CO 2 -enhanced gas recovery (CO 2 -EGR) offers a dual benefit of boosting natural gas production while enabling carbon sequestration. However, the pore-scale displacement dynamics of CO 2 -CH 4 interactions remain poorly understood due to limitations in conventional observation methods. This study develops an integrated low-field nuclear magnetic resonance (NMR) workflow that combines T 2 spectra, stratified T 2 spectra, 1D profiling, and NMR imaging to quantitatively monitor and visualize, in real time, the CO 2 -driven displacement of CH 4 in sandstone cores. This integrated method enabled multi-scale visualization of the CO 2 -EGR process. T 2 spectra analysis revealed that CH 4 was primarily stored in mesopores, with signal amplitude and peak area reflecting CH 4 content and recovery. Stratified T 2 spectra offered spatially resolved insights, confirming piston-like migration fronts across core layers. 1D profiling captured axial CH 4 distribution and displacement front progression, identifying CO 2 breakthrough timing. NMR imaging provided intuitive visualization of displacement morphology, revealing piston-like fronts and CH 4 redistribution. This integrated low-field NMR approach proves to be a powerful, non-destructive tool for real-time, multi-scale characterization of gas–gas displacement processes. Results show that CH 4 recovery is significantly affected by both permeability and injection rate: high-permeability cores with higher injection rates achieved greater recovery, while low-permeability cores showed limited performance. These findings validate NMR as an effective, non-destructive tool for real-time monitoring and contribute to a better understanding of CO 2 -EGR mechanisms in heterogeneous reservoirs. • First use of low-field NMR to visualize CH 4 displacement during CO 2 injection in real time at the core scale. • Integrated T 2 and stratified T 2 spectra characterized CH 4 distribution and quantified recovery during CO 2 injection. • 1D profiling and imaging visualized the displacement front, revealing piston-like behavior and identifying break through time.
Cao et al. (Mon,) studied this question.