Propane simulated in silica pores: Adsorption isotherms, molecular structure, and mobility (2023)

Shale oil and gas primarily exist in nanoscale pore-fracture networks. Thus, it is essential to clarify the flow behavior of hydrocarbon in confined nanopores. This paper summarizes shale oil and gas distribution and exploration status, reviews the microscopic pore structure characterizing methods and methods of investigating fluid behavior in micro-nanopores. Because the multiphase fluid experiment in nanopores requires sufficiently high precision instruments, and the temperature and pressure of the experiment are difficult to indicate the actual formation conditions, molecular dynamics simulation is the inevitable research technique for investigating fluid behavior in shale nanopores. Therefore, the most recent developments and conclusions of molecular dynamics simulation for adsorption, diffusion and flow (displacement) in shale reservoirs are introduced. The simulated substrates include organic matter (graphene, CNT and kerogen), inorganic matter (quartz, calcite, clay, etc.), and mixed matter. The simulated oil and gas component include pure composition and multi-composition. In addition, water, carbon dioxide, supercritical carbon dioxide, and nitrogen are also utilized in the simulation as the displacement or coexistence phases. The temperature, pressure, slit width, water content, oil content and surface characteristics determine the distribution and transportation of fluid in the pore, and consequently impact the production rate. In addition, this paper also discusses the mechanisms of enhanced oil and gas recovery in shale reservoirs, and summarizes the future development of the molecular dynamics simulation of shale oil and gas.

Ions in brine significantly affect EOR. However, the mechanism of EOR with different brine is still controversial. By Combining Molecular Dynamics (MD) method and Quartz Crystal Microbalance with Dissipation (QCM-D) technology to analyze ions distribution and the mechanisms in detaching acidic components on the sandstone, an effective method to determine the detaching capacity was established. The results show that detaching capacity is related to ions distribution and hydration capacity. In the oil/brine/rock system, ions far from the rock are favorable for detaching, while ions near the rock are unfavorable for detaching due to ion bridging effect. The hydrogen bond between water and naphthenic acid is key to detaching. Cations strengthen the detaching by forming hydrated ions with water, and the detaching capacity is negatively correlated with hydrated ions radius and positively correlated with the water coordination number. The detaching determination coefficient was established by considering the ions distribution, ions types, and hydration strength, then verified by QCM-D. The brine detaching capacity with different Ca²⁺/Mg²⁺ ratios was predicted based on MD and detaching determination coefficient, and verified by QCM-D. The optimal Ca²⁺/Mg²⁺ ratio gave 7:3. This study provides theoretical guidance for targeted regulation of brine composition to improve the recovery of tight sandstone reservoir.

We propose an approach that combines classical density functional theory (DFT) and molecular dynamics (MD) simulation to study fluid behavior in nanopores in contact with bulk fluid (macropores). The approach consists of two principal steps: a DFT calculation of fluid composition and density distribution in a nanopore under specified thermodynamic conditions and an MD simulation of the confined system with obtained characteristics. Thus, we investigate an open system in a grand canonical ensemble. This method allows us to identify both structural and dynamic properties of confined fluid at given bulk conditions. Our approach does not require a computationally expensive simulation of a bulk reservoir. This work presents equilibrium density profiles of pure methane, ethane and carbon dioxide and their binary mixtures in slit-like nanopores with carbon walls. Good agreement between the structures obtained by theory and simulation confirms the applicability of the proposed method.

Nanopore networks comprising shale reservoirs exhibit special behaviors in two-phase flow. We use molecular dynamics simulations (MD) to explore the single n-alkanes (nC3-nC10)-water and mixed n-alkanes (nC3, nC6, and nC10)-water flow mechanisms in the confined nanopores of quartz. The results reveal that water molecules form four adsorbed layers near the wall, showing greater viscosity than the bulk water. For single n-alkanes cases, in the n-alkanes-water interface regions (AWIR), n-alkanes molecules tend to be parallel to the pore wall, and the tendency is more pronounced for longer n-alkanes, resulting in a decrease of AWIR width. Moreover, we observed typical liquid–liquid slip in AWIR. Ignoring the slip velocity can underestimate the flow rate of n-alkanes more than 20%. Under the driving force, the bulk n-alkanes exhibit distinct different velocity distributions due to viscosity differences. However, the total fluid velocity distributions in AWIR are quite the same, showing an apparent viscosity independent of the type of n-alkanes and pore size. For mixed n-alkanes, the velocity distribution and apparent viscosity are still the same as single n-alkanes cases due to the unchanged structure of the molecules in AWIR. Our study could provide an in-depth understanding of the oil–water two-phase flow in nanopores, especially the influence of carbon chain length on oil–water interface properties.

Hydrogen storage plays a fundamental role in the future hydrogen energy system, and carbon aerogel is one of the most potential hydrogen storage materials because of its high gravimetric and volumetric density on hydrogen adsorption. In this paper, the amorphous structure of carbon, obtained by a numerical simulation process by using the molecular dynamic and Monte Carlo methods, as well as the primary unit method, was intercepted as a sphere structure for numerical annealing to build a carbon nanosphere, which serves as the basic unit to reconstruct the carbon aerogel's skeleton by the Diffusion Limited Cluster Aggregation (DLCA) method. The hydrogen adsorption in carbon aerogel was simulated by using the self-coding parallel grand canonical Monte Carlo (GCMC) method. The influences of particle diameter, density, temperature, pressure, and specific surface area on the hydrogen adsorbing capacity in carbon aerogel were analyzed in detail. The results showed that the carbon aerogel's hydrogen storage capacity with a specific surface area of 2680m²/g could reach 4.52wt % at 77K and 3.0MPa.

A fundamental understanding of dynamic wettability alteration at complex solid interfaces in varying fluidic environments representative of the subsurface environments is essential for engineering field-scale strategies to enhance energy recovery while reducing environmental impacts. In this study, we characterize dynamic wettability alteration at solid surfaces (e.g., calcite, silica and illite) resulting from the self-assembly and adsorption of polyaromatic macromolecules known as asphaltenes in heptane and toluene solvents at 200bar and 313K using classical molecular dynamics simulations. These conditions are chosen to be representative of the subsurface environments. On adding asphaltenes, the equilibrium contact angles on calcite, silica and illite surfaces which are in the range of 25–34° increased to 69–79°. The adsorption and self-assembly of asphaltene molecules on the solid surfaces and the accumulation on water/hydrocarbon interfaces reduces the spreading distances of water molecules and enhances the hydrophobicity of the solid surfaces. Further, the accumulation of asphaltene molecules at the water/liquid hydrocarbon interfaces results in a reduction of the interfacial tension of water/heptane and water/toluene, thus stabilizing the water droplet. The enhanced surface hydrophobicity contributes to the formation of layered structures of heptane and toluene close to the solid surfaces and reduces the self-diffusivities of these molecules close to the surface.

Fuel, Volume 180, 2016, pp. 718-726

Adsorbed methane in shale organic nanopores is an important factor of the shale gas resource. The major component of shale organic matter, kerogen solids, consists of macro organic molecules and functions as the most important adsorbent. In this work, a novel molecular simulation workflow is proposed to generate organic pores on residue-type kerogen molecules and to simulate the gas adsorption in the pores. The molecular dynamics-based cutter atom pore generation method can construct pores with arbitrary shapes and sizes—approaching microscopy observations—2–50nm shale matrix pores with reasonable physical significance. Grand canonical Monte Carlo simulations for CH₄ and CO₂ are performed on kerogen pores with various pore radii using two categories of molecular potentials. The ensemble averaging density distributions in the pores are calculated and analyzed, which concludes that the free gas state becomes distinguishable from the adsorbed state unless the pore radius exceeds 1nm. Adsorbed layers at the kerogen pore surfaces are heterogeneous because of non-uniformly distributed adsorptive sites on the surfaces. Adsorption isotherms are simulated and thereafter fitted with the Langmuir equation comparing various molecular models and fluid types. The adsorption affinity indicated by the Langmuir pressure for the total adsorbed molecule number in all cases decreases as a function of increasing pore size.

Fluid Phase Equilibria, Volume 458, 2018, pp. 177-185

The effect of nano-confinement on the thermodynamic behavior and saturation properties of petroleum fluids in nanoporous media are important parameters for shale gas/oil production. Confined petroleum fluids can have hysteresis and bubble/dew point behaviors can be very different depending on the thermodynamic route. In this work, we use grand canonical Monte Carlo (GCMC) simulations and engineering density functional theory (DFT) to study the effect of pressure, temperature, and nanopore sizes on the bubble/dew point and hysteresis of hydrocarbons in nanopores. By comparing to GCMC simulations, engineering DFT reliably predicts the vapor-liquid equilibrium of confined hydrocarbon fluids. We also find that dew point of pure confined fluids approaches bulk saturation point as pore size increases, but bubble point can be very different from bulk even for very large pores. As a result, the hysteresis between bubble and dew points increases with pore size. For single-component, the hysteresis decreases as pressure approaches critical pressure. For binary mixtures, the dependence of hysteresis on temperature is dependent on the pore sizes. The cricondentherm point of confined mixtures is shifted toward higher temperatures comparing to the bulk.

Fuel, Volume 253, 2019, pp. 1351-1360

Understanding the enhanced liquid flow behaviors in nano-scale elliptic pores with different surface wettability has tremendous implications in science and engineering, such as water purification and shale oil recovery with abundant irregular circular (elliptic) pores. In this study, the apparent viscosity and enhancement factor model are proposed considering boundary slip velocity and effective viscosity, which is related to the contact angle, pore dimensions and shapes, and the boundary slip velocity depends on the no-slip Hagen-Poiseille equation depending on near-wall water viscosity. In addition, the proposed model is compared and validated with the results from theories and molecular dynamic simulations. Results show that the apparent viscosity decreases with an increasing contact angle, and tends to bulk viscosity with the increase of pore dimension. With contact angle ranging 0 ∼ 180°, the enhancement factor can decrease one order of magnitude with apparent viscosity larger than bulk viscosity, and increase eight orders of magnitude with apparent viscosity smaller than bulk viscosity. With an increasing ratio of semi-long axis a to semi-short axis b of elliptic pores, the apparent viscosity increases first and then decreases down to 0 for a small contact angle, and the enhancement factor decreases first and then increases. In addition, for a larger contact angle, the apparent viscosity gradually decreases and tends to 0, and the enhancement factor increases. The changes of apparent viscosity and enhancement factor are directly related to the changes of boundary velocity and effective viscosity which are caused by the surface wettability, pore dimensions and shapes. The proposed model can be used to describe the liquid flow in nano-scale elliptic pores with different surface wettability, like the shale oil transport in oil-wet organic and water-wet inorganic pores.

Fuel, Volume 171, 2016, pp. 74-86

Understanding the transport of liquid hydrocarbon through nanopores of inorganic minerals is crucial not only to develop liquid-rich shale reservoirs, but also to grasp oil migration from deeply buried extremely low permeability source rocks. We report a molecular study of liquid hydrocarbon (octane) flow through inorganic (quartz) nanopores ranging in size from 1.7 to 11.2nm. Through equilibrium molecular dynamics (EMD), we observe the layering structure of confined octane and conclude that in the center of slits having apertures greater than 3.6nm, the octane properties, e.g., density, self-diffusion coefficient, and viscosity, tend to be bulk-liquid-like. Near the solid–liquid interface, octane molecules diffuse more slowly. Then we use nonequilibrium molecular dynamics (NEMD) to study the pressure-driven flow of octane in quartz slits and present two methods to characterize the behavior: (1) slip length coupled with effective viscosity and (2) apparent viscosity. The Navier–Stokes equation can reasonably describe the flow in quartz nanopores larger than 1.7nm; however, a slip boundary condition or viscosity correction is essential. Although the slip length (∼0.9nm) is small, significant error can be caused in the estimation of overall flux if it is neglected. The variations in slip length and apparent viscosity with driving force, pore size, and temperature can be described by empirical exponential functions. These results can be readily incorporated into existing techniques to estimate apparent liquid permeability of shale—the most fundamental property required for shale exploitation.

Petroleum Exploration and Development, Volume 42, Issue 6, 2015, pp. 844-851

Molecular dynamics simulation was used to study the occurrence state of liquid alkane in pores and slits of shale organic matter. On the basis of OPLS (Optimized Potentials for Liquid Simulation) force field, the alkane densities under different pressures and temperatures were calculated; the comparison with experimental results validated the effectiveness of this approach. With n-heptane as an example, the basic occurrence behaviors of alkanes in the pores and slits of organic matter were analyzed under formation conditions, and the effects of slit aperture, thermal maturity of organic matter, carbon chain length and isomers were also discussed. Results show that: (1) The density distribution of the alkanes across the pores and slits is not uniform, but presents a periodic fluctuation; (2) A “solid-like” alkane layer will form in the vicinity of the solid surface, and its density approximates to 1.9−2.7 times as large as that of the bulk-fluid; (3) Multiple adsorption layers are always shown for liquid alkanes and the thickness of each layer is 0.48 nm; the total number of adsorbed layers is influenced by the slit aperture and fluid composition. Finally, using this approach, the proportion of adsorbed-phase (18.2%) is determined for oil in an organic matter slit.

International Journal of Heat and Mass Transfer, Volume 98, 2016, pp. 675-686

Methane adsorption and its effect on fluid flow in shale matrix are investigated through multi-scale simulation scheme by using molecular dynamics (MD) and lattice Boltzmann (LB) methods. Equilibrium MD simulations are conducted to study methane adsorption on the organic and inorganic walls of nanopores in shale matrix with different pore sizes and pressures. Density and pressure distributions within the adsorbed layer and the free gas region are discussed. The illumination of the MD results on larger scale LB simulations is presented. Pressure-dependent thickness of adsorbed layer should be adopted and the transport of adsorbed layer should be properly considered in LB simulations. LB simulations, which are based on a generalized Navier–Stokes equation for flow through low-permeability porous media with slippage, are conducted by taking into consideration the effects of adsorbed layer. It is found that competitive effects of slippage and adsorbed layer exist on the permeability of shale matrix, leading to different changing trends of the apparent permeability.