PB92-113117
Evaluation of Pilot ESP Performance with Elevated
Loadings from Sorbent Injection Processes
(U.S.) Environmental Protection Agency, Research Triangle Park, NC
4 Sep 91
J
-------
PB92-113117
EPA/600/D-91/244
EVALUATION OF PILOT ESP PERFORMANCE WITH
ELEVATED LOADINGS FROM SORBENT INJECTION PROCESSES
Charles B. Sedman,
Richard E. Valentine,
and
Norman Plaks
U.S. Environmental Protection Agency
Air and Enerqy Engineering Research Laboratory
Research Triangle Park, NC 27711
ABSTRACT
Measurements were made of calcium silicate sorbent (ADVACATE) injected into
a duct to simulate injection into the duct upstream of an electrostatic
precipitator (ESP). The concentration of ADVACATE sorbent submicron
particles (particles S 1 Jim} and projected ESP emissions tended to peak and
began to decrease when the overall particulate matter addition rate to the
gas stream exceeded 12 g/Nm . The submicron fly ash, subjected to the same
duct injection, increased linearly with increased injection rates from 3 to
24 g/Nm3, A possible explanation is in-duct agglomeration of fines by the
coarse particles, similar to observations reported on cyclone performance
evaluations. The duct, flue gas, and sorbent characteristics which affect
agglomeration tendencies probably play a major role in the observations
presented. The majority of the ADVACATE material settled out of the gas
stream. Measurements of the gas-suspended residual particulate matter were
used to model expected ESP performance. The results of the modeling are
encouraging, and suggest that collection of reacted ADVACATE sorbent in an
ESP is manageable. Future work will focus on identifying the process
parameters that will ensure reliable and reproducible control in an ESP
subjected to elevated sorbent loadings.
-------
EVALUATION OF PILOT ESP PERFORMANCE WITH
ELEVATED LOADINGS FROM SORBENT INJECTION PROCESSES
INTRODUCTION
As an integral part of EPA* s program to commercialize the ADV&CATE process
for sulfur dioxide (802) removal, recent research has focused on the effect
of elevated in-duct dust loadings Idue to advanced calcium silicate
{ADVACATE) sorbent] on electrostatic precipitator (ESP) operation. A brief
explanation of the ADVACATE process reveals the many parameters that may
affect sorbent properties and ESP operation [1J.
Figure 1 illustrates the ADVACATE Moist Dust Injection (MDI) process as
proposed for the Shawnee 10 MW_ facility of TVA. In this scheme ADVACATE
sorbent is prepared from the ESP first section dust capture where recycled
dust is split into several fractions: (1) approximately one-third is
slurried in a continuous stirred tank reactor (CSTR) with lime at elevated
temperature to promote the production of high surface area calcium
silicates,* (2) approximately two-thirds is directed to a mixer where it is
blended with product slurry to form a high-moisture, free-flowing powder
suitable for duct injection; and (3) the remainder is added to the bottoms
from the remaining ESP sections and is handled/disposed of as a solid waste.
ADVACATE sorbent is then injected into a flue gas at 150 °C and within half
a second has simultaneously flash dried from the initial 40 to 50% moisture
to about 5%, cooled the flue gas to about 65 °C, and removed a majority of
the S02 and other acid gases present.
This research effort currently focuses on:
• ESP performance of ADVACATE vs. fly ash and fly ash/lime
mixture under comparable conditions.
• ESP performance of ADVACATE at elevated grain loadings.
§ Effect of sorbent surface moisture on ESP performance.
• Effect of SO£ absorption upon ESP performance.
§ Effect of flue gas temperature and moisture upon ESP
performance.
§ Effect of ADVACATE process parameters (e.g., degree of
grinding, reaction tin" , storage, and mixing parameters}
upon ESP performance.
The program being developed to answer these questions has resulted in
substantial renovation of EPA pilot facilities and purchase of new process
equipment. Operatic.i ~»f these facilities in an integrated fashion will
likely not commence -intil mid-1992. In the interim, a shorter-term program
has been implemented in an attemot to determine which areas or parameters
have a major impact on ESP pert: -1- nee. The remainder of this paper is
devoted to results of this interim program, interpreting the results, and
projecting incremental emissions increases due to ADVACATE sorbent injection
on a commercial scale ESP application.
-------
EPA currently operates a. 3400 ro^/h dry flue gas cleaning pilot plant which
consists of a gas-fired heater, spray dryer, numerous dry sorbent
injection/sampling ports, and three particulate matter collectors—pulse-jet
and reverse-air fabric filters and a four-section ESP. It is planned to
install ADVACATE sorbent preparation facilities and modify recycle dust
capabilities to allow an integrated 0.7 MWe ADVACATE system to operate
continuously by mid-1992.
%
In the interim, a second four-section ESP has been modified to examine the
impact of ADVACATE sorbent upon ESP operation, and ADVACATE sorbent has been
prepared in batches for short-term ESP operations. This ESP has previously
been described in reports and technical papers and is briefly described
herein [2J. It is a four-section unit that can be operated in a large
number of configurations with regard to corona wires, wire-to-wire spacing,
type of electrode, and type of energization. The ESP, which had been well
characterized, had predictable performance that could be related to larger-
scale ESPs. Table 1 contains the relevant parameters for all ESP operations
data reported herein.
The ESP was modified externally to allow high dust load operation as shown
in Figure 2. As originally operated, particulate matter was aspirated into
the ESP using modified sand blast guns. The maximum inlet grain loading
with this arrangement was about 4.5 g/tn , which was insufficient for testing
ADVACATE. To the existing 1700 m3/h facility, a sorbent injection system
was added, including a blower, solids metering and feeding, and a 15 m long
10 cm l.D. duct. The added system was designed to deliver 340 nrYh of
ambient air, resulting in a gas velocity of 15 m/s. This stream entered the
ESP, in which the nominal flow velocity was 1.5 m/s, at the centerline,
approximately 3 m upstream of the first ESP section. The entering stream
first encountered a gas disperser cone followed by a coarse egg-crate set of
flow straightening vanes. Considerable effort was made to ensure a flat
profile, both vertical and horizontal, for gas velocity and temperature
entering the first ESP section. Table 2 shows the velocity and temperature
data through the ESP with the main blower and adjunct duct blower both in
operation with no dust injection.
Comparative tests were made with fly ash injected through the aspirators and
the high velocity duct. With the sections energized, the performance, with
aspirator injection, was similar to the historical data for the ESP. The
same loading of fly ash injected through the duct gave performance that
differed from historical data for the ESP, and from normal ESP operation. A
possible reason for the aberrant ESP performance might be the propelling of
particles down the centerline of the ESP by the high velocity stream.
In view of the unusual ESP performance, the test program was altered to:
{1} measure particle size distribution of dust exiting the ESP with all
fields off, (2) measure total particle mass concurrent with particle size,
and (3) model expected ESP performance based on these data. The remainder
of this paper details results of this work.
PARTICLE SIZE MEASUREMENTS
An initial series of tests were run varying the ADVACATE injection rate from
150 to 1000 g/min {equivalent to approximately 3.9 to 26.3 g/Nm3 inlet ESP
loading) to determine relative ESP effects. Sorbent was first prepared by
slurrying pre-ground fly ash from the Meredosia Station of Central Illinois
Public Service with Mississippi calcium hydroxide for 3 hours at 88 °C, then
mixing the slurry with recycle ADVACATE solids for a product having 40-45%
moisture content. The solids were then injected into the duct by a screw
feeder under negative pressure at rates corresponding to 17.8 to 117.8 g/Nm^
in the duct with 1000 ppmv SC^. The partially reacted solids then entered
the ESP and were diluted from 303.5 to 1370.9 Nnrvh by addition of heated
air from the ESP blower, resulting in ESP loadings (not including added
-------
sulfites and sulfates but deleting moisture evaporated) of 3.94 to 26.26
g/Nm . -{Although not measured directly, the maximum impact of SC^
absorption on the calculated ESP loading would be an 18% increase over the
figures in Table 3, Column 2.)
A series of measurements were made as shown in Table 3 where mass and
particle size measurements were made simultaneously on the ESP outlet with
the ESP turned off. The particle size measurements are in terms of the
geometric mass mean diameter (GMMD) and the geometric standard deviation
(GSD), The results in Table 3 show a definite trend in decreasing particle
size scatter or a normalizing of particle size with increasing particle
concentration. Further, a decided drop in particle size diameter is evident
when the feed rate increased from "750 to 1000 g/mirv, concurrent with a
decrease in effective ESP loading from 4.41 to 2,40 g/Nm3.
Another way of looking at the data is to focus only on the submicron
fraction of dust particles. The term submicron fraction denotes that
portion of the particulate matter £ 1.0 jun. From the outlet loading and
particle size data, the submicron fraction was calculated (Table 4), and
shows that the submicron fraction, too, tended to decrease as loading
increased.
A second series of tests were conducted to determine the effect of using
fresh sorbent instead of stored sorbent. For these tests Clinch River fly
ash was slurried with lime and cycled through a Union Process 15-S attritor
or vertical mill for 30 min. The slurry was then reacted for 2 h at 90 °C
and mixed with spent ADVACATE sorbent to yield a 47% moisture solid, which
was then fed immediately into the 15 m duct ESP entry. Results are shown in
Tables 5 and 6.
The contrast between results for old vs. fresh sorbent is significant.
Table 5 shows that fresh sorbent is considerably larger in size, with
substantially more scatter in particle size, as indicated by the larger
standard deviations. The fresh sorbent particles also tended to grow with
increased duct loading, and the ESP outlet loadings were generally much
lower than for old sorbent, indicative of more settling out of sorbent in
the ESP due to gravity. Table 6 shows that the submicron particle loading
is only half that of old sorbent (shown in Table 4).
The overall trends, however, in decreased fines as injection rate increases
are consistent in both cases.
An abbreviated test series of four runs were made with ADVACATE injection at
a point 2 m upstream of the ESP inlet. The ESP outlet loadings in Table 7
are similar to those of the long duct runs, but fines contents are lower, by
50 to 70%. These runs were made at relatively low feed rates and cannot be
compared to the higher rates where duct conditions tended to suppress fines.
As shown in Table 7, the particle size distributions of fresh and old
sorbent were dramatically different; however, the submicron loadings were
similar, again lending credence to the fines content's being a function of
handling rather than preparation.
Long duct runs were made with fly ash as shown in Table 8. The mass
particle loading and submicron particle loading both increased linearly with
injection rate; with submicron loading, an order of magnitude higher than
for ADVACATE (see Tables 4 and 6) at lower rates and two orders of magnitude
higher at the highest rate, corresponding to about 25 g/Nm3 (11 gr/SCF).
Figure 3 compares submicron particles entering the ESP as a function of
injection rate for the ADVACATE tests and fly ash only.
ESP MODELING
With the ESP deenergized, particle size measurements and loadings were
-------
measured in the sample port in the vertical outlet duct of the pilot ESP.
This location was used because measurements taken in the relatively small
cross section area are believed to be more representative than those made in
the ESP casing itself. Particle size was measured with MRI 1 impactors.
Particulate loading was measured using a modified EPA Method 5.
It was assumed that the particulate matter in the vertical outlet duct, with
the ESP deenergized, is the only particulate matter of concern to the ESP.
From Table 5, the majority of the particulate matter introduced into the ESP
falls out prior to reaching the vertical duct. The assumption was that the
particles that settle out would do so regardless of whether or not the ESP
was operating, and do not have to be included in the ESP performance
computations.
The ESP model used for the computer simulation is the new ESPVI Version 4.0,
which is in its final stage of development by Research Triangle Institute
and EPA. ESPVI Version 4.0 can compute V-I values for various electrode
shapes, and then use the computed electrical conditions to predict ESP
performance. Included among the attributes of this model is the ability to
determine the effects of space charge on the ESP, and especially on the
corona current. Determining the effects of space charge is essential when
analyzing ESPs operating with high grain loads.
ESPs of two sizes were modeled to illustrate a range of sizes that might be
encountered in application of the ADVACATE process. The geometries for the
two ESPs are given in Table 9. Also given are the gas conditions for
baseline (149 °C> and ADVACATE (65 °C) operation. The final data presented
in Table 9 are the non-ideal conditions (sneakage and rapping reentrainment)
that were used for the modeling.
The baseline was modeled assuming a fly ash particle size distribution
defined by a GMMD of 15 jun and a GSD of 3. The emissions computed by the
model, for the two ESPs, were placed on Figure 4, and connected by a line.
Because there are only two points connected the line by necessity was
straight, which is not meant to imply that there is a linear relationship
between SCA and emissions.
For the remaining plots the emissions for the two ESPs were modeled
separately for fly ash and various ADVACATE solids loadings at 65 °C.
Appropriate corrections were made to the SCA in going from 149 In 65 °C.
The plots that represent the fly ash plus ADVACATE were obtained by summing
the separate emissions for fly ash and ADVACATE. It should be noted that
obtaining an emission for a composite substance such as fly ash plus
ADVACATE, by summing separately computed emissions, is not a rigorous
approach. The ESP electrical conditions are affected by the space charge,
which, in turn, is established by the particle loading and size distribution
J3]. Thus the electrical conditions for the individual particulate
materials would not be the same as they would be for their sum. This tends
to introduce a slight error in the emissions. However, for the particle
concentrations for which the modeling is being done, the error is not
significant. A further point noted is that the SCA increases when cooling
from 149 to 65 "C. Therefore an SCA value on the baseline plot line is not
equivalent to the SCA for the plot lines.
The Fly Ash Plus ADVACATE Low plot on Figure 4 represents modeled ESP
emissions for pilot data from the 1000 g/min injection rate test {Table 3),
assuming that the particulate matter of concern to the ESP does not increase
with increasing ADVACATE feed rate. This assumption is based on
extrapolation of data in Figure 3 to approximately 80 g/Nm feed rates,
discussed previously. The Fly Ash Plus ADVACATE High plot assumes that the
particulate matter of concern to the ESP, contrary to the experimental
findings, increases linearly with additional ADVACATE feed, or increases
three-fold as feed rate is increased from 26 to 80 g/Nm5. The Fly Ash Plus
ADVACATE No Duct plot is for ADVACATE injection directly into the ESP (no
-------
duct residence time for agglomeration to occur) and is a worst-case
approach. This assumption is magnified by the limited "no-duct" data at low
ADVACATE injection rates, since the 6.57 g/Nm3 injection rate data were
multiplied by 12 to approximate the expected commercial rate of
approximately 80 g/Nm . For all the modeling simulations presented in
Figure 4, the particle size distribution measured and repeated in Tables 3-1
was used.
DISCUSSION OF RESULTS
The obvious conclusions from the data are that (1) ADVACATE material
generates less fines per unit mass than fly ash alone, and (2) some
phenomenon is suppressing ADVACATE fines as the feed rate is increased.
Since the ADVACATE material is prepared by solids' blending with slurry, the
sorbent itself has only very coarse agglomerates, typically hundreds of
micrometers in diameter. The questions are (I) do the agglomerates generate
fines during flash .rying, and (2) what is the effect of preparation,
handling, duct gc:..i *ry, and flue gas conditions on generated fines?
Although the present work does not subject the sorbent to exact field
conditions, an attempt has been made to bound the expected emissions created
by drying and handling. Our current conclusion is that the small {10 cm)
duct and ESP configuration is host to a new phenomenon which apparently
mitigates the majority of fines from in-duct drying, mechanical forces, and
perhaps fines inherent in the flue gas.
This phenomenon has been previously observed in mechanical collector
performance, Muschelknautz (4), Mothes and Loffler (5), and Hoffman, et al.
16] have reported the apparent improvement in cyclone performance with
increased dust loading. Muschelknautz introduced the critical loading
concept which proposes that a moving gas stream contains sufficient
turbulent energy to support only a critical load of dust, and that any
amount of dust exceeding this critical value will spontaneously fall out of
the gas stream unclassified, or without respect to particle size, Mothes
and Loffler explained the same observation as due to sweeping of the smaller
particles to the cyclone wall by larger particles. Hoffman, et al, noted
that the efficiency improvement of cyclones also increased with decreasing
gas velocity; hence, their conclusions were that (1) the improvements were
unrelated to the cyclone parameters, and (2) results were consistent with
agglomeration at the cyclone inlet but an additional sweeping effect in the
cyclone may explain the difference between model predictions and actual
observations based on agglomeration alone.
Considering the large quantities of ADVACATE material that will be injected
into the gas stream, the modelling projections are encouraging. ESP
operation is extremely sensitive to changes in inlet loading and particle
size distribution 13] . Small changes to the particle size distribution of
ADVACATE, again considering the large quantities being injected, provide a
very high potential for profound changes in ESP performance. Variation in
ADVACATE particle characteristics, from batch to batch, has been observed.
An important component of future work on ADVACATE will be to determine the
effects of improvements in S02 capture and modifications to the process on
proper ESP operation.
We cannot overemphasize the limited nature of data presented in Tables 3-8,
however. Duct dimensions, gas velocities, moisture, and a host of other
variables must be examined in future work before the practical significance
of these observations is known. However, the conditions reported suggest a
marked improvement in expected ESP performance at high inlet loadings that
can be explained only by the sequestering and removal of Al VACATE fines in
the duct and ESP transition section.
Planned work in 1992 will expand these observations to higher and lower
loadings, gas velocities, temperatures, moisture, and duct dimensions. In
-------
addition to the question of general applicability, specific questions to be
resolved are:
(1) Will increased sorbent loading continue to show improved
sequestering of fines? Is the upper limit strictly one of
economics?
(2) Will the increased sorbent load remove furnace-borne aerosols
similarly to sorbent fines? Can overall {fly ash + sorbent)
emissions be reduced to levels lower than for fly ash alone?
(3) Will the current projection trends of ESP performance
continue to hold? Will process development modifications
of ADVACATE affect ESP performance?
(4) Can ESP performance be improved by simple recycle of coarse ash
from the first section's ESP bottoms? What will be the effects
on SO2 capture?
(5) How important is addition of hygroscopic materials (such as
ADVACATE sorbent) in the above observations?
(6) What process parameters are important in producing consistent
ADVACATE material?
CONCLUSIONS
The large quantities of sorbent injected into the ESP by the ADVACATE
process provide a potential for ESP problems. However it appears that
addition of increasing loadings of hygroscopic sorbent material such as
calcium silicates (ADVACATE sorbent) tends to suppress any fines generated
by injection, drying, or mechanical process on the sorbent. The phenomenon
is best explained by in-duct agglomeration of fines into larger particles.
The measured quantity and size distribution of the ADVACATE particles that
remain suspended in the gas stream, and would consequently be of importance
to ESP operation, were modeled to provide a prediction of ESP performance.
The results of the modeling were encouraging and suggest that particulate
matter control for an ADVACATE retrofitted process is manageable.
-------
REFERENCES
Hall, B.W., Singer, C., Jozewicz, W.f Sedman, C.B., and Maxwell, M.A.,
Current status of ADVACATE process for flue gas desulfuri^^tion,
presented at the 1991 AWMA Annual Meeting, Vancouver, B.C., Canada,
June 20, 1991.
Lawless, P.A., Daniel, B.E., and Ramsey, G.H., Characterization of the
EPA/IERL-RTP pilot-scale precipitator, EPA-600/7-79-052 (NTIS PB
292820), February 1979.
Plaks, N., The effects on electrostatic precipitation of changes
in grain loading, size distribution, resistivity, and
temperaturer presented at the Ninth Particulate Symposium,
Williamsburg, VA, October 15-18, 1991.
Muschelknautz, E., Die berechnung von zyklonabscheiden fur gase,
Chemie-Ing-Techn, jT4 No.5, 1970.
Mothes, H. and Loffler, F., "Motion and deposition of particles in
cyclones," Ger. Chetn. Eng., 8., (1985).
Hoffman, A.C.r Arends, H., and Sie, H., An experimental investigation
elucidating the nature of the effects of solids loading on cyclone
performance, Filtration and Separation, pp. 188-193, May-June 1991.
-------
TO M/XEfl TO MIXER
STACK
DISPOSAL
Figure 1. ADVACATE/MDI Process
-------
N
_J X
I
LU
o
S
H2O
O
O
O
so
Hf
o
SAMPLE PORTS
\
JO.
L4-
o
o
o
XL
TO SCRUBBER
MEZZANINE /
SAMPLE PORTS
FIELD 1
GAS BURNERS
O
O
O
FIELD 2
XL
O
O
O
FIELDS
O
O
o
FIELD 4
O
O
O
od
'77
Figure 2. EPA ESP Facility, Research Triangle Park, NC
-------
CO
E
0
O
_j
LU
0
DC
Q_
O
CC
O
2
CO
CO
0.35
0.3
0.25
0.2
0.15
0.1
0.05
n
" >
_
;
-
»,•"*
i A
i ,•'
-
* .*
^
.•
^ _^
' 4 e- _^fl-
a". "O -i , , "rtJ
10 20
INJECTION RATE, g/Nm3
ADVACATE, fresh
- ADVACATE, old
• FLY ASH
30
Figure 3. Submicron Particle Mass Vs. Injection Rate
-------
^
s
CO"
o
CO
CO
Ul
1UU
90
80
70
60
50
40
30
20
10
n
~" V
\
\
\
\
\
\
\
\
\
\
\
\
\
'-•-,.. \
" "^^^•^•••...^
i t i "7*^* i ,
20
40 60
SCA, m /m /s
60
FLY ASH
FLY ASH+ADVACATE LOW
FLY ASH+ADVACATE HIGH
FLY ASH+ADV NO DUCT
100
Figure 4. Projected ESP Performance-ADVAGATE Process Impact
-------
Table 1
PILOT ESP DESCRIPTION
No. of Fields
Plates per Field
Plate Spacing, cm
Effective Plate Area, m2
Hires per Field
Hire Diameter, cm
Hire Spacing, cm
SCA , m2/m3/s
At 23 cm plate spacing, 1.5 m/s gas velocity
4
2
22.9
1.5
5
0,32
22.9
28
Table 2
PILOT ESP GAS VELOCITY AND TEMPERATURE PROFILE AFTER MODIFICATION
Top
T, °C
v, m/s
Middle
T, °C
v, m/s
Bottom
T, °C
v, m/s
154.4
1.64
154.4
1.89
153.5
1.12
Between
Fields
1 & 2
151.7
1.48
148.3
1.74
152.8
1.26
Between
Fields
263
149.4
1.61
150.6
1.74
149.4
1.35
Between
Fields
3 & 4
148.9
1.65
147.8
1.61
147.8
1.20
ESP
Outlet
147,8
1.28
148.9
1.93
137.8
1.28
13
-------
Table 3
PARTICLE SIZE ANALYSES WITH ESP OFF
Outlet
Duct Feed ESP Inlet, q/Nm-3 ESP Outlet,
Rate, q/min Calculated Measured ESP Off, q/NmJ
150 3.94 3.69 2.04
500 13.13 8.94 3.80
750 19.70 12.14 4,41
1000 26.26 14.83 2.40
Table 4
Particle Size
GMMD, Urn GSD
21.8 4.00
22.3 3.67
20.0 3.31
13.4 3,03
SUBMICRON PARTICLE BEHAVIOR VS. ADVACATE INJECTION RATE
Outlet
ADVACATE Injection Particle Loading
Rate, a/min ESP Off, a/NmJ
150 2.04
500 3.80
750 4.41
1000 2.40
Table 5
FRESH SORBENT PARTICLE SIZE
Duct Feed ESP Inlet, ESP Outlet,
Rate, a/min Calculated, q/Nnr ESP Off, q/NmJ
150 3.94 0.878
300 7.88 1.060
750 19.70 0.534
1000 26.26 0.355
Submicron
Particle Loading
ESP Off, q/Nm3
0.0233
0.0315
0.0297
0.0240
Outlet
Particle Size
GMMD, urn GSD
65.6 6.77
63.1 6.10
81.4 12.15
112.3 15.35
14
-------
Table 6
SDBMICRON PARTICLE LOADING WITH FRESH SORBENT
ADVACATE Injection
Rate, q/min
150
300
750
1000
Outlet
Particle Loading
ESP Off,Q/Nmj
0,878
1.060
0.534
0.355
Submicron
Particle Loading
ESP Off. g/Nm3
0.014
0,013
0.021
0.012
Table 7
SHORT DUCT INJECTION OF ADVACATE AT 250 q/min
ESP Outlet
Particle Size,
Particle Loading
Sorbent
Fresh
Old
GMMD, Urn
54.5
52.3
100.7
134.6
GSD
5.54
5.80
6.32
7.33
ESP Off, q/Nm-»
0.463
0.318
0.426
0.338
Submicron
Particle Loading
ESP Off,
0.0057
0.0053
0.0046
0.0039
Table 8
SUBMICRON PARTICLE BEHAVIOR VS. FLY ASH INJECTION RATE
Fly Ash
Injection Rate
g/ndn
Outlet Particle
75 (3.28)
150 (6.57)
300 (13.13)
550 (24.07)
Outlet
Particle Size
Submicron
Particle Loading
o/NmJ
1.607
3.147
6.883
8.445
GMMD, ton
8.43
10.23
13.23
9.81
GSD
2.79
3.34
3.91
3.69
ESP Off, a/Nm3
0.0306
0.0881
0.2065
0.3378
15
-------
Table 9
ESP MODELED
Small Large
ESPGeometry
Sections 3 5
Section Length, m 2.29 2."74
Height, m 9.14 9.14
Section Area, mz 2070 2484
Plate Spacing, m 0.229 0.229
Hire Spacing, m 0.229 0.229
Hire Diameter, mm 2.8 2.8
Wires per Lane 10 10
Lanes per Section 50 50
Gas Conditions
High Temperature Case
Temperature, °C 149 149
Volumetric Flow, actual nr/s 159.3 159.3
Velocity, m/s 1.52 1.52
SCA, mVm3/s 39.0 78.0
Low Temperature Case
Temperature, °C 65 65
Volumetric Flow, actual m3/s 129.8 129.8
Velocity,,m/s 1.24 1.24
SCA, mVnrVs 47.9 95.8
Non-ideal Conditions
[same for all)
Sneakage 0.05
Rapping Fraction 0.06
Velocity Standard Deviation 0.15
16
-------
AEERL-P-855
TECHNICAL REPORT DATA
(Pleatc read liulmctions on the rtrcttt before tomplel'
1. RE'ORT NO.
EPA/600/D-91/Z44
PB92-113117
4. TITLE AND SUBTITLE
Evaluation of Pilot ESP Performance with Elevated
Loadings from Sorbent Injection Processes
S. REPORT DATE
G. PERFORMING ORGANIZATION CODE
1, AUTHORtSl
Charles B. Sedman. Richard E. Valentine, and
Norman Plaks
13. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
1O, PROGRAM ELEMENT NO.
See Block 12
11. CONTRACT/GRANT NO.
NA (Inhouse)
J2. SPONSORING AGENCY NAME A«D 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
Published paper; 4~9/91
14, SPONSORING AGENCY CODE
EPA/600/13
is.SUPPLEMENTARY NOTES AEERL project officer is Charles B. Sedman, Mail Drop 4, 919/
541-7700. Presented at 9th Symposium on Participate Control, Williamsburg, VA,
10/15-19/91.
16.ABSTRACT^pj^g paper gives results of an evaluation of pilot electrostatic pi-ecipitalor
(ESP) performance with elevated loadings from the advanced silicate (ADVACATE)
sorbent injection process. Measurements were made of a calcium silicate sorbent
injected into a duct upstream of an ESP. The concentration of ADVACATE sorbent
submicron particles (=/< 1 micrometer) and projected ESP emissions tended to peak
and began to decrease when the overall particulate matter addition rale to the gas
stream approached and then exceeded 12 g/Nm3. The submicron fly ash, subjected
to the same duct injection, increased linearly with increased injection rates from 3
to 24 g/Nm3. A possible explanation is in-duct agglomeration of fines by the coarse
particles, similar to observations reported on cyclone performance evaluations.
The duct, flue gas, and sorbent characteristics that affect agglomeration tendencies
probably play a major role in the observations presented. Most of the ADVACATE
material settled out of the gas stream. Measurements of the gas-suspended residual
particulate matter were used to model expected ^SP performance. The encouraging
results of the modeling suggest that collection of reacted ADVACATE sorbent in a
ESP is manageable, c:
17.
KEY WORDS AUO DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENtlFIERS/OPEN ENDED TERMS
Pollution
Sulfur Dioxide
Sorbents
Calcium Silicates
Flue Gases
Particles
Mathematical Models
Electrostatic Precipi-
tators
Pollution Control
Stationary Sources
Sorbenl Injection
ADVACATE
Particulate
c. COSATI Fkld/Gioup
13 B
07 B
11G
21B
14G
12 A
13J
19 DISIt IBUTtON SI A1LMFNT
St 10 Public-
is SC CORtTV CLARE f/ftl(
I P.Cl.iSSl
71 NO Ol
w
?O SLCumi V CLASS
nrl- .--".sifiPif
------- |