Systematic Study of Heavy Oil Emulsion Properties ... - ACS Publications

Subscriber access provided by GAZI UNIV

Article

A Systematic Study of Heavy Oil Emulsion Properties Optimized with New Chemical Formulation Approach: Particle Size Distribution Nabijan Nizamidin, Upali P. Weerasooriya, and Gary A. Pope Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.5b01818 • Publication Date (Web): 26 Oct 2015 Downloaded from http://pubs.acs.org on October 30, 2015

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 57

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

A Systematic Study of Heavy Oil Emulsion Properties Optimized with New Chemical Formulation Approach: Particle Size Distribution Nabijan Nizamidin*, Upali P. Weerasooriya, Gary A. Pope The University of Texas at Austin, Department of Petroleum and Geosystems Engineering, 200 E. Dean Keeton St. Stop C0300, Austin TX 78712-1585

ABSTRACT: The purpose of this research was to create very polydisperse concentrated heavy oil-in-water emulsions by optimizing the co-solvents, surfactants, alkali, and electrolytes in the chemical formulation with respect to the droplet size distribution. Novel co-solvents that have shown superior performance in chemical formulations used for enhanced oil recovery have been tested. Droplet size distributions that resulted in a lower emulsion viscosity were found to have a higher mean droplet diameter d32 and a bimodal droplet size distribution with diameter ratio d32,L/d32,S>6 and volume fraction ⁄( + ) = 0.2-0.3, where is the volume fraction of the dispersed phase and the subscripts L and S correspond to the larger and smaller peaks in the bimodal distribution. We report the effects of various chemical formulations on the droplet size distribution of heavy oil-in-water emulsions with a particular emphasis on the d32 and for the first time, the maximum packing fraction of oil droplets. A novel one-step preparation procedure is proposed to prepare concentrated multimodal oil-in-water emulsions with a new chemical formulation approach. We were able to formulate stable oil-in-water emulsions with as high as 0.95, which is ~0.30 higher than the theoretical = 0.64 of random close packed 1

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

Page 2 of 57

monodisperse spheres. We observed the optimal particle size distribution of concentrated heavy oil-in-water emulsions prepared with co-solvents for maximum packing at approximately 75% of the Na+ concentration necessary to reach the oil-in-water to water-in-oil inversion point for anionic surfactants. An important application of this study is the transport of heavy oils in pipelines.

1. Introduction Heavy crude oils often need to be transported over long distances from the production sites to the refineries. However, the heavy crude oils are too viscous (103-107 mPa·s) for pipeline transport. A widely studied method to reduce the viscosity is to create an oil-in-water (O/W) emulsion. Such emulsions can be transported in pipelines if their viscosity is less than about 350 mPa·s. The ultimate objective of preparing heavy oil-in-water emulsions for pipeline transport is to reduce the volume of water in the emulsion as much as possible while still maintaining low emulsion viscosity and good emulsion stability. Many heavy oils and bitumen contain significant quantities of naphthenic acid components.1 Such oils can be reacted with alkali with a sufficiently high pH to generate soap and reduce interfacial tension. The quantity and type of soap depends on the composition of the heavy crude oil and the pH of the alkaline solution.1 Alkali is used for this purpose in enhanced oil recovery (EOR) methods. The most common EOR process is to use alkali with a surfactant and a watersoluble polymer (ASP flooding).

More recently, Fortenberry et al.

2

found that heavy oil

emulsions had a lower viscosity when co-solvents were used with the alkali instead of surfactants and this was advantageous with respect to EOR. Fortenberry et al. used several novel cosolvents such as phenol-16EO. These and other novel ethoxylated co-solvents used in this study 2


Page 3 of 57

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

to optimize heavy oil emulsions have shown superior performance compared to conventional alcohol co-solvents.2,3 One of the major advantages of these ethoxylated co-solvents is their ability to be easily tailored to the crude oil and brine by modifying the ethylene oxide (EO) number, which affects their hydrophilic-lipophilic balance. Scaling equations have been proposed that model the rheological properties of emulsions with the hard sphere packing assumption. Pal proposed a model based on the Krieger and Dougherty equation that fits a variety of experimental emulsion viscosity data available in the literature for a dispersed phase concentration () in the range of 0-60%:4

/

= 1 −

.

,

(1)

where is the ratio of emulsion viscosity to continuous-phase viscosity ( ), K is the ratio of the dispersed-phase viscosity ( ! ) to the continuous phase viscosity, φ is the dispersed-phase volume fraction, and is the maximum packing volume fraction of the dispersed-phase without deformation of the droplets. The is a fitting parameter in Eq. (1) and the range of values obtained from the model for the set of experimental data Pal analyzed is approximately 0.55-0.74. One limitation of Eq. (1) is goes to infinity when φ approaches , which is only accurate for hard sphere suspensions. However, emulsion dispersed-phase droplets are soft unlike hard spheres and have been shown to deform into a polyhedral shape as ≫ with finite viscosity.5–8

An improved version9 of Eq. (1) as well as a viscosity equation for

multimodal spherical suspensions as a product of individual unimodal fractions10,11 can be found in the literature, but with the same limitations. Thus, a different model is needed to accurately describe the rheological properties of emulsions when > . 3


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

Seth et al.

Page 4 of 57

developed and tested a model for emulsions with > and ranged from 0.7-

12

0.9. $

%$= 1+&'

() *

(% + ,

//

. ∗

,

(2)

where 0, is the shear stress, 01 is the yield stress, & is the numerical coefficient, 2) is the shear rate, is the continuous-phase viscosity, 21 is the yield strain, and 3 ∗ is the contact modulus. Foudazi et al. rewrote Eq. (2) for a polydisperse emulsion:13 0 = 01 + 185

() * !6+ // Г

,

(3)

where 0 is the shear stress, 01 is the yield stress, 2) is the shear rate, is the continuous-phase viscosity, 8 is the Sauter mean droplet diameter, 5 is the shear modulus, and Г is the interfacial tension. Eq. (1) and Eq. (3) imply that when , emulsion viscosity is shear-thinning and is dependent on 8 5,13–18 for unimodal emulsions. This observation is consistent with experimental data in the literature. The change in the regime for unimodal emulsion rheological properties modeled by Eq. (1) and Eq. (3) occurs at of ~0.55-0.65 and is also when the emulsions start exhibiting highly viscous behavior as well as non-Newtonian properties such as yield stress and shear-thinning behavior.5,14 The observed range of appears to be close to for monodisperse hard sphere suspensions. is theoretically defined as the glass transition point, 9 = 0.58, the random loose packing point, :; = 0.60, or the random close packing point,

;

= 0.64.19 The 9 is when a hard sphere is only able to relax within a cage formed by 4


Page 5 of 57

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

its nearest neighbors, thus limiting diffusion and flow20, :; is the packing volume fraction of uniform hard spheres packed by hand in a random manner, and

;

is the highest packing

volume fraction of uniform hard spheres packed by the vibration method in a random manner. Since the high viscosity, yield stress, and shear-thinning behavior of concentrated emulsions are caused by increasing particle to particle interactions as the result of the deformation of the droplets when > , the droplet size distribution of the emulsions can be tuned to mitigate the increase in viscosity of concentrated emulsions in two ways (1) increase the of concentrated emulsions to achieve because cannot be increased further, increase 8 as high as possible while maintaining emulsion stability. The purpose of this research was to create multimodal concentrated emulsions by optimizing the co-solvents, surfactants, alkali, and electrolytes in the chemical formulation with respect to the droplet size distribution.

A two-step method21,22 of preparing two concentrated unimodal

emulsions, one with smaller mean droplet diameter and one with larger droplet diameter, and mixing them at the optimum ratio to obtain concentrated bimodal emulsions of lower viscosity has been proposed. Our procedure offers the advantage of one-step preparation of concentrated multi-modal emulsions with optimized chemical formulations, which requires only one mixing tank, whereas the two-step method requires three mixing tanks. The of the emulsions was calculated using the Groot and Farr method

23

to compare and contrast the effectiveness of the

various chemical formulations used to prepare concentrated emulsions from a purely droplet size distribution point of view.

Section 2 describes the chemicals used, concentrated emulsion

preparation procedure, the equipment and procedures utilized to measure and analyze particle 5


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

size distributions, as well as the method used to measure oil and emulsion rheological properties. Section 3 goes into the theories behind how the particle size distribution fitting parameters such as the mean diameter, standard deviation, and polydispersity affect the of concentrated emulsions, literature review of experimental data of concentrated emulsions, and relationship between phase behavior studies of oil/surfactant/water systems and particle size distributions. Section 4 discusses the results of the experiments based on the effects the chemical formulations have on the d32 and in terms of the heavy oils, co-solvents, alkali, electrolytes, and synthetic surfactants. Section 5 concludes the paper with the major findings of the study.

2. Experimental 2.1 Materials 2.1.1 Crude oils. Four heavy crude oils were used in this study. The properties of the oils are illustrated in Table 1. The crude oil samples are identified with letters A, B, C, and D. The crude oils were picked to represent a broad range of viscosities from a variety of geological settings. The viscosities of the heavy crude oil samples as a function of temperature are shown in Fig. 1. The viscosities of the heavy crude oils were measured using TA Instruments Advanced Rheometric Expansion System (ARES) LS-1. The cone & plate geometry was used to measure the heavy crude oil samples because of the extremely high viscosity of the oils at low temperature and the couette geometry was used for high temperature. The modified Walther Equation 24 was found to be an acceptable fit for the data for all four heavy crude oil samples. 2.1.2 Co-solvents and surfactants. A variety of co-solvents were tested. The co-solvents were ethoxylated isobutyl alcohol (IBA-nEO), triethylene glycol monobutyl ether (TEGBE), 6


Page 6 of 57

Page 7 of 57

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

alkoxylated phenol (Ph-mPO-nEO), and ethoxylated diisopropylamine (DIPA-nEO). The chemical structures are shown in Fig. 2. The chemicals were obtained from Harcros Chemicals, Taminco, Aldrich Chemicals, and Huntsman Corporation. The most commonly used non-ionic surfactant to prepare heavy crude oil emulsions, nonylphenol-ethoxylate (NPE), was used to prepare heavy O/W emulsions as a reference point for our emulsion samples. NPE was obtained from Harcros Chemicals. 2.1.3 Salts and alkali. Aqueous solutions of sodium hydroxide (NaOH), sodium carbonate (Na2CO3), and sodium metaborate (NaBO2) were used to saponify the naphthenic acids and precursors in the heavy crude oils. Sodium chloride (NaCl) was used to adjust the salinity of the water. These chemicals were obtained from Fisher Scientific.

2.2 Preparation procedure of heavy crude oil in water emulsions All emulsion samples were created using the following procedure unless otherwise noted. 1. The aqueous solution consisting of deionized water (DI), NaCl, co-solvents, alkali and surfactants is mixed at room temperature. All chemicals are measured and reported as a weight percent of the aqueous solution (w/w). 2. A mixture of the aqueous solution and a heavy crude oil is poured into a volumetric vial to prepare emulsions with different concentrations of oil (i.e., 20%, 40%, 60%, 80%, 85%, and 90%). The concentration of the crude oil in an emulsion is reported as a volume percent of the total volume of an emulsion at room temperature (v/v). 3. The mixture is sealed and placed in a 95-100oC oven. No light end losses were observed.

7


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 8 of 57

4. After heating to the oven temperature, the sample is vigorously hand-shaken for 10 seconds every 30 minutes for several hours. 5. The sample is taken out of the oven and cooled down to a room temperature of 23oC ±2oC overnight before experiments are conducted. The compositions of the emulsions are listed in the figure descriptions.

The droplet size

distributions of all concentrated emulsions were best described by lognormal distributions. However, unlike the data reported in the literature, most of our concentrated emulsions formed polydisperse distributions that can best be described with bimodal or trimodal lognormal droplet size distributions instead of narrow unimodal droplet size distributions. A likely explanation for the observed difference with the droplet size distribution of our emulsions compared to those described in the literature is the different method of concentrated emulsion preparation used in this study and the way we optimized the chemical formulations. The HIPR (high internal phase ratio) method25 is a commonly used method of preparing concentrated emulsions ( φ> 0.7). The HIPR method uses low shear (10-100 s-1) mixing for less than 5 minutes to create emulsions with a very narrow droplet size distribution at less than 60oC

25

. We prepared the emulsions with

brief, high shear (100-1,000 s-1) vigorous hand mixing at a higher temperature of 95-100oC.

2.3 Particle size distribution of emulsion droplets Two methods were used to ascertain the particle size and shape of the emulsion samples. 2.3.1 Zeiss Axiovert Fluorescent Light Microscope. The fluorescent light microscope was used to take photomicrographs of the emulsion samples. The samples were doped with fluorescent dyes. Very low concentration of dyes (10-20 ppm) were used to minimize the effects 8


Page 9 of 57

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

of dyes on emulsion properties. The photomicrographs provide a visual evidence of the range of the emulsion droplet size for each samples. 2.3.2 Malvern Mastersizer 3000 Static Light Scattering. The particle size distribution of the emulsion samples were measured using the Malvern Mastersizer 3000 static light scattering equipment. After mixing the samples thoroughly, the emulsion samples were diluted to the necessary concentration utilizing 0.2% NaOH and 0.1% NaCl solution. The alkali is necessary to maintain the pH and to keep the naphthenic acids deprotonated. All measurements were conducted at ambient conditions. The Mie theory was used to calculate the oil droplet size distribution based on how light is scattered by the spherical particles.26 The required refractive indices of the oils and aqueous solutions for the Mie theory were measured using a refractometer. Since such low concentration of diluted emulsions (0.01-0.05% volume) are required to accurately measure the particle size distribution, the Mastersizer 3000 accessory with sample volume of 600 ml was used. The large volume accessory necessitates a larger quantity of concentrated emulsion samples to be diluted, resulting in a more accurate representation of the particle size distribution of the entire samples in some cases of sample inhomogeneity. To calculate the of emulsions, the droplet size distributions must be accurately fitted to a distribution model.

Emulsion droplet size distributions can be modeled accurately by a

lognormal distribution for unimodal distributions and a combination of lognormal distributions for multimodal distributions: ?@(8; , B) = ∏D/ ?D !E

/

F √ H

+ ⁄ E + F

I (:DJ!KF )

,

9


(4)

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 57

where fv is the probability density function of volume, n is the number of peaks in the distribution, d is the diameter of droplets, B is the natural logarithm standard deviation of the droplet diameters, is the natural logarithm of the mean of the droplet diameters, and f is the volume fraction of a peak in the distribution over the entire droplet volume D ⁄∑D/ D . The droplet size distributions are fitted to the lognormal probability density function of volume with the least square method weighted toward larger diameters since the larger droplets contribute significantly more to the volume fraction of the droplets. A closed analytical equation that calculates the of spheres modeled by a unimodal lognormal distribution has recently been derived by Brouwers.27 However, the equation cannot handle a multimodal lognormal distribution and the authors have no knowledge of an analytical equation that is capable of calculating the random close packing fraction of spheres with a bimodal lognormal distribution. The only method of calculating the of polydisperse spheres is to directly simulate the packing of spheres using numerical simulations. Farr and Groot

23

used a

fast 1D algorithm for accurately estimating the of polydisperse hard spheres and compared it to the more computationally expensive 3D algorithms28,29 with favorable results. The Farr and Groot algorithm was used to estimate the of concentrated emulsions in this paper from the lognormal distribution model parameters. Note that we estimated from the droplet size distribution since the

;

;

of the

can be estimated relatively easily for polydisperse

emulsions. While we do not claim that the

;

best represents the over 9 or :; ,

Spangenberg et al.29 showed that =~0.93±0.005

;

in Eq. (1) gave the best fit of

experimental data from the literature. Note that for simplicity, we assumed = ; .

10


Page 11 of 57

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

2.4 Rheological Measurements The viscosities of the emulsions were measured using TA Instruments Advanced Rheometric Expansion System (ARES) LS-1 at ambient conditions. The parallel geometry was used with cross-hatched plates to eliminate slip at the wall. The plate dimensions are diameter d=50 mm and the gap between plates h=1 mm unless otherwise mentioned. The samples were measured at varying gap widths to verify the elimination of slip at the wall. We observed no slip in any of the measurements when cross-hatched plates with a gap between 0.5 to 3 mm were used.

3. Theory The theory section is divided into three main parts: (1) Parameters that influence the of spherical droplets; (2) Literature review of experimental results of concentrated bimodal emulsions of varying dL/dS and M ⁄(M + ) on the emulsion rheology; and (3) Phase behavior study of oil/surfactant/brine mixtures and particle size distribution. 3.1 Parameters that influence the NO of spherical droplets. The of monodisperse hard spheres depends on the method of packing, random (:; =0.60 or ; =0.64) or ordered (~0.74). The = ~0.58-0.64 is exactly the range of observed with emulsions where the emulsion viscosity starts to increase dramatically for narrow unimodal emulsions5,14, suggesting emulsion rheology is best described by random packing of droplets29. Two parameters affect the packing fraction of binary solid spheres where two monodisperse spheres of varying sizes are mixed in varying ratios. The variables are the sphere diameter ratio (dL/dS), where d is the sphere diameter and subscripts S and L correspond to the small and large 11


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 12 of 57

component of the binary mixture and the volume fraction of the small monodisperse spheres with respect to the total monodisperse sphere volume (M ⁄(M + )). Furnas30 introduced the concept of saturated, non-interacting binary monodisperse spheres, which is defined as the smaller spheres filling the void created by the larger spheres without affecting the packing of the larger spheres expressed as dL/dS=∞ and M ⁄(M + )=0.36. Brouwers27 summarized Furnas’ observations and concluded that the of combined mixtures of saturated non-interacting monodisperse spheres can be described by Eq. (5): DMPQ = 1 − (1 − ; )D ,

(5)

where n is the number of non-interacting monodisperse size groups, DMPQ is the of saturated non-interacting spheres of n monodisperse size groups, and

;

is the random close packing

fraction of monodisperse spheres (~0.64). For a binary mixture of saturated non-interacting monodisperse spheres where n=2 and 3, the theoretical value of MPQ = ~0.87 and MPQ = ~0.95. Experiments have revealed that non-interacting binary monodisperse spheres are approximated when dL/dS is greater than 7-10 since realistic experimental emulsion samples with dL/dS=∞ are not possible.30,31 Recent numerical simulation studies by Hopkins et al.28 show that for dL/dS=310, the maximum packing density is observed at M ⁄(M + )=~0.20-0.25. The of theoretical unimodal lognormal distributions estimated using the Farr and Groot model as a function of the standard deviation is depicted in Fig. S1(Supporting Information) to show how the standard deviation affects the . We have observed values for the standard deviations in the range of 0.4 to 1 for polydisperse emulsions. The benefit of emulsions with higher standard deviation of the particle diameter becomes apparent when Eq. (5) is modified to include the effect of standard deviation on the of emulsions: 12


Page 13 of 57

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

DMPQ = 1 − ∏D/(1 − ,D ) ,

(6)

where n is the number of non-interacting unimodal groups, DMPQ is the of saturated noninteracting spheres of n unimodal size groups, and ,D is the of individual unimodal groups estimated based on the standard deviation. Table 2 shows the DMPQ as a function of B , BR , and B for theoretical non-interacting saturated spheres. Very broad bimodal emulsions can have even higher DMPQ than narrow trimodal emulsions. 3.2 Literature review of experimental data of concentrated bimodal emulsions. The rheological benefits of increasing the of concentrated emulsions by preparing mixtures of two narrow unimodal emulsions of varying mean diameter have been extensively explored.5,7,13,16,17,32 Narrow unimodal concentrated emulsions of varying 8 were prepared using the “High Internal Phase Ratio” (HIPR) method developed by Chirinos et al.25 and mixtures of dL/dS = 1.5-15 with M ⁄(M + ) = 0-1 were prepared and tested in the literature.5,16,17 A summary of the literature data is displayed in Table 3. Fig. 3 shows the estimated and theoretical optimum M ⁄(M + ) calculated using the Farr and Groot model assuming standard deviations of B =B =0.3 for the dL/dS in Table 3. Table 3 shows that a concentrated bimodal emulsion with dL/dS = 1.5 did not lower the emulsion viscosity compared to the unimodal emulsion with dL.17 A value of dL/dS = 3.6 showed a 15 fold reduction in emulsion viscosity at the optimum M ⁄(M + ) = 0.36 compared to the unimodal emulsion of the large component at 0=0.9 Pa, but no minimum in emulsion viscosity was observed at 0=30 Pa.16 For =0.7, as the dL/dS increased from 5 to 10, the emulsion viscosity decreased from 100 mPa·s to 70 mPa·s at the optimum M ⁄(M + ), which can be explained by the increase from ~0.8 to ~0.85 compared to the unimodal coarse emulsion viscosity of 1,000 mPa·s.5 When 13


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 57

the dL/dS increased from 10 to 15, the emulsion viscosity decreased from 70 mPa·s to 30 mPa·s while the only increased from ~0.85 to ~0.86, thus the decrease in the viscosity can be attributed mainly to the dL increase to 30 µm from 20 µm.5 The effect of emulsion mean diameter was observed as the unimodal emulsion viscosity decreased from ~4,000 mPa·s to 550 mPa·s when the mean diameter increased from 2 µm to 30 µm for =0.7. Foudazi et al.13 also observed similar effects of dL/dS and M ⁄(M + ) on emulsion viscosity for W/O emulsions of =0.85. The experimental data in Table 3 show that when > for bimodal emulsions, larger dL/dS, larger 8 , and optimum M ⁄(M + ) significantly lowered the emulsion viscosity of bimodal emulsions compared to unimodal emulsions of the same . Shewan et al.33 used Eq. (1) to calculate the hard sphere suspension viscosity for a range of materials, fluids, and from the data available in the literature as well as their own experiments. The was estimated using the Farr & Groot model instead of being a fitting parameter from the particle size distribution of both monodisperse and bimodal suspensions of 8S = 0.3-250 µm, dL/dS = 0.0027, and M ⁄(M + ) = 0-1 and the data collapsed into one curve when @T. ⁄ was plotted.33 Calculating the of emulsions from the droplet size distribution and incorporating it into Eq. (1) appeared to accurately describe the effects of the droplet size distribution on the hard sphere packing viscosity when < and in the presence of purely repulsive inter-droplet interactions. 3.3 Phase behavior study of oil/surfactant/brine mixtures and particle size distribution. The composition of the chemical formulations used to prepare O/W emulsions influences the particle size distribution as well as the d32. Phase behavior studies of oil/surfactant/brine mixtures are often performed to test surfactants in chemical EOR research to identify ultra-low interfacial 14


Page 15 of 57

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

tension at the desired electrolyte concentration and temperature.3,34–36 Baldauf et al.37 conducted phase behavior experiments to study the relationship between microemulsions and concentrated macroemulsions. The optimum condition of a Winsor type III bicontinuous microemulsion is defined as the point where the interfacial tension is equal at the oil/microemulsion interface and the water/microemulsion interface (Fig. 4). Observations show that at this same condition, the volumes of water and oil solubilized in the microemulsion are also equal and the coalescence rate is a maximum.38 A minimum in the average droplet diameter was found near but not at the optimum condition.39–41 A high coalescence rate appears to be a key to the formation of multimodal droplet distributions.40 The time needed for crude oil/surfactants/brine mixtures to reach thermodynamic equilibrium decreases dramatically when co-solvent is added to the mixture. Co-solvents change the interfacial properties of the micelles by disrupting the ordered packing of surfactant molecules, resulting in a more fluid interface and lower interfacial viscosity.3 Thus, to form concentrated emulsions with high , emulsions should be prepared with optimized chemical formulations containing a co-solvent near the optimum condition. Pérez et al.39 observed that a bimodal droplet size distribution formed a small distance from the optimum condition when alcohol was used with a surfactant. dos Santos et al.42 also recently reported that addition of light/medium alcohol co-solvents to surfactants resulted in a bimodal distribution of emulsion droplets. We expanded upon the observations of both Pérez et al. and dos Santos et al. and quantified the effects of co-solvents, surfactants, electrolyte, and crude oil on the particle size distribution, represented by , with a simplified one-step mixing process instead of mixing multiple unimodal emulsions of varying d32 and combining them at an optimal 15


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 57

ratio to generate bimodal/trimodal emulsions. The characterization of multimodal concentrated emulsions in terms of when prepared with the one-step chemical formulation approach similar to chemical EOR phase behavior scans has not been published to our knowledge. 4. Results and discussions Photomicrographs of two 80% oil D emulsions are shown in Fig. 5. The only difference between the two emulsions is the NaCl concentration of 0% and 1%. The photomicrograph of the emulsion sample with 0% NaCl (Fig. 5a), showed a unimodal droplet size distribution and spherical droplet shapes with the bigger droplets showing a slight deformation at static conditions, reflected by low φm=0.68. The photomicrograph of the emulsion sample with 1% NaCl (Fig. 5b), showed a bimodal droplet size distribution with no observable deformation of the droplets, reflected by higher φm=0.78. Fig. 5b shows that the smaller droplets surround the bigger droplets.

We have observed that emulsions samples with φm>0.75 do not cause

significant deformation of the oil droplets at static conditions when φ=0.8. The conditions we used to prepare the emulsions were varied to observe the effects of mixing temperature, speed, and duration on the particle size distributions. The modified capillary number, Ca, has been shown to control the mechanism of droplet breakup and droplet diameter for concentrated emulsions.43 VW = X Y:MZ[D 2) \/Г

(7)

where µemulsion is the viscosity of the emulsion at the mixing shear rate, 2) is the shear rate of mixing, R is the radius of droplets, and Г is the interfacial tension. The Grace Curve and concept of critical capillary number (Cacritical) vs. λ=µd/µemulsion has been used to accurately predict if 16


Page 17 of 57

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

droplets in concentrated emulsion will breakup into smaller droplets or not.43,44

The key

limitation of the Grace Curve and Cacritical is the assumption of a very slow coalescence rate. While not the focus of this paper, we conducted a brief study of mixing speed on the particle size distribution of concentrated emulsions. The capillary number analysis showed that with the shear-thinning behavior of concentrated emulsion viscosity, n=0.512,13,45, a two orders of magnitude change in 2) (1-100 s-1) resulted in only one order of magnitude change in Ca (Ca(2) =100s-1)/Ca(2) =1s-1) = ~10) and one order of magnitude changed in λ (λ (2) =100s-1)/ λ (2) =1s-1) = ~10). The effect of 2) on d32 is dampened because of the shear-thinning behavior of concentrated emulsions. To illustrate this, the effect of mixing speed was observed by making optimized emulsions from 80% oil B and 20% of an aqueous phase with 1.6% phenol-15EO, 0.2% NaOH and 1.4% NaCl. The emulsions were mixed at 60 oC by gently tilting the sample vial upside down (~1-10 s-1) and by vigorously shaking (~100-1,000 s-1) the samples 10 seconds every 30 minutes for 4 hours. Both procedures resulted in d32=15±0.2 µm and =0.8±0.01. We explored the effects of the frequency of mixing on the and d32 of emulsions as illustrated in Fig. S2 (Supporting Information). Mixing the samples more frequently resulted in same within experimental uncertainty and a smaller d32 when the mixing frequency was 0.90 exhibited significantly lower viscosity compared to unimodal emulsions with < 0.75. This is essential for the transport of heavy oil emulsions in pipelines. Millions of combinations of surfactants, co-solvents, alkali, and electrolytes over a wide range of concentrations can be used to make heavy oil emulsions using different preparation methods and variables. Most of these emulsions do not have the desired properties for emulsion transport. It was essential for our purposes to develop a systematic procedure for selecting and testing the best chemical formulations. Concentrated (>0.64) heavy oil-in-water emulsions with as high as 0.95 were prepared in a simplified one-step mixing process by optimizing the particle size distribution based on fundamental principles of interfacial activity and rheology. The viscosity of multimodal concentrated emulsions ( > 0.85) prepared with our method showed comparable if not lower viscosity than the heavy oil emulsions prepared by the common two-step process (mixing two unimodal emulsions) reported in the literature.

Extensive rheological

characterization of the optimized emulsions prepared by this method is beyond the scope of this paper, but will be presented in a future paper. 28


Page 29 of 57

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

Supporting Information. The effect of standard deviation of an unimodal lognormal distribution (S1) and mixing frequency (S2) on . Fitting parameters for the particle size distribution of emulsions in Fig. 7 (S3). pH of alkali in DI water (S4). This material is available free of charge via the Internet at http://pubs.acs.org. Corresponding Author *Tel.: 1-706-266-7336. Fax: 512-471-1006. Email: [email protected] Notes The authors declare no competing financial interest. ACKNOWLEDGMENT This material is based upon research supported by the National Science Foundation Graduate Research Fellowship Program under Grant No. 2011128365. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. The authors would also like to express their gratitude to the industrial sponsors of The University of Texas at Austin Chemical EOR Industrial Affiliates Project, who lent important financial support for this research. Also, special thanks to the Morphodynamics Laboratory in the Jackson School of Geosciences at The University of Texas for the use of the Malvern Mastersizer 3000. Nomenclature ASP

Alkali surfactant polymer

Ca

Capillary number 29


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

3∗

Contact modulus

Continuous-phase viscosity

Cacritical

Critical capillary number at which droplet breakup occurs

8

Diameter of the group of droplets with larger 8 in a bimodal/binary mixture

8

Diameter of the group of droplets with smaller 8 in a bimodal/binary mixture

!

Dispersed-phase viscosity

Dispersed-phase volume fraction of emulsions

R

Droplet radius

X Y:MZ[D

Emulsion viscosity

EOR

Enhanced oil recovery

EO

Ethylene oxide

9

Glass transition point

HIPR

High internal phase ratio

Г

Interfacial tension (IFT)

Maximum packing volume fraction ( ) of dispersed-phase possible without deformation of the spherical dispersed-phase

,]

Maximum packing volume fraction ( ) at the low salinity (high interfacial tension) where the doesn’t vary much as a function of salinity

DMPQ

Maximum packing volume fraction ( ) of saturated non-interacting spheres of n monodisperse size groups

B

Natural logarithm standard deviation of droplet diameter

Natural logarithm mean diameter of droplets

d

Mean diameter of droplets within a bin width from histogram data

e

Number of non-interacting monodisperse size groups

k

Numerical coefficient

O/W

Oil-in-water

?f

Probability density function of volume

PO

Propylene oxide 30


Page 30 of 57

Page 31 of 57

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

;

Random close packing of monodisperse, hard spheres

:;

Random loose packing of monodisperse, hard spheres

λ

Ratio of dispersed-phase viscosity to emulsion viscosity, ! / X Y:MZ[D

Ratio of emulsion viscosity to

K

Ratio of ! to

8

Sauter mean droplet diameter

G

Shear modulus

2)

Shear rate

2)

Shear strain rate

2)1

Shear strain rate at the wall

0

Shear stress

Na+

Sodium concentration in wt. % in aqueous solution

Na+inversion

Sodium concentration in wt. % in aqueous solution at the O/W to W/O transition sodium concentration

Volume fraction of the group of droplets with larger 8 in a bimodal/binary mixture with respect to the total dispersed-phase volume

Volume fraction of the group of droplets with smaller 8 in a bimodal/binary mixture with respect to the total dispersed-phase volume

Volume fraction of smaller droplet volume to total droplet volume, ? ⁄( + ) 8S

Volume-weighed mean diameter

W/O

Water-in-oil

01

Yield stress

REFERENCES (1) (2)

Speight, J. G. High Acid Crudes; Gulf Professional Publishing, 2014. Fortenberry, R. P.; Kim, D. H.; Nizamidin, N.; Adkins, S.; Pinnawala Arachchilage, G. W. P.; Koh, H. S.; Weerasooriya, U. P.; Pope, G. A. Use of Co-Solvents to Improve AlkalinePolymer Flooding. Soc. Pet. Eng. 2013. 31


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

(3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22)

Page 32 of 57

Taghavifar, M. Enhanced heavy oil recovery by hybrid thermal-chemical processes. Ph.D. Thesis [online], University of Texas at Austin, Austin, TX, 2014. http://repositories.lib.utexas.edu/handle/2152/24796 (accessed March 15, 2015). Pal, R. Novel viscosity equations for emulsions of two immiscible liquids. J. Rheol. 2001, 45, 509–520. Nuñez, G. A.; Sanchez, G.; Gutierrez, X.; Silva, F.; Dalas, C.; Rivas, H. Rheological Behavior of Concentrated Bitumen in Water Emulsions. Langmuir 2000, 16, 6497–6502. Romero, N.; Cárdenas, A.; Rivas, H. Creep compliance-time behavior and stability of bitumen in water emulsions. J. Rheol. 2000, 44, 1247–1262. Romero, N.; Cárdenas, A.; Henrı́quez, M.; Rivas, H. Viscoelastic properties and stability of highly concentrated bitumen in water emulsions. Colloids Surf. Physicochem. Eng. Asp. 2002, 204, 271–284. Acevedo, S.; Gutierrez, X.; Rivas, H. Bitumen-in-Water Emulsions Stabilized with Natural Surfactants. J. Colloid Interface Sci. 2001, 242, 230–238. Pal, R. Viscous behavior of concentrated emulsions of two immiscible Newtonian fluids with interfacial tension. J. Colloid Interface Sci. 2003, 263, 296–305. Farris, R. J. Prediction of the Viscosity of Multimodal Suspensions from Unimodal Viscosity Data. Trans. Soc. Rheol. 1968, 12, 281–301. Brouwers, H. J. H. Viscosity of a concentrated suspension of rigid monosized particles. Phys. Rev. E 2010, 81, 051402. Seth, J. R.; Mohan, L.; Locatelli-Champagne, C.; Cloitre, M.; Bonnecaze, R. T. A micromechanical model to predict the flow of soft particle glasses. Nat. Mater. 2011, 10, 838–843. Foudazi, R.; Masalova, I.; Malkin, A. Y. The rheology of binary mixtures of highly concentrated emulsions: Effect of droplet size ratio. J. Rheol. 2012, 56, 1299. Pal, R. Shear Viscosity Behavior of Emulsions of Two Immiscible Liquids. J. Colloid Interface Sci. 2000, 225, 359–366. Malkin, A. Y.; Masalova, I.; Slatter, P.; Wilson, K. Effect of droplet size on the rheological properties of highly-concentrated w/o emulsions. Rheol. Acta 2004, 43, 584– 591. Pal, R. Effect of droplet size on the rheology of emulsions. AIChE J. 1996, 42, 3181– 3190. Pal, R. Rheology of high internal phase ratio emulsions. Food Hydrocoll. 2006, 20, 997– 1005. Masalova, I.; Foudazi, R.; Malkin, A. Y. The rheology of highly concentrated emulsions stabilized with different surfactants. Colloids Surf. Physicochem. Eng. Asp. 2011, 375, 76– 86. Mewis, J.; Wagner, N. J. Current trends in suspension rheology. J. Non-Newton. Fluid Mech. 2009, 157, 147–150. Sollich, P.; Lequeux, F.; Hébraud, P.; Cates, M. Rheology of Soft Glassy Materials. Phys. Rev. Lett. 1997, 78, 2020–2023. Núñez, G. A.; Briceño, M.; Mata, C.; Rivas, H.; Joseph, D. D. Flow characteristics of concentrated emulsions of very viscous oil in water. J. Rheol. 1996, 40, 405–423. Nuñez, G. A. N.; Mata, C. E.; Blanco, C.; Chirinos, M. S.; Sánchez, G. A.; Colmenares, T. R.; Rivas, H. J.; Silva, F. A. Manufacture of stable bimodal emulsions using dynamic mixing. US6903138 B2, June 7, 2005. 32


Page 33 of 57

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

(23) Farr, R. S.; Groot, R. D. Close packing density of polydisperse hard spheres. J. Chem. Phys. 2009, 131, 244104. (24) Mehrotra, A. K.; Eastick, R. R.; Svrcek, W. Y. Viscosity of cold lake bitumen and its fractions. Can. J. Chem. Eng. 1989, 67, 1004–1009. (25) Chirinos, M. L.; Taylor, A. S.; Taylor, S. E. Preparaton of HIPR emulsions and transportation thereof. US4934398 A, June 19, 1990. (26) Mie, G. Beiträge zur Optik trüber Medien, speziell kolloidaler Metallösungen. Ann. Phys. 1908, 330, 377–445. (27) Brouwers, H. J. H. Packing fraction of particles with lognormal size distribution. Phys. Rev. E 2014, 89, 052211. (28) Hopkins, A. B.; Stillinger, F. H.; Torquato, S. Disordered strictly jammed binary sphere packings attain an anomalously large range of densities. Phys. Rev. E 2013, 88, 022205. (29) Spangenberg, J.; Scherer, G. W.; Hopkins, A. B.; Torquato, S. Viscosity of bimodal suspensions with hard spherical particles. J. Appl. Phys. 2014, 116, 184902. (30) Furnas, C. C. The relations between specific volume, voids, and size composition in systems of broken solids of mixed sizes; US Department of Commerce, Bureau of Mines, 1928. (31) McGeary, R. K. Mechanical Packing of Spherical Particles. J. Am. Ceram. Soc. 1961, 44, 513–522. (32) Otsubo, P. Y.; Prud’homme, R. K. Effect of drop size distribution on the flow behavior of oil-in-water emulsions. Rheol. Acta 1994, 33, 303–306. (33) Shewan, H. M.; Stokes, J. R. Analytically predicting the viscosity of hard sphere suspensions from the particle size distribution. J. Non-Newton. Fluid Mech. 2015, 222, 72–81. (34) Lu, J.; Liyanage, P. J.; Solairaj, S.; Adkins, S.; Arachchilage, G. P.; Kim, D. H.; Britton, C.; Weerasooriya, U.; Pope, G. A. New surfactant developments for chemical enhanced oil recovery. J. Pet. Sci. Eng. 2014, 120, 94–101. (35) Flaaten, A.; Nguyen, Q. P.; Pope, G. A.; Zhang, J. A Systematic Laboratory Approach to Low-Cost, High-Performance Chemical Flooding. SPE Reserv. Eval. Eng. 2009, 12, 713– 723. (36) Levitt, D.; Jackson, A.; Heinson, C.; Britton, L. N.; Malik, T.; Dwarakanath, V.; Pope, G. A. Identification and Evaluation of High-Performance EOR Surfactants. SPE Reserv. Eval. Eng. 2009, 12, 243–253. (37) Baldauf, L. M.; Schechter, R. S.; Wade, W. H.; Graciaa, A. The relationship between surfactant phase behavior and the creaming and coalescence of macroemulsions. J. Colloid Interface Sci. 1982, 85, 187–197. (38) Bourrel, M.; Graciaa, A.; Schechter, R. S.; Wade, W. H. The relation of emulsion stability to phase behavior and interfacial tension of surfactant systems. J. Colloid Interface Sci. 1979, 72, 161–163. (39) Perez, M.; Zambrano, N.; Ramirez, M.; Tyrode, E.; Salager, J.-L. Surfactant-Oil-Water Systems Near the Affinity Inversion. XII: Emulsion Drop Size Versus Formulation and Composition. J. Dispers. Sci. Technol. 2002, 23, 55. (40) Tolosa, L.-I.; Forgiarini, A.; Moreno, P.; Salager, J.-L. Combined Effects of Formulation and Stirring on Emulsion Drop Size in the Vicinity of Three-Phase Behavior of Surfactant−Oil Water Systems. Ind. Eng. Chem. Res. 2006, 45, 3810–3814. 33


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 34 of 57

(41) Salager, J.-L.; Perez-Sanchez, M.; Garcia, Y. Physicochemical parameters influencing the emulsion drop size. Colloid Polym. Sci. 1996, 274, 81–84. (42) dos Santos, R. G.; Bannwart, A. C.; Briceño, M. I.; Loh, W. Physico-chemical properties of heavy crude oil-in-water emulsions stabilized by mixtures of ionic and non-ionic ethoxylated nonylphenol surfactants and medium chain alcohols. Chem. Eng. Res. Des. 2011, 89, 957–967. (43) Jansen, K. M. B.; Agterof, W. G. M.; Mellema, J. Droplet breakup in concentrated emulsions. J. Rheol. 2001, 45, 227–236. (44) Grace, H. P. Dispersion Phenomena in High Viscosity Immiscible Fluid Systems and Application of Static Mixers as Dispersion Devices in Such Systems. Chem. Eng. Commun. 1982, 14, 225–277. (45) Meeker, S. P.; Bonnecaze, R. T.; Cloitre, M. Slip and flow in pastes of soft particles: Direct observation and rheology. J. Rheol. 2004, 48, 1295–1320. (46) Sahni, V.; Dean, R. M.; Britton, C.; Kim, D. H.; Weerasooriya, U.; Pope, G. A. The Role of Co-Solvents and Co-Surfactants in Making Chemical Floods Robust. Soc. Pet. Eng. 2010. (47) Chang, L. Y. New correlation for predicting the best surfactant and co-solvent structures to evaluate for chemical EOR. M.S. Thesis [online], University of Texas at Austin, Austin, TX, 2014. http://repositories.lib.utexas.edu/handle/2152/28300 (accessed May 10, 2015). (48) dos Santos, R. G.; Bannwart, A. C.; Loh, W. Phase segregation, shear thinning and rheological behavior of crude oil-in-water emulsions. Chem. Eng. Res. Des. 2014, 92, 1629–1636. (49) Coulaloglou, C. A.; Tavlarides, L. L. Drop size distributions and coalescence frequencies of liquid-liquid dispersions in flow vessels. AIChE J. 1976, 22, 289–297. (50) Gutierrez, X.; Silva, F.; Morles, A.; Pazos, D.; Rivas, H. The Use of Amines in the Stabilization of Acidic Hydrocarbons in Water Emulsions. Pet. Sci. Technol. 2003, 21, 1219–1240. (51) Verzaro, F.; Bourrel, M.; Garnier, O.; Zhou, H. G.; Argillier, J.-F. Heavy Acidic Oil Transportation By Emulsion In Water. Soceity Pet. Eng. 2002. (52) Rihan, R.; Shawabkeh, R.; Al-Bakr, N. The Effect of Two Amine-Based Corrosion Inhibitors in Improving the Corrosion Resistance of Carbon Steel in Sea Water. J. Mater. Eng. Perform. 2014, 23, 693–699. (53) Seth, J. R.; Locatelli-Champagne, C.; Monti, F.; Bonnecaze, R. T.; Cloitre, M. How do soft particle glasses yield and flow near solid surfaces? Soft Matter 2012, 8, 140. (54) Ashrafizadeh, S. N.; Kamran, M. Emulsification of heavy crude oil in water for pipeline transportation. J. Pet. Sci. Eng. 2010, 71, 205–211. (55) Abdurahman, N. H.; Rosli, Y. M.; Azhari, N. H.; Hayder, B. A. Pipeline transportation of viscous crudes as concentrated oil-in-water emulsions. J. Pet. Sci. Eng. 2012, 90–91, 139– 144. (56) Hasan, S. W.; Ghannam, M. T.; Esmail, N. Heavy crude oil viscosity reduction and rheology for pipeline transportation. Fuel 2010, 89, 1095–1100. (57) Ahmet, N. S.; Nassar, A. M.; Zaki, N. N.; Gharieb, H. K. Stability and Rheology of Heavy Crude Oil-in-Water Emulsion Stabilized by an Anionic- Nonionic Surfactant Mixture. Pet. Sci. Technol. 1999, 17, 553–576. 34


Page 35 of 57

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

(58) Wylde, J. J.; Leinweber, D.; Botthof, G.; Oliveira, A. P.; Royle, C.; Kayser, C. Heavy oil transportation: Advances in water-continuous emulsion methods. World Heavy Oil Congr. 2012, WHOC12-333. (59) Zaki, N.; Butz, T.; Kessel, D. Rheology, Particle Size Distribution, and Asphaltene Deposition of Viscous Asphaltic Crude Oil-in-Water Emulsions for Pipeline Transportation. Pet. Sci. Technol. 2001, 19, 425–435.

Tables Table 1: Heavy crude oil properties TotalC (A) Zuata (B) Unknown

PRB (C) Ugnu (D)

Venezuela Canada

Alaska

Dynamic viscosity (mPa·s) at 25oC & 10 s-1 310,000

93,000

62,500

9,000

Specific gravity at 25oC

1.01

0.99

1.02

0.97

API gravity at 25oC

8-9

10-11

7-8

14-15

Total acid number (mg KOH/g oil)

6.40±0.1

3.85±0.2

Origins

35


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 36 of 57

Table 2: Saturated non-interacting DMPQ estimated as a function of standard deviation using the Groot and Farr model (Assumed B = BR = B )

Unimodal

Bimodal

Trimodal

B

/MPQ

MPQ

MPQ

0

0.64

0.87

0.96

0.1

0.65

0.88

0.97

0.5

0.71

0.91

0.98

1.0

0.80

0.96

0.99

36


Page 37 of 57

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

Energy & Fuels

Table 3: Bimodal emulsions are prepared by mixing unimodal emulsions of varying 8]. at various volume fractions in the literature. dL, dS, , experimental optimum ⁄( + ), and are reported in the literature. and calculated optimum ⁄( + ) are estimated using the Farr & Groot model with standard deviation of B =B =0.3. *at shear stress = 0.9 Pa **at shear rate = 10 s-1 Dispersed-Phase

ContinuousPhase

Mineral Oil

Water

Light Crude Oil

Water

Heavy Crude oil

Water

Saturated solution of NH4NO3 in water

Hydrocarbon oil

Surfactants

Octylphenol ethoxylate (10EO) Octylphenol ethoxylate (10EO) Nonylphenol ethoxylate (17.5EO)

PIBSA-Urea

gj lg k

gh.i (um)

N

NO

NmnopOqO =

Nr lN + N r j

dL = 1.95 dS = 1.28

1.50

0.94

~0.68

Range tested 0-1.00

dL = 18 dS = 5

3.60

0.75

~0.76

0-1.00

~0.36

~0.20-0.35

dL = 20 dS = 4

5.00

0.70

~0.80

0-1.00

0.25-0.30

~0.20-0.30

dL =20 dS = 2

10.00

0.70

~0.85

0-1.00

0.25-0.30

~0.20-0.30

dL = 30 dS = 2

15.00

0.70

~0.86

0-1.00

0.25-0.30

~0.20-0.30

dL = 20 dS = 2

10.00

0.80

~0.85

0-1.00

0.25-0.30

~0.20-0.30

dL = 30 dS = 2

15.00

0.80

~0.86

0-1.00

0.25-0.30

~0.20-0.30

dL = 16.9 dS = 8.2

2

0.85

~0.71

0-1.00

0

~0.20-0.40

dL = 16.9 dS = 5.6

3

0.85

~0.74

0-1.00

0

~0.20-0.30

dL = 16.9 dS = 2.7

6

0.85

~0.83

0-1.00

0.15-0.20

~0.20-0.30

37


Experimental optimum 0

Calculated optimum ~0.35-0.50

s ( mPa·s ) at th ru

Ref.

= [;QZ Y

17

=~1,500*

=>10,000,000*

[;QZ Y =~100*

=~1,000

[;QZ Y =~100

16

=~1,000

=~4,000

[;QZ Y =~70

=~550

=~4,000

[;QZ Y =~30

=~2,200

>7,000

[;QZ Y =~900

=~1,400

>7,000

[;QZ Y =~500

=~4,000**

>10,500**

[;QZ Y =

=~4,000**

>19,000**

[;QZ Y =

=~4,000**

>57,000**

[;QZ Y =~2,800**

5

5

5

5

5

13

13

13

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 38 of 57

Figures

Fig. 1. Viscosity of four heavy crude oils at 10s-1. The lines are Modified Walther Equation


38

Page 39 of 57

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

Fig. 2. Structures of a) phenol-mPO-nEO, b) IBA-nEO, and c) diisopropylamine-nEO.


39

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 40 of 57

Fig. 3. The O binary lognormal distribution as a function of theoretical R=dL/dS and ⁄( + ). σS= σL=0.3 used in the Farr and Groot model to calculate the points. The lines are to guide the eyes only.


40

Page 41 of 57

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

Fig. 4. Interfacial tension (left) and coalescence rate (right) of emulsions. Modified from Baldauf et al.37


41

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 42 of 57

Fig. 5. Photomicrographs of 80% oil D emulsions. a) d32=14.4 µm, φm=0.68, and aqueous composition (1.6% Ph15EO, 0.2% NaOH, 0% NaCl). b) d32=14.1 µm, φm=0.78, and aqueous composition (1.6% Ph15EO, 0.2% NaOH, 1% NaCl).


42

Page 43 of 57

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

Fig. 6a. The d32 of the entire lognormal distribution (primary axis) and O (secondary axis) of emulsions made from 80% oil B and 20% aqueous solution (0.4% NaCl and 0.2% NaOH) vs. the mixing temperature of the samples hand-shaken for 10 seconds every 30 minutes for 4 hours. O/W stands for oil-in-water emulsions. pH =9.14-9.16 for all emulsions. The lines are to guide the eyes only.


43

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 44 of 57

Fig. 6b. The d32 of the entire lognormal distribution (primary axis) and O (secondary axis) of emulsions made from 80% oil B and 20% aqueous solution (1.6% phenol-15EO, 0.4% NaCl and 0.2% NaOH) vs. the mixing temperature of the samples hand-shaken for 10 seconds every 30 minutes for 4 hours. O/W stands for oil-in-water emulsions. pH =9.14-9.16. The lines are to guide the eyes only.


44

Page 45 of 57

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

Fig. 7. Lognormal distribution fitting of particle size distribution of emulsions prepared with 80% oil and 20% aqueous solution (1.6% phenol-15EO, 0.8% NaCl, 0.2% NaOH).


45

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 46 of 57

Fig. 8a. The d32 of the entire lognormal distribution of emulsions made from 80% oil and 20% aqueous solution (0.8% NaCl, 0.2% NaOH, and 0-3.2% phenol-15EO) vs. the wt. % of phenol15EO in the aqueous solution. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 96oC. O/W and W/O stand for oil-in-water and water-in-oil emulsions. The lines are to guide the eyes only.


46

Page 47 of 57

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

Fig. 8b. The O of the entire lognormal distribution of emulsions made from 80% oil and 20% aqueous solution (0.8% NaCl, 0.2% NaOH, and 0-3.2% phenol-15EO) vs. the wt. % of phenol15EO in the aqueous solution. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 96oC. O/W and W/O stand for oil-in-water and water-in-oil emulsions. The lines are to guide the eyes only.


47

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 48 of 57

Fig. 9a. The d32 of the entire lognormal distribution of emulsions made from 80% oil B and 20% aqueous solution (0.8% NaCl, 0.2% NaOH, and 0-2.4% co-solvent) vs. the wt. % of co-solvent in the aqueous solution. The co-solvents are phenol-1PO-5EO, phenol-6EO, IBA-15EO, and phenol-15EO from the least hydrophilic to most hydrophilic as indicated in the direction of the arrow. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 96oC. The lines are to guide the eyes only.


48

Page 49 of 57

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

Fig. 9b. The O of the entire lognormal distribution of emulsions made from 80% oil B and 20% aqueous solution (0.8% NaCl, 0.2% NaOH, and 0-2.4% co-solvent) vs. the wt. % of cosolvent in the aqueous solution. The co-solvents are phenol-1PO-5EO, phenol-6EO, IBA-15EO, and phenol-15EO from the least hydrophilic to most hydrophilic as indicated by in direction of the arrow. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 96oC. The lines are to guide the eyes only.


49

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 50 of 57

Fig. 10. The d32 of the entire lognormal distribution (primary axis) and O (secondary axis) of emulsions made from 80% oil B and 20% aqueous solution (0-0.8% NaCl, 0.2% NaOH, and 0 & 1.6% phenol-15EO) vs. the wt. % of NaCl in the aqueous solution. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 96oC. pH = 9.9-10.1 for all emulsion samples. The lines are to guide the eyes only.


50

Page 51 of 57

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

Fig. 11a. The d32 of the entire lognormal distribution of emulsion made from 80% oil and 20% aqueous solution (0.2/0.4/0.6% NaOH, and 1.6% phenol-15EO) vs. Na+/Na+inversion in the aqueous solution. NaCl was used to vary the Na+. Oil A emulsions were prepared with 0.2% NaOH and 3% phenol-15EO. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 96oC. pH = 9.9-10.1 for 0.2% NaOH, 10.3-10.5 for 0.4% NaOH and 10.911.1 for 0.6% NaOH for oil B emulsions. The lines are to guide the eyes only.


51

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 52 of 57

Fig. 11b. The O /O,h of the entire lognormal distribution of emulsion made from 80% oil and 20% aqueous solution (0.2/0.4/0.6% NaOH, and 1.6% phenol-15EO) vs. Na+/Na+inversion in the aqueous solution. NaCl was used to vary the Na+. Oil A emulsions were prepared with 0.2% NaOH and 3% phenol-15EO. O,h=0.755, 0.73, and 0.685 for oil A, B, and D. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 96oC. pH = 9.9-10.1 for 0.2% NaOH, 10.3-10.5 for 0.4% NaOH and 10.9-11.1 for 0.6% NaOH for oil B emulsions. The lines are to guide the eyes only.


52

Page 53 of 57

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

Fig. 12a. The d32 of the entire lognormal distribution of emulsions made from 80% oil B and 20% aqueous solution (0.8% NaCl, alkali, and 1.6% phenol-15EO) vs. the wt. % of N+ in the aqueous solution. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 96oC.


53

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 54 of 57

Fig. 12b. The O of the entire lognormal distribution of emulsions made from 80% oil B and 20% aqueous solution (0.8% NaCl, alkali, and 1.6% phenol-15EO) vs. the wt. % of N+ in the aqueous solution. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 96oC.


54

Page 55 of 57

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

Fig. 13. The d32 of the entire lognormal distribution (primary axis) and O (secondary axis) of emulsions made from 80% oil B and 20% aqueous solution (up to 3.2% DIPA-15EO, 0.4 & 0.8% NaCl) vs. the wt. % of DIPA-15EO in the aqueous solution. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 96oC. The lines are to guide the eyes only.


55

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 56 of 57

Fig. 14. The d32 of the entire lognormal distribution (primary axis) and O (secondary axis) of emulsions made from 80% oil B and 20% aqueous solution (0.25-1.5% NPE-12EO, 0.8% NaCl, 0.2% NaOH, 0 & 1.6% phenol-15EO) vs. the wt. % of NPE-12EO in the aqueous solution. The sample was mixed for 10 seconds every 30 minutes over a period of 4 hours at 75oC. The lines are to guide the eyes only.


56

Page 57 of 57

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

Fig. 15. The viscosity of emulsions made from 80% oil B and 20% aqueous solution (1.6% phenol-15EO, 0.1-1.4% NaCl, 0.4% NaOH, 96oC mixing) measured with the cross-hatched parallel plate (r = 25 mm, h = 1 mm). The measurements were conducted at 21.5oC ± 0.5oC. Inset: The d32 and O calculated from the lognormal particle size distribution of the emulsions


57

Systematic Study of Heavy Oil Emulsion Properties ... - ACS Publications

Recommend Documents