Ultrasonics Sonochemistry 35 (2017) 541–546

Contents lists available at ScienceDirect

Ultrasonics Sonochemistry
journal homepage: www.elsevier.com/locate/ultson

Effect of ultrasonic frequency on separation of water from heavy crude

oil emulsion using ultrasonic baths
Fabiane G. Antes a, Liange O. Diehl b, Juliana S.F. Pereira c, Regina C.L. Guimarães d, Ricardo A. Guarnieri d,
Bianca M.S. Ferreira d, Erico M.M. Flores e,⇑
a
Empresa Brasileira de Pesquisa Agropecuária, Embrapa Suínos e Aves, 89700-991 Concórdia, SC, Brazil
b
Empresa Brasileira de Pesquisa Agropecuária, Embrapa Gado de Corte, 79106-550 Campo Grande, MS, Brazil
c
Departamento de Química Inorgânica, Universidade Federal do Rio Grande do Sul, 91501-970 Porto Alegre, RS, Brazil
d
Centro de Pesquisas e Desenvolvimento Leopoldo Américo Miguez de Mello – CENPES/PETROBRAS, 21941-915 Rio de Janeiro, RJ, Brazil
e
Departamento de Química, Universidade Federal de Santa Maria, 97105-900 Santa Maria, RS, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: In this work, a comprehensive study was performed for the evaluation of ultrasound (US) frequency for
Received 17 November 2015 demulsification of crude oil emulsions. Experiments were performed using ultrasonic baths operating at
Received in revised form 30 March 2016 the following frequencies: 25, 35, 45, 130, 582, 862 and 1146 kHz. Synthetic water-in-oil emulsions with
Accepted 31 March 2016
12%, 35% and 50% of water and medians of droplet size distribution (DSD, D(0.5)) of 5, 10 and 25 lm were
Available online 1 April 2016
prepared using a heavy crude oil (API density of 19). Crude oil demulsification was achieved at frequen-
cies in the range of 25–45 kHz for all tested emulsions. When frequencies higher than 45 kHz were
Keywords:
applied, no changes in the characteristics of the crude oil emulsions were observed. Demulsification effi-
Crude oil
Emulsion
ciencies of about 65% were achieved at a frequency of 45 kHz after 15 min of US application (emulsions
Ultrasound with original water content of 50% and D(0.5) = 10 lm). An important aspect is that no addition of
Demulsification chemical demulsifiers was performed, and the demulsification efficiency was considered high, taking into
Water removal account that the results were obtained using a non-conventional crude oil. Contrary to the normal appli-
Low frequency ultrasound cation of low-frequency US that has been used for emulsification, the proposed approach seems to be a
promising technology for water removal from crude oil emulsions.
Ó 2016 Elsevier B.V. All rights reserved.

1. Introduction The destabilization of water-in-oil emulsions is not a simple

task, especially in the case of crude oil emulsions. Several demulsi-
Crude oil is a complex mixture that, in general, is produced fication processes have been developed for the petroleum industry
together with salted water to form stable emulsions. The co- or for the laboratory scale [5–7], such as filtration by membranes
produced water is emulsified when the fluids flow through pipes [8], centrifugation [9] and microwave radiation [10,11]. In spite
and choke valves, forming emulsions (commonly called water-in- of the relatively good efficiency of these procedures for light crude
oil emulsion) that could be stabilized by compounds naturally pre- oils, they are time consuming, and the addition of chemical demul-
sent in crude oil such as asphaltenes, resins, wax and solids [1,2]. sifiers is almost mandatory (especially for heavy crude oils), which
The presence of water during crude oil production is undesirable increases treatment costs.
mainly due to pipeline corrosion and the additional cost of trans- The use of ultrasound (US) has been also studied to promote the
portation [1,3]. In this sense, crude oil that is virtually free of water demulsification of emulsions with different physical–chemical
is required for pipeline flow and especially for refining operations. properties, as reported in several works [12–18] and patents
Before being used in refining operations, the crude oil should be [19]. In recent years, US has been applied in industrial applications
treated in order to promote the water and consequently the salt of crude oil and related products [20–26]. Especially for demulsifi-
removal. After demulsification, a relatively large amount of water cation processes, two effects attributed to US have been pointed
is separated from crude oil and, for most of refineries, the amount out such as the formation of standing waves field and cavitation
of water in oil before the refining step should be lower than 0.5% [12,20,21]. According to Gardner and Apfel [12], depending on
[4]. the conditions of US application, cavitation can strongly affect sys-
tems containing drops in immiscible liquids, either causing emul-
⇑ Corresponding author. sification or the coalescence of drops. Using high-power US,
E-mail address: ericommf@gmail.com (E.M.M. Flores). cavitation could break drops into finer droplets by rupturing their

http://dx.doi.org/10.1016/j.ultsonch.2016.03.031
1350-4177/Ó 2016 Elsevier B.V. All rights reserved.
542 F.G. Antes et al. / Ultrasonics Sonochemistry 35 (2017) 541–546

surfaces and producing microjets. In this case, emulsions can be The water content was determined in W/O emulsions and in the
formed [27]. On the other hand, the coalescence of drops followed demulsified crude oil by Karl Fisher titration, following the ASTM
by the separation of water from crude oil can be stimulated by cav- D-4377 method [31] and using an automatic titrator (model 836,
itation if just enough energy is transmitted to the drops’ surfaces to Metrohm, Herisau, Switzerland) equipped with a platinum elec-
break down the barriers between the drops without shattering the trode (model 8.109.1306, Metrohm).
drops themselves [12]. In general, most studies using US for
demulsification have involved the treatment of other emulsions 2.2. Reagents
than crude oil (e.g., water and vegetal oils, organic solvents, etc.)
[12,16]. In the case of crude oil, demulsifiers are normally required, Purified water from a Milli-QÒ system (Milipore, Billerica, USA)
increasing treatment costs and contaminating petroleum products was used to prepare all solutions. Reagent Composite 5 (Riedel-de
(e.g. diesel oil), since demulsifiers can remain after the refining Häen, Seelze, Germany) was used for Karl Fisher titration. A 3:1
process [12,18]. mixture of toluene (Vetec Química Fina Ltda., Rio de Janeiro, Brazil)
Recently, a demulsification process using an indirect application and methanol (Carlo Herba Reagents, Milan, Italy) was used for
of US to remove water from crude oil emulsions was developed sample dissolution for the further determination of water content.
[28,29]. The advantage of an indirect application of US, which uses Sodium chloride solution, used for the preparation of synthetic
water as a propagation medium, is that the amount of energy emulsions, was prepared by the dissolution of the salt (Vetec,
delivered is high enough to destabilize the emulsion and cause Brazil) in water.
the separation of oil and water phases, but not too high to result
in emulsification. However, only the frequency of 35 kHz was 2.3. Preparation and characterization of W/O emulsions
investigated.
In the present study, US was applied for the demulsification of Water-in-oil emulsions were prepared using a Brazilian heavy
heavy crude oil (API density of 19), using water as a propagation crude oil with an API density of 19 and with water and salt content
medium. The effect of several ultrasonic frequencies was evaluated below 0.5% and 10 lg g
1, respectively. API density, kinematic
for the first time. Ultrasonic baths with variable frequencies of 25, viscosity (45 °C) and dynamic viscosity (45 °C) of the crude oil
35, 45, 130, 582 and 1146 kHz were used. A comprehensive study sample used for emulsion preparation were determined as 19,
was performed using synthetic crude oil emulsions, and the fol- 133.4 mm2 s
1 and 122.9 mPa s, respectively.
lowing parameters were evaluated: time of exposure to US During the preparation of the synthetic emulsions, a known
(5, 15, 30, 45 and 60 min), water content (12%, 35% and 50%) and amount of 100 g l
1 NaCl solution was added to the crude oil.
median droplet size distribution (DSD, D(0.5) = 5, 10 and 25 lm). The mixture was shaken in an oven adapted with a mechanical
Demulsification efficiencies were calculated by taking into account stirrer to incorporate the water into the oil phase at 80 °C. The mix-
the difference between initial and final water content (after ultra- ture was then homogenized using a PolytronÒ stirrer with the
sonic treatment). selected frequency and time to yield emulsions with a monodis-
perse DSD profile and median values, D(0.5), close to 10 lm. After
preparation, the emulsions were characterized by their water con-
2. Materials and methods tent and DSD. Emulsions with different characteristics were pre-
pared: (i) D(0.5) of 10 lm and different water contents (12, 35
2.1. Instrumentation and 50%) and (ii) 35% water and D(0.5) of 5 and 25 lm. During
the development of this work, the W/O emulsions were prepared
Ultrasonic baths with temperature control and variable fre- daily.
quencies of 25 or 45 kHz (model Transonic TI-H 5) and 35 or
130 kHz (model Transonic TI-H 10) from Elma Ultrasonic (Singen, 2.4. US-assisted demulsification experiments
Germany) were used. According to the manufacturer, the effective
ultrasonic power of these baths were 100 and 200 W for the mod- For all experiments using ultrasonic baths, 20 g of W/O emul-
els TI-H 5 and TI-H 10, respectively. In order to avoid the increase sion were transferred to a conic glass vessel, and the vessel was
in temperature during US application, this parameter was kept introduced into the ultrasonic bath considering the position that
constant by circulation of water using pipes and the available drain higher demulsification efficiencies were obtained.
systems of the ultrasonic baths. A multifrequency bath system that Experiments with ultrasonic baths were performed to evaluate
can operate at high frequencies such as 582, 862 and 1146 kHz (US the influence of different frequencies on crude oil demulsification
Transducer E805TM, Meinhardt Ultraschalltechnik, Leipzig, using an emulsion with 35% water and D(0.5) = 10 lm. The tem-
Germany) was also investigated. This system is similar to conven- perature of the ultrasonic baths was kept at 45 °C and the ampli-
tional ultrasonic baths, but its bottom is itself the transducer, tude was set to 100% (conditions selected according to previous
which is different from normal baths where there is a metallic work) [28].
cover over the transducer. The temperature was kept constant Before each experiment, the vessels with the crude oil emulsion
within a jacketed cell that enabled cooling during US application. were kept in the ultrasonic baths until the emulsion achieved the
An oven adapted with a mechanical stirrer (model 400-DE, selected temperature for each respective test. After exposure to
Nova Etica, Brazil) was used to prepare synthetic emulsions by US, the conic glass vessels were taken out of the ultrasonic baths
mixing crude oil and NaCl solution. A Polytron mechanical stirrer and transferred to a water bath at the same temperature used in
(model PT 3100 D, Littau-Lucerna, Switzerland) was also used for the US experiment. The conic glass vessels were kept in a water
the preparation of synthetic emulsions [29]. DSD in emulsions bath for 30 min to allow the water to settle. For all experiments
was determined with a laser diffraction technique using a particle in which water separation from the emulsion was visually
size analyzer, Mastersizer 2000 (Malvern Instruments, Malvern, detected, the efficiency of demulsification (ED) was calculated by
United Kingdom). White mineral oil (Alpha Química Ltda., São taking into account the initial water content (determined by Karl
Paulo, Brazil) was used as a diluent, and 4 to 7 drops of crude oil Fischer titration) in the emulsion before US (IWC) and the final
or water-in-oil (W/O) emulsion were dispersed in about 40 mL of water content obtained after exposure to US (FWC), according to
the diluent for the measurement of DSD. the equation:
 F.G. Antes et al. / Ultrasonics Sonochemistry 35 (2017) 541–546 543

IWC
FWC It would be expected that the distribution of acoustic power in

ED;% ¼  100
IWC an ultrasonic bath cavity could not be homogeneous due to the
position of the transducer. Therefore, experiments were initially
To determine the water content in the crude oil obtained after
developed to evaluate whether the position in which the vessel
US and settling time, the oil phase was carefully removed using a
containing crude oil is inserted into the bath would affect the
syringe and it was homogenized. For the experiments in which
demulsification efficiency. For these experiments, the temperature
water separation was not visually observed, the emulsion was
and amplitude of the ultrasonic baths were kept at 45 °C and 100%,
gently homogenized in the glass vessel, and an aliquot was sam-
respectively, and US was applied for 15 min (the best condition
pled for DSD determination. In this case, the resulting DSD profile
previously established for demulsification at frequency of
obtained was compared with that of the emulsion without US
35 kHz). The results obtained for all frequencies and positions eval-
treatment. For all experiments, at least 3 replicates were per-
uated are shown in Fig. 2.
formed and the results are given as mean ± standard deviation.
It is possible to observe (Fig. 2) that higher demulsification effi-
Statistical analysis was performed for the comparison of results
ciencies were obtained when the emulsion was positioned in the
using student t-test or one-way ANOVA (minimum significance
center of ultrasonic bath (position 1) or in position 4 or 5, and sim-
level for P < 0.05), according to the number of mean results.
ilar results were obtained for these positions (ANOVA, minimum
The effect of US at different frequencies on emulsions with D
significance level for P < 0.05). Therefore, position 5 was arbitrarily
(0.5) = 10 lm was also studied for crude oil emulsions with 12
selected for all subsequent experiments. It is important to mention
and 50% water content. Times of exposure to US were evaluated
that in all experiments performed at 25, 35 and 45 kHz, a water
from 5 to 60 min for all emulsions (water content of 12%, 35%
layer could be observed on the bottom of conical glass vessel after
and 50% and D(0.5) of 10 lm). Finally, emulsions with D(0.5) = 5
the settling time, and the volume of separated water was equiva-
and 25 lm were prepared and submitted to different ultrasonic
lent to the calculated demulsification efficiency shown in Fig. 2.
frequencies. Fig. 1 shows the overall procedure and the sequence
On the other hand, the separation of water from the emulsion
of optimizations performed.
was not visually observed at 130 kHz in any of the evaluated posi-
tions in the ultrasonic bath. Therefore, DSD was determined to
3. Results and discussion evaluate possible changes in comparison to the original W/O emul-
sion. The emulsion not treated with US was kept in a water bath at
3.1. Comparison of different frequencies of US for the demulsification 45 °C for 15 min, and it was observed that the DSD profile of emul-
of crude oil sion exposed to US at 130 kHz was similar to the DSD profile of the
original emulsion (without US treatment). Subsequent experi-
In a previous study, only the frequency of 35 kHz was used for ments were performed at 130 kHz, decreasing the US amplitude
crude oil demulsification [29]. In the present work, the feasibility power to 50%. Emulsions were exposed to US for 15 min in differ-
of using US for demulsification purposes was evaluated using ent positions in the ultrasonic bath. Again, after the settling time
two ultrasonic baths that could operate at frequencies of 25 or no separation of water was visually detected. After the determina-
45 kHz and 35 or 130 kHz. Furthermore, experiments using higher tion of DSD, no difference was observed between graphics of
frequencies (582, 862 and 1146 kHz) were also performed using a untreated and US-treated emulsions. Other tests were carried out
multifrequency bath system. Additionally, the effect of US for the at 130 kHz at different temperatures (25 and 60 °C) and with dif-
demulsification of crude oil emulsions with variable characteristics ferent exposure times (up to 60 min of US), but no demulsification
(water content and droplet size distribution) was evaluated. or changes in the DSD profile were observed in any test. Therefore,

Ultrasound assisted demulsification

Evaluated parameters
20 g of synthetic W/O emulsion Time of US (5 to 60 min)

Ultrasound (frequencies of 25, 35,

Water content (12, 35 and 50%)
45, 130, 582, 862 or 1146 kHz)

DSD D(0.5) (5, 10 and 25 µm)

Separation of water?
Yes No

Water settling (30 DSD

min in a water bath) determination

Removal of oil
phase

Determination of
water content

Calculation of
demulsification
efficiency

Fig. 1. Experimental workflow for general demulsification procedure. All demulsification experiments on ultrasonic baths were performed at 45 °C.
544 F.G. Antes et al. / Ultrasonics Sonochemistry 35 (2017) 541–546

separation from the crude oil emulsion was observed using any of
Position 2 Position 3 these frequencies after 15 min of US exposure. Therefore, DSD was
25 kHz: 25.2 ± 4.5 25 kHz: 25.4 ± 4.0 determined and the profile obtained was similar to that of the
35 kHz: 29.5 ± 3.8 35 kHz: 41.2 ± 2.8 untrated crude oil emulsion.
45 kHz: 41.8 ± 6.0 45 kHz: 29.6 ± 2.3 The comparison of results obtained at different US frequencies
Position 1 showed that demulsification was achieved at frequencies up to
25 kHz: 45.9 ± 5.4 45 kHz, but no effect on crude oil emulsion was observed at fre-
35 kHz: 43.9 ± 4.9 quencies of 130 kHz or higher. According to the literature [30],
45 kHz: 53.8 ± 4,7 the mechanical effect of acoustic cavitation in an aqueous system
Position 4 Position 5 increases when the frequency of US decreases. In this sense, prob-
25 kHz: 44.0 ± 5.3 25 kHz: 50.1 ± 3.3 ably at ultrasonic frequencies up to 45 kHz, the mechanical effect
35 kHz: 44.1 ± 2.8 35 kHz: 51.0 ± 3.0 of US (cavitation) is strong enough to destabilize the emulsions
45 kHz: 55.3 ± 3.4 45 kHz: 60.0 ± 2.6 and cause the coalescence of water drops with consequent separa-
tion of water from the oil phase [12].
It is already known that to make the separation of water from
Fig. 2. Demulsification efficiencies (%) obtained by US treatment for W/O emulsion
(water content of 35% and D(0.5) = 10 lm) inserted in different positions on crude oil possible, the interfacial film that stabilizes the emulsion
ultrasonic bath (top view). Temperature of bath: 45 °C, amplitude: 100%, time of must be removed or ‘‘broken” [32,33]. Therefore, the mechanism
exposition to US: 15 min (n = 3). of demulsification at low ultrasonic frequencies is related to the
mechanical effect of US (shock waves, microjets and turbulence),
which is responsible for breaking intermolecular interactions and
50
A allowing the coalescence of water drops [12]. On the other hand,
it is also known that US can be applied to prepare emulsions
40 [34]. Therefore, the authors agree that there is a limit for the fre-
quency and energy delivered by US that can cause demulsification
30 or emulsification effects. Yang et al. [18] observed that the demul-
sification of crude oil increased when a higher ultrasonic power
was used (until to reach a determined value). Then, for a higher
20
ultrasonic power, a decrease of demulsification efficiency was
observed. In this sense, according to the results obtained in the
10
present work, it is possible to suggest that the turbulence caused
Efficiency of demulsification, %

by cavitation has to be maintained at a specific level to allow for

0 the breaking of the interfacial film, which is responsible for the sta-
5 15 30 bilization of the W/O emulsions. However, if turbulence becomes
stronger than this level, water drops could be broken into smaller
70 B drops and consequently increase the stability of the emulsion. On
60 the other hand, in the case of frequencies of 130 kHz and higher,
the turbulence caused by cavitation is not strong enough to break
50
the interfacial film and cause demulsification.
40 It is also important to mention that a standing wave field can be
created at all ultrasonic frequencies. This field could be responsible
30 to approximate water droplets and therefore improve the coales-
20
cence. It is possible that some effect of standing waves could occur
in the experiments performed in this work, although suitable
10 experimental conditions, as previously described [21], were not
used here. Therefore, it is not possible to state that the demulsifi-
0
5 15 30 cation of crude oil obtained in this work had no relation to the
Time, min standing wave field. However, the mechanical effect of US was
probably more important for demulsification than a standing wave
Fig. 3. Efficiency of demulsification using US baths for heavy crude oil emulsion field because experimental conditions more suitable to allow this
with initial water content of (A) 12% and (B) 50%, D(0.5) = 10 lm, after different effect were not employed (i.e. the specific dimension of the
times of exposure to US. Frequencies: h 25 kHz, 35 kHz, j 45 kHz. Temperature: apparatus to allow for the reflection of acoustic waves to create a
45 °C, amplitude: 100% (n = 3).
standing wave field with maximum intensity).

the frequency of 130 kHz was not suitable for promoting crude oil 3.2. Effect of US exposure time on the demulsification of crude oil
demulsification.
In order to evaluate frequencies higher than 130 kHz, a multi- The results obtained at variable US frequencies showed that
frequency bath system was used. The same experimental condi- demulsification occurred at 25, 35 and 45 kHz. Therefore, all fur-
tions used on low-frequency ultrasonic baths (kind of emulsion, ther experiments were performed at these frequencies. The emul-
temperature and time) were employed, and the amplitude was sions were exposed to ultrasonic frequencies during the shortest
kept at 70% (the maximum allowed according to the manufac- amount of time possible to achieve maximum demulsification effi-
turer’s recommendations). An indirect application of US was eval- ciency. It was observed during the experiments by using transpar-
uated by positioning the conical glass vessel in the center of the ent glass vessels that the coalescence of water drops started after
bath and using water as the US propagation medium. All allowed the early minutes of US exposure. Therefore, tests were performed
individual frequencies – 582, 862 and 1146 kHz – were tested. As to evaluate the exposure time of emulsion to US in the range of
observed when the frequency of 130 kHz was evaluated, no water 5–60 min. After these experiments, the glass vessels were kept in
 F.G. Antes et al. / Ultrasonics Sonochemistry 35 (2017) 541–546 545

Table 1 Table 2
Demulsification efficiencies obtained by ultrasound for W/O emulsion with D(0.5) Demulsification efficiencies (%) obtained for heavy crude oil emulsions with DSD D
= 10 lm and water content of 35% at different times of exposition to ultrasound. (0.5) of 5, 10 and 25 lm (water content of 35%). Amplitude of ultrasound: 100%,
Temperature of ultrasonic bath: 45 °C, amplitude: 100% (n = 3). ultrasound exposition time of 15 min, temperature of 45 °C (n = 3).

Time, min Efficiency of demulsification, % Frequency, kHz Efficiency of demulsification,%

25 kHz 35 kHz 45 kHz 5 lm 10 lm 25 lm
5 11.6 ± 0.9 16.8 ± 3.1 19.0 ± 5.4 25 28.5 ± 6.2 50.1 ± 3.3 55.0 ± 1.5
15 50.1 ± 3.3 51.7 ± 3.0 60.0 ± 2.6 35 49.6 ± 4.4 51.7 ± 3.0 40.0 ± 3.1
30 50.5 ± 2.4 50.9 ± 4.6 58.2 ± 2.8 45 56.1 ± 1.9 60.0 ± 2.6 36.2 ± 1.4
45 49.9 ± 2.4 48.3 ± 4.9 60.4 ± 1.3
60 48.8 ± 2.0 51.7 ± 2.4 58.6 ± 4.0

The results obtained in this work showed that the demulsifica-

a water bath at 45 °C for 30 min to allow the water drops to settle. tion efficiency for emulsions with different initial water content
The demulsification efficiencies are shown in Table 1. was in agreement with results described in the literature [28,29].
As can be observed in Table 1, demulsification efficiencies lower It is expected that increasing the water content of emulsion would
than 20% were obtained after 5 min of exposure to US at the differ- also increase the demulsification efficiency, because if more water
ent frequencies evaluated. On the other hand, after 15 min, effi- drops are in the emulsion, these drops could more easily come
ciencies higher than 50% were obtained for all of the frequencies together, leading to coalescence [32].
used. Statistical analysis of the results (one-way ANOVA, minimum
significance level for P < 0.05) showed that the demulsification effi-
3.4. Evaluation of US on W/O emulsions with different median drop
ciencies obtained using US exposure times longer than 15 min
sizes
were not statistically different at the same frequency. It could be
observed through the transparent glass vessels that the coales-
All results for the demulsification of crude oil presented up to
cence of water drops started after the early minutes of emulsion
this point were obtained for emulsions prepared with D(0.5) close
exposure to US, and an exposure length of 15 min was necessary
to 10 lm. The DSD of the emulsion samples can also influence the
to achieve maximum demulsification efficiency at the experimen-
efficiency of the demulsification process, since the coalescence
tal conditions employed in the present work.
rates and the sedimentation velocities of the droplets depend clo-
sely on the droplets’ diameters. Usually, emulsions with smaller
3.3. Evaluation of the effect of US on heavy crude oil emulsions with
water drops are more stable [4]. Therefore, experiments were per-
different water contents
formed using emulsions with D(0.5) of 5 and 25 lm (both with
water content of 35%). The results were compared with the previ-
Crude oil emulsions with different water contents and DSD are
ous results obtained for emulsions with D(0.5) of 10 lm and are
explored worldwide, and even within the same oil field, emulsions
shown in Table 2.
with different rheological characteristics could be produced. After
According to Table 2, a higher demulsification efficiency was
the optimization of the main US parameters, experiments were
obtained for emulsions with D(0.5) of 25 lm using 25 kHz. This
performed to evaluate the effect of US on emulsions with variable
result would be expected at all frequencies because, as discussed
water contents (lower and higher than 35%). For these experi-
before, emulsions with larger drops are usually less stable than
ments, the amplitude was set to 100% and the temperature was
emulsions formed by smaller drops. However, using 45 kHz,
set at 45 °C (frequencies of 25, 35 and 45 kHz were evaluated).
demulsification efficiency was higher for emulsions with D(0.5)
The time of exposure to US was evaluated again, because this
of 5 lm than for emulsions with D(0.5) of 25 lm. There is probably
parameter could have a different influence on emulsions according
a relationship between US frequency and droplet size. In this way,
to their water contents. Results are shown in Fig. 3.
depending on the frequency, the US field could affect the surfaces
Water separation from W/O emulsions with an initial water
of water droplets with different intensities, resulting in different
content of 12% and 50% was observed after 5 min of US application
demulsification efficiencies when emulsions with different D(0.5)
(Fig. 3). For the emulsion having an initial water content of 12%
are evaluated.
(Fig. 3A) and using a frequency of 25 kHz, demulsification effi-
Experimental results obtained in this work showed that US at
ciency was significantly lower after 30 min when compared to
frequencies up to 45 kHz has an important effect on the destabi-
15 min. On the other hand, at frequencies of 35 and 45 kHz, no sta-
lization of typical W/O emulsions normally explored by the crude
tistical difference was observed between the results obtained after
oil industry (with D(0.5) ranging on 5–25 lm). Turbulence and
15 min and 30 min (student t-test, minimum significance level for
shock waves caused by cavitation or acoustic streaming can prob-
P < 0.05). By using 45 kHz, the demulsification efficiency was
ably break the interfacial film that stabilizes the emulsion, and
higher and the final water content was 7.9 ± 0.05% after 15 min
immediately after this the same forces could bring water droplets
of US exposure (corresponding to a 36.3% demulsification effi-
together, leading to coalescence. However, more detailed experi-
ciency). For the emulsion with an initial water content of 50%
ments are necessary to confirm our hypothesis, using, for example,
(Fig. 3B), demulsification efficiency still increased after 15 min of
electronic microscopy and related techniques to evaluate the effect
US application. Therefore, it could be concluded that 30 min of
of US on W/O emulsions. Regarding the minimum time required to
exposure to US is necessary to achieve maximum demulsification
achieve the maximum demulsification efficiency, it was observed
efficiency. Due to the higher water content (50%) in comparison
that this parameter has to be evaluated individually according to
with emulsions containing 12% and 35% of water, a longer US expo-
the characteristics of each crude oil emulsion.
sure time is probably necessary to achieve maximum demulsifica-
tion efficiency. It is also possible to observe in Fig. 3B that similar
demulsification efficiencies were obtained at frequencies of 35 4. Conclusions
and 45 kHz (up to 65%), with the final water content for each emul-
sion being 16.8 ± 1.4% and 17.1 ± 3.5%, respectively. It is important The indirect application of US at frequencies of 25, 35 and
to mention that in all these experiments, no chemical demulsifiers 45 kHz to crude oil emulsions with variable characteristics allowed
were added. the demulsification after 15 min of exposure to US with efficiencies
546 F.G. Antes et al. / Ultrasonics Sonochemistry 35 (2017) 541–546

up to 60%. No demulsification was obtained for frequencies of centrifuged after passing between air of electrodes, U.S. Patent 2002166820
AI, 2002.
130 kHz or higher, leading to the conclusion that the indirect appli-
[10] L.O. Diehl, D.P. Moraes, F.G. Antes, J.S.F. Pereira, M.F.P. Santos, R.C.L. Guimarães,
cation of ultrasound causes the destabilization of W/O emulsions J.N.G. Paniz, E.M.M. Flores, Separation of heavy crude oil emulsions using
when frequencies up to 45 kHz are employed. The mechanical microwave radiation for further crude oil analysis, Sep. Sci. Technol. 1358–
effect of acoustic cavitation in aqueous systems increases when 1364 (2011).
[11] W. Tan, X.-G. Yang, X.-F. Tan, Study on demulsification of crude oil emulsions
the frequency of US decreases, so the effect of US on droplet surface by microwave chemical method, Sep. Sci. Technol. 42 (2007) 1367–1377.
and coalescence could be associated with the frequency employed. [12] E.A. Gardner, R.E. Apfel, Using acoustics to study and stimulate the coalescence
The results obtained in the present work are important for the pur- of oil drops surrounded by water, J. Colloid Interface Sci. 159 (1993) 226–237.
[13] Y. Guoxiang, L. Xiaoping, P. Fei, H. Pingfang, S. Xuan, Pretreatment of crude oil
poses of the crude oil industry because demulsification was by ultrasonic-electric united desalting and dewatering, Chin. J. Chem. Eng. 16
achieved for a heavy crude oil in a short amount of time (2008) 564–569.
(15 min) and at a relatively low temperature (45 °C) without using [14] R.J. Schoeppel, A.W. Howard, Effect of ultrasonic irradiation on coalescence
and separation of crude oil-water emulsions, in: SPE/Society of Petroleum
a chemical demulsifier. Therefore, low-frequency US could be a Engineers of AIME, Dallas, USA, 1966. paper 1507.
promising technology for industrial crude oil treatment. Although [15] S. Nii, S. Kikumoto, H. Tokuyama, Quantitative approach to ultrasound
the mechanism involved in the demulsification process is still emulsion separation, Ultrason. Sonochem. 16 (2009) 145–149.
[16] G.D. Pangu, D.L. Feke, Acoustically aided separation of oil droplets from
not clear, it is possible to suggest that the effect is related to the aqueous emulsions, Chem. Eng. Sci. 59 (2004) 3183–3193.
turbulence in the emulsion caused by cavitation or acoustic [17] H.G. Nasiri, Demulsification of gas oil/water emulsion via high-intensity
streaming. The turbulence could be intense enough to break the ultrasonic standing wave, J. Dispersion Sci. Technol. 34 (2013) 483–489.
[18] X.-G. Yang, W. Tan, X.-F. Tan, Demulsification of crude oil emulsion via
interfacial film that stabilizes the emulsion. Afterwards, US could
ultrasonic chemical method, Petrol. Sci. Technol. 27 (2009) 2010–2020.
bring water droplets together, causing coalescence. It is also possi- [19] P.E. Gramme, P. Gramme, E. Gramme, G.P. Eivind, Installation on sea bed, for
ble to conclude that to achieve maximum demulsification effi- separation of oil production fluids, comprises pipe separator that may form
ciency, exposure time has to be evaluated according to each part of transport pipeline, WO Patent 2003033872 A1 (24/04/2003).
[20] G. Ye, X. Lu, P. Han, F. Peng, Y. Wang, X. Shen, Application of ultrasound on
individual emulsion’s water content. Additionally, in the range of crude oil pretreatment, Chem. Eng. Process. 47 (2008) 2346–2350.
frequencies where demulsification was possible (25–45 kHz), dif- [21] G. Ye, X. Lu, P. Han, X. Shen, Desalting and dewatering of crude oil in ultrasonic
ferent efficiencies were obtained according to the median DSD. In standing wave field, J. Petrol. Sci. Eng. 70 (2010) 140–144.
[22] G. Ye, X. Lu, F. Peng, P. Han, X. Shen, Pretreatment of crude oil by ultrasonic-
this sense, it seems that 25 kHz is more suitable for emulsions with electric united desalting and dewatering, Chin. J. Chem. Eng. 16 (2008) 564–
D(0.5) of 25 lm, while 45 kHz enables higher demulsification effi- 569.
ciency for emulsions with D(0.5) of 5 lm. According to the author’s [23] E.C. Gaudino, D. Carnaroglio, L. Boffa, G. Cravotto, E.M. Moreira, M.A.G. Nunes,
V.L. Dressler, E.M.M. Flores, Efficient H2O2/CH3COOH oxidative
knowledge, this is the first time that a study involving the indirect desulfurization/denitrification of liquid fuels in sonochemical flow-reactors,
application of US at different frequencies was performed to evalu- Ultrason. Sonochem. 21 (2014) 283–288.
ate crude oil demulsification. [24] P.A. Mello, F.A. Duarte, M.A.G. Nunes, M.S. Alencar, E.M. Moreira, M. Korn, V.L.
Dressler, E.M.M. Flores, Ultrasound-assisted oxidative process for sulfur
removal from petroleum product feedstock, Ultrason. Sonochem. 16 (2009)
Acknowledgements 732–736.
[25] F.A. Duarte, P.A. Mello, C.A. Bizzi, M.A.G. Nunes, E.M. Moreira, M.S. Alencar, H.
N. Motta, V.L. Dressler, E.M.M. Flores, Sulfur removal from hydrotreated
The authors are grateful to CENPES/PETROBRAS for donation of
petroleum fractions using ultrasound-assisted oxidative desulfurization
samples and financial support and also to CNPq, CAPES and process, Fuel 90 (2011) 2158–2164.
FAPERGS. [26] F. Chemat, Zill-e-Huma, M.K. Khan, Applications of ultrasound in food
technology: processing, preservation and extraction, Ultrason. Sonochem.
813–835 (2011).
References [27] C.A. Bizzi, E.I. Müller, E.M.M. Flores, F.A. Duarte, M. Korn, M.A.G. Nunes, P.A.
Mello, V.L. Dressler, Handbook on Applications of ultrasound: Sonochemistry
[1] A. Bhardwaj, S. Hatland, Dynamics of emulsification and demulsification of for Sustainability, chapter 21: Ultrasound-Assisted Industrial Synthesis and
water in crude-oil emulsions, Ind. Eng. Chem. 33 (1994) 1271–1279. Processes, 535–580, Taylor & Francis Group, EUA, 2012.
[2] J. Djuve, X. Yang, I.J. Fjellanger, J. Sjöblom, J. Pelizzetti, Chemical destabilization [28] R.C.L Guimarães, A.R. Guarnieri, B.S.F. Machado, E.M.M. Flores, V.L. Dressler, D.
of crude oil based emulsions and asphaltene stabilized emulsions, Colloid. P. Moraes, F.G. Antes, J.S.F. Pereira, J.S. Barin, L.O. Diehl, P.B. Hein, Method for
Polym. Sci. 279 (2001) 232–239. separating phases of emulsions of the water-in-oil type by indirect ultrasound
[3] R.A. Mohammed, A.I. Bailey, P.F. Luckham, S.E. Taylor, Dewatering of crude oil application. WO Patent 2014008559 A1, 2012.
emulsions 1. Rheological behavior of the crude oil-water interface, Colloid. [29] F.G. Antes, L.O. Diehl, J.S.F. Pereira, R.C.L. Guimarães, R.A. Guarnieri, B.M.S.
Surf. A 80 (1993) 223–235. Ferreira, V.L. Dressler, E.M.M. Flores, Feasibility of low frequency ultrasound
[4] J. Sjoblom, N. Aske, I.H. Auflen, O. Brandal, T.E. Havre, O. Saether, A. Westvik, E. for water removal from crude oil emulsions, Ultrason. Sonochem. 25 (2015)
E. Johnsen, H. Kallevik, Our current understanding of water-in-crude oil 70–75.
emulsions. Recent characterization techniques and high pressure performance, [30] T.J. Mason, A.J. Cobley, J.E. Graves, D. Morgan, New evidence for the inverse
Adv. Colloid Interface 100 (2003) 399–473. dependence of mechanical and chemical effects on the frequency of
[5] M. Al-Otaibi, A. Elkamel, T. Al-Sahhaf, Experimental investigation of crude oil ultrasound, Ultrason. Sonochem. 18 (2011) 226–230.
desalting and dehydration, Chem. Eng. Commun. 65–82 (2003). [31] Standard test method for water in crude oil by potentiometric Karl Fischer
[6] Petrobras Procedure. N-2499, Dehydration of Crude Oil by Distillation. SC-20, titration, Annual Book of ASTM Standards, American Society for Testing and
Técnicas analíticas de laboratório, Rio de Janeiro, 1993. Materials, ASTM International, West Conshohocken, EUA, 2006.
[7] D.P. Moraes, F.G. Antes, J.S.F. Pereira, M.F.P. Santos, R.C.L. Guimarães, J.S. Barin, [32] K. Kumar, A.D. Nikolov, D.T. Wasan, Mechanisms of stabilization of water-in-
M.F. Mesko, J.N.G. Paniz, E.M.M. Flores, Microwave-assisted procedure for crude oil emulsions, Ind. Eng. Chem. Res. 40 (2001) 3009–3014.
salinity evaluation of heavy crude oil emulsions, Energy Fuels (2010) 2227– [33] J.D. McLean, P.K. Kilpatrick, Effects of asphaltene solvency on stability of
2232. water-in-crude-oil emulsions, J. Colloid Interface Sci. 189 (1997) 242–253.
[8] A. Ezzati, E. Gorouhi, T. Mohammodi, Separation of water in oil emulsions [34] M. Ashokkumar, Ultrasonic preparation of stable flax seed oil emulsions in
using microfiltration, Desalination 185 (2005) 371–382. dairy systems – Physicochemical characterization, Food Hydrocolloid 39
[9] C. Noik, J. Tray, C. Noiek, Effluent separation device comprises at least one (2014) 151–162.
centrifuging element including helical channel, where the effluent is