Experimental and numerical study on gas Huff-n-Puff EOR in shale oil reservoirs

Gas huff-n-puff injection which avoids the viscous fingering phenomenon or gas channeling between connected wells is a potential method to efficiently enhance oil recovery (EOR) in shale oil reservoirs. In this dissertation both experimental study and simulation work were performed to investigate gas huff-n-puff in shale oil cores and reservoirs, including the feasibility of gas huff-n-puff in shale reservoirs. The key work includes nanopore effect on shale oil production, MMP in shale reservoirs, upscale from lab experiment to field production, gas penetration depth, and the EOR performance of different injection gases. This research aims to understand and apply gas huff-n-puff EOR in shale oil reservoirs. The laboratory study examines the performance of CH4 huff-n-puff, core size effect on gas huff-n-puff, the effect of injection pressure, and the performance of different injection gases. The pore size distribution in shale cores is obtained by a mercury injection experiment. CT scan test was also implemented to check the oil saturation and cumulative oil recovery for each cycle. Minimum miscible pressure (MMP) was estimated for Wolfcamp crude oil and CO2 system by using slimtube tests for conventional cores and designed CO2 huff-n-puff tests for shale cores. The experiment results show that CH4 huff-n-puff is an effective EOR method in shale reservoirs. Under the same operation schedule, the larger scale core will result in a lower oil recovery. 93.7% of pore diameters in the Wolfcamp shale core samples are less than 10 nm. The effective gas huff-n-puff injection pressure was 200 psi higher than the measured conventional MMP. Then the lab simulation models were built and history matched with the experiment data. The calibrated numerical models were then employed to perform a series of sensitivity studies to investigate the effects of operation parameters on oil recovery in shale oil cores, such as the number of injection cycles, molecular diffusion, soaking time, and operation schedule. Field simulation models were built and history matched with field production data. The gas penetration depth in fractured reservoir was investigated. The EOR performance of different injection gases in field scale study was also discussed in this study. The results illustrate that the CO2 penetration depth is about 105.6 ft during 100 days’ huff time in the first huff-n-puff cycle, covering about 36% of the SRV region, reducing the oil viscosity by 30% in the gas penetrated region. A wet gas has better performance during the huff-n-puff EOR process than that of a dry gas. The capillary pressure curve in a real shale reservoir with a multicomponent Wolfcamp oil is the first time generated using the modified PR-EOS model combining the effects of changed critical properties and pore size distribution. The results show that for the Wolfcamp oil, the bubble point pressure was suppressed by 17.3 % when r is10 nm and suppressed by 63.8% when r is 1.5 nm. The nanopore confinement suppresses the bubble point pressure, which may predict a long-lasting flat producing GOR in the Wolfcamp oil reservoir. A new expression of dimensionless pressure was generated to describe the huff-n-puff efficiency, and a new version of dimensionless time was derived to describe the performance of different scales. Several validation tests were conducted for different fluid and rock parameters, well constraints, and operation schedules. A type curve was developed for different scales which demonstrates that all sizes yield a similar relationship between oil recovery and dimensionless operation time. It proves that the cumulative oil recovery of gas huff-n-puff EOR in shale oil reservoirs can be predicted.