Effect of Hydrogen Charging on the Stress Corrosion Cracking Behavior of X70 Steel in Simulated Deep Seawater Environment

Abstract

The effects of hydrogen charging on the electrochemical and stress corrosion cracking (SCC) behavior of X70 steel in a simulated deep seawater environment were investigated by using electrochemical measurements, slow strain rate tensile (SSRT) tests, and corrosion morphology characterization through SEM. The results showed that the concentrations of the adsorbed hydrogen in X70 steel after precharging under different hydrostatic pressures increased gradually and tended to be steady with the charging time. High hydrostatic pressures promoted the hydrogen permeation of X70 pipeline steel by promoting the permeating rate and quantity. The SCC susceptibility of X70 steel decreased first and then increased with the hydrogen-charging current density. The area reduction loss (I_ψ) and true strain loss (I_ε) exhibited the lowest SCC susceptibility at the 25 mA/cm² hydrogen-precharging current density. The elongation rate loss (I_δ) exhibited the lowest SCC susceptibility at the 50 mA/cm² hydrogen-precharging current density.

1. Introduction

Oil and gas will continue to be the dominant source of energy resources across the world in the future. Due to the rapid exploration and development of offshore oil and gas, oil and gas resources in deep sea have been considered the major production sources in the exploitation field, and they have led to a large scale of pipeline steels in service. Due to the aggressive marine environment and the complexity of oil and gas production, there may be a high risk of failure for subsea pipelines caused by the combined factors of corrosion, stress corrosion cracking (SCC), hydrostatic pressure, waves, and currents, as well as some emergency events. As a result, huge economic losses and serious environmental pollution occur.

SCC is one of the main failure modes because of the offshore pipeline usually laid directly on the seabed or suspended to the offshore platform. Therefore, changes in a number of aspects including seabed topography, ocean currents, and geological factors produce additional structural tensile stresses. Moreover, high-level tensile stress within local pipeline segment, due to various superpositions of structural stress/strain, hoop stress, and residual stress in weld joints, may exceed the critical level in the occurrence of SCC [1]. Meanwhile, the disbanding and peeling of coatings lead to the direct contact of corrosive media on the surface of pipeline steels. Under these synergetic effects, it can be strongly confirmed that it is highly risky of SCC failure for subsea pipelines.

Although quite a few SCC accidents have been recognized in deep-sea environments, a number of accidents caused by the SCC of oil/gas pipelines in land soil have been characterized in recent years [2,3,4,5,6,7], which are mainly divided into high-pH SCC and near-neutral-pH SCC [8,9,10,11]. High-pH SCC often occurs in places where cathodic protection (CP) is strong with localized pH values higher than nine, while near-neutral-pH SCC usually takes place in weak-CP or no-CP areas [12,13]. The above situations can be more severe, as the service condition is much more complex in subsea environments. Moreover, the penetration of seawater into the gaps will promote the peeling of coatings, creating a more complex corrosion environment. Therefore, it is believed that high-pH or near-neutral-pH environments with a higher concentration of chloride ions can form under peeled coatings, and SCC is easily initiated in such situations [14,15].

Many studies have categorized the mechanisms of SCC into three ways, i.e., anodic dissolution (AD), hydrogen embrittlement (HE), and their mixed influences [16,17]. Like terrestrial pipelines, subsea pipelines were also installed with CP. Therefore, once the coating is destroyed or cathodically disbanded by over-CP, seawater will permeate into the gaps and lead to corrosion [18]. The SCC caused by HE is a severe environmental type of failure that affects almost all metals and alloys [19]. Studies indicated that the cathodic reaction of X70 steel in deep-seawater environments is hydrogen evolution [20,21]. The precipiation of hydrogen may be due to corrosion in aqueous media (including pickling), excessive CP, electroplating (without baking), welding by using damp electrodes, hydrogen gas (if moist) used as a coolant in generators, etc. Hydrogen can initially be present, either externally or internally, within the bulk of a structural alloy. The introduced hydrogen atoms can be adsorbed on the electrode surface and diffused to the internal metal matrix. As a result, the slip resistance of the dislocation decreases, and the dislocation density increases simultaneously. When the hydrogen concentration reaches the threshold value, the main dislocations are amalgamated at the aggregation source, resulting in a local shear stress and promoting the initiation of cracks. Therefore, the cathodic hydrogen evolution plays a significant role in the SCC susceptibility of X70 steel in the simulated seawater environment. Although HE has been extensively studied by many researchers, neither a clear interpretation nor a unified recognition of the mechanics has been achieved yet.

In recent years, the most recognized mechanisms of HE mainly includes stress-induced hydride formation and cleavage [22], hydrogen-enhanced localization plasticity (HELP) model [23], and hydrogen-enhanced decohesion (HEDE) [24,25]. Generally, the HELP model is more suitable for plastic fracture, and HEDE is more suitable for brittle fracture. Although various mechanisms have been proposed, an accurate understanding of the effect of hydrogen on the SCC behavior of X70 pipelines is still insufficient. Electrochemical methods based on the oxidation of reduced hydrogen absorbed in metals and the slow strain rate test method are still powerful tools for studying the effect of hydrogen on the SCC behavior of metals.

In this paper, the effect of hydrogen on the SCC process was investigated with the electrochemical hydrogen permeation method. A great deal of attention has been paid to the effect of the hydrogen current density on the SCC behavior under high hydrostatic pressure. The critical hydrogen concentration of the SCC susceptibility transition point of X70 pipeline steel was confirmed by calculating the average hydrogen density on the surface at various hydrogen-charging current densities.

2. Materials and Methods

2.1. Material and Samples

Slow strain rate tensile (SSRT) test specimens were manufactured with a rolled X70 steel sheet. The X70 steel was provided by Nanjing Iron & Steel United Co., Ltd., Nanjing, Jiangsu, China. The schematic diagram is represented in Figure 1. The chemical compositions were 0.045 wt% C, 1.53 wt% Mn, 0.013 wt% P, 0.0036 wt% S, 0.20 wt% Si, 0.048 wt% V, 0.030% wt% Cr, 0.0031 wt% Cu, 0.020 wt% Ni, 0.15 wt% Mo, and Fe-balanced. After polishing and cleaning, a dehydrogenation procedure was performed, by annealing the X70 specimens at 200 °C for 2 h and slowly cooling them to room temperature in an automatic furnace with a nitrogen inert environment. Prior to each test, the SCC specimens were ground in a longitudinal direction by a successive emery paper up to 2000 grit and then embedded in an epoxy resin with a working area of 0.75 cm². Then, the specimens were rinsed with deionized water, degreased in dehydrated ethanol and acetone with ultrasonic cleaning apparatus and subsequently dried in air.

2.2. Cathodic Hydrogen Charging

Cathodic hydrogen-charging specimens were prepared in duplicate for hydrogen concentration measurements and SSRT tests. Hydrogen charging tests were carried out in a 0.2 mol/L NaOH solution and poisoned with 0.25 g/L thiourea as a hydrogen recombination inhibitor at room temperature. The hydrostatic pressures for hydrogen charging were 0.1 and 5 MPa, which were pressurized by high-purity nitrogen. Before each experiment, the electrolyte was de-oxygenated by pure nitrogen bubbling for 1 h. The charging current densities employed were in the range of 10–1000 mA/cm², while the charging time varied from 2 to 24 h.

2.3. Hydrogen Concentration Measurement

After hydrogen charging was finished, some specimens were cleaned with a 0.2 mol/L NaOH solution once and then quickly transferred to a 0.2 mol/L NaOH solution to test the hydrogen concentration. Before each experiment, the electrolyte was deoxygenated by pure nitrogen bubbling for 1 h and kept until the test process was terminated (throughout the whole test process). The tests were carried out using a three-electrode electrochemical cell with a saturated calomel electrode (SCE) as the reference electrode and a platinum sheet as the auxiliary electrode. An additional potential of +168 mV was applied to the work electrode to monitor the anodic current variation. To obtain the hydroxide current, this anodic current should be corrected by subtracting the background current, which was obtained from the blank experiments with uncharged specimens.

2.4. SSRT Test

The SSRT method was used to investigate the SCC behavior of X70 steel after cathodic hydrogen charging in a simulated 500 m deep seawater environment. SSRT tests were conducted on an electrochemical slow strain rate machine (LF-100-201-V-304, Thermo Electron Corporation, Waltham, MA, USA), with a strain rate of 1.2 × 10⁻⁶ s⁻¹. The control loop of the slow strain rate machine was crosshead displacement-driven by a servo motor. The solution used in the tests was artificial seawater with pH of approximately 8.0 and pressurized by high-purity nitrogen to about 5 MPa. The test temperature was set at the ambient environment (approximately 25 °C). All the tensile tests were performed without the use of an extensometer. In order to calculate the strain, the gauge lengths of the specimens were marked and measured before and after fracture. During the test, the stress–strain curve was recorded to compare the ductile loss of the specimens. After failure, the percentage of elongation and the area reduction of each specimen were calculated after removing corrosion products. In addition, the fracture morphologies were observed using a scanning electron microscope. Each test was reproduced at least three times to ensure the reliability of experimental data. Afterwards, the cross-sectional morphologies of the fracture were observed to investigate the fracture location and cracking mode.

2.5. Hydrogen Analysis

The content of hydrogen in the metal matrix can be determined by the electrochemical oxidation method. When the charged sample is in an anodic electric field and in a NaOH solution, the hydrogen atoms in a subsurface diffuse from the metal matrix to the electrode surface. It can be oxidized to H⁺ to form a hydroxide current density (i_H) on the electrode surface shown in Equation (1):

$$H\to H^{+}+e^{-}.$$

(1)

The i_H has a proportional relationship with the hydrogen atoms density on the metal surface. The hydrogen damage susceptibility can be evaluated by the adsorbed hydrogen concentration (C) as calculated with i_H and the quantity of electric charge (Q), where Q can be acquired from the integration value of the i_H–t curve. The relationship of C with Q is shown in Equation (2), which is an established method, and it was firstly introduced by Yan et al. [26]:

$$C=\frac{Q}{zFv},$$

(2)

where F is the Faraday constant, v is the effective volume of the specimen, and z is the electronic number.

Investigations indicate that there is a square root relationship between the adsorbed hydrogen concentration and the hydrogen-charging current density, shown as following:

$$C\propto i_c^{1/2}.$$

(3)

Cathodic hydrogen charging is widely used as an efficient way for the hydrogen charging of metallic materials. The hydrogen retention ratio represents the relationship between the inner hydrogen diffusion rate and the outer hydrogen-biding rate, and it is shown in Equation (4):

$$R=\frac{Q}{i_c{t}_0},$$

(4)

where R is the hydrogen retention ratio and t₀ is the hydrogen-precharging time. From Equations (2)–(4), we can get the square root relationship [26] between R and i_c, as shown in Equation (5):

$$R\propto i_c^{-1/2}.$$

(5)

An anodic potential of +168 mV was applied to the blank specimen to test the background current i₀. The hydroxide current i_H was obtained from the anodic current at +168 mV by removing the effect of the background current, as illustrated in Figure 2.

3. Results

3.1. Effect of the Hydrostatic Pressure on the Hydrogen Permeation Behavior of X70 Steel

Figure 3 and Figure 4 depict the variation of the hydroxide current i_H of the X70 specimens with the time at the condition of different cathodic hydrogen-charging times at 0.1 and 5 MPa hydrostatic pressures. From the results, we can see that the hydroxide current i_H increased with the charging time under different hydrostatic pressures. The electric quantity Q can be obtained by integrating every single i_H–t curve, and the adsorbed hydrogen concentration C can be calculated from Equation (2). Figure 5 shows the variation of the adsorbed hydrogen concentration with the time under 0.1 and 5 MPa hydrostatic pressures. With the extension of the hydrogen-charging time, the concentration of the adsorbed hydrogen increased gradually and tended to be stable. Comparing the hydrogen permeation kinetics curves under different hydrostatic pressures, we can see that both the hydrogen permeation speed and the quantity were promoted by the hydrostatic pressure.

3.2. Effect of the Hydrogen-Charging Current Density on the Hydrogen Permeation of X70 Steel under High Hydrostatic Pressure

Based on the analysis above, the hydrostatic pressure promoted the hydrogen permeation rate and the permeation quantity of X70 steel under a constant hydrogen-charging current density. In this study, the effect of the cathodic hydrogen-charging current density on the hydrogen permeation behavior was also studied. Figure 6 depicts the variation of i_H with the time under various current densities of 10, 25, 50, 100, 200, and 1000 mA/cm² for 24 h.

Figure 7 shows the variation of the adsorbed hydrogen concentration C with the charging current density of X70 steel. It depicts that the average hydrogen concentration increased with the charging current density, and a linear relationship can be found with the square root value of the charging current density. It was in association with Equation (6):

$$C=\left(1.730\pm 0.223\right)+\left(1.047\pm 0.223\right)\times i_c^{1/2}.$$

(6)

The hydrogen retention rate (R) can be calculated through Equation (4). Figure 8 represents the variation of R with the charging current density. It depicts that the R decreased as the i_c increased. Besides, a linear relationship with the reciprocal square root value of the charging current density can be detected, and it was associated with Equation (7):

$$R=\left(-2.533\pm 0.216\right)+\left(1.094\pm 0.216\right)\times i_c^{-1/2}.$$

(7)

Through the analyses above, we drew the conclusion that the average hydrogen concentration rose with the hydrogen-precharging current density. Meanwhile, the efficiency was reduced with the increases of the hydrogen-precharging current density.

3.3. Effect of the Hydrogen Charging on the SCC Susceptibility of X70 Steel under High Hydrostatic Pressure

3.3.1. SSRT Tests

Figure 9 shows the SSRT results of X70 steel under a simulated 500 m deep seawater environment with different current densities of 10, 25, 50, 100, 200, and 1000 mA/cm². The elongation (δ), area reduction (ψ), and true strain (ε) of the SSRT experiments are listed in

Table 1. The elongation rates (δ), area reductions (ψ), and true strains (ε) of the slow strain rate tensil (SSRT) after 24 h hydrogen charging with different current densities.

Parameter	Hydrogen-Charging Current Density (mA/cm2)
	0	10	25	50	100	200
δ (%)	36.7 ± 0.6	35.8 ± 0.3	36.9 ± 0.4	40.6 ± 0.5	33.3 ± 0.6	30.9 ± 0.9
ψ (%)	82.1 ± 0.2	82.5 ± 0.1	82.6 ± 0.2	81.8 ± 0.2	78.2 ± 0.5	72.7 ± 0.6
ε (%)	1.72 ± 0.01	1.74 ± 0.01	1.75 ± 0.01	1.70 ± 0.02	1.52 ± 0.04	1.30 ± 0.05

. The parameters were obtained through the following equations:

δ = ΔL/L × 100%,

(8)

ψ = ΔA/A0 × 100%,

(9)

ε = ln(1 + ΔL/L),

(10)

where ΔL is the length variation after fracture, ΔA is the cross-sectional area after fracture, and L is the gauge length of the specimen.

Comparing the series test results, we can see that hydrogen charging had a few effects on the yield strength (σ_s) but a distinct effect on the elongation of X70 steel under the 500 m seawater environment. No obvious change of δ can be observed for the hydrogen-precharging current density lower than 25 mA/cm². As the hydrogen-precharging current density reached 50 mA/cm², δ increased. However, when the current density continued to increase to higher than 100 mA/cm², δ decreased again.

From the result, we can see that the elongation rate (35.8%) at the 10 mA/cm² hydrogen-precharging current density was slightly reduced compared to that of the specimen without hydrogen precharging (36.7%). The elongation rate (35.9%) at the 25 mA/cm² hydrogen-precharging current density was similarly to that of the specimen without hydrogen precharging (36.7%). When the hydrogen-charging current density reached 50 mA/cm², the elongation rate rose to 40.6%. As the current density kept increasing to 100 mA/cm², the elongation rate was reduced as the current density increased. The variations of the area reduction (ψ) and the true strain (ε) with the hydrogen-charging current density were in association with each other. When the hydrogen-precharging current density was lower than 25 mA/cm², both the area reduction (ψ) and the true strain (ε) had a slightly shift to the more positive side. When the hydrogen-precharging current density was higher than 50 mA/cm², both the area reduction (ψ) and the true strain (ε) had an obvious shift to the more negative side.

The elongation rate loss (I_δ), area reduction loss (I_ψ), and true strain loss (I_ε) were calculated to analyze the variation of the SCC susceptibility with the hydrogen-charging current density. Generally, the greater these indexes values are, the greater the SCC susceptibility is. Three SCC susceptibility parameters were calculated use Equations (11)–(13), respectively:

$$I_{\mathsf{\delta }}=\left(1-\frac{{\mathsf{\delta }}_s}{{\mathsf{\delta }}_0}\right)\times 100\%,$$

(11)

$$I_{\mathsf{\psi }}=\left(1-\frac{{\mathsf{\psi }}_s}{{\mathsf{\psi }}_0}\right)\times 100\%,$$

(12)

$$I_{\mathsf{\epsilon }}=\left(1-\frac{{\mathsf{\epsilon }}_s}{{\mathsf{\epsilon }}_0}\right)\times 100\%,$$

(13)

where δ_s, ψ_s, and ε_s are the elongation rate, area reduction, and true strain after hydrogen charging, respectively; δ₀, ψ₀, and ε₀ are the elongation rate, area reduction, and true strain without hydrogen charging. The three SCC susceptibility parameters results are depicted in Figure 10.

When the hydrogen-precharging current density was relatively lower than 25 mA/cm², slight decreases of I_ψ and I_ε were observed and a slight fluctuation of the I_δ was detected. When the hydrogen-precharging current densities were relatively higher than 25 mA/cm², all the three parameters exhibited an obvious shift to the more positive side, except a valley-like decrease of I_δ at 50 mA/cm² detected. As the analyses above, the SCC susceptibility of X70 steel decreased first and then increased with the hydrogen-charging current density. I_ψ and I_ε exhibited the lowest SCC susceptibility at the 25 mA/cm² hydrogen-precharging current density. The I_δ exhibited the lowest SCC susceptibility at the 50 mA/cm² hydrogen-precharging current density.

3.3.2. SEM Morphologies of the Fracture Surface

The SEM morphology of the fracture surface in each situation is shown in Figure 11. As shown in Figure 11(a1,a2), an obvious necking with a certain degree of a brittle fracture zone could be observed for the specimen without hydrogen charging. The necking increased gradually and the brittle fracture zone extended continuously, when the hydrogen-charging current densities were 10 and 25 mA/cm², as shown in Figure 11(b1,b2). With the current density increased, the necking shrunk gradually and the fracture surface tended to be smoother, as depicted in Figure 11(c1,f2), which represented a more noble brittleness characteristic of the X70 steel. Additionally, the fracture surfaces were more likely to display the quasi-cleavage morphologies. Tearing edges and dimples, as well as secondary cracks, can be detected almost in all fracture surfaces.

The SEM morphologies of the adjacent side face of the fracture in each situation are shown in Figure 12. Microcracks can be found in the side faces of all the specimens. The crack development orientation was not intergranular, like zigzag but approximately a straight line, which indicated the transgranular development mode of all these cracks. Thus, the SCC development mode of X70 steel with hydrogen precharging in the simulated seawater environment was mostly transgranular stress corrosion cracking (TGSCC).

Some microcracks across the whole corrosion pits were detected during the process of observing the side face cracks, which are shown in Figure 13. These cross-microcracks revealed that the cracks were initiated at the corrosion pits in priority and then extended. Furthermore, more cracks in sizes and amounts in the zone concentrated with more corrosion pits. This illustrated that hydrogen promoted the anodic dissolution of X70 steel in the simulated seawater environment. Meanwhile, the synergistic effects of the hydrogen and anodic dissolution led to the SCC.

4. Discussion

Wang et al. [27] researched the effect of hydrogen on the fracture toughness K_IQ of X70 steel using the electrochemical hydrogen precharging method. Their result shows that when the hydrogen-precharging current density was relatively lower than 25 mA/cm², hydrogen promoted K_IQ for a slightly increase. They defined this current density as the critical hydrogen-charging current density. When the current density was relatively lower than this value, the hardening effect could increase the K_IQ by 11.8% at maximum. However, when the current density was relatively higher than this value, K_IQ decreased, as the hydrogen charging current density increased. As high as the current density increased up to 1000 mA/cm², K_IQ decreased by 13.7%. This kind of effect was also observed in pure iron [28]. Other researches [29,30] also represented that decreasing of the fracture toughness K_IQ of X series steel (X60, X70, X80, and X100) needs a sufficiently high hydrogen-precharging current density or a sufficiently long hydrogen-precharging time. Therefore, the hydrogen embrittlement of pipeline steel needs a sufficiently high hydrogen density. It is in association with our experiment result that the variation of the SCC susceptibility decreased first and then increased with the rising of the hydrogen-charging current density, which means the non-linearity relationship of the hydrogen effect on the SCC susceptibility. There exists a critical hydrogen current density that corresponds with the inflection point of the SCC susceptibility. This can be interpreted by using the hydrogen-induced local plastic model [31,32]. The diffused hydrogen in the metal may concentrate in the dislocation or elastic stress field and form into a Cottrell air mass. The distribution of the Cottrell air mass departs the dislocation and the elastic stress field and reduces their synergistic effect simultaneously. Therefore, the slip resistance of dislocation can be reduced, and the dislocation density can be promoted. As the hydrogen concentration is high enough, the main dislocation would accumulate and integrate at the dislocation source. As a result, the local shear stress is produced, and it increases the SCC susceptibility by promoting the initiation and growth of the crack. The fundamental principle behind the hydrogen embrittlement mechanism is that the presence of hydrogen in the lattice reduces the bond strength, and this leads to rupture at lower levels of stress [33,34].

The fracture morphologies depicted the brittle fracture zone was continuously extended as the hydrogen-charging current density increased. All the fractures depicted quasi-cleavage caused by the intensification of hydrogen embrittlement and the inhibition of anodic dissolution. It could be attributed to the residual hydrogen atoms concentrated in the traps or grain boundaries. As the hydrogen content reached the threshold value, it could lead to hydrogen embrittlement by the hydrogen-induced hardening impact [35,36]. According to the result of SSRT shown in Figure 8, the critical hydrogen-charging current densities I_ψ and I_δ of the SCC susceptibility can be obtained as 25 and 50 mA/cm², respectively. Generally, the value of 25 mA/cm² should be taken as the hydrogen-precharging current density for the SCC susceptibility reflected by I_ψ was always better than for I_δ. However, whether the value was actually the minimum of the SCC susceptibility is unclear. To obtain the minimum and more precise result, more experimental samples and more accurate measurements are needed. Therefore, the critical charging current value should be in the range of 25–50 mA/cm² in our tests. Additionally, the critical value means a competition mechanism between hydrogen-induced plasticity and hydrogen embrittlement [37,38,39]. It is often considered that hydrogen traps are expected to mitigate hydrogen embrittlement by reducing the local hydrogen concentration at crack initiation sites [40,41]. Therefore, for a relatively low concentration of hydrogen atoms, the trap sites play the main role in controlling the elongation and area reduction loss of the X70 steels.

5. Conclusions

In this study, the effect of the hydrogen permeation on the SCC behavior of X70 steel was studied by electrochemical tests, SSRT tests, as well as morphology characterizations. A possible mechanism about the effect of hydrogen on the SCC behavior of X70 steels was proposed. The main conclusions are as follows:

(1) The concentrations of the adsorbed hydrogen in X70 steel after hydrogen precharging under different hydrostatic pressures increased gradually and tended to be steady with the charging time. A high hydrostatic pressure promoted the hydrogen permeation of X70 pipeline steel by promoting the permeating rate and quantity.

(2) The adsorbed hydrogen concentration increased with the increase of the hydrogen-precharging current density, satisfying the relationship of $C\propto \left(1.047\pm 0.223\right)\times i_c^{1/2}$. The hydrogen retention rate was reduced with the increase of the hydrogen-precharging current density, satisfying the relationship of $R\propto \left(1.094\pm 0.216\right)\times i_c^{-1/2}$.

(3) The SCC susceptibility of X70 steel decreased first and then increased with the hydrogen-charging current density. The area reduction loss (I_ψ) and the true strain loss (I_ε) exhibited the lowest SCC susceptibility at the 25 mA/cm² hydrogen-precharging current density. The elongation rate loss (I_δ) exhibited the lowest SCC susceptibility at the 50 mA/cm² hydrogen-precharging current density. The critical charging current value should be in the range of 25–50 mA/cm².