Find your next favorite book

Become a member today and read free for 30 days
Handbook of Materials Failure Analysis with Case Studies from the Oil and Gas Industry

Handbook of Materials Failure Analysis with Case Studies from the Oil and Gas Industry

Read preview

Handbook of Materials Failure Analysis with Case Studies from the Oil and Gas Industry

ratings:
3/5 (1 rating)
Length:
851 pages
6 hours
Released:
Sep 1, 2015
ISBN:
9780081001264
Format:
Book

Description

Handbook of Materials Failure Analysis: With Case Studies from the Oil and Gas Industry provides an updated understanding on why materials fail in specific situations, a vital element in developing and engineering new alternatives.

This handbook covers analysis of materials failure in the oil and gas industry, where a single failed pipe can result in devastating consequences for people, wildlife, the environment, and the economy of a region.

The book combines introductory sections on failure analysis with numerous real world case studies of pipelines and other types of materials failure in the oil and gas industry, including joint failure, leakage in crude oil storage tanks, failure of glass fibre reinforced epoxy pipes, and failure of stainless steel components in offshore platforms, amongst others.

  • Introduces readers to modern analytical techniques in materials failure analysis
  • Combines foundational knowledge with current research on the latest developments and innovations in the field
  • Includes numerous compelling case studies of materials failure in oil and gas pipelines and drilling platforms
Released:
Sep 1, 2015
ISBN:
9780081001264
Format:
Book

Related to Handbook of Materials Failure Analysis with Case Studies from the Oil and Gas Industry

Related Books
Related Articles

Book Preview

Handbook of Materials Failure Analysis with Case Studies from the Oil and Gas Industry - Elsevier Science

China

Preface

Failure analysis and prevention are important issues for all engineering materials and industrial structures. One of the most important engineering fields which plays an essential role in failure analysis is Materials Engineering. Whether the failure happens in service or during the production process (manufacturing defects), failure analysis must be performed in order to prevent it from happening again in future. Another important factor which must be precisely investigated is to determine whether or not the metallic components and structures are being used well.

There have been several reports of catastrophic accidents in the Oil and Gas industry in countries including the United Kingdom, Kuwait, the United States of America, Venezuela, etc. Most, if not all of these accidents, were due to materials failure and were thought to be the cause of a number of workers being killed, in addition to the environmental crises. For instance, in 1993, a natural gas pipeline running along a highway in Venezuela, exploded. The rapid combustion of the spreading gas caused a hellish situation that lead to the deaths of at least 50 people. It should be noted that a lot of factors, including bad design, improper manufacturing, low-grade raw materials in the pipe production, improper joining of the pipes and/or corrosion, can be the cause of such incidents.

This handbook covers analysis of materials failure in the Oil and Gas industry, where a single failure can result in devastating consequences for people, wildlife, the environment, and the economy of a region. The book combines introductory sections on failure analysis with numerous real-world case studies of pipelines and other types of materials failure in the Oil and Gas industry, including joint failure, leakage in crude oil storage tanks, failure of glass fiber-reinforced epoxy pipes, and failure of stainless steel components in offshore platforms, amongst others.

This handbook contains many failure real-world cases and case studies covering a wide spectrum of materials failure related to petroleum, petrochemicals, oil, and gas applications. The editors thank all the contributors for their excellent chapter contributions to this handbook, and for their hard work and patience during preparation, and production of the handbook. We sincerely hope that the publication of this handbook will help people from Industry and Academia to get the maximum benefits from the experience contained in the published chapters.

Summer 2015

Abdel Salam Hamdy Makhlouf

Mahmood Aliofkhazraei

Chapter 1

Failure analysis of oil and gas transmission pipelines

Brad James; Alexander Hudgins    Exponent, Menlo Park, California, USA

Abstract

While pipelines offer a safe, cost-effective way to move natural gas and hazardous liquids, pipeline leaks and ruptures can result in disastrous consequences. Pipelines can be susceptible to different metallurgical failure mechanisms including but not limited to manufacturing defects, third party damage, and corrosion. This chapter outlines common pipeline failure mechanisms and provides selected case studies, which highlight typical issues that can result in pipeline leaks and ruptures. Understanding these failure mechanisms is critical to mitigating risk of future incidents and managing the future integrity of the pipeline.

Keywords

Hazardous liquid

Pipeline

Transmission

Distribution

Failure

Corrosion

Damage

Seam

Gas

Chapter Outline

Introduction   1

Mechanical Damage   3

Longitudinal Seam-Weld Defects   6

3.1 Lap-Weld Defects   8

3.2 Lap-Weld Case Study   9

3.3 ERW Defects   12

3.4 Flash Weld Defects   14

3.5 Submerged-Arc Weld Defects   14

3.6 Submerged-Arc Weld Defect Case Study   15

3.7 Shielded Metal Arc Weld Defects   20

Corrosion   21

4.1 General Corrosion   22

4.2 Stress Corrosion Cracking   22

4.3 High-pH SCC Case Study   23

4.4 Near-Neutral pH SCC Case Study   24

4.5 Hydrogen-Stress Cracking   24

4.6 HSC Case Study   28

4.7 Grooving Corrosion   31

Fatigue   33

Conclusion   36

References   36

1 Introduction

There are over 2.3 million miles of pipelines in the United States that carry natural gas and liquid-petroleum products [1]; approximately 60% of which were installed before 1970 [2]. These oil and gas pipelines are the backbone of our energy-dependent society. Statistics indicate that oil and gas pipelines are a far safer, more efficient, cost-effective way to move these hazardous products than any other method of transportation [1]. However, when pipelines do rupture or leak, disastrous human, environmental, and business consequences can result. The most common issues that can result in pipeline leaks and ruptures include mechanical damage, seam weld defects, stress corrosion cracking (SCC), hard spots, fatigue, and corrosion. This chapter is intended to serve as a primer to help failure analysts to understand and interpret these more common pipeline failure mechanisms and provides specific case studies. The more we understand how and why pipelines can fail, the better we can maintain our aging infrastructure and prevent accidents in the future.

Internal pressure results in hoop stress in a pipeline and is typically quantified using Barlow’s formula [3]. These hoop stresses act to favor crack initiation and growth in a radial and a longitudinal direction. Hoop stresses and the corresponding preferential crack growth directions in pipe are shown schematically in Figure 1.1. Thus, piping anomalies that are most detrimental to integrity are typically those that are oriented in a longitudinal direction.

Figure 1.1 Schematic pipeline showing the orientation of hoop stress resulting from internal pressure and the preferential orientation of crack growth.

Stresses other than pressure-induced hoop stress can affect pipeline integrity. Residual stresses caused by inadvertent dents or gouges can contribute to crack initiation and growth. Welding can also be a source of residual stresses. Another of the more common stresses that can cause pipeline failures are those caused by geological forces, such as landslides, earthquakes, or other large-scale soil or pipeline movement. These can result in axial stresses that cause failures transverse to the pipe’s longitudinal direction, such as at girth welds.

Data from 2011 show that material and weld failures constitute the largest incident type for both hazardous liquid and gas transmission systems (Figures 1.2 and 1.3). Other authors have indicated that third party damage is the most common cause of incidents in gas transmission systems [4]. The following sections outline specific pipeline failure mechanisms and provide examples to illustrate key phenomena. A summary of failure mechanisms presented in this chapter is shown in Table 1.1, which also serves as an outline for the content presented in this chapter.

Figure 1.2 The distribution of incidents in hazardous liquid lines is shown between the years of 2006 and 2010 [ 20 ].

Figure 1.3 The distribution of incidents in gas transmission lines is shown between the years of 2006 and 2010 [ 20 ].

Table 1.1

Summary of Failure Mechanisms Presented in this Chapter

Mechanisms

Mechanical damage

Lap-weld defects

Electric-resistance weld defects

Flash-weld defects

Submerged-arc weld defects

Shielded metal arc weld defects

General corrosion

High-pH stress corrosion cracking

Near-neutral pH stress corrosion cracking

Hydrogen stress cracking

Grooving corrosion

Fatigue

2 Mechanical Damage

Mechanical damage occurs when a pipeline is struck by mechanical equipment, such as a backhoe. Mechanical damage has been characterized by some of the most common causes of pipeline rupture [5]. If the mechanical strike does not immediately rupture the pipe, the formation of the dent or gouge will often result in local stress concentration and residual stresses. API 579 and ASME B31.8 provide guidance for the assessment of the dent or gouge on pipeline integrity [6,7]. An undented pipe does not develop any through-wall bending stresses when pressurized because of the smooth, axi-symmetric curvature of the wall. However, inward dents in a mechanically damaged pipe locally invert the nominal pipe wall curvature. Internal pressurization tends to evert the inward dent, developing locally high through-thickness bending stresses in the dent with the peak tensile stress in the root of the dent on the outer diameter (OD) surface [8,9]. In the absence of pressure-induced stresses, yielding of the pipe steel during inward indentation introduces residual bending stresses in the pipe wall that remain after the indenter is removed. These residual stresses, which can exceed yield-strength levels, are typically tensile in the root of the dent on the OD side and compressive on the inner diameter (ID) side [8]. Axial cracks are often observed in larger pipeline gouges, as shown in Figure 1.4. These mechanically induced cracks can provide sufficient stress intensity to initiate either progressive or overload cracking, depending on pipeline loading conditions. Metallographic examination of these dent-induced cracks indicates a characteristic 45° orientation with respect to the radial direction; examples are shown in Figure 1.5. These 45° shear cracks are formed by shear stresses caused during indenter-induced gouging and then eversion upon removal [5]. Finite element analysis can be conducted to calculate indentor forces required to cause the dent or gouge, and may be able to help determine which types of equipment may have caused the damage [8].

Figure 1.4 Example axial crack caused by mechanical damage.

Figure 1.5 Example shear cracks in a pipeline caused by mechanical damage.

On June 10, 1999, a 16-in. gasoline pipeline ruptured in Bellingham, Washington, and released over 5000 barrels of gasoline into Hannah Creek and eventually Whatcom creek. Tragically, two boys and a young man were killed as a result of the release. The pipeline was installed in 1964, had a nominal wall thickness of 0.312 in., and was made from electric-resistance weld (ERW) American Petroleum Institute (API) X-52 steel. Over the years following installation, several construction activities were conducted in the proximity of the leak site. The following reference provides a more detailed failure analysis investigation of this tragic event [10].

Visual inspection of the pipeline following removal from the accident scene indicated significant mechanical damage, consistent with repeated strikes from mechanical equipment, as shown in Figure 1.6. In all, over 30 dents and gouges were characterized on the subject pipe section. Visual analysis further indicated that the rupture initiated at a longitudinal gouge, shown by the ruptured section in Figure 1.6. The rupture occurred at approximately the 11 o’clock position (looking downstream), well away from the seam weld. A flat area approximately 3.5 in. in length was identified as the fracture origin (Figure 1.7). An area of roughly 45° fracture surface was observed along the pipe OD at the base of the gouge, within the origin area (Figure 1.7). Radial lines indicated that the fracture initiated along this angled area at the base of the gouge. The flat fracture (origin area) transitioned to ductile shear fracture (tearing) along the rest of the rupture, as indicated by the entire fracture thickness oriented along a 45° plane.

Figure 1.6 Photographic montage of the mechanical damage and subsequent failure of the 16-in. gasoline line in Bellingham, Washington.

Figure 1.7 Photograph of the fracture origin area.

Fractographic analysis using a scanning electron microscope (SEM) indicated the shear fracture area exhibited relatively featureless fracture morphology, consistent with corrosion damage over several years. Microvoid coalescence was observed at the base of the shear crack, which then transitioned into cleavage fracture morphology, shown in Figure 1.8. No evidence of progressive fracture morphology, such as fatigue or SCC, was observed. Metallographic cross-sections taken transverse to the pipe longitudinal axis through the rupture origin revealed the significant plastic deformation associated with the gouge and metal transfer from the impacting mechanical tool (Figure 1.9). Most importantly, the metallographic examination confirmed the existence of shear cracks associated with mechanical damage [5,8,9], both at and away from the rupture origin (Figure 1.10). The primary fracture initially followed a roughly 45° angle, and an adjacent, angled shear crack is also observed.

Figure 1.8 SEM image of the overall fracture surface (a) and a high magnification SEM image (b) showing microvoid coalescence at the transition between the preexisting shear crack and overload areas.

Figure 1.9 Metallographic image of the gouge surface, and fracture surface showing the preexisting shear crack and metal transfer from the tooling that struck the pipe (approx. 500 × mag).

Figure 1.10 Metallographic cross-section showing shear cracks in locations other than the primary fracture surface (approx. 25 × mag).

The subject rupture was caused when the stress intensity at a preexisting shear crack caused by mechanical damage exceeded the fracture toughness of the damaged pipe. Previous cyclic pressures were insufficient to initiate and grow a fatigue crack.

3 Longitudinal Seam-Weld Defects

Given their orientation with respect to hoop stresses and the orientation of preferential crack propagation (Figure 1.1), longitudinal seam-weld defects and discontinuities can greatly affect pipeline integrity. Since the advent of modern steel pipe in the 1930s [11], several different longitudinal seam-welding processes have been used: some of the more common processes include lap, flash, electric resistance, and submerged-arc welding (SAW). When properly completed, each of these processes result in good quality, high-strength welds. However, each of these welding processes can exhibit specific defects when done improperly. The following will describe specific defects and anomalies commonly associated with lap, flash, electric resistance, and submerged-arc weld seams.

3.1 Lap-Weld Defects

Some of the first processes used to make long seams in line pipe were the butt- and lap-weld processes, which were used prior to 1930 to make pipe up to 30 in. [11]. Both butt- and lap-welding were performed as batch processes, where flat plates 22-26 ft long were rolled into a tubular shape and then joined individually [11]. Later, a continuous process for making butt-welded pipe was developed, increasing efficiency [12]. These processes utilized a forging operation at high temperature to join the two sides of the pipe [12]. After the plate was formed into a tubular shape, it was reheated to approximately 1350 °C and forced between a die and a welding ball, which applied pressure to the two sides being joined [11]. While butt-welded joints are square, such that the joint area in the cross-section is equivalent to the wall thickness of the pipe, the edges of the pipe in lap-welded joints were angled to achieve a greater bonding area. The process of trimming the edges of the pipe to the specified angle is called scarfing.

Both butt- and lap-welded pipes are known to have had issues that resulted in decreased long-seam strength. Given the manufacturing similarities between butt- and lap-welded pipe, these issues are common to both types of pipe. For example, the welding, or forging is done at high temperature, around 1350 °C, which creates a relatively small operating window, given that the steel will begin to melt around 1455 °C [11]. If the forging temperature is too low, oxides can become entrapped in the joint area. Oxides in the joint provide a brittle region for relatively easy crack propagation in pipe long seams. Another known failure mechanism of these seam types is known as burning, where sulfides form on austenite grain boundaries due to excessive temperatures during processing [11]. Similar to oxides, sulfides that form on grain boundaries provide areas of weakness and relatively easy crack propagation.

3.2 Lap-Weld Case Study

A 12.5-in. nominal OD natural gas transmission pipeline ruptured during hydrotesting. The X-42 pipeline had a nominal wall thickness of 0.25 in. and was fabricated in 1947 with a lap-welded seam. The hydrotest rupture occurred along the seam at a pressure equal to 42% specified minimum yield strength (SMYS). The maximum allowable operating pressure of the line was 29% SMYS.

Visual inspection indicated little plastic deformation associated with the hydrotest rupture, as shown in Figure 1.11. The rupture ran approximately 8 ft along the weld seam with a maximum opening of roughly 1 in. (arrow, Figure 1.11). One half of the rupture was sectioned from the pipe for fracture surface analysis, shown in Figure 1.12. Given that the break sat for several weeks following the hydrotest, the fracture surface was scrubbed with detergent-based solution and a nylon brush. The cleaned rupture surface displayed a mix of dark oxide, red oxide, and shiny areas, shown in the bottom image of Figure 1.12.

Figure 1.11 An example is shown of a pipe which ruptured due to a lap-weld defect. Little plastic deformation is shown.

Figure 1.12 One side of the fracture surface was extracted from the main body of the pipe for analysis. The piece is shown (top), as well as a higher magnification view of the fracture surface (bottom).

SEM-Energy Dispersive Spectroscopy (EDS)-based examination of the lap-weld rupture surface confirmed that the darker-colored fracture surface locations were covered in oxide. The fracture surface regions that appeared to be shiny during visual and optical inspection exhibited little ductility, and no evidence of ductile morphology, shown in Figure 1.13a and b. Fractographic inspection revealed cubic-shaped pits along the fracture surface, shown in Figure 1.13c and d. EDS analysis of these pits indicated high levels of manganese and oxygen, consistent with manganese oxides.

Figure 1.13 SEM images of the fracture surface showing the presence of both pit and brittle fracture in (a) and (b). The fracture surface is shown again in (c) and (d) where the presence of pits is shown, many of which are shown to have sharp and/or square edges.

Metallographic specimens were cut from intact and fractured portions of the subject lap weld. These metallographic specimens were cut, polished, and etched to reveal microstructural features transverse to the longitudinal weld, shown in Figure 1.14. Areas of oxide were observed along broken weld surfaces, shown in Figure 1.15. Similar areas of oxide as well as lack-of-fusion were observed along intact weld surfaces. Relatively small, cubic-shaped, inclusions were observed within close proximity to the bond line of intact and broken lap welds. EDS analysis of these inclusions indicated that they contained elevated levels of manganese and oxygen, consistent with the manganese oxide containing pits on the fracture surfaces in Figure 1.13. Microhardness testing showed minor hardness variation along the weld seams.

Figure 1.14 Transverse metallographic mounts of the long-seam lap weld at three different locations along the length of pipe.

Figure 1.15 Metallographic cross-sections of the lap weld showing the presence of oxides in the lap-weld area.

The rupture occurred due to incomplete fusion along the weld seam that was created during original manufacture of the pipe. Planar oxide deposits were observed along fractured and intact portions of the subject lap weld. Oxide entrapment during lap-welding is associated with poor temperature control [11]. If welding temperatures were insufficient to melt the oxide, it can remain along the bondline and result in a weak weld. There was no indication that the subject pipe had exhibited any sort of leak over its 60 + years of service, and no evidence of progressive cracking, such as fatigue or SCC, was observed. The rupture only occurred when the internal pressure exceeded a critical value. The subject pipe had not likely ever been subjected to a pressure that exceeded 42% SMYS over its lifetime.

3.3 ERW Defects

ERW pipe is relatively common in the United States, as the basic process has been in place since the 1920s [13]. Despite being developed nearly 100 years ago, the ERW process is still used today in modern pipe manufacturing. The ERW process involves forming the pipe section from a flat plate, and the creation of a longitudinal weld seam by a combination of heating using electrical resistance and force (upset) between the two surfaces to be welded. Like butt and lap welds, the ERW process is autogenous, meaning it does not require additional weld or filler metal. ERW pipe made prior to the early 1970s was typically made by either using a direct current (DC) or low-frequency AC process (<360 cycles/s, [13]), while post-early 1970s ERW pipe has been typically fabricated using a high-frequency process.

Low-frequency ERW pipe has generally been more susceptible to weld seam issues than high-frequency ERW pipe [11,13]. Weld seam issues that can affect ERW pipe include, but are not limited to: lack-of-fusion (cold welds), stitching, hook cracks, pinholes, and inclusions [11,13–15].

One of the common defect types in low-frequency ERW welds is lack-of-fusion. Lack-of-fusion in ERW pipe results in unbonded areas along the weld seam. Lack-of-fusion can be caused by either insufficient upsetting force, or the application of inadequate electric current during the welding process. Inadequate current results in insufficient heating, i.e., cold welds [13]. Fractographic examination of ERW lack-of-fusion areas typically shows the presence of dark oxides along the weld surface, as shown in Figure 1.16. Metallographic examination of ERW lack-of-fusion shows either incomplete upset, evidenced by lack of microstructural deformation, as shown in Figure 1.17, or incomplete heating (from interrupted welding current), shown by diffuse heat-affected zones (in non-post weld heat-treated welds). Pinholes are a series of short unwelded areas along the long seam which can cause leaks in the gas or liquid line.

Figure 1.16 Example fracture surface of an ERW pipe, which experienced lack-of-fusion, as evidenced by the dark oxide.

Figure 1.17 Example metallographic cross-section of an ERW weld, which exhibited lack-of-fusion. Little plastic deformation is observed in the weld area, in what should be the upset region of the weld where the metal structure would normally flow toward the outer and inner diameter.

Another defect type found in ERW welds includes stitching, which is a variation in properties and brittleness along the long seam caused by variations in heat input as the weld progresses along the pipe. Stitching is only visible when fracture occurs along the bond line [11].

Hook cracks in ERW welds are formed when manganese sulfide stringers are deformed along with the microstructure during the ERW upset process [5]. Manganese sulfide stringers are commonly found in plate steels and are elongated along the rolling direction of the plate during the steel making process. During the upset process in ERW welding, the stringers, normally elongated along the length of the pipe, deform toward the OD and ID. Dissolution of the manganese sulfide stringer during welding can result in the formation of a hook crack.

3.4 Flash Weld Defects

Flash welds share many similarities with ERW welds; they are both autogenous welds formed by a forging process, i.e., a combination of mechanical force and heating by electrical resistance. Flash welds specifically were done as a batch process on individual pieces of pipe, whereas the ERW process is often performed as a continuous process [13]. The flash-welding process was only performed between the years of 1930 and 1969 [11].

Given the similarities between flash welds and ERW welds, they can be susceptible to many of the same types of defects. For instance, flash welds can also show hook cracks and lack-of-fusion [4,11,13,16]. Similar to ERW welds, flash welds can also be susceptible to selective seam (grooving) corrosion [16]. The same plate stock or skelp was often used for both ERW and flash welds, so the resulting pipes had the same chemistry and were therefore similarly susceptible to manganese sulfide formation and/or selective seam corrosion. As will be shown later, flash-welded pipe from specific manufacturers has been known to have a propensity for hard spots—which can result in integrity concerns due to susceptibility to hydrogen-stress cracking (HSC), although these failures are unrelated to the long-seam welds.

However, it should be noted that although flash welds can exhibit similar defects as ERW seams, the problems seen in the early ERW seams have not found to be as prevalent in the flash welds. This is attributed to the fact that flash-welded pipe was typically cold-expanded and mill-tested, which likely caught many of the more significant defects prior to the pipe being placed into service [16]. Flash welds have a 1.0 joint factor [16].

3.5 Submerged-Arc Weld Defects

SAW was first performed on line pipe in 1930 [11]. It is a process whereby the edges of the plate are brought together and arc welded using a consumable electrode and flux. Unlike ERW and flash welds, heating occurs with an electric arc between an electrode and the metal pipe [12]. Also unlike lap, ERW or flash welds (which are autogenous), SAW uses added filler metal. The arc and weld pool are submerged in a flux to protect the molten weld metal from the atmosphere where it may otherwise pick up extraneous gasses and result in nonmetallic inclusions or porosity. The SAW process is not a forging process like ERW, and no mechanical pressure is applied during the welding process. The first submerged-arc welded pipes were welded on the OD only and are known as single submerged-arc welds. Around 1948, Consolidated Western first developed the double submerged-arc welding process, (DSAW), where the pipe was welded on both OD and ID [11]. The SAW process is versatile, as many different pipe diameters and wall thicknesses can be joined [12]. The SAW process and DSAW specifically remain one of the most common methods of pipe manufacture today [11].

Although SAW is generally thought to be a robust process, it can suffer from many of the same welding issues as other processes such as: (not limited to) lack-of-fusion, inadequate penetration, porosity, and undercut. In addition to these somewhat common problems, submerged-arc welded pipes can also exhibit cracking in the weld metal, which is caused by the movement of plates prior to solidification of the weld metal [11], and toe cracks, which can form at the root, or toe, of the weld either on the ID or on the OD. Toe cracks can be caused by cold expansion after the welding process when the pipe is out of round, or if the weld bead is excessive causing locally high stresses at the toe of the weld as the pipe is forced into a round shape [11,13]. Toe cracks, along with inadequate penetration and weld cracks (solidification cracks), are the most common failure mechanisms in submerged-arc welds

You've reached the end of this preview. Sign up to read more!
Page 1 of 1

Reviews

What people think about Handbook of Materials Failure Analysis with Case Studies from the Oil and Gas Industry

3.0
1 ratings / 0 Reviews
What did you think?
Rating: 0 out of 5 stars

Reader reviews