IMPACTS OF HYDROGEN EMBRITTLEMENT ON DUCTILE PROPERTIES OF NATURAL GAS PIPELINE STEELS

Abstract

The mechanical integrity of pipeline steels is significantly compromised by hydrogen embrittlement (HE), particularly in hydrogen-rich environments. While numerous studies have explored the effects of hydrogen on steel using tensile tests, fracture toughness evaluations, and fatigue loading, research involving hydrogen–natural gas blends remain limited. Most existing investigations have assessed hydrogen in isolation, without accounting for the potential influence of natural gas and its constituent components. Furthermore, there is currently no empirical or physics-based model capable of reliably predicting the ductility of carbon steel in hydrogen-containing environments. This study addresses critical research gaps by investigating the effects of HE on the tensile properties of X52, X60, and X70 pipeline steels. Specifically, the reduction of area (RA) and elongation were evaluated under exposure to pure hydrogen at pressures ranging from 0 to 6.9 MPa and to hydrogen–natural gas blends with hydrogen concentrations from 0% to 100% by volume. The influence of impurities such as oxygen and the effect of temperature were also examined to assess their roles in HE susceptibility. Most experiments, excluding those with variable temperature, were conducted at ambient conditions to isolate the effects of gas composition and pressure. A machine learning (ML) model was developed in parallel with the experimental work. A comprehensive database was constructed by combining the experimental results from this study with published data, which was then used to train and test the ML model for predicting HE-induced ductility degradation. The results from pure hydrogen exposure reveal a systematic decline in ductility with increasing pressure, as evidenced by reductions in RA and elongation. This trend underscores the critical need for HE mitigation strategies in hydrogen-exposed pipeline materials. In contrast, the mixed gas experiments exhibited distinct RA and elongation variations with increasing hydrogen concentration. The highest RA value in pure hydrogen was observed at zero hydrogen pressure, whereas in the blended gas environment, the maximum RA was recorded at approximately 7.5% hydrogen concentration. Comparative analysis under equivalent hydrogen partial pressure conditions indicates that mixed gas embrittlement is less severe than pure hydrogen at low partial pressures (~0.52 MPa). A similar trend was observed in elongation measurements. However, at hydrogen concentrations exceeding 20%, embrittlement effects in mixed gas surpassed those observed in pure hydrogen under equivalent partial pressure conditions. Oxygen exhibited an inhibitive effect on HE, with as little as 100 ppm effectively mitigating HE and preserving the RA value to a level comparable to the 0% hydrogen condition. Temperature demonstrated a complex influence, affecting metal ductility both through thermal softening and HE. RA decreased for all three steels at the lower temperature of 10°C, primarily due to the thermal softening effect. A similar reduction at 50°C suggests a shift toward HE-dominated degradation at elevated temperatures. Among the six machine learning models evaluated, the gradient boosting algorithm demonstrated superior performance, delivering highly accurate predictions of RA under diverse conditions. Key features identified as having the greatest influence included hydrogen pressure, yield strength, oxygen content, and concentrations of phosphorus and iron, aligning well with existing literature and this study's outcomes. The model reliably predicted RA values across a wide range of hydrogen pressures, yield strengths, and oxygen levels, highlighting its robustness and generalization capability. These findings offer critical insights into the interaction between hydrogen exposure and tensile properties, supporting the development of hydrogen embrittlement management strategies through gas composition optimization, the introduction of inhibitive gaseous impurities, and a deeper understanding of the roles of temperature, operating pressure, and elemental composition.