AIR-BORNE PARTICULATE EMISSIONS FROM COTTON GINNING OPERATIONS waaam •V.V.V.V The Robert A. Taft Sanitary Engineering Center TECHNICAL REPORT A60-5 U. S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE ------- AIR-BORNE PARTICULATE EMISSIONS FROM COTTON GINNING OPERATIONS Prepared by Engineering Research and Development Unit Air Pollution Engineering Research U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service Bureau of State Services Division of Engineering Services Robert A. Taft Sanitary Engineering Center Cincinnati, Ohio ------- INTRODUCTION Modern cotton ginning installations using pneumatic conveying equipment, air blast clean- ing equipment, and seed cotton conditioning equipment have increased the output of normal ginning operations and improved cotton fiber quality. This gain in output has been accom- panied by a major increase in the volume of air bearing lint and dust discharged to the atmos- phere. The increased proportion of machine- picked cotton supplied to ginning operations has further increased the volume of waste ma- terials. Present day ginning practices may produce up to 1, 000 pounds of waste material for each bale of cotton ginned, a significant por- tion of which will be discharged to the atmos- phere. Although the discharge of waste material to the atmosphere from the individual cotton gin- ning operation generally results in only a local air pollution problem, the number of such es- tablishments and the gradual urban encroach- ment into the areas of ginning operations is great enough to warrant consideration of suit- able control methods. Figures 1 and 2 indicate the pollution effects and the pollutants involved in the problem. In order to consider possible solutions to these problems, field and labora- tory studies of particulate emissions from cot- ton gins were conducted by the Public Health Service, in cooperation with the Agricultural Research Service. -tLl •v4 -A1 Figure 1. LOCALIZED POLLUTION PROBLEM Figure 2. EXAMPLES OF GIN DISCHARGES ------- 2 AIR-BORNE PARTICULATE EMISSIONS FROM COTTON GINNING OPERATIONS OBJECTIVES The objectives of these studies were three- fold: first, to assess the quantity of particulate emissions from cotton ginning operations in terms of its air pollution potential; second, to evaluate the effectiveness of standard laboratory quality evaluation test equipment for predicting these emission values; and third, to develop an economically feasible control technique and suggest control equipment for the reduction of particulate emissions to the atmosphere. From the standpoint of the air pollution po- tential of cotton ginning particulate emissions, it is necessary to differentiate portions of the emissions by particle size. Those which are of sufficiently small size to carry beyond the prem- ises of the ginning operations under average weather conditions constitute a local air pollu- tion problem. That portion having a large enough particle size to deposit on the gin prem- ises is a lesser problem. Based on an average wind speed of 10 miles per hour and for par- ticles with a density equal to the average den- sity of particulate emissions from ginning oper- ations, it is assumed in this report that parti- cles 100 microns in diameter and larger will deposit on the premises and that particles smaller than 100 microns will carry beyond the gin premises to the surrounding community. BACKGROUND Field sampling of cotton gin effluent and lab- oratory investigation of procedures for the eval- uation of air pollution materials discharged to the atmosphere from the ginning of cotton were conducted at the Cotton Ginning Laboratory, Agricultural Research Service, U.S. Depart- ment of Agriculture, Stoneville, Mississippi. The investigation was made at this laboratory, since this operation was considered to be repre- sentative of ginning operations in general, and the cotton harvested by both machine-and hand- picked methods and processed during testing was representative of the cotton from the areas in question. The flow diagram, Figure 3, indicates the following major sources of particulate emis- sions: unloading fan, six cylinder cleaner, stick and bur machine, gin stand, separator No. 2, seven cylinder cleaner, separator No. 3, and the condenser. The unloading fan, discharge point No. 1, supplies the air for the transfer of cotton from the storage bins or from a wagon to the first separator. These wastes are carried to the dust house by the moving air stream and consist mostly of sand, dirt, and other fine materials. The cotton then passes from the first sepa- rator onto the feed control, into the tower dryer, through a boll trap, and then to the six cylinder cleaner which opens and cleans the boll cotton. The waste discharge from the six cylinder cleaner, discharge point No. 2, is carried to the dust house by a moving air stream and con- sists of fine particles of leaf trash, dirt, sand, stems, and small sticks. (See Glossary) From the cleaner the cotton is moved to the stick and bur machine which removes burs, sticks and stems, together with fine trash not removed by the cylinder cleaner. The discharge duct from the stick and bur machine joins with the air discharge duct from the gin stand and wastes from these sources are carried to the dust house, designated as discharge point No. 3. Samples were taken from the combined dis- charges because individual samples from these two emission sources were unobtainable due to the inaccessibility of the separate ducts. From the stick and bur machine the cotton passes to a second separator, then to a stub tower dryer, and then to a seven cylinder clean- er. The discharge from the second separator at discharge point No. 4 is emitted directly to the atmosphere outside the building. These wastes, carried through the separator, con- sisted mainly of fine particles of leaf trash, ------- Field Sampling Equipment and Techniques 3 NO. Z DISCHARGE TO OUTSiOC pr. « WAGON NO. 40 90AR0MAN e&H SEVEN CYLINDER 'no «SPhElPSFAN © 1925 RPM ' DISCHARGE TO' CONDENSER DISCHARGE 'TO DUST MOUSE DISCHARGE COMBINED NO. 35 CLARAGE FIGURE 3. FLOW DIAGRAM OF U. S, DEPARTMENT OF AGRICULTURE COTTON GIN, STONEVILLE, MISSISSIPPI The waste discharge from the seven cylinder cleaner, discharge point No. 5, is carried to cyclones. Tests include the use of either the standard or the alternate (parallel) cyclones. These wastes consist mainly of pin and pepper trash and dirt. The seed cotton then passes from the seven cylinder cleaner on a belt distributor to the extractor-feeder and then to the huller front gin stand. The trash from the gin stand, burs, sticks, stems, motes, and pin and pepper trash is combined, as previously noted, with the waste discharge from the stick and bur machine and blown to the dust house, discharge point No. 3. From the gin stand the cotton is transferred first to a separator which removes fine leaf particles, motes, dust, and sticks, which are discharged directly to the atmosphere outside the building, discharge point No. 6. The cotton next travels to the lint cleaner and then to the condenser, which discharges pin and pepper trash to the dust house, discharge point No. 7, then to the baler, and out as a finished product. Field Sampling Equipment and Techniques Initially it was intended to take simultaneous samples at each discharge point to differentiate between total particulate and air pollution par- ticulate emissions to the atmosphere, i. e., particles less than 100 microns in size. Dupli- cate sampling trains were to be used, with one train preceded by a settling chamber to remove the particulate 100 microns or over in diameter. Due to weather conditions, insufficient cotton was on hand at the Stoneville Laboratory to allow sampling by both techniques at every point of discharge as originally planned. To further conserve cotton, samples were taken simulta- ------- 4 AIR-BORNE PARTICULATE EMISSIONS FROM COTTON GINNING OPERATIONS inary evaluations were made at each discharge point to determine those locations containing non-air-pollution particulate, and the settling chamber, Figure 4, was employed at only those emission points. The sampling system, Figure 5, consisted of a sampling probe, a high effi- ciency cyclone used during high loadings only, a glass-fiber filter, a pump, an orifice flow meter, an integrating bellows gas meter, and various wet- and dry-bulb thermometers. All points of discharge to the atmosphere were sampled. Whenever possible, sampling positions were selected along straight runs of duct at least 12 pipe diameters in length, and samples were taken at a distance equal to 8 pipe diameters(l) from the start of the run. Where it was necessary to sample ducts not having 12 pipe diameters of straight flow, sam- ples were obtained at a point two-thirds of the longest straight flow distance downstream from a point of gas flow disturbance such as a bend, damper or valve. Four to sixteen point sampling traverses were run, when possible, at each sampling point using isokinetic sampling procedures. (1) When the settling chamber was used, the probe was fixed at a point in the duct with a velocity equal to the average velocity through the duct, as it was impossible, because of the bulky na- ture of the equipment, to move the probe for sampling traverses. GAS EXIT GAS- INLET I. SAMPLING PROBE 2.SETTLING CHAMBER 3. PROBE CONNECTED TO SAMPLING SYSTEM NO. 2 4. BLOWER 5. MOTOR 6. FLOW VALVE 7. PITOT TUBE 8. INCLINED MANOMETER FIGURE 4. SETTLING CHAMBER USED IN CONJUNCTION WITH SAMPLING SYSTEM NO. 2 (1) "Methods for Determination of Velocity, Volume, Dust and Mist Content of Gases", Bulletin ------- Emission from Ginning Operations 5 GAS EXIT GAS INLET 8a 8b 1. SAMPLING PROBE 2. HIGH EFFICIENCY CYCLONE 3. FIBER-GLASS FILTER 6 HOLDER 4. PUMP 5. PUMP BYPASS VALVE S LINE 6. DRY BULB THERMOMETER 7. ORIFICE 8a MANOMETER TO REGISTER 8b. AP ACROSS ORIFICE SURGE BOTTLE 9. BELLOWS GAS METER 10. DRY BULB THERMOMETER I I. WET BULB THERMOMETER FIGURE 5. SAMPLING SYSTEM NO. 2 Emissions from Ginning Operations The data on atmospheric emissions, Table 1, are limited in that replicate samples were ob- tained during ginning of both hand- and machine- picked cotton for only the cyclone discharges. There is, therefore, no indication of the varia- tions at the other points of discharge due to the method of picking. Table 2 presents data on the amount of material collected by the settling chamber and the sampling train, expressed in percent by weight of cotton ginned. Table 3 contains calculated particle discharges to the atmosphere for the total emissions and for the air pollution portion of the emissions, i. e. , that portion of the discharged particulates less than 100 microns in diameter. The evaluation of the particulate emission data from the ginning operation was made, as previously discussed, on two bases: first, on a consideration of the total particulate emis- sion; and, second, on a consideration of only that portion which constitutes a local air pol- lution problem, i. e. , particles of less than 100 microns in diameter. The data in Table 3 show that the calculated air pollution portion of the total particulate dis- charge (i.e. , particles less than 100 microns in diameter) from the gin to the atmosphere with no air pollution control devices utilized was approximately 7. 0 pounds per bale (1. 39%) of cotton ginned. However, the gin had in- stalled two types of cyclones on the discharge from the seven cylinder cleaner. One was an 84" diameter standard cyclone with tapered top and bottom sections and the other a pair of 34" diameter high efficiency design operating in ------- TABLE 1 SUMMARY OF FIELD DATA U. S. DEPARTMENT OF AGRICULTURE COTTON GINNING LABORATORY STONEVILLE, MISSISSIPPI Sample Point Description Unloading Fan Pipe Size 16" D Average Velocity FPM 4680 Volume SCFM 6280 Type Cotton, Picking Method Machine Sam Settling Chamber GR/SCF pies Collected Sampling Train GR/SCF 0. 359 Total GR/SCF 0. 359 Average Emission Rate lbs/hr 19.32 Six Cylinder Cleaner 16" D 3150 3878 Hand 0. 078 0. 037 0. 115 3.82 Stick & Bur Ma- chine Discharge Combined With Huller Front & Mote Discharge from Gin Stand 11" D 4320 2783 Hand 0. 524 0. 026 0. 550 13.12 No. 2 Separator 12" D 4450 3260 Machine 0. 065 0. 065 1.82 Seven Cylinder Cleaner to Cyclones Standard Cyclone Inlet* 18" D 4340 7030 Hand Machine 0.0957 0.0957 0.0124 0.0103 0.1081 0.1060 6.51 6.39 Standard Cyclone Outlet* 19-1/4" D 3890 7030 Hand Machine 0.0070 0.0150 0.0070 0.0150 0.42 0.90 Alternate Cy- clone Inlet* 18" D 4180 6731 Hand Machine 0.0347 0.0346 0.0186 0.0495 0.0533 0.0841 3.08 4.85 Alternate Cy- clone Outlet* 14" D 3680 6731 Hand Machine 0.0019 0.0054 0.0019 0.0054 0.11 0.31 Separator No. 3 21" D 1610 3865 Hand 0.0440 0.0440 Condenser 42" D 1240 11523 Hand 0.0223 0.0223 1.46 2.20 ------- Emissions from Ginning Operations TABLE 2. EFFLUENT SUMMARY . (Percent by Weight of Cotton Ginnedf Sample Point No. Sample Point Description Settling Chamber (S) Sampling Train (A) Total (T) Unloading Fan 0.966 0.966 Six Cylinder Cleaner 0.130 0.062 0.192 Stick & Bur Machine Dis- charge Combined With Hul- ler Front & Mote Discharge From Gin Stand 0.625 0.031 0.656 Separator No. 2 0.091 0.091 Seven Cylinder Cleaner To Cyclones 'Buffalo" Cyclone Inlet 0.289 0.034 0.323 'Buffalo" Cyclone Outlet 0.033 0.033 "A.E.C." Cyclone Inlet 0.100 0.097 0.197 "A.E.C." Cyclone Outlet 0.010 0.010 Separator No. 3 0.073 0.073 Condenser 0.106 0.106 TABLE 3. ATMOSPHERIC PARTICLE EMISSIONS Description of Condition Fractions Under Conditions* Totals % by Weight of Cotton Ginned Pounds per bale A. Total Discharge, No Control T +T +T +T +T +T +T +T 1 2 3 4 5A 5C 6 7 2 2.34% 11.7 B. Total Discharge, Installed Cyclones in Use T1+T2+T3+T4+T5B+T5D+T6+T7 2 2.11% 10.5 C. Total Discharge of Air Pollution Material (Cyclones not in Use) A1+WA4+A5A+A5C+VA7 2 1.39% 7.0 D. Total Discharge of Air Pollution Material (Cyclones in Use) A1+A2+A3+a4+A5b+A5d+A6+A7 2 1.35% 6.8 ------- 8 AIR-BORNE PARTICULATE EMISSIONS FROM COTTON GINNING OPERATIONS efficiencies from 85 to 93% and 95 to 96% respec- tively, but were applied only to the discharge from the seven cylinder cleaner, the average air pollution discharge to the atmosphere, when these cyclones were being utilized, was 6. 8 pounds per bale (1. 35%) of cotton ginned. The total particulate discharged to the atmosphere from the ginning operation, when no control equipment was used, was approximately 11. 7 pounds per bale (2. 34%) of the cotton processed. With the cyclones in use for the seven cylinder cleaner, the total particulate discharge was 10. 5 pounds per bale (2. 11%) of the cotton ginned. Since the gin sampled was selected as typical of such operations, these values may be assumed to show generally the order of dis- charges from cotton gins. Laboratory Quality Evaluation Test Equipment Two standard laboratory quality evaluation test devices were considered as having potential application in predicting air pollution particu- late emissions from the cotton ginning opera- tion. The first of these was the fractionator, Figures 6 and 7. This equipment, normally used to determine the quantity of foreign ma- terial contained in cotton, consists essentially of a sheet metal box-like container in which the cotton is agitated by air jets. In the operation of this device, a sample of non-ginned cotton is prepared for fractionation by hand removal of unopened bolls, seeds, and trash. Weights of cotton and removed trash are obtained. The prepared cotton is then placed in the fractiona- tor and a compressed air supply, connected to inlets at the bottom rear of the fractionator, is turned on. The cotton is lifted by the air stream, follows the curved top of the fractionator, drops onto the air discharge slots in the bottom front of the fractionator, and is carried across the bottom to be once again picked up by the air jets, repeating the cycle. The agitation of the cotton by the air stream is continued for a period of ten minutes. As the cotton is churned by the air stream, the foreign material is dislodged and is carried along with the fractionator effluent. The test equipment used to determine the waste material discharged from the fractionator sampled the entire air flow of the fractionator, and consist- ed of a settling chamber, a cyclone, and a glass fiber filter which was followed by a high volume blower to eliminate excessive back pressure on the fractionator. After fractionation, the cotton and the fractions collected by the sampling sys- tem were weighed to determine the amounts of foreign material removed from the cotton. Figure 6. EXTERIOR VIEW Figure 7. INTERIOR VIEW ------- Laboratory Quality Evaluation Test Equipment 9 A second standard laboratory quality evalua- tion test device used to predict the amount of air pollution particulate emission from cotton gin operations was the Shirley Analyzer, Figure 8. Basically, this equipment is used to deter- mine the amount of foreign material remaining in the clean lint after ginning. Its principle of operation is similar to that of a lint cleaner in that it uses a similar saw arrangement for the removal of foreign material. It is more effi- cient than lint cleaners in use commercially, but causes a greater loss of usable fibers. As a method for collecting the fine fraction of the foreign material removed from a cotton sample by the Shirley Analyzer was not avail- able, it was necessary to determine this value by a material balance; that is, the weight of cotton placed in the analyzer must equal the weight of the cleaned sample plus the collected waste material plus the fine fraction quite ap- propriately designated as noncollected waste. Table 4 contains fractionator sampling train data for cotton obtained from the Stoneville, Mississippi, area. Table 5 contains similar Figure 8. SHIRLEY ANALYZER fractionation data on cotton from the Clemson, South Carolina, area. Data obtained from the tests performed by the Shirley Analyzer for cotton from the Stoneville, Mississippi, area are presented in Table 6. TABLE 4. FRACTIONATION DATA ON 20 SEED COTTON SAMPLES FROM THE STONEVILLE, MISSISSIPPI AREA Sample Number Large Foreign Material Removed by Hand Before Fractionation Seed Cotton After Frac- tionation Settling Chamber Catch Cyclone Catch Material Collected On Filter Wet Dry Wet Dry gms gms gms gms gms gms gms Hand-Picked 1 2 3 4 5 6 7 8 9 10 0.2 0.3 0.4 0.6 0.2 0.8 0.2 0. 5 0. 6 0.8 179. 0 206. 0 176. 0 166.0 170.0 171.0 159.0 171.0 170.0 172. 0 1. 09 .86 2. 00 .77 ,85 5.32 4.92 6.88 5.35 5.01 .95 .75 1.86 .70 .75 4.95 4.60 6.45 5.05 4.70 1.19 1.82 1.44 1. 18 1. 39 1.47 1.72 1.11 2. 15 1.43 1.07 1.73 1.36 1. 10 1.30 1.42 1.61 1.06 2.02 1.37 0. 083 0. 096 0. 099 0.055 0. 064 0. 083 0. 091 0. 075 0. 156 0. 088 Machine-Picked 11 12 13 14 15 16 17 18 19 20 5.6 8.1 5.6 4.6 4. 6 9.6 7.7 7. 0 10.3 6.2 169. 0 140.0 172.0 158. 0 154.0 159.0 146.0 168.0 197. 0 150. 0 10.83 6. 00 9. 38 7.74 7.89 9.05 9.61 12.99 13. 59 9.75 10. 00 5.33 8.43 6.97 7. 19 8.23 8.65 11.81 12.34 8.86 .90 1.45 1.85 1.82 1.55 1. 55 1.24 1.43 2.10 1.20 .83 1.32 1.65 1.66 1.43 1.43 1.08 1. 30 1.96 1.10 0. 086 0. 095 0. 130 0. 083 0. 092 0. 252 0. 125 0. 155 0.220 ------- TABLE 5 FRACTIONATION DATA ON 20 SEED COTTON SAMPLES FROM THE CLEMSON, SOUTH CAROLINA AREA Sample Large Foreign Material Removed by Hand Before Fractionation Seed Cotton After Settling Chamber Catch Cyclone Catch Material Collected On Filter Number Fractionation Wet Dry Wet Dry gms gms gms gms gms gms gms i Cd O » % H HH O ci f > H W M CO Cfl ft O 3 3 o o o -3 H O 52 a z a o M ss > H o 52 w Hand- Picked 21 22.4 186.0 12.38 11.03 1.81 1.40 0.126 22 17.8 169.0 8.78 8.44 .96 .90 0.113 23 19.1 158.0 12.57 12.25 1.38 1.30 0.149 24 28.9 144.0 11.70 11.25 1.43 1.34 0.150 25 36.1 122.0 10.63 10.22 1.00 .96 0.118 26 30.0 170.0 14.23 13.76 1.21 1.14 0.175 27 29.2 176.0 15.37 15.00 1.34 1.31 0.154 28 35.6 184.0 13.08 12.70 1.15 1.08 0.138 29 18.8 150.0 13.09 12.75 1.39 1.31 0.133 30 26.4 170.0 15.23 14.81 1.00 .94 0.136 Machine-Picked 31 5.0 166.0 4.11 3.78 .95 .86 0.080 32 5.4 153.0 5.64 5.18 .96 .87 0.061 33 2.0 196.0 6.56 6.10 .70 .63 0.063 34 5.9 141.0 5.99 5.43 .91 .82 0.096 35 3.7 159.0 7.74 7.23 .66 .59 0.072 36 10.4 169.0 9.81 9.16 .78 .70 0.062 37 3.2 162.0 6.99 6.43 .89 .81 0.079 38 11.9 156.0 8.80 8.11 .82 .75 0.083 39 13.2 134.0 17.41 16.30 .80 .73 0.068 40 13.9 145.0 15.63 14.59 1.22 1.12 ------- Laboratory Quality Evaluation Test Equipment 11 TABLE 6 SHIRLEY ANALYZER AND DRYING DATA ON 19 SAMPLES OF LINT COTTON FROM THE STONEVILLE AREA Sample Classi- fication Wet Sample of Lint Fed Dried Sample of Lint Fed Wet Samples Dried Samples Noncollected Waste Cleaned Lint Collected Waste Cleaned Lint Collected Waste gms gms gms gms gms gms % by Weight of Dried Cotton* Hand-Picked Without Lint Cleaner 1 100.00 92.70 94.40 2.34 89.23 2.13 1.446 2 100.00 93.17 94.53 2.60 89.29 2.37 1.621 3 100.00 92.73 94.82 2.46 90.28 2.14 0.334 4 100.00 94.17 94.88 2.54 90.74 2.25 1.253 5 100.00 92.13 94.42 2.56 89.04 2.31 0.847 Hand-Picked With Lint Cleaner 6 100.00 93.04 96.28 1.32 91.51 1.18 0.376 7 100.00 93.62 96.14 1.28 91.78 1. 14 0.748 8 100.00 92.78 96.64 .87 91.83 .83 0.129 9 100.00 93.24 96.65 1.03 91.38 .97 0.955 Machine-Picked Without Lint Cleaner 10 100.00 94.51 93.76 4.90 88.62 4.29 1.693 11 100.00 94.55 91.73 6.33 86.86 5.53 2.285 12 100.00 94.28 94.67 3.88 89.34 3.45 1.580 13 100.00 94.61 95.00 3.63 89.71 3.21 1.786 14 100.00 94.04 93.14 4.96 88.81 4.31 0.978 Machine-Picked With Lint Cleaner 15 100.00 94.26 94.19 4.27 89.65 3.72 0.944 16 100.00 94.48 93.28 4.91 88,83 4.30 1.429 17 100.00 94.02 94.37 3.92 89.50 3.60 0.979 18 100.00 94.50 94.71 3.26 89.58 2.90 2.138 19 100.00 94.44 94.48 3.68 90.13 3.31 1.059 ------- 12 AIR-BORNE PARTICULATE EMISSIONS FROM COTTON GINNING OPERATIONS Applicability of Fractionator as Air Pollution Indicator A statistical evaluation (Appendix 1) of the fractionator discharge data, Tables 4 and 5, indicates it is not possible to separate the ef- fects of the picking method from the effects of the area from which the cotton was obtained. It is, therefore, not possible, based on fraction- ator data, to make a general statement that one picking method results in less air pollution ma- terial than the other, independent of the area from which the cotton was obtained. By a com- parison of the mean percent air pollution ma- terial for each picking method and each area separately, it was observed that the Clemson area machine-picked cotton yielded significantly less air pollution material than did hand- picked cotton from the Clemson area, and both hand- and machine-picked cotton from the Stoneville area. The latter three did not differ significantly from each other. Based on the fractionator data for cotton from the Stoneville area, the ranges of expected amounts of air pollution material from individual cotton samples during fractionation were es- timated using statistical techniques. (2) The limits of the estimated ranges are such that the probability is less than one in twenty that not more than 5% of the individual samples will fall beyond these limits. All data were used in com- puting these limits, since statistical analysis failed to detect a difference between the amount of waste removed by either method of picking for cotton from the Stoneville area. A comparison of field data (Table 3) with the limits of the estimated ranges based on the fractionator data is as follows: Air pollution material dis- charged to the atmosphere from the ginning operation (Percent by weight of cot- ton ginned - Table 3). . . . Limits of estimated range of air pollution material (Percent by weight of cot- ton ginned after removal of seeds) It is obvious that the field test data fall with- in the fractionator range of air pollution ma- terials discharged from the ginning operation; hence fractionator data may be used to predict the amount of air pollution material that would be discharged during the ginning of cotton with- out control equipment. Further, it is possible by statistical methods to estimate the discharge of air pollutant material (expressed as percent by weight of ginned cotton) from the fractionator data (expressed as percent by weight of cotton fiber after fractionation). The accuracy of es- timation increases as the number of fraction- ator samples increases, e.g., based on a single fractionator sample, the estimated discharge of air pollutant material lies somewhere between 40% to 160% of the value obtained by the frac - tionator, whereas for 20 fractionator samples, the estimated discharge lies between 86% to 114% of the mean of the fractionator data. Inter- mediate values are indicated below: No. of Fractionator Estimation Limits in Samples Percent of Mean 1 40 - 160 5 74 - 126 10 81 - 119 20 86 - 114 Since these analyses show that the fraction- ator data are indicative of actual field sampling of air pollution materials discharged during the ginning process, this device could be used to evaluate on a nation-wide basis air pollution originating from cotton gins. Samples of cot- ton, both hand- and machine-picked, could be selected on a random basis, and sent to a cen- tral location for fractionation testing and data analysis. 1.39% 1.00% - 4.00% (2) "Introduction to Statistical Analysis, " Dixon, W.J., and Massey, F. J., Jr., p. 110, 1951, ------- Applicability of Shirley Analyzer as Air Pollution Indicator 13 Applicability of Shirley Analyzer as Air Pollution Indicator Ten samples of ginned cotton (5 each hand- and machine-picked) taken just prior to passing through the lint cleaner and nine samples (5 machine- and 4 hand-picked) which had been processed through the cleaner were analyzed using the Shirley Analyzer to determine the ap- plicability of this device for evaluating the amount of air pollutant material resulting from the lint cleaning process, Table 6. Results indicate that samples which had passed through a lint cleaner contained on the average about 70% as much particulate (noncollected waste) as did samples which had not passed through the cleaner. Statistical evaluation of the tests performed (Appendix 2) indicates a large variation in test results, due either to sampling variation or variation inherent in the analyzer. This varia- tion is so great that the probability is less than one in twenty that an actual difference of 30% would be detected using the Analyzer. It may be concluded then, that the Shirley Analyzer is not practical for evaluating the amount of pollu- tion material resulting from the lint cleaning operation. By analyzing an even larger number of samples in the Shirley apparatus, it is prob- able that the difference of 30% noted above would be determined to be statistically significant. However, from the practical standpoint it is not desirable to employ a laboratory device re- quiring an excessive number of samples. This device, however, is satisfactory for evaluating differences in pollution from the lint cleaning process due to methods of picking. Statistical analysis of this same data indicates that significantly less particulate (noncollected waste) was obtained for hand-picked cotton, i. e., about 58% of that for machine-picking. As previously noted, the analysis of fractionator data showed no significant difference between hand-picked and machine-picked cotton. There are two possible explanations for this variation. First, the Shirley Analyzer is more efficient in cleaning cotton than either the gin or fraction- ator, and second, before a cotton sample is placed in the fractionator, all large foreign materials, such as sticks, stems, leaves, hulls and seeds are removed by hand. In the ginning process there is a possibility that some of the above material is reduced in size, thereby cre- ating air pollution particulates from waste nor- mally not considered air pollution material. Since, as may be seen from the fractionator data in Tables 4 and 5, machine-picked cotton does contain more hulls than hand-picked, it is possible that the ginning operation created the difference between hand-picked and machine- picked cotton. In this event the Shirley Analy- zer, which uses ginned cotton, would show this difference whereas the fractionator, which uses unginned cotton, would not. Since this difference does not appear at pres- ent to affect the comparison of gin test data and fractionator data, further consideration will not be given to this point. Control Equipment Investigation Control techniques and equipment for reduc- tion of particulate emissions must be within the economic limitation imposed by the marginal operation of the cotton ginning industry. To satisfy this economic limitation, several con- ditions must be met: 1. Moderate initial cost, and reasonable operation, cleaning, and maintenance costs. 2. Maximum pressure drop not to exceed about 0. 5 inches of water, so as to allow use of low cost air moving equip- ment with nominal power consumption. 3. Relatively simple means for cleaning and disposal of waste. 4, High collection efficiency for all par- ticle sizes and for operation under variable flow rates and dust loadings. After considering specifications and costs of control devices currently available, it was ap- parent that no standard equipment would meet all the requirements. Because of the relatively high concentration of lint fibers in the effluent discharge, it was ------- 14 AIR-BORNE PARTICULATE EMISSIONS FROM COTTON GINNING OPERATIONS burlap (which is readily available at cotton gins) as the filter media would best satisfy all of the requirements. Burlap was chosen as the fab- ric because of its low cost and high porosity; hence, its low resistance and ability to main- tain a high filter velocity. The burlap fibers initially act as a filtering media for the lint. Later the lint fibers collected on the burlap create a closely arranged cotton-fiber filter surface on the loosely woven coarse burlap fil- ter cloth. Such a technique results in high ef- ficiency filtration and still meets the required economic limitations. To test this technique, a pilot model was constructed as shown in Figure 9. This device consisted essentially of a box, four sides of which were covered with burlap. Two high volume samplers provided the suction to pull the discharge from the gin through the inlet duct into the filter chamber, through the burlap filters, and then discharge the clean air to the atmosphere. Total discharge from the pilot filter unit was sampled using two high efficiency glass sampling filters attached to the top of the pilot unit. The filter area of this pilot de- vice could be varied by removing a section of sheet metal and substituting a strip of filter material. By obtaining the total weight of ma- terial settled within the device and collected on the burlap, and that collected on the glass fiber sampling filters, it was possible to compute the efficiency of the pilot unit. Pressure taps were inserted on both sides of the burlap filter to indicate the pressure loss across the pilot unit. . HIGH VOLUME SAMPLER & FLOW METERS SUCTION CHAMBER BURLAP FILTER 11 FILTER 1 .CHAMBER" SAMPLING LINE ------- Control Equipment Investigation 15 The pilot device was tested on effluent from a commercial gin in Mississippi wherein all effluents were conducted to a single point of discharge to a dust house. Two filtration ve- locities were used, one at 30 ft/min. and the other at 60 ft/min. Formation of the filter matrix was excellent as shown in Figure 10. Initial pressure drop and increase in pressure drop with time were low for both filtration ve- locities studied. As indicated in Figure 11, the initial pressure drop for each filtration velocity was less than 0.02 inches of water andincreased during a two-hour period to 0. 10 inches of wa- ter for the 30 ft/min. filtration velocity and to 0. 47 inches of water for the 60 ft/min. velocity. Particulate loadings in the discharge and the total mass of particulate collected by the pilot unit are shown in Table 7. Assuming that inlet Figure 10. BURLAP FILTER WITH LINT LOADING loading was uniform during all four quarters of the total sampling period, the loading during each quarter, and the collection efficiency of the unit during that quarter were calculated. As may be seen from the results, the tech- nique appears to be worthy of full-scale trial. Such a control unit would possess low initial cost, low pressure drop with a correspondingly low power requirement, and high collection efficiency. It is suggested that a trial full-scale unit be designed to operate at an approximate filter velocity of 30 ft/min. Such a design would have a low pressure drop increase with time and would require cleaning only every four to eight hours. For minimal initial cost, the unit should consist of a wooden frame covered with open wire screen to support the burlap, be open sided and roofed with a sufficient overhang to protect the filters against rain. Cleaning would be done manually by entering the chamber and removing the collected lint and dust mass from the burlap. I .60 40 FILTER AREA = I SO. FT. FILTER VEL0CITY=60FT. PER MIN.- m .20 FILTER AREA'2 SQ FT FILTER VELOCITY * 30 FT PER MIN.—\ UJ .10 20 40 SAMPLING TIME (MINUTES) 60 80 100 120 NUTES) Figure 11. ANTICIPATED PRESSURE ------- TABLE 7 BURLAP FILTER TEST RESULTS TEST A Filter Area = 2 sq. ft. Filter Velocity = 30 ft. /min. TEST B Filter Area = 1 sq. ft. Filter Velocity = 60 £t./min. Time Interval (min.) Loading to Settling Cham- ber and Burlap Filter During Time Interval (gms) Material Col- lected on Glass- Fiber Filter (gms) Efficiency of Settling Chamber and Burlap Filter Loading to Settling Cham- ber and Burlap Filter During Time Interval (gms) Material Col- lected on Glass- Fiber Filter (gms) Efficiency of Settling Chamber and Burlap Filter 0-30 5.025 0.469 91.5% 4.45 0.280 94.1% 30 - 60 5.025 0.159 96.9% 4.45 0.068 98.5% 60 - 90 5.025 0.093 98.2% 4.45 0.053 98.8% 90 - 120 5.025 0.071 98.6% 4.45 0.025 99.4% TOTAL 20.100 0.792 Avg= 96.2% 17.80 0.426 ------- Appendix 1 17 Summary The quantity of particulate emissions from a typical cotton gin and laboratory devices for predicting such emissions have been evaluated. Although the data obtained are limited in scope, they are believed to be adequate for a general assessment of the problem. It has been shown that the standard laboratory fractionator is ef- fective in establishing the amount of air pollu- tion material that will be discharged to the at- mosphere during ginning operations. From the limited number of areas used in this study, air pollution emissions from cotton ginning appear to be about 1-1/2% of the total weight of proc- essed cotton. Since the gin sampled was se- lected as typical of such operations, this value may be assumed to show generally the order of such discharges from ginning operations. A pilot model of a device to control particu- late discharges from cotton gins was designed based on two common principles employed in control equipment; namely, gravity settling and fabric filtration. Tests in the field indi- cate the method to be much more efficient (96%-97% removal) than equipment presently used in ginning operations. A prototype control device based on the pilot tests is suggested as a means for economically controlling particu- late discharges from ginning operations. This device should be trial tested to establish its feasibility under full-scale field conditions. Personnel The Engineering Research & Development Unit, Air Pollution Engineering Program, Public Health Service, carried out the field sam- pling and laboratory studies for this report. Messrs. C.M. Merkel, J.V. Shephard, and E. A. Harrel of Cotton Ginning Laboratories of the U. S. Department of Agriculture at Stone- ville, Mississippi, made the necessary arrange- ments for sampling in the field and assisted in obtaining field and laboratory data. Appendix 1 Statistical Evaluation of Cotton Gin Fractionator Data In this evaluation, the variable X is the amount of air pollution material contained within cotton expressed as a percent of dry cotton by weight. The factor "percent dry cotton" was determined experimentally as the ratio of the weight of a 30-50 gram sample of seed cotton after drying to its weight before drying. Using data from Tables 5 and 6, a total of 40 such val- ues of X were available. A preliminary analysis showed that, for cot- ton from a given area picked by a given method, the variation between means of different sam- ples of cotton was not significantly different from the variation within measurements ob- tained from replicate analysis of the same sam- ple. This then established that sampling intro- duced no complications in subsequent analyses. The analysis of variance, therefore, reduced to a determination of the effect of two factors, namely area (Clemson and Stoneville) and pick- ing method (hand and machine). The results of ------- 18 AIR-BORNE PARTICULATE EMISSIONS FROM COTTON GINNING OPERATIONS The results of the analysis of variance computations may be summarized in tabular form as follows: TABLE 8 MEAN PERCENTAGE AIR POLLUTION VALUES (AVERAGE OF 10 SAMPLES) Machine-Picked Cotton Hand-Picked Cotton Area X* 95% CI* X* 95% CI* Clem son 0.54%** 0.44%- 0.64% 0.79% 0.69%-0.89% Stoneville 0.91% 0.81%- 1.01% 0.84% 0.74%-0.94% COMBINED 0.72% 0.81% **This value is significantly lower than the other three values, which average 0.85%. Symbols: X - estimated mean %-air pollution (average of 10 samples) 95% CI - 95% confidence interval for true mean. TABLE 9 ANALYSIS OF VARIANCE Source of Variation „Sum of Degrees of variance — —— Squares Freedom — I Between Methods of Pick- ing (M) 0.078234 1 0.0782 II Between Areas (A) 0.444577 1 0.4446 EI A x M Interaction 0.267813 1 0.2678 IV Between Replicate Cotton Samples 0.232243 10 0.0232 V Between Replicate Analyses (Within Samples) 0.674013 26 0.0259 Total 1.696880 39 VI Poll IV and V 0.906256 36 0.0252 Symbols: N.S. - not significant at 0. 05 probability level. * - significantly greater than unity at .01 probability level. F - Ratio I/m = 0.29 N.S. n/in - 1.66 N.S. m/VI = 10.63* ------- Appendix 2 19 Appendix 2 Statistical Analysis of Shirley Analyzer Data A total of nineteen 100 gram samples of ginned cotton from the Stoneville area, Table 7, nine of which had been processed through the gin lint cleaner and ten of which had not, were dried for one hour at 210° F. Each sample was weighed and then separated by the use of a Shirley Analyzer into "cleaned lint" and "col- lected waste" components which in turn were also weighed. "Non-collected waste" (the air pollution fraction) was expressed as a percent- age of the weight of dried cotton and used as the response variable X, in the analysis described below. The analysis of variance technique was used to evaluate the effect of two factors, name- ly picking method (hand and machine) and lint cleaner (i.e., whether or not the lint cleaner was used). The results of this analysis are presented in Tables 10 and 11. TABLE 10 MEAN PERCENTAGE NON-COLLECTED WASTE VALUES Machine-Picked Cotton Hand-Picked Cotton n X 95% CI n X 95% CI Total X Lint Cleaned 5 1.31% 0.90% - 1.72% 4 0.55% 0.09% - 1.01% 0.93%* Lint Not Cleaned 5 1.66% 1.25%- 2.08% 5 1.10% 0.69% - 1.51% 1.38%* COMBINED 10 1.49% 1.20%- 1.78% 9 0.86% 0.55% - 1.17% ~Ratio 0.93/1.38 = 68% (about 70%) Symbols: ji_- number of samples X - estimated mean % non-collected waste (average of n samples) 95% CI - 95% confidence interval for true mean TABLE 11 ANALYSIS OF VARIANCE , . .. Sum of Degrees of Variance Source of Variation squares Freedom Estimate I Between Methods of Pick- ing (M) 1.883304 1 1.8833 II Lint Cleaner (L) 0.793547 1 0.7935 HI M x L Interaction 0.188635 1 0.1886 IV Between Replicates (Exp. Error) 3.361341 15 0.2241 TOTAL 6.226827 18 V Pool in and IV 3.549976 16 0.2219 Symbols: N. S. - not significant at .05 probability level. ** - significantly greater than unity at .01 probability level. F - Ratio I/m = 8.49** H/V = 3.58 N.S. ------- 20 AIR-BORNE PARTICULATE EMISSIONS FROM COTTON GINNING OPERATIONS Appendix 3 Glossary Boll trap - a gravity settling type device. Condenser - a machine designed to collect ginned lint into smooth, endless "Bat". Dust house - rather large tall structure open on top and sometimes also open on bottom. Acts as settling chamber and elutriator. Extractor-feeder - a device which extracts burs, stems, whole leal, and other trash from seed cotton and then feeds the cotton to the huller front gin stand. Feed control - consists of a sealed paddle-wheel arrangement to maintain constant flow of material through the gin. Huller Front Gin Stand- an arrangement of saws and air blasts that removes seeds, hulls, and motes from the cotton, and in addi- tion, the gin fronts remove appreciable quantities of leaves and stems, even when elaborate cleaning and extracting proc- esses have preceded them. Leaf trash - small bits of leaf contained in the lint. Lint Cleaner - a device designed primarily to remove fine leaf particles, motes, dust, sticks, and shale fragments. Motes - a small group of fibers held together by a portion of the covering of the seed. Pin and pepper trash - small pieces of dirt, sand, etc., that appear as pepper in the cleaned cotton. Seed cotton - lint cotton from which the seeds have not been removed. Separator - mechanical device which separates the carrier air stream from the carried lint. Example: wire screen with rotary rake. Seven Cylinder Cleaner - contains seven spiked- drum cylinders which carry and "scrub" the seed cotton over wire mesh or perfo- rated metal screens through which fine particles of leaf trash, dirt or sand, stems and small sticks are expelled from the cotton. Six Cylinder Cleaner - similar to seven cylin- der cleaner. Stick and Bur Machine - employs a revolving saw cylinder. The teeth of the saw hold the locks of seed cotton, subject them to a carding and cleaning action as the cot- ton is spread across the surface of the saw cylinder, and strip them of heavy trash and burs. Stub tower dryer - a device in which the cotton follows a tortuous path while being sub- jected to heated air. Tower dryer - a longer version of the stub ------- |