Environmental cracking (EC) susceptibility of low-alloy steels with a specified minimum yield strength of 655 MPa (95 ksi) and 758 MPa (110 ksi) manufactured by quenching and tempering heat treatments was investigated in high H2S partial pressures (more than 1.0 MPa) using four-point bend tests in autoclaves. The H2S partial pressures and testing temperatures varied from 1.0 MPa to 10 MPa and 24°C to 150°C, respectively. Materials of grades 95 ksi and 110 ksi containing high Cr and Mo showed no macrocracking under all tested conditions. Localized corrosion occurred at several locations after exposure for 1 month under high H2S pressure and high-temperature conditions. It was concluded that the localized corrosion did not form macrocracking even after long-term (3 months) immersion tests. On the other hand, 110 ksi grade material containing low Cr and Mo suffered from sulfide stress cracking at low temperatures (below 66°C) and at an H2S pressure of 1.0 MPa. The material also showed EC at an H2S pressure of 10 MPa and temperature from 107°C to 150°C. The difference of EC susceptibility among the materials is discussed based on corrosion reactions, hydrogen absorption, and morphologies of the corrosion products on the steel surface.

High-pressure/high-temperature (HPHT) oil and gas fields are being actively developed to meet the worldwide energy demand. High-strength steel is required for the HPHT well application to withstand both high-formation pressure and their own weight. Some of the recent HPHT developments are characterized with high levels of H2S, like the conditions found in the Middle East, China, or the Caspian Sea. It is widely recognized that high-strength low-alloy steel suffers from sulfide stress cracking (SSC), which can lead to material catastrophic failure caused by the hydrogen absorbed from wet H2S environments, which are the so-called sour environments. Therefore, high-strength low-alloy steel with proven resistance to environmental cracking (EC) including SSC is required for HPHT applications.

An aqueous solution containing 5% NaCl and 0.5% CH3COOH saturated with 0.1 MPa H2S specified in NACE TM0177 has been widely used for SSC evaluation of low-alloy steels.1  This standard test environment is useful for testing of quality assurance. Additionally, SSC evaluation by simulating the actual field conditions has recently been used to qualify materials for the specific environments, which are less or more severe than the NACE standard test conditions. The test is called fitness-for-purpose or fitness-for-service testing. There are numerous studies on environmental factors affecting SSC, such as temperature, pH, and H2S partial pressure under ambient pressure. However, there are a limited number of studies on the effects of environmental factors under high H2S pressure beyond the ambient pressure on SSC, and there is a need for assessing the SSC susceptibility of materials under high H2S pressure conditions, considering their application to extreme HPHT oil and gas environments.

It is appropriate to review the phases of H2S to understand the variety of corrosion environments under high H2S field conditions. Figure 1 shows the pressure/temperature phase diagram for the H2S/H2O binary system.2  The line “ac” shows a phase boundary between the liquid and gas phases of H2S. The line terminates at a point “c,” which is called the critical point, where the liquid and gaseous phases become indistinguishable, in what is known as a supercritical fluid. The line “de” shows a boundary between liquid H2S and its hydrate. The definition of a sour environment is condensed water or formation water in equilibrium with the gaseous phase of H2S. The regions with H2S gas and H2S supercritical fluid with an aqueous solution correspond to the sour environment. Ikeda, et al., investigated the behavior of hydrogen adsorption into line pipe steels and the susceptibility to hydrogen-induced cracking (HIC) of the line pipe steels in aqueous solutions saturated with high-pressure H2S and CO2.3-4  They reported that absorbed hydrogen concentration and HIC susceptibility increased with the increase in partial pressure of H2S and CO2. They showed that the aqueous solution saturated with mixed gas of 1 MPa H2S and 1 MPa CO2 is more severe than the standard test conditions specified in NACE TM0177. Kimura, et al., reported test results on hydrogen permeability and susceptibility to HIC of line pipe steels in aqueous solutions containing 5% NaCl saturated with high H2S partial pressure.5  They introduced hydrogen permeation test equipment for high-pressure evaluation. They concluded that high-pressure H2S accelerates the hydrogen ingress into the steel. However, the effect reaches the maximum at 0.5 MPa of H2S and then decreases with increasing H2S pressure. Miyasaka, et al., investigated a correlation between high-pressure H2S environments and aqueous solutions saturated with 0.1 MPa H2S from a viewpoint of sour severity using three-point bent-beam tests for low-alloy steel oil country tubular goods (OCTG).6  They concluded that the severity of NACE standard test condition with 0.1 MPa H2S at room temperature can cover the high H2S pressure conditions. Omura, et al., showed that the critical stress intensity factor, KISSC, values of 758 MPa (110 ksi) grade low-alloy steel OCTG derived by double-cantilever beam (DCB) evaluation decreased when H2S pressure increased from 0.1 MPa to 1.5 MPa at ambient temperature.7  Thebault, et al., showed similar test results for 655 MPa (95 ksi) grade low-alloy steel OCTG.8  However, some of today’s HPHT oil and gas fields contain higher H2S partial pressure, exceeding 1.5 MPa at elevated temperatures. There is no publication detailing corrosion and EC susceptibility of low-alloy steel OCTG in extremely high H2S partial pressure and high-temperature severe sour environments.

FIGURE 1.

Pressure–temperature phase diagram for the H2S/H2O system.2 

FIGURE 1.

Pressure–temperature phase diagram for the H2S/H2O system.2 

In this study, EC susceptibility of low-alloy steel OCTG was evaluated in highly pressurized H2S environments. A four-point bend (4PB) test was used for evaluations in aqueous solutions in equilibrium with H2S pressure up to 10 MPa and temperatures up to 150°C. Corrosion weight loss, absorbed hydrogen concentration in the steel, and corrosion products formed on the steel surface were evaluated. Effects of H2S partial pressure, testing temperature, and test duration on cracking susceptibility were investigated; moreover, the results are discussed based on correlations among corrosion reactions, hydrogen ingress, and protectiveness of corrosion products.

Material and Test Specimen

Low-alloy steel OCTG with a specified minimum yield strength (SMYS) of 655 MPa (95 ksi) and 758 MPa (110 ksi) manufactured by quenching and tempering heat treatments were used in this study. All materials exhibited tempered martensitic microstructure. The chemical compositions and yield strength (YS) are listed in Tables 1 and 2, respectively. The materials 95A and 110A were positioned as sour service grade that does not suffer from SSC under the standard test conditions of 0.1 MPa H2S and 24°C specified in NACE TM0177. On the other hand, the material 110B was positioned as non-sour grade that suffers from SSC under the standard test conditions. Flat plate specimens with 2.0 mm thickness, 10 mm width, and 75 mm length were machined from the pipe samples along the rolling directions. The surfaces of the specimens were ground and degreased prior to the corrosiontests.

TABLE 1

Chemical Compositions of Material (wt%)

Chemical Compositions of Material (wt%)
Chemical Compositions of Material (wt%)
TABLE 2

Yield Strength of Materials

Yield Strength of Materials
Yield Strength of Materials

Test Conditions

The 4PB test was used for investigating the EC susceptibility of materials in high-pressure H2S sour environments. Test conditions are listed in Tables 3 and 4. Their conditions were located in the H2S gas region or the H2S supercritical fluid region with aqueous solution, as shown in Figure 1. However, test condition 7 in equilibrium with 1.5 MPa H2S at 24°C is an unstable environment that may hydrate H2S because of the high pressure of H2S and the low temperature. NaCl aqueous solutions (5.0 wt%) were used for all tests in this study. The pH was not adjusted through addition of any acid. The ratio of the test solution volume to the total surface area of specimens was more than 20 mL/cm2.

TABLE 3

Test Conditions (720 h)

Test Conditions (720 h)
Test Conditions (720 h)
TABLE 4

Test Conditions (2,160 h)

Test Conditions (2,160 h)
Test Conditions (2,160 h)

High-pressure immersion tests were performed using autoclave vessels. Prior to the immersion tests, the autoclave was vacuumed and circulated with nitrogen for several cycles to remove oxygen after the insertion of the specimens. Then, the deaerated test solution was transferred into the autoclave and heated up to the required temperature. For test conditions 1 to 6 in equilibrium with 1.0 MPa H2S, an autoclave was pressurized by H2S gas after heating the vessel. The total pressure in the autoclave was the sum of the water vapor pressure that depends on the testing temperature and H2S partial pressure, as shown in Table 3. In test conditions 7 to 13 in equilibrium with over 1.0 MPa H2S, the required amount of liquid H2S was transferred prior to heating. Testing temperatures ranged from 24°C to 150°C. Omura, et al., and Thebault, et al., indicated that additional pressurization of H2S gas for several days or application of excessive partial pressure of H2S, which was more than the targeted test conditions, was needed for ambient temperature tests to compensate for the H2S partial pressure loss by the dissolution of H2S in the test solution during the test.7-8  However, the total pressure and the H2S pressure remained unchanged under all tested conditions in this study because of the low capacity of the vessels and the low solubility of H2S in the test solution at elevated temperatures.

The 4PB tests were conducted according to European Federation of Corrosion (EFC) 16 in duplicate for the three tested materials and for each test condition.9  The specimens were stressed to 90% of the actual yield strength (AYS) at room temperature in accordance with America Society for Testing and Materials (ASTM) G39.10  The test duration was 720 h for the test conditions shown in Table 3. In addition, long-term tests for 3 months (2,160 h) were conducted under two test conditions at 150°C, as shown in Table 4. After exposure, the specimens were rinsed with deionized water and ethanol, and completely dried to prevent rust formation.

In the case of high-temperature conditions, the applied stress on the specimen relaxes by the change of the mechanical properties of the materials. The Young’s modulus and YS decrease with increasing temperature. In this study, up to 3% of stress drops were expected. On the other hand, the ratio for YS at the testing temperature increased up to 5% because the drop of YS with temperature is bigger than that of Young’s modulus. At 150°C, it was estimated that the applied stress was 87% AYS at room temperature or 95% AYS at 150°C. In addition, the specimen and the loading jig expand at high-temperature conditions. However, because the linear expansion coefficient is small, the contribution to the applied stress is considered to be small in this study.

One specimen was used for weight loss measurement and an assessment of whether cracking was present. After removing the corrosion product by means of a brush, the specimens were weighed. After the weight loss measurements, the specimen was mounted in an epoxy resin, and cross-sectional images were observed using an optical microscope and scanning electron microscope (SEM). The other specimen was used for hydrogen concentration measurement, corrosion product analysis, and micro-check. The specimens were immediately refrigerated after the completion of the immersion test in liquid nitrogen to prevent hydrogen release from specimens. Pieces with 10 mm length were cut from the edge of the 4PB specimens. The corrosion product was removed and cleaned prior to hydrogen analyses. The absorbed hydrogen concentration in the cut sample was measured by thermal desorption analysis (TDA). The sample was heated at a constant heating rate of 10°C/min in a chamber. Degassed hydrogen from the heated sample was detected by quadrupole mass spectroscopy. The remaining portion of the second specimen was used for the analyses of corrosion products. The piece with 5 mm length containing corrosion products was cut and embedded in an epoxy resin to analyze the cross-sectional element distribution in the corrosion products. Electron probe microanalysis (EPMA) was used to determine the distributions. From the remaining portion of the sample, corrosion product powder was prepared by brushing the scale. X-ray diffraction (XRD) measurements were obtained for the scale powder using CoKα radiation to determine the crystal structure of the corrosion products.

Environmental Cracking Susceptibility

In this study, the pH was not adjusted through addition of any acid. Additionally, measuring the pH of the test solution under high H2S partial pressure was experimentally difficult. Therefore, the pH of the solution was calculated using FactSage thermochemical software and databases instead of direct measurement of in site pH.11-12  At 1.0 MPa H2S from 24°C to 150°C, the calculated solution pH ranged from 3.4 to 3.5. At 5.0 MPa H2S from 80°C to 150°C, the pH ranged from 3.0 to 3.1.

The 4PB test results of the overall test matrix as a function of temperature and H2S partial pressure are summarized in Figures 2 through 4. Each duplicate specimen showed the same result. No macrocracking was observed in all tested conditions for materials 95A and 110A (Figures 2 and 3). However, pitting or fissures were observed at high temperature and high H2S partial pressure. They were observed in the stressed areas both near the edge and near the center of the specimen. Material 110B showed cracking in both the low temperature and the H2S supercritical fluid regions (Figure 4). Cracking observed at less than 66°C is considered to be SSC because of a hydrogen embrittlement (HE) mechanism. On the other hand, the cracking observed in the supercritical fluid region may not be SSC.

FIGURE 2.

EC susceptibility of 95A as a function of temperature and H2S partial pressure (circle: 720 h, diamond: 2,160 h). *: pitting. **: fissure. Number: corrosion rate (g/m2/h).

FIGURE 2.

EC susceptibility of 95A as a function of temperature and H2S partial pressure (circle: 720 h, diamond: 2,160 h). *: pitting. **: fissure. Number: corrosion rate (g/m2/h).

FIGURE 3.

EC susceptibility of 110A as a function of temperature and H2S partial pressure (circle: 720 h, diamond: 2,160 h). *: pitting. **: fissure. Number: corrosion rate (g/m2/h).

FIGURE 3.

EC susceptibility of 110A as a function of temperature and H2S partial pressure (circle: 720 h, diamond: 2,160 h). *: pitting. **: fissure. Number: corrosion rate (g/m2/h).

FIGURE 4.

EC susceptibility of 110B as a function of temperature and H2S partial pressure (circle: 720 h, diamond: 2,160 h). **: fissure. Number: corrosion rate (g/m2/h).

FIGURE 4.

EC susceptibility of 110B as a function of temperature and H2S partial pressure (circle: 720 h, diamond: 2,160 h). **: fissure. Number: corrosion rate (g/m2/h).

The cross-sectional observations of each material after testing in the supercritical fluid conditions are shown in Figures 5 and 6. The corrosion products on the material were removed in advance. In the case of materials 95A and 110A, several fissures were observed. The depth of the fissures was sufficiently small (less than 20 μm). The sizes were less than those of non-metallic inclusions contained in the steel. Therefore, they are defined as localized corrosion fissures in this paper. A comparison between the fissures developed at 107°C for 720 h (Figure 5) and those developed at 150°C for 2,160 h (Figure 6) clearly shows that a rise in temperature and a prolongation of the test duration do not influence the depth of fissures. The fissure becomes wide instead of being elongated, and then stress concentration at the tip area decreases. It was confirmed that these pits or fissures did not propagate even if the samples were exposed to the environment for a long time. This means that these pits or fissures did not develop macrocracking in the case of materials 95A and 110A. On the other hand, material 110B did not show pits or fissures but showed macrocracking in supercritical fluid conditions, as shown in Figures 5 and 6. In addition, the corrosion product was observed on the crack surface in the case of the conditions at 150°C for 2,160 h (Figure 6).

FIGURE 5.

Cross-sectional observations of 4PB test specimens tested in 10 MPa H2S at 107°C for 720 h.

FIGURE 5.

Cross-sectional observations of 4PB test specimens tested in 10 MPa H2S at 107°C for 720 h.

FIGURE 6.

Cross-sectional observations of 4PB test specimens tested in 10 MPa H2S at 150°C for 2,160 h.

FIGURE 6.

Cross-sectional observations of 4PB test specimens tested in 10 MPa H2S at 150°C for 2,160 h.

As a summary, it is concluded that the tested materials 95A and 110A are applicable for sour environments with pressure up to 10 MPa H2S and temperature up to 150°C, including H2S supercritical fluid conditions. The estimated pH was below 3.5. On the contrary, it was found that material 110B had cracking susceptibility not only at near ambient temperatures but also at elevated temperatures under high H2S partial pressures, although the strength of 110B is the same as that of 110A.

Corrosion Rate and Hydrogen Concentration

The numbers in Figures 2 through 4 indicate corrosion rate, measured by the weight loss, under each test condition. The corrosion rates remarkably changed depending on the test conditions. As a typical example, the effect of testing temperature on corrosion rates at 1.0 MPa H2S partial pressure is shown in Figure 7. The result shows two different temperature dependencies regardless of the materials. The corrosion rate decreased when the testing temperature was increased up to 80°C. On the contrary, the corrosion rate increased when the temperature was increased from 80°C to 150°C. The corrosion rates at 1.5 MPa H2S and 24°C varied according to the materials. This is considered to be caused by the unstable test conditions which may generate the H2S hydrate. If the hydrate had been generated on the test specimen, the corrosion rate might remarkably decrease.

FIGURE 7.

Effect of testing temperature on corrosion rate at 1.0 MPa H2S partial pressure.

FIGURE 7.

Effect of testing temperature on corrosion rate at 1.0 MPa H2S partial pressure.

The absorbed hydrogen concentrations (mass ppm) of material 110A measured by TDA are shown in Figure 8. The hydrogen concentration decreased with increasing temperature at 1.0 MPa H2S. This implies that the increase in temperature may decrease the risk of SSC from the viewpoints of the hydrogen concentration. Under the conditions of high H2S partial pressure, the hydrogen concentrations were higher than those under 1.0 MPa H2S conditions. This means that the increase in H2S partial pressure increased the risk of SSC.

FIGURE 8.

Effect of testing temperature on absorbed hydrogen concentration of 110A.

FIGURE 8.

Effect of testing temperature on absorbed hydrogen concentration of 110A.

Corrosion Product Analyses

After the immersion tests, corrosion product films mainly comprising sulfide covered the steel surface. Appearances of the corrosion products are shown in Figure 9. The corrosion product formed at 1.0 MPa H2S and 24°C condition was a porous, deposit type. The increase in H2S partial pressure and temperature changed the surface morphologies of the corrosion products to a tight, adherent type.

FIGURE 9.

Surface morphologies of corrosion products on material 95A after corrosion test for 720 h.

FIGURE 9.

Surface morphologies of corrosion products on material 95A after corrosion test for 720 h.

EPMA was performed to investigate the chemical compositions of the corrosion products. The distribution of Fe, S, O, Cr, and Mo in the layer of corrosion products formed on the materials immersed in various sour conditions is shown in Figures 10 through 12. The corrosion products mainly comprised iron sulfide. However, the morphology and distribution of elements were quite different depending on the test conditions. In the case of 1.0 MPa H2S conditions, the corrosion products seemed to be divided into an outer layer and an inner layer (Figure 10). The outer layer comprised porous deposit-type iron sulfide, and the inner layer comprised iron sulfide and iron oxide including Cr and Mo at 24°C. The increase in testing temperature at 1.0 MPa H2S increased the thickness of the inner oxide layer. In addition, highly-concentrated Cr and Mo were observed in the inner layer. When H2S partial pressure was increased from 1.0 MPa to 5.0 MPa or 10 MPa, the double-layer structure could not be observed (Figure 11). A very thin inner oxide layer, an iron sulfide middle layer, and an outer Cr–Mo oxy-sulfide layer were observed. Figure 11 also shows the EPMA of corrosion products formed on material 110B containing low Cr and Mo under 1.0 MPa H2S and 107°C conditions for 720 h. The corrosion product comprised a single iron sulfide layer and a porous structure, which was unlike that formed under the same conditions on material 110A. The increase in test duration to 2,160 h increased the thickness of the sulfide layer and reduced the thickness of the inner oxide layer, especially at 10 MPa H2S (Figure 12).

FIGURE 10.

Sectional SEM observations and alloying element distribution across the corrosion products on material 110A in 1.0 MPa H2S for 720 h.

FIGURE 10.

Sectional SEM observations and alloying element distribution across the corrosion products on material 110A in 1.0 MPa H2S for 720 h.

FIGURE 11.

Sectional SEM observations and alloying element distribution across the corrosion products in 5.0 MPa or 10.0 MPa H2S for 720 h.

FIGURE 11.

Sectional SEM observations and alloying element distribution across the corrosion products in 5.0 MPa or 10.0 MPa H2S for 720 h.

FIGURE 12.

Sectional SEM observations and alloying element distribution across the corrosion products in 5.0 MPa or 10.0 MPa H2S for 2,160 h.

FIGURE 12.

Sectional SEM observations and alloying element distribution across the corrosion products in 5.0 MPa or 10.0 MPa H2S for 2,160 h.

XRD analyses of the corrosion products were performed to investigate the correlation between the crystal structure and the protectiveness against corrosion. The diffraction patterns and classification of crystal structures are shown in Figures 13 through 15.Several kinds of corrosion products were observed depending upon the tested conditions. Figure 13 shows the effect of testing temperature on corrosion products after exposure for 720 h under 1.0 MPa H2S conditions. At 24°C, the corrosion product comprised a mixture of mackinawite (tetragonal Fe9S8) and cubic FeS. At 107°C, it comprised a mixture of mackinawite and pyrrhotite (Fe7S8). At 150°C, only the XRD intensity peaks for pyrrhotite could be observed. In the case of 5.0 MPa H2S, the major corrosion product was pyrrhotite for exposure for 720 h, as shown in Figure 14. The corrosion products transformed to enriched S phases, pyrite (FeS2), after exposure of 2,160 h. In the case of 10 MPa H2S, the XRD patterns were similar to those at 5.0 MPa H2S. Additional peaks for the slight formation of greigite (Fe3S4) were observed. The results of the XRD analyses are summarized in Figure 16. The increase in H2S partial pressure and temperature increased S concentration in iron sulfides. Similar corrosion products were observed on materials95A and 110B.

FIGURE 13.

XRD patterns of corrosion products on material 110A after corrosion test saturated with 1.0 MPa H2S for 720 h.

FIGURE 13.

XRD patterns of corrosion products on material 110A after corrosion test saturated with 1.0 MPa H2S for 720 h.

FIGURE 14.

XRD patterns of corrosion products on material 110A after corrosion test saturated with 5.0 MPa H2S.

FIGURE 14.

XRD patterns of corrosion products on material 110A after corrosion test saturated with 5.0 MPa H2S.

FIGURE 15.

XRD patterns of corrosion products on material 110A after corrosion test saturated with 10 MPa H2S.

FIGURE 15.

XRD patterns of corrosion products on material 110A after corrosion test saturated with 10 MPa H2S.

FIGURE 16.

XRD analyses of corrosion products on material 110A after the corrosion test (those under conditions of 45°C and 65°C with 1.0 MPa H2S, referred to a report by Omura, et al.).7 

FIGURE 16.

XRD analyses of corrosion products on material 110A after the corrosion test (those under conditions of 45°C and 65°C with 1.0 MPa H2S, referred to a report by Omura, et al.).7 

Material 95A and 110A showed no macrocracking in the wide range of conditions from 1.0 MPa H2S at 24°C to 10 MPa H2S at 150°C, as shown in Figures 2 and 3. However, small pitting or fissures were observed under high H2S pressure and high-temperature conditions. On the contrary, material 110B suffered from EC in both low temperature and H2S supercritical fluid regions, as shown in Figure 4. Therefore, the EC mechanisms and the difference of EC susceptibility among the materials are discussed based on corrosion reaction, hydrogen absorption, and morphologies of corrosion products on the steel surface in the following sections.

Morphologies of Corrosion Products

The corrosion products comprised an outer iron sulfide layer and an inner Cr–Mo-enriched oxide layer at 1.0 MPa H2S and 24°C for 720 h, as shown in Figure 10. The double-layer structure was also observed by Thebault, et al., under similar test conditions.8  The formation of the inner oxide layer means that H2O can permeate through the outer iron sulfide layer because of its poor protectiveness. The formation of the inner oxide layer also implies reduction in H2S activity underneath the outer layer resulting from consumption of S by the outer sulfide layer. It is well known that similar double-layer protective films are formed on Ni-based alloys in high H2S partial pressure environments.13  The protective film comprises an inner Cr-rich oxide layer and an outer Ni sulfide layer. The increase in temperature at 1.0 MPa H2S over 80°C increased the thickness of the inner oxide layer, as shown in Figure 10. This corresponds to the increase in corrosion rate with increasing temperature at 1.0 MPa H2S over 80°C, as shown in Figure 7. It is estimated that the inner oxide layer formed under 1.0 MPa H2S was insufficient to suppress the corrosion reaction.

The increase in H2S partial pressure reduced the thickness of the inner oxide layer and changed its morphology to a tighter one. This means high partial pressure of H2S enhanced protectiveness of the outer sulfide layer and prevented H2O ingress through the outer sulfide layer. This corresponds to the decrease in corrosion rate with increasing H2S partial pressure, as shown in Figures 2 through 4. After exposure for 2,160 h, the thickness of the inner oxide layer got remarkably reduced and the corrosion products mainly comprised the sulfide layer, as shown in Figure 12. This means that the increase in the test duration also had a beneficial effect in stabilizing the protectiveness of the outer sulfide layer.

The crystal structure of the corrosion products also changed depending upon environmental parameters, as shown in Figures 13 through 16. Mackinawite and cubic FeS were detected by XRD analyses at 1.0 MPa H2S from 24°C to 107°C after 720 h exposure, as shown in Figure 13. Cansio, et al., also reported the formation of the two types of corrosion products after atmospheric pressure tests in conventional NACE TM0177 solution A or solution B.14  Omura, et al., confirmed that only mackinawite was identified at 1.0 MPa H2S below 65°C for 336 h, while both mackinawite and cubic FeS were found after exposure for 720 h.7  The increase in testing temperature changed the iron sulfide to S-enriched structure (pyrrhotite). The change in the structure seems to correspond with the tendency of change of the corrosion rate, as shown in Figure 7. The corrosion rate decreased when temperature was increased up to 80°C. The major iron sulfide structure in the corrosion products seems to be mackinawite. On the other hand, the corrosion rate remarkably increased when temperature was increased over 80°C with generation of pyrrhotite. Additionally, the increase in H2S partial pressure and testing duration also changed the sulfide to S-enriched structures.

Literature on various iron sulfide species formed on low-alloy steels and their stability and solubility was reviewed by Smith and Joosten.15  Mackinawite is generally viewed as an initial metastable phase, which is formed within a short time period.16-20  A higher corrosion rate was observed when significant levels of mackinawite were present compared with presence of other species because of the formation of non-protective, porous scale. The sequential transformation from mackinawite to cubic FeS, to pyrrhotite (Fe7S8), and finally to pyrite (FeS2) was observed in distilled water without NaCl in equilibrium with H2S, and only mackinawite was identified as the primary phase in 5% NaCl solution in equilibrium with H2S, which meant that the Cl attack prevented the formation of protective S-enriched sulfide scale.16  Ueda reported that the increase in temperature promoted the transformation from mackinawite (tetragonal alpha Fe9S8) to troilite (hexagonal beta FeS) in a 5% NaCl solution saturated with 0.1 MPa H2S with increasing in the corrosion rate.21 

In this study, lower S corrosion products, mackinawite and cubic FeS, were mainly observed in the 5% NaCl aqueous solution in equilibrium with 1.0 MPa H2S at 24°C, which was in agreement with the results of previous studies.16-20  However, it was found that an increase in H2S partial pressure beyond 1.0 MPa, an increase in temperature, and a longer test duration accelerated formation of protective S-enriched sulfide scale, as shown in Figure 16. The protectiveness of the S-enriched scale can be explained by its adhesion and its thermodynamic stability corresponding to insolubility and lower dissolution rate in the solution.

The effects of environmental parameters on morphologies of corrosion products are summarized as follows. The increase in H2S partial pressure and test duration had a beneficial effect by promoting the change of the sulfide microstructure to a stable one, which resulted in the suppression of corrosion. Temperature had both detrimental and beneficial effects. On one hand, the increase in temperature at lower H2S partial pressure (under 1.0 MPa) promoted corrosion because of the poor protectiveness of the sulfide layer. On the other hand, the increase in temperature at high H2S partial pressure (beyond 1.0 MPa) promoted the formation of sulfur-enriched stable sulfide, as summarized in Figure 16.

Environmental Cracking Mechanism

Macrocracking observed on the material 110B at lower temperatures in this study, from 24°C to 66°C, was SSC. It is well known that SSC susceptibility decreased with increasing environmental temperature. One of the reasons of the temperature dependence is the decrease in absorbed hydrogen concentration at higher temperatures, as shown in Figure 8. The effect of temperature on the hydrogen accumulation behavior also affects the SSC susceptibility. The decrease in the temperature increases the local hydrogen concentration at the stress concentration area.22  Therefore, SSC did not occur at 5.0 MPa H2S and 80°C, although the hydrogen concentration was sufficiently high. Additionally, there is a possibility that the corrosion behavior or corrosion products affect the SSC susceptibility because no SSC was observed at 3.0 MPa H2S and 66°C for material 110B. A number of studies on the influences of metallurgical factors on SSC susceptibility exist. It is well known that SSC susceptibility is affected by the strength, alloying elements, impurities, tempering temperature, etc. Only material 110B was considered to have susceptibility to SSC because of these complex factors. Strictly speaking, the YS and the applied stress (90% AYS) of material 110B were higher than those of other materials in this study. Therefore, the high SSC susceptibility might depend on the strength and stress effect.

It is difficult to conclude the mechanism for cracking that occurred at elevated temperatures and in the H2S supercritical region. It is quite unlikely that SSC occurred at high temperature based on the discussion on SSC susceptibility. One possibility of the mechanism is active pass corrosion (APC) type EC. The pitting or fissures are generated at the initial stage. Then, the pitting or fissures propagate by a localized corrosion mechanism assisted by tensile stress. In the case of this mechanism, the general corrosion has to be suppressed by, for example, the protective corrosion product layer because only localized corrosion should progress. The lower corrosion rates at 10 MPa H2S and 107°C compared with that at lower H2S partial pressure, as shown in Figures 2 through 4, support this mechanism. In addition, the corrosion product observed on the cracking, as shown in Figure 6, also supports this mechanism. The corrosion product seems to be formed under high-temperature environments. Another possibility is HE type EC after cooling. HE susceptibility at high temperature is quite low for the same reasons as for SSC. However, hydrogen should be absorbed in the specimen in equilibrium with the H2S partial pressure. When testing temperature quickly drops after testing without release of hydrogen, HE may occur because of accumulation of hydrogen to the stress concentration area at low temperature. The sharp cracking, as shown in Figure 5, looks like HE. However, the initiation site of cracking may be pitting, fissures, or small cracking generated by the first mechanism.

The estimated mechanism of cracking at elevated H2S partial pressure and elevated temperature including the effect of alloying elements is illustrated in Figure 17. In the case of materials 95A and 110A containing high Cr and Mo, localized corrosion such as pitting or fissures was generated at the initial stage of exposure. The localized corrosion did not propagate because of sulfur-enriched corrosion products and sufficient amount of the alloying elements, which enriched the corrosion product layer and suppressed further corrosion. In the case of material 110B containing low Cr and Mo, pitting or fissures should be generated during the exposure like in the case of 110A. However, the insufficient amount of alloying elements to produce the protective corrosion product layer led to the propagation of the localized corrosion or cracking resulting from APC type EC mechanism. When the test duration is sufficiently long, cracking is clearly observed, as shown in Figure 6. On the other hand, when the test duration is short, the cracking is also short. It may cause HE type EC near ambient temperature after the completion of the tests depending on the cooling conditions.

FIGURE 17.

Mechanisms of EC which occurred at elevated temperatures in material 110B containing low Cr and Mo.

FIGURE 17.

Mechanisms of EC which occurred at elevated temperatures in material 110B containing low Cr and Mo.

EC susceptibility of low-alloy steels with SMYS of 655 MPa (95 ksi) and 758 MPa (110 ksi) was investigated in high H2S partial pressures using 4PB tests in autoclaves. The H2S partial pressures and testing temperatures varied from 1.0 MPa to 10 MPa and 24°C to 150°C, respectively. EC mechanisms and the difference of EC susceptibility among the materials are discussed based on corrosion reaction, hydrogen absorption, and morphologies of corrosion products on the steel surface. The results are recapped as follow:

  • 95 ksi grade material and 110 ksi grade material containing high Cr and Mo showed no macrocracking in all tested conditions. Localized corrosion such as pitting or fissures that occurred during the initial stage of exposure did not propagate after long-term (3 months) immersion.

  • 110 ksi grade material containing low Cr and Mo suffered from SSC at low temperature (below 66°C) and 1.0 MPa H2S. The material also showed APC type EC at 10 MPa H2S from 107°C to 150°C.

  • The increase in H2S partial pressure and test duration promoted the change in the morphology of the corrosion products to a tight, adherent type, resulting in the suppression of corrosion.

  • The increase in temperature at 1.0 MPa H2S partial pressure promoted the corrosion reaction because of the poor protectiveness of the iron sulfide layer.

  • Two mechanisms were suggested for cracking observed on 110 ksi grade material containing low Cr and Mo at elevated temperatures. One is APC type cracking resulting from the insufficient protectiveness of the sulfide film. The other is HE type cracking after cooling caused by absorbed hydrogen at elevated temperatures.

Trade name.

The authors thank Dr. Elizabeth Trillo of Southwest Research Institute (SwRI) for performing the high-partial pressure H2S tests. The authors also wish to thank Nippon Steel & Sumitomo Metal Corporation (NSSMC) for allowing publication of this study. The assistance of Ms. Akiko Tomio of NSSMC for calculating pH using thermodynamic software and coworkers in the laboratories is gratefully acknowledged.

1.
ANSI/NACE TM0177
,
“Standard Test Method, Laboratory Testing of Metals for Resistance to Sulfide Stress Cracking and Stress Corrosion Cracking in H2S Environments”
(
Houston, TX
:
NACE International
,
1996
,
2005
,
and
2016
).
2.
F.T.
Selleck
,
L.T.
Carmichael
,
B.H.
Sage
,
Ind. Eng. Chem. Fundam.
44
(
1952
):
p
.
2219
2226
.
3.
A.
Ikeda
,
F.
Terasaki
,
M.
Takeyama
,
I.
Takeuchi
,
Y.
Nara
,
“Hydrogen Induced Cracking Susceptibility of Various Steel Line Pipe in the Wet H2S Environment,”
CORROSION 1978, paper no. 43
(
Houston, TX
:
NACE
,
1978
).
4.
A.
Ikeda
,
T.
Kaneko
,
F.
Terasaki
,
“Influence of Environmental Conditions and Metallurgical Factors on Hydrogen Induced Cracking of Line Pipe Steel,”
CORROSION 1980, paper no. 8
(
Houston, TX
:
NACE
,
1980
).
5.
M.
Kimura
,
N.
Totsuka
,
T.
Kurisu
,
T.
Hane
,
Y.
Nakai
,
“Effect of Environmental Factors on Hydrogen Permeation in Linepipe Steel,”
CORROSION 1985, paper no. 237
(
Houston, TX
:
NACE
,
1985
).
6.
A.
Miyasaka
,
K.
Denpo
,
H.
Ogawa
,
SPE Production Engineering
3
,
4
(
1988
):
p
.
578
583
.
7.
T.
Omura
,
K.
Kobayashi
,
M.
Ueda
,
“SSC Resistance of High Strength Low Alloy Steel OCTG in High Pressure H2S Environments,”
CORROSION 2009, paper no. 102
(
Houston, TX
:
NACE
,
2009
).
8.
F.
Thebault
,
L.
Delattre
,
M.
Horstemeier
,
C.
Bosch
,
J.
Bernard
,
B.
Orlans
,
C.
Linne
,
“Sulfide Stress Cracking Resistance of High Strength Low Alloy Steels for High H2S Pressure Wells: Improvements of the H2S Gas Loading Procedure of the Autoclave and Results for T95 and C110 Steel Grades,”
CORROSION 2012, paper no. 1726
(
Houston, TX
:
NACE
,
2012
).
9.
Guidelines on Materials Requirements for Carbon and Low Alloy Steels for H2S-Containing Environments in Oil and Gas Production, no. 16
, 3rd ed. (
London, England
:
European Federation of Corrosion Publications
,
2009
).
10.
ASTM G39
“Standard Practice for Preparation and Use of Bent-Beam Stress-Corrosion Test Specimens”
(
West Conshohocken, PA
:
ASTM International
,
2016
).
11.
C.W.
Bale
,
P.
Chartrand
,
S.A.
Degterov
,
G.
Eriksson
,
K.
Hack
,
R.
Ben Mahfoud
,
J.
Melançon
,
A.D.
Pelton
,
S.
Petersen
,
Calphad
26
(
2002
):
p
.
189
228
.
12.
C.W.
Bale
,
E.
Belisle
,
P.
Chartrand
,
S.A.
Decterov
,
G.
Eriksson
,
K.
Hack
,
I.-H.
Jung
,
Y.-B.
Kang
,
J.
Melançon
,
A.D.
Pelton
,
C.
Robelin
,
S.
Petersen
,
Calphad
33
(
2009
):
p
.
295
311
.
13.
M.
Ueda
,
Corrosion
62
(
2006
):
p
.
856
867
.
14.
M.J.
Cansio
,
M.L.
Latio
,
T.E.
Perez
,
“Corrosion and Hydrogen Insertion in UNS G41300 Steels in Brine Acid Solutions Saturated with Hydrogen Sulfide,”
CORROSION 2012, paper no. 1426
(
Houston, TX
:
NACE
,
2012
).
15.
S.N.
Smith
,
M.W.
Joosten
,
“Corrosion of Carbon Steel by H2S in CO2 Containing Oilfield Environments,”
CORROSION 2006, paper no. 115
(
Houston, TX
:
NACE
,
2006
).
16.
F.H.
Meyer
,
O.L.
Riggs
,
R.L.
McGlasson
,
J.D.
Sudbury
,
Corrosion
14
,
2
(
1958
):
p
.
109t
115t
.
17.
J.B.
Sardisco
,
W.B.
Wright
,
E.C.
Greco
,
Corrosion
19
,
10
(
1963
):
p
.
354t
359t
.
18.
J.B.
Sardisco
,
R.E.
Pitts
,
Corrosion
21
,
8
(
1965
):
p
.
245
253
.
19.
J.B.
Sardisco
,
R.E.
Pitts
,
Corrosion
21
,
11
(
1965
):
p
.
350
354
.
20.
D.W.
Shoesmith
,
P.
Taylor
,
M.G.
Bailey
,
D.G.
Owen
,
J. Electrochem. Soc.
127
,
5
(
1980
):
p
.
1007
1015
.
21.
M.
Ueda
,
Corros. Eng.
44
(
1995
):
p
.
142
150
.
22.
K.
Kobayashi
,
T.
Omura
,
M.
Ueda
,
K.
Nakamura
,
“Effect of Testing Temperature on SSC Properties of Low Alloy Steel,”
CORROSION 2006, paper no. 127
(
Houston, TX
:
NACE
,
2006
).