Stress-Based Design of Pipelines


10
Stress-Based Design of Pipelines


Mavis Sika Okyere


Ghana National Gas Limited Company, Accra, Ghana


10.1 Introduction


In stress-based design, the pipeline is designed so that the stress on the pipeline is maintained below a prescribed limiting value, the specified minimum yield stress (SMYS), by a safety factor. In this chapter, a reference is made to standards that are used in different jurisdictions, and some comparisons are made among British, Canadian, European, international, and U.S. standards. For more detailed information, readers should refer to the current editions of the relevant standards.


As discussed in CSA Z662:23, Section 4.3.5.2 [1], the design formula for a straight pipe is


(10.1)upper P equals StartFraction 2 italic upper S t Over upper D EndFraction times upper F times upper L times upper J times upper T

where P is the design pressure (MPa), S is the specified minimum yield strength (MPa), t is the design wall thickness (mm), D is the outside diameter of the pipe (mm), F is the design factor, L is the location factor, J is the joint factor, and T is the temperature factor.


10.2 Design Pressure


The design pressure is the pressure that is used in all equations and stress calculations and is the maximum internal pressure of the pipeline during its design life. The maximum allowable operating pressure (MAOP), maximum operating pressure (MOP), and surge pressure should be considered with the design pressure. Unduly high design pressures require the use of an excessively thick pipe.


For a gas pipeline, the pressure in the pipeline does not vary greatly due to elevation, but where the elevation is extreme (hundreds of meters) below the inlet point, the highest pressure in the pipeline and the location of the highest pressure need to be determined. For liquid pipelines, there is always a need to consider elevation, especially when the pipeline is below the inlet point.


Normally, it is better to keep a fixed design pressure for the design of the whole pipeline. However, for pipelines with large positive elevation change (i.e., hilly and mountainous zones), the design pressure can be reduced with elevation increases. This will result in different pipes for different sections, complicating the line pipe order and construction, and it makes any future uprating by adding additional pumping stations limited or impossible.


Normally, for gas pipelines, the maximum operating pressure does not exceed the design pressure.


10.2.1 Maximum Allowable Operating Pressure


As per Clause 805.214 of ASME B31.8, the maximum allowable operating pressure is defined as the maximum pressure at which a gas pipeline system is allowed to operate. The value of the MAOP varies and depends on the location and prescribed test pressure [2].


Due to lack of surge pressure in a gas pipeline, the MAOP should not be higher than the design pressure but is often set slightly (5% or less) below the design pressure. For liquid pipelines, the MAOP can be 10% or more below the design pressure due to surge problems. The difference between the MAOP and the design pressure permits shutdown alarms and other protective devices to be set to ensure that the pipeline does not exceed the design pressure [29].


(10.2)MAOP equals StartFraction test pressure left-parenthesis upper T upper P right-parenthesis Over pressure factor left-parenthesis upper P upper F right-parenthesis EndFraction

Table 10.1 Percentage of Overpressure Permitted [19]































ASME B31.8 ASME B31.4 PD 8010 IGE/TD/1 ISO 13623 EN 1594 CSA Z662
Relating to N/A DP DP MOP MAOP MOP MOP
% 0 10 10 10 10 15 10

10.2.2 Maximum Operating Pressure


The maximum operating pressure is equal to the maximum pressure to which the piping system will be subjected in operational conditions. This includes static pressure and the pressure required to overcome friction [2].


Above MOP, a warning alarm might be set, but this will not shut down the pipeline until the pressure increases above the MAOP. To increase the MOP above the MAOP, there is the need to retest the pipeline to a higher pressure in order to increase the MAOP.


10.2.2.1 Overpressure


The allowable overpressure varies between design codes. Table 10.1 outlines the percentage of overpressure that is permitted.


10.2.3 Surge Pressure


Surge pressures in a liquid pipeline are produced by a change in the velocity of the moving stream that results from shutting down of a pump station or pumping unit, closing of a valve, or blockage of the moving stream. Surge pressure decreases in intensity as it moves away from its point of origin [2].


The surge pressure depends on the density of fluid, velocity of fluid, pipe length, speed of closure or shutdown, fluid pressure, and sonic velocity of the fluid. The design codes have limiting values for surge pressure to be added to the MAOP. Surge pressure calculations should be made, and adequate controls and protective equipment should be provided.


Gas pipelines do not suffer excessively from surge due to compressibility within the fluid. However, this can become a major issue for liquid pipelines that have a relatively high fluid velocity (>2.5 m/s) or are subject to sudden closure of valves or pumps.


10.2.4 Test Pressure


Test pressure is set by the design codes to verify that the pipeline is fit for purpose and free from material or construction defects. The setting of the pressure test levels needs to follow the methodology of the design code (e.g., Clause 847.2 of ASME B31.8), but care also needs to be taken not to overstress the pipeline either at its lowest point or when the design calculations have used the empty weight of the pipe, as in gas pipes, and not the temporary weight of the hydrotest water [2].


Pipelines in mountainous or hilly regions are often sectioned into different lengths for testing.


(10.3)Test pressure equals upper M upper O upper P times upper P upper F comma 1.1 less-than upper P upper F less-than 1.4

The test pressures used in different situations are listed in Table 10.2.


10.3 Design Factor


Design factors (Table 10.3) have been used since pipeline design codes were first established to provide a defined level of safety and mechanical strength. The design factor to be used is sometimes limited by legislation in a particular country [29].


Table 10.2 Test Pressures


Source: Adapted from [2].




















Test Pressure (TP) Pressure Factor (PF)
Installed pipeline system TP = MAOP × 1.25 1.25
Offshore platform piping TP = MAOP × 1.4 1.4
Offshore pipeline risers TP = MAOP × 1.4 1.4

Table 10.3 Design Factor [29]




























ASME B31.8 ASME B31.4 PD 8010 IGE/TD/1 ISO 13623 EN 1594
Design factor (liquid) 0.72 0.72 0.77
Design factor (gas) 0.40–0.80 0.30–0.72 0.30–0.80 0.45–0.83 0.72

The design codes state where in this range a specific design factor should be used but generally allow a higher factor to be used, provided that this action is supported by a safety evaluation or is subject to scrutiny by the regulatory safety authority.


It is appropriate to specify the design factor at a level that takes into consideration the boundary between defect arrest and propagation, also known as the leak/break boundary.


The two important design factors are



  • 0.30 for pipelines operating in locations where a line break is not acceptable;
  • 0.80 for pipelines operating in an open country where a line break is unlikely but could be tolerated.

10.4 Determination of Components of Stress


10.4.1 Hoop and Radial Stresses


Hoop stress, σH, is the stress in a pipe of wall thickness t acting circumferentially in a plane perpendicular to the longitudinal axis of the pipe, produced by the pressure P of the fluid in a pipe of diameter D and is determined by Barlow’s formula [2]:



where σH is the hoop stress (MPa), P is the internal design pressure (gauge) (MPa), t is the pipe wall thickness (m), and D is the pipe diameter (m).


10.4.1.1 Thick Cylinders


The full Lamé equations are simplified for the design of thick-walled pipelines. A pipeline with D/t < 20 is known as a thick-walled pipeline. Taking into consideration a thick-walled pipe, subjected to an internal pressure, Pi, with zero external pressure [10],



(10.6)radial stress comma sigma Subscript normal r Baseline equals upper A minus StartFraction upper B Over r squared EndFraction

The two well-known conditions of stress that allow the Lamé constants A and B to be determined are


StartLayout 1st Row a t r equals upper R 1 sigma Subscript r Baseline equals upper P Subscript normal i Baseline 2nd Row a t r equals upper R 2 sigma Subscript normal r Baseline equals 0 EndLayout

(10.7)upper L a m ModifyingAbove normal e With ́ constant comma upper A equals minus upper P Subscript normal i Baseline left-parenthesis StartFraction upper R 1 squared Over upper R 1 squared minus upper R 2 squared EndFraction right-parenthesis


Therefore, from Equations (10.510.8), the radial and hoop stresses are calculated as


(10.9)sigma Subscript normal r Baseline equals minus upper P Subscript normal i Baseline left-parenthesis StartFraction upper R 1 squared Over upper R 2 squared minus upper R 1 squared EndFraction right-parenthesis left-parenthesis 1 minus StartFraction upper R 2 squared Over r squared EndFraction right-parenthesis

(10.10)sigma Subscript normal upper H Baseline equals minus upper P Subscript normal i Baseline left-parenthesis StartFraction upper R 1 squared Over upper R 2 squared minus upper R 1 squared EndFraction right-parenthesis left-parenthesis 1 plus StartFraction upper R 2 squared Over r squared EndFraction right-parenthesis

The maximum radial and hoop (circumferential) stresses occur at r = R1 when σr = Pi. The negative sign indicates tension.


(10.11)therefore sigma Subscript normal upper H Baseline equals minus upper P Subscript normal i Baseline left-parenthesis StartFraction upper R 1 squared plus upper R 2 squared Over upper R 2 squared minus upper R 1 squared EndFraction right-parenthesis

(10.12)sigma Subscript normal r Baseline equals upper P Subscript normal i Baseline left-parenthesis StartFraction upper R 1 squared Over upper R 2 squared minus upper R 1 squared EndFraction right-parenthesis

where σH is the hoop stress (MPa), σ is the radial stress (MPa), R1 is the internal radius (m), R2 is the external radius (m), r is the radius at point of interest (measured from the pipeline center), Pe is the external pressure (gauge) (MPa), and Pi is the internal pressure (gauge) (MPa).


10.4.1.2 Thin-Walled Pipeline


A pipeline with D/t > 20 is known as a thin-walled pipeline. A basic approach is known as thin wall hoop stress theory. Since the maximum hoop stress is normally the limiting factor, it is this stress that will be considered [1012].


It is predictably accurate for D/t > 20. The hoop stress is then calculated as follows:


(10.13)sigma Subscript normal upper H Baseline equals left-parenthesis upper P Subscript normal i Baseline minus upper P Subscript normal e Baseline right-parenthesis StartFraction upper D 2 Over 2 t EndFraction comma upper D slash t greater-than 20

Hoop stress developed in the pipe wall at the internal design pressure is given by


(10.14)sigma Subscript normal upper H Baseline equals left-parenthesis upper P Subscript normal i Baseline right-parenthesis StartFraction upper D 2 Over 2 t EndFraction comma upper D slash t greater-than 20 comma upper P Subscript normal e Baseline equals 0

where σH is the hoop stress (MPa), Pi is the internal design pressure (gauge) (MPa), Pe is the external pressure (gauge) (MPa), t is the design thickness (m), D1 is the inside pipe diameter (m), and D2 is the outside pipe diameter (m).


10.4.2 Longitudinal Stress


The estimation of the longitudinal stress in a section of the pipeline requires the individual stress components to be identified knowing external restraining conditions.


The axial (longitudinal) stress in a pipeline depends wholly on the limiting conditions (imposed boundary condition) experienced by the pipeline, that is, whether the pipeline is unrestrained, restrained, or partially restrained. The boundary conditions can include the effects of soil reaction loads, anchor restraints, line pipe bend resistance, and residual pipelay tension forces [1012]. The longitudinal stress in a thin cylindrical shell is calculated as half of the hoop stress.


The total longitudinal stress should be the sum of the stresses arising from the following (see Sections 10.4.2.110.4.2.7):



  • pressure;
  • bending;
  • temperature;
  • weight;
  • other sustained loads;
  • occasional loadings.

A pipeline should be considered totally restrained when axial movement and bending resulting from temperature or pressure change are totally prevented.


10.4.2.1 Fully Restrained Pipeline


A fully end constrained boundary condition can occur at an anchor block or pig trap and pipeline end manifold (PLEM) or pipeline end termination (PLET) sled. For a fully end constrained pipeline (Figure 10.1), the longitudinal strain (ε1 = 0) and deflection (Δ = 0) components are 0, and the longitudinal stress response can be determined from Equation (10.15), assuming a constant uniform temperature field [1012].


Piping in which soil or supports prevent axial displacement of flexure at bends is restrained. Restrained piping may include the following [2, 6]:



  • straight sections of buried piping;
  • bends and adjacent piping buried in stiff or consolidate soil;
  • sections of above-ground piping on rigid supports.
    A diagram shows a beam with sections labeled restrained, partially restrained, and unrestrained. In the restrained section, forces are indicated as E subscript alpha T 2 minus T 1 and V subscript sigma H. Equation appears below.

    Figure 10.1 Longitudinal tensile stresses in a fully restrained pipe section.


The net longitudinal compressive stress in a restrained pipe is calculated based on Clause 419.6.4 of ASME B31.4 as



where σLR is the restrained longitudinal stress (MPa), α is the thermal linear coefficient of expansion (mm/mm/°C), T2 is the operating temperature (maximum or minimum metal temperature) (°C), T1 is the installation temperature (°C), v is the Poisson’s ratio (v = 0.30 for steel), σH is the hoop stress (MPa), and E is the modulus of elasticity (GPa).


Piping and equipment should be supported so as to prevent or reduce excessive vibration and should be anchored sufficiently to prevent undue strains on connected equipment. Supports, hangers, and anchors should be so installed as not to interfere with the free expansion and contraction of the piping between anchors. Suitable spring hangers, sway bracing, and so on should be provided where necessary [2].


Anchor blocks are used to stop axial movement of a pipeline. Anchors are normally required when the pipeline comes above ground, prior to pig hatches, other branches, or manifolds. Connection of the pipeline to the anchor block is normally done by the addition of slip-on flanges to the pipeline, which are fillet welded in place. A large concrete block is then constructed around the pipeline to resist the expansion forces.


Restrained portions are always prevented from moving by installing anchors and guides, but in a buried line, a large portion is fully restrained by soil friction only [12].


The axial compressive force required to restrain a pipeline can be calculated as follows:



(10.17)Thick wall colon upper F equals upper A left-parenthesis italic upper E alpha left-parenthesis upper T 2 minus upper T 1 right-parenthesis plus StartFraction upper S Subscript normal h Baseline Over k squared plus 1 EndFraction minus v left-parenthesis upper S Subscript normal h Baseline minus StartFraction upper P Over 10 EndFraction right-parenthesis right-parenthesis

where F is the axial force (N), E is the modulus of elasticity (GPa), T1 is the installation temperature (°C), Sh is the hoop stress (MPa), v is the Poisson’s ratio (0.3 for steel), A is the cross-sectional area of the pipe wall (m2), α is the coefficient of thermal expansion (°C−1), T2 is the maximum or minimum metal temperature (°C), k is the ratio of the outside diameter to inside diameter, and P is the internal design pressure (gauge) (MPa).


10.4.2.2 Unrestrained Pipeline


An end-free boundary condition can occur at locations where no physical longitudinal restraint exists, for example, at a riser bend extending from the seabed to the production platform. Since the pipeline is not restrained axially, the Poisson effect and thermal expansion component do not produce stress in the line pipe. The pipeline longitudinal stress is only due to the end cap effect [25].


Piping that is free to displace axially or flex at bends is unrestrained. Unrestrained piping may include the following:



  • aboveground piping that is configured to accommodate thermal expansion or anchor movements through flexibility;
  • bends and adjacent piping buried in soft or unconsolidated soil;
  • an unbackfilled section of an otherwise buried pipeline that is sufficiently flexible to displace laterally or that contains a bend;
  • pipe subject to an end cap pressure force.

For unrestrained sections of a pipeline, the longitudinal tensile stress should be calculated as follows (refer to BS 8010, Part 2, Section 2.8, and Clause 2.9.3.2):


(10.18)sigma Subscript upper L upper U Baseline equals StartFraction sigma Subscript normal upper H Baseline Over k squared plus 1 EndFraction plus StartFraction 1000 upper M Subscript normal b Baseline times i Over upper Z EndFraction

For a thin-walled pipeline, use k = 1.


When the internal pressure is greater than the external pressure, the stress will be positive and tensile, as expected.


The unrestrained longitudinal stress can generally be expressed as



Expanding Equation (10.19), the expression for an approximate thin wall is given as


(10.20)sigma Subscript upper L upper U Baseline equals StartFraction upper D 1 squared left-parenthesis upper P Subscript normal i Baseline minus upper P Subscript normal e Baseline right-parenthesis Over 4 t left-parenthesis upper D 2 minus t right-parenthesis EndFraction proportional-to StartFraction upper D 2 left-parenthesis upper P Subscript normal i Baseline minus upper P Subscript normal e Baseline right-parenthesis Over 4 t EndFraction comma upper D slash t greater-than-or-equal-to 20

where σLU is the unrestrained longitudinal stress (MPa), Mb is the bending moment (N m), D1 is the inside diameter (m), D2 is the outside diameter (m), Z is the pipe section modulus (m3), σE is the end cap stress (MPa), i is the stress intensification factor (see Section 10.6.1), and k is the ratio D2/D1.


10.4.2.3 Partially Restrained Pipeline


At the free end of a pipeline, the total friction force acting on the pipeline increases as one progresses from the free end. This friction is passive and acts against the forces that try to displace the line [2, 4, 5].


For a constant frictional force per unit length of the line, the rate of change of longitudinal stress is constant. The expression for the longitudinal stress experienced by a pipeline that is partially restrained is given as


(10.21)sigma Subscript upper L upper P Baseline equals sigma Subscript normal upper E Baseline minus mu StartFraction upper W Subscript normal s Baseline upper X Over upper A Subscript normal s Baseline EndFraction comma 0 greater-than-or-equal-to upper X greater-than-or-equal-to upper L Subscript normal upper A Baseline

(10.22)sigma Subscript normal upper E Baseline equals upper D 1 squared StartFraction left-parenthesis upper P Subscript normal i Baseline minus upper P Subscript normal e Baseline right-parenthesis Over left-parenthesis upper D 2 squared minus upper D 1 squared right-parenthesis EndFraction

Expanding the above, the expression for an approximate thin wall is given as [5]


(10.23)sigma Subscript upper L upper P Baseline equals StartFraction upper D 2 left-parenthesis upper P Subscript normal i Baseline minus upper P Subscript normal e Baseline right-parenthesis Over 4 t EndFraction minus mu StartFraction upper W Subscript normal s Baseline upper X Over upper A Subscript normal s Baseline EndFraction comma upper D slash t greater-than-or-equal-to 20

where μ is the seabed coefficient of friction, Ws is the submerged weight of the pipeline, X is the distance from the free end of the pipeline (m), As is the area of steel (As = πDt) (m2), and Lx is the length between anchors (m).


Note


ModifyingAbove ModifyingBelow StartLayout 1st Row upper X equals 0 when sigma Subscript upper L upper P Baseline equals sigma Subscript normal upper E Baseline equals sigma Subscript upper L upper U Baseline 2nd Row upper X equals upper L Subscript x Baseline when sigma Subscript upper L upper P Baseline equals sigma Subscript upper L upper R Baseline EndLayout With bar With bar

10.4.2.4 Bending Stress


A pipe must sustain installation loads and operational loads. In addition, external loads such as those induced by waves, current, uneven seabed, trawl-board impact, pullover, and expansion due to temperature changes need to be considered. A pipe subjected to increasing bending may fail due to local buckling/collapse or fracture, but it is the local buckling/collapse limit state that usually dictates the design [13].


The pipe (beam) bending is analyzed using the engineer’s theory of bending (ETB) and simple beam theory. The theory associates the bending stress at a point to the moment imposed on the section or the curvature experienced. It is used to calculate the bending stress at any point in a pipe.


Assumptions



  • Transverse planes before bending remain transverse after bending, with no warping.
  • The pipe material is homogenous and isotropic and obeys Hooke’s law with E the same in tension or compression.
  • The pipeline (beam) is straight and has a constant or slightly tapered cross-section.
  • Load does not cause twisting or buckling. This is satisfied if the loading plane coincides with the section’s symmetry axis.
  • The pipeline is subject to pure bending. This means that the shear force is 0 and that no torsional or axial loads are present.

The equation is valid for elastic pure bending and does not take into account the geometrical deformations that occur, particularly in hollow cylinders (ovality) under bending. The equation is expressed as


(10.24)StartFraction sigma Subscript x Baseline Over y EndFraction equals StartFraction upper M Over upper I EndFraction equals StartFraction upper E Over upper R EndFraction

This is called the engineer’s theory of bending. The standard form of this equation is




where σx is the bending stress (0 ≤ σxσy) (MPa), M is the bending moment (N m), y is the distance from the neutral axis (0 ≤ y ≤ OD/2) (m), I is the moment of inertia (kg m3), E is the Young’s modulus of elasticity (GPa), and R is the radius of curvature (m).


Therefore, knowing the applied bending moment and the location of the centroid, the second moment of area and the stresses along the depth of the pipe section can be calculated.


In the simple theory of bending of beams, it is assumed that no appreciable distortion of the cross-section takes place so that there is no displacement of the material either toward or away from the neutral axis. For a thin-walled pipe subjected to bending, movement of the fibers toward the neutral axis does occur [14].


The maximum bending moment and hence the bending stress at any point in the cross-section can be determined by assuming a beam configuration.


Beam Configurations

A fixed–fixed beam under a uniformly distributed load gives a maximum bending moment (Mff) at the fixed ends of the beam.


(10.27)upper M Subscript f f Baseline equals StartFraction italic upper W upper L squared Over 12 EndFraction

A beam with pinned supports under a uniformly distributed load gives a maximum bending moment (Mpp) at midspan.


(10.28)upper M Subscript p p Baseline equals StartFraction italic upper W upper L squared Over 8 EndFraction

It is commonly accepted that a real beam configuration for an unrestrained pipeline resting on the seabed is somewhere between the fixed–fixed and the pinned–pinned cases and is given a value as follows:


(10.29)upper M Subscript f p Baseline equals StartFraction italic upper W upper L squared Over 10 EndFraction

where Mff is the fixed–fixed bending moment, Mpp is the pinned–pinned bending moment, W is the uniformly distributed load (per meter), L is the length of pipeline span, and Mfp is the bending moment halfway between the fixed–fixed and the pinned–pinned cases.


The stages of longitudinal stresses connected with bending are as follows:



  • Elastic bending up to yield at the outer fiber. At this point, the bending moment is equal to the first yield moment.
  • Elastoplastic bending up to the development of a full plastic hinge. When the plastic hinge is developed, it is known as the full plastic moment.

On the other hand, this can be solved by drawing the bending moment diagram, that is, if the value of the bending moment, M, is determined at various points of the beam and plotted against the distance x measured from one end of the beam. It is further facilitated if a shear diagram is drawn at the same time by plotting the shear, V, against x.


This approach facilitates the determination of the largest absolute value of the bending moment in the beam.


10.4.2.5 Stress due to Sustained Loads


Sustained loads are the sum of dead weight loads, axial loads caused by internal pressure, and other applied axial loads that are not caused from temperature and accelerations [3].


In accordance with ASME Boiler and Pressure Vessel Code, Section III, Subsections NC and ND, the calculated stresses due to pressure, weight, and other sustained mechanical loads must meet the allowable 1.5Sh, that is,


(10.30)StartFraction upper B 1 italic upper P upper D 0 Over 2 t EndFraction plus StartFraction upper B 2 upper M Subscript normal upper A Baseline Over upper Z EndFraction less-than-or-equal-to 1.5 upper S Subscript normal h

where T is the temperature derating factor, P is the internal design pressure (MPa), D0 is the outside diameter of the pipe (m), Z is the section modulus of the pipe (m3), MA is the resultant moment loading on the cross-section due to weight and other sustained loads (N m), Sh = 0.33SuT, at the maximum installed or operating temperature (MPa), and Su is the specified minimum ultimate tensile strength (N/m2).


10.4.2.6 Stress due to Occasional Loads


Occasional loads are loads such as wind, earthquake, breaking waves or green sea impact loads, and dynamic loads such as pressure relief, fluid hammer, or surge loads. In accordance with ASME Boiler and Pressure Vessel Code, Section III, Subsections NC and ND, the calculated stress due to pressure, weight, other sustained loads, and occasional loads must meet the allowable stress as follows [3]:


(10.31)StartFraction upper B 1 upper P Subscript max Baseline upper D 0 Over 2 t EndFraction plus StartFraction upper B 2 left-parenthesis upper M Subscript normal upper A Baseline plus upper M Subscript normal upper B Baseline right-parenthesis Over upper Z EndFraction less-than-or-equal-to italic k upper S Subscript normal h

where Pmax is the peak pressure (MPa) and MB is the resultant moment loading on the cross-section due to occasional loads, such as thrusts from relief and safety valves, loads from pressure and flow transients, and earthquake, if required. For earthquake, use only one-half of the range. Effects of anchor displacement due to the earthquake may be excluded if they are included under thermal expansion. kSh = i.85Sh for upset condition but not greater than 1.5Sy, 2.25Sh for the emergency condition but not greater than i.85Sy, and 3.0Sh for the faulted condition but not greater than 2.0Sy. Sh = 0.33SuT, at the maximum installed or operating temperature (MPa), Su is the specified minimum ultimate tensile strength (N/m2), and Sy is the material yield strength at a temperature consistent with loading under consideration.


10.4.2.7 Stress due to Thermal Expansion


Thermal expansion may be detrimental for the pipe itself, flanges and bolts, branch connections, pipe supports, and connected equipment such as pumps and compressors. Sufficient pipe flexibility is necessary to prevent such detrimental loads [2].


Stresses due to expansion for those portions of the piping without substantial axial restraint shall be combined in accordance with the following equation (refer to Clause 833.8 of ASME B31.8):



(10.33)upper M Subscript normal upper E Baseline equals left-bracket left-parenthesis i Subscript normal i Baseline upper M Subscript normal i Baseline right-parenthesis squared plus left-parenthesis i Subscript normal o Baseline upper M Subscript normal o Baseline right-parenthesis squared plus upper M Subscript normal t Superscript 2 Baseline right-bracket Superscript 1 slash 2 Baseline left-parenthesis upper N m right-parenthesis

The cyclic stress range SESA, where


(10.34)upper S Subscript normal upper A Baseline equals f left-bracket left-parenthesis 1.25 upper S Subscript normal c Baseline plus 0.25 upper S Subscript normal h Baseline right-parenthesis minus upper S Subscript normal upper L Baseline right-bracket

If Equation (10.32) is not met, the piping may be qualified by meeting the following equation:


(10.35)StartLayout 1st Row StartFraction italic upper P upper D 0 Over 4 t EndFraction plus StartFraction 0.75 italic i upper M Subscript normal upper A Baseline Over upper Z EndFraction plus StartFraction italic i upper M Subscript normal upper E Baseline Over upper Z EndFraction less-than-or-equal-to upper S Subscript normal h Baseline plus upper S Subscript normal upper A Baseline 2nd Row left-parenthesis 0.75 i should not b e less than 1.0 right-parenthesis EndLayout

where SE is the stress due to expansion (MPa), SA is the allowable stress range for expansion stress, f is the stress range reduction factor, ME is the range of resultant moment due to thermal expansion (N m), also includes moment effects of anchor displacements due to the earthquake if anchor displacement effects were omitted from occasional loadings, Sc = 0.33SuT, at the minimum installed or operating temperature (MPa), Sh = 0.33SuT, at the maximum installed or operating temperature (MPa), Su is the specified minimum ultimate tensile strength (N/m2), SL is the unrestrained longitudinal stress (N/m2), T is the temperature derating factor, i is the stress intensification factor (see Section 6 or Appendix E of ASME B31.8), Mi is the in-plane bending moment (N m), Mt is the torsional moment (N m), Mo is the out-of-plane bending moment (N m), io is the out-of-plane stress intensification factor (refer to Appendix E of ASME B31.8), and ii is the in-plane stress intensification factor (refer to Appendix E of ASME B31.8).


10.4.3 Shear Stress


Shear stress in a pipeline should be minimized. The shear stress should be calculated from the torque and shear force applied to the pipeline using the following equation:


(10.36)tau equals StartFraction 1000 upper T Over 2 upper Z EndFraction plus StartFraction 2 upper F Subscript normal s Baseline Over upper A EndFraction

where τ is the shear stress (N/m), T is the torque applied to the pipeline (N m), Fs is the shear force applied to the pipeline (N), A is the cross-sectional area of the pipe (m2), and Z is the section modulus of the pipe (m3).


Sections 10.4.3.1 and 10.4.3.2 show how to determine shear stress due to torsion and spanning, which makes up the total shear stress.


10.4.3.1 Shear Stress due to Torsion


Torsion is the twisting of a straight bar when it is loaded by twisting moments or torques that tend to produce rotation about the longitudinal axes of the bar. When subjected to torsion, every cross-section of a circular shaft remains plane and undistorted, and the bar is said to be under pure torsion [15].


The shear stress on a uniform cylindrical shaft that is under a uniform torsion is given by


(10.37)tau equals StartFraction italic trace Subscript normal x Baseline Over upper J EndFraction equals StartFraction italic trace Subscript normal x Baseline Over 2 upper I Subscript normal x Baseline EndFraction equals StartFraction upper T Over 2 upper Z EndFraction

For a thin cylindrical shaft (or thin-walled tube) with t < R/10,


(10.38)upper J equals upper R squared left-parenthesis 2 italic pi r t right-parenthesis equals 2 pi upper R cubed t equals StartFraction pi Over 4 EndFraction upper D cubed t

(10.39)tau Subscript max Baseline equals StartFraction italic upper T upper D Over 2 left-parenthesis left-parenthesis pi slash 4 right-parenthesis upper D cubed t right-parenthesis EndFraction equals StartFraction 2 upper T Over pi upper D squared t EndFraction

The maximum shear stress due to torsion can be calculated as follows:


(10.40)tau Subscript max Baseline equals StartFraction italic upper T upper D Subscript normal o Baseline Over 2 upper J EndFraction equals StartFraction italic upper T upper D Subscript normal o Baseline Over 4 upper I Subscript x Baseline EndFraction equals StartFraction 16 italic upper T upper D Subscript normal o Baseline Over pi left-parenthesis upper D Subscript normal o Superscript 4 Baseline minus upper D 1 Superscript 4 Baseline right-parenthesis EndFraction

where τ is the shear stress (N/m); T is the torque or twisting moment (N m); R is the radial distance from the longitudinal axis (m); Ix and Iy are the moments of inertia about the x– and 33 y-axis, respectively (kg m3); Z is the section modulus (m3); and J is the polar moment of inertia.


10.4.3.2 Shear Stress due to Spanning


The shear stress due to spanning is composed of vertical shear and a longitudinal shear due to the bending. The maximum shear force acting on a simple span is equal to the maximum support reaction. This is in turn equal to the change in shear force at the reaction. The maximum vertical shear stress is defined as the force per unit area. The maximum vertical shear stress is calculated as follows [15]:



where Fs is the shear force applied to the pipeline (N), As is the cross-sectional area of the pipe (m2), and Rmax is the maximum vertical reaction on the pipe (N).


10.4.4 Equivalent Stress


Pressure and temperature as well as other operating conditions such as bending can create expansion and flexibility problems, and therefore stress criteria are specified in all codes, limiting the level of combined stresses allowed in a pipeline.


The design factor relates only to hoop stress; if other stresses are significant, then these could contribute to the pipeline steel, exceeding its yield stress. Design codes vary in the way they calculate the combined or equivalent stress, but the following equation is typical [29].


The equivalent stress corresponds to the total stress in the pipeline resulting from a combination of all the stresses.


The equivalent stress can be calculated using the following equation (refer to DNV 2012, Clause 103):


(10.42)sigma Subscript normal e Baseline equals StartRoot sigma Subscript normal upper H Superscript 2 Baseline plus sigma Subscript normal upper L Superscript 2 Baseline minus sigma Subscript normal upper H Baseline sigma Subscript normal upper L Baseline plus 3 tau squared EndRoot

where σe is the equivalent stress (N/m2), σH is the hoop stress (N/m2), σL is the longitudinal stress (N/m2), and τ is the shear stress (N/m2).


In accordance with Clause 833.4 of ASME B31.4, the maximum allowable equivalent stress is 90% of the SMYS.


10.4.5 Limits of Calculated Stress


Pipe structures should be designed by considering the limit states at which they would be unfit for their intended use by applying appropriate factors.


With reference to BS 8010, Part 3, Clause 4.2.5.4, stress in the pipeline system should satisfy the following inequality:



The allowable stress depends on the pipe material used, the location of the pipe, the operating conditions, and other limitations imposed by the designer in conformance with the code used. The allowable stresses for various grades and types of material are tabulated in Table 402.3.1(a) of ANSI B31.4, 1992 edition.


In accordance with ASME B31.4, for an unrestrained pipeline, the allowable effective stress is 0.72Sy, and for a restrained line, the allowable stress is 0.9Sy.


With reference to Clause 833.3 of ASME B31.8, for a restrained pipe, the allowable longitudinal stress is 0.9SyT, where Sy is the specified minimum yield strength (MPa) and T is the temperature derating factor.


Based on Clause 833.6 of ASME B31.8, for an unrestrained pipe, the allowable longitudinal stress is σAL ≤ 0.75SyT, where Sy is the specified minimum yield strength (MPa) and T is the temperature derating factor [2].


10.4.5.1 Allowable Hoop Stress


The allowable hoop stress may be calculated using the following equation (refer to Clause 805.234 of ASME B31.8 and Clause 201.4.1 of API RP 1111):


(10.44)sigma Subscript a upper H Baseline equals f times e times upper T times upper S Subscript normal y

where σaH is the allowable hoop stress, Sy is the specified minimum yield strength, f is the design factor, e is the weld joint factor, and T is the temperature derating factor.


The allowable hoop stress for the cold-worked pipe is 75% of the above value (Clause 201.4.4 of API RP 1111). The design factor is 0.72 for pipelines and liquid risers, 0.60 for gas risers, and 0.50 for gas platform piping [2].


10.4.5.2 Allowable Equivalent Stress



In accordance with Clause 833.4 of ASME B31.4, the maximum allowable equivalent stress is 90% of the SMYS.


(10.46)sigma Subscript a e Baseline equals 0.9 upper S Subscript normal y

where σae is the allowable equivalent stress and Sy is the specified minimum yield strength.


10.4.5.3 Limits of Calculated Stress due to Sustained Loads


The sum of the longitudinal stresses due to pressure, weight, and other sustained external loads shall not exceed 0.72SA, where SA = 0.75Sy (Sy is the specified minimum yield strength) [6].


10.4.5.4 Limits of Calculated Stress due to Occasional Loads


The sum of the longitudinal stresses produced by pressure, live and dead loads, and other sustained loadings and of the stresses produced by occasional loads, such as wind or earthquake, may be as much as 1.33Sh [6].


10.4.5.5 Limits of Calculated Stress due to Expansion Loads


The computed displacement stress range (expansion stress range) SE in a pipeline should not exceed the allowable displacement stress range SA [6].


(10.47) upper S Subscript normal upper E Baseline less-than upper S Subscript normal upper A Baseline equals f left-parenthesis 1.25 upper S Subscript normal c Baseline plus 0.25 upper S Subscript normal h Baseline right-parenthesis

When Sh is greater than SL, the difference between them may be added to the term 0.25Sh; in that case, the allowable stress range is calculated as


(10.48)upper S Subscript normal upper A Baseline equals f left-bracket left-parenthesis 1.25 left-parenthesis upper S Subscript normal c Baseline plus upper S Subscript normal h Baseline right-parenthesis minus upper S Subscript normal upper L Baseline right-bracket

where SE is the expansion stress range = left-parenthesis upper S Subscript normal h Superscript 2 Baseline plus 4 upper S Subscript normal t Superscript 2 Baseline right-parenthesis Superscript 1 slash 2 (MPa), ME is the resultant bending stress =[(iiMi)2 + (ioMo)2]1/2/Z (MPa), St is the torsional stress = Mt/2Z (MPa), Mi is the in-plane bending moment (N m), Mo is the out-of-plane bending moment (N m), Mt is the torsional moment (N m), ii is the in-plane stress intensification factor, io is the out-of-plane stress intensification factor, Z is the section modulus of the pipe (m3), Sc = 0.33SuT, at the minimum installed or operating temperature (MPa), Sh = 0.33SuT, at the maximum installed or operating temperature (MPa), Su is the specified minimum ultimate tensile strength (MPa), and f is the stress range reduction factor (obtained from Table 302.3.5 of ASME B31.3) or calculated as follows:


(10.49)f equals 6.0 left-parenthesis upper N right-parenthesis Superscript negative 0.2 Baseline less-than-or-equal-to 1.0

where N is the equivalent number of full displacement cycles during the expected service life of the piping system.


10.5 Fatigue


Fatigue is a structural damage that occurs when a pipe material is subjected to cycles of stress or strain [16]. Such stresses are normally concentrated locally by structural discontinuities, geometric notches, surface irregularities, damage defects, and so on.


ASME B31.4 shows how to design the pipeline against fatigue failure.


Pipelines can vary in pressure over hourly, daily, or yearly cycles [16]. Pressure cycling can cause small weld defects to grow in time to a critical size and can be a major factor in determining the fatigue life of welded steel gas pipelines, particularly those pipelines designed for use as line-pack storage. (Gas can be stored temporarily in the pipeline system through a process called line packing.). Fatigue is more critical in oil pipelines than gas pipelines because limited compressibility enhances pressure cycling.


10.5.1 Fatigue Life


ASTM defines fatigue life as the number of stress cycles of a specified character that a specimen sustains before failure of a specified nature occurs. Fatigue life may be affected by cyclic stress state, geometry, surface quality, material type, residual stresses, size and distribution of internal defects, air or vacuum, direction of loading, grain size, environment, temperature, and crack closure [17].


To avoid fatigue failure of a pipeline, the following should be adhered to [8]:



  • One stress cycle per day of 125 N/mm2 could lead to failure after 40 years (15,000 cycles). This is very important for gas pipelines operating under line-pack conditions.
  • Determine the number of stress cycles and the stress range expected during the design life of a pipeline. If the value of stress cycles comes within the reach of 15,000, then the pipe material should be changed or the frequency of cycling or the stress range of cycling should be reduced over the pipe life.
  • Revalidate the pipeline by hydrostatic testing, if the number of stress cycles reaches 15,000.
  • A crack detection tool can be used to determine the condition of the pipeline.
  • Keep the log of pressures and cycles throughout the design life of the pipeline.

10.5.2 Fatigue Limit


The significance of the fatigue limit is that if the material is loaded below this stress, then it will not fail, regardless of the number of times it is loaded [16]. In accordance with IGE/TD/1, 15,000 cycles at 125 N/mm2 has been set as the maximum permissible fatigue life.

A graph shows stress S versus several cycles N on a log scale. Two curves depict material behavior: one labeled In air trending downwards and leveling off, indicating a fatigue limit; the other labeled in corrosive environment drops more steeply without leveling off.

Figure 10.2 Typical generic S–N curves for steel.


10.5.3 S–N Curve


A very useful way to visualize time to failure for a specific material is with the S–N curve. This is a graph of the magnitude of a cyclic stress (S) against the logarithmic scale of cycles to failure (N).


A fatigue life test should be conducted for the pipe material and an S–N curve drawn [16]. For instance, a specimen of the pipe material is placed in a fatigue testing machine and loaded repeatedly to a certain stress, σ1. The loading cycles are continued until failure occurs and the number, n, of loading cycles to failure is noted. Then, the test is repeated for a different stress, say σ2; if σ2 is larger than σ1, the number of cycles to failure will be less. If it is smaller, the number of cycles to failure will be more [16].


Eventually, enough data are accumulated to plot an S–N curve. Such curves have the general shape shown in Figure 10.2, where the vertical axis is usually a linear scale and the horizontal axis is a log scale.


From Figure 10.2, the smaller the stress, the larger is the number of cycles to produce failure. Fatigue strength curves (S–N curves) for a particular material or structural weldment formation can be found in design standards such as ASME Boiler and Pressure Vessel Code, Section VIII, Division 2, Appendix 5, and BS 7910.


10.6 Expansion and Flexibility


Avoid expansion bends and design the entire pipeline to take care of its own expansion. Maximum flexibility is obtained by placing supports and anchors so that they will not interfere with the natural movement of the pipe. Allow for stress intensification factors in components [10].


Expansion joints may be used to avoid pipeline bending (flexure) stress due to the movement of supports or the tendency of the pipe to expand under temperature change. The following should be considered for the use of expansion joints [2]:



  • Select the expansion joint carefully for the maximum temperature range (and deflection) expected so as to prevent damage to expansion fitting.
  • Provide guides to limit movement at the expansion joint to direction permitted by the joint.
  • Provide adequate anchors at one end of each straight section or along their middle length, forcing movement to occur at the expansion joint, yet providing adequate support for the pipeline.
  • Mount expansion joints adjacent to an anchor point to prevent sagging of the pipeline under its own weight and do not depend upon the expansion joint for stiffness (it is intended to be flexible).
  • Give consideration to effects of corrosion since the corrugated character of expansion joints makes cleaning difficult.

Formal flexibility analysis for an unrestrained piping system may not be required.


10.6.1 Flexibility and Stress Intensification Factors


The stress intensification factor (SIF) is defined as the ratio of the maximum stress state (stress intensity) to the nominal stress, calculated by the ordinary formulas of mechanics [2].


In piping design, this factor is applied to welds, fittings, branch connections, and other piping components where stress concentrations and possible fatigue failure might occur. Usually, experimental methods are used to determine these factors.


The structural analysis of risers, expansion loops, or tee assemblies entails the use of flexibility stress intensification factors applicable to accurately model the structural behavior of bends and tees within the system.


It is recognized that some of the SIFs for the same components are different for different codes. In some cases, different editions of the same code provide different SIFs for a given component. The way that the SIFs are applied to moment loadings is also different for different codes. The B31.1 and ASME Section III codes require that the same SIF be applied to all the three-directional moments, while the B31.3, B31.4, B31.5, and B31.8 codes require that different SIFs be applied to the in-plane and out-of-plane moments, with no SIF required for torsion.


Therefore, the stress analyst has to ensure that the appropriate SIFs from the applicable code are used [2].


Flexibility is added in a pipe system by changes in the run direction (offsets, bends, and loops) or by use of expansion joints or flexible couplings of the slip joint, ball joint, or bellow type. In addition, more or less, flexibility can be added by changing the spacing of pipe supports and their function (e.g., removal of a guide close to a bend to add flexibility).


Another way to increase the flexibility is to change the existing piping material to a material with a higher yield or tensile strength or to a material quality that does not need additional corrosion and erosion allowance and thereby obtains a reduction in the wall thickness that again gives more flexibility since the moment of inertia is reduced with a reduction of the pipe wall thickness.


The flexibility factors and stress intensification factors that can be used are listed in standards, for example, ASME B31.8, Table E1, and CSA Z662:23, Table 4.8.


10.7 Corrosion Allowance


Additional wall thickness is added to account for corrosion when water is present in a fluid along with contaminants such as oxygen, hydrogen sulfide (H2S), and carbon dioxide (CO2). This is one of the several methods available to mitigate the effects of corrosion and often the least recommended [18].


Corrosion allowance is made to account for corrosion loss during service, damage during fabrication, transportation, and storage. A value of 1/16 in. may be appropriate. A thorough assessment of the internal corrosion mechanism and rate is necessary before any corrosion allowance is taken. Refer to BS 8010, ASME codes, ISO, API 5L, and other governing codes for the use of corrosion allowance in the design of pipelines [19].


10.7.1 Internal Corrosion Allowance


There is no need for internal corrosion allowance if the substance being transmitted is noncorrosive, for example, dry natural gas [8]. An internal corrosion allowance should be added to the wall thickness calculation when corrosive substances such as wet gas, hydrocarbon liquid, or two-phase flow are being transported through the pipeline [19].


10.7.2 External Corrosion Allowance


A wall thickness allowance for corrosion is not required if the pipe and components are protected against corrosion in accordance with the requirements and procedures prescribed in ASME B31.4, ASME B31.8, and other governing codes (i.e., coated and cathodically protected) [29].


10.7.3 Formulas


This section shows how to add corrosion allowance to the calculation of the nominal wall thickness of a pipeline, using any of the engineering codes listed below [29].



  1. Using BS 8010:
    (10.50)t equals StartFraction upper P times upper D Over 20 upper S Subscript normal y Baseline times upper E times upper F EndFraction

    (10.51)t Subscript n o m Baseline equals left-parenthesis t plus t Subscript corr Baseline right-parenthesis times manufacturing tolerance

  2. Using ASME B31.8:
    (10.52)t equals StartFraction upper P times upper D Over 20 upper S Subscript normal y Baseline times upper E times upper F EndFraction plus t Subscript corr

  3. Using ISO 13623:
    (10.53)t equals StartFraction upper P times upper D Over left-parenthesis 20 upper S Subscript normal y Baseline times upper E times upper F right-parenthesis plus upper P EndFraction

    (10.54)t Subscript n o m Baseline equals left-parenthesis t plus t Subscript corr Baseline right-parenthesis times manufacturing tolerance

where tcorr is the corrosion allowance (m), tnom is the nominal wall thickness (m), P is the design pressure (MPa), Sy is the specified minimum yield strength (MPa), t is the design wall thickness (m), D is the outside diameter of the pipe (m), F is the design factor, and E is the joint factor.


10.8 Pipeline Stiffness


With reference to DNV 2012, Clause 205, the possible strengthening effect of weight coating on a steel pipe is not normally taken into account in the design against yielding. Coating that adds significant stiffness to the pipe may increase the stress in the pipe at discontinuities in the coating. When appropriate, this effect should be taken into account.


For the buried pipe, resistance to external loading is a function of pipe stiffness and passive soil resistance under and adjacent to the pipe.


The overall stiffness of a long section of the pipeline should be calculated from the value for the moment of inertia and should be used to calculate the overall deflections and induced bending moments for the concrete-coated pipe [20].


10.8.1 Calculation of Pipeline Stiffness


Under load, the individual components of the pipe wall (steel, mortar lining, and, when applicable, mortar coating) act together as laminated rings. The combined action of these elements increases the overall moment of inertia of the pipe, over that of the steel pipe alone [20].


The pipe wall stiffness is the sum of the stiffness of the bare pipe, lining, and coating.


The pipe wall stiffness (EI) is the sum of the stiffness of the bare pipe, lining, and coating.


(10.55)left-parenthesis upper E upper I right-parenthesis equals upper E Subscript normal s Baseline upper I Subscript normal s Baseline plus upper E Subscript normal upper L Baseline upper I Subscript normal upper L Baseline plus upper E Subscript normal c Baseline upper I Subscript normal c

(10.56)upper I equals StartFraction t cubed Over 12 EndFraction

where t is the wall thickness of the pipe, lining, or coating; (EI) is the pipe wall stiffness per inch of pipe length (in./lb); ELIL is the stiffness of the lining; EcIc is the stiffness of the coating (e.g., concrete); EsIs is the stiffness of the steel pipe wall; E is the modulus of elasticity (207 GPa for steel and 27.6 GPa for cement mortar); and I is the transverse moment of inertia per unit length of the pipe wall (in.3 or mm3).


The stiffness of each of the laminar rings (i.e., steel pipe, cement mortar lining, and cement mortar coating) is calculated using the modulus of elasticity of the component in GPa and the moment of inertia as a per unit length value, defined as t3/12.


10.8.1.1 Deflection


Pipe stiffness and passive soil resistance of backfill play a significant role in predicting deflection. M.G. Spangler [21] of Iowa State University published the Iowa formula in 1941.


(10.57)Pipe deflection equals StartFraction load o n pipe Over pipe stiffness plus soil stiffness EndFraction

Deflection of a pipeline is calculated using the modified Iowa deflection formula as follows:


(10.58)normal upper Delta x equals StartFraction upper D 1 italic upper K upper W upper R cubed Over left-parenthesis upper E upper I right-parenthesis plus 0.061 upper E prime upper R cubed EndFraction

where Δx is the horizontal deflection of the pipe (m), D1 is the deflection lag factor (1.0–1.5), K is the bedding constant (0.1), R is the pipe radius (m), (EI) is the pipe wall stiffness per meter of pipe length (in./lb, m/kg), E ′ is the modulus of soil reaction (GPa), and W is the external load per unit length of the pipe (dead load (earth load) + live load).


Some data on loading are presented in Table 10.4.


Spangler hypothesized that if the lateral movement of various points on the pipe ring was known, the distribution of lateral pressures could be determined by multiplying the movement of any point by the modulus of passive resistance, E′. For mathematical convenience, this lateral pressure was assumed to be a simple parabolic curve embracing only the middle 100° arc of the pipe (see Figure 10.3).


Table 10.4 Standard HS-20 Highway and E-80 Railroad Loading


Source: [20]/AMERICAN (American Cast Iron Pipe Company).



















































Highway HS-20 Loading Railroad E-80 Loading
Height of Cover (ft) Load (psi) Height of Cover (ft) Load (psi)
1 12.5 2 26.4
2 5.6 5 16.7
3 4.2 8 10.1
4 2.8 10 7.6
5 1.7 12 5.6
6 1.4 15 4.2
7 1.2 20 2.1
8 0.7 30 0.7

He also assumed that the total vertical load was uniformly distributed across the width of the pipe, and the bottom vertical load was distributed uniformly over the width of the pipe bedding [20].


The following terms can be introduced to describe the three separate factors that affect the pipe deflection:



  • load factor (D1KW);
  • ring stiffness factor (EI/R3);
  • soil stiffness factor (0.061 E ′). Values of E ′ are listed in Table 10.5.

The modified Iowa formula can be represented as a load factor


(10.59)normal upper Delta upper X equals StartFraction load factor Over ring stiffness factor plus soil stiffness factor EndFraction

Information on the load factor, ring stiffness factor, soil stiffness factor, bedding constant, and deflection lag factor can be obtained from standards of American Concrete Pipe Association [2022].


The steel pipe is designed as a flexible conduit; considerable deflection can occur without damaging the pipeline. Deflection limitations are a function of the rigidity of the specific lining and coating being used.

A diagram of an ellipse showing internal angles and forces. The ellipse has a central angle of 100 and two 40-degree angles. External arrows represent pressures: P a, P b, and P h. Labels indicate force equations with variables, as well as a subgrade.

Figure 10.3 Spangler assumptions for pressure distribution [21].


(Published with permission of American Concrete Pipe Association.)


Table 10.5 E Modulus of Soil Reaction


Source: [20]/AMERICAN (American Cast Iron Pipe Company).






























































































Type of Soil Depth of Cover (ft) E Modulus of Soil Reaction (psi)
85%a 90%a 95%a 100%a
Fine-grained soils with less than 25% sand content (CL, ML, and CL—ML) 0–5 500 700 1000 1500
5–10 600 1000 1400 2000
10–15 700 1200 1600 2300
15–20 800 1300 1800 2600
Coarse-grained soil with fines (SM and SC) 0–5 600 1000 1200 1900
5–10 900 1400 1800 2700
10–15 1100 1700 2300 3300
15–20 1300 2000 2700 3800
Coarse-grained soil with little or no fines (SP, SW, GP, and GW) 0–5 700 1000 1600 2500
5–10 1000 1500 2200 3300
10–15 1050 1600 2400 3600
15–20 1100 1700 2500 3800
Crushed stone N/A 3000 3000 3000 3000

a Standard AASHTO relative compaction.


The calculated deflection is limited to 5%, although larger deflections may not affect pipe performance. Limits for pipe deflection for various forms of coating or lining should be obtained from governing standards and codes [20].


10.8.2 Calculation of the Induced Bending Moment


Coating that adds significant stiffness to the pipe may increase the stress in the pipe at discontinuities in the coating. When appropriate, this effect should be taken into account.


Maximum stresses will develop along the pipeline where there is no concrete coating (i.e., at the field joint) and may be calculated for a known curvature from Equation (10.60).



where σB is the bending stress, C is the outside radius of the steel pipe, R is the imposed bending radius, Io is the overall moment of inertia, Is the moment of inertia of bare pipe, and E is the modulus of elasticity.


10.9 Pipeline Ovality


Ovality is generally associated with bending of pipes and affects the integrity of pipe bends. The control of ovality in pipe bending is discussed in this section [13].


The ovality of a pipe section depends on the dimensional tolerances imposed during the pipe bending while installing the pipe to fit the terrain [14].


Based on DNV 2012, Clause 202, ovality of a pipeline is defined in Equation (10.61). This will affect the structural capacity of the pipeline and shall be taken as the maximum ovality prior to loading. The advantage of ovality less than 0.5% is not allowed. Ovality in excess of 3% shall be assessed in line with DNV 2012, Clause D900. Ovalization caused during the construction stage should be included in the total ovality to be used for design [7]. Pipe ovality is calculated as



The out of roundness is calculated as


(10.62)upper O equals upper D Subscript max Baseline minus upper D Subscript min

where Dmax is the maximum measured inside or outside diameter, Dmin is the minimum measured inside or outside diameter, D is the nominal outside diameter, f0 is the ovality, and O is the out of roundness.


With reference to DNV 2012, Clause D1100, a maximum allowable ovalization of 3% applies for the pipeline as the installed condition. Under DNV 2012, Clause D400, a minimum ovalization of 0.5% is to be accounted for in the system collapse check and the combined loading [7].


In accordance with DNV 2012, Clause D1100, the ovality of a pipeline exposed to bending strain may be calculated using Equation (10.63).



where f prime 0 is the total ovalization due to unidirectional bending and external pressure, εc is the characteristic bending strain resistance, f0 is the initial ovalization, Pe is the external pressure (MPa), Pc is the characteristic collapse pressure (MPa), and t is the nominal wall thickness of the pipe (not corroded) (m).


10.9.1 Brazier Effect


The ovalization mechanism results in loss of stiffness in the form of limit point instability, referred to as “ovalization instability” or Brazier effect. Brazier effects describe the influence of ovalization on the buckling of thin shells. Ovalization can be enhanced by applying bending moments and internal or external pressure [23].


Brazier projected that the ovality was related to longitudinal bending strain by a correlation as follows [5, 24]:


(10.64)upper F Subscript normal b Baseline proportional-to left-parenthesis StartFraction epsilon Subscript normal b Baseline upper D Over t EndFraction right-parenthesis squared

Based on the work of Calladine, the correlation to be used in resolving the flattening induced by bending is calculated as [5, 23, 24]


(10.65)upper F Subscript normal b Baseline equals three eighths left-parenthesis StartFraction epsilon Subscript normal b Baseline upper D Subscript normal o Baseline Over t EndFraction right-parenthesis squared

where ε is the percentage bending strain, D is the nominal diameter (m), and t is the nominal wall thickness (m).


This correlation may be used for the diameter-to-wall thickness ratio less than 35.


10.9.2 Ovality of a Buried Pipeline


A buried pipe will ovalize under the effects of earth and live loads. The modified Iowa deflection formula may be used to calculate the pipe ovality under earth and live loads [20, 22].


(10.66)StartFraction normal upper Delta y Over upper O upper D EndFraction equals StartFraction upper D Subscript normal l Baseline italic upper K upper P Over left-parenthesis italic upper E upper I right-parenthesis Subscript e q Baseline slash upper R cubed plus 0.061 upper E prime EndFraction

where OD is the pipe outside diameter (m), Δy is the vertical deflection of the pipe (m), D is the deflection lag factor (1.0–1.5), K is the bending constant (0.1), P is the pressure on the pipe due to soil load Pv plus live load Pp (MPa), R is the pipe radius (m), (EI)eq is the equivalent pipe wall stiffness per inch of pipe length (in./lb, m/kg), and E is the modulus of soil reaction (MPa).


The pipe wall stiffness, (EI)eq, is the sum of the stiffness of the bare pipe, lining, and coating [22].


(10.67)left-parenthesis upper E upper I right-parenthesis Subscript e q Baseline equals upper E upper I plus upper E Subscript normal upper L Baseline upper I Subscript normal upper L Baseline plus upper E Subscript normal c Baseline upper I Subscript normal c

(10.68)upper I equals StartFraction t cubed Over 12 EndFraction

where ELIL is the stiffness of the lining, EcIc is the stiffness of the coating, and t is the wall thickness of the pipe, lining, or coating.


10.10 Minimum Pipe Bend Radius


In engineering, the minimum bend radius of a material is a measure of how tightly a piece can be bent before it breaks. Materials that can withstand a high degree of curvature are preferred in construction because they are more versatile, lending themselves to a wider array of industrial applications [11].


Many factors affect the pipe minimum bend radius, and knowing these factors is crucial to building sound structures. In pipe fitting and sheet metal construction, the minimum bending radii of materials depend on their thickness, composition, and skill of the fabricator [25].


10.10.1 Minimum Pipe Bend Radius Calculation Based on Concrete


(10.69)upper R equals StartFraction upper E times upper C Over sigma Subscript normal upper B Baseline EndFraction

where R is the bending radius (m), C is the pipe radius + enamel thickness + concrete thickness (m), σB is the bending stress (MPa), and E is the modulus of elasticity for concrete (103 MPa).


10.10.2 Minimum Pipe Bend Radius Calculation Based on Steel


(10.70)sigma Subscript normal upper L Baseline plus sigma Subscript normal upper B Baseline less-than-or-equal-to left-parenthesis upper F right-parenthesis left-parenthesis upper S Subscript normal y Baseline right-parenthesis

(10.71)sigma Subscript normal upper B Baseline equals left-parenthesis upper F right-parenthesis left-parenthesis upper S Subscript normal y Baseline right-parenthesis minus sigma Subscript normal upper L


StartLayout 1st Row sigma Subscript normal upper L Baseline equals StartFraction upper P upper D Over 4 t EndFraction 2nd Row upper R equals StartFraction upper E times upper C Over sigma Subscript normal upper B Baseline EndFraction EndLayout

(10.72)therefore upper R equals StartFraction upper E times upper C Over left-parenthesis left-parenthesis italic upper F upper S Subscript normal y Baseline right-parenthesis minus left-parenthesis italic upper P upper D slash 4 t right-parenthesis right-parenthesis EndFraction

where σL is the longitudinal stress (MPa), σB is the bending stress (MPa), Sy is the pipe specified minimum yield strength (MPa), P is the design pressure (MPa), D is the pipe outer diameter (m), t is the pipe wall thickness (m), R is the bending radius (m), E is the modulus of elasticity (103 MPa), C is the pipe radius (m), and F is the stress factor.


Generally, the minimum pipe bend radius is required during the installation load case and the operational (in-service) load case [11].


10.10.3 Installation Condition


It is usually mandatory to specify the minimum radius of curvature permitted during the installation stage of a pipeline, whether it is S-lay, J-lay, or Tow-out, reeling of a pipeline [11].


The minimum bend radius for the installation condition should be calculated as


StartLayout 1st Row upper R equals StartFraction upper E times upper C Over sigma Subscript normal upper B Baseline EndFraction 2nd Row sigma Subscript normal upper L Baseline plus sigma Subscript normal upper B Baseline less-than-or-equal-to sigma Subscript a e EndLayout

From Equation (10.45),


StartLayout 1st Row sigma Subscript a e Baseline equals italic upper F upper S Subscript normal y Baseline 2nd Row sigma Subscript normal upper L Baseline plus sigma Subscript normal upper B Baseline less-than-or-equal-to italic upper F upper S Subscript normal y Baseline 3rd Row sigma Subscript normal upper B Baseline equals left-parenthesis upper F right-parenthesis left-parenthesis upper S Subscript normal y Baseline right-parenthesis minus sigma Subscript normal upper L Baseline 4th Row left-parenthesis sigma Subscript normal upper B Baseline right-parenthesis Subscript perm Baseline equals left-parenthesis italic upper F upper S Subscript normal y Baseline right-parenthesis minus sigma Subscript normal upper L Baseline 5th Row sigma Subscript normal upper L Baseline equals StartFraction italic upper P upper D Over 4 t EndFraction 6th Row therefore left-parenthesis sigma Subscript upper B Baseline right-parenthesis Subscript perm Baseline equals left-parenthesis italic upper F upper S Subscript normal y Baseline right-parenthesis minus StartFraction italic upper P upper D Over 4 t EndFraction EndLayout

(10.73)upper R Subscript perm Baseline equals StartFraction upper E times upper C Over left-parenthesis sigma Subscript normal upper B Baseline right-parenthesis Subscript perm Baseline EndFraction

where (σB)perm is the permissible bending stress (MPa).


All other parameters are as defined in Section 10.10.2.


10.10.4 In-Service Condition


(10.74)upper R Subscript normal a Baseline equals StartFraction upper E times upper C Over upper E Subscript perm Baseline EndFraction

where Ra is the minimum as-laid radius and Eperm is the maximum permissible permanent elastic strain.


10.10.4.1 Pipeline Located Offshore


On the other hand, for a pipeline located offshore, the minimum bend (curvature) in order to prevent slippage is calculated by the following equation:


(10.75)upper R Subscript s l Baseline equals StartFraction upper T Subscript normal o Baseline Over upper W Subscript normal s Baseline upper F EndFraction

where Rsl is the minimum bend to prevent slippage (limiting slippage curvature), To is the normal on bottom tension, Ws is the submerged unit weight, and F is the lateral friction coefficient or the on-bottom friction coefficient.


upper T Subscript normal o Baseline equals upper T Subscript normal i Baseline minus italic upper W upper H

where T0 is the maximum tension, H is the water depth, and Ti is the tension at the inflexion point.


10.11 Pipeline Design for External Pressure


10.11.1 Buried Installation


10.11.1.1 Check for Buckling


Buried pipelines supported by a well-compacted, granular backfill will not buckle due to vacuum (i.e., when the gauge pressure is below atmospheric pressure) in the pipeline system. To confirm the stability of a pipeline, an analysis of the external loads relative to the pipe stiffness can be performed.


If the soil and surface loads are excessive, the pipe cross-section could buckle. The ring buckling depends on limiting the total vertical pressure load on the pipe [20, 22].


(10.76)sigma Subscript r b Baseline equals StartFraction 1 Over upper F upper S EndFraction StartRoot 32 upper R Subscript normal w Baseline upper B prime upper E prime StartFraction left-parenthesis italic upper E upper I right-parenthesis Subscript e q Baseline Over upper D cubed EndFraction EndRoot

where σrb is the allowable buckling pressure (MPa), FS is the factor of safety, C is the depth of soil cover above the pipe (m), D is the outside diameter of the pipe (m), Rw is the water buoyancy factor = 1 − 0.33(hw/C), 0 < hw < C, hw is the height of the water surface above top of the pipe, (EI)eq is the equivalent pipe stiffness, E ′ is the modulus of soil reaction (GPa), and B ′ is the empirical coefficient of elastic support.


(10.77)upper B prime equals StartFraction 1 Over 1 plus 4 normal e Superscript left-parenthesis minus 0.065 upper C slash upper D right-parenthesis Baseline EndFraction

In steel pipelines, buckling occurs when the ovality reaches about 20%.


The sum of the external loads should be less than or equal to the pipe’s allowable buckling pressure, σrb.


Confirming the resistance of a pipeline to bucking involves an analysis of the external loads relative to the pipe stiffness [20]. The total external load acting on a pipe must be less than or equal to the allowable buckling pressure of the pipe and can be calculated using the following equation:


(10.78)gamma Subscript normal upper W Baseline h Subscript normal upper W Baseline plus upper R Subscript normal w Baseline StartFraction upper W Subscript normal c Baseline Over upper D EndFraction plus upper P Subscript normal v Baseline less-than-or-equal-to sigma Subscript r b

where σrb is the allowable buckling pressure (MPa), γW is the specific weight of water (0.0361 lb/in.3), hw is the height of water above the pipe (in., m), Rw is the water buoyancy factor, Wc is the vertical soil load on the pipe per unit length (lb/in., kg/m), D is the outside diameter (m), and Pv is the internal vacuum pressure (MPa).


The total live load acting on a pipe must be less than or equal to the allowable buckling pressure of the pipe. When analyzing the possibility of potential buckling of a pipeline, there is the need to determine the total live loads acting on a pipeline, using the following equation:


(10.79)gamma Subscript normal upper W Baseline h Subscript normal upper W Baseline plus upper R Subscript normal w Baseline StartFraction upper W Subscript normal c Baseline Over upper D EndFraction plus StartFraction upper W Subscript normal upper L Baseline Over upper D EndFraction less-than-or-equal-to sigma Subscript r b

where WL is the live load on the pipe per unit length (lb/in., kg/m).


When the allowable buckling pressure is not sufficient to resist the buckling loads, the soil envelope should first be investigated to increase the allowable E ′ [20].


10.11.2 Above-Ground or Unburied Installation


It is recommended to use propagation criteria for pipeline diameters under 16 in. and a collapse criterion for pipeline diameters above or equal to 16 in.


The propagation criterion is out of date and should be used where optimization of the wall thickness is not required or for pipeline installation methods not compatible with the use of buckle arrestors such as reel and tow methods [20].


It is generally economical to design for propagation pressure for diameters less than 16 in. For greater diameters, the wall thickness penalty is too high. When a pipeline is designed based on the collapse criteria, buckle arrestors are recommended.


10.11.2.1 Collapse Criterion


When a pipeline installed above ground is subjected to vacuum, the wall thickness must be designed to resist collapse due to the vacuum. Analysis should be based on pipe functioning in the open atmosphere, absent of support from any backfill material [20]. Collapse pressure, pc, is the pressure required to buckle a pipeline.


The collapse pressure should be calculated using Timoshenko’s theory for collapse of a round steel pipe as follows [20]:


(10.80)upper P Subscript normal c Baseline equals StartFraction 2 upper E Subscript normal s Baseline left-parenthesis t Subscript normal s Baseline slash d Subscript normal n Baseline right-parenthesis cubed Over left-parenthesis 1 minus v Subscript normal s Superscript 2 Baseline right-parenthesis EndFraction plus StartFraction 2 upper E Subscript normal upper I Baseline left-parenthesis t Subscript normal upper I Baseline slash d Subscript normal n Baseline right-parenthesis cubed Over left-parenthesis 1 minus v Subscript normal upper I Superscript 2 Baseline right-parenthesis EndFraction plus StartFraction 2 upper E Subscript normal c Baseline left-parenthesis t Subscript normal c Baseline slash d Subscript normal n Baseline right-parenthesis cubed Over left-parenthesis 1 minus v Subscript normal c Superscript 2 Baseline right-parenthesis EndFraction

where Pc is the collapsing pressure (MPa), ts is the steel cylinder wall thickness (m), tI is the cement coating thickness (m), dn is the diameter to the neutral axis of the shell (m), Es is the modulus of elasticity for steel (30 × 106 psi, 207 × 103 MPa), EI and Ec are the moduli of elasticity for cement mortar (4 × 106 psi, 27.6 × 103 MPa), vs is the Poisson’s ratio for steel (0.30), and vI and vc are the Poisson’s ratios for cement mortar (0.25).


The mode of collapse is a function of D/t ratio, pipeline imperfections, and load conditions. A safety factor of 1.3 is recommended. When a pipeline is designed using the collapse criterion, a good knowledge of the loading conditions is required.


10.11.2.2 Propagation Criterion


Propagating pressure, Pp, is the pressure required to continue a propagating buckle. A propagating buckle will stop when the pressure is less than the propagating pressure (refer to DNV 2012, Clause 501). The recommended formula for calculating the propagation criterion is the latest given by AGA [18].


(10.81)upper P Subscript normal p Baseline equals 33 upper S Subscript normal y Baseline left-parenthesis StartFraction t Subscript n o m Baseline Over upper D EndFraction right-parenthesis Superscript 2.4

The nominal wall thickness (tnom) should be determined such that



where Pp is the propagation pressure (MPa) and Pe is the external pressure (MPa).


The recommended safety factor of 1.3 is to account for uncertainty in the envelope of data points used to derive Equation (10.82).


10.12 Check for Hydrotest Conditions


In order to check that the pipeline is fit for the purpose for which it was designed, a pressure test of a harmless fluid prior to commissioning is required by most design codes and safety legislation. Most pressure tests of subsea pipelines are done with water, but on some occasions, nitrogen or air has been used.


The minimum hydrotest pressure for gas pipelines is equal to 1.25 times the design pressure for pipelines [18]. Codes do not require that the pipeline be designed for hydrotest conditions but sometimes give a tensile hoop stress limit 90% of the SMYS.


The pressure test level should be based on design codes and is aimed at



  • showing the integrity of the pipeline;
  • removing defects;
  • locating the presence of small leaks and pinholes.

10.13 Summary


This chapter on stress-based design of pipelines has provided a summary of the following:



  • The information needed for a stress-based design of pipelines and design consideration.
  • What to include in the pipeline design in order to ensure pipeline integrity.
  • Standardized methods to achieve safe and reliable pipeline design.
  • Pipeline integrity management approach that would meet industry needs.

Pipeline design, materials, and construction techniques are documented in design codes, client standards, handbooks, research papers, and so on.


Based on research, there is regular improvement in pipeline design, construction methods, and materials, and so it is important to access the most recent information on currently accepted and proven technologies.


Appendix 10.A.1 Case Study


This case study was part of a project consisting of an offshore pipeline system, a gas processing plant, and an onshore pipeline system that included a main line and a branch line. The offshore pipeline system included a 10-in. flexible riser installed to a FPSO (floating production storage and offloading) system resting on a PLET landing porch, a pipeline end termination (PLET), a 58-km pipeline with 12 in. outside diameter, and a pig launcher/receiver installed at both ends (see Figure 10.4).


The objective of this study was to establish that the pipeline meets or exceeds the requirements of standards.


The technical details and specification of the offshore gas export pipeline are listed in Tables 10.A.110.A.3.


Determination of Pipe Diameter


From process and hydraulic analysis, the flow rate, pipe diameter, operating temperature, and design pressure are as shown in Tables 10.A.110.A.3.


Wall Thickness Design


Based on the requirement of the ASME B31.8 code,


t equals StartFraction upper P times upper D Over 20 upper S Subscript y Baseline times upper E times upper F EndFraction plus t Subscript corr Baseline equals StartFraction 238 times 324 Over 20 times 450 times 1 times 0.72 EndFraction equals 10.9 m m

t Subscript n o m Baseline equals left-parenthesis t plus t Subscript corr Baseline right-parenthesis equals left-parenthesis 10.9 plus 3 right-parenthesis equals 14.9 m m
A graph shows water depth versus distance from F P S O. It starts at 0 meters depth, descending steeply to negative 2000 meters at 14.7 kilometers. There is a P L E T at 2.7 kilometers and a 10-flexible riser. A shallow water section extends 41.86 kilometers, followed by a 2 2-kilometer approach to the seashore, ending at 60 kilometer.

Figure 10.4 Profile of the offshore pipeline system.



Table 10.A.1 Pipe Material Properties




















































Description Value Unit
Material designation API-5L-X65
Material SMYS 450 MPa
Material SMTS 535 MPa
Steel density 7850 kg/m3
Young’s modulus 207 GPa
Poisson’s ratio 0.3
Shear modulus 80000 MPa
Coefficient of thermal expansion 0.00001125 1/°C
Design factor (F) 0.72
Joint factor (E) 1
Corrosion allowance, (tcorr) 3 mm

Table 10.A.2 Pipeline Specification
















Property Rigid pipeline
Nominal diameter 12 in.
Outer diameter (mm) 323.9
Roughness (mm) 0.04572

Table 10.A.3 Technical Specification of Pipeline

























Item Offshore Gas Export Pipeline
Design pressure (barg) 238
Operating pressure (barg) 207
Design temperature (°C) 70
Operating Temperature (°C) 50
Length (km) 58
Normal operating flow rate (MMSCFD) 120

From API 5L Standardized Pipe Schedule Chart, at 12 in. outer diameter and calculated t = 14.9 mm, tnom = 15.9 mm is selected.


therefore bold-italic nominal wall thickness equals bold 15.9 bold m m equals bold 0.625 bold in bold period

italic inner diameter equals italic outer diameter minus left-parenthesis 2 times italic nominal wall thickness right-parenthesis

italic inner diameter equals 323.9 m m minus left-parenthesis 2 times 15.9 m m right-parenthesis equals 292.1 m m

All parameters are defined in Table 10.A.1.


Determination of Components of Stress


The structural analysis of pipeline systems is concerned with the determination of stress or strain state of a pipeline and the subsequent check of the stress or strain state against a fatigue limit or allowable stress. The general term of structural analysis is indeed very wide, and this chapter considers the more important aspects in relation to pipeline systems in detail.


Hoop Stress


Hoop stress is calculated using the Barlow’s formula (Equation 10.4) as


sigma Subscript normal upper H Baseline equals StartFraction upper P upper D Over 2 t EndFraction equals StartFraction 23.8 times 0.3239 Over 2 times 0.0159 EndFraction equals 242.4157 upper M upper P a

bold-italic sigma Subscript bold upper H Baseline bold-italic equals bold 242.4 bold upper M upper P a

Thin-Walled Pipeline

With reference to Section 10.4.1.2, the standard dimension ratio is


StartFraction upper D Over t EndFraction equals StartFraction 0.3239 Over 0.0159 EndFraction equals 20.37 greater-than 20

Therefore, the offshore pipeline is a thin-walled pipeline.


Compressive Hoop Stress


For a deep-water application, the external hydrostatic pressure should be accounted for by using ∆P instead of P for hydrostatic conditions,


upper P 0 equals rams horn times StartFraction upper H Over 144 EndFraction

upper P Subscript normal e Baseline equals upper P 0 equals 64 times StartFraction 2624.67 Over 144 EndFraction equals 1166.52 upper P s i equals 8.04288 upper M upper P a

sigma Subscript normal upper H Baseline equals StartFraction left-parenthesis upper P Subscript normal i Baseline minus upper P Subscript normal e Baseline right-parenthesis upper D 2 Over 2 t EndFraction equals StartFraction left-parenthesis 23.8 minus 8.04288 right-parenthesis times 0.3239 Over 2 times 0.0159 EndFraction equals 160.4947 upper M upper P a

bold-italic sigma Subscript bold upper H Baseline bold-italic equals bold 160.5 bold upper M upper P a

where, H is the pressure head and ɤ is the unit weight of water. All other parameters are as defined in Sections 10.4.1.1 and 10.4.1.2.


Longitudinal Stress


Restrained Longitudinal Stress

A fully end constrained boundary condition occurs at pipeline end manifold (PLEM) or pipeline end termination (PLET) sled of the offshore pipeline, 2-km section buried and restrained by soil friction. In accordance with section 833.2 of ASME B31.8, the net longitudinal stress is calculated as


sigma Subscript normal upper L Baseline equals sigma Subscript upper L upper P Baseline plus sigma Subscript upper L upper R Baseline plus left-parenthesis sigma Subscript x Baseline right-parenthesis Subscript max

Longitudinal Stress due to Internal Pressure


The longitudinal stress due to internal pressure in a fully restrained pipeline is calculated as


sigma Subscript upper L upper P Baseline equals 0.3 sigma Subscript normal upper H Baseline equals 0.3 times StartFraction upper P upper D Over 2 t EndFraction equals 0.3 left-parenthesis StartFraction 23.8 times 0.3239 Over 2 times 0.0159 EndFraction right-parenthesis equals 72.724 upper M upper P a

bold-italic sigma Subscript bold upper L upper P Baseline bold-italic equals bold 72.72 bold upper M upper P a

Longitudinal Stress due to Thermal Expansion


The net longitudinal stress due to thermal expansion in a fully restrained pipeline is calculated as


sigma Subscript upper L upper R Baseline equals italic upper E alpha left-parenthesis upper T 1 minus upper T 2 right-parenthesis equals italic upper E alpha left-parenthesis upper T 2 minus upper T 1 right-parenthesis equals 207000 times 0.00001125 left-parenthesis 27 minus 50 right-parenthesis equals minus 53.561 upper M upper P a

bold-italic sigma Subscript bold upper L upper R Baseline bold-italic equals minus bold 53.56 bold upper M upper P a

All parameters are as defined in Section 10.4.2.


Bending Stress


The bending stress is calculated using the engineer’s theory of bending in Section 10.4.2.4 and Equation (10.26).


Radius of curvature of the bent pipe, R


upper R equals 1000 upper D equals 1000 times 0.3239 equals 323.9 normal m

Bending stress


sigma Subscript x Baseline equals StartFraction italic upper E y Over upper R EndFraction equals StartFraction 207000 times left-parenthesis 0.3239 slash 2 right-parenthesis Over 323.9 EndFraction equals 103.5 upper M upper P a

Maximum bending stress


left-parenthesis sigma Subscript x Baseline right-parenthesis Subscript max Baseline equals StartFraction upper M upper D Subscript normal o Baseline Over 2 upper I EndFraction equals StartFraction upper E upper D Subscript normal upper O Baseline Over 2 upper R EndFraction equals StartFraction 207000 times left-parenthesis 0.3239 right-parenthesis Over 2 times 323.9 EndFraction equals 103.5 upper M upper P a

left-parenthesis bold-italic sigma Subscript bold-italic x Baseline right-parenthesis Subscript bold max Baseline bold equals bold 103.5 bold upper M upper P a

All parameters are defined in Section 10.4.2.4.


Total Longitudinal Stress for the Restrained Section


sigma Subscript normal upper L Baseline equals sigma Subscript upper L upper P Baseline plus sigma Subscript upper L upper R Baseline plus left-parenthesis sigma Subscript x Baseline right-parenthesis Subscript max Baseline equals 72.724 plus StartAbsoluteValue negative 53.561 EndAbsoluteValue plus 103.500 equals 122.663 upper M upper P a

bold-italic sigma Subscript bold upper L Baseline bold-italic equals bold 122.7 bold upper M upper P a

Axial Compressive Force

The axial compressive force required to restrain a thin-walled pipeline can be calculated Equation (10.16) as follows:


Cross-Sectional Pipe Wall Area


The cross-sectional wall area or area of the piping material can be calculated as


upper A equals StartFraction pi left-parenthesis d Subscript normal o Baseline Superscript 2 Baseline minus d Subscript normal i Baseline Superscript 2 Baseline right-parenthesis Over 4 EndFraction

upper A equals StartFraction 3.142 left-parenthesis 0.3239 squared minus 0.2921 squared right-parenthesis Over 4 EndFraction equals 0.015387 normal m squared

Axial compressive force

upper F equals upper A left-parenthesis italic upper E alpha left-parenthesis upper T 2 minus upper T 1 right-parenthesis plus 0.5 italic sigma upper H minus italic v with hook sigma upper H right-parenthesis

upper F equals 0.015387 times left-bracket 207000000 times 0.00001125 left-parenthesis 50 minus 70 right-parenthesis plus left-parenthesis 0.5 times 242416 right-parenthesis minus left-parenthesis 0.30 times 242416 right-parenthesis right-bracket equals 29.361 k upper N

bold-italic upper F bold-italic equals bold 29.36 bold k upper N

All parameters are defined in Section 10.4.2.1.


Unrestrained Longitudinal Stress

An end-free boundary condition can occur at locations where no physical longitudinal restraint exists; for example, at a riser bend extending from the seabed to the production platform or PLET (see Figure 10.4). The unrestrained longitudinal stress in an operational pipeline can generally be expressed as


sigma Subscript normal upper L Baseline equals sigma Subscript upper L upper P Baseline plus sigma Subscript upper L x Baseline plus left-parenthesis sigma Subscript x Baseline right-parenthesis Subscript max

Longitudinal Stress due to Internal Pressure


The longitudinal stress due to internal pressure in an unrestrained pipeline is calculated as


sigma Subscript upper L upper P Baseline equals 0.5 sigma Subscript normal upper H

sigma Subscript upper L upper P Baseline equals 0.5 times StartFraction upper P upper D Over 2 t EndFraction equals 0.5 left-parenthesis StartFraction 23.8 times 0.3514 Over 2 times 0.0487 EndFraction right-parenthesis equals 42.933 upper M upper P a

bold-italic sigma Subscript bold upper L upper P Baseline bold-italic equals bold 42.93 bold upper M upper P a

Bending Stress


The bending stress is calculated using the engineer’s theory of bending in Equation (10.26) as


left-parenthesis sigma Subscript x Baseline right-parenthesis Subscript max Baseline equals StartFraction upper M upper D Subscript normal o Baseline Over 2 upper I EndFraction equals StartFraction upper E upper D Subscript normal upper O Baseline Over 2 upper R EndFraction comma y equals upper O upper D slash 2

Radius of curvature, R


upper R equals 1000 upper D equals 1000 times 0.3514 equals 351.4 normal m

upper R equals 351.4 normal m

R = radius of curvature of the bent pipe, m


Bending stress,


With reference to Section 10.4.2.4 and Equation (10.25),


sigma Subscript x Baseline equals StartFraction italic upper E y Over upper R EndFraction equals StartFraction 207000 times left-parenthesis 0.3514 slash 2 right-parenthesis Over 351.4 EndFraction equals 103.5 upper M upper P a

Maximum bending stress,


left-parenthesis sigma Subscript x Baseline right-parenthesis Subscript max Baseline equals StartFraction upper M upper D Subscript normal o Baseline Over 2 upper I EndFraction equals StartFraction upper E upper D Subscript normal upper O Baseline Over 2 upper R EndFraction equals StartFraction 207000 times left-parenthesis 0.3514 right-parenthesis Over 2 times 351.4 EndFraction equals 103.5 upper M upper P a

left-parenthesis bold-italic sigma Subscript bold-italic x Baseline right-parenthesis Subscript bold max Baseline bold-italic equals bold 103.5 bold upper M upper P a

All parameters are defined in Section 10.4.2.4.


Total Longitudinal Stress for the Unrestrained Section


In accordance with Section 10.4.2, the total longitudinal stress is calculated as


sigma Subscript normal upper L Baseline equals sigma Subscript upper L upper P Baseline plus sigma Subscript upper L x Baseline plus left-parenthesis sigma Subscript x Baseline right-parenthesis Subscript max

sigma Subscript upper L u Baseline equals 42.933 plus 0 plus 103.500 equals 146.433 upper M upper P a

bold-italic sigma Subscript bold upper L u Baseline bold equals bold 146.4 bold upper M upper P a

where


sigma Subscript italic upper L upper U Baseline is the unrestrained longitudinal stress comma left-parenthesis upper M upper P a right-parenthesis period

Equivalent Stress


In accordance with Section 10.4.4, and Equation (10.43), equivalent stress in the pipeline is


sigma Subscript normal e Baseline equals StartRoot left-parenthesis 242.416 right-parenthesis squared plus left-parenthesis 122.663 right-parenthesis squared minus left-parenthesis 242.416 times 122.663 right-parenthesis EndRoot

bold-italic sigma Subscript bold e Baseline bold equals bold 209.9 bold upper M upper P a

All parameters are defined in Section 10.4.4.


Limit of Calculated Stress


Design factor, F = 0.72


Specified minimum yield strength for X65M, Sy = 450 MPa


Pipe material grade = API 5L X65M PSL2


italic Allowable stress equals italic design factor times italic specified minimum yield strength

italic Allowable stress equals 0.72 times 450 equals 324 upper M upper P a

Allowable Longitudinal Stress


Restrained Section

For a restrained pipeline, the allowable longitudinal stress is 0.9 Sy × T, (refer to Equation (10.43) and Section 10.4.5)


sigma Subscript upper A upper L Baseline equals 0.9 upper S Subscript normal y Baseline times upper T equals 0.9 times 450 times 1 equals 405 upper M upper P a

sigma Subscript upper L upper R Baseline less-than sigma Subscript upper A upper L

Therefore, the requirements of the code are met with respect to longitudinal stress in the restrained section.


All parameters are defined in Section 10.4.5.


Unrestrained Section


Based on Clause 833.6 of ASME B 31.8, for an unrestrained pipe, the allowable longitudinal stress is σAL ≤ 0.75 Sy × T, where Sy is the specified minimum yield strength, (MPa), and T is the temperature derating factor


sigma Subscript upper A upper L Baseline equals 0.75 upper S Subscript normal y Baseline times upper T equals 0.75 times 450 times 1 equals 337.5 upper M upper P a

sigma Subscript upper L upper U Baseline less-than sigma Subscript upper A upper L

Therefore, the requirements of the code are met with respect to longitudinal stress in the unrestrained section.


Allowable Hoop Stress

In accordance with Equation (10.41),


sigma Subscript a upper H Baseline equals f times e times upper T times upper S Subscript normal y Baseline equals 0.72 times 1 times 1 times 450 equals 324 upper M upper P a

The calculated hoop stress σH = 160.49474 MPa


therefore sigma Subscript normal upper H Baseline less-than sigma Subscript a upper H

Therefore, the requirements of the code are met with respect to hoop stress


All parameters are defined in Section 10.4.5.1.


Allowable Equivalent Stress

The maximum allowable equivalent stress is 90% of the SMYS (refer to Section 10.4.5, and Equation (10.45))


sigma Subscript a e Baseline equals 0.9 upper S Subscript normal y Baseline equals 0.9 times 450 equals 405 upper M upper P a

sigma Subscript normal e Baseline less-than sigma Subscript a e

Therefore, the requirements of the code are met with respect to equivalent stress.


All parameters are defined in Section 10.4.5.2.


Flexibility Analysis


As per clause 833.7 of ASME B 31.8, no formal analysis is required in systems which are of uniform size, have no more than two points of fixation, no intermediate restraints, and fall within the empirical equation below.


upper K greater-than-or-equal-to StartFraction italic upper D upper Y Over left-parenthesis upper L minus upper U right-parenthesis squared EndFraction

where



  • D = outside diameter of the pipe
  • Y = the resultant of total displacement strains to be absorbed by the piping system
  • L = developed length between anchors
  • U = anchor distance, straight line between anchors
  • Note:

Calculation of L and U is based on Figure 10.4.


K = 208.3 For SI units or K = 0.03 for FPS units


Displacement Strain

Calculate displacement due to thermal expansion. The entire offshore system is considered, i.e., both riser and the entire offshore pipeline system.

Thermal Strain, ∆L

The change in length of the thin-walled pipeline is determined from the thermal strain.


increment upper L equals alpha left-bracket upper T 2 minus upper T 1 right-bracket upper L

upper L equals 2.7 k m plus 58 k m equals 60.7 k m equals 60700 normal m

increment upper L equals 0.00001125 left-bracket 50 minus 27 right-bracket left-parenthesis 60700 right-parenthesis equals 15.7 normal m

increment upper L equals 15.7 m

Flexibility Check


upper K greater-than-or-equal-to StartFraction italic upper D upper Y Over left-parenthesis upper L minus upper U right-parenthesis squared EndFraction

upper L equals 2.7 plus 14.14 plus 41.86 plus 2 equals 60.7 k m

upper U equals StartRoot left-parenthesis 2.7 right-parenthesis squared plus left-parenthesis 14.14 right-parenthesis squared plus left-parenthesis 41.86 plus 2 right-parenthesis squared EndRoot equals 46.162 k m

StartFraction italic upper D upper Y Over left-parenthesis upper L minus upper U right-parenthesis squared EndFraction equals StartFraction 0.3239 times 15.706 Over left-parenthesis 60700 minus 46162 right-parenthesis squared EndFraction equals 2.407 times 10 Superscript negative 8

The flexibility criterion is less than 208.3; therefore, no further or formal flexibility analysis is required. That is, bold-italic upper K greater-than StartFraction bold-italic upper D upper Y Over left-parenthesis bold-italic upper L minus bold-italic upper U right-parenthesis squared EndFraction


External Pressure check


The ASME code does not provide a formula to check for collapse resistance; thus, the API RP-1111 is used.


Collapse Criterion

Yield Pressure and Collapse


upper P Subscript normal y Baseline equals 2 upper S Subscript normal y Baseline left-parenthesis StartFraction t Over upper D EndFraction right-parenthesis equals 2 times 450 left-parenthesis StartFraction 0.0159 Over 0.3239 EndFraction right-parenthesis equals 44.180 upper M upper P a

bold-italic upper P Subscript bold y Baseline bold approximately-equals bold 44.2 bold upper M upper P a

Elastic Collapse Pressure


upper P Subscript normal e Baseline equals 2 upper E StartFraction left-parenthesis StartFraction t Over upper D EndFraction right-parenthesis cubed Over left-parenthesis 1 minus v squared right-parenthesis EndFraction equals 2 times 207000 StartFraction left-parenthesis StartFraction 0.0159 Over 0.3239 EndFraction right-parenthesis cubed Over left-parenthesis 1 minus 0.3 squared right-parenthesis EndFraction equals 53.817 upper M upper P a

bold-italic upper P Subscript bold e Baseline bold-italic equals bold 53.82 bold upper M upper P a

Collapse Pressure


upper P Subscript normal c Baseline equals StartFraction upper P Subscript normal y Baseline upper P Subscript normal e Baseline Over StartRoot upper P Subscript normal y Superscript 2 Baseline upper P Subscript normal e Superscript 2 Baseline EndRoot EndFraction equals StartFraction 44.18 times 53.817 Over StartRoot 44.18 squared times 53.817 squared EndRoot EndFraction equals 34.147 upper M upper P a

bold-italic upper P Subscript bold c Baseline bold-italic equals bold 34.15 bold upper M upper P a

f 0 times upper P Subscript normal c Baseline equals 0.70 times 34.147 equals 23.90323

upper P 0 equals gamma times StartFraction upper H Over 144 EndFraction

upper P Subscript normal e Baseline equals 64 times StartFraction 2624.67 Over 144 EndFraction equals 1166.52 upper P s i equals 8.04288 upper M upper P a

StartAbsoluteValue upper P Subscript normal e Baseline minus upper P Subscript normal i Baseline EndAbsoluteValue equals 8.04288 minus 0 equals 8.043 upper M upper P a left-parenthesis italic during installation left-parenthesis italic empty pipe right-parenthesis right-parenthesis

StartAbsoluteValue upper P Subscript normal e Baseline minus upper P Subscript normal i Baseline EndAbsoluteValue equals 8.04288 minus 23.8 equals minus 15.75 upper M upper P a left-parenthesis italic during operation right-parenthesis

StartAbsoluteValue upper P Subscript normal e Baseline minus upper P Subscript normal i Baseline EndAbsoluteValue Subscript max Baseline equals 8.043 upper M upper P a

StartAbsoluteValue bold-italic upper P Subscript bold e Baseline minus bold-italic upper P Subscript bold i Baseline EndAbsoluteValue Subscript bold max Baseline bold-italic less-than-or-equal-to bold-italic f bold 0 bold-italic times bold-italic upper P Subscript bold c Baseline bold Pipeline design passes collapse criterion check

Propagation Criterion


The recommended formula for calculating the propagation criterion is the latest given by AGA (1990)


upper P Subscript normal p Baseline equals 33 upper S Subscript normal y Baseline left-parenthesis StartFraction t Over upper D EndFraction right-parenthesis Superscript 2.4 Baseline equals 33 times 450 left-parenthesis StartFraction 0.0159 Over 0.3239 EndFraction right-parenthesis Superscript 2.4 Baseline equals 10.717 upper M upper P a

bold-italic upper P Subscript bold p Baseline bold-italic equals bold 10.72 bold upper M upper P

0.8 upper P Subscript normal p Baseline equals 8.574 upper M upper P a

StartAbsoluteValue bold-italic upper P Subscript bold-italic e Baseline minus bold-italic upper P Subscript bold-italic i Baseline EndAbsoluteValue Subscript bold-italic max Baseline bold-italic less-than-or-equal-to bold 0.8 bold-italic upper P Subscript bold-italic p Baseline comma bold-italic therefore bold-italic n o bold-italic buckle arrestor is required

1.3 upper P Subscript normal e Baseline equals 1.3 times 8.043 equals 10.46 upper M upper P a

upper P Subscript p Baseline greater-than 1.3 upper P Subscript e Baseline comma italic therefore the pipeline design passes the propagation italic criterion check

Bending Buckling Check


epsilon Subscript normal b Baseline equals StartFraction t Over 2 upper D EndFraction equals StartFraction 0.0159 Over 2 times 0.3239 EndFraction equals 0.02454

g left-parenthesis delta right-parenthesis equals left-parenthesis 1 plus 20 delta right-parenthesis Superscript negative 1 Baseline equals left-parenthesis 1 plus 20 times 0.01 right-parenthesis Superscript negative 1 Baseline equals 0.833

During installation,


StartFraction epsilon Over epsilon Subscript normal b Baseline EndFraction plus StartFraction upper P Subscript normal e Baseline minus upper P Subscript normal i Baseline Over upper P Subscript normal c Baseline EndFraction less-than-or-equal-to g left-parenthesis delta right-parenthesis

StartFraction epsilon Over epsilon Subscript normal b Baseline EndFraction plus StartFraction upper P Subscript normal e Baseline minus upper P Subscript normal i Baseline Over upper P Subscript normal c Baseline EndFraction equals StartFraction 0.005 Over 0.02454 EndFraction plus StartFraction 8.04288 Over 34.14747 EndFraction equals 0.44

During operation,


StartFraction epsilon Over epsilon Subscript normal b Baseline EndFraction plus StartFraction upper P Subscript normal e Baseline minus upper P Subscript normal i Baseline Over upper P Subscript normal c Baseline EndFraction less-than-or-equal-to g left-parenthesis delta right-parenthesis

StartFraction epsilon Over epsilon Subscript normal b Baseline EndFraction plus StartFraction upper P Subscript normal e Baseline minus upper P Subscript normal i Baseline Over upper P Subscript normal c Baseline EndFraction equals StartFraction 0.003 Over 0.02454 EndFraction plus StartFraction negative 15.75712 Over 34.14747 EndFraction equals negative 0.66

StartFraction bold-italic epsilon Over bold-italic epsilon Subscript bold b Baseline EndFraction bold-italic plus StartFraction bold-italic upper P Subscript bold e Baseline minus bold-italic upper P Subscript bold i Baseline Over bold-italic upper P Subscript bold c Baseline EndFraction bold-italic less-than-or-equal-to bold-italic g left-parenthesis bold-italic delta right-parenthesis italic Therefore comma italic the pipeline meets the requirements of the code italic for bending buckling

Where,ε is bending strain, P0 is external pressure, Pi is internal pressure, fo is collapse factor, (0.7 for seamless or ERW pipe), v is Poisson’s ratio (0.30 for steel), ɤ is sea water density, and Pp is propagation pressure.


Conclusion


The stress analysis shows that the design meets the requirements of ASME B31.8, BS 8010, API RP-1111, DNV—OS- F101, and IGE/TD/1.


References



  1. 1 Canadian Standards Association (2023) CSA Z662:2023: Oil and Gas Pipeline Systems, Canadian Standards Association, Mississauga, Ontario, Canada.
  2. 2 American Society of Mechanical Engineers (2003) ASME B31.8-2003: Gas Transmission and Distribution Piping Systems. American Society of Mechanical Engineers, New York.
  3. 3 American Society of Mechanical Engineers (2000) ASME B31.3-2000: Code for Pressure Piping, American Society of Mechanical Engineers, New York.
  4. 4 British Standard Institute (1992) BS 8010-1992: Code of Practice for Pipelines, British Standard Institute, UK.
  5. 5 JP Kenny Group (1994) Engineering Design Guideline, JP Kenny Group, UK.
  6. 6 American Society of Mechanical Engineers (2006) ASME B31.4-2006: Pipeline Transportation Systems for Liquid Hydrocarbons and Other Liquids, American Society of Mechanical Engineers, New York.
  7. 7 Det Norske Veritas (2012) DNV-OS-F101-2012: Submarine Pipeline Systems—Offshore Standard, Det Norske Veritas.
  8. 8 Institution of Gas Engineers (2001) IGE/TD/1-2001: Steel Pipelines for High Pressure Gas Transmission, 4th ed., Institution of Gas Engineers, London, UK.
  9. 9 Det Norske Veritas (2008) DNV-RP-D101-2008: Structural Analysis of Piping Systems, Det Norske Veritas, Norway.
  10. 10 Symonds, J., Vidosic, J.P., Hawkins, H.V., and Dodge, D.D. (1996) Strength of materials. In: Marks’ Standard Handbook for Mechanical Engineers, 10th ed., McGraw-Hill, New York, Section 5, 5-2–5-67.
  11. 11 Stewart, M. (1997) Pipeline Engineering 2, International Training and Development, Jakarta.
  12. 12 Chung, P.L. (1978) Stress analysis methods for underground pipe lines. OT Pipe Line Industry, April/May, Parts 1 and 2.
  13. 13 Hauch, S. and Bai, Y. (1999) Bending moment capacity of pipes. Proceedings of Offshore Mechanics and Arctic Engineering (OMAE), Houston, TX, Paper No. PL-99-5033.
  14. 14 Alexander, C. (2012) Evaluating the effects of ovality on the integrity of pipe bends. Proceeding of the 9th International Pipeline Conference, American Society of Mechanical Engineers, Calgary, Alberta, Canada, Paper No. IPC2012-90582.
  15. 15 Roylance, D. (2000) Shear and Torsion, Massachusetts Institute of Technology, Cambridge.
  16. 16 Timoshenko, S.P. and Gere, J.M. (1999) Fatigue. In: Mechanics of Materials, 3rd ed., Stanley Thornes Publishers Ltd, UK, p. 116.
  17. 17 Chernilevsky, D.V., Lavrova, E.V., and Romanov, V.A. (1984) Fatigue. In: Mechanics for Engineers, Moscow:MIR Publishers, Moscow, Russia, pp. 347–381.
  18. 18 Guo, B. and Ghalambor, A. (2005) Natural Gas Engineering Handbook, 2nd ed., Gulf Publishing, Houston, TX, Chapter 10.
  19. 19 Hopkins, P. (2003) The Structural Integrity of Oil and Gas Transmission Pipelines: Comprehensive Structural Integrity, Elsevier, Vol. 1.
  20. 20 American Spiral Weld Pipe (2013) ASWP Manual—2013. Steel Pipe Design, American Spiral Weld Pipe, North Carolina, USA.
  21. 21 American Concrete Pipe Association (2001) ACPA-2001. Buried Facts: Structural Design Consideration, American Concrete Pipe Association, USA.
  22. 22 American Lifeline Alliance (2001) ALA-2001: Guidelines for the Design of Buried Steel Pipe, American Lifeline Alliance, USA (with addenda through February 2005). Available at http://www.americanlifelinesalliance.com/pdf/Update061305.pdf.
  23. 23 Calladine, C.R. (1989) Theory of Shell Structures: The Brazier Effect in the Buckling of Bent Tubes, Cambridge University Press, Cambridge, UK, p. 595.
  24. 24 Ellinas, C.P., Raven, P.W.J., Walker, A.C., and Davies, P. (1987) Limit state philosophy in pipeline design. Proceedings of Offshore Mechanics and Artic Engineering (OMAE).
  25. 25 Emerson, E. (2010) Factors that affect the bending radius of pipe cable. Available at www.had2know.com.

May 10, 2025 | Posted by in General Engineer | Comments Off on Stress-Based Design of Pipelines
Premium Wordpress Themes by UFO Themes