Experimental and Molecular Modeling Study of Bubble Points of

Subscriber access provided by Fudan University

Article

Experimental and Molecular Modeling Study of Bubble Points of Hydrocarbon Mixtures in Nanoporous Media Manas Pathak, Hyeyoung Cho, and Milind D. Deo Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.6b02422 • Publication Date (Web): 12 Jan 2017 Downloaded from http://pubs.acs.org on January 13, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Experimental and Molecular Modeling Study of Bubble Points of Hydrocarbon Mixtures in Nanoporous Media Manas Pathak, Hyeyoung Cho and Milind Deo Department of Chemical Engineering, University of Utah, Salt Lake City, UT, USA *

Corresponding author e-mail: [email protected] Keywords: bubble point, confinement

Abstract

The shale play resources have played a key role in increasing oil production in the last decade in the United States. The sizes of pores in shales storing the oil are believed to be of the order of nanometers. It is believed that the fluids present in such small nanometer scale pores have different properties compared to properties measured in the bulk. Fluid saturation pressures at given temperatures – bubble points for oils and dew points for condensates in the nanopores are affected by the influence of pore walls in the vicinity of the fluid molecules. Approach to bubble point or dew point influences the proportion of liquid or gas produced from a given well, and thus impacts the economic viability. Hence an accurate measure of saturation pressures is important. In this paper, we describe experiments in well characterized synthesized mesoporous materials and present Gibbs Ensemble Monte Carlo (GEMC) simulations understanding possible reasons for observations made in the experiments. The experimentally measured saturation pressure of a mixture of decane-methane in confined spaces of the mesoporous material are observed to be less than the saturation pressure of mixture in the bulk state. The GEMC simulations were performed to investigate fluid phase equilibrium in confined pores and find possible reason(s) behind the suppression of bubble points in the confined spaces. The simulations show that reduction in critical properties of the nano-confined fluids lead to the suppression of bubble point pressures of the fluids mixtures in confined pores.

ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Introduction The oil and gas production from shale plays have significantly increased the overall oil and gas production in the United States1. Of the total U.S. crude oil production in the year 2015, roughly 52% or about 4.5 million barrels per day of crude oil were produced directly from shale and other tight rock resources according to U.S. Energy Information Administration (EIA). Figures 1 and 2 shows the recent total production from shale gas and tight oil (majorly shale oil) plays, respectively, in the U.S.

Figure 1: U.S. Energy Information Administration (EIA) official shale gas production data through July 2016. The figure shows current gas production of over 40 BCF of gas per day from US Shale gas plays. Source: U.S. Energy Information Administration (EIA).


Page 2 of 26

Page 3 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 2: U.S. Energy Information Administration (EIA) official tight oil production data through July 2016. The figure shows current oil production of over 4.0 million barrels of oil per day from US tight oil plays. Tight oil in US is produced majorly from low-permeability shale formations but also include sandstones and carbonates. Source: U.S. Energy Information Administration (EIA).

Even with successful shale exploitation in the U.S. in the last decade, there is still a constant challenge of increasing the recovery from the shale reservoirs. Much of the decades of understanding from conventional oil and gas exploration fails to increase primary recovery over 10% in shale oil2,3 and 20% in shale gas plays. In some cases, the shale rocks are also the source of oil and gas found in these reservoirs and contain organic matter, known as kerogen4,5, in them. Kerogen is believed to be the precursor to oil and gas found in shales. The nanometer scale pores (1-20 nm) in organic (kerogen) and non-organic parts of shales have been identified earlier6 and would be referred as ‘nanopore’ in rest of the paper. Such small pore dimensions add another complexity in understating the behavior of fluids present in shales. Currently, there is still a huge gap in the understanding of storage and transport of hydrocarbons in shales. When the dimensions of pores are comparable to the mean free path of the fluid molecules, the collision with the wall of the pores cannot be neglected. At the nanoscale, in the confined pores of shales, phase behavior of fluids not only depends on fluid-fluid interactions, as in the bulk state, but also


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 26

depends on fluid – pore wall interactions7,8. The pore area per unit volume increases as pore dimensions decrease, therefore the surface forces are not negligible at nanoscale9.

This

interaction between fluid molecules and the pore wall is thus required to be included in any calculation of thermodynamic fluid properties or transport properties in shales. However, the traditional equation of state fails to add the complexity of pore wall interaction in calculating thermodynamics properties of confined fluids. The current work aims to study the effect of pore wall – fluid interaction on pressure-volume-temperature (PVT) properties of the fluids residing in nanopores of the shale rocks. In recent years, there has been major interest among researchers in the oil and gas field in looking at the fluid thermodynamic properties in confined pores7–20. The current work employs molecular modeling techniques to look at the equilibrium between co-existing vapor-liquid phase in the nanopores and compare it to the experiments conducted on artificially created meso (~3.5 nm pore size) porous material. The potential use of a technique known as Gibbs Ensemble Monte Carlo simulations (GEMC)21,22 has been explored in this paper for understanding phase equilibrium of hydrocarbon fluids under confinement in shale systems. In GEMC, the phase equilibrium between liquid and vapor phases is directly established inside a nanopore and thus, relates well to the physical description of confined fluids in the nanopores of shales. GEMC method directly gives the densities of the coexisting phases inside a nano pore. Sheng11 has demonstrated through experimental work that there are two phases of fluids inside the nanometer scale pores – the free confined phase and the adsorbed phase. Such small pores are also abundant in shales. However, in author’s knowledge, a fundamental understanding for the difference in thermodynamic properties of free confined fluid and bulk fluid has not been well characterized. In order to understand the phase behavior of confined fluids, this work aims to study the vapor-


Page 5 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

liquid equilibrium of confined fluid while ignoring the other possible effects such as adsorption. The free but confined phase has properties different than the bulk fluid because of the interactions with the pore wall. The free-confined fluid phase is different than the adsorbed phase, however the two phase may remain in thermodynamic equilibrium. The adsorbed part of fluids stays in the form of mono to multiple layers of fluid molecules adsorbed on the pore wall inside the pore. The vapor part of adsorbed fluid may condense beyond a threshold temperature and pressure into condensed liquid phase leading to the confined capillary condensation phenomena19. Such condensation usually happens at the temperature and pressure below the phase boundary of the bulk fluid. However, for the current research, the effect of adsorption and possible capillary condensation is ignored to focus on the effect of confinement on the behavior of free fluid phase in the nanopores11. Further, the effect of high capillary pressure in smaller pores is also ignored in the current work. Each nanopore in shale has its own vapor-liquid equilibrium established within the pore. Therefore, a series of Gibbs Ensemble Monte Carlo (GEMC)21,22 simulations are performed to understand fluid phase equilibrium within a nanometer scale pore. The key PVT properties affected by the confinement of fluids include phase boundaries, phase compositions and saturation pressures (bubble points and the dew points), interfacial tensions, fluid viscosities and fluid densities. For this work, effect on densities, phase boundaries and the bubble points are studied using a combination of experiments and molecular modeling. The experiments were conducted to measure the bubble points of the liquid hydrocarbon mixture (decane-methane in 90-10% molar ratio) confined in nanometer range pores. For the experimental part, synthesized mesoporous materials with perfectly calibrated pore sizes in nanometers were used to represent the porous system in shale reservoirs. Nitrogen adsorption/desorption isotherms, and transmission electron microscopy (TEM) were


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

used to characterize the samples. Measurements show that the synthesized materials have a highly ordered hexagonal mesostructure and a narrow pore size distribution. Bubble point pressure measurements of mixtures of decane-methane were performed in these mesoporous materials. The experimental results are complemented with the results from molecular simulations (GEMC) to understand the dynamics at pore scale and investigate reasons behind suppression in the bubble point pressure observed in the experiments. Simulations methodology Gibbs Ensemble Monte Carlo (GEMC)21,22 simulations are the kind of molecular-scale simulations that have widely being used by researchers to study vapor-liquid equilibrium including in confined spaces19–23. There have not been wide applications of GEMC simulations to study the effect of confinement in ultra-tight oil and gas formations such as shales. The GEMC simulations technique is used in the current work to understand the effect of confinement in graphite slit pores, of fluid molecules on fluid density and phase behavior of the fluid mixture. GEMC simulations were performed for pure decane and a binary mixture of decane-methane in 90-10 % molar ratio confined in graphite slit pores of 3.5 nm slit width. The 3.5 nm was specially selected as the average pore size in the synthetic SBA-15 porous media on which experiments were performed was 3.5 nm. The GEMC simulations work on the principle of fulfilling the following four thermodynamic criteria to establish phase co-existence between two regions21,22: 1. The two regions should be in internal equilibrium 2. Temperature of the two regions should be same 3. Pressure of the two regions should be same 4. Chemical potential of all components in the two regions should be same


Page 6 of 26

Page 7 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Once the simulations are set to be performed at a constant temperature, following three Monte Carlo moves are performed in the Gibbs Ensemble to achieve thermodynamic equilibrium between two co-existing phases: 1. Displacements of particles within each region to ensure equality of internal energy 2. Fluctuations in the volume of two regions to ensure equality of pressure 3. Transfer of particles between two regions to ensure equality of chemical potential The figure 3 shows the schematic of Gibbs Ensemble method and explains the three kinds of Monte Carlo moves mentioned earlier.

Region I

Region II

Figure 3: The Gibbs Ensemble Monte Carlo Simulations establishes the equilibrium between two fluid phases by employing three kinds of move in the two regions- Region I, II. The two regions are taken from the two simulation boxes, each containing two different phases of the fluid. The three moves performed in the two regions taken are particle displacement, volume change and particles transfer and are shown in Figure 3. The figure is redrawn from the work by Panagiotopoulos22

In the current work, GEMC simulations were performed on pure decane (Figure 4) and a binary mixture of the 90-10% molar ratio of decane-methane mixture. The set-up of GEMC simulations performed between two regions of liquid and vapor phases of decane respectively is


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

shown in Figure 4. To include the effect of confinement, the GEMC simulations were performed on the fluid(s) confined in the slit pore geometry made up of graphite walls (Figure 4 and 5). Such a slit pore is shown in the form of the molecular model in Figure 5.

Figure 4: The Gibbs Ensemble Monte Carlo (GEMC) simulations were performed on pure decane and a binary mixture of decane and methane (90-10% molar ratio) confined by graphite walls. Figure 4 shows set up of the GEMC simulation performed for pure decane. The three kinds of moves in the GEMC simulations were performed in the two regions made up of graphite walls.


Page 8 of 26

Page 9 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 5: The molecular model showing fluid mixture confined in graphite slit pore. H is the height between the graphite layers forming a pore. The H will correspond to the pore dimensions in the experiments discussed later in the paper. This H is also represented in the wall-fluid potential discussed later in the text.

The molecular simulation methodology involved performing canonical (nvt) GEMC simulations at constant temperature. In canonical (nvt) ensemble, the number of molecules (n), volume of the system (v) and temperature of the system (t) is kept constant throughout the simulations. As the simulations are performed at different temperatures, phase equilibrium established between liquid and vapor phases shift. Such a temperature dependent shift in equilibrium between two phases is evident by the different mole fractions of each component in the two phases and different density of each phase as the temperature is changed. Using series of simulations, vapor-liquid equilibrium is simulated at different temperatures in this work. At every isothermal phase equilibrium simulation, densities and mole fraction of each component in the two phases is monitored.


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 26

The 12-6 Lennard-Jones25 (L-J) potential was used to account for non-bonded interactions between the fluid molecules. The TraPPE-UA force field parameters were used in L-J potential which is described as: ∅ = 4

− --------------------(1)

where is the depth of the potential well, σ is the finite distance at which the inter-particle potential is zero, r is the distance between the particles. The Lorentz-Berthelot26,27 combining rules are used for calculating the parameters for unlike molecules

= + --------------------(2) = --------------------(3) The pore wall-fluid interaction potential ( was calculated from Steele Potential28,29 (10-4 Lennard-Jones wall): '( +

= 2!" Δ $% &

)

*

−&

'( , )

-

'( * − &./)0+./ 2 *3--------------------(4)

where , , Δ are the well depth of the potential, molecular diameter and distance between layers of fluids (f) and graphite (w), respectively. For a given pore width (H) as shown in Figure 3, the total potential energy inside the pore is given as23: 567 = + 8 − --------------------(5) where z is the coordinate axis perpendicular to the pore wall. The slit distance (H) is fixed at 3.5 nm to complement the experiments performed with SBA-15 that has pore dimension of 3.5 nm. The objective of the simulations was to find a possible cause for changes in the bubble points observed in the experiments, not necessarily compare the changes calculated by modeling and observed by experiments. Steele Potential can be used rigorously to represent fluid-graphite


Page 11 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

interaction. The values of , , and Δ for graphite were taken to be 28.0 K(kb), 0.340 nm, 0.335 nm respectively, from earlier works on Steele Potential23,28,29. The fluid parameters for methane and decane ( , were chosen based on TraPPE-UA force field developed by Martin and Siepmann30. The values of were calculated using the combining rules

discussed above. = --------------------(6)

= + --------------------(7) Experimental Study Artificial mesoporous material was fabricated to perform experiments to measure the saturation pressures of fluids confined in nanometer scale pores. SBA-15 is one of mesoporous silica materials which has highly ordered hexagonal structure. SBA-15 can be synthesized in the presence of triblock poly(ethylene oxide)–poly(propylene oxide)–poly(ethylene oxide) (PEO– PPO–PEO) copolymers. SBA-15 has high BET surface areas (690 – 1040 m2/g) and small aligned pores (4.6 – 30 nm)31. This material was used to represent the porous structure of shale reservoirs. The saturation pressures were then measured for the bulk fluid (methane and decane mixture with 90:10 mole ratio) and for the same mixture confined in the pores of synthesized SBA-15.

Synthesis procedures of mesoporous materials (SBA-15) Four grams of Pluronic P123 block copolymer (EO20PO70EO20, Aldrich) are dissolved in 120 mL of 2M HCl and 30 mL of deionized water then stirred at room temperature. Once fully dissolved, 9.1 mL of tetraethylorthosilicate (TEOS, Aldrich) is added to the solution. Then, the prepared solution is covered and placed in the oven at 37oC for 20 hours. The temperature of the


Energy & Fuels

oven is raised to 70oC for 16 hours for hydrothermal treatment. Further, the synthesized silica is filtered and washed using a mixture of water and ethanol (1:1 volume ratio). After drying, it is placed into the oven and the temperature is raised to 500 oC for 3 hours with a 1 oC/min ramp rate. The synthesized silica is characterized using TEM and the pore size was determined by the N2 adsorption. The nitrogen isotherm (Figure 6) of the synthesized SBA-15 has a hysteresis loop in the relative pressure range between 0.4 to 0.6. This type IV isotherm is generally shown in mesoporous materials. The Brunauer–Emmett–Teller (BET)32 specific surface area of the synthesized SBA-15 is 752 m2/g, and the average pore size calculated by Barret-Joyner-Halenda (BJH)33 method is 3.5 nm. The 3.5 nm slit width in molecular model described earlier corresponds to average pore size of SBA-15 calculated by BJH method.

18

Quantity adsorbed (mmolg)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 26

16 14 12 10 8 6 4 2 0 0

0.2

0.4

0.6

0.8

1

1.2

Relative pressure (P/P0) Adsorption

Desorption

Figure 6. Nitrogen isotherms of the synthesized SBA-15. A hysteresis loop in the relative pressure range between 0.4 to 0.6 is observed which is typical of type IV isotherm mesoporous materials.


Page 13 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

The pore images of SBA-15 were examined by Transmission Electron Microscopy (TEM). TEM images (Figure 7) show a highly ordered two-dimensional mesoporous structures. Also, the average pore size of SBA-15 in TEM images are less than 5 nm, which makes it a good candidate to represent porous shale reservoirs.

Figure 7. TEM images of the synthesized SBA-15. The average pore size is less than 5 nm.

Measurement of bubble point pressure Bubble point pressure measurements were performed using a customized apparatus that was designed at the University of Utah (Figure 8). A 6.25 inches long stainless steel with 3/8inch outer diameter and 1/32-inche wall thickness was connected to two high-pressure syringe pumps for pumping decane and water, two gas cylinders of methane and nitrogen, and a vacuum pump. This system was placed in an oven. The pressure and volume of the system were recorded using Labview synced to a data acquisition system which is connected to the custom-built apparatus described above. The amount of moles of the decane placed in the cell were calculated using its molecular weight (142.28 g/mol) and its density (0.73 g/ml). Then, moles of methane were calculated for 90:10 decane-methane molar ratio. The pressure of methane was calculated


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 26

using ideal gas law. For this calculation, the volume of the methane was calculated as the system volume subtracted by decane volume. Based on this calculation, a specific volume of decane was added to the system, and the methane was pumped with the specific pressure into the system from the bottom of the system in order to increase the contact area and equilibrium. The temperature of the oven was raised to 100 oF. After the temperature in the oven reaches steady state and no longer fluctuates, the pressure of the system was raised by adding water continuously to the system at a rate of 0.1 mL/min. Once the pressure reached 1200 psi, the pressure of the system was decreased by withdrawing water at a low rate of 0.1 mL/min to make sure that equilibrium is established at each pressure. 1200 psia pressure was chosen as it is adequately above the bubble point and allows capturing enough data when pressure decreases at faster rate with increasing volume, when the system is at pressures above the bubble point pressure. From the extrapolated PV graph, the bubble point pressure was estimated. The variations in bubble points were ± 5% as evidenced by repeated measurements. To conduct experiments with the mesoporous materials, the above described procedure was repeated with SBA-15 packed in the stainless steel tube with filters and filter papers at the both sides.


Page 15 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 8. Schematic of the customized apparatus built at University of Utah for measuring the bubble point pressure of the hydrocarbon fluids.

Discussion and Results The experimental bubble point pressures were estimated from the pressure-volume (PV) plot from the saturation pressure measurement experiment. As the first bubbles of gas emerge, the relative increase or decrease in pressure per unit volumetric change decreases in comparison to when the system was bubble free. The bubble point pressure in conventional thermodynamic measurements was interpreted from slope change of the PV graph34. Figure 9 shows the experimental results of bulk (no porous medium) and the one with the mesoporous materials (column packed with SBA-15). MATLAB was used to fit two parts of the curve and detect the point of intersection. The MATLAB fit had an R2 value higher than 0.98. The bubble point pressure of the methane and decane mixture with 90:10 mole ratio in the absence of porous medium was 374 psia at 100 oF (310 K), with SBA-15 was 242 psia at 100 oF (310 K). Thus, the experimentally measured bubble point pressure with SBA-15 was lower than the bubble point pressure measured for the bulk fluid.


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 26

Figure 9. Experimentally derived PV graph showing different saturation pressures for the bulk fluid and fluid confined in SBA-15. A suppression of 132 psia is observed at 310 K in the bubble point of confined fluid from the original bubble point of 374 psia of bulk fluid.

The GEMC simulations were performed for pure decane for a range of temperatures. At each isothermal simulation of decane (vap) – decane (liq) equilibrium, the density of each phase was monitored. The density for each phase was plotted against temperature to get temperaturedensity plot in Figure 10. The simulations were performed for two cases- one considering pore wall – fluid interaction, the other without the effect of pore wall – fluid interaction. The pore wall – fluid interaction was captured using Steele Potential discussed above. The density plots were extrapolated from near-critical region to estimate the critical temperature of pure decane in bulk and confined state8,30. As system approach critical point, simulations performed on a finite system cannot capture the divergence of the correlation length that characterizes critical points35 which may lead to abnormal values of critical properties (finite size effect35–38). Hence, the critical density and critical temperature were calculated by extrapolation from far-critical region in this paper. The critical temperature of bulk decane was calculated to be 705K compared to the true critical temperature of 618K39 of bulk decane. The difference from true critical temperature


Page 17 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

is attributed to the finite size effect described earlier and is also observed by researchers previously35–38. A suppression of 125 K in critical temperature was calculated for the confined decane which is close to the observations made by other researchers20,40. The difference between the two values (bulk and confined) is more relevant than the actual values.

Figure 10: Temperature-density plot for pure decane. The densities from near-critical region were extrapolated with good R2 values to estimate the critical temperature of the decane in confined and bulk state. Suppression of approximately 125 K is calculated in critical temperature of decane confined in nanometer scale pores, which is roughly similar to the observations made by other researchers20,41. vap_wall and liq_wall in the legend refer to the use of Steele pore wall potential to account for pore wall – fluid interaction. vap_no_wall and liq_no_wall represent the bulk state when pore wall – fluid interactions were not included.

In order to calculate the corresponding change in critical pressure of the confined fluid, Antoine42 equation was used because of its applicability for wide range of temperatures43


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

including near critical region for decane. The Antoine equation in following form was used to calculate the suppression in the critical pressure of confined decane: ln= = > −

?

@0A

--------------------(8)

where P and T are in kPa and OC; A, B and C are constants and their values are based primarily on data presented by Poling44. The pressure-temperature (PT) plot for pure decane is shown in Figure 11. It should be noted that the bulk values of critical properties estimated from the GEMC model were used in this plot. A suppression in vapor pressure is calculated for the confined decane. Such a suppression happened due to change in the critical properties of the pure decane confined in nanometer range pores. This is a clear indication of changes in the thermodynamic properties of fluids confined in the nanopores. This changes the phase behaviour of free confined fluid compared to the same fluid in bulk state. A potential reason for change in fluid thermodynamic properties and thus, phase behaviour of confined fluid is pore wall – fluid interaction that are dominant at nanometer scale. These surfaces forces cannot be neglected at the nanometer scale as the surface area per unit volume is very high at such small scale.


Page 18 of 26

Page 19 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

Figure 11: The PT plot for pure decane. A suppression in vapor pressure is observed for decane confined in the slit pore made up of graphite.

GEMC simulations based on workflow discussed earlier were again performed for the binary mixture of decane – methane (90-10% molar ratio) to complement the findings from the experiments. The motivation of simulations was to understand the dynamics at pore scale and contribute in the fundamental understanding of phase behavior of fluid in confined spaces of shale rocks. The simulations helped in capturing a potential reason for suppression in the bubble point of the binary mixture of decane – methane that was observed from the experiments. However, this understanding can be utilized in the study of phase equilibrium of other hydrocarbon mixtures confined in nanometer scale pores. The GEMC simulations were performed at different temperatures to simulate the phase equilibrium between vapor and liquid of decane-methane mixture. The densities at each isothermal step was monitored and plotted to


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 26

get temperature-density plot for the binary mixture (Figure 12). The densities from near-critical region were

Figure 12: Temperature-density plot for mixture of decane-methane (90-10% molar ratio). A suppression of about 85 K in critical temperature is calculated for binary mixture of decane – methane in 90-10% molar ratio. liq_wall and vap_wall in the legend refer to the use of Steele pore wall potential to account for pore wall – fluid interaction. liq_no_wall and vap_no_wall represent the bulk state when pore wall – fluid interactions were not included. A suppression in the critical properties of the confined fluids is attributed to the pore wall – fluid interaction that becomes dominant at the nanometer scale.

extrapolated with good R2 values to estimate the critical temperature of the mixture. A suppression of about 85 K in the critical temperature of the binary mixture is calculated for


Page 21 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

confined mixture compared to the bulk mixture. When methane was used in the decane-methane mixture, the suppression in critical temperature of confined mixture wasn’t as large as it was for pure decane in confinement. It is probably because the pore wall has less effect on gas phase. This is also seen by other authors that the suppression in critical temperature is larger for longer chain hydrocarbons20,40. The critical density is also suppressed by 0.55 g/mL in the confined mixture. This change in the fluid thermodynamic properties leads to the corresponding change in phase behaviour of the confined fluids. By performing the molecular modeling using GEMC simulations, it was seen that the fluids behave like a bulk fluid if Steele pore wall potential was not used in simulating the phase behaviour. This indicates that the pore wall – fluid interaction is one of reasons for the suppression in the critical properties and vapor pressure of the confined fluids. The true critical properties of bulk decane and confined critical properties of decane from GEMC were used to calculate phase envelop in PT plot (Figure 13) using WinPropTM commercial simulator (from computer modelling group) for the binary mixture of decane and methane (90-10 molar ratio). The critical properties of methane were not changed in the confined state as it only constitutes 10% of the binary mixture. The PT plot shows a suppression of 190 psia comparable to 132 psia observed in the experiments at around 100 oF. However, the difference could be because the surface forces are different in simulations and experiments. The objective of the simulations was to find a possible explanation for suppression seen in the experiments.


Energy & Fuels

600

500

400 Pressure (psia)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 26

300

200

100

0 -100

0

100

200

300 400 Temperature (°F)

Decane-methane_bulk

500

600

700

Decane-methane_confined

Figure 13: Pressure-Temperature plot calculated by simulations showing the suppression in the bubble point of confined mixture (90-10 decane-methane binary mixture).

Conclusions An artificial mesoporos material (SBA-15) was synthesized to measure the change in the bubble points of hydrocarbon fluids in the confined state. The bubble point pressures in the bulk were found to be higher than the bubble points of same fluids confined in the nanometer scale pores. A suppression of 132 psia in the bubble point decane-methane (90-10% molar ratio) was observed for the mixture confined in the 3.5 nm pores of SBA-15. GEMC simulations were performed to diagnose this suppression of bubble points for confined fluids. It was found from the GEMC simulations that the pore wall – fluid interactions at the nanometer scale are dominant and cause suppression in the critical properties of fluids


Page 23 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

confined in nanometer scale pores. The estimation of correct fluid thermodynamic properties including pressure-volume-temperature (PVT) properties is essential to make better prediction of reserves and the rates of recovery of these reserves from shale plays. The conventional PVT simulators don’t account for pore wall – fluid interaction and fail to capture the suppression in critical properties of the confined fluids. The conventional simulators should be updated with models like the one proposed here in order to capture the dynamics at the pore scale.

References (1)

Pathak, M.; Deo, M.; Craig, J.; Levey, R. In Proceedings of the 2nd Unconventional Resources Technology Conference; American Association of Petroleum Geologists: Tulsa, OK, USA, 2014.

(2)

Swindell, G. S. In SPE Annual Technical Conference and Exhibition ; San Antonio, Texas, USA, 2012.

(3)

Sheng, J. J. J. Nat. Gas Sci. Eng. 2015, 22, 252–259.

(4)

Pathak, M.; Deo, M. D.; Panja, P.; Levey, R. A. In SPE/CSUR Unconventional Resources Conference; Society of Petroleum Engineers, 2015.

(5)

Pathak, M.; Pawar, G.; Huang, H.; Deo, M. D. In Carbon Management Technology Conference; Carbon Management Technology Conference, 2015.

(6)

Nagarajan, N. R.; Honarpour, M. M.; Arasteh, F. In Unconventional Resources Technology Conference, Denver, Colorado, 12-14 August 2013; Society of Exploration Geophysicists, American Association of Petroleum Geologists, Society of Petroleum Engineers, 2013; pp 1905–1910.


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(7)

Pitakbunkate, T.; Balbuena, P. B.; Texas, a; Moridis, G. J.; National, L. B. 2014, No. 2004, 1–16.

(8)

Singh, S. K.; Sinha, A.; Deo, G.; Singh, J. K. J. Phys. Chem. C 2009, 113 (17), 7170–7180.

(9)

Devegowda, D.; Sapmanee, K.; Civan, F.; Sigal, R. SPE Annu. Tech. Conf. Exhib. 2012, 1–22.

(10)

Haider, B. A. Impact of Capillary Pressure and Critical Properties Shift Due To Confinement on Hydrocarbon Production, Stanford University, 2015.

(11)

Sheng, L.; Hadi, N.; Lutkenhaus, J. L. AIChE J. 2016, 0 (0).

(12)

Dong, X.; Liu, H.; Hou, J.; Wu, K.; Chen, Z. Ind. Eng. Chem. Res. 2016, 55 (3), 798–811.

(13)

Wang, L.; Parsa, E.; Gao, Y.; Ok, J. T.; Neeves, K.; Yin, X.; Ozkan, E. In SPE Western North American and Rocky Mountain Joint Regional Meeting; Denver, Colorado, USA, 2014.

(14)

Ma, Y.; Chen, J.-H.; Jamili, A. In Unconventional Resources Technology Conference (URTeC); 2016.

(15)

Firincioglu, T. Bubble point suppression in unconventional liquids rich reservoirs and its impact on oil production, Colorado School of Mines.

(16)

Nagy, S.; Siemek, J. Arch. Min. Sci. 2014, 59 (4), 1005–1022.

(17)

Gelb, L. D.; Gubbins, K. E.; Radhakrishnan, R.; Sliwinska-Bartkowiak, M. Reports Prog. Phys. 1999, 62, 1573–1659.

(18)

Jin, Z.; Firoozabadi, A. Soc. Pet. Eng. J. 2016, 190.

(19)

Barsotti, E.; Tan, S. P.; Saraji, S.; Piri, M.; Chen, J.-H. Fuel 2016, 184, 344–361.

(20)

Xiong, X.; Devegowda, D.; Civan, F.; Sigal, R. F. In Unconventional Resources Technology Conference (URTeC) ; Society of Petroleum Engineers (SPE) , 2013.

(21)

Panagiotopoulos, A. Z. Mol. Phys. 1987, 62 (3), 701–719.


Page 24 of 26

Page 25 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

(22)

Panagiotopoulos, A. Z. Mol. Simul. 1992, 9 (1), 1–23.

(23)

Sliwinska-Bartkowiak, M.; Gras, J.; Sikorski, R.; Radhakrishnan, R.; Gelb, L.; Gubbins, K. E. Langmuir 1999, 15 (18), 6060–6069.

(24)

Lastoskie, C.; Gubbins, K. E.; Quirke, N. Langmuir 1993, 9 (10), 2693–2702.

(25)

Jones, J. E. Proc. R. Soc. A Math. Phys. Eng. Sci. 1924, 106 (738), 463–477.

(26)

Berthelot D. C. R. Hebd. Seanc. Acad. Sci., Paris 1898, 126, 1703.

(27)

Lorentz, H. A. Ann. Phys. 1881, 248 (1), 127–136.

(28)

Steele, W. A. Surf. Sci. 1973, 36 (1), 317–352.

(29)

Steele, W. A. The interaction of gases with solid surfaces; Pergamon Press: Oxford, U.K, 1974.

(30)

Martin, M. G.; Siepmann, J. I. J. Phys. Chem. B 1998, 1998, 102, 2569–2577.

(31)

Zhao, D.; Huo, Q.; Feng, J.; Chmelka, B. F.; Stucky, G. D. J. Am. Chem. Soc. 1998, 120 (24), 6024–6036.

(32)

Brunauer, S.; Emmett, P. H.; Teller, E. 1936, 407 (1), 309–319.

(33)

Barrett, P. E.; Joyner, L. G.; Halenda, P. P. J. Am. Chem. Soc., 1951, 73 (1), 373–380.

(34)

McCain, W. D. The properties of petroleum fluids; Petroleum Pub. Co., 1973.

(35)

Panagiotopoulos, A. Z. In Observation, Prediction and Simulation of Phase Transitions in Complex Fluids; Baus, M., Rull, L. R., Ryckaert, J. P., Eds.; Kluwer Academic Publishers: Dordrecht, Netherlands, 1995; Vol. 460, pp 463–501.

(36)

Caccamo, C.; Costa, D.; Fucile, A. J. Chem. Phys. 1997, 106 (1), 255.

(37)

Dinpajooh, M.; Bai, P.; Allan, D. A.; Siepmann, J. I. J. Chem. Phys. 2015, 143 (11), 114113.

(38)

Khedr, M. B.; Al-Busaidy, M. S.; Osman, S. M. J. Phys. Condens. Matter 2005, 17 (28),


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

4411–4422. (39)

NIST Webbook.

(40)

Singh, S. K.; Singh, J. K. Fluid Phase Equilib. 2011, 300 (1), 182–187.

(41)

Singh, S. K.; Saha, A. K.; Singh, J. K. J. Phys. Chem. B 2010, 114 (12), 4283–4292.

(42)

Antoine, C. Compt. Rend. 1888, 107, 681–684.

(43)

Thomson, G. W. Chem. Rev. 1946, 38 (1), 1–39.

(44)

Poling, B. E.; Prausnitz, J. M.; O’Connell, J. P. The Properties of Gases and Liquids, 5th ed.; McGraw-Hill Education.


Page 26 of 26

Experimental and Molecular Modeling Study of Bubble Points of

Recommend Documents