DEVELOPMENT OF A COMPATIBILITY ASSESSMENT MODEL FOR EXISTING PIPELINES FOR HANDLING HYDROGEN-CONTAINING NATURAL GAS

Abstract

Hydrogen blending in existing natural gas pipelines offers a practical pathway for using established infrastructure in emerging low-carbon energy systems. However, hydrogen exposure can degrade pipeline steels by reducing ductility and fracture toughness and accelerating fatigue crack growth, thereby raising important questions about the compatibility and remaining life of existing pipeline assets. Conventional integrity assessment methods often rely on conservative material limits or simplified assumptions and may not fully capture the combined influence of hydrogen concentration, steel grade, defect geometry, and cyclic loading on crack-driven failure. This study develops an integrated compatibility assessment model for existing pipelines transporting hydrogen-containing natural gas by coupling machine-learning-based material degradation predictions with fracture-mechanics-based integrity and lifetime assessment. The proposed framework uses machine-learning surrogate models to predict hydrogen-sensitive degradation metrics, including reduction of area (RA), fracture toughness (FT), elongation, and fatigue crack growth behavior for API 5L pipeline steels. These predictions are integrated into stress-intensity-factor-based fracture mechanics models to simulate crack growth evolution under cyclic internal-pressure loading. The complete workflow is implemented as the University of Oklahoma Hydrogen Embrittlement Assessment Tool (OU HEAT), a graphical user interface that combines property prediction, stress intensity factor calculation, fatigue crack growth simulation, sensitivity analysis, model comparison, and integrity-relevant output visualization within a single assessment platform. Case study evaluations were conducted on API 5L X52, X60, and X70 pipeline steels under varying hydrogen concentrations and operating conditions. The results show that predicted RA and FT generally decrease with increasing hydrogen concentration, with degradation trends approaching a plateau at higher blend levels. Remaining fatigue life also decreases significantly with increasing hydrogen concentration, underscoring the importance of explicitly accounting for hydrogen-assisted fatigue crack growth in pipeline compatibility assessments. Sensitivity analysis further demonstrates that hydrogen concentration, material properties, defect geometry, and operating pressure can strongly influence predicted crack growth and lifetime. Model comparison with standard fatigue-curve approaches shows that the integrated framework provides physically consistent lifetime predictions across different cyclic pressure scenarios. Overall, this study presents a scalable and practical assessment methodology for screening existing pipelines intended for hydrogen-natural gas service. By linking hydrogen-induced material degradation to crack driving forces, flaw tolerance, and remaining life, the developed framework supports more informed pipeline integrity management, inspection planning, and risk-based decision-making for hydrogen-exposed pipeline systems.