Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353

Contents lists available at ScienceDirect

Process Safety and Environmental Protection

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

Non-isothermal modeling of simultaneous CO2

and SO2 removal in a semi-dry spouted bed reactor

Mohamad Ali Fakhari, Amir Rahimi, Mohammad Sadegh Hatamipour ,

Ali Fozooni
Chemical Engineering Department, College of Engineering, University of Isfahan, Hezar Jerib Blvd., Isfahan, Iran

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

Article history: In the present study, a comprehensive non-isothermal model is developed to study the
Received 19 March 2015 performance of a spouted bed reactor (SBR), in which CO2 is removed at the presence of SO2
Received in revised form 19 August by using NaOH solution. For this aim, the stream-tube model is applied for hydrodynamics
2015 of solid and gas phases, and then by using the conservation laws of mass and energy, the
Accepted 4 September 2015 governing equations for gas and solid phases are derived and solved numerically. The effects
Available online 11 September 2015 of variation of different operating parameters and process conditions are evaluated, and by
comparing the model results with the gathered experimental data, the maximum, minimum
Keywords: and average error are obtained. The results indicate that the CO2 removal efciency increases
Non-isothermal mathematical by increasing the inlet CO2 concentration and by decreasing the inlet SO2 concentration,
model ratio of supercial gas velocity to minimum spouting velocity and inlet gas temperature.
Spouted bed reactor Also, the modeling overall results indicate that by increasing the bed diameter and static
Simultaneous CO2 + SO2 removal bed height, CO2 absorption efciency increases.
streamtubes 2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1. Introduction Using spouting beds is a good alternative to existing pro-

The emissions of CO2 and SO2 from power plants that burn fos-
sil fuels cause changes in the global climate and form acid rain
in the atmosphere, respectively, and result in serious environ-
mental problems. CO2 is the major atmospheric contaminant
gas leading to increase of the earth temperature. Acid rain
and the acidication of the environment emerged as a seri-
ous global problem during recent decades (Haghnegahdar
et al., 2010; Rahimi et al., 2010). Therefore, development of
new technologies to reduce or eliminate the emissions of
carbon dioxide and sulfur dioxide into the environment is
essential.
Carbon dioxide is usually removed by wet scrubbers, u-
idized and xed beds and spray towers. Various technologies
for ue gas desulfurization FGD can be classied into three dif-
ferent types: wet scrubbers, semi-dry processes such as spray
drying and sorbent injection, and dry processes (Ma et al.,
2001; Xu et al., 2000).
cesses for the removal processes of acidic gaseous pollutants
from combustion gases. In this process, a slurry sorbent con-
taining very ne absorbent particles such as soluble alkali
powders (like NaOH, Ca(OH)2 and Mg(OH)2 ) is fed continu-
ously into a spouted bed in which coarse particles are spouted
with hot gas containing CO2 and SO2 . The reaction between
CO2 /SO2 and the sorbent and the drying of the slurry take place
in the bed simultaneously. Finally, the dried and reacted sor-
bents are entrained out of the bed and then collected in a bag
lter (Ma et al., 2000; Moeini and Hatamipour, 2008; Nakazato
et al., 2004).
Several investigators have already used spouted beds to
remove acidic gaseous pollutants showed higher efciency
of SBR in removing gaseous pollutants than conventional
processes (Haghnegahdar et al., 2010; Ma et al., 2000, 2001;
Nakazato et al., 2004; Xu et al., 2000).
Mathematical modeling of spouting beds was rst
proposed by Mathur and Lim (1974). They dened a

Corresponding author. Tel.: +00 98 31 3793 4047; fax: +00 98 31 3793 4031.
E-mail address: hatami@eng.ui.ac.ir (M.S. Hatamipour).
http://dx.doi.org/10.1016/j.psep.2015.09.001
0957-5820/ 2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353 343

Nomenclature
s spout region
a specic surface area of solids per unit volume SO2 SO2 concentration
(m2 /m3 ) t total
A cross sectional area (m2 ) T terminal
Ar Archimedes number () v water vapor
Cp specic heat (J/kg/K) w Water
CO2,in inlet CO2 concentration (ppm)
D Diameter (m) Superscripts
E CO2 removal efciency (%) * saturation
g gravity constant (m/s2 )
G mass ux of dry air (kg/m2 ) Greek letters
h convective heat transfer coefcient (w/m2 /s) porosity constant ()
H height(m) hc enthalpy of reactions between CO2 NaOH (J/kg)
Hc static bed height (cm) hcs enthalpy of reactions between SO2 NaOH (J/kg)
Hm maximum spoutable bed height (m) T approach to saturation temperature of the bed
K streamtube numerator in annulus () (K)
ky mass transfer coefcient of water vapor (m/s) z longitudinal element (m)
m mass of water on the surface of solid particle porosity ()
(kg) H enthalpy (J/kg)
Na/C molar ratio of NaOH to CO2 ()  heat of vaporization (J/kg)
Nu Nusselt number ()  viscosity (Pa s)
Pr Prandtl number ()  density (kg/m3 )
q convective heat transfer (w/m2 )
Qg,in inlet gas ow rate (m3 /h)
r thickness of liquid on particles (m)
rc ux of CO2 consumption (kg/m2 /s)
rcs ux of SO2 consumption (kg/m2 /s)
rw water evaporation ux (kg/m2 /s) one-dimensional isothermal model considering spout and
Re Reynolds number () annulus regions and assuming that the gas travels in plug ow
S axial mass ux of solid particles (kg/m2 /s) in both regions. Recently some researchers developed math-
Sc Schmidt number () ematical models to analyze the performance capability of a
Sh Sherwood number () spouted bed in various processes (Ghalavand et al., 2010, 2012;
SO2,in inlet SO2 concentration (ppm) Haghnegahdar et al., 2011; Jeng et al., 2001; Liu and Kato, 2000;
t Time (s) Moeini and Hatamipour, 2008; Nieto et al., 2007; Niksiar et al.,
T Temperature (K) 2013; Olazar et al., 2005; Rahimi et al., 2010; Sanchez et al.,
U gas velocity (m/s) 2000; Silva et al., 2011; Tao et al., 2010). Haghnegahdar et al.
V average axial velocity of the solid particles (m/s) (2011) proposed a non-isothermal mathematical model for the
W mass ow rate (kg/s) performance of spouted bed in chemical absorption of CO2 .
X CO2 concentration (ppm) Their model was based on stream tubes model and mass and
y mass fraction (ppm) energy conservation equations. Besides to spout and annu-
Y mass fraction of water vapor (kg H2 O/kg dry air) lus regions, their model included an additional area called
z vertical distance from gas inlet (m) fountain. They investigated the effect of some of operat-
Z SO2 concentration (ppm) ing parameters such as supercial gas velocity, Ca/C ratio and
static height of coarse particles bed on CO2 absorption ef-
Subscripts ciency and then compared the predictions of their model with
0 Reference experimental results.
a annulus region In ue gas of power plants, reneries, petrochemical and
air dry air other heavy industries, both CO2 and SO2 gases present. These
c Column two gases react simultaneously with sorbent slurry and sig-
CO2 CO2 concentration nicantly impact to each other removal efciency. To date no
f fountain region mathematical modeling has been reported to investigate the
g gas phase simultaneous absorption of CO2 and SO2 reaction in spouted
in Inlet bed reactors.
k number of stream tubes Hence a non-isothermal reactive absorption mathematical
mf minimum uidization model is developed in the present study to analyze mass and
mix gas mixture heat transfer between the phases for CO2 removal in presence
ms minimum spouting of SO2 from ue gas. The stream tubes hydrodynamic model
out Outlet (Sanchez et al., 2000) was used, and the governing equations of
p Particle gas and solid phases in different areas of the spouted bed were
pg particle to gas obtained. Also the effect of operational and design parame-
r radial direction ters such as column diameter and static bed height on the
performance of SBR was investigated.
344 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353
Fig. 1 Schematic diagram of the experimental apparatus.
2. Process description
Spouted bed is used for Geldart D type particles and uidiza-
tion is not effective (Mathur and Epstein, 1974). In this process,
sorbent solution which contains NaOH is fed drop wise man-
ner into the spouted bed of coarse particles. The spouting uid
is ue gas containing CO2 and SO2 . The coarse particles sup-
port slurry lm and enhance surface area for contacting the
reactants (Lin et al., 2008).
The reactor is a glass cylinder column (Pyrex) with 850 mm
height, 94 mm internal diameter and 3 mm wall thickness. The
lower cone-shaped part had a 45 angle and 68 mm height and
its orice inlet diameter was 7 mm. To prevent heat loss, the
bed wall was covered with an insulator. Spherical glass parti-
cles of 0.50.8 mm diameter with average diameter of 0.65 mm
and density of 2400 kg/m3 were used as spouted medium. Air
supplied by a central air compressor, was passed through an
oil lter and mixed with CO2 and SO2 from cylinders to form
the (synthetic) ue gas. Commercial CO2 and SO2 gases with
>99% purity were purchased from the market in 20 and 40 kg
cylinders, respectively. Supercial gas velocity in the reactor
was regulated by a velocity meter (Testo, 615), and CO2 and
SO2 ow rates were adjusted with rotameters. Mixture of air,
CO2 and SO2 after preheating by an electronic heater was fed
into the SBR from the bottom. CO2 and SO2 concentrations at
the inlet and outlet of SBR were measured by an infra-red CO2
gas monitor (03000 ppm, Guardian Plus, Edinburgh Instru-
ments Limited) and digital SO2 analyzer (03000 ppm, Testo
350-XL). The temperature of the inlet gas was measured by
a thermocouple at the reactor entrance and was controlled
by an On/Off controller (TPT979H model, Terminator Com-
pany). Pelleted sodium hydroxide (NaOH) with purity of 99%
was purchased from Merck agent in Iran. Sodium hydrox-
ide solution was continuously fed into the spouted bed by
a dosing pump (DLX MA/AD, ETATRON D.S. manufactured
by Etatron Company). The slurry feed line was suspended
in the reactor at the height of 60 cm from the reactor inlet.
Schematic diagram of the experimental apparatus is shown in
Fig. 1.
3. Mathematical model
In this model, the non-uniformity of spouted bed reactor is
considered by dividing the bed to three homogenous regions
based on hydrodynamic properties and porosity (Sanchez
et al., 2000). Three regions include:
(a) Dilute central core (spout region) with the upward ow of
gas and particles.
(b) Dense annular (annulus region) with upward gas ow and
downward ow of particles.
(c) Upper dilute region (fountain region) in which the outlet
solid particles are returned from the spouted region into
the annulus region. In this region the gas and solid phase
are in the form of a mixed ow reactor.
A schematic representation and characterizations of
stream-tube model is shown in Fig. 2.
In the present study, the equations are extended based
on hydrodynamic model of Sanchez et al. (2000) with the
exception that due to excessive mixing, fountain region is con-
sidered as a mixed ow reactor. The basic assumptions of this
model are as follows:
(a) In the annulus region the stream tubes model is used. Each
tube is considered as a plug ow reactor with counter-
current streams of gas and particles. No dispersion and
no mixing occur radially between stream tubes or axially
along the stream-tube length.
(b) In the spout region, the gas ow is considered to be plug
ow.
(c) All of outlet gas of spout region enters the fountain.
(d) Pressure in all parts of the reactor is considered constant.
(e) In the annulus region, heat transfer between adjacent
stream tubes is performed through convection.
(f) Heat of reaction is assigned to the solid particles. This
energy is then exchanged between gas and solid by heat
conduction mechanism in all regions.
 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353
Fig. 2 Stream tube model scheme.
(g) Because of the turbulent nature of gas phase in the SBR,
no mass transfer resistance is considered between gas and
solid phases.
(h) Because of low temperature variations of solid coarse
particles, solid thermo-physical properties are assumed
constant.
(i) Inlet slurry to the reactor is distributed among the parti-
cles of fountain region equally.
(j) Variations of mass caused by chemical reaction
are neglected (Hook et al., 1992; Sanchez et al.,
2000).
(k) In the fountain region, the gas ow is considered as in a
mixed ow reactor.
3.1. Governing equations
According to the proposed assumptions and by using mass
and energy conservation laws, governing equations of the
system performance can be extracted from each region sep-
arately. As previously mentioned, annulus region is divided
into several stream tubes and each stream tube consists of an
annular tube with countercurrent gas and solid ows equiv-
alent to a plug ow reactor. Thus, for each stream tube,
mass and energy balance equations are separately devel-
oped. In this region, a longitudinal element is used in each
stream tube (Fig. 3c). Also, in the fountain region, because
of excessive mixing, mass and energy balance equations in
this region are derived assuming this region as a mixed ow
rector. (Fig. 3b). In the spout region, according to cocurrent
gas and solid particles ow and by considering a longitudi-
nal element (Fig. 3a) mass balance equations for CO2 , SO2 ,
water vapor and liquid water on the coarse particles sur-
face and energy equation for gas and solid phases is derived.
In the derived equations, subscript s represents the spout
region.
345
3.1.1. Total gas mass balance
If Gst be mass ux of gas (dry air + water vapor + CO2 + SO2 ) and
the Gs be mass ux of dry air (which is constant throughout
the tower). The following relations are derived:
Gst = Gs + Gs Y + Gs X + Gs Z = Gs (1 + Y + X + Z) (1)
where X, Y and Z are mass ratio of CO2 , mass ratio of water
vapor, mass ratio of SO2 , respectively. Thus we have:
Gs A |z Gs A |z + z Gsr d (Ar ) = 0 (2)
d Gsr
(Gs ) = (3)
dz D/4
where Gsr is radial mass ux of dry air and A and d(Ar ) are
normal and lateral cross sectional area of the spout and D is
diameter of the spout.
3.1.2. Mass balance equation for CO2
Under steady state condition, CO2 mass balance equation for
the element dz in spout can be written as follows:
Gs Xs A |z Gs Xs A |z + z Gsr Xs d (Ar ) rc aAz = 0 (4)
where a is the specic surface area of solids per unit volume
of the spout region and rc is rate of CO2 consumption per unit
area of sorbent. After combining Eqs. (3) and (4), the resultant
mass balance equation for CO2 in spout is as follows:
dXs rc a
= (5)
dz Gs
3.1.3. Mass balance equation for SO2
Under steady state condition, SO2 mass balance equation for
the element dz in spout can be written as follows:
Gs Zs A |z Gs Zs A |z + z Gsr Zs d (Ar ) rcs aAz = 0 (6)
where rcs is rate of SO2 consumption per unit area of sorbent.
After combining Eqs. (3) and (6), the resultant mass balance
equation for SO2 in spout is as follow:
dZs rcs a
= (7)
dz Gs
3.1.4. Mass balance equation for water vapor
As the inlet gas is almost dry and moisture source is due to
inlet NaOH solution, the amount of water on solid coarse par-
ticles is low close to the reactor inlet and even in some points
there is no water in the gas phase for evaporation. So, there is
a humidity distribution in the gas and solid particles surface
at different locations on the bed. The reaction rate constant
is inuenced by gas humidity. For this purpose, by applying
the humidity mass balance on the desired element, the water
evaporation rate can be arrived as follows:
Gs Ys A|z Gs Ys A|z+z Gsr Ys d (Ar ) + rw aAz = 0 (8)
rw = ky g (Ys Ys ) (9)
where rw , ky and Y* s are water evaporation rate per unit area of
solid particles, mass transfer coefcient of water vapor in gas
phase and saturation concentration of water vapor in the solid
particles temperature. After combining relationships (3), (8)
346 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353

Fig. 3 Selected elements for spout (a), annulus (c), and Selected shape for fountain (b) regions.

and (9), the differential equation of water vapor concentration
variations in spout region is obtained:
dYs ky g a (Ys Ys )
= (10)
dz Gs
Mass balance equation for water on the surface of particles:
Considering one of the solid particles in the spout region,
the differential mass conservation equation of the water
present on surface of coarse particle is as follows:
gas phase, humidity in the gas phase at each point of reactor
should be calculated. As the relative humidity is a function
of gas temperature, and evaporation water is consistently
accompanied by the high rate of heat transfer (enthalpy of
vaporization) it is necessary to develop the energy balance
equation for gas and coarse particles separately. Energy bal-
ance equation for gas phase is as follow:
Gs Hmix A|z Gs Hmix A|z+x Gsr Hmix d(Ar ) qpg aAz + rw Hv aAz = 0

(17)
dm
= 4rp2 rw (11)
dt

where m and rp are the mass of water on the surface of solid mix = Cpair (Tg T0 ) + Ys 0 + Ys Cpv (Tg T0 ) = Cpg (Tg T0 ) + 0 Ys
particle and the radius of one of solid particles including the (18)
thickness of water on it, respectively. So m could be obtained
as follow:

4  3 
m=  rp rs3 w (12) qpg = hpg (Tg Tp ) (19)
3

Also, dt could be replaced by v = 0 + Cpv (Tg T0 ) 0 (20)

dz where T0 , Tg and Tp are the reference temperature, gas tem-

dt = (13)
Vs perature and temperature of solid particles, respectively. 0 is
enthalpy of water vaporization,mix and v are enthalpy of
In the above equations, rs , w and Vs are the radius of the
gas mixture and enthalpy of water vapor, respectively. qpg and
solid particles, water density and the average axial velocity of
hpg are convective heat ux and convective heat transfer coef-
the solid particles, respectively. Finally, the differential equa-
cient between gas and solid particles, respectively, and Cp is
tion of the existing water on particles surface in the spout
heat capacity. After substituting Eqs. (18)(20) in (17), differen-
region is obtained by combining Eqs. (9) and (11)(13).
tial equation of gas temperature variations in the spout region
drps ky g could be obtained as Eq. (21):
= (Y Ys ) (14)
dz Vs w s
dTgs hpg a
= (Tgs Tps ) (21)
Mass balance equation for coarse particles with the water on dz Gs Cpg
their surfaces:
If S and Sr be the axial and lateral mass ux of solid parti- 3.1.6. Temperature Gradient in solid phase
cles, we have: The energy balance equation for solid particles including the
water on their surfaces is as follows:
Ss A|z Ss A|z+z Ssr d (Ar ) rw aAz = 0 (15)
Ss p A|z Ss p A|z+z + Ssr pr dAr + qpg aAz+
(22)
Then the mass balance equation for coarse particles with rc hc aAz + rcs hcs aAz rw v aAz = 0
the water on the their surfaces is

dSs Ssr p = Cpms (Tps T0 ) (23)

= rw a (16)
dz D/4
pr = Cpms (Tpa T0 ) (24)
3.1.5. Temperature gradient in gas phase
According to the description mentioned above and the In these equations, Tpa is inlet solid particles temperature
dependence of the reaction rate on the relative humidity of from the annulus region,p and pr are enthalpy of the solid
 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353 347

particles with water layer and enthalpy of inlet solid parti- a) CO2 reaction rate in absence of SO2 :
cles from the annulus with water layer, respectively. hc and
hcs are the enthalpy of reactions between CO2 NaOH and rc = 1.0118yCO2 0.3996 T 0.9437 (26)
SO2 NaOH, respectively. Cpm is the effective heat capacity of
particles and water layer. By substituting Eqs. (23) and (24) into b) CO2 reaction rate in presence of SO2 :
(22), the nal differential equation of temperature variations
of solid particles in spout region could be obtained as follows: rc = 7.7733yCO2 1.5202 ySO2 0.2731 T 0.1666 (27)

S Cp
r ms
dTps (Tps Tpa ) + hpg a (Tgs Tps ) + rc ahc + rcs ahcs
1
= D/4 (25)
dz Ss Cpms
ky g a (Ys Ys ) 0 + ky g a (Ys Ys ) Cpms (Tps T0 )

The governing equations of the annulus and fountain

which are shown in Table 1 could be derived in the same c) SO2 reaction rate:
manner.
rcs = 0.0787yCO2 0.3087 ySO2 0.4727 T 0.5385 (28)

3.2. Model parameters

where rc and rcs are the ux of CO2 consumption and SO2 con-
sumption in (kg/(m2 s)), respectively. yCO2 and ySO2 are CO2 and
Table 2 shows the models main parameters. Required param-
SO2 concentrations in ppm, respectively, and T is approach
eters in the model are calculated by the equations proposed
to saturation temperature of the bed in K.
in the literature. The equations for calculating mass and heat
transfer coefcients in the different regions of reactor are pre-
3.5. Numerical solution
sented in Table 3.

All the equations summarized in Table 1 are rst order

3.3. Boundary conditions
ordinary differential equations (ODEs). The important point
regarding these coupled equations is the dependency of them
Due to the special conditions of the reactor and the unique gas
and the way that these are interrelated. Another complexity
ow and circulation of solid particles in the SBR, the system
for solving these coupled equations is countercurrent direc-
has a complex boundary condition which is discussed in the
tions of gas ow and particles in annulus region. To solve
following sections.
governing ODEs in numerical method avoiding divergence of
responses and considering the complexity of the system, the
3.3.1. Spout region
trial and error method as explained in appendix is used.
In spout region, CO2 , SO2 and water vapor concentrations in
gas and gas temperature are known at the entrance of the
4. Results and discussion
reactor. However, the temperature and thickness of water layer
on solid particles are not obvious in all the boundaries in this
4.1. Hydrodynamic evaluation
region initially. In addition, due to the radial movement of solid
particles from annulus region into spout region, temperature
Using stream tubes model in the mathematical modeling of
prole and water layer thickness prole on coarse particles
SBR signicantly reduces the volume of mathematical oper-
along the border of annulus and spout regions are required.
ations and calculations. Therefore, for the system studied
Initial values of temperature and water layer thickness proles
which includes gasliquidsolid phases and mass and heat
are the outlet data derived from annulus region equations.
transfer phenomena with complicated calculation and equa-
tions, it is necessary to use a model that could obviate
3.3.2. Fountain region momentum equations (Haghnegahdar et al., 2011).
Fountain is extension of spout and annulus regions. Then,
In order to assess the accuracy of the model, it is necessary
the conditions of top of the spout and annulus regions (out-
to evaluate the prediction of the hydrodynamic model. Fig. 4
let data obtained by solving governing equations of spout and
shows the locus of stream tubes calculated and assumed to
annulus) show the boundary conditions of fountain region.
form the annulus region is shown. Because of coned bottom
As mentioned above, because of the mixed reactor behavior,
shaped of SBR, the stream tubes at the upper parts of reac-
there isnt any local distribution in this region.
tor have been obtained with a uniform slope but at the lower
parts, a sudden change is seen in the slope of the shape. If the
3.3.3. Annulus region bed studied becomes uniformly conical, the locus of stream
In the annulus region, the conditions of inlet gas ow to each tubes obtained is not affected by this sudden change. The
stream tube are the conditions of outlet gas ow of the spout accuracy of prediction the hydrodynamic model is compara-
region. The solid particles entering the annulus area are pro- ble with other stream tubes modeling of SBRs (Haghnegahdar
vided by the return and fall of the spout coarse particles. et al., 2011; Lim and Mathur, 1976; Niksiar et al., 2013).
Therefore, the outlet conditions of fountain particles are con-
sidered as the inlet conditions for annulus region. 4.2. Model verication

3.4. Rate of reactions At this point, to verify the model predictions, the experimental
results obtained by the authors are used. Experimental condi-
The rate of reaction between CO2 /SO2 and sodium hydroxide, tions and the column specications are presented in Table 4.
as proposed by Fakhari (2014) could be obtained as follows: First, a comparison is made between the experimental results
348 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353

Table 1 Summarized governing equations for three regions of the SBR.

Region name Conservation equation Equation

Spout Mass CO2 (dXs /dz) = (rc a/Gs )

balance SO2 (dZs /dz) = (rcs a/Gs )
Water vapor (dYs /dz) = + (ky g a (Ys Ys ) /Gs )
Energy Gas phase (dTgs /dz) = (hpg a/Gs Cpg ) (Tgs Tps )
balance (dTps /dz) =
Solid phase

(Sr Cpms / (D/4)) (Tps Tpa ) + hpg a (Tgs Tps ) + rc ahc + rcs ahcs
(1/Ss Cpms )
ky g a (Ys Ys ) 0 + ky g a (Ys Ys ) Cpms (Tps T0 )
Fountain Mass CO2 Gf Xf Af,in Gf Xf Af,out (1/3) rc af Af Hf = 0
balance SO2 Gf Zf Af,in Gf Zf Af,out (1/3) rcs af Af Hf = 0
Water vapor Gf Yf Af,in Gf Yf Af,out + (1/3) rw af Af Hf = 0
   
Energy Gas phase Gf Hm Af,in Gf Hm Af,out qpg aAf Hf /3 + rw Hv aAf Hf /3 = 0
balance
   
Sf Hp Af,in Sf Hp Af,out + qpg aAf Hf /3 + rc hc aAf Hf /3
Solid phase    
+ rcs hcs aAf Hf /3 rw Hv aAf Hf /3 = 0
Annulus Mass CO2 (dXak /dz) = (rc a/Gak )
balance SO2 (dZak /dz) = (rcs a/Gak )

   
Water vapor (dYak /dz) = + ky a Yak Yak g /Gak
   
Energy Gas phase dTakg /dz = (hpg a/Gak Cpg ) Takg Tfp
  
balance   hpg a Takg Takp + rc ahc z
Solid phase dTakp /dz = (1/Sak Cpms )

 
+rcs ahcs z ky g a Yak Yak 0

Table 2 Expressions for calculation of uid dynamics parameters of the bed.

Parameter name Correlation Reference
   1/4
Spout diameter Ds = Dc 2.028g U/ (p g ) 1 mf gDc Hook et al. (1992)
Spout porosity s (z) = 1 (z/H) Smith et al. (1982)

  
     2
Minimum Umf = 42.4 1 mf g /g dp 1+3.111 104 3mf Ar/ 1 mf 1 Smith et al. (1982)
uidization
velocity
   2
2/3
Maximum spoutable Hm = D2c /dp (Dc /din ) (700/Ar) 1+35.9 106 Ar 1 Wu et al. (1987)
bed height  3

Gas velocity in Ua = Umf 1 (1 (z/Hm )) Mamuro and Hattori (1968)
annulus

  
     2
Ums = 42.857 1 mf g /g dp 1 + 3.111 104 3mf ArC / 1 mf 1
 
C = 1 E Hc B2 /Dc [(2 (E + B 1.1) 3.24/ /2 (E + B 1.1) + 3.24/ )]
Minimum spouting E2 + (2B 2.2 + 3.24/ ) E + [(B 2) (B 0.2) 3.24B/ ] = 0 Hook et al. (1992)
velocity = 7.18 (A din /Dc ) + 1.07
   
A = g Umf UT / (p g ) gdin
B = (1/ (1 + Hc /Dc ))
0.379 0.892 3.49 2.75  2.13
Hf = 46.4din ((Uin /Ums ) 1) (Hc /Hm ) (p g /g ) (dp /din ) (din /Dc ) A
Fountain height Day (1990)
A = g Umf UT / (p
 g ) gdin   
3
Annulus total mass Wagt = g Ums (/4) D2C D2S 1 (1 (Z/Hm )) Mamuro and Hattori (1968)
ow rate

with mathematical model to predict and validate the effect
of operational parameters on the removal efciency of carbon
dioxide as the main response of the study. Then the models
prediction for effect of some reactor design parameters on CO2
removal efciency investigated. Based on a good agreement
between model predictions and experimental results, it was
found that the proposed mathematical model correctly pre-
dicts the behavior of the system. For process simulation the
user should specify the value of the mentioned parameters in
Table 5. The program could calculate the removal efciency of

Table 3 Correlations used for calculating the heat and mass transfer coefcients in three different regions of SBR:.
Region Phenomena Coefcient Reference

Spout Mass transfer Sh = 1.81Re0.5

p Sc
0.33
Wu et al. (1987)
1.46
 1.3
Heat transfer Nu = 0.0005(dp Ums g /g ) Uf /Ums Wu et al. (1987)
Annulus Mass transfer Sh = 1.45Re0.5
p Sc
0.33
Cecen (1994)
Heat transfer Nu = 2 + 0.39Re0.5
p Pr
0.33
Cecen (1994)
Fountain Mass transfer Sh = 1.81Re0.5
p Sc
0.33
Wu et al. (1987)
1.46
 1.3
Heat transfer Nu = 0.0005(dp Ums g /g ) Uf /Ums Wu et al. (1987)
 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353 349

Fig. 4 The locus of stream tubes calculated by the model.

Table 4 Experimental conditions and the column

specications, Fakhari (2014).
Column internal diameter 94 mm

Column height 850 mm

Column cone angle 45
Average diameter of coarse particles 650 m
Density of coarse particles 2400 kg/m3
Inlet diameter of the cone 7 mm

Table 5 Required factors for specication by the user

for simulation.
Factor Unit Range
Fig. 5 Effect of inlet SO2 concentration on the CO2 removal

Inlet Gas temperature C 200250 efciency.
Inlet CO2 concentration ppm 9001400
Inlet SO2 concentration ppm 0200
NaOH with a stronger alkaline property has more afnity to do
Na/C molar ratio 2.2
Gas ow rate m3 /h 13.5519.35 a neutralization reaction with sulfuric acid resulted from SO2 .
Accordingly, by increasing the inlet SO2 concentration, sorbent
further reacts with SO2 and less sorbent remains for reac-
CO2 /SO2 for each case. In this study, carbon dioxide removal
ting with CO2 . So, the CO2 removal efciency is reduced. Also,
efciency (E) is dened as follows:
the negative power of SO2 concentration in the kinetic equa-
Xin Xout tion (which was previously presented) shows that increasing
E (%) = 100 (29) in SO2 concentrations naturally leads to CO2 lower removal
Xin
efciency. As Fig. 5 shows, the model results are in good agree-
where X is the CO2 concentration in ppm. ment with experimental results and the minimum, maximum
and average values of relative error are 5, 14 and 8%, respec-
4.2.1. Effect of inlet SO2 concentration tively.
Fig. 5 shows the effect of inlet SO2 concentration into the
reactor on CO2 removal efciency. The model results are com- 4.2.2. Effect of inlet CO2 concentration
pared with the corresponding experimental data. As it could Fig. 6 shows the experimental data and mathematical model
be seen, CO2 removal efciency reduces by increasing the SO2 predictions for the effect of the concentration of carbon diox-
concentration over the whole range of the experiments. The ide in the inlet gas on the removal efciency of the reactor.
maximum CO2 removal efciency is at zero SO2 concentration As can be seen, increasing the inlet CO2 concentration will
which only carbonation reaction takes place. When the reac- increase the reactors removal efciency. The experiment at
tors inlet gas contains SO2 along with CO2 , both gases react to zero concentration of CO2 corresponds to single reaction
the sorbent and consequently a competitive absorption take between SO2 and NaOH resulting in the production sodium
place. When SO2 gas dissolves in water on coarse particles, sulte powder. Due to the competitive reaction of CO2 and
it produces a stronger acid (sulfuric acid), compared to CO2 SO2 with absorbent solution, by an increase in CO2 concentra-
which produces carbonic acid. Therefore, in reaction time, tions, the chemical reaction rate of CO2 increases, and thus the
350 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353
Fig. 6 Effect of inlet CO2 concentration on the CO2 removal
efciency.
Fig. 7 Effect of inlet gas temperature on the CO2 removal
efciency.
progress of the reaction and the yield of reactor in CO2 removal
increase. Also the positive power of CO2 concentration in
the kinetic equation shows that increasing CO2 concentration
naturally leads to increasing the mass ux of this gas, and ulti-
mately increases removal efciency. As Fig. 6 shows, the model
results are in good agreement with experimental results and
the minimum, maximum and average values of relative error
are 0.5, 4 and 3%, respectively.
4.2.3. Effect of inlet gas temperature
Fig. 7 shows the comparison between the experimental results
and mathematical model ones on effect of inlet gas tempera-
ture on the reactor removal efciency. As it could be seen, the
removal efciency of carbon dioxide is reduced by an increase
in inlet gas temperature. Different inlet gas temperatures will
affect the reaction equilibrium, reaction rate, and the trans-
formation and diffusion between CO2 and absorbent (Chen
et al., 2005). But the most important effect of inlet gas tem-
perature is on the evaporation rate of water in the reactor.
As previously mentioned, rst, CO2 and SO2 are dissolved in
the water covered on coarse particles and ionized in liquid
phase and then react with alkaline absorbent. Increasing the
inlet gas temperature causes more evaporation of the liquid
phase within the reactor and reduces the possibility of the
reaction between carbon dioxide and NaOH. In other words, it
could be said that by an increase in the inlet gas temperature,
the sorbent is dried sooner and a dry layer is formed on the
coarse particle surface and accordingly the removal efciency
of CO2 is reduced. As Fig. 7 shows, the model results are in
good agreement with experimental results and the minimum,
maximum and average values of relative error are 2, 4 and 3%,
respectively.
Fig. 8 Effect of gas ow rate on CO2 removal efciency.
Fig. 9 Variation of CO2 concentration in annulus region.
4.2.4. Effect of supercial gas velocity
In a SBR, supercial gas velocity is an important process
parameter. In Fig. 8, the effect of U/Ums ratio (supercial gas
velocity to minimum spouting velocity) on the reactor removal
efciency is shown. Decreasing U/Ums ratio (in a constant bed
height) prolongs the residence time of ue gas in the reac-
tor, and hence increases the contact time or reaction time
between CO2 gas and sorbent in the reactor (Ma et al., 2000).
However, by a reduction in gas velocity, the system capacity is
reduced. So, larger equipment (reactor) is needed that causes
an increase in the xed capital costs. It should be noted that
for spouted beds, below minimum spouting velocity, the bed
could not be operated normally (Wu et al., 1987). Therefore, the
ratio of supercial gas velocity to minimum spouting velocity
in the SBR could not be lower than 1. As Fig. 8 shows, the model
results are in good agreement with experimental results and
the minimum, maximum and average values of relative error
are 0.5, 8 and 4%, respectively.
4.3. Parametric study
With regard to an acceptable agreement between the exper-
imental results and mathematical model predictions, the
sensitivity analysis of model is considered. In this regard, rst,
the variations of some parameters in SBR are studied. Then,
the effect of different designing parameters on the efciency
of the reactor is investigated.
4.3.1. Variations of CO2 /SO2 concentrations in SBR
Fig. 9 shows the variations of CO2 concentration in three dif-
ferent stream tubes in the annulus region. It should be noted
that the numbering of the stream tubes starts from the glass
walls of the reactor and ends with the spout region. In other
words, the rst stream tube locates in the vicinity of wall of
 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353 351

Fig. 10 Variations of CO2 /SO2 concentrations during the length of spout region.

SBR and the tenth stream tube is located in the vicinity of the
spout region.
As it could be seen, because of the greater residence time
of gas through the rst stream tube, the variations of CO2
concentration in the tube are higher. Approaching the cen-
ter of the bed (the next numbers are referred to the tubes),
the tubes become shorter and the residence time of the gas
in the tubes reduces. Therefore, the outlet CO2 concentra-
tion from such tubes is higher than that from the longer
tubes. Obviously, CO2 variation curve in a tube in terms of
height starts from a height in which the tube begins from
the annulus region. It is clear that because of the reduction
in concentration and consequently reduction in absorption
rate, the slop of concentration variation decreases. It is worth Fig. 12 Effect of static bed height on CO2 removal
noting that the variations in CO2 concentration are low in efciency.
spout region. So, in Fig. 9, the inlet CO2 concentration into
4.3.2. Effect of designing parameters on SBR efciency
the tubes is almost equal. For a better analysis, the varia-
Fig. 11 shows the models prediction for the variations of the
tions of CO2 and SO2 concentrations during the length of spout
CO2 removal efciency in the reactor with the bed diameter.
region are shown in Fig. 10. The results shown in Figs. 9 and 10
As it could be seen, the variations trend is almost linear. It is
are in good agreement with Moeini and Hatamipour (2008)
clear that, by increasing the bed diameter, gas residence time
results.
and hence the contact time of the gas and liquid phase rise.
By comparing Figs. 9 and 10, it was found that due to the
Increasing the two-phase contact time increases the chem-
low residence time of gas in the spout region, in comparison to
ical reaction progress and thus the SBR removal efciency
the annulus region, concentration variations in spout region is
increases.
much less than the annulus region. Note that the difference in
In Fig. 12, the models prediction for the CO2 removal ef-
residence time depends on both tube volume and gas ow rate.
ciency variations in the reactor with bed height is shown. The
For example, numerical calculations show that the volume of
bed static height is increased by increasing the amount of solid
the rst tube in the annulus is about 7 times greater than the
particles within the bed. By addition of the particles, the bed
volume of the spout but the average gas ow rate in a spout
weight increases, and consequently the gas ow needs more
region is 8 times greater than the gas ow rate of the rst tube
momentum to spout the particles. Increasing the inlet gas
in annulus region.
momentum is possible by increasing the inlet gas ow rate.
In brief, increasing the static height increases the inlet gas
ow rate. On the other hand, by increasing the bed height,
gas residence time, and hence gas and liquid contact time
increases. Increasing the contact time of the two-phase helps
more chemical reaction progress and thus increases the CO2
removal efciency in SBR. Fig. 12 has good agreement with
other researcher results (Haghnegahdar et al., 2011; Moeini
and Hatamipour, 2008; Rahimi et al., 2010; Sanchez et al., 2000).

5. Conclusion

Simultaneous removal of CO2 and SO2 using hydroxide

sodium solution in a semi-dry spouted bed reactor was studied
by using combined stream tubes concept, the laws of con-
servation of mass and energy and non-isothermal reactive
Fig. 11 Effect of column diameter on CO2 removal absorption model. Ability of the model to predict the hydro-
efciency. dynamic behavior of the system besides the removal efciency
352 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353

of CO2 is examined by comparison of the results with the

experimental data. Also the effect of some design parameters
is studied. The major ndings from the present study may be
summarized as follows:

It was found that inlet SO2 and CO2 concentrations and

U/Ums ratio have considerable effects on CO2 removal
efciency, while inlet gas temperature do not inuence per-
formance of the reactor signicantly.
In Addition, CO2 removal efciency increases by increasing
the inlet CO2 concentration and by decreasing the inlet SO2
concentration, U/Ums ratio and inlet gas temperature.
Considering maximum, minimum and average values of
errors of the model results with existing experimental ones
reveals good agreement between the experimental and
model predictions.
Besides, SBR designing parameters have signicant effect
on system performance which by increasing bed static
height and bed diameter, CO2 absorption efciency
increases.
The developed model could be used for the design of indus-
trial SBRs.

Appendix A.

For calculating the gas ow path in the annulus, the vertical

height of the annulus could be divided into M equal intervals
and the width of the annulus at the top into N equal intervals
(see Fig. 13). Let each interval at the annulus top represent
one path of gas ow and let Q(J 1) be the volumetric ow
rate between streamlines J 1 and J (Lim and Mathur, 1976).
In our mathematical model N and M are equal to 10 and 400,

Fig. 14 Solution procedure for the mathematical model.

respectively. For instance, energy balance calculation proce-

dure in the SBR is explained in Fig. 14 briey.

References

Cecen, A., 1994. The maximum spoutable bed height of ne

particles spouted with air. Can. J. Chem. Eng. 27, 792797.
Chen, J.C., Fang, G.C., Tang, J.T., Liu, L.P., 2005. Removal of carbon
dioxide by a spray dryer. Chemosphere 59, 99105.
Day, J.Y., 1990. The fountain height and particle circulation rate in
a Spouted Bed. Chem. Eng. Sci. 45, 29872990.
Fakhari, M.A., 2014. Experimental Study and Mathematical
Modeling of Simultaneous Removal of CO2 and SO2 in a
Fig. 13 Coordinate system and grid points for streamline Spouted Bed Reactor. Chemical Engineering Department,
calculations (Lim and Mathur, 1976). University of Isfahan, Isfahan (MSc. Thesis).
 Process Safety and Environmental Protection 9 8 ( 2 0 1 5 ) 342353
Ghalavand, Y., Hatamipour, M.S., Rahimi, A., 2010. Experimental
and parametric study on drying of green peas in spouted bed.
Int. J. Food Sci. Technol. 45, 25462552.
Ghalavand, Y., Rahimi, A., Hatamipour, M.S., 2012. Experimental
and mathematical modeling of green pea drying in a spouted
bed. Dry. Technol. 30, 128137.
Haghnegahdar, M.R., Hatamipour, M.S., Rahimi, A., 2010. Removal
of carbon dioxide in an experimental powder particle spouted
bed reactor. Sep. Purif. Technol. 72, 288293.
Haghnegahdar, M.R., Hatamipour, M.S., Rahimi, A., 2011.
Mathematical modeling of CO2 capture in a semi-dry spouted
bed reactor. Sep. Purif. Technol. 80, 509518.
Hook, B.D., Littman, H., Morgan, M.H., 1992. A priori modeling of
an adiabatic spouted bed catalytic reactor. Can. J. Chem. Eng.
52, 966968.
Jeng, R.S., Altwicker, E.R., Morgan, M.M., Wilk, R.D., 2001. Propane
combustion in a spouted bed combustor. II: a modeling study
in the bed region. Combust. Sci. Technol. 170 (1), 119149.
Lim, C.J., Mathur, K.B., 1976. A ow model for gas movement in
spouted beds. AIChE J. 22, 674680.
Lin, C.C., Chen, B.C., Chen, Y.S., Hsu, S.K., 2008. Feasibility of a
cross-ow rotating packed bed in removing carbon dioxide
from gaseous streams. Sep. Purif. Technol. 62, 507512.
Liu, Y., Kato, K., 2000. Process analysis of semidry
desulphurization process in powder particle spouted bed. J.
Chem. Eng. Jpn. 33 (2), 223.
Ma, X., Kaneko, T., Tashimo, T., Yoshida, T., Kato, K., 2000. Use of
limestone for SO2 removal from ue gas in the semidry FGD
process with a powder-particle spouted bed. Chem. Eng. Sci.
55, 46434652.
Ma, X., Kaneko, T., Xu, G., Kato, K., 2001. Inuence of gas
components on removal of SO2 from ue gas in the semidry
FGD process with a powder particle spouted bed. Fuel 80,
673680.
Mamuro, T., Hattori, H., 1968. Flow pattern of uid in spouted
beds. J. Chem. Eng. Jpn. 1, 15.
Mathur, K.B., Epstein, N., 1974. Spouted Beds. Academic Press Inc,
New York, NY.
353
Mathur, K.B., Lim, C.J., 1974. Vapor phase chemical reaction in
spouted beds: a theoretical model. Chem. Eng. Sci. 29, 789797.
Moeini, M., Hatamipour, M.S., 2008. Flue gas desulfurization by a
powder-particle spouted bed. Chem. Eng. Technol. 31, 7182.
Nakazato, T., Liu, Y., Kato, K., 2004. Removal of SO2 in semi-dry
ue gas desulfurization process with a powder particle
spouted bed. Can. J. Chem. Eng. 82, 110115.
Nieto, O., Nino, M., Martinez, R., Tailleur, R.G., 2007. Simulation of
a spouted bed reactor for solid catalyst alkylation. Fuel 86,
13131324.
Niksiar, A., Sohrabi, M., Rahimi, A., 2013. Comparative evaluation
of existing correlation to predict spouted bed hydrodynamics.
Dry. Technol. 31, 975989.
Olazar, M., Aguado, R., Velez, D., Arabiourrutia, M., Billbao, J.,
2005. Kinetics of scrap tire pyrolysis in a conical spouted bad
reactor. Ind. Eng. Chem. Res. 44, 39183924.
Rahimi, A., Hatamipour, M.S., Gholami, M., Haghnegahdar, M.R.,
2010. Non-isothermal modeling of the ue gas
desulphurization process using a semi-dry spouted bed
reactor. Chem. Eng. Res. Des. 89, 777784.
Sanchez, I., Mazza, G., Flamant, G., Gauthier, D., 2000. A
streamtube non-isothermal spouted-bed reactor
mathematical model. Chem. Eng. Sci. 55, 193202.
Silva, D.O., Tamiozzo, L.M., Duarte, C.R., Murata, V.V., Barrozo,
M.A.S., 2011. Modeling of seed coating in a spouted bed. Dry.
Technol. 29, 286294.
Smith, K., Arkun, Y., Littman, H., 1982. Studies on modeling and
control of spouted bed reactor-I: Reactor modeling. Chem.
Eng. Sci. 37, 567579.
Tao, M., Jin, B., Zhong, W., Yang, Y., Xiao, R., 2010. Numerical and
experimental study on ue gas desulfurization in the
underfeed circulating spouted bed. Chem. Eng. J. 159, 149158.
Wu, S.W.M., Lim, C.J., Epstein, N., 1987. Hydrodynamics of
spouted beds at elevated temperatures. Chem. Eng. Commun.
62, 251268.
Xu, G., Guo, Q., Kaneko, T., Kato, K., 2000. A new semidry
desulphurization process using a powder particle spouted
bed. Adv. Environ. Res. 4, 918.