Answers are grounded in this paper and cite their sources.
Simulation reveals hydrogen purity changes with blending ratios in gas reservoirs, indicating optimized storage strategies.
This study presents a quantitative simulation analysis of hydrogen-enriched methane (HENG) storage with nitrogen as the cushion-gas in a depleted gas reservoir by varying three key operational parameters: the injection/withdrawal period, hydrogen blending ratio (5–20%), and injection depth. Ten injection–withdrawal cycles were modeled for each scenario, and performance was evaluated using cycle-averaged and cumulative hydrogen purity, recovery factors, and the mixing zone size. Extending the injection period increased hydrogen purity to 20.00–20.26% and reduced nitrogen to 0.001–0.003%, but recovery decreased from 65.63 to 53.83–41.09% due to enhanced dispersion and residual trapping. The blending ratio was the dominant control: 20% blending yielded 19.9–20.0% purity with nitrogen as low as 0.00–0.03%, whereas 5–10% blending produced lower purity but minimized nitrogen production to 0.97–1.08%. Injection depth affected nitrogen recovery more than purity, increasing from 0.72–1.20% (upper) to 1.46–1.61% (lower), along with thicker mixing zones. Final mixing zone size ranged from 3176 to 5546 blocks, with smaller zones consistently linked to higher purity and lower nitrogen breakthrough. The shut-in period further reduced nitrogen recovery from 6.49 to 1.33% and stabilized mixing behavior. Overall, minimizing late-cycle mixing zone thickness is essential for maintaining HENG storage performance. Although this study provides quantitative insights into HENG operational strategies, the use of a homogeneous grid and simplified fluid properties limits representation of geological heterogeneity and reactive processes. Future work will incorporate heterogeneity and reaction modeling into field-scale simulations to validate and refine these operating strategies for practical deployment.
No takes yet. Share an insight, caveat, or question.
Kim et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: