
Weld Strength Matching and Tensile Strain Capacity of Pipeline Girth Welds
Introduction
Following reports of unexpected pipeline girth weld failures in the USA (1), APGA convened a workshop to discuss the technical requirements and structural integrity of pipeline girth welds of similar welded pipelines in Australia.
The workshop was attended by approximately 80 pipeline personnel from across Australia, representing the custodians of our major pipeline assets, APGA Board members, Australian Standards representation, and Australian State Pipeline regulators.
Invited presentations from leading Australian experts in steel, welding, failure analysis, engineering, including a distinguished Canadian welding engineer were designed to provide an understanding of the technical factors controlling weld joint strength in modern high
strength steels along with a critical review of the reported girth weld failures in the USA.
Of particular concern to the Australian industry is that the USA failures were claimed to be:
1. Mostly grade X70 pipe welded with cellulosic consumables, E6010/E8010, and
2. USA code compliant, which could imply that similar Australian welded pipelines are subject to such failures. We note that Australian Standards are different to the US Code and compliance with the US Code does not indicate compliance with Australian Standards.
Girth weld integrity, or more precisely girth “weld strength matching” (2) detailed by Dr Garcia, is an interplay of many factors which include the pipe strength, weld strength, profile and the complex interaction of weld imperfections, which are an inherent artifact of welding, along with the stress and strain which can occur during construction and operation. The critical component of stress needs to be very carefully understood in terms of “construction” where there is no internal pressure, but the pipe is subject to handling stress, and “operation” where internal pressure is constant but movement
Assessment of USA Failure Reports with Australian experience
General information of the USA failures has been published in a number of reports (3; 4; 5) which reveal that there have been two waves of girth weld failures. The first wave occurred 2007-2009 and was confirmed to be associated with Hydrogen Assisted Cold Cracking (HACC) and were predominantly related to tie-in welds. A PRCI report, initiated by PHMSA and undertaken by DNV was published (6) to comprehensively address the requirements for the industry to control hydrogen during and after welding and also the influence of stress during and immediately after welding had on the initiation of such cracking. The second wave of these girth weld failures, which totalled 10 occurred in 2015-2020 (7; 4; 3) . Analysis of these girth weld failures by USA researchers suggested that weld strength undermatching and heat affected zone (HAZ) softening were primarily responsible. The 2023 APGA workshop was based on a metallurgical analysis of the USA failures, utilizing broad Australian experience into defect acceptance levels in girth welds (8), advanced numerical modelling (9) along with the extensive research into HACC (10) the latter of which occurred over many decades incorporating a number of Government sponsored CRC’s and leading industry groups. The Australian analysis has identified anomalies and omissions, which require re-assessment of the conclusions drawn regarding the root cause of the USA failures:
- Five (5) out of the ten (10) girth weld from the second wave of failures were tie-in welds, similar to the first wave of girth weld failures,
- A large number of the second wave girth weld failures possessed features such as vertical fractures with an absence of any visible strain, which is not typical of weld strength undermatching, where significant plastic strain (collapse) even on a small scale would be expected. Vertical sections on a fracture face are typical of low strain failure, not unlike brittle fracture, typically associated with HACC.
- The majority of weld cross sections presented in the USA publications showed the fracture propagation event and not the fracture initiation site and therefore do not provide any evidence of the metallurgical root cause for failure.
- Fatigue fracture was not identified on one of the girth weld failures (3) and the extreme damage to the pipe (8) in the USA failure publication. (3)
- Absence of any Weld Procedure Specifications (WPS) and Weld Procedure Qualification Reports (WPQR) of the pipeline construction welds.
- Hardness mapping results which were utilized to highlight differences in strength matching were inconsistent:
- Differences in hardness matching maps varied significantly for each weld failure, which is insufficient to conclude that undermatching as the root cause.
- b. Indeed, some isolated high hardness results could only be attributed to unfavourable microstructures which are known to increase susceptibility to HACC,
- c. The evaluation of adjacent, intact girth welds is tenuous evidence, especially where no WPS is provided,
- Finite element modelling provided to support the conclusion of weld strength undermatching, in the absence of ANY weld defect, can result in low (gross) strain failure, and is inconsistent with the results of Lu & Michal (4; 9) and Ted Anderson (11).
- Reported failure investigations did not adequately consider the influence of weld defects, such as HACC, which is difficult to identify and can cause failure over small levels of strain, in a similar manner to weld strength undermatching.
- More specifically, the published conclusions (1, 3, 6) relating steel alloy design as a factor to these pipeline girth weld failures are at odds with evidence that the pipe associated with second wave of girth weld failures possessed a wide range of both carbon contents and carbon equivalents. Low carbon contents and carbon equivalents were not a feature of the pipe from the second wave of failures, as with the first wave of failures.
A review of Australia Standard AS2885, and comparison with API 1104 was presented by the chair of the Part 2 committee on welding, Mark Lackenby. The key features of each standard in relation to the elements addressing weld strength matching was summarised, as shown in
Figure 1.
Figure 1: Main differences between USA API 1104 and Australia AS2885.
These points are self-explanatory, however the distinguishing feature of the standards, as explained by David Horsley was Australia’s “fixation on HACC, for good reason if you don’t preheat”. Moreover, David was complimentary that the Australian standard AS2885 was the benchmark in terms of a risk-based approach to management of pipeline integrity. Phil Venton, an Australian veteran of pipeline design and construction, explained the aspect of Pipeline Risk Assessments, which was incorporated into AS2885.1 some 25 years ago in 1997.
The licensee is responsible for pipeline safety, and that use of the “Safety Management System (SMS) is a universal obligation through the pipeline life cycle”. A distinction between these Standards, of relevance to the outcomes of an industry-based investigation of the USA failures (12), was the development and control of pipe girth weld procedure specification, which is conducted for each pipeline construction project, on the pipe to be used in construction. This is sound logic when the pipe for each pipeline project is specifically designed in terms of steel chemical composition and rolling process. The steel supplier is an essential variable, as identified in the WPS section of AS2885, and this controls performance of the final welded pipeline. This is not the situation in the USA where both steelmaker and also weld heat input are omissions in terms of essential variables to the WPS. The latter of which has a major influence on weld HAZ softening, particularly in the absence of any weld procedure specification.
In terms of pipeline movement, which was noted to be a key feature of many USA failures, an informative update on new techniques that diminish the uncertainty around external pipeline loads during operation was explained by Barry Foster (Baker Hughes) and Daniel Schneke (Rosen). Unexpected displacement is known to occur around river crossings and flood zones; but the greatest concern is the unexpected displacements due to slow ground shift. Current In-Line Inspection (ILI) and Inertial Measuring Unit (IMU) technology has now been developed which can define pipeline condition and position in terms of pipe strain along with global position. The historical concerns of landslip and subsidence can now be better understood, risk assessed and mitigated.
Presentations by Australia’s largest pipeline owners and operators provided a detailed summary of the controls and practices adopted in response to the knowledge of these USA girth weld failures. It was explained that practices are aligned with well-established maintenance procedures, and consistent with the standard risk process and SMS through the pipeline life cycle. This is all ably supported by International collaborative research programs. High level screening of X70 pipelines constructed using cellulose consumable procedures have been undertaken to assess risk, particularly in identified suspect regions and mitigated by design as required. An example of a recent pipeline project utilizing API 5L PSL2 X70 pipe, welded using E6010/E8010 weld procedure with an MAOP 15.3MPag and 0.8 design factor was presented as an example. In some cases, new pipeline construction projects have been limited to grade X60 to ensure weld strength matching until further details of full-scale girth weld tension tests are completed by the collaborative Industry & Government sponsored Future Fuels Cooperative Research Centre (FFCRC).
The current research activities within Australia are coordinated through the FFCRC, which has initiated a project to quantify the complex parameters defining girth weld strain capacity. This program of research will utilize both wide plate tests and full-scale pipe tension tests of grade X70 pipe welded with cellulose E6010/E8010 consumables and incorporating a range of artificial weld defects. A detailed literature review has been completed along numerical models and stu ies, referred to above. The first batch of full-scale tests have been completed and a further set of associated & confirmatory tests are planned over the next few months. The research outcomes from this extensive program of work will be presented to the Australian industry through FFCRC and APGA research meetings and the international community through our joint membership of EPRG/PRCi/APGA tripartite meetings and international conferences and journals (4).
Concluding Comments
have not been adequately evaluated and that a causal relationship with weld strength undermatching and HAZ softening has not been metallurgically confirmed. The details of each girth weld failure do not provide adequate evidence to support the conclusion that weld
strength undermatching and HAZ softening are the universal factors that have contributed to those failures.
Moreover, the conclusions drawn from the June 2023 APGA Workshop review was that the two USA waves of girth weld failures (which occurred in quick succession) exhibited very similar characteristics and are believed to exhibit fracture features consistent with the
presence of HACC. The metallurgical analysis of the failures presented in this workshop and the detailed modelling by Lu (4; 9) and Anderson et al (11), provides strong circumstantial evidence that the culmination of external loading in the presence of a defect has exacerbated
weld strength undermatching in the USA girth weld failures, as depicted in Figure 2.
It is well known that weld strength undermatching is a feature of cellulosic welding processes in X70 grade pipe. Weld procedural qualification would identify such deviations in pipe strength, provided of course, that weld qualification was conducted on representative p project pipe for each pipeline construction project, as mandated in Australia Standards AS2885. It is important to note that this requirement is not a mandatory requirement in API1104 and is generally not employed by the majority of USA pipeline construction companies, where the use of “heritage weld procedures” is the norm.
Actions moving forward
In agreement with the outcomes of the 2023 Canadian Energy Regulator (CER) Girth Weld Workshop (13), the APGA workshop confirms that a better understanding of the U A girth weld failures is required before any mandatory actions are undertaken here in Australia.
The Australian Standard AS2885 has served the industry well, as demonstrated by our established incident database. Despite past performance, continuous improvement is essential to maintain safety in pipeline operation. An outcome from the investigative workshop by APGA on these reported USA girth weld failures is the opportunity for the AS2885 Standard to be further improved by reinforcement of specific items with respect to weld integrity. In parallel with, or even prior to, the initiation of a revision to the relevant Part of the AS2885 Standard, APGA could also publish guidance document(s) based on the outcomes of the Workshop.
For new pipelines opportunities for improvement have been identified for the relevant
sections of Australian Standard AS 2885.
AS2885.1 (Design and Construction) sub-committee to consider:
1. Reinforcement or enhancement of existing text in Appendix K6.4, or moving to clause
5.10.2, a requirement to identify locations of land instability and taking suitable design
actions.
2. Requirements for pipe strength design.
AS 2885.2 (Welding) sub-committee to consider:
Improved configuration, interpretation, and acceptance criteria of the cross weld tensile test in the Weld procedure qualification testing (WPQT).
2. The traditional girth weld hardness traverse (profile) to be replaced by a quantitative hardness mapping technique.
3. Although it is well known that HACC must be designed out the weld procedure, it is possible that AUT could be deployed in high-risk areas and also during WPQT to further ensure elimination of HACC.
4. Reduced spread in pipe steel strength and inclusion of longitudinal tensile properties in future pipeline procurement.
5. Consolidation / strengthening of design requirements.
6. Engineering analysis to ensure weld strength matching requirements,
7. Mandatory reports of each weld consumable batch,
8. Utilize the outcomes of the FFCRC initiated full scale test program for quantitative analysis of weld strength matching requirements under both pipeline construction and operation conditions.
9. Field weld monitoring to ensure compliance with welding procedures should be stressed.
For existing pipelines there are opportunities identified for the Australian Standard AS 2885.3
(Operations and maintenance) sub-committee to consider:
1. Ongoing monitoring and risk management of pipeline operations,
2. Effective use of ILI / IMU pipeline monitoring in conjunction with available geotechnical reports. 3. Review of landform/geotechnical landscapes for high strength (>X65) pipelines, welded throughout with cellulosic consumables, particularly in areas of potential ground instability and uneven loading such as crossings.
4. Undertake additional inspection of girth welds during routine pipeline digs to ensure girth weld integrity, with respect to the potential existence of HACC, particularly in areas of ground instability.
In the meantime, in advance of a formal revision being initiated, Pipeline Licensees can address these items immediately, as a pro-active safety measure. Additionally, pipeline licensees and the industry should consider the development of a standardised approach to investigation of pipeline incidents (i.e., data gathering, report contents).