United States
Environment*! Protection
Industrial Environmental Research
Laboratory
Cincinnati OH 45268
EPA-600/2-80-141
June 1980
Research and Development
Evaluation of
Technology for
Control of Arsenic
Emissions at the
Campbell Red Lake
Gold Smelter
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9 Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EVALUATION OF TECHNOLOGY FOR
CONTROL OF ARSENIC EMISSIONS
AT THE CAMPBELL RED LAKE GOLD SMELTER
by
G.H. Marchant and R.L. Meek
Southern Research Institute
Birmingham, Alabama 35205
Grant No. R 8Q4955
Project Officer
John 0. Burckle
Industrial Pollution Control Division
Industrial Environmental Research Laboratory
Cincinnati, Ohio 45268
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environ-
mental Research Laboratory (Cincinnati), U.S. Environmental
Protection Agency, and approved for publication. Approval does
not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute
endorsement or recommendation for use.
11
-------
FOREWORD
When energy and material resources are extracted, pro-
cessed, converted, and used, the related pollutional impacts
on our environment and even on our health often require that
new and increasingly more efficient pollution control methods
be used. The Industrial Environmental Research Laboratory-
Cincinnati (lERL-Ci) assists in developing and demonstrating
new and improved methodologies that will meet these needs both
efficiently and economically.
This report presents the results of a field test and
evaluation of a highly effective system employed on a non-
ferrous metals application for the control of arsenic trioxide
air emissions generated in the roasting of arsenic sulfide
bearing gold ores. The Nonferrous Metals and Minerals Branch
of the Energy Pollution Control Division should be contacted
for further information.
David G. Stephan
Director
Industrial Environmental Research Laboratory
Cincinnati
111
-------
ABSTRACT
Since arsenic is a significant component of ores processed
in the nonferrous metals industries, it is of interest to char-
acterize and evaluate control strategies which have demonstrated
the potential for lowering the emission rates of arsenic and
other hazardous effluents from smelter operations. The Campbell
Red Lake Mines Gold Smelter at Balmerton, Ontario, Canada, has
developed and implemented a successful control strategy for
arsenic emissions from a nonferrous smelting operation. The Red
Lake smelter uses cyclones and a hot electrostatic precipitator
to recover metal values from roaster dusts with subsequent air
quenching to condense (or desublime) arsenic trioxide which is
recovered in a low-temperature baghouse. A test program was con-
ducted at Red Lake to characterize the control systems and to
evaluate the potential for transferring the technology to non-
ferrous smelting operations in the United States.
This report presents the results of the test program con-
ducted at the Red Lake gold smelter during the period September
18-28, 1978. The overall efficiency for the control of particu-
late emissions using a combination of a hot electrostatic pre-
cipitator, an air quench in a mixer-cooler, and a cold baghouse
exceeded 99.9%. Collection of particulate arsenic in the bag-
house was greater than 99.95%; however, overall arsenic collec-
tion efficiency in the baghouse was slightly less due to passage
of As203 vapors. Total arsenic emissions from the system were
only 11 mg/dscm (6.8 x 10~7 Ib/dscf or 0.20 Kg/hr (0.44 Ib/hr).
IV
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables viii
Acknowledgements ix
1. Introduction 1
2. Conclusions 2
3. Process Description 3
4. Sampling Program 9
5. Particulate Measurements 13
6. Gas Sampling and Analysis 25
7. Other Analyses 32
8. Resistivity Analyses 41
9. Technology Transfer 46
References 49
Appendix 50
-------
FIGURES
Number Page
1 Simplified flow diagram 4
2 Roasting and gas cleaning process 5
3 Gas mixer-cooler 8
4 Sampling locations 10
5 Average ESP inlet cumulative size distribution . 14
6 Average ESP inlet cumulative percent versus
particle diameter 15
7 Average ESP outlet cumulative size distribution. 16
8 Average ESP outlet cumulative percent versus
particle diameter 18
9 Average baghouse inlet cumulative size distri-
bution 19
10 Average baghouse inlet cumulative percent versus
particle diameter 20
11 Average baghouse outlet cumulative size distri-
bution 21
12 Average baghouse outlet cumulative percent versus
particle diameter 22
13 Baghouse particulate collection efficiency ... 23
14 Trace element gas sampling train 26
15 SEM photograph of ESP solids 35
16 SEM photograph of Baghouse solids 35
17 Energy dispersive X-ray analysis of ESP solids . 36
VI
-------
FIGURES (continued)
Number
18 Energy dispersive X-ray analysis of baghouse
solids 37
19 X-ray powder diffraction pattern for EPS solids . 38
20 X-ray powder diffraction pattern for baghouse
solids. 39
21 Typical secondary voltage-current curve 42
22 Clean and dirty plate voltage-current curves
from in situ point-to-plane resistivity probe . 43
A-l Modified Brink BMS-11 cascade impactor 51
A-2 Andersen Mark III cascade impactor 52
A-3 University of Washington Mark III cascade
impactor 53
A-4 DM/DlogD versus particle diameter for ESP inlet
data 54
A-5 DM/DlogD versus particle diameter for ESP putlet
data '. . . . 55
A-6 DM/DlogD versus particle diameter for baghouse
inlet data 56
A-7 DM/DlogD versus particle diameter for baghouse
outlet data 57
VII
-------
TABLES
Number Page
1 Electrostatic Precipitator Descriptive
2
3
4
5
6
7
8
9
10
11
Parameters
Baghouse Descriptive Parameters
Plant Operating Data
Impactor Mass Loadings
Gas Stream Measurements and Trace Elements
Distribution
Gas Flow Data
Solid and Vapor Distribution of As
Trace Element Analyses
As2O3 Partial Pressures
SSMS Analyses
In Situ Resistivity - Baghouse Inlet
7
7
12
24
28
29
30
33
33
40
45
Vlll
-------
ACKNOWLEDGEMENTS
This test program was sponsored by the Metals and Inorganic
Chemicals Branch of EPA's Industrial Environmental Research
Laboratory, Cincinnati, Ohio, John 0. Burckle, Project Officer.
Assistance in arranging the test was provided by Environment
Canada's Air Pollution Control Directorate, Ottawa, Ontario,
W.A. Lemmon, Chief, Mining, Mineral and Metallurgical Division.
The gas sampling and analysis were conducted by Radian
Corporation, Austin, Texas under the direction of Dr. John C.
Terry.
We would like to give special thanks to Stewart Reid,
General Manager, Ken Dickson, and Scott Roberts, Campbell Red
Lake Mines Ltd., Balmerton, Ontario, Canada who were very coop-
erative in providing access to the plant and in making modifica-
tions required to conduct the test program.
IX
-------
SECTION 1
INTRODUCTION
The Campbell Red Lake Mines operation at Balmerton,
Ontario, Canada has developed and implemented a successful con-
trol strategy for arsenic emissions from a nonferrous smelter.
In their process, a gold-bearing arsenopyrite ore is roasted in
a two-stage fluid-bed roaster to eliminate arsenic, antimony,
and sulfur. The hot roaster calcines are recovered in a series
of cyclones followed by a hot electrostatic precipitator, with
the precipitator and cyclone catches being returned to the gold
recovery system. Hot gases from the precipitator are quenched
with ambient air to condense (or desublime) arsenic trioxide
which is then collected in cold bag filters before the particu-
late-free gas is exhausted to the atmosphere. Overall collection
efficiency of the system exceeds 99.9% for all particulate and
99.8% for arsenic trioxide. AS2O3 emissions from the system are
almost exclusively in the gaseous state and amount to only about
eleven milligrams per dry standard cubic meter of exhaust gas.
Since arsenic is a significant component of many ores
processed in this country by the nonferrous metals industry, the
characterization and evaluation of control strategies which have
demonstrated capabilites for reducing the emission rates of
arsenic and other hazardous emissions from smelters is of inter-
est. The test program at Red Lake was conducted to evaluate the
performance of their "hot ESP/cold baghouse" system and to
assess the potential for transferring the technology to non-
ferrous smelting operations in the United States.
-------
SECTION 2
CONCLUSIONS
Field tests conducted at the Campbell Red Lake Mines gold
smelter at Balmerton, Ontario, Canada have demonstrated and con-
firmed that the use of a hot electrostatic precipitator (operated
above the dewpoint of As203) followed by quenching with ambient
air to condense (or desublime) As203 with subsequent collection
of condensed particulate in a baghouse operated at a low tempera-
ture (at which the As203 vapor pressure is very low) is poten-
tially viable technology for control of arsenic emissions from
nonferrous smelters.
Operation of the hot electrostatic precipitator at tempera-
tures above the dewpoint of As203 provides a mechanism for effi-
cient recovery of metal values for recycle to the smelting opera-
tion without significant condensation of arsenic. Subsequent
quench cooling with ambient air effectively condenses the As203
to fine particulate which can be collected in a baghouse with
low emissions of arsenic if the cooler-baghouse system is at a
low temperature (circa 100-110°C).
The particulate collection efficiency of the hot ESP in
use at Red Lake was about 98.3%; the particulate collection
efficiency of the cold baghouse was greater than 99.9%; and the
overall collection efficiency for arsenic was about 99.8%, with
the major part of the emission being as volatile As2O3. The
arsenic emission was a direct function of the vapor pressure of
As2O3 at the baghouse outlet.
Based on the tests conducted at Red Lake, it was concluded
that the system is a potential candidate for technology transfer
to other nonferrous smelters that process arsenious ores. How-
ever, the technology is probably not applicable to smelting
processes where the smelter gases are either utilized for pro-
duction of sulfuric acid or water scrubbed before being released
to the atmosphere.
A preliminary evaluation of potential technology transfer,
to a particular smelter stream can be made with only a limited
amount of data including the arsenic input and distribution
within a system, the gas flows and temperatures in existing (or
proposed) control devices, the subsequent treatment and disposi-
tion of the gas, and the vapor pressure of As2O3.
-------
SECTION 3
PROCESS DESCRIPTION
A simplified flowsheet of the overall operation of the
Campbell Red Lake Mines Limited installation at Balmerton,
Ontario, Canada, is shown in Figure 1. Ore from an underground
mine is subjected to crushing, grinding, ball milling, flotation,
and tabling to obtain three cuts for further processing. The
first cut (tabling overflow) is amalgamated directly for recovery
of gold, the second (flotation tails and underflow) is relatively
low in arsenic and sulfur and can be treated by cyanidation to
liberate the gold, but the third cut requires roasting to produce
a leachable calcine and eliminate arsenic, antimony, and sulfur
which interfere with cyanidation. Two-stage fluid-bed roasting
effectively removes the arsenic, antimony, and sulfur from the
concentrate. The calcines from the roasters, cyclones, and hot
electrostatic precipitator can then be treated by cyanidation to
recover the gold. The hot gaseous effluent from the electrostatic
precipitator is quenched with cold air to condense AsaOs which is
recovered in a cold baghouse.
A more detailed flow diagram of the roasting, gas cleaning,
and arsenic recovery system at Campbell Red Lake Mines is shown
in Figure 2. The concentrate (nominally 9% As, 13% S, 23% Fe-
design basis) is fed to the first fluid-bed roaster as a slurry
containing about 80% solids. The design capacity of the system
is 2500 Kg/hr (2.7 tons/hr)(dry basis). Most of the primary
roaster air is fed to the first roaster which is operated at
about 540-565°C (1000-1050°F) where most of the As, Sb, and S
are oxidized and volatilized. The overflow and underflow from
the first roaster are further treated in the second fluid-bed
roaster at at about 500-525°C (925-975°F) to assure maximum
elimination of As, Sb, and S from the calcines. The gases from
the second roaster are fed through 2 parallel sets of 3-stage
cyclones operated at about 400°C (750°F), and the calcine catch
from the cyclones is combined with the roaster bed underflow for
return to a cyanidation circuit for recovery of gold.
The gases from the two banks of cyclones are combined and
diluted with air to reduce the gas temperature to 370°C (700°F)
before being fed to the electrostatic precipitator which consists
of 2 units in parallel with 2 chambers in each unit. The pre-
cipitator is double-walled, and electrically-heated air is
-------
FROM MINE
COARSE ORE
STORAGE
I
CRUSHfNG
CIRCUIT
FINE ORE
STORAGE
GRINDING
CIRCUIT
FLOTATION
FLOTATION TAILS
TO CYANIDATION
ROASTER
AIR
TABLING
FLUID BED
ROASTERS
UNDERFLOW TO
AMALGAMATION
COOLING
AIR
COOLING
AIR
CYCLONES
HOT
ELECTROSTATIC
PRECIPITATOR
DILUTION
COOLER
COLD
BAGHOUSE
T
,A»2O3 TO UNDERGROUND
STORAGE
GASES
TO STACK
Figure 1. Simplified flow diagram.
-------
CYCLONES
400°C
(750°F)
AIR 7.4 m3/min
(260 SCFM)
ELECTROSTATIC
PRECIPITATOR
(2 IN PARALLEL)
370°C (700°F)
COOLING AIR
300 rr>3/m
(10,650 SCFM)
CONCENTRATES
2400 Kg/hr
(5400 Ib/hr)
SOLIDS AS
80%SLURRY
MIXING CHAMBER
107°C(225°F)
AIR
50 m-Vmin
(1730 SCFM)
AIR
26 rr>3/min
(940 SCFM)
CALCINES
TOCYANIDATION
2157 Kg/hr STACK GAS
(4750 Ib/hr) 395 mj/min
.—.(13,930 SCFM)
STACK
BURNER
93°C
(200°F)
FAN
BAGHOUSE
( 2 BANKS OF 2 )
UNDERGROUND
STORAGE
253 Kg/hr
(560 Ib/hr)
STACK
Figure 2. Roasting and Gas Cleaning Process (Design Flows).
-------
circulated through the outer shell to maintain the wall tempera-
ture above the condensation point of As203. As will be shown by
the data in this report, very little sublimed arsenic (primarily
As203) is condensed in the precipitator and only particulate
arsenic (perhaps as As2S2 carryover, a combined complex, or as
As2Os) is removed in the ESP with the other entrained particulate
from the hot cyclones. The hoppers from the electrostatic pre-
cipitator are emptied periodically, and the catch is combined
with the roaster and cyclone calcines for return to the gold
recovery system.
The hot precipitator off-gases are then mixed with ambient
air to reduce the temperature to about 107°C (225°F) to condense
the sublimed arsenic (As203). This particulate is caught in a
4-chamber baghouse before the gases are sent to a double-walled
stack. The arsenic trioxide dust from the baghouse is removed
periodically and conveyed to underground storage. The burner
shown in Figure 2 was provided to reheat the gases to obtain
plume rise; however, the burner has not been needed in normal
operation.
Descriptive parameters for the precipitator are given in
Table 1. The precipitator is a double-walled unit, and elec-
trically-heated air is circulated in the shell to maintain tem-
perature and avoid wall condensation. One transformer rectifier
powers all four chambers of the precipitator. During the test
period, the roaster system was operated at around 90% of design
capacity so the actual SCA during the test was about 60 m2/(m3/
sac) or about 300 ft2/1000 acfm.
Parameters for the baghouse are shown in Table 2. The
baghouse is insulated, and the stack is also a double-walled unit
to avoid wall condensation.
A key feature of the Red Lake System is the mixing chamber
used for cooling the hot gases from the precipitator to condense
or desublime the AS20s. A simplified diagram of this mixer-
cooler is shown in Figure 3. Ambient air is fed into the center
of the mixer, and the hot gases from the precipitator are fed
tangentially to create a swirling mixing action. Ideally, con-
densation (or desublimation) of As2Oa occurs only at the juncture
of the hot and cold gas streams. Wall condensation is avoided
by keeping the hot gases on the outer periphery and by heat
tracing and insulating the walls. Air flow to the mixer-cooler
is controlled to maintain a baghouse temperature below 110°C
(225°F).
The overall roasting and gas treatment system at Red Lake
was designed by Hatch Associates Ltd., Toronto, Canada1'2,
the hot-side precipitator was designed by Joy Manufacturing Co.
(Canada) Ltd., and the baghouse was designed by Wheelabrator
Corporation of Canada Ltd.
-------
TABLE 1. ELECTROSTATIC PRECIPITATOR
DESCRIPTIVE PARAMETERS
Item
Collection electrode area
(4 chambers-2 sets parallel)
No. of fields per chamber
Collection electrode spacing
Collection electrode dimensions
Corona electrode dimensions
(round wire)
Wire spacing
Number of gas passages
(per chamber)
Gas passage length
Volume flow rate (design)
Design temperature
Design efficiency
Design specific collection
area (SCA)
Measured efficiency (particulate)
Arsenic capture (total)
Metric
187 . 3m2
1
25.4 cm
1.83 x 1.83m
0.268 cm
22.86 cm
7
1.83 m
3.44 m3/sec
371°C
98%
54.37 m2/(m3/sec
98.3%
15%
English
2016 ft2
1
10 in.
6 x 6 ft
0.1055 in.
9 in.
7
6 ft
7300 acfm
700°F
98%
276.2 ft2/
1000 acfm
98.3%
15%
TABLE 2. BAGHOUSE DESCRIPTIVE PARAMETERS
No. of compartments
No. of bags (per compartment)
Bag material
Bag diameter
Bag length
Air-to-cloth ratio
(actual)
Measured efficiency (particulate)
Arsenic capture (total)
230
(10 rows of 23)
Draylon T (acrylic)
10.16 cm (4 in.)
2.44 m (8 ft)
9.36 x 10~3 '(m3/sec)/m2
(1.84 ft3/min/ft2)
99.9+%
99.8%
-------
TO BAGHOUSE
HOT GAS
AMBIENT AIR
HOT GAS
Figure 3. Gas mixer-cooler. *
* U.S. Patent 4,126,425 and Canadian Patent 993,368 assigned to Hatch
Associates Ltd., Toronto, Canada.
-------
SECTION 4
SAMPLING PROGRAM
During the period of September 18-28, 1978, personnel from
Southern Research Institute and Radian Corporation carried out a
test program in the Campbell Red Lake Mines Ltd. gold smelter
at Balmerton, Ontario, Canada.
Radian Corporation personnel operated an integral sampling
system designed to obtain particulate and gas samples for
elemental analysis, and a gas sampling system for determining
concentrations of sulfur oxides.
Southern Research Institute personnel operated cascade
impactors for determination of the particle size distribution of
the particulate in each gas stream, and a total particulate
sampling system for determination of total particulate loading
in each stream. SRI personnel also monitored the in situ resis-
tivity of the particulate entering the baghouse^ (for possible
modeling of a precipitator which could replace the baghouse).
The primary purpose of the test program was to evaluate the
performance of the gas treatment system which consists of a hot
electrostatic precipitator for recovery of metal values, a
mixer-cooler for condensation (desublimation) of As20sand Sb203,
and a cold baghouse for removal of fine condensation particulate.
A simplified sketch indicating sampling points is shown in
Figure 4.
Sample Point 1 at the precipitator inlet was under positive
pressure, and some difficulties were experienced in isokinetic
sampling. The duct was vertical with an upward gas flow and an
internal diameter of 0.43 m (17 in.).
Sample Point 2, which was one leg of the precipitator out-
let, was highly inaccessible and was not regarded as a satis-
factory sampling point for several reasons: (1) Sampling of only
one leg of the precipitator required an assumption that both
sides in the parallel configuration were operating identically,
(2) Possible dust accumulation in the bottom of the duct could
affect flow measurements, and (3) The location was too close to
converging and diverging ducts for uniform particulate sampling.
-------
ROASTERS
TEMPERING
AIR
ELECTROSTATIC
f) PRECIPITATOR
AIR
REHEAT
BURNER
BAG FILTER
V
FAN
A^Os
(TO STORAGE)
Figure 4. Sampling locations.
-------
Access to this sample location was also very difficult due to
physical restraints within the building housing the equipment.
However, since the precipitator and mixer-cooler are very cdosely
coupled, samples were taken from this 0.25 m (10 in.) square
duct since this was the only possible sampling point for esti-
mating particulate at the ESP outlet/mixer-cooler inlet.
Sample Point 3 at the outlet of the mixer-cooler provided
data for the baghouse inlet. Gas flow was vertically downward
through the 0.71 m (28-in.) duct.
Sample Point 4 in the stack provided data for the baghouse
outlet. The internal diameter of the stack was 0.76 m (30-in.).
Insofar as possible, all samples were taken isokinetically
to assure that representative measurements were made. There were
some variations in operation during the test period that affected
the sampling program; however, adequate data were obtained to
evaluate the overall performance of the electrostatic precipita-
tor and the baghouse.
A summary of plant operating data for the test period is
shown in Table 3. These data show that the plant was in reas-
onably stable operation during the tests even though there was
one operational shut-down during the test period, and the pre-
cipitator had an electrical malfunction near the end of the test
period. (Data omitted).
11
-------
TABLE 3. PLANT OPERATING DATA
Freeboard Temperatures
Reactor No. 1, °C
OF
Reactor No. 2, °C
°F
Air Flow-to-Wind Box
Reactor No. 1, m3/min (cfm)
m/sec
Reactor No. 2, m3/min (cfm)
m/sec
ESP and Baghouse Temperatures
ESP Inlet, °C
°F
ESP Outlet, °C
op
Baghouse Inlet, °C
Ave
549
1021
511
951
34 (1200)
34.0
26 (920)
26.1
392
737
308
586
118
245
Baghouse Outlet, °C 111
°F 232
Stack, °C 102
°F 216
Precipitator*
AC Amps 10.1
AC Volts 418
DC Milliamps 40
DC Volts x 1000 57.3
* Excluding data at end of test period.
532-558
990-1037
503-508
938-946
33.7-34.6 (1190-1220)
33.7-34.6
24.4-28.3 (860-1000)
24.4-28.3
378-409
713-768
277-327
531-620
109-121
229-250
109-113
228-236
100-104
212-219
9.5-10.5
413-423
35-45
56.2-58.6
12
-------
SECTION 5
PARTICULATE MEASUREMENTS
Modified Brink cascade impactors were operated at the ESP
inlet and baghouse inlet (Sample Points 1 and 3), Andersen
Mark III cascade impactors were operated at the precipitator
outlet (Sample Point 2), and University of Washington Mark III
(Pilat) cascade impactors were operated at the baghouse outlet
(Sample Point 4). The collection substrates (glass fiber) used
in the Brink and Andersen impactors were conditioned by acid
washing in the laboratory to decrease the amount of substrate
weight gain due to gas phase reactions. The collection sub-
strates used in the University of Washington (Pilat) impactors
were greased stainless-steel substrates which were prepared in
the laboratory prior to use in the field. All impactor data
obtained at Red Lake were reduced using a computer program
described in EPA Report Number 600/7-78-042, A Computer-based
Cascade Impactor Data Reduction System, by J.W. Johnson, G.I.
Clinard, L.G. Felix, and J.D. McCain, March 1978. Details of
the cascade impactors are given in the Appendix, Figures A-l,
A-2, and A-3.
The impactor data collected at the ESP inlet sampling loca-
tion are presented in Figure 5 as cumulative mass loading vs.
particle size. Figure 6 presents the same data as cumulative
percent vs. particle diameter. As can be seen from Figure 6,
50% of the particulate at the ESP inlet was less than 16 ym. A
maximum particle diameter of 25 ym was used in the impactor data
reduction program for each sampling location, since the size of
the largest particle found in the cyclone catches of the Brink
impactors operated at the ESP and baghouse inlets was 25 ym.
As indicated earlier, satisfactory particulate sampling
could not be carried out at the ESP outlet; however, a limited
number of impactor runs were made at the ESP outlet to approx-
imate the loading of particulate less than two microns would
be present at the outlet of the condenser in addition to that
resulting from condensation (or desublimation). Figures
7 and 8 present data on the cumulative mass vs. particle size
and cumulative percent vs. particle size at the ESP outlet.
Figures 9 and 10 show the same data for the baghouse inlet and
Figures 11 and 12 for the baghouse outlet.
Figure 7 shows that the mass loading of particulate below
2 microns diameter at the ESP outlet was less than 7 x 102ing/ACM,
13
-------
U
rn 10s
104-:
MG/ACM = MILLIGRAMS/ACTUAL CUBIC METER
GR/ACF = GRAINS/ACTUAL CUBIC FOOT
^lO3
\
CD
M
a
8]
-"-ID
"1
H—i i i mil 1—i i 1111H 1—i i i mil
1CT1 10° 101
PARTICLE DIAMETER (MICROMETERS)
Figure 5. Average ESP inlet cumulative size distribution.
14
-------
99.99
99.8
99.5
UJ
h-
98
95
80
70
BO
50
40
30
EO
lir
0-01
I
10
1-1
*—i ii mH 1—i > i mil 1—i > i mil
101
PARTICLE DIAIv€TER (MICROMETERS)
Figure 6. Average ESP inlet cumulative percent versus
particle diameter.
15
-------
MG/ACM = MILLIGRAMS/ACTUAL CUBIC METER
GR/ACF = GRAINS/ACTUAL CUBIC FOOT
M
a
_j
103--
TlO1
fa
—I—I I HUH 1—I I I HH| 1—I * HUH
1O"1 10° 101
PARTICLE DIAMETER (MICROMETERS)
Figure 7. Average ESP outlet cumulative size distribution.
16
-------
and Figure 8 indicates that this amounted to about 60% of the
total mass at the ESP outlet. The data for the baghouse inlet,
after addition of quench air in the mixer-cooler, in Figures 9
and 10 show that the mass loading of particulate below 2 microns
at the baghouse inlet was about 7 X 103 mg/ACM and also amounted
to about 60% of the total mass at this point. Figures 11 and 12
show that the particulate below 2 microns at the baghouse outlet
was less than 5 X 10~2 mg/ACM or about 10% of the total particu-
late emission. Data from Figures 9 and 11 show (for example)
that the baghouse collection efficiency for particulate below 2
microns was greater than 99.99%. This is further illustrated in
Figure 13 which shows the baghouse collection efficiency and
penetration as a function of particle size.
Figure 13 shows that a minimum collection efficiency of
about 99.8% occurred at about 5 microns with another "minimum
point" at about 0.5 microns. This bimodal distribution is not
abnormal for baghouse performance. Other workers in the field
have attributed the increased penetration (decreased collection
efficiency) at around 0.5 microns to the lack of inertial or
diffusional collection mechanisms, and the increase around 5
microns has been attributed to seepage3.
One concern in the evaluation of data from the ESP and
baghouse was that carryover from the ESP would partially obscure
performance of the mixer cooler and provide erroneous data on
particulate at the inlet to the baghouse. However, data in
Figure 7 shows that the mass loading of particulate below 2
microns diameter at the ESP outlet was less than 7 X 102 mg/ACM
at a gas flow rate of 3888 dscm/hr (from Table 5). Figure 9
shows that the mass loading of particulate below 2 microns
diameter at the baghouse inlet was about 7 X 103 mg/ACM at a gas
flow rate of about 16,779 dscm/hr. The ratio of these numbers
(7 X 102 X 3888 -r 7 X 103 X 16,779) indicates that less than five
percent of the particulate entering the baghouse was particulate
at the ESP outlet; therefore, greater than 95 percent of the
particulate entering the baghouse resulted from condensation in
the mixer-cooler, and the data shown in Figures 9 to 13 are
accurate reflections of the particulate formed in the mixer-
cooler and removed by the baghouse. The mass loadings from all
the impactor runs used in the data reduction programs are pre-
sented in Table 4. Based on these data, the average particulate
collection efficiency of the baghouse exceeds 99.9%. Data from
the mass sampling trains and the trace element trains also indi-
cate an average efficiency above 99.9% for the baghouse.
Impactor data for the precipitator show a particulate
collection efficiency of 98.3%. Trace element data (collected
independently by Radian) show a somewhat lower precipitator
efficiency of 97.1%. However, all data agree that the overall
particulate collection efficiency of the "hot ESP/cold baghouse
system" exceeds 99.9%.
17
-------
h-
bJ
a!
Q!
U
M
g
33.33
99!s:
99.5-
33 1
95:
90^
BO-
ZO ^
60^
501
4O1
10-
Bi
1]
O.E:
o.oi:
1C
»
'
I /
...-»'
r
f
*
*
; *
»**
•
r1 10° lo1 ic
K *^K.^^^ ^^ i^^^« ^M» ^i^ MM* • fc kT^M*«M^B^0^ ^ K A ••• ••^^^•^^••^tt M»B»^VV^^^*^^h^B>^ «.
PARTICLE DIAMETER (MIO^WETERS)
Figure 8. Average ESP outlet cumulative percent versus particle diameter.
18
-------
\
1O4::
id3--
M
<
MG/ACM = MILLIGRAMS/ACTUAL CUBIC METER
GR/ACF = GRAINS/ACTUAL CUBIC FOOT
u
\
M
a
-LID
ri
H—i < i mil 1—i i 11 ml 1—i i i mil
icr1 10°
PARTICLE DIAMET
(MICRDMET
Figure 9. Average baghouse inlet cumulative size distribution.
19
-------
LU
U
CK
LJ
Q.
U
M
h-
a
ra.aa
ggls
99.5-
99-
98-
95-
90-
8O^
"7O "
S\J 1
60^
5°i
•4O "^
Sj
E:
0.5^
O-E:
n. rn _
[
L
; j*
: I
r *
*
; •
: *
**
1 1 1 1 1 Illl 1 1 1 1 i II N 1 1 1 1 1 1 1 |M|
ID'1 1O7 101
PARTICLE DIAMETER (MICROMETERS)
/OL Average baghouse inlet cumulative percent versus
particle diameter.
20
-------
ID"
1—i i i HIM 1—i i i niH 1—i i i mil
1O'1 ±CP 1O1
PARTICLE DIAMETER CMICRDMET
Figure 11. Average baghouse outlet cumulative size distribution.
a:
CD
a
,-7
21
-------
h-
z
UJ
g
tf
>
M
|
99.99
99-95
3979
99.8
99.5
99
98-
95-
90-
80-
70-
60^
50^
4O1
EO-
ei
li
0.5^
O.Ei
o.oi:
1C
L
: i
• I
; X
: X
- •
• *
*
V
ft
1 *
X
X
r1 10° lo1 ic
PARTICLE DIAMETER CMICRDMET
Figure 12. Average baghouse outlet cumulative percent versus
particle diameter.
22
-------
a
iu-=
>• ^^x^L
1 1 j ™
100;
lO'1-
10"2i
•
•
•
«
•
•
•
•
10"4-
1C
«• %
• _
• •
• .
• •
» •
» •
» •
• .
• •
• _
• •
• «
* •
• •
• ' •
* •
» •
• •
» «
• •
• •
• •
» •
t m
• •
» •
•
•
X
• M
! » :
•
» «
• m
• m
• *
» . *n
: * « :
• V •
; •• ;
* -
* *
«»
» *
m * v
•
* *
* «
: « i :
»i
X
!
J* 10P 101 1C
w
r3O.O
r33.0
5
: bJ
M
U
U_
:33-3 fa
:
»
:
j-33.333
3
-99.3399
-?
PARTICLE DIANCTER (MICROMETERS)
Figure 13. Baghouse paniculate collection efficiency.
23
-------
TABLE 4. IMPACTOR MASS LOADINGS
to
*«.
Location
ESP Inlet
ESP Inlet
ESP Inlet
ESP Inlet
Bghs Inlet
Bghs Inlet
Bghs Inlet
Bghs Inlet
Bghs Inlet
Bghs Inlet
Bghs Inlet
Bghs Inlet
Bghs Inlet
Bghs Inlet
Bghs Inlet
Bghs Outlet
Bghs Outlet
Bghs Outlet
Bghs Outlet
Bghs Outlet
ESP Outlet
ESP Outlet
Run#
RLI-3
RLI-6
RLI-7
RLI-8
RLI-12
RLI-13
RLI-14
RLI-16
RLI-17
RLI-18
RLI-19
RLI-21
RLI-22
RLI-23
RLI-24
RLO-1
RLO-2
RLO-3
RLO-5
RLO-6
RLO-8
RLO-9
Baghouse efficiency =
Date
9/19/78
9/20/78
9/20/78
9/20/78
9/21/78
9/21/78
9/21/78
9/25/78
9/25/78
9/25/78
9/25/78
9/26/78
9/26/78
9/26/78
9/26/78
9/19/78
9/20-21/78
9/20-21/78
9/25/78
9/25/78
9/26/78
9/26/78
99.99+%
Sample
Time, min
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
360
960
960
480
480
4
4
Impactor
oF
600
580
555
560
244
230
225
220
220
225
230
228
243
228
243
228
226
226
228
228
575
575
Temp.
°C
316
304
291
293
118
110
107
104
104
107
110
109
117
109
117
109
108
108
109
109
302
302
Mass Loading
gr/SCF
18.73
114.82
50.36
58.82
5.25
5.07
5.34
4.56
5.04
5.14
5.80
4.55
5.19
5.97
6.70
0.00024
0.00029
0.00028
0.00027
0.00029
1.017
1.079
mg/SCM
4.29x10"
2. 63x10 5
1. 15x10 5
1. 35x10 5
1.20x10"
1.16x10"
1.22x10"
1.04x10"
1.15x10"
1.18x10"
1.33x10"
1.04x10"
1.19x10"
1.37x10"
1.53x10"
0.550
0.657
0.642
0.621
0.667
2. 33x10 3
2.47xl03
ESP efficiency = 98.3%
-------
SECTION 6
GAS SAMPLING AND ANALYSIS
Two types of modified EPA Method 5 sampling trains were
used at each sampling site. These two trains differed from each
other only in the number of impingers and the solutions in them.
An illustrative gas sampling train is shown in Figure 14.
The trace element train consisted of 7 modified Smith-
Greenburg impingers operated in an ice bath. Impingers 1 and 2
contained 10% HNOs/ impinger 3 was empty, impingers 4 and 5 con-
tained 10% NaOH, impinger 6 contained 6% H202, and impinger 7
contained silica gel. Basic and neutral species were absorbed
in the HNOs and acid species in the NaOH. The gas was then
scrubbed in H202 and dried by the silica gel and passed into the
control console. Lear Siegler PM-100 stack sampling consoles
were used. These contained all flow controls, temperature and
pressure indicators, a pump, and a dry gas meter.
The sulfur oxide train contained only five impingers. The
first contained 80% isopropanol for absorption of SOs- The
second and third contained 6% H2O2 for absorbing S02, the fourth
was empty, the fifth contained silica gel.
Both trains contained in-stack and out-of-stack filters
which were used to collect particulate (in-stack) and vaporous
arsenic trioxide which condensed in the probe after removal
of other particulate (out-of-stack).
After sampling, extensive cleaning of the sampling train
was required to remove As20a which had condensed in the probe
and glassware. The nozzle, probe, and glassware were rinsed with
H20, acetone, and NaOH solution to dissolve As20s. Since As2Os
is condensable and adheres to walls of sampling equipment, a
rigid washing protocol was used to assure maximum recovery from
the sampling trains.
Solid (particulate) from the in-stack filter and gaseous
components from the out-of-stack filter and the trace element
sampling train were analyzed for four elements: arsenic,
antimony, lead, and selenium. Gas flow velocities, gas analyses,
temperatures, and particulate grain loadings were also determined
at each sample point.
25
-------
FILTER
TEMPERATURE
SENSOR
S-TYPE
PITOT TUBE
TEMPERATURE TEMPERATURE
SENSOR 1Q% HNQ3 DRY 10% NaOH 6% H2O2 SENSOR
(\\ IMPINGERS IMPINGER IMPINGERS IMPINGER ~
VACUUM
LINE
SILICA GEL
DESSICCANT
TEMPERATURE SENSORS
ORIFICE GAUGE
Figure 14. Trace element gas sampling train.
-------
The trace element analyses (for As, Sb, Pb, and Se) were
based on atomic absorption (AA) using flame aspiration and a
graphite furnace attachment. Solid samples were digested' in
HF/HN03/HC1 acid mixtures. Liquid samples and digested solids
were aspirated directly into the air-acetylene flame of the AA
using standard calibration solutions'* and heating programs5.
Gas analyses were based on the use of the sulfur oxide
(SOX) absorption train (for S02 and S03) and the use of Fyrite
solutions (for O2 and CO2). The impinger solutions from the SOX
train were analyzed by liquid ion chromatography6 and moisture
was determined by weight gain (corrected for S02 absorption) in
the impinger train. Since Fyrite also absorbs S02 along with
CO2, values shown for C02 are corrected.
A summary of the gas stream measurements, gas analyses,
and trace element analyses are shown in Table 5. Individual
measurements of gas flows and grain loadings are shown in Table
6. It should also be noted that the difference in arsenic levels
shown in Table 5 for the ESP outlet and baghouse inlet reflect
variations in process conditions (not on-stream changes) since
the ESP and baghouse samplings were not made simultaneously.
The grain loading data in Table 5 show a particulate
removal efficiency of about 97.1% for the hot ESP and better than
99.9% for the baghouse (based on in-stack measurements used with
the trace element trains). Further, based on all particulate
in the system (combining ESP inlet and condensed particulate at
the baghouse inlet), the overall particulate removal efficiency
of the combined ESP-baghouse system was 99.99+%.
Lead and selenium analyses in Table 5 indicate removal
efficiencies for these elements of only about 90%; however, the
amounts of lead and selenium present in the system were very low
and their combined emissions were less than 0.02 Kg/hr (0.04
Ib/hr). Antimony removal efficiency was about 99.5%, but the Sb
emissions from the baghouse were only about 0.01 Kg/hr. For the
more prevalent trace element, arsenic, the overall efficiency of
the baghouse and ESP combined was about 99.9%, with the major
portion emitted being in the gaseous (rather than particulate)
form.
A more detailed analysis of the solid and vapor distri-
bution for arsenic is shown in Table 7. This table shows that
most of the arsenic (As203) entering the precipitator is in a
volatile form and relatively little is caught in the ESP. Par-
ticulate arsenic at the ESP inlet is shown to be only about 3%
of the total arsenic in the stream and most of this is collected
in the precipitator. Comparison of the data for arsenic in the
ESP inlet and outlet indicates that about 15% is collected in
the ESP; however, as indicated earlier, measurements at the ESP
outlet were not considered to be highly reliable. Other data
27
-------
TABLE 5. GAS STREAM MEASUREMENTS AND
TRACE ELEMENTS DISTRIBUTION
Temperature
°C
oF
Velocity
m/sec
ft/sec
Grain Loading
grams/dscm
grains/dscf
Flow Rate
dscm/hr
dscf/hr
Trace Elements
As-Kg/hr
-Ib/hr
Pb-Kg/hr
-Ib/hr
Sb-Kg/hr
Ib/hr
Se-Kg/hr
-Ib/hr
Gas Analysis
%02
%S02
%H20
%C02
ppm SO 3
ESP
Inlet
380
716
21.3
70.0
35.2
15.4
3899
137,680
(total)
179
393
0.14
0.32
2.1
4.7
0.0045
0.010
11.3
8.2
22.8
0.3
336
ESP
Outlet
327
621
21.9
71.9
1.01
0.442
3888
137,314
152*
334
0.05
0.11
0.40
0.88
0.0017
0.0037
13.9
5.9
21.8
1.8
791
Baghouse
Inlet
116
240
16.5
54.0
11.40
4.98
16,779
592,526
105*
232
_
-
_
-
—
-
19.0
1.4
5.5
0.8
187
Baghouse
Outlet
113
236
16.9
55.3
0.00094
0.00041
19,817
699,833
0.20
0.44
0.015
0.034
0.011
0.025
0.0004
0.0009
19.5
1.4
5.6
0.0
<2
* Not measured simultaneously.
28
-------
TABLE 6. GAS FLOW DATA
(Trace element trains)
Site
ESP Inlet
it
M
ii
ti
ii
ii
Avg. ESP Inlet
ESP Outlet
ii
ii
ii
ii
M
ii
ii
ii
Avg. ESP Outlet
Bghs. Inlet
ii
M
ii
ii
ii
ii
Avg. Bghs. Inlet
Bghs. Outlet
it
ii
ii
ii
Avg. Bghs. Outlet
Date
9-21
9-21
9-25
9-25
9-26
9-26
9-26
-
9-18
9-19
9-20
9-20
9-21
9-21
9-25
9-25
9-25
-
9-18
9-19
9-20
9-20
9-27
9-27
9-27
-
9-26
9-26
9-26
9-27
9-27
_
Temperature
°F
704
712
715
715
720
728
718
716
665
643
639
633
611
638
570
590
600
621
249
242
236
246
233
236
237
240
227
238
238
-
242
236
°C
373
378
379
379
382
387
381
380
352
339
337
334
322
337
299
310
316
327
121
117
113
119
112
113
114
116
108
114
114
-
117
113
Velocity
ft/sec
67.3
66.3
69.5
59.8
74.7
76.4
76.1
70.0
69.0
62.3
72.3
77.1
76.3
76.7
69.5
70.0
73.9
71.9
61.2
54.0
55.5
51.8
53.7
51.5
50.0
54.0
54.0
51.1
57.4
59.6
54.2
55.3
m/sec
20.5
20.2
21.2
18.2
22.8
23.3
23.2
21.3
21.0
19.0
22.0
23.5
23.3
23.4
21.2
21.3
22.5
21.9
18.7
16.5
16.9
15.8
16.4
15.7
15.2
16.5
16.5
15.6
17.5
18.2
16.5
16.9
Grain Loading
gr/dscf
_
16.16
13.68
17.20
16.73
13.35
15.4
_
—
_
0.301
0.584
0.494
0.376
0.454
-
0.442
_
-
-
4.437
6.273
5.189
4.007
4.98
-
0.00036
0.00019
0.0007-1
0.00037
0.00041
grams /dscm
—
37.0
31.3
39.4
38.3
30.5
35.2
_
—
_
0.69
1.34
1.13
0.86
1.04
1.01
_
-
-
10.15
14.35
11.87
9.17
11.40
0.00082
0.00043
0.00162
0.00085
0.00094
29
-------
TABLE 7. SOLID AND VAPOR DISTRIBUTION OF As
(Trace element trains)
Phase
SOLID
gm/dscm
Ib/dscf
Kg/hr
Ib/hr
VAPOR
gm/dscm
Ib/dscf
Kg/hr
Ib/hr
TOTAL
Kg/hr
Ib/hr
ESP
Inlet
1.21
7. 58x10" 5
4.7
10.4
44.6
2. 7 8x10" 3
174
383
179
393
ESP
Outlet
2.68xlO~2
1.67xlO~6
0.10
0.23
39.0
2. 4 3x10" 3
152
334
152
334
Baghouse
Inlet
6.20
3. 8 7x10" "
104
229
7. 82x10" 2
4. 89x10" 6
1.31
2.90
105
232
Baghouse
Outlet
1.8xlO~ 3
l.lxlO"7
0.02
0.07
9.14xlO~3
5.71xlO~7 .
1. 81x10" 1
3. 99x10" :
0.20
0.44
30
-------
(analysis of the ESP dust) also indicate that the particulate
arsenic collected in the ESP was less than 2-3% of the total in
the system. In any event, since the primary function of'the ESP
is to recover metal values with minimum collection of" arsenic,
the data obtained during the test program confirmed that the
system was operating effectively.
Data in Table 7 also show that most of the volatile
arsenic (as As2C>3)that passed through the hot ESP was condensed
in the mixer-cooler so that nearly 99% was in particulate form
at the baghouse inlet, and the collection efficiency of particu-
late arsenic in the baghouse was greater than 99.95%. Overall
arsenic collection efficiency in the baghouse was somewhat lower
at 99.8% due to the passage of volatile As2O3. The effectiveness
of the baghouse for recovery of particulate arsenic is further
illustrated by the data for the baghouse outlet which shows a
particulate emission of only 0.02 Kg/hr. Total arsenic emissions
(including both volatiles and particulates) were about 0.2 Kg/hr.
31
-------
SECTION 7
OTHER ANALYSES
In addition to the data obtained with the trace element
trains (reported in the previous section), analyses were made
of the roaster feed, calcine, the ESP dust, and the baghouse
dust. Typical analyses are shown in Table 8. Initially, it was
anticipated that these values could be used in conjunction with
the in-stack particulate analyses and the trace-element-train
analyses to develop detailed material balances over the entire
system. However, since the discharge rates from hoppers of the
ESP and baghouse were not measured, no direct measurements of
the precipitator and baghouse catches could be made that were
directly comparable to the trace-element-train measurements.
Indirect material balances can be made using the grain loading
and flow rates from Table 5 and the analyses from Table 8. The
calculated precipitator catch based on these data was about
135 Kg/hr (300 Ib/hr) with about 2 Kg/hr (5 Ib/hr) as arsenic.
Similarly, the calculated baghouse catch was about 190 Kg/hr
(420 Ib/hr) with about 130 Kg/hr (287 Ib/hr) as arsenic. These
calculations further illustrate that the control system was
operating essentially as designed, since less than 2% of the
total arsenic was collected in the hot ESP and the major portion
was collected in the baghouse.
The partial pressures of As203 in the various gas streams
are shown in Table 9 in comparison to the vapor pressures of
As2O3 at the temperatures of the sample points. The data show
that the partial pressures of As2O3 at the inlet and outlet of
the ESP are substantially below the equilibrium vapor pressure
which confirms that no significant condensation should occur in
the ESP at normal operating temperatures. The calculated par-
tial pressure at the baghouse inlet, however, exceeds the
equilibrium vapor pressure at this point. Since this sampling
point was immediately following the mixer-cooler, it is possible
that the As2O3 volatiles at this stage were supersaturated but
reached equilibrium in passage through the baghouse. In any
event, the final partial pressure (and emission) of volatile
As2O3 from the baghouse is exactly what would be predicted from
vapor pressure data.
Samples of the ESP and baghouse dusts were also examined
using an ETEC Omniscan scanning electron microscope (SEM) and
32
-------
TABLE 8. TRACE ELEMENT ANALYSES
Arsenic, % Lead, ppm
Roaster Feed 2.3 283
Calcine 1.1 353
ESP Dust 1.5 518
Baghouse Dust 68.3 11
Ant imony , ppm
2115
1530
3650
4580
Selenium, ppm
6
8
53
<2
TABLE 9. As 2C>3 PARTIAL PRESSURES
Location
Temperature
Partial Pressure*
(mmHg)
Vapor Pressure**
(mmHg)
ESP Inlet
ESP Outlet
Baghouse Inlet
Baghouse Outlet
380
327
116
113
5.3
4.7
9.4 x 10~3
1.0 x 10~3
300
80
1.5 x KT3
1.0 x 10
-3
* Based on data in Table 7
** CRC Handbook of Chemistry and Physics, 50th Edition (1969) and Behrens,
R. G. and Rosenblatt, G. M., "Vapor Pressure and Thermodynamics of
Octahedral Arsenic Trioxide (Arsenolite)," J. Chem. Thermodynamics, 4_,
175-190 (1972)
33
-------
were qualitatively analyzed using a Princeton Gamma Tech energy
dispersive X-ray analyzer. Figures 15 and 16 are SEM photographs
of the ESP and baghouse solids respectively. The right side of
each photograph is an enlargement of the indicated left field.
The ESP solids appear to contain a variety of crystal
sizes, shapes, and colorations (as would be expected from the
nature and composition of the particulate). By contrast, the
baghouse solids appear to be uniform in color and particle size,
except for agglomeration of small particles.
Scans of the energy dispersion X-ray analyses for the
ESP and baghouse solids are given in Figures 17 and 18, respec-
tively. The major peaks in the ESP solids were identified as
As, Al, Si, S, K, Ca, and Fe. The only major element found in
the baghouse solids was As.
To obtain information on the crystalline species in the
dusts, the ESP and baghouse solids were analyzed by X-ray powder
diffraction using a Siemens X-ray Diffractometer with nickel
filtered CuKa radiation.
The X-ray diffraction pattern obtained for the ESP sample
is given in Figure 19. The peaks on the diffraction pattern
were identified as AS203, SiO?, Fe30if, Fe2O3 , or as unknown.
The d-spacings in Angstroms, A, for the peak assignments are
included on the diffraction pattern. The ESP sample diffraction
pattern for crystalline phases showed the presence of Fe30it
(magnetite), FeaO3 (hematite), SiOa (alpha quartz), and minor
amounts of AsaOs (arsenolite). The three major peaks at 9.40,
3.11, and 2.88 A that could not be positively assigned are
believed to be calcium aluminum silicate.
The diffraction pattern for the baghouse dust is given
in Figure 20. The only crystalline phase identified was arseno-
lite. Several small peaks at 6.18, 3.93, 3.53, and 3.35 A
could not be assigned.
Spark Source Mass Spectrometry (SSMS) was also used to
provide semiquantitative analyses of the ESP and baghouse dusts.
These data were to be used as a guide for analysis of other
potentially volatile materials (other than As, Sb, Pb, and Se);
however, no additional quantitative analyses were made since
the analyses for antimony (Table 5), which was shown to be a
major component in both streams by SSMS, showed very low concen-
trations. SSMS data metal components in the dusts at concen-
trations above 1 ppm are given in Table 10.
34
-------
Figure 15. SEM Photograph of ESP Solids (1410X Magnification]
Figure 16. SEM Photograph of Baghouse Solids (1190X Magnification)
35
-------
u>
3000 T
2500-
2000-
S 1500-
1000-
500
SI
Al
As
Ca
Fe
As
8
i
10
12 14 16 18
ENERGY (KEV.)
Figure 17. Energy dispersive X-ray analysis
of ESP solids.
-------
3000 -i
2500-
1500-
As
§
CJ
2000-
co
1000 -
500-
8 10
ENERGY (KEV.)
Figure 18. Energy dispersive X-ray analysis
of baghouse solids.
-------
Leqend
U)
03
o
X
•s
OJ
-t->
3
C
•r—
E
v>
HI
a.
E
c
QJ
A -
B - S102
C - FesOi,
D - FezOs
X - Unknown
(100) Indicate major peak
Peak assignments Include crystal d
Two Theta Angle (Degrees), CuKa Radiation
Figure 19. X-ray powder diffraction pattern
for ESP solids.
-------
*
o
x »•
3
C
•t—
E
-------
TABLE 10. SSMS ANALYSES
(concentration in ppm weight)
Aluminum
Antimony
Arsenic
Barium
Boron
Cadmium
Calcium
Cerium
Cesium
Chromium
Cobalt
Copper
Gallium
Germanium
Gold
Iron
Lanthenum
Lead
Lithium
Magnesium
Manganese
ESP
Dust
MC
MC
MC
48
8
1
MC
5
1
150
520
MC
7
2
50
MC
4
460
11
MC
510
Bghse.
Dust
63
MC
MC
2
<1
<1
860
-
0.6
9
6
50
1
0.2
-
MC
-
9
1
870
8
Mercury
Molybdenum
Neodymium
Nickel
Potassium
Rubidium
Scandium
Selenium
Silicon
Silver
Sodium
Strontium
Tellurium
Thallium
Tin
Titanium
Tungsten
Vanadium
Yttrium
Zinc
Zirconium
ESP
Dust
NR
3
2
MC
MC
9
5
55
MC
34
MC
36
17
1
50
MC
30
61
15
MC
16
Bghse.
Dust
NR
2
-
63
120
-
0.2
6
650
6
54
0.8
-
• -
0.4
18
3
19
-
120
NR - Not reported
MC - Major component, >1000 ppm
40
-------
SECTION 8
RESISTIVITY ANALYSES
As a part of the test program at Red Lake, voltage-current
(V-I) curves were obtained for the transformer-rectifier set
used on the electrostatic precipitator. A typical V-I curve is
shown in Figure 21. The measurements showed no evidence of a
highly resistive particulate at the hot precipitator operating
at a temperature of about 380°C (720°F) . This again confirms
that the use of the hot precipitator to recover metal values
without condensation of AsaOs is reliable.
We also made measurements of the resistivity of the dust
at the baghouse inlet using SRI's point- to-plane probe. Data
in Table 11 show that the average resistivity of the particulate
entering the baghouse was 1.9 x 1013 ohm-cm at 116°C using the
"spark" or plate-to-plate measurement technique. Figure 22
shows clean plate and dirty plate VI curves from the in-situ
resistivity probe and clearly indicates a highly resistive
particulate.
The measurements at the baghouse inlet were made so that
the feasibility of using an electrostatic precipitator to con-
trol As203 emissions could be assessed using the SRI-EPA mathe-
matical model of electrostatic precipitation7 and data on mass
loading, particle size distribution, gas temperature, electrical
resistivity, and effective ion mobility.
Data from the Red Lake tests showed that conditions at the
baghouse inlet were:
Mass loading - 1.2 x 10* mg/SCM (from Table 4)
Mean particle size - 1.0 micron (from Figure 8)
Gas temperature - 116 °C (from Table 11)
Electrical resistivity - 1.9 x 1013 ohm-cm (from Table 11)
The particulate mass loading is fairly high, and the mass
medium particle diameter of 1.0 ym reflects a very fine particu-
late. This combination of a fairly high mass loading and very
fine particle size distribution would pose significant problems
in collection by electrostatic precipitation. High currents
(ion densities) would have to be achieved to effectively charge
the high number density of particles. The high particulate
number density could quite possibly result in the quenching of
41
-------
EB.O--
C_J
c_n
o_
CD
1B.O--
1E.O-
c_n
CD
LU
CfcT
12.0--
B.O-
2&
041
OPERATING POINT
al
a
a
a
a
a
« a 54
VOLTAGE ,KV
Figure 21. Typical secondary voltage-current curve.
go
42
-------
00
0.17
0.16
0.15
0.14
0.13
0.12
0.11
^ 0.10
u 0.09
cc
3 0.08
o
0.07
0.06
0.05
0.04
0.03
0.02
0.01
10 11 12 13
VOLTAGE, kV
14
15
16
17
Figure 22. Clean and dirty plate voltage-current curves from in situ
point-to-plane resistivity probe.
5650
3.4 x 10'4
.068
1.2 x
ohm-cm
-------
the corona discharge. In addition, even if a steady corona
current could be maintained, sparking would occur at lower than
desirable voltages due to particulate space charge effects which
would result in high values of electric field in the interelec-
trode space.
The high electrical resistivity would also severely limit
the voltages and currents which could be effectively utilized in
the precipitation process. Currents would be limited to 2 nA/cm2
or less for an assumed electrical breakdown strength of 20 kV/cm
with a large susceptibility to back corona and sparking. The
poor electrical conditions which would be imposed due to the
high resistivity of the collected particulate would require the
utilization of a very large precipitator to achieve acceptable
overall mass collection efficiency.
In this application to collect As20s particles, the prop-
erties of the particulate are such as to prohibit the use of a
precipitator as an effective control device unless the particu-
late properties can be modified. The effects due to high
values of mass loading and electrical resistivity and the very
fine particle size distribution would all combine to produce
very poor electrical conditions which would likely be very
unstable. This conclusion is consistent with results reported
by Porle and Lindquist8 who found that back corona conditions
existed in electrostatic precipitators collecting As2Os at
130°C unless the dust was conditioned with sulfuric acid.
44
-------
TABLE 11. IN SITU RESISTIVITY - BAGHOUSE INLET
en
Temperature
Test #
1
2
3
4
5
6
°C
239
241
244
241
232
241
°C
115
116
118
116
111
116
Cell Depth, mm
1.
2.
1.
0.
1.
1.
24
48
00
68
34
18
Resistivity,
Spark
7.
2.
1.
3.
1.
69xl012
19x10 13
—
84X101 3
11x10 13
62x10 : 3
6.
6.
1.
1.
1.
1.
ohm -cm
VI
25x10 12
49x10 12
17x10 : 2
40x10 l 3
49xl012
16x10 ' 3
E , kv/cm
26.
30.
36.
52,
28.
35.
6
2
0
9
0
6
Average
240
116
1.9xl013 6.8xl012
-------
SECTION 9
TECHNOLOGY TRANSFER
One of the objectives of the test program at Red Lake was
to determine whether their system for recovery of metal values
in a hot electrostatic precipitator, followed by quench cooling
with air with collection of condensed AS203 particulate in a
cold baghouse, has potential for technology transfer and use
for As2Os control in other nonferrous applications. The test
program clearly demonstrated and confirmed that the Campbell
Red Lake Mines technology is effective for control of arsenic
emissions from a nonferrous smelter. Modification of the system
for use in other nonferrous smelters (e.g. copper smelters)
appears to be technically feasible if no major difficulties are
encountered in scale-up from the relatively small operation at
Red Lake to a large smelting operation.
The design data for Red Lake are based on a hot-gas flow
to the precipitator of about 93 SCMM (3273 SCFM) or 207 ACMM
(7300 ACFM) at 370°C (700°F). After removal of low-arsenic
particulate in the ESP, the gas is diluted with 302 SCMM
(10,650 SCFM) ambient air in the mixer-cooler so that the gas
flow to the cold baghouse is 395 SCMM (13,923 SCFM) or 518 ACMM
(18,300 ACFM) at 107°C (225°F). In their case, the dilution
ratio from the ESP to the baghouse was about 4.25 (i.e. on a
standard basis, the gas flow through the baghouse was 4.25 times
the flow through the ESP). This significant increase in gas
flow is acceptable for the Red Lake installation; however, this
may be a limitation in larger smelters. In any event, major
considerations in potential application of this control tech-
nology to other nonferrous operations would involve design scale-
up and sizing of the dilution mixer-cooler system and an economic
evaluation of the baghouse system for handling the larger gas
volumes that would be required in major nonferrous smelting units.
Overall, it appears that the hot ESP/mixer-cooler/cold
baghouse technology may be applicable to control of arsenic
emissions from systems such as copper reverberatory furnaces,
electric furnaces, and converters. In some cases, such as
reverberatory or electric furnaces now using hot electrostatic
precipitators, control of arsenic emissions may be improved by
addition of a mixer-cooler system and a cold baghouse to existing
installations. Detailed heat and material balances would be
46
-------
required for each particular situation; however, a preliminary
evaluation of the potential applicability of the system can be
made with only a limited amount of data concerning the arsenic
input and distribution within the system, the gas flows and
temperatures in existing (or proposed) control devices, the
subsequent treatment and disposition of the gas, and the vapor
pressure of
For example, in primary copper smelters processing ores
containing only small quantities of arsenic (e.g. less than 0.1
percent) in the feed, it is probable that the partial pressure
of As203 in any process stream would be less than the vapor
pressure at the exit temperature of the control device (e.g.
7 x 10~" mmHg at 107°C) . In such cases, the hot ESP/air dilu-
tion/cold baghouse system would clearly have no value. In
other smelters where the process stream is sent to an acid plant,
current particulate collection and scrubbing practices are
generally adequate for removal of arsenic from the gas stream
so that atmospheric emissions can be controlled. Also, if the
process stream is to be sent to an acid plant, excessive air
dilution would be undesirable so that the hot ESP/air dilution/
cold baghouse control technique would not be feasible.
However, in those cases where off-gases from a smelting
operation are vented to the atmosphere, the Red Lake technology
for control of arsenic emissions may warrant at least a pre-
liminary evaluation. For example, the gas flow from an elec-
trostatic precipitator controlling particulate emissions from
a copper reverberatory furnace may be about 70 m3/sec at 315 °C
(150,000 ft3/min at 600°F) . Air dilution and cooling of this
gas stream to 107°C (225°F) would require about 90 m3/sec
(193,000 ft3/min) of ambient air at 25°C (77°F). The vapor
pressure of As203 at 107°C can be calculated from the data of
Behrens and Rosenblatt9 as follows:
log p (atm) = - 6067/T(°K) + 9.905
For an AsaOa vapor pressure of 7 x 10"4 mmHg, the calculated
emissions would be about 8 mg/SCM(5 x 10~7 Ib/SCF). The
emissions from the reverberatory furnace control system would
then be about 3.6 Kg/hr (8 Ib/hr) as As203 or 2.7 Kg/hr (6 Ib/hr)
as arsenic. For some copper smelters, this emission level for
arsenic may be an improvement over existing systems; however,
potential applicability of the hot ESP/air dilution/cold bag-
house technology to any smelter operation would require a
detailed evaluation and comparison with other systems such as
scrubbers or spray chamber/ESP or baghouse collector combina-
tions.
Alternatively, the use of the ESP/air dilution/cold bag-
house concept may be combined with a water spray chamber to
47
-------
decrease the gas loading through the baghouse or to lower the
temperature attainable with a spray chamber without encountering
dewpoint problems in the dust collector. These latter approaches
are beyond the scope of the demonstrated control technology in
use at Red Lake, however.
48
-------
REFERENCES
1. Goodfellow, H. and Gellender, M., "Arsenic Poses Tricky
Recovery Task," Canadian Chemical Processing, pp. 26-27,
February 1978.
2. Goodfellow, H. et al, "Arsenic Removal from Roaster Off-
Gases," Fourth International Clean Air Congress (1978).
3. Ensor, D.S., Hooper, R.G., and Scheck, R.W., "Determination
of the Fractional Efficiency, Opacity Characteristics,
Engineering and Economic Aspects of a Fabric Filter Operat-
ing on a Utility Boiler," EPRI FP-297, November 1976.
4. Perkin-Elmer, Analytical Methods for Atomic Absorption
Spectrophotometry. Norwalk, CT, 1973.
5. Perkin-Elmer, Analytical Methods for Atomic Absorption
Spectroscopy Using the HGA Graphite Furnace. Norwalk, CT,
March 1973.
6. Dionex Corporation, Analytical Ion Chromatography, Models
10 and 14, Operation and Maintenance Manual. Palo Alto,
CA, Jan. 1976.
7. Gooch, J.P., McDonald, J.R., and Oglesby, S., A Mathematical
Model of Electrostatic Precipitation. Environmental Pro-
tection Technology Series, Publication No. EPA-650/2-75-037
(April 1975).
8. Porle, K. and Lindquist, B., "Electrostatic Precipitators
(ESP's) in Two Stages Used for Arsenic Recovery at the
Ronnskar Copper Works," Symposium Preprint Control of
Particulate Emissions in the Primary Nonferrous Metals
Industries (March 1979).
9. Behrens, R.G. and Rosenblatt, G.M., "Vapor Pressure and
Thermodynamics of Octahedral Arsenic Trioxide (Arsenolite)",
J. Chem. Thermodynamics, 4^, 175-190 (1972).
49
-------
APPENDIX
NOZZLE
PRECOLLECTION
CYCLONE
JET STAGE
(7 TOTAL)
COLLECTION
PLATE
SPRING
Figure A-1. Modified Brink BMS-11 cascade impactor.
-------
JET STAGE (9 TOTAL)
SPACERS
GLASS FIBER
COLLECTION
SUBSTRATE
NOZZLE
INLET
BACKUP ^
FILTER -^
PLATE
HOLDER-
CORE
Figure A-2. Andersen Mark III cascade impactor.
51
-------
JET STAGE O-RING
COLLECTION PLATE
INLET
\
FILTER HOLDER
COLLECTION
PLATE (7 TOTAL)
JET STAGE
(7 TOTAL)
Figure A-3. University of Washington Mark III cascade impactor.
52
-------
103-
:I
I
I
I
t i Mini 1—i < i mil 1—i ii iitH
PARTICLE DIAMETER (MICROMET
Figure A-4. DM/DlogD versus particle diameter for ESP inlet data.
53
-------
103--
i
.
ii
—i—i i i inH 1—i i > mil 1—i i i MM!
icr1 icP lo1
PARTICLE DIAMETER (MICROMETERS)
Figure A-5. DM/DlogD versus particle diameter for ESP outlet data.
54
-------
ID4-
a
CD
§
§
a
103::
102::
101-
I
i i mill—i i i UHI|—i i i Mini—i \ i HUH
UDP
PARTICLE DIAMETER (MICROMETERS)
Figure A-6. DM/DlogD versus particle diameter for baghouse inlet data.
55
-------
1 1D~^ r
1O
r3.
I
i i m i H - ( — i i I HIM
t I i niH
ICT1 10°
PARTICLE DIAMETER (MICRDMET
Figure A-7. DM/DlogD versus particle diameter for baghouse outlet data.
56
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
FPA-ftnO/?-80-141
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Evaluation of Technology for Control of Arsenic Emission
at the Campbell Red Lake Gold Smelter
5. REPORT DATE
JUNE 1980 ISSUING DATE.
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
G. H. Merchant and R. L. Meek
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Southern Research Institute
2000 Ninth Avenue, South
Birmingham, Alabama 35205
10. PROGRAM ELEMENT NO.
1AB604
11. CONTRACT/GRANT NO.
R-804955
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Laboratory
Office of Research and Development
U. S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
f)np nf
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
Project Officer: John 0. Burckle
16. ABSTRACT
The Campbell Red Lake Mines Gold Smelter at Balmerton, Ontario, Canada, has developed
and implemented a successful control strategy for arsenic emissions from a nonferrous
smelting operation. The Red Lake smelter uses cyclones and a hot electrostatic
precipitator to recover metal values from roaster dusts with subsequent air quenching
to condense (or desublime) arsenic trioxide which is recovered in a low-temperature
baghouse. This report presents the results of the test program conducted at the Red
Lake gold smelter during the period September 18-28, 1978. The overall efficiency for
the control of particulate emissions using a combination of a hot electrostatic
precipitator, an air quench in a mixer-cooler, and a cold baghouse exceeded 99.9%.
Collection of particulate arsenic in the baghouse was greater than 99.95%; however,
overall arsenic collection efficiency in the baghouse was slightly less due to passage
of AS203 vapors. Total arsenic emissions from the system were only 11 mg/dscm
(6.8 x 10~7 Ib/dscf or 0.20 Kg/hr (0.44 Ib/hr).
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
ixhaust Emissions
Smelting
Trace Elements
Pollution
arsenic
arsenic trioxide
air pollution control
gold roaster/smelter
fabric filter/baghouse
13B
8. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport}
Unclassified
21. NO. OF PAGES
67
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77)
PREVIOUS EDITION IS OBSOLETE
•ts U.S. SnvEfWFNT PRINTING nn-itt: 1980--657-165/"007
57
------- |