r
EMISSION TESTING REPORT
PROJECT NUMBER 72 MM Q5
STAUFFER CHEMICAL COMPANY
TARPON SPJRI.NG-S. „ FLORIDA
PEDCo ENVIRONMENTAL
-------
EMISSION TESTING REPORT
PROJECT NUMBER 72 MM 05
STAUFFER CHEMICAL COMPANY
TARPON SPRINGS- FLORIDA
Prepared by
Richard W. Gerstle, P.E.
Robert S. Amick
Contract No. 68-02-0237
Task 4
EPA Project Officer: John Wilkens
Gene Riley
PEDCo-Environmental Specialists, Inc.
Cincinnati, Ohio
-------
TABLE OF CONTENTS
SECTION PAGE
LIST OF FIGURES ii
LIST OF TABLES iii
I. INTRODUCTION 1
II. SUMMARY AND DISCUSSION OF RESULTS 6
III. PROCESS DESCRIPTION AND OPERATION 29
IV. LOCATION OF SAMPLING POINTS 33
V. SAMPLING OF ANALYTICAL PROCEDURES 41
APPENDIX A - FLUORIDE, P^O,. , AND PARTICULATE
SAMPLING DATA AND SAMPLE CALCULATIONS
APPENDIX B - GASEOUS RESULTS
APPENDIX C - ORIGINAL FIELD DATA
APPENDIX D - SAMPLING PROCEDURES
APPENDIX E - ANALYTICAL PROCEDURES AND RESULTS
APPENDIX F - SAMPLE IDENTIFICATION LOG
APPENDIX G - TEST LOG
APPENDIX H - PROJECT PARTICIPANTS
-------
LIST OF FIGURES
FIGURE PAGE
1 Simplified Flow Diagram of Elemental Phosphorus 4
Manufacturing Process
2 Slag Tap Scrubber Inlet Sampling Site 34
3 Slag Tap Scrubber Outlet Sampling Site, 35
Stauffer Chemical Co., Tarpon Springs, Fla.
4 Slag Tapping Sampling Sites 36
5 Kiln Scrubber Inlet Sampling Site, Stauffer 37
Chemical Co., Tarpon Springs, Florida
6 Kiln Scrubber Outlet Sampling Site, Stauffer 38
Chemical Co., Tarpon Springs, Florida
7 Spray Chamber Inlet and Outlet Sampling Sites 39
8 Fluoride and PpCv Sampling Train . 43
9 Particulate Sampling Train 45
10 Sulfur Dioxide Sampling Train 47
ii
-------
LIST OF TABLES
TABLE PAGE
1 Summary of Program Measurements Made at Stauffer 2
Chemical Co. in Tarpon Springs, Florida
2 Slag Tapping—Scrubber Inlet Fluoride and P^Or 7
Emissions, Summary
3 Slag Tapping—Scrubber Outlet Fluoride and P^OC 8
Emissions, Summary
4 Scrubber Water Analysis—Slag Tap Process Venturi 9
Scrubber
5 Nodulizing Kiln Tests—Stack Gas Volumetric Flow 13
Rates
6 Fluoride and P^O^. Emission Data Summary Nodulizing 14
Kiln—Scrubber Spray Chamber Inlet
7 Fluoride and P^CV Emission Data Summary Nodulizing 15
Kiln—Scrubber Spray Chamber Outlet B
8 Fluoride and P-jCV Emission Data Summary N dulizing 16
Kiln—Scrubber Spray Chamber Outlet C
9 Particulate Emission Data Nodulizing Kiln—Scrubber 17
Spray Chamber Inlet
10 Particulate Emission Data Nodulizing Kiln—Scrubber 18
Spray Chamber Outlet
11 Nodulizing Kiln Summary of Gaseous Testing 19
12 Air Return Line—Fluoride and PpO- Data Summary 21
13 Scrubber Water Analyses—Nodulizing Kiln Spray 22
Chamber
14 Fluoride Content of Various Process Streams 30
111
-------
I. INTRODUCTION
Atmospheric emissions of fluorides, phosphorus pent-
oxide, total particulate, and sulfur dioxide from Stauffer
Chemical Company's elemental phosphorus plant at Tarpon
Springs, Florida, were sampled to form a data base for
New Source Performance Standards as authorized by the Clean
Air Act of 1970.
Tests were made to determine fluoride and phosphorus
pentoxide concentrations before and after the venturi
scrubber serving a furnace slag tapping operation. Fluoride,
phosphorus pentoxide, total particulate, and sulfur dioxide
concentrations before and after the spray chamber serving
the feed nodulizing kiln were also measured. In addition,
the fluoride and P^Oj. concentrations in the air return line
and the CO line leading to the kiln were measured. All
field testing was conducted by PEDCo-Environmental
Specialists, Inc. and all sample analyses were performed
by the Environmental Protection Agency's Office of Air
Quality and Planning. Table 1 summarizes the test locations,
dates, and the measurements made. In addition, Stauffer
Chemical Company conducted fluoride tests at the nodulizing
kiln spray chamber outlet during this test series using both
manual and continuous bubbler systems.
This test series was interrupted by Hurricane Agnes.
Sampling therefore occurred during two different periods.
-1-
-------
TABLE 1.
SUMMARY OF PROGRAM MEASUREMENTS MADE AT STAUFFER CHEMICAL COMPANY
IN TARPON SPRINGS, FLORIDA
DATE
1972
6/13
6/14
7/18
7/20
7/21
7/21
6/13
, 6/14
v> 7/18
6/13
6/14
7/18
7/20
7/21
7/21
6/15
7/19
6/16
6/21
6/21
6/16
6/21
6/21 ,
TEST
NO.
1AK
2AK
3AK
1
2
3
1BK
2BK
3BK
1CK
2CK
3CK
1
2
3
1-CO
1,2,3
IS
2S
3S
IS
2S
3S
TEST
SITE STACK GAS PARAMETERS EMISSIONS
Kiln Spray Velocity Temp. % H00 Fluorides & P-,0^
Chamber Inlet
\ /
Kiln Spray a
Chamber Outlet, B
1
Kiln Spray
Chamber Outlet C
5.7' Above Site B
Kiln Spray
Chamber Outlet, B
\
1
Particulate & SO^T
fL
1
Fluorides & P^Oj.
\
f-> ~>
f
Particulate
v i \
w
CO Line b Fluorides & P2°5
Air Return Line Velocity • 1
Slag Tap Scrubber
Inlet
1
Slag Tap Furnace
Outlet
1 I
( \
*
Fluorides & P-?0^
i
,
a) Velocity too low to measure and was calculated based on gas volume
measured at inlet site and dilution air drawn into duct.
b) No velocity measured due to dangerous conditions (explosion) and small
sampling port.
c) Also sampled for SO on July 19 at inlet and outlet site
-------
In all cases, the tests were run to determine average
emission concentrations and rates under normal operating
conditions. Whenever possible, testing was conducted
during normal operating conditions.
Slag Tapping Operation
At this plant, the slag tapping ports in the electric
arc furnace were covered by a movable hood system which
vented fumes during slag tapping through a venturi scrubber
and out the stack. Stack gases were sampled before and after
the scrubber as shown by Points E and F in Figure 1. For the
three test runs sampling was conducted only, during the
tapping operation to determine concentrations of fluoride
and phosphorus pentoxide (P^O-). Samples of the scrubber
£ Ij
water were also collected at Points g and h as shown in
Figure 1.
Nodullzing Kiln
Gases from the feed nodulizing kiln pass through cyclone
collectors and then a large diameter spray chamber which also
serves as the exit stack. Stack gases were sampled before
and after the spray chamber at Points A, and B and C, respec-
tively, as shown in Figure 1. The C location was 5.7 ft.
above Site B and was sampled to determine the effect of
additional residence time of the gaseous emissions in the
-3-
-------
PROCESS
WATER
t
SOLID OR LIQUID STREAMS
-'- GAS STREAMS
"^SAMPLED STREAMS
r
/^
X PHOSPHATE
ORE
\ 1>
STOCKPILES \-.
I 1
I
L
f
i
ORE
DRYER
t
co
^
/ / FUEL --*
/
j /
DUMP V
OR
Figure 1.
FLARE
1 OTHlR *~ "
! 1
NODUL- I
IZING V
KILN
r
1
t I
k — *- 1
1
I
Air Return
Line
COOLER
t
1
1
1
,._ : £.
^TOCK
BINS
^
1 Sampling Port
| AIR
CO line
1 C°
f
f
I
1
— — r
1ENTAL
PHOSPHOROUS
ESP
f
FEED
BINS
CO + P4
^ - • _
!'
. L
PROCESS
WAT
i
ER
g 4
r^
M
H
M
1 ^
- _ zfe.
^ SCRUBBER
\
/ ^^ h
<
\
. FURNACE
\
bAI UKAIWV > ' «• -V
>^
E *
S~-
X
^ WATER
i
L HOOD ./'
xy
TAPHOLE &
H
^ RUNNER
/
' ~~~^T~-~-Wk. c. r-
\ ^^ SLAG
i»
WATER
PROCESS
WATER
-------
large stack. Sampling site A was located before the spray
chamber. Scrubber water samples were taken at Points c and
d.
Simultaneous testing was done at the three emission test
sites, A, B, and C to determine concentrations of fluoride
and PpOj-. Triplicate tests were made. Simultaneous tests
were performed at two test sites, A and B, to determine
concentrations of particulate and sulfur dioxide. Triplicate
tests were made for these constituents. During the
fluoride and particulate tests, gas samples were als,o taken
for molecular weight determinations.
Raw material and product samples were also collected
during the test periods and analyzed for fluorides, P^O,.,
and trace metals. Complete analysis of these samples is
included in EMB file 72-MM-05.
-5-
-------
II. SUMMARY AND DISCUSSION OF RESULTS
Slag Tapping Operation
Slag tapping was an intermittent operation which occured
approximately every 70 minutes. Each tap lasted for approxi-
mately 12 minutes. Sampling was conducted simultaneously
before and after the venturi scrubber only during the
tapping periods to determine the concentrations of fluorides
and P2°5*
Tables 2, 3, and 4 summarize the stack gas and scrubber
water data obtained during these tests. Sampling procedures
are described in Section V. More detailed sampling and
analytical results are included in the appendices to this
report. Emission data are reported on a concentration (grains
per standard cubic foot) basis, and also as pounds per hour
emitted during the slag tapping periods.
The higher gas flows measured at the outlet site indicate
leakage of outside air into the duct system around the fan.
Measured moisture content of the outlet stream showed that the
gas was not saturated with water. This may be due to some
reheating of the gas because of the hot slag being dumped near
the duct leading to the outlet site, and also incomplete
saturation in the venturi scrubber.
-6-
-------
TABLE 2. SLAG TAPPING - SCRUBBER INLET
rjjUUKj.u£i & *2 5
Run Number
Date
Volume of Gas Sampled - DSCFa
Percent Moisture by Volume
Average Stack Temperature - F
Stack Volumetric Flow Rate - DSCFM
Stack Volumetric Flow Rate - ACFMC
Percent Isokinetic
Percent by Volume CO-
Percent by Volume O-
Percent by Volume CO
Percent by Volume N2
Fluoride Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions, % insoluble
PjOj. Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
Ib/hr
PjO,. Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
P20c Emissions, % insoluble
^iYLLtiSJ-UiNO
1
6-16-72
29.78
2.60
219
14,000
18,900
75.8
21.0
<1
79.0
107
0.0555
0.0410
6.66
107.1
0.0555
0.0410
6.66
0.093
509
0.264
0.195
31.7
509.6
0.264
0.195
31.7
1.18
ouL"int\Ki
Composite
6-21-72
26.00
1.82
173
14,600
18,200
107
21.0
<1
79.0
4 2'. 4
"0.0251
0.0201
3.14
44.2
0.0262
0.0209
3.28
4.07
453.2
0.269
0.215
33.6
465.8
0.276
0.221
34.5
2.71
3
6-21-72
20.10
2.15
189
13,100
16,900
113
21.0
<1
79.0
35.5
0.0272
0.0212
3.06
35.8
0.0275
0.0213
3.08
0.838
527,6
0.405
0.316
45.4
532.3
0.408
0.318
t5.8
0.883
Weighted
Average
25.29
2.19
194
13,900
18,000
98.6
21.0
<1
79.0
61.6
0.0376
0.0289
4.48
62.4
0.0381
0.0293
4.53
1.28
496.6
0.303
0.234
36.1
502.4
0.306
0.235
36.5
1.16
Dry standard cubic feet at 70 F, 29.92 in. Hg.
Dry standard cubic feet per minute at 70 F, 29.92 in. Hg.
Actual cubic feet per minute.
During slag tapping only.
-7-
-------
TABLE 3. SLAG TAPPING - SCRUBBER OUTLET
FLUORIDE & P-,Or EMISSIONS SUMMARY
^ 3
Run Number
Date
Volume of Gas Sampled - DSCFa
Percent Moisture by Volume
Average Stack Temperature - °F
Stack Volumetric Flow Rate - DSCFMb
Stack Volumetric Flow Rate - ACFMC
Percent Isokinetic
Percent by Volume CO_
Percent by Volume 0-
Percent by Volume CO
Percent by Volume N_
Fluoride Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions, % insoluble
P-O- Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
Ib/hr
P-O- Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
P-O- Emissions, % insoluble
a Dry standard cubic feet at 70 °F,
Dry standard cubic feet per minute
1
6-16-72
28.95
7.56
121
18,800
22,500
106
<1
21.0
<1
79.0
3.21
0.00171
0.0143
0.276
4.25
0.00227
0.0189
0.365
24.5
57.5
0.0307
0.0256
4.94
63.8
0.0340
0.0256
5.48
9.87
29.92 in.
at 70 °F,
2
6-21-72
27.29
5.62
120
17,200
20,000
107
<1
21.0
<1
79.0
1.70
0.000960
0.0-00824
0.142
2.43
0.00137
0.00118
0.202
30.0
61.3
0.0346
0.0298
5.11
62.4
0.0352
0,0298
5.21
1.76
Hg.
29.92 in.
3
6-21-72
26.71
5.86
119
17,200
20,000
105
<1
21.0
<1
79.0
1.40
0.000808
0.000692
0.119
1.58
0.000912
0.000781
0.134
11.4
82. 1
0.0474
0.0406
6.98
83.9
0.0484
0.0406
7.14
2.15
Hg.
Weighted
Average
27.65
6.35
120
17,900
20,800
106
<1
21.0
<1
79.0
2.10
0.00117
0.000995
0.179
2.75
0.00153
0.00130
0.233
22.0
67.0
0.0376
0.0325
5.68
69.9
0.0390
0.0325
5.94
4.15
Actual cubic feet per minute.
During slag tapping only.
-8-
-------
TABLE 4 . SCRUBBER WATER ANALYSIS - SLAG TAP PROCESS VENTURI SCRUBBER
I
vo
I
Date
1972
6/16
6/21
6/21
6/21
6/21
Test No.
1 - Discharge
2 - Discharge
3 - Discharge
Slag Pita
Slag Pita
Soluble Fluorides
ing/liter
282
366 avg = 375
477
365
340 avg = 359
373
596
452 avg = 491
426
243
238
P2°5
mg/ml
1.4
1 . 7 avg =1.7
1.9
1.7
1 .6 avg =1.6
1.4
2.5
1.8 avg =1.9
1.5
1.4
1.0
pH
2.78
2.50
2.32
2.62
2.75
2.88
2.15
2.43
2.61
3.48
3.56
avg = 2.53
avg = 2.75
avg =2.40
a) This water is used for scrubber water make-up.
-------
R.
Flexible TefIon^-^tubing was used between the probe
and first impinger for all tests at both the inlet and out-
let sites.
Fluoride Data Evaluation - The measured values for
the slag tapping appear to be representative of process
emissions except for Test 1 (inlet and outlet) which has
approximately twice the emission rate as compared to Tests
2 and 3. The following are possible reasons for the high
fluoride measurement on Test 1:
1. During the first slag tap the sampling crew on
the inlet side was driven off the platform by
the high concentration of acid-gas fumes (probably
escaping from under the furnace fume hood). The
EPA Project Officer, John Wilkens, had delayed
the first slag tap for approximately two hours.
2. The analytical results in Appendix E, show that
for the inlet Test 1 the soluble fluorides were
107 ing/sample as compared to 42.4 mg/sample and
35.5 mg/sample of soluble fluorides respectively
for Tests 2 and 3. Sample volumes were approxi-
mately the same for all three tests.
3. The percent isokinetic for Test 1 inlet was only
76% as compared to 107% and 113% for Tests 2
and 3. This would result in the possibility of
a higher grain loading. Only four traverse ports
were used during Test 1 at the inlet. During
Tests 2 and 3 16 traverse points were used. The
P^OC- concentration was however not out-of-line
for this run.
-10-
-------
Based on these observations, Test 1 inlet does not
appear to be representative of normal operation. Water
soluble fluoride emissions were in the range of 0.025
to 0.027 grains/DSCF at the scrubber inlet and averaged
0.0012 grains/DSCF at the outlet. The scrubber collection
efficiency, based on pounds of total fluoride emitted,
averaged 95%.
P 0 Data Evaluation - The Test 3 inlet concentration
of 0.405 grains/DSCF for the water soluble P^c is approxi-
mately 35% higher than the concentration measured in Tests
1 and 2. However, no reason is apparent for this, and an
average of 0.303 grains/DSCF on the inlet appears to be
representative of normal operation. The outlet concentrations,
averaging 0.038 grains/DSCF and ranging from 0.031 to 0.048
grains/DSCF are also considered in the range of normal
operation. Average scrubber collection efficiency for
water soluble PO^C was 88%..
Nodulizing Kiln Operation
Tables 5 to 13 summarize the data obtained on the
nodulizing kiln spray chamber. Tests for fluorides, P2°s'
total particulate, and sulfur dioxide were conducted before
and after the spray chamber which was used to reduce atmos-
pheric emissions. Since the outlet flow from the spray
-11-
-------
TABLE 5. NODULIZING KILN TESTS -
STACK GAS VOLUMETRIC FLOW RATES
RUN
NO.
Fluoride 1
Fluoride 2
Fluoride 3
Particulate 1
Particulate 2
Particulate 3
DATE
6/13/72
6/14/72
6/22/72
7/18/72
7/20/72
7/21/72
7/21/72
INLET
GAS
VELOCITY
FPMa
6,490
6,230
6,800
4,700
5,960C
5,440
5,840C
5,230
5,390C
GAS
VOLUME
SCFMb
23,900
24,300
24,300
18,100
22,700C
20,700
22,200°
20,200
20,700°
OUTSIDE TEMPERING AIR
GAS GAS PORT
VELOCITY VOLUME SITE
FPMa SCFMb
Lower
Upper
Lower
Upper
2,730 2,520
2,730 2,520 Lower
Upper
Lower
Lower
Lower
(B)
(C)
(B)
(C)
(B)
(C)
(B)
(B)
(B)
OUTLETS
GAS
VELOCITY
FPMa
86.4
80.0
94.7
91.9
95.9
85.8
81.2
81.6
74.1
GAS
VOLUME
SCFMb
26,400
26,400
26,800
26,800
26 ,800
26,800
25,200
24,700
23,200
to
I
b
Feet per minute, stack conditions.
Standard cubic feet per minute, dry basis at 70°F, 29.92 in. Hg.
Represents flow measurements determined by performing a separate pitot traverse before
or after each particulate run. Normal inlet sampling covered only 12 traverse points
and was considered not as representative as the separate pitot traverses that covered
24 points.
-------
TABLE 6. FLUORIDE & P-O- EMISSION DATA SUMMARY
NODULIZING KILN - SCRUBBER SPRAY CHAMBER INLET
Run Number Composite 2 3
Date
Volume of Gas Sampled - DSCFa
Percent Moisture by Volume
Average Stack Temperature - °F
Stack Volumetric Flow Rate - DSCFMb
Stack' Volumetric Flow Rate - ACFM°
Percent Isokinetic
Percent by Volume CO_
Percent by Volume 02
Percent by Volume CO
Percent by Volume N2
Fluoride Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions, % insoluble
P_O5 Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
Ib/hr
P-.O,- Emissions - total
£ o
mg Collected
gr/DSCF
gr/ACF
lb/hrd
P2°5 Emissions, % insoluble
6-13-72
64.03
19.9
702
23,900
67,000
102
1.7
20.3
0.87
77.1
7300.8
1.76
0.626
360
8171.1
1.97
0.703
404
"10.7
85.1
0.0205
0.00731
4.18
149.6
0.0360
0.0128
7.36
43.1
6-14-72
28.65
19.1
653
24,300
64,300
105
12.4
13.3
<1
74.3
3201.5
1.72
0.651
358
3382.6
1.84
0.694
382
5.35
243.1
0.131
0.0528
27.2
466.7
0.251
0.100
52.2
47.9
7-18-72
32.70
20.7
729
24,300
70,200
120
16
8
2.2
73.8
5401.4
2.54
0.880
530
5546.2
2.61
0.898
540
2.61
317.5
0.150
0.0516
31.1
d
d
d
d
d
Weighted
Average
41.79
19.9
695
24,200
67,200
109
5301
1.95
0.703
405
5700
2.10
0.757
434
6.68
215
0.0795
0.0293
16.5
-
-
_
-
-
Dry standard cubic feet at 70 °F, 29.92 in. Hg.
Dry standard cubic feet per minute at 70°F, 29.92 in. Hg.
Actual cubic feet per minute.
Analysis of ^2^5 -i-nsoiublo portion impossible due to gel being formed
when sodium hydroxide was combined with sample.
-13-
-------
TABLE 7.
NUiJUijl^lNU K.11.N - bCKUtiBEj
Run Number
Date
Volume of Gas Sampled - DSCFa
Percent Moisture by Volume
Average Stack Temperature - °F
Stack Volumetric Flow Rate - DSCFM
Stack Volumetric Flow Rate - ACFMC
Percent Isokinetic
Percent by Volume C0_
Percent by Volume O-
Percent by Volume CO
Percent by Volume N_
Fluoride Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions, % insoluble
P-Og Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
Ib/hr
P2°5 Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
P_05 Emissions, % insoluble
a Dry standard cubic feet at 70 °F,
Dry standard cubic feet per minute
K bFKAX Cti
1
6-13-72
43.07
10.0
138
26,400d
33,100
74.1
13.3
13.4
0.2
73.1
8.8
0.00315
0.00251
0.689
9.45
0.00338
0.00270
0.742
6.88
23.9
0.00855
0.00683
1.94
26.6
0.00951
0.00760
2.20
10.2
29.92 in.
at 70 °F,
lAMBJiK UUT1
2
6-14-72
39.97
16.7
140
26,800d
36,300
67.8
6.0
17.9
<1
76.1
4.7
0.00181
0.00134
0.404
5.3
0.00204
0.00151
0.457
.11.3
15.1
0.00582
0.00431
1.33
17.8
0.00686
0.00508
1.56
15.2
Hg.
29.92 in.
j-tiT a
3
7-18-72
58.67
19.0
135
26,800d
36,800
62.6
24.5
0.00643
0.00468
1.47
25.0
0.00656
0.00478
1.50
2.00
67.0
0.0176
0.0128
4.04
70.4
0.0185
0.0135
4.23
4.84
Hg.
Weighted
Averaqe
47.24
15.2
138
26,700d
35,400
68.2
12.7
0.00414
0.00313
0.947
13.3
0.00434
0.00328
0.992
4.51
35.3
0.0115
.0.00859
2.64
38.3
0.0125
0.00933
2.86
7.83
Actual cubic feet per minute.
Outlet volumetric flows based on inlet flow measurements plus
additional tempering air entering prior to the outlet sampling site.
-14-
-------
TABLE 8. FLUORIDE & P205 EMISSION DATA SUMMARY
NODULIZING KILN - SCRUBBER SPRAY CHAMBER OUTLET C
Weighted
Run Number
Date
Volume of Gas Sampled - DSCFa
Percent Moisture by Volume
Average Stack Temperature - °F
Stack Volumetric Flow Rate - DSCFMb
Stack Volumetric Flow Rate - ACFMC
Percent Isokinetic
Percent by Volume CO_
£t
Percent by Volume O2
Percent by Volume CO
Percent by Volume N2
Fluoride Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions, % insoluble
P20j- Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
Ib/hr
P205 Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
PjOg Emissions, % insoluble
a Dry standard cubic feet at 70 °F,
Dry standard cubic feet per minute
1
6-13-72
44.78
3.17
136
26,400d
30,700
74.7
13.3
13.4
0.2
73.1
4.1
0.00141
0.00122
0.320
4.42
0.00151
0.00130
0.343
7.2
11.5
0.00395
0.00340
0.898
13.2
0.00454
0.00391
1.03
12.9
29.92 in.
at 70 °F,
2
6-14-72
39.42
14.1
140
26,800d
35,200
65.8
6.0
17.9
<1
76.1
6.4
0.00250
0.00191
0.576
7.6
0.00297
0.00226
0.684
15.8
17.1
0.00668
0.00510
1.54
19.7
0.00770
0.00588
1.77
13.2
Hg.
29.92 in
~»
Composite
7-18-72
27.07
9.01
138
26,800d
32,900
105
11.1
0.00631
0.00514
1.45
11.6
0.00660
0.00537
1.52
-.4.31
25.5
0.0145
0.0118
3.33
26.4
0.0150
0.0122
3.45
3.41
. Hg. .
_,
Averaae
37.09
8.76
138
26,700d
32,900
81.8
7.20
0.00299
0.00243
0.684
7.87
0.00327
0.00266
0.748
8.51
18.0
0.00747
0.00605
1.72
19.8
0.00822
0.00662
1.88
9.09
Actual cubic feet per minute.
Outlet; volumetric flows based on inlet flow measurements plus
additional tempering air entering prior to the outlet sampling
. -15-
site.
-------
TABLE 9. PARTICULATE EMISSION DATA
NODULIZING KILN-SCRUBBER SPRAY CHAMBER INLET
RUN NUMBER
Date
Volume of Gas Sampled-DSCFa
Percent Moisture by Volume
Average Stack Temperature- °F
Stack Volumetric Flow Rate-
DSCFMb
Stack Volumetric Flow Rate-
ACFMC
Percent Isokinetic
Percent Excess Air
Feed Rate-ton/hr
Particulates-probe , cyclone,
and filter catch
mg
gr/DSCF
gr/ACF
Ib/hr
Particulates-total catch
mg
gr/DSCF
gr/ACF
Ib/hr
Percent impinger catch
1
7-20-72
23.02
20.1
652
18,100
48,500
114
167
1069
0.715
0.266
111
1382.9
0.925
0.345
143
22.7
2
7-21-72
25.006
18.4
682
20,700
56,100
108
39.4
1054.6
0.650
0.240
115
1196.7
0.738
0.272
131
11.9
3
7-21-72
24.81
18.1
670
20,200
54,000
110
44.7
1114.5
0.692
0.259
120
1347.9
0.838
0.313
145
17.3
WEIGHTED
AVERAGE
24.28
18.9
668
19,700
52,900
111
83.7
1079
0.685
0.254
116
1309
0.831
0.309
140-
17.6
Dry standard cubic feet at 70°F, 29.92 in. Hg.
Dry standard cubic feet per minute at 70°F, 29.92 in. Hg,
Actual cubic feet per minute
-16-
-------
TABLE 10. PARTICULATE EMISSION DATA
NODULIZING KILN-SCRUBBER SPRAY CHAMBER OUTLET
RUN NUMBER
Date
Volume of Gas Sampled-DSCFa
Percent Moisture by Volume
Theoretical Saturated per-
cent moisture
Average Stack Temperature- °F
Stack Volumetric Flow Rate-
DSCFMb
Stack Volumetric Flow Rate-
ACFMC
Percent Isokinetic
Feed Rate-ton/hr
Particulates-probe ,
and filter catch
mg
gr/DSCF
gr/ACF
Ib/hr
Particulates-total catch
mg
gr/DSCF
gr/ACF
Ib/hr
Percent impinger catch
1
7-20-72
105.57
19.6
12.0
122
25,200
31,200
119
386.4
0.0564
0.0455
12.3
634.5
0.0926
0.0749
20.1
39.1
2
7-21-72
109.38
22.7
13.5
126
24,700
31,300
127
417
0.0587
0.0466
12.5
799.3
0.113
0.0892
23.9
47.8
3
7-21-72
105.51
20.1
11.5
120
23,200
28,400
130
300.3
0.0439
0.0360
8.76
630.2
0.0922
0.0754
18.4
52.3
WEIGHTED
'AVERAGE
106.82
20.8
12.5
123
24,400
30,300
125
368
0.0531
0.0426
11.1
688
0.0993
0.0797
20.8
46.5
Dry standard cubic feet at 70°F, 29.92 in. Hg.
Dry standard cubic feet per minute at 70°F, 29.92 in. Hg.
Q
Actual cubic feet per minute calculated from saturated
eras conditions.
NOTES: All outlet volumetric flows are based on inlet duct gas
flow plus 2530 SCFM for dilution air entering the gas
stream prior to the outlet sampling point.
-17-
-------
TABLE 11. NODULIZING KILN
SUMMARY OF GASEOUS TESTING
A
RUN NO.
Fluoride 1
Inlet
Outlet
Fluoride 2
Inlet
Outlet
Fluoride 3
Inlet
Outlet
Particulate 1
Inlet
Outlet
Particulate 2
Inlet
Outlet
Particulate 3
Inlet
Outlet
DATE
6-13-72
6-14-72
7-18-72
7-20-72
7-21-72
7-21-73
CARRIER GASES
GAS COMPOSITION (VOLUME
CO,
CO
12.4 .13.3 <1
6.0 17.9 <1
16.0
8.0 2.2
DRY)
*T~
1.70 20.3 0.87 77.1
13.3 13.4 0.20 73.1
74.3
76.1
73.8
18.6 12.1 1.93 67.4
18.3 6.47 2.00 73.3
12.2 10.7 2.90 74.2
18.4 6.70 1.43 73.4
11.3 10.8 2.30 75.6
RUN NO.
1
2
3
B. SULFUR DIOXIDE EMISSIONS
DATE SPRAY CHAMBER INLET.' SPRAY CHAMBER OUTLET
7-19-72
7-20-72
7-20-72
ppm SOC"
2540
2100
787
ppm
144
210
415
Weighted Average 1610 Weighted Average 235
-18-
-------
TABLE 11. Continued
C. FLUORIDE EMISSIONS
(CO LINE TO KILN)
RUN NO. DATE FLUORIDE EMISSIONS, (GR/DSCFd)
WATER SOLUBLE TOTAL
6-15-72 0.000630 0.000630
Orsat analysis
Determined by differences, includes other gases
Q
Parts per million by volume
Grains per dry standard cubic fee€
-19-
-------
TABLE 12.
AIR RETURN LINE
DATA SUMMARY
Run Number
Date
Volume of Gas Sampled - DSCFa
Percent Moisture by Volume
Average Stack Temperature - °F
Stack Volumetric Flow Rate - DSCFMb
Stack Volumetric Flow Rate - ACFMC
Percent Isokinetic
Percent by Volume C0_
£,
Percent by Volume O-
Percent by Volume CO
Percent by Volume N-
Fluoride Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
Fluoride Emissions, % insoluble
P2°5 Emissions - water soluble
mg Collected
gr/DSCF
gr/ACF
Ib/hr
P-Oj. Emissions - total
mg Collected
gr/DSCF
gr/ACF
lb/hrd
P20,- Emissions, % insoluble
1
7-19-72
5.73
2.90
700
16,900
36,400
-
0
21
0
79
79. 5d
0.214d
0.0995d
31.1
79. 6d
0.215d
0.0996d
31. ld
0.13d
1.2
0.00323
0.00149
0.48
1.7
0.00457
0.00211
0.68
29
2
7-19-72
5.72
3.36
' 700
16,800
36,400
See note
0
21
0
79
42.6
0.115
0.0530
16.5
42.7
0.115
0.0531
16.6
0.23
1.3
0.00350
0.00161
0.51
1.9
0.0051
0.0023
0.74
31.6
3
7-19-72
5.62
3.26
700 '
16,800
36,400
-
0 )
21 )
0 )
79 )
63.4
0.174
0.080
25.2
63.5
0.174 ;
0.080 .
25.2
0.16
1.4
0.00384
0.0018
0.55
2.0
0.0055
0.0025
0.79
30
Weighted
Average
5.68
3.18
700
16,800
36,400-
-
Assumed
values
based
on air
61.8
0.168
0.0775
24.3
61.9
0.168
0.0776
24.3
0.16
1.3
0.00352
0.00163
0.51
1.9
0.00506
0.00230
0.74
31.6
Dry standard cubic feet at 70 °F, 29.92 in. Hg.
Dry standard cubic feet per minute at 70°F, 29.92 in. Hg.
Actual cubic feet per minute.
' uoservod iaLerlur^nce in analysis (see Sampling & Analytical
Procedures Section).
Note: Run not conducted using isokinetic sampling procedures
due to small sampling port. -20-
-------
TABLE 13. SCRUBBER WATER ANALYSES - NODULIZING KILN SPRAY CHAMBER
Date
1972 Test No.
6/13 1 - Intake
6/13 1 - Discharge
6/14 2 - Intake
6/14 2 - Discharge
7/18 3 - Intake
7/18 3 - Discharge
Soluble Fluorides
mg/liter
27.8
38 avg =32.3
31
1200
480 avg = 779
657
42
40 avg = 41
40
549
610 avg = 602
647
21
33 avg =24
18
976
22 avg = 482
447
Soluble P-0-
• mg/ml b
0.1
0.02 avg = 0.05
0.02
0.04
0 .03 avg = 0.033
0.03
0.02
0 .02 avg =0.02
0.02
0.1
0.05 avg = 0.10
0.16
0.01
0.01 avg =0.01
0.01
0.09
0.04 avg = 0.06
0.05
PH
7.65
9.25
8.60
2.12
2.18
2.12
8.39
8.80
9.01
2.49
2.34
2.34
5.50
6 .00
5.60
2.40
3.50
2 .75
avg =8.50
avg =2.14
avg = 8.73
avg = 2.39
avg =5.70
avg = 2.88
-------
chamber was too low to measure with a pitot tube, the value
was calculated and based on the sum of the inlet flow and
the outside tempering air which entered the duct just before
the induced draft fan (see Table 5). Two outlet test sites
were used to measure fluorides and P^O-. These sites,
designated by test numbers BK and CK were 5.7 ft. apart
near the top of the spray tower (see Figure 6 , Page 38).
Sampling ports B, D, F, and H were associated with test
site BK, and Ports A, C, E, and G were associated with the
upper (downstream) site CK.
Table 5 summarizes the measured gas velocities at the
inlet, and the calculated velocities and volumes at the outlet.
Fluoride Evaluation (Water soluble) - Tests 1 and 2
with inlet loadings of 1.76 grains/DSCF and 1.72 grains/
DSCF, respectively, .as seen in Table 6, seem to be repre-
sentative of normal kiln operation. Carbon monoxide from
the furnace supplying fuel to the kiln and residual fuel
oil were normal during testing, and pH control of the
scrubber water by addition of lime was controlled satis-
factorily as shown by the data in Table 13. Testing was
interrupted by inclement weather (Hurricane Agnes) and
Test 3 was conducted a month later on July 18, 1972. The
following observations are pertinent to Test 3:
-22-
-------
1. Due to equipment limitations at the inlet site,
four points on the 12 point traverse were not
sampled and the remaining eight points were
used in obtaining the stack sample (see Appendix
C, Field Data).
2. Table 13 shows that the pH of the inlet scrubber
water was not well controlled during Test 3.
Measurements of 5.5 to 6.0 pH for Test 3 were
obtained as compared to 7.65 to 9.00 pH measure-
ments for Tests 1 and 2.
Inlet Test 3 is therefore not too representative of
normal operation, but is indicative of emission control
fluctuations which can be expected. Average grain loadings
for the water soluble fluorides averaged 1.95 grains/DSCF
before the spray chamber, and 0.0041 at the outlet. Total
fluorides measured at the inlet site averaged 434 pounds/
hour. The spray chamber was very effective in reducing
fluoride concentrations and achieved an average collection
efficiency of more than 99% with an average outlet loading
of 0.99 pound/hour at site BK.
The attempt at relating increased residence time of
the gas stream in the spray chamber to a reduction in
emissions was not conclusive since only slightly lower
values were measured on the average at the downstream
sampling site. The close proximity of the downstream
sampling site to the stack exit caused the flow and probably
the pollutant concentrations to be affected by the wind.
-23-
-------
P 0 Evaluation - Outlet B, Tests 1 and 2 appear repre
sentative of normal operation with average water soluble
P205 loadings of 0.0058 to 0.0086 grains/DSCF. Inlet
loadings of 0.02 to 0.13 grains/DSCF were measured at
this time. Carbon monoxide from the furnace supplying
fuel to the kiln and residual fuel oil usage were normal
during testing. The scrubber water pH was satisfactorily
adjusted to an alkaline range to control emissions. For
Test 3, the inlet total P-jO- loading is not reported. The
insoluble portion consisted of a very large amount of
particulate matter and upon fusion with sodium hydroxide in
the analysis procedure a gel was formed. The soluble
PpOj- for this test was slightly higher than for Tests 1
and 2.
Concentrations of water soluble P~0r averaged 0.08
grains/DSCF at the inlet and 0.009 at the spray chamber
outlet and appear to be representative of emissions for
this unit over extended periods of time. The spray chamber
collection efficiency for P^O,- removal was approximately
87% based on the data at Site B.
Table 8 summarizes the data obtained at the downstream
outlet site, CK. Average values for both fluorides and PpCv
are approximately 25% lower than those measured at Site B.
-24-
-------
However, one cannot state whether these lower concentrations
are caused by the increased residence time in the spray
chamber or by dilution with outside air which blew into the
chamber.
Isokinetic sampling rates at the outlet were generally
low for the fluoride runs. This was due to a variety of
factors including; sampling rates so low that orifice pressure
could not be read accurately; higher than expected moisture
content; and not accounting for dilution air added to vent
gas flow between inlet and outlet sites. This low isokinetic
sampling rate would tend to bias the test results on the high
side.
Particulate Evaluation - Tables 9 and 10 present the
particulate emission data obtained for the inlet and outlet
of the spray chamber.• Only the lower (BK) outlet sampling
site was used for particulate measurements and samples were
taken at a single point approximately six feet into the
stack at each port for a total of four sampling points.
The definition of true particulate in this type of gas
stream when sampled with a train using a glass probe and
fiber glass filter is very difficult because of the re-
activity of gaseous fluorides with the sampling equipment.
-25-
-------
The inlet tests are representative of uncontrolled
emission rates and agreed fairly well from test to test
and averaged 0.69 grains/DSCF for the front portion of the
sampling train (probe, cyclone, and filter). The "total
particulate" concentration averaged 0.83 grains/DSCF. The
outlet samples also agreed fairly well from test to test.
All of these tests were run at a sampling rate which was
approximately 25% too high; thus biasing the results on
the low side.. Outlet particulate concentrations 'averaged
0.053 and 0.099 grains/DSCF respectively for the front portion
(probe and filter) and total train fractions. The impinger
contents averaged 46.5% of the total particulate collected.
Excessively high moisture content of the stack gas
confirms the observed presence of entrained water droplets.
The spray chamber removed an average of 90% of the
particulate as determined by the front portion of the
sampling train.
Sulfur Dioxide Evaluation - Table 11 presents the SO
data obtained at the spray chamber inlet and outlet. Inlet
concentrations averaged 1610 ppm and the outlet averaged
235 ppm yielding a scrubber efficiency of approximately
85%. This efficiency was, however, very variable.
-26-
-------
Presented in Table 11 also are the fluoride results
from the CO line tests. Due to the high carbon monoxide
content of this line, and the potential presence of trace
amounts of phosphorus, a severe explosion hazard existed
when the sampling port was opened. For this reason, only
one test was run at this location. A very low fluoride
concentration of 0.00063 grains/DSCF was found.
Air Return Line
Table 12 summarizes the data obtained on the air
return line. These samples were taken by sampling at a
constant rate with a straight probe since little, if any,
visible particulate matter was present in this gas stream.
The concentrations of soluble fluorides in the air
return line range from 0.115 to 0.214 grains/DSCF. Soluble
phosphorus pentoxide concentrations range from 0.00323 to
0.00384 grains/DSCF.
Scrubber Water Samples
Presented in Table 13 are the analytical results of
the scrubber water samples, collected during the fluoride
and P?0r sampling at the nodulizing kiln spray chamber.
The lower pH of the scrubber water intake during Run 3 (5.7
-27-
-------
compared to 8.5 and 8.7 for Runs 1 and 2) did not result
in any significant change in scrubber efficiency for
fluorides and P?0r-
Three separate samples were taken during each test
period at approximately one-hour intervals.
-28-
-------
III. PROCESS DESCRIPTION AND OPERATION
Elemental phosphorus is produced from phosphate rock
by reduction in an electric arc-furnace. Typical ores
contain 10-13% phosphorus so that about 10 tons of rock
must be processed per ton of phosphorus produced. Con-
siderable quantities of coke, silica, and recycled materials
are fed to the furnace with the beneficiated ore. Table 14
presents a typical analysis for the fluoride content of
various materials used in the elemental phosphorus manu-
facturing process (see Figure 1).
Prior to being fed to the furnace, the ore is agglom-
erated and heat-hardened in a direct-fired rotary kiln.
The partially fused product is cooled and crushed to a
specified .size before being fed to the electric-arc furnace.
Aside from some particulates, the major emissions from
this feed preparation step are fluorides.
At Stauffer's Tarpon Springs Plant, the rotary kiln
is fired with a combination of CO from the phosphorus
furnace and residual fuel oil. The gaseous kiln effluent
is'passed through low pressure drop cyclones, diluted with
quench air, blown through a fan and scrubbed with water
in a spray tower. The scrubber water has an inlet design
pH of 7 to 7.5 which is maintained by lime addition.
Particulate collected in the cyclones is recycled to the
kiln feed.
-29-
-------
TABLE 14. FLUORIDE CONTENT OF
VARIOUS PROCESS STREAMS
PROCESS STREAM
FLUORIDE ANALYSIS
mg/gram
Kiln Feed
Coarse
Fine
Kiln Product
Furnace Phosphate
Rock Feed
Residual Fuel Oil
Kiln Nodules
Coke
Silica Sand
Furnace Slag
Ferro Phosphorus
Phosphorus Condenser Water
3.5 - 45
32
31
38
0.06
33
0.14
0.06
28.5
0.10
276 mg/liter pH 1.63
-30-
-------
Kiln operation is usually erratic. Not only do
mechanical and operating problems affect its performance,
but furnace problems are also reflected at the kiln. The
kiln uses CO directly from the furnace as a partial fuel
source. Because of the erratic furnace operation, the kiln
did not operate smoothly during testing. However, this is
representative of "usual" operating procedures and produc-
tion rates were reasonably close to design capacity.
Rock fed to the kiln is of two types. The two materials
»
are weighed (moist) into the kiln. An undetermined, variable
amount of undersized product is returned to the kiln from the
cyclone collectors.
The furnace feed is carefully proportioned with silica
and coke before being transferred to feed bins directly above
the furnace. The feed mixture then moves by gravity from the
bins down into the furnace as the furnace feed is consumed.
The reaction within the furnace is approximated by the
following equation:
(2300-2700°F)
2 Ca3 (P04)2 + 10 C + 6 Si02 — > P4 + 10 CO + 6 CaSiO
Elemental phosphorus and carbon monoxide leave the furnace
as a gas. Dust is removed from the stream by an electrostatic
precipitator and the phosphorus vapor is condensed out in
-31-
-------
direct-contact water condensers. Waste CO gas is used as a
fuel in the kiln operation or flared. Fuel oil is also used
for kiln fuel when the furnace is down, or insufficient CO is
available.
The non-gaseous by-products of slag and ferrophos (FeP.)
are periodically tapped from the furnace into open-pits and
quenched with water. Ferrophos is tapped into molds.
Furnace tapping operations are a source of particulate
emissions in the form of P2°5' coke smoke, and fluorides.
Hooding over the front of the furnace and over the slag runner
is satisfactory. However, the hood over the metal catch pot
is unsatisfactory since it is too small and poorly placed.
There is no hood over the ferrophos molds. Vent gases from
the hood system pass through a water saturator and then
through a venturi scrubber.
Make-up water for the scrubber is fed into the circulating
pump discharge. Total instrumentation on this system is a
pressure gauge on the water line to the venturi. The lack of
instrumentation combined with the make-up water feed method
makes estimates of the total water circulation rate difficult.
-32-
-------
IV. LOCATION OF SAMPLING POINTS
Slag Tapping
Figures 2, 3, and 4 show the slag tapping scrubber inlet
and outlet sampling sites. As shown, 16 points were used at
the horizontal inlet duct (8 along each diameter), and 36
points (18 along each diameter) at the outlet site. Both ducts
were circular.
Nodulizing Kiln
Figures 5, 6, and 7 show the sampling port locations on
the kiln spray chamber inlet and outlet respectively for the
Tarpon Springs Plant. The inlet site was located in a vertical
round section leading down to the induced-draft fan. Twelve
points were sampled along each diameter. The outlet sites were
in the top of the spray chamber, approximately 78 feet above
the inlet line. The two outlet sites were identical except
that site CK was 5.7" downstream from site BK. Sixty sampling
points were to be used at each outlet site. However, due to
probe length limitations caused by the narrow platform, Point
15 in each port was not sampled.
Particulate was sampled only at the lower (BK) site at the
spray chamber outlet, and since a rigid probe was used, only a
single point approximately 6' into the chamber was sampled at
each of the four ports.
-33-
-------
CLEAN OUT PORTS
B
o
H'6"
DUCT FROM
FURNACE HOOD
PLATFORM
I
OJ
TRAVERSE POINT
DISTANCES FROM
INSIDC OF STACK
(INCHES)
1.
2.
3.
4.
5.
6.
7.
8.
1.0
2.5
4.6
7.6
16.0
19.0
21.1
22.6
SATURATOR
TO VENTURI
SCRUBBER
Figure 2. Slag tap scrubber inlet samolina site
-------
SAMPLING
PORTS
TRAVERSE POINT
DISTANCE FROM
INSIDE OF STACK
(INCHES)
1 .
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.'
15.
16.
17.
18.
1.0
1 .6
2.7
3.9
5.2
6.7
8.4
10.6
13.7
22.1
25.2
27.3
29.1
30.5
31.9
32.1
34.3
34.9
Figure 3. Slag tap scrubber outlet sampling site
Stauffer Chemical Co., Tarpon Springs, Florida.
-35-
-------
Sampling
Ports
Outlet duct leading
from venturi scrubber.
Sampling ports were
later installed 3' from
top.
Sampling
Ports
Inlet duct leading from
hood inside building to
saturator
Figure 4. Slag tapping sampling sites.
-36-
-------
21'
i
CO
-J
I
ROTARY KILN .
CYCLONE DUST
COLLECTORS
-43.5".
O-O
W N
• SITE AK_
SAMPLING PORTS
10.5'
"\
OUTSIDE AIR INLET
N
t
TRAVERSE POINT
DISTANCE FROM
INSIDE OF STACK
(INCHES)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
1.0
2.9
5.1
7.7
10.9
15.4
28.1
32.6
35.8
38.4
40.6
42.5
TO
SPRAY CHAMBER
FAN
Figure 5. Kiln scrubber inlet sampling site
Stauffer Chemical Co., Tarpon Springs, Florida.
-------
TRAVERSE POINT
DISTANCE FROM
INSIDE OF STACK
(INCHES)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
2.1
6.9
11 .7
16.4
21.7
26.1
32.9
38.7
45.2
52.2
59.9
68.4
78.4
90.1
108.5
SAMPLING PORTS
SAMPLING
SITECK
A-_J 2 4 6 8 10 12 14
^••••t«» •••••• •
B:~ll 357911 13 15
H G
Figure 6. Kiln scrubber outlet sampling site
Stauffer Chemical Co., Tarpon Springs, Florida.
-38-
-------
Spray chamber. Outlet
sampling ports were
later located 3' and
8.7' from top of chamber
Inlet site showing
existing ports near
floor. Cyclone hoppers
seen in right foreground
Figure 7. Spray chamber inlet and outlet sampling sites
-39-
-------
Gas samples taken for Orsat analyses were collected by
traversing across the spray chamber inlet and outlet ducts.
Sulfur dioxide samples were taken at a single point in the
spray chamber outlet, and at two locations (one in each port)
at the inlet.
-40-
-------
V. SAMPLING AND ANALYTICAL PROCEDURES-
All sampling procedures were selected by EPA prior to
field sampling. All analyses of collected samples were also
performed by EPA. Details of the sampling procedures are
presented in Appendix D. At all sampling sites, narrow work
space and poor accessibility hindered completion of the work.
In addition, interruptions due to process malfunctions and
thunderstorms further delayed the test program.
Velocity and Gas Temperature
All gas velocities were measured with a type S pitot tube
and inclined draft gage. In all cases, velocities were measured
at each sampling point across the stack diameter to determine
an average value according to procedures described in the
Federal Register - Method 1. Temperatures were measured by
long stem dial thermometers, except at the spray chamber outlet
sites where a chrome1-alumel thermocouple was used because of
the large diameter.
Molecular Weight
An integrated sample of the stack gases was collected
D
during each run by pumping gas into a Mylar bag at the rate of
approximately 0.5 liter per minute. This bag sample was then
analyzed with an Orsat apparatus for C0?, O_ and CO.
Federal Register, Vol. 36, No. 247, December 23, 1971
-41-
-------
Fluorides and P00C
" •-——• -'^i—j
The basic train for all total fluorides and phosphorus
pentoxide samples is shown in Figure 8 and consisted of three
standard Greenburg-Smith impingers containing 100 ml each of
10% sodium hydroxide, one empty, straight tip impinger, a 3"
or 4" unheated Whatman #1 paper filter, and an impinger
containing approximately 200 grams of accurately weighed silica
gel. The impingers were contained in an ice-water bath. A
stainless steel nozzle and glass lined probe were used in all
cases. A flexible Teflon® line, approximately 7' long, connected
the probe to the first impinger when sampling at the slag tap
inlet and outlet sites, and when sampling for fluorides and
P^Oc at the kiln spray chamber outlet. A flexible connector
*— *J
was used since traversing with a rigid train was not possible
at these sites.
In m'd'st cases sampling was conducted under isokinetic
conditions either by calculating an average velocity and
sampling rate, or by monitoring the velocity with a pitot
tube and adjusting the sampling rate accordingly.
Train clean-up consisted of measuring the volume increase
of the impinger contents and silica gel weight gain, a triple
water rinse of all components from nozzle tip to filter holder,
and triple acetone rinse of all components. The impinger
a) This was required at the kiln outlet where the sampling
velocity was too low to measure.
-42-
-------
Glass Probe
Whatman No. 1 Paper Filter
CO
I
10%
NaOH
Solution I
in
impingers
100' ml 10° ml 10° ml
Greenburg- Smith Impingers
Dial Thermometers
Calibrated Orifice
Fine
Adjustment
Valve
Coarse
Valve
Vacuum
Gauge
Silica Gel
Ice Bath
Figure 8. Fluoride and
samPlin9 train.
-------
contents, filter, and water rinse were combined in one glass
sample jar and the acetone rinse was placed in a separate jar.
Particulates
Method 5 as described in the Federal Register, December
23, 1971, was used to measure particulate matter. This sampling
train is shown in Figure 9. At the kiln inlet, a cyclone was
used just before the filter and the duct was traversed for
only six of the 12 points along each diameter. At the outlet
site, one point was sampled in each of the four ports at the
'B1 test site. This sampling point was about 6' inside the
stack wall.
The train clean-up procedure included measuring the
water collected and weighing the silica gel to determine
moisture content. The probe, cyclone, and front half of the
filter holder were then rinsed with acetone and placed in a
container. The rear half of the train consisting of the rear
half of the filter holder, impingers, and connectors was
rinsed with water and the water then added to the impinger
contents. The rear half of the train was then rinsed with
acetone and the washings placed in a third sample jar. The
filter was placed in a separate container.
-44-
-------
FILTER
Ul
i
THERMOMETER
HEATED
GLASS
PROBE
HEATED I
SECTION I-
| J_1PJLRL_ OF WAILR
THERMOMETERS
UMDILICAL
CORD
CALIBRATED ORIFICE
MANOMETER
Figure 9. Particulate sampling train.
-------
Sulfur Dioxide
EPA Method 6 i-n the Fede r al Re giste r, December 23, 1971,
was used to measure S02. The sampling train is shown in
Figure 10 and consisted of a heated glass probe, a midget
bubbler (fritted glass tip) containing 15 ml of 80% isopropyl
alcohol, glass wool, two midget impingers containing 15 ml each
of 3% H202, and a fourth dry impinger. The probe washings and
the bubbler contents were discarded after each run. The midget
impinger contents were placed in a glass sample container.
These impingers were then rinsed with distilled water and
the washings placed into this same container.
-46-
-------
Heated Glass Probe
Glass Wool
80% Isopropanol
Volume-
15 ml
H2°2
Volume -
15 ml each
Empty
To Pump and Meter Assembly
Midget Impingers
Ice/Water Bath
Figure 10. Sulfur dioxide sampling train.
------- |