Case Studies in Engineering Failure Analysis 9 (2017) 63–70

Contents lists available at ScienceDirect

Case Studies in Engineering Failure Analysis

journal homepage: www.elsevier.com/locate/csefa

Short communication

Failure analysis of leakage caused by perforation in an L415 steel

T
gas pipeline
⁎
Zhiyong Liua, , Wenju Liaoa, Wei Wua, Cuiwei Dua, Xiaogang Lia,b
a
Corrosion and Protection Center, University of Science and Technology Beijing, Beijing 100083, PR China
b
Ningbo Institute of Material Technology & Engineering, Chinese Academy of Sciences, Ningbo 315201, Zhejiang, PR China

AR TI CLE I NF O AB S T R A CT

Keywords: The reasons for the failure of a buried pipeline perforated during construction were investigated
L415steel by a chemical composition analysis; a metallographic test; macromorphology observation;
Buried pipeline characterization of the corrosion products by scanning electron microscopy, energy-dispersive X-
Pitting ray spectroscopy, and X-ray diffraction; field medium characterization; and an analysis of the
Perforation
working conditions. The results revealed that the material composition and organization of the
steel pipe conformed to API Specification 5CT. However, the reason for the perforation of the
L415 steel pipe was an ultrahigh growth rate of pitting corrosion, as high as 14 mm per year. We
confirmed that the synergistic effect of a high partial pressure of oxygen introduced by an im-
proper packing process and concentrated Cl− in the corrosion product layer, which originated
from groundwater with a high salt concentration that was used for the water pressure test, were
responsible for the failure process.

1. Introduction

With the rapid growth of demand for petroleum and gas, an increasing number of underground pipelines have been constructed to
transport oil and natural gas resources [1]. However, accidents caused by the leakage of pipelines occur frequently, leading to serious
social consequences and economic losses [2–4]. Corrosion damage is one of the main threats to the safety and durability of pipelines
[5]; as many as 33% [6] to 51% [7] of oil and/or gas pipeline failures have been attributed to corrosion. In the case of internal
corrosion, extensive research focusing on the effects of aggressive media and various factors on the corrosion modes and their
behaviors has been reported. Liu et al. [8] researched the corrosion behavior of two high-strength low-alloy steels in pressurized NaCl
solution and found that the corrosion types of both steels changed from uniform corrosion to localized corrosion with increasing
pressure. De Waard et al. [9] investigated the relationship between CO2 partial pressure and pH under aqueous conditions with high
CO2 pressures and found that an increase of the CO2 partial pressure reduces the pH and thus accelerates the corrosion of materials.
Kermaniet al. [10] further revealed the existence of a threshold pressure value of approximately 0.02 MPa, above which pitting and
serious localized corrosion rather than uniform corrosion will occur to various degrees with a shift in the CO2 partial pressure.
Previous studies have also shown that when sufficient oxygen is present, the corrosion products mainly comprise low-porosity
FeOOH, which provide a strong protective effect for metal electrodes [11]. However, when insufficient oxygen is present, a large
amount of Fe3O4 is generated on the specimens and is easily separated from the substrate, thereby making the original uniform oxide
film porous and greatly reducing the protective effectiveness of the film [12]. Thus, the pressure and the oxygen content have been
demonstrated to strongly influence the corrosion rates of materials [13–16]. However, the literature contains few reports on the effect

⁎
Corresponding author.
E-mail address: liuzhiyong7804@ustb.edu.cn (Z. Liu).

http://dx.doi.org/10.1016/j.csefa.2017.07.003
Received 13 March 2017; Received in revised form 9 June 2017; Accepted 31 July 2017
Available online 05 August 2017
2213-2902/ © 2017 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
Z. Liu et al. Case Studies in Engineering Failure Analysis 9 (2017) 63–70

Table 1
Results of chemical analysis (wt.%).

Element C Si Mn P S V Mo Cr Ni Fe

Substrate 0.10 0.22 1.22 0.012 0.0033 0.015 ≤0.01 ≤0.01 0.032 balance

of the partial pressure oxygen on the localized corrosion of pipe steels under various conditions.
The failed natural gas pipeline in this case is approximately 104 km in length, with dimensions of ϕ508 × 7.1 mm2. The material
is L415 steel. According to onsite construction records, the pressure test, sweeping water, and packing work were completed in turn
after the pipeline was completed in July 2012. However, the water for pressure test was pumped from a coastal river nearby, which
contains higher concentration of corrosive specises, such as choloride ion, etc. After sweeping water, the inner pipe was not com-
pletely in a dry state, wet rust layer existed until the failure. Afterword, compressed air as high as 4.5 MPa instead of nitrogen gas as
the pressure medium had been used in the packing process for three months. The above procedures were only 7 months in total, and
then the pipeline leak was detected. The rate of perforation was as high as 14 mm/a; this phenomenon is rare in corrosion cases.
According to preliminary analyses this pipeline, rapid perforation is caused by a high-pressure oxygen-enriched atmosphere, which
considerably accelerates pitting corrosion inside the pipeline [17,18].
In this study, we systematically analyzed the characteristics and causes of failure of the aforementioned pipeline in a high-
pressure, oxygen-enriched, humid environment. The results not only provide a basis for the prevention of similar accidents but also
give us deep insights into the mechanism by which high-pressure O2 affects localized corrosion.

2. Analytical method

2.1. Analysis of substrate

A small amount of iron chips was drilled from the substrate after the anti-corrosion coating had been removed and the substrate
had been cleaned with acetone, distilled water, and dried in air. The chemical composition of these chips was analyzed; the results are
presented in Table 1. Microstructural tests of the substrate were carried out according to Chinese Standard GB/T13298. All specimens
were wet ground to a 300 grit finish, polished with 0.5-μm diamond polishing paste, and eroded with a nitric acid alcohol solution
after mechanical polishing. They were then rinsed with distilled water, dried in air, and observed under a stereomicroscope.

2.2. Analysis of corrosion characteristics

The macroscopic and microscopic corrosion morphologies of the failed pipeline after removal were observed using a digital
camera and a scanning electron microscope, respectively. The corrosion products were characterized by energy-dispersive X-ray
spectroscopy (EDS) and X-ray diffraction (XRD). Across-section of the rust layer was analyzed using a line sweep to determine the
distribution of elements in the rust layer. The corroded surfaces of the pipeline were examined by scanning electron microscopy
(SEM) after the corrosion products were removed using a descaling solution (a mixture of 500 mL of concentrated hydrochloric acid,
500 mL of deionized water, and 6–10 g of hexamethylenetetramine). To study the development of a perforation in the pipeline, cross-
sections of the perforation area were observed. Corrosion products in pits were also analyzed by EDS.
The characteristics and cause of failure are discussed together with the corrosion characteristics, surrounding environmental
factors, and working conditions of the pipeline.

3. Results

3.1. Microstructure

The microstructure of the substrate was composed of ferrite and pearlite (Fig. 1). The transverse and longitudinal structures were
essentially consistent. The grain size was fine and evenly distributed.

3.2. Corrosion morphology observation

The macroscopic corrosion morphologies of the pipeline are presented in Fig. 2. The anti-corrosion coating was intact, and a few
damage points resulting from the leakage of pressurized gas were observed at the 7-8 o'clock directions of the pipeline (Fig. 2a). No
obvious corrosion was observed on the outer wall of the pipeline after the anti-corrosion coating was removed. After the pipeline was
opened, the inner wall of the pipe exhibited a moisture condition. The rust layer on the bottom of the pipe was very damp; in contrast,
the top of the inner pipe was smooth and dry, without heavy rust (Fig. 2b). The rust layer at the foot of the inner pipe was moister
than that at the top surface, and a large number of scrap-like product layers were observed in addition to the uniform rust layer
(Fig. 2c). In addition, extensive pitting corrosion was observed under scraps on the inner wall (Fig. 2d).
The scrap-like corroded surfaces were examined by SEM (Fig. 3). The surface was covered by a thick oxide film with cracks on top
of the bulges (Fig. 3a). The cracks may have formed because of the high-pressure,oxygen-enriched atmosphere, which markedly

64
Z. Liu et al. Case Studies in Engineering Failure Analysis 9 (2017) 63–70

Fig. 1. Microstructure of the matrix in (a) base metal, (b) weld beam and (c) heat affected zone.

Fig. 2. Macro-morphology of the wall of the pipeline, (a) external coating at a perforation, (b) internal morphology of rust layer within 9–3 o’clock, (c) internal
morphology of rust layer within 4–8 o’clock, (d) after removed the scrap-like rust layer at the position as marked in (c).

65
Z. Liu et al. Case Studies in Engineering Failure Analysis 9 (2017) 63–70

Fig. 3. Micromorphology of the inner wall of the pipeline (a: the top of drum package; b: the bottom of drum package; c: one of the leaking perforation).

accelerated the corrosion rate as well as the internal stress generated between corrosion products because of rapid accumulation.
These cracks provided channels for the corrosion medium to penetrate to the metal surface and increase the possibility of localized
corrosion underneath. After removing the coverings, we found numerous large pits underneath; the pits were filled with loose rust
(Fig. 3b). One of the perforation-caused leaks is shown in Fig. 3c. Given that the time before the leakage occurred was short, the pits
developed rapidly after nucleation in the high-pressure, oxygen-enriched atmosphere.
The predominant corrosion feature observed after the corrosion products were removed is shown in Fig. 4. The corrosion mor-
phology was characterized by comprehensive corrosion with a high density of small pits and a much lower density of large per-
forations (maximum size of 3 mm, Fig. 4a). A large number of tiny pits were observed at the bottom of the pits (Fig. 4c), which
indicates a tendency for the development of deep pits.

3.3. Analysis of corrosion products

The results of chemical analysis of the corrosion products on the perforation, i.e., the scrap-like scale, are shown in Fig. 5. The
corrosion products mainly contained iron, oxygen, chlorine, and other alloying elements. The chlorine content of the corrosion
products was 0.51 wt.%. To analyze the composition in the corrosion products, we performed XRD analysis after scraping the rust
layer from the inner wall of the pipe and milling it in a mortar. The diffraction pattern is presented in Fig. 6, which shows that the rust
was mainly composed of Fe3O4 and Fe2O3, although α-FeOOH was also detected. These components are very similar to those formed
in moist air on mild steel, which indicates that the oxygen-enriched atmosphere in the pipe did not greatly shift the corrosion
mechanism but might have substantially accelerated the cathodic processes and the rust formation processes [18].
To determine the distribution of the elements in the rust layer, we used EDS to analyze across-section of the rust layer; the results
are illustrated in Fig. 7. The cross-sectional morphology of the rust layer on the inner wall was relatively dense and thick, with no
substantial delamination. Moreover, pits were observed on the metal surface. The area containing pits was selected for line-scanning
analysis, and the results show that the iron and oxygen contents in the rust layer were high. Thus, the corrosion products were mainly
iron oxides. Meanwhile, elemental chlorine was concentrated on both outer and inner surfaces of the rust layer.
According to the aforementioned results, the rust layer on the inner wall was relatively dense and thick, providing a protective
effect for the pipe by slowing the corrosion of metals. However, iron oxide could easily be damaged by chloride [17], and cracks on
the top of the corrosion products could provide channels for the medium to reach the metal surface. Therefore, pits could rapidly
nucleate and develop on the inner wall, resulting in perforation failure.

Fig. 4. Micromorphology of the inner wall of pipeline after the removal of the corrosion products: (c) is Partial view of (b).

66
Z. Liu et al. Case Studies in Engineering Failure Analysis 9 (2017) 63–70

Fig. 5. EDS spectra of corrosion products on the inner wall of the pipeline.

Fig. 6. XRD spectra of the corrosion products.

3.4. Cross-sectional analysis of pits

To study the development of pitting on the inner wall, we observed a cross-section of pits by SEM and analyzed the elemental
composition of corrosion products at the bottom of the pipe by EDS. The SEM and EDS results (Fig. 8) reveal that pits formed in an
elongated strip with a substantial vertical trend; the pit depth was 4–5 mm, and the width was two to three times greater than the
depth. The chemical analysis results show that chlorine accumulated at the bottom of the pipe,which was approximately 25 times
greater than that in the corrosion products on the surface. This enrichment of chlorine promoted the development of pits, which led to
pipe failure by perforation.

67
Z. Liu et al. Case Studies in Engineering Failure Analysis 9 (2017) 63–70

Fig. 7. Results of cross-sectional analysis of rust layer.

Fig. 8. SEM image of a crosssection of a pit and EDS spectra of corrosion products at the bottom.

3.5. Analysis of corrosion medium

After cutting the pipeline specimen, we observed that the inner wall was relatively humid. To obtain the chemical composition of
the water film, we cut a specimen of 100 × 130 mm2 from an typical corrosion region to characterize the products on its surface after
the specimen was cleaned with acetone. The corrosion products were milled in a mortar, vibrated in an ultrasonic cleaner for 30 min,
and soaked in 50 mL of deionized water. The filtered water samples were analyzed for ion composition; the results are reported in
Table 2 as detected values. The ion concentration in the aforementioned water film was calculated with consideration of the specimen
size and rust-layer thickness (measured as shown in Fig. 6, approximately 350 μm); the result is reported in Table 2 as the estimated
value, which represents the average content of ions in the rust layer. However, the concentration of Cl− in the pits was observed to be
much higher than that in the rust layer; the concentration of Cl− in the pits exceeded the critical pitting value and thereby accelerated
the pitting.

Table 2
Results of leaching solution of corrosion products.

Ion F− (mg/L) Cl− (mg/L) SO42− (mg/L) NO3− (mg/L) Na+ (mg/L) K+ (mg/L) Ca2+ (mg/L) Mg2+ (mg/L)

Detected value 0.6493 4.995 18.5654 0.4292 42.6956 2.625 3.4358 1.5326
Estimated value 7.135 54.890 204.015 4.716 469.182 28.846 37.756 16.842

68
Z. Liu et al. Case Studies in Engineering Failure Analysis 9 (2017) 63–70

4. Discussion

During the construction of the pipe, groundwater was wrongly used as the pressure water, which led to serious contamination of
chloride and other salts in the pipeline. In addition, the groundwater level of the area where failure occurred was high and the
temperature in the pipeline was low, allowing steam to easily condense in the rust layer. Moreover, the medium of the packing
process was air and the pressure was 4.5 MPa, resulting in an oxygen partial pressure in the pipe as high as 0.9 MPa. Therefore, a
high-pressure, oxygen-enriched, humid environment formed in the pipeline,which provided conditions that favored rapid corrosion.
In this case, the reaction was mainly oxygen absorption into the inner wall [19]:

Fe-2e → Fe2+

O2 + H2O + 4e → 4OH−

Fe2+ generated on the anode combined with OH−generated on the cathode:

Fe2+ + 2OH− → Fe(OH)2

Because of the high oxygen partial pressure in the pipeline, Fe(OH)2 continued to oxidize:

Fe(OH)2 + 1/2O2 + H2O → Fe(OH)3

Fe(OH)3 → FeOOH + H2O

8FeOOH + Fe2+ + 2e → 3Fe3O4 + 4H2O

The corrosion products were further oxidized to Fe2O3,resulting in the formation of a brown corrosion-product film on the inner
wall (Figs. 2 and 6).
The above procedures generate continuous corrosion product layer on the metal surface, providing a measure of protection for the
pipe wall. However, the existence of cracks (Fig. 3a) and the accumulation of Cl− at the 3–9 o'clock directions of the pipeline, which
could damage the rust layer, allowed the medium, including oxygen, to penetrate to the metal surface and form pits [20–23].
Within the area of pitting nucleation, a corrosion cell comprising a large cathode and a small anode formed between the surface
and the bottom of the pits. The bottom was the anode which substantially increased the concentration of Fe2+. Fe2+continued to
hydrolyze as

Fe2+ + H2O → FeOH+ + H+

Additionally, because of more Fe3+ sites generate in active points under the rust layer, Cl− will diffuse from the surroundings to
continuously concentrate on local metal surface. Thus, the Cl− concentration at the bottom of the pits increased substantially (Fig. 8).
This promotes the anodic reaction of local active sites and the hydrolysis of Fe3+, which lead to the further acidification of the media
of active sites:
Fe3++ Cl-+ OH- ↔ FeOCl + H+
The accumulation of Cl− acidified the solution and accelerated the electrochemical dissolution of metal [24–27]. Thanks to the
hydrolysis reaction and Cl−, the bottom of the pits keep in a state of activation in high corrosion rate.
According to all processes above, there will be a stronger autocatalytic phenomenon in pits. On one hand, in the bottom of pits,
the sum of the above reaction synthesis further promotes the anodic reaction of local active sites:
Fe +2Fe3+ + 6OH- → 3Fe(OH)2

Fe +2H+ → Fe2 + + 2Had → H2 ↑ .

On the other hand, when the corroded Fe2+ sites transported from the bottom to the oxygen-arrival positions in the pits, they
were then more flexible to be oxidized and further to form hydrolysis of Fe3+ to accelerate the dissolution there (Fig. 2c). Thus,
because the oxygen content is much higher than that in air, the equilibrium current is much higher than that under normal oxygen-
content environment. The above processes indicate that the increase of oxygen concentration will greatly promote the corrosion rate
of pits.
In summary, the high-pressure, oxygen-enriched environment considerably accelerated the corrosion of the pipeline, forming a
thick corrosion product film. However, the formation of cracks and damage to the rust layer caused by Cl− deposition provided a
path for the medium to reach and remain on the metal surface (Fig. 3), thereby causing pitting. The rate of development was much
higher than the rate of nucleation, resulting in the formation of a long strip of pits inside the material and ultimately leading to
perforation.

5. Conclusions and Recommendations

• The air used as the medium in the packing process and the high underground water level formed a high-pressure, oxygen-enriched
atmosphere in the pipeline. The steam in air condensed into a thin film attached to the inner wall of the pipe, and Cl− was
concentrated at the 3–9 o'clock directions of the pipeline. These events caused serious pitting corrosion.

69
Z. Liu et al. Case Studies in Engineering Failure Analysis 9 (2017) 63–70

• Given the influence of high pressure, oxygen enrichment, and Cl −

concentration on the bottom of the pits, the rate of pitting
development was much higher than the rate of nucleation. Pits rapidly developed into long strips and then led to perforation
failure; the perforation rate reached as high as 14 mm/a.
• It is that the construction process is not in accordance with the relevant engineering standards caused the accident, which
confirmed that the importance of complying with the construction code cannot be overemphasized.
• Once corrosive pollution occurred in the pipe, timely cleaning and drying must be done as soon as possible, and additionally,
humidity or dew point temperature should be detected and controlled in the subsequent use of the pipe.

Acknowledgements

This work was supported by the National Nature Science Foundation of China (No. 51471034) and China Scholarship Council
(CSC) (No. 201606465051).

References

[1] Pacheco JL, Sisak W. Advances in wet gas pipeline technology. International Petroleum Technology Conference. 2007.
[2] Li MC, Cheng YF. Corrosion of the stressed pipe steel in carbonate-bicarbonate solution studied by scanning localized electrochemical impedance spectroscopy.
Electrochim Acta 2008;53:2831–6.
[3] Hardie D, Charles EA, Lopez AH. Hydrogen embrittlement of high strength pipeline steels. Corros Sci 2006;48:4378–85.
[4] King F, Jack TR, Chen W, Wang S, Elboujdaini M, Revie W, Worthingham R, Dusek P, Corrosion’ 2001, NACE, Houston, 2001 (paper no. 1214).
[5] Li XG, Zhang DW, Liu ZY. Materials science: share corrosion data. Nature 2015;527:441–2.
[6] Stanley RK. An analysis of failure in coiled tubing. IADC/SPE Drilling Conference. 1998.
[7] Crabtree AR. CT-Failure monitoring:a decade of experience. SPE/ICOTA coiled tubing and well inter vent ion conference and exhibition. 2008.
[8] Liu J, Li XB, Wang J. Effect of hydrostatic pressure on the corrosion behaviors of two low alloy steels. Acta Metall Sin 2011;6:697–705. in Chinese.
[9] Waard DC, Lotz U, Dugstad A. Influence of liquid flow velocity on CO2 corrosion. Corrosion/95. 1995.
[10] Kermani MB, Smith LM. Corrosion control in oil and gas production design considerations. the Institute of Materials.
[11] Park JJ, Pyun SI, Na KH, et al. Effect of passivity of the oxide film on low-pH stress corrosion cracking of API5L X65 pipeline steel in bicarbonate solution.
Corrosion 2002;4:329–36.
[12] Li YJ, Wu JJ, Zhang D, et al. The electrochemical reduction reaction of dissolved oxygen on Q235 carbon steel in alkaline solution containing chloride ions. J
Solid State Electrochem 2010;9:1667–73.
[13] Xiao RJ. Corrosion failure cause analysis and evaluation of corrosion inhibitors of Ma Huining oil pipeline. Eng Fail Anal 2016;68:113–21.
[14] Grincova A. Failure analysis of conveyor belt in terms of impact loading by means of the damping coefficient. Eng Fail Anal 2016;68:210–21.
[15] Bitterlin M. Cracking mechanisms in large size ingots of high nickel content low alloyed steel. Eng Fail Anal 2016;68:122–31.
[16] Rahmani K. Evaluation of inhibitors and biocides on the corrosion, scaling and biofouling control of carbon steel and copper-nickel alloys in a power plant
cooling water system. Desalination 2016;393:174–85.
[17] Meng G. Effect of Cl− on the properties of the passive films formed on 316 l stainless steel in acidic solution. J Mater Sci Technol 2014;30:253–8.
[18] Zhao YF. Failure analysis of a natural gas pipeline. Eng Fail Anal 2016;63:61–71.
[19] Cao CN. Theory for corrosion electrochemistry. first ed. Beijing: Chemical Industry Press.
[20] Szklarska-Smialowska Z. Pitting corrosion of metals. TX: NACE International Houston.
[21] Liu JC, Park SW, Nagao S, Nogi M, Koga H. The role of Zn precipitates and Cl- anions in pitting corrosion of Sn-Zn solder alloys. Corros Sci 2015;92:263–71.
[22] Liu ZY, Li XG, Cheng YF. Understand the occurrence of pitting corrosion of pipeline carbon steel under cathodic polarization. Electrochim Acta 2012;60:259–63.
[23] Nazeri MFM, Ismail AB, Mohamad AA. Effect of polarizations on Sn-Zn solders alloys in alkaline electrolyte. J Alloy Compd 2014.
[24] Moayed MH, Laycock NJ, Newman RC. Dependence of the critical pitting temperature on surface roughness. Corros Sci 2003;45:1203–16.
[25] Zhu LK, Yan Y, Qiao LJ, Volinsky AA. Stainless steel pitting and early-stage stress corrosion cracking under ultra-low elastic load. Corros Sci 2013;77:360–8.
[26] Chen W, Boven GV, Rogge R. The role of residual stress in neutral pH stress corrosion crackingof pipeline steels. Part I: pitting and cracking occurrence. Acta
Mater 2007;55:29–42.
[27] Frankel GS, Sridhar N. Understanding localized corrosion. Mater Today 2008;11:38–44.

70