-------
Table 1
PILOT ESP RESULTS
MASS MEASUREMENTS, BASELINE
6/24/89	6/26/89
SYSTEM INLET
Temp., °F	308	315
gr/scf	4.199	4.018
lb/MMBtu	9.258	8.296
DSCFM	8553	9017
OUTLET®
Temp., °F	276	277
gr/scf	0.0045	0.0033
lb/MMBtu	0.012	0.008
EFFICIENCY11, %	99.87	99.90
SCA, ft2/kacfm	231	215
OMEGA K, cm/sec	97.1	112.7
aNo Rapping
bBased on lb/MMBtu
10

-------
Table 2
PILOT ESP RESULTS
MASS MEASUREMENTS, E-SOX CONDITIONS
10/22/89	10/25/89
POST DID INLET
Temp., °F
gr/scf
lb/MMBtu
DSCFM
170
3.891
7.867
9267
160
3.891"
7.867
8493
180
3.891"
7.867
8493
OUTLETb
# of Runs
Start Time
End Time
2
1116
1527
1
1002
1110
1
1430
1546
Temp., °F
gr/scf
lb/MMBtu
175
0.1288
0.313
162
0.6352
1.552
184
0.0464
0.113
EFFICIENCY0, %
96.02
80.27
98.56
SCA, ft2/kacfm
257
287
278
OMEGA K, cm/sec
20.5
4.7
32.9
aInlet Data of 10/22/89 Used
bNo Rapping
cBased on lb/MMBtu
11

-------
Table 3
AVERAGE ELECTRICAL OPERATING CONDITIONS


Average

Average
Current

Voltage,
Density,
Condition
kV
nA/cm2
Baseline, 6/25/89, 281°F
46.1
56.2
E-SOX Slurry, 10/25/89, 180°F
45
48.9
E-SOX Slurry, 10/25/89, 160°F
42.9
50.6
12

-------
scale with a maximum of 0.5, whereas the 160°F segment was recorded with a
maximum scale reading of 2.0.
ESP performance changed from excellent under baseline conditions (0.012 or 0.008
lb/MMBtu out) to unacceptable (0.113 to 1.55 lb/MMBtu out) with slurry injection.
Note that inlet data for slurry conditions were obtained downstream of the DID
array in a low velocity region, and therefore the efficiency data in Tables 1 and
2 are not directly comparable. Average process conditions on 10/22 and 10/25
were as follows: Ca/S ratio = 1.44; S02 removal = 48.5%; and the estimated total
mass loading of dried, partially sulfated sorbent and fly ash was 10.2 gr/scf.
Using this estimate as a basis, the mass train traverse in the low velocity
region recovered about 38% of the total mass.
"Omega K" values in Tables 1 and 2 represent an ESP performance parameter that
provides a semi-quantitative means of comparing performance under various con-
ditions.2 As points of reference, omega k values of full scale ESPs collecting
ash downstream of spray dryers have been reported to range from 27 to 62 cm/sec.
The recent paper by Durham, et al.3 provided data on the Shawnee TVA ESP Spray
Dryer pilot plant which indicated omega k values ranging from 32 to 53 cm/sec.
Thus, the highest efficiency "no-rap" data from the E-S0X system at 180°F
indicate a performance parameter in the lower portion of the range reported for
spray dryer applications.
The decrease in outlet emissions measured by the mass train with increasing
temperature on 10/25/89 is confirmed by the P5A trace shown in Figure 6. This
trend toward higher emissions at lower operating temperatures was observed at
earlier times in the test program, and proved to be a reproducible phenomenon.
The large spikes appear to represent non-rapping reentrainment occurring on a
massive scale at the lower temperatures.
A related observation concerns the appearance of rapping spikes on the P5A output
which coincides with rapping of the DID array. This rapping process is expected
to produce relatively large particles which would be charged and be driven to the
collecting electrodes quickly with the observed voltages and currents. However,
large rapping spikes were observable at the outlet with a time lag corresponding
to the gas transit time between the DID array and the P5A sampling point. This
observation indicates that a large fraction of the sorbent/ash mixture is
instantaneously and repeatedly reentrained by electrical forces as it travels
through the electrical fields.
The electrical operating points in Table 3, and the voltage-current curves in
Figures 3 through 5, reveal no anomalies which would explain the extremely high
outlet emissions with sorbent injection. As will be shown in a subsequent dis-
cussion of observed vs predicted performance, the measured voltages and currents
indicated that very high electrical migration velocities and collection
efficiencies would be predicted under both baseline and slurry conditions. It
is also of interest to note that a significant change in voltages and currents
did not occur as temperature was increased from 162 to 181°F, although mass
emissions decreased by a factor of 13.7.
13

-------
An extensive trouble shooting effort was performed during the test program in an
attempt to locate possible mechanical problems that might be responsible for the
excessive particulate emissions with slurry injection.
Specifically:
•	Hopper fluidizing air was turned off and on;
•	The ash removal screw conveying system was turned off and on;
•	Nozzles and the DID array were periodically cleaned both on line
and while the system was down for short-term repairs; and
•	Voltages were held to values below those which would cause
excessive sparking during the test periods.
The above items related to mechanical issues did not result in a reduction in
outlet emissions; excessive sparking did, however, increase outlet emissions.
Also, the No. 4 precharger was energized and de-energized, with no apparent
effect on the outlet mass monitor.
Overflowing hoppers are one source of emissions which could not be directly ruled
out by observation. However, the ash removal system was monitored to ensure that
it was operating. Furthermore, the reproducible change in outlet emissions with
temperature is not explainable by hopper overflow since the lower emission data
were obtained later during the same test day.
In view of potential flow disturbances due to the presence of sorbent injection
nozzles and the DID array, temperature traverses were conducted during the test
program. A velocity traverse with air flow was also performed after the test
program with the perforated plate downstream from the DID array in an uncleaned
condition. Tables 4 and 5 illustrate the inlet temperature distribution on
October 25 when the average inlet temperatures were 160 and 180°F, respectively.
Table 6 contains the inlet velocity traverse. Note that high velocity and low
temperature areas coincide near the bottom of the ESP. This combination of
conditions, where reentrainment by electrical forces is greatest in the region
of highest velocity, would be expected to exacerbate the ESP performance
problems. Also, during the test program, an average temperature of 155°F was
obtained from a traverse at the outlet of the fourth field of the ESP, while the
average inlet temperature was 173°F. As expected, the lowest outlet temperatures
occurred near the bottom of the ESP.
The test program data clearly indicated that operation in the 180°F region
improved ESP performance. However, operation at temperatures far above the
adiabatic saturation point is not an acceptable solution for excessive
particulate emissions because of the adverse effects on S02 removal.
Particulate Characterization
Figure 7 contains cumulative inlet size distributions obtained by impactors at
the E-SOx pilot facility, and Figure 8 illustrates baseline and E-SOx size-
dependent efficiencies obtained from impactor data. Also shown on Figure 8 are
14

-------
104 p
I "'I I I I I I I I
I I I I I I I I
I	I I I I I I I
10J
CO
V)
<
u
>
(-
<
-i
3
D	9
O 10'
10 1
0.1
~ BASELINE
• E-SOX (POST DID)
n -| nV mfc «
1.0
10
100
PARTICLE SIZE./xm
Figure 7. Inlet Cumulative Mass vs. Particle Size for Baseline and Post-DID E-SOX
Conditions
15

-------
PARTICLE SIZEf(u.m
Figure & Measured and Modeled Penetration vs. Particle Size, E-SOX Demonstration Tests
16

-------
Table 4
TEMPERATURE DISTRIBUTION AT ESP INLET
AVERAGE INLET GRID TEMPERATURE EQUALS 160°F
Position	Position	Position	Position
One	Two	Three	Four
op	op	op	op
Top Row 172	175	176	177
Third Row 168	165	164	164
Second Row 157	153	156	150
Bottom Row 148	147	147	146
GRID IS VIEWED IN	DIRECTION	OF GAS FLOW
17

-------
Table 5
TEMPERATURE DISTRIBUTION AT ESP INLET
AVERAGE INLET GRID TEMPERATURE EQUALS 180°F
Position	Position	Position	Position
One	Two	Three	Four
°F		°p	°p	°p
Top Row 195	186	190	196
Third Row 187	182	184	187
Second Row 176	171	172	175
Bottom Row 173	166	165	172
GRID IS VIEWED IN	DIRECTION	OF GAS FLOW
18

-------
Table 6
GAS FLOW MEASUREMENTS AT PILOT ESP INLET
VELOCITIES IN FEET PER MINUTE
Position Position Position Position
Top Row
Third Row
Second Row
Bottom Row
One
ft/min
388
385
282
655
Two
ft/min
275
235
205
520
Three
ft/min
335
245
230
455
Four
ft/min
400
325
465
640
GRID IS VIEWED IN DIRECTION OF GAS FLOW
1)	Perforated Plate Uncleaned From
Operability Test
2)	Mass Gas Flow 44,000 lb/hr, Temp. 98°F
3)	Grid Average is 378 ft/min.
19

-------
ESP model projections for baseline and E-SOx conditions. These projections
include the effects of 10% sneakage and reentrainment with a gas velocity
standard deviation of 0.25. The model projections will be discussed later.
Total mass loadings obtained with the impactor traverses are presented in
Table 7.
If it is assumed that each slurry droplet produces one agglomerated particle upon
drying, a size distribution can be estimated for dried and partially sulfated
sorbent. Figure 9 contains an estimated size distribution of the dried sorbent
downstream of the DID array that was supplied by B & W.* A Bahco-derived size
distribution of ash/sorbent mixture obtained from the ESP hopper is also provided
on Figure 9.
The impactor data in Figure 7 indicate no significant difference in total inlet
loading between the E-SOx and baseline conditions in the size range resolved by
the impactors (below about 8 /*m diameter). An examination of the dried
agglomerate size distribution in Figure 9 reveals that only 15% of the slurry
residue would be expected to consist of particles 8 jim in diameter and smaller,
which is consistent with the lack of increase that was observed in this size
range by inlet impactors. The low total mass loading obtained with the impactors
under E-S0X conditions results from the fallout and impaction which occur in the
spray chamber and on the DID array, and from the difficulties in obtaining a
representative total mass sample in the low velocity region behind the DID array.
A comparison of the baseline and E-S0X fractional efficiency curves in Figure 8
indicates a large drop in efficiency under E-S0X conditions across the entire
size range. Total average mass loading obtained with the outlet impactors on
10/20 and 10/23 at 170°F (Table 7) is similar to that obtained with the mass
trains at 180°F on 10/25 (0.0497 vs. 0.0464 gr/scf). It is of interest to note
that the penetration ratio of E-S0X to baseline conditions at 3 /*m diameter is
similar to the results of Durham3 for the Shawnee spray dryer ESP (13 for E-SOjj
and 10 for Shawnee).
A comparison of the ESP model projections with baseline data in Figure 8 shows
good agreement between model projections and measured results. In contrast, the
E-S0X measured data exhibit a large decrease in efficiency instead of the
increased performance projected by the ESP model. This indicates that ESP
performance with sorbent is dominated by factors which are not represented in the
existing model. All of the model projections are "no rap" cases, since the pilot
ESP rapping system was not energized during sampling.
Penetration vs particle size curves such as those in Figure 8 are calculated by
taking the ratio of outlet to inlet particle mass in a given size interval.
However, the dried slurry residue is an agglomeration of smaller particles which
originated from the lime slaking process, and the potential exists for
deagglomeration to occur in the ESP. It has been hypothesized that slurry
agglomerates could be broken apart due to internal forces arising from the
relatively high values of charge acquired by particles in the interelectrode
region. These forces must overcome the cohesive forces between the individual
particles. The ID fan upstream of the outlet sampling ports would not be
expected to provide sufficient shear forces to deagglomerate particles in the
size range of interest. If deagglomeration or decrepitation did occur, the
20

-------
Figure 9. Bahco and Predicted Slurry Residue Size Distribution
21

-------
Table 7
PILOT ESP IMPACTOR DATA, BASELINE
Inlet	Outlet
Date 6/25 & 6/27	6/25
Mass Loading
gr/acf 1.7834	0.0034
gr/scf 3.0160	0.0053
Avg. Temp., °F 306	270
Avg. MMD, pra 19.9	5.1
Avg. SCA, ft2/kacfm	208
PILOT ESP IMPACTOR DATA,	E-SOX CONDITION
System Inlet	ESP Inlet	Outlet
Date 10/20	10/23	10/20 & 10/23
Mass Loading



gr/acf
2.163
1.114
0.0372
gr/scf
3.405
1.567
0.0497
Avg. Temp., °F
264
170
170
Avg. MMD, (xm
21
17
4.3
Avg. SCA, ft2/kacfm
272


22

-------
fractional efficiency curves would represent the net of the collection and
decrepitation processes.
Evidence that a significant degree of slurry residue decrepitation did occur is
presented in Figures 10 and 11. These figures contain plots of the signal ratio
of calcium to iron and calcium to silicon from an energy dispersive X-ray (EDX)
analysis of inlet and outlet impactor substrate samples. There is a large change
in the relative amounts of calcium to iron and silica from the inlet to the
outlet. This change is consistent with the hypothesis that slurry residue
agglomerates were broken apart, so that the relatively fine, calcium-rich
particles dominate the smaller size fractions in the outlet samples.
Photomicrographs of outlet impactor stages 4 and 6 are presented in Figure 12 for
baseline and E-S0X conditions. Also illustrated is a large agglomerate captured
in the cyclone stage of the inlet impactors. The outlet stages illustrate that
the impactors were classifying the sampled particles, and no evidence of gross
reentrainment of larger particles from the upper stages was observed. If
reentrainment occurred in the impactors to a significant degree, the change in
composition with impactor stage could not be attributed to compositional
differences as a function of particle size.
Further evidence that deagglomeration of slurry residue can occur is given in
Figure 9, in which it can be observed that a Bahco size distribution of a sample
obtained from the ESP hoppers contains more fine particles on a relative basis
than the estimated slurry residue distribution entering the ESP. Figure 13,
which provides composition of the Bahco size fractions as a function of particle
size, illustrates that the finer size fractions are dominated by calcium-rich
material.
Chemical analyses of samples collected from several points in the system are con-
tained in Table 8, and coal compositions of representative samples are presented
in Table 9. It is of interest to note that the outlet sample composition is very
similar to that obtained at the ESP inlet. This observation provides further
evidence that the slurry droplet residues were not retained in the ESP as
predicted. Since the agglomerated slurry residue has a small fraction of fine
particles, theoretical collection efficiency vs particle size relationships
predict that the ESP would selectively collect the slurry residue so that the
outlet particle mass would contain a significantly smaller fraction of calcium
compounds.
Reliable values for the electrical resistivity of sorbent/ash mixtures are
difficult to obtain with standard methods. The in situ point-plane resistivity
probe has been reported to selectively reentrain low resistivity sorbent
particles,3 and this process is likely to have biased data obtained with the
probe during this test series. In addition, samples collected at various points
in the system at different times during the slurry injection test program
exhibited significantly different resistivity vs temperature curves, as Figure
14 illustrates. The laboratory data were obtained using a modified procedure
that was adapted for samples containing calcium sorbents.5
An examination of the in situ and laboratory data for 10/24 and 10/25 (Figure 14)
indicates resistivity values in the 1010 to 1011 ohm-cm range at 160-180°F, which
23

-------
Figure 10. EDX Derived Calcium to Silicon Ratios vs. Particle Size for E-SOX Impactor
Substrate Samples
24

-------
0.1	0.2	0.5	1.0	2.0	5.0	10.0
PARTICLE SIZE,/xm
Figure 71. EDX Derived Calcium to Iron Ratios vs. Particle Size for E-SOX Impactor
Substrate Samples
25

-------
10
15
10
10
14
13
10
12
£
o
£
3
>
>
i—
«
10
11
10
10
10b
108
I
11
II
I
107
LABORATORY
• ESP; 6/26; 6.7% H20; 7 ppm SO3
O ESP; 10/3; 10.9% H20
~ ESP; 10/24: 9.8% H2O
O ESP; 10/26: 9.8% H20
IN SITU
~ BASELINE
A E-SOX CONDITION
3.0
60
140
2.8
84
183
2.6
112
233
2.4
144
291
2.2
182
359
2.0
227
441
1.8
283
541
1000/K
°C
°F
TEMPERATURE
Figure 14. Laboratory and In Situ Resistivity Measurements, E-SOX Demonstration Tests
28

-------
Table 8
BURGER PILOT SYSTEM E-SOX SOLIDS

Mass


Mass Train


Train
HoDDersb
After

%
Inlet8
Trans.
ESP
DIDb
Outlet1
Li20
0.03
0.02
0.02
0.01
0.01
Na20
0.3
0.2
0.2
0.2
0.1
K20
1.9
1.0
0.9
0.8
0.7
MgO
1.1
1.4
1.5
1.5
1.3
CaO
3.0
30.6
40.2
39.2
38.3
Fe203
22.8
13.9
7.7
6.6
6.4
A1203
22.4
11.9
10.1
9.0
8.5
Si02
47.0
26.0
20.9
17.9
17.3
Ti02
1.3
0.6
0.6
0.5
0.4
p2o5
0.5
0.1
0.1
0.2
0.2
S03
0.6
13.6
16.2
21.0
24.7
a.	Ash only
b.	Ash plus sorbent
29

-------
Table 9
COAL COMPOSITIONS FROM BASELINE AND E-SOX TEST SERIES
ANALYSES BY CT&E
%		Baseline	E-SOX Condition
H20	8.25	6.67
C	64.83	67.89
H	4.25	4.32
N	1.24	1.33
CI	0.00	0.06
S	2.64	2.67
Ash	12.18	11.02
Volatile	33.50	34.78
Fix. C	46.08	47.54
Btu/lb	11627	12120
30

-------
should not result in low resistivity reentrainment. However, the samples of 10/3
indicate resistivity values below 1010 ohm-cm at 170°F, and exhibit a very steep
slope which would result in resistivity values of less than 108 ohm-cm at 140°F.
It should be noted that the laboratory atmosphere is only a static simulation of
the dynamic environment which exists in the ESP. It could be argued that the
presence of sulfur oxides and surface moisture in the actual environment is
likely to produce a lower real-time resistivity value with a representative
sample than would be obtained in the laboratory air-water vapor environment with
samples obtained from system hoppers.
Baseline resistivity data are also shown on Figure 14. The probe provided data
in good agreement with the laboratory data in equilibrium with 7 ppm S03. This
value of S03 was estimated to result from the sulfur content of the coal. The
observed fact that neither the probe nor the ESP experienced any difficulty in
collecting an ash with a resistivity of 109 ohm-cm suggests that low resistivity
may not be the only property of a dust responsible for excessive reentrainment.
Ash cohesivity is also expected to be relevant in efforts to quantify factors
responsible for excessive reentrainment. Measurements of cohesivity of fly ash
and ash/sorbent mixtures were performed on fly ash alone and on the ash/sorbent
mixtures from the ESP with slurry injection. These measurements produced the
surprising result that the E-S0X solids at 145°F exhibited a cohesivity in the
low end of the range (40.3°, angle of internal friction6) measured for a large
number of fly ash samples. Furthermore, the angle of internal friction of the
fly ash at 285°F was 45.5°, which is significantly higher than the E-S0X solids
at 145°F. Low cohesivity would be expected to aggravate a reentrainment tendency
resulting from low dust resistivity.
Discussion
Table 10 contains a summary of measured and model predictions of outlet mass
loading and efficiencies for the E-S0X pilot facility. Since all measurements
were conducted with rappers off, the model estimates are all "no rap" values.
The modeling factors represent the fraction of sneakage/reentrainment which is
assumed to occur over four stages, and the sigma g value is the normalized
standard deviation of the gas velocity distribution. The value of 0.25 is
assumed for the baseline test; the value of 0.36 is based on the measurements
presented in Table 6.
A comparison of measured and predicted results shows that the model failed to
predict the performance trends as well as the absolute value of outlet emissions.
Performance improvements were predicted because of increased electrical migration
velocities at the lower temperatures with slurry injection. However, outlet
emissions increased with slurry injection from predicted values by factors
ranging from 28 to 390. Model output could be forced to match the measured
results by assigning reentrainment values per stage ranging from 40 to 85%.
These results are qualitatively similar to those of Durham,3 in which model pre-
dictions of the Shawnee spray dryer ESP were off by a factor of 80, and a reen-
trainment factor of 60% was required to match measured and modeled penetrations
under spray dryer conditions. However, there are significant quantitative
differences in that omega k values for the E-SOx unit ranged from a low of
31

-------
Table 10
MODELING RESULTS - BURGER PILOT ESP
Measured Performance		Modeled Performance
Mass Loading
lb/MMBtu
Efficiency
%
Mass Loading
lb/MMBtu
Efficiency
%
Modeling
Factors8
Baseline 0.008
99.92
0.009
99.91
0.10/0.25
E-SOX (anticipated)

0.004
99.94
0.10/0.25
E-SOX (180°F) 0.113
98.56
0.100
98.7
0.40/0.36
E-SOX (170°F) 0.313
96.02
0.274
96.5
0.55/0.36
E-SOX (160°F) 1.55
80.27
1.8
76.9
0.85/0.36
aCombined sneakage & reentrainment/<7B
32

-------
4.7 cm/sec to a high of 32.9 cm/sec. In contrast, the Shawnee data indicated
omega k values ranging from a low of 32.3 cm/sec with a two field configuration
to 51 to 53 cm/sec with three or four fields.
The extreme temperature sensitivity of the E-SOjj ESP performance is believed to
result from the combined high-velocity/low-temperature regions near the bottom
of the precipitator. These conditions would magnify the process of electrical
reentrainment which also appears to be occurring to some degree at the higher
temperatures. Since the particulate emission levels with slurry injection were
unacceptable, additional testing was performed after B & W improved the inlet gas
flow. Results from these measurements are provided in the following section.
33

-------
SECTION 3
ADDITIONAL MEASUREMENTS WITH IMPROVED INLET GAS FLOW CONDITIONS
Introduction
The gas velocity and temperature distribution data in Table 6 indicated a
coincidence of the high velocity and low temperature regions near the bottom of
the precipitator. B & W personnel modified the gas flow distribution prior to
the performance of additional precipitator testing. The objective of this
additional testing (termed the E-SOx ESP Performance Extension Test) was to
determine whether the modifications to the ESP inlet gas distribution would
permit achievement of the performance goals for the program. These goals are 50%
S02 removal with less than 0.1 lb per million BTU particulate emissions.
Table 11 presents gas velocity data obtained by B & W personnel following
modifications to the inlet baffling. These data indicate a significant
improvement from the distribution illustrate in Table 6. For example, the V/Vavg
ratio for the bottom row is only 1.18, whereas the comparable measurements from
Table 6 indicate a V/Vavg ration of 1.50.
Similar improvements were obtained in temperature distributions. For example,
B & W data taken with slurry injection after the modifications show that the
average bottom row temperature is 160.0°F when the inlet average is 160.5°F.
Site preparation for the additional field work began on July 30, 1990, and
testing ended on August 18, 1990. When the pilot facility was operational, the
ESP was operated during the daylight hours and kept warm during the night with
the system in a closed loop configuration. SRI obtained inlet mass measurements
during the baseline tests, and outlet mass measurements during the baseline and
slurry injection tests. Particle size measurements were obtained only on August
17, 1990 at the outlet sampling location and after the DID (Droplet Impingement
Device) at the ESP inlet. Voltage-current data were gathered during the various
tests conditions when the system was considered stable.
Mass Concentration Measurements
Mass measurements were obtained on August 8 and 9, 1990, at the system inlet and
outlet sampling locations under baseline conditions and these data are presented
in Table 12. Table 13 contains the data in Table 12 but averaged by day and
includes efficiency, SCA, and omega K calculations and baseline data from the
1989 test program. The specific collection area (SCA) and precipitation rate
parameter (omega K) calculations assume that all four fields are operational.
As can be seen from the data in Table 12, the outlet emissions tended to decrease
during the day when the conditions of the ESP were set to remain constant for the
test day. If the average inlet mass loading is used for the ninth, the
efficiency of the ESP went from 98.44% to 99.7% over an eight and one half hour
period. The ESP did not have sufficient time to equilibrate under the daily
operating conditions of the test program. The increase in efficiency with
operating time suggest non-rapping reentrainment emissions decreased as the
temperature of the electrodes increased from night time closed loop operating
34

-------
Table 11
E-SOx FIELD PILOT DEMONSTRATION
ESP FLOW DISTRIBUTION TESTS 8/2/90
(DATA OBTAINED BY B&W)
Conditions: Atomizing Air at 121-130 psig
Perforated Plate Cleaned
Horizontal Perf Plate at Bottom of DID Covered
6" Side Baffles Installed
Grid shown is view looking with the gas flow
Column
Number
1
2
3
4
5
6
Row Ave
Row 10
1.47
1.19
1.20
1.19
1.25
1.48
1.30 V/Vavg

373
301
303
300
317
374
328 feet/min
Row 9
1.38
1.19
1.19
1.19
1.15
1.21
1.22

350
301
300
300
292
305
308
Row 8
1.20
1.16
1.08
1.13
0.81
0.75
1.02

304
293
272
287
204
191
259
Row 7
0.74
1.01
1.05
1.15
1.02
0.63
0.93

186
255
265
290
259
159
236
Row 6
0.58
0.83
1.01
1.02
0.90
0.56
0.82

248
210
256
257
227
141
207
Row 5
0.50
0.49
0.81
1.03
0.76
0.74
0.72

126
124
205
260
192
188
183
Row 4
0.74
0.76
0.88
0.95
0.84
0.84
0.83

186
192
223
241
212
213
211
Row 3
1.13
0.80
0.82
0.90
0.79
0.99
0.91

287
202
207
227
200
251
229
Row 2
1.28
1.04
0.88
0.96
1.05
1.25
1.08

325
262
222
243
265
317
272
Row 1
1.41
1.12
1.04
1.04
1.07
1.40
1.18

356
283
264
262
270
355
298
Col Avg
1.04
0.96
0.99
1.05
0.96
1.01


264
242
252
267
244
256







Grid
253 feet/min






Avg
4.2 feet/sec
IGCI ESP
INLET VELOCITY DISTRIBUTION
STANDARDS



99% of points with V/Vavg <
1.40 %
of points
with V/Vavg < 1,
. 40 Vavg 1
85% of points with V/Vavg <1.15 % of points with V/Vavg <1.15 Vavg
91
80
35

-------
Table 12
E-SOx PILOT PRECIPITATOR DATA, BASELINE
INLET MASS TRAIN MEASUREMENTS




Gas
Flow

Mass Loading®

Date
Run ID
%
o2
Avg.
Gas
Temo.
acfra
dscfm
gr/acf
gr/dscf
lbs/
Million
Btu
Condition
8-8-90
DBIN-1
6.6
291
8,936
5,505
2.2081
3.5849
7.3502
Baseline
8-8-90
DBIN-2
6.6
290
9,353
5,642
2.6068
4.3215
8.8605
Baseline
8-9-90
DBIN-3
7.3
291
11,341
6,912
2.8242
4.6341
9.9904
Baseline
8-9-90
DBIN-4
7.3
286
11,112
6,812
2.6588
4.3377
9.3516
Baseline
OUTLET MASS TRAIN MEASUREMENTS




Gas
Flow

Mass Loading




Avg.




lbs/



%
Gas




Million

Date
Run ID
q2
TemD.
acfm
dscfm
gr/acf
cr/dscf
Btu
Condition
8-8-90
DBOT-1
9.2
259
9,572
6,498
0.0131
0.0193
0.0484
Baseline
8-8-90
DBOT-2
9.2
261
9,752
6,510
0.0058
0.0087
0.0219
Baseline
8-9-90
DBOT-3
10.2
259
11,315
7,609
0.0370
0.0551
0.1510
Baseline
8-9-90
DB0T-4
10.2
267
11,378
7,545
0.0104
0.0157
0.0429
Baseline
8-9-90
DBOT-5
10.2
259
11,426
7,682
0.0072
0.0107
0.0294
Baseline
aAll mass loadings were obtained, without electrode rapping
36

-------
AVERAGE
8/8/90
SYSTEM INLET
Temp., °F	291
gr/scf	3.953
lb/MMBtu	8.105
DSCFM	5574
Table 13
BASELINE RESULTS
8/9/90	6/24/89	6/26/89
289	308	315
4.486	4.199	4.018
9.671	9.258	8.296
6862	8553	9017
OUTLET3
Temp., °F	260
gr/scf	0.014
lb/MMBtu	0.035
EFFICIENCY1*, %	99.57
SCA, ft2/kacfm°	367.5
OMEGA K, cm/sec	41.0
262 276	277
0.027	0.0045	0.0033
0.074	0.012	0.008
99.23	99.87	99.90
299.3 231	215
40.2 97.1	112.7
aNo Rapping
bBased on lb/MMBtu
°Inlet Gas Flow Data Used
37

-------
conditions (approximately 200°F) to the gas temperature maintained during the
test series.
The average baseline efficiencies for August 8 and 9, 1990 were 99.57% and
99.23%, respectively. These data are considerably different from those of the
1989 baseline tests (see Table 2, last two columns) in that the SCA's are higher
and the omega K's are much lower (these calculations are based on four field
operation). It should be noted that the baseline data taken during the June 1989
test period were obtained with flue gas flowing through the ESP 24 hours per day.
Slurry testing began during the afternoon of August 11, 1990, but operational
problems with the facility delayed further testing until the August 15. Table
14 contains the mass data from the outlet sampling location during the slurry
injection tests. The first two mass measurements of August 15, 1990 were single
port tests of only 20 minutes duration. These tests were used to establish the
flow and temperature conditions of the slurry tests program so that the
objectives of the program could be met. The last test on the 15th was also of
20 minutes duration. This test was conducted with the average ESP inlet
temperature at 158.9°F in order to estimate emissions level of future tests.
Table 15 presents the data obtained by B & W for the various conditions tested,
along with the mass data obtained by SRI during these test segments.
On August 18, 1990, the average ESP inlet temperature was set at approximately
160°F for the additive tests. The first two outlet mass measurements were
determined with a one hour run time; all other measurements on the 18th were for
30 minutes. Each of the three sampling ports were traversed during each test on
the 18th. When alum (Al2(SO<,)3*xH20 where x is approximately 14) was added to the
slurry, the outlet emissions increased under the conditions tested. It was
expected that the outlet emissions would decrease with the use of the additive,
but this did not occur. After the second test with alum, the average inlet
temperature to the ESP was raised to determine whether the outlet emissions would
decrease. Past experience indicated that emissions decreased with increasing
temperature, if there were no contributing difficulties other than the expected
low resistivity and/or deagglomeration of lime slurry particulate within the ESP.
Once again, the outlet emissions decreased when the average temperature was
increased to 171.5°F. These data provide no evidence of a beneficial effect due
to the use of alum. This same conclusion was reached when calcium chloride was
added to the slurry for the last two tests of the extended test program. The
elevated emissions of the calcium chloride tests may have been in part due to the
length of the test day during which the ESP performance deteriorated throughout
the 160°F test day.
Figure 15 presents the data in Table 14 plotted as omega K vs temperature. This
temperature-emissions relationship was evident during the 1989 test program as
it was during the 1990 extended test program. The data in Figure 15 displayed
as solid symbols were obtained when the temperature (average ESP inlet) was
quickly changed to check emissions and the system was not given time to reach
thermal stability. Note that the ESP performance exhibits a drastic
deterioration as the inlet temperature approaches 160°F.
38

-------
E-SOx PILOT ESP
OMEGA K vs TEMPERATURE
AUGUST 1990 SLURRY TESTS
45
40
O 1990 SLURRY TESTS
• DAY END TEST, 8/15/90
A QUICK DELTA T TEST
35
30
25
20
15
o
o
10
3
0
j	i i I i i i
J	I	I	L
J	L
155	160	165	170
AVG. INLET TEMPERATURE, DEG. F
175
Figure 15. Omega K vs. Temperature for August 1990 E-SOX Extension Test.
39

-------
Table 14
E-SOx PILOT PRECIPITATOR EXTENSION TEST
OUTLET MEASUREMENTS - SLURRY INJECTION
Date
Run ID
8/11/90 DBOT-6
Inlet
Gas
Temp.
Gas
Flow
acfm
160.8°F 11202
Mass
Loading
lbs/MMBtu
0.3888
Reference
SCAa
299.9
Collection
Efficiencyk
91.92
Omega K,
cm/sec
10.7
Test Conditions
Lime Slurry,
Low Temp
8/15/90
•o
o
DB0T-7
DBOT-8
DBOT-9
DBOT-10
-167
167.7
164.5
158.9
12293
11896
10248
11481
0.1760
0.1372
0.0651
0.1187
273.3
282.4
327.9
292.7
96.34
97.15
98.65
97.53
20.3 Lime Slurry
22.8 Lime Slurry
28.7	Lime Slurry
Lime Slurry,
23.8	Lowered Temp
8/16/90 DBOT-11 167.6
DBOT-12 168.6
9906
9830
0.0682
0.0413
DBOT-13 166.7 11234 0.0663
DB0T-14 164.3 10192 0.0722
339.2
341.8
299.1
329.7
98.58
99.14
98.62
98.50
27.1	Unit 8, Lime
Slurry
33.6 Unit 7, Lime
Slurry
31.2	Lime Slurry
27.2 Lime Slurry
(continued)

-------
Table 14 (continued)
E-SOx PILOT PRECIPITATOR EXTENSION TEST
OUTLET MEASUREMENTS - SLURRY INJECTION
Date	Run ID
8/18/90 DBOT-15
DBOT-16
DBOT-17
¦P-
h-1
DBOT-18
DBOT-19
DBOT-20
DBOT-21
Inlet
Gas
Temp.
159.3
162.0
160.5
171.5
158.1
160.0
158.5
Gas
Flow
acfm
9995
10071
10256
10314
10289
10104
10222
Mass
Loading
lbs/MMBtu
0.4129
0.7306
1.1190
0.1360
0.8772
1.3893
2.7798
Reference
SCAa
336.2
333.6
327.6
325.8
326.6
332.5
328.7
Collection
Efficiency13
91.42
84.81
76.74
97.17
81.77
71.12
42.22
Omega K,
cm/sec
9.1
5.4
3.3
19.8
4.5
2.4
0.5
Test Conditions
Lime Slurry, Low
Temp
Lime Slurry
+ 1 wt. % Alum
Lime Slurry
+ 1 wt. % Alum
Lime Slurry
+ 1 wt. % Alum
170°F Set Pt.
Lime Slurry
Lime Slurry
+ 1 wt. % CaCl2
Lime Slurry
+ 1 wt. % CaCl2
aUsed outlet acfm for estimate of flow
bEstimated inlet loading to ESP is 4.811 lbs/MMBtu

-------
Imoactor Measurements
Particle size distributions were obtained at the ESP inlet, after the DID, and
at the ESP outlet on August 17, 1990. The condition established for this test
day was for the ESP to operate at approximately 165°F and the Ca/S ratio to be
approximately 1.2. These conditions were expected to produce 50% removal of S02
and less than 0.1 pounds per million Btu of particulate exiting the ESP.
Figures 16 through 19 present the inlet particle size data as cumulative mass
loading, cumulative percent, DM/DLOG D and DN/DLOG D vs. particle size,
respectively. Figures 20 through 23 present these data for the outlet particle
mass concentrations. The average inlet mass loading, using the impactor
loadings, was 1.576 gr/acf or 2.28 gr/scf. The average of the outlet data
resulted in a loading of 0.0249 gr/acf or 0.0343 gr/scf (data from 1989 were
0.0497 gr/scf). This equaled an emission level of 0.083 lbs/MMBtu at an SCA of
approximately 310 ft2/kacfm (four field operation). Past comparisons of
calculated outlet emissions for mass trains and impactors have indicated that
impactor measurements usually provide a mass concentration average which differs
by 25% or less from that obtained with mass trains. This difference is due to
the inability of the impactor to sample isokinetically at each sample point.
Figure 24 presents the fractional collection efficiency for the pilot ESP on
August 17, 1990. As past data have indicated, the apparent minimum in collection
efficiency occurs at approximately one micron particle diameter. The average
overall efficiency for the 17th, using the impactor data on a lbs/MMBtu basis,
was 98.18% (these data were obtained without the rappers operating during the
impactor measurements).
Voltage-Current Data
Table 16 contains the averages of secondary voltage and current readings taken
during the various conditions stated in the table. Figures 25, 26, and 27
present voltage vs current density curves for the baseline and slurry injection
conditions. These curves were obtained at the end of the test day indicated,
where conditions were considered to be fairly stable. The fourth field voltage-
current readings were abnormal. The fourth field transformer was connected to
the third field to check for possible misalignment, but the readings did not
change, indicating a problem with the TR set or the associated instrumentation.
It was thought that the current metering loop was in error by approximately a
factor of ten. There was no equipment on site to check this, nor time after the
test program was terminated to investigate the problem. Figure 27 also has a
curve included that is a fourth field curve assuming the current for the voltage
read on the 17th of August 1990 would have corresponded to that of October 25,
1989. The data in Table 16, for the eighteenth of August, do not indicate that
there were any significant differences in the voltage and current data under the
various tests conditions of that day.
Electron Microscopy
Impactor substrates from an inlet impactor (Brink) and an outlet impactor
(University of Washington) were subjected to Energy Dispersive X-Ray (EDX)
analysis. These data, presented as calcium to silica ratio vs. particle size,
44

-------
90 % CONFIDENCE LIMITS
D1LLIES BOTTOM. AUGUST 1990
3H0 - 2.42 GM/CC MASS < 0.49 MICRONS INCLUDED IN FIT
t! 0 '
i o V
1 0J"
o
<
\
o
o
Q
<
o
1 02-
CO
% 10'
LU
i 
-------
90% CONFIDENCE LIMITS
D1LLIES BOTTOM. AUGUST 1990
RHO - 2.42 GM/CC MASS <0.49 MICRONS INCLUDED IN FIT
99.99 _
99.95
99. 9
99.8
ttt¥
99.5
99
98
95
90
80
70
60
50
40
30
20
1 0
5
2
1
0-5
0.2
0. 1
0.05
0.01
1
P
I
i
i
H	1 I I I I I l|	1	1 I I I I I 11	1	1 I I I I I
0"'	10°	10'	102
PARTICLE DIAMETER (MICROMETERS)
Figure 17. Inlet Cumulative Percent vs. Particle Diameter for the E-SOX Extension Test,
A ugust 17,1990.
46
-------
90 % CONFIDENCE LIMITS
D1LL.IES BOTTOM. AUGUST 1990
SHO - 2.42 GM/CC MASS < 0.49 MICRONS INCLUDED IN FIT
1 0 4x
1 0 -r
ro
H
Q
\
o
Q
CD
O
_l
Q
\
1 02-:
01"
0°
1 0
J11
£
*X**

I if

I I I I I M[	1-
1 0
-H-H-)	1	1 MINI)
10'	10
PARTICLE DIAMETER (MICROMETERS)
Figure 18. Inlet DM/DLOGD vs. Particle Diameter for the E-SOX Extension Test,
A ugust 17,1990.	. 7
-------
90 7. CONFIDENCE LIMITS
DILLIES BOTTOM. AUGUST 1990
RHO - 2.42 GM/CC MASS < 0.49 MICRONS INCLUDED IN FIT
1 O13^-
1 0,2i
2IO"i
o
W 1 o 104-
LL) 1 u *
1,0.,
<
Q_
O
1 08==
o 1 07±
o
a
\
Q
i o6*
1 0 =r
10"
1 0
1

* T
i-
i ! | *
j- I i ''
! i
1 j
	1	1 I I I I 111	1	1 1 I II M|	1	1 I II
'	10°	10'	102
PARTICLE DIAMETER (MICROMETERS)
Figure 19. Inlet DN/DLOGD i/s. Particle Diameter for the E-SOX Extension Test,
A ugust 17, 1990.
-------
90 % CONFIDENCE LIMITS
D1LL1ES 30TT0H (OUTLET). AUGUST 1990
RHO - 2.55 GM/CC MASS < 0.24 MICRONS INCLUDED IN FIT
0
04t
103::
1 O2"
O
<
\
o
o
Q
<
o
CO
< 1 0
LU
1 0°x
o
1 0
-i
1 0
-1


fi1
-10°
-I 0
T
I
T
i

o
<
\
fy
CD
O
Q
<
O
<
LlJ
1 n ~ 3 <
i o	,
T
T
T
T
J_1 n -4
ZD
C ^
l i i i 111
H	1 I I I II 11	1	1 I I I I it
0
1 0
1 0
PARTICLE DIAMETER (MICROMETERS)
Figure 20. Outlet Cumulative Mass vs Particle Diameter for the E-SOX Extension Test,
August 17, 1990.
49
-------
90% CONFIDENCE LIMITS
D1LL1ES BOTTOM (OUTLET). AUGUST 1990
99.99
99. 95 iL
99. 9 1L
99.8 ::
99. 5 iL
99 1L
98 ][
95 ::
90 I
80
70
60
50
40
30
20
1 0
5
2
1
CO iL
0.2
0. 1
0.05
0-01
3HO - 2.55 GM/CC MASS < 0.24 MICRONS INCLUDED IN FIT
t ¥
xi

n
r
x
.i
**•
1 0
	1—I I I III l[	1—I I I 11 III	1—I I I 11 111
10°	10'	102
PARTICLE DIAMETER (MICROMETERS)
Figure 21, Outlet Cumulative Percent vs. Particle Diameter for the E-SOX Extension Test,
August 17, 1990.
50
-------
90 % CONFIDENCE LIMITS
DILL1ES BOTTOM (OUTLET). AUGUST 1990
RHO - 2.55 GM/CC MASS <0.24 MICRONS INCLUDED IN FIT
1 03t
1 O2-:
ro
~
\
O
o
o
o
_l
Q
\
1 0 1 -r
0°-:
1 0

H	1 I I I I I I
H	1 I I I I I 11
f- I I 1 I I 11
1 o2
0"'	I0U	10'
PARTICLE DIAMETER (MICROMETERS)
Figure 22. Outlet DM/DLOGD vs. Particle Diameter for the E-SOX Extension Test,
A ugust 17,1990.
51
-------
90 % CONFIDENCE LIMITS
D1LUES BOTTOM (OUTLET) , AUCUST 1990
SHO - 2.55 GH/CO MASS < 0.24 MICRONS INCLUDED IN FIT
1 o'2¥
1 oni
10,0i
ro
O . ~9
OS:
CO
UJ
" 10ai
h-
oc
£ io7*
o
5 1 06i
Q
O
° 105i
Q
\
Q 1 04i
i o^=
1 02
1 0
X J.
If.
I

-1
H	1 1 I I 1 11|	1	1 1 I 1 I 11|	1	1 I II 1 I l|
0°	¦ 10'	10
PARTICLE DIAMETER (MICROMETERS)
Figure 23. Outlet DN/DLOGD vs. Particle Diameter for the E-SOX Extension Test,
August 17, 1990. ^
-------
PENETRATION-EFFICIENCY
90 7. CONFIDENCE LIMITS
D1LL1ES BOTTOM. AUGUST 1990
RHO -2.42 GM/CC
1 02t
1 0 1 -r
1 o ° —
1 O"1"
1 0
-2
1 0
TII
TntlT
¦M]
qj _
[)

I
if T
t'ti
H	1 I I I
i o°
1 0
1
tO.O
190.0
-99. 0
+ 99-9
i
-L
-L
-L
H	1 I 1 I I 111	1	1 I I I
1 0
99. 99
2
PARTICLE DIAMETER (MICROMETERS)
Figure 24. Fractional Collection Efficiency, E-SOX Extension Test, August 17, 1990.
53
-------
CSj
E
o
«T
c
CO
z
LU
Q
I-
Z
UJ
cc
cc
3
O
E-SOx PILOT ESP
VOLTAGE-CURRENT CURVES
AUGUST 9, 1990
10
20	30	40
SECONDARY VOLTAGE, kV
50
Figure 25. Voltage-Current Curves for August 9, 1990; Baseline.
54
-------
E-SOx PILOT ESP
VOLTAGE-CURRENT CURVES
AUGUST 11, 1990
" I 8"" I " '1	1 I I I I
40 -
35
o
FIELD 1
•
FIELD 2
&
FIELD 3
A
FIELD 4
CM 30
E
o
<"
c
jf
CD
Z
LJJ
Q
I—
LLI
DC
DC
ZD
o
15 -
10 -
5 "
10
30	40
SECONDARY VOLTAGE, kV
Figure 26, Voltage-Current Curves for August 11, 1990; Lime Slurry.
55
-------
40
CM
£
o
-------
Table 16
E-SOx PILOT PRECIPITATOR
AVERAGE OF VOLTAGE AND CURRENT READINGS
AUGUST, 1990
Date
TR #1
kV	mA
TR #2
kV mA
TR #3
kV mA
TR #4
kV mA
8/9/90 35.4 7.5 42.5 21.5 48.1 28.6 41.4 3.2
Conditions
Baseline
Flyash
8/11/90 38.5 7.5 39 21.9 38 29 38.3 3.3
8/11/90 38.3 7.6 39.3 21.8 37.7 30 37.7
8/16/90 42.75 17.8 38.8 30.5 37.8 30 34.3
40.4 18
8/17/90 41.2 18
8/18/90 41	18
45.25 18
44.75 18
45.5 18
38.8 30 38.3 30 35.5
39.7 31.2 40.3 30 36.7
40 28 41 29 39
44 17.9 41 29.7 42.8 29.7 39.7
4.2
4.2
4.1
4.2
42 30 42 29.5 39.5
41 29 42.5 30 40
41.3 29.2 42.9 29.8 40.5 4.1
Baseline
Zeroed Meters
Ca/S =1.05
160°F
Unit 8
Ca/S =1.33
Unit 7
Ca/S =1.15
Impactors Today
Ca/S =1.2
165°F
Slurry @
Ca/S =1.45
160°F
Slurry + alum
Ca/S =1.35
160°F
Slurry + alum
Ca/S =1.35
170°F
Slurry
Ca/S =1.36
160°F
Slurry + CaClo
Ca/S =1.35
160°F
57
-------
are contained in Figure 28, while Figure 29 presents the data as calcium to iron
ratio vs. particle size. These data are similar to those from the 1989 test
series in that they demonstrate an enrichment of calcium in the outlet fine
particle size bands. This observation again suggests deagglomeration of slurry
residue within the ESP.
Discussion
With the modifications to the ESP inlet gas flow and temperature distribution,
data in Table 15 indicate that 50% removal of S02 and emissions of less than 0.1
lb/MMBtu were attained during one of the tests and possibly during the impactor
test day. Steady state conditions were never achieved during the pilot ESP
extension tests due to the fact that the ESP was placed in a closed loop
arrangement each night. Because of this, long term conclusions cannot be drawn
from these results. Although the temperature maldistribution at the inlet of the
ESP was solved, the dependence of the outlet emissions on temperature is still
quite evident. The data in Table 14 and Figure 15 indicate clearly the effect
of temperature and outlet emissions as related to the E-S0X process at the Burger
Station facility.
The additives, alum and calcium chloride, were expected to reduce the outlet
emissions by increasing the tensile strength properties of the particulate layer,
thus reducing reentrainment within the ESP. Others have reported the positive
effect of calcium chloride7 used in spray dryer FGD systems followed by an ESP.
The EPRI article states that S02 removal and the removal efficiency of the ESP
were enhanced with chloride addition. Neither of these increases were observed
during the limited time the chloride additive was injected at the E-SOx facility.
The dependence of ESP collection efficiency on inlet temperature, while slurry
was injected, was again demonstrated, as the data in Figure 15 illustrate. The
outlet emissions were reduced while alum was being added by raising the average
inlet temperature 10°F.
The voltage-current data indicated a possible problem with the fourth field
transformer, but there were no spare transformer-rectifiers available for
substitution during the test program.
58
-------
E-SOx PILOT ESP
EDX SIGNAL RATIOS vs PARTICLE SIZE
Ca/Si RATIOS FROM IMPACTOR SUBSTRATES
O OUTLET
• INLET
'	'	¦ ¦ i i ¦ .
J	I	' ' ' '
10
-1
10u
PARTICLE SIZE, micrometers
10
Figure 28. Ca/Si EDX Ratios vs. Particle Size; August 1990.
59
-------
0
LL
oj
O
1-
<
cc
(D
CO
X
a
LLI
E-SOx PILOT ESP
EDX SIGNAL RATIOS vs PARTICLE SIZE
Ca/Fe RATIOS FROM IMPACTOR SUBSTRATES
150 "
125 -
100 -
75 -
50 -
25
0 -
O OUTLET
• INLET
I I I I I |
10
-1
I I I I 1 I I I
' ' 1	' ) ' '
J	I	1—1	¦ ' '
10u
PARTICLE SIZE, micrometers
101
Figure 29. Ca/Fe EDX Ratios vs.Particle Size; August 1990
60
-------
SECTION 4
CONCLUSIONS
1.	Analysis of particle size fractions collected on impactor stages at the
inlet and outlet of the E-SOx ESP showed a large increase in the relative
calcium content of the finer size fractions across the ESP.
2.	Massive reentrainment of ash/sorbent mixtures could be induced without
electrode rappers in service by lowering the operating temperature of the
ESP inlet. The reentrainment could be reduced by elevating the average
inlet operating temperature 10 to 20°F with no accompanying change in
secondary voltages and currents.
3.	ESP performance for the E-SOx process, as evaluated at the Burger station
with the coal, lime and conditions present during testing, is dominated by
two factors not represented in the existing EPA-SRI versions of the
mathematical model of ESP performance. These factors are instantaneous
reentrainment of low resistivity ash/sorbent particles and deagglomeration
of slurry residue within the ESP.
4.	Significant improvement of the velocity and temperature profiles downstream
from the DID array allowed outlet particulate emissions to be reduced to
less than 0.1 lb/106 Btu with 50% S02 removal. However, the severe
reentrainment problem within the ESP was still present, especially at
temperatures below 160°F.
5.	Additional work would help develop a quantitative understanding of the
chemical and physical properties of slurry residues which result in poor ESP
performance. Slurry additives designed to increase dust layer tensile
strength and reduce reentrainment showed no beneficial effects during the
brief test periods that were possible in the current program. Additional
testing with these additives could include longer term and more stable
process operating conditions.
61
-------
SECTION 5
REFERENCES
1.	K. E. Redinger, et al. "Results from the E-SOx 5 MWe Pilot Demonstration."
In: Proceedings: 1990 S02 Control Symposium, Volume 4, EPA-600/9-91-015d
(NTIS PB91-197244), May 1991.
2.	S. Maartmann. "Experience with Cold Side Precipitators on Low Sulfur
Coals." In: Symposium on the Transfer and Utilization of Particulate
Control Technology, Volume I, EPA-600/7-79-044a (NTIS PB295226), February
1979.
3.	M. D. Durham, et al. "Low-Resistivity Related ESP Performance Problems in
Dry Scrubbing Applications." J. Air Waste Manage. Assoc. 40:112 (1990).
4.	K. E. Redinger (B & W Alliance Research Center) to G. H. Marchant, Jr.
(Southern Research Institute) Personal communication, January 23, 1990.
5.	R. P. Young, J. L. DuBard, and L. S. Hovis. "Resistivity of Fly Ash/Sorbent
Mixtures." In Proceedings: Seventh Symposium on the Transfer and
Utilization of Particulate Control Technology, Volume I, EPA-600/9-89-046a
(NTIS PB89-194039), May 1989.
6.	0. Molerns. "Theory of Yield of Cohesive Powders." Powder Technology
12:259, 1975.
7.	Richard Rhudy. "Chloride Addition Improves S02 and Particulate Removal in
SD/FGD Systems." ESC UPDATE 20, 1990.
62
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before complet
1 . REPORT NO.
.EPA-600/R-92-196
3.
4. TITLE AND SUBTITLE
Effects of E-SO,
Technology on ESP Performance
5. REPORT DATE
October 1992
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
G. H. Marchant, Jr., J. P. Gooch, and
M. G. Faulkner
8. PERFORMING ORGANIZATION REPORT NO.
SRI-ENV-91-89-6790
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Southern Research Institute
P. O. Box 55305
Birmingham, Alabama 35255-5305
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
CR814915-01-0, Task 6790
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 8/90 - 11/90
14. SPONSORING AGENCY CODE
EPA/600/13
15. supplementary notes ^EERL project officer L. S. Hovis is no longer with the Agency;
for details concerning this report, contact Samuel L. Rakes, Mail Drop 4, 919/541-
2828.
16. abstractrep0rt gives results of an evaluation of the E-SOx process at Ohio Edi-
son's Burger Station. Adequate sulfur dioxide (SC2) removal and acceptable particu-
late emission levels from the electrostatic precipitator (ESP) were the prime objec-
tives of this investigation. The report describes limited ESP performance testing
under both baseline and E-SOx conditions. The ESP data collected under E-SOx con-
ditions, which give the required 50% SG2 removal, show evidence of ESP perfor-
mance dominated by factors not represented in existing versions of ESP perfor-
mance models. Analyses of particle size fractions from impactor stages revealed
that the relative calcium content of the finer size fractions increased from inlet to
outlet. These analyses and other considerations indicate that the factors which do-
minate under the conditions tested are a combination of instantaneous reentrainment
of low resistivity ash/sorbent particles and deagglomeration of slurry residues with-
in the ESP. These observations may be important to other sorbent injection proces-
ses as well as to E-SOx. Improvement of the gas velocity and temperature distri-
butions at the ESP inlet improved the ESP performance, but performance was still
dominated by the reentrainment process and was therefore lower than mathematical
model predictions.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Pollution
Electrostatic Precipitators
Sulfur Dioxide
Particles
18. DISTRIBUTION STATEMEN
Release to Public
b.IDENTIFIERS/OPEN ENDED TERMS
Pollution Control
Stationary Sources
E-SOx Process
Particulate
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page)
Uncla ssified
c. COSATI Field/Group
13 B
131
07B
14G
21. NO. OF PAGES
71
22. PRICE
EPA Form 2220-1 (9-73)
i
-------