Fuel 79 (2000) 10471055

www.elsevier.com/locate/fuel

Wax deposition from Middle East crudes

A.M. Elsharkawy a,*, T.A. Al-Sahhaf b, M.A. Fahim b
a

Petroleum Engineering Department, College of Engineering & Petroleum, Kuwait University, P.O. Box 5969, Safat 13060, Kuwait
Chemical Engineering Department, College of Engineering & Petroleum, Kuwait University, P.O. Box 5969, Safat 13060, Kuwait

Received 11 May 1999; received in revised form 10 October 1999; accepted 20 October 1999

Abstract
This paper reports measurements of wax content by acetone precipitation techniques as well as wax appearance temperature (WAT) by
viscosity measurements and differential scanning calorimetry (DSC) of eight different stock-tank crude oils from the Middle East. Comparison of WAT measured by DSC and viscosity indicates that the viscosity method overestimates the WAT. Crude oil gravity measured, by
digital density meter, and molecular weight, by vapor pressure osmometer, were used to characterize the plus fraction and predict WAT and
amount of wax formed at a given condition by the thermodynamic model. Comparison between predicted and measured results shows that
measured WAT by DSC compares very well with that predicted from the model for most crudes. Generally, wax contents measured by the
modified UOP method 46-64 and that predicted by the thermodynamic model are in good agreement. 2000 Elsevier Science Ltd. All rights
reserved.
Keywords: Wax; Differential scanning calorimetry; Oil viscosity; Wax appearance temperature; Wax dissolution temperature

1. Introduction
Crude oils produced from the formations are transported
and processed at low temperatures where solid wax particles
may precipitate. These solid particles cause additional pressure drop in production tubing, pipeline, and processing
equipment and eventually cause plugging. Great saving in
the cost of operating such equipment can be achieved from
accurate modeling of wax appearance temperature (WAT)
and the amount of wax formed at given conditions. Crude
oils contain mixtures of light and heavy hydrocarbons that
can be classified as paraffins, naphthenes and aromatics as
heavy as C60 or C70 in addition to polars and asphaltenes [1].
The lighter parts of the crude oils keep the heavier parts,
wax and asphaltene, in solution. This solubility depends on
pressure, temperature and composition of crude oils. The
presence of light ends increase, the solubility of wax in
crude oil. Asphaltene acts as inhibitor for wax formation
[2]. When the temperature of crude oil drops as it occurs
in production tubing of oil wells and pipe lines, the solubility of the heavy fractions may be sufficiently reduced to
cause precipitation of solid particles of wax and asphaltenes.
Two types of wax are commonly encountered in crude oils.
The first is the macrocrystalline wax composed of mainly
* Corresponding author. Fax: 965-484-9558.
E-mail address: asharkawy@kuc01.kuniv.edu.kw (A.M. Elsharkawy).

straight-chain paraffins (n-alkanes) with varying chain

length (about C20 C50). The second is the microcrystalline
or amorphous waxes containing high portion of isoparafins
(cycloalkanes) and naphthenes with a molecular weight
ranges from C30 to C60. The presence of these solid particles
causes a change in the flow behavior of crude oil from
Newtonian to non-Newtonian [3]. Thus, resulting in a
marked increase in crude oil viscosity and hence additional
pressure drop which increases the cost of pumping and
decreases pipeline and tubing efficiency because of potential
plugging [410]. Great saving in the cost of operating
processing equipment, pipeline and production tubing can
be achieved from accurate modeling of WAT and the
amount of wax formed at given conditions. Several methods
have been presented to the industry for coping with wax
related problems [1113]. These methods fall in three
different categoriesthermal, mechanical, and chemical.
All of these methods have their disadvantages; they all
increase operating expenses. Therefore, it is essential to
be able to predict the amount and conditions of wax precipitation in order to reduce operating expenses.
Several models have been presented in the literature for
the modeling of WAT and amount of wax formed at different temperatures [1416]. These models generally overestimate the amount of wax formed below the cloud-point
temperature [17]. Won [18] and Won and Daniel [19] used
the regular solution theory to describe the non-idealities in

0016-2361/00/$ - see front matter 2000 Elsevier Science Ltd. All rights reserved.
PII: S0016-236 1(99)00235-5

1048

A.M. Elsharkawy et al. / Fuel 79 (2000) 10471055

Nomenclature

f
s
r

phase fugacity
solubility parameters
the density at standard condition of normal
paraffins
Gibbs free energy
DGi
DH
enthalpy
DS
entropy
f
fugacity
the standard state fugacity
fioS
MW
molecular weight
p
pressure
R
the gas constant
T
temperature
melting temperature
Tf
V
molar volume
x
phase mole fraction
the mole fraction of wax in pseudozs
component
Superscripts
L
liquid
S
solid
oil and wax phases. Hansen et al. [20] applied the polymer
solution theory of Flory [21] for the description of the oil
phase while the wax phase was assumed to be an ideal. The
cloud points obtained using the model of Won were somewhat higher than those measured. Countinho et al. [22,23]
proposed a model in which the solid state is described by
local composition. Ungerer et al. [24] presented a hypothesis that each component of the heavy fraction of crude oil
can crystallize pure wax leading to several solid wax phases.
To consider the solid phase as an ideal, although used before
by a number of authors, is not justifiable. It is well established that the paraffinic solid phase is highly non-ideal
giving rise to multiple solid phases in equilibrium. Finally,
Pedersen et al. [25] and Pedersen [26] presented a thermodynamic model based on the SRK equation of state to
predict the wax formation at different conditions. Results
reported by Pedersen [26] show that the model successfully
matches the experimental data for wax deposition for the
North Sea crudes. For this reason, the model presented by
Pedersen [26] is considered in this paper to model wax
deposition from Middle East crudes. Hamouda et al. [27]
reported that the wax deposited in the pipeline at a higher
temperature, than those measured in the laboratory since
pipeline wall roughness and/or the presence of nucleation
sites, such as solid, corrosion products, play a great role for
deposition of wax from undersaturated fluids. Models
accounting for molecular diffusion and shear dispersion
were also presented to predict wax formation under dynamic
conditions [28].
Determination of quantity of wax precipitated at different
temperatures depends on the definition of wax and method

of measurement [2932]. Differential scanning calorimetry

(DSC), polarized microscopy, low-resolution pulsed nuclear
magnetic resonance (NMR), and viscosity measurements
have been used to determine WAT for North Sea crude
oils. DSC and NMR have been used to determine the
amount of wax precipitation from crude oils at different
temperatures. The acetone precipitation technique has also
been used to determine the total amount of wax.
This paper is aimed at characterizing Middle East crudes,
measuring the temperature at which the solid wax phase
starts to crystallize and total wax content, and using the
measured data to predict the amount of wax precipitated
at various temperatures and atmospheric pressures.
2. Thermodynamic model
A modification of the thermodynamic model presented by
Pedersen [26] has been used in this study to predict the
WAT and the amount of wax formed at a given temperature.
The model uses the vaporliquidsolid equlibrium for
predicting phase equlibria of oil mixtures taking into
account the possible formation of a wax phase
"
!#
DHif
T
S
s oL
fi xi fi p exp
1 f
1
RT
Ti
may be found
The liquid phase fugacity coefficient foL
i
using the SoaveRecdlichKwong equation of state on
the pure component i at the temperature and pressure of
the system. If the equation of state is the same as that
used to describe the liquidvapor equilibrium this ensures
a consistent representation of the liquid phase with respect
to the types of phase equilibria (vaporliquid and liquid
solid).
To be able to account for the possible formation of the
wax phase, it is necessary to have a procedure for estimating
the melting temperature, enthalpy of diffusion, and the
molar volume. For the defined fractions, these properties
are well documented in the literature. However, for the
plus fraction (C7), these properties are estimated from
correlations. Enthalpy of fusion, Dhfi in cal/mol of component i, is calculated using the following expressions [18]:
Dhfi 0:1462Tif MWi

where MWi is the molecular weight. We used data of nparaffins up to C30 that is collected from literature to correlate the melting temperature Tif in Kelvin and molar volume
Vi ; in cm 3/mol, for components of the C7 fraction
Tif 382:72 20242:593=MWi

Vi 45:865 0:1641MWi lnMWi

Erickson et al. [14] found that the wax forming components

of the plus fraction to be mainly n-paraffins. To split the C7
fraction into several pseudo-components forming the wax, a
procedure is required. In this study, the procedure used by

A.M. Elsharkawy et al. / Fuel 79 (2000) 10471055

1049

out by the viscosity method and DSC. Details of these

measurements are discussed below.

Table 1
Molar composition data for Middle East Crudes
Oil

N2
CO2
H2S
C1
C2
C3
i-C4
n-C4
i-C5
n-C5
C6
C7
C8
C9
C10

0.00
0.03
0.13
0.02
0.34
1.62
0.81
3.28
2.19
4.10
7.19
6.27
5.78
5.33
62.71

0.00
0.05
0.19
0.02
0.51
2.32
1.06
4.09
2.35
4.46
7.37
4.73
4.42
4.15
64.28

0.00
0.10
0.06
0.00
0.73
2.72
0.72
3.59
1.85
3.36
5.92
5.65
5.25
4.89
65.16

0.00
0.01
0.00
0.00
0.38
3.11
0.81
4.71
1.47
3.87
7.90
2.36
2.30
2.24
70.84

0.00
0.06
0.00
1.40
0.28
0.64
0.37
2.33
2.09
3.69
7.02
4.32
4.10
3.88
69.82

0.00
0.02
0.00
0.02
0.46
1.52
0.61
2.43
2.09
3.67
8.03
4.67
4.40
4.16
67.92

0.00
0.04
0.00
0.00
0.81
2.17
0.74
2.32
1.58
2.95
8.24
4.21
3.99
3.80
69.15

3.1. Viscosity measurements

The capillary viscometer was calibrated using standard
calibration fluid. The maximum error in viscosity measurements was calculated as 0.13% for a very viscous fluid
having a viscosity of 1200 cp as shown in Table 2. Before
any viscosity measurements were carried out, the stock-tank
oil sample was first heated to 80C to ensure that all wax
particles are dissolved. The sample was later charged to an
appropriate capillary viscometer and cooled to the desired
temperature. The time required to flow a standard volume
through a selected capillary tube of given size was calculated and the oil viscosity is determined. At each temperature, an average of three runs was recorded and the test is
repeated every 10C from 80 to 10C. The repeatability of
viscosity measurements was calculated to be about 1%.

Pedersen et al. [33] is used. The plus fraction is divided into

several pseudo-components with the carbon number fraction up to C80. The mole fraction zsi of the potentially wax
forming part of the pseudo-component is calculated from
the following expression [26]
"
!c #
ri rpi
s
total
1 A BMWi
5
zi zi
rpi

3.2. Differential scanning calorimetry measurements

WAT during cooling and wax dissolution temperature
(WDT) during heating were measured using the scanning
calorimeter (Mettler Toledo TA 4000 system). DSC was
chosen for measuring the amount of wax because it is superior to Nuclear Magnetic Resonnance (NMR). NMR is not
very sensitive for low amount of wax, i.e. less than 5%.
While, DSC is faster and more accurate for low amount of
wax [10]. It is a standard procedure in most research laboratories that the calibration of DSC for temperature and heat
flow determination is carried out using the melting point and
heat flow of high purity metal or compounds [29,31]. In the
present study, DSC instrument was adjusted to both
temperature and heat flow using pure Indium (see Table
3). A series of high purity normal parrafins were also used
for low temperature adjustment of the DSC in the temperature range of 50 to 95C [30]. These materials were
chosen because they have known melting temperatures
and enthalpies of fusion (see Table 3). Published data for
the melting temperature and melting enthalpy for Indium
were used to program the DSC to draw the baseline. Difference in melting temperature (mp) and enthalpies of fusion
(DH) for normal alkanes between literature data and
measured ones were of the order of 0.33 to 4.16C and

where MWi is the molecular weight in g/mol and ri the

density in g/cm 3 at the standard condition of components
i, and A, B and C are constants (0.8824, 5:353 104 ; and
0.1144, respectively). rpi is the density at the standard condition of normal paraffins with the same molecular weight as
the pseudo-component i, calculated from the following
expression:

rpi 0:3915 0:0675 ln MWi

3. Experimental
The WAT and the amount of wax formed at a given
temperature have been measured for eight stock-tank
crude oils from major producing oil fields in the Middle
East. The compositional analyses of these samples are
given in Table 1. These wax measurements were carried
Table 2
Calibration data for capillary viscometer
Fluid no.

Reported viscosity (cp)

Flow time (s)

Tube no.

Tube constant

Measured viscosity (cp)

Error (%)

1
2
3
4
5
6
7

1200
992
475
95.9
48
9.8
4.5

632.43
361.68
344.12
373.43
472.10
593.62
524.93

600
450
400
300
200
100
75

19
2.745
1.381
0.2575
0.1017
0.01651
0.00857

12016
992.8
475.2
96.2
48.01
9.80
4.50

0.13
0.08
0.04
0.31
0.02
0.01
0.00

1050

A.M. Elsharkawy et al. / Fuel 79 (2000) 10471055

Table 3
DSC measurements on normal paraffins and Indium
Standard sample Published data

Indium
Indium
n-C6
n-C7
n-C8
n-C16
n-C24

DSC measurements

Melting temperature (C)

Melting enthalpy (J/g) Gradient (C/min) Range (C)

156.6
156.6
95.32 a
90.85 a
56.77 a
18.158 d
50.6

28.5
28.5
151.78 b
140.21 b
181.65 c
235.6

10
10
10
10
10
10
10

30180
30180
30 to 120
30 to 120
30 to 100
10120
30120

Onset (C)

DH (J/g)

DT (C)

D (DH) (J/g)

157.8
157.9
99
89
57
17
53.2

27.1
27.4
160.4
156.1
194.8
238.5
263.3

1.2
1.3
3.68
1. 85
0.33
1.158
2.6

1.4
1.1
8.62
15.88
13.23
2.9

Daubert TE, Danner RP. Technical data-book-petroleum refining. 4th ed. Washington, DC: American petroleum institute, extant 1987.
Thermodynamic Research center. Selected values of properties of hydrocarbon and related compounds. American Petroleum Institute research project 44.
Texas A&M University, College Station, TX, loose-leaf data sheet, extant 1980.
c
Stull DR, Westrum EF Jr., Sinke GC. The chemical thermodynamics of organic compounds. New York:Wiley, 1969.
d
Finke hL, Gross ME, Waddington G, Huffman HM. Low temperature thermal data for the nine normal paraffin hydrocarbons from octane to hexadecane. J
Am Chem Soc 1954;76:333.
b

15.88 to 2.4 J/gm, respectively. Based on these results, a

correction factor of 2C was used to correct wax crystallization onset temperature (WAT) and dissolution temperature measured by DSC for crude oils. Prior to any
measurements, the crude oil sample is heated to a temperature of 80C for 60 min and shaken thoroughly to ensure that
all wax crystals have been dissolved. The oil is later transferred to a tarred sample capsule and weighed. The filled
and closed sample is heated and shaken again to 70C. An
aluminum capsule of 40 ml capacity was used for sample
measurements and references. During the cooling process,
the sample was cooled at a rate of 10C/min from 70 to
140C using liquid nitrogen. A cooling rate of 10C/min
might produce an undercooling effect. However, a very low
scanning rate would give reduced sensitivity as the DSC
signal is the time derivative of the heat flow [30]. Therefore
we used a cooling rate of 10C/min. At the end of each run,
the sample compartment was purged with nitrogen gas. The
crystallization onset temperature (WAT) was determined as
the onset of the exothermal peak corresponding to the
liquidsolid transition. Under heating conditions, the
sample was heated from 140 to 80C at a rate of 10C/
min. The dissolution temperature on heating was taken as

A
B
C
D
E
F
G
H

After heating

Wax (%)

Density (@ 15.6C)

Molecular weight

0.8859
0.8964
0.8810
0.8752
0.9495
0.9084
0.8851
0.9026

279
301
288
238
419
324
297
320

3.3. Molecular weight and density of the C7 fraction

Average molecular weight of the C7 fraction was
measured by a vapor pressure osmometer (Knauer model
A0280). The osmometer was calibrated using various
concentrations of benzene in chloroform. A digital
frequency density-meter (model DMA 48) was used to
measure the sample density. The density and the molecular
weight were measured at 15.6C after heating the sample to
70C. The molecular weight and density of the C7 fraction
are reported in Table 4.
3.4. Wax content
The wax content is determined by the modified UOP
method 46-64 (Burger et al. [28]) that is based on precipitation of wax by acetone. We modified the method to remove
fine grain sands, sediments, and asphaltene from crude oil.
The details of the procedure is given in Appendix A.
4. Results and discussion

Table 4
Characterization of Middle East crudes
Oil

the temperature at which the solidliquid exotherm reaches

the baseline. The accuracy of DSC measurements were
calculated to be about /2C.

5.54
9.67
6.50
3.63
14.69
5.02
6.41
7.10

4.1. WAT from viscosity measurements

Viscosity for most pure liquids and many suspensions at
given temperature and pressure is a well-defined property
that is independent of shear stress and velocity gradient
provided the flow is laminar (Newtonian behavior). Fluid/
particle interaction, asymmetry, and orientation of suspension of particles cause deviation from Newtonian behavior.
Fig. 1 shows the plot of viscosity versus temperature for one
of the eight crude oil samples aforementioned. Other crude
oils show similar behavior. Viscosity has an exponential

A.M. Elsharkawy et al. / Fuel 79 (2000) 10471055

1051

Fig. 1. Effect of wax precipitation on viscosity of crude oil.

dependence on temperature. Escobedo and Mansoori [34]

have shown that viscosity measurements can be used to
detect the onsets of colloidal asphaltene flocculation in
paraffinic and asphaltene micellalization in aromatics. The
viscositytemperature in the Newtonian range can be
considered as a simple exponential Arrhenius equation, as
reported by Pedersen et al. [25]:

m A eEa =RT

where m is the Newtonian dynamic viscosity, Ea is viewed

as an activation energy of viscous flow, A is a constant, R is
the Universal gas constant, and T is the temperature. Realizing the fact that oil viscosity obeys the exponential Arrhenius equation until solid particles of wax starts to
precipitate, wax precipitation temperature can be determined from the viscosity temperature plot such as Fig. 1.
4.2. DSC measurements
4.2.1. WAT and WDT
Figs. 2 and 3 show DSC results during cooling and heating temperature scanning in the temperature range of 140
to 70C, respectively, for one of the eight crude oil samples.
Fig. 2 is a typical thermogram showing the exothermal peak
during cooling, the glass transition temperature, Tg, and the
processing baseline. However, Fig. 3 shows the endothermal peak during heating. Most DSC curves show the sharp
spike of the endothermal peak and/or exothermal peak for
Middle East crudes, except crude oil sample E. The exothermal peak corresponds to the crystallization and hence precipitation of solid particles of wax during the cooling process.

The endothermal peak corresponds to dissolution or melting

of all wax crystals during the heating process. DSC curve of
oil sample E shows the baseline intersecting the exotherm.
Similar phenomenon was reported by Hansen et al. [30].
Table 5 shows the WAT, sometimes referred to as the
onset crystallization temperature, the WDT, and the glass
transition temperature (Tg) for the Middle East crudes that
are described in this paper. WAT is taken as the temperature
at which the first crystal of wax appears. On the DSC curve,
it is taken as the onset temperature of exthothermal peak.
During the cooling process, the solvating power of the oil
matrix decreases resulting in precipitation of solid particles
of wax. The Middle East crudes described in this study have
a WAT ranging from 32 to 38C. Further cooling of the
crude oil sample in the range of 100 to 150C results
in crystallization of the oil matrix into solid material like
glass. This transition is marked by increase in viscosity and
decrease in heat capacity. The temperature at which this
transition occurs is defined as glass transition temperature
Tg. Crude oil considered in this study has a glass transition
temperature ranging from 106 to 148C. Above Tg a
wax crystal and liquid oil matrix exist. Below, Tg only a
crystalline wax phase and glassy oil matrix exist. Our
experimental results show no correlation between Tg and
WAT or between Tg and wax content. This is in agreement
with Hansen et al. [30].
It is clear from Table 5 that the WDT is always 10 to 29C
higher than the wax precipitation temperature. Rnningsen
et al. [3] reported that WDT is 528C higher than WAT for
North Sea crudes. The difference between WAT and WDT
might be caused by undercooling and overheating that result

1052

A.M. Elsharkawy et al. / Fuel 79 (2000) 10471055

Fig. 2. DSC thermogram of oil F during cooling.

in non-equilibrium conditions during fast temperature scanning (10C/min.). To avoid this, Hansen et al. [30] recommended using very low temperature scanning. In this study,
however, it was found that the low temperature scanning
reduces the DSC sensitivity. Another reason for the difference between WAT and WDT might be the way the processing baseline is drawn.
WDT is taken as the temperature at which all the precipitated wax has been dissolved into the oil matrix during the
heating process, as shown in Fig. 3. It is the temperature
corresponding to the endothermal peak on the DSC curve
during heating temperature scanning. The crude oils
described in this study have a WDT ranging from 42 to
67C.
4.2.2. Enthalpy of precipitation and melting
Table 5 also shows measured enthalpy of precipitation
and enthalpy of dissolution for the eight crude oil samples.
The total energy released during cooling or heating process
is proportional to the area between the base line and the
exothermal peak or endothermal peak, respectively. Middle
East crudes have enthalpy of precipitation ranging from

47.5 to 66.3 J/g and melting enthalpy from 60.5 to 78.6 J/g.
Generally, the dissolution enthalpy is higher than the
corresponding enthalpy of precipitation. This enthalpy
difference could not be accounted for thermodynamically.
It must be the result of experimental uncertainty. Differences between enthalpy of precipitation and dissolution
may reflect the differences in oil composition (i.e. molecular
weight and molecular structure) and their waxy content.
Linear relationship between transition enthalpies and paraffin
content has been observed [30]. Our results do not show
such relationships.
4.2.3. Comparison of measured and predicted results
Table 6 shows comparison between WAT from viscosity
measurements, DSC measurements, and that predicted from
the thermodynamic model. The WAT from viscosity
measurements is 28C higher than that of WAT from
DSC measurements for all Middle East crudes, except
crude oil C. Uncertainty in measurements of WAT as high
as 10F (6C) was reported by Hunt [11]. Other investigators
found similar differences between WAT from viscosity and
DSC. Measured WAT by DSC compares with those

Table 5
DSC measurements on Middle East crudes
Oil

WAT (C)

Enthalpy of precipitation (J/g)

WDT (C)

Enthalpy of dissolution (J/g)

Glass transition temperature (C)

A
B
C
D
E
F
G
H

36
38
33
38
36
37
32
33

66.3
61.0
58.3
66.2
60.3
47.5
59.4
51.1

51
67
60
53
50
56
42
45

77.1
78.6
70.7
72.5
66.3
60.5
64.9
62.2

148
132
127
138
111
106
145
133

A.M. Elsharkawy et al. / Fuel 79 (2000) 10471055

1053

Fig. 3. DSC thermogram of oil F during heating.

predicted by the thermodynamic models. Maximum

difference between predicted and measured WAT by
DSC was in the order of 7C. Fig. 4 shows comparison
between measured and predicted WAT and wax content
for one of the eight crude oils considered in this paper.
WATs compare very well with that predicted from the
model for most crudes. However, wax contents
measured by modified UOP method 46-64 and that
predicted by the thermodynamic model are in good
agreement for four samples out of the eight samples
considered in this paper. This might be due to entrapment of liquid in solid residue.
Some investigators have found correlation between
WAT, pour point and wax content. WAT increases with
increasing wax content. However, results reported by
Rnningsen et al. [3] showed no correlation between minimum pour point and wt.% wax at minimum pour point. Our
results also show no correlation between WAT and wax
content.

5. Conclusions
In this study, eight stock-tank crude oil samples from major
producing oil fields in the Middle East have been characterized. The characterization includes measurements of crude oil
viscosity to estimate WAT. It also comprises DSC measurements of WAT, WDT, enthalpy of precipitation and dissolution, glass transition temperature and total wax content.
Our measurements indicate that the WDT is somewhat
higher than the WAT for all of the crudes. No correlation
was found between glass transition temperature and total
wax content of the crude. However, enthalpy of precipitation of wax during cooling is always lower than the
corresponding dissolution enthalpy during the heating
process. Comparison of WAT from viscosity measurements and DSC indicates that the former is always
higher. Thus, it is more appropriate to use WAT from
viscosity measurements for designing of pipe lines and
production equipment.

Table 6
Measurements and prediction of WAT, dissolution temperature (WDT) and wax content
Viscosity method

DSC method

Predicted

UOP 46-64 modified

Oil

WAT (C)

WAT (C)

WDT (C)

WAT (C)

Wax (%)

Wax (%)

A
B
C
D
E
F
G
H

36
33
37
43
38
45
36
39

36
38
34
38
36
37
32
33

51
67
60
53
50
56
42
45

37.3
38.8
39.8
31.0
39.4
39.5
40.2
40.0

5.6
5.06
6.4
5.3
1.6
4.1
6.1
4.6

5.5
9.6
6.5
3.6
14.6
5.0
6.4
7.1

1054

A.M. Elsharkawy et al. / Fuel 79 (2000) 10471055

Fig. 4. Measured and predicted weight % wax at atmospheric pressure.

WAT and wax content predicted from the thermodynamic model compares well with our measurements for
most crudes.

Appendix A
Measurement of wax content
1. Crude oil samples were heated to 70C to ensure
complete dissolution of all solid phases.
2. The oil was then mixed with pentane (1:40, v/v), left
overnight, and filtered to remove asphaltene and other
sediments.
3. A representative sample, 5 gm for example, was taken
into a tarred bottle of suitable size.
4. A given volume of petroleum ether (35 cm 3) was added
to the sample and stirred until the sample is thoroughly
dissolved.
5. A given volume of acetone (about 110 cm 3) was added
and stirred well.
6. The sample was placed into a deep freeze at 20 to
30C and allowed to come to temperature for about 2 h.
7. The following items were pre-cooled to 20C: Buchner
porcelain filtering funnel, Whatman No. 934 glass fiber
filters, vacuum flask, and a mixture of three parts acetone
plus one part ether.
8. Before filtering the cold sample/solvent mixture, the filter
was seated in the filter funnel by wetting the filter with
the cold solvent mixture and evacuating the assembled

apparatus. The sample was filtered by pouring it slowly

into the funnel, using the stirring rod as a guide. The
stirring rod, bottle, and filter cake were washed well
with the cold solvent mixture. The vacuum was disconnected, the filter was removed with a forceps, and placed
in the original tarred bottle. The wax crystals in the filter
funnel and the stirring rod were washed into the bottle
with toluene.
9. The toluene was evaporated to dryness and then the
bottle was re-weighed. The difference between the tar
and the final bottle weight, minus the weight of the filter
used, is the weight of the wax crystals contained in the
original 5 g sample.
References
[1] Mazee WM. In: Hobson GD, Pohl W, editors. Modern petroleum
technology, 4. Barking: Applied Science Publishers, 1973. p. 782
803.
[2] Misra S, Baruah, Singh KP. Paraffin problems in crude oil production
and transportation: a review. SPE Prod Facilities 1995;February:50
54.
[3] Rnningsen HP, Bjrndal B, Hansen AB, Pedersen WB. Wax precipitation from North Sea crude oils. 1. Crystallization and dissolution
temperature, and Newtonian and non-Newtonian flow properties.
Energy and Fuels 1991;5:895908.
[4] Bern PA, Withers VR, Cairns JR. Wax deposition in crude oil pipelines. Paper EUR 206 presented at the European Offshore Petroleum
Conf. & Exhib. London, 2124 October 1980. p. 57178.
[5] McClaflin GG, Whitfill DL. Control of paraffin deposition in production operations. SPE paper no. 12204 presented at the 58th Ann. Tech.

A.M. Elsharkawy et al. / Fuel 79 (2000) 10471055

[6]

[7]
[8]
[9]

[10]

[11]
[12]

[13]

[14]

[15]

[16]
[17]
[18]

[19]
[20]

Conf. & Exhibit. of the Soc. Petrol. Eng. San Francisco, CA, 58
October 1983.
Weingarten JS, Euchner JA. Methods for predicting wax precipitation
and deposition. SPE paper no. 15654, presented at the 61st Ann. Tech.
Conf. & Exhibit. of the Soc. Petrol. Eng. New Orleans, LA, 58
October 1986.
Bott TR, Gudmiundsson R. Deposition of paraffin wax from flowing
systems. Institute of Petroleum. Report No. IP 77-007, 1989.
Carnahan NF. Paraffin deposition in petroleum production. J Petrol
Technol 1989;October:10245.
Agrawal KM, Khan HU, Suriananarayanan M, Joshi GC. Wax deposition of Bombay high crude oil under flowing condition. Fuel
1990;69:7946.
Calange S, Meray VR, Behar E, Malmaison R. Onset crystallization
temperature and deposit amount for waxy crudes. SPE paper no.
37239 presented at the Int. Symp. on Oilfield Chem. Houston, TX,
February, 1997.
Hunt A. Uncertainties remain in predicting paraffin deposition. Oil
Gas J 1996;7(29):96103.
Bernadiner MG. Advanced asphaltene and paraffin control technology. SPE paper no. 25192 presented at the Int. Symp on Oilfield
Chem. New Orleans, LA, 25 March 1993.
Ferworn KA, Hummami A, Ellis H. Control of wax deposition: an
experimental investigation of crystal morphology and an evaluation
of various chemical solvents. SPE paper no. 37240 presented at the
Int. Symp. on Oilfield Chem. Houston, TX, 1821 February 1997.
Erickson DD, Niesen VG, Brown TT. Thermodynamic measurement
and prediction of paraffin precipitin in crude oil. SPE paper no. 26604
presented at the 68th Ann. Tech. Meet. & Exhibit. of the Soc. Petrol.
Eng. Houston, TX, 36 October 1993.
Niper TC. Thermodynamic modeling for organic solid precipitation.
SPE paper no. 24851, presented at the 67th Ann. Tech. Conf. &
Exhibit. of the Soc. Petrol. Eng. Washington, DC, 47 October 1992.
Majeed A, Bringedal B, Overa S. Model Calculates wax deposition
for N. Sea Oils. Oil Gas J 1990;6(18):6369.
Lira-Galeana C, Firoozabadi A. Thermodynamics of wax precipitation in petroleum mixtures. AIChE J 1996;42(1):248329.
Won KW. Thermodynamic calculation of cloud point temperature
and wax phase composition of refined hydrocarbon mixtures. Fluid
Phase Equilibria 1989;53:37796.
Won KW, Daniel F. Thermodynamic model of liquidsolid equilibria
for natural fats and oils. Fluid Phase Equilibria 1993;82:26173.
Hansen JH, Fredenslund A, Pedersen KS, Rnningsen HP. A thermo-

[21]
[22]

[23]
[24]

[25]

[26]
[27]

[28]
[29]

[30]

[31]

[32]

[33]
[34]

1055

dynamic model for predicting wax formation in crude oils. AIChE J

1988;34(12):193742.
Flory PJ. Principle of polymer chemistry. Ithaca, NY: Cornell University Press, 1953.
Countinho JAP, Andersen SI, Stenby EH. Evaluation of activity coefficient models in prediction of alkane solidliquid equilibria. Fluid
Phase Equilibria 1995;103:2339.
Countinho JAP, Knudsen K, Andersen SI. A local composition model
for paraffinic solid solution. Chem Engng Sci 1996;51(12):327382.
Ungerer P, Faissat, Leibovic C, Zhou H, Behar E, Moracchini G. High
pressure-high temperature reservoir fluids: investigation of synthetic
condensate gases containing a solid hydrocarbon. Fluid Phase Equilibria 1995;111:287311.
Pedersen SK, Skovborg P, Hans PR. Wax precipitation from North
Sea Crude oils. 4. Thermodynamic modeling. Energy and Fuels
1991;5:92432.
Pedersen KS. Prediction of cloud point temperature and amount of
wax precipitation. SPE Prod Facilities 1995;February:4649.
Hamouda AA, Viken BK. Wax deposition mechanism under highpressure and in presence of light hydrocarbons. SPE paper no. 25189
presented at the Int. Symp. on Oilfield Chem. New Orleans, LA, 25
March 1993.
Burger ED, Perkins TK, Straggler jH. Studies of wax deposition in the
Trans Alaska pipeline. J Petrol Technol 1981;6:107586.
Claudy p, Letoffe JM, Chague B, Orrier J. Crude oils and their distillates: characterization by differential scanning calorimetry. Fuel
1988;67(1):5861.
Hansen AB, Larsen E, Pedersen WB, Nielsen AB, Rnningsen HP.
Wax precipitation from North Sea crude oils. 3. Precipitation and
dissolution of wax studied by differential scanning calorimetry.
Energy and Fuels 1991;5:91423.
Letoffe JM, Claudy P, Kok MV, Garcin M, Volle JL. Crude oils:
characterization of waxes precipitated on cooling by d.s.c. and thermomicroscopy. Fuel 1995;74(6):8107.
Pedersen WB, Hansen AB, Larsen E, Nielsen A, Rnningsen HP.
Wax precipitation from North Sea crude oils. 2. Solid-phase content
as function of temperature determined by pulsed NMR. Energy and
Fuels 1991;5:90813.
Pedersen KS, Thomssen P, Fredenslund Aa. Characterization of gas
condensate mixtures. Adv Thermodynamics 1989;1:13752.
Escobedo J, Mansoori GA. Viscsometric principle of onsets of colloidal asphaltene flocculation in paraffinic oils and asphaltene micelization in aromatics. SPE Prod Facilities 1997;May:11622.