Abstract

CO2 injection enhanced oil recovery has become one of the most important approaches to develop heavy oil from reservoirs. However, the microscopic displacement behavior of heavy oil in the nanochannel is still not fully understood. In this paper, we use CO2 as the displacing agent to investigate the displacement of heavy oil molecules confined between the hydroxylated silica nanochannel by nonequilibrium molecular dynamics simulations. We find that for heavy oil molecules, it requires more much higher displacing speed to fully dissipate the residual oil which is found related to the decreased CO2 adsorption on the silica nanochannel. A faster CO2 gas injection rate will lower the CO2 adsorption inside the nanochannel, and more CO2 will participate in the displacement of the heavy oil. The results from this work will enhance our understanding of the CO2 gas displacing heavy oil recovery and design guidelines for heavy oil recovery applications.

1. Introduction

Enhanced oil recovery [19] is critical for the petroleum industry to recover the residual oil. Among many oil recovery techniques, CO2 flooding [1017] is one of the most effective and successful methods to extract the heavy oil from reservoirs. For the past decades, due to the increasing computational power and more advanced algorithms, molecular dynamics has become an essential tool to explore underlying physics for CO2 flooding in the oil recovery process [1823]. Yan et al. [24] studied the oil inside a nanopore throat transport phenomenon employing molecular dynamics simulation. They revealed the different roles of scCO2 slug and water slug in CO2 WAG injection and investigated CO2 WAG injection-based enhanced oil recovery (EOR) mechanism. Fang et al. [25] in 2019 studied oil displacement in CO2/N2 slug flooding at the molecular and atomic levels and they proposed that the key factors for oil exploitation are the CO2 swelling effect and N2 propelling effect. They have also further calculated the mass transfer process [26]. Zhang et al. [27] used molecular dynamics to characterize the migration of the oil/gas mixture in channels and they further uncovered molecular insights about the oil charging and migration through different diameter lengths of hydrophilic nanopore throats. Since heavy oil molecules are more likely to be staying onto the surface, numerous studies have also investigated the adsorption of heavy oils on the silica surface, such as Wu et al. [28], who investigated the sorption, diffusion, and distribution of heavy oil on the silica surface using molecular dynamics simulation in 2013, and they later evaluate the diffusion and partitioning of SARA fractions of heavy oil on the soil organic matter-coated quartz surface. Zhong et al. [29] investigated the adsorption behavior of decane and other components of heavy oil as different polarity components in heavy oil on the water-wet silica surface.

In this study, nonequilibrium molecular dynamics (NEMD) simulations were used to study the oil transport phenomena inside the silica nanochannel with different CO2 pushing speed. The oil dynamic molecular behavior along the pushing direction was characterized to obtain the fundamental understanding of different oil recovery behaviors. Simulations reveal that asphaltene molecules require ultrahigh pushing speed to initiate the displacement process. Due to the large interactions between asphaltene molecules, the asphaltene molecules tend to aggregate during the entire displacement process and higher pushing speed results in smaller asphaltene molecule distribution near the silica surface and faster decaying interaction with each other. The results from this research will enhance the understanding of the heavy oil recovery behavior in the nanoconfined hydroxylated silica channel and design guidelines for heavy oil recovery enhancement applications.

2. Methods and Simulation Model

In this work, as shown in Figure 1(a), the silica nanochannel was created by cleaving the α-quartz silica along the crystallographic orientation (100) with dimensions of 215 Å×27 Å×11 Å. The hydroxylated silica nanochannel was fixed during the simulation except the hydrogen atoms. 140 asphaltene (C27H23NO2) molecules as shown in Figure 1(b) were used to simulate the heavy oil and placed inside the silica nanochannel. 2100 TraPPE [30] CO2 molecules as shown in Figure 1(c) were randomly placed outside the silica nanochannel, and a rigid graphene sheet was used to push the CO2 to displace the oil. Periodic boundary conditions were applied in all three dimensions. Nonequilibrium molecular dynamics (NEMD) simulations were performed to simulate the oil displacement process. The chosen timestep is 1 fs. The microcanonical ensemble (constant number of atoms, volume, and energy) integration was applied to the system with Langevin thermostats at 300 K to represent the thermal vibration of the oil and gas. After the structures are fully relaxed, a vacuum space at the ends is created by enlarging the simulation cell size in the pushing direction (x-axis). This allows enough space for the CO2 to travel due to the periodic boundary condition. The nonbond interactions between different types of atoms are simulated using the Lennard-Jones interaction: E=4εσ/rij12σ/rij6, where ε and σ are the energy and length constants and rij is the distance between different types of atoms. Important Lennard-Jones parameters are shown in Table 1. The L-J cutoff of the L-J interactions is 10 Å. The PPPM (particle-particle particle-mesh) is used to calculate the long-range electrostatic interaction in the system with an accuracy of 1×105. The large-scale atomic/molecular massively parallel simulator (LAMMPS) simulations are used throughout the simulations [31]. The chosen time step size is 1 fs.

3. Results and Discussion

3.1. Illustration of Oil Displacement under Different Gas Injection Rates

To study the effects of the CO2 injection rate on the asphaltene molecules inside the nanochannel, we varied the gas injection rate at 2, 4, 10 to 100 m/s. More different injection rate cases will be shown in the supporting information. The simulated asphaltene displacement processes under the 10 m/s and 100 m/s CO2 injection rates are shown in Figures 2 and 3. In both cases, the asphaltene displacement process reveals three stages: squeezing, pushing, and fully detached state. We found that the lower CO2 injection cases cannot displace all of the asphaltene molecules out of the nanochannel, while the 100 m/s injection case can easily displace all of the asphaltene molecules to the vacuum space. Such a finding is somewhat surprising, since it would be more intuitive to postulate for a proper range of the gas injection rate; under the slower injection rate, there will be more asphaltene molecules dissolved with the CO2 molecules; hence, it will reduce the interaction with the silica nanochannel.

3.2. Interaction Energy of CO2, Oil, and Silica under Different Gas Injection Rates

To further illustrate the physical process of oil displacement, we calculate the interaction energy between oil, CO2, and the silica nanochannel during the displacement process. As shown in Figures 4(a) and 4(b), a 10 m/s case and a 100 m/s injection case both show the three-stage oil displacement process: the initial squeezing, pushing, and final detachment. During the squeezing process, the interaction between the CO2 and asphaltene molecules is gradually increasing while the oil and the silica nanochannel interaction maintains the same which indicates that the real oil displacement process is not yet initiated. As the CO2 gas molecules cannot be compressed anymore, it started to push the asphaltene molecules in the nanochannel, hence the interaction between the asphaltene molecule and the silica nanochannel between to decrease. Since at the 10 m/s lower gas injection rate, there are still remaining asphaltene molecules inside the nanochannel, the final interaction between the asphaltene molecules and the silica nanochannel is around 200 kcal/mol while the higher gas injection rate (100 m/s) is around 0 which indicates that all of the asphaltene molecules are being displaced out the silica nanochannel.

3.3. Adsorption of CO2 under Different Gas Injection Rates

The local time evolution CO2 distribution near the silica interface can also be characterized by the radial distribution function (RDF) (Figure 5) calculated as gr=nr/4πr2ρr, nr is the number of atoms in a shell of thickness, Δr at a distance r from the reference atom, and ρ is the average CO2 atom number density. We take the Si atom in the silica nanochannel as the reference atom. As expected, the time evolution RDF shows that from a smaller gas injection rate to a large gas injection rate, the time CO2 saturation time becomes smaller as shown in Figures 5(a)–5(d). Under different gas injection rates, the peak location does not shift which indicates that the absorption CO2 layer structure is already stable. At the same time, the second peak for different gas injection rates becomes higher as simulation time increases which shows that more and more CO2 molecules are being displaced into the nanochannel. We further calculate the number of residual oil molecules in the nanochannel after all the CO2 molecules are being pushed into this nanoconfined space.

We have further plotted the steady-state radial distribution function of silica atoms on the nanochannel surface with respect to the carbon atoms in the carbon dioxide at different gas injections for comparison. As shown in Figure 6, there are two peaks occurring at 4 Å and 8 Å. As the gas injection rate increases, the first layer of peaks decreases which indicates that there is less CO2 adsorbed on the silica surface. The higher gas injection rate will weaken the CO2 adsorption and more CO2 will participate in the oil displacement process to remove the residual heavy oil molecules.

3.4. Number of Residual Oil Molecules under Different Gas Injection Rates

As shown in Figure 7, the larger the gas injection rate, the smaller the residual oil molecules left in the nanochannel and none are left until 90 and 100 m/s which is much higher than the light oil as reported by Liu et al. [32] which only requires 10 m/s. The large difference is mainly due to the stronger interaction of the asphaltene oil molecules with the silica nanochannel when compared with light dodecane oil molecules. The stronger the interaction between the oil molecules and the nanochannel, the larger the friction will be for these oil molecules, thus requiring a larger gas injection rate to fully displace these oil molecules.

The faster and full oil displacement can be attributed to the larger pushing force at higher pushing speed. As shown in Figures 8(a) and 8(b), both low pushing speed and high pushing speed scenarios exhibit a similar force interaction trend between carbon dioxide, oil, and the silica nanochannel. The interaction force between the carbon dioxide and the silica nanochannel gradually grows and decays because all the carbon dioxide has been pushing into the vacuum space. And the interaction force between the oil and the silica nanochannel decays to almost zero which indicates that almost all of the oil molecules have been displaced out of the nanochannel.

4. Conclusion

In summary, we have performed NEMD simulations to study the oil displacement behavior inside silica nanochannel under different CO2 pushing speeds. We observe that asphaltene molecules require ultrahigh pushing speed to initiate the displacement process. Due to the large interactions between asphaltene molecules, the asphaltene molecules tend to aggregate during the entire displacement process and higher pushing speed results into smaller asphaltene molecule distribution near the silica surface and faster decaying interaction with each other. A faster gas injection rate will lower the CO2 adsorption ability on the silica nanochannel surface; hence, more CO2 will participate in the displacement of the heavy oil. The results from this work will enhance the understanding of the heavy oil recovery behavior in the nanoconfined hydroxylated silica channel and design guidelines for heavy oil recovery enhancement applications.

Data Availability

Data is available upon request.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

This work was financially supported by the National Natural Science Foundation of China (no. 51604293), the Shandong Provincial Natural Science Foundation, China (no. ZR2016EEB30), the Fundamental Research Funds for the Central Universities (no. 17CX02009A), the Qingdao Applied Basic Research Program (no. 17-1-1-32-jch), the Scientific Research Foundation of China University of Petroleum for Talent Introduction (no. YJ201601093), and the National Science and Technology Major Project (2016ZX05031-002).

Exclusive Licensee GeoScienceWorld. Distributed under a Creative Commons Attribution License (CC BY 4.0).