f Inhibition of Hydrogen Embrittlement Effects in Pipeline Steel - Technical Report

National Gas has identified the need to understand the effects of hydrogen on various pipeline materials commonly found in the UK National Transmission System (NTS) to determine their potential for a hydrogen use case. Accordingly, National Gas is interested in investigating the potential use of oxygen as a mitigation of hydrogen embrittlement (HE), the detrimental effect of hydrogen on the mechanical properties of metallic materials. This report presents the laboratory findings part of the project “Inhibition of hydrogen embrittlement effects in pipeline steels”, in which ROSEN has been requested to investigate the potential use of oxygen as a mitigation of hydrogen effects.
Three pipe materials were investigated, which included a ERW X52 commissioned in 1993, a DSAW X60 commissioned in 1973 and a DSAW X80 constructed in 2004. The test programme aimed to characterise the effectiveness of oxygen on mitigating hydrogen embrittlement and included an in-air baseline characterisation and three main hydrogen tests: threshold stress intensity factor KIH, fracture toughness and two types of fatigue crack growth rate tests (frequency scans and Paris curve). Fracture toughness and fatigue crack growth rate tests were performed in pure hydrogen and at various oxygen concentrations ranging from 50 to 1000 ppm.
The baseline in-air characterisation showed that some of the mechanical properties of the X52 and X60 were below the T/SP/PIP/1 requirements. For the X52, the yield strength of transverse parent material specimens was below the 360 MPa required by its designation. For the X60, both the longitudinal and transverse yield strength of the parent material were below the 415 MPa required for this grade. Furthermore, CVN energies were also below the T/SP/PIP/1 requirements for parent metal and seam weld metal for the X60 pipe. The X80 was within specifications.
Threshold stress intensity factor KIH tests were performed on the three investigated materials on base and weld metal in pure hydrogen. Crack extension was not seen in any of these tests for the applied stress intensity factors as high as 73 MPa√m. The absence of crack growth was verified by means of SEM examination which showed the direct transition from the fatigue pre-crack region to the final fracture region. For this reason, KIH tests with oxygen additions were not conducted.
Fracture toughness testing showed that the fracture resistance was reduced when testing in pure hydrogen. This was seen in X60 specimens that had a fracture resistance JQ of 31 kJ/m2, 35% of that in air. The addition of oxygen resulted in an increase of the average JQ, which at 250 ppm O2 was 85% of the in air value. At concentrations above this value, the fracture resistance continued to approach in-air values, although at a decreased rate. Increasing the oxygen concentration also resulted in smaller standard deviations of the fracture resistance, decreasing from 13 to 2 kJ/m2 as the O2 concentration was increased from 50 to 500 ppm. These findings indicate there is a smaller variability of the performance when oxygen is added. The recovery of the fracture resistance was consistent with the SEM examination of fracture surfaces of tested specimens. In pure hydrogen, fracture surfaces were consistent with a quasi-cleavage fracture characterised by planar facets with visible river marks. As oxygen was increased, the fracture surface became more dimpled and areas with signs of plastic strain became evident. At concentrations of 250 ppm O2, the fracture surface resembled those obtained in air, suggesting the same ductile failure mode. Obtaining quantitative fracture toughness data for the X52 and X80 specimens in oxygen additions was not possible due to material related challenges. However, fractographic examination showed that in pure hydrogen and low oxygen concentrations e.g. 50 ppm, the failure mode was predominantly brittle while at 250 ppm O2 and in air, the failure mode was ductile and was accompanied of limited crack growth.
The inhibiting effects of oxygen were also seen in fatigue crack growth rate tests. Frequency scan tests on X60 showed that the exposure to hydrogen resulted in high crack growth rates, up to 2 orders of magnitude above the BS7910 in-air mean, depending on the testing conditions. Crack growth rates were affected by the choice of the frequency. For the test performed with a Kmax of 45 ksi√in (49 MPa√m), crack growth rates continued to increase above 1 mm/cycle as the frequency was reduced to 1E-4 Hz. For the test performed with a Kmax of 38 ksi√in (42 MPa√m), crack growth rates reached a plateau at a frequency of 0.1 Hz. The additions of oxygen resulted in a reduction of crack growth to values a few times above the BS7910 in-air mean. Furthermore, the effect of frequency was visibly reduced in 250 ppm O2, as reducing the frequency 40,000 times resulted only in a 2 times increase of the crack growth rates. The effect of oxygen was also seen in Paris curve types of tests. In pure hydrogen, crack growth rates were comparable with ASME B31.12 and Sandia National Laboratories reference curves and were in general up to an order of magnitude above the BS7910 in-air mean, depending on ΔK and other parameters. The additions of oxygen at concentrations as low as 50 ppm resulted in reductions in fatigue crack growth rates to values closer to those in air. Further increases in the oxygen concentration resulted in slight reductions of crack growth rates that were more noticeable at higher ΔK and Kmax. The recovery of the fatigue performance was consistent with the fractography observations. The addition of oxygen resulted in an overall increase of the plastic strain seen on the fracture surfaces. In pure hydrogen, fracture surfaces were consistent with quasi-cleavage failure and had planar cleavage facets with river marks and very fine striations that were mostly visible at high frequencies and ΔK under high magnifications (55,800x). The fracture surfaces of specimens tested in presence of oxygen resembled those seen in the in-air fatigue pre-crack region, especially at high frequencies and high ΔK. At low frequencies and low ΔK, the fracture surfaces resembled that of quasi-cleavage fracture although they had a ‘fibrous’ aspect with visible signs of plastic strain. Striations were readily visible on the fracture surfaces of all specimens tested in presence of oxygen, and were more defined than those seen in pure hydrogen.
Based on the data generated in this work, a concentration of 250 ppm O2 is recommended as the minimum value to achieve inhibition of hydrogen effects. This concentration provided a recovery of 85% of the in-air fracture resistance and resulted in crack growth rates close to in-air levels. The fracture surfaces of specimens tested at this oxygen concentration generally showed ductile features consistent with higher toughness failure mechanisms.