Video: Capacity Assessment of CO2 Storage and Enhanced Oil Recovery in Residual Oil Zones
- Bailian Chen (Los Alamos National Laboratory) | Rajesh Pawar (Los Alamos National Laboratory)
- Document ID
- Society of Petroleum Engineers
- Publication Date
- Document Type
- 2018. Copyright is retained by the author. This presentation is distributed by SPE with the permission of the author. Contact the author for permission to use material from this video.
- 7.6.6 Artificial Intelligence, 5 Reservoir Desciption & Dynamics, 5.4.2 Gas Injection Methods, 5.4 Improved and Enhanced Recovery, 5.4 Improved and Enhanced Recovery
- Machine learning, Enhanced oil recovery, CO2 storage, CO2 flooding, Capacity assessment
- 0 in the last 30 days
- 2 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 5.00|
|SPE Non-Member Price:||USD 10.00|
Residual oil zones (ROZs) are defined as those zones where oil is swept over geologic time period (natural flush) and exists at residual saturation. ROZs are increasingly being commercially exploited using CO2-enhanced oil recovery (EOR) (in Permian Basin). In this study, CO2 storage potential, long-term CO2 fate and oil recovery potential in ROZs are characterized. We use numerical simulations of CO2 injection with a reservoir model based on data from the Permian Basin. The changes of CO2 storage capacity and potential oil recovery with amount of CO2 injection are investigated. The effects of different well patterns (five-spot and line drive) and well spacing on fraction of CO2 retained in reservoir and cumulative oil production are also investigated. Furthermore, the effect of different CO2 injection modes, i.e., continuous CO2 injection and water-alternating-gas injection (WAG), on the CO2 storage and EOR potential are evaluated and compared. After the preliminary characterization of CO2 storage and EOR potential in ROZs, we next develop empirical models that can be used for estimating the CO2 storage capacity and oil production potential for different ROZs. A supervised machine learning algorithm, Multivariate Adaptive Regression Splines (MARS, (Jamali et al.)) is used for developing the empirical models.
Results show that CO2 retention efficiency and oil recovery vary non-linearly with amount of CO2 injected. It is observed that long-term CO2 fate is a function of CO2 injection amount and significant fraction of reservoir CO2 resides in hydrocarbon phase. Five-spot well pattern results in more oil production and larger amount of CO2 retained in reservoir than line-drive well pattern. During the investigation of well spacing, we observe that less number of wells actually results in higher CO2 retention and oil recovery, and less number of wells can also result in less probability of wellbore leakage. In comparison of WAG and continuous CO2 injection modes, it is observed that WAG injection has higher fraction of injected CO2 retained in reservoir, but with slightly lower cumulative oil production. In the study of empirical models for the capacity assessment of CO2 storage and EOR, results show that MARS can generate high-fidelity empirical models that can be used to predict the cumulative CO2 storage capacity and cumulative oil production for different ROZs.