journal of pipeline engineering incorporating the journal of

March, 2014 Vol.13, No.1

Great Southern Press Clarion Technical Publishers

Journal of Pipeline Engineering

incorporating The Journal of Pipeline Integrity

Sample

copy

- not

for di

stribu

tion

Journal of Pipeline Engineering

Editorial Board – 2014

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, Rosen Engineering, Karlsruhe, Germany

Jorge Bonnetto, Operations Director TGS (retired), TGS, Buenos Aires, ArgentinaDr Andrew Cosham, Atkins, 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, Department of Civil and Environmental Engineering, Faculty of Engineering and Design, Carleton University, Ottawa, ON, 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, DNV Kema, Groningen, Netherlands

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

Patrick Vieth, President, Dynamic Risk, The Woodlands, TX, USADr Honggang Zhou, Center for Reliable Energy Systems, Dublin, OH, USA

Dr Joe Zhou, Technology Leader, TransCanada PipeLines Ltd, Calgary, CanadaDr Xian-Kui Zhu, Principal Engineer, Edison Welding Institute, Columbus, OH, USA

� � �

Sample

copy

- not

for di

stribu

tion

1st Quarter, 2014 1

The Journal of Pipeline Engineeringincorporating

The Journal of Pipeline IntegrityVolume 13, No 1 • First Quarter, 2014

ContentsProf Andrew Palmer, Jiexin Zheng, Paul Brunning, and Gerry Lim ...................................................................... 5

Fishing trawl pull-over across pipelines

Richard Kania, Ralf Weber, and Stefan Klein ..................................................................................................... 13On the assessment of low-frequency electric-resistance-welded linepipe defects by EMAT and CMFL in-line inspection

Dr Yong-Yi Wang and Dr Ming Liu ..................................................................................................................... 21Real world considerations for strain-based design and assessment

Prof. Dr Dimitri G Pavlou .................................................................................................................................. 29Pressure-wave propagation in multi-layered fibre-reinforced polymeric pipelines due to hydraulic hammer

Dr Maciej Witek ................................................................................................................................................. 37An assessment of the effect of steel pipeline wall losses on the maximum allowable operating pressure of a gas pipeline

Mohamed Dafea, Dr Phil Hopkins, Roland Palmer-Jones, Patrick de Bourayne, and Lionel Blin ........................ 49An investigation into the failure of a 40-in diameter crude oil pipeline

Dr Chris Alexander, Alexander Aalders, William Bath, Brent Vyvial, Rhett Dotson, and Danny Seal .................. 63Evaluating anchor impact damage to the subsea Canyon Chief pipeline using analysis and full-scale testing methods

❖ ❖ ❖

OUR COVER PHOTO shows a 24-in diameter Rosen EMAT tool being prepared for launch. Photo courtesy of Rosen Engineering.

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

for di

stribu

tion

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: © 2014 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 2014 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: [email protected]• web: www.j-pipe-eng.com• www.pipelinesinternational.com

Editor: John Tiratsoo• email: [email protected]

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: [email protected]

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

for di

stribu

tion

1st Quarter, 2014 3

Editorial

IN THIS ISSUE three significant papers look at aspects of what goes wrong when an on- or offshore

pipeline is damaged. The opening paper, by Prof. Andrew Palmer and Jiexin Zheng from the National University of Singapore, and their co-authors from Subsea 7, re-examine the issue of what happens when a fishing trawl board is pulled over a pipeline.

Fishing trawls can damage pipelines on the seabed, and it is important to be able to predict the force on a pipeline when trawlgear is pulled over it. Analysis and comparison with full-scale measurements indicate that the conventional calculation is incomplete, but that it is usually conservative. The pull-over load has more than one component, and the components depend on the trawl velocity in different ways.

Trawlgear interaction with a pipeline is a complex sequence of processes that has more than one stage. The first impact stage occurs immediately after the trawl board or trawl beam makes contact with the pipeline, and the forward movement of the gear then pauses briefly. Shortly afterwards, the ‘pull-over’ stage begins, and a little later the gear climbs over the pipeline, sometimes after sliding along it (if the gear does not approach at a right angle). A number of research studies have been carried out into this issue, and have included analyses, model tests, and full-scale tests and, as a result of these, DNV published a Recommended Practice (DNV-RP-F11110 in 2010.

The authors of this paper conclude that Equations 4.2 and 4.3 in the DNV Recommended Practice are not fully consistent with observations in many tests, most of them at full scale. In particular, the observed velocity effect is weaker than it is in the equations. A more complete analysis recognizes that the pull-over force has more than one component, and that the different components depend on velocity in different ways; the pull-over force needs to take account of the weight of the trawlgear and its motion across the pipeline. That said, the analysis represented by the two DNV RP equations appears generally to give conservative results. Non-conservative results are found only when the equations are applied to very low velocities, too slow to occur in trawling practice, or to extremely long and flexible towing warps.

THE SECOND paper covering issues surrounding pipeline failure (on pages 49-62) has been prepared

by Mohamed Dafea of GL Noble Denton (now DNV GL) and co-authors from Penspen, Trail, and Société du Pipeline Sud Européen. Its subject is an investigation into the failure of a 40-in diameter crude oil pipeline.

In August, 2009, there was a 2.5-m long rupture in the longitudinal seam weld of a crude oil pipeline in France, causing a spillage of approximately 2000 cum in a protected area. This rupture caused the authorities to withdraw the permit to operate a 260-km long section of this pipeline. The pipeline had had a similar failure in August, 1980 which was attributed to a fatigue crack initiating at the inner side of the longitudinal weld. There was evidence of ‘roof topping’ along this weld.

Penspen Ltd was contracted by the pipeline operator to carry out an independent investigation into the cause of the 2009 failure, review and confirm the actions needed for safe short-term operation to allow internal inspection, and determine a safe future life for the pipeline. The failed pipe specimen was not immediately available for inspection, and an initial analysis indicated that a purely analytical evaluation would not provide conclusive results, due to variability in material properties, geometry, and loading. It was decided that a better understanding of the behaviour of defects in the pipe, and the fatigue performance, was required.

A detailed laboratory programme of burst and fatigue testing on a section of the linepipe was recommended, and these tests were carried out on a combination of ‘defect-free’ ring specimens, and ring specimens with artificially machined slits to represent crack-like defects. These tests showed that the failure pressures of the burst tests could be predicted using a recognized industry model: in this case, roof topping, laminations and inclusions, and toughness variations were found to have no noticeable effect on the defect size at failure. The corresponding fatigue-test results showed that the defect-free rings had a fatigue life one order of magnitude longer than those containing a machined slit, and that the fatigue life of a defect-free ring could be predicted using a standard S-N method. In addition, it was found that crack growth was conservatively predicted using standard fatigue fracture mechanics.

The fatigue-test results showed that the 1980 and the 2009 failures were caused by a combination of cyclic pressure loading, roof topping, and a pre-existing weld defect (probably present when the pipeline went into service in 1972). As a result of this failure investigation, a conservative methodology for predicting the remaining fatigue life of the pipeline has been developed and validated by testing, and the pipeline is now satisfactorily back in operation.

IN THE THIRD paper on the topic of pipeline failure in this issue, Dr Chris Alexander f Stress

Engineering Services and co-authors from Williams

Sample

copy

- not

for di

stribu

tion


The work associated with this study represents one of the more comprehensive efforts conducted to date in evaluating damage to a subsea pipeline. The results of the analysis and testing work provided Williams with a solid understanding of the behaviour of the damage inflected to the pipeline and what level of performance can be expected from the repaired pipeline during future operation. After the engineering analysis and testing phases of this work were completed, the deepwater pipeline was repaired using a grouted subsea repair sleeve. In reviewing the associated body of work presented in this paper, the authors make the following observations:

A horizontally-split sleeve can be safely installed by an ROV on a live pipeline by use of a buoyancy module and pull-down winches.Cement grout can be prepared on a surface vessel and then pumped down a long hose to fill a grout sleeve on the seafloor.

A purpose-built metrology tool with mechanical gauges and simple fabrication techniques can produce an accurate model of the shape of a damaged pipeline on the seafloor.

Using analysis and full-scale testing techniques prior to deployment of repair methods improves confidence in the design and ensures that stresses in the damaged region of the pipeline are reduced to acceptable levels.

Midstream, Saipem America, and GL Noble Denton International, evaluate anchor-impact damage to the subsea Canyon Chief pipeline using analysis and full-scale testing methods.

As the authors point out, this paper presents findings from a study conducted as part of a joint-industry effort to evaluate the severity of damage inflicted to Williams’ subsea 18-in x 0.875-in wall thickness, Grade X-60 Canyon Chief gas export pipeline due to an anchor impact at a water depth of 2,300 ft. The 158-km long natural gas gathering pipeline connects Williams’ Devils Tower platform and its Blind Faith natural gas lateral to MP 261A on the Gulf of Mexico’s continental shelf, where connection is made to Williams’ Transco pipeline. The natural gas from this system is primarily delivered to Williams’ Mobile Bay gas-processing plant in Coden, Alabama, and the system parallels the Mountaineer oil pipeline for around half its route.

The phases of work described in this paper included an initial assessment after the damage to the deepwater pipeline was detected, evaluating localized damage using finite-element analysis based on in-line inspection data, and full-scale destructive testing including burst tests. The final efforts included the design and evaluation of a subsea-deployed repair sleeve, and the study included modelling Saipem’s repair sleeve design accompanied by full-scale destructive testing. Strain gauges were used to measure strain in the reinforced dent beneath the sleeve: these were then compared to prior results for the unrepaired dent test results.

Sample

copy

- not

for di

stribu

tion

1st Quarter, 2014 5

FISHING TRAWLS CAN damage pipelines on the seabed. It is important to be able to predict the force on a pipeline when trawlgear is pulled over it. Analysis and comparison with full-scale measurements

indicate that the conventional calculation is incomplete, but that it is usually conservative. The pull-over load has more than one component, and the components depend on the trawl velocity in different ways. The factors that need to be included in more complete models are discussed.

*Corresponding author’s contact details:tel: +65 6516 4601email: [email protected]

by Prof Andrew Palmer*1, Jiexin Zheng1, Paul Brunning2, and Gerry Lim2

1 Department of Civil and Environmental Engineering, National University of Singapore, Singapore

2 Subsea7, Singapore

Fishing trawl pull-over across pipelines

FISHING VESSELS DRAG otter trawls and beam trawls across the seabed. If the trawlgear strikes a

pipeline, it can damage the pipe, by indenting the pipe wall, fracturing the concrete weight coating, damaging the anti-corrosion coating, or moving a section of the pipeline bodily across the seabed.

The North Sea is intensively fished, and the sizes and weights of trawlgear continue to increase. Bottom trawling is moving into deeper and deeper water: as an example, the pipeline route surveys in 2004 for the Medgaz pipeline crossing from Algeria to Spain picked up trawl scars 0.3 m deep in 602 m of water at KP170 and in 401 m at KP174, within the upper part of the continental rise to the Spanish coast at Cabo de Gata [1].

At the very beginning of pipeline development in the North Sea, it was thought that all pipelines had to be trenched to protect them from trawlgear, though in shallow water closer to shore trenching was also for other reasons such as stability. Later engineers came to think that trenching was not always necessary, and in the late 1970s Shell devoted a substantial research effort to demonstrating that it would be unnecessary to trench its FLAGS gasline over most of its length. The industry concluded that only pipelines smaller than 16 in (406.4 mm) in diameter would need to be trenched. A number of research studies [2-9] were carried out, and included analysis, model tests, and full-scale tests, and DNV published a Recommended Practice DNV-RP-F111 [10] in 2010. Some efforts have been made to lower the diameter limit below 16 in.

Trawlgear interaction with a pipeline is a complex sequence of processes that has more than one stage. The first impact stage occurs immediately after the trawl board or trawl beam makes contact with the pipeline. The forward movement of the gear then pauses briefly. Shortly afterwards the pull-over stage begins, and a little later the gear climbs over the pipeline, sometimes after sliding along it (if the gear does not approach at a right angle).

Igland [11] developed a finite-element model in ANSYS to analyse a heavy clump weight dragged over the seabed by a warpline and pulled over the pipeline. The output of that trawlgear-interaction analysis was the pull-over force magnitude, duration, and the impact response on the pipeline. The pull-over loads and duration compared with the loads amplitude and duration given by DNV-RP-F111, and results showed that the actual stiffness of pipeline and the pipe-soil interaction would reduce trawl loads; the method in DNV-RP-F111 was therefore conservative.

A group of researchers at the Norwegian University of Science and Technology (NTNU) used the SIMLA software to simulate the trawlgear pipeline interaction. Johnsen cited unpublished work by Moller which focused on pull-over loads from polyvalent trawl boards with different span heights and trawling velocities. Comparisons were made with DNV-RP-F111, and it was found that the forces were similar only when the free span was high. The velocity change from 2 m/s to 3 m/s had a smaller effect in the simulation than in DNV-RP-F111. Johnsen [12] and Longva [13] focused on trawl board-pipeline interaction, and Johnsen developed a new hydrodynamic load model using SIMLA for the simulations. The seabed proximity and forward-speed effects of the trawl board were considered in the

Sample

copy

- not

for di

stribu

tion


the DNV design loads are generally higher than those given by model tests. She also pointed out that those design loads have higher pull-over forces and longer pull-over times than those observed in full-scale tests, especially for small-diameter pipes.

Problem

An assessment of impact damage requires knowledge of the magnitude of the force between the trawlgear and the pipeline. The DNV-RP-F111: 2010 standard [10] says that the maximum horizontal force Fp applied to the pipe is given by its equations (4.2) and (4.3) in clause 4.3, that is:

Trawl boards

1/2( )P F t wF C V m k= (1)

Beam trawls

1/2[( ) ]P F t a wF C V m m k= + (2)

where

V is the trawling velocitykw is the warp line stiffnessmt is the steel mass of the board or beam with shoesma is the hydrodynamic added mass and the mass of entrained water, andCF is a dimensionless function of “…trawlgear type and some geometrical parameters…” and is given in equations (4.4) through (4.8) in the RP as a function of the pipeline diameter, the trawlgear height, and the span height (if the line is in a span).

Equations 1 and 2 have appeared many times. The simple argument that leads to them is to say that immediately before the initial impact the trawlgear kinetic energy is (1/2)mtV2, and that immediately after the impact the gear is stationary and all that energy is transferred to the elastic strain energy of the towing warp. If the warp has tension Fp and constant stiffness kw (tension/elongation), the strain energy in the warp is (1/2)Fp

2/kw. Equating the gear kinetic energy before impact to the warp strain energy afterwards gives Equn 1 at once.

The equations are by themselves an incomplete model of a complex situation. Johnson [16] wrote about it straightforwardly, and one cannot do better than repeat his argument:

“It is common in courses in Strength of Materials to estimate the greatest stress in a long uniform vertical rod due to the impact of a falling rigid

finite-element model to calculate the pull-over force. Johnsen looked into the oblique trawl board crossing and concluded that a perpendicular crossing did not predict the largest pull-over load. A comparison with DNV-RP-F111 showed that the RP over-predicts the pull-over loading.

Longva et al. [14] improved the contact elements, and the comparison with experimental results showed good consistency. The finite-element model was validated against various experiment results, and a parametric study was conducted. The authors concluded that if the ratio of pipe diameter to trawl board height is less than 0.3, it is acceptable to ignore the effect of the pipeline on the behaviour of the board. It was also found that non-perpendicular crossings at high span heights induced the maximum pull-over forces.

Maalo [15] worked on the experimental model test and the simulation of clump weight-pipeline interaction of a fixed pipe section at low span heights. The main finding is that the increasing in pipeline flexibility resulted in a decrease in pull-over force. The pull-over loads were lower in the simulations than in DNV-RP-F111 (Johnsen [12]).

Johnsen [12] investigated the clump-weight trawlgear interaction with submarine pipelines with FE simulations. She mentioned that the pull-over load of clump weight calculation methods in DNV-RP-F111 was based on an experimental model test executed at MARINTEK in 2004. In that test programme, 139 tests were done, and all were model tests with scaling according to Froude’s law. The scale factor was 1:10, and three pipelines with diameters 350 mm, 530 mm, and 840 mm were modelled. Six different span height were used, and three different pipe end conditions. The modelled pipeline length was 25 m. Four different velocities were used, 1.45 m/s, 1.75 m/s, 1.95 m/s, and 2.18 to 2.45 m/s for different clump weight types. The author found that

Fig.1. Dynamic tension induced by a falling weight (redrawn from Fig.1.22 in Ref.11).

Sample

copy

- not

for di

stribu

tion

1st Quarter, 2014 7

tell us that if the velocity is infinitesimally small, the pull-over force is also infinitesimally small. For several reasons, though, very small velocities do not occur in trawling practice; in particular, a low velocity will not hold the net open. Similarly, if the towing warp is very long (or is in a catenary or carries weights), the warp stiffness kw is very low and the pull-over force becomes small.

The coefficient CF given in the RP is between 3 and 4 for beam trawls with hoop bars and between 3.5 and 5 without hoop bars, coefficients that are much greater than unity. The interaction between trawlgear and a pipeline is extremely complex, and the equations in the RP represent only part of the processes involved. They have been calibrated by the inclusion of the CF factors in such a way that the forces predicted by the equations are generally conservative.

Observations in tests

The same conclusion can be reached from tests. A relevant question is what exactly the tests measure, a notoriously difficult question with impact tests. Most tests site a load cell at the lower end of the towing warp, immediately before the trawlgear; some tests also have accelerometers on trawl boards.

As long ago as 1980, Moshagen and Kjeldsen [2] reported field tests on 16-in and 36-in pipelines, and laboratory tests on models of 48-in lines (scaled-up by applying Froude scaling). Figure 2 replots their pull-over force data for conventional beam trawls. It can be seen that the pull-over force is not proportional to velocity, and that the velocity effect is much weaker. They also observed a weak or absent velocity effect for otter trawls.

mass M1, see Fig.1.22, by assuming that all the kinetic energy of M1 is absorbed as uniformly distributed strain energy in the rod. Then the stress attained is given by:

2

22

2 1 01.

2 2A M v

Eσ

=

(3)

22 21 0 1

2 0 2 22 2 2/

M v ME v EM M

σ ρρ

= =

(4)(1.51 in Johnson)

“A2, ℓ2, ρ2, E2, and v0 denote the cross-sectional area of the bar, its length, its density, and its Young’s modulus, and v0 is the speed of impact of M1 on the weightless collar at the bottom of the rod. M2 is the mass of the rod.

“Now the dynamic stress initiated (Johnson’s italics) at impact, σ0, is really given by

2 220 2 0 0 2 0 2

2

orEv v Eσ ρ σ ρρ

= =

(5)(1.52 in Johnson)

σ0 is independent of M1.”

Equation 5 is a well-known result from impact mechanics: see, for example, Kolsky [17], Goldsmith [18], or Johnson [16].

The equations 1 and 2 predict that pull-over force is proportional to impact velocity, and are applicable in the range where they are customarily applied. They are not applicable at very small velocities, because they

Fig.2. Pull-over forces from beam trawls on 16-in, 36-in, and 48-in pipelines (redrawn from Figs 13 and 14 in Ref.2).

Sample

copy

- not

for di

stribu

tion


effect is to generate an impulsive tension F in the towing warp, given by:

F V mk= (6)

where:

m is the mass per unit length of the towing warpk is the axial stiffness (axial force/axial strain) of the towing warp

Note that k in Equn 6 is not the same as the warp stiffness described in the RP, which has different dimensions. The impulsive increase of tension moves up the towing warp with velocity

kcm

= (7)

The RP10 gives an equation (4.9) for “the warp stiffness of one single wire (typically 32-38 mm diameter)” which implies that k is 3.5 × 107 N. Klinke [20] gives the mass per unit length of 1.375-in (34.9-mm) IWRC wire rope as 3.5 lb/ft (5.21 kg/m). If therefore the trawl velocity V is 3 knots (1.545 m/s), then from Equns 6 and 7 the impulsive tension F is 21 kN (2.1 tonnes), and the increase of warp tension moves with a velocity of 2590 m/s. This corresponds to the impact component of the warp tension.

A tension stress wave is reflected from a rigid boundary as a second tension wave with equal magnitude. If no other factors were present, the superposition of the transmitted wave and the reflected wave would double the force. The RP correctly points out that a

Similar conclusions have been reached by other research. Trevor Jee Associates [7] carried out tests on a 1/18 scale model otter trawl. They divided the total warp force into components, and defined a ‘delta force’ as the difference between the baseline force before the impact and the peak force during the impact. Figure 3 replots their data, and shows that the delta force depends hardly at all on velocity, whereas the baseline force is closely proportional to (velocity)2, as we would expect if the effective drag coefficient is nearly independent of the Reynolds number.

Later tests [8] pulled an otter trawl across the Skene pipeline bundles. In the words of the report:

“Figure 7-4 describes the impact warp tensile forces as a function of velocity…This shows that there is no apparent effect of velocity on the loads produced at 45º and shallower angles of impact and that there is a suggestion that the impact force may be reduced at the lowest velocity measured for the 90º angle impact.”

Our experiments [19] support the same conclusion. Figure 4 plots the results of pull-over tests on a model beam trawl pulled at right angles to a pipeline. The broad range of measurements at 2 m/s reflects the much larger number of tests at that velocity. Once again the delta force is independent of velocity.

Analysis

If we idealise the collision as between a rigid mass moving with velocity V (representing the trawlgear) and a rigid, stationary, and immovable pipeline, the

Fig.3. Warp forces for ‘perfect’ trawl boards in otter trawls (redrawn from Fig.34 in Ref.7).

Sample

copy

- not

for di

stribu

tion

1st Quarter, 2014 9

shoe of a beam trawl in contact with a 416.4-mm outside diameter pipeline (16-in OD + 5-mm coating) slightly embedded in the seabed. The tension in the warp required to start to lift and rotate the shoe can be calculated by taking moments about the contact point C, taken to be 150 mm above the mudline, corresponding to an embedment of 66.4 mm, about 15% of diameter. Assuming the mass of the half the beam to be 860 kg (Ref. 2, table 1), corresponding to a submerged weight of 750 kgf (7.4 kN), the warp attachment point at the location shown, and the warp at 20º to the horizontal, the warp tension comes out to be 1520 kgf (14.9 kN). The corresponding horizontal force applied to the pipe is 14.0 kN, and the tangential force applied to the pipe is 2.3 kN. The ratio between the normal and tangential reactions at C is 0.16, and so the shoe will roll on the surface of the pipe without sliding, if the coefficient of friction between the shoe and the pipe is greater than that (as it almost always will be).

This calculation shows that a significant warp tension is required to roll and slide the trawlgear across the pipeline, however slowly the trawlgear is moving, and the tension is independent of velocity and comparable in magnitude to the impact tension calculated above. The calculation explains why only part of the measured warp tension is velocity-dependent. It is based on a severely idealised two-dimensional model, and could be made more complicated, but the conclusion is clear.

If the shoes at either end of a beam trawl do not contact the pipeline simultaneously, the beam and both shoes have to be lifted and rotated by the warp tension applied to the first shoe to hit, and this will approximately double the additional warp tension.

catenary warp line will have a somewhat lower stiffness. Moreover, the dynamically increased tension in the warp line will induce transverse movement, which will generate additional hydrodynamic forces and dissipate energy. The reflected stress wave can be expected to be significantly smaller that the transmitted wave, and Ref. 7 confirmed that the warp tension peaks measured at the towing vessel are much smaller that the peaks measured at the trawlgear, and attributed that to the catenary.

Comparison with tests and with previous analysis

The results in the previous sections help us to understand why the velocity effect is weaker than Equns 1 and 2 would lead us to expect. The impulsive component of tension in the towing warp is given by Equn 6, possibly with a multiplier of between 1 and 2 to allow for reflection. The calculated impulsive tension is significantly smaller than the measured tension. This suggests that the velocity effect in Equn 6 applies to only a part of the pull-over force, and that in turn is consistent with Figs 2 to 4.

Additional components of warp tension

Measured total warp tensions during pull-over are much higher than the impact component alone. Where does the rest of the tension come from?

Imagine first that the gear is pulled very slowly, so that the loading is quasi-static and there are no dynamics. The trawlgear has to rotate and lift as it begins to cross the pipeline. Figure 5 is an end view of one

Fig.4. Pull-over forces from model tests on beam trawls [14].

Sample

copy

- not

for di

stribu

tion


effect is weaker than it is in the equations. A more complete analysis recognizes that the pull-over force has more than one component, and that the different components depend on velocity in different ways. It needs to take account of the weight of the trawlgear and its motion across the pipeline. That said, the analysis represented by equations 4.2 and 4.3 in DNV-RP-F111 appears generally to give conservative results. Non-conservative results are found only when the equations are applied to very low velocities, too slow to occur in trawling practice, or to extremely long and flexible towing warps.

Acknowledgement

The authors thank Subsea7 for permission to publish this paper, and for continued support of the larger research programme of which this forms part. They thank Trevor Jee, Richard Verley and colleagues at Subsea7 for helpful comments, and Sun Shu for help with the tests. It is emphasized that the conclusions put forth reflect the views of the authors alone, and not necessarily those of Subsea7.

References

1. Medgaz: the first direct pipeline from Algeria to Europe. Slide presentation (undated).

2. B.H.Moshagen and S.P.Kjeldsen, 1980. Fishing gear loads and effects on submarine pipelines. Proc. Offshore Technology Conference, Houston, paper OTC3782.

3. R.L.P.Verley, B.H.Moshagen, N.C.Moholdt, and I.Nygaard, 1992. Trawl forces on free-spanning pipelines. Int. J. of Offshore and Polar Engineering, 2, 1.

A real trawl is, of course, in motion, and several other dynamic and hydrodynamic effects are present. The tension require to lift the shoe and impart a rotational acceleration has an additional dynamic component proportional to the angular acceleration. Its significance can again be assessed by a rough calculation. Suppose that the moment of inertia for rotation in the plane of the shoe and referred to the centre of gravity is 34 kgm2, corresponding to the 860 kg mass and a mean square radius of 0.2 m. The moment of inertia referred to C is 316 kgm2. If the angular acceleration is uniform as the shoe rotates through a quarter of a revolution, without sliding, bringing the contact point to the top of the pipeline in 500 ms, the angular acceleration is 12.6 rad/s2 and the required torque about C is 4.0 kNm. The warp tension required to induce that angular acceleration is 12.4 kN (1.27 tonnes). That is again comparable with the other components of warp tension, and cannot be neglected. It is velocity-dependent.

The trawlgear comes momentarily to a stop when it impacts on the pipeline, but some of the water surrounding the gear is still moving, and so the hydrodynamic added mass effect exerts a forward force on the gear, some of which will be transmitted to the pipeline. In addition, the moving water induces a hydrodynamic drag force that acts directly on the pipeline. The trawl net itself is continuing to move and so the tension in the sweepline between the net and the beam or otter board must decrease, increasing the net forward force.

Conclusions

Equations 4.2 and 4.3 in DNV-RP-F111 are not fully consistent with observations in many tests, most of them at full scale. In particular, the observed velocity

Fig.5. Impact of a beam trawl shoe on a pipeline.

Sample

copy

- not

for di

stribu

tion

1st Quarter, 2014 11

12. I.B.Johnsen, 2013. Clump-weight trawl gear interaction with submarine pipelines. Norwegian University of Science and Technology, Trondheim.

13. V.Longva and S.Sævik, 2012. A penalty-based body-pipeline contact element for simulation of pull-over events. ASME 2012 31st Int. Conf. on Ocean, Offshore and Arctic Engineering. Rio de Janeiro, Brazil. Vol.3: Pipeline and Riser Technology: 241-250.

14. V.Longva, S.Sævik, et al., 2013. Dynamic simulation of subsea pipeline and trawl board pull-over interaction. Marine Structures, 34(0): 156-184.

15. K.Maalø, H.S.Alsos, et al., 2010. Detailed analysis of clump-weight interference with subsea pipelines. ASME 2012 31st Int. Conf. on Ocean, Offshore and Arctic Engineering. Rio de Janeiro, Brazil. Vol.3: Pipeline and Riser Technology: 725-732.

16. W.Johnson, 1972. Impact strength of materials. Edward Arnold, London.

17. H.Kolsky, 1953. Stress waves in solids. Clarendon Press, Oxford.

18. W.Goldsmith, 1960. Impact. Edward Arnold, London.

19. J.Zheng, 2014. Unpublished PhD dissertation, National University of Singapore.

20. J.A.Klinke, 2008. Rigging handbook. ACRA Enterprises, Stevensville, MN,USA.

4. R.L.P.Verley, 1994. Pipeline on a flat seabed subjected to trawling or other limited duration point loads. Proc. 4th International Offshore and Polar Engineering Conference, Osaka, 128-134.

5. T.Mellem, J.Spiten, R.L.P.Verley, and M.Moshagen, 1996. Trawl board impacts on pipelines. Proc. Offshore Mechanics and Arctic Engineering Conference, 5, 165-177.

6. O.Fyrileiv, J.Spiten, T.Mellem, and R.L.P.Verley, 1997. DNV’96 acceptance criteria for interaction between trawlgear and pipelines. Proc. Offshore Mechanics and Arctic Engineering Conference, 5, 91-98.

7. Trevor Jee Associates (now Jee Ltd), 2003. Overtrawling large diameter pipelines: flume tank tests interpretative report.

8. Idem., 2004. Skene bundle overtrawling trials; offshore trials interpretation report.

9. D.Ø.Askheim and O.Fyrileiv, 2006. New design code for interference between trawlgear and pipelines – DNV RP-F111. Proc. Offshore Mechanics and Arctic Engineering Conference, paper OMAE2006-92127.

10. Det Norske Veritas, 2010. Interference between trawlgear and pipelines. Recommended Practice DNV-RP-F111, Høvik.

11. R.T.Igland and T.Soreide, 2008. Advanced pipeline trawl gear impact design. ASME Conference Proceedings (48203): 271-277.

Sample

copy

- not

for di

stribu

tion

The definitive conference forpipeline professionals. Don’t miss

this five-day experience!

INTERNATIONAL PIPELINE CONFERENCE & EXPOSITION

THE CONFERENCE

September 29 - October 3, 2014The Hyatt Regency Hotel & TELUS Convention Centre

Calgary, Alberta, Canada

Conference Patrons

internationalpipelineconference.com

THE EXPOSITION

Participate at the world’s leading marketplace for pipeline technology and services!

Major Exposition SponsorsMedia Sponsor

September 30 - October 2, 2014TELUS Convention CentreCalgary, Alberta, Canada

internationalpipelineexposition.com

Attend 13 Technical Tracks with 300+ presentations, keynotes and tutorials.

300Network with over 1,500 industry professionals at sessions and official functions.

1,500Join representatives from 45 countries spanning the globe.45

Visit 200 companies exhibiting today’s

foremost technologies, products and services.

200Connect with 5,000

registrants— technical leaders and business

executives.5,000

Source from a global marketplace converging

from 30+ countries. 30

CHALLENGING PIPELINE PERFORMANCE

& EXPOSITIONINTERNATIONAL PIPELINE CONFERENCE

@IPC_Calgary @petroleumshow

Sample

copy

- not

for di

stribu

tion


THE OCCURRENCE OF low-frequency electric-resistance-welded (LF-ERW) or electric-flash-welded (EFW) linepipe imperfections has been the root cause of many integrity-management initiatives to

minimize and mitigate the risk of pipeline failure across the oil and gas pipeline industry. Since their first appearance in the 1920s, defects in or near the LF-ERW and EFW seam repeatedly lead to either hydrostatic test or in-service failures. Where in the past in-line inspection (ILI) technologies might have experienced limitations in addressing vintage ERW linepipe defects, modern smart ILI technologies show enhanced capabilities. High-resolution electro-magnetic transducer (EMAT) and circumferential magnetic-flux leakage (CMFL) ILI technologies have advanced in the recent years enabling more challenging inspections. This paper summarizes the inspection results of defects detected and reported by EMAT and CMFL in a 22-in ERW linepipe. Correlation of ILI and manual NDE data enables evaluation of current ILI capabilities and improvement of current defect-assessment methods.

This paper was presented at the Pipeline Pigging and Integrity Management Conference held in Houston, TX, USA, in February, 2014.

*Corresponding author’s contact details:tel: +49 5919 1365 16email: [email protected]

by Richard Kania1, Ralf Weber2, and Stefan Klein*3

1 TransCanada Pipelines, Calgary, Canada2 ILI Consulting, Karlsruhe, Germany3 Rosen Technology & Research Centre, Lingen, Germany

On the assessment of low-frequency electric-resistance-welded linepipe defects by EMAT and CMFL in-line inspection

VINTAGE PRE-1970 low-frequency electric-resistance-welded (LF-ERW) and electric-flash welded (EFW)

linepipe are known to show a variety of imperfections along the bonding line inherent to the manufacturing process and pipe material [1-3]. For LF-ERW pipe, hot-rolled and coiled strip steel was usually cold-formed into a round can and passed through a welding device. An alternating current – typically of 120 Hz – was applied to heat the edges after which they were mechanically forced together to achieve bonding without melting the steel. For direct-current-welded ERW (DC-ERW) only the nature of the welding current was different [2]. For EFW pipe a direct electrical current was also applied, but the entire weld was formed in one stage. EFW pipe is often referred to as A.O.Smith pipe, since this process was only applied by this one manufacturer

in North America. ERW pipe can be identified by its typical flat-topped fin of extra metal visible at the OD and ID of the seam which is trimmed away, leaving an almost smooth surface for LF-ERW pipe. Since the introduction of micro-alloying processes to modern linepipe steel production, the variance of steel properties significantly improved in comparison to vintage low-carbon linepipe steel, and inclusions are unlikely to occur. Today, steel is designed for specific application purposes by careful selection of alloys and the linepipe manufacturing processes outlined above are either replaced, for example, by applying an alternating current at high-frequencies of more than 400 Hz for high-frequency ERW, or suspended, as for EFW.

A variety of anomalies present in pre-1970 ERW pipe have been identified as the root cause for in-service or hydrostatic pressure test failures in the past [2]. Among these, lack of fusion (LOF) or cold welds, J-shaped hook cracks, and selective seam corrosion (SSC) are likely to be the most frequent and prominent types of anomaly. In contrast to other linepipe defects, seam-

Sample

copy

- not

for di

stribu

tion


Transmission and reflection signals are captured by means of two separate receiver sensors within the EMAT sensor arrangement. Due to the limited propagation distance of the waves between the measuring elements, this design ensures high signal-to-noise ratios as a basis both for sensitivity and also accurate determination of the position and dimensions of features.

Waves which propagate from transmitter to receiver through the pipe wall without hindrance are used to assess the external pipe coating. The ultrasound, attenuated by intact coatings and lower signal amplitude, is captured by the receiver; in the case of coating disbondment or coating holidays, attenuation is reduced. Pipe anomalies situated in the sensitive EMAT measurement area reflect part of the ultrasonic wave, and information on the frequency, time of flight, and modes is used for discrimination of cracks and volumetric features and for determining the length and depth of features. The quality of the EMAT signals obtained from crack and coating measurements is supported by the quantification of possible lift-off effects and magnetization measurements. Further details on the concept of the high-resolution EMAT technology are described in Refs 6-8.

The magnetic-flux-leakage method has been established in the ILI industry for several decades in axial (AMFL) and, more recently, circumferential (CMFL) directions. A sketch of the CMFL arrangement is shown in Fig.2: CMFL has been introduced to detect and characterize axially oriented anomalies. The CMFL application is more complex, and a magnetization excitation of at least 10 kA/m must be established around the entire circumference. The need for an excitation level to saturate the pipeline has been addressed in several publications [9-11]: an appropriate magnetization level significantly reduces the spurious effects caused by, for example, pipe stresses or tool-velocity effects. Furthermore, the signal-to-noise ratio for signal amplitudes is improved and was confirmed by measurements obtained at an in-house test facility.

Validation and inspection programme

In a recent joint collaboration project, the capabilities of high-resolution EMAT and CMFL ILI technologies for assessing the integrity management of ERW pipe were investigated. Validation of the EMAT technology prior to inspection of the pipeline was undertaken to determine the EMAT detection and sizing capabilities of targeted defects located in the ERW seam. After validation of the ILI capabilities, the pipeline has been inspected and cut-outs performed for NDE and further analysis. The present contribution summarizes the results achieved so far, with special emphasis on the correlation of NDE and ILI results to demonstrate state-of-the-art ILI capabilities.

weld anomalies are located in an area where the pipe geometry and material can be significantly different from the properties in, for example, the pipe body. Local properties of the material are mostly unavailable, preventing formal integrity assessment of the asset and prediction of critical anomaly dimensions [4]. In the past linepipe integrity has therefore been mostly secured by hydrostatic pressure testing at intervals of the order of (typically) several years. By development of technologies for non-destructive examination (NDE) and, more recently, smart in-line inspection (ILI) tools, technologies have become available that can successfully acquire rich information to examine complex seam-weld irregularities. Nowadays, and in the absence of either material properties and seam-weld anomaly-assessment criteria, hydrostatic testing, direct assessment, and ILI data are incorporated into pipeline integrity-management programmes. The inevitable procedure and industry practice is to repair all detected crack-like seam-weld anomalies, which has in all likelihood resulted in the unnecessary repair of numerous anomalies [4].

EMAT and CMFL ILI technology

Figure 1 shows a schematic representation of the EMAT principle. A single EMAT probe inspects a small, well-defined, area between transducer and receiver.

Fig.1. Sensor arrangement for the high-resolution EMAT. An ultrasonic shear wave is generated in the pipe wall, travelling from transducer to receiver. An obstacle, such as a crack situated in the sensitive area of the EMAT sensor, results in reflection of the ultrasound.

Fig.2. Magnetic flux in the circumferential direction for circumferential magnetic-flux-leakage (CMFL) technology.

Sample

copy

- not

for di

stribu

tion


Results and discussion

The EMAT and CMFL ILI data were analysed and reported as shown in Table 2. The reported features were prioritized by reported lengths and depths for field verification. After initial in-the-ditch confirmation of the ILI-reported features, corresponding joints were

Prior to inspection, an original section of the 22-in pipe was removed from the pipeline and pull-tested. The pipe was specified as DC-ERW welded X-52 steel of 0.281-in (7.2-mm) wall thickness manufactured at the Youngstown Sheet & Tube Co (YS&T) pipe mill. Pre-1970 YS&T ERW pipe has been referenced before [2, 5].

The weld surface of the spools removed from the pipeline was prepared and examined by wet black-and-white magnetic-particle inspection (MPI). For reference and validation, 12 electrical-discharge-machined (EDM) notches of rectangular cross-section were subsequently machined into the bonding line of the ERW seam weld, and Table 1 shows the corresponding notch dimensions. The pull-through tests were carried out using the EMAT tool to validate detection capabilities and signal characteristics of the reference reflectors. Figure 3 shows the corresponding normalized EMAT frequency-reflection signal of the notches of one of the pull-through tests: all 12 notches were successfully detected. The weld itself shows a reflection within the frequency range of higher-order shear-wave modes (SHn). Clear discrimination of the surface-breaking artificial defects from the weld and irregularities was enabled by the high signal-to-noise (S/N) ratio for the zero-order mode (SH0).

After validation of the EMAT ILI tool, the pipeline was inspected by EMAT and CMFL tools and additional joints were cut out based on combined EMAT and CMFL ILI data analyses.

Notch ID (#) Length (mm) Depth (%)

1 50 100

2 40 100

3 25 100

4 20 100

5 15 100

6 10 100

7 25 40

8 50 40

9 50 60

10 50 60

11 25 60

12 25 60

Table 1. Dimensions of EDM notches in the ERW bonding line used as reference reflectors for ILI tool-validation purposes. Wall thickness of the pipe spool is 0.281 in.

Fig.3. EMAT frequency-reflection signal of the reference EDM notches machined into the bonding line of the ERW weld. Notch IDs and dimensions are given in Table 1.

Sample

copy

- not

for di

stribu

tion


in this case, inadequate. The reporting and assessment strategy needs to be customized to reflect the true severity and enable advanced integrity management of linear anomalies.

In order to assess the capability of the ILI technologies used with regard to natural defects, and accurately to measure a targeted defect profile, the ILI raw data need to be considered and evaluated in detail. Thus, captured ILI data were correlated to PAUT profiles of the seam welds under consideration. Figures 4 to 7 show detailed correlations of the different data: reported indications are indicated by vertical lines, and the figures are organized by showing the recorded EMAT frequency of the reflection signal of an individual sensor capturing the seam weld in a B-scan representation on top. The colour-coded B-scan is used during the analysis of the data. The manual PAUT depth profile is shown at an absolute scale. As mentioned above, in the case where a surface-breaking defect was visible, but no depth could be determined, a depth of 1 mm was recorded. In the case where there were coupling issues for the ultrasound, no depth was recorded. The EMAT frequency response shows the normalized reflected EMAT frequency signal. The CMFL magnetization shows the pipe magnetization of single sensor with respect to an offset correction to visualize the data. The sensor data displayed were manually selected by the highest S/N ratio, and all ILI datasets were used during the data analysis of the tool runs.

The defect shown by Fig.4 has been identified as a LOF at the bonding line of the seam. The length of the defect of 15 mm (0.590 in) is significantly smaller than the wall thickness of the pipe 7.2 mm (0.281 in). For such penetrators, as for EDM notches, the lower frequency reflection clearly indicates the reported defect. The CMFL signal indicates the defect as well. An increase in the recorded magnetization is detected. MPI and PAUT confirmed the presence of

replaced. The cut-outs were cleaned and the coating was removed. The spool surface was then sandblasted and prepared for wet black-and-white MPI. As the industry standard practice, manual phased-array ultrasonic testing (PAUT) NDE examination of the seams was selected as a benchmark for the ILI results. Unless otherwise stated, the NDE reported lengths refer to the lengths as being measured on the OD of the pipe surface, and reported depths refer to the manual PAUT result.

The entire ERW seams were examined by MPI to detect and map all surface-breaking defects, and the seams were then tested using PAUT by applying a grid of equally spaced measurements, separated by 10 mm. The individual measurements were recorded and traced in order to achieve a depth profile of seam anomalies, such as LOF. Using PAUT, the minimum detectable and reliably measureable defect depth was given as 1 mm according to the contracted NDE vendor. In the case where a surface-breaking defect was detected at a certain location, but the depth could not be determined at the same location, the minimum detectable defect depth was recorded. This procedure is justified, since grinding of shallow or superficial features confirmed the inaccuracy of the PAUT for depths of less than 1 mm. In case no measurement was possible due, for example, to inefficient coupling of the ultrasound into the pipe wall, no depth was recorded.

In general, ILI results for volumetric defects are reported by lengths, widths, and depths in order accurately to describe the geometry of a feature. For linear non-volumetric or planar anomalies, only the lengths and depths are of interest, since the anomaly width is of the order of microns to one millimetre. However, the complexity of depth profiles of continuous planar anomalies that extend over greater distances up to joint lengths cannot be successfully described by two individual and uncoupled measurements. For example, the assumption of elliptically shaped cross-sections is,

Joint IDILI PAUT NDE

Type Max. depth (mm)

Depth (%)

Length (mm)

Type Max. depth (mm)

Depth (%)

Length (mm)

1 LSWA 3.8 70 LOF 4.1 15

2 LSWA 3.9 62 LOF 3.8 40

3 LSWA 3.2 1512 Hook crack 2.9 1710

3 LSWA 2.7 108 LOF 2.7 62

4 LSWA 3.6 12828 Hook crack 3.9 12060*

LSWA: linear seam weld anomaly* Entire cut-out length

Table 2. Combined EMAT and CMFL ILI analyses results vs manual PAUT NDE for four reported spools. Corresponding data are shown in Figs 4 to 7.

Sample

copy

- not

for di

stribu

tion


distribution and a significant overlap of individual measurements in the axial direction, the extension of the defect is enlarged. The same phenomenon can be recognized for EDM notches. However, this adds additional conservatism to the assessment of the defect.

Figure 5 shows a similar LOF defect in the bonding line of 40 mm (1.575 in) length. Both EMAT and CMFL detected the LOF, and the reported depth of 3.8 mm (0.149 in) matches the PAUT depth. The length is overestimated to a lesser extent, and the defect was detected by MPI on the OD of the pipe.

the penetrator. The length of 15 mm (0.590 in) is well below the specified minimum defect length for the EMAT and of the order of the shortest EDM notch tested upfront. The reported ILI depth matches the NDE depth and is within the specified accuracy. Although no destructive-testing result available at this point, it is assumed, that the defect-depth profile is similar to that of a machined EDM notch. The ILI overestimates the length of the defects with respect to the reference MPI length, and this can be explained by the extent of the measurement area of the EMAT sensor. Assuming a Gaussian EMAT sensor-sensitivity

Fig.4. ILI and manual PAUT NDE data for spool no.1. High correlation of the PAUT depth profile and EMAT frequency response is visible for defects that exceed 1 mm depth. Further description is given in the text.

Fig.5. ILI and manual PAUT NDE data for spool no.2. High correlation of the PAUT depth profile and EMAT frequency response is visible for defects that exceed 1 mm depth. Further description is given in the text. Sam

ple co

py - n

ot for

distr

ibutio

n


low S/N ratio, and consequently the POD is reduced for such defects. Both defects can also be detected in the CMFL data, although the increase in magnetization is different, which is supposed to originate from a wider opening of the downstream defect. Consequently the volume magnetized decreases and the proportional magnetization increases.

Figure 7 shows the most complex defect. MPI examination of the seam indicated a LOF defect at the bonding line for the entire ERW spool. As discussed above, although the maximum ILI depth and length, and NDE

After the PAUT measurement indicated the presence of the defect, the surface-breaking LOF was detected on the ID of the pipe.

Figure 6 shows a non-continuous LOF defect along the ERW seam, where the anomaly on the downstream end of the spool has been confirmed as being internally surface-breaking. The EMAT frequency response shows good correlation to the PAUT depth profile. For sections of deeper depth, the processed EMAT signal shows a higher signal, and vice versa. Surface-breaking defects below 1 mm (0.039 in) of depth appear at a

Fig.6. ILI and manual PAUT NDE data for spool no.3. High correlation of the PAUT depth profile and EMAT frequency response is visible for defects that exceed 1 mm depth. Further description is given in the text.

Fig.7. ILI and manual PAUT NDE data for spool no.4. High correlation of the PAUT depth profile and EMAT frequency response is visible for defects that exceed 1 mm depth. Further description is given in the text.Sam

ple co

py - n

ot for

distr

ibutio

n


the real depths and lengths. Present detection and sizing capabilities, and their limitations, can then be ultimately assessed.

Acknowledgements

The authors gratefully acknowledge the support of TransCanada Pipelines Ltd, and also the assessment, verification, and correlation work done in-field and in shops by Spencer Blomquist and his team from Applus RTD in Edmonton, Canada.

References

1. J.F.Kiefner et al., 1996. History of line pipe manufacturing in North America, ASME.

2. Idem, 2012. ERW and flash weld seam failures. Kiefner and Associates, Inc., Final Report to Battelle, 2012 (http://primis.phmsa.dot.gov/matrix/PrjHome.rdm?prj=390).

3. M.Mohitpour et al., 2010. Pipeline integrity assurance – a practical spproach. 1st Edn.

4. B.O.Hart et al., 2004. Early generation seam welds. Final Report R3017-01R, PRCI.

5. B.N.Leis, 2013. Compare-contrast analysis of inspection data and failure predictions versus burst-test outcomes for ERW-seam defects. Final Interim Report to Battelle, (http://primis.phmsa.dot.gov/matrix/PrjHome.rdm?prj=390).

6. J.D.Achenbach, 1975. Wave propagation in elastic solids. In: Applied mathematics and mechanics, 16, Elsevier.

7. T.Beuker et al., 2004. SCC detection and coating disbondment detection improvements using the high resolution EMAT ILI-technology. Proc. IPC, pp IPC04-0697.

8. M.Klann et al., 2006. Pipeline inspection with the high resolution EMAT-ILI tool: Report on full-scale testing and field trials. Proc.IPC, pp IPC06-10156.

9. T.Beuker, B.Brown, et al., 1999. Advanced magnetic flux leakage signal analysis for detection and sizing of pipeline corrosion: field evaluation program. GRI report GRI-00/0109.

10. A.E.Crouch, T.Beuker, B.Brown, 20003. Flux leakage signals from corrosion defects in pipeline subjected to bending loads. Pipeline Pigging & Integrity Technology, 3rd Edn, pp 267-282.

11. H.J.M.Jansen et al., 1994, Magnetization as a key parameter of magnetic flux leakage pigs for pipeline inspection. Insight, 36, pp 672-677.

12. R.Kania et al., 2012. Validation of EMAT technology for gas pipeline crack inspection. Proc. IPC, pp IPC12-90240.

results, are within acceptable tolerances, the complex depth profile cannot be captured by two uncoupled measurements. As for the previously discussed defect, both the PAUT and ILI data show good correlation. The measured PAUT depth varies for different location along the seam, and this bears rich information when compared to the EMAT frequency response. The EMAT signal is increased at locations where the defect depth exceeds 1 mm (0.039 in): however, it does not appear to be linearly dependent on the anomaly depth. A more-complex function is required to determine the absolute depth at a certain location, and this is in accordance with the experience obtained from previously reported SCC situated in the pipe body of similar pipe [12]. Furthermore, it can be seen, that a minimum effective cross section is required to allow the defect to be detectable. This is in accordance with the defect and data shown in Fig.4, and implies an increasing POD for increasing defect depths of constant length. The CMFL signal does not resolve all indications shown in the PAUT depth profile, and the shortest indications are partially not well resolved. Variation in the opening of the surface-breaking planar defect and corrosion at or in the seam-weld area might effectively influence the CMFL signal.

Conclusion and outlook

In a joint effort of the pipeline operator and ILI vendor, EMAT and CMFL ILI technologies have been used to detect, identify, and size anomalies of vintage pre-1970 ERW linepipe. Manual PAUT was used as a benchmark with regard to detection and sizing capabilities, and correlation of the PAUT depth profiles of LOF defects in the weld and ILI data has been performed. From the different dimensions in terms of lengths and depths of the investigated defects, the capabilities have been demonstrated. The applied EMAT and CMFL show high sensitivity to defects in the ERW weld. The EMAT successfully resolved different defects ranging from exemplary penetrators to continuous seam-weld anomalies of varying depths. The same holds for the CMFL, where the signal of short penetrators is supposed to be effectively influenced by volumetric defects at, or in, the seam weld area. Consequently the signals are superimposed, as demonstrated for one example.

The present EMAT and CMFL IL technologies’ capabilities have been successfully applied to detect anomalies of the ERW seam-weld. To assess and confirm the presented results, additional joints will be tested. Consecutive destructive testing of cut-outs will clarify the nature of the defects and help to determine

Sample

copy

- not

for di

stribu

tion

.

www.clarion.org

ORGANIZED BY

CALL FOR PAPERS NOW OPEN

21 – 24 October 2014 Berlin, Germany

ABSTRACT SUBMISSIONProposals should include an abstract of no more than 200 words,

accompanied by the author’s complete contact information and affiliation, and should be sent to:

John Tiratsoo, Director, Tiratsoo Technical e-mail: [email protected]

SCHEDULEAbstracts deadline: 16 June, 2014 | Final papers: 22 September, 2014

FPP_2014_FP_CallforPapers_A4.indd 1 4/04/2014 9:35 am

Sample

copy

- not

for di

stribu

tion


STRAIN-BASED DESIGN and assessment (SBDA) focuses on potential failures driven by high longitudinal strains. Pipeline failures driven by longitudinal stresses or strains are relatively rare events in comparison

to failures driven by hoop stresses. Longitudinal strains are often associated with ground movement or other unusual upsetting events. SBDA is performed by comparing strain demand with strain capacity. Strain demand may be obtained from direct measurement or pipe-soil interaction models. Strain capacity is typically estimated using suitable models supported by experimental test data.

There are gaps between the present approaches to SBDA and field conditions under which SBDA is applied. For instance, linepipes are delivered with a range of tensile properties. Welds are produced by a variety of processes with a range of tensile and toughness properties. Pipe dimensions, mechanical properties, and soil conditions affect both strain demand and strain capacity.

The overall process of SBDA is introduced first. It can be seen that conditions assumed in models can be quite different from those in actual field applications. For instance, there can be considerable variations in the tensile strength of linepipes. The potential impact of strength variations on the measured/reported strain demand and strain capacity is described. The overall approaches to SBDA with appropriate consideration of actual field conditions are suggested. Some unresolved issues related to SBDA are described, particularly in the context of characteristics of modern linepipes.

This paper was presented at the Pipeline Technology Conference held in Ostend, Belgium, in October, 2013.


by Dr Yong-Yi Wang* and Dr Ming LiuCenter for Reliable Energy Systems, Dublin, OH, USA

Real world considerations for strain-based design and assessment

THIS IS THE THIRD PAPER in a three-paper series on the strain-based design and assessment (SBDA).

The first paper focuses on tensile strain design [1], and introduced the background of SBDA. Three key elements of tensile strain design, linepipe specifications, girth-weld qualification, and tensile strain models are also covered in the paper. The second paper focuses on compressive strain design [2], and a set of refined compressive-strain models is presented. These models are evaluated against experimental data. The current paper builds on the foundations established in the first two papers and explores practical issues in the application of SBDA.

The overall process of strain-based design

The key components of SBDA and their relationship are schematically shown in Fig.1. The strain demand (alternatively termed ‘applied strain’) may be estimated

from pipe operational and environmental conditions, such as geotechnical features and soil conditions. The compressive and tensile strain capacities may be determined from their respective models supported by experimental test data. The strain demand and the strain capacity are then compared to determine if a design has a sufficient margin of safety against conditions that cause high longitudinal strains. The characteristics and variations of pipe and weld mechanical properties, geometric dimensions, and imperfections, flaw dimensions, etc., can impact all components in the SBDA process.

Variation of tensile properties and its implications

Observed tensile property variations

In a US DOT PHMSA and PRCI co-funded project aimed at developing tensile-strain models, extensive tensile-property testing was conducted for the linepipes that were used to fabricate full-scale test specimens [3]. The tensile properties of an X-65 pipe with 12.75-in OD and 0.5-in WT are shown in Figs 2 to 5. The tensile tests were performed using machined round-bar

Sample

copy

- not

for di

stribu

tion


stress-strain curves is representative of the variations of the actual material properties. Even with that understanding, the strength variation is 4-5 ksi (28-34 MPa) in the longitudinal direction and 8-9 ksi (55-62 MPa) in the hoop direction.

Impact of tensile property variation on the measured tensile strain capacity

Flaw behaviour and remote strain measure

To examine the relationship between the girth-weld flaw behaviour and the remote strain, a pipe section loaded in longitudinal tension is shown in Fig.6. The pipe section has a girth weld with a planar flaw: the cross-sectional plane where the flaw is located is termed the ‘flawed plane’. The regions where the remote strain is measured are referred to as ‘uniform-strain zones’. The total longitudinal load acting on the flawed plane and the uniform-strain zones is the same. Assuming the cross-sectional areas are the same at these locations, the nominal stresses in those regions are the same. The loads in the remote regions are ‘transmitted to the flawed plane through the equivalence of the nominal stress.

If the material in the two uniform-strain zones has slightly different stress-strain behaviour, the strains in those zones have to be different in order to keep the nominal stress the same in both regions. Figure 7 shows three slightly different stress-strain curves: the baseline stress-strain curve is taken from the full-scale test 1.18 of a 12-in pipe [3], and two additional stress-strain curves are ‘created’ by adding or subtracting the strength of the baseline curve by 0.5 ksi (3.5 MPa) at strains greater than 0.5%.

specimens from four pipe joints of the same heat. For the longitudinal properties shown in Figs 2 and 3, the strength variation on the plastic portion of the stress-strain curves (the parts of the curves to the right of the knee of the stress-strain curves) is in the range of 4-5 ksi (28-34 MPa). However, due to the difference in the elastic slopes of the curves and non-linearity before the knee of the stress-strain curves, the reported yield strength at the customary total strain of 0.5% varies from 74.6 ksi to 82.2 ksi, resulting in a maximum difference of 7.6 ksi (52 MPa).

For the hoop properties shown in Figs 4 and 5, the strength variation on the plastic portion of the stress-strain curves is in the range of 8-9 ksi (55-62 MPa). However, due to the difference in elastic slopes of the curves and non-linearity before the knee of the stress-strain curves, the reported yield strength at the customary total strain of 0.5% varies from 58.3 ksi to 79.5 ksi, resulting in a maximum difference of 21.2 ksi (146 MPa). The reported yield-strength values are particularly sensitive to the accuracy of the strain measure: for example, the accuracy of elastic slope and any deviation from linearity in the stress-strain response before the knee of the stress-strain curves. The large reported yield-strength variation in the hoop properties is not considered representative of normal material-property variations. However, the plastic portion of the stress-strain curves is insensitive to the accuracy of the strain measure of the test specimens. The strength values in this part of the stress-strain curves are a direct reflection of the measured load during testing. Assuming the load cells are in calibration, the strength variation in the top portion of the

Fig.1. Key components of SBDA and their relationship.

Sample

copy

- not

for di

stribu

tion


Observed variation in strains

A large number of curved-wide-plate (CWP) tests were conducted under a US DOT PHMSA and PRCI sponsored project [4]; the consequent pipe and girth-weld mechanical properties have been extensively covered by Gianetto et al. [5]. The non-uniformity of the strain distribution in a CWP specimen is shown in Fig.9 in a photoelastic image [6]. The corresponding remote strains in the base pipe sections (above and below the girth weld) are given in Fig.10.

The range of the strain values corresponding to the same stress value due to the slight strength difference is shown in Fig.8. The baseline curve establishes a one-to-one correlation; the other two curves provide the upper- and lower-bound strain values corresponding to the same stress produced by the baseline curve at strains greater than 1.0%. The result shows that when a flaw is subjected to a stress level corresponding to the stress at 4.0% strain of the baseline stress-strain curve, the strains in the uniform strain zones can vary from 2.5% to 8% (material’s uniform strain) if the stress-strain curves vary by ±0.5 ksi from the baseline stress-strain curve.

Fig.2. Longitudinal stress-strain curves of four joints of “high Y/T” X-65 linepipe of the same heat,

Fig.4. Hoop tensile stress-strain curves of four joints of “high Y/T” linepipe of the same heat.

Fig.6. Schematic illustration of the remote regions where the strains are measured in test specimens and the flawed plane where the flaw failure events are initiated.

Fig.3. Top portion of Fig.2.

Fig.5. Top portion of Fig.4.

Sample

copy

- not

for di

stribu

tion


It is evident that the strains in the top and bottom pipe sections start to deviate upon yielding, and this difference is attributable to small variations of the linepipe properties. Using the load, cross-sectional area, and the displacements measured by the LVDTs located at the top and bottom part of the CWP, the nominal engineering stress vs engineering strain curves may be constructed, as shown in Fig.11. The strength difference between the top and bottom pipe sections is 10 MPa (1.45 ksi) and 20 MPa (2.9 ksi) at 0.5% and 2.0% strains, respectively. At the termination of the test, the nominal stress is at 823 MPa (119.4 ksi) on both sides, but the remote strains in the top and bottom pipe sections differ by a factor of approximately 1.5 (3.34% vs. 2.04%), see Fig.12. The range of strength variation is small and quite reasonable. Due to the rather flat stress-strain relations in the plastic range of the stress-strain curves, the small difference in the stress-strain curves results in a large difference in strains while both sides are at the same applied stress level. Tensile strain capacity – model prediction vs experimental

measurement

X-65 pipes and girth welds

In the full-scale tests of 12-in OD X-65 pipe and girth welds, the measured tensile-strain capacity (TSC) from nominally identical pipe and girth welds having identical flaw sizes and locations was shown to differ by as much as a factor of two or more [3]. The experimentally measured and the predicted TSCs of the high Y/T pipes are compared in Fig.13 for all 19 tests: the predicted TSCs using the median values of the pipe and weld tensile properties are given as the x values, while the corresponding measured TSCs are given as the y values. Three lines are also given in the plot: the middle line represents the one-to-one, i.e., perfect agreement between the measured and predicted TSCs using median tensile property, while the upper line represents the remote strain value needed to produce the same stress level as the baseline curve when the strength is 1 ksi (6.9 MPa) weaker than the baseline curve. By the same process, the lower line represents the remote strain when the strength is 1 ksi (6.9 MPa) stronger than the baseline curve. The figure shows that the middle line of the prediction goes through the centre of the cluster of the test data. The upper- and lower-bound curves due to a ±1 ksi strength variation follow the upper and lower bounds of the experimental data. In other words, the entire variation of the test data can be explained by a ±1 ksi strength variation in the tensile properties. This variation is considerably less than the measured variation as shown in Figs 2 to 5.

X-80 and X-100 pipes and girth welds

Figure 14 compares the predicted TSC with the measured TSC from CWP specimens [7]. The tensile

Fig.7. Stress-strain curves of slightly different strength levels.

Fig.8. Strain values corresponding to the three stress-strain curves at the same stress level as a function of the strain of the baseline stress-strain curve.

Fig.9. Photoelastic image showing the non-uniform deformation on both sides of a girth weld.

Sample

copy

- not

for di

stribu

tion


bounding lines represent the possible strain range from a strength variation of 3 ksi (21 MPa) which is consistent with observed strength variation shown in Figs11 and 12.

Strategy for SBDA

Girth weld response by category

Wang et al. have proposed the concept of four categories of response when a girth weld with planar flaws is subjected to longitudinal tensile strain, as shown in Fig.16 [6]. If a girth weld flaw is well ‘protected’ by a wide and smooth weld cap with sufficient weld strength, the flaw opening would eventually saturate before reaching a critical value, as represented by the Category I behaviour. Conversely, if the weld strength is low and/or a high level of misalignment is not protected by a wide weld cap, a gross strain concentration would occur in the weld, resulting in a large flaw opening immediately after the material in the flawed plane becomes plastic. This would lead to low strain capacity

strain capacity (TSC) values of X-80 and X-100 girth welds are represented by the circles and triangles, respectively. The flaw locations are represented by either the filled or open symbols. There is generally a good agreement between the model prediction and experimentally measured TSC. X-100 pipes and girth welds

The experimentally measured TSC of 30 CWP specimens are compared with the Level 2 prediction of the PRCI-CRES models [4] in Fig.15. The middle point in each bar represents the averaged strain value between the top and bottom parts of the CWP specimens, while the ends of the bars are the strain values in either the top or bottom part of the same specimen. The length of the bar is the spread of the strain values between the top and bottom parts of the specimen. As the averaged strain goes up, the spread tends to become large because the stress-strain curves become flattened in comparison to the stress-strain response at lower strain values. The dashed

Fig.10. Strains above and below the girth weld as a function of crack-mouth-opening displacement (CMOD).

Fig.12. Top portion of the stress-strain curves of Fig.11, showing the difference in measured strain between the top and bottom parts of the CWP.

Fig.11. Engineering stress-strain curves constructed from the LVDTs located on the base pipe at the top and bottom parts of the CWP.

Fig.13. Comparison of measured and predicted TSCs of the high Y/T pipes. The predicted TSCs are from Level 2 PRCI-CRES models. The possible ranges of the TSCs from the strength variation of ±1 ksi are also shown.

Sample

copy

- not

for di

stribu

tion


6. Develop and implement strain-demand monitoring systems. When needed, verify the reliability and accuracy of the monitoring systems.

7. Develop and implement continuous evaluation and mitigation plans if the strain demand and material properties are time-dependent.

Tailored SBDA procedures may be further refined depending on conditions specific to the project or pipeline of interest.

Some unresolved issues in SBDA

Consistency of strain demand and strain capacity

In a typical SBDA scenario, the strain demand is either estimated using pipe-and-soil-interaction models or IMUs (inertial-mapping units). The strain output from either the models or IMUs can be of different gauge length from the gauge length used to report and quantify strain capacity. Consequently the same value of strain demand and strain capacity may not represent the same state of the materials of interest. Other sources of inconsistency may include the fact that the material models used in the strain-demand calculation may be different from those used in the strain-capacity calculation.

Assessment of anomalies in the presence of high longitudinal strains

Most pipelines are subjected to corrosion and mechanical damage in their lifetime. The current methodologies for assessing and mitigating those anomalies were developed under the premise that hoop stress is higher than longitudinal stress, and thus hoop stress is the primary driver for potential failures. Under conditions typical of SBDA, the stress and strain in the longitudinal direction are higher than those in the hoop direction, and the validity of the current assessment methods is not known under such conditions.

as represented by Category IV behaviour. Categories II and III behaviours are bounded by those two extreme cases: flaws would open with increasing applied stress, initially forming blunted flaw tips and eventually starting to grow macroscopically.

Principal considerations for SBDA

The overall SBDA involves the following critical considerations.

1. Determine the nature of the strain demand. Features to consider may include whether the event is one-time, cyclic, and/or time-dependent.

2. Determine a set of target strain-capacity levels by introducing appropriate safety factors to the estimated strain demand.

3. Conduct a preliminary assessment of the possible responses for all postulated failure events. This may involve estimating strain capacity using models and a range of possible material, dimensional, and operational parameters. What-if scenarios may be exercised to determine the feasibility of achieving certain material-property requirements. For a small project, in which the material-property variations can be well defined, the strategy may be achieving Category I behaviour. Possible conditions leading to Category IV behaviour should be precluded.

4. Develop material specifications and qualification processes based on the requirements of strain capacity. The material-qualification processes must be accompanied by well-defined test procedures that cover all aspects of testing, from specimen position in pipes/welds to post-test data analysis and reporting.

5. Develop and implement field-quality assurance and monitoring process.

Fig.14. Comparison of experimentally measured TSC with the Level 2 prediction of the PRCI-CRES models. CTODA is computed from upper-shelf Charpy energy.

Fig.15. Comparison of measured TSC and the TSC predicted by Level 2 PRCI-CRES models.

Sample

copy

- not

for di

stribu

tion


Nanney of PHMSA, David Horsley (previously with TransCanada and BP), and Maria Quintana of Lincoln Electric, are gratefully acknowledged.

References

1. Y.-Y.Wang and M.Liu, 2013. Status and applications of tensile strain capacity models. Proc. 6th Pipeline Technology Conference, Ostend, Belgium, October 7-9.

2. M.Liu, Y.-Y.Wang, F.Zhang, X.Wu, and S.Nanney, 2013. Refined compressive strain capacity models. Ibid.

3. Y.-Y.Wang, M.Liu, X.Long, M.Stephens, R.Petersen, and R.Gordon, 2011. Validation and documentation of tensile strain limit design models for pipelines. PRCI Project ABD-1, US DOT Agreement DTPH56-06-T000014, Final report, 2 August http://primis.phmsa.dot.gov/matrix/PrjHome.rdm?prj=200.

4. Y.-Y.Wang, H.Zhou, M.Liu, W.Tyson, J.Gianetto, T.Weeks, M.Richards, J.McColskey, M.Quintana, and V.B.Rajan, 2012. Weld design, testing, and assessment procedures for high-strength pipelines. Summary report 277-S-01, US DOT Agreement DTPH56-07-T-000005, January.

5. J.Gianetto, W.Tyson, Y.-Y.Wang, J.Bowker, D.Park, and G.Shen, 2010. Properties and microstructure of weld metal and HAZ regions in X100 single and dual torch girth welds. Proc. 8th International Pipeline Conference, paper IPC2010-31411, 27 September – 1 October, Calgary.

6. Y.-Y.Wang, M.Liu, D.McColskey, T.Weeks, and D.Horsley, 2010. Broad perspectives of girth weld tensile strain response. Ibid., paper IPC2010-31369.

7. Y.-Y.Wang, F.Zhang, M.Liu, W.Cho, and D.Seo, 2012. Tensile strain capacity of X80 and X100 welds. Proc. 31st International Conference on Ocean, Offshore and Arctic Engineering, OMAE 2012-84240, 1-6 July, Rio de Janeiro, Brazil.

Post-buckling response of wrinkles

Wrinkles are typically not considered an immediate integrity threat. However, with the continued decrease of a material’s effective strain-hardening capacity, as represented by very flat stress-strain curves in the plastic range of the material response, and the possibility of high levels of HAZ softening in the vicinity of seam welds of microalloyed linepipe steels, hoop strains generated by wrinkles may be sufficient to cause thinning or even rupture of seam welds. Such a possibility should be considered even though seam integrity is typically not considered a part of SBDA.

Concluding remarks

There has been considerable progress in the development of SBDA technology in the last decade. Strain-capacity models, both tensile and compressive, have become more sophisticated and capable of correlating a multitude of parameters affecting strain capacity.

Against such advances, there are still considerable gaps between conditions assumed in the models and real-world field conditions. For instance, pipe-strength variations in a single joint or among multiple joints are a reality which is not sufficiently addressed in current models. Understanding the impact of strength variation is critical for at least two reasons. First, modern linepipe steels can have very low strain hardening; secondly, test procedures for tensile properties are not sufficiently detailed and specific to preclude the influence of variations in test procedures. These factors together can lead to strain concentrations in pipe joints of lower strength, and those joints may not have sufficient strain-hardening capacity to compensate for the lower strength. Consequently, a small number of joints can experience much higher strains than the averaged strain over a long segment.

From the viewpoint of material properties, having good strain hardening is an absolute necessity for SBDA. Similarly, having an appropriate weld profile to preclude gross strain concentration and having upper-shelf toughness properties are also necessary.

The premises upon which current assessment models are built may not be applicable any more under typical SBDA conditions, and many basic assumptions about pipeline integrity and integrity assessment need to be reassessed.

Acknowledgment

The work reported here represents multi-year efforts supported by US DOT PHMSA (US DOT Contract No. DTPH56-06-T000014 and DTPH56-10-T000016), PRCI, and CRES. The guidance and support of numerous individuals involved in the efforts, particularly Steve

Fig.16. Four categories of response expressed in crack-driving force (CFD) vs remote-strain relations.

Sample

copy

- not

for di

stribu

tion

www.pipeinst.org

• Recognizing your skills and status

• Providing a professional network

• Promoting pipeline engineering

The global organization for oil and gas

pipeline engineers

Sign up today!

THEPROFESSIONALINSTITUTE OF

PIPEINE ENGINEERS

PIPE_A4_FP.indd 1 9/09/13 12:20 PM

Sample

copy

- not

for di

stribu

tion


A FLUID-PIPE INTERACTION model for multi-layered fibre-reinforced polymeric (FRP) pipelines subjected to hydraulic hammer is presented. Taking into account the flow parameters and the anisotropic elasticity

properties of the pipe wall, the pressure oscillations and the corresponding speed of the pressure shock due to sudden or fast reduction of discharge of a liquid is determined. A representative example for pipes made by E-Glass/epoxy is studied and the effect of material’s fibre orientation as well as the pipe’s diameter into pressure shock magnitude and speed is discussed.

Author’s contact details:tel: +30 6973 750530email:[email protected]

by Prof. Dr Dimitri G PavlouMechanical Engineering Department, Technological Institute of Halkida, Psahna, Halkida, Evoia,

Greece

Pressure-wave propagation in multi-layered fibre-reinforced polymeric pipelines due to hydraulic hammer

A SUDDEN CHANGE OF discharge in a multi-layered FRP pipeline may result in stresses of

sufficient magnitude to exceed the design stresses [1]. During sudden or fast closing of a valve, a pressure shock wave which travels with high speed in up- and downstream directions is developed as a result of the Newton’s impulse-momentum equation. The creation of this pressure shock in the fluid due to sudden or fast flow reduction should be considered as an important parameter for pipeline design.

Even though formulae providing the amplitude and the speed of the pressure wave in isotropic pipes (for example, steel pipes) are well known [2], according to the author’s knowledge fluid-pipe dynamic-interaction models for multi-layered FRP pipes are missing from the bibliography. Since both the flow parameters and the material properties of the anisotropic composite pipe wall control the dynamic deflection of the pipe, a description of the hydraulic-hammer-induced wave propagation is based on principles of unsteady flow and anisotropic elasticity.

In the present study, a model providing the pressure increase and the corresponding speed of the pressure wave due to hydraulic hammer in multi-layered FRP

pipelines is derived. The analysis is based on the continuity relationship between the pressure wave, Newton’s impulse-momentum law for a fluid, and the mechanical behaviour of a multi-layered anisotropic laminate composing the pipe’s wall.

Mechanical behaviour of a multi-layered FRP material

Hooke’s law for an anisotropic laminate

For a filament-wound multi-layered pipe, the coordinate system of a FRP layer is shown in Fig.1. Let E1, E2 be the moduli of elasticity in the directions x1, x2 respectively, and ν12, G12 the corresponding Poisson’s ratio and shear modulus. Hooke’s law in the principal coordinate system x1-x2-x3 for an anisotropic layer is given [3-5] by:

1 11 12 1

2 12 22 2

12 66 12

00

0 0

Q QQ Q

Q

σ εσ ετ γ

= (1)

where:

111

12 211EQv v

=− (2)

Sample

copy

- not

for di

stribu

tion


( ) ( )3 311 12 66 12 22 6626 2 2Q Q Q Q n m Q Q Q nm= − − + − + (12)

( ) ( )2 2 4 411 22 12 66 6666 2 2Q Q Q Q Q n m Q n m= + − − + + (13)

The parameters m and n are functions of the fibre orientation angle θ given by m = cos θ and n = sin θ.

Mechanical behaviour of a multi-layered anisotropic laminate

According to the classical lamination theory, the force (per unit length) Nx, Ny, Nxy, and moment (per unit length) Mx, My, Mxy resultants of a multi-layered laminate (Fig.2) are correlated [3-5] with the mid-plane strains and curvatures by the following relationship:

x

y

xy

x

y

xy

NNNMMM

11 12 16 11 12 16

12 22 26 12 22 26

16 26 66 16 26 66

11 12 16 11 12 16

12 22 26 12 22 26

16 26 66 16 26 66

B B BB B BB B B

B B B D D DB B B D D DB B B D D D

Α Α Α Α Α Α Α Α Α

=

0

0

0

0

0

0

x

y

xy

x

y

xy

kkk

εεγ

(14)

where:

11

( )N

ij k kijkk

A Q z z −=

= −∑ (15)

2 21

1

1 ( )2

N

ij k kijkk

B Q z z −=

= −∑ (16)

12 2 21 112

12 21 12 211 1v E v EQ

v v v v= =

− − (3)

222

12 211EQv v

=− (4)

66 12Q G= (5)

221 12

1

EE

ν ν= (6)

Assuming that the fibre orientation is θ (Fig.1), the corresponding Hooke’s law for the fibre in the global coordinate system x-y-z is [3-5]:

11 12 16

12 22 26

16 26 66

x x

y y

xy xy

Q Q Q

Q Q Q

Q Q Q

σ εσ ε

τ γ

=

(7)

where:

( )4 2 2 411 12 66 2211 2 2Q Q m Q Q n m Q n= + + + (8)

( ) ( )2 2 4 411 22 66 1212 4Q Q Q Q n m Q n m= + − + + (9)

( ) ( )3 311 12 66 12 22 6616 2 2Q Q Q Q nm Q Q Q n m= − − + − +

(10)

( )4 2 2 411 12 66 2222 2 2Q Q n Q Q n m Q m= + + + (11)

Fig.1. Coordinate system x1-x2-x3 for fibres with respect to the global coordinate system x-y-z.

Sample

copy

- not

for di

stribu

tion


For a multi-layered long FRP pipe (Fig.3) subjected to internal pressure δp, only the resultant Ny is predominant. Therefore, Equn 18 yields

022y ya Nε = (20)

The resultant Ny can easily be obtained from the equilibrium equation for a half pipe [6]:

12yN p Dδ= (21)

Therefore,

2212

oy a p Dε δ= (22)

or

22oy a p rε δ= (23)

3 31

1

1 ( )3

N

ij ijk k kk

D Q z z −=

= −∑ (17)

In the above equations, ijkQ (17a) is the parameter ijQ (17b) given by the Equns 8-13 corresponding to the k-th layer, zk-1 and zk are the distances of the end surfaces of the k-th layer measured from the interior surface of the pipe, and N is the number of layers making-up the pipe wall. Inversion of the matrix Equn 14 provides the relationship between the strains and curvatures with the force and moment resultants:

0

0

0

0

0

0

x

y

xy

x

y

xy

kkk

εεγ

11 12 16 11 12 16

12 22 26 21 22 26

16 26 66 61 62 66

11 12 61 11 12 16

12 22 62 12 22 26

16 26 66 16 26 66

a a a b b ba a a b b ba a b b bb b b d d db b b d d db b b d d d

α

=

x

y

xy

x

y

xy

NNNMMM

(18)

The 6 x 6 matrix consisting of the components αij is given by Equn 19 (below).

11 12 16 11 12 16

12 22 26 21 22 26

16 26 66 61 62 66

11 12 61 11 12 16

12 22 62 12 22 26

16 26 66 16 26 66

a a a b b ba a a b b ba a b b bb b b d d db b b d d db b b d d d

α

11 12 16 11 12 16

12 22 26 12 22 26

16 26 66 16 26 66

11 12 16 11 12 16

12 22 26 12 22 26

16 26 66 16 26 66

B B BB B BB B B

B B B D D DB B B D D DB B B D D D

Α Α Α Α Α Α Α Α Α

=

(19)

Fig.2. Nomenclature for (a) force resultants, and (b) moment resultants [3].

Sample

copy

- not

for di

stribu

tion


Taking into account the definition of density and assuming uncompressible liquid, the following relation can be obtained:

mV

δρ δ = (29)

or

2

VmVδδρ = − (30)

yielding:

VVδδρ ρ= − (31)

Combining the above equation with the following definition of the bulk modulus of elasticity K:

( )/pK

V Vδ

δ= − (32)

the following formula can be obtained:

pK δρδρ

= (33)

With the aid of the Equns 28 and 33, Equn 24 can now be written in the following form:

1 1p A pc cK A cδ δ δ

ρ = + + −

(34)

or

21 1

1

A ppA Kc

A p AA K A

δ δδρ

δ δ δ

+ + =

+ + (35)

Since δA/A << 1 and δp/K << 1, the above equation yields:

2 pc p AK A

δρ δ δ=+ (36)

Since A = πr2, the ratio δA/A can be written in the form:

2A rA rδ δ

= (37)

Unsteady flow model

Consider a pipe with cross section A, in which liquid with density ρ flowing from left to right at speed u is brought to rest by a pressure wave moving from right to left. Let the pressure wave travel towards the left with speed c relative to the oncoming liquid. The continuity equation across the wave for the above fluid model is given [2] by:

( ) ( ) ( )A c A A c uρ ρ δρ δ= + + − (24)

The equilibrium condition of the forces across the wave can be written:

{ }( )dmA p c u cdt

δ− = − − (25)

Taking into account that the mass flow rate dm/dt is given by:

dm Acdt

ρ= (26)

Equn 25 yields:

p c uδ ρ= (27)

or

puc

δρ

= (28)

Fig.3. Diagram of a filament-wound multi-layered pipe.

Sample

copy

- not

for di

stribu

tion


The velocity c*of a pressure wave in liquids of infinite extent is a constant given by:

* Kcρ

= (43)

Therefore, Equn 42 can be written in the following form:

22

1* 1 2

cc a r K

=+ (44)

The pressure amplitude due to the pressure wave can be obtained by Equn 27, i.e.:

22

**1 2

cp ua r K

δ ρ=+ (45)

Thus, the total liquid pressure in the composite pipe will oscillate within the range of

*tot op p pδ= ± (46)

where p0 is the static pressure of the pipe during steady-state flow.

The required time period T for the shock wave to travel back to the point where the sudden stopping of the flow happens:

2 LTc

= (47)

On the other hand, since the length of the perimeter of the pipe is 2πr, the strain in the middle of the wall’s thickness is:

( )22

oy

rr

δ πε

π=

(38)

or:

oy

rrδε = (39)

Combining Equns 37, 39, and 23, the following formula can be obtained:

222A a p rAδ δ= (40)

Therefore, Equn 36 yields:

2

222

pc p a p rK

δδρ δ ρ

=+

(41)

or

22

1

2c

a rKρ ρ

=+ (42)

Pressure shock speed [mls]

Material: E-Glass/EpoxyD=0.10mL=1000m

800

600

400

200

10 20 30 40 50Number of Plies

0=±75

0=±450=±300=±15

0=±60˚

Fig.4. Water-hammer-induced pressure wave speed vs number of layers (plies) for several values of filament winding angle θ for pipes made from e-glass/epoxy laminates.

Sample

copy

- not

for di

stribu

tion


Implementation of the model

Consider a multi-layered FRP pipe made from e-glass/epoxy laminate conveying a fluid with density ρ = 1000 kg/m3 and bulk modulus of elasticity K = 219 x 107 N/m2 (water). The length of the pipe L = 1000 m, and the speed of the flowing fluid u = 3 m/s. The mechanical properties of the layers composing the pipe’s multi-layered wall are E1 = 39 x 109 N/m2, E2 = 8.6 x 109 N/m2, ν12 = 0.28, and G12 = 3.8 x 109 N/m2, while the layer thickness h = 0.15 mm. Taking into account Equn 43, the velocity of the shock wave in the liquid is c* = 1480 m/s.

In order to estimate the effect of the fibre’s orientation and the number of layers composing the pipe wall on

where L is the length of the pipe as far as the valve which stops the flow.

In the case where the time of closure of the valve is not zero but Tv, and in the specific case where Tv ≥ T, the maximum overpressure can be obtained from the formula:

2*v

Lp uT

δ ρ (48)

In the case where Tv ≤ T, the rapid closure is considered as being equivalent to instantaneous closure, and the shock wave will reach its maximum value p* given by Equn 45.

Fig.5. Water-hammer-induced pressure vs number of layers (plies) for several values of filament winding angle θ for pipes made from e-glass/epoxy laminates.

Fig.6. Water-hammer-induced pressure vs number of layers (plies) for various pipe diameters for pipes made from e-glass/epoxy laminates.

Pressure [Pa]

Number of Plies

2.5106

2.0106

1.5106

1.0106

500000

10 20 30 40 50

0=±75

0=±60˚

0=±450=±300=±15

Material: E-Glass/EpoxyD=0.10mL=1000m

Pressure [Pa]

Number of Plies

1.4106

1.2 106

1.0106

800000

600000

400000

200000

10 20 30 40 50

Material: E-Glass/Epoxy0=±45˚L=1000m

D=0.90m

D=0.70m

D=0.50mD=0.30mD=0.10m

Sample

copy

- not

for di

stribu

tion


the hydraulic-hammer-induced pressure increase and pressure wave speed in pipes made from multi-layered FRP materials.

2. The model presented is based on the continuity relationship between the pressure wave, Newton’s impulse-momentum law for a fluid, Hooke’s law for an FRP anisotropic layer, and the compliance matrix of the multi-layered laminate composing the pipe’s wall.

3. An example of a pipe made from e-glass/epoxy laminate has been used, and the influence of the filament-winding angle, the number of layers (plies), and the pipe’s diameter on the water-hammer-induced overpressure and the corresponding pressure shock wave speed have been studied.

4. The results indicate that the hydraulic-hammer-induced overpressure and its wave speed increase when fibre orientation, number of layers, and diameter also increase.

References

1. D.G.Pavlou, 2012. Composite materials for piping applications. DESTech Publications.

2. B.S.Massey, 2006. Mechanics of fluids. 8th Edn, Taylor and Francis.

3. M.W.Hyer, 2009. Stress analysis of fibre-reinforced composite materials. DESTech Publications.

4. J.N.Reddy, 2004. Mechanics of laminated composite plates and shells. 2nd Edn, CRC Press.

5. L.P.Kollár and G.S.Springer, 2003. Mechanics of composite structures. Cambridge University Press.

6. A.C.Palmer and R.A.King, 2008. Subsea pipeline engineering. 2nd Edn, Pennwell.

the pressure increase, as well as the corresponding shock wave speed due to a sudden reduction of flow (for example, the rapid closing of a valve), we consider a pipe with a mean diameter D = 0.10 m. With the aid of Equns 2-6 the parameters Qij can be derived. Using the above values, the material properties are correlated with the fibre orientation θ using Equns 8-13. For pipes composed from laminates with a number of layers N = 10, 20, 30, 40, or 50, the stiffness parameters Aij, Bij, and Dij are determined by Equns 15-17. Therefore, the compliance parameters aij, bij, and dij can be obtained by Equn 19. The compliance parameter a22 is used for the determination of pressure wave speed as well as the overpressure by using Equns 44 and 45, respectively. In Figs 4 and 5 the above results are shown for pipes composed from layers with fibre orientations θ = ±15o, ±30o, ±45o, ±60o, and ±75o.

From Figs 4 and 5 it can be concluded that both the pressure shock wave speed and hydraulic-hammer-induced overpressure are increased by increasing the filament winding angle θ and the number of layers composing the pipe’s wall.

In order to detect the effect of the diameter on the overpressure due to the hydraulic hammer, we can keep the fibre-orientation angle constant. For θ = ±45o, the overpressure is correlated with the number of layers for five classes of pipe diameter, namely D = 0.10 m, 0.30 m, 0.50 m and 0.90 m. These results are shown in Fig.6, indicating that high water-hammer-induced overpressure is developed for high values of pipe diameters.

Conclusions

1. Taking into account the anisotropic mechanical properties and unsteady flow parameters, a fluid-pipe interaction model has been derived for estimating

Sample

copy

- not

for di

stribu

tion

SUBSCRIBE TODAY

DON’T MISS AN ISSUE!

Purchase your Pipelines International subscription online at pipelinesinternational.com/subscribe

PIN_Subscription_ad_FP_no_promo.indd 1 7/06/13 5:02 PM

Sample

copy

- not

for di

stribu

tion


Author’s contact details:tel: +48 22 76 70 984email: [email protected]

by Dr Maciej WitekDepartment of Heat and Gas Engineering Systems, Warsaw University of Technology, Warsaw,

Poland

An assessment of the effect of steel pipeline wall losses on the maximum allowable operating pressure of a gas pipeline

INTERNAL AND EXTERNAL wall defects in steel pipelines with a maximum operating pressure (MOP) above 1.6 MPa can be classified as follows:

• material losses resulting in a reduction of wall thickness, for example as a consequence of electrochemical corrosion caused by oxidation, or caused by either direct or alternating current at locations of damaged insulation;

• geometric shape defects caused by mechanical damage, such as dents or ovality;

• cracks caused, for example, by stress corrosion or fatigue;

• other types of defect caused during pipe manufacture or welding imperfections, such as delamination or inadequate weld penetration in the joints.

In practice, more than a single defect can be found at a particular location on a steel pipeline: wall damage such as corrosion, as well as different types of defect, such as a dent with wall thinning, may occur simultaneously. The majority of pipeline wall defects detailed above can be detected in the operating phase by internal examination performed using various inspection devices. Among the inspection methods used for pipelines with an MOP greater than 1.6 MPa on a commercial scale [1], inspections using intelligent tools provide the most data, with the proviso that the individual technologies used are each designed for the detection of different types of defect. The selection of a testing method depends on the experience gained from previous inspections, and particularly on the type of wall damage anticipated based on the available operational data. In some instances, it is justifiable to combine two inspection technologies during a single

tool run, for example the combination of magnetic and gyroroscopic technologies.

The present state of gas pipeline inspection technologies using inspection tools

At the present stage of development of diagnostics using inspection tools, the following technologies can be used by a gas network operator for the assessment of the technical condition of high-pressure gas pipelines in the operation phase, in compliance with the rules set out in Ref.2:

• internal geometry examination with a calibration tool operating on the swing-arm deflection principle, or by the eddy-current-based contactless method;

• inspection using the magnetic-flux leakage (MFL) method, either in standard resolution (SR) or high resolution (HR);

• examination with magnetic-flux leakage transverse-field inspection (TFI) tools, which have been developed especially for detecting defects in factory-made longitudinal welds, opened cracks, and material losses oriented in parallel to the pipe axis, with a circumferential dimension (width) of up to 10% of the wall thickness, which cannot be detected with the MFL HR technologies;

• examination with eddy-current-based shallow-internal-corrosion (SIC) tools, developed specially for the detection of internal gas pipeline wall defects up to 10 mm deep, which are detectable with a lower level of confidence using the MFL HR technology;

• inspection by guided waves using electromagnetic-acoustic transducers (EMAT), designed for the detection of wall cracks in gas pipelines, and stress-corrosion cracking (SCC) and fatigue cracks in particular;

Sample

copy

- not

for di

stribu

tion


It should be noted that no diagnostic tools have been developed so far – on a commercial scale – which would enable the detection of insulation defects. Also, with the use of inspection tools, it is not possible to determine the pipeline depth of cover or soil coverage thickness, but only to calculate the pipe burial depth by comparing the coordinates obtained from the gyroscopic inspection tool with the ground elevation determined by geodesic methods. In the Polish gas network, internal-geometry survey tools and MFL tools, primarily detecting gas pipeline wall material losses, have been employed to date: the assessment of their effect on the gas pipeline operation parameters is subjected to detailed analysis in this paper.

• survey with ultrasonic inspection tools to detect cracks in gas pipeline walls and welds, and in particular SCC and fatigue cracks, whose use in gas pipelines involves a liquid coupling medium between the sensors and the steel surface;

• examination with leak-detection tools, enabling the detection and location of gas leaks by acoustic or pressure-gradient measurement methods;

• examination with gyroscopic inspection tools to determine, using lasers, the gas pipeline running profile in the horizontal and vertical planes by determining the coordinates in the GPS (x,y,z) system.

Fig.1. Photograph of a wall defect formed during pipe manufacture, detected using a magnetic-inspection tool.

Fig.2. Photograph of a corrosion defect detected with a magnetic-inspection tool.

Sample

copy

- not

for di

stribu

tion


The calculation results for specific sizes of material loss were verified by laboratory tests by comparing them with the results of destructive pressure tests of pipe lengths containing defects. Recent issues of the standard documents that provide procedures for calculating the acceptability of various flaw types for steel gas pipelines are as follows:

• American Society of Mechanical Engineers Standard ASME B31G:2009 Manual for determining the remaining strength of corroded pipelines: supplement to ASME B 31:2009 Code for pressure piping [3];

• British Standard BSI BS 7910:2005 Guide to methods for assessing the acceptability of flaws in metallic structures [8].

In the assessment of the acceptability of a single wall-loss defect, the following equations (1-3), as provided in [3], are used to determine the circumferential stresses corresponding to a gas pipeline failure:

1f

e

1.1MRσ −=

for a defect in the entire wall cross-section, i.e. when d = t;

f

1e

2131.1 21 M

3

dt

dRt

σ−

−=

− for a partial gas pipeline wall loss

defect, when 0.1t < d < 0.8t and M ≤ 4.12;

Analysis of gas pipeline wall-loss defects

This type of damage, which involves a localized reduction in the pipeline wall thickness, mainly comprises the effects of the corrosion process and wall defects occurring during pipe manufacture, which are primarily detectable using MFL technology. An example of a wall loss in an operating gas pipeline, formed during pipe manufacture and detected with a magnetic inspection tool, is shown in Fig.1, while an example of a defect caused by corrosion is illustrated in Fig.2. A cross-section of a wall-loss defect in a gas pipeline is shown in Fig.3.

Calculation methods for determining the significance of a wall-loss defect in a gas pipeline subjected to static pressure can be classified as follows:

1. Analysis of limiting stress states based on the allowable strains resulting from the wall material’s yield stress values [3, 4, 5, 6, 7], with the proviso that the pipes should have relatively smooth contours (bottom profile), as is normally the case when considering the effects of electrochemical corrosion.

2. Use of finite-element stress analysis (FESA) for the non-linear numerical computation of the components of permissible strains and stresses, which makes it possible to take into account complex load states, as well as analysing the atypical shapes of wall losses that may occur at a single location on a gas pipeline [6, 8, 9].

3. Employing fracture-mechanics’ methods using the FAD (failure-assessment diagram), while regarding the wall loss as a notch [8, 10] in a model approach.

dt

t -- gas pipeline wall thickness

d 23 d

0.85 d

t

d

t

Fig.3. Cross-section of a model wall-loss defect in a gas pipeline.

Fig.4. Shape models of the wall-loss defect taken for calculation.

Sample

copy

- not

for di

stribu

tion


burst pressure, Pf, can be determined from the transformed equation:

ff

2

Z

tPDσ× ×

= (3)

There are many modifications of the above-mentioned limit-state method which enable the determination of a wall-loss defect’s acceptability, and where the calculation procedures are based on identical assumptions with the equations only differing in coefficients and the defect-shape model [10].

Later on in the paper, allowable-operating-pressure calculation results are provided for several defects involving wall-thickness reductions, which were obtained using the standardized calculation method from ASME B31G:2009, analysis level 1 [3]. For defects found as a result of the inspection of a DN-700 gas pipeline operated in Poland, and made of G-355 steel to the standard PN-79/H-74244 [11] with a yield strength of Re = 355 MPa, calculations based on Equns 1-3 are summarized in Table 1, where Dw is the gas pipeline’s internal diameter (mm), and the other designations are the same as in Equns 1-3.

An analysis of the relationships given in Equns 1 and 2 indicates that the circumferential stresses, σf, corresponding to the failure pressure of a gas pipeline with a defect, when M < 4.12, may not exceed 110% of the pipe steel’s yield strength, Re. At the same time, with the defect sizes given in Table 1, the effect of defect depth and defect length on the circumferential

f

e

1 dR tσ

= − for a partial wall-loss defect, when

0.1t < d < 0.8t and M > 4.12;

where:

2

Z

M 1 0.8 LD t

= + ⋅ (2)

in which:

σf = circumferential stress corresponding to gas pipeline failure pressure (MPa);Re = yield stress of the wall material of a given gas pipeline length (MPa);L = axial defect length (mm);d = defect depth (mm);Dz = pipe external diameter (mm);t = design rated wall thickness of a given gas pipeline length (mm).

The material loss depth, d, is taken for calculation depending on the defect bottom shape, which is assumed in the ASME B31G method as parabolic or – in its modified version – as rectangular, as shown in Fig.4. The values of the reduced depth of a single defect are indicated on the right-hand side of each of the shape models. Knowing the circumferential stress, σf, as determined from Equns 1 and 2, corresponding to the failure of a gas pipeline with a wall-loss defect, the static

Defect no. Dz (mm) d (mm) t (mm) d/tDw

(mm) L (mm) σf (MPa) Pf (MPa)

D1 711 6.7 10.0 0.67 691 24 380.99 10.72

D2 711 4.0 10.0 0.40 691 15 388.74 10.94

D3 711 3.7 10.0 0.37 691 49 376.59 10.59

D4 711 3.6 10.0 0.36 691 37 382.15 10.75

D5 711 3.2 10.0 0.32 691 13 389.51 10.96

D6 711 3.2 10.0 0.32 691 22 387.75 10.91

where:Dw – gas pipeline inner diameter (mm);The other designations are the same as in Equns 1-3.

Table 1. Calculation results for the defects detected by in-line MFL inspection of a gas pipeline with Dz = 711 mm, MOP = 5.5 MPa.

Sample

copy

- not

for di

stribu

tion


pressure-safety factor λ being chosen by the gas pipeline operator, the expression for the critical value of the relative depth of a defect meeting the conditions 0.1t < d < 0.8t and M ≤ 4.12 can be obtained, and will have the form:

ze

12

ze

z

MAOP1.65 0.75

MAOP1.1 1 0.82

kr

DRd tt

D LRt D t

λ

λ−

× ×− = × ×

− + (5)

where the pressure-safety factor is defined by the relationship:

f

MAOPPλ =

(6)

In the Polish regulations concerning the technical conditions to be met by gas pipeline networks [12], the MAOP means the maximum value of pressure to which a gas pipeline may be subjected. In view of the above, it can be assumed that the MAOP value will either exceed the MOP by the pressure increment resulting from the flow dynamics in an unsteady state, or will have a value below the MOP, if this results from the failure-pressure calculation, assuming the pressure-safety factor accepted by the network operator. For a typical high-pressure gas pipeline operating with circumferential

stress corresponding to the failure pressure of a given gas pipeline length and, as a consequence, on the static burst pressure, is relatively small, as demonstrated by the Pf pressure values, which are contained in a narrow range of 10.6-11.0 MPa. The limiting coefficient value of M = 4.12, given in Equns 1 and 2, for a pipeline diameter of Dz = 711 mm and wall thickness of t = 10 mm, corresponds to a defect length equal to L = 380 mm. Only for defects exceeding this length and the resulting value of M > 4.12 will the effect of the defect depth on the failure pressure be decisive. The significance of gas pipeline wall losses is best illustrated in the diagram of relative defect depth d/t versus defect length L.

For a partial wall-loss defect, i.e. 0.1t < d < 0.8t and M ≤ 4.12, a diagram can be plotted of the allowable relative defect depth d/t as a function of the so-called normalized defect length L/(Dz x t) 0.5, where, for a gas pipeline length with a constant diameter and wall thickness, the variable is L. To this end, Equns 1-3 can be transformed to:

f ze

12

f ze

z

1.65 0.75

1.1 1 0.82

P DRd tt

P D LRt D t

−

×− = ×

− + (4)

After substituting the maximum-allowable operating pressure (MAOP) in Equn 4, with the acceptable

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

00.00 0.50

d/t

1.00

D1

D2

D5D6

D4D3

λ=1.10

λ=1.25

λ=1.39

MS4

.12

1.50 2.00 2.50

L/(D2t)0.5

3.00 3.50 4.00 4.50 5.00

Fig.5. Diagram of the acceptability of 0.1t < d < 0.8t and M ≤ 4.12 flaws for partial wall-loss defects of a DN 700 gas pipeline made of G-355 steel.

Sample

copy

- not

for di

stribu

tion


The direct assessment of defects involves the field detection of the location of the defect and performance of a visual examination of the damage with non-destructive tests, normally ultrasonic tests. All flaws qualified for repair – based on the DA – should be immediately rectified, or the pipeline operating pressure should be reduced. It is also set out in the standard [13] that a gas pipeline that has defects from this group can still be operated without being repaired under a specific MAOP equal to the MOP, depending on the results of a direct assessment of the defects, of which none shall exceed the critical size within the specified time until the next inspection.

B. Wall defects of a gas pipeline with the λ factor in the range of 1.1-1.25 should be periodically assessed by the operator using DA until the next in-line inspection, with the frequency depending linearly on the λ value. In this case, the time interval until the next inspection in accordance with ASME B31.8S:2010 [13] shall not exceed five years.

C. Wall defects of a gas pipeline with the λ factor in the range 1.25-1.39 should be periodically assessed by the operator using DA until the next inspection with an inspection tool, with the frequency depending linearly on the λ value. In this case, the time interval until the next inspection in accordance with ASME B31.8S:2010 [13] shall not exceed ten years.

The above method of analysing the acceptability of wall-loss sizes depending on the maximum-allowable operating pressure is called the fitness-for-purpose assessment (FPA) of a gas pipeline, where the purpose is to transport the gas at the required pressure and volume. In the FPA analysis of a given high-pressure gas pipeline operated with circumferential stresses of magnitudes above 50% of the yield strength, the curves corresponding to the

stresses in the range of 50-72% of the yield strength, the λ factor for material-loss defects is taken from the range of (1.1-1.39) as the limiting values resulting from ASME B31.8S:2010 [3]. The diagram of the permissible relative defect depth d/t as a function of L/(Dz × t)0.5

for the defects from Table 1 is shown in Fig.5, with values of λ of 1.1, 1.25, and 1.39. The data from Table 1 have been plotted in Fig.5: all flaws lying below the (d/t)kr line for the λ factor value assumed by the operator should be regarded as acceptable. The higher the acceptable pressure-safety factor value assumed by the gas pipeline operator, the narrower the flaw-acceptability region, which is limited at the top by the curve with the position resulting from the assumed λ value. When assuming a higher pressure-safety factor value for permissible gas pipeline operating pressure analyses, such as λ = 1.39 (the inverse of the maximum design factor value of fo = 0.72 according to PN-EN 1594:2011 [14]), the diagram of acceptability for the defects from Table 1 will shift downwards, as shown in Fig.5. The recommendations for the acceptability of flaws of a high-pressure gas pipeline operated at circumferential pressures in the range of 50-72% of the yield strength, depending on the pressure-safety factor, are given in paragraphs A-C below, based on the standard [13].

A. The operator shall be obliged to perform a direct assessment (DA) of the defects within a period not exceeding five days if diagnostic examination finds corrosion defects in the wall of a gas pipeline with a design pressure-safety factor of λ ≤ 1.1, or should other flaws be detected which might cause leaks or result in a burst of the pipeline within a short time, as well as in case of finding any metal losses in contact with factory-produced longitudinal pipe seams made by one of the following methods:

• direct-current welding;• low-frequency electric-resistance welding; or• electric-flash welding.

Standard designation

Range of the required minimum yield strength in MPa

235 – 245 289 – 295 317 – 320

355 –360 385 – 390

413 – 415 445 – 450 480 – 485 553 – 675

API Spec 5L/ISO 3183

BBNBMBQ

X42X42NX42MX42Q

X46X46NX46MX46Q

X52X52NX52MX52Q

X56X56NX56MX56Q

X60X60NX60MX60Q

X65X65NX65MX65Q

X70X70NX70MX70Q

X80X80NX80MX80Q

PN-EN 10208-2

L245NBL245MB

L290NBL290MB

– L360NBL360MBL360QB

– L415NBL415MBL415QB

L450MBL450QB

L485MBL485QB

L555MBL555QB

Table 2. A summary of steels designed for manufacturing pipes according to API/ISO and European EN standards [15, 16].

Sample

copy

- not

for di

stribu

tion


as determined by the Charpy-V method for a single specimen at a temperature of 0oC. Table 2 details steels currently used for constructing gas pipelines according to American and European standards.

In view of the significant improvement in recent years in the quality of materials used for the manufacture of pipes intended for steel gas pipelines, the question arises: to what extent are the results of calculations for the static pressure corresponding to a pipe burst at the site of a defect, based on the ASME B31G:2009 [3], valid for L485MB-grade high-strength thermo-mechanically rolled steel to PN-EN 10208-2:2011 [15] and X70M according to ISO 3183:2007 [16]?

In the following, a fitness-for-purpose assessment will be made for two gas pipelines made from L480MB steel conforming to PN-EN 10208-2:2011, with the following parameters:

• Pipeline no. 2: Dz = 610 mm, MOP = 8.4 MPa, t = 11.3 mm (the pipes made with no negative wall thickness deviation), Re0.5min = 485 MPa; these parameters indicate that the gas pipeline will operate at circumferential stresses of up to 48% of the yield strength, which means the design factor fo < 0.5.

• Pipeline no. 3: Dz = 914 mm, MOP = 8.4 MPa, t = 13.4 mm (the pipes made with no negative wall thickness deviation), Re0.5min = 485 MPa; these parameters indicate that the gas pipeline will operate at circumferential stresses of up to 61 % of the yield strength, which means the design factor fo < 0.6.

pressure-safety factor values detailed above constitute the reference values for the network operator, which can be defined as below:

λ = 1.1 – burst curveλ = 1.25 – repair curveλ = 1.39 – defect-acceptability curve.

Gas pipelines made of thermo-mechanically rolled steel

In the previous section, results are provided for the circumferential stress corresponding to a failure and the burst pressure of a high-pressure gas pipeline made of alloy steel according to the 1979 material standard formerly applicable in Poland. The steels currently used in Europe for the manufacture of Class B linepipes intended for high-pressure gas pipelines conform to the PN-EN 10208-2:2011 [15] standard, and are designated with the letter ‘L’ and the number indicating the minimum yield strength value expressed in MPa, for example ‘360’. The subsequent letters have the following meaning:

N – normalized or normalizing-formed steel; M – thermo-mechanically rolled steel; and Q – toughened steel.

The final letter ‘B’ indicates that the steel meets the enhanced quality requirements for Class B pipes. Modern steels used for high-pressure gas pipelines should be fully normalized and constructed using a technology appropriate to fine-grained steel, and should meet high requirements for important strength parameters, for example the impact energy of a value above 30 J,

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

00 0.5

d/t

1

D8

D7

λ=1.39

λ=1.65

λ=2.00

MS4

.12

1.5 2 2.5

L/(Dzt)0.5

3 3.5 4 4.5 5

Fig.6. Diagram of the acceptability of 0.1t < d < 0.8t and M ≤ 4.12 flaws for partial wall-loss defects of a DN 600, MOP = 8.4 MPa, gas pipeline made of L-485-MB steel.

Sample

copy

- not

for di

stribu

tion


different depths which are, respectively, 35% and 60% of the wall thickness. Considering the design-pressure safety factors for λ > 1.39 (see the red curve in Fig.6), neither defect requires rectification.

Concerning the time intervals for the inspection of the gas pipeline operated at circumferential stresses within 30-50% of the yield strength, the reference λ value for a re-inspection time interval of five years is λ = 1.39, and for a re-inspection time interval of 10 years, the minimum pressure safety factor value is λ = 1.65, in accordance with Table 3 of ASME B.31.8S:2010, while for a 15-year re-inspection time interval, it is 2.0.

The different values of the design factor for the above two gas pipelines entail different assumptions for the fitness-for-purpose assessment based on ASME B31.8S:2010 [13].

For the DN 600 gas pipeline no. 2 made of L485MB-grade steel operated at circumferential stresses lower than 50% of the yield strength, the diagram of the permissible relative defect depth (d/t) as a function of L/(Dz × t) 0.5, shown in Fig.6, has been plotted for values of λ = 1.39, 1.65, and 2.00, taken from Table 3 of ASME B31.8S:2010 [13]. The defects, denoted by D7 and D8, have a similar length, but significantly

Defect no.

Re0.5min (MPa)

DN d

(mm)t

(mm)d/t

DZ (mm)

L (mm)

M σf

(MPa)Pf

(MPa)

MAOPdop (MPa)

λ = 1.39

MAOPdop (MPa)

λ = 1.65

MAOPdop (MP)

λ = 2.00

D7 485 600 4.10 11.3 0.35 610 210 2.47 450.54 16.69 12.01 10.12 8.35

D8 485 600 6.80 11.3 0.60 610 200 2.38 384.30 14.24 10.24 8.63 7.12

Defect no.

Re0.5min (MPa)

DN d

(mm)t

(mm)d/t

DZ (mm)

L (mm)

M σf

(MPa)Pf

(MPa)

MAOPdop (MPa)

λ = 1.39

MAOPdop (MPa)

λ = 1.65

MAOPdop (MP)

λ = 2.00

D9 485 900 4.00 13.4 0.30 914 300 2.62 462.38 13.56 12.33 10.85 9.75

D10 485 900 6.00 13.4 0.45 914 300 2.62 422.25 12.38 11.26 9.90 8.91

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

d/t D10

D9

λ=1.10

λ=1.25

λ=1.39

MS4

.12

L/(Dzt)0.5

0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

Table 3. A summary of defect-strength calculations for gas pipeline no.2, Dz = 610 mm, MOP = 8.4 MPa.

Table 4. A summary of defect-strength calculations for gas pipeline no.3, Dz = 914 mm, MOP = 8.4 MPa.

Fig.7. Diagram of the acceptability of 0.1t < d < 0.8t and M ≤ 4.12 flaws for partial wall-loss defects of a DN 900, MOP = 8.4 MPa, gas pipeline made of L485 MB steel.

Sample

copy

- not

for di

stribu

tion


followed by a destructive-strength test, and the results were compared to the actual burst pressures Pt determined experimentally. A summary of the results, based on Ref. 6, is given in Table 5.The relative deviation δ of the failure-pressure determination was calculated from Equn 7:

t f

t

100 %P PP

δ −= × (7)

The burst-pressure value obtained during the destructive strength test, Pt, is, in each case, higher than the value of Pf value calculated according to ASME B31G:2009 [3]; the relative deviation values are given in the last column of Table 6. The actual pressure-safety factor, λrz, relates to the value of the burst pressure Pt; the λ values determined from Equn 6 for the defects summarized in Table 6 are lower in each case.

Summary

For gas pipeline no.1 operated in Poland with circumferential stresses above 50% of the yield strength, and constructed from G-355 steel conforming to the standard [11], a defect assessment was made of the

The diagram of the permissible relative defect depth (d/t) as a function of the normalized defect length L/(Dz × t)0.5, for values of λ of 1.1, 1.25, and 1.39, in the case of the DN 900, MOP = 8.4 MPa, L485MB-grade steel gas pipeline no. 3 operating at circumferential stresses within 50-72% of the yield strength is shown in Fig.7. Table 4 and Fig.7 indicate that the pipe defects denoted as D9 and D10 lie in the acceptable region, considering the pressure safety factor λ = 1.39 as determined from Equn 6. The interpretation of the calculation results is similar to that for gas pipeline no. 1 operated at circumferential stresses within 50-72% of the yield strength.

The MAOPdop values given in the last three columns of Tables 3 and 4 are calculated from Equn 6 assuming the limiting values of the λ factor in accordance with Table 3 of ASME B31.8S:2010, and are the possible values of the MAOP to be maintained in a given gas pipeline during its operation phase.

Verification of failure-pressure calculations for gas pipelines made of thermo-mechanically rolled steel

For this purpose, within the experiment, artificial defects of the sizes given in Table 5 were made on pipe lengths made of thermo-mechanically rolled steel,

Defect no. DN d

(mm)t

(mm)d/t

DZ (mm)

L (mm)

σf

(MPa)Pf

(MPa)

D11 600 4.1 11.3 0.36 622.6 210 448.66 16.29

D12 600 6.8 11.4 0.60 622.8 200 386.82 14.16

D13 900 4.0 13.4 0.30 926.8 300 462.61 13.38

D14 900 6.0 13.4 0.45 926.8 300 422.57 12.22

Defect no. DN d

(mm)t

(mm)DZ

(mm)Pf

(MPa)

Test burst pressure, Pt

(MPa)

Actualpressure safety

factor, λrz

Relative deviation,

δ (%)

D11 600 4.1 11.3 622.6 16.29 21.4 2.67 23

D12 600 6.8 11.3 622.8 14.16 15.0 1.87 5

D13 900 4.0 13.4 926.8 13.38 17.3 2.06 23

D14 900 6.0 13.4 926.8 12.22 13.3 1.58 9

Table 5. A summary of defects and failure-pressure calculation results for the gas pipeline made from X-70M steel pipe conforming to ISO 3183:2007.

Table 6. Results of the determination of the relative pipe specimen failure-pressure deviation.

Sample

copy

- not

for di

stribu

tion


assessment, adequate knowledge by the analyst, and proven software, while on the other hand it makes it possible to analyse more-complex load states, as well as atypical defect shapes and different defect types occurring in the same place.

References

A. M.Witek, 2002. Preventive operation of a gas transmission network. Gaz Woda i Technika Sanitarna, 3/2002. SIGMA NOT Publishing House (in Polish).

B. API, 2005. Publication 1163: In-line inspection systems qualification standard. American Petroleum Institute Publishing Service, 1st Edn, Washington.

4. ASME, 2009. B31G:2009 Manual for determining the remaining strength of corroded pipelines: supplement to ASME B 31 Code for pressure piping. American Society of Mechanical Engineers, New York.

5. P. Hopkins, 2001. Defect assessment in pipelines. Training material from the Prague Conference, 6-8 March.

6. M.Witek, 2007. Inspections of high pressure gas pipelines under Polish conditions in the aspect of relevant legislation and regulations. Gaz Woda i Technika Sanitarna 10/2007. SIGMA NOT Publishing House (in Polish).

7. F.Dewint, 2011. Validation of the ASME B31G and RSTRENG methodologies for the determination of the burst pressure of corroded pipes in API 5L X70 / EN 10208-2 L485. 3rd ASME India Oil & Gas Pipeline Conf., February, (material available as per Aug. 15th, 2012 on the website www.novanumeric.com/samples.php?CalcName=B31G).

8. C.T.Belachew, C.I.Mokhtar, and K.Saravanan, 2009. Evaluation of available codes for capacity assessment of corroded pipelines. Petronas Universiti Teknologi, Mechanical Engineering Department, Bandar Sri Iskandar, Malaysia, February.

9. BSI, 2005. BS 7910:2005 Guide to methods for assessing the acceptability of flaws in metallic structures, British Standards Institution, London.

10. H.Moustabchir, Z.Azari, S.Hariri, and I.Dmytrakh, 2010. Experimental and numerical study of stress-strain state of pressurised cylindrical shells with external defects. Engineering Failure Analysis, 17, pp506-514.

11. G.Pluvinage, M.Allouti, C.Schmitt, and J.Capelle, 2011. Assesment of gouge, a dent, or a dent plus a gouge, in a pipe using limit analysis or notch fracture mechanics. J.Pipeline Engineering, 3rd Quarter, Great Southern Press.

12. PKN Warsaw, 1979. PN-79/H-74244: Steel seamed line pipes (in Polish).

13. The Regulation of the Minister of the Economy on the technical conditions to be met by gas networks. Dz. U. of 2001, No.97, Item 1055).

effect of the pipe wall material loss, as detected by inspection with an MFL intelligent tool, on the maximum-allowable operating pressure. Calculations based on the ASME B31G:2009 [3] methodology showed that the gas pipeline defects lay within the acceptable region with the assumed pressure safety factor of λ = 1.39, taken as the inverse of the maximum value of the design factor fo = 0.72 resulting from the Polish engineering and building codes [12] and the standard PN-EN 1594:2011 [14].

For two gas pipelines of DN 600 and DN 900, and MOP = 8.4 MPa made of L480MB thermo-mechanically rolled steel conforming to the standard PN-EN 10208-2:2011 [15], the assessment of the effect of the pipe-wall loss of material on the maximum-allowable operating pressure was made. By making calculations similar to those for gas pipeline no.1, it was demonstrated that individual defects might lie in a region above the limiting curve with a pressure-safety factor of λ = 2.0, although all of the defects determined were contained within the acceptable region if a pressure-safety factor of λ = 1.39 was assumed.

For X-70M thermo-mechanically rolled steel linepipe made in conformance with ISO 3183:2007 [16], the burst-pressure value obtained from the destructive hydrostatic test and the defect-failure pressure calculated based on the ASME B31G:2009 [3] methodology were compared, yielding results relatively deviating by up to 25%. In all of the cases examined, the destructive-strength test pressure was higher than the design value.

Calculations of the effect of pipeline wall-loss defects on the maximum-allowable operating pressure using the algorithms provided in ASME B31G:2009 [3] is now regarded as a conservative approach due to a considerable underestimation of the design pipeline burst pressure for some defects, as demonstrated by the comparison results provided in the present paper and in other publications [7, 10]. In addition, the defect-acceptability analysis based on limiting stress states [3, 8] is limited to a single type of wall defect in a given location without the possibility of examining different types of defect occurring at that location.

In recent years, the assessment of the acceptability of gas pipeline material-loss defects has been more and more often made using numerical methods relying on the non-linear finite-element method, an example of which is given in Ref.9, or by fracture-mechanics’ methods based on the FAD diagram, where a defect is treated as a pipeline wall notch [10]. The use of modern numerical methods is not standardized, though accepted by BSI BS 7910:2005 [8], which, on the one hand, requires an individual approach to

Sample

copy

- not

for di

stribu

tion


16. PKN Warsaw, 2011. PN-EN 10208-2:2011 Steel line pipes designed for combustible media – Technical specifications of delivery – Part 2: Pipes of the requirement Class B, PKN Warsaw 2011.

17. ISO, 2007. 3183:2007 Petroleum and natural gas industries – Steel pipe for pipeline transportation systems, Switzerland.

14. ASME, 2010. B31.8S:2010 Managing system integrity of the pipelines: Code for pressure piping – B31 Supplement to ASME B31.8, American Society of Mechanical Engineers, New York.

15. PKN Warsaw, 2011. PN-EN 1594:2011 Gas delivery systems – pipelines of a maximum operating pressure above 16 bar – Functional requirements, PKN Warsaw 2011.

Sample

copy

- not

for di

stribu

tion

What if you could have the acclaimed Pipeline Engineering courses from Clarion Technical Conferences and Tiratsoo Technical

right on your desktop?

NOW YOU CAN!In partnership with Penspen Integrity, Clarion and Tiratsoo Technical are pleased to announce the immediate availability of the popular course via online distance-learning:

> Pipeline Defect Assessment: a self-paced learning and reference resource.

Purchase the entire course or targeted groups of modules of most interest. All complete course purchases receive a 20% discount.

COURSE ALUMNI: If you attended the public course offered by Clarion and Tiratsoo Technical, you are eligible for a 10% discount on any of the four targeted groups of modules available, or 25% off the complete course.

For more information and to purchase the online courses, visit www.tiratsootechnical.com

OnlineCourses_PDA_2013_FP.indd 1 4/04/2014 9:45 am

Sample

copy

- not

for di

stribu

tion


IN AUGUST 2009, there was a 2.5-m long rupture in the longitudinal seam weld of a crude oil pipeline in France. The failure caused a spillage of approximately 2000 cubic metres in a protected area. This rupture

caused the authorities to withdraw the permit to operate a 260-km long section of this pipeline.

The pipeline had a similar failure in August, 1980. The 1980 failure was attributed to a fatigue crack initiating at the inner side of the longitudinal weld. There was evidence of ‘roof topping’ along this weld.

Penspen Ltd was contracted by the pipeline operator Société du Pipeline Sud Européen (SPSE) to carry out an independent investigation into the cause of the 2009 failure, review and confirm the actions needed for safe short-term operation to allow internal inspection, and determine a safe future life for the pipeline.

The failed pipe specimen was not immediately available for inspection. An initial analysis indicated that a purely analytical evaluation would not provide conclusive results, due to variability in material properties, geometry, and loading. It was decided that a better understanding of the behaviour of defects in the pipe, and the fatigue performance, was required. A detailed laboratory programme of burst and fatigue testing on a section of linepipe was recommended. The tests were carried out on a combination of ‘defect-free’1 ring specimens, and ring specimens with initial electric-discharge-machined (EDM) slits to represent crack-like defects. These tests showed:

1. Burst tests: The failure pressures of the burst tests could be predicted using a recognized industry model. Roof topping, laminations and inclusions, and toughness variations were found to have no noticeable effect on the defect size at failure.

2. Fatigue tests: The fatigue test results showed that defect-free rings had a fatigue life one order of magnitude longer than those containing an EDM slit, and that the fatigue life of a defect-free ring could be predicted using a standard S-N method. In addition, it was found that crack growth was conservatively predicted using standard fatigue-fracture mechanics.

The fatigue test results showed that the 1980 and the 2009 failures were caused by a combination of cyclic pressure loading, roof topping, and a pre-existing weld defect (probably present when the pipeline went into service in 1972).

This paper provides an overview of the failure investigation, with a discussion of the ring testing, supporting tests, and fracture surface inspection.

This paper was presented at the 9th International Pipeline Conference held on 24-28 September, 2012, in Calgary, AB, Canada.


by Mohamed Dafea*1, Dr Phil Hopkins2, Roland Palmer-Jones2, Patrick de Bourayne3, and Lionel Blin4

1 GL Noble Denton, Loughborough, UK2 Penspen Ltd, Newcastle-upon-Tyne, UK3 Trapil, Paris, France4 Société du Pipeline Sud Européen, Fos sur Mer, France

An investigation into the failure of a 40-in diameter crude oil pipeline

1. ‘Defect-free’ means that any deviation from sound pipe or weld material is within workmanship limits [1, 2].

Sample

copy

- not

for di

stribu

tion


testing was carried out by GL Noble Denton (GLND). The ring tests, of linepipe cut from the pipeline, included 11 static (‘burst’) tests and 13 fatigue tests.

Following completion of the ring tests, Penspen used a two-step approach to analyse the test results and explain the 2009 failure:

• S-N analysis: an empirical approach based on fatigue tests. This has been used to investigate the cause of the 1980 and the 2009 failures2.

• Analytical models based on fracture mechanics: this has been used to establish a conservative approach for determining the remaining fatigue life of the pipeline. The approach used the Battelle ‘flow-stress-dependent’ equation [5]), to determine defect limits, and BSI 7910:2005 [6] to model seam-weld fatigue-crack growth from an initial crack size to the critical size, where there is a geometry defect increasing the local stresses (e.g. roof topping and weld misalignment).

Nomenclature

D: pipe diameter (mm)EDM: electric-discharge machinedNlowerbound: lower-bound allowable number of cycles for fatigue design from an S-N curve, representing a 2.3% probability of failureNmean: allowable number of cycles to reach the mean S-N curve, representing a 50% probability of failureNupperbound: upper-bound number of cycles from an S-N curve, representing a 97.73% probability of failure∆P: pressure range (bar).SCF: stress-concentration factor due to geometry defects, such as roof topping, weld misalignmentt: pipe-wall thickness (mm)∆σhoop: hoop stress range resulting from internal pressure (MPa)∆σs: bending stress due to due to geometry defects, such as roof topping, weld misalignment (MPa).∆σtotal: total stress range taking into account stress elevation due to geometry defects, such as roof topping, weld misalignment (MPa)

Small-scale material tests

A series of small-scale material tests was carried out on three pipe samples, obtained from the pipeline adjacent to the 2009 failure, to ensure a complete

2. The analysis was based on both the 13 fatigue-ring tests carried out between 2010 and 2011, and the small-scale fatigue tests carried out in 1980.

SOCIÉTÉ DU PIPELINE Sud Européen (SPSE) owns and operates a 40-in diameter pipeline,

approximately 700 km long, which delivers crude oil to several refineries in France. The pipeline was commissioned in 1972 [3], and is constructed of API 5L X60 [2] linepipe, with wall thickness varying between 8.74 mm and 12.7 mm. The pipeline is pressure cycled due to variations in the demand for the oil.

1980 failure

In August, 1980, the pipeline ruptured at a longitudinal weld, and evidence of inclusions and laminations was present at the fracture surface of the failed section of pipe. A failure investigation study [3] attributed the failure to a fatigue crack initiating at the inner side of the weld, combined with a geometry defect known as ‘roof topping’, see Fig.1.

Roof topping is caused by failure to crimp the plate edges sufficiently before the ‘U’ and ‘O’ pipe-forming process, leaving a peak along the seam weld. Roof topping may decrease the fatigue life of a pipeline by concentrating the local stresses at the seam weld.

A failure investigation in 1980 focused on the following:

• theoretical modelling of the stress concentration caused by roof topping;

• defect depth to cause failure; and,• fatigue crack growth during service.

Operational limits were set for the maximum allowable roof topping height, and maximum defect depth, based on this work. SPSE carried out numerous repairs, between 1980 and 2009, taking into account the 1980 failure investigation results and operational experience.

2009 failure

In August, 2009, another longitudinal weld failure occurred (see Fig.3 [4]), and evidence of inclusions and laminations was again present at or around the fracture surface of the failed weld.

Penspen Ltd was again contracted by SPSE to investigate the cause of failure and estimate the remaining fatigue life of the pipeline. The failed pipe section was not available for inspection or testing. Initial analytical work indicated that the failure may have been related to the presence of a crack-like seam weld defect in 1972, when the pipeline went into operation, or it might have been caused by a combination of cyclic loading and roof topping.

It was decided to conduct a programme of small-scale material tests and ‘ring testing’ (Fig.2) to quantify the burst and fatigue performance of the pipeline, and this

Sample

copy

- not

for di

stribu

tion


test rig can also be pressure cycled to investigate the effects of fatigue loading. The ring sample is then hydraulically expanded until the sample fails. Notches can be machined into the sample and a fatigue crack can be grown from the edge of the notch by repeated pressure cycling in the rig, to investigate the effects of such cracks.

record of material properties was available. The following tests were carried out:

• Longitudinal and transverse tensile testing of the parent pipe material.

• Macro examination and hardness testing of the seam weld.

• Chemical composition of the parent pipe material.• Crack-tip-opening displacement (CTOD) fracture-

toughness testing of the parent pipe material.• Sulphur-print testing of parent pipe material.• Charpy impact testing of the parent pipe material

and the longitudinal pipe seam weld.

This small-scale testing results indicated the following:

• Samples 1 and 2: the pipe met the API 5L specifications in terms of tensile strength, Charpy impact toughness, hardness, and chemical composition, for X-60 pipe material.

• Sample 3: » The pipe met the API 5L specifications in terms of tensile strength, seam-weld Charpy impact toughness, and seam-weld hardness, for X-60 pipe material.

» Charpy tests of the parent material failed to meet the current minimum impact energy requirement of API 5L3 and recorded low percentage shear areas (12% and 15%), which can indicate low toughness.

Ring tests

A ring-tension test rig is shown in Fig.2. A short ‘ring’ of steel (approximately 76 mm in width) is cut from a pipeline and placed inside the test rig. A ring-tension

3. With an average of 17 J versus minimum required of 30 J.

long seam weld Peak or ‘roof top’

Fig.1. Illustration of the seam weld and roof-topping effect.

Fig.2. A ring-tension-test machine.

Fig.3. Failed section of pipe (2009).

Sample

copy

- not

for di

stribu

tion


objective was to understand the failure behaviour, i.e. whether the failure predominantly ductile or brittle.

Fatigue tests

Thirteen fatigue ring tension tests were also carried out. These tests consisted of two sets:

• 11 fatigue tests with initial EDM slits. • two fatigue tests without initial EDM slits.

Table 3 gives a summary of these fatigue-test results.

EDM slits were chosen, as it is very difficult to grow fatigue cracks to a prescribed depth and precise location in a seam weld with roof topping. The EDM slits behave similarly to fatigue pre-cracks in this material (see Figs 14 and 15).

The key objective of the fatigue tests was to understand the development and growth of fatigue cracks at the longitudinal seam weld toe.

The results of this testing programme were analysed in order to investigate the cause of previous failures, and estimate the pipeline remaining fatigue life.

Static tests

Eleven static ring-tension tests were carried out. These tests consisted of three sets:

• one static test on a ring on the ‘as-received’ pipe (i.e. it did not contain any machined defect).

• six static tests with EDM slits4. • four static tests with EDM slits and fatigue

cracks at the base of the slit.

Tables 1 and 2 provide a summary of these static-test results.

The key objective of these static tests was to be able to predict ‘critical defect sizes’ – defects that would cause failure at the pipeline’s operating pressure. A secondary

Ring no.Maximum roof topping height

(mm)

Measured EDM slit depth (mm)

Final crack average depth at

failure (mm)

Failure pressure (barg)

Comment

1 3.0 N/A 0.00 102.817 No EDM

2 3.1 1.05 1.05 95.0 EDM slit only

3 3.5 1.85 1.85 76.7 EDM slit only

4 3.6 2.85 2.85 61.2 EDM slit only

5 3.8 3.65 3.65 54.9 EDM slit only

6 4.2 5.35 5.35 43.2 EDM slit only

21 4.1 3.65 3.65 64.8 EDM slit only

Table 1. Summary of static test results on rings with an EDM slit only.

Table 2. Summary of static test results on rings with an EDM slit and a fatigue pre-crack.

Ring no.Maximum

roof topping height (mm)

Measured EDM slit

depth (mm)

Fatigue pre-crack initiation EDM slit + fatigue pre-

crack average depth (mm)

Failure pressure (barg)

Pressure range (bar)

Number of cycles

Fatigue pre-crack average depth (mm)18

7 4.3 0.85 10 to 40 18000 0.0019 0.85 98.3

8 4.4 0.9510 to 29 28000

3.40 4.35 54.010 to 40 6000

9 4.5 1.65 10 to 29 16800 3.08 4.73 47.7

14 4.8 2.6510 to 29 8500

2.1420 4.79 53.221 to 29 5000

17. Note this ring did not fracture but coned.18. The average depth of the fatigue pre-crack was measured from the fracture surface of the ring, after the test.19. No measurable crack initiated after the ring was fatigue cycled before the static test.20. Further fatigue cycling (2,000 cycles at a pressure range of 10 to 29 bar, and 4,000 cycles at a pressure range of 21 to 29 bar) was applied, but final fatigue pre-crack depth was not measured prior to the static test.4. An EDM slit has a width of 0.15 mm, and resembles a sharp defect such as a planar weld defect.

Sample

copy

- not

for di

stribu

tion


that total fracture of the material may occur below its maximum strength properties. Typical sources of cyclic loading in pipelines are:

• Cyclic loading during construction (for example, transportation of line pipes can cause ‘railroad cracking’).

• Cyclic loading during service (such as internal pressure cycling).

Fatigue is not a major cause of failures in oil and gas pipelines, but when it does occur, it is usually associated with longitudinal seam-weld defects, or damage such as dents.

A summary of this analysis is explained in the sections below.

Failure investigation

The 1980 and 2009 failures were attributed to fatigue cracks initiating at the inner surface of the pipe, at the toe of the seam weld, where there were geometry defects known as roof topping.

Fatigue is the progressive damage caused to a material when subjected to cyclic loading, and involves the formation and growth of a crack. Under repeated cycles, the crack may propagate to such an extent

Table 3. Summary of fatigue-test results.

Ring no.Roof topping height

(mm)Measured EDM slit

(mm)Pressure range (bar) Total number of cycles

10 4.3 2.7010 to 29 3,500

18 to 32 12,024

11 4.4 3.6510 to 29 1,200

10 to 40 683

12 4.4 2.55 10 to 29 10,305

13 4.1 2.6010 to 29 6,300

18 to 32 12, 460

15 4.6 2.0010 to 29 12,310

17 to 29 14,000

17 4.4 4.25 10 to 29 1, 565

18 4.4 1.3510 to 29 16,753

17 to 29 7, 000

19 4.4 1.3510 to 29 14,000

17 to 29 668

20 4.5 N/A 15 to 40 164,493

22 5.2 2.70

10 to 29 5,500

18 to 32 20,000

15 to 35 11, 173

23 5.0 2.8010 to 29 2,000

15 to 35 7, 513

24 3.8 2.8010 to 29 4376

21 to 29 4, 000

25 N/A21 N/A15 to 40 230,000

10 to 80 10, 165

21. Ring had no roof topping; however, the ring was oval and the weld was misaligned.

Sample

copy

- not

for di

stribu

tion


by the pressure variations in the pipeline6, and the equivalent number of cycles for various pressure ranges has been calculated using Miner’s7 rule [8]. This stress range is an approximate representation of the pipeline minimum and maximum operating pressures (10 bar and 29 bar), and a roof topping height of 4.5 mm.

The method used to calculate the equivalent number of cycles at this reference stress range for any ring is based on the conversion of a pressure range, ΔP, into a stress range, Δσ, taking into account the stress-concentration factor resulting from geometry defects5 (such as roof topping and weld misalignment).

S-N assessment of rings with EDM slits

Eleven fatigue tests were carried out on rings with EDM slits with depths between 1.3 mm and 4.3 mm8. Of these 11 rings, two had laminations, and similar slit depths at the start of the fatigue tests; however, their fatigue lives were dissimilar: one ring had a fatigue life three times longer than the other ring. The test results suggests that laminations do not

S-N assessment procedure

The limit to the number and amplitude of stress cycles which a defect-free material can survive is generally plotted on an S-N curve. The procedures are based on the quantitative relationship between the fatigue strength and number of cycles corresponding to a specific probability of failure.

S-N curve

The procedures described in BS 7608:1993 [7] have been used to assess the results of the fatigue tests and operational fatigue loading due to internal pressure cycling against S-N curve limits at a ‘reference stress range’, σref, see Fig.4:

• The lower-bound allowable number of cycles for design (Nlowerbound), representing a 2.3% probability of failure (POF).

• The allowable number of cycles to reach the mean curve (Nmean), representing a 50% POF.

• The upper-bound number of cycles (Nupperbound), representing a 97.7% POF.

The weld classification chosen to represent a double-sided full-penetration submerged-arc weld is Class E [7].

‘Reference stress range’

A ‘reference stress range’ of 317 MPa has been selected to represent the cyclic stress range caused

Fig.4. Class E S-N assessment curves.

5. Stress calculations take into account the elevation in stress resulting from roof topping and weld misalignments. These were calculated using BS 7910 for each ring, and validated using finite-element (FE) stress analysis.6. From most recent operating regime.7. Assumes linear damage, i.e. the combined effect of all cycles will cause the structure to fail when the total damage = 1.8. All EDM slits were positioned at the internal surface of the weld toe, on the misaligned side of the weld to match previous pipeline failures.

Class E, Mean Curve

1000

1000

100

10

1

1.E+02

Nlowerbound

1.E+04

N, Cycles (Log Scale)

Str

ess,

Mpa

(Log

Sca

le)

1.E+06 1.E+08

Class E, -2 Std Class E, +2 Std Reference Stress

Nupperbound

Nmean

Sample

copy

- not

for di

stribu

tion


is expected as the S-N curves do not account for the presence of a defect such as an EDM slit.

S-N assessment of rings without initial EDM slits

Two ‘defect free’ rings were fatigue tested and assessed against BS 7608 S-N curves; of these, only one ring had roof topping. The other ring had no roof topping; however, the ring was oval and the weld was misaligned. These rings have fatigue lives predicted by the S-N curve5, see Fig.6.

shorten the fatigue life, but the effect of laminations on fatigue life is difficult to predict, as it depends on its location relative to the pre-existing defect, as well as its shape and continuity.

Figure 5 shows the results of the S-N assessment of all fatigue tests on rings with EDM slits (the EDM slit depth is given in brackets after the ring number).

All rings with EDM slits had a fatigue life shorter than that of the lower bound predicted by BS 7608. This

Fig.5. Class E S-N curves and failures of rings with EDM slits.

Fig.6. Class E S-N curves and failures of ‘defect-free’ rings.Sam

ple co

py - n

ot for

distr

ibutio

n


All samples, except F1, had a fatigue life higher than the upper-bound limit given in BS 7608, so the S-N curve provides a conservative prediction of the fatigue lives of these ‘defect-free’ specimens, and are similar to the defect-free rings in Fig.6.

S-N fatigue assessment of previous failures

Pipeline pressure-history data was available for the period between 1994 and 2008. This data was analysed10, taking into account the stress elevation caused by roof topping, and an estimated number of completed pressure cycles at the reference stress range was calculated for the 1980 and the 2009 failures. This data was compared with the BS 7608 Class E S-N curve, Fig.8.

Figure 8 indicates that the fatigue loading applied up to the time of the 1980 and 2009 failures was not sufficient to cause failure if the welds contained only workmanship anomalies (i.e. no pre-existing defect was in the weld), despite the roof topping:

• It has been estimated that 40 years of service would have been required for the section of pipe with 7.5 mm of roof topping that failed in 1980 to reach a 50% probability of failure. In fact it failed after eight years.

• For the section of pipe that failed in 2009, which had 4.5 mm roof topping, 104 years of service would have been required, rather than 36 years.

Further analysis of the fracture surface of one of these rings indicated that the ring experienced a mixture of brittle and ductile fracture just after the crack had grown through the middle third of the remaining wall (an area of banded inclusions and laminations).

S-N assessment of the 1981 fatigue tests

Eight small-scale fatigue tests (four-point bending tests) had been carried out by Centre des Matériaux de l’Ecole des Mines in 1981 following the 1980 failure [9] on two samples (designated F and J) of the pipeline. Sample F had a large roof topping, offset of the external weld to the internal weld, and also contained micro cracks with depths between 15 µm and 100 µm (within weld workmanship limits). Sample J did not have any roof topping.

The samples were cycled between 100 MPa and 500 MPa. Specimens F1, F8, and J1 were halted prior to failure because there was a crack approximately 3 mm in length. Specimen F7 was tested at a different stress range which was not provided in the test report.

Figure 7 shows the comparison of these 1981 test results with the Class E S-N curve5. The number of cycles was corrected so that the data could be plotted at the reference stress range9.

Fig.7. Class E S-N curves and the 1980 fatigue tests.

5. Stress calculations take into account the elevation in stress resulting from roof topping and weld misalignments. These were calculated using BS 7910 for each ring, and validated using finite-element (FE) stress analysis.9. Note that variations in roof topping across the samples were not accounted for in this analysis, as they were tested directly in bending.

10. Pressure cycles were counted using the reservoir cycle counting method, see Ref.7.

Sample

copy

- not

for di

stribu

tion


• The fatigue life of the two operational failures (1980 and 2009) coincide with ring tests containing EDM slots.

The results of the fatigue testing indicate that there were pre-existing weld-toe defects present at the locations of the 1980 and 2009 failures. These defects were probably present when the pipeline went into service, and this suggests that the 1980 and the 2009 failures were not caused by roof topping alone: a combination of pressure cycling, roof topping, and a pre-existing defect caused these failures.

Remaining fatigue life calculations

Introduction

Conservative engineering models are needed in order to establish a conservative methodology to evaluate the remaining fatigue life of the pipeline. These engineering models must:

• Conservatively predict the defect size to cause pipeline failure.

• Conservatively predict the fatigue loading required to grow a defect from an initial known or assumed size to a size to cause failure (the ‘critical’ size).

Model to determine critical defect size

The recorded failure pressures and defect depths prior to failure of the ring specimens were compared with the values predicted using a recognized model from the pipeline industry: the Battelle ‘flow-stress-dependent’

Figure 9 shows that defect-free rings have a fatigue life approximately one order of magnitude greater than rings with defects (EDM slits).

The 1980 and 2009 failures have fatigue lives similar to those of rings with EDM slits. This suggests that there were pre-existing defects (for example, weld defects) at these failure locations. It is noted that:

• The cycles to failure for the 2009 failure (a lamination present, and possibly, a pre-existing defect) was 28,623 cycles11.

• The cycles to failure for one ring (initial EDM of 2.7 mm and a lamination) was 29,256 cycles11.

S-N assessment of previous failures

The equivalent number of years of operation at the reference stress range (317 MPa) for each ring was calculated from testing based on the pipeline pressure history between 1994 and 2008, see Fig.10. This figure shows when each specimen would be expected to fail, if it had been in the pipeline:

• Defect-free rings have a fatigue life one order of magnitude higher than rings with EDM slits.

• The reduction in fatigue life due to the bending stress introduced by roof topping is at least one order of magnitude12.

11. At a reference stress range of 317 MPa.12. The 1980’s fatigue tests were four-points bending tests. These tests will not capture the effect of stress elevation due to roof topping.

Fig.8. Class E S-N curves and the 1980 and the 2009 failures.

Sample

copy

- not

for di

stribu

tion


predicted by BSI 7910:2005 [6]. This was to establish a conservative method for prediction of remaining fatigue life of an existing defect in the pipeline.

Burst assessment of static test results on rings with EDM slits

Seven rings with EDM slits, and one ring without an EDM slit, were tested. Of these, one ring had a mid-wall lamination and an EDM slit, and was tested to evaluate the effect of laminations on the failure pressure. All rings had roof topping, see Table 1.

All rings showed a predominantly ductile failure, although a small area of brittle fracture is evident near the outer surface of each ring. In addition, inclusions

equation [5]. This is to establish a relationship between the depth of a defect in the toe of the weld and the failure stress of the pipe sample. The Battelle equation was developed in the USA in the 1960s by the Battelle Memorial Institute to predict the failure behaviour of linepipe containing defects. Battelle carried out a large number of full-scale tests (over 300) and developed a semi-empirical model. This model is now widely-used to predict the failure pressure of a part-wall defect in thin-walled (low-constraint), ductile linepipe steel.

Model to determine fatigue life

Six rings were monitored during fatigue testing, and the number of cycles to failure (excluding number of cycles for crack initiation at the slit) was compared with the values

Fig.9. Fatigue-life comparison.

Fig.10. Operational assessment of fatigue tests and failures.

Sample

copy

- not

for di

stribu

tion


present in the steel are evident in the mid-thickness. A typical example is given in Fig.11 for the fracture surface of a ring, with the relevant micrographs of the surface in Fig.12.

The results of burst tests on EDM slits are illustrated in Fig.13.

The results on EDM slits match the failure pressures predicted using the Battelle equation. These ring tests show that the critical defect depth for the pipeline at a pressure of 40 bar (resulting in a hoop stress of 233 MPa) is approximately 5 mm (see Fig.13).

Burst assessment of static test results on rings with EDM slits and fatigue pre-cracks

Three rings with EDM slits and fatigue pre-cracks14 were tested. The fracture surfaces of the tests with fatigue cracks grown from the EDM slit were similar to those from tests on EDM slits only, i.e. predominately ductile with a small brittle area near the outer surface.

The results of burst tests on EDM slits and fatigue cracks are illustrated in Fig.14. These results show that the failure pressures of rings with EDM slits and fatigue cracks in this linepipe material are both reasonably predicted by the Battelle model.

Burst assessment of fatigue tests

The ring tests used to assess the effect of fatigue were investigated to give more burst test results: this was achieved by estimating the fatigue crack depth at failure. The results are shown in Fig.15, and add more predictions for cracks in the weld: they show that the failure pressures of rings with EDM slits or fatigue cracks are again reasonably predicted by the Battelle model.

The burst test results show that:

• The failure pressures of the ring tests were reasonably predicted using a recognized industry failure model.

• Roof topping and laminations have no noticeable effect on the burst pressure, or critical defect size for failure.

• The pipe failure is predominately ductile, and the failure pressure depends on the tensile strength of the steel, rather than the toughness.

Fig.11. View of the fracture surface of a ring from static-ring tests.

Fig.12. Electron micrographs of the ring.

14. Fatigue cycles were applied to these rings to generate initial fatigue pre-cracks. The rings were then pressurized to failure, and the average depth of the fatigue pre-crack was measured from the fracture surface.

Analytical modelling of fatigue test results: future operation

Fatigue testing of each ring was carried out at different pressure ranges in order to simulate operational conditions, see Table 3. A retrospective study of fatigue-crack growth on six rings was undertaken in accordance with BSI 7910, using the TWI CrackWise software [10]. These rings were monitored during testing, by regular visual checks, and the number of cycles to initiate a crack evident at the edge of the ring was recorded. Consequently, it was possible to exclude the cycles for crack initiation from the fatigue assessment, and only assess the growth of the fatigue crack.

The following assessment procedure was followed:

• Calculate the critical crack depth at the maximum operating pressure using the Battelle flow stress dependent equation.

• Convert pressure ranges into hoop stress ranges as [11]: ∆σhoop = (∆P x D)/(20t)

• Calculate SCFs for each ring, resulting from roof topping at the minimum and maximum pressures of a pressure range using BS 7910.

Sample

copy

- not

for di

stribu

tion


made using this assessment approach is shown in Table 4. The results of this comparison show that the fatigue life of a section of this pipe with roof topping and a crack-like seam-weld defect is conservatively predicted using BSI 7910 (with a minimum safety factor of 3.6, see Table 4). This methodology can therefore be used to predict the fatigue life of known defects in the pipeline16, or defects that could be present, but are below the detection threshold of any in-line inspection system used in the future.

• Calculate the bending stress range using BS 7910 (Annex D) as:

• ∆σs = (σhoop (SCF -1))@min. pressure – (σhoop (SCF -1))@max. pressure• Calculate the number of cycles required to grow

a crack from an initial depth to a final depth at the loading conditions above (hoop and bending stress ranges) using BSI 791015.

A comparison between the numbers of cycles recorded during the ring-fatigue tests and the fatigue-life predictions

Fig.13. Ring-test results for rings with both no EDM slit and with an EDM slit.

Fig.14. Test results for rings with EDM slits and fatigue pre-cracks.

15. Based on the upper bound values of BSI 7910 for Paris fatigue crack growth law (constants: m = 3, and A = 2.3 x 10-12).

16. The information on critical defect sizes will also be useful when planning future inspections and evaluating the crack detection thresholds of different tools.

Sample

copy

- not

for di

stribu

tion


Fig.15. All ring-test results, with EDM slits or fatigue cracks.

Table 4. Retrospective studies on fatigue-crack growth.

Ring no.

EDM slit + pre-crack

depth22 (mm)

Crack depth after

fatigue growth (mm)23

Crack growth (mm)

Pressure range (barg)

Hoop stress range (MPa)

Bending stress range (MPa)

No. of cycles from

BS 7910

No. of cycles from test

Safety factor24

10 3.2 5.23 2.03 18 to 32 80 126.8 1,524 12,024 7.9

12 3.05 5.25 2.20 10 to 29 108.10 200.3 623 5,305 8.5

133.10 5.20 2.10 10 to 29 109.10 187.8 616 3,800

> 65.20 5.20 025 18 to 32 80.42 121.3 0 12,460

15

2.50 4.09 1.59 10 to 29 107.99 203.9 1000 1,000

> 11

4.09 5.26 1.17 17 to 29 68.2 117.4 633 7,000

5.26 5.26 025 10 to 29 107.99 203.9 0 3,500

5.26 5.26 025 17 to 29 68.20 117.4 0 7,000

5.26 5.26 025 10 to 29 107.99 203.9 0 1,810

181.85 2.88 1.03 17 to 29 67.97 117.1 7,000 7,000

3.62.88 5.28 2.40 10 to 29 107.61 202.3 766 2,735

243.30 3.81 0.51 21 to 29 45.72 63.4 4,000 4,000

4.33.81 5.22 1.41 10 to 29 108.57 172.6 1,176 275

22. From test.23. Predicted using BSI 7910.24. Number of cycles from test / number of cycles predicted by BSI 7910.25. Predicted to fail.

Sample

copy

- not

for di

stribu

tion


in particular, Wilson Santamaria, of ExxonMobil, UK, for valuable support and discussions throughout the course of this work and permission to publish. The diligent testing carried out by GLND, UK, is also gratefully acknowledged.

References

1. A.Cosham and P.Hopkins, 2004. An overview of the pipeline defect assessment manual (PDAM). 4th International Pipeline Technology Conference, Ostend, Belgium.

2. API, 2007. Specification for Line Pipe. ANSI/API Specification, API 5L, 44th Edition.

3. F.Hourlier and A.Pineau, 1981. Bursting of a Bergrohr tube of the SPSE pipeline: Final report. Ministere de L’Industrie, Ecole National Superieure des Mines de Paris, Centre des Materiaux.

4. Anon., 2010. Troncon Fissure No43440A Reception (5 Planches). Appendices of the 2009 Failure Investigation Report, Report No.: 4823115-001-1 PC.

5. J.F.Kiefner, W.A.Maxey, R.J.Eiber, and A.R.Duffy, 1973. The failure stress levels of flaws in pressurised cylinders. ASTM STP 536, American Society for Testing and Materials, Philadelphia, pp 461-481.

6. BSI, 2005. Guide to methods for assessing the acceptability of flaws in metallic structures. BSI 7910: 2005, incorporating Amendment 1 (2007), British Standards Institution, London, UK.

7. BSI, 1995. Code of Practice for the fatigue design and assessment of steel structures, incorporating Amendment No. 1, BS 7608:1993, British Standards Institution, London, UK.

8. M.A.Miner, 1945. Cumulative damage in fatigue. J. Appl. Mech., 12, Trans. ASME, 67, pp A159-A164.

9. Anon., 1980. Inspection and tests performed on a welded pipe burst in operation. Institut de Soudure Test Report, 62 920.

10. Anon., 2009. Crackwise software; compiled by The Welding Institute, Version 4.1.

11. ASME, 2010. Gas transmission and distribution systems. ASME Code for pressure piping B31, ASME B31.8-2010, Revision of ASME B31.8-2007, The American Society Of Mechanical Engineers.

Conclusions

Static ring test results show:

• The failure pressures of the ring tests were reasonably predicted using a recognized industry model.

• Roof topping and laminations have no noticeable effect on the burst pressure, or critical defect size for failure.

• The pipe failure is predominately ductile, and the failure pressure depends on the tensile strength of the steel.

Fatigue ring test results show:

• The fatigue life of the pipeline can be conservatively predicted using a recognized industry model.

• The reduction in fatigue life due to the bending stress introduced by roof topping is at least one order of magnitude.

• Rings with pre-existing defects have a fatigue life one order of magnitude less than ‘defect-free’ rings.

Causes of pipeline failures:

• This study has shown that the 1980 and the 2009 failures of the SPSE pipeline in France were a combination of pressure cycling, roof topping, and a pre-existing defect.

Future operation:

• A conservative methodology for predicting the remaining fatigue life of the pipeline has been developed and validated by testing.

Acknowledgments

The authors would like to acknowledge the support of SPSE, the SPSE shareholders (Total SA, ExxonMobil Corporation, Société de Participations dans l’Industrie et le Transport du Pétrole, BP France, Société des Pétroles Shell, BASF SE, ConocoPhillips Germany GmbH), and Sam

ple co

py - n

ot for

distr

ibutio

n


THIS PAPER PRESENTS findings from a study conducted as part of a joint industry effort involving engineers from Williams Midstream, Stress Engineering Services, Inc., GL Noble Denton, and Saipem

America. The purpose of this study was to evaluate the severity of damage inflicted to Williams’ subsea 18-in x 0.875-in wall thickness, Grade X-60 Canyon Chief gas export pipeline due to an anchor impact at a water depth of 2,300 ft. The phases of work included an initial assessment after the damage to the deepwater pipeline was detected, evaluating localized damage via finite-element analysis based using in-line inspection data, full-scale destructive testing including burst tests, and final efforts included the design and evaluation of a subsea-deployed repair sleeve. The study included modelling Saipem’s repair sleeve design accompanied by full-scale destructive testing. Strain gauges were used to measure strain in the reinforced dent beneath the sleeve, which was then compared to prior results for the unrepaired dent test results.

The work associated with this study represents one of the more comprehensive efforts conducted to date in evaluating damage to a subsea pipeline. The results of the analysis and testing work provided Williams with a solid understanding of the behaviour of the damage inflicted on the pipeline, and what level of performance can be expected from the repaired pipeline during future operation. After the engineering analysis and testing phases of this work were completed, the deepwater pipeline was repaired.

This paper was presented at the 9th International Pipeline Conference held on 24-28 September, 2012, in Calgary, AB, Canada.


by Dr Chris Alexander*1, Alexander Aalders2, William Bath3, Brent Vyvial1, Rhett Dotson1, and Danny Seal4

1 Stress Engineering Services, Inc., Houston, TX, USA2 Williams Midstream, Houston, TX,USA3 Saipem America, Houston, TX, USA4 GL Noble Denton International, Inc., Houston, TX, USA

Evaluating anchor impact damage to the subsea Canyon Chief pipeline using analysis and full-scale testing methods

THE WILLIAMS Canyon Chief 18-inh diameter pipeline was hooked by an anchor in late 2005 at

a water depth of 2,300 ft and the resulting damage pulled the pipeline laterally 1,500 ft from its original path. Inspection efforts using ROVs at the time of the accident indicated that the pipeline was not leaking. However, in the interest of safety, the pipeline pressure was lowered to approximately 800 psi (15% SMYS pressure, where SMYS is the Specified minimum yield strength of the pipe material) and allowed to continue operation while a remediation method was developed

(the repair was made at a reduced pressure). The intent, after remediation work was completed, was that the pipeline would be returned to the full 3,200-psi (55% SMYS) operating pressure.

A minimum level of information was available; however, the clearly-defined objective from Williams was to develop a reinforcing solution to restore integrity to the damaged pipeline that involved a dent having material loss in a bent section of pipe. Sources of information included ROV video footage, in-field measurements using ROV-assisted tools, and in-line inspection (ILI) data that provided the three-dimensional geometry of the dent. Figure 1 is a photograph taken using an ROV showing the geometry of the dent. As observed in this figure, the coating was relatively intact, although the ILI tool did detect metal loss in the vicinity of the dent.

Sample

copy

- not

for di

stribu

tion


defect was selected and evaluated for further study. The assessment included detailed modelling, as well as full-scale destructive testing. In a parallel effort, Williams retained the services of Saipem America to assist in the design, assessment, construction, and deployment of a repair sleeve.

The main focus of the testing programme was to experimentally quantify the severity of damage inflicted to the pipeline by the anchor snag. The limited finite-element modelling supported the experimental work, primarily to size the indenter geometry. ILI data provided by Rosen was used to generate a representative dent defect, including the associated metal loss. The repair sleeve, designed by Saipem America, was tested as part of the programme, with results being compared between the reinforced and unreinforced dent geometries to evaluate the effectiveness of the repair.

Figure 2 includes a sonar image of the bend in the pipeline, which was overlaid with scale circles used to provide an estimate of the radius of bend. As shown, the radius of the bend was between 35 and 80 ft.

After the initial inspection efforts were completed, Williams contracted the services of Stress Engineering Services, Inc., to perform an assessment of the pipeline damage. Finite-element analysis, along with full-scale destructive testing, was used to evaluate the damage inflicted to the 18-in x 0.875-in, X-60 Canyon Chief gas export pipeline. At the time of the incident, the operating pressure was 1,450 psi (25% SMYS), while the maximum allowable operating pressure (MAOP) is 3,600 psi (62% SMYS). Multiple defects were detected in the pipeline and measured during the ILI, some involving dents with combined metal loss. From among the identified defects, the most severe dent

Fig.1. ROV-assisted measurement of the dent using a straightedge.

Fig.2. Sonar image with radius-of-curvature estimation of the pipeline.

Sample

copy

- not

for di

stribu

tion


evaluated using finite-element analysis (FEA), although the primary focus was a 7.4%-deep dent, identified by Rosen as Dent ID 339968. Figure 3 provides a stress contour plot from the FEA for the dented region at a pressure of 100% SMYS (a hoop stress of 60,000 psi). As noted in Fig.3, the maximum principal stress is 221.4 ksi, corresponding to a stress-concentration factor of approximately 3.7. FEA was also used to calculate stresses in the sleeve, and also estimate the level of strain reduction in the dented region due to the presence of the sleeve. Figure 4 is a schematic diagram showing the layout for this particular FEA model, while Fig.5 plots strains in the dented region with and without reinforcement. As observed in this latter figure, the sleeve acted to reduce strain in the dent by a factor of approximately five at a pressure of 1,450 psi (25% SMYS).

The sections of this paper that follow provide details on the finite-element modelling work, experimental assessment efforts, and design/fabrication/deployment of the sleeve technology to reinforce the damaged Canyon Chief pipeline. The authors of this paper were able to participate in all phases of this project, spanning the initial assessment of the dent in question after its discovery to actually designing and deploying the repair technology.

Finite-element modelling

An ILI conducted by Rosen detected the presence of multiple dents, including one having a depth of 7.4% of the outside diameter of the pipe. Additionally, this dent damage included a localized metal loss having a depth of 11 percent of the pipe’s nominal wall thickness. A total of 16 dents were

Fig.3. Stress-contour plot for the dented region 100% SMYS (hoop stress of 60,000 psi).

Fig.4. FEA model geometry of the repair sleeve.

Sample

copy

- not

for di

stribu

tion


Sub-scale testing efforts

Several sub-scale tests were performed using 8-in nominal diameter pipe material to evaluate the performance of two load-transfer filler materials in a cold-water environment. The sub-scale tests were designed to address the following questions that were raised during the early phases of this project:

• Is it possible for a filler material to effectively fill the annulus between the outside surface of the pipe and the inside of the sleeve?

• Will a cement-filler material properly reinforce the dent within the clamp, especially in light of concerns in using an epoxy filler material that might not cure in deep water at 40oF?

• How much reinforcement is actually provided by a steel clamp (i.e. how much will the steel sleeve reduce strain in the dented region)?

• Will the cement filler material properly cure in a cold-water subsea environment?

Figure 6 shows the cross-section of bolt-on sleeve for sub-scale testing, while Fig.7 is a schematic diagram showing the set-up for sub-scale dry test with repair sleeves. Figure 8 shows strain-range measurements and the associated stress-concentration factors from the sub-scale tests.

Five full-scale burst tests were conducted that included a range of defect types including dents and material losses, and Fig.9 is a schematic diagram showing the layout for the different test samples. The purpose in conducting burst tests involving the different dent/gouge combinations was to address the potential

An FEA model was also constructed of the pipeline-grout sleeve system and determined that expansion of the pipeline due to increased internal pressure would apply an expansion force through the grout that the sleeve would see as an internal pressure of 750 psi. However, it was recognized that the seals in the sleeve would not see a pressure due to expansion of the pipeline. Instead they would only see the pressure of the cement grout during injection. Therefore, it was accepted that some leakage of the seals during the hydrotest could be allowed. The grout sleeve itself is not a pressure-containment but is simply a strong-back for the grout. This fact simplified sealing of the sleeve on the pipeline and allowed a novel end-seal arrangement which kept the corners of the end seals at the horizontal split line pulled back away from the pipeline until after the sleeve was closed around the pipe.

Testing methods and results

In addition to the finite-element modelling work that was used to evaluate the relative severity of the dent, experimental investigations were undertaken to determine the relative severity of the dents with metal loss and also evaluate the feasibility of the proposed repair solutions. While the primary focus of the experimental work involved full-scale destructive testing, sub-scale testing using smaller-diameter pipe was also conducted to evaluate the level of reinforcement provided by select filler materials that had to be deployed from a boat to the repair being made at a water depth of 2,300 ft.

The sections of this paper that follow provide details on the testing methods and results associated with the sub-scale and full-scale tests, respectively.

Fig.5. FEA-calculated strains in the dented region with and without sleeve reinforcement.

Sample

copy

- not

for di

stribu

tion


Prior to denting, longitudinally oriented EDM notches were installed in Samples 1 through 3, while an axisymmetric groove was machined in Sample 4 to simulate metal loss in the dented region of the pipe (in the absence of actual data, a gouge length of 6 in was used). The most severe defect involved the dent with a 21% deep V-notch metal loss. One of the five tests involved the pipe being reinforced using the repair sleeve designed by Saipem (the defect in this pipe sample was the most severe combination: a 7.4% deep dent with a 21% gouge). As with the unrepaired samples, strain gauges were placed near the dent beneath the repair to measure stresses due to internal pressure loading. Cyclic pressure was also applied to all unreinforced and reinforced samples prior to burst testing to simulate future years of service.

defect severity that might exist in the pipeline. All measurement devices have an inherent uncertainty, and ILI tools are no exception. Williams Midstream and Stress Engineering Services, Inc., evaluated the range of damage scenarios, especially with regard to metal loss in the dent, and determined that the four following defects best represented the potential damage that might have been inflected to the pipeline. All dents involved a 7.4% deep dent (measured as a percentage of the pipe’s outside diameter), while all groove and gouge depths are measured as a percentage of the pipe’s nominal wall thickness.

• Sample #1: dent with 11% deep axial groove• Sample #2: dent with 21% deep axial gouge• Sample #3: dent with 11% deep axial gouge• Sample #4: dent with 11% deep circumferential

groove

Fig.6. Cross-section of bolt-on sleeve for sub-scale testing. Fig.7. Schematic diagram showing set-up for sub-scale dry test with repair sleeves.

Fig.8. Strain-range measurements from the sub-scale tests.

Sample

copy

- not

for di

stribu

tion


• Sample 3 failed at 9,986 psi (171% SMYS) » Dent with 11% deep axial gouge » Failed in the seam weld (away from the dent)

• Sample 4 failed at 9,530 psi (163% SMYS) » Dent with 11% deep circumferential groove » Failed in the seam weld (away from the dent)

• Repaired sample failed at 9,486 psi (163% SMYS) » Dent with 21% deep axial gouge » Failed in the base pipe outside of the repair

In contrast to the unreinforced samples where the maximum hoop strain was measured to be 11,166 με, the maximum hoop strain in the reinforced dent was measured to be 1,060 με (located 6 in axially from the centre of the dent). The significance of this comparison is that the Saipem sleeve was effective in reducing strain in the dented region to levels equivalent to those associated with hoop strains measured in the undamaged base pipe. Figure 11 is a photograph showing the burst test for the repaired sample: note that the failure occurred on the base pipe outside of the reinforcing sleeve.

The minimum failure pressure for any of the burst tests was 9,485 psi (2.6 times the MAOP of 3,600 psi, and 6.5 times the current operating pressure of 1,450 psi). In this sample, the hoop strain in the base pipe at 3,600 psi was measured to be 1,226 με (microstrain, where 10,000 με equals 1% strain), while the maximum hoop strain was measured to be 11,166 με located 2 in axially from the centre of the dent. Figure 10 is a photograph showing the burst test for Sample 2 (the 21% deep axial gouge). Of the unreinforced dents, this is the only on that failed in the dented/gouged region. Listed below (and provided in Table 1) is a summary of the burst test results, including results for the reinforced pipe sample.

• Sample 1 failed at 9,485 psi (163% SMYS) » Dent with 11% deep axial groove » Failed in the seam weld (away from the dent)

• Sample 2 failed at 9,739 psi (167% SMYS), the only failure to occur in the dent defect » Dent with 21% deep axial gouge » Failed in the dent/gouge area

Sample # Sample description Burst pressure (psi) Failure description1 Dent with 11% deep axial groove 9,485 Failed in seam weld (away from dent)2 Dent with 21% deep axial gouge 9,739 Failed in dent/gouge area*3 Dent with 11% deep axial gouge 9,986 Failed in seam weld (away from dent)4 Dent with 11% deep circumferential

groove9,530 Failed in seam weld (away from dent)

Repaired Dent with 21% deep axial gouge 9,486 Failed in base pipe outside of repair

* Note: This was the only failure to occur in the defect region of the pipe.

Fig.9. Schematic diagram showing the layout for the two test samples.

Table 1. Summary of burst-test results.

Sample

copy

- not

for di

stribu

tion


Grout sleeve design

An ROV-installable grout sleeve design was developed that would accept the full range of bend radii of 35 to 80 ft. The 28-in inside diameter of the sleeve was selected to provide a minimum of 3 in of grout around the pipeline at the minimum estimated radius. The ends of the sleeve were then cut at an angle to match the likely exit angle of a 45-ft radius. Openings in the end plates were also offset to improve the fit on the pipeline. The sleeve was then split horizontally so that vertically-positioned, cone-head screws would

Field deployment

The analysis and testing efforts were fundamental in supporting the design and deployment of Saipem’s grouted repair sleeve. The parameters that were studied in the prior phases of work were an integral part in determining what and how would be installed subsea. The sections below provide in-depth discussions on the various phases of work associated with the design and deployment of the sleeve. The pipeline was operating at approximately 900 psi at the time that the sleeve was installed.

Fig.11. Photograph showing the burst test for the repaired sample with the failure outside the sleeve.

Fig.10. Photograph showing the burst test for Sample 2 (21% deep axial gouge).

Sample

copy

- not

for di

stribu

tion


a fixed object on the sea floor. To mitigate this risk, the grout sleeve was deployed with a buoyancy module and a suppressor skid (Fig.12). The buoyancy module was sufficient to suspend the grout sleeve above the sea floor, and this method allowed the assembly to be landed on the sea floor next to the pipeline and then quickly disconnected from the vessel. Once this package was on the sea floor, the ROV was in complete control of the installation. A pair of ROV-operated winches, mounted on the spreader bar, was then used to pull the grout sleeve down onto the pipeline, and this step is illustrated in Fig.13. Later these winches were used to provide a controlled assent for the spreader bar and buoyancy module after the grout sleeve was disconnected.

Metrology tool

A special metrology tool was built to measure the pipeline curvature in the area where the grout sleeve would be installed. This ROV-operated tool was deployed on a skid that was disconnected from the lowering line after it landed on the sea floor. The ROV then docked with the tool and positioned it on the pipeline by centring it on the dent. The ROV powered two pairs of hydraulic arms on the tool to clamp the tool onto the pipeline; it could then disconnect from the tool and make a video record of the readings on 20 equally spaced mechanical gauges that rested against the pipeline at five selected positions along the length of the tool. After the tool was recovered to the workshop, short sections of pipe were fitted into the tool and positioned to duplicate the gauge readings from the sea floor. These short sections were then welded together, removed from the tool, and laid into the open grout sleeve to confirm the clearance around the pipe and the exit angle of the pipe from the sleeve.

Offshore installation

The installation of the grout sleeve required that an access hole be dredged under the pipeline. This hole needed to be approximately 20 ft long and 4 ft deep to provide clearance for the bottom half of the sleeve to swing closed under the pipeline (see Fig.13). It was recognized that the weight of the sleeve plus the cement grout would be approximately 12,000 lbs when the buoyancy module was released. Therefore, to prevent the pipe from sagging into the hole, ROV-operated pipe-support frames were installed approximately 30 ft back from either side of the hole. Each frame consisted of two mud mats and a single, motor-operated, grab. These frames did not lift the pipeline but simply added mud-mat support. The support frames were used during installation so that the pipe functioned as a simple beam: they did not lift the pipeline – rather, they supported it, so that

provide an ROV-friendly method of closing the sleeve around the pipeline. It was recognized that the centre of gravity of a horizontally-split sleeve would change significantly as the sleeve opened and closed. In order to minimize this effect, the spreader bar connecting legs were attached at specific positions, with one pair of legs attached to the top half of the sleeve and one pair attached to the bottom half of the sleeve. This articulated arrangement moved the spreader bar horizontally as the sleeve opened and closed.

To hold the sleeve closed, 48 screws were required, and a torque wrench was specially designed to tighten the screws using an ROV. After the sleeves were made up, the grout-fill pipe was mounted at the bottom centre of the bottom half of the sleeve. The top half of the sleeve incorporated three ROV-operated vent valves through which, during grout filling, the water inside the sleeve was forced out. When the ROV observed good, clean, grout exiting the vent, that valve was closed. This method allowed confirmation that the sleeve was completely filled with cement grout.

Deployment plan

It was also recognized that the most critical time during an offshore operation is the time when the crane or lowering line on the vessel is connected to

Fig.12. Grouted sleeve with suppressor skid and buoyancy module rigged for deployment.

Sample

copy

- not

for di

stribu

tion


After the grout sleeve was installed, the ROV pulled a fabric grout bag into the hole under the sleeve and this bag was then filled with cement grout to fill the vacant space below the grout sleeve. After the grout was allowed to cure for a few hours, the ROV disconnected the spreader bar from the sleeve and it, along with the buoyancy module and suppressor weight, was recovered to the surface. The pipeline-support frames were also released and recovered. The only components that remained on the sea floor were the grout sleeve and the grout bag supporting it. The final configuration is shown in Fig.14 in a photograph taken by an ROV. This remediation action cleared the pipeline for full operating pressure operation.

it could not deflect into the mud. A simple beam calculation was made that showed that the pipeline could support itself between the pipe-support frames. The weight of the sleeve was supported by the buoyancy until after the grout bag was installed in the hole, and the pipeline did not see any significant loading during the repair. While the sleeve, spreader bar, and buoyancy module were connected to the pipeline, there was a net upward force of approximately 1,000 lbs on the pipeline. After the sleeve was filled with grout, there was a net downward force of approximately 1,000 lbs. After the grout bag was filled, we believe the pipeline was fully supported and there was no force acting on the pipeline.

Fig.13. The ROV powers the spreader-bar winches to pull the grout sleeve down onto the pipeline.

Fig.14. The grout sleeve and the grout bag installed on the pipeline.

Sample

copy

- not

for di

stribu

tion


Bibliography

1. V.Alessandro, B.Bath, C.Alexander, A.Aalders, and D.Seal, 2012. Application of a grouted sleeve to remediate a damaged subsea pipeline. Paper OTC 23454, Offshore Technology Conference, Houston, Texas, 30 April – 3 May.

2. D. Raghu, R.Swanson, and C.Alexander, 2008. Methodology to establish the fitness for continued service of a hurricane damaged export pipeline in 1000 m of water. Paper OTC-19653-PP, Offshore Technology Conference, Houston, Texas, 5-8 May.

3. C.Alexander, 2009. Evaluating damaged subsea pipelines using an engineering-based integrity management program. Proc. IOPF, Paper IOPF2009-6002, ASME International Offshore Pipeline Forum, Houston, Texas, 28-29 October.

Closing comments

The information presented in this paper detailed an effort involving four organizations to restore full-service capacity to a deepwater subsea pipeline damaged by an anchor. The work involved a combination of analysis and testing techniques to support the design, fabrication, and deployment of a grouted subsea repair sleeve. In reviewing the associated body of work presented in this paper, the following observations are made:

• A horizontally split sleeve can be safely installed by an ROV on a live pipeline by use of a buoyancy module and pull-down winches.

• Cement grout can be prepared on a surface vessel and then pumped down a long hose to fill a grout sleeve on the sea floor.

• A purpose-built metrology tool with mechanical gauges and simple fabrication techniques can produce an accurate model of the shape of a damaged pipeline on the sea floor.

• Using analysis and full-scale testing techniques prior to deployment of repair methods improves confidence in the design and ensures that stresses in the damaged region of the pipeline are reduced to acceptable levels.

Sample

copy

- not

for di

stribu

tion

TRAINING

2014

Training courses – 2014

Are you up to speed?10000 2 4

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 offshore.

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

www.clarion.org

Onshore Pipeline Engineering31 March – 4 April 2014, Houston, Texas28 April – 2 May 2014, Newcastle, United KingdomPresented by Professor Phil Hopkins & Dr Roger King

This course is a formal introduction to pipeline engineering and will cover everything from pipeline engineering basics to material selection, pipeline corrosion and the repair and integrity management of those pipelines

Defect Assessment in Pipelines29 April – 1 May 2014, Newcastle, United KingdomPresented by Professor Phil Hopkins

This course will present the latest defect-assessment methods to pipeline engineers and managers. These methods will range from simple, quick, assessment methods, to the more-detailed fitness-for-purpose analysis.

Practical Pigging Operations31 March – 4 April 2014, Houston, TexasPresented by Penspen, Mears and Greene’s Energy Group

This hands-on training course provides a wide-ranging overview of all aspects of pigging operations.

The course takes place at Greene’s Energy Group’s water-driven pipeline test loop facility in Houston. During the five days, the participants typically participate in up to nine different runs of various tools.

Onshore Practical Pigging Training17-19 June 2014, Tricht, the NetherlandsPresented by Penspen & A.Hak Industrial Services

This first-hand and practical training course provides a wide-ranging overview of all aspects of pigging operations for onshore pipelines.

The site is A.Hak Industrial Services’ test loop facility in Tricht, The Netherlands. The facility has a 15m, 12 inch water-driven pipeline, and a second 150 m, 24 inch line under construction.

Subsea Practical Pigging Training17-19 September, 2014, Bergen, NorwayPresented by Penspen & KTN

A three-day training course covering the practical aspects of pipeline pigging. This training course, being held at KTN’s facility in Bergen, Norway, will provide a wide-ranging overview of all offshore aspects of pigging operations.

The syllabus includes both hands-on exercises using the KTN test loop, as well as classroom instruction, and full documentation.

Practical Pigging Training1-5 September 2014, Rio de Janeiro Presented by Penspen & CTDUT

This training course provides the opportunity to get first-hand knowledge experience and a wide-ranging overview of all aspects of pigging operations. The course will be held at the CTDUT facility adjacent to Petrobas’ Duque de Caxias refinery, near Rio de Janeiro.

The course uses CTDUT’s 14 inch, 120 m long, water and nitrogen/air-driven test loop at the site.

During the five days, the participants typically perform up to nine different runs of various tools.

Training_Ad_2014.indd 1 4/04/2014 9:37 am

Sample

copy

- not

for di

stribu

tion

Sample

copy

- not

for di

stribu

tion

journal of pipeline engineering incorporating the journal of

Documents