Recommendations on Transmission and Founded 1863 Distribution Practice Royal Charter 1929 IGE/TD/12 Edition 2 Patron Communication 1681 Her Majesty The Queen Pipework stress analysis for gas industry plant Price Code: T6 The Institution of Gas Engineers and Managers 12 York Gate London NWi 4QG Tel: 020 7487 0650 Fax:020 7224 4762 Email: general@igem.org.ukIGE/TD/12 EDITION 2 COMMUNICATION 1681 2003 The following amendments (January 2005) apply to all copies of IGE/TD/12 Edition 2 published in 2003. Inner Front Cover Footer each page ‘Sub- Section 1.10 Clause 2.1 Clause 5.2.4, Clause 8.5.7.1 ADL A2.2.2 A2.3 (Not essential) Amend to: Price Code: T6 The Institution of Gas Engineers and Managers Charnwood Wing Holywell Park Ashby Road Loughborough Leicester LE11 3GH Tel: 01509 282728 Fax: 01509 283110 Email: general@igem.o! (Not essential) Amend to: IGEM, Charnwood Wing, Holywell Park, Ashby Road, Loughborough, Leicestershire, LE11 3GH ‘Website: www.igem.org.uk, (Not essential) Address amended to: IGEM, Charnwood Wing, Holywell Park, Ashby Road, Loughborough, Leicestershire, LE11 3GH 4® Bullet. Delete “IGE/GM/1”. Substitute: IGE/GM/8 Delete formula entirely. Substitute: PDX. fon = 0S In 1* equation. Delete neg. Substitute: neq 2° Bullet. Delete “Distribution mains”. Delete Note. ‘Substitute: Steel and PE pipelines for gas distribution Delete 1* bullet. Substitute: © IGE/GM/8 on gas meter installations ‘Note: Publication anticipated in 2005. 1% Bullet. Add a note: Note: Now obsolete, Superseded by BS EN 13480, metalic industry piping. monaA331 A3.3.2 A3.3.3.1 A3.3.3.2 A4.10.2 A4,10.4 AS2.1 Inner rear cover Delete Equation for Sq. Substitute: Delete all three equations. Substitute: Sy Ba 1, 22M Oa aT p.*-,] a * eta") rd.?-4?] Delete 4" line (1% equation for S,). Substitute: & 1 BBs 5, 22MePe xo,*-0¥] * "| Delete 6" line (2"! equation for S,). Substitute: 32M, 0, yj 4F wb-0'] * * peo] Delete 7" line (equation for S,). Substitute: 16 Dy Me Smt * {{p.*-0, +h peal Jl Table 13. 1* column, 2" row. Inse 11.0 Table 15. Note 1. Delete entirely, Subs! te: IL: 2.0.5,fF, Ty . then re can be taken as the radius to the centre Of Ts and Ty = Ts, otherwise Fe fm which isthe radius to the centre of Ts.and T= T's Pg 80. Table. 5™ column, 1* row. Amend “2.0” to: 1.0 (Not essential) Delete address. Substitute: IGEM, Charnwood Wing, Holywell Park, Ashby Road, Loughborough, Leicestershire, LE11 3GHRear (Not essential) Delete “12 York Gate, London, NW 46. cover Substitute: IGEM, Chamwood Wing, Holywell Park, Ashby Road, Loughborough, Leicestershire, LE11 3GH Rear (Not essential) Delete list entirely, Substitute: cover Item No. Title Date of Comm. Price Edition No. Code IGE/TD/1 Steel pipelines for high pressure Edition 4 gas transmission 2001 1670 T14 IGE/TD/1 Handling, transport and storage of Edition 4 steel pipe, bends and fittings 2001 1671 T2 Supplement 1 IGE/TD/3 Steel and PE pipelines for gas Edition 4 distribution 2003 1677 T10 IGE/TD/3 Handling, transport and storage of Edition 4 PE pipe and fittings 2003 1682 3 Supplement 1 IGE/TD/4 Gas Services 1994 1562 3 Edition 3 IGE/TD/12 Pipework stress analysis for gas Edition 2 industry plant 2003 1681 16 IGE/TD/13_ Pressure regulating installations for transmission and distribution 2001 1672 ™ IGE/TD/15 Services and metering installations for a gas flow not exceeding 6 m? h” at supply MOP exceeding 75 mbar but not exceeding 2 bar 2002 1679 3 IGE/TD/101 Adoption of pipe systems by a GT = management of UIP activities 2002 1674 Tu IGE/TD/1 Combined Amendments 2005 - - Edition 4 IGE/TD/3 Edition 4 IGE/TD/12 Edition 2 IGE/TD/13 END OF AMENDMENTS TO IGE/TD/12 EDITION 2. ronanIGE TO/12 Esition 2 CONTENTS SECTION PAGE 1 Introduction i 2 Scope 3 3 ‘Competency and quality assurance 5 23a ‘Competency 5 32 Quality assurance 5 © 3.2.41 General 5 + 3.2.2 Co-ordination 5 + 32.3. Software 5 + 3.2.4 Analysis 5 + 3.2.5 Review of analysis 5 + 3.2.6 Construction 5 + 32.7 Design appraisal 5 4 Materials 6 e4a General 6 242 Elastic modulus (E) 6 243 Coefficient of thermal expansion (: 6 ° 44 Poisson's ratio (v) 7 245 Proof stress (Re) 7 2 46 Fatigue 7 5 Load and restraint conditions 8 5a General 8 5.2 Pressure 8 e 53 Temperature 8 + 5.3.1 Design temperatures 8 + 5.3.2 Tie-in temperature 8 + 5.3.3 Ambient temperature 8 + 5.3.4 Thermal expansion and contraction 9 254 Weight effects 8 55 Construction loads 9 56 Occasional, exceptional and other loads 9 = 5.6.1 | Shock 8 + 5.6.2 Wind 9 + 5.6.3. Discharge reactions 9 + 5.6.4 Vibration 3 + 5.6.5 Earthquake 8 + 5.6.6 Traffic 10 5.7 ‘Support, anchors and applied displacements 10 58 Soil 10 6 Performance u e64 General ut 2 62 Internal pressure i (© IGEM , 12 York Gate, London, NWI 4QG. Website: wuw-igem.org.ukIGE 70/12 Eaton 2 2 63 Normal sustained loads ° 64 Abnormal sustained loads * 65 Hydrostatic test + 66 Shakedown ° 67 Fatigue 7 Wall thickness and selection of pipework components ea General 272 Straight pipe ° 73 Fittings 274 Thickness allowances, 8 Stress and fatigue analysis 81 General procedure * 82 ‘Scope of pipework for analysis, * 83 Pipework modelling = 83.1 General + 8.3.2 Longitudinal pressure strain + 8.3.3. Thickness for analysis + 8.3.4 Estimation of soil restraint o 84 Load case table + 8.4.1 Compilation + 84.2 Use * 85 Stress analysis procedure and acceptance limits © 85.1 General + 8.5.2 Analysis of sustained conditions + 8.5.3 Acceptance limits for sustained conditions + 8.5.4 Analysis of shakedown conditions = 8.5.5 Acceptance limits for shakedown conditions + 8.5.6 Analysis of cyclic conditions * 8.5.7 Acceptance limits for cyclic conditions * 36 Reactions + 8.6.1 Supports = 8.6.2 Loads on connected equipment + 8.6.3 Strain-sensitive equipment * 387 Displacements + 38 Buckling 9 Resolving stress exceptions eon General © 92 New build © 9.2.1 Wall thickness + 9.2.2 Forged tees to BS 1640 + 9.2.3. Soil restraint = 9.2.4 Supports + 9.2.5 Branch flexibility modelling + 9.2.6 — Design-by-analysis (DBA) 293 Existing installations 9.3.1 Design conditions = 9.3.2 Operating history 9.3.3 Temperature conditions © IGEM , 12 York Gate, London, NWI 4Q6. Website: wnw.igem.org.uk Fea it 12 12 12 14 14 14 15 15 16IGE TO/12 Edition 2 + 9.3.4 Site modifications 10 Verification © 10.1 General © 10.2 Operating conditions © 10.3. Other design parameters and assumptions # 10.4 Action to be taken u Documentation © 111 Stress analysis records + 11.1.1 Commissioning and operational records 2) 11:12. Situations where analysis is not required © 112 Stress analysis report + 11.21 General = 11.2.2. Design brief + 11.2.3. Register + 11.2.4 Stress model drawing . Load case table . Soll information * 11.2.7 Isometric plot + 11.2.8 Geometric and material properties + 11.2.9 Stress exceptions + 11.2.10 Computer program input/output data + 11.2.1 Conclusions and recommendations APPENDIX 1 Glossary, acronyms and notation © ALI —— Glossary and acronyms © AL2 Notation 2 References © A241 Legislation ©9022 IGEM + A.2.2.1 Recommendations on Transmission and Distribution Practice + A2.2.2 Gas Measurement Procedures + A2.2.3 Gas Legislation Guidance A230 BSE © A24 Overseas © -A25 Miscellaneous 3 Stresses © 3.1 Notation © 3.2 Description ¢ 3.3. Individual stress components + A3.3.1. Straight pipe, wall thickness transitions, concentric and eccentric reducers + A3.3.2 Bends, mitre bends and welding elbows A3.3.3 Weldolets, long weld neck flanges, weldoflanges, sweepolets and tees © IGEM , 12 York Gate, London, NW1 4Q6. Website: www.igem.org.uk 26 27 27 27 27 27 28 31 31 34 36 38 38 38 40 40 40 40IGE TD/12 Edition 2 4 Stress concentration factors (SCFs) © A41 Scope @ 4.2 Dimensions © A4.3 Notation © A44 Straight pipe © A4.5 Wall thickness transition © A46 Concentric and eccentric reducer © A4.7 Bends and welding elbows © A48 Closely spaced mitre bend © A4.9 Single or widely spaced mitre bend © A4.10 — Set-on welded branch connections © A4.10.1 Weldolet ‘+ A4.10.2 Long weld neck flange = A4.10.3 Weldoflange + A4.10.4 Fabricated tee © A4.11 Other branch connections A4.11.1 Weldolets ‘4.11.2 Long weld neck flange 4.11.3 Weldoflange Ad.11.4 Fabricated tee A4.11.5 Forged tee to BS 1640 A4.11.6 Drawn/extruded and forged tee not to BS 1640 A4.11.7 Sweepolet & welded-in contour insert + A4.11.8 Full encirclement tee A4.11.9 Fabricated tee with pad or saddle 5 Fatigue analysis 6 Design-by-analysis (DBA) method 7 Example load case table 8 Worked example TABLES 1 Minimum nominal wall thickness of steel pipe 2 Design factor (f) 3 Maximum permitted sustained stresses FOr Se, Saati, Sssbay vonnrs Susi 4 ‘Shakedown factor, Kso 5 Fatigue analysis thresholds for carbon and austenitic steels 6 Straight pipe SCFs 7 Wall thickness transition SCFs 8 Concentric and eccentric reducer SCFs 9 Bend and welding elbow SCFs © IGEM , 12 York Gate, London, NWI 4QG. Website: www.iger.org.uk 73 78 79 14 45 20 20 24 46 47 48 49IGE TO/12 Féltion 2 10 iL 12 13 14 15 16 7 18 19 20 21 22 23 24 25 26 27 28 FIGURE Closely spaced mitre bend SCFs Single or widely spaced mitre bend SCFs Weldolet SCFs Long weld neck flange SCFS Weldoflange SCFs Fabricated tee SCFs Weldolet (others) SCFs Long weld neck flange (others) SCFs Weldoflange (others) SCFs Fabricated tee SCFs Forged tee to BS 1640 SCFs Drewn/extruded and forged tee not to BS 1640 SCFs ‘Sweepolet & welded-in contour insert SCFs Full encirclement tee SCFs Fabricated tee with pad or saddle SCFs Fatigue classes Equation constants for S-N fatigue design curves Correction factors for Se and Nr Example load case table for shakedown, normal and abnormal sustained and fatigue Adjacent branch limitations of set-on branch connections S-N Fatigue curves for E = 209 x 10° N/mm? Basic nomenclature of set-on branch connections © IGEM , 12 York Gate, Landon, NWI 4Q6. Website: ww.igem.org.uk 50 51 53 54 55 56 59 60 61 62 63 64 65 66 67 69 70 78 52 69 72IGE TD/12 Edition 2 SECTION 1 : INTRODUCTION aa 12 13 Ld 15 1.6 17 18 @IGEM , 12 York Gate, Landon, NWI 4QG. Website: In 1975, the Institution of Gas Engineers (IGE) published IGE/TD/9 for offtakes and pressure regulating installations (PRIs) operating at pressures exceeding 7 bar. Within IGE/TD/9, Section 9 and the associated Appendix 4 on pipework design and stressing outlined methods for the calculation of stresses within a piping system. IGE/TD/9 has been superseded by IGE/TD/13. In 1985, IGE published IGE/TD/12 (Edition 1), extending the scope of these methods to cover a wider application than that given in IGE/TD/9. Amendments to IGE/TD/12 (Edition 1) were issued in 1990. ‘This edition of IGE/TD/12, which supersedes Edition 1 (which is now obsolete), has been drafted by a Panel appointed by the Institution of Gas Engineers and Managers’ (IGEM's) Gas Transportation Committee, subsequently approved by that Committee and published by the authority of the Council of IGEM. It reflects improvements and developments through calculation and analysis. The aim of these Recommendations is to ensure that piping systems are designed to withstand the stresses within them and to ensure that forces and moments applied to connected equipment and support structures are considered. This edition provides a revised content and improved structure compared to Edition 1, Sections on stress calculation equations, stress concentration factors {SCFs) and fatigue are now presented in appendices. Sections on pressure testing and construction have been excluded since these aspects are more appropriate to IGE/TD/13. A worked example using the guidance on stresses, SCFs and fatigue is now included. This provides greater clarity on the sustained, shakedown and fatigue assessments. Note: Although IGE/TO/13 covers only PRIs, it may be deemed appropriate to use it for ‘construction and testing of other installations. Although much of the technical content of Edition 1 has been retained, there are a number of important changes: © design for creep conditions is no longer considered. The revised scope is controlled by a limit placed on the maximum design temperature © pipework stress analysis Is now based on the use of a matching wall thickness for both flexibility and stress calculations. This reflects common practice in other codes © the fatigue assessment now adopts the weld classification system and material S-N curves of PD 5500 and BS 7608. ‘A new section has been included on resolving code stress exceptions. This includes guidance on branch flexibility modelling and on “design-by-analysis” (DBA) using three-dimensional finite element (FE) modelling. Materials and methods of analysis are subject to continuous evolution and it Is intended to review these Recommendations from time to time. Regard will be paid to the publications and findings of appropriate organisations with which IGEM maintains liaison. These Recommendations reflect the need to ensure adequate reliability and continuity of supply at pressures that are safe for the downstream system and equipment. These Recommendations do not attempt to make the use of any method or specification obligatory against the judgement of the responsible engineer. Where new and better techniques are developed and proved, they should be 1 .gem.org ukIGE TD/12 Edition 2 19 adopted without waiting for modification of these Recommendations. Amendments to these Recommendations will be issued when necessary and their publication will be announced in the Journal of the Institution and other publications as appropriate. The primary responsibility for compliance with legal duties rests with the employer. The fact thet certain employees, for example “responsible engineers”, are allowed to exercise their professional judgement does not allow employers to abrogate their primary responsibilities. Employers must: (a) Have done everything to ensure, so far as is reasonably practicable, that there are no better protective measures that can be taken other than relying on the exercise of professional judgement by “responsible engineers”. (b) Have done everything to ensure, so far as is reasonably practicable, that “responsible engineers” have the skills, training, experience and personal qualities necessary for the proper exercise of professional judgement. (©) Have systems and procedures in place to ensure that the exercise of Professional judgement by “responsible engineers” is subject to appropriate manitoring and review. (d) Not require “responsible engineers” to undertake tasks which would necessitate the exercise of professional judgement that is beyond their competence. There should be written procedures defining the extent to which “responsible engineers” can exercise their judgement. when “responsible engineers” are asked to undertake tasks that deviate from this, they should refer the matter for higher review. Note: The responsible engineer is @ suitably quetifed, competent and experienced engineer or 3 suitably qualified, competent and experienced person acting under his or her supervision, ‘appointed to be responsible for the application ofall or part of these Recommendations. Requests for interpretation of these Recommendations in relation to matters within their scope, but not precisely covered by the current text, may be addressed to Technical Services, IGEM, 12 York Gate, London, NW1 4QG, em: general@igem.org.uk, and will be submitted to the relevant committee for consideration and advice, but in the context that the final responsibility is that of the engineer concerned. If any advice is given by or on behalf of IGEM, this does not imply acceptance of any liability for the consequences and does not relieve the responsible engineer of any of his or her obligations. 2 © IGEM,, 12 York Gete, London, NW 4QG. Website: www.igem.ora.ukIGE TO/12 Edition 2 SECTION 2 : SCOPE 24 2.2 2.3 24 25 2.6 These Recammendations cover the stress analysis and calculation of loadings for welded and flanged pipework used in gas industry installations on pipelines. In general, an installation on a pipeline comprises one or more of the following: ‘© compressor station line valve branch connection meter installation (see also IGE/GM/1 or IGE/GM/4, as appropriate) multi-junction station with or without in-line metering and which may include pig traps and/or line valves overhead pipe crossing pigging station PRI controlling gas flow and/or pressure from one system to another eoee terminal other offtakes. ee eee ‘These Recommendations apply to pipework (as outlined in Sub-Section 2.1) on. land, including water crossings. For pipework offshore, additional or alternative guidance may be required. However, many of the recommendations will remain valid. Note: Offshore pipework Is on the seaward side of the fow water mark ar special boundaries drawn at bays and estuaries. These Recommendations are entirely dependent upon the application of other relevant codes and standards, appropriate to the system being designed, for example IGE/TD/1, IGE/TD/13, etc., to ensure overall plant integrity. Note: The information covering materials (pressure testing and construction) Is included to quality the stress analysis recommendations and dram attention to important areas of Zonsieration ‘These Recommendations are entirely dependent upon the correct selection of materials compatible to the appropriate duty. These Recommendations are not intended for general application to pipelines, which are covered by IGE/TD/1, IGE/TD/3 and IGE/TD/4 respectively. Where larger diameter pipework is above ground for a significant distance, these Recornmendations may be relevant. These Recommendations cover stress analysis for pipework carrying gas at a maximum operating pressure (MOP) not exceeding 100 bar and at a temperature between -25°C and 120°C inclusive. Note: Specifically, these Recommendations apply to the range of conditions expected to be ‘encountered In the types of plant specified in Sub-Section 2.5 These Recommendations cover pipework handling 1 and 2% family gases as defined in BS 1179, odorised or unodorised. ‘These Recommendations apply only to carbon and austenitic steels whose permanent elongation at tensile fracture at room temperature is greater than 14% when measured using the method outlined In BS EN 10002 Part 5. Note 4: The use of pie manufactured from non-metallic, cast metallic or other brittle material is rot within the scope of these Recommendations. Note 2: Materiis operating above thelr creep threshold are not covered. 3 © IGEM , 12 York Gate, London, NW1 4QG. Website: www.igem.org.ukIGE 70/42 Esition 2 27 ‘These Recommendations apply to pipework containing gas at a positive gauge pressure, i.e. internal pressure not less than external pressure. 28 Pipes with a diameter to wall thickness ratio (D/t) of less than 10 require special consideration. For pipes of D/t greater than 96, the possibility of buckling of the pipework needs to be considered. 29 All pressures are gauge pressures unless otherwise stated, 4 © IGEM , 12 York Gate, London, NW1 4QG. Website: wwn.igem.org.ukIGE TO/12 Eaton 2 SECTION 3 : COMPETENCY AND QUALITY ASSURANCE 34 3.2 3.24 3.2.2 3.2.3 3.24 3.2.5 3.2.6 3.2.7 COMPETENCY Any person engaged in pipework stress analysis should be competent to carry out such work. This may be achieved by an appropriate combination of education, training and practical experience. The analyst should have documented proof of competence. QUALITY ASSURANCE General A suitable Quality Management System to BS EN ISO 9000 should be in piace for any design activity. Note: Guidetines far Quality Plans are provided in BS 150 10005. Co-ordination The analysis (see clause 3.2.4) should be carried out by, or under the supervision of, a responsible engineer who can, by virtue of status, ensure that effective liaison is established and maintained between the various functions which contribute to the design and construction of a plant, including procurement of materials. Software ‘Software used to perform analyses should be checked and verified. Analysis Any model used in the analysis of pipework should be a reasonable representation of the design for construction and any existing plant. ‘The model should represent pipework geometry, material properties, loading and restraint conditions. Where assumptions or approximations are required in the model, details and-their justification should be documented. Reference should be made to Section 8. Review of the analysis A review of the analysis should be carried out by an independent analyst. The scope of the review should include a check of load cases applied, boundary conditions, pipework geometry, material properties and compatibility with the proposed design, including any existing pipework, etc. Construction Arrangements should be made to ensure that materials and workmanship are in accordance with the approved construction design. Reference should be made to Section 10 if variations from the construction design exist. Design appraisal The design for construction and supporting analysis should be subject to a design appraisal, for example as outlined in IGE/GL/5. 5 (© IGEM , 12 York Gate, London, NW1 4QG, Website: wivwr.igem.org.ukIGE TO/12 Eeition 2 SECTION 4 : MATERIALS at ait 412 443 414 41.5 44.6 4.1.7 4.18 42 43 GENERAL, Materials used in pressure-containing pipework should conform to nationally listed specifications covering chemical, physical and mechanical properties, method and process of manufacture, heat treatment and quality control applied. Materials should be selected having regard to the process involved and be resistant to any adverse chemical effects of service conditions. ‘Note: An example, In this respect, is the phenomenon of stress corrasion cracking (SCC). Expert ‘advice is recommended when there isa risk of SCC occurring. For carbon or low alloy steel pipework, either the specified minimum yield strength (SMYS), or the 0.2% proof stress, should not be greater than 85% ultimate tensile strength (UTS), throughout the temperature range for which the pipework will be used. For austenitic stainless steel pipework, the 1.0% proof stress (Rp) should not be greater than 60% UTS at room temperature. Materials should have adequate fracture toughness properties throughout the design temperature range. Note: Detailed guidance is given in PD $500 and ASHE 831.3. Consideration should be given to the requirements for weldability and fracture toughness, as well as stress, when deciding the suitable pipework material, particularly where low temperatures are to be encountered in service. The weldability of pipe and fittings should be confirmed and a welding procedure proved. Note: Detaited guidance ls given in Reference A2.5.8, ASME B31.3 and BS 4515. Any limitation an the use of materials should be taken inte account. Note: Ranges of materials are tsted In piping syste codes such as ASME IIT, ASME 831.3 and BS 806. The quoted values of material properties mey need to be adjusted to eliminate {actors speaifc to the selected cade in order to derive values which are necessary for these Recommendations. Gaskets and welding materials should comply with _nationally-listed specifications and be entirely compatible with the other elements comprising the pipework system. Note: This section does not address gaskets and walting materials. ELASTIC MODULUS (£) E should be taken as the value for the material at 10°C or the value appropriate to the lowest temperature of operation below 10°C. COEFFICIENT OF THERMAL EXPANSION («) x will vary with temperature. For the purposes of determining forces and moments for assessment using these Recommendations, a single value may be taken. 6 © IGEM , 12 York Gate, London, NWT 4QG. Website: www.igem.org.ukIGE TD/A2 Edltion 2 44 POISSON'S RATIO (v) v should be considered in determining forces and moments for assessment using these Recommendations. Note: For carbon or low alioy steel, v may be taken as 0.3. 45 PROOF STRESS (Ry) R, is the stress at which the non-proportional elongation is equal to the specified percentage of the original geuge length. Reference should be made to 8S EN 1002-1, 4.6 FATIGUE Reference should be made to Sub-Section 6.7 and Appendix 5 for information on fatigue properties of materials. 7 © IGEM, 12 York Gate, London, NWI 4QG. Website: wiw.igem.org.ukIGE TO/12 Edition 2 SECTION 5 : LOAD AND RESTRAINT CONDITIONS 5a 5.2 5.21 5.2.2 5.2.3 53 5.3.1 5.3.2 5.3.3 GENERAL Pipework should be designed for the most severe conditions of coincident pressure and temperature as specified in the design brief, Notes This section defines the pressure, temperature and various forces applicable to the desian of pipework systems. PRESSURE The following formula for determining minimum wall thickness should be used: PDX 205 tein rrinimum wall thickness of pipe (mm) design pressure (bar) ‘outside diameter of pipe (mm) ‘design factor determined from Table 2 Spectied minimum yield strength (N/mm) ALE" and Is an overpressure factor. one po ‘The design pressure should be the target MOP. The maximum incidental pressure (MIP) should include overpressure above the design pressure and should be applied to shakedown analysis and normal sustained analysis. For fatigue cyclic loading where stress ranges are evaluated between two load cases at different pressures, the overpressure factor, X, should be applied only in the load case which has the higher pressure value (see A3.2). TEMPERATURE Design temperatures Design temperatures should take account of any foreseeable fault condition or temperature reduction due to de-pressurisation of the pipework, Ties temperature Tie-in temperature should be taken as the average pipework metal temperature at the time of final attachment of the sub-assembly. ote: This isthe reference tempersture used for thermal loacing assessments. Ambient temperature Due account should be taken of the degree of exposure of pipework to ambient temperature. Where pipework is not protected from the sun, an estimate should be made of the metal temperature due to solar gain, This temperature and the maximum. gas temperature should be considered under clause 5.3.1 and the higher value should be used in the design. Where pipework is not protected from ambient loss, the lower temperature considered under clause 5.3.1 should be the minimum’ ambient air temperature that is likely to be sustained or the gas temperature, whichever is the lower. 8 © IGEM , 12 York Gate, London, NW 4QG. Website: www.igem.org.ukIGE TO/A2 Edition 2 9.3.4 54 5.5 5.6 5.6.1 5.6.2 5.63 5.6.4 5.6.5 ‘Thermal expansion and contraction The following thermal effects should be taken into account: ‘© thermal loads due to restraints Note! These loads cansist of forces and moments that arise when pipework is prevented Tram free thermal expansion or contraction because of restraints. * loading due to differences in thermal expansion characteristics Note: These loads result from differences in thermal expansion where materials with Aifferent thermal expansion coefficients are Joined, for example in bi-metatic piping Uihere pipe is lined or coated with a metal of afferent properties. WEIGHT EFFECTS The following weight effects should be taken into account: © live loads Note: Live loads comprise the weight of the gas transported, ar the medium used for test or ‘lesning, in adclivon to snow and ie loads in facalties where such conditions exist. It may also be necessary to take into account the effect of trafic on buried pipework. © dead loads Note: Dead loads comprise the weight ofthe pipework, Including valves, strainers, insulation ‘and other superimposed permanent loads supported by the pipework, including sof! loading in areas nhere the buried pinenork has poor bed support (which may occur in the area of buried supports. CONSTRUCTION LOADS Where there is a likelihood of significant construction stresses, these should be assessed accordingly. Where pre-strain due to cold pull is stipulated in the design, this should be incorporated into the analysis. OCCASIONAL, EXCEPTIONAL AND OTHER LOADS Shock Effects caused by sudden changes in external or internal conditions, for example hydraulic shock, reductions in pressure and surge conditions, should be taken into account in the analysis. Wind ‘The effect of wind loading should be taken into account in the analysis of pipework exposed to the wind. The loading should be calculated in accordance with an appropriate code such as BS 6399 Part 2. Discharge reactions Reaction forces due to discharge of the contained fluid, including temperature changes that may occur, should be taken into account in the analysis. Vibration Consideration should be given to the effects of vibration, the analysis of which is beyond the scope of these Recommendations. Expert advice should be sought. Earthquake Consideration should be given to the effects of earthquake phenomena, the analysis of which is beyond the scope of these Recommendetions. Reference should be made to technical literature, for example References A2.5.1 and A25.2. 9 (© IGEM , 12 York Gate, London, NWA 4QG. Website: waw.igem.org.ukIGE TO/12 Edttion 2 5.6.6 Traffic Consideration should be given to the possibility of "live" loading due to vehicles on buried pipework. Note 1: The seisction of the appropriate leading for design will depend on the surface type and usage ebove the buried pipework and on eny fixed barriers for control of vehicie access. Note 2: The avallabe guidance for transient loading on buried rigid pipes, for exemple References 42.5.3 and 42.5.4, is considered particularly relevant. 57 SUPPORTS, ANCHORS AND APPLIED DISPLACEMENTS: Pipe quides and anchors will reduce free movement. The rigidity of either type of resistance against pipe movement should be included In the analysis. Additionally, the effects of movernent of pipe supports and connected equipment should be taken into account. 5.8 SOIL The soil around buried pipework will act as a restraint on the movement and Geflection of the pipe structure and, therefore, should be taken into account. Note: Soll restraint wi reduce movement of pipework by a combination of frictional resistance on ibe coating and by transverse restraint in the vicinity of bends and tees. Bending ‘moments at bends and tees may ba increased due to reduced system faxfoilty. 10 © IGEM , 12 York Gate, London, NW 496, Website: www.igem.org.ukAGE 1/22 Edition 2 SECTION 6 : PERFORMANCE 6.4 6.1L 6.1.2 6.1.3 6.2 6.21 6.2.2 63 64 GENERAL ‘The mechanical design of pipework should ensure safety with adequate performance for the design life. Note: This can be achieved by verifying, through analysis, that material stress levels are within ‘specified acceptance limits and that displacements, internal forces and reactions can be ‘accommodated by supports and connected equipment. ‘Analysis should involve the initial determination of pipework wall thickness requirements based on rules for basic pressure design. Overall pipework response should then be assessed to verify structural and pressure-retaining integrity under the design loading conditions. This assessment should demonstrate that pipework will not fail by rupture due to internal pressure and ‘other sustained loading or by progressive plastic deformation or fatigue cracking due to cyclic loading. System deflections and load transfer at supports should be checked to confirm support design and performance assumptions used in the modelling. Similar checks should be performed for any connected equipment. INTERNAL PRESSURE ‘The primary stresses in pipework components due to direct loading from internal pressure should be controlled according to the area classification of the installation and SMYS of the selected material at the maximum design temperature. Note: The area classification controls the hoop stress to 30%, 50% or 67% of the appropriate SMYS. The wall thickness values calculated to control pressure stress are considered @ minimum and should be Increased to accommodate allowances for manufacturing under-tolerances, corrosion and erosion. Pipework components should then be selected to meet or exceed the total required wall thickness values. Note: The setection of pipework components according to pressure (including atiowances) provides the intial condition for subsequent stressing ceiculations or analysis under other syste loads NORMAL SUSTAINED LOADS The pipework stress state, system deflection and load transfer to supports and equipment should be acceptable under the action of all normal sustained loads. They include internal pressure, MIP, weight and normal wind loading, if applicable (see clause 5.6.2). If the design includes for the pre-loading (cold- pull) of connections/joints during construction, these should be also treated as sustained loads, Note: For this analysis, MIP is considered to be @ normal sustained load. Acceptance should be based on keeping the von Mises equivalent stress throughout the system to a defined proportion of SMYS. ABNORMAL SUSTAINED LOADS Pipework and components may experience temporary or occasional sustained loads which are either additional to, or higher than, the normal sustained load iL @ IGEM,, 12 York Gate, London, NWI 4QG. Website: wwu.igem.org.ukIGE TO/12 Edtion 2 65 6.6 66.1 6.6.2 67 674 case, These may include "live" loading due to vehicles, ground settlement, abnormal wind, snow and ice loading, and seismic loading. ‘Acceptance should be based on keeping the von Mises equivalent stress throughout the system to a defined proportion of the SMYS. ote: The acceptance stress level for abnormal foading is approximately 12% higher than for the normal sustained case. HYDROSTATIC TEST ‘The hydrostatic pressure test condition should be assessed as an abnormal sustained loading case. ote: This includes the additonal weight loading from water and an internal pressure in excess of the design pressure. ‘Tis load case 1s likely to be the most onerous In relation to plpework components and ‘equipment suppart design, elthough adcitional temporary supports can be specified Acceptance should be based on keeping the hoop stress to within 90% of the SMYS at the test temperature and the von Mises combined stress to within the requirements of the abnormal sustained case (see also clause 8.5.3). SHAKEDOWN The maximum stress range experienced by the pipework and components due to variations in coincident pressure and temperature should be controlled in order to avoid progressive plastic deformation (also termed Incremental collapse or ratchetting) Shakedown assessment should be performed between pairs of load cases and the limiting pair identified. Acceptance should, generally, be based on keeping the maximum von Mises stress, based on component stress ranges, to within a multiple (typically 1.8 to 2.0 (see Table 4)) of the material mean SMYS over the temperature range associated with the pair of load cases. Note: This ensures elastic shakedown occurs and thst the fatigue design curves are valid for use. If the maximum von Mises stress based on component stress ranges exceeds the elastic allowable range, it may still be possible to demonstrate that progressive deformation does not occur. Alternative calculation methods, such as finite element analysis (FEA), should be used to determine if the region that exceeds the elastic allowable range is sufficiently localised so that global elastic behaviour occurs without any progressive deformation. Hence, the localised region will have achieved shakedown to elastic-plastic cyclic behaviour due to the surrounding elastic body. ‘Note: Resolving code stress exceptions is described In Section 9, FATIGUE ‘The maximum stress range experienced by the pipework and components due to variations in coincident pressure, temperature and other loadings should be assessed in order to avoid local structural failure due to fatigue cracking. Acceptance should be based on keeping the maximum principal stress range to below the value from the relevant fatigue design curve for the number of cycles anticipated over the design life. 12 © IGEM , 12 York Gate, London, NW 4QG. Website: www igem.org.uie1GE TO/12 Edition 2 6.7.2 For varying stress range amplitudes over the service life of the pipework, the fatigue assessment should involve @ cumulative damage calculation 6.7.3 The fatigue design curves assume that shakedown to an elastic cycle has been demonstrated. Where shakedown to elastic-plastic cyclic behaviour occurs, and has been demonstrated to be acceptable against progressive plastic deformation, the subsequent fatigue analysis should include the use of plasticity correction factors (see Appendix 5). 13 @ IGEM , 12 York Gate, London, NW 4QG, Website: wwwrigem.org.ukIGE 70/12 Euition 2 SECTION 7 : WALL THICKNESS AND SELECTION OF PIPEWORK 74 72 724 COMPONENTS GENERAL Calculations should be made for pipe and fitting wall thickness for sizing ‘components for the intended main load. IGE/TD/13 details recognised standards used for component selection and should be consulted prior to performing any stress analysis, Note: The wall thickness calculations given in IGE/*D/13 are only applicable for pipework components whose main ioad Is internal pressure. STRAIGHT PIPE ‘The thinnest manufactured wall thickness for the specified pipe should be equal to or greater than the sum of the calculated minimum wall thicknesses and, when applicable, the corrosion and erosion allowances stated in Sub-Section 7A. ‘The formula from clause 5.2.1 and the value from Table 1 should be used to determine minimum wal! thickness. NOMINAL MINIMUM NOMINAL DIAMETER OF PIPE) — WALL THICKNESS, (mm) cm) 50 3.9 80 478 100 4.78 150 4.78 200 6.35 250 6.35 300 6.35 400 6.35 450 6.35 500 7.92 600 7.92 750 9.52 900 9.52 1050 1191 4220 12.7 Note: At lower pressures, these minimum nominal wall thicknesses are necessary to withstand hhandling during construction and while in service ‘TABLE 1 - MINIMUM NOMINAL WALL THICKNESS OF STEEL PIPE 14 @ IGEM , 12 York Gate, London, NW1 4QG. Website: wrw.igem.org.ukIGE TO/12 Eeition 2 7.3 7.34 7.3.2 74 7.441 7.4.2 ‘The classification of Area Types, should be as shown in Table 2. SITE DESCRIPTION AREA TYPE | MAXf Process plant manned by process All 0.67 personnel Locations protected against external R 0.67 interference and unmanned S 05 T 0.3 Locations unprotected against all 0.3 external interference and unmanned Notes: Type R Rural areas with @ population density not exceeding 2.5 persans per hectare. 2 Types areas intermediate in character (between Types R and T) in which the population density exceeds 2.5 persons per hectare and which may be developed extensively With residential properties, schools, shops efc. + TyeT Central areas of towns oF cies, with 2 high population density, many multi-storey Buildings, dense traffic ang numerous underground services. TABLE 2 - DESIGN FACTOR (f) FITTINGS IGE/TD/13 details recognised standards used for selection of fittings, i.e. bends and tees, and should be consulted before proceeding with stress analysis. Existing fittings should be checked to determine whether they meet the minimum design requirements of the appropriate standard THICKNESS ALLOWANCES Reduction in wall thickness caused by chemical action, oxidation, scaling, abrasion or erosion should be taken into consideration in the selection of wall thickness. Where thinning is expected to occur, the minimum nominal wall thickness of pipework components should be increased by an allowance (C) to compensate for the resultant expected reduction during the desian life. 15 © IGEM , 12 York Gate, London, NW1 496. Website: wit igem.org.ukIGE TD/L2 Eaton 2 SECTION 8 : STRESS AND FATIGUE ANALYSIS 84 8.14 8.1.2 8.13 B14 8.15 8.1.6 8.2 8.3 8.3.1 8.3.1.1 GENERAL PROCEDURE The flexibility of pipework Is made up from the flexibilities of pipe, fittings and ‘equipment. Equipment should be assumed rigid unless a flexibility Factor (k) can be assigned. Note: The exibilties oF pipework fittings are given in Appendix 4, together with associated potes. A method should be used that calculates the following for the pipework and components: pressure stresses that are produced in both the axial and hoop directions solely to contain internal pressure ‘* bending stresses, shear stresses and axial stresses induced by the combined effects of temperature and pressure in the pipework, by applied end thrusts, externally applied loads, weight of pipework/fittings and contents and effects due to soil restraint stresses from occasional loads such as hydrostatic test, wind, earthquake, traffic etc,, if deemed necessary. An analysis should be performed on pipework between points where boundary conditions are known or sufficiently remote as to not affect the pipework of interest. Analyses to cover sustained, shakedown and fatigue conditions should be performed to satisfy the design brief and the performance criteria outlined in Section 6. The minimum wall thickness of any pipe, fitting or equipment should be determined in accordance with Section 7, after considering the appropriate design factor for the area type where the item is to be installed. Any pipework component that does not meet the design requirements of Sub- Section 8.5 may be re-examined and this should be carried out using techniques outlined in Section 9, ‘SCOPE OF PIPEWORK FOR ANALYSIS All pipework should be considered for analysis, unless documentary proof is already avaliable to show that the pipework, or equivalent pipework, already satisfies this edition or IGE/TD/12 Edition 1 (see also clause 11.1.2). PIPEWORK MODELLING General A pipework model should be established in accordance with good modelling practice, with particular attention being given to the considerations outlined in clauses 8.3.1.1 to 8.3.1.7. Geometric and material properties of components The correct combination of geometric and material properties should be applied. Nore: This is Mustrated particulerly for set-in or set-cn tees, where the tee material may be different tthe attached pipe. 16 (© IGEM , 12 York Gate, London, NWI 4QG. Website: www.igem.org.ukIGE TD/12 Edition 2 8.3.1.2 8.3.1.3 83.14 83.15 83.1.6 8.3.1.7 8.3.2 Element type and length ‘The appropriate element type should be selected to represent the appropriate mass, stiffness and rotational degree of freedom of the component. The element size should be sufficiently refined in the model to ensure that any non-linear behaviour is represented accurately. For straight pipe which is close to components of interest, such as bends, tees, etc., element lengths should be reduced. Note: In particular, this fs important for buried pipework to enable the soil restraints to be ‘realistic, especially at branch connections, Boundary conditions Boundary conditions should be applied carefully to ensure their influence does not invalidate the stress analysis results of items that may be affected by them. Pipework should be modelled as a complete system. Where this is not possible, consideration should be given to the use of overlapping madels. If this approach is inappropriate and the system has to be broken down into discrete models, representative boundery conditions should be applied. Supports and restraints Supports and restraints should be modelled to realistically represent their behaviour. fiote: Further details are given in clause 9.2.4 Flexibility of branch connections The flexibility factors stated in Appendix 4 for branch connections are assumed to be unity, which is conservative. If a better estimate of the forces and moments 2t these connections is required, it should be made by improved modelling of the flexibility of branch connections, Note: This latter approach can be applied t0 the Initial pipework model, or applied where the ‘analysis results show stress exceptions in branch connections. Further details are given in etouse 9.2.5 Temporary pipework Consideration should be given to the inclusion of temporary pipework in the stress model. Note: Where such pipework fs to be installed for a significant period of time, a fatigue analysis: ‘may be appropriate in addition to an assessment of strass levels. Model verification The model should be verified to ensure, as far as reasonably practicable, that the model represents the design of the pipework system, or that which exists. Longitudinal pressure strain Changes in the length of pipework will induce loads into restrained pipework systems. Such changes, due to internal pressure and offset by the Poisson effect of hoop expansion, should be taken into account. 17 © IGEM , 12 York Gate, Londen, NWI 4Q6. Website: www. igem.org.ukIGE TO/12 Eaton 2 833 8.3.4 84 84.1 Thickness for analysis Where pipework components are used which comply with “recognised industry standards’, for example those listed in IGE/TD/13 (thus ensuring that manufacturing tolerances on wall thickness are adequately controlled) the nominal wall thickness less any corrosion allowance should be used in performing the flexibility and stress analyses. Where this is not possible, the flexibility analysis and stress analysis should be performed using both maximum (including corrosion allowance) and minimum, (less any corrosion allowance) wall thickness. In each case, the same wall thickness values should be used in the flexibility analysis and for the stress calculations, Estimation of soil restraint ‘The calculation of soll restraint values for stress analysis requires the application of several algorithms based on theoretical and empirical techniques. These require input parameters that define the installation geometry and pipe dimensions, soil mechanical properties, water-table level and pipe coating frictional properties. The reliability of soil restraint modelling will be influenced by the pipe/soil structural modelling capability, the choice and performance of predictive algorithms, and the ability to select appropriate geo-technical parameters for a specific site. The minimum capability for routine analysis should be elastic flexibility analysis using a line element representation with forces and restraint springs applied at element-to-element nodal points. For soil restraint simulation, the springs should offer, as 2 minimum, a bi-linear force/displacement relationship and each buried pipe node should have four springs in multiple directions with different stiffness values (where appropriate) to cater for the directional dependence of soil restraint. Both upper and lower bound soil restraint conditions should be considered. Note: Guidance on appropriate predictive techniques and soll property requirements can be found In NEW 3650 Code 1 and in ASCE Publication 2. In some cases, site-specific investigation of existing installations will allow load cases to consider a single soil restraint condition. For this to also apply to new piping proposals, the site-specific ground investigation should be supported by construction requirements and controls (see References A2.5.3 and A2.5.4). LOAD CASE TABLE ‘An example load case table is given in Appendix 7. ‘Compilation The validity of a stress analysis depends on the modelling of the system and the proper selection of feasible combinations of operating conditions. The load case table should be identified at the outset of any pipework stress analysis based on the design brief. Note: Subsequent modifications to 2 load case table will require the stress and fatigue analysis to be upaated. 18 © IGEM , 12 York Gate, London, NW 4QG. Website: wwitigem.org.ukIGE TD/12 Edition 2 8.4.2 85 B51 B51 85.1.2 8.5.2 853 Use Piping analysis for the load cases covering the following shoutd, normally, be performed for both a lower bound and an upper bound soil restraint condition: © operating conditions (for shakedown and fatigue) © normal sustained (pressure and weight) © stresses from occasional loads, for example hydrostatic test, wind, earthquake, traffic, etc. if deemed necessary. Note: The range between bounds will reflect the uncertainty over instalation and geo-technicel conditions. STRESS ANALYSIS PROCEDURE AND ACCEPTANCE LIMITS General Commercially-available software for the static and dynamic analysis of pipework systems, and their support structures, is readily available. However, the software used in the analysis of the pipework should be supported by appropriate verification documentation. Hand calculations may be undertaken as an alternative to the use of software but care should be taken in the choice of any simplifying assumptions that are made and checks should be carried out on results, The procedure should include the calculation of stresses for the following conditions: * Sc= maximum equivalent stress induced by internal pressure, pipe weight, weight of contents and insulation, steady applied loads, effects’ of soil and normal winds and long range thermal stresses that Include significant elastic follow up * Sy= maximum von Mises equivalent stresses from the component stress ranges due to cyclic load which will allow the pipework to shakedown * Sq = principal stress ranges causing fatigue in the pipework, Note ir In certain circumstances, it may also be necessary for stresses to be calculated for ‘abnormat sustained conditions, These are denoted by Sui, Suz... Sar Ses, Su? Which equate f0 maximum equivalent siress, es for Se, but including effects due to each of the occasional abnormal loads, 4 tof in turn. Note 2: The formulae to calewlate the principal stresses and the equivatent von Mises stresses are given in Appendix 2. The individual stresses used in these formulae are defined in Appendix 3. Analysis of sustained conditions ‘The maximum sustained stress throughout the pipework should be calculated using the method given in Appendix 3 and should satisfy the criteria given in ciause 8.5.3. If the design includes pre-loading (cold-pull) of connections/joints during construction, these loads should be treated as sustained loads. If the temperature at assembly of the pipework is outside the range of operating conditions, this temperature should be taken into account, When calculating the sustained stresses Ss and S..s, SCFs should be included from Appendix 4. The weight of any liquid hydrostatic test should also be included in determining this stress. Acceptance limits for sustained conditions The maximum permitted sustained stresses for normal and abnormal conditions should be taken from Table 3. In addition, reference should be made to Sub- Section 6.5 for the assessment of hydrostatic test conditions, 13 © IGEM , 12 York Gate, London, NW1 406. Website: wwis.iger.org.ukIGE TO/12 Edition 2 8.5.4 8.5.5 8.5.6 NORMAL (S,) ABNORMAL (Ssapt, Seab2. .. Seati) SMYS| SMYS SMYS) SMYS £07. >0. SINS 0.74 | SUVS. 0. f urs $074 | “urs 7974 urs £974 | “rs 7°74 0.67 [0.80 SMYS [0.34 (SMYS +UTS) [0.90 SMYS | 0.38 (SMYS +UTS) 0.60 SMYS | 0.25 (SMYS +UTS) | 0.67SMYS | 0.28 (SMYS +UTS) 0.36 SMYS_| 0.15 (sms +uTs) | 0.40 sms __| 0.17 (SMYS +UTS: Note: SMYS and UTS values are obtained at the maximum design temperature ar, if known, the maximum temparature appropriate ta the sustained load case being analysed, TABLE 3 - MAXIMUM PERMITTED SUSTAINED STRESSES for Sz, Szspi/ San: 1 Senbi Analysis of shakedown conditions The maximum von Mises equivalent stress (Sym) based on the component stress ranges (Sz, Sr, Sq) Should be evaluated by the method outlined in Appendix 3 for each component in the pipework. The calculation should be based on the range of coincident pressure and temperature, from minimum to maximum pressure and from minimum to maximum temperature and should also include any changes in dead weight and external loads. If the temperature at assembly of pipework is outside the range of operating conditions, this temperature should be taken into account. Loadings should be investigated to ensure that no greater stress ranges exist at Intermediate pressure and temperature conditions its for shakedown conditions Acceptance i The calculated maximum von Mises equivalent stress (Syy) based on the component stress ranges (S., Sh, Sq), should not exceed Spx defined by: Se = 0.5 Kao(Sy + Si) sm = maximum von Mises equivalent stress range cue to cyclic load to ensure elastic shakedown of the pipework. Sm = maximum von Mises equivalent stress range from the component stress renges due to cyclic loss. Sy = SMYS or, for materials which do not exhibit a clear yield point, the stated proof stress of the material at rcom temperature. Sw SHYS of, for matetiis which da not exhibit a clear yield point, the stated proof sess of the material at maximum temperature. Keo = shakedown factor for the material (see Table 4). IF evidence exists to show that this factor is higher than that givan in Table 4, tha higher value may be used MATERIAL K Carbon steel__| 1.8 ‘Austenitic steal_| 2.0 TABLE 4 - SHAKEDOWN FACTOR, Kso Analysis of cyclic conditions ‘The maximum principal stress range (Sx) at each component in pipework should be evaluated using the method described in Appendix 3 for each combined pressure and temperature range during service life, If the conditions in clause 8.5.7 are not satisfied, a cumulative damage ‘assessment should be carried out for the predicted stress cycles due to variable amplitude loading according to Appendix 5 20 © IGEM , 12 York Gate, London, NW: 4Q6. Website: www.igem.org.ukIGE TO/12 Edition 2 8.5.7 85.7.1 8.5.7.2 8.6 86.1 8.6.2 8.6.3 Acceptance limits for cyclic conditions If the Sp levels from Table 5 are not exceeded, the pipework component may be considered satisfactory without the need for a fatigue analysis. Alternatively, If the Sx levels from Table 5 are exceeded, but the total number of equivalent stress cycles occurring in the design life does not exceed the number specified in Table 5, the pipework system can also be considered satisfactory without a fatigue analysis. This latter threshold only applies when the component loading is due to pressure only and the elastic shakedown limits are met. Otherwise, a detsiled fatigue analysis should be carried out as described in Appendix 5. An equivalent stress cycle is defined as the number (Mea) of maximum operating pressure range cycles that cause the same damage as the summation of (n)) cycles of pressure range (Pi) for each transient from the load case table. eg = Ma [Py / MOP] + np Py /MOP]” = nj /MOP]” m= see Table 26 If a detalled fatigue analysis is necessary, the calculated fatigue damage should not exceed unity. FATIGUE FATIGUE ANALYSIS NOT REQUIRED IF EITHER OF DESIGN _|_ FOLLOWING THRESHOLD VALUES ARE NOT EXCEEDED CURVE PRINCIPAL STRESS NUMBER OF EQUIVALENT (see Appendix 5) RANGE Sp STRESS CYCLES (N/mm?) (PRESSURE ONLY) Class D 53 100 Class E 47 100 Class F 40 100 Class G 29 100, Class W 25 100 TABLE 5-FATIGUE ANALYSIS THRESHOLDS FOR CARBON AND AUSTENITIC STEELS REACTIONS Supports For the analysis of pipework, pipe supports are usually often modelled by applying boundary conditions to the pipe. The forces and moments reacted at these boundary conditions should be used to check that the design of the supports is acceptable. Loads on connected equipment The forces and moments applied by pipework to connected equipment should be checked against manufacturers’ allowable limits or, if these are not available, against calculated limits. Strain-sensitive equipment The forces and moments imposed by pipework on strain-sensitive equipment should be checked against manufacturers’ allowable limits. 21 @IGEM , 12 York Gate, London, NWI 406. Website: www. igem.ora.ukIGE 70/12 Eaton 2 87 88 DISPLACEMENTS Displacements calculated by the pipe stress model should be reviewed to ensure that contact between components, structures or ground, which has not been Idealized in the model, is not occurring, ‘Note: Intermittent contect during trensients leads to non-tinearities. If non-linearities are suspected to be occurring, and they cannot be demonstrated as being insignificant, the pipe stress model should be modified to Include the non-linear effects and new results obtained. BUCKLING ‘Where significant compressive stresses are present, either during operation or under hydrostatic test conditions, the possibility of buckling should be investigated. Empirical limits are provided in Section A3.5 of PD 5500 for compressive general primary membrane stress. Note: For cases where compressive stresses are generated due to highly concentrated external joads, an acceptable procedure is to lt the total compressive stress fn any direction to 0.9 SMYS, 22 © IGEN, 12 York Gate, London, NWI 4QG. Website: ww igem.org.ukIGE TO/12 Eetion 2 SECTION 9 : RESOLVING STRESS EXCEPTIONS O41 9.2 9.2.4 9.2.2 GENERAL When @ component is found to violate any of the design criteria in Sections 6 and 8, the component is termed a “stress exception”. It Is possible to reduce the stress exceptions to an acceptable level by a variety of methods and improved modelling techniques. Some guidance is provided below detailing how stress exceptions can be reduced by: * obtaining component-specific or site-specific data, for example “as-built” wall thickness incorporating refined modelling techniques, for example branch flexibility using detailed FEA, for example DBA modifying the layout of the pipework, for example by improved flexibility replacing the component modifying constraint, for example soft fill refinement of other input data. The list is not exhaustive and, where new and improved techniques become available, they should be considered. These factors should be considered in resolving stress exceptions to determine the most efficient route in terms of both time and cost. The following sections detail how improved data, modelling techniques and new design methods can be used for new-build sites and those sites where a change in design conditions is to be made. NEW-BUILD The following component, site-specific data and modelling techniques can be used in resolving shakedown, sustained and fatigue stress exceptions, Wall thickness In those cases where pipes and fittings have already been procured on the basis of design for pressure, the flexibility and stress analysis (including calculation of SCFs) may utilise revised wall thickness values based on actual recorded measurements For piping components that comply with recognised industry standards, for example those listed in IGE/TD/13, the minimum recorded "as-built" ‘wall thickness values (less any corrosion allowance) should be used. This will be of benefit only if the minimum recorded "as-built" wall thickness is greater than the nominal wall thickness used in the initial analysis (see clause 8.3.3). For pipework components that are not manufactured to recognised specifications, the stress analysis (including calculation of SCFs) should be performed for both the maximum recorded wall thickness values (without any corrosion allowance) and for the minimum recorded wall thickness values (less any corrosion allowance). In each case, the same recorded wall thickness values should be used in the flexibility analysis and for the stress calculations, Forged tees to BS 1640 ‘The bending SCF for forged tees to BS 1640 is based on the crotch thickness being 1.5 times the nominal thickness of the attached pipe. If the following criteria can be met, the minimum recorded wall thickness of the tee, rather than 23 © IGEM , 12 York Gate, London, NWI 4QG. Website: mww.igem.crg.ukGE TO/42 Edition 2 9.2.3 9.2.5 the attached pipe, can be used to calculate the bending SCF, ly, noting that the mean radius (Rn) will also then relate to the tee header: © the crotch thickness is greater than 1.5 times the minimum tee header thickness: ‘© the lengths of the header and branch meet the requirements of BS 1640, ‘Only when both these criteria are satisfied can the minimum recorded tee wall thickness be used to re-calculate the bending SCF, is. Nevertheless, the minimum recorded tee wall thickness values can be used In the stress analysis and to re-calculate the pressure and thrust SCFs, i; and i respectively, as indicated in clause 9.2.1, Soil restraint For new-build situations, the designer has the option to control the selection and placement of backfill to pipework below ground. This can reduce the uncertainty ‘over the soil restraint conditions adopted in the piping stress analysis. Engineered solutions for tow or high restraint can also be adopted In order to affect the buried pipework's response to loading. These can include: bagged polystyrene beads lightweight granular fill cement/bentonite grout compacted coarse granular material concrete. ‘Note: The placement of large places of elecrica! insulating material adjacent (© buried pipework ‘snot preferred as it will reduce the effectiveness of cathodic protection. For existing pipework, the option exists to perform a ground investigation at the site and to arrange for appropriate classification and mechanical testing of representative material. The scope of the specified work will dictate the extent to which clarification is achieved on groundwater level, pipe trench backfill, and natural ground. A detailed study would confirm the seasonal water level range, the natural ground properties and the nature and condition of backfill (including its extent below and adjacent to the pipework). Supports A refinement in support modelling, a re-positioning of supports or a change in support type may be sufficient to resolve stress exceptions. Examples of refined support modelling include: © directional restraint, for example lift-off © improved support stiffness (normal, transverse & rotational) © sliding friction. ‘A re-positioning of supports or a change In support type will, generally, require 2 clear understanding of the piping system response to loadings. Branch flexibility modelling For branch connections, the flexibility factors quoted in Appendix 4 are assumed to be unity, which is conservative. By Improved modelling of the branch connection flexibility, 2 better estimate of the forces and moments at these connections can be determined. The method Is known as "the weightless rigid link and point spring” and is detailed in References A2.5.5 and A2.5.6. 24 12 IGEN , 12 York Gate, London, NWI 4QG, Webske: wura.igem.crg.ukIGE TO/12 Edition 2 9.2.6 9.3 9.3.1 9.3.2 9.3.3 ‘The modelling technique requires a weightless rigid fink from the centre of the main (header) pipe to the surface of the main (header) and a point (zero length) spring representing the flexibility of the branch, The spring stiffness is calculated using empirical formulae. The allowable limits of the empirical formulae for the above method should be checked. The following checks should be made before proceeding with this method: © the modelling technique is only applicable to set-on branches i.e, weldolets, long weld neck flanges, etc. the axis of the branch connection is normal to the main (header) pipe surface the main (header) pipe radius-to-thickness ratio is less than 50 the branch pipe radius to main (header) pipe radius ratio Is less than 0.5. To avoid any ambiguity regarding the Implementation of modelling the flexibility using the above method, the designer should refer to References A2.5,5 and A256. Design-by-analysis (DBA) The DBA method is described in Appendix 6, Where Appendix 6 is used in resolving a stress exception, the limits for all four of the basic stress categories in A6.5 should be satisfied, in which case the component may be deemed fit-for- purpose. Where the DBA limits are not satisfied, modifications should be made to either the layout of the pipework, the component geometry, the component material strength or 3 combination of these options. EXISTING INSTALLATIONS Design conditions The methods and modelling techniques stated in clauses 9.2.1 to 9.2.6 can also be used in resolving sustained, shakedown and fatigue stress exceptions related to the change in design conditions of installations, for example pressure, temperature and cyclic duty. ‘The following additional data can be used in resolving specific stress exceptions, Operating history Where Information on actual pressure and temperature fluctuations seen by an installation is available, this can be used in an attempt to resolve a fatigue stress exception, The information should be used to determine a revised estimate of the fatigue damage accumulated prior to the change in design conditions. Subsequent pipe stress analyses should use this revised fatigue damage summation when estimating the total predicted fatigue damage at the end of design life for the change in design conditions. Temperature conditions Where temperature conditions vary, for example at a compressor station, metal temperature decay profiles along the pipeline length can be taken into account. This can be derived, for example, by using heat transfer modelling techniques. Where actual pressure and measured metal temperatures exist, these may be used to validate the use of a heat transfer mathematica! model. 25 (© IGEM , 12 York Gate, London, NWI 406. Website: enwwigem.crg.ukIGE TO12 Esttion 2 9.3.4 Where any shakedown or fatigue stress exceptions exists, it may be possible to use the predicted thermal decay profile to resolve the exceptions. In this case, a validated model should be available and the decay profile taken at the change in design conditions, However, the model should consider all other operating conditions to ensure that bounding data (due to the change in design conditions) is used. Site modifications Ik should be determined whether any site modifications have an Impact on the integrity of existing pipework. Where it is not possible to demonstrate the effect of the modifications oni existing pipework, a revised stress analysis and assessment of the modifications and existing site should be carried out. The methods and modelling techniques stated in clauses 9.2.1 to 9.2.6 can be used in resolving any stress exceptions. 26 1 IGEM , 12 York Gate, London, NWI 4Q6. Website: www. igem.org.ukIGE TD/12 Edition 2 SECTION 10 : VERIFICATION 10.4, 10.2 10.3 10.4 GENERAL Pipework system designs should be verified at appropriate times during the Iife of the plant, to ensure that all parameters used in the design and stress analysis process remain valid. It should, therefore, be ensured that: © actual operating conditions remain within those used for design and stress, analysis purposes © parameters and assumptions made during all stages of the design process remain valid. OPERATING CONDITIONS Within an appropriate period from the date of commissioning, the plant operational records should be examined to determine whether the actual operating conditions are within those used for design and stress analysis purposes, in particular: © maximum and minimum pressure © maximum and minimum temperature © cyclic operation, i.e. number and nature of pressure and temperature cycles. OTHER DESIGN PARAMETERS AND ASSUMPTIONS: Within an appropriate period from the date of commissioning, the plant should be surveyed to determine whether the original design parameters and assumptions, such as the following, remain valid: * corrosion allowance settlement of supports pipework movement ground conditions area type protection against external interference. ee eee ACTION TO BE TAKEN If, after examination, there are variations between the design and actual conditions that are considered to be significant, continued operation of the plant should be justified by either: * amending the pipework stress analysis input data, re-analysing and checking code compliance, or © examining the original analysis results to assess the sensitivity of the stress levels to the changed parameters and/or the magnitude of the stress levels likely to be affected by the changed parameters. For example, the stress levels may be relatively low, so will remain code-compliant, even with the amended parameters. 27 @ IGEM , 12 York Gate, London, NW 406. Website: wwrw.igem.org.ukIGE TO/t2 Edivon 2 SECTION 11 : DOCUMENTATION ia Waa ALL2 11.2 24 STRESS ANALYSIS RECORDS The purpose of maintaining pipework design documentation is two-fold. Firstly, it is intended to enable the trensmission of relevent data between various interested persons and disciplines during the design period and to record the progress of the design. Secondly, it records, for future assessment of the components, methods, data and decisions used in formulating the design. The availability of this information is important when there is 8 need, for example, for verification as set out in Section 10 Prior to commencing any pipework design based on these Recommendations, a formalised documentation procedure should be introduced to ensure that the various steps and requirements of these Recommendations are carried out and recorded in the correct sequence. The documentation method should guarantee that the relevant facts required for adequate system design are presented in a manner that will make for ease of assessment and that all factors are taken into account before making design decisions. 1n addition, the design documentation systems should not only dovetail together effectively, but should be part of an overall comprehensive project documentation system that covers all aspects of the project, including design, materials, component supply and construction. Note: Computer-aided systems involving intec-siated databases have many benefits to offer in this held and their use i recommended. "As-built” drawings should be prepared to ensure that a full record is made of the as-built situation. These drawings should record, in a precise manner, the actual location of all equipment, pipe runs, valves and instruments, piping spools, etc, This as-built data should, as a matter of importance, be checked against the design documentation to ensure that all aspects of construction conform to the design requirements. ‘Commissioning and operation: records Comprehensive operational records should be maintained which detail the conditions which the plant has experienced during its life, including commissioning, subsequent start-up and shut-down. They should be available in such a form as to facilitate the verification procedures outlined in Section 10, to enable future changes in design conditions to be accommodated and possible stress exceptions to be resolved. Records for situations where analysis is not required The decision not to carry out a comprehensive analysis, and the reasoning, records and calculations supporting this decision should be documented (see ‘Sub-Section 8.2). STRESS ANALYSIS REPORT General Sufficient documentation should be compiled In arder to provide an auditable trail of what has been analysed, how it has been analysed and the outcome of the analysis. This documentation will normally take the form of a Stress Analysis Report, which should comprise the items identified in clauses 11.2.2 to 11.2.11, 28 @DIGEM , 12 York Gate, London, NWA 4QG, Website: ww.igem.org.UukIGE TO/12 Edition 2 41.2.2 11.2.3 11.24 11.2.5 11.2.6 1.2.7 11.2.8 11.2.9 14.2.10 As a minimum, the Report should comprise the following: Design brief The design brief is an important part of the stress analysis process to demonstrate that the analysis clearly meets its requirements. Register The register of all documentation is used for creating the stress model, for example general arrangement drawing, detail drawings, material take-offs, material specifications, soils information, etc. Stress model drawing The stress model drawing shows the extent of the stress model and how the system extremities have been modelled, i.e. type of boundary condition applied and where applicable, how interfaces with other piping systems have been modelled. The location of strain sensitive equipment should also be identified. Load case table ‘The load case table provides details of the specific installation being analysed and includes the derivation of data used, for example design and operating parameters, cyclic requirements. Soils information Soils information includes: © raw soils data used as a basis for developing data for use in the pipe stress model methodology data used in the stress model supporting information demonstrating how variables such as soll type, pipework diameter, depth of cover, pipework coefficient of friction between pipe and soil interface etc. have been addressed Isometric plat ‘The isometric plot identifies the extent of the stress mode! and shows points of interest, such as boundary points, tees, bends, etc., using node/label/element numbers. Geometric and material properties This shows details of all geometric and material properties for all component types included in the analysis. It also shows whether the components comply with recognised industry specifications. Detail such as outside diameter, wall thickness, corrosion allowance and material grade are included as @ minimum. Stress exceptions Where any stress exceptions remain on completion of the analysis, details are provided to enable further work to be carried out as appropriate. Computer program input/output data All data is provided either in hard copy, floppy disk, CD, etc. as appropriate, identifying the program identity and status, 29 (© TGEM , 12 York Gate, Loncon, NW 4Q6. Website: waw.lgem.org.ukIGE 70/12 Eaition 2 112.11 Conclusions and recommendations This comprises a statement of the findings of the analysis, whether complete code compliance has been achieved and whether any further work is required. 30 © IGEM , 32 York Gate, London, NWI 4QG. Website: wi. igem.org.ukIGE 1D/22 Edition 2 APPENDIX 1 : GLOSSARY, ACRONYMS AND NOTATION ALA GLOSSARY AND ACRONYMS ACOP austenitic stainless steel corrosion allowance DBA design brief design factor (f) design pressure design temperatures elastic follow up equipment equivalent stress FE FEA flexibitity factor (k) Approved Code of Practice. Chrome-nickel steel with low carbon content, commonly known as 18/8 stainless steel. Additional pipewall thickness provided where corrosion or eresion is expected to be of significance during the life of the pipework system. Design-by-analysis. A brief provided by the client to define the requirements of the work. ‘The Initial selection of a pipe wall thickness is based only on hoop: stress. Design factor ensures that the wall thickness has an allowance to accommodate additional loads, usually initially unknown, other than just internal pressure. The magnitude varies with the area type at which the site Is located and dictates the allowable sustained stress limits to be applied. The safety margin on sustained stress is greatest in heavily populated areas compared to rural areas. ‘The pressure on which design calculations are based. The maximum and minimum temperatures that a section of pipework could experience in operation. A secondary stress is self limiting in that it will reduce as a result of distortion, i.e. deformation-contralled. lastic follow up refers to instances where the elastic displacement is sufficiently large that the stresses appear to be load-controlied even though distortion has occurred. Metallic parts and’ assemblies joined together by pipes and components, for example regulators, valves, compressors, and pumps. ote: Stress analysis of equioment is not covered by these Recommendations, A value of stress calculated from the equation developed by von Mises defining the criterion of yield using the principal stresses or individual stresses, Finite element. Finite element analysis. The introduction of pipe bends or branch junctions into a piping system results in the occurrence of additional flexibility. For bends, k is the ratio of the rotation that occurs between the bend ends to that which occurs between the ends of a straight pipe of identical length as the bend, For branch junctions, k Is the additional rotation, expressed as an additional fictitious length of branch pipe with’ a rigid juncture that occurs at the branch junction. a4 © TGEM , 12 York Gate, Landon, NW 4QG. Website: wwow.igem.orgiuk1GE TO/12 Edition 2 hoop stress IGE IGeEM individual stresses long range thermal stresses, maximum operating pressure (MOP) maximum incidental pressure (MIP) MiP Mop NAFEMS pressure regulating installation (PRI) PRI principal stresses proof stress (Ry) sec scr scl scp shakedown Stress in a pipe wall in the circumferential direction. Institution of Gas Engineers. Institution of Gas Engineers and Managers. Stresses along one axis in pipe/components induced by forces, bending moments and internal pressure. Stresses induced from axial thermal expansion of a long straight pipeline from a uniform temperature rise. The stresses behave as primary stresses (see A6.4) due to elastic follow up. Hence, long range thermal stresses are distinctly different from localised thermal stresses, for example due to through wall temperature gradients, that are secondary in nature. ‘The maximum pressure at which system can be operated continuously under normal conditions. ‘The maximum pressure which a system is permitted to ‘experience under fault conditions, limited by safety devices. Mote: This valve is permitted to exceed MOP. Maximurn incidental pressure. Maximum operating pressure. ‘A non-profit making international membership association for the Engineering Analysis Community. It includes organisations using, developing or teaching various forms of numerical analysis. (waww.nafems.org), ‘An assembly of equipment designed to regulate, or reduce, the pressure of gas. Comprises all pressure-containing and associated equipment between the upstream face of its inlet isolation valve and the downstream face of its outlet isolation valve. Pressure regulating installation, ‘At any point in the pipework, these are the maximum and minimum compressive or tensile stresses at right angles to one another, calculated by combining the individual hoop, axial and shear stresses at that point. The analysis in these Recommendations is two-dimensional and the principal stresses {ie in the surface of the pipework material. For materials which do not exhibit a sudden clear yield point, stress at which a non-proportional elongation, equal to @ specified percentage of the original gauge length, occurs. Stress corrosion cracking. ‘Stress concentration factor. Stress classification line. Stress classification plane. A condition of elastic system response once initial localised yielding has occurred under the limits of coincident pressure and temperature, 32 © IGEM , 12 York Gate, London, NWI 4QG. Website: wwin-igem.org uleIGE TD/12 Edition 2 shear stress singularities sMYS stress concentration factor (SCF) stress exception stress range sub-assembiies sustained stress tie-in UTS weldability Note: Locelisad regions may achieve shokedown to elastic-plastic eycie behaviour, without progressive deformation, i surrounded by an elastic body. A stress resisting the relative slippage of cross-sectional planes. ‘The stress is produced by the shear force and/or torsional moment acting on the pipe or component cross-section. Undefined or infinitely large values often resulting from a division by zero. Note: One meaning in FE refers to the occurrence of an infinite stress or strain ta sharp internal corner or under a point joad. AS the mesh Is refined the calculated stress wil Increase towards the theoretical vatue of infinity. Specified minimum yield strength. The ratio of local stress to a remote reference stress. Stresses induced in components and pipes are increased by geometric changes and material discontinuities in pipework. The ratio of this increased stress to the remote reference stress is the SCF. The derivation of the value of SCF is different for sustained and fatigue conditions, and for bending, thrust, torsion, shear and pressure stresses. Note: In IGE/FD/2 Edltion 2, the SCFs do not Include localised peak stresses ‘since these are already included in the fatigue design curves. A term used when the calculated stress on a component/fitting is found to excead any of the acceptance criteria for sustained, shakedown or fatigue ioading. The algebraic difference of stresses at a particular point in the pipe or component, due to variations in the operating conditions, Arrangements of pipework and equipment fabricated and assembled independently from the main pipework configuration often away from site at a manufacturer's premises. Stress induced in the pipe wall or component which is not changed by distortion or movement. ‘The point where significant sections and sub-assemblies are finally attached to one another. wrt ate tensile strength (also known as tensile strength). The ability of a material to be welded satisfactorily. 33 © IGEM , 12 York Gate, London, NW1 406, Website: ww igem.org.ukIGE TO/12 Edition 2 AL2 NOTATION ‘The following notation applies to the main sections of the Recommendations. Appendices include separate notation Sean: Ss Sy Sr tee Thinning allowance. Outside diameter of pipe or component. Diameter to wall thickness ratio. Elastic modulus. Design factor. Flexibility factor. Shakedown factor. slope of fatigue design curve. Expected number of cycles associated with pressure range P,. Equivalent number of cycles. Design pressure. Mean radius. Proof stress. Abnormal sustained stress arising from an abnormal load. Maximum von Mises equivalent stress range due to cyclic load to ensure elastic shakedown of the pipework. Principal stress range resulting from changes in normal operating conditions causing fatigue in the pipework, Normal sustained stress induced by Interna! pressure, pipe weight, weight of contents and insulation, steady applied loads, effects of soil and normal winds. ‘Maximum von Mises equivalent stresses from the component stress ranges due to cyclic load. SMYS or, for materials which do not exhibit a clear yield point, the stated proof stress of pipe material at room temperature. SMYS or, for materials which do not exhibit a clear yield point, the stated proof stress of pipe material at maximum temperature. Thickness of pipe, slow taper or branch of tee. Minimum wall thickness. Overpressure factor. cont. 34 © IGEM , 12 York Gate, London, NW2 4QG. Website: www.igem.org.ukIGE TD/12 Eaton 2 & Coefficient of thermal expansion. v Poisson's ratio. 35 © IGEM,, 12 York Gate, London, W1 4QG, Website: winw.igem.org.ukIGE TH/1? Edition 2 APPENDIX 2 : REFERENCES These Recommendations are set out against 2 background of legislation in force in Great Britain at the time of publication. Similar considerations are likely to apply in other countries, where reference to appropriate national legislation is necessary, ‘All relevant legislation must be complied with and relevant Approved Codes of Practice (ACoPs), official Guidance Notes and referenced codes, standards, etc. should be taken into account. Where British Standards etc. are quoted, equivalent national or international Standards ete. equatly may be appropriate. Care needs to be taken to ensure that the latest editions of the relevant documents are used. ADA LEGISLATION © Construction (Design and Management) Regulations © Gas Act ‘+ Health and Safety at Work, etc. Act ‘* Management of Health and Safety at Work Regulations © Pipelines Safety Regulations ‘© Pressure Systems (Safety) Regulations. Az IGEM A221 Recommendations on Transmission and Distribution Practice © IGE/TD/1 Steel pipelines for high pressure gas transmission Edition 4 © IGE/TO/3 Distribution mains Edition 4 Note: Publication anticipated in 2003. Use Edltion 3 until pubished. © IGE/TO/4 Gas services Edition 4 Note: Publication anticipated in 2003. Use Edttion 3 unti pubished. © IGE/TD/S Offtakes and pressure regulating installations for inlet pressures between 7 and 70 bar (obsolete) e IGE/TD/12 Pipework stress analysis for gas industry plant Edition 1 (obsolete) © IGE/TD/13 Pressure regulating installations for transmission and distribution systems. A222 Gas Measurement Procedures © IGE/GM/t Gas meter installations for pressures not exceeding Edition 2 100 bar © IGE/GM/4 Flowmetering practice for pressures between 38 and 250 bar. £2.23 Gas Legislation Guidance * IGE/GL/S Plant modification procedures. 36 © IGEM , 12 York Gate, London, NW1 4G. Website: wwew.igem.org.ukIGE 10/12 Edition 2 A23 A248 ALS A2SA 02.5.2 ALS ALS. 42.5.6 A287 A2S8 BSI (abbreviated titles) * Bs 806 Specification for design and construction of ferrous piping instaltations for and in connection with land boilers, © BS 1179 Glossary of terms used in the gas industry © Bs 1640 Steel butt-welding pipe fittings * Bs 4515 Welding of steel pipelines on tand and offshore © BS6299Pt2 —_Loading for buildings ~ wind loads © BS 7608 Fatigue design and assessment of steel structures © PD 5500 Unfired fusion welded pressure vessels * BS EN ISO 9000 Quality systems © BS EN 10002-1 Tensile testing of metallic materials. Method at ambient temperature BS EN 10002-5 Tensile testing of metallic materials. © BS ISO 10005 Quality plans. . OVERSEAS (abbreviated titles) © ASME, Boiler and Pressure Vessel Code, Section III Division 1, Sub-section NB ~ Class 1 components, 1998 Edition ASME B31.3, 1999 Process Piping NEN 3650 Code 1. Requirements for Steel Pipeline Transportation Systems. Nederlands Normalisatie Institute. Delft. 1992 * ASME VIII, Division 2, Boller & Pressure Vessel Code, American Society of Mechanical Engineers © ASCE Publication 2. Guidelines for the Seismic Design of Oil and Gas Pipeline Systems. MISCELLANEOUS Lilwal, RC Seismicity and seismic hazard in Britain. Seismol Bull, Insst. Geology. Sci., No 4. Dowrick, D.J. Earthquake risk and design ground motions in the UK offshore area. Proc. Inst, Civ. Enars., Part 2, 1981, 71, June. Young 0.€. and O'Reilly M.P. A guide to design loadings for buried rigid pipes. TRRL. DoT. 1983. Trott J.J. and Young ©.C, Buried rigid pipes: Structural design of pipelines. Elsevier, 1984, Welding Research Council Bulletin 297, "Local Stresses in Cylindrical Shells due to External Loadings on Nozzles - Supplement to WRC Bulletin No. 107", August 1984, Welding Research Council Bulletin 329, “Accuracy of Stress Intensification Factors For Branch Connections", December 1987. EEMUA Publication No. 153:1996. Hechmer, J.L. & Hollinger, G.L., "3D Stress Criteria Guidelines For Application”, Welding Research Council Bulletin 429, February 1998. 37 © IGEM , 12 York Gate, London, NWI QS, Website: www igem.crg.ukIGE TO/12 Edicion 2 APPENDIX 3 : STRESSES ABL NOTATION The terms are defined as follows for this Appendix. = outside diameter of straight pise/transition/reducer/bend/fitting header 2t section considered (mm). For full encirclement tees, this includes the feinforcement > = inside diameter or straight pipe/transition/reducer/band/fitting header at section considered (mm) = outside ameter of ting branch {mm}. This [s dependent on what location ofthe fitting Is used to define "t". See sketches in Appendix + Inside diameter oF Aiting branch (mm) thickness oF reinforcement for full encirclement tees (mmm) for full encrciement tees, this 's te outside clameter of the header pipe, Da = 2t (mm). For all other fttings, Dy = De 'SCFs for pressure, bending, thrust, torsion and shear (see Appencix 4) bending moment er bending moment range (N mm} in plane bending mament (N ram} Jn piane bending maments for load cases w and w (N mm) ‘out of plane bending moment (W am) ‘out of plane bering mament for load cases «and y (N mm) loys tal = torsional moment of torsional moment range (N mm) torsional moments for load cases w and w (N mm) = Wall thickness of straight pipe/transtion/ceducer/nitting branch at section considered (rm) For connections defined in Sub-Appendix A4.10, t= te For connections defined in Sub-Appendices A4.31.1'to A&.11.¢, t= Ty For connections defined in Sub-Appendices Ad.11.5 to A4.11.9, t= Ty 1 = Wall thickness of bend/fiting header at section considered (mmm). For full fencircement tees, is the sum of the main pipe thickness and the reinforcing thickness, te ws in piane shear force (1) Vn 2h Vie In piane shear forces for load cases w and w (N) ve ‘out of plane shear force (N} Vou and Von ‘out of plane shear force for load cases y ands (N} ve applied trangverse force ar shear force range (N) St ‘axa stress (Kiem?) 5h hoop stress (N/r Sa shear stress (N/mm) Sk = principle. stress range resulting from changes in normal operating ‘conditions causing fatigue in the pipework Sim = maximum von Mises equivalent stresses from the component stress anges due to cyclic load = internat pressure load or pressure load range (N/mm?) Pu and Pr. = internal pressures for load cases w ard w (Wem) F = sum of axial ferce oF force range (N) Fuand fe axel forces for load cases w and n (N) w = factor to be applied to the empirically derived branch bending SCF, In ‘The branch bending effective SCF is kw where W = UT DESCRIPTION Stresses in pipework are made up from individual stress components in the axial and hoop directions and from shear stresses. The individual stresses are given in ‘43.3 for pipe and a range of pipework fittings including bends. Include SCFs for bending (ia), axial force/end thrust (i), internal pressure (ip) torsion (J,) and shear loading (i,) when calculating the maximum stresses in pipework, bends and fittings as required in clauses 8.5.2, 8.5.4 and 8.5.6, The 2ppropriate SCFs are given in Appendix 4, For each of the combinations of conditions given in clause 8.5.1, calculate the hoop, axial and shear stresses (S,, S, and Sq) from the internal pressure (P), bending moment (M,), torsional moment (M,), applied shear force (Vz) and applied axial force (F) acting at that connection or branch, using the equations in A33. 38 (© IGEM , 12 York Gate, London, NW1 4QG. Website: wwew.igem.org.UkIGE 70/12 Eeltion 2 Evaluate the sustained stress for an individual set of conditions relevant to a single load case. For the single load case, the bending moment and shear force are the resultants, given by: Evaluate the shakedown stress for one or more pairs of load cases. The calculation Is based upon the effect of changes that occur in mechanical or ‘thermal loadings with time. Hence, a stress range is evaluated between two load cases, for example denoted nand respectively, The moments and forces from the load cases are combined before determining the resultant values. These are given by: My = Vin — Mag)? [Mou ~ Mo)? Vy = ylVin ~ Misr)? + orn ~ Vora)? Moreover, the resultant pressure, axial force and torsional moment load ranges (between toad cases y and 4) for the shakedown assessment are given by: PaPy-Py Fay ~ Fe My = May ~ Maa When a stress range Is to be evaluated between two load cases at different pressures, assign the load case identifier subscript « to the load case which has the higher pressure value, Hence, (P = Py ~ Px) will always be a positive value. ‘The principal stress range in any plane can be calculated for any set of conditions from: sy = $[[5 +8. 5 -8.F+ 482] Use this to establish the maximum principal stress range, paying due regard to the direction and sign, for use in the fatigue analysis. The von Mises equivatent stress or stress range is given by: Sv = IS +8,8, +38. ‘The permitted values of von Mises equivalent stress for sustained and von Mises equivatent stress ranges for shakedown loading are given in clauses 8.5.3 and 8.5.5 respectively. @ IGEM , £2 York Gate, London, RWI 406. WebsitIGE TD/12 Edition 2 A.3 A331 3.3.2 43.3.3 INDIVIDUAL STRESS COMPONENTS Use the following equations, in consistent units, to calculate the maximum tensile (positive) stress and the maximum compressive (negative) stress. Give due regard to the direction of the axial force, F, which is defined as the net axial force in the pipe wall taking account of the forces induced by thermal effects, dead load effects, restraints to movements and externally imposed axial loads but excluding the axial force generated in the pipe wall by internal pressure, ‘The equations assume pressure stresses due to unrestrained pipe behaviour. The equations for axial stress are marginally conservative for fully restrained pipes, where the axial pressure stress is due to the Poisson effect of the hoop stress. In practice, pipes are rarely “fully” restrained, even when buried and, therefore, it is conservative to assume that the axial pressure term is Po/at. Straight pipe, wall thickness transitions, concentric and eccentric reducers PD, Sy =I, zt 32M, De [D,*-D,*] 4Vr “1? - DF] 1, be 32 My Dy 27? 2 [Dat ~ BT seu [fe ae _)], wa2mo, 4t 2 [D.?-D/]}} x[D,* -b4] 16D, M, ‘i 4Vr xD, =D]" af,’ - D1 Sq Hi Weldolets, long weld neck flanges, weldoflanges, sweepolets and tees This includes fabricated, full encirclement, pad or saddle reinforced tees, forged tees to BS 1640, drawn/extruded and forged tees not to BS 1640 and welded-in contour inserts. Note: it's acknowledged that the structural behaviour of spit tees is very diferent to that of full fencirclement tees, Split tees cannat be accurately represented in 3 pipe stress model due to the pressure in the annulus effectively evering the fitting From the header pipe, causing complex stresses at the fillet welds. Seok advice when modeling spit tees in a pipe stress mode! where the intention Is t0 extract forces end maments for application to a FE model representing # spilt tee. 40 @IGEM , 12 York Gate, London, NWL 406. Websites vw.igem.org.ukIGE 10/12 Edition 2 A33.3.1 3.3.3.2 Branch pipe analysis Use the larger value oft PD. Pd Sp=ip Pe and Sy =i, 58 Use the larger value of: 32.My dy 27 nfd.* -4"] wd ~ 4°] and 32My dy S,=h eth Ws 47” Id." = 47] # [d.? - 61 iw Is the effective branch bending SCF, where w = /T. i.w is never tess than 1.0 for sustained, or either 1.5 or 2.0 for cyclic loading, as specified in ‘Appendix 4. 36 dy Ma_ 4 Ve rId.*—d*]* Udo" ~ ls Header pipe analysis Use the larger value oft Use the larger value oft 6. <1, 2204), 22MoPo 4 ee aT TD 08] © 2D? - D7] and 5, =/, Pedy ,, 32MyDo " 4F 27S pF + (Dy? —87] 16D.Ma_ 45 w[D)?-*] * #[D,? -D7] D la Do-2t, {for full encirclement toes only) Ds {far all other types of Fiting) at @IGEM , 32 York Gate, Londen, NW 4QG. Website: wrr.igem.org.ukIGE TD/12 Edition 2 APPENDIX 4 : STRESS CONCENTRATION FACTORS (SCFs) Aa SCOPE Use the SCFs given in this Appendix with the equations given in Appendix 3. Use the flexibility factors (k) given in this Appendix for the various pipework components for the flexibility analysis. ‘The values of SCFs and k are tabulated for commonly-used pipework components. Unless specific data, referenced in the Stress Analysis Report, exist that would warrant lower SCFs or higher flexibility factors than those tabulated in this Appendix, apply these stated data strictly. For pipework components not covered in this Appendix, SCF and k can be determined by experimental analysis or theoretical analysis such as detailed FEA. Include test data or theoretical analysis in the Stress Analysis Report. A4.2 DIMENSIONS Use nominal dimensions for calculating the numerical values of SCF and k. 43 NOTATION ‘The terms are defined as follows for the various sections in this Appendix: Section Geometry Aaa Aas Aa. a7 Straight pipe wall thickness transition Concentric and eccentric reducer Bend and welding elbow Closely spaced mitre bend Symbot D. ° D, 2 Definition Outside diameter of pipe Nominat wall thickness of pipe Outside diameter of pipe Nominat wall thickness of pipe Outside diameter at large end of concentric or eccentric reducer Outside diameter at small end of concentric or eccentric reducer Transition radius at lerge diameter end ‘Transition radius at small diameter end Nominal wall thickness at large end of ‘concentric or eccentric reducer Nominal wall thickness at small end of ‘concentric or eccentric reducer Cone angle Modulus of elasticity Internal pressure Centre-line radius of the bend Mean radius of matching pipe Nominal wall thickness of fitting Bend angle Effective bend radius = Scot 2 Mean radius of matching pipe Mitre spacing at pipe centre lines Nominal wall thickness of fitting One half of angle between adjacent mitre axes 42 © IGEM , 12 York Gate, London, Wi 4QG. Website: www. igem.org.0k Units mm mm min mm mm mm mm mm Degrees Nem? Némm? mam mm mm Radlans mm mm mm RadiansIGE T0/22 Edition 2 4.9 Single or widely spaced mitre bend Ad.10.1 Weldolet FefRy < 0.5 4.10.2 Long _ weld neck flange PefBy < 05 4.10.3. Weldoflange FefRy < 0.5 FP T Te Effective bend radius felt cot) Mean radius of matching pipe Mitre spacing at pipe centre lines Nominal wall thickness of fitting One half of angle between adjacent mitre axes Outside diameter of main pipe Mean diameter of main pipe Inside diameter of branch pipe Outside diameter of branch pipe Effective mean diemeter of reinforced or unreinforced branch = (d, +7 + 22) or (d, + Ty) depending on reinforcement Height of branch reinforcement Mean radius of main pipe Inside radius of branch pipe Mean radius of branch pipe Effective mean radius of reinforced or unreinforced branch = dn/? Outside radius of reinforced branch Fillet radi Nominal wall thickness of main pipe Wall thickness of branch reinforcement Nominal wall thickness of attached branch pipe Angle between vertical and slope Inside taper offset Outside diameter af main pipe Mean diameter of main pipe Inside diameter of branch pipe Outside diameter of branch pipe Effective mean diameter of reinforced or unreinforced branch = d) + Te+ 22 Height of influence of flange Mean radius of main pipe Inside radius of branch pipe Mean radius of branch pipe Effective mean radius of reinforced or unreinforced branch = dn/2 Outside radius of reinforced branch Fillet radii Nominal wall thickness of main pipe Wall thickness of branch reinforcement Nominal wall thickness of attached branch pipe Inside taper offset Outside diameter of main pipe Mean diameter of maln pipe Inside diameter of branch pipe Outside diameter branch pipe Effective mean diameter of reinforced or unreinforced branch = (d, + Tp + 2z) or (4, + T4) depending on reinforcement Height of branch reinforcement Mean raius of mein pipe Inside radius of branch pipe Mean radius of branch pipe Effective mean redius of reinforced or unreinforced branch = 9/2 Outside radius of reinforced branch Fillet radii Nominal wall thickness of main pipe Wall thickness of branch reinforcement 43 @ IGEN , 12 York Gate, Londen, NW1 4G. Website: ww.igam. org.uk mm mm mm men Radians mm. 33 mm mm. mm mm mm mm mm mm mm mm Degrees mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mmIGE TO/12 Edition 2 4.10.4 Fabricated tee F/R < 0.5) 4.11.1 Weldolet not meeting Section A4.10.1, A411.2 Long weld neck flange not meeting Section A.10.2 A411.3 Weldoflange not meeting Section 4.10.3, A4.11.4 Fabricated tee A411.5 Forged tee to BS 1640 D, De a do Nomina! walt branch pipe Angle between vertical and slope Inside taper offset thickness of attached Outside ciameter of main pipe Mean diameter of main pipe Inside diameter of branch pipe Outside diameter of branch pipe Effective mean diameter of reinforced or unreinforeed branch = (dj + Ty) oF (6, + T's) depending on reinforcement Height of branch reinforcement Mean radius of main pipe Inside radius of branch pipe Mean radius of branch pipe Effective mean racius of reinforced or uunreinforced branch = dnf2 Outside radius of reinforced branch Fillet re Nominal wall thickness of main pipe Wall thickness of branch reinforcement Nominal wall thickness of attached branch pipe Inside diameter of main pipe Outside diameter of main pipe Inside diemeter of branch pipe Outside diameter of branch pipe Mean radius of main pipe Nominat wall thickness of main pipe Nominat wall thickness of attached branch pipe Inside diameter of main pipe Outside diameter of main pipe Inside diameter of branch pipe Outside diameter of branch pipe Mean radius of main pipe Nominal wall thickness of main pipe Nominal wall thickness of attached branch pipe Inside diameter of main pipe Outside diameter of main pipe Inside diameter of branch pipe Outside diameter of branch pipe Mean radius of main pipe Nominal wall thickness of main pipe Nominal wall thickness of attached branch pipe Inside diameter of main pipe Outside diameter of main pipe Inside diameter of branch pipe ‘Outside diameter of branch pipe Mean radius of main pipe Nominal wall thickness of main pipe Nominal wall thickness of attached branch pipe __ Inside diameter of tee header Outside diameter of tee header Inside diameter of tee branch ‘Outside diameter of tee branch 44 8 IGEM , 12 York Gate, Lendon, NWI 4QG. Website: www.igem.org.uk mm degrees mm mm mm mm mm, mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mm mam mm mm mm mm mm mmIGE 10/22 Edition 2 441.6 — Drawn/ extruded and forged tee not to BS 1640 4.11.7 Sweepolet and welded- In contour insert AgiLa Full encirclement tee A4.11.9 Fabricated tee with pad or saddle © GEM , 12 York Gate, London, NW 4Q6. Outside diameter of attached branch pipe Mean radius of main pipe “Tee crotch radius Wall thickness of tee header Wall thickness of tee branch Tee crotch thickness Nominal wall thickness header pipe of attached Inside diameter of tee header Outside diameter of tee header Inside diameter of tee branch Outside clameter of tee branch Outside diameter of attached branch pipe Half width reinforcement length on the header Reinforcement length on branch from the outer surface of the header Mean radius of tee header Tee crotch radius Wall thickness of tee header Wail thickness of tee branch Inside diameter of main pipe Outside diameter of main pipe Inside diameter of branch pipe Outside diameter of branch pipe Mean radius of main pipe Fitting crotch radius Noriinal wail thickness of main pipe Nominal wali thickness of attached branch pipe Fitting crotch thickness Inside diameter of main pipe Outside diameter of main pipe Inside diameter of branch pipe Outside diameter of branch pipe Mean radius of main pipe Sum of nominal wall thickness of main pipe + encirclement thickness Nominal wall thickness of attached branch pipe Encirclement thickness Inside dlameter of main pipe Outside diameter of main pipe Inside diameter of branch pipe Outside diameter of branch pipe Mean radius of main pipe Nominal wall thickness of main pipe Nominal wall thickness of attached branch pipe Pad or saddle thickness 45 Website: wwr.igem.org.uk mm mma mma mm mm mm mm mm mm mm mm mm mm mm mim mm mm mn ram mm mm mm mm mm mm men mm mm mm mm mm mmIGE TD/12 Edition 2 Aaa STRAIGHT PIPE Sketch Validity limits kK The wall thickness of | 1.0 butting pipes is between ;— |0.875T and 1.1T for an axial distance jo;7 from centreline of weld pep For straight pipe remote from any welds or other discontinuities, take both the sustained and cyclic SCF as 1.0. Welds can exist in both the as-welded condition or ground flush. The requirements for a flush weld are: ‘the total thickness of the weld cap (reinforcement) should not exceed 0.17 © no concavity on either the interior or exterior surfaces of the weld should exist © the finished weld cap profile does not have a slope greater than 7° (where the angle is measured from a tangent to the surface of the pie). Where one of these requirements is not observed, classify the weld SCF as- welded. For straight pipe containing butt welds, take the sustained SCF as 1.0. Take the cyclic SCFs for the pressure, bending, thrust, torsion and shear loadings as those given in Table 6. CYCLIC SCF TYPE ip ie h ty te Pressure | Bonding | Thrust | Torsion_| Shear As-welded | Seam welded | 2.0 13 13 | 1.0 2.0 Seamless 13, 13 13 |10 2.0 ‘Seam welded | 2.0 Ld i | 1.0 2.0 Flush ‘Seamless La Lat 11 __|10 2.0 TABLE 6 - STRAIGHT PIPE SCFs 46 © IGEM , 32 York Gete, London, NW1 4QG. Website: wwve.igem.org.ukIGE TO/12 Edition 2 Aas WALL THICKNESS TRANSITION Sketch Validity limits Kk Wall thickness of} 1.0 butting pipes _is between 0.875T and 1.17 for an axial distance \Pat from centre line of weld Inside "Tanai en Pipe centreline For transitions meeting the validity limits, take the sustained SCF for all loadings as 1.0. ‘Take the cyclic SCF for the pressure, bending, thrust, torsion and shear loadings as those given in Table 7. CYCLIC SCF TYPE ip li ke ike i Pressure|Bending |Thrust_ {Torsion |Shear ‘As-welded 18 24 2.0 [1.0 [2.0 Flush 18 17 2.0 [1.0 2.0 TABLE 7 - WALL THICKNESS TRANSITION SCFs 47 (© IGEM , 12 York Gate, London, NW 4QG. Website: wiww.igem.org.ukIGE TD/L? Edition 2 AS6 CONCENTRIC AND ECCENTRIC REDUCERS: Validity k Sketch “te Cone angle o < 60° Wall thickness throughout the body of the reducer is | 1.0 ‘not less than the minimum thicknesses required to | resist the 2 design pressure. iting oertraine For reducers meeting the validity limits, take the sustained SCF for all loadings as 1.0 for cone angles < 30°, and 2.0 for cone angles from 30° to 60°. Take the cyclic SCF for the pressure, bending, thrust, torsion and shear loadings as those given in Table 8. TYPE eycuic scr Pressure Bending e Reducer +:0.0056e P= aan oho with Ry and Re | 1* 0905/7, 1.040.3600°C79) 1.0 | 1.0 | 20 20.1 Dy Reducer ; at 5 ith et gy, [BO 8.008801 ta) | 1.52 (ten) | 57] A% tm (4+ cot B) Zz —— TYPE SUSTAINED SCF CYCLIC SCF i, ip Ie ig i i, iy is Pressure|Bending|Thrust|Torsion| Shear |Pressure| Bending Shear ver less| Iwever fess] lenan 4.0 than 1.5 18 1.8 Seam 10 vy | 10] 10 | 10) 20 Se of 13 | 10 | 20 welded 0 nfs 18 18 Seamiess| 1.0 pe f to] we) a0} 23 7 13 | 10 | 20 Motes: Where flanges ae atached 1 one or oth ends, multiply te values of k and ein By the ctor C. 2. Parone end fanged, ¢ = n7B 2 Rorbath ends fanged, ¢ = 478 TABLE 11 - SINGLE OR WIDELY SPACED MITRE BEND SCFs St @ IGEM , 12 York Gate, London, NW 4QG. Website: www.igem.org.ukIGE TD/12 Edition 2 M10 SET-ON WELDED BRANCH CONNECTIONS ‘The SCFs in this section are only applicable for branch connections that satisfy both fabrication and geometrical criteria. The fabrication limitations are: (a) Set-on welded (b) The axis of the branch connection is normal to the run pipe surface (©) The arc distance measured between the centres of adjacent branches along the outer surface of the run pipe is not less than three times the ‘sum of the two adjacent branch inside radii in the longitudinal direction, or is not less than two times the sum of the two adjacent branch radii along the circumference of the run pipe (see Figure 1). ‘The geometrical limitations for typical set-on welded geometries are detailed in Sections A4.10.1 to A4.10.4, Satisfy these limitations before proceeding. Typical set-on welded geometries where the SCFs and flexibility factors are applicable are: © weldotet © long weld neck flange © weldoflange @ fabricated tee with the fillet weld radii specified, Details of flexibility factors and SCFs for the above branch connections are given in the following sub-appendies, Where the limitations are not satisfied, use the SCFs given in A4.11 in the first instance. When neither A4.10 or Aé.11 SCFs are appropriate, experimental or theoretical analysis can be used to either evaluate appropriate SCFs or to justify the use of the following SCFs. Include such data in the Stress Analysis Report. FIGURE 1 - ADJACENT BRANCH LIMITATIONS OF SET-ON BRANCH CONNECTIONS 52 © IGEM , 12 York Gate, London, NW1 4Q6. Website: wwigem.org.ukIGE TO/12 Edition 2 A410... WELDOLET ‘Sketch Validity fimits Kk Ra/T < 50 io TefRm < 0.50 (see Note’? for re value) For dy > 114.3mm then ry > Q.1T but <0.5T For dy < 114.3mm then ry Is not required i, can be divided by 2.0 if ra 2 Ty/2 or T/2 whichever is the larger If rz is unknown then ts shail be taken as Imm oo 15 is not less than the Main pipe. larger of 0.0020d, or 2(siney’ x (te- (r+ To) (See Note 4) T= Tp Or Ty (see Note 1 for Tavatue) ‘SUSTAINED SCF i, fe fg i TYPE Pressure Bending Thrust Torsion) shear aver fess than 20 (See Note 3) Main pipe 10 1.0 10 | 10 Branch pipe 10 1.0 10 | 10 CYCLIC SCF i is i i ie Pressure Bending Thrust Torsion) Shear Hever less than £2 | Never less than 5 (See Note 3) Main pipe 10 | 20 nch 10 | 20 Notes. 1. by 20,5y,Ty , then re can be token as the radius to the centre of To and Tr = Toy otherwise re= tm whlch isthe radius to the centre of Manel Ty = Ps 2 Bereftg> 12,1 = 22. 3 The product iw (see Appendix 3) is never less than 1.0 for sustained or 1.5 for eyetc assesomants 4, Tas nol tPickness of bronch reinforcement TABLE 12 - WELDOLET SCFs 53 (© IGEM, 12 York Gate, London, NW1 4G, Website: wwivlgem.org.ukIGE TD/12 Edition 2 4.10.2. LONG WELD NECK FLANGE Sketch Validity limits k RT < 50 10 e/Rey < 0-50 (see Note 1 for r. value) For dy > 114.3mm then ry > 0.1T but <0.5T For dy < 114.3mm then r, is not required Branch pipe can be divided by 2.0 if ra ,/2 oF 7/2 whichever is the larger 7 If ris unknown then rp shall be 7 taken as Imm Re Th= Ts 20.55% When this limit is not satisfied the SCF need to be altered to take into account the effect of the fienge SUSTAINED SCF | i, i i] TYPE Pressure Thrust —_[Torsion| Shear! Main pipe 10 1.0 x0 | 10 Branch pipe f3)"(2] AE) 10 to | 10 CVCLIe SCF 5 ty i moe Pressure Bending Thrust —_[Torsion| Shear| Hever less than L2 Wever less than 25 (See Note 3). in pi ofa Pry) + vane woo TOTO | aah [set] = ain vive | 45) (5) ae} aanayf” | 20 | 20 Branch pipe ode) TEL) | o of ey] 10 | 20 Notes: 4, els taken as the radius to the centre 2 Heyhty > 12, rT = 12. 3B assessments. of Tn ‘TABLE 13 - LONG WELD NECK FLANGE SCFs 54 @IGEM , 12 York Gate, London, NWI 4QG, Website: wwrw.igem.org.uk The product iw (see Appendix 3) is never less than 1.0 for sustained or 1.5 for eyeIGE TD/12 Edition 2 4.10.3 WELDOFLANGE Sketch Validity limits k R/T < 50 1.0 , _Fienge afm < 0.50 ia t (see Note’? for re value) i | For d, > 114,3mm then r, > 0.17 | but <0.5T pt For d, 114.3mm then ry is not oe Branch pipe required ‘inforcement iy can be divided by 2.0 if rz = “Te/2 oF T/2 whichever is the larger If r2 is unknown then rz shall be taken as Imm. R. ry Is not less than the larger of " 0.0020d, or 2(sind)°x (Fp - (1 + Main pipe T) To= Ty OF Ty (520 Not# 1 below for Ty valve) ‘SUSTAINED SCF te i, i TYPE Pressure Bending Thrust [Torsion Never less than 0 (See Note-3) Main pipe 1.0 ovsed"e)" 1.0 10 | 10 Branch pipe 1.0 1.0 10 | 1.0 i i i Pressure Bending Thrust |Torsion| shear Waver less than'i.2 | Never less than 1.5 (See Note 3) main pipe |.(%) "(g) [Ey ( 10 | 20 Branch pipe |, (0. )"(4,)"/ 7)". 10 | 20 Mela EG Notes: 1. IL, 2 O.5{aT, then re can be taken as the redius to the centre of Tp and Ty wT, atheriise re n Which Is the rads tothe centre of Tp and Tr = T's 2 Wgfly = 12 tien fly = 12. 3. THe photic ie Ge’ appendc 3) never iss than 1.0 for sustaned or 15 for eee 4. Ty swell thickness of branch reinforcement. TABLE 14 - WELDOFLANGE SCFs. 55 (© IGEM , 12 York Gate, London, NWI 4QG. Website: yrww.igem.org.ukIGE 10/12 Editon 2 A4.10.4 FABRICATED TEE Sketch Validity k Ral < 50 1.0 qT ar tY/Ry < 0.50 (See Note J for r, value) For dy > 114.3mm then Branch pipe ts: > 0.17 but <0.5T : For dy < 114.3mm then r; is not | required |, can be divided by 2.0 If r, > | ‘Tp/2 or T/2 whichever is the larger Tf m2 is unknown then r2 shall be taken as imm Ty= Ty OF To Main pipe (See Note 1 for Ty vatue) SUSTAINED SCF ip iy ig is TYPE Pressure Bending Thrust Torsion| Shear Hvar less than 1.0 (See Note 3) Main pipe 1.0 1.0 1.0 Branch pipe 1.0 1.0 1.0 i, ig is Pressure Torsion] Shear Waver less than Lz a Main pipe fe) Gy 10 | 20 vey Branch pipe |1.4{%] (é] (e) J 1.0 2.0 Notes: 1. IL, 2re O8,f;Tp . can be taken as the rads to the centre of Ts and Ty = Ts, otherwise re ro hich 5 the radius to the centre of T's and To = 2 try 2 ea. 3. ‘me product fw’ (eee Appendix 2) is never fees than 4.0 fer sustained or 2.5 for cyclic assessments Th = wall thickness of branch reinforcement. ‘TABLE 15 - FABRICATED TEE SCFs 56 © IGEM , 12 York Gate, London, NWW1 4QG. Website: www.igem.org.ukIGE TD/12 Edition 2 A4dd OTHER BRANCH CONNECTIONS This section details the SCFs and flexibility factors for branch connections not covered by A4.10 or not meeting the geometrical limitations defined in A4.10. The branch connection types covered in this section include the following: © 4.111 Weldolet © 4.11.2 Long weld neck flange © 4.11.3 Weldoflange © 4.114 — Fabricated tee © 4.11.5 — Forged tee to BS 1640 @ 4.11.6 — Drawn/extruded and forged tee not to BS 1640 © 4.11.7 Sweepolet and welded-in contour insert © 4.11.8 — Full encirclement tee © 4.11.9 Fabricated tee with pad or saddle. Where SCFs for 2 particular branch connection are not detailed in this sub- appendix, experimental testing or theoretical analysis using FEA can be used to generate this data. Consider all possible loadings and include the results in the Stress Analysis Report. ‘The SCFs for the extruded tee are higher than those for the forged tee. Hence, if the analyst is unsure whether the tee is forged or extruded, or if the crotch thickness requirements of the forged tee cannot be met, the SCFs for the extruded tee would be used. 57 © IGEM , 12 York Gate, London, NWA 4Q6, Website: wwi..gem.org.ukIGE TD/1? Edition 2 Agii.i WELDOLET Sketch B k 1.0 Main pipe ‘SUSTAINED SCF ~| 7; ip 7 i i TYPE Pressure Bending Thrust [Torsion] Shear Wever tess than 70 (See Notes) : 18 Main pipe 1.0 B 1.0 10 | 10 : 18 Branch pipe 1.0 we 10 10 | 10 i, i Hrorsion] shear Wain,pipe 10 | 20 For dD, > 0. h Branch pipe | For d/D,< 0.7 aE “ 8 aig For d/D, > 0.7 A 10 | 20 arf 4, ete] areas 1. Far branch connections in the range Of 0.5 d/De < 1.0, multinyie by 1.5 to compensate ‘or possible non-conservative out-of-plane SCF 2 The praduer nw (ase Spent 3) is never less than 1.0 for sustained or 3.5 for cya assessments. TABLE 16 - WELDOLET (OTHERS) SCFs 58 © IGEM , 12 York Gate, London, nW1 496. Website: www. iger. org.ukIGE TO/12 Edition 2 A4AL2 LONG WELD NECK FLANGE Sketch h k Flange 332. 1.0 cope - Branch pipe : Main pipe Re ‘SUSTAINED SCF iy ig ia i TYPE Pressure Bending Thrust —_|Torsion| Shear aver fess than LO (See Notes) Ls ° Main phy 1.0 “a 1.0 1 1.0 pipe ne 18 Branch pipe 10 1.0 10 | 10 ‘CYCLIC SCF ip ip ik Va is Pressure Bending Thrust Torsion] Shear Wéver less than £5 (See Notes) Main pipe For d/>, 20.7 ] 1 ve a 1.2) hTe/2e) T 1.0 2.0 For d/D, > 0.7 Wf “8 @) ela] Branch pipe For d//D, < 0.7 18 1.0 2.0 For d/D > 0.7 me aiztal] DL ote: fe For branch connections in the range of 0.5 s d/Ds s 1.0, mutiply is by 4.5 to compensate for possible non-conservative aut-of plane SCF. 2 The pracct Inw (see appendi« 3} is never ess than £.0 for sustained or 1.5 for eva assesoments TABLE 17 - LONG WELD NECK FLANGE (OTHERS) SCFs 59 @ IGEM , 12 Yerk Gate, London, NW 496. Website: www.igem.org.ukGE TD/12 Edition 2 A611.3 WELDOFLANGE Sketch i K Flonoe 1.0 | Branch pipe ! T : Main pipe ‘SUSTAINED SCF i i, TYPE Bending Thrust —_{Torsion!| [Never fess than 1.0] (See Notes) Main pipe 1.0 10 | 10 Branch pipe 1.0 10 | 10 i Pressure |Torsion| Shear Main pipe For d/0, 30.7 aztal]™ 1.8 1.0 2.0 |Branch pipe For d/D, < 0.7 eell” ts im 1.0 2.0 For d/D, > 0." he st Notes: a a For branch connections in the range of 0.5 + de/Dz < 1.0, muitily f by 1.5 to compensate far possible non-conservative out-of-plane SCF. The product i.W (See Appendix 3) Is never less than 1.0 for sustained or 1.5 for cyclic assessments, TABLE 18 - WELDOFLANGE (OTHERS) SCFs 60 © IGEM , 12 York Gate, London, NWI 4QG. Website: wrens gem. org.ukIGE TD/12 Ealtion 2 A414 FABRICATED TEE Sketch h Kk i Tr 1.0 | Rin i | I | | | | 1 i Mein pipe SUSTAINED SCF 7 iy i i TYPE Pressure Bending Thrust |Torsion| Shear (Never less than 7.0 (Soe Note 1) 1.8 Main pipe 1.0 wh 4.0 10 | 10 18 Branch pipe 1.0 w 1.0 10 | 10 cyclic Scr ip Ty ie i, iy Pressure Bending Thrust Torsion! Shear Naver fess than 23 (See Note 1) [Main pipe |For d/D, = 0.7 ell) Ls sn 1.29272 (2a) | 1.0 2.0 For d/D; > 0.7 wa AG) petal] Branch pipe For d/d,<0.7 5 dela it” air 1s z 10 | 20 For d/D, > 0.7 h Note: The product inv (se Appendix 3) fs never less than 1.0 for sustained or 1.5 for oyclic assesements, TABLE 19 - FABRICATED TEE SCFs 6t © IGEM , 12 York Gate, London, NW 4QG. Website: werw.igem.org.ukIGE TH/22 Edition 2 Ad115 FORGED TEE TO BS 1640 Sketch Validity limits [kc re20.125 dx | 1.0 TE LST) (See Note 1) ‘SUSTAINED SCF ip ie ny iy a TYPE Pressure Bending Thrust | Torsion | Shear ‘Naver fess than 2.0} (See Note 2) L Main pipe 1.0 10 | 4.0 Branch pipe 1.0 1.0 | 1.0 Pressure Torsion | Shear [Never fess than 2.0 (See Note 2) Main pipe For 0/0, < 0.7 ziel] a 0.67 = For d/D, > 0.7 (=, aziel] Branch pipe For d/D; < 0.7 feria yh zit] For d/D; > 0.7 isl” 10 | 20 osfta) joefsgil| 20 | 29 Notes: 1. Tpis the pipe wal! thickness not the tee thickness. 2. The product fuw (see Appendix 3) is never less than 1,0 for sustained or 2.0 for cyclic assessments, TABLE 20 - FORGED TEE TO BS 1640 SCFs 62 © IGEM , 12 York Gate, London, NWI 4Q6. Website: www.igem.org.ukIGE TD/12 Edition 2 AGILE DRAWN/EXTRUDED AND FORGED TEE NOT TO BS 1640 ‘Sketch Validity fn Kk Himits 0.05¢q¢te < | TT) 10 (O.1do +12.5) | RTO * Re Ly 2 0.707, Led, SUSTAINED SCF ie i TYPE Pressure Bending Thrust Shear Naver loss than 70 (See Note Main pipe 1.0 1.0 1.0 1.0 Branch pipe 1.0 1.0 10 | 10 i k iy i Pressure Bending Thrust —_|Torsion| Shear Never less than 15) (See Nove) Main pipe For d/D, < 0.7 eal] La 10 | 20 For d/D, > 0.7 we ‘ ° Branch pipe For d/D, s 0.7 asf? se] r ela] Ls 10 | 20 For d/D, > 0.7 ne alae etal] Note: The product iw (sea Appendix 3) is never less then 1.9 for sustained oF 1.5 for cyclic. assessments, ‘TABLE 21 - DRAWN/EXTRUDED AND FORGED TEE NOT TO BS 1640 SCFs 63 @IGEM , 12 York Gate, London, RWI 496. Website: wwn.igem.crg.ukIGE TO/L2 Edition 2 4.11.7 SWEEPOLET & WELDED-IN CONTOUR INSERT Sketch Validity its A k . r re 20.125 do 1.0 ; vr T.2 LST al to * te SUSTAINED SCF i, iy iy ig is TYPE Pressure Bending Thrust Torsion] Shear Wavar fess than 1.0) ‘| {See Notes) an of 18 Main pipe 1.0 na 1.0 1.0 1.0 ; czy Branch pipe 1.0 ne 1.0 1.0 1.0 CYCLIC SCF ip i iy iy Pressure 9 Thrust Torsion| Shear Maver tess than £5 (See Notes) Main pipe 18 4 1.0 2.0 Branch pipe | For d/D,<0.7 a fe o i saa] 20 | 20 For d/D; > 0.7 of) Notes: 1 ran branch connections inthe range of 0.5 x dD 1.0, multiply ly by 1.5 to compensate for possible non-conservative outatpane SCP 2. The prodiet inv cee Appendbe 3) & never fess than 1.0 for sustined or 1.5 tor eyelc TABLE 22 - SWEEPOLET & WELDED-IN CONTOUR INSERT SCFs 64 © IGEM , 12 York Gate, London, NW 4Q6. Website: www.iger.org.ukIGE TD/12 Edttion 2 ASALS FULL ENCIRCLEMENT TEE ht k cr-ost,)% | 1° Ra(T =Te) (See Mots 1) I SUSTAINED SCF ib ty te iy B TYPE Pressure Bending Thrust |Torsion| Shear Tavares tho 8 “ee Note 2) imps 18 Main pipe 10 10 10 | 10 18 Branch pipe 1.0 We 1.0 1.0 1.0 eveLie SCF i, ip i ig iy Pressure Bending Thrust Torsion) Shear Neverlass hag tS “Gee Note) Main pipe For d/D, 30.7 cai rls Ls sey] | 20 | 20 For d/D, > 0.7 wa ore 2f2fayy” dels] Branch pipe For di/D) < 0. 18 a 10 | 20 For d/D, > 0.7 W 2g 2T] oe | Esl] Notes: 1 soe sum of the man poe and encclement thciress 2 The pode! ace Spends 9) 5 eve” less a for sustoined or 1.5 fer ey essssment, ‘TABLE 23 - FULL ENCIRCLEMENT TEE SCFs 65 (© 1GEM, 12 York Gate, London, NWI 496. Website: www.igem org.ukA4.11.9 FABRICATED TEE WITH PAD OR SADDLE ‘Sketch h k 1.0 le |Torsion |Shear Main pipe [1.0 10 [Lo Branch pipe [1.0 ie 1.0 re) 1.0 ‘CveLie SCF i, ie le ig le lPressure IBenaing Thrust ‘Torsion |Shear Inever fess than 1.5 (See Note 2) For d/D, 3 0.7 pater i o 2.0 Main pipe ror dy, > 0.7 h " For d/D, = 0.7 nga aly” fel, . Branch pipe Jror d/D, > 0.7 ne ote: The product ia.w (S02 Appendix 3) is never less than 1.0 for sustained or 1.5 for cyclic assessments TABLE 24 - FABRICATED TEE WITH PAD OR SADDLE SCFs 66 © IGEM , 12 York Gate, London, NW1 4QG. Website: wwv.igem org.ukIGE TD/32 Edition 2 APPENDIX 5 : FATIGUE ANALYSIS ASL INTRODUCTION ASL Changes In pressure, temperature and externalty applied loads will cause corresponding changes in stress. Fatigue cracking can arise under such conditions. For this reason, it is necessary to ensure that all pipework systems have adequate fatigue lives. A512 High frequency vibration-induced fatigue may occur in a piping system, particularly when the natural frequency of the system matches the source frequency. This problem is not addressed in the following, therefore specialist advice is needed where it is considered that it may occur. ASL3 Welded joints are particularly sensitive to fatique and guidance Is given for Welds that are most commonly encountered in gas industry plant. In order to assess the fatigue life of the pipework and its components, it is necessary to identify, for the anticipated design life, all variations in pressure, temperature and load which aive rise to ranges of principal stress, Sa, calculated from Appendix 3. AS.2 FATIGUE CLASSES A524 Fatigue classes are provided based on the behaviour for welds that are most commonly encountered in gas industry plant. These fatigue classes are identified in a consistent manner with that adopted by PD 5500 and BS 7608. This is to ensure that the source of the date remains transparent and capable of being re~ appraised to confirm the validity of the recommended fatigue classes in the event of any non-standard applications arising. The application of the fatigue classes to existing components is shown in Table 25. COMPONENT FATIGUE | NOTES CLASS ‘Straight welded pipe (flush) Class D | See Appendix 4 for definition ‘as-welded) Class & _| of fush and as-welded. Wall thickness transitions (flush) Class | See Appendix 4 for definition (as-welded) {Class €__| of flush and as-welded, ‘Concentric and eccentric reducers Class E Bends and welding elbows (seam welded) | Class E (seamless) __| Class D Witre bends (single, widely or closely spaced) Class & Weldolet Class F Long weld neck flange: Class F Weldoftange Class F Fabricated tee ‘Class F Forged tee to BS 1640 Glass F Drawn/Extruded and forged tee nat to BS 1640 Class F ‘Sweepolet and welded-in contour insert | Class F Full encirclement tee Class F Pad or saddle reinforced tee Class F ‘TABLE 25 - FATIGUE CLASSES 5.2.2 For partial penetration joints, such as the filtet welds on full encirclement tees, assume possible failure from the weld root as Class W. 5.2.3 For components other than those listed in Table 25, or for partial penetration Joints, assume the fatigue class as Class G unless it can be demonstrated by reference to PD 5500 or BS 7608 that another, less onerous, fatigue class is appropriate. 67 GIGEM , 12 York Gate, London, NWI 4QG, Website: wiv. igem.ora.ukwoe t/t cow A524 A5.3 AS The fatigue design curve equations assume a lognormal distribution and are defined as belng two stendard deviations below the mean curve fitted to the test data. This infers a nominal probability of failure of 2.3% and is consistent with the application of PD 5500 and BS 7608. The fatigue design curves are represented in Table 26 and Figure 2. ‘The analysis also recognises that the fatigue strength of welded joints is, to some extent, dependent on plate thickness. Guidance is given in A5.5 for calculating the fatigue behaviour of materials thicker than 22 mm. EXCLUSION FROM FATIGUE ANALYSIS In order to assess the fatigue life of the pipework and its components, it is necessary to identify all variations, including excursions, in pressure, temperature and loads in the design life and calculate that Ss that they produce. Carry out a detailed fatigue analysis in accordance with AS.4 if the conditions of clause 8.5.7 are not satisfied, FATIGUE ANALYSIS ‘The S-N curve shown in Figure 1 extends below the stress range values quoted in Table 5, This is required for systems subjected to 2 number of different stress ranges where some ranges exceed the values quoted in Table 5 and some are below. For such situations, the S-N curve is assumed to extrapolate indefinitely into the high ¢ycle regime but at a reduced slope after 10” cycles, as illustrated in Figure 1. Likewise, the S-N curves shown in Figure 2 extend back to a number of stress cycles less than those quoted in Table 5. This is required for systems which may Include some components subjected to localised regions of high stress range but are judged safe from incremental collapse even though the shakedown criteria for efastic action are not met at these highly localised regions. In these instances, seek specialist advice because the S-N curve will become non-linear and material-dependent. Methods for deriving elastic-plastic correction factors are available when shakedown to clastic-piastic action occurs at localised regions, Empirical equations are provided In Annex C of PD 5500. Alternatively, limit load analysis methods, as described in A6.7, can be used. For pipework components subjected to changes in stress and cycles in excess of the threshold values of Table 5, the maximum permitted number of cycles N for a correspanding Sq can be calculated using the constants from Table 26 in the folowing equation: N=A/S,” N = number of cycles ‘Sa= maximum principe stress range (N/mm?) Aland m are constants from Table 26. 68 © IGEM , 12 York Gate, Londen, KWL 4Q6. Website: www. igern.org.ukc1GE TD/12 Edition 2 Class Constants of S-N curve Stress range at (for E=209x 10° N/mm?) N =10’cycles ForN=10’cycles | For N > 10’cycles (N/mm?) ™ A ™ A DB 3.0 152x107 |'5.0 4.1810" [53 N/mm? E 3.0 1.04x107 [5.0 2.29x10°° | 47 N/mm? F 3.0 6.33x10" [5.0 1.02x10* | 40 N/mm G 3.0 2.50x10" | 5.0 2.05x10™ [29 N/mm’ Ww 3.0 1.58x10" [5.0 9.77x10 | 25 N/mm’ ‘TABLE 26 - EQUATION CONSTANTS FOR S-N FATIGUE DESIGN CURVES 1N Fatigue Curvas for €=209000 Nin? 1000 100 ie Stross Range, $ , (Mmm?) > i 0 1502 1603 LE 105 1S ENOL EKS Number of Cycles, N FIGURE 2 - S-N FATIGUE CURVES FOR E=209 x 10° N/mm? 69 WIGEM , 12 York Gate, London, NWI 4QG. Website: www.lgem.ora.ukIGE TO/12 Edition 2 ASS ASS CORRECTIONS FOR PLATE THICKNESS ON FATIGUE LIFE In pipework and components with thicknesses greater than 22 mm, a worsening of fatigue performance with increasing thickness is to be expected, and corrections to the principal stress range, S, or life, N, need to be made when fatigue calculations are carried out. For a life of N cycles, reduce the maximum allowable principal stress range Sa, calculated from the equation in AS.4 and constants in Table 26, to stress range Spr according to the equation: 05 Sar = Se 2] ty ‘Ser the thickness corrected stress range for materials thicker than 22 mm 1b = nominal thickness (rm), elther tor T as appropriate for Fatigue conditions Alternatively, the maximum number of permitted cycles Nr for material thicker than 22 mm with stress range, Sp, can be calculated from N and the following ‘equations where N is the maximum number of permitted cycles for stress range, Se, from the equation in A5.4 and constants in Table 26: 075 w= [2 for N < 10 125 wan [2] for N > 107 Values of correction factors for thicknesses ranging from 22 to 100 mm are given in Table 27. THICKNESS t, | FACTOR ON STRESS | FACTOR ON LIFE (mm) Ns 10’ N> 10; Up to 22 1.0 1.0 10 22 to 32 0.91 0.75 0.63 32 to 45 0.83 0.58 0.41 45 to 60 0.78 0.47 0.29. 60 to 80 0.72 0.38 0.20 ‘80 to 100 0.68. 0.32 0.15 TABLE 27 - CORRECTION FACTORS FOR Sx AND Ny CORRECTIONS FOR MODULUS OF ELASTICITY ON FATIGUE LIFE ‘The fatigue design curves are applicable for 2 modulus of elasticity (E) equal to 209 x 10? n/mm?. Where the grade of carbon or austenitic steel used is known to have a different modulus of elasticity, make corrections to the principal stress range, Sa of life, N, when fatigue calculations are carried aut. These corrections are additional to those in AS.5, if applicable. For a life of N cycles, correct the maximum allowable principal stress range, Sa, calculated from the equation in A5.4 and constants in Table 26, and possibly correct to Sar in AS.5, £0 stress range Spe according to the equation: ba=S_ ‘Sar = the modulus of elasticty corrected stress range: E'= of elasticity (N/mm?) at the maximum operating temperature 70 © IGEM , 12 York Gate, London, NWI 4Q6. Website: vwirger.org.ukIGE 10/12 Edttion 2 AS7 ASB AS.B.1 AS.8.2 Altematively the maximum number of permitted cycles, Ne, can be calculated from N, or Nr if corrected in AS.5 earlier, and the following equations where N (or No) is the maximum number of permitted cycles for stress range S, (or S.) from the equation in A5.4 and constants in Table 26: a a forN < 10” 208 x 10 s e - a for N > 30" 209 x10 CUMULATIVE FATIGUE DAMAGE For pipework and components subjected to repeated changes of stress with differing stress ranges, assess the cumulative fatique damage. Where the stress range pattern is complex, the use of the reservoir method may assist (see BS 7608). For Sg, establish the number of permitted cycles N using the relevant fatigue design curve constants from Table 26 with the equation from AS.4. Make a correction for thickness t as described in A5.5 if appropriate. Evaluate the cumulative fatigue damage using the number of cycle repetitions (n.) for each principal stress range as follows: Fatigue damage = n/N; + 12/Nz + na/Ns + wn. + Y/N A fatigue damage summation of unity or less is acceptable. Where the summation exceeds unity, the following alternative actions, in no particular order of priority, can be considered: © revisit the assessment to see if any simplifying pessimistic assumptions can be modified, which will enable the values of Sq to be reduced (see Section 9). * use DBA (Appendix 6) to enable more accurate values of Sa to be evaluated. ‘* modify the component design to enable values of Sq to be reduced. BRANCH CONNECTION CROTCH CORNER FATIGUE DAMAGE For set-on welded branch connections, the timiting location for fatigue is assumed to be the weld toe in the header pipe. However, where the load case table is dominated by pressure cycling, rather than cycles due to system forces and moments, the crotch corner location will usually become more limiting for fatigue than the header weld. The crotch corner location is not of concern for failure modes other than fatigue such as plastic collapse (due to sustained loading) or progressive distortion (due to exceeding elastic shakedown limits). Hence, the SCFs specified in Appendix 4 are for the weld toe, in the header pipe, where’ all the above failure modes are credible, See Figure 3 for basic nomenclature of set-on branch connections, Taking into consideration the different fatigue design curves for crotch corner parent material (Class D) and header weld material (Class F), providing the Cumulative fatigue damage of the weld toe in the header pipe does not exceed 0,25, it can be assumed that the crotch comer fatigue damage Is also below unity. This conservatively assumes that the header weld stress range is cycling at the elastic shakedown limit and that plasticity correction factors are applied to the crotch corner focation. na © IGEM , 12 York Gate, London, NWI 4QG. Website: www.igem.crg.ukIGE TD/12 Edition 2 A583 If the header weld fatigue usage factor is predicted to exceed 0.15, it is recommended that a crotch corner fatigue analysis is undertaken using stresses evaluated by FEA. A584 In the case of weldolets, long weld neck flanges, weldoflanges and fabricated tees, if the SCFs used are from A4.10, the value of 0.15 is only applicable when stresses have been evaluated assuming the default weld radius of rp = 1.0 mm or less. If stresses are being evaluated based on a weld radius greater than 1.0. mm, it is recommended that a separate crotch corner fatigue analysis is undertaken using stresses evaluated by FEA. Branch Pipe Branch Weld Header Weld Weld Toe Crotch Corner. i Header Pipe I I FIGURE 3 - BASIC NOMENCLATURE OF SET-ON BRANCH CONNECTIONS 72 © IGEM , 12 York Gate, London, NWI 4Q6. Website: wwrrigem.org.ucIGE TD/12 Edition 2 APPENDIX 6 : DESIGN-BY-ANALYSIS (DBA) METHOD Specialist advice an DBA will be required. IGEM's website provides details of Consultants who may be able to provide this advice. AG.1 62 A6.3 INTRODUCTION ‘Any pipework component which does not meet the acceptance limits in Sub- Section 8.5, when using the IGE/TD/12 Edition 2 simplified DBA method, can be assessed using alternative calculation methods. One alternative is to use a more detailed DBA method such as prescribed in traditional pressure vessel design codes. The IGE/TD/12 Edition 2 simplified method is, henceforth, referred to as design by formula" so as to avoid confusion with the tracitionai pressure vessel DBA methods. ‘These alternative, and more established, DBA methods are described in a selection of well-known literature (see A2.5) where further detail can be found. The following sections do, however, provide a brief overview of some of the key aspects associated with DBA. One of the main areas of conservatism in the “design by formula" method is that it assumes that the SCFs due to the individual loadings of pressure, bending, thrust, torsion and shear are all co-incident. Alternative modelling techniques, such as FEA, enable this conservatism to be removed. However, this introduces additional complexity to the assessment through the use of DBA. DESIGN-BY-ANALYSIS APPROACH ‘An acceptable DBA assessment may be achieved by use of: © a recognised FE code provided that the modelling and solution procedures adhere to the guidelines set out by recognised international organisations such as NAFEMS or * a database of known stress intensities provided the geometric parameters lie within the range of applicabitity. Ensure the stress intensities obtained from the DBA do not exceed the acceptance criteria given by A6.5 for the appropriate load combinations considered. To undertake a DBA assessment, extract the forces and moments from the pipe stress model for the relevant stress exception with due care being taken to identify the axial, in-plane and out of plane directions, Details of the component geometry are required for a FE model to be generated. For example, cetalled information such ag the crotch radii (on branch connections) is erucial since the predicted stresses are particularly sensitive to this parameter. Nevertheless, when component details are not available, sensitivity studies can still be undertaken, using various FE models, to establish bounding values. Where components are declared fit-for-purpose using DBA, include in the DBA documentation all details relating to the geometry, FE mesh, boundary conditions, loadings, material properties and the DBA assessment. Keep this DBA documentation with the Stress Analysis Report. CONDITIONS ‘The stresses obtained from the alternative modelling techniques need to be post-processed so it can be shown that the component will: © not collapse (the sustained loading assessment) © not progressively distort (the shakedown assessment) 73 @IGEM , 12 York Gate, London, NW 4QG. Website: www.igem.org uleIGE TO/1? Eaition 2 A684 AGAL 46.4.2 6.4.3 6.4.4 6.4.5 16.4.6 A647 AGAB © not fe due to cyclic loading (the fatigue assessment). TERMINOLOGY The following terminology is used in DBA (see also PD 5500 Annex A), Stress intensity The stress intensity is twice the maximum shear stress, i.e. the difference between the algebraically largest principal stress and the algebraically smaltest principal stress at a given point. Tensile stresses are considered positive and compressive stresses are considered negative. Gross structural discontinuity A gross structural discontinuity is a source of stress or strain intensification that affects a relatively targe portion of a structure and has a significant effect on the ‘overall stress or strain pattern or on the structure as a whole. Examples of gross structural discontinuities are nozzles and branch connections. Local structural discontinuity A local structural discontinuity is 2 source of stress or strain intensification that affects a relatively small volume of material and does not have a significant effect on the overall stress or strain pattern or on the structure as a whole. Examples of local structural discontinuities are small fillet radii, small attachments and partial penetration welds. Normal stress The normal stress is the component of stress normal to the plane of reference - also referred to as direct stress. Membrane stress The membrane stress is a component of the normal stress that is uniformly distributed and equal to the average value of stress across the thickness of the section under consideration. Bending stress The bending stress is the component of the normal stress that is proportional to the distance from the centre of the wall thickness. Shear stress ‘The shear stress is the component of stress acting in the plane of reference. Primary stress A primary stress is a stress produced by mechanical loadings and so distributed in the structure such that no redistribution of load occurs as a result of yielding. It is @ normal or a shear stress developed by the imposed loading that is necessary to satisfy the simple laws of equilibrium. The basic characteristic of this stress is that it Is not self-limiting. Primary stresses that considerably exceed yield strength will result in feilure, or at least in gross distortion. ‘Thermal expansion-induced stresses that involve “elastic follow up", such that only a small redistribution of load occurs as a result of yielding, can also be considered as primary stresses. Primary stress is divided into ‘general’ and “lecal’ categories. An example of 2 general primary stress is the stress in a 74 @ TGEM, 12 York Gate, London, NWT 4QG. Website: ww. igem.org.ukIGE TD/12 Ecition 2 A649 6.4.10 A6.5 6.5.1 cylindrical shell due to internal pressure. An example of a primary local stress is the membrane stress in a shell produced by external load and moment at a permanent support or at a nozzle connection. The lacat category of stress is considered to have some characteristics of a secondary stress but is classified as being primary to be conservative. Secondary stress A secondary stress is a normal or shear stress developed by the constraint of adjacent parts or by self-constraint of a structure. The basic characteristic of 2 secondary stress is that it is self-limiting. Local yielding and minor distortions can satisfy the conditions that cause the stress to occur. Failure from one application of the stress is not to be expected. An example of a secondary stress Is the bending stress at a gross structural discontinuity. Peak stress The basic characteristic of a peak stress is that it does not cause any noticeable distortion and is objectionable only as a possible source of @ fatigue crack or @ brittle fracture. Peak stress does not need to be considered in a fatigue assessment if the fatigue design curves already allow for peak stress effects. The fatigue design curves in Appendix 5 do not require peak stress to be considered. STRESS CATEGORIES, FAILURE MODES AND LIMITS: A calculated stress depending upon the type of loading and/or the distribution of the stress will fall within one of the following four basic stress categories. Ensure the stress intensities derived fall within the limit detailed for each category to demonstrate an acceptable DBA, Note: The stress latensity Hits detalled in this Sub-Appendl are applicable for carbon, carbon manganese and {ow alloy steels up to and including 50°C. For other materials, end temperatures in excess of 50°C, detalied guidance fs available in Annex & znd Annex K oF PD 5500. General primary membrane (Pm) The principal failure mode is ‘collapse’ In the sense that collapse includes tensile instability and ductile rupture (under short term loading). ‘The limiting value of the stress intensity for this category is: sx SMYS. or sx (SMYS + UTS) ‘Is. constant depending on the design factor and the ratio of SHYS/NTS. The values of s are specified in Table 3, reproduced below. NORMAL (S.) ABNORMAL (Szan1, Seana... Seavi) ‘SMYS SMYS ‘SMYS ‘sMYS <0.74 | SWS 5, <0 > 0. Te <0. er o.7d 0.74 | HOS > 0.74 0.80 SMYS__| 0.34 (SMYS +UTS) | 0.9: 0.60 SMYS | 0.25 (SMYS +UTS) | 0.6: .30 | 0.36 SMYS__| 0.15 (SMYS +UTS) | 0.4 SMYS | 0.28 (SMYS +UTS) sys __| 0.17 (SMYS +UTS) lasol— is in | ERS) S 0 SMYS | 0.38 (SMYS +UTS) 7 0 Note: SMYS and UTS values are obiained at the maximum design tempersture or, if known, the ‘maximum temperature appropriate to the sustained load case being analysed. 75 @ IGEM , 42 York Gate, London, NW 4Q6. Website: wewrigem.orguleIGE TD/22 Edition 2 AG.5.2 76.5.3 6.5.4 A6.6 Local primary membrane (P,) The failure modes for P,, also apply to P. The difference is that P, has some characteristics of a secondary stress and, therefore, this Is recognised in the limiting valve being set higher than that for Pj. The limiting value of the stress intensity for this category is the lower of SMYS or 1.5(UTS/2.35). Note: SMYS and USS values are obtained at the maximum design temperature or, If known, the maximum temperature appropriate ta the sustained load case being analysed. Primary membrane plus primary bending (Pm + Ps oF Pi + Po) The failure modes for Py also apply to Py. The difference is that through thickness ylelding (shape factor effect) requires a stress higher than SMYS for collapse to occur. Hence, the principal failure mode for primary membrane plus primary bending is excessive plastic deformation, The limiting value of the stress intensity for this category is the lower of SMYS or 1.5(UTS/2.35). Mote: SMYS and UTS values are obtained at the maximum design temperature or, if known, the maximum temperature appropriate to the sustained load case being analysed. Primary plus secondary (P + Q) ‘The limit on primary plus secondary assures that the stress will shakedown to an elastic cycle. This is fundamental to assuring that the elastic fatigue analysis, hhas valid stresses. it also precludes significant plastic strain ratchet. The limiting value of the stress intensity for this category is the lower of 2.0 SMYS or 3.0(UTS/2.35) for both carbon and austenitic steels, The shakedown factor of 2.0 is considered acceptable when using DBA, since the failure criteria adopted Is based upon Tresca maximum shear stress theory. This failure criterion, for predicting the onset of yielding In a complex system of principal stresses, is conservative compared to the von Mises shear strain eneray failure criterion for which a shakedown factor of 1.8 is used in the design by formula sections for carbon steel, Table 4. Use of elther option is judged to be similar in net terms. Note: SMYS and UTS vaiues are obtained at the maximum design tempersture or, IF known, the maximum temperatures appropriate to the load case being analysed. CALCULATION OF STRESS INTENSITIES: ‘The results of FEA should be analysed using Stress Classification Lines (SCLs). ‘The use of SCLs Is described in WRC 429 which provides guidance on selecting valid SCLs and on the linearisation of stresses when they are constructed. The membrane and bending stresses are calculated from the component stresses, not principal stresses. Local non-linearities caused by stress-raising features can be included at an SCL location and an SCL may originate or end at a singularity, Singularities in @ FE model preclude accurate stresses at the singularity location but they, or stress-raising features, do not invalidate the linearisation. Principal stresses have to be calculated using linearised component stresses. Hence, stress differences and stress intensities can be computed. Calculation of stress ranges at the component stress level eliminates the need to check for rotating principal stress directions. The use of SCls, for evaluation of membrane and bending stresses, is appropriate for most geometries but was originally devised for use with 2D axisymmetric models. Some particular geometries are conducive to use of a Stress Classification Plane (SCP). These are appropriate for many plate-type components, such as external loads on lugs. SCLs can be conservative when used with 3D models. For exemple, an SCL taken through the crotch corner of a 76 © IGEM , 12 York Gate, London, NWI 4QG. Website: www.igem.org.uk