Hydrogen Environment-assisted Cracking Research Articles

Hydrogen assisted cracking driven by cathodic protection operated at near −1200 mV CSE – an onshore natural gas pipeline failure

A 22-inch onshore natural gas transmission pipeline in operation since 1960 experienced an in-service rupture in January of 2018. The rupture was caused by an axial crack located 150 km (93 miles) downstream of the closest compression station. While the root cause analysis of the rupture was being carried out, excavations of indications reported by a Hard Spot In-Line Inspection approximately 42 km (26 miles) downstream of the rupture found leaks from predominantly circumferential crack colonies mixed with spiderweb ones not associated with hard spots.The conclusion from this investigation is that the cracking mechanism for the in-service rupture and leaks is mainly hydrogen environment assisted cracking (HAC), with the hydrogen being generated by the cathodic protection. The root cause for the failures is the combination of an aged and defective coating and the cathodic protection (CP) system operated for decades at near the pipe-to-soil potential limit of −1200 mV CSE.An extensive literature search indicates that this investigation is the first fully documented case of an onshore transmission pipeline failure due to HAC driven by CP, with the cracks occurring in the pipe body of X52 steel, away from welds and not related with hard spots. The paper summarizes the approach used for the root cause analysis and the findings, where the implications in the management of the pipeline integrity of the pipeline system are discussed. The pipeline operator implemented a series of preventive and corrective measures, which included pressure testing, cathodic protection improvement, coating rehabilitation programs and In-Line Inspection programs.

On the hydrogen environment-assisted cracking resistance of a compositionally complex Co–Ni–Cr–Fe–Mo–Ti alloy

The hydrogen environment-assisted cracking (HEAC) behavior of a compositionally complex Co–Ni–Cr–Mo–Fe–Ti–Al alloy (MP98T) proposed for marine fastener applications was evaluated via slow-rising stress intensity testing while immersed in 0.6 M NaCl at applied potentials ranging from −1.0 to −1.2 VSCE (vs. saturated calomel electrode). Results indicate that MP98T is resistant to HEAC at applied potentials representative of cathodic protection. In particular, minimal crack growth was observed across the tested applied potential range, which represented conditions known to promote aggressive HEAC in legacy Ni-based fastener alloys, such as Monel K-500. Measurement of the diffusible hydrogen concentration (CH,Diff) of MP98T as a function of applied potential in 0.6 M NaCl suggests this improved susceptibility may be linked to a reduced efficacy for hydrogen uptake in MP98T at hydrogen overpotentials where Monel K-500 is susceptible. Specifically, the CH,Diff of MP98T at a given applied potential was consistently observed to be less than 50% of that measured for Monel K-500.

Elucidating the loading rate dependence of hydrogen environment-assisted cracking in a Ni-Cu superalloy

The effect of loading rate (dK/dt) on the hydrogen environment-assisted cracking (HEAC) behavior of Monel K-500 immersed in 0.6 M NaCl and polarized to −950 and −1200 mVSCE was assessed via a slow-rising stress intensity (K) testing framework at dK/dt ranging from 0.2 to 20 MPa√m/hr. Comparative testing on Monel K-500 in an inert environment (dry N2 with RH < 5%) at dK/dt ranging from 0.1 to 4 MPa√m/hr revealed a clear dK/dt-dependence of local plasticity effects on the direct current potential difference (dcPD) signal. A method for decoupling this “false” signal caused by local plasticity from the ‘true’ crack growth kinetics is presented. Analysis of the corrected data showed that the threshold stress intensity for the onset of HEAC (KTH) was nominally invariant with dK/dt. The HEAC growth rates (da/dt) were only mildly dependent on dK/dt under severe environmental conditions (−1200 mVSCE). However, for moderate conditions (−950 mVSCE), a strong influence of dK/dt on HEAC growth rates (da/dt) was observed. For this latter testing environment, two characteristic regimes of behavior were noted depending on the applied dK/dt: (1) a ‘plateau’ regime where da/dt is independent of dK/dt was observed for dK/dt > 8 MPa√m/hr and (2) a ‘linear’ regime where log(da/dt) linearly scales with log(dK/dt) was observed for dK/dt < 8 MPa√m/hr. These trends are interpreted using resistance curves calculated from the measured da/dt vs. KJ data. The results of this analysis demonstrate that diffusion-mediated crack growth is only observed for dK/dt > 10 MPa√m/hr at −950 mVSCE and all evaluated dK/dt at −1200 mVSCE, while mechanics-based factors govern the da/dt response for dK/dt < 10 MPa√m/hr at −950 mVSCE. The influence of dK/dt is considered in the context of current testing standards for evaluating HEAC susceptibility; specifically, rising K-based frameworks consistently yield conservative HEAC metrics.

Multiscale Assessment of Deformation Induced by Hydrogen Environment-Assisted Cracking in a Peak-Aged Ni-Cu Superalloy

A multiscale approach leveraging electron backscatter diffraction (S-EBSD), high resolution EBSD (HR-EBSD), and transmission electron microscopy (TEM) was employed to assess the deformation induced proximate to the crack path by hydrogen environment-assisted cracking (HEAC) in peak-aged Monel K-500. Kernel average misorientation (KAM) results calculated from S-EBSD indicate that the deformation pertinent to HEAC is localized to within 25 μm of the crack path. Geometrically necessary dislocation (GND) density maps calculated from HR-EBSD confirm this localization. The evaluation of the deformation distribution in three separate grains along the crack path using complementary HR-EBSD and TEM suggest a qualitative similarity in dislocation density between the two techniques, though HR-EBSD is unable to spatially resolve the fine dislocation structures observed via TEM. However, non-negligible differences in dislocation patterning were observed in the three evaluated grains, highlighting the importance of a multiscale approach for characterizing deformation to understand the governing microstructural and mechanical factors.

The effect of isothermal heat treatment on hydrogen environment-assisted cracking susceptibility in Monel K-500

The effect of isothermal heat treatment on hydrogen environment-assisted cracking (HEAC) susceptibility in a single heat of Monel K-500 was evaluated via slow-rising stress intensity (K) testing while immersed in 0.6 M NaCl solution and exposed to applied potentials ranging from −1000 to −1200 mVSCE (vs. saturated calomel electrode). Four heat treatments, corresponding to the non-aged, under-aged, peak-aged, and over-aged conditions, conducted at a constant aging temperature of 923 K, were examined in this study. Crack growth kinetics and fractography demonstrate that alloys heat-treated to the under-aged and peak-aged conditions exhibit increased susceptibility to HEAC relative to alloys treated to the non-aged and over-aged conditions, respectively. Microstructural variables that are expected to vary under this isothermal heat treatment protocol are grain size, grain boundary character, grain boundary impurity concentration, yield strength, strain hardening, and precipitate morphology. A detailed assessment of grain size, grain boundary character, and grain boundary impurity concentration evolution with aging time suggests these variables do not explain the observed variation in HEAC susceptibility of Monel K-500. Differences in the crack tip hydrostatic stress field induced by differences in yield strength and strain hardening were directionally consistent with the observed HEAC susceptibility, but do not capture the enhanced susceptibility of the under-aged heat treatment relative to the peak-aged condition. However, transmission electron microscopy of each heat treatment revealed the presence of planar slip in the under-aged and peak-aged specimens, as compared to wavy slip in the non-aged and over-aged specimens, suggesting that the propensity for strain localization correlates well with HEAC susceptibility.

Hydrogen Environment Assisted Cracking of Modern Ultra-High Strength Martensitic Steels

Martensitic steels (Aermet®100, Ferrium®M54™, Ferrium®S53®, and experimental CrNiMoWV at ultra-high yield strength of 1550 to 1725 MPa) similarly resist hydrogen environment assisted cracking (HEAC) in aqueous NaCl. Cracking is transgranular, ascribed to increased steel purity and rare earth addition compared to intergranular HEAC in highly susceptible 300M. Nano-scale precipitates ((Mo,Cr)2C and (W,V)C) reduce H diffusivity and the K-independent Stage II growth rate by 2 to 3 orders of magnitude compared to 300M. However, threshold K TH is similarly low (8 to 15 MPa√m) for each steel at highly cathodic and open circuit potentials. Transgranular HEAC likely occurs along martensite packet and {110}α′-block interfaces, speculatively governed by localized plasticity and H decohesion. Martensitic transformation produces coincident site lattice interfaces; however, a connected random boundary network persists in 3D to negate interface engineering. The modern steels are near-immune to HEAC when mildly cathodically polarized, attributed to minimal crack tip H production and uptake. Neither reduced Co and Ni in M54 and CrNiMoWV nor increased Cr in S53 broadly degrade HEAC resistance compared to baseline AM100. The latter suggests that crack passivity dominates acidification to widen the polarization window for HEAC resistance. Decohesion models predict the applied potential dependencies of K TH and da/dt II with a single-adjustable parameter, affirming the importance of steel purity and trap sensitive H diffusivity.

Hydrogen Environment Assisted Cracking of a Modern Ultra-High Strength Martensitic Stainless Steel

A modern martensitic stainless steel (Ferrium® PH48S™) resists hydrogen environment assisted cracking (HEAC) in aqueous NaCl at ultra-high yield strengths (1,400 MPa to 1,600 MPa). HEAC is transgranular, because of increased steel purity and La addition, compared to severe intergranular HEAC in Custom® 465-H900 without rare earth elements. Minimum threshold for HEAC (KTH) is low (8 MPa√m to 17 MPa√m) for each steel under substantial cathodic polarization. Transgranular HEAC occurs along martensite packet and {110}α′-block interfaces in PH48S, likely a result of H decohesion enabled by localized plasticity. Martensite transformation produces a large area of coincident site lattice interfaces in the refined microstructure of PH48S. However, a susceptible network of random packet/block interfaces is connected in 3D to limit interface engineering. Nanoscale strengthening precipitates in PH48S reduce effective H diffusivity to the mid-10−10 cm2/s range, because of reversible H trapping with a binding energy of 12 kJ/mol. This diffusivity reduces the Stage II growth rate by 1 to 3 orders of magnitude compared to C465 and carbide strengthened ultra-high strength steels. PH48S and C465 are nearly immune to HEAC when cathodically polarized by 50 mV to 500 mV, attributed to a minimum in occluded-crack tip overpotential for H production. The breadth of this protective-potential window increases with decreasing steel strength. Increased Cr does not degrade HEAC resistance, suggesting that crack passivity dominates cation acidification to reduce H production and/or uptake. A quantitative decohesion model effectively predicts the potential dependence of da/dtII using crack tip H solubility reverse calculated from a KTH model.

The Effect of Microstructural Variation on the Hydrogen Environment-Assisted Cracking of Monel K-500

The influence of microstructural variation on hydrogen environment-assisted cracking (HEAC) of Monel K-500 was evaluated using five nominally peak-aged lots of material tested under slow-rising stress intensity loading while immersed in NaCl solution under cathodic polarizations. Minimal variation in HEAC resistance among material lots was observed for an applied potential of −950 mVSCE (E app, vs saturated calomel), whereas lot-to-lot variability in the fracture morphology demonstrates a significant difference in the HEAC resistance at the less negative potential of −850 mVSCE, suggesting that relatively severe H environments produce sufficient crack-tip H to minimize the impact of metallurgical differences. Sensitivity analyses accomplished by varying the inputs used in decohesion-based, micromechanical models imply significant variations in HEAC resistance are possible for realistic changes in grain boundary toughness, hydrogen uptake behavior, and yield strength. Grain size, impurity segregation (including the effects of gettering elements), grain boundary character/connectivity, and crack path tortuosity are also considered in the context of HEAC susceptibility. Yield strength, global hydrogen content, as well as impurity segregation to grain boundaries, especially boron and sulfur, are speculatively considered to be the dominant contributions in determining HEAC resistance. Modifications that would incorporate the effects of grain boundary segregation are proposed for the K TH model; detailed validation of such changes require high-fidelity and quantitative inputs for the degree of grain boundary segregation. Regardless, fracture mechanics-based HEAC results, detailed microstructural characterization, and micromechanical modeling were successfully coupled to gain insights into the influences governing the microstructure-dependent HEAC susceptibility of Monel K-500.

Measurement and Modeling of Hydrogen Environment-Assisted Cracking in a Ni-Cu-Al-Ti Superalloy

This research improves H decohesion mechanism-based modeling of intergranular stress corrosion cracking in a Ni-Cu superalloy, Monel K-500. New cracking data plus improved model parameters lead to accurate predictions of the cathodic potential dependencies of KTH and H diffusion-limited da/dtII for Monel K-500 under slow-rising K in 0.6 M NaCl solution. Experiments and modeling demonstrate that IGSCC is eliminated for applied potentials more positive than a critical level between −900 and −840 mVSCE, but slow-subcritical cracking persists by a microvoid-based mechanism.

Erratum to: Measurement and Modeling of Hydrogen Environment-Assisted Cracking in Monel K-500

Fig. 17—(Color online) Model predicted (Eq. [2]) and measured Stage II crack growth rate (at elastic K = 50 MPa m, Table III; Fig. 9) vs crack tip CH-Diff (bottom axis) and stress enhanced CHr (top axis) for aged Monel K-500 (ATI Allvac, j; Table I) stressed in 3.5 pct NaCl at EA between 1100 and 700 mVSCE. The rH/rYS is 12.0 in Eq. [3]. Measured da/dtII are included for a Special Metals lot of Monel K-500 (m) using the corresponding H uptake law (Footnotes * and **). Fig. 18—(Color online) Model predicted (Eq. [2]) and measured Stage II crack growth rate (at elastic K = 50 MPa m, Table III; Fig. 9) vs applied cathodic potential for aged Monel K-500 (ATI Allvac, j; Table I) stressed in 3.5 pct NaCl. The rH/rYS equals 12.0 in Eq. [3]. Measured da/dtII values are included for a Special Metals lot of Monel K-500 (Footnote *, m).

Measurement and Modeling of Hydrogen Environment-Assisted Cracking in Monel K-500

Hydrogen environment-assisted cracking (HEAC) of Monel K-500 is quantified using slow-rising stress intensity loading with electrical potential monitoring of small crack propagation and elastoplastic J-integral analysis. For this loading, with concurrent crack tip plastic strain and H accumulation, aged Monel K-500 is susceptible to intergranular HEAC in NaCl solution when cathodically polarized at −800 mVSCE (E A, vs saturated calomel) and lower. Intergranular cracking is eliminated by reduced cathodic polarization more positive than −750 mVSCE. Crack tip diffusible H concentration rises, from near 0 wppm at E A of −765 mVSCE, with increasing cathodic polarization. This behavior is quantified by thermal desorption spectroscopy and barnacle cell measurements of hydrogen solubility vs overpotential for planar electrodes, plus measured-local crevice potential, and pH scaled to the crack tip. Using crack tip H concentration, excellent agreement is demonstrated between measurements and decohesion-based model predictions of the E A dependencies of threshold stress intensity and Stage II growth rate. A critical level of cathodic polarization must be exceeded for HEAC to occur in aged Monel K-500. The damaging-cathodic potential regime likely shifts more negative for quasi-static loading or increasing metallurgical resistance to HEAC.

Effect of Variable Stress Intensity Factor on Hydrogen Environment Assisted Cracking

Fitness-for-service evaluations of engineered components that are subject to environment assisted cracking (EAC) often require analyses of potentially large crack extensions through regions of variable stress intensity. However, there are few EAC data and models that directly address the effects of variable stress intensity factor on EAC crack growth. The model developed here is used to evaluate stress corrosion cracking (SCC) data that were obtained on a high-strength beta-titanium alloy under conditions of variable crack mouth opening displacement (CMOD) rate. SCC of this Ti alloy in ambient temperature, near-neutral NaCl aqueous solution is thought to be due to hydrogen environment assisted cracking (HEAC). As the model equations developed here do not admit to a closed form solution for crack velocity as a function of applied stress intensity factor, K, a semiquantitative graphical solution is used to rationalize the crack growth data. The analyses support a previous suggestion that the observed crack growth rate behavior can be attributed to the effect of crack tip strain rate on rates of mechanical disruption and repair of an otherwise protective crack tip oxide film. Model elements introduced here to HEAC modeling include (1) an expression relating corrosion-active surface area to crack tip strain rate and repassivation rate, (2) an expression relating the critical grain boundary hydrogen to the applied stress intensity factor, and (3) an expression relating CTSR to both applied and crack advance strain rate components. Intergranular crack advance is modeled assuming diffusive segregation of corrosion-generated hydrogen to grain boundary trap sites causing embrittlement of the fracture process zone (FPZ). The model equations developed here provide a quantitative basis for understanding the physical significance of K-variation effects and, with additional development, will provide an engineering tool for analysis of crack growth in a variable K field.

Measurement and Modeling of Hydrogen Environment–Assisted Cracking of Ultra-High-Strength Steel

Modern precipitation-hardened ultra-high-strength AERMET 100 steel (Fe-Co-Ni-Cr-Mo-C) is susceptible to severe transgranular hydrogen environment–assisted cracking (HEAC) in neutral 3.5 pct NaCl solution. The threshold stress intensity for HEAC, K TH , is reduced to as low as 10 pct of K IC , and the stage II subcritical crack growth rate, da/dt II, is up to 0.5 μm/s. Low K TH and high da/dt II are produced at potentials substantially cathodic, as well as mildly anodic, to free corrosion. However, a range exists at slightly cathodic potentials (–0.625 to –0.700 VSCE), where the crack growth rate is greatly reduced, consistent with reduced crack-tip acidification and low cathodic overpotential for limited H uptake. Short crack size (250 to 1000 μm) does not promote unexpectedly severe HEAC. High-purity AERMET 100 is susceptible to HEAC because martensite boundary trapping and high crack-tip stresses strongly enhance H segregation to sites that form a transgranular crack path. The stage II da/dt is H diffusion rate limited for all potentials examined. A semiquantitative model predicts the applied potential dependence of da/dt II using reasonable input parameters, particularly crack-tip H uptake reverse calculated from measured K TH and a realistic critical distance. Modeling challenges remain.

湿潤空気中におけるアルミニウム合金切欠き板材の水素脆化特性

For 6061-T6 and 7075-T6 aluminum alloy plates, SSRT (slow strain-rate technique) tests in humid air with 90% relative humidity were carried out using a notched tensile specimen to evaluate the susceptibility to hydrogen embrittlement and to appreciate the role of notch in the process of HEAC (hydrogen-environment-assisted cracking). 7075-T6 showed a significant embrittlement due to intergranular cracking more inspired by the existence of notch, while in contrast, 6061-T6 exhibited a trend of an enhanced plastic deformation by hydrogen effects, accompanied with a trace of nonconventional TTS facet in fracture surface and an increase in notch opening displacement in the early stage of still low stress level. The SSRT behaviors of these alloys are explained to result from the mechanism of hydrogen-enhanced localized plasticity, where the process of hydrogen population is supported to include both effects of trapping at microstructural defects such as inclusions and stress-induced diffusion at macroscopic defects such as notch.

湿潤空気中における７０７５および６０６１アルミニウム合金の水素脆化特性

For 6061-T6 and 7075-T6 aluminum alloys as candidate liner materials for high-pressure hydrogen container, SSRT (slow strain-rate technique) tests in humid air with 90% relative humidity were carried out to evaluate the susceptibility to hydrogen embrittlement (HE) and to reveal the process of HEAC (hydrogen-environment -assisted cracking). 7075-T6 showed a significant susceptibility to HE, resulting in intergranular cracking, while in contrast, 6061-T6 exhibited an excellent resistance required as liner materials. HEAC of 7075-T6 is confirmed to progress as follows; at the first stage micro-cracks originate at inclusions of Al–Fe–(Cu) phase on the specimen surface and then advance to intergranular cracking. 6061-T6 would not give rise to micro-cracks because of less localization of slip deformation enhanced by hydrogen effects, mainly due to the precipitation microstructure.

836 水素マイクロプリント法による高強度アルミニウム合金の HEAC (水素環境助長割れ)挙動の観察

836 水素マイクロプリント法による高強度アルミニウム合金の HEAC (水素環境助長割れ)挙動の観察