FP914

SSL Engineering Design Manual for Subsea Applications Rev. 1.
1 March 2003
Table of Contents
1. INTRODUCTION 1
2. ADVANTAGES IN THE USE OF SSL AS A SUBSEA PIPELINE 1
3. SSL PIPE AND FITTING CONFIGURATIONS 2

3.1 Pipe Wall 2
3.1.1 Description 2
3.1.2 Pressure Rating 3
3.1.3 Temperature De-rating Factors 4
3.2 Coil-Lock Joining Systems and Adapters 4

3.2.1 Coil-Lock 4
3.2.2 Threaded Alloy or Coated Steel Coil-Lock Adapters 6
3.3 Pipe Configurations 7

3.3.1 Coil-Lock® Bell by Coil-Lock® Spigot SSL Pipe 7
3.3.2 Coil-Lock® Spigot by Coil-Lock® Spigot SSL Pipe 7
3.3.3 Repair Pipe 8
3.4 Fitting Configurations 9
4. DESIGN OF SUBSEA PIPELINES 11

4.1 Introduction 11
4.2 Directional Changes for Subsea Pipelines 11

4.2.1 Field Bends Using Elbows or Installed Bend Radius in Subsea Pipelines 11
4.2.2 Snake-Lay for Thermal Expansion Using “Flexure of the Pipe” 11
4.3 Shore Approach for SSL Subsea Pipelines 13

4.3.1 Transitions from Shore Approach to Above Ground Pipelines 13
4.3.2 Thrust Reactions 14
4.3.2.1 Reaction Force from Internal Pressure Induced Thrust 14
4.3.2.2 Reaction Force from Internal Pressure Induced Thrust and Thermal Expansion 15
4.4 Risers 17
4.5 Bottom Stability 18

4.5.1 Weight Coating and Concrete Collars 18
4.5.2 Weight Blankets 19
4.5.3 Poured in Place Anchors 20
4.5.4 Mechanical Anchors 20
4.5.5 Chain Weights 22
5. INSTALLING SUBSEA SSL PIPELINES 22

5.1 Overview 22
i ©Ameron 2003 - FP914 03/03

SSL Engineering Design Manual for Subsea Applications Rev. 1.1 March 2003
5.2 S-Lay 22
5.2.1 Comparison of Installation Techniques for SSL and Carbon Steel pipe Using S-Lay 23
5.2.2 Allowable Axial Tension and Bend Radius for SSL and Carbon Steel pipe Using S-Lay 24
5.2.3 Advantages and Disadvantages of SSL vs. Carbon Steel pipe Using S-Lay 27
5.2.4 Discussion of the Use of SSL versus Carbon Steel Pipe When Using S-Lay 28
5.3 J-Lay 30
5.4 Off Bottom Tow 31
5.5 Surface Tow 31
5.6 Mid Depth Tow 32
5.7 Relative Installation Rates and Costs of SSL versus Carbon Steel 33
6. SSL SUBSEA PIPELINE ANALYSIS USING FEA PROGRAMS 34

6.1 Introduction 34
6.2 Definition of Partial FActors f1, f2, f3, A1, A2 and A3 35
6.3 Equivalent Allowable Design Stress Envelope 39

6.3.1 Introduction 39
6.3.2 Design Envelope for the GRE Component of SSL 39
6.3.3 Equivalent Design Envelope for the Overall SSL Pipe Wall 39
6.4 General Outline of the FEA Analysis Method for an SSL Pipeline 44
6.5 Limitations in Existing FEA Analysis Software 45
6.6 Equivalent Elastic Properties of SSL 46

6.6.1 Equivalent Modulus, Poisson’s Ratio, Thermal Coefficient of Expansion and Density for the
SSL Pipe Wall 46
6.6.2 Equivalent “EI” Values for the SSL Pipe Wall 50
6.6.3 Equations for Equivalent “EI” for the SSL Pipe Wall in the Hoop Direction 50
7. SSL SUBSEA PIPELINE ANALYSIS USING AMERON PROPRIETARY

MATHCAD® SOFTWARE 53
7.1 Introduction 53
7.2 FLow Diagrams for Structural Design 53
7.3 Inputs Required for the MathCAD Analysis 56
7.4 Analytical Approach and Equations Utilized in the MathCAD Analysis 60

7.4.1 Pressure Containment 60
7.4.1.1 GRE Stress per ISO 14692 60
7.4.1.2 Average Pipe Wall Stress Containment per DNV-OS-F101 62
7.4.2 System Collapse 64
7.4.2.1 System Collapse Check for Lateral Perssure Only, per DNV Recommended Equations 64
7.4.2.2 System Collapse Check 65
ii ©Ameron 2003 - FP914 03/03

7.4.2.3 System Collapse Check for Lateral Perssure Only, per NASA SP-8007 Recommended
Equations 66
7.4.2.4 Buckling Arrestors at the Coil-Lock Joint 68
7.4.3 Pipe Soil Interaction 68
7.4.4 Stress from Internal Pressure Plus Unsupported Span 71
7.4.5 Vortex Induced Vibration at the Unsupported Span 73
7.4.6 Stress from Internal Pressure Plus Installed Bend Radius in the Horizontal and Vertical Plane 76
7.4.7 Stress from Internal Pressure Plus Unsupported Span Plus Installed Bend Radius in the Vertical
Plane – Section 12 of the MathCAD Workbook 78
7.4.8 Analysis of Installation Loads Using the J-Lay Method 79
7.4.9 Design Summary 87
8. OTHER CONSIDERATIONS IN SSL PIPELINE DESIGN 92

8.1 SSL PIPELINE HYDRAULIC CALCULATIONS 92
8.1.1 Definitions 92
8.1.2 Design Methods and Assumptions 92
8.1.3 Design Equations 93
8.1.4 Determination of an Analytical Equation for Kinematic Viscosity 95
8.2 INTEGRITY Monitoring 101

8.2.1 Verification of Electrical Continuity. 101
8.2.2 Monitoring a Subsea SSL Pipe 102
8.3 Repairs 104

8.3.1 Temporary Repair of Minor Local Damage 104
8.3.2 Repair by Pipe Replacement 105
8.4 Erosion Rates with Liquids and Abrasive Fluids 105
8.5 Transient Pressure (water hammer) 106
9. OTHER REFERENCE DOCUMENTS 108

9.1 Product Literature 108
9.2 Test reports 108
9.3 Analysis and technical reports 108
9.4 Footnotes 108
10. APPENDIX 109

10.1 Equivalent Allowable Design Envelopes 109
10.2 Allowable Axial Tension and Bend Radius Combinations for S- and J-Lay 127
10.3 Guidelines for the UKOOA Analysis of a Pipeline as Implemented in CAESAR II® 141
iii ©Ameron 2003 - FP914 03/03

Table of Figures
Figure 1 - SSL Pipe Wall 2
Figure 2 - Standard Coil-Lock® Joint for SSL Pipe 5
Figure 3 - Seal Designs Available for the Coil-Lock® Joint 5
Figure 4 - Coil-Lock Male Threaded Steel End Adapter 6
Figure 5 - Coil-Lock Female Threaded Steel End Adapter 6
Figure 6 - Coil-Lock® Bell by Coil-Lock® Spigot SSL Pipe 7
Figure 7 - Coil-Lock® Spigot by Coil-Lock® Spigot SSL Pipe 7
Figure 8 - Repair Pipe 8
Figure 9 - Repair Spool 8
Figure 10 - Assembly of Flange by Flange Steel Elbow with Two Flange by Coil-Lock
Threaded Adapters 9
Figure 11 - Assembly of Flange by Coil-Lock Steel Elbow with One Flange by Coil-Lock
Threaded Adapter 9
Figure 12 - Assembly of Flanged Steel Tee with Three Flange by Coil-Lock Threaded
Adapters 10
Figure 13 - Assembly of Flanged by Coil-Lock Steel Tee with One Flange by Coil-Lock
Threaded Adapter 10
Figure 14 - Input Dimensions for Analysis of a “Snake Lay” Subsea Pipeline 12
Figure 15 - Gradual Transition for a Shore Approach to an Above Ground Pipeline 13
Figure 16 - Weight Coated SSL Using a Concrete Weight Coat 18
Figure 17 - Concrete Collars Bolted to the Outside of SSL with Locating Rings Integrally
Wound into the SSL Pipe Wall 19
Figure 18 - Weight Blankets Placed On Top of SSL Pipe to Stabilize Position 19
Figure 19 - Poured in Place Anchor Weight 20
Figure 20 - Screw Anchor for Holding SSL to the Seabed 21
Figure 21 - Spade Anchor Holding SSL to the Seabed 21
Figure 22 - Typical S-Lay Configuration 23
Figure 23 - Failure, Strength and Design Envelopes for SSL3 to SSL10 at 21°C (70°F) 26
Figure 24 - Allowable Combinations of Axial Tension and S-Lay Bend Radius for 12-inch
SSL 27
Figure 25 - Design Envelops for SSL5 versus X-42 Carbon Steel 29
Figure 26 - Allowable Combinations of Axial Tension of 12-inch SSL vs X-42 Steel Pipe 29
Figure 27 - Typical "J-Lay" for SSL Pipe 30
Figure 28 - Typical Off Bottom Tow Installation of SSL Pipe 31
Figure 29 - Typical Surface Tow Installation of SSL Pipe 32
Figure 30 - Typical Mid Depth Tow Installation of SSL Pipe 32
Figure 31 - Typical Graph Showing the Cyclic De-rating Factor A3, at the Specified
Number or Cycles and Ratio of Maximum and Minimum Cyclic Operating
Pressures 38
Figure 32 - Relationship Between the Short-Term Failure Stress and the Long-Term Design
Stress of the GRE Component of SSL 40
Figure 33 -Regression Analysis of the GRE Component of SSL Pipe at 66°C 41
Figure 34 - GRE Design Envelope at f2=0.67 41
iv ©Ameron 2003 - FP914 03/03

Figure 35 - Equivalent SSL Design Envelope at f2=0.67 and Operating Temperature at 21°C 42
Figure 36 - Equivalent SSL Trapezoidal Design Envelope at f2=0.67 and Operating
Temperature at 21°C 42
Figure 37 - Logic Diagram for Design of Subsea Pipelines per DNV-OS-F101 54
Figure 38 - Logic Flow Diagram for Design of Subsea SSL Pipelines Laying on the Bottom 55
Figure 39 - Typical Output Graph for a Subsea Pipeline - Stress Condition as a Closed End
Pressure Vessel 61
Figure 40 - Minimum Hoop Yield and Tensile Strength of the Overall SSL Pipe Wall 63
Figure 41 - Most Common Buckling Mode Used for SSL in NASA SP-8007 66
Figure 42 - Typical Output Graph for Bearing Capacity Factors for Subsea Soil 69
Figure 43 - Typical Output Graph for Subsea Pipelines - Stress Condition for Internal
Pressure plus Unsupported Span 72
Figure 44 - Typical Graph of Calculated Spans for VIV versus Input Span 75
Pressure plus Installed Bend Radius 77
Figure 46 – Typical Output Graph for Subsea Pipelines – Stress Condition for Internal
Pressure plus Installed Bend Radius in the Vertical Plane plus Unsupported Span 79
Figure 47 - Jay Lay Geometry During Installation 80
Figure 48 - J-Lay Installation Diagram 82
Figure 49 - Combined Axial Stress at the J-Lay Installation 84
Figure 50 - S-Lay Installation Diagram 87
Figure 51 - Typical Summary of Inputs for Product Dimensions, Operating Conditions,
Subsea Conditions and Installation Conditions 88
Figure 52 - Typical Summary of Inputs for Subsea Soil Conditions and Design Factors 89
Figure 53 - Typical Summary of Inputs for Design Factors (continued) and Deflections and
Vortex Induced Vibration Coefficients 91
Figure 54 - Typical output of the flow calculations using the Mathcad® workbook 96
Figure 55 - Graph of Interpolated Kinematic Viscosities 99
Figure 56 - Graph of the Linear Coefficients 100
Figure 57 - Graph of the Exponential Coefficients 100
Figure 58 - Permanent Internal and External Integrity Monitoring Setup for Subsea Pipelines103
Figure 59 - Detail of the Monitoring Station for Internal and External Integrity Monitoring 103
Figure 60 - Expansion Plug, Backed-Up by Metal Clamp 104
Figure 61 - Allowable Design Stress for the GRE Component of SSL (English Units), with
f2=0.67 111
Figure 62 - Allowable Design Stress for the GRE Component of SSL (SI Units), with
f2=0.67 112
Figure 63 - Allowable Design Stress for the GRE Component of SSL (English Units), with
f2=0.83 113
Figure 64 - Allowable Design Stress for the GRE Component of SSL (SI Units), with
f2=0.83 114
Figure 65 - Equivalent Allowable Design Stress for the Overall SSL Pipe Wall at 70°F
(English Units) with f2=0.67 115
v ©Ameron 2003 - FP914 03/03

Figure 68 - Equivalent Allowable Design Stress for the Overall SSL Pipe Wall at 21°C (SI
Units) with f2=0.67 118
Figure 77 - Allowable Axial Load vs. Bend Radius Combinations for 8-inch SSL During
Installation (J-Lay or S-Lay) 127
vi ©Ameron 2003 - FP914 03/03

Figure 91 - SSL Pipe Configuration File 141
Figure 92 - Special Execution Parameters 142
vii ©Ameron 2003 - FP914 03/03

Table of Tables
Table 1 - SSL Pressure Ratings for 3- to 10-Strip Pipe 3
Table 2 - Temperature De-Rating Factors 4
Table 3 - Thrust Reactions for SSL Pipe at Rated Pressure installed at 21°C and operated at
21°C (70°F) 15
Table 4 - Thrust Reactions for SSL Pipe at Rated Pressure, Installed at 21°C (70°F) and
Operated at 66°C (150°F) 16
Table 5 - Thrust Reactions for SSL Pipe at Rated Pressure, Installed at 21°C and Operated at
93°C (200°F) 17
Table 6 - Maximum Axial Stress That Can Be Applied to SSL During S-Lay at 21°C 26
Table 7 - Typical Assembly and Test Times for SSL Coil-Lock Joints 33
Table 8 - Relative Lay Rates and Savings for SSL vs. Carbon Steel 34
Table 9 - Coordinates for Long-Term Trapezoidal SSL Design Envelope in English Units 43
Table 10 - Coordinates for Long-Term Trapezoidal SSL Design Envelope in SI Units 43
Table 11 - Equivalent Elastic Properties of the SSL Pipe Wall at 21°C (70°F) 47
Table 14 - Typical Long-Term EI Values (Hoop Direction) for the SSL Pipe Wall at Various
Temperatures 50
Table 15 - Boundary Condition Coefficients for VIV Calculations 59
Table 16 - Minimum Hoop Tensile and Yield Strength of SSL Pipe 63
®
Table 17 - Typical Summary of Calculation Outputs from MathCAD Analysis, Including
Design Checks 92
Table 18 - Viscosity data used for the derivation of the viscosity equation 95
Table 19 - Relative Erosion Rate Values for Selected Materials 106
viii ©Ameron 2003 - FP914 03/03

1. INTRODUCTION
This is a special guide for the design and installation of Steel Strip Laminate (SSL) subsea
pipelines. A separate guide is available from Ameron to aid pipeline engineers in the design of
on shore pipelines. As with all piping materials, use of SSL pipe requires careful design of the
pipeline under operating and installation conditions. To simplify this process, Ameron
provides special MathCAD® software. This software assists the pipeline engineer in
determining loads during installation and under operating conditions.
Methods for determining the equivalent elastic properties of the SSL pipe wall structure are
also presented. These properties allow use of existing Finite Element Pipeline Analysis (FEA)
programs. The limitations and advantages of both software approaches are discussed to aid the
pipeline engineer in their use.
2. ADVANTAGES IN THE USE OF SSL AS A SUBSEA

PIPELINE
SSL has many unique advantages over conventional carbon steel pipe in subsea applications as
follows:
1. SSL is corrosion resistant on both the inside and outside of the pipe, providing long
maintenance free service life.
2. No inhibitor injection required to control corrosion rates.
3. SSL does not require external cathodic protection such as sacrificial anodes.
4. The available SSL jointing systems are very quick to assemble, substantially reducing the
barge time to install a subsea line.
5. Because of the high lay rates possible with SSL, the installed cost can be less than carbon
steel.
6. The seals on the SSL Coil-Lock joint can be quickly checked at assembly, before laying
the pipe over the side of the lay barge.
7. The inside diameter of SSL pipe is very smooth, resulting in significantly lower pressure
losses for the same nominal pipe size. Often a smaller pipe will have equal or higher flow
rates.
8. The GRE component of SSL is a natural insulator that can delay temperature loss of the
transported crude oil. This will also allow for lower viscosity and higher flow rates over
longer distances. Additional integral insulation can be applied to the outside of SSL for
special applications. Consult your Ameron or Centron representative with your special
requirements.
9. A proprietary resin can be utilized in the liner of SSL that delays wax and hydrate buildup.
10. The external collapse resistance of SSL can be made to any desired level, while
maintaining a constant pressure rating, by properly selecting the ratios of GRE and steel.
This is particularly useful in deep applications.
11. The Coil-Lock joint is a natural collapse arrestor in deep subsea applications.
12. SSL is substantially lighter than conventional steel pipe. This can be useful in unstable
muddy bottoms where the pipeline may have a tendency to settle into the mud.
1 ©Ameron 2003 - FP914 03/03

3. SSL PIPE AND FITTING CONFIGURATIONS
3.1 PIPE WALL
3.1.1 Description
A cross section of Steel Strip Laminate (SSL) pipe wall construction is shown in Figure 1. The
pipe combines conventional Glass Reinforced Epoxy (GRE) composite on the inside and
outside surfaces of the pipe wall, with high-strength steel strip layers in between. The GRE
inner and outer jackets provide corrosion resistance equivalent to conventional GRE pipe and
steel provides high strength reinforcement. The joining system (described later) encapsulates
the steel strips at both ends of every pipe section to provide complete corrosion protection for
the steel.
Figure 1 - SSL Pipe Wall
2 ©Ameron 2003 - FP914 03/03

3.1.2 Pressure Rating

Table 1 gives the pressure rating of SSL pipe at 21°C (70°F) with Coil-Lock® joints. The GRE
component of SSL pipe is designed to be in compliance with API- Specification 15HR(1), while
the steel component is in compliance with the design stresses specified in API-5L(2), ASME
B31.4(3), B31.8(4) and DNV-OS-F101 for subsea.
Table 1 - SSL Pressure Ratings for 3- to 10-Strip Pipe
Pipe SSL3 SSL4 SSL5 SSL6 SSL7 SSL8 SSL9 SSL10
Diameter (bar) (bar) (bar) (bar) (bar) (bar) (bar) (bar)
(mm)
200 146.0 180.8 215.5 250.1 284.8 319.5 354.1 388.7
250 116.0 143.7 171.2 198.8 226.3 253.9 281.4 308.9
300 97.2 120.4 143.5 166.6 189.7 212.8 235.8 258.9
350 88.6 109.7 130.8 151.8 172.9 193.9 214.9 236.0
400 77.4 95.9 114.3 132.7 151.1 169.4 187.8 206.2
450 70.3 87.1 103.8 120.5 137.2 153.8 170.5 187.2
500 63.3 78.4 93.4 108.4 123.5 138.5 153.5 168.5
550 56.1 69.5 82.8 96.2 109.5 122.8 136.1 149.5
600 52.7 65.3 77.8 90.4 102.9 115.4 127.9 140.4
700 43.6 53.9 64.3 74.6 85.0 95.3 105.6 116.0
750 40.7 50.3 60.0 69.7 79.3 89.0 98.6 108.3
800 38.1 47.2 56.3 65.3 74.4 83.4 92.5 101.5
900 33.9 42.0 50.0 58.1 66.1 74.2 82.2 90.2
1000 30.5 37.8 45.0 52.3 59.5 66.7 74.0 81.2

Diameter (psi) (psi) (psi) (psi) (psi) (psi) (psi) (psi)
(in)
8 2,118 2,622 3,125 3,628 4,131 4,633 5,136 5,638
10 1,683 2,083 2,483 2,883 3,283 3,682 4,081 4,481
12 1,410 1,746 2,081 2,416 2,751 3,086 3,420 3,755
14 1,285 1,591 1,897 2,202 2,507 2,812 3,117 3,422
16 1,123 1,391 1,657 1,924 2,191 2,457 2,724 2,990
18 1,020 1,263 1,505 1,747 1,989 2,231 2,473 2,715
20 918 1,137 1,355 1,573 1,791 2,009 2,227 2,444
22 814 1,008 1,201 1,395 1,588 1,781 1,974 2,168
24 765 947 1,129 1,310 1,492 1,674 1,855 2,036
28 632 782 932 1,082 1,232 1,382 1,532 1,682
30 590 730 870 1,010 1,150 1,290 1,430 1,570
32 553 685 816 948 1,079 1,210 1,341 1,473
36 492 609 725 842 959 1,076 1,192 1,309
40 442 548 653 758 863 968 1,073 1,178
3 ©Ameron 2003 - FP914 03/03

3.1.3 Temperature De-rating Factors

Table 2 gives the temperature de-rating factors for SSL pipe operating at temperatures up to
93°C (200°F). The temperatures highlighted in purple are those for which long-term regression
data exists. The de-rating factors have been linearly interpolated between these temperatures,
per API 15HR(1). The exact de-rating factor changes slightly between 3- and 10-steel strips,
due to the load sharing between the steel and GRE. The exact de-rating factor is automatically
calculated in the MathCAD® workbook presented in Section 7. For temperatures over 93°C
(200°F) contact your Ameron or Centron representative.
Table 2 - Temperature De-Rating Factors
Temperature °C Temperature °F De-rating Factor
-40 to 21 -40 to70 1.000
25 77 0.984
30 86 0.963
35 95 0.943
40 104 0.923
45 113 0.902
50 122 0.882
55 131 0.862
60 140 0.841
66 150 0.817
70 158 0.798
75 167 0.775
80 176 0.752
85 185 0.729
90 194 0.706
93 200 0.692
3.2 COIL-LOCK JOINING SYSTEMS AND ADAPTERS
3.2.1 Coil-Lock
Figure 2 shows the standard Coil-Lock® Joint. It comes with three types of seal configurations
as follows:
1. Single O-ring for water and other non-hazardous materials
2. Double O-ring for oil or potentially hazardous material (Standard)
3. Double O-ring, with injected adhesive sealant, for critical applications such as high-
pressure sour gas.
4 ©Ameron 2003 - FP914 03/03

The double O-ring, with injected sealant, is shown in Figure 3. Advantages of the standard
Coil-Lock® joint include the following:
1. O-ring seals can be pressure tested with air at assembly, to assure leak tight integrity.
2. Outstanding leak tight integrity when adhesive sealant is utilized. Can be used for critical
applications such as sour gas.
3. Can be disassembled, even when adhesive sealant is used.
4. Fast, easy installation.
Figure 2 - Standard Coil-Lock® Joint for SSL Pipe
Figure 3 - Seal Designs Available for the Coil-Lock® Joint
5 ©Ameron 2003 - FP914 03/03

3.2.2 Threaded Alloy or Coated Steel Coil-Lock Adapters

Figure 4 and Figure 5 show male and female Coil-Lock by ANSI steel flange end adapters.
These Coil-Lock adapters can be manufactured using customer specified alloy and can be
welded to any specialized steel fitting required to interface with ridged or flexible risers.
Proprietary connectors for remote operated vehicles (ROV) assembly can also be adapted to
SSL using the male or female Coil-Lock steel adapter.
Figure 4 - Coil-Lock Male Threaded Steel End Adapter
Figure 5 - Coil-Lock Female Threaded Steel End Adapter
6 ©Ameron 2003 - FP914 03/03

3.3 PIPE CONFIGURATIONS

The intent is to construct SSL subsea pipelines using the standard configurations shown below.
If specials are required, consult your Ameron or Centron representative.
3.3.1 Coil-Lock® Bell by Coil-Lock® Spigot SSL Pipe

Figure 6 shows the standard bell-by-spigot pipe. This is the most common configuration
required to assemble an SSL pipeline.
Figure 6 - Coil-Lock® Bell by Coil-Lock® Spigot SSL Pipe
3.3.2 Coil-Lock® Spigot by Coil-Lock® Spigot SSL Pipe

Figure 7 shows a Coil-Lock® spigot by Coil-Lock® spigot SSL pipe. This pipe is typically
utilized at one end of a subsea pipeline to in order to utilize the female threaded end adapter,
shown in Figure 5, at both ends of the pipeline.
Figure 7 - Coil-Lock® Spigot by Coil-Lock® Spigot SSL Pipe
7 ©Ameron 2003 - FP914 03/03

3.3.3 Repair Pipe

Figure 8 shows an SSL repair pipe. This pipe can be inserted into an existing line, in order to
replace a pipe that has been damaged. Replacement is accomplished by cutting the damaged
pipe from the line and disassembling the joints. The repair pipe can then be inserted into the
line and screwed into the bell and spigot, by telescoping the middle section. A second method
of repair is shown in Figure 9. This repair is completed by threading on alloy steel Coil-Lock
adapters as shown in Figure 4 and Figure 5 to the ends of the undamaged pipe and then closing
the pipeline with a fabricated alloy steel spool.
Figure 8 - Repair Pipe
Figure 9 - Repair Spool
8 ©Ameron 2003 - FP914 03/03

3.4 FITTING CONFIGURATIONS

Alloy or epoxy coated flanged steel elbows and Tees can be bolted directly to Coil-Lock Male
or Female steel adapters as shown in Figure 10, Figure 11, Figure 12 and Figure 13. Male or
female Coil-Lock® steel end adapters can be utilized at either end of the elbow or run on the tee
to maintain lay direction.
Figure 10 - Assembly of Flange by Flange Steel Elbow with Two Flange by Coil-Lock
Threaded Adapters
Figure 11 - Assembly of Flange by Coil-Lock Steel Elbow with One Flange by Coil-Lock
Threaded Adapter
9 ©Ameron 2003 - FP914 03/03

Figure 12 - Assembly of Flanged Steel Tee with Three Flange by Coil-Lock Threaded
Adapters
Figure 13 - Assembly of Flanged by Coil-Lock Steel Tee with One Flange by Coil-Lock
Threaded Adapter
10 ©Ameron 2003 - FP914 03/03

4. DESIGN OF SUBSEA PIPELINES
4.1 INTRODUCTION
This section defines the practical aspects of designing an SSL pipeline for subsea service. The
analysis method for determining the stresses for subsea pipelines, using CAESAR II® or other
FEA programs, is outlined in Section 6. Sections 7.2 to 7.4 describe analysis methods using
Ameron proprietary MathCAD® programs to determine stresses in subsea SSL pipelines.
4.2 DIRECTIONAL CHANGES FOR SUBSEA PIPELINES
4.2.1 Field Bends Using Elbows or Installed Bend Radius in Subsea Pipelines
SSL pipe cannot be field bent by yielding the pipe wall. All field bends must be done using
elbows, or flexing of the pipe elastically to make a gradual directional change or use of a short
flexible jumper like Co-Flexip. Vertical elevation changes in subsea pipelines can usually be
accommodated by flexing the pipe to follow the bottom contour. Horizontal changes will
usually require gradual flexing of the pipe and can typically be built into the lay path of the
pipeline. When this is not possible, an alloy fitting with field welding to the alignment is
required. The stresses caused by bending to the installed bend radius must be included in the
axial loads in order to determine the total combined axial loading conditions on the SSL pipe
wall. For subsea pipelines, the axial stress caused by pressure induced thrust, any axial
constraint imposed on the pipeline, axial stress at unsupported spans and axial stress from
installed bend radius must all be added. This changes the ratio of hoop to axial stress and will
usually require lowering the allowable hoop stress and therefore the pressure rating of the pipe.
See Section 7.4. A general rule-of-thumb for a subsea pipeline is one or two additional steel
strips, over the minimum pressure requirements, will be needed to account for axial stress
caused by flexural bending from installed bend radius and unsupported span.
4.2.2 Snake-Lay for Thermal Expansion Using “Flexure of the Pipe”

Thermal expansion of subsea pipelines is sometimes accommodated by laying pipeline on the
seabed in a snake pattern. This approach allows for thermal expansion, but will also require
additional layers of steel in the pipe wall to allow for the additional axial stress caused by
flexing the pipe into the snake lay pattern. The amount of steel required will be a function of
the “MinBendRadius” chosen for the snake lay. The equations below, along with Figure 14,
help define the dimensions for the offset of the “snake lay” relative to the “installed bend
radius” with some appropriate safety factor for installation error.
11 ©Ameron 2003 - FP914 03/03

A typical analysis of the primary dimensions for a snake lay are as follows:
MinBendRadius := 500⋅ m This is the minimum installed bend radius of the "Snake Lay"
L2 := 100⋅ m This is the width of the straight section beween "S-curves"
SafetyFactor := 2 This is a safety factor on installed Bend Radius due to

difficulties in maintaining a constant bend radius during the
pipe laying operation. This insures the minimum bend radius
is not exceeded locally. Suggested value is 2.
YTotal := 50⋅ m This is the total lateral offset in the "S-curve"
0.5
  YTotal  
2
 2 
L1 := ⋅
( SafetyFactor MinBendRadius ) − ( SafetyFactor MinBendRadius ) 
⋅ − 
   2  
L1 =
L3 := 2⋅ L1
L3 = This is the length of the "S-Curve"
LTotal := 4⋅ L1 + L2
LTotal = This is the total length of "Snake"
 YTotal  This is the angle at the inflection nodes. Can be

Theta := atan   Theta = deg
 2⋅ L1  used to aid field alignment.
The above dimensions for L1, L2 and L3 are shown in Figure 14.
Figure 14 - Input Dimensions for Analysis of a “Snake Lay” Subsea Pipeline
12 ©Ameron 2003 - FP914 03/03

4.3 SHORE APPROACH FOR SSL SUBSEA PIPELINES

This section defines the practical aspects required for design of an approach to shore.
Shore approaches should consider the following:
1. The approach should be as straight as possible and perpendicular to the shore to minimize
wave action. This will minimize axial loads required to later pull the SSL pipe into the
shore approach encasement.
2. The SSL pipe should be incased in a concrete coated steel pipe to protect it from wave
damage.
3. The SSL pipe should be pulled into place after installation of the encasement. Roller
centralizers can be utilized to reduce axial loads during installation.
4. Some form of anchoring will be required at the entrance to the encasement. This is
necessary to prevent excessive pipe movement. See Section 4.3.2 for reaction forces.
5. Some approaches may require horizontal directional drilling, due to a cliff or abrupt
elevation changes at the shoreline.
4.3.1 Transitions from Shore Approach to Above Ground Pipelines

It is recommended that the long, gradual transitions, shown in Figure 15, be used for transitions
from the shore encasement to an above ground condition whenever possible to reduce cost,
bend radius and axial force required to pull the SSL pipe into the encasement. Please note, the
pipe used in the gradual transition may require additional steel strips to account for flexing of
the pipe in the transition and the additional axial stress from pressure induced thrust.
Figure 15 - Gradual Transition for a Shore Approach to an Above Ground Pipeline
13 ©Ameron 2003 - FP914 03/03

4.3.2 Thrust Reactions
4.3.2.1 Reaction Force from Internal Pressure Induced Thrust

The MathCAD® program calculates the thrust force produced by a fully anchored pipeline.
Table 3 gives thrust reactions for pipe installed and operated at 21°C. The equation for the
thrust assumes a long pipeline that is “virtually anchored” by its length:
−1 ( K15 + K16 − K13 − K14 − K10 + K11 − K12 + K9 )
AxialLoad ⋅π ⋅D m ⋅
4 (
− ν GREHoopAxial ⋅ K1 + ν GREHoopAxial ⋅ K2 ⋅ν Stl − ν Stl ⋅ K6 + K5 )
Where :
P Internal pressure at operating temperature

EGREHoop Modulus of GRE in hoop direction at the specified
operating temperature
EGREAxial Modulus of GRE in axial direction at the specified

operating temperature
ν GREHoopAxial Poisson's ratio for the GRE in the hoop to axial direction at
the specified operating temperature
EStl Modulus of steel. TLiner Thickness of GRE Liner

ν Stl Poisson's Ratio for
TJacket Thickness of GRE Jacket
the steel layers
ID Inside Diameter of Pipe TStl N ⋅TStlStrip Overall thickness of

steel layers
N Number of steel strips TGRE TLiner + TJacket Overall thickness of

GRE layers
TStlStrip Thickness of each Dm ID + TGRE + TStl Pipe mean diameter

steel strip
TStl ⋅ν GREAxialHoop ν Stl
K1 + K9 P ⋅ D m ⋅ K5
TGRE ⋅ EGREHoop EStl
K10 P ⋅ D m ⋅ ν GREHoopAxial ⋅ K1
1 TStl
K2 +
EStl EGREAxial ⋅ TGRE K11 P ⋅ D m ⋅ ν GREHoopAxial ⋅ K2 ⋅ ν Stl
P ⋅D m K12 P ⋅ D m ⋅ ν Stl ⋅ K6
K3
2 ⋅ TGRE ⋅ EGREHoop
K13 4 ⋅ K3 ⋅ TGRE ⋅ EGREAxial ⋅ ν GREAxialHoop ⋅ K5
1 TStl
K5 + K14 4 ⋅ K3 ⋅ TGRE ⋅ EGREAxial ⋅ K2 ⋅ ν Stl
EStl TGRE ⋅ EGREHoop
K15 4 ⋅ K3 ⋅ TGRE ⋅ EGREAxial ⋅ K1

TStl ⋅ν GREHoopAxial ν Stl
K6 +
TGRE ⋅ EGREAxial EStl K16 4 ⋅ K3 ⋅ TGRE ⋅ EGREAxial ⋅ ν GREAxialHoop ⋅ ν Stl ⋅ K6
14 ©Ameron 2003 - FP914 03/03

Table 3 - Thrust Reactions for SSL Pipe at Rated Pressure installed at 21°C and operated
at 21°°C (70°F)

Diameter (Newton) (Newton) (Newton) (Newton) (Newton) (Newton) (Newton) (Newton)
(mm) (Pound) (Pound) (Pound) (Pound) (Pound) (Pound) (Pound) (Pound)
(inch) x105 x105 x105 x105 x105 x105 x105 x105
200 1.905 2.396 2.892 3.391 3.896 4.406 4.921 5.441

(8) (0.428) (0.539) (0.650) (0.762) (0.876) (0.991) (1.106) (1.223)
250 2.336 2.934 3.538 4.145 4.756 5.375 5.997 6.624
(10) (0.525) (0.660) (0.795) (0.932) (1.069) (1.208) (1.348) (1.489)
300 2.794 3.508 4.227 4.948 5.677 6.408 7.147 7.889
(12) (0.628) (0.789) (0.950) (1.112) (1.276) (1.441) (1.607) (1.774)
350 3.447 4.325 5.154 6.015 6.842 7.713 8.59 9.431
(14) (0.775) (0.972) (1.159) (1.352) (1.538) (1.734) (1.931) (2.120)
400 3.914 4.911 5.858 6.811 7.804 8.771 9.744 10.67
(16) (0.880) (1.104) (1.317) (1.531) (1.754) (1.972) (2.191) (2.399)
450 4.396 5.519 6.584 7.655 8.795 9.876 10.96 12.05
(18) (0.988) (1.241) (1.480) (1.721) (1.977) (2.220) (2.464) (2.709)
500 4.908 6.129 7.276 8.546 9.783 10.94 12.11 13.28
(20) (1.103) (1.378) (1.636) (1.921) (2.199) (2.459) (2.722) (2.985)
550 5.365 6.737 8.03 9.356 10.65 12.04 13.25 14.56
(22) (1.206) (1.515) (1.805) (2.103) (2.394) (2.707) (2.979) (3.273)
600 5.874 7.329 8.712 10.2 11.62 13.15 14.46 15.77
(24) (1.321) (1.648) (1.959) (2.293) (2.612) (2.956) (3.251) (3.545)
700 6.647 8.333 10.01 11.71 13.4 15.09 16.81 18.51
(28) (1.494) (1.873) (2.250) (2.633) (3.012) (3.392) (3.779) (4.161)
750 7.121 8.91 10.72 12.52 14.33 16.16 17.98 19.8
(30) (1.601) (2.003) (2.410) (2.815) (3.222) (3.633) (4.042) (4.451)
800 7.584 9.512 11.43 13.35 15.28 17.21 19.15 21.09
(32) (1.705) (2.138) (2.570) (3.001) (3.435) (3.869) (4.305) (4.741)
900 8.664 10.74 12.92 15.09 17.26 19.36 21.59 23.82
(36) (1.948) (2.414) (2.905) (3.392) (3.880) (4.352) (4.854) (5.355)
1000 9.461 11.82 14.22 16.6 19.01 21.4 23.82 26.22
(40) (2.127) (2.657) (3.197) (3.732) (4.274) (4.811) (5.355) (5.894)
4.3.2.2 Reaction Force from Internal Pressure Induced Thrust and Thermal Expansion
The MathCAD® program for subsea SSL pipelines also calculates the total thrust produced
from internal pressure and thermal growth. The analysis is based on an assumed very long
pipeline that is “virtually anchored” away from shore by its own weight. Table 4 and Table 5
give thrust eactions for pipe installed at 21°C (70°F) and operating at 66 (150°F) and 93°C
(200°F) respectively. The equation for total thrust is:
15 ©Ameron 2003 - FP914 03/03

−1 ( K9 − K10 + K11 − K12 − K13 − K14 + K15 + K16 − K17 + K18 + K19 − K20 )
AxialLoad ⋅π ⋅Dm ⋅
4 ( )
− ν GREHoopAxial ⋅K1 + ν GREHoopAxial ⋅K2 ⋅ν Stl − ν Stl ⋅K6 + K5
Where :
K17 4 ⋅TGRE⋅EGREAxial⋅K1 ⋅K6 ⋅β SSL ⋅EStl ⋅Temp β SSL = Thermal Coefficient of Expansion
K18 4 ⋅TGRE⋅EGREAxial⋅K1 ⋅K6 ⋅β SSL ⋅EStl ⋅TempInstall Temp= Operating Temperature
K19 4 ⋅TGRE⋅EGREAxial⋅K2 ⋅β SSL ⋅EStl ⋅Temp ⋅K5 Temp Install = Installation Temperature
K20 4 ⋅TGRE⋅EGREAxial⋅K2 ⋅β SSL ⋅EStl ⋅TempInstall ⋅K5
All other Ki variables are the same as those defined in Section 4.3.2.1 above.
Table 4 - Thrust Reactions for SSL Pipe at Rated Pressure, Installed at 21°C (70°F) and
Operated at 66°C (150°F)

(inch) x105 x105 x105 x105 x105 x105 x105 x105
200 3.098 3.935 4.741 5.552 6.371 7.195 8.026 8.863

(8) (0.696) (0.885) (1.066) (1.248) (1.432) (1.618) (1.804) (1.992)
250 3.824 4.851 5.84 6.834 7.835 8.844 9.857 10.88
(10) (0.860) (1.091) (1.313) (1.536) (1.761) (1.988) (2.216) (2.446)
300 4.657 5.797 6.976 8.159 9.35 10.55 11.75 12.96
(12) (1.047) (1.303) (1.568) (1.834) (2.102) (2.372) (2.642) (2.914)
350 5.415 6.826 8.245 9.639 11.01 12.41 13.79 15.14
(14) (1.217) (1.535) (1.854) (2.167) (2.475) (2.790) (3.100) (3.404)
400 6.198 7.793 9.353 10.92 12.54 14.07 15.61 17.12
(16) (1.393) (1.752) (2.103) (2.455) (2.819) (3.163) (3.509) (3.849)
450 6.977 8.771 10.52 12.77 14.09 15.8 17.52 19.24
(18) (1.568) (1.972) (2.365) (2.871) (3.168) (3.552) (3.939) (4.325)
500 7.723 9.69 11.6 13.65 15.57 17.5 19.36 21.23
(20) (1.736) (2.178) (2.608) (3.069) (3.500) (3.934) (4.352) (4.773)
550 8.446 10.67 12.75 14.95 17.08 19.21 21.19 23.34
(22) (1.899) (2.399) (2.866) (3.361) (3.840) (4.319) (4.764) (5.247)
600 9.253 11.6 13.87 16.28 18.6 20.93 23.17 25.32
(24) (2.080) (2.608) (3.118) (3.660) (4.181) (4.705) (5.209) (5.692)
700 10.81 13.56 16.29 19.04 21.67 24.34 27.15 29.69
(28) (2.430) (3.048) (3.662) (4.280) (4.872) (5.472) (6.104) (6.675)
750 11.56 14.51 17.44 20.37 23.3 26.26 29.07 31.73
(30) (2.599) (3.262) (3.921) (4.579) (5.238) (5.903) (6.535) (7.133)
800 12.25 15.48 18.59 21.71 24.84 27.92 31.01 33.93
(32) (2.754) (3.480) (4.179) (4.881) (5.584) (6.277) (6.971) (7.628)
900 13.8 17.23 20.67 24.32 27.77 31.23 34.8 38.16
(36) (3.102) (3.873) (4.647) (5.467) (6.243) (7.021) (7.823) (8.579)
1000 15.42 19.28 23.17 27.04 30.95 34.83 38.35 42.65
(40) (3.467) (4.334) (5.209) (6.079) (6.958) (7.830) (8.621) (9.588)
16 ©Ameron 2003 - FP914 03/03

Table 5 - Thrust Reactions for SSL Pipe at Rated Pressure, Installed at 21°°C and
Operated at 93°C (200°F)

(inch) x105 x105 x105 x105 x105 x105 x105 x105
200 3.642 4.569 5.511 6.46 7.415 8.359 9.318 10.28

(8) (0.819) (1.027) (1.239) (1.452) (1.667) (1.879) (2.095) (2.311)
250 4.539 5.649 6.829 7.989 9.157 10.33 11.5 12.67
(10) (1.020) (1.270) (1.535) (1.796) (2.059) (2.322) (2.585) (2.848)
300 5.45 6.797 8.168 9.563 10.95 12.34 13.75 15.15
(12) (1.225) (1.528) (1.836) (2.150) (2.462) (2.774) (3.091) (3.406)
350 6.292 7.933 9.533 11.16 12.75 14.34 15.97 17.58
(14) (1.414) (1.783) (2.143) (2.509) (2.866) (3.224) (3.590) (3.952)
400 7.164 9.052 10.85 12.67 14.51 16.34 18.17 20.01
(16) (1.611) (2.035) (2.439) (2.848) (3.262) (3.673) (4.085) (4.498)
450 8.046 10.18 12.21 14.23 16.32 18.37 20.42 22.52
(18) (1.809) (2.289) (2.745) (3.199) (3.669) (4.130) (4.591) (5.063)
500 8.947 11.21 13.51 15.84 18.13 20.37 22.62 24.9
(20) (2.011) (2.520) (3.037) (3.561) (4.076) (4.579) (5.085) (5.598)
550 9.813 12.38 14.87 17.39 19.86 22.4 24.83 27.34
(22) (2.206) (2.783) (3.343) (3.909) (4.465) (5.036) (5.582) (6.146)
600 10.78 13.47 16.19 18.97 21.66 24.43 27.06 29.75
(24) (2.423) (3.028) (3.640) (4.265) (4.869) (5.492) (6.083) (6.688)
700 12.52 15.69 18.84 22.01 25.19 28.33 31.51 34.73
(28) (2.815) (3.527) (4.235) (4.948) (5.663) (6.369) (7.084) (7.808)
750 13.41 16.79 20.19 23.55 26.95 30.35 33.75 37.14
(30) (3.015) (3.775) (4.539) (5.294) (6.059) (6.823) (7.587) (8.349)
800 14.26 17.88 21.49 25.05 28.74 32.35 36.1 39.61
(32) (3.206) (4.020) (4.831) (5.631) (6.461) (7.273) (8.116) (8.905)
900 16.09 20.08 24.23 28.22 32.39 36.4 40.46 44.6
(36) (3.617) (4.514) (5.447) (6.344) (7.282) (8.183) (9.096) (10.026)
1000 17.88 22.21 26.75 31.34 35.94 40.35 44.92 49.38
(40) (4.020) (4.993) (6.014) (7.046) (8.080) (9.071) (10.098) (11.101)
4.4 RISERS
In shallow water, it is generally recommended that a flexible jumper such as Co-Flexip be used
to make the final connection. In deeper waters, SSL can be used as a cantenary riser and in the
middle range, where the depth is too shallow to make the turn from vertical to horizontal, an
alloy elbow, similar to that shown in Figure 10 or Figure 11 will be required to make the
connections. Consult your Ameron or Centron representative for recommendations.
17 ©Ameron 2003 - FP914 03/03

4.5 BOTTOM STABILITY

The calculations for on bottom stability are beyond the scope of this manual since these
calculations are too subject to the local conditions and design practice. Brief descriptions of the
available options for anchoring SSL pipe to the seabed are given below.
4.5.1 Weight Coating and Concrete Collars

SSL can be weight coated in a manner similar to conventional steel pipe. The SSL pipe
sections will usually have to be coated on an individual basis, due to the upsets at the Coil-Lock
bell and spigot. Continuous coating is possible if the spacing on the continuous conveyor
support rollers is set to miss the Coil-Lock joint and the lead per revolution is long enough to
avoid the joint. Figure 16 shows a typical SSL pipe with a concrete weight coating. The
advantage of factory weight coating is reduced cost, but the extra weight on the pipe has to be
considered in the calculations for installation loads.
Concrete collars can also be bolted to SSL during the laying of the pipe. Integral locating rings
can be wound into the SSL pipe wall to locally reinforce the pipe for the collar and to help
locate and maintain collar position. As with conventional weight coating, the advantage of
concrete collars is reduced cost, but the extra weight on the pipe has to be considered in the
calculations for installation loads.
Figure 16 - Weight Coated SSL Using a Concrete Weight Coat
18 ©Ameron 2003 - FP914 03/03

Figure 17 - Concrete Collars Bolted to the Outside of SSL with Locating Rings Integrally
Wound into the SSL Pipe Wall
4.5.2 Weight Blankets

The main advantage for weight blankets is the ability to lay the pipe first, with reduced weight
and loads on the pipe and then later place blankets over the pipe to stabilize the pipe on the
bottom. Cost will be higher for setting the weight blankets after the pipe is laid and some form
of ROV or divers will be usually required to guide the blankets onto the pipe.
Figure 18 - Weight Blankets Placed On Top of SSL Pipe to Stabilize Position
19 ©Ameron 2003 - FP914 03/03

4.5.3 Poured in Place Anchors

The main advantage of the poured in place anchor is the pipe can be installed with no
additional weight during installation. The “poured in place” anchors are blow molded
polyethylene molds that are flooded with water while a diver places them on the pipe. In this
condition, they are essential neutral buoyancy and therefore easy for the diver to manipulate.
Once locked in place on the pipe, a hose can be fitted to the polyethylene mold to fill it with
concrete. The heavier concrete naturally displaces the water and the hose can be easily
manipulated to the fill point of the mold. Concrete is batched and mixed at the surface and
either pumped or gravity fed to the mold. See Figure 19.
Figure 19 - Poured in Place Anchor Weight
4.5.4 Mechanical Anchors

The main advantages for mechanical anchors are 1) the ability to lay the pipe first, with
reduced weight and loads on the pipe and 2) the low weight of the anchor compared to weight
coatings or blankets in unstable soils such as muddy bottoms. Weight coatings and blankets
may cause the pipe to settle into the mud while mechanical anchors will not add weight, but
will stabilize the pipe. As with weight blankets, the cost will be higher for setting the anchors
after the pipe is laid and some form of ROV or divers will usually be required to set the
anchors. Anchors can take many forms. A typical screw anchor is shown in Figure 20, while a
typical spade type anchor is shown in Figure 21.
20 ©Ameron 2003 - FP914 03/03

Figure 20 - Screw Anchor for Holding SSL to the Seabed
Figure 21 - Spade Anchor Holding SSL to the Seabed
21 ©Ameron 2003 - FP914 03/03

4.5.5 Chain Weights

The chain weights used in Off-Bottom Tow, Surface Tow and Mid-Depth Tow installation
methods can also be used to stabilize the pipe once it is at the desired location. See Sections
5.4, 5.5 and 5.6.
5. INSTALLING SUBSEA SSL PIPELINES

Most standard lay procedures utilized for carbon steel pipe can be easily adapted for SSL. A
detailed description of all the aspects of installing subsea pipelines is beyond the scope of this
manual and many of the details are proprietary to the pipe lay companies. However, a brief
description of the available techniques is presented below to show how SSL can be easily
adapted to existing installation methods. The advantages and disadvantages of installing SSL
versus carbon steel pipe are also discussed.
5.1 OVERVIEW
Several ways of laying SSL subsea are available to the installer. They are:
1. S-Lay
2. J-Lay
3. Off-Bottom Tow
4. Surface Tow
5. Mid Depth Tow
Reel installation is not possible for SSL, since the pipe cannot be yielded in the axial direction,
without destroying the pipe structure.
5.2 S-LAY
The S-Lay approach takes its name from the shape of the pipeline between the stinger and the
seabed. This approach is usually utilized in shallow water, but S-Lay can be used up to 1650
meters depth, when using state of the art barges and planned yielding of the steel pipe in the
axial direction during the lay operation. SSL cannot be bent beyond yield in the axial direction
without severe damage to the structure, so the S-Lay approach is limited to shallower depths for
SSL compared to carbon steel pipe. S-Lay requires that axial tension be applied to the pipe, in
order to minimize the bending stresses caused by the weight of the pipe and the S-shape during
the laying operation. Figure 22 shows the start up of a typical “S-Lay”. During startup, the
barge is stationary and the pipe is winched into place. Once the pipe is at the seabed pile, the
barge will begin moving as it continues to lay pipe with the “S-Curve” between the seabed and
the stinger on the back of the lay barge.
22 ©Ameron 2003 - FP914 03/03

Figure 22 - Typical S-Lay Configuration
5.2.1 Comparison of Installation Techniques for SSL and Carbon Steel pipe
Using S-Lay
In conventional S-Lay, the steel pipe sections are typically pre-welded into double or quad
length pipe to maximize the lay rate. This pre-assembly requires a long and large barge, but
allows for the time required to complete the following operations:
1. Welding
2. X-ray or ultrasonic inspection of the weld
3. Weld repair (if it does not pass inspection)
4. Repair of the liner at the weld (if one is utilized)
5. Application of the exterior coating at the welded joint
6. Leak testing (if required)
7. Installation of the sacrificial anodes.
It is obvious from the above list that substantial time is required to complete all the required
tasks and compromises are often necessary that affect long-term performance or cost. For
example, if the pipeline has an epoxy liner to help slow internal corrosive attack, the liner must
be repaired at the weld and inspected in a relatively short period of time. This usually requires
a special coupling to protect the liner at the weld in order to eliminate this problem. These
couplings are expensive and protrude into the inside of the pipe, making pressure testing during
the lay operation difficult and later pigging problems can result.
23 ©Ameron 2003 - FP914 03/03

After the pipe sections have been pre-assembled into double or quad pipe and inspected, they
are rolled into position to be welded to the back end of the pipeline being laid. The same set of
tasks and problems are now encountered at this station as well and effectively establish and
limit the pipe lay rate.
For SSL pipe, the operation is similar, but many of the time consuming steps, such as welding
plus liner and coating repair are eliminated or reduced in time. Assembly of an SSL subsea
pipeline for S-lay requires the following tasks.
1. SSL pipe can be pre-assembled into double or quad pipe prior to installation into the back
of the pipeline being laid. The required operations are as follows:
a) Pipe sections are assembled using the standard installation procedures as follows:
i) Clean and inspect the threads and sealing surfaces
ii) Install the nylon locking key into the bell
iii) Lubricate all surfaces with soap
iv) Inspect the O-rings and stretch them onto the spigot
v) Align the joints for assembly and insert axially as far as possible.
vi) Rotate the pipe by hand until resistance is felt.
vii) Tighten the joint using a chain or strap wrench until the threads bottom out.
b) Inspection can be accomplished several ways as follows:
i) Inspect the position of the makeup marks at the back end of the male threads to
ensure complete assembly.
ii) Each joint seal can be tested by applying air pressure to the annular space between
the two O-ring seals as shown in Figure 3.
iii) Steel end caps can be threaded onto the ends of the pre-assembled pipe sections and
the assembly can be tested using air or high pressure water.
2. Final assembly is accomplished by rolling the tested double or quad pipe into position for
final assembly into the back end of the pipeline being laid.
a) Assembly is completed by rotating the pipe until bottoming out the threads
b) Final inspection includes verifying the position of the makeup marks and air testing the
annular gap between the double O-ring seals as shown in Figure 3..
The entire assembly process for SSL only takes a few minutes. The final assembly (items 2a
and 2b) can be completed in less than a minute with proper automation for handling (assumes
the joints have been pre-cleaned and the O-rings plus locking keys have been installed).
Resulting pipe lay rates can be substantially faster than steel pipe. See Section 5.7 for a
comparison of lay rates and installation cost.
5.2.2 Allowable Axial Tension and Bend Radius for SSL and Carbon Steel
pipe Using S-Lay
During the lay operation, the pipe must be tensioned axially to minimize the bending stresses in
the “S-curve” caused by the weight of the pipe.
Please note, that special grips may be required to apply the axial load to the SSL pipe wall. If
standard grips are used, it is suggested the pipe be grabbed at the Coil-Lock joint. The joint is
very thick and can take the clamping loads without damage. An alternate approach is to wind
gripping points into the pipe wall during manufacture. Consult the factory for details.
24 ©Ameron 2003 - FP914 03/03

The axial failure stress for SSL has been studied and reported in work done at the University of
Houston Composites Engineering Application Center (CEAC). See Section 9.3. Testing by
Ameron has confirmed these values. The failure and strength envelope for 3 to 10 steel strips
at 21°C is presented in Figure 23. Values presented are the average stress across the entire
SSL pipe wall. Table 6 gives a summary of the axial failure stress and maximum allowable
axial stress during installation of an SSL subsea pipeline. These values can be utilized to
graph all possible combinations of allowable bend radius versus axial tension for any SSL pipe
diameter and number of steel strips, using the following equation:
 ( ID + 2⋅ t)  2  ID  2

( )
MaxAxialLoad := σMaxAxial − σSSLBending ⋅ π⋅  
 2   2  
 −   Maximum allowable
axial load
Where :
σMaxAxial Maximum allowable axial stress during

inatallation from Table 6.
σSSLBending := ESSLAxial⋅  
ID + 2⋅ t Axial stress from bending the pipe to the

 2⋅ BendRadius  installed bend radius
ESSLAxial Axial modulus of the overall SSL pipeline
BendRadius install Bend radius of the SSL pipeline during

installation
ID Internal diameter of SSL pipe
t Structural wall thickness of SSL as defined in

Table 6.
Figure 24 shows a typical graph of allowable combinations of axial load and installation bend
radius for 12-inch SSL. Appendix 10.2 contains graphs of allowable combinations of axial
load and installation bend radius for all other diameters of SSL pipe. Please note that SSL can
be specially designed for the installation loads, so consult the factory for your special
installation requirements if standard SSL does not meet your needs.
25 ©Ameron 2003 - FP914 03/03

Table 6 - Maximum Axial Stress That Can Be Applied to SSL During S-Lay at 21°C
Description Nominal Axial Stress

Maximum Allowable Safety Design
at Failure
Axial Stress During Factor to Wall
Installation at 1.5 Times Failure Thickness
Long-Term Rating (1)
(MPa) (psi) (MPa) (psi) (inch)
SSL3 220.7 32,000 96.5 13,990 2.29 0.236
SSL4 252.8 36,660 111.0 16,100 2.27 0.256
SSL5 280.4 40,660 123.1 17,850 2.28 0.276
SSL6 303.4 44,000 133.4 19,340 2.28 0.295
SSL7 326.2 47,300 142.2 20,620 2.29 0.315
SSL8 340.2 49,330 149.8 21,720 2.27 0.335
SSL9 358.6 52,000 156.5 22,690 2.29 0.354
SSL10 372.4 54,000 162.3 23,540 2.29 0.374
(1) Use this value for design purposes. The design wall thickness only includes the thickness
of the GRE reinforcing layers and the steel layers. It does not include any resin between the
steel layers or any excess resin on the OD of the pipe. Actual pipe wall will be slightly thicker.
Hoop Stress (MPa)

0 300 600 900 1200
100
o Allowable Design [3] Expt'l data
T=70 F
( 54) Ply Matrix Cracking (n=3)
Steel Strip/Epoxy Yielding (n=4) 600
Final Failure
80
4=2.35:1
Axial Stress (MPa)

3=1.90:1
1=0.70:1
2=1.80:1
=2:1
480
Axial Stress (ksi)
60
360
40
5=5.00:1
8 9
10 240
5 6 7
4
n=3
20
120
n=3 4 5 67 891
0 0
0 50 100 150 200
Hoop Stress (ksi)
Figure 23 - Failure, Strength and Design Envelopes for SSL3 to SSL10 at 21°C (70°F)
26 ©Ameron 2003 - FP914 03/03

12-inch SSL at 70F
4000
SSL3
3500
SSL4
3000 SSL5
Bend Radius (Feet)
SSL6
2500 SSL7
SSL8
2000
SSL9
1500 SSL10
1000
500
0
0.00 50.00 100.00 150.00 200.00 250.00 300.00 350.00
Axial Load (kips)
Figure 24 - Allowable Combinations of Axial Tension and S-Lay Bend Radius for 12-inch
SSL
5.2.3 Advantages and Disadvantages of SSL vs. Carbon Steel pipe Using S-
Lay
The advantages and disadvantages of the S-Lay approach for installing steel pipelines versus
SSL pipelines are as follows:
1. Advantages:
a) SSL has a significant advantage over steel pipe in lay rate, due to the mechanical Coil-
Lock joint. See Section 5.7 for a comparison.
b) An S-Lay barge can usually install welded steel pipe faster than a J-Lay barge can lay
welded steel pipe. This is because the work required to weld, inspect and coat the joints
can be easily divided over several work stations along the length of the barge. All
workstations can perform their individual tasks in parallel, reducing the overall cycle
time. For example, if the overall time to weld, inspect, X-ray, test and coat the joint
takes 45 minutes and this work can be evenly divided over 5 stations, the cycle time to
lay each 40-feet pipe section is 45/5 = 9 minutes.
c) Most barges have been modified to S-Lay, due to the faster lay rate, for welded steel
pipe, over the J-Lay approach. Therefore, there is a greater availability of these barges
around the world.
2. Disadvantages:
a) The big disadvantage for SSL, when using S-Lay, is the higher bending loads and
resulting axial stresses applied to the pipe during the lay operation compared to J-Lay.
This can be a big disadvantage for SSL, since SSL does not have the axial load caring
27 ©Ameron 2003 - FP914 03/03

capability of ordinary steel pipe. The reason for this is SSL has been designed to
optimize performance as a closed end pressure vessel, with the allowable axial stress
equal to half the allowable hoop stress. This is possible for SSL, since it is a composite.
In contrast, steel has equal strength in all directions and the allowable stress in the axial
direction is equal to that in the hoop direction. See Figure 25 for a comparison of the
allowable design envelops.
b) When installing SSL, special grips may have to be designed into the axial tensioner to
avoid damage to the SSL pipe wall. As an alternative, local buildups can be wound into
the SSL pipe wall for standard grips or the SSL pipe can be gripped at the Coil-Lock
bell using standard or special grips that take advantage of the cone shape of the bell.
5.2.4 Discussion of the Use of SSL versus Carbon Steel Pipe When Using S-
Lay
The mechanical Coil-Lock joint on SSL provides much faster lay rates compared to carbon
steel pipe when utilizing the S-Lay approach. However, SSL does not have as high an average
axial tensile strength as a typical steel pipe, so this can cause some limitations during
installation.
If S-Lay must be utilized, the steel in SSL has very high yield strength, which allows a fairly
small bend radius compared to normal carbon steel pipe and the light weight can reduce the
axial loads required to maintain the S-curve. This largely offsets the axial stress limitation of
SSL, but careful analysis is required for S-Lay. See Figure 26 for an example of this. The
design criterion for the pipe in Figure 26 is a 12-inch pipeline with a pressure rating of 2000 psi
at 150°F (66°C). This pressure and temperature rating will require at least an SSL6 and an X-
42 steel pipe with a wall thickness of 0.422-inch, when designed at 72% of SMYS for pressure
containment. It is assumed the X42 can be installed at up to 90% of yield in the axial direction.
Figure 26 shows that 12-inch SSL6 can be installed at a smaller bend radius for axial loads up
to 100,000 lbs. Above 100,000 lbs of axial load, the X-42 pipe can be installed at a smaller
bend radius than SSL, however, the X-42 pipe will weigh about 2.5 times the weight of the 12-
inch SSL6 pipe, so will require much more axial load to maintain the same S-curve radius. If
larger axial loads are required during installation (i.e. to resist cross currents), a 12-inch SSL10
can be installed at a smaller bend radius than the X-42 pipe at axial loads up to 240,000 lbs and
the steel pipe will still weigh 1.7 times the weight of 12-inch SSL10.
28 ©Ameron 2003 - FP914 03/03

70
SSL5 Composite Design Stress Envelope

SSL5 Composite Failure Stress Envelope
60
Plastic Yielding of SSL Steel Strips
X42 Steel Pipe Design at 72% of SMYS
50 X42 Steel Pipe Yield (SMYS)

X42 Steel Pipe Tensile Strength
Axial Stress, ksi
40
30
20
10
0
0 20 40 60 80 100 120 140
Hoop Stress, ksi
Figure 25 - Design Envelops for SSL5 versus X-42 Carbon Steel
12-inch SSL at 70F
3000 X-42 with 0.422-inch Wall

SSL6
2500 SSL7
SSL8
Bend Radius (Feet)
2000 SSL9
SSL10
1500
1000
500
0
0.00 100.00 200.00 300.00 400.00 500.00
Axial Load (kips)
Figure 26 - Allowable Combinations of Axial Tension of 12-inch SSL vs X-42 Steel Pipe
29 ©Ameron 2003 - FP914 03/03

5.3 J-LAY
The J-Lay approach also takes its name from the shape of the pipeline during installation. See
Figure 27 for a typical J-Lay installation. This approach is usually utilized in deeper water, but
J-Lay can be utilized in shallower water if the initial angle of the installation derrick is properly
set. The J-Lay curve is similar to a cantenary, so unlike the S-Lay, it does not require as much
axial tension to maintain the shape and minimize the bending stresses. The J-Lay installation
makes the most sense for SSL installation, when it is possible. The Coil-Lock joint is very easy
to assemble in this configuration and is similar to running a down hole tubing string.
Combined axial tension and bend radius must be carefully reviewed, but they are typically
much lower than the loads applied to a pipe using S-Lay at the same water depth. The
equations in Section 5.2.2 and the graphs in Appendix 10.2 are useful for both S-Lay and J-Lay
in determining allowable combinations of axial load and installation bend radius.
The big advantage in the S-Lay approach over the J-Lay approach is faster installation rates
when installing welded carbon steel pipe. The mechanical Coil-Lock joint on SSL overcomes
this limitation, providing similar lay rates when utilizing either the S-Lay or J-Lay approach.
For the above reasons, Ameron does not recommend the S-Lay approach when a J-Lay can be
substituted
When using J-Lay, the SSL pipe can be pre-assembled and tested into double, triple or quad
length pipe on the deck of the barge before being tilted up to the derrick for final assembly to
the back end of the previously installed pipe. See Section 5.7 for comparative lay rates.
Figure 27 - Typical "J-Lay" for SSL Pipe
30 ©Ameron 2003 - FP914 03/03

5.4 OFF BOTTOM TOW

Figure 28 shows a typical Off Bottom Tow method of installation. The stability of the pipe
assembly off the seabed floor is maintained by a combination of ballast chains and buoys. The
weight of the chains drag on the seabed until a natural buoyancy balance is achieved. This
form of installation makes the most sense, when the water depth is shallow and port access to a
large assembly area is available for assembling the long sections for tow. A key advantage of
this approach is the long assembled pipe sections can be pre-tested, prior to installation, to
insure leak tight joints. Loads applied to the pipe during transport and installation can also be
very low compared to J- or S-Lay.
Alloy steel Coil-Lock adapters are required at the ends of each long towed pipe assembly, to
assist in underwater assembly. Special metallic fittings can also be utilized on the end of the
towed assemblies, to aid in remote piloted vehicle assembly. See Section 3.2.2 for the various
adapters available.
Figure 28 - Typical Off Bottom Tow Installation of SSL Pipe
5.5 SURFACE TOW

Figure 29 shows a typical Surface Tow method of installation. The stability of the pipe
assembly on the surface is maintained by ballast buoys and the weight of the pipe. The pipe
can be flooded or filled with air to maintain proper buoyancy. If filled with air, care must be
taken during flooding to avoid air pockets in the middle of the pipe length that may be very
hard to control and remove while sinking the pipe to the seabed floor.
This form of installation makes the most sense, when the water is deep and port access is
available to a large area for assembling and testing the long sections for tow. A key advantage
of this approach is the long assembled pipe sections can be pre-tested, prior to installation, to
insure leak tight joints. Loads applied to the pipe during transport and installation can also be
low compared to J- or S-Lay if the seas are not rough and the currents are low.
As in Off Bottom Tow, alloy steel Coil-Lock adapters are required at the ends of each long
towed pipe assembly, to assist in underwater assembly. Special metallic fittings can also be
utilized on the end of the towed assemblies, to aid in remote piloted vehicle assembly. See
Section 3.2.2 for the various adapters available.
31 ©Ameron 2003 - FP914 03/03

Figure 29 - Typical Surface Tow Installation of SSL Pipe
5.6 MID DEPTH TOW

Figure 30 shows a typical Mid Depth Tow method of installation. The stability of the pipe
assembly is maintained by a combination of ballast buoys, the weight of the pipe and the
tension applied by the towing vessels. The pipe can be flooded or filled with air to maintain
proper buoyancy. If filled with air, care must be taken during flooding to avoid air pockets in
the middle of the pipe length that may be very hard to control and remove while sinking the
pipe to the seabed floor.
Figure 30 - Typical Mid Depth Tow Installation of SSL Pipe
32 ©Ameron 2003 - FP914 03/03

5.7 RELATIVE INSTALLATION RATES AND COSTS OF SSL VERSUS

CARBON STEEL
Table 7 shows typical times for assembling a Coil-Lock joint. The tasks have been broken
down into 3 stations to balance the workload, while maximizing the installation rates. Some
automation is assumed in the cycle times shown, but they are conservative if sufficient
automation is included or larger crews provided. Review of Table 7 shows a cycle time of 4.0
minutes per 40-foot section of 12-inch SSL pipe, if no adhesive sealant is utilized. This time
includes a 1.0 minute air test of the annular space between the double O-ring seals at the joint.
If the adhesive sealed joint is utilized, the cycle time increases to 4.5 minutes. Dividing the
assembly sequence into 6 stations would allow as little as 2 minute per 40-foot pipe section.
Table 7 - Typical Assembly and Test Times for SSL Coil-Lock Joints
Description1, 2 Time Estimate in Minutes

8-Inch 16-Inch 24-Inch
Station 1 – Inspection and Loading
1. Attach Straps to SSL pipe 1.0 1.0 1.0
2. Pick up SSL pipe and place in assembly station 1.5 1.5 2.0
3. Clean spigot and bell ends with brush 1.0 1.0 1.5
4. Inspection of spigot and bell ends 0.5 0.5 0.5
Subtotal Time for Station # 1 4.0 4.0 5.0
Station 2 – Preparation for Assembly

5. Light sanding if adhesive sealant is used and 1.0(1) 1.5(1) 2.0(1)
needed*
6. O-ring inspection 0.5 0.5 1.0
7. O-ring installation and joint lubrication 1.0 1.5 2.0
8. Install nylon locking key into threads 1.0 2.0 3.0
Subtotal Time for Station # 2 w/o Adhesive Sealant 2.5 4.0 4.0
Subtotal Time for Station # 2 with Adhesive Sealant 3.5 5.5 6.0
Station 3 – Final Assembly and Test

9. Transfer pipe to assembly station 0.5 0.5 0.5
10. Align and stab spigot into bell 0.5 0.5 0.5
11. Screw pipe until joint is made up. 0.5 0.5 0.5
12. Visual inspection of assembly 0.5 0.5 0.5
13. Optional Seal Check. Attach fittings and air check 2.0(1) 2.0(1) 2.0(1)
14. Injection of Sealant if configuration 3 is chosen 1.5(1) 2.5(1) 3.5(1)
Subtotal Time for Station # 3 w/o Seal Check and 2.0 2.0 2.0
Adhesive Sealant
Subtotal Time for Station # 3 with Seal Check 4.0 4.5 5.5
Subtotal Time for Station # 3 with Seal Check and 4.5 7.0 9.0
Adhesive Sealant:
1. Optional Procedures
33 ©Ameron 2003 - FP914 03/03

Table 8 attempts to show the potential savings during installation of SSL. The faster joint
assembly time can save in overall installed cost. The values shown are maximum possible lay
rates with no problems for wind, waves, etc. Savings can be much more in poor conditions.
The table is only indicative and does not include barge mobilization time, connection to risers,
etc. Level of automation in the barge and barge rental rates will vary, affecting the outcome.
Table 8 - Relative Lay Rates and Savings for SSL vs. Carbon Steel
Lay Method Max Lay Rate Max Lay Rate Barge Cost Potential
for Steel (1) for SSL (2) Savings for 10-
km Line
S-Lay 3,456 m/day 5,069 m/day $350k/day 1 day @ $350k
J-Lay (3) 1,728 m/day 2,534 m/dau $350k/day 2 days @ $350k
(1) Based on a cycle time of 8 minutes and an overall efficiency of 80% when laying pre-assembled double
random length steel pipe (24-meter assembled length).
(2) Based on a cycle time of 4 minutes and an overall efficiency of 80% when laying pre-assembled double
random length SSL pipe (22-meter assembled length).
(3) Assumes laying single random pipe when using J-Lay. Double and even quad length pipe can be installed if
the barge has a tall enough derrick.
6. SSL SUBSEA PIPELINE ANALYSIS USING FEA

PROGRAMS
6.1 INTRODUCTION
It is possible to utilize existing FEA analysis programs, such as CAESAR II®, for designing
SSL subsea pipelines. This is accomplished by inputting equivalent elastic properties for the
overall SSL pipe wall and then doing a conventional analysis. The results are then compared to
an allowable design envelope that defines all allowable axial and hoop stress combinations. If
all the calculated values are inside the allowable design envelope, the design is acceptable.
A proprietary MathCAD® workbook is available from Ameron to assist the pipeline engineer in
properly analyzing and selecting SSL for a subsea pipeline. Additional advantages obtained in
utilizing the proprietary MathCAD program includes analysis of other loading conditions such
as earth reaction loads, buoyancy, external collapse, vortex induced vibration (VIV) and flow
analysis. See Section 7 for details on Ameron supplied MathCAD workbooks.
Section 6.2 provides the definitions of design factors f1, f2 and f3 plus de-rating factors A1, A2
and A3 and how they are used in determining the allowable SSL pipeline design stresses per
ISO 14692. Section 6.4 provides a general outline of the FEA analysis method for SSL
pipelines. Section 6.3 provides the background for the origin of the SSL design envelope for
reference purposes. Section 6.5 provides information on the limitations in existing FEA
analysis programs, specifically CAESAR II®. Finally, Section 6.6 provides elastic and other
properties of SSL pipe that are necessary inputs for CAESAR II® and other FEA analysis
programs.
34 ©Ameron 2003 - FP914 03/03

6.2 DEFINITION OF PARTIAL FACTORS F1, F2, F3, A1, A2 AND A3

SSL is a composite structure made up of both GRE and steel layers. As such, SSL is designed
like composite pipe and utilizes data for the GRE component in determining the limiting design
stress. Many pipeline engineers are not familiar with the concepts utilized in ISO 14692(8) in
determining the allowable design stresses. This section provides background into the safety
and de-rating factors utilized in the determining the allowable design stresses. See Section 6.3
for more background into the sizing and shape of the long-term design envelope.
1. The partial factor f1 is the safety factor between the nominal regression line for predicting
long-term hydrostatic stress (LTHS) and the lower confidence limit (LCL). The
relationship is as follows:
LCL = f1*LTHS
The default value for f1 is 0.85. However, if data exists for the LCL as defined by testing
per ASTM D2992(9) or ISO 14692, that value must be utilized in place of the default value.
2. The partial factor f2 for GRE pipe, or the GRE component of SSL, is the safety factor
between the long-term LCL failure stress (after 20 years of service) and the long-term
design stress. The relationship is as follows:
Long-Term Design Stress = f1*f2*LTHS = f2*LCL
The value for part factor f2 is equal to 0.67 for sustained loads, such as pressure or
unsupported span. The corresponding concept for steel pipe in ASME B31.4(3) is design
stress at 72% of the Specified Minimum Yield Strength (SMYS). The part factor f2
changes, depending on the type of load. For loads that are self-limiting, such as thermal
expansion and installed bend radius, the partial factor f2 is increased to 0.83. Again, a
similar concept exists in ASME B31.4(3), where thermal expansion stresses are designed
up to 90% of SMYS (Design Stress = 0.90*SMYS). For intermittent and cyclic loads such
as surge pressure, field proof testing with water, etc. ISO 14692(8) allows f2 to be as high as
1.0 but limits the total number of “cycles” to 7000 during the anticipated life of the
pipeline.
3. The partial factor f3 is really a design factor utilized by CAESAR II® and UKOOA to
define the long-term allowable hoop stress as a function of combined axial stress from
pressure and bending. In GRE or SSL, the allowable axial stress is a function of the
applied hoop stress and vice versa. See Section 6.3 and Appendix 10.3 for a detailed
explanation of how CAESAR II® and UKOOA define this factor in predicting the
allowable long-term hoop stress. The relationship is a follows:
Long-Term Design Stress (with bending) = f1*f2*f3*LTHS = f2* f3*LCL
The pipeline engineer does not have to calculate factor f3 or input a value for it into the
analysis when using CAESAR II® or Ameron’s proprietary MathCAD analysis. CAESAR
II® will calculate the factor f3 automatically, based on combined bending and axial induced
thrust loads to determine the allowable “code” stresses when using the UKOOA code.
35 ©Ameron 2003 - FP914 03/03

This will not happen in CAESAR II® when selecting the other available codes such as
ASME B31.4. (3) Ameron’s MathCAD programs calculate combined axial stress based on
the allowable long-term GRE design envelope, thus eliminating the need for the design
factor f3. See Section 6.6.2 and Appendix 10.3 for more details.
4. A1 is the partial factor for temperature de-rating. See Section 3.1.3. The value for A1 is
the ratio of the allowable design stress at the specified operating temperature to the value
for allowable design stress at a reference temperature, in this case 21°C. The allowable
stress at the specified operating temperature is extrapolated linearly between known values
for design stress determined by regression testing using ASTM D2992-96 for the testing
protocol and analyzed per ISO 14692, Annex K.
5. A2 is the partial factor for chemical resistance. For oil field service, this value is usually
considered 1.0.
6. A3 is the partial factor for cyclic performance. A3 is calculated using the equations in the
Dec 98 Shell DEP and not the equations in ISO 14692. The reason for this is the equations
in ISO do not correlate with known test data and are known to need revision, while the Dec
98 Shell DEP values are in reasonable agreement with the known long term cyclic test
data. The cyclic rating is known to be a function of the operating temperature, the number
of cycles and the ratio of the minimum to maximum cycle pressure (Rcyclic). The first step
in calculating A3 is to calculate the ratio of static to cyclic performance. The ratio is
calculated using the nominal value for static regression at 100,000 hours to the nominal
value for cyclic regression at 150,000,000 cycles. Both data sets must be tested at the
same temperature. These values are chosen because they both represent exposure of
100,000 hours (i.e. it takes 100,000 hours to reach 150,000,000 cycles at the specified
cycling rate in ASTM D2992). This is done so the degradation due to environmental
effects (i.e. temperature and chemical exposure during the test) for both sets of test data are
treated equally and the additional degradation for cycling can be separated out.
σStatic100000 := 158.29MPa
⋅ These are projected stress values at 100000 hours and 150M
cycles respectively, used to determine the factor for calculating
σCyclic150M := 52.44⋅ MPa the LTHScyclic in Section 6.
σStatic100000
fc :=
σCyclic150M
fc = 3.018
Using the value for fc, the value for LTHScyclic at 20 years and 150,000,000 cycles can
now be calculated as follows:
LTHSstatic
LTHScyclic :=
fc
LTHScyclic = 35.622MPa
36 ©Ameron 2003 - FP914 03/03

The value for LTHSmixed can now be calculated as follows:
Pmin
Rcyclic := Rcyclic = 0.405 Cyclic pressure ratio as defined is Shell DEP98.
Pmax
LTHSmixed := LTHScyclic if Rcyclic ≤ 0.4

0.5
LTHScyclic2⋅  1 − Rcyclic2  + LTHSstatic 2⋅ Rcyclic2 if 0.4 < Rcyclic < 0.8
   
LTHSstatic otherwise
LTHSmixed = 54.066MPa Long-Term Hydrostatic Strength for cyclic-static service

after 150 million cycles
The LTHScyclic is determined from 150 million cycles, thus the de-rating factor must be
determined at the specified number of cycles. The equation for the cyclic degradation
line shown in Figure 31 is determined from the LTHSmixed and LTHSstatic. The
equation of a line for the de-rating factor will take the form of y = m*log(x) + b, where
the values for b and m can be calculated as follows:
 LTHSmixed    LTHSmixed  
 LTHS − 1   − 1 
   LTHSstatic   ⋅ log( 7000)
mmixed :=
static
b mixed := 1 − 
(
log 150⋅ 10
6 ) − log( 7000)  ( 6 ) 
 log 150⋅ 10 − log( 7000) 
mmixed = −0.07 b mixed = 1.268
Finally, the cyclic de-rating factor is calculated using the following equation:
A 3 := 1.000 if NCycles < 7000 O
otherwise O
((mmixed⋅ log( NCycles ) + bmixed)) (
if (mmixed⋅ log ( NCycles ) + b mixed) > mcyclic⋅ log 150⋅ 10
6 ) ...
+ b cyclic
( )
 mcyclic⋅ log 150⋅ 106 + b cyclic otherwise
 
In the above example and for 8,000,000 cycles and Rcyclic = 0.405, A3 is:
A 3 = 0.651
Note that no de-rating factor is required for less than 7000 cycles and the de-rating factor
is not reduced beyond the minimum value for full cyclic rating of 1/fc, reflecting the
asymptotic nature demonstrated in the test data. The MathCAD workbook automatically
provides a graph of the cyclic de-rating factor as shown in Figure 31.
37 ©Ameron 2003 - FP914 03/03

Cyclic Derating Factor

1.1
0.8
StaticCurveY
Derating Factor, A3
OpPtY
0.6
f ( x)
h( x)
0.4
0.2
0
0
3 4 5 6 7 8 9 10
1 .10 1 .10 1 .10 1 .10 1 .10 1 .10 1 .10 1 .10
1000 StaticCurveX , OpPtX , x, x 10
10
Cycles
0.8 < R < 1
Operating Point
0 < R < 0.4
R
Figure 31 - Typical Graph Showing the Cyclic De-rating Factor A3, at the Specified
Number or Cycles and Ratio of Maximum and Minimum Cyclic Operating Pressures
7. Finally, the allowable long-term design stress at R=2 can be calculated by:
Allowable Long-Term Design Stress = f1*f2*A1*A2*A3*LTHS = f2* A1*A2*A3*LCL
This value is used to size the allowable design envelopes (at R=2) in the MathCAD
workbooks. The factor f3 is not included, since the shape of the allowable design envelope
automatically accounts for this factor. Please note the allowable design envelopes in
Appendix 10.1 are based on the following equation:
Allowable Long-Term Design Stress = f1*f2*A1*LTHS = f2* A1*LCL
The part factor f3 is not included in the above equation because it is taken care of by the
shape of the allowable design envelope. The part factors f2 and A1 are clearly marked on
the allowable design graphs in Appendix 10.1. If additional de-rating is required for part
factors A2 and A3, the pipeline engineer must calculate these factors separately as noted
above and superimpose them on the allowable design stress envelope.
38 ©Ameron 2003 - FP914 03/03

6.3 EQUIVALENT ALLOWABLE DESIGN STRESS ENVELOPE
6.3.1 Introduction
The sections below describe the method for determining the “equivalent allowable design stress
envelope” for the overall SSL pipe wall. This “equivalent allowable design stress envelope” is
required to compare calculated FEA output stress to allowable values. The “size” of the
“equivalent allowable design envelope” is set to ensure the GRE stress, within the SSL pipe
wall, is within its allowable long-term design stress limit for the GRE component of SSL. This
is accomplished by determining the average overall pipe wall stress (steel and GRE) at the time
the GRE stress is at the maximum allowable long-term value for the specific ratio of GRE hoop
to GRE axial stress (R-ratio). Because of the influence of the GRE component, the resulting
long-term hoop stress for the overall SSL pipe wall is a function of applied axial stress and visa
versa.
6.3.2 Design Envelope for the GRE Component of SSL

The long-term design envelope for the GRE component of SSL is developed from the
following test data:
1. Long-term regression testing per ASTM D2992(9) and analyzed per ISO 14692, Annex K, at
rated operating temperature.
2. The shape of the short-term failure envelope as defined in Ameron Technical Bulletin
“Failure Characterization of Bondstrand Glass Fiber Reinforced Epoxy Pipe”.
ISO 14692(8) defines the long-term design stress envelope for GRE pipe as similar in shape to
the short-term failure envelope. This relationship, for the GRE component of SSL, is shown in
Figure 32. Figure 33 gives the long-term design stress from regression testing at 66°C and
Figure 34 shows how this long-term design stress is utilized to size the GRE design envelope.
Figure 61 to Figure 64 in Appendix 10.1 give the allowable design stress envelopes per ISO
14692, for the GRE component of SSL in SI and English units at 21°C (70°F), 66°C (150°F)
and 93°C (200°F) for part factors f2=0.67 and f2=0.83. Combined loads on the SSL pipeline
must result in GRE stresses that are inside these design envelopes at the specified operating
temperature.
6.3.3 Equivalent Design Envelope for the Overall SSL Pipe Wall
In order to utilize existing FEA programs, the limiting GRE stress defined above must be
converted to an “equivalent overall wall stress” for both the GRE and steel components. An
example of this type of graph is given in Figure 35. For reference and design purposes, Figure
65 to Figure 76 in Appendix 10.1 give design envelopes for the equivalent allowable design
stress for the overall SSL pipe wall in SI and English units at 21°C (70°F), 66°C (150°F) and
93°C (200°F) for part factors f2=0.67 and f2=0.83. Figure 65 to Figure 76 are a convenient way
to treat the individual steel and GRE stresses in the SSL pipe wall, by converting them to an
equivalent, homogeneous material. This way, conventional FEA analysis programs can be
utilized to calculate the allowable pipeline stresses. Additional graphs are required for other
operating temperatures, in order to properly include the temperature de-rating factor and the
39 ©Ameron 2003 - FP914 03/03

changes in elastic properties of the GRE. Consult your Ameron representative if you need
these graphs for your specific operating conditions.
Finally, the ISO 14692(8) standard defines a simplified long-term trapezoidal design envelope.
The long-term trapezoidal design envelope, including the part factor f2, can be defined by the
coordinates given in Table 9 or Table 10. The trapezoidal design envelope is a simplified
alternate design envelope compared to those presented in Appendix 10.1. Please note, the
simplified trapezoidal design envelope does not include all allowable design stress
combinations given in Figure 65 to Figure 76. However, these coordinates may be useful as
inputs to CAESAR II® or other programs using the UKOOA design code. The coordinates
given in Table 9 or Table 10 can also be utilized, by pipeline engineers, to manually or
automatically plot the FEA outputs against this simplified design envelope. An example of the
plot for the simplified trapezoidal design envelope is given in Figure 36.
Please Note: The MathCAD® workbook, for analyzing an SSL subsea pipeline, automatically
assures the GRE component of SSL has resulting stresses that land inside the long-term design
envelope. See Section 7.4.
FIBERGLASS DESIGN ENVELOPE AT VARIOUS TEMPERATURES

35,000
3400 Series Pipe Failure Envelope at 21°C
SSL Pipe Fiberglass Design Envelope at 21°C
30,000 SSL Pipe Fiberglass Design Envelope at 66°C
SSL Pipe Fiberglass Design Envelope at 93°C
Fiberglass Axial Stress, psi
25,000
20,000
15,000
10,000
5,000
0
0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000
Fiberglass Hoop Stress, psi
Figure 32 - Relationship Between the Short-Term Failure Stress and the Long-Term
Design Stress of the GRE Component of SSL
40 ©Ameron 2003 - FP914 03/03

Least Square Calculation for Determining the Long-Term Hydrostatic Pressure

ISO 14692-2:2002 (Annex K)
100-mm 3400 Series Pipe at 66°C
Data Time Stress Regression Results: Log Stress = 2.6046 -0.0810 x Log Time
Point hours MPa Log Time = 32.143 -12.341 x Log Stress
1 17.7 351.67 R= -0.880
2 333.0 253.35 Projected Stress, MPa Time, hours 95% Lower Confidence Limit, MPa
3 561.0 209.25 484.90 0.1 371.72
4 876.0 216.11 277.05 100 218.49
API LTHS Design Stress

5 1,218.0 241.02 229.89 1,000 181.66
HDB Design Stress

6 1,644.0 230.49 190.76 10,000 150.44
7 2,051.0 263.24 158.29 100,000 124.09
8 2,062.0 269.58 151.26 175,200 118.34
9 2,105.0 207.17 140.43 438,000 109.463
10 2,116.0 236.35 131.35 1,000,000 101.988
11 2,226.0 195.67 79.145 Design Stress 79.289
12 2,546.0 226.59
13 3,337.0 225.79 1000
14 5,291.0 189.88
15 6,113.0 216.60
16 6,552.0 199.29
17 7,420.0 174.09
18 19,868.0 167.80
19 12,319.0 172.64
20 1.7 348.19 Stress, MPa
21 0.7 350.30 100
22
23
24
25
26 Data Points
27 Projected Stress
28 95% Lower Confidense Limit
29
10
30
1.00E-01 1.00E+00 1.00E+01 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06
31
Time, hours
32
Figure 33 -Regression Analysis of the GRE Component of SSL Pipe at 66°°C

(with part factor of 0.67)
70
3400 Series Pipe Design Envelope at 21°C
60
Design Stress for Hoop-Loaded 3-Strip SSL Pipe at 21°C
Design Stress for 3400 Series Pipe at 66°C
Design Stress for Bidirectional-Load 4-Strip SSL Pipe at 93°C
50
Fiberglass Axial Stress, MPa
Hoop Stress = 79.3 MPa

Axial Stress = 39.7 MPa

40
30
20
10 Hoop Stress = 127.5 MPa

77.2 MPa 91.2 MPa 106.9 MPa
0
0 15 30 45 60 75 90 105 120 135 150 165
Fiberglass Hoop Stress, MPa
Figure 34 - GRE Design Envelope at f2=0.67
41 ©Ameron 2003 - FP914 03/03

Design Hoop Stress vs. Axial Stress for SSL Pipe as A Pressure Vessel at 21 °C
(with part factor = 0.67 )
210
Total Axial Stress as Pressure Vessel, MPa

180
150
120
90
60
3 Steel Strips 4 Steel Strips

30
0
0 60 120 180 240 300 360 420
SSL Composite Hoop Stress, MPa
Figure 35 - Equivalent SSL Design Envelope at f2=0.67 and Operating Temperature at

21°°C
200
180
160
140
Axial Stress MPa
120
100
80
SSL 3 SSL 4
60
SSL 5 SSL 6
40 SSL 7 SSL 8
SSL 9 SSL 10
20
0
0 50 100 150 200 250 300 350 400
Hoop Stre ss MPa
Figure 36 - Equivalent SSL Trapezoidal Design Envelope at f2=0.67 and Operating

Temperature at 21°°C
42 ©Ameron 2003 - FP914 03/03

Table 9 - Coordinates for Long-Term Trapezoidal SSL Design Envelope in English Units
70ºF 150ºF 200ºF

Axial Only Loading 2:1 Loading Axial Only Loading 2:1 Loading Axial Only Loading 2:1 Loading
Hoop Axial Hoop Axial Hoop Axial Hoop Axial Hoop Axial Hoop Axial
Stress Coordinates, psi, with Part Factor = 0.67
SSL 3 0 9370 35705 17695 0 8712 29319 14530 0 8129 24883 12332
SSL 4 0 10789 40419 20015 0 10076 33013 16392 0 9445 28083 13907
SSL 5 0 11984 44387 21962 0 11224 36290 17956 0 10552 30777 15228
SSL 6 0 13002 47775 23619 0 12203 39012 19287 0 11495 33078 16353
SSL 7 0 13879 50701 25046 0 13046 41364 20433 0 12308 35066 17322
SSL 8 0 14643 53255 26286 0 13779 43417 21430 0 13015 36801 18164
SSL 9 0 15313 55504 27374 0 14423 45225 22304 0 13636 38329 18903
SSL 10 0 15905 57499 28334 0 14993 46830 23077 0 14184 39685 19556
SSL 3 0 11607 39144 19400 0 10792 36321 18000 0 10070 30826 15277
SSL 4 0 13366 49735 24628 0 12483 41099 20307 0 11701 34790 17228
SSL 5 0 14845 54987 27207 0 13905 44957 22244 0 13072 38127 18865
SSL 6 0 16106 59184 29259 0 15116 48328 23893 0 14240 40977 20258
SSL 7 0 17193 62809 31027 0 16161 51242 25313 0 15247 43440 21459
SSL 8 0 18139 65973 32563 0 17070 53786 26547 0 16123 45590 22502
SSL 9 0 18970 68759 33910 0 17868 56025 27631 0 16892 47483 23417
SSL 10 0 19704 71230 35101 0 18573 58013 28588 0 17572 49162 24226
Note: The values in Green font are calculated from the other value in the pairs based on R ratio of 2:1.
Table 10 - Coordinates for Long-Term Trapezoidal SSL Design Envelope in SI Units
21ºC 66ºC 93ºC

Axial Only Loading 2:1 Loading Axial Only Loading 2:1 Loading Axial Only Loading 2:1 Loading
Hoop Axial Hoop Axial Hoop Axial Hoop Axial Hoop Axial Hoop Axial
Stress Coordinates, MPa, with Part Factor = 0.67
SSL 3 0 65 246 122 0 60 202 100 0 56 172 85
SSL 4 0 74 279 138 0 69 228 113 0 65 194 96
SSL 5 0 83 306 151 0 77 250 124 0 73 212 105
SSL 6 0 90 329 163 0 84 269 133 0 79 228 113
SSL 7 0 96 350 173 0 90 285 141 0 85 242 119
SSL 8 0 101 367 181 0 95 299 148 0 90 254 125
SSL 9 0 106 383 189 0 99 312 154 0 94 264 130
SSL 10 0 110 396 195 0 103 323 159 0 98 274 135
Stress Coordinates, MPa, with Part Factor = 0.83
SSL 3 0 80 270 134 0 74 250 124 0 69 213 105
SSL 4 0 92 343 170 0 86 283 140 0 81 240 119
SSL 5 0 102 379 188 0 96 310 153 0 90 263 130
SSL 6 0 111 408 202 0 104 333 165 0 98 283 140
SSL 7 0 119 433 214 0 111 353 175 0 105 300 148
SSL 8 0 125 455 225 0 118 371 183 0 111 314 155
SSL 9 0 131 474 234 0 123 386 191 0 116 327 161
SSL 10 0 136 491 242 0 128 400 197 0 121 339 167
Note: The values in Green font are calculated from the other value in the pairs based on R ratio of 2:1.
43 ©Ameron 2003 - FP914 03/03

6.4 GENERAL OUTLINE OF THE FEA ANALYSIS METHOD FOR AN

SSL PIPELINE
A typical FEA analysis is accomplished as follows:
1. For subsea pipelines, look up the pressure rating in Table 1 and multiply it times the
temperature de-rating factor in Table 2 to determine the minimum SSL pipe wall that will
meet the required pressure rating at the specified operating temperature. Note that the
required pressure rating is the net pressure from internal pressure less the minimum external
pressure at the minimum water depth. For subsea lines where the pipe is laying on the
seabed and is not axial constrained, one or two additional steel strips are usually required
over the minimum pressure rating to handle the additional axial stresses caused by installed
bend radius or other externally applied axial loads. Additional checks are required to
satisfy DNV-OS-F101 as well. See Section 7.4 for more details.
2. Next convert the elastic response of the overall SSL pipe wall to equivalent elastic
properties at the specified operating temperature. Values are presented in Section 6.6;
Table 11, Table 12 and Table 13 for 21°C, 66°C and 93°C respectively. A MathCAD®
program, “Equivalent Elastic Properties” is available for automatically calculating elastic
properties at temperatures in between those presented in Table 11 to Table 13 or for special
SSL pipe wall constructions. Please note, the elastic properties are a function of both the
number of steel layers and the operating temperature, so if you need values between those
presented in Table 11 to Table 13, they will have to be calculated or extrapolated between
values in the tables.
3. Look up other related inputs from Table 11 to Table 13, such as thermal coefficient of
expansion, pipe wall density, etc, that may be necessary for inputs to the CAESAR II® or
other FEA analysis program. As an alternate, the MathCAD® program “Equivalent Elastic
Properties” will do this for you automatically for temperatures not presented in Table 11 to
Table 13.
4. Input the elastic properties and other required inputs into the CAESAR II® program and
calculate the stresses in the pipeline for a pipe installed at ambient and operated at the
specified operating temperature, using the ASME B31.4(3) or ASME B31.8(4) design code.
5. Manually add axial bending stresses from the installed bend radius to the calculated axial
stress from pressure and from bending. See Section 6.5 for a detailed explanation why the
bending stress from “installed bend radius” must be manually added and compared to the
allowable design envelope. The stress from pressure is titled “Axial Stress” and the stress
from unsupported span and bending in the expansion loop is titled “Bending Stress” in the
stress report from CAESAR II®. The bending stress from “installed bend radius” can be
calculated from the following equations:
OD
2
ε BendRadius
BendRadius
σ SSLBendRadius E SSLAxial ⋅ ε BendRadius
Total axial stress can now be calculated from the CAESAR II® stress report as follows:
σ AxialTotal σ Axial + σ Bending + σ SSLBendRadius
44 ©Ameron 2003 - FP914 03/03

6. Finally compare the total axial stress (as calculated in item 5 above) and hoop stress
combinations at each node to the graphs in Appendix 10.1, using the appropriate operating
temperature and part factor f2 (f2=0.67 for sustained loads of pressure and unsupported span
and f2=0.83 for sustained loads plus self limiting loads such as thermal growth or installed
bend radius). See Section 6.3 for a definition of the origin and design basis of the
“Equivalent Allowable Design Stress Envelope”. Do not use the values reported for Von
Mises stress or overall calculated stress for the above comparison to the equivalent
allowable design envelope. The reason is the design envelope is plotted using principle
stress in the hoop and axial direction and the overall calculated stress may be design code
driven such as using a 2-D Mohr’s circle in the British Standard. Von Mises stress is a
strain energy criteria for materials with ductility above 5% that does not apply to GRE or
SSL materials. Therefore the combined stress and Von Mises stress will be meaningless
when comparing to the equivalent allowable design envelopes in Appendix 10.1. If the
stresses at all nodes land inside the envelope for the initially chosen pipe (say SSL5), the
analysis is complete and the selected pipe is capable of handling all the loading conditions.
If any of the stresses at the nodes lays outside the specified design envelope, repeat the
process with the next higher number of steel strips in the SSL pipe wall until all stress
combinations, at all nodes, lay within the envelope for the selected pipe.
7. The entire procedure, outlined above, must be repeated twice. The first analysis uses a part
factor f2 set equal to 0.67, with the pipeline installed and operated at elevated temperature.
Flexural bending stresses are not added in this analysis. The second analysis, utilizes
design envelopes with the part factor f2 = 0.83. In this analysis, the pipeline will be
installed at ambient temperature and operated at elevated temperature. Bending stress must
first be added to calculated axial stress before comparing to the allowable stress envelopes
in Appendix 10.1. See Appendix 10.3 for guidelines in using CAESAR II® FEA analysis.
6.5 LIMITATIONS IN EXISTING FEA ANALYSIS SOFTWARE

A number of limitations exist in the design methods utilized in CAESAR II® and other pipeline
FEA analysis programs, as listed below:
1. Installed Bend Radius. All the current design codes assume no additional stress in the
pipeline from field bends, if the node positions are input directly from the line lay. It is
assumed that a steel pipeline is bent by yielding during installation and therefore, no
additional stress from flex bending is applied during operation. This assumption is not
valid for GRE or SSL pipelines. The GRE component of SSL is a limited ductility
material and must be designed within the allowable long-term design envelope as shown in
Figure 61 to Figure 64 of Appendix 10.1. Installed bend radius, from flexing the pipeline,
must be included in the analysis of combined stresses. The installed bend radius can be
included in the CAESAR II® analysis of stress by first installing the pipe in a straight line
(layout of nodes) and then flexing the pipeline to the line lay position (displace nodes to
the final installed pipeline lay position). This is a difficult and time-consuming task and is
generally not recommended (i.e. it is easier to add the axial stress from installed bend
radius manually to the FEA output before comparing to the long-term design envelope).
2. Part Factor f2. Two separate runs of the CAESAR II® pipeline analysis program must be
made, in order to compare actual stress to allowable, using the two different part factors for
the different loading conditions. In the first run, the pipe is assumed to be installed with no
45 ©Ameron 2003 - FP914 03/03

bending stress from installed bend radius, at elevated temperature and operated at elevated
temperature. This will determine the loads for a part factor f2=0.67. For the second run,
the part factor f2 is set to 0.83 and the pipe is installed in a bent condition, at ambient and
then allowed to grow thermally, as the temperature increases to operating conditions. This
way, the thermal induced and installed bend radius stresses are compared against the
higher allowable design stress, resulting from 0.83*LCL. See Section 6.2.
3. UKOOA. Use of the UKOOA standard, within CAESAR II®, is recommended for above
ground pipelines but should not be used for subsea pipelines where compression stress is
possible. The UKOOA standard was developed for above ground piping systems and does
not include the possibility of axial compression.
4. ASME B31.4(3) and ASME B31.8(4). ASME standards calculate the principle stresses in
the hoop and axial direction, and are therefore useful in calculating the stress in SSL pipe.
Calculated values can then be manually compared to the allowable stress at the actual R-
ratio. The use of ASME B31.4(3) and B31.8(4) is recommended for analyzing subsea
pipelines where axial compression is a possibility. This is because the ASME code is not
limited to positive axial stress, as is the UKOOA standard. The ASME B31.4(3) design
code analysis does not include axial stress from installed bend radius, as noted above. If
there is an installed bend radius required to conform to terrain, the axial stress from
installed bend radius must first be manually added to calculated values at each node, when
checking the design using a part factor f2 = 0.83. The combined stress must then be
manually checked against the allowable stress, based on R-ratio as defined in Figure 65 to
Figure 76. Do not use the “allowable stress” column in the stress report as this value does
not utilize the trapezoidal design envelope and will give misleading results. See Section
6.3 and 10.3 for more details. The pipeline engineer can also “flex” the pipeline into the
line lay position for the nodes as noted in item 1 above. Using this technique, the pipeline
design engineer will not have to manually add the bending stress from installed bend
radius as this will be automatically calculated by CAESAR II®.
5. British Standard 7159. Ameron does not recommend the use of the British Standard.
The “Allowable Stress” column of the stress report in the British standard calculates a
“combined stress” utilizing a 2-D Mohr’s circle to determine maximum combined stress.
This approach is useless in calculating allowable stresses for GRE or SSL and will give
misleading results.
6.6 EQUIVALENT ELASTIC PROPERTIES OF SSL
6.6.1 Equivalent Modulus, Poisson’s Ratio, Thermal Coefficient of

Expansion and Density for the SSL Pipe Wall
Table 11 to Table 13 give equivalent elastic properties for SSL pipe wall at 21°C (70°F), 66°C
(150°F) and 93°C (200°F). A MathCAD® program titled “Equivalent Elastic Properties” is
available for calculating these values at other temperatures, or for custom pipe designs. If you
need these values or the MathCAD workbooks, contact your Ameron representative.
46 ©Ameron 2003 - FP914 03/03

Table 11 - Equivalent Elastic Properties of the SSL Pipe Wall at 21°C (70°F)
Description SSL3 SSL4 SSL5 SSL6 SSL7 SSL8 SSL9 SSL10

Poisson’s Ratio
0.344 0.335 0.329 0.325 0.322 0.319 0.317 0.316
(Axial to Hoop)
Poisson’s Ratio
0.309 0.307 0.306 0.306 0.305 0.305 0.304 0.304
(Hoop to Axial)
Axial Modulus
9.623 11.20 12.55 13.72 14.74 15.64 16.45 17.17
(106 psi)
Axial Modulus
6.635 7.721 8.651 9.458 10.16 10.79 11.34 11.83
(1010 MPa)
Hoop Modulus
10.73 12.21 13.48 14.58 15.55 16.40 17.15 17.83
(106 psi)
Hoop Modulus
7.398 8.420 9.295 10.05 10.72 11.30 11.83 12.29
(1010 MPa)
Thermal
Coefficient of
7.184 7.022 6.915 6.840 6.783 6.740 6.705 6.677
Expansion
(10-6 in/in°F)
Thermal
Coefficient of
1.293 1.264 1.245 1.231 1.221 1.213 1.207 1.202
Expansion
(10-5 mm/mm°C)
Wall Thickness
0.236 0.256 0.276 0.295 0.315 0.335 0.354 0.374
(in)
Wall Thickness
6.000 6.500 7.000 7.500 8.000 8.500 9.000 9.500
(mm)
Density (lb/in3) 0.123 0.135 0.146 0.155 0.163 0.170 0.177 0.182
3
Density (10
3.399 3.744 4.039 4.296 4.520 4.718 4.894 5.051
kg/m3)
EAxial
⋅ ν AxialHoop 0.309(1) 0.307(1) 0.306(1) 0.306(1) 0.305(1) 0.305(1) 0.304(1) 0.304(1)
EHoop
SHi (2) (psi) 52796 59209 64617 69240 73238 76729 79804 82534
SHi (2) (MPa) 364 408 446 477 505 529 550 569
Ri (3) 0.532 0.542 0.549 0.554 0.557 0.560 0.563 0.565

1. This value is required for the CAESAR II® software input table. Please note that Caesar II uses reversed
subscript convention as opposed to the convention utilized by Ameron in the SSL literature and other FEA
analysis software packages.
2. This value is required by CAESAR II® to define the hoop stress from f1*LTHS or LCL
3. This value is required by CAESAR II® and is the ratio of the axial stress at zero hoop stress to the axial stress
at a ratio of 2:1 for hoop to axial stress in the long-term design or failure envelope.
47 ©Ameron 2003 - FP914 03/03

Table 12 - Equivalent Elastic Properties of the SSL Pipe Wall at 66°°C (150°°F)

Poisson’s Ratio
0.353 0.342 0.335 0.330 0.326 0.323 0.321 0.319
(Axial to Hoop)
Poisson’s Ratio
0.318 0.315 0.313 0.311 0.310 0.309 0.308 0.308
(Hoop to Axial)
Axial Modulus
9.408 11.00 12.36 13.55 14.58 15.49 16.30 17.03
(106 psi)
Axial Modulus
6.486 7.584 8.524 9.339 10.05 10.68 11.24 11.74
(1010 MPa)
Hoop Modulus
10.45 11.95 13.24 14.35 15.33 16.19 16.96 17.65
(106 psi)
Hoop Modulus
7.202 8.238 9.127 9.897 10.57 11.17 11.69 12.17
(1010 MPa)
Thermal
Coefficient of
7.120 6.968 6.869 6.799 6.748 6.708 6.676 6.650
Expansion
(10-6 in/in°F)
Thermal
Coefficient of
1.282 1.254 1.236 1.224 1.215 1.207 1.202 1.197
Expansion
(10-5 mm/mm°C)
Wall Thickness
0.236 0.256 0.276 0.295 0.315 0.335 0.354 0.374
(in)
Wall Thickness
6.000 6.500 7.000 7.500 8.000 8.500 9.000 9.500
(mm)
Density (lb/in3) 0.123 0.135 0.146 0.155 0.163 0.170 0.177 0.182
Density (10-3
3.399 3.744 4.039 4.296 4.520 4.718 4.894 5.051
kg/cm3)
EAxial
⋅ ν AxialHoop 0.318(1) 0.315(1) 0.313(1) 0.311(1) 0.310(1) 0.309(1) 0.308(1) 0.308(1)
EHoop
SHi (2) (psi) 47085 52984 57958 62210 65887 69097 71925 74435
SHi (2) (MPa) 325 365 400 429 454 476 496 513
Ri (3) 0.602 0.617 0.627 0.635 0.641 0.645 0.649 0.652
®
1. This value is required for the CAESAR II software input table. Please note that Caesar II uses reversed
48 ©Ameron 2003 - FP914 03/03

Table 13 - Equivalent Elastic Properties of the SSL Pipe Wall at 93°C (200°F)

Poisson’s Ratio
0.360 0.347 0.339 0.333 0.329 0.326 0.323 0.321
(Axial to Hoop)
Poisson’s Ratio
0.321 0.317 0.315 0.313 0.312 0.310 0.310 0.309
(Hoop to Axial)
Axial Modulus
9.172 10.78 12.16 13.36 14.40 15.33 16.15 16.88
(106 psi)
Axial Modulus
6.324 7.433 8.385 9.209 9.930 10.57 11.13 11.64
(1010 MPa)
Hoop Modulus
10.27 11.79 13.09 14.22 15.20 16.07 16.85 17.54
(106 psi)
Hoop Modulus
7.084 8.130 9.026 9.803 10.48 11.08 11.62 12.09
(1010 MPa)
Thermal
Coefficient of
7.046 6.907 6.817 6.754 6.708 6.672 6.644 6.621
Expansion
(10-6 in/in°F)
Thermal
Coefficient of
1.268 1.243 1.227 1.216 1.207 1.201 1.196 1.192
Expansion
(10-5 mm/mm°C)
Wall Thickness
0.236 0.256 0.276 0.295 0.315 0.335 0.354 0.374
(in)
Wall Thickness
6.000 6.500 7.000 7.500 8.000 8.500 9.000 9.500
(mm)
Density (lb/in3) 0.123 0.135 0.146 0.155 0.163 0.170 0.177 0.182
Density (10-3
3.399 3.744 4.039 4.296 4.520 4.718 4.894 5.051
kg/cm3)
EAxial
⋅ ν AxialHoop 0.321(1) 0.317(1) 0.315(1) 0.313(1) 0.312(1) 0.310(1) 0.310(1) 0.309(1)
EHoop
SHi (2) (psi) 40470 45563 49855 53523 56694 59463 61902 64066
SHi (2) (MPa) 279 314 344 369 391 410 427 442
Ri (3) 0.661 0.681 0.695 0.704 0.712 0.718 0.722 0.726
® ®
1. This value is required for the CAESAR II software input table. Please note that CAESAR II uses reversed
49 ©Ameron 2003 - FP914 03/03

6.6.2 Equivalent “EI” Values for the SSL Pipe Wall

Because SSL is a composite structure combining material with vastly different elastic
properties, an equivalent “EI” must be calculated to apply standard engineering equations for
external loads. Table 14 summarizes typical “EI” values for the SSL pipe wall at 21, 66 and
93°C. “EI” for special pipe wall configurations and other than listed temperatures can be
calculated as described in Section 6.6.3.
Table 14 - Typical Long-Term EI Values (Hoop Direction) for the SSL Pipe Wall at
Various Temperatures
Number of Steel 21°C (70°F) 66°C (150°F) 93°C (200°F)
Strips N-mm x 105 N-mm x 105 N-mm x 105
(lb-in x 103) (lb-in x 103) (lb-in x 103)
6.071 5.613 5.337
SSL3
(5.373) (4.968) (4.724)
8.539 7.945 7.587
SSL4
(7.558) (7.032) (6.715)
11.82 11.08 10.63
SSL5
(10.462) (9.807) (9.408)
16.08 15.17 14.61
SSL6
(14.232) (13.427) (12.931)
21.49 20.38 19.71
SSL7
(19.020) (18.038) (17.445)
28.19 26.87 26.08
SSL8
(24.950) (23.782) (23.083)
36.34 34.81 33.88
SSL9
(32.164) (30.809) (29.986)
46.12 44.34 43.27
SSL10
(40.820) (39.244) (38.297)
6.6.3 Equations for Equivalent “EI” for the SSL Pipe Wall in the Hoop
Direction
SSL has unique capability in design for external collapse. The pipe wall can be designed for
the best combination of resistance to internal pressure and external collapse, by changing the
ratio of steel and GRE during manufacture. Since the collapse resistance is proportional to the
cube of the wall thickness, small changes in wall thickness make large changes in collapse
resistance. This design flexibility is not possible for steel pipelines. Consult your Ameron
representative for optimum design, based on required combinations of external and internal
pressure resistance.
Equivalent EI, for a custom pipe wall construction, is calculated by summing the EI values for
the GRE liner and jacket along with the steel. The basic equation is I + Area*Y2 for each
50 ©Ameron 2003 - FP914 03/03

element. The relative values of the Modulus E for steel versus GRE must be accounted for in
the determination of the individual I+Area*Y2 values and resulting centroid of the overall SSL
pipe wall. This is done, by adjusting the calculated area of the GRE layers by the ratio of the
modulus of the steel versus the modulus of the GRE. This makes the GRE area look smaller,
but gives the GRE element the same weighting in the centroid calculation as the steel by
adjusting it to the same modulus. See the “Area” calculations below. The “I” values are
determined assuming unit length of the pipe wall in the axial direction, so the resulting units are
in unit length rather than area.
Steel layers in SSL are bonded with a resin layer between each strip. For simplicity, the
thickness of steel layers is modified to include resin, and the modulus is modified by the theory
of mixtures as follows:
E Stl ⋅ T StlStrip ⋅ N
E StlMod
N ⋅ T StlStrip + ( N − 1 ) ⋅ T Resin
T StlM N ⋅ T StlStrip + ( N − 1 ) ⋅ T Resin
Where:
EStl Modulus of the steel
TStlStrip Thickness of the steel strip
N Number of steel layers
TResin Thickness of the resin layers between the steel strips
The Equivalent EI is calculated as follows:
Y
Section Area
EGREHoop YLiner := 0⋅ mm
GRE Liner A GREL := ⋅ TLiner
EStlMod
TLiner + TJacket
EGREHoop YJacket := + TStlM
GRE Jacket A GREJ := ⋅ TJacket 2
EStlMod
EStlMod  TLiner + TStlM 

Steel Wall A Stl := ⋅ TStlM YStl :=  
EStlMod  2 
The sum of the areas can be calculated as follows:
SUM Area := A GREL + A GREJ + A Stl
51 ©Ameron 2003 - FP914 03/03

The sum of the areas times the distance from the reference centroid can be calculated as
follows:
SUM YArea := A GREL⋅ Y Liner + A GREJ⋅ Y Jacket + A Stl ⋅ Y Stl
The distance from the centroid of the SSL pipe wall to the middle of the GRE liner (the
reference position for the initially assumed centroid) can be calculated as follows:
SUM YArea
Y SSL :=
SUM Area
The moment of inertia of the individual layers about their own neutral axis can be calculated
per unit of axial length as follows:
3 3 3
TLiner TJacket TStlM
ILin := IJac := ISteel :=
12 12 12
The sum of the EI value for all the layers is calculated as follows:
EILiner := EGREHoop⋅  ILin + TLiner⋅ (YSSL − YLiner) 

2
EIJacket := EGREHoop⋅  IJac + TJacket⋅ (YSSL − YJacket) 

2
EIStl := EStlMod ⋅  ISteel + TStlM ⋅ (YSSL − YStl ) 

2
EISSL := EILiner + EIJacket + EIStl
Correction for initial pipe stiffness must be included when compared to long-term values. This
is accounted for by a “shear lag factor” as follows:
ShearLagFactor := 0.85
EISSLLongTerm := EISSL ⋅ ShearLagFactor
52 ©Ameron 2003 - FP914 03/03

7. SSL SUBSEA PIPELINE ANALYSIS USING AMERON

PROPRIETARY MATHCAD® SOFTWARE
7.1 INTRODUCTION
This manual is focused on subsea pipelines and will therefore only present information on the
analysis of subsea SSL pipelines using proprietary Ameron MathCAD software.
Ameron’s MathCAD program for subsea pipelines is based on designing within the criteria
established in ISO 14692(8) for the GRE component of SSL and DNV-OS-F101 for both the
steel component of SSL and the overall subsea pipeline. The steel component of SSL also
conforms to API-5L(2) and the GRE component conforms to long-term allowable stresses
defined in API 15-HR and/or ISO 14692 under combined loading conditions. The GRE
component is checked against allowable hoop and axial stress as shown in Figure 61 to Figure
64. Stresses are calculated for combined loads, including pressure, unsupported span, soil
loading, vortex induced vibration (VIV), thermal expansion, and installed bend radius.
Other supporting calculations are preformed for system collapse, local buckling and installation
loads.
7.2 FLOW DIAGRAMS FOR STRUCTURAL DESIGN

Figure 37 represents a logic flow diagram for the required design checks and the associated
equations presented in DNV-OS-F101. This diagram applies to the Load Controlled Condition
(LCC). SSL is not a ductile material. The pipe wall cannot be loaded beyond yield without
damaging the composite, so SSL must be designed using the LCC diagram. Please note, the
DNV-OS-F101 standard states “when the Displacement Controlled Condition (DCC) cannot
be determined, it can be substituted by the LCC”. This is permissible because the LCC is more
conservative.
The logic diagram in Figure 37 has been modified to eliminate unnecessary steps that do not
apply to SSL. Figure 38 shows the modified logic diagram, utilized in the proprietary
MathCAD workbook. The workbook assumes the pipeline is laying on the seabed and is
unconstrained. Some partial axial constraint does exist along the length of the pipeline, due to
the weight of the pipe, so the calculated axial stresses will be conservative. A Finite Element
Analysis (FEA) program will be required for more accurate analysis of any partially
constrained area of the pipeline. Very long pipelines will be fully constrained in the middle.
The axial constraint reduces the axial stress applied to the pipe. The reduced stress will allow a
smaller installed bend radius without exceeding the allowable axial stress in service, so an FEA
analsys may be very useful in selecting a pipe with reduce pressure rating.
53 ©Ameron 2003 - FP914 03/03

Start
Criteria Not OK
Pressure Containment
Equation 5.14
OK
System Collapse
Equation 5.22
Local Buckling -
Combined Loading Displacement Controlled
Equation 5.23/24
Local Buckling – Combined Loading

Equation 5.25/26
έp < 0.3 % ECA

D1000
Ep > 2.0 %
Supplementary Requirement
P required
Propagating Buckling Consider Buckle

Equation 5.27 Arrestors
Fatigue Ovalisation
Ratcheting
Special Consideration
Stop
Figure 37 - Logic Diagram for Design of Subsea Pipelines per DNV-OS-F101
54 ©Ameron 2003 - FP914 03/03

Start
Pressure Containment
Section 6
System Collapse
Section 7
Local Buckling Combined

Loading
Section 7.1 to 7.3
Buckling Arrestors
@ Coil-Lock Joint
Section 7.4
Pipe Soil Interaction
Special Section 8
Considerations
Pressure + Unsupported Span

Section 9
Vortex Induced Vibration at

Unsupported Span
Section 10
Pressure + Installed Bend Radius

Section 11
Pressure + Installed Bend Radius +

Unsupported Span - Section 12
J-Lay and S-Lay Installation

Loads - Section B
Stop
Figure 38 - Logic Flow Diagram for Design of Subsea SSL Pipelines Laying on the Bottom
55 ©Ameron 2003 - FP914 03/03

7.3 INPUTS REQUIRED FOR THE MATHCAD ANALYSIS

The MathCAD® program requires the following manual inputs:
1. Product Dimensional Inputs

• ID Pipe inside diameter
• TLiner Thickness of the GRE liner
• N Number of steel strips in the pipe wall
• TStlStrip Thickness of the steel strip
• TJacket Thickness of the GRE jacket
• TWeight Thickness of the weight coating
• ρWeight Density of the weight coating
• LJoint Length of the GRE Coil-Lock joint on the SSL pipe
• Lpipe Overall length of the SSL pipe
2. Operating Conditions
• Pop Operating pressure
• PVacMin Minimum vacuum pressure
• Pmax Maximum cyclic pressure (usually the same as the operating pressure)
• Pmin Minimum cyclic pressure
• NCycles Number of cycles anticipated during the life of the pipeline
• TempInstall Installation temperature (ocean temperature)
• Temp Operating temperature
• BendRadV Installed bend radius, to follow local contour in the vertical plane
• BendRadH Installed bend radius in the horizontal plane to follow a “snake lay” or
the general pipeline path.
• ρContent Density of the fluid inside the pipe
3. Subsea Conditions
• Hmin Minimum depth of water at the installation location
• Hmax Maximum depth of water at the installation location
• g Gravitational constant at installation site
• ρSeawater Seawater density
• υk Seawater kinematic viscosity
• U100 Sea current velocity at pipeline unsupported span for the 100 year storm
• Span Anticipated span of unsupported pipe in the installed condition
• e Anticipated gap between the seabed and the pipeline at the unsupported
span
4. Installation Conditions
• Lp Touch down distance behind the barge when J-Laying
• FhBarge Horizontal force at the installation barge when using the J-Lay method
• LSting Stinger length at the S-Lay installation
• Φ Stinger setup angle (in deg.)
• Ntr True external axial force (from installation) at the lay barge when using
J-Lay
56 ©Ameron 2003 - FP914 03/03

5. Subsea Soil Conditions (See Tables 2.3.1, 2.3.2 and 2.3.3 from the MathCAD workbook for
guidance on the input values for subsea soil design factors)
• φSand Angle of friction for submerged sand
• γSand Submerged weight of sand
• υSand Poisson’s ratio for sand
• esSand Void content of sand
• RSpanShoulderSand Ratio of shoulder length to free span length for sand
• Su Undrained shear strength of clay
• γClay Submerged weight of clay
• υClay Poisson’s ratio for clay
• esClay Void content of clay
• RSpanShoulderClay Ratio of shoulder length to free span length for clay
• v Vertical penetration of the pipe in the sea soil (determined by
preliminary tests)
• K0 Coefficient of earth pressure at rest (usually 0.5)
The following tables can be used to estimate the soil properties for the pipe-soil interaction
calculations. These estimations shall be used only if soil test data is not available for the
actual installation conditions.
Table 2.3.1 Typical geotechnical parameters for sand

γsand
Soil Type φs 3 νs ess
(kN/m )
Loose 28-30 8.5-11.0 0.35 0.7-0.9
Medium 30-36 9.0-12.5 0.35 0.5-0.8
Dense 36-41 10.0-13.5 0.35 0.4-0.6
Table 2.3.2 Typical geotechnical parameters for clay

φs γclay
Soil Type νc esc
(kN/m2) (kN/m3)
Very soft <12.5 4-7 0.45 1.0-3.0
Soft 12.5-25 5-8 0.45 0.8-2.5
Firm 25-50 6-11 0.45 0.5-2.0
Stiff 50-100 7-12 0.45 0.4-1.7
Very stiff 100-200 10-13 0.45 0.3-0.9
Hard >200 10-13 0.45 0.3-0.9
57 ©Ameron 2003 - FP914 03/03

Table 2.3.3 provides guidance data for the support length ratio, where:
L is the length of the free span (also noted as Span in the Section 2.1 of this Mathcad
workbook;
Lsh is the shoulder length before the span
Table 2.3.3 Support length ratios
Ratios for Sand Ratios for Clay
Sand Type Lsh/L Clay Type Lsh/L
Loose 0.3 Very soft 0.5
Medium 0.2 Soft 0.4
Dense 0.1 Firm 0.3
Stiff 0.2
Very stiff 0.1
Hard 0.07
6. Design Factors Necessary for ISO 14692 and DNVOS-F101

• αU Material strength factor (normal value = 0.96)
• αA Anisotropy factor. Normally set to 0.95
• γSC Pressure containment factor. Equals 1.046 for low class, 1.138 for
normal class and 1.308 for high class
• γm Material resistance factor for limit states SLS/ULS/ALS. Usually set to
1.0 for FLS and 1.15 for all other limit states
• γC Pipeline condition load factor. Set to 1.07 for uneven seabed or snaked
lay condition, 0.82 for continuously stiff support and 0.93 for system
pressure test. Otherwise this value is set to 1.0.
• γinc Incidental design pressure. Set at 1.10. Can be reduced to 1.05 if the
requirements for pressure safety system are met.
• αfab Maximum fabrication factor. Usually set to 0.93 to account for
Bauschinger effects on the steel component of SSL
• fo Ovalisation factor. Usually set to 0.01 for SSL. Maximum permissible
value is 0.03
• σStatic100000 Nominal static regression stress at 100000 hours
• σCyclic150M Nominal cyclic regression stress at 150,000,000 cycles and 100000 hours
• fc Ratio of the nominal regression stress for long-term static (at 100,000
hours) to long-term cyclic (at 100,000 hours and 150,000,000 cycles)
• k∆T Themal conductivity factor for composites (default value in ISO 14692 is
0.85)
7. Vortex Induced Vibration Safety Factors
• γf Safety factor for fatigue. Normally has a value of 1.20, but can be 1.15 if
data is available for span length and gap.
• γCF Safety factor for the screening criteria which is specific for the first
vibration mode. Normally has a value of 1.30.
• δ Logarithmic decrement of structural damping. Typically the value of the
coefficient is 0.05 for steel. SSL can be set to 0.10 due to the inherent
damping in the composite component of SSL.
58 ©Ameron 2003 - FP914 03/03

• C1 Coefficient used for the calculation of the fundamental natural frequency

of the pipeline. See Table 15 for guidance.
• C2 Coefficient required for the calculation of the static deflection of the
pipeline. See Table 15 for guidance.
• C3 Coefficient used for the calculation of the fundamental natural frequency
of the pipeline. See Table 15 for guidance.
• C4 Coefficient used in the calculation of the unit diameter stress amplitude.
See Table 15 for guidance.
• C5 Coefficient used in the calculation of the static bending moment for free
spanning pipelines. This is a preliminary value, and the coefficient may
be calculated using the static soil stiffness. See Table 15 for guidance.
• C6 Coefficient is required for the calculation of the static deflection of the
pipeline. See Table 15 for guidance.
Table 2.5.1 Boundary conditions coefficients

Coefficient Pinned-Pinned Fixed-Fixed Single span on seabed
C1 1.57 3.56 3.56
C2 1.00 0.25 0.25
C3 0.8 0.2 0.4
2
L 
C4 4.39 14.1 Shoulder: 14.1 eff 
 L 
Midspan: 8.6
1
Shoulder: 2
L 
C5 1/8 1/12 18 eff  − 6
 L
Midspan: 1/24
C6 5/384 1/384 1/384
Table 15 - Boundary Condition Coefficients for VIV Calculations
59 ©Ameron 2003 - FP914 03/03

7.4 ANALYTICAL APPROACH AND EQUATIONS UTILIZED IN THE

MATHCAD ANALYSIS
Each of the logic steps in Figure 38 are calculated as described in the following sections.
7.4.1 Pressure Containment

Section 6 of the MathCAD workbook calculates both ISO 14692 and DNV requirements for
pressure containment.
7.4.1.1 GRE Stress per ISO 14692

A series of equations have been derived to calculate the stress state in the GRE and steel layers
in the SSL pipe wall. The criteria deriving the equations is strain compatibility between the
steel and GRE layers. The predicted values have been checked experimentally and correlate
very well. The data and analysis are contained in the reports summarized in Section 9.3 and
can be obtained on request.
The stress state in the GRE component of the SSL pipe wall is checked graphically to ensure
the stresses are inside the allowable long-term design envelop. The applied pressure is
calculated as follows:
P inc := γ inc⋅ P op
P ext_min := ρ Seawater⋅ g ⋅ H min
P int_min := ρ Content ⋅ g ⋅ H min
P := P inc + P int_min − P ext_min
Figure 39 gives a typical graph of the GRE stress state at the specified operating temperature
and pressure, loaded as a closed end pressure vessel. This is the first analytical check, and
determines maximum pressure rating of the pipe per ISO 14692(8). Note the steel stresses
reported in Figure 39 are well within the allowable design stress. This is typical for SSL where
the steel stress is usually less than 50% of the SMYS when the GRE component reaches the
maximum allowable level.
60 ©Ameron 2003 - FP914 03/03

Graph 6.1. Maximum Pressure Rating as a Closed End Pressure Vessel
GRE Design Envelope

40
35
30
GRE Axial Stress (MPa)
25
20
15
10
0
0 20 40 60 80 100
GRE Hoop Stress (MPa)
ID = 208.8mm P = 103.202bar Temp = 45
TLiner = 3 mm N=5 TJacket = 1.5mm
σGREHoop = 53.502MPa σGREAxial = 24.855MPa RRatioGRE= 2.153
StlDesignStress = 828MPa σStlAxialGap = 222.464MPa σStlHoop = 349.118MPa
A 1 = 0.922 A 2 = 1.000 A 3 = 0.651 f2 = 0.67
The dashed graph above is the allowable long term design envelope under all loading conditions for the
GRE at the specified operating conditions. The solid diagonal line is at the R-Ratio for the calculated
hoop to axial stress in the GRE. The short crossing lines, on the R-Ratio line, are at the calculated
stress for the GRE in the hoop and axial direction. For a proper design, the calculated GRE stress
should be within the allowable design stress envelope.
Figure 39 - Typical Output Graph for a Subsea Pipeline - Stress Condition as a Closed
End Pressure Vessel
61 ©Ameron 2003 - FP914 03/03

7.4.1.2 Average Pipe Wall Stress Containment per DNV-OS-F101

The average stress in the overall SSL pipe wall is compared separately, in Section 6.2, at the
specified operating temperature and pressure, to allowables per DNV-OS-F101.
The applied pressure is the incidental pressure defined in Section 6 of the MathCAD workbook
as:
P := P inc − P ext_min
The wall thickness, t1, is defined by the following equation per Section 5 C300 of [10]. It is not
required to consider a corrosion allowance for the SSL pipe wall. Table 16 gives the design
wall thickness for 3 to 10 steel strip SSL pipe.
t 1 := T Stl + T Liner + T Jacket
DNV defines equations for both a yielding limit state and a bursting limit state, including
design safety factors. The containment pressure at the yielding limit state and the bursting limit
state must be calculated and compared to determine which is larger as follows:
t1 2
Pbs := 2⋅ ⋅ fy ⋅ This is the yielding limit state per DNV OS-F101 Section 5, C403,
Dm 3 equation 5.16.
t1 fu 2
Pbu := 2⋅ ⋅ ⋅ This is the bursting limit state per DNV OS-F101 Section 5, C403,
Dm 1.15 3 equation 5.17.
Pb := min(Pbs , Pbu) The pressure containment criterion is determined by considering

the smaller of the yielding and bursting limit states, equation 5.15
from the DNV standard.
The values required in the above equations for the minimum yield and tensile strength of the
overall SSL pipe wall are contained in Table 16 and shown graphically in Figure 40.
Finally, Pb is compared with pressure P. Please note: equation 5.14 in the DNV standard takes
into account the necessary partial factors.
62 ©Ameron 2003 - FP914 03/03

Table 16 - Minimum Hoop Tensile and Yield Strength of SSL Pipe

Description Nominal Axial Minimum Hoop
Nominal Hoop Minimum Hoop Design
Stress at Yield Stress
Stress at Failure Tensile Stress at Wall
Failure (R=0) (R=2)
(R=2) Failure (R=2) Thickness
(Steel Stress of (90% of nominal) (1)
1150 MPa)
(MPa) (psi) (MPa) (psi) (MPa) (psi) (MPa) (psi) (inch)
SSL3 220.7 32,000 578.6 83,900 520.8 75,510 423.0 61,335 0.236
SSL4 252.8 36,660 632.4 91,700 569.2 82,530 478.9 69,440 0.256
SSL5 280.4 40,660 676.6 98,100 608.9 88,290 526.8 76,385 0.276
SSL6 303.4 44,000 716.6 103,900 644.9 93,510 568.4 82,420 0.295
SSL7 326.2 47,300 751.7 109,000 676.6 98,100 604.7 87,680 0.315
SSL8 340.2 49,330 784.1 113,700 705.7 102,330 636.8 92,335 0.335
SSL9 358.6 52,000 814.5 118,100 733.0 106,290 665.3 96,470 0.354
SSL10 372.4 54,000 841.4 122,000 757.2 109,800 690.8 100,165 0.374
(1) Minimum wall thickness for design purposes. Actual wall will be thicker
900
900
800
Minimum Hoop Stress (MPa)
700
SusuUG ( x)
SusuYG( x)
yu
600
500
400 400
2 3 4 5 6 7 8 9 10 11
2 x, x, x s 11
No. of Steel Strips
Minimum Tensile Stress Curve Fitting
Minimum Yiled Stress Curve Fitting
Test Point
Figure 40 - Minimum Hoop Yield and Tensile Strength of the Overall SSL Pipe Wall
63 ©Ameron 2003 - FP914 03/03

7.4.2 System Collapse

Section 7 of the MathCAD workbook calculates system collapse. Three different approaches
are calculated within the MathCAD workbook as follows:
1. DNV recommended Haagsma and Schaap (1981) equation for local buckling per Section
12, F700 of [11].
2. Timoshenko derivation for cylinders with free ends and small elliptical deformations.
Equation is from Roark, Table 34, Condition 8.
3. NASA SP-8007 recommended equations
The answers from the above approaches can be quite different. It is up to the pipeline engineer
to decide which of these values is appropriate for the intended service. The MathCAD
workbook defaults to the value calculated by NASA SP-8007 in Section 7.3 of the workbook,
since this appears to give the most reasonable answers. See Section 7.4.2.3 below for an
example of those calculations.
7.4.2.1 System Collapse Check for Lateral Pressure Only, per DNV Recommended
Equations
Section 7.1 of the MathCAD workbook calculates system collapes using the Haagsma and
Schaap (1981) equation using the following form:
3 2  2 Dm  2
Pc − Pel⋅ Pc −  Pp + Pel⋅ Pp ⋅ f0⋅  ⋅ Pc + Pel⋅ Pp 0
 t1 
where: Pc Is the characteristic collapse pressure
Pel Is the elastic collapse pressure.
Pp Is the plastic collapse pressure
f0 Is the ovality factor of the pipeline. Per [2] this values has
to be smaller than 0.03
In order to solve the cubic polynomial equation above, the following separate quantities are
determined (see Section 12, F700 from [11]):
3
 t1 
2⋅ E SSLHoop⋅  
P el :=
 Dm
1 − ν SSLAxialHoop⋅ ν SSLHoopAxial
 t1 
P p := 2⋅ StlYield⋅ α fab⋅ 
 Dm
B := −P el
64 ©Ameron 2003 - FP914 03/03

 2 Dm
C := − P p + P p ⋅ P el⋅ f 0⋅
 t1 
2
∆ := P el⋅ P p
⋅  − ⋅ B + C 
1 1 2
U :=
3  3 
⋅  ⋅ B⋅ C + ∆ 
1 2 3 1
V := ⋅B −
2  27 3 
Φ := acos 
−V 
 3
 −U 
Y := −2⋅ −U⋅  cos 

  Φ + 60⋅ π  

 3 180  
The characteristic collapse pressure according to Haagsma and Schaap is then calculated as
follows:
1
P c := Y − ⋅B
3
Previous experience indicates that the above value may be an overstatement of the buckling
collapse pressure of the pipeline under lateral pressure loading only. For this reason the
calculations in Sections 7.4.2.2 and 7.4.2.3 are introduced below.
7.4.2.2 System Collapse Check

Section 7.2 of the MathCAD workbook calculates the critical collapse pressure using the
Timoshenko derivation for cylinders with free ends and small elliptical deformations as
follows:
3⋅ EISSLLongTerm Equation from Roark, Table 34, Condition 8

PCritical :=
 ID (TStl + TLiner + TJacket) 
3
 + 
2 2 
Where: The value for EISSLLongTerm can be calculated using the equations in Section 6.6.3 or
taken from the values listed in Table 14. The above value for PCritical is typically the most
conservative of the collapse values calculated in Sections 7.4.2.1 to 7.4.2.3.
65 ©Ameron 2003 - FP914 03/03

7.4.2.3 System Collapse Check for Lateral Pressure Only, per NASA SP-8007
Recommended Equations
Section 7.3 of the MathCAD workbook calculates system collapse using the method provided
by NASA SP-8007. This method takes into account the buckling mode of the pipe as well as
the overall geometry.
For the SSL pipe it can be assumed that the Coil-Lock joint has the function of a stiffening
ring. With this assumption, based on previous experience, as well as on finite element studies
of the pipe the most common buckling mode of the SSL pipe is anticipated to be as shown in
Figure 41.
Figure 41 shows this buckling mode is characterized by either 2 or 3 complete circumferential

waves and 1 complete axial wave. The 3 complete circumferential waves should be more
common for the pipe laying on the sea floor, while the 2 complete circumferential waves may
be more common for a free spanning pipe. With these observations the following quantities are
defined:
ma := 1 This is the number of complete axial waves of the buckling mode.
n c := 3 This is the number of complete circumferential waves of the buckling mode.
γ := 0.9 This is a correlation factor that accounts for differences between classical
theory and the results of the experimental testing. Normaly, the values for
this factor are 0.8 to 0.9.
Figure 41 - Most Common Buckling Mode Used for SSL in NASA SP-8007
66 ©Ameron 2003 - FP914 03/03

(LPipe − LJoint)2 This is the curvature parameter.

Z := ⋅ 1 − ν SSLAxialHoop⋅ ν SSLHoopAxial
Dm
⋅ (TGRE + TStl)
2
1 n c 2(LPipe − LJoint ) This is the parameter that defines the

β := ⋅ ⋅ buckle aspect ratio.
π 2ma Dm
 
( )2 + 12 ⋅
2 2
1 2 γ ⋅Z
ky := ⋅ 1 + β  This is the buckling coefficient of the pipe
β 

2
π
4
( 1+ β
2 )
2

subject to lateral pressure.
Using the above values, the critical buckling pressure (for the chosen buckling mode) is
calculated using the following formula:
E SSLHoop⋅ (T GRE + T Stl)

2 3
π
P Critical := k y ⋅ ⋅
12 D m
(
⋅ (L Pipe − L Joint ) ⋅ 1 − ν SSLAxialHoop⋅ ν SSLHoopAxial
2
)
2
It is estimated that this is the most accurate calculation from Sections 7.4.2.1, 7.4.2.2 and
7.4.2.3. Therefore this is the critical pressure value used in the subsequent calculations within
the MathCAD workbook. The pipeline engineer will have to use his own judgment regarding
which value calculated is the most appropriate for the intended application.
The maximum installation depth can then be calculated as follows:
1  P Critical 
H MaxAllowable := ⋅  P VacMin +
ρ Seawater⋅ g  γ inc⋅ γ SC⋅ γ m 
The above value is then automatically checked against the installation depth input in Section
2.1 of the MathCAD workbook.
67 ©Ameron 2003 - FP914 03/03

7.4.2.4 Buckling Arrestors at the Coil-Lock Joint

Section 7.3 of the MathCAD workbook calculates the propagation bucking arresting capability
of the Coil-Lock joint. Propagation buckling cannot be initiated unless local buckling has first
occurred. If the external pressure exceeds the propagation buckling criterion, buckle arrestors
have to be installed at determined spacing. To make this determination the following quantities
are calculated:
StlYield This is the average stress in the SSL pipe wall when the steel
fy := σSSLHoop⋅ stress reaches yield in the hoop direction. All values in the
σStlHoop
equation were calculated in previous sections.
2.5
 t1  Propagation buckling pressure per Section 5 D510 of the
Ppr := 35⋅ fy ⋅ α fab⋅  
 Dm  DNV-OS-F101.
The calculated propagation pressure in then automatically compared to the external pressure
input in Section 2.1 of the MathCAD workbook.
7.4.3 Pipe Soil Interaction

Section 8 of the MathCAD workbook calculates the pipe soil interaction, in order to determine
the effective free span length at any unsupported area of the pipeline. This span is used in
Sections 9 and 12 of the MathCAD workbook in order to calculate the combined loading
condition from bending in the pipe wall, from the unsupported span and installed bend radius
on the sea floor, plus the loads from internal pressure. For accurate calculations the interaction
between pipe and sea floor has to be considered (10) (15). The bearing capacity factors of the
subsea soil can be calculated using the following equations:
π
φs := ⋅ φSand
180
2
  π φs  
NqSand := expπ⋅ (tan (φs)) ⋅  tan  + 
  4 2 
( ( ))
NcSand := ( NqSand − 1)⋅ cot φs
( ( ))
NγSand := 1.5⋅ ( NqSand − 1)⋅ tan φs
The output of these equations is shown graphically in the MathCAD workbook in Graph 8.1.
An example of this output is given in Figure 42.
68 ©Ameron 2003 - FP914 03/03

Graph 8.1. Bearing Capacity Factors of the Subsea Soil (sand only)
100
80
Bearing Capacity Factors
60
40
20
0
0 5 10 15 20 25 30 35 40 45 50
Friction Angle
Nq
Nc
Ngama
Figure 42 - Typical Output Graph for Bearing Capacity Factors for Subsea Soil
69 ©Ameron 2003 - FP914 03/03

The MathCAD workbook then calculates the effective span of the pipe resting on the seabed,
using the following equations:
b :=  2 ( OD − v ) ⋅ v  if v < 0.5⋅ OD This is the load distribution width as a

function of the penetration, v.
OD otherwise
RvSand := γ Sand⋅ b ⋅ ( NqSand⋅ v + 0.5⋅ NγSand⋅ b ) This is the static vertical reaction of the sea
floor soil if the soil is sand (as defined in
Section 2.2 of this workbook).
W Submerged
⋅  1 + ⋅ RSpanShoulderSand
1 1
σSand := ⋅ (1 + K0)⋅ b ⋅ γ Sand +
2 3⋅ b  2 
The effective mean stress in the soil at the
span supports, if soil is sand.
6 kN
⋅ (3 − esSand)
2
4⋅ 10 ⋅
2
m
GSand := ⋅ σSand This is the shear modulus for sand
1 + esSand
0.88⋅ GSand
KVSand := This is the relevant static soil stiffness, in
1 − ν Sand the vertical direction, if the soil is sand.
0.75
 EWtCoat ⋅ IWtCoat  This is a factor to account for the concrete
CSF := kc⋅  
 ESSLAxial⋅ ISSLAxial coating stiffening effect on the pipe wall.
 4 
KVSand⋅ Span 
β Sand := log
 ( 1 + CSF) ⋅ ESSLAxial⋅ ISSLAxial
 
4.73
RSpanSand := if β Sand ≥ 0.27
2
−0.066⋅ β Sand + 1.02⋅ β Sand + 0.63
This is the ratio of the effective
4.73 free span length over the input
otherwise
2 span length in order to consider
0.036⋅ β Sand + 0.61⋅ β Sand + 1.0 the span as fully fixed, when the
soil is sand.
Span EffSand := Span ⋅ RSpanSand Effective span length for sand
The MathCAD workbook performs the above calculations for both sand and clay to later
calculate the vortex induced vibration frequency of the unsupported pipe in the defined cross
flow
70 ©Ameron 2003 - FP914 03/03

7.4.4 Stress from Internal Pressure Plus Unsupported Span

Section 9 of the MathCAD workbook calculates axial stress from the combined loading
condition of internal and external pressure induced thrust plus bending from the unsupported
span. The bending stress from unsupported span is calculated using the effective span
calculated in Section 8 of the MathCAD workbook plus the net axial force on the installed pipe
as follows:
2 2
π⋅ ID π⋅ OD
S eff := N tr + P op ⋅ − P ext_max⋅
4 4
Seff is the net axial force on the pipeline, including residual installation forces plus internal and
external thrust.
2
π ⋅ E SSLAxial⋅ I SSLAxial
P EulerSand :=
2
Span EffSand
PEulerSand is the Euler column buckling load
The moment applied to the unsupported span can be calculated as follows:
(Span EffSand)2W Submerged

SpanMoment Sand := C 5
S eff
1 + C 2⋅
P EulerSand
The maximum strain in the GRE from the unsupported span can then be calculated as follows:
SpanMoment ⋅ 
 ID + T 
GRE + T Stl 
ε MaxGRE :=
 2 
E SSLAxial⋅ I SSLAxial
Note: The overall SSL pipe wall values (ESSLAxial and ISSLAxial) are used in the above
equation in order to get the strain from unsupported span for the overall SSL pipe structure.
Finally, the stress in the GRE component can be calculated as follows:
σ GREAxialSpan := ε MaxGRE⋅ E GREAxial
The axial stress from unsupported span must then be added to the stress from thrust as follows:.
σ GRETotAxialSpan := σ GREAxial + σ GREAxialSpan
The total axial stress in the GRE is then checked graphically, as shown in Figure 43 against the
allowable long-term GRE design envelope. The GRE design envelope is sized for f2 = 0.67,
per Table 3 of ISO 14692(8).
71 ©Ameron 2003 - FP914 03/03

Graph 9.1. Internal pressure, plus axial stress from free span (tensile stress side).
GRE Design Envelope

40
35
30
25
20
15
10
0
0 10 20 30 40 50 60 70 80 90 100
ID = 208.8mm N=5 P = 103.202bar Temp = 45
TLiner = 3 mm TJacket = 1.5mm Span = 22m
σGREHoop = 53.502MPa σGRETotAxialSpan = 26.569MPa RRatioGRE= 2.014
StlDesignStress = 828MPa σStlTotAxGapSpan = 246.108MPa σStlHoop = 349.118MPa
A 1 = 0.922 A 2 = 1.000 A 3 = 0.651 f2 = 0.67
The dashed graph above is the design envelope under all loading conditions for the GRE at the specified
operating temperature. The solid diagonal line is at the R-Ratio for the calculated hoop to axial stress in
the GRE. The short crossing lines, on the R-Ratio line, are at the calculated stress for the GRE in the
hoop and axial direction. For a proper design, the calculated GRE stress should be within the allowable
design stress envelope.
Pressure plus Unsupported Span
72 ©Ameron 2003 - FP914 03/03

7.4.5 Vortex Induced Vibration at the Unsupported Span

Section 10 of the MathCAD workbook does the on-bottom analysis of the pipeline following
the additional requirements of the DnV Recommended Practice titled “Free Spanning
Pipelines” and the DnV Classification Notes No. 30.5 titled "Environmental Conditions and
Environmental Loads". The DnV documentation recommends the use of Finite Element
Methods (FEM) for this calculation. In order to perform these calculations using closed form
solutions, it is necessary to determine an approximate equation that relates the natural
frequency of the SSL pipe to the maximum allowable free span length. This method is
considered acceptable in the literature and is available in a separate MathCAD workbook. The
following quantities are determined:
The dynamic mass, including the weight coating can be calculated as follows. Note the
dynamic mass includes the weight of the displaced water per DNV.
π
( ) ⋅  ID ⋅ ρ Content + ODWtCoat ⋅ ρ Seawater
2 2
M dyn := AREAStlAxial⋅ ρ Stl + AREAGREAxial⋅ ρ GRE +
4
and then also using the information from the previous sections, the natural frequency of the
pipeline which also accounts for the pipe-soil interaction, can be determined using the
following equation:
  DeflFreeSpan Sand  
2
E SSLAxial⋅ I SSLAxial S eff
f nSand := C 1⋅ 1 + CSF⋅ ⋅ + 2⋅
1 C + 3⋅ 
C  
4  P EulerSand  OD WtCoat  
M dyn ⋅ Span EffSand 
A similar calculation for a clay bottom is also done in the MathCAD workbook.
U100⋅ ODWtCoat
Re := Calculated Reynolds number of the cross flow at the 100 year
νk storm condition.
Knowing the Reynolds number, the Strouhal number can be determined from the graphs
provided in the DnV documentation. For the purpose of computerized calculations an
approximation function is used in the Mathcad® workbook to determine this number.
−1
Cm := 1 +  10⋅ + 1 
e
This is the added mass coefficient, calculated as a function of the
 ODWtCoat  gap ratio.
2
δ + 1−1
  This is the calculated ratio of structural damping for the given
ζ := π⋅
π value of the logarithmic decrement coefficient of structural
δ damping, δ. The most common value is 0.008.
2  W Pipe + W WtCoat  This equation calculates the ratio of structural damping. (Only dry
ks := π ⋅ ζ⋅  − Cm  weight of pipe is considered).
 W Submerged 
73 ©Ameron 2003 - FP914 03/03

3 W Pipe + W WtCoat 
 π + C m
V R :=
 W Submerged 
2 3 W Pipe + W WtCoat  2
1.5 + St ⋅ π ⋅  + C m  − 1.5⋅ k s 
  W Submerged  
The above equation is the reduced velocity of the cross flow. Crossflow excitations may occur
when 3< VR<16 for all Reynolds numbers, but the maximum response is found in the range
4.8< VR<8. The designer using this Mathcad workbook can reduce the severity of the inflow
oscillations by changing the structural properties. In practice other methods are available such
as the addition of aerodynamic devices such as strakes or spoiling devices which reduce the
strength of the vortex-induced forces.
St⋅ U100
fs := ⋅ γ CF⋅ γ f Calculated frequency of the vortex shedding at the 100 year storm
ODWtCoat (cyclonic) condition.
The maximum allowable free span length is calculated in two different ways. A separate
MathCAD workbook explains the derivation of these equations.
0.7⋅ C 1 E SSLAxial⋅ I SSLAxial

Span VIV := ⋅
fs M dyn
This is a simple verification based on the criteria that no excitations occur if fs<0.7*fn.
Finally, a more conservative value for the span is calculated using the accurate formula for the
natural frequency of the pipeline as follows:
C1 E SSLAxial⋅ I SSLAxial
SpanSand VIV := ⋅
f nSand M dyn
It is important to note that in the above equation for the free span, the modulus and the moment
of inertia values are for the SSL pipe wall only.The weight coating is assumed to crack, and its
contribution to the axial stiffness of the pipe is accounted for in the calculation of the natural
frequency.
The MathCAD workbook also automatically calculates spans for clay for comparison purposes.
The calculated spans are then compared graphically to the input span and the full range of
potential fixity constants (C1) as shown in Figure 44.
74 ©Ameron 2003 - FP914 03/03

Graph 10.1. Natural Frequency vs. VIV Pipe Span
4.5
3.5
3
Natural Frequency (Hz)
2.5
1.5
0.5
0
10 15 20 25 30 35 40 45 50
Span Length (m)
Fixed-Fixed
Average Operating Condition
Simple-Simple
Calculated Operating, Sand
Calculated Operating, Clay
Static Span
m
Span = 22m U100 = 0.75 e = 0.9m
sec
2
mm 5
ν k = 1.05 Re = 1.734 × 10 VR = 4.173
sec
fnSand = 1.737Hz fnClay = 1.582Hz fs = 0.957Hz
Span VIV = 17.309m SpanSand VIV = 15.352m SpanClay VIV = 16.086m
Figure 44 - Typical Graph of Calculated Spans for VIV versus Input Span
75 ©Ameron 2003 - FP914 03/03

7.4.6 Stress from Internal Pressure Plus Installed Bend Radius in the
Horizontal and Vertical Plane
Section 11 of the MathCAD workbook calculates stress from the combined loading condition
of thrust from internal and external pressure plus installed bend radius in both the horizontal
and vertical planes.
The strain and stress from installed bend radius, in the vertical plane, can be calculated by
geometry as follows:
ID
+ T GRE + T Stl
2
ε BendRadV :=
BendRad V
σGREBendRadV := EGREAxial⋅ ε BendRadV This is the stress in the GRE component of

the SSL pipe from the installation bend
radius in the vertical plane.
Likewise, the strain and stress from installed bend radius, in the horizontal plane, can be
calculated by geometry as follows:
ID
+ T GRE + T Stl
2
ε BendRadH :=
BendRad H
σGREBendRadH := EGREAxial⋅ ε BendRadH This is the stress in the GRE component of

the SSL pipe from the installation bend
radius in the horizontal plane.
Finally, the combined stress from bending can be calculated as follows:
This is the total stress in the GRE

2 2
σGREBendRadius := σGREBendRadV + σGREBendRadH component of the SSL pipe from the
installed bend radius.
The above calculated axial stress from the combined horizontal and vertical installed bend
radius must then be added to the stress from thrust as follows:
σ GRETotAxialBendRad := σ GREBendRadius + σ GREAxial
The total axial stress in the GRE component of SSL is then checked graphically, as shown in
Figure 45 against the allowable long-term GRE design envelope. The GRE design envelope is
sized for f2 = 0.83, per Table 3 of ISO 14692(8).
76 ©Ameron 2003 - FP914 03/03

Graph 11.1. Internal pressure, plus installed bend radius in the vertical and horizontal planes
(tensile stress side).
GRE Design Envelope

45
40
35
30
25
20
15
10
0
0 20 40 60 80 100 120 140
ID = 208.8mm N=5 P = 103.202bar Temp = 45
σGREHoop = 53.502MPa σGRETotAxialBendRad = 28.976MPa RRatioGRE = 1.846
StlDesignStress = 828MPa σStlHoop = 349.118MPa
BendRad V = 550m BendRad H = 1200m
A 1 = 0.922 A 2 = 1.000 A 3 = 0.651 f2 = 0.83
Pressure plus Installed Bend Radius
77 ©Ameron 2003 - FP914 03/03

7.4.7 Stress from Internal Pressure Plus Unsupported Span Plus Installed
Bend Radius in the Vertical Plane – Section 12 of the MathCAD Workbook
Section 12 of the MathCAD workbook calculates stress from the combined loading condition
of thrust from internal and external pressure plus bending stress from unsupported span and
installed bend radius in vertical plane.
It is assumed in the following calculations; there is no installed bend radius in the horizontal
plane at the unsupported span. The reason for this assumption is the inability of the subsea soil
to apply the necessary horizontal bending moment to maintain the installed bend radius if the
pipe is not touching the seabed. The total axial stress in the GRE can be calculated using the
following equation and the previously calculated values.
σ GRETotAxialBendRadV := σ GREBendRadV + σ GREAxial + σ GREAxialSpan
The above axial stress from unsupported span and from installed bend radius (vertical direction
only) is than checked graphically, as shown in Figure 46, against the allowable long-term GRE
design envelope. The GRE design envelope is sized for f2 = 0.83, per Table 3 of ISO 14692(8).
78 ©Ameron 2003 - FP914 03/03

Graph 12.1. Internal pressure, plus free span, plus installed bend radius in the vertical plane
GRE Design Envelope

45
40
35
30
25
20
15
10
0
0 20 40 60 80 100 120 140
ID = 208.8mm N=5 P = 101.55bar Temp = 45
σGREHoop = 52.645MPa σGRETotAxialBendRadV = 30.267MPa RRatioGRE= 1.739
StlDesignStress = 828MPa σStlTotAxGapBendRadV = 299.702MPa σStlHoop = 343.529MPa
BendRadV = 570m BendRadH = 1200m
A 1 = 0.922 A 2 = 1.000 A 3 = 0.658 f2 = 0.83
Figure 46 – Typical Output Graph for Subsea Pipelines – Stress Condition for Internal
Pressure plus Installed Bend Radius in the Vertical Plane plus Unsupported Span
79 ©Ameron 2003 - FP914 03/03

7.4.8 Analysis of Installation Loads Using the J-Lay Method

The main assumption for this section is that the preferred offshore installation method for SSL
pipelines is a J-lay type of installation. The schematic shown in Figure is a simplified
representation of the J-lay installation method. The equations used in this calculation were
derived using the J-lay schematic and catenary theory. They are available in a separate
MathCAD workbook on request.
V T
Beta
y(x), h
Alpha
H
T
x, p
Figure 47 - Jay Lay Geometry During Installation
The geometry and mechanics of the J-lay can be calculated as follows:
W Pipe := (AREAJacketAxial + AREALinerAxial)⋅ ρ GRE + AREAStlAxial⋅ ρ Stl ⋅ g This is the weight of the
dry pipe, per unit length of
pipeline.
π 2
This is the weight of the pipe content (at the installation of the
W ContInst := ⋅ ID ⋅ ρ Seawater⋅ g
4 pipeline) per unit length of pipeline. During installation the common
practice is to flood the pipeline with sea water.
π 2 This is the weight of the displaced water, per unit length of pipeline,
W DSea := ⋅ ODWtCoat ⋅ ρ Seawater⋅ g and including the weight covers. It is used to calculate the apparent
4
weight of a submerged object, per Archimedes law.
W SubmInst := W Pipe + W WtCoat + W ContInst − W DSea Apparent (submerged) weight of pipe

during installation.
80 ©Ameron 2003 - FP914 03/03

Given
  Lp  
Hmax a⋅  cosh  −1
  a  
RadInst := Find( a)
180   Lp  
β := ⋅ atan  sinh   This is the J-lay angle as shown in the above schematics.
π   a 
α cat := 90 − β This is the stinger setup angle.
 Lp 
Lcat := a⋅ sinh   This is the actual length of the catenary shaped pipeline during
 a  installation. This length is measured to the TDP only.
FhBarge := a⋅ W SubmInst This is the horizontal reaction force at the installation barge.
FvBarge := Lcat⋅ W SubmInst Vertical reaction force at the installation derrick.
The above calculation can be verified by checking the installation angle from the calculated
reactions as follows:
180  Fh Barge 
α ver := ⋅ atan 
π  Fv Barge 
Finally, the "true" tension at the pipeline during installation using the J-lay method can be
calculated as follows:
2 2
T := Fh Barge + Fv Barge
Assuming a "Part Factor" for f2 = 1.00, the maximum axial stress during installation is:
σ MaxGREAxialInst = 29.4 MPa
Using the value for the axial modulus for GRE at the installation temperature, then the
minimum allowable bend radius can be calculated as follows:
4
E GREAxialInst = 1.93 × 10 MPa
E GREAxialInst⋅  + T GRE + T Stl 

ID
Rad Instmin :=  2 
σ MaxGREAxialInst
81 ©Ameron 2003 - FP914 03/03

Finally, the MathCAD workbook plots thet J-Lay installation diagram for both minimum and
maximum installation depth versus distance from the lay barge as given in Figure 48.
Graph B.1. J-Lay Installation Diagram
11.43
32.86
Installation Depth (m)
54.29
75.71
97.14
118.57
50 0 50 100 150 200

Installation Length (m)
Catenary for maximum depth
Catenary for minimum depth
RadInstmin = 73.118m RadInst = 183.779m
FhBarge = 41.838kN FhBargeAllow = 200kN
α ver = 37.227 Vertical angle (deg) at the installation derrick.
Lcat = 241.882m Actual length of the string for the given installation parameters.
Figure 48 - J-Lay Installation Diagram
82 ©Ameron 2003 - FP914 03/03

In order to determine the combined axial stress in the J-Lay installation the catenary equations
are approximated with parabola equations. It can be demonstrated that the errors introduced by
such an approximation are approximately 1%. Most of the equations are similar except the
equations describing the variation of the radius of curvature and tension:
3
2
 2
rad ( x) := ap ⋅  1 +  x  This function describes the variation of the installation radius during
a 
  p  the J-Lay installation. The pipeline is modeled as a parabola.
2
T( x) := T0⋅ 1 + 
x 
 This function describes the variation of the tension during the
 ap  J-Lay installation. The pipeline is modeled as a parabola.
T( x)
σ AxialTens( x) :=
AREA GREAxial + AREA StlAxial
σ AxialBend( x) := E SSLAxial⋅ ε BendRad( x)
σ TotAxial( x) := σ AxialTens( x) + σ AxialBend( x)
The equations above show the calculation of the combined axial stress at any point of the
installation string at the J-Lay installation. The functions of the stress from tension, stress from
bending and the combined axial stress are shown in Figure 49.
Similar to the catenary evaluation the remaining geometry of the installation are calculated
using the following equations:
180  Lp  This is the tangent angle at the installation.

β par := ⋅ atan  
π  ap 
Finally the actual length of the installation string can be determined using the equation for the arc length of
the parabola. The differential equation for the arc length is:
2
d
s
d 
1 +  y By simple integration and using the boundary conditions for the given
dx  dx  geometry of the J-Lay installation, this yields:
 L 2 
 p 
ap  Lp   Lp 
Lpar := ⋅   ⋅ 1 +   + asinh  
2  ap   ap   ap 
.
83 ©Ameron 2003 - FP914 03/03

Graph B.2. Combined Axial Stress in the Pipeline during the J-Lay Installation using the
Parabola Model
140
120
100
Total Axial Stress (MPa)
80
60
40
20
0
0 50 100 150 200
Combined Axial Stress
Axial Stress from Bending
Axial Stress from Tension
Allowable Axial Stress
xMin = 200m σAxialMin = 27.608MPa
xMax = 0 m σAxialMax = 65.621MPa σAxialAllow = 123.1MPa
RadInst = 183.779m RadInstpar = 166.667m RadInstmin = 73.118m
α cat = 37.227 α par = 39.806
Lcat = 241.882m Lpar = 240.869m
Depending on the mechanical and geometrical conditions at installation, the function for the combined
axial stress in the pipe can have no more than two extreme points. In most occurrences, the graph of the
function is uniform, either ascending or descending.
From the results it can be noticed that indeed the parabola model is a close approximation of the catenary
model for the J-Lay installation. Furthermore, the parabola model is also a more conservative approach.
Figure 49 - Combined Axial Stress at the J-Lay Installation
84 ©Ameron 2003 - FP914 03/03

7.4.9 Analysis of Installation Loads Using the S-Lay Method

For the situations when the installation angle as determined in the J-Lay method is not
practical, an S-Lay installation can be used. For the S-Lay method there are certain generally
accepted requirements for “good engineering practice”. These requirements are as follows, and
it is assumed that they are met:
• The stinger setup ensures that the pipeline overbend radius is larger than the minimum
allowable installation radius for the pipeline. Common values for the overbend radius
are about 200 m, but such that the strain at the overbend region (most likely the most
stressed area) stays within the elastic limit. The pipeline inflection point is virtually
coincident with the stinger tip. This determines the tip angle.
• The tension force at the hydraulic tensioner controls the sagbend curvature. The radius
of curvature at the sagbend has to be larger than the minimum allowable installation
radius as well.
• The tension force at the hydraulic tensioner also controls the contact between the pipe
and the stinger tip (usually between the last and the third from the last roller). This is
required in order to minimize the risk of very high, uncontrolled, dynamic bending
loads. If all the rollers are in contact with the pipe, the tension is increased, while it is
decreased if too few rollers are in contact.
• The sagbend section of the pipeline follows a catenary shape. The parabola
approximation of the catenary is acceptable and conservative.
• The S-Lay calculation determines static loads only. Dynamic loads from wave and
current actions are ignored in this calculation, as they are beyond the scope of the
Mathcad workbook.
Based on these assumptions the equations used for the determination of this installation are as
follows:
1 2
f ( x) − ⋅x This is the parabola equation for the stinger portion of the installation string
2⋅ aSting (the overbend section). The origin of the coordinate system is at the
tensioner.
1 2
g ( x) ⋅ (x − h) + k This is the parabola equation for the sagbend section of the installation
2⋅ aparS string. Parameters "h" and "k" determine the position of this section in
relation to the origin of the coordinate system.
LSting This is the characteristic parameter of the parabola equation for the stinger
aSting :=
tan (φ) portion of the installation. As mentioned before, values smaller than 200 m
are not practical.
219 798
1 2
HSting := ⋅ LSting This is the depth of the pipeline at measured at the stinger tip.
2⋅ aSting
85 ©Ameron 2003 - FP914 03/03

2
 aSting  This is the characteristic parameter of the parabola equation for the
aparS := 2⋅ (Hmax − HSting)⋅   sagbend section. This equation is derived using the boundary
 LSting  condition for the inflection point.
aSting
LpS := 2⋅ (Hmax − HSting)⋅ This is the characteristic installation length for the sagbend section
LSting of the installation. This equation is derived using the boundary
condition for the inflection point.
For the geometry of the installation calculated as shown above, the forces and stresses in the
installation string can be determined using the equations:
T hMin := W SubmInst⋅ a parS
3
2
  x  
2 This is the function for the radius of curvature of the overbend
rad Sting( x) := aSting⋅  1 + a  section of the installation string, and also represents a setup
  Sting   parameter of the stinger.
T hMin OD
σ overbend( x) := + E SSLAxial⋅
AREA GREAxial + AREA StlAxial 2⋅ rad Sting( x)
And finally the total length of the installation string:
The length of the installation string is determined (for the maximum installation depth) by the sum of the
arc lengths of the overbend and the sagbend sections, as follows:
 L 2 
aSting  Sting   LSting   LSting 
Lpar1 := ⋅   ⋅ 1 +   + asinh  
2  aSting   aSting   aSting 
 L 2 
aparS pS   LpS   LpS 
Lpar2 := ⋅  ⋅ 1 +   + asinh  
2  aparS   aparS   aparS 
L stringS := L par1 + L par2
Similar to the J-Lay analysis, the MathCAD workbook plots the S-Lay installation diagram for
both minimum and maximum installation depth versus distance from the lay barge as given in
Figure 50.
86 ©Ameron 2003 - FP914 03/03

Graph B.3. S-Lay Installation Diagram
300
200
100
Installation Depth (m)
100
200
300
700 600 500 400 300 200 100 0
Stinger Section
Subsea Section, Maximum Depth
Subsea Section, Minimum Depth
Inflection Point
Figure 50 - S-Lay Installation Diagram
7.4.10 Design Summary

The MathCAD workbook provides a summary of the inputs and output calculations at the end
of the workbook. Figure 51, Figure 52 and Figure 53 show typical input summaries.
Table 17 is a typical calculation summary table, showing stresses under various operating
conditions, with a design check for each loading condition. The design checks are referenced
to the graphical checks shown in Figure 39 to Figure 46. It can be seen by review of Table 17,
plus Figure 46, the limiting GRE are for the combined loading condition of internal and
external thrust, plus bending stress from installed bend radius and unsupported span.
87 ©Ameron 2003 - FP914 03/03

C. SUMMARY OF INPUTS AND CALCULATIONS
Product Dimensional Inputs
ID = 208.8mm N=5
TStlStrip = 0.5mm TLiner = 3 mm TJacket = 1.5mm
kg
TWeight = 10mm ρ Weight = 2242
3
m
LPipe = 12.19m LJoint = 400mm
Operating Conditions
Pop = 95bar PVacMin = 1 bar
9
NCycles := 8⋅ 10 Pmin = 38.5bar Pmax = 95bar
TempInstall = 21 Temp = 45
BendRadV = 550m BendRadH = 1200m
kg
ρ Content = 800
3
m
Subsea Conditions
Hmin = 90m Hmax = 120m
3 kg
ρ Seawater = 1.027× 10
3
m
2
mm
ν k = 1.05 Kinematic viscosity of sea water in cross flow
sec
m Sea current velocity at pipeline span for the 100 year storm. Usual
U100 = 0.75 values (cyclonic condition) are 0.9 to 1.2 m/sec.
sec
Span = 22m Expected free spanning of the pipeline on the sea floor.
e = 900mm Gap between pipeline and sea floor.
Installation Conditions
Lp = 200m Installation length behind the barge (using a J-lay installation

method).
FhBarge = 41.838kN Horizontal force at the installation barge (using the J-lay method).
Figure 51 - Typical Summary of Inputs for Product Dimensions, Operating Conditions,

Subsea Conditions and Installation Conditions
88 ©Ameron 2003 - FP914 03/03

LSting = 80m This is the stinger length at the S-Lay installation
φ = 0.349 This is the stinger setup angle
Ntr := 0⋅ kN "True" axial force (from installation) in the pipe wall
Subsea Soil Conditions
φSand = 32 Angle of friction for submerged sand. This parameter applies only to
cohesionless soils (will not apply to clay).
kN This is the submerged unit weight of sand.

γ Sand = 10.5
3
m
ν Sand = 0.35 This is the Poisson's ratio for sand.
esSand = 0.65 This is the void ratio for sand
kN This is the undrained shear strength of clay. This parameter applies

Su = 28
2 only to cohesive soils (will not apply to sand).
m
kN This is the submerged unit weight of clay.
γ Clay = 8.5
3
m
ν Clay = 0.45 This is the Poisson's ratio for clay.
esClay = 1.2 This is the void ratio for clay.
v = 60mm This is the vertical penetration of the pipe in the see soil,
determined from preliminary tests. If no data is available a
conservative value of 20-40% of the diameter will be considered.
K0 = 0.5 Coefficient of earth pressure at rest, usually 0.5.
RSpanShoulderSand = 0.2 Ratio of shoulder length to free span length if the sea floor is sand.
RSpanShoulderClay = 0.3 Ratio of shoulder length to free span length if the sea floor is clay.
Design Factors
α U = 0.96 Normal value of factor. Shall be 1.00 for system pressure test.
α A = 0.95 Anisotropy factor. Value shown for axial direction, due to relaxed
testing requirements in specification (Table 6-3) Otherwise is 1.00.
γ SC = 1.308 Pressure containment safety class factor. For low class is 1.046,
for normal class is 1.138, and for high class is 1.308
γ m = 1.15 Material resistance factor for limit states SLS/ULS/ALS. Shall be

1.00 for FLS and 1.15 for all other limit states.
γ C = 1.07 Pipeline condition load effect factor. Shall be 1.07 for uneven
seabed or in a snaked condition, 0.82 for continuously stiff
supported, 0.93 for system pressure test, and 1.00 otherwise.
Figure 52 - Typical Summary of Inputs for Subsea Soil Conditions and Design Factors
89 ©Ameron 2003 - FP914 03/03

γ inc = 1.10 Incidental to design pressure ratio, normally 1.10. If the

requirements for the pressure safety system are met, the value
can be set to 1.05, but not less.
α fab = 0.93 This is the maximum fabrication factor. Considering the specifics of
the fabrication of the SSL pipe compared to the requirements in
Section 5, B600 of [2] this partial factor can be 0.93 or (more
conservative) 0.85, to account for Bauschinger effects on the steel
component.
f0 = 0.03 This is the ovalisation factor. The maximum permissible value is

0.03. If one assumes there is no ovalisation at the pipe, a
mandatory check is required for 0.005.
σ Static100000 = 158.29MPa These are projected stress values at 100000 hours and 150
million cycles respectively, used to determine the factor for
σ Cyclic150M = 52.44MPa calculating the LTHScyclic in Section 6.
Cyclic Long-Term Hydrostatic Strength factor as determined by

fc = 3.018 ISO 14692-2:2002 Appendix K. In absence of data, default value
of 4 shall be used as specified by Shell DEP 98.
k∆T = 0.85 This factor accounts for the low thermal conductivity of GRE.
Deflections and Vortex Induced Vibration Coefficients
γ f = 1.15 Safety factor for fatigue.
γ CF = 1.3 Cross flow screening safety factor.
Cm = 1.026 Added mass coefficient.
δ = 0.05 Logarithmic decrement of structural damping. Typically the value of

this coefficient is 0.05.
C1 = 2.56 This coefficient is used for the calculation of the fundamental

natural frequency of the pipeline, Section 10.
C2 = 0.62 This coefficient is required for the calculation of the static deflection
of the pipeline, Section 10.
C3 = 0.5 This coefficient is used for the calculation of the fundamental

natural frequency of the pipeline, Section 10.
C4 = 8.6 This coefficient is used in the calculation of the (unit diameter)

stress amplitude.
1
C5 = This coefficient is used for the calculation of the static bending
10 moment for free spanning pipelines. This is a preliminary value, and
the coefficient may be calculated using the static soil stiffness.
5
C6 = This coefficient is required for the calculation of the static deflection
768 of the pipeline.
90 ©Ameron 2003 - FP914 03/03

Figure 53 - Typical Summary of Inputs for Design Factors (continued) and Deflections
and Vortex Induced Vibration Coefficients
Design Check Summary
CHAPTER A. OPERATING CONDITIONS

Description
Section 3.1. Temperature Derating Factor 0.922
Section 3.1. Chemical Derating Factor 1.000
Section 3.3. Cyclic Load Derating Factor 0.658
Calculated Calculated Max Hoop

Description (continued) GRE Hoop GRE Axial R-Ratio Part Factor Stress at R- Design Check
Stress (MPa) Stress (MPa) Ratio (MPa)
Section 6.1. Closed End Pressure Vessel 52.65 24.46 2.153 0.67 64.49 OK
Section 9. Internal Pressure Plus Free

52.65 26.65 1.975 0.67 55.04 OK
Span on Seafloor
Section 11. Internal Pressure, Plus
52.65 28.46 1.850 0.83 59.50 OK
Installed Bend Radius
Section 12. Internal Pressure, Plus
Installed Bend Radius (vertical plane) Plus 52.65 30.27 1.739 0.83 52.77 OK
Free Span
Design Allow able
Section 6.2. DNV Pressure Containment Criteria, Stress, (MPa) 101.55 274.28 OK
Section 7.3. System collapse check for lateral pressure, Max Depth (m) 120.00 325.47 OK
Section 7.4. Propagation Buckling, (bar) 12.09 22.46 OK
Section 10.1. Pipeline Natural Frequency, Sand (Hz) 2.34 N/A OK
Section 10.1. Pipeline Natural Frequency, Clay (Hz) 2.08 N/A OK
Section 10.2. Vortex Shedding Reduced Velocity 4.17 N/A OK
Section 10.2. Frequency of Vortex Shedding (Hz) 0.96 Check done on follow ing line
Section 10.2. Span from Vortex Induced Vibration (m) 17.00 17.31 OK
Section 10.2. Span (conservative check), Vortex Induced Vibration, Sand (m) 17.00 13.23 Check Spoiling
Section 10.2. Span (conservative check), Vortex Induced Vibration, Clay (m) 17.00 14.02 Check Spoiing
CHAPTER B. INSTALLATION
Calculated Allow able
Section B.1. Installation radius, J-Lay (m) 183.78 73.12 OK
Section B.2. Maximum Combined Axial Stress, J-Lay (MPa) 65.62 123.10 OK
Section B.3. Minimum Overbend (stinger) radius, S-Lay (m) 219.80 73.12 OK
Section B.3. Maximum Combined Axial Stress at Overbend, S-Lay (MPa) 120.06 123.10 OK
91 ©Ameron 2003 - FP914 03/03

Table 17 - Typical Summary of Calculation Outputs from MathCAD® Analysis, Including

Design Checks
8. OTHER CONSIDERATIONS IN SSL PIPELINE DESIGN
8.1 SSL PIPELINE HYDRAULIC CALCULATIONS
8.1.1 Definitions
Definitions of the terms used in this section are presented below. It is assumed that the user has
some basic understanding and knowledge of terms and concepts utilized in mechanics of fluids.
µ Dynamic (or absolute) viscosity, as defined by Newton’s law and expressed in Poise
(P) or stokes (St) in CGS units. It is important to note that the multiples of these
units should not be used together with SI units. Conversion factors are presented in
the literature.
ν Kinematic viscosity defined as the ratio of dynamic viscosity to mass density. Fluid
is assumed to be Newtonian , which means that it exhibits a linear relation between
the applied shear stress and the resulting rate of deformation. When this
relationship is non-linear the fluid is called non-Newtonian, and a typical example is
shale oil or very thick (wax filled) hydrocarbons.
ε Absolute roughness of the interior pipe wall. Expressed in the same units as the
pipe diameter.
f Friction factor characterizing the flow.
H Head of fluid is used to denote system energy. There usually have to be considered
a static (or elevation) head, velocity head, and friction head.
Hf Frictional head loss represents the ratio of pressure loss to specific gravity of the
fluid as defined by the Darcy-Weisbach equation. (See following paragraph for
details and form of the equation.)
Re Reynold’s number
8.1.2 Design Methods and Assumptions

The dominant aspect to be considered in the hydraulic design is the friction head loss. There
are three methods used in this type of calculation:
• Darcy-Weisbach
• Fanning
• Hazen-Williams
The former two equations require a determination of the friction factor as a function of the
Reynolds number. The Hazen-Williams equation is an equation entirely based on empirically
determined coefficients. The engineering literature provides useful charts and tables to be used
with any of these equations.
However, the goal of this manual is to provide an analytical approach that does not require
reference to other design manuals or graphs. For this reason it has been decided that the
Darcy-Weisbach equation is best suited for this purpose. Therefore only this method will be
discussed, and an entirely analytical method will be provided for the design. It is strongly
recommended though, that for demanding or critical installations the results provided by this
92 ©Ameron 2003 - FP914 03/03

method should be verified against other conventional methods, or when available, against
specialized fluid flow analysis software methods.
8.1.3 Design Equations

The following pages outline a complete flow and head loss calculation for a common pipeline.
The notational conventions are the same as those used in the Mathcad® workbook. This
workbook is available from AMERON, and will automatically provide a complete flow
analysis.
Input Data
ID Inside diameter of SSL pipe, in.
IDStl Inside diameter of the steel pipe, in.
ε Surface roughness of the interior wall of SSL. Depends on the chosen pipe material,
in.
ε Stl Surface roughness of the interior wall of the steel pipe. Depends on the chosen pipe
material, in.
FR Flow rate per operating day, lb/sec.
Temp Operating temperature, oF.
ρ Density of the transported fluid at the operating temperature, lb/ft3.
ν Kinematic viscosity of the transported fluid at the operating temperature, ft2/sec. It
is very important not to confuse this value with the absolute viscosity.
L Length of pipe line, ft. It is assumed that the pressure drop and the frictional head
loss are calculated at the end of this length.
Calculation Formulas
The preliminary calculations use the following formulas:
π
Area = ⋅ ID 2 Flow section area. (1)
4
FR
w= Flow velocity. (2)
ρ ⋅ Area
ID ⋅ w
Re = Reynolds’s number. Dimensionless. (3)
ν
It is generally accepted that a laminar flow is characterized by Reynolds’s numbers equal to or
smaller than 4000. Larger values represent a transition or turbulent flow. With this number
calculated, the general practice is to determine the transitional and turbulent Darcy-Weisbach
friction factor using available graphs and tables. The alternate method is to use the following
equations, consistent with an entirely analytical design:
For laminar flow, with Re ≤ 4000 , use the equation
64
f =
Re
For transitional or turbulent flow, use the Colebrook equation. This equation has been
demonstrated to represent the analytical basis for the Moody diagrams, normally used
for the hand calculations.
93 ©Ameron 2003 - FP914 03/03

1 ε 2.51
= −0.86 ⋅ ln( + ) (4)
f 3.7 ⋅ ID Re⋅ f
1
Using the substitution x = the equation (4) becomes:
f
ε 2.51
x = −0.86 ⋅ ln( + ⋅ x)
3.7 ⋅ ID Re
or in the generalized form,

x = a ⋅ ln(b + c ⋅ x)
The solution for this equation is given by:

−1 a
− b ⋅ c ⋅W ( ⋅ exp(− )) − a
x= b⋅c b⋅c
b
where W is the so-called Lambert’s W function. This type of solution usually requires
series expansions which make the calculations quite elaborate. Therefore, AMERON
recommends computer assisted techniques for solving equation (4). The Mathcad®
workbook for SSL pipeline analysis, and the special workbook for flow analysis are
capable of performing this calculation
With the friction coefficient determined, the pressure loss can be calculated as follows:
w2 L
∆p = f ⋅ ρ ⋅ (5)
2 ⋅ g ID
lb ft
expressed in 2
, and where g = 32.174 ⋅
in sec 2
The Mathcad® workbooks available from AMERON International use the above formulas to
compare a steel pipeline with the equivalent SSL pipeline. A sample output of such a
calculation is presented in Figure 54 for a pipeline designed to transport crude oil with a 50%
water-cut.
The input data was as follows:
ε Steel = 0.0018 ⋅ in Surface roughness is considered equivalent to a

commercial cast iron pipe. Experience indicates
that a steel pipeline, after corrosion has initiated,
will have a comparable surface roughness.
94 ©Ameron 2003 - FP914 03/03

ε GRE = 0.000084 ⋅ in Surface roughness of the GRE pipeline.

lb
ρ = 50.85 Density of crude oil with a 50% water-cut.
ft 3
ft 2
ν = 0.01St = 1.076391 ⋅ 10 −5 Kinematic viscosity of crude oil with a 50%
sec
water-cut.
It can be seen from this sample analysis, that a 10 inch diameter SSL pipeline can easily replace
a 12 inch diameter steel pipeline. This has a very important economic significance to the
pipeline installer as well as the operator, and clearly shows that the direct comparison of the
SSL pipelines to the steel ones is not possible without such an analysis.
8.1.4 Determination of an Analytical Equation for Kinematic Viscosity

In the calculation method presented above, the user only needs to research the engineering
literature for the value of the kinematic viscosity. This may not be an available option in
certain conditions. Therefore, the remainder of this section presents the method used to derive
an analytical equation that can be used to calculate the value of the viscosity.
Interpolated Viscosities Oil Tem p.

Water 10 20 30 40 50 60 70 80 90 (F)
2.40E-05 4.66E-05 6.92E-05 9.18E-05 1.14E-04 1.37E-04 1.60E-04 1.82E-04 2.05E-04 2.27E-04 0.00025 25
1.40E-05 2.76E-05 4.12E-05 5.48E-05 6.84E-05 8.20E-05 9.56E-05 1.09E-04 1.23E-04 1.36E-04 0.00015 50
1.00E-05 1.85E-05 2.70E-05 3.55E-05 4.40E-05 5.25E-05 6.10E-05 6.95E-05 7.80E-05 8.65E-05 9.50E-05 75
7.30E-06 1.35E-05 1.96E-05 2.58E-05 3.20E-05 3.82E-05 4.43E-05 5.05E-05 5.67E-05 6.28E-05 6.90E-05 100
5.90E-06 1.06E-05 1.53E-05 2.00E-05 2.47E-05 2.95E-05 3.42E-05 3.89E-05 4.36E-05 4.83E-05 5.30E-05 125
4.80E-06 9.02E-06 1.32E-05 1.75E-05 2.17E-05 2.59E-05 3.01E-05 3.43E-05 3.86E-05 4.28E-05 4.70E-05 150
4.00E-06 7.50E-06 1.10E-05 1.45E-05 1.80E-05 2.15E-05 2.50E-05 2.85E-05 3.20E-05 3.55E-05 3.90E-05 175
3.60E-06 6.54E-06 9.48E-06 1.24E-05 1.54E-05 1.83E-05 2.12E-05 2.42E-05 2.71E-05 3.01E-05 3.30E-05 200
Table 18 - Viscosity data used for the derivation of the viscosity equation
95 ©Ameron 2003 - FP914 03/03

No Flow Rate Flow Rate Pressure Drop Pressure Drop Pressure Drop Velocity
12 in. Steel Pipe 12. in SSL Pipe 10. in SSL Pipe 10 in. SSL Pipe
(lb/sec)_ (barrel/day) (lb/in2) (lb/in2) (lb/in2) (ft/sec)
1 50 12857 0.255433 0.123 0.287 1.523
2 100 25714 0.969462 0.429 1.004 3.045
3 130 33429 1.615535 0.690 1.616 3.959
4 150 38571 2.136843 0.894 2.096 3.959
5 200 51429 3.757076 1.508 3.541 6.090
6 250 64286 5.830024 2.265 5.324 7.613
7 300 77143 8.35563 3.161 7.436 9.136
8 330 84857 10.086029 3.763 8.857 10.049
9 350 90000 11.331719 4.192 9.869 10.658
10 400 102857 14.762765 5.355 12.616 12.181
11 450 115714 18.646406 6.648 15.673 13.704
12 495 127286 22.528647 7.921 18.686 15.074
13 500 128571 22.982637 8.069 19.036 15.226
Pressure Drop 12in. Steel Pressure Drop 10in. SSL Pressure Drop 12in. SSL
25
20
Pressure Drop (lb/sq in)
15
10
0
0 20000 40000 60000 80000 100000 120000 140000
Flow Rate (barrels/day)
Figure 54 - Typical output of the flow calculations using the Mathcad® workbook
96 ©Ameron 2003 - FP914 03/03

The viscosity equation will be determined for the discrete data available at AMERON for the
sp grade 0.86 crude oil as shown in Table 18. AMERON International has the capability to
determine an analytical equation for any data presented in a similar form to the one in the
mentioned table.
The equation for the kinematic viscosity of a liquid has the form of a generalized exponential
function producing a family of curves:
ν ( Percent, Temp ) = (a1 ⋅ Percent + b1 ) ⋅ exp(Temp ⋅ (a 2 ⋅ exp( Percent ⋅ b2 ) + c 2 )) +

(6)
+ (a 3 ⋅ Percent + b3 )
Then, in order to determine the required equation, we need to calculate the parameters ai , bi .
This can be accomplished by a series of curve fittings as shown in Figure 55, Figure 56 and
Figure 57:
• Figure 55 – represents the curve fitting of the data shown in Table 1 with a family of
curves as described by equation (3) above;
• Figure 56 – represents the curve fitting of the linear portion of the equation represented
by the sets of parameters a1 ,b1 and a3 ,b3 ;
• Figure 57 – represents the curve fitting of the exponential portion of the equation
represented by the set a 2 , b2 , c 2 .
The graphs can also be used to determine the accuracy of the derived equation. It is beyond the
scope of this section to detail such calculations, but AMERON can provide the necessary
information with respect to accuracy for a given set of experimental data.
For the curve fittings described above the values of the parameters can be easily calculated
using available techniques. The determined values are:
a1 = 3.6275 * 10 −6
b1 = 3.6515 * 10 −5
a 2 = 6.6981 * 10 −4
b2 = −0.075042
c 2 = −0.024549
a 3 = 2.9965 * 10 −7
b3 = 3.6608 * 10 −6
97 ©Ameron 2003 - FP914 03/03

Combining the above constants into equation (6) results in the following equation for viscosity
as a function of temperature and percentage of water-cut in crude oil.
ν ( Percent, Temp) = (3.6275 ⋅ 10 −6 ⋅ Percent + 3.6515 ⋅ 10 −5 ) ⋅

⋅ exp(Temp ⋅ (6.6981 ⋅ 10 − 4 ⋅ exp( Percent ⋅ (−0.075042)) − 0.024549)) + (7)
−7 −6
+ (2.9965 ⋅ 10 ⋅ Percent + 3.6608 ⋅ 10 )
Using equation (7) for a pipeline transporting crude oil with 30% water-cut at an operating
temperature of 150oF, the equation yields:
ν = (3.6275 ⋅ 10 −6 ⋅ 30 + 3.6515 ⋅ 10 −5 ) ⋅ exp(150 ⋅ (6.6981 ⋅ 10 −4 ⋅ exp(30 ⋅ (−0.075042)) −

− 0.024549)) + (2.9965 ⋅ 10 −7 ⋅ 30 + 3.6608 ⋅ 10 −6 )
or
ft 2
ν = 1.634636 ⋅ 10 −5
sec
The calculated value has an error of 7% compared to the tabulated data, and as seen from the
graphs, it can be assumed to represent an average error.
Combining equations (1), (2) and (3) the Reynolds’s number can be calculated as follows:
4 ⋅ FR
Re = (8)
π ⋅ ρ ⋅ ID ⋅ν ( Percent, Temp)
where the viscosity is a calculated value like the one in the previous example.
Using the Reynolds’s number as calculated in equation (8), the friction factor can be calculated
using equation (4) and finally pressure loss can be calculated using equation (5)
A standalone workbook is available from AMERON International. This workbook is also
included as a section of the general MathCAD® workbooks created for the calculation of
buried and above ground SSL pipelines .
In conclusion, a completely analytical method is provided in order to perform basic flow
calculations for an SSL pipeline. This makes possible the implementation of such calculations
in a MathCAD® workbook requiring only basic data input from the user.
As shown in Figure 54, due to the smooth internal wall of the SSL pipe, an SSL pipe that is one
diameter smaller, usually has comparable flow to the steel alternative.
98 ©Ameron 2003 - FP914 03/03

3
1 .10
yw
yo
4
e0( x) 1 .10
e10 ( x)
e20 ( x)
Kinematic Visacosity
e30 ( x)
e40 ( x)
e50 ( x)
e60 ( x)
e70 ( x)
e80 ( x) 5
1 .10
e90 ( x)
eo( x)
6
1 .10
0 25 50 75 100 125 150 175 200 225 250
xw , xo , x
Temperature, deg. F
Figure 55 - Graph of Interpolated Kinematic Viscosities

General engineering experience shows that the kinematic viscosity behaves like an exponential
function of temperature. This is considered to be true for water, crude oil and any mixture of
water and crude oil. The blue curve represents the viscosity function for water. The red curve
represents the viscosity function for oil. The magenta curves represent 10% increments of
water-cut in crude oil.
99 ©Ameron 2003 - FP914 03/03

4
Figure 56 - Graph of the 4 .10
4
Linear Coefficients 3.6 .10
4
The red line represents the 3.2 .10
4
y0 2.8 .10
curve fitting for the
set of parameters . 4
y2 2.4 10
a1 ,b1 (first linear 4
.
portion of the l0( x) 2 10
4
equation). The blue 1.6 .10
l2( x) 4
line represents the 1.2 .10
curve fitting of the set of 8 .10
5
parameters a 3 ,b3 (second 5

4 .10
linear portion of the
0
equation). 0 10 20 30 40 50 60 70 80 90 100
vx, vx2, x
Figure 57 - Graph of the Exponential Coefficients

The graph represents the curve fitting of the function used to determine the coefficients of the
exponential portion of the equation for kinematic viscosities, the parameters set a 2 , b2 , c 2 .
0.023
0.02317
0.02334
0.02351
0.02368
y1
0.02385
e1( x)
0.02402
0.02419
0.02436
0.02453
0.0247
20 10 0 10 20 30 40 50 60 70 80 90 100
vx, x
100 ©Ameron 2003 - FP914 03/03

8.2 INTEGRITY MONITORING

Figure 58 and Figure 59 show a typical integrity monitoring setup for a subsea SSL pipeline.
This type of installation can be used for ongoing monitoring the integrity of the internal and
external GRE liner and jacket in a subsea SSL pipeline, while the pipeline is producing and
without disturbing flow inside the pipe.
The test station in Figure 59 contains four lead wires from each group of pipes to minimize the
number of lead wires required. A typical “group of pipe” could be 1-km long and would
include two stainless steel reference probes. The four lead wires from each group include one
black wire from each of the two ends of the group of pipe, a green wire from the inside
stainless reference probe and a red wire from the outside stainless reference probe. All the pipe
in a group are jumpered together at the joints to so they are electrically continuous with each
other and can be all monitored at one time. If some form of damage is detected by the integrity
monitoring station, diver or ROV inspection will be required to determine the individual pipe
with a problem, but the effort will be significantly reduced and continuous ongoing monitoring
is possible.
Two types of reference probes are required. The first reference probe monitors for inside
corrosion and must be installed through the Coil-Lock bell. The inside reference probes must
be immersed in a conductive fluid on the inside of the pipe to be effective. In an oil/water
emulsion, where the entire fluid is conductive, the probe can be placed at the top of the pipe. In
stratified oil/water or gas/oil/water flows, the probe needs to be placed at or near the bottom of
the pipe, so that it is in contact with the conductive water layer. The second probe monitors for
outside corrosion. This probe can be strapped anywhere to the outside of a pipe, preferably in
the middle of the “group” of pipe to be monitored. The sea water is sufficiently conductive to
provide for monitoring a group of pipe.
8.2.1 Verification of Electrical Continuity.

Electrical continuity should be checked prior to integrity monitoring of the protective GRE
layers. This can be accomplished by verifying the electrical connections are continuous and the
reference probes are operating properly as follows:
1. Verify continuity of the black twisted pair copper cable connection to the monitored group
of pipe. This is accomplished by connecting the test leads from a high input impedance
voltmeter, to the two black binding posts corresponding to the pipe section, as shown in
Figure 58. The voltmeter should have an impedance >10 mega ohms, with increments of
0.01 mV.
2. Switch on the voltmeter and set it to read resistance. Resistance reading should be <100
ohm. This indicates that the connection of the cable to the pipe and the jumpers from pipe
to pipe are electrically continuous.
3. Verify the stability of the potential between the stainless steel reference probes by
connecting the test leads from the voltmeter to two green binding posts corresponding to the
inside stainless steel reference probes on adjacent pipe groups.
4. Switch on the voltmeter and set the range to 0.01 mV. If the potential does not deviate
more than 0.01 mV during the initial 2 to 5 seconds of connection, the potential is
considered stable and you can proceed to step 1 for internal monitoring. If the potential
101 ©Ameron 2003 - FP914 03/03

changes by more than 0.03 mV during the initial 2 to 5 seconds of connection, the potential
is considered unstable and the source of the instability must be determined and the lack of
continuity fixed prior to continuing. Possible sources of the instability are a bad electrical
connection between the green binding posts and the stainless steel probes.
5. Verify the stability of the potential between the stainless steel reference probes by
connecting the test leads from the voltmeter to two red binding posts corresponding to the
inside stainless steel reference probes on adjacent pipe groups.
6. Switch on the voltmeter and set the range to 0.01 mV. If the potential does not deviate
more than 0.01 mV during the initial 2 to 5 seconds of connection, the potential is
considered stable and you can proceed to step 1 for external monitoring. If the potential
changes by more than 0.03 mV during the initial 2 to 5 seconds of connection, the potential
is considered unstable and the source of the instability must be determined and the lack of
continuity fixed prior to continuing. Possible sources of the instability are a bad electrical
connection between the green binding posts and the stainless steel probes.
8.2.2 Monitoring a Subsea SSL Pipe

For a subsea pipelines, internal or external integrity of the protective GRE layers can be
monitored, on groups of pipe, as outlined below.
Monitoring of the Internal GRE Protective Liner:

1. Connect a lead wire from a high input impedance (>10 mega ohms) voltmeter, with
increments of 0.01 mV, to the black binding post which is connected to the subsea pipe
group to be monitored as shown in Figure 58.
2. Connect a lead wire from the other terminal of the voltmeter to the green binding post,
which is connected to the “inside monitoring” stainless steel reference probe for the group
of pipe to be monitored (the probe connected though the Coil-Lock bell to the inside fluid).
3. Switch on the voltmeter. If the potential changes by more than 0.03 mV during the initial 2
to 5 seconds of connection, the potential is considered “unstable”. The internal steel
surface of the SSL pipe is being adequately protected by GRE liner. If the potential does
not deviate more than 0.01 mV during the initial 2 to 5 seconds of connection, the potential
is considered “stable” and the internal GRE liner is no longer protecting the steel in the SSL
pipe wall.
Monitoring of the External GRE Protective Jacket:

1. Disconnect the lead wire from the “inside monitoring” stainless steel reference probe (green
terminal) and connect the lead wire to the red terminal for the “outside monitoring”
stainless reference probe (the one strapped to the outside of the pipe).
2. Connect the other terminal of the voltmeter to the group of SSL pipes (black terminal)
being monitored.
3. Switch on the voltmeter. If the potential changes by more than 0.03 mV during the initial 2
to 5 seconds of connection, the potential is considered “unstable” and the external GRE
jacket is protecting the steel in the SSL pipe wall. If the potential does not deviate by more
than 0.01 mV during the initial 2 to 5 seconds of connection, the potential is considered
“stable” and the external GRE jacket is no longer protecting the steel in the SSL pipe wall.
102 ©Ameron 2003 - FP914 03/03

Figure 58 - Permanent Internal and External Integrity Monitoring Setup for Subsea
Pipelines
Figure 59 - Detail of the Monitoring Station for Internal and External Integrity
Monitoring
103 ©Ameron 2003 - FP914 03/03

8.3 REPAIRS
8.3.1 Temporary Repair of Minor Local Damage

Temporary repair of minor damage can be accomplished as follows:
1. Drill through the affected area to make a round hole as shown in item 1 of Figure 60.
2. Install an expansion plug, as shown in item 2 of Figure 60. The plug is backed up by a steel
collar to prevent blowout (items 3 and 4 of Figure 60). Tight expansion of the plug will
stop internal leakage, but this should not be considered a permanent repair.
1. Drill through effected area to make 2. Install clamp with attached

a round hole. expansion plug.
3. Assemble clamp with provided bolts. 4. Tighten plug nut and hand-
tighten clamp nuts.
Figure 60 - Expansion Plug, Backed-Up by Metal Clamp
104 ©Ameron 2003 - FP914 03/03

8.3.2 Repair by Pipe Replacement

Figure 8 shows a typical repair pipe. The pipe can be collapsed axially, for lateral insertion
into the existing pipeline. Repair is accomplished as follows:
1. Cut a 3-foot section from the center of the damaged pipe. This will allow removal of the
bell and spigot without having to disturb the location of the undamaged pipeline.
2. Remove the bell and spigot, of the damaged pipe, from the existing pipeline. If the line has
been in service for an extended period, the bell and spigot may have to be split
longitudinally, in order to remove.
3. Disassemble the repair pipe by removing split flanges on fixed side. Thread one half of
repair pipe to the existing pipeline, slide collar laterally past the end of now assembled
repair pipe segment. Thread the other half of the repair pipe into the existing line. Slide
collar back to seal both ends of repair pipe and torque bolts to place the pipeline in tension.
Come-along may have to be used to move the collar back and forth.
Figure 9 shows an alternate form of repair. Repair using this technique is similar to that
outlined above and is accomplished as follows:
1. Cut a 3-foot section from the center of the damaged pipe. This will allow removal of the
bell and spigot without having to disturb the location of the undamaged pipeline.
2. Remove the bell and spigot, of the damaged pipe, from the existing pipeline. If the line has
been in service for an extended period, the bell and spigot may have to be split
longitudinally, in order to remove.
3. Thread on flange by Coil-Lock adapters (Figure 4 and Figure 5) to both ends of the
undamaged pipeline. Measure the distance between the flanges. Fabricate a pipe spool to
these dimensions and insert it between the flanged adapters on the ends of the undamaged
pipeline as shown in Figure 9
8.4 EROSION RATES WITH LIQUIDS AND ABRASIVE FLUIDS

SSL has abrasion resistance similar to conventional GRE pipe. It is difficult to predict erosion
rates. Erosion rates are affected by the fluid velocity, particle size, density and shape. As a
general guideline, liquid flow velocities should be held below 5 meters/sec, while slurry
velocity should be held below 2.5 meters/sec. Table 19 gives relative erosion rates for various
materials. The following equation, in conjunction with the values contained in Table 19, can be
used to roughly estimate relative erosion rates for SSL compared to steel pipe in similar
application.
105 ©Ameron 2003 - FP914 03/03

n
General Equation: Erosion Rate = K(V)f(B)
Where: K = a constant associated with the slurry material

V = partical velocity in meters/sec
n = a constant associated with the pipe or lining material
f(B) = a function associated with the angle of impingement
V = 2 meter/sec V = 3 meter/sec V = 4 meter/sec V = 5 meter/sec V = 6 meter/sec

Wear Wear Wear Wear Wear
Rate Rate Rate Rate Rate
Relative Relative Relative Relative Relative
to 6Mo to 6Mo to 6Mo to 6Mo to 6Mo
n n n n n
Material n V Steel V Steel V Steel V Steel V Steel
6Mo Steel 2.67 6.36 1.00 18.79 1.00 40.50 1.00 73.49 1.00 119.58 1.00
Epoxy GRE 3.70 13.00 2.04 58.26 3.10 168.90 4.17 385.65 5.25 757.11 6.33
Aluminum 3.80 13.93 2.19 65.02 3.46 194.01 4.79 452.99 6.16 905.68 7.57
Stainless Steel 2.20 4.59 0.72 11.21 0.60 21.11 0.52 34.49 0.47 51.51 0.43
HDPE Lined Steel 3.31 9.93 1.56 38.04 2.02 98.65 2.44 206.58 2.81 377.87 3.16
Polyurea Lined GRE 2.32 4.98 0.78 12.73 0.68 24.79 0.61 41.55 0.57 63.38 0.53
Ceramic Lined GRE 3.31 9.93 1.56 38.04 2.02 98.65 2.44 206.58 2.81 377.87 3.16
The values for n for all materials except HDPE, ceramic lined GRE and the new GRE liner are obtained from the literature. The values of n fo
new GRE liner, ceramic lined GRE and HDPE are set to achieve the relative wear rates obtained during the initial testing. Relative wear rate
HDPE and ceramic lined GRE are half that compared to standard GRE, while the new liner has a wear rate approximately 1/12 of that for sta
unlined GRE at particle velocities around 6 meters/sec.
Table 19 - Relative Erosion Rate Values for Selected Materials
8.5 TRANSIENT PRESSURE (WATER HAMMER)

Water hammer results from abrupt changes in fluid velocity in the piping system. Transient
pressure is the rapidly moving wave that can increase or decrease pressure in a piping system
resulting in rupture or collapse. Causes of water hammer include sudden closure of a valve.
The magnitude of water hammer is a function of:
1. Fluid properties
2. Fluid velocity
3. Pipe wall modulus of elasticity
4. Pipe wall thickness
5. Length of the pipeline
6. Speed at which the momentum of the fluid changes.
The magnitude of the pressure wave for SSL is much higher than conventional GRE, due to the
much higher equivalent modulus of the steel reinforced pipe wall. The Talbot equation
calculates surge pressure due to changes in velocity as follows:
a SG
Ps ⋅
2
⋅∆v Equation 4-21 of AWWA Manual M45. This is the
gc ft⋅in
2.3 ⋅
maximum surge pressure, assuming full instantaneous change
lb in velocity.
106 ©Ameron 2003 - FP914 03/03

Where:
in
12 ⋅
ft Equation 4-22 of AWWA Manual M45
a
ρ Fluid  1 ID 
⋅ + 
gc  k ESSLHoop ⋅twall 
ID Pipe inside diameter, in

Ps Surge pressure, psig
a Wave velocity, ft/s
∆v Change in flow velocity, ft/s
ρ
Fluid Customer specified fluid density, lbm/ft3
SG Fluid specific gravity, dimensionless
k Fluid bulk modulus, psi. Water is 300,000 psi.
gc Gravitational constant, 32.2 lbm-ft/lbf-s2
ESSLHoop Hoop modulus of the SSL pipe wall, psi
t wall Total thickness of the SSL pipe wall, in.
The MathCAD workbook automatically calculates the surge pressure. Maximum allowable
surge pressure is defined as follows:
P s ≤ 1.25P Rating − P Operating
107 ©Ameron 2003 - FP914 03/03

9. OTHER REFERENCE DOCUMENTS

The following documents are available for additional information on SSL:
9.1 PRODUCT LITERATURE

1. Bondstrand Product Data Sheet for SSL. FP 353 A (07/99)
2. Installation Guide for SSL. FP815 (4/99)
3. Bondstrand Guide Specification for SSL. FP819 (1/99)
9.2 TEST REPORTS

1. Report on Cold (-40°C) Burst Test of 249 mm Diameter SSL Pipe No. 201 384-TR/1/E000
by British Aerospace Defense Limited, Royal Ordnance, Rocket Motors Division (1996)
2. Impact Testing of a 4-Layer RO Pipe Section by British Aerospace Defense Limited, Royal
Ordnance, Rocket Motors Division (1995)
3. PDVSA INTEVEP Technical report on evaluation of SSL pipe under H2S Aggressive
Environment (November 1988)
4. PDVSA INTEVEP Evaluation of two polymers used in Coil-Lock Joining system (October
2000)
5. PDVSA INTEVEP Field Test Evaluation of SSL Pipe (December 1999)
6. Bondstrand Technical Bulletin FP588 – Failure characterization of Bondstrand glassfiber
reinforced epoxy pipe (1995)
7. Steel Strip Laminate (SSL) Pipe a New, High-Performance, Hybrid Composite Pipe
9.3 ANALYSIS AND TECHNICAL REPORTS

1. Preliminary Status report on SSL Pipe (August, 1996)
2. Leakage resistance and Structural Integrity of SSL Composite Pipe and its Joint under
Combined Axial Loading,, Internal Pressure and Makeup Turns by Professor S.S. Wang
Composites Engineering and Applications Center (CEAC, Houston, TX, February 2001)
3. Structural Integrity of SSL Composite Pipe for High Pressure Applications: Deformation,
Failure Prediction and Design Verification. (CEAC, Houston, TX, November 2000)
9.4 FOOTNOTES
1. American Petroleum Institute, Specification for High Pressure Fiberglass Line Pipe,
Second Edition, April 1, 1995.
2. American Petroleum Institute, Specification for Line Pipe, Fortieth Edition, November 1,
1992.
3. The American Society of Mechanical Engineers, Pipeline Transportation Systems for
Liquid Hydrocarbons and other liquids, ASME B31.4 – 1998 Edition.
4. The American Society of Mechanical Engineers, Gas Transmission and Distribution Piping
Systems, ASME B31.8 – 1999 Edition.
5. Shell International Petroleum, Glass-Reinforced Plastic Pipelines and Piping Systems,
DEP 31.40.10.19-Gen, December 1998.
6. The American Society for Testing and Materials, Classification of Soils for Engineering
Purposes, D 2487 –93, 1998 Annual Book Volume 04.08.
108 ©Ameron 2003 - FP914 03/03

7. American Water Works Association, Glass-Fiber-Reinforced Thermosetting-Resin

Pressure Pipe, ANSI/AWWA C950-95, October 1, 1996.
8. The International Standard Organization, Petroleum and natural gas industries – Glass-
reinforced plastic (GRP), November 2001.
9. The American Society for Testing and Materials, D 2992 – 96, Practice for Obtaining
Hydrostatic or Pressure Design Basis for “Fiberglass” (Glass-Fiber-Reinforced
Thermosetting-Resin) Pipe and Fittings.
10. "Assessment and Analysis of Unsupported Subsea Pipeline Spans", U.S. Department of the
Interior, Minerals Management Service, December 23,1997
11. DNV-OS-F101, "Submarine Pipeline Systems", Det Norske Veritas Standard, Norway,
2000
12. DNV-OS-F201, "Dynamic Risers", Det Norske Veritas Standard, Norway, 2002
13. DNV-RP-C202, "Buckling Strength of Shells", Det Norske Veritas Recommended
Practice, Norway, 2000
14. DNV-RP-E305, Onbottom Stability Design of Submarine Pipelines", Det Norske Veritas
Recommended Practice, Norway, October 1998
15. DNV-RP-F105, "Free Spanning Pipelines", Det Norske Veritas Recommended Practice,
Norway, March 2002
16. DNV-CN-30.5, "Environmental Conditions and Environmental Loads", Det Norske
Veritas Classification Notes, Norway, March 2000
17. "Assessment of AMERON SSL Pipe for Wonnich Pipeline", Australian Marine and
Offshore Group PTY LTD, April 1998
18. Antonio Carlos Fernandes, "Up-to-date Program for Preliminary Analysis of Flexible and
Steel Catenary Risers Using Catenary Concepts", Proceedings of the 10th International
Symposium on Offshore Engineering, University of Rio de Janeiro, September 1997
19. S. Hauch, Y. Bai, "Bending Moment Capacity of Pipes", Offshore Mechanical and Arctic
Engineering, July 11-16, 1999.
20. NASA SP-8007, "Buckling of Thin-Walled circular cylinders", NASA Space vehicle
Design Criteria, August 1968
21. “Subsea Pipeline Systems and Risers”. American Bureau of Shipping (ABS), 2001
10. APPENDIX
10.1 EQUIVALENT ALLOWABLE DESIGN ENVELOPES

The following design envelopes are required to check FEA analysis output against allowable
stress combinations versus operating temperature and number of steel layers.
109 ©Ameron 2003 - FP914 03/03

NOTE: This page intentionally left blank.
110 ©Ameron 2003 - FP914 03/03


12,000
10,000 Design Stress for Hoop-Loaded 3-Strip SSL Pipe at 21°C
Hoop Stress = 11,500 psi

8,000
Axial Stress = 5,750 psi

6,000
4,000
2,000 Hoop Stress = 18,484 psi

13,229 psi 15,498 psi

11,189 psi
0
0 4,000 8,000 12,000 16,000 20,000 24,000 28,000 32,000
Figure 61 - Allowable Design Stress for the GRE Component of SSL (English Units), with f2=0.67
111 ©Ameron 2003 - FP914 03/03


70
60
50

40
30
20

77.2 MPa 91.2 MPa 106.9 MPa
0
0 15 30 45 60 75 90 105 120 135 150 165
Figure 62 - Allowable Design Stress for the GRE Component of SSL (SI Units), with f2=0.67
112 ©Ameron 2003 - FP914 03/03


12,000
10,000 Design Stress for Hoop-Loaded 3-Strip SSL Pipe at 21°C

8,000 Axial Stress = 7,123 psi

6,000
4,000
2,000 Hoop Stress = 22,898 psi

13,860 psi 16,388 psi 19,199 psi
0
0 4,000 8,000 12,000 16,000 20,000 24,000 28,000 32,000
Figure 63 - Allowable Design Stress for the GRE Component of SSL (English Units), with f2=0.83
113 ©Ameron 2003 - FP914 03/03

80
70
60

40
30
20

10
95.6 MPa 113.0 MPa 132.4 MPa
0
0 20 40 60 80 100 120 140 160 180 200 220
Figure 64 - Allowable Design Stress for the GRE Component of SSL (SI Units), with f2=0.83
114 ©Ameron 2003 - FP914 03/03

30,000
Total Axial Stress as Pressure Vessel, psi
25,000
20,000
15,000
10,000
5,000 3 Steel Strips 4 Steel Strips

0
0 10,000 20,000 30,000 40,000 50,000 60,000
SSL Composite Hoop Stress, psi
Figure 65 - Equivalent Allowable Design Stress for the Overall SSL Pipe Wall at 70°F (English Units) with f2=0.67
115 ©Ameron 2003 - FP914 03/03

25,000
20,000
15,000
10,000
5,000
0
0 10,000 20,000 30,000 40,000 50,000
116 ©Ameron 2003 - FP914 03/03

20,000
18,000
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
0 10,000 20,000 30,000 40,000 50,000
117 ©Ameron 2003 - FP914 03/03

210
180
150
120
90
60

30
0
0 60 120 180 240 300 360 420
Figure 68 - Equivalent Allowable Design Stress for the Overall SSL Pipe Wall at 21°C (SI Units) with f2=0.67
118 ©Ameron 2003 - FP914 03/03

180
150
120
90
60
30 3 Steel Strips 4 Steel Strips

0
0 60 120 180 240 300 360
119 ©Ameron 2003 - FP914 03/03

150
120
90
60
30
0
0 60 120 180 240 300
120 ©Ameron 2003 - FP914 03/03

40,000
35,000
30,000
25,000
20,000
15,000
10,000

0
0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000
121 ©Ameron 2003 - FP914 03/03

30,000
25,000
20,000
15,000
10,000

0
0 10,000 20,000 30,000 40,000 50,000 60,000 70,000
122 ©Ameron 2003 - FP914 03/03

25,000
20,000
15,000
10,000
5,000
0
0 10,000 20,000 30,000 40,000 50,000 60,000
123 ©Ameron 2003 - FP914 03/03

270
240
210
180
150
120
90
60
0
0 60 120 180 240 300 360 420 480 540
124 ©Ameron 2003 - FP914 03/03

210
180
150
120
90
60

30
0
0 60 120 180 240 300 360 420 480
125 ©Ameron 2003 - FP914 03/03

180
150
120
90
60

0
0 60 120 180 240 300 360 420
126 ©Ameron 2003 - FP914 03/03

10.2 ALLOWABLE AXIAL TENSION AND BEND RADIUS COMBINATIONS FOR S- AND J-LAY
8-inch SSL Pipe at 70F

3000
SSL3
2500 SSL4
SSL5
Bend Radius (Feet)
2000 SSL6
SSL7
1500 SSL8
SSL9
1000 SSL10
500
0
0.00 50.00 100.00 150.00 200.00 250.00
Axial Load (kips)
Figure 77 - Allowable Axial Load vs. Bend Radius Combinations for 8-inch SSL During Installation (J-Lay or S-Lay)
127 ©Ameron 2003 - FP914 03/03

3500
3000 SSL3
SSL4
2500 SSL5
Bend Radius (Feet)
SSL6
2000 SSL7
SSL8
1500 SSL9
SSL10
1000
500
0
0.00 50.00 100.00 150.00 200.00 250.00 300.00
Axial Load (kips)
128 ©Ameron 2003 - FP914 03/03

12-inch SSL at 70F
4000
SSL3
3500
SSL4
3000 SSL5
Bend Radius (Feet)
SSL6
2500 SSL7
SSL8
2000
SSL9
1500 SSL10
1000
500
0
0.00 50.00 100.00 150.00 200.00 250.00 300.00 350.00
Axial Load (kips)
129 ©Ameron 2003 - FP914 03/03


4500
4000 SSL3
SSL4
3500
SSL5
Bend Radius (Feet)
3000 SSL6
SSL7
2500
SSL8
2000 SSL9
1500 SSL10
1000
500
0
0.00 50.00 100.00 150.00 200.00 250.00 300.00 350.00 400.00
Axial Load (kips)
130 ©Ameron 2003 - FP914 03/03


5000
4500 SSL3
4000 SSL4
SSL5
3500
Bend Radius (Feet)
SSL6
3000 SSL7
2500 SSL8
2000 SSL9
SSL10
1500
1000
500
0
0.00 50.00 100.00 150.00 200.00 250.00 300.00 350.00 400.00 450.00
Axial Load (kips)
131 ©Ameron 2003 - FP914 03/03


6000
SSL3
5000 SSL4
SSL5
Bend Radius (Feet)
4000 SSL6
SSL7
3000 SSL8
SSL9
2000 SSL10
1000
0
0.00 100.00 200.00 300.00 400.00 500.00
Axial Load (kips)
132 ©Ameron 2003 - FP914 03/03


6000
SSL3
5000 SSL4
SSL5
Bend Radius (Feet)
4000 SSL6
SSL7
3000 SSL8
SSL9
2000 SSL10
1000
0
0.00 100.00 200.00 300.00 400.00 500.00 600.00
Axial Load (kips)
133 ©Ameron 2003 - FP914 03/03


7000
SSL3
6000
SSL4
SSL5
5000
Bend Radius (Feet)
SSL6
4000 SSL7
SSL8
3000 SSL9
SSL10
2000
1000
0
0.00 100.00 200.00 300.00 400.00 500.00 600.00
Axial Load (kips)
134 ©Ameron 2003 - FP914 03/03


8000
SSL3
7000
SSL4
6000 SSL5
Bend Radius (Feet)
SSL6
5000
SSL7
4000 SSL8
SSL9
3000
SSL10
2000
1000
0
0.00 100.00 200.00 300.00 400.00 500.00 600.00 700.00
Axial Load (kips)
135 ©Ameron 2003 - FP914 03/03


9000
8000 SSL3
SSL4
7000
SSL5
Bend Radius (Feet)
6000 SSL6
SSL7
5000
SSL8
4000 SSL9
3000 SSL10
2000
1000
0
0.00 100.00 200.00 300.00 400.00 500.00 600.00 700.00 800.00
Axial Load (kips)
136 ©Ameron 2003 - FP914 03/03


10000
9000 SSL3
8000 SSL4
SSL5
7000
Bend Radius (Feet)
SSL6
6000 SSL7
5000 SSL8
4000 SSL9
SSL10
3000
2000
1000
0
0.00 100.00 200.00 300.00 400.00 500.00 600.00 700.00 800.00
Axial Load (kips)
137 ©Ameron 2003 - FP914 03/03


10000
9000 SSL3
8000 SSL4
SSL5
7000
Bend Radius (Feet)
SSL6
6000 SSL7
5000 SSL8
4000 SSL9
SSL10
3000
2000
1000
0
0.00 100.00 200.00 300.00 400.00 500.00 600.00 700.00 800.00 900.00
Axial Load (kips)
138 ©Ameron 2003 - FP914 03/03


12000
SSL3
10000 SSL4
SSL5
Bend Radius (Feet)
8000 SSL6
SSL7
6000 SSL8
SSL9
4000 SSL10
2000
0
0.00 200.00 400.00 600.00 800.00 1000.00
Axial Load (kips)
139 ©Ameron 2003 - FP914 03/03


14000
SSL3
12000
SSL4
SSL5
10000
Bend Radius (Feet)
SSL6
8000 SSL7
SSL8
6000 SSL9
SSL10
4000
2000
0
0.00 200.00 400.00 600.00 800.00 1000.00 1200.00
Axial Load (kips)
140 ©Ameron 2003 - FP914 03/03

141 ©Ameron 2003 - FP914 03/03

10.3 GUIDELINES FOR THE UKOOA ANALYSIS OF A PIPELINE AS

IMPLEMENTED IN CAESAR II®
The explanations in this section are limited in detail, assuming that the user has some basic
knowledge of the CAESAR II® software.
STEP 1: Configuration of the Model

Choose (or define) the appropriate *.FRP file. The files have the following naming convention:
AxxSSLyy.FRP, and a sample file is shown in Figure 91. In this notation, xx determines the
pipe size and yy determines the number of steel layers. For instance, the file A16SSL10.FRP is
for a 16-inch diameter pipe with 10 steel layers, and the file A08SSL05.FRP designates an 8-
inch pipe with 5 steel layers. For the user’s convenience, and also for a consistent and unitary
usage of the software we also provide AMERON.FIL file. This file is necessary to ensure the
correct conversion of the metric to the English units internally used by CAESAR II®. The rest
of the configuration is standard for FRP pipe and self explanatory by browsing the tabs of the
pull-down menu.
Figure 91 - SSL Pipe Configuration File

STEP 2: Define the Geometry
In this step the CAESAR II® geometry is defined. The analysis can be run without the
“allowable” stresses or with the “allowable” stresses. If the allowable code stresses are not
defined the user will have to determine the compliance with code manually as explained in
Section 6.4. Caution should be exercised in this case, as the calculated code stress, is not
correct.
141
STEP 3: Execute Analysis
After the geometry is complete, but before running the analysis, open the “Special Execution
Parameters” from the “Kaux” menu dialog screen. Please verify that the input values on this
dialog screen correspond with input values in the configuration file used in STEP 1 above.
This input dialog screen can also
become very useful in running
“what if” type analysis.
NOTE: If the dialog box for the

ambient temperature is left blank,
the value will default to 21°C.
This value is the “installation”
temperature and will have to be set
to operating temperature when
running the analysis for f 2 =0.67.
With this done, the model is ready

for pre- and post-processing, and
then the interpretation of the
results.
Figure 92 - Special Execution Parameters
142
Field testing Pipe system is designed for field testing with water at ambient temperature at 125% of
rated pressure.
Surge pressure Maximum allowable surge pressure is 125% of rated pressure at ambient
temperature
Conversions
1 psi = 6895 Pa = 0.07031 kg /cm2
1 bar = 105Pa = 14.5 psi = 1.02 kg /cm2
1 MPa = 1 N/mm2 = 145 psi = 10.2 kg /cm2
1 inch = 25.4 mm
1 Btu.in/ft2h°F = 0.1442 W/mK
°C = 5/9 (°F-32)
Important notice This product literature and the recommendations for usage it contains are
based on test data reasonably believed to be reliable. It is intended that this
literature be used by personnel having specialized training in accordance with
currently acceptable industry practice and normal operating conditions.
Variation in environment, changes in operating procedures, or extrapolation of
data may cause unsatisfactory results. We recommend that your engineers
verify the suitability of this product for your intended application. Since we
have no control over the conditions of service, we expressly disclaim
responsibility for the results obtained or for any consequential or incidental
damages of any kind incurred.
AMERON
INTERNATIONAL
Fiberglass Pipe Group Fiberglass Pipe Division Fiberglass Pipe Division Fiberglass Pipe Division
Headquarters The Americas Europe Asia
PO. Box 801148 PO. Box 878 Ameron B.V. Ameron (Pte) Ltd
Houston, TX 77280 Burkburnett, TX 76364 P.O. Box 6 No. 7A, Tuas Avenue 3
U.S.A. U.S.A. 4190 CA Geldermalsen Jurong
Phone: (+1) 713 690 7777 Phone: (+1) 940 569 1471 The Netherlands Singapore 639407
Fax: (+1) 713 690 2842 Fax: (+1) 940 569 2764 Phone: (+31) 345 587 587 Phone: (+65) 861 6118
Internet: www.ameron.com Fax: (+31) 345 587 561 Fax: (+65) 862 1302/861-7834
143

FP914

Uploaded by

FP914

Uploaded by

SSL Engineering Design Manual for Subsea Applications Rev. 1.

2. ADVANTAGES IN THE USE OF SSL AS A SUBSEA PIPELINE 1

3. SSL PIPE AND FITTING CONFIGURATIONS 2

3.2 Coil-Lock Joining Systems and Adapters 4

3.3 Pipe Configurations 7

3.4 Fitting Configurations 9

4. DESIGN OF SUBSEA PIPELINES 11

4.2 Directional Changes for Subsea Pipelines 11

4.3 Shore Approach for SSL Subsea Pipelines 13

4.5 Bottom Stability 18

5. INSTALLING SUBSEA SSL PIPELINES 22

i ©Ameron 2003 - FP914 03/03

5.4 Off Bottom Tow 31

5.5 Surface Tow 31

5.6 Mid Depth Tow 32

6. SSL SUBSEA PIPELINE ANALYSIS USING FEA PROGRAMS 34

6.2 Definition of Partial FActors f1, f2, f3, A1, A2 and A3 35

6.3 Equivalent Allowable Design Stress Envelope 39

6.5 Limitations in Existing FEA Analysis Software 45

6.6 Equivalent Elastic Properties of SSL 46

7. SSL SUBSEA PIPELINE ANALYSIS USING AMERON PROPRIETARY

7.2 FLow Diagrams for Structural Design 53

7.3 Inputs Required for the MathCAD Analysis 56

7.4 Analytical Approach and Equations Utilized in the MathCAD Analysis 60

ii ©Ameron 2003 - FP914 03/03

8. OTHER CONSIDERATIONS IN SSL PIPELINE DESIGN 92

8.2 INTEGRITY Monitoring 101

8.3 Repairs 104

8.4 Erosion Rates with Liquids and Abrasive Fluids 105

8.5 Transient Pressure (water hammer) 106

9. OTHER REFERENCE DOCUMENTS 108

9.2 Test reports 108

9.3 Analysis and technical reports 108

9.4 Footnotes 108

10. APPENDIX 109

iii ©Ameron 2003 - FP914 03/03

iv ©Ameron 2003 - FP914 03/03

v ©Ameron 2003 - FP914 03/03

vi ©Ameron 2003 - FP914 03/03

vii ©Ameron 2003 - FP914 03/03

viii ©Ameron 2003 - FP914 03/03

2. ADVANTAGES IN THE USE OF SSL AS A SUBSEA

1 ©Ameron 2003 - FP914 03/03

3. SSL PIPE AND FITTING CONFIGURATIONS

3.1 PIPE WALL

Figure 1 - SSL Pipe Wall

2 ©Ameron 2003 - FP914 03/03

3.1.2 Pressure Rating

Pipe SSL3 SSL4 SSL5 SSL6 SSL7 SSL8 SSL9 SSL10

3 ©Ameron 2003 - FP914 03/03

3.1.3 Temperature De-rating Factors

3.2 COIL-LOCK JOINING SYSTEMS AND ADAPTERS

4 ©Ameron 2003 - FP914 03/03

Figure 2 - Standard Coil-Lock® Joint for SSL Pipe

Figure 3 - Seal Designs Available for the Coil-Lock® Joint

5 ©Ameron 2003 - FP914 03/03

3.2.2 Threaded Alloy or Coated Steel Coil-Lock Adapters

Figure 4 - Coil-Lock Male Threaded Steel End Adapter

Figure 5 - Coil-Lock Female Threaded Steel End Adapter

6 ©Ameron 2003 - FP914 03/03

3.3 PIPE CONFIGURATIONS

3.3.1 Coil-Lock® Bell by Coil-Lock® Spigot SSL Pipe

Figure 6 - Coil-Lock® Bell by Coil-Lock® Spigot SSL Pipe

3.3.2 Coil-Lock® Spigot by Coil-Lock® Spigot SSL Pipe

Figure 7 - Coil-Lock® Spigot by Coil-Lock® Spigot SSL Pipe

7 ©Ameron 2003 - FP914 03/03

3.3.3 Repair Pipe

Figure 8 - Repair Pipe

Figure 9 - Repair Spool

8 ©Ameron 2003 - FP914 03/03

3.4 FITTING CONFIGURATIONS

9 ©Ameron 2003 - FP914 03/03

10 ©Ameron 2003 - FP914 03/03

4. DESIGN OF SUBSEA PIPELINES

4.2 DIRECTIONAL CHANGES FOR SUBSEA PIPELINES

4.2.2 Snake-Lay for Thermal Expansion Using “Flexure of the Pipe”