Kinetic Study of Sulfur Dioxide Elimination by Limestone
through the Lab Scale Circulating Fluidized Bed Combustor
Dowon Shun*, Dal-Hee Bae *, In-Kyu Jang**, Keon-Hee Park** and Seung Kyu Park**,†
*Greenhouse Gas Research Center,
Korea Institute of Energy Research, 71-2 Daejon 305-343, Korea
**Department of Chemical Engineering,
Hoseo University, Asan 336-795, Korea
Email :
Abstract—Characteristics of sulfur dioxide emission from
coal and petroleum coke combustion were examined in a lab
scale circulating fluidized bed (CFB) combustor. The rate
constant of the first order rate expression for the absorption
SO2 on the CaO surface was similar regardless of the origin of
the limestone, the particle size and the initial SO2
concentration. However, the total SO2 absorption capacity was
different depending on the origin of the limestone. The
breakability of the particle which provides new surface for the
reaction seems to play a major role in the absorption
characteristics.
Keywords-Sulfur Dioxide (SOx); Limestone; Circulating
Fluidized Bed Combustor; Emission; Kinetics; First order
reaction
1. Introduction
Sulfur oxides (SOx) gases form when the coal and heavy
oil are burned. The SO2 readily dissolves in water vapor and
leads to the formation of acid and interacts with other gases
and particles in the air to form sulfates and other products that
can be harmful to people and environment [1,2]. SO2 pollution
is thought to promote wheezing, bronchial constriction,
shortness of breath, and exacerbation of asthma [2]. Nitrogen
oxides (NOx) are highly reactive gases that contain nitrogen
and oxygen in varying molecular combinations. The major
source of NOx is the combustion of fossil fuels such as coke in
electric power plants or petroleum in vehicle engines [3].
Many of the nitrogen oxides are colorless, but nitrogen
dioxide (NO2) combined with particles in the air can cause a
reddish-brown haze. The presence of NOx leads to a variety of
environmental problems such as ground level ozone, acid rain
and the deforestation by acid rain. They are poisonous for the
respiratory system, provoking both lung infection and
respiratory allergies. Both sulfur dioxide and nitrogen oxides
contribute to acid rain [4]. Just a few decades ago, SOx and
NOx pollution had gotten so bad that acid rain had damaged
countless buildings, monuments, car finishes [5]. One member
of the NOx family, nitrous oxide (N2O), is a potent greenhouse
gas [6]. Air pollution caused by SOx and NOx, which are
largely the result of industrial processes, may also produce
environmental impacts. For these reasons, SOx and NOx
emissions should be eliminated.
In the field of fluidized bed combustion with high thermal
capacities, circulating fluidized bed (CFB) reactors are
considered to be the most efficient commercial utility which
burns various solid fuels, including coals, with a minimum
operation and maintenance cost [7-9]. The reactors are usually
lack of sulfur capture facilities, such as flue gas treatment,
since the in-situ SO2 capture by the limestone injection into the
combustor itself is sufficient to comply with the regulation.
Although the best operation parameters for the boilers could be
collected from the actual experience of the commercial scale
boilers [10], the information is not always intuitive since the
parameters of the commercial boilers in operation are normally
interrelated. It is quite difficult to understand the actual effects
of specific variables on the performance of the commercial
boiler in operation. On the other hand, many experimental
results from previous researches were carried out with packed
bed units in small scale and turned out to be impractical to be
applied to the commercial CFB boilers [3]. To investigate the
emission characteristics of CFB systematically, a CFB
combustor was built in-house grade. Many researchers have
been performed in conventional air-combustion fluidized bed
boilers to eliminate the SOx by limestone flux, and much
information has been accumulated through the kinetic and
spectroscopic studies, many basic questions still remain to be
answered clearly. The emission of SOx and NOx gases during
the combustion of coal or petroleum pitch in the CFB has been
studied. The aim of this study is to elucidate the absorption of
SO2 by three kinds of limestone produced from Tanyang,
Samchuk and Jinsan in Korea. Emission of SOx and NOx in the
CFB and reaction of them with limestone have been
investigated.
2. Experimental
A. Circulated fluidized bed
The circulated fluidized bed (CFB) combustor was
prepared with quarts tubes [8-10]. The CFB combustor consists
of a riser (combustion chamber) and a cyclone, a loop seal and
an ash classifier. The silica sand in the riser was heated to more
than 900 ºC and was blown upward by the combustion air from
Advances in Materials Physics and Chemistry
Supplement: 2012 world Con
g
ress on En
g
ineerin
g
and Technolo
gy
Cop
y
ri
g
ht © 2012 SciRes.189

the bottom of the riser to the top. Since the heat generated from
the combustion of coals was not sufficient to sustain the
desired combustor temperature, multi sets of electric heaters
were installed around the outer side of the test unit.
B. Sulfur dioxide capture by limestone and
analysis
Test coal samples were screened with the particle size
between 0.1-0.7 mm. In the combustion experiments were
tested five different coal samples;
1) Blair Athol coal: a bituminous coal from Australia
2) Herbei coal : a high sulfur bituminous coal widely used in
the power industries of China
3) Shenhwa coal : a low sulfur, high heating value coal from
China
4) Tokye coal : a Korean anthracite
5) Petroleum coke : a byproduct from a Korean refinery
The combustor installed with bed materials was heated with air
only to the pre specified temperature by electric heaters. As the
temperature of the combustor reached the pre-set value, the
coal was fed into the bottom of the riser. While the coal was
burned inside of the riser, the ashes were entrained to the top of
the combustor and led to the cyclone. The particles except very
fine ashes were collected by the cyclone and recycled to the
riser through the return leg. The particles were circulated
between the riser and the return leg until the combustion is
completed and the particle size is reduced enough to flow out
of the cyclone to the bag filter. The effluent from the cyclone to
the bag filter was analyzed by the URAS Model 14-1, 14-2 of
ABB Hartman and Braun to measure the levels of SO2, NOx,
N2O, CO, CO2, and O2. The limestone was feeding by 50g to
CFB reactor. The limestone was calcined in a muffle furnace at
900 ºC for 2 hours.
3.Results and Discussion
The design specifications of the CFB are listed in Table 1.
In the combustion test of coal samples, only the air/fuel ratio
was varied while the operation temperature is fixed. However
both the operation temperature and the air/fuel ratio were fixed
for the investigation of the sulfur capture kinetics.
Furthermore the variation of the reactor temperature and the
air/fuel ration were restricted to be varied within the precision
limit of the facility while the coal feed rate was varied
between 0.1 and 0.2 kg/min.
Table 1. Design specification of the quartz CFB.
Parameter Minimum Maximum
Coal, kg/h 0.16 0.24
Air flow LPM @ stp 22 28
O2 in flue gas, % 2 7
Velocity of gas, m/s 5 6
Average bed temp, K 1123 1173
Table 2 presents the compositions of each Korean
limestone utilized in the experiment. All three samples are
typical commercial products and currently used for the
commercial CFB boilers in Korea. The CaO contents were
between 51-55%, and the purities were similar. Limestones
were activated after the calcinations in the CFB combustor.
Figure 1 shows the SEM images of the Jinsan limestone
before and after the calcination. The limestone was calcined in
a muffle furnace at 900 ºC for 2 hours. After the calcination,
CO2 on the CaCO3 was detached and the CaO surface was
exposed with the generation of many small pores. The
wrinkled CaO surface was known to provide SO2 absorption
sites on which reactions occur [3, 4].
Table 2. The main composition of three limestones.
Limestone
Component Tanyang Samchuk Jinsan
CaO 53.13 54.25 51.6
MgO 0.94 0.36 0.39
CO2 42.72 42.97 40.92
(a)
(b)
Figure 1. SEM image of T limestone (a) before and (b) after
calcination.
Table 3 presents the BET surface analysis by a Quantachrome
Autosorb 1 with the absorption of nitrogen gas at its boiling
temperature.
Table 3. BET surface analysis of limestone.
Sam- CaCO3 CaO after calcination
190 Cop
y
ri
g
ht © 2012 SciRes.

ple Area
(m2/g)
Vol.
(cm3/
g)
Pore
dia.
(ȝm)
Area
(m2/g)
Vol.
(cm3/
g)
Pore
dia.
(ȝm)
Tan-
yang 0.53 0.15 1.13 6.77 2.21 1.30
Sam-
chuk 0.28 0.09 1.30 11.24 3.66 1.30
Jin-
san 0.41 0.12 1.17 1.85 0.59 1.28
(a)
0
100
200
300
400
500
01234567
O2[%]
E
m
i
s
i
s
o
n
[
p
p
m
@
3
.
5
%
O
2
]
NOx
N2O
SO2
(b)
0
100
200
300
400
500
600
0123456
O2[%]
E
m
i
s
i
s
o
n
[
p
p
m
@
3
.
5
%
O
2
]
7
N2O
NOx
SO2
Figure 2. The effects of the aeration on the emissions of SO2,
NOx and N2O from the combustions of (a) the Blair Athol coal
and (b) the Tokye coal.
In the Blair coal combustion, about 280 ppm of SO2 emission
was slightly decreased to 200 ppm as the oxygen content
increases. Meanwhile, the N2O emission was slightly
increased from 110 ppm to 310 ppm as the oxygen feeding
was increased. In the Tokye coal combustion, the SO2
emission gradually was decreased from 500 ppm to 400 ppm
as the oxygen feeding increases. Figure 3 shows the variation
of CaO conversion rate of Tanyang limestone as the reaction
goes by. The comparison was made among different average
particle sizes. The maximum conversion rates were observed
when CaO conversion, XCaO, was 0 - 0.1. After that it decayed
rapidly.
0
0.5x10-5
1.0x10-5
1.5x10-5
2.0x10-5
2.5x10-5
00 .0 50 .10 .1 50 .20 .2 5
X
C
a
O
1.4-0.2
0.2-0.08
0.08-0
d
F
S
O
[
m
o
l
/
s
]
2
Figure 3. Variation of CaO conversion rate with respect to the
particle size of Tanyang limestone.
CaSO4 layer is formed after SO2 gas is absorbed on the CaO
surface, the surface larger of CaSO4 molecules will block
further absorption of SO2 to inner CaO site. So the ash
diffusion model is excluded. The reaction must be either
controlled by the gas film diffusion or chemical kinetics. In
this experiment the large mass of CaCO3 is introduced to
provide infinite reaction surface. The external transport rate
will be same for all the experiment and will not affect the
reaction rate. The gas film diffusion control mechanism and
the chemical reaction control mechanism were compared and
the chemical reaction control showed better fit to the data.
A second order absorption kinetics was proposed to analyze
the experimental data.
CaOSO CkCr 2
Where,
r
; SO2 absorption rate [gmol SO2/gmol limestone-s]
; 2nd order rate constant [liter/gmol CaO-s] k
; Concentration of CaO
CaO
C
[gmol CaO/ gmol limestone]
; Concentration of SO2 at the combustor exit
[gmol SO2/liter]
2
SO
C
The kinetic equations can be restated as;
CaOSOSO FkCdF 22 
Where CaOSO dFdF
2 [gmol SO2/s]
Since the excess amount of CaO was injected in the reactor.
The concentration of CaO in the circulating fluidized bed
combustor is considered to be constant and the reactor is
considered as a mixed reactor regarding CaO. The reactor
equation is the first order only to the concentration of SO2.
2
'SO Ckr Where, ; 1st order rate constant [1/s]. 'k
When the SO2 absorption rate of limestone with different
origin is compared, the Jinsan limestone showed the highest
rate constant and that of Samchuk showed the lowest constant.
Cop
y
ri
g
ht © 2012 SciRes.191

Sulfur elimination by limestone under CFB combustion
conditions is the net effect of a competition between sulfur
capture and sulfur release during which the composition of the
Ca surface changes continuously between CaO, CaS and
CaSO4. Therefore, we conclude that the limestone feeding is
enough to capture all SO2 gases, so the reactions become
depend on the concentration of SO2. And the reaction can be
disrupted by the formation of CaSO4 after reaction goes by.
For understanding of SO2 gas adsorption onto limestone, the
chemical reaction model should be studied [5, 12]. For the
comparison of the rate controlling step, we postulate that the
particle is a globular shape at the shrinking core model during
the adsorption of gas onto the particle.
G
At the reaction : A (gas) + B (soild) -> Solid Products
4. Conclusion
The CaO contents of three kinds of limestone were between
51-55%. The BET surface area of Samchuk limestone was
increased about 40 times after calcinations. The surface area of
Samchuk CaCO3 was increased from 0.28 m2/g to 11.24 m2/g
after the calcinations to CaO at 900 oC for 2 hours at the CFB
combustor. Based on the kinetic study of SO2 gas adsorption to
CaO solid, the CaO conversion curve fits with respect to the
reaction time. Since the slope of time versus is linear
and close to 1, it follows gas film diffusion control. We suggest
that the reaction is the first order. Since the limestone feeding
is enough to capture all SO2 gases, the reaction become depend
on the concentration of SO2. And the reaction can be disrupted
by the formation of CaSO4 after reaction goes by.
In the case of kinetic control, REFERENCES
[1]F. Garcia-Labiano, A. Rufas, L. F. de Diego, M. de las Obras-
Loscertales, P. Gayan, and J. Adanez, Fuel, 90, 3100 (2011).
Therefore, [2]G. Ozkan, and G. Dogu, Chem Engineering and Processing, 41, 11
(2002)
[3]S. K. Park, V. Kurshev, Z. Luan, C. W. Lee, and L. Kevan, Micropor.
Mesopor. Mater., 38, 255 (2000).
[4]D. R. M. Wright, F. Pinto, L. Armesto, M. A. Caballero, M. P. Aznar, A.
Cabanillas, Y. Huang, C. Franco, I. Gulyurtle, and J. T. McMullan, Fuel
Processing Technol. 87, 793 (2006).
Figure 4 shows the CaO conversion curve fitting with
respect to the reaction time. It is clearly seen that the slope of
time versus is linear and close to 1, the reaction
follows gas film diffusion control.
[5]M. Higashi, S. Uchida, N. Suzuki, and K. I. Fujii, IEEE Trans. on
Plasma Sci., 20, 1 (1992)
[6]D. Shun, D. H. Bae, J. Y. Paek, and Y. S. Park, Korean J. Chem. Eng.,
21, 890 (2004).
[7]D. Shun, D. H. Bae, K. H. Han, S. H. Cho, and S. Y. Lee, Hwahak
Konghak 40, 345 (2002).
G
[8]D. R. M. Wright, F. Pinto, L. Armesto, M. A. Caballero, M. P. Aznar, A.
Cabanillas, Y. Huang, C. Franco, I. Gulyurtle, and J. T. McMullan, Fuel
Processing Technol. 87, 793 (2006).
[9]K. Redemann, E. U. Hartge, and J. Werther, Fuel, 87, 3669 (2008).
[10]D. Shun, H. S. Chang, T. S. Park, D. H. Bae, and G. T. Jin, Korean J.
Chem. Eng., 18, 630 (2001).
[11]P. Gayan, J. Adanez, L. F. de Diego, F. Garcia-Labiano, A. Cabanillas,
A. Bahillo, M. Aho, and K. Veijonen, Fuel, 83, 277 (2004).
[12] L. Jia, J. Wang, and E. J. Anthony, Chem. Eng. J., 94, 147 (2003).
Figure 4. CaO conversion curve fitting to the shrinking core model.
192 Cop
y
ri
g
ht © 2012 SciRes.