$12 MoCabe-Thicle Graphical Method for Trayed Towers 269 EXAMPLE 7.1 Distillation of a Mixture of Benzene and Toluene. Four hundred and fifty tools (204 kml) of a mixture of 60 mol benzene (LK) and 40 mol% toluene (HIK) isto be separated into a quid distillate anda liquid bottoms product of 95 mol" and 5 mol%e benzene, respectively. The fed enters the column with a mola per- cent vaporization equal to the distillate-to-eed ratio. Use the McCabe-Thiele method to compute, at 1 atm (101.3 KPa) 8) Nain (©) Rau, and (6) numberof equilibrium stages N, for R/R. and the optimal fed-stage location, Also, compare the results with tote from a process simlator, Solution First calculate D and 8, An overall material balance on benzene ives 0.60(450) = 0 95D 4 0.058 w Atotalbalance gives 450=D +B @) Combining (1) and (2) and solving, D = 275 Ibmothh, B = 175, Ibmoly and D/F 0.611. Thus, the molar vaporization ofthe feed gine Le_(P=Vr)_ From (7-26), 0.389 “oa the slope ofthe qline is = 0.687 ai (a) In Figure 7.13, where y and x refer to benzene, xp = 0.95 and Xp = 0005, the minimum stages are stepped olf between the quilibrium curve and the 45° line, giving Naa = 6.7 © In Figure 7.14, « gine is drawn that as a slope of ~0.637 and passes through the [eed composition (z- ~ 0,60) on the 45° line. For Rin an operating line for the rectifying section passes through the poiat x = xp = 0.95 on the 45° line and through the point of intersection ofthe g-line and the equilibrium curve (= 0.684, x = 0.465). The slope of this operating lin is 0.55, which from (7-9) equals R/(R-+ 1). Therefore, Rain = 1.22. The operating refx ratio i 1.3 Rous = 1.3(1.22) = 1.59, From (7-9), the steipping-section operating-ine slope is R159 Rel isost © ois The two operating lines and the g-line are shown in Figure 7.15, where the stripping-section operating line is drawn to piss through the point x= xp = 0.05 on the 45° line and through the intersection ofthe gline with the reetifying-section operating line. The equilibrium stages are stepped off, fist, between the rectifying section operating line andthe equilibrium curve, and then erween te strpping-section operating ine and the equiibium curve, starting ftom point A (dsillatecompesi- ‘on and finishing at point B (bottoms composition). For the opti- tal feed stage, the transfer from the rectifying section operating line to the stripping-section operating line takes place at point P, giving N= 13.2 equilibrium stages the feed going into stage 7 from the top, and N/Nqas = 13.2/6.7 = 1.99. The bottom igure 7.13 Determination of minimum stages for Example 7. Figure 7.14 Determination of minimum refx for Example 7.1. igure 7.15 Determination of numberof equilibrium stages and feed-stage location for Example 7.270 Chapter? Disllation of Binary Mixcores stage isthe pata boiler, leaving 122 equilibrium stages inthe column. If the plate efficiency were 08, 16 trays would be needed (4) Using a process simulator (for thee different feed leeation), as discussed in Chapler 10, the following results are obtained, which show good agreement with the MeCabe- Thiele method, Method Total FeedStage Benzene Benzene Stages fromTop Xp Xe MeCabeThiele 1327 0950 0.050 Simulator B 1 0947 0085 Simulator 4 1 0953 004s Simulator 4 8 0950 0.050 §73_ EXTENSIONS OF THE MCCABE-THIELE METHOD §73.1 Column Operating Pressure Column pressure and condenser type are established by the algorithm shown in Figure 7.16, which is formulated to achieve, if possible, a reflux-drum pressure, Pp, between 0 and 415 psia (2.86 MPa) at a minimum temperature of 120°F (49°C), corresponding to the use of water as condenser cool- ant, Pressure and temperature limits depend on economic factors, Columns can operate at a pressure higher than 415, sia if it is well below the critical or convergence pressure of the mixture, To obtain the bottom pressure, a condenser pres- sure drop of 0 to 2 psi (0 to 14 kPa) and an overall column pressure drop of 5 psi (35 kPa) may be assumed, When the number of trays is known, more refined computations give sun Distiate and bottoms Gaelne PUBBI-PS] pons pata MPs) set Pt 30 sla approximately 0.1 psiftray (0.7 kPa/tray) pressure drop for lalmospheric and superatmospherie pressure operation, and 0.05 psutray (0.35 kPa/tray) pressure drop for vacuum opera tion, A bottoms bubble-point calculation should be made to censure that conditions are not near-critical or above product- decomposition temperatures. As the algorithm indicates, if the bottoms temperature is 100 igh, a lower temperature is, ‘mandated. The pressure in the reflux drum is then reduced, and the calculation of bottoms pressure and temperature is repeated, until they are acceptable. This may result in vac~ ‘uum operation and/or the need for a refrigerant, rather than, cooling wate, fr the condenser. §73.2 Condenser Type ‘Types of condensers are shown in Figure 7.17. A total con- denser is suitable for reflux-drum pressures to 215 pia (1.48, MP2). A partial condenser is appropriate trom 215 psia to 365, pia (2.52 MPa) but can be used below 215 psia to give a vapor distillate. A mixed condenser can provide both vapor and liq uid distillates. A refrigerant is often used as cootant above 365, psia when components are difficult to condense, As illustrated in Example 7.2, a partial condenser provides an additional stage, based on the assumption that liquid reflux leaving the reflux drum isin equilibrium withthe vapor distillate §73.3, Subcooled Reflux Although most distillation columns ate designed so that reflux is a saturated (bubble-point liquid, such is not always: the case. For an operating column with a partial condenser, the reflux is a saturated liquid unless heat losses occur. For & total condenser, however, the reflux is often a subcooled lig- uid if the condenser is not tightly designed, thus resulting in the distillate bubble-point temperature being much higher than the inlet cooling-water temperature, If the condenser Py 218 pal ty bottoms Sresa ot | 2H eae iaovries'c1__ | Condenser a p> 268 psi Ty bottoms slp 86 MPa) Lower pressure Fy Figure 7.16 Algorithm for extablishing disillation-columa, pressure and condenser type$74 Extension ofthe McCabe-Thiele Method 271 aitiate ut " outlet pressure is lower than the top tray pressure, the reflux will be subcooled for all three types of condensers. ‘When subcooled reflux enters the top tray, it causes some vapor entering the tray to condense. The latent enthalpy of condensation of the vapor provides the sensible enthalpy to heat the subcooled reflux tothe bubble point. In that case, the internal reflux ratio within the rectifying section of the col- ‘umn is higher than the external reflux ratio from the reflux, drum. The MeCabe-Thiele construction should be based on the internal reflux ratio, which can be estimated from the fol- lowing equation, which is derived from an approximate energy balance around the top tray: Roun = R( I+ (7-30) where Cr, and AH? are per mole and AT scooting is the degrees of subcooling. The internal reflux ratio replaces R, the external reflux ratio in (7-9). Ifa correction is not made for subcooled reflux, the calculated number of equilibrium, stages is somewhat more than required. Thus, subcooled reflux is beneficial EXAMPLE 7.2 McCabe-Thiele Method When Using a Partial Condenser. One thousand kml of 30 mol% hexane and 70% n-octane is Aistiled in acoluma consisting of apatalreboiler, one equilibrium plate, and a partial condenser, all operating at 1 atm. The feed, 2 bbubbie-poin liquid, is fed to the reboiler, from which a liquid bot toms is withdrawn, Bubble-point reflux ftom the partial condenser is returned to the plate. The vapor distilate contains 80 mol% hex- ane, and the reflux ratio, L/D, is 2. Assume the partial rebole, plate, and partial condenser are equilibrium stages. (a) Using the McCabe-Thiele method, calculate the bottoms composition and ‘kmol/n of distillate produced. (b) If «is assumed tobe 5 (actualy, 43 atthe reboiler and 6.0 atthe condenses), calculate the bottoms composition analytically Solution First determine whether the problem is completely specified. From Table 5.de, Np =C+2N+6 degrees of freedom, where N’ inches the partial reboiler and the stages in he column, but not the partial condenser With N ~2 and C— 2, Np ~ 12, Specified ae Le > Laue iatilate Figure 7.17 Condenser types: (4) total condenser; (b) partial condenser, (¢) mixed © condenser. Fed stream variables 4 Plate and reboiler pressures 2 Condenser pressure 1 0(=0) for plate 1 [Number of sages 1 Feed-stage location 1 Reflux ratio, L/D 1 Distillate composi 1 2 Total 1 ‘Thus, the problem is flly specified and ean be solved. (2) Graphical solution. diagram of the separator is given in Figure 7.18, as isthe MeCabe-Thiele graphical soluion, which is consracted inthe following manner 4. The point yp line $8 atthe partial condenser is located on the 2, Because xq (telox composition) is in equilibrium with yo, ‘the point (xx, yp) is located on the equilibrium curve. 3. Since (L/¥) = 1 —1/[1 + (£/D)] = 2/3, the operating line with slope L/V = 2,5 is drawn through the point yp = 0.8 fon the 45° line until it imersects the equilibrium curve Because the feed is introduced into the partial reboile, there is no sniping section 44. Three stages (partial condenser, plate 1, and partial reboiler) ate stepped off, and the botomts composition xy = 0.135 is read. ‘The amount ofdisilate is determined ftom overall mate- al balances. For hexane. 2pF = ypD +x9B. Therefore, (0.3)(1,000) = (0.8)P + (0.135)B. For the total flow, B~1,000~D. Solving these two equations simulta. neously, D = 248 kmol/ (0 Anabel whan, For = 5 equim qld compen th tn pen bya remem of 3) » — 1 vray @ Frese Gree iad iene ee 1. The guid refx at xp is caleulated from (1) for y=39=08 08 v0 2. yidetermined by a material balance abou the conden: Voi =Dyp tne with D/V=1/3 and L/V =2/3 1/3)(0.8) + (2/3)(0.44) on 4 56mm (Chapter 7 Distillation of Binary Mixtures i own Jie : mses i of pen i pool ramet "9 01 02 09 04 05 06 07 08 09 10 Figure 7.18 Solution to Mot tctonsthoare ne ai Exec 56 — 3. Fem 0) frie pape O ons 4 rome pa ae 4, By material balance around plate 1 and the partial condenser, “dary Sl 0402) one Ye Dyot in By approximating te oliniam sure wih aaa O48 obanedeshn 0195 onda pt), sod 79 = (1/3)(0.8) + (2/3)(0.202) = 0402, For a large number of plates, part (b) can be computed with a treat EXAMPLE 73. McCabe-Thise Method fora Colamm oan with Only a Feed Plate. Solve Example 7.2: (a) graphically, assuming thatthe feed is into duced on plate I rather than into the reboiler, as in Figure 7.19; () by determining the minimum numberof sages required to cary ‘out the separation; (c) by determining the minimum reflux ratio, (a) The solution given in Figure 7.19 is obtained as follows 1. The point x, yn is located on the equilibrium line 2, The operating line fr the rectifying section is drawn through the point y = x= 0.8, witha slope of L/V = 2/3, 3. Intersection of the gline, xy = 0.3 (which, fora saturated liquid, i a vertical line), with the entiching-section operating — ae ender os 4 orf Pen Fe oe) Fos é Fees ef raw é we orf Yona" | 207 02 03 a4 05 08 07 OF DS 10 Figure 719 Solution Example 73,a ZO Figure 7.20 Solution for total eux in Example 73 line is at point P. The stripping-sction operating line must also pass through P, but its slope and the point xy are not known, 4. The slope ofthe stripping-section operating line is found by ‘eial and error to give three equilibrium contacts inthe col- ‘umn, with the middle stage involved in the switch from fone operating line to the other. If the middle stage is the ‘optimal feed stage, the result is xp = 0.07, as shown in Fig- wre 7.19, From hexane and total material balances (0.3)(1,000) = (0.8) +-0.07(1,000~ D). Solving, D = 315 koh ‘Comparing this result to that obtained in Example 7.2 shows that the bottoms purty and distillate yield are improved by feeding to plate 1 rather than to the reboler. ‘This could have hen anticipated ifthe gline hed been con- structed in Figute 7.18, The partial reboiler is thus not the ‘optimal feed stage, (b) The construction corresponding to total reflux (L/V ro products, no feed, minimum equilibrium stages) is shown in Figure 7.20, Slightly more than two staget are required for an xp of 0.07, compared to the three stages previously require. (6) To determine the minimsam refux ratio, the vertical gine in Figure 7.19 is extended from point P until the equilibrium curve is intersected, which i the point (071,03). The slope, (L/ Van of the operating line forthe rectifying section, which connects this point tothe point (0.8, 0.8) onthe 45” line, is 0.18. Thus (L/D)pig = (C/Vrun){1 ~ C/V] = 0:22. This is consider sly less than the L/D = 2 specified $73.4 Reboiler Type Reboilers for industrial-size distillation columns are usually external heat exchangers of either the kettle or the vertical thermosyphon type, shown in Figure 7.21. Hither can provide the large heat-transfer surface required. In the former case, liquid leaving the sump (reservoir) al the bottom of the col- ‘umn enters the ketle, whete itis partially vaporized by wans- for of heat from tubes carrying condensing steam or some other heat-transfer fluid. The bottoms product liquid leaving. §73 Extensions ofthe MeCabe-Thiele Method 273 the reboiler is assumed to be in equilibrium with the vapor returning to the bottom tray. Thus, a kettle reboiler, which is sometimes located in the bottom of @ column, is a partial, reboiler equivalent to one equilibrium stage Vertical thermosyphon reboilers are shown in Figures 7.21b and 7.21c. In the former, bottoms product and rcboiler {eed are both withdrawn from the column bottom sump. Cir- culation through the reboiler tubes occurs because ofa differ- ence in static heads of the supply liquid and the partially, ‘vaporized fluid in the rcboiler tubes. The partial vaporization, provides enrichment in the exiting vapor. But the exiting lig- tid is then mixed with liquid leaving the bottom ay, which contains a higher percentage of volatiles, This type of reboiler thus provides only a fraction of a stage and it is best to take no credit fori. In the more complex and less-common vertical thermosy- phon reboiler of Figure 7.2le, the reboiler liquid is with- drawn from the bottom-tray downcomer. Partially vaporized liguid is returned to the column, where the bottoms product, from the bottom sump is withdrawn, This type of reboiler functions as an equilibrium stage. ‘Thermosyphon reboilers are favored when (1) the bottoms product contains thermally sensitive compounds, (2) bottoms pressure is high, (3) only a small AT is available for heat transfer, and (4) heavy fouling occurs. Horizontal thermosy- phon reboilers may be used in place of vertical types when ‘only smal static heads are needed for cisculation, when the surface-area requirement is very large, and/or when frequent tube cleaning is anticipated. A pump may be added to a ther- mosyphon reboiler to improve citculation. Liquid residence ‘ime in the column bottom sump should be at least 1 minute and pethaps as much as 5 minutes or more. Large columns may have a 10-foot-high sump. §7.3.5 Condenser and Reboiler Heat Duties For saturated-liquid feeds and columns that fulfill the MeCabe-Thiele assumptions, reboiler and condenser duties are nearly equal. For all other situations, it is necessary to establish heat duties by an overall energy balance: Fhy + Qn = Dhy + Bhy + Qc + Qs (rat) Except for small and/or uninsulated distillation equipment, iow, can be ignored, With the assumptions of the MeCabe— ‘Thiele method, an energy balance for a total condenser is Qc = D(R + 1)AH™* (7-32) where AN"? — average molar heat of vaporization, Fora pat- ial condenser, Qc = DRAH™? (73) Fora partial reboier, Oy = BVA (7-38) For a bubble-point feed and total condenser, (7-16) becomes: BV, =L+D=D(R+1) (7.35)274 Chapter? Disillation of Binary Mixtures + condensate Je steam Figure 7.21 Reboilers for plan-rize Aisillation columns: (a) Ketle-type reboler; = condensate (©) vertical thermosyphon-type reboiler, ___ * © For partially vaporized feed and a total condenser, the reboiler duty is less than the condenser duty, and is given by (7-36) oer If saturated steam is the beating medium for the reboiler, the steam rate required is given by the energy balance (7-37) ‘where m, ~ mass flow rate of steam, Ms — molecular weight, of steam, and AH,"” = molar enthalpy of vaporization of The cooling water rate for the condenser is 9. 73) Tea Te) (8) where icy = mass flow rate of cooling water, Cre = specific heat of water, and Ty, Tig = cooling water tempers ture out of and into the condenser. Because the cost of reboiler steam is usually an order of ‘magnitude higher than the cost of cooling water, the feed is, frequently partially vaporized to reduce Qr, in comparison to Qc. as suggested by (7-36) reboile iguid withdfamn from bottom sump; (©) vertical thermosyphon-type reboiler, reboiler liquid withdrawn ftom boton-teay downcomer $73.6 Feed Preheat Feed pressure must be greater than the pressure in the column, at the feed tray. Excess feed pressure is dropped across a valve, which may cause the feed to partially vaporize before centering the columa. Erosion on feed trays can be a serious problem in column operations. Second-law thermodynamic efficiency is highest if the feed temperature equals the temperature in the column at the feed tray. It is best to avoid a subcooled liquid or superheated, vapor by supplying a partially vaporized feed. This is achieved by preheating the feed with the bottoms product or fa process stream that has a suitably high temperature, to censure a reasonable AT driving force for heat transfer and a sufficient available enthalpy. $73.7 Optimal Reflux Ratio A distillation column operates between the limiting condi tions of minimum and total reflux. Table 7.3, which is, adapted from Peters and Timmerhaus [6], shows that as R increases, NV decreases, column diameter increases, and reboiler steam and condenser cooling-water requirements inerease, When the annualized fixed investment costs for the column, condenser, reflux drum, reflux pump, and reboiler are added to the annual cost of steam and cooling water, an§74 Extensions ofthe McCabe-Thicle Method 275 Table 73. Effect of Reflux Ratio on Annualized Cost of a Distillation Operation ‘Annualized Cost, Siye Reboiler Duty, Condenser Duty Cooling ‘Total Annualized Mayy Actual Diam. fe Btw Bul Equipment Water Steam Cost $iyr 100 Infinite «67 9,510,160 9.416000 Tnfinte «17,340 (132,900 Tainite 105 2» 68 9,776,800 9,680,000, 44640 171820 136,500 198,960 14 ea 20 10,221,200 10,120,000 38100 18,600 142.500 199.200 123 Is uM 10,665,500 10,360,000 36480 19,410 148,800, 204,690 132 6 3 1,110,000 11,000,000 35,640 20.220 155,100 210,960 149 4 1 11,998,800 11,880,000 35940 21,870 167,100 224.910 175 8 50. 13,332,000 13,200,000 36870 24,300 185,400 246,570 (Adapted rom an example by Peters and Timmeshaus 6)) optimal reflux ratio of R/Ryig = 1.1 is established, as shown, in Figure 7.22 for the conditions of Table 7.3. ‘Table 7.3 shows that the annual reboiler-steam cost is almost eight times the condenser cooling-water cost, At the optimal reflux ratio, steam cost is 70% of the total. Because {he cost of steam is dominant, the optimal reflux ratio is sen- sitive to the steam cost. At the extreme of zero cost for steam, the optimal R/ Ris is shifted from 1.1 to 1.32, This example assumes that the heat removed by cooling water in the con- denser has no value, ‘The accepted range of optimal to minimum reflux is from 105 to 1.50, with the lower value applying to a dificult sepa- ration (€.g., a = 1.2). However, because the optimal reflux. zatio is broad, for flexibility, columns axe of reflux ratios greater than the optimum, §73.8 Large Numbers of Stages The MeCabe-Thiele construction becomes inviable when, relative volatility or product purities are such that many Figure 7.22 Optimal rellux ratio for a representative distillation operation, stages must be stepped off. In that event, one of the following. five techniques can be used: 1, Use separate plots of expanded scales and/or larger dimensions for stepping off stages at the ends of the yor diagram, eg., an added plot covering 0.95 to L . As described by Horvath and Schubert [7] and shown in Figure 7.23, use a plot based on logarithmic coor- dinates for the bottoms end of the y-x diagram, while for the distillate end, turn the log-log graph upside down and rotate it 90°. As seen in Figure 7.23, the operating lines become curved and must be plotted from (7-9) and (7-14). The 45° line remains straight and the equilibrium curve becomes nearly straight at the two ends, . Compute the stages at the two ends algebraically in the ‘manner of part (b) of Example 7.2. Ths is readily done with a spreadsheet, 4, Use a commercial MeCabe-Thiele program, . Combine the MeCabe-Thiele graphical construction with the Kremser equations of §5.4 for the low and/or high ends, where absorption and stripping factors are almost constant. This preferred technique is illustrated in Example 7.4 EXAMPLE 7.4 McCabe-Thicle Method for a Very Sharp Separation. Repeat part (¢) of Example 7.1 for benzene distillate and bottoms purities of 99.9 and 0.1 mol%, respectively, using a reflux ratio of 1.88, which is about 305% higher than the minimamn refax of 1.44 Al the top ofthe colums, «= 2.52: at the bottom, « = 2.26. Solution Figure 7.24 shows the construction for the region x ~ 0028 to (0.956, with stages stepped off in two directions, starting from the feed stage, In this region, thee ate seven stags above the feed and cight below, fora total of 16 including the feed stage. The Kremser ‘uation is used Yo determine the remaining stages276 Chapter? Disillation of Binary Mixtures on 6.9998 mMT-I TF oot sss y 001 as ‘operating c Figure7.28 Use of log-log ao0os asl coordinates for MeCabe—Thiele 0007 aor 00 0 ey a28oa90,lasees construction: (a bottoms end of ee : colunn;() disilate end of “i column, Adliional stages for the rectifying section: With respeet to Figure 5:82, counting stages from the top down, from Figure 724: Using (1-3). for Cr Drnrne = 0.956, Onsteeune =0.982 and (yy s)yinae = 0.018 ‘Aso, eae 0999 at (ue 001 Combining (555), (5-24), (5-19), (5-48), and (550) 1,2) (yes = 0k sole (a) GS] a Toga Or Dessee Oren (739) where Nj = aldtional rectifying seston stages. Since this is ke an absorption section, it is best to apply (7-39) to toluene, the heavy key, Because a = 252 atthe top of the column, where Kyeqsne i close t0 1, take Kune = 1/252 = 0397. Since R= 1.88, L/V =R/(R+ 1) = 0.658. ° 32 34 Mole fraction of benzene inthe aula, = Figure 7.24 McCahe-Thiele construction for Example 7.4 from 1028 tox = 0956. Therefore, Avinme = L/(VK ee) = 0:653/0.397 = 1.64, whic is assumed constant in the uppermost part of the rectifying section, ‘Therefore, rom (7-38) for toluene, + (1-1) (2018 0.001(0.97) (rei) (coor=ooore397)| 5 Tog Lot Additional stages for the stripping section: As in Figure 5.8b, counting stages from the botiom up we have from Figure 7.24: Cevst)vaaos = 0.048, AI, (2t)eaer = (trae ‘Combining the Kremser equations fora sripping section gives Ns ei 40) Toa(t/Aa) y ‘where, Nj = additional equlirium stages forthe stripping section and A — absorption factor in the stripping section = Z/KV. ‘Benzene is stripped in the stripping section, so itis best to apply (7-40) to the benzene. At the bottom, where Kine 38 apprOX rately 1.0, @ = 2.26, and therefore Kine = 2.26. By material balance, with flows in Ibmolhh, D=270.1. For R= 1388, L= 5078, and V = 270.1 4 507.8~7779. From Example 7.1, Vp =D= 270.1 and Ly = 450-2701 = 1799. Therefore, Z [+ by = 5078 +1799 = 687.7 thmolh and V=V—Vs ‘71779 210.1 = 507.8 lomo. L/V = 6877/5078 = 354, 354/2.26 = 0.599 Substitution into (7-4) gives ons rose 050 (208 =0001228) 9 S001 0001/3265) | 86 Toa(170 599) ‘This value includes the partial reboiler. Accordingly, the total num ber of equilibrium stages starting from the bottom is: partial reboiler 594-8 + foed sage +7 4 5.0 =269. $73.9 Use of Murphree Vapor Efficiency, Eyyy In industial equipment, itis not often practical to provide the ‘combination of residence time and intimacy of contactrequired to establish equilibrium on each stage. Hence, con- centration changes are less than those predicted by equilibrium, ‘The Murphree vapor efficiency, introduced in $6.5.4, describes tray performance for individual components in, either phase and is equal to the change in composition divided by the equilibrium-predicted change. When applied to the vapor phase in a manner similar to (6-28) (rat) where Eyy is the Murphree vapor efficiency for stage n,n 1is the stage below, and y3 is the composition in the hypo- thetical vapor phase in equilibrium with the liquid leaving ‘The component subscript in (7-41) is dropped because values of Eyyy ate equal for two binary components. In stepping off stages, the Murphree vapor efficiency dic: tates the fraction of the distance taken from the operating line ‘o the equilibrium line. This is shown in Figure 7.25a for the cease of Murphree efficiencies based on the vapor phase. In, Figure 7.25b, the Murphree tray efficiency is based on the liquid. In effect, the dashed eurve for actual exit-phase com- position replaces the thermodynamic equilibrium curve for & particular set of operating lines, In Figure 7.258, Ey = EF/EG = 0.7 for the bottom stage. $73.10 Multiple Feeds, Sidestreams, and Open Steam ‘The McCabe-Thiele method for a single feed and two prod: ‘ucts is extended to multiple feed streams and sidestreams by adding one additional operating line for each additional steam, A multiple-feed arrangement is shown in Figure 7.26. In the absence of sidestream Ls, this arrangement has §74. Extensions ofthe McCabe-Thiele Method 277 no effect on the material balance in the section above the upperfeed point, F\. The section of column between the upper-feed point and the lower-feed point, F2 (in the absence of feed F), is represented by an operating line of slope 1'/V', which intersects the rectifying-section operating line. A’ similar argument holds for the stripping section Hence itis posible to apply the MeCabe~Thiele construction shown in Figure 7.27, where feed F is a dew-point vapor, while feed Fis a bubble-point Liquid. Feed F and sidesteam Lz of Figure 7.26 are not present, Thus, between the two feed points, the molar vapor flow rate is V'=V~ Fy and Lai + Fy = L + Fy. For given x9,24,. 2. Xp and L/D. the three operating lines in Figure 7.27a are constructed A sidestzeam may be withdrawn from the rectifying section, the stripping section, or between multiple feed points, as a saturated vapor or saturated liquid. Within ‘ateral-alance constraints, Lg and xs can both be specified. In Figure 727, a saturated-liquid sidestream of LK mole fraction xs and molar flow rate Ls is withdrawn from the ree- tifying section above feed F. Inthe section of stages between the side steam-withdrawal stage and the feed stage, L=L~ Ly, while V! =V. The McCabe-Thiele construc- tions determine the location of the sidestream stage. However, IF is not located dicey above xs, the reflux ratio must be varied until it does For certain distllations, an energy source is introduced directly into the base of the column. Open steam, for example, can be used if one of the components is water or if water can form a second liquid phase, thereby reducing the boiling point, as inthe stam disillaon of fat, where there is no reboiler and heat is supplied by superheated steam. Commonly, the feed contains Water, which is removed as bottoms. In that application, Ox of Figue 7.26 is replaced by a steam of com- position y = 0 (pure steam), which, with x = xp, becomes @ point on the operating line, since these passing streams exist at the end of the column, With open steam, the bottoms low rate w igure 7.25 Use of Murphiee = Feito — None cave at Mare even) plate efficiencies in McCabe ‘Thiele consirction,278 Chapter7 Distillation of Binary Mixtures Condensete ec A Side stream a Borroms Figure 7.26 Complex disillaion column with multiple feeds and sidestream, ‘Saurated vapor assumed, \ is increased by the flow rate of the open steam, The use of ‘open steam rather than a reboiler for the operating condition, F = F; = Ls = O's vepresented graphically in Figure 7.27c. EXAMPLE 7.5 McCabe-Thiele Method for Column with a Sidestream. ‘A coluran equipped with pata reboiler and total condenser, ope ating at steady slate with a saturatedliquid feed, has liquid side stream in the rectifying section. Using the McCabe~Thiele assumptions: (a) derive the two operating lines in the rectifying sec sion; (6) find the point of intersection ofthe operating lines; (6) find the intersection of the operating line between F and Ls with the diagonal; and (d) show the constuction on 23+ diagram, Solution (a) By material balance over Section 1 in Figure 7.28, Ve-iy4-1 Inka + DXp, Fot Section 2, Vs-a¥e a= L's 1X5-1 + laxs +Drp, The (wo operating lines for conditions of constant molar overflow become: Figure 7.27 MoCabe-Thicle construction or complex columns: (a) two feeds (saturated liquid and saturated vapor: (6) one feed, one sidesteam (saturated liquid); () use of openSection Figure 7.28 Distillation column with sidestream for Example 75 Lp LL tsxs~Dxp yaEertay and y= Exp EP (by Eaquating the two operating lines, the intersection occurs a (L—L)yx = Lsxs and since L—L! = Ly, the pont of itersec- Vion Becomes x = xs Figure 7.29 McCabe-Thiele diagram for Example 75. §74 Estimation of Stage Eficiency for Distillation 279 (6) The intersection ofthe lines LL, faxs + Dry y= Lx + Dp and y= xovoutsaty = BEA TSE (@) The y-x diagram is shown in Figure 7.29. §7.4 ESTIMATION OF STAGE EFFICIENCY FOR DISTILLATION Methods for estimating stage efficiency for binary distillation, are analogous to those for absorption and stripping, with one major difference, In absorption and stripping, the liquid phase is often rich in heavy components, and thus liquid vis- cosity is high and mass-transfer rates arc low. This leads to low stage efficiencies, usually less than 50%, For binary dis Lllation, particularly of close-boiling mixtures, both compo- nents are near their boiling points and liquid viscosity is low, with the resull that stage efficiencies for well-designed trays: are often higher than 70% and can even be higher than 100% for large-diameter columns where a crossilow effect is present. §7.4.1 Performance Data ‘Techniques for measuring performance of industrial distilla- tion columns are described in AIChE Equipment Testing Procedure [8]. Overall column efficiencies are generally measured at conditions of total reflux to eliminate transients due to fluctuations from steady state that are due to feed vari- ations, etc. However, as shown by Williams, Stigger, and Nichols (9), efficiency measured at total reflux (L/V = 1) ccan differ from that at design reflux. A significant factor is hhow closely to flooding the column is operated. Overall effi- ciencies are calculated from (6-21) and total reflux data. Indi vvidual-tray, Murphree vapor efficiencies are calculated using (6-28). Here, sampling from downcomers leads to variable results. To mitigate this and other aberrations. it is best to ‘work with near-ideal systems, These and other equipment- specific factors are discussed in $6.5 Table 7-4, from Gerster et al. [10]. lists plant data for the distillation at total reflux of a methylene chloride (MC ethylene chloride (EC) mixture in a 5.5-ft-diameter columa, containing 60 bubble-cap trays on 18-inch tray spacing and ‘operating at 85% of flooding at total reflux. EXAMPLE 7.6 Tray Efficiency from Performance Data, Using the performance data of Table 74, estimate: (a) the overall, ‘cay efficiency forthe section of trays from 35 t0 29 and (b) Ewv for cay 32, Assume the following values for acc280 Chapter? Distillation of Binary Mixtures Table 7.4 Performance Data for the Distillation of a Mixture of ae wane Jue from 7-3) ‘Methylene Chloride and Ethylene Chloride 0.00 355 0.00 Company Eastman Kodak 010 361 0.286 Location Rochester, New York 020 370 ast Column diame! 5.51 (65.5 inches LD) 030 376 0617 No, of trays 6a 040 383 omg ‘Tray spacing 18 inches 050 391 0.796 ‘Type try 10:0Ws of 3-nch-diameter bubble 0.60 400 0857 caps on 47-inch wiangulat 070 4.03 0.904 centers 115 capsiray 0x0 4.08 0.942 Bubbling rca 2ore 0.90 4a7 0974 ‘Length of liquid eave 29 inches 1.00 425, 1.00) Ouilet-weir height 225 inches Downcomer clearance 15 inches Liquid rate 245 galimin-t = 1,115.9 Thin Vapor Ffactor 1.31 fs aba’)? Percent of flooding 8 ee Pressure, top tay 33.8 sia (@) The above x-a-y data ate ploted in Figure 7:30. Four equli-. Pressure, bottom try 22.0psia brium stages are stepped off from xs) = 0.898 to x39 = 0.0464 Liquid composition, mole % methylene chloride: for total reflux, Since the actual number of tages is also 4,E, From tray 33 398 from (6-21) is 100% Prom try. 26 (©) Attota reflux conditions passing vapor and liguid streams have From tay 29, 4.64 the same composition, +0 the operating line is the 45° line, <§ ————————________________ Source: FA, Gerster, AB. Hill, NHL Moehgrot, nd DB. Robinson, Tray Using this, ogether with the above performance dala and the equilibrium curve in Figure 7.30 for methylene chlorie, with ‘ways counted from the bottom up fon 6.28 (Ew) == Methylene ehlorde= = Figure 7.30 McCabe-Thiele diagram for Example 7. Bffcences in Disillacion Colinas, Final Report from the Unversity of Delaware, AICHE, New York 1958, §7.4.2 Empirical Correlations of Tray Efficiency Based on 41 sets of data for bubble-cap-tray and sieve-tray ccolumns distilling hydrocarbons and a few water and misci- ble organic mixtures, Drickamer and Bradford [11] corre lated E, in terms of the molar-average liquid viscosity, 1, Of the lower feed at average tower temperature, The data covered temperatures from 157 to 420°F, pressures from 14.7 to 366 psia, feed liquid viscosities from 0.066 to 0.355 eP, and overall tray efficiencies from 41% to 88%. “The equation Ey = 133 ~668l0g 742) where Ey isin % and wis ine, fs the data with average and tmaximum percent deviations of 5.0% and 13.0%. A plot of the Drickamer and Bradford correlation, compared fo per formance data for dsillaton, i given in Figure 731. Equa- tion (7-42) is resticted to the above range of data and is, intended mainly for hydrocarbons $65 showed that mass-tansfer theory predicts that over a wide range of a, the importance of liquide and gas-phase mass-transfer resistances shifts. Accordingly, O’Connell [12] found that the Drickamer-Bradford formulation is in- adequate for feeds having a large a. O°Connell also devel ‘oped separate correlations in terms of 4a for fractionators and for absorbers and strippers. Ax shown in Figure 7.32 Lockhart and Legget [13] were able to obtain a single corre. lation using the produc of liguid viscosity and an appropriate69 overall stage effiieney, % Figure 7.31 Drickamer and Bradford's corelation for plate efficiency of distillation columns volatility as the correlating variable. For fractionators, x41 was used; for hydrocazbon absorbers, the volatility was taken, as 10 times the K-value of a key component, which must be distributed between top and bottom products, Data used by O'Connell cover a range of a ftom 1.16 to 205. The effect on E, of the ratio of liquid-to-vapor molar flow rates, L/V, for eight different water and organic binary systems in, a 10-inch-diameter column with bubble-cap ways was reported by Williams etal. (9]. While L/V did have an effect, it could not be correlated. For fractionation with L/V nearly equal to 1.0 (ce., total reflux), their distillation data, which, are included in Figure 7,32, are in reasonable agreement with the O'Connell correlation, For the distillation of hydrocar- bons in a 0.45-m-diameter columa, Zuiderweg, Verburg, and Gilissen [14] found the differences in E,, among bubble-cap, sieve, and valve trays to be insignificant at 85% of flooding Accordingly, Figure 7.32 is assumed to be applicable to all, Uhree tray types, but may be somewhat conservative for well-designed trays. For example, data of Fractionation Distiletion of water solutions Absorption of hydrocarbons 2 Disitetion ante of Willams a3] i Disiletion data of FR for valve trey (24) 874 Estimation of Stage ficiency for Distillation 281 Research Incorporated (FRD for valve trays operating with hydrocarbon systems, also included in Figure 7.32, show effi- ciencies 10% fo 20% higher than the comelation. For just the distillation data plotted in Figure 7.32, the (O'Connell corzelatin fits the empirical equation Ey = 50.3(an)°™* (7-43) where E, is in %, w is in ¢P, and ois at average colomn conditions. The data in Figure 7.32 are mostly for columns with Liquid flow paths from 2 to 3 ft, Gautreaux and O'Connell [15] showed that higher efficiencies are achieved for longer flow paths because the equivalent of two or more completely mixed, successive liquid zanes may be present Provided that jit les between 0.1 and 1.0, Lockhart and Leggett [13] recommend adding the increments in Table 7.5 to the value of ,, from Figure 7.32 when the liquid flow path is greater than 3 ft, However, at high liquid flow rates, long liguid-path lengths are undesirable because they lead to excessive liquid gradients and cause maldistribution of vapor Table 75 Comection to Overall Tray Etficiency for Length of Liquid Flow Path (0.1 < pa < 10) Length of| Value to Be Added to Liguid Flow Pat, f B, fom Figure 7.32, % 3 ° 4 10 5 1s 6 20 5 2 10 25 1s 2 17 Meet, Jp, Bae, Advances in Prtroleam Chemistry and Reining. Vol 1 nescence, New Yor, pp. 323-226 (0950, Figure 7.32 Lockhatt and Leggett version ofthe O° Connell correlation for overall ay efficiency of frationators, sbsorbers, and tippers. [Adapted from FY. Lockhart and CW. Lee fen KA. Kobe and 1. MeKet, J, Es ‘Advances ix Petroleum Chemisty and Ref ing, Iersienee, New Yor, Vol 1. pp. 323— 526 1958).282 Chapter? Distillation of Binary Mixtures flow, in which case multipass trays, shown in Figure 6.15 and discussed in $6.5.3, are preferred, EXAMPLE 7.7 Estimation of Stage Efficiency from Empirical Correlations. For the benzene-toluene distillation of Figure 7-1, use the Drckamer— Bradford and O'Connell correlations to estimate E, andthe number of ‘actual plates cequired. Obtain the column height, assuming 24-inch teay spacing with 4 ft above the top tay for removal of entrained liguid and 10 ft below te botiom tay for bots surge capacity. The separation requites 20 equilibrium stages plus a partial reboler that acts as an equilibrium stage, Solution ‘The liquid viscosity is determined at the feed-stage condition of 220°F, assuming a liquid composition of $0 mol% benzene; of benzene = 0.10 cP; 4 of toluene = 0.12 ¢P; and average 4 — O.11 cP From Figure 7.3, the everage ais 29 + ~T Average: From the Drickamer-Bradford correlation (7-42), By (66 log(0.11) ~ 17%. Therefore, N, ~ 20/0.77 ~ 26. Column height = 4-+ 2(26 1) + 10 = 64 From the O’Connell correlation, (7-43), Ey = 50.32.39) (oy? = 8%, For a Siiameter column, the Iength of the liguid flow path is, bout 3 ft fora single-pass tray and even lest for a two-pass tray. From Table 7:5, the efiiciency correction is zero. Therefore [Ny = 20/0.68 = 29.4, or round up to 30 trays. Column height = 42(30-1) +10=72f §7.4.3 Semitheoretical Models for Tray Efficiency In §6.5.4, semitheoretical tray models for absorption and stripping based on Eyyy (6-28) and Eoy (6-30) are developed, ‘These are also valid for distillation. However, because the equilibrium line is curved for distillation, X must be taken as, ‘mV/L (not KV/L = 1A), where m = local slope of the equi- librium curve = dy/dx. In §6.6.3, the method of Chan and Fair (16) is used for estimating Egy from mass-transfer con~ siderations. Eyyy can then be estimated. The Chan and Fair correlation is applicable to binary distillation because it was developed from data for six binary mixtures, §7.44 Scale-up from Laboratory Data [Experimental pilot-plant or laboratory data are rarely neces- sary prior to the design of columns for ideal or nearly ideal binary mixtures. With nonideal or azeotrope-forming solu- tions, use of a laboratory Oldershaw column of the type dis- cussed in §6.5.5 should be used to verily the desired degree of separation and to obtain an estimate of Egy. The ability to predict the efficiency of industrial-size sieve-tray columns {rom measurements with [-inch glass and 2-inch metal diam- eter Oldershaw columns is shown in Figure 7.33, from the work of Fair, Null, and Bolles [17], The measurements are for eyclohexane/n-hey 7.33a) and near-atmospheric conditions (Figure 7.336), and for the isobutane/n-butane system at 11.2 atm (Figure 7.33) ‘The Oldershaw data are cortelated by the solid lines. Data for the 4-ft-diameter column with sieve tays of 8.39% and 13.7% ‘open area were obtained, respectively, by Sakata and Yanagi [18] and Yanagi and Sakata [19], of FRI. The Oldershaw col- ‘umn is assumed to measure Eoy. The FRI column measured at vacuum conditions oo Senor ain Percent loos Percent ood 7 iz arm 1a oo Ho Figure 7.33 Comparison of Oldershaw column efficiency with point efficiency in 4 diameter FRI column with seve tray (a) cyclohexane Percent food im heptane syste: () cyelohexane/n-heptane system: (©) isobutanefn-butane system,E,, bul the relations of $6.5.4 were used to convert the FRI dala to Foy. The data cover % flooding from 10% to 95%. Data from the Oldershaw column ate in agreement with the FRI data for 14% open area, except at the lower part of the flooding range. In Figures 7.33b and 7.33c, FRI data for 8% ‘open area show efiiciencies as much as 10% higher, §7.5 COLUMN AND REFL DIAMETERS UX-DRUM ‘As with absorbers and strippers, distillation-column diameter is calculated for conditions atthe top and bottom trays of the tower, using the method of $6.6.1. Ifthe diameters differ by 1 {or less, the larger diameter is used for the entire column, If the diameters differ by more than 1 ft itis often more eco- rnomical to swage the column, using the different diameters computed for the sections above and below the feed, $75.1 Reflux Drums {As shown in Figure 7.1, vapor flows from the top plate to the condenser and then to a cylindrical reflux drum, usually located near ground level, which necessitates a pump to lift, the reflux to the top of the column. Ifa partial condenser is, used, the drum is oriented vertically to facilitate the separa- lion of vapor from liquid —in effect, acting asa flash drum, Vertical Drums Vertical refx and flash drums are sized by calculating @ rminimum drum diameter, Dy, to prevent liquid carryover by entrainment, using (6-44) in conjunction with the curve for 2é-inch tray spacing in Figure 6.23, along with values of Faq = 1.0 in (6-42), f =0.85, and Ay = 0. To handle pro- cess fluctuations and otherwise facilitate contol, vessel vol- ‘ume, Vy, is determined on the basis of liquid residence time, 1, which should be at least 5 min, withthe vessel half full of liguid [20} 2LMit % Where Lis the molar liguid flow rate leaving the ve suming a vertical, cylindrical vessel and neglecting bead vol- ume, the vessel eight H is Ww. Vv =D} (7-48) However, if H > 4Dr, itis generally preferable to increase Drand decrease H to give H = 4D. Then H_(v\"" Dr a 7-46 G ) (7-46) ‘Akeight above the liquid level ofa least 4 ft is necessary for feed entry and disengagement of liquid droplets from the vapor. Within this space, its common to install a wite mesh pad, which serves as a mist eliminator. 975 Column and Refiux-Drum Diameters 283 Horizontal Drums When vapor is totally condensed, a cylindrical, horizontal reflux drum is commonly employed to re sate, Equations (7-44) and (7-46) permit estimates of the drum diameter, Dy, and length, H, by assuming a near- ‘optimal value for H/Dy of 4 and the liquid residence time suggested for a vertical drum. A horizontal drum is also used following a partial condenser when the liquid flow rate is appreciably greater than the vapor flow rate. cive the conden- EXAMPLE 7.8 Diameter and Height of a Flash Drum. Equilibrium vapor and liquid streams leaving a flash drum supplied by a partial condenser are as follows: Component Vapor Liquid Lime Hel 492 os Benzene 118s 314 Monochlorobenzene is: 1785 Total 2392 260.7 Lot 19,110 26.480 TF 270 270 P.psia 35 35 Density, os7t 57.08 Determine the dimensions ofthe Mash dru Solution Using Figure 6.24 26,480 (0371) _ 5 1 fw =o (arm) ~ 0 anu Cp ata 24-inch way spacing is 034, Assume, in (42), that c-G From (640, ‘57.08 — 0.371\ °° vp =034(2852572)" ga —.200n From (6-44) with Ag/A From (7-44), with £= 5 minutes = 0.0833 b (26.480}(0.0833) nae (57°08) 3 From (7-43), (40773) OTS) 938 Gje26) However, H/Dr = 193/2.26 ~ 854 > 4, Therefore, redimension Vor H)Dy =284 Chapter? Distillation of Binary Mixtures From (7-46), my" pr = (ZS) = 2918 and H=4 Dr = (4)(2.91) = 11.648 Ga) ue Height above the liquid level is 11.64/2 = 5.82 8, which is ade quate. Alternatively. with a height of twice the minimum dis. engagement height, H ~ 8 ftand Dy — 3.58 §7.6 RATE-BASED METHOD FOR PACKED DISTILLATION COLUMNS Improvements in disteibutors and fabrication techniques, and ‘more economical and efficient packings, have led to increas- ing use of packed towers in new distillation processes and the retrofitting of existing wayed towers to increase capacity and reduce pressure drop. Methods in $6.7 and $6.8 for estimat- ing packed-column parameters and packed heights for absorbers are applicable to distillation and are extended here, for use in conjunction with the McCabe~Thiele diagram, [Both the HETP and HTU methods are covered, Unlike dilute solution absorption or stripping, where values of HETP and HTU may be constant throughout, values of HETP and HT ccan vary, expecially across the feed entry, where appreciable changes in vapor and liquid traffic occur. Also, because the equilibrium line for distillation is curved, equations of $6.8 ‘must be modified by replacing X = KV/L with ‘mV _ slope of equilibrium curve TL slope of operating line where m = dy/dx varies with location. The efficiency and amass-transfer relationships are summarized in Table 7.6. 87.6.1 HETP Method for Distillation In the HETP method, equilibrium stages are first stepped off on a McCabe-Thiele diagram, where equimolar counter- diffusion (EMD) applies. At each stage, 7, P, phase-flow ‘Table 7.6 Modified Efficiency and Mast-Transfer Equations for Binary Distillation mvt. m= dyjdx Etciency: Equations (6-31) to (6-37) hold if is defined by (7-47) Mase tans: Loa 7) local slope of equilibrium curve aed ors) mb og mPMtlo, a9) kia ue +e 7-50) roid Kaa Ka kya on Hoc =H + Ni, asp ETP ~ Hog lnd/04~ 1) sy ratio, and phase compositions are noted. A suitable packing ‘material is selected, and the column diameter is estimated for ‘operation at, say, 70% of flooding by one of the methods of $638. Mass-transfer coefficients for the individual phases are estimated for the stage conditions from correlations in $6.8, From these coefficients, values of Hog and HETP are esti- ‘mated for each stage and then summed to obtain the packed heights of the rectifying and stripping sections. If experimen- tal values of HETP are available, they are used directly. In ‘computing values of Hog from Hg and Hy, ot K, from k, and ze, (6-92) and (6-80) must be modified because for binary distillation, where the mole fraction of the LK may range from almost 0 at the bottom of the column to almost I at the top, the ratio (») ~ y*)/(2x1 ~ x) in (6-82) is no longer a con- stant equal to the K-value, but the ratio is dy/d equal to the slope, m, of the equilibrium curve, The modified equations are included in Table 7.6, EXAMPLE 7.9 Packed Height by the HETP Method. For the benzene-toluene distillation of Example 7.1, determine packed heights ofthe rectifying and stipping sections bated onthe following values for the individual HTUs, Included are the LV val ‘ues for each section from Example 7. Ht Lv Rectfying section 116 as, 02 ‘Sipping section 0.90 053 140. Solution Equilibrium curve slopes, m from Figure 7.15 and values of bate from (7-47), Hog for each stage is from (7-52) in ‘Table 7.6 and HETP for each stage is from (7-53). Table 7.7 shows ‘that only 0.2 of stage 13 is needed and that stage 14 isthe patil, reboiler. From the results in Table 7.7, 10 ft of packing should be sed in each section, Table 7.7_ Results for Example 7.9 Sagem SormV/E Hoot HETR.A 1 oa7 076 132. 174 2 oss os 136 170 3 061 098, 162 164 4 ost 108 168 162 5 on 116 im 159 6 080 129 mn 156 Total for rectifying section 935 1 090 068 132 164 8 098. 070 128 132 9 Lis os 134 147 10140 1.00 143 143 m0 121 153 140 2 190 136 182 138 1220 137 173 13102)=0.27 Total for stripping section on Total packed height 18.96§7.6.2, HTU Method for Distillation In the HTU method, stages are not stepped off on a McCabe Thiele diagram. Instead, the diagram provides data to per form an integration over the packed height using mass-transfer coefficients or transfer units Consider the packed distillation column and its McCabe Thiele diagram in Figure 7.34. Assume that V,L, V. and L are constant. For equimolar countercurrent diffusion (EMD), the rate of mass transfer of the LK from the liquid to the vapor phase is n= kya(x— x1) = kay, —y) (7-58) Rearranging, ney (7-55) In Figure 7.34, for any point x,y) on the operating line, the interfacial point (xj, y;) on the equilibrium curve is obtained by drawing a line of slope (—eya/ya) from point (x,y) t0 the point where it intersects the equlivium curve ‘By material balance over an incremental column height, with constant molar overflow, Vy = kyaly; ~»)5 dl Lax ~ kya(x~ x)S dl (756) (751) where S is the cross-sectional area of the packed section. In- tegrating over the rectifying section, ain pe Vidya tee [dm [ray I. se (rss) Mody _'® Huds «Oi [ee Eo Or Lo ih ea 7 ® ® Figure 7.34 Disillation in a packed column. 916 Rate-Based Method for Packed Disilation Columns 285 Integrating over the stripping section, vy. fille) Pe Vdy ede (hs a a ie KaS(x— x) (7-60) Hedy l Hyde or y)s= [” He® = 7-61) Oe [S-Las Values of k, and ke vary over the packed height, causing the slope (—kya/kya) to vary. If kea > kya resistance to mass twansfer resides mainly in the vapor and, in using (7-61). itis most accurate to evaluate the integrals in y. For kya > Kea, the integrals in x are used. Usually, itis suficient to evaluate k, and k, at three points in each section (o determine their variation with x. Then by plotting their ratios from (1-55), @ locus of points P can be found, from which values of (9; ~ ») for any value of y, oF (x — 33) for any value of x, eam be read for use with (7-58) to 7-61). These integrals can be evaluated either graphically or numerically EXAMPLE 7.10 Packed Height by the HTU Method. Two hundeed and fity kmoUh of saturated-liquid feed of 40 mol% ‘sopropyl ether in isopropanol is distilled in a packed columa oper- sting at | atm to obtain distillate of 75 mol% isopropyl ether and 3 bottoms of 95 mol‘ isopropanol. The rellux ratio i 1. times the ‘minimum and the column has a total condenser and partial reboiler, The mass-transfer coeficients given below have been estimated from empirical correlations in §6.8. Compute the packed heights of the rectifying and stripping sections Solution From an overall material balance on isopropyl ether, 0.40(250) = 0.750 + 0.05(250 ~ D) Solving, D = 125kmoUhand B = 250 ~ 125 = 125kmolh ‘The equilibrium curve at 1 atm is shown in Figure 7.35, where io propyl eter isthe LK and an azeotrope is formed at 78 mol ether 020408 Mole faction of isopropyl ether in the lau, = moe 19 Figure 7.35 Operating lines and minimara reflux line for Example 710,286° Chapter? Distillation of Binary Mixtures Mole fraction of isopropyl Figure 7.36 Mass-transfer driving forces for Example 7.10 The distillate composition of 75 mol% is safely below the s200- tropic composition, Also shown in Figure 7.35 ar the q-line and the reclification-section operating line for minim relux. The slope of the later is measured tobe (11/V) nj, = 039. From (7-27), Rain =0.39/(1 039) =0.64 and R=15 Rox =0.96 = RD = 0,96(125) = 120kmolh and LD = 120+ 125 = 245 kmol Li Ly = 120 § 250 = 370 kmolh Va V~Vp= 245 ~0 = 245 kmoln Rectification operaing-ine slope = L/V = 120/245 = 0.49 This line and the stipping-section operating line are ploted in Figure 7.35. The partial reboler is stepped off in Figure 7.36 to tive the following end points for determining the packed heights re the symbols refer to Figure 7.34 of the two sections, wh Supping Section Rectfying Section Top (xp = 0.40, yp = 0577) (x, = 0.75, yp = 075) Bottom (41 =0.135, 9, =0.18) (x = 040, yp = 0577) ‘Mase-ransercoelliciente at three values of x areas follows ka ka x __kmol/nhmole faction) kmol/m?tmole fraction) Stripping section: ous 305 1,680 02s 300 1760 035 335 1.960 Rectfying section 045 1s 610 0.60 180 610 ors 165 165 ‘Mass-ansfer-coeffieient slopes are computed for each point x fon the operating lie using (7-55) and are drawn from the operating, line to the equilibrium line in Figure 7.36. These are tie lines ‘because they tie the operating line tothe equilibrium line. Using the sie lines as hypotenuses, ight tiangles are constructed, as shown in Figure 7.36, Dashed lines, AB and IC, are then drawn through the points a the 90° triangle comers. Additional ie lines can, as needed, be added to the thee plotted lines in each section to give better accuracy, From the i lines, values of (9) —)) ean be tabulated for ‘opetating-line y-values. Colman diameter isnot given, ste packed volumes are determined irom rearrangements of (7-58) and (7-60), with V— Sip vid ri ve fre 7.82) Lee q Vdy 83) v,- [Yo 7-3) |, Bao, ae Values of ba are interpolated ar ncroser, Reels we Vor) o ka OTN > ® Bao ‘Sipping section 018 os 307 55 025 0150303 54 035 013 300 57 04s 0103320 14 osm 0.030 350 23 Recttying section: 0377 0030187 437 0.60 0033185 401 065 oom 182 498 070 oo 175 523 075 0010 165 185 By numerical integration, Vp = 8.6m and Vp = 123m! §7.7_ INTRODUCTION TO THE PONCHON- SAVARIT GRAPHICAL EQUILIBRIUM-STAGE, METHOD FOR TRAYED TOWERS ‘The McCabe~TThiele method assumes that molar vapor and liquid flow rates are constant, This, plus the assumption of no heat losses, eliminates the need for stage enezgy balances. ‘When component latent heats are unequal and solutions non- ideal, the MeCabe-Thiele method is not accurate, but the Ponchon-Savarit graphical method [21, 22], which includes energy balances and utilizes an enthalpy-concentration dia- {gram of the type shown in Figure 7.37 for n-hexane/n-octane, is applicable. This diagram includes curves for enthalpies of saturated-vapor and liquid mixtures, where the terminal points of tie lines connecting these two curves represent equi librium vapor and liquid compositions, together with vapor and liquid enthalpies, at @ given temperature. Isotherms above the salurated-vapor curve represent enthalpies of sup- cerheated vapor, while isotherms below the saturated-liquid$7.7 Inteoduction tothe Ponchon-Savarit Graphical Hquiibrium-Stage Method for Trayed Towers 287 Figure 7.37 Enthalpy-concentraion digra for mehexane/ curve represent subcooled liquid. In Figure 7.37, a mixture of 30 mol% hexane and 70 mol% octane at 100°F (Point A) isa subcooled liquid. By heating it to Point B at 204°F, it be- comes @ liquid atts bubble point. When a mixture of 20 mol% hexane and 80 mol% octane at 100°F (Point G) is heated to 243°F (Point B), at equilibrium it splits into a vapor phase at Point F and a liquid phase at Point D. The liquid phase con- tains 7 mol% hexane, while the vapor contains 29 mel% hexane. Application of the enthalpy-concentration diagram to cequilibrium-stage calculations is illustrated by considering a single equilibrium stage, n 1, where vapor from stage 7 2 is mixed adiabatically with liquid from stage m to give an overall mixture, denoted by mole fraction 2, and then brought to equilibrium. The process is represented in two steps, mix- ing followed by equilibration, at the top of Figure 7.38. The energy-balance equations for stage n — 1 are Mixing: Vy-alln-2 | Lgl =(Vn-z | En)he (7-64) Eyuilibration: — (Vy-x + Inte = Vaya + Lathe (7-68) where H and h are vapor and liquid molar enthalpies, respee~ tively. The material-balance equations for the light compo- nent are Mixing: Yn-2Vnn2 + Xuln = 2(Vn2 + (7-66) Equilibration: 2(Vn-2 + Ln) = Yn Vent + Xn-aln. (7-87) Simultaneous solution of (7-64) and (7-6) gives Ho. ot (7-68) Yao Fe which isthe three-point form of a straight line plotted in Fig- ture 7.38, Similarly, solution of (7-65) and (7-67) gives ahs hem hy Dei == (7-69) Which is also the equation for a straight line. However, inthis case, y,, and %,-1 ate in equilibrium and, therefore, the points Hy.1,3q 1) and (Hy-1,35 4) must lie on opposite ends of the te line that passes dhrough the mixing point (f., 2), a8 shown in Figure 7.38, ‘The Ponchon-Savarit method for binary distillation is an extension of the constructions in Figure 7.38 to counter- current cascades. A detailed description of the method is not, given here because it has been superseded by rigorous, com- puter-aided calculation procedures in process simulators, Which are discussed in Chapter 10. A detailed presentation of the Ponchon-Savarit method for binary distillation is given, by Henley and Seader [23] Enthalpyinit quant, i igure 7.38 Two-phase mixing and equilibration on an enthalpy-concentration diagram288 Chapter? Disullation of Binary Mixcures SUMMARY 1. A binary mixture can be separated economically into two nearly pure products by distillation if « > ~1.05 and no azeotrope forms. 2, Distillation is the most mature and widely used separa- tion operation, with design and operation practices well established. 3. Product purities depend mainly on the number of equili brium stages in the rectifying section and the stripping section, and to some extent on the reflux ratio. However, both the number of stages and the reflux ratio must be greater than the minimum values corresponding to total reflux and infinite stages, respectively. The optimal RoR is usually in the Fange of 1.10 to 1.50. 4, Distillation is conducted in trayed towers equipped with sieve or valve trays, or in columns packed with random or structured packings. Many older towers are equipped with bubble-cap trays 5. Most distillation towers have a condenser, cooled with water, to provide reflux, and a reboiler, heated with steam, for boilup. 6. When the assumption of constant molar overflow is valid, the McCabe-Thiele graphical method is conve- nient for determining stage and reflux requirements ‘This method facilitates the visualization of many aspects of distillation and provides a procedure for locating the feed stage. 7. Design of a distillation tower includes selection of opet- ating pressure, type of condenser, degree of reflux sub- cooling, type of reboiler, and extent of feed preheat. REFERENCES 1, Forbes, RA, Short History ofthe Ar of Disltion, 1. Bel, Leiden cases) 2. Mix, TW, 1S. Dweck, M. Weinberg, nd B.C. Armstong, Chom ng Pron, THs), 48-55 (1978) Kiser, HZ, Dist Design, McGrail, Now York (199) 4. Kiste, HZ, Distilation Operation, MeCraw-Hil (4990, 5. McCabe, WIL, and EW. Thiel, Ind. Eng. Chem, 17, 605-611 4925), New York 6 Peters MS.,and KD, Timmerhavs, lant Design and Economics for Chemical Engincer, sth 8, MeGra-Hl, New York (199), 7. Horvath, PJ, and RF. Sebuber, Chem. Eng 6(3), 129-182 (1958) 8. AICHE Equipment Testing Procedure, Tray Dislation Columns. 20d ALCRE, Now York 098, 9. Wiliams, GC. BK. Stigger and LHL Nichol, Cham. Brg. Progr 46 7-16 4950, 10. Gerster, 1A, AB. Til, NH, Hochgofané DB Robinson, Tray Ef ences m Distillation Colunns, Final Report from th Univery of Del ware, AICAE, New York (1959) 11. Drickame, HLG., and IR, Beaord, Trans, AICHE, 39, 319-360 (494s) 12. O'Connell ILE, Trane. AICHE, 42, 781-755 (1948) 8. The McCabe-Thiele method can be extended to handle Murphree stage efficiency, multiple feeds, sidestreams, open steam, and use of imterreboilers and intercondensers, 9. For tayed columns, estimates of overall stage efficiency, defined by (6-21), can be made with the Drickamer and. Bradford, (7-42), or O'Connell, (7-43), correlations. More accurate procedures use data from a laborstory ‘Oldershaw column or the semitheoretical mass-transfer ‘equations of Chan and Fair in Chapter 6. 10. Tray diameter, pressure drop, weeping, entrainment, and downcomer backup can be estimated by procedures in Chapter 6 1, Reflux and flash drums are sized by a procedure based ‘on Vapor entrainment and liquid residence time. 12, Packed-column diameter and pressure drop are det ‘mined by procedures presented in Chapter 6, 13, The height of a packed column is established by the HETP method or, preferably, the HTU method. Appli- cation to distillations parallels the methods in Chapter 6 for absorbers and strippers, but differs in the manner in which the curved equilibrium line, (7-47), is handled, 14, The Ponchon-Savatit graphical method removes the assumption of constant molar overflow in the McCabe ‘Thiele method by employing energy balances with an centhalpy-concentration diagram. However, it has been largely supplanted by rigorous programs in process simulators 18, Lockhar. J. and CW. Leggetin KA. Kobe and 13. MeKetta, J Eds, Advances i Perolcane Chemistry and Refing Val. 1, Interscience [New York pp. 325-526 (1958, 1A. Zoiderwog, FD. Veburg ad FAM. Giten, Pro International, ‘Symposium om Dilation, Itivion of Chem, Eng, London. 202-207 (4960) 1S. Gusreax, ME, and ILE, O'Coanel, Chem. Brg. Prog. $4) 232- 237.955). 16 Chan, Hand JR Pair Ind ng Chem. Process Des Dev, 28, 814 819.4988). 17, Fair, JR, HR Nol, and WL, Boles, Ind Eng. Chem, Process Des. Der, 22,5354 (1989), 18, Sakata Mand". Yanag, 1 Chem. # Symp. Sr, 6,221 (1079) 19. Yanagi, and M. Sakata, Ind. Bg. Chom. Process Des. Devel. 2, 7121982} 20. Younger, AM, Chom Eng, 625), 201-202 (1855, 21, Ponchon, M, Tech Modeme 18,20, 55 (1921), 22, Sava R, Arts et Meiers pp. 65 142,178, 241,266,307 (1922), 28, Henley. BY. and ID. Seader. EqullorianStage Separation Oper "ons in Chomacal Beginecring Toba Wiley & Sons, New York (198), 24, Gliveh Ballas Tay, Bulletin 159, Pitz W. GMich and Sons, Dallas (Gram FRI reper of September 3, 1958).STUDY QUESTIONS 74. What equipment is included ina typical dstilaton operation? 72. What determines the operating pressure of a distillation column? 73. Under what conditions docs « distillation column have to operate under vacuum? 74. Why ate distillation columns arranged for countercurrent ow of liquid and vapor? 7S. Why isthe MeCabe-Thiele graphical method useful inthis era of more rigorous, computer-aided algebraic methods used in process simulators? 7.6. Under what conditions does the McCabe-Thiele assumption of constant molar overfiow hold? 717. In the McCabe-Thiele method, between which two lines is the staircase constructed? 78. What is meant by the reflux ratio? What is meant by the boilyp ratio? 79. Whatis the qline and how is it related tothe feed condition? EXERCISES Note: Unless otherwise stated, the usual simplifying assumptions of saturated-liquid eux, optimal feed-stage location, no heat losses. Heady stale, and constant molar liquid and vapor flows apply to cach exercise Section 7. 7A. Differences stripping. Listas many differences between (1) absorption and distillation and (2) stripping end distillation as you can, 72. Popularity of packed columns. Priot tothe 1980s, packed columns were rarely used for distila- tion unless column diameter was les than 2.51 Explain why, in recent years, some tryed towers are being retrofitted with packing ‘snd some new large-diameter columns are being designed for pack ing rather than trys. Detween absorption, distillation, and 73. Use of cooling water ina condenser. ‘A mixture of methane and ethane is subject to distillation. Why can't water be used as a condenser coolant? What would you use? 74. Operating pressure for distilation, 'A mixture of ethylene and ethane isto be separated by distilla tion. What operating pressure would you suggest? Why? 7S. Laboratory data for distillation design, ‘Under wht cizcurnstances would it be advitable to conduct lab oratory or plotplant teste of a proposed distillation? 716. Bomomic trade-off in distillation design Explain the economic trade-of between (ays and ello. Section 7.2 717. MeCabe-Thiele Method, Tithe 50 years following the development by Sorel in 1894 of mathematical model for continuous, steady-state, equiibrumstage distillation, many noncomputerized methods were proposed for solving the equations graphically or algebraically. Today, the only method from that era that remains in widespread use is the Exerises 289 7.40. What aze the five possible feed conditions? TAAL. Inthe MeCabo-Thiele method, are the stage stepped of from the top down or the bottom up? In either caze, when is best, during the stepping o switch from one operating line tothe other? Why? 7.42. Can column be operated a total reflux? How? 7.13. How many stages are necessary for operation at minimum sellux ratio? 744. What is meant by a pinch point? Is i always located atthe feed tage? AS. What is meant by subcooled reflux? How does it affect the amount of reux inside the column? 7.16, Ist worthwhile to pret the feed oa distillation column? 7.A7. Why isthe stage efficiency in distillation higher than that in absorption? 78. What kind ofa small laboratory column is useful for obain- ing plate efficiency data? McCabe-Thiele graphical method, What stibutes ofthis method are responsible for is continuing popularity?” 78. Compositions of countercurrent cascade stages For the cascade in Figure 7.39a, calculate (a) compositions of streams Vqand L by assuming 1 atm pressure, salurateiguid and ‘vapor feeds, and the vapor-liquid equilibrium data below, where ‘compositions are in mole %,(b) Given the feed compositions in cas cade (a), how many stages are required to produce « Vs containing '85 mol alcohol? (c) For the cascade configuration in Figute 7.38, with D = 50 mols, what are the compositions of D and L;? (@) For the configuration of cascade (b), how many sages are required to produce aD of 50 mal alcohol? EQUILIBRIUM DATA, MOLE-FRACTION ALCOHOL x O01 030s 07 09) y 0205 oss 0829s Sat leahot on H3 ‘ om sleahal ay | soko « 8 Figure 7.39 Data for Exercise 78290 Chapter? Disullaion of Binary Mixcores 79. Stripping of air. Liquid air is fed tothe top ofa penfrated-tray reboiledstipper operated at I atm. Sixty % of the oxygen inthe feed i to be draw off in the bottoms vapor product, whieh is to contain 0.2 mol nitogen. Based on the assumptions and equilibrium data below, cal culate: (a) the male % Ns in the vapor fom the cop plate, (the ‘vapor generated in the sil per 100 moles of feed, and (c) the num ber of stages required Assume constant molar overflow equal the moles of feed. Lig uid air contains 20.9 mol% O, and 79.1 mol% Ns, Te equilibrium data (Chem Met Eng. 35, 622 (1928) a atm ae: Mole-Percent —_-Mole-Peveent ‘Temperature, K Nein Liquid Nr in Vapor n35 100.00 100.00 7198 90.00 914 1873 73.0 93.62, Mas 7000 90.31 4033 60,00 5591 8135 50.00 x06 3254 40.00 73.50 8394 30.00 405 85.62 20.00 sos 87.67 10.00 31.00 90.17 0.00 0.00 7210. Using operating data to determine reflux and distillate ‘composition. mixture of A (more volatile) and B is separated ina plate dis: \illation coluasn, Tn two separate tess run with a saturated-liguid feed of 40 mol% A, the following compositions. in mol% A, were ‘obtained for samples of liquid and vapor steams from thee consec- Uutive stages between the feed and total condenser atthe top Molt A Test Tex? Liquid Vapor ‘Liquid 795 680 750680 MHL mo so 80s Mw 619 sla 605 30 Determine the rellux ratio and overhead composition in each cae, assuming tha the column has more than thee sages. TAL, Determining the best distilation procedure. A saturated-liguid mixture of 70 mol% benzene and 30 mol toluene, whose relative volatility is 2.5, is tobe distilled at 1 atm to produce a distillate of 80 mol’ benzene. Five procedures, described below, ate under consideration. For each procedure, calculate and (abulate: (a) moles of distillate per 100 moles of feed, (b) moles of {otal vapor generated per mole of dsl, and (c) mol% benzene in the residue. (d) For each part, constuet ayx diagram. Oa tis indi cae the compositions of the overhead produc, the reflux, and the ‘composition ofthe residue () Ifthe objective is to maximize ota benzene recovery which, if any, ofthese procedures is prefered? “The procedures are as follows: 1. Continuous distillation followed by partial condensation. The feed is sent to the ditec-heated sil pot, rom which the residue is continuouely withdrawn. The vapors eater the top of a hel ‘coiled partial condenser that discharges into a trap, The liq tid is returned (refluxed) tothe stil. while the residual vapor is condensed as a product containing 80 mol% benzene, The molat ratio of reflux fo product is 05. 2. Continuous distillation in a column containing one equilibrium plate, The fed is tent to the drect-heated sll, roms which res due is withdrawn continuously, The vapors Sram the plate enter the top of a helically coiled partial condenser that discharges into a wap. The liquid from the tap is retuned to the plate, while the uncondensed vaper is condensed o form a dsillate contain ing 80 mol% benzene. The molae ratio of reflux to product is 05 ‘3. Continuous distillation in a column containing the equivalent of ‘wo equilibrium plates, The fed is sent tothe dicetcheated sil, fom which residue is withdrawn continuouely. The vapors from the top plate enter the top of a helically coiled partial condenser that discharges into a tap. The ligud from the Cap is returned to the top plate (fetuxed), while the uncondensed vapor is con densed to a distillate containing 80 mol% benzene. The molar ratio of refi to producti 0.5 4. The operation isthe same as for Procedure 3, except that liguid fom the tap is retuened to the bottom plate 5. Continuous distillation in a column with the equivalent of one cquilibrium plate. The feed at its boiling point is introduced on the plate. The residue is withdrawn from the direct-heated still pot. Te vapors from the plate enter the (op of a paral condenser that discharges into a trp, The liguid from the trap is retumed to the plate, while the uncondensed vapor is condensed toa dis tillate of 0 mol% benzene. The molar ratio af reflux to product, 05: 7412, Evaluating distillation procedures A saturated-liguid mixture of 50 mol% benzene and toluene is Aislled at 101 KPa in an apparatus consisting of a sll pot, one the ‘retical plate, anda total condenser, The still pot is equivalent to an ‘equilibrium stage. The apparatus isto produce a distillate of 75 mol % benzene, For each procedure below, calculate, if possible, the moles of distillate per 100 moles of feed. Assume an a of 25 Procedures: (a) No reflux with feed (0 the sill pot.) Feed to the sil pot with reiux ratio = 3.(c) Feed tothe plate with a reflux ratio of 3, (d) Feed tothe plate witha eeux ratio of 3 from a partial condenser. c) Part (6) using minimum reflux. (0) Part () using total reflux 713. Separation of benzene and toluene. ‘A column at 101 kPa isto separate 30 kg/h of a bubble-point solution of benzene and toluene containing 0.6 mase-fracton tol ene info an overhead product of 0.97 mass-raction benzene and ‘bottoms product of 0.98 mast-fraction toluene at a reflux ratio (of3.5. The feeds sent to the optimal tay, and the relax i at satus sion temperature. Determine the: (a) top and bottom products and (8) number of sages using the following vapor-liguid equilibrium data, EQUILIBRIUM DATA IN MOLE- FRACTION BENZENE, 101 KPA. 021 037 051 06% 072 0.79 086 091 095 0.98 x 01 02 03 04 05 06 07 08 09 0:957.14. Calculation of products. ‘A mixture of 54.5 mol’ benzeae in chlorobenzene at its bubble ‘point is fed continuously tothe bottom plate of a column containing {wo equilbcium plates, with a paral reboiler anda total condenser. Sulficient heat is supplied tothe reboiler to give V/F = 0.855, and the reflux ratio /V in the top ofthe column is constant at 050, Under these conditions using the equilibrium data below, what are the compositions of the expected products?” EQUILIBRIUM DATA AT COLUMN PRESSURE, ‘MOLE- FRACTION BENZENE, x 0.100 0200 0.300 0.400 0.500 0.600 0.700 0.800 y 0314 0508 0.640 0.734 0805 9862 0905 0.943, TAS. Loss of trays ina dist ‘A continuous distillation with areux ratio (L/D) of 35 yields a distillate containing 97 wi% B (benzene) anda botioms of 98 wie T (Coluene), Due to weld failures, the 10 stripping plats in the bottom fection ofthe column are ruined, bt the 14 upper reciying plates fe intact. Ii suggested thatthe column sil be wsed, with the feed (Pas saturated vapor atthe dew point, wih F = 13,600 kgfh con- taining 40 wi B and 60 we T. Assuming thatthe plate efficiency remains unchanged at 50%: (2) Can this column sill yield a disl- late containing 97 wt% B? (b) How much distillate is there? (©) What isthe residue composition in mole 7 For vapor-liquid equilibrium data, see Exercise 7.13. tion column, 746. Changes toa distillation operation, ‘A disillation column having eight theoretical stages (seven stages + paral reboiler + total condenser) separates 100 kmol/h of, satuated-Liguid feed containing 50 mol% A into a product of 90, mol% A. The liquid-to-vapor molar ratio atthe top plate is 0.75. The saturaed-liquid feed enters plate $ from the (op. Determine: (@) the bottoms composition. (6) the Z/V ratio in the striping sec Won, and () the moles of bottoms per hou Unknown (othe operators. the bolts holding plates 5, 6, and 7 rust chrough, and the plats fll into the still pot. What is the new bottoms composition? Tis suggested that, instead of returning reflux tothe top plate, an equivalent amount of liguid product ftom another column be used as, reflux. If that product contains 80 mel% A, wht i naw the compo- ston of) the distillate and (b) the bottoms? EQUILIBRIUM DATA, MOLE FRACTION OF A y 019 037 05 062 O71 O78 08 09 096 01 02 03 04 05 08 07 08 09 LAT. Eifect of diferent feed conditions. ‘A istillation unit consists of a partial reboler, 2 column with seven equilibrium plates, anda total condenser, The feed is «50 mol ‘5 mixture of benzene in toluene. Iie desired to produce a disillate containing 96 mol% benzene, when operating at 101 KPa (a) With saturated-tiquid feed fed to the ith plate from the top, cal culate: (1) minimum reflux ratio (Lp/D)ni: (2) the bottoms ‘composition, using a reflux ratio (fg/D) of twice the minimum; and (3) moles of product per 100 moles of fed, () Repeat part (a) fora saturated vapor fed to the fi the top. ate from Bxerises 291 (©) With saturated-vapor feed fod to the reboiler and a refi ratio (L/¥ of 08, calculate: (1) bottoms composition, and (2) moles of product per 100 moles of feed ‘Equilibrium data ate in Exercise 7.13, 78. Conversion of distillation to stripping. A valve-tay column containing eight theoretical plates, a pata reboiler, anda total condenser separates a benzene-toluené mixture containing 36 mol% benzene at 101 kPa. The reboiler generates 100 kmoUh of vapor. A request has been made for very pure toluene. and itis proposed to an this column at «stripper, with the sata rated-lquid feed tothe top plate. exmploying the same boilup atthe sil and returning no reflex to the column. Equilitvium data are given in Exercise 7.13, (a) What isthe minimum feed rate under the proposed conditions, and what is the corresponding composition of {he liquid in the reboiler at the minimum feed? (b) Ata feed rte 25% above the minimum, whal is the rale of production of toluene, fand what ate the compositions in mol of the product and distillate? 7.19. Poor performance of distillation. Filly mol% methanol in water at 101 kPa is continuously dis- tilled in a seven-plate perforated-tray column, with « total con- denser and partial reboier heated by seam, Normally, 100 kml of feed is inoduced on the third plate from the bottom. The over- head product contains 90 mol3 methanol, andthe bottoms 5 mol%. (One mole of reflux is returned for each mole of overhead product. Recently it has been impossible to maintain the product purity in spite of an increase in the rellux rai, The following test data were obtained molt mol alcool 100 3 62 2 53 80 94 = ‘What isthe most probable cause of thit poor performance? Whst further tests would you make to establish the reason fr he rouble? CCould some 90% product be obtained by futher increasing the re- ‘lux rato, while keeping the vapor rate constant? ‘apor-liquid equilibrium data at 1 atm (Chem. Eng. Prog.. 48, 192 (1952)] in mole-fraction methanol are x 00321 00523 0075 0.158 0225 0349 0813 0918 y 0.1900 02940 0.352 0516 0593 0.703 0.918 0.963, 7.20. Bifect of feed rate reduction operation. ‘A fractionating column equipped with a steam-heated partial reboiler and total condenser (Figure 7.40) separates a mixture of 50 mol% A and 50 mol% B into an overkead product contaiing 90 mol% A and a bottoms of 20 mols A. The column has three theoretical plates, and the reboiler is equivalent to one theoretical plate. When the system is operated at L/V = 0.75 with the feed as 2 Saturated liquid to the bottom plate, the desited products are obtained, The steam (o the reboiler is controlled and remains con- stant. The reflux to the column also remains constant. The feed to the column is normally 100 kmol/a, but it was inadvertently cut back to 25 kmolh, What willbe the composition ofthe reflux and the vapor leaving the rebiler under these new conditions? Assume292 Chapter? Distillation of Binary Mixcores rp 3. Fees Steam flow — T cortralar Figure 7.40 Data for Exercise 7.20. thatthe vapor leaving the eebollr isnot superheated, Relative volo ity is 3.0. 721. Stages fora binary separation AN saturated vapor of maleic anhydride and benzoic acid contain ing 10 mol% acid is a byproduct of the manufacture of phthalic anhydride Iis distilled at 13.3 KPa to give a product of 99.5 mol% maleic anfyéride and a bottoms of 0.5 mol, Calculate the number of theoretical plates using an L/D of 1.6 times the minimum using the date below. VAPOR PRESSURE, TORR: Temperaure,"C 1050100200400 Maleic anhydride 78.7 1168 135815591795, Benzoic acid 116 1678 1850 2058 27 7.22, Caleulation of stages algebraically. ‘A bubble-point feed of 5 mol% A in B is tobe distilled to give a Aisilate containing 35 mol% A and a bottoms containing 0.2 mol, ‘The column has a paral reboier and a partial condenser. Ifa = 6, calculate the following algebraically: (a) the minimum number of equilibrium stage: (b) the minimus beilup ratio V/B: (c) the actual ‘number of stages fora boilup ratio equal to 1.2 times the minimum, 7.23, Distillation with a subcooted feed. ‘Methanol (M) isto be separated from water (W) by distillation, as showin in Figure 7.41, The feed is subcooled:g = 1,12. Determine the fed-stage location and the number of stages required, Vapor Tiguid equilibrium data ae given in Exercise 7.19. 7.24, Calculation of distillation graphically and analytically, A saturated-liquid feed of 69.4 mol% benzene (B) i toluene (T) isto be disilled at | atm to produce a distillate of 90 mol% benzene with a yield of 25 moles of distillate per 100 moles of feed. Te feed is sent oa stears-heated reboiler, whee bottoms is withdrawn con ‘unuouely. The vapor from the reboiler goes toa partial condenser and then to a phase separstor that return the liguid reflux tothe reboiler. The vapor fom the separatr. which sin equilibrium with the liquid relux, is sent to « tral condenser to produce distillate. At ‘quilibriu the mole ratio of B to Tin the vapor from the reboiler i 25 times the mole ratio of B to T in the bottoms, Calculate WwW toato Figure 7.41 Data for Exercise 7.23 analytically and graphically the total moles of vapor generated in the reboiler per 100 mo of feed, 725, Operation at total reflux ‘A plant has 100 kimol ofa liguid mixture of 20 mol benzene and $0 mol% chlorobenzene, which isto be distilled at 1 atm fob ‘ain bottoms of 0.1 mol% benzene, Astume « ~ 4.13, The plant has 1 column containing four theoretical plates, a total condenser, a teboier, ad a ceflux drum to collet condensed overhead, A tun i to be made a total reflux. While steady state is being approached, a Sinite amount of dsillate is held ina reflux tap. Wren the steady state i reached, the bottoms contain 0.1 mol benzene. What yield ‘of botoms ean be obtained? The liquid holdup inthe column is neg lipible compared to that inthe reboiler and relux dru Section 7.3 7126, ‘Trays for a known Murphree efficiency. ‘A 50 mol% mixture of acetone in isoprepane isto be distilled to produce a distillate of 80 mol%® acetone and a bottoms of 25 mol. ‘The feed isa saturated liquid, the column is operated with a refx ratio of 05, and the Murphree vapor eliency is $0%. How many crays are required” Assume a total condenser, partial reboiey, st rated-lguid rex, and optimal feed stage. The vaporliguid equi ‘rium data are: EQUILIBRIUM DATA, MOLE- PERCENT ACETONE Liquid 0 26 54 117 207 297 341 440 520 Vapor 0 $9 174 315 456 557 601 687 743, Liquid 639 746 803 865 902 925 957 1000 Vapor 815 870 894 923 942 955 974 1000 727. Minimum reflux, boilup, and number of trays for known cfliciency, ‘A mixture of 40 mol’ carbon disulfide (CS) in carbon tetra chloride (CCL) is continuously distilled. The feed is 50% vaporized (q = 0.5). The product from a total condenser i 95 mol" CS, and the bottoms ftom a partial reboiler is 5 mol% CS, The column, operates with a reflux ratio, L/D, of 4 to 1. The Murphree vapor efficiency is 80%. (a) Calculate graphically the minimum reflux, the ‘minimum boilup ratio from the reboiler, V/B. and the minimumnumber of stages (inchuding the reboiler) (6) How many ways, ace required for the actual columa at 80% Murphree vapor-tray tficiency by the McCabe-Thiele method? The vepor-liquid equilibrium data at columa pressure in tems of CS; mole ¥ 005 01 02 03 04 05 06 07 08 09 y 0.135 0245 0.42 0.545 0.64 0.725 0.79 O85 0905 0.955 728. Reboiler duty for a distillation. {A distillation unit consists ofa partial seboiler, a bubble-cap co! umn, and a total condenser. The overall plate eliciecy is 65%. The feed isa bubble-point liquid of 50 mol% benzene in toluene, which is fed tothe optimal plate. The column is to produce a distillate con- taining 95 mol benzene anda bottoms of 95 mol% toluene. Celeu- late for an operating pressure of I atm the: (2) minimum reflux ratio {Duis (0) minissum number of actual plates; (©) number of actual plates needed fora reflux ratio (L/D) of 50% more than the minimum; (4) kgfb of distillate and bottoms, if the feed is 907.3, Kesh; and (¢) saturated steam at 273.7 kPa required in kg/h for the reboler using the enthalpy data below and any assumptions nec- asay, (0) Make a rigorous enthalpy balance on the reboies, using the enthalpy data below and assuming ideal solutions. Enthalpies se in Btu/Tbmol a reboiler temperature hi hy benzene 4900 18,130 toluene 8.080 21,830 ‘Vaporiquid equilibrium data are given in Exercise 7.13, 7.29. Distillation of an azeotrope-forming mixture. ‘A continuous distillation uni, consisting ofa perforated-teay col- ‘uma witha partial eeboiler and a total condenser, is to be designed to separate ethanol and water at 1 alm, The bubble-point feed con= tains 20 mol% alcobol The distillate is to contain 85 mol% alcohol, and the recovery isto be 97%. (a) What isthe molar concentztion ofthe bottoms? (b) What is the minimum value ofthe reflux ratio, LV. the teflux ratio L/D, and the boil ratio V/B? (¢) What isthe minimum number of theoretical stages and the number of actual plates ifthe overall plate efficiency is 55% (@) Ifthe L/V is 0.80, how many acal plates wil be required? ‘Vaporliguid equilibrium for ethano!—water at 1 atm in terms of role fractions of ethanol ae Und. Eng. Chem. 24, 881 (1932)} x y ne | = y Te dois 0170 95:50 | 03273 05826 8150, oorat 0891 89.00 | 03965 06122 $0.70, 0.0955 04375 86:70 | 05079 0.46564 1980, 0.1238 0470a 85.30 | 05198 0.6599 79.70, o1sst oss sto | 5732 ossar 79.30, 02337 sais 2.70 | 06763 07385 78.74, 02508 05580 82.30 | 07472 O78I5 7841 sss 089437815 7.30. Multiple feeds and open steam. Solvent iso be separated from water by dilation to produce 4 distillate containing 95 mol% A at a 95% recovery. The feed is Exercises 293 available in two saturatedliqid streams, one containing 40 mol ‘A aad the other 60 mol A. Each steam will provide 50 kmolh of Component A. The « it 3 and since the lesr-volatile component is water iis proposed to supply the necessary reboiler heat in the form of open steam. Por the preliminary design, the operating reflux ratio, L/D, is 133 times the minimum, using a total condenser. The overall plate efficiency is estimated :0 be 70%. How many plates ‘will be required, and what willbe the bottoms composition? Deter- rine analytically the points necessary to locate the operating line. ach feed should enter the column at its optimal locaton, 731. Optimal feed plate location. [A sauratedliquid feed of 40 mol% r-hexane (EH) end 60 mol% reoctane isto he separated into a distillate of 95 mol Ff and a bot toms of $ mol% H. The reflux ratio L/D is 0.5, and a cooling coil, submerged in the liquid of the second plate from the top removes suiceat heat to condense 50 mol’ of the vapor rising from the third plate down from the top. The x-y data of Figute 7.37 may be ‘sed, (a) Derive the equations needed to locate the operating lines. () Locate the operating lines and determine the required numberof ‘theoretical plates if the optimal feed plat location i used. 732. Open steam for alcohol distillation, One hundred kiolM of a saturted-lguid mixture of 12 mol% ‘hy! alcohol in water is dill continuously using open steam at 1 atm introduced diecty to the bottom plate, The distillate required {is 85 mol% aleoho, representing 90% recovery of the aleobol inthe fcod. The reflux is saturated liquid with L/D = 3. Feed is on the optimal stage. Vapor-liquid equilitsium data are given in Exercise 1/20, Calculate: (a) the steam requirement in kml; (b) he number of theoretical stages; (c) the optimal feed sage; and (@) the mini sium reflux ratio, (L/D 7.33. Distillation of an azeotrope-forming mixture using open steam, ‘A 10 mol% isopropano!-in-water mixture tits bubble point isto be distilled at 1 atm fo produce a distillate containing 67.5 mol% isopropanol, with 98%: recovery. Ata reflux ratio L/D of 1.5 umes the minimum, how many theoretical sages will be required: (a) if partial reboileris used? (b) if no reboiler is used and saturated steam at 101 KPa is introduced below the bottom plate? (&) How many stages are required at total refux? “Vapor-liguid data in mole-fraction isopropanol a 101 KPa ate TG 9300 M02 9212 129 som 016 9028 1ST y 02195 04620 05242 0.5686 05928 06821 07821 09160, X _OO1I8 0841 04974 03496 04525 06704 0.7693 0.94 Notes: Composition ofthe azcotrpe is x point of szeotrope = 80.22°C, 74. Comparison of partial reboiler with live steam. An aqueous solution of 10 mol% isopropanol at its bubble point, Js fed tothe top ofa stripping column, operated at I atm, to produce 8 vapor of 40 mol% isoprepanol. Two schemes, both involving the same heat expenditure, with V/F (moles of vaporimole of feed) = (0.246, ae under consderstion. Scheme I uses a paral reboiler et the bottom of & stripping column, with steam condensing inside a cloted col. In Scheme 2, live steam is injected directly below the bottom plate, Determine the numberof stages required in each case, Eguiibrium data are given in Exercise 7.33, = 0.6854. Boing feed stages for two feeds © optmal-stage locaton for each feed and the num- ber of theoretical stages required for the distillation separation294 Chapter? Distillation of Binary Mixcures LD = 12D, ‘eat a ps Feed, 60 mole ts vaporizes 26 malts acti aid steams Figure 7.42 Data for Exercise 7.35. shown in Figure 742, using the following vapor-tiquid data in mole actions of wat WATER (WYACETIC ACID (A), 1 ATM. Xw 0.0055 0.053 0.125 0.205 0297 0.510 0.689 0.803 0,959 yy 00112 0.133 0.240 0.338 0437 0.630 0.751 0.866 09725 7.36. Optimal sidestream location, Determine the number of equilibrium stages and optimal-stage locations for the feed and liquid sidestreams ofthe distillation pro cess in Figure 7.43, assuming that methanol (M) and ethane! (B) orm aa ideal solution, 731: ‘A mixture of n-heptane (H) and toluene (T) is separated by extractive distillation with phenol (P). Distillation is then used to recover the phenol for recycle, as shown in Figure 74a, where the ‘mall amount of n-beptane inthe feed is ignored. For the conditions shown in Figure 7.4da, determine the number of stages required, Note that heat must be supplied tothe reboler ata high temperature Use of an interreboiler isle, DTZ (UD) gy SO MORE amelh yu R Liguid 15 kametn eo [sis seam 80 mole Be mole Figure 7.43 Data for Exercise 7.36. because of the high boiling point of phenol, which causes second- law inefficiency, Therefore, consider the scheme in Figure 7.44b, ‘where an interreoiler, located midway between the bottom plate and the feed stage, provides 50% ofthe boilup used in Figure 74a, ‘The remainder of the boilup is provided by the reboier. Determine the stages tequited for the case with the interrebolee and the tem perature ofthe interreboler stage. Unsmoothed vaporliquid equii- Diu data at I atm [Trans. AICHE, 41, 555 (1945)] at x ye |r ye ons 03410 17270 | 06512 0.9260 12000 0082 05120 13940 | 07400 0.9453 119.70 01248 06250 14940 | osorz 0.9545. 115,60 02190 0785014220 | ossaa 0.9750 112.70 0.2750 08070 13380 | 09394 09851 113.30 oaoso 08725 12830 | 0970 0.9948 111.10 04800 osv01 12570 | 09910 0.9980 11.10 05898 09159 12220 | 09973 0.9993 110.50 a SP saat kgmolty “ kgmolin no Esco c Revoiler Reboiler “srenel steam se ari Figure 7.44 Data for Erercise » 731738. Addition of intercondenser and interreboi A distillation columa to separate n-butane from n-pentane was recently put on line in a refinery. Apparently, there was a design err because the column did not make the desied separation, as shown below [Chem Eng. Prog, 61(8), 79 (1965) Tis proposed to add an intercondenser inthe rectifying section to generate more reflux and an interreboiler inthe stripping section to prodice additional boilup. Show by use of « MeCabe-Thile dia: 7am how this might improve the operation, Design Actual Specification Operation Mol nC in distate 026 1349 Mol nin bottoms 046 428 7.39. Use of Kremser method to exte rmcthod. Chlorobenzene production by chlorination of benzene produces two close-boiling isomers, para-dichlorobenzene (P) and ortho= dichlorobenzene (0), which ae separated by distillation. The feed to the columa is 62 mol% P and 38 mol% 0, The pressure atthe bottom and top of the column are 20 psia and 15 psi, respectively. The disillate isto be 98 mol% P, and the bottoms 96 mel% O. The feed is slightly vaporized with y = 0.9. Caleulate the number of equilibrium stages requited for an R/Ryun = 1.15. Base your ealcula- tions on an average a = 1.163 obtained asthe arithmetic average between the column top and column bottom using vapor pressure data and the assumption of Raoult's and Dalton’s laws. The McCabe-Thiele construction should be supplemented at the two ends by ute ofthe Kremser equations as in Example 74 ‘MeCabe-Thiele 7.40. Air separation using a Linde double column. ‘0, and N; are obtained by distillation of ar using the Linde dou- ble column, shown in Figure 745, which consists of «lower column at elevated pressure surmounted by an atmospheric-pressure co!~ ‘un, The boiler of the upper column is also the reflux condenser for both columns. Gaseous air plus enough liquid to compensate for heat leak into the column (more liquid if liqud-oxygen product is withdrawn) enters the exchanger al the base ofthe lower column and condenses, giving up hea othe biling liquid and ths supply ing the columa vapor flow. The liguid air enters an intermediate Point in this column. The rising vapors are parially condensed to form the reflux, and the uncondensed vapor passes to an outer row of tubes and is totally condenced, the liquid nitrogen collecting in fan annulus, as shown, By operating tis column at 4 to atm, the liguid oxygen boiling at 1 atm is cold enough to condense pure nitrogen, The liquid in the bottom of the lower column contains bout 45 mol’ O, and forms the feed for the upper column, This double column can produce very pure O> with high O. recovery, and relatively pure N;, Ona single McCabe—Thiele diagram—using equilibrium lines, operating lines, ylines, a 43° line, stepped-off stages, and other illsteative aids—show qualitatively how stage requirements can be computed Section 74 7.41. Comparison of tray efficiency. Performance data for a distillation tower separating a 50/50 by weight pecent mixture of methanol and water areas follows Feed rte Ted 45,438 Ib; feed condition pressure ‘bubble-point liquid at Exerises 295 ic ‘reat “eam + rvotovae Figure 7.45 Data for Exercise 7.40. Wee methanol in dsillate = 1.00; Reflux ratio = 0.947; relax condition — saturated liquid ‘Boilup ratio = 1.138; pressure in reflux drum = 14.7 psia: 95,04; Wish methanol in bottoms Type condenser = total: type reboiler = partial Condenser pressure drop = 0.0 psi; tower pressure drop 0 psi Trays above feed tray = 5; trays below feed tay = 6: Total trays = 12: tray diameter = 6 ft: type tray = single-pass sieve ways flow pathlength = 50.5 inches; Weir length = 42.5 inches; hole afea = 10%; hole size = 3/16 inch Weir height Viscosity of feed 20 dynelem: Surface tension of bottoms 2inehes: ay spacing = 24 inches: (0.34 P; Sucface tension of distil 58 dynefem; Temperature of top tray = 154°F: temperature of bottom tcay 207 F ‘Vapor-liquid equilibrium data at column pressure in mole fac- tion of methanol aze 0412 0156 0.379 0578 0.675 0.729 0.792 0.915 x 000565 0.0246 0.0854 0.205 0315 0398 0518 0.793 Based on the above data: (2) Determine the overall tray efficiency sssuming the reboiler is equivalent to a theoretical stage () Estimate the overall tay efficiency from the Drickamer-Bradford correlation. (¢) Estimate the overall ray ellcieny frm the O’Conne correlation, accountng for length of flow path (2) Estimate the Mu Pree vaporray eiciency by the method of Chan and Fit296 Chapter? Disllation of Binary Mixcores igure 7.46 Data for Exercise 7.43 742. Oldershaw column efficiency. For the conditions of Exercise 7.41, a laboratory Oldershaw col- mn measures an average Murphree vapor-point efficiency of 65%. Eximate Eyvand E,, Section 75 743, Column diameter Figure 7.46 shows conditions forthe top tay ofa distillation col uma, Determine the columa diameter at 85% of flooding for a valve ray, Make whatever assumptions nesessary. 744, Column sizing. 47 depicts a propylene/propane distillation. Two seve uray columns in series are used because a 270-ray column poset stractural problems, Determine column diameters, tay efficiency using the O° Connell correlation, number of actual trays, and column eights 748. Sizing a vertical lash drum. Determine the height and diameter of a vertical ash drum for the conditions shown in Figure 7.48. ner o 1281 Ibmoline C3" igure 7.47 Data for Exercise 7.44 Tomah ne 1878 Torvaln ny Mea meh 2258 Figure 7.48 Data for Exercise 7.45. 274. Ibmelin dence 0. timoln monachior oer 7.46, Sizing a horizontal ash drum. Determine the length and diameter of a horizontal reiux drum forthe conditions shown in Figure 7.49 TAT. Possible swaged column, Results of design calculations for @ methanol-water distillation operation are given in Figure 7.50. (a) Calculate the column diame ter at the top and at the bottom, assuming sieve trays, Should the column be swage? (b) Calculate the length and diameter of the br ‘zontal tefox drum. 748, ‘Tray calculations of flooding, pressure drop, entrain- ‘ment, and froth height. For the conditions given in Exercise 7.41, estimate forthe top tay andthe bottom tay (a) % of loading; () ery pressure deep in pi; (©) whether weeping will occur, (d) entrainment rate; and (6) feots height in the downcome, 749, Possible retrofit to packing. Exercise 7.41 is increased by 30%, with conditions—except for tower pressure drop—remaining the same, estimate for the top and bottom trays: (@) % of flooding () tray pressure drop in pi; (6) entrainment rate; and (@) froth, eight in the downcomser. Will the new operation be acceptable? IF not, should you consider a reroit with packing? If zo, should both sections of the column be packed, or could just one section be packed to achieve an acceptable operation? ne) 0.01 Figure 7.49 Data for Exercise 7.46. Seiurated 1.91 molt methanol Figure 7.50 Data for Exercise 7.47Section 7.6 750. HETP calculation, "A mixture of benzene and dichloroethane is used to obtain HET data fora packed column that contains 10 of packing and operates adiabatically at atmospheric pressure. The liquid is charged tothe reboiler, and the column is operated at total efx until equiibsium is established. Liquid samples from the distillate and roboiler give for benzene Xp = 0.653, xg = 0.298. Calculate HETP in inches for this packing. What aze the limitations of sing thie calculated value for design? Data for x-y at J atm (in benzene mole fractions) are x 010 y oul 020 030 040 050 060 970 080 090 022 0325 0426 0526 0.6825 0.720 0815 O91 751. Plate versus packed column. ‘Consider a dilaton column for separsting ethanol from water at L atm, The feed is @ 10 mol% ethanol bubble-point ligui, the bottoms contains I mol% ethanol, and the distillate is 80 mol% ethanol, R/Rrn~ 15. Phase-equilibrium data ate given in Exercise 1.29, and constant molar overliow applies. (a) How many theoretical plates ate required above and below the feed if a plate column is ‘sed? (b) How many tansfer units are required above and below the feed if a packed column is used? (c) Assuming the plate efficiency is 80% and the plate spacing is 18 inches, how high is the plate column? (d) Using an Hog, valve of 1.2f, how high isthe packed columa? (@) Assuming that HTU data ae available only on the ben- zene-foluene system, how would one go about applying the data to obtain the HTU for the ethanol-water system? Table 78 Methanol Water Vapor-Liguid Equilibrium and Eathalpy Data for 1 atm (MeOH Exercises 297 7.52. Design of random and structured packed columns. Plant capacity for the methanol-water distillation of Exercise 7.41 is to be doubled. Rather than installing « second, identical ‘eayed tower, a packed column is bein considered, This Would have ‘feed location, product purities, reflux ratio, operating pressure, and capacity identical tothe preseat wayed tower. Two packings are being considered: (1) 0-mum plastic Nor-Pac rings (a random pack ing), and (2) Montz metal B1-300 (a structured packing) ‘For each ofthese packings, design a column to operate at 70% of fooding by calculating for cach section: (a) liguid holdup, () coenn diameter (©) Hoe (8) packed height, and (@) pressure drop. What are the advantages, if any, of each ofthe packed-columin designs over a second trayed tower? Which packing, if either, i preferable? 7.83. Advantages ofa packed column. For the specifications of Example 7.1, design « packed column sing S0-mm metal Hilow rings and operating a 70% of Nooding by calculating for each section: (a liquid holdup, (b) column diame= ter (©) Hos (@ packed height, and (¢) pressure drop. What ate the advantages and disadvantages of a packed column as compared toa leayed toner fr this service? Section 72.7 7.84, Use ofan enthalpy-concentration diagram. Figure 7.37 is an enthalpy-concentration diagram for mherane (€D, and moctane (0) at 10] KPa. Use thie diagram to determine the: (a) mole-raction composition ofthe vapor when a liquid con- taining 30 mol% His heated from Point A to the bubble-point tem- () energy required to vaporize 60 mol of & Methyl Alcohol) Enthalpy above O° Bllbmol Solution Vapor Liquid Equilibrium Data Saturated Mol Mol% Saturated Vapo Liquid MeOH in MeOH Temperature, yorx re by re hy Liquid Vapor “c ° 100 20,720 100 3240 o o 100 5 989 20,520 928 3070 20 4 964 10 917 20,340 817 2950 40 935 15 962. 20,160 Bad 2850 60 912 20 oa, 20,000 817 2.160 40 v3 30 916 19,640 780 2,620 100 si7 40 382 19310 3 2.540 150 sa 50 49 18970 BA 470 200 si7 0 309) 18,650 m2 2,410 300 780 70 166 18310 3 2,370 400 753 80 n2 17.980 6 2.330 soo 1 90 os 17.680 65.0 2.290 00 2 100 ots 17390 665 2.250 709 03 x00 678 500 550 950 50 1000 os: Source: 16, Dunlop, "Vapor-Ligud Equliviam Data." MS. thesis, Brooklyn Polyechne Ts298 Chapter? Distillation of Binary Mixcures ‘mixta iilly at 100°F and containing 20 mol H (Point G); and (6) compositions of the equilibrium vapor and liquid resulting from part), 7.58. Use ofan enthalpy-concentration diagram, ‘Using the enthalpy-conceniration diegram of Figure 7.37, deter mine the following fora mixture of n-hexane (ID and n-actane (O) at J atm: (a) temperature and compositions of equilibrium liquid and ‘vapor resulting ftom adiabatic mixing of 950 Ibfh ofa mixture of 30 smol% Hin © at 180°F with 1,125 Ibh of a mixture of 80 mol H in (© at 240°F; (b) energy required to partially condense, by cooling, a vixtae of 60 mol% Hin O ftom an inital temperate of 260°F to 1200°F. What are the comporitions and amounts of the resulting ‘vapor and liquid phases per Ibmol of original mixture? (c) If the ‘vapor from part (b) is further cooled to 180°, determine the compo- sitions and relative amounts ofthe resulting vapor and liquid 7.56, Plotting an enthalpy-concentration diagram for disilla- tion calculations. ‘One hundeed Ibmolh of 60 mol% methanol in water at 30°C and 1 aim is to be separated by distillation into a igui distillate contain ing 98 mol% methanol and a bottoms containing 96 mol% water. Enthalpy and equilibrium data for the mixture at | atm are given in Table 7-8. The enthalpy of the feed mixture is 765 Bruflbmol. {@) Using the given data, plot an enthalpy-concentraton diagram, () Devise a procedure to determine, from the diagram of part (a). ‘he minimum number of equlibuium stages forthe condition af total reflux and the required separation. (c) From the procedure devel- ‘oped in pat (b), determine Nin. Why isthe value independent of the feed condition? (é) What are the temperatures of the distillate and bottoms?