@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Applied Science, Faculty of"@en, "Chemical and Biological Engineering, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Bergeron, Georges"@en ; dcterms:issued "2011-12-15T02:15:57Z"@en, "1963"@en ; vivo:relatedDegree "Master of Applied Science - MASc"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """A study has been made of sampling techniques in a liquid-liquid extraction spray column using first a bell-shaped and a hook-shaped probe for the dispersed phase and for the continuous phase respectively. Later a piston method was attempted for the dispersed phase. The main interest in this research was the point concentration inside the column. At first, the time to reach steady state was considered in the absence of sampling. Later on, the rate of purging and sampling was varied for the probes up to 14.2 cc./min. for the continuous phase and 28.2 cc./min. for the dispersed phase. These rates were not sufficient to disturb the steady state. The measured point concentration was studied as a function of rate of sampling. Coalescence at the dispersed phase (bell-shaped) probe entrance did not take place. Finally, a piston sampler was set up and used in conjunction with the continuous phase (hook-shaped) probe as a second way to obtain point concentrations of the dispersed phase to compare with the results obtained with the bell-shaped probe. From these experiments, it can be concluded that sampling rate, varied from zero to 34.0 cc./min. for the continuous phase and from zero to 28.0 cc./min. for the dispersed phase, does not influence the point concentrations measured for column flows of 54.8 ft³/hr.-ft² and 72.4 ft³/hr.-ft² for the water and ketone phases respectively. The point concentration of the dispersed phase measured with the piston do not check definitively the results obtained with the bell-shaped probe; they do indicate that such agreement is fairly probable."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/39716?expand=metadata"@en ; skos:note "SAMPLING OF THE PHASES WITHIN A LIQUID-LIQUID EXTRACTION SPRAY COLUMN by GEORGES BERGERON B. Sco A 0, University of Montreal, 1961. A.THESIS SUBMITTED IN PARTIAL FULFILMENT THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE i n the Department of CHEMICAL ENGINEERING We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA November, 1963 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that the L i b r a r y s h a l l , make i t f r e e l y a v a i l a b l e f o r reference and study. I f u r t h e r agree that per-m i s s i o n f o r extensive copying of t h i s t h e s i s f o r . s c h o l a r l y purposes may be granted by the Head of my Department or by h i s r e p r e s e n t a t i v e s . I t i s understood that copying, or p u b l i -c a t i o n of t h i s t h e s i s f o r f i n a n c i a l gain s h a l l not be allowed without my w r i t t e n p e r mission. Department of CHEMICAL ENGINEERING The U n i v e r s i t y of B r i t i s h Columbia, Vancouver 8 , Canada. Date NOVEMBER, 1963. i i ABSTRACT A study has been made of sampling techniques i n a l i q u i d -l i q u i d extraction spray column using f i r s t a bell-shaped and a hook-shaped probe for the dispersed phase and for the continuous phase respectivelyo Later a piston method was attempted for the dispersed phaseo The main interest i n t h i s research was the point concentration inside the column. At f i r s t , the time to reach steady state was considered in the absence of sampling* Later on, the rate of purging and sampling was varied for the probes up to 14„2 cc./min. for the continuous phase and 28.2 cc 0/min. for the dispersed phase<> These rates were not s u f f i c i e n t to disturb the steady state. The measured point concentration was studied as a function of rate of sampling. Coalescence at the dispersed phase (bell-shaped) probe entrance did not take place. F i n a l l y , a piston sampler was set up and used in conjunction with the continuous phase (hook-shaped) probe as a second way to obtain point concentrations of the dispersed phase to compare with the results obtained with the bell-shaped probe. From these experiments, i t can be concluded that sampling rate, varied from zero to 34 00 cc./min. for the continuous phase and from zero to 28 o0 cc./min. for the dispersed phase, does not influence • •' 3 the point concentrations measured for columhi..-. flows of 54.8 f t / h r . - f t 3 2 and 72.4 f t / h r . - f t for the water and ketone phases respectively. The point concentration of the dispersed phase measured with the piston do not check d e f i n i t i v e l y the res u l t s obtained with the bell-shaped probe; they do indicate that such agreement i s f a i r l y probable. ACKNOWLEDGEMENTS The author would l i k e to express sincere thanks to S. D. Cavers for the assistance and constructive c r i t i c i s m s offered throughout the course of this project. Thanks go also to Canadian Chemical Co. of Edmonton, for supplying 1480 pounds of methyl isobutyl ketone. The production of a covering for the piston, to permit i t to operate without leaks, was s o l e l y the work of R. Muelchen. TABLE OF CONTENTS page INTRODUCTION 1 EXPERIMENTAL METHODS 11 Preliminary work 11 Procedure: 1.- probe method 19 2.- piston method 26 RESULTS 28 DISCUSSION 76 A) Study of time needed to obtain steady state 76 B) Effect of purging rate on the minimum purge time 77 C) Influence of sampling rate on steady state 78 D) Drop coalescence at the dispersed phase probe entrance.. 82 E) Effect of sampling rate on measured point concentration. 84 F) Duplication of runs 88 G) Concentration study i n the Elgin head 88 H) Piston results 90 SUMMARY 99 RECOMM EN DAT IONS 101 NOMENCLATURE 103 LIST OF REFERENCES 105 APPENDICES 107 I. Rotameter ca l i b r a t i o n s « 108 II . Runs data ..112 II I . Study of time needed to obtain steady state. Runs data .118 IV. Study of minimum purge time. Runs data 125 V. Point concentration versus sampling rate. Runs data ....135 i v T A B L E OF CONTENTS c o n t . V I . J e t c h a r a c t e r i s t i c d a t a 138 V I I . S a m p l e c a l c u l a t i o n f o r some k e t o n e i n t h e w a t e r p r o b e s a m p l e 140 V I I I . P o s s i b l e s o u r c e s o f e r r o r s 141 V LIST OF TABLES Table Page 1 Key to Figure 1 13 2 Location of Stainless Steel sampling probes 23 3 Concentration study in the Elgin head 57 4 Piston r e s u l t s .. • 73 6 Probe results corresponding to piston results of Table 4 74 5A Piston and probe results and the i r corresponding equilibrium values 75 6 Concentration in both phases for Run 8 and Choudhury's Run 65 89 7 Calibration of rotaaieter 108 8 Calibration of rotameter 110 9 Runs data 113 10 Over-all transfer data 115 10A Raw data of concentration p r o f i l e s for Runs 5D and 8 . 117 11 Study of the time needed to obtain steady state i n Runs 6 I l f GJ, 6M, 9A, 9C, and 91 119 12 Study of the time needed to obtain steady state in Runs 6K, 6L, 9B, and 9H 122 13 Summary of the steady state studies 124 14 to 23 Study of the minimum purge time for several runs 126 24 Summary of the minimum purge time 134 25 Point concentration versus sampling rate. Runs 9H, 7E, 7F, 5F, 5G, 5H, 5B and ID 136 v i LIST OF FIGURES Figure Page 1 Schematicjflow diagram 12 2 Stainless Steel tanks 15 3 Piston with polyethylene coating 18 4 Lever to push the piston . ... 20 5 Tip patterns for ketone nozzle.... 22 6 Steady state study, Run 61^ 29 6A Steady state study, Run 9A 31 7 Steady state study, Run 6 L 32 8 Summarizing plot of the steady state results 33 9 Study of the minimum purge time, Run 2 3G 10 Minimum purge time versus purging rate 37 11 Influence of sampling rate on point concentration,Run 9H 39 12 Influence of sampling rate on point concentration,Run 9H 40 13 Influence of sampling rate on point concentration,Run 5G 42 14 Influence of sampling rate on point concentration,Run 5H 43 15 Influence of sampling rate on point concentration,Run 45 16 Influence of sampling ra te on point concentration,Run 5F 46 17 Influence of sampling rate on point concentration,Run 7E 48 17A Influence of sampling rate on point concentration,Run 7F 49 18 Influence of sampling rate on point concentration,Run ID 50 19 Influence of sampling rate on point concentration,Run ID. 51 20 Influence of sampling rate on steady state, Run 7F „. • • 53 21 Influence of sampling rate on steady state, Run 7E .. » 0 54 22 23 Duplication < Duplication < of Choudhury*s of Choudhury's Run 65, Run 65, 56 58 v i i LIST OF FIGURES cont. Figure Page 24 Concentration study i n the Elgin head,.^ .-.,. 59 25 Effect of piston sampling on steady state, Run 9C 62 26 Effect of piston sampling on steady state, Run 9D 63 27 Effect of piston sampling on steady state, Run 9F 65 28 Calibration of ketone rotameter 69 29 Behaviour of the drops at ketone probe entrance 83 29A Two piston samples having d i f f e r e n t holdup 93 30 Rotameter cali b r a t i o n s 109 31 Rotameter cali b r a t i o n s I l l 1 INTRODUCTION The l a s t decade has seen the unit operation c a l l e d l i q u i d -l i q u i d extraction having a rapid growth. These new developments were due f i r s t l y to the i n d u s t r i a l users asking for more information which showed the ne*ed for research. Secondly, the interest of researchers in mass transfer operations caused the spray column type of l i q u i d - l i q u i d extractor to be analyzed by many workers (1,2, 3,4,5,6,7,8,9,10,11,12,13). Mass transfer c o e f f i c i e n t s were determined in general by using the i n l e t and outlet concentrations and calculating the logarithmic mean drivin g force. Several assumptions were made in doing such calculations. One of these was that the two l i q u i d s involved were not back-mixing. Another, was that no mass transfer took place during drop formation or during drop coalescence at the end of the column, or in other words, that end ef f e c t s were absent. It can be pointed out that d i l u t e solutions were assumed also (or constant flows)and constant slope of the equilibrium curved These were the p r i n c i p a l assumptions made before the work of many \\. investigators mentioned l a t e r , i n this survey. In 1950, Geankoplis and Hixon (1) employed a movable sampling device to remove internal samples of the continuous phase during operation of a spray tower. This sampler consisted primarily of a 5 mm. outside diameter (0.12 i n . inside diameter) glass tube, which extended into the extraction section and occupied approximately 1.7% of the column cross sectional area. By means of a hook at the end of the sampler, the descending continuous phase was withdrawn; the sampler tube touched the wall of the column. These workers determined the continuous phase concentration p r o f i l e throughout the column and located a large end effect at the i n l e t of the continuous phase. No end effe c t was found at the dispersed phase i n l e t . Later on, Geankoplis, Wells and Hawk (2), using the same method, found the same large end effe c t s at the continuous phase i n l e t , and also that these depended neither on the type of system employed, nor the di r e c t i o n of solute transfer. They also proved that the effe c t of inter n a l sampling i s small on the material balances. Also, the effe c t of internal sampling was small since the deviation in the ov e r - a l l material balance, calculated as follows: N = A L (C - C ) w w w^ N t = A L t ( c t 2 - \\ > N = N + N w t_ 2 % Deviation = N - N w N * 100 averaged 4 % or less for a l l runs i n the towero In 1952, Newman (3) entered the discussion and maintained that the end effects found by Geankoplis and Hixson (1) and by Geankoplis, Wells and Hawk (2), were the results of v e r t i c a l mixing of the continuous phase due to the movement of the drops. He showed that the results of Geankoplis and coworkers (1,2) were midway between the results to be expected for t r u l y countercurrent, unmixed flow of the continuous phase, and the completely uniform concentration in that phase which would b^ e produced by very e f f i c i e n t s t i r r i n g . One year l a t e r , Geankoplis with Kreager (4) studied the effec t of the column height on the mass transfer c o e f f i c i e n t . Their conclusions were the same as to end effectsfat< the\" . .C continuous phase i n l e t which Geankoplis and Kreager (4) suggest to be due to the cocurrent flow of continuous phase in the form of \"atmospheres\" of continuous phase surrounding and t r a v e l l i n g with the drops of dispersed phase. In the same year, Gier and Hougen (5) used two sampling techniqueso In the column i t s e l f , nine holes were d r i l l e d 6 inches apart along the length of a 6 i n . I.D. column to provide an entrance for 3 i n . hypodermic needles which sampled the continuous phase. Each needle had a thin-walled brass tube bent into a hook shape so that the open end of the hook faced in the d i r e c t i o n of drop: movement. These sampling needles f i t t e d 29 mL hypodermic springes. Later, provision was made for sampling the dispersed phase by d r i l l i n g eight holes along the length of the column, opposite to and midway between the needle holes already mentioned,, The dispersed phase samplers each consisted of an inverted 1 i n . glass funnel connected by Tygon tubing to a suitably formed length of V* i n . copper tubing. They found the concentration p r o f i l e s for the continuous phase si m i l a r to those ' measured by Geankoplis and co-workers (1,2,4) and showed that the material balance equation was not applicable when written as follows: dN = 1 dC , c c although, i t was true that: dN = L. dC^ d d . (dN being the mass transfer across the interface between phases i n a height dh of column)o This statement i s correct even i f the same equations, integrated, both hold for terminal conditions. The reason for the f i r s t equation f a i l i n g i s a serious bulk mixing of the continuous phase. The second equation holds because the drops do not back-mix 0 They also concluded that because of the mixing, the measured height of a transfer unit ( as defined l a t e r by Miyauchi (6)$ must be determined graphically from internal conditions. Furthermore, i t i s pointed out by Smoot and Babb (7) that i f longitudinal mixing i s extensive, even a graphical integration of the r e s u l t i n g p r o f i l e to obtain the measured Nox may not y i e l d the \"true\" Noxo This i s equivalent to stating that i n the equation: Nox = (Nox),. + Correction term even a graphical integration using the experimental concentration p r o f i l e and this equation: (Nox).. = dx M — — — — . X - X e w i l l not y i e l d the \"true\" 'value of Nox as defined by this equation: Nox = K a h x V x unless the t h i r d term of the f i r s t equation mentioned here, i s negligible.. But this t h i r d term i s a term which corrects for back-mixing i n the x phase. Then, i f the x phase i s the dispersed phase, Nox and (Nox),. are i d e n t i c a l , since the correction term goes to zero M because of no mixing i n that phase.. In 1954, Heertjes, Holve and Talsma (ti) measured the concentration of the continuous phase of a spray column by draining the column in stages at the end of runs and sampling the effluents. The concentration of the continuous phase appeared to be nearly constant over the column's height. These results appeared to show the existence of good v e r t i c a l a x i a l mixing but they do not agree with those of Geankoplis and co-workers (1,2,4) where the concentration of the continuous phase was not constant with distance. This difference i n the res u l t s may be due to mixing, taking place before draining and caused by the currents produced by the f i n a l r i s i n g of drops a f t e r 5 flow i s shut off; perhaps the difference may be also due in part to mixing caused by draining, and also to convection due to density differences. The foregoing summarizes the information available on sampling techniques before the star t of the research i n Canada by Cavers and Ewanchyna (9). These workers studied c i r c u l a t i o n i n the continuous phase of a spray column and end ef f e c t s . Their techniques consisted of using movable v e r t i c a l probes based on those of Geankoplis and co-workers (1,2(4) and those of Gier and Hougen (5). The f i r s t set of inter n a l sampling tubes was constructed of Pyrex tubing, l / 8 - i n . OoD0 and 0.08-in. I.D.. The continuous phase sampling tube had one end curved i n the form of a hook with a 3/16-in,, I.D„ radius. The probe entrance faced upward, away from the r i s i n g drops. The dispersed phase sampling tube had one end flared out to 34-in. O.D. i n order to catch the ketone d r o p s o The maximum percent of the internal cross-sectionnal area of the column occupied by the two probes taken together was approximately 4.2%6 (The area of two 1/8-in. diameter c i r c l e s and one Ki-in. diameter c i r c l e i s 4.2% of that of a 1.5-in. diameter c i r c l e ) . These Pyrex sampling tubes did not prove to be very s a t i s f a c t o r y because of breakage problems; however, the i r » functioning i n other respects was good. A second set of sampling tubes was made out of 1/8-in. O.D. and 0.020-in. wall thickness Type 304 sta i n l e s s s t e e l seamless tubing. The continuous phase sampling tube had one end curved i n the form of a hook with a 3/32-in. I.D. radius. The opening faced the top of the column as before. The dispersed phase sampling tube was f i t t e d into a flared out section of 2l/64-in 0 maximum 0.Do. In th i s case, the maximum percent of the internal column's cross-sectionnal area occupied by the two probes 6 taken together was 6 02% 0 These probes were lowered into the column from the top. This approach therefore was si m i l a r to that used for the continuous phase only ?by Geankoplis and co-workers (1,2,4), but s l i g h t l y d i f f e r e n t from that of Gier and Hougen (5). Cavers and Ewanchyna (9) confirmed that the presence of sampling tubes i n the column did not change the column operation. They also demonstrated that the dire c t i o n of sampling ( i . e . from top to bottom or bottom to top of the column) did not influence the measured point concentrations. It was also proved that the location of the tubes in the horizontal cross section had no effect on the concentrations of samples withdrawn. These workers found no end effects .where the dispersed phase entered the column, but did find end effects at the continuous phase i n l e t to the column, i . e . d i s c o n t i n u i t i e s i n the concentration p r o f i l e s . For the case of transfer of acetic acid from a continuous aqueous phase to a dispersed, methyl isobutyl ketone (M.I.B.K.) phase, they explained that the discontinuity in the water phase p r o f i l e could be broken into two parts: one representing the effect of drop agitation at the interface, and the other the effe c t of back-mixing in the aqueous phaseo The discontinuity i n the ketone phase concentration p r o f i l e was attributed to the agitation e f f e c t . For transfer of acetic acid in the reverse d i r e c t i o n , the discontinuity in the ketone phase p r o f i l e was absent, and that in the water phase was attributed s o l e l y to back-mixing i n the continuous phase. Some years l a t e r , the work of Cavers and Ewanchyna (9) was continued by Choudhury (10) who used an apparatus s i m i l a r to that of Ewanchyna(11)o He did work also on the sampling technique. He investigated the minimum purging time required to remove material of the wrong concentration from the probes. Furthermore, he made these measurements at various rates of flow through the probes. However, he did not get an adequate plot for the minimum purge time, but instead only a few scattered pointso Choudhury mentioned also that the drops sometimes coalesced at the ketone sampler's entrance. He had planned to check the effe c t of such coalescence on the results for the dispersed phase sampler by varying the sampling rate for the dispersed phase. Also, a material balance has to be made i n interpreting the results of analyses of dispersed phase samples. The problem arises as to whether or not the hook probe gives a representative sample of the continuous phase concentration for use in this balance. (Presumably the b e l l probe, i n removing drops, p r e f e r e n t i a l l y removes continuous phase from near the drops). Varying the sampling rate of both phases might change the re s u l t s by changing the average continuous phase concentration taken i n by one or both of the probes. Choudhury varied the sampling rate of $fc>th phases to measure i t s influence on the concentrations measured at a point. These experiments were made at a cross-section where the phases were not at equilibrium even i f Hawrelak (12) thought so. This statement i s made because the point concentrations reported at 6.16 f t from the nozzle were 8.2% from equilibrium as compared with 19.7%, the maximum deviation between the equilibrium concentration p r o f i l e (plot of C^*) (10) and the i concentration p r o f i l e obtained. Choudhury's results were night, and they did throw l i g h t on the problem of C being representative for W i the conditions under study. However, the range of rates covered by Choudhury (10) was too small. Choudhury (10) found concentration p r o f i l e s which had the same pattern as those measured by the e a r l i e r workers (1,2,4,9,11) * This percentage i s calculated as ( C, . - C. . ) x 100 f a k i k i 8 He fixed for himself the maximum permissible purging or sampling rate to be 15 cc./min., without checking the effect on steady state. Hawrelak (12) designed and constructed a piston-type sampler used in conjunction with hypodermic needles, or with the continuous phase (hook-shaped) probe, to sample the continuous phase, as a second, but not completely independent, way to obtain point concentration of the dispersed phase to compare, with:the results obtained with \" v • the bell-shaped probe. The hypodermic syringes were used to take samples of the continuous phase immediately above and below the piston block; the needles entered the column through theasbestos gaskets sealing the glass column to the piston block. It was discovered l a t e r that an appreciable volume of ketone phase was entrained i n the continuous phase samples taken by the syringes; the syringe samples were discontinued because of the near imp o s s i b i l i t y of correcting for transfer between two phases in the continuous phase sample. It was decided that the water phase probe sampler, which showed no ketone entrainment, could be used i n place of the syringe samples, i f i t was assumed that the probe sample gave a representative sample of the water phase. This probe was to be used to sample at the axis of the piston. A minimum leakage rate around the piston of one to two mis/ min. was encountered a f t e r a few seconds; but this leakage occurred only when the piston was shoved to the l e f t i n the block. Sampling was accomplished by shoving the piston from i t s right to i t s l e f t hand position i n the block. The l'/fc-in. I.I), hole i n the piston that was to s l i d e into l i n e with the column was f i l l e d with either water phase which had leaked from the piston, or with outlet water phase from the column i f there was no such leakage. In t h i s way when a piston sample was taken the column continued to operate with no appreciable disturbance. 9 It was inte r e s t i n g to observe that when a piston sample was taken (portion of column contents removed) a gap i n the ketone phase occured, which appeared to move up the column for one to two feeto The same d i f f i c u l t y ( as spec i f i e d e a r l i e r ) arose about the material balance on which the interpretation of the dispersed phase samples depends. To be more s p e c i f i c , does the hook probe gives a representative sample of the continuous phase, the concentration of which has to be used i n this balance? In addition, a n a l y t i c a l d i f f i c u l t i e s were encountered due to the small dispersed phase holdup obtained. (Holdup = volume percent dispersed phase in the column.) This d i f f i c u l t y arose in using the material balance equation written as follows for the piston sampler: C = C - V ( C - C ) k i kf w wi wf Vk It i s evident that the volume of the ketone phase in a sample should be large so that the equilibrium concentration of solute i n the water phase i s considerably dif f e r e n t from the i n i t i a l value. Only then does one avoid the error re s u l t i n g from having to substract quantities which are of about the same magnitude i n applying the equation. (This approach assumes one analyzes the piston sample at equilibrium. Other p o s s i b i l i t i e s are discussed later.) The maximum holdup used by Hawrelak (12) was 12.7%, 3 3 corresponding to a value of (C . - C . ) * of only 1.2 lb.-moles/ft xlO ° wi wf 3 3 where C _ was approximately / 40.0 lb.-moles/ft.xlO . It i s d i f f i c u l t wf to compare Hawrelak's results with the piston, with those obtained with the bell-shaped probe, because of these a n a l y t i c a l d i f f i c u l t i e s , because of 'th© leakage mentioned e a r l i e r , and also because of an error hie made 10 in positioning the continuous phase probe„ This was one inch higher than the axis of the piston i n several runs. However, the piston method appeared to give the calculated i n i t i a l ketone concentration always lower than the concentration measured with the bell-shaped probe. The present work i s a continuation of that of Ewanchyna (11), that of Choudhury (10), and that of Hawrelak (12). A l l three were concerned, at least i n part, with sampling of the phases. It was planned to attack the problem by f i r s t varying the sampling rate used with both sampling probes and noting any change i n the dispersed phase concentration r e s u l t s , and secondly by operating, the piston sampler under conditions of very high holdup of dispersed phase in the column, when the a n a l y t i c a l problems associated with the piston sampler should be comparatively small 0 The p o s s i b i l i t y of the hook-shaped probe not providing representative samples must be considered with respect to both sampling methods, and i f these samples were not representative, varying the sampling rates used with each of the probes might produce changes in the r e s u l t s . However, i f i s considered unrepresentative, and i t could only be too high, then C^/ comes out too low: C. , = C, . - V ( C . - C . ) k i kf w wi wf Vk Hence, (C. .) . , i s too low: then i t w i l l not check with the probe k i piston sampling resu l t s which were obtained at high holdup where the correction term, V /V, (C . - C „), i s less important. (Recall that the piston w k wi wf i samples were obtained with low holdup). It i s interesting to*note that i f values were higher, a l l the water p r o f i l e s calculated using piston flow (10), would be higher, indicating more backmixing than that assumed present heretofore. 11 EXPERIMENTAL METHODS A) Preliminary work. The apparatus arrangement (Fig. 1) was s i m i l a r to the one used by previous workers (10, 11, 12, 13) except for a few modifications to be mentioned l a t e r , A general reorganisation of the apparatus and a cleaning were needed because the equipment had been p a r t i a l l y disasembled to f a c i l i t a t e the move of the Department of Chemical Engineering into a new building. A rotameter had to be reconditionned. The aluminum tanks used by Choudhury (10) and Rocchini (13) were corroded. Before replacing them a survey was made of possible materials to contain M.I.B.K. and aqueous solutions of acetic acid. It was found that Pyrex and stainless steel were the only suitable materials which were also r e a d i l y available at a reasonable p r i c e 0 Taking into account the safety requirements, four stainless steel tanks were chosen and made as spe c i f i e d on Figure 2. To prevent any leakage of M.I.B.K. (flash point 75°F (15)) from the feed and storage tanks, stainless steel tubing and f i t t i n g s were used to connect both tanks to the pumps. Polyethylene tubing (with some stainless steel f i t t i n g s and some Sa ran f i t t i n g s ) was used for the water phase. Corrosion information was obtained from i n d u s t r i a l suppliers of materials for tanks and tubing. The suppliers reported that polyethylene was not recommended to be used with M.I.B.K. except where the M.I.B.K,, phase was flowing intermittently. A stagnant M.I.B.K. phase can destroy the properties of polyethylene by taking out the p l a s t i c i z e r . 12 13 TABLE lo KEY TO FIGURE 1 A - Continuous phase feed tank B - Continuous phase receiver and storage tank. C - Dispersed phase receiver and storage tank D - Dispersed phase feed tank E - Continuous phase constant head tank F - Dispersed phase constant head tank G - Continuous phase rotameter H - Dispersed phase rotameter I - Continuous phase i n l e t sample valve , - Continuous phase flow rate control valves K , K - Dispersed phase flow rate control valves L - Dispersed phase i n l e t sample valve M - 6-in. I.D. Pyrex top end section N - Continuous phase i n l e t s t a i n l e s s steel pipes 0 - Drain valve for top end section P^t - Centrifugal feed pumps for continuous and dispersed phases respectively Q - Level of interface R - Column proper l1/&-in. I.D. Pyrex S - Dispersed phase nozzle T l ' T 2 ' T 3 ' T4 \" T h e r m o m e t e r s U - Special Pyrex reducer 3-in. I.D. to lV&-in. I.D. V - Vent connected to a l i n e going outside building W - Pressure equalizing vent X - Control for interface l e v e l Y - Valve for draining the column Z j , - Outlet sample valves for continuous and dispersed phases respectively 14 TABLE lo CONTINUED PTS - Piston type sampler PS^, PS^ - Piston sample exit ports a - Dispersed phase sample probe b - Continuous phase sample probe c - T r a v e l l i n g block from which sampling probes are suspended d - Guide on framework for block \"c\" e - Continuous phase sampling rate control valve e' - C a p i l l a r y tubing f - Dispersed phase sampling rate control valve f - C a p i l l a r y tubing h - Continuous phase sample bottle k - Dispersed phase sample bottle m~- Mercury manometer n - Water aspirator (vacuum controlled by a i r vent at the bottom of a mercury column and also by \"r\") r - Valve for releasing vacuum 15 FIGURE 2o STAINLESS STEEL TANK 16 On the grounds that polyethylene was less expensive than s t a i n l e s s s t e e l , and that polyethylene was probably at least f a i r l y r esistant to M. 1.13 .K., polyethylene tubing was i n s t a l l e d to replace Saran tubing which had been attacked by M.I.B.K. (13) 0 However, to prevent mechanical breakage and to give more operating freedom to the author, s t a i n l e s s s t e e l tubing was used l a t e r , i n places where the M.I.B.K. phase was flowing continuously when the solution were being mixed or reci r c u l a t e d with the column proper not i n operation. A l l rubber tubing was removed from the sampling l i n e s to prevent any runnjng back of l i q u i d , possibly rubber contaminated, into the column. F i n a l l y , polyethylene tubing was i n s t a l l e d a l l the way from the sample receiving flasks to the stainless s t e e l probes. I n i t i a l l y , the M.I.B.K. had to be d i s t i l l e d as sp e c i f i e d by the pamphlet \"Ketones\" (14) due to po l l u t i o n from the use of Saran tubing by e a r l i e r workers (12,13). Also as spec i f i e d by the pamphlet, the M.I.B.K. d i s t i l l e d over was collected only between the temperatures of 114°C and 117°C measured just before the vapor l e f t the d i s t i l l i n g flask by the side opening. A l l the available M.I.B.K. was p u r i f i e d i n this way. A black precipitate was l e f t i n the d i s t i l l i n g f l a s k . By doing a chromatograph test, the purity of the d i s t i l l e d material was obtained. Preceding this v e r i f i c a t i o n , a chromatographic analysis was run for the M.I.]\\K. reserved from the previous drums, sp e c i f i e d as being 99.0% pure by the suppliers. One peak exists for the pure material (as received) while three peaks were present for the d i s t i l l e d material, which was composed of M.I.B.K., acetic acid, and d i s t i l l e d water. This result showed that no undesirable impurities were l e f t in the d i s t i l l e d material, i f the assumption i s made that the observed peaks corresponded to the three substances mentioned. This assumption was not tested. 17 The repairing of a rotameter was necessary. New s t a i n l e s s s t e e l plates that compress the packing around the tube were made to replace those of mild s t e e l which had been corroded by leaking m a t e r i a l 0 Also, the scales had been shif t e d r e l a t i v e to the tubes when scales were removed and replaced. Both meters had to be recalibrated and the resu l t s are recorded i n Appendix I. The second part of this work was done using a piston sampler designed and constructed by Hawrelak (12) i n i t i a l l y . The piston-type sampler, PTS i n Figure 1, was flanged to the glass column by means of standard Corning Type I flanges and hard asbestos gaskets. In a l l cases the piston axis was 1.59-ft. above the nozzle t i p s . In normal operation of the column, the phases pass through one of the holes i n the piston. The p r i n c i p l e of this sampling method consists of taking out four inches of the column's phases by cutting through the operating column with the piston which contains v e r t i c a l l y d r i l l e d passages of the same inside diameter as the column. Piston samples were collected at points PS^ and PS 2 i n Figure 1, into volumetric flasks by slamming the piston from one side to the other of i t s t r a v e l . Past experience with a hard-chromed phosphor bronze cylinder block and a brass piston coated with soft solder, and also with an aluminum piston, was that there were always leaks (12). An attempt to stop this leakage was made by covering the piston with a polyethylene sheet, holding i t mechanically as shown on Figure 3. This approach solved the problem of leaks between the piston and piston block; also this made i t unnecessary to consider periodic replacement of the soft solder piston surface due to corrosion or mechanical damage. The polyethylene sheet was attached mechanically because cements can not 18 1/8 HIGH MOLECULAR W POLYETHYL SHEET THICKENED SECTION OF POLYETHYLENE COVER T 6.0 ALUMINUM PISTON 0.0015 OD : 4.0805 0.0015 1/4^1 Jut > 1 / 2 / \" \" \\ \\ / z <= H II r i / 4 I I 11 1 1 I ' 1 1 I t 1A. -l-ll I rl I I I I I I I I L . . I FIGURE 3. PISTON DESIGN 19 be used due to the danger of solution contamination, p a r t i c u l a r l y with surface-active materials; i n addition, most cements are soluble in M.I.B.K. (15) 0 A lever was b u i l t to replace the hand driving handle. The piston diameter tolerances had been reduced by increasing the piston diameter with the polyethylene sheet to help reduce the leaks; but as much f r i c t i o n had to be overcome', a lever mechanism to drive the piston was required 0 Figure 4 i s a sketch representing the lever used to push the piston into the block. B) Procedure 1- Probe method In a l l the runs performed, mass transfer took place from the continuous phase to the dispersed phase. The continuous phase flowed by gravity downward from the top of the column while the dispersed phase was fed to the bottom of the column through chamfered nozzles as droplets. The required flow rates were set by means of the rotameters; during the time needed to reach steady state, the interface c o n t r o l l e r was adjusted to hold the interface at one p a r t i c u l a r l e v e l . Throughou a l l the runs, the interface remained two'.tofour inches below the top of the top plate of the Elgin head. During each run, the interfac elevation remained within a range of %-in. but varied over one inch range from run to run. Thus the height of the column (nozzle t i p s to interface) i s reported as 7-ft. and 41/£-in. £ i/i-in0 for a l l runs. Throughout the course of a run, numerous checks were made on the interface height and on the flow meter settings. It was found that l i t t l e or no readjustment was necessary. FIGURE 4. LEVER TO PUSH THE PISTON The velocity through each of the nozzle ti p s was held at a constant value of 0.357 f t . / s e c , except for a few runs. To keep this l i n e a r v e l o c i t y of the dispersed phase through the t i p s constant with varying t o t a l ketone flow rates through the column, some of the t i p s were blocked out by using Teflon caps. Accordingly, the number of open tip s changed with ketone flow rate. Figure 5. shows the various t i p patterns used. The continuous phase flow rate was held constant throughout this work. The jets of f l u i d obtained at the nozzle ti p s would produce uniform drops according to the Johnson and B l i s s c o r r e l a t i o n (16), as reported in Appendix VI 0 These authors claimed that two dif f e r e n t kinds of drop formation can take place at the nozzle depending on the v e l o c i t y through the nozzle. They mentioned that two regions exist: one region below which drops cease forming at nozzle ti p s and another above which drops cease being uniform. The corr e l a t i o n of Johnson and B l i s s (16) was used as a guide to determine whether the v e l o c i t y used i n this work was low enough to be within the reported region of Uniform drop s i z e . However, the work of Rocchini (13), i n which a nozzle t i p v e l o c i t y of 0.362 Ft./sec. was used, shows that i n the present study various drop sizes must have been present. The d i s t r i b u t i o n was not measured but would be expected to be close to that of Rocchini (13). The probe method of obtaining point concentrations used by e a r l i e r workers (10,11, and 12) needed to be checked. Samples in the present work were taken i n the upward di r e c t i o n : from nozzle t i p s to interface, except when the purging and sampling rate studies were done The sampling was done at distances from the nozzle which were the same as those used by Choudhury (10) and shown in Table 2. * T h i 6 v e l o c i t y was not used for probe samples taken for use with piston samples. BLOCK OUT 2 2 l_K = 72 .7 F T ^ H R F T 2 V*= 0.357 F T / S E C '0 p^:q q o p O Q ; o L K = 120.3 F T / / H R F T 2 V*= 0.338 F T / S E C FIGURE 5o TIP PATTERNS FOR KETONE N0ZZLE„ V* Linear v e l o c i t y through t i p s i n ketone nozzle ft./sec» TABLE 2 LOCATION OF STAINLESS STEEL SAMPLING PROBES (TUBES). Point number Distance above nozzle t i p s , f t . 1 0.078 1A 0.161 2 0.445 2A 1.161 2B 0.911 3 0.755 3A 2.161 3B 1.661 4 1.060 4A 3.161 4B 2.411 5 1.379 5A 1.355 5B 4.161 5C 3.786 6 5.161 7 6.161 8 7.286 24 This approach was decided on to permit easy comparison with the work of Choudhury (10) which had the object of obtaining concentration p r o f i l e s . The probe sample taken were received into clean dry Erlenmeyer f l a s k s . These were closed immediately to prevent as much evaporation as possible. Each sample was analyzed either on the same day as the sample was taken, or, at the worst, not l a t e r than the next day 0 The volumes of the phases were measured by pouring the mixture co l l e c t e d i n a volumetric flask into a graduate. From each of the continuous phase samples, a volume of 10 ml. was measured with a pipette and analyzed by t i t r a t i o n with approximatively 0.1 N sodium hydroxide with phenolphthalein indicator. A s i m i l a r method was used for the dispersed phase samples, but before t i t r a t i o n of 10 ml. from each ket one layer, 25 ml. of SDAG-IK mixture were added. (SDAG-1K mixture i s made i n d u s t r i a l l y by mixing 100 gallons of dehydrated ethyl alcohol with 5 gallons of dehydrated methyl alcohol.) The result was a homogeneous, single phase in which the end point could be determined e a s i l y (or much more e a s i l y than would have been true i f two phases had been present). Sometimes a f t e r the analysis had been completed, a ketone layer was v i s i b l e on the surface of the mixture i n the f l a s k . Perhaps a larger volume of alcohol should have been used. A blank solution was prepared for the case of the dispersed phase analyses. The blank solution was the same as a t y p i c a l solution analyzed except that the 10 ml. dispersed phase were not added. Less than a drop of 0.1 N sodium hydroxide was needed to change the color of the blank so l u t i o n . This amount of t i t r a t i n g agent was considered ne g l i g i b l e and not substracted from the t o t a l amount needed to t i t r a t e the dispersed phase samples. Sodium hydroxide and g l a c i a l acetic acid were reagent grade 25 ( A.C.S. s p e c i f i c a t i o n ) and were obtained from Nichols Chemical Co., Ltd, Montreal. Laboratory d i s t i l l e d water was used for a l l runs in this projecto The M.I.B.K. was technical grade furnished by the Canadian Chemical Co., Edmonton, except that i n the e a r l i e r runs (Run 1 up to and including Run 3) the d i s t i l l e d ketone, described i n the preliminary work, was used, because the suppliers were not ready to deliver fresh M.I.B.K.. However, as soon as the fresh chemical arrived, the apparatus including the tanks were washed with t h i s M.I.B.K.. For a l l runs, the rate of transfer of acetic acid across the interface i n the'column was calculated i n lb.-moles/hr. by two di f f e r e n t equationso One was based on the t o t a l change i n concentration of the water phase, and the other on the corresponding t o t a l change i n concentration of the ketone phase. The equations were: N = L A ( C - C ) 1 W W W^ W,j \\ - *1c A ( C k 2 - V . 2 These were applied without including the volume of samples taken out by the probes. Values of and N^ were s l i g h t l y d i f f e r e n t in general, and an average value was determined by use of the following expression: N s N + N, 3 w k 2 This value was estimated to be better than either N or N, for further w k ca l c u l a t i o n . The percentage deviation was calculated for each run as a measure of the quality of the experimental work. The equation used was: Percentage deviation = (N - N, ) A N * The fresh M.I.B.K. then was used for a l l runs from and including run 3A. 26 The point concentrations i n the units of lb.-moles/ft^ were re a d i l y obtained by t i t r a t i o n . With these t i t r a t i o n r e s u l t s the following equation was used to calculate back to the dispersed phase concentration at the time of sampling: C. . = C. _ - V ( C . - C _ ) 5 k i kf __w wi wf Vk As far as the probe method i s concerned, the general calculations stopped at th i s point because the emphasis was put on the point concentration, rather than on H. T. U. as in previous work. 2. Piston method. When the piston method was used, the extraction process was the same as mentioned e a r l i e r and used with the probe method. But, when the piston sampler was incorporated into the apparatus, the procedure was s l i g h t l y d i f f e r e n t . When steady state conditions were achieved, the probes were lowered to the elevation of the axis of the piston where purging and sampling of both phases took place simultaneously. Purging and sampling rates and purging times were used as prescribed by the f i r s t part of the present work and recorded on Figure 10. Inlet and outlet samples of both phases were taken regularly from the start to the end of a run and consequently when the probe samples were being taken. When s u f f i c i e n t volumes of both phases had been obtained, the probes were removed from the path of the piston. The l 1/^-in. I.D. hole i n the piston that was to s l i d e into l i n e with the column was f i l l e d with outlet water phase from the column. In thi s way, when a piston sample was taken, the column continued to operate with no appreciable disturbance. After a piston sample had been taken, s u f f i c i e n t time was allowed to restore the steady state, i 27 which time had been studied as a part of t h i s work. The p r i n c i p a l change made in the extraction process consisted in increasing the dispersed phase flow rate to give a larger holdup of dispersed phase in the column, and, consequently, in the piston sample. Figure 5. shows the setup to the nozzle necessary for this change of flow. It was planned to use the l i n e a r v e l o c i t y of 0.357 ft. / s e c . as mentioned e a r l i e r . However, the flow rate for 21 t i p s to give this velocity, was calculated using the nominal nozzle diameter of 0.10-in. instead of the actual average diameter of 0.1029-in. (10). The actual velocitiescalculated on the basis of the l a t t e r diameter are recorded in Table 10, Appendix I I . An acetic acid material balance has to be made between the i n i t i a l conditions prevailing i n the column at the time of sampling with the piston, and the f i n a l concentration existing i n the removed piston sample at the time of analysis. The equation used was the same as Equation 5 just given. The quantity C ^ i s found by using the concentration given by the continuous phase probe at the axis of the piston just before taking a piston sample. When the l a t t e r sample i s taken, C, „ and C _ can be measured and also V, and V . Then C. . can be ' kf wf k w kx calculated: the quantity r e a l l y needed. This c a l c u l a t i o n i s permissible on the basis that: k i kf k and V . = V = V . wi wf w These l a s t two statements probably are very close to the truth, and are assumed to be correct within the experimental error of a run. It should also be understood that the concentration i s assumed to vary l i n e a r l y inside the four inches long piston. This assumption i s based on previous results(10,11) which show that the concentration versus the 27A distance from the nozzle i s almost l i n e a r over seven feet of the column's length. 28 RESULTS A)S.teady state requirements: It was discovered that the concentration inside the column influenced the length of time needed to achieve a steady state i n making an extraction run. The concentration 6\"f the column f i l l e d with the continuous phase feed solution d i f f e r e d from that existing inside the column which had been operated before; t h i s difference was due to the extraction process which took place. Then, two ways of s t a r t i n g a run needed to be investigated. The f i r s t one consisted i n using a column f i l l e d only with the c o n t i -nuous phase at feed concentration and with no dispersed phase above the interface. ( In run 61^ (Fig. 6) a s l i g h t variation i n this procedure took place: some d i s t i l l e d water had been l e f t i n the Elgin head from the back-washing operation which preceded). The second way was the start-up of a run using a column f i l l e d with the continuous phase l e f t i n the column at the end of a previous run and with the corresponding dispersed phase above the interface. To obtain the time needed to reach a steady state condition for the f i r s t way of startup investigated, the column f i r s t was f i l l e d with continuous phase feed solution only, up to the interface l e v e l i n the Elg i n head. A flow of dispersed phase then was started. Samples of both feed solutions were taken three times during the extracting operation whereas the outgoing solutions were sampled every five minutes-. Figure 6 shows the results of a run using an i n i t i a l continuous phase concentration s l i g h l y d i f f e r e n t from 50.4 lb . moles,'of ^ .acetic acid/ cu. f t . of water due to the presence of d i s t i l l e d water i n the Elgin head before the column was f i l l e d with ro* 27 • I — CO w 2 6 o 3 25 O I 24 O 23h 3 -J O o 22 0 10 DISTILLED WATER IN THE COLUMN BEFORE RUN S I. WATER - O O WATER PHASE KETONE PHASE j L J L 1 2 0 30 40 50 TIME, MINUTES 60 70 FIGURE 6. Steady state study, Hun 6 1^ 80 £0 «9 continuous phase feed solution. (Consequently, Run 61^ i s not included in the results given on Figure 8.) The continuous phase feed 3 concentration, of 50.4 lb.-mdles/cu. ft.xlO , was used i n other runs which were done using the same conditions except for the Elgin head which was drained before beginning to f i l l the column. Figure 6A shows an example of the results of one such run. A l l these results are recorded in Table 11, given in Appendix I I I . To r e a l i z e the second way of beginning a run (as just mentiomed) the column was not operated for one or two days a f t e r an e a r l i e r run had been completed. This time was allowed to elapse i n order to permit the concentrations within the column to approach more uniform and, i n addition, lower values than those which applied at the end of the e a r l i e r run; extraction continued at a low rate beyond the end of the e a r l i e r run. A new extraction run then was started with the object of finding the time required to reach steady state for these new i n i t i a l conditions. The method of sampling the i n l e t and toutl'et solutions was the same as mentioned e a r l i e r . Figure 7 refers to a run started with a column where the concentration was lower than 50.4 lb,-moles/cu. f t . xlO due to a previous run. Table 12, given i n Appendix III shows the r e s u l t s of four others runs performed using the same procedure. This investigation was carried out for a constant continuous 3 2 phase flow rate of 54.8 f t . / h r . - f t . while the dispersed phase flow rate 3 2 3 2 was varied from 72.4 f t . / h r . - f t . up to 208.0 f t . / h r . - f t . . Figure 8 summarizes a l l the results obtained. (Table 13, given i n Appendix I I I , l i s t s the same results.) These runs were a l l performed to give the resu l t s described without using any internal method of sampling. However, a f t e r very long 32 ro O 28 ro Q O < CO L J CD L d O z o o h-LU _ J h-3 O 27 26 25 o < 24 23 22 - O O W A T E R PHASE -Q-ChKETONE PHASE 10 20 30 40 TIME , MINUTES. 50 i t CUPS 7. Steady state study, Run 6L* 50 40 LLI LLi 20 Q \\ \\ L = CONSTANT = 54.8 F T 3 / H R . \" F T ^ \\ / - F E E D WATER PHASE /\\\" \\ / FILLING THE COLUMN. ^COLUMN OPERATED \" PREVIOUSLY. ^ \\ \\ \\ i , i . i . i - i I J L_ 0 50 100 150 200 250 300 350 DISPERSED PHASE FLOW R A T E , FT . /.HR.-FT. FIGURE 3. Suimnarizing p l o t of the steady state r e s u l t s . 34 times when the steady state had been reached, some of the runs were continued, and int e r n a l samples were taken, as part of the study of the effe c t of such sampling on the steady state. B) Purging study: The purging time i s the time needed to achieve a uniform concentration a f t e r the position of the probes i n the column has keen changed. The purging time depends on the amount of vacuum applied to the probes, and the resultant rate of purging which i s controlled also by the in s e r t i o n of lengths of c a p i l l a r y tube inside the sampling l i n e s . The vacuum, and these c a p i l l a r i e s condition also the rate of sampling, obviously. The procedure followed was to establish a steady state operation of the extraction column, and then to place the probes at some convenient distance from the nozzle t i p s . The probes were f i l l e d at t h i s position for f i f t e e n minutes at the purging rate to be studied. The probes then were moved to a second position where purging was ca r r i e d out for measured times. The positions of the probes mentioned here were s p e c i f i c a l l y those of sampling positions 2, 2A, 3B, 4A, and 7 as given i n Table 2. (On Figures sampling position i s abbreviated to \"Position\" or\"Pos'.*) In each run, several samples were taken at various sampling rates. After the probes had been moved to a second (or to a subsequent) posit i o n , a sample was taken every minute. The concentration of the samples obtained a f t e r s u f f i c i e n t purging i n a l l the runs done for t h i s purpose (Runs lG,2,2A,2B,2C,7,7B,7C,and 7D) check within approximately 0.5% at the worftt i n the respective phase. The sampling rates used for each run in the present work appear i n Table 24 (Appendix IV) which also shows the minimum purging time for each p a r t i c u l a r value of the purging rate. A t y p i c a l example of what was obtained for a p a r t i c u l a r run i s shown i n Figure 9. A l l the result s obtained for each run concerning the minimum purge time appear i n Tables 14, 15, 16, 17, 18? 19,120? 121!, 22, and 23, located i n Appendix IV. A summarizing plot of the results of a l l these runs i s given as Figure 10 following. This Figure shows reasonably accurately the minimum purge time to obtain uniform concentration at various sampling rates. To be on the safe side for normal extracting operations, at least 2 to 3 minutes should be allowed beyond the value of the purge time obtained from Figure 10. Table 24 corresponding to Figure 10 i s given in Appendix IV. C) The influence of sampling rate on point concentrations: A l l research work done i n the past, using the sampling probe method, appeared uncertain because of the underlying assumption: that the measured water phase probe concentrations, C , are representative samples of the aqueous phase for use i n the material balance for cal c u l a t i n g An attempt was made to establish conclusively the continuous phase concentration p r o f i l e . The method used consisted in placing the probes at various sampling positions i n the columm where concentrations were reasonably far from equilibrium conditions. Incidentally, Choudhury's equilibrium curve (10), available i n the laboratory plotted to a large scale, was used throughout the work for getting equilibrium concentrations. The s p e c i f i c distances used were 1.59 and 1.66 f t . from the nozzle t i p s . At these locations experiments were made to find any apparent effect of sampling rate on point concentration. 36 4 0 - P O S . 7 35 !~ ^ O Ll_ 2 5 CO LU o CD O z o o LU _ J Z> O UJ CD O cr 2 0 Q i PURGING RATE = 13.9 CC/MIR P 0 S . 4 A PURGING RATE=II.2 CC/MIN. P 0 S . 2 A P 0 S . 7 POS 4 A \\ PURGING R A T E = 15.9 CC/MIN ~ • POS. 4 A PURGING RATE=I4.7 CC/MIN O -POS. 2 A - O O WATER P H A S E - O - O K E T O N E P H A S E 1 '0 2 4 6 8 PURGING T IME, MIN FIGURE 9. Study of the miniunm purge tiiao, Rum 2. 10 PURGING R A T E , CC./MIN. FIGURE 10. Minlciu® purge ilaje versus jkwrging rate . 38 Two d i f f e r e n t kindsof sampling procedure were followed. In one the probes were purged for 10 minutes at a certain rate. Following this period this rate was maintained for an additional period long enough to provide a reasonable volume of sample for analysis* When the rate sampling^was changed, the sampling l i n e s were purged again for 10 minutes at the new sampling rate before taking the new sample. This procedure was i d e a l * for obtaining concentration values which would d i f f e r from each other i f the concentration varies greatly with the sampling rate. Run 9H was performed following this sampling procedure for both phases. Figures 11 and 12 ( data i n Tabic 25, Appendix V) give the results of th i s run and show that great differences i n point c concentrations do not exist even i f the rates varied from 6.2 to 16.0 cc./min. for the water probe and from 8.2 to 15.6 cc./min. for the ketone probe. However, i t was discovered l a t e r that these samples were taken at a position where the concentrations were only 5% (% =(C* - C )xl00) KX ^ ,.„f£*j^ Cki away from the equilibrium concentrations. Perhaps even i n t h i s case, some changes i n concentration would be possible for high rates of sampling. Run 5 G was performed i n a manner si m i l a r to the method used for Run 9H, except that the purge time was applied as prescribed by Figure 10 i n a l l cases. Other differences were that only the ketone probe*sampled for a l l the run, and that the concentrations at the sampling position were far from equilibrium concentrations. (This statement i s also true for a l l the other runs, with the possible exception ID where concentrations were s t i l l noticeably away from equilibriumj The water probe sampled just at the beginning and the * except that i n one case a longer purge time would have been required to provide asafety factor i n addition to the requirements of Figure 10. ro • O x ro CO L J 17 CD I-3 O UJ CD O 16 8 1 5 14 SAMPLING POSITION 1.59FT WATER PHASE 1 8 10 12 14 16 SAMPLING R A T E , CC/MIN. FIGURE IIo Influence of sampling rate ©a point concentration, Kun 9H» 18 20 Ci* JO o r o * r— LK. >^ CO IE 9 -o CD* - J 8 -• O , z o o 7-h-o 6 -UJ CO o 5 -6 SAMPLING POSITION 1.59 FT - Q J L KETONE P H A S E 8 10 12 14 16 18 20 SAMPLING R A T E , C C / M I N . FIGURE 12. Influence•of sampling rat© ©n point coseeatr^t'iofs., tea ©BU end of the run, and at only one sampling rate. The water probe was not used for the intervening f i v e samples because i t was planned to study whether or not the ketone probe concentration varied when samples\" were not being removed with the water probe. Figure 13 records the results obtained i n Run 5G for the ketone probe and Table 25 in Appendix V gives the corresponding data. Another Run, 5H, was performed to study further the question of any variations i n the ketone probe concentration with sampling rate when the water probe did not operate. At the beginning of th i s Run, the sampling li n e s were purged for 15 minutes. Then a sample was taken for 3 minutes. The lines were not purged as prescribed by Figure 10 for the next two samples, but were for the remaining three samples. • This meant that four samples were obtained according to the f i r s t sampling procedure and two according to the second procedure to be described shortly. Figure 14 records the results of Run 511. (See Table 25, Appendix V for the tabulated results of Run 5H.) The method of pl o t t i n g t h i s figure i s that used for Figure 15 and described when that Figure i s presented. The second sampling procedure consisted i n placing the probe at location 3B (1.66 f t from the nozzle tips) and purging for a s u f f i c i e n t time to purge the probe. After this operation, one sample was taken at the same rate used for purging. Then, instead of purging the probe again, another sampling rate was set, and used to sample at the same location. This sampling procedure was employed i n Runs 5B^, 5F, * 7E, 7F and ID „ Each of these runs must be described here separately. In Run 5B^, a purge of nine minutes was carried out before taking the f i r s t sample of each phase. Then, without purging, another * In Run ID a probe location of 6.16 f t . was used. ro X ro (f) L d Z> O LU QQ O (XL CL 15 m 14 o I3h 12 I I SAMPLING POSITION 3B • KETONE PHASE j L 8 10 -12 14 16 SAMPLING R A T E , CC./MIN. 18 20 FIGUHE 13o Xafltteace ®t ssBBpiing rate oa p o i a t c o n c e n t r a t i o n , tea §6 X CO LU CD o z o o I— o LU CD O cr Q . 14 13 I I SAMPLING POSITION 3 B • o o a o o K E T O N E PHASE i 8 10 12 14 16 SAMPLING R A T E , CC./MIN 18 2 0 FIGURE 14. Influence of s a i l i n g rate on point concentration, Run SH, 44 sample was taken at a d i f f e r e n t rate and only with the ketone probe. Two further samples were obtained i n a s i m i l a r way. Except f o r the f i r s t sample which was done according to the f i r s t procedure, those taken a f t e r were made up of mixtures of solutions sucked out from the column at d i f f e r e n t rates. However, i f concentration variations due to variations of sampling rate are present, the r e s u l t of this procedure would be a change in concentration from sample to sample. It i s true that concentrations corresponding to d e f i n i t e sampling rates were not obtained; nevertheless, an absence of•concentration change would indicate no ef f e c t of sampling rate on measured concentrations. Figure 15 shows the r e s u l t s of Run 5B^. Each square*is plotted at the sampling rate being used when the respective samples passed into the sample fl a s k . The l i n e s drawn from each square* to the axis of abscissas reach that axis at the sampling rates which were in use when the l i q u i d s found i n the respective sample actually passed into the probe entrance from the column. The percentage marked on each of these l i n e s i s the approximate volume percentage of the sample which passed into the probe entrance at the rate indicated by the l i n e . The triangles represent the concentration corresponding to an average rate of passage into the probe entrance. This average rate was obtained by weighting the rate indicated by each of the l i n e s according to the corresponding volume percentage. (These r e s u l t s also are recorded i n Table 25, Appendix V.) Run 5F was performed exactly as Run 5B^. Figure 16 (and Table 25) shows these r e s u l t s . The second sampling procedure also was used to perform Run 7E. No purging took place before sampling was begun with the result that the f i r s t two samples in Table 25 could not be used. Ten minutes samples of each phase were taken simultaneously at various rates ranging up to * For Figures 17, and 18 c i r c l e s are used instead of squares. SAMPLING R A T E , CC./MIN. FIGURE 15. Influence of sampling rate on point concentration, Run 5B SAMPLING RATE, CC/MIN 47 14.2 cc./min. for the water probe and from 5.4 to 9.8 cc./min. for the ketone probe. Each sample represents the results corresponding to more than one rate of sampling into the probes. Figure 17 shows the results for the continuous phase probe; the same method of p l o t t i n g was used as Figure 15. Results for both probes appear i n Table 25, Appendix V. (The ketone phase results have not been plotted because only a narrow range of ketone sampling rates was investigates.) No s i g n i f i c a n t variations in point concentration were found. Run 7F was performed as Run 7E, The f i r s t two samples i n Table 25 could not be used because of f a i l u r e to purge. Several samples of each phase were taken at various rates ranging up to 28.2 cc./min. for the ketone phase probe. Simultaneous samples were taken with the water probe at a constant rate of 2.8 cc./min.. The results are shown on Figure 17A and recorded in Appendix V, Table 25. (The water phase results have not been plotted because only a single rate of sampling was investigated.) Another Run, ID, was done also by using the second procedure (except, of course, that the f i r s t sample again was taken, in e f f e c t , by the f i r s t procedure). In Run ID, the probes were placed this time 6.16 f t . from the nozzle t i p s . The sampling li n e s were purged 20 minutes before taking the f i r s t sample of each of the phases. Samples of each phase were taken at the same time at various rates for 5 minutes, again without purging between samples. However, as mentioned e a r l i e r , the concentrations measured under these conditions should vary i f the sampling rate influenced the point concentration. No s i g n i f i c a n t changes i n the concentrations were found even though the water sampling rate was varied from 11 to 34 cc./min. and the ketone sampling rate from 12.9 to 28.4 c c o / m i n . o Figures 18 and 19 record the resul t s for each SAMPLING R A T E , CC./MIN. PI602SS 17. Influence ©£ sampling rate on point concentration, Sun 73. ro -O X ro UJ - J O O Z o o O LU QQ O cc CL h SAMPLING POSITION 4A 19 18 17 16 15 0 8 12 K E T O N E P H A S E S A M P L I N G R A T E , CC. /MIN. FIC*U£?B 17Ao Influence o f eanpling rate- on point concentration, Saa».7F. 4 9 SAMPLING POSITION 7 WATER PHASE 16 2 0 2 4 2 8 3 2 SAMPLING R A T E , C C . / M ! FIGURE 18 o Influence of sampling rate on point concentration. Run ID. 3 6 4 0 8 ro -g x ro CO w 25 CD - J b z o o o LU QQ O cr 24 23 22 2 I 12 SAMPLING POSITION 7 KETONE PHASE 14 16 18 2 0 22 2 4 SAMPLING R A T E , CC./MIN FIGURE $9i. laflueaee « f sampling rate on.point eoQeeatratioxt, Him IB, 52 as does Table 25 i n Appendix V. D) The influence of sampling procedure on the steady state concentrations. During Runs 7£ and 7F of the investigation of sampling rate v a r i a t i o n for both phases described previously, the column i n l e t and outlet concentrations were measured. The outlet samples were taken to investigate the influence of the rate of sampling on the steady state concentrations. At f i r s t , the continuous phase sampling rate was kept constant while the dispersed sampling rate was Varied over a wide range of rates. Figure 20 records the concentrations obtained in r e l a t i o n to the sampling rate in,Run 7F. The next operation consisted in varying the continuous phase sampling rate and holding the other one constant. These r e s u l t s are shown on Figure 2.1 for Run 7E. The procedure just mentioned was done to study i f the concentration p r o f i l e of either phase would change i f either rate i s varied with the other one held constant. The concentration p r o f i l e s were not measured i n d e t a i l . However, change i n the outlet concentration would indicate a change i n the p r o f i l e . On Figures 20 and 21, the column's length (nozzle ti p s to interface) has been included, as well as the sampling rate, shown as a percentage of the respective phase flow. A maximum deviation i n either run i n the outlet concentrations was 2.3% from the mean value. The i n l e t concentrations were constant because they were fixed before beginning a run. (A maximum rate of 14.2cc./min. was used on the continuous phase probe while the maximum used on the dispersed phase probe was 28.2 cc./min..) (Note also that the probes were in operation only for the times shown. There was no additional purging.) S3 INTERFACE LEVEL OF 7 - 5.5\" A-SAMPLING TIME , POS. 4 A. W=,K=. SAMPLING RATE AS % OF PHASE FLOW. MAXIMUM DEVIATION OF CONCENTRATION FROM AVERAGE : W = 2 . 3 % , K=l.2%. W=0.9% K =5.0% 8 25.5! FIGTOJE 20. Inflwouc© of 40 50 6 0 saapllgg pate on steady state, tea, ?p. 70 80 9 0 100 TIME FROM START , MIN. 54 o ori o o o H _ l I — o o o A 4.2\" 2 6 . 2 INTERFACE LEVEL OF 7'-4.7\". V) 2 6 . 0 UJ 2 5 . 8 M 2 5 . 6 2 5 . 4 2 5 . 2 2 5 . 0 2 0 STEADY STATE i 1 SAMPLING TIME.POS. 4 A. W= , K= SAMPLING RATE AS % OF PHASE FLOW. 1% = 3.1 CC./MIN. OF WATER = 3.5 CC./MIN. OF KETONE MAXIMUM DEVIATION OF CONCENTRATION' W = 1.8 % W=I.O % W=2.0% K = 2 . 0 % K = l . 8 % W=3.6°/« K = l . 6 % i WATER PHASE W=2.6 % K=l.5 % 1 1 W=4.4% k = 2 . 8 % J L 3 0 4 0 5 0 6 0 7 0 8 0 9 0 T I M E , MINUTES. (FROM START) 100 110 120 130 140 150 FIGURE 21. Influence of sampling rate on steady state, Run 7E. E) Concentration p r o f i l e s : As mentioned i n the l i t e r a t u r e survey, Choudhury's work did not include the steady state requirements, or an adequate study of the minimum purge time to change the solution inside the probe a f t e r a relocation. Furthermore Choudhury assumed that the continuous phase probe sample was representative. The present work, was >•..•.' performed to study these effects and consequently i t became obvious that the duplication of at least one of Choudhury's runs was a necessity. The f i r s t attempts to duplicate the r e s u l t s obtained by Choudhury were made af t e r knowing the steady state requirements and also the time needed to change the concentration i n the probe when relocation took place. Run 65 of Chouchury's thesis (10) was repeated five times. A comparison of the results obtained with those of Choudhury are shown for Run 5D only, i n Figure 2 2 . (Table 10A gives the data of this run.) There was a s i g n i f i c a n t difference in r e s u l t s . It i s believed that the difference was caused mainly by an error i n a n a l y t i c a l procedure made by the present investigator. At the beginning of the present work, two 10 ml. pipettes were calibrated and used to measure the samples to be analyzed. The c a l i b r a t i o n of both pipettes showed that the difference in volume from each other and from 10 ml. was n e g l i g i b l e . After three months of experiments, a calibrated pipette was broken and replaced by a new one. However, due to the previous experience, the new pipette was not cal i b r a t e d . However, aft e r long use this pipette was calibrated, and found to de l i v e r 5% more than those used i n i t i a l l y . A second cause of error consisted of an evaporation of some dispersed phase from the feed ss L = 51.1 F T 3 / H R - F T w HEIGHT, FT. FIGURE 22.: Duplication oS Caoudnury^s Run 65, Run 57 tank through the vent system,, This evaporation caused a gradual increase i n the dispersed phase feed concentration. These errors were re a d i l y corrected, and Run 8 was performed to recheck Run 5D and Choudhury's Run 65. Figure 23 shows the results of both Run 8 and of Choudhury's Run 65. These resul t s are given also in Appendix I I , Table 10A. F) Concentration study i n the Elgin head: It should be interesting to mention here an attempt which was made to study solute concentrations i n the Elgin head. A glass tube was introduced into a vent l i n e i n s t a l l e d on the Elgin head, close to the center of the column. From th i s location, several samples of water phase and of ketone phase were taken at d i f f e r e n t distances from the column interface. Table 3 and Figure 24 give the results obtained. Since these concentrations were not related to the point concentrations inside the column, th i s study was not carried any further. TABLE 3 Concentration study i n the Elgin head. Height of Distance Concentration, ^ Located on column. from nozzle. Ib.-moles/ft 3xl0 . Figure 24 f t . f t . as: 7'- 3.8\" 7'- 0.25\" 34.45 A 7'- 3.8\" 7'- 3.30\" 34.42 B 7'- 3.8\" 7«- 3.55\" 34.20 C 7'- 3.8\" 7'- 4.05\" 15.85 D 7'- 3.8\" 7«- 4.30\" 15.00 E 7'- 3.8\" 7'- 6.80\" 14.86 F S3 L w = 5 4 . 8 FT /HR.-FT. L k = 7 2 . 7 F T 3 / H R . - F T 2 INTERFACE KETONE PHASE PROFILE - - - - CHOUDHURY J I I l l _ FIGURE 23. I 2 HEIGHT, FT. Duplication of Choudhury's Run 65, RUES 8< FIGURE 24. Concentration study in the Elgin head 60 G) Piston sampling. L - Influence on steady state: Three runs were carried out to determine the length of time required to achieve steady state concentrations under conditions of 3 2 normal operation using a dispersed phase flow rate of 120 o2 f t / h r . - f t 3 2 and a continuous phase flow rate of 54.2 f t / h r . - f t . In addition, the influence on steady state was investigated of the removal of a portion of the contents of the column and simultaneous insertion of d i s t i l l e d water, or, on other occasions, of outlet water phase, by means of the piston. For the f i r s t run, the Elgin head was drained completely, the column was f i l l e d with continuous phase feed solution, and the extraction process started using the flow conditions just mentioned. Both i n l e t solutions were sampled three times each during the run, while both outlet solutions were sampled every f i v e minutes from the s t a r t . At fifty-two minutes from the beginning of the extracting operation, the piston was operated with d i s t i l l e d water in the hole which was to s l i d e into l i n e with the column proper. D i s t i l l e d water was used to show the effect of replacing a column section by a water solution which w i l l give the maximum disturbance of the steady state. Just a f t e r taking a piston sample, samples of the outlet solutions were taken every two minutes for 10 minutes. Another sample of each outlet solution was taken eighteen minutes aft e r the piston sample had been obtained. Another piston sample was taken 20 minutes a f t e r , and the same procedure for sampling the outlet solutions was repeated. From this run, i t was found that the time to obtain steady state a f t e r the s t a r t of the run was comparable to the previous r e s u l t s . Taking a piston sample using d i s t i l l e d water in the piston hole which was moved into l i n e with the 61 column proper resulted i n 15 minutes being required for reestablishing the steady state for the water phase. The dispersed phase outlet concentration was not changed by the piston sample being taken using d i s t i l l e d water to replace the column's section. These results are given in Figure 25. • Run 9D was performed s t a r t i n g with a column which had been operated previously. The flow rates were the same as for Run 9C. A si m i l a r procedure was applied to obtain samples of i n l e t and outlet solutions. At thirty-two minutes from the beginning of the run, a piston sample was taken, re s u l t i n g i n the replacement of apportion of the column by a portion of d i s t i l l e d water. As expected, this run gave a shorter result, for the length of time needed for the column to achieve steady state when samples were not being taken. In t h i s Run some disturbance of the outlet ketone phase concentration apparently took place; however, only one sample showed a deviation from the normal steady value, and the result may not be s i g n i f i c a n t . The disturbance of the water phase exit concentration was much more pronounced ( as expected from the previous r e s u l t s ) . Steady state was restored i n a l i t t l e \" over 10 minutes. Figure 26 records the result s of Run 9D. Run 9F was performed using exactly the same conditions of flow as those of 9D and 9C. In Run 9F, the probes were used to sample at the piston axis with a purging and sampling rate of 10.0 cc./min. for the continuous phase and 9.4 cc./min. for the dispersed phase. The purging time was 10 minutes. Referring to Figure 10 which sp e c i f i e d the minimum purging times for several purging rates, i t can be seen that 6.2 minutes were the minimum needed to purge the continuous phase probe at a rate of 10.0 cc./min., and 5.2 minutes were the minimum COLUMN OUT. CONC, LB. MOLES/FT xlO: 89''-. X TIME , MINUTES. FIGURE 26. Eff e c t of piston sampling on steady state, 3 Run 9D.; M 64 needed to purge the dispersed phase probe at a rate of 9.4 cc./min.. It was decided to use 10 minutes to be on the safe side as mentioned e a r l i e r . A piston sample was taken half a minute aft e r the probe sample. The piston hole which had to s l i d e into l i n e with the column proper was f i l l e d t his time with the outlet continuous phase solution. (This was the solution to be used i n the dispersed phase sampling runs.) The sampling procedure was repeated three, times. Figure 27 shows the results which indicate that the steady state i s not disturbed by a piston sample when the piston hole s l i d i n g into l i n e with the column i s f i l l e d with outlet continuous phase solution. The carrying out of probe sampling did not affect the steady state, as would have been expected from the results given under the heading of sampling rate influence on steady state. 2.- Piston point concentrations: Five runs were performed to obtain information about the sampling procedure by means of a piston sampler. For the f i r s t Run, 9K, flow rates of 120.0 ft°/hr.-ft 2 for the 3 2 dispersed phase and 54.8 f t / n r . - f t for the continuous phase were used. Samples were taken at the axis of the piston by means of the probes; ha l f a minute l a t e r a piston sample was taken. Three of these samples were available at the end of th i s Run. The purging and sampling rates used with the water probe was 12.0 cc./min. and with the ketone probe 13.0 cc./min. The purging time used was 10 minutes as compared with 5.2 minutes and 3.8 minutes required to purge adequately the water and ketone probes respectively according to Figure 10 for the rates just mentioned. The average volume percent ketone i n the ketone probe sample was 89.6% while that for the piston sample was 14.5%. 25 23 21 19 17 I 5 tro-KETONE PHASE J i I u _ — L PURGING TIME. —. SAMPLING TIME. PISTON SAMPLE TAKEN T 1 I 1 1 WATER PHASE L 10 20 30 40 50 60 70 80 90 TIME , MINUTES. FIGURE 27. Effect of piston sampling on steady state„ Run 9F 66 A second Run, 9F, was performed to check Run 9E. The 3 2 dispersed phase flow rate was 119.8 f t / h r . - f t and the continuous phase 3 2 flow rate, 54.8 f t / h r c - f t 0 As before, the probes were used to sample at the axis of the piston; but, after the second sample had been taken, i t appeared desirable to check whether the probes were r e a l l y at the piston axis when sampling. It was found that they were one inch too low. This meant that the probe samples of run 9E and those of 9F obtained to this point were not suitable for comparison with the piston samples. The error was corrected and Run 9F was continued. The purging and sampling rates used were 10.0 cc./min. for the continuous phase probe and 9.1 cc./min. for the dispersed phase probe. The purging time was 10 minutes whereas the minimum purging time was 5.2 minutes for the ketone probe and 6.2 minutes for the water probe according to Figure 10 for the rates mentioned above. Out of four piston samples taken in Run 9F, the resul t s of the l a s t two were almost exactly the same as those taken with the dispersed phase probe. Run 9E and the part of Run 9F i n which the probes were i n the wrong position were not discarded. This course was taken because the piston results obtained were comparable with one another and with those of Run 9G which was performed using the same conditions. Table 4 records the r e s u l t s , from which i t can be seen that the dispersed phase concentrations of the piston samples of Runs 9E and 9F at the time of analysis, C j ^, are equal to close approximation for a l l the samples. The next Run to be described, Run 9G, was a r e p e t i t i o n of 9F. Run 9G was done to v e r i f y the results of the previous runs but with the e a r l i e r wrong probe locations avoided. Four piston res u l t s were obtained i n this Run; they reproduce previous dispersed phase concentrations at time of analysis very c l o s e l y as shown i n Table 4. 67 A l l the runs performed to obtain piston results to t h i s point were obtained for a drop v e l o c i t y through the nozzle t i p s of 0,338 f t . / s e c . Reference to Table 4 shows that a maximum volume percent ketone i n the column of 14 08 was found using this dispersed phase linear v e l o c i t y i n Runs 9E, 9F, and 9G. It was suggested that t h i s ketone holdup be increased for the purpose of getting more ketone phase i n a piston sample 0 Other things being equal, the analysis ( i f done at equilibrium) would be less c r i t i c a l , and, consequently, the piston dispersed phase concentrations would be more accurate. To increase the holdup, i t became necessary to increase the v e l o c i t y i n the nozzle tip s well above the values used so f a r . Run 9H was done using the maximum capacity of the ketone rotameter available at the beginning of the work; 21 noz/.le t i p s were used, and a v e l o c i t y of 0.475 f t . / s e c . i n them. For Run 9H, the 3 2 dispersed phase flow rate was 169.5 f t . / h r . - f t . and the continuous phase flow rate, 54,8 f ? . / h r . - f t ? . In t h i s Run, the probes were used again to sample at the piston axis. The piston sample was taken h a l f a minute a f t e r the probe sample. The purging and sampling rate used with the dispersed phase probe was 8.7 cc./min. and with the continuous phase probe 6.2 cc./min.. The purging time was 10 minutes. Reference to Figure 10 shows that the rates just mentioned require a minimum purging time of 5.4 minutes for the ketone probe and 10 minutes for the water probe. Thus the 10 minutes allowed was ample for the ketone probe but included no safety factor i n the case of the water probe. The safety factor was not added i n t h i s case because the solutions i n the feed tanks were running out very quickly. The volume percent ketone i n a piston sample now was 22.2%. (The volume percent ketone for the corresponding dispersed phase probe samples averaged 92aG%.) The dispersed phase piston concentrations, found at the time of analysis €. _, were lower than previously because of higher dispersed phase flow xv X rate used. These results also are recorded i n Table 4. For the f i r s t time some ketone was present with one continuous probe phase,sample out of four as a result of the high ketone holdup. A correction to the water concentration was made by writing Equation 5 for each probe. The value of for the ketone present i n the water was taken to be the equilibrium value. The two equations r e s u l t i n g were then solved simultaneously to find the value of C .. (This J wi cal c u l a t i o n i s given i n Appendix VIL) A new rotameter was i n s t a l l e d and calibrated for the purpose of being able to further increase the dispersed phase flow rate and therefore the column (and piston sample) holdup. The c a l i b r a t i o n of t h i s rotameter i s given i n Figure 28, following. Run 91 was performed using a dispersed phase flow rate of 3 2 3 2 208.0 f t . / h r . - f t . and a continuous phase flow rate of 54.8 f t . / h r . - f t . . A v e l o c i t y of 0.584 f t . / s e c . was used through the 21 t i p s of the nozzle. The probes were used to sample at the piston axis using a dispersed phase probe rate of 7.3 cc./min. and a continuous phase probe rate of 6.5 cc./min.. Based on Figure 10 minimum purge times of 6.4 minutes and 9.6 minutes were required for the ketone and for the water prober, respectively; however, 10 minutes were allowed as purging time. For the second time, some ketone was present with two continuous phase samples out of a t o t a l of five as a re s u l t of the high Ketone holdup i n the column, i n spite of low purging and sampling rates. As done for Run 9H, corrected values were calculated using Equation 5 for each probe. This c a l c u l a t i o n was done for Samples 3 and 5, but only in 69 of 6.0 x 5.0 O K E T O N E P H A S E R O T A M E T E R 72 ° F 1 i 0.0 0.1 0.2 0.3 0.4 R O T A M E T E R READING 0.5 FIGURE 28. C a l i b r a t i o n of the rotameter, s e r i a l Number 20,789 f o r a ketone phase having a c o n c e n t r a t i o n o f 6.43 l b . - n o l e s / f t 70 Sample 3 was athene any change i n the figures? (See Appendix VIL) Even —3 3 in Sample 3 the correction was very small, +0.03 xlO lb.-moles/ft. for C ., and n e g l i g i b l e for C. .. The ketone was only 1.2% of the t o t a l wi k i the volume of water in Sample 3, and was 0.9% of Sample 5. The highest holdups as measured by the piston sampler were obtained in t h i s Run 91, the average holdup for f i v e samples being 28.4%. The volume percent ketone i n the dispersed phase probe samples for t h i s Run averaged 96.5%. The dispersed phase piston concentrations at time of analysis, again were found to be lower than before due to the increase i n the dispersed phase flow rate. It was found l a t e r that the phases were close to equilibrium. This i s probably why the i n i t i a l dispersed phase concentration (determined by using the material balance as applied to the piston samples) checked the probe dispersed phase resul t s within less than 2.0% deviation as defined i n connection with Table 4„ ( For the case of one sample out of five i n t h i s Run, the % deviation was higher. The r e s u l t s of t h i s Run appear in Table 4 also.) Table 5 shows in d e t a i l the probe data corresponding to each piston sample. In a few cases there was too l i t t l e aqueous phase in the dispersed phase sample for i t to be possible to obtain C by analysis. The value of C „ was then assumed to be the equilibrium value wf corresponding to C^f' a n < * ^n:*-s equilibrium value was used as C i n the material balance c a l c u l a t i o n of C. .. In these cases, the correction k i applied to C ^ to produce was always small, never exceeding 1.5% of Cki° Subtracting from the C . values of Table 5 the C , values of ° wi wf Table 4, produces differences (C . - C 3 which do not increase with ' wi wf increased holdup as was expected; i n fact t h i s difference did not change greatly except for Run 91 where i t was comparatively low. * Change or correction from the values obtained i f the ketone i n the water probe sample was neglected .altogether. 71 Increasing the holdup was supposed to help to make the difference between C , and C _ bigger, but unfortunately the holdup lowers the wi wf 0 0 mass transfer resistance i n the column by increasing the i n t e r f a c i a l area between continuous phase and drops. The result i s that equilibrium between the phases i s approached at the location of the sampler as the holdup increases. Table 5A records the probe i n i t i a l r e s u l t s , the piston results at time of analysis, and t h e i r corresponding equilibrium values. It also gives the mass transfer driving forces, at the location of the piston calculated from the probe r e s u l t s . Although the holdup was doubled between Run 9G and Run 91, these driving forces were cut down be a factor of 7 a7 for the water phase, and 5.6 for the ketone phase. (These factors should be very close to each other for constant slope of the equilibrium curve.) As shown by Table 5A, t h i s average driving force for the water phase i n Run 91 was 0.44, whereas for Runs 9E, 9F, and 9G i t was 3.4. The average driving force for the ketone phase in Run 91 was 0.32, whereas for Runs 9E, 9F, and 9G i t was 1.8. It should be noted that the values of the ketone phase concentration at time of analysis*, for the piston ((C, _) . . , Table 5A) a l l seem to be J ' ^ kf piston approximately 2% beyond the equilibrium values (C£ ) calculated from (C „ ) . , by means of the equilibrium curve. A s i m i l a r comment can wf piston J 1 be made, of course, with reference to the values of (C .) . , . These ' ' wf piston are a l l less than the equilibrium values (C* ) calculated from M wf (0, _) , . again by means of the equilibrium curve, kf piston The probe resul t s of Table 5A show that both phases within the column were almost at equilibrium at 1.59 f t from the nozzle t i p s * For the piston samples, the analyses a l l were done a f t e r the phases had reached equilibrium. 72 for Run 91. This result was caused by the extraction taking place mainly i n the upper part of the column. The large amount of ketone present extracted much more acid from the water entering the column than that removed i n the runs at lower holdup. The concentration of acetic acid i n the water phase when i t reached the piston then was small, and nearly in equilibrium with the concentration i n the ketone phase. Further d e t a i l s w i l l be given i n the discussion on th i s subject, l a t e r i n the thesis. 73 TABLE 4 Piston r e s u l t s . Run Volume Volume Concentrations Average % deviatii number. of percent lb .-raoles/ftV xlO piston of pistol sample ketone and probe results cc. results from average. V. V c. „ c „ c. . c, s k w kf wf k i k i piston r e s u l t s probe *9E 17.0 100.0 14.5 10.34 20.50 6.22 8.34 7.28 -14.3 *9E 17.0 101.0 14.4 9.87 19.50 9.51 7.98 8v75 + 8,7 *9E 17.5 100.0 14.8 9.90 19.68 10.59 7.98 9.29 +14.0 *9F 17.0 100.0 14.5 10.54 20.96 9.48 8.61 9.04 + 4.8 *9F 17 cO 100.0 14.5 10.45 20.79 10.45 8.44 9.45 +10.6 9F 17.0 100.0 14.5 10.34 20.55 8.28 8.43 8.34 - 0.7 9F 16.0 101.0 13.7 10.22 20.55 7.38 8.08 7.73 - 4.5 9G 16.0 101.0 13.7 10.22 20.30 6.45 8.36 7.41 -13.0 9G 16.0 101.0 13.7 10II5 20.35 6.30 8.30 7.30 -13.7 9G 17.0 100.0 14.5 10.15 20.15 7.44 8.33 7.89 - 5.7 9G 17.0 100.0 14.5 10.15 20.30 8.33 8.28 8.31 + 0.2 a9H 26.0 91.0 22.2 8.00 16.00 8.21 7.20 7.71 + 6.5 J9H 26.0 91.0 22.2 7.77 15.80 7.25 7.33 7.29 - 0.5 9H 26.0 91.0 22.2 7.65 15.39 6.64 7.05 6.85 - 3.7 9H 26.0 91.0 22.2 7.65 15.53 6.00 7.10 6.55 - 8.4 91 33.5 83.5 28.6 7.59 15.24 7.02 7.15 7.09 - 1.0 91 33.5 83.5 28.6 7.09 14.25 5.94 6.14 6.34 - 6.9 :9i 33.0 84.0 28.2 7„09 14.25 6.58 6.76 6.67 - 1.4 91 33.0 84.0 28.2 7.07 14.39 6.84 6.76 6.80 + 0.6 91 33.0 84.0 28.2 7.07 14.39 6.84 6.82 6.83 + 0.2 * Runs performed with probes l . - i n . below the axis of the piston a Runs which had a negative (C .- C „) difference ° wi wf I Correction was applied for the ketone i n the water probe sample. 74 TABLE 5 Probe r e s u l t s corresponding to piston results of Table 4. Run Volume of Volume Concentrations number. sample cc. percent lb.-moles/ft? xlO (ketone probe) ketone. \\ V w C k f Cwf C . wi ketone probe water *9E 61.5 S7.5 89.2 8.76 17.75 8.34 21.20 *9E 38.0 4.0 90.5 8.29 16.64 7.98 19.56 *9E 31.0 3.0 91.2 8.29 16.35 7.98 19.56 *9K 23.7 1.0 96.0 8.73 #18.20 8.61 21.14 *9F 28.0 2.0 93.4 8.64 #18.00 8.44 20.79 9F 81.0 5.0 94.1 8.64 17.52 8.43 20.90 9F 42.0 5.0 89.3 8.70 15.75 8.08 21.00 9G m.o 4.0 90.9 8.70 17.51 8.36 20.90 9G 48.5 4.0 92.3 8.58 17.51 8.30 20.96 9G 51.5 4.0 92.8 8.58 17.37 8.33 20.61 9G 53.5 5.0 91.5 8.58 17.37 8.28 20.61 9H 39.0 3.0 95.2 7.30 14.60 7.20 15.94 J9H 15.6 0.4 97.5 7.30 14.95 7.33 15.95 9H 38 „0 3.0 92.7 7.15 14.31 7.05 15.68 9H 19.0 1.5 92,7 7.18 #15.00 7.10 16.00 91 41.5 2.5 94.3 7.18 #15.00 7.15 15.47 91 36.0 2.0 94.8 6.77 #14.15 6.74 14.71 :9i 28 oO 1.0 96.5 6.77 #14.15 6.76 14.45 91 28.0 1.0 96.5 6.77 #14.15 6.76 14.48 91 40.0 2.0 95.2 6.83 #14.20 6.82 14.48 * Runs performed with probes l . - i n . below the axis of the piston J Correction was applied for the ketone i n the water probe sample. # Cwf results are equ i l i b r i u m values corresponding to C R f . The volume was too small to be analyzed. TABLE 5A Piston and probe results and t h e i r corresponding equilibrium values. Run Piston Results Probe Results Driving number. C k f wf Cwf C* V C k i C* . Wl C . force probe at l b . -moles/ft\"? x 10 J l b . -moles/ft°x 10° locat C*.-C . Wl Wl ion. C* -C k i * 9£ 10.34 21.30 20.50 9.90 8.34 17.38 21.20 10.25 3.8 1.9 9E 9.87 20 c 50 19.50 . 9 „5Q 7.98 16.65 19.56 9.45 2.9 1.5 9E 9.90 20.60 19.68 9.55 7.98 16.65 19.56 9.45 2.9 1.5 9F 10.54 21.70 20.96 10.30 8.61 17.91 21.14 10.25 3.2 1.6 9F 10.45 21.60 20.79 10.00 8.44 17.60 20.79 10.15 3.2 1.7 9F 10.34 .21.30 20.55 10.00 8.43 17.60 20.90 10.20 3.3 1.8 9F 10.22 20.90 20.55 10.05 8.08 16.85 21.00 10.20 4.2 2.1 9G 10.22 20.90 20.30 9.85 8.36 17.40 20.70 10.20 3.5 1.8 9G 10.15 20.70 20.35 9.90 8.30 17.30 20.96 10.20 3.7 1.9 9G 10.15 20.70 20.15 9.80 8.33 17.33 20.61 10.00 3.3 1.8 9G 10.15 20.70 20.30 9.85 8.28 17.28 20.61 10.00 3.3 1.7 9H 8.00 16.50 16.00 7.70 7.20 15.00 15.94 7.70 0.9 0.5 9H 7.77 16 J.5- 15.80 7,53 7.28 15ol5 15.95 7.G5 0.7 0.4 9H 7.65 15.95 15.39 7.38 7.05 14.70 15.68 7.55 1.0 0.5 9H 7 a65 15.95 15.53 7.40 7.10 14.80 16.00 7.75 0.6 ... ,0.3 91 7.59 15.80 15.24 7.30 7.15 14.90 15.47 7.40 0.6 0.3 91 7.09 14.75 14.25 6.80 6.74 14.08 14.71 7.10 0.6 0.4 91 7.09 14.75 14.25 6.80 6.76 14.09 14.42 6.90 0.3 0.1 91 7.06 14.70 14.39 6.90 6.76 14.08 14.48 7.20 0.4 0.4 91 7.06 14.70 14.39 6.90 6.82 14.20 14<:»8 7.20 0.3 0.4 # The starred values are equilibrium values corresponding to the values of the preceeding column of the Table. 76 DISCUSSION A) Study of time needed to obtain steady state; The previous work (10,11,12,13) on a l i q u i d - l i q u i d extraction spray column, did not include an adequate study of the length of time required to reach a steady extracting operation. It i s evident that a steady state must be achieved before r e l i a b l e data can be obtained. In the present work, an investigation was performed to determine accurately t h i s length of time. As would be expected, there i s a r e l a t i o n between the state of the column at the beginning of a run, the dispersed phase flow rate, and the minimum length of time necessary for steady state concentrations to be reached. (The continuous phase flow rate was not varied in t h i s study but would also have an important bearing, of course^ As would have been supposed for the flow rates and the d i r e c t i o n of mass transfer studied here, the higher the concentration inside the column at the beginning of a run, the longer i s the time needed to achieve steady state. The time required, of course, depends on the dispersed phase flow rate when the continuous phase flow rate i s kept constant as summarized in Figure 8 given e a r l i e r . There are two factors which influenced t h i s time to reach steady state. The f i r s t one i s the i n i t i a l concentration prev a i l i n g inside the column mentioned above. A reduction of approximately 20 minutes in the time to reach steady state at a dispersed phase flow rate 3 ° of 72.5 f t . / h r . - f t T i s achieved when.the column has been operated before so that low concentrations exist to begin with in the continuous phase. This time reduction i s reduced l i n e a r l y as the dispersed phase flow rate i s increased and f i n a l l y reaches a reduction of two to three minutes as flooding i s approached. These resul t s a l l are for a constant continuous phase flow rate of 54.8 f t . / h r . - f t . . The second factor was the precision of l e v e l control at the interface achieved by the interface c o n t r o l l e r . The length of time to reach steady state i s very long when the column as started contains only continuous phase feed solution, perhaps partly because the interface takes between t h i r t y to forty minutes to s t a b i l i z e at a reasonable l e v e l . When the column has been used before, the interface i s already at this reasonable l e v e l . This second factor was of operational nature and could be solved by the i n s t a l l a t i o n of an automatic l e v e l c o n t r o l l e r . B) Effect of purging rate on the minimum purge time to achieve a uniform concentration i n the probe samples: As expected, the minimum purging time was reduced when the rate of purging was increased as shown i n Figure 10 which gave the res u l t s obtained from the investigation of the minimum length of time needed to have a uniform concentration i n each of the probe samples afte r relocation of the probes. On the Figure 10 just mentioned, Choudhury's r e s u l t s were included. These are shown by means of dashed l i n e s . The present study covered a wider range of rates: from 2 to 13 cc./min. for the water phase and from 2 to 16 cc./min. for the ketone phase. A purging rate of 2 cc./min was considered to be the minimum at which sampling could proceed because about 22 minutes purge time i s needed even before c o l l e c t i o n of a sample can begin. The upper ends of the ranges were, respectively, a rate of 13 cc./min. for the water phase, and 16 cc./min. for the ketone phase. These rates were considered to be f a i r l y high because at th i s time i n the research, i t was not known i f high rates perhaps could influence the point concentration obtained and also disturb 78 the steady state. After the studies of the rate of sampling versus the point concentration, and of steady state disturbance, i t was proved that the rates, just mentioned, were s u f f i c i e n t l y high for the purpose of determining concentration p r o f i l e s . The present resu l t s are s i m i l a r to those of Choudhury for the continuous phase, but more complete. The present dispersed phase resul t s in Figure 10 are low compared to those of Choudhury. Choudhury's high values for the dispersed phase minimum purge time were probably due to d i f f e r e n t tube lengths i n the sampling l i n e s from those used here. The curves of Figure 10 were used i n the present work. For safety, two or three minutes o r d i n a r i l y were added to the indicated minimum purge time as mentioned e a r l i e r . Sometimes by design when feed solution supplies were low v or 'sometimes inadvertently, t h i s extra amount of time was not always included. However, the purge time was never less than the minimum recommended i n Figure 10 except for specijil cases where the sampling rate was varied to see i t s influence on the point concentration. C) Influence of sampling rate on steady state: The data of Figures 20 and 21, showed that sampling rates from * 3 ° 0.3 to 8.75% of the dispersed phase flow of 72.8 f t . / h r . - f t T and from /* * 3 / a 0.9 to 4.5% of the continuous phase flow of 54.8 f t . / h r . - f t . , caused no r e a l v a r i a t i o n of outlet concentration; and therefore no r e a l effect on the steady state. For most samples the sampling time used was 10 minutes, and, following t h i s , 20 minutes usually was allowed to elapse before beginning to take the next sample. During the ten minutes sampling period the volume of ketone phase i n the Elgin head (2.5.-in. of ketone * 0.3% = 1.1 cc./min. and 8.75% = 28.2 cc./min. ** 0.9% p 2.9 cc./min. and 4.50% =14.2 cc./min. 80 i n a 6.-in. I.D. cylinder) would be replaced four times, and any changes in outlet concentration (caused by sampling the ketone phase at a given rate) therefore should have shown up, i f not in the 10 minutes sampling period,then near the beginning of the time between samples. For the water phase a time of 6.7 minutes i s needed to sweep out by piston flow the volume of continuous phase i n the column from the l e v e l of the probe down to the column bottom plus that i n 10 f t . of 3/8.-in. tubing and i n the interface c o n t r o l l e r . The amount in the column was calculated as the volume of the whole empty column below the probe including that of the expanded portion, minus the volume of the nozzle for dispersing ketone, and minus the volumes occupied by the drops. For the flow rates used i n Runs 7E and 7F, Choudhury (10) has shown that the value of the driving force factor for back-mixing, F^ (10), i s about 0.9, so that, under the conditions of these Runs, back-mixing i s not very serious. Therefore during the time of 10 minutes used for sampling, or near the beginning of the period following, before the next sample was taken, any change i n concentration caused by sampling would be expected to show up i n the analysis of the water leaving the column. In Runs 7E and 7F the concentration of a water phase i n equilibrium with the leaving ketone phase was about 5% different', i n concentration from the actual entering water phase. Had the phases at the top of the column been i n equilibrium l i t t l e change might have been expected i n the outlet ketone concentration as a result of removing samples of the phases within the column. The small disturbances of the steady state found in the study seem to be caused by the interface c o n t r o l l e r . The concentrations vary 3 to 4% due to an interface change of l . - i n . This statement i s made on the basis of the steady state study, shown in Tables 11 and 12, where a i the steady state i s not reached i f the interface varies byithat much. Then smaller interface changes can produce the concentration variations found as part of the work described in the present section. As mentioned e a r l i e r , an automatic device should be i n s t a l l e d to regulate the interface l e v e l and hold i t at a fixed distance from the nozzle t i p s . It i s interesting to remark here that the steady state concentrations can not be disturbed due to the amount of l i q u i d taken out by both probes i n Runs 7E and 7F i f purging and sampling rate of 14.2 cc./min. i s not exceeded for the water phase and 28.2 cc./min. for the dispersed phase. It i s to be understood that these sampling rates are conditioned by the phase flow rates, and, as just mentioned only one was investigated for each phase. The sampling rate of 14.2 and 28.2 cc./min. are safe for these flows. Higher sampling rates than 28 cc./min for the dispersed phase and 14.2 cc./min. for the continuous phase might well have produced noticeable disturbance of steady state conditions. Indeed, increasing either sampling rate i n d e f i n i t e l y would be bound to produce t h i s effect at some sampling rate, and beyond this l i m i t the effect on steady state would, of course, be more and more evident. The present resu l t s did not extend to values as high as some of those used i n studying the influence of the sampling rates on the point concentrations. However, the high sampling rates used to study i f the point concentrations Kary with the sampling rate, were applied only for that study. These rates were not and should not be used i n normal runs because of possible influences on the steady state. Nevertheless, the present resu l t s do show that the steady state condition would not be disturbed by the sampling rates used by the present author i n determining concentration p r o f i l e s i n the column, and 82 i n studying the piston method of sampling, i n both of which the rates of 14.2 cc./min for the water phase, and that of 28.2 cc./min. for the ketone phase, were not exceeded, or, indeed, approached even cl o s e l y , i n the case of the ketone phase. The present resul t s suggest also that the sampling rates used by Choudhury (10) would have no effect on the steady state i n almost a l l of his runs. D) Drop coalescence at the dispersed phase probe entrance: When the column was in f u l l operation the sampling process was carried out using several rates. At the time of sampling, the drops entering the bell-shaped probe were observed c a r e f u l l y . Three models, of behavior can be observed depending on the rate of sampling. F i r s t o f ' a l l , i f the sampling process i s not going on, an interface exists at the end of the probe. When the sampling rate i s low, drops gather at the entrance and come into the probe slowly. Sometimes two or three drops were pushed out ©f the way by the drops t r a v e l l i n g upward into the probe. However, no coalescence was ever observed during operation of the probe. When the sampling rate i s increased, the drops come d i r e c t l y into the dispersed phase probe without touching each other, and without stopping at the probe entrance. This observation was made i n answer to Hawrelak's statement (12) that the residence time of ketone drops at the ketone probe entrance i s too long. Hawrelak also said that an interface was \"sometimes\" created there. Figure §9 shows the result s of the present study of the behaviour of drops at the ketone probe entrance and shows what r e a l l y happens there. Never, during purging or sampling was an interface observed. The statement of Hawrelak i s not believed by the present worker. Hawrelak apparently did not have detailed information on which 83 FIGURE &9:».- Behaviour of the drops at ketone ps?©ls© entrance at various sampling rates* to base h i s statement. Thus he does not say for what conditions an interface was created at the ketone probe entrance. In the present work, observations were made also at much faster rates and consequently ever a bigger range of rates than those used by Hawrelak; absolutely no change from the behaviour sketched in Figure 29 occurred at the ketone probe entrance. The d i f f i c u l t y of the residence time being too long, mentioned by Hawrelak, may be something of a problem for the lower rate cases as Figure 29 implies. This problem arises as a result of the accumulation of drops. But drops accumulate only at very low sampling rates i . e . from 2.0 cc»/min 0 up to 5.0 cc./min.; for the range between 5.0 cc./min. and 18 cc./min. a clearing up of accumulated drops i s observed and the sit u a t i o n i s intermediate between that of the second sketch in Figure 29 and the thi r d sketch (which represents a sampling rate beyond the normal operation). In other words, the residence time of a drop at the ketone probe entrance i s not a serious problem as £bng as the rate i s kept above 5.0 cc./min.. As mentioned l a t e r , an upper value perhaps would be needed to avoid the problem of false point concentrations due to other l i q u i d coming from above or under the sampling position should this happen. This p a r t i c u l a r matter w i l l be discussed l a t e r . E) The effe c t of sampling rate on measured point concentrations: The eff e c t was studied of changing each of the probe sampling rates on the premise that, i f changing either sampling rate produced a change i n the corresponding measured concentration, then the results with the corresponding probe would be suspect. (The absence of such a change would not, however, of i t s e l f prove that the respective probe was producing the correct r e s u l t f o r certain.) Thus such experiments with 85 the continuous phase probe would result i n an apparent decrease i n concentration as the sampling rate increased i f by increasing the sampling rate the regions near drops made up a larger percentage of the sample. In the case of the dispersed phase probe, increasing the sampling rate would have been expected to lower the concentration i f coalescence at the probe mouth had been taking place at low sampling rates with a corresponding mass transfer into the drops during coalescence. (However, coalescence i s not observed.) I f increasing the dispersed phase sampling rate resulted in continuous phase distant from the drops making a larger contribution to the continuous phase taken into the dispersed phase probe along with the drops, then increasing the dispersed phase sampling rate would have been expected to produce a larger volume of higher average concentration aqueous phase as part of the dispersed phase sample. Intthe equation used for cal c u l a t i n g C. . , ° k i ' k i kf _w wi wf \\ the term C, . would have been larger, the factor (C .- C .) would have kf ' wi wf been smaller (since C ;,s given by the continuous phase probe, presumably would have stayed constant), and the factor V^/V^ would have been larger. Depending on the r e l a t i v e magnitude of the effects of the changes i n these two factors some change in C , either up or down, would be> K X expected as a result of the combination of th e i r product with the increased P a r t i a l anwers to the questions posed above are given by the resul t s obtained by varying the sampling rates and observing the effects on the measured point concentrations. Looking at the resul t s presented on t h i s subject shows that l i t t l e or no variation of concentration occurred even i f the sampling 86 rates were as high as 28 cc./min for the dispersed phase probe and 34 cc./min. for the continuous phase probe. One interesting aspect of the behaviour of the ketone probe was that, as predicted e a r l i e r , the volume percent of ketone contained i n a ketone sample (probe) decreases as the sampling rate increases. This ef f e c t does not present too serious a problem with the ketone probe rates used in t h i s work, because the volume percent ketone in the ketone sample was always above 90.0% except for some higher rates of sampling where the volume percent ketone was around 85.0% of the ketone sample. The important point i s that when the volume percent of ketone i s t h i s high, the correction for mass transfer a f t e r sampling i s only small, and whether the water sample i s representative i s of comparatively small importance. As indicated i n Table 4 and as discussed l a t e r , the dispersed phase concentrations obtained with the piston deviate from the average of the probe and piston results within better than £ 1 per cent for two r e s u l t s obtained, one i n each of the Runs 9F and 9G, which Huns ' were not close toaequilibrium conditions. (Runs 9H and 91 were f a i r l y close to equilibrium, 91 more so than 9H.) The other nine res u l t s for these two Runs show deviations from the average of up to - 14%; however, four .results • in addition to the two already mentioned exhibit markedly better agreement than t h i s . (The arithmetic average of the percentage deviation i s calculated to be -1.24% for Runs 9E,9F and 9G. I f 9H i s included also the r e s u l t i s -1.31%.) These resul t s do not prove d e f i n i t i v e l y that the piston r e s u l t s check the probe ones, but they do indicate that such agreement i s f a i r l y probable. I f t h i s agreement i s assumed for the moment, and i f i t i s r e c a l l e d that i n the probe samples the volume percent water i s very low, 87 so that the problem of whether the water sample i s representative i s not too important, then agreement between results from the prober, and result s from the piston, where the volume percent water i s very much higher, does seem to indicate that the water probe gives samples which are representative of the water part of the piston sample. Furthermore, and as a separate argument, i f the resul t s for the piston and for the probe agree with one another, the inference i s very strong that the continuous phase probe gives water concentrations which are representative for use in the cal c u l a t i o n of (C. .) . . After a l l , * kx probe. the hook probe p r e f e r e n t i a l l y samples the main body of the continuous phase. The values of C so obtained would seem to be more appropriate for use with the piston samples, where ketone holdup i s low, than with the probe samples, where ketone holdup i s high, and where the water phase probably i s derived on the average from nearer the drops. Agreement between (C, .) , and (C, .) . , , when the same C .is used ° kx probe kx piston' wi in evaluating each of these, indicates that the region close to the drops does not contribute s u f f i c i e n t solute to the water part of the probe sample to make inappropriate for use i n getting ^ ^ ^ p r o b e One should r e a l i z e that with coalescence at the probe entrance not a problem, the comparison of a piston sample and probe sample where the ketone holdups were the same i n each would produce l i t t l e or no confirmation of either sampling method. Here i t i s convenient to mention that the drops are of diff e r e n t sizes and therefore do not a l l r i s e at the same rate. Therefore, at any elevation i n the column, the concentration of a drop should depend on i t s s i z e . Fortunately the small drops are only a small percentage of the dispersed phase (13) and the large drops are in a f a i r l y small size range (13), a l l r i s i n g at nearly the same speed. F) Duplication of Runs: Figure 23 represents schematically what the author would l i k e to c a l l good reproduction of Choudhury's results considering that feed concentrations are always a b i t hard to copy, and r e p r o d u c i b i l i t y i s rare l y perfect when mass transfer i s the subject treated. As mentioned e a r l i e r , the duplication appeared to be much better a f t e r the small errors of a pipette de l i v e r i n g wrong volume and of evaporation taking place from the contents of the ketone feed tank, had been corrected. Thus, for Run 8, which i s most nearly comparable with Choudhury's Run 65, the average deviation of the result s from those of Choudhury was 1.6% for the continuous phase concentrations, and 2.2% for the dispersed phase concentrations. (These deviations are given without regard to sign.) Table 6 records the point concentrations obtained i n Run 8 and also those of Choudhury. Also, included are the deviation and the percent deviation of the result s of Run 8 from Choudhury's Run 65, at various distances from the nozzle t i p s . G) Concentration study in the Elgin head: The r e s u l t s obtained i n the b r i e f survey of the concentrations i n the Elgin head showed that the concentration of the water phase changed from the value applicable at the point of entry of the water phase into the column, to a lower value, before t h i s phase reached the interface or entered the column proper. However, i t should be noted that Choudhury (10), and also Ewanchyna (11), assumed that the water entered at i t s i n l e t concentration at the interface and that t h i s concentration changed over almost zero height of column to some lower value. These workers also assumed that no further r i s e in ketone concentration occurred above the interface. Both assumptions are not TABLE 6 Concentrations in both phases (smoothed values) for Run 8 of t h i s work and Run 65 o£ Choudhury. Phase. Concentration of acetic acid i n the phase, lb.-moles/ft? x 10 3at distance from nozzle t i p s ( f t . ) of 0.0 O.d 0.5 1.0 2.0 3.0 4.0 5.0 6.0 7.0 Inlet Height of column, f t . Run 8 Water 25.93 26.30 28.20 30.80 36.30 40.50 43.50 45.80 47.40 49.00 50=40 7.383 Choudhury Run 65 Water 26.30 27.00 29.60 32.50 36.90 40.40 43.60 46.20 47.70 49.10 50.40 7.383 Deviation of Run 8 from Run 65 -0.37 -0.70 -1.40 -1.70 -0.60 +0.10 -0.10 -0.40 -0.30 -0.10 0.00 % deviation 1*4 2.6 4.7 5.2 1.6 0.2 0.2 0.9 0.6; 0.2 0.0 Outlet Run 8 Ketone 6.75 7.20 9.00 11.10 15.25 17.80 20.50 22.20 23.40 24.30 24.96 7.383 Choudhury Run 65 Ketone 6.50 7.10 9.30 11.40 15..00 17.80 19.90 21.60 22.90 24.00 24.30 7.383 Deviation of Run 8 from Run 65 +0.25 +O.10 -0.30 +0o30. +0.25 0.0 +0.60 +0.60 +0.50 +0.30 +0.66 % deviation 3.5 1.4 3.2 2.6 1.7 0.0 3.0 2.8 2.2 1.3 2.7 oo CO 90 in accord with the present experimental r e s u l t s . In fact, the concentration of the ketone phase appears to decrease above the interface. It appears that both assumptions used by the previous investigators are over-simplifications. H) Piston r e s u l t s : 1.- Effeot of a piston sample on the steady state: The f i r s t operation done with the piston consisted of finding the steady state requirements: mainly the influence of a piston sample on the outlet concentrations. In Figures 25, 26, i t can be seen that a piston sample disturbs the steady state for at least 15 minutes i f d i s t i l l e d water i s used as a l i q u i d for replacing the column l i q u i d taken out by the sampler. Figure 27 shows that l i t t l e or no disturbances exist when outlet continuous phase i s used instead of d i s t i l l e d water. The f i r s t test was done to learn the maximum deviation from the steady state l i k e l y to be encountered, that occurring when a solution of minimum concentration: d i s t i l l e d water, was put into the piston hole through which the column was not operating. In normal operation of the piston sampling device, the f i l l i n g up of this piston hole with outlet continuous phase gives a maximum deviation of less than 1% in the exit concentration of the continuous phase and of zero in that of the dispersed phase. With the use of d i s t i l l e d water, the deviation was always 7% for the continuous phase and zero for the dispersed phase (except as already noted for one measurement shown on Figure 26). These deviations were obtained using a continuous phase flow rate of 54.8 f t ? / h r . - f t ? and a dispersed phase flow rate of 120.5 f t ? / h r . - f t ? . Knowing the ef f e c t of a piston sample on the steady state permits one to operate the sampler only when steady state has been restored a f t e r the taking of any preceding^ sample. 2.- Comparison of piston and probe dispersed phase samples: Runs 9F, 9F 9 9G, 9H and 91 were made with the purpose of getting comparisons between C . given by the piston and C given by the p r o b e o Table 4 shows the results obtained in these Runs. The author would l i k e to mention here that these Runs were made at various dispersed phase flow rates with a view to have?--; a holdup of dispersed phase which increased in a stepwise fashion as runs progressed. The same continuous phase flow rate of 54 08 f t ? / h r . - f t ? was used for a l l the runs i n Table 4. Cut, unfortunately, and as learned l a t e r , the la s t Run, 91 was performed with the solutions near equilibrium conditions in the column at the axis of the piston. Due to th i s fact, the five r e s u l t s of Run 91, have not been considered in th i s discussion. The fact that the flow rate of the dispersed phase was always increased caused the mass transfer resistance in the column to be lowered as a result of the increased i n t e r f a c i a l area. The lowered resistance created the equilibrium conditions mentioned above. In the Run preceding 91, Run 9H, conditions also were not as far from equilibrium as might be desirable. Run 91 perhaps was close to flooding conditions. According to Sherwood and Pigford (17) (P. 442 Figure 212, curve D) the flooding 3 , 2 velo c i t y for the dispersed phase should be roughly 245.0 f t . / h r . - f t . 3 2 for a continuous phase v e l o c i t y of 54.8 f t . / h r . - f t . . The highest 3 , 2 dispersed phase flow rate used i n this work was 208.3 f t . / h r . - f t . . The increase of the holdup, as mentioned e a r l i e r , was supposed to make the difference (C . - C _) bigger. The following r wi wf 0 0 example explains why this procedure was adopted. Suppose two samples are taken with the piston sampler, but 92 with d i f f e r e n t ketone flow rates to the column so that the holdups i n the column (and i n the corresponding sample) are differento Suppose also that the feed concentrations to the column are adjusted so that i n both runs the values of C, . and C . at the sampler location are the same. k i wi 1 (Such adjustments were not carried out i n the present experiments with the r e s u l t that i n Table 4, C. ., and i n Table 5, C ., are seen to ' k i ' ' wi' decrease as the holdup increases.) Suppose that the two phases of the piston are allowed to come to equilibrium and are then analyzed. As mentioned e a r l i e r , Equation 5 i s used, C. . = C. , - V (C , - C .) 5 k i kf w wi wf Vk Now suppose that of the two samples, numbered one and two respectively, sample 1 has the lower ketone holdup. Hence V < V w2 ^ wx k2 k l and The s i t u a t i o n i s as shown in Figure-29A. For the experiments of this work transfer i s out of the water and into the ketone phase. Hence wf Now at equilibrium: Hence Furthermore, r e c a l l that: wi C C k i ^ kf Cwf * 2 C k f wi k i (C .).=(C . ) _ wi 1 wi 2 ®3 A/, KI SAMPLE I SAMPLE 2 FIGURE 29A<» TETO piston samples having different holdup and Considering.ythe r e l a t i v e volumes of water and ketone in samples 1 and 2 shows that there i s more t o t a l solute available i n sample 1 than i n satmple 2, and, therefore, (C .)_> (C wf 1 wf 2 Then (C . - C < (C . - C J 0 wi wf 1 wi wf 2 and analysis i s less c r i t i c a l for sample 2 where the difference i s greater. Compare now two further runs, one at low holdup and one at high holdup for which the feed concentrations to the column are manipulated so that when the holdup i s increased the difference (C .-C r r wi w i s maintained constant„ For these two runs, i t i s evident that the correction term Vw (C . - C J „ • wi wf k i n Equation 5 becomes less important at the higher holdup because V ~~V i s lower. k As mentioned before the s i t u a t i o n as between samples in the present work i s not as simple as in either of the two i l l u s t r a t i o n s just discussed. Thus in increasing the holdup, C . and C, both • wi k i decreased and so did t h e i r difference. This point w i l l be considered again following a look at the effect on the results of when the analysis i s done: before the phases of the piston sample have reached equilibrium, or a f t e r they have done so. Although the piston samples were analyzed at equilibrium as assumed i n the discussion of the l a s t paragraph, the accuracy i s not affected i f analysis i s done before this condition i s reached. Suppose that the phases of the piston sample can be separated very quickly, just a f t e r the sample has been taken. Then C . and C . w i l l be wx wf p r a c t i c a l l y i d e n t i c a l , and even though the difference (C - C ) can wi w f be determined only very inaccurately the value of w i l l be known quite accurately, because there i s almost no correction to be applied to in determining from i t . (Refer to Equation 5.) On the othe hand, i f equilibrium i s reached before the phases are separated and analyzed, (C , - C _) w i l l be larger, and much more accurately known, wi wf ° as i t w i l l have to be i n order that the comparatively large correction which i s now necessary can be applied to C, _,, and an accurate C. . kf k i obtained. For intermediate cases where equilibrium has not been reache but separation of the phases has taken place some time a f t e r sampling, the correction w i l l be obtainable with only intermediate accuracy. However, since the correction w i l l be of only intermediate s i z e , such accuracy i s appropriate. Evidently, then, and as one would have i n s t i n c t i v e l y assumed, whether or not equilibrium i s reached between the phases of the piston sample before they are separated should not a f f e c t the accuracy of the f i n a l value desired: C, . . But what of the * • • kx eff e c t of increasing the holdup i n the series of piston runs? When one examines the difference (C . - C „), obtainable from Tables 4 and 5, i t wx wf becomes apparent that as the holdup increased t h i s difference remained of the same order of magnitude. This result i s due to the lower levels of concentrations C , and C, . mentioned e a r l i e r , which correspond i n wi k i r these experiments to the phases in the column being c l o s e r to equilibrium than they were at lower holdup. In fact, the phases were almost at equilibrium in the column at the piston location for Run 91. 96 The idea of increasing the holdup to decrease the e f f e c t of small errors i n analysis on the f i n a l valuesof C, . was seized on with k i enthusiasm but without allowing for other results of increasing the holdup, in p a r t i c u l a r that whereby the phases come more nearly to equilibrium at the piston location i n use. Thus, as the holdup i s increasing the mass transfer resistance i s decreasing and f i n a l l y one gets equilibrium conditions i n the lower region of the column. (Most of the extraction of acid from the water took place i n the upper p a r t ! J of the column as noted e a r l i e r . ) The present work teaches that as ketone holdup i s increased, arrangements must be made to feed the water phase at a concentration high enough so that the phases are not in equilibrium at the sampler location. An alternative would be to move the sampler toward the top of the column. The concentration p r o f i l e s in the a x i a l d i r e c t i o n of the column are not known for the high holdup runs of the present work, and therefore the conditions of the phases with respect to equilibrium toward the top of the column are not known i n any d e t a i l . However, the water entering the column was not near equilibrium with the ketone leaving the column for Run 91. Referring to Table 4, which gives the results obtained with both methods of sampling the dispersed phase: piston and probe, i t can be seen that the probe figures and the piston figures are very di f f e r e n t indeed in f i v e cases. In these r e s u l t s , the percent deviation, as defined in connection with Table 4, exceeded 10%. But, the arithmetic average deviation of the results for Runs 9E, 9F, 9G and 9H i s calculated to be -1.31%. Also, i f the dispersed phase concentration at time of analysis, (C, „) . , , i s observed, i t i s seen that these kf piston'. res u l t s of each Run are i n reasonable agreement. The difference between (C, . ) . and (C, . ) . , appearing in Table 4 probably are due to k i probe k i piston ' r ° . 97 three major errors caused by the manipulations associated with the analyses, as shown in Appendix VIII. An error of 21.3% can be obtained i f the errors are supposed to be cumulative. This result explains why some values of the dispersed phase concentrations obtained with both methods are so d i f f e r e n t . None of the new values of (C, .) determined with the piston k i are comparable with those obtained by Hawrelak, because di f f e r e n t flow rates were employed by each worker. However, i t i s possible to compare Hawrelak's deviations and the deviations obtained i n this work. The values under ten percent deviation i n this work are comparatively numerous (10 out of 15 values i f only Ruin 91 i s excluded, or 6 out of 11 values i f both of Runs 9H and 91 are excluded). Hawrelak had only 3 values out of 20 in this range of from zero to 10% deviation. Another interesting remark concerning Hawrelak's resul t s as shown in his thesis page 65, i s that as many as 6 piston results are compared with a single probe r e s u l t . In other words, Hawrelak unfortunately did not take a probe sample each time he took a piston sample. (He even uses a probe sample from one run with a piston sample from another run done on a d i f f e r e n t day.) An examination of his data books confirms these remarks. He was of course operating on the assumptions that he was able to maintain true steady state, that he could reproduce his p r o f i l e s very accurately, and that he could take probe samples very reproducibly. However, Hawrelak's Run 84, performed using a water flow rate of 90.0 f t ^ / h r . - f t ? , and a ketone flow rate of 90.5 f t ? / h r . - f t ? , to give a volume percent ketone of 11.6, shows that the average difference ' 3 3 (C . - C . ) , 1.0 lb.-moles/ft.xlO , was bigger i n this run than in his wi wf other runs. This average difference i s also bigger than those of the * Should be halved for comparison with the figures of Table 4. 98 present work. However, the percentage deviation of his piston results from the average of his probe and piston re s u l t s i s generally higher for Run 84 than for h i s other Runs, (except for Run 79) and also generally higher than the present deviations. (None of Ilawrelak's values are below 10% i n Run 84.) This i s another indication that the difference (C . - C „) must be made higher ( by increasing the i n l e t concentration wi wf of the water phase), but that also the volume r a t i o , ^^/V , in Equation 5 must be smaller. Considering the present r e s u l t s , the author believes that the check between the two methods of sampling has been put on the right track. It i s now known that the difference between C . and C „ must be wi wf increased by increasing the column feed water concentration. Then at high holdups agreement between the two methods of sampling could be expected assuming that the uncertainties of the analysis can be overcome. 99 SUMMARY The following statements can be made based on the measurements described i n t h i s thesis: 1. - A r e l a t i o n was found between the dispersed phase flow rate and the time to reach the steady state. This was done for an average continuous 3 2 phase flow rate of 54.8 f t . / h r . - f t . The time depended on the i n i t i a l concentration l e v e l inside the column and on whether an interface with ketone above i t existed in the Elgin head. 2. - A c a l i b r a t i o n was done to get the r e l a t i o n between the purging rate and the minimum purging time required to change the solution present i n the sampling probes. 3. - No coalescence can be seen at the dispersed phase probe entrance. Such coalescence takes place only when not sampling. 4. - The steady state seems to be influenced more by the interface c o n t r o l l e r than by the sampling rate used with the probes. No r e a l e f f e c t of sampling on the steady state was found for sampling rates up to 8.8% of a dispersed phase flow of 72.8 f t ? / h r . - f t 2 and 4.4% of a 3 2 continuous phase flow rate of 54.8 f t . / h r . - f t . . (These sampling rates in cc./min. were 28.2 and 14.2 respectively.) 5. - The point concentration of either phase as measured by the respective probe did not vary with sampling rate even i f rates as high as 34 cc./min. for the water probe and 28.4 cc./min. for the ketone probe were reached. 6. - Choudhury's Run 65 was duplicated within 2,2% (average deviation) for the dispersed phase concentrations and 1,6% (average deviation) for the continuous phase concentrations. It should be pointed out that the* i n l e t concentrations were always as close as possible to those of Choudhury, but not exactly the same. 7. - The dispersed phase i n i t i a l concentrations obtained by the piston 100 method v e r i f y those obtained by means of the probes within 2% deviation for three samples (not including Run 91) and within 10% deviation for ten other resul t s out of fifteen,, A f i n a l check of the probe sampler by the piston sampler has not been obtained„ However, much closer agreement has been demonstrated i n the present work than had appeared in the e a r l i e r investigation. I f a large difference between and can be obtained at the same time as a high holdup of ketone then a check of (C, .) . . and (C, . ) ' i s expectedo But the holdup should k i piston k i probe * not be as high as in the probe sample :then, as noted e a r l i e r , the methods become i d e n t i c a l . 101 RECOMMENDATIONS After having worked for almost two years with the apparatus described previously, i t can be recommended that an automatic le v e l c o n t r o l l e r for the interface would re s u l t i n a great improvement of the operation. I n s t a l l a t i o n of a str i p p i n g column would permit work with the spray column to be carried out on a more regular basis. The lengthy back-washing procedure could be carried out while data were being obtained. As far as the apparatus i s concerned, these two suggestions would f a c i l i t e the research work: considerably. Concerning the theor e t i c a l aspect of the research, more study with the piston i s needed. A study could be made by varying the continuous phase concentration and flow rate, so that the phases are not i n equilibrium at the location of the sampler even i f the dispersed phase flow rate i s also high. In the l i g h t of the previous suggestions, an attempt to increase the holdup of the column would be welcome, i n order to get more and more volume percent ketone i n a piston sample. However, to do t h i s , new dispersed phase pump would be needed because the maximum capacity of the present pump was reached in Run 91. It should be borne i n mind, however, that the column may have been operating close to the flooding condition i n that run, and much higher holdups may not be possible. Along with these suggestions, a recommendation i s made that in measuring the volumes of the phases of the piston samples the procedure of pourring from a flask into a graduate be changed back to the method of Hawrelak, where the phases of each piston sample were colle c t e d d i r e c t l y i n a graduated f l a s k . Such flasks should be graduated over a larger range than were those of Hawrelak. The larger range i s necessary to provide for higher holdups than those encountered 102 by him. The graduated portion should extend from 0 to 40 ml. About the analysis technique, i t i s also recommended that a bigger pipette should be used for purpose of getting the least possible error due to measurement of the volume of the samples. 103 NOMENCLATURE Except where noted otherwise, the following nomenclature wafe used throughout the Thesis: Symbols 2 A Cross-sectionnal area of column, f t . . a 2 3 I n t e r f a c i a l area per unit volume of extraction column, f t . / f t . . C* Phase solute concentration which could be i n equilibrium with 3 3 the concentration of the other phase, lb.-moles/ft. xlO . 3 3 C Solute concentration, lb.-moles/ft. x 10 . C, . I n i t i a l concentration of solute i n ketone phase during internal k i r ° 3 3 sampling, lb.-moles/ft. x 10 C ^ Concentration of solute i n ketone phase of ketone sample as measured at time of analysis, lb.-moles/ft? x 10? C . I n i t i a l concentration of solute in water phase during internal wi sampling, lb.-moles/ft°. x 1Q^. C . Concentration of solute i n water phase of ketone sample as wf 3 3 measured at time of analysis, lb.-moles/ft. x 10 , h Effective height of extraction section of column, f t . , (measured from nozzle t i p s to interface.) K. Over-all mass transfer c o e f f i c i e n t , lb.-moles (hr.) ( f f r ) (lb.-moles/ft?) Ka Over-all extraction capacity c o e f f i c i e n t , hr L Phase flow rate, f t ? / h r . - f t ? . N Amount of solute transferred based on i n l e t and outlet concentrations, lb.-moles/hr.. NN\"/A Not applicable. N Number of o v e r - a l l transfer units based on x phase, ox V Volume of ketone phase i n the ketone.sample, cc. IV V Volume of -Water phase i n the ketone*, sample, cc, w d Integral sign. D i f f e r e n t i a l sign. Subscripts 1 Inlet. 2 Outlet. k Ketone phase. w Water phase. c Continuous phase. d Dispersed phase. M Measured. X Phase x. i i n i t i a l . f F i n a l . t Toluene. Superscript * Equilibrium value. 105 LIST OF REFERENCES 1. Geankoplis, C. J. and Hixson, A. N., Ind. Eng. Chem. 42:1141,1950. 2. Geankoplis, C. J., Wells, P. L., and Hawk, E. L., Lnd. Eng. Chem. 43 : 1848, 1951. 3. Newman, M. L,, Ind, Eng. Chem. 44 : 2457, 1952. 4. Geankoplis, C. J. and Kreager, R. M,, Ind. Eng. Chem. 45: 2156, 1953. 5. Gier, T. E. and Hougen, J. 0., Ind. Eng. Chem. 45 : 1362, 1953. 6. Miyauchi, T., U. S. At. Energy Comm. Kept. UCRL 3911, AUGUST 1957. 7. Smoot, L. D. and Babb, A. L. , IS EC Fund. jL : 93, May 1962. 8. Heertjes, P. M., Holve, W. A., and Talsma, H., Chem. Eng. S c i . 3 : 122, 1954. 8. Cavers, S. D. and Ewanchyna, J. E., Can. J. of Chem.Eng. 35 : 113, 1957. 10. Choudhury, P. R., M. A. Sc. Thesis, University of B r i t i s h Columbia, 1959. 11. Ewanchyna, J. E. , M. A. Sc. Thesis, University of Saskatchewan, 1955. • 12. Hawrelak, R. A., M. A. Sc. Thesis, University of B r i t i s h Columbia, 1960. 13. Rocchini, R. J., M. A. Sc. Thesis, University of B r i t i s h Columbia, 1961. 14. Carbide and Carbon Chemicals Co., Pamphlet \"Ketones\",; 34, 1953. 15. Kirk, R. E., and Othmer, 0. F. , Encyclopedia of Chem. Tech. 1st Sup. : 317, 1957. 16. Jonhson, H. F. and B l i s s , H., Trans. Am. Inst. Chem. Engrs. 42 : 331, 1946. 106 17. Sherwood, T. K. , and I'igford, R. L. , ABSORPTION and EXTRACTION, 2nd ed. : 442. Mc Graw-flill Book Co., Inc., New York, 1952. A P P E N D I C E S 108 APPENDIX I RotametersCalibrations. The rotameters used throughout t h i s work were calibrated a f t e r the cleaning and re-assembly of the apparatus at the beginning of the research. The continuous phase rotameter was calibrated using d i s t i l l e d 3 3 water (saturated with M.I.U.K.) and containing 50.2 lb.-moles/ft. xlO of acetic acid. For the dispersed phase rotameter, M.I.B.K. (saturated with d i s t i l l e d water) and containing 6.54 lb.-moles/ft? xlO^ of acetic acid was used. Both ca l i b r a t i o n s were carried out at room temperature. The re s u l t s are given i n Table 7 and plotted on Figure 30. Since the smallest divisions of the scales of the rotameters just mentioned were one quarter of the next largest d i v i s i o n s , i t became obvious that a change of scale to cm. and mm. div i s i o n s would inprove the precision of the readings since then the decimal system could be used d i r e c l y ; in addition, the smallest divisions of the scale would be smaller i n actual magnitude than previously. A second c a l i b r a t i o n was done and the result s are shown i n Table 8 and plotted on Figure 31. TABLE 7 Rotameter Calibrations. Water Rotameter. S e r i a l number 14143. Room and l i q u i d temperature: 72.5 °F Liquid: d i s t i l l e d water saturated with M.I.B.K. and containing 50.2 lb.-moles/f t\"? xlO of acetic acid. Rotameter Reading. Rate of flow, ft?/hr. 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.1615 0.3415 0.5080 0.6675 0.8315 0.9915 1.1420 109 KETONE P H A S E R O T A M E T E R WATER PHASE ROTAMETER 72.5 F. 0.04 0.08 0.12 0.16 ROTAMETER READING 020 FIGURE 30. Rotameter Calibrations. Rotameter S e r i a l No.: 14143. No.: 14142. Ketone Rotameter. S e r i a l number 14142. Room and l i q u i d temperature: 72.5 °F Liquid: M.I.B.K. saturated with d i s t i l l e d water and containing 6.53 lb.-moles ft\"? x 10^ of acetic acid. Rotameter Reading, Rate of flow, ft?/h 0.02 0.2320 0.04 0.4150 0.06 0.6150 0.08 0.7850 0.12 1.1850 0.16 1.5510 TABLE 8 Rotameter Calibrations. Water Rotameter. S e r i a l number 14143. Room and l i q u i d temperature: 74 °F. Liquid: d i s t i l l e d water saturated with M.I.B.K. and containing 50.0 lb.-moles/ft^ x 10^ of acetic acid. 3 Rotameter Reading. Rate of flow, f t . / h r . mm. 2S 0.0739 50 0.1680 75 0.2682 100 0.3885 125 0.5115 150 0.6273 175 0.7500 200 0.8840 225 1.0200 Ketone Rotameter. S e r i a l number 14142. Room and l i q u i d temperature: 73 °F. Liquid: M.I.B.K. saturated with d i s t i l l e d water 3 3 and containing 6.53 lb.-moles/ft. x 10 of acetic acid. 3 Rotameter Reading.(mm.) Rate of flow, f t . / h r . 50 0.2120 75 0.3830 100 0.5800 125 0.7860 150 1.0220 175 1.2780 200 1.5150 225 1.7980 I l l ROTAMETER READING MM. F I G U R E 31. Rotameter Calibrations. Mm Scale., APPENDIX II Run data. The following Tables ( 9 and 10) show the purpose of each of the runs and the o v e r - a l l transfer data. A l l Runs done in the present work, which were used to produce the results reported i n t h i s thesis are l i s t e d here. Also included are the following c h a r a c t e r i s t i c s which concerned the sasupling probes: sampling rate for both phases, the sampling rate as a percentage of the phase flow rate, the purging time, and the temperature at which the run was performed. Also included in Table 10A are the data of Runs 5D and 8 f£©m which the concentrations p r o f i l e s were obtained. 113 TABLE 9. Run i number. Purpose of Run. Sampling Sampling Purging Tern] rate , cc./min . as % of time. °F phase min. (ave, Water Ketone flow rate. phase. phase. Water phase; Ketone phase. *1C Sampling rate study V V V V V 77 ID Sampling rate study V V V V V 78 *1E Sampling rate study V V V V V 78 •IF Sampling rate study V V V V V 77 1G Sampling rate study V V V V V 75 2 Sampling rate study V V V V V 76 2A Sampling rate study V V V V V 74 2B Sampling rate study V V V V V 78 2C Sampling rate study V V V V V 78 2D Sampling rate study V V V V V 72 *3 Sampling rate study V V V V V 74 *3A Duplication of Run 65 of Choudhury 9.0 13.0 2.9 3.9 9.0 75 *3B Duplication of Run 65 of Choudhury 9.0 13.0 2.3 3.3 9.0 75 *3C Point concentration versus rate V V V V V 74 4 Duplication of Run 66 of Choudhury 8.2 12.4 4.1 3.8 8.0 72 *5A Reproduction of Run 3A 8.0 14.0 2.8 4.2 9.0 72 5B, Duplication of Run 1 65 of Choudhury 9.3 13.8 3.2 4.2 9.0 76 5D Duplication of Run 65 of Choudhury 8.9 14.3 3.0 4.3 9.0 76 *5E Duplication of Run 65 of Choudhury 8.9 13.0 2.9 3.9 9.0 78 5F Point concentration versus sampling rate V V V V V 71 5G Point concentration versus sampling rate V V V V V 74 5H Point concentration versus sampling rate V V V V V 72 Steady state study N/A N/A N/A N/A N/A 74 6 J 1 Steady state study N/A N/A N/A N/A N/A 73 *6H Steady state study N/A N/A N/A N/A N/A 76 *6I 0 Steady state study N/A N/A N/A N/A N/A 73 *6G\" Steady state study N/A N/A N/A N/A N/A 72 6K Steady state study N/A N/A N/A N/A N/A 69 6L Steady state study N/A N/A N/A N/A N/A 74 6M Steady state study N/A N/A N/A' N/A N/A 73 * Runs which were not used i n this thesis. V = Varied. N/A = Not applicable. 114 TABLE 9 Cont.. Run Purpose of Run. Sampling Sampling Purging Temp, number. rate, cc./min. as % of time. o F phase min. (ave.) flow rate. Water Ketone Water Ketone phase, phase. phase, phase. 7 Sampling rate study V V V V V 75 *7A Sampling rate study V V V V V 75 7B Sampling rate study V V V V V 74 7C Sampling rate study V V V V V 75 7D Sampling rate study V V V V V 72 7E Steady state versus sampling rate V V V V V 75 7F Steady state versus sampling rate V V V V V 75 8 Duplication of Run 65 of Choudhury 10.0 10.0 3.2 2.5 8.0 75 *8A Point concentration versus sampling rate 7.1 V V 10.0 76 *8B Point concentration versus sampling rate 7.9 V 2.5 V 10.0 75 9A Effect of a piston sample on steady state N/A N/A N/A N/A N/A 76 9B Effect of a piston sample on steady state N/A N/A N/A N/A N/A 76 9C Effect of a piston sample on steady state N/A N/A N/A N/A N/A 76 9D Effect of a piston saiisple on steady state N/A N/A N/A N/A N/A 74 9E Piston samples taken 12 o0 13.0 3.8 2.3 10.0 76 9F Piston samples taken 10.0 9.4 3.2 1.7 10.0 76 9G Piston samples taken 9.6 8.8 3.0 1.6 10.0 75 9H piston samples taken 6.2 8.2 1.9 1.1 10.0 77 91 Piston samples taken 6.5 7.6 2.1 0.8 lOcO 73 * Runs which were not used in t h i s thesis. V = Varied. N/A = Not applicable. TABLE 10 Over-all transfer data. Run Concentrations, number lb.-moles/ft? x 10? Flow rates f t 3 / h r . - f t ? Water phase. Ketone phase. Water. Ketone, wl w2 k l k2 w Linear Velocity through the nozzle t i p s , f t / s e c . Over-all a c e t i c acid transfer rates. lb.-moles/ ft? x 10 3. N w N. Percent deviation Nw\" Nk xlOO 1C 50.2 26.2 6.3 24.9 55.8 68.7 0.338 16.4 15.7 4.2 ID 50.7 26.5 6.1 24.8 54.6 67.4 0.332 16.2 15.4 5.1 -4-E 53.7 28.9 6.7 26.3 54.6 67.4 0.332 16.6 16.2 2.4 IF 37.9 21.5 5.4 18.5 54.8 70.6 0.347 11.0 11.3 2.7 1G 39.8 22.5 5.8 17.3 54.1 65.9 0.324 11.5 10.9 5.4 2 37.6 20.4 5.0 18.4 54.6 68.6 0.335 11.5 11.3 1.8 2A 38.3 22.2 5.5 18.8 55.1 68.7 0.335 10.9 11.2 2.7 2B 39.4 20.5 5.2 19.0 54.6 72.7 0.357 12.6 12.3 2.4T 2C 40.3 21.8 5.8 19.6 54.6 $9.0 0.338 12.* 11.§ §.6 2D 39.5 19.9 4.9 19.1 52.9 71.7 0.352 12.7 12.5 1.6 3 39.7 21.3 5.7 19.3 52.9 71.8 0.353 11.9 12.0 0.8 3A 38.4 19.5 4.8 18.6 52.7 71.7 0.352 12.2 12.1 0.8 3B 38.4 19.5 4.9 18.8 52.4 72.7 0.357 12.2 12.4 1.6 3C 34.7 17.9 4.7 16.4 54.7 71.8 0.353 11.3 10.3 9.3 4 34.4 12.1 4.9 14.9 36.3 71.5 0.352 9.9 8.8 11.8 5A 50.6 25.0 6.6 25.0 52.7 72.7 0.357 16.5 16.4 0.6 5C 50.4 24.8 6.5 24.1 51.2 72.7 0.357 16.1 15.8 1.9 5D 51.6 24.9 6.5 24.5 51.1 71.7 0.352 16,7 15.8 5.5 5E 51.0 24.9 6.5 24.2 51.2 71.7 0.352 16.4 15.6 5.0 5F 51.1 25.6 6S5 24.5 52.9 72.7 0.357 16.5 16.0 3.1 5G 50.3 26.3 6.5 24.4 54.1 72.7 0.357 15.9 15.9 0.0 5H 51.4 25.6 6.4 24.6 52.9 71.7 0.352 16.7 16.0 4.3 6 I l 49.9 24.4 6.5 24.6 54.9 72.9 0.358 17.1 16.2 5.4 6 J 1 49.9 25.7 6.6 24.8 54.9 73.6 0.362 16.3 16.4 0.6 6H 49.7 24.6 6.5 24.3 54.9 72.9 0.358 16.9 16.3 3.6 6 I 2 6G 49.9 25.1 6.6 24.9 54.9 73.1 0.359 16.0 16.4 2.4 50.1 25.1 6.6 24.9 54.9 72.9 0.358 16.8 16.3 3.0 6K 50.3 25.1 6.8 25.1 54.9 72.9 0.358 16.9 16.4 3.0 6L 50.7 25.3 6.7 25.3 54.8 72.8 0.358 1711 16.6 3.0 OI TABLE 10 Cont. Over-all transfer data. Run Concentrations, Flow rates number lb.--moles/ft? x 10 3 f t 3 / h r . - f t ? Water phase. Ketone phase. Water. Ketoi Cwl Cw2 C k l Ck2 L w \\ 6M 50.7 25.5 6.7 25.5 54.8 73.2 7 48.7 24.4 614 24.3 54.8 72.8 7A 49.7 24.8 6.9 24.6 54.8 72.8 7B 49.6 25.0 6.6 24.5 54.8 72.8 50. 49.5 25.1 6.7 24.6 54.6 72.8 7D 50.7 25.9 7.0 25.3 54.8 72.8 7E 51.3 25.4 6.6 25.5 54.8 72.8 7F 51.9 25.8 7.0 25.8 54.8 72.8 8 50.4 25.9 6.8 25.0 54.8 72.7 8 A 51.6 27.2 7.8 25.8 55.0 72.8 8B 51.5 27.1 8.0 25.7 S5.Q 73.2 9A 43.5 16.5 6.8 18.3 54.3 120.0 9B 43.4 16.7 6.8 18.3 54.2 119.6 9C 51.4 17.6 6.8 21.8 54.2 120.5 9D 51.4 17.6 6.8 21.5 54.2 120.5 9E 49.5 16.5 6.4 21.1 54.8 120.5 9F 49.9 16.9 6.4 21.3 54.8 119.6 9G 50,6 17.6 6.5 21.7 54.8 120.5 9H 50.8 14.4 6.6 18.3 54.8 169.2 91 50.5 13.9 6.7 16.3 54.8 208.0 Linear Velocity Over-all acetic Percent through the acid transfer deviation nozzle t i p s . rates. lb.-moles/ f t . / s e c . ft?x 10 3. N N, N -N. w k w k : N 0.360 16.9 16.9 0.0 0.358 16.5 16.0 1.9 0.358 16.7 15.8 5.5 0.358 16.5 16.0 3.1 0.358 16.3 16.0 1.9 0.358 16.6 16.3 1.8 0.358 17.4 16.9 2.9 0.358 17.5 16.8 4.1 0.357 16.5 16.2 1.8 0.358 16.4 16.1 1.8 0.360 16.4 15.9 3.1 0.337 17.9 16.9 5.7 0.335 17.7 16.8 5.2 0.338 22.4 22.1 1.4 0.338 22.4 21.7 3.2 0.338 22.2 21.7 2.3 0.335 22.2 21.8 1.8 0.338 22.2 22.4 0.9 0.475 24.4 24.3 0.4 0.584 24.6 24.5 0.4 TABLE 10A. Data of concentration p r o f i l e s for Runs 5D and 8. (not smoothed). Run 5D Run 8 neignt Volume in a toncemrations, Height Volume i n a Concentrations, of the ketone sample. lb.-moles/ft3x 10. 3 of the sample.(ketone) lb.-moies/ft 3xlO? probes. probes. f t . \\ \\ Cwf C k f C k i C . wi f t . V w V k Cwf C k f C k i C . wi 0.078 0.0 48.0 - 6.57 6.57 26.40 0.078 0.8 32 .6 - 7.11 7„11 26.25 0.445 5.4 37.6 18.90 9.24 9.11 27.78 1.161 2.3 29 .3 24.50 12.00 11.44 31.60 1.060 5.0 37.0 22.40 12.00 10.78 31.35 2.161 2.2 28 .6 31.55 15.90 15.49 37.20 1.661 5.0 36.0 26.85 13.86 12.97 33.32 3.1G1 2.2 28 .2 35.65 18.20 17.81 40.60 2.411 4.8 36.2 30.30 15.90 15.00 37.20 4.161 2.4 28 .4 39.90 20.40 20.04 44.15 3.161 4.0 35.5 33.45 17.52 16.80 39.84 5.161 2.0 28 .2 41.30 22.08 21.75 46.00 4.161 *1.0 34.8 36.45 18.87 18.23 42.00 6.161 2.4 26 .4 45.30 23.50 23.28 47.75 5.161 3.8 33.7 90.35 18.87 20.69 46.29 7.286 2.0 26 . 2 45.15 24.70 24.36 49.60 6.161 4.2 31.4 42.20 23.16 22.25 48.99 7.286 4.0 30.0 43.50 24.96 23.86 51.72 Average i n l e t water concentration: 51.61 Average outlet water concentration: 24.70 Average i n l e t ketone concentration: 6.50 Average outlet ketone concentration: 24.50 Average i n l e t water concentration: 50.40 Average outlet water concentration: 25.95 Average i n l e t ketone concentration: 6.75 Average outlet ketone concentration: 24.96 1 APPENDIX III Study of time needed to obtain steady state inside the column. As mentioned e a r l i e r i n the thesis, two di f f e r e n t concentra-tion p r o f i l e s exist inside the column before beginning a run. These concentrations depend on the l i q u i d f i l l i n g the column at that time. If the column i s f i l l e d with continuous phase feed before the run s t a r t s , and the column hasn't been operated before, the time needed to reach the steady state i s longer than i t i s when the column has been operated before and contains both phases. This result i s due to the higher average concentration which exists i n the column i n the former case and also increased d i f f i c u l t i e s i n steadying out the interface. Table 11 records the results obtained i n f i v e runs* where the conditions of the former case applied. The concentrations of both outlet solutions: continuous phase, and dispersed phase, were sampled as functions of time. The height of the column has been noted in the same Table \\\\, The i n l e t concentration for the continous phase feed 3 3 solution averaged 50.4 lb.-moles/ft. xlO • and, for the dispersed phase 6.6 lb.-moles/ft? xlO^. As just mentioned, i f the concentration p r o f i l e s inside the column resulted from a previous run, the time needed to reach the steady state i s shorter, due to the smaller average concentration existing at the s t a r t of the run, and due to the existence of a l i q u i d l i q u i d interface at the s t a r t of the run. Table 12 records the result obtained for th i s case. A l l the runs of th i s sort. Time, min. *Run 6 1 . Concentrations lb.-moles/ft? xlO? TABLE 1 1 . Steady state study. •Run 6 J . Column Concentrations, height, lb.-moles/ft 3 xlO? Column height c C c C f t . C , C „ C, n f t . wl w2 k l k 2 wl w2 k l k 2 0 4 9 . 9 0 — 6 . 5 3 _ 4 9 . 8 6 . 5 0 _ 5 1 9 . 2 2 3 . 4 29.1 2 5 . 3 1 0 2 3 . 0 2 4 . 2 2 9 . 1 2 5 . 3 1 5 2 4 . 6 2 4 o 5 2 7 o 2 2 5 . 1 2 0 2 5 . 0 2 4 . 6 2 6 . 7 2 5 . 0 2 5 2 5 . 3 2 4 . 8 4 9 . 9 2 6 . 3 6 . 6 2 4 . 9 7 . 4 6 3 0 2 5 . 2 2 4 . 8 2 6 . 4 2 4 . 9 7 . 4 4 3 5 2 5 . 3 2 4 o 9 2 5 o 9 2 4 . 8 7 . 4 3 4 0 2 5 . 2 2 4 . 9 7 . 4 6 2 5 . 7 2 4 . 8 7 . 4 2 4 5 2 5 . 1 2 4 . 9 7 . 4 4 2 5 . 7 2 4 . 8 7 . 4 1 5 0 4 9 . 9 2 5 . 1 6 . 5 3 2 4 . 9 2 5 . 7 2 4 u 8 7 / 3 8 5 5 2 5 . 2 2 4 . 9 7 . 4 3 2 5 . 5 2 4 . 7 7 . 3 8 6 0 2 5 . 1 2 4 . 9 4 9 . 9 2 5 . 4 6 . 6 2 4 . 7 7 . 3 8 6 5 2 5 . 1 2 4 . 9 5 5 . 4 3 2 5 . 2 2 4 . 6 7 . 3 7 7 0 2 5 . 1 2 4 . 9 7 . 4 3 2 5 . 3 2 4 . 6 7 . 3 6 7 5 2 5 . 1 2 4 . 9 7 . 4 2 8 0 2 5 . 1 2 4 . 9 8 5 9 0 4 9 . 9 2 4 . 9 6 „ 4 7 2 4 . 8 7 . 4 3 * Runs started with continuous phase feed f i l l i n g the column. In the blank spaces, no readings were taken. + In t h i s run there was d i s t i l l e d water i n the annular space of the Elgi n head TABL6 11 Cont. Steady state study. Time, *Run 6M min. Concentrations, lb.-moles/ft? xlO 3. wl w2 k l k2 Column height, f t . •Run 9A Concentrations, lb.-moles/ft 3 xlO? wl w2 k l k2 Column height, f t . 0 50.5 5 10 15 20 25 30 50.7 35 40 45 50 55 60 65 70 75 51.0 6.6 31.9 28.3 27.3 26.7 26.3 26.2 26.2 25.9 25.7 25.5 25.3 25.5 25.2 25.7 6.8 26.3 26.2 25.9 25.7 25.3 25.5 25.7 25.5 25.5 25.5 25.5 25.5 25.2 7.46 7145 7.43 7.42 7.41 7.41 43.6 43.3 43.4 27.9 21.6 20.6 19.4 17,7 17.0 16.6 16.5 16.5 16.5 6.7 6.8 6.8 21.8 21.0 20.4 20.2 18.7 18.2 18.2 18.3 18.3 18.3 7.29 7.24 7.24 7.25 7.27 7.28 7.29 7.30 6.7 * Run started with continuous phase feed f i l l i n g the column. The blank spacesmean that no reading was taken. to o TABLE 1 1 Cont. Steady state study. Time, 'Run 9 C . min. Concentrations, lb.-moles/ft 3x 1 0 ? wl w2 k l k 2 Column height, f t . •Run 9 1 . Concentrations, lb.-moles/ft 3x 1 0 3 . wl Wi k 2 Column height, f t . 0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 0 5 5 6 0 6 5 7 0 7 5 5 1 . 3 5 1 . 5 5 1 . 4 2 8 . 9 6 . 8 2 Q . 9 2 0 . 8 1 9 . 2 1 8 . 8 1 8 . 5 1 6 . 3 1 8 . 2 1 8 . 0 1 8 . 0 6 „ 8 6 . 8 2 5 . 2 2 3 . 9 2 3 . 6 2 2 . 9 2 2 . 5 2 2 . 2 2 2 . 2 2 2 . 2 2 2 . 0 2 2 o 0 7 . 3 6 7 . 4 2 7 . 4 0 5 0 . 4 5 0 . 4 1 4 . 2 1 3 . 9 1 4 . 0 1 4 . 2 1 4 . 2 1 3 . 9 1 3 . 9 1 3 . 8 1 3 . 9 1 3 . 9 1 3 . 6 1 3 . 8 1 3 . 9 1 3 . 9 1 3 . 9 6 . 7 6 . 7 1 4 . 6 1 6 . 0 1 7 . 2 1 7 . 6 1 7 . 4 1 6 . 3 1 6 . 1 1 6 . 2 1 6 . 3 1 6 . 2 1 6 . 3 1 6 . 3 1 6 . 2 1 6 . 4 1 6 . 4 a 7 . 2 8 7 . 2 9 7 . 3 7 ? Run started with continuous phase feed f i l l i n g the column, a Interface c o n t r o l l e r raised, l . - i n . a f t e r t h i s reading. The blank spacesmean that no reading was taken. to TABLE 12. Stendy state study. Time, min. 'Run 6K. 0 5 10 15 20 25 30 35 40 45 50 Concentrations, lb.-moles/ft 3xl0 3. wl 50.1 50.3 50.3 viJi 26 04 24.8 24.8 25.0 25.0 25.2 25.1 25.0 25.1 25.1 k l 6.5 6.8 6.8 Column height, **Run 6L. Concentrations, lb.-moles/ft 3x 10 3. Column height, f t . f t . C C C C C k2 Cwl w2 k l k2 7.45 50.8 6.5 23.6 7.45 26.6 23.2 7.40 24.4 7.45 24 o 5 24.6 7.39 24.8 7.45 24.6 25.1 7.39 25.0 7.40 25.0 25.3 25.0 7.40 25.2 25.3 7.38 24.9 7.40 50.6 25.2 6.7 25.4 25.0 7.40 25.5 25.2 7.38 25.2 7.40 25.4 25.3 25.1 7.44- 25.3 25.3 7.38 25.1 7.41 50.5 25.2 6.7 25.3 7.38 with a column which operated before. Blank space means that no reading was tnken. TABLE 1 2 Cont. Steady state study. a Interface con t r o l l e r lowered 1 . - i n . a f t e r t h i s reading. Blank spaces mean that no reading was taken. Time, ••Run 9B ••Run 9 H min. Concentrations Column Concentrations, Column lb. -moles/ft 3x 1 0 3 . height, lb .-moles/ft 3x 1 0 3 . height, f t . f t . C G c c C . C „ C, , C, „ 0 5 wl w2 k l k 2 wl w2 k l k 2 4 3 . 5 1 6 . 6 6 . 8 1 8 . 2 7 . 3 3 5 0 . 6 1 8 . 4 6 . 6 1 7 1 6 a 7 . 3 7 1 0 1 6 . 7 1 8 . 3 1 4 . 7 1 7 . 5 1 5 1 6 . 6 1 8 . 3 7 . 3 4 1 4 . 1 1 8 . 4 2 0 1 6 . 6 1 8 . 3 2 5 1 6 . 5 1 8 . 3 7 . 3 5 3 0 4 3 o 4 1 6 . 6 6 . 8 1 8 . 3 3 5 1 6 . 7 1 8 . 3 7 . 3 7 1 4 . 4 1 8 . 7 4 0 4 3 . 2 6 . 8 4 5 1 7 o 0 1 8 . 5 7 . 3 8 1 4 . 4 1 8 , 7 5 0 1 6 . 5 1 8 . 4 5 5 1 6 . 5 1 8 o 5 7 . 3 8 1 4 . 4 1 8 . 4 7 . 3 7 6 0 6 5 5 0 . 9 6 . 6 7 . 3 7 * * Run started with a column which had been operated before. to OJ Table 13 summarizes the results obtained in a l l these runs. This Table shows, for each di f f e r e n t way of s t a r t i n g a run, the time i n minutes which one should wait to get steady state using the dispersed phase flow rate mentioned. The continuous phase flow rate used was at a value of 54A8 f t ^ / h r . - f t ? TABLE 13. Summary of the time needed to obtain steady state depending on the concentration in the column . L = constant = 54.8 f? . / h r . f t ? w Continuous phase feed f i l l i n g the column before the st a r t of a run and no dispersed phase present. Runs started from a column which had been operated before. Run Time, Dispersed Run Time, Dispersed number. min, phase flow number. min. phase flow rate, L^ rate, 1^ 6J 40.0 73.60 6K 20.0 72.90 6M 45.0 73.20 6L 25.0 72.80 9C 30.0 120.00 9B 15.0 120.20 9A 35.0 120.00 911 12.5 169.40 91 12.5 208.00 Note: The time values were taken from graphs made of the results given in Table 11 and 12. For safe operation, 5 minutes more should be added. Run 6 1^ has not been included; t h i s run was begun with d i s t i l l e d water i n the Elgin head. 125 APPENDIX IV Study of the minimum purge time for changing the concentration of the solution in the probes using various rates. The following Tables, 14 to 25, give the results of eleven Runs (1G, 2, 2A, 2D, 2C, 2D, 7, 7tf, 7C and 7D) performed to study the time required to change the concentration of the material i n the probes using various rates of purging. It can be remarked that in these Tables (14 to 25), for some Runs a complete analysis of the ketone sample i s not given; this was due to the presence of only a small volume of water i n the ketone sampleo The smallness of this volume was caused by the low purging rate which col l e c t e d 90 to 95% by volume of ketone phase i n the dispersed phase sample. The eff e c t of such a small volume of water on the resul t s would be less than 2% at the worst. A summarizing Table (number 24) i s given at the end of this Appendix. The purge time was found to f a l l as the purging rate increased as would be expected. TABLE 14. Study of minimum purge time. Run 1G. Note: Probe f i l l e d with l i q u i d at sample position 4A and moved at zero time to sample position 7. Average continuous phase probe sampling rate 4.9 cc/min. Average dispersed phase probe sampling rate 13.7 cc./min. Time, Volume of phases Concentrations, min. (a) in ketone sample lb.-moles/ft? x 10 3. cc. From to V C , C, _ C . w k wf kf k i wi 0 1 2„2 11.3 27.9 16.2 15.2 32.8 1 2 2.2 12.0 27.5 15.7 14.7 33.1 2 3 2.3 11,8 33.1 18.1 18.1 -3 4 1.9 11.6 34.2 19.0 18.5 37.2 4 5 2.1 11.6 34.8 19.0 18.5 37.6 5 6 2.4 11.4 34.1 19.1 18.3 37.8 6 7 2.1 11.8 34.5 19.0 lcs.5 — TABLE 15. Study of the minimum purge time. Run 2 Note: Probe f i l l e d with l i q u i d at sample position 7 and moved at zero time to sample position 4A. Average continuous phase probe sampling rate: 13.8 cc./min. Average dispersed phuise probe sampling rate: 15.7 cc./min. Time, Volume of phases Concentrations, min. (a) i n ketone sample lb.-moles/ft 3 x 10 3. cc, From to V, C _ C, „ C . fe k wf kf k i wi 0 1 2.4 13.3 33.1 18.0 1C.6 36.4 1 2 2.4 13.3 32.3 17.5 16.7 36.2 2 3 2.4 13.2 27.3 14.6 13.5 33.4 3 4 2„6 13.3 27.2 14.5 13.5 32.1 4 5 2.4 13.6 27.0 14.4 13.6 31.7 5 6 2.5 13.4 27.2 14.5 13.7 31.7 6 7 2.6 13.8 26.4 14.5 13.5 31.7 7 lit 7.4 38.0 27.6 14.4 13.6 31.8 (a) one minute sample. TABLE 15 Cont. Study of the minimum purge time. Run 2 Note: Probes f i l l e d with l i q u i d at sample position 4A and moved at zero time to sample point 2A. Averaged continuous phase sampling rate: 10.7 cc./min. Averaged dispersed phase sampling rate: 14 05 cc./min. Time, Volume of phases Concentrations, min. (a) i n ketone sample lb.-moles/ft? x 10 3. cc. From to V w \\ Cwf C k f C k i C . wi 0 1 2.4 12.0 26.9 14.6 13.6 32.0 1 2 2.4 12.6 27.6 14.7 13.5 31.9 2 3 2.3 13.0 19.5 10.4 8.6 29.9 3 4 2.4 12.4 19.7 1011 80S 26.a 4 5 2.4 12.8 20.2 10.1 9.1 25.8 5 6 2.0 12.0 20.5 10.2 9.4 25.6 6 7 2.5 19.6 10.3 9.1 25.6 7 i a 11.0 62.5 20.2 10.1 9.2 25.6 TABLE 16. Study of the minimum purge time. Run 2A. Note: Probes f i l l e d with l i q u i d at position 2A and moved at zero time to position 4A. Averaged continuous phase probe sampling rate: 3.8 cc./min. Averaged dispersed phase probe sampling rate: 8.7 cc./min. Time, Volume of phases Concentrations, min. (a) in ketone cc sample • lb.-moles/ft? x 10°. •From to V V, C „ C, B, c , • C . w k b wf kfi. k i wi 0 1 1.1 8.2 .11. Z.:. 10-45 10.45 25,32 1 2 1.2 8.0 10.13 10.13 25 o*0 2 3 1.0 8.3 10.39 10.39 25.50 3 4 1.1 8.0 13.62 13.62 25.52 4 5 1.1 6.1 14.35 14.35 25.70 5 6 1.0 7.3 14.46 14 .,48 26.50 6 7 1.0 7.3 14.10 14 .10 26.25 7 8 0.9 7.8 14.55 14.55 29.68 8 9 0.9 7.4 14. 55 14.55 29.72 9 10 0.8 7.6 14.61 14.61 31.10 10 11 0.8 7.5 14.61 14.61 31.00 11 12 0.9 7.4 14.54 14 „ 54 31.10 12 13 0.9 7.5 14.54 14.54 30.85 TABLE 1 6 Cont. RUN 2 A cont. Note: Probes f i l l e d with l i q u i d at sampling position 4 A and moved at zero time to sampling position 7 . Averaged continuous phase probe sampling rate: 1 0 . 5 cc./min. Averaged dispersed phase probe sampling rate: 1 1 . 7 c c o/min. Time, Volume of phases Concentrations, min. (a) in ketone sample l b.-moles/ft V x 1 0 ° . From to V w • V k ( b ) C w f °kf C k i C . Wl 0 1 1 . 8 9 . 6 1 5 c 4 0 1 5 . 4 0 3 2 . 4 1 2 1 . 8 1 0 . 2 1 5 . 5 5 1 5 . 5 5 3 2 . 3 2 3 1 . 6 1 0 . 4 1 6 . 1 8 1 6 . 1 8 3 3 . 6 3 4 l o 4 1 0 . 1 1 8 . 2 0 1 8 . 2 0 3 4 . 4 4 5 1 . 6 1 0 o O l b . 2 0 1 8 . 2 0 3 6 . 7 5 6 1 . 4 1 0 o l 1 8 c 2 0 1 8 . 2 0 3 6 o 7 6 7 1 . 4 1 0 . 6 1 8 . 3 2 1 8 0 3 2 3 6 . 8 7 8 1 . 7 1 0 . 1 1 8 . 1 5 1 8 . 1 5 3 6 . 6 8 9 1 . 5 9 c 8 1 8 . 8 3 1 8 . 8 3 3 6 . 4 (a) one minute sample (b) Not enough water i n the ketone sample for analysis. TABLE 1 7 . RUN 2 B . Study of the minimum purge time. Note: Probes f i l l e d with l i q u i d at sampling position 4 A and moved at zero time to sampling position 7 . Averaged continuous phase probe sampling rate: 6 . 5 cc./min. Averaged dispersed phase probe sampling rate: 7 . 0 cc./min. Time, Volume of phases Concentrations, min. (a) in ketone cc. sample lb.-nioles/ftV x 106' From to V w \\ (b)°wf C k f C k i C . Wl 0 1 0 . 7 5 c 2 1 4 . 7 1 4 . 7 3 1 c 8 1 1 . 0 9 . 0 1 4 . 4 1 4 . 4 3 1 . 4 2 3 1 . 2 7 . 9 1 4 . 6 1 4 . 6 3 1 . 9 3 4 0 . 7 7 . 6 } 5 . 2 1 5 . 2 3 3 c 9 4 5 0 c 4 4 . 6 1 8 . 0 1 8 . 0 3 5 . 9 5 6 1 . 0 8 . 7 1 8 . 5 1 8 . 5 3 6 . 5 6 7 0 , 5 5 . 1 1 8 . 3 1 8 . 3 3 7 . 2 7 8 0 . 5 3 . 5 1 8 . 3 1 8 . 3 3 7 . 2 8 9 0 . 3 3 o 5 1 9 . 0 1 9 . 0 3 7 . 3 (a) One minute sample. ( b ) Not enough water in the ketone sample for analysis. TABLE 1 8 . Study of the minimum purge time. RUN 2C Note: Probes f i l l e d with l i q u i d at sampling position 4A and moved at zero time to sampling position 2A. Averaged continuous phase probe sampling rate: 7 . 7 cc./min. Averaged dispersed phase probe sampling rate: 1 1 . 9 7 cc./min. Time, Volume of phases Concentrations, min. (a) in ketone cc sample • lb.-moles/ft? x 1 0 ° . From to V V, C „ C, , c, . C . w k , . wf kf ( b ) - 1 4 , 2 k i wi 0 1 1 . 3 ' 1 0 . 2 1 4 . 2 3 2 . 4 1 2 1 . 3 1 1 . 2 1 3 . 6 1 3 o 6 3 1 o 6 2 3 1 . 0 1 1 . 1 1 2 c 4 1 2 . 4 3 1 . 0 3 4 1 . 2 1 1 . 2 9 . 5 9 . 5 2 9 . 1 4 5 1 . 2 1 1 . 2 9 . 5 9 . 5 2 6 . 6 5 6 1 . 4 1 1 . 4 9 . 6 9 1 6 2 5 . 3 6 7 1 . 2 1 0 . 2 9 . 6 9 . 6 2 4 . 6 7 8 1 . 2 9 . 5 9 . 7 9 . 7 2 4 . 4 (a) One minute sample. (b) Not enough water in the ketone sample for analysis. Note: Probes f i l l e d with l i q u i d at sample positiinaSSdand moved at zero time to sampling position 4A. Averaged continuous phase probe sampling rate: 1 . 6 cc./min. Averaged dispersed phtise probe sampling rate: 5 . 6 cc./min. Time, Volume of phases Concentrations, min. (a) in ketone sample lb.-moles/ft\"^ x 1 0 ° . cc. From to V V, C „ C, , c, • C . w k wf kf k i Wl 0 1 0 . 1 3 , 9 1 8 . 7 1 8 . 7 3 5 . 9 1 2 0 . 0 3 . 8 1 8 . 9 1 8 . 9 3 6 . 3 2 3 0 . 0 7 . 0 J 8 . 5 1 8 . 5 3 4 . 7 3 4 0 . 0 6 . 2 1 8 . 6 1 8 . 6 3 6 . 9 4 5 0 . 0 4 . 4 1 9 . 0 1 9 . 0 3 4 . 9 5 6 0 . 3 5 . 8 1 7 c 6 1 7 . 6 3 6 . 0 6 7 0 . 1 6 . 2 1 5 . 6 1 5 . 6 3 6 . 6 7 8 0 . 5 6 . 0 1 4 . 8 1 4 . 8 3 6 . 4 8 9 0 . 2 5 . 6 1 4 . 9 1 4 . 9 3 6 . 3 9 1 0 0 . 2 6 . 6 1 4 . 9 1 4 . 9 3 6 . 0 (a) One minute sample. (b) Not enough water i n the ketone sample for analysis. TABLE 18 Cont. RUN 2C cont. Note: Probes f i l l e d with l i q u i d at sampling position 2A and moved at zero time to sample position 4A. Averaged continuous phase sampling rate: 9.0 cc./min. Averaged dispersed phase sampling rate: 11.5 cc./min. Time, Volume of phases Concentrations, min. (a) in ketone cc. sample lb.-moles/ftV x 10°. From to V w Vk (b) Cwf °kf C k i C . wi 0 1 1.2 10.4 10.8 10.8 27.3 1 2 1.2 10.4 11.0 11.0 27.4 2 3 1.2 10.4 11.8 11.8 28.2 3 4 0.0 3.8 15.8 15.8 30.7 4 5 0.8 7.2 15.5 15.5 32.9 5 6 1.6 11.6 15.8 15.8 33.9 (a) One minute sample. (b) Not enough water i n the ketone sample for analysis. i TABLE 19. Study of the minimum purge time. RUN 2D Note: Probes f i l l e d with l i q u i d at sampling position 4A and moved at zero time to sampling position 2A. Averaged continuous phase sampling rate: 10.4 cc./min. Averaged dispersed phase sampling rate: 12.4 cc./min. Time, Volume of phases Concentrations min. (a) in ketone sample lb.-moles/ft? x 10 3. From to V w '' Vk (b?wf °kf C k i C . wi 0 1 1.4 11.6 13.6 13.6 -1 2 1.1 9.9 13.8 13.8 29.9 2 3 1.4 10.6 12.7 12.7 29.1 3 4 1.4 10.4 9.3 9.3 26.7 4 5 1.6 11.2 9.2 9.2 24.9 5 6 1.5 11.3 9.2 9.2 24.0 6 7 1.6 11.7 9.5 9.5 24.3 7 8 1.6 11.6 9.4 9.4 24.3 (a) One minute sample. (b) Not enough water i n the ketone sample for analysis. TABLE 19 Cont. RUN 2D cont. Note: Probes f i l l e d with l i q u i d at sampling position 2A and moved at zero time to sample position 4A. Averaged continuous phase sampling rate: 1.9 cc./min. Averaged dispersed phase sampling rate: 6.5 cc./min. Time, Volume of phases Concentrations, min. (a) i n ketone sample lb.-moles/ft? x 10 3. cc. From to V w \\ (b) Cwf C k f C k i C . Wl 0 1 1 . 0 5 . 2 - 1 0 o 9 1 0 . 9 2 2 . 2 1 2 1 . 2 5 . 2 - 1 0 . 9 1 0 . 9 2 4 . 4 2 3 1 . 0 5 . 4 - 1 0 . 8 1 0 . 8 2 3 . 7 3 4 1 . 0 5 . 0 - 1 0 . 8 1 0 . 8 2 0 . 2 4 5 1 . 1 4 . 8 - 1 0 . 1 1 0 . 1 2 4 . 2 5 6 0 . 8 6 . 6 - 1 2 . 3 1 2 . 3 2 6 . 3 6 7 0 . 4 6 . 9 - 1 4 . 0 1 4 . 0 2 4 . 0 7 8 0 . 6 6 . 4 - 1 4 . 0 1 4 . 0 2 4 . 5 8 9 0 . 5 6 o 3 - 1 3 . 9 1 3 . 9 2 4 . 1 9 1 0 0 . 0 6 . 8 - 1 4 . 0 1 4 . 0 2 4 . 3 1 0 1 1 0 . 8 6 . 0 1 4 . 0 1 4 . 0 2 5 . 1 (a) One minute sample. (b) Not enough water in the ketone sample for analysis. TABLE 20 Study of the minimum purge time. RUN 7. Note: Probes f i l l e d with l i q u i d at sampling position 4 and moved at zero time to sampl*np 0pnsition 4A. Averaged continuous phase probe sampling rate: 2.8 cc./min. Averaged dispersed phase probe sampling rate: 5.4 cc./min. Time, Volume of phases Concentrations, min. (a) i n ketone sample lb.-moles/ft? x 10 3. cc. . From to V V, ..^C _ C C. . C . w k (b) wf k f k i wi 0 3 0.8 16.2 - 11.7 11.7 29.8 3 6 0.8 16.4 - 12.2 12.2 29.6 6 9 0.8 15.1 - 17.7 17.7 29.7 9 12 0.8 14.8 - 17.7 17.7 32.5 12 15 0.8 15.0 - 17.7 17.7 36.4 15 18 2.0 24.0 - 17.4 17.4 38.9 18 21 2.0 26.0 - 17.4 17.4 38.7 (a) Three minutes sample. (b) Not enough water i n the ketone sample for analysis. 132 TABLE 21. Study of theminimum purge time. RUN 7B. Note: Probes f i l l e d with l i q u i d at sampling position 1A and moved at zero time to sampling position 4A. Averaged continuous phase probe sampling rate: 2.8 6c./min. Averaged dispersed phase probe sampling rate: 2.1 cc^/min. Time, Volume of phases Concentrations, min. (a) i n ketone sample lb.-moles/ft? x 10 3. cc. From to V v, (b>Cwf C k f C . w k k i wi 0 3 2.6 3.1 12.4 12.4 30.8 3 6 0.4 6.4 12.4 12.4 30.5 6 9 0.0 7.3 11.8 11.8 30.9 9 12 0.0 6.8 11.5 11.5 32.6 12 15 0.3' 5.7 12.4 12.4 38.1 15 itz 0.0 4.2 17.0 17.0 38.8 17 19 0.0 4.0 18.0 18.0 39.3 19 21 0.0 4.0 18.0 18.0 39.3 21 23 0.0 3.8 18.0 18.0 39.4 (a) Three or two minutes samples. (b) Not enough water i n the ketone sample for analysis. TABLE 22. Study of the minimum purge time. RUN 7C. Note: Probes f i l l e d with l i q u i d at sampling position 4A and moved at zero time to sampling position 1A. Averaged continuous phase probe sampling rate: 5.0 cc./min. Averaged dispersed phase probe sampling rate: 4.8 cc./min. Time, Volume of phases Concentrati ons, min. (a) in ketone sample lb.--moles/ft. x 10 3. From to cc V w • Vk Cwf C k f C k i C . wi 0 3 0.6 8.2 - 18.0 18.0 39.6 3 6 0.8 10.8 - 18.2 18.2 38.6 6 9 0.6 13.2 - 17.5 17.5 33.0 9 11 0.3 8.4 - 16.0 16.0 30.0 11 13 0.4 8.8 - 11.7 11.7 30.6 13 15 0.3 7.9 - 11.8 11.8 30.6 15 17 0.6 8.2 - 11.8 11.8 30 o 6 17 19 0.3 8.0 - 11.8 11.8 30.8 (a) Three or two minutes sample. (b) Not enough water in the ketone sample for analysis. 133 TABLE 22 Cont. RUN 7C cont. Note: Probes f i l l e d with l i q u i d at sampling position 1A and moved at zero time to sampling position 4A«. Time, min. (a) Volume of phases in ketone sample cc. Concentrations, lb.-moles/ft 3 x 10 3. From to V V, C . w k wf kf k i Wl 0 3 0.6 12.8 11.9 11.9 30.7 3 6 0.6 13.5 11.7 11.7 33.8 6 10 0.8 16.0 15.7 15.7 39.0 10 12 0.6 7.4 18.0 18.0 39.0 12 14 0.6 9.6 17.7 17.7 39.7 14 16 0.5 9.7 18.0 18.0 39.7 16 18 1.0 20.2 18.0 18.0 39.7 (a) Four, three and two minutes samples. (b) not enough water i n the ketone sample for analysis. TABLE 23. Study of the minimum purge time. RUN 7D. Note: Probes f i l l e d with l i q u i d at sample position 1A and moved at zero time to sample position 4A. Averaged continuous phase probe sampling rate: 3.7 cc./min. Averaged dispersed phase probe sampling rate: 2.5 cc./min. Time, min. (a) From 0 3 6 9 11 13 15 17 ' 19 to 3 6 9 11 13 15 17 19 21 Volume of phases in ketone sample cc. V w 0.6 0.7 0.0 0.0 010 0.0 0.0 0.0 0.0 k 6.6 7.8 4.8 4.8 4.9 5.2 4.5 5.0 4.9 Concentrations, lb.-moles/ft 3x 10 3. wf kf 12.1 12.2 12.1 12.8 12.8 17.1 18.8 18.8 18.8 k i 12.1 12.2 12.1 12.8 12.8 17.1 18.8 18.8 18.8 C . wi 31.2 31.5 34.1 38.7 40.1 40.4 40.7 40.7 40.7 (a) Three or two minutes sample (b) Not enough water in the ketone sample for analysis. 1 3 4 TABLE 2 4 . Summary of a l l results to deternime the minimum purge time. (Smoothed values). Run number. Continuous phase. Minimum Sampling rate, Dispersed phase. Sampling Minimum purge time, rate, purge time cc./min min. cc./min min 1 G 1 2 . 8 5 . 0 1 3 . 8 3 . 5 2 1 1 . 0 6 . 0 1 5 . 9 3 . 0 2 1 3 . 0 5 . 0 1 4 . 9 3 . 5 2 A 3 . 8 _ * 8 . 7 5 . 5 2 A 1 0 . 5 6 . 0 1 1 . 7 4 . 0 2 B 8 . 5 7 . 0 8 . 6 5 . 5 2 C 7 . 7 1 2 . 0 4 . 0 2 C 9 . 0 5 . 0 1 1 . 5 4 « 0 2 C 1 . 6 _ * 5 . 6 8 . 0 2 D 1 . 0 _ * 6 . 5 7 . 0 2 D 1 0 . 4 7 . 0 1 2 . 4 5 . 0 7 2 . 8 2 0 . 0 5 . 4 9 . 0 7 B 2 . 8 1 9 . 0 2 . 1 1 8 a 3 7 C 5 . 0 1 2 . 0 4 . 1 1 3 . 0 7 C 6 . 3 1 0 . 0 4 . 5 1 2 . 0 7 D 3 . 7 1 6 . 0 2 . 5 1 6 . 8 * The minimum purge time was not obtained i n these runs^ the time was too short for the concentration curve to f l a t t e n out. 135 APPENDIX V. Point concentration versus sampling rate. The following Table, 25, represents thfe res u l t s of the runs done to demonstrate that the sampling rate did not influence the point concentration. TABLE 25. Point Concentration Versus Sampling Kate. Run number. 9H 7F 7E 5H Purge time, min. Sampling time, min. Minimum Purge Time according to Figure 10. min. Water Ketone phase. phase. Rate of Sampling, cc./min. Water phase. Ketone phase. Volume of Concentrations^ phases i n lb.-moles/ftVxlO J, ketone sample. 66. V, C w wf k f k i Probe location. wi 10 5 6.0 5.6 10.4 8.4 3.0 39.0 14.6 7^ 3 7„2 15.8 1.59 10 5 4.0 3.0 16.0 15.6 7.0 69.0 15.0 7.3 7.2 16.2 1.59 10 5 10.0 5.6 6.2 8.2 3.0 38.0 13.3 7.2 7. 1 15.7 1.59 10 2 6.4 4.8 9.4 10.3 1.5 19.0 al5.0 7.2 7. 1 16.2 1.59 0 10 18.6 25.0 2.8 1.1 0.6 10.4 28.4 19. 1 18.9 41.4 4A 0 10 18.6 18.6 2.8 2.0 1.8 17.8 28.0 18.8 f a . 4 41.2 4A 0 10 18.6 9.4 2.8 5.2 2.0 50.0 27.7 18, 4 -18. 3 41.3 4A 0 414 18.6 1.0 2.8 28.2 24.0 103.0 28.6 18, 4 |8.5 41.3 4A 0 8 18.6 3.0 2.8 16.2 19.0 111.0 28.5 19<.: Ik 48.4 41.4 4A 0 10 18.6 6.8 2.9 7.0 14.0 56.0 29.2 19.6 #16.6 41.2 4A 0 10 9.8 7.4 6.4 6.2 4.0 58.0 27.7 18.6 17.7 41.2 4A 0 10 5.2 8.6 11.3 5.7 3.0 54.0 27.7 18.6 17.8 41.3 4A 0 10 7.6 9.2 8.1 5.4 3.0 50.5 27.7 18.6 17.7 41.2 4A 0 10 4.3 5.0 14.2 9.8 : 6.0 82.0 27.3 18.3 17.3 41.2 4 A 15 5 6.8 3.6 9.0 13.8 5.0 64.0 26.7 13.5 ^2.9 33.3 3B 5 5 13.8 4.8 4.5 10.2 4.0 47.0 27.0 13.8 13.2 34.8 3B 5 5 * 6.6 7.2 2.5 33.3 27.0 14.0 13.5 - 3B 8 6 * 9.4 - 5.3 0.0 26.5 - 14.0 14.0 - 3B 7 3 f0.6 3.4 6.0 14.5 5.0 38.5 28.2 14.4 13.7 35.0 3B 10 3 3.0 - 15.5 5.5 41.0 28.6 14.5 13.9 - 3B 5 3 * 2.0 — 18.7 7.0 48.2 28.6 14.5 13.9 - 3B OJ TABLE 25 Cont. Run Purge number. time, 5G 5F 5B, ID Sampling time, Minimum Purge Time according Rate of Sampling, cc./min. Volume of phases i n Concentrations, lb.-moles/f&10 3. Probe location. min. min. to Figure 10. ketone sampl e. min. Water Ketone Water Ketone V,_ V, C . C . phase. phase. k k IV f kf k i Wl phase. phase 18 3 6.6 3.6 9.3 13.7 4.0 37.0 27.3 14.1 13.2 35.4 3B 7 3 * - 4.1 . - 12.3 3.8 33.2 27.1 13.8 12.8 - 3B 8 3 * 5.6 • — 8.3 2.0 23.0 25.5 13.9 13.0 - 3B 9 3 * 5.6 — 8.3 2.0 23.0 23.4 13.7 12.7 - 3B 5 3 * 3.2 — 15.0 5.8 39.2 25.8 14.1 12.6 - 3B 5 3 » 2.6 — 17.0 7.0 44.0 28.0 14.3 13.1 . - 3B 6 2 6.4 0.5 - 20.3 5.0 35.5 27.3 14.7 13.6 35.7 3B 9 3 7.0 3.6 8.8 13.7 4.0 37.0 26.1 14.0 ^2.9 35.5 3B 0 3 13.2 4.6 4.7 10.7 3.0 29.0 27.6 14.0 13.2 35.6 3B 0 3 * 4.0 - 12.7 3.0 35.0 26.7 13.9 13.1 - 3B 0 9% * 3.2 - 15.0 14^0 125.0 28.4 14.1 13.4 - 3B 0 6 • 2.6 - • 17.0 9.0 109.8 28.1 14.2 13.5 - 3B 0 7. * 2.3 - 17.6 17.0 105.2 29.0 14.3 13.2 - 3B 9 3 6.6 3.7 9.3 13.6 5.5 36.3 24.6 12.6 ?2.5 32.1 3B 0 5 13.4 4.8 4.6 18.4 6.0 46.0 24.6 12.5 12.4 32.2 3B 0 5 * 2.7 - 16.6 11.3 71.0 25.9 12.8 12.7 - 3b 0 5 7.0 8.9 - - - - - - 32.0 3B 0 5 * 0.5 - 23.3 20.0 96.6 27.4 13.3 12.4 - 3B 0 5 6.6 * 9.2 - - - - - • - 3a.7 3B 4 5 0.5 1.0 34.0 28.4 27.7 104,3 45.7 23.8 23.1 47.7 7 0 5 1.0 1.0 21.0 21.0 25.0 80.0 45.6 23.8 23.1 47.8 7 0 5 5.2 3.8 11.0 12.9 10.2 54.0 44.2 23.5 22.6 47.9 7 0 5 4.0 2.6 16.0 17.0 12.2 72.6 45.1 23.5 23.1 47.9 7 a = Equilibrium value, not analyzed. * = Not sampled. + = Average water concentration was used to calculate these values. # = F i r s t concentration of water used to do the c a l c u l a t i o n s . CJ •si 138 APPENDIX VI Jet c h a r a c t e r i s t i c data. (Jonhson and B l i s s ) . Reference to Rocchini's thesis (13) page 63, showed that he used a v e l o c i t y through the nozzle of 0.3623 ft. / s e c . However, Choudhury (10) used a v e l o c i t y of 0.357 f t . / s e c . as reported i n his thesis, page 26. It was discovered that this difference was caused by the use of s l i g h t l y d i f f e r e n t flow rates of dispersed phase. After recalculating the v e l o c i t y of these previous workers (10,13) and checking the above values to close approximation, i t was decided to use a value of 0.357 f t . / s e c . based on the average diameter of the nozzle t i p s presently available. This i s 0.1029.-in. (10). It became necessary to check i f this v e l o c i t y through the nozzle would be expected to give uniform drops. Referring to Jonhson and B l i s s (16), i t was discovered that the v e l o c i t y used in t h i s work was high enough so that the drops would have ceased forming at the nozzle ti p s but not so high that the drops would have ceased being uniform. In other words, the v e l o c i t y of 0.357 ft . / s e c . u t i l i z e d would give uniform drops. Although a l i n e a r interpolation of Table II of the paper by Jonhson and B l i s s (16) indicates that drops should, for a nozzle of 0.103.-in. I.D., cease being uniform when the v e l o c i t y i n nozzle ti p s reached 0.350 f t . / s e c , according to their curve in t h e i r Figure 3, the v e l o c i t y plotted at 0.11-in. I.D. appears to be too low. (This plotted v e l o c i t y i s the one from the Table.) In fact, at 0.111-in. I.D. nozzle diameter they draw t h e i r curve considerably above the plotted point. It i s on the basis of t h i s fact that the statement i s made that the t i p v e l o c i t y of 0.357 ft . / s e c . should produce uniform drops + Not always used. * It was observed that the drops formed straight (not sinuous) jets except for one t i p ; at'' this''tip the jet was sinuous. A l l jets were app. Vfe-in. in length. 139 according to Jonhson and B l i s s ( l 6 ) . APPENDIX VII Sample cal c u l a t i o n of the material balances used in connection with Run 9H water Sample no. 2 which contained ketone. As a result of a high flow rate of dispersed phase and perhaps of too high a sampling rate for such a flow, the following calculations have to be made to correct for the presence of ketone i n the water probe sample. Equation 5 i s applied f i r s t to the sample obtained with the water probe: (C, .) = (6, J . - (V /V, ) . (C . - (C .) ' ) k i kf water w k water wi wf water probe probe probe The same equation i s applied also to the ketone probe sample: (C. .) = (C, _). . - (V /V, ), . (C . - (C ,). . ) k i kf ketone w k ketone wi wf ketone probe probe probe Where the volume of one phase in the sample i s small, as compared with the volume of the other phase, an equilibrium value i s taken instead of aking use of analysis. In the two equations only two unknowns, C m and C ., are l e f t . These, of course, can be solved for. wi In Run 9H, Sample number 2, the water probe sample contained 2.6 cc. of ketone and 10.2 cc. of water. From the values l i s t e d in Table 5 the following are obtained af t e r substitution: (C, .) = 7.62 - 10.2/2.6 (C . - 15.88) k i wi and (C, .) 7.36 - 0.4/15.6 (e . - 14.95) k i wi Solving for C ... i t was found that: ° wi C . = 15.95 lb.-moles/ft 3x 10 3 wi Using this i n i t i a l value for the water probe gives a values of 7.33 for for the probe dispersed phase. Using of 15.95 i n the material balance for the piston sample produces a C ^ of 7.25. 141 APPENDIX VIII Possible sources of errors causing discrepancies between piston and probe dispersed phase results. Problems a r i s i n g in the analysis of the solutions i n the piston samples were suspected by Hawrelak (12) as being the sources of much error i n the values of (C, .) . , . In Equation 5, the factor k i piston (C^^ - S f ^ a P P e a r s i n the second term to the right of the equal sign a The concentrations involved are of about the same order of magnitude as mentioned e a r l i e r , and substracting them tends to make a n a l y t i c a l errors important. An example would i l l u s t r a t e Better the d i f f i c u l t y just mentioned. In Sample no. 2 of Run 9G done with the piston, the phase concentrations at time of analysis were (C _,) . , , and (C. _) . . , wf piston; kf piston* 3 3 3 3 20„35 lb.-moles/ft. x 10 and 10.15 lb.-moles/ft.x 10 respectively. This sample contained 101 cc. of continuous phase and 16 cc. of dispersed phase. The analysis of the continuous phase taken with the 3 3 probe, or (C .) . , was 20.96 lb.-moles/ft. x 10 . Now, substitution 1 ' wi probe in Equation 5: (c, .) . . = « ; . _ ) • . . - v /v, ((c .) . - (c _) . , ) k i piston kf piston w k wi probe wf piston produces, (C ) k i piston = 10016 - 101/16(20.96 - 20.35) (C ) 3 3 k i piston = 6.30 lb.-moles/ft. x 10 . This concentration has to be compared with the one obtained 3 3 by the probe method which was 8.30 lb.-moles/ft. x 10 , for which the volumes were 4 cc. of water and 48.5 cc. of ketone. (C, .) . = 8.58 - 4/48.5(20.96 - 17.52) k i probe 142 It seems l o g i c a l now to consider i f i t i s possible that the dispersed phase concentrations obtained by the two methods do not check because of possible manipulative errors in analysis. For example, what can bo the effect of a pipetting error on the f i n a l results? Based on the calibrations of the 10-ml. pipettes, a possible error of - 0.4% in the volume delivered can be expected. IVhat can be the e f f e c t of errors i n measuring the volumes of the phases in a piston sample? At the worst, - 0.5 ceo of water can be taken as possible errors i n V ^ , and corresponding to these errors, - 0 05 cc. i n V ^ . F i n a l l y , what can be the effect of adding one drop past the end point in the t i t r a t i o n s of the probe and of the piston samples? Calculations have been done to reveal the effects of the possible errors mentioned above. F i r s t of a l l , to be able to r e a l l y see t h e i r influence on some of the analysis values, the e f f e c t of each error has been considered separately; the results are presented in four cases: F i r s t case: suppose that i n pipetting 10 ml. of water from the continuous phase of the probe sample, 10.04 ml. a c t u a l l y was delivered, and, s i m i l a r l y , 9.96 ml from the water phase of the piston sample. A l l other quantities measured in the analysis were assumed to have the values used to calculate the r e s u l t s of Sample 2 of Run 9G given e a r l i e r i n this Appendix. Replacing i n Equation 5 for the piston sample: C, . = 10.16 - 101/16 (20.88 -20.43) k i 3 3 A value of 7.32 lb.-moles/ft. x 10 i s found for (C, .) . . as compared k i piston with 6.30 as given e a r l i e r . (The difference i s 16.4% of 6.30.) Replacing i n Equation 5 for the probe sample: C, . = 8.58 - 4/48.5 (20.88 - 17.52) k i A value of 8.30 lb.-moles/ft? x 10 3 i s found for (C, .) , k i probe as compared 143 with 8.30 calculated e a r l i e r . Second case: suppose that in pipetting 10 ml. of water from the continous phase of the probe sar.ple, 9.96 ml. actually was delivered, and, s i m i l a r l y , 10.04 ml. from the water phase of the piston sample. As before a l l other quantities measured in the analysis are taken to be unchanged. In Equation 5, for the piston sample, then: C, . = 10.16 S 101/16 (21.04 - 20.27) k i 3 3 Then a value of 5.30 lb.-moles/ft.x 10 i s found for (C, . ) . . as k i piston compared to 6.30. (The difference i s 15.9% of 6.30J Equation 5 for the probe sample then i s C, . = 8.58 - 4/48.5 (21.04 - 17.52) k i 3 3 A value of 8.29 lb.-moles/ft. x 10 i s found for (C, .) , k i probe as compared with 8.30 calculated e a r l i e r . Third case: suppose that the volume of water i n the piston sample was actually 100.5 cc. instead of 101.0 as measured and suppose that the corresponding volume of ketone was r e a l l y 16.5 cc. instead of 16.0 c c . A l l other quantities measured i n the analysis were as used in the calc u l a t i o n given at the beginning of t h i s Appendix. Equation 5 for the piston sample becomes for these conditions: C, . = 10.16 - 100.5/16.5 (20.96 - 20.35) k i The r e s u l t i n g (C , . ) . . would have been 6.44 lb.-moles/ft? x 10^ ° k i piston instead of 6.30 the o r i g i n a l value. (The difference i s 2.2% of C.30.) Fourth case: suppose that the t i t r a t i o n s of each phase of the piston sample, one drop was added past the end point i n each analysis. (In the t i t r a t i o n , i t would be much easier to overun the end point then to add too l i t t l e 144 sodium hydroxide.) Once again, a l l the other measurements involved i n the basic calculations are assumed correct. In Equation 5, then: -6 = 10.13 - 101/16(20.96 - 20.32) 3 3 Then the re s u l t i n g (C, ,) . , would have been 6.09 lb.-moles/ft. xlO 0 k i piston as compared with 6.30 obtained o r i g i n a l l y . (The difference i s 3.3% of 6.30.) A c a l c u l a t i o n was made of the effect of several errors of the sort discussed i n cases 1 to 4 taking place at the same time i n connection with one piston sample and the probe sample corresponding to i t . ' An error of 21.3% over the o r i g i n a l value of 6.30 lb.-moles 3 3 / f t . xlO for (C. .) . . i s calculated i f due to pipetting error, the k i piston 1 * & * volume analyzed for the probe continuous phase had been 10.04 ml. instead of the 10 ml. assumed and, i f both phases of the piston sample analyzed had been 9.96 ml. instead of the 10 ml. assumed and, i f the measured volume of the phases of the piston sample had been measured too low by -0.5 cc. and too high by +0.5 cc. for the water and ketone phases respectively, and f i n a l l y i f 1 drop of 0.1 N NaOH had been added past the end point i n the t i t r a t i o n of the water of the continuous phase sample. Equation 5 for t h i s case i s (for the piston sample): C k i = 1 0 o 2 ° \" 100.5/16.5(20.85 - 20.43) For the probe sample an error of 0.4% over the o r i g i n a l of 3 3 8.30 lb.-moles/ft. xlO i s cumulated i n this way: a) i f the volume analyzed had been 10.04 ml. instead of the 10 ml. assumed for the probe continuous phase sample. b) i f the volume analyzed had been 9.96 ml. instead of the 10 ml. assumed for the ketone of the probe dispersed phase sample. 145 c) i f the volume of water of the continuous phase probe had been t i t r a t e d one drop past the end point. Equation 5 for th i s case i s : C = 8„61 - 4 / 4 8 . 5 ( 2 0 o d 5 - 17 = ,52 ) I f one can eliminate these possible errors by analyzing bigger volumes and by using the highest accuracy to measure the volumes of the phases of a piston sample, then i t would appear that analysis d i f f i c u l t i e s should no longer be a problem,, o "@en ; edm:hasType "Thesis/Dissertation"@en ; edm:isShownAt "10.14288/1.0059160"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Chemical and Biological Engineering"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Sampling of the phases within a liquid-liquid extraction spray column"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/39716"@en .