JOURNAL OF COLLOID AND INTERFACE SCIENCE 189, 242253 (1997)

ARTICLE NO. CS974807

Effects of Asphaltene Solvency on Stability
of Water-in-Crude-Oil Emulsions
Joseph D. McLean and Peter K. Kilpatrick
1
Department of Chemical Engineering, North Carolina State University, Raleigh, North Carolina 27695-7905
Received July 25, 1996; accepted January 31, 1997
(wash water) is deliberately emulsied into the oil to
The formation of stable and persistent emulsions and foams in
contact hydrophilic materials, which then partition into the
the production and rening of crude petroleum is a challenge
aqueous droplets (10, 11). These desalter emulsions must
which has deed broad and generic resolution for several decades.
subsequently be broken down to recover the clean crude
Rational and systematic approaches to demulsication have been
oil. However, a considerable proportion of these emulsions
slow to develop due to a lack of fundamental understanding of
are very stable and lead to sludge generation (12). As the
the molecular origins of emulsion stabilization and the full range
EPA continues to redene specic petroleum-derived com-
of factors which govern emulsion stability. Several studies have
ponents which will be targeted as listed wastes, the need
shown the importance of resins and asphaltenes, which have the
to minimize the formation of petroleum emulsions and the ability to organize and form rigid lms at the oil /water interface.
We have developed a molecular model in which we propose that impetus to develop new methods for effective treatment of
the integrity of these lms and thus their ability to stabilize water-
them have become more acute.
in-crude-oil emulsions are sensitive to a variety of crude solvency
Despite years of research, there is a lack of a fundamental
parameters, such as aromaticity, resin-to-asphaltene ratio, and po-
understanding of the mechanisms governing the stability of
lar functional group concentration. This model was tested by corre-
oil-continuous emulsions. With crude oil emulsions, the im-
lating the stability of emulsions formed from a variety of crude
portance in long-term stability of a rigid and protective lm
oils Arab Berri ( Extra Light ) , Arab Heavy, Alaska North Slope,
surrounding the water droplets is clear (1318). The de-
and San J oaquin Valley in which the resin and asphaltene con-
tailed properties of this lm, together with a fundamental
tents vary, as well as their specic characteristics. The results of
knowledge of the chemistry of the interfacially active com-
the elemental and functional group characterization of these
ponents in the crude oil, are far from understood (19). How-
crudes and their fractions and the techniques utilized to obtain
ever, studies have established the signicance of components them were presented previously. Detailed quantitative protocols
for gauging relative emulsion stability have been developed to such as asphaltenes and resins, which, because of their sur-
further evaluate the proposed model by blending solvents of vary-
face-active and structure-forming properties, accumulate at
ing aromaticity and by doping isolated resins from different crudes
the interface and hence have pronounced effects on emulsion
into solvent-modied crudes. Dramatic destabilization of emul-
stabilization and breaking (11, 1316, 1931). Others have
sions was accomplished by modifying the crude solvency in either
also shown the importance of other components which may
fashion. Simple physical and chemical techniques for minimizing
be present in the crude such as waxes and inorganic solids
emulsion formation such as basic crude blending and solvent-
(8, 18, 19, 24, 3142). The focus of this study will be on
recycle schemes will also be discussed. 1997 Academic Press
the role of the asphaltenes and their interactions with the
resins and surrounding crude medium in forming these in-
terfacial lms and the resultant effects on emulsion stability.
INTRODUCTION
Therefore, it is of interest to isolate and characterize these
specic crude oil components.
Waste minimization and cleanup have become the focus
It has been suggested that the asphaltenes tend to form
of environmental policies in the petroleum industry (28).
aggregates which have an adsorbed covering sheath of aro-
As a preliminary step during renery treatment, crude oils
matic resins as a stabilizing layer (4345). These proposed
are normally desalted to remove species such as chloride
interactions between the resins and asphaltenes are illus-
salts which poison renery catalysts and enhance corrosion
trated in Fig. 1. Strong indications have been obtained that
of overhead distillation columns (9). In this process, water
emulsions stabilized by individual asphaltene molecules,
which are dissolved in the oil phase in the form of a molecu-
lar solution, are of little signicance for crude oils, when
1
To whom correspondence should be addressed. E-mail: Kilpatrick@
che.ncsu.edu. compared with stabilization by colloidally dispersed asphal-
242
0021-9797/97 $25.00
Copyright 1997 by Academic Press
All rights of reproduction in any form reserved.
AID JCIS 4807 / 6g24$$$$41 04-29-97 21:49:06 coida
243 STABILITY OF WATER-IN-CRUDE-OIL EMULSIONS
Asphaltene association is also determined in part by the
balance of acidic to basic functional groups within its chemi-
cal structure and also that of the resins.
It is evident that several fundamental research questions
need to be addressed: (i ) What key physical properties of
the petroleum and contacting water in the renery dictate
the ultimate stability of the emulsion? Asphaltene and resin
content? Water pH and ionic strength? (ii ) What are the
chemical functional groups in the resins and asphaltenes
which contribute to their interfacial activity? (iii ) What in-
teractions between the interfacially active components (res-
ins and asphaltenes) are responsible for the stability of these
emulsions? The answers to these and related questions serve
as starting points for development of emulsion destabiliza-
tion and minimization techniques.
The purpose of this study is to determine the effect of the
solvency state of the asphaltenes on the stability of water-in-
crude-oil emulsions. An emulsion stabilization mechanism
is presented and discussed in which the solvency of the
asphaltenes is mediated by the resins and the surrounding
FIG. 1. Schematic of the proposed mechanism of interaction between crude medium. The state of asphaltene solubility is proposed
resins and asphaltenes.
to be governed by the characteristics (e.g., carbon type and
polar functional group content ) of each of these components
tenes (15, 28, 4648). As a rule, asphaltenes are not soluble
which in turn determine the nature and extent of the interac-
in the distillate part of the crude or residual oil, but they do
tions between them. These effects will be investigated from
dissolve to varying extents when associated with resins (43,
two different angles: (i ) by correlating the stability of emul-
4951). It has been further determined that asphaltenes sta-
sions generated from four different crude oils with character-
bilize water-in-oil emulsions only if they are near or above
ization parameters which are anticipated to be indicative of
the point of incipient occulation, which suggests that their
the crude solvency state according to the proposed mecha-
mode of action is to collect at the interface in the form of
nism of emulsion stability, and (ii ) by modifying the state
nely divided solid particles or aggregates (28, 46, 52).
of asphaltene solvency directly through the addition of resins
From these studies, it can be inferred that modifying the
isolated from different crude types and through the addition
solvency of the asphaltenes from a state of dispersion to a
of solvents of varying aromaticity and then determining the
state of dissolution should decrease the ability of asphaltenes
resultant effect on the stability of the emulsions produced
to stabilize emulsions.
from these modied crude oils. These studies will also pro-
Since the resins and asphaltenes in crude oil are known
vide some insight into potential methods of emulsion resolu-
to contain acidic and basic groups, the pH of the water in
tion and minimization.
an emulsion can be expected to affect both the quantities
and types of materials constituting the interfacial lm. The
EXPERIMENTAL
aqueous-phase pH was indeed found to signicantly affect
the physical properties of the adsorbed lms (e.g., rigidity
Materials. The crude oils chosen for this study (Arab
and mobility), which led to laboratory tests of pH adjustment
Berri, Arab Heavy, Alaskan North Slope, and San Joaquin
in emulsion breaking (16, 53).
Valley) were selected because of their extensive use in oil
Several testable hypotheses were formulated from this ex-
reneries and because they represent extremes in gravity,
tensive and in-depth survey of the literature:
resin and asphaltene contents, and emulsion-forming tenden-
cies. The general properties of the crude oils selected for Asphaltenic colloids solvated with resinous material
play an important role in emulsion stabilization. this study are presented in Table 1. Arab Heavy (AH) is a
high sulfur crude produced from an off-shore eld, Safaniya, Adsorption of these components forms lms of consid-
erable mechanical strength at the droplet interface and ren- known to be the worlds largest offshore eld (54, 55). Arab
Berri (i.e., Arab Berri Extra Light, AB) is a high API ders the droplet stable to coalescence.
The state of dispersion of the asphaltenes (molecular gravity, low sulfur crude produced from Upper Jurassic Arab
zone reservoirs in Berri, Saudi Arabia (55, 56). The nitro- vs colloidal ) is critical to the strength or rigidity of these
lms and hence to the stability of petroleum emulsions. gen-rich San Joaquin Valley (SJV) crude oil (a blend of
AID JCIS 4807 / 6g24$$$$42 04-29-97 21:49:06 coida
244 MCLEAN AND KILPATRICK
TABLE 1
General Properties of Crude Oils
Property Test methods SJV ANS AB AH
Gravity, API ASTM 12.2 27.5 38.6 27.4
D 287
Viscosity,
a
SUS at 100F ASTM 15,025 72 41 118
D 445
Nitrogen, wt% GC combustion 0.90 0.31 0.07 0.25
Sulfur, wt% GC combustion 1.20 1.42 1.52 3.16
a
Viscosity of distilled water at 68F is 31 SUS.
crudes from San Joaquin Valley, California) is a crude of air interface. Immediately after homogenizing, the emul-
sions were transferred to 10-mL polypropylene centrifuge low API gravity and unusually high viscosity (57). Alaskan
North Slope (ANS) is produced from elds in the North tubes; most emulsions poured readily immediately following
Slope of Alaska, predominantly in the Prudhoe Bay region. emulsication. The amounts of oil and water to be emulsied
(including the W/ O ratio), the geometry of emulsifying
Methods. To ensure homogeneity of the oil samples, the
equipment (e.g., 15-mL jar, 6-mm rotor/ stator congura-
whole crudes were mixed thoroughly with the use of a Harbil
tion), and the amount of energy input to the system specied
GQM high-speed paint mixer for 3 min. The aqueous phase
in the emulsication protocol were determined in prelimi-
(i.e., deionized water) was prepared by adjusting the pH
nary experiments to ensure reproducibility, complete emulsi-
using diluted NaOH and HCl. The emulsions were produced
cation of materials, and droplet size distributions typical
according to a standard protocol outlined in schematic form
of emulsions produced in the reneries (550 mm).
in Fig. 2. Six milliliters of deionized water (at specied pH)
Referring again to Fig. 2, the fresh emulsions were then and 4 mL of the whole crude oil were pipetted into a 15-
allowed to stand under the inuence of gravity for 24 h mL polypropylene jar. This mixture was then processed with
before the amount of water which separated was determined the use of a Virtishear Cyclone IQ homogenizer with a 6-
since a period of rapid coalescence is known to immediately mm rotor/ stator conguration (gap width, 0.127 mm) at
follow the formation of an emulsion (15). The volume of 15,000 rpm for 3 min at the oil / water interface, for 2 min
at the bottom of the jar, and for 1 min just below the oil / oil and water resolved, if any, after gravitation for 24 h
FIG. 2. Schematic of the emulsion generation and standard/modied stability protocols.
AID JCIS 4807 / 6g24$$$$42 04-29-97 21:49:06 coida
245 STABILITY OF WATER-IN-CRUDE-OIL EMULSIONS
was measured via visual inspection by placing the emulsion
samples next to a graduated centrifuge tube. The emulsions
were then centrifuged using a RC5C refrigerated centrifuge
from Sorvall Instruments at 15,000 rpm (28,700g) and 30C
for 30 min. The volumes of water and oil separated due
to centrifugation were determined by decanting the oil and
pipetting the water into separate graduated tubes. The
amount of resolved water is the most appropriate gauge of
emulsion stability in water-in-crude-oil emulsions since co-
alescence of the droplet phase is the limiting step in the
demulsication process (53).
To study the effects of changes in system variables (e.g.,
crude oil types, added resins, added solvents of varying aro-
matic carbon content, etc.), a modied stability protocol was
employed which resolved approximately half the water from FIG. 3. Effect of crude oil types and pH of aqueous phase on emulsion
stability.
the whole crudes (i.e., base cases). As shown in Fig. 2, the
modied stability protocol simply adds heating and recentri-
fugation steps to the standard stability protocol to attain the
ity sedimentation (i.e., all of the water resolved after gravita-
desired water resolution. When the modied protocol was
tion for 24 h). This was thought to be due to its high viscos-
used, the emulsion samples were heated in a water bath to
ity at room temperature, which inhibits the shearing of water
a specied temperature for 18 h after the initial centrifuga-
droplets to the micrometer-level droplet sizes necessary to
tion and were then recentrifuged at the previously noted
form a stable emulsion. Attempts to form an emulsion with
specications. By changing the amount of heat input into
SJV at a temperature of 100C were successful to some
the system between centrifugations, one can control and de-
extent. The heating of the SJV crude oil to 100C would
termine the amount of water resolved for each of the base
be sufcient to melt any parafnic and/ or microcrystalline
cases in a particular study. This enables the investigator to
waxes present in the crude and consequently lower the vis-
broaden the window of observed trends in relative emulsion
cosity of the continuous oil phase enough to allow intimate
stability. Preliminary results indicated the standard protocol
mixing of the oil and water. However, these emulsions were
was sufcient to obtain the desired water resolution (50%)
still not very stable (approximately half the water resolved
for Alaskan North Slope crude, while Arab Berri and Arab
after settling under gravity for 24 h) compared to the emul-
Heavy crudes required heating to 45 and 70C, respectively,
sions formed from the other crudes in this study at room
and recentrifugation.
temperature. An emulsion formation and stability protocol
The crude oils under investigation were fractionated ac-
which allows analysis at higher temperatures and pressures
cording to their resin and asphaltene contents and were sub-
is currently being developed. With this in mind, the stability
sequently characterized with respect to their elemental, func-
of emulsions produced from these crudes in order of decreas-
tional group, and aromatic carbon contents in a prior study.
ing rank are as follows:
The instruments and techniques used to obtain the composi-
tional analysis of these crudes are described in a previous
Arab Heavy Arab Berri Alaska North Slope
paper (1).
San Joaquin Valley.
RESULTS AND DISCUSSION
Effect of aqueous-phase pH. As depicted in Fig. 3, aque-
ous-phase pH affected emulsion stability markedly. With Effect of crude type. The emulsion stability results for
the types of crude oils in this study which formed emulsions AH and AB crudes, emulsions prepared with water at pH
10 and 12 were relatively unstable. The polar functional stable to gravity sedimentation are presented as a function
of both crude type and aqueous-phase pH in Fig. 3. It should groups contained in the resinous and asphaltenic material
originally present in the interfacial lm would most likely be noted that emulsion stability, according to this operational
denition, increases as the amount of resolved water de- be ionized at these extreme pH values. The ionization of
these polar groups would be capable of creating high surface creases (i.e., they are inversely related). It is apparent that
Arab Heavy forms the most stable emulsions (05% water charge densities that would drastically change the lm prop-
erties by introducing an internal repulsion in the lm, thereby resolved, except at basic pHs 10 and 12) and that Arab
Berri and Alaska North Slope form considerably weaker destroying the mechanical properties that give rise to protec-
tion against coalescence. The restored or enhanced stability emulsions (5090% water resolved). San Joaquin Valley
crude formed emulsions which were totally unstable to grav- observed in AB and AH emulsions at low pHs reveals the
AID JCIS 4807 / 6g24$$$$42 04-29-97 21:49:06 coida
246 MCLEAN AND KILPATRICK
ous phase upon emulsication. These results of the effects
of varying aqueous-phase pH suggest that these acidic com-
pounds are more effective at stabilizing emulsions when they
are not ionized.
Proposed mechanism of emulsion stability. Before we
can attempt to relate the observed trends in the stability of
the emulsions produced from the crude oils in this study to
certain characteristics of these crudes, we must consider how
the polar, surface-active constituents of the crude (e.g., res-
ins and asphaltenes) interact with each other and with the
surrounding crude medium and how these interactions could
affect the resultant emulsion stability. As mentioned in the
literature review, resins help to solubilize the asphaltenes by
forming a resin-solvated asphaltenic aggregate (illustrated
FIG. 4. Effect of ionization of interfacially adsorbed components from
in Fig. 1) and thus tend to diminish the surface-active nature
different crude types on pH of resolved water.
of the asphaltenes. However, as depicted in Fig. 5, these
resin-solvated asphaltenic colloids might be able to lower
their free energy by shedding the solvating resins on one
potential absence of strongly basic polar groups in these
side of the colloid to form a partially solvated aggregate
particular crude oils.
which is interfacially active and adsorbs at the water/ oil
Also, the water resolved when the crude emulsions are
interface. As pointed out in the literature review, the extent subjected to centrifugation and/ or heating was characterized
to which the asphaltenes are solvated is the controlling factor with respect to pH. This residual aqueous-phase pH is
in determining the surface-active nature of these colloidal plotted as a function of its initial pH for ANS and AB in
aggregates. In fact, asphaltenes which are molecularly dis- Fig. 4. For neutral and basic pHs, the residual aqueous-phase
solved are ineffective (with regard to their ability to stabilize pH is consistently and substantially more acidic than the
emulsions) compared to precipitated or colloidally dispersed initial pH by as much as 45 pH units for all the crudes
asphaltenes. studied. This suggests that the interfacially active com-
What are the dominant contributors to asphaltene solubil- pounds in the crude adsorbed at the oil / water interface are
acidic in nature and some ionize or dissociate into the aque- ity? If we consider, for example, a crude oil which is rich
FIG. 5. Mechanism of emulsion stabilization: Adsorption of asphaltenic aggregates.
AID JCIS 4807 / 6g24$$$$42 04-29-97 21:49:06 coida
247 STABILITY OF WATER-IN-CRUDE-OIL EMULSIONS
in solvating resins relative to its asphaltene content (i.e., a
crude oil which has a high resin-to-asphaltene, R/ A, ratio),
we would expect the asphaltenes in this particular crude oil
to be molecularly dissolved and therefore unable to stabilize
emulsions. Aside from the resin-to-asphaltene ratio, there
are at least two other characteristics that merit consideration:
(i ) Asphaltenes are known to be the most aromatic portion
of the crude oil; therefore we would expect that as the aroma-
ticity of the resins and surrounding oil medium increased
with respect to the aromaticity of the asphaltenes, the asphal-
tenes would become more molecularly dissolved and less
able to stabilize emulsions. (ii ) Hydrogen-bonding interac-
tions between the resins and asphaltenes could enhance sol-
vation of the asphaltenes. It stands to reason that if the
asphaltenes from a particular crude oil contain high concen-
trations (relatively speaking) of C|O groups and if the
corresponding resin fraction is rich in carboxylic acid or
other proton-donating polar functional groups, then intermo-
lecular hydrogen bonding could lead to strong solvation of
asphaltenes by resin molecules. The higher the concentration
and polarity of these functional groups, the stronger the inter-
actions between them, thus diminishing the ability of these
colloidal aggregates to shed their sheath of solvating resins
in order to become surface-active.
Apparently, then, the solubility of asphaltenes in crude
oil, their tendency to aggregate, and their tendency to adsorb
at oil / water interfaces should be controlled by: (i ) the ratio
FIG. 6. Schematic of dominant contributors to asphaltene solubility and
of resins to asphaltenes, (ii ) the aromaticity ratios of the resultant effects on their surface activity.
resins and the crude medium to the asphaltenes (as dened
in Fig. 7), and (iii ) the concentrations of the polar functional
interactions with the resin molecules than a crude oil of
groups (e.g., carbonyls, carboxylic acids, pyrolles, amides,
lower functional group content. Thus, the crude oil com-
and phenols) contained in the resin and asphaltene fractions
posed of resins and asphaltenes of high functionality would
of a given crude. These three modes of asphaltene solvation
contain asphaltenic aggregates that nd it difcult to shed
are illustrated in Fig. 6, in which the properties which deter-
their solvating resins and are consequently less surface-ac-
mine the extent of asphaltene solvation and their resultant
tive and have a lower propensity to form stable emulsions.
surface activity are highlighted. The difference in properties
Of course, each of these contributions is not isolated from
shown in Fig. 6 could either be borne out in a comparison
the other in determining the solubility state of the asphaltenes
of two different crude types or effected through chemical
within a given crude oil. In fact, one could also envision
addition in a given crude oil. These assertions also suggest
combining some or all of these ideas such as the addition
possible means for emulsion minimization and/ or resolution.
of resin molecules, which are both very aromatic and highly
The asphaltenes would become less surface-active with the
functional, into a particular crude to molecularly dissolve
increasing availability of solvating resins (at higher R/ A
the indigenous asphaltenes and minimize the formation of
values) whether through chemical addition or as a function
stable emulsions during processing.
of crude type. Increasing the aromaticity (i.e., aromatic car-
The results of the effect of the different crude types on
bon content ) of the crude medium through the addition of
emulsion stability are reviewed and the anticipated correla-
aromatic solvents should solvate strongly surface-active as-
tions between the characterization parameters of the crudes
phaltenes by creating an environment that is more like the
under study and their observed emulsion stabilities are sum-
nature of the condensed, polyaromatic asphaltene molecules.
marized in Fig. 7. Note that all of these parameters are
This in effect increases the aromaticity ratio of the crude
expected to be inversely related to the emulsion stability
medium with respect to the asphaltenes. Finally, a crude oil
according to the proposed mechanism.
which has a high polar functionality (i.e., high concentration
of functional groups) in its resin and asphaltene molecules Discussion of characterization results. With this in
mind, the yields and pertinent characterization results of the should form asphaltenic colloidal aggregates which are more
strongly solvated through the increased number of H-bond polar fractions isolated from these crudes are presented in
AID JCIS 4807 / 6g24$$$$43 04-29-97 21:49:06 coida
248 MCLEAN AND KILPATRICK
(1), the intensities of the aromatic carbon ring-breathing
modes (at 1600 cm
01
) are very sensitive to the polarity
of aromatic substituent groups. Therefore, the apparently
high aromatic carbon content of SJV polar fractions (as
measured by FTIR spectroscopy) is likely an artifact of the
strong polarity of these fractions, as indicated by the polar
functional group content already cited. While this might
seem to give rise to fractions which are more surface-active
in SJV than in AH, the increased polarity may actually cause
the asphaltenes in SJV to be more strongly solvated by the
resins and thus less surface-active as discussed previously.
Other distinctive features of the SJV polar fractions were the
H/ C ratio and the aromatic carbon content of the asphaltene
fraction. The H/ C ratio of the SJV asphaltenes was consider-
ably higher than those from the other crudes (1.17 vs 1.02
1.11), and the aromaticity value was lower (46% vs 49
FIG. 7. Results of emulsion stability analysis on different types of crude
55%). These results suggest either that SJV crude has more
oils and summary of anticipated effects of the proposed mechanism on
aliphatic character (less fused polyaromatic ring structure)
emulsion stability.
than other crudes or that the asphaltene fraction still contains
some strongly adsorbed resinous compounds. In either case,
Table 2. These results were compiled from a previous study the SJV polar fractions would likely be less capable of stabi-
(1). The fractionation was performed using a method similar lizing emulsions than the polar fractions of the other crudes,
to the SARA technique which employs precipitation of as- which was indeed found to be the case in this study.
phaltenes with heptane followed by chromatographic isola-
Correlation of R/ A and asphaltene contents with emulsion
tion of resins. Characterization of the polar fractions by ele- stability. The R/ A value and asphaltene contents of the
mental analysis and by FTIR spectroscopy revealed that AH four crudes in this study (in order of increasing emulsion
asphaltenes were richest in sulfur and considerably lower in stability) are presented graphically in Fig. 8. It is evident
oxygen content and carbonyl functional group content than that Arab Heavy has the lowest R/ A value and is also richest
SJV crude. These results reveal that the concentrations of in asphaltenes, as would be expected from the proposed
carboxylic/ carbonyl and pyrrolic functional forms are 1.5 stabilization mechanism. However, AB has the highest R/
3 times larger in SJV polar fractions than in AH polar frac- A value but forms the second most stable emulsion. The R/
tions, indicating that SJV polar fractions are considerably A value denitely tells us part of the story, but it is
obvious that one cannot depend solely on this parameter as more acidic and polar. Additionally, the aromatic carbon
a predictor of relative emulsion stability. content of the SJV asphaltenes as measured by FTIR spec-
troscopy was much higher than the true value obtained Effect of addition of resins on emulsion stability. To see
a more direct and comprehensive effect of the R/ A value via NMR spectroscopy. As discussed in a previous paper
TABLE 2
Summary of Characterization Analysis of Resins and Asphaltenes Isolated via Precipitation and Extrography
Resins Asphaltenes
Characterization
parameter SJV ANS AB AH SJV ANS AB AH
% Yield 20.26 9.47 3.49 9.56 4.57 3.35 0.68 8.27
H/C 1.38 1.38 1.36 1.39 1.17 1.08 1.02 1.11
% N 1.89 0.94 0.67 0.66 2.44 1.11 0.77 0.94
% S 1.90 3.29 4.76 5.95 1.60 3.06 5.84 7.17
% O 2.30 2.91 2.44 2.39 3.52 2.67 2.49 1.46
% Arom. C (NMR) 38.7 38.1 38.8 37.9 45.9 51.9 54.8 49.2
% Arom. C (FTIR) 17.0 11.9 11.7 18.4 62.8 38.2 46.9 28.7
% Carbonyl 2.3 1.5 1.7 1.3 1.5 1.0 1.7 0.6
% Pyrrole 0.6 0.3 0.3 0.3 0.5 0.2 0.2 0.1
V (ppm) 187.7 117.9 62.8 205.0 463.8 516.4 352.8 667.6
Ni (ppm) 268.0 64.4 43.0 85.8 873.1 386.0 130.5 352.1
AID JCIS 4807 / 6g24$$$$43 04-29-97 21:49:06 coida
249 STABILITY OF WATER-IN-CRUDE-OIL EMULSIONS
toluene, so-called heptol ) in increasing amounts (1050%).
Figure 10 shows the effects of blending the three different
solvents with AB and AH (prior to homogenization), respec-
tively, on emulsion stability as a function of the solvent
content in the oil phase. It is clear that addition of toluene
and heptol destabilizes the emulsions, while with sufcient
addition of n-heptane (3050%), the emulsions are actually
stabilized. This is fully consistent with the notion that in-
creased aromaticity of the crude medium helps to further
solubilize the asphaltenes. Moreover, with heptol and tolu-
ene, there may be a viscosity reduction effect. With n-hep-
tane, however, it is clear that the stabilization of the emulsion
must be due to reduced molecular solubility of the asphaltene
fraction of the crude. Despite the fact that 30% added n-
heptane is clearly not sufcient to precipitate the asphaltenes,
there must be reduced molecular solubilization resulting in
aggregation and surface activity.
Extent of effect of toluene blending and pH adjustment.
FIG. 8. Total asphaltene content and resin-to-asphaltene (R/ A) ratio
Since the blending of toluene had the greatest destabilizing
as a function of increasing emulsifying propensity (from left to right ) of
effect on the stability of emulsions produced from AB (e.g.,
four crude oils.
complete resolution at 30% toluene) as determined by the
modied protocol at 45C, an aqueous-phase pH scan with
on emulsion stability, we explored the option of adding or
a 30% blend of toluene was performed to determine the
doping of resins to the crude oil. This study has been
water resolution attained solely by gravity sedimentation
performed with most combinations of crude types plus resins
since this is the most practical method utilized in the rener-
from different crude types dissolved in heptol (to ensure
ies. The results in Fig. 11 show that the emulsions produced
complete dissolution of the resin fraction within the crude
with the crude/ toluene blend are stable under gravity for
oil ) to gain further insight into the interactions between these
acidic and neutral pHs (pH 28), but exhibit signicant
interfacially active constituents in the presence of the indige-
water resolution (70%) for basic pHs (pH 1012). No
nous crude oil. The effect of adding resins from AH, AB,
water resolution from AB crude emulsions without added
and SJV to solvent-modied AH crude oil (30% v/ v heptol
toluene is observed after the emulsion was allowed to settle
in crude oil ) using the modied protocol at 50C as a func-
under gravity; therefore, this result suggests that solvent-
tion of R/ A is presented in Fig. 9. The added AH resins
modied AB emulsions could be broken or even prevented
destabilize the emulsion appreciably as the R/ A is raised
from forming simply by controlling the pH of the aqueous
from 1.1 to 2.1 and then appears to have little effect thereaf-
phase.
ter. However, the addition of AB resins, which are relatively
higher in sulfoxide content, appears to have a more dramatic
and consistent destabilizing effect over the entire range of
R/ A values up to 3.7. Adding SJV resins, which are rich
in pyrollic and carboxylic functional groups, has the most
dramatic destabilizing effect when the R/ A value is raised
from 1.1 to 2.2. It is obvious that resins isolated from differ-
ent crude types have varying impacts on the resultant emul-
sion stability, which is exactly what we would expect based
on the proposed mechanism.
Effect of addition of solvents on emulsion stability. The
hypothesis that asphaltene solvency is directly responsible
for emulsion stability led to another study aimed at determin-
ing the effect of changing the aromatic nature of the crude
medium by blending solvents of varying amounts and aroma-
ticity with the whole crude oils. In our rst experiment,
FIG. 9. Effect of adding resins isolated from three different crude oils
crudes were systematically modied by addition of a purely
on stability of emulsions produced from a solvent-modied crude oil (Arab
aliphatic solvent (n-heptane), an aromatic solvent (tolu-
Heavy with 30% heptol ). Percentage water resolved versus R/ A of solvent-
modied crude oil. ene), or a mixture of the two (70% n-heptane and 30%
AID JCIS 4807 / 6g24$$$$43 04-29-97 21:49:06 coida
250 MCLEAN AND KILPATRICK
FIG. 12. Effect of blending solvents of varying molecular structure and
aromaticity on stability of emulsions produced from Arab Heavy.
stability as mediated by the state of dispersion of the asphal-
FIG. 10. Effect of blending solvents (prior to homogenization) of vary-
tenes in the crude medium. The results from this study are
ing aromaticity on the stability of emulsions produced from Arab Berri and
Arab Heavy. plotted in Fig. 12 as a function of the amount of solvent
added to AH crude oil prior to homogenization. It is evident
that all of the solvents are sufciently aromatic to destabilize
Effects of solvent aromaticity and molecular structure.
the emulsions and moreover that the most aromatic solvents
Solvents of varying molecular structures and aromaticities
(i.e., solvents of highest aromatic carbon content ) are more
including benzene, toluene, xylene, ethylbenzene, tert-butyl-
effective in resolving the emulsions. To see this effect more
benzene, cumene, and cymenewere also utilized to further
clearly, the slopes of these water resolution curves for the
gauge the inuence of the solvent power on emulsion
various solvents (at the onset of destabilization, 15%)
were plotted as a function of the solvent aromaticity as pre-
sented in Fig. 13. There is a denite trend toward greater
destabilization effectiveness as the molecular aromaticity of
the solvent increases. Since asphaltenes are known to be
condensed polyaromatic-ringed compounds, naphthalene
and phenanthrene were dissolved in benzene (at concentra-
tions just under their solubility limits of 0.37 and 0.35 g/
mL of benzene, respectively) to see if this would effect any
further destabilization of the emulsions produced with these
solvent-modied crudes. This was indeed found to be the
case, as Fig. 12 shows a considerable destabilization over
and above what was observed with the other pure solvents
over the entire range of added solvent.
Summary of correlation between crude oil characteristics
and their resultant emulsion stability. In summary, the re-
FIG. 11. Effect of aqueous-phase pH on stability of emulsions produced
sults from the characterization of these crudes in terms of
from a solvent-modied crude oil (Arab Berri with 30% toluene) as gauged
by gravity settling (i.e., no centrifugation or heating). asphaltene content, R/ A, the aromaticity ratios of the crude
AID JCIS 4807 / 6g24$$$$43 04-29-97 21:49:06 coida
251 STABILITY OF WATER-IN-CRUDE-OIL EMULSIONS
the effect of the very low value of the aromaticity ratio of
the oil to the asphaltenes in this particular crude. From the
solvent experiments just presented, the low aromaticity of
the AB crude medium with respect to the aromaticity of its
asphaltenes would be expected to play a large role in stabiliz-
ing emulsions produced from this crude oil. Therefore, even
though there does not appear to be a single, universal param-
eter to predict emulsion stability, there is preliminary evi-
dence to support the utility of a phase map or model which
includes some or all of these parameters in predicting relative
stability of emulsions produced from different crude oils.
CONCLUSIONS
The results from this study tend to support the proposed
FIG. 13. Effect of solvent molecular structure and aromaticity on effec-
mechanism in which emulsion stability is governed primarily
tiveness of destabilization of emulsions produced from Arab Heavy. Desta-
by the state of solubility of the asphaltenes in the crude
bilization effectiveness versus aromatic carbon content of the added solvent.
oil. Assertions from the literature stipulate that asphaltenes
which are at or near the point of precipitation are more
surface-active than those which are sufciently solvated or medium and resins to the asphaltenes, the concentration of
polar functional groups within the resins and asphaltenes, molecularly dispersed. The characteristics of the crude oil
which should play a role in determining the solubility state and how they relate to the stability of the emulsions produced
from these crudes are presented in Table 3. It is evident that of the asphaltenes include the resin-to-asphaltene ratio, the
aromaticity ratios of the crude medium and resins with re- SJV, which is the weakest emulsion former (at ambient
temperature) of the crudes investigated in this study, pos- spect to the asphaltenes, and the concentration of polar func-
tional groups in the resins and asphaltenes. These parameters sesses the highest concentration of polar functional groups
and the highest aromaticity ratios for both the oil and the were tabulated and correlated with the stability of the emul-
sions produced from these crude oils. The observed trends resins. AH on the other hand, which forms the most stable
emulsion, has the lowest R/ A and the lowest functional suggest that a model which incorporates some or all of these
parameters (and possibly others) could be utilized to predict concentrations of the crudes in this study. Both of these
observations fall in line with the proposed mechanism which relative emulsion stability.
This research also suggests a number of possible emulsion predicts an indirect correlation between all these characteris-
tics of the crude oil and the resulting emulsion stability. minimization approaches in the renery which could be im-
plemented at relatively low cost. One obvious possibility is However, AB appears to be an oddball considering that it
exhibits the second highest emulsion-forming propensity of the blending of crudes to achieve crude aromaticity and/ or
R/ A values which yield minimal stability emulsions. For the crudes in this study, yet it contains the least amount of
asphaltenes, the highest R/ A value, and the second highest example, when SJV is blended with AH crude in proportions
exceeding 30% SJV, the resulting blended crude forms very amount of polar functional groups. One must also consider
TABLE 3
Summary of Parameters Suspected to Control Emulsion Stability
Increasing emulsion stability
SJV ANS AB AH
Crude type whole whole whole whole
Asphaltene content
a
4.57 3.35 0.68 8.27
R/A 4.43 2.83 5.13 1.16
Arom. oil/arom. asphaltenes
b
0.45 0.45 0.31 0.43
Arom. resins/arom. asphaltenes
b
0.84 0.73 0.71 0.77
% Polar functional groups
c
4.44 1.64 1.98 1.21
a
Wt%.
b
As determined by
13
C NMR spectroscopy.
c
Wt% of total asphaltenes and resins as determined by FTIR spectroscopy.
AID JCIS 4807 / 6g24$$$$43 04-29-97 21:49:06 coida
252 MCLEAN AND KILPATRICK
2. Liem, A. J. S., and Woods, D. R., A. I. Ch. E. Symposium Series
unstable emulsions in our laboratory which settle upon the
70(144), 823 (1974).
action of gravity. Another straightforward implementation
3. Menon, V. B., and Wasan, D. T., in Surfactant Science Series
of the results of this research would involve the blending of
(K. J. Lissant, Ed.), Vol. 13, p. 162, Dekker, New York, 1983.
aromatic distillation side streams either into the desalting
4. Menon, V. B., Nagarajan, R., and Wasan, D. T., Sep. Sci. Technol.
unit or into a desalter knockout tank in order to solubilize 22(12), 22952322 (1987).
5. Lissant, K. J., Colloids Surf. 29, 15 (1988).
adsorbed asphaltenes and destabilize emulsions. Another op-
6. Curran, L. M., in A. I. Ch. E. Conference on Pollution Prevention for
tion that could be implemented in conjunction with these
1990s, Washington, DC, unpublished.
approaches is the adjustment of the wash water pH to an
7. Berridge, S. A., Thew, M. T., and Loriston-Clarke, A. G., J. Inst. Pet-
optimum pH, which varies from crude to crude, for which
rol. 54(539), 333357 (1968).
the resultant emulsion exhibits minimum stability. In this
8. Bridie, A. L., Wanders, T. H., Zegveld, W., and Heijde, H. B. v. d.,
Mar. Pollut. Bull. 11, 343348 (1980). study, both AB and AH crude emulsions were destabilized
9. Obah, B., Erdohl Kohle. Erdgas. Petrochem. 41(2), 7174 (1988).
at basic pHs (pH 1012). Adjustment of aqueous-phase pH
10. Menon, V. B., and Wasan, D. T., Sep. Sci. Technol. 23(12 & 13),
is less expensive than many other demulsication methods,
21312142 (1988).
such as the use of surfactants, and can appreciably reduce
11. Taylor, S. E., Chem. Ind. 20, 770773 (1992).
the requirements of more complex and expensive processes
12. Kilpatrick, P. K., Basic Principles and Control of Renery Emulsion
Formation, Petroleum Environmental Research Forum (PERF) Proj- needed to achieve the desired resolution of crude oil emul-
ect 88-07, 1991.
sions in the reneries.
13. Little, R. C., Fuel 53(10), 246252 (1974).
Further questions remain. How general are the results of
14. Shetty, C. S., Nikolov, A. D., and Wasan, D. T., J. Dispersion Sci.
the research reported here? What effect would temperature
Technol. 13(2), 121133 (1992).
and pressure changes have upon the stability of emulsions
15. Mackay, G. D. M., McLean, A. Y., Betancourt, O. J., and Johnson,
produced from the crudes in this study, particularly with B. D., J. Inst. Petrol. 59(568), 164172 (1973).
16. Strassner, J. E., J. Petrol. Technol. 20, 303312 (1968).
crudes of high wax contents? How critical is the role of
17. Dodd, C. G., J. Phys. Chem. 64(5), 544550 (1960).
asphaltenic polarity in affecting asphaltene solubility? What
18. Graham, D. E., Stockwell, A., and Thompson, D. G., in Chemicals
more insightful techniques are available to isolate and mea-
in the Oil Industry (P. H. Ogden, Ed.), pp. 7391. Royal Chem.
sure the polarity of these surface-active components? Are
Soc., London, 1983.
there hidden variables which also play an important role in
19. Johansen, E. J., Skjarvo, I. M., Lund, T., Sjoblom, J., Soderlund, H.,
and Bostrom, G., Colloids Surf. 34, 353370 (1989). asphaltene solubility, such as molecular weight? Does the
20. Cratin, P. D., Ind. Eng. Chem. 61(2), 3545 (1969).
minimization of emulsion stability formed from whole
21. Oren, J. J., and MacKay, G. D. M., Fuel 56(10), 382384 (1977).
crudes by solubilizing asphaltenes result in an enhancement
22. Mansurov, I. R., Ilyasova, E. Z., and Vygovskoi, V. P., Chem. Technol.
of emulsion stability as formed by adsorption of crude sur-
Fuels and Oils 23(12), 9698 ( 1987).
face-active materials on ultrane particulates, such as iron
23. Siffert, B., Bourgeois, C., and Papirer, E., Fuel 63(6), 834837
oxides or silica? Can phase maps of crudes be created in (1984).
24. Menon, V. B., and Wasan, D. T., Sep. Sci. Technol. 19(8 & 9), 555
which aromaticity, asphaltene and resin contents and asphal-
574 (1984).
tene polarity are the key variables in which emulsion stability
25. Mukherjee, S., and Kushnick, A. P., ACS Div. Petrol. Chem. 33(1),
can be identied as clear regimes on the map? Is it more
205210 (1988).
advantageous to solubilize asphaltenes and carry them with
26. Isaacs, E. E., Huang, H., Babchin, A. J., and Chow, R. S., Colloids
the crude into the catalytic cracking unit or is it preferable
Surf. 46, 177192 (1990).
27. Eley, D. D., Hey, M. J., Symonds, J. D., and Willison, J. H. M., J. to remove them in a desalter emulsion unit and deal with
Colloid Interface Sci. 54(3), 462466 (1976).
them separately? These questions combine molecular-level
28. Waarden, M. v. d., Kolloid Z. Z. Polymer 156(2), 116122 (1958).
thinking with process considerations and constitute a fruitful
29. Hassander, H., Johansson, B., and Tornell, B., Colloids Surf. 40, 93
area for ongoing research.
105 (1989).
30. Papirer, E., Bourgeois, C., Siffert, B., and Balard, H., Fuel 61(8),
ACKNOWLEDGMENTS 732734 (1982).
31. Aveyard, R., Binks, B. P., Fletcher, P. D. I., and Lu, J. R., J. Colloid
We are grateful to the Petroleum Environmental Research Forum (PERF)
Interface Sci. 139(1), 128138 (1990).
through Grants 91-05 and 95-02 and to the National Science Foundation
32. Thompson, D. G., Taylor, A. S., and Graham, D. E., Colloids Surf. 15,
(NSF) through a predoctoral fellowship to Joseph D. McLean for supporting
175189 (1985).
this research. We also acknowledge the work of a few outstanding under-
33. Sjoblom, J., Mingyuan, L., Hoiland, H., and Johansen, E. J., Colloids
graduate researchers who made signicant contributions to this study: Brian
Surf. 46, 127139 (1990).
Phillips, Samuel Boyce, and Aaron Hughes.
34. McMahon, A. J., Interfacial Aspects of Water-in-Crude Oil Emulsion
Stability, in EmulsionsA Fundamental and Practical Approach (J.
Sjoblom, Ed.), pp. 135156. Kluwer Academic, Dordrecht Norwell,
REFERENCES
MA, 1992.
35. Bobra, M., Fingas, M., and Tennyson, E., Chem. Tech. April, 236 1. McLean, J. D., and Kilpatrick, P. K., Comparison of Precipitation and
241 (1992). Extrography in the Fractionation of Crude Oil Residua. Energy and
Fuels, 1996. 36. Briggs, T. R., J. Ind. Eng. Chem. 13(11), 10081010 (1921).
AID JCIS 4807 / 6g24$$$$44 04-29-97 21:49:06 coida
253 STABILITY OF WATER-IN-CRUDE-OIL EMULSIONS
37. Dodd, C. G., Moore, J. W., and Denekas, M. O., Ind. Eng. Chem. 47. Jones, T. J., Neustadter, E. L., and Whittingham, K. P., J. Can. Petrol.
Technol. 17(2), 100108 (1978). 44(11), 25852590 (1952).
48. Sjoblom, J., Soderlund, H., Lindblad, S., Johansen, E. J., and Skjarvo,
38. Lucassen-Reynders, E. H., and Tempel, M. V. d., J. Phys. Chem. 67,
I. M., Colloid Polymer Sci. 268, 389398 (1990).
731734 (1963).
49. Swanson, J. M., J. Phys. Chem. 46, 141150 (1942).
39. Mizrahi, J., and Barnea, E., Br. Chem. Eng. 15(4), 497503 (1970).
50. Moschopedis, S. E., and Speight, J. G., Fuel 55, 187192 (1976).
40. Gelot, A., Friesen, W., and Hamza, H. A., Colloids Surf. 12, 271303
51. Burke, N. E., Hobbs, R. E., and Kashou, S. F., J. Petrol. Technol. No-
(1984).
vember, 14401456 (1990).
41. Menon, V. B., and Wasan, D. T., Colloids Surf. 23, 353362 (1987).
52. Eley, D. D., Hey, M. J., and Symonds, J. D., Colloids Surf. 32, 87
42. Tambe, D. E., and Sharma, M. M., J. Colloid Interface Sci. 157, 244
101 (1988).
253 (1993).
53. Sjoblom, J., Urdahl, O., Borve, K. G. N., Mingyuan, L., Saeten, J. O.,
43. Al-Jarrah, M. M. H., and Al-Dujaili, A. H., Fuel Sci. Technol. Int. 7(1),
Christy, A. A., and Gu, T., Adv. Colloid Interface Sci. 41, 241271 (1992).
6988 (1989).
54. Ali, M. F., and Ali, M. A., Fuel Sci. Technol. Int. 6(3), 259290
44. Reynolds, J. G., and Biggs, W. R., Fuel Sci. Technol. Int. 4(6), 749
(1988).
777 (1986).
55. Ali, M. F., Hasan, M. U., Bukhari, A. M., and Saleem, M., Hydrocar-
45. Nghiem, L. X., Hassam, M. S., Nutakki, R., and George, A. E. D., Soc.
bon Process. 64(2), 8386 (1985).
Petrol. Eng. 5, 375384 (1993).
56. Hasan, M. U., Ali, M. F., and Arab, M., Fuel 68(6), 801803 (1989).
57. Boduszynski, M. M., ACS Div. Petrol. Chem. 30, 626640 (1985). 46. Neumann, H.-J., Petrochem. 18, 776779 (1965).
AID JCIS 4807 / 6g24$$$$44 04-29-97 21:49:06 coida