Scouring pigging of oil and gas pipelines is used to maintain flow efficiency by abrading the inner surface to remove precipitates such as wax and scale. There has been uncertainty over whether pigging of corrosion resistant alloy pipelines used in sour service could adversely affect their resistance to localized corrosion and stress corrosion cracking (SCC). In this work, a novel environmental chamber was developed to enable scouring pigging to be simulated on pipeline steels while under tensile stress and exposed to deoxygenated brine. Using this configuration, representative pigging damage was reproduced on Type 316L stainless steel cladding with both ground and as-received finish prior to four-point bend testing in a range of environments. Under conditions in which the alloy is considered to be resistant to SCC, simulated pigging had no adverse effect on the resistance of the alloy to localized corrosion or cracking. Under more severe test conditions, in which the alloy is considered to be susceptible to SCC, specimens with an as-received surface experienced preferential pitting and cracking located at pigging scratches, whereas no such effect was observed for ground specimens. Electrochemical impedance spectroscopy measurements demonstrated that the abraded surface had a thinner passive film and a lower corrosion resistance than the as-received surface. However, this detrimental effect of surface abrasion, which is associated with the removal of the pickled and passivated surface, is taken into account by the conservatism of SCC test programs, in which ground specimens are typically included. These findings support the perspective that under normal operating conditions scouring pigging is not expected to increase the risk of pitting or SCC.

Pipelines used in the production and transportation of oil and gas are susceptible to the formation of internal precipitates, such as wax and scale, which deposit on the inner surfaces. While for carbon steels there is often a beneficial effect of scaling on their corrosion resistance, for corrosion resistant alloys (CRAs) these deposits provide no additional protection and can have deleterious effects on both production efficiency and production capacity. In suitable pipelines, this problem is addressed by using a pig to scour the inner surface with an array of steel bristles or studs.1-2  This is a mechanically aggressive process and visual inspection has confirmed that it leads to abrasive wear of the pipe surface. In this respect, scouring pigging can be considered a form of surface treatment, and consequently assessment of its potential to influence the susceptibility of CRAs to localized corrosion and cracking is of particular importance to the oil and gas industry.

The effect of surface condition on the stress corrosion cracking (SCC) and pitting resistance of CRAs has been the focus of many papers over the past few decades.3-11  It is well established that surface finish, near-surface residual stress, and the presence of microstructural defects and surface inhomogeneities can all influence the resistance of CRAs to SCC. Surface inhomogeneities such as mechanical defects and inclusions are favorable sites for pit initiation and have a dominant influence over pitting resistance when compared with more commonly used metrics such as the average surface roughness (Ra).5-7  Studies of the effect of residual surface stress on crack initiation/propagation have established that, in general, processes that induce residual tensile stresses tend to lower a material’s resistance to cracking, while those that place the surface in a state of compressive stress, e.g., shot peening, can improve it.6,8-10 

Given that scouring pigging affects both the surface topography and the near-surface residual stress state of a pipeline, determining the combined effect on SCC resistance is a complex problem. In a preliminary study, the effect of hard wire pigging on the SCC resistance of super 13 Cr and duplex stainless steels was investigated and it was shown that, for ground specimens manufactured from girth-welded tubulars, simulated pigging damage did not enhance the susceptibility to SCC.11  Furthermore, results showed that crack initiation and propagation on super duplex stainless steel was inhibited by the presence of pigging scratches, owing to the presence of local compressive residual stresses. The question was then raised as to whether these observations would hold true for austenitic alloys, in particular Type 316L stainless steel (SS, UNS S31603(1)) cladding, which is often used in combination with carbon steel because of its significantly lower cost compared to a whole-pipe CRA solution. The rationale here is that the hot roll bonding process used to produce most clad pipelines may alter the surface properties significantly when compared to those of seamless pipe. A further consideration is the surface state of the material, as testing of an as-received internal surface (rather than a ground specimen) would be more representative of the pigging process in service. In this work the effect of scouring pigging on the SCC resistance of Type 316L SS cladding, in both the ground and as-received state, is investigated using a novel experimental procedure for simulation of pigging scratches in more representative oilfield environments.

Materials and Specimens

Material was supplied by BP in the form of longitudinally-welded carbon steel pipe with Type 316L SS cladding, the chemical composition of which is shown in Table 1. Four-point bend specimens were manufactured from the cladding by cutting 120 mm × 18 mm strips and grinding away the carbon steel backing material, thereby reducing the thickness of the specimens to approximately 2.5 mm. The copper sulfate test12  was used to confirm that all traces of the carbon steel backing material had been removed. This manufacturing process resulted in specimens with one as-received surface and one machined surface. Herein, “as-received specimens” refers to those in which four-point bending placed the as-received surface in tension. “Ground specimens” were produced by grinding the machined surface of the specimens to a 600 grit (P1200) finish using silicon carbide paper; these specimens were loaded in four-point bending to place the finished surface in tension. Finally, all specimens were degreased in acetone in an ultrasonic bath for 30 min, rinsed in deionized water, and dried in air prior to use.

TABLE 1

Chemical Composition of the Type 316L SS Cladding Material

Chemical Composition of the Type 316L SS Cladding Material
Chemical Composition of the Type 316L SS Cladding Material

Simulated Pigging

Specimens were flexed to a constant displacement in air to a load at room temperature equivalent to 100% σ0.2 at the test temperature. The exact procedure for achieving this has been described previously, and further detail is not provided here.11  To simulate the damage sustained during pigging, a novel configuration was designed that allowed a conical stainless steel stylus to be rastered across the surface of a specimen while it was loaded in a four-point bend jig and exposed to typical conditions that a pipeline may be exposed to during pigging; a schematic of this setup is shown in Figure 1. To achieve this, simulated pigging was performed on specimens while submerged in a solution containing 50,000 ppm chloride and buffered to either pH 4.0 or pH 4.5. The solution was saturated with 1 bar (100 kPa) CO2 at ambient temperature with dissolved oxygen levels maintained at <10 ppb. Deaeration of the solution took 3 h to achieve and was validated by measurement of the dissolved oxygen concentration using an Orbisphere optical oxygen sensor accurate to 0.5 ppb. All scratches were produced by a single pass of the stylus across the specimen surface at a nominal load of 100 N, which was the maximum load of the scratch tester used (Teer Coatings ST-3001). The choice of load was designed to replicate the relatively severe damage observed in service and produced scratches approximately 0.3 mm in width and 30 μm in depth. Multiple scratches were made in parallel, spaced at 0.5 mm and 1.0 mm. Transverse scratches were 10 mm long and made in groups of 15, while longitudinal scratches were 50 mm long and made in groups of 10; hence scratches covered between 0.4% and 2% of the total tensile surface. Scratches applied longitudinally were oriented parallel to the applied tensile stress to simulate a pipeline under axial stress. Scratches applied in the transverse direction were perpendicular to the applied tensile stress to simulate a pipeline under hoop stress. Once scratched, specimens were left for 1 h to allow full repassivation to take place before being transferred to a vacuum desiccator, where they were stored prior to autoclave testing.

FIGURE 1.

Configuration used to simulate pigging damage.

FIGURE 1.

Configuration used to simulate pigging damage.

Close modal

Stress Corrosion Cracking Testing

Four-point bend testing was performed in accordance with NACE TM 0316.13  Autoclaves were assembled as shown in Figure 2 and were deaerated by purging overnight with high-purity nitrogen. Nitrogen pressure was then used to drive deaerated solution into the autoclave and the gas flow was maintained for a further hour, following which the test solution was charged overnight with the appropriate H2S/CO2 gas mixture at a rate of 100 mL/min. To achieve the desired H2S/CO2 partial pressure, the autoclave temperature was increased to 58°C at a constant total pressure of 1 bar (with the gas mixture still flowing); this was maintained for 1 h to allow equilibration of the liquid/vapor phases. The autoclave was then sealed and heated at constant volume to the test temperature (110°C). The pressure was monitored daily to ensure that it remained constant over the duration of the test, confirming that the H2S was not being depleted and the vessel was fully sealed. After 30 d, the test was terminated by reducing the temperature and purging with nitrogen for 24 h at a flow rate exceeding 100 mL/min.

FIGURE 2.

Autoclave configuration used in the SCC tests.

FIGURE 2.

Autoclave configuration used in the SCC tests.

Close modal

Test conditions are summarized in Table 2. Three different test conditions were selected to encompass conditions in which the alloy is (i) susceptible to SCC, (ii) resistant to SCC, and (iii) close to the pass/fail boundary. In each test, nine four-point bend specimens were tested: three unscratched control specimens reflecting typical performance of the cladding material, three longitudinally scratched specimens, and three specimens with transverse scratches.

TABLE 2

Test Conditions for SCC Tests on Type 316L SS Cladding Material

Test Conditions for SCC Tests on Type 316L SS Cladding Material
Test Conditions for SCC Tests on Type 316L SS Cladding Material

Test solutions representing produced water (50,000 ppm chloride) were prepared from analytical grade NaCl in deionized water. The solutions were adjusted to the required pH under 1 bar of CO2. A pH of 4 was achieved by addition of 0.4 g/L of sodium acetate (C2H3NaO2) followed by addition of dilute hydrochloric acid (HCl) as required. A pH of 4.5 was achieved by addition of NaHCO3. The test solutions were deaerated by purging overnight with high-purity nitrogen at a rate of 100 mL/min. Following each test the pH of the solution was re-measured under 1 bar of CO2 to confirm that it had not drifted during the test.

Post Test Analysis

Following SCC testing, the specimens were rinsed with deionized water and dried in air. They were unloaded from the four-point bend jigs and inspected for cracks using low-powered optical microscopy at a magnification of ×100. If no cracking could be observed on the surface, the specimen was cross-sectioned and prepared metallographically for inspection with high-powered optical microscopy at a magnification of ×500. Selected cross sections were also analyzed using scanning electron microscopy and electron backscatter diffraction (EBSD).

Electrochemical Impedance Spectroscopy

Electrochemical impedance spectroscopy (EIS) was used to explore the effect of surface defects on the corrosion properties of the clad surface. Coupon specimens were produced from the cladding material and nickel wire encased in polyetheretherketone (PEEK) tubing was spot-welded onto them to provide electrical contact. The non-exposed area of each electrode was then sealed with a high-temperature epoxy adhesive leaving an exposed area of approximately 4 cm2. The relatively low density of surface defects on the as-received surface meant that determining their influence on the impedance signal could have presented a challenge. To address this, artificial defects were produced by coarsely abrading the as-received surface in air with 60 grit SiC paper. This preparation technique produced defects comparable to those that were present on the supplied pipe. Specimens prepared in this way were then compared to unabraded as-received specimens that were free of pre-existing defects. To prevent localized corrosion from taking place, which would prevent analysis of the passive surface, EIS tests were performed in the absence of H2S and at a reduced test temperature. The corrosiveness of the solution was then varied by incrementally decreasing the pH.

The test setup comprised a three-electrode electrochemical cell consisting of a cladding coupon (working electrode), a saturated calomel reference electrode, and a platinum wire counter electrode. The electrolyte used was a 50,000 ppm Cl solution saturated with CO2 and maintained at 70±0.5°C. The pH was initially adjusted to 4.5 using NaHCO3, and then following each completed frequency scan the solution was acidified by a further 0.5 pH units using dilute HCl. Impedance spectra were recorded at each pH upon stabilization of the corrosion potential (Ecorr); this was repeated until measurements had been made at pH 3.0. In each instance the frequency was scanned from 1 kHz to 0.1 Hz at an amplitude of ±10 mV from Ecorr using a VMP 2 AC potentiostat (Princeton Applied Research). Data fitting was performed using EC-LAB® V10.39.

Stress Corrosion Cracking Testing

The results of the SCC tests are summarized in Table 3. Consistent with the previous study on supermartensitic and duplex stainless steels,11  simulated pigging did not lead to SCC in any of the as-received specimens tested within the generally accepted safe operating envelope of the alloy (Test 1) and in tests performed at the upper limit (Test 2), close to the pass/fail boundary. This was determined from detailed microscopic (×100) inspection of the specimens in plan-view, and subsequent metallographic preparation of cross sections inspected using high-powered optical microscopy (×500). Some shallow pitting (<20 μm in depth) was observed in Test 2; however, these pits were considered to be non-propagating and did not form with any tendency toward pigging scratches. The absence of cracking of Type 316L SS in these conditions is consistent with a previous study that investigated the SCC resistance of ground Type 316L SS specimens in identical environmental conditions.5 

TABLE 3

Results of the SCC Tests

Results of the SCC Tests
Results of the SCC Tests

In contrast, as-received specimens tested under the most severe conditions (Test 3) exhibited multiple instances of localized corrosion and cracking. On the control specimens, irregular pits exceeding 500 μm in width formed preferentially at pre-existing surface defects. On the pigged as-received specimens, large pits formed preferentially on both pigging scratches and pre-existing defects, as illustrated in Figure 3. Typically, the pits that formed on pigging scratches were around 300 μm in diameter, appearing to be limited only by the width of the scratch on which they formed. Regions of the as-received surface that were free of defects and scratches showed no evidence of either propagating pits or SCC, but this may be partly attributable to relative difficulty in detecting cracks on the dull as-received surface compared to the highly reflective scratched surface.

FIGURE 3.

Localized corrosion of pigging scratches (top) and pre-existing defects (bottom), tested at 110°C in 0.1 bar H2S, pH 4.0, and 50,000 ppm Cl.

FIGURE 3.

Localized corrosion of pigging scratches (top) and pre-existing defects (bottom), tested at 110°C in 0.1 bar H2S, pH 4.0, and 50,000 ppm Cl.

Close modal

Preferential cracking at longitudinal scratches was observable in plan-view under low magnification (×10) and appeared to have initiated from pits, as shown in Figure 4. Cracks associated with transverse scratches were only identifiable upon examination of metallographically prepared cross sections of the specimens. In these occurrences, cracks were not associated with pits but formed at the edge of scratches and appeared to extend through the radial borderline between the bulk microstructure and a band of near-surface strain-induced nanocrystallization. Examples of this type of cracking are presented alongside an annotated EBSD grain reference orientation color scale map of a scratch cross section in Figure 5.

FIGURE 4.

Examples of localized cracking of longitudinal pigging scratches on as-received specimens tested at 110°C in 0.1 bar H2S, pH 4.0, and 50,000 ppm Cl.

FIGURE 4.

Examples of localized cracking of longitudinal pigging scratches on as-received specimens tested at 110°C in 0.1 bar H2S, pH 4.0, and 50,000 ppm Cl.

Close modal
FIGURE 5.

EBSD grain reference orientation color scale map superimposed on a gray scale image quality map of the cross section of a transverse scratch (top left), and optical micrographs of cross sections of transverse scratches exhibiting SCC (top right and bottom).

FIGURE 5.

EBSD grain reference orientation color scale map superimposed on a gray scale image quality map of the cross section of a transverse scratch (top left), and optical micrographs of cross sections of transverse scratches exhibiting SCC (top right and bottom).

Close modal

Ground specimens (Test 4) presented a different trend to the as-received specimens and exhibited no evidence of preferential pitting within scratches. Longitudinal scratches showed evidence of pitting but not cracking, either in plan-view or on prepared cross sections. In contrast, transverse scratches exhibited evidence of cracks that had been broadened by subsequent corrosion at the edge of the scratch profile. All of the ground specimens demonstrated evidence of pitting across the entire surface, with numerous pits exceeding 75 μm in diameter and 50 μm in depth. Inspection of these pits using confocal microscopy revealed that they had an open geometry and did not appear to have initiated any SCC.

Electrochemical Impedance Spectroscopy

Nyquist plots of the fitted spectra acquired during the EIS measurements are shown in Figure 6. These were fitted using two commonly used equivalent circuits for stainless steels in aqueous environments, as illustrated in Figure 7. The first is a relatively simple circuit that considers only the behavior of a single layer passive film. This is described by a charge transfer resistance in parallel with a constant phase element (CPE). The impedance of a CPE (ZCPE) is described by Equation (1) and accounts for non-ideal capacitive behavior. This basic circuit was used to provide a general comparison of the trends of the spectra while avoiding errors that can arise when fitting data to a circuit with a large number of elements.

formula

where Q = constant with units of F·m−2·s(a−1) equivalent to capacitance when a is unity; a = deviation from pure capacitance, unity representing a pure capacitor; and ω = frequency of AC perturbation.

FIGURE 6.

Nyquist plots of the fitted EIS measurements for specimens with as-received (top) and abraded (bottom) surfaces, immersed in 50,000 ppm (Cl) solution at 70°C at varying pH.

FIGURE 6.

Nyquist plots of the fitted EIS measurements for specimens with as-received (top) and abraded (bottom) surfaces, immersed in 50,000 ppm (Cl) solution at 70°C at varying pH.

Close modal
FIGURE 7.

Equivalent circuits for a single layer passive film (top) and a bilayer passive film (bottom).

FIGURE 7.

Equivalent circuits for a single layer passive film (top) and a bilayer passive film (bottom).

Close modal

The second circuit was used to produce estimates of passive film thickness and charge transfer resistance. This circuit is more complex than the first and assumes that the passive film is made up of a bilayer comprising an outer porous film and a dense inner film. Rp represents the resistance of the porous film, including the resistance within the pores, and CPEp is its non-ideal capacitance. Rct is the charge transfer resistance of the compact inner layer and CPEdl is the non-ideal capacitance of the electrical double layer.14-15  Because of the higher number of degrees of freedom associated with this circuit, a good fit can be achieved for a wide range of values for each element, which introduces some uncertainty in the derived values. For this reason, analysis of the EIS results focuses on the general trends observed in the data rather than the precise quantitative values of the outputs.

The Brug model was used to determine the effective capacitance of each CPE (Equation [2]).16  Previously, this method has been used to calculate passive film thicknesses for Type 304 stainless steels (UNS S30400) and produced values that were consistent with ellipsometry measurements.15  

formula

where Ceff = effective capacitance of passive film (F), and Rs = solution resistance (Ω).

The results were then used to calculate the passive film thickness for each of the spectra using Equation (3).

formula

where d = passive film thickness; ε0 = permittivity of vacuum (8.85 × 10−12 F/m); ε = relative permittivity of oxides of steel, taken as 12;17  and A = geometric surface area of electrode (m2).

Fits using each of the two equivalent circuits displayed the same general trend; the passive film on the as-received surface exhibited a greater charge transfer resistance and a lower effective capacitance (Ceff) than that of the abraded surface at all pH values tested. A summary of the EIS fitting parameters for the bilayer equivalent circuit is presented in Figure 7, demonstrating that the as-received surface has a greater corrosion resistance and a slightly thicker passive film. It should be noted that all calculations were based on the geometric surface area of the specimens and the real surface area will be larger as a result of surface roughness. This implies that actual differences in corrosion resistance and passive film thickness would be greater than indicated in Table 4, as the as-received surface (∼1 μm Ra) was considerably rougher than the abraded surface (∼0.5 μm Ra). The different properties of the as-received and abraded surfaces may be explained by considering the manufacturing process of the pipeline. The as-received surface of the cladding is typically pickled and passivated during manufacture to remove mill scale, enhance passivation, and dissolve surface inclusions, which results in a substantial increase in its corrosion resistance and a more protective passive film relative to the abraded surface.18 

TABLE 4

EIS Fitting Parameters and Calculated Passive Film Thickness for As-Received and Abraded Specimens

EIS Fitting Parameters and Calculated Passive Film Thickness for As-Received and Abraded Specimens
EIS Fitting Parameters and Calculated Passive Film Thickness for As-Received and Abraded Specimens

Electron Backscatter Diffraction

EBSD characterization of the scratch morphology on the tested four-point bend specimens showed that abrasion of the surface, whether originating from simulated pigging, grinding during specimen preparation, or pre-existing scratches on the as-received surface, produces near-surface residual strain and nanocrystallization (Figure 8). Previous studies that have investigated the direct effect of nanocrystallization on the corrosion properties of austenitic stainless steels have found that, in general, a more refined grain size offers improved resistance to initiation of both general corrosion and pitting when the alloy is in the passive state but is more susceptible to propagation once the alloy is depassivated.19 

FIGURE 8.

EBSD grain reference orientation color scale maps superimposed on a gray scale image quality maps showing cross sections of the unscratched as-received surface (top left), the scratched as-received surface (top right), the scratched and unscratched ground surface (bottom left), and a pre-existing defect on the as-received surface (bottom right).

FIGURE 8.

EBSD grain reference orientation color scale maps superimposed on a gray scale image quality maps showing cross sections of the unscratched as-received surface (top left), the scratched as-received surface (top right), the scratched and unscratched ground surface (bottom left), and a pre-existing defect on the as-received surface (bottom right).

Close modal

This implies that the preferential initiation of pits at scratches on as-received specimens, tested in the failure regime, is not attributable to the microstructural changes imposed during the scratching process. Instead, it is proposed that the preferential pitting at scratches and defects can be best explained by consideration of the influence of abrasion on the chemistry of the pickled and passivated surface. Abrasion removes the as-received passive film and repassivation then takes place forming a thinner and less corrosion resistant passive film. Abrasion also exposes fresh inclusions that can act as pit initiation sites. Hence, pigging or any other form of local abrasion of the as-received surface will produce a region of reduced corrosion resistance that will corrode preferentially in appropriately severe environments. This is not observed for ground specimens because the grinding process has already removed the pickled and passivated surface prior to any subsequent form of abrasion. In this case pigging only affects the microstructure and has little influence on the properties of the passive film.

It should be noted that relatively little cracking was observed on the ground specimens despite being tested well outside the safe operating envelope for the alloy. This may be attributable to the presence of compressive residual stresses that were produced during the machining process that was used to remove the carbon steel backing. Cross sections of these specimens display thick bands of near-surface residual strain and nanocrystallization that were not present on the unscratched as-received surfaces, as shown in Figure 8. Normally the strained material is removed by progressively grinding the surface with decreasing abrasive severity. In the present study, the use of progressive grinding was restricted because of the use of thin specimens (2.5 mm). Instead, the desired surface finish was produced using SiC paper, which resulted in only a minimal loss of thickness and hence did not remove the strained sub-surface material. These observations support the perspective that producing four-point bend specimens without progressive grinding can influence the outcome of SCC testing.

In summary, pigging of as-received specimens introduces surface scratches that are more susceptible to localized corrosion than the unscratched surface. However, testing at the upper limit of the safe operating envelope of the alloy did not shift them into a regime in which they were susceptible to SCC. This may be rationalized by considering the typical test procedures used to establish the performance limits of alloys for sour service. Test programs to evaluate resistance to SCC typically include specimens with a ground finish, i.e., with the pickled and passivated surface removed, which provides a conservative test with respect to service application.

The observations reported in the present study support the perspective that pigging of Type 316L SS clad lines should not negatively impact their resistance to SCC. It is important to consider that this study simulated typical conditions in which pigging is performed in the absence of H2S and exposure to H2S only occurs after pigging. Under these conditions, scratching the surface of the alloy results in depassivation, which is followed by rapid repassivation at the point the scratching is stopped. If pigging were performed in the presence of a significant concentration of H2S, it is possible that the surface may not repassivate and this could lead to localized corrosion. Investigation of this effect is beyond the scope of this study, but could form the basis of future work.

  • A novel test setup has been developed that can be used to simulate pigging damage in downhole oil and gas production environments.

  • No increase in susceptibility to SCC was observed for as-received Type 316L SS cladding as a result of simulated pigging, even in tests performed at the upper limit of the safe operating envelope of the alloy.

  • In conditions where the alloy is susceptible to SCC, preferential pitting and cracking initiated at both pigging scratches and pre-existing defects on as-received specimens.

  • Abrasion of as-received Type 316L SS cladding reduced its corrosion resistance as a result of removal of the surface that had been subjected to pickling and passivation. However, this effect is taken into account by the conservatism of SCC test programs, in which ground specimens are typically used.

(1)

UNS numbers are listed in Metals and Alloys in the Unified Numbering System, published by the Society of Automotive Engineers (SAE International) and cosponsored by ASTM International.

Trade name.

1.
R.M.
Roehner
,
J.V.
Fletcher
,
F.V.
Hanson
,
N.F.
Dahdah
,
Energy Fuels
16
(
2001
):
p
.
211
217
.
2.
J.
Quarini
,
S.
Shire
,
Proc. Inst. Mech. Eng. Part E
221
(
2007
):
p
.
1
10
.
3.
G.T.
Burstein
,
P.C.
Pistorius
,
Corrosion
51
(
1995
):
p
.
380
385
.
4.
M.H.
Moayed
,
N.J.
Laycock
,
R.C.
Newman
,
Corros. Sci.
45
(
2003
):
p
.
1203
1216
.
5.
G.
Hinds
,
L.
Wickström
,
K.
Mingard
,
A.
Turnbull
,
Corros. Sci.
71
(
2013
):
p
.
43
52
.
6.
A.
Turnbull
,
K.
Mingard
,
J.D.
Lord
,
B.
Roebuck
,
D.R.
Tice
,
K.J.
Mottershead
,
N.D.
Fairweather
,
A.K.
Bradbury
,
Corros. Sci.
53
(
2011
):
p
.
3398
3415
.
7.
Z.
Szklarska-Smialowska
,
Pitting Corrosion of Metals
(
Houston, TX
:
NACE International
,
1986
).
8.
A.B.
Rhouma
,
H.
Sidhom
,
C.
Braham
,
J.
Lédion
,
M.E.
Fitzpatrick
,
J. Mater. Eng. Perform.
10
(
2001
):
p
.
507
514
.
9.
Y.
Sano
,
M.
Obata
,
T.
Kubo
,
N.
Mukai
,
M.
Yoda
,
K.
Masaki
,
Y.
Ochi
,
Mater. Sci. Eng. A,
417
(
2006
):
p
.
334
340
.
10.
Y.F.
Al-Obaid
,
Eng. Fract. Mech.
51
(
1995
):
p
.
19
25
.
11.
J.
Hesketh
,
G.
Hinds
,
R.
Morana
,
Corrosion
72
(
2016
):
p
.
439
448
.
12.
ASTM Standard A380
,
“Standard Practice for Cleaning, Descaling, and Passivation of Stainless Steel Parts, Equipment, and Systems”
(
West Conshohocken, PA
:
ASTM International
,
2013
).
13.
NACE Standard TMO316
,
“Four-Point Bend Testing of Materials for Oil and Gas Applications”
(
Houston, TX
:
NACE
,
2016
).
14.
N.
Mahato
,
M.M.
Singh
,
Portugaliae Electrochim. Acta
29
(
2011
):
p
.
233
251
.
15.
F.
Mohammadi
,
T.
Nickchi
,
M.M.
Attar
,
A.
Alfantazi
,
Electrochim. Acta
56
(
2011
):
p
.
8727
8733
.
16.
G.J.
Brug
,
A.L.G.
Van Den Eeden
,
M.
Sluyters-Rehbach
,
J.H.
Sluyters
,
J. Electroanal. Chem. Interfacial Electrochem.
176
(
1984
):
p
.
275
295
.
17.
W.
He
,
O.Ø.
Knudsen
,
S.
Diplas
,
Corros. Sci.
51
(
2009
):
p
.
2811
2819
.
18.
D.
Wallinder
,
J.
Pan
,
C.
Leygraf
,
A.
Delblanc-Bauer
,
Corros. Sci.
42
(
2000
):
p
.
1457
1469
.
19.
R.K.
Gupta
,
N.
Birbilis
,
Corros. Sci.
92
(
2015
):
p
.
1
15
.