United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EMB Report 85-CHM-12
June 1985
Air
Chromium Screening
Study Test Report
Harbison-Walker
Refractories
Baltimore,
Maryland
-------
EMISSION TEST REPORT
HARBISON-WALKER REFRACTORIES
BALTIMORE, MARYLAND
ESED 85/02
EMB 85CHM-12
Prepared by
Entropy Environmentalists, Inc.
Post Office Box 12291
Research Triangle Park, North Carolina 27709
Contract No. 68-02-3852
Work Assignments No. 18, 21, and 23
PN: 3023
EPA Task Manager
Dennis Holzschuh
U. S. ENVIRONMENTAL PROTECTION AGENCY
EMISSION MEASUREMENT BRANCH
EMISSION STANDARDS AND ENGINEERING DIVISION
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
November 1985
-------
CONTENTS
Figures iv
Tables v
1.0 INTRODUCTION 1-1
2.0 PROCESS OPERATION 2-1
2.1 Process Description 2-1
2.2 Air Pollution Control System 2-4
2.3 Process Conditions During Testing 2-4
3.0 SUMMARY OF RESULTS 3-1
3.1 Particulate Matter, Hexavalent Chroniun and Total Chromium 3-3
3.1.1 Rotary Dryer Exhaust 3-4
3.1.2 Fabric Filter Outlet 3-7
3.1.3 Tunnel Kiln Stack 3-9
3.2 Particle Size Distribution 3-10
3.3 Emissions in Units of Process Rate and Control Equipment
Collection Efficiency 3-12
3.4 Sunmary of Analytical Results for Hexavalent and Total Chromium 3-14
3.5 Visible Emissions Observation Data 3-16
4.0 SAMPLING LOCATIONS AND TEST METHODS 4-1
4.1 Rotary Dryer Exhaust (Sampling Location A) 4-1
4.2 Fabric Filter Outlet (Sampling Location B) 4-5
4.3 Fabric Filter Dust Hopper (Sampling Location C) 4-7
4.4 Tunnel Kiln Stack (Sampling Location D) 4-7
4.5 Rotary Dryer Feed (Sampling Location E) 4-9
4.6 Velocity and Gas Temperature 4-9
. 4.7 Molecular Weight 4-9
4.8 Particulate Matter 4-10
4.9 Particle Size Distribution 4-10
4.10 Hexavalent Chromiun Content 4-11
4.11 Total Chrcraiun Content 4-11
5.0 QUALITY ASSURANCE 5-1
ii
-------
CONTENTS (Continued)
APPENDICES
A TEST RESULTS AND EXAMPLE CALCULATIONS A-1
Particulate, Hexavalent Chromium and Total Chromiun A-3
Example Particulate Test Calculations A-9
Particle Size for Total Particulate, Hexavalent Chromiun A-32
and Total Chromiun
Hexavalent Chromiun Particle Size Analytical Data A-58
Total Chrccniun Particle Size Analytical Data A-60
Total Chromium Analysis Calculation A-62
Explanation of Total Chromium Analysis Calculation Table A-64
B FIELD AND ANALYTICAL DATA B-1
Particulate Matter B-3
Particle Size Distribution B-30
Total Particulate Analysis B-45
Particle Size Distribution Analysis B-55
Hexavalent Chromiun Analysis B-62
Total Chromium Analysis B-67
C VISIBLE EMISSION OBSERVATION DATA C-1
D SAMPLING AND ANALYTICAL PROCEDURES D-1
Determination of Total Particulate Emissions D-3
Determination of Hexavalent Chromium Emissions D-8
Determination of Total Chromium Content D-15
Determination of Particle Size Distribution D-22
Grab Samples D-29
E CALIBRATION AND QUALITY ASSURANCE DATA E-1
F MRI PROCESS DATA F-1
Process and Emission Capture Efficiency Observations of Rotary
Dryer Baghouse During Testing F-3
Process Observation of the No. 1 Tunnel Kiln During Testing F-7
G TEST PARTICIPANTS AND OBSERVERS G-1
ill
-------
FIGURES
Number Page
2-1 Flow Diagram for the Manufacture of Chromiun Refractories 2-2
4-1 Process Air Flow Schematic of Rotary Dryer and Tunnel Kiln 4-2
4-2 Rotary Dryer Exhaust Duct (Sampling Location A) 4-4
4-3 Fabric Filter Outlet Stack (Sampling Location B) 4-6
4-4 Tunnel Kiln Stack (Sampling Location D) 4-8
IV
-------
TABLES
Number Page
2.1 Final Operating Parameters for the Rotary Dryer and Tunnel
Kiln No. 1 2-6
3.1 Testing Schedule for Harbison-Walker 3-2
3.2 Sumnary of Flue Gas Conditions 3-5
3.3 Sunmary of Participate, Hexavalent Chromiun, and Total
Chromium Emissions 3-6
3.4 Smmary of Particle Size Distribution 3-11
3.5 Summary of Emission Rates in Units of Process Rate and Efficiency 3-13
3.6 Sunmary of Analytical Results for Hexavalent and Total Chromiun 3-15
3.7 Sunmary of Analytical Results for Hexavalent and Total Chromiun
Quality Assurance Samples 3-17
3.8 Sunmary of Visible Emissions Data for Fabric Filter for
Harbison-Walker 3-18
4.1 Sampling Plan for Harbison-Walker 4-3
5.1 Particle Size Blank Filter and Reactivity Filter Analysis 5-2
5.2 Audit Report Chromiun Analysis 5-4
-------
1.0 INTRODUCTION
During the week of June 24, 1985, Entropy Environmentalists, Inc. conducted
an emission measurement program at Harbison-Walker Refractories plant located
in Baltimore, Maryland. The purpose of this program was to provide data for a
screening study to determine the quantity and form of chromiun emissions
associated with the manufacture of refractories.
Comprehensive testing was conducted on a rotary dryer controlled by a
cyclone and a fabric filter and an uncontrolled tunnel kiln.
The rotary dryer and the tunnel kiln at this plant were selected for source
testing for the following reasons:
The rotary dryer and tunnel kiln are typical process equipment for
the refractories industry. Both the chromite ore processed and the
chromiun refractory bricks produced at this plant are typical for the
industry. Thus, uncontrolled emissions measured at this plant are
expected to be representative of those at other plants in the
industry.
The rotary dryer is operated and used to dry chromiun-containing raw
materials at least 8 to 12 hours per day. Dryers at other plants are
operated far less often.
The fabric filter used to control the dryer and the degree of control
achieved are typical of those at other plants in the industry. Thus,
the controlled emissions measured at this plant are expected to be
representative of those at other plants in the industry.
None of the tunnel kilns used in the industry are controlled.
Participate concentrations and mass emission rates were measured at the
rotary dryer exhaust duct, fabric filter outlet, and tunnel kiln stack using
U. S. Environmental Protection Agency (EPA) Reference Method 5.* Total chrcmiun
*40 CFR 60, Appendix A, Reference Method 5, July 1, 1981.
1-1
-------
concentrations and hexavalent chromium concentrations were measured at the same
locations by further analysis of the Method 5 samples using the alternate
sample preparation and analytical procedures as described in Appendix D. Flue
gas flow rates, temperature, moisture content, and composition [oxygen (02),
carbon dioxide (002), and carbon monoxide (CO)] were measured in conjunction with
the particulate tests. In addition, the particle size distribution of
particulate matter emissions at the rotary dryer exhaust, the fabric filter
outlet, and tunnel kiln stack was determined along with hexavalent and total
chromium distribution by particle size.
Representative samples of the dust collected by the fabric filter were
collected during the particulate tests for determination of the hexavalent and
total chromiun content of the material entering the fabric filter.
Mr. Michael Maul [Midwest Research Institute (MRI)] monitored process
operation throughout the test period. Mr. Dennis Holzschuh (EPA Task Manager)
of the Emission Measurement Branch (EMB) observed the test program. Mr. A. H.
Clark, Assistant Plant Manager - Engineering and Pat McDermott, Engineering,
served as the plant contacts and Mr. Ralph Crawford, Manager of Engineering,
served as the corporate contact for Harbison-Walker.
This report is organized into several sections addressing various aspects
of the testing program. Immediately following this introduction is the "Process
Operation" section which includes a description of the process and control
device tested. Following this is the "Summary of Results" section which
presents table summaries of the test data and discusses these results. The next
section, "Sampling Locations and Test Methods" describes and illustrates the
sampling locations for emissions testing and grab sampling and then explains the
sampling strategies used. The final section, "Quality Assurance," notes the
1-2
-------
procedures used to ensure the integrity of the sampling program. The
Appendices present the complete Test Results and Example Calculations
(Appendix A); Field and Analytical Data (Appendix B); Visible Emissions
Observation Data (Appendix C); Sampling and Analytical Procedures (Appendix D);
Calibration Data (Appendix E); MRI Process Data (Appendix F); and Test
Participants and Observers (Appendix G).
1-3
-------
2.0 PROCESS OPERATION
2.1 PROCESS DESCRIPTION
Harbison-Walker Refractories is a diversified refractories producer
with several plants throughout the country. The Baltimore, Maryland,
plant produces basic refractory brick and unformed refractories (specialty
refractories). Magnesite-chromium refractory brick is the predominant
basic refractory brick produced at the plant. The Baltimore plant is
listed by the U.S. Bureau of Mines as one of the major users of chromium
in the refractory industry. This plant is among the larger suppliers of
chromium-containing refractories. The chromite ore content in the
magnesite-chromium refractory brick ranges between 0 and 80 percent.
The process equipment used to produce basic brick at the Baltimore plant
is typical of that used in the industry.
Figure 2-1 is a flow diagram for the manufacture of refractories
that contain chromium at the plant. Chromite ore, which is imported, is
stored outdoors in stockpiles adjacent to stockpiles of magnesite ore.
Magnesite-chromium brick that are rejected for sale (called "bats") are
also typically stored outdoors until they are reprocessed. A front-end
loader is used to transfer magnesite ore and the bats to a vibrating
grizzly, which feeds into a gyrating crusher. The crushed bats are
transported by belt conveyor to the rotary dryer that was tested, which
is used to dry raw materials containing chromium. The crushed magnesite
ore is transported by belt conveyor to a rotary dryer used to dry only
magnesite ore. The chromite ore, which does not require crushing, is
placed directly into a hopper next to the chromium rotary dryer by a
front-end loader.
The chromium rotary dryer is approximately 6 feet (ft) in diameter
and 60 ft in length. The burner is located at the discharge end of the
dryer and is fired by natural gas, although No. 2 fuel oil can be used.
The dryer operates at temperatures between 500° and 600°F and has a
rated production capacity of 25 tons/h. The actual production rate is
about 14 to 20 tons/h, however, because of limitations in kiln production
rate. Moisture content of the chromite ore is approximately 2 percent
prior to drying and less than 0.5 percent after drying.
2-1
-------
PRESS
I
J 1
\
ROTARY DRYER
(CHROMIUM)
ROTARY DRYER
(MAGNESITE)
\
PACKAGING AND
SHIPPING
MIXEU
TUNNEL KILN
STORAGE AND HOLDING BINS
fiATCM
RAILCAR
SCREENS/ BALL HILL/
AIR DRY PANS
SEPARATOR
PACKAGING AND
SHIPPING
Figure 2-1. Flow diagram for the manufacture of chromium refractories.
-------
The hot, dried materials, which are typically 350° to 400°F in
temperature at the dryer discharge end, are sent through a rotary cooler
located next to the rotary dryer. The rotary cooler consists of a
horizontal, cyclindrical steel shell measuring approximately 6 ft in
diameter and 40 ft in length. Cooling is achieved by the countercurrent
flow of ambient air into the cooler, which reduces the temperature of
the dried materials to 150°F. The rotary cooler is used only during
summer months to hasten cooling so that binding agents added later in
the production process will not evaporate.
After drying, the chromite ore and the other raw materials are
ground to specified sizes. The different size raw materials are weighed
out in various proportions and mixed with a liquid binder and other
minor additives. The resultant mixture is then fed to a mechanical
brick press where various refractory shapes are produced. Any scrap
from the presses is recycled to the grinding, screening, and mixing
stages of the process.
Although some of the pressed brick are dried, packaged, and shipped
without further processing, most of the pressed brick are placed on a
refractory-protected railcar and sent through either tunnel kiln No. 1
or No. 2 where they are fired at high temperatures. Each kiln measures
approximately 320 ft in length, 12 ft in height, and 10 ft in width.
The kilns are equipped with 18 burners mounted along the middle section
of both sidewalls and are fired by natural gas.
Each kiln has three zones, a preheating zone at the entrance of
the kiln, a firing zone in the middle section of the kiln, and a cooling
zone. The hot combustion gases are drawn countercurrent to the flow of
brick through the firing and preheating zones by an induced draft fan.
The kiln exhaust gases, which total approximately 46,000 acfm at 660°F,
are ducted out through a stack. The kiln firing temperature zone is
typically kept at 3200°F. The entrances to the tunnel kilns are double
sealed with sliding doors to prevent air from being drawn in and disrupting
the countercurrent airflow. Approximately 48 hours are required for the
brick to travel the entire length of a kiln. Brick exiting the kilns
are cooled and subsequently packaged and shipped.
2-3
-------
2.2 AIR POLLUTION CONTROL SYSTEM
Emission points in the production process are controlled by baghouses
for raw material recovery and for air pollution control. The chromite
ore dryer is equipped with a single compartment, negative-pressure
baghouse preceded by a cyclone. The baghouse is located above the
rotary dryer and .inside the building that houses the dryer. The baghouse
contains 456 woven glass bags that each measure 5 inches in diameter and
9 ft in length, resulting in a cloth area of 5,372 square feet. The
baghouse is designed to handle 9,000 acfm of gas at an air-to-cloth
ratio of 2.5:1 and a pressure drop of 6 inches w.c. The bags are cleaned
once every 2 hours using vibration created by sonic horns. Prior to
entering the baghouse, the temperature of the heated gas stream is about
350°F. The baghouse exhaust is discharged at a temperature of about
140°F through a rectangular stack that extends from the baghouse outlet
up through the roof of the building. Tests were conducted at both inlet
and outlet to the chromite ore dryer.
The rotary cooler is controlled by a negative-pressure, continuous-
cleaning pulse jet baghouse. The baghouse has 1 compartment with 144
polyester felt bags. The bags measure 6 inches in diameter and 8.3 feet
in length. The total cloth area is 1,887 square feet. The baghouse is
designed to handle 10,000 acfm of gas at an air-to-cloth ratio of 5:1.
The stack dimensions are the same as those of the rotary dryer baghouse
stack. The cooler baghouse was not tested.
Neither tunnel kiln is equipped with a control device. Combustion
gases and any entrained particulate matter are emitted directly to the
atmosphere through a stack at an exhaust gas temperature of between 700°
and 800°F. Tests were conducted at the exhaust stack of the No. 1 kiln.
2.3 PROCESS CONDITIONS DURING TESTING
The rotary dryer, the rotary dryer baghouse, and the No. 1 tunnel
kiln were monitored to ensure normal operation throughout the test. The
dryer process parameters monitored included type of chromium product
dried, percent Cr203 in the product, production rates, and operating
procedures. The kiln process parameters monitored included type of
chromium product dried, percent Cr203 in the product, production rates,
2-4
-------
operating procedures, draft pressure, temperatures in the firing zone,
exhaust gas temperatures, and fuel flow to the kiln. Observations of
fugitive emissions escaping the dryer were made about every 15 minutes
of operation, and estimates of emission capture efficiency were recorded.
In addition, system wide pressure drop for the baghouse was monitored,
as were visible emissions from the baghouse stack.
Table 1 lists the material throughput and characteristics for the
dryer and No. 1 tunnel kiln during testing. Process parameters and
observations recorded during the test program are presented in Appendix F
Processes were operated within normal limits throughout the tests.
Observations of the dryer operation during testing were made about every
15 minutes. No gauges were available to monitor the dryer firing tempera-
ture. The flame temperature is adjusted based on operator judgment.
Because of the steady-state nature of the tunnel kiln, plant personnel
only record the firing zone temperature every 2 hours, the draft pressure
and exhaust temperature once every 4 hours, and the fuel flow once
each shift.
During the first test on Tuesday, June 25, three different chromite
ores were processed by the dryer. For the last 20 minutes of the run
during which the second chromite ore was being processed, the dryer
flame was turned off because there was sufficient heat in the dryer to
dry the remaining material adequately. This flame shutdown is considered
normal operation by the plant and was done during each test run. Twenty
minutes into the run during which the third chromite ore was being
processed, the chain on the rotary dryer slipped. The dryer was shut
down, and the test was discontinued for the day. The baghouse was not
cleaned during this test run.
On Wednesday, June 26, the second test covered processing of one
type of chromite ore. Testing continued during the cleaning of the
baghouse. Visible emissions ranging between 0 and 35 percent opacity
were noted during cleaning and immediately after the baghouse came back
on-line.
The third test on June 26 consisted entirely of recycled bats.
Because the brick typically contains less moisture than the ore, the
intensity of the burner flame was decreased from that used for ore
2-5
-------
TABLE 2-1. FINAL OPERATING PARAMETERS FOR THE ROTARY
DRYER AND TUNNEL KILN NO. 1
Test
run No.
1
1
2
3
4
4
1
2
3
Chromite Process
product weight, tons
Hammond-10
T-chrome
BC-3
Nu-60 bats
T-chrome
Hammond-10
0.65 Nu-60b
0.35 HW31-73
0.70 Nu-60
0.30 other brick
0.70 Nu-60
0.30 other brick
Rotary dryer
35.00
3.56
27.90
37.20
12.30
21.35
Tunnel kiln No.
7.35
8.59
8.03
Percent
Cr203
32
56.5
46.3
14.5
56.5
32.0
1
14.5
10. 0C
14.5
10.0
14.5,
10. Ob
Cr203
throughput
11.20
2.01
12.92
5.39
6.95
6.83
0.95
1.13
1.06
Process weight during inlet testing only. Process weight during outlet
testing is similar since inlet and outlet testing were done as close
.together as possible.
Type of chromite product averaged over test run.
CApproximate average of Cr203 in the variety of chromite products.
2-6
-------
drying. More fugitive emissions escaped the dryer from the feed exit
during some of this test than during the tests for chromite ore processing.
Plant personnel indicated that a partial blockage in the ductwork leading
to the baghouse might be decreasing the airflow, rendering the baghouse
less effective for emission capture at the point of material transfer.
The crushed brick had more fine lightweight material than the chromite
ores and apparently overburdened the capture system at this point on
occasion. Testing was not stopped because it was estimated that 98 percent
of rotary dryer emissions were still being captured by the baghouse
despite the increased fugitive emissions. Two-thirds of the way through
the third test, the conveyor belt transporting the recycled brick to the
dryer jammed. The dryer flame was turned down to low, and the rotation
of the drum stopped for about 5 minutes. Testing was stopped temporarily
until the operation was back to normal. Testing continued through the
cleaning cycle of the baghouse. Visible emissions ranging between 0 and
5 percent opacity were noted during cleaning and immediately after the
baghouse came back on-line.
The fourth dryer test on June 26 was done during the processing of
two different chromite ores. The second chromite ore processed needed a
higher temperature flame to ensure sufficient drying. The dryer was
stopped rotating for about 3 minutes during testing because of a
bottleneck downstream in the production process. Because the burner was
firing and chromite ore was in the dryer, testing was continued during
the short time the dryer was not rotating. The baghouse was not cleaned
during this test run.
The rotary dryer baghouse and tunnel kiln operated continuously
without interruption throughout the testing.
2-7
-------
3.0 SUMMARY OF RESULTS
Paniculate matter and particle size distribution tests were conducted at
the rotary dryer exhaust (uncontrolled rotary dryer emissions), the fabric
filter outlet (controlled emissions from rotary dryer and fugitive dust from
materials handling), and the tunnel kiln outlet (uncontrolled tunnel kiln
emissions). The fugitive dust from the five material handling pick up points
entered the duct to the fabric filter downstream of the rotary dryer exhaust
sampling location and was not quantified. The testing schedule for
Harbison-Walker is shown in Table 3.1.
In brief, the uncontrolled emissions from the rotary dryer averaged 28.9
pounds per hour of particulate, 0.0003 pounds per hour of hexavalent chromium
and 3.5 pounds per hour of total chromium. The controlled emissions fron the
fabric filter controlling the rotary dryer and five fugitive dust material
transfer points averaged 5.3 pounds per hour of particulate, 0.0002 pounds per
hour of hexavalent chromium, and 0.64 pounds per hour of total chromiun. The
overall collection efficiency of the fabric filter was greater than 80 percent
by weight for particulate emissions, greater than 35 percent by weight for
hexavalent chromiun, and greater than 80 percent by weight for total chromium.
The collection efficiency values for both hexavalent chromiun and total
chromiun are not accurate representations of the actual collection efficiencies
since more than half the emissions exiting the fabric filter are materials
collected prior to the testing and held in the filter cake. For this test
program, the material on the bags of the fabric filter which was collected
prior to testing was higher in hexavalent chromium concentration, thereby
making the hexavalent chromiun collection efficiency appear lower.
3-1
-------
TABLE 3.1. TESTING SCHEDULE FOR HARBISON-WALKER
I
to
Date
(1985)
6/25
6/26
6/27
6/28
Sample Type
Particul ate
Particle size
Particul ate
Participate
Particul ate
Particle size
Particle size
Particle size
Particul ate
Particul ate
Particle size
Particle size
Particle size
Particul ate
Particle size
Rotary Dryer Exhaust
Run
No.
1
SI
3
5
7
S5
S9
Sll
Test Time
24 h clock
0815-1053
1117-1137
0813-1016
1124-1423
1623-1852
1019-1026
1508-1523
1856-1911
Fabric Fil ter Outlet
Run
No.
2
S4
4
6
8
S6
S8
S10
Test Time
24 h clock
0820-1110
0836-1101
0810-1033
1107-1432
1514-1829
0823-1006
1125-1418
1519-1807
-
Tunnel Kiln Stack
Run
No.
9
10
S12
S13
S14
11
S15
Test Time
24 h clock
1050-1401
1525-1833
0949-1150
1254-1654
1733-1933
0732-1050
0716-0916
-------
The uncontrolled emissions from the tunnel kiln averaged 3.3 pounds per
hour of particulate matter, 0.009 pounds per hour of hexavalent chromium, and
0.14 pounds per hour of total chromium.
The particle size distribution tests for the rotary dryer showed that about
24 percent of the uncontrolled particulate matter emissions were less than
10 ym and 52 percent of the controlled emissions were less than 10 ym in
diameter. Further analysis of the combined particulate samples showed that
about 64 percent of the hexavalent chromium emissions were less than 10 ym and
75 percent of the controlled hexavalent chromium emissions were less than
10 ym. The particle size distribution tests for the tunnel kiln showed that
84 percent, 84 percent, and 93 percent of the emissions by weight were less
than 10 ym for the uncontrolled emissions of particulate, hexavalent chromium
and total chromium, respectively.
In the following sections, the results addressed above and additional
results are presented and discussed in detail according to the emission type
and sampling location. The computer printouts of the emission calculations can
be found in Appendix A. The original field data sheets and the analytical data
are located in Appendix B.
3.1 PARTICULATE MATTER, HEXAVALENT CHROMIUM, AND TOTAL CHROMIUM
Particulate matter tests (EPA Method 5) along with the determination of the
associated flue gas flow rates were conducted at the rotary dryer exhaust,
fabric filter outlet, and tunnel kiln outlet. The particulate matter samples
were initally analyzed using gravimetric techniques to determine the mass of
particulate matter. Then the samples were further analyzed for hexavalent and
total chromiun. Complete descriptions of each sampling location and the
sampling and analytical procedures used are given in Chapter 4.
3-3
-------
3.1.1 Rotary Dryer Exhaust
The emission measurements made at the rotary dryer exhaust represent the
uncontrolled emissions from the rotary dryer.
Flue Gas Conditions and Isokinetic Sampling Rate - A sunmary of the flue
gas conditions at the rotary dryer exhaust is presented in Table 3.2. The
volumetric flow rate for the four runs averaged 6,800 actual cubic meters per
hour (238,000 actual cubic feet per hour) with a flue gas temperature of
151°C (304°F), and a moisture content of 12.6 percent and composition of
16.2 percent oxygen, and 2.7 percent carbon dioxide. The volunetric flow rate
at standard conditions averaged 4,100 dry standard cubic meters per hour
(145,000 dry standard cubic feet per hour). Standard conditions are 20°C
(68°F), 760 mm Hg (29.92 in. Hg) and dry.
The isokinetic sampling rate was within the allowable range for all runs.
Particulate Emissions - The particulate emissions from the rotary dryer
(see Table 3.3) were variable. The variability is believed to be the normal
variability related to the process operations. The particulate emissions
averaged 3170 milligrams per dry standard cubic meter (1.39 grains per dry
standard cubic foot) and 13.1 kilograms per hour (28.9 pounds per hour).
Hexavalent Chromixro Emissions - The hexavalent chromiun emissions were
variable when compared to the corresponding particulate run and averaged 9, 11,
20, and 6 micrograms of hexavalent chromium per gram of particulate matter
emissions. The hexavalent chromiun emissions averaged 0.033 milligrams per dry
standard cubic meter (15 x 10~ grains per dry standard cubic foot) and
0.00014 kilograms per hour (0.00032 pounds per hour). The emission results
were in the quanifiable range, therefore, the variability is probably due to
the variability in the process and materials processed.
3-4
-------
TABLE 3.2. SUMWRY OF FLUE GAS CONDITIONS
i
Ln
Run
No.
Date
(1985)
Test Time
24 h clock
Volumetric Flow Rate
Actual8
acmh
x 106
acfh
x 106
Standard
dscmh
x 106
dscfh
x 106
Stack
Temperature
°C
°f
Moisture
%
°2
co2
CO
%
Isok 1 net Ic
%
Rotary Dryer Exhaust
1
3
5
7
6/25
6/26
6/26
6/26
Average
0815-1053
0813-1016
1124-1423
1623-1852
0.0067
0.0070
0.0063
0.0070
0.0068
0.237
0.248
0.223
0.246
0.238
0.0036
0.0038
0.0049
0.0041
0.0041
0.127
0.135
0.171
0.146
0.145
176
188
84
156
151
349
371
184
313
304
18.2
13.7
5.6
12.8
12.6
15.8
14.6
19.6
15.0
16.2
3.2
3.4
0.8
3.4
2.7
0.0
0.0
0.0
0.0
0.0
109.9
100.3
92.5
100.2
Fabric FlIter Outlet
2
4
6
8
6/25
6/26
6/26
6/26
Average
0820-1 110
0810-1033
1 107-1432
1514-1829
0.0156
0.0146
0.0153
0.0157
0.0153C
0.550
0.514
0.540
0.556
0.540C
0.0126
0.0117
0.0137
0.0135
0.0129C
0.446
0.412
0.484
0.476
0.454C
71
72
49
58
62C
160
161
121
136
144C
5.6
5.7
1.5
3.5
4.lc
19.2
18.0
19.9
19.7
19.2C
1.3
0.9
0.5
0.7
0.85C
0.0
0.0
0.0
0.0
0.0
95.3
97.5
80.7
96.4
Tunnel Kiln Stack
9
10
11
6/27
6/27
6/28
Average
1050-1401
1525-1833
0732-1050
0.0982
0.0975
0.0989
0.0982
3.47
3.44
3.49
3.47
0.0394
0.0389
0.0404
0.0396
1.39
1.38
1.43
1.40
416
416
409
414
781
781
768
777
5.5
5.9
4.8
5.4
16.5
16.7
17.0
16.7
2.3
2.3
2.5
2.4
0.0
0.0
0.0
0.0
93.1
102.4
93.0
Volumetric flow rate In actual meters per hour (acmh) and actual cubic feet per hour (acfh) at stack conditions.
Volumetric flow rate In dry standard cubic meters per hour (dsmh) and dry standard cubic feet per hour (dscfh).
These values show that the ambient dilution air was approximately twice the Inlet flue gas.
-------
TABLE 3.3. SUMM/RY OF P/ftTICULATE, HEXAVALENT CWOMIUM, AND TOTAL CWOMIUM EMISSIONS
Run
No.
Date
(1985)
Parti cul ate
Concentration
mg/dscm
gr/dscf
Mass Emissions
kg/h
Ib/h
Hexavalent Chromium
Concentration
mg/dscm
gr/dscf
x 10~3
Mass Emissions
kg/h
Ib/h
Total Chromium
Concentration
mg/dscm
gr/dscf
Mass Emissions
kg/h
Ib/h
Rotary Dryer Exhaust
1
3
5
7
6/25
6/26
6/26
6/26
Average
2)91
2675
2804
5027
3170
0.957
1.169
1.225
2.197
1.387
7.87
10.22
13.61
20.74
13.11
17.35
22.52
30.00
45.73
28.9
0.0196
0.0298
0.0554
0.0288
0.0334
0.00858
0.01304
0.02421
0.01260
0.01461
0.000071
0.000114
0. 000269
0.000119
0.000143
0.000156
0.000251
0. 000593
0.000262
0.000316
366
551
258
424
400
0.160
0.241
0.113
0.185
0.175
1.31
2.10
1.25
1.75
1.60
2.90
4.64
2.76
3.85
3.54
U)
I
en
Fabric Fl Iter Outlet
2
4
6
8
6/25
6/26
6/26
6/26
Average
163.6
241.9
186.7
157. 1
187a
0.0715
0.1057
0.0816
0.0686
0.0818a
2.064
2.826
2.556
2.115
2.39
4.551
6.230
5.636
4.664
5.27
0.00460
0. 00642
0.00644
0.00717
0.006163
0.00201
0.00281
0.00281
0.00313
0.00269a
0. 000058
0. 000075
0. 000088
0.000097
0.000080
0.000128
0.000165
0.000194
0.000213
0.000175
21.9
29.0
17.7
22.6
22. 8a
0.0096
0.0127
0.0077
0. 0099
0.010a
0.276
0.339
0.242
0.304
0.29
0.609
0.747
0.533
0.670
0.64
Tunnel Kl In Stack
9
10
1 1
6/27
6/27
6/28
Average
46.94
42.70
24.37
38.0
0. 0205
0.0187
0. 01 06
0.0166
1.848
1.663
0.984
1.50
4.073
3.666
2.170
3.30
0.1034
0. 1 1 79
0. 0903
0. 1039
0.0452
0.0515
0.0395
0.0454
0.00407
0.00459
0. 00365
0.00410
0.00897
0.01012
0. 00804
0. 00904
1.09
2.36
1.48
1.64
0.00048
0.00103
0.00065
0.00072
0.043
0.092
0.060
0.065
0.095
0.203
0. 132
0.143
The pollutant concentrations have been biased low by a factor of approximately three due to the fact that the dilution air was about
twice the amount of the Inlet flue gas.
-------
Total Chromiun Emissions - The total chromiun emissions for each test run
(see Table 3.3) were variable when compared to the corresponding particulate
results, and averaged 17, 21, 9, and 8 percent by weight total chromiun. As
for the hexavalent chromiun emissions, the variability is probably due to the
variability in the process. The total chromium emissions averaged 400
milligrams per dry standard cubic meter (0.175 grains per dry standard cubic
foot) and 1.60 kilograms per hour (3.54 pounds per hour).
3.1.2 Fabric Filter Outlet
The fabric filter outlet emissions represent the controlled emissions from
(1) the rotary dryer and (2) five material handling pick up points (fugitive
dust). Following the inlet sampling location were the insersections of an
emergency bypass stack with damper and the five ducts used to control fugitive
emissions from the material handling points. The volume of ambient air
entering the system from the ducts and the damper was almost double the volume
of flue gas fron the rotary dryer. This resulted in a measured outlet
volumetric flow (at standard conditions) of approximately three times that of
the measured inlet flow. This did not cause any problems in sampling; the net
result is, however, pollutant concentrations that are biased low by a factor
of three, a volumetric flow rate biased high by a factor of three, and a true
value for the mass emission rate.
Flue Gas Conditions and Isokinetic Sampling Rate - A summary of flue gas
conditions at the fabric filter outlet is presented in Table 3.2. The
volumetric flow rate (which includes the dilution air) was very consistent for
all runs and averaged 15,300 actual cubic meters per hour (540,000 actual cubic
feet per hour) with a flue gas temperature of 62°C (144°F) and a moisture
content of 4.1 percent and composition of 19.2 percent oxygen and 0.85 percent
3-7
-------
carbon dioxide. The volunetric flow rate (including dilution air) at standard
conditions averaged 12,900 dry standard cubic meters per hour (454,000 dry
standard cubic feet per hour). Standard conditions are 20°C (68°F), 760 mm Hg
(29.92 in. Hg) and dry.
The isokinetic sampling rate was well within the allowable range for three
of the four sample runs. The other run (at 81 percent) was less than 90
percent of the isokinetic sampling rate. This would have very little effect on
the total results since the flue gas parameter values for this run were close
to the averages of the other runs.
Participate Emissions - The particulate emissions from the fabric filter
were fairly consistent (see Table 3.3) and averaged 187 milligrams per dry
standard cubic meter (0.082 grains per dry standard cubic foot) uncorrected for
dilution air and 586 milligrams per dry standard cubic meter (0.26 grains per
dry standard cubic foot) corrected for dilution air. The emission rate
corresponding to both the uncorrected and corrected concentrations was 2.4
kilograms per hour (5.3 pounds per hour).
Hexavalent Chrominu Emissions - The hexavalent chromiun concentrations in
the particulate catch were much higher at the fabric filter outlet than at the
rotary dryer exhaust and averaged 28, 27, 34, and 46 micrograms of hexavalent
chromiun per gram of particulate matter for Runs 2, 4, 6, and 8, respectively.
This higher concentration is a result of the fact that the material previously
collected on the bags which was released during the testing was several times
greater in chromiun concentration than the uncontrolled emissions. The
hexavalent chromiun emissions averaged 0.006 milligrams per dry standard cubic
meter (2.7 x 10~ grains per dry standard cubic foot) and 0.00008 kilograms
per hour (0.00018 pounds per hour).
3-8
-------
Total Chromium Emissions - The total chromiun emissions for each test run
(see Table 3.3) were fairly consistent when compared to the corresponding
particulate results and averaged 13, 12, 9, and 14 percent by weight total
chromiun. The total chromiun emissions averaged 22.8 milligrams per dry
standard cubic meter (0.01 grains per dry standard cubic foot) and 0.29
kilograms per hour (0.64 pounds per hour).
3.1.3 Tunnel Kiln Stack
The tunnel kiln emissions represent the uncontrolled emisisons from the
tunnel kiln. No air pollution control equipment is installed to collect these
emissions, therefore, these emissions are discharged to the atmosphere.
Flue Gas Conditions and Isokinetic Sampling Rate - A summary of the flue
gas conditions from the tunnel kiln is presented in Table 3.2. These were very
consistent from run to run; the volunetric flow rate averaged 98,200 actual
cubic meters per hour (3,470,000 actual cubic feet per hour), with an average
flue gas temperature of 414°C (777°F), a moisture content of 5.4 percent,
and a gas composition of 16.7 percent oxygen and 2.4 percent carbon dioxide.
The volunetric flow rate at standard conditions averaged 39,600 cubic meters
per hour (1,400,000 dry standard cubic feet per hour). Standard conditions are
20°C (68°F) , 760 mm Hg (29.92 in. Hg) , and dry.
The isokinetic sampling rate was within the allowable range for all runs.
Particulate Emissions - The particulate emissions from the tunnel kiln were
fairly consistent from run to run (see Table 3.3), and averaged 38 milligrams
per dry standard cubic meter (0.017 grains per dry standard cubic foot) and 1.5
kilograms per hour (3.3 pounds per hour).
Hexavalent Chromium Emissions - The hexavalent chromiun concentration in
the particulate catch was variable from run to run, averaging 2200, 2800, and
3700 micrograms of hexavalent chromium per gram of particulate emissions. The
3-9
-------
hexavalent chrcmiun emissions averaged 0.10 milligrams per dry standard cubic
meter (0.045 x 10~ grains per dry standard cubic foot) and 0.004 kilograms per
hour (0.009 pounds per hour).
Total Chromium Emissions - The total chrcmiun emissions for each test run
(see Table 3.3) were variable when compared to the corresponding particulate
results, and averaged 2.3, 5.5, and 6.1 percent by weight total chrcmiun. The
total chromium emissions averaged 1.64 milligrams per dry standard cubic meter
(0.00072 grains per dry standard cubic foot) and 0.065 kilograms per hour
(0.143 pounds per hour).
3.2 PARTICLE SIZE DISTRIBUTION
Particle sizing runs were conducted at all locations tested for particulate
emissions. The first of the four runs at each location was conducted at a
point of average velocity. The other runs at the same location were conducted
at a point with a velocity similar to the first run. This sampling procedure
was followed to ensure that the particle cut-size for all four runs would be
the same on corresponding impactor stages.
The total mass of particulate matter collected on each stage was determined
using a gravimetric technique. Stages were then combined in such a manner as
to obtain a quantifiable amount of hexavalent and total chromiun and determine
the particle size distribution of these chrcmiun species. The particle size
distribution results are presented in Table 3.4 and the corresponding
calculations and graphs can be found in Appendix A. The particle size
distribution for the uncontrolled rotary dryer emissions showed that
approximately 24 percent of the particulate, 64 percent of the hexavalent
chromiun, and 19 percent of the total chromium, by weight, were less than 10 ym
in diameter. The controlled emissions from the rotary dryer and fugitive dust
sources showed that approximately 52 percent of the particulate, 75 percent of
the hexavalent chromiun, and 42 percent of total chromiun, by weight, were less
than 10 ym in diameter. 3-10
-------
TABLE 3.4. SUMM/R Y OF P/WTICLE SIZE DISTRIBUTION
OJ
i
Run
No.
Date
(1985)
Test Time
24 h clock
Part Icul ate
wt. less than size, %
1 y m | 5 ym
10 y m
Hexavalent Chromium
wt. less than size, %
1 ym | 5 ym
10 ym
Total Chromium
wt . less than si ze, %
liym
5 ym
10 ym
Rotary Dryer Exhaust
SI
S5
S9
S11
6/25
6/26
6/26
6/26
1117-1 137
1019-1026
1508-1523
1856-191 1
Average
0.6
3
0.5
0.8
1.2
9
18
14
12
13
17
30
30
20
24
3.5*
39*
64*
0.8*
9*
19*
Fabric F!Iter Outlet
S4
S6
SB
SIO
6/25
6/26
6/26
6/26
0836-1 101
0823-1006
1125-1418
1519-1807
Average
3.5
5.5
0.5
3.5
3.2
33
39
19
33
31
54
63
39
52
52
8*
52*
75*
1.7*
23*
42*
Tunnel Kl I n Stack
S12
S13
S14
S15
6/27
6/27
6/27
6/28
0949-1 150
1254-1654
1733-1933
0716-0916
Average
72
71
72
70
71
79
78
80
74
78
87
85
85
81
84
71*
81*
84*
84*
91*
93*
'Values calculated from composites of all runs.
-------
The particle size distributions for the tunnel kiln emissions are also
presented in Table 3.4 and show that the majority of the particulate,
hexavalent and total chromium emissions were less than 1 vm in diameter.
The particle size distribution samples contained a smaller amount of
hexavalent and total chromium than the particulate samples and were therefore
subject to a higher degree of analytical error. The results are, however,
believed to be representative and show a hexavalent chromium particle size
dependency toward the smaller particle sizes; results from two of the three
locations sampled showed that the majority of the hexavalent chromiun was
present in the particles less than 5 ym in diameter, which is to be expected.
3.3 EMISSIONS IN UNITS OF PROCESS RATE AND CONTROL EQUIPMENT COLLECTION
EFFICIENCY
The emission rates in units of the process rate are expressed in terms of
grams of pollutant emissions per tons of chrcmite product processed and are
presented in Table 3.5
To determine the collection efficiency of the fabric filter, the un-
controlled and controlled emissions measured were used; no actual measurements
were made on the mass removal rates from the collector.
Emissions from the rotary dryer are actually controlled by both a cyclone
and a fabric filter; fugitive emissions from the material transfer pick up
points are also controlled by the same fabric filter. However, during the test
program, the open bottom cyclone showed no collected material. When the rotary
dryer exhaust emissions were to calculate collection efficiency, that of the
fabric filter (see Table 3.5) averaged about 80 percent by weight for particu-
late matter, 35 percent by weight for hexavalent chromiun, and 80 percent by
weight for total chromiun. The actual efficiency would be somewhat higher
depending upon the amount of the unquantified fugitive particulate emissions.
3-12
-------
TABLE 3.5. SUMNVRY OF EMISS ION RATES IN UNITS OF PROCESS RATE AND EFFICIENCY
Run
No.
Process Rate
tons/h
Uncontrolled Emissions
part Icul ate
g/ton
hexavalent
chromium
g/ton x IO"3
total
chromium
g/ton
Controlled Emissions
partlcul ate
g/ton
hexavalent
chromium
g/ton x IO"3
total
chromium
g/ton
Collection Efficiency
partlcul ate
%
hexavalent
chromium
%
total
chromium
%
Rotary Dryer Fabric Filter (+ Cyclone)
1,2
3,4
5,6
7,8
19.28
13.95
21.26
16.83
Average
408
732
641
1232
753
3.68
8.18
12.65
7.08
7.90
68.0
151
58.8
104
95.5
107
203
120
126
139
3.00
5.38
4.15
5.76
4.57
14.3
24.3
11.4
18. 1
17.0
>73.8
>72.3
>81.3
>89.8
>79.3
>18.5
>34.2
>67.2
>18.6
>34.6
>79.0
>83.9
>80.6
>82.6
>81.5
U)
I Tunnel Kl In
\->
U)
9
10
1 1
3.68
4.30
4.02
Average
502
388
246
379
1108
1068
910
1029
11.7
21.4
14.9
16.0
N/A
N/A
N/A
N/A
N/A
N/A
-------
As previously noted, an accurate collection efficiency cannot be determined
for hexavalent chromium and total chromium for a fabric filter if the materials
typically collected are variable with respect to concentration of these
pollutants. Based on other studies, it appears that about 50 to 80 percent of
the emissions from the fabric filter consist of materials previously collected
on the bags. The collection efficiency for hexavalent chromium is apparently
lower due to the fact that material previously collected by the fabric filter
was much higher in hexavalent chroniun content. The concentration of
uncontrolled hexavalent chromium emissions averaged about 11 yg of Cr per
gram of particulate. The material collected by the fabric filter and then
discharged through the hoppers averaged 76 yg of Cr per gram of
particulate. The emissions controlled by the fabric filter averaged 34 yg of
+6
Cr per gram of particulate.
3.4 SUMMARY OF ANALYTICAL RESULTS FOR HEXAVALENT AND TOTAL CHROMIUM
The summary of analytical results for hexavalent chromium and total
chromiun for all samples collected is presented in Table 3.6. The analytical
data sheets are contained in Appendix B. The results shown in Table 3.6 for
hexavalent and total chromiun are the results obtained by the EPA tentative
method for "Determination of Hexavalent Chromiun Emissions frcm Stationary
Sources" and the "EPA Protocol for Emissions Sampling for Both Hexavalent and
Total Chromiun" (see Appendix D). When, for total chromiun analysis, the table
indicates that the sample "residue" was analyzed, then the values presented for
total chromiun content are a sun of (1) the hexavalent chromiun in the sample
filtrate from the extraction of the sample and (2) the chromiun in the residue
from the extraction as measured by Neutron Activation Analysis. When the table
indicates that the "total" sample was analyzed, then the values presented for
3-14
-------
TABLE 3.6. SUMM/RY OF ANALYTICAL RESULTS FCR HEXAVALENT AND TOTAL CHROMIUM
Run
No.
Samp 1 e
Type
Samp 1 e
No.
Analyzed
Amount of
Sample
Analyzed
Hexavalent Chromium
Results
ii g
Concentration
ug/g
Amount of
Sampl e
Anal yzed
Total Chromium
Results
mg
Concentration
mg/g
Rotary Dryer Exhaust
1
1
3
5
7
51,5,9,11
SI, 5, 9,11
SI, 5, 9, 11
Part leu late Front Half
Impl nger Contents
Participate Front Half
Part Icul ate Front Half
Partlculate Front Half
Particle Size, Large
Particle SI ze. Medium
Partlc le SI ze, Smal 1
C-296
C-307
C-297
C-298
C-299
C-326
C-327
C-328
2989.1 mg
Total
3540.9 mg
3806.8 mg
7181 .6 mg
678.4 mg
257.9 mg
105.4 mg
26.8
< 0.2
39.5
75.2
41.2
3.8
5.0
3.7
8.97
negl Iglble
11.2
19.8
5.74
5.60
19.4
35.1
Residue
175 ml
Residue
Residue
Residue
Residue
Residue
Residue
499
0
729
350
605
103
19.4
9. 12
166.9
0
205.9
91.9
84.2
151.8
75.2
86.5
Fabric Fl Iter Outlet
2
2
4
6
8
54,6,8,10
54,6,8, 10
54,6,8,10
Partlculate Front Half
Impl nger Contents
Partlculate Front Half
Partlcul ate Front Hal f
Partlculate Front Hal f
Particle SIza, Large
Partlc le SI za. Medium
Particle SI ze, Smal 1
C-300
C-31I
C-301
C-302
C-303
C-329
C-330
C-331
598.2 mg
Total
1009.4 mg
672.6 mg
438.1 mg
191 .6 mg
191 .0 mg
1 09 . 2 mg
16.8
< 0.2
26.8
23.2
20.0
4.2
7.0
8.1
28.1
neg 1 Iglble
26.6
34.5
45.7
21.9
36.6
74.2
Residue
175 ml
Residue
Residue
Residue
Residue
Residue
Residue
80
0
121
63.6
62.9
25.4
14.8
8.38
133.7
0
119.9
94.6
143.6
132.6
77.5
76.7
Tunnel Kl In Stack
9
9
10
1 1
S12, 13, 14, 15
512,13,14,15
512,13,14,15
Partlculate Front Half
Impl nger Contents
Partlculate Front Hal f
Partlculate Front Hal f
Particle Size, Large
Particle Size, Medium
Particle SI ze, Smal 1
C-304
C-315
C-305
C-306
C-332
C-333
C-334
54.5 mg
Total
53.9 mg
29.0 mg
19.4 mg
8.0 mg
75.2 mg
120
< 0.2
148.8
107.5
1.7
1.7
9.7
2202
negl Iglble
2761
3707
87.6
212
129
Residue
175 ml
Residue
Residue
Residue
Residue
Residue
1.27
0
2.98
1.76
0.27
0.49
4.31
23.3
0
55.3
60.7
13.9
61.3
57.3
Grab Samples
1
2
3
4
1
2
3
Ore Feed
Ore Feed
Ore Feed
Ore Feed
Fabric Filter Hopper
Fabric Filter Hopper
Fabric Filter Hopper
C-319
C-320
C-321
C-322
C-323
C-324
C-325
4.4
1.8
19.2
1.4
75.3
87.0
65.7
1 12.8 g
103.3 g
1 19.6 g
115.9 g
139.9 g
119.2 g
116.1 g
33.9
35.8
14.0
47.0
20.5
18.8
18.0
300.5
346.6
117.1
405.5
146.5
157.7
155.0
Blank Samples
Particle SI ze Fl Iters
Dlstl 1 led H20 Blank
C-336
C-318
~
1.0
<0.2
negl Iglble
Residue
100 ml
0
0
-------
total chromium content are from the direct analysis of the total sample for
total chromium by Neutron Activation Analysis. A table showing the total
chromium calculations for each sample can be found at the end of Appendix A of
th is repo rt.
The hexavalent chromiun concentration was somewhat variable for most
sampling locations. The variability of results for the particle size
distribution tests emissions may reflect some analytical imprecision due to the
small amount of hexavalent chromium analyzed. The other samples had a
sufficient quantity of participate, hexavalent chromium and total chrcmiun, and
should therefore be representative of the sample analyzed. Overall, the goals
of obtaining quantifiable emissions were obtained.
Quality assurance audit samples were analyzed for both the hexavalent and
total chromium methods. As shown in Table 3.7, no bias was present and the
results are considered acceptable.
3.5 VISIBLE EMISSIONS OBSERVATION DATA
Visible emission observations were performed at the fabric filter stack
exit by the EMB Task Manager (see Table 3.8 and Appendix C). Readings were
made for two or three 6 to 7 minute sets for Runs 2, 4, and 6. Under normal
operating conditions the opacity at the fabric filter stack was observed to be
0 percent; however, during the manual cleaning cycles used between product
runs, the opacity levels increased to an average of 3 to 10 percent with a
maximun range of 30 percent.
3-16
-------
TABLE 3.7. SUMWRY OF ANALYTICAL RESULTS FCR HEXAVALENT AND TOTAL OflOMIUM QUALITY ASSIRANCE SANPLES
Run
No.
Samp 1 e
Type
Sampl e
No.
True
Value
Hexavalent Chromium
Results
pg/ml
%
Dev.
Total Chromium
Resul ts
ug
%
Dev.
Qual Ity Assurance Samples
Qual Ity Assurance
Qual Ity Assurance
Qual Ity Assurance
Qual Ity Assurance
C-335
QA-19
QA-20
QA-21
50 yg/ml Cr*6
50 ug Cr
100 p g Cr
200 vi g Cr
50.8
+ 1.6
52.30
97.22
200.1
+4.6
-2.8
+0.05
I
I-1
-J
-------
TABLE 3.8. SUMMARY OF VISIBLE EMISSIONS DATA FOR FABRIC FILTER
HARBISON-WALKER
Date
Time
Range
Avg. % Opacity
Average Over
All Sets
Run Nos. 1 and 2
6/25
0906-0913
1007-1014
0
0-15
0
5
2.50
Run Nos. 3 and 4
6/26
0847-0854
0952-0959
1024-1033
0
0
0-30
0
0
10.14
4.08
Run Nos. 5 and 6
6/26
1321-1328
1425-1434
1630-1640
0
0-5
0-10
0
0.9
3.2
1.60
3-18
-------
4.0 SAMPLING LOCATIONS AND TEST METHODS
This section describes the sampling locations and test methods vised to
characterize emissions from the rotary dryer and tunnel kiln at Harbison-Walker
Refractories in Baltimore, Maryland. A total of five sampling locations were
used in the emission testing program. At three sampling locations, emissions
testing was conducted for particulate matter, total chromiun content, hexa-
valent chromiun content, and particle size distribution and chromium
distribution with respect to particle size. At the fourth and fifth sampling
locations, grab samples of the dust collected by the fabric filter and the
rotary dryer ore feed were taken for hexavalent and total chromiun analysis.
The relative positions and the type of testing conducted at each location are
shown in the simplified process flow diagram (see Figure 4-1) and accompanying
Table 4.1. The subsections which follow further describe each sampling
location and applicable test methods.
4.1 ROTARY DRYER EXHAUST (SAMPLING LOCATION A)
Particulate matter, hexavalent chromiun, total chromiun, particle size
distribution, and chromiun distribution with respect to particle size
distribution were measured in the rotary dryer exhaust duct. A schematic of
this sampling location is shown in Figure 4-2. A 3.5 by 17 inch slot for
sampling was cut in one side of the 17-inch square, vertical duct. This
sampling slot was located 67 inches {3,9 equivalent duct diameters) downstream
of a bend in the duct and 17 inches (1 equivalent duct diameter) upstream of
another bend in the duct to the fabric filter.
4-1
-------
To atmosphere
Sampling Location D
I. D. Fan
Formed
Refractory
Brick
Tunnel
Kiln
Sampling Location B
Sampling Location A
Sampl ing
Location E
Ore Feed
To atmosphere
I. D. Fan
Fabric Filter
cyclone
Sampling
Location C
*^ Hopper Dust
Fugitives
from
Materials
Handling
(5 Ducts)
Emergency Bypass Duct
Damper
Rotary
Dryer
Figure 4-1. Process Air Flow Schematic of Rotary Dryer and Tunnel Kiln.
4-2
-------
TABLE 4.1. SAMPLING PLAN FOR HARBISON-WALKER
Sample Type
Samp!ing
Locations
Number
of Samples
Methods
Particulate matter
Hexavalent chromium
Total chromium
Particle size distribution
Hexavalent and total chromium
distribution by particle size
Hexavalent chromium, total
chromium
A, B, D
A, B, D
A, B, D
A, B, D
A, B, D
C, E
4 at A & B
3 at D
4 at A & B
3 at D
4 at A & B
3 at D
4 at A & D
5 at B
3 grab at C
4 grab at E
EPA Method 5
EPA 5 using Tentative
EPA Method for Hexavalent
Chromium
EPA 5 using EPA Protocol
for Total Chromium
Impactor (Andersen)
Impactor using Tentative
EPA Method for Hexavalent
Chromium and EPA Protocol
for Total Chromium
Tentative EPA Method for
Hexavalent Chromium,
EPA Protocol for Total
Chromium
4-3
-------
17"
17*
4
t
I
I
T
I
c
B
A
SECTION F-F
TRAVERSE POINTS
4 AXES
4 POINTS/AXIS
16 TOTAL POINTS
TO
FABRIC FILTER
EMERGENCY
EXHAUST
r
67"
" ->
t
FROM
ROTARY DRYER
SIDE ELEVATION
DAMPER
(CLOSED)
EMERGENCY
EXHAUST
t
A B C D
4-17"
t
SLOT
(3.5" x 17")
FRONT ELEVATION
FIGURE 4-2. ROTARY DRYER EXHAUST DUCT (SAMPLING LOCATION A)
4-4
-------
For the Method 5 testing (used for particulate matter, hexavalent chromium,
and total chromium determinations), a total of 16 points (4 axes, 4 points per
axis), as per Method 1, were sampled. Each point was sampled for 7.5 minutes
for a total sampling time of 120 minutes per run (one exception was Run 5 where
a process upset resulted in several points not being sampled and a shorter
total sampling time of 105 minutes).
The particle size testing (including hexavalent and total chromium
distribution by particle size), was conducted at a single point. Runs S1, S5,
S9, and S11 were 20, 7.:52, 15, and 15 minutes in duration, respectively.
4.2 FABRIC FILTER OUTLET (SAMPLING LOCATION B)
Particulate matter, hexavalent chromium, total chromium, particle size
distribution, and hexavalent and total chromium distribution with respect to
particle size were measured at the fabric filter outlet. A schematic of this
sampling location is shown in Figure 4-3. A 3.5 by 22 inch slot for sampling
was cut in the long side of the 19 by 22 inch rectangular, vertical duct. This
slot was located 110 inches (5.4 equivalent duct diamteters) downstream of the
fabric filter I. D. fan and 100 inches (4.9 equivalent duct diameters) upstream
of the fabric filter stack exit.
For the EPA Method 5 sampling (used for particulate matter, hexavalent
chromium, and total chromium determinations), a total of 16 points (4 axes,
4 points per axis), as per Method 1, were sampled. For the first run each
point was sampled for 15 minutes, however, a process upset prevented sampling
all the points resulting in a total run time of 170 minutes. In addition, the
heavy particulate catch from this run suggested a shorter sampling time, so for
the remaining runs each point was sampled for 7.5 minutes. The total sampling
4-5
-------
TRAVERSE POINTS
4 AXES
4 POINTS/AXIS
16 TOTAL POINTS
9"
T ^ r
*
*
*
«
«
«
«
A B C D
SECTION T-T
SLOT (3.5" x 22")
22">
t
AB CD
- 100'
-110"
ELEVATION VIEW
FROM I.D. FAN
FIGURE 4-3. FABRIC FILTER OUTLET STACK (SAMPLING LOCATION B>
4-6
-------
times for these runs were 136, 121, and 120 minutes. For the 136 minute run,
the EPA Task Manager requested that sampling continue beyond the 120 minute
mark.
The particle size testing (including hexavalent and total chromium
distribution by particle size) was conducted at a single point. Runs S4, S6,
SB, and S10 were 100 minutes in duration.
Visible emissions observations of the effluent from the fabric filter stack
were conducted by the EMB Task Manager for several 6 (or 7) minute sets during
Runs 2, 4, and 6.
4.3 FABRIC FILTER DUST HOPPER (SAMPLING LOCATION C)
Grab samples representative of the material collected by the fabric filter
were taken from the fabric filter hopper during each set of test runs. Each
grab sample was a composite of material from each of the three hopper sections
of the filter and each was analyzed for hexavalent chromium and total chromium
content.
4.4 TUNNEL KILN STACK (SAMPLING LOCATION D)
Particulate matter, hexavalent chromium, total chromium, particle size
distribution, and hexavalent and total chromium distribution with respect to
particle size were measured at the tunnel kiln stack.
A schematic of this sampling location is shown in Figure 4-4. Two sampling
ports were installed at a 90° angle in the 42 inch diameter vertical stack.
These ports were located 96 inches (2.3 duct diameters) downstream of the
induced draft fan and 380 inches (9.0 duct diameters) upstream of the stack
exit.
4-7
-------
TRAVERSE POINTS
2 AXES
12 POINTS/AXIS
24 TOTAL POINTS
42" DIA.
SECTION X-X
42"
t
A
O-
t
- 380"
B
'"I
96" X
ELEVATION VIEW
FROM KILN
FIGURE 4-4. TUNNEL KILN STACK (SAMPLING LOCATION D)
4-8
-------
For the Method 5 testing (used for particulate matter, hexavalent chromium,
and total chromium determinations), a total of 24 points (2 axes, 12 points per
axis), as per Method 1, were sampled. Each point was sampled for 5 minutes for
a total sampling time of 120 minutes per run.
The particle size testing (including hexavalent and total chromium
distribution by particle size) was conducted at a single point. Runs S12, S14,
and S15 were 120 minutes in duration, Run S13 was 240 minutes in duration.
4.5 ROTARY DRYER FEED (SAMPLING LOCATION E)
Grab samples representative of the feed material (ore) entering the rotary
dryer were taken frcm the feed belt as it went to the dryer. Samples were
taken at the beginning, middle, and end (if possible because of process
operations) of each particulate test series. The samples from each test series
were combined into a single sample which was analyzed for hexavalent chromium
and total chromim content.
4.6 VELOCITY AND GAS TEMPERATURE
A type S pitot tube and an inclined draft gauge manometer or two differen-
tial pressure gauges in-paralled were used to measure the gas velocity pressure
(A?)- Velocity pressures were measured at each sampling point across the duct
to determine an average value according to the procedures outlined in Method 2
of the Federal Register.* The temperature at each sampling point was measured
using a thermocouple and digital readout.
4.7 MOLECULAR WEIGHT
Flue gas composition was determined utilizing procedures described in
Method 3 of the Federal Register.* A bag sample was collected during each
particulate test run. The bag contents were analyzed using an Orsat Gas
Analyzer.
*40 CFR 60, Appendix A, Reference Methods 2, 3, and 5, July 1, 1980.
4-9
-------
4.8 PAKTICULATE MATTER
Method 5, as described in the Federal Register,* was used to measure
particulate grain loading at locations A, B, and D. All tests were conducted
isokinetically by traversing the cross-sectional area of the stack and
regulating the sample flow rate relative to the flue gas flow rate as measured
by the pitot tube attached to the sample probe. A sampling train consisting of
a heated, glass-lined probe, a heated 87 mm (3.4 inches) diameter glass fiber
filter (Gelman A/E), and a series of Greenburg-Smith impingers was employed for
each test. An acetone rinse of the nozzle, probe, and filter holder portions
of the sample train was made at the end of each test. The acetone rinse and
the particulate caught on the filter media were dried at room temperature,
desiccated to a constant weight, and weighed on an analytical balance. Total
filterable particulate matter was determined by adding these two values. See
Appendix D for detailed sampling procedures.
4.9 PARTICLE SIZE DISTRIBUTION
Particle size samples were obtained using Andersen Mark III Cascade
Impactors. These in-stack, multistage cascade impactors have a total of eight
stages, followed by a back-up filter stage and particle size cut-offs ranging
nominally from 0.5 to 15 microns. Substrates were 64 mm diameter glass fiber
filters. A constant sampling rate was maintained through the test period.
Sampling rates were set for isokinetic sampling as long as the sampling rate
did not exceed the recommended flow rate for the impactor. See Appendix D for
detailed sampling procedures.
Four impactor runs each were conducted at the rotary dryer exhaust, the
fabric filter outlet and the tunnel kiln stack. At the locations sampled, a
* 40 CFR 60, Appendix A, Reference Methods 2, 3, and 5, July 1, 1980.
4-10
-------
single point was sampled. With the exception of selection of the sampling
point locations, the procedures used followed those recommended in the
"Procedures Manual for Inhalable Particulate Sampler Operation" developed for
EPA by the Southern Research Institute.*
4.10 HEXAVALENT CHROMIUM CONTENT
Hexavalent chromiun content was determined utilizing procedures described
in the tentative EPA Method "Determination of Hexavalent Chromiun Emissions
from Stationary Sources" (see Appendix D). The Method 5 filter catch collected
and weighed for each Method 5 run was taken and analyzed for hexavalent
chromiun content using this method. If was also used to determine the
hexavalent chromiun content of representative portions of the fabric filter
hopper dust and rotary dryer ore feed samples.
4.11 TOTAL CHROMIUM CONTENT
Total chromiun content was determined using procedures described in the
" EMB Prototcol for Sample Preparation and Emission Calculation of Field Samples
for Total Chromiun" in combination with Neutron Activation Analysis (NAA) (see
Appendix D). Samples collected during Method 5 runs and first submitted to
analysis for hexavalent chromium were then analyzed for total chromiun using
this method. The total chromiun content of the fabric filter hopper dust and
the rotary dryer ore feed samples were also determined using these procedures
using a representative portion of the sample.
Prepared for EPA under Contract No. 68-02-3118, November 1979.
4-11
-------
5.0 QUALITY ASSURANCE
Because the end product of testing is to produce representative emission
results, quality assurance is one of the main facets of stack sampling.
Quality assurance guidelines provide the detailed procedures and actions
necessary for defining and producing acceptable data. Two such documents were
used in this test program to ensure the collection of acceptable data and to
provide a definition of unacceptable data. These documents are: the EPA
Quality Assurance Handbook Volume III, EPA-600/4-77-OZ7 and Entropy's "Quality
Assurance Program Plan" which has been approved by the U. S. EPA, EMB.
Relative to this test program, the following steps were taken to ensure
that the testing and analytical procedures produce quality data.
« Calibration of field sampling equipment. (Appendix E describes
calibration guidelines in more detail.)
Checks of train configuration and on calculations.
On-site quality assurance checks such as sampling train, pitot
tube, and Orsat line leak checks, and quality assurance checks of
all test equipment prior to use.
Use of designated analytical equipment and sampling reagents.
In addition to the pre- and post-test calibrations, a field audit was
performed on the meter boxes used for sampling. Entropy used the procedures
described in the December 14, 1983 Federal Register (48FR55670). In addition,
the analytical balance used for filter weighing was audited with Class "S"
weights.
As a check on the reliability of the method used to analyze the filters
for particle size tests, sets of filters that had been preweighed in the lab
were resubmitted for replicate analysis. Table 5.1 summarizes these results.
5-1
-------
TABLE 5.1. PARTICLE SIZE BLANK FILTER AND REACTIVITY FILTER ANALYSIS
Sample type
Particle size
blank run filters
B472
A472
B473
A473
B474
A474
B475
A475
SF192
Particle size
reactivity run
filters
B525
A525
B526
A526
B527
A527
B528
A528
SF205
Original tare
weight, mg
166.68
145.49
166.12
146.85
164.70
147.69
164.46
148.40
275.34
164.18
146.00
164.06
147.62
163.31
146.17
165.68
145.91
274.34
Bl ank weight,
mg
166.69
145.53
166.22
146.85
164.71
147.70
164.49
148.42
275.34
164.18
146.02
164.09
147.66
163.32
146.17
165.68
145.92
274.30
Net weight,
mg
0.01
0.04
0.10
0.00
0.01
0.01
0.03
0.02
0.00
0.00
0.02
0.03
0.04
0.01
0.00
0.00
0.01
-0.04
5-2
-------
Audit solutions prepared by the EPA were used to check the analytical
procedures of the laboratories conducting the hexavalent and total chrcmivm
analyses. Table 5.2 presents the results of these analytical audits. The
audit tests show that the analytical techniques were good.
The sampling equipment, reagents, and analytical procedures for this test
series were in compliance with all necessary guidelines set forth for accurate
test results as described in Volune III of the Quality Assurance Handbook.
5-3
-------
TABLE 5.2. AUDIT REPORT CHROMIUM ANALYSIS
Plant:
Task No.:
Date samples received:
Sample analyzed by:
Reviewed by:
Date analyzed:
Date of review:
Sample
Number
C-335"
3/^/5
5/f'2&
QA-2/
yg/ml
Cr+b or Cr
SO Mq InJLC^
So ^ Cr
J J
/CO ^ Cs
2.QDP1 t
\/£S
N&5
X6S
Audit
Value
5^>-6
5"^.3O
17. ZZ.
2.00. \
Relative
error, %
+ I.G>
+ +.U
-j.e
^o.osr
5-4
------- |