Electrochemical and Sulfide Stress Corrosion Cracking Behaviors of Tubing Steels in a H₂S/CO₂ Annular Environment

13Cr and P110 steels, the chemical compositions and mechanical properties of which are shown in Table 1, were used in this experiment. Based on a survey on the chemical composition of the solution medium in a real tubing annular environment, the test solution was composed of NaHCO₃ (2.71 g/L), NaCl (6.15 g/L), and Na₂SO₄ (0.33 g/L) to simulate actual working conditions. The solution pH was then adjusted to 4 with 5% acetic acid. Given that an inhibitor is widely used in actual operational activities to control the corrosion of steels (Ref 16, 17), a 1000 ppm inhibitor (a commercial imidazoline-based product) was added to the solution under several test conditions to insure that test results are as close to real conditions as possible.

All the samples used for electrochemical tests were coated with epoxy, and an exposure area of 10 × 10 mm² served as the working surface. The working surface was ground using 1200 grit sandpaper, degreased with acetone, and rinsed with deionized water. High-pressure tests were performed using an autoclave with a capacity of 1 L. After the specimens and test solution were placed in the autoclave, the solution was de-aerated by purging nitrogen for 1 h. The autoclave was then pressured with H₂S to the required levels, which ranged from 0 to 0.3 MPa. CO₂ and N₂ were pressured to 4 and 9 MPa, respectively. The test conditions are shown in Table 2.

An EG&G M2273 electrochemical test system was used for electrochemical measurements. A three-electrode test cell was used with a platinum plate as the counter electrode, a Ag/AgCl electrode as the reference electrode, and the steel specimen as the working electrode. Potentiodynamic polarization curves were obtained by changing the electrode potential automatically from the cathodic branch to the anodic branch at a scan rate of 0.5 mV/s. Electrochemical impedance spectroscopy (EIS) was measured at open circuit potential with a sinusoidal potential excitation of 10 mV amplitude in the frequency range of 100 kHz to 10 mHz. The impedance data were fitted with ZsimpWin software using equivalent circuits. Experiments for each H₂S concentration were conducted at least thrice to insure that the results were reproducible and reliable.

U-bend immersion tests are useful in studying the SCC mechanism and propagation mode of materials under static load and static strain in a specific environment (Ref 18). Therefore, this study used U-bend immersion tests to investigate the effect of H₂S on SSCC susceptibility under a high-pressure CO₂/H₂S environment. Immersion tests were performed in a high-pressure autoclave with a capacity of 3 L. The working conditions (such as the test solution, order for pressured gases, and specific partial pressure parameters of the gases) for static load U-bend immersion tests were identical to those used for the electrochemical test. Three parallel specimens for each type of material were placed in the test solution for every test, which lasted for 720 h at room temperature.

After the corrosion tests, the immersed U-bend specimens used for surfacing analysis were removed and rinsed with deionized water. A wire cutter was used to cut the top parts of two parallel specimens. A heavy stress was applied to these parts, which were then immersed in descaling solution (500 mL of HCl + 500 mL of H₂O + 3.5 g of (CH₂)₆N₄) in an ultrasonic container for 3 min. Another specimen was retained for backup. The specimens were then degreased with acetone and dried with cool air, and their surface morphologies were observed by SEM.

Figure 1 shows the polarization curves of 13Cr stainless steel and P110 steel in solutions containing 4 MPa CO₂ with/without 0.05 MPa H₂S. Two curves for 13Cr stainless steel showed obvious passive characteristics; by contrast, the curves of P110 steel showed active dissolution characteristics. Compared with the H₂S-free condition, addition of 0.05 MPa H₂S significantly increased the passive current density (i _pass) of 13Cr stainless steel but slightly decreased the i _corr of P110 steel. For 13Cr stainless steel, the increase in i _pass implies that both anodic iron dissolution and cathode reduction were accelerated because of the presence of H₂S. For P110 steel, the decrease in i _corr demonstrates that the corrosion process was, to some extent, inhibited by the presence of H₂S. Compared with sweet corrosion, the presence of H₂S moved the corrosion potentials strongly toward the negative direction in both 13Cr and P110 steels.

The EIS curves of 13Cr stainless steel and P110 steel measured at open-circuit potential mode in solutions containing 4 MPa CO₂ with/without 0.05 MPa H₂S are shown in Fig. 2. The Nyquist plots exhibited characteristic capacitive semicircles, and the diameter of the capacitive semicircle of 13Cr stainless steel was significantly larger than that for P110 steel. Compared with the H₂S-free condition, addition of 0.05 MPa H₂S slightly decreased the diameter of the capacitive semicircle of 13Cr stainless steel but significantly increased the diameter of the capacitive semicircle of P110 steel. This result demonstrates that the presence of H₂S degrades the corrosion resistance of 13Cr stainless steel but improves that of P110 steel under the test conditions. The results so far are consistent with the results obtained from the polarization curves. Moreover, the shape of the Nyquist plots (capacitive loop) of each steel did not change with/without H₂S, which indicates the same mechanism for steel corrosion throughout the entire test.

Figure 3 shows the polarization and EIS curves of 13Cr and P110 steels under different \( {P}_{{{\text{H}}_{ 2} {\text{S}}}}\). An equivalent circuit shown in Fig. 4 was used to fit the EIS data to quantify the electrochemical parameters. In this equivalent circuit, R _s represents the solution resistance, Q _f represents the capacitance of the corrosion product film, R _f represents the resistance of the corrosion product film, Q _dl represents the double layer capacitance, and R _ct represents the charge transfer resistance. As both R _f and R _ct represent the resistances of the corrosion process of carbon steel, a parameter named polarization resistance, R _p (R _p = R _f + R _ct), was used to evaluate the corrosion rate of carbon steel in a H₂S/CO₂ environment (Ref 19).

Figure 5 shows the relationships of the EIS fitting results R _p, the corrosion current density i _corr of P110, and the passive current density i _pass of 13Cr with different H₂S partial pressures. The parameters i _pass (i _corr) and 1/R _p are in good agreement with each other. In Fig. 5(a), the R _p of 13Cr stainless steel without H₂S and 0.05 MPa \( {P}_{{{\text{H}}_{ 2} {\text{S}}}}\) exhibited significantly higher values than those obtained at higher H₂S partial pressures (0.1 to 0.3 MPa \( {P}_{{{\text{H}}_{ 2} {\text{S}}}}\)), and the i _pass increased greatly in the presence of H₂S. However, i _pass remained stable as the H₂S partial pressure increased from 0.05 to 0.3 MPa. In Fig. 5(b), the regularity of the change in 1/R _p of P110 steel was consistent with that of i _corr; both 1/R _p and i _corr significantly decreased with increasing H₂S partial pressure. The lowest R _p and largest i _corr both occurred during sweet corrosion, and the presence of H₂S significantly decreased the i _corr of the steel.

After the high-pressure immersion tests, the U-bend specimens were subjected to surface analysis after the removal of corrosion scales. Figure 6 shows micrographs of the surfaces of 13Cr stainless steel. Uniform corrosion was observed during sweet corrosion, while cracks on the surface of the sample could be observed during sour corrosion. Figure 6(b-d) shows that the width and depth of the cracks on the surface of 13Cr steel increased with increasing partial pressure of H₂S, which indicates that increases in H₂S enhance susceptibility toward SSCC. Moreover, the U-bend samples subjected to 0.2 and 0.3 MPa H₂S partial pressures were fractured (no failure specimen was observed at 0 and 0.1 MPa H₂S partial pressure). The fracture cross-sections of the specimens are shown in Fig. 6(e) and (f). SSCC cracks were initiated on the steel surface and continued to propagate until the ligament of the specimen could no longer bear the applied load.

Figure 7 shows the morphology of the surface of P110 tubing steel obtained under various \( {P}_{{{\text{H}}_{ 2} {\text{S}}}}\) after the removal of corrosion products. Corrosion pits with diameters of ~25 μm and small-sized cracks were observed without H₂S. At 0.1 MPa \( {P}_{{{\text{H}}_{ 2} {\text{S}}}}\), slight corrosion and a few small cracks were observed on the matrix. These generated cracks are related to hydrogen evolution during the cathodic process in the presence of H₂S. Hydrogen atoms that penetrate into the substrate facilitate the nucleation of cracks. As the H₂S partial pressure increased to 0.2 and 0.3 MPa, the depth and number of cracks significantly increased, which indicates that the cathodic process was significantly accelerated and that hydrogen ingress into the matrix was enhanced.

Figure 8 shows the SEM morphology of U-bent specimens of 13Cr stainless steel and P110 steel obtained without corrosion inhibitor after the removal of corrosion products. In corrosive medium without inhibitor, the metal comes into direct contact with anions and subsequent dissolution of the matrix occurs at the metal-solution interface. Therefore, severe corrosion occurs in metals in acidic solution without inhibitor. Upon addition of 1000 ppm inhibitor to the test solution, inhibitors function by adsorption of ions or molecules onto metal surface. The inhibitor prevents direct contact between the matrix and the corrosive medium and can thus reduce the corrosion rate by decreasing the anodic and/or cathodic reaction (Ref 17). Thus, compared with that in the solution without inhibitor, the corrosion process decreased greatly in the solution with inhibitor, as shown in Figs. 6(a) and (c) and 7(a) and (c). These results indicate that the inhibitor decreases the corrosion rate of the matrix significantly.

CO₂ oil displacement technology, a type of technology for enhanced oil recovery that requires the injection of CO₂ into underground reservoirs, has attracted worldwide attention for its many advantages, which include low cost, non-toxicity, environmental protection, and good miscibility with crude oil (Ref 20, 21). In natural gas and oil wells, packers are used to insure that the exterior of the tubing steel and the interior of the casing steel do not touch the corrosion medium. However, some defects may exist in the tubular steel, and these defects could cause widespread leakage of production steel. Corrosion at defect sites during service worsens the leakage problem. In CO₂ injection wells, the high pressure of CO₂ injected into the underground natural gas reservoirs may result in the following:

CO₂ then dissociates into the annular environment protective liquid: The chemical dissociation of CO₂ behaves similar to a weak acid and decreases the solution pH (the solution pH was ~4). This reaction accelerates the dissolution of steel during the anodic process, and the hydrogen evolution reaction in the cathodic process occurs as follows:

Promotion of the cathodic hydrogen evolution process will increase the susceptibility of the tubular steel toward SCC because hydrogen atoms generated in the cathodic process could penetrate into the steel substrate.

The annular environment between the tubing steel and the casing steel is full of protection liquids. This stationary, enclosed, de-aerated environment should not contain H₂S. However, our investigation on the solution medium extracted from the actual annular environment showed that the protective liquid contains a high concentration of H₂S. Thus, the outer face of the tubing steel suffered from sulfate-reducing bacteria (SRB) corrosion. SRBs are anaerobic bacteria that reduce sulfate to sulfide. The presence of SRBs in the annular environment can consume sulfate radicals though cathodic depolarization, thereby converting sulfate to sulfide and promoting H₂S formation in an acid environment (Ref 22, 23):

The synergistic effects of leakage and presence of SRB promote the formation of a complex annular environment in CO₂ injection wells featuring low temperature, high pressure CO₂/H₂S, and low pH (~4), as shown in Fig. 9.

In a high-pressure CO₂/H₂S environment, a significant difference exists between the effects of varying \( {P}_{{{\text{H}}_{ 2} {\text{S}}}}\) on the corrosion behaviors of tubular steels (Ref 24, 25). Pots et al. (Ref 26) claimed that the effects of H₂S on the corrosion behavior of steel in a CO₂/H₂S environment depend on the CO₂/H₂S ratio, which determines the nature of the scales and the corrosion mechanism. Although no specific critical H₂S concentration provides inhibition, the inhibitory effect may be related to the formation of a corrosion product film (Ref 15, 27). Choi (Ref 28) suggested that the precipitation of an iron sulfide and iron carbonate film is possible in acidic solutions (pH = 3 or 4) because of local supersaturation in regions immediately above the steel surface. Corrosion product films form on the steel surface during the corrosion of the steel under a CO₂/H₂S environment:

During sweet corrosion, the corrosion product film is composed of FeCO₃, which is a typical product of CO₂ corrosion. However, with increases in the partial pressure ratio of H₂S/CO₂, the corrosion process gradually switches to H₂S and CO₂ mix control, and the corrosion product film of Fe _x S is dominant under these conditions (Ref 15).

In 13Cr stainless steel, CO₂ corrosion dominates the corrosion process in conditions without H₂S, and a layer of passive corrosion product film consisting of Cr(OH)₃ and FeCO₃ covers the steel surface (Ref 15). However, this layer of corrosion product film is not tightly adhered to the matrix and can be easily removed from the steel surface without the protection of the inhibitor. Some particles, such as carbides, act as cathodic sites, while the rest of the metal acts as an anode, and then many micro galvanic cells were formed (Ref 14). These heterogeneities sites disrupt the corrosion product film, such that severe pitting occurs without the inhibitor, as shown in Fig. 8(a) and (b). As the inhibitor is added to the solution medium, the metal matrix becomes well protected under the combined effects of the corrosion product film and the inhibitor. Thus, uniform corrosion with no cracks is observed in Fig. 6(a). In the presence of H₂S, chemically dissociated ions, such as HS⁻ and S²⁻, from the additional H₂S are strongly adsorbed on the surface of the steel. On one hand, these HS⁻ and S²⁻ ions react with the passive film to form soluble corrosion products that promote the dissolution of the original passivity film. On the other hand, the FeS _x film generated by the S ^x− and Fe²⁺ cannot protect the matrix since this Fe _x S film competes with the inhibitor covering the surface of steel (Ref 29‐31). Thus, the passive surface film is severely damaged, which would explain why the polarization resistance R _p significantly decreases, as shown in Fig. 5(a). The part of the matrix without film or inhibitor protection will come into contact with the corrosive medium directly and become greatly corroded.

A layer of thin corrosion product film FeCO₃ covers the P110 steel surface under CO₂ corrosion. The severe pitting shown in Fig. 7(a) occurs during sweet corrosion because of the uneven distribution of the corrosion product film formed on the surface of the steel (Ref 14). During sour corrosion, the presence of H₂S dissociates in solution and facilitates the formation of a layer of Fe _x S, which is more compact than the FeCO₃ film and inhibits ion migration (Ref 2). Compared with sweet corrosion, i _corr greatly decreases during sour corrosion because of the protection endowed by the Fe _x S film, which is reflected by the significantly increased R _p shown in Fig. 5(b). In contrast to the 13Cr stainless steel, the layer of Fe _x S film on the P100 steel does not compete with the inhibitor, and the matrix is well protected under the combined effects of the corrosion product film and the inhibitor. The different performances of the corrosion product film Fe _x S and the inhibitor cause 13Cr stainless steel and P110 steel to exhibit different corrosion behaviors in a sweet corrosion environment, i.e., the presence of H₂S accelerates the corrosion of 13Cr stainless steel but inhibits the corrosion of P110 steel.

The reduction of hydrogen ions is one of the main cathodic reduction reactions (Ref 3‐5) in a solution of pH 4. Hydrogen atoms generated from the cathodic process can penetrate into the steel matrix, which results in hydrogen-induced cracking (HIC) and contributes to the increased HIC susceptibility of steels. Thus, SSCC of the steels in a CO₂/H₂S environment may be expected to be based on hydrogen, i.e., HIC. As shown in Figs. 6 and 7, for the 13Cr stainless steel surface shows no cracks during sweet corrosion, whereas the P110 steel surface exhibits several small cracks. These cracks reveal that P110 steel has higher susceptibility toward SSCC than 13Cr stainless steel in a high-pressure CO₂ environment. SSCC cracks prevail under conditions with H₂S, and the 13Cr stainless steel specimen fails under 0.2 and 0.3 MPa H₂S. This result is attributed to high-pressure H₂S enhancing hydrogen ingress into the steel as well as HIC susceptibility, which is related to the increase in absorbed hydrogen content with increasing partial pressure of H₂S (Ref 25). The regularity of change in morphology (SSCC susceptibility) of the 13Cr stainless steel, as revealed by the immersion test, is consistent with that observed in the electrochemical test results (i _pass). Both SSCC susceptibility and i _pass were promoted by increasing \( {P}_{{{\text{H}}_{ 2} {\text{S}}}}\). For P110 carbon steel, however, the increase in SSCC susceptibility seems to be in conflict with the decrease in i _corr as the \( {P}_{{{\text{H}}_{ 2} {\text{S}}}}\) increases from H₂S-free to 0.3 MPa. As previously stated, the decrease in i _corr may be attributed to the FeS _x film covering the steel surface; this layer of film provides protection to the steel. However, hydrogen atoms can still penetrate into the matrix because of their size, regardless of the presence or absence of a film layer. As the H₂S partial pressure increases from 0.1 to 0.3 MPa, more cracks appear on the specimen surface and susceptibility toward SSCC is enhanced with increasing \( {P}_{{{\text{H}}_{ 2} {\text{S}}}}\).

Overall, 13Cr stainless steel exhibited general corrosion while the P110 steel exhibited severe pitting under a high-pressure CO₂ environment. Such results may be due to differences in the protective performance of corrosion product scales on each surface. The presence of H₂S in CO₂-containing environments can promote corrosion risk by facilitating corrosion in the 13Cr steel (increase i _pass) at a rate greater than that induced by pure CO₂ corrosion. By contrast, the corrosion risk of P110 (decrease i _corr) can be decreased by preferentially forming a layer of FeS _x on the film surface; this type of film is more compact than FeCO₃. H₂S also increases SSCC risk by promoting hydrogen atom ingress into the substrate in for 13Cr stainless steel and P110 steel. Development of SSCC is of particular concern because it degrades steel and results in cracking.