ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF ENFORCEMENT
EPA-330/2-78-008
Cotton Gin Emission Tests
Marana Gin
Producers Cotton Oil Company
Marana, Arizona
(NOVEMBER 2-1 9, 1 977)
NATIONAL
ENFORCEMENT INVESTIGATIONS CENTER
DENVER. COLORADO
AND
REGION IX, SAN FRANCISCO I
MAY 1978
-------
Environmental Protection Agency
Office of Enforcement
EPA 330/2-78-008
COTTON GIN EMISSION TESTS
MARANA GIN
PRODUCERS COTTON OIL COMPANY
Marana, Arizona
[November 2 - 19, 1977]
May 1978
National Enforcement Investigations Center - Denver
and
Region IX - San Francisco
-------
CONTENTS
I INTRODUCTION 1
II SUMMARY AND CONCLUSIONS 5
III PROCESS DESCRIPTION 7
IV CONTROL EQUIPMENT DESCRIPTION 12
LINT CAGES 12
CYCLONES 16
V TEST PROCEDURES 19
CYCLONE SAMPLING PROCEDURES 19
LINT CAGE SAMPLING PROCEDURES 24
AMBIENT AIR SAMPLING PROCEDURES 27
VISIBLE EMISSION OBSERVATION PROCEDURES 30
PROCESS OBSERVATION PROCEDURES 30
VI TEST RESULTS 32
CYCLONE RESULTS DISCUSSION 41
LINT CAGE RESULTS DISCUSSION 42
REFERENCES 44
FIGURES
1 Marana Gin Plot Plan 2
2 Short-Staple Cotton Gin Process Flow 9
3 Long-Staple Cotton Gin Process Flow 10
4 Short-Staple Gin Emission Points 13
5 Long-Staple Gin Emission Points 14
6 Lint Cage Diagram 15
7 Cyclone Involute System 22
8 Lint Cage Sampling Arrangement 26
9 Hi-Vol Sampling Locations 28
TABLES
1 Lint Cage Construction Details 17
2 Cyclone Design Data 18
3 Sampling Location Data 20
4 Process Data Summary 33
5 Emission Data Summary, Short-Staple Gin 34
6 Emission Data Summary, Long-Staple Gin 35
7 Visible Emission Observation Summary 37
8 Ambient Air Data Summary 39
9 Cotton Analytical Results 40
-------
APPENDICES
A Presurvey Inspection Report
B Calibration Procedures and Data
C Sampling Train Construction Details
D NEIC Analytical Procedures and Data
E Lint Cage Dimensions, Flow Data and Calculations
F High-Volume Sampler Reference Procedures
and Calibration Data
G Cotton Gin Production Data
H Agricultural Research Service Trash Analysis
Analytical Procedure and Data
I Chain-of-Custody Procedures and Records
J Emission Data Summary Sheets
K Field Data Sheets and Example Calculations
-------
-------
I. INTRODUCTION
The Environmental Protection Agency (EPA), Region IX, requested
the National Enforcement Investigations Center (NEIC) to source test
the Marana Gin of Producers Cotton Oil Company, Marana, Arizona.
NEIC was to determine if the gin was in compliance with the Pima County
Regulation II, Rule 2 which limits the particulate emissions based on
process weight, i.e., seed cotton feed rate.
The Marana Gin consists of two gins—one for processing short-
staple cotton,* and the other for processing long-staple (Pima)
cotton—with production capacities of 10 and 4 bales/hr, respectively
[Figure 1]. The two gins are similar, each being separated into a
high-pressure and a low-pressure side, so named because of the process
fan capacities. The high-pressure side includes the cleaning, drying
and ginning of seed cotton. Particulate emissions from these sources
are controlled by cyclones. The low-pressure side includes cleaning,
compressing and baling of lint cotton, with the emissions being con-
trolled by lint cages.
NEIC performed a presurvey inspection of the Marana Gin on
November 17, 1976, to determine the feasibility of source sampling,
and to evaluate process operations and control devices [Appendix A].
This inspection determined that: 1) the cyclones could be sampled if
the outlets were modified to provide acceptable sampling locations;
Short-staple cotton fibers range in length from 2.4 to 2.7 cm
(0.94 to 1.06 in) as compared to long-staple cotton fibers which
range from 2.8 to 3.8 cm (1.09 to 1.5 in).
-------
N
/I
storage
Wagon
Storage
Area
Lint Cages
Long Staple Cotton Gin
cyclones
seed
storage
Figure 1
Wagon Storage Area
cyclones
storage
cyclones
Unloading
Lint Cages
© ©
Short Staple Cotton Gin
©
cyclones
Bale
Cotton
Storage
Area
seed
storage
Producers Cotton Oil Company
Marana Gin Plot Plan
-------
3
2) the seed and mote* cyclones could not be sampled because of their
inaccessibility; and 3) the low-pressure side lint cages could not be
source tested using EPA Method 51 procedures.
During the preliminary stages of developing the source test pro-
gram, questions arose as to whether a suitable sampling methodology
existed. A meeting was held in Dallas, Texas on January 5, 1977 be-
tween the U.S. Department of Agriculture (USDA), Agricultural Research
Service (ARS); the Cotton Ginning Council of America; and the EPA
(Region IX and NEIC) to discuss sampling procedures for measuring
cotton gin cyclone emissions. It was decided that NEIC and ARS would
undertake a comparability study between the EPA Method 5 and the ARS
high-volume sampling procedures. The results of this comparability
study2 showed that Method 5 could be used to test cotton gin emissions,
provided a large sampling nozzle (>1.3 cm or 0.5 in) was used. The
Aerotherm Corporation's high-volume source sampler (HVSS) was selected
to test the cyclone emissions because it meets all the EPA Method 5 testing
requirements using a large sampling nozzle.
The HVSS was not considered suitable to sample the lint cage
emissions because the air-flow rates were below that detectable by an
S-type pitobe assembly, and the HVSS does not have the geometric flexi-
bility required. The Rader** high-volume (HV) sampler was selected
to test the lint cage emissions because it does have the flexibility
requi red.
On September 27, 1977, EPA personnel met with Marana Gin officials
to discuss the modifications required to enable testing of the cyclone
outlets and to discuss sampling procedures for the lint cages.
Motes are a group of fibers attached to a portion of seed hull.
Brand name.
-------
4
Between November 2 and 19, 1977, the short- and long-staple
cotton gins were source tested. In addition to the emission testing,
ambient air particulate sampling was performed with high-volume
samplers (Hi-Vols) to determine if the gin emissions had an impact on
the air quality in immediacy jiTiiaoions were
evaluated in accordance with Pima County Regulation II, Rule 1 which
limits the emissions to 40% opacity.
-------
II. SUMMARY AND CONCLUSIONS
The lint cages and cyclones of the short- and long-staple cotton
gins of Producers Cotton Oil Company, Marana, Arizona, were source
tested to determine compliance with Pima County Regulation II, Rule 2.
During the November 2 to 19, 1977 survey the short- and long-staple
gins operated at 86 and 85% capacity, 8.6 and 3.4 bales/hr, respec-
tively. The seed cotton that was ginned was machine-picked cotton.
According to Rule 2 the short-staple gin is allowed to emit par-
ticulate at 5.1 kg (11.2 lb)/hr. The cyclone and lint cage emissions,
listed below, totaled 13.4 kg (29.5 lb)/hr, 164% greater than the
allowed emissions.
Particulate
Emissions
Actual
Allowed
Source
kg/hi
r lb/hr
kg/hr
lb/hr
Cyclones
7.23
16.0
-
-
Lint Cages
6.13
13.5
-
-
Total
13.4
29.5
5.1
11.2
The short-staple gin emissions determined by the source testing
are conservative, because: 1) the mote cyclone emissions were not
tested, therefore not included, and 2) the lint cage testing was
conducted over the acceptable isokinetic sampling rate, thus emission
results are less than the actual emissions.
-------
6
Particulate emissions from the long-staple gin are limited to
2.7 kg (6.02 lb)/hr. Lint cage and cyclone emissions totaled 3.40 kg
(7.50 lb)/hr, 25% greater than the allowable.
Particulate Emissions
Actual Allowed
Source kg/hr lb/hr kg/hr lb/hr
Cyclones 1.84 4.06
Lint Cages 1.56 3.44
Total 3.40 7.50 2.7 6.02
The source testing results of the long-staple gin emissions are
also conservative because the seed cyclone emissions were not included
in the total, and the lint cage emissions results are less than the
actual emissions for the reasons discussed above.
Visible emissions observed from the short-staple gin cyclones were
less than the Pima County Regulation II, Rule 1, opacity limit (40%).
-------
III.
PROCESS DESCRIPTION
Both short- and long-staple gins process seed cotton in a similar
manner. Pneumatic or air conveying is the principal means of moving
cotton within each gin. Approximately 725 kg (1,600 lb)* of seed
cotton is required to produce a bale of lint cotton (227 kg or 500 lb).
The weight difference (498 kg or 1,100 lb) is the weight of trash and
cottonseed removed during processing.
Seed cotton is usually delivered to a gin in a wagon, and unloaded.
This seed cotton is alternately dried and cleaned twice before being
ginned. Lint cotton, cottonseed and trash are separated during ginn-
ing, after which the lint cotton is cleaned again before being baled.
The condition of the seed cotton, a function of the picking pro-
cedures, affects process operations and emissions. Cotton is generally
picked twice from the defoliated stalks which accounts for 90 to 95%
of the cotton ginned. Picked cotton contains some trash (stalks,
leaves and dirt). Subsequent to the second pick, a vacuum sweeper is
used to pick up cotton including trash that has fallen to the ground
(ground-cotton). Second-pick cotton, which has slightly more trash
than first-pick but much less than ground cotton, was reportedly
ginned during the November 2-19 survey.
At the Marana Girt it was computed to take 662 kg (1,460 lb)
of seed cotton to produce a bale of lint cotton.
-------
8
The seed cotton moisture content also affects the process opera-
tions and emissions. The moisture content is a function of the mode
of harvesting (i.e., picked or ground cotton) and the time of harvest-
ing (i.e., morning or afternoon, and before or after a rain). A 6 to
8% moisture content is considered ideal for processing; less than 6%
results in cotton being removed with the trash, increasing the loading
to the cyclones. Conversely, only a small amount of trash will be
removed when the moisture content is greater than 8%, resulting in a
poor grade of bale cotton. Detailed process information for both
Marana gins follows.
The seed cotton is off-loaded from a wagon with a vacuum system
and collected in a separator [Figures 2 and 3]. The cotton then enters
a natural-gas-fired reel drier. The reel drier is a horizontal rotary
drier that also removes some trash as the cotton passes through the
unit.
This cotton is air-conveyed to a 7-cylinder inclined cleaner
(the long-staple gin has a 5-cylinder inclined cleaner). The cleaner
consists of 7 (5) spiked drum cylinders which carry and "scrub" the
seed cotton to remove fine particles such as leaves, dirt and sand.
Next, a bur extractor removes heavy trash and burs from the cotton
with a revolving saw cylinder. The teeth of the saw hold the seed
cotton and subject it to a carding and cleaning action as the cotton
is spread across the surface of the cylinder.
The cotton is then fed into a 24-shelf natural-gas-fired tower
drier. The cotton, while falling from shelf to shelf, is dried by a
concurrent stream of warm gases. The cotton moisture content at the
exit is normally between 6 and 8% This dry cotton is air-conveyed
to second-stage inclined cleaners (the short-staple gin has two cleaners
-------
9
Baled Motes
~
i
i
Seed Cotton
Baler
Separator
Baled
Cotton
Inclined
Cleaner
Reel
Drier
Baler
! Mote
iCyclone
Inclined
CIoanor
¦> -
Condenser
Bur
Extractor
Tower
Drier
LEGEND:
Cotton—
Inclined
Cleaner
Inclined
Cleaner
Motes
Lint Cage
Saw Gin
"eeder
(5 Feeder - Gin Units)
Figure 2. Short-Staple Cotton Gin Process Flow
-Producers Cotton Oil Company
Marana Gin
-------
10
Seed LEGEND:
Cotton
I Cotton
Baled Cotton
2nd Stage
Lint
Cleaner
1st Stage
Lint
Cleaner
Feed
Feed
Battery
Condenser
Baler
Separator
5 Roller
Gins
5 Roller
Gins
Inclined
CIeaner
Reel
Drier
Tower
Drier
Bur
Extractor
Figure 3. Long-Staple Cotton Gin Process Flow
Producers Cotton Oil Company
Marana Gin
-------
11
in parallel, the long-staple gin has one). These inclined cleaners
are identical to the one previously described.
Then, the cotton is distributed to five extractor feeders in
parallel (the long-staple gin has ten) where additional burs, stems,
whole leaves and other trash are removed. The cotton is fed at a
constant rate to five 80-saw-gin stands (10 roller gins in the long-
staple gin). Excess cotton is returned to the second stage cleaners.
The saw-gin stands use saws and air blasts to separate seeds and
trash from the cotton; the roller gins use revolving cylinders to
separate the seeds and trash from the lint. The resulting lint cotton
is air-conveyed to two lint cleaners operated in parallel (two sets
of two lint cleaners are used in the long-staple gin). The lint
cleaners remove fine particles, motes, dust and sticks. The clean
lint cotton is air-conveyed to the condenser where it is formed into
a smooth endless mat and fed into the baler. The bales are wrapped
in burlap, tied with metal bands and stored.
The motes and lint cotton removed with the trash from the lint
cleaners and condenser are recovered in the mote cyclone. The motes
are then cleaned in an inclined cleaner, and baled. Motes are not
recovered in the long-staple gin.*
The long-staple cotton reportedly contains only 1 to 2% motes
while short-staple contains 10 to 12%.
-------
IV. CONTROL EQUIPMENT DESCRIPTION
All the trash removed from the cotton is pneumatically conveyed
out of the gin building where cyclones and lint cages remove the trash
from the air streams. Eight lint cages control the short- and long-
staple gin low-pressure side emissions [Figures 4 and 5]. The trash
collected by the lint cages is air conveyed to the mote cyclone or
Station 2306 cyclones (short- and long-staple gins, respectively).
Twenty-eight cyclones control the short- and long-staple gin
high-pressure side emissions [Figures 4 and 5]. These cyclones
operate in groups of twos and threes. In the short-staple gin, all
collected trash is conveyed to three cyclones (Station 2211) which
have a screw conveyor at their base to dump the collected trash into
a wagon.
LINT CAGES
The lint cage consists of a screened cylinder [Figure 6] where
the gases enter from the top and vent through the entire cylindrical
circumference with a swirling flow. The collected particulate (mostly
lint) sloughs off the inside of the screen and falls to the bottom of
the lint cage.
The lint cages are supported from the ground and, with the excep-
tion of Station 2203 [Figure 4], are located against the outside wall
of the gin buildings, exposed to the weather. Station 2203 lint cage
is partially enclosed by a shed on three sides and on top, so that
the gases are vented through a 2 x 2.2 m (6.5 x 7.2 ft) opening.
-------
SN 2211
1
1
1
1
1
1
1
1
i
1
I
1
I
1
i
i
Station
Numbers
2204
2205
2205
2207
2203
[oooooo oo oo
O
Hote cyclone
©]
SN 2203
©
SN 2202
Short staple gin building
LEGEND:
O - cyclone-.
© - lint cages
V - cyclone pair
VV - cyclone tnpl
SN - station numbe
N
/I
SN 2201
L°^°JSM 2209
Figure 4. Short-Staple Gin Emission Points
Producers Cotton Oil Company
Marana Gin
-------
LEGEND:
f—
Csl
ro
LO
0
O
0
O
O
0 -
**0
ro
ro
ro
ro
Oi
CNJ
CM
CM
C\J
V -
:z:
SI
2:
s:
3-
SN -
CO
00
00
CO
co
©
©
©
©
©
cyclone pair
Pima (long staple) cotton gin buildi
ng
N
A
C >
oi- r-v co cr> o
•r- QJ o O CD ,—
+-> -Q ro ro co ro
JO E CM CVI N CM
<-n 2: i—A A A
"loo
00000 o
00
Seed cyclone
o
ro
C\J
Figure 5. Long-StaDle Gin Emission Points
Producers Cotton Oil Company
Marana Gin
-------
Isometric
View
Access Door
Inlet
V
Screened Secti
/
Outlet
Figure 6. Lint Cage Diagram
-------
16
The three short-staple gin lint cages, each have 28 sections
approximately 1.2 x 0.4 m (3.9 x 1.2 ft) in two bands of 14 sections
each. The 3.6 m (11.7 ft) high cage has a third band of solid metal
between the two screened bands [physical dimensions of each lint cage
are listed in Table 1].
The five lint cages of the long-staple gin are 1.1m (3.6 ft)
above the ground, 0.5 to 0.9 m (1.5 to 3 ft) away from any cage or
building wall. Each lint cage has 8 sections, approximately 1.2 x
0.4 m (3.9 x 1.2 ft), in a single band [Table 1].
CYCLONES
A cyclone is an inertial separator in which particles are sep-
arated from the gas streamlines by means of centrifugal force created
in a vortex flow, driving the particulates suspended in the gas to
the collector wall. From there, the particulates settle to the bottom
of the cyclone and are pneumatically or mechanically carried away.
The cyclones at the Marana Gin are supported so that with the
cyclone outlets are 4.6 to 7.3 m (15 to 24 ft) above the ground. The
cyclone outlets range from 43 to 48 cm (17 to 19 in) in diameter and
are 10 cm (4 in) tall. Table 2 summarizes the cyclone design gas
flow rates and geometry data provided by the Company.
-------
17
Table 1
LINT CAGE CONSTRUCTION DETAILS
PRODUCERS COTTON OIL COMPANY
Marana, Arizona
Cage Cage No. of
Lint Cage Process Controlled Height Diameter Sections
Station No. m ft m ft
Short-Staple Cotton Gin [Figure 4]
SN 2201
Lint Cleaner
2.5
7.8
1.6
5.1
28
SN 2202
Condenser
3.6
11.7
1.6
5.1
28
SN 2203
Lint Cleaner
2.4
7.8
1.6
5.1
28
Long-Staple
Cotton Gin [Figure 5]
SN 2301
1st Stage Lint
1.5
5
0.9
3
8
Cleaner
SN 2302
1st Stage Lint
1.5
5
0.9
3
8
Cleaner
SN 2303
2nd Stage Lint
1.5
5
0.9
3
8
Cleaner
SN 2304
2nd Stage Lint
1.5
5
0.9
3
8
Cleaner
SN 2305
Condenser
1.5
5
0.9
3
8
-------
18
Table 2
CYCLONE DESIGN DATA
PRODUCERS COTTON OIL COMPANY
Marana, Arizona
Cyclone Group
Station No.
Process Controlled
Cyclone Size
(Diameter)
cm in
Design Gas Flow
Rate
m3/min ft3/min
Short-Staple Cotton Gin [Figure 4]
SN 2204
No. 1 inclined cleaner
97
38
226
8,000
reel drier
SN 2205
Bur extractor
97
38
170
6,000
SN 2207
No. 2 & 3 inclined
97
38
282
10,000
cleaners, gin stand
and feeder trash
conveyor
SN 2208
Gin stand feeder
97
38
170
6,000
overflow, mote in-
clined cleaner trash
SN 2206
Mote cyclone bypass
97
38
226
8,000
Mote
No. 1 & 2 lint cleaners
270
104
226
8,000
and lint cages. Con-
denser lint cage
SN 2209
Separator (unloading)
97
38
170
10,000
SN 2211
Cyclone trash pickup
97
38
Unknown
Long-Staple Cotton Gin [Figure 5]
SN 2310
No. 1 inclined cleaner
97
38
226
8,000
bur extractor
SN 2308
Separator (unloading)
97
38
226
8,000
SN 2307
Gin Trash No. 2 in-
97
38
170
6,000
clined cleaner
SN 2306
Lint cleaner trash,
97
38
170
6,000
lint cage trash
SN 2309
Reel drier
97
38
85
3,000
Seed
Seed conveying
61
24
42
1,500
a Data from Company process flow diagrams.
b SN 2206 cyclones are only used when the mote cleaning system is
not operating and the mote cyclone is being bypassed.
-------
V. TEST PROCEDURES
CYCLONE SAMPLING PROCEDURES
The cyclone exits were 10 cm (4 in) tall and could not be sampled
according to Methods 11 and 21 without outlet modifications. To accom-
plish sampling the outlets were modified with an involute system
[Figure 7] to eliminate the cyclonic gas flow, combine the emissions
of the cyclone pairs and triplets, and direct the gas flow toward the
ground. Sampling was performed in the 51 cm (20 in) diameter involute
system ducting at two perpendicular test ports about 1.5 m (5 ft)
from the ground. Because an involute system was used for each cyclone
pair or triplet, the sampling locations were similar in each case.
All particulate sampling in conjunction with the involute system
was performed according to Method 5. The stack gas molecular weight
was calculated using the average analyses of three gas samples. Gas
samples were obtained November 17 by a grab sample technique and an-
alyzed with Fyrite* type combustion gas analyzers. Stations 2209 and
2211 were not sampled, but were assumed to have the same gas compo-
sition as the other Stations. Past source tests had shown that cotton
gin emissions, even natural-gas-fired drier combustion gases, have
had gas compositions of ail 21% oxygen and 79% nitrogen; therefore
these values were assumed for Stations 2209 and 2211.
The number and location of the sampling points were determined
according to Method 1. Table 3 gives the lengths of straight ducting
Brand name.
-------
Table 3
SAMPLING LOCATION DATA
PRODUCERS COTTON OIL COMPANY
Marana, Arizona
. _ Distances3 s „ Po1nts Total Sampling Time
Station Upstream (A) Downstream (B) Per Run
No" m ft dia. m ft dia. Required*3 Used (min)
Short-Staple gin
2204
3
10
6
0.8
2.5
1.5
16
16
64
2205
3
10
6
0.8
2.5
1.5
16
16
64
2207
3
10
6
0.8
2.5
1.5
16
16
64
2208
3
10
6
0.8
2.5
1.5
16
16
64
2209
3
10
6
0.8
2.5
1. 5
16
16
64
2211
5.5
18
10.8
1.1
3.5
2.1
8
12
60
Long-
Staple Gin
2306
7.0
23
13.8
1.1
3.5
2. 1
8
12
60
2307
5.5
18
10.8
1.1
3.5
2.1
8
12
60
2308
5.5
18
10.8
1.1
3.5
2.1
8
12
60
2309
5.5
18
10.8
1.1
3.5
2. 1
8
12
60
2310
5.5
18
10.8
1.1
3.5
2.1
8
12
60
a For A and B see Figure 7.
b Method 1, Figure 1-1
-------
21
upstream and downstream of the test ports [Figure 7], the number of
diameters upstream and downstream, the required and actual number of
sampling points used, and the total sampling time per run.
Moisture content of the gas stream being sampled was determined
from the volume of water collected in the impingers and the weight
gain of the silica gel (Method 41).
The probe and oven (filter) were not heated. Past source tests
and the comparibility study2 show that the gas moisture content is
low (<3%). Method 5 allows this deviation when condensation is not a
problem.
At least three sampling runs were performed at all sampling loca-
tions. Prior to each run, the sampling train was leak checked at 380
mm (15 in) Hg vacuum or more. At the completion of the run, a second
leak check was conducted at the highest vacuum recorded during the
test. These checks were considered acceptable if the leak rate did
not exceed 0.0057 m3/min (0.02 acfm), which is 4% of the sampling
rate (5 acfm) as specified in Method 5.
All pitobe assemblies (pitot tube and probe), dry gas and orifice
meters used in these tests had been calibrated before leaving Denver,
Colorado and were recalibrated upon return [Appendix B]. All calibra-
tions were within the Method 5 specifications.
The HVSS, manufactured by Acurex Aerotherm [Appendix C] and used
for the cyclone testing, was arranged as follows:
a. Stainless steel (316) nozzle
b. Stainless steel (316) lined probe
c. Flexible Teflon-lined probe
d. Stainless steel (316) cyclone
e. Glass fiber filter (15.2 cm diameter) in
stainless steel filter holder
-------
Involute
Cap
Top View
inlet
90° Bend
Involute
Cap
Side View
Involute
Cap
inlet
Cyclone
Ll
Test Ports
Outlet
Figure 7. Cyclone Involute System
-------
23
f. Impingers (0 to 2)*--modified Greenburg-Smith, empty
g. Impinger—modified Greenburg-Smith with approximately 300
grams of si 1ica gel.
An NEIC mobile laboratory, on plant property during the tests,
was used for all sample train preparation and sample recovery.
Sample recovery proceeded as follows:
a. The filter was removed, folded, and inserted in a rigid
card and sealed in an envelope. The appropriate filter
identification information was recorded on the outside of
the envelope.
b. The nozzle, steel probe, Teflon probe, cyclone and front
portion of the stainless steel filter holder were brushed
and washed with acetone. The washings from each run were
collected in a glass jar with a Teflon-lined cap, and the
liquid level marked.
c. The volume of water collected in the empty impingers was
volumetrically measured, as part of the moisture determina-
tion, and discarded.
d. The silica gel impinger was weighed to determine the mois-
ture gain, and the silica gel was discarded.
e. All initial and final volumes and weights were recorded on
a sample cleanup sheet along with other pertinent data.
All samples were returned to the NEIC laboratories for
particulate analyses which were performed according to the procedures
described in Method 5 [Appendix D].
The number of empty impingers varied between zero and two.
-------
24
LINT CAGE SAMPLING PROCEDURES
Samples were collected around the circumferences of the seven
unenclosed lint cages. At Stations 2201 and 2202, sections of the
lint cage closest to the building wall could not be sampled because
the clearance between the wall and cage was not adequate for the
sampling nozzle.
The number of sample points was determined by the number of
screened sections on the lint cage. One point was sampled at the
center of each screened section. Since the screened sections had
similar areas [Appendix E], this procedure could be considered an
equal area method.
Gas vented from Station 2203 lint cage was discharged through an
opening in the shed wall. This opening was divided into 16 equal
areas (4x4 matrix) and the center of each area was sampled.
With the exception of Station 2203, each lint cage was sampled
three times. The Station 2203 lint cage was sampled six times, three
times in conjunction with an ambient air high-volume sampler to deter-
mine if emission results of the two methods (Rader HV vs Hi-Vol) were
comparable.
The vent gas molecular weight was calculated assuming the gas
composition as air, 21% oxygen and 79% nitrogen. This assumption is
considered valid because ambient air is used to convey wastes to the
lint cages.
Moisture content of the gas being sampled was determined by the
wet-bulb dry-bulb procedures given in Method 4. The moisture deter-
minations were performed after the lint cage sampling was complete.
As the lint cotton moisture was between 6 to 8% and the carrier gas
-------
25
was ambient air, the moisture content in the exhaust gas at the time
of and subsequent to sampling was assumed to remain the same.
The lint cage emissions could not be sampled isokinetically because
*
the gas velocity was below the sensitivity of the pi tot tube, and
the gas flow was swirling and non-perpendicular to the screened
circumference. Therefore, the lint cages were sampled at a constant
rate of 0.85 m3 (30 ft3)/min with the sample nozzle facing the
screened circumference. The nozzle was located 7.6 cm (3 in) from the
screening [Figure 8].
Before and after each run, the Rader HV sampler was leak tested
by plugging the nozzle and measuring the flow across the orifice meter.
These checks were acceptable if the orifice meter reading was zero.
The orifice meter used in these tests was calibrated according
to the Rader HV sampler operating manual, before leaving Denver, in
the field and upon return [Appendix B]. Because the gas flow could
not be measured leaving the lint cage, the volumetric flow rate was
determined by measuring the incoming gas flows. Only three of the
eight lint cages had good locations for a velocity traverse, and these
three locations (Stations 2302, 2304 and 2305) were checked once during
the survey. All eight lint cage inlets had previously been traversed
by the USDA-ARS [Appendix E] when no cotton was being processed. The
NEIC and ARS flow data for the three locations (Stations 2302, 2304
and 2305) were compared and it was found that the ARS data were 4 to
19% greater than the NEIC data. It was concluded that this difference
in gas flows was caused by the processing of cotton. Therefore, before
the ARS flow data at the five locations not traversed by NEIC were
used, the ARS data were adjusted to be representative of when cotton
was being processed (Test Results Section - Lint Cage Results Discussion).
The S-type and standard pitot tubes have sensitivities of 0.06
cm (0.025 in) of velocity pressure or about 160 m/min (500 fpm).
-------
26
Wall of Building
Top
View
Nozzle
Rader HV Sampler
Screened Circumference
Figure 8. Lint Cage Sampling Arrangement
-------
27
The Rader HV sampler, manufactured by Rader Company, Inc.
[Appendix C], and used to sample the lint cages, was arranged as
follows:
a. Aluminum nozzle (4.7 and 8.9 cm-diameter)
b. Aluminum probe - adjustable length
c. Aluminum filter housing for 20 x 25 cm (8 x 10 in)
glass fiber filter
d. Orifice with dial thermometer
e. Solenoid valve connected to control section and
control computer
f. Flexible hose
g. Suction blower
The Rader sampler nozzle and probe were cleaned at the sampling
location according to the Rader operating manual. This necessitated
turning on the suction blower and brushing both nozzle and probe until
the piece was visually free of dust and lint. The freed dust was
collected on the filter. No acetone washing was performed.
The filter holder (with filter) was sealed (corked) and taken to
the mobile laboratory where the filter was removed and the filter
holder brushed to move any remaining particulate onto the filter. The
filter was folded, inserted in a rigid card and sealed in an envelope.
The samples were handled and analyzed as described for the cyclone
samples.
AMBIENT AIR SAMPLING PROCEDURES
*
Two General Metals High-Volume samplers were used to measure the
ambient air particulate concentration at six locations [A-F, Figure 9]
in and around the two gin process buildings. The Hi-Vols were located
upwind and downwind of the gins as described below:
Brand name.
-------
Wagon Storage Area
N
/I
storage
cyclones
storage
© |cvclonp<; I
Wa-
Sto
Ar.
^ Lint Cages
Long Staple Cotton Gin
cyclones
©
seed
storage
Prevai1lng
Wi nd
Directi on
Unloading
Q>
Lint Cages
©© ©
Short Staple Cotton Gin
cvclones
Figure 9.
Producers Cotton Oil Company
Marana Gin Plot Plan
HiVol Sampling Locations
(2>—©- Hi Vol s (Ambient)
©
Bal e
Cotton
Storage
Area
seed
storage
ro
00
-------
29
Location No. Description
12 to 15 m (40 to 50 ft) south of lint
cage SN* 2201 on a concrete slab which
supports the seed conveying system
Inside the shed that protects lint cage
SN 2203, approximately 2 m (6.5 ft)
from the lint cage
West end of the cyclone SN 2204 support
base, 3 m (10 ft) north of the shed opening
for lint cage SN 2203
1.6 m (5.3 ft) north of the shed
opening for lint cage SN 2203
1.8 m ( 6 ft) north of the northwest
corner of the long-staple gin building,
approximately 6.1 m (20 ft) west of lint
cage SN 2301
East end of the cyclone SN 2310 support
base about 4.6 m (15 ft) from the long-
staple gin building
The Hi-Vols were operated according to the procedures specified in
40 CFR 50,** Appendix B---"Reference Method for the Determination of
Suspended Particulates in the Atmosphere (High-Volume Method)"
[Appendix F]. In most cases, sampling air flow was continuously re-
corded.
A single orifice plate was used to calibrate the Hi-Vols (designed to
maintain a constant sampling rate) before each sampling run. (The orifice
plate was calibrated with an NBS traceable Roots*** meter.)
Station number.
Code of Federal Regulations, Title 40, Part 50.
Brand name.
-------
30
Most (12 of 15) Hi-Vol samplers were scheduled to run 24 hours
(midnight to midnight). However, the particulate buildup for four
filters (locations B-3 Runs and D-l run) was sufficient to reduce the
air sampling rate and sampling was terminated. Two of three sampling
runs, performed in conjunction with the Rader HV at Station 2203, had
sampling times of 64 minutes, while a third run was approximately 10
hours.
The samples were recovered by removing the filters from the holder,
folding the filters in half, placing them in a rigid card and sealing
them in an envelope. Flow data, sampling time, and appropriate identi-
fication information were recorded on the envelope.
The samples were analyzed by the analytical procedures previously
described.
VISIBLE EMISSION OBSERVATION PROCEDURES
Visible emission observations (VEO's) were performed according
to Method 91 specifications. At this plant, with the predominantly
southeast wind and the majority of the cyclones grouped together, the
emissions intermingled and care was necessary to ensure a representa-
tive observation. VEO's on the lint cage emissions were not possible
because of interference between the lint cage emissions and wind-blown
dust.
PROCESS OBSERVATION PROCEDURES
The gin process was monitored by recording the number of wagons
unloaded and the number of bales produced. Production rates were
checked by comparing the wagon and/or bale numbers to the clock time,
-------
31
and calculating the throughput. The production data obtained from
the gin office [Appendix G] were used to determine the allowable pro-
cess weight regulation emissions.
Because the gin emissions were dependent on the type and condition
of the cotton—for example, pick of cotton—random grab samples of
seed cotton were taken directly from the wagons and turned over to
the USDA-ARS, Southwestern Cotton Ginning Research Laboratory for
analysis of trash content. Four samples of short-staple (Upland)
cotton and five samples of long-staple (Pima) cotton samples were
analyzed.
The analysis consisted of separating the hulls, sticks, motes
and fines (leaves) from the seed cotton and determining the weight of
each, gravimetrically [Appendix H]. The condition of the cotton was
established by its trash weight fraction of raw cotton.
-------
VI. TEST RESULTS
Between November 2 and 19, 1977, the short- and long-staple
cotton gins were tested to determine their particulate emission
rates. All cyclones were source tested three times with the
Aerotherm HVSS. The seed, mote and Station 2206 cyclones were not
sampled because they were either inaccessible or not operating. The
lint cages were sampled three times with the Rader HV sampler.
Station 2203 lint cage was sampled an additional three times in con-
junction with ambient air Hi-Vol sampling to determine if the test
results could be correlated. All samples were collected and analyzed
according to NEIC chain-of-custody procedures, except as noted in
Appendix I.
During the testing period, the average production rate of the
short-staple gin was 8.6 bales/hr, 86% of the rated capacity (10 bales/hr).
The long-staple gin average production rate of 3.4 bales/hr was 85%
of the rated capacity (4 bales/hr). Continuous gin operations were
maintained during each sampling period. From the process weight table
in Pima County Regulation II, Rule 2 and the daily cotton feed rate
[Table 4], the allowable emission rates were calculated as 5.1 kg
(11.2 lb)/hr and 2.7 kg (6.02 lb)/hr, respectively, for the short-
and long-staple gins.
The total gin particulate emissions are the sum of the individual
sources, e.g., lint cages and cyclones. Tables 5 and 6 present the
average emissions from each source and the cumulative results. Test
Station 2308, Run No. 1 isokinetics were 89.3%.
-------
Table 4
PROCESS DATA SUMMARY
PRODUCERS COTTON OIL COMPANY
Marana, Arizona
Process Weight Allowed Emissions
Date lb/daya kg/daya ton/hr m. ton/hr lb/hr kg/hr
Short-staple cotton gin
11/5
225,130
102, H8
5.00
4.54
11/6
243,860
110,613
5.52
5.0"l
11/7h
289,690
131,401
6.44
5.84
1 l/8b
61,920
13,651
-
-
11/9
316,380
143,507
7.03
6.38
11/10
297,525
134,955
6.61
6.00
11/11
307,4*10
139,439
6.83
6.20
11/12
283,720
128,693
6.30
5.72
11/13
296,770
134,613
6.53
5.92
Average
-
-
6.28
5.70
Long-Staple Cotton
gin
11/12
31,210
14,157
1.39
1.26
11/13
48,420
21,963
2. "15
1.95
11/14
117,930
53,492
2.62
2.38
11/15
109,710
49,763
2.44
2.21
11/16 .
100,180
45,441
2.23
2.02
n/i7°
-
-
2.70
2.45
n/1^
-
-
2.26
2.05
11/19
-
-
2.76
2. 5"l
Average
-
-
2.32
2.10
a Process day is 22.5 hrs, gin shuts down for two 45-minute
lunches per day.
b No sampling occurred on this day. Data not included in averages,
c Process day - 8 to 10 hours.
d Gin process Summary not available. Process weight calculated
from time period required to unload wagons.
-------
Table 5
EMISSION DATA SUMMARY
SHORT-STAPLE GIN
PRODUCERS COTTON OIL COMPANY
Marana, Arizona
Average Average ^
Concentration8 Mass Emissions
Station No. gr/cf mg/m3 lb/hr kg/hr
Cyclones
2204 0.044 100 3.33 1.51
2205 0.011 25.7 0.38 0.17
2207 0.024 54.6 1.35 0.61
2208 0.015 33.8 0.50 0.23
2209 0.048 110 2.14 0.97
2211 0.19 437 8.25 3.74
Subtotal - - 16.0 7.23
Lint Cages
2201 0.039 89.1 4.32 1.96
2202 0.010 23.0 0.88 0.40
2203 0.051 118 8.31 3.77
Subtotal - - 13.5 6.13
Total - - 29.5 13.4
Emission Limitation 11.2 5.1
a At standard conditions,
b Adjusted for blockage effects.
-------
35
Cyclones
Lint Cages
Table 6
EMISSION DATA SUMMARY
LONG-STAPLE GIN
PRODUCERS COTTON OIL COMPANY
Marana, Arizona
Station No.
Average
Concentration1
gr/scf
mg/nr
Average ^
Mass Emissions
lb/hr
kg/hr
2306
2307
2308
2309
2310
Subtotal
0.0083
0.0094
0.064
0.043
0.017
18.8
21.0
146
98.8
38.9
0.51
0.40
1.17
1.16
0.82
4.06
0.23
0.18
0.53
0.53
0.37
1.84
2301
2302
2303
2304
2305
Subtotal
0.033
0.015
0.016
0.027
0.017
76.4
35.4
35.7
61.7
38.8
1.15
0.55
0.34
0.66
0.74
3.44
0.52
0.25
0.15
0.30
0.34
1.56
Total
Emission Limitation
7.50
6.02
3.40
2.7
a At standard conditions,
b Adjusted for blockage effects.
-------
36
data from each source run are summarized in Appendix J. Field data
sheets and example calculations are presented in Appendix K.
The short-staple gin emissions were 13.4 kg (29.5 lb)/hr, 164%
greater than the allowed emissions. The cumulative emission results
of the short-staple gin are conservative because they do not include
the mote cyclone emissions* and include lint cage emissions which
were conservatively calculated (see Lint Cage Results Discussion). A
substantial (28%) part of the total emissions was from the trash
cyclones (Station 2211).
Emissions from the long-staple gin were 3.40 kg (7.50 lb)/hr,
which is 25% greater than the allowed 2.7 kg (6 lb)/hr. The seed
cyclone emissions are not included in the long-staple cumulative re-
sults. By not including these emissions, and including conservative
lint cage results (see Lint Cage Results Discussion), the overall
emission rate reported is lower than actual.
The short-staple gin visible emissions were observed and recorded.
Of these, no individual VEO readings exceeded 20% opacity and no average
reading of 24 observations exceeded 10% opacity [Table 7]. The VEO's
did not exceed the Pima County Regulation II Rule 1 which limits emissions
to less than 40% opacity.
The emissions from the long-staple gin cyclones were generally
not visible, therefore no VEO's were performed. No VEO's were made
of the mote and seed cyclone emissions, despite the fact some visible
emissions were observed from the mote cyclone. Observation of the
mote cyclone visible emissions was largely impeded because of the
dispersion caused by the rain cap.
The mote cyclone outlet was not accessible
-------
37
Table 7
<
VISIBLE EMISSION OBSERVATION SUMMARY
PRODUCERS COTTON OIL COMPANY
Marana, Arizona
Range of Readings Average of
Station No. Date (Low - High %) 24 Observations (%)
2209
11/9
0b
0
2207
11/10
0-5
0.5
2205
11/10
0b
0
2204
11/10
ob
0
2208
11/10
0-5
0.5
2211
11/10
10b
10
2211
11/10
5-20
10
2209
11/10
0-5
2
a Involute system not in place during observations,
b All values were the same.
-------
38
No visible emission observations could be performed at the lint
cages because of the interferences from the wind and nearby lint cages.
However, a white haze in the vicinity (1 to 3 m) of these devices
necessitated the wearing of dust masks.
Ambient air particulate data are presented in Table 8. Due to
weather conditions, vehicle movement in the plant and process operating
changes, the results vary widely. All the data exceeded the National
Primary Ambient Air Quality Standard of 0.26 mg/m3 maximum 24-hr concen-
tration, and all but one value (Location A, November 11 sample) exceeded
the "never to be reached" value of 1 mg/m3 specified in the Federal
Air Episode Criteria.3 Particulate concentrations of 1 mg/m3 are
designated as substantial endangerment levels. Therefore, the gin
emissions have an adverse impact on the air quality of the plant
employees' work environment.
The data from Locations B, C and D (downwind of short-staple
gin) were generally greater than at Location A (upwind). This was
true for the long-staple gin (Locations E and F) also. No correlation
could be made between the ambient air data at Location D and the Rader
sampling data at Station 2203, because of the limited number of data
points.
Table 9 summarizes the seed cotton analytical results. As
mentioned in the Process Description, it takes about 725 kg (1,600
lb) of seed cotton to produce a bale of lint cotton. According to
the Cotton Ginners Handbook4 [Tables 4-15], the average weight of
trash removed during ginning of machine-picked cotton is 92 kg (203
lb)/bale. Therefore, trash normally accounts for about 13% (92 kg/725 kg)
of the seed cotton weight to the gin. The trash weight percentage of
the short-staple and long-staple seed cotton samples obtained at the
Marana Gin were 9.8 and 9.6%, respectively. Because these sample
results were close to the Handbook value (13%), the cotton processed
was considered typical of machine-picked seed cotton.
-------
39
Table 8
AMBIENT AIR DATA SUMMARY
PRODUCERS COTTON OIL COMPANY
Marana, Arizona
Particulate
(Date) Sample Volume Particulate Catch Concentration
Location November m3 gm mg/m3
A
9
77.3
1.12
14.5
A
11
1,770
0.88
0.50
B
4
895
80.0
89.4
B
5
899
76.4
85.0
B
6
458
48.0
105
C
7a
1,146
11.16
9.74
C
10
1,043
116.4
112
D
9
761
1.14
1.50
D
9
77.3
8.14
105
D
9
78.1
5.72
73.2
E
12
1,757
6.84
3.89
E
14
1,771
10.38
5.86
E
16
1,772
13.98
7.89
F
13
1,561
2.46
1.58
F
15
1,553
4.24
2.73
a 11/7-raining, 11/8-gin not operating.
-------
40
Table 9
COTTON ANALYTICAL RESULTS
PRODUCERS COTTON OIL COMPANY
Marana, Arizona
Trash Identification
Percent of
Raw Short-
Staple Cotton
Percent of
Raw Long-
Staple Cotton
Hul 1 s
Sticks & Stems
Motes
Leaves and Fines
Total Trash
1.5
2.0
0.7
0.8
0.7
0.2
6.9
6.6
9.8
9.6
a Average of four samples,
b Average of five samples.
-------
41
The filters and acetone wash
sampling runs on the short-staple
were analyzed for four phosphorus
thion, malathion and dimethoate.
[Appendix D] is quoted below:
samples collected during the four
gin trash cyclones (Station 2211)
pesticides—methyl and ethyl para-
The analytical results summary
"The four filters contained no detectable amounts of the pes-
ticides of interest. All four of the acetone washes contained
methyl parathion, ranging from 0.1 to 1.8 pg. Three of the
acetone washes contained ethyl parathion in amounts ranging from
0.1 to 0.5 pg. Malathion was not found in any of the acetone
washes. The analytical methods used to analyze ethyl and methyl
parathion and malathion prohibited analyzing the acetone wash
samples for dimethoate."
Based on the acetone wash results, the average emission concentra-
tions of methyl and ethyl parathion were 0.18 and 0.052 pg/m3, respect-
ively. The methyl and ethyl parathion emission rates were 1.64 and
0.47 mg/hr, respectively.
CYCLONE RESULTS DISCUSSION
The involute system was used for all cyclone testing. Because
the diameter of the involute system ducting (0.5 m or 1.7 ft) was
less than 0.9 m (3 ft), the Scientific Glass* sampling probe (no ex-
ternal sheath) caused some flow blockage. From NEIC experimental
data on the relationship between blockage and the pitot tube calibration
coefficient (Cp), the data show that as blockage increases, Cp decreases
[Appendix B, Figure B~2]. The average probe blockage when sampling
at the centerpoint in the 51 cm (20 in) duct was 4.7%, resulting in a
Cp decrease of 3.3%.
Manufacturer's name.
-------
42
This decrease changed two parameters: 1) the cyclone volumetric
flow rates, and 2) the isokinetic sampling rates. A 3.3% Cp decrease
reduced the cyclone volumetric flow rates and thus the emission rates
by 3.3%. A 3.3% decrease in the cyclone emission rates changed the
total gin emission rate about 1.8%.* The data presented in Tables 5
and 6 reflect the 3.3% decrease in the cyclone data contained in
Appendix J [Table J-20].
The isokinetic sampling rate increased by 3.3% as Cp decreased
by 3.3%. Of the 33 runs,** 28 were within the range of 90 to 110%.
The five remaining runs had isokinetic sampling rates between 110 and
112.3%. Since the isokinetic sampling rates were greater than the
allowable extreme (110%), the emission rates for these 5 runs are
conservati ve.
An attempt was made to calculate the effect of the involute
system on measured emissions. It can be concluded from engineering
principles that the involute system effectively reduced the cyclone
exit flow because of friction losses caused by the expansion joint
90° bend, and additional straight section of pipe. However, the
effect of the involute system on cyclone collection efficiencies
could not be computed. Thus, the involute system effect on cyclone
emissions could not be estimated.
LINT CAGE RESULTS DISCUSSION
The eight lint cage volumetric gas flow rates were measured in
the inlet ducting because the outlet velocities were below the
Both the short- and long-staple gin cyclone results account for
55% of the total emissions. 55% of the 3.3% change in the
cyclone emissions reduces the total emissions by 1.8%.
33 runs - 11 cyclone groups each sampled 3 times.
-------
43
sensitivity of the pitot tube. NEIC measured three locations
(Stations 2302, 2304 and 2305) during the survey; ARS had previously
measured the velocities at eight locations (Stations 2201, 2202,
2203, 2301, 2302, 2303, 2304 and 2305) under two different operating
conditions, i.e., cotton was or was not being processed. Before
using the ARS flow data at the five locations not traversed by NEIC,
the ARS data was adjusted to reflect the operating condition during
which the NEIC data was obtained, i.e., processing of cotton.
In order to do this, the NEIC and ARS flow data available for
both operating conditions for Stations 2302, 2304 and 2305 was graphed
[Appendix E, Figure E-l]. Assuming the flow reduction observed at
the three locations occurred at the other five locations (Stations
2201, 2202, 2203 2301 and 2303), the graph was used with the ARS data
to obtain the adjusted ARS flow data for these five locations.
The NEIC measured and the ARS adjusted flow data are less than
the ARS actual flow data. Using these lower values results in the
calculation of conservative emission data.
Because the gas velocities from the lint cages were below the
sensitivity of the pitot tube, the sampling was performed at a con-
stant rate of 0.85 m3 (30 ft3)/min. This resulted in the short-staple
gin lint cage average isokinetic sampling rates of 400, 1,530* and
100% at Stations 2201, 2202 and 2203, respectively [Appendix E, Table
E-l]. The long-staple gin lint cage average isokinetic sampling rates
ranged from 312 to 466%. According to Method 5 sampling at a rate
exceeding the acceptable isokinetic sampling range results in par-
ticulate concentrations, and thus emissions, which are conservative,
that is, less than the actual values.
4.7 in (1.86 in) nozzle used in place of 8.9 cm (3.505 in) nozzle.
-------
44
REFERENCES
1. Code of Federal Regulations, Title 40, Part 60, Standards of
Performance for New Stationary Sources, Appendix A - Reference
Methods, August 18, 1977.
2. Comparability Study of EPA Method 5 and a High-Volume Sampler
for Particulate Testing of Cotton Gin Emissions, EPA-330/1-77-006,
June 1977 (Draft).
3. Code of Federal Regulations, Title 40, Part 51.16, Prevention of
Air Pollution Emergency Episodes.
4. Cotton Ginners Handbook, Agriculture Handbook No. 503, Agricultural
Research Service, U. S. Department of Agriculture, July 1977,
Figure 4-30.
5. A Method of Interpreting Stack Sampling Data, Smith, W.S.,
Shigehara, R.T., and Todd, W.F., Paper Presented at the 63rd
Annual Meeting of the Air Pollution Control Association, St.
Louis, Missouri, June 14-19, 1970.
-------
APPENDICES
A Presurvey Inspection Report
B Calibration Procedures and Data
C Sampling Train Construction Details
D NEIC Analytical Procedures and Data
E Lint Cage Dimensions, Flow Data and Calculations
F High-Volume Sampler Reference Procedures
and Calibration Data
G Cotton Gin Production Data
H Agricultural Research Service Trash Analysis
Analytical Procedure and Data
I Chain-of-Custody Procedures and Records
J Emission Data Summary Sheets
K Field Data Sheets and Example Calculations
-------
Appendix A
Presurvey Inspection Report
-------
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF ENFORCEMENT
NATIONAL ENFORCEAAENT INVESTIGATIONS CENTER
BUILDING 53, BOX 25227, DENVER FEDERAL CENTER
DENVER, COLORADO 80225
to Chief, Field Operations Branch December 17, 1976
from Paul R. dePercin
subject Presurvey Inspection of the Marana Gin, Producers Cotton Oil Company,
Marana, Arizona
On November 17, 1976, Mr. Daniel Yee, EPA, Region IX; Mr. H. Poole,
Pima County Health Department; and the writer inspected the Marana Gin of
Producers Cotton Oil Company, Marana, Arizona, to determine the feasibility
of performing source testing and to evaluate process operations, control
devices and the emissions.
The Marana Gin consists of two ginning operations: 1) the processing
of short staple cotton, and 2) the processing of Pima or long staple cotton
[Figure 1]. The short staple cotton gin processes 10-11 bales* of cotton/
hour; the Pima cotton gin processes 3-4 bales/hour. The seed cotton (i.e.,
unprocessed cotton) is harvested and brought to the gin by the farmers.
The gin separates the trash, cottonseed and cotton lint. The cottonseed
and bale cotton is returned to the farmer, who sells them. The gin is
not involved with the growing, harvesting or sale of the cotton.
The plant representatives contacted were Messrs. R. Prikosovi t.s, Gin
Maintenance Superintendent, and Jack Collins, Gin Manager.
Process Description
The seed cotton is delivered to the gin in a wagon which holds about
2.2 m. tons (2.5 tons). It normally takes about 725 kg (1,600 lbs) of
seed cotton to produce a bale of cotton (227 kg or 500 lbs). The weight
difference (500 kg or 1,100 lbs) is the weight of trash and seed removed
during processing.
The condition of the seed cotton, a function of the picking procedures,
affects process operations and emissions. Cotton is generally picked twice
from the defoliated stalks which accounts for 90-95% of the cotton ginned.
Picked cotton contains some trash (stalks, leaves and dirt). Subsequent
to the second pick a vacuum sweeper is used to pick up cotton including
* A bale weighs 227 kg (500 lbs).
-------
3 L/'aj T~
sc/:]£
3
/0 C/clo/v e S
f $L=, r. T sr,>/>lc ro7Tc/u J '/u
/J~J/o s~J fo>oT",<\r Z- 5'
Pi &?*> $ Co 7?o $2
pffoCe f
/rncTe cyc.L<,»e
Sh ofcf QTtipl* G
$e c p
cyc.Lo/l-»t~>e3
pi.-^A colto *-> J**aj
£ sr*ifj?° AJ C0KiTZioLs
&eev>
S ef/i AAtb /2 /=s>~i'i fterO
Co/^TfioL 5
see®
STarttiy e
&/0J
P &o®VC&$ CoWbfi* @fb> Co•
Figure 1: Gin Plot Plan
-------
-2-
trash that has fallen to the ground (i.e., ground cotton). The pollution
control devices are not designed to remove this increased trash load,
therefore, visible emissions occur in the exhaust gases. Visible emissions
are seldom observed when the picked cotton is processed.
The seed cotton moisture content is a function of the mode of harvesting
(i.e., picked or ground cotton) and when the cotton was harvested (i.e.,
morning or afternoon, before or after a rain), and affects the process
operations and emissions. A 6-8% moisture content is ideal for processing.
A cotton moisture content less than 6% results in cotton being removed with
the trash. Only a small amount of trash will be removed when the moisture
content is greater than 8%, resulting in a poor grade of bale cotton. The
driers adjust the cotton moisture to the desired range during processing.
The cleaner (i.e., inclined cleaners) efficiencies vary with the moisture
content, thus the trash loading to the cyclones is a function of the cotton
moisture content.
The specific processing steps of each gin are described below:
1. Short Staple Cotton Gin [Figure 2]
The seed cotton is off-loaded from the wagon with a vacuum
system and collected in the separator where the cotton is re-
moved from the carrier air stream. The air stream is vented to
three cyclones (11/12 and 13)* and the cotton enters a natural
gas fired reel drier.
The reel drier is the initial process used to reduce the
moisture content to between 6 and 8%. The horizontal rotary
drier performs a small amount of cleaning.
After the reel drier the cotton is air-conveyed to and
processed by a 7 cylinder inclined cleaner. The cleaner
consists of 7 spiked drum cylinders which carry and "scrub"
the seed cotton to remove the fine particulates (e.g., leaf,
dirt and sand). The reel drier combustion gases and the
carrier air is vented to two cyclones (1 and 2). The trash
collected in both the reel drier and inclined cleaner is
collected in two additional cyclones (9 and 10).
The cotton passes to a bur extractor, which removes heavy
trash and burs with a revolving saw cylinder. The teeth of
the saw hold the seed cotton and subjects them to a carding
and cleaning action as the cotton is spread across the surface
* Nuinbe in parentheses refer to Figure 2 and the table of control equipment
(Control Equipment Description).
-------
_
8"i
1 and 2 1
Cyclones1** - _
Seed
Cotton
Reel
Dri er
V
¦
/
/
'
/
V
No.
Incl
f.lp
1
ined
finer
X.
Separator
i
111,12 and 13
Ofc) Cyclones
oo 3 and 4
Cyclones
Cotton
Baler
«y
Burr
Extractor
\
— i ..J
'
Tower
Drier
\
--- K
Lint
Condenser
-.Screen
i
O
No. 1
Lint Cleaner
No. 2
Lint Cleaner
v ! Lint
Screen
No. 2
Inclined
PIaa nor
No. $
Inclined
Clean?]:
Cyclone
ovo
9 and 10
Cvclones
Feedei
Figure 2
Short Stap
Cotton Proces
Flov/
Feeder
Legend:
Cotton
Trash and Air
eeder
7 and 8 r\/rinn^c
-------
-3-
of the cylinder. The trash and burs removed from the cotton are
pneumatically conveyed to two cyclones (3 and 4) which remove the
trash and burs from the air stream.
The cotton is then fed into the tower drier. This natural gas
fired drier is made up of 24 shelves. The cotton is subjected to
the war in gases while falling from shelf to shelf. The cotton
moisture centent is adjusted to the desired 6 to 8%. The dry
cotton is air-conveyed to the second set of two inclined cleaners
(operated in parallel).
These inclined cleaners are identical to the one previously
described. The carrier air is ducted into two cyclones (5 and 6).
The trash is collected by the same cyclones (i.e., 9 and 10) which
receive the reel drier and No. 1 inclined cleaner trash.
The cotton is then distributed to five extractor-feeders
where the burs, stems, whole leaf and other trash are removed.
The trash drops to a conveyor belt and the cotton is fed at a
constant rate to the 80-saw gin stands.
The gin stand uses these saws along with air blasts to
separate the seeds and trash from the cotton. The resulting
lint cotton is air-conveyed to two lint cleaners. The trash
is dropped to the feeder trash conveyor belt. The conveyor
belt carries the combined feeder and gin stand trash to a
point where the trash is picked up by an air stream and carried
to the two cyclones (5 and 6) that also control the No. 2 in-
clined cleaner carrier air emissions.
The lint cleaners (operated in parallel) remove the fine
leaf particles, motes, dust and sticks. The carrier air and
trash of the two lint cleaners are controlled by different
systems. The Mo. 1 lint cleaner emissions are vented to a
lint screen, where the collected lint is drawn to the cyclones
(3 and 4) that also collect the trash from the bur extractor.
The f.'o. 2 lint cleaner emissions are ducted to another lint
screen. The lint drawn off the bottom of this lint screen is
conveyed to the mote cyclone. During upset conditions the
mote cyclone is bypassed and the lint is removed from the air
stream by the two cyclones (9 and 10) that collect the trash
from the inclined cleaners and reel drier.
The lint cotton is air-conveyed to the condenser where the
ginned lint is formed into a smooth endless mat. The cotton
mat is fed into the baler. The bales are wrapped in burlap,
-------
-4-
tied with metal bands and stored outside for shipment. The
air stream from the condenser is vented to a lint screen. Trie
collected lint is drav/n to the two cyclones (3 and 4) that
control the bur extractor and No. 1 lint cleaner emissions.
2. Pima Cotton Gin [Figure 3]
The seed cotton is off-loaded from the wagon with a vacuum
system and collected in the separator where the cotton is re-
moved from the carrier air stream. The air stream is ventei
to two cyclones [3 and 41* and the cotton enters a natural gis
fired reel drier.
The horizontal rotary drier reduces the cotton moisture
content to 6 to 8%.
After the reel drier the cotton is air-conveyed to and
processed by a five cylinder inclined cleaner. The cleaner
consists of five spiked drum cylinders which carry and "scrub"
the seed cotton to remove the five particulates (e.g., leaf*
dirt and sand). The trash removed is collected in two
cyclones [1 and 2].
The cotton passes to a bur extractor which removes heavy
trash and burs with a revolving saw cylinder. The teeth of
the saw hold the seed cotton and subjects them to a carding
a.id cleaning action as the cotton is spread across the surface
of the cylinder. The trash and burs removed from the cottar
are pneumatically conveyed to the two cyclones [9 and 10] tfhat
also control the emissions from the reel drier.
The cotton is further processed in a 7 cylinder inclined'
cleaner. The 7 cylinder inclined cleaner operates as the 5
cylinder inclined cleaner described previously. The trash iis
removed by two cyclones [7 and 8].
The cotton is then distributed to the three roller gin
stands which use a revolving cylinder to separate the seeds
and trash from the lint without breaking or damaging the loig
lint fibers. The seeds are pneumatically conveyed to a cyclone
and storage pile. The trash is removed in two cyclones [5 and 6].
The resulting lint cotton is air-conveyed to two lint
cleaners (operated in parallel) which remove the fine leaf
* Numbers in brackets refer to Figure 3 and the table of control
equipment (Control Equipment Description).
-------
Seed
Cotton
Cotton
Bales
Saler
Cyclones
i
PLint Screen
— ->
9 and 10
Cyclones
Cyclone:
Lint Cleaners
-O—
i nt Screen
Cotton
Roller
Gin
Trash and air _>
Roller
Gin
Roller
Gin
Trash
Cotton Seeds
s/
oo
7 and 8
Cyclones
Burr
Extractor
Battery
Condenser
No. 1
Inclined Cleaner
No.?
Inclined Cleaner
Separator
Reel Drier
O O 5 and 6 Cyclon=0V Scsd Cyc1one
Figure 3. Pima Cotton Process Flow
-------
-5-
particles, motes*, dust and sticks. The lint cleaners vent
the carrier air to individual lint screens. The collected
lint** is drav/n ofF the two lint screen bottoms and collected
in the same cyclones [7 and 8] that collect the trash from the
7 cylinder inclined cleaner.
The lint cotton is air-conveyed to the battery condenser
where it is formed into a smooth endless mat. The cotton mat
is fed into the baler. The bales are wrapped in burlap, tied
with metal bands and stored outside for shipment. The exhaust
gases are vented to a lint screen. The collected lint* is
removed in the same cyclones [7 and 8] that collect the trash
and lint from the No. 2 inclined cleaner and lint cleaners.
The trash is collected by the cyclones and hauled away.
The seed is hauled to a cottonseed oil mill.
Control Equipment Description
Cyclones are used to control the process emissions and collect the
trash removed from the cotton. Two or three cyclones are used to collect
the particulate from a single source. Below is a summary of these
process control devices and their associated gas flow rates.
1. Short Staple Cotton Gin
Cyclone Size
(Diameter)
Cyclone*** Process Controlled cm (in)
1 & 2 No. 1 inclined cleaner 97 (38)
reel drier
3 & 4 Bur extractor, No. 1 97 (38)
lint cleaner
Battery condenser
5 & 6 No. 2 & 3 inclined 97 (38)
cleaners, gin stand
and feeder trash
conveyor
7 & 8 Gin stand feeder 97 (38)
overflow
* Motes are a group of fibers attached to a portion of seed hull.
** Only a small amount of poor grade lint is collected and recovery
is not considered economical.
*** See Figure 2.
Gas Flow
Rate
nr/min (ACFM)
226 (8,000)
170 (6,000)
282 (10,000)
170 (6,000)
-------
-6-
Cyclone Size Gas Flow
(Die eter) , Rate
Cyclone Process Controlled cm (in) m /min (ACFM)
9 & 10 No. 1, 2 & 3 inclined 97 (38) 226 (8,000)
cleaners-trash, No. 2
lint cleaner when mote
cyclone bypassed, reel
drier
Mote No. 2 lint cleaner 270 (104) 226 (8,000)
11, 12 Separator 97 (38) 170 (10,000)
6 13
Pima Cotton Gin
Cyclone Size Gas Flow
(Diameter) g Rate
Cyclone* Process Controlled cm (in) m /min (ACFM)
1 & 2 No. 1 inclined cleaner 97 (38) 226 (8,000)
3 & 4 Separator 97 (38) 226 (8,000)
5 & 6 Trash feeder conveyor 97 (38) 170 (6,000)
of roller gin overflow
7 & 8 No. 1 & 2 lint cleaners 97 (38) 170 (6,000)
50-in. battery condenser
No. 2 inclined cleaner
9 & 10 Bur extractor 97 (38) 85 (3,000)
Reel drier
Seed Seed conveying 61 (24) 42 (1,500)
*See Figure 3.
With the exception of the mote and seed cyclones, all the cyclone outlets
are 43 cm (17 inches) in diameter and 10 cm (4 inches) in height. The mote
cyclone outlet has a diameter of .9 m (3 ft) and height of 15 cm (6 inches).
The seed cyclone has an outlet diameter of 30 cm (12 inches) and outlet
height of 5 cm (2 inches). Only the mote cyclone has a rain cap.
All the lint screens used in the plant (six total) are identical, .9 m
(3 ft) in diameter and 1.2 m (4 ft) in height. The lint screen is a
cylindrical cage mode of 14 to 18 mesh wire screening. The contaminated
gases enter the lint screen from the top and vent through the screening.
The lint collected on the screening sluffs off when the weight of the lint
is greater than the air exhaust pressure.
-------
-7-
The exhaust gases from the cyclones and lint screens are composed
primarily of air with temperatures ranging from ambient to 200°F and
moisture from 0-3%. The gas velocities at the cyclone outlets range
from 12-18 m/sec (40-60 ft/sec) and cyclonic flow is present at the
outlets. The exhaust gases can contain dust, dirt and lint.
Source Sampling Feasibility
There are^ 31 emission points in the M.arana Gin; 25 cyclones and 6
lint screens. * Twenty-three cyclones emit the majority of these emissions.
The mote and seed cyclones are not considered major emission sources.
Neither the cyclones nor the lint screens have stacks or ducts that
can be sampled according to EPA, Method I1 criteria (minimum requirements—
2.5 diameters of straight ducting). The gas flow patterns from the
cyclones and lint screens do not meet the EPA, Method 21 requirements,
which state that sampling must be performed at a location where the gas
flow is straight and uniform, not swirling or cyclonic.
There are no usable sampling platforms and little structural
support for platforms. The cyclones can be source sampled if modifi-
cations are made. The lint screens, however, cannot be sampled.
The gin has 25 cyclones; 14 control emissions from the short staple
cotton gin and the remainder control the Pima cotton gin emissions.
Cyclones in sets of two or three control the process emissions. Assuming
that each cyclone in the set (i.e., two or three) has identical emission
rates, only one cyclone for each processing operation would have to be
sampled. This assumption must be checked by determining the gas flow
rate of each cyclone. A minimum of 11 cyclones would be sampled to
obtain the total short staple gin and Pima cotton gin emissions ("i.e., 6
and 5 cyclones respectively). The seed and mote cyclones would not be
sampled because both are minor pollution sources. The seed cyclone
separates the cottonseed from the carrier air stream with 100% efficiency.
The mote cyclone also recovers 100% of the lint and motes from the waste
air stream.
In order to sample the cyclone emissions, each cyclone outlet will
have to be modified by attaching a stack extension which meets the EPA
Method 1 and 2 requirements. An involute stack extension has been used
by the Agriculture Research Service, USDA, to obtain such a sampling
location. The involute-shaped cap is connected to the cyclone outlet
thereby changing the vertical gas flow to horizontal with little inter-
ference to the volumetric gas flow rate and reducing the swirling of the
exhaust gases. Where involute caps are connected to two or more cyclones
the emissions can be combined in a single duct. The combined emissions
will require fewer source tests and the test results will be more indicative
of the process emissions.
-------
130 DEGREE BENDS
STRAIGHTENING
VANES
Sample Ports
BUILDING
Trash
Figure 4 SCHEMATIC DIAGRAM OF THE CYCLONE .UNITS AND SAMPLING DUCTS
-------
-8-
The cyclone exhaust gases contain dirt, dust and lint. The number
of large lint fibers (1.3 cm or .5 inches) are minimal. The nozzle
sizes normally used by the EPA Method 51 range from 0.5 to 1.3 cm (0.19
to .50 inches;) and will collect the dirt and dust particles, but large
lint fibers will not be collected thereby biasing the results low. The
high-volume source sampler3 is an alternative to the EPA Method 5 sampling
train. The high-volume source sampler is designed to collect large
particulate in a gas stream using large nozzles (5 cm or 2 inches). The
procedure, however, is still being tested and has not been approved.
The process operating conditions depend on the condition of the
seed cotton. Ground cotton, v/ith its large percentage of trash, has
greater particulate emissions during processing than first or second
pick. But the ground cotton accounts for only 10% of the total cotton
processed. Further the moisture content of the cotton in each wagon load
varies. From 3 to 5 and 1 to 2 wagon loads of cotton will be processed,
respectively, during a sample run by the short staple and Pima cotton
gins. To eliminate the variations in cotton conditions, a single pick
of cotton must be processed during sampling.
The lint screens cannot be sampled as the waste air stream is
discharged in all directions through the screen. Any enclosure used to
confine these emissions would affect the lint screen performance.
Summary and Conclusions
Representative results of the emissions can be determined with the
use of stack extensions and by processing a single pick of cotton. Six
tests or 18 sampling runs will determine the short staple cotton gin
emissions, and 5 tests or 15 sampling runs will determine the Pima
cotton gin emissions. There would possibly be some bias in the Method 5
results, but the bias should be insignificant (i.e., less than 10%).
The lint screens cannot be source sampled. The geometry of the
control device precludes testing according to EPA Methods 1 and 2. Lint
screens are major emission sources which contribute significantly to the
gin emissions.
The seed and mote cyclones cannot be sampled because of their in-
accessibility and should not be sampled because they are minor sources.
-------
BIBLIOGRAPHY
1 40 CFR 60 - Standard of Performance of Mew Stationary Sources, Appendix A,
Reference Method 1 through 5.
2 Air Pollution .Emission Test, Delta and Pine Land Company Cotton Gin,
Scott, Mississippi. Report No. 72-MM-16, EPA, Office of Air Quality
Planning and Standards, Emissions Standards and Engineering Division,
November 1974.
3 Emissions from Cotton Gin at Valley Gin Company, Peoria, Arizona. Project
Report No. 72-MM-20, EPA, Office of Air Quality Planning and Standards,
Research Triangle Park, N.C.
-------
Appendix B
Calibration Procedures and Data
-------
Calibrations
The Aerotherm HVSS dry gas meter was calibrated by the NEIC pro-
cedures except that in place of a wet test meter a Roots* meter was
used. The fioots meter is traceable to an NBS** standard.
When determining the orifice coefficient (AHa) a 50 cycle timer
was used on 60 cycle electrical current to time the sample volume
This gave an erroneous AHa which had to be redetermined in the field.
The 3-2 pitot tube B leg (the leg used) had a calibration co-
efficient (cp) of 0.79 at the beginning of the Marana Survey and 0.78
at the end. This 1.3% change is insignificant and since Method 2
(40 CFR PART 60) (Appendix A) makes no mention of using the post-
survey pitobe calibration coefficient, it is inferred the initial co-
efficients should be used in all calculations, i.e., for isokinetic
and emission rate determinations.
The initial coefficient, however, was affected by probe blockage
in the 41 cm (20 in) duct. In this Appendix is the Blockage Data and
Graph, which were used to adjust the Sampling Data. This data was
developed experimentally by NEIC.
* Brand Name.
**National Bureau of Standards.
-------
NEIC PROCEDURE FOR
CALIBRATION OF DRY GAS METER
AND ORFICE METER
Dry gas meters are used in source testing units to accurately
measure sample volumes drawn during testing. A critical orfice is
also installed to provide a known sampling rate so that isokinetic
sampling can be maintained. These units will be calibrated before
and after each sampling trip.
Calibration is accomplished by making simultaneous total volume
measurements with a calibrated wet test meter and the dry gas meter.
The wet test meter must be previously calibrated from a primary standard.
Calibration is performed follows:
1. Level wet test meter and adjust the water level to the
proper point.
2. Level and zero the manometer on sampling control unit.
3. Leak check unit and air hoses at 15 inch Hg (leakage rate must
be zero). Assemble vacuum line to the wet test meter.
(Caution: NO NOT Leak Check System by Plugging the Inlet to
the Wet Test Meter, this will cause internal damage to the
meter,)
A. Warm up control unit by operating vacuum pump for 30 minutes
with wet test meter connected in series.
5. Close the course valve and open the fine adjust (by-pass) valve.
6. Turn or vacuum pump, open course adjust valve and turn the fine
adjust valve until manometer reads 0.5" 1^0 (AH).
-------
-2-
7. Simultaneously record the dry gas meter reading, wet test
meter reading and time. Record temperature of wet test
meter, inlet and outlet temperature of dry gas meter and
atmospheric pressure during the test run.
8. Allow pump to run until the wet test meter indicates
exactly 5 cubic feet of air have passed through the system
(10 cubic feet when a AH of 2, 3 and 4 inches H^O are used)
and record time.
9. Repeat steps 5-9 for AH of 1", 2" 3" and 4" H^O.
10. Calibration record will be kept in a permanent file at NEIC.
Copies will be made for field use.
Calculations
Calculate the accuracy of the dry gas meter (y) as follows:
Vw Pb (td + 460)
If = Vd (Pb + AH (tw + 460)
13.6)
Where: ^
V = Volume of gas metered, wet test meter, ft.
w
3
= Volume of gas metered, dry gas meter, ft.
P, = Atmospheric pressure, inches Hg
D
t^ = Dry gas meter temperature, °F (tj in — t^ out)
tw = Wet test meter temperature, °F
If y 4 1-00 (+0.02) then gas meter will be taken to Public Service
Company of Colorado gas meter shop for adjustment and/or repair.
Orfice meter coefficient (AH@ = 0.317 AH
P (td+460)
b
(tw+460) 9
Vw
-------
-3-
Where:
3
Vw = Volume of gas metered, wet test meter, ft
Pjj = Atmospheric pressure
= Dry gas meter temperature, °F
tw = Wet test meter temperature, °F
9 = Time elapsed, minutes
-------
Orifice Meter Calibration
Date ^ Box flo.
Barometric pressure, Pu= in. Hg Dry gas meter No.
29. <7/
Ori fi ce
Manometer
setting,
AH
in. Ho0
Gas volume
wet test
meter
vw>
ft3
Gas volume
dry gas
meter
Vd>
ft3
Temperature
Time
0,
min
l.ol
Y
JO?-
AH@
Wet Test
Dry gas meter
Meter
Inlet
Outlet
Average
V
°F
*di'
°F
tdo'
°F
td>
°F
0.5
¥> 9
.*rro
'7 0
,—
rx
—
2 60
f.c 0
.093.F
1.0
V
f.o
7o
—
rL/
—
J. ?o
/ OP
.o9?Z
2.0
3~- 0
r. &
9o
—
ry
.
/ c/o
>.07
,/cS
3.0
9.9
/& O
*/>T>
—
rv
¦—
7 2°/
ifo(
Jo 3
4.0
/&¦ 0
/0¦ 0
?D
') )
(fo-K/C r>) C,
p
(5~-0 ) (it/ 9/ /¦ .O'iCt }{-S-PCib)
ct! (?PL
y. <7 J
1.0
0.0737
(>/, C/Y?r;Jf fv i> Vf;
f.oW7)([\ (Co.) / 91
(0 /- ^A>)
¦?u.r'/ f
ya J
3.0
0.219
(9 t Yl f/)f V'/ r 'sr...)
(0 ?/7)f n
*->Ys J'r
C fn J) Co'/ c-/ -f- (/&v\
/Vv/^V
J
4.0
0.294
f'&){2 '/. t>/)(y£j -hVCti)
f. 0 >/ yJ ( 9)
\ .. j")
(/u>) C¦>'/
.?(/ / (7cf*(/c^) L /o 1
c O 5 & I
Where: V,,
vU
Tw =
^ =
Pb =
0 =
Volume, wet test meter Cali
Volume Dry gas meter
Temperature, Wet Test Meter
Temperature, Dry Gas Meter
Atmospheric Pressure, Inches Hg
Time, minutes
bration by:
Checked by:
r ~
Remarks:
4/24/77
.{A"'l «¦
J
. '1 1 t'1 /
{ • ' ' -1 V r"|
i*"" )
^ ,
5." .,<<> e*t>
/ . /,
-------
Orifice Meter Calibration
Date /P-Zv 7~7 Box Mo.
~T
jH.'iZ
Barometric pressure, P^= in. Hg Dry gas meter No.
Ori fi ce
Manometer
setting,
AH
in. H9O
Gas -volume
wet test
meter
V
ft3
Gas volume
dry gas
meter
Vd>
ft3
Temperature
Time
0,
min
y
AH@
Wet Test
Dry qas meter
Meter
Inlet
Outlet
Average
\r
°F
*di'
°F
V
°F
td»
°F
—£
0.5
5
£jo
i,n
¦-
/
—
fl
.0./?
1.0
5
£H
(-V
-
U>
-
L^i
7?
C &
2.0
10
fOJO
LX
—
U
-
f.oo
. o(?L
3.0
10
/DO
i¥
13-
-
/.?3
f,o!
.067
4.0
10
??t
b
—
7?
-
/•Lo
f.o'X
,0 U7
Average
/,oo
,b(nll
AH
0.5
aH
13.6
V., Pu (t^ + 460)
. /13.6 N
0.0358 4-^Q)
I &0 (lH C'i -r 4 U.C.Q
0.0317AH
Ph (tH+ 460)
(t„, + 460)0
- Vw
0,o'j>\~\ % v.ij
¦if
ItcVc?/
iT
O o3["?^/,o \ YC'fi1 .-j"'/ 7*-
1.0
0.0737
-ry.11-! dr-(7t td^>>
/c ((mv-w«,/'/?XGz.-f
0.0-71^^^0 fYGi-t^c^z. 1L
3.0
0.219
/C *-? t/.M5Y'7'Z-VMbo')
/o o /V'-U ?-r.Z-\'t) (CA^L-.O)
i it—^ i.
?ilK( '1 It LfbO)
£
/o
(» g>^Q *3.0 fV^A ^
7"2--v^u^L /o J
4.0
0.294 j /rwi-'/ >7/,Y
I n?-^ ui^Yb'z *- wud)
P31~? /¦ o z-t >4!,c>, i ¦UH'!-
^<{Le)L . >o J
Where:
v3
Tw
Id
Pb
e
Volume, wet test meter Cali
Volume Dry gas meter
Temperature, Viet Test Meter
Temperature, Dry Gas Meter
Atmospheric Pressure, Inches Hg
Time, minutes
bration by: Q V^uA
Checked by:
+/£-
Remarks:
4/24/77
-------
J !
i
101
100
P0
oj;
2. 0
1 . 0
I-.
--L-
a
> - !
1 * I '
1-
M
_.l _
—!a ecu Hag
i !
I.
¦f
I
reijir
e-
111
i.
-l-L-i--. ^
jPEDVin R TEST
"•r DlRESSE'lkiMIIAsb-R^MII
j....J. -j -HbU.yi'ON.-tEXAS-
Mr.Tt:H! INFORMATION . I
Mi.clrl j ! JA7(25 \ j""
Hi i*t¦ 11 No. i ''76£0S2b j
Hill -of Xlat'lr- C> -
Order No. j _ \ds/.330I
i D(Ua pbln
|" Pdjulp'pnen
i.
NT!
-I
med us
r diimcr
~1""
nj?
IT rover ;Used:
jTcsl Da'ter'
3-/7-77
I- - : - U
D. w, Bxaho./-/*
[Tested hy:^_
1 ¦ k : •-/¦ y/\
•Certified 15 yQ U f
I ' '
. J.
.1
I
1
i
nt proving'
ic r*!alir{>n.il|
vithiairiat
peraturje.
... ...
- ; 1.
i^ERENpAL CURV.E
i -
(•? A
- P3 i i | ] n t i |\i I 1" I i OM ill I I (lAi
-------
NEIC Procedure for Pitot Tube Calibration
Introduction
The Type-S pitot tube is used by NEIC to measure stack gas
velocity during source sampling. The pitot tube coefficient (Cp)
of this instrument is determined by calibration against a trace-
able National Bureau of Standards (NBS) standard pitot tube. The
Type-S pitot tube is calibrated on a probe sheath with a h inch dia
nozzle atta'ched. All pitot tubes are calibrated from 305 m/min
(1000 ft/min) to 1524 m/min (5000 ft/min). Pitot tubes used during
tests will subsequently be recalibrated at a minimum of 3 points
within the velocity range observed during testing. Tubes which have
been damaged or suspected of being damaged during field use will be
recalibrated over the entire range (i.e. 305 to 1524 m/min).
I. Equipment Required
A. Flow System - Calibration is performed in a frlow system
meeting the following minimum requirements:
(1) The air stream is confined in a well-defined cross
sectional area, either circular or rectangular.
The minimum size is 30.5 cm (12 inches) diameter
for circular ducts and at least 25 cm (10 inches),
as the shortest dimension for rectangular ducts.
(2) Entry ports provided in the test section, shall be a
minimum of 8 duct diameters downstream and 2 diameters
upstream of any flow disturbance, e.g. bend, expansion,
contraction, opening, etc.
-------
(3) The flow system must have the capacity to generate over
the range of 305 m to 1524 m (1000 ft. - 5000 ft.)/min.
Velocities in this range must be constant with time to
guarantee steady flow during calibration.
B. Calibration Standard
A standard type pi tot tube either calibrated directly
by N.B.S. or traceable to an M.B.S. standard shall be
the calibration standard.
C. Differential Pressure Gauge
An inclined or expanded scale manometer shall be
used to measure velocity head (aP). Such gauges 3iall be
capable of measuring AP to within + 0.13 mm (0.005 inches)
'^0. A micro-manometer capable of measuring with 0.013 mm
(0.0005 in) ^0 will be used to measure aP of less than
13 (0.5") H20.
D. Pi tot Tube Lines
Flexible lines made of tygon or similar tubing shall
be used.
E. Thermometer
A mercury in glass or other type thermometer checked
agains a mercury in glass thermometer is considered suitable.
F. Barometer
A mercury column barometer shall be available to determine
atmospheric pressure.
II. Physical Check
1. The openings are sharp and do not have a rolled edge.
2. The impact planes of sides A & B are perpendicular to
the Traverse Tube axis [Figure 2].
-------
-3-
3. The impact planes are parallel to the longitudinal tube axis
[Figure 3].
III. Calibration Procedure
The Type-S pi tot tube shall be assigned an identification
number. The first digit of the number is the effective length of
the tube, followed by a dash and consecutive numbers for the number
of tubes of the same effective length, i.e. 5-1 signifies a five
foot pitottube and is the number one tube. Calibration proceeds
as follows.
A. Fill manometer with clean oil of the proper specific gravity.
Attach and leak check all pitot tube lines.
B. Level and zero monometer.
f
C. Position the standard pitot tube in the test section at
the calibration point. If the flow system is large enough
and does no interfere with the Type-S tube the standard
tube may be left in the system.
D. Insert the Type-S tube into the flow system.
E. Checks for the effect of turbulance are made as follows:
1. Read AP on both Type-S and standard pitot tubes with
the standard pitot tube in place and compare with read-
ings when the standard tube is withdrawn from system.
2. Read aP on the Type-S tube at centerline of flow system,
then take readings while moving the tube to the side
of the system. This will define the boundary turbulance
layer.
3. Position the Type-S tube so that there impact openings
are perpendicular to the duct cross sectional area 2nd
-------
-4-
check for null (zero) reading. Absence of a null reading at
this position indicates non-laminar flow conditions.
F. Read AP and record on data table.
G. With the Type-S "A" leg orientated into the flew read aPs
and record on data table.
H. Repeat steps F and G until three sets of velocity data
have been obtained.
I. Remove Type-S pi tot tube and rotate probe nozzle until it
aligns with side "B" impact openings.
J. Insert the Type-S pi tot tube and proceed as in steps F through
K, Adjust flow system to new volocity and repeat F-J.
f
L. Record air temperature in the test system and barometric
pressure during testing.
IV. Calculations
1. At each "A"-side and "8"-side velocity setting, calculate
the three valves of Cp (s) as follows:
pi tot tubing inches H^O
APS - Velocity heaa", measured by the Type-S
pi tot tube, inches ^0
2. Calculate Cp, the average (mean of the three Cp(s)
valves.
H.
Where:
CPs ~ Type-S pi tot tube coefficient
Cp gj-j - Standard pi tot tube coefficient {NBS)
fiP - Velocity head, measured by Standard
-------
-5-
3. For each CP calculated in step 2, calculate o, the average
deviation from the mean as follows:
a(Side "A" or "B") = 2]cp (s) - Cp (A or B) |
3
4. The pi tot is acceptable if:
(a) The "A" and "B" side average deviations calculated by
• equation 2 are <_ 0.01.
(b) The difference of the "A", and "B" sides Cp calculated
1 \ •*
by equation 1 is <_ 0.01 for each individual velocity.
5. Calculate the test section velocity as follows:
V = KCp /TAP std
V pm
Where:
V = Average test-section velocity, ft/min
K = 5130 (constant)
Cp = Coefficient of standard pi tot tube
T = Temperature of.gas stream °R
P = Barometric pressure, inches Hg
M = Molecular weight of air = 29.0
aT std = Average of the three standard pi tot
tube readings, inches ^0
V. Record Keeping
Flow system data and information on each pitot tube shall
be recorded in a bound book.
The flow system data shall include:
1. The tunnel cross-sectional area and length
up-stream and down-stream of the test site )ft.)
from disturbances.
-------
check for null (zero) reading. Absence of a null reading at
this position indicates non-laminar flow conditions.
F. Read AP td and record on data table.
G. With the Type-S "A" leg orientated into the flew read &PS
and record on data table.
H. Repeat-steps F and G until three sets of velocity data
have b'een obtained.
I. Remove Type-S pi tot tube and rotate probe nozzle until it
aligns with side "B" impact openings.
J. Insert the Type-S pi tot tube and proceed as in steps F through
K. Adjust flow system to new volocity and repeat F-J.
L. Record air temperature in the test system and barometric
pressure during testing.
Calculations
1. At each "A"-side and "B"-side velocity setting, calculate
the three valves of Cp (s) as follows:
Cps - Type-S pi tot tube coefficient
Cp std - Standard pi tot tube coefficient
-------
-6-
2. Time tunnel used (hrs)
3. Air temperature (°F) in flow system and barometric
pressure (inches Hg).
4. All checks for turbulance and flow distribution.
5. Velocity range (ft/min).
The pi tot tube information shall include:
1. I,D. number
2. Checks for physical dctfpages, errors noted and
modifi cations.
3. Dates and surveys pi tot tubes were used.
4. Date of calibrations, coefficient and dates of
re-calibration.
The calibration records will be kept on file at NEIC. Copies of
the appropriate calibration dates will be furnished for each source
test project.
-------
c
7"
•A
Figure 1. Measurement of Type-S pitot tubo length (dimension "a"') and impact-picr.e
separation distance (dimension "b").
TRANSVERSE
TUBE AXIS
IMPACT
PLANES
'Figure 2
e-S pitotitube, endi
pendlcular to transverse tube axis..
lA^ID'EPLAIJE
Ldhgitud^al
tube axis
B-SIOE PLANE
w CiiH. i
Figure 3. Type-§ tube, top view; impact-open
¦ing planes parallel to longitudinal tube axis.
From "A TYPE-S PITOT TUBE CALIBRATION STUDY" by
Robert F. Vollaro, October 15, 1975
-------
US Fnvironmental Protection Agency
Nat:onal Enforcement Investigations Center-Denver
Calibration Pitot Tube:ID Number AJ AS Cp C ¦ c/ /
Type-S Pitot Tube ID Number: -
p / £'c
6 - 1l<,
i^7 ' / A >-i
0 ¦ u
ft, t fro
^>,/K
n
a
/) r s2 7
r)-u z
O.c/3
o 19 a
6 .0^
( J
f>. 3
p. ?
7S«
0< 3.7
o • V*
O V-iU
6.7 V
o
1
0, 9/
/?, 9/
6 l%c
0 >7&>
^7'c/, ^ ~Zj^ {)t{^C
/), 9/
O . 3s
c 7^'
O- 7fc/
/' 3S~
/' 3 o
A.Ill
O
/.
/.**
7,,rs
6,m
o^l
/' fo
g-i*t
0, 7^7
/'tjr
o 111
6.111
/
*
ro
Ojn /
•J / /A7n
OOlo
Leg Average Cp
.u
probe sheath attached *
nozzle attached l.u . v
okji
nozzle attached > '/jl
sampling'isokjnetical 1y
//
Performed By:
Calibration Date:
9 -23 - 77
if''' "
-------
US I'nviroiir.-nL?! Protection Agency
National Lnforces?nt Investigations Center-Denver
Calibration Pi tot Tube: ID flumbor / , 7~-» Cp / /
Type-S Pi tot Tube ID [lumber: ',? '/ ^
1 > v_"7 j> | C n-?^- C -
^ ^ j
A -•— / f.
^P
Standard
Pilot
Ap S--Tvpe Pi tot
CP
Cor.Tents
A leq
B leq
A
B
AO ¦
/' £2.
/• 7r b>
,777
.170
/ C
/ 0 2
/ t-;
¦7£
, 37
nYx
,170
3-/C/K
. ¥/ o
,T6L!
,7 7't
O 7-T
//<~
O vl
. i ,
,799
."7 %.%
// G>
. \
, 1%
r?x&
O 7"~
¦ /!C
,1%
,17?
/<
,7?X
.776
Leg Average Cp
During Pi tot Calibration:
probe sheath attached /c-•;
nozzle attached
sampling isokinetically .
j /' / ^ f
Performed By: ' "f /K~^' ' "c- Calibration Date: .v /.-/ J, ,
I /
-------
1 1 I t 1 ' 1
1 2 3 4 5 6
Decrease 'in Pitot Tube Coefficient, Percent
FigureB-2. Plot of Blockage - Percent vs. Decrease in Pitot Tube Coefficient - Percent
-------
C/3I 7- 8 £ ft JZ£ a.<>
AFi G!{/
yr
o f
/\Jd>ZZ/!tL
SV^7
3. V7f
3. 5/y
--3/4^7
. /.
£
7
C4
7
/! r<£
/, * .
(jl?^.
. /,_*7_3
J< ^... /..
_/. _
. / '.^
.!'.
-------
Appendix C
Sampling Train Construction Details
-------
Aerotherm HVSS Description
The Acurex/Aerotherm High Volume Source Sampler (HVSS) consists
of a control unit, vacuum unit and a sampling unit. The units are con-
nected together with quick disconnect electrical and air lines, and
umbilical cords. Not all the equipment used was of Aerotherm manufac-
ture and this equipment is described below the item it replaced.
A) The HVSS control unit contains the following (figure 1):
1. All temperature and electrical switches and
controls.
2. Dry gas and orifice meters to accurately
determine the sample volume and sampling rate.
3. Magnehelic gauges to indicate the pressure
drop of the orifice meter and pi tot tube.
The magnehelic gauges were not used during
the Marana survey. Inclined manometers re-
placed the magnehelics.
4. Digital Temperature Indicator (DTI) which
gives an instant readout from several points;
stack, oven, impinger outlet, meter inlet,
meter outlet by use of a selector switch.
The DTI was only used to monitor the meter
inlet temperature. The duct temperature was
measured with a dial or glass thermometer.
B) The vacuum unit (pump) is capable of drawing a high vacuum
(65 CM Hg) and a high volume (280 1pm - free flow) of air. The
pump is a rotary fiber vane type which does not require lubrica-
tion, but oil bath filters are used for pump protection.
Attached directly to the pump are the flow control and bypass
valves for adjusting sampling rates.
C) The sampling unit is made up of three distinct sections; impinger
case (figure 2), oven and probe. All three units can be converted
to form one sampling unit or can be separated for unusual sampling
conditions. Below are the individual component descriptions:
*Information included in this description is from the report,
"Operating and Service Manual, Source Assessment Sampling System,"
D. Blake, Aerotherm Report UM-77-80.
-------
DRY GAS METER DIAL
ELAPSED TIME INDICATOR
MAGNlMfUC.
r\ AEROTHERM
' ACUREX Corporation
FAN ON/OFF SWITCH
PITOT INLETS
SAMPLE INLET
SAMPLE
EXHAUST
115V 15 AMP
POWER INPUTS-
POWER TO
PROBE AND
OVEN HEATERS
OVEN FAN
THERMO-
COUPLE
INPUTS —
MAGNEHELICS
SHOWING AP
FOR PITOT
TUBES
MAGNEHELIC SHOWING
£P FOR ORIFICE
PROBE HEATER
CONTROLS
OVEN HEATER
INDICATOR LIGHT
OVEN HEATER
ON/OFF SWITCH
OVEN HEATER
CONTROLS
•SELECTOR SWI TCH KEY
— MAIN POWER
ON/OFF SWITCH
MAIN POWER
INDICATOR LIGHT
-DIGITAL TEM-
PERATURE
INDICATOR
PROBE HEATER
INDICATOR
LIGHT
PROBE HEATER
ON/OFF SWITCH
Figure 1. Control unit.
-------
Figure 2 Impinger train.
-------
1. Impinger case - an uninsulated fiberglass case
capable of holding four plastic impingers (1
liter capacity) in an ice bath.
2. Oven - an insulated, double walled, stainless
steel (S.S.) box that can hold the cyclone and
- filter holder. The S.S. cyclone and filter-
holder connect with S.S. fittings and Teflon seals.
The filter support is a S.S. screen.
3. Probe - a S.S. lined, external sheathed probe.
The sheath, which contains the liner, pi tot
tube, and thermocouple connections, is 6.4 cm
(2.5 in) in diameter and connects directly to
the cyclone inlet.
The HVSS probe was replaced with a Scientific
Glass (S.G.) Inc. AP-5000 S.S. lined probe dur-
ing the Marana survey. The S.G. probe was used
because it has less duct blockage than the HVSS
probe. A flexible Teflon probe connected the
S.G. probe to the cyclone inlet.
-------
Rader HV Sampler Description*
"The Rader Hi-Volume Sampler provides a means of determining par-
ticulate matter in emissions. It has developed over several years, and
results obtained from a variety of sources have proven it to be a
versatile and reliable sampler.
"The Sampler is illustrated by Fig. 1. It consists of four assem-
blies. (1) The Filter Holder Assembly houses the filter support and
filter. (2) The Inlet Extension Section is clamped to one side of the
Filter Holder. The 1-7/8" inlet nozzle is recommended for velocities
below 2500 FPM, additional nozzle adapters are available for higher
velocities. (3) The Control Section is clamped to the opposide side
of the Filter Holder. It consists of the flow sensors, control valve
and suction blower attached with flexible hose. (4) The Control Com-
puter which performs all needed calculations to run the stack test.
"These operating instructions have been developed to enable the
operator to collect an accurate sample with a minimum of effort.
Briefly, a pi tot traverse is performed to determine the velocity at
the sampling points. In some instances, the velocity may be calcu-
lated with sufficient accuracy to select the nozzle size. The sampler
automatically regulates to achieve an isokinetic sampling rate, but the
pi tot traverse may be necessary to determine volume flow. Samples are
then collected on pre-weighted filters using blanks for monitoring any
changes to the filter tare weights.
"Since particulate will accumulate in the inlet probe during a
test run, the procedure provides for the collection and inclusion of
this particulate in the sampling results."
The HV sampler is all aluminum with rubber gaskets and seals. The
maximum sampling rate is about 90 cfm.
*Information included in this description is from the instruction
manual for the Rader Hi-Volume Sampler (Automatic) distributed by
Rader Companies, Inc.
-------
SUCTION BLOWER
CONTROL SECTION
CONTROL COMPUTER
FILTER HOUSING
INLET SECTION
RADER MODEL A-2000 AUTOMATIC STACK SAMPLER
FSG. 1
-------
The Rader HV sampler operates in the following manner:
1. A standard type pitot tube, adjacent to the sampling nozzle,
senses the stack gas velocity and this pitobe differential
pressure is transmitted to the control computer via air lines.
2. Simultaneously the unit sampling rate is measured by the
orifice meter and the orifice differential pressure is
transmitted via air lines to the computer.
3. Based on the velocity pressure, the computer adjusts the
solenoid valve until the orifice meter pressure corresponds
to the sampling rate necessary for isokinetic flow. The
computer uses the sampling rate at the orifice meter and
the sampling time period to indirectly determine the
sample volume.
The Rader HV sampler does not meet the following Method 5 requirements:
1. No moisture determination is performed.
2. No direct measurement of the sample volume is made.
3. Stack temperature readout is not available.
-------
Appendix D
NEIC Analytical Procedures and Data
-------
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF ENFORCEMENT
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
BUILDING 53, BOX 25227, DENVER FEDERAL CENTER
DENVER, COLORADO 80225
to Mr. Paul dePercin, Field Coordinator DATE January 6, 1977
Marana Cotton Gin Study
from Chief
Chemistry Branch
subject Results of Particulates Analyses
Attached is a summary of the results of particulate analyses of filters and
acetone washes collected for the Marana Cotton Gin Study.
In addition, several samples are being analyzed for phosphorus pesticides
and those results will be sent to you shortly.
Theodore 0. Meiggs
Attachment
cc: Harp
Young
Stager
-------
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF ENFORCEMENT
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
BUILDING 53, BOX 25227, DENVER FEDERAL CENTER
DENVER, COLORADO 80225
to Chief DATE January 6, 1977
Chemistry Branch
from w. E. Stager
(Reviewed by D. Vietti)
subject Results of Particulates Analyses for the Marana Gin Study
Analytical Procedures
Filters
The filters to be tared were desiccated at 20 +_ 5.6°C (68 + 10°F) and ambient
pressure for 24 hours and weiqhed to constant weiqht. The tare weights were
recorded to the nearest 0.1 mg. During each weighing, the filters were not
exposed to the laboratory for more than two minutes.
After sample collection the filters were returned from the field folded in
field data cards in sealed envelopes. The data cards and filters were re-
moved from the envelopes and placed into a desiccator after the circle charts
had been stapled to the cards. Indicating Drierite, which removes uncombined
water from the filters, was used as the desiccant. The filters were desiccated
at 20 + 5.6°C (68 + 10°F) and ambient pressure for 24 hours and weiqhed to a
constant weight, i.e., a difference of no more than < 0.5 mq, or 1% of the
gross weight minus the tare weight, whichever was greater, between two con-
secutive weighings.
The single pan analytical balance was calibrated against Class "S" weights
before weighing the filters. Additionally, desiccator and weiqhinq room
temperature and relative humidity readings were recorded. All handling of
the filters was performed with forceps.
Acetone Wash
The acetone probe washes were received in quart jars with Teflon lined lids.
The level of liquid in the containers was noted and no leakage had occurred.
It was noted, however, that the volume of acetone in the field blanks was
considerably less than the average sample volume. The volume of acetone in
each sample was measured volumetrically to within + 1 ml and recorded on
the bench sheets.
The samples were mixed to suspend the solids therein and transferred to
tared 250 ml beakers, as was the acetone used to rinse the jars. The beakers
were placed in an aluminum foil tunnel-designed to prevent particulate con-
tamination of the sample, yet allow efficient air flow for escape of acetone
vapors-in a hood. The hood door was kept closed and empty tared beakers were
used as blanks to verify that the samples did not become contaminated.
-------
- 2 -
After a minimum of 24 hours in the evaporating tunnel, the beakers were
transferred into a desiccator having Drierite as the desiccant. After 24 hours
the beakers were weighed to constant weiqht with at least 6 hours between
consecutive weighings.
Field and Laboratory Blanks
For both sizes of filters-6" and 8 x 10"-and for acetone washes-both field
and laboratory blanks were collected and weighed at the minimum rate of one
blank of each type for every ten sameles.
Statistical treatment of the field and lab blank data were used to determine
the detection limit utilizing the formula:
D.L. = x + 2^
-------
ANALYTICAL DATA REPORTING FORM v Page
NAK.E OF SURVEY iV. o\uy hi h 7
SAMPLE NO.
STATION DESCRIPTION
TIME
ANALYSES PERFORMED
"SituL 4aHj;.s
Acf ItfYoC U'c^l-t
/) 'i v 1»ri ¦ i^O
t
iaf)-6MloW
/V] v-cwiet S-w
/ 5-00
V * | ^32*7.
o !
>T.||-0> - IIOS"
h /I
/03O
37
) S~3 £>
11,11 -&3-M D5~
M .]
/3 7.53^
id
»
a 5 jo
>> ;>
Ha O
13
¦V7c//-c2- not
>> >,
13clO
H
H c(
-"°7
" >>
}$ ZO
7
S'7
1T-C3 -ff- // C ^
» >¦
08^5"
/
1 3-0
7tit,f
n
)&<5-
I
I1e'' *
< 1
m
-Vc f
" h
ISHO
< i
t,-lc^-g'J • luc
\
>>
Dbsy
1
: / 86
}to?-c,l' H lo
7,
) 25"2-
.5.5 jr
y>v?< L^~ i| /c,
" h
lb CO
?
H70
2-11>7 -c^> n if
" I
3
fjO0!
VU.7-i.W- j//f
'
1 ol t
0
Ho3
»
1
J
|
-------
ANALYTICAL DATA REPORTING FORM Page 2-.
IAME OF SURVEYt\UYfty,t £,n f)» DATES COVERED n In In >ill(J?7
'AMPLE NO.
STATION DESCRIPTION
TIME
ANALYSES PERFORMED
Eftw
fW<- Wo ll/r->t\
'vvcy
/Vl /t fici G-1 n'
Kl/5"
0
1
\£%
Oicr-cl-i/;J
11 *'
/5"00
]
IHC
T.
6% 33
< 1
III
S-ef -'111—
loso
3
L'JX
V-v1-} cL-mt
rvr(->
H
JII
1.T-C 4 -o J - M 12-
I*
. /520
6
6ZZ
2-7oH-u ''
0^7-
17
ctl)
2-3 lO'Ol -//'S
' 1 '/
)7M8
16
I 70
'/ *
W8
43
/76
7-Vc-oJ-''11
'/
IL^
23
! .a/^
llo'f-ol -II if
• t
;
o?or
¦2-7
MX
}*.v9-c2- ii it
' ' '/
) /cu
31
HRZ I
7_V?03"
13 30
\S
2Ln
¦/
2-7
7-h7
"C| - II 1 L
V
'/
bZHo
1
o
I Ol
"
lo? 5~
>—
ilti
?5c7.i3-uiG
!2c&
2-
137
|
-------
ANALYTICAL DATA REPORTING FORM Page
E' OF SURVEYhrw.\x ft tuning 4rr;wt ^ DATES COVERED ulnln ft If 9 {77
?LE NO.
' STATION DESCRIPTION
TIME
ANALYSES PERFORMED
3tv k A"/5t5
Ace-U^
vw*
^"kC-ol -"O
Tviflta fhlv:'
o*'-ys-
U
K 1
/HS-'
1 > '>
/o;
I
\S'o°l
7-3oS'-o^-';,5
*) »t
/•5"3D
•%/ ¦
H6G
2 Jo .5-03
t' i>
0132/
£S
19B
;?ii— "oM
/?«- li 0 fiu-i ic.
/5-0G
ar
' c>.^
>H| -I/O
QfOO
/. ^
0' /
}i«S /Ic^c
1,
J 3 c/3~
0.7
0-3
3-TO— /ID
'<
Ic^-y
0, 1
?.7cM 'II11.
I. ¦/
j^y-o
/. f
0- 6
') >1
o.&
0 -5
1307 >UC
n "
1 <4'] 5—
—
OM '
uh^jii
LrJtr /V UP fc - p , ( /7>
-------
rillMi*! UyMU un'ih i^wi ^nl ih* v*' i ^ j
NAME OF SURVEY ff)aY&71 ct Ay^oha DATES COVERED //A//77 ft I (f! 71
SAMPLE
NUMBER
STATION DESCRIPTION
TIME
ANALYSES PERFORMED
w;-ln Filf-fj-
uj+. <*-
J.3.0l-o( - f/D^
L t «"1~ Co. SS
?->Lc>l- ol ->loci
•t M i, ,;iy
o%lo
\,)%
yio\-o \ -1111
'• " " llv
30
0>$%
~X7o^- 0! - McH
Liw 4- Co ft
n9fo
1sio~i-- ol- ho M
J
M "
\Z7o
1> XH
%-U>2-o^- »1 0<-|
'¦ 1 •
/r/4
%co3
11 c3 -0/- i/c7
c-ha- C L**d~ C-Gw^r )
!oo%
5". 31
P-Zo3~t> "2- ~ llol
u
<1 M
U«< e
5,0 6
it /1
'4(7.
7,7 /
7-2^- C>M * lloff
M '1
<59 TG
3 11 0
7 7-Oj - 0 ^T-J ( 0 ^
• M i(
/3 2-2.
6.ssr
JioV cb- iloc/
H 1 /
; <-] 10
4.^7
^2o3 -o| - 1(0f
Lvvvi" Cc*.o^ /"/ V
/ cc-"3-
hll
V7o3-cT--M Orf
«/
II '»
g.iK
'Z'Z -c *1) " U o rf
1' ,' l'
IHIO
5.7 X.
-------
NAME OF SURVEY Chm DATES COVERED /»/«//77- n I/l(77
SAMPLE
NUMBER
STATION DESCRIPTION
TIME
ANALYSES PERFORMED
Hi'Vnl /f-fv-
1
UJ+ • T <
jl2o"3-o I - ||o4
J
RiOl 0
- )i0 $"
»» '' '»
^2&V-o| - 1 f o£>
M " !¦
¦—
a
W?.0 '
LtoS-o 1 - ) u 1
''
—
11116
£.2©Vo 1 - 111 0
1
"
—
B
life.4
?0 / -c ("III *^—
)Vl £»" V" C^U CL. *" ft lw.C-Uv^"
6.
32bf-o| " 'I' M
11 " II
—
to.SST
1 - :| lb
" " II
—
X3ol-ol -h ot
l~ wvf" Co-O?
/oc? X
3,87
^3c'i-01-- ;i } V
— 5—J
>,
/fl t
sr
L3ol -o3 - li/l
. / ''
/5-SO
4, 00
l3c.1.-o.I -ii)3
II '!
PS
M?
5-S02-O7--1113
>1 >1
to}<1
1.45
W3- -03-" ,:i
1!
>1
mv
a. 3.5
'-3o3- cl - l| I ?>
M |i
IHIL
1.4 a.
^ovo7." Ill 3
" 1
j..? ^
>3C3-C3 "///<-/
( .
D % "W
0^1
- K // ik'r c
'¦ ( r' • 1-Gx J- /v rl(~cb tv,< f'(C&V
t/L>. J.
¦Jr ,aUtrt
rfilfhli (M.1 |i^3 /;0-
^ kJ- Cc_
1
-------
OF SURVEY^ hl'.l „
J
— o * 'T *vw^
(/
J
J
— —
— ——
' ¦¦ ¦ ¦
-------
NAME OF SURVEY -Damci***. on q=- DATES COVERED nh^jn-— ////W77 ^
v ~
SAMPLE
.VJM3ER
STATION DESCRIPTION
TIME
ANALYSES PERFOR
•1 E D
Hi Vol Fr /ffv
to4~. ct,
"X 2o<-l-c\ ~ || m
L 1 n\~ C&ct-c
09 ^6
u
,6a
730H-0I -11/4
J
»l *1
\,%y
710V-C I-/Iff
imo
G.m
23c -c/-iiiLj
M )|
mn
/,"7/
230?-o2--ii 15"
fl '/
o*r®Q
O-70
7-3or-o"3 -1115
// )'
Id3H
2-31 b-o I - 1113
jO w\.C-uav" ft-, r -Ifr \)x}f
—
9-3/O-CM -11/S~
t * 11 (1
.
¦ H.3W
-------
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF ENFORCEMENT
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
BUILDING 53, BOX 25227, DEN1 ER FEDERAL CENTER
DENVER, COLORADO 80225
T0 Mr. Paul dePercin, Field Coordinator CATE January 23, 1978
Marana Cotton Gin Study
h»cm chief
Chemistry Branch
subject Resuits of Pesticide Analyses
Attached is Ms. Carlberg's report which summarizes the results of analyzing
four different samples from Station 2211 at the Marana Cotton Gin for phos-
phorus pesticides. Trace amounts of methyl and ethyl parathion were found
in the acetone washes ranging from 0.03 to 0.8 ug/kg (ppb). The filters
did not contain sufficient particulate material to detect these low concen-
trations which appear to be too low to be of concern.
Theodore 0. Meiggs
Attachment
cc: Young
Harp
Carl berg
-------
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF ENFORCEMENT
NATIONAL ENFORCEMENT INVESTIGATIONS CENTER
BUILDING 53. BOX 25227, DENVER FEDERAL CENTER
DENVER, COLORADO 80225
Chief 0ATE January 19, 1978
Chemistry Branch
K. A. Carlberg
SUBJECr Analysis of Samples from Marana Cotton Gin in Arizona for Pesticides
Background
Eight samples from the Marana Cotton Gin in Arizona were submitted for
pesticides analyses. These samples consisted of four filters and four
particulates/acetone washes. The acetone wash samples had been taken
to dryness and resuspended in acetone before they were submitted for
pesticide analysis. The pesticides of concern were ethyl parathion,
methyl parathion, malathion, and dimethoate.
Results
The four filters contained no detectable amounts of the pesticides of
interest. All four of the acetone washes contained methyl parathion,
ranging from 0.1 - 1.8 ug. Three of the acetone washes contained ethyl
parathion in amounts ranging from 0.1 - 0.5 ug. Malathion was not four.d
in any of the acetone washes. The analytical methods used to analyze
ethyl and methyl parathion and malathion prohibited analyzing the acetone
wash samples for dimethoate.
The results are given in the table following. In addition, the weight
of particulates found on each filter and in each acetone wash, as re-
ported in a memo by W. E. Stager dated January 6, 1977, is given. It
is not surprising that pesticide levels in the acetone wash samples
were greater than those on the filters due to the much higher particulate
levels in the acetone wash samples.
-------
- 2 -
Table of Results
FILTERS
ug on
Filter
Station
mg Particulates
Methyl
Ethyl
v Run §
Date
on Filter
Parathion
Parathion
Malathion
Dimethoate
2211
1
11/04/77
30
<0.05
<0.05
<0.1
<0.01
2211
2
11/05/77
37
<0.05
<0.05
<0.1
<0.01
2211
3
11/05/77
53
<0.05
<0.05
<0.1 '
<0.01
2211
4
11/05/77
78
<0.05
<0.05
<0.1
<0.1
PARTICULATE/ACETONE WASHES
ug
in Acetone Wash
Station
mg
Particulates
Methyl
Ethyl
H Date
Time
Sequence
in Wash
Parathion
Parathion
Malathion
2211
11/04/77
1500
01
3327
0.6
0.1
<1.0
2211
11/05/77
1030
02
1536
0.1
<0.1
<1.0
2211
11/05/77
1345
03
2429
1.8
0.5
<1.0
2211
11/05/77
1534
04
2530
1.5
0.5
<1.0
Methodology
A. FILTERS: Each filter was extracted with 150 ml of acetone for 1 hour
using a wrist action shaker. The extracts were then dried with sodium
sulfate and concentrated to 10 ml in a Kuderna-Danish evaporative concen-
trator. The concentrated extracts were analyzed using a gas chromatograph
equipped with an alkali-flame ionization detector. (GC-AFID)
B. PARTICULATE/ACETONE WASHES: Each acetone wash was filtered through
Whatman #1 filter paper and concentrated to 10 ml in a Kuderna-Danish
evaporative concentrator (KD). The concentrated extracts were then ana-
lyzed on a GC-AFID. The extracts were too dirty at this point to discern
the presence or absence of the pesticides of interest. The concentrated
extracts, therefore, were cleaned up using an Analytical Biochemistry
Laboratories Gel Permeation Chromatograph (GPC) equipped with a column of
SX-3 resin and eluted with 15% methylene chloride in cyclohexane. The
cleaned-up extracts were concentrated to 5 ml in a KD, since 5 ml was
injected onto the GPC. The extracts were then analyzed on a GC-AFID. At
this point it appeared possible that all of the pesticides of interest were
in each of the extracts. In order to confirm this, the extracts were ana-
lyzed on a gas chromatograph equipped with an electron capture detector.
Once again, because of the extreme sensitivity of the EC detector, the
extracts were too dirty to be analyzed. Therefore, the samples were sub-
jected to a Florisil column cleanup. The cleanup used was the one de-
scribed in the FDA, Pesticide Analytical Manual, Vol. 1, Section 211.14d,
-------
3 -
for cleanup of organochlorine and organoDhosphorous pesticides. Unfor-
tunately, dimethoate does not elute from this Florisil column and therefore
confirmation of the presence of dimethoate was not possible. The peak
which eluted at the same retention time as dimethoate in the GC-AFID
chromatograms appears in an area of the chromatoqram traditionally sub-
ject to many interference peaks. Therefore, in my judqment, the presence
of dimethoate in the acetone wash samples is doubtful.
The Florisil cleanup referred to above involves elutinq each sample through
4 inches of activated Florisil topped by one-half inch of sodium sulfate.
The column is eluted with 200 ml of 6% ethyl ether in petroleum ether,
followed by 200 ml of 15% ethyl ether in petroleum ether, 200 ml of 50%
ethyl ether in petroleum ether and finally 200 ml of ethyl ether. Methyl
and ethyl parathion elute in fraction 2 (15% EtO) while malathion elutes
in fraction 3 (50% EtO). This Florisil cleanup proved to be insufficient
to allow the samples to be analyzed by EC, however, using the AFID, it was
found that 2 peaks still appeared in fraction 2 in the area of elution of
methyl and ethyl parathion. However, fraction 3 was free of any peak
corresponding to malathion, thus eliminating malathion as a constituent
of the samples.
Fraction 2 of the extracts, which contained peaks suspected of beinq
methyl and ethyl parithion, were submitted to an alumina column cleanup.
The extracts were cleaned up on a 15 cm column of neutral alumina, deac-
tivated with 3% water which was eluted with three 50 ml portions of
benzene. Methyl parathion elutes from this colu.nn in fractions 2 and 3,
while ethyl parathion elutes completely in fraction 2. After this
cleanup, the extracts were clean enough to be analyzed on the EC-GC.
The peaks suspected of being methyl and ethyl parathion were confirmed as
such by proper retention times on the EC-GC and by their proper elution
pattern from the neutral alumina column.
Kathleen A. Carlberg
-------
Lint Cage
and
Appendix E
Dimensions, Flow Data
Calculations
-------
L.i&T' Loe~> CoL c*cs6-^ 7~~/o aj J~
/t£J /tfb reCociTy 7-*#ve*se.
U/cKG. ts*et> C"S 7~A y-Ae ^ /*s /**e
P&efcr->T;ct> /*« fA-c 7-74B£~c- & c. £-0c~> •'
srAtto/u
JVO -
rt\/e*A?e
t>ocT~ feC +ct'Ty
PVC+
Ot'Ar~?r?£
, . .C/^.
/ifiS PaTA
tFP?)
/ufrc, qaTa
CFfif)
wtfi.j-vsrcp
FLe>&> ?
2^0/
Z9.Z
ZL5~* t
aao
/jr. 0
/3.0
r*
9*>*
3 7*0
Z3&/
T/V
*/3. 0
IT
Z?o 2
TZf
^3.9
—
/?
2.3* 3
3f 9
—
X7-o
I T
230^
s/,r
—•
IX
*SoS~
ccJ
«T3k. £
ft
-------
l/j— VeLoc, i T~y
fs.V? Cp l/Zp\! TJ C* s .9?
r r i n 1
t>- r/y= S9/ R
|/;: rr.vp ^9/ fh ^ zr.cl ,v* /f;
]&*o y)(*rr) UJ
**T /Lt> -#n
~ SiJ. Ji - A'O Cm/ApS —
bct*j p f>&* ce-rseP V&p - 0>9S ^aj
"/Ac tfvL*cc7y /> Ab 'TQJVcb /=b< r*<
0 ^ Ct>77~QAJ tffi-Aph £""/.)
/?v A*f VeUc*1~y oS**
\
is & * of t/ALea*7~ ?-*> a rcCo c»?~y °^
2S\/ FPS /Vr be'*} f*»cerjeb g/ajzcT]
Li«Sr C, (S o Lv A*-e 7~C
^ = 6t>C/— l/s A Cs Tj~)C^ & * *0
- ^ r /-. - /; £*-¦«> r* ^ (§$F)(-H£)
w /3 to o Ar/i.
/t*>e*/*bct> C/=~^ /* i«.
- yj, tf>oc> $c/=s>^
-------
/gyro
-------
Table E-1
LINT CAGE ISOKINETIC RATES*
PRODUCERS COTTON OIL COMPANY
MARANAj ARIZONA
Sampling
Station Rate
Number ft3/min
Nozzle
Area
ft2
Nozzle
Velocity
ft/mi n
Gas Flow
ft3/min
Lint Cage
Area
ft2
Screen
Velocity
ft/min
Isokinetic
%
Short-Staple Cotton Gin
2201 26.5
.0670
396
13,000
129.5
100
396
2202 27.1
.0189
1,430
10,300
no
93.6
1 ,532
2203 27.5
.0670
410
19,000
46.8
406
101
Long-Staple Cotton Gin
2301 27.2
.0670
406
4,000
32.7
122
333
2302 27.2
.0670
406
4,150
31.9
130
312
2303 27.1
.0670
404
2,880
31.7
90.9
444
2304 27.0
.0670
403
2,880
33.3
86.5
466
2305 27.0
.0670
406
5,060
39.7
128
317
* Average values for three runs} except for Station 2203 which is an average of six runs.
-------
i/cl- t'1/ V.-- <¦'
STjt M130Z
j>iT //'c) iY/-.
coe.sr /J
/
— , i/x
/
¦—
— *\
— ¦ ¥Cv4lV5 Uo ft-1. _
st^ko/^}- J_ "/& t~ «
. ////5/77 0^// . . ...
f^ycr PA ejfc-> ^ //e s
^ T^ p>-w 1
-2 £>*>/...~. . 5-J' 5>y /7% .
S 3 6 2 S<£ . .2.2.4
.5.^0 o' .. -5? ?3 /&
53 / „„,5Y t-/ / 7%
;i ;•> ^ ,5' .... c5-7 ;/•> .
-. /)6tr«: f' & <¦-/> ft _ 7v /_-l's- ^ 5''/' //G-
. Pf'iCtf,
-------
5> 7~/9
L(
• r "
5 <0 — AJ o
cTc/ 5 /~b
to 4» ~ 1"
/
¦—
/ - .33
r ^ ^
5 - 'J2 2.
7
3
5 . 8<3
y
/ ^6
y .. /Y
.5"
- . < /V
.* , '1 ..
. ..
. .., /7. _
t .. .., /'/
7 -
- « t7
7 . ./?
S-.
. A?
S' .. .,11. .
- %- -
_ .. , _ ..
... ? , /d .. .
- — — -
— —
sr/tf".. -<93.o^_ ... .
- .... .. .
- - •
- _ _. .
, "A .Aic>t . .
. /.--
— >£S
/
,Ld_ .
5 ~
. «M % -
JZ —
,7X
3 -
- . . £ 7 .. .
¦72 ..._. .
7
.ȣ> y
-- - - -? "—
,5;_. ~
7L_.
- - 5*. —
-'b 'f - -
f -5^2.
,^7
rr
i ._?. 5/^.
.;¦$' 5
'/ —
i
i
i
nV
i
I
S'
.5^
.5 —
. _.. .,/. -
^...*rrr
- \j nice li'l ^ 5 t /<'a' A
fb/o 7~~(Zfx «o4 .-) *4^,2.
6l-7
7///^///' .
-------
•-J7
I, If
t.DO
f, /V
/,M
}¦!'/
2
£(¦ -
& ~
7-
°l "
10 ^
I / —
I
a -
.13-
.../Y~
/,5 "
I'J 7
... .17
1.(7
1,/e
f //
/W
/•/4
/?- //
¥
/>
/-2
//
lb
0
1
7
—)
/
/
o
r
'/• //
/./j"
/./^
i'tf
i,i jp
l,cA
I./1
U€
hi' i
%2o /
V x -*r
~A %,/'
t> o c>
>/,/£
-2'/'
/""/<•/
-------
I'
c>
£
8-) -
J3
fN.
2 2 <2^-)
f/>//y\* Cf*-> 1 ' v z"-3 S
" A C* ^
/' C2
3. 7£
/•?-/
f.*._
A /6
J.'^l _..
. /• / f
/ , I 3 .
/' / 7
/'. U-
l. { &
/1 tl_<6
> w
/'/6
'->7
/¦ is.
/- /? •
X 3- 77...
X -V '
-------
r? rrf 2 3 o / "'"7 ,;''
3-7/'
/
//,2
> 32.3
1
A/A
3
/-/I-
V- "37 3
4
/ ./(>
<5 7 3
6'
A/7.
?< S. o 2
- XtW
.4
A./^
. ^3 3 5" .
7
/3LkO
f ~
/, A*
< f
/7" ^
/'/<£.
.&,/3
6 "
t. ¦J0
. .... .3.-77 .
... V'/y7. . .
../,_/3
11
-7.,
__ cJ -.
/_z./3
'/
-
- —
-------
5 ?"^ .5303
K)FJ~C O
/ ~ A/3,
3 - /'/4
3 ~ /'S3
'/ ~ ( '0>76 '
x _ a
/
j
3
Lt
s
./
. 7_
X
/,//
/./3
/.6$'
/,07
/./& .
-l-£
• -A
S7A. y^_ .^3^) y
3*,.
3- * &
y. / 3t>.
4,%. y f
o (y /
f, //"/
/' 5 //... .
3/.6?
¦V-/VS
y.2£o ...
f 67X _
^.6
•/. 5 7 3 .
-3.., ?f y.._
-------
/J £7~£ & '> ' A
_ -t /21 ft;)-"
5 ;- /- •""• 3-3 & S ¦¦¦"¦< '-•¦*'
/-
-79
' 7f-
-
¦ ?f
1, 07
.7?
¦ ty .
. >'!
_• _ _
. .L?y....
t ./£ .
,7/
A' 3 , 79
K -&>/£
Z.YO
£.$60
A .&/£
JZ , xoO
ZL 3&o
. . &<* 5
*. 5 7?
X. $05
J\ • S-? ->
2.6>l*
/& 7
¦QM>'
-------
m
<$f/t AGRICULTURAL
Z*ll RESEARCH
|