Abstract: Energy supply chains face uncertainties due to complex international geopolitics. Gas storage, as a crucial component of the national energy reserve, buffers fluctuations in natural gas prices and mitigates supply disruptions, ensuring stable supply to residential and industrial sectors and safeguarding national energy security. For efficient and safe operation of gas storage under multi-component, multi-cycle injection-production operations, accurate understanding of the mixing and diffusion of working and cushion gases in the reservoir is essential. However, experimental studies on gas diffusion have largely focused on shale, coal, and tight formations, leaving the diffusion behavior of multi-component gases in carbonate reservoirs poorly understood. In this study, carbonate samples from the Upper Carboniferous Huanglong Formation (upper member) of the Wolonghe Gas Field were examined, and their pore size distribution was characterized by nuclear magnetic resonance and high-pressure mercury intrusion porosimetry. Diffusion experiments were conducted on gas mixtures containing CH₄ with CO₂, N₂, and O₂. Based on comparative analysis of the experimental data, an appropriate mathematical model for gas diffusion coefficients in multi-scale carbonate reservoirs was identified. The carbonate samples exhibit a pronounced multi-scale pore structure. Under identical temperature and pressure conditions, higher porosity and permeability correspond to larger diffusion coefficients for all gas species; moreover, the binary diffusion coefficient of the CH₄-CO₂ pair exceeds that of the N₂-CO₂ pair. In a multi-component system, O₂ exhibits the largest diffusion coefficient, followed by CH₄, while N₂ and CO₂ have the smallest; furthermore, the presence of O₂ affects the sensitivity of the CH₄ and N₂ diffusion coefficients to changes in the volumetric fractions of CO₂ and N₂. The calibrated diffusion coefficient model can be extended to predict diffusion coefficients under different temperature and pressure conditions. These findings provide experimental and computational methodologies for the accurate prediction of gas storage operations and the rational design of operational strategies.

Key words: Carbonate rock, Diffusion coefficient, Multi-component, Diffusion experiment, Mathematical model