TEST NUMBER 72-MM-26
  FMC CORPORATION
  POCATELLO,  IDAHO
       PEDCo ENVIRONMENTAL

-------
                   PEDCo- ENVIRONMENTAL.
                        SUITE 8  •  ATKINSON SQUARE
                              CINCINNATI. OHIO  45246
                                        51 3 / 77 1-433O
        TEST NUMBER 72-MM-26

           FMC CORPORATION
          POCATELLO, IDAHO
            Prepared by:
        Norman Kulujian, P.E.
   and Richard W. Gerstle, P.E,
PEDCo-Environmental Specialists, Inc,
          Cincinnati, Ohio
  Contract No. 68-02-0237, Task 15

-------
                      TABLE OF CONTENTS

                                                Page Number

  I.  INTRODUCTION                                   1

 II.  SUMMARY AND DISCUSSION OF RESULTS             3

III.  PROCESS DESCRIPTION AND OPERATION            12

 IV.  LOCATION OF SAMPLING POINTS                  18

  V.  SAMPLING AND  ANALYTICAL PROCEDURES           21

      APPENDIX A -  DATA SUMMARY OF PARTICULATE,
      FLUORIDE AND  P?0  RESULTS WITH EXAMPLE
      CALCULATIONS

      APPENDIX B -  GASEOUS RESULTS WITH
      EXAMPLE CALCULATIONS

      APPENDIX D -  FIELD DATA

      APPENDIX F -  LABORATORY REPORT

      APPENDIX G -  TEST LOG

      APPENDIX I -  PROJECT PARTICIPANTS AND TITLES

-------
I.  INTRODUCTION




     Under the Clean Air Act of 1970, as amended, the




Environmental Protection Agency is charged with the estab-




lishment of performance standards for stationary sources




which may contribute significantly to air pollution.  A




performance standard is based on the best emission reduction




systems which have been shown to be technically and eco-




nomically feasible.






     In order to set realistic performance standards, accurate




data on pollutant emissions must be gathered from the station-




ary source category under consideration.






     The FMC Corporation plant in Pocatello, Idaho, was




designated as a well-controlled stationary source in the




elemental phosphorus reduction industry and was thereby




selected by GAP for an emission testing program.  This




plant has two briquet calcining kilns at this location,




which provide feed to four electric reduction furnaces.  Each




calciner is equipped with two parallel, low-energy scrubbers.






     Testing was conducted before and after the No. 1 scrubber




serving the No. 2 calcining kiln during the period of




September 20 to October 3, 1972.  Samples were collected to

-------
determine the filterable and total particulate emissions,



fluorides, P90 , and nitrogen oxides at each sampling
            ^ D


location.

-------
II.  SUMMARY AND DISCUSSION OF RESULTS




     A total of seven runs were made during the testing




period for the determination of total and filterable




particulates,  fluorides, P 05/ and nitrogen oxides.  A run




consisted of the simultaneous collection of an isokinetic




sample at the inlet and outlet of the No. 1 scrubber serving




the No. 2 calciner, and the collection of a sample in an




evacuated flask at each location for the determination of




nitrogen oxides.  The first three runs were for the deter-




mination of particulate emissions, while the remaining four




were for the determination of fluoride and P?0c emissions.






     The results of particulate, fluoride, and P 0  sampling




are summarized in Tables 1 through 4.  Scrubber efficiencies




are calculated in Table 5.  Table 6 presents nitrogen oxide




results.






     There is no explanation for the large difference between




inlet and outlet flows.  No pattern was established such as




the inlet or outlet consistantly displaying the higher value.




No adjoining duct was positioned between the inlet and outlet




sampling sites that allowed entry or exit of an air stream.






     The calciner is a continuous operation, but in some




instances problems arose with the operation or testing procedure




A discussion of events is described in further detail as follows

-------
                     TABLE 1.  SUMMARY OF PART±CULATE RESULTS
 Run  Number
Inlet, Scrubber No.1,  Calciner No.  2
                  1             2
 Date, 1972
 Volume  of  Gas  Sampled-DSCFa
 Percent Moisture  by  Volume
 Average Stack  Temperature-°F     ,
 Stack Volumetric  Flow  Rate-DSCFM
 Stack Volumetric  Flow  Rate-ACFM
 Percent Isokinetic
                 9/20
                57.68
                 0.16
                 539
               81,500
              182,000
                 105
  9/20
  56.81
  11.1
   518
 68,600
169,000
   122
  9/21
  49.57
   9.2
   517
 68,200
167,000
   107
                                 Avg.
 68,500
168,000
 Particulates-probe,  cyclone
       and  filter  catch

 mg
 gr/DSCF
. gr/ACF
 Ib/hr
 Ib/ton feed

 Particulates-total .catch

 mg
 gr/DSCF
 gr/ACF
 Ib/hr
 Ib/ton feed

 Percent  impinger  catch
                683.4
                0.183
               0.0817
                 128
               884.4
                0.237
                0.106
                 165
                 23.2
 3213.8
  0.873
  0.355
   513
 3679.1
  0.999
  0.406
   588
   12.6
 2444.1
  0.76
  0.311
   445
  2545
  0.791
  0.328
   463
   3.9
  0. 816
  0.333
    479
  0.675
  0.365
    526
  8.3
  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.
  Results from Run #1 not included in average

-------
             TABLE 2.  SUMMARY OF RESULTS, PARTICULATE EMISSIONS

                    Outlet, Scrubber 1, Calciner No.  2 .
Run Number
Date' 1972
Volume of Gas Sampled-DSCFa
Percent Moisture by Volume
Average Stack Temperature-°F    ,
Stack Volumetric Flow Rate-DSCFM
Stack Volumetric Flow Rate-ACFM
Percent Isokinetic '

Particulates-probe, cyclone
      and filter catch

mg
gr/DSCF
gr/ACF
Ib/hr
Ib/ton feed

Particulates-total .catch

mg
gr/DSCF
gr/ACF
Ib/hr
Ib/ton feed

Percent impinger catch
   9/20
  45.53
   16.5
   140
 64,300
101,000
   99.2
  199.9
  0.0676
  0.0428
   37.3
  415.6
  0.141
 0.0893
   77.7
   52.0
   9/20
  56.31
   16.7
   140
 85,700
135,000
   92.2
  591.6
  0.162
  0.102
   119
1043.4
0.285
0.181
 210
   43.1
              9/21
             60.43
              20.6
              140
            79,900
           134,000
              106
             187.4
             0.0478
             0.0284
              32.8
 568.6
 0.145
0.0865
  99.5
              66.9
                                                Avg.
                    140
                  76,600
                 123,000
                   0.0925
                   0.0577
                    63.0
                                0.190
                                 0.119
                                  129
                    54.0
  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.      .                 .

-------
              TABLE 3.   SUMMARY OF RESULTS,  FLUORIDE AND P  0   LOADING

                       Inlet,  No.  1.  Scrubber, .Calciner  No. 2
  Run Number

  Date,  1972
  Volume of Gas  Sampled-DSCFa  '
  Percent Moisture by Volume
  Average Stack  Temperature-°F    ,
  Stack  Volumetric Flow Rate-DSCFM^
  Stack  Volumetric Flow Rate-ACFM
  Percent Isokinetic
  Fluorides,  Total

  mg
  gr/DSCF
  gr/ACF
  Ib/hour
  P205,  Total

  mg
  gr/DSCF
  gr/ACF
  Ib/hour
9/22
52.21
8.13
473
72,100
167,000
. 107
12.22
0.00361+.
0.00156+
2.23+
24.30
0.00718+
0.00310+
4.44 +
10/02
29.16
4.46
576
77,600
190,000
N/A*
62.9
.03328
.01357
22.1
53.70
0.0284
0.0116
18.9
10/03
27.438
14.6
609
69,500
196,000
133
280.2
0.1576
0.5590
93.9
51.3
0.0289
0.0102
17.2
10/03
18.31
12.3
593
63,100
171,000
97.9
139.9
.11793
.04353
63.8
64.4
0.0543
0.0200
29.4
Aver a ere
    563
 70,600
181,000
  ,10294
  ,20537
  59.93
  .0372
  .0139
  21.83
b Dry standard cubic feet at 70°F, 29.92 in. Hg.
c Dry standard cubic feet per minute at 70°F, 29.92 in. Hg,
  Actual cubic feet per minute

* Two nozzle sizes .used-during test.
+ Not included in average calculations,  see discussion.

-------
           TABLE 4.  SUMMARY OF RESULTS, FLUORIDE AND P 0- EMISSIONS,

                     Outlet, Scrubber'No. 1, Calciner No. 2
Run Number

Date, 1972
Volums of Gas Sampled-DSCFa
Percent Moisture by Volume
Average Stack Temperature-°F    ,
Stack Volumetric Flow Rate-DSCFM^
Stack Volumetric Flow Rate-ACFM
Percent Isokinetic
Fluorides, Total

mg
gr/DSCF
gr/ACF
Ib/hour
P205, Total

mg
gr/DSCF
gr/ACF
Ib/hour


1
9/22
20.4
140
76,300
128,000
102
0.
0.
0.
0.
6.72
00187
00111
1.23
21.2
00591
00351
3.87
2
10/02
22.2
140
58,400
99,700
105
12.76
.00444
.00160
2.22
15.00
0.00522
0.00306
2.62
3
10/03
22.8
142
75,800
130,000
113
33.14
0.00893
0.00519
5.81
27.4
0.00738
0.00429
4.80
    4 .

  10/03

   22.4
    142
 55,600
 95,500
   94.3
   7.19
0.00297
0.00172
   1.42
                                                                       17.8
                                                                    0.00735
                                                                    0.00428
                                                                       3.50
                                                                                   Avg,
                                                                                   141
                                                                                66,500
                                                                               113,000
                                                                                 ,00647
                                                                                 00241
                                                                                  2.67
             00647
             00379
              3.70
  Dry standard cubic feet per minute at 70°F, 29.92 in. Hg.
  Actual cubic feet per minute

-------
             TABLE 5.  SCRUBBER EFFICIENCY DATA SUMMARY
Constituent
Run No.  Inlet(Ib/hr)  Outlet(Ib/hr)
Particulate,
total catch



1
2
3

165*
588
463

77.7
210
99.5
Efficiency
Fluorides,
total



P2O5, total


1
2
3
4 .
1
2
3
4
2.23*
22.1
93.9
63.8
4.44*
18.9
17.2
29.4
1.23
2.22
5.81
1.42
3.87
2.62
4.80
3.50

90.0
93.8
97.8
Average 93.9
86.1
72.1
88.1
                                                  Average  82.1
  Not included in average, see discussion.

-------
              TABLE 6.  SUMMARY OF NO  RESULTS, CALCINER NO.  2
TEST NO.
1A
IB
2A
2B
3A
3B
4A
4B
DATE
'1972
9/22
9/22
10/2
10/2
10/3
10/3
10/3
10/3
SITE
Inlet
Outlet
Inlet
Outlet
STACK GAS
VOLUME
SCFMa
72,100
76,300
77,600
'. ... .5:8. ,4,0.0 . ;. . .
Inlet | 69,500
g
Outlet | 75,800
i
Inlet
Outlet
63,100
55,600
NO EMISSIONS
PPM ^
272
260
271
396 .
308
294
280
86
LBS/HR
140
141
149
165
153
159
126
34b
Average

Average
Inlet

Outlet
142

155
At standard conditions of 70°F and 29.92" Hg. , dry basis.
Not included in average, process down when sample taken.

-------
     Particulate' Test 1                  •

     The inlet sampling team had problems with supporting
long glass probes in sample box in addition to thermal
expansion of probe.  Sample box clamp arrangement altered
to give probe more stability.  Broken probe may have allowed
ambient air to enter sample stream.  Results of inlet sample
are questionable.  When comparing moisture and particulate
loadings to results from Tests 2 and 3,  the values were
appreciably lower.  Therefore results from Test 1 are not
included in the summary (Table 5).

     Particulate Test 2

     Calcining kiln was down for 20 minutes during test.

     Particulate Test 3

     This test was run without any problems.

     Fluoride Test 1

     The original sampling train reagent, containing 10
percent NaOH, crystallized in the outlet impingers.  The
filters and impingers were replaced to complete the test.
Because of this difficulty, the results of this test are
felt to be in error and are not included in the averages.

     Fluoride Test 2

     The NaOH concentration was reduced to 2.5 percent to
avoid crystallization.  Substitution of a smaller diameter
nozzle on the inlet probe allowed reduced sampling rate
and eliminated the clogging problem.  The glass probe was
still breaking due to duct vibration or the thermal expan-
sion between the stainless steel and glass.  It was im-
possible to pinpoint the time of breakage.

     Fluoride Test 3

     No complications arose during this test.  However,
sampling rate was 33% too high at the inlet location.

     Fluoride Test 4

     Scrubber went down at beginning of test.  Flow rates
in the stack increased during the test.   The sampling rate
on the outlet train could not be maintained.  Filters were
repeatedly changed in an attempt to pull an adequate flow
rate.  Acetone rinse portion of the outlet sample was lost
in laboratory.
                            10

-------
     NO  Tests 1, 2, 3, 4

     No problems arose during the NO  tests except for
the invalid results obtained at the outlet on Test 4.
The EPA process engineer inadvertently told plant operator
that testing was complete and the process was shut down
before the sample was taken.  The scrubber did not remove
any nitrogen oxides.
                           11

-------
III.  PROCESS DESCRIPTION AND OPERATION




     The FMC Corporation plant at Pocatello, Idaho produces




elemental phosphorus for shipment to other locations for burning




or production of phosphorus trichloride.  The plant has two




calciners which provide feed to four electric reduction furnaces




in the form of calcined briquets.  In addition to phosphate ore




( 24% P^OC-) the furnaces are fed bituminous coke and silica.



Plant capacity is about 145,000 tons per year of P..






     Moist, crushed furnace feed is briquetted and fed onto a




continuous grate calciner where it is dried, partially fused,




and volatile components are driven off.  Most of the fluorides




present are driven off in this feed-preparation step.  Briquets




are carried at 3-5 feet per minute along the calciner.  Burning




CO gas  (from the reduction furnaces), and supplemental LNG as




needed, impinges on the 18-inch deep bed of briquets and heats




them to about 2400°F.  After going through the approximately




80-feet long calcining section, the briquets pass over a cooling




section of about the same length.  Air is pulled through the




bed of hot briquets and either returned to the calcining




section as preheated secondary air or exhausted-










     Each of the two calciners is controlled by two low-energy




scrubbers in parallel.  The two scrubbers are the same except




that excess secondary combustion air from the primary cooling




fan is vented to the No. 1 scrubber  at a point before the
                           12

-------
calcining fan.  Since the No. 1 scrubber serving the No.




2 calciner was tested, the following information applies




only to it.






     A 500 hp calciner fan pulls heated air and combustion




products through the bed of briquets.  The pressure drop




across the bed is 4-6 inches H?0 , depending on the feed




characteristics.  Exhaust air then enters the scrubbers at




a rate of 77-92,000 SCFM.  The scrubber operates essentially




at atmospheric pressure.  Each scrubber is followed by a




demister.  A 250 hp scrubber booster fan follows the demister,




and final exhaust is through an approximately 90 foot stack.






     The scrubber is equipped with three internal rotors  which.




throw the scrubbing solution up into the moving gas stream.




The CaO scrubbing solution enters the system at the demister




tower and immediately behind the No. 3 rotor.  Make-up water



is introduced under each rotor-and mixing of water-and slurry




is accomplished by the action of the scrubber rotors.






     The scrubber lime slurry is prepared at the No. 1 calciner




which is some distance from the scrubber tested.  The lime is




fed from a silo into a mixing tank, mixed with water, and




pumped to the scrubbers.  A timer-controlled valve on the




slurry lime cycles the scrubber feed.  Make-up water flow is
                           13

-------
continuous.  Attempts are made to hold the scrubbing liquor

pH at 5 to 8.  This is controlled by the speed of the slurry

feed timer cycle.  The control of the pH was not very well

attended during testing.


     In order to protect the lining in the scrubber, water

sprays are provided on the inlet ductwork to the scrubbers.

These sprays are activated when the temperature of the inlet

gas stream gets too high.  These sprays were off during all

except the first test.


     A total of 7 sampling runs were made.  A run consisted of

the simultaneous collection of a sample at the inlet and outlet

location.  Three runs were made to determine particulate

emissions and four were made to determine fluoride emissions.

One of these four runs was not completed and was discarded.

Nitrogen oxide samples were collected at each location during

each run.  The following discussion covers the process operation

during each run.


     Run 1.  September 20, 1972

     Just prior to the run, the No. 2 burner fan was taken
down for repair.  This unit was back in operation by the time
testing began.  From the beginning of the run at about 1100
hours until 1500 hours, the emergency water sprays in the
inlet scrubber duct were on.  Estimated flow rate was about
10 gpm.  From 1500 to 1530 the sprays were off.  The scrubber
pH was 2.3 until about 1430 when the slurry feed equipment
was repaired.  The scrubber pH then rose to above 7.'
                          14

-------
     Run 2, September 20, 1972

     At 1955 the grate drive motor overloaded and kicked
out for about 15 minutes.  Sampling was discontinued for
about 5 minutes of this period.  During this minor upset,
the sampling crew on the inlet side noticed a noisy passage
of coarse particles.  Also particulate collected on the
sampling train filter at this time appeared to be larger.
In such upsets, the calciner is not shut down.  Only the
burners are cut back to prevent overheating.  Consequently,
there is only a short lead time in returning to normal
when the grate is again put into operation.

     Run 3, September 21, 1972

     The pH when checked this morning was low.  Although
the operator's log did not reflect this, it had evidently
been low for some time.  Discussing this with plant personnel,
it seems that the lime slurry circuit- is prone to plugging
and that the plug is not discovered quickly.  After the low
pH was brought to the plant's attention, it was brought up
to 6.0

     The calciner was shut down for repairs to the feeder
at 1230.  The calciner was back in service at 1530.  Sampling
was begun at 1630 when feed rate and operating temperatures
were back to normal.

     During the test, the scrubber pH fluctuated from 2.3
to 5.5+.  During most of the run, the pH was 5.5+.  Litmus
paper was used for these pH measurements.

     Fluoride Run 1, September 22, 1972

     Scrubber pH was spot checked at 0830 and found to be
5.5+.  Operator's log showed the pH was 3-4.5 during the
preceding night shift.  Calciner was down due to malfunction
from 1030-1100.  The system was down again from 1130-1300
when a pallet jumped the track.  At 1310, the calciner was
back in operation.  Testing began at 1330.

     Fluoride Run 2  (discarded), September 26, 1972

     At 1555, testing was begun.  Although the process
appeared to be operating normally, there were constant
problems with the sampling trains.  Apparent clogging of
the sampling lines and erratic velocity profiles were
unexplained.  Because of these difficulties and inclimate
weather, the tests were terminated and the run was discarded.
                          15

-------
     Fluorides, Run 2 (Repeated),  October 2, 1972

     Scrubber water make-up had to be increased part way
through the test because no overflow was occurring.   Loss
of water around the No.  3 rotor shaft leaked enough make-
up water such that the rate was increased from 100 gpm to
140 gpm.  This rotor also appeared to be turning slower.

     A .small amount (less than 1%) of natural gas was fired
during this test.

     Litmus paper was used to periodically check the scrubber
water pH during testing.  While this indicated a comfortable
pH of 4, a check with a pH meter indicated it to be only
1.8-2.2.
     Fluorides Run 3 , October 3, 1972

     The pH was checked with litmus paper and it again
appeared to be high.  The No. 3 rotor was not working due
to either a frozen-up bearing or slipping belt.  In addition
water was being fed into the scrubber under the No. 2 rotor.
Normally it was fed in through the demister and a lower flow
is put in under the rotors.  The water line under the No. 1
rotor was cracked and useless.  These flow changes should
not significantly affect the scrubber performance.  Testing
began at 0945.  The pH was found to be low again.  Problem
was traced to the lime feed being shut off.  Feed was
restored at 1120.  A belt for No. 3 rotor was installed
half-way through the test, but the rotor was now buried in
sludge and would not turn.  The scrubber was taken down to
free the No. 3 rotor after testing was concluded.  At this
time a bearing was replaced.

     The only process upset occurred at 1145 when the grate
speed and heating was cut back 15 minutes due to a feed loss.

     Run 4, October 3, 1972

     All rotors were working and the leakage around the No.
3 rotor had been reduced.  Scrubber water pH was still
difficult to maintain.  First, the lime slurry pump went
out.  When this was repaired, the pH on the tested scrubber
did not increase.  In contrast, the pH in the No. 2 scrubber
was 7.6.  Having confined the problem to the scrubber, the
operator spent over an hour trying to get the pH up.  Success
was limited to keeping it at a barely acceptable level.
Towards the end of the run, the pH was being maintained
above 5.  the scrubber pH for the last three fluoride runs
are summarized below:
                          16

-------
TABLE 7.  SCRUBBER SOLUTION pH

Run No .
2
3
4

30
3.40
2.90
5.25
Minutes
60
2.25
3.50
4.50
into
90
2.35
2.25
3.25
Run
Composite
2 . 20
2.20
4.90
               17

-------
IV.  LOCATION OF SAMPLING POINTS




     Figure 1 shows the sampling locations at the calcining




kiln scrubber inlet and outlet.  Sampling points are noted




in Figure 2.  As shown, at the inlet of the calcining kiln




scrubber, the inlet samples were collected at 40 points, 10




points horizontally along each of 4 ports in the rectangular




horizontal duct.  The outlet samples were collected simul-



taneously at 32 points (16 along each diameter)  in the 79.5




inch diameter vertical exhaust stack.
                          18

-------
EXHAUST FROM
CALCINING KILN
NO. 2
                  18'
                   98",
                  15
4'
50.75"!
k

0
O
O
0
\SAN
one
                                         ROTATING  DISC
                                         SCRUBBER  NO.l
                             PORTS  (A.)
                            GROUND LEVEL
                 INLET SITE
                                                                               52'
                                                                                NE
                                                                                     O
                                                                                     NW
                                                        79.5"
                                                                                         SAMPLING
                                                                                         PORTS  (B
                                                OUTLET SITE
          Figure 1.   Calcining kiln scrubber sampling sites,  F.M.C.  Corporation,
                     Pocatello, Idaho.
          , : . .-,"»fe-.,»J^'' *«•

-------
SAMPLING TRA\
PORTS
. 6.25-^J
A 	 T 	
12.75"
+ _

-<
12.75"
+_

—
12.75"
D — L—


1
o


0


o


o
1
2
o


o


o


0

3
o


o


0


0

4
o


o


o


o

5
0


o


0


0

6
0


0


o


o

7
o


0


o


o

0
o


o


o


o

9
o


o


o


o

• 0
o


o


o


o

FROf
i




f

\
c

50. 75"








                                                                     TRAVERSE POINT DISTANCES
                                                                     FROM INSIDE OF STACK, inches
                                                                             4.9     6.   53.9
                                                                        2.   14.7   ,  7.   63.7
                                                                        3.   24.5     8.   73.5
                                                                            34.3     9.   83.3
                                                                        5.   44.1  '  10.'  -93.1
                        CROSS-SECTION OF KILN SCRUBBER INLET
to
o
'*•;.. .
                                                  TRAVERSE  POINT  DISTANCES
                                                  FROM  INSIDE  OF  STACK,  inches
                                                 1
                                                 2
                                                 3,
                                                 4.
                                                 5.
                                                 6,
                                                 7.
                                                        1 .3
                                                        3.9
                                                        6.8
                                                        9.9
                                                       13.4
                                                       17.5
                                                       22.5
                                                8.  29.8
 9.
10.
11 .
12.
13.
14.
15.
16.
49.7
57.0
62.0
66.0
69.6
72.2
75.6
78.2
           CROSS-SECTION OF  KILN SCRUBBER  OUTLET

            Figure 2.  Calcining kiln scrubber sampling  sites with  location  of  sampling
                                       points--F.M.C.  Corporation.

-------
V.  SAMPLING AND ANALYTICAL PROCEDURES




     Particulate sampling and analysis was conducted according




to Method 5, Federal Register, V. 36, N. 247, December 23, 1971.




In addition, the particulates collected in the impinger portion




of the train were recovered and analyzed for mass.  As specified




in this procedure, the stack gas velocity was determined by




Method 1, Federal Register, V. 36, N. 247, December 23, 1971,




and the average gas molecular weight was determined by Method




3 using an Orsat apparatus.






     Because of poor sampling accessibility at the outlet




location, a flexible unheated Teflon-connector, approximately




9 feet long, was used to connect the sampling probe to the




filter holder.  The connector was cleaned with the same brush




used to clean the heated glass probes at the outlet.






     The EPA project officer requested a water and acetone wash




on the front and back half of the sampling train.  This



added a probe water wash to the conventional method of recovering




the sample.






     Nitrogen oxides were determined by Method 7, F.R. December




23, 1971.  Fluorides and P-CU were determined by isokinetically




extracting a sample of the stack gas using the basic train and




procedure as described in Method 5 for particulates.  However,
                          21

-------
this train was modified such that the probe was followed by



three Greenburg-Smith impingers containing 100 ml of 10% NaOH,*



an empty straight-tip impinger, an unheated three-inch



Whatman #1 filter paper, and finally an impinger containing



approximately 200 gm of preweighed silica gel.  The impinger



train was placed in an ice-water bath.  As was the case during



the particulate sampling, a flexible Teflon connector was used



to connect the probe to the first impinger at the outlet



sampling location.  A diagram of this train is shown in Figure 3,






     Sample recovery consisted of measuring the volume increase



of the impinger contents and the silica gel weight gain, a



triple water rinse of all components from nozzle tip to filter



holder, and a triple acetone rinse of all components.  The



impinger contents, filter, and the water rinse were combined in



one glass sample jar, and the acetone rinse was placed in



another sample jar.






     In some instances the pump used to obtain an Orsat sample



was not properly working.  The EPA project officer agreed to



assume a molecular weight based on previous samples.





     Fluorides in the sample were determined by the Spadns-



Zirconium-Lake Method, and the P?0,_ content was determined by



the ammonium molybdate procedure.
*100 grams NaOH mixed with a liter of water.
                          22

-------
U)
STACK WALL

 Heated
 Glass  Probe
                                          Flexible Teflon
                                             Connector (outlet
                                                   train  only)
                                         Paper
                                         Filter
                                                        Thermometer
                              Manometer
                                                          Modified
                                                       G-S Impingers
G-S Impingers
                                           Thermometers
                                                               Control
                                                                    Valves    s—^Vacuum
                                                                                   Gauge
                                                                                              Check
                                                                                              Valve
                     Manometer ' =
                                                                   Umbilical
                                                                      Cord
                             Figure 3. Fluoride and P20C. Sampling train.

-------
                                  FILTER
                                        /   FLEXI
BLE TEFLON CONNECTOR
HEATED
GLASS
PROBE
                                             (OUTLET TRAIN ONLY)
                                                                        THERMOMETER

                               |_	I 100 ml .  OF WATER
                                       THERMOMETERS
                                          y
                                                                    L&-2K

                                                               ILICA GEL^
                             UMBILICAL
                             CORD
        CALIBRATED ORIFICE
      CONTROL
      VALVES
                                                                         VACUUM
                                                                         GAUGE

                                                                         -Hh
            MANOMETER-IE
                          Figure  4.   Particulate  sampling  train.

-------