73-FRT-7
(REPORT NUMBER)
AIR POLLUTION EMISSION TEST
U.S.S. AGRI-CHEMICALS
(PLANT NAME;
NASHVILLE, TENNESSEE
(PLANT ADDRESS)
U. S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Water Programs
Office of Air Quality Planning and Standards
Emission Standards and Engineering Division
Emission Measurement Branch
Research Triangle Park, N. C. 27711
-------
EPA REPORT NUMBER 73-FRT-7
U.S.S. Agri-Chemicals
NASHVILLE, TENNESSEE
Submitted to
Environmental Protection Agency
Office of Air Programs
Contract No. 68-02-0225
Task No. 15
Submitted by
Engineering-Science, Inc.
7903 Westpark Drive
McLean, Virginia 22101
November 1974
-------
I TABLE OF CONTENTS
Section Title Page
II INTRODUCTION 1
III SUMMARY AND DISCUSSION OF RESULTS 2
IV PROCESS OPERATION 12
V LOCATION OF SAMPLING PORTS 15
VI SAMPLING AND ANALYTICAL PROCEDURES 22
-------
LIST OF TABLES
Table Title Page
III-l Corresponding Changes of Ammonia Concentrations in
Exhaust Gas and Scrubber Liquor 7
III-2 Kiln Dryer/Cooler Baghouse Emission Summary 8
III-3 Ammoniator Scrubber Emission Summary 9
III-4 Kiln Dryer/Cooler Baghouse Flow Data 10
III-5 Ammoniator Scrubber Flow Data 11
LIST OF FIGURES
Number Title Page
IV-1 Schematic Diagram of Process, U.S.S. Agrichemicals,
Nashville, Tennessee 14
V-l Kiln Inlet to Baghouse 16
V-2 Kiln Baghouse Stack 17
V-3 Location of Sampling Ports: Ammoniator Scrubber Inlet
and Outlet 20
V-4 Traverse Point Locations 21
-------
II INTRODUCTION
Source emission tests were conducted at the U.S.S. Agri-Chemical phosphate
fertilizer plant, Nashville, Tennessee, during the week of May 21, 1973. The
purpose of the tests was to gather data upon which the U. S. Environmental Pro-
tection Agency can establish emission standards for new sources in the phosphate
fertilizer industry under Section 111 of the Clean Air Act of 1970.
This report describes the methods used and results obtained in the test
series. In total,twelve (12) samples were obtained for each of the following
emissions; particulates, ammonia and total fluorides. Samples were obtained from
two process operations, the dryer kiln/cooler baghouse and the ammoniator
scrubber. Simultaneous sampling of both the inlet and outlet gas streams of
each control device was conducted. Thus three samples were collected for
each of the pollutants at each of the following locations: kiln baghouse
inlet, kiln baghouse exit, ammoniator scrubber inlet and ammoniator scrubber
exit. The sampling locations are shown in Figure IV-1. In addition to these
gas stream samples, grab samples of scrubber liquid, phosphate rock and granu-
lated fertilizer product were periodically taken for analysis by EPA. Analyses
of the fluoride and ammonia samples were done by EPA chemists. An analysis of
the particulate samples was done by ES.
Section III of this report contains a discussion and summary of the
results of the tests. Subsequent sections describe the location of the sampling
ports and the sampling and analytical procedures. The Appendices contain all
the raw data collected on-site as well as the complete analysis of this data.
Mr. Robert M. Martin and Mr. J.E. McCarley, of the Office of Air Quality
Planning and Standards, Environmental Protection Agency, coordinated the tests
and arranged for the test facilities.
1
-------
Ill SUMMARY AND DISCUSSION OF RESULTS
The data obtained during the test was reduced to computer format for
analysis. The concentrations of particulate, fluoride, and ammonia in the
exhaust gases along with their respective emission rates were calculated.
Complete particulate results, along with a sample calculation, appear in
Appendix A, and complete results for fluoride and ammonia appear in Appendix
B. All results are expressed both in English and metric units. In accordance
with EPA's letter of June 20, 1973, included in Appendix J, emission rates are
reported in Ib/hr and Ib/ton feed. Summaries of the data appear in Table III-l
through III-5, and are discussed in more detail later in this section.
When the fertilizer plant changes production to a different type of
fertilizer, input feed rates must also change. While ES was testing, two
types of fertilizers were produced: 10-10-10 and 6-12-12, the numbers
standing for proportionate amounts of constituents. Feed rates for each type
of product, as reported by EPA in their letter of December 10, 1973, are
as follows:
NH- Feed Rates
10-10-10 1.20 tons/hr
6-12-12 2.45 tons/hr
P-0 Feed Rates
10-10-10 0.465 tons/hr
6-12-12 2.90 tons/hr
TOTAL FEED RATE
25 tons/hr
-------
Fertilizer 10-10-10 was produced from tests number 1 through 6 and 6-12-12
was produced during the remainder of the tests.
As described in Section V, the scrubber inlet (horizontal stack, see
Figure V-3) had three inches of water flowing in the lower quarter. For
flow rate calculations, an effective cross-sectional area was derived by sub-
tracting the area occupied by the water. An effective stack diameter was
then calculated from this derived area. The real diameter, 18 inches, was
reduced to 15.4 inches.
When reviewing the values for flow rate at the kiln and ammoniator,
discrepancies on the order of 25-35% between inlet and outlet were noticed.
EPA methods and standards were stringently adhered to and all calculations were
carried out properly. No evidence is available to formulate an explanation
for these discrepancies at the kiln. However, at the scrubber, system leakage
could have allowed dilution air to enter the gas stream.
When reviewing isokinetic values, a few are noted that are greater than
10% from isokinetic. For most cases where the isokinetic value was greater
than 110%, the average orifice pressure drop is approximately 0.40 in. H_0.
At this rate, the meter AH0 is about 1.7 instead of the value of 1.85 which
was used to set up the nomograph.
A. Particulate Analysis
Test No. 11-1, at the scrubber inlet, must be disregarded because a
quantity of mud and water, which was later found to be flowing in the stack,
had been inadvertently sampled. Test No. 11-0, at the scrubber stack must
be disregarded due to upset conditions at the scrubber which occurred suddenly
during the test allowing a large amount of particulate matter to be caught
by the sample train before the train could be shut down.
-------
The particulate emissions do not appear to be influenced by the type of
fertilizer being produced. Controlled filterable particulate emissions at the
baghouse stack range from .65 to 1.38 Ibs/hr. Controlled total emissions range
from 3 to 14 Ibs/hr. Controlled emissions as a function of feed rate range
from 0.13 to 0.56 Ibs/ton feed. Comparing inlet and outlet emission data at
the baghouse it is noted that filterable particulates are reduced greater than
99.9% in the baghouse; total particulates are reduced 99.7%. At the inlet
to the baghouse only 0.43% of the total particulate catch is found in the
impingers. At the outlet 86% of the total catch passes through the filter
and is caught by the impingers.
At the scrubber stack controlled filterable particulate emissions range
from 2.73 to 3.24 Ibs./hr. Controlled total emissions range from 4.25 to 7.73
Ibs/hr. Controlled emissions from the baghouse as a function of feed rate
range from 0.17 to 0.31 Ibs/ton feed. The scrubber is 97% efficient in re-
moving filterable particulates and 95% efficient in removing the total particu-
late load. At the scrubber inlet only 2% of the particulate catch is found in
the impingers, but at the outlet 47% of the catch passes through the filters
and is caught by the impingers.
B. Fluoride Analysis
Test No. 4-1 at the kiln baghouse inlet was voided because the sample
was spilled during analysis.
Controlled emission rates at the baghouse stack show a relationship to the
type of fertilizer being produced. Production of 10-10-10 fertilizer with a
P90_ feed rate of 0.465 ton/hr yields a controlled fluoride emission rate of
about 2.7 Ib/hr, while 6-12-12 production, with a larger P^ feed rate of
2.90 tons/hr yields smaller controlled fluoride emissions of 0.57 to 0.78
Ibs/hr. Controlled emission'rates as a function of feed show this same rela-
tionship with 5.8 Ibs/ton P205 feed being emitted during 10-10-10 production
4
-------
and 0.20 to 0.27 Ibs/ton ?205 feed being emitted during 6-12-12 production.
Controlled fluoride emission rates from the scrubber stack during 6-12-12
production range from 0.02 to 0.25 Ibs/hr, or 0.006 to 0.087 Ibs/ton P20_ feed.
No tests were conducted at the scrubber stack during 10-10-10 production,
therefore, no comparison can be made.
Checking inlet and outlet data, it is observed that the baghouse is
96% efficient in removing fluoride; at the scrubber, 46% of the fluoride
load is removed.
C. Ammonia Analysis
Ammonia samples were drawn non-isokinetically and no attempt was made to
measure stack velocity or flow rate. To calculate ammonia concentrations and
emission rates, the flow rates calculated from parallel tests were used. If
no parallel tests were made at the same time, then figures from tests conducted
on the same day were used. If no tests were conducted on the same day at that
particular test location, then an average value for flow rate was used.
During 6-12-12 production, controlled ammonia emission rates are higher
than during 10-10-10 production. At the baghouse stack, controlled ammonia
emissions are about 2.1 Ib/hr or 1.8 Ib/ton NH_ feed with an NH^ feed rate
of 1.20 tons/hr during 10-10-10 production, and 10 to 19 Ibs/hr or 4 to 8
Ibs/ton NH- feed with a higher feed rate of 2.45 tons/hr during 6-12-12
production. At the scrubber stack, controlled ammonia emissions during
10-10-10 production are 116 to 142 Ibs/hr or 97 to 118 Ibs/ton NH3 feed,
while during 6-12-12 production, controlled emissions are about 177 Ibs/hr
or 72 Ibs/ton NH3 feed.
Comparing the inlet and outlet data, it is noted that the baghouse reduces
ammonia emission by 88%. Negative efficiencies were noted at the scrubber.
-------
Engineering-Science, Inc. had followed the EPA test methodology for sampling
ammonia in exhaust gas streams. The results presented in this study indicate
that more gaseous ammonia was liberated from the ammoniator scrubber than was
measured in the inlet gas stream. In other words, by our calculations the
ammoniator scrubber seemed to add ammonia to the gas stream instead of re-
ducing or minimizing these emissions. ES has reviewed the data thoroughly and
are convinced that errors in transcribing, calculations, laboratory reporting,
and final writing did not result in the negative efficiencies projected.
The most probable cause for the results obtained for ammonia emissions
is the scrubber operation since gaseous ammonia can be absorbed in water or
some acidic solution. Changes in the inlet concentration may change the
equilibrium of the scrubber solution and thus may cause ammonia to be liber-
ated. The net effect of this situation the liberation of ammonia from
the scrubber water under certain conditions could have occurred during
the three tests for ammonia made on the ammoniator scrubber. As shown in
Table III-l, the change in ammonia concentration in the exhaust gas corres-
ponded to the change in ammonia concentration of the scrubber liquor. When
the concentration of ammonia in the exhaust gas decreased from inlet to
outlet, the ammonia concentration in the scrubber liquor increased. When
the concentration in the exhaust gas increased, the concentration in the
scrubber liquor decreased. The scrubber was probably operating at near
NH_ saturation level. It must be noted that the scrubber was primarily
designed to remove particulates, not ammonia.
-------
TABLE III-l
CORRESPONDING CHANGES OF AMMONIA CONCENTRATIONS
IN EXHAUST GAS AND SCRUBBER LIQUOR
Test No.
2
6
9
Liquor HN3 Gas NH3
Concentration Concentration
(mg/1) (gr/scfd)
In Out In Out
48.7 49.4 2.14 1.99
39.9 52.1 3.22 2.57
116.8 108.7 1.71 3.22
-------
TABLE III-2
KILN DRYER/COOLER BAGHOUSE EMISSION SUMMARY
Run No.
1
5
AVGa
7
4
8
10
AVGa
3
12
13
AVG
Pollutant
PT
PT
PT
PT
FL
FL
FL
FL
NH3
NH3
KH3
Product
10-10-10
10-10-10
10-10-10
6-12-12
10-10-10
6-12-12
6-12-12
6-12-12
10-1 10
6-12-12
6-12-12
6-12-12
FRONT HALF
In
gr/scfd
19.85
9.59
14.72
11.76
Out
gr/scfd
.008
.005
.007
.004
In
Ib/hr
4532
2096
3314
2817
Out
Ib/hr
1.376
.814
1.095
.649
In
Ib/ton
181.3
83.8
132.6
112.7
Out
Ib/ton
.055
.033
.044
.026
Eff
99.9
99.9
99.9
99.9
TOTAL
In
gr/scfd
19.91
9.65
14.78
11.79
*
.122
.045
.084
.052
.426
.697
.562
Out
gr/scfd
.079
.021
.050
.050
.015
.004
.003
.004
. .011
.112
.060
.086
In
Ib/hr
4547
2109
3328
2825
*
27.7
10.3
19.0
12.17
101.3
164.2
132.8
Out
Ib/hr
13.95
3.36
8.66
8.92
2.71
.778
.572
.675
2.11
18.86
10.45
14.66
In
Ib/ton
181.9
84.4
133.2
113.0
*
9.54
3.57
6.56
10.14
41.36
67.02
54.19
Oat
Ib/tor.
.558
.134
.346
.357
5.83
.268
.197
.233
1.76
7.70
4.27
5.99
Eff
99.7
99.8
99.8
99.7
*
97.2
94.4
95.8
82.7
81.4
93.6
87.5
do
*Sadple spilled during analysis.
- Average for specified product.
-------
TABLE III-3
AMMONIATOR SCRUBBER EMISSION SUMMARY
Run No.
11
14
15
AVG
16
17
18
AVG
2
6
AVG
9
Pollutant
PT
PT
PT
PT
FL
FL
FL
FL
NH3
NH3
NH3
Product
6-12-12
6-12-12
6-12-12
6-12-12
6-12-12
6-12-12
6^12-12
6-12-12
10-10-10
10-10-10
10-10-10
6-12-12
FRONT HALF
In
gr/scfd
__«._..
2.95
3.18
3.07
Out
gr/scfd
.__-»___
.068
.053
.061
In
Ib/hr
_«___
95.0
133.4
114.2
Out
Ib/hr
_____..
3%.24
2.73
2.99
In
Ib/ton
TEST
3.80
5.34
4.57
Out
Ib/ton
VOIDED
.130
.109
.120
Eff
.«.»_...
96.6
98.0
97.3
TOTAL
In
gr/scfd
r. . -.-
3.07
3.20
3.14
.005
.005
.005
.005
2.15
3.23
2.69
1.72
Out
gr/scfd
_.._««-*»..
.162
.083
.123
.001
.004
.001
.002
2.04
2.65
2.35
3.2?
In
Ib/hr
-.
98.9
134.2
116.6
.21:
.18;
.19!
.19?
84.0
113.0
98.5
66.0
Out
Ib/hr
7.73
4.25
5.99
.03!
.25'
.01;
.10:
115.9
142.0
129.0
177.5
In
Ib/ton
3.96
5.37
4.67
.073
.064
.069
.069
70.0
102,5
86.3
26.9
Out
Ib/ton
.31
.17
.24
.013
.087
.006
.035
96.6
118.3
107.5
72.4
Eff
X
92.2
96.8
94.5
82.0
-36. 8b"
91.5
45.6
-38.0
-25.7
-31.9
-169.
Although negative efficiency, the value was included in average.
-------
TABLE III-4
KILN DRYER/COOLER BAGHOUSE FLOW DATA
Test
No.
1-1
1-0
4-1
4-0
5-1
5-0
7-1
7-0
8-1
8-0
10-1
10-0
AVG-I
AVG-0
Flow Rate
(acfm)
38,949
29,576
38,931
29,628
37,949
27,334
40,002
29,094
39,629
28,915
38,193
28,752
38,942
28,883
Flow Rate
(scfm)
26,650
20,604
26,822
21,040
25,510
18,834
27,961
20,646
26,417
20,518
26,641
20,353
26,667
20,333
Moisture
C%)
8.71
9.85
8.95
8.85
10.63
10.43
7.47
8.44
11.68
9.16
7.36
10.31
Stack
Temp.(°F)
224
210
214
202
218
210
216
205
219
203
221
196
Isokinetic
(%)
105.7
101.1
116.4
96.7
114.2
102.5
110.4
101.0
118.0
101.6
111.3
99.2
Impinger
Catch (%)
.33
90.13
.65
75.76
.30
92.70
0.43
86.2
10
-------
TABLE 111-5
AMMQNIATOR SCRUBBER FLOW DATA
Test
No.
11-1
11-0
14-1
14-0
15-1
15-0
16-1
16-0
17-1
17-0
18-1
18-0
AVG-I
AVG-0
Flow Rate
(acfm)
5,066
7,579
5,348
7,753
6,148
8,115
6,056
7,965
6,003
8,795
5,577
8,073
5,700
8,046
Flow Rate
Cscfm)
4,036
5,567
3,755
5,558
4,899
6,004
4,687
5,744
4,785
6,808
4,391
6,132
4,426
5,969
Moisture
C%)
12.21
13.34
22.64
16.02
12.21
12.94
12.85
15.02
9.54
10.91
10.53
13.16
Stack
Temp. (°F)
110
143
110
138
110
141
123
142
129
130
130
126
Isokinetic
(%)
|
99.8
113.5
104.7
107.0
105.1
105.9
107.7
102.8
107.8
95.7
Impinger
Catch (%)
^«i
3.90
58.1
0.55
35.8
2.22
47.0
11
-------
IV PROCESS OPERATION
This plant produces different grades of fertilizer with ammonium sulfate as
the source of ammonia. The sulfate is shipped from Gary, Indiana, where it is
produced as a by-product of coke oven operation. Anhydrous ammonia and nitro-
gen solutions are also used as nitrogen sources to react with the phosphoric
acid. Since the phosphoric acid might not react with all the ammonia, sulfuric
acid is added to pick up excess ammonia.
The mixing of these materials (plus some others) is carried out in the
ammoniator which is basically a reaction vessel. Ammonia not absorbed in the
mixture escapes into the air causing air pollution. Other emissions from the
ammoniator are flourides and particulates.
The granules produced in the ammoniator are dried in a rotary dryer.
Cooling air is introduced into the rotary cooler to cool the hot material. The
cooler gases charged with particulates are drawn by a fan located downstream of
a cyclone which serves to remove the coarse material. The gases drawn by this
fan are discharged into the plenum of the dryer.
The dryer is operated by natural gas and has its own blower for supplying
air to the burner. Gases and particulates are withdrawn by an induced fan via
a baghouse and discharged into the atmosphere. The fan is centrifugal, driven
by a 100 HP, 1725 RPM Reliance Motor. The pulley ratio is 1:2, resulting in
about 3000 RPM on the fan shaft. The ID of the stack is 32.5 inches.
The granules are sized in two closed (not hooded) screens. The first,
called the hot screen, classifies the granules leaving the dryer. The fine
material is introduced into the cooler and coarse material into a crusher for
reduction in size and returned for screening to the same hot screen.
12
-------
From the cooler, the granules are discharged over a fine screen. Those
not passing through this screen represent the product. The fines going
through are recycled to the ammoniator. Raw material is fed to the ammoniator
at 25 TPH. The recycle-to-feed ratio varies between 0.5:1 to 1.5:1. That
means that total material handled in the ammoniator, recycle plus feed, ranges
between 40 and 60 TPH for a production rate of 25 TPH.
A schematic diagram of the process appears in Figure IV-1. This and
the process description were provided by EPA.
13
-------
SCHEMATIC DIAGRAM OF PROCESS, USS AGRICHEMICALS NASHVILLE, TENNESSEE
Test Locations
Test Locations
Phosphoric Acid
Ammonia
Ammonium Sulfate
H2S04
tr>
c/i
C5
Product
Ai r i
___ product Screen
Hot Screen^
Fines
Recycle
CD
73
-------
V LOCATION OF SAMPLING PORTS
Four different sampling locations were used during the tests; therefore,
each will be discussed individually.
A. Granulating Kiln Baghouse Inlet
Two 2-1/2 inch diameter sampling ports were located in the granulating
kiln effluent duct. The ports were located 90° apart and approximately 16 feet
downstream from the kiln discharge point. Figure V-l is a line sketch showing
the location and configuration of the sampling ports.
The kiln effluent duct to the baghouse is constructed of 1/8 inch steel
plate, 37 inches inside diameter. A small access platform was erected to
support the sample train during the horizontal traverse. The vertical port
was traversed by suspending the sample box on ropes and pulleys from the
overhead roof supports. During the vertical traverse, the sample box was
modified so that the probe extended downward vertically. Referring to
Method 1 (Federal Register; Vol. 36; n. 247; December 23, 1971), a total of
12 sampling points (6 on each axis) were selected.
B. Granulating Kiln Baghouse Exit
Two 2-1/2 inch diameter sampling ports were located in the baghouse
effluent stack, 90° apart approximately 30 feet above ground. Figure V-2 is
a line sketch of the baghouse, showing the location and configuration of the
sampling ports.
The baghouse effluent stack is constructed of 1/8 inch steel plate,
32-1/2 inches inside diameter. A3' x 8' platform extension from the upper
baghouse platform was constructed specifically for the project. Access to
the platform was provided by a vertical ladder to the ground.
15
-------
FIGURE V-l.
KILN INLET TO BAGHOUSE
Sampli ng
Port
2 1/2" Dia.
16
ENGINEERING-SCIENCE, INC.
-------
FIGURE V-2
KILN BAGHOUSE STACK
09
Motor
in
CM
32-1/2
r
Sampl1ng
Port
2 1/2" Dia.
Baghouse
17
ENGINEERING-SCIENCE, INC.
-------
The ports were located approximately 25 feet downstream from the baghouse
induced draft fan and 8 feet upstream from the effluent exit point. According
to Method 1, a total of 12 sampling points (6 on each axis) were selected.
C. Ammpniator Scrubber Inlet
Two 2-1/2 inch diameter sampling ports were located in the horizontal
section of the scrubber inlet duct about 25 feet above ground. Figure V-3
is a line sketch of the venturi scrubber, separation tank, fan and stack
showing the location and configuration of the sampling ports. The horizontal
orientation of the duct necessitated vertical sampling in one of the ports.
A system of ropes and pulleys was used to support a standard EPA sample train
box which had been modified for this purpose.
The scrubber inlet duct is constructed of 1/4 inch fiberglass, 18 inches
inside diameter. A 4' x 4' working platform was constructed specifically for
the project. Platform access was provided by a vertical ladder down to a
previously existing platform on the kiln baghouse structure. Access to this
platform was by a vertical ladder to ground level.
The ports were located 4 feet upstream from a 90° elbow and 10 feet
I
downstream from the nearest turn. According to Method 1, a total of 12 sampling
points (6 on each axis) were selected, Later, this was reduced to only the
6 points on the horizontal access due to the extremely large quantity of water
found to be flowing down the lower quarter diameter of the duct. The water was
found upon inspection to be at least three inches deep in the bottom of the duct,
thus negating the possibility of sampling the lower 3 points on the vertical
axis. Testing continued after the EPA project officer approved the plan to
utilize only the six horizontal axis points. Figure V-4 shows all traverse
point locations for all stack ports. In all cases, the points are numbered
consecutively, from 1 to 6 with the first point nearest the sampling port.
18
-------
D. Ammoniator Scrubber Outlet
Two 2-1/2 inch diameter sampling ports were located in the ammoniator
scrubber stack, 90° apart and approximately 30 feet above ground. The loca-
tion and configuration of the sampling ports are also shown in the line sketch,
Figure V-3. The ammoniator scrubber stack is constructed of 1/4 inch fiber-
glass, 24 inches inside diameter. A 61 x 61 working platform at the building
roof line already existed before the test and no modifications or additions
were required to the structure. Access to the platform was provided by
vertical ladders to a lower roof top and then to the ground.
The ports were located 27 feet downstream from a forced draft fan and
15 feet upstream from the effluent exit point. According to Method 1, a
total of 12 sampling points (6 on each axis) were selected.
19
-------
FIGURE V-3
LOCATION OF SAMPLING PORTS
AHHONIATOR SCRUBBER INLET AND OUTLET
1/2" Dia
Induced Draft Fan
20
ENGINEERING-SCIENCE, INC.
-------
FIGURE V-4
TRAVERSE POINT LOCATIONS
KILN OUTLET
to Baghouse
KILN STApK
from Baghouse
SCRUBBER INLET
SCRUBBER STACK
21
ENGINEERING-SCIENCE, INC.
-------
VI SAMPLING AND ANALYTICAL PROCEDURES
The sample and velocity traverse points were selected as described in
Section V. All participate and fluoride tests, with the exception of the
ammoniator scrubber inlet, included a traverse of both axes in the stack with
ten-minute sampling times at each point. Sampling data, however, was reported
every five minutes, twice for each sample point. As described in Section V,
the excessive quantity of water flowing down the ammoniator scrubber inlet
duct necessitated sampling on only the horizontal axis. Sampling times and
sequences were suggested or approved by the EPA field officer.
Heavy particulate loading was experienced at times, expecially on the
scrubber inlet and scrubber stack, necessitating a test shutdown, change of
filter, and a test restart. This is noted in the test log in Appendix G.
A. Particulate Sampling
The isokinetic sampling of particulate matter was performed using an EPA
approved sampling train. The probe tip in test 11-1 was changed to a smaller
size in the middle of the test to keep the test isokinetic. The probe size
used in calculations was a weighted average. Water was used instead of acetone
for the front half wash after each particulate test. Since fertilizer is
more soluble in water than acetone, a water wash would lead to a more efficient
recovery of particulate from the sample train. Particulate sample catch was
determined by Method 5, Standards of Performance for New Stationary Sources.
B. Fluoride Sampling
The sampling of fluorides was conducted isokinetically with a modified
sampling train. An unheated filter holder was placed between the third and
fourth impingers and contained'a Whatman No. 1 paper filter. The rest of the
22
-------
train remained the same as a standard particulate train. Sampling times were
two hours with data points recorded every 5 minutes. AH fluoride catch was
emptied into one bottle. Analysis consisted first of fusing insoluble fluoride
with sodium hydroxide and distilling the samples from sulfuric acid to remove
interferences, then a portion of the distillate was reacted with SPADNS
reagent. The fluoride was determined spectrophotometrically by its bleaching
effect upon the reagent dye. This procedure is described further in Appendix E.
C. Ammonia Sampling
Ammonia sampling was carried out non-isokinetically with the probe in-
serted halfway into the stack port. Sampling times were 60 minutes with data
recorded every 15 minutes. A modified sampling train was used in which the
ammonia was absorbed with dilute sulfuric acid in the first two sample
train impingers. The rest of the train remained the same as a standard
particulate train. A modified Kjeldahl distillation procedure recovers the
ammonia as ammonium borate, and subsequent titration with standard sulfuric
acid solution yields the acid equivalent to isolated ammonia. This procedure
is described further in Appendix E.
23
------- |