mm)
Load Combination
DLSA using Flexcom ALSS GLS
1-year wave + 10-year current 155-160 161 70
10-year wave + 1-year current 170-180 230 79

78
 CHAPTER 6: CONCLUSIONS AND
RECOMMENDATIONS

6.1. Introduction
This thesis is concerned with the design of pipeline on-bottom stability which is one of the
most important design tasks and represents the central part of the subsea pipeline design
process. The on-bottom stability analysis checks the stability of as-laid pipeline against
hydrodynamic loads arising from wave and current loadings. Increasing the CWC is
considered as the most effective stabilization method used to ensure a satisfactory submerged
pipeline weight, and hence can keep the pipeline in equilibrium against the hydrodynamic
loads.

The following section describes the conclusions that are drawn based on the results that are
obtained using different on-bottom stability design methods. These design methods are
applied on a pipeline that has been laid in Gabon. This pipeline was selected because of the
availability of its data, and the difficulty of getting such data from a local case study in Egypt.
In the following sections also, some directions for future work are presented.

6.2. Conclusions
6.2.1. General
The on-bottom stability design parameters such as hydrodynamic loads and pipe-soil
interaction were discussed. The methodologies commonly used for the assessment of pipeline
on-bottom stability were described, and a case study was presented and analyzed using three
different design approaches and summarized as follows:

• Static analysis using absolute lateral static stability method (ALSS).

• Calibrated methods using generalized lateral stability method (GLS) and AGA/PRCI
Level 2.
• Dynamic analysis using Flexcom commercial software package based on dynamic
lateral stability analysis guidelines (DLSA) of RP-F109 (DNVGL, 2017a).

On-bottom stability software is built and designed with a graphical user interface (GUI) using
MATLAB programming language (The MathWorks, 2018), and based on RP-F109 (DNVGL,
2017a) to calculate the CWC required for pipeline stability under different combined loading
conditions. The latter arises from waves and currents during installation (empty pipe
condition) and operation (minimum product density) phases. The developed software was
used to calculate the optimum concrete thickness using ALSS and GLS design methods.

Based on the results obtained using different design methodologies stated in AGA/PRCI
(PRCI, 2008) and RP-F109 (DNVGL, 2017a), it can be concluded that the concrete coating
requirements extremely depend on the pipeline design data such as load combinations from
waves and currents, wave and current characteristics, type of soil, and the fluid-pipe-soil
interaction.

6.2.2. Static Analysis

79
ALSS method was used to perform the static analysis to ensure the absolute stability of the
pipeline under extreme design waves. The concrete thickness provided using this method is
higher than that provided using the calibrated methods which may be considered conservative
in some areas at which the pipeline movements are allowed.

ALSS method is particularly relevant where the environmental loads are current dominated,
and no pipeline movement is required.

6.2.3. Calibrated Methods

GLS method and AGA/PRCI Level 2 software were used as calibrated methods. Both
methods provided a concrete thickness of less than the ALSS method. GLS has two design
criteria (0.5D and 10D allowable displacement) that provide more flexibility and less concrete
thickness.

Increasing the allowable displacement has a significant influence on the CWC required for
stability as well as the density of the product being transported during the operation condition.
For example, “zero” concrete thickness has been calculated during operation condition using
GLS 10D allowable displacement criterion.

Wave hitting angle could affect the stability weight requirements, e.g., a perpendicular wave
hitting angle to the pipeline provided conservative CWC.

Wave characteristics were the major contributor for determining the CWC, particularly in
shallow water depths where the waves were dominated by currents. The CWC has increased
with increasing the significant wave heights.

6.2.4. Dynamic Analysis

The DLSA of a concrete-coated subsea pipeline was performed using one of the available
offshore simulation software, namely Flexcom (Wood Group, 2018). Flexcom was used to
model the pipeline in the realistic sea-state, which is composed of a combination of irregular
waves and steady currents. The soil resistance was modeled using the pure friction term in
between the soil and the pipeline, which was represented by the Coulomb friction model.
However, it did not consider the passive resistance of the soil due to the embedment of the
pipeline, as a result of the present capabilities of the software. The time-domain simulation of
the DLSA of a subsea pipeline introduced a better understanding of its response, and a clear
prediction of the limits of its lateral displacements and the associated stresses. Comparing to
the conventional pipeline stability approaches, this type of analysis is very time-consuming,
and generally requires many details to accurately model the in-situ subsea pipeline, using a
finite element model. A brief discussion of the main conclusions that may be aggregated from
this analysis are:

• If the joint probability distribution of the waves and currents are unavailable, then the
most-worse loading combination of return period values (RPV) of waves and currents
shall be applied as recommended by RP-F109 (DNVGL, 2017a).
• Hydrodynamic loads arising from drag, inertia, and lift forces generated on the
pipeline are increased by increasing the total water particle velocity induced from the
summation of wave-induced particle velocity and current velocity.
• Ignoring the passive soil resistance term decreases the soil resistance dramatically and
pipeline weight becomes the major factor in the assessment of on-bottom stability

80
 analysis, especially when considering the pipe-soil interaction in terms of pure friction
term only. Increasing pipeline weight increases the contact force on the seabed and
hence increases soil resistance to the lateral displacement. Therefore, all mathematical
models, which are based on the pure friction term, mostly result in conservative
weights of stability.
• Ignoring the passive soil resistance term has limited the acceptance criteria for lateral
displacement to 0.5D instead of 10D because of the uncontrolled lateral displacement
and the simplicity of the soil model that is based on the Coulomb friction model.
• In the critical cases in which the lateral displacements exceed their allowable limits,
the stresses induced by such excessive lateral displacements should be thoroughly
examined to ensure pipeline integrity. Excessive stresses (e.g., bending stresses), will
result in collapsing the pipe wall, and consequently result in losing its integrity.
• Random seed numbers assigned to wave components have an impact on confirming
the choice of the concrete coating thickness. Seven simulation runs with seven
randomly chosen seed numbers should be performed as recommended by RP-F109
(DNVGL, 2017a).

6.3. Recommendations
Since pipeline design is constantly evolving and design hypotheses are continuously
changing, there is always a need to improve the design methods and understanding the
parameters that may affect them.

Some suggestions for future studies on this topic are presented as follows:

• Application of the new soil models presented in the updated RP-F114 (DNVGL,
2017b) within the developed MATLAB software, and comparing the updated results
with the present results. These new models replaced the old ones presented in RP-
F109 (DNV, 2007) and applied in this thesis.
• Apply the on-bottom stability design, using the presented methods of stability, on a
local case study in Egypt to study the effect of different environmental loads and soil
properties on the calculated concrete thickness.
• Validation of the calibrated methods’ results with AGA/PRCI Level 1, optimization of
the results using Level 3, and adopting appropriate assumptions corresponding to the
same environmental loads during the analysis.
• Validation the dynamic analysis results obtained from Flexcom software with another
FE software package that includes the passive soil resistance term.
• Taking the advantage of the new feature that has been added by Flexcom developing
team in the recent version “Flexcom 8.13.3” which called “User Solver Variables”.
This feature will allow for changing the soil resistance coefficient during the
simulation to simulate the effect of passive soil resistance. Flexcom software
developers will be working on adding the interface of soil resistance in this new
feature, and it will be available in the future versions. User has to write his own code
including all involved parameters to calculate the value of soil resistance during the
simulation.

81
 REFERENCES

Abankwa, N. O., Johnston, S. J., Scott, M., & Cox, S. J. (2015). Ship motion measurement
using an inertial measurement unit. 2015 IEEE 2nd World Forum on Internet of Things
(WF-IoT), 375–380. https://doi.org/10.1109/WF-IoT.2015.7389083

Allen, D. W., Lammert, W. F., Hale, J. R., & Jacobsen, V. (1989). Submarine Pipeline On-
Bottom Stability: Recent AGA Research. Offshore Technology Conference, 121–132.
https://doi.org/10.4043/6055-MS

Allseas Company. (2021). Pipeline Protection. https://allseas.com/activities/pipelines-and-

subsea/pipeline-protection/

Amlashi, H. (2017). On-Bottom Stability Design of Submarine Pipelines – a Probabilistic

Approach. International Journal of Coastal and Offshore Engineering, 1(3), 29–40.
ijcoe.org/article-1-59-en

Anderson, B., Shim, E., Zeitoun, H. O., & Chin, E. J. (2017). Approach to Lateral Buckling
and On-Bottom Stability Interaction Assessment. Proceedings of the ASME 2013 32nd
International Conference on Ocean, Offshore and Arctic Engineering, 4A.
https://doi.org/10.1115/omae2013-10250

ANSYS Inc. (2021). ANSYS Release 2021 R1 (2021 R1). ANSYS, Inc.
https://www.ansys.com

API. (2018). API Specification 5L, Line Pipe.

Atteris Pty Ltd. (2011). CORUS-3D On-bottom Stabilisation Analysis Software for Submarine
Pipelines.

Bai, Q. (2013). On-bottom Stability Design of Subsea Pipelines - Lecture Notes from
University of Stavanger.

Bai, Q., & Bai, Y. (2014). Subsea Pipeline Design, Analysis, and Installation. Elsevier.

Bai, Y. (2001). Pipelines and Risers (Vol. 3). Elsevier.

Bai, Y., & Bai, Q. (2005). Subsea Pipelines and Risers. Elsevier.

Bai, Y., Xu, W., Ruan, W., & Tang, J. (2014). On-bottom stability of subsea lightweight
pipeline (LWP) on sand soil surface. Ships and Offshore Structures, 12(7), 954–962.
https://doi.org/10.1080/17445302.2014.962249

Bai, Y., & Yu, Z. (2011). Pipeline On-Bottom Stability Analysis Based on FEM Model.
Proceedings of the ASME 2011 30th International Conference on Ocean, Offshore and
Arctic Engineering, 4, 329–333. https://doi.org/10.1115/omae2011-49384

Braestrup, M., Andersen, J. B., Andersen, L. W., Bryndum, M. B., Christensen, C. J., &
Nielsen, N.-J. R. (2009). Design and Installation of Marine Pipelines. Wiley-Blackwell.

82
Brennodden, H., Lieng, J. T., & Sotberg, T. (1989). An Energy-Based Pipe-Soil Interaction
Model. Offshore Technology Conference, 147–158. https://doi.org/10.4043/6057-MS

Bruton, D. A. S., Bolton, M., Carr, M., & White, D. J. (2008). Pipe-Soil Interaction With
Flowlines During Lateral Buckling and Pipeline Walking - The SAFEBUCK JIP.
Offshore Technology Conference.

Dassault Systèmes. (2021). Abaqus FEA (No. 2021X). Dassault Systèmes.

https://www.3ds.com/products-services/simulia/products/abaqus/

DNV. (1981). Rules for Submarine Pipeline Systems.

DNV. (1988). On-Bottom Stability Design of Submarine Pipelines (DNV-RP-E305).

DNV. (2007). On-Bottom Stability Design of Submarine Pipelines (DNV-RP-109).

DNVGL. (2017a). On-Bottom Stability Design of Submarine Pipelines (DNVGL-RP-F109).

DNVGL. (2017b). Pipe-Soil Interaction for Submarine Pipelines (DNVGL-RP-F114).

DNVGL. (2017c). Submarine Pipeline Systems (DNVGL-ST-F101).

Ghebreghiorghis, D. A. (2014). On-Bottom Stability Analysis of Subsea Pipelines According

to DNV-RP-F109. University of Stavanger.

Guo, B., Song, S., Chacko, J., & Ghalambor, A. (2005). Offshore Pipelines. Elsevier.

Hale, J. R., Lammert, W. F., & Jacobsen, V. (1989, May 1). Improved Basis for Static
Stability Analysis and Design of Marine Pipelines. Offshore Technology Conference.
https://doi.org/10.4043/6059-ms

Hasselmann, K., Barnett, T. P., Bouws, E., Carlson, H., Cartwright, D. E., Enke, K., Ewing, J.
A., Gienapp, H., Hasselann, D. E., Kruseman, P., Meerburg, A., Muller, P., Olbers, D. J.,
Richter, K., Sell, W., & Walden, H. (1973). Measurements of Wind-Wave Growth and
Swell Decay during the Joint North Sea Wave Project (JONSWAP).
http://resolver.tudelft.nl/uuid:f204e188-13b9-49d8-a6dc-4fb7c20562fc

Hibbitt, Karlsson, & Sorensen. (1988). ABAQUS, Ver. 5.8 (5.8).

Holthe, K., Sotberg, T., & Chao, J. C. (1987, April 27). An Efficient Computer Model for
Predicting Submarine Pipeline Response to Waves and Current. Offshore Technology
Conference. https://doi.org/10.4043/5502-MS

Jacobsen, V., Bryndum, M. B., & Bonde, C. (1989). Fluid Loads on Pipelines: Sheltered or
Sliding. Offshore Technology Conference, 133–146. https://doi.org/10.4043/6056-ms

Jas, E., O’Brien, D., Fricke, R., Gillen, A., Cheng, L., White, D., & Palmer, A. (2012).
Pipeline stability revisited. The Journal of Pipeline Engineering, 12(4), 259–268.
http://eprints.soton.ac.uk/id/eprint/419914/

Kien, L. K., Ming, L. S., & Badaruddin, M. F. B. (2010). Dynamic On-Bottom Stability of
Shallow Water Pipeline - A Case Study. Proceedings of the ASME 2010 29th

83
 International Conference on Ocean, Offshore and Arctic Engineering, 5, 813–825.
https://doi.org/10.1115/omae2010-20850

Kyriakides, S., & Corona, E. (2007). Mechanics of Offshore Pipelines - Vol. 1: Buckling and
Collapse. Elsevier.

Lambrakos, K. F., Remseth, S., Sotberg, T., & Verley, R. L. P. (1987). Generalized Response
of Marine Pipelines. Offshore Technology Conference, 211–222.
https://doi.org/10.4043/5507-ms

Lammert, W. F., Hale, J. R., & Jacobsen, V. (1989). Dynamic Response of Submarine
Pipelines Exposed to Combined Wave and Current Action. Offshore Technology
Conference, 159–170. https://doi.org/10.4043/6058-ms

McMaster, S. Y., O’Brien, D., Scholtz, D. E., & Ryan, J. R. (2012). On-Bottom Stability
Analysis for a Pipeline on a Mobile Seabed. Proceedings of the ASME 2012 31st
International Conference on Ocean, Offshore and Arctic Engineering, 3, 225–233.
https://doi.org/10.1115/omae2012-83291

Morison, J. R., Johnson, J. W., & O’Brien, M. P. (1953). Experimental Studies of Forces on
Piles. Coastal Engineering Proceedings, 1(4), 340–370.
https://doi.org/10.9753/icce.v4.25

Moshagen, H., & Kjeldsen, S. P. (1980). Fishing Gear Loads and Effects on Submarine
Pipelines. The 12th Annual Offshore Technology Conference, 383–392.
https://doi.org/10.4043/3782-MS

Ose, B. A., Bai, Y., Nystrom, P. R., & Damsleth, P. A. (1999). A Finite-Element Model for
In-Situ Behavior of Offshore Pipelines on Uneven Seabed and Its Application to On-
Bottom Stability. Proceedings of the Ninth International Offshore and Polar
Engineering Conference, 2, 132–140. https://onepetro.org/ISOPEIOPEC/proceedings-
abstract/ISOPE99/All-ISOPE99/ISOPE-I-99-137/24702

Palmer, A. C., & King, R. A. (2006). Subsea Pipeline Engineering. PennWell Corporation.

Pierson, W. J., & Moskowitz, L. (1964). A Proposed Spectral Form for Fully Developed
Wind Seas Based on the Similarity Theory of S. A. Kitaigorodskii. Journal of
Geophysical Research, 69(24), 5181–5190.

PRCI. (2008). Submarine Pipelines On-bottom Stability Volume 1 & 2; Project No. PR-178-
04405 (PRCI PROJECT NO. PR-178-04405).

Robertson, M., Griffiths, T., Viecelli, G., Oldfield, S., Ma, P., Al-Showaiter, A., & Carneiro,
D. (2015). The Influence of Pipeline Bending Stiffness on 3D Dynamic On-Bottom
Stability and Importance for Flexible Flowlines, Cables and Umbilicals. Proceedings of
the ASME 2015 34th International Conference on Ocean, Offshore and Arctic
Engineering, 5A. https://doi.org/10.1115/OMAE2015-41646

Seatools Company. (2021). Subsea Rock Installation. https://www.seatools.com/subsea-

solutions/subsea-rock-installation/

The MathWorks. (2018). MATLAB (No. R2018a). The MathWorks, Inc.

84
Tian, Y., & Cassidy, M. J. (2008). Modeling of Pipe–Soil Interaction and Its Application in
Numerical Simulation. International Journal of Geomechanics, 8(4), 213–229.
https://doi.org/10.1061/(ASCE)1532-3641(2008)8:4(213)

Tian, Y., & Cassidy, M. J. (2010). The challenge of numerically implementing numerous
force-resultant models in the stability analysis of long on-bottom pipelines. Computers
and Geotechnics, 37(1–2), 216–232. https://doi.org/10.1016/j.compgeo.2009.09.004

Tian, Y., & Cassidy, M. J. (2011). Incorporating uplift in the analysis of shallowly embedded
pipelines. Structural Engineering and Mechanics, 40(1), 29–48.
https://doi.org/10.12989/sem.2011.40.1.029

Tian, Y., Cassidy, M. J., & Chang, C. K. (2015). Assessment of Offshore Pipelines using
Dynamic Lateral Stability Analysis. Applied Ocean Research, 50, 47–57.
http://dx.doi.org/10.1016/j.apor.2015.01.001

Tørnes, K., Zeitoun, H. O., Cumming, G., & Willcocks, J. (2009). A Stability Design Rational
- A Review of Present Design Approaches. Proceedings of the ASME 2009 28th
International Conference on Ocean, Offshore and Arctic Engineering, 3, 717–729.
https://doi.org/10.1115/omae2009-79893

Verley, R. L. P., Lambrakos, K. F., & Reed, K. (1987). Prediction of Hydrodynamic Forces
on Seabed Pipelines. Offshore Technology Conference, 171–180.
https://doi.org/10.4043/5503-ms

Verley, R. L. P., & Lund, K. M. (1995). A Soil Resistance Model for Pipelines Placed on
Clay Soils. Proceedings of the ASME 1995 14th International Conference on Ocean,
Offshore and Arctic Engineering, 5, 225–232.
https://doi.org/https://www.osti.gov/biblio/205471

Verley, R. L. P., & Sotberg, T. (1994). A Soil Resistance Model for Pipelines Placed on
Sandy Soils. Journal of Offshore Mechanics and Arctic Engineering, 116(3), 145–153.
https://doi.org/10.1115/1.2920143

Wagner, D. A., Murff, J. D., Brennodden, H., & Sveggen, O. (1987). Pipe-Soil Interaction
Model. Offshore Technology Conference, 181–190.
https://doi.org/10.1061/(ASCE)0733-950X(1989)115:2(205)

Wood Group. (2018). Flexcom (8.10.2). John Wood Group PLC.

https://www.woodplc.com/capabilities/digital-and-technology/software,-applications-
and-analytics/flexcom

Yang, H., & Wang, A. (2013). Dynamic Stability Analysis of Pipeline Based on Reliability
Using Surrogate Model. Journal of Marine Engineering and Technology, 12(2), 75–84.
https://doi.org/10.1080/20464177.2013.11020279

Youssef, B. S. (2011). The Integrated Stability Analysis of Offshore Pipelines (Issue

December). The University of Western Australia.

Youssef, B. S., & O’Brien, D. (2017, May). On-Bottom Stability Analysis of Submarine
Pipelines, Umbilicals and Cables Using 3D Dynamic Modelling. Offshore Technology
Conference. https://doi.org/10.4043/27727-ms

85
Zeetech Engineering B.V. (2013). On-Bottom Stability Analysis.

Zeitoun, H. O., Tørnes, K., Cumming, G., & Brankovic, M. (2008). Pipeline Stability - State
of the Art. Proceedings of the ASME 2008 27th International Conference on Offshore
Mechanics and Arctic Engineering, 3, 213–228. https://doi.org/10.1115/omae2008-
57284

Zeitoun, H. O., Tørnes, K., Li, J., Wong, S., Brevet, R., & Willcocks, J. (2009). Advanced
Dynamic Stability Analysis. Proceedings of the ASME 2009 28th International
Conference on Ocean, Offshore and Arctic Engineering, 3, 661–673.
https://doi.org/10.1115/omae2009-79778

86
APPENDIX A RESULTS OF ON-BOTTOM STABILITY
ANALYSIS

A.1 Static Analysis Results

Table A.1: Full results of ALSS method
Depth Installation Phase (mm) Operation Phase (mm)
(m) 1C10W 10C1W Max 10C100W 100C10W Max
6.1 152 117 152 136 89 136
8 118 95 118 86 59 86
10 99 82 99 60 43 60
15 76 42 76 33 24 33
20 132 102 132 108 70 108
25 110 89 110 75 50 75

A.2 Calibrated Methods Results

A.2.1 Generalized Lateral Stability Method (0.5D Displacement)
Table A.2: Full results of GLS method (0.5D displacement)
Depth Installation Phase (mm) Operation Phase (mm)
(m) 1C10W 10C1W Max 10C100W 100C10W Max
6.1 63 69 69 24 25 25
8 57 67 67 15 23 23
10 57 63 63 12 19 19
15 57 54 57 10 11 11
20 60 68 68 19 23 23
25 56 65 65 14 20 20

A.2.2 Generalized Lateral Stability Method (10D Displacement)

Table A.3: Full results of GLS method (10D displacement)
Depth Installation Phase (mm) Operation Phase (mm)
(m) 1C10W 10C1W Max 10C100W 100C10W Max
6.1 43 42 43 1 0 1
8 42 42 42 0 0 0
10 42 42 42 0 0 0
15 42 42 42 0 0 0
20 42 43 43 0 0 0
25 42 42 42 0 0 0

87
A.2.3 AGA/PRCI Level 2 Software
Table A.4: Full results of AGA/PRCI Level 2
Depth Installation Phase (mm) Operation Phase (mm)
(m) 1C10W 10C1W Max 10C100W 100C10W Max
6.1 86 67 86 41 31 41
8 70 58 70 27 22 27
10 61 53 61 19 13 19
15 51 49 51 7 3 7
20 77 62 77 34 21 34
25 68 57 68 21 17 21

A.2.4 Sensitivity Analysis Results between AGA/PRCI Level 2 and GLS

0.5D
Table A.5: Full results of GLS 0.5D for the sensitivity analysis
Depth Installation Phase
(m) 1C1W 1C10W 10C1W 10C10W
6.1 100 133 108 138
8 85 106 91 113
10 75 91 81 98
15 61 71 68 77
20 57 62 68 68
25 58 56 65 67
30 56 57 61 67
35 53 57 56 63
40 50 55 52 60

Table A.6: Full results of AGA/PRCI Level 2 for the sensitivity analysis
Depth Installation Phase
(m) 1C1W 1C10W 10C1W 10C10W
6.1 151 183 150 185
8 121 154 120 155
10 107 131 107 130
15 80 94 81 93
20 63 79 65 81
25 55 68 57 70
30 50 59 52 60
35 46 53 49 55
40 45 49 47 52

88
 APPENDIX B FLEXCOM INPUT FILES

B.1 Initial Static Analysis

$PREPROCESSOR
*PARAMETERS
WaterDepth, <20m>
Do, <0.559m>
WT, <9.5mm>
Di, =[Do-2*WT]
t_corr, <3.2mm>
t_conc, <160mm>
Db, =[Do+2*t_corr+2*t_conc]
rho_conc, 3040
rho_corr, 950
rho_sea, 1030
rho_cont, 850
Lm, 0.05

$MODEL
*OCEAN
=[WaterDepth], =[rho_sea], 9.81

*LINES
LINE=Pipeline, 250
START=Pipeline Start, 0.5, 0, 0
END=Pipeline End, 0.5, 250, 0
5, , , RATIO=FINE

*GEOMETRIC SETS
OPTION=RIGID
SET=Pipeline
<207GPa>, <80GPa>, =[Do], =[Di], 7850.0, , , , =[Db], =[Db], =[Db]

*HYDRODYNAMIC SETS
DIAMETER=BUOYANCY
SET=All
0.7, 0.0, 3.29, 0.0, , 0.9

*POISSON
SET=Pipeline
0.3

*COATINGS
SET=Pipeline
TYPE=EXTERNAL
NAME=Corrosion Coating, THICKNESS==[t_corr], DENSITY==[rho_corr]
SET=Pipeline
TYPE=EXTERNAL
NAME=Concrete Coating, THICKNESS==[t_conc], DENSITY==[rho_conc]

*SEABED PROPERTIES
TYPE=ELASTIC

89
SET=All
STIFFNESS=300000
FRICTION=0.6, 0.6, =[Do*Lm], =[Do*Lm], 0.0
LATERAL=0.0
SUCTION=0.0

*LINE LOCATIONS
LINE=Pipeline
LABEL=Pipeline Mid Node, 125

$LOAD CASE
*NAME
OBS - Installation Phase; Initial Static Analysis

*ANALYSIS TYPE
TYPE=STATIC

*BOUNDARY
TYPE=CONSTANT
{Pipeline Start}, 2, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline Start}, 3, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline Start}, 4, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline Start}, 5, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline Start}, 6, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline End}, 2, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline End}, 3, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline End}, 4, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline End}, 5, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline End}, 6, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
GEN={Pipeline Start}, {Pipeline End}, 5, 1, 0.0, FIXATION=ABSOLUTE

*TIME
STEP=FIXED
0, 1

B.2 Quasi-Static Analysis

$LOAD CASE
*NAME
OBS - Installation Phase; Quasi-Static Analysis

*ANALYSIS TYPE
TYPE=QUASI-STATIC

90
*BOUNDARY
TYPE=CONSTANT
{Pipeline Start}, 2, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline Start}, 3, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline Start}, 4, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline Start}, 5, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline Start}, 6, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline End}, 4, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline End}, 5, 0.0, FIXATION=ABSOLUTE
TYPE=CONSTANT
{Pipeline End}, 6, 0.0, FIXATION=ABSOLUTE

*TIME
STEP=VARIABLE
0, 500, 5
0.01, 0.001, 5, 0.055

*RESTART
LAST="Inst-Initial-Static", TYPE=DYNAMIC/STATIC

*DATABASE
TIME=ALL
*DAMPING
TYPE=STANDARD
SET=All
0.0, 0.0

B.3 Static Analysis for Current Loadings

$LOAD CASE
*RESTART
LAST="Inst-Quasi-Static", TYPE=DYNAMIC/STATIC

*ANALYSIS TYPE
TYPE=STATIC

*CURRENT
TYPE=UNIFORM
0.203, 90

*NAME
OBS - Installation Phase; Current Static Analysis

*TIME
STEP=FIXED
0, 1

*BOUNDARY

91
TYPE=CONSTANT
{Pipeline Start}, 1, 0.0, FIXATION=RELATIVE
TYPE=CONSTANT
{Pipeline Start}, 2, 0.0, FIXATION=RELATIVE
TYPE=CONSTANT
{Pipeline Start}, 3, 0.0, FIXATION=RELATIVE
TYPE=CONSTANT
{Pipeline Start}, 4, 0.0, FIXATION=RELATIVE
TYPE=CONSTANT
{Pipeline Start}, 5, 0.0, FIXATION=RELATIVE
TYPE=CONSTANT
{Pipeline Start}, 6, 0.0, FIXATION=RELATIVE
TYPE=CONSTANT
{Pipeline End}, 4, 0.0, FIXATION=RELATIVE
TYPE=CONSTANT
{Pipeline End}, 5, 0.0, FIXATION=RELATIVE
TYPE=CONSTANT
{Pipeline End}, 6, 0.0, FIXATION=RELATIVE

*DATABASE
TIME=ALL

*DAMPING
TYPE=STANDARD
SET=All
0.0, 0.0

B.4 Dynamic Analysis for Wave Loadings

$PREPROCESSOR
*PARAMETERS
T, <13.9s>
T_Storm, 10800
TimeStep, 0.01
RecTimeStep, 0.5
RampTime, 100

$LOAD CASE
*NAME
OBS - Installation Phase; Dynamic Analysis

*ANALYSIS TYPE
TYPE=DYNAMIC, SOLUTION=NONLINEAR

*RESTART
LAST="Inst-Current-Static", TYPE=DYNAMIC/STATIC

*WAVE-JONSWAP
FREQUENCY=GP, SPEC=HSTP
3.02, =[T], 0.5, 0.02, 1, 90, 2

*RAMP
TYPE=LINEAR

92
*TIME
STEP=VARIABLE
0, =[T_Storm], =[RampTime]
=[TimeStep], 0.005, 0.2, 0.055

*WAVE-GENERAL
SEED=573259391

*DAMPING
TYPE=STANDARD
SET=All
0.0, 0.0

$DATABASE POSTPROCESSING
*STATISTICS
TYPE=KINEMATIC, SET=All
ENVELOPE, 3, PARA=M, 0, 1.0, UNITS=m
TITLE=Lateral Displacement Envelope
TYPE=KINEMATIC, SET=All
ENVELOPE, 1, PARA=M, 0, 1.0, UNITS=mm
TITLE=Vertical Displacement Envelope
TYPE=FORCE, SET=All
ENVELOPE, 13, 0, 1.0, , UNITS=MPa
TITLE=Von Mises Stress Envelope

*TIMETRACE
TYPE=KINEMATIC
START=0, END==[T_Storm]
{Pipeline Mid Node}, 3, 1.0, PARA=M, UNITS=m
TITLE=Mid Node Lateral Displacement
TYPE=KINEMATIC
START=0, END==[T_Storm]
{Pipeline End}, 3, 1.0, PARA=M, UNITS=m
TITLE=End Node Lateral Displacement
TYPE=KINEMATIC
START=0, END==[T_Storm]
{Pipeline Mid Node}, 1, 1.0, PARA=M, UNITS=mm
TITLE=Mid Node Vertical Displacement
TYPE=KINEMATIC
START=0, END==[T_Storm]
{Pipeline End}, 1, 1.0, PARA=M, UNITS=mm
TITLE=End Node Vertical Displacement
TYPE=KINEMATIC
START=0, END==[T_Storm]
{Pipeline Mid Node}, 3, 1.0, PARA=V, UNITS=m/s
TITLE=Mid Node Lateral Velocity

93
‫وجدير بالذكر أن اتةاه األموا والتيارات المائية قد أعر بشاكل كبير علي تحديد الوزن المطلوب الساتقرار‬
‫فط االناابياب‪ ،‬إذ كلماا إقترب إتةااهماا من التعااماد علي فط االناابياب‪ ،‬كلماا زادت االحماال الهيادرودينااميكياة‬
‫التي تتطلاب م قااوماة اكبر عن طريق زياادة الوزن الاذي بادوره يزياد من مقااوماة الترباة لازاحاة الةاانبياة‬
‫الناتةة من االحمال الهيدروديناميكية‪ .‬كما تأعر أيضااااا اسااااتقرار فط األنابيب بعمق المياه‪ ،‬حيث أظهرت‬
‫النتائل ارتفاع ملحوظ ذي سام طبقة غال ال رساانة المطلوب لتحقيق اساتقرارها ذي المياه الضاحلة‪ ،‬كما‬
‫أظهرت أيضااا تناقص ذي ساام ال رسااانة المطلوب لتحقيق إسااتقرار فط األنابيب كلما زاد عمق المياه‬
‫بسبب تالشي تأعير األموا علي فط األنابيب‪.‬‬
‫علي ىااااعياد فر‪ ،‬ذاإن التحليال الادينااميكي قاد أمادناا بفهم اكبر لرد ذعال فط األناابياب وإزاحتاه عن طريق‬
‫تمثيال محااكااة لبيئاة ممااعلاة للبيئاة الحقيقياة المزمع تواجاد فط األناابياب ذيهاا‪ .‬إال إن هاذا النوع من التحليال‬
‫يسااتلزم وقت أطول مقارنة باألساااليب التقليدية التي تم ذكرها سااابقا‪ ،‬وأيضااا يتطلب تواذر معلومات دقيقة‬
‫لتمثيل البيئة الحقيقية الزمع وضع وانزال فط االنابيب ذيها‪.‬‬
‫أظهرت النتائل بإست دام التحليل الديناميكي انه كلما زادت سرعة جزيئات المياه الناتةة من حركة األموا‬
‫والتياارات الماائياة عناد مسااااتوي فط األناابياب‪ ،‬زادت معهاا القوي الادينااميكياة المؤعرة علي فط األناابياب‪.‬‬
‫أظهرت أيضاااا النتائل ان وزن فط األنابيب يمثل العامل ذو التأعير األكبر ذي تقييم اساااتقرار فط األنابيب‬
‫البحرياة علي القااع‪ ،‬فااىااااة عناد األفاذ ذي االعتباار تاأعير معاامال االحتكااك بين فط االناابياب والترباة ذقط‪،‬‬
‫مع اهماال قوي المقااوماة المةهولاة من الترباة جراء ان ماار فط األناابياب ذيهاا‪ .‬ذي بعض الحااالت الحرجاة‬
‫التي نتل عنها زيادة اإلزاحة الةانبية ل ط األنابيب عن الحد المساااامول به‪ ،‬ذإنه يةب التأكد ان اإلجهادات‬
‫النااتةاة عن تلا اإلزاحاة ماا زالات اقال من الحادود القصااااوي المساااامول بهاا‪ ،‬تةنباا ً لتحميال فط األناابياب‬
‫إجهادات مفرطة بما قد يسبب إنهيار جدرانها‪.‬‬
‫للتحقق (بما ال يدع مةال للشا ) من عبات اساتقرار فط األنابيب علي القاع كما ذمكر ذي الممارساة الموىي‬
‫بها "‪ ،"DNVGL-RP-F109‬ذالبد من افتبار اساااتقرار فط األنابيب من فالل محاكيات لظرو بيئية‬
‫م تلفة من حاالت البحر العشاااوائية التي قد تتسااابب ذي ت يير االحمال الهيدروديناميكية وزيادة االزاحات‬
‫الةانبية‪.‬‬
‫وأفيرا‪ ،‬تم إعاداد دراسااااة مقاارناة بين النتاائل المساااات لصااااة من التحليال الادينااميكي مع كلتاا الطريقتين‬
‫المذكورتان ذي الممارسااة الموىااي بها "‪ "DNVGL-RP-F109‬أال وهما الطريقة المعايرة التي تساام‬
‫بحد ضااائيل من اإلزاحة (نصاااا قطر األنبوب) و طريقة الثبات المطلق‪ ،‬وتبين من الدراساااة مدي أهمية‬
‫األفذ ذي االعتباار مقااومة التربة الساااالبياة ذي زيادة مقااومة التربة الكلياة لألحماال الهيادرودينااميكياة الواقعة‬
‫عليهاا والتي من شااااأنهاا العمال علي عباات واسااااتقرار فط األنابياب وبالتاالي تقليال كالً من الوزن المطلوب‬
‫لالسااتقرار وساام ال طاء ال رساااني‪ .‬كما تبين ايضااا ً من الدراسااة أهمية اجراء التحليل الديناميكي حيث‬
‫أظهرت النتاائل ان فاا ملحوظ ذي الوزن المطلوب لالسااااتقرار باإساااات ادام التحليال الادينااميكي وهاذا ماا‬
‫يتطلب بناء نموذ دقيق لتمثيل الحالة الواقعية ل ط األنابيب ذي الظرو البيئية الفعلية المحيطة به‪.‬‬

‫ز‬
 ‫الملخص‬

‫ت معزي أهمية إجراء تحليل إساتقرار فطوط األنابيب البحرية علي القاع إلي ضامان اساتقرارها أعناء ظرو‬
‫مرحلتي التركيب والتشااا يل من فالل تحديد الوزن الالزم لزيادة مقاومة التربة لألحمال الهيدروديناميكية‬
‫دون زيادة إجهادات فطوط األنابيب عن الحد المساامول بها‪ .‬وذي هذا الصاادد‪ ،‬يتم إجراء تحليل اسااتقرار‬
‫فطوط األنابيب البحرية علي القاع بدراساة القوي المؤعرة عليها لضامان اساتقرارها الرأساي والةانبي‪ .‬ذيما‬
‫يتعلق باالسااتقرار الةانبي ل طوط األنابيب البحرية علي القاع‪ ،‬تمساات دم بيانات األرىاااد الةوية ال اىااة‬
‫بموجات العواىاااا والتيارات المائية لتوقع القيم التصاااميمية لسااارعة جزيئات المياه وتساااارعها ذي قاع‬
‫البحر‪ .‬أما ذيما يتعلق باإلساتقرار الرأساي ل طوط األنابيب البحرية علي القاع‪ ،‬ذيتم التأكد من تحقيق الوزن‬
‫الكاذي للت لب علي تأعير الطفو الناجم من إزاحة فطوط األنابيب لمياة البحر‪.‬‬
‫ذي هذه الرساااالة‪ ،‬تم مناقشاااة العوامل التي تؤعر على اساااتقرار فطوط األنابيب البحرية‪ ،‬وشااارل الطرق‬
‫التقليدية والمتقدمة لتحليل اسااااتقرارها على القاع‪ ،‬وأفيرا ً تم إجراء دراسااااة حالة بإساااات دام عالعة طرق‬
‫تصميمية‪ :‬طريقة الثبات المطلق‪ ،‬طريقة الطرق ال ممعايرة ‪ ،‬وطريقة التحليل الديناميكي‪.‬‬
‫لتصااااميم إسااااتقرار فطوط االنابيب البحرية علي القاع باساااات دام طريقة الثبات المطلق وطريقة الطرق‬
‫ال ممعاايرة‪ ،‬تم بنااء و تطوير برناامل باإساااات ادام ل اة البرمةاة "‪ "MATLAB‬بماا يتواذق مع المماارسااااة‬
‫الموىااااى بهاا "‪ ."DNVGL-RP-F109‬تم اساااات ادام طريقاة معاايرة أفر المادرجاة ذي برناامل‬
‫"‪ "AGA/PRCI Level 2‬الااذي تاام تااطااويااره باامااعاارذااة " ‪Pipeline Research Council‬‬
‫)‪ "International (PRCI‬والاااتاااي كاااانااات ذاااي الساااااااباااق جاااز ًءا مااان " ‪American Gas‬‬
‫(‪ ."Association )AGA‬تم ايضااااا ً إجراء دراساااااة مقاارناة بين كلتاا الطريقتين المعيااريتين اللتين تم‬
‫است دامهما من "‪ "DNVGL‬و "‪ "AGA/PRCI‬باإلضاذة إلى إجراء تحليل حساسية لعدة اعماق‪.‬‬
‫تم إجراء تحليل االساتقرار الديناميكي ل ط أنبيب بحري علي القاع باسات دام برنامل "‪( "Flexcom‬حزمة‬
‫برامل العناىر المحدودة من‪ )John Wood Group PLC‬بواسطة محاكاة زمنية مع األفذ ذي اإلعتبار‬
‫حاالاة البحر ذي ظال ظرو تحميال م تلفاة من األموا والتياارات الماائياة ومقااوماة االحتكااك مع الترباة و‬
‫الساالوك غير ال طي ل ط األنابيب‪ .‬ومع ذحص الحد األقصااى لازاحة الةانبية ل ط األنابيب واإلجهادات‬
‫المؤعرة عليها والناتةة عن افتال سم طبقة ال ال ال رساني‪ .‬وأفيرا‪ ،‬تم إجراء سبعة عمليات محاكاة‬
‫زمنية م تلفة للتحقق من استقرار فط األنابيب تحت تأعير حاالت بحر عشوائية م تلفة‪.‬‬
‫أظهرت دراسااة الحالة ان متطلبات وزن طبقة غال ال رسااانة الالزمة لتحقيق اسااتقرار فط األنابيب ذي‬
‫القاع يعتمد علي األحمال التي تتشااكل من تةمع األموا والتيارات المائية‪ ،‬وعلي فصااائص تل األموا‬
‫والتيارات المائية وأيضااا علي نوع التربة الموضااوع عليها فط االنابيب‪ .‬كما أظهرت النتائل ان إساات دام‬
‫ساااام غطااء فرساااااني كبير‪ ،‬حيث تعتمد هذه الطريقة علي تحقيق‬ ‫طريقاة الثباات المطلق دائماا ما تعطي م‬
‫االسااتقرار ضااد اقوي االحمال الهيدروديناميكية ال ممحتملة من األموا والتيارات المائية‪ ،‬دون الساامال ألي‬
‫ازاحة جانبية ل ط االنابيب‪.‬‬
‫أما بالنسابة للطرق ال ممعايرة المسات دمتين ذي دراساة الحالة‪ ،‬وجد أنهما تعطيان سام فرساانة اقل من ذل‬
‫المعطي من طريقة الثبات المطلق‪ .‬حيث منتحت الطرق ال ممعايرة المشااار إليها ذي الممارسااة الموىااي بها‬
‫"‪ "DNVGL-RP-F109‬سماحيات إزاحة جانبية ل طوط االنابيب تتراول ما بين نصا القطر ال ارجي‬
‫لألنبوب إلي عشارة أمثاله كحد اقصاي؛ ذي حين منتحت الطريقة ال ممعايرة من "‪"AGA/PRCI Level 2‬‬
‫تقليص سم ال رسانة المطلوب لتحقيق االستقرار‪.‬‬

‫و‬
 ‫المستخلص‬

‫فطوط األنابيب هي الوساااايلة الرئيسااااية لنقل منتةات البترول‪ .‬يتم نقل المنتةات من البئر أو ال زان إلى‬
‫مناطق تةميع تحت البحر أو الشاااطأ أو منصااات اإلنتا للمساات دمين النهائيين‪ .‬للوىااول إلى التصااميم‬
‫األمثل لنظام فطوط األنابيب البحرية‪ ،‬هناك حاجة إلى العديد من الدراسااااات الهندسااااية المتطورة‪ ،‬والتي‬
‫تشامل إفتيار حةم فطوط األنابيب‪ ،‬وافتيار المواد المصانعه لها‪ ،‬وتحديد مساار فطوط األنابيب‪ ،‬وتصاميم‬
‫االستقرار علي القاع‪ ،‬إلخ‪.‬‬
‫يعد تصاااميم االساااتقرار علي القاع أحد أهم المراحل األسااااساااية المكونة لعملية تصاااميم فطوط األنابيب‬
‫البحرياة‪ .‬حياث يممثال اسااااتقراهاا الرأسااااي والةاانبي علي القااع أهم ضاااارورياات مقااوماة األحماال‬
‫الهيادرودينااميكياة النااجماة من األموا والتياارات الماائياه عليهاا ومقااوماة الترباة المالىااااقاة لهاا‪ .‬ذي أغلاب‬
‫األحيان‪ ،‬تضااا طبقة فرسااانية فارجية كوزن إضاااذي إلى فط األنابيب بهد الوىااول إلي حل ذعال‬
‫(إقتصاديا ً وزمنياً) لضمان استقراره‪.‬‬
‫ذي هذه الرساااالة‪ ،‬تم مناقشاااة العوامل التي تؤعر على اساااتقرار فطوط األنابيب البحرية‪ ،‬وشااارل الطرق‬
‫التقليادياة والمتقادماة لتحليال اسااااتقرارهاا على القااع‪ ،‬وأفيرا إجراء دراسااااة حاالاة بااساااات ادام عالعاة طرق‬
‫تصااااميمياة‪ :‬طريقاة الثباات المطلق‪ ،‬طريقاة الطرق ال ممعاايرة‪ ،‬وطريقة التحليل الديناميكي‪ .‬تتناول الرسااااالة‬
‫أيضاااا شااارل البرنامل الذي تم بناءه وتطويره بإسااات دام ل ة البرمةة "‪ "MATLAB‬لحسااااب الوزن‬
‫المطلوب الساتقرار فطوط األنابيب البحرية علي القاع وأيضاا حسااب سام ال رساانة الالزم لتحقيق ذل‬
‫الوزن اعتمادا علي طريقة الثبات المطلق و كلتا الطرق ال ممعايرة اللتان تم ذكرهما ذي الممارسااة الموىااي‬
‫بها "‪ "DNVGL-RP-F109‬اىاادار ‪ .2017‬كما تم اساات دام طريقة معايرة افري المدرجة ذي برنامل‬
‫طور بواسااطة " ‪Pipeline Research Council International‬‬ ‫”‪ “AGA/PRCI Level 2‬الذي ّ‬
‫)‪ ،"(PRCI‬باإلضااذة إلي ذل ‪ ،‬تم تتويل تطبيق الطرق الثالعة بدراساة مقارنة ذيما بينها إليضاال تأعير كل‬
‫طريقة ومت يراتها علي اسااتقرار فطوط األنابيب‪ .‬تم أيضااا اساات دام برنامل "‪( "Flexcom‬حزمة برامل‬
‫العناىاار المحدودة من ‪ )John Wood Group PLC‬لتحليل اسااتقرار فط األنابيب ذي القاع بواسااطة‬
‫محااكااة لنطااق زمني مع األفاذ بعين االعتباار حاالاة البحر ذي ظال مةموعاات تحميال م تلفاة من األموا‬
‫والتيارات المائية ومقاومة االحتكاك مع التربة والسلوك ال ير فطي ل ط األنابيب‪.‬‬

‫ه‬
 ‫إقــــــرار‬

‫أقر أنه ال يوجد أي جزء من هذه الرسااالة قد ساابق تقديمه لنيل درجة أفري ذي هذه الةامعة أو أية جامعة‬
‫أو مؤساااساااة تعليمية أفري بمصااار او بال ار ‪ ،‬وأنها تتواذق مع الميثاق األفالقي للبحث العلمي بةامعة‬
‫اإلسكندرية وعلي األفص األمانة العلمية‪.‬‬

‫االساام‪ :‬محمد عمرو محمد عبدالسالم‬

‫التوقيع‪:‬‬

‫د‬
 ‫كلية الهندسة‬
‫قسم الهندسة البحرية وعمارة السفن‬

‫تصميم استقرار خطوط األنابيب البحرية علي القاع‬

‫في المياه الضحلة‬

‫رسالة علمية مقدمة ضمن متطلبات الحصول علي درجة ماجستير العلوم‬

‫في‬
‫الهندسة البحرية وعمارة السفن‬

‫مقدمة من‬
‫محمد عمرو محمد عبدالسالم‬

‫بكالوريوس الهندسة البحرية وعمارة السفن‬

‫كلية الهندسة‪ ،‬جامعة اإلسكندرية‪2013 ،‬‬

‫‪2022‬‬