United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EMB Report 79-SOD-1
May 1979
Air
v>EPA Sodium Carbonate
Emission Test Report
Texasgulf
Granger, Wyoming
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
EMISSION MEASUREMENT BRANCH
MAIL DROP 13
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
DRAFT REPORT
EMISSION TEST PROGRAM: SODIUM CARBONATE MANUFACTURING PLANT
CONDUCTED AT
TEXASGULF, INC.
GRANGER, WYOMING 82934
CONTRACT NUMBER 68-02-2819
TASK ASSIGNMENT 14
PROJECT NUMBER 79-SOD-l
YORK PROJECT NUMBER 1-9517-14
AUGUST 1, 1979
-------
TABLE OF CONTENTS
Page
LIST OF FIGURES -i-
LIST OF TABLES -ii-
1.0 INTRODUCTION 1
2.0 SUMMARY 1
3.0 SAMPLING METHODS 29
3.1 Test Port Locations and Sampling 29
Point Determination
3.2 Gas Velocity x 30
3.3 Gas Composition 30
3.4 Particulate 30
3.5 Organics 37
3.6 Sulfur Dioxide 38
3.7 Particle Size Distribution 38
3.8 Visible Emissions 40
4.0 ANALYTICAL METHODS 43
4.1 Particulate 43
4.2 Sulfur Dioxide 43
4.3 Particle Size Distribution 43
5.0 APPENDIX 46
5.1 Complete Computer Data Printouts 47
5.2 Field Data Sheets 84
5.3 Laboratory Data 133
5.4 Calibration Sheets 143
5.5 Sample Calculations 156
-------
LIST OF FIGURES
Figure 1-1
Figure 1-2
Figure 2-1
Figure 2-2
Figure 2-3
Figure 2-4
Figure 4-1
Figure 4-2
Figure 4-3
Figure 4-4
Figure 4-5
Figure 4-6
Figure 4-7
Figure 4-8
Product Dryer Process Diagram
Trona Dryer Process Diagram
Andersen Particle Size Analysis
Product Dryer Inlet
Bahco Particle Size Analysis
Product Dryer Inlet
Andersen Particle Size Analysis
Trona Dryer Inlet
Bahco Particle Size Analysis
Trona Dryer Inlet
Product Dryer Scrubber Inlet
Sampling Point Locations
Product Dryer Scrubber Outlet
Sampling Point Locations
Trona Dryer Cyclone Inlet
Sampling Point Locations
Trona Dryer Electrostatic Precipi-
tator Outlet Sampling Locations
Modified Particulate Sampling
Train
S02 Sampling Train
Andersen Sampling Train
Andersen Stack Sampler
Page
2
3
25
26
27
28
31
32
33
34
35
39
41
42
-------
-11-
LIST OF TABLES
Page
Table 2-1
Table 2-2
Table 2-3
Table 2-4
Table 2-5
Table 2-6
Table 2-7
Table 2-8
Table 2-9
Table 2-10
Table 2-11
Table 2-12
Summary of Emission Test Results -
Product Dryer - English Units
Summary of Emission Test Results -
Product Dryer - Metric Units
Summary of Emission Test Results -
Trona Dryer - English Units
Summary of Emission Test Results -
Trona Dryer - Metric Units
Summary of S02 and Organic Emission
Test Results -
Trona Dryer - English Units
Summary of Opacity Observations -
Product Dryer
Summary of Opacity Observations -
Product Dryer
Summary of Opacity Observations -
Product Dryer
Summary of Opacity Observations -
Product Dryer
Summary of Opacity Observations -
Trona Dryer
Summary of Opacity Observations -
Trona Dryer
Summary of Opacity Observations -
Trona Dryer
10
11
' 12
13
14
15
16
17
18
19
20
21
-------
-in-
Table 2-13
Table 2-14
Table 2-15
Summary of Opacity Observations
Trona Dryer
Summary of Opacity Observations
Trona Dryer
Summary of Opacity Observations
Trona Dryer
Page
22
23
24
-------
-1-
1.0 INTRODUCTION
York Research Corporation (YRC) under contract 68-02-2819 was
requested by the United States Environmental Protection Agency
(USEPA) to perform an emission test program at a sodium carbonate
manufacturing plant. The test program was conducted at the
Texasgulf Corporation located in Granger, Wyoming. Sampling
was performed from May 21, 1979 to May 24, 1979. The sampling
locations included:
• Product Dryer Inlet
• Product Dryer Outlet
• Trona Calciner Inlet
• Trona Calciner Outlet
Figures 1-1 and 1-2 show the sampling locations in the process.
Samples were collected for solid particulate organics, sulfur
dioxide and particle size distribution. The objective of the
test program was to determine the emission levels of controlled
sodium carbonate production facilities to support planned source
emission standards in the sodium carbonate industry.
The test team consisted of the following individuals:
Name Affiliation Title
Dennis Holzschuh USEPA Technical Manager
Roger A. Kniskern YRC Project Manager
William J. Cesareo YRC Project Director
John Breger YRC Test Engineer
Keith Synnestvedt YRC Test Technician
Albert Burton YRC Test Technician
Laurie Behr YRC Test Technician
Joseph Kuntz YRC Chemical Technician
Bruce Wuebber YRC Test Technician
2.0 SUMMARY AND DISCUSSION OF TEST RESULTS
Tables 2-1 through 2-19 and Figures 2-1 through 2-4 summarize
the results of the emission test program. These tables present
the results of tests for the following parameters:
-------
PRODUCT
DRYER
SCRUBBER
INLET i
TEST j
POINT !
OUTLET
TEST
POINT
i
N)
I
PRODUCT DRYER PROCESS DIAGRAM
FIGURE 1-1
-------
TRONA
DRYER
CYCLONE
PRECOLLECTOR
INLET
TEST
POINT
ELECTROSTATIC
PRECIPITATOR
OUTLET
TEST
POINT
TRONA DRYER PROCESS DIAGRAM
FIGURE 1-2
-------
-4-
• Particulate
• Particle Size Distribution
• Visible Emissions
• Organics
• Sulfur Dioxide
Sample analysis for all parameters were performed at YRC lab-
oratories in Stamford, Connecticut or Denver, Colorado with the
exception of organics. A portable gas chromatograph with a
flame ionization detector was set up at the test site for
analysis of organic samples.
The following table details the isokinetic ratio results of the
particulate tests conducted at the product dryer.
Test No. 1
Scrubber Inlet 146.5%
Scrubber Outlet 111.0%
2
149.1%
94.2%
3
120.9%
115.8%
Only outlet test 2 meets the isokinetic requirement (100 + 10%)
of the reference method.
In the majority of sampling situations encountered, the stack
gas moisture can be estimated with an adequate degree of confi-
dence so that isokinetic sampling can be easily achieved. This
is the case when the moisture content of the gas stream is less
than 30 percent by volume and is relatively constant. However,
in those cases where the moisture content is greater than 30
percent, or when the moisture content varies significantly,
such as is the case for the product dryer, a single average
value is usually inadequate. Since the moisture content is not
a linear variable in the isokinetic equation, the estimation
error that is tolerable decreases as the absolute value of the
moisture content increases or varies.
-------
-5-
\
The moisture content of the flue gas at this product dryer
sampling location varied from 51.1% to 61.1-% at the inlet test
location and from 31.7% to 52.3% at the outlet test location.
These variations in moisture content were the cause of the
anisokinetic sampling conditions at the sampling locations.
The anisokinetic conditions at each test location were such that
the velocity in the nozzle was greater than the velocity in the
stack (Vn > Vs). Under such conditions the measured concentra-
tion of particulate is less than the actual concentration in the
stack gas. This is due to the inertial properties of the larger
particles - they tend to pass the nozzle while the gas and the
smaller particles are drawn into the nozzle. As a result, fewer
particles are collected per unit volume. This being the case,
it is important to sample isokinetically in streams where there
are predominately large particles.
Correction factors for anisokinetic sampling are shown in
Exhibit A. The following table detailing the test results is
corrected for anisokinetic sampling.
Test No. 1 2 3
Inlet Particulate Concentration (gr/SCFD)
Measured 32.25525 29.98808 33.62483
Correction Factor 0.6875 0.6745 0.8246
Actual 46.91673 44.45972 40.77714
Outlet Particulate Concentration (gr/SCFD)
Measured 0.03761 0.04252 0.03789
Correction Factor 0.99 1.0000 0.99
Actual 0.03780 0.04252 0.03827
Scrubber Efficiency (%)
Based on Measured 99.88 99.86 99.89
Based on Actual 99.92 99.90 99.91
-------
-6-
EXHIBIT A
Appendix C
Errors due to Anisokinetic Sampling
Failure to withdraw a sample from a flow-
ing stream at the same velocity as that which
exists locally in the stream will result in non-
representative sampling. If the sampling rate
is much higher than the local stream velocity,
a greater fraction of smaller rather than larger
panicles will be drawn into the probe. If sam-
pling is much lower than the stream velocity,
large panicles will be impacted into the col-
lecting probe.
Although theoretical and experimental data
are available, a comprehensive study of these
errors has not been made. The data is almost
entirely empirical and reflects different tech-
niques using different panicles. The influence
of probe shape and size has yet to be fully eval-
uated.
In Table Cl the errors due to anisokinetic
sampling rates are given, which represent data
composited from several workers' experi-
ments.4
Panicles used in the studies yielding the
data shown were coal dust, dibutyi phthalate.
and fungus spores, all of which are relatively
low density materials, ranging from 1.3 for
coal dust to about 1 or somewhat less for the
spores. Since panicle density will materially
increase the inertial effects, the sampling error
could be considerably larger than, the tabled
values for a given panicle size. The last column
*GREEN. H. L.; LANE. w. R.
Dusts, Smokfi, and .WisM. London: £. and F. X. Span,
Ltd. 1964. 2nd ed. p 272.
Table Cl
Ratio of Observed to Actual Concentration of Panicles when Sampled
at Various Fractions and Multiples of Isokinetic Flow
C Observed concentration
in sompie
_...,. C., Actual concentration
U Prooe miet velocity
i.'.. Duct velocity d,, = 4um df— 12am .: * O.J.'B 0.:>S
.2 o.i s O.M« ij.»5
.J 0.: 7 (I «4 0 :)4
i IV: 7 •).:!•: 0.93
..i !>.! « i.).i» •> V!3
n O.r.i 0.3:)
.7 0-4 0 "3
..-? 0 : 2 i).72
i :• o • n o.riS
••" . :>'iK
1.33
1.23
1.14
1.06
1.03
l.-iii
0.-5
'J.:l'2
0.55
O.J3
1.46
1.41
1.32
1.16
i. or
l.OO
0.-J3
0.37
0.34
0.31
0."6
0.7^
0.71
•.'.no
O.sii
•J.->4
Limit
for Very
Larste
Panicies
2.00
1.67
1.44
1.25
l.ii
:.jo
O.r'O
O.Jo
0.77
0.72
0.57
J.n3
0.59
0.55
'X 33
O.iO
T!T" :!•<•« ,"-r •.•nvwr This r:
-------
-7-
APPENDIX
of the table shows the limit of the concentra-
tion ratio for very large or dense particles. If
the sampling probe inlet velocity is only 50
percent of the duct velocity and all large lor
dense) particles in the projected area of the
probe inlet are impacted into the probe, twice
as many will be collected as should have been.
Similarly, should the sampling probe inlet
velocity be twice the duct velocity and the
N13.1
particle inertia be such that only those parti-
cles approaching in the projected area of the
probe are collected, then the observed con-
centration would be one-half the actual con-
centration.
It must be remembered that panicles gen-
erated by most natural processes vary widely
in size, and the sampling error will be the com-
posite effect of all particle sizes present.
39
-------
-8-
The correction factor was applied as follows:
Fc _ C_
FC ~ Ca
where: Fc = Correction Factor
C = Measured Concentration
Ca = Actual Concentration
r
therefore: Ca = =r-
The correction factor for the inlet concentrations was for large
particles since the Bahco analysis showed that 83% of the par-
ticles were greater than 63y. The correction factor for the
outlet concentration was for dp = 4y as only small particles
are expected to be present.
As shown by the analysis, correcting for anisokinetic sampling
increased the scrubber efficiency by 0.034%.
The scheduled test program was not completed at the trona dryer
due to two major upsets. These upsets occurred due to blasting
in the adjacent calciner and a chemical spill which made it im-
possible to remain in the area of the inlet test location.
Hence, the second inlet particulate test was aborted and the
remainder of the test program aborted when a major breakdown in
a conveyor gearbox occurred. The only tests not completed were
two inlet sulfur dioxide tests and the outlet particle size
distribution tests.
Particle size distribution tests were to be conducted at each
test location. However, due to the high moisture at the product
dryer test locations it was impossible to obtain a representa-
tive sample. The moisture dissolved the material as it was
collected on the substrates. Sufficient sample was collected
at the inlet such that an aliquot from each test was taken and
composited for a sieve and Bahco particle size analysis. It
is not feasible to conduct a particle size analysis on any
outlet filters because of the small amount of particulate.
-------
-9-
Any scraping of the filter would bias the particle size by intro-
ducing fiberglass material into the sample; also, any extrac-
tion of the filter by use of solvents such as water would
dissolve the particulate matter and the subsequent crystalli-
zation would not be representative of actual test conditions.
The sulfur dioxide test results from the trona dryer were at
the lower limit of detection of the analytical method. Since
there is no process data available a material balance cannot
be performed to verify these results.
-------
Location
Date
TABLE 2-1
SUMMARY OF EMISSION TEST RESULTS
PRODUCT DRYER
ENGLISH UNITS
Run 1
Run 2
Run 3
Average
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Inlet
5/21
5/21
5/21
Volume of Gas Sampled (DSCF)
Percent Moisture by Volume
Average Stack Temperature ( F) .
Stack Volumetric Flow Rate (DSCFM)
Percent Isokinetie
22.55
52.7
187.0
16875.
146.5
33.34
44.8
160.0
14076.
111.0
20.79
51.1
186.8
15292
149.1
32.82
31.7
160.0
16331.
94.2
24.67
61.1
188.5
12514.
120.9
30.64
52.3
160.8
12396.
115.8
_
55.0
187.4
14894.
138.8
_
42.9
160.4
14267
107.0
Total Particulate -
Filter Catch and Front Half
Acetone Wash
gr/DSCF
Ib/hr
Collection Efficiency,
Percent
32.25525 0.03671
4665.61 4.43
99.86
29.98808 0.04252
3930.81 5.95
99.85
33.62483 0.03443
3606.65 3.66
99.89
31.95605 0.03789
4067.69 4.68
99.87
Dry Standard Cubic Feet at 68 F, 29.92 inches Hg.
Dry Standard Cubic Feet per minute at 68°F, 29.92 inches Hg.
-10-
-------
TABLE 2-2
SUMMARY OF EMISSION TEST RESULTS
v_ PRODUCT DRYER
METRIC UNITS
Location
Date
Run 1
Run 2
Run 3
Average
Inlet
Outlet Inlet
Outlet Inlet
Outlet
Inlet
5/21
5/21
5/21
Outlet
Volume of Gas Sampled (DNm )
Percent Moisture by Volume
Average Stack Temperature, ( C) , .
Stack Volumetric Flow Rate (DNm /min)
Percent Isokinetic
0.64
52.7
86.1
478
146.5
0.94
44.8
71.1
399
111.0
0.59
51.1
86.0
433
149.1
0.93
31.7
71.1
462
94.2
0.70
61.1
86.9
354
120.9
0.87
52.3
71.6
351
115.8
_
55.0
86.3
422
138.8
_
42.9
71.3
404
107.0
Insoluble Particulate -
Filter Catch and Front Half
Acetone Wash
mg/DNm
Kg/hr
Collection Efficiency, Percent
73812.19
2116.32
84.01
2.01
68624.05
1783.01
97.31
2.70
76946.32
1635.97
78.78
1.66
73127.52
1845.10
86.70
2.12
99.86
99.85
99.89
99.87
a Dry Normalized Cubic Meters at,20°C, 760mm Hg
Dry Normalized Cubic Meters at 20°C, 760mm Hg
-11-
-------
TABLE 2-3
SUMMARY OF EMISSION TEST RESULTS
TRONA DRYER
ENGLISH UNITS
Location
Date
Volume of Gas Sampled (DSCF)a
Percent moisture by Volume
Average Stack Temperature ( F) .
Stack Volumetric Flow Rate (DSCFM)
Percent Isokinetic
Run 1
Inlet
5/23
19.23
19.8
400.9
103845
90.4
Outlet
54.78
17.6
403.8
114857
99.3
Run 2
Inlet
5/24
18.81
21.5
388.4
98926
92.8
Outlet
50.51
16.9
400.8
104257
100.8
Run 3
Outlet
5/24
55.82
18.4
402.9
111243
104.4
Inlet
20.6
394.7
101386
91.6
Average
Outlet
17.6
402.5
110119
101.5
Total Particulate -
Filter Catch and Front Half
Acetone Wash
gr/DSCF
Ib/hr
Collection Efficiency, Percent
51.20223
465575.65
0.03402
33.49
99.93
53.11937
45042.41
0.02686
24.01
99.94
0.00684
6.53
52.16080
45309.04
0.02258
21.34
99.94
a Dry Standard Cubic Feet at 68°F, 29.92 inches Hg.
b Dr6 Standard Cubic Feet per minute at 68°F, 29.92 inches Hg.
-12-
-------
TABLE 2-4
SUMMARY OF EMISSION TEST RESULTS
TRONA DRYER
METRIC UNITS
Run 1
Run 2
Run 3
Average
Location
Date
Volume of Gas Sampled (DNm )a
Percent Moisture by Volume
Average Stack Temperature (°C) 3 .
Stack Volumetric Flow Rate (DNm /Min)
Percent Isokinetic
Inlet
5/23
0.54
19.8
205.0
2941.
90.4
Outlet
1.55
17.6
206.5
3252
99.3
Inlet
5/24
0.53
21.5
19B.O
2801
92.8
Outlet
1.43
16.9
204.9
2952
100.8
Outlet
5/24
1.58
18.4
206.1
3150
104.4
Inlet
_
20.6
201.5
2871
91.6
Outlet
_
17.6
205.8
3118
101.5
Insoluble Particulate -
Filter Catch and Front Half
Acetone Wash
mg/DNm
Kg/hr
Collection Efficiency, percent
117170.00
20673.12
77.85
15.19
99.93
121557.19
20431.24
61.47
10.89
99.94
15.66
2.96
119363<63
20552.18
51.66
9.68
99.94
a Dry Normalized Cubic Meters at 20°C, 760mm Hg
b Dry Normalized Cubic Meters at 20°C, 760mm Hg
-13-
-------
Table 2-5
Location
Date
SC>2 Emissions
ppm
Ibs/hour
Organic Emissions
ppm
SUMMARY OF S02 AND ORGANIC EMISSION TEST RESULTS
TRONA DRYER
English Units
Run 1
Inlet Outlet
5/23
30
5/23
0.80
0.91
28
Run 2
Run 3
Inlet Outlet Inlet Outlet
5/24
1.61
1.58
22
Average
Inlet Outlet
5/23
0.82
0.85
32
5/24
0.77
0.85
1.61
1.58
26
0.80
0.87
30
-------
-15-
TABLE 2-6
SUMMARY OF OPACITY OBSERVATIONS
PRODUCT DRYER
1515-1614
5-21-79
Six -Minute Interval
1
2
3
4
5
6
7
8
\
9
10
Average Opacity (%)
0
0
0
0
0
0
0
. 0
0
0
-------
-16-
TABLE 2-7.-
SUMMARY OF OPACITY OBSERVATIONS
PRODUCT DRYER
1615-1715
5-21-79
Six Minute Interval Average Opacity (%)
1 0
2 0
3 ...-. 0
4 ' 0
5 . 0
6 0
7 0
8 0
9 0
10 0
-------
-17-
TABLE 2-8
SUMMARY OF OPACITY OBSERVATIONS
PRODUCT DRYER
0900-1000
5-22-79
Six Minute Interval
1
2
3
4
5
6
7
8
9
10
Average Opacity (%)
0
0
0
0
0
0
0
0
0
0
-------
-18-
TABLE 2-9
SUMMARY OF OPACITY OBSERVATIONS
PRODUCT DRYER
1000-1100
5-22-79
Six -Minute Interval
1
2
3
4
5
6
7
• - — - --8 - - •
9
10
Average Opacity (%)
0
0
0
0 ^ I
0
0
0
o
0
0
-------
-19-
TABLE 2-10
SUMMARY OF OPACITY OBSERVATIONS
TRONA DRYER
1118-1218
5/23/79 '
Six Minute Interval Average Opacity (%)
1 3-
2 2
3 3
4 2 •
5 0
6 2
7 °
8 2.
9 1-
10 2
-------
-20-
TABLE 2-11
SUMMARY OF OPACITY OBSERVATIONS
.TRONA DRYER
1218-1248
5/23/79
Six -Minute Interval Average Opacity (%)
1 .3 •
2 2 .»
3 .2
4 2
5 2 =
6 0
7
8
9
10
-------
-21-
TABLE 2-12
SUMMARY OF OPACITY OBSERVATIONS
... TRONA DRYER
1612-1712
5/23/79
Six Minute Interval Average Opacity (%)
1 2-
2 3-
3 1
4 0
5 . 0
6 0
7 2
8 ' 1-
9 1
10 1
-------
-22-
TABLE 2-13
SUMMARY OF OPACITY OBSERVATIONS
' ' TRONA DRYER
"••1712-1812
5/23/79
Six Minute Interval Average Opacity (%)
1 0
2 ' 0
3 0
4 0
5 0
6 0
7 0
8 0
9 0
10
-------
TABLE 2-14
SUMMARY OF OPACITY OBSERVATIONS
TRONA DRYER
0815-0915
5-24-79
Six -Minute Interval Average Opacity (%)
1 0
2 0
3 0
4 0
5 0
6 0
7 0
8 0
9 0
10 0
-------
-24-
TABLE 2-15
SUMMARY OF OPACITY OBSERVATIONS
TRONA DRYER
0915-1015
5-24-79
Six-Minute Interval Average Opacity (%)
1 0
2 0
3 0
4 0
5 0
6 0
7 0
8 0
9 0
10 0
-------
-25-
PARTICLE SIZE DISTRIBUTION
.99.99 99.999.8 9998 95 90 80706050403020
10
1 OS 02 0.1 0.05 0.01
100.0
ANDERSEN PARTICLE SIZE ANALYSIS
TEXASGULF, INC.
PRODUCT DRYER INLET
FIGURE 2-1
0.1
0.01 0.05 0.1 0.2 OS 1
10
20 30 40 50 60 70 80
90 95
98 99
99.8 99.9
O.I
CUMULATIVE PER CENT BY WEIGHT LESS THAN(Dp)
-------
-26-
PARTICLE SIZE DISTRIBUTION
100.0
90.0
99.99 99.9 99.8 99 98 95 90 80 70 60 50 40 30 20 10 5 2 1 0.5 0.2 0.1 0.05 0.01
100.0
BAHCO PARTICLE SIZE ANALYSIS
TEXASGULF, INC.
PRODUCT DRYER INLET (COMPOSITE SAMPLE
3 TESTS)
FIGURE 2-2
o.e
0.1
0.01 O.OS 0.1 0.2 0.5 1 2
10
20304050607080 90 95 9699
99.8 99.9 99.99
CUMULATIVE PER CENT BY WEIGHT LESS THAN(Dp)
-------
-27-
PARTICLE SIZE DISTRIBUTION
100.0
99.99
99.9 99.8
99 98 95
90
80
ANDERSEN PARTICLE SIZE ANALYSIS
TEXASGULF, INC.
TRONA DRYER INLET
FIGURE 2-3
0.1
0.01 0.05 0.1 0.2 0.5 1
10
20 30 40 SO 60 70 80
90 95
98 99
99.8 99.9 99.99
CUMULATIVE PER CENT BY WEIGHT LESS THAN(Dp)
-------
-28-
PARTICLE SIZE DISTRIBUTION
0 005 "•<» 100.0
99.999.8 99 98 95 90
BAHCO PARTICLE SIZE ANALYSIS
TEXASGULF, INC.
TRONA DRYER (CALCINER) INLET (COMPOSITE
SAMPLE 2 TESTS)
FIGURE 2-4
20 30 40 SO 60 70 80
0.01 0.05 0.1 0.2 0.5 1
90 95
98 99
99.8 99.9 99.99
O.I
CUMULATIVE PER CENT BY WEIGHT LESS THAN(Dp)
-------
-29-
3.0 SAMPLING METHODS
3.1 Test Port Locations and Sampling Point Determination
The location of the test ports and sampling points at each
location was determined in accordance with guidelines out-
lined in EPA Method 1 (Sample and Velocity Traverses for
Stationary Sources).
The sampling ports at the Product Dryer Scrubber Inlet
are located in the straight run of duct immediately fol-
lowing the transition at the outlet of the product dryer.
The dimension of the duct at this location is 48 inches
in diameter. Two ports are installed 90° to each other. Or-
iginally 24 traverse points were to be sampled in each port,
but due to extremely high grain loading, ten points were
sampled for three minutes each resulting in a total test
time of sixty minutes. (Figure 4-1)
The sampling ports at the Product Dryer Scrubber Outlet
are located in the stack which vents the exhaust gases
from the scrubber.to the atmosphere. Two sampling ports
are located 90° intervals around the stack. The stack
diameter at this location is 51 inches;six traverse points
were sampled in each port for five minutes each, resulting
in a test time of 60 minutes. (Figure 4-2).
^
The Trona Dryer Cyclone Inlet ports are located in the duct-
work between the fan and the cyclone inlet and the duct
dimensions at this location are 12.21 feet by 5.04 feet.
Eight ports were utilized and four points sampled at two
minutes each were used because of the high grain loading.
The resulting test time was 64 minutes. (Figure 4-3).
The sampling ports at the Trona Dryer Electrostatic Pre-
cipitator Outlet are located in the stack that vents the
exhaust gases from the precipitator to the atmosphere. Four
sampling ports were installed at 90° apart around the stack,
whose diameter measures 114 inches. A total of 12 traverse
points were sampled for 8 minutes each resulting in a test
time of 96 minutes. (Figure 4-4)
-------
-30-
3.2 Gas Velocity
The gas velocity at each location was determined in accor-
dance with guidelines outlined in EPA Method 2 (Determination
of Stack Gas Velocity and Volumetric Flow Rate). A precali-
brated type "S" pitot tube and thermocouple were rigidly
attached to each sampling probe. The velocity pressure was
measured on an inclined manometer, and the temperature on a
pyrometer. Readings were recorded at each traverse point.
3.3 Gas Composition
The gas composition was determined in accordance with guide-
lines outlined in EPA Method 3 (Gas Analysis for Carbon
Dioxide, Oxygen, Excess Air and Dry Molecular Weight). Since
there is no combustion involved at the product dryer, the gas
composition was assumed to be air. A check was made with Fyrite
analyzer for carbon dioxide and oxygen content.
Since the trona dryer is fired with coal, combustion does take
place and the gas composition is not air. The carbon dioxide
and oxygen content of the flue gas was measured with, a Fyrite
analyzer at several points in the duct. The final values are
an average of the readings taken at each point.
3.4 Particulate
The particulate concentrations were determined in accordance
with guidelines outlined"in EPA Method 5 (Determination of
Particulate Emissions from Stationary Sources).
The sampling train consisted of a nozzle, stainless steel
probe, heated sample box that contained the filter, impingers,
vacuum pump, dry gas meter and calibrated orifice (Figure
4-5). The nozzle was rigidly connected to the probe. The
probe consisted of 5/8" O.D. tubing which is wrapped with
heater tape and attached to one end is a ground balljoint.
Attached to the probe were an "S" type pitot tube and
thermocouple used for monitoring the velocity pressure and
temperature. The probe and heater box were attached to the
impinger train by means of a flexible sample line.
-------
-31-
SAMPLING POINT
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
DISTANCE FROM STACK WALL (INCHES)
1.
1.
2.
3.
5.
6.
7.
9.
11.
13.
15.
19.
23.
32.
34.
36.
38.
40.
41.
42.
44.
45.
46.
47.
00
54
64
79
04
34
73
31
04
06
50
10
90
50
94
96
69
27
66
96
21
36
46
00
PRODUCT DRYER SCRUBBER INLET SAMPLING
POINT LOCATION
FIGURE 4SJL-:
-------
-32-
SAMPLING POINT
1
2
DISTANCE FROM STACK WALL (INCHES)
2.24
7.45
15.10
PRODUCT DRYER SCRUBBER OUTLET SAiMPLING POINT
LOCATIONS
FIGURE 4^2*
-------
-33-
12.21'
5.04'
Sampling Point
1
2
4
5
Distance From Stack Wall (inches)
5.04
15.12
25.20
35.28
45.56
55.64
TRONA DRYER CYCLONE INLET SAMPLING POINT LOCATIONS
ORIGINAL FIGURE - NOT USED FOR TEST)
FIGURE ,:4-3-
-------
-34-
Sampling Point
1
2
3
Distance From Stack Wall' (inches)
5.02
16.64
33.74
TRONA DRYER ELECTROSTATIC PRECIPITATOR OUTLET SAMPLING
POINT LOCATION
FIGURE -"4-4-
-------
MODIFIED PARTICIPATE SAMPLING TRAIN
2^,
U>
Ul
FIGURE 4-5
ES-093
-------
The impinger train consisted of five Greenburg-Smith im-
pingers connected in series. The first impinger was modi-
fied by replacing the impinger tip with a blank stem. This
t
impinger was initially filled with 100 milliliters of dis-
tilled water. The second impinger was a standard Green-
burg-Smith impinger containing 100 milliliters of distilled
water. The third and fourth impingers were identical to
the first and were left dry while the fifth contained 300
grams of indicating type silica gel.
The remainder of the train consisted of a check valve,
vacuum gauge, dry gas meter and calibrated orifice.
From the fifth impinger, the effluent stream flowed through
a check valve, flexible rubber vacuum tubing, a vacuum
gauge, a needle valve, a leakless vacuum pump and a dry
gas meter.
A calibrated orifice completed the train and was used to
measure instantaneous flow rates. The dual manometer across
the calibrated orifice was an inclined verticle type gradu-
ated in hundredths of an inch of water from 0 to 1 inches
and in tenths from 1 to 10 inches.
During the test the following data was recorded at each
traverse point:
• Traverse Point
• Sampling Time
• Clock Time
• Gas Meter Reading (cf)
• Velocity Pressure (in H20)
• Orifice Pressure Drop (in. 1^0)
• Stack Temperature (°F)
• Dry Gas Meter Temperature - Inlet and Outlet (°F)
• Pump Vacuum (in. Hg)
• Sample Box Temperature (°F)
• Impinger Temperature (°F)
-------
-37-
The relationship of AP reading with the AH reading is a
function of the following variables:
• Orifice Calibration Factor
• Gas Meter Temperature
• Moisture Content of Flue Gas
• Ratio of Flue Gas Pressure to Barometric Pressure
• Stack Temperature
• Sampling Nozzle Diameter
A nomograph was used to correlate all the above variables
such that a direct relationship between AP and AH was de-
termined by the sampler and isokinetic conditions could
be maintained.
At the completion of the test, the samples were recovered
in the following manner:
Container #1: The filter was removed from the filter
holder and placed in its original con-
tainer arid sealed.
Container #2: The nozzle, probe, cyclone bypass and
front half filter holder were rinsed
with acetone. The acetone was placed in
a glass jar and sealed.
Container #3: The silica gel was returned to its ori-
ginal container and sealed.
3.5 Organics
A gaseous sample was withdrawn from the source using a
heated, glass lined stainless steel probe. Samples were
drawn into a prepurged, evacuated, heated (to above the
dew point of the sample gas) 250 ml glass grab flask until
a positive pressure was obtained in the flask. The sample
was injected into an AID model 621 portable Gas Chromato-
graph (GC) directly from the heated grab flask using the
positive pressure obtained while sampling to fill the sample
loop of the G.C. Two injections per flask were made to
-------
-38-
determine the reproducability of the sample.
The samples were analyzed for Total Hydrocarbons as methane
using a 1 cc gas sample loop and a flame ionization detec-
tor. The column temperature was 125°C. The G.C. was
standardized employing a range of certified analyzed
methane standards prepared in helium. Comparison of peak
heights of the sample gas with those of the standards
gave the following results:
Calculations
SF = The Sensitivity Factor = - ppm Standard -
peak ht(mm) x (attenuation x range)
ppm total Hydrocarbons = SF x ?eak ht « of samPle x
(attenuation x range)
3.6 Sulfur Dioxide
Sulfur oxide emissions were determined in accordance with
EPA Method 6. A glass lined heated probe was attached
to the sample train by means of a three way stopcock tee.
The sample train consisted of a midget bubbler containing
15 ml of 80% isopropanol , a blank impinger to catch carry-
over, two midget impingers each with 15 ml of 3% hydrogen
peroxide and a fifth blank impinger for carryover from the
two previous. A dry gas meter was used to meter the sample
volume and a roto-meter monitored the sample rate. The
sample was taken over a 20 minute period at a rate of 0.1
CFM. After sampling, the train, (Figure 4-6) , was purged
with activated charcoal filtered ambient air for 20 minutes
3. 7 Particle Size Distribution
The particle size distribution samples were collected using
an Andersen Cascade Impactor. The impactor consists of
multiple stages which collect .different particle sizes
-------
S08 SAMPLING TRAIN
VARIAC
PITOT
TUBE
PITOT
MANOMETER
SILICA GEL
DRYING TUBE
VACUUM
GAUGE
MIDGET
BUBBLER
MIDGET
IMPINGERS
NEEDLE
VALVE
I
u>
PUMP
DRY GAS
METER
-
RATE
METE
y
R
FIGURE 4-6
ES-028
-------
-40-
(Figure 4-7). Each stage consists of an orifice of a speci-
fic diameter above a collection plate. The orifice sizes
of each stage are different and are arranged in descending
order, the largest being stage 1. The sampling system was
set up as shown in Figure 4-8. The stack conditions were
determined and the sample was extracted isokinetically.
As the sample flows through each orifice, and is deflected
off a glass fiber substrate filter placed on the collec-
tion plate, particles of a specific size become impacted
on the substrate while the remaining particles entrained
in the gas stream proceed to the next collection stage.
The range of particle sizes retained on the substrate varies
according to the velocity of the gas (as determined by the
sampling rate and orifice diameter), the gas viscosity and
the particle density. Since the orifices are arranged in
descending diameters, the gas velocity increases and the
particle size collected on each stage decreases.
During the sampling a cyclone preseparator was used to
precut particles above 10 microns and avoid overloading
the collection substrates. At the completion of each test
the contents of the preseparator and an acetone wash were
placed in a sample bottle. The glass fiber substrate
filters were returned to their original containers and
sealed.
3.8 Visible Emissions
The visible emissions were determined in accordance with
guidelines outlined in EPA Method 9 (Visual Determination
of the Opacity of Emissions from Stationary Sources).
-------
ANDERSEN SAMPLING TRAIN
ANDERSON
SAMPLER
IMPINGERS
PUMP
GAS METER
ORIFICE
MANOMETER
FIGURE 4-7
ES-094
-------
-42-
ANDERSEN STACK SAMPLER
JET STAGE (9 TOTAL)
SPACERS
NOZZLE
GLASS FIBER
COLLECTION
SUBSTRATE
BACKUP
FILTER
PLATE ^
HOLDER
INLET
\
CORE
ES-095
FIGURE 4-8
-------
-43-
4.0 ANALYTICAL PROCEDURES
4.1 Particulate
Each sample from the particulate test was analyzed
in the following manner:
Container #1: The filter was removed from its sealed
container and placed on a tared watch
glass. The filter and watch glass were
dessicated with anhydrous CaSC>4 and
weighed to a constant weight. The weight
was recorded to the nearest 0.01 mg.
Container #2: The acetone washings were transferred to
a tared beaker. The acetone was then evap-
orated at ambient temperature and pressure,
dessicated and weighed to a constant weight.
The weight was recorded to the nearest 0.01 mg.
Container #3: The silica gel was weighed to the nearest
0.1 gram on a beam balance.
4. 2 Sulfur Dioxide
The contents of the first two impingers along with washes
were discarded and the contents of the last three impingers
along with washes were saved for analysis for SC>2. The
sulfur oxide samples were analyzed titrimetrically using
barium perchlorate. Calculations for sulfur oxide emissions
appear in the Appendix.
4.3 Particle Size Distribution
The fiberglass substrate filters were dessicated and weighed
to a constant weight. The net weight gain was recorded to
the nearest 0.01 mg.
The acetone rinse of the cyclone preseparator was trans-
ferred to a tared beaker. The beaker was heated to a tem-
perature well below the boiling point until the water was
evaporated. The beaker was then dessicated and weighed
to a constant weight. The net weight gain was recorded to
-------
-44-
the nearest 0.01 mg. Bahco analysis was also performed
on the inlet samples due to the high moisture content re-
stricting use of the Andersen Impactor.
-------
Prepared by:
Reviewed by;
Approved by:
-45-
PROJECT 01-9517-14
William J. Cesreo
Project Direstar
ger A. Kniskern
Senior Project Manager
Anthony j Buonicore
General Manager
Air Pollution Services Division
------- |