To clarify how injection media affect oil replacement depth and pore mobilization in tight oil reservoirs, this study conducted three huff-and-puff cycles with \(\hbox {CO}_2\) , \(\hbox {CO}_2\) +active water, and \(\hbox {N}_2\) via laboratory displacement tests, combining macro-pressure dynamics, recovery data, and micro-NMR pore analysis. Experimental materials included two core groups: low-permeability cores (A-group: 14.3 \(-\) 19.59 mD, 2.5 \(\times \) 30 cm) and tight cores (B-group: 0.0153 \(-\) 0.0255 mD, 2.5 \(\times \) 5 cm); fluids were 0.5% active water and high-purity \(\hbox {CO}_2\) / \(\hbox {N}_2\) . Results show \(\hbox {CO}_2\) performs best: first-cycle recovery 20%, oil exchange rate 0.214 ml/ml, pore threshold 0.008 \(\mu \) m, via miscible displacement and adsorbed oil desorption. \(\hbox {N}_2\) (16% first-cycle recovery) relies on pressure drive with limited desorption. \(\hbox {CO}_2\) + active water (7% first-cycle recovery, 0.01 \(\mu \) m threshold) is constrained by active water viscosity but selects adsorbed oil. This study supports injection media optimization, highlights \(\hbox {CO}_2\) ’s micro-pore advantage for EOR, aiding carbon utilization in tight reservoirs and carbon neutrality.
