EPMOO/2-77-107d
July 1977
Environmental Protection Technology Series
SOURCE ASSESSMENT:
MECHANICAL HARVESTING OF COTTON
State of the Art
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental Protection
Agency, have been grouped into five series. These five broad categories were established to
facilitate further development and application of environmental technology. Elimination of*
traditional grouping was consciously planned to foster technology transfer and a maximum
interface in related fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION TECHNOLOGY
series. This series describes research performed to develop and demonstrate instrumenta-
tion, equipment, and methodology to repair or prevent environmental degradation from point
and non-point sources of pollution. This work provides the new or. improved technology
required for the control and treatment of pollution sources to meet environmental quality
standards.
EPA REVIEW NOTICE
1 his report has been reviewed by the U.S. Environmental Protection Agency, and approved
for publication. Approval does not signify that the contents necessarily reflect the views and
policy of the Agency, nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
1'his document is available to the public through the National Technical Information Service,
Springfield, Virginia 22161.
-------
EPA-600/2-77-107d
July 1977
SOURCE ASSESSMENT:
MECHANICAL HARVESTING OF COTTON
State of the Art
by
J. W. Snyder and T. R. Blackwood
Monsanto Research Corporation
1515 Nicholas Road
Dayton, Ohio 45407
Contract No. 68-02-1874
ROAP No. 21AXM-071
Program Element No. 1AB015
EPA Project Officer: Dale A. Denny
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
The Industrial Environmental Research Laboratory (IERL) of
EPA has the responsibility for insuring that pollution con-
trol technology is available for stationary sources to meet
the requirements of the Clean Air Act, the Federal Water
Pollution Control Act and solid waste legislation. If control
technology is unavailable, inadequate, uneconomical or social-
ly unacceptable, then financial support is provided for the
development of the needed control techniques for industrial
and extractive process industries. The Chemical Processes
Branch of the Industrial Processes Division of IERL has the
responsibility for investing tax dollars in programs to
develop control technology for a large number (>500) of
operations in the chemical industries.
Monsanto Research Corporation (MRC) has contracted with
EPA to investigate the environmental impact of various indus-
tries which represent sources of pollution in accordance with
EPA's responsibility as outlined above. Dr. Robert C. Binning
serves as MRC Program Manager in this overall program entitled,
"Source Assessment," which includes the investigation of sources
in each of four categories: combustion, organic materials,
inorganic materials, and open sources. Dr. Dale A. Denny of
the Industrial Processes Division at Research Triangle Park
serves as EPA Project Officer. Reports prepared in the Source
Assessment Program are of two types: Source Assessment
Documents, and State of the Art Reports.
111
-------
Source Assessment Documents contain data on emissions from
specific industries. Such data are gathered from the litera-
ture, government agencies and cooperating companies. Sampling
and analysis are also performed by the contractor when the
available information does not adequately characterize the
source emissions. These documents contain all of the infor-
mation necessary for IERL to decide whether a need exists to
develop additional control technology for specific industries.
State of the Art Reports include data on emissions from
specific industries which are also gathered from the litera-
ture, government agencies and cooperating companies. However,
no ext.ensive sampling is conducted by the contractor for such
industries. Sources in this category are considered by EPA
to be of insufficient priority to warrant complete assessment
for control technology decision making. Therefore, results
from such studies are published as State of the Art Reports
for potential utility by the government, industry, and others
having specific needs and interests.
This study was undertaken to provide information on air
emissions from the mechanical harvesting of cotton. In this
project, Mr. D. K. Oestreich served as EPA Project Leader.
IV
-------
CONTENTS
Section
Preface
Figures
Tables
Symbols
I
II
III
IV
V
VI
Introduction
Summary
Source Description
A. Source Definition
B. Process Description
1. Mechanical Cotton Pickers
2. Mechanical Cotton Strippers
3. Mechanical Cotton Gleaners
C. Factors Affecting Emissions
1. Harvesting Machines
2. Trailer Loading
3. Field Transport
D. Geographical Distribution
Emissions
A. Composition and Hazard Potential
1. Major Emissions
2. Minor Emissions
B. Emission Factors and Emission Burdens
C. Definition of Representative Source
D. Source Severities
Control Technology
A. State of the Art
B. Future Considerations
Growth and Nature of the Industry
A. Present Technology
B. Emerging Technology
C. Industry Production Trends
Page
iii
vii
viii
x
1
2
6
6
8
8
13
17
18
18
20
21
21
28
28
28
30
34
39
41
47
47
48
50
50
50
52
v
-------
Section
VII
VIII
IX
X
XI
CONTENTS (Continued)
Unusual Results
A. Seasonal Nature of the Industry
B. Cotton Stripper Emission Rates
C. Relation of Field Size to Severity
Appendixes
A. Comparison of Stripped and Overall
Cotton Yield
B. Sampling and Analysis Procedures
and Results
C. Emission Rate and Emission Factor
Calculations
D. Source Severity Calculations
Glossary of Terms
Conversion Factors and Metric Prefixes
References
Page
53
53
55
55
57
58
60
71
82
95
99
101
VI
-------
FIGURES
Figure Page
1 Mechanical Cotton Picker 10
2 Key Section of Mechanical Cotton Picker 11
3 Two-row Brush-type (Rollers with Brushes) 14
Cotton Strippers
4 Two-row Cotton Stripper with Green Boll 15
Separator Elevator and Tractor-Mounted Basket
5 Four-row Self-propelled Cotton Stripper with 15
Stationary Fingers and Slits on Stripping
Head: (a) Harvesting; (b) Loading Cotton
Trailer
6 Cotton Harvested, 1969 21
7 Representative Field for Calculating Cotton 43
Harvesting Severities
8 U.S. Cotton Area, Yield, and Production; 52
1957-74
9 Usual Start of Cotton Harvest Season 53
B-l Flow Chart for Atmospheric.Stability Class 62
Determination
B-2 Field Data Form 63
VII
-------
TABLES
Table Page
1 Respirable Particulate Emission Factors 3
for Representative Cotton Harvesting
Operations
2 1971-73 Average Cotton Production by State 23
3 Cotton Production by Harvest Method for 24
Cotton-type Farms in 1971
4 1971-73 Average Cotton Area Harvested by 26
Harvest Method
5 State and Weighted Average Population Den- 27
sities for 12 Leading Cotton-Producing States
6 TLV's of Pesticides Applied to Cotton Crops 31
7 Agricultural Chemical Concentrations in 33
Cotton Gin Emissions and Trash
8 Respirable Particulate Emission Factors for 35
Cotton Harvesting Operations
9 Free Silica Emission Factors for Cotton 35
Harvesting
10 Estimated Maximum Pesticide, Defoliant, and 37
Desiccant Emission Factors for Cotton
Harvesting
11 Annual Respirable Particulate Emissions and 38
Emission Burdens from Cotton Harvesting
12 Maximum Severities for Total Suspended 44
Particulates, "Inert" Dust, and Raw Cotton
Dust from Representative Cotton Harvesting
Operations
13 Maximum Severities for Free Silica and 45
Agricultural Chemical Residues from Repre-
sentative Cotton Harvesting Operations
14 Usual Cotton Harvesting Dates, by State 54
A-l Lint Cotton Yield for Primary Stripping 59
Districts in Texas, 1971-73
A-2 Lint Cotton Yield for Cotton-type Farms in 59
Primary Stripping Counties in Oklahoma, 1971
B-l Explanation of Terms on Field Data Form 61
B-2 Continuous Function for Lateral Atmospheric 65
Diffusion Coefficient, a
B-3 Continuous Function for Vertical Atmospheric 65
Diffusion Coefficient, a
' Z
viii
-------
TABLES (Continued)
Table Page
B-4 Mass Emission Rates from Cotton Harvesting 67
Samples - Computer Input and Results
B-5 Composition Analysis Results 69
C-l Mass Emission Rates from Cotton Harvesting 74
Operations, with 95% Confidence Limits
C-2 Respirable Particulate Emission Factors and 80
Supporting Data for Cotton Harvesting
Operations
D-l Tabular Severity Calculations for Total 93
Suspended Particulates and Inert Dust
D-2 Raw Cotton Dust Severities 94
D-3 Free Silica Severities 94
IX
-------
SYMBOLS
Symbol Definition
A Harvested area required to fill one harvester basket
with seed cotton
A Harvested area required to fill one cotton trailer
with seed cotton
A Harvest rate, area harvested per unit time
(CL) _ Confidence limit for two-row strippers with baskets
i I D
(CL) 2 T Confidence limit for two-row strippers pulling
trailers
(CL) i, Confidence limit for four-row strippers
(CL) 95% confidence limit
(CL) Confidence limit for stripping
Q
C Cotton trailer capacity, weight of lint cotton
D Representative distance from source to receptor,
or representative field transport distance
DT (x) Dose of pollutant from basket dumping at distance
x from trailer
D (x) Total dosage in plume at distance x from source
Eu Harvester emission factor
n
ET Trailer loading emission factor
J_i
E Field transport emission factor
E Field transport emission factor based on distance
ETOT Total emission factor
f2 B Fraction of strippers that are two-row models
with baskets
f2 T Fraction of strippers that are two-row models
pulling trailers
fi+ Fraction of strippers that are four-row models
f Fraction of time that receptor is exposed to
" harvester plume in one harvester pass
f_ Fraction of time that receptor is exposed to
transport plume in one transport pass
F Hazard factor
Fe._ Free silica hazard factor
ID Inert dust
L Length of representative field
x
-------
SYMBOLS (Continued)
Symbol
M
mph
n
n
B
"
N
ppm
Q
Q
Qi
Q
QI
Q,
H
TR
r
RCD
S
S1
Si02
SH
SL
SRCD
-i
'Total
STR
TLV
TSP
Definition
Chemical dose lethal to 50% of a population of
test animals
Diffusion model designated as 1, 2, or 3 for point,
line, or dose model, respectively
Miles per hour
Number of samples
Number of harvesting and basket dumping cycles com-
pleted in 1 day
Number of harvester baskets required to fill one
cotton trailer
Number of harvester passes required to fill one
harvester basket
Number of cotton trailers filled in 1 day
Number in total population
Parts per million
Mass emission rate, mass per unit time
Average emission rate
.th
sample
Emission rate from the i
Harvester emission rate
Emission mass from dumping one harvester basket
Field transport emission rate
Number of rows
Raw cotton dust
Source severity of pollutant
Atmospheric stability class designated as A, B, C,
D, E, or F
Silicon dioxide; free silica
Severity for harvester
Severity for trailer loading
Raw cotton dust severity
Total source severity
Severity for field transport
Threshold limit value
Total suspended particulate
XI
-------
SYMBOLS (Continued)
Symbol Definition
± Values of "Student t" distribution between which
95% of the area lies
tj Time required to fill and dump one harvester basket
t Time of receptor exposure to pollutant puff
t_, Time consumed in dumping one harvester basket
D
t,. Time base of hazard factor
r
tT Time required to harvest one length of representative
11 field
to Daily operating time
(t0)R Total basket dumping time in 1 day
(to)f[ Total harvester operating time in 1 day
(to)T Total time devoted to field transport in 1 day
t Time for one harvester pass
t Sampling time
5
t. Time for harvester to turn after each pass
u Mean wind speed
VD Mean harvester speed
n
v Mean cotton trailer field transport speed
w Width of plume
P F
w Row spacing
w Swath width
s
x Distance downwind from source in the direction of
the mean wind
y Crosswind distance
Y Lint cotton yield, mass per unit harvested area
A Difference between converted concentration and
background concentration
a Estimated population standard deviation
a Standard deviation in the crosswind direction of
^ the plume concentration distribution
a Standard deviation in the vertical direction of the
plume concentration distribution
X Pollutant concentration, mass per unit volume
xii
-------
SYMBOLS (Continued)
Symbol
X(x)
x(x,y)
XA(X)
Xe(D)
Xe(x)
max
Xp(D)
Xp(x)
(xmax) o
(7 )
Amax TR
Definition
Pollutant concentration in plume at distance x
from source
Pollutant concentration located at point (x,y) in
plume
Crosswind integrated plume concentration at distance
x from source
Average maximum receptor concentration during
basket dumping
Concentration during exposure to pollutant puff at
distance D from source
Average concentration at receptor during exposure
to pollutant puff
Maximum receptor concentration during exposure to
instantaneous puff
Concentration in plume at distance D from source
Average ground level concentration in plume at
distance x from source
Maximum ground level concentration in the plume at
at edge of representative field
Time-averaged maximum pollutant concentration at
edge of representative source
Time-averaged maximum receptor concentration from
basket dumping
Time-averaged maximum receptor concentration from
harvester emissions
Maximum average receptor concentration during source
operation
Time-averaged maximum receptor concentration from
field transport
Xlll
-------
SECTION I
INTRODUCTION
Cotton harvesting includes the removal and collection of
seed cotton (fibers with seeds) from mature plants and the
transport of this cotton from the fields. Machines now
account for more than 99% (by weight) of all cotton harvested,
and machine and field transport activities cause air pollu-
tion in the form of respirable dust.
The objective of this work was to assess the environmental
impact of mechanical cotton harvesting activities and to
produce a State of the Art Report summarizing available data
on air emissions from this source. This document was pre-
pared by acquiring and analyzing information on: (1) the
cotton harvesting process and equipment; (2) source locations
and distribution; (3) mass emissions, state and nationwide;
(4) effects on air quality; (5) air pollution control tech-
nology; and (6) projected growth and anticipated techno-
logical developments of the industry.
Emission information was developed from a limited sampling
program involving cotton harvesting operations at three
farms characteristic of the industry. Resulting emission
rates and factors were used to estimate air quality impact
and state and nationwide emissions.
-------
SECTION II
SUMMARY
The 1971-73 average annual production of lint cotton was
2.70 x 106 metric tons (2.97 x 106 tons) harvested from
49,200 km2 (12.1 x 106 acres) in 19 states, predominantly
south of the 36th parallel. The four leading states were:
Texas, which harvested 31% of the national production from
41% of the cotton area harvested; Mississippi, which
harvested 15% of the cotton from 12% of the area harvested;
California, which harvested 12% from 7% of the area; and
Arkansas, which produced 10% from 10% of the area.
More than 99% of both national production and cotton area
were harvested mechanically. The two principal harvest
methods; were machine picking, with 70% of the harvest from
61% of the area, and machine stripping, with 29% of the harvest
from 39% of the area. Picking is practiced throughout the
cotton regions of the U.S., while stripping is practiced
chiefly in the dry plains of Texas and Oklahoma. Mechanical
gleaning, a cleanup operation after picking, was not
Lint cotton is cotton that has been cleaned and had the seeds
removed; it is the product of cotton gins. Production figures
are usually presented by weight of lint cotton rather than
seed cotton, which is raw cotton before ginning.
1 metric ton = 106 grams = 2,205 pounds = 1.1 short tons
(short tons are designated "tons" in this document); other
conversion factors and metric system prefixes are presented
in Section X.
-------
considered in this study, since it accounts for less than
1% of the national production and is similar to stripping.
Three unit operations are involved in mechanical harvesting
of cotton: harvesting, trailer loading (basket dumping), and
transport of trailers in the field. Respirable particulate
(<7 ym in diameter) emission factors for these operations
in representative picking and stripping are presented in
Table 1. All of the emissions from harvesting and trailer
loading in this table are raw cotton dust, which is associated
with byssinosis. Raw cotton dust is predominantly plant
fragments but it also contains some free silica due to soil
dust. Emissions from field transport are all soil dust.
Free silica accounts for 7.9% by weight of the total emis-
sions from picking operations; the figure for stripping
operations is 2.3%. Maximum pesticides content in emissions
from cotton harvesting operations, mainly consisting of
organochlorines such as Endrin, is estimated as 220 ppm,
maximum defoliant content is 17 ppm of DBF (tributylphosphoro-
trithioate) for picking, and maximum desiccant content is
2,200 ppm of arsenic for stripping.
Table 1. RESPIRABLE PARTICULATE EMISSION FACTORS FOR REPRESENTATIVE
COTTON HARVESTING OPERATIONS
kg emitted/km2 harvested (Ib emitted/1,000 acres harvested)
Type of harvester
Mechanical
Mechanical
picker
stripper
Unit operation
a
Harvesting
0.455 ± 0.738
(4.06 ± 6.58)
2.30 ± 0.82
(20.5) ± 7.32)
Trailer.
loading
0.0699
(0.624)
0.0918
(0.819)
a
Transport
0.427 ± 0.119
(3.81 ± 1.06)
0.279 ± 0.078
(2.49 ± 0.70)
Total
0.952
(8.49)
2.67
(23.8)
Confidence limits at the 95% confidence level.
Insufficient data to assign confidence limits.
-------
Cotton harvesting operations accounted for 0.002% of the
national respirable particulate emissions in 1972. The
highest state contributions were 0.046% in Texas and 0.025%
in Oklahoma. Both are predominantly stripping states, with
95% of total cotton harvesting emissions caused by stripping.
The air quality impact of cotton harvesting emissions was
determined by the maximum severity from representative
picking and stripping sources. Source severity is defined
as the ratio of the time-averaged maximum ground-level
pollutant concentration to the pollutant hazard factor.
The hazard factor for total suspended particulate is the
24-hour primary national ambient air quality standard; for
other cotton harvesting pollutants it is the threshold limit
value (TLV®) divided by a safety factor of 100.
The representative cotton farm is defined as harvesting
0.786 km2 (194 acres) from one square field, with the down-
wind side subject to public exposure. Solid planting (no
rows shipped) with a row spacing of 1.02 m (40 in.) is the
representative row pattern, and cotton is harvested with one
mechanical harvester 8 hr/day during the harvest season,
which lasts an average of 10 weeks in any specific area.
The operating parameters for picking and stripping are
different, so the two harvest methods were treated separately.
The representative yield from mechanical picking is 63.0 metric
tons of lint cotton per km2 harvested (562 Ib/acre) with chemi-
cal defoliation practiced. The representative picker harvests
two cotton rows simultaneously at 1.34 m/s (3.0 mph) and has
a harvester-mounted basket for collecting the picked cotton.
Representative stripper yield is 41.2 metric tons of lint cot-
ton per km2 harvested (368 Ib/acre) with chemical desiccation
-------
practiced. The representative stripper harvests two rows
of cotton simultaneously at 2.23 m/s (5.0 mph) and has a
harvester-mounted cotton collection basket. Six stripper or
picker baskets are required to fill a cotton trailer with
654 kg (1,440 Ib or 3 bales9) of lint cotton. For both
picking and stripping, filled cotton trailers are trans-
ported from the field at a speed of 4.47 m/s (10.0 mph).
The highest source severity is for raw cotton dust, at
0.00703 for picking operations and 0.0350 for stripping
operations. Severities for other pollutants are less than
0.001.
No emission control technology is applied to cotton harvesting
operations, and no voluntary future application is expected.
Average annual cotton production in the U.S. is not expected
to increase more than a few percent by 1978. Average annual
harvested area is expected to remain constant. Future cotton
production will be influenced by the price of petroleum, used
to make synthetic fibers. The trend is to fewer but larger
cotton farms, resulting in increased use of multiple
harvesters with higher harvest rates.
al bale = 480 Ib (net weight) of lint cotton.
-------
SECTION III
SOURCE DESCRIPTION
A. SOURCE DEFINITION
i
Cotton, the principal fiber crop of the U.S., is best
known for the thread and cloth made from its fibers. By-
products from cotton, however, are a large portion of its
crop value. The seeds are pressed for their oil which is
used in vegetable oils, margarine, lubricants, soaps, and
paints. The remaining cottonseed cake is ground for high
protein cattle feed or fertilizer. Cottonseed hulls are
used for low grade cattle roughage, paper, fiberboard and
fertilizer.1 Cotton linters, the short fibers left on the
seeds after ginning, are used for cellulose in making
rayon.2
More than 99% of all cotton grown in the U.S. is of the
American upland varieties.3 Most of the remainder is
American-Pima (formerly called American-Egyptian) cotton.l
M. S. Production of Field Crops, 6th Edition.
New York, McGraw-Hill Book Company, 1970. p. 447-483.
2Linton, G. E. Natural and Manmade Textile Fibers. New
York, Duell, Sloan and Pearce, 1966. p. 208-219.
Agricultural Statistics, 1973. Washington, U.S. Depart-
ment of Agriculture, Yearbook Statistical Committee (U.S.
Government Printing Office Stock Number 0100-02841).
1973. p. 58-75.
-------
Upland cotton has a staple (fiber) length of 19 mm to 38 mm,
and typical lint (ginned) cotton yield is about 53 metric
tons of lint cotton/km2 harvested although irrigated areas
of Arizona and California may yield more than twice that
amount.3'11 American-Pima cotton is grown in irrigated areas
of the Southwest, has very strong fibers from 38 mm to 44 mm
long,1 and yields about 52 metric tons of lint cotton/km2
harvested.3
Use of machines to harvest cotton has increased rapidly in
the last two decades, stimulated by increasingly scarce
and expensive agricultural labor. An average worker can
hand pick roughly 70 kg to 90 kg of seed cotton a day if
the yield is good.1 Since about 75% of seed cotton weight
is the seed, it requires 2,300 to 3,100 man-days to hand
pick 1 km2 of cotton with an average lint yield of 53 metric
tons/km2. By contrast, a single harvesting machine can
harvest the same area in 4 to 20 days. Production of
mechanical harvesters was limited before 1950; by 1953,
22% of all cotton in the U.S. was machine harvested; in 1963,
the figure rose to 72%.5 Hand labor is now used mainly at
the start of the harvest season or on farms with a few
thousand square meters or less of cotton.6
4Crop Production, 1974 Annual Summary. U.S. Department of
Agriculture, Statistical Reporting Service, Crop Reporting
Board. Washington. Publication No. CrPr 2-1(75).
January 16, 1975. 64 p.
5Colwick, R. F., et al. Mechanized Harvesting of Cotton.
U.S. Department of Agriculture, Agricultural Research Service.
Beltsville. Southern Cooperative Series, Bulletin No. 100.
March 1965. 70 p.
6Voelkel, K. E. Texas Cotton Review, 1973-74. The University
of Texas at Austin, Natural Fibers Economic Research. Austin.
Research Report No. 104. July 1974. 143 p.
-------
In 1969, 137,000 UoS. farms harvested 2.18 x 106 metric
tons of lint cotton from 44,800 km2. Although cotton-type
farms (farms with cotton sales of at least $10,000 or 50% of
the total value of all farm products sold) accounted for 29%
of all farms growing cotton, they harvested 52% (by weight)
of all cotton on 48% of all land devoted to cotton.7 Mech-
anical harvesters accounted for 99% of all cotton harvested
from cotton-type farms in 1971: mechanical pickers har-
vested 79%; mechanical strippers, 19%; and mechanical
gleaners or ground harvesters, 1%.8
Emissions in the form of respirable dust are generated by
mechanical cotton harvesting. As discussed in this document,
cotton harvesting refers to the removal of cotton from plants,
the collection of this cotton, and its transport from the
fields. Differences in the characteristics of machine picking,
machine stripping, and machine gleaning make it necessary to
consider each method as a source subtype in assessing emis-
sions from mechanical harvesting of cotton.
B. PROCESS DESCRIPTION
1. Mechanical Cotton Pickers
Mechanical cotton pickers, as their name implies, selectively
pick locks of seed cotton from open cotton bolls, and leave
7Census of Agriculture, 1969. Volume II, General Report.
Chapter 8, Type of Farm. Washington, U.S. Bureau of the
Census, 1973. 287 p.
8Census of Agriculture, 1969. Volume V, Special Reports.
Part 3, Cotton. Washington, U.S. Bureau of the Census,
1973. 184 p.
8
-------
the empty burs and unopened bolls on the plant. Typical
modern pickers are self-propelled and require only one
operator to harvest two rows of cotton simultaneously at
a speed of 1.1 m/s to 1.6 m/s. A mechanical cotton picker
is shown in Figure 1.
The key section of a picker is presented in Figure 2.9 Gear-
driven rotating spindles are mounted in vertical rows on cam-
oriented picking bars which make the spindles enter and leave
the plants at right angles to the rows. The spindles are
tapered and barbed, straight and toothed, or fluted so that
the seed cotton locks wrap around them as they pass through
the plants. The doffers pull the cotton off the spindles by
evenly spaced discs with uneven surfaces mounted on rotating
vertical cylinders. Blower-forced air moves the cotton from
the doffers into the pneumatic conveyor ducts which carry it
into the picker-mounted basket. The doffed spindles are
cleaned of sticky residues by passing over water-moistened
pads before entering the cotton row again.9
When the picker basket is filled with seed cotton, the
machine is driven to a cotton trailer at the edge of the
field. A cotton trailer is simply a flatbed wagon with the
sides and ends enclosed by slats or wire screen. The basket
is raised and tilted hydraulically, the top swings open,
and the cotton falls into the trailer. To make maximum use
of the trailer volume, the cotton is spread out and sometimes
compacted by tramping. When the trailer is full it is pulled
from the field, usually by pickup truck, and taken to a
cotton gin.
9Kelly, C. F. Mechanical Harvesting. Scientific American,
217(2):50-59, August 1967.
-------
Figure 1. Mechanical cotton picker
Courtesy of International Harvester Company.
10
-------
Figure 2. Key section of mechanical cotton picker
From "Mechanical Harvesting" by Clarence F. Kelly. Copyright^
August 1967 by Scientific American, Inc. All rights reserved.
11
-------
To maximize harvesting efficiency and minimize cotton trash,
leaf stain, and moisture content, picker-type cotton is
usually defoliated before machine harvesting. Chemical de-
foliants are applied by airplane or ground rig when 60% or
more of the bolls are open.5'10 The defoliants cause an
abscission layer to develop on the leaf stem where it is
attached to the branch, and the leaves subsequently become
detached and fall.11 With the proper defoliant application
and favorable weather conditions, the cotton is ready for
machine picking about 2 weeks later. This time span allows
the leaves to fall and defoliant residues to be reduced to
allowable limits set by the Department of Agriculture,10
and normally precludes leaf regrowth problems. A second
application is sometimes used if heavy rain follows the first
application, if the cotton is especially tall or has lush
foliage, or if excessive leaf regrowth occurs.
In a good season cotton is often machine picked a second
time, about 2 weeks after the first, which can add 50% to
the yield.12 In this case the defoliant dose must be low
enough to prevent killing the plants or damaging the green
and unopened bolls. The process of defoliant application
and emissions thereof are discussed in another document.13
10Elliott, F. C. Cotton Defoliation Guide for Texas. Texas
A&M University, Texas Agricultural Extension Service.
College Station. Bulletin No. L-145. 1969.
^Addicott, F. T., and R. S. Lynch. Defoliation and Desic-
cation: Harvest-Aid Practices. In: Advances in Agronomy,
Volume IX. New York, Academic Press Inc., 1957. p. 69-93.
12Personal communication. Dr. R. B. Metzer, Texas Agri-
cultural Extension Service, Texas A&M University, College
Station. October 24, 1975.
13Peters, J. A., and T. R. Blackwood. Source Assessment:
Defoliation of Cotton, State of the Art. Monsanto Research
Corporation, EPA Contract 68-02-1874. Dayton. Preliminary
document submitted to the EPA, February 1976. 124 p.
12
-------
2. Mechanical Cotton Strippers
Mechanical cotton strippers remove opened and unopened bolls
along with the burs, leaves, and stems from cotton plants
and leave only bare branches. Stripping is necessarily a
once over operation although strippers can be used as the
final harvest operation after machine picking. The principal
advantage of mechanical stripping is economy since the
typical cost to own and operate mechanical strippers is half
that for cotton pickers.5 Strippers also harvest faster
with lower field losses. They are tractor-mounted, tractor-
pulled, or self-propelled; require one operator; strip
plants with pairs of rotating rolls, rolls with stationary
picking bars, or stationary fingers and slits; and harvest
from one to four rows of cotton at speeds of 1.8 m/s to
2.7 m/s. Some common types of strippers are shown in
Figures 3, 4, and 5.
When stripping rolls are used, their axes are parallel to
the rows and tilted at approximately 0.52 rad (30°) to the
horizontal with the front end lower. The rolls are smooth
or machine-roughened, or equipped with fingers or longitudinal
strips, and are made of steel, rubber, or bristle-brush
material. The gap between pairs of rolls allows cotton plants,
but not bolls, to pass through. Bolls are torn from the
plants by the rollers. The arrangement for a single stripper
roll is similar, except that the gap is between the roll and
a stationary stripping bar.5 When stationary fingers are
used they are mounted in the direction of the rows on strip-
ping heads. The assembly looks much like a comb. As the
machine moves forward, the cotton plants follow the slits
between fingers which are spaced so that they strip the plant
branches.
13
-------
Figure 3. Two-row brush-type (rollers with brushes)
cotton strippers5
14
-------
Figure 4. Two-row cotton stripper with green boll
separator elevator and tractor-mounted basket5
(a)
(b)
Figure 5. Four-row self-propelled cotton stripper
with stationary fingers and slits on stripping head
(a) harvesting; (b) loading cotton trailer
15
-------
After the cotton is stripped, it enters a conveying system
which carries it from the stripping unit to an elevator.
Most conveyors utilize either augers or series of rotating
spiked-tooth cylinders and also accomplish some cleaning by
moving the cotton over perforated, slotted, or wire mesh
screens. Dry plant material (burs, stems, and leaves) is
crushed and falls through the openings to the ground. Blown
air is sometimes used to assist cleaning.5 Many strippers
are also equipped to remove the burs from the seed cotton.
The elevator is a belt or chain with cross-flights, a
duct carrying blower-forced air, or both. Most pneumatic
elevators are also designed as green boll separators. Air
carries the fluffy seed cotton up the elevator and through
a spout into a pulled cotton trailer or harvester-mounted
basket. The heavier unopened green bolls fall or roll by
gravity into a collection bin. They are dumped on piles
at the edges of the field, since many will later open and
can be salvaged.5
Since mechanical strippers are designed to leave only bare
plant branches, efforts are made to minimize trash and
moisture content and maximize cotton grade and harvesting
efficiency by reducing plant foliage. The easiest and
cheapest method is to delay harvesting until after the first
frost or freeze in the fall; frost causes leaves to defoliate,
and a freeze desiccates (kills and dries) them. However,
this method can be practiced only in areas where the freeze
occurs soon after cotton maturity, since wind, rain, long
standing time, or a combination of these will result in
excessive crop loss and grade reduction. Another method is
chemical defoliation, as discussed for picker-type cotton.
The mcst widely practiced method of harvest-aid treatment
prior to stripper harvesting is chemical desiccation.
16
-------
Chemicals which stop plant growth and kill and dry the
foliage are applied when 75% to 90% of the bolls are open.
Desiccant application and associated emissions are discussed
in another document.13 Some leaves drop but most stay on
the plants and dry out, an effect similar to that of natural
freezing.5'10'11 The cotton is often dry enough to harvest
in about a week, but 2 weeks are often needed for desiccant
residues to drop to allowable levels.10'13 Some of the dry,
brittle leaves are shaken from the plants during stripping
arid some leaves and burs are removed from the cotton by the
strippers. Much trash remains with the seed cotton, but
gins in stripping areas are equipped to adequately remove
dry leaves, burs, and stems.
3. Mechanical Cotton Gleaners
Gleaners are used to salvage cotton left in the fields
after conventional harvesting, usually machine picking. They
are used primarily in arid cotton-growing areas where dry
ground and air prevent serious lint damage by rotting. Some
gleaners are independent machines, but there are also glean-
ing attachments for pickers. The most common gleaner designs
use notched belts to collect the cotton and include equip-
ment for partial cleaning. Gleaners harvest about 13,000
m2/hr at ground speeds of 1.3 m/s to 2.2 m/s. Approximately
„ fr>tVR.i'ctor*
0.15 km^ will yield 1 "kg of lint cotton at gleaning effi-
ciencies near 50%.5
Mechanical cotton gleaners were not specifically studied
for this report, due to their minor role in cotton produc-
tion. Less than 1% of the U.S. cotton crop is harvested
with gleaning machines.8 Since their principles of operation
are similar, gleaners and strippers are assumed to have the
same emission factors in calculating total emissions from
cotton harvesting.
17
-------
C. FACTORS AFFECTING EMISSIONS
1. Harvesting Machines
The ma.jor factors affecting emissions from mechanical cotton
harvesters are: (1) type of harvester (picker or stripper)
and design; (2) type (variety) of cotton; (3) preharvest
treatment and condition of crop; (4) harvest rate; and (5)
extent of machine cleaning (trash removal). Each of these
interrelated characteristics varies widely; however, by
categorizing the process by harvester type, the ranges of
variables are reduced.
A quantitative study of the factors affecting emissions
was not attempted. No data have been published and a
massive sampling project would be required to establish the
quantitative functional dependence of emissions on the
affecting variables. The following qualitative discussion
offers some insight into how the factors influence composi-
tion and rate of emissions.
Emissions from strippers and gleaners are higher than those
from pickers because more cotton trash is handled. Machine-
picked cotton averages about 0.2 kg of trash (hulls, sticks,
stems, leaves, and dirt) excluding seed weight, per kg of
lint cotton. By contrast, machine-stripped cotton averages
1.1 kg/kg, and machine-scrapped (gleaned) cotton averages
1.8 kg/kg. 1 ** Some trash is crushed into fine particles and
emitted by harvester conveying and cleaning systems.
ltControl and Disposal of Cotton-Ginning Wastes. U.S.
Public Health Service, National Air Pollution Control
Administration. Raleigh. Publication No. 999-AP-31.
1967. 103 p.
18
-------
Stripper cotton varieties generally produce smaller, hardier
plants than picker cotton. Stripper plants, usually less
than 0.9 m high, were developed for varying degrees of
storm resistance, depending on the areas for which they were
intended, so that cotton is not lost before harvest.
Picker cotton varieties are generally more than 0.9 m high;
they have lusher foliage and wide-opening bolls to facilitate
picking, and give higher yields than stripper cotton.
Varieties and harvesting methods are usually matched. How-
ever, strippers are sometimes used to harvest common picker
varieties, especially for final harvest, resulting in higher
emissions, particularly if the preharvest treatment is
desiccation rather than defoliation.
The condition of the crop depends on several factors, in-
cluding cultural practices, fertilization, precipitation and/
or irrigation, weather during growing season, pest damage
and control, preharvest treatment, and harvest timing. All
affect the yield and trash content, and hence the quantity
of emissions, of machine-harvested cotton. Pest control
and preharvest treatment also affect the composition of
emissions. Thousands of pesticides are registered with EPA
for use on cotton to control hundreds of different insects,
worms, fungi, weeds, and other pests.15 Pesticide application
ranges from none to once every 3 or 4 days with amounts per
application varying by as much as an order of magnitude.
Residues on the plants during harvest depend on the chemical,
time and rate of application, and weather (particularly
precipitation) prior to harvest. Chemical defoliants, desic-
cants, and regrowth inhibitors used in varying amounts also
leave residues on the cotton.
15Personal communication. Elgin G. Fry, Office of Pesti-
cides Programs, U.S. Environmental Protection Agency,
Washington. October 29, 197.5.
19
-------
In general, increased harvest rate increases emission rate
due to the rate of material handling. For instance, a two-row
stripper harvesting at the same speed as a one-row stripper
of the same design causes a higher emission rate. Design dif-
ferences can override harvest rates in affecting emissions.
Harvester speeds are usually too slow for entrainment of soil
dust unless the soil is especially dry or windspeed is high.
Seed cotton trash removal is the prime cause of dust emis-
sions from cotton harvesters. The purpose of harvester
cleaning systems is to leave trash in the fields so cleaner
cotton can be taken to the gin, resulting in higher income
to the farmer through lower ginning fees and higher cotton
grade. Fine particles in the trash become airborne when
exhausted from the harvesters, the major source being
pneumatic elevators. One of the reasons for using blower-
forced air to carry cotton to wire-screened or slatted
cages or trailers is to gain trash separation.16 Fine
trash remains airborne while the seed cotton is caught by
the screen. Higher wind speed enhances dust entrainment,
especially when the harvester is traveling into the wind.
2. Trailer Loading
When harvester-mounted baskets are used, the cotton is
dumped into transport vehicles, usually trailers (wagons),
for hauling to gins. The basket is hydraulically raised
and tilted, allowing the cotton to fall by gravity into the
trailer. Air movement from wind and the falling cotton
causes a puff of dust composed of soil and trash. Wind
speed, amount and type of trash in the cotton, height of
fall, and type of trailer affect the amount of dust generated.
16Elliott, F. C. Keep Cotton...Dry-Loose-Clean. Texas A&M
University, Texas Agricultural Extension Service. College
Station. Publication No. MP-297. 8 p.
20
-------
3. Field Transport
Emissions from field transport of cotton depend on vehicle
types and speed, type of surface traveled, and surface
moisture. Field "road" surfaces range from grass, to dirt
tracks in grass from repeated travel, to bare soil. For
given transport conditions, cotton yield and trash content
determine the number of trailers pulled to and from the
field and hence the total emissions generated in trans-
porting the harvested cotton.
D.
GEOGRAPHICAL DISTRIBUTION
Virtually all cotton in the United States is grown south
of the 36th parallel17 where 19 states produce cotton. The
distribution of cotton harvesting areas is shown in Figure 6.18
UNITED STATES TOTAL
46,526km2
\( 11,496,320 ACRES)
MACHINE STRIPPED AREA'^s4 (5.000ACRES) X
Figure 6. Cotton harvested, 196918
17Burkhead, C. E., R. C. Max, R. B. Karnes, and E. Reid.
Field and Seed Crops - Usual Planting and Harvesting Dates
by States in Principal Producing Areas. U.S. Department
of Agriculture, Statistical Reporting Service. Washington,
Agriculture Handbook No. 283. 1972. p. 10-12.
18Census of Agriculture, 1969. Volume V, Special Reports.
Part 15, Graphic Summary. Washington, U.S. Bureau of the
Census, 1973. p. 125.
21
-------
The principal cotton-producing areas are apparent from this
map: the plains of western Texas, the Blacklands of eastern
Texas,, the Mississippi Valley, and the San Joaquin Valley of
California. The region encompassing the plains of eastern
New Mexico, western Texas, and southwestern Oklahoma and
the Texas Blacklands is primarily stripper harvested while
cotton in the rest of the country is machine picked.
Mechanical gleaning is practiced mainly in Arizona, but
represents only 10% by weight of the state harvest and 1%
of the national harvest.
Table 2 presents average cotton production figures for the
1971-1973 harvest seasons by state. Averages for the 1971-
1973 seasons, rather than the 1972 season alone, are shown
so that localized effects (mostly due to weather) during a
particular season are deemphasized. Texas is the leading
cotto;i state, accounting for 31% of U.S. total weight and
41% of: U.S. total harvested area. The top 12 states harvest
97% of the national total of 2.70 x 106 metric tons of lint
cotton from 97% of the total 49,173 km2 harvested.
Table 3 shows the breakdown of cotton harvesting methods by
state and the contribution of each state to the total pro-
duction by each harvest method based on cotton-type farms
in 1971. Machine picking is the predominant harvest method
in the U.S., accounting for 70% of the national production
and greater than 97% of the production in 9 of the 12 leading
cotton states. Cotton stripping accounts for 29% of the
national harvest, and predominates only in Texas and Okla-
homa. Mechanical gleaning is important only in Arizona, and
accounts for less than 1% of the national harvest.
22
-------
Table 2. 1971-73 AVERAGE COTTON PRODUCTION BY STATE3'4
Produc-
tion
rank
1
2
3
4
5
6
7
8
9
10
11
12
State
Texas
Mississippi
California
Arkansas
Louisiana
Arizona
Alabama
Tennessee
Georgia
Missouri
Oklahoma
South Carolina
b
All other
U.S. TOTAL
Lint cotton
production3
103 metric tons
842
401
336
270
133
132
120
110
81
74
68
63
69
2,699
Percent
of U.S.
production
31.2
14.9
12.5
10.0
4.9
4.9
4.4
4.1
3.0
2.7
2.5
2.3
2.6
100
Area
harvested,
km2
20,250
5,767
3,438
4,759
2,276
1,223
2,223
1,823
1,608
1,203
1,932
1,288
1,383
49,173
Percent of
U.S. area
harvested
41.2
11.7
7.0
9.7
4.6
2.5
4.5
3.7
3.3
2.5-
3.9
2.6
2.8
100
Yield,
metric tons/km2
41.6
69.5
97.7
56.7
58.4
107.9
54.0
60.3
50.4
61.5
35.2
48.9
49.9
54.9
to
U)
Cotton after cleaning and seed removal by ginning.
All other cotton producing states including North Carolina, New Mexico, Florida, Kentucky,
Virginia, Nevada, and Illinois.
-------
Table 3. COTTON PRODUCTION BY HARVEST METHOD FOR COTTON-TYPE FARMS IN 19718
Produc-
tion
rank
1
2
3
4
5
6
7
8
9
10
11
12
State
b
Texas
Mississippi
California
Arkansas
Louisiana
Arizona
Alabama
Tenessee
Georgia
Missouri
Oklahoma
South Carolina
All other0
U.S. TOTAL
Percent of state production
Machine
picked
15
99
100
99
99
89
99
97
98
99
16
97
92
_d
Machine
stripped
85
1
1
84
5
_d
Machine
gleaned
10
2
_d
Hand
labor
1
1
1
2
1
1
2
1
_d
Percent of U.S. total production
Machine
picked
5
15
12
10
5
4
4
4
3
3
3
2
70
Machine
stripped
27
2
29
Machine
gleaned
i
1
Hand
labor
As defined by the Specialized Survey of Cotton Operations, 1971, cotton-type farms are those which had:
(a) $10,000 or more in cotton sales; or (b) 50% or more of total farm sales from cotton, excluding farms
with total sales under $2,500.8
Based on Texas Agricultural Extension Service data.6
C . . -
All. other cotton producing states including North Carolina, New Mexico, Florida, Kentucky,
Virginia, Nevada, and Illinois.
d
Not applicable.
Note: Blanks indicate <0.5%.
-------
No data were found which show explicitly the breakdown on
area harvested by each method. However, reasonable esti-
mates can be inferred from the production data of Table 3.
Machine-picked and hand-harvested areas should have approxi-
mately the same harvest yields (mass per unit area harvested).
Machine gleaning is a cleanup operation performed after
machine picking; hence gleaned area is included in machine-
picked area and does not cause a split in total harvested
area. Therefore, only machine stripping will cause noticeable
differences between proportions of total production and total
harvested area. Examination of cotton production data for
Texas6/19 and Oklahoma8 shows differences of less than 1
metric ton/km2 between machine-stripped and overall yield for
each state (see Appendix A). Therefore, it is reasonable to
assume that yields for all harvest methods, except gleaning,
within a state are equal, and that the proportion of area
harvested by each method is the same as the proportion of
production. For gleaning, a reasonable estimate is obtained
by applying the fraction of picker farms which also glean to
the total machine-picked area in each state.
Table 4 shows the resulting estimated average 1971-73
cotton area harvested by each method for the 12 leading cotton
states and the total U.S. Of the total U.S. area harvested,
61% is machine picked, 39% is machine stripped, and less
than 1% is harvested by hand labor; 4% of the machine-picked
area is also machine gleaned.
19Caudill, C. E., P. M. Williamson, M. D. Humphrey, Jr.,
L. P. Garrett, and L. Canion. 1973 Texas Cotton Statis-
tics. Texas Crop and Livestock Reporting Service, Texas
Department of Agriculture. Austin. Bulletin 113.
June 1974. 21 p.
25
-------
Table 4. 1971-73 AVERAGE COTTON AREA HARVESTED
BY HARVEST METHOD
(km2)
Produc-
tion
rank
1
2
3
4
5
6
7
8
9
10
11
12
State
Texas
Mississippi
California
Arkansas
Louisiana
Arizona
Alabama
Tennessee
Georgia
Missouri
Oklahoma
South Carolina
All otherb
U.S. TOTAL0
Machine
picked
3,000
5,700
3,400
4,700
2,300
1,200
2,200
1,800
1,600
1,200
300
1,200
1,300
30,000
Machine
stripped
17,200
1,600
100
19,000
Hand
labor
100
i
200
Machine
gleaned3
200
100
700
100
1,200
Based on proportion of machine-picked farms which also
glean cotton.8
All other cotton producing states including North Carolina,
New Mexico, Florida, Kentucky, Virginia, Nevada, and Illinois
r
Data may not add to totals shown because of independent
rounding.
Note: Blanks indicate <50 km2.
26
-------
Table 5 shows the average population density for the 12
leading cotton states, based on the 1970 Census.20 To find
the overall average population density for cotton harvesting
the state densities were weighted according to the fraction
of total harvested cotton area within the respective states.
The resulting weighted average of 22 persons/km2 is the
representative population density for cotton harvesting
states in the U.S.
Table 5. STATE AND WEIGHTED AVERAGE POPULATION DENSITIES
FOR 12 LEADING COTTON-PRODUCING STATES
State
Texas
Mississippi
California
Arkansas
Louisiana
Arizona
Alabama
Tennessee
Georgia
Missouri
Oklahoma
South Carolina
TOTAL
1970
population
density, 20
persons/km2
16.5
18.1
49.3
14.3
31.3
6.0
26.2
36.6
30.5
26.2
14.4
33.1
Percent of
total U.S.
cotton area
harvested3
41.2
11.7
7.0
9.7
4.6
2.5
4.5
3.7
3.3
2.5
3.9
2.6
97.2
Weighted
population
density, b
persons/km2
6.99
2.18
3.55
1.43
1.48
0.15
1.21
1.39
1.04
0.67
0.58
0.89
21.6
1971-73 average.
For example, Texas weighted density = Q-'? x 16.5 = 6.99,
201970 Census & Areas of Counties and States. In: 1975
World Almanac & Book of Facts. New York, Newspaper Enter-
prise Association, Inc., November 1974. p. 183-201.
27
-------
SECTION IV
EMISSIONS
A. COMPOSITION AND HAZARD POTENTIAL
1. Major Emissions
Emissions from mechanical cotton harvesting operations are
in the; form of solid particulates (dust) composed of cotton
plant fragments and soil dust. Respirable particulate
(<7-ym mean aerodynamic diameter) emissions are of particular
interesst, because it is this fraction of total particulates
that has the largest potential effect on human health. The
respirable particulates are composed mainly of raw cotton
dust and soil dust, which contains free silica.
Particulate matter is an EPA criteria pollutant, so national
air quality standards have been established for it. The
24-hour primary standard for total suspended particulates
(TSP) is 260 yg/m3.21 In addition, the American Conference
of Governmental Industrial Hygienists (ACGIH) has established
a TLV of 10 mg/m3 for inert (less than 1% by weight free
21Code of Federal Regulations, Title 42 - Public Health,
Chapter IV - Environmental Protection Agency, Part 410
National Primary and Secondary Ambient Air Quality
Standards, April 28, 1971. 16 p.
28
-------
silica) dusts.22 TLV's refer to "time-weighted concentra-
tions for a 7- or 8-hour workday and 40-hour workweek"
recommended by ACGIH to protect worker health.
All of the particulate matter emitted by harvesting machines
and trailer loading operations is considered to be raw
cotton dust (RCD). Prolonged exposure to RCD has been iden-
tified as causing a chronic bronchial disease called bys-
sinosis, often referred to as "Monday sickness." The disease
has been recognized in cotton textile mill workers since the
early nineteenth century.23 Symptoms are Monday morning
chest tightness and shortness of breath, which are more
obvious when exposure to cotton dust is resumed after a short
period (a weekend, for example) of nonexposure. Cotton plant
debris from bracts, stems, and leaves is believed to be the
harmful component. The specific harmful agent has not been
identified, but studies indicate that it is water soluble
and has the properties of a polyphenol. 2l*'25 The TLV for
raw cotton dust is 200 yg/m3, as measured with a vertical
elutriator that has a 15-ym theoretical particle size cut-
off.22'26
22TLV's® Threshold Limit Values for Chemical Substances
and Physical Agents in the Workroom Environment with In-
tended Changes for 1975. American Conference of Govern-
mental Industrial Hygienists. Cincinnati. 1975. 97 p.
23Kilburn, K. H., G. G. Kilburn, and J. A. Merchant.
Byssinosis: Matter from Lint to Lungs. American Journal
of Nursing. 7_3:1952-1956, November 1973.
2UMerchant, J. A., J. C. Lumsden, K. H. Kilburn, V. H.
Germino, J. D. Hamilton, W. S. Lynn, H. Byrd, and D.
Baucom. Preprocessing Cotton to Prevent Byssinosis.
British Journal of Industrial Medicine. 30^237-242, 1973.
25Hamilton, J. D., G. M. Halprin, K. H. Kilburn, J. A.
Merchant, and J. R. Ujda. Differential Aerosol Challenge
Studies in Byssinosis. Archives of Environmental Health.
26:120-124, March 1973.
29
-------
Emissions from mechanical cotton harvesting also contain
free silica (SiC>2) from soil dust. Silicosis can result
from prolonged exposure to free silica particles. The TLV
for reispirable particles (<10 ym in diameter) with free
silicc: content greater than 1% is given in mg/m3 by:22
TLV = — (1)
(% SiO2) + 2
where the "% Si02" is the percentage of respirable free
silicci, measured as quartz, in the dust. This relationship
was used for the £?-ym fraction of particles sampled in
this £>tudy. Analysis of particulate samples showed free
silica content of 7.9% for mechanical cotton picking and
an average of 2.3% for mechanical cotton stripping
(Appendix B).
2. Minor Emissions
Small quantities of chemical pesticide, defoliant, and desic-
cant residues are present in the particulates emitted by cotton
harvesting. Thousands of pesticides are registered with EPA
for use on cotton.15 Table 6 lists agricultural chemicals
commonly used on cotton.27 Where TLV's have not been estab-
lished, they were estimated from the LDso level which is the
26Neefus, J. D. Cotton Dust Sampling: I Short Termed
Sampling. American Industrial Hygiene Association
Journal. ^: 470-476, June 1975.
27Rawlings, G. D., and R. B. Reznik. Source Assessment:
Cotton Gins. Monsanto Research Corporation, EPA Contract
68-02-1874. Dayton. Preliminary document submitted to
the EPA, December 1975. 97 p.
30
-------
Table 6. TLV'S OF PESTICIDES APPLIED TO COTTON CROPS27
Type .of pesticide
Inorganic fungicides
Copper sulfate
Organic fungicides
Dithiocarbamates
Zineb
Phthalimides
Captan
Dinocapf dodine, quinones
Phenols
Organic herbicides
Arsenicals
Phenoxys
2,4-D
Phenyl ureas
Diuron
Linuron
Fluometuron
Amides
Alanap
Alachlor
Carbamates, see insecticides
Dinitro group
Triazines
Other organics
Trifluralin
Nitralin
Dalapon
Norea
Synthetic organic insecticides
Organochlorines
Strobane
DDT
Endrin
Dieldrin
Toxaphene
Organophosphorus
Disulfoton
Bidrin
Methyl parathion
Parathion
Trichlorfon
Azinphosmethyl
Phorate
Ethion
Carbamates
Carbaryl
Methomyl
Miticides
Dicofol
Chlorobenzilate
Omite
Pumigants
Dibromochloropropane
Telone
Defoliants and desiccants
Arsenic acid
DBF
Folex
Sodium chlorate
TLV,a
mg/m3
1.0
(5.0)
(5.0)
0.4
19.0
0.5
10.0
(5.0)
(5.0)
(5.0)
(5.0)
(5.0)
(0.1)
(5.0)
(10.0)
(5.0)
(1.0)
(5.0)
0.5
1.0
0.10
0.25
0.5
(0.1)
(0.1)
0.2
0.1
(1.0)
0.2
(0.1)
(0.1)
5.0
(0.1)
(1.0)
1.0
(5.0)
' (0.1)
(0.1)
0.25
(1.0)
(5.0)
(5.0)
Acute oral
I-Dso, mg/kg
300
>5,200
9,000
300 to 1,000
3,400
1,500 to 4,000
8,900
8,200
1,200
10 to • 60
3,000 to 5,000
>10,000
2,000
970
2,000
220
113 to 118
5 to 17.8
46
80 to 90
12.5
15 to 22
14 to 24
3.6 to 13
560 to 630
11 to 13
1.1 to 2.3
27 to 65
500 to 850
17
809
960
2,200
173
250 to 500
48 to 100
350
1,272
1,200
avalues in parentheses are assumed TLV's based on their
according to the following schedule:
TLV = 0.1 if LD50 <300 rag/kg
TLV = 1.0 if 300 50 <1,000 rag/kg
TLV = 5.0 if 1,00*0 10,000 mg/kg
31
-------
dose, in mg of compound/kg of body weight, that is lethal
to 50% of a population of test animals, usually male rats.
The pesticides most commonly sprayed on cotton crops in
1971 ware DDT (which has since been banned from use by EPA) ,
Toxaphene, and methyl parathion. Sodium chlorate, tributyl-
phosphorotrithioite (Folex) and tributylphosphorotrithioate
(DEF) accounted for more than 90% of all defoliants used on
cotton in 1971. The predominant desiccants applied to cotton
in 1971 were arsenic acid and paraquat, used in Texas and
Oklahoma.28
No data on the agricultural chemical content of cotton
harvesting dust were found in the literature. Maxima and
minima, of concentrations found in cotton gin trash and emis-
sions are summarized in Table 7.1^/29-32 Field samples of
28And3:ilenas, P. A. Farmers' Use of Pesticides in 1971...
Quantities. U.S. Department of Agriculture, Economic
Research Service. Washington. Publication No. ERS-536.
February 1974. 35 p.
29Fea:irheller, W. R. , and D. L. Harris. Particulate Emis-
sion Measurements from Cotton Gins, Delta and Pine Land
Co.,, Scott, Mississippi. Monsanto Research Corporation.
Dayton. Report No. MRC-DA-358. Environmental Protection
Agency, EMB Project Report No. 72-MM-16. November 1974.
239 p.
30Feairheller, W. R., and D. L. Harris. Particulate Emis-
sion Measurements from Cotton Gins, Bleckley Farm Service
Co., Cochran, Georgia. Monsanto Research Corporation. Day-
ton. Report No. MRC-DA-357. Environmental Protection Agency,
EMB Project Report No. 72-MM-23. November 1974. 265 p.
3 Emissions from Cotton Gin at Valley Gin Company, Peoria,
Arizona. PEDCo-Environmental Specialists, Inc. Cincinnati.
Environmental Protection Agency, EMB Project Report No.
72-MM-20. 1973. 37 p. plus Appendix.
32Durrenberger, C. Cotton Gin Report. Texas Air Control
Board. Austin. May 31, 1974. 50 p.
32
-------
Table 7. AGRICULTURAL CHEMICAL CONCENTRATIONS IN COTTON GIN EMISSIONS AND TRASH
Emission/trash
Concentration, ppm by weight
Maximum - where measured
Minimum - where measured
co
Organochlorine pesticides
p,p '-DDT
o,p-DDT
p,p'-TDEb
p,p'-DDEb
Toxaphene
Endrin
Dieldrin
TOTAL
Organophosphorus pesticides
Parathion
Methyl parathion
Merphos
Diazinon
TOTAL
Defoliants
DBF
Desiccants
Arsenic acid (as arsenic)
53.0
5.94
2.60
3.91
136
0.05
0.05
202
14.8
5.1
0.6
0.04
20.5
inclined cleaner trash29
inclined cleaner trash29
- inclined cleaner trash29
- inclined cleaner trash29
- inclined cleaner trash29
- inline filter inlet31
- inline filter inlet31
total gin trash31
total gin trash31
total gin trash31
inline filter inlet31
17.1 - total gin trash
31
2,200
- air downwind from gin
32
<0.01 - inline filter inlet31
1.0 - green leaf and stick extractor29
<0.01 - inline filter inlet31
0.02 - inline filter inlet31
0.32 - inline filter inlet31
<0.01 - total gin trash3o
0.01 - inline filter inlet31
0.14 - inline filter inlet31
0.01 - inline filter inlet31
c
<0.02 - inline filter inlet31
0.06 - inline filter inlet31
5.67 - gin emission sample
31
Minimum of detectable concentrations.
b
Degradation product'of DDT.
c
Only one sample taken.
-------
cotton harvesting emissions contained 2.94 ppm parathion,
1.47 ppm methyl parathion, and 4.90 ppm DBF in picker har-
vesting emissions and <0.07 ppm arsenic in stripper harvesting
emissions (see Appendix B). These concentrations are within
the ranges of those at cotton gins, so it is reasonable to
assume that Table 7 represents ranges that could be expected
in harvesting emissions.
B. EMISSION FACTORS AND EMISSION BURDENS
Table 8 lists respirable particulate (<7-ym diameter) emis-
sion factors for the principal types of cotton harvesting
operations in the U.S. The factors are based on average
machine speed, basket capacity, trailer capacity, lint cotton
yield, transport speed, and mass emission rates for the
respective harvester types. The confidence limits are based
on those for emission rates at the 95% confidence level.
Details of the calculations are shown in Appendix C. The
weighted average stripper factors are based on estimates
that 2% of all strippers are four-row models with baskets,
and 40% and 60% of the remainder are two-row models pulling
trailers and two-row models with mounted baskets, respec-
tively.33 Raw cotton dust emission factors are the sum of
those from harvesting and trailer loading, since field trans-
port emissions are soil dust only.
The free silica emission factors presented in Table 9 are
calculated from the total emission factors in Table 8 and
the free silica content as measured in emission samples
from picker and stripper harvesting.
33Personal communication. Dr. R. B. Metzer, Texas Agri-
cultural Extension Service, Texas A&M University, College
Station. January 13, 1976.
34
-------
Table 8. RESPIRABLE PARTICULATE EMISSION FACTORS FOR
COTTON HARVESTING OPERATIONS
(kg emitted/km2 harvested)
Type of harvester
Picker
Two- row,
Stripper
Two-row ,
Two-row,
Four-row
Weighted
with basket
pulled trailer
with basket
, with basket
C
average
Unit operation
Harvesting
0.455 ± 0.738
7.37 ± 16.7
2.30 ± 0.82
2.31 ± 0.82
4.28 ± 6.53
Trailer
loading
0.0699
_b
0.0918
0.0918
0.0560
Transport
0.427 ± 0.119
0.279 ± 0.078
0.279 ± 0.078
0.279 ± 0.078
0.279 ± 0.078
Total
0.952
7.65
2.67
2.68
4.61
No confidence limits, since factors based on only one sample.
Not applicable.
"Based on proportions of stripper types.
Table 9. FREE SILICA EMISSION FACTORS FOR COTTON HARVESTING
(kg emitted/km2 harvested)
Type of harvester
Picker
Two-row, with basket
Stripper
Two-row, pulled trailer
Two-row, with basket
Four-row, with basket
Weighted average
Emission factor
0.0752
0.176
0.061
0.062
0.106
Based on proportions of stripper types.
35
-------
Table 10 presents estimated maximum emission factors for
pesticides, defoliants, and desiccants based on Tables 7 and
8. Totals for organochlorine and organophosphate pesticides
are given since application of specific chemicals varies
widely.
Annual respirable particulate emissions from cotton harvesting
and their contribution to total emissions31* are listed by
state in Table 11. The emissions were calculated from the
emission factors in Table 8 and the 1971-73 average area
harvesited in Table 4. it was assumed that picker area was
picked, twice, excluding gleaned area, since it is common prac-
tice to harvest picker cotton twice in one season.12 Stripping
emissions were calculated from the weighted average total
emission factor for stripper operations, which was also
assumed applicable to gleaning. Using Texas as an example,
emissions from picking are:
0.455 ]^T (2 x 3,000 - 200)km2 + (0.0699 + 0.427)^-3,000 km2
= 4,100 kg = 4.1 metric tons
In the equation above it is not necessary to apply the factors
for trailer loading and field transport to twice the harvested
area, since those factors are based on total annual yield
(Appendix C). Annual emissions from Texas stripping and
gleaning operations are:
4.61 j^M (17,200 + 200)km2 = 80,200 kg = 80.2 metric tons
3**1972 National Emissions Report. U.S. Environmental Pro-
tection Agency. Research Triangle Park. Publication
NO. EPA-450/2-74-012. June 1974. 422 p.
36
-------
Table 10. ESTIMATED MAXIMUM PESTICIDE, DEFOLIANT, AND DESICCANT
EMISSION FACTORS FOR COTTON HARVESTING
(mg emitted/km2 harvested)
Type of harvester
Picker
Two -row,
Stripper
Two- row,
Two -row,
Four-row
Weighted
with basket
pulled trailer
with basket
, with basket
c
average
Pesticides
Organp-
chlorines
192
1,545
539
541
931
Organo-
phosphates
19.5
157
54.7
54.9
94.5
Harvest-aid chemicals
Defoliants
16.3
Desiccants
16,800
5,870
5,900
10,100
U)
-J
Assuming DEF.
Assuming arsenic.
r
Based on proportions of stripper types,
Note: Blanks indicate not applicable.
-------
Table 11. ANNUAL RESPIRABLE PARTICULATE EMISSIONS AND EMISSION BURDENS
FROM COTTON HARVESTING
State
Texas
Mississippi
California
Arkansas
Louisiana
Arizona
Alabama
Tennessee
Georgia
Missouri
Oklahoma
South Carolina
d
All other
U.S. TOTAL
(all states)
Respirable particulates from
cotton harvesting, metric tons/yr
Picking9
4.1
8.0
4.7
6.6
3.2
1.4
3.1
2.5
2.3
1.7
0.4
1.7
1.8
41.5
Stripping and gleaning"
80.2
0.5
3.2
7.4
0.9
92.2
State total
particulate
3U
emissions ,
metric tons/yr
549,399
168,355
1,006,452
137,817
380,551
72,685
1,178,643
409,704
404,574
202,435
93,595
198,767
1,928,310
17,872,000
Emission burden from
cotton harvesting, %c
0.046
0.014
0.002
0.014
0.003
0.019
0.001
0.002
0.002
0.003
0.025
0.003
0.002
oo
oo
Assuming picker area picked twice, excluding gleaned area (see text).
b
Using weighted average factor for stripping and assuming it applies to gleaning.
Assuming one-third of total particulates is respirable.
All other cotton producing states including North Carolina, New Mexico, Florida, Kentucky,
Virginia, Nevada, and Illinois.
Note: Blanks indicate <0.2 metric ton/yr.
-------
The emission burden is the fraction of total particulate
emissions contributed by cotton harvesting operations.
Since only respirable particulate emissions were sampled
(Appendix B), the emission burden was calculated by assuming
that one-third of a state's total particulate emissions are
respirable (<7 ym in diameter). Emission burdens are shown
in Table 11.
C. DEFINITION OF REPRESENTATIVE SOURCE
According to the 1969 Census of Agriculture, "cotton-type"
farms accounted for 29% of all U.S. farms growing cotton,
but harvested 52% of all lint cotton from 48% of all land
devoted to cotton. A "cotton-type" farm is defined as one
with annual cotton sales of at least $10,000 or 50% of the
total value of all farm products sold.7 In this document,
the representative cotton harvesting source is defined as a
cotton-type farm harvesting the average area of cotton which,
in 1971, was 0.786 km2.8 By defining the representative
source as a cotton-type farm, farms with a few thousand
square meters or less in cotton are eliminated from considera-
tion. These small farms are likely to harvest cotton by hand
and are not characteristic of mechanical harvesting opera-
tions. Solid planting (no rows skipped) was the predominant
(74% of all area) planting pattern on cotton-type farms in
1971.8 Representative row spacing is 1.02 m (40 in.).
To maintain good cotton quality, the seed cotton must not
contain too much moisture. Seed cotton moisture is con-
trolled by the relative humidity in the field, which is
highest before dew evaporates in early morning and when dew
begins to form in late afternoon. For this reason, cotton
is usually harvested from midmorning until late afternoon,
or approximately from 9 a.m. or 10 a.m. until 6 p.m. or
39
-------
7 p.m. on warm sunny days.14'16 Acceptable hours are shorter
on cool or cloudy days. On this basis, 8 hr was chosen as
the length of a representative cotton harvesting day.
Mechanical cotton pickers and cotton strippers have distinctly
different operating characteristics. Thus, it is necessary
to treat picker and stripper operations separately. Repre-
sentative picker yield is 63.0 metric tons of lint cotton/km2
harvested (from Tables 2, 3 and 4). The representative picker
is self-propelled, harvests two cotton rows simultaneously
at a speed of 1.34 m/s, and has a harvester-mounted basket
for collecting the picked cotton.12 Six harvester baskets
of cotton are required to fill a representative cotton
trailer, which has a capacity of 654 kg of lint (ginned)
cotton.35 The previously presented emission factors for
picking operations apply to the representative picker source.
Cotton-type farms that machine picked cotton in 1971 averaged
one picker per farm.8
An estimated 98% of all cotton strippers are two-row models,
^60% of which have mounted baskets and 40% pull trailers to
collect the harvested cotton.33 Hence, the representative
stripper is defined as a two-row model with a mounted basket.
Average speed while harvesting is 2.23 m/s.12 Representa-
tive stripper yield is 41.2 metric tons of lint cotton/km2
harvested (from Tables 2, 3 and 4). Six harvester baskets
are required to fill the representative cotton trailer with
654 kg of lint cotton. Cotton-type farms that stripped
cotton in 1971 averaged one stripper per farm.8
35Personal communication. Dr. R. B. Metzer, Texas Agri-
culture Extension Service, Texas A&M University, College
Station. December 16, 1975.
40
-------
For both pickers and strippers, 0.3 min is required for
turning at the ends of rows, 2 min are required for the
basket dumping operation, and filled cotton trailers are
transported in the field at a speed of 4.47 m/s.a
D. SOURCE SEVERITIES
Source severity is intended to be a relative measure of the
air quality impact of an air pollution source when compared
to other sources. Of particular interest is the maximum
severity from the representative source, defined as:
S = (2)
where S = maximum severity from the representative source
x" = maximum time-averaged pollutant concentration
to which the general public is exposed due to
emissions from the representative source
F = pollutant "hazard" factor
The concentration must be averaged over the same time base
as the hazard factor. The same units should be used for the
concentration and the hazard factor; the resulting source
severity value is dimensionless. For total suspended particu-
lates (TSP) the hazard factor is defined as the national
24-hr primary ambient air quality standard of 260 yg/m3; hence
the 24-hr average concentration must be used to calculate TSP
severity. For other pollutants from cotton harvesting, the
hazard factor is an adjusted TLV. The representative cotton
harvesting day is 8 hr long, which is the time base for
TLV's, so no time adjustment is necessary. However, TLV's are
From observation of representative operations.
41
-------
designed for industrial workers, so the hazard factor is
defin«;d as TLV/100 to account for exposure to the general
public. Eight-hour average concentrations are used to calcu-
late severities from these TLV-based hazard factors.
To calculate maximum concentrations, the representative
harvest area (0.786 km2/farm) was combined into one square
field, as shown in Figure 7. The representative distance
from emission sources to the receptor is one-half of a side
of the field. The receptor is located so as to be exposed
to the maximum pollutant concentration due to all opera-
tions. Maximum concentrations were calculated from average
respirable particulate emission rates (developed in Appen-
dix D) and operating parameters for the representative
sources by applying appropriate Gaussian diffusion models.36
The national annual average wind speed of 4.5 m/s and Class
C atmospheric stability were used. A dose model was applied
to trailer loading, and a modified point source cross-wind
integrated concentration model was applied to harvesters and
field transport. The concentrations were then averaged over
the representative operating time of 8 hr; 24-hr average
concentrations are found by dividing the 8-hr concentrations
by three. The detailed calculations are contained in
Appendix D.
36Turner, D. B. Workbook of Atmospheric Dispersion Estimates,
Revised 1970. U.S. Environmental Protection Agency,
Office of Air Programs. Publication No. AP-26. July 1971.
84 p.,
42
-------
1 WIND DIRECTION i
1 1 1 1 1
HARVESTER PATH
TRAILER LOADING
xaXX LOCATION
^TRAILER"
TRANSPORT
'PATH
Q
i _
t
D = 443m
586m
-W2J\ RECEPTOR
TOTAL AREA -0. 786 krn
Figure 7. Representative field for calculating cotton
harvesting severities
Calculated time-averaged maximum respirable particulate
concentrations for the representative sources are listed with
the resulting source severities for total suspended particu-
lates (TSP), "inert" dust, and raw-cotton dust in Table 12.
The 24-hr average concentration was used for TSP severities;
the 8-hr average was used for other pollutants. The highest
severities are from raw-cotton dust (which represents all
emissions from harvesting and trailer loading): 0.00703 for
mechanical picking operations (47% from harvesting and 53%
from trailer loading), and 0.0350 from mechanical stripping
operations (77% from harvesting and 23% from trailer loading),
Source severities for other types of strippers are calculated
in Appendix D.
Source severities for free silica and agricultural chemical
residues in the respirable particulate emissions are listed
in Table 13. Free silica severities are less than 0.01 and
the sums of all maximum agricultural chemical severities are
less than I0~k for representative picking and stripping
operations.
43
-------
Table 12. MAXIMUM SEVERITIES FOR TOTAL SUSPENDED PARTICULATES, "INERT" DUST,
AND RAW COTTON DUST FROM REPRESENTATIVE COTTON HARVESTING OPERATIONS
Harvesting operation
Mechanical picking
Time-averaged maximum concentration
24-hr avg
8-hr avg
Severity, S
Total suspended particulate
"Inert" respirable dustb
Raw cotton dustc
Mechanical stripping
Time-averaged maximum concentration
IIicLX
24-hr avg
8-hr avg
Severity, S
Total suspended particulate
"Inert" respirable dust"
Raw cotton dustc
Unit operation
Harvesting
0.00222
0.00665
0.00000854
0.0000665
0.00332
0.0181
0.0542
0.0000696
0.000542
0.0271
Trailer
loading
0.00247
0.00742
0.0000095
0.0000742
0.00371
0.00525
0.0158
0.0000202
0.000158
0.00790
Transport
0.00416
0.0125
0.000016
0.000125
0.00440
0.0132
0.00440
0.0132
Total
0.00885
0.0266
0.000034
0.000266
0.00703
0.0278
0.0832
0.000107
0.000832
0.0350
3Hazard factor, F = 260 yg/m3, 24-hr average.
Hazard factor, F = 100 ug/m3, 8-hr average; applies to dust with <1% free silica.
CHazard factor, F = 2 ug/m3, 8-hr average.
Note: Blanks indicate no cotton dust emissions from transport.
-------
Table 13. MAXIMUM SEVERITIES FOR FREE SILICA AND AGRICULTURAL
CHEMICAL RESIDUES FROM REPRESENTATIVE COTTON HARVESTING OPERATIONS
Pollutant
Fraction of total
respirable particulates
Source
severity
u
Free silica
Picking
Stripping
Agricultural chemicals
Pesticides
C
Organochlorines
Picking
Stripping
d
Organopho sphate s
Picking
Stripping
g
Total pesticides
Picking
Stripping
Defoliants (picking only)
DBF
Desiccants (stripping only)
Arsenic acid (as arsenic)y
Total agricultural chemicals
Picking
Stripping
7.9%
2.3%
202 ppm
20.5 ppm
222 ppm
17.1 ppm
2,200 ppm
2,440 ppm
0.00263
0.00357
0.00000537
0.0000168
0.000000545
0.0000017
0.00000592
0.0000185
0.000000046
0.00000732
0.00000597
0.0000258
3For agricultural chemicals, the
(see Table 7).
TLV given by Equation 1.
P
Minimum TLV for this group is 0,
Minimum TLV for this group is 0.
Minimum TLV for this group is 0.
f
-------
The field size given as representative is approximately the
largest area that can be harvested in a day. Since the
source severities are all very small (<0.01) and cotton
fields are rarely larger than the representative size,
severity distributions are not presented in this report.
Affected population from a representative source is defined
as the number of persons exposed to a severity >1.0. Since
maximum severities for all pollutants are less than 0.1, the
affected population from representative cotton harvesting
operations is zero.
46
-------
SECTION V
CONTROL TECHNOLOGY
A. STATE OF THE ART
None of the current cotton harvesting equipment or practices
provide for control of emissions. In fact, equipment designs
and operating practices tend to maximize emissions. Harvester
conveying and cleaning systems are designed to remove cotton
trash and leave it in the field. Harvester collection baskets
and cotton trailers are purposely designed to let blower air
and/or wind carry trash and dust away from the cotton.5'16
Trash removal in the field increases cotton grade and reduces
ginning costs.
Preharvest treatment (defoliation and desiccation) and harvest
timing practices are used to minimize moisture and trash in
the harvested cotton and maximize harvest efficiency. These
practices, especially desiccation which leaves plant parts
dry and brittle, increase cotton grade and yield, but also
tend to maximize emissions.
Soil dust emissions from field transport can easily be re-
duced by reducing vehicle speed. However, vehicle speed is
usually governed by the condition of the field or road, since
emissions are usually of minor concern (if any) to the drivers,
47
-------
B. FUTURE CONSIDERATIONS
Control of emissions from mechanical cotton harvesters
would involve enclosing cotton cleaning, conveying, and
collection systems and venting them through a particulate
collection device. This would require a fan or blower to
move the particulate-laden air. Practical collection devices
would be cyclones, skimmers, and filters or baghouses.
Filters and baghouses are the least likely to be employed
because the volume of trash collected would require very
frequsnt cleaning or huge sizes. Cyclones and skimmers
remove large particles (>10 pm) most efficiently and are not
effective in removing respirable particles. Their efficiencies
can be improved by the addition of a wetting system, but this
would require hauling water on the harvesters. Any emissions
control would be most easily applied to harvesters with
mounted baskets.
Use of defoliants and desiccants is not likely to be reduced
unless the cotton harvester manufacturing industry has un-
expected success in designing machines that can efficiently
harveist cotton from plants with full green foliage. Harvest
timing practices are also unlikely to change, since minimizing
moisture is one of the key requirements in maximizing cotton
grade.
In addition to reducing vehicle speed, emissions from field
transport of cotton can be reduced by watering or chemical
treatment with dust suppressants or oil. Such chemical treat-
ment is practical only if transport is on a field road, and
even then is undesirable due to soil residues and leaching
during rain.
Unless government agencies impose emission regulations on
cotton harvesting operations, it is unlikely that emission
48
-------
control technology will be applied, mainly because of the
high cost to farmers. Due to competition from foreign
suppliers and lower priced synthetic fibers, a primary aim
of U.S. cotton farmers is to reduce costs. This is the
chief reason that cotton harvesting in the U.S. has become
almost completely mechanized.
49
-------
SECTION VI
GROWTH AND NATURE OF THE INDUSTRY
A. PRESENT TECHNOLOGY
The equipment and practices currently employed in mechanical
cotton harvesting are described in Section III. More than
99% each of national cotton production and harvested area are
presently harvested mechanically. Stripping is a dry-weather
practice, and has evolved as the predominant harvest method
in the moderately dry and arid areas of the Southwest.
Though more expensive and less efficient, mechanical pick-
ing ramains the chief harvest method in the Midsouth and
Southeast. Stripping has been mostly experimental in these
areas. Problems caused by shallow-rooted and tender plants,
by loose sandy soil, and by excessive cotton losses due to
rainy and windy weather while waiting for all bolls to open
indicate that strippers are not well suited to humid climates
and silty lowland soil.5
B. EMERGING TECHNOLOGY
Use of cotton strippers may rise as plant varieties are
developed that can.adapt to humid areas and still retain
the characteristics necessary for good stripper harvesting.
Finger-type strippers are replacing roller-type strippers
in some areas where plant uprooting is not a problem. The
50
-------
finger-type strippers are mechanically simpler, require
less maintenance, and are less expensive to own and operate.
They also have less potential fine particulate emissions,
since there are no rollers to crush the dry leaves, stems,
and burs.
Due to a decreasing number of gins and shorter ginning seasons,
transport to the gin is a rising cost to the cotton farmer.
Backup of loaded trailers at the gin forces farmers to have
more and larger trailers available which has resulted in
increased use of harvester-mounted baskets and larger trailers
(with up to doubled capacity), and decreased use of harvester-
pulled trailers. "Module" transport is also increasing,
wherein the cotton-filled baskets are removed from the
harvester and hauled to the gin on a truck, eliminating the
need for trailers.
The trend toward fewer but larger cotton farms is increasing
the number and size of harvesters. On larger farms it is be-
coming common practice to use two or more harvesters in a
field simultaneously. Machines capable of harvesting up to
four rows of cotton at a time are appearing.
Ginners are encouraging farmers to bring cleaner cotton to
the gins. Less trash in the cotton means higher grade and
price to the farmer and lower cleaning cost to the ginner.
Harvester-mounted trash removal equipment adds to the har-
vester purchase price and increases maintenance and power
requirements. The extra cost is high considering that the
harvesters are used only 1 or 2 months a year. A possible
alternative is a separate trash removal machine for use in
the field. This would centralize trash removal and resulting
emissions, making it more practical to apply emission control
51
-------
equipment. Field extraction of trash is not now widely
practiced.2 7
C. INDUSTRY PRODUCTION TRENDS
Cotton production, harvested area, and yield trends from
1957 to 1974 are shown in Figure 8.4'37 After a poor 1967
harvest year, production and harvested area rose and then
decreased slightly from 1972 to 1974. U.S. cotton con-
sumption per capita has been decreasing steadily since 1965,
and cotton exports increased 100% from 1968 to 1972.37
Increasing competition from foreign cotton producers and
synthetic fibers have suppressed the growth of domestic
cotton production. Unless petroleum prices rise drastically,
causing a concomitant rise in synthetic fiber prices, U.S.
cotton production is not expected to increase much, if any,
above average 1971-73 levels by 1980. Any production in-
crease will probably be due to increased yield rather than
an increase in harvested area. Production, yield, and
harve:sted area in any one season are completely dependent
on weather conditions.
1957 I960 1963 1966 1969
. YEAR BEGINNING AUGUST 1
1972
1975
Figure 8. U.S. cotton area, yield, and production; 1957-744'37
371973 Handbook of Agricultural Charts. Agricultural Hand-
book No. 455. Washington, U.S. Department of Agriculture,
October 1973. 152 p.
52
-------
SECTION VII
UNUSUAL RESULTS
A.
SEASONAL NATURE OF THE INDUSTRY
As an agricultural operation, cotton harvesting is seasonal
in nature. The U.S. cotton harvest season runs from August
through January, but harvesting in any one area is concen-
trated in a much shorter period. The average cotton gin
operates only 10 weeks per year;27 the area around the average
gin is harvested over the same period. Figure 9 and Table 14
show the usual cotton harvesting dates in the U.S.
BEFORE AUG. 20
AUG. 20 - SEPT. 9
SEPT. 10 - SEPT. 30
'OCT. 1 OCT. 20
AFTER OCT. 20
Figure 9. Usual start of cotton harvest season
53
1 7
-------
Table 14. USUAL COTTON HARVESTING DATES, BY STATE17
State
Illinois
Missouri
Virginia
North Carolina
South Carolina
Georgia
Florida
Kentucky
Tennessee
Alabama
Mississippi
Arkansas
Louisiana
Oklahoma
Texas
New Mexico
Arizona
Nevada
California
Usual harvesting dates
Begin
Sep 15
Sep 15
Sep 15
Sep 15
Sep 1
Sep 1
Aug 15
Sep 15
Sep 15
Sep 5
Sep 20
Sep 15
Aug 25
Oct 15
Aug 1
Sep 10
Sep 1
Oct 15
Oct 1
Most
Sep 30
Oct 1
Sep 25
Oct 1
Sep 20
Sep 15
Sep 15
Oct 1
Sep 25
Sep 20
Oct 5
Oct 1
Sep 15
Nov 10
Nov 1
Oct 15
Oct 15
Oct 25
Oct 15
active
- Oct 25
- Nov 1
- Nov 1
- Nov 15
- Nov 1
- Oct 15
- Oct 15
- Oct 25
- Nov 15
- Dec 1
- Nov 5
- Nov 10
- Nov 15
- Dec 5
- Dec 1
- Nov 15
- Dec 10
- Dec 15
- Dec 1
End
Nov 5
Dec 15
Dec 1
Dec 10
Dec 1
Nov 15
Oct 30
Dec 1
Dec 5
Dec 20
Dec 10
Dec 15
Dec 1
Dec 15
Dec 20
Dec 15
Jan 15
Jan 1
Jan 15
In assessing emissions from cotton harvesting, it is im-
portant to consider the length of the season. The emission
burdens presented in Section IV, for example, would be more
illustrative if they could be based on a time span shorter
than a year. The length of the season is also important
in evaluating the impact on air quality.
54
-------
B. COTTON STRIPPER EMISSION RATES
Results from field sampling (Appendix B) showed that the
mean mass emission rate from the sampled four-row stripper
was only 63% of the rate from the sampled two-row stripper
( Appendix C). This result is surprising, considering that
the four-row harvest rate is approximately twice the two-row
harvest rate. Based on field observation, the major particu-
late emission point is located where the cotton discharges
from the harvester conveying system, which is usually
pneumatic. The sampled four-row stripper had a mounted
basket, while the two-row model pulled a trailer to collect
the seed cotton. It was concluded that harvesters with
mounted baskets have lower mass emission rates due to the
shorter distance and time in which the cotton and trash are
airborne before reaching the collection basket.
It is possible that the anomaly in emission rates was also
partly caused by the difference in stripper types. The two-
row model used rotating rolls to remove the cotton from the
plants, while the four-row model used stationary fingers and
slits. The roll-type model has a higher potential for
emitting fine particles since the rollers crush the dried
leaves,stems, and burs.
C. RELATION OF FIELD SIZE TO SEVERITY
It is a common inclination to expect larger sources to
cause higher pollutant concentrations (and hence severity)
than smaller ones. Such is not the case with cotton harvest-
ing. The reason is that increasing field size increases the
distance to the nearest receptor and increases the area over
which pollutants are emitted. With constant operating
parameters and emission rate, pollutant concentration (and
severity) is inversely proportional to D1'814, where D is
55
-------
half of the width of the field and represents the average
distance to the nearest receptor (Appendix D).
The above discussion indicates that the trend toward fewer
and larger cotton farms may decrease the maximum severity
for representative harvesting operations. The use of multiple
and larger harvesting machines will negate the farm size
effect to some degree.
56
-------
SECTION VIII
APPENDIXES
A Comparison of Stripped and Overall Cotton Yield
B Sampling and Analysis Procedures and Results
C Emission Rate and Emission Factor Calculations
D Source Severity Calculations
57
-------
APPENDIX A
COMPARISON OF STRIPPED AND OVERALL COTTON YIELD
Infornation on harvested cotton yield is necessary to deter-
mine the area harvested by each harvest method when only pro-
duction weight data are available. As discussed in Section
III, only stripper-harvested yield should cause a noticeable
difference between proportions of weight and area harvested
among machine picking, stripping, and gleaning. To deter-
mine stripper yield, production data from Texas and Oklahoma
were examined; these two states together harvest 99.6% of
the total weight of stripped cotton.
Texas estimates were based on data from seven crop reporting
districts which accounted for 95.7% of the total stripping
machines in the state but only 1.8% of the total picking
machines.5 Data from the 1971-73 seasons were averaged to
negate effects of particularly good or poor harvests. The
data and results are shown in Table A-l. The average yield
for the seven districts was 41.2 metric tons/km2 harvested
compared with the average overall state yield of 41.6 metric
tons/km2.
Oklahoma estimates were based on data from seven counties in
which machine picking accounted for less than 10% of the
harvested weight from cotton-type farms in 1971.8 The data
and results are shown in Table A-2. The yield for primary
stripper counties was 27.2 metric tons/km2 harvested compared
with a state yield of 27.9 metric tons/km2 for all cotton-
type farms.
Since the difference between estimated stripper yield and
overall state yield is less than 1 metric ton/km2 for both
states, it is reasonable to assume that yields within a state
are the same for machine picking, machine stripping, and hand
labor harvesting.
58
-------
Table A-l.
LINT COTTON YIELD FOR PRIMARY STRIPPING DISTRICTS
IN TEXAS, 1971-73
District
1-N
1-S
2-N
2-S
3
4
7
Area harvested,
km2
19716
1,651
7,374
1,716
2,064
105
3,096
243
Total for 16,249
above
districts
19726
1,619
7,596
1,817
2,214
138
3,136
356
16,876
197319
1,708
9,166
2,104
2,226
138
2,857
360
18,559
Lint cotton production,
10 3 metric tons
19716
47.5
225.8
44.7
43.8
1.3
67.8
8.1
439.0
19726
89.1
382.7
83.9
90.4
3.3
107.0
12.0
768.4
197319
94.6
513.0
105.0
100.2
4.4
93.1
12.6
922.9
1971-73 Avg.: 17,228
1971-73 Avg. yield:
1971-73 Avg.: 710.1
41.2 metric tons/km2
For Texas 1971-73 Avg.: 20,250
1971-73 Avg. yield;
1971-73 Avg.: 841.8
41.6 metric tons/km2
Table A-2. LINT COTTON YIELD FOR COTTON-TYPE FARMS IN
PRIMARY STRIPPING COUNTIES IN OKLAHOMA, 19718
County
Caddo
Custer
Grady
Greer
Kiowa
Tillman
Washita
Area harvested,
km2
12.4
16.3
10.7
13.4
32.9
55.9
84.0
Lint cotton production,
metric tons
318
528
307
379
950
1,241
2,411
Total for above
counties
225.6
1971 yield:
6,134
27.2 metric tons/km2
Total for
Oklahoma
438.6
1971 yield;
12,232
27.9 metric tons/km2
59
-------
APPENDIX B
SAMPLING AND ANALYSIS PROCEDURES AND RESULTS
1. RKSPIRABLE PARTICULATES
A GCA Model RDM 101-4 portable respirable dust monitor3 was
used to sample emissions from cotton harvesting operations.
With this monitor, digital readout of mass concentration is
obtained from electronic measurement of beta ray absorption
of the collected sample. A small cyclone removes particles
of ^7-jjm mean aerodynamic diameter ahead of the collection
chamber, so that only respirable concentrations are measured.
With the cyclone removed, the instrument is capable of
measuring total particulate mass concentrations. However,
the high concentration of large plant fragments from cotton
harvesting presented the danger of plugging the sampling
orifices and, consequently, total particulate concentrations
were not measured. The instrument is capable of measuring
dust concentrations from 20 yg/m3 to 10,000 pg/m3 with
sampling times of 4 min or longer. The accuracy of the
instrument is defined by the manufacturer as "...within ±25%
of the measurements obtained from a companion simultaneous
gravimetric respirable mass sample for 95% of the samples."
Dust concentrations from harvesting were measured by follow-
ing each harvesting machine through the field at a constant
distance directly downwind from the machine while staying in
the visible plume centerline. The respirable dust monitor
was carried overhead with arms extended to minimize ground
effects and body interference. Distance downwind was
aGCA Corporation; GCA/Technology Division; Bedford,
Massachusetts.
60
-------
visually estimated, using the cotton row spacing and number
of rows as a reference. The procedure for trailer loading
was the same, but since the trailer is stationary while
being loaded, it was necessary only to stand a fixed distance
directly downwind from the trailer while the plume or puff
passed over. Readings taken upwind of all field activity
gave background concentrations.
Wind speeds were measured with an anemometer assembled by
MRC which has cup-type rotors attached to the shaft of a
small electric motor. Generated current is read in micro-
amperes from an ammeter and converted to wind speed with a
calibration chart. Readings of %15 s were taken before and
after each particulate sample, and the average was recorded.
Atmospheric stability class was determined from wind speed
and atmospheric conditions by using the chart of Figure B-l.38
Field data were recorded on the form shown in Figure B-2.
The terms on the form are explained in Table B-l.
Table B-l. EXPLANATION OF TERMS ON FIELD DATA FORM
Term
Meaning
Read., mg/m3
Cone., yg/m3
R/T
BCD, pg/m3
A, yg/m3
Q, g or g/s
S1
M
Concentration reading
Converted concentration for sampling times greater
than 4 min (lower right hand corner)
respirable reading
total mass reading
R
T
Background concentration
The difference between the converted concentration
and the background
Calculated emission rate
Stability for the time of day the unit operation
was sampled
The diffusion model used, referenced as 1, 2, or 3
(point, line, or dose, respectively)
38Blackwood, T. R., T. F. Boyle, T. L. Peltier, E. C. Eimutis,
and D. L. Zanders. Fugitive Dust from Mining Operations.
Monsanto Research Corporation. Dayton. Report No. MRC-DA-
442. (EPA Contract 68-02-1320, Task 6.) May 1975. p. 34.
61
-------
to
ATMOSPHERIC
CLASS ISO
RADIATION INDEX =
-2
I NO
RADIATION INDEX =
- I
TIME OF DAY
INSOLATION
CLASS
wnnuriMF
LATE AM. EARLY PM
MID AM, MID PM
EARLY AM, LATE PM
4
3
2
1
Figure B-l. Flow chart for atmospheric stability class determination38
-------
MODEL: POINT-1
LINE = 2
SOURCE TYPE
DATE
BY_
(Ti
OJ
UNIT OPERATION
WIND
SPEED
MPH
DIS1
X
FANC
Y
E, FT
Z
TIME
MIN.
READ.
mg/m^
CONC
TIME OF DAY
ATM.STABILI1Y
R/T
BGD,3
A,
Q,
g
s1
M
COMMENTS
TOTAL SAMPLING TIME MULTIPLY READING BY
4 MINUTES 1
8 MINUTES 0.46
16 MINUTES 0.23
20 MINUTES 0.184
30 MINUTES ' 0.122
37 MINUTES 0.1
Figure B-2. Field data form
-------
Mass emission rates were calculated from the field data
with appropriate Gaussian plume diffusion models. Emission
rates for harvester emissions were calculated with the model
for ground level concentration at the plume centerline from
a stationary ground level point source:36
*(x) =
where x = ground level concentration at plume center-
line, g/m3
x = distance downwind from source, m
Q = source mass emission rate, g/s
IT = 3.14
0,0 = standard deviation of plume concentration
y distribution horizontal and vertical to the
plume, respectively, m
u = mean wind speed, m/s
This model assumes that the plume spread has a Gaussian
distribution in the horizontal and vertical planes, that
there is total reflection of the plume at the earth's sur-
face (no deposition or reaction), and that sampling time is
a few minutes. The standard deviations, o and o , are
calculated from Tables B-2 and B-3, developed from published
empirical plots of o and o versus downwind distance from
the source o 39' 1+0 Although the harvester was a moving source,
the receptor was also moving; thus the sampling procedure
employed enables treatment as a stationary point source.
39Eimutis, E. C., and M. G. Konicek. Derivations of Con-
tinuous Functions for the Lateral and Vertical Atmospheric
Dispersion Coefficients. Atmospheric Environment.
£:£!59-863, November 1972.
^Martin, D. V., and J. A. Tikvart. A General Atmospheric
Diffusion Model for Estimating the Effects on Air Quality
of One or More Sources. Presented at 61st Annual Meeting
of the Air Pollution Control Association, for NAPCA,
St,, Paul, 1968. 18 p.
64
-------
Table B-2. CONTINUOUS FUNCTION FOR LATERAL
ATMOSPHERIC DIFFUSION COEFFICIENT, o 39
0 =
where x = downwind distance from source
Stability class
A
B
C
D
E
F
a
0.3658
0.2751
0.2089
0.1471
0.1046
0.0722
Table B-3.
CONTINUOUS FUNCTION FOR VERTICAL ATMOSPHERIC
DIFFUSION COEFFICIENT, a 40
az = ex + f
Usable range
>1,000 m
100-1,000 m
<100 m
Stability
class
A
B
C
D
E
F
A
B
C
D
E
F
A
B
C
D
E
F
Coefficient
0.00024
0.055
0.113
1.26
6.73
18.05
C2
0.0015
0.028
0.113
0.222
0.211
0.086
0.192
0.156
0.116
0.079
0.063
0.053
2.094
1.098
0.911
0.516
0.305
0.18
d2
1.941
1.149
0.911
0.725
0.678
0.74
0.936
0.922
0.905
0.881
0.871
0.814
-9.6
2.0
0.0
-13
-34
-48.6
£2
9.27
3.3
0.0
-1.7
-1.3
-0.35
0
0
0
0 '
0
0
65
-------
Mass emission rates for cotton trailer loading (harvester
basket dumping) were calculated with the Gaussian model for
ground level dosage at the plume centerline from a stationary
ground level point source with a finite mass release:36
QT
DT(x) = - ^- (B-2)
where D (x) = total dosage, g-s/m3
Q = total mass release, g
The total dosage can be estimated as
DT(x) £ X(x)-ts (B-3)
where x(x) = measured concentration from sampling, g/m3
t = sampling time, s
s
In using Equation B-2 care must be taken that a and a are
representative of the release time and that the plume path is
accura.tely known. Sample time was 4 min and wind speed and
direction were steady during sampling, so the dose model
should: give reasonable results.
Mass emission rates for each sample were calculated by com-
puter. Data were screened for validity and void samples were
eliminated before calculations were performed. The input
data and calculated emission rates are presented in Table B-4.
2. COMPOSITION
Particulate samples for analysis of selected compounds
were collected with a General Metal Works GMWL-2000 high
66
-------
Table B-4. MASS EMISSION RATES FROM COTTON HARVESTING
SAMPLES - COMPUTER INPUT AND RESULTS
Unit operation
Machine picking
ii
"
Machine stripping
Two-row, pulling trailer
M
H
Four-row, with basket
ii
n
n
n
n
Trailer loading
Four-row stripper
u, mph
7
5
3
3
7
7
20
20
10
10
10
10
10
x, ft
20
35
40
100
60
110
30
30
50
40
40
40
40
t , min
s
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
"'x, yg/m3
190
30
50
420
250
500
610
1,750
550
480
680
740
70
S1
B
B
B
C
C
C
D
D
C
C
C
C
B
M
1
1
1
1
1
1
1
1
1
1
1
1
3
Q
0.002171 g/s
0.00068 g/s
0.0008677 g/s
0.02068 g/s
0.01140 g/s
0.06825 g/s
0.01032 g/s
0.02961 g/s
0.02578 g/s
0.01503 g/s
0.02128 g/s
0.02317 g/s
0.9718 g/dump
CTi
-------
volume: air sampler and Nuclepore filters. To aid transport
and hc.ndling, the motor and filter assembly were removed from
the sampler housing and mounted on a modified tripod stand.
The sampler was placed in or adjacent to the cotton fields
downwind of harvesting activity in a position to collect the
highest possible concentration. The sampler was powered by
a portable gasoline-engine electric generator placed approxi-
mately 20 m downwind from the sampler. Air flow rate was
adjusted to about 0.8 m3/min to 1.2 m3/min, measured with a
flowmeter (visi-float) calibrated with the sampler and a
clean Nuclepore filter.
All sanples were analyzed for total sample mass and free
silica, content. Analyses for pesticides and harvest-aid
chemicals (defoliants and desiccants) were selected based on
discussions with the respective farm operators of what chemi-
cals had been applied to the crop. Samples from cotton
picking were analyzed for parathion, methyl parathion, and
DEF; samples from cotton stripping were analyzed for arsenic.
Picker and stripper cotton plant samples were also analyzed
for the respective chemicals for comparison with particulate
samples. A top soil sample was also analyzed for free silica.
Parathion, methyl parathion, and DEF were analyzed by gas
chromatography after extraction with pesticide-quality
hexane. Free silica analysis was performed by low temperature
ashing followed by dilution with potassium bromide and measure-
ment by infrared spectophotometry. The arsine generator
method vras used for arsenic analysis. Samples were split
when necessary. Results are presented in Table B-5.
General Metal Works, Inc., Cleves, Ohio.
Nuclepore Corporation, Pleasanton, California.
68
-------
Table B-5. COMPOSITION ANALYSIS RESULTS
vo
Mechanical cotton.picking
Filter samples
Total particulate concentration
Parathion
Methyl parathion
DBF
Free silica
Cotton samples
Parathion
Methyl parathion
DEF
Mechanical cotton stripping
Filter Samples
Total particulate concentration
Arsenic
Free silica
Seed cotton sample
Arsenic
Cotton foliage sample
Arsenic
Texas Blackland top soil
Arsenic
Free silica
161 yg/m3
2.94 ppm (wt)
1.47 ppm (wt)
4.90 ppm (wt)
7.9% (wt)
0.219 ppm (wt)
0.0820 ppm (wt.)
0.417 ppm (wt)
Sample 1
337 yg/m3
<0.07 ppm (wt)
3.4% (wt)
Sample 2
202 yg/m3
<0.07 ppm (wt)
1.2% (wt)
22.6 ppm (wt)
2.31 ppm (wt)
<0.3 ppm (wt)
13% (wt)
-------
The total particulate concentrations cannot be used to
calculate emission rates because of inconsistent harvesting
patterns and source location, and multiple source interference.
70
-------
APPENDIX C
EMISSION RATE AND EMISSION FACTOR CALCULATIONS
Neither specific data nor estimated emission rates or emis-
sion factors for cotton harvesting operations were found in
the literature. Therefore, values were obtained from a
limited sampling program supplemented with information from
the literature and contacts with cotton experts. The sources
sampled were: one two-row mechanical picker with harvester-
mounted basket, one tractor-mounted two-row brush-type stripper
with a tractor-pulled trailer, one self-propelled four-row
finger-type stripper with harvester-mounted basket, and
trailer loading from the four-row stripper. Emissions from
field transport were based on results from sampling of grain
harvesting activities.kl Although the number of sources
sampled was small, care was taken to choose operations repre-
sentative of the cotton harvesting industry so that the
resulting emission estimates would be similarly representative.
The rest of this appendix shows how emission rates and emission
factors were derived.
1. EMISSION RATES
Emission rates were derived in the form:
Q = Q±(CL)g5% (C-l)
where Q = mass emission rate
Q = arithmetic mean emission rate from all samples
(CL)g5% = confidence limit at 95% confidence level
41Wachter, R. A., and T. R. Blackwood. Source Assessment:
Harvesting of Grain, State of the Art, Monsanto Research
Corporation, EPA Contract 68-02-1874. Dayton. Preliminary
document submitted to the EPA, December 1975. 81 p.
71
-------
The confidence limits are given by:
t
- n
(C-2)
whens t
95%
a =
n =
N =
± values between which 95% of the area under
the "Student t" distribution lies, with n - 1
degrees of freedom
estimated population standard deviation,
estimated from the samples
number of samples
number in population
The estimated population standard deviation is given by:
(2Q7T2~I
a =1
n
?
n
(C-3)
N
where Q. = emission rate from the i
.th
sample
For cotton harvesting, the total population is large (in the
thousands) and the number of samples was small (<6). There-
fore 3 is estimated by dropping the last square root term in
Equation C-3. Similarly the confidence limits are estimated
by dropping the last square root term in Equation C-2. Hence,
the equations used to calculate the confidence limits were:
(CL)
t95%g
95%
(C-4)
where
a =
te
n
(Qi
(SQ. )
2 -
n
(C-5)
Emission rates were calculated as described above for cotton
pickers, two-row strippers pulling trailers, and four-row
72
-------
strippers with harvester-mounted brackets, using the data
from Table B-4.
It was assumed that emissions from a given type of harvester
are proportional to the rate of material handling. Two-row
strippers travel at about the speed of four-row strippers.12
Therefore, the emission rate from two-row strippers with
harvester-mounted baskets was estimated as half the rate from
four-row strippers.
Travel speed for field transport was estimated as 4.47 m/s
(10 mph). The emission rate for field transport was calculated
from the emission factor for field transport in grain har-
vesting at that speed.
Emissions from harvester basket dumping were assumed pro-
portional to the quantity of cotton dumped, the height of
drop, and the fine trash (fine leaf and dirt trash) content
of the cotton. The sample result shown in Table B-4 was
obtained from the dumping of one basket on a four-row stripper
into an empty trailer. This value was halved to account for
an average trailer being half full. Baskets on two-row
strippers are about half the size of four-row stripper baskets,
and emissions for dumping the smaller baskets were assumed to
be half as much. Picker baskets are about the same size as
two-row stripper baskets, but picked cotton contains only
half as much fine trash, 1J* so emissions from picker basket
dumping were assumed to be half those from a two-row stripper
basket. Wind speed also affects emissions during trailer
loading, but the sample in Table B-4 was taken with a wind
speed of 4.47 m/s, which equals the national annual average
wind speed. Confidence limits could not be assigned for
trailer loading emissions since only one sample was taken.
73
-------
Resulting mass emission rates are shown in Table C-l. These
rates apply only while the unit operations are being per-
formed and cannot be totaled to obtain an overall emission
rate.
Table C-l. MASS EMISSION RATES FROM COTTON HARVESTING
OPERATIONS, WITH 95% CONFIDENCE LIMITS
Type of harvester
Picker
Stripper
Two- row,
pulling trailer
Two -row,
with basket
Four-row,
with basket
Harvesting,
mg/s
1.24 ± 2.01
33.4 ± 75.8
10.4 ± 3.7
20.9 ± 7.4
Basket
dumping,
mg
121
a
243.
486
Transport,
mg/s
22.4 ± 2.2
22.4 ± 2.2
22.4 ± 2.2
22.4 ± 2.2
Not applicable.
2. EMISSION FACTORS
Emission factors were developed in terms of mass emitted
per unit area harvested. Four chief reasons explain the
choice of area harvested as the emission factor base.
(1) To be useful, emission factors must be based on
t
easily quantified variables. The most readily
attainable data for cotton harvesting are weight
of lint cotton harvested and area harvested.
(2) Emission factors should be based on the variables
that exert the strongest influence on total emissions,
Since the time and resources available precluded
74
-------
quantifying the effects of several variables, it
was necessary to choose a single variable for the
emission factor base. Emissions from a given cotton
harvesting machine with specified crop conditions
are determined by the amount of trash handled.
Trash is composed mainly of plant fragments and
soil dust. It was concluded that the amount of trash
handled is best represented by the number of cotton
plants harvested, and hence area harvested under
representative planting patterns, rather than the
amount of cotton harvested from those plants. There-
fore, it is felt that area harvested is the variable
which best represents total harvester emissions
within a harvest method.
(3) Total emissions from cotton trailer loading (harvester
basket dumping) and field transport may be better
represented by the weight of cotton produced than
by the area harvested, since the total production
determines the number of times these unit operations
are performed. However, if emission factors are
derived from emission rates for representative condi-
tions - as they are in this document - yield, and hence
production, is inherently included in the derivations.
Moreover, if trailer loading and field transport
emission factors are put on a common basis with
harvester emission factors, the factors within a
harvest method can be added to obtain an overall
total emission factor.
(4) It is common practice to harvest picker cotton twice
in the same season. The second harvest may be picked,
stripped, or gleaned. Emissions from this harvest
area overlap are easily calculated from an area-based
emission factor.
75
-------
Derivations of numerical emission factors are described below.
a. Harvesters
Emission factors for cotton harvesters were derived from the
basic formula:
(C-6)
where; E = harvester emission factor, mg/m2 or kg/km:
harvested
Q-. = harvester emission rate, mg/s
rl
A = harvest rate, m2/s
Each factor in Equation C-6 actually represents the mean from
a representative source, as does each quantity shown in the
rest of this appendix.
The machine harvest rate is determined by
A = VH • wg (C-7)
where v = harvester speed while harvesting, m/s
£1
w = width of harvester swath, m
o
and the width of the harvester swath is simply
w = r • w (C-8)
S JT
where r = number of cotton rows in one swath, an integer
w = cotton row spacing, m
76
-------
The representative cotton row spacing is 1.016 m (40 in.).42
Substituting this value in Equation C-8 and combining Equa-
tions C-6, C-7, and C-8, the harvester emission factor is
given by:
°H
EH ~ vu - r - (1.016 m) (C~9)
n
b. Trailer Loading
Emission factors for cotton trailer loading were derived
from:
where E = trailer loading emission factor, mg/m2 or kg/km2
harvested
QT = emission mass from one harvester basket dump, mg
LI
A^, = area harvested in filling one harvester basket, m2
£5
The area harvested in filling one harvester basket is deter-
mined by
AT
A = (C-ll)
where A = area harvested in filling one cotton trailer, m2
nB = number of harvester baskets to fill one trailer,
an integer
and A is calculated as
A
42Personal communication. Mr. Dan Pustejovsky, Hillsboro,
Texas. October 1, 1975.
77
-------
where C = harvested cotton capacity of one trailer, kg
Y = yield of harvested cotton, kg/m2
Combining Equations C-10 through C-12 gives:
Q ' Y ' n
c. Field Transport of Cotton in Trailers
The area-based emission factor for field transport of cotton
is derived from a distance-based emission factor developed
for field transport of grain.41 The basic formula for cal-
culating the field transport emission factor is:
D
ET '
where E = field transport emission factor based on area
harvested, mg/m2 or kg/km2 harvested
ETn = field transport emission factor based on dis-
tance traveled, mg/m
D = distance traveled in transporting one trailer
of harvested cotton from the field, m
A = area harvested in filling one cotton trailer, m2
In Equation C-14 the factor of two accounts for the fact that
the travel distance is covered twice - once in bringing the
empty trailer into the field and once in taking it out
loadedo The distance traveled, D, is the mean field trans-
port distance for a representative cotton harvesting operation,
and equals one-half the side of a representative harvesting
operation area: 443 m (see Section IV and Appendix D).
Equation C-12 is used to find A .
78
-------
The distance-based emission factor, E , is 5.0 ± 1.4 mg/m
(95% confidence level) at a vehicle speed of 4.47 m/s
(10 mph),1*1 considered the mean vehicle speed for field
transport of cotton.
Substituting for ET[), D, and AT gives the equation used to
calculate field transport emission factors:
fc (4,430 ± 1,240 mg) . Y ((,_15)
1 ^
The emission factors and base data used in deriving them are
presented in Table C-2. Average yields for picking and
stripping were calculated from the U.S. total production
(Tables 2 and 3) divided by the U.S. total area harvested
(Table 4) for the respective harvest methods. Other data
are from field observations and discussions with cotton
experts. The 95% confidence limits shown for the emission
factors are based on the confidence limits of emission rates
(for field transport, the confidence limits of the distance-
based emission factor). All quantities other than emission
rates were considered absolute.
The average emission factors for stripper harvesting in
Table C-2 are weighted averages based on the proportions of
stripper types. It is estimated that 2% of all strippers
are four-row models with mounted baskets, 59% are two-row
models with mounted baskets, and 39% are two-row models
pulling trailers (i.e., 60% of two-row strippers have mounted
baskets and 40% pull trailers).33 The confidence limits
associated with the weighted averages were calculated as:
79
-------
Table C-2. RESPIRABLE PARTICULATE EMISSION FACTORS AND
SUPPORTING DATA FOR COTTON HARVESTING OPERATIONS
Parameter
Q , mg/s
(I ,
v , m/s
r
EH, mg/m2 or
kq/km2 harvested
QL,'''e mg
Y, kg/m2
harvested
d
CT,!1 kg lint
cotton
E , ' mg/m2 or
kg/km2 harvested
E , mg/m2 or
kg/km2 harvested
E ,6 mg/m2 or
kg/km2 Harvested
Harvester type
Picker,
two- row,
with basket
1.24 ± 2.01C
1.34
2
0.455 ± 0.738
121
0.0630
6
654
0.0699
0.427 ± 0.119
0.952
stripper
Two- row
pulled trailer
33.4 ± 75.8
2.23
2
7.37 ± 16.7
0.0412
654
0.279 ± 0.078
7.65
Two- row,
with basket
10.4 ± 3.7
2.23
2
2.30 ± 0.82
243
0.0412
6
654
0.0918
0.279 ± 0.078
2.67
Four- row,
with basket
20.9 ± 7.4
2.23
4
2.31 ± 0.82
486
0.0412
3
654
0.0918
0.279 ± 0.078
2.68
Average
4.28 ± 6.53
0.0560
0.279 ± 0.078
4.61
Weighted average calculated as explained in text.
From Table C-l.
All confidence limits are for the 95% confidence
level (see text).
From field data and Reference 12.
Lack of data precludes assignment of
confidence limits.
From Tables 2,3, and 4.
From Reference 35.
Note: Blanks indicate not applicable.
80
-------
where (CL) = overall confidence limit for
stripping
f^, f2 , f2 = fractions of total strippers
' ' that are four-row models,
two-row models with baskets,
and two-row models pulling
trailers, respectively
(CL) , (CL)2 , (CL)2 = confidence limits for four-
' ' row strippers, two-row
strippers with baskets, and
two-row strippers pulling
trailers, respectively
Equation C-16 is not statistically correct,43 but it provides
a means for estimating overall confidence limits for the
weighted averages.
^Serth, R. W. (Monsanto Research Corporation, Dayton).
Error Propagation Formulas. Internal publication.
22 July 1975. 20 p.
81
-------
APPENDIX D
SOURCE SEVERITY CALCULATIONS
1. DEFINITION
Source severity is a measure of the air quality impact of
source emissions. Of particular interest is the maximum
severity from a representative source, defined earlier (in
Section IV) as:
S -=
where S = maximum severity from representative source
X = maximum time-averaged pollutant concentration
to which the public is exposed due to emis-
sions from the representative source, yg/m3
F = pollutant hazard factor, yg/m3
The hazard factor, F, is the short-term national primary
ambient air quality standard, when it exists. Only one
such standard applies to cotton harvesting: the 24-hr
standard for total suspended particulates, 260 yg/m3.21
For other pollutants, the hazard factor is the TLV adjusted
for time of exposure compared to a normal 8-hr workday, and
divided by 100 to account for the hazard potential to the
general public compared to that for industrial workers. The
representative cotton harvesting source operates a maximum
of 8 hr/day, so no time adjustment is necessary. Thus the
cotton harvesting hazard factors for pollutants other than
total suspended particulates are given by:
F = 0.01(TLV) (D-2)
82
-------
The concentration, x r in Equation D-l must be time-
ma x
averaged for the same period at the time base of the appli-
cable hazard factor. Since the unit operations in cotton
harvesting (harvesting, trailer loading, and field transport)
do not operate continuously, the time-averaged concentration
is determined by:
where (x ) ., = maximum average receptor concentration
max o , . .
during source operation
tQ = total time that source operates during
a day
t = time base of hazard factor, 24 hr for total
suspended particulates and 8 hr for other
pollutants
The concentrations during operation are calculated with
appropriate air pollution diffusion models.
2. CONCENTRATION CALCULATIONS
a. Harvesting and Transport
Harvesting machines and field transport of cotton are moving
point sources that make it difficult to model pollutant
concentrations at stationary receptors. However, with the
coordinate system attached to the source, it becomes a
stationary point source with a moving receptor. The integrated
cross wind concentration is used to model this case, starting
with the Gaussian point source model for ground level concen-
tration from a ground level source:36
x(x,y) = —y exp ^—\ (D-4)
•no o u
Y z
83
-------
where x(x»Y) = concentration at any point (x,y) in the
plume, g/ra3
x = distance from source on plume centerline, m
y = distance from plume centerline, parallel
to ground, m
Q = source emission rate, g/s
o,a = horizontal and vertical standard deviations,
^ respectively, of the plume concentration
distribution from Tables B-2 and B-3, m
u = average wind speed, assumed constant, m/s
To find the average ground level concentration on a line
parallel to the y-axis at a distance, x, from the source,
Equation D-4 is first integrated with respect to y across
the plume:
00 00
C f 0 /- 2 \
XA(x) = I x(x,y)dy = I ^—- expj^—J dy (D-5)
J J TTCJ a u \2a 2 I
—oo —oo y z \ y
XA(x) is called the crosswind integrated concentration, in
g/m2, and is the area under the concentration distribution
curve for ground level. In Equation D-5, Q, ir, and u are
constants, and a and a are functions of x only* Also, the
concentration distribution curve is symmetric, so the integral
in Equation D-5 is twice the area under one side of the curve.
Thus,
(D-6)
The integral in Equation D-6 is solved by integration in
the conplex plane. Substituting the value of the integral44
Standard Mathematical Tables, 14th Edition. Selby, S.M.
(ed.). Cleveland, The Chemical Rubber Co., 1965. p. 345,
84
-------
and simplifying,
— (D-7)
To find the average ground level concentration across the
plume, XA(x) i-s divided by the plume width. The above
derivation assumes that the plume is infinite in width with
asymptotic concentrations approaching zero at ± «°. In
reality, the plume has finite width. To estimate the width,
the plume edges are defined as those points between which
95% of the area under the concentration distribution curve
is contained. These points are at a distance ±1.96 a from
the plume axis, so the plume width is 3.92 a . Hence, the
average ground level concentration in the plume is
0.95 XA(X) 0.95 XA(X)
3.92ay
where x (x) = average ground level concentration at distance
" x from source, g/m3
w = plume width, m
or, substituting Equation D-7 into Equation D-8,
The maximum receptor concentration is defined at the
representative distance, D, from the source to the edge
of the representative harvesting operation. Evaluating o
and a for Class C stability (national average) at x = D
z
(for 100 m
-------
= X(D) = '
max o
max p n i . B 1
where (x ) = maximum ground level concentration in the
" plume at the edge of the representative
field, g/m3
This concentration must be time adjusted by the fraction
of time that the receptor is actually exposed to the plume.
The adjustment will be shown later in this appendix.
b. Trailer Loading
Emissions from cotton trailer loading can be modeled with
the Gaussian dose model for maximum ground level concen-
tration from a ground level source:36
QT
DL(X) = ±— (D-ll)
where D (x) = total dose at the plume centerline at
distance x from the source, g-s/m3
QT = total pollutant release, g
LI
a ., a , u = as defined in Equation D-4
The average concentration over the time of exposure to the
puff is estimated as:
DT (x)
where x (x) = average concentration at receptor during
exposure, g/m3
t = time of receptor exposure to the puff, min
86
-------
Substituting the representative distance, D, for x, evalu-
ating
D-12,
ating a and a as above, and combining Equations D-ll and
0.05 Q
pmax P
e max e i . e
D . t
e
where (x«)« = maximum receptor concentration during
6 lUaX
exposure
The time required for the harvester to dump one basket, tn,
13
is greater than the time of exposure, t , because of the
time consumed in raising and lowering the basket and other
related activities. Therefore, the average concentration
during the basket dumping operation is
(*B>max = (*e>max IT (D
D
where (XB) = average maximum receptor concentration
max during basket dumping
t_, = time consumed in dumping one basket
13
or, combining with Equation D-13,
0.05 Q
B
The further time correction that is necessary to account for
the fraction of time spent in performing the basket dumping
operation will be shown below.
87
-------
3. TIME AVERAGING
a. Harvesting and Trailer Loading
One harvester pass is completed each time the harvester
travels the length of one side of the representative field,
L. The number of harvester passes required to fill one
harvester basket is found by:
C/n
n =
where n = number of harvester passes required to fill
^ one harvester basket
C = capacity of one representative cotton trailer,
kg lint cotton
nR = number of baskets required to fill one cotton
trailer
L = length of one side of representative field, m
r = number of rows in harvester swath
wr = row spacing, m = 1.016 m
Y = lint cotton yield, kg lint cotton/m2 harvested
The time required for one harvester pass is:
tp =
where t = time for one harvester pass, min
tT = time to harvest one length, L, of field,
LI
mm
t. = time for harvester to turn after each pass,
min s?0. 3 mina
From field observation.
88
-------
and tT is calculated as:
LI
n
where vu = harvester speed while harvesting, m/s
n
The time to complete one cycle of harvesting and dumping
one harvester basket is:
where tl = time to fill and dump one harvester basket, min
tR = time to dump one harvester basket or change
trailers, min
The number of these cycles completed in one 8-hr work day is
The total harvester operating time in a day, (t0) , is
H = ni(n
and the total basket dumping time in a day, (tQ)B, is:
(t0)B = nlts (D-22)
The concentrations derived previously can now be time
averaged. For harvesting, the concentration in Equation D-10
is first multiplied by the fraction of time that the receptor
is exposed to the plume in one harvester pass, or:
89
-------
w /60 v
P
fp = t (D-23)
v
where w = plume width = 3.92 a , m
VH = harvester speed, m/s
The time averaged maximum concentration from harvesting is
calculated by combining Equations D-3, D-10, D-21, and D-23
1.82 Q / w /n,n
H
H
where ^Xmax)jT = time-averaged maximum receptor concentration
due to harvesting, g/m3
Q = harvester emission rate while operating, g/s
For basket dumping, the time-averaged maximum concentration
is calculated by combining Equations D-3, D-15, and D-22:
^0.05 Q \ n,
(xmax)B =1 T- (D-25)
max B \ . 8 ! k ] t
where (x )B = time-averaged maximum receptor concentration
max due to trailer loading, g/m3
Q = total pollutant release in dumping one
harvester basket, g
b. Field Transport
The nunber of cotton trailers filled in 1 day is:
As shown in Equation D-8. a calculated from Table B-2
at x = D. Y
90
-------
nl
nT = — (D-26)
B
where n = number of trailers filled in one day
nx = number of harvester baskets dumped in one day,
from Equation D-20
n = number of harvester baskets required to fill
one trailer
The fraction of time, f , that the receptor is exposed to
the plume in one transport pass is:
fT = (D-27)
where w = plume width = 3.92 a evaluated at x = D, m
D = representative transport distance = one-half of
the length of one side of the representative
field, m
The total time devoted to transport in 1 day is:
2 nm D nm D
(t°> -
where v^ = transport vehicle speed, m/s
The factor 2 in Equation D-28 accounts for the fact that two
passes are made for each trailer filled: one to pull the
empty trailer onto the field and one to pull the filled
trailer out. Combining Equations D-3, D-10, D-27, and D-28:
1.82 Q w n
(xmax)TR = TR P T (D-29)
max TR 3Q Di.8n» v t
91
-------
where (x )mn = time-averaged maximum receptor concentra-
max
field transport f g/m3
QTR = transport emission rate while operating, g/s
4. RESULTS
Source severities were calculated by combining Equations
D-24,, D-25, and D-29 with Equation D-l, using the supporting
equations presented and representative source parameters.
The calculations are presented in tabular form in Table D-l
for total suspended particulates (TSP) and inert dust (ID) .
Severities can be calculated for other pollutants by applying
their hazard factors to the appropriate 8-hr average particulate
concentrations in Table D-l, using the pollutant composition
fraction. Care must be taken to apply pollutant composition
fractions to the proper unit operations. For example, raw
cotton dust is 100% of the composition of harvester and
trailer loading emissions but is not present in transport
emissions, which are all soil dust. Another case is free
silica, where the only available measurement of composition
fraction is for total particulate from all harvesting opera-
tions .
Raw cotton dust severities, S , are presented in Table D-2.
£\L*U
Eight-hour average respirable particulate concentrations from
Table D-l were divided by the cotton dust hazard factor of
2 yg/n3 .
Free silica (S.02) severities are presented in Table D-3.
The 8-hr average respirable particulate concentrations from
Table D-l were divided by the respirable free silica hazard
factor, given by:22
= 100 yg/m3 (D
VO, (% S) + 2 (L>
92
-------
Table D-l. TABULAR SEVERITY CALCULATIONS FOR TOTAL
SUSPENDED PARTICULATES AND INERT DUST
Parameter3
Operating Basis
"B
r
Y, kg/m2
n
P
VH, m/s
t. , rain
L
t , rain
P
tfl, rain
,tj, rain
nl
Harvesting
QH, mg/sb
(*n.ax>H' "3/m3
24-hr average
8-hr average
TSPC
ID"
Trailer Loading
QL, mge
24-hr average
8-hr average
SL
TSP
ID
Field Transport
nT
QTR- rag/a
24-hr average
8-hr average
STR
TSP
ID
Total
STOTAL
TSP
ID
Type of harvester
Picker,
two-row
6
2
0.0630
0.961
1.34
11.0
11.3
2
12.9
37.2
1.24 1 2.01
0.00222
0.00665
8.54 x 10"6
1.65 x 10"5
121
0.00247
0.00742
9.50 x ID"6
7.42 x 10"s
6.20
22.4 i 6.3
0.00416
0.0125
1.60 x lO"5
1.25 x 10-"
3.40 x 10"5
2.66 x 10""
Stripper
Two- row,
pulled trailer
1
2
0.0412
8.82
2.23
6.62
6.92
5
66.0
7.27
33.4 t 75.8
0.0644
0.193
2.48 x 10""
1.93 x 10~3
7.27
22.4 t 6.3
0.00488
0.0146
1.88 x 10~5
1.46 x 10""
2.67 x 10-"
2.08 x 10~3
Two- row,
with basket
6
2
0.0412
1.47
2.23
6.62
6.92
2
12.2
39.3
10.4 ± 3.7
0.0181
0.0542
6.96 x 10~5
5:42 x 10""
243
0.00525
0.01575
2.02 x ID"5
1.58 x 10""
6.55
22.4 t 6.3
0.00440
0.0132
1.69 x 10~5
1.32 x 10""
1.07 x 10""
8.32 x 10""
Four-row,
with basket
3
4
0.0412
1.47
2.23
6.62
6.92
2
12.2
39.3
20.9 t 7.4
0.0363
0.1089
1.40 x 10""
1.09 x 10"3
486
0.0105
0.0315
4.04 x 10"5
3.15 x 10""
13.1
22.4 ± 6.3
0.00880
0.0264
3.38 x 10~5
2.64 x 10-"
2.14 x 10""
1.67 x 10"3
The following quantities are common to
calculations for all types of harvesters:
CT = t>:>4 Kg
L = 886 m
wr = 1.016 m
tt = 0.3 rain
D = 443 m
w = 3.92(0.2089) x
P D0.9031 = 201 m
U* "' = t>.JJ.B X 1U
VT - 4.47 m/s
(V8-hr = «° min
(V 24-hr = 1'"° min
FTSP = 26° "9/1"3
FJD = 100 pg/m3
With 95% confidence limits.
cTotal suspended particulates.
Inert dust.
Confidence limits not available.
Note: blanks indicate not applicable.
93
-------
Table D-2. RAW COTTON DUST SEVERITIES
Harvester type
Picker, two-row
Stripper
Two-row, trailer
Two-row, basket
Four-row, basket
Weighted average
Unit operation
Harvesting
0.00332
0.0965
0.0271
0.0544
0.0547
Trailer
loading
0.00371
0.00790
0.0158
0.00498
Total
0.00703
0.0965
0.0350
0.0702
0.0597
Based on 59% two-row with trailer, 39% two-row
with basket, 2% four-row with basket.
Note: Blanks indicate not applicable.
Table D-3. FREE SILICA SEVERITIES
Harvester type
Picker
Stripper
Two -row, trailer
Two -row, basket
Four-row, basket
Weighted average
Free silica
conteVit, mass %
7.9
2.3
_b
_b
_b
Hazard
factor,
vig/m3
10.1
23.3
_b
_b
_b
Source
severity
0.00263
a
0.00891
0.00357
0.00716
0.00572
Not applicable.
b ., ,_,
Not available
c
2% four-row with basket.
94
-------
SECTION IX
GLOSSARY OF TERMS
ABSCISSION - The process by which leaves detach from plants
by cell division or cleavage; defoliation.
AMERICAN-EGYPTIAN - A type of cotton grown in the U.S.,
predominantly in arid areas of the Southwest. It's lint is
dark cream or buff in color with strong fibers 38 to 44
millimeters in length. Also called American-Pima.
BALE - Product unit of cotton gins. A bale is 218 kilograms
(480 pounds) net weight of compressed lint cotton.
BASKET - Screen cage mounted on cotton harvester to collect
harvested cotton.
BASKET DUMPING - Operation of unloading harvested cotton from
a harvester basket into a cotton trailer.
BOLL - Capsule containing the seed and lint of a cotton plant;
the end result of ovary development.
BRACT - Small scalelike modified leaf growing at the base of
a flower, or in cotton, at the base of the boll.
BUR - Split wall of open boll.
BYSSINOSIS - Chronic bronchial disease caused by prolonged
exposure to raw cotton, flax, or hemp dust. Symptoms are
chest tightness and shortness of breath; symptoms are worst
upon renewed exposure after a few days of nonexposure. Bract,
stem, and leaf fragments believed to be source of causal
agents.
CONFIDENCE LEVEL - Probability that a value falls within
specified limits.
CONFIDENCE LIMITS - Plus-or-minus (±) value defining the
limits between which a value is expected to lie, with a spe-
cified confidence level.
95
-------
COTTON GIN - Facility which separates trash and seeds from
lint cotton and compresses it into bales.
COTTON-TYPE FARM - A farm with at least $10,000 of annual
cotton sales or at least 50% of all farm sales from cotton,
excluding farms with total annual sales under $2,500.
CYCLONE - Device used to separate particles from a gas stream
with centrifugal force.
DEFOLIANT - Chemical applied to plants to cause abscission.
DEFOLIATION - Process that induces abscission.
DESICCANT - Chemical applied to plants to cause desiccation.
DESICCATION - Process that kills leaves, permitting tissue
drying and inhibiting abscission.
DOFFER - Part of a mechanical cotton picker which removes
cotton from the picking spindles.
DOSE - Pollutant concentration multiplied by exposure time.
ELEVATOR - Mechanism of a cotton harvester which lifts har-
vested cotton and deposits it in a basket or trailer.
EMISSION BURDEN - Fraction of total annual mass of emissions
contributed by a specific source type, by state or nationwide.
FIELD TRANSPORT - The movement of empty cotton trailers into
a field and loaded trailers out of a field, within the con-
fines of the field.
FREE SILICA - As used in this document, crystalline silicon
dioxide as quartz.
GLEANER - Mechanical harvester used to salvage cotton from
the ground after mechanical cotton picking.
GRADE - Classification of the quality of cotton based on
color, leaf fragments and other foreign matter, and ginning
preparation.
HARVEST EFFICIENCY - Fraction of cotton in a field that a
harvester retrieves.
HARVEST RATE - Area harvested per unit time.
HAZARD FACTOR - Concentration used to denote the relative
toxicity of a pollutant.
96
-------
INERT DUST - Particulate matter with less than 1% quartz
that does not produce disease or toxic effects when exposure
is kept under reasonable control; also called nuisance dust.
LINT - Cotton fibers.
LINT COTTON - Harvested cotton from which seeds and trash
have been removed by a cotton gin.
LINTERS - Short fibers remaining on seeds after cotton
ginning.
LOCK - Distinct, separate tuft of cotton lint and seeds
developed in a locule formed by the two halves of adjacent
carpels in a boll. Bolls of most cultivated varieties of
cotton contain three to five locks.
PICKER - Mechanical cotton harvester which picks locks of
cotton from open bolls by means of rotating spindles that
poke into the plants perpendicular to the plant row.
POINT SOURCE - Pollutant emitter with a definable pollutant
outlet point.
RAW COTTON DUST - Dust from mechanical handling of cotton
which has had no chemical or physical treatment other than
mechanical cleaning.
RECEPTOR - As used in dispersion modeling, a hypothetical
sensor of pollutant concentration.
RESPIRABLE PARTICULATE - In this document, particles less
than 7 ym in diameter.
ROW SPACING - Centerline to centerline distance between
planted rows.
SEED COTTON - Harvested cotton prior to ginning.
SEVERITY - Ratio of pollutant concentration to a hazard
factor.
SILICOSIS - Chronic lung disease caused by prolonged expo-
sure to free silica dust.
SKIMMER - Device which removes particles from a gas stream
by passing the stream through a short radius IT radian (180 )
bend and "skimming" off the particles thrown to the outside
of the bend.
97
-------
SPINDLE - One of many barbed, toothed, or fluted and tapered
or straight spinning cylinders that a mechanical cotton
picker uses to pick locks from open bolls.
STAPLE - Fiber of cotton, wool, flax, etc., with reference
to length and fineness.
STRIPPER - Mechanical cotton harvester which strips bolls
from plants with pairs of rotating cylinders or stationary
fingers and slits.
TOTAL SUSPENDED PARTICULATE - Airborne particles collected
on a glass fiber filter with a high-volume air sampler.
TRAILER - Wagon with screen or slatted sides used to collect
harvested cotton and transport it from the field to a gin.
TRAILER LOADING - Process of dumping harvested cotton from a
harvester basket into a trailer.
TRASH - Plant fragments, soil dust, and other foreign matter
in harvested cotton.
UPLAND - Cotton species comprising more than 99% of all cotton
grown in the U.S. It's lint is almost pure white with fibers
19 to 38 millimeters long that adhere strongly to the seeds.
YIELD - Weight of lint cotton harvested per unit area.
98
-------
SECTION X
CONVERSION FACTORS AND METRIC PREFIXES
.it 5
CONVERSION FACTORS
To convert from
to
gram/kilogram (g/kg)
gram/meter3 (g/m3)
gram/second (g/s)
kilogram (kg)
kilogram (kg)
kilogram/kilometer2 (kg/km2)
kilogram/kilometer2 (kg/km2)
kilogram/metric ton
kilometer2 (km2)
kilometer2 (km2)
meter (m)
meter (m)
meter2 (m2)
meter2 (m2)
meter3 (m3)
meter/second (m/s)
metric ton
metric ton
metric ton
metric ton/kilometer2
milligram (mg)
r\
person/kilometerz
r\
person/kilometer^
radian (rad)
grain/pound
grain/foot3
pound/hour
bale (480-pound net weight)
pound-mass (Ib mass
avoirdupois)
pound/acre
pound/mile2
pound/ton
acre
mile2
foot
mile
acre
foot2
foot3
mile/hour
bale (480-pound net weight)
pound (mass)
ton
bale/acre
grain
person/acre
person/mile
degree (angle)
Multiply by
7.000
4.370 x 10~1
7.937
4.593 x 10~3
2.205
8.922 x
5.710
2.000
2.470 x
3.861 x
3.281
6.215 x
2.470 x
1.076 x
3.531 x
2.237
4.594
2.205 x
1.102
1.860 x
1.543 x
4.047 x
2.590
5.730 x
-3
10
102
10
,-1
10
10
-It
-It
101
101
103
io-2
io-2
10-3
IO1
'Metric Practice Guide. American Society for Testing and Materials.
Philadelphia. ASTM Designation: E 380-74. November 1974. 34 p.
99
-------
METRIC PREFIXES
Multiplication
Prefix Symbol factor Example
micro y 10~6 1 ym = 1 x 10~6 meter
milli m 10~3 1 mm = 1 x 10~3 meter
kilo k 103 1 kg = 1 x 103 grams
100
-------
SECTION XI
REFERENCES
1. Kipps, M. S. Production of Field Crops, 6th Edition.
New York, McGraw-Hill Book Company, 1970. p. 447-483.
2. Linton, G. E. Natural and Manmade Textile Fibers.
New York, Duell, Sloan and Pearce, 1966. p. 208-219.
3. Agricultural Statistics, 1973. Washington, U.S.
Department of Agriculture, Yearbook Statistical Committee
(U.S. Government Printing Office Stock Number 0100-
02841), 1973. p. 58-75.
4. Crop Production, 1974 Annual Summary. U.S. Department
of Agriculture, Statistical Reporting Service, Crop
Reporting Board. Washington. Publication No. CrPr
2-1(75). January 16, 1975. 64 p.
5. Colwick, R. F., et al. Mechanized Harvesting of Cotton.
U.S. Department of Agriculture, Agricultural Research
Service. Beltsville. Southern Cooperative Series,
Bulletin No. 100. March 1965. 70 p.
6. Voelkel, K. E. Texas Cotton Review, 1973-74. The
University of Texas at Austin, Natural Fibers Economic
Research. Austin. Research Report No. 104. July
1974. 143 p.
7. Census of Agriculture, 1969. Volume II, General Report.
Chapter 8, Type of Farm. Washington, U.S. Bureau of the
Census, 1973. 287 p.
8. Census of Agriculture, 1969. Volume V, Special Reports.
Part 3, Cotton. Washington, U.S. Bureau of the Census,
1973. 184 p.
9. Kelly, C. F. Mechanical Harvesting. Scientific
American. 217(2):50-59, August 1967.
101
-------
10. Elliott, F. C. Cotton Defoliation Guide for Texas.
Texas A&M Univeristy, Texas Agricultural Extension
Service. College Station. Bulletin No. L-145. 1969.
11. Addicott, F. T., and R. S. Lynch. Defoliation and
Desiccation: Harvest-Aid Practices. In: Advances in
Agronomy, Volume IX. New York, Academic Press Inc.,
1957. p. 69-93.
12. Personal communication. Dr. R. B. Metzer, Texas
Agricultural Extension Service, Texas A&M University,
College Station. October 24, 1975.
13. Peters, J. A., and T. R. Blackwood. Source Assessment:
Defoliation of Cotton, State of the Art. Monsanto
Research Corporation, EPA Contract 68-02-1874. Dayton.
Preliminary document submitted to the EPA, February
1976. 124 p.
14. Control and Disposal of Cotton-Ginning Wastes. U.S.
Public Health Service, National Air Pollution Control
Administration. Raleigh. Publication No. 999-AP-31.
1967. 103 p.
15. Personal communication. Elgin G. Fry, Office of
Pesticides Programs, U.S. Environmental Protection
Agency, Washington. October 29, 1975.
16. Elliott, F. C. Keep Cotton ... Dry-Loose-Clean.
Texas A&M University, Texas Agricultural Extension
Service. College Station. Publication No. MP-297. 8 p.
17. Burkhead, C. E., R. C. Max, R. B. Karnes, and E. Reid.
Field and Seed Crops - Usual Planting and Harvesting
Dates by States in Principal Producing Areas. U.S.
Department of Agriculture, Statistical Reporting Service.
Washington. Agriculture Handbook No. 283. 1972. p. 10-12
18. Census of Agriculture, 1968. Volume V, Special Reports.
Part 15, Graphic Summary. Washington, U.S. Bureau of
the Census, 1973. p. 125.
19. Caudill, C. E., P. M. Williamson, M. D. Humphrey, Jr.,
L. P. Garrett, and L. Canion. 1973 Texas Cotton Statis-
tics. Texas Crop and Livestock Reporting Service, Texas
Department of Agriculture. Austin. Bulletin 113.
June 1974. 21 p.
20. 1970 Census & Areas of Counties and States. In: 1975
World Almanac & Book of Facts. New York, Newspaper
Enterprise Association, Inc., November 1974. p. 183-201.
102
-------
21. Code of Federal Regulations, Title 42 - Public Health,
Chapter IV - Environmental Protection Agency, Part 410 -
National Primary and Secondary Ambient Air Quality
Standards, April 28, 1971. 16 p.
22. TLV's® Threshold Limit Values for Chemical Substances
and Physical Agents in the Workroom Environment with
Intended Changes for 1975. American Conference of
Governmental Industrial Hygienists. Cincinnati. 1975.
97 p.
23. Kilburn, K. H., G. G. Kilburn, and J. A. Merchant.
Byssinosis: Matter from Lint to Lungs. American
Journal of Nursing. 73:1952-1956, November 1973.
24. Merchant, J. A., J. C. Lumsden, K. H. Kilburn, V. H.
Germino, J. D. Hamilton, W. S. Lynn, H. Byrd, and
D. Baucom. Preprocessing Cotton to Prevent Byssinosis.
British Journal of Industrial Medicine. 30:237-242, 1973,
25. Hamilton, J. D., G. M. Halprin, K. H. Kilburn, J. A.
Merchant, and J. R. Ujda. Differential Aerosol
Challenge Studies in Byssinosis. Archives of Environ-
mental Health. 26:120-124, March 1973.
26. Neefus, J. D. Cotton Dust Sampling: I Short Termed
Sampling. American Industrial Hygiene Association
Journal. 36:470-476, June 1975.
27. Rawlings, G. D., and R. B. Reznik. Source Assessment:
Cotton Gins. Monsanto Research Corporation, EPA
Contract 68-02-1874. Dayton. Preliminary document
submitted to the EPA, December 1975. 97 p.
28. Andrilenas, P. A. Farmers' Use of Pesticides in 1971...
Quantities. U.S. Department of Agriculture, Economic
Research Service. Washington. Publication No. ERS-536.
February 1974. 35 p.
29. Feairheller, W. R., , and D. L. Harris. Particulate
Emission Measurements from Cotton Gins, Delta and Pine
Land Co., Scott, Mississippi. Monsanto Research
Corporation. Dayton. Report No. MRC-DA-358. Environ-
mental Protection Agency, EMB Project Report No. 72-MM-
16. November 1974. 239 p.
30. Feairheller, W. R., and D. L. Harris. Particulate Emis-
sion Measurements from Cotton Gins, Bleckley Farm Service
Co., Cochran, Georgia. Monsanto Research Corporation.
Dayton. Report No. MRC-DA-357. Environmental Protection
Agency, EMB Project Report No. 72-MM-23. November 1974.
265 p.
103
-------
31. Emissions from Cotton Gin at Valley Gin Company, Peoria,
Arizona. PEDCo-Environmental Specialists, Inc., Cincinnati.
Environmental Protection Agency, EMB Project Report No.
72-MM-20. 1973. 37 p. plus Appendix.
32. Durrenberger, C. Cotton Gin Report. Texas Air Control
Board. Austin. May 31, 1974. 50 p.
33. Personal communication. Dr. R. B. Metzer, Texas
Agricultural Extension Service, Texas A&M University,
College Station. January 13, 1976.
34. 1972 National Emissions Report. U.S. Environmental
Protection Agency. Research Triangle Park. Publication
No. EPA-450/2-74-012. June 1974. 422 p.
35. Personal communication. Dr. R. B. Metzer, Texas
Agricultural Extension Service, Texas A&M University,
College Station. December 16, 1975.
36. Turner, D. B. Workbook of Atmospheric Dispersion
Estimates, Revised 1970. U.S. Environmental Protection
Agency, Office of Air Programs. Publication No. AP-26.
July 1971. 84 p.
37. 1973 Handbook of Agricultural Charts. Agricultural
Handbook No. 455. Washington, U.S. Department of
Agriculture, October 1973. p. 121.
38. Blackwood, T. R., T. F. Boyle, T. L. Peltier, E. C.
Eimutis, and D. L. Zanders. Fugitive Dust from Mining
Operations. Monsanto Research Corporation. Dayton.
Report No. MRC-DA-442. (EPA Contract 68-02-1320,
Task 6.) May 1975. p. 34.
39. Eimutis, E. C., and M. G. Konicek. Derivations of
Continuous Functions for the Lateral and Vertical
Atmospheric Dispersion Coefficients. Atmospheric
Environment. 6:859-863, November 1972.
40. Martin, D. V., and J. A. Tikvart. A General Atmospheric
Diffusion Model for Estimating the Effects on Air Quality
of One or More Sources. Presented at 61st Annual Meeting
of the Air Pollution Control Association, for NAPCA,
St. Paul, 1968. 18 p.
41. Wachter, R. A., and T. R. Blackwood. Source Assessment:
Harvesting of Grain, State of the Art. Monsanto Research
Corporation, EPA Contract 68-02-1874. Dayton. Prelimi-
nary document submitted to the EPA, December 1975. 81 p.
42. Personal communication. Mr. Dan Pustejovsky, Hillsboro,
Texas. October 1, 1975.
104
-------
43. Serth, R. W. (Monsanto Research Corporation, Dayton),
Error Propagation Formulas. Internal publication.
22 July 1975. 20 p.
44. Standard Mathematical Tables, 14th Edition. Selby,
S.M. (ed.). Cleveland, The Chemical Rubber Co., 1965,
p. 345.
45. Metric Practice Guide. American Society for Testing
and Materials. Philadelphia. ASTM Designation:
E 380-74. November 1974. 34 p.
105
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. DEPORT NO.
EPA- 600/2- 77 -107d
2.
3. RECIPIENTS ACCESSION NO.
4. T.TLE AND SUBTITLE SOURCE ASSESSMENT: Mechanical
Harvesting of Cotton--State of the Art
5. REPORT DATE
July 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J. W. Snyder and T.R. Blackwood
8. PERFORMING ORGANIZATION REPORT NO.
MRC-DA-684
9. PERFORMING ORGAN ZATION NAME AND ADDRESS
Monsanto Research Corporation
1515 Nicholas Road
Dayton, Ohio 45407
10. PROGRAM ELEMENT NO.
1AB015; ROAP 21AXM-071
11. CONTRACT/GRANT NO.
68-02-1874
12. SPONSORING AGENCV NAME AND ADDRESS
'EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 2/75-4/76
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES jgRL-RTP project officer for this report is Dale A. Denny, Mail
Drop 61, 919/541-2547.Similar previous reports are in the EPA-600/2-76-03Z series.
16. ABSTRACT
The report summarizes reported data on air emissions from the mechanical
harvesting of cotton, including the machine removal and collection of seed cotton from
mature plants and the transport of this cotton from the field. Machine harvesting and
field transport cause air pollution in the form of respirable dust, from soil and raw
cotton, and agricultural chemicals contained in the cotton dust. Mechanical cotton
harvesting accounted for 0.002% of the national respirable particulate emissions in
1972. Highest state contributions were 0.046% in Texas and 0.025% in Oklahoma. The
air quality impact of cotton harvesting emissions was assessed in terms of source
severity (the ratio of the time-averaged maximum pollutant concentration from defined
representative picking and stripping sources to the primary ambient air quality stan-
dard for particulate, or to a corrected TLV for noncriteria pollutants). The highest
source severity was for raw cotton dust: 0.00703 for picking, and 0.035 for stripping.
Severities for other pollutants were less than 0.001. No emission control technology
is applied to cotton harvesting, and no voluntary future application is expected.
Avesrage annual harvested area is expected to remain constant.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Assessments
Cotton Plants
Cottonseed
Dusi:
Harvesting
Soils
Agricultural Chem-
istry
Agricultural Machi-
nery
Air Pollution Control
Source Assessment
Raw Cotton
Cotton
Particulate
Source Severity
13B 08M,08G
14B
02D,06C 02A
11G
02C
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport/
Unclassified
21. NO. OF PAGES
118
20. SECURITY CLASS (Thispage\
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
106
------- |