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.

-------