!| i‘ WIN!!!” Ii u 114 296 __THS STUDY OF PACKiNG METHODS AND ?HE§R EFFECT UPQN UQUID DISTREBUfiCDN EN 3UMPED 53ACKED ‘E‘OWERS I‘Esoszs in? “site Degree a! M. S. :VZQCHfiGAN STATE UNNERSS'E‘Y Saks: F. Knees: z?$3 THESIS This is to certify that the thesis entitled Study of Packing MEthods and Their Effect Upon Liquid Distribution in Dumped Packed Towers presented by John F. Knoop has been accepted towards fulfillment of the requirements for Masters degree in Ch-E- WWW Major professor Date May 13) 1963 0-169 LIBRARY Michigan State University ABSTRACT BTUDK OF BACKING METHODS AND THEIR EFFECT UPON LIQUID DISTRIBUTION IN DUMPED PACKED TOWERS by thn E. Khoop The liquid distribution using water with two packed towers having inside diameters of 2.5 and 5.0 inches were measured for glass spheres and glass Raschig rings dumped by various methods. Nb gas flow was used and the liquid distribution was found by dividing the liquid flow into four parts, each from twenty-five per cent of the cross sectional area. In.most cases,~ a multipoint feed distributor was employed. It was found that a packed depth equivalent to four tower diameters with this distributor was suf- ficient to give results that were reproducible for greater packed depths. For glass spheres, no relationship could be found to indicate that the liquid distribution was altered by the packing method. Large "wall effects" (liquid cone centration near the wall) were noted for column to packing ratios of 5 to l. The effect was suppressed but not eliminated by increasing the ratio to 10 to 1. For glass Raschig rings, one condition was feund where it appeared that the liquid distribution was affected to a small extent by the packing method employed. John F. Knoop In that case, random dumping methods resulted in liquid flow to the center of a column. When the column was packed from both the center to the wall and wall to center, the distribution became more uniform. This was for a column to packing ratio of 20 to l. Ebr a ratio of 10 to 1, large “wall effects" again resulted. A comparison was also made between gas distribution data reported in the literature and the liquid distribu- tions which were found experimentally. From this data, it appeared that the flow rates were proportional in the center of the tower and inversely proportional near the wall. For all the packings and towers, the liquid distribu- tion varied with the flow rate. However, no trend was noted. Approved 10W Jaw. STUDY OF PACKING METHODS AND THEIR EFFECT UPON LIQUID DISTRIBUTION IN DUMPED PACKED TOWERS By John F. KnOOp A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemical Engineering 1963 Acknowledgement I would like to thank Dr. Randall Ludt for his many suggestions and dedicated help without which I would not have been able to complete this thesis. Similarly, I would like to extend thanks to Mr. William Clippinger who so capably and expertly aided in the construction of the experimental towers. 11 Table of Contents Page Title Page 1 Acknowledgement 11 Table of Contents iii List of Tables iv List of Figures vi Introduction 1 Historical Background 2 Equipment h Procedure 7 a) Operating 7 b) Correlation of Data 10 Results 11 Discussion of Results 17 a) Spherical Packing 21 b) Raschig Ring Packing 27 c) Comparison of Liquid and Gas Distributions 33 Conclusions 35 Appendix 36 References “3 iii List of Tables Table 1. Tabulated Results from 2. 5 Inch Diameter Column Packed with 0. 5 Inch Glass Spheres to a Depth of 36 Inches. Table 2. Tabulated Results from 2.5 Inch Diameter Column Packed with 0.25 Inch Glass Spheres to a Depth of 20 Inches. Table 3. Tabulated Results from 2.5 Inch Diameter Column Packed with 0.25 Inch Glass Raschig Rings to 3 Depth of 42 Inches. Table H. Tabulated Results from 5.0 Inch Column Packed with 0.5 Inch Glass Spheres by Method DRP. Table 5. Tabulated Results from 5. 0 Inch Diameter Column Packed with O. 25 Inch Glass Raschig Rings to 8 Depth of 40 Inches. Table 6. Tabulated Results from 5. 0 Inch Diameter Column Packed with O. 5 Inch Glass Raschig Rings to 3 Depth of 40 Inches. Table 7. Average Liquid Distribution for Spherical Pa Oking 0 Table 8. Average Liquid Distribution for Raschig Rings. Table A. Experimental Data from 2. 5 Inch Diameter Column Packed with 0. 5 Inch Glass Spheres to 8 Depth of 36 Inches. Table B. Experimental Data from 2.5 Inch Diameter Column Packed with 0.25 Inch Glass Spheres to a Depth of 20 Inches. Table C. Experimental Data from 2. 5 Inch Diameter Column Packed with 0. 25 Inch Glass Raschig Rings to 8 Depth of 42 Inches. Table D. Experimental Data from 5.0 Inch Diameter Column Packed with 0.5 Inch Glass Spheres by Method DRP. Table E. Experimental Data from 5. 0 Inch Diameter Column Packed with 0. 25 Inch Glass Raschig Rings to a Depth of 40 Inches. iv Page 11 12 13 l“ 15 16 24 3O 37 38 39 40 N1 List of Tables (Continued) Page Table F. Experimental Data from 5.0 Inch Diameter Column Packed with 0.5 Inch Glass Raschig Rings to a Depth of 40 Inches. M2 List of Figures Page Figure 1. Experimental packed column. 5 Figure 2. Packing support and liquid separator. 6 Figure 3. Effect of initial distribution on final liquid distribution. 2.5 inch tower with 0.5 inch spheres. 18 Figure N. Effect of packing depth on liquid distribution. 5.0 inch tower with 0.5 inch spheres. 20 Figure 5. Liquid distribution for three different methods of packing a 2.5 inch tower with 0.5 inch glass Spheres. 22 Figure 6. Liquid distribution for three different methods of packing a 5.0 inch tower with 0.5 inch glass spheres. 23 Figure 7. Liquid distribution for three packings of the same method in a 2.5 inch tower with 0.25 inch spheres. 25 Figure 8. Liquid distribution for seven methods of packing a 2.5 inch tower with 0.25 inch Raschig rings. 28 Figure 9. Liquid distribution for several methods of packing a 5.0 inch tower with 0.25 inch Raschig rings. 29 Figure 10. Liquid distribution for 0.5 inch Raschig rings in a 5.0 inch tower. 31 Figure 11. Comparison of gas and liquid distributions. 3M vi Introduction Packed towers have been employed throughout the chemical process industry for a long period of time in a great many operations as a means of contacting liquids and vapors, as a medium for chemical reactions, as a facility for liquid-liquid extraction, and as a mode of mixing. However, the scale-up and sizing of these operations is carried out with little confidence. Even for columns of the same dimension containing identical packing there is a lack of agreement. A possible explanation might be found if the packing arrangement were studied because of the effect that the packing can have on the liquid and gas flow patterns within a tower. Therefore, since packing arrangement could be a primary factor in determining the tower's operating efficiency, it was proposed to determine if the packing arrangement could be varied by changing the method by which a column was packed. The liquid distribution from four equal and concentric area sections at the bottom of a tower was used as a parameter to indicate any change in packing arrangement. Historical Background The design of packed towers is still largely a matter of experience partly because proven empirical relationships have not been developed through experimental work. Leva1 has pointed out that a large amount of the data that has been collected is contradictory and thus might add to the difficulty of defining variables that are important to the design of packed towers. 0f the design tools that are available, many have been based upon theoretical derivations which may very well be valid if controlling assumptions are correct. In most cases, it is assumed that both liquid and gas flows are uniform over the column's cross sectional area. Thus, since the flows are not uniform, and this has been proven,“’5 the relative gas~liquid rates (reflux for distillation) will vary considerably within the column and will be reflected in the operating efficiency of a column. Specifically, Norman2 found a large difference in the operating efficiency of two packed columns under identical conditions. He concluded that maldistribution (non-uniformity) of liquid was the cause of the difference. Other experimenters such as Morales, Spinn, and Smith3 have studied the distribution of gas in a 2 inch diameter tower packed with Raschig rings. With no liquid flow, the gas velocity too varied considerably over the cross 2 3 sectional area with a low velocity occurring in the center and at the walls. They also found that the gas distribution varied with depth of packing. Baker, Chilton, and Vernon” studied the liquid distribution of various packings at various depths. They found that the liquid concentrated at the walls of small towers and not with large towers. This led them to conclude that the liquid would concentrate on the wall unless the column diameter was greater than 8 times the packing diameter. They also found that the liquid distribution remained unchanged once it had traversed a depth equivalent to 10 column diameters for a single center point feed. These men also concluded that the vapor velocity had little effect upon liquid distribution except near flooding cenditions and that moderate changes in liquid rate had little effect upon liquid distribution. Another experimenter, J. N. Muiiins5, worked on the same problem of liquid distribution and substantiated the fact that the liquid tends to concentrate at the wall unless the prOper sized packing is used. However, he concluded that the "wall effect" could be eliminated if the column diameter were 12 or more times the packing diameter. Equipment The experimental work was carried out in two packed columns. One of the columns as shown in Figure 1 was U8 inches long and had an inside diameter of 2.5 inches. The tower was cut from a piece of transparent plastic pipe. The pipe was held firmly between the bottom packing support plate and an upper plate with four steel rods. The rods were threaded and equipped with wing nuts for easy dismantling. The packing support as shown in Figure 2 was constructed from brass bar stock and tubing in such a manner that the cross sectional area between each of the four concentric cylinders would trap the liquid flow from 25 per cent of the column cross sectional area. The drains from each of the column cross sections were connected to tubing so that the flow from each could be collected in 4000 ml flasks. A rotameter on the water feed line was used as a flow indicator. A stop watch was used to measure the length of time for each run. The second column was identical to the first except that the inside diameter was 5.0 inches. The column was supported with a similar plate and all auxiliary equip- ment was the same. The columns were packed with l/u inch diameter glass spheres, 1/2 inch glass spheres, 1/A inch glass Raschig rings, and 1/2 inch glass Raschig rings. Figure 1. Experimental lacked column. UPPER SUPPORT *4L1 PLATE \ \ u \_ PERFORATED fig! mom DISTRIBUTOR PLATE 1 mTAMErER AREA STEEL ROD J / rLAs TIC COLUMN F I SUPPORT / PLATE T W T LINE DRAIN ZONE 1 DRAIN / DRAIN ZONE 2 ZONE 3 DRAIN ZONE 4 Packing Support and Liquid Separator. Figure 2 . SUPPORT BCREEN DRAIN ZONE 3 r' ' ZONE 4 |\—DRAIN Procedure a) Operating Since it was desired to measure the effect of random packing methods on liquid flow patterns, it was necessary to pack each column several times. To start, the plastic column was fitted between the upper and lower brass support plates with rubber gaskets. The steel guide wires were put in place and tightened to hold the plates and column in place. Next, the column was clamped to a laboratory support rack and leveled. A one-quarter inch screen was placed over the packing support so that small packing would not wedge in the separator. The column was then packed by dumping the packing into the column using one of the methods listed on the following page. In most cases, the column was packed to a depth of 40 inches. Next, a liquid distributor was placed on tOp of the packing and the liquid flow started. No gas flow was used. The liquid flow was increased until the packing became flooded. This rate was observed on the rotameter. The flow was reduced by approximately half and the column was allowed to attain a state of equilibrium. In the meantime, four H000 ml flasks were placed in a rack so that they could be placed under the drains at the same time. A stop watch was started as the flasks in the rack were placed beneath the drains. After a large quantity of water had been collected, the rack and flasks were 8 removed as the watch was stopped. The run time was recorded along with the volume of water collected at each drain. This procedure was followed for three different water flow rates on each packing. After obtaining this data, the column was dismantled and the packing removed. After reassembling the column, packing was again dumped into the tower by another method. This procedure was repeated for all of the packings. The following methods were followed in packing the two columns: (1) Dry random pack (DRP) Packing was dumped into a dry column evenly over the cross sectional area. (2) Packed from center (PCF) Packing was dumped into the center of a dry column. (3) Packed from wall (OSD) Packing was dumped into annular area near the wall of a dry column. (u) Wet random pack (RPW) Packing was dumped into a water-filled column evenly over the cross-sectional area. (5) Moving water pack (MCW) Packing was dumped into water—filled tower with the water being moved down the column at a superficial velocity of 0.15 feet per second or O.b3 feet per second. (6) Rotating water pack (ROD) Packing was dumped into water-filled tower with water being rotated with a stirrer so that the packing was thrown to wall of column. (7) Moving rotating water pack (BMW) Packing was dumped into water—filled tower with water being rotated and being moved down at a velocity of O.h3 feet per second. (8) Packed from wall in water (OSW) Packing was dumped into annular space near the wall of a water-filled column. (9) Packed with tube in center (PTC) Packing was dumped into wateréfilled tower that had a 7/16 inch diameter tube in the center of the tower. Tube was removed after tower was packed. (10) Packed with a tube being raised in center (RTC) Packing was dumped into a water-filled tower that had a 7/16 inch diameter tube being raised in the center of column. Tube extended into packing l/2 inch. (11) Packed in tube (PIT) Packing was placed in 1 inch diameter tube. Tube was placed in dry tower and slowly raised. Packing was replaced in tube as packing drained into the tower. b) Correlation of Data The experimental results were used to calculate the percentage of total liquid flow that was collected from each of the four equal and concentric areas of the column cross section. Graphically, these data were correlated to the over-all liquid mass flow rate which was expressed in pounds per hour per square foot of empty-cross sectional area. In each case, the data were plotted as an accumulative percentage of the flow collected from successive annular zOnes, starting with the annular area adjacent to the wall and ending with the flow from the center section. For example, the percentage shown in Figure 4 for a mass flow rate of 15,000 pounds per hour per square foot indicate that the area near the wall (Zone 1) collected 3M% of the flow; the area adjacent to Zone 1 (Zone 2) collected 58-3u or 2am; the next area (Zone 3) collected 77-58 or 19%; and the center area (Zone u) collected 100-77 or 23%. Had the percentages all been 25, the flow distribution would have been perfectly uniform and the three points on the graph would have been at 25, 50, and 75 instead of 3M, 58, and 77. Therefore, if as in the example the curves are displaced to the right, the flow is concentrated at the wall. Similarly, a shift to the left would indicate a con- centration in the center. 10 Tabulated Results from 2.5 Inch Diameter Column Packed with 0.5 Inch Glass Spheres to 3 Depth of Table 1. 36 Inches. Run Method No. of Packing l 1 DRP 2 1 DRP 3 1 DRP u 1 DRP 5 1 DRP 6 1 DRP g 1 DRP 1 DRP 9 1 DRP 10 2 DRP ll 2 DRP l2 2 DRP 13 2 DRP 1“* 2 DRP 15* 2 DRP 16* 2 DRP 17 2 DRP l8 1 POP 19 1 PCF 20 l POP 21 l OSD 22 l OSD 1 OSD *These runs were with center point distribution, all others Mass Flow Lbs./Hr.-Sq.Ft. “2,500 “2,200 “2,200 “1,900 5“,100 56,800 57.000 81,000 81,900 “1,000 “2,000 56,“00 56,200 81,500 “1,600 57,600 80,500 “2,000 56,000 81,300 “2,500 58,200 85,000 -were multipoint. 11 Percent of Flow Zone Zone 1 U) .D'U" Ira-mm tU‘lUO .lz'Jt‘U'IUWUJUU cream .t‘Jz'U1 \OmoomHWM-QWNHHUWU‘HNNOCDKDN CD:A»~J#HU#JOKD¢>twothUthquOChomflCD Zone 2 18.4 20.9 21.7 20.4 23.“ 22.3 21.6 22.1 oooxmww-q O “3 CW [0 l—‘\O\O CID-DU'I mm CL» OO 2“. 262 N 0 4:5 0 U0 .t'Jz‘ONU'IUUKO NOONONKO CDKOOKOU'IONN 0 N00\ HHHHHHHHHHHHHHHHHHHHHHH Hmmmww 1:0) cmoooouow-q (Dz-cm tun-4:60 o Table 2. Tabulated Results from 2.5 Inch Diameter Column Packed with 0.25 Inch Glass Spheres to a Depth of 20 Inches. Run Method Mass Flow Percent of Flow No. of Lbs./Hr.-Sq.Ft. Zone Zone Zone Zone Packing 1 2 3 “ 2“ 3 DRP 52,000 27.“ 27.1 30.6 1“.9 25 “ DRP “9,800 27.1 27.0 22.1 23.8 26 “ DRP 32,300 3“.1 28.2 25.6 12.1 2% “ DRP “1,500 3“.6 25.1 19.3 21.0 2 “ DRP 55,900 33.0 19.7 26.9 20.“ 29 5 DRP “6,300 20.8 3“.2 21.0 2“.0 30 5 DRP 31,200 33.6 28.8 20.9 15.7 31 5 DRP 38,“00 29.8 25.2 23.7 21.3 32 5 DRP “7,100 3“.2 22.“ 22.2 21.2 13 Table 3. Tabulated Results from 2.5 Inch Diameter Column Packed with 0.25 Inch Glass Raschig Rings to a Depth of “2 Inches. Run Method No. of Packing 33 l RPw 3“ l RPw 35 l RPW 36 2 RPw 37 2 RPW 38 2 RPW 39 3 RPW “0 3 RPW “l 3 RPW “2 “ RPW “3 “ RPW ““ “ RPw “5 1 NOW “6 1 MCW “ l MCW “ 2 Mcw “9 2 MON 50 2 MCW 51 3 MCW 52 3 MCW 53 3 NOW 5“ 1 ROD 55 l ROD 56 l ROD 5 1 BMW 5 1 BMW 59 1 BMW 60 1 osw 61 1 OSW 62 l osw 63 1 PTC 6“ 1 PTC 65 l PTC 66 l RTC 6 1 RTC 6 l RTC Mass Flow Lbs o/Hro-Squt. 35.“00 52,000 79.300 3“, 300 53, ,000 ,300 36, 700 53.500 76,100 3“,800 52,000 76,600 35.000 52.500 79.700 36, 00 52, 00 79,000 3“,800 53.000 73.“00 35.700 50.500 71,200 36,000 52,800 7 ,200 3 ,200 5“,000 76.900 39.800 53.200 76,600 37.500 52.700 7“,6oo 13 Zone 1 O OKOkmLU OKO 4:1» H NH CDCDCD-D‘Jr-Q O\CDUJN O\U"| CID-x] HUD-x] 0 4:10 OUT-\‘IKO O ONUJU) #:00me O m-qwxn [\3 oomooc-m OUT 1‘: wwzkttzw.lrlrwthkkmeWtwwwmk-wwktwctw Nmomm o o o o o o o o WONl—‘O 35.9 Zone 2 tmmmmmm-qqmml—‘wmfi-D‘UOWWUTCRWI-‘LDWCDNNO\'\10\O\H-l:'-I:U1 000000000000. 00 o o to. o WQDQWNOWNWmml-‘OWKOmwmmNONNWOWHNmm-D’mbmOW NNNNNNNNNl—‘I—‘NNHNNMNNNNNNNNNNNNNHHNNNN Percent of Flow Zone 3 18.3 1 . 16.1 21.2 19. 19. 19.7 23.1 22.1 28.1 21.8 19.“ 21. 2 18. 3 19. “ 25. O 22.1 21. 8 23. 7 2“.O 22. 6 25.3 28.6 22.0 28. 5 27. 2 2“. 6 16.0 17- 7 19. 6 17- 7 20.1 22.6 22.7 19.7 22.0 Zone 18.2 l“.9 16.5 20.0 15. 5 l6. “ 12.6 15-9 16.7 19.1 Tab le “. Tabulated Results from 5.0 Inch Diameter Column Packed with 0.5 Inch Glass Spheres by Method DRP. Run No. 69 7o 71 72 73 7“ Depth of Packing “o" “o" “o" 20" 20" 20" 30" 3O" 30" “0" “o" “0" By Method PCF 81 82 83 By 8“ 85 86 no" no" “0" Method OSW 140" no" “0" Mass Flow Lbs./Hr.-Sq.Ft. 25.900 2“,200 33.400 15.900 2“,900 3“,1OO 1“,500 2“,2oo 33.900 1“,900 26,100 33.100 15,700 2“,900 33.600 15.700 25,600 32,“00 1“ Zone 1 3331 20.8 35-3 25.7 22.“ 28.“ 25-7 21.5 37.“ 25.7 23.5 {01000 mmoo O ODUTCD NNLU mm“ o 0 CD30 Zone 2 26.0 26.3 23.7 21.2 27.0 21.6 19.8 26.3 22.7 18.“ 25.7 22.3 Percent of Flow Zone 3 21.8 22.8 25.7 21.2 23.1 2“.9 21.“ 2“.3 2“.0 27.0 20.7 22.3 13.3 23.2 2“.7 20.3 26.“ 23.“ Zone Table 5. Tabulated Results from 5.0 Inch Diameter Column Packed with 0.25 Inch Glass Raschig Rings to 2 Depth of “0 Inches. Run Method Mass Flow Percent of Flow N0. of Lbs./Hr.-Sq.Ft. Zone Zone Zone Zone Packing l 2 3 “ 8g 1 RPw l“,100 20.8 23.8 2“.O 31.“ 8 1 RPw 23,“00 23.2 23.2 2“.7 28.9 89 1 RPw 3“,000 21.5 21.5 22.5 3“.5 90 2 RPw 15,150 19.5 22.0 22.“ 36.1 91 2 RPw 23,600 22.8 22.5 23.7 31.0 92 2 RPW 30,900 23.1 22.7 2“.0 30.2 93 3 RPW 1“,900 18.5 2“.“ 2“.“ 32.7 9“ 3 RPw 25,000 22.“ 2“.8 23.6 28.7 95 3 RPw 31,800 21.9 23.1 23.1 31.9 96 1 MCW l“,200 21.2 26.1 22.1 29.6 9 1 wow 25,800 2“.8 23.“ 2“.8 27.0 9 1 MCW 28,500 2“.“ 21.“ 21.2 33.0 99 1 ROD 13,900 27.8 25.8 19.8 26.6 100 1 ROD 20, 00 2“.“ 26.2 2“.7 2“.g 101 1 ROD 33, 00 25.9 2“.2 23.1 26. 102 2 ROD 15,100 22.2 29.9 16.9 31.0 103 2 ROD 21,600 22.5 26.0 2“.“ 27.1 10“ 2 ROD 32,“00 2“.“ 29.“ 23.8 28.0 105 1 me 16,800 21.8 33.0 19.8 25.“ 106 1 BMW 2“,800 26.1 26.7 22.5 2“.7 107 1 BMW 33,100 2“.l 2“.9 2“.2 26.7 108 2 BMW 19,“00 25.“ 39.0 20.8 23.“ 109 2 RMw 25,300 21.9 23.0 21.9 33.2 110 2 RMW 30,000 2“.O 2“.8 21.0 30.2 111 l PIT 17,500 22.6 30.“ 19.“ 27.6 112 1 PIT 23,600 23.5 23.8 26.3 26.3 113 1 PIT 28,500 23.7 2“.5 23.7 28.1 15 Table 6. Depth of “0 Inches. Run Method N0. of Packing 11“ l RPW 115 l RPw 116 1 RPW 11 2 RPW 11 2 RPW 119 2 RPw Mass Flow Lbs./Hr.—Sq.Ft. 18,600 23,800 3“, 00 18, 00 2“,“00 31,800 16 Zone 27.0 2'2 28:6 30.1 32.3 Zone 2 23.6 25.1 21.3 20.2 17.8 15.3 Tabulated Results from 5.0 Inch Diameter Column Packed with 0.5 Inch Glass Raschig Rings to 2 Percent of Flow Zone 3 N O :3 (D .1:- l—‘LJU [DUO NUT wmwmmm NOON-EKOUT Discussion of Results Primarily, the purpose of this study was to determine if the packing arrangement of dumped beds could be altered enough by employing various dumping methods to change the liquid distribution. Because there are several other variables which could affect the liquid distribution, it was desired to define these variables experimentally so that their effect could be minimized. Three variables which were considered to be important were feed distribution, bed depth, and liquid rate. The first variable that was studied was liquid feed distribution. Baker, Chilton, and Vernon had presented conclusive evidence that the liquid distribution does not change once the liquid has traversed about 10 column diameters of depth when a single center point feed was employed. The 2.5 inch column was packed to a depth of 36 inches or 1“ column diameters and, therefore, should have been in a range where the liquid distribution would be unaffected. Data was collected and correlated to flow rate for a center point feed and a multipoint feed. From the plot of this data in Figure 3, it can be seen that the liquid distribution was unaffected by the initial distribution at that depth. When the single center point feed was used, it was observed visually that the liquid had not spread to the wall until a depth of 15 inches was reached. With 21 inches of the wall wetted, 17 FIGURE 3. EFFECT OF INITIAL DISTRIBUTION ON FINAL LIQUID DISTRIBUTION. 2.5 INCH TOWER WITH 0.5 INCH SPHERES. 90,000 x-SINGLE CENTER FEED o-MULTIPO INT FEED 80,000» 7 1 12 LIQUID 70,000- ZONE ZONE ZONE ZONE mass 1 2 3 4 GEN- 60.000' WALL TER FLOW 6 50,000- RATE “0,000- + LBS. HBO-SQ, OFT o 30,000+ 20,000 ‘ an . - 0 20 “0 6O 80 100 PERCENT OF LIQUID FLOW 18 r“ 1‘ 19 it would appear that a depth less than 9 column diameters would be required when a multipoint distributor was used. To prove that this would be the case, data were obtained from the 5 inch column equipped with a multipoint distributor and packed to depths of 20, 30, and “0 inches with 0.5 inch glass spheres. This data is shown in Figure “ and it can be seen that the distribution was almost identical for all three depths. Therefore, it was concluded that a packed depth equivalent to “ column diameters for a column equipped with a multipoint liquid feed distributor would be more than sufficient to insure that the initial distribution was not affecting the final liquid distribution. In all of the above cases, the liquid distribution was plotted against the liquid rate so that this affect could be shown. As can be seen in each case, the liquid distribution varied considerably with rate. It should also be noted that this rate is of such a high value that its significance is not fully understood. Theoretically, the flow should become more uniform as the rate increases and in most cases it was found that it does. However, on one or two packings this was not the case. Possibly, the liquid rates in these columns were not as close to a loading point as the rate might indicate. FIGURE 4. EFFECT OF PACKING DEPTH ON LIQUID DISTRIBUTION. 5.0 INCH TOWER WITH 0.5 INCH SPHERES. 45,000 0-20" DEPTH x-30" DEPTH ‘-“o“ DEPTH 40, 000 - LIQUID 35, 000 - + MASS 30, 000 - ZONE ZONE ZONE ZONE FLOW 1 2 3 4 25'00° ' WALL CENTER RATE 20, 000 ~ LBS . WEE-503T": . 15,000 - 1 ° ’r‘ 10.0006—1t—WW—mo PERCENT OF LIQUID FLOW 20 a) Spherical Packing Having concluded that the use of a multipoint feed distribution above a packed bed of at least 20 inches would be sufficient, work was directed to determine the effect of packing methods on the liquid distribution. Both towers were packed with glass spheres. In the case of the 2.5 inch tower packed with 0.5 inch spheres, the column was packed seven times using three different methods. The first method which was a random dumping in a dry tower was repeated four times. The other two methods were similar, except in one case the column was packed by dumping in the center and the other by dumping near the wall. The liquid distribution was found for each dumping and is shown in Figure 5 as a plot of distribution versus mass flow rate. In no case could any variation in distribution be contributed to the method by which the column was packed. The same 0.5 inch glass spheres were packed in the 5.0 inch tower using three methods. Again, the tower was packed by random dumping and dumping in the center of a dry tower. The third method to be employed was that of dumping near the wall of a water-filled column. The liquid distribution for these methods are shown in Figure 6. Again, no change in distribution was found. The 2.5 inch tower was also packed with 0.25 inch glass spheres. In all cases, the column was packed by 21 FIGURE 5. LIQUID DISTRIBUTION FOR THREE DIFFERENT METHON OE PACKING A 2.5 INCH TOWER WITH 0.5 INCH GLASS SPHERE. 90,000 80,000- LIQUID 70, 000 I- ZONE MAss 1 60, 000 .. "m 510w 50,000. RATE “0,000. LB . 30,000. 20,000 - . 4. - 0 20 “0 6O 80 100 PERCENT OF LIQUID EIOW 22 FIGURE g. LIQUID DISTRIBUTION FOR THREE DIFFERENT METHODS 0E PACKING A 5.0 INCH TOWER WITH 0.5 INCH CLASS SPHERES. “5,000 “0,000 .- LIQUID 359 000 P MASS 30, 000 . “in now; ZONE CENTER 3 25, 000 -- ZONE RATE ZONE 2 20,000- 1 LBS _ _. WALL HRo-BQOHO x 15,000» x 10,000. 1 4 A - 0 20 “0 6O 80 100 PERCENT OF LIQUID FIOVT 23 2“ random dumping into a dry column. No other methods were used since the distribution could not be reproduced for the same packing. It was noted that the liquid flow from some of the zones fluctuated when the column was being fed at a constant rate. Since there was a possibility that the packing support screen might be producing the fluctuations, the screen was removed. However, the fluctuation did not cease as illustrated in Figure 7. Therefore, the average liquid distribution of each packing was found and is shown in the following table. Table 7. Average Liquid Distribution for Spherical Packing. Packing Diameter - d, Column Diameter - D. d D D/d Method (Wall) Zones (Center) (Inches) (Inches) 1 '2"" 3 “ 1/2 2.5 5 DRP “3.9 22.219.3 1“.6 1/2 2.5 5 PCF “6.7 19.8 §.“ 13.1 1/2 2.5 5 OSD “5.5 20.61.8 12.0 1/“ 2.5 10 DRP 30.5 26.5 23.6 19.“ 1/2 5.0 10 DRP 27.3 23.“ 23. 6 19.“ 1/2 5.0 10 PCF 29.“ 25.10.“ 25.2 1/2 5.0 10 OSD 29.1 22.7 23. “ 25.2 Here, as with the plotted data, no effect of packing method on liquid distribution could be found. With this data 1t can be seen that a large "wall effect" occurred at this low ratio of column to paCking diameter. Essentially, this is the same conclusion that was drawn by Baker, Chilton, and Vernon. From their studies, it was concluded that the "wall effect" could be eliminated FIGURE 7. LIQUID DISTRIBUTION FOR THREE PACKIIGS OF THE SAME WTHOD IN. A 2.5 INCH TOWER WITH 0.25 INCH SEERFB. 60,000 55,000- LIQUID _ 50,000.- ' MASS “5,000- 0—— -p FLOW . ZONE ZONE I , “0,000.. 1 2 I’ F A... g . RATE 1' 35'0“” “:1 : LBS. HR.-SQ.F1'_'. 30,000- 25,000L . . . . 0 20 40 60 80 100 PERCENT OF LIQUID m 25. 26 by keeping the diameter ratio equal to or above 10. However, as shown by this data the "wall effect" was only suppressed, not eliminated by using a diameter ratio of 10. Since a higher ratio was not used,it cannot be said what the minimum ratio would be; however, Norman indicated that the minimum ratio should be 12. Therefore, it appears that this data would be comparable to his 12 to 1 ratio. b) Raschig Ring Packing Theoretically, it should be much more difficult to obtain similar packing arrangements for dumped beds of irregularly shaped packing such as Raschig rings. However, on the other hand, it should be easier to shift the arrangement by employing a specific dumping method. As a result, the 2.5 inch column was packed with 0.25 inch Raschig rings by seven methods. These methods included a random dumping into a water-filled tower; random dumping into a water-filled tower that was being drained to give the packing a faster fall velocity; dumping into a water-filled tower with water being rotated so that packing would be driven to and settle near the wall; dumping into a centered tube extending to the packing in a water-filled tower so that the packing would slide from center to wall during dumping; and dumping in annular area between the wall and a tube extending to the packing so that the packing would slide from wall to center in dumping. The liquid distribution for each case was plotted in Figure 8. In no caSe could any variation in distribution be contributed to the dumping method. Similarly, the 5.0 inch column was packed with 0.5 inch Raschig rings using five of the above packing methods. The liquid distribution for each case is shown in Figure 9. From this plot it was Shown that random 27 FIGURE 8. LIQUID DISTRIBUTION FOR SEVEN METHODS OF PACKING A 2.5 INCH- TOWER WITH 0.25 INGEE RASCHIG RINGS .. 90,000 80,000 . x {x I I X x LIQUID ’ x 70,000 b ZONE ZONE ZONE ZONE MASS . 1 2 3 4 60.000 - WALL CEN- TER 1mm x 3‘ fig ‘ T 50, 000 . i; RATE. .. 1 x “0,000 x I x x x LBS. - xx x HR“.-S"'Q""".EE. 30,000L 20,000 . . L . 0 20 “0 60 80 100 28 FIGURE 9. LIQUID DISTRIBUTION FOR SEVERAIIMETHODS OF’ PACKING A 5.0 INCHZTOWEfiiWITH 0.25 INCH RASCHIG RINGS. “5,000 METHODS 2. x-RPW‘, MOW. 0-ROD, RMw, PIT “0,000 . LIQUID , 35.000 - J ,J ,1 MASS , 1 4 7F 30,000 - I ZONE | ZONE 1‘ ZONE 1‘ ZONE now . 1 2 3 4 25.000 ' WILL I ° ° CENTER 0 RATE. Ii 9 I ' | O O 20,000r | l O ' O 188. | c ' H' R.‘ .-S" Q‘.'ET. | g 0 15,000- 4’ o '51: $ . o 1 10,000 ‘ 1 l I O 20 40 60 80 100 PERCENT. OF LIQUID MOW 29 30 packing methods resulted in liquid flow concentrating in the center. Using methods that allowed the packing to build up on the wall or in the center resulted in a fairly uniform distribution. The 5.0 inch column was also packed with 0.5 inch glass Raschig rings. However, only two packings were made since poor agreement was found even though the same method was used. This data is shown in Figure 10. Since there was some variation, an average liquid distribution was found and listed in the following table. Table 8. Average Liquid Distribution for Raschig Rings. Packing Diameter - d, Column Diameter - D. d D D/d Method (Wall) Zones (Center) (Inches) (Inches) 1 2 3 “ 1/“ 2.5 10 RPW 39.“ 23.6 20.8 16.2 1/“ 2.5 10 MCW 37.5 23.6 22.0 16.9 1/“ 2.5 10 ROD 39.2 22.6 2“.6 13.5 1/“ 2.5 10 BMW “0.“ 20.5 26.8 12.“ 1/“ 2.5 10 osw 39.9 26.2 17.8 16.“ 1/“ 2.5 10 PTC “1.“ 2“.6 20.1 13.6 1/“ 2.5 10 PIT 36.5 25.5 21.5 16.2 1/2 5.0 10 RPW 30.2 20.5 21.2 28.1 1/“ 5.0 20 RPW 21.5 23.1 23.7 31.7 1/“ 5.0 20 NOW 23.5 23.6 22.7 29.9 1/“ 5.0 20 ROD 2“.5 26.9 22.1 27.3 1/“ 5.0 20 BMW 2“.o 28.9 21.6 27.2 1/“ 5.0 20 PIT 23.2 26.2 23.1 27.3 FIGURE 10. LIQUID DISTRIBUTION FOR 0.5 INw RASCHIG RIMS. IN A 5.0 INm‘ TOWER. 45,000 “0,000 LIQUID 359.000 " X x X 1 Russ ‘ ‘ 30,000 I ZONE ZONE ZONE ZONE DION 1 2 3 4 25.000 1' WALL x I x I GEM 1 RATE 20,000 - LEE :5: x x HR.'§Q.ET. 15,000 b 10,000 L l ‘ . O 20 “-0 60 80 100 PERCENT OF LIQUID FLOWS 31 32 Here, it can be seen that a diameter ratio of 10 - 1 resulted in a "wall effect" or wall flow. When the ratio was increased to 20 - l, the liquid flowed to the center for random dumping methods. Again, the averaged data indicates that the center flow could be decreased slightly by employing a method that allowed the packing to build up on the wall or in the center of the tower. Thus, it was concluded that the wall effect could be eliminated with a diameter ratio CD/d) of 20 - 1. However, at this condition, a "center flow effect" occurred which might be suppreSsed by using the proper packing method. Since the variation in liquid distribution due to packing method was small, it would be recommended that these packing methods be checked on larger diameter towers employing larger packing. 0) Comparison of Liquid and Gas Distributions Morales, Spinn, and Smith have found the gas distribution for a 2 inch tower for 1/8, l/“, and 1/2 inch Raschig rings at various depths of packing. They observed that low velocities occurred in the center and near the wall of the tower as well as finding that the distribution varied with depth of packing. Since their 1/“ inch packing in the 2 inch tower should be comparable to 1/“ inch rings in this 2.5 inch tower, an attempt was made to correlate the distributions. This was done by determining the relative flow rate at a point in reference to the column radius (r0) for both the gas and liquid. The relative gas rate was found by dividing the point velocity by the average velocity (V/V avg.). The relative liquid rate was found by dividing the actual liquid mass rate for a section (G) by the average mass rate (G avg.) for the whole section. These adjusted rates were plotted and shown in Figure 11. The gas and liquid rates appeared to be proportional in the center of the tower and inversely proportional near the walls. Since this was based on only one comparison, it is recommended that the gas distribution be found for larger towers before any firm conclusions are drawn about the relation of gas to liquid distribution. 33 EIGURE II. COMPARISON. OF GAS AND LIQUID DISTRIBUTIONS FROM SEPARATE TOWERS PACKED WITH 0.25 INCH RASCHIO RIM-18. GAS DISTRIBUTION WAS EROM DATA BY MORAIES,, SPINN, AND SMITH FOR A SUPEREICIAL GAS VELOCITY OF 1.21 PT-./SEC. AND NO LIQUID DION: INA 2 INCH TOWER. LIQUID DISTRIBUTION WAS FOR A LIQUID MASS FLOW RATE or 15,000 IBS./ER.-SQ.IT. AND NO GAS FLOW. IN THE 2.5 INCH TOWER. 1.750 RATIO 1.500 _ 0E 1.250 - ACTUAL. 1.000 1- - GAB 0 Flow. RATE 0 T0 0.500 - AVERAGE 0.250 s mow RATE 1 1 OLA: . 4 1 ° 10° .8 O6 04 .2 o .2 O4 O6 08 1.0 RADIUS PROM CENTER/ RADIUS OF TOWER 3“ It?“ Conclusions The following conclusions were reached as a result of the preceding experimental work: 1) For the packing methods employed, it was not possible to change the liquid distribution by any large amount. 2) Liquid concentrates near the wall of towers having a ratio of column to packing diameters of 10 - 1. 3) Flow rate affects the liquid distribution; however, no trend was observed. “) Liquid flow reaches a fixed distribution within a depth equivalent to “ column diameters below a multipoint feed distributor. 5) Liquid distribution did not change as long as liquid rate was constant for most packings; however, 0.25 inch Spheres produced a distribution that fluctuated. 6) Data from two independent columns, one for measuring gas distribution with no liquid flow and another for measuring liquid distribution With no gas flow, the gas and liquid rates appeared to be proportional in the center and inversely proportional near the wall. 35 I. Table A. Experimental Data from 2.5 Inch Diameter Column Packed with 0.5 Inch Glass Spheres to a Depth of 36 Indhes. Run. Method N0. 01' Packing 1 l DRP 2 1 DRP 3 1 DRP 4 1 DRP 5 1 DRP 6 1 DRP 7 l DRP 8 l DRP 9 l DRP 10 2 DRP 11 2 DRP 12 2 DRP 13 2 DRP 14 2 DRP 15 2 DRP 16 2 DRP 17 2 DRP 18 1 POP 19 1 POP 20 l PCF 21 1 OSD 22 l OSD 23 1 OSD Run Time (Seconds) 43.0 45.5 43.4 41.8 40.0 36.0 36.7 30.0 29.0 41.0 40.0 37.0 36.5 27.3 41.5 34.0 31.0 41.8 32.2 27.0 36.3 35-0 27.5 Zone 1 4070 3900 3840 3800 3520 3700 3740 3670 3600 3730 3700 3960 3820 3610 3860 3660 4100 3850 3730 3830 3330 4050 4000 37. Zone 2 1440 1650 1710 1530 2170 1800 2000 2260 2230 1620 1680 2040 2090 2190 1780 2020 2480 1490 1480 1940 1260 1800 2220 Zone 3 1250 1230 1200 1140 2180 2000 1930 2550 2650 870 865 1710 1690 2240 740 1610 2580 1180 1500 2460 1200 1800 2690 water Collected (Milliliters) Zone “ 1080 1100 1120 1045 1410 1280 1300 1960 1730 1000 1000 1240 1200 1500 1040 1110 1600 1000 1040 1180 820 1090 1120 v-11111 it _____'__1 Table B. Experimental Data from 2.5 Inch Diameter Column Packed with 0.25 Inch Glass Spheres to a Depth of 20 Inches. Run Method Run Time Water Collected (Milliliters) N0. of (Seconds) Zone Zone Zone Zone Packing 1 2 3 4 24 3 DRP 42.5 2600 2570 2910 1400 25 4 DRP 46.7 2700 2690 2200 2380 26 4 DRP 35.8 1680 1400 1270 610 27 4 DRP 35.8 2200 1600 1230 1340 28 4 DRP 27.6 2180 1300 1780 1350 29 5 DRP 28.0 1160 1900 1170 1330 1 30 5 DRP 30.0 1350 1190 840 630 r‘r‘ 31 5 DRP 31.0 1530 1290 1210 1080 i 32 5 DRP 31 .9 2200 1440 1430 1370 g 38 Table 0. Run NO. 33 3“ 35 36 37 38 39 40 41 “2 43 44 45 “6 47 48 “9 50 51 52 53 5“ 55 56 57 58 59 60 61 62 63 6“ 65 66 67 68 HHHHHHHHHHHHHHHUUUNMMHHH###UUWMNMHHH Experimental Data from 2.5 Inch Diameter column Packed with 0.25 Inch Glass Raschig Rings to a Depth of 42 Inches. Method of Packing RPW RPW Run Time (Seconds) 34.2 36.2 26.5 35.6 32.2 26.5 34.2 34.0 28.3 33.6 34.3 27.0 34.2 35.4 25.“ 38.0 32.5 30.2 31.7 28.0 25.6 32.0 30.8 22.0 30.0 18.2 15.0 30.2 21.0 20.0 30.0 24.6 16.3 34.5 34.6 16.6 Zone 1 1940 3510 3700 1950 3540 4080 1940 2940 “000 1440 2780 3350 1780 3510 3340 1670 2800 4260 1540 2640 3490 1680 2640 2680 1570 1760 2100 1930 1960 2660 2160 2360 2130 2000 2930 1960 39 Zone 2 1330 1950 2200 1130 1200 1500 1500 2040 2040 1370 2140 2080 1220 1690 1790 1550 1900 2440 1100 1500 2000 1200 1200 1550 1010 820 940 1370 1330 1520 1320 1420 1220 1410 2080 1300 Zone 3 970 1400 1630 1110 1440 1760 1060 1800 2040 1410 1670 1730 1090 1460 1680 1480 1630 2230 1120 1530 1820 1240 1900 1480 1320 1120 1150 790 860 1290 910 1130 1210 1260 1540 1170 water Collected (Milliliters) Zone 4 950 1200 1480 1050 1140 1560 880 1000 1170 800 1060 1730 1050 1330 1660 1210 1040 1320 960 700 740 780 920 1000 730 420 520 860 720 1120 740 700 770 880 1260 880 Tab1e D. Experimental Data from 5.0 Inch Diameter Column Packed with 0.5 Inch Glass Spheres by Method DRP. Run Depth Run Time Water Collected (Millilitere) N0. of (Seconds) Zone Zone Zone Zone Packing 1 2 3 4 69 40“ 19.2 2020 2220 1860 2440 70 40" 21.1 2400 2300 2000 2060 71 40" 17.6 2100 2400 2600 3000 72 20“ 20.8 2000 1200 1200 1270 73 20" 19.6 2150 2260 1930 2030 74 209 17.5 2300 2220 2550 3180 75 304 24.4 2330 1200 1300 1240 76 30" 17.2 1840 1880 1740 1700 77 30” 17.2 2150 2270 2400 3180 78 40" 23.4 2240 1100 1240 1400 79 40" 17.9 2060 2060 1660 2240 80 40? 15.0 2000 1900 1900 2720 By Method PCF 81 40“ 22.0 2360 1450 800 1400 82 40" 21.0 2380 2340 2080 2180 83 40" 14.0 1840 2000 2000 2250 By Method 08w 84 40" 21.0 2100 1160 1160 1260 85 40“ 19.8 2300 2100 2300 2000 86 40" 14.0 1850 1750 1820 2360 40 Table E. Run Method N0. of Packing 87 1 RPW 88 1 RPW 89 1 RPW 90 2 RPW 91 2 RPW 92 2 RPW 93 3 RPw 94 3 RPW 95 3 RPW 96 1 MCW 97 1 MOW 98 1 MOW 99 1 ROD 100 1 ROD 101 1 ROD 102 2 ROD 103 2 ROD 104 2 ROD 105 1 RMW 106 1 RMW 107 1 RMW 108 2 RMW 109 2 RMW 110 2 RMW 111 l PIT 112 1 PIT 113 1 PIT Experimental Data from 5.0 Inch Diameter column Packed with 0.25 Inch Glass Raschig Rings to a Depth of 40 Inches. Run Time (Seconds) 21.0 21.0 12.8 20.0 18.2 15.6 21.5 20.7 14.3 23.0 15.5 15.“ 21.1 19.8 15.4 20.2 22.2 20.3 19.8 16.9 15.6 20.0 18.9 20.3 20.5 17.2 16.0 Zone 1 1060 1960 1600 1010 1680 1910 1020 2030 1700 1200 1700 1840 1400 1680 2310 1160 1850 2750 1250 1880 2140 1540 1800 2520 1400 1640 1860 41 Zone 2 1210 1960 1600 1140 1660 1880 1350 2200 1800 1480 1600 1620 1300 1800 2160 1560 2130 2750 1900 1920 2200 1840 1890 2600 1880 1660 1920 Zone 3 1220 2080 1680 1160 1750 1980 1350 2100 1880 1250 1700 1600 1000 1700 2060 880 2000 2620 1140 1620 2140 1260 1800 2200 1200 1840 1860 water Collected (Milliliters) Zone 4 1600 2440 2570 1880 2300 2500 1800 2550 2480 1680 1860 2500 1340 1700 2400 1620 2230 3180 1460 1780 2360 1420 2730 3160 1700 1840 2200 Table F. Experimental Data from 5.0 Inch Diameter Column Packed with 0.5 Inch Glass Raschig Rings to a Depth of “0 Inches. Run Method N0. 01' Packing 114 l RPW 115 1 RPW 116 1 RPW 117 2 RPW 118 2 RPW 119 2 RPW Run Time (Seconds) 23.1 17.4 16.0 20.0 42 Zone 1 2000 2100 3160 1840 2140 2740 Zone 2 1750 1780 2000 1300 1260 1300 Zone 3 1770 1600 2040 1200 1360 1800 Water Collected (Milliliters) Zone 4 1880 1630 2200 2100 2340 2640 References l. Leva, "Tower Packings and Packed Tower Design", Second Edition, 65-69. Norman, Trans. Inst. Chem. Engrs., London 29, 2, 226, (1951). Morales, Spinn, and Smith, Ind. Eng. Chem. 43, 225 (1951). Baker, Chilton, and Vernon, Trans. Am. Inst. Chem. Engrs. 31. 295. (1935). J. w. Mullins, Ind. Chemist 33, 408-17, (1957). 43 3604'? USE We?” zllllillllllllllllIlllliilwlllllljlfllijlm||1|)|H||||||||I