ECAs: Lifting the lid of the Black Box

Ninth Planet Engineering Limited, Newcastle upon Tyne, UK

Kenneth A Macdonald 
University of Stavanger, Stavanger, Norway

Isabel Hadley and Philippa Moore 
TWI Ltd, Granta Park Great Abington, Cambridge, CB21 6AL, UK

Proceedings of the ASME 2017 36th International Conference on Ocean, Offshore and Arctic Engineering OMAE2017 June 25-30, 2017, Trondheim, Norway

Abstract

An engineering critical assessment (ECA) is commonly conducted during the design of a pipeline in order to determine the tolerable sizes for flaws in the girth welds. An ECA is a method for assessing the acceptability of a flaw in a structure, i.e. to demonstrate fitness-for-service. API 579-1/ASME FFS-1 2016 and BS 7910:2013+A1:2015 Incorporating Corrigenda Nos. 1 and 2 give guidance for conducting fitness-for-service assessments of cracks and crack-like flaws. Appendix A of DNV-OS-F101, 2013 provides detailed procedures for evaluation of the fracture limit state of girth welds, based on BS 7910:2005, but prescribes a number of amendments and modifications. API 579-1/ASME FFS-1 and BS 7910:2013 will give similar, but slightly different, results.

An ECA can be something of a black box because of the use of software to conduct the calculations. Software can have the unfortunate side-effect of obscuring the understanding of the calculations. ECAs are also seen as a thing that is very complicated.

In an attempt to lift the lid on the black box that is, or is perceived to be, the ECA, the assessment of circumferentially orientated, surface-breaking crack-like flaw (a planar flaw) in a girth weld is illustrated through a comparison of the relevant stress intensity factor and reference stress solutions in API 579-1/ASME FFS-1 and BS 7910:2013; and through illustrating the effect of the choice of the Level or Option of the assessment, and assumptions made in the assessment with respect to constraint, misalignment, residual stresses, etc. Also illustrated are the implications of defining the fracture toughness in terms of the crack tip opening displacement or J-integral, and of using a single value of fracture toughness, based on the initiation of stable tearing or the first attainment of a maximum force plateau (maximum load), or of using tearing resistance data.

A comparison is also made with the notionally simplified fitness-for-service procedures given in Annex A of API Standard 1104, Annex K of CSA Z662-15, and the EPRG guidelines on the assessment of defects in transmission pipeline girth welds – Revision 2014.

The result is an outline of what might be considered to be good practice for when conducting an ECA during the design of a pipeline.

Introduction

An engineering critical assessment (ECA) is commonly conducted to calculate the tolerable size of flaws in pipeline girth welds. The flaw acceptance criteria that are used during installation or construction are then informed by these tolerable flaw sizes, the capabilities (probability of detection and sizing accuracy) of the inspection method and other practical considerations. An ECA is a method for assessing the acceptability of a flaw in a structure, i.e. to demonstrate fitness-for-service.

Pipeline welding codes and standards, e.g. API Standard 1104 [1], BS 4515-1 [2], CSA Z662-15 [3] and DNV-OS-F101 [4], specify workmanship acceptance levels for welding defects in pipeline girth welds. These acceptance levels represent what a ‘good’ welder should be able to achieve. They are not fitness-for-service defect limits. Girth welds are inspected using radiography or ultrasonics. Automatic ultrasonic inspection of pipeline girth welds is becoming the dominant method of inspection in modern pipelines. Radiographic inspection gives an indication of the type of defect and a measurement of the length of the defect, but no direct measurement of the defect height. Automated ultrasonic inspection gives information on the height, length and position of a defect, although it gives only a limited indication of the defect type. Workmanship acceptance levels are historically based on the assumption that the welds will be inspected using radiography, and hence are specified in terms of the length of the defect and the type of welding defect. An ECA can also be used to define acceptance limits, i.e. limits based on fitness-for-service. Acceptance limits are specified in terms of defect length and height. Acceptance limits that are based on fitness-for-service calculations are normally larger than workmanship acceptance levels. That said, the limits might be smaller if the loading or environment are severe. The reasons for conducting ECAs are twofold: (1) to avoid unnecessary repairs, or (2) to determine if workmanship acceptance levels are themselves fit-for-service. The use of an ECA should not be viewed as an alternative to good workmanship.

API 579-1/ASME FFS-1 2016 [5] and BS 7910:2013+A1:2015 Incorporating Corrigenda Nos. 1 and 2 [6] give guidance for conducting fitness-for-service assessments of cracks and crack-like flaws. API 579-1/ASME FFS-1 2016 is less commonly used in the pipeline industry, at least in the context of calculating the tolerable size of flaws in pipeline girth welds. Appendix A of DNV-OS-F101 provides detailed procedures for evaluation of the fracture limit state of girth welds (i.e. an engineering critical assessment). It is based on BS 7910:2005 [7], but prescribes a number of amendments and modifications. It gives additional guidance on the assessment of girth welds subject to high strains (greater than 0.4 percent). DNV-RP-F108 [8] was developed to provide additional guidance for ECAs of girth welds subject to cyclic plastic strains during installation. Annex A of API Standard 1104 and Annex K of CSA Z662-15 give notionally simplified procedures for calculating the tolerable size of flaws in girth welds. Annexes A and K are based on similar principles to API 579-1/ASME FFS-1 2016 and BS 7910:2013, but simplified partly by virtue of their limited scope of application. The European Pipeline Research Group has produced guidelines on the assessment of defects in transmission pipeline girth welds [9]. The guidelines (hereafter the Guidelines) give defect acceptance levels, based on workmanship considerations, denoted Tier 1 (based on API Standard 1104 and BS 4515-1), and defect acceptance limits, based on fitness-for-service, denoted Tier 2 and Tier 3.

The assessment of circumferentially orientated, surface-breaking crack-like flaw (a planar flaw) in a girth weld is illustrated through a comparison of the relevant stress intensity factor and reference stress solutions in API 579-1/ASME FFS-1 and BS 7910:2013; and through illustrating the effect of the choice of the Level or Option of the assessment, and assumptions made in the assessment with respect to constraint, misalignment, residual stresses, etc.

API 579-1/ASME FFS-1 2016 and BS 7910:2013

API 579-1/ASME FFS-1 2016 Fitness-For-Service [5] provides guidance for conducting fitness-for-service assessments using methodologies specifically prepared for pressurised equipment. API RP 579 was published in January 2000. It was written as a recommended practice rather than as a mandatory code or standard. The American Petroleum Institute and the American Society of Mechanical Engineers established the Fitness-For-Service Joint Committee in 2001 and published the first edition of API 579-1/ASME FFS-1 Fitness-For-Service in June 2007. The second edition was published in June 2016. API 579-1/ASME FFS-1 is an American National Standard.

BS 7910:2013+A1:2015 Incorporating Corrigenda Nos. 1 and 2 Guide to methods for assessing the acceptability of flaws in metallic structures [6] gives guidance for assessing the acceptability of flaws in all types of (metallic) structures and components. The British Standards Institution first published guidance for the assessment of structural integrity of welded structures containing flaws, designated PD 6493, in 1980. It was extensively revised in 1991, incorporating more advanced fracture assessment methods from the nuclear industry. A British Standard was first published in 1999 under the direction of Technical Committee WEE/37 Acceptance levels for flaws in welds [10]. It was subsequently revised in 2005 and 2013. BS 7910:2013 is a full revision of the standard, and incorporates elements of FITNET [11,12] and R6 [13].

The fracture clauses in API 579-1/ASME FFS-1 2016 (hereafter API 579-1) and BS 7910:2013 (hereafter BS 7910) are broadly similar. Part 9 Assessment of Crack-Like Flaws in API 579-1 and Section 7 Assessment for fracture resistance in BS 7910 are directly comparable . Part 9 and Section 7 are both based on the Failure Assessment Diagram (FAD). Table 1 summarises and compares the Levels in Part 9 and the Options in Section 7. Levels 2 and 3A in Part 9 are approximately equivalent to Option 1 in Section 7. Level 3B is equivalent to Option 2.

A compendium of stress intensity factor solutions and reference stress solutions is given in API 579-1 (Annexes 9B and 9C) and BS 7910 (Annexes M and P). A number of the solutions are the same. The solutions in API 579-1 are generally applicable to a wider range of geometries.

A fitness-for-service assessment in accordance with one or the other of API 579-1 and BS 7910 will notionally give similar, but not necessarily identical, results. Larrosa and Ainsworth, 2015 [14] illustrate the differences using Example Problems 9.5 and 9.6 in API 579-2/ASME FFS-2 2009 [15], an axial internal surface flaw in cylinder and a fully-circumferential outside surface flaw in cylinder, respectively.

The illustrative assessment described herein is limited to the assessment of the fracture resistance of a circumferentially orientated, external surface crack-like flaw in a girth weld (so, Part 9 and Section 7).

Failure Assessment Diagram

The Option 1 FAD for a material that does not exhibit a yield discontinuity, with a yield strength equal to 450 N.mm-2 and a tensile strength equal to 535 N.mm-2, is given in Fig. 1 . (For comparison, a curve for a material that does exhibit a yield discontinuity is also shown in Fig. 1. The difference between the two curves illustrates the reasoning behind Clause 7.1.3.6 in BS7910, that “If there is uncertainty, a Lüders plateau should be

assumed for L_r ≥ 1 and no yield plateau for L_r < 1.”) The Level 2 FAD in API 579-1 is also given in Fig. 1.

Fig. 1 illustrates that the Level 2 and Option 1 failure assessment lines in API 579-1 and BS 7910 will lead to slightly different results.

An illustrative Option 2 FAD is given in Fig. 2. The Option 2 FAD requires detailed stress-strain data (a stress-strain curve). The Option 1 FAD is more conservative.

Stress Concentration Factor due to Axial Misalignment

The stress concentration factors due to axial misalignment in API 579-1 and BS 7910, listed in Table 2, are compared in Fig. 3. Table 8.9 Centerline – Circumferential Joint, Centerline Offset and Table D.1 d) Axial misalignment at girth welds in tubes, pipes, vessels and at seams in spheres, with or without thickness changes are based upon the simple theoretical formula for axial misalignment between flat plates of equal thickness (derivable using beam theory). The formulae in API 579-1 and BS 7910 are similar, differing only by a factor of (1-ν²). Table D.1 d) in BS 7910 gives the larger stress concentration factor.

Formulae for calculating the stress concentration factor due to axial misalignment are also given in DNV-OS-F101 App A. D205, Eqns. A.2-A.6 [4] and DNVGL-RP-0005 2.10.1, Eqn. 2.10.1 [20]. The stress concentration factor calculated using either of these formulae is lower than that calculated using Table 8.9 or Table D.1 d).

Stress Intensity Factor and Reference Stress Solutions

The stress intensity factor and reference stress solutions for surface and embedded circumferential flaws in a cylinder given in in API 579-1 and BS 7910, are listed in Table 3 and Table 4. The solutions for embedded flaws default to flat plate solutions.

The stress intensity factor and reference stress solutions are compared in Fig. 4 and Fig. 5. The load is a uniform membrane stress. The aspect ratio, a/c equals 0.25 and 1.0.

The stress intensity factor solutions for internal and external surface flaws are broadly similar. A larger difference is apparent for a/c = 0.25 and for an internal flaw. The solutions for an embedded flaw are also similar.

The reference stress solutions for internal and external surface flaws are significantly different. The solutions are different because the collapse modes are different. P.10.2 and P.10.4 are local collapse solutions. 9C.5.14 is a net section (global) collapse solution. The solutions for an embedded flaw are identical (and both are local collapse solutions).

Fig. 4 and Fig. 5 illustrate that, depending on the specific case, the use of the stress intensity factor and reference stress solutions in API 579-1 and BS 7910 will lead to similar results or significantly different results.

Note:

1.API 1104 and CSA Z662-15 give a stress intensity factor solution for a surface flaw, but not for an embedded flaw.

Fig. 4  Stress intensity factor solutions for circumferentially orientated flaws in API 579-1 and BS 7910

Notes:

1. The reference stress is calculated with Pb = 0.0 and km = 1.0.
2. 9C.5.14 (and 9C.5.13) is not directly comparable with either P.10.2 or P.10.4 because the former is a net section collapse solution and the latter are local collapse solutions.
3. 9C.5.19 and P.10.6 are identical.
4. API 1104 and CSA Z662-15 give a reference stress solution for a surface flaw, but not for an embedded flaw.

Fig. 5  Reference stress solutions for circumferentially orientated flaws in API 579-1 and BS 7910

Software

An ECA is commonly perceived to be a thing that is very complicated. Initial impressions of API 579-1 and BS 7910 will indeed be coloured by the fact that the number of pages runs to 1,320 and 493, respectively. Annex A of API Standard 1104 and Annex K of CSA Z662-15 are significantly shorter (22 and 18 pages, respectively), being specific rather than general documents and, consequently, easier to navigate. Nevertheless, the equations are still perhaps somewhat daunting. API 579-1 states that “The level or amount of education and experience … shall be commensurate with the complexity, rigor, requirements and significance of the overall assessment.” and that “The Engineer … shall be competent to perform the level of assessment required.” BS 7910 states that “the execution of its provisions will be entrusted to appropriately qualified and experienced people, for whose use it has been produced.”

An ECA can be something of a black box because of the use of software to conduct the calculations. Hand-calculations are possible, but not practical. Software removes the drudgery. Software could take the form of macros or scripts to conduct the calculations, or the automation of the assessment procedures, e.g. CrackWISE and IntegriWISE [16,17]. A black box obscures understanding. The ability to use software is not the same as competence. Competence is obtained through training, mentoring and experience [18].

The illustrative assessment, below, illustrates the effect of the choice of the Level or Option of the assessment, and assumptions made in the assessment with respect to constraint, misalignment, residual stresses, etc., and, so, irrespective of the incidental detail of how the calculations are conducted, partially lifts the lid on the black box.

An Illustrative Assessment

An assessment of the fracture resistance of a circumferential, external crack in a girth weld in cylinder subject to a tensile longitudinal load is presented. It is representative of a typical ECA of a girth weld in pipeline (albeit slightly simplified).

The diameter of the (hypothetical) pipe is 406.4 mm (16 inch) diameter and the wall thickness is 17.5 mm. The grade of the line pipe steel is Grade L450 (equivalent to API 5L X65). The specified minimum yield strength (SMYS) of Grade L450 is 450 N.mm^-2 and the specified minimum tensile strength (SMTS) is 535 N.mm^-2. The strain at the yield strength is taken to 0.5 percent (consistent with the definition of the yield strength in BS EN ISO 3183:2012 [19]). The elastic modulus, E, is 205,000 N.mm^-2 (from Table 3 in BS 7910) and Poisson’s ratio, ν, is 0.3. The material is assumed not to exhibit a yield discontinuity³.

The applied longitudinal stress is equal to 450 N.mm^-2 (consistent with Tiers 2 and 3 of the Guidelines, see below).

The girth weld is assumed to overmatch the parent metal in terms of both the yield strength and the tensile strength.

The fracture toughness of the girth weld is variously defined in terms of the crack tip opening displacement and the J-integral, in order to illustrate the effect of various assumptions⁴. The crack tip opening displacement determined using a single edge notch bend specimen is equal to 0.15 mm (consistent with Tier 3 of the Guidelines, see below). The J-integral determined using the same specimen is equal to 125 N.mm^-1. The J-integral determined using a single edge notch tension specimen is equal to 250 N.mm^-1.

The ECA is conducted in accordance with BS 7910:2013. The results of the calculations are presented in terms of the tolerable flaw size (flaw height versus flaw length). The results are compared with EPRG guidelines on the assessment of defects in transmission pipeline girth welds – Revision 2014, and also Annex A of API Standard 1104 and Annex K of CSA Z662-15.

The EPRG guidelines on the assessment of defects in transmission pipeline girth welds – Revision 2014

The EPRG guidelines on the assessment of defects in transmission pipeline girth welds – Revision 2014 gives defect acceptance limits, based on fitness-for-service, denoted Tier 2 and Tier 3. Tier 2 is based on curved wide plate tests. Tier 3 is based on curved wide plate and full-scale tests. The defect limits are semi-empirical.

The Guidelines are very straightforward, in comparison to API Standard 1104, etc. but a number of requirements must be satisfied in order to use the defect limits. The Guidelines require that the girth weld is overmatched. Tier 2 is applicable to applied strains less than or equal to 0.5 percent. Tier 3 is applicable to applied stresses less than or equal to the yield strength of the pipe. Tier 3 requires an average crack tip opening displacement (CTOD) greater than or equal to 0.15 mm (and a minimum value greater than or equal to 0.1 mm) at the minimum design temperature. The Guidelines are not applicable to situations in which significant fatigue loading is present.

Tier 2 permits a planar flaw with a height of 3 mm and a length up to seven times the wall thickness (7t), i.e. 3x122.5 mm. It is also permits shorter, deeper flaws. The defect limits from Tier 2 of the Guidelines are plotted in Fig. 8. Tier 3 permits a planar flaw with a height of 3 mm and (in this geometry) a length equal to approximately 15 percent of the circumference (approximately 3x190 mm).

API Standard 1104, Annex A

Annex A Alternative Acceptance Standards for Girth Welds in API Standard 1104 presents three Options for defining defect limits. Option 1 is the simplest. Defect limits are given for two values of crack tip opening displacement: greater than or equal to 0.25 mm (0.01 inch), and greater than or equal to 0.10 mm (0.004 inch) and less than 0.25 mm (0.01 inch). Option 3 is a reference out to alternative procedures (such as BS 7910).

Option 3 is intended for cases where the fatigue loading exceeds the limit specified in Annex A. Option 2 is a slightly simplified engineering critical assessment. It is based on a Failure Assessment Diagram (identical to Level 2A in BS 7910:2005). Option 1 was calculated using Option 2.

A plot of allowable imperfection height versus allowable imperfection length is given in Fig. 6. The Tier 2 defect limits from the Guidelines (see above) and Case Nos. IV, V, VI and VIII (see below) are also plotted. The allowable imperfection length includes a factor of safety of 1.5 on the calculated length. The allowable flaw size is smaller than Tier 2 and Case No. IV (although it allows flaws that are shorter and deeper than Tier 2). It is smaller than that calculated using Option 2 in Annex K of CSA Z662-15, because of the different reference stress solutions.

Option 2 is similar to API 579-1 or BS 7910. The stress intensity factor and reference stress solutions are different. The stress intensity factor calculated using the solution given in Annex A is lower than that calculated using Annex M of BS 7910 (see Fig. 4). The reference stress is higher than that calculated using Annex P (see Fig. 5). Misalignment and residual stresses are not directly considered. The effect of residual stress is notionally accounted for through the factor of safety on length (Clause A.2.5).

CSA Z662-2015, Annex K

Annex K Standards of acceptability for circumferential pipe butt welds based upon fracture mechanics principles in CSA Z662-15 presents two Options for defining defect limits. Option1 is based on the crack tip opening displacement design curve, as in PD 6493:1980. The allowable imperfection size is

the lesser of that to prevent failure by brittle fracture or plastic collapse (similar to Level 1 in BS 7910:2005). Option 2 is a slightly simplified engineering critical assessment. It is based on a Failure Assessment Diagram (identical to Level 2A in BS 7910:2005).

Option 2 in Annex K is identical to Option 2 in Annex A apart from the use of a different (less conservative) reference stress solution and a different (more conservative) set of formulae for converting from the J-integral to the crack tip opening displacement. The reference stress solution is similar to that given in Annex P of BS 7910 (both are based on the plastic collapse solution in Kastner et al., 1981 [21]).

A plot of allowable imperfection height versus allowable imperfection length is given in Fig. 7. The Tier 2 defect limits from the Guidelines (see above) and Case Nos. IV, V, VI and VIII (see below) are also plotted. The allowable imperfection length includes a factor of safety of 1.5 on the calculated length and the allowable imperfection height includes and an addend of 0.5 mm on the calculated height⁶. The allowable flaw size is similar to Tier 2 and slightly larger than Case No. IV (and it allows flaws that are shorter and deeper than Tier 2). It is larger than that calculated using Option 2 in Annex A of API Standard 1104, because of the different reference stress solutions.

Option 2 is similar to API 579-1 or BS 7910. The stress intensity factor solution is different. The stress intensity factor calculated using the solution given in Annex A is lower than that calculated using Annex M of BS 7910 (see Fig. 4). Misalignment and residual stresses are not directly considered.

The ECA (BS 7910:2013)

The illustrative assessment is of the fracture resistance of a circumferential, external surface flaw in cylinder subject to a tensile longitudinal load. Failure is the loss of containment. The ECA is conducted in accordance with BS 7910:2013. The Option 1 FAD is used. Homogeneous material properties are assumed. This is a conservative assumption provided that the girth weld overmatches the parent metal. The assessment is based on a single value of toughness, typically representing the lowest value from a set of three similar specimens. Lower bound tensile properties (the specified minimum yield and tensile strength) are used. The applied load is assumed to represent an upper bound. Partial safety factors of other factors of safety are not applied.

The stress intensity factor solution is M.7.3.4 External surface flaws oriented circumferentially (which defaults to M.4.1 Surface flaws in plates). The reference stress solution is P.10.4 External surface flaw in thin-walled pipe/cylinder under combined tension, bending and pressure⁷ .

The effect of the assumptions made in the assessment with respect to residual stresses, constraint, misalignment, etc. is illustrated through Case Nos. I-XI, below. The tolerable flaw sizes are plotted in Fig. 8. Acceptance limits would take into account the capabilities of the inspection method and other practical considerations.

Case No. I does not include the effects of the local stress concentration at the weld toe or residual stresses. Case Nos. II and III include the effect of residual stresses. Case No. III assumes some global relief of the residual stresses. Case No. IV includes the effects of the local stress concentration at the weld toe and residual stresses. The fracture toughness used in Case Nos. I-V is that determined using a single edge notch bend (SENB) specimen (a_o/W=0.5), i.e. a high constraint test specimen. In Nos. VI-X, it is determined using a single edge notch tension (SENT) specimen (a_o/W=0.5), i.e. a low constraint test specimen. The crack-tip constraint in a tension specimen is a closer approximation to that at a crack in a girth weld subject to a tensile load. In Case Nos. I-IV, the fracture toughness is expressed in terms of the crack tip opening displacement. In Case Nos. V-VIII, it is expressed in terms of the J-integral. In Case No. XI, the fracture toughness is expressed in terms of a resistance curve.

Case Nos. VII-X include the effect of the stress concentration due to misalignment.

Case Nos. III-VI and XI illustrate what would usually be considered good practice when assessing a girth weld in the as-welded condition (albeit with no misalignment).

1. A primary membrane stress equal to the yield strength of the parent metal, i.e. P_m = 450 N.mm^-2. (P_b = 0 N.mm^-2.) A fracture toughness determined using a single edge notch bend specimen, δ_mat = 0.15 mm.
2. A uniform tensile residual stress due to welding equal to the yield strength of the parent metal, i.e. Q_m = σ_Y. The tolerable flaw size decreases significantly. (K_r increases. L_r does not change.) Otherwise, as No. I.
3. Global relief of a uniform tensile residual stress due to welding is assumed as a result of the primary loading applied to the as-welded structure, as per Clause 7.1.8.2, i.e. Q_m = (1.4-σ_ref/σ_f)σ_Y. The tolerable flaw size increases significantly. (K_r decreases. L_r does not change.) Otherwise, as No. I.
4. Weld toe stress intensity magnification factor for a semi-elliptical surface crack at the weld toe (Annex M, Clause M.11.1.3 Solution based on 3D finite element modelling). The width of the weld cap (the attachment length) is 10 mm, i.e. L = 10 mm. The tolerable flaw size decreases slightly. (K_r increases. L_r does not change.) Otherwise, as No. III.
5. A fracture toughness determined using a single edge notch bend specimen, J_mat = 125 N.mm^-1. The tolerable flaw size increases slightly. (K_r decreases. L_r does not change.) Otherwise, as Nos. III and IV.
6. A fracture toughness determined using a single edge notch tension specimen, J_mat = 250 N.mm^-1. The tolerable flaw size increases significantly. (K_r decreases. L_r does not change.) Otherwise, as Nos. III and IV.
7. Stress magnification (concentration) factor due to axial misalignment (Annex D, Table D.1). Axial misalignment equal to 1.5 mm, i.e. e = 1.5 mm. The tolerable flaw size is zero. (K_r increases. L_r increases. L_r is greater than L_r,max.) Otherwise, as No. VI.
8. Stress magnification (concentration) factor due to axial misalignment (Annex D, Table D.1), but the local bending stress due to misalignment is classified as a secondary stress. The tolerable flaw size (relative to No. VI) decreases slightly. (K_r increases. L_r does not change. L_r is less than L_r,max.) Otherwise, as Nos. VI and VII.
9. Stress magnification (concentration) factor due to axial misalignment (Annex D, Table D.1), but the local bending stress due to misalignment is classified partially (15 percent) as a secondary stress. The tolerable flaw size (relative to No. VI) decreases. (K_r increases. L_r decreases. L_r is less than L_r,max.) Otherwise, as Nos. VI and VII.

I. A primary membrane stress equal to the yield strength of the parent metal. A fracture toughness determined using a single edge notch bend specimen, δ_mat = 0.15 mm.
II. A uniform tensile residual stress.
III. Global relief of a uniform tensile residual stress due to welding.
IV. Weld toe stress intensity magnification factor for a semi-elliptical surface crack at the weld toe.
V. A fracture toughness determined using a single edge notch bend specimen, J_mat = 125 N.mm^-1.
VI. A fracture toughness determined using a single edge notch tension specimen, J_mat = 250 N.mm^-1.
VII. Stress magnification (concentration) factor due to axial misalignment.
VIII. -IX. P_m(k_m ^-1) is classified as a secondary stress or partially as a secondary stress.
X. P_m(k_m ^-1) is classified as a primary stress, but apply Neuber’s Rule as per App. A D205 in DNV-OS-F101.
XI. A resistance curve (J-R) determined using single edge notch tension specimens.

Fig. 8  The tolerable flaw size calculated in accordance with BS 7910:2013, using the Option 1 approach without a ductile tearing assessment (Case Nos. I-X) and with a ductile tearing assessment (Case No. XI)

• Stress magnification (concentration) factor due to axial misalignment (Annex D, Table D.1) and apply Neuber’s Rule as per App. A D205 in DNV-OS-F101. The local bending stress due to misalignment is classified as a primary stress. The tolerable flaw size (relative to No. VI) decreases slightly. (K_r increases. L_r increases. L_r is less than L_r,max.) Otherwise, as Nos. VI and VII.

• A tearing resistance curve determined using single edge notch tension specimens. The initiation toughness (Δa = 0.2 mm) is 166.7 N.mm^-1, the toughness at maximum load (Δa = 0.3 mm) is 250 N.mm^-1, i.e. J0.2 = 166.7 N.mm^-1, J_m = 250 N.mm^-1, and the exponent of the curve, x, is 0.5. The tolerable flaw size changes slightly. (K_r decreases. L_r does not change.) Otherwise, as No. VI.

A circumferential, external surface crack-like flaw in a girth weld is described by Case No. IV. It includes the effect of the local stress concentration at the weld toe (the weld toe stress intensity magnification factor) and the effect of residual stresses. The effect of the local stress concentration at the weld toe is relatively small (Nos. III and IV). The effect of residual stresses is significant (Nos. I, II and III) and the effect of assuming some global relief is also significant (Case Nos. II and III). Residual stresses may be assumed to be uniform or non-uniform. The simplest approach is to assume that the residual stresses are uniform and then to assume some global relief of the residual stresses, as per Clause 7.1.8.2 (as in Case No. III). Annex Q in BS 7910 gives non-uniform residual stress profiles, but these require knowledge of the heat input during welding. (Annex 9D in API 579-1 similarly gives non-uniform residual stress profiles, but does not have an equivalent to Clause 7.1.8.2.) Case Nos. III and IV illustrate the importance of including the local stress concentration at the weld toe and residual stresses. It is also noteworthy that local stress concentration at the weld toe and residual stresses have no effect if the toughness is sufficiently high such that failure is controlled by plastic collapse (L_r is independent of M_km and M_kb or Q_m and Q_b). The effect of constraint is significant (Case Nos. IV and V), see below. The effect of axial misalignment is significant (Case Nos. VI, VII and VIII-X), see below.

Case No. VI describes a circumferential, external surface crack-like flaw in a girth weld with a fracture toughness determined using a low constraint specimen, as compared to Case Nos. IV and V with a fracture toughness determined using a high constraint, deep-notched bend specimen. Case Nos. IV and V are lower than Tier 2. Case No. VI is similar to Tier 2, but lower than Tier 3. Tier 2 of the Guidelines is based on wide plate test data. Tier 3 is based on wide plate and full-scale tests data. The Guidelines are semi-empirical. That Case No. VI is similar to Tier 2 is indicative that the different approaches are broadly aligned. Case No. VI is, in fact, still conservative because explicit consideration of over-matching and ductile tearing would tend to lead to a larger tolerable flaw size (albeit noting that the effect of misalignment is to reduce the tolerable flaw size). The use of Option 2 (or even Option 3) would also lead to a larger tolerable flaw size.

Crack Tip Opening Displacement, J-integral and Constraint

The effect of determining the fracture toughness in terms of the J-integral rather than the crack tip opening displacement is illustrated by Case Nos. IV and V. δ_mat can be converted into J_mat, as per Clause 7.1.4.6⁹, but the calculated value of J_mat will be lower than that determined directly from the test record because the conversion is a conservative approximation.

The effect of constraint is illustrated by Case No. V and VI. The single edge notch bend test specimen has a high level of constraint at the crack tip. A crack in a girth weld is typically subject to a lower level of constraint than that implied by a single edge notch bend. A pipeline is a relatively thin-wall structures subject to predominantly membrane loading. The single edge notch tension test specimen has a lower level of constraint at the crack tip. The fracture toughness determined using an SENT specimen will be higher than that determined using an SENB specimen. A specimen with a lower level of constraint can be used provided that the constraint at the crack-tip generated in the specimen is not less than that generated by the flaw in the component to be assessed.

The effect of defining the fracture toughness in terms of a single value, based on the first attainment of a maximum force plateau (maximum load), or of using tearing resistance data is illustrated in Case Nos. VI and XI. In Case No. VI, J_m = 250 N.mm^-1. In Case No. XI, J_0.2 = 166.7 N.mm^-1 and J_m = 250 N.mm^-1. The tolerable flaw size increases at shorter lengths and decreases slightly at longer lengths. The tolerable flaw size would increase, had, for example, J_0.2 = 250 N.mm^-1, i.e. initiation at or close to maximum load.

Axial Misalignment

The effect of axial misalignment is illustrated by Case Nos. VI, VII and VIII-X. Axial misalignment introduces an additional local bending stress (equal to P_m(k_m ^-1), where k_m is the stress concentration factor). k_m is calculated using Table D.1, d) Axial misalignment at girth welds in tubes, pipes, vessels …, with or without thickness changes. The effect of axial misalignment is significant. The local bending stress due to misalignment is considered to be a primary bending stress and so it contributes to both fracture and plastic collapse (i.e. K_r and L_r, respectively). It is the effect on L_r that is particularly significant. Case No. VII illustrates that the formulation of the reference stress solution is problematic (the effect is particularly pronounced in Case Nos. VII-X because P_m is equal to the yield strength). The notional effect of P_b on L_r is arguably too large.

Axial misalignment introduces an additional local bending stress (equal to P_m(k_m ^-1)). The default interpretation is that this additional bending stress is a primary stress, and so contributes to both fracture and plastic collapse (meaning that it is included in the calculation of both K_r and L_r). Case No. VII follows the default interpretation. Wei, 2010 [22] and Pisarski, 2011 [23] note that it might be more appropriate to treat this local bending stress as a secondary stress (meaning that it is only included in the calculation of K_r). In Case No. VIII, P_m(k_m ^-1) is classified as a secondary stress. In Case No. IX, 0.85xP_m(k_m ^-1) is classified as a secondary stress and 0.15x85xP_m(k_m ^-1) is classified as a primary stress. K_r increases. L_r does not change or increases slightly. The tolerable flaw size increases if the local bending stress is treated as a secondary stress (Case No. VIII) or partially as a secondary stress (Case No. IX). The effect is significant.

k_m is an elastic stress concentration. Appendix A of DNV-OS-F101 states that Neuber’s Rule may be applied. Neuber’s Rule is not included in either Section 7 of BS 7910 or Part 9 of API 579-1. The effect of applying Neuber’s Rule is to reduce the value of k_m. P_m(k_m ^-1) is classified as a primary stress, but km is smaller. K_r and L_r are calculated using the reduced value of k_m. (K_r is an elastic term, so it is questionable whether it is strictly appropriate to use the reduced value of k_m). The effect of applying Neuber’s Rule is illustrated by Case No. VIII. The effect is significant.

The issues illustrated by Case Nos. VII and VIII-X only arise if the primary membrane stress is large (near yield). A pipeline might be subject to large axial stresses and strains, but in a broader context it is an edge case. The default interpretation is Case No. VII. It is conservative. Case Nos. VIII-X highlight alternative interpretations that should be considered if the default interpretation is problematic.

what are the lessons to be learnt?

An engineering critical assessment (ECA) is more than simply conducting the calculations. The calculations are more than a fracture assessment. Software is simply a tool to be used to conduct the calculations. The illustrative assessment of the fracture resistance of a circumferential, external crack in a girth weld in cylinder subject to a tensile longitudinal load only illustrates part of the calculations. Calculations are moot if the data that is used is inappropriate or incorrect. Nevertheless, the assessment does illustrate some broader lessons.

API 579-1 and BS 7910 are intimidating documents. Annex A in API Standard 1104 and Annex K in CSA Z662-15 have taken all of the legwork out of navigating a route through either API 579-1 or BS 7910. In addition, API Standard 1104 and CSA Z662-15 (and BS 4515-1) present additional requirements regarding testing, qualification and approval of welding procedures and welders, inspection and testing of welds, etc. However, the simplicity of API Standard 1104 and CSA Z662-15 is notional. Option 2 in API Standard 1104 and CSA Z662-15 requires a similar level of calculation to an assessment conducted in accordance with either API 579-1 or BS 7910. The stress intensity factor and reference stress solutions are not significantly better than those in either API 579-1 or BS 7910 (see Fig. 4 and Fig. 5). Therefore, aside from obviating the need to navigate through the larger documents, the benefits of using Option 2 are open to question. The allowable imperfection size (the flaw acceptance criteria) are, at least in the case of the illustrative assessment, similar to or smaller than the tolerable flaw size calculated using BS 7910. The allowable imperfection size includes factors of safety, but that on length is at least partially there because of the simplified treatment of residual stresses.

Option 1 in Annex A of API Standard 1104 is, relatively speaking, very simple and straightforward to apply.

Tier 2 of the Guidelines is also very simple and straightforward to apply. Tier 2 will, for equivalent loading, etc., give larger tolerable flaw sizes than Option 1 in Annex A.

A simple and pragmatic approach would be to use Tier 2 of the Guidelines or Option 1 in Annex A of API Standard 1104 if and when all of the respective requirements are satisfied. Fatigue loading during installation or operation is the obvious load case that would prevent their use.

API 579-1 or BS 7910 should be used if and when the requirements of the Guidelines, etc. are not satisfied. A preference for either API 579-1 or BS 7910 is largely subjective and also an issue of familiarity. The stress intensity factor solutions in API 579-1 tend to have a wider range of applicability (and these solutions could be used instead of those in BS 7910). The treatment of uniform residual stresses in BS 7910 is arguably better than that in API 579-1. The reference stress solution for a surface flaw in BS 7910 is arguably more appropriate. The crack tip opening displacement, whether determined using SENB or SENT, is commonly used in Welding Procedure Qualification (largely for historical reasons). The crack tip opening displacement and the J-integral can both be determined independently from the test record, and both parameters have their uses. The J-integral determined from the test record (if available) should be used in an assessment (the use of the crack tip opening displacement would be slightly more conservative). Fracture toughness testing using SENB and SENT is standardised in BS 7448 [24-26] and BS 8571 [27], respectively¹⁰.

In the spirit of taking some of the legwork out of navigating a route through BS 7910:2013 (analogous to API Standard 1104 and CAS Z662-15), and only considering the calculations required for a fracture assessment, a few signposts to the relevant sections in BS 7910 are provided below:

7.1.4 Fracture toughness [the definition of K_mat]
(Equation 15) [J_mat]
7.3.2 The cut-off value of L_r [the definition of L_r,max]
(Equation 25)
7.3.3 Option 1 [the definition of the failure assessment line]
(Equations 26-28)
7.3.6 The fracture ratio, K_r [the definition of K_r]
(Equation 38 or 39)
7.3.7 The load ratio, L_r [the definition of L_r]
(Equation 40)
7.1.8.2 As-welded structures [residual stresses]
“… global relief of [uniform] residual stresses may be assumed as a result of the primary loading applied to the as-welded structure.” (Equations 24a & b)
Annex D Stress due to misalignment [k_m]
Table D.1d), or DNV-OS-F101 App A. D205, Eqs. A.2-A.6 or DNVGL-RP-0005 2.10.1, Equation 2.10.1
[DNV-OS-F101 App A. D205 “… The Neuber approach may be applied. …”]
Annex M Stress intensity factor solutions
internal and external surface flaws oriented circumferentially
M.7.3.2 & M.7.3.4 (M.4.1)
M.11.1
embedded flaws
M.7.3.6 (M.4.3)
Annex P Compendium of reference stress and limit load solutions for homogeneous and strength mis-matched structures
internal or external surface flaw in a thin-walled pipe/cylinder
P.10.2 & P.10.4 (Equation (P.23)
embedded flaws in a thin-walled pipe/cylinder
P.10.6 (P.6.3)
Annex R Determination of plasticity interaction effects with combined primary and secondary loading [a correction to account for plasticity interaction effects, in Equations 38 and 39]
R.2 Simplified procedure …

A Final Aside…

An ECA is undertaken during design in order to calculate the tolerable sizes of flaws in the girth welds. The flaw acceptance criteria should be informed by the results of the ECA, but not dictated by these results. The capabilities of the inspection method, etc., should be considered. An ECA should complement good quality workmanship.

The goal is to consistently fabricate high quality girth welds at a reasonable rate of production. It is not to produce the largest possible flaw acceptance criteria.

Conclusion

An engineering critical assessment (ECA) is more than calculations and it is more than the software used to conduct the calculations. A black box obscures understanding. An investigation of the effect of assumptions made in the assessment with respect to constraint, misalignment, residual stresses, etc., illustrates the effect of the various assumptions. Curiosity is a step towards understanding. Software is then a tool to remove drudgery rather than a black box.

API 579-1 and BS 7910 are intimidating documents. Annex A in API Standard 1104 and Annex K in CSA Z662-15 are less intimidating, but arguably, either API 579-1 or BS 7910 are preferable to Option 2 in either Annex A or Annex K.

Option 1 in Annex A Alternative Acceptance Standards for Girth Welds of API Standard 1104 or Tier 2 of the EPRG guidelines on the assessment of defects in transmission pipeline girth welds – Revision 2014 should be used if and when all of the respective requirements are satisfied.

API 579-1/ASME FFS-1 and BS 7910:2013 include compendia of stress intensity factor solutions and reference stress solutions. The solutions for circumferentially orientated surface and embedded flaws in a cylinder are different (apart from the reference stress solutions for embedded flaws). The differences are both small and large. An understanding of the reasons for the differences would reduce the scope for significant differences between assessments using the different standards.

The effect of the stress concentration due to misalignment on the reference stress (and limit load) warrants further investigation.

References

1. American Petroleum Institute, 2013. API Standard 1104 Welding of pipelines and related facilities. Twenty-First Edition, September 2013. Errata 3, July 2014. Addendum 2, May 2016. Washington, D.C., USA: API Publishing Services.
2. British Standards Institution, 2009. BS 4515-1:2009 Specification for welding of steel pipelines on land and offshore. Part 1 Carbon and carbon manganese steel pipelines. London, UK: British Standards Institution.
3. Canadian Standards Association, 2015. Z662-15 Oil and gas pipeline systems. Revised March 2016. Toronto, Ontario, Canada: CSA Group.
4. Det Norske Veritas, 2013. Offshore Standard DNV-OS-F101 Submarine pipeline systems. Høvik, Norway: Det Norske Veritas.
5. American Petroleum Institute, 2016. API 579-1/ASME FFS-1 2016, Fitness-For-Service. New York, N.Y., USA: American Society of Mechanical Engineers & Washington, D.C., USA: API Publishing Services.
6. British Standards Institution, 2015. BS 7910:2013+A1:2015, Incorporating Corrigenda Nos. 1 and 2 Guide on methods for assessing the acceptability of flaws in metallic structures. London, UK: British Standards Institution.
7. British Standards Institution, 2007. BS 7910:2005, Incorporating Amendment No. 1 Guide to methods for assessing the acceptability of flaws in metallic structures. London, UK: British Standards Institution.
8. Det Norske Veritas, 2006. Recommended Practice DNV-RP-F108 Fracture control for pipeline installation methods introducing cyclic plastic strain. Høvik, Norway: Det Norske Veritas.
9. Andrews, R.M., Denys, R.M., Knauf, G. and Zarea, M., 2015. EPRG guidelines on the assessment of defects in transmission pipeline girth welds – Revision 2014. In The Journal of Pipeline Engineering, 14(1): 9-21.
10. Hadley, I., Wiesner, C.S. and Maddox, S.J., 2000. PD 6493 becomes BS 7910; what's new in fracture and fatigue assessment? In J.B. Wintle, editor. Flaw Assessment in Pressure Equipment and Welded Structures. London, UK, 8 June 1999. London, UK: Institution of Mechanical Engineers.
11. Koçak, M., Webster, S., Janosch, J.J., Ainsworth, R.A. and Koers, R., 2008. FITNET Fitness-for-Service (FFS) Procedure. Vol. I. Geesthacht, Germany: GKSS Research Centre.
12. Koçak, M., Hadley, I., Szavai, S., Tkach, Y. and Taylor, N., 2008. FITNET Fitness-for-Service (FFS) Annex. Vol. II. Geesthacht, Germany: GKSS Research Centre.
13. R6 Panel, 2001. R6: Assessment of the integrity of structures containing defects. Revision 4. Gloucester, UK: British Energy Generation Ltd.
14. Larrosa, N.O. and Ainsworth, R.A., 2015. Comparisons of the solutions of common FFS standard procedures to benchmark problems. In Procedia Engineering, 130(2015): 1327-1342.
15. American Petroleum Institute, 2009. API 579-2/ASME FFS-2 2009, Fitness-For-Service Example Problem Manual. New York, N.Y., USA: American Society of Mechanical Engineers & Washington, D.C., USA: API Publishing Services.
16. < http://www.twisoftware.com/software/crackwise/> [Accessed 02 January 2017]
17. < http://www.twisoftware.com/software/integriwise/> [Accessed 02 January 2017]
18. Unger, M. and Hopkins, P., 2016. Training and Education: The great competency divide… Paper No. IPC2016-64500. In Proceedings of the 2016 11th International Pipeline Conference. IPC2016. Calgary, Alberta, Canada, 26-30 September 2016. New York, USA: American Society of Mechanical Engineers.
19. British Standards Institution, 2012. BS EN ISO 3183:2012 Petroleum and natural gas industries – Steel pipe for pipeline transportation systems. London, UK: British Standards Institution.
20. DNV GL, 2014. Recommended Practice DNVGL-RP-0005:2014-06 RP-C203: Fatigue design of offshore steel structures. Høvik, Norway: DNV GL.
21. Kastner, W. Roehrich, E., Schmitt, W. and Steinbuch, R., 1981. Critical crack sizes in ductile piping. In International Journal of Pressure Vessel and Piping, 9(3): 197-219.
22. Wei, L., 2010. The effects of classification of misalignment-induced stresses in engineering critical assessments of welded joints with some misalignment. Paper No. PVP2010-25335. In Proceedings of the ASME 2010 Pressure Vessels and Piping Division PVP Conference. PVP 2010. Bellevue, Washington, USA, 18-22 July 2010. New York, USA: American Society of Mechanical Engineers.
23. Pisarski, H., 2011. Assessment of Flaws in Pipe Girth Welds. In International Conference on Welding of High Strength Pipeline Steels. Araxá, Brazil, 28-30 November 2011. Araxá, Brazil: CBMM & Warrendale, USA: The Minerals, Metals & Materials Society.
24. British Standards Institution, 1991. BS 7448-1:1991 Fracture mechanics toughness tests. Method for determination of KIc, critical CTOD and critical J values of metallic materials. London, UK: British Standards Institution.
25. British Standards Institution, 1997. BS 7448-2:1997 Fracture mechanics toughness tests. Method for determination of KIc, critical CTOD and critical J values of welds in metallic materials. London, UK: British Standards Institution.
26. British Standards Institution, 1997. BS 7448-4:1997 Fracture mechanics toughness tests. Method for determination of fracture resistance curves and initiation values for stable crack extension in metallic materials. London, UK: British Standards Institution.
27. British Standards Institution, 2014. BS 8571:2014 Method of test for determination of fracture toughness in metallic materials using single edge notch tension (SENT) specimens. London, UK: British Standards Institution.
28. Hutchison, E.K., Moore, P.L. and Bath, W.P., 2015. SENT stable tearing crack path deviation and its influence on J. Paper No. PVP2015-45480. In Proceedings of the ASME 2015 Pressure Vessels and Piping Conference. PVP 2015. Boston, Massachusetts, USA, 19-23 July 2015. New York, USA: American Society of Mechanical Engineers.

Table 1   Assessment levels in API 579-1/ASME FFS-1 and options in BS 7910:2013

API 579-1/ASME FFS-1 2016	BS 7910:2013
Level 1 a simple assessment based on screening curves; limited to cracks and cracks-like defects in pressurised cylinders, spheres or flat plates away from all structural discontinuities	no comparable option
Level 2 a material non-specific FAD without an assessment of ductile tearing Level 3A as Level 2, but with partial safety factors Level 3D as Level 3A with an assessment of ductile tearing	Option 1 a material non-specific FAD; with or without an assessment of ductile tearing (with or without partial safety factors)
Level 3B a material specific FAD without an assessment of ductile tearing, with partial safety factors Level 3D as Level 3B with an assessment of ductile tearing	Option 2 a material specific FAD, requiring a uniaxial true stress-strain curve; with or without an assessment of ductile tearing (with or without partial safety factors)
Level 3C a material, geometry and loading specific FAD without an assessment of ductile tearing, with partial safety factors Level 3D as Level 3C with an assessment of ductile tearing	Option 3 a material, geometry and loading specific FAD, defined using finite element analysis; with or without an assessment of ductile tearing (with or without partial safety factors)
Level 3E the use of other recognised assessment procedures	no comparable option

Notes:

• The Level 2 FAD in API 579-1 is similar to, but not the same, as the Option 1 FAD in BS 7910:2013. It is identical to the Level 2A FAD in BS 7910:2005. The FADs used in Option 2 in Annex A of API Standard 1104 and in Option 2 in Annex K of CSA Z662-15 are also identical to the Level 2 FAD in API 579-1 and the Level 2A FAD in BS 7910:2005.
• The Level 2A FAD in BS 7910:2005 was taken from R6 Revision 3. The Option 1 FAD in in BS 7910:2013 is taken from R6 Revision 4 and FITNET.
• The Level 3B FAD in API 579-1 is identical to the Option 2 FAD in BS 7910:2013.

Table 2 Stress concentration due to misalignment in Part 8 and Annex D

API 579-1/ASME FFS-1 2016	BS 7910:2013
Table 8.9 Equations For The Ratio Of Induced Bending Stress To Applied Membrane Stress For The Circumferential Joints Of A Cylinder With Centerline Offset And Angular Misalignment Centerline – Circumferential Joint, Centerline Offset	Table D.1 Formulae for calculating the bending stress due to misalignment in butt joints d) Axial misalignment at girth welds in tubes, pipes, vessels and at seams in spheres, with or without thickness changes

Note:

• The formulae in API 579-1 and BS 7910 are similar, differing only by a factor of (1-ν²).
• Formula for calculating the stress concentration factor due to axial misalignment are also given in DNV-OS-F101 App A. D205, Eqs. A.2-A.6 and DNVGL-RP-0005 2.10.1, Equation 2.10.1.

Table 3 Stress intensity factor solutions in Annex 9C and Annex M

API 579-1/ASME FFS-1 2016	BS 7910:2013
9B.5.14 Cylinder – Surface Crack, Circumferential Direction – Semi-Elliptical Shape, Through-Wall Fourth Order Polynomial Stress Distribution with a Net Section Bending Stress (KCSCCE2)	M.7.3.2 Internal surface flaws oriented circumferentially M.7.3.4 External surface flaws oriented circumferentially M.4.1 Surface flaws in plates
9B.5.13 Cylinder – Surface Crack, Circumferential Direction – Semi-Elliptical Shape, Internal Pressure and Net-Section Axial Force (KCSCCE1)
9B.5.19 Cylinder – Embedded Crack, Circumferential Direction – Elliptical Shape, Through-Wall Fourth Order Polynomial Stress Distribution (KCECCE) 9B.3.9 Plate – Embedded Crack, Elliptical Shape, Through-Wall Fourth Order Polynomial Stress Distribution (KPECE2)	M.7.3.6 Embedded flaws M.4.3 Embedded flaws in plates
9B.3.8 Plate – Embedded Crack, Elliptical Shape, Through-Wall Membrane and Bending Stress (KPECE1)

Note:

• 9B.5.13 and 9B.5.14 are identical given a uniform membrane stress.
• The stress intensity factor solutions for a surface flaw in API Standard 1104 and CSA Z662-15 are the same, and different to those in either API 579-1 or BS 7910.

Table 4 Reference stress solutions in Annex 9D and Annex P

API 579-1/ASME FFS-1 2016	BS 7910:2013
9C.5.14 Cylinder – Surface Crack, Circumferential Direction – Semi-Elliptical Shape, Through-Wall Fourth Order Polynomial Stress Distribution with a Net Section Bending Stress (RCSCCE2) [net section]	P.10.2 Internal surface flaw in thin-walled pipe/cylinder under combined tension, bending and pressure [local] P.10.4 External surface flaw in thin-walled pipe/cylinder under combined tension, bending and pressure [local]
9C.5.13 Cylinder – Surface Crack, Circumferential Direction – Semi-Elliptical Shape, Internal Pressure and Net-Section Axial Force (RCSCCE1) [net section]
9C.5.19 Cylinder – Embedded Crack, Circumferential Direction – Elliptical Shape, Through-Wall Fourth Order Polynomial Stress Distribution (RCECCE) [local] 9C.3.7 Plate – Embedded Crack, Infinite Length, Through-Wall Fourth Order Polynomial Stress Distribution (RPECL) [local]	P.10.6 Embedded flaws in thin-walled pipe/cylinder [local] P.6.3 Embedded flaws in plates under combined tension and bending [local]

Note:

• 9C.5.13 and 9C.5.14 are identical if the net section bending moment (M) is equal to zero.
• The reference stress solution for a surface flaw in API Standard 1104 is different to that in either API 579-1 or BS 7910. The reference stress solution for a surface flaw in CSA Z662-15 the same as that in BS 7910.