Sulﬁde Stress Cracking of C-110 Steel in a Sour Environment

: The scope of this study includes modeling and experimental investigation of sulﬁde stress cracking (SSC) of high-strength carbon steel. A model has been developed to predict hydrogen permeation in steel for a given pressure and temperature condition. The model is validated with existing and new laboratory measurements. The experiments were performed using C-110 grade steel specimens. The specimens were aged in 2% (wt.) brine saturated with mixed gas containing CH 4 , CO 2 , and H 2 S. The concentration H 2 S was maintained constant (280 ppm) while varying the partial pressure ratio of CO 2 (i.e., the ratio of partial pressure of CO 2 to the total pressure) from 0 to 15%. The changes occurring in the mechanical properties of the specimens were evaluated after exposure to assess material embrittlement and SSC corrosion. Besides this, the cracks developed on the surface of the specimens were examined using an optical microscope. Results show that the hydrogen permeation, and subsequently SSC resistance, of C-110 grade steel were strongly inﬂuenced by the Partial Pressure Ratio (PPR) of CO 2 when the PPR was between 0 and 5%. The PPR of CO 2 had a limited impact on the SSC process when it was between 10 and 15 percent.


Overview
In a sour environment, Sulfide Stress Cracking (SSC) corrosion often degrades downhole tubulars by embrittlement which leads to premature failure. This study is aimed at understanding the mechanism of hydrogen diffusion in metals in presence of H 2 S, determining the factors that influence the vulnerability of tubulars to SSC corrosion, and formulating a model to predict SSC corrosion susceptibility of metals. Often SSC corrosion studies [1][2][3][4] are performed at pressures less than 6 bar. The assessment of SSC corrosion is commonly performed applying the standard NACE test Method A with cylindrical specimens. Besides this, other methods have been developed in the industry to assess the SSC corrosion resistance of metals [5]. Cernocky et al. [6] developed an experimental setup for SSC corrosion testing of a minipipe specimen subjected to triaxial stress loading conditions. The test setup simulated the loading conditions that are similar to the actual wellbores. Results were validated with measurements obtained from full-scale experiments.
The life of oil wells critically depends on casings, which protect the integrity of the wellbore. With increasing depth of wells, casing materials with high strength are required. Even though high-strength materials can handle high pressure and stress levels, they could be susceptible to Sulfide Stress Cracking (SSC) corrosion in a sour environment [7]. This type of corrosion rapidly degrades downhole tubulars by embrittlement, and subsequently cracking that causes early failure. SSC is corrosion that is caused often when stressed tubulars are exposed to an H 2 S environment. Because of SSC, tubulars usually fail at stress levels well below their yield stress [7]. As a result, SSC is affected by different environmental and metallurgical factors, including material type, microstructure, pressure, temperature, and the composition and pH of the surrounding solution [7][8][9][10]. Figure 1a displays the operating conditions for Q-125. The area on the left side of the green line represents the safe conditions (i.e., conditions with low partial pressures and high solution pH) in which the vulnerability of the metal to SSC corrosion is minimal while the area on the right side of the red line shows the unsafe conditions in which the material is highly prone to the SSC corrosion. The area in between the two lines represents less certain conditions in which there is a moderate level of vulnerability to the SSC  Figure 1a displays the operating conditions for Q-125. The area on the left side of the green line represents the safe conditions (i.e., conditions with low partial pressures and high solution pH) in which the vulnerability of the metal to SSC corrosion is minimal while the area on the right side of the red line shows the unsafe conditions in which the material is highly prone to the SSC corrosion. The area in between the two lines represents less certain conditions in which there is a moderate level of vulnerability to the SSC corrosion. Comparing the vulnerability of different steels, P-110 demonstrated the highest window of applicability following Q-125. The SSC corrosion resistance of P-110 was better than Q-125 when they were tested at the same conditions. Thus, as anticipated, the grade with higher yield strength was more susceptible to SSC corrosion.
A study conducted by Vera and Case [12] examined the effects of pH and temperature on SSC corrosion resistance of high strength API carbon steel (P-110) varying pH (2.5 to 3.5) at room temperature and changing temperature (25 to 95 • C) at a pH of 2.5. Increasing pH and temperature improved the SSC corrosion resistance of the metal. Other studies [7,[13][14][15][16][17] reported similar observations indicating improvement of the SSC corrosion resistance of steel with an increase in solution pH and temperature.
Effects of Microstructure and Cold Work: A number of investigations [18,19] were performed to study the impact of microstructure on SSC corrosion. A study [19] conducted on materials with varying martensite and bainitic contents indicated that materials with a high martensite content exhibit a higher SSC corrosion resistance. These findings concur with the observations of an earlier study [18]. A more recent SSC corrosion study [20] performed on API-grade pipeline steel demonstrated the impact of microstructure on SSC corrosion. Due to the presence of hard phase cementite at grain boundaries, the ferrite-bainitic microstructure exhibited a lower SSC corrosion resistance as compared to the acicular ferrite microstructure.
Besides microstructure, cold work and the presence of inclusion clusters influence the SSC corrosion resistance. Increasing inclusion clusters in metals exacerbates their susceptibility towards SSC corrosion [21]. Moreover, acicular ferrite exhibits higher SSC resistance than ferrite degenerated pearlite. These findings are in agreement with the observations of a similar study [20]. Furthermore, cold work considerably influences the SSC corrosion resistance of metals. Cold work is regularly conducted on metals to improve their mechanical characteristics such as yield strength, Young modulus, and hardness. When cold work is performed, the grain structure of the material could get distorted, increasing the number of structural dislocations. This causes the loss of ductility and reduction in SSC corrosion resistance [13]. Often heat treatment such as tempering is applied after cold work to recover the SSC corrosion resistance of metals.
Effects of Alloying Element, Hardness, and Fluid Chemistry: In addition to heat treatment, different alloying elements are introduced into the formulation of metals and alloys to improve their metallurgical characteristics including SSC corrosion resistance. Chromium (Cr), Molybdenum, and Nickel are important alloying elements used for improving SSC corrosion resistance. Chromium lessens the rate of hydrogen diffusion into metals [22]. When Nickel content is less than 1%, it does not affect the SSC corrosion resistance of metals [23,24]. However, when it is more than 2%, the yield strength of metals increases, reducing their SSC corrosion resistance [25]. The addition of Molybdenum significantly enhances the SSC corrosion resistance of steels by decreasing the depassivation pH [26]. Metals with a reduced depassivation pH perform better in severe sour environments.
Hardness is one of the commonly applied criteria for selecting steels used in sour environments. Increasing hardness exacerbates the vulnerability of metals towards SSC corrosion. Even if the Rockwell hardness (HRC) limit of 22 is one of the criteria recommended by NACE, it is not a sufficient precondition to prevent SSC corrosion in sour environments. Predominantly, the SSC threshold stress reduces with hardness ( Figure 2), even though there is an abnormal threshold stress trend in the hardness range of 22-24 [27]. The SSC threshold stress is a stress level below which metal vulnerability to failure due to SSC corrosion is minimal in standard corrosive solution. Some case studies [28,29] presented in the literature showed the SSC corrosion failures of soft metals that had a hardness level of less than the specified limit (22 HRC). Corros [27]).
In addition to the characteristics of metals, the surrounding fluid chemistry is a key factor that affects their SSC corrosion resistance. The presence of a corrosion inhibitor significantly increases the critical stress intensity factor, which represents the minimum stress value required to propagate a crack on SSC corroded metal [30]. The effect of salt content on SSC corrosion resistance of steel is often limited [31].

Mechanism of SSC Corrosion
SSC corrosion is a hydrogen embrittlement phenomenon resulting from the combined effects of stress and the diffusion of atomic hydrogen into the lattice structure of metals following corrosion by wet H2S. The requirements for SSC corrosion include (i) tensile stress (imposed by either applied loads or residual stresses); (ii) susceptible material with a structural defect, and (iii) entry of atomic hydrogen through exposure to a corrosive environment containing hydrogen sulfide [32]. SSC corrosion occurs due to the combined effect of a corrosive environment containing hydrogen sulfide and tensile stress, which leads to the formation of cracks when its level exceeds the threshold value. The presence of H2S promotes the penetration of hydrogen atoms into the lattice structure of the metal. Because of the stretching of the metal, the corrosion protective layer regularly breaks and detaches from the metal surface. As the stretching continues, a small molecular opening develops on the surface of the metal through which hydrogen atoms can diffuse into it. In presence of a corrosive solution, the atoms of the metal get oxidized leaving electrons on the surface of the metal. These electrons that are absorbed by hydrogen ions form free H atoms, which could convert to H2 molecules. The following chemical reactions describe the process: Hydrogen sulfide plays a major role in SSC corrosion by impeding the reaction represented by Equation (3). This increases the concentration of hydrogen atoms near the metal surface [33]. Subsequently, the hydrogen atoms enter the metal lattice through openings developed on the surface of the metal due to the stretching of the metal. Consequently, the presence of H2S and tensile stress promotes the entry or permeation of In addition to the characteristics of metals, the surrounding fluid chemistry is a key factor that affects their SSC corrosion resistance. The presence of a corrosion inhibitor significantly increases the critical stress intensity factor, which represents the minimum stress value required to propagate a crack on SSC corroded metal [30]. The effect of salt content on SSC corrosion resistance of steel is often limited [31].

Mechanism of SSC Corrosion
SSC corrosion is a hydrogen embrittlement phenomenon resulting from the combined effects of stress and the diffusion of atomic hydrogen into the lattice structure of metals following corrosion by wet H 2 S. The requirements for SSC corrosion include (i) tensile stress (imposed by either applied loads or residual stresses); (ii) susceptible material with a structural defect, and (iii) entry of atomic hydrogen through exposure to a corrosive environment containing hydrogen sulfide [32]. SSC corrosion occurs due to the combined effect of a corrosive environment containing hydrogen sulfide and tensile stress, which leads to the formation of cracks when its level exceeds the threshold value. The presence of H 2 S promotes the penetration of hydrogen atoms into the lattice structure of the metal. Because of the stretching of the metal, the corrosion protective layer regularly breaks and detaches from the metal surface. As the stretching continues, a small molecular opening develops on the surface of the metal through which hydrogen atoms can diffuse into it. In presence of a corrosive solution, the atoms of the metal get oxidized leaving electrons on the surface of the metal. These electrons that are absorbed by hydrogen ions form free H atoms, which could convert to H 2 molecules. The following chemical reactions describe the process: Hydrogen sulfide plays a major role in SSC corrosion by impeding the reaction represented by Equation (3). This increases the concentration of hydrogen atoms near the metal surface [33]. Subsequently, the hydrogen atoms enter the metal lattice through openings developed on the surface of the metal due to the stretching of the metal. Consequently, the presence of H 2 S and tensile stress promotes the entry or permeation of hydrogen atoms into the metal lattice. The permeated atoms occupy void spaces present in the lattice structure resulting in the embrittlement and failure of the metal. Moreover, the presence of H 2 S facilitates the localized crevice corrosion caused by chloride ions [34].

Modeling SSC Corrosion
A new mathematical model has been developed based on an existing model [35] to predict the susceptibility of materials towards SSC corrosion under different environmental conditions. The model predicts hydrogen atom concentration in the steel matrix. The existing model has been improved to extend its range of applicability. The new model predicts H 2 S concentration using modified Raoult's law which is applicable for a wide range of pressures (0 to 6.2 MPa) while the existing model predicts the solubility based on Henry's law which is valid for low pressures (less than 0.21 MPa). The new model also accounts for the impacts of salt content and solution pH on SSC corrosion.

Model Formulation
The SSC corrosion susceptibility model includes three major calculation steps: (i) determination of H 2 S and CO 2 concentrations in the surrounding solution using solubility model (Appendix A), (ii) predicting of the surrounding solution pH (Appendix B), and (iii) computation of the hydrogen concentration in steel using hydrogen permeation model (Appendix C). The critical hydrogen concentration in steel is utilized as a threshold condition required for the determination of SSC susceptibility of the metal. Along with the solubility of H 2 S, the solubility model predicts CO 2 solubility at a specific pressure and gas-phase composition. The model is appropriate up to 80 • C. The commonly-used API grade steels (Q-125, T-95, and C-110) are not susceptible to SSC corrosion above this temperature [11]. The model is applicable for H 2 S and CO 2 partial pressures of up to 6 and 15 MPa, respectively. The details of the model are presented in Appendix A.

Numerical Procedures
The flowchart presented in Figure 3 displays the calculation steps that are performed to obtain a reliable and unique numerical solution to the SSC corrosion model. The model first (Step 1) determines the solubility of CO 2 in brine solutions at a given CO 2 partial pressure, temperature, and salt concentration applying Appendix A (Equation (A7)). Then, using the concentration of CO 2 obtained from the solubility model, the solution pH is computed as presented in Appendix B (Equation (A25)). Subsequently, the model computes the solubility of H 2 S in brine at a given H 2 S partial pressure, temperature, and salt concentration as presented in Appendix A (Equation (A6)). Using H 2 S concentration calculated in Step 3 and pH calculated from Step 2, the model predicts the concentration of hydrogen atom inside the steel matrix using Appendix C (Equation (A28)). Using yield strength of the metal, the critical hydrogen atom concentration is determined using Equation (A28). Then, comparing the hydrogen atom content of the steel with the critical hydrogen content, the susceptibility of the material towards SSC corrosion is determined.

Validation of SSC model
The new model has been validated extensively by comparing its predictions with published measurements [11,36,37]. The measurements of Morana and Nice [11] and Skogsberg et al. [38] were obtained by performing the NACE Method A tests on API grade carbon steels (P-110 and Q-125) while varying the pH of the corrosive environment (aqueous solution) and the partial pressure of H 2 S in a mixed gas, which was in equilibrium with the solution. The SSC model has been applied to forecast the vulnerability of the materials in these environments. The model predictions are in agreement with published measurements (Figure 4). Some of the specimens which passed the NACE Method A are predicted to be prone to SSC in the given environments. This is because the model provides a conservative estimate since it considers the maximum hydrogen permeation rate as a criterion for SSC failure.

Validation of SSC model
The new model has been validated extensively by comparing its predictions with published measurements [11,36,37]. The measurements of Morana and Nice [11] and Skogsberg et al. [38] were obtained by performing the NACE Method A tests on API grade carbon steels (P-110 and Q-125) while varying the pH of the corrosive environment (aqueous solution) and the partial pressure of H2S in a mixed gas, which was in equilibrium with the solution. The SSC model has been applied to forecast the vulnerability of the materials in these environments. The model predictions are in agreement with published measurements (Figure 4). Some of the specimens which passed the NACE Method A are predicted to be prone to SSC in the given environments. This is because the model provides a conservative estimate since it considers the maximum hydrogen permeation rate as a criterion for SSC failure.    Furthermore, the SSC corrosion model can be utilized to predict the vulnerability of metal towards SSC-related failures under various environmental conditions having different concentrations of CO2 and H2S. Masouri and Zafari [37] presented measurements showing the failure of API grade carbon steel (L-80) under different sour conditions, varying partial pressures of CO2 and H2S in the gas phase. Using the experimental conditions and exposure time, the model is used to predict the susceptibility of the material to SSC failure. The predictions of the model show good agreement with experimental results ( Figure 5).

Experimental Study
In addition to modeling, an experimental study was performed to investigate SSC corrosion of API carbon steel. All experiments were conducted at a temperature of 38 °C and high pressure 41.4 MPa (inside the minipipe). Two different experimental setups (tensile testing apparatus and SSC corrosion test setup) were utilized to evaluate the embrittlement of API grade C-110 carbon steel. The Tensile Strength Testing (TST)

Partial Pressure of H 2 S, kPa
Failed [37] No SSC zone SSC prone zone

Experimental Study
In addition to modeling, an experimental study was performed to investigate SSC corrosion of API carbon steel. All experiments were conducted at a temperature of 38 • C and high pressure 41.4 MPa (inside the minipipe). Two different experimental setups (tensile testing apparatus and SSC corrosion test setup) were utilized to evaluate the embrittlement of API grade C-110 carbon steel. The Tensile Strength Testing (TST) apparatus and SSC corrosion test setup were used in sequence to evaluate the embrittlement of the material.

SSC Corrosion Evaluation Method
Minipipe specimens ( Figure 6) that were cut from an API casing were used for testing SSC corrosion and the associated embrittlement of the test material. To increase accuracy and minimize machining-induced defects, water-jet cutting and milling machines were utilized in the manufacturing of the specimens. The following test procedure was used to evaluate embrittlement occurring after simultaneously exposing the specimens to the corrosive environment and constant stress load: Step 1. To test its fitness for the SSC experiment, the specimen was stretched three times to 80% of its elastic limit using the TST apparatus and examined for cracks and other mechanical defects.
Step 2. The specimen was thoroughly (inside and outside surfaces) cleaned with methyl ethyl ketone and exposed to a sour environment in the SSC corrosion test setup for one week while being subjected to a stress level of 85% of its yield stress and inner part over-pressurization of 13.8 MPa.
Step 3. The specimen was recovered from the SSC corrosion setup and strained to failure in the air using the TST apparatus to determine the mechanical properties of corroded specimen material.
Specimen failure during the corrosion test or change in mechanical properties after exposure indicates if the specimen experiences embrittlement. stress and inner part over-pressurization of 13.8 MPa.
The specimen was recovered from the SSC corrosion setup and strained to failure in the air using the TST apparatus to determine the mechanical properties of corroded specimen material.
Specimen failure during the corrosion test or change in mechanical properties after exposure indicates if the specimen experiences embrittlement.

TST Apparatus
The TST apparatus (Figure 7) was used to strain specimens to failure. The device consists of: (i) structural frame made of four stud bolts and three rectangular flanges; (ii) double-acting hydraulic cylinder; (iii) two minipipe holders used at the top and bottom; (iv) syringe pump with speed controller; (v) measuring instrumentation (displacement sensor and pressure transmitter); and (vi) data acquisition card. During the experiment, a minipipe specimen was screwed to the top and bottom holders. The top holder was directly attached to a flange. The bottom holder was connected to the piston rod of the

TST Apparatus
The TST apparatus (Figure 7) was used to strain specimens to failure. The device consists of: (i) structural frame made of four stud bolts and three rectangular flanges; (ii) double-acting hydraulic cylinder; (iii) two minipipe holders used at the top and bottom; (iv) syringe pump with speed controller; (v) measuring instrumentation (displacement sensor and pressure transmitter); and (vi) data acquisition card. During the experiment, a minipipe specimen was screwed to the top and bottom holders. The top holder was directly attached to a flange. The bottom holder was connected to the piston rod of the hydraulic cylinder using a coupler. The displacement sensor was used to measure the change in specimen length occurring during stretching, which is used to determine the strain. The pressure transmitter was utilized to measure oil pressure during the test. The stress developed in the minipipe during the experiment is calculated using the oil pressure data. According to the ASTM standard, the stress loading rate was maintained at 379 MPa per minute. The stress rate was accurately controlled by a computer equipped with a data acquisition and control system. The computer controlled the rate of stress applied on the minipipe by manipulating the syringe pump speed using a data acquisition system. The air injection line was used to actuate the cylinder for adjusting the piston position during the mounting of the minipipe on the setup.

SSC Corrosion Test Setup
The SSC corrosion test was conducted by simulating the loading conditions in the field. Hence, the test specimen was subjected to complex loading conditions by applying tensile load and inner part over-pressurization at the same time. Both internal and external parts of the specimen were exposed to the corrosive environment. To perform the experiments, a new test setup has been designed and built. The schematic of the setup is shown in Figure 8. The setup consists of: (i) a jacketed SSC corrosion cell; (ii) two gas chambers (GC1 and GC2) used to store 250 mL gas during the test; (iii) gas injection cylinder needed for boosting the supply gas pressure; (iv) heating system, which circulates heating medium through the jacket of the corrosion cell; (v) pneumatic cylinder to apply a constant tensile load; (vi) measurement and instrumentation system to monitor and recover test pressure and temperature; and (vii) data acquisition system. stress developed in the minipipe during the experiment is calculated using the oil pressure data. According to the ASTM standard, the stress loading rate was maintained at 379 MPa per minute. The stress rate was accurately controlled by a computer equipped with a data acquisition and control system. The computer controlled the rate of stress applied on the minipipe by manipulating the syringe pump speed using a data acquisition system. The air injection line was used to actuate the cylinder for adjusting the piston position during the mounting of the minipipe on the setup.

SSC Corrosion Test Setup
The SSC corrosion test was conducted by simulating the loading conditions in the field. Hence, the test specimen was subjected to complex loading conditions by applying tensile load and inner part over-pressurization at the same time. Both internal and external parts of the specimen were exposed to the corrosive environment. To perform the experiments, a new test setup has been designed and built. The schematic of the setup is shown in Figure 8. The setup consists of: (i) a jacketed SSC corrosion cell; (ii) two gas chambers (GC1 and GC2) used to store 250 mL gas during the test; (iii) gas injection cylinder needed for boosting the supply gas pressure; (iv) heating system, which circulates heating medium through the jacket of the corrosion cell; (v) pneumatic cylinder to apply a constant tensile load; (vi) measurement and instrumentation system to monitor and recover test pressure and temperature; and (vii) data acquisition system. During the test, the specimen was first mounted in the SSC corrosion cell. The top side of the specimen was attached to the cell lid. The bottom side of the specimen was connected to the piston rod of the pneumatic cylinder. The inner part of the specimen and the annular space between the cell and specimen were filled with brine (2% NaCl solution). The cell was sealed by tightening the lid. Gas injection lines that were connected to the inner and outer parts of the specimen were attached to the cell. Test gases (methane, carbon dioxide, and hydrogen sulfide) required for the experiment were injected into the inner and outer parts of the minipipe simultaneously using the injection cylinder. This was performed by opening valves V15 and V16 and injecting the gas. During the injection, the gas chambers which were directly connected to the inside and outside of the minipipe During the test, the specimen was first mounted in the SSC corrosion cell. The top side of the specimen was attached to the cell lid. The bottom side of the specimen was connected to the piston rod of the pneumatic cylinder. The inner part of the specimen and the annular space between the cell and specimen were filled with brine (2% NaCl solution). The cell was sealed by tightening the lid. Gas injection lines that were connected to the inner and outer parts of the specimen were attached to the cell. Test gases (methane, carbon dioxide, and hydrogen sulfide) required for the experiment were injected into the inner and outer parts of the minipipe simultaneously using the injection cylinder. This was performed by opening valves V15 and V16 and injecting the gas. During the injection, the gas chambers which were directly connected to the inside and outside of the minipipe were pressurized. The chambers served as gas accumulators because the volumetric capacities of the minipipe and SSC corrosion cell were insufficient to hold the gas phase in addition to the corrosive solution. The SSC corrosion tests were conducted maintaining the inside and outside of the minipipe at different pressures. The pneumatic cylinder was installed below the SSC cell to apply tensile load on the minipipe. The cylinder rod is directly connected to the bottom side of the specimen.

Corrosion Test Procedure
A specimen was first stretched three times to 80% of its elastic limit using the TST apparatus and examined for cracks and other mechanical defects. Then, the specimen was scrubbed using methyl-ethyl ketone to remove dirt materials such as oil and grease. After cleaning, the specimen was mounted on the piston rod and filled with brine (2% NaCl solution). Subsequently, the cell was filled with brine. The specimen was lowered into the cell to the appropriate position and the cell lid was assembled. Then, the cell cover was placed and bolted. The gas injection lines were connected to the cell. To deoxygenate the corrosive solution, the inner and outer parts of the specimen were purged with nitrogen at 13.8 MPa for 30 min. After purging, test gases were injected at the desired pressure into the inner and outer parts of the specimen. The inner part of the specimen was maintained at a higher pressure than its outside. During pressurization, both parts of the specimen were first pressurized simultaneously to the desired minipipe external pressure (27.6 MPa). Then, the outer part of the specimen was isolated and more gases were injected into its inner part to increase the pressure to 41.4 MPa. A differential pressure (13.8 MPa) was maintained between the outer and inner parts of the specimen to detect its failure during the test and also to simulate the actual wellbore condition with a certain level of differential pressure load. A tensile load that produces the stress level of 85% of the yield strength of the specimen material was applied using the pneumatic cylinder. The specimen was left in the cell for one week.
Materials that are vulnerable to SSC corrosion are expected to display cracks within one week of exposure [39]; therefore, the one-week time interval was selected for the experiments. Test parameters such as inner and outer specimen pressures, temperature, and pneumatic cylinder pressure were continuously monitored and recorded using the data acquisition system. Test pressures tend to slightly drop because of gas consumption due to the corrosion process and leaks occurring at the inlet connections. Hence, to maintain constant pressure, gas was injected into both sides of the specimen as needed. After seven days of exposure, the cell was depressurized and the specimen was recovered from the cell and examined for cracks. Table 1 shows the test conditions for SSC corrosion experiments. The minipipe specimen used in Test 2 was the baseline, which was not exposed to the corrosive environment. The baseline specimen was tested using the TST machine to obtain reference (uncorroded material) mechanical properties. The rest of the specimens were exposed to different sour conditions as shown in Table 1. All experiments were conducted at constant temperature (38 ± 1 • C) and H 2 S concentration (280 ± 20 ppm).

TST Test Procedure
Specimens that did not fail during SSC corrosion experiments were stretched to failure using the TST apparatus. For this test, a specimen was mounted on the TST apparatus and stressed gradually at a controlled stress rate until failure. The ultimate tensile strength (UTS) and plastic strain to failure (PSF) were measured (Figure 9). PSF represents the non-elastic part of the total strain before failure [40]. The broken specimen was placed under an optical microscope for examination of the protective layer and the type of failure.

Mechanical Properties
During the investigation, the PPR of CO2 was varied to examine its effect on SSC susceptibility. UTS and PSF are used to assess the susceptibility of material towards SSC in sour conditions. A number of studies [41][42][43][44][45] demonstrated the use of PSF for the assessment of SSC. Figures 10 and 11 present the UTS and PSF of the specimens as a function of the PPR of CO2. The UTS and PSF of the specimens showed an exponential reduction with the PPR of CO2. The relationships are expressed as: PSF = 1.14 + 3.63 .
where xc represents the PPR of CO2 in percentage. The correlations are valid for the PPR of CO2 ranging from 0 to 15%. Test 3 was conducted four times because two of the specimens (Specimens 3 and 10) were broke inside the SSC cell. The UTS and PSF reduced

Mechanical Properties
During the investigation, the PPR of CO 2 was varied to examine its effect on SSC susceptibility. UTS and PSF are used to assess the susceptibility of material towards SSC in sour conditions. A number of studies [41][42][43][44][45] demonstrated the use of PSF for the assessment of SSC. Figures 10 and 11 present the UTS and PSF of the specimens as a function of the PPR of CO 2 . The UTS and PSF of the specimens showed an exponential reduction with the PPR of CO 2 . The relationships are expressed as: where x c represents the PPR of CO 2 in percentage. The correlations are valid for the PPR of CO 2 ranging from 0 to 15%. Test 3 was conducted four times because two of the specimens (Specimens 3 and 10) were broke inside the SSC cell. The UTS and PSF reduced rapidly (exponentially) as the PPR of CO 2 was increased from 0 to 5%. As the PPR of CO 2 was continued to increase above 5%, UTS measurements were stabilized while the PSF data exhibited a gradual reduction. All specimens exposed to the corrosive environment showed a reduction in UTS loss of mechanical strength. This could be explained by considering the effects of hydrogen embrittlement and on the change in mechanical properties (ductility and ultimate tensile strength) of the material. Hydrogen embrittlement is expected to reduce both UTS and PSF [46,47]. These reductions could be because of hydrogen permeation and the associated microstructural changes.  In addition to H2S, the presence of CO2 induces embrittlement in high-strength steels [48]. Hence, the presence of these two gases synergized the embrittlement process leading to significant reductions in UTS and PSF when small quantities of these gases (280 ppm of H2S and 0 to 5% of CO2) were introduced into the corrosive environment. However, when the PPR of CO2 was increased above 5%, the reduction in UTS with PPR vanished. The reduction of PSF with the PPR of CO2 also diminished gradually. These observations can be explained using the hydrogen permeation model. Figure 12 shows the model predicted atomic hydrogen concentration inside the material as a function of the PPR of  In addition to H2S, the presence of CO2 induces embrittlement in high-strength steels [48]. Hence, the presence of these two gases synergized the embrittlement process leading to significant reductions in UTS and PSF when small quantities of these gases (280 ppm of H2S and 0 to 5% of CO2) were introduced into the corrosive environment. However, when the PPR of CO2 was increased above 5%, the reduction in UTS with PPR vanished. The reduction of PSF with the PPR of CO2 also diminished gradually. These observations can be explained using the hydrogen permeation model. Figure 12 shows the model predicted atomic hydrogen concentration inside the material as a function of the PPR of CO2. Experimental data and model show a consistent trend for SSC susceptibility. At high In addition to H 2 S, the presence of CO 2 induces embrittlement in high-strength steels [48]. Hence, the presence of these two gases synergized the embrittlement process leading to significant reductions in UTS and PSF when small quantities of these gases (280 ppm of H 2 S and 0 to 5% of CO 2 ) were introduced into the corrosive environment. However, when the PPR of CO 2 was increased above 5%, the reduction in UTS with PPR vanished. The reduction of PSF with the PPR of CO 2 also diminished gradually. These observations can be explained using the hydrogen permeation model. Figure 12 shows the model predicted atomic hydrogen concentration inside the material as a function of the PPR of CO 2 . Experimental data and model show a consistent trend for SSC susceptibility. At high PPRs of CO 2 (greater than 3%), the pH of the corrosive environment is expected to reduce significantly, facilitating the hydrogen permeation process and leading to a high concentration of atomic hydrogen in the metal. As a result, the UTS and PSF of the material reduced sharply with the PPR of CO 2 at low PPRs (between 0 and 5%) and exhibited minor variation at high PPRs (greater than 10%). Using the model, the critical hydrogen atom concentration (Hc) for C-110 at 38 °C is 14.7 ppm. It indicates the critical PPR of CO2 above which the material is expected to fail due to SSC corrosion. Hence, an SSC failure is anticipated when the PPR of the corrosion environment is greater than 9.7%. Five specimens (Specimens 3, 6, 7, 8, and 10) tested above this CO2 PPR value and 40% of them (Specimens 3 and 10) failed due to SSC. This demonstrates that the model is conservative, as indicated earlier. However, no specimen tested below 9.7% PPR of CO2 failed due to SSC.

Analysis of Crack and Scale Characteristics
Ductile and brittle failures have different crack characteristics. Near the crack region, ductile failures display shear deformation and appear fibrous while brittle failures do not exhibit these characteristics [49,50]. Thus, examination of crack characteristics could be useful to assess the embrittlement of a specimen. The micrographs of the specimens (Figures 13 and 14) were taken using a digital microscope after they were strained to failure during tensile strength test or broken in the SSC cell when they were exposed to a corrosive environment while strained to 85% of their yield stress. The crack tips of a specimen (Figure 13a) tested for 15 days without the presence of CO2 and H2S showed noticeable necking and fibrous regions. Besides this, the cracks propagated with shear deformation that characterizes ductile failure. The crack tips of the uncorroded specimen ( Figure 13b) exhibited similar ductile characteristics. On the other hand, a specimen tested at 10% PPR of CO2 (Figure 13c) did not exhibit necking or shear deformation around the crack region, indicating the absence of ductile failure. The crack displayed almost no plastic deformation of the material at the crack tip confirming the presence of brittle failure. Using the model, the critical hydrogen atom concentration (H c ) for C-110 at 38 • C is 14.7 ppm. It indicates the critical PPR of CO 2 above which the material is expected to fail due to SSC corrosion. Hence, an SSC failure is anticipated when the PPR of the corrosion environment is greater than 9.7%. Five specimens (Specimens 3, 6, 7, 8, and 10) tested above this CO 2 PPR value and 40% of them (Specimens 3 and 10) failed due to SSC. This demonstrates that the model is conservative, as indicated earlier. However, no specimen tested below 9.7% PPR of CO 2 failed due to SSC.

Analysis of Crack and Scale Characteristics
Ductile and brittle failures have different crack characteristics. Near the crack region, ductile failures display shear deformation and appear fibrous while brittle failures do not exhibit these characteristics [49,50]. Thus, examination of crack characteristics could be useful to assess the embrittlement of a specimen. The micrographs of the specimens (Figures 13 and 14) were taken using a digital microscope after they were strained to failure during tensile strength test or broken in the SSC cell when they were exposed to a corrosive environment while strained to 85% of their yield stress. The crack tips of a specimen (Figure 13a) tested for 15 days without the presence of CO 2 and H 2 S showed noticeable necking and fibrous regions. Besides this, the cracks propagated with shear deformation that characterizes ductile failure. The crack tips of the uncorroded specimen (Figure 13b) exhibited similar ductile characteristics. On the other hand, a specimen tested at 10% PPR of CO 2 (Figure 13c) did not exhibit necking or shear deformation around the crack region, indicating the absence of ductile failure. The crack displayed almost no plastic deformation of the material at the crack tip confirming the presence of brittle failure. The micrographs of Specimens 4, 5, 7, and 9 that were taken after exposure or straining to failure are presented in Figure 14. Specimens 4, 5, and 9, which were tested at low PPRs of CO2 (≤5%), exhibited shear deformation and fibrous regions near the crack tips. Specimen 4, which was tested at 0 PPR of CO2 displayed the highest plastic deformation before failure. However, Specimen 7, which was corroded at 15% PPR of CO2, demonstrated strong brittle behavior with a low PSF value and lack of shear deformation and fibrous regions near the crack tip. The micrographs of the corrosion scales of Specimens 3, 4, 5, 7, and 9 are presented in Figure 15. The images were obtained considering a square area (2 mm × 2 mm) on the surface of the specimens as shown in Figures 13 and 14. Specimen 3 and 5 were mostly covered with thick flake-like and fragile corrosion scales. Scattered structures that look like fish-eyes were observed on the surface of Specimen 3. Corrosion scales formed on the surface of Specimens 4, 7, and 9 were thin and scattered. The micrographs of Specimens 4, 5, 7, and 9 that were taken after exposure or straining to failure are presented in Figure 14. Specimens 4, 5, and 9, which were tested at low PPRs of CO 2 (≤5%), exhibited shear deformation and fibrous regions near the crack tips. Specimen 4, which was tested at 0 PPR of CO 2 displayed the highest plastic deformation before failure. However, Specimen 7, which was corroded at 15% PPR of CO 2 , demonstrated strong brittle behavior with a low PSF value and lack of shear deformation and fibrous regions near the crack tip. The micrographs of Specimens 4, 5, 7, and 9 that were taken after exposure or straining to failure are presented in Figure 14. Specimens 4, 5, and 9, which were tested at low PPRs of CO2 (≤5%), exhibited shear deformation and fibrous regions near the crack tips. Specimen 4, which was tested at 0 PPR of CO2 displayed the highest plastic deformation before failure. However, Specimen 7, which was corroded at 15% PPR of CO2, demonstrated strong brittle behavior with a low PSF value and lack of shear deformation and fibrous regions near the crack tip. The micrographs of the corrosion scales of Specimens 3, 4, 5, 7, and 9 are presented in Figure 15. The images were obtained considering a square area (2 mm × 2 mm) on the surface of the specimens as shown in Figures 13 and 14. Specimen 3 and 5 were mostly covered with thick flake-like and fragile corrosion scales. Scattered structures that look like fish-eyes were observed on the surface of Specimen 3. Corrosion scales formed on the surface of Specimens 4, 7, and 9 were thin and scattered. The micrographs of the corrosion scales of Specimens 3, 4, 5, 7, and 9 are presented in Figure 15. The images were obtained considering a square area (2 mm × 2 mm) on the surface of the specimens as shown in Figures 13 and 14. Specimen 3 and 5 were mostly covered with thick flake-like and fragile corrosion scales. Scattered structures that look like fish-eyes were observed on the surface of Specimen 3. Corrosion scales formed on the surface of Specimens 4, 7, and 9 were thin and scattered.

Conclusions
The following conclusions can be drawn based on experimental and modeling investigations performed in this study:

Conclusions
The following conclusions can be drawn based on experimental and modeling investigations performed in this study:

•
Variation in the PPR of CO 2 affected the embrittlement of steel considerably when the PPR was between 1%-5% at H 2 S concentration of 280 ppm. When the PPR of CO 2 was between 10 and 15 percent, the deterioration of mechanical properties by the CO 2 content became saturated.

•
The trend of hydrogen atom concentration with the PPR of CO 2 predicted by the SSC model is consistent with those of measured UTS and PSF. The existence of CO 2 decreases the pH of the surrounding solution and expedites the SSC process.

•
In the absence of CO 2 , the pH of the corrosion environment is expected to be close to neutral, and only limited hydrogen ions present in the solution. As a result, experiments did not show noticeable embrittlement of specimens tested at low PPRs of CO 2 in seven days of exposure. • The SSC model predictions are mostly in agreement with existing and new measurements even though it provides a conservative forecast of failure.

Acknowledgments:
The authors would like to thank the Bureau of Safety and Environmental Enforcement (BSEE) for sponsoring this project (E12PC00035) and the University of Oklahoma for providing the necessary support. The technical contribution of the industry advisory board members is also highly appreciated.