close

Вход

Забыли?

вход по аккаунту

?

The Role of Ammonium-Sulfur Aerosols on Nitrogen Oxides Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation with Ammonia and Ozone Injection.

код для вставкиСкачать
Dev. Chem Eng. Mineral Process., 8(5/6),pp.483-503,2000.
The Role of Ammonium-Sulfur Aerosols on
Nitrogen Oxides Removal by Pulsed Corona
Enhanced Wet Electrostatics Precipitation
with Ammonia and Ozone Injection
C.-H. Tseng, T.C. Keener*,S.-J. Khan$ and J.-Y. Lee'
Department of Civil and Environmental Engineering, and 'Department
of Chemical Engineering, University of Cincinnati, Cincinnati,
Ohio 45221, USA
The removal of nitrogen oxides (NO and NO2)has been investigated in a bench-scale
pulsed-corona enhanced wet electrostatic precipitator (wESP). The 70 H z pulsed
voltage is applied up to 60 kV. Simulated flue gases with NO, concentration up to
1200 ppm have been used to determine the feasibility of NO, removal in she wESP.
NO must be oxidized to NO2 before any removal takes place. NO, removal eficiency
increased with gas residence time, inlet NO concentration and applied corona power.
In the air stream with I0 seconds gas residence time, up
to
20% of 1000 ppm NO
(22% of NO,) was removed at a 20 watt pulsed corona. The amount of in-situ ozone
was not enough to be considered as a major NO, removal mechanism in this wESP.
Ozone injection improved the NO, removal for both the corona and non-corona cases
by oxidizing NO. With 300 ppm of ozone, the removal of 750 pprn NO was increased
from 18% to 50% by oxidation, but total NO, removal was improved only to 25% .
* Author for correspondence. (email: tkeener@uceng.uc.edu).
483
C.-H.Tseng, T.C.Keener, S.-J. Khang and J.-Y. Lee
However, very high NO, removals were measured in simulated flue gas that
contained ammonia, sulfur dioxide and ozone. For instance, in a 3% oxygen, I I%
COzsimulated flue gas with 800 ppm NO and 70% relative humidity at 8.6 sec of gas
residence time, 30 kV corona discharge, the removal eficiency of N O was only 5%.
Injecting NHj only ( N H f l O x ratio 1) at 32 watts corona power, NO removal was
increased to 10%. With the co-presence of 2400 ppm SO2, 200 p p m ozone injection
(no ammonia) increased NO removal to 36% by oxidation, but total NO, removal
only to 17%. Afrer co-injecting 312 ppm ozone and 2900 ppm NH3 (stoichiometry
ratio 0.53), total NO, removal was increased to 79%. It was determined that the
ammonium salt aerosols produced fiom the reaction of ammonia and sulfirr dioxide
enhanced the NO, removal substantially. The significance of this catalyzing effect by
the in-situ ammonium salt aerosols is addressed in this paper.
Keywords: wet electrostatic precipitator, nitrogen oxides, denitrGcation, pulsed
corona, ammonium-sulfur aerosols, ozone.
Introduction
The economical removal of NO, still represents a significant technical challenge that
could ultimately prevent the use of certain types of fossil fuels for energy production.
Alternative postcombustion cleaning technologies have been developed, including an
innovative method utilizing a Iow-cost wet electrostatic precipitator (wESP)to remove
gaseous pollutants as well as particulate matter. Simultaneous removal of NO, and
SO, by corona discharge in an electrostatic precipitator requires smaller installation
spaces and investment costs than conventional combinations of scrubbing De-SO, and
catalytic De-NO, processes.
Wet ESPs have the following advantages over dry systems: No dust layer can be
built, especially for the aerosol liquid, tars, and oil mists contained in flue gases.
Therefore, back-corona, spark-over, and dust re-entrainment can be avoided. The
absorption of ammonia can reduce ammonia leakage.
In addition, wESPs
experimentally performed better than the dry ESPs in the removal of NO, [ 11 because
the chemical reactions involving water and its radicals help the removal of NO, 121.
484
NOx Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation
The objective of this research is to study the removal of NO, from combustion gas
in a pulsed corona enhanced wet electrostatic precipitator. This paper presents the
experimental results of a pulsed corona enhanced wet electrostatic precipitator to
remove nitrogen oxides.
Literature Review
Pulsed Corona Power Technologies
Pulsed corona technology is a cost-effective means for the control of many gaseous
pollutants, including NO,, SO,, mercury vapor, freon and other organic compounds
[3]. Corona discharges generate low-energy electrons (-20 eV) and the electrons are
accelerated from a very low level of lunetic energy as they drift along the high voltage
region (corona region) until they collide with a gas molecule and immediately lose
energy by excitation, attachment, dissociation, or ionization. After transferring its
energy to the molecule, the electron is re-energized by the electrical field.
Studies revealed that a pulsed corona exhibits higher removal efficiency than a DC
corona for the simultaneous removal of NO, and SO, [4, 51. Only electrons enable
non-elastic collisions with neutral molecules and produce active radicals.
In
conventional plasma processes, low gas pressures were used to minimize gradual heat
transfer from electrons to ions and molecules through collision processes. By pulsing
the high voltage, only electrons can be accelerated to gain sufficient energy to
generate radicals, whereas ions and molecules having much larger mass cannot be
sufficiently accelerated to get a concurrent energy loss. Furthermore, more energetic
electrons are present in discharge plasmas, which produce more free radicals for
pollutant removal or oxidation.
Masuda's work of Pulse-induced Plasma Chemical Process (PPCP) [5] showed that
De-NO, can be achieved by both positive and negative pulsing. NO is oxidized by
PPCP to NO?, which is removed again by PPCP. The efficiency of NO removal in
negative pulsing is a function of the specific power of pulsing (P/Q) divided by the
inverse of the square root of the gas residence time. Positive pulsed corona showed
higher efficiency for NO, removal 121. A positive corona produces longer streamers
485
C.-H. Tseng, T.C. Keener, S.-J. Khang andJ.-Y. Lee
fully filling the gas, ionizing larger volume at the same energy level, which results in a
larger active volume and higher energy. Negative coronas appear only in a small
region around the wires. The available corona space in a positive discharge is about
10 times longer than a negative corona [3]. However, as far as an energy-based
efficiency is concerned, there is no difference between polarities [3].
Removal Mechanisms
( i ) Electron Attachment
In a corona-discharge field with voltage range 3-15 kV, low-energy electrons are
generated. When electrons attach on NO molecules, negative ions (NO.) are formed
and then separated by electrical field. The first study of electron attachment of NO
was by Bradbury [6] who found that negative ions are formed with low velocity
electrons in low pressure. The attachment probability for NO linearly increases with
an increase in pressure. It was suggested that three-body attachment occurs since an
increase in pressure increases the fraction of all collisions.
Three-body attachment: e' + AB + M + AB' + M + energy
( i i ) Oxidation by Ozone
Generally, ozone is an unwanted by-product of ESPs. However, it is possible to use
high concentrations of ozone as a strong chemical oxidizer, resulting in the oxidation
of NO and the following removal of NO,. Ozone quickly oxidizes NO to NO2 or NO3
that is easier to dissolve into water or to form aerosols with ammonia.
Ozone is generated in a corona discharge [7]. NO can be oxidized to NO2 by free
oxygen and ozone in less than 0.1 sec [8].
NO + 0 + NO2
(1)
-
486
NO + O3 + NO2 + 02;log k = 10 11 cm3/mol-sec @300K [9]
(2)
3 NO + 0 3 + 3 NO2
(3)
[8]
NOx Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation
Ozone also oxidizes NO, to nitrogen trioxide or dinitrogen pentoxide [7]:
NO2 + O3 + NO3 + 0 2 ; log k = 7.5
2 NO + 3 03 + N 2 0 5
- 11 cm3/mol-sec @300K [9]
+3 0 2
(4)
(5)
These chemicals easily react with water to form nitric acids, thus removing the
original pollutants from the flue gas stream.
Simachev et al. [8] and Lozovskii et al. [lo] demonstrated that with the injection
of ozone/ammonia in a wet scrubber of a pilot plant, with an ozone/NO concentration
ratio less than 1, NO is oxidized to NO2 in the gas phase in an S02/N0 gas stream.
Ozone oxidized NO first and the oxidation of SO2 was 10% lower. The oxidation of
NO by ozone in gas scrubbing systems was independent to the liquid spraying
conditions because the oxidation is fast and in the gas phase alone.
(iii) Oxidation by Radicals
When a corona discharge (DC, AC or pulsed) is applied to the flue gas, energetic
electrons are created, and transfer energy to the dominant gas molecules (N2. 02,H20,
C02) by collisions, resulting in the formation of primary radicals (0,N, OH), positive
and negative ions and excited molecules. The subsequent electron-ion, ion-ion
reactions and electron detachments create more secondary radicals
(Hot,etc.) [ 111.
Large amount of 0, O;,OH, H radicals are easily generated in coronas. The radicals
either oxidize NO,, or react with them to form aerosols. Since the formation energy of
the radicals is in the order of 10 eV, the energy of the electrons in a corona discharge
is sufficient to produce the radicals.
In Ohtsuka's [5,7] DC corona discharge experiment, NO was oxidized to NOZand
N205when the corona field exceeds the ordinary level of precipitator operation, E =
4.5 - 9.0 kV/cm, with the existence of 0 2 and H20. 67% NO was oxidized by ozone;
33% by OH radical, 0, 0 , 0; etc.
Detailed reaction mechanisms were not
mentioned. These results show that DC corona oxidation of NO in a dry ESP could
have quite a high technical potential, but would require a high-energy consumption.
487
C.-H. Tseng, T.C. Keener, S.-J. Khang and J.-Y. Lee
NO, can be reduced to N2 and H20 by NH and NHz radicals in a dry corona
discharge [ 12, 131. However, increasing the concentration of 0 2 reduced the De-NO,
efficiency.[141 Therefore, NO can only be reduced under very low 0 2 concentration,
which may not be practical in the application of flue gas treatment. In Mizuno's [ 11
wESP tests, half of the removed NO was dissociated into N2 and 02,and the rest was
absorbed by water.
(iv) Ammonia in the Gas Phase
Ammonia (NH3) is widely used in the conventional selective catalytic reduction (SCR)
process as a reducing agent and in the electron beam method to convert NO, into
ammonium nitrate (NH4N03) aerosols. In the dry pulsed corona method, in addition
to forming ammonium salts, ammonia molecules are also dissociated to form NH and
NH2 radicals in the corona. The NH and N H 2 radicals react with NO and NOz during
De-NO, processes [2,4, 121.
Masuda et al. [ 5 ] mentioned that ammonia enhanced the removal of NO2, but not
NO nor SO2. This enhancement is possibly due to its NO2-scavenging effect from the
gas phase, while it does not enhance NO oxidation.
Mizuno et al. [l] showed that in a dry ESP from room temperature to 15OoC,
ammonia has effects only on the De-N02 efficiency, not on the De-NO efficiency.
This is because the conversion of NO to NOz does not depend on ammonia but mainly
on the concentration of free oxygen radicals [12]. In the co-presence of ammonia and
water vapor, however, both NO and NOx removal efficiency were enhanced in
proportion to the temperature and power input [ 11.
Veldhuizen et al. [ 151 showed that the NO conversion was strongly increased with
the addition of SO2 or NH3. The gas compositions used was 6% 02,8% C 0 2 , 16%
H 2 0 , 300 ppm NO, and 300 ppm SO2 at 80°C. The best cleaning result was 80% NO
and 95% SO2 removal with 3 ppm NH3 leak at residence time of 30 seconds when 600
ppm NH3 was introduced much later than Sol.
Since dry ESPs performed better with the addition of ammonia and water vapor for
De-NOx, wESP with ammonia injection also improves NO, removal [l]. In the
488
NOx Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation
system of NOX/NH3/H20 during corona discharge, NO, was converted mainly to
N&N03 aerosols [ 12, 163, which was confirmed later by infrared spectroscopy [2].
Experimental Description
A wESP was used to measure the NO, removal in a continuous flow system [17] as
shown in Figure 1. The wESP consisted of a mixing chamber, inlet and outlet ducts, a
collecting area in the top section for NO, removal studies, and a bottom section for the
collection and sampling of the water.
Figure 1. Schematic of the Wet-ESP Experimental System.
Table 1 following contains a summary of wESP experimental parameters.
489
C.-H.
Tseng, T.C. Keener, S.-J. Khang and J.-Y. Lee
Table 1. Summary of wESP Experiment Parameters.
Properties
Plate-plate spacing
I
1
Typicalvalue
0.2
I
1
Varible Range
0.05 0.35
-
I
I
Unit
m
Pure NO was mixed with laboratory compressed air or pure nitrogen in the mixing
chamber at room temperature. A~rflowrates ranged from 75 to 132 Urnin (ideal gas
residence times from 8.6 to 15 sec) with initial NO concentrations ranged from 400 to
1000 ppm were used in this study. In addition, air-nitrogen-COz mixtures at a airnitrogen volumetric ratio of 1:6 with 70% relative humidity were made to simulate
3%-02 and 6%-02 flue gases with 11% COZ,ideal gas residence times of 11.5 sec at
25°C. After a transition duct is a diffusion screen and an orifice where ammonia was
injected. The top part of the wESP carried the gas and contained the electrically
isolated stainless steal electrode wires and the grounded collection plates.
Plate-to-plate spacing was 20.3 cm, with discharge electrode wires spaced 10 cm from
the inlet and outlet and 7.6 em apart. Pressure inside the top box was slightly higher
than the atmospheric pressure to prevent dilution from outside air.
The gaseous
pollutants were removed to the water film running uniformly over the grounded plates
with a water flow rate 3.8 L/min. NO, concentrations were sampled at both the inlet
and outlet duct by a chemiluminescence NO-NOz-NO,
analyzer (Thermo
Environmental Instruments 42H analyzer). A Welsbach ozonator MD408 was used to
generate the ozone, and it was measured at the end of the collection plates by an
Ozone Analyzer 1003AH (Dasibi Environmental Cop., Glendale, California).
490
NOx Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation
The bottom half of the ESP collected the water and was divided into six sections
which separated the water into six phases. Each phase represented a step by step
reaction occurring in the wESP box. The first phase corresponded to the first sixth of
the gaseous residence time in the box; etc. For each experiment, water sampling from
the 6 water compartments was conducted 20 minutes after the gas concentrations
reached a steady state in order to ensure that the water concentrations reached a steady
state as well. EDTA solution was immediately added to the samples as a complexing
agent to inhibit nitrite (NO;) oxidation by metal ions. The samples were filtered,
sealed and refngerated before analysis.
Nitrate (NO<) and nitrite (NOi)
concentrations were determined by an ion chromatograph unit (IC, Dionex DX- 120,
Dionex Corp., Sunnyvale, California). The anion analysis of the raw tap water used
for the wESP experiments indicated that the nitrate ranged from 5.75-6.96 mgL.
The corona discharge was produced by a commercial high voltage transformer
(PS/WR 100 R2.5-11 Series WR, Glassman High Voltage Inc., Whitehouse Station,
New Jersey) which has the capability to produce both positive and negative direct
current voltage up to 100 kilovolts.
A maximum voltage for the system was
determined by experiment to be -60 kV due to the limitation of spark over.
A
pulsing module was developed and added between the DC power supply and the
discharge electrodes. High voltage was pulsed up to 90 Hz by a distributor that was
rotated by a variable speed motor. Pulsing allows the power level to be increased
without undue sparking.
A chiller-filter system was installed between the wESP outlet sampling point and
the on-line analyzers. The chiller system is used to condense aerosols and water vapor
to prevent water vapor from entering the analyzers. It consists of two single pass
double-pipe condensers in series. The sampled exhaust gas was cooled to -1°C by
flowing through the inner tubes of the condensers from bottom to top with coolant
flowing in the opposite direction. A filter system was installed in order to remove
aerosols before sampling. Two filter holders were installed in parallel between the
chiller system and the on-line analyzers. Filter papers with 0.22nm pore diameter
were used and switched as needed as the pressure head through one of the filters
increased to an unacceptable level.
491
C.-H.Tseng. T.C.Keener, S.-J.Khang and J.-Y. Lee
Results and Discussion
Initial tests were conducted to determine the effect of the electrostatic field on NO
removal under dry and wet conditions. Figure 2 shows that the corona discharge
caused a momentary drop in the outlet NO concentration, which returned to its
original value after a few minutes.
900
800
,
4
E
Q
C
.700
Corona on
8 32 k V . 1 4 watt
0
E
Outlet concentration
u
0
4
8
12
Time, min
Figure 2. NO Concentration in Dry ESP.
(3%02,800pprn NO, gas residence time 11.4 sec (2OOC. 70% RH),
without water; positive pulsed corona)
This demonstrated that NO molecules could be charged and moved to the
collection plate. In this dry system, the charge was quickly dissipated and the NO
molecules re-entered into the gas stream. After the system reached steady state,
equilibrium was established between the attraction and re-entrainment of NO, no more
net removal of NO was measured.
492
NOx Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation
I. Nitrogen Stream
Tests were conducted in the air stream and in the nitrogen stream to determine the
effect of oxidation on NO removal. In a pure N2 stream (tests #1-3 of Table 2), less
than 4% of NO is removed even though there are both corona discharge and water
film in the wESP. These results suggest that NO has to be oxidized to NO2 (or other
more soluble nitrogen oxides) before any significant removal takes place. The water
vapor from the evaporation in the wESP (41.8
- 43.2% relative humility for 3.8 Umin
water flow) oxidizes or absorbs only 0.8% of NO. A corona level of 31 kV (12 watts)
increased NO removal very little in a nitrogen stream. Therefore, NO, removal can
only be initiated after NO was oxidized by either
0 2
or the oxidizing radicals
developed from O2under the corona discharge.
Table 2. Test Removal of Nitrate Oxide in a Nitrogen Stream and in Air.
Test number
Carrier Gas
I
1
#1
I
#2
N7
I
#3
1
#4
1
I
#5
I
#6
Air
*Experimental conditions: 25'C air or nitrogen, 12°C water.
11. Air Stream
In air (test #4-6 of Table 2) approximately one tenth of the NO is oxidized to NOz in
10 seconds at room temperature. There was 17% removal of NO in the air stream
without corona discharge, which is attributed to the oxidation of NO by air and
ultimate NOr removal by water. The corona discharge increased the removal.
In Figures 3 and 4, tests were conducted to measure the NO, removal when the
power, gas residence time, and input NO, concentration were adjusted.
493
C.-H.Tseng, T.C. Keener, S,-J. Khang and J.-Y. Lee
Corona Voltage, kV
30
35
40
25%
i
-
2
10%
E
2
I
5%
rn 8.6 sec Res.tirne
0% I
10
5
0
15
20
25
Corona Power, watt
Figure 3. NO Removal Eficiency vs. Corona Power.
(in air (25°C 43%RH), 1000 ppm NO, 3.8 Umin water (IOOC);positive pulsed
corona)
30% 1
$
10%
I
1
0
a
0
2
f
15 sec Res.Tirne
t 10 sec Res.Time
+8.6 sec Res.Tirne
-
NO removal
Total NOx removal
Figure 4. NO and NO, Removal in Air vs. Input Concentration and Gas Residence
Time. (in air (25"C, 43%RH), 3.8 Umin water (12"C),positive corona discharge)
In non-corona tests, NO was oxidized to NO2 by
0 2
and absorbed by water in the
wESP. That resulted in 6-8% removal efficiency. In Figure 3, the application of
power increased the NO removal level by another 10 %. However, 20% seems to be
494
NOx Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation
the limit of NO removal efficiency when the residence time is less than 10 seconds.
In Figure 4, higher inlet NO concentrations led to higher removal efficiencies,
which are consistent with SO2 removal in the wESP [18]. This is explained by the
formation of a large gas concentration gradient at the gas-side boundary layer which
results in a new, higher equilibrium level not based on the bulk gas concentration.
Longer residence times lead to higher removal efficiencies as well.
(a) With Ozone Injection
The ozone concentration generated by the positive corona discharge was 0.23 ppm
under the same experimental condition in the wESP. Since the reaction between
ozone and NO, is at a one-to-one molar ratio, the amount of in-situ ozone would not
be enough to be considered a major removal mechanism in this process. Therefore,
ozone was generated externaIly and introduced into the gas phase in the wESP to
study the effects of ozone on the removal of NO,.
100
200
300
400
Ozone Concentration, ppm
Figure 5. The NO and NOx Removal vs. Ozone Injection in Air-NOx Stream.
(in air (25"C, 43%RH), 717-776 ppm NO (797-878 ppm NOx), 8.6 sec gas residence
time, 3.8 Umin water (12"C), positive pulsed corona.
The NO, removal results of ozone injection in air-NO, stream under positive
495
C.-H.Tseng, T.C.Keener, S.-J. Khang and J.-Y. Lee
corona are shown in Figure 5. NO, removal efficiency is lower than NO removal
efficiency because the NO2 level was increased at the wESP outlet. This NO2 increase
was from the oxidation of NO by ozone. These results show that ozone improved
both NO and total NO, removal. The presence of 300 ppm ozone improves both NO
removal (from -18% to -50%) and NO, removal (from -10% to -25%) in both the
corona and non-corona discharge cases. Corona has a limited effect on both De-NO
and De-NO, efficiency.
The NO, removal results of ozone injection in air-N0,-S02 stream are shown in
Figure 6. Experiments were conducted under positive corona discharge with ozone
injection into the corona discharge region. These results show that under positive
corona discharge, the presence of ozone improves the NO, removal with and without a
corona discharge, and for both air-NO, and air-SO2-NO, stream. Moreover, in an airS02-N0, stream, lower NO removal but higher total NO, removal was obtained with
the presence of SOZ. Therefore, the presence of SO2 improved the NO2 removal and
total NO, removal.
60%
1
60%
I
naR-NO%
\
I
T
I
--fm
.$ 40%
2
E
a
0%
I
J
100
200
300
Ozone niedim. pprn
400
m an-NO%
20%
88 OkV
-I.
-
,
100
200
300
4w
Ozone nlecllon,pprn
Figure 6. The NO and NO, Removal vs. Ozone Injection in Air-NO,and Air-S02-N0,
Streams.
(in air-NO, and air-S02-NOXstreams (25"C, 43%RH). -1100 ppm NO (-1230 ppm NO,), 10
sec gas residence time, 3.8 Umin water (12OC),positive pulsed corona)
(b) With Ammonia Injection
The NO removal results of ammonia injection in NO,-air stream under positive corona
are shown in Figure 7. The results show that ammonia did not improve the NO
removal in either the corona or non-corona cases in air (Figure 7a). However when
496
NOx Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation
SOz was present (Figure 7b), the catalyzing effect of the in-situ ammonium salt
aerosols improved the total NOx removal by almost 20%. This catalytic effect will be
discussed in detail in the following section.
40%
-
40%
in air-NOx stream
(no aerosol)
-in air-NOx-SOP stream
(Aerosol formed)
De-NOx
0-
i
w.
g 30%
?
Z
1
-
m
c3-
0
I
W
," 30%
-W
x
0
0
z
w
c
;10% +No Corona
+corona 30 k V
+Corona 46 k V
0%
-f
z
Corona 30 kV
0%
7
(7a) in Air-NOx stream
(7b) in Air-NOx-S02 stream
Figure 7, NO and NO, removal vs. Ammonia Injection in Air.
(in air (25°C. 80%RH), 714-776 pprn NO (762-878 ppm NO,), 8.6 sec gas residence time, 3.8
Umin water (12"C),positive pulsed coronal
111. Simulated Flue Gas without Ammonium Sulfur Aerosols
Humidified air-N~-CO~-NO,mixtures were made to simulate 396-02 and 6%-02flue
gases with ideal gas residence times of 11.5 sec at 25OC. Results are summarized in
Table 3 and Figure 8.
In Figure 8, the removal efficiency of 800 ppm NOx was reduced from 20-30%
(in air) to 5% (in 3%-02 flue gas) and 12% (in 6%-02 flue gas) because of the low
oxygen level. Corona power enhances the SO2 removal, but was not so helpful for
NO, removal. When a corona discharge was applied to a simulated flue gas, the
improvement of NO, removal efficiency was usually less than 3% before voltage
reached the operation limit.
497
C.-H. Tseng, T.C.Keener, S.-J. Khang and J.-Y. Lee
Table 3 Maximum Removal Eficiency in air and in simulated flue gas in the wESP.
I
Without in-situ ammonium sulfur
I
In-situ ammonium sulfur aerosols formed
Pulsed
WetESP
30 kV
Corona
Discharge
Ammonia
0
Ozone
Maximum SO2
Removal
0
50
55%
MaximumNO
Removal
(in Air)
MaximumNOx
Removal
(in Air)
MaximumNO
Removal
(3%-02 flue gas)
MaximumNOx
Removal
(3%-02 flue gas)
28%
-
-
25%
18%
14%
6%
-
39
43%
51
57%
36%
74%
82%
80%
78
-
80%
17%
13%
5%
72%
-
77%
79%
72
76%
\
(in air and in 3%02,1 ’% C02 humidified simulated flue gas (2SoC), -700 ppm NO,-2400
pprn SO*, 8.6 sec gas residence rime, 3.8 Umin water (12OC))
15%
- NO removal
- ToIal NOx removal
I
z
--
%10%
i
16M)ppm NH3 ( 2 1 )
-m
J
E
E
a
E
4
0
5
.................. ..-
B
5%
4
......................r m NH, injestion
_._._.-.-.-+.
TI
5
- NO removal
- Total NOx removal
5%
4
4
n simulated flue gas: 3% 4
11.4 Res. time
and 11%
n simulatedHue gas: 6%4 and
114G02.9.6 sec Res. time
a.
0%
0%
10
20
Corona Power, watt
30
10
20
30
Corona Power, wan
(8a) 3%-02 Flue Gas
(8b) 670-0~
Flue Gas
Figure 8. NO, Removal with Ammonia Injection in Simulated Flue Gases. (in 3%-02
and 6%-02,11% COz humidi$ied simulatedjlue gas (25°C). 800 ppm NO, no SOz, 11.4 sec gas
residence time, 3.8 Umin water (12’C), positive pulsed corona)
498
I
NOx Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation
Ammonia was injected upstream of the wESP to provide a reaction time of 16.5
seconds before entering the corona discharge region.
In non-corona cases, NO
removal was increased by only 2% at a stoichiometric injection ratio 1:1. When the
ammonia injection was doubled, no significant increase of De-NO and De-NO,
efficiency was observed because no additional NO2 was generated. Ammonia only
helps the removal of NO2, not NO [ 1, 5, 121.
In corona discharge cases, however, removal efficiency increased with ammonia
concentration because additional NO was oxidized to NO2 in a corona discharge
although the amount of oxidation is very small (-2%). These results indicate that
oxygen sources are critical to the oxidation of NO, which leads to the NOx removal in
wESPs.
The corona power in a Wet-ESP is not enough to drive significant chemical
reaction to NO.
The low oxygen content in the flue gas can not oxidize NO
sufficiently to achieve a significant NOx removal. Therefore, ozone was injected into
simulated 3%-02 flue gas as an oxidizer. Without ammonia injection, 200 ppm ozone
injection increased the NO removal efficiency from 6% to 36% by oxidizing NO to
NOz (see Table 3). However, total NOx removal increased only from 5% to 17%,
which indicated NO2 can not be removed efficiently in the wESP.
IV. Simulated Flue Gas with Ammonium Sulfur Aerosols: Catalyzing Effect of
In-situ Aerosols
Very high NO, removals were measured when the in-situ ammonium sulfur aerosols
were formed in simulated flue gas that contained ammonia, sulfur dioxide and ozone
(see Table 3). With the co-presence of 2400 ppm S02, 200 ppm ozone and 2500 ppm
NH3 (ammonia to pollutants stoichiometry ratio 0.45), total NO, removal increased to
66%.
Further increasing the additives to 312 ppm ozone and 2900 ppm NH3
(stoichiometry ratio 0.53), total NO, removal increased to 79%.
Figure 9 clearly shows the catalyzing effect of in-situ aerosols. Without aerosols,
ozone improved the total NO, removal very little. When aerosols were formed, 200
ppm of ozone resulted in the maximum total NO, removal. Higher ozone input did
not further improve the removal. Doubling NH3 did not further increase the amount of
499
C.-H. Tseng, T.C. Keener, S. -J. Khang and J. - Y.Lee
aerosols and the NOx removal. Since the 2900 ppm of input NH3 is about the
concentration of removed SOz (98% of 2400 ppm) and NO, (72% of 700 ppm), the
formed ammonium salt aerosols were more likely to be ammonium bisulfite
(NI-&HS03)or bisulfate (N&HS04), alone with ammonium nitrate (NH4N03).
100%
no NH3 injection
A NH3 2900 ppm
8 80%
0 NH3 5800 ppm
2
C
.-a
\with
NH,
(with aerosol)
60%
W
J
9
40%
dx
4
/
20%
0%
w
,-.h
tiout
NH3
(without aerosol)
T
0
50
100
150
200
250
300
350
Ozone Injection, ppm
Figure 9. The Improved NO, Removal vs. Ammonium and Ozone in Simulated Flue
Gas. (in 3% 0 2 , 11 % C02 humidified simulated flue gas (25"C), 700 ppm NO, 2400
ppm ,902,
8.6 see gas residence time, 3.8 Umin water (12"C),positive pulsed corona,
30-40 kV)
Ta
NOx Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation
The aerosols were collected and analyzed by CHNS analyzer for C, H, N contents
and total sulfur analyzer for the sulfur content. The oxygen content was estimated by
assuming the remaining content is oxygen. The samples were analyzed twice and the
results are listed in Table 4. Assuming 98% of 2400 ppm SO2 and 72% of 700 ppm
NOx were converted to NhHSO4 and NhN03, the aerosols should be consist of
8 7 % of
~ W H S 0 4 and 1 3 % of
~ N&N03. The ultimate contents of this estimated
aerosol mixture are close to the analysis results (see Table 4).
In the cases without ozone, the catalyzing effect of aerosols was limited.
Therefore, ozone is still a necessary oxidizer. In the 200 ppm ozone cases, aerosols
enhance NO removal by at least 30%. It is believed that ozone, NO, H20 and NH3
were adsorbed to the surface of aerosols. NO was oxidized and formed NH4N03 at
the same time. This explains why not only NO2 removal, but also NO removal was
improved when the aerosols were present.
It was determined that the ammonium salt aerosols produced from the reaction of
ammonia and sulfur dioxide enhanced the NO, removal substantially. The aerosol
formation itself is a SO2 removal process that improved the SO2 removal approaching
to 100%. The in-situ aerosols with sub-micro diameter were well spread in the flue
gas. These aerosols serve as highly efficient adsorbent and provide tremendous
surface area to catalyze the De-NO,chemical reactions. After the SO2 and NOx were
separated from gas, these small aerosols can be easily removed from gas in a wESP.
Conclusions
1.
A bench-scale pulse-enhanced wESP was constructed to test the removal of NO,.
Simulated flue gases with NO, concentration up to 1200 ppm were used to
determine the feasibility of NO, removal in the wESP.
2.
There is no appreciable amount of NO removal in a pure N2 stream, although
there are corona discharge and water film in the wESP. NO has to be oxidized to
NO2 before any removal takes place.
3.
It was experimentally found that higher input NO concentration resulted in higher
501
C.-H.Tseng, T.C.Keener, S.-J. Khang and J.-Y. Lee
removal efficiencies. NO, removal efficiency increased with gas residence time,
inlet NO, concentration, and applied corona power. In the air stream with 10
seconds gas residence time, up to 20% of loo0 ppm NO (or 22% NO,) was
removed using a 20 watts pulsed corona discharge.
4. Ammonia injection did not improve the NO, removal in both corona and noncorona cases in air. The amount of in-situ ozone is not enough to be considered
as a major NO, removal mechanism in a wESP. NO and total NO, removal are
improved with in-situ ozone injection because NO is oxidized by ozone. The
-
presence of 300 ppm ozone improves both the NO (from 18% to 50%) and NO,
removal (from -10% to 20%).
5.
In simulated flue gas (3% 02,11% C02, 800 ppm NO), the removal efficiency of
NO, is only 5%. Adding NH3 (NHflO, ratio 1, no ozone) at 32 watts corona
discharge, NO, removal was increased to 10%. In 6%-O2simulated flue gas, NH3
injection (NHJNO, ratio 1) increased NOx removal from 10% to 13%. 200 ppm
ozone injection (no ammonia) increased NO removal from 13% to 36% by
oxidation, but total NO, removal from 10%to 17%.
6. High NO, removals were measured when the in-situ ammonium sulfur aerosols
were formed in simulated flue gas that contained NH3, SO2 and ozone. With the
2400 ppm SOz, 200 ppm ozone and 2500 ppm NH3 (ammonia to pollutants
stoichiometry ratio 0.45),total NO, removal increased to 66% in 3%-02
simulated flue gas. Further increasing the additives to 312 ppm ozone and 2900
ppm NH3 (stoichiometry ratio 0.53), total NO, removal increased to 79%.
7. The ammonium sulfur aerosols may serve as highly efficient adsorbents with
tremendous surface area and catalyze the De-NO, chemical reactions. It is
believed that ozone, NO, HzO and NH3 were adsorbed to the surface of aerosols.
NO was oxidized and formed NI&N03 at the same time.
References
1.
Mizuno. A.; Shimizu, K.; Matsuoka, T. and Furuta, S. 1995. Reactive Absorption of NO, using Wet
Discharge Plasma Reactor. JEEE Transactions on Industry Applications, 3 1(6), 1463-1467.
2. Chakrabmi, A.; Mizuno, A.; Shimizu, K.; Matsuoka. T. and Furuta, S. 1995. Gas Cleaning with
Semi-Wet Type Plasma Reactor. EEE Trans. on Industry Applications, 31, 500-506.
502
NOx Removal by Pulsed Corona Enhanced Wet Electrostatics Precipitation
3. Masuda, S. 1990. Non-Equilibrium Plasma Chemical Process PPCP and SPCP for Control of NOx.
SOXand Other Gaseous Pollutants. 4th International Conference on Electrostatic Precipitation, Chap.
9, 615-623.
4. Dinelii, G.; Civitano, L. and Rea, M. 1990. Industrial Experiments on Pluse Corona Simultaneous
Removal of NOx and SO2 from Flue Gas. IEEE Transactions on Industry Applications, 26(3), 535-41.
5. Masuda, S. and Nakao, H. 1990. Control of NOx by Positive and Negative Pulsed Corona Discharges.
E E E Transactions on Industry Applications, 26(2), 374-83.
6. Bradbury, N.E. 1934. Formation of Negative Ions in Gases by Electron Attachment, Part I. NH3, CO,
NO, HCI and Clz. J. Chemical Physics, 2,827.
7. Ohtsuka, K.; Yukitake, T.and Shimoda, M. 1985. Oxidation Characteristics of Nitrogen Monoxide in
Corona Discharge Field. Proc. Inst. Electrostat Japan, 9(5), 346-351.
8. Simachev. V.Y.; Novoselov, S.S.; Svetlichnyi, V.A.: Gavrilov, A.F.; Gorokhov, M.V.; Semenov, V.I.;
Ryzhikov, V.A. and Demchuk, V.V. 1988. Investigation of the Ozone-Ammonia Method of
Simultaneous Desulphurisation and Denitrification of Flue Gases when Burning Donetsk Coals.
Thermal Engineering, 35(3), 171-175.
9. Kerr, J.A. and Moss, S.J. 1982. CRC Handbook of bimolecular and Termolecular Gas Reactions, CRC
Press, Inc., Boca Raton, Florida
10. Lozovskii, V.A.; Novoselov, S.S.; Svetlichnyl, V.A.; Ryzhikov, V.A.; Simachev, V.Y. 1988. The Role
of Ozone in Oxidizing Reactions when it Interacts with the Flue Gases from Thermal Power Stations.
Thermal Engineering, 35(8), 442-444.
11. Eichwald, 0.; Yousfi, M.; Hennad, A. and Benabdessadok, M.D. 1997. Coupling of Chemical
Kinetics, Gas Dynamics, and Charged Particle Kinetics Models for the Analysis of NO Reduction
from Flue Gases. J. Applied Physics, 82,4781-94.
12. Chang, J.S. 1989. Role of H2O and NH3 on the Formation of NbNO3 Aerosol Particles and De-NOx
under the Corona Discharge Treatment of Combustion Flue Gases. J. Aerosol Science, 20(8). 10871090.
13. Masuda, Senichi; Hosokana, S.; Tu. X.L.;Sakakibara, K.;Kitoh, S. and Sakai, S. 1993. Destruction of
Gaseous Pollutants by Surface-Induced Plasma Chemical Process (SPCS). IEEE Transactions on
Industry Applications, 29(4), 781-786.
14. Helfritch, D.J. Aug. 22, 1997. Plasma Technologies Applied to Air Pollution Control. Workshop on
Applications of Electrostatics for Control of Gas Phase Air Pollutants, The University of Cincinnati,
Cincinnati, Ohio.
15. Veldhuizen, E.M.; Zhou. L.M. and Rutgers, W.R. 1998. Combined Effects of Pulsed Discharge
Removal of NO, S02, and NH3 from Flue Gas. Plasma Chemistry and Plasma Processing, 18(1). 9 1111.
16. Ohkubo, T.;Kanazawa, S.;Nomoto, Y.;Chang, J.S. and Adachi, T. 1994. NOx Removal by a Pipe
with Node-Plate Electrode Corona Discharge System. IEEE Transactions on Industry Applications,
30(4), 856-861.
17. Tseng, C.H.; Keener, T.C. and Khang, S.J.June 20-24, 1999. Nitrogen Oxides Removal by Pulsed
Corona Enhanced Wet Electrostatics Precipitation, Air & Waste Management Association's 92&
Annual Meeting, St. Louis, Missouri, USA.
18. L a r h n , K.. Tseng, C.H.; Keener, T.C.and Khang, S.J.June 8-13, 1997. Sulfur Dioxide Removal by
Enhanced Elecuostatics. Air & Waste Management Association's 9 0 Annual
~
Meeting, Toronto,
Ontario, Canada.
503
Документ
Категория
Без категории
Просмотров
1
Размер файла
886 Кб
Теги
ammonia, precipitation, sulfur, ammonium, wet, aerosol, pulse, electrostatic, removal, oxide, injections, ozone, nitrogen, role, enhance, corone
1/--страниц
Пожаловаться на содержимое документа