Objective Long-term shutdown of a supercritical CO₂ pipeline typically results in gas-liquid two-phase coexistence inside the pipeline. Direct restart brings considerable safety risks. Owing to engineering constraints, the inlet station is equipped only with compressors.For restart under gaseous CO2 charging, systematic experiments and the governing laws are still lacking. This study examines coupled temperature-pressure changes and phase evolution under gaseous CO₂ charging. It quantifies how initial liquid holdup and charging flow rate affect restart time and CO₂ consumption. It also develops a restart-time prediction formula for this condition.

Methods A fully transparent high-pressure CO₂ phase evaluation experimental setup is self-designed and constructed. Gaseous CO₂ charging and restart experiments are carried out with initial liquid holdup values of 0.6, 0.7 and 0.8 and charging flow rates of 25 mL/min and 45 mL/min. The phase change process is recorded in real time by a high-speed camera system, and temperature and pressure data are collected synchronously. On the basis of mass conservation and the Peng-Robinson (P-R) equation of state, a theoretical prediction formula for full liquefaction time is derived. The theoretical formula is further modified by considering the pressure change of CO₂ inside the pipeline and the density change of CO₂ charged at the pipeline inlet.

Results Under gaseous CO₂ charging, CO₂ followed a sequential path: gas-liquid two-phase → liquid → dense phase. Pressure showed three stages: stable, slow rise, and sharp rise. At 25 mL/min, raising initial liquid holdup from 0.6 to 0.8 shortened complete liquefaction time from 1,098 s to 155 s. This was an approximately 86％ reduction. CO₂ consumption fell from 457.5 mL to 64.6 mL. At an initial liquid holdup of 0.6, raising the flow rate from 25 to 45 mL/min shortened complete liquefaction time by about 49.5％. CO₂ consumption decreased by about 9.2％. However, an excessively high flow rate accelerates pressurization and introduces safety risks. For the modified formula, the mean absolute percentage error against experimental data decreased to 5.79％. The theoretical formula gave 8.02％.

Conclusion Gaseous CO₂ charging enables the gradual liquefaction and slow pressurization of CO₂ from the initial gas-liquid two-phase state. Initial conditions and charging flow rate are important factors affecting the full liquefaction process. The proposed prediction model provides a theoretical basis for estimating the restart time of supercritical CO₂ pipelines during restart under gaseous CO₂ charging.