2022 Volume 62 Issue 4 Pages 740-749
TMCP (thermo-mechanical controlled process) linepipes have long been used in severe sour environments, but recently sulfide stress cracking (SSC) caused by local hard zones has become a concern. In order to clarify the hardness threshold that leads to SSC, four-point bend (4PB) SSC tests as specified in NACE TM0316 were conducted under several H2S partial pressure conditions. At H2S partial pressures of 1 bar and higher, the surface hardness threshold (at 0.25 mm from the surface) observed in 4PB SSC specimens without SSC cracking was approximately correlated to a maximum acceptable hardness level of 250 HV0.1. A stable low surface hardness of 250 or less HV0.1 was achieved by suppressing hard lath bainite (LB) and obtaining a soft granular bainite (GB) microstructure, resulting in a superior SSC-resistant property. It was found that a SSC crack propagated when the surface hardness increased due to an increasing volume fraction of the LB microstructure. Under a 16 bar H2S partial pressure condition, the crack growth rate increased in the sour environment, and hydrogen embrittlement by H2S was promoted. However, in the 4PB SSC test at 16 bar, the stress concentration and the transition to a crack were suppressed because the shape of localized corrosion was semicircular due to low localized corrosivity. This can be the reason why the SSC susceptibility was similar to 1 bar condition, especially in the 4PB SSC test using the samples with lower surface hardness level of 250 or less HV0.1.
In drilling of offshore oil and gas fields, linepipe (UOE steel pipe) manufactured from steel plates is used in pipelines for oil and natural gas transportation.1) In addition to strength and toughness, corrosion resistance is a required property of linepipes.2) Especially in wet sour environments containing corrosive gases such as hydrogen sulfide (H2S) and carbon dioxide (CO2), sufficient resistance against hydrogen induced cracking (HIC) and sulfide stress cracking (SSC) is required.3,4,5)
SSC originated in Alberta, Canada in the late 1940s.6) In the 1980s, measures against SSC were taken in linepipes made from thick steel plates by the thermo-mechanical controlled process (TMCP) utilizing controlled rolling and controlled cooling.5) Since that time, many studies have been carried out, and it is known that SSC is mainly affected by the three factors of material,7,8,9) sour environment,10,11) and applied stress.12,13) In order to avoid SSC, a hardness limit is applied to the material in the design of sour linepipes, as provided in the NACE MR 0175/ISO 15156-1 standard, and a hardness upper limit of 22 HRC (about 250 HV10) is specified for carbon steel and low alloy steel.14) API Spec 5L, which is a general linepipe standard, specifies that the hardness at positions 1.5 mm from the inner and outer surfaces of the pipe shall be 250 HV10 or less. These hardness specifications have been applied to many sour linepipes, and safety against SSC has been demonstrated in tests in conventional H2S environments having a H2S partial pressure of 1 bar or less. However, SSC tests have rarely been conducted in severe H2S environments above 1 bar, especially for low alloy steel linepipes, and many points concerning material compatibility remain to be clarified.
In recent years, SSC has occurred in H2S environments which were more severe than the conventional environment and exceeded 1 bar. As one reason for this, the focus was placed on the formation of hardened portions with depth of less than 1 mm in the pipe inner surface region of the linepipes manufactured by TMCP, and it became necessary to conduct a new investigation into the effects of the surface hardness and H2S partial pressure on SSC.15) SSC resistance cannot be evaluated only by the hardness HV10 at 1.5 mm from the surface, which is the conventional hardness measurement position, and the micro Vickers hardness HV0.5 or HV0.1 at the 0.25 mm or 0.5 mm positions near the surface layer has been proposed.16,17) Since the indentation size becomes smaller as the load of the Vickers hardness measurement decreases, the hardness in the microscopic region near the surface layer can be determined. In the past, the influence of local hardening near this surface layer on surface corrosion behavior and SSC crack initiation and propagation was unclear.
The mechanism of SSC is considered to be either Hydrogen Embrittlement (HE)-dominant18) or a combined form of HE and Active Path Corrosion (APC),19) but the detailed mechanism is not adequately understood when SSC is divided into initial corrosion and crack formation. In particular, the H2S partial pressure condition in a wet sour environment would presumably change the contributions of HE and APC, but the effects of the H2S partial pressure condition on the SSC mechanism remain unclear. To clarify this issue, it is necessary to separate and consider the initial corrosion behavior and the process of crack formation including crack initiation and propagation.
In this study, first, the effect of the surface hardness distribution of the inner surface of a low alloy linepipe (in particular, the micro hardness HV0.1 at the 0.25 mm position of the inner surface of the pipe) and the H2S partial pressure condition on the SSC resistance is described. The purpose of this study was to clarify the effect of the H2S partial pressure condition on SSC resistance, including the SSC mechanism, assuming mainly the X65 grade GB microstructure. Concretely, the effect of the H2S partial pressure on corrosion behavior in sour environments was confirmed by a four-point bend (4PB) SSC test, and the above-mentioned results were verified by electrochemical measurement (polarization measurement). In addition, the effect of the H2S partial pressure on crack propagation characteristics was also evaluated by a notched SSRT test and strain increment fracture test in sour environments.
X65 grade linepipe steels with thicknesses of 20 to 30 mm were used. Coupons with a size of 300 mm square were cut out from pipes produced with different surface cooling rates under TMCP conditions in the production of thick steel plates, and aging treatment simulating coating at 250°C for 1 h was applied. Figure 1 shows the typical microstructures of the inner pipe surface at various cooling rates. In X65 grade linepipes, the surface microstructure usually forms a bainitic microstructure. The bainite microstructure is generally based on bainitic ferrite (BF) formed in a lath-like shape, and is often classified as upper bainite if carbides precipitate at the BF interface or does not precipitate, and lower bainite if carbides precipitate within the BF grains.20) However, in low carbon steel such as linepipe steel, BF grows into a granular form when the cooling speed is slow, and therefore is classified here as lath bainite (LB) or granular bainite (GB), depending on the morphology of the BF. A high surface cooling rate of more than 200°C/s results in a hard LB microstructure, while a low cooling rate of less than 50°C/s results in a soft granular bainite GB microstructure, and an intermediate cooling rate of about 100°C/s leads to a mixed microstructure of LB and GB. Figure 2 shows an example of the hardness distribution on the inner surface of the pipe. Pipes made from steel plates cooled at a high cooling rate of 100°C/s or more tended to show high hardness over a wide range both in the thickness direction and in the width direction. On the other hand, when the cooling rate was 50°C/s or less, the hardness was low over a wide range in both directions. In particular, at this low cooling rate, hardness is stably below 250 HV0.1 at 0.25 mm from the surface. Specimens of 5 mm thickness were taken from the inner surface of these aged pipes, and 4PB SSC tests were conducted.
Effect of surface cooling rate on surface microstructure.
Effect of surface cooling rate on surface hardness distribution.
Figure 3 shows an example of the hardness distribution measured from the inner surface of the pipe in the pipe thickness direction. Regardless of the cooling rate of the surface layer, the HV0.1 hardness decreased in the inner pipe at positions within 1.0 mm in the surface layer, the hardness was on almost the same level, and the microstructure was mainly GB. Round bar tensile test pieces, notched round bar tensile test pieces, and CT test pieces, which will be described later, were taken from this steady hardness position and used in the respective tests.
Hardness distribution in thickness direction of pipes with three different cooling rates.
In order to evaluate SSC performance, 4PB SSC tests were conducted under various H2S partial pressure conditions in accordance with the NACE TM 0316 standard.21) 4PB test pieces of 5 mm thickness were taken from the inner surface of a pipe as received condtion with surface iron oxide. The SSC test conditions are given in Table 1. The severity map of the sour environment in Fig. 4 also shows the SSC test conditions. In a study using the NACE TM 0177A solution (Sol. A),22) the partial pressure of H2S was in the range of 1 to 16 bar. For the H2S partial pressures of 8 and 16 bar, a gas mixture with 5 bar CO2 was used. In the test using the NACE TM0177B solution (Sol. B),22) the partial pressure of H2S was in the range of 0.07 to 0.13 bar and the total gas pressure was adjusted to 1 bar by mixing CO2. The NACE TM 0177 buffered solution was used to target a partial pressure of 0.15 bar for H2S, a pressure of 0.85 bar for CO2, and a starting pH of 3.1. Figure 5 shows a schematic diagram of the jig for the 4PB SSC test. The inner surface side of the pipe was set as the tensile side of four-point bending, and stress of 90% AYS (Actual Yield Strength) (460 to 486 MPa) of the pipe before aging was applied. The test time was 720 h. After the test was completed, the central part of the test piece was cut and polished to evaluate the presence of cracks. Hardness HV0.1 at 0.25 mm from the surface was measured at a pitch of 1 mm using the same cut test pieces, and was evaluated at the maximum value of a 30 point measurement. The weight loss after the test was measured to evaluate the amount of corrosion after the test. The corrosion products were collected after the test, and the surface corrosion products were analyzed by XRD (X-ray diffraction) measurement. The X-ray diffraction condition was 40 kV-200 mA of the CuKα line. Phase identification of the corrosion products by Raman spectroscopy was also performed. An excitation wavelength of 532 nm was used in the Raman spectroscopy, and was measured through a 100× objective lens.
| Test solution (NACE TM0177) | pH (start/final) | Partial pressure (bar) | Duration (hr) | |
|---|---|---|---|---|
| H2S | CO2 | |||
| Solution A (5.0 wt%NaCl+0.5 wt%CH3COOH) | 2.6 to 2.8/<4.0 | 1 | – | 720 |
| 8 | 5 | |||
| 16 | 5 | |||
| Solution B (5.0 wt%NaCl+2.5 wt%CH3COOH+0.41 wt%CH3COONa) | 3.4 to 3.6/<4.0 | 0.07 | 0.93 | 720 |
| 0.10 | 0.90 | |||
| 0.13 | 0.87 | |||
| 1.3 | 3.5 | |||
| 3.3 | 5.1 | |||
| Buffered solution chemistry (5.0 wt%NaCl+5.0 wt%CH3COOH+0.40 wt%CH3COONa) | 3.0 to 3.2/<3.5 | 0.15 | 0.85 | 720 |
SSC severity map and test conditions.
Schematic illustration of 4-point bend loading jig.
In order to evaluate the effect of the H2S partial pressure on the initial corrosion behavior, electrochemical tests were conducted under constant tensile load conditions, and polarization curves were measured according to NACE TM 0177 Method A.22) A pipe after aging with a high cooling rate exceeding 200°C/s was used. The specimen geometry is shown in Fig. 6. A round bar tensile test piece was taken at 5 mm from the inner surface of the pipe. The surface of the test piece was covered so that only a certain area was exposed. The test conditions are given in Table 2. A constant load of 90% of AYS was applied, and tests were carried out under the conditions of 0.15 bar H2S, NACE Sol. B, pH 3.1 and 1 bar H2S, NACE Sol. A, pH 2.7. Polarization was measured at +500 mV or −500 mV from the open circuit potential (OCP) at a sweep rate of 20 mV/min based on the OCP at that time point after holding at the OCP for a period of 72 h from saturation of the test gas. In order to investigate the change in corrosion behavior over time, a polarization measurement was also performed only on the +500 mV (vs. OCP) anode side from the OCP at a sweep rate of 20 mV/min based on the OCP at that time point after a short immersion period of 0.5 h. A saturated KClAg/AgCl electrode was used as the reference electrode, and a platinum electrode was used as the counter electrode.
Specimen geometry.
| Test solution (NACE TM0177) | pH (start/final) | Partial pressure (bar) | Duration (hr) | |
|---|---|---|---|---|
| H2S | CO2 | |||
| Solution A (5.0 wt%NaCl+0.5 wt%CH3COOH) | 2.6 to 2.8/<4.0 | 1 | – | 72 |
| 0.5 | ||||
| Solution B (5.0 wt%NaCl+2.5 wt%CH3COOH+0.41 wt%CH3COONa) | 3.0 to 3.2/<4.0 | 0.15 | 0.85 | 72 |
| 0.5 | ||||
In order to clarify the contribution of APC and HE as the driving forces in the crack propagation process, constant potential SSRT (Slow Strain Rate Testing) tests were conducted according to NACE TM0198.23) SSRT is a test method in which continuous plastic deformation is applied to a test piece by tension at a low strain rate to break the material. The SSRT method has both the aspect of an accelerated test method for environmentally assisted cracking including SSC, and the aspect of a test method mainly for elucidation of the crack propagation process, as a method for controlling the generation rate of slip step at a crack tip. A pipe after aging at a high cooling rate exceeding 200°C/s was used. The schematic of the notched round bar tensile test specimen is shown in Fig. 7. Specimens were taken at 5 mm from the inner surface of the pipe. The test conditions were a solution environment of 0.15 bar H2S + 0.85 bar CO2, NACE Sol. B (pH 3.1), that is, a sour environment, or air, and a strain rate of 4 × 10−7/s. In the SSRT test, the phenomenon of the crack propagation process itself is the same as that in the 4PB SSC test, but since a constant strain rate is applied continuously, the SSRT test is a more severe environment than the 4PB SSC test. Here, a low H2S environment was chosen because the effect of the potential change may not be clear in a high H2S environment. The test was also performed with some test pieces while applying a constant potential of +50 mV, +100 mV, −50 mV, or −100 mV to the natural immersion potential (−644 mV vs. SCE). A saturated calomel electrode was used as the reference electrode, and a platinum electrode was used as the counter electrode.
Schematic of notched round bar tensile specimen.
To evaluate SSC propagation characteristics, the ratio of the fracture time in the sour environment to the fracture time in air was measured. For example, an early fracture means the material has high susceptibility to embrittlement in a sour environment. The dominant driving forces (APC and HE) in the crack propagation process can be determined from the relationship between the constant potential condition and the fracture time.
2.5. Strain Increment Fracture Toughness Test MethodIn order to investigate the effect of the sour environment on the crack initiation and propagation process, the strain increment type fracture toughness test was carried out. Focusing on the H2S partial pressure, the test gas was set at 0.15 bar-H2S + 0.85 bar-CO2, the standard condition of 1 bar-H2S, or 16 bar-H2S + 5 bar-CO2. The sample was a pipe after aging with a slow cooling rate of less than 50°C/s. A schematic of the CT (Compact Tension) specimen according to ASTM E 1820/2 is shown in Fig. 8. The specimen thickness was 12.7 mm. Specimens were taken from the center of the pipe, and the crack growth direction was in the pipe circumferential direction. The test solution was NACE Sol. A, and the pH was set to 2.7. The loading rate dK/dt until crack initiation was assumed to be constant at 0.005 Nmm−3/2/s, which was the lowest value of fracture toughness in tests performed in advance at various loading rates. The test was conducted using a single sample method, and the test pieces were pre-immersed in the solution for 4 days in order to obtain a steady concentration of hydrogen near the crack tip. The crack length was monitored by measuring the direct current potential drop (DCPD), which was then converted to the crack length by Johnson’s equation (ASTM E 1457-0724)). After the test, the specimens were fractured, and the actual initial and final crack lengths were measured under a stereomicroscope and corrected according to ASTM E 1457-07.24) The crack mouth opening displacement (CMOD) was calculated, and stress intensity factor K and fracture toughness value J were obtained.
Schematic of CT specimen.
Figure 9 shows the results of the 4PB SSC test. The figure shows the maximum value of hardness HV0.1 at 0.25 mm from the surface after the test and the H2S partial pressure and occurrence of cracks for each specimen. At H2S partial pressures of 1 bar or more, the limit of surface hardness at which no crack occurs in the 4PB SSC test was approximately 250 HV0.1. At 0.13 bar, the limit was about 270 HV0.1, and no SSC was observed at 0.07 bar. SSC occurred predominantly in samples with high cooling rates above 200°C/s, and at 16 bar, SSC occurred in some samples with cooling rates of 100°C/s. Samples with a high cooling rate exceeding 200°C/s showed a high hardness of more than 250 HV0.1 over a wide area in the thickness direction (approximately 1 mm) and width direction (approximately 10 mm), indicating that cracks tend to propagate in the thickness direction when SSC occurs. Among the samples with a high cooling rate exceeding 200°C/s, cooling of the samples indicated by the square marks was stopped in the low temperature region of <350°C, and the hardness HV0.1 at the 0.25 mm position tended to be higher than when cooling was stopped at 350°C or more (samples shown by diamond marks), and SSC occurred in the former samples except under the condition of 0.07 bar.
Effect of H2S partial pressure and surface hardness on SSC in 4-point bend test.
Summarizing these results, when the hard LB microstructure was suppressed and the GB microstructure was formed as the main component, it was found that the hardness at the 0.25 mm position of the surface layer was 250 HV0.1 or less, and good SSC resistance performance was obtained independent of the H2S partial pressure condition. However, under the H2S partial pressures of 1 bar or more, when the volume fraction of the LB microstructure increased and the surface hardness of the pipe increased to more than 250 HV0.1, SSC occured easily with increasing H2S partial pressures. The relationship between the hardness HV0.1 at 0.25 mm from the surface and SSC resistance generally agreed with the relationship7) between the hardness HV10 at 1.5mm from the surface and the SSC resistance under the conventional condition of a H2S partial pressure of 1 bar or less. However, as a new finding, this research revealed that SSC resistance performance differs depending on the difference of the microstructure and the distribution of hardness HV0.1 in the range within 1.5 mm of the surface layer, suggesting that control of the microstructure and hardness of this surface layer is very important for SSC resistance.
3.1.2. Evaluation of Corrosion Products after Four-Point Bend SSC TestPhotographs of the surface and cross section of the specimen after the 4PB SSC test are shown in Fig. 10. A thin corrosion product layer with a thickness of 10 μm or less was deposited on the steel surface under the condition of 0.15 bar H2S. On the other hand, a large amount of corrosion products were deposited on the surface of the steel under the conditions of 1 bar and 16 bar H2S, and corrosion product layers with thicknesses of 100 to 200 μm and 50 to 100 μm, respectively, were observed. The effect of the H2S partial pressure on the post-test corrosion loss and maximum localized corrosion depth is shown in Fig. 11. Corrosion loss and the local corrosion depth were highest at 0.15 bar and lowest at 1 bar, indicating that 0.15 bar is the most severe and 1 bar is least severe condition for both total and local corrosion. Subsequently, XRD and Raman analyses of the corrosion products after the 4PB SSC test were conducted. The results are shown in Fig. 12. In the 0.15 bar H2S environment, the corrosion product was found to be mainly γFe2O3, while at 1 bar H2S, the composition was mainly FeS, and a slight amount of Fe3S4 was also observed. At 16 bar H2S, a further increase in the ratio of Fe3S4 the 0.25 mm position from the peak intensity. It is generally known that FeS-based corrosion products exhibit corrosion protection.25) Based on the fact that the thickest corrosion product layer was deposited under the 1 bar H2S environment (Fig. 10), the low corrosion amount in the 1 bar H2S environment is estimated to be due to the protective effect of this thick corrosion product layer mainly composed of FeS.
Results of surface and cross-sectional observation under each H2S partial pressure condition.
Corrosion weight loss and maximum corrosion depth under each H2S partial pressure condition.
Results of XRD and Raman analysis.
The results of polarization measurements obtained with retention times of 72 h and 0.5 h in the 0.15 bar and 1 bar H2S sour environments are shown in Figs. 13 and 14, respectively. The result of the polarization measurement at 72 h showed no clear difference in the cathodic polarization curve under the two H2S partial pressure conditions. However, the anodic polarization curve indicated that the anodic current was suppressed under the condition of 1 bar H2S. From the comparison of the anodic polarization curves measured at 0.5 h and 72 h, suppression of the anodic current increased with the longer immersion time. In particular, the anode suppression by the long 72 h immersion was remarkable at 1 bar. The results of this polarization measurement are consistent with the relationship of the H2S partial pressure to corrosion loss (Fig. 11) in the 4PB SSC test described in the previous section. In other words, in comparison with the 0.15 bar environment, it was found that the formation of corrosion products with a higher protection property in the 1 bar environment suppresses the anode reaction, and consequently, the amount of corrosion is also suppressed.
Polarization property in sour environments (0.15 and 1.0 bar H2S).
Effect of immersion time on polarization property in sour environment (0.15 and 1.0 bar H2S).
Figure 15 shows the load-time curve in the constant potential SSRT test. Significant premature fracture was observed in the sour solution environment compared with the air environment. Figure 16 shows the relationship between the fracture time ratio (time to fracture in sour environment/air) and the potential in the constant potential SSRT test. Without potentiostatic operation (0 mV), the fracture time ratio was 31%. The fracture time ratio showed a minimum value at ±50 mV on both the cathode side and anode side, and the fracture time was prolonged when a larger potential (±100 mV) was applied. This result means that both the cathode reaction (hydrogen generation reaction) and the anode reaction (iron dissolution reaction) are driving forces of SSC propagation, and the propagation process is controlled by the balance of the two reactions. In addition, since the fracture time was shortest under the cathode condition of −50 mV, it was found that the cathode reaction contributed to the propagation phenomenon more dominantly than the anode reaction. In other words, the SSC crack growth phenomenon was found to be a combined failure phenomenon involving both APC and HE, with HE as the predominant force. It is considered that the prolonged fracture time at +100 mV was due to the effect of a decrease in HE type crack growth resulting from a decrease in the quantity of the hydrogen generation reaction, rather than APC type crack growth due to the promotion of the iron dissolution reaction at the crack tip. On the other hand, because HE is the dominant embrittlement mode of crack propagation, the prolonged fracture time at −100 mV cannot be explained only by the reduced contribution of APC under the cathode overvoltage condition. As a reason for this prolonged operation, it is considered possible that the driving force of HE did not increase sufficiently under the condition of an excessive cathode overvoltage of −100 mV because the amount of hydrogen entering the steel material did not increase greatly. H2S is a cathode reactive species26) in a sour corrosive environment, and is also known to be a catalyst27) that promotes the penetration of hydrogen atoms formed on the steel surface into the steel. That is, under a large cathode overvoltage condition of −100 mV, the amount of hydrogen generated increases, but the amount of H2S consumed by the cathode reaction also increases. As a result, it is somewhat likely that the concentration of H2S on the steel surface decreased and the amount of hydrogen entering the steel itself decreased, so that HE crack propagation was not promoted.
Load-time curve in constant potential SSRT test.
Effect of potential on fracture time in constant potential SSRT test.
Figure 17 shows the relationship between the fracture toughness value J in the strain increment type fracture toughness test and the crack length. The fracture toughness J at the onset of crack growth was similar under the 0.15 bar H2S and 1 bar H2S conditions, but was much lower under the 16 bar H2S condition. As the crack grew, the order of the J value was 16 bar condition < 1 bar condition < 0.15 bar condition, i.e., the fracture toughness value decreased at higher H2S partial pressures. Thus, this test proved that SSC crack propagation in a sour environment is heightened by a high H2S partial pressure. While the strain increment type fracture toughness test considers only the crack initiation/propagation process, the 4PB SSC test also includes the local corrosion process. In other words, the strain increment type fracture toughness test assumes the existence of an initial crack and does not consider the formation process of that initial crack (localized corrosion). Moreover, because the stress intensity factor at the crack tip is controlled in an environment where it is constantly increasing, this is a more severe test of SSC propagation resistance than the actual condition.
Effect of H2S partial pressure on resistance to SSC crack propagation.
Based on the results up to the previous section, it is inferred that the SSC phenomenon consists of both local corrosion initiation/growth (crack initiation and growth) and crack initiation/propagation processes.19,28,29) Investigation of the SSC mechanism based on the 0.15 bar H2S environment revealed that the generation process was dominated by anodic dissolution (local dissolution process: APC) and that local dissolution was promoted by plastic stress concentration and material hardening.30,31) Due to the high corrosiveness of the hard microstructure on the steel surface, the hard microstructure is preferentially dissolved, and the area where the hard microstructure is dissolved becomes a stress concentration area in the presence of stress. Therefore, introduction of dislocations by stress acts as a driving force, and the part is fixed as a preferential dissolution site. In the localized corroded part thus formed, concentration of anions (Cl−, HS−, etc.) in the solution environment progresses.30) That is, when corrosion starts, anions migrate toward the anode (electrophoresis), and as a result, Cl − ions and Fe2+ ions form iron chloride at the anode (inside the pitting portion), and iron hydroxide Fe (OH)3 and hydrochloric acid HCl are formed by a hydrolysis reaction, lowering the pH of the environment in the local corrosion zone.32) This localized lowering of pH causes further dissolution of iron and provides an additional driving force for localized corrosion growth. The results in Section 3.3 confirmed that the propagation process of SSC was based on the manifestation of hydrogen embrittlement (HE). As described above, the series of estimated SSC mechanisms is shown in Fig. 18, in which local corrosion occurs with stress as a driving force, further stress concentration and deterioration in the corrosive environment occur as local corrosion progresses, leading to hydrogen embrittlement. The following discusses the effect of the H2S partial pressure on the SSC mechanism in each process of local corrosion initiation/growth and crack initiation/propagation.
Schematic diagram of estimated SSC mechanism.
As shown in Fig. 18 (Phase 1 to 3), concentration of hydrogen sulfide ions in localized corrosion and lowering of pH due to the progress of localized corrosion increase hydrogen penetration into the steel and cause hydrogen accumulation at the corrosion tip, where stress is concentrated, resulting in the transition to a hydrogen embrittlement crack. When a hard phase exists at the corrosion tip, a hydrogen embrittlement crack is easily generated owing to its high hydrogen trapping property (internal dislocation). As the crack grows, solution exchange between the crack tip and the bulk environment becomes difficult, and the pH in the crack continues to decrease. The stress concentration also accelerates active dissolution and hydrogen embrittlement at the crack tip, as shown in Phase 4 in Fig. 18, and the crack propagates at an increasing crack growth rate with crack growth.
Figure 19 shows a schematic illustration of the effect of the H2S partial pressure on the SSC mechanism during the local corrosion initiation/growth and crack initiation/propagation processes. In local corrosion occurrence/growth (formation of the crack initiation point), formation of protective FeS is difficult at the 0.15 bar H2S partial pressure. Under this condition, protection by FeS cannot be expected in the corroded part, and the part which was preferentially corroded as a result of stress concentration is fixed as a corroded part. Therefore, local corrosion resistance is high. On the other hand, the 1 bar H2S condition is an environment in which highly protective FeS is formed, so that the corrosion resistance of the steel increases together with corrosion, and formation of local corrosion is difficult. (However, once a fully grown localized corrosion part is formed, it is likely to be fixed as a preferential localized corrosion site.) In addition, since the corrosion product is easily formed on the steel surface, it is considered that the corrosion product is sufficiently present even inside the local corroded part, but due to the high closure of the inside of the local corroded zone, the acidic environment of the locally corroded area is maintained and solubility is further increased. Furthermore, based on this closure, the solution resistance inside the localized corroded part may increase, causing potential degradation due to IR drop.33) This degradation may cause hydrogen to evolve at the tip of the localized corroded part, and thus may facilitate the transition to HE at the localized corroded part. Under the high pressure 16 bar H2S partial pressure condition, corrosion products containing FeS are deposited on the surface, but because these corrosion products have a brittle nature, the protective property is not obtained as compared with the 1 bar condition. Moreover, owing to the high H2S concentration, the corrosion environment itself is severe. As a result, the corrosion rate is sufficiently high even at sites other than preferential corrosion sites, and overall corrosion resistance is high.
Effect of H2S partial pressure on local corrosion process and SSC crack initiation/propagation process.
As described in Section 3.4, crack initiation and propagation are accelerated as the H2S partial pressure increases if localized corrosion is not considered. This can be interpreted as a result of promotion of HE due to increased hydrogen penetration into the steel by H2S. If local corrosion is considered, the semicircular shape of the local corrosion under the 16 bar H2S condition makes the occurrence of stress concentration difficult, and is a difficult condition for the transition to a hydrogen embrittlement crack. Therefore, it is considered that SSC sensitivity (at the same SSC hardness threshold value) was equivalent to that under the 1 bar H2S condition, especially in the 4PB SSC test of the GB based microstructure. On the other hand, in the region with a hardness exceeding 250 HV0.1 in the LB + GB mixed microstructure, HE sensitivity was high, and as shown in Fig. 19, and it can be inferred that the 16 bar H2S environment facilitated the transition to a HE crack. As described above, in a 1 bar environment, based on the IR drop, it is estimated that hydrogen generation is high inside the localized corrosion zone. In other words, the 1 bar condition shows high SSC sensitivity in comparison to 0.15 bar not only because of the increase in the amount of hydrogen introduced into the steel due to the high concentration of H2S in the environment, but also because of the increase in the amount of hydrogen generated in the localized corrosion zone itself, which may contribute to ease of the transition to a HE crack.
It should be noted that this experimental examination of the SSC mechanism was based on a soft GB main microstructure with a Vickers micro hardness of less than 250 HV0.1 as shown in Fig. 3, because the test pieces were taken away from the surface layer. From the 4PB SSC test and other test results, it is expected that the effect on the initial APC and the effect on HE which contributes to SSC crack initiation and propagation will be more remarkable with a hard GB + LB mixed microstructure or a LB main microstructure. The effects of these detailed microstructures will be examined further in the future.
The effects of the surface hardness distribution and H2S partial pressure on the sulfide stress corrosion cracking (SSC) behavior of low alloy steel linepipes in sour environments were investigated in the 4PB SSC test.
• In the 4PB SSC test, the limit value of the hardness of the surface for SSC decreased with increasing H2S partial pressures from 0.07 bar to 1 bar, and remained constant at 250 HV0.1 above 1 bar.
• By reducing the cooling rate of the surface layer, thereby suppressing the LB microstructure and forming a GB main microstructure, it was possible to stably ensure a low hardness of 250 HV0.1 or less, and SSC resistance was improved even at H2S partial pressures of 1 bar or more.
• When a GB + LB mixed microstructure with hardness exceeding 250 HV0.1 was formed, SSC susceptibility increased at higher H2S partial pressures.
• Regarding the effect of the H2S partial pressure on the corrosion process, both corrosion loss and the local corrosion depth were largest at 0.15 bar, and corrosion was suppressed under the 1 bar condition. Tight FeS was formed as a corrosion product on the material surface under the 1 bar condition, and suppression of the iron dissolution reaction by this corrosion product was confirmed electrochemically. It was suggested that H2S affected the formation of the protective corrosion product FeS, and this became an influencing factor for sour corrosion.
• From 4PB SSC test results, it was found that SSC in sour environments was enhanced by a high H2S partial pressure over 1 bar H2S. As the reason for this, because hydrogen embrittlement is one of the driving forces for crack propagation at a 0.15 bar H2S partial pressure, it is estimated that hydrogen embrittlement was promoted by H2S.
• From constant potential SSRT test results, SSC propagation was controlled by the balance of the cathodic reaction (hydrogen evolution reaction) and the anodic reaction (iron dissolution reaction). Both HE and APC are the driving force of propagation, but the contribution of HE was larger than APC. It is considered that crack propagates as HE mechanism due to hydrogen accumulation at the corrosion pit or groove after localized corrosion.
• In the strain increment type fracture toughness test, when the H2S partial pressure was changed from 0.15 bar (+0.85 bar-CO2) to 1 bar or 16 bar, both crack initiation and propagation tended to be promoted.
