Guidance on the Use, Specification, and Anomaly Assessment of Modern Linepipes

Transcription

1 Guidance on the Use, Specification, and Anomaly Assessment of Modern Linepipes Yong-Yi Wang and Dan Jia Center for Reliable Energy Systems 5858 Innovation Dr. Dublin, OH October 24, 2018

2 Overview Scope and objectives Observations and preliminary guidance Basis for those observations Incident analysis Pipe MTRs Test of specially made welds Tensile strain capacity analysis Guidance Near term vs. long term Difficult vs. more difficult Path forward No time for details and substantial supporting information 2

3 Scope and Objectives Dealing with two major issues Unexpected girth weld incidents in newly constructed pipelines Potential issues related to the tolerance to mechanical damage in new linepipes Solutions from multiple angles Linepipe specifications Testing and criteria for test data Welding options Welding procedure qualification requirements Reduction of loads/stresses on girth welds (not a focus within MATH-5-3B) Operation phases Pipelines already in service Near-term new pipeline projects Pipe replacement projects New pipelines to be built in the future 3

4 Girth Weld Incidents ~10 known incidents in last 2-3 years in US Most were in-service failures Some were hydrostatic failures (leaks) Pipes ERW pipes (12, 20 and 24 OD) SAWH (spiral pipes) 30+ OD Grades X52, X65 (small to medium diameter ERW) Most were X70 (medium to large diameter, ERW and SAWH) One X70 and one X80 (transition weld) Welding all manual SMAW: E6010 root, E8010 fill and cap passes X70/X80: SMAW/E6010 root, FCAW fill and cap passes 4

5 Failure Mechanism Technical Representative hardness values Pipe: 235 Hv Root pass: 165 Hv (70% of pipe hardness) Fill pass: 205 Hv (87% of pipe hardness) HAZ: 185 Hv (79% of pipe hardness) Contour of plastic strain. Note that the plastic strain in the HAZ is much higher than that in the pipe and upper fill passes. 5

6 Non-contributors Contributors to the Incidents Linepipes met the requirements of API 5L or equivalent standards Welding procedures were qualified to API 1104 Field inspection followed industry practice/standards In most cases, no or very small weld flaws Minimum amount of high-low misalignment Contributors Weld strength undermatching HAZ softening Soft root Weld bevel (manual SMAW/FCAW) Bending loads from normal ground settlement and other sources 6

7 Weld Strength Undermatching Impact of Linepipe Manual welding processes and consumables did not change Strength of SWAH and ERW pipes progressively moved up relative to earlier SAWL pipes Large range permitted in API 5L Test method systematically under-represents actual yield strength Changes in steel making, test method, and definition of yield strength lead to large variations in reported yield strength of the same joint of pipe Mills have to increase the averaged strength to meet the minimum strength requirements. Industry response to expanded pipes due to understrength pipes PHMSA ADB The negative consequence of higher strength was not well recognized. 7

8 Weld Strength Undermatching Impact of Linepipe Review of linepipe produced recently Strength of SAWH and ERW pipes is 5-6 ksi higher than SAWL pipes of the same grade Examples X70 ERW pipes: YS in lower 90s ksi, UTS as high as 108 ksi X52 ERW pipes: YS in low to mid 80s ksi 8

9 Steel Making and Response to Girth Welding Thermal Cycles Leaner chemistry: move towards lower carbon content and lower hardenability X70 Earlier days SAWL pipes: C: ~0.08%, Pcm: ~ Present day SAWH pipes: C: ~ , and Pcm: ~ Some present day ERW pipes: C: as low as 0.02%, Pcm: as low as 0.12 Greater reliance on rolling and accelerated cooling to gain strength Consequence Greater level of HAZ softening relative to the actual strength of the pipe HAZ strength could be lower, and/or HAZ strength being the same, but greater level of softening due to pipe strength being higher It s the relative strength mismatch that controls the strain concentration once the applied strain is beyond ~0.2%. 9

10 Alternative Welding Options - Welding Matrix Examine the impact of the following paramters Weld cap size Consumable for the root pass Heat input Weld No. Priority Consumables Cap Reinforcement Heat Input Purpose 1 High Group B, E6010/E8010 Regular cap, no extra reinforcement Mid-range, or nominal Baseline, established practice 2 High Group B, E6010/E8010 Regular cap, no extra reinforcement High end 3 High Group B, E6010/E8010 Regular cap, no extra reinforcement Low end Baseline, established practice, high heat input to see to the impact on HAZ softening, to a lesser extent, any impact on weld metal strength Baseline, established practice, low heat input to see to impact on HAZ softening, to a lesser extent, any impact on weld metal strength 4 High Group B, E6010/E8010 Larger cap, with extra reinforcement Mid-range, or nominal Impact of larger cap than Weld No. 1, nominal heat input 5 High Group B, E6010/E8010 Larger cap, with extra reinforcement High end Impact of larger cap than Weld No. 2, high heat input to generate the maximum amount of HAZ softening 6 High Group C, E8010/E8010 Regular cap, no extra reinforcement Mid-range, or nominal Baseline, stronger root than Weld No. 1 7 Medium Group C, E8010/E8010 Larger cap, with extra reinforcement Mid-range, or nominal Impact of larger cap, nominal heat input, stronger root than Weld No. 4 8 Medium Group C, E8010/E8010 Larger cap, with extra reinforcement High end Impact of larger cap, high heat input to generate the maximum amount of HAZ softening, stronger root than Weld No. 5 10

11 Alternative Welding Options - Regular vs. Wider Cap 11

12 Alternative Welding Options Pipe Chemical Composition Pipe chemical composition from MTR of adjacent pipes Pipe ID C Mn P S Si Al Cb V Ti N A B Pipe ID Cr Mo Cu Ni B Ca Sn CE A B Pcm

13 Alternative Welding Options Weld Marcos Weld macros heat input, consumables Weld 1 nominal, E6010/E8010 Weld 2 high, E6010/E8010 Weld 3 low, E6010/E8010 Weld 4 nominal, E6010/E8010 Weld 5 high, E6010/E8010 Weld 6 nominal, E8010/E8010 Weld 7 nominal, E8010/E8010 Weld 8 high, E8010/E8010 Weld 9 nominal, E6010/E

14 Alternative Welding Options - Microhardness Weld 1 Nominal heat input E6010/E8010 consumables 14

15 Alternative Welding Options - Microhardness Weld 1 Nominal heat input E6010/E8010 consumables Weld 4 Nominal heat input E6010/E8010 consumables 15

16 Alternative Welding Options - Microhardness Weld 1 Nominal heat input E6010/E8010 consumables 16

17 Alternative Welding Options - Microhardness Weld 1 Nominal heat input E6010/E8010 consumables Weld 6 Nominal heat input E8010/E8010 consumables 17

18 Alternative Welding Options - Microhardness Weld 1 Nominal heat input E6010/E8010 consumables 18

19 Alternative Welding Options - Microhardness Weld 1 Nominal heat input E6010/E8010 consumables Weld 7 Nominal heat input E8010/E8010 consumables 19

20 Tensile Strain Capacity (TSC) Analysis Manual welds without flaws Parameter Range/Values Number of Cases Pipe wall thickness Thin and medium (3/8'' and 5/8'', & mm) 2 Pipe strength & strain hardening Low, middle and high 9 Weld root consumable E6010 and E Fill/cap consumable E HAZ softening level 10% and 20% 2 HAZ width 2.5 mm and 4.5 mm 2 Bevel geometry Standard 1 Cap reinforcement - width 2.8, 6.4, 9.0 and 12.0 mm beyond the original bevel 4 Cap reinforcement - height 0.8, 1.6, 2.1 and 3.0 mm 4 High-low misalignment 1/16'' 1 Values of all parameters, when possible, were taken from actual test data. About 170 conditions have been analyzed. 20

21 Weld Profile in FE Models Three Cap Width Cap width = 6.4 mm Pipe HAZ Weld HAZ Pipe Root Cap width = 9.0 mm Weld HAZ HAZ Pipe Pipe Root Cap width = 12.0 mm Pipe HAZ Weld HAZ Pipe Root 21

22 Engineering Stress (MPa) Pipe Tensile Properties Three Y/T ratios for each UTS level. Pipe Bound HAZ HAZ Pipe Softening YS UTS YS UTS Y/T % ksi ksi ksi ksi Y/T Lower U, YS = 103 ksi U, YS = 93 ksi U, YS = 82 ksi M, YS = 91 ksi M, YS = 84 ksi M, YS = 75 ksi L, YS = 82 ksi L, YS = 75 ksi L, YS = 70 ksi Engineering Strain (%) Middle Upper

23 TSC (%) TSC Analysis Group 1, WT=3/8 2.5 Bar color Cap reforecement height (mm) Cap reinforcement width (mm) Blue Orange > 2.00 > 2.00 > > > > 2.00 > 2.00 > > % 0.0 Condition No Pipe Strength L L L L L L M M M M M M H H H H H H Pipe Y/T Ratio L L M M H H L L M M H H L L M M H H HAZ Softening (%) HAZ Width (mm) UTS WFC /UTS Pipe YS WFC /YS Pipe Root Strength

24 TSC Without Extra Cap Reinforcement Order of impact on TSC 1: Weld strength mismatch 2: HAZ (level of softening and width) 3: Weld root strength The parameters in lower order can play a more prominent role if the higherorder parameters don t dominate. Impact of pipe strength When pipe strength is in the upper bound, most TSC would not meet the requirement. When pipe strength is in the lower bound, most TSC would meet the requirement. When pipe strength is in the middle, TSC is affected more prominently by parameters other than pipe strength (mismatch). 24

25 TSC Extra Cap Reinforcement Extra cap width is an important parameter. If the undermatching and HAZ softening are moderate, extra cap width is effective. If the undermatching/haz softening level is high, both width and height are needed to compensate for the low strength in the weld area. If HAZ softening level is very high and the width is wide, cap reinforcement may not be sufficient. 25

26 Girth Weld Strain Tolerance Adequate strain tolerance level in most cases: ~0.50% With a safety factor, the target tensile strain capacity: ~0.75% X70 pipe and girth welds Permitted UTS level of PSL 2 pipes: ksi E6010/E8010 consumables If (a) no excessive HAZ softening and (b) pipe UTS < ksi failure outside weld area, good strain tolerance If (a) no excessive HAZ softening and (b) pipe UTS > 100 ksi failure in weld area, poor strain tolerance If (a) no excessive HAZ softening and ksi < pipe UTS < 100 ksi failure in weld area, strain tolerance may still be OK. Excessive HAZ softening without weld strength undermatching Could fail in weld area, leading to poor strain tolerance 26

27 Impact of Yield Strength Yield strength has impact on strength tolerance and failure location. Why is YS not used in initial screening? YS in longitudinal direction is usually not measured. YS in hoop direction from flattened strap systematically under-represent true yield strength. YS measured in customary tests tends to have large variations. UTS in both longitudinal and hoop directions tends to be close. In many cases, yield strength can be estimated using a Y/T ratio of

28 Girth Welding Options (Preliminary) If pipe UTS ksi, E6010/E8010 is likely adequate. If < pipe UTS < 100 ksi, E8010/E8010 may be OK? (need further analysis and testing) If pipe UTS > 100 ksi, need weld strength > E8010/E8080, or Weld cap reinforcement All of the above would be valid if there is no excessive HAZ softening. 28

29 Pipeline in Service Integrity Assessment Information needed Pipe MTRs WPS / PQR Strain tolerance level High Low In-between If the strain tolerance is not high, further screening Strain demand Experience, e.g., terrain, likelihood of settlement/landslides, construction practice Location, e.g., tie-in, crossings Risk tolerance Site-specific assessment 29

30 Observed trends Linepipe Specifications Standard deviation in reported YS: 3-4 ksi, say 3.5 ksi ± 3D = 21 ksi, 99.1 Percentile SMYS + 3SD = 80.5 ksi ± 2D = 14 ksi, 94.7 Percentile SMYS + 2D = 77 ksi Average under-reporting of YS by current tests = 12% targeting at least 79 ksi for X70 79 ksi + 2SD = 86 ksi Actual YS of X70 Mean: low-mid 80 ksi Range = ~77-95 ksi Actual UTS of X70 Mean: ~91-95 ksi Range: ~

31 Linepipe Specifications Hoop vs. Longitudinal Properties ERW and SAWL pipes Strength in longitudinal direction is lower than that in hoop direction SAWH pipes Tests from non-flattened specimens: very little difference between hoop and longitudinal. Tests from flattened specimens: YS in hoop is 5-7 ksi lower than YS in longitudinal. This difference is largely due to test methods. 31

32 To Achieve Weld Performance With the current test methods and test data scatter, it s difficult to keep UTS < 95 ksi. To keep using E6010/E8010, need to reduce the strength of the pipe Eliminate/reduce the systematic under-representation of yield strength Reduce the scatter in test data Possible solutions Test specimens without flattening Yield strength defined at a strain greater than 0.5% total strain Impost more robust test protocol If pipes were to be tested and qualified the same way, need to change welding High strength consumables Cap reinforcement Low heat input to derive high strength from the same consumable (feasibility?) Conclusion: work from both sides, plus reduced level of HAZ softening 32

33 Path Forward None of the potential solutions are easy and quick. Low-hanging fruits Premises: having X70 pipes with upper bound UTS at 95 ksi is difficult Approaches E8010 root or similar Extra cap reinforcement Reduction in heat input, specify more passes 33

34 Path Forward Next step engage pipe mills and labs to try out new test procedures Next step understanding HAZ softening Next step lower the upper-bound strength Impose lower limits on pipe strength Hoop vs. longitudinal YS, UTS, or both Requiring longitudinal tensile tests a current proposal Longitudinal test: limit upper bound Hoop test: limit lower bound Consequence Mills have smaller window to work with Can lead to disqualifying materials purely due to imprecisions in test methods 34

35 Long term Better tests for pipes More precise representation of pipe strength Path Forward Reduction in the reliance on rolling and accelerated cooling to achieve strength Welding options can be selected based on the deliverate selection of pipe strength Industry-wide initiatives Continue to engage API 5L and 1104 and similar standard organizations for incremental improvements Engage ASME on updates in design requirements PRCI efforts to identify gaps in our current practice, particularly inconsistences among design, materials specifications, welding procedure qualification, and operations and maintenance. 35

36 Thank You Q&A 36