Changes to Integrity Management Reinspection Intervals for Pipelines in Hydrogen Service

This Task 4 report was developed by Kiefner and Associates, Inc. (Kiefner) for the U.S. Department of Transportation (USDOT) Pipeline and Hazardous Materials Safety Administration (PHMSA) as part of the research project “Changes to Integrity Management Reinspection Intervals for Pipelines in Hydrogen Service” under Contract 693JK32310011POTA. The objective of the work was to evaluate how hydrogen service, including hydrogen-blended natural gas service, may affect remaining life (RL) predictions and, consequently, integrity management (IM) reinspection intervals (RI) relative to conventional natural gas service.

The assessment focused on crack-like features in the pipe body and welds, including longitudinal, circumferential, and off-axis flaws, as well as mechanical damage such as dents. Since hydrogen accelerates fatigue crack growth and reduces fracture resistance, flaw populations that might be slow-growing or benign in natural gas service can become more concerning in hydrogen service. Consequently, this study assessed RL across a range of flaw sizes, fracture toughness values, pipe geometries, and pressure cycle severities, including flaw sizes near or below the practical detection limits of in-line inspection (ILI) and non-destructive examination (NDE) tools.

Fracture mechanics-based fatigue analysis in hydrogen was performed using advanced fatigue models that capture crack growth behavior across the full fatigue spectrum, including the threshold, stable growth, and rapid growth regions. Crack growth was predicted using the Harter T-method, implemented in AFGROW software, together with fatigue crack growth rate (FCGR) data generated by Sandia National Laboratories (SNL) in hydrogen environments. By using experimentally measured FCGR data, rather than relying solely on upper-bound design curves, the analysis reduced conservatism and provided a more realistic estimate of RL in hydrogen service. Results were also benchmarked against the fatigue design curves formalized in ASME B31.12 Code Case 220. Case studies were performed assuming 100% hydrogen, as experimental results indicated that even small hydrogen concentrations can have detrimental effects on material properties, including FCGR and fracture toughness.

Hydrogen-induced cracking can occur with or without applied stress. However, crack growth rate data for hydrogen-induced cracking (HIC) are not well documented, and thus, time-dependent life calculations cannot be performed. This report focuses on hydrogen-accelerated fatigue as hydrogen FCGR data is available and can be used directly in fracture mechanics analyses to quantify RL and RIs. For the purposes of this study, a fatigue-based framework provided a practical and technically defensible basis for comparing hydrogen service with conventional natural gas (NG) service.

Across the cases studied, fatigue life in hydrogen gas was approximately one order of magnitude lower than in air, and this trend held across different pipe geometries and pressure cycle severities. For longitudinal crack-like flaws, long fatigue lives in hydrogen service were predicted under mild pressure cycling for small flaw sizes representative of flaws that may remain undetected by ILI or NDE. For such cases, conservative upper-bound design curves remain adequate for screening; based on the current results, use of ASME B31.12 upper-bound curves appears sufficient. For larger flaws (with depth greater than 40% of the pipe wall thickness (WT)), the use of actual FCGR test data predicts longer life; in some cases, approximately double that predicted by the upper-bound design curves. Under aggressive pressure cycling, the results in hydrogen service indicated that the integrity assessment of larger flaws benefits from using more realistic FCGR inputs than using the ASME B31.12 upper-bound curves.

The sensitivity studies also showed that fatigue life in hydrogen service is strongly affected by flaw depth, pressure cycle severity, and WT. Pipes with thinner WT consistently exhibited shorter fatigue life because of the smaller remaining ligament for a given normalized flaw depth. More severe pressure cycles produced shorter remaining life. Larger-diameter, thicker-wall pipes having the same outside diameter (OD)/WT ratio showed approximately twice the life, or more, for larger flaw sizes. Most longitudinal flaw failures in hydrogen service occurred only after the crack became effectively through-wall, representing a leak-before-break scenario. This is favorable from an IM standpoint because it allows a greater opportunity for detection and intervention before rupture.

Fracture toughness effects in hydrogen service were also evaluated. Reducing plane strain fracture toughness from the experimentally hydrogen tested base-case average value of 90 ksi√in to the ASME B31.12 minimum of 50 ksi√in produced only a minor reduction in fatigue life for many longitudinal flaw cases. However, further reductions in toughness, below 50 ksi√in, led to much shorter remaining life, demonstrating that very low-toughness conditions in hydrogen service remain a significant concern, particularly in vintage materials and weld regions.

Girth welds in hydrogen service introduce additional sensitivity to flaw size, weld residual stress, and geometric stress concentrations, such as hi/lo. The combined effects of higher-strength pipe grade, residual stress, and weld misalignment were shown to reduce RL. These findings indicated that girth welds should receive greater attention in hydrogen service, particularly in high-grade pipe segments, non-post weld heat-treated (non-PWHT) welds, locations with known fit-up challenges, and areas subject to higher cyclic pressure fluctuations.

The study also confirmed that dents require more conservative assessment in hydrogen service. Dents, gouges, and wrinkles can promote crack initiation during formation, fatigue cracking under cyclic loading, and reduced failure pressure in hydrogen-embrittlement materials. Hydrogen further increases failure susceptibility by reducing the plastic deformation required for crack initiation and increasing the severity of crack driving forces at dented locations. As a result, dents that may have been considered non-injurious in natural gas service may require re-evaluation and, in some cases, more advanced assessment methods and higher safety factors (SFs) for pipelines in hydrogen gas service.

From the work in this study, the following future studies are recommended to reduce uncertainty and better support implementation within IM programs for pipelines in hydrogen service:

• Probabilistic fatigue assessment: To explicitly account for uncertainty in flaw size, probability of detection (POD), probability of identification (POI), ILI/NDE sizing error, fracture toughness, FCGR scatter, residual stress, and pressure history.
• Hard spots in hydrogen and hydrogen-blended service: To quantify the relationship among local hardness, microstructure, through-wall extent, fracture resistance, and hydrogen-assisted crack initiation/growth.
• Hydrogen concentration: To quantify the effect of low and intermediate hydrogen (1% to 20%) concentrations in blended-gas service on FCGR, including the roles of hydrogen partial pressure, mixed-gas composition, and possible impurities, and to develop engineering crack-growth models for hydrogen blends rather than assuming only 100% hydrogen conditions.[i]^,[ii]

Low-cycle fatigue: To include low-cycle fatigue associated with relatively infrequent but high-severity transient events, including start-ups, shutdowns, station recycling, blowdowns, and other non-steady operating events.

[i] Brongers, M. (2024, July 26). Literature review on reinspection intervals for pipelines carrying hydrogen or hydrogen/natural gas blends (Task 1 Final Report No. 24-30994-5000932R). Kiefner and Associates, Inc.; U.S. Department of Transportation Pipeline and Hazardous Materials Safety Administration.

[ii] Brongers, M. P. (2025, June 23). Evaluation of analyses for different hydrogen/natural gas blends (Task 5 Final Report No. 25-30994-6000754R). Kiefner and Associates, Inc.; U.S. Department of Transportation Pipeline and Hazardous Materials Safety Administration.