Available online at www.sciencedirect.

com

Fuel 87 (2008) 12411248

www.fuelrst.com

Measuring Salinity in crude oils: Evaluation of methods and

an improved procedure
Montserrat Fortuny a,b, Elisangela B. Silva b, Agnaldo C. Filho a,b, Rosana L.F. V. Melo b,
Marcio Nele c, Raquel C.C. Coutinho d, Alexandre F. Santos a,b,*
b

a
Programa de Engenharia de Processos/Universidade Tiradentes, Av. Murilo Dantas 300, Aracaju 49032-490 SE, Brazil
Laboratorio de Engenharia de Petroleo/Instituto de Tecnologia e Pesquisa, Av. Murilo Dantas 300, Aracaju 49032-490 SE, Brazil
c
Escola de Qumica/Universidade Federal do Rio de Janeiro, Cidade Universitaria, Rio de Janeiro 21945-970 RJ, Brazil
d
CENPES/PETROBRAS, Cidade Universitaria Q.7, Rio de Janeiro 21949-900 RJ, Brazil

Received 1 February 2007; received in revised form 26 June 2007; accepted 12 July 2007
Available online 7 August 2007

Abstract
In this work, a novel procedure is developed to determine the amount of salt for crude oils in which the available laboratory techniques are unable to provide accurate salt content values. The technique is based on measuring the salinity of the wash water that is
vigorously mixed with the crude oil (forming an emulsion) and then separating it from the water-in-oil emulsion by irradiating microwave energy during few minutes. This three step process involves the emulsication of the wash water exempt of salt in crude oil, followed by the recovery of the water by microwave heating and, nally, the analysis of the salt content using the classical Mohrs method
titration procedure. The novel technique presented here was able to provide reliable information about the salt content on Brazilian
crude oils, in a rapid manner, without need of unsafe time-consuming solvent extraction procedures.
2007 Elsevier Ltd. All rights reserved.
Keywords: Salt content; Crude oils; Emulsions; Desalting; Microwave

1. Introduction
In the petroleum industry, there is a great variation in
the salt content of crude oils depending mainly on the
source and, possibly, on the producing wells or zones
within a eld. The amount of mineral salts varies with the
geologic formation and can be as high as 200,000 ppm
[1]. In addition, at the renery, salt water introduced during shipment by tanker may contribute to the total salt
content [2]. In almost all cases, the salt content of the crude
oil consists of salt dissolved in small droplets of water that
are dispersed in the crude [3]. The chemical composition of
these salts varies, but the major portion is nearly always
*
Corresponding author. Address: Programa de Engenharia de Processos/Universidade Tiradentes, Av. Murilo Dantas 300, Aracaju 49032-490
SE, Brazil. Tel.: +55 79 32182115; fax: +55 79 32182190.
E-mail address: [email protected] (A.F. Santos).

0016-2361/$ - see front matter 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.fuel.2007.07.013

sodium chloride with lesser amounts of calcium and magnesium chlorides. In fact, the presence of salt in the crude
oil leads to several problems during transporting and the
rening processes, including corrosion of lines, fouling,
and also the deactivation of catalysts employed at the renery. As a consequence, the renery sites are generally
equipped with desalting systems which are able to perform
the desalination of the crude oil sent to the renery. The
desalting is considered a critical operation at the renery
due to the importance of meeting the specications of the
acceptable quantities of salt and water in the treated oil.
For these reasons, measurements of salt and water content
in crude oils are very important in all oil industry operations including crude oil production, processing, transportation and rening.
Concerning water content measurements, several procedures have been described in the literature including on-line
monitoring devices and laboratory test methods, as

1242

M. Fortuny et al. / Fuel 87 (2008) 12411248

reviewed by Pal [4]. Depending upon the facilities available,

standard laboratory methods may be applied which can
involve DeanStark distillation (ASTM D-4006) [5], centrifuge separation [6], Karl Fischer titration [7], among others. Dierent meters have been found on the petroleum
literature for on-line monitoring of water content on pipelines or vessels. The main principles behind these meters
include generally electrical, microwave, radiation scattering, spectroscopic, ultrasonic, or density (or viscosity)
based techniques [4]. Most of them detect water by a measurement of an electric property of the uid mixture,
including capacitance, conductance or dielectric properties
[8]. In particular, one of the most popular sensing devices
used to follow the water content on crude oil streams is a
capacitance probe, which is widely known as a BSW monitor where BSW means the percentage of Basic Sediment
and Water in the ow stream, and consists in free water
and sediment and/or emulsied water and sediment [3].
Once the dielectric constants of crude oil and water are
about 2 and 80, respectively, it is possible to use a concentric probe, which senses the apparent dielectric constant of
a uid stream by measuring its capacitance between the
probe electrodes. Since dielectric constants of oil and water
vary with temperature, a temperature compensation
scheme is required.
On the other hand, the evaluation of salt content on
crude oils is not a trivial task. This occurs because the salt
determinations are based on indirect methods, generally
involving the evaluation of electrical properties of crude
oil samples that are also inuenced by the water content.
Besides, as crude oil quite often presents poor electrical
properties, salt extraction procedures are generally
required in order to distinguish the ionic/electric activity
associated to the saline species come from the crude oil.
Otherwise, sharper electric properties may be attained by
diluting the crude oil into an appropriated solvent. As
the extraction procedures also involve the use of a proper
solvent (the extracting uid) such as toluene, xylene, and
others, one can conclude that the crude oil salt characterization depends clearly on the preparation of mixtures.
As a consequence, a good homogeneous condition in the
mixture must be satised. Homogeneity is therefore the
key feature for measurements based on electrical properties, and also is crucial for a right sampling in terms of having a representative sample.
Basically, three standard lab-scale test methods have
been employed to measure salt contents in crude oils: (i)
extraction and titration by classical Mohrs method [9],
which involves the salt content determination by extraction
and the volumetric titration method [10]; (ii) Potentiometric method [11], that comprises the use of an electrically
heated extraction apparatus in which a weighed crude oil
aliquot dissolved previously in xylene is added. Specied
volumes of alcohol, acetone and water are also added into
the apparatus as the extracting uid. A portion of aqueous
extract is analyzed for total halides by potentiometric titration (silver nitrate as the titrant) using a silver electrode and

a glass reference electrode; (iii) Electrometric Method [12]

which is based on the evaluation of the conductivity of a
solution of crude oil in a mixed alcohol solvent, using a
beaker and a set of electrodes. The chloride (salt) content
is obtained by reference to a calibration curve of conductivity versus chloride concentration of known mixtures.
Reasonable results for petroleum products with low viscosity can be achieved by those three methods presented
before. However, some diculties arise when one deals
with high water content emulsions or crude oil with higher
viscosities; hence, sample handling needs to be improved
beyond the existing standards. Considering the standard
tests involving extraction procedures (the Mohrs method
and the potentiometric one), the increased viscosity of the
sample implies a totally dierent mixing energy needed to
keep the homogeneity at the same desired level. Besides,
the extraction operation involving heating of very viscous
materials may represent a risk to the safety of the installation and experimental setup. In the case of the electrometric method, mixing problems with very viscous
samples can be reduced by warming until they are reasonably uid before the analysis; however, according to the
method, no sample shall be heated more than is necessary
to lower viscosity to a manageable level. Moreover, the
electrometric method is specically designed for crude oils,
with no remarks about the maximal water content for
which the method stands. Since the water inuences the
conductivity, one should expect increased uncertainties
on the salt content measured by this method when the
crude oil of interest contains signicant amounts of emulsied water.
From the previous discussion, although the existing
standard laboratory methods seem to be useful for determining the salt contents of crude oils, the use of these methods implies the taking of careful considerations in the
installation and the chemicals involved. As a matter of fact,
standard methods tend to fail when we are to analyze viscous crude oils or samples with high water contents.
In the work described herein, an improved procedure
was developed in order to determine the amount of salt
for those crude oils in which the available laboratory standard methods are unable to provide accurate salt content
values. The technique is based on measuring the salt content of the wash water that was separated from the crude
oil using microwave heating, after being articially emulsied using high shear rates. The novelty of this procedure
relies on the joint use of an emulsication step followed
by a demulsication one based on the irradiation of microwave energy. A number of experiments were performed in
order to demonstrate the feasibility of this technique. Standard laboratory procedures generally used for salt content
determination were also conducted for comparisons. The
novel procedure developed here was able to provide reliable information about the salt content on Brazilian crude
oils, in a rapid manner, without need of time-consuming
procedures or solvents preparations. It is believed that this
type of procedure can be very useful for the measurement

M. Fortuny et al. / Fuel 87 (2008) 12411248

of salt content for a number of crude oils and may serve to

inspire the development of a new (and more general) standard method in the future.
1.1. Microwave irradiation
The microwave irradiation oers a clean, inexpensive,
and convenient method of heating which often results in
higher yields and shorter reaction times [13,14]. The heating of liquids using microwaves can be explained by the
interaction of matter with the electric eld of the incident
radiation, causing the movement of ions as well as that
of induced or permanent molecule dipoles. The movement
of such species can cause heat generation (the so-called
dielectric heating) by two main mechanisms: dipole rotation, which is related to the alignment of molecules that
have permanent or induced dipoles with the electric eld
component of the radiation, and ionic conduction, which
refers to the migration of dissolved ions with the oscillating
electric eld [15].
Besides the quick heating of the materials, some authors
consider that the microwaves can provide specic eects
(not purely thermal) generally connected to the selective
absorption of microwave energy by polar molecules
[16,17]. The occurrences of these eects have not been
explained yet [18]. In spite of that, the microwave irradiation has become an interesting tool for a number of applications in the crude oil industry. Gunal and Islam [19]
studied the role of microwave irradiation in altering rheological properties of crude oils during heating. The authors
observed an irreversible eect on crude oil viscosity, which
was not detected in the absence of asphaltenes. This result
was supposed to be linked to the interaction of the microwave with the asphaltenes, suggesting the existence of a reorientation of their molecular structure. Similar ndings
were obtained by Bjorndalen and Islam [20] who observed
that at higher concentration of asphaltenes, microwave
irradiation leads to greater decline in viscosity, apparently
related to some sort of re-orientation of molecules combined with breakdown of some of the asphaltene particles.
In the current work, microwave irradiation is employed
for demulsication purposes. Since the pioneering work of
Wolf [21], a lot of activity has been developed concerning
the use of microwave irradiation for the demulsication
of water-in-oil and oil-in-water emulsions, providing
improved results when compared to the conventional heating procedure [2225]. In short, high rates are observed for
the separation of water and oil phases using microwaves,
which have been attributed to the following mechanisms:
(i) microwaves quickly heat emulsions, reducing the viscosity of the continuous phase (oil), thus favoring the water
droplets contact [22,23], (ii) the temperature increase can
cause reduction in the viscosity of the rigid lm formed
by natural surface-active species such as asphaltenes, resins, oil-soluble organic acids, solids, and waxes, in the
wateroil interface, making easier the coalescence between
dispersed water droplets [23,24], (iii) the microwave-

1243

induced molecular rotation neutralizes the zeta potential

of the dispersed droplets, thus reducing the stabilization
oered by the ionic surface-active species [24], and (iv)
the preferred absorption of microwaves by water droplets
in water-in-oil emulsions causes intense internal pressure
in these droplets, resulting in the expansion of the dispersed
phase and in the reduction of the thickness of the interfacial lm [23,25]. However, this latter mechanism is not
yet clearly proved. On the other hand, the breaking of
chemical bonds cannot be induced by the mere absorption
of microwaves in view of the low energy of photons as
compared to the chemical bonds-associated energies [26].
The energy range of hydrogen bonds, ionic bonds and
Van der Waals-like interactions vary between 0.04 and
4.51 eV, while the microwave photons are associated to
energies varying between 1.24 106 and 1.24 104 eV
according to the microwave frequency [15].
2. Experimental
2.1. Crude oil characterization
An experimental study was conducted to evaluate the
performance of the dierent techniques used to measure
the salinity of crude oils. Three dierent Brazilian crude
oils were used in this study. The characteristics of these
crude oils are summarized in Table 1 and the techniques
used in their characterization are described as follows.
The dynamic viscosity of the dierent crude oils was
determined using a rotational rheometer (Physica Rheolab
MC1, from Anton Paar) equipped with a heating jacket.
The analysis consisted in measuring the shear stress as a
function of the shear rate keeping the sample at 30.0 C
(0.3 C).
The water content (WC) of the crude oils was determined by the Karl Fischer (KF) reagent method, in accordance with ASTM D-1744 procedures [27]. The solvent
used during the analysis was a mixture of dry methanol
and chloroform (20%, v/v). For standardization of the
KF reagent, distilled water was solubilized into the solvent
cited before. A Metrohm KF titrator (model 836 Titrando)
equipped with a double platinum electrode was employed
during the water content determination tests.
The asphaltene content was determined gravimetrically
after recovering the n-heptane insoluble fraction from the
oil, in accordance with the ASTM D-6560 test method
[28]. Besides containing asphaltenes, this n-heptane insoluble fraction may also contain waxy substances and inorganic material. The asphaltenes are then recovered from
this fraction by successive washing with n-heptane and toluene, followed by solvent (toluene) evaporation. The total
acid number (TAN) was determined by potentiometric
titration of the crude oil with alcoholic potassium hydroxide (KOH) solution, in accordance with the ASTM D-664
test method [29]. Prior to each titration, crude oil samples
were dissolved in a 50% v/v toluene/isopropanol solution.
The same automatic titrator used for Karl Fischer analysis

1244

M. Fortuny et al. / Fuel 87 (2008) 12411248

Table 1
Summary of crude oils properties
Oil

API

Viscosity, cP
(30 C)

WC%a

% Asphaltene

Acid numberb

A
B
C

19.5
19.1
20

284
290
338

0.45
1.31
2.10

3.02
2.95
<0.5

1.31
1.32
1.22

a
b

WC means weight percentage of water content in the mixture.

Acid number is reported as mgKOH/g crude oil.

(Metrohm 836 Titrando) was employed for the acid number determination, however, equipped with a combination
electrode suitable for nonaqueous titrations. Such an electrode consisted in a glass electrode combined with a silver/
silver chloride (Ag/AgCl) reference electrode built in the
same electrode body. The acid number is then described
by the mass of KOH used to titrate sample per weight of
crude oil (mgKOH/g oil). Note that the crude oils studied
here are moderately acidic (as indicated by its TAN value)
and possesses a signicant salt content, as usually encountered in the inlet stream of a typical industrial desalting
plant section, which should reduce the water and salt content of the crude oil.

2.2. Experimental techniques for the crude oil salinity

characterization
2.2.1. Extraction and volumetric titration method
(adaptation of IP 77)
The crude oil salt removal was carried out using a glass
extraction apparatus from Petrotest which is specially
designed in accordance with the IP 77 [10]. The extraction
apparatus consists of a boiling ask of 500 cm3 capacity
which contains on the bottom an electric resistance for
the heating of the samples and, on the top, a reux condenser for the recovery of volatile compounds. The extraction runs were conducted for the dierent crude oils. In
most experiments, 100 cm3 of water was initially charged
to the ask followed by the addition of 100 cm3 of the
crude oil. Since the crude oils studied here were viscous,
all crude oil aliquots were mechanically stirred during
2 min to ensure homogeneity before sampling. Finally,
100 cm3 of toluene and 25 cm3 of acetone were transferred
to the ask. The mixture was then heated and kept at boiling conditions for at least 30 min, completing one wash
cycle of the experiment. Although the boiling interval
and the number of wash cycles established by the IP 77 test
method are 15 min and one cycle, respectively, higher salt
recoveries were obtained for our crude oils when higher
boiling times and further wash cycles were employed. Thus,
at least four wash cycles were adopted here, which means
overall boiling times of at least 120 min. After heating,
the mixture was left to cool and to separate in the ask,
resulting in an aqueous phase on the bottom and an oil
phase. The design of the ask enabled the removal of the
aqueous phase from the bottom of the ask to the salt

characterization step. The classical volumetric titration

method based on the Mohr method [9] was employed to
assess the chloride contents in the aqueous phase. The
Mohr titration method uses chromate ions as an indicator
(K2CrO4) in the titration of chloride ions with a silver
nitrate (AgNO3) standard solution. After all the chloride
has been precipitated as white silver chloride, the rst
excess of titrant results in the formation of a silver chromate precipitate, which signals the equivalence point.
Titration tests were performed here using a 0.01 N standard AgNO3 solution in the presence of a few drops of a
50 g dm3 K2CrO4 solution. The equivalence point was
reached when the color of the solution changed from yellow to brick-red due to the formation of a silver chromate
precipitate. The salt content was calculated from Eq. (1)
and reported as milligrams of sodium chloride per dm3.
X

VA
V1

VV B2
VS

N 58500 V E

where VA stands for the volume of silver nitrate solution

used to titrate the extract (in cm3), VB is the volume of silver nitrate solution used to titrate the blank, V1 is the volume of the extract aliquot used in the titration, V2 is the
volume of the aliquot used in the titration in blank, VE is
the total volume of the extract, VS is the volume of the sample specimen and N stands for the normality of the silver
nitrate solution.
2.2.2. Electrometric method (ASTM D-3230)
To determine the salt content of crude oils in accordance
with ASTM D-3230 procedures, a digital thermo-salt-incrude analyzer (LABCOMP SCTO 2100, from Petrolab)
was employed. This instrument performs conductivity measurements of crude oil dissolved in a mixed alcohol solvent,
using a beaker and an electrode assembly. Besides, the
instrument is equipped with a temperature probe that performs corrections on the conductivity signal automatically.
Both temperature and conductivity measurements are used
to automatically calculate salt concentration in g m3
directly from the internally stored calibration le, mirroring the standard sample concentrations outlined in ASTM
D-3230. The electrode assembly is composed by two parallel platinum boards of 25 50 mm exactly opposed and
gapped at 6 mm, and electrically separated by a non-conducting spacer.
Concerning the calibration procedure, the instrument
was calibrated against standard solutions of a rened neutral oil (chloride-free oil) and mixed chloride solutions in
xylene and an appropriated mixed alcohol solvent, always
in accordance with ASTM D-3230 test method. The mixed
alcohol solvent was prepared by initially mixing 37% v/v of
dry methanol in 1-butanol. Finally, to each litre of this
mixture, 3.0 cm3 of water was added and the resulting mixture was labeled as a stock mixed alcohol solvent. The
mixed chloride solution consisted in NaCl/MgCl2/CaCl2
(70/20/10) solution diluted in the mixed alcohol solvent,

M. Fortuny et al. / Fuel 87 (2008) 12411248

reaching a total salt concentration of 0.1 g dm3. A

homogenized aliquot of neutral oil (10 cm3) was carefully
dissolved in an organic solution containing 40 cm3 of
xylene and 50 cm3 of mixed alcohol solvent. The solution
was then transferred into a 100 cm3 test beaker in which
the conductivity measurements were performed. To cover
the range of chlorides contents of interest (0
430 mg dm3), known fractions of that 50 cm3 mixed alcohol solvent should be replaced by standard amounts of the
mixed chloride solution. During the procedure, the instrument records the corresponding conductivity signal for
each standard chloride concentration enabling to build an
internal calibration curve.
Once the calibration curve was established, the salt content of the crude oil was determined by dissolving 10 cm3 of
the crude oil sample in the organic solution (xylene/mixed
alcohol solvent). The mixture was then hand-shaken vigorously during 1.5 min, as recommended by ASTM D-3230
test method. After allowing the mixture to stand for
approximately 5 min, the mixture was placed in the test
beaker for salt content measurements.
2.2.3. New method: emulsication, microwave-assisted
demulsication and volumetric titration method
The new procedure developed here involved three
stages. The rst stage consisted in the emulsication of
the crude oil under analysis with wash water exempt of
salts. To achieve high mixing levels, a known amount
of water was added stepwise to the hydrocarbon phase,
while the mixture was hand-shaken to completely incorporate the water, resulting in a pre-emulsion with a total volume of 300 cm3. This mixture was then homogenized using
an Ultra-Turrax T-25 homogenizer tted with a S25-25G
dispersing tool (IKA). The ratio of water and crude oil in
the mixture was set to make up an emulsion with a range
of 3550% (v/v) of water content. The homogenization
conditions, such as the stirring frequency and the time of
homogenization, were kept constant at 17,500 rpm and
10 min, respectively, yielding emulsions with droplet size
distributions (DSD) with mean values in the range of
320 lm, as measured by laser diraction technique
(Malvern Instrument Model 2000 Mastersizer).
The second stage consisted in the demulsication of the
mixture for water removal using microwave irradiation.
This microwave assisted demulsication stage was conducted using a commercial microwave reactor system
(Anton Paar Synthos 3000). This instrument is equipped
with two magnetrons able to provide 1400 W of continuous
microwave power, a rotor system in which eight vessels
with an 80 cm3 capacity can be inserted at one time, and
magnetic stirrer for agitation of the sample in each vessel.
Also, an infrared sensor placed on the rotor bottom
enables us to assess the surface temperature of the vessel
and an immersing temperature probe (gas bulb thermosensor) placed at one reference vessel can be applied to follow
the temperature evolution of that sample during tests. The
vessels were 80 cm3 in capacity and made of quartz glass,

1245

resisting temperatures up to 300 C and pressures up to

80 bar. The control system supplied with the instrument
enables us to perform heating ramps with power and temperature conditions programmed prior to the run. A builtin forced-air cooling system allows for a fast cooling of the
vessels. The standard procedure for demulsication runs
involved the use of just two quartz glass vessels, each of
them lled with 40 g of emulsion. These samples were stirred throughout the tests using the magnetic stirrer, to facilitate the contact among the water droplets in the emulsion.
The total microwave irradiation time was set to 15 min for
each test. During this irradiation time, the temperature was
programmed to increase from ambient to the setpoint value
(90 C) in 2 min, being kept at this value during the subsequent 13 min. In this way, the microwave power supplied
by the instrument was automatically manipulated to ensure
the temperature programmed for the samples during the
heating ramp. After the irradiation time, samples were
quenched by the cooling system during 10 min and the
water phase formed during the demulsication test was
characterized in terms of salt contents. Note that, since
high mixing levels (and, ultimately, homogeneity) had been
achieved in the rst stage, the salt concentration of the separated water would not depend on the demulsication eciency. In a broader sense, these emulsication and
demulsication stages in this new method reproduce what
goes on during actual desalting operations: the rst stage
simulates the mix-valve action in the inlet stream of the
desalter vessel and the second stage reproduces the recovery of a water phase containing increased salt content
and a crude oil containing little unresolved water. However, it should be pointed out that the use of microwave
irradiation allows for the separation of the water phase
at relatively short time (15 min), as shown in a previous
work of this group [13].
Finally, the third stage consisted in the evaluation of the
salt content in the water phase obtained in the previous
stage. The water phase was ltered (for oil droplets
removal) and its salt content was characterized by the volumetric titration method based on the Mohr as described
before.

3. Results and discussion

3.1. Extraction and volumetric titration method (adaptation
of IP 77)
Extraction and titration tests were performed according
to the procedures described before in order to determine the
salt content of each crude oil. During these tests, at least
four wash cycles were employed to reach higher salt recoveries for the same crude oil sample. Tests were performed in
duplicate. The properties of the crude oils used during the
tests are reported in Table 1. The results of cumulative salt
contents obtained using this procedure are displayed in
Table 2. In all cases, the salt content increased markedly

1246

M. Fortuny et al. / Fuel 87 (2008) 12411248

Table 2
Cumulative salt content results for crude oils A, B and C from stepwise
extraction tests
Wash cycle

Salt content (mg dm3)

Oil A
(1)

1
2
3
4
5
6
7
8
9

29.6
37.4
45.6
53.4

Oil B
(2)
29.2
37.9
45.4
53.2

(1)
35.2
46.1
51.6
54.6
75.8
88.8
92.7

Oil

Oil C
(2)
23.7
47.8
51.3
60.1
67.8
75.2
86.5
93.4
99.4

(1)
15.8
20.0
23.8
25.8

Table 3
Salt content results for crude oils A, B and C from electrometric method
and related values for the Formation Water (calculated in mg of NaCl per
dm3 of water)
WC (%)

(2)
13.7
17.8
22.5
24.2

with the number of wash cycles, meaning that the extraction

procedure was unable to remove the salt from the crude oils
eciently. Tests for oils A and C yielded reproducible values of salt content per wash cycle. However, for these oils,
it was impossible to carry on additional wash cycles beyond
the 4th cycle due to the very small volume of organic phase
produced. Conversely, some uctuations were observed in
the salt content measured for oil B along the wash cycles.
Also, greater amounts of organic phase were available after
each wash cycle, allowing for running additional wash
cycles for oil B samples.
Clearly, reported values in Table 2 show that the cumulative salt content increase continuously with the number
of wash cycles for both oils A and B, suggesting that more
wash cycles would be needed. It seems that the salt content
reached a steady value only in the tests with oil C. As a
consequence, results provided by this method are likely
to be underestimated, which means that, even after adjusting the method including stepwise extraction schemes with
higher boiling interval, the current method yielded unreliable values of salt contents for those crude oils.
For the sake of comparison with other test methods, let
us calculate the salt content provided by this method for
each crude oil. Considering the average and deviation values for each oil reported in Table 2, the resulting salt content using this method was 53.3 0.1 (oil A), 96.1 4.7
(oil B) and 25.0 1.1. It is nevertheless likely that these
values are not entirely true (and analysis based on a dierent test method procedure should clarify this point).
3.2. Electrometric method (ASTM D-3230)
The electrometric method described before was applied
to evaluate the salt content of crude oils A (lots 1 and 2),
B (lot 1) and C. Table 3 presents the results of water content determined by the KF reagent method and the salt
content for the dierent crude oils. These data represent
average values of two independent tests and were used to
calculate the respective salt content of the formation water
emulsied in the crude, also included in Table 3, which
stands for the amount of salt (in mg) in the crude oil sam-

A (lot-1)
A (lot-2)
B (lot-1)
C
a

0.45
0.83
0.76
2.10

Salt content (mg dm3)

Crude oila

Formation water

69 3
153 7
279 10
432 9

15,378
18,386
36,750
20,581

Average value deviation.

ple per volume of water phase present in the crude assessed

by KF analysis. Confronting these results with those
obtained before by extraction procedures, note that much
higher salt contents were detected by the electrometric
method for all crude oils. Concerning the oil A, note that
two dierent lots were used in the tests, with sharp dierences in the WC. In fact, changes of the WC in the crude
oil generally lead to variations of the salt content in the
crude oil. As a consequence, an increase in the WC may
raise the salt amount per total volume of crude. It is nevertheless interesting to observe that the calculated salt content in the formation water may serve as a further proof
of validity (or not) of the values obtained through this
method. This occurs because the concentration of salt in
the formation water is supposed to vary between crude oils
coming from dierent sources or explored using dierent
methods. However, for samples coming from the same
crude oil, but with dierent WC, the salt content of this
water entrained in the crude is expected to be similar to
the salt content of the formation water, since the only difference between these samples has been the storage time. In
this way, the long storage time of crude oil samples produces water segregation and a natural increase of the WC
is generally observed. According to Table 3, one may
observe that the salt contents of the formation water for
the two lots of oil A are very dierent (an increase of
19.6% from lot 1 to the lot 2). Recall that these values of
salt content in the formation water were calculated supposing that the salt contents measured by the instrument for
the crude oil were true. Among the reasons that could
explain that dierence, one may consider that the calibration curve of the salt analyzer did not take into account
a number of variables that can play a role on the conductivity values measured by the instrument, such as the WC
and the water droplet size. Thus, further studies are being
conducted in the lab to investigate the use of the electrometric method in order to build calibration curves on a
broader range of conditions, with the inclusion of these
variables, which means that the current electrometric standard method should be reviewed.
On the other hand, the electrometric method is inappropriate to analyze samples with relatively high salt contents
or with salt contents beyond the range of calibration, as in
the case of oil C. For this crude oil, the salt content was
assessed by extrapolating the calibration curve.

M. Fortuny et al. / Fuel 87 (2008) 12411248

3.3. New method: emulsication, microwave-assisted

demulsication and volumetric titration method

Table 4
Salt content results for crude oils A, B and C from the new method and
related values for the Formation Water (calculated in mg of NaCl per dm3
of water)
Oil

A (lot-1)
A(lot-2)
B (lot-1)
B (lot-2)
C
a

Salt content (mg dm3)

WC (%)

0.45
0.83
0.76
1.31
2.10

Crude oila

Formation water

108 5
186 7
313 13
530 16
524 15

23,978
22,458
41,211
40,489
24,929

Average value deviation.

Table 5
Salt content results for crude oils A, B and C from dierent evaluation
methods (calculated in mg of NaCl per dm3 of crude oil)
Oil

WC (%)

A (lot-1)
A(lot-2)
B (lot-1)
B (lot-2)
C
a

0.45
0.83
0.76
1.31
2.10

500
450
Salt Content (mg.dm-)

The new method based on emulsication and microwave-assisted demulsication described before was applied
to evaluate the salt content of crude oils A (lots 1 and 2), B
(lots 1 and 2) and C. Table 4 presents the results of water
content determined by the KF reagent method, the salt
content for the dierent crude oils and for the respective
formation water emulsied in the crude. These data represent average values of two independent tests. Note that the
new method provided salt content values much higher than
those obtained by the previous methods, which seems to be
related to the high mixing level achieved during the emulsication stage. As expected, the increase of the WC of the
crude oil between lots (e.g., A and B) yielded an increase
of the salt content per total volume of crude oil. However,
even when the WC increased between crude oil lots, the
estimated values of salt content for the formation water
presented slight negative uctuations of 6.3% for oil A
and 1.8% for oil B, which are found to be very low when
compared to the increase of 19.6% observed for the electrometric method discussed before. This may be seen as a
further proof of reliability of the emulsication/microwave-assisted procedure adopted here.
Table 5 summarizes the average results of salt content
for the three crude oils using the dierent evaluation methods, indicating that both extraction and electrometric
methods underestimated the salt content of the samples.
Fig. 1 illustrates this point showing the average salt content
values obtained by the dierent methods for the rst lot of

1247

400
350
300
250
200
150
100
50
0
lC
Oi

Ne
lB
Oi

wM

eth

Ele
lA
Oi

Ex

ctro

od

me

tra

ctio

tric

Fig. 1. Salt content values of oils A, B and C using dierent evaluation

methods. Crude oils containing WC of 0.45% (oil A), 0.76% (oil B, except
for the extraction method which involved WC = 1.31%) and 2.10% (oil C).

oil A (WC = 0.45%), oil B (WC = 0.76%, except for the

extraction method which involved WC = 1.31%) and oil
C (WC = 2.10%).
An important issue concerning this new method is
related to the demulsication time and eciency. Clearly,
if high mixing levels are achieved during the emulsication
stage, the main concern for the following stages will be the
separation of an amount of resolved water large enough to
perform the salt content characterization. Besides, even if a
signicant amount of aqueous phase remains entrained in
the crude after the demulsication stage, the concentration
of salt in both resolved and unresolved water should be
essentially the same, since sample homogeneity had been
ensured in the beginning. As a consequence, this method
does not require very high demulsication eciencies to
provide accurate values of crude oil salt contents. This is
comforting as it implies that we can select the microwave
irradiation conditions (including the irradiation time and
temperature) that provide a minimum amount of separated
water for titration purposes. For all the tests related in
Table 4, microwave irradiation times of 15 min and temperature of 90 C were enough to provide proper amounts
of separated water for titration purposes. Thus, the emulsication/microwave-assisted demulsication procedure presented here may be viewed as a promising tool to perform
reliable salt content evaluations in a rapid manner.
4. Conclusions

3

Salt content (mg dm )

Extraction

Electrometric

New method

53.3 0.1

69 3
153 7
279 10

432 9

108 5
186 7
313 13
530 16
524 15

96.1 4.7
25.0 1.1

Average value deviation.

The aim of this study was to evaluate dierent techniques for the determination of the salt content of heavy
crude oils. In this direction, an experimental study was carried out based on three Brazilian crude oils whose salt contents were determined in accordance with the techniques
already studied in the literature established by the standard
laboratory methods (IP 77 and ASTM D-3230). Results

1248

M. Fortuny et al. / Fuel 87 (2008) 12411248

show that the existing techniques based on solvent extraction procedures or calibration curves present some limitations. First, even after including additional wash cycles,
the extraction method provided much underestimated values of salt contents for the three crude oils. Besides, this
method involved cumbersome and time-consuming experimental procedures.
The electrometric method also seemed to underestimate
the salt contents in all cases. Also, the calibration curve as
recommended by the standard procedure did not include
values of salt content beyond 430 mg dm3, as in the case
of oil C; hence, an extrapolation of the curve calibration
was required, but resulted in unreliable values. Moreover,
the water content in the crude led to corrupted values of
salt content of the formation water, which means that care
must be taken when one applies this standard methodology
in the eld (although no consideration concerning water
content limits in the crude oil is stated by this standard
method).
Finally, an improved procedure combining an emulsication stage with the microwave-assisted demulsication
and the volumetric titration was proposed. Results
revealed that this new method is able to deal with samples
containing dierent water content, providing accurate values of salt contents for viscous crude oils without time-consuming procedures or calibration curves. Therefore,
besides being fast, this microwave strategy presented shows
little sensitivity to the disturbances generally encountered
during actual crude oil processing. Consequently, these
results are encouraging and show that the microwave strategy presented is robust enough and may serve to inspire the
development of a new (and more general) standard method
in the future.
Acknowledgments
The authors thank Petrobras (Petroleo Brasileiro S.A.
Brazil) for supporting the present work. M.F. and M.N.
also thank CNPq and FINEP (Brazilian funding agencies).
References
[1] Mohameda AMO, Gamala M, Zekri AY. Eect of salinity and
temperature on water cut determination in oil reservoirs. J Petrol Sci
Eng 2003;40:17788.
[2] Speight JG. Handbook of petroleum analysis. New York: WileyInterscience; 2001.
[3] Smith HV, Arnold KE. Crude oil emulsions. In: Bradley HB, editor.
Petroleum engineering handbook. 3rd ed. Richardson: Social of
Petroleum Engineers; 1992. p. 19.119.34.
[4] Pal R. Techniques for measuring the composition (oil and water
content) of emulsions a state of the art review. Colloid Surface A
1994;84:14193.

[5] ASTM D-4006. Standard test method for water in crude oil by
distillation.
[6] ASTM D-4007. Standard test method for water and sediment in
crude oil by the centrifuge method (laboratory procedure).
[7] ASTM 4377. Standard test method for water in crude oils by
potentiometric Karl Fischer titration.
[8] Jakoby B, Vellekoop MJ. Physical sensors for water-in-oil emulsions.
Sensor Actuat A 2004;110:2832.
[9] Skoog DA, West D, Holler FJ. Fundamentals of analytical chemistry.
7th ed. Belmont: Thomson; 1996.
[10] IP 77. Determination of salt content extraction and volumetric
titration method.
[11] ASTM D-6470. Standard test method for salts in crude oils
(potentiometric method).
[12] ASTM D-3230. Standard test method for salts in crude oil (electrometric method).
[13] Fortuny M, Oliveira CBZ, Melo RLFV, Nele M, Coutinho RCC,
Santos AF. Eect of salinity, temperature, water content, and pH on
the microwave demulsication of crude oil emulsions. Energ Fuels
2007;21:135864.
[14] Roussy G, Pearce JA. Foundations and industrial applications of
microwave and radio frequency elds: physical and chemical
processes. Chichester: John Wiley & Sons; 1995.
[15] Thuery J. Les microondes et leurs eets sur la matie`re applications
industrielles, agro-alimentaires et medicales. Paris: Technique et
Documentation; 1983.
[16] Perreux L, Loupy A. In: Loupy A, editor. Microwaves in organic
synthesis. Weinheim: Wiley; 2002. p. 61114.
[17] Caddick S. Microwave assisted organic reactions. Tetrahedron
1995;52:1040332.
[18] Miklavc A. Strong acceleration of chemical reactions occurring
through the eects of rotational excitation on collision geometry.
Chem Phys Chem 2001;8/9:5525.
[19] Gunal OG, Islam MR. Alteration of asphaltic crude rheology with
electromagnetic and ultrasonic irradiation. J Petrol Sci Eng
2000;26:26372.
[20] Bjorndalen N, Islam MR. The eect of microwave and ultrasonic
irradiation on crude oil during production with a horizontal well. J
Petrol Sci Eng 2004;43:13950.
[21] Wolf NO. (Conoco Inc.). Use of microwave radiation in separating
emulsions and dispersions of hydrocarbons and water. US Patent
4,582,629; 1986.
[22] Fang CS, Chang BKL, Lai PMC, Klaila WJ. Microwave demulsication. Chem Eng Commun 1988;73:22739.
[23] Nilsen PJ, Kornfeldt A, Nygren T, Fdhila RB. (ABB Research Ltd.).
Method for separating an emulsion by using microwave radiation.
Patent WO 0112289; 2001.
[24] Fang CS, Lai PMC, Chang BKL, Klaila WJ. ;Oil recovery and waste
reduction by microwave radiation. Environ Prog 1989;8:2358.
[25] Xia L, Lu S, Cao G. Stability and demulsication of emulsions
stabilized by asphaltenes or resins. J Colloid Interf Sci
2004;271:5046.
[26] Fini A, Breccia A. Chemistry by microwaves. Pure Appl Chem
1999;71:5739.
[27] ASTM D-1744. Standard test method for determination of water in
liquid petroleum products by Karl Fischer reagent.
[28] ASTM D-6560. Standard test method for determination of asphaltenes (heptane insolubles) in crude petroleum and petroleum products.
[29] ASTM D-664. Standard test method for acid number of petroleum
products by potentiometric titration.