March, 2013 Vol.12, No.1

Great Southern Press Clarion Technical Publishers

Journal of Pipeline Engineering

incorporating The Journal of Pipeline Integrity

Sample

copy

not fo

r dist

ributi

on

Journal of Pipeline Engineering

Editorial Board - 2013

Dr Husain Al-Muslim, Pipeline Engineer, Consulting Services Department, Saudi Aramco, Dhahran, Saudi Arabia

Mohd Nazmi Ali Napiah, Pipeline Engineer, Petronas Gas, Segamat, MalaysiaDr-Ing Michael Beller, Landolt Steuer & Unternehmensberatung AG, Luzern, Switzerland

Jorge Bonnetto, Operations Director TGS (retired), TGS, Buenos Aires, ArgentinaDr Andrew Cosham, Atkins Boreas, Newcastle upon Tyne, UK

Dr Sreekanta Das, Associate Professor, Department of Civil and Environmental Engineering, University of Windsor, ON, Canada

Leigh Fletcher, Welding and Pipeline Integrity, Bright, AustraliaDaniel Hamburger, Pipeline Maintenance Manager, Kinder Morgan, Birmingham, AL, USA

Dr Stijn Hertele, Universiteit Gent – Laboratory Soete, Gent, BelgiumProf. Phil Hopkins, Executive Director, Penspen Ltd, Newcastle upon Tyne, UK

Michael Istre, Chief Engineer, Project Consulting Services, Houston, TX, USA

Dr Shawn Kenny, Memorial University of Newfoundland – Faculty of Engineering and Applied Science, St John’s, Canada

Dr Gerhard Knauf, Salzgitter Mannesmann Forschung GmbH, Duisburg, GermanyProf. Andrew Palmer, Dept of Civil Engineering – National University of Singapore, Singapore

Prof. Dimitri Pavlou, Professor of Mechanical Engineering, Technological Institute of Halkida , Halkida, Greece

Dr Julia Race, School of Marine Sciences – University of Newcastle, Newcastle upon Tyne, UK

Dr John Smart, John Smart & Associates, Houston, TX, USAJan Spiekhout, Kema Gas Consulting & Services, Groningen, Netherlands

Prof. Sviatoslav Timashev, Russian Academy of Sciences – Science & Engineering Centre, Ekaterinburg, Russia

Patrick Vieth, President, Dynamic Risk, The Woodlands, TX, USADr Joe Zhou, Technology Leader, TransCanada PipeLines Ltd, Calgary, Canada

Dr Xian-Kui Zhu, Senior Research Scientist, Battelle Pipeline Technology Center, Columbus, OH, USA

� � �Sam

ple co

py

not fo

r dist

ributi

on

1st Quarter, 2013 1

The Journal of Pipeline Engineeringincorporating

The Journal of Pipeline Integrity

Volume 12, No 1 • First Quarter, 2013

ContentsIn memoriam

Alan Reece, 1927-2012 .................................................................................................................................................... 5

Dr Ted L AndersonRecent innovations in pipeline seam-weld integrity assessment ......................................................................................... 7

Michael J Rosenfeld and Rick W GailingPressure testing and recordkeeping: reconciling historic pipeline practices with new requirements ................................... 15

Wim N Schipaanboord, Jan Marquering, Bart G Koppens, and Jan SpiekhoutWelding on in-service gas pipelines using low-yield electrodes .......................................................................................... 29

Dr Julien Capelle and Professor Guy PluvinageEvaluation of failure risk due to use of high-strength steels in pipelines .......................................................................... 51

Hossein Ghaednia, Jorge Silva, Sara Kenno, Prof. Sreekanta Das, Rick Wang, and Richard KaniaPressure tests on 30-in diameter X65 grade pipes with dent-crack defects ....................................................................... 61

❖ ❖ ❖

OUR COVER PHOTO shows a recent 1.5-km long pull-in of a 24-in pipe-in-pipe district-heating pipeline in Rotterdam at a depth of 50m. A second horizontal directionally-drilled pipe will be installed before the end of March, 10m below the first. The contractor is Visser and Smit Hanab, on behalf of Rotterdam’s district-heating company Warmtebedrijf Rotterdam.

The Journal of Pipeline Engineering has been accepted by the Scopus Content Selection & Advisory Board (CSAB) to be part of the SciVerse Scopus database and index.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering2

1. Disclaimer: While every effort is made to check the accuracy of the contributions published in The Journal of Pipeline Engineering, Great Southern Press Ltd and Clarion Technical Publishers do not accept responsibility for the views expressed which, although made in good faith, are those of the authors alone.

2. Copyright and photocopying: © 2013 Great Southern Press Ltd and Clarion Technical Publishers. All rights reserved. No part of this publication may be reproduced, stored or transmitted in any form or by any means without the prior permission in writing from the copyright holder. Authorization to photocopy items for internal and personal use is granted by the copyright holder for libraries and other users registered with their local reproduction rights organization. This consent does not extend to other kinds of copying such as copying for general distribution, for advertising and promotional purposes, for creating new collective works, or for resale. Special requests should be addressed to Great Southern Press Ltd, PO Box 21, Beaconsfield HP9 1NS, UK, or to the editor.

3. Information for subscribers: The Journal of Pipeline Engineering (incorporating the Journal of Pipeline Integrity) is published four times each year. The subscription price for 2013 is US$350 per year (inc. airmail postage). Members of the Professional Institute of Pipeline Engineers can subscribe for the special rate of US$175/year (inc. airmail postage). Subscribers receive free on-line access to all issues of the Journal during the period of their subscription.

4. Back issues: Single issues from current and past volumes are available for US$87.50 per copy.

5. Publisher: The Journal of Pipeline Engineering is published by Great Southern Press Ltd (UK and Australia) and Clarion Technical Publishers (USA):

Great Southern Press, PO Box 21, Beaconsfield HP9 1NS, UK

• tel: +44 (0)1494 675139• fax: +44 (0)1494 670155• email: jtiratsoo@gs-press.com• web: www.j-pipe-eng.com• www.pipelinesinternational.com

Editor: John Tiratsoo• email: jtiratsoo@gs-press.com

Clarion Technical Publishers, 3401 Louisiana, Suite 110, Houston TX 77002, USA

• tel: +1 713 521 5929• fax: +1 713 521 9255• web: www.clarion.org

Associate publisher: BJ Lowe• email: bjlowe@clarion.org

6. ISSN 1753 2116

THE Journal of Pipeline Engineering (incorporating the Journal of Pipeline Integrity) is an independent, international, quarterly journal, devoted to the subject of promoting the science of pipeline engineering – and maintaining and

improving pipeline integrity – for oil, gas, and products pipelines. The editorial content is original papers on all aspects of the subject. Papers sent to the Journal should not be submitted elsewhere while under editorial consideration.

Authors wishing to submit papers should do so online at www.j-pipeng.com. The Journal of Pipeline Engineering now uses the Aires Editorial Manager manuscript management system for accepting and processing manuscripts, peer-reviewing, and informing authors of comments and manuscript acceptance. Please follow the link shown on the Journal’s site to submit your paper into this system: the necessary instructions can be found on the User Tutorials page where there is an Author's Quick Start Guide. Manuscript files can be uploaded in text or PDF format, with graphics either embedded or separate. Please contact the editor (see below) if you require any assistance.

The Journal of Pipeline Engineering aims to publish papers of quality within six months of manuscript acceptance.

Notes

v v v

www.j-pipe-eng.comis available for subscribers

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 3

Editorial

THE PROGRAMME AT the recent – and 25th – Pipeline Pigging and Integrity Management conference

in Houston contained 22 presentations of considerable significance to the hydrocarbons’ pipeline industry, and we hope to publish a number of these in the Journal of Pipeline Engineering. A number of new aspects of current technology were discussed, along with papers of particular current significance such as those about the SmartBall ILI acoustic leak-detection equipment, the impact of shale plays on pigging, and the use of electro-magnetic acoustic technology (EMAT) pigs to detect coating conditions.

To choose to highlight here a few papers out of such an interesting and wide-ranging selection is inevitably invidious. While the three chosen here cover widely different aspects of the conference’s subject, all are nevertheless of general significance and importance to the industry as a whole: first-run success, corrosion growth rates, and record keeping.

The financial cost of an inspection run failure has probably been underestimated by the industry, although it is estimated that it could be as high as 30% of the total contracted costs for an in-line inspection (ILI). For some projects, the costs associated with a failed run can be far greater than the original project costs (for example, additional vessel-support costs during offshore operations). A failed run can also result in a delayed inspection and an associated increased risk, as well as potentially compromising compliance with regulatory requirements. As pointed out by Stephen Gower of BP in his introduction to a panel discussion on this subject, the consequences of run failure vary in severity and can be presented as a pyramid similar to the typical representation of safety statistics: a stuck tool requiring intervention, or a pipeline failure as a result of an incorrect inspection report, would be at the top of the pyramid. The lower tiers would capture technical failures, and the effectiveness of cleaning. As part of BP’s continuous improvement process, ILI suppliers and internal stakeholders have been brought together for a facilitated workshop to understand the factors affecting first-run success rates, and a ‘guidance note’ was prepared which was then shared with the Pipeline Operators Forum (POF); this has now been further developed as a POF Guidance Document. As Mr Gower pointed out, at the very least “successful ILI requires good communication between all parties”.

Assessing how quickly defects may corrode can be complex due to various factors including non-linearity in corrosion growth rates, differing corrosion mechanisms for different kinds of defect, and differences in inspection tool tolerances. To simplify such assessments, a single corrosion growth rate may be applied to all defects in a pipeline, and this

approach is frequently taken as the basis for determining a re-inspection date by intelligent pigging. The Pipeline Operator’s Forum provides guidance on classifying defects based on pipeline wall thickness, a defect’s axial length, and its circumferential width. Examples of this include pitting-type defects, general-corrosion-type defects, and pinholes. The paper by Wood Group’s Toby Fletcher examined the use of these classifications to estimate corrosion growth rates by fitting the dimensions of defects to different statistical distributions. He reports that the use of this method can help to refine estimates of corrosion growth rates and corrosion-management strategies, leading to development of more-accurate intelligent pig re-inspection intervals and consequent savings.

Based on its investigation of the 2010 San Bruno, California, gas pipeline incident, the US’ National Transportation Safety Board’s recommendations to pipeline operators and regulators has focused industry attention on validation of gas pipeline maximum operating pressures of pipelines installed prior to federal or state regulations. The industry is in the process of establishing the viability of the continued operation of these so-called ‘grandfathered’ pipelines as well as of pipelines that lack the records necessary to give full confidence in the quality of installation. A full understanding of the implications requires knowledge of how industry practices and code requirements in the areas of pressure testing and record keeping have evolved over time. The paper by Mike Rosenfeld and Rick Gailing (respectively, of Kiefner & Assocs and Sempra Utilities) which is published on pages 15-28 of this issue, explored these issues in gas pipeline standards and regulations, historically and currently, and both nationally and in the state of California. The authors note that it is likely that similar considerations may develop in other states, and the paper describes the evolution of pipeline pressure-testing requirements, what records have been specifically required, how those records relate to establishing the maximum allowable operating pressure of a pipeline, why grandfathered pipelines have existed, and the significance of recently articulated criteria for records’ accuracy.

As the authors point out, the costs for California’s gas pipeline operators to perform hydrostatic testing and/or replacement of hundreds of miles of grandfathered pipelines would be enormous, and the question to be settled – they suggest, by public debate – is how to share this enormous cost between shareholders and rate payers in an equitable manner. The authors do not dispute the need to modernize the pipeline infrastructure, to verify the integrity of grandfathered pipelines as well as any other category of pipeline, and to enhance data in order

First-run success, corrosion growth, and recordkeeping

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering4

record-keeping requirements, did the regulator recognize a contravention in the past? If so, what was done about it? If not, has a reasonable ‘statute of limitations’ rendered it expired?) Finally, they point out that customers should pay rates that fully reflect the cost of providing the goods or services, and these rates should therefore reflect the cost of complying with any new regulatory requirements. This includes the cost of retesting or replacing pipe not thought to meet modern standards of integrity, where those activities are necessitated by new regulations.

These are clearly highly challenging issues, and will reflect on pipeline networks everywhere, as they age: as in other branches of technology, and particularly computing, the Californian community can be considered to have led the way. It will be interesting to see how this leadership is manifested in the pipeline world.

to manage risk as well as is practical. They believe that pipeline operators should be accountable for failures to meet regulatory obligations. “However”, they continue, “we further believe that regulators should not give in to the temptation to penalize operators retrospectively for practices that previously were not considered deficient as judged by contemporary standards. Past practices must be gauged against an accurate and reasonable interpretation of historic standards, regulations, and accepted practices. Moreover, proceedings to establish financial penalties for past failures to comply with regulations should be separated from ratemaking proceedings designed to regulate future behaviour.” When weighing financial penalties, the authors say consideration should also be given to the intensity of the regulator’s past focus on regulatory provisions deemed in retrospect to have been contravened. (In other words, if the regulatory violation was a failure to comply with

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 5

Alan Reece, the pioneer of the ploughing method for trenching marine pipelines, passed away on 31

December, 2012, at the age of 85. He was an extraordinary character, who will not be forgotten by anyone who was lucky enough to meet him, and he made a lasting contribution to the industry.

Alan’s early work was in agriculture, and as a student he took part in the legendary volunteer work camp that built a railway in Yugoslavia in the late 1940s. At the University of Newcastle he became a Lecturer and later a Reader, and carried out research in soil mechanics and vehicle-soil interaction.

His engagement with pipeline trenching began in 1976. Jetting was slow and expensive, and did not form a well-shaped trench. Mechanical cutting was equally slow, and subject to frequent breakdowns. Bob Brown at RJ Brown and Associates (RJBA) had the idea that ploughing could be much faster and more economical, and that it would cut a trench that would protect a pipeline better. The immediate demand was a trench that would protect a 36-in loading line for the Statfjord field in the Norwegian sector of the North Sea, in stiff 150-kPa clay. Bob persuaded Mobil Exploration Norway that a plough would be the way to cut the trench. At RJBA, Ed Vermeulen – latterly the Chairman of the DNV Pipelines Committee – started the detailed design, but it was soon realized that some specialist knowhow would be valuable. Many sceptics predicted that the plough would fail. The writer approached the UK National Institute of Agricultural Engineering, but it was less than helpful. He then talked to Andrew Schofield, who had recently returned to Cambridge University as Professor of Geotechnics, and he recommended Alan, characteristically adding as a positive factor that Alan had read Schofield’s book on critical state soil mechanics. We phoned Alan, and a couple of days later Vural Dolen and the writer went to Newcastle to see him. He was clearly intrigued, and remarked that he had always thought ploughs “too difficult”.

The problem with trenching by a cutting blade in soil is that it easily becomes unstable. Either the blade rises out of the soil and scrapes along the surface without making much of a trench, or else it digs in further and further until it becomes an anchor. A few days after the first meeting Alan came to RJBA with a cardboard model of a plough, based on a concept applied in Finland to huge ploughs for cutting drainage ditches in that country. That design solves the stability problem: the front end of a long beam runs on wheels or skids that hold it at fixed level above the ground surface, and the rear end carries the share, with underneath it a heelplate that runs on the bottom the

trench that the share has just cut: how that stabilizes the cutting depth is described in another paper [1]. The next step was a 1/16-scale steel model, and it worked perfectly in a clay pit. From there, the development accelerated, first to a quarter-scale model and more tests.

There were hiccups. Alan always refused to compromise, and refused to do anything that he considered boring: “if it isn’t enjoyable, it isn’t worth doing”. He walked out of a meeting with RJBA’s Vice-President rather than sign away his ownership of intellectual property. He argued that most patents are not worth the paper they are written on, but that a few ideas are good ones, and they deserve to be protected. Alan and Bob Brown were in some ways similar: both were lively and creative engineers, with well-developed egos, strong opinions, and strong appetites. More than once their egos came close to collision, but they respected each other and problems were avoided.

One evening we arrived at Newcastle for a test of the quarter-scale plough, planned for the next day. Mobil was represented by a tall German in a brilliant white trench coat and an equally tall Texan in spectacular boots: both were somewhat sceptical. We ate and drank, and Alan was in particularly fine form, with a fund of stories, not all of them suitable for a family magazine like the Journal of Pipeline Engineering. After three convivial hours, the German turned to Alan: “This is all very interesting, Dr Reece, but can you tell us about the test tomorrow?” “Ah, well”, replied Alan, “I have a confession to make. The plough broke in two at three o’clock this afternoon.”

In memoriam Alan Reece, 1927-2012

Fig.1. Alan Reece with the first full-scale test plough, designed to cut a 1.2-m deep trench in stiff clay.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering6

If only there were a video of the four faces! The next morning we were able to see that the plough had made a short trench before it broke. The development might well have cratered there and then, but miraculously it continued. Mobil was heroically supportive, the Norwegian authorities agreed to moderate their demands for deep trenching, and the full-scale plough to cut a 1.2-m deep trench in stiff clay was designed, built, and tested. Figure 1 shows Alan in the test trench. In June, 1977, the Statfjord trench was ploughed successfully.

More ploughs followed. In the Canadian Arctic, a deep 1.5-m trench was needed to protect the Panarctic Drake F-76 flowline bundle against ice gouging. The seabed soil was totally different from the seabed at Statfjord, about 40 times weaker, and it was important that the weight of the plough would not cause the trench sides to collapse behind it. The plough could not be built and tested in time to catch the summer 1977 sealift, and so it had to be designed so that it could be taken to pieces and flown to Melville Island. That plough, too, operated correctly in March, 1978. A sign of success was that the engineers who had contemptuously scoffed at the idea of a plough now began to pretend that it had been their idea all along.

RJBA and Alan both developed ploughs further. RJBA wanted to design ploughs but not to build them, whereas Alan wanted his company Soil Machine Dynamics to build and supply complete plough systems. Progressively, competition led the companies to drift apart, happily without personal animosity. Alan secured his first order to build a plough, and signed the first contract on his kitchen table. He brought in the research students who had been with him since the beginning, Tony Trapp and Tim Grinsted, and together they built the company, initially in an office in Stocksfield loud with Italian opera. He moved into cable ploughs and mine ploughs, which run in front of tracks of a tank driving across a minefield, and lift and move aside mines before the tracks reach them (preferably without turning them over). His fabrication company Pearsons took off.

Alan pursued one of his principal convictions, that the decline of manufacturing industry in the UK must be arrested. He talked to Margaret Thatcher, which must have been another memorable encounter between strong and vivid characters. Alan was sceptical of government support of research, and famously remarked that government support is for second-rate ideas, because first-rate ideas can find support from investors. Later Tony and Tim left SMD and formed their own company, the Engineering Business, which became another success story. Alan gave a major endowment to the manufacturing group at Cambridge University, and made possible its new building, now named after him (Fig.2).

Alan was a delightful character, and made a huge contribution to the industry, far more than a hundred plodders (or a thousand code-bashers). It is tempting to opt for the conventional piety, and to say that “we shall not see his like again”. But of course we shall, and our goal must be to protect and nurture their creativity, and to accept and indeed welcome occasional eccentricities.

Ann Beresford is writing a more extended account of Alan’s life.

References

1. A.C.Palmer, J.P.Kenny, R.Perera, and A.R.Reece, 1979. Design and operation of an underwater pipeline trenching plough. Géotechnique, 29, 305-322.

Prof. Andrew PalmerCentre for Offshore Research and Engineering,

National University of Singapore

Fig.2. Alan Reece on 19 November, 2009, at the opening of the eponymous building for Cambridge University’s manufacturing group.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 7

THE INTEGRITY OF pipelines with longitudinal seam welds has received renewed interest by operators and regulators, due primarily to a number of high-profile incidents. Most operators and pipeline-

integrity consultants have assessed seam weld flaws with methodologies that have changed very little over the past 30 years. However, the status quo is no longer viable, given the heightened public scrutiny and regulatory pressure.

This paper presents a number of recent innovations in assessment technology that can lead to improved reliability and a more optimal use of finite integrity budgets.

This paper was presented at the Fixing Pipeline Problems conference held in Berlin in October 2012, and organized by Clarion Technical Conferences and Tiratsoo Technical.

Author’s contact details:tel: +1 303 938 3010email: t.anderson@questintegrity.com

Quest Integrity Group, Boulder, CO, USA

Recent innovations in pipeline seam-weld integrity assessment

by Dr Ted L Anderson

THERE HAVE BEEN a number of catastrophic pipeline incidents in recent years, the most notable of which

was the 2010 explosion of a Pacific Gas & Electric (PG&E) pipeline in San Bruno, California. The belief among the general public is that pipeline failures are more frequent than in the past. It is debatable whether this belief has a factual basis (i.e. more frequent failures are the result of the aging pipeline infrastructure), or whether the Internet and the 24-hr news cycle cause an inflated sense of the actual threat1. Nevertheless, perception is reality in politics, so both Congress and PHMSA are instituting tighter regulations on the pipeline industry. There is increasing pressure on operators to demonstrate that they are taking steps to improve the integrity of their pipelines. Simply adhering to the status quo in the form of existing integrity plans is no longer an option.

Governmental bodies are paying particular attention to seam-welded pipes because several prominent releases, including the PG&E explosion, occurred at longitudinal seams. In the past year there has been much discussion of the seam-weld issue by various parties, including state and local government agencies, pipeline operators, industry organizations, and consultants. Some have advocated a large influx of research funding to address the perceived gaps in technology. In a number of cases, however, the necessary technology already exists, so further R&D might result in reinventing the wheel.

For example, the pipeline industry currently relies on flaw-assessment methods that are over 30 years old, but improved models are available. There have been significant advances in fracture mechanics, fitness-for-service assessment, and remaining-life models in the past 30 years. The pipeline industry can benefit by adopting methodologies that have been successfully applied in other industries, including oil and gas production, refinery, chemical, petrochemical, and power generation.

This paper presents a sample of innovative technology that can be applied to the integrity management of seam-welded pipe. The focus of this article is on cracks and other planar flaws, but innovative approaches for other anomaly types are also available.

Can a ‘conservative’ approach be unsafe?

Although many pipeline operators are open to innovations in integrity management, some adopt the attitude of “if it ain’t broke, don’t fix it.” One of the most common arguments against using advanced flaw-assessment methods is that the traditional approaches are ‘conservative’, in that they tend to underestimate burst pressure and critical flaw size. There are at least two counter arguments to this point of view:

• When a ‘conservative’ failure model is used in conjunction with a hydrostatic test, the simplified model underestimates the maximum flaw size that could have survived the test. As a result, larger-than-expected flaws are often left in the pipe after a hydrostatic test. Since large cracks grow faster than small cracks, the ‘conservative’ approach can overestimate remaining life. Using a supposedly conservative model in reverse is analogous to looking through the wrong end of a telescope.

1 Historical data from PHMSA shows a decreasing rate of serious pipeline incidents overall, but there were an unusually high number of incidents in 2010.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering8

• false negative – no action is taken on an anomaly that poses a risk, either because it is not detected by ILI or because inaction is a conscious decision

In an ideal world, only true positives and true negatives occur; false positives result in unnecessary expenditures, and false negatives result in higher risk.

Figure 1 illustrates the effect of knowledge of the pipeline condition on the cost of integrity. Consider two theoretical extremes: perfect ignorance and perfect knowledge. The former corresponds to the situation where nothing is known about the condition of the pipe, and digs and repairs are made randomly. Perfect knowledge means that the operator has advance knowledge of the exact timing and location of failures, which enables just-in-time action to prevent incidents. Real pipelines, of course, operate between these two extremes, and the cost of 100% integrity decreases with knowledge about the presence of flaws and their associated risk because remediating false positives costs money. Advanced-assessment methods improve the state of knowledge about the risk that various flaws pose, which reduces the cost of achieving 100% reliability.

Figure 2 illustrates the case where the integrity budget is fixed. Given a finite budget, a ‘conservative’ assessment can result in less reliability compared to what could be achieved with an advanced assessment. The reason for the diminished reliability is that false positives consume the integrity budget, so fewer resources are available to inspect and assess other areas that may contain dangerous flaws. In other words, false positives can lead to false negatives, as described below.

Consider the hypothetical scenario of an operator who owns six similar pipelines, and has a total integrity budget of $6 million. Assume that the ILI cost is $0.5 million per line. Figure 3 illustrates the case where a ‘conservative’ assessment is applied. A large number of false positives cause unnecessary

• When a ‘conservative’ flaw-assessment model is used make decisions on digs and repairs, a large number of anomalies are remediated unnecessarily. This would not be a problem if integrity budgets were infinite. However, funds spent on unnecessary repairs are not available to address real areas of concern.

The first point was addressed in a 2010 PPIM paper [1]. The second argument is explored in more detail below.

Given a pipeline that contains a number of anomalies, there are four possible outcomes for each anomaly2:

• true positive – an anomaly that poses a risk is remediated

• true negative – action on an anomaly that doesn’t pose an immediate risk is deferred

• false positive – an anomaly that doesn’t pose an immediate risk is remediated

2 These terms are traditionally used with respect to flaw detection, but the present meaning is different. In present case, the terminology of true positive, false positive, etc., pertains to the decision on whether or not to remediate a particular anomaly.

Fig.1. The relationship between the cost of achieving 100% and knowledge about the condition of the pipeline.

Fig.2. Reliability of a pipeline (or a system of lines) versus knowledge, given a fixed integrity budget.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 9

Fig.3. Allocation of resources when anomalies are assessed by a ‘conservative’ methodology. Repairs and remediation of false positives result in the integrity budget being consumed without inspecting Lines 5 and 6.

Fig.4. Allocation of resources when an advanced assessment is applied. Reduction or elimination of false positives results in an optimum use of resources.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering10

Fig.5. Finite-element model of an OD crack in an ERW seam. The model is ¼ symmetric.

Fig.6. Critical flaw curves for three failure models, (top) Charpy V-notch specimen; (bottom) fracture-mechanics’ specimen.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 11

Recently, Keifner [3] has published a modified version of the ln-sec equation. This modified model does not reflect the significant advances in the field of fracture mechanics that have occurred since the 1970s. Rather it merely incorporates a correction factor that eliminates the non-physical trend for zero-depth flaws. The modified ln-sec model represents an improvement over the original method, but more rigorous and accurate approaches are available.

The American Petroleum Institute (API) and the American Society for Mechanical Engineers (ASME) have jointly published a standard that covers fitness-for-service assessment of pressure equipment, including pressure vessels, storage tanks, piping, and pipelines [4]. This standard, informally

digs and repairs, which results in the budget being consumed without inspecting Lines 5 and 6. Figure 4 illustrates how advanced assessment can result in an optimum use of the integrity budget. The advanced assessment costs slightly more than the traditional approach, but a significant cost saving results from reducing the number of digs and repairs in each line. Consequently, sufficient budget is available to ensure reliable operation of all six lines.

Failure models for cracks

The assessment of cracks and other planar flaws is a perfect illustration of new technology offering a distinct advantage over traditional pipeline-centric models. A comparison old and new approaches is given below.

The pipeline industry has traditionally used the NG-18 equation to estimate burst pressure and critical flaw size. This approach, which is also known as the ln-sec model, was developed in the early 1970s [2] and is a semi-empirical equation that was calibrated to burst-test data. This equation contains a known glitch, in that it is overly conservative for long, shallow flaws. Most of the flaws in the original burst test dataset were less than 6 in long, but a few burst tests on pipes with longer flaws were outliers in the empirical correlation. The problem with this model can be seen more directly when a plot of burst pressure versus flaw length is generated: if the flaw depth is set to zero, the calculated burst pressure decreases with flaw length, which obviously does not comport with reality.

Fig.7. Laboratory specimens for measuring material toughness.

Fig.8. Close-up of a Charpy notch at an ERW bond line.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering12

for longer flaws. This trend is expected, since the original ln-sec equation was calibrated to burst-test data for mostly shorter cracks; the ln-sec method has serious shortcomings for longer cracks, as previously discussed. Note that the original ln-sec model predicts burst for a 14-in flaw with zero depth: in other words, a superficial scratch 14-in long should fail the hydrostatic test, according to the original ln-sec model. To put this in perspective, a 1500-psi test pressure in a 16-in X52 pipe with 0.25-in wall thickness corresponds to 92% of SMYS, so the original ln-sec model is obviously incorrect. The modified ln-sec model gives improved predictions for long flaws, but it still underestimates the critical crack depth for flaws greater than 4 in long in this case.

Recently, PRCI completed a research project whose aim was to develop an improved failure model for cracks in longitudinal seam welds [5]. This project consisted of a 3D finite-element parametric study, where a range of crack sizes, pipe dimensions, and material properties were analysed. The FEA results were then fit to a series of equations. Thus the new PRCI crack-assessment model should approximately match the FEA-based critical flaw curve in Fig.6.

Quantifying toughness of pipeline steels

Both the original and modified ln-sec equations use Charpy impact energy to characterize the toughness of the pipeline steel. Since this model was first published, more sophisticated material tests based on the principles of fracture mechanics have been developed and standardized. Figure 7 shows photographs of the Charpy V-notch and fracture-mechanics’ specimens: a key difference between the two specimen types is that fracture-mechanics’ specimens contain sharp fatigue cracks. This distinction is particularly important when testing ERW seams. The photograph in Fig.8 is a close-up of a Charpy notch at an ERW bond line: the notch radius is significantly larger than the width of the bond line, and consequently the measured toughness reflects an average for the various microstructures in the vicinity of the bond line rather than the material at the bond line. The sharp fatigue crack in the fracture-mechanics’ specimen can be precisely located in the relevant microstructure, so it is better able to discriminate the variations in material properties from the bond line to the parent metal.

Real-time pressure cycle fatigue analysis

Seam-welded pipelines that are in cyclic service can experience fatigue failure if not properly managed. Planar flaws that are introduced at manufacture can grow over time due to pressure cycling, and eventually a growing crack will lead to a leak or rupture if it is not remediated.

Pressure-cycle-fatigue analysis (PCFA) is a technique that has been used by the pipeline industry to manage the risk associated with seam-weld flaws that may grow in service.

known as API 579, includes assessment procedures for crack-like flaws. A variety of analysis options are available in API 579, ranging from simplified methods that can be implemented in a spreadsheet, to sophisticated computer simulations. The most-accurate predictions of burst pressure and critical crack size are obtained from 3D finite-element analyses (FEA); Fig.5 shows a typical FEA model of a crack in a longitudinal seam weld.

Figure 6 compares a critical-crack-size curve inferred from 3D FEA with curves computed from the original and modified ln-sec equations. Curves of this type are used to infer the maximum flaw sizes that survive a hydrostatic test: cracks that fall below the critical curve are expected to survive the test, while flaws above the curve are expected to result in a leak or rupture. The three curves agree reasonably well for cracks less than 4 in long in this case, but there is a significant difference

Fig.9. Overall architecture of Quest Integrity’s automated PCFA system.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 13

Pressure data are typically collected at pumping stations (in liquid lines) and stored in a PI data historian or similar system. Periodically, pressure readings at discrete time intervals are exported to a CSV file or spreadsheet. These data are processed through a rain-flow cycle-counting algorithm, which quantifies the number and magnitude of pressure cycles in the form of a histogram. The histogram is then input into a fracture-mechanics’ model to predict the growth of actual or postulated flaws in the pipeline. The PCFA is used to make decisions on the retest interval or re-inspection interval in cases where the integrity-management plan calls for hydrostatic testing or ILI, respectively.

The PCFA process is fairly time consuming and labour intensive. In a typical case, a pipeline operator sends pressure data to a consultant, who then submits a report to the operator two or three months later. A PCFA is usually performed annually because more-frequent intervals are not practical.

Quest Integrity is currently developing a software system for automatically performing PCFA, and Fig.9 illustrates the system architecture. At initial set-up for a given pipeline, the user enters basic data, such as pumping station locations, pipe dimensions, elevations, and material properties. Once the system is online, it periodically imports pressure data from the PI data historian, and then processes it through the rain-flow and fracture-mechanics’ algorithms. Both the pressure data input and the processed output are stored in a database and reports are generated at regular intervals based on user settings. Because the system is automated, it is possible to obtain virtually real-time updates on pressure cycling. For example, an operator may choose to generate PCFA reports on a weekly or monthly basis. It is also possible to track the growth of thousands of flaws in multiple pipelines.

Conclusions

1. Heightened public awareness of pipeline safety issues and increasing regulatory pressure mean that the maintaining the status quo on integrity-management is not an option.

2. Although the pipeline industry currently relies on flaw-assessment technology that is over 30 years old, improved models are available.

3. Using traditional approaches that are ‘conservative’ may actually lead to more risk because integrity budgets are finite, and money spent on unnecessary digs and repairs is not available to address more-critical areas.

4. Failure models for crack-like flaws that are based on finite-element analysis (FEA) are vastly superior to the ln-sec method, which has traditionally been used in the pipeline industry. A new PRCI method based on a curve-fit of FEA results provides a simpler alternative to performing custom FEA for each situation that is encountered.

5. Fracture-mechanics’-based toughness tests are more reliable than Charpy tests, especially for ERW seams.

6. Pressure-cycle-fatigue analysis (PCFA) has traditionally been time consuming, so it is typically performed no more than annually, and at a limited number of locations in a pipeline. However, an automated PCFA system that is currently under development makes real-time monitoring of pressure cycling and crack growth at thousands of locations feasible.

References

1. T.L.Anderson, 2010. Advanced assessment of pipeline integrity using ILI data. Pipeline Pigging and Integrity Management Conference, Houston, February.

2. J.F.Kiefner, W.A.Maxey, R.J.Eiber, and A.R.Duffy, 1973. Failure stress levels of flaws in pressurized cylinders. ASTM STP 536, American Society for Testing and Materials.

3. J.F.Kiefner, 2001. Modified equation helps integrity management. Oil and Gas Journal, 6 October, pp 64-66.

4. API and ASME, 2007. API579-1/ASME FFS-1, Fitness-for-service. Jointly published by the American Petroleum Institute and the American Society for Mechanical Engineers, June.

5. G.G.Chell, 2008. Criteria for evaluating failure susceptibility due to axial cracks in pressurized line pipe. PRCI Project MAT-8 Final Report, December.

Sample

copy

not fo

r dist

ributi

on

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 15

THE NATIONAL Transportation Safety Board’s recommendations to pipeline operators and regulators, based on its investigation of the 2010 San Bruno, California, gas pipeline incident, has focused industry

attention on validation of gas pipeline maximum operating pressures of pipelines installed prior to federal or state regulations. The industry is in the process of understanding the viability of continued current operation of ‘grandfathered’ pipelines and pipelines that lack records necessary to give full confidence of the quality of installation. A full understanding of the implications requires knowledge of how industry practices and code requirements in the areas of pressure testing and recordkeeping have evolved over time.

This paper1 explores these issues in gas pipeline standards and regulations, historically and currently, nationally and in the state of California. It is likely that similar considerations may develop in other states. The paper describes the evolution of pipeline pressure testing requirements, what records have been specifically required, how those records relate to establishing the maximum allowable operating pressure (MAOP) of a pipeline, why so-called ‘grandfathered’ pipelines have existed, and the significance of recently articulated criteria for records accuracy.

This paper was presented at the 25th annual Pipeline Pigging and Integrity Management conference held in Houston in February 2013, and organized by Clarion Technical Conferences and Tiratsoo Technical.

*Corresponding author:tel: +1 614 410 1602email: mrosenfeld@kiefner.com

1. Based substantially on Application No. 11-11-002, Rebuttal Testimony of Southern California Gas Company and San Diego Gas & Electric Company in Support of Proposed Natural Gas Pipeline Safety Enhancement Plan, Before the Public Utilities Commission of the State of California, prepared testimony of Michael J. Rosenfeld, July 18, 2012.

1 Kiefner & Associates Inc./Applus-RTD, Worthington, OH, USA2 Southern California Gas Co, Los Angeles, CA, USA

Pressure testing and recordkeeping: reconciling historic pipeline practices with new requirements

by Michael J Rosenfeld*1 and Rick W Gailing2

THE INVESTIGATION by the National Transportation Safety Board (NTSB) of the 2010 gas pipeline incident

in San Bruno, California, determined that the pipeline rupture originated in one of several short ‘pups’ that was not manufactured in a manner consistent with known linepipe manufacturing processes. Who manufactured the pups, why they were manufactured as they were, how they came into the operator’s possession, and how they came to be installed, was not established conclusively and may never be known beyond speculation. Based on its findings, the NTSB issued recommendations to the industry to review its

pipeline records to better understand whether they support their respective MAOPs, particularly for ‘grandfathered’ pipelines operating in Class 3 and 4 locations and in high-consequence areas (HCAs) in Class 1 and 2 locations.

In response, the California Public Utilities Commission (CPUC) issued Decision (D.)11-06-017 directing all operators of natural gas pipelines in the state of California to replace or pressure test all pipelines that have not been pressure tested to “modern standards”2. This is interpreted to include all grandfathered pipelines, as well as pipelines having insufficient records to substantiate that a pressure test was conducted. Which pipelines must be included in the operator’s testing plans is less clear than it at first seems, since pressure-testing requirements and record-keeping requirements have evolved over time. In the proceedings before the CPUC, a broad spectrum of difficult questions have been raised, such as:

• Does a verified test to lesser performance and documentary requirements than the 1970 49 CFR 192, Subpart J acceptably qualify the MAOP?

2. http://docs.cpuc.ca.gov/PUBLISHED/NEWS_RELEASE/136948.htm

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering16

development of the standard. This desire was further stimulated by a widely publicized gas distribution system incident in Rochester, New York, in 1950, and concern for a consequent regulatory response3. In response, Section 8 of the 1951 B31.1 addressing only natural gas pipelines was approved and published as a stand-alone document in 1952. Although it drew largely on the technical requirements for gas and air piping in Section 2, and selected fabrication details from Section 6 of the 1951 B31.1 standard, the publication separately from B31.1 provided the platform for further development of a more comprehensive pipeline-specific technical standard.

The 1955 edition of Section 8, designated B31.1.8, represented a significant technical advance in requirements for natural gas transmission and distribution piping systems. It incorporated a risk-informed design basis in the form of a location-class scheme based on the density of development near the pipeline, significantly more guidance relevant to the design and installation of cross-country transmission pipelines and gas distribution systems, and rigorous new pressure-testing requirements. It was thought that a well-conceived technical standard for pipelines could be useful to state pipeline-safety regulations4. Elements of the 1955 standard are still evident in the current edition. The standard was revised and republished as B31.8 in 1958, 1963, and 1968 prior to the issuance of pipeline-safety regulations by Department of Transportation (DOT) in 1970. Addenda were issued in some years between editions, and B31.8 continued to be revised and periodically republished from 1974 to the present time.

The CPUC enacted General Order 112 (GO 112) in 1961, specifying minimum rules for the design, construction, operation, and maintenance of natural gas pipelines within the state of California; 14 other states had also established regulations for natural gas pipelines by that time. GO 112 incorporated substantial portions of the 1958 edition of B31.8, omitted portions in conflict with CPUC requirements, and added language where necessary to accomplish its goals as the utilities’ regulator. The incorporation of suitable portions of B31.8 into GO 112 was consistent with ASA’s purpose in publishing its standard. Subsequent editions of GO 112 in 1964 and 1967 incorporated significant portions of the most-current edition of B31.8 until DOT issued its gas pipeline regulations in 1970. Subsequently, GO 112 incorporated DOT regulations.

In response to a significant gas pipeline incident in Natchitoches, LA, in 1965, the Natural Gas Pipeline Safety Act (NGPSA) of 1968 authorized DOT to create the Office of Pipeline Safety (OPS, predecessor to the present Pipeline and Hazardous Materials Safety Administration – PHMSA),

• Could a probable test supported by incomplete or indirect documentation suffice?

• Is an operator “imprudent” for failing to ensure an unbroken chain of documentation from a time prior to when specific record-keeping requirements came into effect? What if the records’ loss occurred after record-keeping requirements were in effect?

• Are complete pressure testing records necessary to operate a pipeline in a prudent and safe manner, and if not, should an operator then be held responsible for failing to maintain records intact, retrospectively?

• Is it possible to not follow voluntary standards while still remaining prudent?

• Do existing regulations acknowledge and accept the possibility of incomplete records?

Whether rate payers or shareholders should bear the burden of applying new criteria for MAOP validation of gas transmission pipelines is a current issue in California and depends on the answers to these and other questions. The authors are not aware of this issue arising in other states at this time; however, PHMSA is evaluating whether similar requirements may be appropriate for the interstate gas transmission pipeline system. It may be only a matter of time before the question of who pays for what arises outside of California.

As regulators and operators contemplate what criteria should apply retrospectively, and at whose expense any corrective action is taken, the history of practices with respect to pressure testing and recordkeeping should be recognized and accounted for. This paper is intended to provide information toward that objective.

Standards and regulations development

Evolutionary steps

The evolution of modern gas pipeline standards can be traced to the B31 Code for Pressure Piping, Standard B31.1, first published as a tentative standard by the American Standards Association (ASA), a predecessor to the American National Standards Institute (ANSI), with sponsorship of the American Society of Mechanical Engineers (ASME). This standard covered the materials, design, and fabrication of piping systems with industry-specific sections for power piping, gas and air piping, oil piping, and district-heating piping. The scope of Section 2 covering gas and air piping systems included city-gas distribution systems, and cross-country gas pipelines and compressor stations. ASA B31.1 was updated and republished in 1942, 1947, and 1951.

The gas pipeline industry desired to further develop the standard to better address the technical requirements for buried natural gas pipelines, which differ substantially from the technical issues associated with piping systems within power and process facilities that tended to dominate technical

3. L.L.Elder, 1995. The history of the gas piping standards/technology committee. GPTC/GPSRC 25th Anniversary Meeting, 17-20 July.4. F.A.Hough, 1954. The gas industry has approved its new safety code. Gas Magazine, November.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 17

embodied in Title 49 of the Code of Federal Regulations Part 192 (49 CFR 192) was intended to prescribe the level of performance that must be met, while leaving industry free to develop the specific means of meeting the prescribed level of performance10. In other words, regulations prescribe “what” while industry standards describe “how”.

Even though technical provisions in the regulations (GO 112 and 49 CFR 192) have their origins in technical provisions in the standard (B31.8), there are many areas in which the regulations and the standard do not agree, both historically and today. These include matters of pressure design, material properties, hydrostatic pressure test requirements, valve spacing, record keeping, and various elements of operation and maintenance.

History of gas pipeline pressure-test requirements

Advent of hydrostatic testing

Hydrostatic pressure testing11 is a standard practice for commissioning a pipeline today, but this was not always the case. The concept of pressure testing as a means of establishing the ability of pipe to safely contain pressure in operation was adopted from the vessel industry, which had begun to implement that practice prior to 1900. However, pressure testing with water a natural gas pipeline that is many miles long is much more difficult than filling a vessel with water; these differences posed serious challenges to early pipeline operators, for a couple of reasons. One is that the large quantity of clean water necessary to fill a cross-country pipeline was difficult to obtain and manage in any location, and particularly so in dry-climate regions where many early large pipelines were constructed. The second problem was dewatering, since the methods and tools to accomplish that had yet to be developed. (This was also the case with other complications such as bleeding trapped air, rupture isolation/containment, and refilling or transferring test water from section to section.) Similar limitations affect gas distribution systems: the quantities of water required are still large, the networked nature of the systems complicates dewatering, and residual water in distribution piping is a problem for customers. Consequently, through the 1940s, if a pressure test was performed at all, it was usually accomplished using the transported commodity, natural gas in the case of gas pipelines, or crude oil or petroleum products in the case of liquid transmission pipelines. Owing to concerns for the consequences of a failure when testing with product (loss of product in the case of liquids, and loss of extensive quantities of pipe due to fracture propagation in the case

enact interim safety standards within three months consisting of existing state safety standards, and issue federal pipeline safety regulations within 24 months. Interim regulations comprised of existing standards were imposed until complete regulations were adopted as Part 192, effective from 1 July, 1970. A review of the technical content of Part 192 shows a clear influence of B31.8, with revisions in language and additional content for clarity and enforcement. Part 192 does not make specific reference to B31.8 on most technical matters because it was the belief of the then-director of OPS that a regulation may be potentially compromised by referring to industry-developed standards5.

The NGPSA also required the establishment of the Technical Pipeline Safety Standards Committee (TPSSC). The purpose of the TPSSC was to review all proposed pipeline regulations for “technical feasibility, reasonableness, and practicality”6.

In 1970, by agreement with OPS, ASME began publishing language from Part 192, supplemented with practices from B31.8 and other sources, to guide operators in meeting the regulatory requirements. The publication was prepared by the Gas Piping Standards Committee (GPSC) and was known as the GPSC Guide. In 1982, the administrative support was transferred to the American Gas Association (although it continued to be published by ASME), the committee name changed to the present Gas Piping Technology Committee (GPTC), and it acquired recognition as ANSI Z3807.

Standards are not regulations

The foregoing discussion explains the origin of present-day regulations in contemporary industry-developed standards. Standards exist to provide technical guidance and promote uniformity in practices. In particular, ASME B31.8 was intended to be a statement of what is generally accepted to be good practice8, written by engineers for an audience of other engineers, designers, managers, and regulators. Hence the standard cannot include practices that are not generally accepted even if they are superior, nor should it include practices that are considered unnecessary. The requirements set forth in B31.8 are considered adequate under normally encountered conditions, while unusual conditions are not specifically provided for. Also, the standard is not law. The standard was intended to improve public safety through compliance by pipeline operators voluntarily and in good faith9.

A regulation is a legally enforceable requirement as a government response to a problem. Regulations are written by regulators for an audience of inspectors and the regulated entities, for the purpose of enforcement. The regulation

5. W.C.Jennings, 1971. The regulator’s handbook. June.6. Federal Register, 1970. Vol. 35, No. 161, Wednesday, 19 August, pg. 13256.7. Elder idem.8. F.A.Hough, 1955. The new gas transmission and distribution piping code (ASA B31 Section 8). Gas Magazine, Series in 8 Parts, January through September.9. Hough, idem.

10. Jennings, idem.11. ‘Hydrostatic testing’ means conducting a pressure test of a pipe or vessel using water as the pressurizing medium. However, the term has often been used historically and today to refer to pressure testing using any fluid including gaseous media such as air, nitrogen, or natural gas. In this document, ‘hydrotesting’ is used in the incorrect but colloquial form to indicate a pressure test using any fluid except where a distinction is made with respect to the test medium.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering18

benefits, limitations, and mechanics of hydrotesting20-22. Over time, other operators began to adopt the practice of pressure testing with water to higher stress levels than had previously been customary.

The evolution of test requirements for commissioning a new pipeline system as they pertain to transmission pipelines constructed from steel pipe is summarized briefly below. Testing requirements are not discussed herein for low- and high-stress distribution piping, mains, and service lines; piping fabricated from plastic or cast iron pipe; testing for purposes of uprating; and testing to accommodate changes in location classes. The reason for omitting these requirements is they introduce significant complexity in details that are not central to the issue at hand. Figure 1 gives a timeline for major changes in pressure testing requirements.

The B31.8 standard, predecessors and sequels

Two important eras can be defined with respect to pressure-testing requirements in B31.8: pre-1955 and 1955-and-later, because the 1955 edition of the Code marks the first time that pressure testing a pipeline after construction was made a requirement for complying with the standard. The basis for this conclusion is discussed below.

ASA B31.1 prior to 1955

Section 2 of the 1935 B31.1 defined two categories of pipe based on location: Division 1 piping was air or gas piping constructed within power plants, gas plants, or manufacturing plants, or within the boundaries of cities or villages; Division 2 piping was constructed in compressor stations, installed cross-country, or outside boundaries of cities or villages. Within both divisions, before installation, valves and fittings were to be “capable of withstanding a hydrostatic

of testing with natural gas), operators typically limited test pressures to between 5 psig and 50 psig above, or at most 10% above, their intended operating pressure12-15.

The first large-scale use of proof testing long-distance gas pipelines with water was carried out by the Texas Eastern Transmission Corp. (TETCO) in 195016. In 1947, TETCO acquired the two War Emergency Pipelines built to transport crude oil and fuel from Texas to New Jersey during World War II, and converted them to transport natural gas. TETCO experienced many service failures due to original pipe manufacturing defects that may have enlarged while in petroleum transportation service, and also due to corrosion because parts of the line were installed uncoated to save time. In 1950, TETCO completed an ambitious programme to revalidate the integrity of the pipelines by pressure testing them with water to levels well above the MAOP and in some cases up to yielding. TETCO was able to do this because it had already developed cleaning pigs that were inserted into traps and propelled by gas pressure to sweep accumulated liquids out of the line as part of the process of converting the lines from liquid to gas17. Although the company experienced hundreds of pipeline breaks during testing18, the tested pipelines were reliable in subsequent years and portions of them are still in service today19. As a result of TETCO’s experience, the industry performed scientific studies between 1953 and 1968 to better understand the

12. J.F.Kiefner, 2001. Section 4, Hydrostatic Testing. GRI guide for locating and using pipeline industry research. Prepared by Kiefner & Associates, Inc. for Gas Research Institute, GRI-00/0192.04, March.13. Hough, 1955, idem.14. W.B.McGehee, 1998. Maximum allowable operating pressure (MAOP) background & history. Report for Gas Research Institute, 5 March.15. T.M.Shires and M.R.Harrison, 1998. Development of the B31.8 Code and federal pipeline safety regulations: implications for today’s natural gas pipeline system. GRI-98/0367.1, December.16. C.J.Castaneda and J.A.Pratt, 1993. From Texas to the east: a strategic history of Texas Eastern Corporation, Texas A&M University Press.17. Castaneda and Pratt, idem.18. S.A.Bergman, 11974. Why not higher operating pressure for lines tested to 90% SMYS? Pipeline and Gas Journal, December.19. Kiefner, idem.

20. McGehee, idem.21. Kiefner, idem.22. Bergman, idem.

Fig.1. Timeline of pipeline hydrostatic pressure-test requirements.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 19

not required for Division 2 piping. Most pipeline operators made this same interpretation until such time as testing became a clearly stated requirement in the 1955 edition27.

The 1942 edition slightly revised the post-installation testing to requirements to be “capable of withstanding a test pressure” of 150% of the service pressure for Division 1 piping, or 50 psig greater than the maximum service pressure for Division 2 piping. A test after installation “may be made with air or gas” which “need not exceed 120% of the maximum allowable working pressure” for Division 1 piping, or “shall not exceed 120% of the maximum allowable working pressure” for Division 2 piping. Clearly, these test limits conflict with an interpretation that a requirement to be “capable of withstanding a pressure test” is synonymous with an actual requirement to carry out such a test, to 150% of the service pressure. The duration of a pressure test, if performed, was not specified. The Code stated that “where an actual internal pressure test is made” (implying the existence of places where an “actual internal pressure test” was not made), the test pressure should be maintained for long enough to inspect the joints and connections. This requirement implies that the test’s primary purpose was a leak test of flanged, threaded, or welded connections, not a proof of the strength of the pipe. Nowhere were working pressures established on the basis of a post-installation test. They were based on the mill test or an engineering calculation.

The 1947 Addendum to the 1942 B31.1 standard did not change the testing requirements for gas and air piping. The 1951 B31.1 standard slightly revised the post-installation testing provision to read “where an internal fluid pressure test is made, it shall not exceed” 150% of the maximum allowable working pressure for Division 1 piping, and for Division 2 piping, 120% of or 50 psig greater than the maximum allowable working pressure, whichever was greater. The language still only required a capability for withstanding a test, not the performance of an actual test. If a test was performed using any fluid (liquid or gaseous), the maximum test level was limited, and no minimum test duration was prescribed other than that it be long enough to inspect joints and connections for leaks.

With the 1952 stand-alone gas-only Code, the Division 1 and Division 2 designations were replaced with description of the systems in paras 807(c)(2)(a) and 807(c)(1)(a), respectively. Pressure-testing requirements were found in Chapter 5, ‘Requirements after installation’, and were identical to the 1951 Code. However, for cross-country pipelines working pressures did depend on whether a post-installation test was performed. For pipelines installed prior to 1952, the allowed working pressure was either 80% of the pipe mill-test pressure or a maximum of 72% of the yield strength (Y) multiplied by a joint efficiency factor (E). For steel pipe outside of compressor stations, the allowed working pressure was either 80% of

shell test” to designated pressures based on pressure-rating classes similar to present-day pressure ratings for valves and flanged fittings. Pipe used in Division 1 service was also to be “capable of meeting the hydrostatic test requirements” contained in listed pipe-product specifications, but pipe used in Division 2 was to be “subjected to and safely withstand a mill pressure test” in accordance with the pipe-product specification (but not in excess of 90% of the yield point or yield strength of the material).

Some parties have suggested that the words “capable of withstanding” a pressure test to some level mean that a post-construction test was required23. However, the Code used clearly different language to indicate an actual test requirement (for example, pipe in the mill) as opposed to a capability for withstanding a test, which is a design requirement that is met through specifying an adequate combination of wall and metal strength. This is consistent with language used in contemporaneous standards for wrought fittings that also required that items be “capable of withstanding a pressure test to 1.5 times the working pressure”, followed immediately by language stating that an actual test of each item was not required24, 25. This is further supported by recent interpretations of similar pressure-capability language in PHMSA regulations26.

After installation, Division 1 piping systems containing welded joints were to be “capable of withstanding a hydrostatic test” to 1.5 times the service pressure, with the test to be applied “where practical”. It was further stated that “if a test is performed” welds were to be struck by hammer blows to jar them during the hydrostatic pressure test, something that can only be done with exposed piping, not a buried pipeline. With Division 2 piping, there were no pressure-test requirements post-installation because such a test was deemed unnecessary: fittings were designed to be as strong as matching pipe, and pipe was required to have been tested at the mill. The working pressure was 80% of the pipe mill test pressure, or a percentage of the yield strength calculated as the seam-joint efficiency factor divided by 1.4.

In no case was the working pressure established in the 1935 Code on the basis of a post-installation pressure test. The 1935 Code was understood to mean that testing of the pipe after installation was discretionary for Division 1 piping and

23. Notably, but not limited to CPUC Division of Ratepayer Advocates (DRA) in briefs commenting on proposed pipeline safety enhancement implementation plans submitted by PG&E, Southern California Gas Co, and San Diego Gas & Electric.24. American Standards Association, 1940. Steel butt welding fittings. ASA B16.9.25. American Society for Testing and Materials, 1940. Standard specifications (tentative) for factory-made wrought carbon-steel and carbon-molybdenum-steel welding fittings. ASTM A234.26. J.A.Gale, 2005. Office of Hazardous Materials Standards, PHMSA, letter to M.Fox, Chemical Accidents Reconstruction Services, Inc., response to inquiry concerning 49 CFR 173.306, Ref. No. 04-0203, 18 March 18. “Section 173.306(a)(3)(ii) requires a metal aerosol container to be capable of withstanding without bursting a pressure of one-and-one-half times the pressure of the content at 130 degrees F. The [Hazardous Materials Regulations] do not specify a method for demonstrating that the container is capable of withstanding the specified pressure. You may demonstrate that the container meets the standard by testing or design specifications.” 27. Hough, 1955, idem.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering20

were observed, the pipe was not operated at more than 80% of the test pressure, and the pipe had a seam-joint efficiency factor of 1.00.

Paragraph 841.5 ‘Safety during tests’ advised the user to give due regard to the safety of employees and the public during pressure tests. When air or gas is the test medium, steps were required to remove persons not involved in conducting the test when the test hoop-stress level exceeds 50% SMYS.

Pressure-test requirements in the 1958, 1963, 1967, 1968, 1975, and 1982 standards and their addenda were the same as in the 1955 standard.

The 1984 Addendum to the 1982 edition specified that the pressure test of all piping intended to operate at hoop-stress levels of 30% SMYS or greater be held for a minimum duration of two hours. This was the first occurrence of a specified minimum test duration in B31.8. Test levels were the same as previously. The pressure test requirements in the 1986 edition were the same as the 1984 Addendum.

The 1989 standard introduced a new operating-stress level in excess of the traditional maximum operating-stress level of 72% SMYS in Class 1, up to a maximum operating stress of 80% SMYS. Pipe in this category was referred to as Class 1, Division 1, and was to be pressure tested to a minimum stress level of 100% SMYS, with water as the only permitted test fluid. The traditional maximum operating stress of 72% SMYS was referred to as Class 1, Division 2. The same test requirements applied for Class 1, Division 2, and for Classes 2, 3, and 4 as in previous editions. The requirements from the 1989 edition remained unchanged in the 1992, 1995, 1999, 2003, and 2007 editions.

Important revisions were made to the pressure-testing requirements with the 2010 edition. The minimum test ratio for Class 1, Division 2 pipe (with a maximum operating stress level up to 72% SMYS) was raised to 1.25, regardless of test medium, and the minimum test ratio for Class 3 and 4 piping was raised to 1.50. Also, significant additional guidance on test planning, execution, and risk mitigation is provided.

CPUC General Order 112 and sequels

The CPUC introduced regulations governing the design, construction, operation, and maintenance of natural gas pipelines within the state of California under General Order (GO) 112, first issued in 1961. The pressure-testing requirements in GO 112 are discussed below. Of course, a year other than 1961 may represent a regulatory threshold in other states.

CPUC General Order 112 incorporated significant portions of the 1958 B31.8 standard, with certain changes to the pressure-testing requirements. Among those changes were: the pressure testing requirements were extended to pipe operating at hoop stresses of 20% or more of SMYS (rather

the pipe mill test pressure or 60% of Y times E for pipe not tested after installation, or 85% of Y times E for pipe tested after installation. Working pressure limits dependent upon whether a post-installation test was performed conclusively indicates that a post-installation test was not a requirement.

B31.8 post-1955

The 1955 Code introduced the concept of four location class factors based on density of land development adjacent to the pipeline, each with different maximum allowable operating stress levels, and different pressure-test requirements following installation. The precise definitions of the classes in terms of building or dwelling counts, and the dimensions of the reference area, were somewhat different from today’s, but the intended meanings of the classes were the same as today (for example, Class 1 being rural, and so on) and the allowed operating stresses were also the same.

Testing requirements were stated in para. 841.3 ‘Testing after construction’. All mains and services were to be tested, except tie-ins where individual test sections were eventually joined after testing. This was the first time in the gas-piping standard that testing after installation became a firm requirement, but no minimum test duration was specified. The design requirement for a capability to withstand a pressure test was moved to Chapter 3 ‘Piping system components and fabrication requirements’, para. 831 ‘Piping system components’, where components were to be designed to withstand the system pressure test without failure, leakage, or impairment of their serviceability. Moving the “capability to withstand” language to Chapter 3 further substantiates the fact that a post-construction test was not intended or required by that wording.

Pressure-test requirements were given in para. 841.4 ‘Test requirements’. All pipelines and mains to be operated at a hoop stress of 30% or more of the specified minimum yield strength (SMYS) “shall be given a field test to prove strength after construction and before being placed in operation”. Piping installed in Class 1 areas was to be tested with air or gas to 1.1 times the maximum operating pressure, or hydrostatically tested to at least 1.1 times the maximum operating pressure; piping installed in Class 2 areas was to be tested with air to 1.25 times the maximum operating pressure, or hydrostatically tested to at least 1.25 times the maximum operating pressure; and piping installed in Class 3 and 4 areas was to be hydrostatically tested to at least 1.4 times the maximum operating pressure.

The hydrotest requirement for Class 3 and 4 piping was waived if the ground temperature at the time of the test was at, or might fall below, a temperature of 32˚F, or water of satisfactory quality was not available in sufficient quantity. In that case, an air test to 1.1 times the maximum operating pressure could be performed and the test pressure ratio of 1.4 did not apply. Air testing of Class 3 and 4 pipe was allowed in any case provided that: strict hoop-stress limits

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 21

in accordance with Subpart L. The origin and basis are described in the Preamble to the first full issuance of Title 49 ‘Transportation’ published in the Federal Register29.

In the original proposal for Part 192, no recognition was given for piping installed prior to 195530 on the basis of very loose testing requirements, and for piping already operating at hoop-stress levels greater than 72% SMYS. The Federal Power Commission (FPC) wrote to OPS pointing out that there were thousands of miles of pipeline already in service, installed in accordance with prevailing standards and practices, that could not continue operating at their-then current levels and comply with the proposed regulations. The FPC also stated that based on a review of the operating records of interstate pipelines, no improvement in safety would be gained by reducing the operating pressures of existing pipelines “which have been proven to be capable of withstanding present operating pressures through actual operation”. In response, OPS included a ‘grandfather’ clause to permit continued operation of pipelines at the highest operating pressure the pipeline had experienced in service during the five years preceding 1 July, 1970 (even if the pipe had previously been subjected to a hydrostatic-pressure test to qualify a higher MAOP but the pipe had not operated at that level during the specified five-year interval).

GO 112 already had set a regulatory precedent for the grandfathering of untested pipelines. Gas pipelines placed in service after 1 July, 1961, were required to be pressure tested, but those installed before that date were exempted from pressure test requirements31. The CPUC was likely guided by provisions in para. 804.6 of the 1955 B31.1.8 and its sequels that the standard was not intended to be applied retroactively to existing facilities insofar as design, installation, establishing the operating pressure, and testing were concerned. Consistent with these exemptions, the concept that new or evolving requirements concerning materials, design, construction, and the establishment of the MAOP are not retroactive to existing facilities that are already in operation was recognized in the federal pipeline regulations from the outset. This concept is embodied in para. 192.13 and is fully expressed in the discussion of the retroactive effect on existing pipelines in the Preamble to Part 19232.

than 30% or more of SMYS); the test margin for Class 1 pipelines was increased to 1.25; the test margins for Class 3 and 4 pipelines were increased to 1.5; and the test pressure was required to be maintained until it was stabilized and for a period of not less than one hour.

GO 112-A and GO 112-B were published in 1964 and 1967, respectively. The requirements on pressure testing were the same as those in 1961.

Following the issuance of 49 CFR 192, the 1971 GO 112-C replaced content from B31.8 with content from Part 192 with some additional requirements. The content from Part 192, Subpart J ‘Test requirements’, was incorporated verbatim. The 1979 GO 112-D incorporated the content from Part 192 issued in 1978. Since Subpart J remained relatively static in subsequent years, few changes in actual requirements occurred in GO 112.

49 CFR 192

The first full set of federal pipeline regulations was issued in 1970. Subpart J ‘Test requirements’, paras 192.501 to 192.517, set forth requirements for pressure testing of pipelines after construction. An important new requirement relative to those contained in preceding or contemporaneous editions of B31.8 or GO 112 was para. 192.505(c) stipulating maintaining the strength test pressure for at least eight hours. As originally proposed, the specified minimum test duration was 24 consecutive hours, a practice that was observed by some – but not all – pipeline operators. This was reduced to eight hours on the recommendation of the TPSSC because there was no evidence that a longer test was a superior test28.

Aside from limitations based on maximum hoop stress levels, maximum-allowable operating pressure was based on dividing the pressure test by a minimum specified factor, given in Subpart L ‘Operations’, Clause 192.619(a)(2)(ii). For pipe installed after 11 November, 1970, test pressure ratios were 1.1, 1.25, and 1.5 in Classes 1, 2, and 3 or 4, respectively. For pipe installed and tested prior to 12 November, 1970, the test ratio for Classes 3 and 4 was 1.4, based on the requirements in the interim Federal standard between 1968 and 1970, which were the same as B31.8, and based on B31.8 being the de facto national standard prior to 1968 (except in California and perhaps a few other states).

These requirements for testing after construction have remained static in subsequent years.

Grandfathered pipelines

The term ‘grandfathered’ pipelines refers to those pipelines for which the operating pressure was established on the basis of operating history rather than pressure testing

28. Fed. Reg., pg. 13255.

29. Fed. Reg., pg. 13248-13276.30. In its comments to the original docket, the TPSSC referred to 1952 as the first year that the ASME B31.1.8 gave minimum test pressures. However, that new test requirement occurred in 1955, not 1952. The TPSSC comments are interpreted accordingly herein.31. CPUC, Rulemaking 11-02-019, Findings of Fact No. 5, pg. 27.32. Fed. Reg., pg. 13250, on the subject of the retroactive effect on existing pipelines, quotes the Natural Gas Pipeline Safety Act, Section 3(b): “Not later than 24 months after the enactment of this Act, and from time to time thereafter, the Secretary shall, by order, establish minimum Federal safety standards for the transportation of gas and pipeline facilities. Such standards may apply to the design, installation, inspection, testing, construction, extension, operation, replacement, and maintenance of pipeline facilities. Standards affecting the design, installation, construction, initial inspection and initial testing shall not be applicable to pipeline facilities in existence on the date such standards are adopted.”

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering22

The 1955 edition was the first B31 piping standard to extend its scope beyond design, construction, and commissioning of the piping system to include operation and maintenance. Accordingly, additional recordkeeping language was introduced in Chapter V, ‘Operating and maintenance procedures’. “Basic requirements” therein stated that “each operating company having gas transmission or distribution facilities … shall: (a) have a plan covering operating and maintenance procedures…(c) keep records necessary to administer the plan properly”. Further, records “should” be made of pipeline inspections for external or internal corrosion, listing several items of potential interest, and records “should” be made covering leaks and repairs. In addition, leakage-survey records, line-patrol records, and other records relating to routine or unusual inspections “should” be kept on file as long as the section of line remained in service. The operator was required to have plans for inspecting pipe-type and bottle-type gas holders, and to keep records detailing the inspection and test work done and the results.

The terms “shall” and “should” were used throughout B31.1.8 and its sequels. “Shall” is understood to mean an action is required, while “should” is understood to mean an action is recommended but not required. Records adequate to effectively execute the pipeline operation and maintenance were required, but specific records were merely recommended and what was actually required was left to the operator. The possibility was not precluded that data different from – or in addition to – what the standard said “should” be recorded might be necessary in order to fulfil the requirement to “keep records necessary to administer” the operation and maintenance plan. Note also that the Code has historically given leave to not follow specific requirements where the operator can show by experience, testing, or analysis that an alternative is safe and reliable34. An operator could conceivably set forth a position that maintaining some kinds of record is unnecessary based on experience.

Record-keeping requirements 1961 to 1970

The 1958, 1963, and 1968 editions of ASME B31.8 did not differ from the 1955 edition with respect to record keeping. The 1968 edition included certain enhancements, such as the weld inspection requirements similar to those introduced by the 1961 GO 112, but without the accompanying weld-inspection record-keeping requirement. On the other hand, the corrosion-inspection and leak-investigation record-keeping provisions became required, not recommended.

California General Order 112 of 1961 incorporated most, if not all, of the 1958 B31.8 standard, with added requirements to better meet the objectives of the CPUC for clarification and for enforcement. Some important additions involved recordkeeping. GO 112 added minimum welding inspections based on location class, and stipulated that a record be made of the results of the tests and the inspection method used. The

History of record-keeping requirements

Record-keeping requirements prior to 1955

Record-keeping requirements specified in engineering standards for gas pipelines prior to 1955 were few, and focused on welding. The 1935 B31.1 standard required employers of welders to maintain records of their welding operators showing dates of employment, results of welding tests, and their assigned identifying mark. (Welders were required to stamp their identifying mark adjacent to welds they made on pipe.) The 1942 B31.1 standard, Appendix I, Part I, required that records of welding-procedure qualification testing and copies of the record for each qualified welder were to be kept by the manufacturer or contractor. No record-retention period was specified, and no other record-keeping requirements were expressed.

No provisions or requirements for record-keeping of any kind dealing with welding or installation were specified in the 1951 B31.1. Similarly, none were given in the 1952 B31.1, Section 8 in its entirety.

It would be reasonable to expect that a variety of documents related to the design and construction of a pipeline facility be retained long-term. However, retention of technical documents was not addressed by the engineering standards of the day. It was generally thought that a copy of the specifications under which the pipeline was built (and supplemented by commercial documents, such as contracts and purchase orders) would generally be adequate to provide evidence of the work that was done33.

Record-keeping requirements 1955 to 1961

The 1955 B31.1.8 Chapter II ‘Welding’ required that records of welding-procedure qualification tests be retained for as long as the welding procedure was in use. Further, the pipeline operator or contractor (or, presumably, whoever employed the welders) was required, during construction, to maintain a record of the welders qualified, their dates of employment, and test results.

Chapter IV ‘Design, installation, and testing’ required maintaining records showing the type of fluid used for pressure testing and the test pressure of pipelines that operate at a hoop stress of 30% or more of SMYS. The specified retention period was the useful life of the facility. This record-keeping requirement was not stated under general testing provisions applicable to all pipelines, nor under subsequent paragraphs that presented separate pressure-test requirements for pipe operating at less than 30% of SMYS but more than 100 psig, leak-test requirements for pipe operating at 100 psig or more, or leak-test requirements for pipe operating at less than 100 psig, respectively. Thus, an operator might reasonably not have retained records for tests performed in accordance with those paragraphs.

33. Hough, 1955, idem.

34. This includes but is not limited to: materials and equipment selection, fittings and components design, above-ground piping design, longitudinal stresses in buried pipelines, valve spacing, and cathodic protection criteria.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 23

or above 100 psig but below 30% of SMYS. The record must indicate the following seven items: (1) the names of the operator, the responsible employee, and the test company (if any); (2) the test medium used; (3) the test pressure; (4) the test duration; (5) pressure readings; (6) elevation variations if they are significant; and (7) leaks or failures. Such records must be retained for the useful life of the facility.

• Subpart K – Uprating, para. §192.553(b): a record is required of each investigation (for example a review of the design, and operating and maintenance history), work done, and each pressure test in connection with the uprate. The record must be retained for the life of the uprated segment.

• Subpart L – Operations, para. 192.619(a): sets forth criteria for establishing the MAOP, as the lowest of the design pressure of the weakest components or pipe based on specified attributes, the pressure obtained by dividing the post-construction test pressure by a specified factor, the highest actual operating pressure during five years preceding 1 July, 1970, for furnace butt-welded pipe a pressure equal to 60% of the mill test pressure, for other pipe a pressure equal to 85% of the highest test pressure the pipe experienced in the field or pipe mill, or the maximum safe pressure determined in consideration of the condition and operating history of the pipeline.

• Subpart M – Maintenance, para. 192.709: a record is required of each leak discovered, repair made, line break, leak survey, line patrol, and inspection of transmission pipelines for as long as the line remains in service. Records have to be retained at least until the next round of inspections (for example, five years).

• Numerous other activities (sampling of odorant, valve maintenance, vault maintenance, distribution leakage surveys, and others) must occur at specified periodic intervals. No record keeping was specified in connection with those activities.

The 1970 issuance of Part 192 added Subpart I on corrosion control, which required installation and criteria for the cathodic protection (CP) of buried steel pipelines, periodic monitoring of the effectiveness of the CP system, monitoring of internal corrosion, and monitoring of atmospheric corrosion. Record-keeping requirements as of 31 July, 1972 are discussed below:

• Subpart I – Corrosion control, para. 192.491(a): each operator was required to maintain records or maps showing the location of cathodically protected pipe, CP facilities (such as rectifiers or anodes), and other structures bonded to the pipe. Also, in para. 192.491(b), each record or map from para. (a) plus records of each test or inspection of the CP system in sufficient detail to show adequacy of corrosion control, were required to be retained as long as the facility is in service.

requirement for pressure testing of pipe that operates at 30% or more of SMYS was extended downward to pipe operating at 20% or more of SMYS. This change in scope included the pressure-test record-keeping requirements, which consisted only of the test fluid and test pressure according to para. 841.417. In Chapter V, recommended patrols and corrosion inspections were made mandatory, and recommended records of corrosion inspections and leak investigations became required.

A Chapter VI ‘Records’ was added consisting entirely of CPUC-added language. Clause 301.1 therein stated that “the responsibility for maintenance of necessary records to establish that compliance with these rules has been accomplished rests with the Utility. Such records shall be available for inspection at all times by the Commission…”. In other words, the Utility must maintain sufficient records to be able to prove on demand that the Utility is complying with all of the rules. This could include design calculations, material-procurement records, and a broad range of construction and installation inspection data, in addition to the operation and maintenance activities described above, and could well have required more record keeping than was the case before GO 112. Also, the specifications for materials and equipment, installation, testing, and fabrication were required to be maintained by the Utility.

A Chapter VII ‘Reports’ was also added that required reporting to CPUC 30 days in advance of any proposed new installation, major reconstruction, or change in MAOP. Specific information to be reported to the CPUC included the purpose or reason for the activity, specifications concerning pipe to be installed, the MAOP, and the test parameters to be used.

GO 112-A of 1964 and GO 112-B of 1967 added no new record-keeping requirements.

Record-keeping requirements post-1970

Complete federal safety standards for gas pipelines were introduced in 1970. Although some technical content was based on the 1968 edition of B31.8, the regulatory provisions went well beyond B31.8 in terms of inspections and recordkeeping. All provisions were required, not merely recommended (“shall”, not “should”). Moreover, many of these requirements exceeded those in effect in GO 112 at that time. These are briefly discussed below:

• Subpart E – Welding, para. 192.243(f), where non-destructive testing (i.e. radiography) of welds is performed: a record is required showing the number of girth welds made, the number tested, the number rejected, and their disposition by location (such as milepost), for the life of the pipeline. Also para. 192.225(c) requires a record of the details of each qualification of a welding procedure, to be retained for as long as the procedure is used.

• Subpart J – Test requirements, para. 192.517: a record is required of each test performed on pipelines operating at a hoop stress of 30% or more of SMYS

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering24

applied at all times prior to the introduction of the new requirements. A timeline of introduction of major record-keeping requirements is shown in Fig.2.

Prudence and policy

Unbroken chain of documentation not the rule

The practical significance of the ‘grandfather’ rule was that it was not necessary for an existing pipeline already in service to have been pressure tested to the minimum specified ratio of the MAOP. In fact, clause 192.619 offered four possible alternatives for establishing the MAOP:

• para. 192.619(a)(1) recognized the design pressure of the weakest component in accordance with Subparts C and D. In this case the MAOP would be based on manufacturer’s component pressure ratings or engineering calculations using specified material strength and wall thickness.

• para. 192.619(a)(3) recognized the highest pressure to which the pipeline had been subjected during the five years preceding 1 July, 1970.

• para. 192.619(a)(4) recognized 85% of the highest test pressure to which the pipe had been subjected, either in the pipe mill or in the field. If no field test was documented, the mill test would govern. The operator could determine the pipe-mill test pressure from the pipe product specification.

• para. 192.619(a)(5) allowed the operator to determine the maximum safe pressure considering the history of the segment, known corrosion, and actual operating pressure. This might be used, for example, with an uncoated pipeline that had experienced general wall thinning due to corrosion. (It is notable that this language existed prior to the use of in-line inspection for conducting integrity assessment, so an operator might not have had complete information about the extent of corrosion.)

Important and extensive new record-keeping requirements were put in place to support operator qualification (OQ) in 1999, integrity-management planning (IMP) for transmission pipelines in high-consequence areas (HCAs) in 2004, and distribution system IMP in 2009, as discussed below:

• Subpart N – Qualification of pipeline personnel, para. 192.807: requires the operator to maintain qualifications of personnel performing covered tasks. The qualification records must include identification of the individuals, the covered tasks each individual is qualified for, the dates of qualification, and the qualification method. The records must be maintained while the person is performing the covered task and for five years after.

• Subpart O – Gas transmission pipeline integrity management, para. 192.947: requires the operator to maintain records demonstrating compliance to Subpart O. The required items listed are (a) a written integrity-management plan; (b) documents supporting the threat identification and risk assessment; (c) a written baseline assessment plan (BAP); (d) documents supporting each decision, analysis or process of each element of the BAP and IMP; (e) personnel training programme and records; (f) prioritized assessment mitigation schedule; (g) documents supporting the direct-assessment (DA) plan; (h) documents supporting the confirmatory direct-assessment (CDA) plan; and (i) verification of notifications made to OPS or any state regulator as required by Subpart O.

• Subpart P – Distribution pipeline integrity management, para. 192.1011: requires the operator to maintain records that demonstrate compliance to the requirements of Subpart P, for at least 10 years. The records must include any superseded copies of the IMP.

Clearly, the regulations are punctuated by several major new additions to record-keeping requirements, and today’s record-keeping requirements cannot be presumed to have

Fig.2. Timeline of significant record-keeping requirements.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 25

Loss of useful records for any reason is not desirable, but past failure to preserve records does not necessarily imply operator imprudence or irresponsibility, neither does operating a pipeline while gaps exist in some records. Not all records are important to safely operating a pipeline day-to-day once the primary purpose of the record has been satisfied; prudence is exercised in making good choices with the information available. Consider that a pressure test of a pipeline following construction had been performed and that all stakeholders (owner, state or federal regulator, lender, insurer) were satisfied that the pipeline had been properly designed, constructed, and commissioned. The MAOP is entered into a ledger, a memo stating the MAOP is issued to control room operator, pressure control set points are confirmed, and operating procedures are updated. Consider next that the actual pressure-test records become lost some years later. How does the loss of that record affect any of the numerous activities a prudent operator is obliged to carry out day after day, such as: controlling pressure within established set points, marking the line for excavators, conducting damage-prevention and public education programmes, periodically testing valves, performing leakage surveys, repairing leaks, conducting line surveillance, maintaining cathodic protection, or training operations personnel, to name a few? The answer is that it does not. Once the MAOP has been correctly established, using any one of the allowed methods, those records have little bearing on day-to-day operation of the line.

The foregoing discussion is not meant to suggest that all records’ losses or data gaps are inconsequential. In fact, accurate and readily available data of some kinds are essential for safe and efficient operation. The authors support the industry’s efforts to respond to the San Bruno incident by evaluating the accuracy of its records. There is value in good records; however, the authors also believe that more has been made of the role of record-keeping as a cause of that incident than may have been warranted. The conclusion that the incorrect identification of the pup that originated the failure as “30-in seamless” in the pipe inventory database would have led any operator to enact a series of decisions culminating in the removal of the pipe decades after its installation stretches credibility.

The industry, including regulators and other stakeholders, should contemplate whether any amount of retrospective records’ analysis can offer complete protection against “unknown unknowns”, particularly where they originated many decades ago37. That argues in favour of the CPUC decision to require replacement or retesting where adequate test records are lacking. It also leans toward ratepayers carrying the financial burden of achieving the added assurance provided by pipe replacement or retesting unless it can be shown that an operator’s behaviour went well beyond the lack of historic documentation.

None of the above methods for establishing the MAOP necessarily required documentation of a prior post-installation pressure test. In fact, the method given in clause (a)(3) requires knowing no information about the specified grade or wall thickness of the pipe. That these alternative methods of establishing MAOP were allowed indicates that OPS accepted that records of testing or of pipe physical attributes were not always available. In particular, an operator was not prohibited from using clause (a)(3) even if a test had been performed and test records had been lost for some reason. These alternatives have been in Part 192 from 1970 to today, so OPS/PHMSA has since 1970 accepted that not all records need necessarily be present, or if present, need necessarily be complete or represent an unbroken chain of traceability.

Can an operator be ‘prudent’ while missing records?

It is not uncommon for pipeline operators to have incomplete or inaccurate data about attributes of portions of their pipeline systems, including specified pipe-material grades, specified nominal wall thicknesses, seam types, pipe manufacturers, coating types, flange or valve pressure classes, installation dates, construction specifications, welding procedures, pressure tests, corrosion control, and historic operating pressures.

The likelihood of records being incomplete increases with the age of the system, particularly with systems built prior to 1970 when the more-extensive records’ requirements of Part 192 came into effect. Nationwide, 37% of natural gas transmission pipelines now in service were installed before 1960, and 61% were installed before 1970; thus, a sizable proportion of existing pipelines was installed at a time when only minimal provisions for record keeping were found in standards and regulations. While the likelihood of gaps in the data increases with age, compromised data exist in systems built in many eras, including those built after 197035. Whether a lack of certain documents constitutes violation of regulations or indicates operator imprudence has become central to whether shareholders or rate payers pay for costly retesting or replacement of pipe.

There are many innocuous causes for loss of records including: an individual not recognizing the importance of a document or collection of documents, change of facility ownership, loss event (fire, flood), clerical mishandling, or misplacement in offsite storage, to list a few. Certainly back-up copies in one form or another can offset the loss of originals, but consider that photocopy technology was not widely available until the mid-1960s, perhaps after some original documents were already lost, and the back-up process is not without risk either36.

35. http://www.phmsa.dot.gov/pipeline/library/data-stats36. An anecdote reported to the authors was an occasion where a clerical worker, instructed to photocopy hydrostatic test records, first separated the pressure charts from the test report forms which had been stapled together into separate piles, irreversibly breaking the link between pressure records and test segments. In another case, documents sent by an operator to a third-party long-term storage facility were misplaced by the storage contractor.

37. Certainly intensive QA auditing efforts soon after construction can potentially reveal incorrect actions or wrongdoing.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering26

P-10-2 issued to Pacific Gas and Electric (PG&E)38. The NTSB recommendation that records “should” (not “shall” or “must”) be traceable, verifiable, and complete applied to “all as-built drawings, alignment sheets, and specifications, and all design, construction, inspection, testing, maintenance, and other related records…relating to pipeline system components, such as pipe segments, valves, fittings, and weld seams for Pacific Gas and Electric Company natural gas transmission lines in class 3 and class 4 locations and class 1 and class 2 high consequence areas that have not had a maximum allowable operating pressure established through prior hydrostatic testing”. The NTSB did not extend the recommendation to all pipeline facilities in all locations, nor to any facilities anywhere that have in fact been pressure tested.

In the gas pipeline context, the terms originated with NTSB as stated above, though NTSB may have assimilated the terms from applications outside the pipeline industry. The NTSB is not a regulatory agency, but rather an independent agency of the United States’ government that has no responsibility for writing regulations and no powers of regulatory enforcement. Based on accident investigations that it is authorized to perform, the Board offers opinions and recommendations that may or may not be observed39. The NTSB recommendations almost certainly were not made in consultation with PHMSA40.

Shortly after issuing SR P-10-2, PHMSA issued ADB-2011-01 advising operators that records they rely on for establishing the MAOP “must be reliable” and that the records “shall be traceable, verifiable, and complete”41. It took PHMSA another 16 months to develop guidance to the industry as to how to interpret the terminology in order to satisfy the new requirements42. In the meantime, the industry attempted to articulate what was necessary to meet these requirements by issuing white papers43 and developing individual company processes in the hope of meeting the regulator’s unspoken criteria. No guidance was found in the GPTC Guide44, a widely used reference guide to the interpretation of and compliance to Part 192. The fact that no guidance could be found in any common external publication is consistent with the position that “traceable, verifiable, and complete” represented new criteria.

Gaps in data that validate the MAOP severely limit an operator’s options for addressing a change in location class, pressure uprate, or request for regulatory waiver or special permit, which is as it should be. Data quality also has implications for integrity management. Certain elements of an IMP, notably the integrity-threat identification and risk-assessment tasks, are facilitated by having reasonably complete and accurate historical and technical data. ASME B31.8S recognizes that data important or useful to these tasks may be missing:

• para. 4.2.1 Data requirements: prescriptive integrity management programs, states that if listed data elements relevant to an integrity threat are not available, the integrity threat must be assumed to apply;

• para. 4.4 Data collection, review, and analysis, states that unavailability of data cannot be used to justify excluding an integrity threat;

• para. 5.9 Data collection for risk assessment, advises that if significant data are not available, the risk model may need to be modified based on an analysis of the impact of the data being unavailable;

• in Appendix A, the paragraph ‘Gathering, reviewing, and integrating data’ states that where the operator is missing data, conservative assumptions shall be used with the risk assessment, or the segment shall be prioritized higher for each integrity threat listed.

Part 192, Subpart O, para. 192.917, requires the operator to perform integrity-threat identification and risk assessment in accordance with B31.8S, Sections 4 and 5, respectively, which incorporate the above provisions concerning how to compensate for unavailable data. By referencing these sections, the regulations clearly contemplate that data important to an IMP may be unavailable.

“Traceable, verifiable, and complete” represents new requirements on record keeping

It has been suggested in the course of public debate that the criteria for document reliability, being “traceable, verifiable, and complete” do not represent new standards for the quality of natural gas pipeline records. While the attributes of “traceable, verifiable, and complete” are certainly desirable, and reasonably expected in modern times, they are not standardized thresholds for data quality for pipelines of all eras, and have no basis in regulation. They represent new documentary criteria.

The words “traceable, verifiable, and complete” appear nowhere in any issues of B31.8, GO 112, or 49 CFR Part 192 prior to SR P-10-2. The words, as applied to documents related to the design, construction, or operation of a natural gas pipeline, did not originate with the federal pipeline regulatory agency, the CPUC, or any other state pipeline regulatory agency. The terminology “traceable, verifiable, and complete”, as used in connection with gas pipelines, originated with the NTSB’s Safety Recommendation (SR)

38. NTSB, Safety Recommendation (SR) P-10-2, January 3, 2011.39. The NTSB’s reports are not admissible in court, 49 USC 1154(b): “No part of a report of the Board, related to an accident or an investigation of an accident, may be admitted into evidence or used in a civil action for damages resulting from a matter mentioned in the report”; although its investigators’ factual reports are, 49 USC 835.40. 49 USC 1131: “The Board shall provide for appropriate participation by other departments, agencies, or instrumentalities in the investigation. However, those departments, agencies, or instrumentalities may not participate in the decision of the Board about the probable cause of the accident.”41. PHMSA Advisory Bulletin. Establishing maximum allowable operating pressure or maximum operating pressure using record evidence, and integrity management risk identification, assessment, prevention, and mitigation. ADB-2011-01.42. PHMSA Advisory Bulletin. Pipeline safety: verification of records. ADB-2012-06.43. American Gas Association, 2011. Industry guidance on records review for re-affirming transmission pipeline MAOPs. October.44. GPTC Guide for gas transmission and distribution piping systems. ANSI Z380.1, various years.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 27

What does this mean for policy?

The costs for California natural gas pipeline operators to perform hydrostatic testing and/or replacement of hundreds of miles of grandfathered pipeline, some located in congested areas, in order to comply with CPUC’s decision have been projected to be $1.2 billion for Sempra and about $1.3 billion for PG&E45. The question to be settled, in public debate or hearings, is how to share the enormous cost between shareholders and ratepayers in an equitable manner.

The authors do not dispute the need to modernize the nation’s pipeline infrastructure, verify the integrity of grandfathered pipelines as well as any other category of pipeline, and enhance data as well as is practical in order to manage risk as well as is practical. We believe that pipeline operators should be accountable for failures to meet regulatory obligations. However, we further believe that regulators should not give-in to the temptation (perhaps driven by public pressure) to penalize operators retrospectively for practices that previously were not considered deficient as judged by standards of the era.

Past practices must be gauged against an accurate and reasonable interpretation of historic standards, regulations, and accepted practices. Moreover, proceedings to establish financial penalties for past failure to comply with regulations should be separated from ratemaking proceedings designed to regulate future behaviour46. When weighing financial penalties, consideration should also be given to the intensity of the regulator’s past focus on regulatory provisions deemed in retrospect to have been contravened. (In other words, if the regulatory violation was a failure to comply with

From ADB-2012-06, “traceable” records are tied to original documents. It should not be surprising that some documents that predated 1961 or even 1970 might not have been retained if there was no regulatory requirement to retain them. It is also possible, with the passage of time, for original documents to become lost through any of the possible loss mechanisms described above. In these or similar circumstances, it becomes impossible to meet the “traceable” test. While not optimal, losses of traceability are not uncommon.

“Verifiable” records are those in which data are confirmed by other separate documentation. ADB-2012-06 appears to require that any record used to establish the MAOP must be confirmed by another record. Nowhere in the historical or current regulatory language reviewed above does agreement between multiple data sources appear as a requirement.

“Complete” records are finalized by a signature or date. ADB-2012-06 gives, as an example: “a complete pressure testing record should identify a specific segment of pipe, who conducted the test, the duration of the test, the test medium, temperatures, accurate pressure readings, and elevation information as applicable.” This example lists two items that are not specified in para. 192.517, namely the specific segment of pipe, and temperatures. Thus meeting PHMSA’s requirements for record keeping since 1970 actually does not meet the test for completeness in ADB-2012-06, so clearly a new requirement has been imposed.

The language of SR P-10-2 is clearly made in reference to grandfathered pipelines that are now in Class 3 or 4 areas. As explained in Part D above, gaps in documentation could well occur in connection with many such pipelines. Therefore, the notion that the criteria of SR P-10-2 represent thresholds of data reliability that have always existed in regulations is inconsistent with established fact.

Fig.3. Pressure-test compliance matrix.

45. www.dra.ca.gov46. CPUC Rulemaking 11-02-019, Exhibit 21, Pacific Gas and Electric, pipeline safety enhancement plan (implementation plan). Prepared testimony of Susan F. Tierney, “Principles to Align Safety and Regulatory Ratemaking Policy”, 28 February, 2012.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering28

that hydrostatic pressure testing after construction was not required by applicable industry standards (ASME B31.8) until 1955. The technology for pressure testing cross-country gas pipelines was not developed until 1950. Prior to 1955, operating pressure was established by the pipe-mill test or an engineering calculation. Post-construction tests were discretionary, and generally for detecting leaks at flanged or welded joints above ground. Industry standards were referenced by some state regulations (for example, GO 112 in 1961) until the issuing of federal safety regulations in 1970.

Record-keeping requirements began with welding quality control. In 1955, hydrostatic-test records and those necessary for executing the operator’s O&M plan were required. Of test records, only test fluid and test pressure were required, and then that requirement only pertained to pipelines operating at a hoop stress of 30% or more of SMYS. It was thought at that time that engineering specifications and commercial documents were adequate to demonstrate practices. Additional record-keeping requirements were imposed by federal safety regulations in 1970, including for hydrostatic-pressure testing.

When federal regulations were issued in 1970, several options were available for establishing the MAOP of existing (grandfathered) pipelines that did not rely on a documented post-construction pressure test. This establishes that for the past 50 years, the regulators have accepted that documents supporting the MAOP could be incomplete. In that context, the tests set forth in ADB-2012-06 represent new standards that should not be used to judge whether an operator has complied with prior standards for that purpose as part of the ratemaking process.

record-keeping requirements, did the regulator recognize an infraction in the past? If so, what was done about it? If not, has a reasonable ‘statute of limitations’ rendered it moot?)

Finally, customers should pay rates that fully reflect the cost of providing the goods or services47. Rates should therefore reflect the cost of complying with new regulatory requirements. This includes the cost of retesting or replacing pipe not thought to meet modern standards of integrity, where those activities are necessitated by new regulations.

Figure 3 presents a matrix of compliance in terms of pressure-test activity versus time. The time scale is divided between pre-1955, 1955 to 1961, 1961 to 1970, and post-1970. (A threshold year other than 1961 may apply in other states.) The pressure-test activities are grouped as ‘no test’, ‘probably tested but no records’, ‘partial test records available’, and ‘full test records available’. Situations where the indicated combination of testing history and documentary completeness are likely to comply with applicable requirements of the era, or not, are suggested in the matrix. Other situations fall in a grey area that should be evaluated making a reasonable accounting of circumstances. The burden of revalidating confidence in the existing pipeline system is likely to be tied, in part, to the perceived degree of compliance to requirements of the era. Figure 3 can serve as a guide to that process.

Summary and conclusions

This paper reviewed the history of industry standards and regulatory requirements in the areas of hydrostatic pressure testing and record keeping. The review shows

47. Tierney, idem.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 29

THIS ARTICLE DISCUSSES the technical background to being able to weld safely on gas pipelines, particularly older pipelines with a higher carbon equivalent.

Making connections in gas pipelines that may not be taken out of service (and, typically, operating at 130 bar and with a flow rate of 10 m/s) requires special techniques. Hot tapping is often applied; split tees to enable this technique are welded to pipelines that are – and remain –completely operational. The split tees are placed on the pipeline and then welded together using longitudinal welds to make one tee piece; after inspection, circumferential welds are made to create the connection between the tee piece and the pipeline.

This same welding technique is also used to repair damage to pipelines. This can be in the form of a buttering-up repair as part of a pipeline-rehabilitation programme, or installing repair sleeves. Installing repair sleeves can also be a follow-up to an in-line inspection that has established that there is damage to the pipeline, or to an activity that has been reported as doing damage as a result of external interference (such as excavation, trenchless techniques for installation of cables or other pipelines, and ploughing).

NV Nederlandse Gasunie has conducted research since 1978 by, among others, TNO (Netherlands Organization for Applied Scientific Research) in order to obtain a sound basis for a reliable and safe way to weld onto gas pipelines. This article incorporates over 30 years’ operational experience with welding on gas pipelines, and addresses regulations, critical aspects of welding on operational gas pipelines, and the control of these critical aspects. The metallurgical backgrounds are also discussed.

This paper was presented at the Fixing Pipeline Problems conference held in Berlin in October, 2012, and organized by Clarion Technical Conferences and Tiratsoo Technical.

*Corresponding author:tel: +31 (0) 50 700 9785email: wim.schipaanboord@dnvkema.com

1 DNV KEMA Energy & Sustainability, Groningen, Netherlands2 Nederlandse Gasunie Repair and Emergency Department, Deventer, Netherlands

Welding on in-service gas pipelines using low-yield electrodes

by Wim N Schipaanboord*1, Jan Marquering2, Bart G Koppens2, and Jan Spiekhout1

Good procedures and quality control

The reasons for welding on operational gas pipelines are primarily economic. Shutting-down pipelines is expensive, because most supply contracts are based on continuous supply. A strict requirement is that no concession whatsoever is made to safety. Good procedures and quality control are essential. The risks are made manageable through meticulous specifications and the application of accurate welding procedures. The consequences of careless working practices can be particularly serious, as can be seen in Figs 1-3.

Regulations

The European standard for welding on gas pipelines, NEN-EN 12732, stated in its first edition – published in 2000 – that welding for the purposes of hot tapping may only be carried out after “sufficient research and development in order to ensure that the safety and the implementation are assured and that the mechanical properties are correct.”

Before the start of welding work on pressurized pipelines and systems, the pipeline operator must analyse whether the design, the materials, and the condition of the pipeline concerned are suitable for welding at the operating pressure. The critical aspects that have to be referred to are also mentioned in the welding procedure specification. There is a list of critical aspects in Appendix D of NEN-EN

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering30

12732; the appendix is informational in character, and gives the most important practical requirements that are set for welding:

• The specified minimum yield stress of the filler metal may not be greater than 400 N/mm².

• The tensile strength must be between 400 and 560 N/mm².

• The strain at failure (A5) must be at least 26%.• The average Charpy V-notched bar impact value of

the filler metal must be at least 47 J; a single value of at least 32 J is acceptable; the test temperature must correspond to the standard that the pipeline owner applies.

• The average of a set of three measurements of the diffusible hydrogen content (HDM) may not be greater than 3 ml, and a single value may not exceed 3.5 ml.

Fig.1. DN1050 circumferential weld of a split tee started to leak after earth movement (see Figs 2 and 3).

Fig.2. Circumferential weld of a split tee leaked after a crack formed - photograph taken after pressure was relieved.

Fig.3. Detail of Fig.2. After a part of the fillet weld had been removed, signs of cracking could be seen at the place of the leakage. The high toughness (200 J at -20ºC) and strain capacity of the filler metal limited the extent of the damage.

Figure 4 shows an example of a hot tap for installing a Stopple tee with by-pass pipe; Figs 5 and 6 give examples of an in situ buttering-up repair.

Appendix D of EN 12732 also summarizes the areas of attention and precautions for preventing cold cracking and burnthrough:

• Normalized steels (from before 1970) are more susceptible to cold cracking. Preheating and reducing the gas flow rate are measures to prevent it.

• A minimum extension length must be maintained at a certain minimum wall thickness so that the advance speed during welding guarantees that there is no local burnthrough.

• Reducing the pressure in the pipeline must be considered below a certain minimum pipe wall thickness.

• A minimum gas flow rate must be maintained below a certain minimum pipe wall thickness.

• A minimum distance must be maintained between the hot-tap fitting to be welded and other welds.

The European standard is functional in nature and gives no details. Other European standardization institutes have also accepted EN 12732, for example in DVGW Arbeitsblatt GW 350.

Pipeline materials

The oldest pipeline materials (dating from the 1960s) are of normalized quality and can be welded satisfactorily when preheated to 150ºC. Since the beginning of the era of large-scale construction of pipelines (1964-1970), the carbon equivalent (Ceq-IIW) has been limited to 0.47 (for Dutch pipelines) and the carbon content is limited to 0.23% by specification. However, in certain cases pipeline materials are still found with values that exceed these limits to a modest degree. Presumably these materials were supplied at the time in accordance with the-then applicable API 5L without additional specification, and

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 31

Fig.4. Implementation of a hot tap on a DN 600 pipeline from 1964 for a by-pass line and a DN 600 Stopple tee welded on the DN 600 pipeline in order to be able to interrupt the gas supply. The welding was done under the prevailing operating conditions (pressure and gas flow). Circumferential and longitudinal seams were inspected magnetically. Both longitudinal seams were also subjected to time of flight diffraction (ToFD) inspection.

Fig.5. In situ weld on/buttering-up repair (3x) carried out on a DN 450 pipeline and two sets of repair sleeves (partially depicted).

Fig.6. Detail of in situ weld-on/buttering-up repair with low-yield electrode on L 415 MB pipeline material.

Nr. Type dimension (in)

Pressure at failure (bar)

Pressure at failure/ SMYS

Pressure at failure/design pressure

Location of failure

1 12 x 4 x 12 250 2.3 3.7

2 18 x 4 x 18 225 >1.2 >3.4

7 12 x 4 x 12 285 >2.4 >4.3

10 24 x 24 x 24 170 2.6 2.5

11 24 x 24 x 24 160 2.5 2.4

12 24 x 16 x 24 182 >1.4 >2.7

19 24 x 16 x 24 182 1.5 2.8

24 8 x 8 x 8 270 2.2 4.1

Figure 7 - Failure location of 8 pipeline sections with split tees cut out after operational use at full operating pressure (Inspection in 1978).

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering32

Fig.8. This pipeline burst when it was pressurized after a split tee had been welded on it. Bursting pressure about 200 bar; 1978 (MAWP = 66.2 bar).

Fig.9. Metallographic examination of longitudinal (left) and circumferential fillet weld (right). The arrows on the fillet weld show the signs of cracking. At that time the welding was done with a standard basic electrode.

Fig.10. Hydrogen-induced cracking between gas pipeline and circumferential fillet weld.

Fig.11. Hardness as a function of cooling rate of TMCP steel with CEQ = 0.37 and 0.42 and normalized steel with Ceq = 0.49.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 33

Fig.12. A cooling time vs temperature diagram for normalized steel with a carbon equivalent of 0.49. The line in front of the perlite nose suggests the cooling while welding on in-service gas pipelines.

Fig.13. As for Fig.12, but now for TMCP steel with a carbon equivalent of 0.37.

in such cases a carbon equivalent of 0.50 can occur. Initially the specified minimum yield stress of these steels was limited to 345 N/mm²; from 1970 onwards, the steel quality St 60.7 was specified, which later changed to StE415.7TM with a specified minimum yield stress of 415 N/mm².

EN 10208-2 also specifies thermomechanically-treated steels (TM), and quenched-and-tempered steels. They have a significantly reduced susceptibility to cold cracking

compared with the normalized quality, but an evaluation of the welding procedure is required because of the much greater strengths (with a specified minimum yield stress of 550 N/mm²). In 1984 the transformation behaviour of both steels was investigated: acting on instructions from Gasunie, TNO used a welding simulator to study both normalized steel (Ceq 0.49, wall thickness 13 mm) and thermomechanically-treated steel (Ceq 0.37, wall thickness 7 mm).

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering34

opportunity to inspect, it is not prudent to repair a weld in an operational pipeline by means of grinding and re-welding, which is very risky from a safety point-of-view. It is much more important to develop a welding process with a very high probability of preventing the occurrence of cold cracking. The following measures have therefore been taken:

• Make longitudinal welds between the sleeve/split tee sections on a mild steel base strip 3 mm thick and conduct weld inspection using ultrasonic testing (beam angles 45˚, 60˚, and 70˚).

• Make circumferential fillet welds with a basic electrode and start an investigation into reducing the hardnesses in the heat-affected zone.

Hardness testing using a welding simulator

The steels that were commercially available in 1978 were TMCP steel (StE 415.7TM) and normalized steel (StE 415.7N). Many pipeline systems were constructed in the

Historical investigation in 1978

In 1978 a testing programme was carried out in which 20 split tees and welded fittings were cut out of pipelines and blanked off to carry out burst tests. It emerged from the burst tests that the pipeline failed first, and before the split tees or welded fittings that had been installed during operation; Fig.7 gives a detailed overview of the failure locations, and Fig.8 gives an example of the result of a burst test. As a rule, a safety factor of about 2.4 was specified, defined as the relationship between the burst pressure and the design pressure. Metallurgical investigation revealed that the weld quality could have been better. Figure 9 shows cross-sections of the longitudinal and circumferential fillet welds. In addition to the fact that the weld quality was unsatisfactory, high hardnesses – with an order of magnitude of 450 HV10 – were observed. An undesirable consequence of high hardnesses can be a hydrogen crack in the buttering region, which is a location that is impossible, or nearly impossible, to inspect; Fig.10 gives an example. Apart from the limited

Fig.14. Pipeline steel manufactured before 1972 – normalized quality – can have carbon equivalents up to 0.52. Spectral analysis has to show whether this steel can be welded without special measures.

Fig.16. Layer structure of buttering layers and fill layers in the determination of the heat flow during experiments in the Bernoulli Laboratory in 1983.

Fig.15. Reference sample of pipeline material dating from 1964 for verification of the spectral analyses carried out on pipelines before 1975.

Fig.17. Placement of thermocouples on a gas pipeline fillet weld for the purposes of determining the heat flow in experiments in the Bernoulli Laboratory, 1983.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 35

level that is still acceptable for EN 15614-1 (this standard deals with describing and approving welding methods and welding method tests). Experimental data do indeed show hardnesses between 450 and 500 HV10 when welding with a cooling time Δt 8/5 of 2-4 secs.

It has emerged from these investigations that it is important to know the chemical composition of the pipeline steel types of normalized quality that were used before 1975.

Spectral analysis

The chemical composition of pipeline materials used before 1975 and with an elastic limit above 345 N/mm² must be verified using a mobile spectrometer, and material samples taken from sections of old pipelines are used in this verification. Obviously, reliable reference material is needed when calibrating the spectral measuring equipment. As part of the ISO 3834 quality management system, Gasunie’s Repair and Emergency Department in Deventer has such reference material: this was taken from a pipeline constructed in 1968, which has a Ceq (IIW) of 0.49. When the old pipeline material is being verified, this reference material is also analysed before and after the measurements of the pipeline being welded on. The settings of the mobile spectrometer can be adjusted if the deviation is too large. A company accredited in accordance with IEC 17025 conducts these spectral analyses.

Netherlands before 1972 using the latter type with normalized final treatment. In the welding simulator, samples were exposed to a range of simulated conditions in order to be able to plot the dilatometer curves. Hardness measurements were also made on the test material. Cooling times between 4 and 100 secs were simulated for the peak temperatures of 1000°C and 1300°C. Figure 11 gives an example of the hardness of three different chemical compositions as a function of cooling time.

At a cooling time (Δt 8/5, i.e. a cooling time between 800˚C and 500˚C) of 8 secs or less, 100% martensite was formed in the normalized steel. Because of the high concentration of carbon (about 0.22%), this type of martensite is susceptible to hydrogen embrittlement and residual welding stresses.

In the TM steel, a bainite transition structure occurred at Δt 8/5 of 2.5 secs or longer. This structure is significantly more favourable than the brittle, cracking-susceptible, martensite in the normalized steel, as illustrated in Figs 12 and 13.

At a peak temperature of 1300˚C, a hardness of 450 HV10 (Vickers hardness) was measured with the normalized steel (Ceq 0.49). The hardness of the thermomechanically-treated steel with a Ceq of 0.37 at a Δt 8/5 of 3 secs was approximately 380 HV10.

If a peak temperature of 1000˚C is selected, the hardness of the normalized steel drops to 380 HV10, which is the

Fig.18. Temperature profiles at the different locations on the gas pipeline and fillet weld in order to determine the heat flow in experiments in the Bernoulli Laboratory, 1983.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering36

Item Chemical analysis – content of different elements in % Ceq

C Si Mn P S Cr Mo Ni Al Cu Nb V

001 0.23 0.29 1.48 0.01 0.02 0.05 0.02 0.05 0.05 0.09 - - 0.5

002 0.04 0.33 0.62 0.01 0.01 0.06 0.01 0.03 0.01 0.03 0.02 0.01 0.16

003 0.11 0.38 1.51 0.02 0.01 0.09 0.02 0.04 0.03 0.04 0.05 0.01 0.39

004 0.15 0.46 1.43 0.01 0.02 0.04 0.01 0.36 0.02 0.17 0.01 0.14 0.46

005 0.16 0.45 1.29 0.02 0.01 0.03 0.03 0.35 0.02 0.11 0.01 0.11 0.43

001: Linepipe steel muster ‘345B’; API 5L X56, 42-in x 14.2-mm

002: Buttering-up low-yield electrode (quasi AWS type E5016), Re < 400 N/mm²003: Linepipe material L485MB; also used for testing as sleeve material004: 18-in sleeve component STE-420N; used for repair sleeves and split tees005: 24-in sleeve component STE-420N; used for repair sleeves and split tees

Table 1. Overview of pipeline material, split tee material, and deposition analysis. Typical values of the carbon equivalents for sleeve and split tee materials are between 0.44 and 0.48.

Fig.19. Measurements in the Bernoulli Laboratory made in 1983 in order to establish the relationship between gas flow rate and cooling rate Δ8/5:

Fig.20. Relationship between heat input and cooling time Δ8/5 at a gas flow rate of 25 m/s.

▼= DN 450 x 8, 60 bar ■ = DN 450 x 8, 64 bar ● = DN 200 x 7.9, 64 bar▲= DN 450 x 8, 60 barSam

ple co

py

not fo

r dist

ributi

on

1st Quarter, 2013 37

usually occurs in the base metal (sleeve sections or split tees). The implant test welding is carried out with parameters that are representative for welding under practical conditions. Longitudinal test strips of the normalized (N) steel and the thermomechanically-treated (TM) steel were taken from the pipe wall; test strips of the normalized steel were also taken radially through the pipe wall. During the execution of the implant tests, the loading level was reduced in steps of 25 N/mm² until no further fracture occurred within a 24-hour period. In this way the critical failure stress was determined as a function of the hydrogen content in the deposition. A higher critical failure stress means limited susceptibility to cold cracking.

These experiments were carried out in 1983 and the critical failure stress was about 525 N/mm² for the ‘modern’ TM steel at 3 ml HDM (hydrogen in melted metal). In the case of normalized steel with a hydrogen content up to 1-1.5 ml HDM, the critical failure stress of the deposition was around 400 to 425 N/mm². During the preparation of EN 12732 Appendix D, Gasunie proposed this value as a maximum where it is still safe to carry out welding on a pipeline material without restrictions on pressure and gas-flow rate. The critical failure stress of normalized steel samples taken radially through the wall was only 380 N/mm². Figure 18 gives an overview of the graphical relationship between the critical failure stress and hydrogen in the filler metal that was established for the three steel qualities investigated.

Fitting materials for tee pieces

The materials used for tapped tees are normalized quality and the minimum specified yield stress is limited to 460 N/mm² and the carbon equivalent (Ceq) is not greater than 0.48; see Table 1 for a comprehensive overview of the materials used. The fitting materials are heated to a maximum of 150˚C before welding.

There are no special metallurgical aspects associated with this with regard to the weldability. It should be pointed out that there is an increased chance of cracking occurring if the diffusible hydrogen content is above about 5 ml/100 g of filler metal particularly at preheat temperatures below 40˚C and when the minimum specified yield stress of the weld filler material is greater than 440 N/mm². An intake inspection of the weld filler material is carried out on the basis of modified CTS tests (see below). Future work will have to show to what extent quenched and tempered material (P 420QH/ P460 QH) show improved weldability.

Weld filler material

CTS tests

The susceptibility of the connection to cold cracking in combination with the casing material of the split tee to be connected can be tested using a modified controlled-

Relationship between gas flow rate and cooling rate

The cooling time between 800˚C and 500˚C (Δt 8/5) is an important parameter. There is forced cooling if gas and water are flowing.

Research was done (Belgraver, 1983; De Haan, 1991, Bernoulli Laboratory – part of former Gasunie Engineering and Technology) into the relationship between gas flow rate and the Δt 8/5 cooling rate. It was established that a Δt 8/5 of 8 secs can occur at a gas flow rate of 3 m/s or higher. Measurements were made at different pipe diameters and under controlled welding conditions. This investigation revealed that there is a greater chance of martensite formation at higher gas flow rates.

Figures 15a-d give an overview of how the relationship between gas flow rate and cooling time Δt 8/5 came about. It can be concluded from Fig.15d that, in the case of normalized steel, cooling takes place before the perlite nose at gas flow rates of 3 m/s or higher.

Heat input versus cooling time

A study was also made of whether raised heat input has a significant effect on the cooling time Δt 8/5. This relationship was investigated for the first and second buttering layers on the gas pipeline using different electrode diameters in the PF(3G) position. In the 1984 investigation, welding was done with a gas flow rate of 25 m/s and a gas temperature of 10.2˚C. Doubling the heat input resulted in a doubling of the cooling time Δt 8/5: see Fig.16. The cooling time rose to barely above 8 secs at the highest heat input of 16 kJ/cm. This means that there were also conditions for forming martensite in the second buttering layer as well.

The question of whether hydrogen, dissolved in the metal matrix, has enough time to diffuse was also investigated. The measured cooling times from 300˚C to 100˚C were found to remain below 10 secs at a heat input of about 8 kJ/cm, and below 20 secs at a heat input of 16 kJ/cm, see Fig. 17. These times are longer than under normal welding conditions. Hydrogen embrittlement can occur if no satisfactory measures are taken.

Implant test

The implant test is used to investigate susceptibility to cold cracking in order to be able to establish the critical failure stress as a function of hydrogen content. Unlike the CTS test and the tensile Y-U groove test described in the literature, with the implant test one can establish a quantitative relationship between the critical failure stress and the quantity of diffusible hydrogen in the filler metal and the heat-affected zone (HAZ).

In this investigation the implant strips had a screw-shaped groove so that they were always in the HAZ. Any cracking

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering38

Fig.21. Relationship between heat input and cooling time Δ3/1 at a gas flow rate of 25 m/s.

Fig.22. Relationship between free diffusible hydrogen in filler metal of the buttering layers and the critical failure stress. The electrode used was a low-yield AWS type E 5016.

Fig.23. Overview of a series of CTS tests on sleeve material with a Ceq of 0.48 and HDM values from 4.5 to 6 ml. The cladding layer of the welding electrodes was deliberately ‘loaded up’ with moisture in a climate-controlled cabinet. The increased moisture content resulted in a greater quantity of diffusible hydrogen in the heat-affected zone. The goal of this experiment was to test the crack susceptibility of the connection.

EN ISO 17642-2 gives a basic procedure for carrying out the CTS tests. A layer was buttered in these modified CTS tests in order to match conditions in practice. In this illustration the welded connection with the sleeve component consists of one bead.

Cold-cracking tests are carried out in order to establish the susceptibility of a combination of pipeline and split tee material and type of welding electrode including the free diffusible hydrogen and the welding parameters.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 39

Decreasing the diffusible hydrogen content brought-in during the welding also plays an important part. Figure 23 is an example of hydrogen escaping immediately after welding on structural steel with a basic electrode, made visible in glycerine.

The tests were intended primarily to investigate and verify the cracking resistance in the transition between the buttering layers and the sleeve components of the split tee. One needs to bear in mind that the relatively high carbon equivalent in the ferritic-perlitic steel can always be susceptible to cracking and impedes rapid hydrogen diffusion.

Bead tensile test

This bead-on-plate test was further developed by Gasunie primarily in order to investigate the crack susceptibility between the buttering layers and the gas pipeline. Old sections of pipeline (item 001 in Table 1) were kept in reserve in order to be able to carry out comparative testing between laboratory conditions and the circumstances in practice (an internal pressure of 66.2 bar and a gas flow rate of 20 m/s). These tests that compare with practice are the ultimate method for simulating reality as closely as possible. The linepipe material with an IIW carbon equivalent of 0.50 (item 001 in Table 1) was carefully worked and kept under a constant load at a level of 85% of the yielding stress. Bead-on-

thermal-severity (CTS) test (see Figs 19 and 20) undertaken on a water-filled pipe and at a constant flow. In the past, a selection of suitable filler materials was made together with Lincoln (previously Smitweld) on the basis of modified CTS tests. There are three critical aspects with regard to the welding of the relatively stiff connection between the sleeve and an in-service gas pipeline in combination with the high cooling rate due to the flowing gas:

• Formation of a hard brittle structure (martensite).• Hydrogen embrittlement as a result of diffusible

hydrogen from crystal-bound water in the electrode cladding layer that ends up in the welding bath and the heat-affected zone.

• High residual stresses in the filler metal.

The diagrammatic layer structure of the modified CTS test is shown in Fig.21, and Fig.22 is a photograph of the CTS test with one bead. Table 1 contains typical compositions of pipeline and split-tee materials, and an analysis of a low-yield electrode deposition (one buttering layer).

The carbon equivalent (IIW-Ceq) of 0.16 of the deposition of the bead from the low-yield electrode is substantially lower than that of the sleeve and the in-service gas pipelines. This relatively ‘soft layer’ is used to reduce the residual stresses.

Fig.24. A CTS test in which cracking was recorded in the sleeve component. The cracking is in the HAZ of the sleeve material. The filler metal is free of cracks. The sleeve component would have to be limited as regards carbon equivalent or the preheat temperature would have to be increased.

Fig.25. Diagrammatic layer structure of the modified CTS test.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering40

plate welds and buttering layers were prepared in the PF and PC welding positions, and the heat input was comparable to the welding procedure qualification (WPS) that was used for the welding in practice; Figs 23a-c and 24 are illustrations of such a test. It is possible to deduce from these tests whether the low-yield electrode has suitable properties to enable the crack-free welding of this steel (Ceq = 0.50). Figure 25 is an overview of these bead-cold cracking tests, in which a just crack-free bead-on-plate was welded with a temper bead under a representative operational load and cooling rate equal to those derived by a gas flow rate of 10 m/s as a function of the heat input for a pipeline material with a high carbon equivalent and diffusible quantity of hydrogen. It should be emphasized that this test is a procedure that was developed by Gasunie itself and is not related to any standards. This is the ultimate test for providing underpinning for the suitability of a weld-filler material as a means for making in situ weld repairs for pipeline rehabilitation and buttering layers for welding split tees and repair sleeves.

In summary, the critical aspects can be controlled using the following remedies:

• Limit the quantity of martensite to a minimum; if possible ensure the tempering of the martensite structure.

• Keep the moisture susceptibility of the electrode buttering-up layer as low as possible, and prevent moisture absorption during exposure of the electrode cladding. Endeavour to have a product where the free diffusible hydrogen (HDM) is lower than 3 ml, and preferably even lower.

• Select a prudent layer structure for the fillet weld and a good welding sequence in a circumferential direction. A more favourable reduction of stress can be achieved by having buttering layers that are twice as long as the thickness of the split tee or repair sleeve.

• Select a core wire made from pure iron with extremely low contaminant concentrations and a low carbon and manganese content.

Improved packaging for the electrodes

In 1980, experiments were carried out in collaboration with electrode manufacturer Smitweld in which a search was conducted for a core wire with extremely low carbon and manganese contents. This is known as the ARMCO core wire. The covering for the buttering-up electrodes is of course full basic type. An investigation conducted by TNO on behalf of Gasunie demonstrated that this combination has a very favourable effect on the quality of the electrode in regard with metallurgical properties. A disadvantage is that the welders need to be thoroughly trained and experienced in handling this electrode for welding in the uphill position: the hydrogen diffuses much faster in the ARMCO iron than in other steels and, as a result, it can migrate quickly to the heat-affected zone of the pipeline and split tee, or repair sleeve. Figure 26,

Fig.26. CTS test with one bead, pipeline material L415MB and sleeve material L485 MB, wall thickness 15.9 mm; item 003 of Table 1, water filled.

Fig.27. Escape of hydrogen gas from the filler metal and zone next to the weld immediately after welding with a basic electrode. This emission is a diffusion process and is temperature dependent.

Fig.28. API 5L X 56 material under a tensile load of 85% SMYS buttered in order to validate the suitable electrode for the repair welds for pipelines in operation.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 41

taken from a dissertation by Pekka Navasmaa [1], gives an overview of the diffusion speeds, from which it can be deduced that the diffusion speed in the ferritic-perlitic steel is about a factor of three lower than in the ARMCO material. The core wire for the low-yield electrode is made from this material. This could be an explanation for why cold cracking occurs more readily in the sleeve material of split tees than in the bainitic ‘modern’ linepipe materials such as L 415 or L 485 MB, see Fig.20. It would seem that modern thermomechanically-treated steel is certainly more tolerant to hydrogen emission after welding than normalized steel, which is currently still being used for repair sleeves and split tees. It is therefore important to be cautious when applying this type of material.

An investigation aimed at limiting the amount of moisture absorption after the packaging is opened was conducted by Gasunie in 1983. These experiments were satisfactory and Gasunie starting putting the electrodes in vacuum packaging itself, based on the idea of the first author of this paper. On the instructions of Gasunie, TNO measured moisture absorption and hydrogen diffusibility, and these measurements confirmed the positive effects of this packaging method. In the meantime, the vacuum packaging of basic electrodes has become generally accepted.

Welding the longitudinal seam

Standard welding procedures are applied: manual welding with clad basic electrodes, in combination with mechanized welding or otherwise. Welding for the longitudinal welds is always done on a steel base strip in order to prevent

Fig.29. During the welding work and for 72 hours thereafter the tension strip is kept under a constant stress of 85% SMYS of the pipeline material to be investigated.

Fig.30. Diagrammatic overview of the testing to be carried out for the CTS test in Fig.25.

Bead-on-plate: A = welds with low heat input; E and D = welds with high heat input.

Influence of temper bead: B and F = welds with variable distance from the melting line.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering42

The longitudinal weld can be made with manual welding and using a mechanized welding process (FCAW), see Fig.30.

Welding the circumferential seam

The circumferential seam is welded in the PF welding position with a low-yield electrode: see the diagrammatic structure in Fig.31. The buttering layer is intended to create a relatively soft layer of filler metal before the adhesion of the sleeve component takes place. A broader foot for the buttering layer is also chosen in order to achieve more favourable stress reduction. The narrowing at the end of the sleeve is structurally beneficial for reducing the stiffness. Figure 32 shows the layer structure of a circumferential weld connection: the qualification of the circumferential weld consists of macros,

contact with the in-service gas pipeline, and this is a requirement that is also referred to in EN 12732. After inspection using the ToFD (time of flight diffraction) ultrasonic technique, the two circumferential welds were made one after the other. The ToFD technique, together with pulse-echo ultrasonic weld inspection, has been the standard inspection method for Gasunie since 1993 for testing the integrity of welds in split tees and repair sleeves. The welding method qualification is carried out in accordance with EN 15614-1.

Figure 27 gives an impression of a split tee welded under full gas pressure (60 bar) and gas flow rate (about 10 m/s). Figures 28 and 29 show the calibration activities for weld inspection with ToFD for the split tee in Fig.27.

Fig.31. Hardnesses at high cooling rates (Δt 8/5 = 2 [s] in steel with Ceq = 0.49). At this cooling rate the conditions pass across in front of the perlite nose, which explains the high hardnesses. Compare with Figs 12 and 13.

Fig.32. Relationship between heat input and cooling time:

x = HI = 0.26 kJ/mm and complete 3D cooling in FCAW welding process; X = situation before FCAW welds on gas pipeline with HI (gross 4 kJ/cm).

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 43

Qualification for circumferential weld connection: the FCAW welding process

The qualification programme for mechanized welding on gas pipelines includes:

• Selection of the filler material• Bead-on-plate tests on a pipe filled with water to

determine hardnesses at and next to the temper bead.• CTS tests to check the cracking susceptibility with

regard to the split-tee material and repair sleeves material for the welding process and weld-filler material combination.

• Bead tensile test to check the cracking susceptibility with regard to the linepipe material for the welding process and weld-filler material combination.

hardness measurements, and bending tests. A magnetic crack inspection is carried out after the welding.

Strength of the fillet-weld connection

Weld calculations were made for the circumferential seam on the basis of internal pressure. A traditional throat cross-section welding calculation was applied because it concerns a weld with good toughness properties. As regards the longitudinal direction, the weld must be able to transmit stresses from the pipeline. Wide-plate tests at Ghent University’s Soete Laboratory showed that the pipeline material starts to yield before the weld and that the weld withstands the load well. In the past, a large-scale bending test was conducted at Exova (at that time Matcon). It showed that the weld did not fail under high loading.

Fig.33. The split tee after welding of the longitudinal seams with a clad electrode. In order to achieve even shrinkage, the welding was done using the ‘pilgrim step’ technique. The full-penetration weld is made on a steel base strip in order to prevent penetration into the gas pipeline.

Fig.34. Calibration of ToFD equipment on calibration plate. Photograph with permission of Applus-RTD, Rotterdam.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering44

Heat input net (kJ/mm)

Interpass temp. (a) ˚C

Cooling time (secs)

T 8/5 T 5/3

0.25 23 2.4 4.7

0.27 26.6 3.5 6.3

0.26 29.5 0.9 1.8

0.3 27 1.5 3.3

Table 2. Measurement and recording of the welding parameter Δ8/5 and interpass temperature between the welding beads are essential for comparison with welding under practical conditions (internal pressure 66.2 bar and gas flow rate 20 m/s).

Fig.35. Calibration scan of the plate demonstrating the five notches according to EN 14751.Material taken from the same plate as used to make the split tees. Photograph with permission of SGS Nederland BV, Spijkenisse.

• Dummy test by welding a set of repair sleeves to a water pipeline and qualification by means of notch-toughness testing, strength of the weld connection, and bending tests.

• Burnthrough tests as part of the safety assessment of the overall welding system.

• Final qualification on a gas pipeline DN 450 x 7 mm L415 MB with repair sleeves of L 485 MB, 15.9 mm thick.

Figures 33a-d give an impression of this last phase of the qualification process.

A qualification programme was conducted during 2010-2012 for the mechanization of the welding process for the circumferential seam. This was recently qualified by TüV-Nord and the Competenz Center of TüV-Hamburg. Figure 34 gives an illustration of mechanized welding.

Fig.36. Mechanized welding of the longitudinal seam of the split tee for horizontal tapping, wall thickness 60 mm; material P 460 NL2. Tapering at the end produces a more favourable stress distribution between the fillet weld and the sleeve component.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 45

Applications

Welding on gas pipelines can be applied in the following situations:

Welded repair sleeves for restoring pipeline integrity as a result of corrosion

Figure 35 shows a case of serious corrosion, probably as a result of MIC (microbiologically induced corrosion). In this case the placement of the repair sleeves was possible (longitudinal welds and circumferential welds). Gasunie has standardized repair sleeves for this.

Buttering-up repairs for restoring pipeline integrity as a result of corrosion

Figures 36 and 37 show some parts of a validation investigation in which buttering-up corrosion damage to

Fig.37. Typical layer structure for the buttering layers and fillet weld for split tees and repair sleeves.

Fig.39. Execution of partially complete circumferential weld using manual welding.

Fig.38. Layer structure of the buttering layer, attachment weld and structure of the filler weld with clad electrode. The core wire is low carbon, approximately 0.02% C. A magnetic crack inspection was conducted after the welding.

Weld/steel/range of DH values (at RT)

Average DH range (cm2/sec)

Armco iron 70 (± 5) x 10-6

Fine-grained steel 25 (± 5) x 10-6

SMAW OK 74.78 5.5-15.5 x 10-6

SMAW OK 75.75 3.5-10.0 x 10-6

FCAW PZ 6138 7.9-17.5 x 10-6

SAW OK Autrod 13.27/OK 10.62

7.0-39.0 x 10-6

Table 3. Overview of diffusion speeds from Pekka Nevasmaa's dissertation [1].

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering46

a circumferential weld connection and adjoining pipeline material was validated. The welding work was done by Gasunie’s Repair and Emergency Department in Deventer and the tests were conducted by Ghent University’s Soete Laboratory in 1988 and 2002.

The API 5L X 56 material is representative, see Table 2, for the pipeline material on which the welding tests referred to above were conducted. The actual yield strength was 389 N/mm² and the notch toughness was 18 and 23 J at 0˚C. The circumferential weld had a yield strength of 470 N/mm² and an average toughness of 55 J at 0˚C.

The wide-plate test revealed the following parameters. At the moment that an indication occurred (the ‘critical’ event) there was overall strain of 4.83% and 4.45%. In the pipeline section, strains of 5.1% and 4.67% were recorded, and in the welded connection strains of 1.75% and 1.9% were recorded. The associated stresses were 538 and 559 N/mm². Figure 38 shows that the final structure fails outside the repair zone. The repairs did not affect in any way the integrity of

Size (in) 2011 2001-2011

Hot tap Stopple Hot tap Stopple

2 197 3 2241 4

3.5 8 0 21 0

4 15 4 239 52

6 15 4 130 53

8 21 15 84 55

10 0 4 9 4

12 16 9 70 36

14 0 1 7 4

16 3 3 23 13

18 2 4 27 11

20 0 0 4 6

24 0 3 11 6

30 0 0 7 2

36 1 0 10 6

42 0 0 4 3

48 0 0 8 1

278 47 2895 253

Table 4. Overview of the welding activities carried out by the Gasunie’s Repair and Emergency Department in Deventer.

Fig.40. During qualification activities the buttering-up and fillet weld are extended in order to enable notch toughness testing (Charpy V-notch) to be done.

Fig.41. Location of welding procedure qualification for repairing pipelines.

the weld connection of the old cellulose weld or the HAZ. This investigation therefore revealed that buttering-up with the low-yield electrode can be carried out such that the integrity of the welded connection and the immediate surroundings are not negatively affected.

This repair method has now been used a number of times:

• 1986: buttering up 8-in pipeline, 150 bar (natural gas supply)

• 1988: dressed weld on 36-in pipeline bend, 66.2 bar, (Fig.39); validation (1) Ghent University, Soete Laboratory

• 1992: buttering-up 12-in pipeline, 40 bar• 1994: buttering-up 18-in pipeline, 66.2 bar• 2002: buttering-up 36-in pipeline, 66.2 bar, validation

(2) Ghent University

Welded split tee for restoring pipeline integrity as a result of mechanical damage

If there has been mechanical damage, it is sometimes possible to avoid placing an infill piece by welding a split tee and cutting out the mechanical damage. The coupon with the dent/gouge is removed and this restores the integrity. This method has limited applicability: the position of the dent/gouge is one of the determining factors.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 47

• Restricting the hydrogen content of the filler metal is the dominant factor in avoiding cracking.

• Shutting-down and the unnecessary venting of gas are avoided, which is beneficial for the environment.

• Cutting through the pipeline is avoided. In the past this led to substantial costs because the cut pipeline broke off. Personnel safety was also an issue here.

• When a pipeline is being cut into, the gas has to be diverted via another route. This is an expensive activity and results in extra natural gas transportation losses.

• Mechanized welding, Fig.41, is a new step towards more efficient and ergonomic ways of operating.

• The application of advanced ultrasonic techniques like ToFD for weld inspection with a high probability of detection (>90%) provides an enhanced level of demonstrable quality of the work.

References

1. Pekka Nevasmaa. Predictive model for the prevention of weld metal hydrogen cracking in high strength multipass welds. Department of Mechanical Engineering, University of Oulu, p114.

2. Gasunie specification CSW-04-N (internal specification).3. W.N.Schipaanboord, B.G.Koppens, and J.Marquering,

2012. Lassen aan gasvoerende leidingen NIL-Lastechniek, March.

Acknowledgements

Repair welds and all related activities were carried out by Gasunie’s Repair and Emergency Department in Deventer. The authors want to thank Jan Woltjer (Manager), and Wytze Sloterdijk and Menno van Os (both executive senior consultants at DNV Kema Gas Consulting en Services, Groningen – formerly Gasunie Engineering and Technology) for their support during the preparation of this paper.

Pipeline repair by Gasunie’s Repair and Emergency Department in Deventer

Between 1964 and today Gasunie’s Repair and Emergency Department in Deventer has made about 2000 welded connections on in-service gas pipelines, and the technology described in this publication formed the basis of these (Figs 40 and 41). This technology has proved to be very trustworthy and makes it possible to implement reliable joints while giving due regard to safety and the environment with minimum gas transportation losses. A group of 14 welders who have been qualified by TüV Nord is available to do this work. This group is trained and qualified annually for welding under operational gas transmission conditions. The Repair and Emergency Department was recently certified under ISO 3834-2. There are now procedures relating to repairing the transmission network for the range between DN15 and DN 1200 by placing infill pieces and valves under conditions where gas may be present (hot tie-ins). Methods have been developed, including measuring underpressure before and during welding, in order to be able to guarantee optimal weld quality.

Welding procedures have been qualified for enabling the welding of repair sleeves, split tees, and weldolets in common welding positions, using both manual and mechanized welding process.

It has become clear that:

• Welding on in-service gas pipelines made from unalloyed steel under pressure while gas is flowing can be carried out safely.

• The use of a low-yield electrode with a basic cladding layer produces reliable connections provided that the quality requirements are met.

See pages 48–49 for Figs 42–53.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering48

Fig.42. DN 450 and DN 200 by-pass for qualifications under full pressure (66.2 bar and gas flow rate 10-20 m/s).

Fig.45. The welding of larger split tees (> DN 600) is done using a mechanized welding system and welds are inspected by means of ToFD and pulse-echo ultrasonics.

Fig.46. Local MIC corrosion, suitable for repair by using welded repair sleeves.

Fig.47. Corrosion could be repaired by buttering-up using low-yield electrode. In this picture, corrosion is simulated by grinding, depth approximately 4 mm.

Fig.43. Double FCAW welding head in position ready for buttering on a DN 450 L415 MB line pipe.

Fig.44. Power supplies and control units for the double FCAW welding heads.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 49

Fig.48. ToFD inspection prior to the repair in order to ensure that no hidden defects lead to instability of the welded connection during repairs under pressure and flow. Graphics with permission from Applus-RTD, Rotterdam.

Fig.49. Buttering-up procedure in a particular sequence in order to distribute the shrinkage evenly.

Fig.51. Failure of pipe-restored weld, and adjacent pipe material.

Fig.52. Pipeline rehabilitation 1988: welded on-line pipe segment from damaged part with 5-mm wall thickness loss as a result of erosion-corrosion. ILI inspection was conducted in 1987. Welding on in-service pipeline carried out in 1988 on DN750 x 17 mm bend of Q and T steel; Re = 524 N/mm² and Rm = 639 N/mm².

Fig.53. Gasunie’s Repair and Emergency Department implements hot taps from DN 15 to DN 1200. Over the last four decades it has made over 2000 welded connections under operational conditions.

Fig.50. Wide-plate testing applied for validation of repair welding by buttering-up (tests performed at Ghent University in 2006; Prof. Denys).

Sample

copy

not fo

r dist

ributi

on

MaY 15–16 2013Marriott Westchase hotel

houston, tX, usa

Exhibition and sponsorship options available, visit

www.clarion.org

Platinum SPonSor GolD SPonSor

UPSF_13_FP_Ad.indd 1 6/02/13 1:31 PM

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 51

THE USE OF THE new generation of pipe steels with high yield stress potentially increases the risk of brittle fracture. In order to evaluate this risk, safety factors associated with a surface crack and operating

pressure have been evaluated for three pipe steels: X52, X70, and X100.This evaluation has been made using a failure-assessment diagram and the SINTAP procedure. This analysis has also been extended to X120 pipe steel. The use of a domain failure assessment indicates that for this steel a risk of elastic-plastic fracture exists; however, for the X52, X70, and X100 pipe steels, failure potentially occurs by plastic collapse.

*Corresponding author:tel: +33 3 8764 0844email: pluvinage@cegetel.net

1 ENIM, Metz, France2 Fiabilité Mécanique, Conseils, Silly sur Nied, France

Evaluation of failure risk due to use of high-strength steels in pipelines

by Dr Julien Capelle1 and Professor Guy Pluvinage*2

AT PRESENT, the requirement for natural gas is rapidly increasing internationally, and pipelines are

used in natural gas transmission over long distances. The amelioration of gas transportation capacity is possible by increasing pipe diameters and operating pressures, gas cooling, decreasing the internal surface roughness, and increasing service reliability. Several studies have shown that the factors that have the most efficient effect on gas-transportation capacity are, in a decreasing order, pipe diameter, operating pressure, distance between compression stations, compression rate, and service temperature. By increasing the operating pressure and pipe diameter, the transportation capacity is increased, and this results in obvious economic advantages. Table1 summarizes the evolution of pipeline operating pressures and diameters over the last century.

Several pipelines have now been constructed with 1420 mm (56 in) diameter, and the use of this large diameter requires high-strength steels in order to avoid wall thicknesses that are difficult to weld as well as to minimize the steel weight. There are significant advantages in using higher-grade linepipes, such as X100 and even X120, when constructing long-distance pipelines, because such materials can improve transportation efficiency of the pipeline by increasing internal transportation pressure, and material cost can be saved correspondingly by reducing the wall thickness of the

pipe body and consumables for girth welding. However, there are still a number of transportation safety problems consequent on using high-strength pipeline steels.

First of all, where pipelines are laid through complex regions with high-risks, such as earthquake zones, gas pipelines in service may endure large displacements and stresses; the maximum flexural deformation can reach 4-5% when on pipelines through such regions.

Secondly, the increased pressure in modern pipelines also results in the danger of running ductile fractures as the results of the high stored energy content of the compressed gas.

For over 70 years, pipe yield stresses have been continuously increasing. However, an important part of the networks remains made from old pipe of Grade A and X52 steel [1]. The evolution of pipe yield stress with time can be seen in Fig.1. An important breakthrough in pipe steel development took place in the mid-1970’s with the introduction of the thermomechanical-rolling process substituting normalization by heat treatment. Higher strengths and better toughnesses have been obtained by further optimization of the thermomechanical-rolling process, and by lowering the C/Mn ratio. Later, in high-toughness X70 pipe steel, vanadium and niobium have added strength through precipitation hardening. Today, the development of X120 pipe steel has accelerated the strength improvement of linepipe; a low-C, high-Mn-Mo-Ti-B, steel has been developed, with a tensile strength as high as 931 MPa, and high toughnesses have been introduced. The submerged-arc welding (SAW) method, with one pass each for the internal and external welds, which had

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering52

can be seen that the yield stress of the studied steels is higher than the standard requirements, and elongation at fracture is strongly reduced when yield stresses increase.

Examination of the microstructure (Fig.3) shows that X52 and X70 steels have a ferrite-perlitic microstructure, while X100 has a bainitic one.

The fracture toughness KIc and δc have been determined using compact-tension specimens according to the French standards NF A 03-180 [2] (KIc) and NF A 03-182 [3] (δc). Specimen dimensions are extracted from three different pipe samples, as given in Table 4.

It can be seen that pre-cracking is in the pipe’s longitudinal direction. The critical load was determined using acoustic emission, which also determined the crack initiation (subscript i). The critical load obtained correlates well with the failure load

been employed for conventional grades of linepipe, has also been adopted for seam welding of the X120 pipe.

Due to the combined use of high-strength steel, high operating pressure, and large-diameter pipe, the risk of brittle failure has increased. By comparing the remaining safety factor due to the presence of crack-like defects, it is possible to describe evolution of this risk versus time through evolution of pipe design, and this has been done in the following by using the failure-assessment diagram (FAD) and, in particular, the SINTAP procedure.

Material

The three pipe steels that have been studied are X52, X70, and X100, and their chemical compositions are given in Table 2. The average values of the steels’ tensile properties are given in Table 3, and typical stress–strain curves in Fig.2. It

Fig.1. Evolution of the average pipe yield stress vs time.

Year Operating pressure (bar)

Diameter(mm)

Annual capacity(m3 x 106)

Power gas consumption

over 6000 km (%)

1910 2 400 80 49

1930 20 500 650 31

1965 66 900 830 14

1985 80 1420 26000 11

Table 1. The evolution of transportation conditions in gas pipelines.

Table 2. Chemical composition of the studied steels.

C Mn Si Cr Ni Mo S Cu Ti Nb Al

X52 0.206 1.257 0.293 0.014 0.017 0.006 0.009 0.011 0.001 <0.03 0.034

X70 0.125 1.68 0.27 0.051 0.04 0.021 0.005 0.045 0.003 0.033 0.038

X100 0.059 1.97 0.315 0.024 0.23 0.315 0.002 0.022 0.022 0.046 0.037Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 53

obtained by the traditional offset procedure: the individual and mean values are listed in Table 5.

The failure-assessment diagram and the SINTAP procedure

The failure-assessment diagram

In a failure-assessment diagram, the basic fracture-mechanics’ relationship with three parameters (the applied stress (σapp), defect size (a) and fracture toughness (KIc or

Table 3. Tensile properties of the studied steels: X52, X70, and X100.

Young’s modulus E (MPa)

Yield stress σy (MPa)

Ultimate strengthσu (MPa)

Elongation at fractureAr en (%)

API 5L X52 194,000 437 616 23.14

API 5L X70 215,000 590 712 18.3

API 5L X100 210,000 866 880 6.75

Fig.2. Stress-strain curves of API 5L X52, X70, and X100 pipe steels.

Fig.3. Microstructure of the three studied steels.

X52 X70 X100

Table 4. Diameter and thickness and material of the three studied pipes.

Steel Diameter (mm) Thickness (mm)

X52 610 11

X70 710 12.7

X100 950 16

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering54

The advantages of using a failure-assessment diagram are:

• the use of a unique tool taht can be applied to any critical situation; the alternative is to use several failure criteria, from LFM, EPFM, and LA;

• to be able to obtain the safety factor Fs for any non-critical situation.

The SINTAP procedure is derived from the initial failure-assessment diagram, although the definitions of the non-dimensional parameters are a little different: the parameter kr is derived from the applied Japp parameter and fracture toughness JIc:

rkapJ

IcJ=

(3)

and the parameter Sr is replace by Lr:

rLP

LP

ref= =

σ

σ 0 (4)

JIc)) are replaced by a two-parameters relationship f(kr, Sr). Stress and defect size are combined into the applied stress-intensity factor Kapp or applied J parameter Japp, and the parameters kr and Sr are non-dimensional according to the following definitions:

rkappK

IcKand rS

app

u= =

σ

σ (1)

where σu is the ultimate strength. In the plane {Sr; kr}, a given relationship kr = f(Sr) defines the safe and the failure zones (Fig.5). Initially, the relationship between the non-dimensional stress-intensity factor kr and the non-dimensional stress S resulted from a plasticity correction that was continuously able to describe any kind of failure, from brittle fracture to plastic collapse.

A typical representation of a failure-assessment diagram is given in Fig.5, in which the load-safety factor Fs is defined according to:

sFOBOC

= (2)

Table 5. Fracture toughness of the studied steels X52, X70, and X100.

Fig.4. CT specimen extraction, geometry, and dimensions.

KI,i KI,imean δi δi,mean

(MPa√m) (MPa√m) (mm) (mm)

X52 CT1 97.5995.54

0.21 0.18

CT2 93.49 0.14

X70 CT1 117.99118.59

0.102 0.112

CT2 119.19 0.123

X100 CT1 159.98151.82

0.125 0.108

CT2 143.66 0.091

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 55

The applied parameter J is obtained by assuming proportionality between Japp and the elastic value of the J parameter Jel. The coefficient of proportionality is derived from the constitutive non-dimensional stress-strain relationship of the material.

The relationship between kr and Lr is considered as a limit curve obtained from experimental data. This limit curve can be considered in practice to be an interpolation curve between brittle fracture and plastic collapse. In this method, failure near plastic collapse is represented by data in the ‘tail’ of the diagram.

The SINTAP procedure

There are several failure-assessment diagram procedures that are similar, i.e. EPRI in the USA, R6 in UK, and RCCMR in France, each with small and more or less conservative

where P is the applied load, and PL the limit load. The material behaviour is assumed to follow the Ramberg–Osgood relationship:

ε

ε

σ

σα

σ

σ0 0 0= + ( )

n (5)

where ε0 and σ0 are, respectively, the reference strain and stress, and n is the strain-hardening exponent. The reference stress is given by:

refP

Pσ σ=

00 (6)

where P0 is the reference load.

Fig.5. A typical failure-assessment diagram (FAD), and the definition of safety factor.

Table 6. The cases studied.

Steel Diameter (mm) Thickness (mm) Operating pressure (bars)

Crack depth (mm)

Crack ratio (a/c)

X 52 610 11 70 2.2 0.4

X 70 710 12.7 70 2.54 0.4

X 100 950 16 100 3.2 0.4Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering56

where f(Lf), Lf, Lfmax, and σY, are respectively an interpolating

function, a non-dimensional loading parameter, the maximum value of non-dimensional loading or parameter, and the yield stress.

Pipe defects and the associated stress-intensity factor

A surface, longitudinal, and semi-elliptical crack in the wall of a pipe was chosen for study. This type of defect can conservatively represent a crack-like defect (Fig.6), which is one of the most common defects detected in pipes.

differences in the safe zone area. The SINTAP procedure [4] is the result of a multi-disciplinary European project designed to obtain a unified multi-level method useful for across the industry, from SMEs to large companies. The level hierarchy depends on the stress-strain curve and fracture toughness: lower levels of the procedure can be used with a simple description of the stress-strain curve but with higher conservatism.

The mathematical expressions for the SINTAP procedure for the lowest and most-conservative level (i.e. the basic level) are shown in Equn 7 below:

(7)

Fig.6. A surface longitudinal semi-elliptical crack in the wall of a pipe.

Fig.7. Failure-assessment diagram and assessment points for X52, X70, and X100 pipe steels.

f(Le , 0 L 1

r

Lr

r6

). .

+ ×

≤ ≤

−

− × 12

0 3 0 7

1 12

212Lr µ

+ ×

× < ≤

−

− −

120 3 0 7. . ,e L 1 L L

where

r

N 12N

r rmaxµ

µ = ×

= −min 0.001 E ,0.6 , L 1

2, N 0.3 1

Yrmax Y U

U

Y

Uσσ σσ

σσ

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 57

of the crack tip (point A in Fig.1) and for the crack mouth on the surface of the cylindrical shell (point B in Fig.1).

Results

Three cases that were studied and correspond to different steels. The operating pressure is considered higher for X100 steel because it is used for the new generation of pipelines which work at higher operating pressures and larger diameters.

The parameter kr has been determined using Equns 1 and 8, and Lr using Equn 1. For each case, an assessment point with coordinates (Lr*, kr*) was reported in a failure-assessment diagram (Fig.7). Each steel has its own failure-assessment diagram because the parameter μ is different for each steel, although the difference is relatively small, particularly for Lr < 0.8. We note that the three assessment points are in the safe zone, i.e. below the failure curve given by Equn 7. Then, using the procedure described in Fig.5, the safety factor was then determined and is reported in Table 7.

It can be seen that the safety factors are greater than 2 for all the steels: accordingly, the pipe is safe in each case, and the defect doesn’t need to be repaired.

The stress-intensity factor for such a crack is given by the general formula:

K pRt

a MIint π  

Φ (8)

where p is the internal pressure, Rint is the internal radius of the pipe, t the wall thickness, a the crack depth, M the geometrical factor correction, and Φ the secondary elliptic integral.

Φ = −−

∫ 102

2 2

22

π

θ θc ac

sin d (9)

An approximate value of Φ is given by:

Φ2 1 651 1 464= + . ( / ) .a c (10)

The geometrical factor M has been computed by the finite-element method [5] and reported for a/c = 0.4 and a/t = 0.2. The values obtained correspond to the chosen defect geometry. The function M is different at the lowest point

Steel X52 X70 X100

Safety factor 3.38 3.87 3.23

Table 7. Safety factors according to pipe steel.

Fig.8. Safety factor values associated with different pipe steels.

Table 8. X120 steel pipe design conditions.

Steel Diameter (mm) Thickness (mm) Operating pressure (bar)

Crack depth (mm)

Crack ratio (a/c)

X 120 1420 23 120 4.6 0.4

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering58

The required CTOD was calculated on an assumption of the existence of a surface-breaking crack of 2 mm depth at a seam weld toe, with possible shape irregularity and stress distribution. As a result, it was concluded that a CTOD of 0.08 mm or more was good enough. Since a defect equal to or larger than 2 mm an be detected by non-destructive inspection, an internal defect up to 4 mm in width will be permissible under the same value of critical CTOD. It can be seen that the safety factor decreases as the yield stress of the pipe steel increases, together with diameter, wall thickness, and operating pressure

The evolution of failure type when increasing the yield stress of pipe steels can be predicted by using a domain-failure-assessment diagram (DFAD) [7]. This is a failure-assessment diagram divided into three zones of potential failure type: brittle fracture, elastic-plastic failure, and plastic collapse. A DFAD is limited by the failure-assessment curve that gives the limit of a safe and an unsafe pipe. The safe area is divided conventionally into three zones:

• Zone 1: if the assessment point lies in this zone, increasing the applied pressure leads to brittle fracture.

• Zone 2: where increasing the applied pressure leads to elasto-plastic fracture.

• Zone 3: where plastic collapse occurs by increasing service pressure.

Discussion

The previous results indicate that the safety factor decreases when the pipe design changes to using high-strength steel such as X100. In that case, the pipe diameter, wall thickness, and operating pressure all increase simultaneously with the material’s yield stress. In order to understand the consequences of designing a new pipe with X1210 steel, the safety factor was determined using the data in Table 8 [6].

The diameter was chosen as the largest readily-available pipe diameter, and the wall thickness is compatible for seam welding the X120 pipe using the same submerged-arc welding (SAW) method (with one pass each for the internal and external welds) which had been employed for pipes of the conventional grades. The operating pressure was chosen to be an anticipated value for future operations.

Due to the unavailability of X120 pipe steel, the mechanical properties (yield stress and ultimate strength) were obtained from [6] and are the mean of the two values reported in Table 3. The fracture toughness was deduced from the two required values of the critical crack-tip-opening displacement (CTOD) δc in the base metal and in the welds at -20°C, given in Table 1. The CTOD is converted into fracture toughness using the following LFM relationship:

K Ec y c= σ δ. . (11)

Table 9. The mechanical properties of X120 steel.

Yield stress (MPa) Ultimate strength (MPa) CTOD - base metal (mm) CTOD - welds (mm)

908 981 0.14 0.08

Fig.9. Domain failure-assessment diagram, and assessment points for the four studied pipe steels.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 59

has an elastic-plastic failure potential risk. In this case, it seems necessary to evaluate the additional risk of a brittle running crack. This risk is associated with high stored energy due to the large pipe diameter and high operating pressure.

References

1. 6th EGIG report, 1970 – 2004, December 2005. Gas pipeline incidents. Doc. no EGIG 05.R.0002.

2. Norme AFNOR, 1981. NF A 03-180: Produits sidérurgique, détermination du facteur d’intensité de contrainte critique des aciers.

3. Norme AFNOR, 1987. NF A 03-182: Mécanique de la rupture, détermination de l’écartement à fond de fissure (CTOD).

4. SINTAP, 1999. Structural integrity assessment procedure. Final revision, EU-Project BE 95-1462, Brite-Euram Programme, Brussels.

5. G.Pluvinage and V.T.Sapunov, 2004. Fuite et rupture des tubes endommagés. Cépaduès Edittion, ISBN 2-85428-6448.

6. Nippon Steel, 2004. Technical report 90: Development of ultra-high-strength linepipe, X120.

7. G.Pluvinage, J.Capelle, C.Schmitt, and M. Mouwakeh, in press. Notch and domain failure assessment diagrams as a tool for defect assessment of gas pipes.

8. C.E.Feddersen, 1970. Evaluation and prediction of residual strength of center cracked tension panels ASTM STP, p50.

Based on the Feddersen diagram [8], the limit of these three zones is defined conventionally as follows:

• Zone I1: 0 < Lr < 0.62 Lr,y • Zone 2: 0.62 Lr,y < Lr < 0.95 Lr,L

• Zone 3: 0.95 Lr,L < Lr < Lr,max

where Lr,y is associated with the yield pressure, and Lr,max is the maximum value of Lr. In Fig.9, in a domain-failure-assessment diagram, are reported the assessment points of the four studied pipe steels. The X52, X70, and X100 grades have a fully ductile failure potential. However, the X120 steel has a more pronounced risk of elastic-plastic failure.

Conclusion

The risk of failure for a steel pipe has been evaluate through a conventional defect type. In the crack-like defect approach, a surface defect represents a type of external interference which often occurs on pipe: its depth, as half of the wall thickness, conservatively gives a maximum of constraint. At the operating pressure, the safety factor always exceeds the conventional value of 2, and it can therefore be concluded that is not necessary, from a fracture-mechanics’ point of view, to repair this defect.

Additionally, the use of a domain-failure-assessment diagram shows the potential of brittle or elastic fracture risk, and it has been seen that the X120 steel

Sample

copy

not fo

r dist

ributi

on

WGlobalWebb®

®

20–23 October 2013, BahrainGulf COnventiOn Centre, Bahrain

www.pipelineconf.com

COnferenCeTechnical streams presented by industry leaders covering a wide range of subjects will run over the two and a half day event.Some of the subjects to be discussed;• Planning, design, construction and materials• Operations and maintenance• Asset integrity management• Inspection and cathodic protection• Repair and rehabilitation• Automation and control• Leak detectionPaper abstracts are now being accepted.

Join leaders in the international pipeline industry as they converge for the Best Practice in Pipeline Operations and integrity Management Conference and exhibition in Bahrain.

exhiBitiOnA comprehensive exhibition will be part of the event, allowing companies from around the world to showcase their products and services. Contact us today to book

your space.

netwOrkinGThroughout the event there will be ample opportunities to network with participants to further your business relationships. Meet with industry leaders from around the world.

registrations will open soon – make sure you attend this landmark event.

organizers

Held under the Patronage of His Excellency Shaikh Ahmed bin Mohamed Al Khalifa, Minister of Finance, Minister in Charge of Oil and Gas Affairs, Chairman of National Oil & Gas Authority, Kingdom of Bahrain

PLaTinUM eLiTe sPonsor siLVer sPonsorsPLaTinUM eLiTe sPonsor

bahrain_conf_feb13_ad.indd 1 26/02/13 12:16 PM

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 61

*Corresponding author’s details: tel: +1 519 253 3000e-mail: sdas@uwindsor.ca

PIPELINES CAN BE exposed to external interference and corrosive environments and, as a result, damage in the form of dents, corrosion, cracks, and gouges can form in the pipe wall. Such damage can reduce

the pressure capacity of the pipeline. Combinations of multiple defects could be even more alarming, and provide serious concerns for the line’s safety and operation. One such combination is the formation of a dent and crack together at the same location of the linepipe. This type of defect is called a ‘dent-crack’ defect in this paper. A long-term research programme is currently underway at the University of Windsor to study the influence of various parameters, such as the level of internal pressure, dent depth, dent shape, crack length, and crack depth, on the pressure capacity of 30-in diameter and X65 grade pipes with D/t of 90. From the study completed so far it has been found that the crack depth of 2 mm or smaller does not have much damaging effect on the pressure capacity of this pipe. This paper discusses the test specimens, test set-up, test procedure, and test results obtained from this study.

1 Centre for Engineering Research in Pipelines, University of Windsor, ON, Canada2 TransCanada Pipelines Ltd, Calgary, AB, Canada

Pressure tests on 30-in diameter X65 grade pipes with dent-crack defects

by Hossein Ghaednia1, Jorge Silva1, Sara Kenno1, Prof. Sreekanta Das*1, Rick Wang2, and Richard Kania2

EXTERNAL INTERFERENCE CAUSES various defects which significantly affect the transportation of oil

and gas in pipelines. Corrosion, cracks, punctures, dents, gouges, and combination of such damage from a variety of external interferences are some common examples of surface damage in pipelines. Gouges, dents, cracks, and punctures that form in the pipe wall as a result of contact and/or impact from foreign objects are often referred to as mechanical damage. Mechanical damage of oil and gas pipelines is one of the major reasons of failure of pipelines in service, and this damage may result in loss of product, explosions, fire, human and/or animal casualties, and pollution. It has been reported that the failure of oil and gas transmission pipelines resulting from mechanical damage ranges proportionately from 55% in the USA to around 70% in Europe [1-4].

Accidental impacts are common in onshore and offshore pipeline although they are formed in a different way. Construction and excavation cause mechanical defects in onshore pipelines, while anchors or trawling gear cause

mechanical damage in offshore pipelines. This type of damage is known as dents and gouges (with or without wall cracking) in pipeline industry. A dent is an inward permanent plastic deformation of the pipe wall which causes a gross distortion of the pipe cross section [5]. A dent also causes stress and strain concentrations and a local reduction in the pipe diameter. A gouge is a metal-loss defect that occurs in the pipeline due to the scraping action of the excavating equipment or due to the rubbing action of the pipeline with a foreign object such as a rock. A gouge and dent can both form at the same time, and a dents can also exist with a wall crack inside it. These cracks occur mostly during the formation of the dent or at later due to corrosion or scratching the surface by an external object. The other mentioned reason for having a crack in the pipeline surface is fatigue or cyclic loading [6-8].

Dent defects in energy pipelines have been a major concern for pipeline operators, and a significant number of investigations have been completed to understand the effect of dents of various shapes and sizes on the behaviour of energy pipes under monotonic and cyclic pressures. It has been found and generally accepted that a dent defect alone is not a threat to the integrity of the pipe structure under monotonic pressure load when the dent depth is up to 24% of the diameter of the pipe [9-11]. Extensive

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering62

Experimental work

Test specimen

This study was accomplished by performing laboratory-based experimental tests. The purpose of performing full-scale tests was to determine experimentally the final burst strength of field-defect pipes due to dent-crack damage and also to determine the critical strain values obtainable from the tests. As a result, X65 steel pipes, of the type mostly used in oil and gas pipelines, were selected as test specimens. Six specimens were fabricated from pipe with 30-in outer diameter and wall thickness of 8.2 mm, and conforming to grade X65. The outer diameter-to-thickness of pipe ratio (D/t) of all the specimens used in this experimental work was 93, and the length of each pipe specimen was 2050 mm. The two ends of the specimen were welded to thick caps for withholding the pressure during the dent and burst test, which will be explained later. A number of tensile specimens were cut from a virgin pipe section and the standard tensile test was carried out to determine the mechanical properties of the pipe material. The properties of the material which were obtained from the tensile test were: yield strength 540 MPa, tensile strength 620 MPa, and modulus of elasticity 200 GPa.

research has also been completed to understand how the burst strength of pipelines with part-wall defects (such as corrosion alone, or a crack alone) changes as the shape of the corrosion or crack dimensions change. Detailed guidelines on predicting the burst strength of such pipes are available in various pipeline deign and maintenance codes and standards, for examples ASME 2004, DNV 2004, and CSA 2007. Numerous studies have also been undertaken by various research groups and individuals to determine the burst-pressure capacity of pipes that have developed a gouge in the dent (often known as dent-gouge defect). Various design guidelines and equations for calculating the burst strength of such pipes are available [6, 12].

The burst strength of pipes with just one of the damage forms such as dent, crack, or corrosion, and also a combined dent-gouge, has been subjected to extensive research. No studies have been undertaken to determine the burst strength of dented pipe that has developed a crack in the dent which will be called a dent-crack defect in the subsequent discussion). Therefore, this study was undertaken to develop a detailed design guideline for determining the safe burst pressure limit for various oil and gas pipelines that have developed a dent-crack.

Specimen Crack length (mm)

Crack depth (mm)*

Dent length (mm)

Dent depth (mm)

Pressure** (psi)

X65-1 100 2 100 30 (3.96%) 507 (30%)

X65-2 100 2 100 28 (3.7%) 920 (50%)

X65-3 100 2 100 28 (3.7%) 1105 (60%)

X65-4 100 2 500 (blunt) 28 (3.7%) 500 (30%)

X65-5 100 4 100 28 (3.7%) 500 (30%)

X65-6 100 4 100 46 (6%) 0

* It was almost impossible to measure the exact crack depth.

** Service pressure during the denting test.

Table 1. Properties of the test specimens.

Fig.1. The two indenters: (a) rectangular indenter (X65-1); (b) blunt indenter (X65-4).

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 63

Denting test

The experimental programme was carried out in the Structural Engineering Laboratory at the University of Windsor. Figure 3 schematically shows the test set-up for the indenting test: as described, the pipe specimens were placed on three thick steel supports and the boundary condition between the pipe and the support plane could therefore be defined as a contact interaction. The denting load was applied on the top surface of the pipe wall using a universal loading actuator.

The total plastic deformation of the pipes due to application of the indenting load was measured using two linear voltage displacement transducers (LVDTs) located at the top of the actuator and at right angles to capture stroke of the indenter. End caps welded at the end of each specimen allowed the pipes to be to pressurized using a hydrostatic pump. The internal pressure and the indenter shape varied for different specimens. The specimens were instrumented with strain gauges to monitor and record the variations of strain around the dent. Four different lines of strain gauges with eight to ten strain gauges on each line were installed. Consequently, a large area around the dent was covered to understand the critical strain location. The strain gauge layout is shown in Fig.3; no strain gauges were installed underneath the indenter because those strain gauges failed as soon as a load was applied through the actuator. Bursting test

Specimens were pressurized with a pressure washer in order to quantify the ultimate bursting strength of a pipe with a dent-crack defect. A number of strain gauges were located inside and around the dented area, and a concentration of load was observed in these regions. The pressure was monitored using a digital pressure gauge that was located on an end cap. Figure 5 shows bursting test set-up for specimen X65-3.

The full-size tests were completed in three different loading steps: fatigue or cyclic loading, dent loading, and bursting, and each of these steps is explained in detail in the following sections. All the pipes have an EDM-cut notch with 100 mm length and depth of 1 to 1.5 mm at the mid-span. This area was subjected to local dent and fatigue loads. The main parameters of this study were service pressure (when the pipe subjected to the dent loads) and the indenter shape. In most of the cases where pipeline in the field is subjected to lateral loading it contains internal pressure. In the tests, the pressure varied from 500-1105 psi, which represents 30% to 60% of the service pressure of a typical oil and gas pipeline. Indenters of two different shapes were used: a rectangular shape and a blunt shape, as shown in Fig.1; details of the test specimens are given in Table 1.

Test set-up and instrumentation

Fatigue test

The pipe specimens rested on three raised thick steel supports. The supports at the ends consisted of a pin and a roller, while to keep specimens in location, a steel collar of the same radius as the pipe was used on the mid-span support. Rubber strips were placed between the pipe and the collar to minimize sliding between them, as can be seen in Fig.2. Also, a small collar was used at the top of the specimens between the actuator and the pipe in order to prevent load concentration. Between the top collar and the pipe, a rubber strip was used as well. Using an EDM machine, a notch was cut with depth of 1 mm and length of 100 mm, and located almost 200 mm below the actuator. The rubber strip was 110 mm from the notch (Fig.2). The universal loading actuator applied cyclic sine loading from 50-100 kN on the pipe. The number of cycles to reach approximately 1 mm crack depth was chosen based on CT tests that had previously been completed at the University of Windsor.

Fig.2. Fatigue test set-up.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering64

acting more stiffly under the indenting load than X65-1 a with shorter, rectangular, dent. In other words, the blunt indenter had a larger area in contact with the pipe than the rectangular indenter. For specimens X65-1, X65-2, and X65-3, the indenter shape and size remained constant while the internal pressure was 0.3py, 0.5 py, and 0.6 py, respectively. From the load-deformation behaviour, the yielding stress was almost same in all of these specimens. During the process of denting, the pipes underwent both elastic and plastic deformation: as the denting load was removed, elastic spring-back occurred. The spring-back influences the dent depth, which is an important parameter influencing the strain concentration in a dent [5]. In Fig.4 it can be seen that for the specimens with the rectangular indenter and varying internal pressure, the amount of spring-back due to removal of the indenting load increased as the internal pressure increased. However, for specimens at the same level of internal pressure and different indenter shapes, the spring-back influence was negligible. For example, 55.5% spring-back was observed for sample X65-1 after the removal of the indenter, while the amounts of spring-back were 60.6%, 60.0%, and 54.1% for samples X65-2, X65-3, and X65-4, respectively.

Strain distribution

Strain data obtained from the strain gauges installed in the circumferential direction are presented in Fig.5. Figure 5a shows a comparison between the tensile circumferential strain between specimens X65-1 (with the rectangular indenter) and X65-4 (with the blunt and longer indenter). Figure 5b compares specimens X65-1, X65-2, and X65-3, with the same rectangular indenter but different internal pressures. In the case of the blunt indenter (X65-4), the maximum tensile circumferential strain was 0.63% obtained from strain gauge 6, located 185 mm from the centre of the notch along the perimeter of the pipe.

Test procedure

For all four specimens, the same procedure was followed whether performing fatigue, denting, or bursting tests. First, the pipe was subjected to sine-cyclic loading using a universal loading actuator. Next, the pipe specimens were filled with water to a certain level of pressure that is a fraction of py, the water pressure required for circumferential yield of the pipe steel. Following this, the universal loading actuator applied an indenting load to the pipe – in several loading and unloading steps – while the internal pressure was maintained at the same level. After completion of each load step – that is, after complete removal of the indenting load – the internal pressure in the pipe was reduced to zero in order to obtain strain data when the pipe was completely unloaded. Finally, the pipe was pressurized using the pressure washer until bursting occurred, so as to find the bursting strength of the dent-crack pipe.

Test results

Load deformation behaviour

For specimens X65-1, X65-2, and X65-3 the rectangular indenter was used, while the blunt indenter was used for specimen X65-4. For some specimens, the indenting load was applied in multiple loading and unloading steps, and the load deformation behaviour of the specimens is shown in Fig.4. It can be seen from this figure that, for the blunt indenter (X65-2), a higher load was required for indenting than that required for the rectangular indenter (X65-1). Both specimens were equally pressurized to 500 psi (0.3py) and have a 2-mm notch-crack depth. Specimen X65-1 showed a total deformation of 10 mm at a load level of 226 kN; however, X65-4 – which involved a blunt indenter – required a load of 382 kN to achieve the same amount of deformation. The reason for this is that specimen X65-4 was loaded with a longer dent, which results in the specimen

Fig.3. The indenting test: (a) test set-up scheme; (b) strain-gauge layout.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 65

that, as the internal pressure increases, the values of the maximum circumferential strain also increase. The highest strain was observed by strain gauge 4, at 130 mm from the notch. The highest circumferential strains for specimens X65-1, X65-2, and X65-3 after unloading were 1.37%, 1.91%, and 2.17%, respectively. None of these specimens showed compressive strain values around the dented area.

However, in the case of the rectangular indenter (X65-1), the maximum tensile strain was 1.37% at strain gauge 4, which was 130 mm away from the indenter centre along the perimeter of the pipe (Fig.3). Hence, the specimen with the smaller indenter (X65-1) was subjected to a higher strain. This strain occurred at a greater distance from the notch than the specimen with the blunt indenter (X65-4). From the strain-gauge readings in Fig.5b it can be observed

Fig.4. Load-deformation behaviour.

Fig.5. Circumferential strain distribution comparisons: (a) between X65-1 and X65-4; (b) between X65-1, X65-2, and X65-3.

Sample

copy

not fo

r dist

ributi

on

The Journal of Pipeline Engineering66

Specimens X65-5 and X65-1 had the same specifications except for the crack depth: that for X65-5 was 4 mm, while that for X65-1 was 2 mm. The burst pressure for specimen X65-5 was 8.96 MPa (1300 psi), and that for X65-1 was 13.96 MPa (2024 psi). Hence, X65-5 sustained 36% less pressure than the same crack-dented pipe with a 2-mm crack depth. In fact, the bursting pressure of X65-5 was 0.7py, which is less than the allowable pressure of 0.8py. Specimen X65-6 had the same bursting strength of 8.96 MPa (1300 psi) as specimen X65-5, and all other specifications were the same except for the dent depths of 46 mm (or 6%) and 28 mm (or 3.7%) respectively. The X65-6 bursting strength was equal to 0.7py.

The only parameter that had a considerable effect on bursting strength of the crack-dented pipe was the crack depth above 4 mm.

Conclusion

After completing a number of full-scale experiments on X65 grade pipe the following conclusions can be drawn.

The indenter shape affects the strain distribution and the maximum magnitude of the dent. The blunt indenter required a larger load to reach the same dent depth.

Pressure during denting affected the strain magnitude away from the dent. The smaller the pressure during indenting, the larger the stress. It must be noted that this pattern was not observed very close to the dent.

Bursting pressure was not significantly affected by dent shape, internal pressure during indenting, or by cracks with depths less than 2 mm. The only parameter to significantly affect bursting pressure was crack depth if it was at least 4 mm.

Figure 6 shows the distributions of strains in the longitudinal direction, showing similar trends to Fig.5. From Fig.6 it is evident that, as in the circumferential direction, the smaller rectangular indenter showed more strain concentration than the longer blunt indenter. For all specimens (X65-1, X65-2, and X65-3) the maximum strain was recorded at strain gauge 18, which was 45 mm away from the edge of the notch. The maximum strains recorded were 3.6% for X65-3, 3.35% for X65-2, and 1.92% for X65-3. It should be noted that all the strain values recorded from the test specimens were tensile.

Bursting strength

The value of py for X65 grade pipe specimen is 12.67 MPa (1837 psi). The bursting pressure for the specimen with a small rectangular indenter (X65-1) was 13.96 MPa (2024 psi), while that for the specimen with a longer indenter (X65-4) was 14.9 MPa (2160 psi). Hence, X65-4 showed a higher bursting pressure by 6.3%. However, both specimens had higher bursting pressures than the yielding pressure of 12.67 MPa (1837 psi) as defined by the codes. Specimens X65-1 and X65-4 had a bursting pressure of 1.10py and 1.18py, respectively. In the case of specimens with different internal pressures during denting (X65-1, X65-2, and X65-3), the bursting pressure did not vary by more than 60 psi: the values of bursting pressure were 13.96 MPa (2024 psi), 14.34 MPa (2080 psi), and 14.3 MPa (2074 psi) respectively for specimens X65-1, X65-2, and X65-3. Put alternatively, the bursting pressure was equal to 1.10py, 1.13py, and 1.13py for specimens X65-1, X65-2, and X65-3 respectively. Therefore, when the crack depth is less than 2 mm, the internal pressure and the indenter shape were not found to have a considerable effect on the final bursting pressure of the crack-dented pipes.

Fig.6. Longitudinal strain distribution comparisons: (a) between X65-1 and X65-4; (b) between X65-1, X65-2, and X65-3.

Sample

copy

not fo

r dist

ributi

on

1st Quarter, 2013 67

6. E.R.Lancaster and S.C.Palmer, 1996. Strain concentrations in pressurized dented pipes. J.Process Mechanical Engineering, Part E, pp29-38.

7. I.B.Iflefel, D.G.Moffat, and J.Mistry, 2004. The interaction of pressure and bending on dented pipe. Int. J. Pressure Vessel and Piping, 82, pp761-769.

8. S.A.Karamanos and K.P.Andreadakis, 2006. Denting of internally pressurized tubes under lateral loads. Int. J. Mechanical Science, 48, pp1080-1094.

9. A.M.Gresnigt, S.A.Karamanos, and K.P.Andreadakis, 2007. Lateral loading of internally pressurized steel pipes. J. Pressure Vessel Technology, 129, pp630-637.

10. E.Dama, S.A.Karamanos, and A.M.Gresnigt, 2007. Failure of locally buckled pipelines. Ibid., 129, pp272-279.

11. B.C.Pinheiro and I.P.Pasqualino, 2009. Fatigue analysis of damaged steel pipelines under cyclic internal pressure. Int. J. Fatigue, 31, pp962-973.

12. W.A.Maxey, 1986. Outside force defect behaviour. Technical report submitted to the Pipeline Research Committee of the American Gas Association, Report No. 162, AGA Catalogue No. L51518.

References

1. R.B.Smith and D.N.Gideon, 1979. Statistical analysis of DOT-OPSO data. AGA 6th Symposium on Linepipe Research, Huston, Texas, ppD.1-D.9.

2. K.C.Wang and E.D.Smith, 1982. The effect of mechanical damage on fracture initiation in linepipes: Part I-dents. Internal report submitted to Physical Metallurgical Research Laboratories of CANMET, Ottawa, Canada, Report No. ERP/PMRL 82-11(TR).

3. M.F.Zarea, D.N.Toumbas, C.E.Philibert, and I.Deo, 1996. Numerical models for static and dynamic puncture of gas transmission linepipe and their validation. International Pipeline Conference, June, Vol 2, pp777-784.

4. E.R.Lancaster and S.C.Palmer, 1996. Burst pressure of pipes containing dents and gouges. J. Process Mechanical Engineering, Part E, pp19-27.

5. A.Cosham and P.Hopkins, 2004. The effect of dents in pipelines: guidance in the pipeline defect assessment manual. Int.J. Pressure Vessels and Piping, Elsevier, 81, pp127-139.

Sample

copy

not fo

r dist

ributi

on

18 - 20 March 2013HCC, Hannover, Germany

8th Pipeline Technology Conference

Supported by

The ptc advisory committee

More information and all previously published papers at: www.pipeline-conference.com

Pipeline Technology Conference 2010

Euro Institute for Information and Technology Transfer

WISSEN für die ZUKUNFTOldenbourg Industrieverlag GmbHVulkan-Verlag GmbH

ptc Ad PIN 12 2012.indd 1 22.10.2012 10:43:39

Sample

copy

not fo

r dist

ributi

on

TRAINING

20137-11 April Pipeline integrity workshop (Banff )

11-12 April Microbiological corosion (Houston)

15-19 April Onshore pipeline engineering (Houston)

17-19 April Inspection of subsea pipelines (Houston)

22-26 April Deepwater riser engineering (Houston)

22-26 April Subsea pipeline engineering (Houston)

6-10 May (provisional) Practical pigging course (Tricht, Netherlands)

15-16 May Unpiggable Pipelines Solutions Forum (Houston)

3-7 June Pipeline defect identifi cation & sizing (Houston)

3-4 June Pigging and inline inspection (Houston)

3-4 June DOT pipeline safety regulations (Houston)

3-5 June Defect assessment in pipelines (Houston)

3-5 June Pipeline integrity management (Houston)

5-7 June Defect assessment calculations workshop (Houston)

5-7 June Advanced risk management (Houston)

6-7 June Stress-corrosion cracking (Houston)

17-18 June CO2 trasnportation by pipeline (Newcastle, UK)

19-20 June 4th International Forum on the Transportation of CO2 by Pipeline (Newcastle, UK)

24-28 June Practical pigging course (Bergen, Norway)

16-20 September Practical pigging course (Rio de Janeiro, Brazil)

23-26 September Rio Pipeline Conference (Rio de Janeiro, Brazil)

1-4 October Subsea production systems (Houston)

6-9 October 6th Pipeline Technology Conference (Ostend, Belgium)

20-23 October 2nd Best Practices in Pipeline Operations and Integrity Management Conference (Bahrain)

28-30 October Defect assessment in pipelines (Kuala Lumpur, Malaysia)

28 Oct-1 Nov Onshore pipeline engineering (Kuala Lumpur, Malaysia)

APR 2013

MAY 2013

JUN 2013

SEP 2013

OCT 2013

Training courses – 2013

ARE YOU UP TO SPEED?102000 2

Working with a faculty of 38 leading industry experts, Clarion and Tiratsoo Technical are privileged to provide some of the best available industry based technical training courses for those working in the oil and gas pipeline industry, both onshore and off shore.

Complete syllabus and registration details for each course are available at:

www.clarion.org

Training_Ad.indd 1 18/03/13 5:30 PM

Sample

copy

not fo

r dist

ributi

on

Sample

copy

not fo

r dist

ributi

on