вход по аккаунту


Experimental and Mechanism Studies on a Different-Velocity Circulating Fluidized Bed for Flue Gas Desulfurization.

код для вставкиСкачать
Dev. Chem. Eng. Mineral Process.,8(3/4),pp. 199-206,2000.
Experimental and Mechanism Studies on a
Different-Velocity Circulating Fluidized Bed
for Flue Gas Desulfurization
Y.H. Wu*, B. Feng, 2. Huang and D.J. Li
Thermal Energy Engineering Research Institute, Southeast University,
Nanjing, 210096, €? R. CHINA
Experimental studies of flue gas desulfurizarion were carneed out in a differentvelociv circulating fluidized bed reactol; using dry slaked lime as the sorbent. SO,
removal efficiency could be as greater than 80% when the CdS molar ratio was 1.1,
and when a small amount of water was added into the process. By considering the
main aspects that influence desulfurization efficiency in the experiment, it was found
that the water sprayed into the reactor plays an important role. In addition, a
preliminary investigation was performed on the desulfirization mechanism. The
desulfurization reaction occurs quickly in the waterfilm on the surface of the sorbent.
Circulating fluidized bed (CFB) technology has been widely used in the
petrochemical industry for many years. However, it was not introduced into the field
of flue gas desulfurization (FGD) until the 1980s by Lurgi Company, Germany [I]. In
contrast to the wet FGD process such as the limestonelgypsum method, the CFB
technology is a dry FGD method that is suitable for small- and medium-size boilers.
This method reduces the SO2 emission with a high SO2 removal efficiency, lower
capital cost and smaller space requirements. A few slight modifications were made
on the traditional CFB reactor. The normally straight reactor is transformed into an
"n" shape so that the height is reduced to half of the original [2, 31. The ascending
pipe is where the main desulfurization occurs and the descending part has two main
functions. First, to increase the reaction time between the sorbent and SO2. Second,
to reduce the height of the particle collector. The reactor consists of six pipes with
two different diameters that were installed alternately, causing a different velocity
distribution in the reactor [4].
* Authorfor correspondence (
Y.H. Wu,B. Feng, 2.Huang and D.J. Li
Experimental Details
The experimental facilities are shown schematically in Figure 1. The temperature of
the flue gas produced from an oil combustor is adjusted down to about 15OOC by
mixing with cold air. A certain amount of fly ash, with 15g ash per m3 flue gas of
concentration, and pure SOzare added into the flue gas to simulate a real coal burning
boiler. The simulated dust-laden flue gas passes through a circulating fluidized bed
reactor, a two-stage cyclone and a bag filter, respectively, and finally it is exhausted
into the air by an induced draft fan. The dust collected by the cyclone goes back into
the reactor through a slant air duct. The dry slaked lime powder is added into the
reactor continuously by a screw feeder with the set CdS molar ratio. The water is
sprayed into the reactor through a two-fluid nozzle.
The pressure drop of the circulating fluidized bed reactor and the two-stage
cyclone are measured by three pairs of differential manometers. There are five
thermocouples placed in different parts of the system. The dust removal efficiency of
the cyclone is measured by the pre-estimating flow velocity method. The
concentration of SO1 is measured online by a multicomponent gas analyzer,
BINOS"1000, which is made in Germany. The SOz removal efficiency is defined as
the concentration difference between the inlet and outlet of the reactor divided by the
inlet SOz concentration.
Figure 1. Schematic diagram of the experimental apparatus.
The flow rate in the reactor can be regulated by changing the diameter and length
of pipes in the reactor. It consists of three segments of pipes with 106mm diameter,
and another three segments of pipes with 186mm diameter. The flow rate was
Experimental and Mechanism Studies on a CFBfor Flue Gas Desulfirization
designed in the range of 300-700m3/h. For a flow rate of 500m3/h, the residence time
of the flue gas in the reactor is about 3.2 seconds.
The slaked lime used as the sorbent is made ftom quicklime by dry slaking on site,
and the slaking temperature is controlled at about 105°C to keep it dry. Two kinds of
quicklime are chosen. One is made in the local limekiln, Yuhua Lime Plant, which is
used mainly as a building material and the purity usually varies from 53% to 80%.
According to the ASTM standard for testing the lime reactivity, putting 25g lime into
a heat-proof container with lOOg water in it, the average temperature rise is only 45°C
after 3 minutes, therefore, Yuhua lime is not very reactive. The other lime is
produced by Baoshan Iron and Steel Group, Shanghai, and with the same test
standard, the temperature nses to 67°C. Therefore, Baoshan lime has a much higher
reactivity than Yuhua lime.
Results and Analysis
I Pressure drop of the system
The quadratic relationship between the total pressure drop CAP) and the flow rate of
flue gas is shown in Figure 2. When the flow rate is at the standard operating mode,
500 Nm3/h, the total pressure drop is about 2400 Pa. However, if the electrostatic
precipitator is used, the total pressure drop of the system will reduce considerably.
100 200
300 400 500
Flow rate { m3h>
Figure 2. Relationship between total pressure drop and the flow rate of the flue gas.
Y.H.Wu, B. Feng, Z. Huang and D.J. Li
N Eflect of CdS raih
Figure 3 shows the relationship between the desulfurization efficiency of two
different sorbents for different CdS molar ratio. The experimental parameters are the
inlet temperature of 15OoC, inlet SO2 concentration of 1160 ppm, the flue-gas flow
rate of 400 m3h, and the water-spray rate of 11 kglh.
C d S molar ratio
Figure 3. Relationship between CdS molar ratio and the desul&nzation efficiency.
As the CdS molar ratio increases, the desulfurization efficiency in the system also
increases. However, when a different slaked lime is used in the experiment, there are
variations between the desulfurization efficiency and the CdS molar ratio. For
Baoshan slaked lime, when the CdS molar ratio is at a relatively low value, the
desulfurization efficiency rises quickly with increasing CdS molar ratio. However,
when the CdS molar ratio reaches higher values, the desulfurization efficiency rises
less rapidly. In contrast to Baoshan lime, it is found that the Yuhua lime shows a
slower (and smoother) increase in desulfurization efficiency with the CdS molar
ratio, due to it having more impurities that can not react with S02.
ZZZ E$ect of the water rate
The relationship between desulfurization efficiency and water flow rate (for different
SO2 inlet concentrations) is shown in Figure 4, with an inlet temperature of 150"C,
flue-gas flow rate of 420 m3h,
and CdS molar ratio of 1.2.
Experimental and Mechanism Studies on a CFBfor Flue Gas Desulfurization
1 0 1 2 1 4
Amount of water (kg/h)
1 0 1 2 1 4
Amount ofwater (
Figure 4. Relationship between desulfurization eficiency and amount of water added.
At a given Ca/S molar ratio, the desulfurization efficiency increases with
increasing amount of water sprayed into the reactor. From Figure 4a, when the water
rate is lower than 12 k g h then the desulfurization efficiency rises rapidly. At higher
rates, the desulfurization efficiency increases smoothly and slowly. This indicates
that desulfurization efficiency does not depend only upon increasing the water rate
when the CdS molar ratio is relatively low. It is obvious that at a certain CdS molar
ratios, there is a reaction limit between Ca(OHh and Sop. In the experiment, an
important phenomenon was found: namely as the water flow rate increases, the
Y.H.Wu,B. Feng, Z Huang and V.J.Li
desulfurization efficiency increases smoothly at first then it becomes quicker and
finally it slows down when the amount of water reaches a certain value. It is thought
that when the water rate is small, the water evaporates as soon as it is sprayed into the
reactor. This evaporation stage is so short that the water cannot form a film on the
surface of the sorbent. During this period, the water only increases the humidity of
the flue gas, so there is no obvious effect on desulfurization efficiency. When more
water is sprayed in, the flue-gas temperature decreases and the time for evaporation is
extended. It has enough time to form a steady-state water film on the surface of the
sorbent, so that the desulfurization reaction will take place in the film and change into
an ionic reaction. The ionic reaction between Ca(OH)2and SO2 is much quicker than
the two-phase (gas-solid) reaction that occurs at lower water flow rates. Therefore,
the desulfurization efficiency rises much more rapidly.
It is also observed from both Figures 4a and 4b that the desulfurizationefficiency
is rather low (only about 15%) if no water is sprayed into the reactor irrespective of
the type of slaked lime used. This indicates that if a higher desulfurization efficiency
is required, then water should be sprayed to form a film on the surface of Ca(OH)*
particles, because SO2 reacts with Ca(OH)2much more quickly in the liquid phase
than in the solid phase. Neathery [5] showed the same results.
IV Eflect of the approach to saturation
Another important factor is the approach-to-adiabatic-saturationtemperature (AT). In
fact, its influence on the desulfurization efficiency is reflected by the water sprayed
into the reactor. The approach-to-adiabatic-saturationtemperature (which decides the
evaporation and desulfurization characteristics of the liquid-film) is the difference
between the average outlet temperatwe (i.e. the reaction temperature) of the flue gas
and the adiabatic-saturation temperature. Desulfurization efficiency increases
exponentially with decreasing AT. Lower AT is better for desdfurization because the
time for the water evaporation will be longer thus prolonging the reaction time
between SO2and the sorbent. Therefore, the efficiency of sorbent utilization and of
the desulfurization both increase, especially for the condition when the inlet SO2
concentration is high. Alternatively, it requires sufficient time for evaporation to keep
the sorbent dry at the outlet of the reactor so that no deposition will exist in the
cyclone. From this point of view, higher AT is required. Therefore, many aspects
such as the coal type, sulfur content, efficiency of desulfurization,sorbent utilization,
flow rate of flue gas, etc., should be considered systematically in order to determine
the value of AT.
Experimental and Mechanism Studies on a CFBfor Flue Gas Desurfurizatwn
There is little difference between the adiabatic-saturation temperature of the flue
gas and the dew point. Therefore, it is thought that the difference between the outlet
temperature and the dew point of the flue gas will be equal to AT. The functional
relationship between the desulfurization efficiency and AT is shown in Figure 5. The
desulfurization efficiency can reach a high level when AT is at a low value, even
when using Yuhua slaked lime that is not very active as a sorbent. However, the
desulfurization efficiency rises more quickly when using Baoshan lime. At the
experimental conditions with flue gas at 500 m3h, 1.1 CdS molar ratio, and 12'C AT,
the efficiencyof desulfurization can reach about 89%.
tYuhua lime
+Baoshea lime
Figure 5. Relationship between the desulfurization eficienq and AT.
V Preliminary mechanism analysis
When the simulated flue gas goes into the reactor, it contacts with the particles in the
lower part of the reactor where a high-density phase area exists. Some particles
themselves are the sorbent and they react with SO2 by means of physical or chemical
adsorption. Some are fly ash plus sorbent or the production of a desulfurization
covering on the surface.
After the water is sprayed into the reactor, the droplet impacts with the particles to
form a film on the surface. In this period, it is considered that the SO1 is absorbed by
the water film and ionizes:
SO, + H,O a H' + HSO;
HSO; @ H' f SO:205
Y.H. Wu,B. Feng, Z. Huang and D.J. Li
The sorbent ionizes in the film at the same time to form:
Ca(OH), # Ca2*+ 20H-
The ionized products then react in the film to form CaS03 by the reaction:
As the flue gas ascends, the speed of absorption reduces gradually due to the
evaporation of the film and the reducing concentration of SOz. Thus, the overall
reaction rate decreases. When the flue gas goes into the two-stage cyclone, the
reaction is re-intensified because of the increasing particle concentration in the
cyclone and the increasing relative slip velocity between gas and solid.
The characteristic of the CFB reactor is that a large amount of sorbent can be
circulated so that the efficiency of desulfurization and sorbent utilization are both
increased. When the sorbent particles have frictional contact and impact each other in
the reactor, the unreacted surface is revealed to the flue gas.
It was found that the SO2 removal efficiency of a circulating fluidized-bed reactor
can achieve efficiencies greater than 80% when using slaked lime as the sorbent with
a small amount of water added. The advantages of this technology are simple
equipment, small space requirement and lower capital cost.
The evaporation of the liquid film plays a very important role in the
desulfurization reaction. When the average reaction temperature is close to the dew
point of the flue gas, the desulfurization efficiency is higher.
1. Huang. Z. 1997. Experimtntal and applied research on circulating fluidized bed flue gas
desulfurizationtechnology. Ph.D. dissertation, Southeast University, China
2. Stouffer. M.R.. Rosenhoover, W.A., and Withum. J.A. 1992. Development of the advanced coolside
desulfurizationprocess. 9' Annual International Pittsburgh Coal Confmnce. pp.874-879.
3. Commission of Chemical Eslgineering. 1982. A Handbook of Chemical Engineering (volume 16):
Drymg. Chinese Chemical Industry Pnss. pp.44-52.
4. Wu, Y.H., Wang, W.L., Huang, Z,
and Li, D.J. 1998. Experimental studies on a circulating fluidized
bed for flue gas desulfurization.Journal of Southeast University, China,lql), pp.62-66.
5. Neathety, J.K. 19%. Model for fluegas desulfurizationin a circulatingdry scrubber. AIChE I., 42(1),
Без категории
Размер файла
375 Кб
experimentov, flue, velocity, mechanism, fluidized, desulfurization, bed, different, gas, studies, circulating
Пожаловаться на содержимое документа