Nonequilibrium Two‐Phase Wellbore Modeling of CO 2 Injection Shut‐In and Restart in Depleted Gas Reservoirs

ABSTRACT CO 2 injection into depleted gas reservoirs can operate close to the vapor–liquid equilibrium envelope because post‐depletion pore pressure is often much lower than in saline‐aquifer storage. This study applies a one‐dimensional transient two‐phase mixture solver with a Lee‐type evaporation–condensation source closure, hybrid Peng–Robinson/Span–Wagner CO 2 properties, transient radial heat transfer, and an injectivity‐coupled sandface boundary to a 1500‐m tapered‐completion CO 2 injector. The base case injects pure CO 2 at 30 kg/s into a depleted reservoir at 9.0 MPa. Three shut‐in durations (1, 10, and 30 min) and three restart durations (10, 30, and 60 min) are evaluated; selected shut‐in cases are benchmarked against RELAP5‐3D under matched conditions. The two solvers agree within approximately 1.7°C on the minimum wellhead temperature. Shut‐in collapses the frictional pressure contribution and leaves the upper wellbore below the local saturation pressure. The resulting vapor‐rich upper section absorbs latent heat and cools by more than 17°C in the upper 300 m. Modeled minimum wellhead temperatures range from −19.1°C for emergency shut‐in to −8.7°C for 30‐min controlled shut‐in. Restart can create a second flash‐cooling event; the 60‐min restart is not necessarily the thermally safest option because it prolongs the subsaturation flashing window. For this case, the 30‐min shut‐in and 30‐min restart give the best balance among low‐temperature response, flashing‐window closure, and ramp severity of the three shut‐in and three restart ramp durations tested. This recommendation is scenario‐specific, not a universal operating limit, and should be supported by sensitivity analysis before material qualification or field approval.