chemical engineering research and design 1 0 9 ( 2 0 1 6 ) 180–189

Contents lists available at ScienceDirect

Chemical Engineering Research and Design

journal homepage: www.elsevier.com/locate/cherd

SO2 removal by seawater in a spray tower:

Experimental study and mathematical modeling

Sh. Darake, M.S. Hatamipour ∗ , A. Rahimi, P. Hamzeloui

Chemical Engineering Department, College of Engineering, University of Isfahan, Hezar Jerib Blvd.,
81746-73441 Isfahan, Iran

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

Article history: In this article, the reactive absorption of SO2 by seawater is studied in a spray tower experi-
Received 14 May 2015 mentally and mathematically. The liquid film formation on the tower wall is implemented in
Received in revised form 26 the model and measured experimentally at different operating conditions. The effect of liq-
November 2015 uid to gas flow rate, initial SO2 concentration in gas phase and initial gas temperature on SO2
Accepted 30 November 2015 removal efficiency is examined. Regarding the importance of liquid droplets hydrodynamics
Available online 25 January 2016 and its effect on the performance of the equipment, the required differential equations for
predicting the trajectory and local velocity of droplets are also developed based on the noz-
Keywords: zle and spray characteristics and solved simultaneously with other governing equations. In
SO2 removal order to survey the effect of nozzle type on removal efficiency, two different types of noz-
Seawater zles are examined. Semi-empirical correlations are proposed for two different nozzles by
Spray tower using experimental data and droplets hydrodynamics model, to predict the amount and the
Flue gas desulfurization (FGD) variation of liquid film mass flow rate on the spray tower wall. Results indicate that neglect-
Mathematical modeling ing the liquid film formation leads to an average of 23% error in predicting the removal
efficiency when nozzle type 1 is used, while the calculated error of model by considering
the film formation is reduced to 4%. By implementation of droplets hydrodynamics model
and applying a modified thermodynamics model for predicting the behavior of the existing
chemical reactions, the capability of the spray tower model in predicting the SO2 removal
efficiency is enhanced.
© 2016 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1. Introduction
Recently, the international committees have adopted several
restrictions for depletion of the released SO2 from indus-
tries (Vidal et al., 2007). In the past few decades, studies
have focused dramatically on the flue gas desulphuriza-
tion (FGD) process. Meanwhile, different equipment such as
venturi scrubbers, spray towers, tray towers, packed beds
(Bandyopadhyay and Biswas, 2006, 2008; Gamisans et al., 2002),
and membrane technology (Sun et al., 2008) have been studied
for FGD process in a vast range.
The most common method for FGD is scrubbing the pol-
luted gas through alkali solvents like urea, dilute NaOH,
limestone slurry, NaCl solution, water, and seawater (Barbooti
et al., 2011; Bokotko et al., 2005; Jeong and Kim, 1997). Accord-
ing to searches in this respect, a few studies were conducted
on using seawater as an alkali absorbent for SO2 removal.
Sun et al. (2008) studied SO2 absorption by seawater in a
hollow fiber membrane contactor and found that the mass
transfer coefficient in seawater is about double the mass trans-
fer coefficient in the NaOH solution content with a pH of
8.35.

Abbreviations: FGD, flue gas desulfurization; ppm, parts per million by volume; ppt, parts per trillion.
∗
Corresponding author. Tel.: +98 313 7934047; fax: +98 313 7934031.
E-mail address: hatami@eng.ui.ac.ir (M.S. Hatamipour).
http://dx.doi.org/10.1016/j.cherd.2015.11.027
0263-8762/© 2016 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
 chemical engineering research and design 1 0 9 ( 2 0 1 6 ) 180–189 181

concentrations and types of the selected additives. Javed et al.

Nomenclature (2006) conducted an experimental study to investigate the
effect of imparting swirl to the axial gas flow on mass transfer
D diameter of tower (m) coefficient through a spray tower. They developed a corre-
d droplet diameter (m) lation for gas phase mass transfer coefficient as a function
Cpg heat capacity of gas (J/kg K) of the gas flow rate and initial droplet SMD1 . Their method
Cpl heat capacity of liquid (J/kg K) was based on the assumption of constant thickness of liq-
Cpv heat capacity of vapor (J/kg K) uid film. Bandyopadhyay and Biswas (2008) carried out an
Qg flow rate of gas (m3 /s) experimental study to identify the important factors affect-
Ql total flow rate of liquid ing on SO2 absorption in a spray tower and survey of liquid
Qd flow rate of droplet flow rate. Their founding revealed that in liquid to gas ratio of
Qf flow rate of liquid film 3 (L/m3 ) and pH of 11.7, the removal efficiency reaches 100% for
HSO2 Henry’s constant inlet gas with SO2 concentration in range of 500–1500 ppmv .
h heat transfer coefficient (W/m2 K) A model based on the penetration theory was developed to
I ionic strength calculate the dynamic absorption rate of sulfur dioxide by
kwd mass transfer coefficient for water between gas limestone slurry in a spray tower by Brogren and Karlsson
and droplet phase (mol/m2 s) (1997). In this model, instantaneous equilibrium reactions and
Kwf mass transfer coefficient for water between gas reaction with finite rate were considered. Bandyopadhyay and
and droplet phase (mol/m2 s) Biswas (2007) presented a simple realistic model in order to
Kgd mass transfer coefficient for SO2 between gas investigate the effectiveness parameters such as the size and
and droplet (mol/m2 s) velocity of droplet, superficial gas velocity, liquid flow rate
Kgf mass transfer coefficient for SO2 between gas and tower height on the performance of a spray absorption
and droplet (mol/m2 s) tower. Besides, the removal efficiency was strongly affected
ki thermodynamic equilibrium constant by spray tower hydrodynamics, flow rates and dimensions.
Mg molecular weight of gas (kg/kmol) Gao et al. (2008) proposed a CFD model for SO2 absorption
Vre relative velocity of droplet by limestone slurry in a spray scrubber. The gas and liquid
Ml molecular weight of liquid (kg/kmol) temperatures were adjusted at 298 K in this work. Besides,
mi concentration of ion they overlooked the liquid film formation onto the tower wall
Nsd molar flax of SO2 between gas and droplet and applied Rosin–Rammler rule for size distribution of initial
(mol/m2 s) droplets. Marocco (2010) developed a CFD model for prediction
Nsf molar flax of SO2 between gas and film of removal efficiency close to the nozzle zone at isothermal
(mol/m2 s) condition, where the interaction between flue gas and slurry
Nwd molar flax of water vapor from droplet is high in spray column. Zhuang et al. (2015) developed a CFD
(mol/m2 s) model for NO2 wet absorption including the chemical absorp-
Nwf molar flax of water vapor from liquid film tion by sodium sulfite in a spray tower.
(mol/m2 s) Rahimi et al. (2002) developed a mathematical model to
ys SO2 concentration at gas bulk study simultaneous heat and mass transfer in hot gas spray
yw water concentration at gas bulk system. Downing (1966) correlations were used to calculate
ywi water concentration at gas–liquid interfacial heat and mass transfer coefficients. Bozorgi et al. (2006) devel-
Zi ion charge oped a mathematical model by consideration of liquid film
CD drag coefficient formation to study the performance of a spray scrubber of an
T0 reference temperature industrial ammonium nitrate plant. Their results showed that
Td droplet temperature the assumption of terminal velocity for the motion of droplets,
Tf film temperature especially for smaller ones is a rough assumption. Keshavarz
et al. (2008) showed that the importance of the consideration of
Greek letters liquid film formation in a spray tower model depends on solu-
i activity coefficient for ion bility of component in gas phase. Their model showed that
Hr heat of SO2 solution in seawater (J/mol) neglecting the liquid film formation causes a smaller error
g density of gas (kg/m3 ) for pollutant with low solubility. In order to predict absorp-
l density of liquid (kg/m3 ) tion height in a spray tower, Zhu et al. (2014) investigated SO2
absorption by limestone–gypsum. They revealed that absorp-
tion height declines with increasing liquid–gas ratio and pH
Al-Enezi et al. (2001) measured the solubility of SO2 in value of the slurry, and increases with increasing droplet
the Persian Gulf seawater through a semi-continuous sys- diameter, gas flow rate, gas temperature, inlet SO2 concen-
tem. Rodriguez-Sevilla (2004) developed a thermodynamic tration and absorption efficiency, respectively.
model for SO2 solubility in seawater as a function of pH Vidal et al. (2007) carried out an experimental study on FGD
and the temperature of seawater (278.15–318.15 K). They used by use of seawater in a catalytic packed tower. They worked in
an extended version of the Debye–Huckel theory (1972) and the range of liquid to gas ratios of 4–15 (L/m3 ) and used active
the Pitzer ion-interaction model (Bromley, 1972; Pitzer, 1991; carbon in packed structure. They showed that a 47% reduction
Rodriguez-Sevilla, 2004). They found that the solubility of SO2 in seawater flow rate could be obtained in the catalytic packed
in seawater is within 20–60% and 6–30% higher than its sol- tower relative to conventionally spray tower.
ubility in distilled water and NaCl solution of similar ionic
strength, respectively. Zhang et al. (2011) investigated the
1
effect of additives on seawater FGD by introducing different Sauter mean diameter.
182 chemical engineering research and design 1 0 9 ( 2 0 1 6 ) 180–189
Ebrahimi et al. (2003) developed a rate-based model of a
counter current reactive process for the absorption of SO2 into
NaHCO3 /Na2 CO3 in a packed bed column. Schultes (1998) con-
ducted a theoretical study on reaction-kinetic process for SO2
in sodium hydroxide solution in packed-column. Their results
revealed that gas side mass transfer resistance is increased
with increasing enhancement factor. Chavez et al. (2012) con-
ducted an experimental study of SO2 absorption using water
in packed tower and represented some correlations for liquid
and gas phase mass transfers based on the experimental data.
As mentioned, several attempts have been made to
investigate SO2 absorption in a spray tower and other equip-
ment experimentally, but the mathematical modeling studies
related to the reactive SO2 absorption have been limited by
alkali solution in a spray tower and so these studies have not
been completed. For instance, Marocco (2010) has not consid-
ered the temperature effects in his commercially CFD code
for simulation of SO2 physical absorption in a spray tower.
In addition, the review of the existing literature reveals that
the number of studies on SO2 removal by seawater and its
scientific bases are limited.
In the present study, a thermodynamic model is used to
survey the reactive absorption of SO2 by seawater while the
non-instantaneous reaction is applied unlike previous works.
Applying a droplet hydrodynamics model in order to involve
liquid film formation in SO2 removal process in a spray tower,
improves the model accuracy as well. Liquid to gas flow rate,
initial SO2 concentration in gas phase and initial gas temper-
ature are the factors that their effects on performance of SO2
removal in a spray tower are examined in this research.
2. Materials and methods
A schematic diagram of the experimental setup of current
study is presented in Fig. 1. This setup consists of a stainless
steel tower (185 cm height, 0.16 cm diameter), air compressor,
an air heater, a water pump, a seawater tank, a SO2 cylinder
and digital indicators. Sulfur dioxide gas with 99.5% purity
from a high pressurized cylinder is used to supply the required
SO2 . The gas cylinder is equipped with a pressure regulator
(0–20 bar) and an air flowmeter (0.2–2 N L/min). A SO2 stream
is injected to the bottom of the tower and mixed completely
with an air stream before entering the reactive zone. An elec-
trical heater is used for air stream preheating to achieve the
required temperature. To avoid the heat loss, the external sur-
face of the tower and the air heater are insulated with 4 cm
thick glasswool.
The simulated seawater of 25 ◦ C ambient temperature is
pumped to the top of tower and is then sprayed into the
gas stream through a nozzle. A conductivity meter (JENWAY
model 4520, England) was used for measurement of the sim-
ulated seawater and pH of the solution is measured by a
pH-meter (model Istek pH-20 N, Korea). The components of
standard seawater at salinity of 35 g/kg are listed in Table 1.
Here, the main components are Na+ and Cl− . Regarding the
lack of access to seawater due to long distance from coast,
the artificial seawater is prepared in the laboratory. Sea salt
is added to water for preparing synthesized seawater with a
suitable conductivity (80 ms) and pH of 8.2 according to Al-
Enezi et al. (2001). Through a variable electrical power, the air
temperature is set at the desired values determined through
the experimental design. The gas temperature at tower exit
is measured by thermal sensors. An analyzer (Testo 350-XL,
Table 1 – The main ions existed in seawater of salinity
35 ppt (g/kg).
Constituent wt% Mmole/kg g/kg
Chloride 55.29 546 19.353
Sodium 30.74 569 10.76
Magnesium 3.69 53 1.292
Sulphate 7.75 28 2.712
Calcium 1.18 10.3 0.412
Potassium 1.14 10.2 0.399
Germany) is used to measure the SO2 concentration at the
tower outlet.
Here, the important point is measuring liquid film flow rate
as the liquid droplets collide to the tower wall. In order to mea-
sure the flow rate of liquid film on the tower wall, a 10 cm
wide, 50 cm long (equal to tower section circumstance) sheet
is welded to the tower inner surface at 25 degree, at 170 cm
from the nozzle at the tower top (Fig. 1). By fixing the outlet
valve on this location, the liquid film flow rate is measured.
According to dependence of spray angle to the total liquid flow
rate, at different liquid flow rates which enters to the nozzle,
different liquid film flow rates are obtained. It should be noted
that the value of liquid film flow rate is measured merely at
the 170 cm from the top of the tower. In addition, two nozzles
with different structures are used to survey the effect of noz-
zle type on liquid film formation as well as the SO2 removal
efficiency in the spray tower. The main difference between the
applied nozzles are the size of droplets and angle of spray at
different liquid flow rate. The diameters of nozzles 1 and 2 are
1200 and 1000 ␮m, respectively. Another important feature of
a nozzle is its 0-angle of spray at different liquid flow rates. In
order to obtain these angles versus liquid flow rate for each
nozzle in the tower, some experiments were carried out. For
calculation of these angles based on the total liquid flow rate
and liquid film formation experimental data at 170 cm from
the nozzle, the correlation developed by Bozorgi et al. (2006) is
used, which is mentioned in the following section.
3. Mathematical model
In current study, a general mathematical model is developed
to determine the non-isothermal absorption of SO2 by sea-
water accompanied with chemical reaction. In a spray tower,
majority of droplets collide to the tower wall and convert to
liquid film after leaving the nozzle and passing a specific dis-
tance. Heat and mass transfer between gas phase and liquid
droplets occur, before the change of free droplets to the liq-
uid film; while, after liquid film formation, liquid film is also
involved in heat and mass transfer phenomena. Therefore,
the removal efficiency could be affected by hydrodynamics of
droplets and liquid film. Droplets’ spraying hydrodynamic is
an important debatable issue in determining the mass trans-
fer rate in a spray tower. In order to calculate the velocity
of droplets in each spatial coordinate at isothermal state a
mathematical model is proposed.
3.1. Hydrodynamic model for dispersed phase
The schematic diagram of the droplets spray pattern consid-
ered in this model is shown in Fig. 2. This figure shows that the
injected droplets at each angle make specified surface (A (i )).
By solving the droplets dynamic equations, the amount of
dispersed phase (droplets) and liquid film in each location
 chemical engineering research and design 1 0 9 ( 2 0 1 6 ) 180–189 183

Fig. 1 – Schematic diagram of experimental setup.

By applying Newton’s second law for a droplet in the z direc-

tion, the equation of droplet axial velocity variation is obtained
through the following equation:

dVdz (l − g ) g 3 g CD (Vg + Vdz ) |Vre |

= − (1)
dt l 4 l d

Similarly, by applying the same law for a droplet in the r

direction, Eq. (2) is obtained for prediction of radial velocity of
droplet variation:

dVdr 3  
= CD g (Vdr ) Vre  (2)
dt 4l d

Fig. 2 – Schematic diagram of droplet spray from nozzle. Axial and radial velocity for each droplet at distinctive
spray angle is calculated through solving Eqs. (1) and (2), simul-
taneously. The distance that droplet has passed in z direction
is determined. In this study, the obtained results verify the at specified time after injection can be calculated by the fol-
fact that the droplets at each angle have different velocities; lowing equation:
therefore, heat and mass transfer coefficients are calculated
for each droplet. The proposed model for hydrodynamics of t
dispersed phase is based on the following assumptions: z= Vdz dt (3)
t=0

- Neglecting the droplet diameter reduction due to evapora-

While, Eq. (4) calculates distance radial:
tion
- Isothermal system
t
- Neglecting the droplets collision with each other
r= Vdr dt (4)
- A mean diameter size is used for all of the droplets
t=0

The first assumption shows that droplets diameter does when the calculated radial distance becomes equal to the
not have any variation due to evaporation along the tower radius of the tower, the droplet collides to the tower wall and
length in the hydrodynamic model. is converted to liquid film.
184 chemical engineering research and design 1 0 9 ( 2 0 1 6 ) 180–189

liquid phase, which causes to reduction in droplets diame-

ter. In order to express the model following assumptions are
considered:

- Steady state condition

- Adiabatic process
- Neglecting gas and liquid flow rate variations due to evapo-
ration
- Neglecting the droplets collision with one another
- A mean diameter size is used for all droplets at each section
of tower
- Constant thermo-physical properties (Sharqawy et al., 2010)
- Neglecting heat transfer resistance in the liquid

3.2.1. Mass balance equations

Fig. 3 – Schematic of liquid film calculation method.
Eq. (5) developed by Bozorgi et al. (2006) calculates amount
of the droplets that are injected in each spray angle (Fig. 2):
tan2 ((i /2) + (i /4)) − tan2 ((i /2) − (i /4))
Qi (i ) = Ql
tan2 (max /2)
The pollutant concentration in gas phase is transferred to the
droplets and liquid film along the height of the spray tower.
Therefore, the variation equation of pollutant concentration
in the gas phase is the main equation in the mathematical
modeling of a spray tower. The total area of droplets in the
gas phase and for an incremental length of the tower, which
(5) is available for heat and mass transfer, is obtained by the fol-
lowing equation:

3.1.1. Numerical solution of hydrodynamic model

The Droplets dynamic equations are solved in each section of


Qd (i ) d2 z
Ad = (7)
the tower. Equations are solved by Euler numerical method in (/6) d3 Vdz
=0
r and z directions. It should be noted that in each section of
tower, the complex of droplets is available in the angle range of The mass balance equation for SO2 in gas phase is
0 to  max degrees that are divided to the groups of  i extent. expressed by the following equation:
Therefore, the equations must be solved for entire droplets
between angle of 0 and  max .  
dys Mg max
Qd (i ) d2
Eq. (4) calculates the distance passed by each droplet group = Nsd + Nsf D (8)
dz g Qg Vdz (/6) d3
in r-direction. When the obtained value from Eq. (4) equals to =0
the tower radius, the travelled distance in z direction is calcu-
lated for each droplet group. If this value becomes more than where Nsd is the molar transfer flux of SO2 between gas and
the tower height it means that droplets doesn’t collide to the droplets, and Nsf is the molar transfer flux between gas and
tower wall. While if this value becomes less than tower height liquid film. These parameters are represented as Eqs. (9) and
it means that, the liquid film is formed on the tower wall. This (10), respectively:
value determines the collision location of each droplet group
to the tower wall. After calculating the droplet flow rate at NSd = Kgd (ys − mSO2 HSO2 ) (9)
each  i using Eq. (5), the liquid film flow rate at each spec-
ified height of the tower is obtained according to Eq. (6) and Nsf = Kgf (ys − mSO2 HSO2 ) (10)
Fig. 3 by adding the previous liquid film flow value and the
droplets flow rate at each angle that collide to the tower wall: Parameter for mSO2 is the solubility of SO2 in seawater.
  Darake et al. (2014), represented the method of solubility cal-
Qfj+1 = Qfj + Qj+1 j+1 (6) culation. At each longitudinal increment of the tower there is
a distinct equilibrium between two phases, which depends on
According to liquid film flow prediction, the model results equilibrium temperature, partial pressure of SO2 in gas phase
are obtained at different operating condition. By applying and seawater composition.
experimental data and model results, the semi-empirical Since the liquid is introduced into the gas phase by evap-
correlations are developed and presented in Section 4 as a oration, a mass balance is required for water vapor in the
function of operating parameters. gas phase. The mass balance equation for water vapor in gas
phase leads to the following equation:
3.2. Gas absorption model  
dyw Mg max
Qd (i ) d2
The mathematical modeling of spray column is developed =− Nwd + Nwf D (11)
dz g Qg Vdz (/6) d3
based on Euler-Lagrange approach. In this study, the mass =0

transfer resistance in gas and liquid phase, effect of droplet

where
and liquid film temperature, the evaporation rate effect on
gas phase and variation of droplet diameter are considered
due to liquid evaporation in the length of the tower. Despite Nwd = kwd (ywi − yw ) (12)
hydrodynamic model, the evaporation term is considered in
this part. The high gas temperature leads to evaporation of Nwf = kwf (ywi − yw ) (13)
 chemical engineering research and design 1 0 9 ( 2 0 1 6 ) 180–189 185

The Antoine equation in the range of 0–100 ◦ C is used to

Table 2 – Parameters and heat and mass transfer
calculate the equilibrium mole fraction of water (ywi ). coefficients correlations.
The diameter of droplets is emerged in Eqs. (8) and (11).
Reference Correlation
This factor is variable along the tower length. Therefore, the
 
total mass balance has been considered for droplets. Evap- Downing (1966) shdg = M 2 + 0.6Re0.5 Sc0.33
 
oration and reaction terms cause the variation in droplet Perry and Green (1997) KSO2 l = 10 DSO2 −H2 O,l /d
diameter. The effect of reaction term relative to evaporation Chun and Seban (1972) Shfg = 0.435Re−0.33
term is insignificant; hence, the effect of chemical reac- Perry and Green (1997) Shfl = 3.41
tion on droplets size variation is neglected. By applying Downing (1966) Nud =  
MN (1/B) Ln (1 + B) 2 + 0.6Re0.5 Pr0.33
Lagrangian approach, the equation of droplet diameter varia-
Chun and Seban (1972) Nuf = 0.435Re−0.33
tion is obtained as follows:
Haider and Levenspiel CD =  
(1989) (24/Re) × 1 + 0.19Re0.65 +
dd 2N M (0.44Re/ (7185 + Re))
= − wd l (14)
dz l Vdz

3.2.3. Numerical solution of gas absorption model

3.2.2. Energy balance equations
The obtained equations are dependent to each other, so, a
The heat effect of SO2 absorption by alkali solution in various
calculation code and a suitable solution method is required
gas scrubbers has not been a subject of studies as it should
for these first order equations. The equations are solved by
be (Marocco, 2010; Bandyopadhyay and Biswas, 2008). Vapor
Runge–Kutta method in MATLAB software environment. The
transmission from liquid surface to the gas phase is due to
tower is divided to 1 mm steps. According to availability of
simultaneous heat and mass transfer.
boundary conditions at the top of the tower, the equations
In this study, the authors have considered the heat effects
are solved from top to bottom. Inlet liquid temperature, pol-
associated with the chemical reaction and liquid evaporation
lutant concentration at the inlet liquid and the initial droplet
from droplets and liquid film by developing the energy balance
diameter are the available boundary conditions at the top of
equations in gas and liquid phases.
tower. While, the outlet gas temperature, pollutant concen-
The energy balance equation for gas phase is obtained by
tration at the outlet gas and the gas humidity are guessed
the following equation:
at the top of the tower. The initial diameter of droplets is
presumed equal to the nozzle mesh diameter. The hydro-
dTg Mg dynamic model equations accompanied with the equations
=
dz (Cpg + Cpv yw ) g Qg of gas absorption model are solved simultaneously. Hence, the

max
Qd (i )d2 1   values of gas temperature, pollutant concentration at the gas
× hd (Tg − Td ) + hf D Tg − Tf (15) phase and gas humidity are obtained. These iterative calcula-
(/6) d3 Vdz
tions are continued until converging these parameters. Table 2
=0  

max
Qd (i )d2
lists the required correlations for determining the heat and
+ Cpv (Tg − T0 ) Nwd + Nwf D mass transfer coefficients.
Vdz (/6) d3
=0

4. Results and discussion

The gas temperature changes due to convective heat trans-
fer and evaporation of liquid droplet and film.
In addition to convective heat transfer and evaporation
terms, the heat of reaction in liquid phase is involved in the
energy balance equation for calculating the droplet tempera-
ture. This equation is presented as follows:
dTd hd (Tg − Td ) + Nsd H − Nwd w0 + Cpl Nwd (Td − T0 )
=   (16)
dz Vdz l Cpl d/6Ml
In this section, the results of hydrodynamic model of droplets
spray in the tower are investigated and compared with the
measured data. Through this procedure, some semi-empirical
correlations are obtained for predicting the liquid film flow
rate. In Section 4.3, the effect of gas and liquid flow rate, SO2
concentration in gas phase and gas temperature on removal
efficiency are studied experimentally. It should be noted that
the removal efficiency is defined based on the following equa-
tion:

where the heat of reaction in the saturated solution is equal ys,inlet − yS,out
× 100 (18)
to 6.7 kCal/mol (Bandyopadhyay and Biswas, 2008). yS,inlet
The energy balance equation in liquid film is expressed by
the following equation: 4.1. Liquid phase hydrodynamics


dTf DMl   The droplet spray is considered as cone-shaped in this model.
= hf Tg − Tf − Nsf Hr Fig. 4 illustrates the variation in radial component of velocity
dz l Cpl Qf
 (Vdr ) with time, for a nozzle with spray angle of 90 degree. As
1 dQf l   the figure shows, after passing specific time the radial veloc-
+ C T − Tf − Nwf w0 (17)
D dz Ml pl d ity of droplets at entire angles will be reduced to zero because
of free fall of the droplets. The vertical component of velocity
  (Vdz ) is increased with time and then reaches a constant value
In the above equation, the term including dQf /dz is equal to the terminal velocity (Fig. 5). It is obvious that the
related to the increase of enthalpy in liquid film due to col- initial velocity value influences the required length for reach-
lision of droplets to the tower wall and their conversion to the ing the terminal velocity and has no effect on its value. The
liquid film. droplets which are injected at zero angle do not convert into
186 chemical engineering research and design 1 0 9 ( 2 0 1 6 ) 180–189
Droplet velocity in r-direction (m/sec)

1.5
Table 3 – Comparison between model results and
experimental data of liquid film at the distance of 1.7 m
pi/2 from nozzle 1.
1
pi/4 Liquid film percent

Total liquid Obtained from Eq. (19) Exp. data Error (%)
0.5 flow (L/min)

1 33.93 32.5 4.4

1.75 44.97 36.36 23.6
0 2.25 51.04 52.42 2.63
0 0.5 1 1.5 2 2.5 3 3.5 4
Time(sec) 2.75 54.46 58.25 3.07

Fig. 4 – Model prediction for droplet radial velocity variation 100

versus time.
80

Removal Efficiency
5
Droplet velocity in z-direction (m/sec)

4.5 60
4
40
3.5 With considering film formation, Model
3 Without considering film, Model
0 20 Experimental data
2.5 pi/2
pi/4
2 0
0 2 4 6 8 10 12 14 16 18 20
1.5 Liquid to Gas flow Ratio (Lit/m3)

1
0 0.25 0.5 0.75 1 1.25 1.5 Fig. 6 – The effect of Liquid to gas flow on SO2 removal
Time(sec)
efficiency in nozzle 1. (Experimental data and model
Fig. 5 – Model prediction for droplet axial velocity variation results, Gas flow: 12 m3 /h, Inlet gas temperature: 80 ◦ C,
versus time. Initial SO2 concentration in gas phase: 1500 ppm).

the liquid film; however, those injected at angles above zero
collide to the tower wall after travelling the specific length
and are converted to the liquid film. Applying a uniform and
constant value for droplets velocity at various angles intro-
duces error in mathematical modeling of absorption in the
spray towers; while the opposite is used in this model. Apply-
ing the spray hydrodynamics and liquid film prediction model
leads to an increase in accuracy of the model for calculating
the transfer coefficients and making a more realistic model.
4.1.1. Semi-empirical correlations for liquid film prediction
Experimental data indicated that in the total liquid flow range
(0.5–3 L/min), 30–60% of the total value is converted to liquid
film at the distance of 170 cm from the nozzles. This affects
strongly the SO2 removal efficiency prediction. Diameter and
angle of spray are the main characteristics of every nozzle. In
additions to nozzle structure, the two later properties depend
on liquid flow rate and pressure of nozzle. Since the noz-
zle pressure is constant during the experiments, the angle of
spray is merely a function of the total liquid flow rate. Based
on the experimental data of liquid film values at the distance
of 170 cm from each nozzle (nozzles 1 and 2), different total
liquid flows are obtained; the spray angles for each liquid flow
rate could be estimated by combining the experimental data
and hydrodynamics model results. Furthermore, the wrought
liquid film percentage could be correlated as a function of total
liquid flow (Ql ) and distance from nozzle (Z) for each nozzle.
The relation for wrought liquid film percentage for nozzle 1 is
expressed as:
%Qf = 25.16Ql 0.5035 Z0.5636
where the parameter Z0 in Eq. (19) is the location of liquid film
commencement in the tower:
Z0 = 1.547 − 0.5599Ql
The correlations for nozzle 2 are as follows:
%Qf = 18.98Z0.5963 Ql0.795 Z > Z0 (21)
Z0 = 1.09 − 0.2811Ql 0.5 < Ql < 3 (22)
In the above equations the Ql is expressed in L/min and Z
in meter. These correlations are appropriate for a spray tower
of the diameter of 16 cm and height of 185 cm.
Comparing the results obtained from the experiment with
the results obtained from the model, indicates an average error
of 8.4% (Table 3).
This proposed hydrodynamic model can be applied to study
the liquid film formation in towers of different sizes. In fact,
this procedure can be useful for different sizes of spray tower
and nozzles with different geometry. However, lack of experi-
mental data prevents the verification of the proposed model.
4.2. Gas absorption model validation
Liquid to gas ratio is the most important operating parameter
in the spray towers. The results of the model’s prediction for
liquid to gas phase ratio effect on removal efficiency of noz-
zles 1 and 2 that are illustrated in Figs. 6 and 7, respectively.
Model results are presented for two different cases including
model without and with considering liquid film formation.
These figures reveal the discrepancy between mathematical
models with and without film. Since the liquid film forma-
tion is not considered in the filmless model, more removal
efficiency is predicted. Indeed, in the presence of total liquid
Z > Z0 (19) in the droplet form, more contact area becomes available in
the tower for mass and heat transfer. The experimental data
reveal that there is a lower tendency for liquid film formation
in nozzle 2 in comparison with nozzle 1. Thus, the discrepancy
between the two models (with film and filmless) for nozzle 2 is
0.5 < Ql < 2.75 (20) lower than nozzle 1. The average absolute error of 4% and 7%
 chemical engineering research and design 1 0 9 ( 2 0 1 6 ) 180–189
100
80
Removal Efficiency
60
40
Without considering film, Model
With considering film formation, Model
20 Experimental data
0 0
0 2 4 6 8 10 12 14
Liquid to Gas flow Ratio (Lit/m3)
Fig. 7 – The effect of Liquid to gas flow on SO2 removal
efficiency in nozzle 2. (Experimental data and model
results, Gas flow: 12 m3 /h, Inlet gas temperature: 80 ◦ C,
Initial SO2 concentration in gas phase: 1500 ppm).
100
90
Model
Experimental Data
Removal Efficiency
80
70
60
50
40
30
800 1000 1200 1400 1600
Droplet Diameter (micrometer)
Fig. 8 – Variation of removal efficiency versus initial droplet
diameter (experimental data and model results, Liquid to
gas ratio: 12 L/m3 , Initial SO2 concentration in gas phase:
1500 ppm).
are observed for model prediction with consideration liquid
film formation in the spray tower for nozzles 1 and 2, respec-
tively. Neglecting the liquid film formation phenomena leads
to a 23% average error in prediction of removal efficiency when
nozzle 1 is used, and the same error is about 9% for nozzle 2.
The effect of droplets diameter on the SO2 removal effi-
ciency in a spray tower is shown in Fig. 8. It is revealed that
the removal efficiency is decreased as the droplets diameter
increases. Droplet diameter is a significant parameter in the
heat and mass transfer phenomena in the spray towers. It is
observed that a decrease in droplet diameter at a constant liq-
uid flow leads to an increase in contact surface area and the
removal efficiency.
It is worth mentioning that the droplets entertainment by
gas phase should be accounted for, since this issue becomes
more important as the droplets size is decreased. Here, it is
recommended that it is better to determine an optimum value
of droplet diameter according to tower dimensions and oper-
ational conditions.
The model prediction for gas phase temperature along the
tower length for three values of inlet temperature is shown
in Fig. 9. The gas temperature is decreased sharply as soon
as it enters the tower and reaches to 27 ◦ C that is its adia-
batic saturation temperature based on inlet conditions. The
model prediction for the adiabatic saturation temperature is
in good agreement with that of the literature (Marocco, 2010).
It should be noted that reaching this temperature does not
take place just as soon as the gas enters the tower. In fact,
in the lower parts of the tower where much of the separa-
tion process takes places, the gas temperature does not reach
187
600
550 Tginlet= 590 K
Gas Temperature (K)
Tginlet= 400 K
Tginlet= 310 K
500
450
400
350
300
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
16 18 20 Distnace from top of the tower (m)
Fig. 9 – Inlet gas temperature variations passing through
spray column. (Initial SO2 concentration in gas phase:
1500 ppm, Inlet liquid flow temperature: 25 ◦ C, Liquid to gas
flow ratio: 12.5 L/m3 ).
1800 2000
Fig. 10 – Experimental data of variation of removal
efficiency versus liquid to gas flow for nozzles 1 and 2.
(Liquid to gas flow ratio: 12 L/m3 , Inlet gas temperature:
80 ◦ C, Initial SO2 concentration in gas phase: 1500 ppm).
its final value (the adiabatic saturation temperature); there-
fore, the non-equilibrium heat and mass exchange between
liquid and gas phases would affect the gas absorption pro-
cess. This phenomenon is especially important in high inlet
gas temperature situation in absorption process.
4.3. Parametric study
One of the objectives of this study is to investigate the nozzle
type effect on the removal efficiency. The effect of liquid flow
rate on the removal efficiency for two types of nozzle is shown
in Fig. 10. Droplet velocity is increased with an increase in
liquid flow rate and this leads to an increase in mass transfer
coefficients and SO2 removal efficiency. The removal efficiency
is increased sharply with an increase in liquid flow rate and
reaches almost a constant value.
According to the above figure, for the liquid flow rate more
than 1.5 L/min the removal efficiency of nozzle 2 is more than
that of nozzle 1. Experimental data indicate that due to the
nozzle structure, at this liquid flow range, the wrought liquid
film on the tower wall for nozzle 2 is lower than that for nozzle
1. Furthermore, the produced droplets from the nozzle 2 are
smaller than that from the nozzle 1. The two later reasons lead
to more SO2 removal through nozzle 2, indicating the better
performance.
For the liquid range of 1–1.5 L/min, the experimental data
indicated that merely 31–36% of the total liquid convert to liq-
uid film in nozzle 1. While, this value is about 36–40% in the
nozzle 2. By comparing these values with each other, it could
be concluded that less liquid film is formed in the tower with
nozzle 1. Therefore, it is reasonable that in this liquid flow
188 chemical engineering research and design 1 0 9 ( 2 0 1 6 ) 180–189
Fig. 11 – Experimental data of variation of removal
efficiency versus liquid flow at various gas flow for nozzle
1. (Inlet gas temperature: 80 ◦ C, Initial SO2 concentration in
gas phase: 1500 ppm).
Fig. 12 – Experimental data of variation of removal
efficiency versus gas flow rate at various inlet2
concentration in gas phase for nozzle 1. (Liquid flow rate:
2.5 L/min, Inlet gas temperature: 80 ◦ C).
rate range, nozzle 1 be more efficient than nozzle 2. In fact,
structure of nozzles merely leads to these results.
The SO2 removal efficiency variation versus liquid flow rate
at two gas flow rates of 8 and 12 m3 /h for nozzle 1 are shown in
Fig. 11. By increasing liquid flow from 1 to 4 L/min the removal
efficiency is increased from 43% to 90% at 12 m3 /h gas flow rate
and at 8 m3 /h gas flow rate the removal efficiency reaches 93%
from 55%. This figure shows that the increase in gas flow rate
reduces the removal efficiency. Increasing the gas flow rate
leads to a decrease in the gas residence time and reduces the
overall mass transfer. By comparing these values, it is found
that the liquid flow rate is a more important parameter than
the gas flow rate in the absorption process in a spray tower.
The gas flow effect on removal efficiency of SO2 concentra-
tion of 500 and 1500 ppm for nozzle 1 is shown in Fig. 12. By
variation of gas flow rate from 6 to 18 m3 /h, the SO2 removal
efficiency for concentration of 1500 ppm and 500 ppm are
reduced by 18% and about 16%, respectively. Here, it is revealed
that the gas flow rate is a more effective parameter than SO2
concentration on removal efficiency process.
The effect of inlet gas temperature on removal efficiency
is shown in Fig. 13. By increasing the inlet gas temperature,
the removal efficiency decreases. At lower gas temperatures,
the liquid evaporation and water vapor mass transfer to gas
phase increase the mass transfer boundary layer thickness,
which in turn leads to an increase in diffusion resistance
against SO2 mass transfer from gas to liquid phase. In fact,
at this temperature range the negative effect of evaporation
82
Model result
80
Removal Efficiency
Experimental data
78
76
74
72
50 100 150 200 250 300 350 400 450
Inlet gas temperature (C)
Fig. 13 – The effect of inlet gas temperature on removal
efficiency in nozzle 1. (Experimental data and model
results, Liquid to gas flow ratio: 12.5 L/m3 , Initial SO2
concentration in gas phase: 1200 ppm).
term overcomes the positive effect of chemical reaction. While
at higher temperatures, the effect of chemical reaction neu-
tralizes the negative effect of evaporation. This temperature
behavior illustrates that the SO2 reaction with seawater has
high activation energy. In this type of reactions, the chemical
reaction effects at higher temperatures are more than that at
lower temperatures.
5. Conclusion
Experimental and theoretical study of sulfur dioxide removal
by seawater is carried out in a spray tower. In order to
calculate the liquid film amount on the spray tower wall, semi-
empirical correlations are presented using experimental data
and hydrodynamic model results. The model predictions show
a good agreement with the obtained experimental data in a
spray tower. Accuracy of the spray tower mathematical model
by considering liquid film formation is investigated for two
different nozzles. Produced droplets by nozzle 1 are larger and
more liquid film is accumulated on the tower wall. Hence,
neglecting liquid film formation in a spray tower with nozzle 1
leads to 16% more error in removal efficiency prediction than
that of nozzle 2. According to the experimental data, the effec-
tive factors in order of importance are the liquid flow, gas flow,
SO2 concentration in gas phase and the inlet gas tempera-
ture. At high gas temperatures, an increase in evaporation rate
leads to an increase in resistance against SO2 transfer from gas
to liquid phase up to a distinctive temperature. The results of
this study (some removal efficiency data that are more than
90) confirm the suitability of seawater for SO2 absorption in
spray towers, especially in the coastal areas.
References
Al-Enezi, G., Ettouney, H., El-Dessouky, H., Fawzi, N., 2001.
Solubility of sulfur dioxide in seawater. Ind. Eng. Chem. Res.
40 (5), 1434–1441.
Bandyopadhyay, A., Biswas, M.N., 2006. Prediction of the removal
efficiency of a novel two-stage hybrid scrubber for flue gas
desulfurization. Chem. Eng. Technol. 29 (1), 130–145.
Bandyopadhyay, A., Biswas, M.N., 2007. Modeling of SO2
scrubbing in spray towers. Sci. Total Environ 383, 25–40.
Bandyopadhyay, A., Biswas, M.N., 2008. Critical flow atomizer in
SO2 spray scrubbing. Chem. Eng. J. 139 (1), 29–41.
Barbooti, M.M., Ibraheem, N.K., Ankosh, A.H., 2011. Removal of
nitrogen dioxide and sulfur dioxide from air streams by
absorption in urea solution. J. Environ. Prot. 2 (2), 175–185.
 chemical engineering research and design 1 0 9 ( 2 0 1 6 ) 180–189
Bokotko, R.P., Hupka, J., Miller, J.D., 2005. Flue gas treatment for
SO2 removal with air-sparged hydrocyclone technology.
Environ. Sci. Technol. 39 (4), 1184–1189.
Bozorgi, Y., Keshavarz, P., Taheri, M., Fathikaljahi, J., 2006.
Simulation of a spray scrubber performance with
Eulerian/Lagrangian approach in the aerosol removing
process. J. Hazard. Mater., B 137, 509–517.
Bromley, L.A., 1972. Approximate individual ion values of ␤ (or B)
in extended Debye-Huckel theory for uni-univalent aqueous
solutions at 298.15 K. J. Chem. Thermodyn. 4 (5), 669–673.
Brogren, C., Karlsson, T., 1997. Modeling the absorption of SO2 in
a spray tower scrubber using the penetration theory. Chem.
Eng. Sci. 25 (18), 3085–3099.
Chavez, R., Flores-Alamo, N., Guadarrama, J., 2012. Experimental
evaluation of sulfur dioxide absorption in water using
structured packing. Int. J. Chem. Eng. 2012, http://dx.doi.org/
10.1155/2012/579381, Article ID: 579381, 6 pages.
Chun, R.K., Seban, A.R., 1972. J. Heat Transf. 94 (4), 432–436.
Darake, Sh., Rahimi, A., Hatamipour, M-S., Hamzehlooie, P., 2014.
Sep. Sci. Technol. 49 (7), 988–999.
Downing, C.G., 1966. The evaporation of drops of pure liquids at
elevated temperature. Am. Inst. Chm. Eng. J. 2 (4), 760–766.
Ebrahimi, S., Picioreanu, C., Kleerebezem, R., Heijnen, J., Van
Loosdrecht, M., 2003. Rate-based modeling of SO2 absorption
into aqueous NaHCO3 /Na2 CO3 solutions accompanied by the
desorption of CO2 . Chem. Eng. Sci. 58, 3589–3600.
Gamisans, X., SarrÃ, M., Lafuente, F.J., 2002. Gas pollutants
removal in a single- and two-stage ejector-venturi scrubber. J.
Hazard. Mater. 90 (3), 251–266.
Gao, X., Huo, W., Luo, Z., Cen, K., 2008. CFD simulation with
enhancement factor of sulfur dioxide absorption in the spray
scrubber. J. Zhejiang Univ. Sci. 9 (11), 1601–1613.
Haider, A., Levenspiel, O., 1989. Powder Technol. 58 (1), 63–70.
Javed, K.H., Mahmud, T., Purba, E., 2006. Enhancement of mass
transfer in a spray tower using swirling gas flow. Chem. Eng.
Res. Des. 84 (6), 465–477.
Jeong, S.M., Kim, S.D., 1997. Enhancement of the SO2 sorption
capacity of CuO/␥-Al2 O3 sorbent by an alkali-salt promoter.
Ind. Eng. Chem. Res. 36 (12), 5425–5431.
189
Keshavarz, P., Bozorgi, Y., Fathikaljahi, J., Taheri, M., 2008.
Prediction of the spray scrubbers performance in the gaseous
and particular removing. Chem. Eng. J. 140, 22–31.
Marocco, L., 2010. Modeling of the fluid dynamics and SO2
absorption in a gas–liquid reactor. Chem. Eng. J. 162, 217–226.
Perry, R.H., Green, D.W., 1997. Perry’s Chemical Engineering Hand
book, seventh ed. Mc Graw-Hill, New York, NY.
Pitzer, K.S., 1991. Ion interaction approach: theory and data
correlation. In: Pitzer, K.S. (Ed.), Activity Coefficients in
Electrolyte Solutions. CRC Press, Boca Raton, FL.
Rahimi, A., Taheri, M., Fathikalaji, J., 2002. Mathematical
modeling of heat and mass transfer in hot gas spray systems.
Chem. Eng. Commun. 189, 959–973.
Rodriguez-Sevilla, J., 2004. Absorption equilibria of dilute SO2 in
seawater. J. Chem. Eng. Data 49 (6), 1710–1716.
Schultes, M., 1998. Absorption of sulphur dioxide with sodium
hydroxide solution in packed columns. Chem. Eng. Technol.
21 (2), 201–209.
Sharqawy, M.H., Lienhard, J.H., Zubair, S.M., 2010.
Thermophysical properties of seawater: a review of existing
correlations and data. Desalin. Water Treat. 16 (1-3), 354–380.
Sun, X., Meng, F., Yang, F., 2008. Application of seawater to
enhance SO2 removal from simulated flue gas through hollow
fiber membrane contactor. J. Membr. Sci. 312, 6–14.
Vidal, B., Ollero, P., Gutierrez Ortiz, J., Villanueva, A., 2007.
Catalytic seawater flue gas desulfurization process: an
experimental pilot plant study. Environ. Sci. Technol. 41 (20),
7114–7119.
Zhang, Y., Gao, Y., Zhou, J., Wang, G., Li, C., 2011. Effect of
additives on seawater flue gas desulfurization. In: Proc. Int.
Conf. Environ. Eng., Singapore.
Zhu, J., Ye, S., Bai, J., Wu, Z., Liu, Z., Yang, Y., 2014. A concise
algorithm for calculating absorption height in spray tower for
wet limestone–gypsum flue gas desulfurization. Fuel Process.
Technol. 129, 15–23.
Zhuang, Z., Sun Ch Zhao, N., Wang, H., Wu, Z., 2015. Numerical
simulation of NO2 absorption using sodium sulfite in a spray
tower. J. Chem. Technol. Biotechnol., http://dx.doi.org/10.1002/
jctb.4669.