SPE 58724

OCCAM: Onset-Constrained Colloidal Asphaltene Model

Sebastiano Correra, SPE, EniTecnologie S.p.A., and Francesca Donaggio, EniTecnologie S.p.A.

Copyright 2000, Society of Petroleum Engineers Inc.

Introduction
This paper was prepared for presentation at the 2000 SPE International Symposium on Asphaltene deposition is a well known problem in oil
Formation Damage held in Lafayette, Louisiana 23–24 February 2000.
production. Typically, asphaltene deposits affect the
This paper was selected for presentation by an SPE Program Committee following review of production of oil from the wellbore to the surface facilities and
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to may also cause inconveniences during storage and pipeline
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at transportation. Asphaltenes may also deposit during enhanced
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
oil recovery processes, especially when miscible flooding with
for commercial purposes without the written consent of the Society of Petroleum Engineers is heavy gases is employed. The economic impact of asphaltene
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous deposition can be relevant; for example, the need for remedial
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. jobs to remove near wellbore formation damage implies
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
downtime and loss of production [1,2]. Last but not least,
asphaltene deposition takes place also downstream in the
Abstract refinery, where it causes fouling of equipments and poisoning
Asphaltene deposits affect the primary production of oil from of catalysts.
the wellbore to the surface facilities and may also cause The need to foresee such problems has prompted many
problems during storage and pipeline transportation. A model researchers to try to explain the asphaltene behavior: different
to predict the occurrence of asphaltene separation is therefore physical pictures of asphaltene precipitation have led to
required in order to avoid or minimize upsets due to several models, based on different approaches.
asphaltene precipitation. Liquid-liquid models. First, Hirschberg et al. [3] proposed
With the aim of identifying the physical nature of the a liquid-liquid equilibrium model based on polymer solution
asphaltene separation, both onset-of-precipitation and thermodynamics. In this model asphaltenes are considered as a
quantity-of-asphaltene experimental data were produced. large macromolecular solute dissolved in a solvent; this
While in general previous researchers have tried to apply or solvent is made up with all the non-asphaltenic components,
adapt well-known models to describe the experimental results, and its properties are evaluated by using proper mixing rules.
the peculiar behavior observed for asphaltenes has been used Asphaltene precipitation is treated as a phase separation
here to gain insight into the mechanism of precipitation. From involving an equilibrium between two liquid phases. Due to
this work a model (Onset-Constrained Colloidal Asphaltene the significant difference in molecular size between
Model) has been developed, allowing predictions of asphaltenes and other components, the Flory-Huggins
asphaltene flocculation from standard PVT data and a limited expression for the free energy of mixing is used. Phase
set of experimental data on stabilized oil. Because it is separation implies the calculation of compositions of both
physically well founded, the model is applicable to any equilibrium phases; Hirschberg further simplifies the
situation in which asphaltene separation is possible; it allows calculations by imposing that a “pure asphaltene phase” is
prediction of both the precipitation onset and the quantity of formed. This model was also used by Burke et al. [4,5],
asphaltene that separates as a function of oil composition, Chaback [6], and de Boer et al. [7].
temperature and pressure. Other models were also based on polymer solution
Model applications show good agreement between predicted thermodynamics:
and experimental data on stock tank oils; moreover upstream ♦ Hirschberg in a following paper [8] added to his
applications are shown which further confirm model results. model an aggregation step;
Besides, the model explains numerous experimental ♦ Cimino et al. [9] adopted the hypothesis that a “pure
observations and physical descriptions of many other solvent phase” (instead of a pure solute phase) is
researchers, so that it is further reinforced by the prior separated at the onset of precipitation;
literature.
♦ Nor-Azlan et al [10] used a similar approach, without
A new insight into the phenomenon of asphaltene precipitation
hypothesis on the equilibrium compositions;
was gained and an original model was built, so that most of
♦ Mansoori et al. [11,12,13] adopted a polymer solution
previous experimental data can be included in this framework.
theory, in the framework of continuum
2 S. CORRERA, F. DONAGGIO SPE 58724

thermodynamics. Similar models where used by • Onset of precipitation. A n-paraffin (antisolvent) is

Chung [14,15]. gradually added to the oil (or to the mixture containing
Still other models describe separation as a liquid-liquid asphaltenes) and the separation start-up is detected.
phase equilibrium, but without using a free energy model The onset corresponds to the quantity of antisolvent
[16,17]. needed to begin the separation.
Solid-liquid models. There are several solid-liquid • Amount of asphaltenes. In this case a quantity of
equilibrium models used to describe both asphaltene and wax antisolvent higher than the onset is added to the
precipitation; in these frameworks, asphaltene separation is mixture. The precipitate is then separated (by filtration
described as a general solid-liquid phase splitting [18-25]. or centrifugation), “washed” and dried.
Thermodynamic-colloidal model. Park et al. [26,27] and We decided to use onset-of-precipitation data because
Leontaritis et al. [28,29] proposed a colloidal model to essential information could be lost in separation and washing
calculate the conditions for asphaltene precipitation; this of asphaltenes in quantity-of-asphaltene measurements [41].
approach is based on a physical interpretation of asphaltene Many researchers have used onset of precipitation experiments
stabilization by resins first proposed by Pfeiffer et al. [30]. In to check asphaltene stability [42-48,50,51] and found
this model asphaltenes are described as insoluble solids empirical relationships that are equivalent to the equation used
dispersed in the surrounding medium by the resins adsorbed in previous studies [9,49]:
on their surface. An equilibrium between asphaltene-adsorbed
resins and oil-solubilized resins is assumed, and precipitation Y = α0 + α1 X.................................................................. (1)
is hypothesized to occur when resins desorb from the
asphaltene surface. Chung [15] compared this model with a Where X = (mass of diluent)/(mass of oil);
polymer solution model, without reaching unequivocal results. Y = (mass of precipitant)/(mass of oil).
Later, Mansoori et al. developed a fractal aggregation
model [31,32,33] to describe asphaltene aggregation steps and α0 and α1 are two coefficients which depend on the
following the onset of precipitation. diluent/precipitant pair.
A further model was developed by Takhar et al. [34] that
appears to be similar to the thermodynamic-colloidal one; These empirical relationships were derived from
however only limited information has been reported for it. experimental results and describe a peculiar behavior: the
Scaling model. Rassamdana et al. [35,36] proposed a onset of precipitation do not depend of asphaltene
model which, rather than investigate the physical nature of the concentration. Buckley [50,51] identified this behaviour as a
system, uses some scaling concepts. In this way they colloidal one and developed a model such that the onset
succeeded in describing the experimental behavior of oils with happens when asphaltene-solvent interactions attain a critical
many different precipitants. value. These interactions were espressed by means of the
Micellisation model. A micellisation model has also been Flory dimensionless interaction parameter.
presented [37-40]: according to these authors, an equilibrium
exists between asphaltene micelles and asphaltene molecules, The physical description. From the experimental
dissolved in the oil. Asphaltene molecules have an extremely observations it is possible to extract three main considerations:
low solubility, so the separation happens when the equilibrium 1. In the literature we have found complex physical models
tends to disaggregate the micelles. to describe a simple empirical behavior (see equation 1).
2. There is an observed behavior on which a general
The existence of such a high number of different models agreement was reached: while doing titration experiments,
shows that the true nature of the phenomenon still must be the asphaltene precipitation onset does not depend of the
understood: there is not yet a model which fully explains the initial concentration [9,42,43,49-51]. Equation 1 is a way
observed behavior. Often, instead of trying to understand the to express this behavior.
physical phenomenon, “available” models (like, e.g., phase 3. All authors dealing with the structure of asphaltene
equilibrium models) are used without a critical analysis of the molecules agree in identifying asphaltenes as compounds
experimental observations. made of a polyaromatic sheet with paraffinic tails. This
Therefore, there is a strong need for a physically-founded structure led many authors to identify asphaltenes as
method for predicting asphaltene deposition, able to correctly sterical stabilized colloids [26-29,32,33]. According to
describe the experimental behavior; the aim of the present them, asphaltene colloidal particles are peptized by
work is to present such a model. adsorbed resins and separation happens when resins
desorb.
The physical model
Experimental observations. Two kinds of experimental data The independence of the onset on concentration is
are reported in the literature: onset of precipitation data and characteristic of many colloidal systems; the steric
quantity of asphaltene data. stabilisation model is one proposal for why this behaviour is
found. As reported in references about colloidal dispersions
SPE 58724 OCCAM: ONSET-CONSTRAINED COLLOIDAL ASPHALTENE MODEL 3

[52], sterically stabilized colloidal systems are stabilized by By substituting eq. 3 in eq. 2 we obtain the model for the
means of chains that do not interpenetrate. By adding an onset of asphaltene flocculation:
antisolvent the overall solvent mixture gets worse, and
(v1/RT) (δ2 − δ1)2 = χθ ................................................... (4)
flocculation occurs when the theta point (of paraffinic tails) is
attained. The theta point is defined as the condition for which Of course properties of pseudo-component 1 are to be
the side-chain/solvent interactions become equivalent to the evaluated as average properties:
side-chain/side-chain interactions: the side-chains can v1 = zovo+ zdvd+ zpvp........................................................ (5)
interpenetrate and flocs are formed. This point only depends
of the solvent quality, and does not depend of the solute
concentration. δ1 = φoδo+ φdδd+ φpδp ....................................................... (6)
So we can explicate the behavior at the onset with this
model: the onset of precipitation corresponds to the
attainment of the asphaltene theta point. where vi = molar volume of component i;
The previously reported “steric-stabilization” models zi = molar fraction of component i;
correctly described the stabilizing mechanism, while they were δi = solubility parameter of component i;
not correct in describing the destabilization: asphaltene onset φi = volumetric fraction of component i.
corresponds to the formation of flocs due to the attainment of
theta conditions and not to the resins desorption from and, with regard to indexes,
asphaltenic micelles. d = of diluent (solvent added)
On the other hand the Buckley model [50,51] now turns o = of the oil;
out to be physically justified: this model is the correct one, as p = of the paraffin (antisolvent);
theta conditions correspond to the attainment of a certain χ 1 = of the overall solvent mixture;
value. 2 = of asphaltenes

The onset model. In order to explain the model, hereafter we Till now we have described the model, that was also
refer to a titration experiment: an antisolvent (a n-paraffin) is previously empirically employed by others [50,51,53,54]; in
added to a stock-tank oil. the following we will demonstrate that such a model can give
We decided to use a pseudo-binary model: asphaltenes are rise to straight lines like eq. 1. We must demonstrate that
considered as a single component (component 2), dissolved in equation 1 is equivalent to equation 4; in order to do so, we
another single component (component 1). The non-asphaltene substitute eqs. 5 and 6 into eq. 4. With some algebraic steps, it
component (component 1) is a mixture of oil, antisolvent, and is possible to obtain the following equation:
other “asphaltenic” molecules, which do not separate at the aYY Y2 + (aXY X + aY) Y + (aXX X2 + aX X + a1) = 0........ (7)
onset. Thermodynamic properties of such a mixture are
average properties: the solubility parameter is volume- The factors used are explained in the Appendix.
averaged, while the molar volume is mole-averaged. The equation 7 is not linear, but if numerical values are
The θ (or onset) condition can be expressed as substituted to the variables a linear trend in the range of
practical interest (0 < X < 10) is observed. It is possible to
χ = χθ .............................................................................. (2) obtain the straight line equation by means of an approximate
where method:
χ = Flory dimensionless interaction parameter; Intercept: by putting X=0 the intercept is obtained as:
χθ = interaction parameter at theta conditions. α0 = (-aY + (aY2 – 4a1aYY)1/2)/(2 aYY) ............................. (8)
A generally accepted value for χθ is 0.5, deriving from the Slope: when X>1 the 2nd order term (in eq. 7) is dominant, so
Flory-Huggins theory for polymeric solutions. We could either that the slope can be calculated as
keep this value or use χθ as a pivoting parameter to better fit
the experimental data; as the latter choice would result in an α1 = (-aXY + (aXY2 – 4 aXXaYY)1/2)/(2 aYY) ...................... (9)
ill-posed estimate problem, we preferred the former. We have thus shown how equation 4 can give rise to the
Now, χ can be expressed in terms of solubility parameters observed straight lines.
by mean of the expression
χ = (v1/RT) (δ2 − δ1)2 ...................................................... (3) Amount of precipitate
As already said, the experimental evidence on the onset of
precipitation led us to identify the separation as a colloidal
where R = gas constant; phenomenon, i.e. the precipitation of a sterically stabilized
T = absolute temperature; dispersion. This means that the “precipitating” phase is
v1 = solvent molar volume; constituted only by asphaltenes. When the titration with
δ2 = asphaltene solubility parameter; antisolvent is carried out after the onset, the precipitation goes
δ1 = solvent solubility parameter. on and higher and higher quantities of asphaltenes separate
4 S. CORRERA, F. DONAGGIO SPE 58724

[35,36], until complete asphaltene separation is accomplished. The OCCAM model

This behavior is due to the asphaltenes polydispersion; and
thus the asphaltenes in the oil have a distribution of solubility Onset model. Equation 2 represents the θ condition which is
parameters, rather than a single value. established at the onset of asphaltene precipitation. It can be
They could be described by a unique distribution function used to predict asphaltene precipitation from live oils of
q(δ2), where (q(δ2)dδ2) is the quantity of asphaltenes with a known composition, at pressures higher than the bubble point
solubility parameter between δ2 and δ2 + dδ2. Now, the (Pb) pressure. When the overall pressure is lower than the
asphaltene titrations are performed by using low solubility Pb pressure, the formation of a triphasic system should be
parameter solvents (n-paraffins), so that δ1 diminishes; by considered. As formerly proposed by other authors [4], in
looking at eq. 3, this means that the asphaltenes fraction which these cases we assume that it is possible to split the calculation
precipitates at the onset is that with the highest solubility in two steps: knowing the overall composition of the mixture,
parameter, δ2max . we perform flash calculations to find the composition of the
As at the onset the χ value is equal to χθ, the corresponding liquid and vapor phases; then the proposed model is used to
value of δ2 for the asphaltenes precipitating at the onset can be describe the asphaltene status in the liquid phase.
calculated as: The approach is summarized in the following steps:
1. Experimental determination of stock tank oil precipitation
RT
δ max
2 = δ1 + χθ ..................................................... (10) thresholds.
v1 2. Regression, by means of the proposed model, of the
where both v1 and δ1 are obtained considering the amount experimental values to find the two parameters δ2max and
of precipitant added till the onset is reached. δο (highest solubility parameter of asphaltene and
If the titration is continued further, the current values of χ solubility parameter of stock tank oil) at the experiment
exceed the 0.5 value for all the asphaltenes for which the temperature.
following equation is true: 3. Calibration of the equation of state (Soave-Redlich-
Kwong) by using the onset data. In using EOS the oil is
RT
δ 2 > δˆ 2 ; where δˆ 2 = δˆ1 +
χθ .................. (11) represented by means of components and pseudo-
v̂1 components, each one characterized by critical
i.e. all the asphaltenes whose solubility parameter is temperature (Tc), critical pressure (pc) and acentric factor
between δ̂2 and δ2max will precipitate. The symbol “^” (ω). For the pseudocomponents, these parameters can be
evaluated from their boiling point and specific gravity by
identifies the current value, corresponding to the current
using the Lee-Kessler relationships [55]. Of course the
amount of precipitant added during the titration.
boiling point of the heaviest component is not
The quantity of asphaltenes precipitated can be calculated experimentally available, so we decided to use it as a
as: pivotable parameter to tune the EOS: the boiling
δ max
2 temperature of the heaviest component (residue) is such
Q= ∫ q(δ2 )dδ2 ............................................................ (12) that the value of the solubility parameter of the stock tank
δˆ 2 oil is equal to the value found in step 2.
So, if we plot the quantity of precipitate Q versus the The bubble pressure and the density at a certain
temperature (typically the reservoir temperature) are used,
solubility parameter δ̂2 , we should obtain a monotonic curve.
if available, in order to better reproduce the PVT behavior
The previous discussion logically leads to the conclusion of the live oil. These data are matched by modifying the
kij interaction parameters and the Peneloux shifting factors
that by experimentally determining at first the function Q( δ̂2 )
[56]; the components, whose values are modified, are
from a restricted amount of experimental data, it is possible to chosen after a sensitivity analysis.
apply it to calculate the quantity of precipitate in other 4. Calculation of δ2max and v2 values at different
conditions. temperature and pressure conditions. We assumed that
δ2max does not depend on pressure, but depends on
In this paper we initially describe how to perform these
temperature as proposed by Hirschberg et al. [3]. The
calculations, and then validate the procedure with some
numerical coefficient of the Hirschberg relationship can
examples on stock tank oils.
be corrected if onset measurements at various temperature
The key role of the onset (theta conditions) induced us to are available. v2 is considered to be insensitive to both
name the model OCCAM, acronym of Onset-Constrained temperature and pressure.
Colloidal Asphaltene Model. 5. Live oil molar volume and solubility parameter
calculations at different conditions. At the same
conditions of step 4 the solubility parameter and the molar
volume of the oil are evaluated by using the modified
SPE 58724 OCCAM: ONSET-CONSTRAINED COLLOIDAL ASPHALTENE MODEL 5

Soave-Redlich-Kwong EOS [56], calibrated in the third 1. Asphaltenes are precipitated by adding the antisolvent
step. mixture to the stock tank oil (volume ratio 30/1).
6. Calculation of the asphaltene phase boundary. At each T, 2. The whole mixture is ultrasonically mixed and filtered by
p condition χ is calculated with equation 3; if χ > χθ then using a Watman 42 filter. A filtrate and a filter cake are so
at these conditions the asphaltenes are not stable. In such obtained.
a way it is possible to obtain a phase envelope in the T,p 3. The filter cake is vacuum-dried at 100°C and weighed.
plane.

Amount of precipitate model Table 1: PVT data for oils #1 and #2

In a similar way it is possible to use the quantity of Oil #1 Oil #2
precipitate measurements on stock tank oil to calculate the Stock tank API gravity 10.6 41.2
amount of asphaltenes which separates at any condition. Reservoir pressure (bar) 328.6 1060
The steps involved are: Reservoir temperature (°C) 100.0 172.0
1. Experimental determination of the maximum
Live oil bubble pressure (bar) 53.9 177.8
asphaltene solubility parameter (δ2max ) from onset
determinations using eq. 3.
2. Experimental determinations of amount of precipitate Component % wt % wt
H2S 0.000 0.004
Q at various δ̂2 values after the onset. A Q vs.
CO2 2.686 0.378
δ̂2 curve can be drawn. N2 0.014 0.633
3. Calculation of v1 and δ1 at different conditions using Methane 0.135 5.535
the equation of state (see the procedure for onset Ethane 0.044 2.300
calculations). Then the correspondent δ̂2 value can be Propane 0.145 2.624
calculated using equation 11, and the amount of i-Butane 0.062 1.363
asphaltenes Q which precipitate in such conditions can n-Butane 0.243 2.219
be easily derived. i-Pentane 0.196 1.554
n-Pentane 0.241 1.829
Hexanes 0.335 4.653
Experimental methods
The onset of precipitation tests and the amount of Heptanes 0.346 4.974
precipitate evaluation are reported for two Italian stock tank Octanes 0.819 6.107
oils. The model has also been tested on other oils. Nonanes 0.746 5.047
In order to find any possible difference, we selected oils of Decanes 0.947 5.207
different API gravities. Undecanes 0.548 2.835
Dodecanes 0.500 3.072
Materials. The compositional analysis of the live oils are Tridecanes+ 91.993 49.667
reported in Table 1; their API gravity is 10.6 °API for oil #1
and 41.2 °API for oil #2.
The obtained asphaltene quantity is referred to the original
Each oil has been diluted with toluene to some selected
sample weight.
toluene/oil ratios (ranging from 0.8 up to 11), and then titrated
with n-C7.
Results
Onset of precipitation. This technique was already described
Onset of precipitation. Figures 1 and 2 show the onset of
[9,49]: a spectrophotometric technique has been employed to
precipitation results for oil #1 and #2, respectively. They
detect the cloud point of stock tank oils. By means of a
report the data in terms of (mass of precipitant/mass of oil)
peristaltic pump, the oil is recirculated through a UV
ratio (y) versus (mass of diluent/mass of oil) ratio (x); by
spectrophotometer (mod. HP 8152A), where the solution
fitting the experimental points, we obtained straight lines,
absorbance at 800 nm is recorded. An antisolvent is added by
whose equations are also reported in the figures.
an automatic titrator to the asphaltene solution, leading to
From these fittings, the solubility parameters of oils and
phase separation (onset of flocculation) which is detected by
their asphaltenes at the titration temperature (22 °C) have been
means of the recorded absorbance.
obtained, as shown in Table 2.
All experiments were carried out at atmospheric pressure
and 22 °C.

Amount of precipitate. The procedure employed to measure

the quantity of asphaltenes is the following:
6 S. CORRERA, F. DONAGGIO SPE 58724

Table 2: Calculated oil and asphaltene solubility worth to note that it is the only case in which we found an
parameters at 22 °C asphaltene solubility parameter lower than that of the stock
Oil #1 Oil #2 tank oil (see Table 2).
Oil solubility parameter (Hil) 10.03 8.31
Asphaltene solubility parameter (Hil) 9.53 9.78
1100 stability boundary
1000
bubble press.
A more extensive discussion on the experimental linear 900
Reservoir
Conditions
behavior is given in [57]. 800

Such data have been used, together with the physical 700

P (Bar)
600
characterization of the oils, to evaluate the solubility 500
parameters of both live oils and asphaltenes at the 400
experimental temperature, using the data as presented in figs. 1 300

and 2. 200
100
0
0 100 200 300 400 500

14 T (°C)
y = 0.7566x - 0.9434
mass of diluent / mass of oil

12 R2 = 0.9871
10
Figure 3: Oil #2 asphaltene deposition envelope
8
As shown in figure 3, at reservoir temperature Oil #2 it is
6 found to be unstable between 5 and 742 bar; the last one is
4 close enough to the pressure, where asphaltene problems
2 occur, to consider the results satisfactory. Moreover, at room
0 temperature the oil is predicted to be unstable at any pressure
0 5 10 15 20 tested; the prediction is confirmed by the presence of a solid
mass of precipitant / mass of oil asphaltenic residue in such conditions.

Table 3: Amount of precipitate for Oils #1 and #2

Figure 1: Oil #1 experimental data and their fitting
Oil Antisolvent mixture Q δ̂ 2
The SRK equation of state has been tuned using the w% hil
reservoir temperature bubble pressure and density; the residue Components 1st comp.
bubble temperature has been corrected to reproduce the vol.fraction
measured stock tank oil solubility parameter. The so tuned 1 nC5 - Metyl Naphtalene 0.840 10.4 9.18
equation was used to calculate the solubility parameter of the nC5 – Tetraline 0.820 11.6 9.14
live oil at different temperatures and pressures: for each (T, P) nC5 – Toluene 0.800 13.5 9.05
point it is therefore possible to evaluate whether the live oil is nC5 – CycloC6 0.727 19.5 9.00
stable or not. nC7 1.000 20.8 8.89
iC8 - CycloC6 0.852 21.5 8.55
mass of diluent / mass of oil

8 2 nC5 - Metyl Naphtalene 0.840 0 9.18

7 y = 1.2645x + 0.465 nC5 - Tetraline 0.820 0.02 9.14
6 R2 = 0.8618 nC5 - Toluene 0.800 0.11 9.05
5 nC5 - CycloC6 0.727 0.24 9.00
4 nC7 1.000 0.25 8.89
3 iC8 - CycloC6 0.852 0.25 8.55
2
1 Oil #2 has already been tested in ref. 9, where only the
0 stock tank composition was reported, obtaining similar results:
0 1 2 3 4 5 6 this only goes to show that the present and the previous
mass of precipitant / mass of oil models give results which are not in conflict with each other,
even if the present one appears to have a much more sound
Figure 2: Oil #2 experimental data and their fitting physical basis.

Oil #1 is a very heavy oil and, as expected, doesn’t present Amount of precipitate. Besides onset of precipitation,
any asphaltene problem during production. In fact, it turned quantity of asphaltene determinations were performed on the
out to be stable in the whole range investigated. Besides, it is oils #1 and #2. In table 3 the quantity of precipitate are
SPE 58724 OCCAM: ONSET-CONSTRAINED COLLOIDAL ASPHALTENE MODEL 7

reported; for each point we also report the calculated Table 4: statistical parameters of
δ̂ 2 parameter. asphaltene distribution
Oil n. δ2 avg. Std. Dev. δ2
It is worthwhile to note that the experiments were
hil hil
performed by using different antisolvents, to show the
generality of the approach. 1 9.17 0.17
2 9.07 0.03

We did not perform any prediction of the amount of

1.2
precipitate on live oils because at present field data are not
available, thus rendering any comparison impossible.
1

Conclusions
0.8
A colloidal model predicting both the precipitation onset
and the amount of asphaltene separated has been developed. It
Q

0.6
yields predictions of live oil behavior using only a few
0.4
experimental data on stock tank oils, two of which are reported
here.
0.2 The model describes asphaltenes as sterically stabilized
colloids, and is consistent with the empirical observations of
0 many authors about the precipitation onset. The amount of
8.40 8.60 8.80 9.00 9.20 9.40 9.60 precipitate is calculated by taking into account the
δ2
polydispersion of asphaltenes properties.
Figure 4: Oil #1 amount of precipitate Its comparatively simple mathematical form, the need for a
limited amount of experimental data on stock tank oils,
together with its agreement with available experimental data
Figures 4 and 5 show the experimental Q vs. δ̂ 2 curves; on live oils, make this model a powerful tool for predicting
they were fitted by assuming that the q(δ2)distribution is a asphaltene problems in the field.
Gaussian one. The best fit results are shown in the figures as a
continuous line, while the average value and the standard Acknowledgments
deviation of the Gaussian distribution are reported in Table 4. We would especially thank Dr. T. P. Lockhart and Dr. C.
Carniani for their helpful suggestions; Mr. E. Radaelli, Dr. A.
1 .2 Delbianco and Mr. F. Stroppa for their experimental work.

1 References
1. Tambini, M., Galoppini, M.: “Asphaltene deposition monitoring
0 .8 and removal treatments: an experience in ultra deep wells”,
Paper SPE 27822, presented at 1994 European Production
0 .6 Operations Conference and Exhibition, Aberdeen, UK, March
Q

15-17.
0 .4 2. Mansoori, G.A.: “Arterial Blockage in the Petroleum and
Natural Gas Industries”, presented at 1996 Conf. On Controlling
0 .2 Hydrates, Asphaltenes and Waxing, Aberdeen, Scotland, Sept.
16-17.
0 3. Hirschberg, A., deJong, L.N.J., Schipper, B.A., Meijer, J.G.:
8 .4 0 8 .6 0 8 .8 0 9 .0 0 9 .2 0 9 .4 0 9 .6 0 9 .8 0 1 0 .0 0
“Influence of Temperature and Pressure on Asphaltene
δ2
Flocculation”, SPE J. (June 1984) 283-293.
Figure 5: Oil #2 quantity of precipitate 4. Burke, N.E., Hobbs, R.D., Kashou, S.F.: ”Measurement and
Modeling of Asphaltene Precipitation From Live Reservoir
Fluid Systems”, SPE 18273 (1988).
5. Burke, N.E., Hobbs, R.D., Kashou, S.F.: “Measurement and
The monotonic behavior of the experimental curve Q vs.
Modeling of Asphaltene Precipitation”, JPT (Nov 1990) 1440-
δ̂ 2 and its “S” shape using different antisolvents could be an 1447.
indication in favor of the presented model. In fact this shape 6. Chaback, J.J.: “Discussion of Measurement and Modeling of
Asphaltene Precipitation”, JPT (Dec 1991) l519-1520.
can be thought at as the integral of a distribution curve.
7. de Boer, R.B, Leerlooyer, K., Eigner, M.R.P., van Bergen,
A.R.D.: “Screening of Crude Oils for Asphalt Precipitation:
Theory, Pratice, and the Selection of Inhibitors”, SPE 24987,
(1992).
8 S. CORRERA, F. DONAGGIO SPE 58724

8. Hirschberg, A., Hermans, L.: “Asphaltene phase behavior: a 29. Kawanaka, S., Leontaritis, K.J., Park, SJ., Mansoori, G.A.:
molecular thermodynamic model. Characterization of Heavy “Thermodynamic and Colloidal Models of Asphaltene
Crude Oils and Petroleum Residues”, presented at 1984 Flocculation”, ACS Symp. Series 396 (1989) 443-457.
Symposium Intemational, Lyon, France, June 25-27, 492-497. 30. Pfeiffer, J.Ph., Saal, R.N.J.: “Asphaltic bitumen as colloid
9. Cimino, R., Correra, S., Sacomani, P.A., Carniani, C.: system”, J. Phys. Chem. (1940) 44, 139-49.
“Thermodynamic Modeling for Prediction of Asphaltene 31. Janardhan, A.S., Mansoori, G.A., “Fractal nature of asphaltene
Deposition in Live Oils”, SPE 28993 (1995). aggregation”, J.Petr.Sci.& Eng. (1993) 9, 17-27.
10. Nor-Azlan, N., Adewumi, M.A.: “Development of Asphaltene 32. Escobedo, J., Mansoori, G.A.: “Solid Particle Deposition During
Phase Equilibria Predictive Model”, SPE 26905 (1993). Turbulent Flow Production Operations” SPE 29488 (1995).
11. Mansoori, G.A., Jiang, T.S.: “Asphaltene Deposition and its 33. Escobedo, J., Mansoori, G.A.: “Asphaltene and Other Heavy-
Role in EOR Miscible Gas Flooding”, presented at 1985 3rd Organic Particle Deposition During Transfer and Production
European EOR, April 16-18, Roma, Italy. Operations”, SPE 30672 (1995).
12. Kawanaka, S., Park, S.J., Mansoori, G.A.: “The Role of 34. Takhar, S., Ravenscroft, P.D., Nicoll, D.C.A.: “Prediction of
Asphaltene Deposition in EOR Gas Flooding: A Predictive Asphaltene Deposition During Producton Model Description
Technique”, SPE/DOE 17376 (1988). and Experimental Details”, SPE 30108 (1995).
13. Kawanaka, S., Park, S.J., Mansoori, G.A.: “Organic Deposition 35. Rassamdana, H., Dabir, B., Nematy, M., Farhani, M., Sahimi,
From Reservoir Fluids: A Thermodynamic Predictive M.: “Asphalt Flocculation and Deposition: I. The Onset of
Technique”, SPE Res. Eng. (May 1991) 185-192. Precipitation”, AIChEJ (Jan.1996) 42(1), 10.
14. Chung, F., Sarathi, P., Jones, R.: “Modeling of asphaltene and 36. Rassamdana, H., Sahimi, M.: “Asphalt Flocculation and
wax precipitation”, Niper Report 498 (1991). Deposition: II.Formation and Growth of Fractal Aggregates”,
15. Chung, F.T.H.: “Modeling of Heavy Organic Deposition”, Niper AIChEJ (Dec.1996) 42(12), 3318.
Report 555 (1992). 37. Victorov, A.I., Firoozabadi, A.: “Thermodynamic Micellization
16. Rydahl, A.K., Pedersen, K.S.: “H.P.Hjermstad, Modeling of Model of Asphaltene Precipitation from Petroleum Fluids”,
Live Oil Asphaltene Precipitation” presented at the 1997 Spring AIChE Journal (1996) 42(6), 1753-1764.
National AIChE Meeting, Houston, TX, March 9-13 38. Pan, H., Firoozabadi, A.: “A Thermodynamic Micellization
17. Szewczyk, V., Thomas, M., Behar, E.: “Prediction of Model for Asphaltene Precipitation: Part I: Micellar Size and
Volumetric Properties and (Multi-) Phase Behavior of Growth”, SPE 36741 (1996).
Asphaltenic Crudes”, Revue de l’IFP (1998) 53(1), 51-58. 39. Pan, H., Firoozabadi, A.: “Thermodynamic Micellization Model
18. Chung, T.-H.: “Thermodynamic Modeling for Organic Solid for Asphaltene Precipitation from Reservoir Crudes at High
Precipitation”, SPE 24851 (1992). Pressures and Temperatures”, SPE 38857 (1997).
19. MacMillan, D.J., Tackett, J.E. Jr., Jessee, M.A., Monger- 40. Wu, J., Prausnitz, J.M., Firoozabadi, A.: “Molecular-
McClure, T.G.: “A unified Approach to Asphaltene Thermodynamic Framework for Asphaltene-Oil Equilibria”,
Precipitation: Laboratory Measurement and Modeling”, SPE AIChE Joumal (1998) 44(5), 1188-1199.
28990 (1995). 41. Rydahl, A.K., Pedersen, K.S., Hjermstad, H.P.: “Laboratory
20. Lindeloff, N., Heidemann, R.A., Andersen, S.I., Stenby, E.H.: data for live oils: Software for simulation of asphaltene P and T
“A Thermodynamic Mixed-Solid Asphaltene Precipitation region”, presented at 1997 Conference on Controlling Hydrates,
Model”, Petr. Sci. & Techn. (1998) 16(3&4), 307-321. Waxes and Asphaltenes, Aberdeen, UK, Oct. 20-21.
21. Nghiem, L.X., Coombe, D.A.: “Modeling Asphaltene 42. Hotier, G., Robin, M.: “Action da divers diluants sur les produits
Precipitation During Primary Depletion”, SPE 36106 (1997). petroliers: mesure, interpretation et prevision de la floculation
22. Nghiem, L.X., Hassam, M.S., Nutakki, R.: “Efficient Modelling des asphaltenes”, Rev. De l’IFP (1983) 38(1), 101-120.
of Asphaltene Precipitation”, SPE 26642 (1993). 43. Heithaus, J.J.: “Measurement and significance of asphaltene
23. Chung, T.H.: “Thermodynamic Modeling for Organic Solid peptization”, J. Inst. Of Petroleum (1962) 48(458), 45-53.
Precipitation”, Niper Report 623 (1992). 44. Andersen, S.I.: “Asphaltene Analysis by Flocculation Titration-
24. Dukhedin-Lalla, L., Cartdlidge, C., Shaw, J.M.: “Prediction of Fundamental Relations and Applications”, presented at 1998 8th
Phase Behavior of Bitumen/Resid Mixtures”, University of Int. Conf. On Properties and Phase Equilibria for Product and
Toronto, DSS Contract 030SQ.23440-3-93-08, Feb. 1995. Process Design, Noordwijkerhout, NL, Apr 26-May 1.
25. Kamath, V.A., Kakade, M.G., Sharma, G.D.: “An Improved 45. Wiehe, I.A.: “The Oil Compatibility Model and Crude Oil
Thermodynamic Model of Asphaltene Equilibria”, in Asphaltene Incompatibility”, presented at 1999 Int. Conf. On Petroleum
Particles in Fossil Fuel Exploration, Recovery, Refining and Phase Behavior and Fouling, Houston, TX, March 14-18.
Production Processes, M.K. Sharma and F.T. Yen Ed., Plenum 46. Wiehe, I.A.: “Prevention of Fouling by Incompatible Crudes
Press, NY (1994). with the Oil Compatibility Model2, presented at 1999 Int. Conf.
26. Park, S.J., Mansoori, G.A.: “Organic Deposition From Heavy On Petroleum Phase Behavior and Fouling, Houston, TX, March
Petroleum Crudes”, presented at Fourth UNITAR/UNDP 14-18.
conference on heavy crude and tar sands, paper 225 (1989). 47. Wiehe, I.A.: “Mitigation of the Plugging of a Hydrotreater with
27. Park, S.J., Mansoori, G.A.: “Aggregation and Deposition of the Oil Compatibility Model”, presented at 1999 Int. Conf. On
Heavy Organics in Petroleum Crudes”, Energy Sources (1988) Petroleum Phase Behavior and Fouling, Houston, TX, March
10, 109-125. 14-18.
28. Leontaritis, K.J., Kawanaka, S., Mansoori, G.A., “Descriptive 48. Wiehe, I.A., Kennedy, R.J.: “A process for blending potentially
Accounts of Thermodynamic and Colloidal Models of incompatible petroleum oils”, Intemational Patent WO
Asphaltene Flocculation”, presented at 1988 Symp. On 98/26026.
Advances in oil field chemistry, ACS Meeting, Toronto, 49. Cimino, R., Correra, S., Del Bianco, A., Lockhart, T.P.:
Canada, June 5-ll. “Solubility and phase behavior of asphaltenes in hydrocarbon
SPE 58724 OCCAM: ONSET-CONSTRAINED COLLOIDAL ASPHALTENE MODEL 9

media”, in Asphaltenes: Fundamentals and Applications, E.Y. Appendix

Sheu and O.C. Mullins ed., Plenum Press, N.Y., (1995). Factor of equation 7 are calculate by using the following
50. Buckley, J.S.: “Microscopic Investigation of the Onset of expressions
Asphaltene Precipitation”, Fuel Sci. & Tech. Intemat. (1996) 14,
55-74 a1 = [(δ2 - δo)/ρo]2 - Eθ/(Moρo)
51. Buckley, J.S., Hirasaki, G.H., Liu, Y., Von Drasek, S., Wang, J-
X., Gill, B.S.: “Asphaltene Precipitation and Solvent Properties aX = 2 [(δ2 - δo)/ρo] [(δ2 - δd)/ρd] - Eθ[1/(Moρd) + 1/(Mdρo)]
of Crude Oils”, Petr. Sci. & Technol. (1998) 16(3&4), 251-285.
52. Napper, D. H.: Polymeric stabilization of colloidal dispersions, aY = 2 [(δ2 - δo)/ρo] [(δ2 - δp)/ρp] - Eθ[1/(Moρp) + 1/(Mpρo)]
Academic Press, London, UK (1983).
53. Savastano, C., Meli, S.: “Polymer solubility theory predicts aXY =2 [(δ2 - δd)/ρd] [(δ2 - δp)/ρp] - Eθ[1/(Mdρp) + 1/(Mpρd)]
efficiency of residual oil extraction by polar solvents”, Fuel
Science and Technology International (1993) 11(3&4), 445-461.
54. Laux, H., Rahimian, I., Schorling, P.: “The Colloidal Stability of aXX = [(δ2 - δd)/ρd]2 - Eθ/(Mdρd)
Crude Oils and Crude Oil Residues”, Petroleum Science and
Technology (1999) 17(3&4), 349-368. aYY = [(δ2 - δp)/ρp]2 - Eθ/(Mpρp)
55. Kesler, M.G., Lee, B.I., “Improve Prediction of Enthalpy of
Fractions”, Hydroc. Proc. (1976) 53, 153. Eθ = χθ RT
56. Soave, G.,: “Application of equation of state and the theory of
group solutions to phase equilibrium prediction”, Fluid Phase
Equil. (1993) 87, 23-25.
57. Donaggio, F., Correra, S., Lockhart, T.P.: “Onset of Asphaltene
Precipitation: Toward the Unification of the Empirical Models”,
presented at 1999 3rd Int. Symp. On Colloid Chemistry in Oil
Production, Huatulco, Oaxaca, Mexico, Nov. 14-17