Startseite The corrosion behavior and cracking susceptibility of the disbonded coating in the X80 steel pipeline by theoretical simulations and experimental measurements
Artikel Öffentlich zugänglich

The corrosion behavior and cracking susceptibility of the disbonded coating in the X80 steel pipeline by theoretical simulations and experimental measurements

  • Bo Zhao EMAIL logo , Zhijuan Zhao , Dong Jin , Na Li EMAIL logo , Linlin Liu , Xin Cao und Binan Shou
Veröffentlicht/Copyright: 5. November 2021

Abstract

The corrosion behavior and cracking susceptibility of the disbonded coating in the X80 steel pipeline were investigated by different methods. The oxygen content in the trapped solution decreased rapidly with the formation of an airtight disbonded area. The airtight system affected the electrode reaction process, resulting in the inhibition of corrosion in the center of the disbonded area and the more refined surface finish of the sample. The bottom of the disbonded area underwent a relatively intense and accelerated corrosion reaction controlled by the diffusion process. The cracking susceptibility of the X80 steel firstly decreased and then increased in the pointing direction.

1 Introduction

A buried pipeline represents one of the most important transportation facilities in the industry (Li et al. 2020). Coating technology is an important method to prevent external steel corrosion in the buried pipelines and improve their service life. However, installation and service of steel induce inevitable damage to the applied coatings and the formation of pinholes, bubbles, and peeling-off marks in the coating (Nguyen Dang et al. 2015). These defects yield disbonded areas between the damaged coating and the base metal. Electrolytes from soil penetrate the disbonded areas during the long-term use of the pipeline. The solutions of soil electrolytes gradually evolve into trapped solutions with completely different physicochemical properties developed via numerous chemical and electrochemical reactions (Xu et al. 2016). The trapped solutions could yield stress corrosion cracking (SCC) of the pipeline steel, dramatically shortening the service life of the buried pipeline. There are many ongoing in-depth research studies on this topic since it is currently one of the hottest research topics in the field of anticorrosion of buried pipelines (Mauricio et al. 2019; Varela et al. 2015; Zhao et al. 2019).

The electrochemical properties and SCC behavior in the disbonded area of the X70 pipeline under an artificial seawater environment were investigated by a simulated disbonded coating device (Ma et al. 2020; Wu et al. 2019). Since coated pipeline steel is often used in conjunction with cathodic protection, a cathodic current is not likely to pass through the narrow disbonded gap when the pipeline is at the cathodic protection state. However, the cathodic protection fails to cover the bottom of the disbonded coating (Chen et al. 2009; Varela et al. 2015; Wang et al. 2016; Xu et al. 2008). The potential distribution under the disbonded coating and the composition of the trapped solution will be the focus of this corrosion study (Yan et al. 2008, 2014). Conventional methods are inconvenient to study the corrosion behavior of the bottom of the disbonded coating. Currently, the transfer of simulated solution to a macro-system coupled with a conventional stress corrosion experiment has been widely employed by considering factors such as local stress (Bueno et al. 2014; Gadala et al. 2014; Luo et al. 2016). Available studies indicate that high-strength pipeline steel has a relatively high SCC susceptibility to acidic soil in China (Liu et al. 2008, 2009a,b). Still, research on this topic in China lags behind other countries, and Chinese researchers directly employ the systems developed abroad, neglecting a significant difference in the soil environment between China and different geographical regions worldwide. Thus, the existing studies on the environments of disbonded areas formed under complex soil environment and their evolution cannot provide precise guidance for Chinese long-distance pipeline projects.

In order to confirm the mechanism of pipeline corrosion and SCC in the system of typical Chinese acidic soil, the corrosion behavior and SCC susceptibility of X80 steel under the disbonded coating were addressed by studying the composition, electrochemical behavior, slow strain rate tensile (SSRT) testing, and morphological observation. These results will shed new light on exploring the mass transfer in the trapped solution and developing corrosion behavior in the corresponding disbonded system. Thus, it will be used to guide the implementation of cathodic protection projects of pipeline steel.

2 Materials and methods

2.1 Materials preparation

X80 steel pipelines were used as a steel substrate, and its chemical composition is as follows (wt%): C 0.036, Si 0.197, Mn 1.771, P 0.012, S 0.002, Cr 0.223, Ni 0.278, Cu 0.220, Al 0.021, Ti 0.019, Mo 0.184, V 0.001, Nb 0.110, and Fe balance. The simulated solution with a 1:1 mass ratio made from the actual soil was used as an experimental solution. This solution was prepared using the following compound of the denoted concentrations: 0.0475 g/L NaCl, 0.0167 g/L CaCl2, 0.0391 g/L KNO3, 0.0151 g/L NaHCO3, 0.0266 g/L Na2SO4, 0.0170 g/L MgCl2 6H2O, and adjusted pH value to 4.5 with a 5% aqueous acetic acid solution (Ma et al. 2020). The ionic composition is shown in Table 1.

Table 1:

Ionic composition of the experimental solution.

pH Anion concentration (%) Cation concentration (%)
Cl SO42− NO3 HCO3 Ca2+ Mg2+ K+ Na+
4.5 0.0024 0.0036 0.0018 0.0011 0.0006 0.0002 0.0004 0.0021

2.2 Simulation method

A simulation device for the disbonded area was built for the experiment, Figure 1, to simulate the physicochemical properties of the trapped solution in the disbonded coating. A detailed description of the test equipment and procedures was reported in the literature (Ma et al. 2020). Ring-shaped plastic gaskets were sandwiched between these plates to adjust the gap between them to 0.24 mm. A reference electrode was designed to measure the trapped solution layer to address the incompatibility of the conventional reference electrode with the simulated disbonded system used in this study. The schematic diagram is shown in Figure 1c. The needles were immersed in the test liquid layer during testing.

Figure 1: 
						Schematic diagram of the device designed for the simulation of the disbonded area: (a) side view of the simulation device, (b) top view of the simulation device, and (c) the reference electrode.
Figure 1:

Schematic diagram of the device designed for the simulation of the disbonded area: (a) side view of the simulation device, (b) top view of the simulation device, and (c) the reference electrode.

2.3 Characterization

2.3.1 Composition testing

The solution composition was tested by the in-situ measurement of parallel specimens of the same size coupled with liquid sampling measurement. The pH value and ionic content of the trapped solution were measured with a Dionex ICS 2100 ion chromatograph analyzer.

2.3.2 Electrochemical measurement

The open-circuit potential and electrochemical impedance spectroscopy (EIS) were measured by a biologic VMP3 electrochemical workstation at a test frequency range of EIS of 100 kHz–10 mHz and an AC signal of 10 mV.

2.3.3 SSRT test

According to the experimental results, the trapped simulation solution was used in the stress corrosion test of the disbonded area at different positions. The specimens in the smooth plate shape were tested with a WDML-3 microcomputer-controlled slow-strain-rate tensile testing machine. The test methods were conducted in accordance with the GB/T15970 standard. After the termination of the fracture test, the specimens were taken out. The corrosion products attached to the surface were removed by resining the specimens with a rust remover. The dimensions of each specimen were measured with a micrometer screw, and their elongation and the section shrinkage were calculated in this experiment.

2.3.4 Morphological observation

The corrosion and fracture morphology were observed by an FEI Quanta 25 scanning electron microscope (SEM). The samples were de-rusted using a descaling solution with 500.0 mL HCl, 500.0 mL H2O, and 3.5 g hexamethylenetetramine. The typical morphology of the gold-sputtered samples’ surface was examined.

3 Results

3.1 Compositional changes

The oxygen content in the gap is closely related to the electrochemical process on the steel surface, and it mainly depends on the oxygen diffusion and redox processes. Generally, the diffusion process is the primary control condition for the oxygen content in the gap (Liu et al. 2009a,b). Figure 2 shows the temporal evolution of dissolved oxygen content in the trapped solution within the disbonded area. There is no distinct difference in the dissolved oxygen content in the trapped solution at various positions, which exhibits a rapid decline with time dropping from 8.3 to 1.0 mg/L on Day 1 before reaching a stable level. The immediate oxygen consumption induces the formation of an anaerobic environment where the effect of oxygen becomes negligible.

Figure 2: 
						The evolution of the dissolved oxygen in the trapped solution within the disbonded area with time.
Figure 2:

The evolution of the dissolved oxygen in the trapped solution within the disbonded area with time.

Figure 3 shows a relatively complex change in the pH value of the trapped solution in the disbonded area over time. The pH value of the trapped solution at each position firstly increases and then decreases. A maximal value is reached on Day 1, and it decreases with the progress of the corrosion reaction. However, the magnitude of the change varies with its distance from the holiday. It should be pointed out that the pH value change of the bottom of the disbonded area has the largest amplitude. The pH value is already lower than that of the bulk solution from Day 7 to 9. Besides, its pH value drops to about 4.0 on Day 25. The trend of the pH value change of the solution under the disbonded coating is consistent with the results reported in the literature (Ma et al. 2020; Wu et al. 2019; Yan et al. 2008), indicating that an acidic solution environment will form at the bottom of the disbonded coating (Qian et al. 2016).

Figure 3: 
						The evolution of the pH value in the trapped solution within the disbonded area.
Figure 3:

The evolution of the pH value in the trapped solution within the disbonded area.

Figure 4 shows the temporal change in the anion concentration of the trapped solution in the disbonded area. The Cl concentration exhibits the highest increase compared to other anions, following a monotonous increasing trend at each position of the disbonded area. The closer the position to the bottom of the disbonded area, the more significant the concentration increase. On Day 25, the Cl concentration at the bottom of the disbonded area (i.e., a distance of 25 cm) reaches 746.65 mg/L, increasing by 108% compared to the bottom of the disbonded area. The increase at the position near the holiday is smaller, being about 45% on Day 25. This difference in the Cl concentration increase is caused by the small diameter of Cl, making it prone to electromigration under the influence of the electric field (Hołyst and Poniewierski 2012). Both NO3 and HCO3 contents show a declining trend. Compared with NO3, the HCO3 content declines with a greater amplitude. On Day 7, the concentration at each position in the disbonded area is close to its minimum, without apparent difference among the positions. Specifically, the concentration at each location is about 10 mg/L, which is nearly 90% lower than the initial concentration. The decrease in the NO3 concentration is relatively small, reaching its minimal value on Day 15 at each position. Specifically, the change in the NO3 concentration at the bottom of the disbonded area is more evident than at the position near the holiday. The maximal decrease rate is about 34%. Similar to the Cl concentration changes, the NO3 ions diffuse from a high concentration to a low-concentration position (i.e., 250.0 mm) in the late stage of corrosion. However, the amplitude of concentration diffusion is relatively low. The SO42− concentration does not show a specific pattern, without a distinct concentration difference between the original bulk solution and the simulated solution collected on Day 25.

Figure 4: 
						The temporal evolution of the anion content in the trapped solution within the disbonded area: (a) Cl−, (b) HCO3−, (c) NO3−, (d) SO42−.
Figure 4:

The temporal evolution of the anion content in the trapped solution within the disbonded area: (a) Cl, (b) HCO3−, (c) NO3−, (d) SO42−.

Figure 5 shows the concentration of Na+, K+, Mg2+, and Ca2+ cations in the trapped solution in the disbonded area over time. Among these cations, the Na+ concentration undergoes the most remarkable change. Besides, the changing pattern varies with the distance from the holiday. The Na+ concentration does not substantially change at the position near the holiday in the early stage of the corrosion reaction, reaching the maximum on Day 7, and then starting to decrease. There is no distinct change in the Na+ concentration at the bottom of the disbonded area during the early stage of the corrosion reaction. A monotonous decrease of Na+ concentration occurs after Day 7. The change in the Na+ concentration at the position located in the center of the disbonded area shows an intermediate level. Since Na+ ions do not participate in the corrosion reaction, the change in its concentration may be related to the formation of the electric field gradient and the concentration diffusion. More significant changes take place at the bottom of the disbonded area (i.e., Position 5). The Na+ concentration drops to about 77% of its original concentration on Day 25. In contrast, the K+ concentration drops to about 83% of its original at the same time point. The increase rate of Mg2+ concentration on Day 7 is about 35%. The decrease in the Ca2+ concentration is more significant than that of the wide slit, being only about 34% of its concentration in the bulk solution.

Figure 5: 
						The temporal evolution of the cation content in the trapped solution within the disbonded area: (a) Na+, (b) K+, (c) Mg2+, and (d) Ca2+.
Figure 5:

The temporal evolution of the cation content in the trapped solution within the disbonded area: (a) Na+, (b) K+, (c) Mg2+, and (d) Ca2+.

3.2 In-situ corrosive electrochemical reaction variation

Figure 6 shows the change in the open-circuit potential at each position of the disbonded area over time. The initial potential of each position is at the same level of about −640 mV. After 2–3 days of corrosion, the open-circuit potential rapidly drops and reaches a relatively stable level. The open-circuit potential at the bottom of the disbonded coating (the holiday at 25 cm) retains a negative trend, indicating that the corrosion environment at the bottom of the disbonded coating is continuously changing. The open-circuit potentials at the holiday and the bottom of the disbonded area on Day 25 are −664 and −748 mV, respectively. The negative shifts from the initial open-circuit potential are about −20 and 108 mV, respectively, which indicates that with the progress of the corrosion reaction in the disbonded area, a large potential difference between the holiday of the disbonded area and the bottom of the disbonded area appears. With the corrosion development in the disbonded area, the holiday acts as a cathode. The bottom of the disbonded area acts as an anode, consistent with the conventional crevice corrosion phenomenon (Chen et al. 2015; Xu et al. 2016).

Figure 6: 
						The change in the open-circuit potential at each position of the disbonded area over time.
Figure 6:

The change in the open-circuit potential at each position of the disbonded area over time.

Figure 7 shows the EIS of X80 steel at each position of the disbonded area over time. The electrode impedance spectra at each position of the disbonded area are basically the same. They are composed of a high-frequency capacitance arc and a low-frequency inductive reactance arc, indicating a strong adsorption effect on the electrode surface. The radius of the arcs is almost the same, indicating the same reactions and reaction degrees (Zhao et al. 2021). The fitted equivalent circuit is presented in Figure 8a. With the progress of the corrosion reaction, the radius of the capacitance arc of each position increases on Day 1, while the inductive reactance arc disappears. This behavior may be related to a layer of protective corrosion products formed on the surface of X80 steel, which inhibits the corrosion reaction. The equivalent circuit is shown in Figure 8b. With the further progress of the corrosion reaction at each point in the disbonded area, the radius of the capacitive reactance arc at positions 1–3, i.e., closer to the holiday, increases monotonously with time. The largest increase in the radius of the capacitive reactance arc is observed in the impedance spectra of Position 3, indicating the highest resistance to the corrosion reaction, i.e., the corrosion products provide the best protection for the substrate at this position. The radius of the capacitive reactance arc at positions 4 and 5 firstly increases and then decreases. The maximal radius of the capacitive reactance arc at position 4 appears on Day 18, while at Position 5, it occurs on Day 3. Meanwhile, Warburg impedance appears in the impedance spectra, and the nature of the electrode reaction gradually shifts from the activation-controlled to the diffusion-controlled reaction (Zhao et al. 2018). The corresponding fitted equivalent circuit is presented in Figure 8c. The electrochemical impedance spectra of the disbonded area have the following characteristics: first, the radius of capacitive resistance arc at the pre-corrosion position, especially in the center of the disbonded area, exhibits a relatively large increase; second, during the post-corrosion stage, the radius of capacitive resistance arc at the bottom of the disbonded area undergoes a distinct decline. This behavior is characteristic of a crevice effect in the disbonded area, which can cause sealing of the disbonded area. This crevice effect inhibits corrosion in the center of the disbonded area but enhances it at the bottom of the disbonded area.

Figure 7: 
						The change in the EIS at each position of the disbonded area over time: (a) point1, (b) point 2, (c) point 3, (d) point 4, and (e) point 5.
Figure 7:

The change in the EIS at each position of the disbonded area over time: (a) point1, (b) point 2, (c) point 3, (d) point 4, and (e) point 5.

Figure 8: 
						The equivalent circuit of EIS, shown in Figure 7, with different features: (a) capacitive loop with inductance, (b) single capacitive loop, (c) capacitive loop.
Figure 8:

The equivalent circuit of EIS, shown in Figure 7, with different features: (a) capacitive loop with inductance, (b) single capacitive loop, (c) capacitive loop.

R ct is the charge transfer resistance (Ω cm2), Rs is the solution resistance (Ω cm2), RL is the inductance resistance (Ω cm2), L is inductance (H/cm2), Zw is Warburg impedance (Ω cm2), Q is a constant phase element.

The constant phase angle element Q in the equivalent circuit in Figure 8 can be calculated using Eq. (1):

(1)ZQ=1Y0(jω)n.

where j is the dimensionless square root of −1, ω (rad s−1) is the angular frequency of the sinusoidal AC wave signal; Y0−1 cm−2 s−1) always has a positive value; n is the diffusion coefficient. It is a dimensionless constant whose value ranges from 0 to 1. Among all these parameters, Y0 and n are two parameters of Q. When n = 0, ZQ equals 1/Y0. Under this circumstance, Q is resistive, and R is impedance; when n = 1, ZQ equals C. Under this circumstance, Q is capacitive, and the impedance equals 1/(jωC). It is generally perceived that n can be expressed as the electrode surface roughness caused by the dispersion effect [23].

By considering the charge transfer resistance Rct, the relation between the two parameters (n and Rct) and two variables (time and disbonding depth) can be illustrated by the curves presented in Figure 9.

Figure 9: 
						The changes in n and Rct in the disbonded area with time and disbonding depth: (a) 3D map of n, (b) heat map of n, (c) 3D map of Rct, and (d) heat map of Rct.
Figure 9:

The changes in n and Rct in the disbonded area with time and disbonding depth: (a) 3D map of n, (b) heat map of n, (c) 3D map of Rct, and (d) heat map of Rct.

Figure 9 shows the changes in n and Rct in the disbonded area with time and disbonding depth. Both n and Rct change in a regular manner. The X80 steel sample has a relatively rough surface. The value of n at the holiday is lower than that at the bottom of the disbonded area; in the center of the disbonded area, n is relatively large, and the sample has a relatively smooth surface. The charge transfer resistance Rct is larger in the center of the disbonded area and increases faster. Rct at the holiday is low, and the increase is slower. At the bottom of the disbonded area, Rct firstly increases and then decreases, and a minimum appears in the later stage of the corrosion reaction.

3.3 SCC susceptibility

According to the test results presented in Section 2.1, the holiday (original solution), near the holiday (i.e., Position 1, at 5 cm from the holiday), the center of the disbonded area (i.e., Position 3, at 15 cm from the holiday), and the bottom of the disbonded area (i.e., Position 5, at 25 cm from the holiday) were chosen as four sampling points in the following experiment. Trapped solutions collected from the four sampling points on Day 25 were subjected to a SSRT test to study SCC susceptibility in different trapped solution environments. The composition of the trapped solution at four sampling points is shown in Table 2. Meanwhile, as the dissolved oxygen content in the solution trapped in the disbonded area is low, and there is a little difference in the distribution of dissolved oxygen, the treatment of deoxidation by bubbling N2 into the solution was adopted before the stress corrosion test experiments. The specific deoxidation step is to pre-deoxygenate the simulated solution by bubbling high-purity nitrogen through the solution for 6 h before the experiment. The nitrogen continuously flew through the simulated solution until the end of the experiment.

Table 2:

The ionic components of the trapped solution.

Position Codename pH Anion concentration (%) Cation concentration (mg/L)
Cl SO42− NO3 HCO3 Na+ K+ Mg2+ Ca2+
Holiday YT0 4.5 0.0036 0.0024 0.0018 0.0011 0.0021 0.0004 0.0002 0.0006
Position 1 YT1 6.0 0.0052 0.0016 0.0006 0.0027 0.0006 0.0002 0.0005
Position 3 YT2 5.5 0.0064 0.0019 0.0003 0.0028 0.0005 0.0002 0.0004
Position 5 YT3 4.0 0.0074 0.0016 0.0004 0.0016 0.0003 0.0003 0.0002

Figure 10 shows the SSRT curves of X80 steel at different positions in the disbonded area. Compared with the air-exposed X80 steel, the fracture strength and elongation of the X80 steel soaked in the solution are reduced to a certain extent. This decline mainly originates from the fact that the simulated solution of the Yingtan district soil is a strong corrosive solution. Its trapped solution system is also corrosive. During fracture failure, the effects of corrosion factors and mechanical factors are coupled, which accelerates the loss of mechanical properties of steel.

Figure 10: 
						The SSRT curves of X80 steel at different positions in the disbonded area.
Figure 10:

The SSRT curves of X80 steel at different positions in the disbonded area.

According to the test results, the breaking strength of the steel at the four positions can be ranked as YT2 > YT1 > YT0 > YT3. In contrast, the corresponding elongation at the break of the steel at the four positions can be sorted in the following order: YT2 > YT1 > YT3 > YT0. From the perspective of fracture strength and fracture elongation, the steel has relatively weak mechanical properties in YT0 and YT3. To illustrate further the relationship between the tensile strength and the SCC susceptibility of the material, we selected the shrinkage (ψ, %) and section elongation (δ, %) as parameters to evaluate the SCC susceptibility. The calculation equations are as follows:

(2)δ=IAI0I0×100%
(3)ψ=S0SAS0×100%

where I0 (mm) is the original gauge length of the tensile specimen; IA (mm) is the gauge length of the tensile specimen after elongation; S0 (mm2) is the original cross-sectional area of the tensile specimen at I0; SA (mm2) is the original cross-sectional area of the fracture site.

The smaller the values of the section shrinkage ψ and the elongation δ, the greater the SCC susceptibility of the material. Figure 11 shows that the shifting of the SCC susceptibility of X80 steel under the disbonded gap reflected by ψ is relatively consistent with that reflected by δ. From the holiday to the bottom of the disbonded area, the SCC susceptibility of the material firstly decreases and then increases. Among the four investigated points, the X80 steel at the holiday has a strong SCC susceptibility. The strongest and weakest SCC susceptibility values are recorded at the bottom of the disbonded area and in the center of the disbonded area, respectively. This finding is consistent with the loss of the material’s mechanical properties, indicating that the two environments at the holiday and the bottom of the disbonded area have a relatively strong SCC susceptibility to X80 steel. The elongation and section shrinkage of X80 steel at the bottom of the disbonded area reaches minimum values, indicating that the sample undergoes sudden fracture during the elongation process. The high SCC susceptibility at the bottom of the disbonded area is mainly attributed to the high concentration of hydrogen atoms at the bottom of the disbonded area (Liu et al. 2008). Some hydrogen atoms penetrate the steel and form Cottrell atmosphere pinning dislocations, which affects the material’s resistance to stress corrosion (Liu et al. 2011; Song et al. 2014).

Figure 11: 
						The SCC susceptibility of X80 steel at different positions in the disbonded area: (a) shrinkage (ψ) and (b) section elongation (δ).
Figure 11:

The SCC susceptibility of X80 steel at different positions in the disbonded area: (a) shrinkage (ψ) and (b) section elongation (δ).

The fracture area was sliced from the fractured tensile specimen. The morphology of the fracture was observed on macro- and micro-scales using SEM. Figure 12 shows macroscopic necking and fibrous zones in the primary fracture of the air-fractured X80 steel. Distinctive dimple structures can be observed from a microscopic scale. The walls of the local dimples feature the characteristics of serpentine slip, typical for ductile fracture. This indicates that X80 steel undergoes ductile fracture during a slow-fracture-rate process and possesses good ductility in the atmosphere environment (Liu et al. 2016). Under the four investigated conditions for the simulated disbonded area, relatively flat macroscopic fractures of the samples are observed. There is a certain tendency to brittle fracture, which belongs to the fracture characteristic of the combined ductile and brittle fracture pattern. YT0 and YT3 exhibit a higher flatness among all the points, indicating a more distinct brittle fracture tendency at the holiday and the bottom of the disbonded area. This feature is consistent with the curves obtained at a slow elongation rate. The microscopic morphology shows that X80 steel does not only undergo dimple fracture in the four different solution environments but also that a certain number of micropores appear in the fracture zones of all four samples. Among them, fewer micropores appear in the simulated environments at YT1 and YT2, located in the center of the disbonded area. Besides, the average diameter of the micropores is relatively smaller than that observed at other points. The fracture zones of the two samples tend to feature the characteristics of ductile fracture. This indicates that X80 steel does not have strong SCC susceptibility in the center of the disbonded area, showing a clear tendency to ductile fracture. The presence of micropores is apparent in the two simulated environments of the holiday (YT0) and the bottom of the disbonded area (YT3). The fracture zones are flatter and feature specific characteristics of brittle fracture. In the YT3 environment located at the bottom of the disbonded area, small cracks appear in the primary fracture of X80 steel. The crack size is larger than the micropores that appear in the simulated YT0 solution. In the YT3 environment, there are obvious cracks around the micropores on the X80 steel surface, indicating higher SCC susceptibility (Ma et al. 2020).

Figure 12: 
						The macro- and micro-scale morphology of the fracture of X80 steel at each positions: (a) in air, (b) in air, (c) YT0, (d) YT0, (e) YT1, (f) YT1, (g) YT2, (h) YT2, (i) YT3, and (j) YT3.
Figure 12:

The macro- and micro-scale morphology of the fracture of X80 steel at each positions: (a) in air, (b) in air, (c) YT0, (d) YT0, (e) YT1, (f) YT1, (g) YT2, (h) YT2, (i) YT3, and (j) YT3.

Figure 13 represents the lateral morphology of the fracture sample in Figure 12. Secondary microcracks do not appear on the lateral surface of the fracture of the control sample, which features the characteristics of plastic deformation (Liu et al. 2016). In the simulated solution environment, the lateral fracture morphology of YT1 and YT2 in the disbonded area exhibits relatively regular plastic deformation. It can be perceived that the material in the center of the disbonded area has a relatively low SCC susceptibility. However, there are some secondary microcracks on the lateral surfaces of the X80 steel fractures at the holiday (YT0) and the bottom (YT3) of simulated environments of the disbonded area. Specifically, the secondary cracks that appear on the steel surface under the YT3 environment are larger and more numerous than under the YT0 environment, which proves that the bottom of the disbonded area has higher SCC susceptibility. These secondary cracks exhibit straight shapes and apparent characteristics of trans-crystalline stress corrosion. It is worth noting that besides the obvious secondary microcracks, YT3 shows more pronounced corrosion morphology, which may be related to acidification of the solution at the bottom of the disbonded area (Mohtadi-Bonab et al. 2015). The trapped solution at the bottom of the disbonded gap is acidic and forms an anaerobic environment. The pH value of the trapped solution is close to 4, so the concentration of hydrogen atoms that can penetrate the base metal and promote the SCC process is significant, which supports the brittle fracture of the material (Wu et al. 2017).

Figure 13: 
						The lateral morphology of the fracture sample in Figure 12: (a) in air, (b) YT0, (c) YT1, (d) YT3, and (e) YT5.
Figure 13:

The lateral morphology of the fracture sample in Figure 12: (a) in air, (b) YT0, (c) YT1, (d) YT3, and (e) YT5.

4 Discussion

4.1 Evolution of solution in the disbonded area

The initiation and development of stress corrosion of pipeline steel under the disbonded anticorrosion layer involve various interfacial chemical/electrochemical reactions between steel and solution. The generation of corrosion products induces the formation of an airtight zone in the disbonded area. Mass transfer and mass diffusion in the disbonded area are affected by the concentration gradient and the electric field (Yan et al. 2015). Newman et al. (1966) proposed that in the calculation or simulation of the flux of a substance in a solution, three factors should be taken into consideration: migration under the influence of the electric field, diffusion under the concentration gradient, and the convection in liquid with a certain velocity. Since the trapped solution is practically completely static, the effect of the convection factor on the mass diffusion is negligible. As shown in Eq. (4), the factors affecting the composition of the solution trapped in the disbonded area include only the mass transfer under electric field and concentration diffusion.

(4)Ni=ziuiFCiφxDiCiii

where Ni is the mass flux (mol/cm2 s), Ui is the ionic mobility (m/V), F is Faraday constant (which is 96,500 C/mol), Zi is electric charges (C), Ci is the molar concentration (mol/L), ϕ is potential (V), and Di is the diffusion coefficient (cm2/s).

In this study, the difference in the compositions of the trapped solution is caused mainly by corrosion reactions. In aqueous solutions, the anodic reactions in the corrosion process of X80 steel mainly include the Fe dissolution reaction (Eq. (5)).

(5)FeFe2++2e

According to different conditions, such as the pH value of the solution, cathodic reactions including hydrogen evolution, Eq. (6), or oxygen absorption reaction, Eq. (7), may occur. In the disbonded area, due to a long and narrow gap, the mass transfer process is blocked. H+ and O2 in the disbonded area are quickly consumed, and OH is generated, yielding a rapid increase in the pH value of the solution trapped during the early stage of the reaction.

(6)2H++2eH2
(7)O2+4H2O+4e4OH

Meanwhile, due to the formation of the airtight environment in the disbonded area, X80 steel undergoes locally unsteady chemical-electrochemical reactions (Liu et al. 2019), yielding a significant difference in the ion composition between the trapped solution and the external bulk solution. The ionization equilibria of H+, OH, H2CO3, HCO3, and CO32− in the trapped solution are shown in Eqs. (8) and (10) (Egbewande et al. 2014; Eslami et al. 2010). In the early stage of the reactions, the H+ ions in the trapped solution are quickly consumed, and the pH value of the trapped solution gradually increases, shifting the equilibria of the reactions presented in Eqs. (8)(10) to the right. The HCO3 ions in the solution are converted into CO32−.

(8)H2CO3H++HCO32
(9)HCO3H++CO32
(10)HCO3+4OHH2O+CO32

In the trapped solution, Ca2+, Mg2+, and Fe2+ ions are generated in the corrosion reaction from CaCO3, MgCO3, and FeCO3 through chemical reactions, yielding a decrease of the ionic content in the trapped solution. As shown in Eqs. (11)(13), besides the reduction of the total HCO3 concentration in the trapped solution, these reactions also cause a decrease in the cation concentration, e.g., such as Ca2+ and Mg2+.

(11)Ca2++CO32CaCO3
(12)Mg2++CO32MgCO3
(13)Fe2++CO32FeCO3

The progress of the corrosion reaction induces the compositional difference of the solution at different positions in the disbonded area because of their different distances from the holiday. Specifically, closer to the holiday (Position 1), H+ and O2 can be partially supplemented by the holiday via the concentration diffusion. As a result, reductive oxygen reactions and hydrogen evolution reactions occur continuously, yielding a high pH value of the trapped solution. Therefore, the electrode potential rises, and this spot becomes the cathode area of the electrochemical reactions. A long and narrow gap prevents the transport of H+ and O2 ions penetrated in the trapped solution through the holiday to the bottom of the disbonded area. Therefore, the cathode reactants are quickly consumed and fail to promote the cathodic reactions. As a result, only anodic reactions occur in this area, and it then becomes the anodic zone of the corrosion reaction. The anodic reaction only involves the Fe dissolution reaction, Eq. (5). Meanwhile, under the anaerobic environment at the bottom of the disbonded area, Fe2+ ions undergo a hydrolysis reaction, and a large amount of H+ ions is generated (Eq. (14)). In the late stage of the corrosion reaction, the trapped solution in the bottom of the disbonded area is acidified due to the hydrolysis reaction, which can be verified by the results of the pH value measurement presented in Figure 4.

(14)Fe2++2H2OFe(OH)2+2H+

Overall, the above analysis shows that the continuous increase of the cation concentration at the bottom of the disbonded area disrupts the electroneutrality equilibrium in this area, so the electric field forms in the trapped solution from the bottom of the disbonded area to the holiday. To maintain electroneutrality, the anions outside of the disbonded area migrate inward. The cations inside of the disbonded area migrate outward (Wan et al. 2019). Na+, K+, and Cl ions exhibit relatively high ion mobility and strong migrating ability, so they mainly undergo long-range migration. In contrast, Mg2+, Ca2+, NO3, HCO3, and SO42− mostly undergo short-range migration. Meanwhile, the acidification of the trapped solution at the bottom of the disbonded area causes a partial dissolution of the corrosion products. Fe2+ ions also migrate to the holiday. The electrode reaction in the disbonded area is gradually dominated by diffusion control, and Warburg impedance appears in the EIS. As the vicinity of the holiday is relatively rich in oxygen, Fe2+ ions that migrate to this area are converted into Fe-containing oxides. These oxides accumulate around the holiday, which further deteriorates the sealing of the disbonded area and aggravates the progress of the unique corrosion reaction in the disbonded area.

The above analysis shows that the electric field and the ion concentration difference in the trapped solution cause the local distribution of ions in the disbonded area. This yields the compositional difference of the trapped solution among different positions, resulting in local differences in the SCC of X80 steel among the positions in the disbonded area. The diagram in Figure 14 shows the formation of the local compositional difference in the trapped solutions. However, a comprehensive analysis of the corrosion in the solution trapped in the disbonded area and the SCC mechanism under applied stress should be performed to illustrate the stress corrosion behavior and the SCC susceptibility of X80 steel.

Figure 14: 
						Schematic diagram of the corrosion reaction in the disbonded area.
Figure 14:

Schematic diagram of the corrosion reaction in the disbonded area.

4.2 SCC susceptibility

In acidic soil solutions, pipeline steel in the disbonded area solution is prone to both anode dissolution and hydrogen embrittlement (HE) (Liu et al. 2016; Mohtadi-Bonab et al. 2015). Figure 12 indicates that the SCC susceptibility at the holiday and the bottom of the bonded area is relatively higher than at the other positions. The potential difference in the disbonded area cannot provide full-scale cathodic protection of the holiday because of the low corrosion resistance of X80 steel. Therefore, the anode corrosion reaction caused by the Fe dissolution also occurs during the holiday (Liu et al. 2008, 2009a,b; Wu et al. 2019). However, the solution at the holiday can be regarded as the bulk solution of the soil. As the pH of this bulk solution is only 4.5, the solution does not exhibit a distinctive acidity. Besides, since the solution is relatively pure, the anode is unlikely to undergo a chemical poisoning reaction. Therefore, hydrogen ions do not play a dominant role during the reaction, indicating that the decrease in the mechanical properties of the material may be caused mainly by the anodic dissolution (AD) resulting from the corrosion reaction.

The bottom of the disbonded area represents the anode section, where electrochemical corrosion, mainly the significant electron donation of iron, occurs. Meanwhile, the pH value in this area drops below 4.0, and the acidity of the trapped solution increases. Besides, as the trapped solution at this position is rich in Cl, the side effect of chemical corrosion may also occur. The hydrogen precipitation increases because of the increase in acidity. Hydrogen ions are very likely to enrich steel defects, such as grain boundaries, inclusions, and dislocations, promoting HE. The coupling of the unsteady-state electrochemical behavior and stress corrosion reaction may result in a physical poisoning reaction, which, together with the increase in the degree of HE, may cause the shifting of the fracture failure mechanism to the mixed anode dissolution mechanism. This behavior and HE are critically important (AD + HE) (Ma et al. 2020; Zhao et al. 2021). The micromorphology, Figures 12e and 13e, can also confirm that the anode reaction at the bottom of the disbonded area has been aggravated.

5 Conclusions

  1. The corrosion behavior and SCC susceptibility of X80 steel under the disbonded coating were investigated by different methods. With the progress of corrosion reactions, the environment under disbonded coating formed an anaerobic environment, which significantly changed the chemical parameters of the solution, i.e., the content of dissolved oxygen, pH, and ionic distribution.

  2. The crevice affected the progress of electrode reactions and showed the area with the highest SCC susceptibility at the bottom of the disbonded area. A relatively severe corrosion reaction also occurred at this position. In contrast, the lowest area was located in the center of the disbonded area. The local difference was caused by the diffusion effect and the chemical/electrochemical reactions, exhibiting the highest SCC susceptibility level from the action of hydrogen.

  3. These findings indicated that the X80 steel used in this study exhibited a stronger SCC susceptibility at the bottom of the disbonded area than at other positions. Therefore, the trapped solution at the bottom of the disbonded area should be in the focus on engineering research.


Corresponding author: Bo Zhao, Institute of Chemistry, China Special Equipment Inspection and Research Institute, Beijing10029, China, E-mail: ; and Na Li, Shanghai Institute of Applied Physics, Chinese Academy of Sciences, Shanghai201800, China, E-mail:

Funding source: China Special Equipment Inspection and Research Institute for Youth Foundation Projects

Award Identifier / Grant number: 2019-02

Award Identifier / Grant number: 2018-11

Funding source: National Natural Science Foundation of China

Award Identifier / Grant number: 51801226

  1. Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.

  2. Research funding: Supported by China Special Equipment Inspection and Research Institute for Youth Foundation Projects 2019-02 and 2018-11, and National Natural Science Foundation of China (51801226).

  3. Conflicts of interest: The authors declare no conflicts of interest regarding this article.

References

Bueno, A.H.S., Moreira, E.D., Siqueira, P., and Gomes, J.A.C.P. (2014). Effect of cathodic potential on hydrogen permeation of API grade steels in modified NS4 solution. Mater. Sci. Eng., A 597: 117–121, https://doi.org/10.1016/j.msea.2013.12.033.Suche in Google Scholar

Chen, X., Li, X.G., Du, C.W., and Cheng, Y.F. (2009). Effect of cathodic protection on corrosion of pipeline steel under disbonded coating. Corrosion Sci. 51: 2242–2245, https://doi.org/10.1016/j.corsci.2009.05.027.Suche in Google Scholar

Chen, X., Wang, G., Gao, F., Wang, Y., and He, C. (2015). Effects of sulphate-reducing bacteria on crevice corrosion in X70 pipeline steel under disbonded coatings. Corrosion Sci. 101: 1–11, https://doi.org/10.1016/j.corsci.2015.06.015.Suche in Google Scholar

Egbewande, A., Chen, W., Eadie, R., Kania, R., Boven, G.V., Worthingham, R., and Been, J. (2014). Transgranular crack growth in the pipeline steels exposed to near-neutral pH soil aqueous solutions: discontinuous crack growth mechanism. Corrosion Sci. 83: 343–354, https://doi.org/10.1016/j.corsci.2014.02.032.Suche in Google Scholar

Eslami, A., Fang, B., Kania, R., Worthingham, B., Been, J., Eadie, R., and Chen, W. (2010). Stress corrosion cracking initiation under the disbonded coating of pipeline steel in near-neutral pH environment. Corrosion Sci. 52: 3750–3756, https://doi.org/10.1016/j.corsci.2010.07.025.Suche in Google Scholar

Gadala, I.M., and Alfantazi, A. (2014). Electrochemical behavior of API-X100 pipeline steel in NS4, near-neutral, and mildly alkaline pH simulated soil solutions. Corrosion Sci. 82: 45–57, https://doi.org/10.1016/j.corsci.2013.12.020.Suche in Google Scholar

Hołyst, R. and Poniewierski, A. (2012). Electrochemical systems. In: Hołyst, R. and Poniewierski, A. (Eds.). Thermodynamics for chemists, physicists and engineers. Springer Netherlands, Dordrecht, pp. 245–263.10.1007/978-94-007-2999-5_11Suche in Google Scholar

Li, J., Liu, Z., Lou, Y., Du, C., and Li, X. (2020). Evidencing the uptake of electrons from X80 steel by Bacillus licheniformis with redox probe, 5-cyano-2,3-ditolyl tetrazolium chloride. Corrosion Sci. 168: 108569, https://doi.org/10.1016/j.corsci.2020.108569.Suche in Google Scholar

Liu, Z.Y., Hao, W.K., Wu, W., Luo, H., and Li, X.G. (2019). Fundamental investigation of stress corrosion cracking of E690 steel in simulated marine thin electrolyte layer. Corrosion Sci. 148: 388–396, https://doi.org/10.1016/j.corsci.2018.12.029.Suche in Google Scholar

Liu, Z.Y., Li, X.G., and Cheng, Y.F. (2011). Mechanistic aspect of near-neutral pH stress corrosion cracking of pipelines under cathodic polarization. Corrosion Sci. 55: 54–60.10.1016/j.corsci.2011.10.002Suche in Google Scholar

Liu, Z.Y., Li, X.G., Du, C.W., and Cheng, Y.F. (2009a). Local additional potential model for effect of strain rate on SCC of pipeline steel in an acidic soil solution. Corrosion Sci. 51: 2863–2871, https://doi.org/10.1016/j.corsci.2009.08.019.Suche in Google Scholar

Liu, Z.Y., Li, X.G., Du, C.W., Lu, L., Zhang, Y.R., and Cheng, Y.F. (2009b). Effect of inclusions on initiation of stress corrosion cracks in X70 pipeline steel in an acidic soil environment. Corrosion Sci. 51: 895–900, https://doi.org/10.1016/j.corsci.2009.01.007.Suche in Google Scholar

Liu, Z.Y., Li, X.G., Du, C.W., Zhai, G.L., and Cheng, Y.F. (2008). Stress corrosion cracking behavior of X70 pipe steel in an acidic soil environment. Corrosion Sci. 50: 2251–2257, https://doi.org/10.1016/j.corsci.2008.05.011.Suche in Google Scholar

Liu, Z.Y., Wang, X.Z., Du, C.W., Li, J.K., and Li, X.G. (2016). Effect of hydrogen-induced plasticity on the stress corrosion cracking of X70 pipeline steel in simulated soil environments. Mater. Sci. Eng., A 658: 348–354, https://doi.org/10.1016/j.msea.2016.02.019.Suche in Google Scholar

Luo, J., Zhang, L., Li, L., Yang, F., Ma, W., Wang, K., and Zhao, X. (2016). Electrochemical corrosion behaviors of the X90 linepipe steel in NS4 solution. Nat. Gas. Ind. 3: 346–351, https://doi.org/10.1016/j.ngib.2016.12.011.Suche in Google Scholar

Ma, H., Zhao, B., Liu, Z., Du, C., and Shou, B. (2020). Local chemistry–electrochemistry and stress corrosion susceptibility of X80 steel below disbonded coating in acidic soil environment under cathodic protection. Construct. Build. Mater. 243: 118203, https://doi.org/10.1016/j.conbuildmat.2020.118203.Suche in Google Scholar

Mauricio, L., Facundo, V., Yongjun, T., and Maria, F. (2019). The effect of ageing on cathodic protection shielding by fusion bonded epoxy coatings. Prog. Org. Coating 134: 58–65.10.1016/j.porgcoat.2019.04.074Suche in Google Scholar

Mohtadi-Bonab, M.A., Szpunar, J.A., Basu, R., and Eskandari, M. (2015). The mechanism of failure by hydrogen induced cracking in an acidic environment for API 5L X70 pipeline steel. Int. J. Hydrogen Energy 40: 1096–1107, https://doi.org/10.1016/j.ijhydene.2014.11.057.Suche in Google Scholar

Newman, J. (1966). Effect of ionic migration on limiting currents. Ind. Eng. Chem. Fundam. 5: 525–529, https://doi.org/10.1021/i160020a016.Suche in Google Scholar

Nguyen Dang, D., Lanarde, L., Jeannin, M., Sabot, R., and Refait, P. (2015). Influence of soil moisture on the residual corrosion rates of buried carbon steel structures under cathodic protection. Electrochim. Acta 176: 1410–1419, https://doi.org/10.1016/j.electacta.2015.07.097.Suche in Google Scholar

Qian, H., Wang, L., Wang, H., Zheng, W., Zhang, D., and Du, C. (2016). Electrochemical behavior and stress corrosion sensitivity of X70 steel under disbonded coatings in korla soil solution. J. Mater. Eng. Perform. 25: 4657–4665, https://doi.org/10.1007/s11665-016-2319-6.Suche in Google Scholar

Song, J. and Curtin, W.A. (2014). Mechanisms of hydrogen-enhanced localized plasticity: an atomistic study using α-Fe as a model system. Acta Mater. 68: 61–69, https://doi.org/10.1016/j.actamat.2014.01.008.Suche in Google Scholar

Varela, F., Tan, M.Y.J., and Forsyth, M. (2015). Understanding the effectiveness of cathodic protection under disbonded coatings. Electrochim. Acta 186: 377–390, https://doi.org/10.1016/j.electacta.2015.10.171.Suche in Google Scholar

Wan, Y., Tan, J., Zhu, S., Cui, J., Zhang, K., Wang, X., Shen, X., Li, Y., and Zhu, X. (2019). Insight into atmospheric pitting corrosion of carbon steel via a dual-beam FIB/SEM system associated with high-resolution TEM. Corrosion Sci. 152: 226–233, https://doi.org/10.1016/j.corsci.2019.03.017.Suche in Google Scholar

Wang, W., Shen, K., Yi, J., and Wang, Q. (2016). A mathematical model of crevice corrosion for buried pipeline with disbonded coatings under cathodic protection. J. Loss Prev. Process. Ind. 41: 270–281, https://doi.org/10.1016/j.jlp.2016.03.024.Suche in Google Scholar

Wu, W., Liu, Z., Hu, S., Li, X., and Du, C. (2017). Effect of pH and hydrogen on the stress corrosion cracking behavior of duplex stainless steel in marine atmosphere environment. Ocean Eng. 146: 311–323, https://doi.org/10.1016/j.oceaneng.2017.10.002.Suche in Google Scholar

Wu, W., Liu, Z., Li, X., and Du, C. (2019). Electrochemical characteristic and stress corrosion behavior of API X70 high-strength pipeline steel under a simulated disbonded coating in an artificial seawater environment. J. Electroanal. Chem. 845: 92–105, https://doi.org/10.1016/j.jelechem.2019.05.044.Suche in Google Scholar

Xu, C., Gao, F., Wang, Y., and He, C. (2016). Transient numerical model for crevice corrosion of pipelines under disbonded coating with cathodic protection. Mater. Des. 89: 196–204.10.1016/j.matdes.2015.09.047Suche in Google Scholar

Xu, C., Li, X., Du, C., and Liang, P. (2008). Crevice corrosion behavior of the steel X70 under cathodic polarization. Acta Metall. Sin. 44: 1431–1438.Suche in Google Scholar

Yan, M., Sun, C., Xu, J., Wu, T., Yang, S., and Ke, W. (2015). Stress corrosion of pipeline steel under occluded coating disbondment in a red soil environment. Corrosion Sci. 93: 27–38, https://doi.org/10.1016/j.corsci.2015.01.001.Suche in Google Scholar

Yan, M., Wang, J., Han, E.-H., Sun, C., and Ke, W. (2014). Characteristics and evolution of thin layer electrolyte on pipeline steel under cathodic protection shielding disbonded coating. Acta Metall. Sin. 50: 1137–1145.Suche in Google Scholar

Yan, M., Wang, J., Han, E., and Wei, K. (2008). Local environment under simulated disbonded coating on steel pipelines in soil solution. Corrosion Sci. 50: 1331–1339, https://doi.org/10.1016/j.corsci.2008.01.004.Suche in Google Scholar

Zhao, B., Wan, H., Peng, Y., Liu, C., Li, J., and Shou, B. (2021). Influence of crevice thickness on corrosion behavior of API X80 steel under disbonded coating in acid soil environment. Int. J. Electrochem. Sci. 16: 150808, https://doi.org/10.20964/2021.01.19.Suche in Google Scholar

Zhao, T., Liu, Z., Chao, L., Dai, C., Du, C., and Li, X. (2018). Variation of the corrosion behavior prior to crack initiation of E690 steel fatigued in simulated seawater with various cyclic stress levels. J. Mater. Eng. Perform. 27: 4921–4931, https://doi.org/10.1007/s11665-018-3585-2.Suche in Google Scholar

Zhao, T., Liu, Z., Xu, X., Li, Y., Du, C., and Liu, X. (2019). Interaction between hydrogen and cyclic stress and its role in fatigue damage mechanism. Corrosion Sci. 157: 146–156, https://doi.org/10.1016/j.corsci.2019.05.028.Suche in Google Scholar

Received: 2021-03-28
Accepted: 2021-08-05
Published Online: 2021-11-05
Published in Print: 2021-12-20

© 2021 Walter de Gruyter GmbH, Berlin/Boston

Heruntergeladen am 16.9.2025 von https://www.degruyterbrill.com/document/doi/10.1515/corrrev-2021-0013/html
Button zum nach oben scrollen