New York |v DISTILLATI ON Matthew Van Winkle Profeanor of Chemical Engineering The Unteeratty of Texan McGraw-Hill Book Company St.Louis San Francisco Toronto London — SydneyDISTILLATION Copyright © 1967 by MeGraw-Hli, Ins, All Rights Revered Printed in the United Siater of Amica. No part of thi pb Photocopying. recording, of otherwise, without the priot writen Perminion ofthe publiser, Library of Congres Cellog Card Number 6113203 07.0671956 8910111219 141516 XPKP Tés21098,Preface This book was written to serve as a guide to fractional distillation desigo. ‘The attempt has been made to treat the subjeet matter in such a manner that the “why” as well as the "how" is clearly delineated, In addition, design ‘methods in common use are given along with working-size charts so that a process feactionater design ean be made with a minimum of cross-reference consultation with other Books and source material. Recommended design methods foreach of the principal types of plate design are given, accompanied by the recommended limits of the design variables, Distillation in general has been the subject of much research and this has resulted in a large accumulation of literature on the subject, Obviously, in one book, itis impossible to incorporate even a minimum of discussion of allof the information available. Therefore, the selection of material covered rust lie with the author, and the selection has been made on the basis of the author's industrial, consulting, and academic experience, with regard to both the understanding of the subject and the production of workable and economical designs. It is hoped that the book will serve as a text where the unit operation of distillation to be taught and, certainly, asa reference for practicing engineers who are concerned with designing distillation equipment. “The many design equations given here can be programmed and used with various machine-computation methods for fractional distillation design ‘Also, many ready-made programs for such caleulations and some very good books on the subject are available elsewhere. As an aid to those who wish to uiilize machine computation, available literature on the subject has been referenced and included in the text. ‘The author wishes to express appreciation to former students who are in the field of industrial distillation design and engineering for their suggestions, to present graduate students who have given constructive suggestions re garding the presentation of the subject matter, and to friends in industry ‘and in the academic field who have given helpand encouragement inthis work. Matthew Van WiaklePART 1 CHAPTER | CHAPTER 2 2A. 22. 23. CHAPTER 3 3 32. 33. 34. 33. Contents Preface ix Introduction 1 VAPOR-LIQUID EQUILIBRIA 6 PHASE EQUILIBRIA—BINARY SYSTEMS 7 Rules and Laws Describing Equilibrium 7 Vapor-Liquid Equilibria—Binary Systems 16 Prediction of Binary Vapor-Liquid Equilibrium Data from Pure-component Properties 30 Consistency Tests for Binary Vapor-Liquid Equilibrium Data 31 Problems 39 Nomenclature 42 References 43 MULTICOMPONENT VAPOR-LIQUID EQUILIBRIA 46 Ideal Multicomponent Systems 46 Ileal Ternary Systems 49 Nonideal Vapor-Liquid Equiibria—Multicomponent Systems 35 Appendix: Relationship between Liand Coull and Bonham Vapor-Liquid Equilibrium Relations 117 Problems 121 Nomenclature 123, References 124 COMPLEX SYSTEM VAPOR-LIQUID EQUILIBRIA 127 ‘True-boiling-point Curve 127 ASTM or Engler-type Distillation 129 Equilibrium Flash Vaporization (EFV) Curves 130 Interrelation between EFV, TBP, and Engler Distillation Curves 131 Determination of EFV Curves—Superatmosphieric Pressure 132xii contexts 36. 37. CHAPTER 4 4.1 42. PART 2 CHAPTER 5 Suh 52. 33. 54. 58. CHAPTER 6 6A 62. CHAPTER 7 my 72. 73. Atmospheric and Subatmospheric EFV Relationships from ASTM and TBP Data 136 Characteristics of Equilibrium Vapor and Liquid 153 Nomenclature 157 References 158 EQUILIBRIUM AND SIMPLE DISTILLATION 160 Equilibrium Distillation 160 Differential Vaporization—Condensation 170 Problems 187 Nomenclature 188 References 189 FRACTIONAL DISTILLATION 192 GENERAL CONSIDERATIONS 193, Reflux 194 ‘The Fractionating Column 195 ‘The Contact Stage—Plate or Tray 196 Design Considerations in Fractionating Processes 197 Equilibrium Stages—General Considerations 202 Problems 243, Nomenclature 245 References 247 BINARY FRACTIONATION—NUMBER OF EQUILIBRIUM STAGES. 248 Analytical Determination of Number of Equilibrium Stages 249 Graphical Methods 256 Problems 280 Nomenclature 284 References 285 TERNARY AND MULTICOMPONENT SYSTEM FRACTIONATION—NUMBER OF STAGES 286 Preliminary Calculations 287 Design Procedure 290 Short-cut Methods 30314. 78. CHAPTER 8 8.1 PART 3 CHAPTER 9 9. CHAPTER 10 to. 10.2, CHAPTER I Wa. n2 13. conrests xiii Computer Rigorous Calculation of Number of Theoretical Plates 304 Graphical Evaluation of the Number of Theoretical Stages for Ternary Systems 321 Problems 329 Nomenclature 330 References 332 COMPLEX SYSTEM FRACTIONATION—ESTIMA- TION OF NUMBER OF EQUILIBRIUM STAGES 334 Pseudocomponent Design Method 336 Problems 376 Nomenclature 377 References 378 AZEOTROPIC AND EXTRACTIVE FRACTIONAL DISTILLATION 380 TECHNIQUES OF SEPARATION OF AZEOTROPES AND OTHER DIFFICULT TO SEPARATE MIXTURES BY FRACTIONAL DISTILLATION 381 Techniques for Separation 382 Nomenctature 389 SELECTION OF ADDITION AGENTS FOR AZEOTROPIC AND EXTRACTIVE DISTILLATION 390 Relative Volatility as an Economic Factor 390 Prediction of Vapor-Liquid Equilibrium Data 392 Nomenclature 431 References 432 AZEOTROPIC AND EXTRACTIVE FRACTIONAL, DISTILLATION 434 Mechanism of Relative Volatility Change 434 Choice of Entrainers or Solvents 436 Selection of an Azeotropiec or Extractive Process 437xiv contents 14 Is, 16 PART 4 CHAPTER 12 121, 122. 123. 124, 125, CHAPTER 13 Ba. 13.2 133. Ba, 13s. CHAPTER 14 14, 142. 143, 144 Design of an Azeotropic Distillation Process 437 Design of an Extractive Distillation Process 446 Solvent Recovery 469 Problems 473, Nomenclature 474 References 475 FRACTIONATING PLATE AND COLUMN DESIGN 478 FRACTIONATION DEVICES—EQUILIBRIUM STAGES 479 ‘The Bubble-cap Plate 480 Perforated Plate 488 Variable-orifice Perforated Plates 492 Other Equilibrium-stage Trays 492 Plate Components 494 References 501 PLATE AND COLUMN HYDRAULICS AND EFFICIENCY 502 Flow-energy Losses in Plate and Column 502 Evaluation of Pressure Drop and Liquid Backup in Downcomers 506 Flow Conditions Causing Inoperability $21 Efficiency Evaluation and Prediction $33 Prediction of Efficiency 550 Problems $61 Nomenclature 562 References 566 PLATE-FRACTIONATING-COLUMN, DESIGN METHODS 570 General Column Design S71 Bubble-cap Tray Design 575 Perforated-tray and Column Design 584 Value-tray Design Methods $98 Problems 600 Nomenclature 600 References 602CHAPTER 15 PACKED COLUMN DESIGN 604 15.1. Packings 604 15.2. “Pseudoequilibrium Stage” Devices 613, 15.3, Uses of Packed Columns 615 154. Differential Mass Transfer 616 15.5. Empirical Design Relationships 620 15.6. Capacity Design of Packed Towers 630 15.7. Design of Packed Towers for Distillation 638 Problems 643 Nomenclature 644 References 646 Appendix A: Tables 649 Appendix B: Vapor Pressures at Various Temperatures 653 Name Index 669 Subject Index 675Toto Necrete 8 fe oe Introduction Fen gy ‘The use of distillation as separation and purification process is very old, Egloff and Lowry [8]t and Underwood [17] delved into the eatly historical writings pertaining to alchemy and science of the past to find where the process of distillation originated and when, and how it developed to its present state of use, Through the writings of early historians (Hoefer (101, Kopp [11], Bertholet (2) it was indicated that the frst recorded description of a batch distillation occurred in Cleopatra's time in Egypt around 50 n.c. However, because of other, older historical descriptions of products essential cils, perfumes, medicines, beverages, ee, it ean be deduced that distillation in some form was known probably as much as 1000 to 2000 years before that time, Fresh water was produced by distillation of sea water by using @ sponge as a condenser around 300 4.p., and about the same time turpentine was distilled fom rosin oil and condensed in a wool mat. ‘A simple condenser consisting of a long tube exposed to the air and leading toa receiver was described in the fourth century, and in the eleventh century the fist record of aleohol by distiliation was made. ‘The beverage Alcohol process became the frst industrial distillation process during the period between the eleventh and fourteenth century, and it only became Eigoificant when the problems of sealing the joints in the still and devising a water-cooled condenser were solved, This was necessary to obtain renson- able yields of products containing high alcohol contents Inthe early sixteenth century distillation was being applied to separation and recovery of alcohol, water, vinegar, essences, oils, and numerous other products. Books appeared on the subject of distillation in the enrly part of the sixteenth century, Brunschwig (4] and Andrew {1] included information on distillation processes developed up to that time, which essentially covered small-scale batch distillations having practically no reflux but using steaight some eases, coiled tubes cooled with water for a condenser. Lonicer 13} deseribed an sir-cooled still head with water-cooled condenser. 1. is interesting to note that Libavivs [12] described a still provided an air-cooled head which enabled simultaneous removal of five side streams. 1 Numbers In beackets pertain to references given at end of each chapter.2 wrnovyction Many sfigenpocer today conader the ale steam topping ealann lo bbe a modern invention. In 1624 Donato é'Eremita (7] described a batch still with a fractionating sill head, and Libavius (12] described batch sills with reboilers and factionating heads. Boyle (3) distilled wood alcohol aud vinegar and recovered various factions according to their boiling points “This was probably the first analytical dstlation, Up until the nineteenth century the stills were of the differential batch type with litle reflux. They were small, 18 to 30 in. in diameter, and 3 to 4 ft high including accessories. The yields were apparently reasonably high and the products usable. Cooper [6] described the “modern” distilling devices used in his time, [Near the end of the eighteenth century innovations were introduced which advanced the state of the distillation act measurably. In 1800 Rumford (15) used open steam a5 a heating agent. The forerunner of the preseatday bubble cap was invented in 1822 by Pervier (14). These caps were used as contact devices for steam introduced below the plate on which the caps were placed, wine was circulated around the caps on the plate, and the steam “ ripped the aleohol from the wine. In 1830 Coffee {S] developed a con- Ainuous still which used perforated plates, feed preheat, and internal reflux. ‘As an interesting sielight, in 1822 it was estimated that approximately 10,000,000 gal of iit liquor were made in the British Isle while only 3,000,000 gal were being taxed. Distillation obviously was an art over the period of its early development, which encompassed about 3500 to 4000 years. The invention of bubble caps, perforated plates, water-cooled condensers, reflux, feed preheat, and the adaptation of the process to continuous operation took place over the telatively short period of 300 years, During the entire period, however, there was apparently noaitempt (osystematize or apply quantitative principles to distillation processes Late in the nineteenth century Hausbrand {9] and Sorel (16) introduced the fist recorded quantitative mathematical discussions applied to frac- tionating sill design. Sorel developed and applied mathematical relations to the fractional separation of binary mixtures~primarily those comprised of alcohol and water, and incorporated such considerations as variable boilup and overflow, variable molal enthalpy, heat losses (rom the column in his heat balances, compositions, rates, reflux, and pressure effects. ‘Thus atthe beginning ofthe present century, Sorel and Hausbrand evolved the basic relations upon which all modern distillation design calevations are based, and all methods used today are expansions and modifications of those original methods The rationalization and organization of mathematical applications to process fractionation design pointed up the necessity for study in several Felated areas,” These were the thermodynamics of vaporiquid equilibria ~mrmopuction 3 and phase behavior in binary and multicomponent nonideal systems; con tact stage improvement in design to obtain the best possible approach to an equilibrium stage; plate and column hydraulics study and the study of the variables which affect efficiency and operability of the fractionating column; and the rationalization of design methods to enable optimization of fractionating process design from the standpoint of economics, ‘These subjects form the basis for the discussion in the following chapters, REFERENCES 1, Andrew, L.(eanslaog): “The Vertuose Boke of Disylaeyon of the Waters ofall Manner of Horbes,” Drunshuig, 1527. 2. Berthole, D.: "Histoie des seiences—h Ocigincs de Fatchemie,” Paris, 185, 3. Boyle, R.: "Philosophical Works.” vol. 11, London, 1738. 4. Brunschwig, IL: “Liber de arte desilland de sieplicibus," Strasburg 1500; de are detllandi de composition,” Strasburg, 1507 S. Coffee, A: British Patent No, $974, 1830, 6. Cooper, A.: "The Complete Distiller," 24 ed, London, 1761 7. Donate ¢ Eremita: “Del elixir vitae,” Naples, 1624 ® 9. chémie au moyen age,” Pari, 1893; “Lee ier Eplofl, G., and C. B. Lowry: Ind. Eng. Chem, 21:920 (1929). ie Wirkungawsie der Reclfer und Disilin Apparate.” Betln, Haurbrand, B. 1493, 10, oeter, Fu “Histoire de fa chémiec" Pais, 1866, 11, Kopp, Ha: “Geschichte der Chemie,” Braunschweig, 18 12. Libavioe, Avs “Alehymia.” Frankfurt, 1606 13, Lonicer,A.t "KOnsliche Conterfeytunge,” Frankfort, 1573 1M Perrier, Sie A.: Drtsh Patent No, 4694, 1822 15, Rumford, B.T.: "Essays, Poitiel, Eeonomical 1902, 16, Sore, B.: “La Rectifeation de Valcoo!,” Pa industriel,” Pars, 1899, 17 Underwood, A.J. Vai Trans Itt, Chem. Engr (London), 13341935, .nd Philosophical,” vol. 11), London, 494; “Disillation et rectientionpart Vapor-liquid equilibriachapter Phase equilibria-binary systems In the distillation process the separation of a mixture of materials to obtain one or more desired products is achieved by selection of conditions of temperature and pressure so that at least a vapor and a liquid phase coexist and a difference in relative concentration of the materials to be separated in the two phases is attained. When the two (or more) phases are in a state of physical equilibrium, the maximum relative difference in concentration of the materials in the phases occurs, Therefore, attainment of equilibrium conditions is desirable in the distillation process, and most design methods use equilibrium as one of the boundary conditions for quantitative design calevlations Because of this, equilibrium data must be available for such caleulations tobe made, Such data may be obtained from technica iterature, they may be determined experimentally, or they may be obtained through the use of correlations, theoretical or empirical, relating pure-componcrt and mixture properties. 1.1. RULES AND LAWS DESCRIBING EQUUABRIUM A number of quantitative rules and laws have been devised to describe and define systems in a state of physical equilibrium. Most thermodynamics and physical chemistry books give detailed development of the rules cited here and, therefore, complete discussions will not be included,8 VAPOR:LIQUID EQUILIBRIA Gibbs Phase Rule J. Willard Gibbs [16] described equilibrium mathematically in terms of the number of components, the number of phases coexisting, and the con- ditions necessary to establish equilibrium. In its simplest form, representing a system at a definite pressure and temperature, the phase rule is expressed as follows: VaC+2-P (quay where V = the degrees of freedom or degrees of variance of the system ia terms of the independent variables, coricentration, temperature, and pressure, which must be fixed in order to define the system at equilibrium C= the number of independent components appearing in all phases, at equilibrium where the number of components represent the Fewest variable constituents from which the composition of each phase can be expressed directly or in terms of an equation P =the number of phases wherein a phase is a physically distinct part of the system having the same composition throughout and which is separated from the other parts of the system by means of an interface Thus for systems at equilibrium, composed of a vapor phase and of @ liquid phase, fixing two variables defines a binary system equilibrium, three variables a ternary system equilibrium, and m variables an n-component system equilibrium, For a system composed of three phases, fixing — 1 variables establishes equilibrium for an mcomponent system. Ideal Mixtures Ideal gases are those whose behavior is described by the ideal-gas law, which can be stated mathematically as vuitt 2) wherein the volume of one mole of gas is related to the conditions of temperature and pressure through the ideal-gas-law proportionality factor A. Ideal gases do not interact chemically by forming new molecular species for’ associate physically because of differences in molecular volumes and nonspecific molecular attractive forces. The same criteria may be said to apply to ideal liquids.BINARY SYSTEMS 9) “The concept of “ideal” ga “igen ea” prs entre, and “ideal” liquid mixtures hay formed Be Bavs Tor many quetiatve describing equilibrium. Of principal interest in the field ef Dalton's iw of partial presues and Raoulls aw teting the pesare exerted by 4 component in the vapor phine of gaseour nator tote concentration ih the liquid phase and its vepor pressure. — Daiton's {I aw sttes hat the total press of & miture of gates equal tothe um ofthe pata pressures ofthe mived gues. These = a Samratntato te (13) Also Dalton postulated that the partial pressure of an ideal gas in a gaseous mixture is proportional to the relative number of rualeeules of that gas in the mixture (or to its mole fraction). ‘Thus Pom HP (1.4) Raoult’ [36] law, relating the partial pressure in the vapot pliase to the liquid phase composition, is expressed as Pom xP as) Combining Datton's and Raoult’s laws results in an expression describing mixtures of ideal vapors and liquids in equilibrium. = Sem Sym = Snr 6) and for a single component, Pym Pe i) The eapor pressure Py of « component is 2 unique property of the com- ponent and is direct function of temperatae-” ‘Thus it ineveases with an Increase in temperature, anda materia! having a higher vapor pressure ata given temperature than anther components sud to be more vole. Wapor presture and temperature are commonly related By means ofthe Antoine 2} equation B logPaA—aoy 8) where A, 5, and C are constants for a particular compound over a relatively narrow temperature range (usually not over 100°C), Valucs of these10 vapon-Liquio EQuiLiMRtA constants for various compounds and families of compounds and the temperature ranges for which the constants apply appear in a number of references, Dreisbach [12], API Project Report No. 44 (3), Perey [34], and others present either the Antoine constants, tabular vapor pressure data, oF both, . Example 1.1 “The following vapor peessute-emperature data are available, “Temperature, “© Pressuee, mm Tig | Ethyltenzene | Ethyl cyclohexane 760 13619 178 100 dat 69.08 30 367 $1.50 Determine the boiling temperatures for both compounds at 100 vim 1g pressure by the Aine equation Solution “The thtee points of data permit the direct solution for a set of Antoine constants by Algebraic means, Aa alternative method descrited in Dreisbach [12] can be used. The ‘constant Cis solved for fist by the empieicalformuls: Co 2390.19 where ty ithe normal boiling point. The three linear equations can then be solved for the Best values of A and B. Fer ethylbenzene, opr a~ 2 Erm Ae C= 239 — 019613619) = 203.12 eee) 02388 TSO) = thy FO” TAT — TT By BT By = 18033 = 1037 Substituting ditetly fr 4, Ay = 698668 Ay = 65868) A = 695656 dus = 6.9567 For ethyl eycloherane the solution i identicalPASE FOUILIMRIA-RINARY SwETEMS 11 ‘Tabvlated Results Ethybenzene Source or method “Antoine constants BP, 400 men Ig 4 a € 9073} 0076 | sea7z2 1427 Dicisbach’s method soser | iar | 21312 1382 Published valves sass | raraass | 213.208 132 Ethyl eyeloherane Source ot method ‘Antoine constants BP, 400 mm Hg. °C 4 a € Direct algebraic sotution | agar | 2900 | 3476 1099) Dreisbach's method sazs | 2 | 21396 top.t1 Published values 6avon | isedors | 215.128 109.12 For convenience in interpolation, vapor pressure charts in the form of Cox [10] and Katz et al. [25], among many others, are commonly used. Relative volatility is a widely used relation in distillation calculations since it is a measure of separability; the larger the value of a,,, the ensier the separation, It is defined by ) For an ideal mixture aga ot (19) Nonides1-Gns Mixtures Actual gases and their mixtures generally do not follow the ideal-gas law because of the formation of different molecular species and because of intermolecular attractive forces. Deviations from the ideal-gas law can be expressed in a number of ways, and probably the most useful from a distill tion standpoint are those of compressibility factor and fugacity coefficient. Za ain12 vapor-tiquip equininata where Z is the compressibility factor representing the ratio of the actual gas volume to that of an ideal gas under the same conditions of temperature and pressure, It approaches a value of unity at zero pressure and is essentially unity for most gases at and around atmospheric pressure, However, extreme deviations can be encountesed at Example 1.2 Assume that the ethylbenzene-ethyl cyclohexane system behaves ideally in a liguid rmintuce and determine the telalive volatility of elbyl eyclohesane refered to ethyl- benzene at (@) $0°C, (8) 150°C. ‘Antoine constants ‘Componente 4 a © Ethylbenzene easi9 | 1424255 + Eihyleyciohesane | 6870 | 1384036 Solution log? = an ‘Components 1] eee | mete] legr | Pmmilg () Ethytbenzene so | 263206 | sii. | 1.34601 130 | 363206 } 392134 | 3.03sas @) Ethyleyetohexane | 30 | 265.128 | s.2202¢ | 1.65015 150 | 365.128 | 3.79085 | 3.07946 At SoC. au, = 1271 = 1407, ‘Most thermodynamics books give methods of determining compressibility factors using critical constants and correlations based on the theorem of corresponding states, For example, Hougen et al, (22] present a compres- sibility chart relating compressibility factor, reduced temperature, and reduced pressure for materials having a critical compressibility Z, = 0.27, By means of tabular data and a correlating equation, the compressibility for compounds having critical compressibility factors other that 0.27 can be determined. This method resulted from the work by Lyderson et al. (31) Fugacity coeflicient » relates the partial pressure exerted by an actual gasTIASE EQUILIARIA-RIVARY fstEES 13 to that which it would exert if it behaved ideally. It is defined as follows: L ral 1.12) (tay where fis the escaping tendency of fugacity defined by Lewis and Randall (26) by a6) (23) wrens aay At lower pressures, around atmospheric nnd below, the fugac essentially equal to the partial pressure ofthe gas and the Fugacity coefficient is essentially unity, whereas at higher pressures it deviates widely from unity. Hougen etal. [22] presented a fugacity coefficient chart based on the work of Lydetson et al. (31) relating the coefficient with reduced temperature and pressure for compounds having a critical compressibility Z, = 0.27, Tabular data and a correction equation enable evaluation of the fugacity coefficient for other materials having different critical compressibility factors. There are many compressibility and fugneity coefficient charts in the literature. These range in type from generalized relationships to specie compound charts correlated on the basis of true reduced conditions, pseudo reduced conditions, and in some cases on actual temperatures and pressures For specific compounds the specific charts are more accurate; for general use the author prefers to use the Hougen et al. [22] generalized charts and tables. For a nonideat-gas mixture the fugacity coefficient enters inte Dalton's Jaw equation as (114) Nonideal-Liquid Mixtures In actual liquids and liquid mixtures the mean distances between molecules fare much less than those in the gaseous state and the forces of attraction between them are much greater, Nonideal behavior is evidenced by non- additivity of volumes when compounds are mixed in solution, and by heats, of mixing when the pure components are mixed at constant temperature and pressure. The extent of deviation from nonideatity of components in liquid mixtures is measured by the activity coefficient y. Agnin the reader is referred to standard thermodynamics texts for derivations aavt definitions of this factor in terms of activity and fugacity. Applying this correction factor to Raoult’ law eesults in Pom vals (113)14 varoa-tiquio EQUILIBRIA Incorporating both the fugacity coefficient and activity coefficient into a corrected Raoult’ law resulls in Eq. (1.16), which can be used as ny Pim veePs (119) a basis for calculation of vapor-liquid equilibria if the intercelationships of the variables are known or can be predicted. (Some thermodynamicists prefer to define three or more factors to correct for nonideal behavior. Here the two described are considered sufficient.) Relative volatility in nonideal systems is shown by Vd voy peer a) wie Pon : It may increase, decrease, or remain constant with increase in temperature, depending upon the nature of the system, Because relative volatility varies Ies with tentperature than does vapor pressure, itis used quite extensively in distilation ealeatations Phase Diagrams hate diagrams are used to dstribe two-component systems by ploting rr —r———C =‘ presture at a constant value ofthe remaining one.” Composiion versus Entalpy (a fonction of temperature) at constant presse, and vapor com. Position versus liquid composition pots at constant pressure areutefu fa oo r—r_—_—auril_nrnrr_sed talclations. Figures Io and L.2'ilusrate common types of binary phate In Tig 1.1, thea ej dlagrams ace typical of regular or norma systems “The 6 fj diagrams ar ypc of minimum Bolling homogeneous azotrope, the gk dngrama of mexianbolng honiogeceotssasoiropes, andthe deh, 1 dingeams of minim boing heterogeneous aestropes. Ta the st thre systems only one liquid phase exists, whereas inthe out, (wo iui phases ean exist at and below the aeotrope fenperatre Effect of Pressure on Phase Equilibria As the total pressure is increased on a binary system, the boiling tempera- tures of the pure components inerease and the boiling points of their mixtures also increase, In addition, the two-phase area on the rx plot decreases untilPHASE EQUILIMRIAMRINARY S¥StIANS 15 @ © a an o o , | 7 , a a oO a igure 1.1. Phase diagrams for various types of binary systerss. it disappears at the critical temperature and pressure of the -nixture. This is indicated in Fig. 13a. ‘A similar behavior is noted on the 2.x plot when the temperature is increased, ns shown in Fig. 1.36, It has been observed that, as the pressure increased, the difference in boiling points of the two components may decrease, increase, of remain essentially constant, depending on the character16 varor-tiquio EQUILIBRIA Erie gen tninciy ot comoes 41000 39.000 37000 33000 § ss000 8 31.000 12,000} 29000 1,000} 4 10,000) zg 250005 9009} 23000 8000] 2000 000) Graven, 18,900 = 6000 OF OF OS 04 05 06 OF OS 0 10 riovne 1.2 Enthalpy-concentration diagram. (From Cut fond Wang, Ind, Eng. Chem, 42:273 (1950). of the components in the mixtures, as long as the eritical temperature (or pressure) of neither component is exceeded. 1.2. VAPOR-LIQUID EQULLIBRIA—BINARY SYSTEMS Heal Vapors and Liquids ‘The experimental criteria for ideal behavior of mixtures are 1, There is no net change in volume when the components are mixed in the liquid or vapor state, ie. the volumes are additive. 2, The heat of mixing is zero.PHASE CQUILIERIA-MINARY syStESSS 17 iouRe 1.3 Binary phase diagrams at different pressures, 3. There is no change in molecular aggregation with respect to the components mixed and uni While in most eases nonideal mixtures can be detected on the basis of the above criteria, it is possible for molecular complexes between like molecules to be broken and others formed between unlike molecules with a zero net volume change. Also it is possible for physical interactions with endo- thermic heat effects and chemical interactions with exothermic hent effects to balance one another with no net effect. The chances of both occurring, ina mixture are remote, but it is recommended that the ideality of mixtures he tested by more than one criterion. ‘The vapor-liquid equilibrium composition data may be ealeulated for a binary system asstimed to be ideal from the vapor pressure-temperature data of the pure components at and around atmospheric pressure. A series of temperatures ranging from the boiling temperature of one component 10 that of the other are selected, and the vapor pressures of the components are evaluated at those temperatures from data, charts, oF from Antoine constants, The vapor pressure data are then used in Eqs. (1.18) and (1.19) to calculate the vapor and liquid compositions in equitibrfum at the18 varoR-tiquio EquiLtoRtA temperature selected. From the combined Dalton’s and Raoult’s law Eq. (1.6), applied to a binary system, the following equations result: ats) (1.19) ‘The liquid-phase boundary line on a Px diagram (at constant temperature) is a stright line for ideal sysiems described by x(Pi— Pd + Pee Py, (1.20) In ideal binary systems and in actual systems whose behavior approximates ideality, the relative volatility is constant or essentially constant, and the ‘equilibrium data may be calculated by (121) and Js x= 1.22) uF ee) Thus the entice range of equilibrium data can be derived from the constant relative volatility or from an average relative volatility, if the a values at the component boiling temperatures do not differ more than a few percent. Nonldeal Vapors and Liquids Although many binary systems, both in the vapor and liquid states, approximate ideal behavior, the greater majority of systems encountered in practice are nonideal in either or both states. For distillation calculations necessary to have reasonably accurate vapor-liquid equilibrium com- position data, ‘These may be obtained by a variety of methods 1. They may be determined experimentally. . 2. The system may be assumed to be ideal in ils behavior, and the data ‘may be calculated from the ideal equations. 3. The data may be calculated from a few experi empirical equations. 4. They may be roughly estimated from physical data on the pure com- ponents alone with the use of empirical relations, satal points by utilizing Since no method of calculating vapor-liquid equilibrium data is completely accurate, actual experimental data should be used if at all possible. ForINARY systrus 19 systems known to be approximately ideal, the use of ideal equations may be satisfactory within the limits of engineering accuracy. Where the system is known to be nonideal, it is necessary to use method 3 if complete experi mental data are not available, Method 4 is to be used only for rough Decanse the fugacity coefficient is a quantitative correction Fretar showing the extent of the departure of actual gases from ideal behavior and because the activity coeficient is a similar correction factor relating the behavior of actual liquids to that of ideal liquids, evaluation of these coefficients as a function of temperature, pressure, and composition is essential 10 the determination of vapor-liquid equilibrium data. Also, because distillations are practically always conducted at “constant™ pressure conditions, only the isobaric data are useful in direct application to design calculations. Thus, the relationship between the coefficients, temperature, and composition at constant pressure is of particular interest. Furthermore, most distillations are conducted at lower pressures where the behavior of the vapor phase is essentially ideal and the fugacity coefficients are approximately unity. Because of this, more effort has been extended to the study of solution nonideality through activity coefficient correlation with composition at constant pressure, Variables Affecting Activity Coefficient Composition The Gibbs-Duhem [15, 13] equation was derived from basic thermodynamic relations for constant temperature and constant pressure, and neglecting all internal energy contributions of surface, magnetic, electrical, etc., effects. (For derivation, sec chemical engineering thermo- dynamics texts.) In terms of activity coefficient and composition, the general equation is (RY) ta CRA tt G8 cao Because of the conditions assumed in the derivation, this equation is strictly applicable only to systems at constant temperature and pressure. It cannot bbe integrated unless some simplifying assumptions are introduced which are valid for the particular system under consideration, Otherwise the com= position data calculated from the integented equation can bs seriously in Thl and Dodge [23] modified the Gibbs-Duhem equation for a binary system at constant pressure and obtained sy (124) aiding + x:dln yy20 varor-tiquio equiLinnia where anyar z-~Snlz), saad and OH =H = nH — xo (1.26) where H, Hf, H1f are molal enthalpies of mixture and components, respec+ tively, as liquids at the temperature and pressure of the equilibrium. The effect of pressure on activity coeficent can be shown by In 2 1 "Yer RT (P= VidelPa = Pad (27) ‘The variation of activity coefficient with temperature is described by -(# oH) [et enah R Ir, 7 OR TAT, es Esomple 13 Assuming ideal behavior of both liquid and vapor for the binary system, moctane- ethylbenzene at 200 mm Hig, compute the following data and plot fx, 1, and xy ‘Antoine constants 4 2B c eam | ussis | 20952 2Eihytbersene | 69st | 1424255 | 219.206 Solution () Vapor pressure data (from Anioine equation): Pressure, en Hg ‘Temperature, *C ‘Octane | Ethyibentene nae 2000 a2 ts 2102 1308 " nse vet ® 223 8.3 a 239.7 3 3 92.67 pst 200.0PHASE EQUILIDRIA-BINARY sySTENS 21 (8) eal compositions: 5 = 200 — PAE, = Py). yu = Poa200. Tengenton, © | B0- | A = [=n [on ne wa] ssa | imo | am | v0 % a | is | cam | imas | con 0 mo | ad | oss | tas | ome ® us| Gee | caer | ‘aaa | oars x ns | ona | ste | tise | onto ner oo | a foo on | ee “ = + b ——|-- elt 09, oats on ora G0 10% da 0a OE GH Mon ttn sine etn cere nd ish ‘ai Exasitie 1.3. n-Octane-ethytbenzene system at 260 mm Hg. (2) Temperature: composition diagram. (tb) Equilibrium diagram, The van Laar Equations Van Laar (44) attempted to relate activity coefficients and compositions by making the following assumptions regarding the thermodynamic changes occurring in mixing pure liquids. He assumed (a) that the change in the excess partial entropy of mixing of a component is zero or the actual change in partial entropy of the component upon mixing is equal to the change in partial entropy of the component upon mixing if it formed an ideal solution, i.e. AS," = 05 (b) that the partial molal volumes of the components remain constant upon mixing; (c) that the van der Waals [43] equation of state applies to both components and mixture as vapors or liquids, Van Laar used the basie thermodynamic relation G).- "G)-i ATONUQUID Gavin and the van der Waals equation of state. Differentiation yielded (1.29) and substitution gave au) _@ Gih- 7 3 Integration from vapor at zer0 pressure to the liquid state yielded aan Using the assumption that Vs 5 and that the constants of the mixture could be obtained by eae = (Of + tt xo «3 and Be = Bb aby Ho aby (33) he derived ar line a3) for component I in a binary system. Defining by Ase (1.35) and : (36) the classic van Laar equations resulted 2 Thay = 139 *" Tae Vy Tiny,» —42. (1.38) OaPHASE CQUILIDRIA-BINARY SySTEMS 23 Woht Equations From basic thermodynamic equations it can be shown that the activity coefficient of any component can be related to partial excess free energy and composition of a solution as follows: | ¢ scr) Om, I rarinee = ATIny, (1.39) Wohl [46, 47] related excess free energy and composition in the liquid phase by the expansion of the excess free-energy equation = Ley + Ltt to (1.40) where (partial molar excess free energy) (11) Dn and qu, gn ete., are effective molal volumes of compen: For example, (effective volume fraction of i) (1.42) and ay, Ox. ete. are constants relating force interaction between combina- tions af unlike molecules. For a binary, Zetey = ten + rien = 2 (149) Since a= an (aay the greater the number of terms in Eq, (1.40) the more accurate the fit to actualdata, Equations involving for example, diy are third-order equations snd thote involving, for example, ay are fourth-order equations. ‘Third-order equations for a binary system were developed by Wohl as follows: ace t — = 222, 32,203 + 3217s Oy23 4 UL — (14) Since the sum of the effective volume fractions is equal to 1,0, any or all terms can be multiplied by 24 ++ 73: ac 2303RT ™ (4 Bea) al2igi(2aun } 3ay F tegen + 3enesd) (1-46)24 varoRLiquiD eQUILIDRIA Letting = 12a + Ban) and B= gilda + 30) Differentiating Eq. (1.39), HAGH/2309RT) _ al(ry + my)AC*/230327) a log ys ngelan) in mslgela) nelgalgn'l— mA 4 2m Bgl) + tn F mata? logy In terms of effective volume fraction =, the Wolil equations become ieee [4 2a (a — 4) % efova(et al logy Osher Equations Correlating Binary Vapor-Liquid Equilibria (a7) (14s) 49) (1.50) asi) (1.52) (1.33) (1.54) (1.55) (1.56) Carlson and Colburn [9] modified the van Laar equations to eliminate the tomperatuce variable. Since van Laar assured q, and gy weee the van dee Waals volume constants hy and hy, the ratio of the effective motal volumes becomes equal to the ratio of the modified constants. Thus, gus = B'/A™ and txt Gia log re st) (1.58)PHASE CQUILIDRIAGDINARY S¥STTANS 25 Generally the “normal” binary systems show positive values of log y over the whole composition range. The minimum boiling azeotropic systems hhave positive values throughout the composition range, and the maximum boiling azeotropic systems exhibit negative values over the entire range of composition. The constants 4’ and 8” in the Carlson-Colburn modified van Laar equation are related to the corresponding constants in the original van Laar equations as follows: — 139) 2.3037 (159) a (2.303AT (1.60) where 4’, BY are the Carlson-Colburn constants and A, B are the original van Laar constants, ‘The constants 4" and B’ in the foregoing equations can be evaluated from experimental data, For a binary system only one point of data would be required, theoretically, for simultaneous solution of pairs of equations ‘This requices an accurately determined point such as that represented by an azeotrope composition and temperature, Other points may be used if they are accurately evaluated and those determined in the middle range of concentration, ie., x = 0.4 to 0.6, are usually more reliable. A common method {41] utilizing the Carlson-Colburn modified van Laar equations for evaluation of isobaric binary vapor-liquid equilibrium data from one experimental point without the use of trial-and-ertor calculations is outlined as follows. Solving for the constants in terms of composition and activity coefficients and utilizing a tes (1 1 pen) ay tops (1 + SPER (162) and using - xi the method is as follows: 1. From the experimental point x,t, Py determine yy a Pu and Pe 2. Caleutate the values of a’ and J from Eqs (1.61) and (1.62), 3. Compute yin, and yyy forthe whole concentration eange ant plot ‘A. Caleutate tx and fy data by the following steps: (e) Assume that Eq. (1.63) will express the vapor presiwee relation of both components with temperatore. Pima eb (163)i Yorgi gviienin (This is valid for'most binary systems over their range of boiling temperature.) (1.64) (1.65) (1.66) (1.67) (1.68) #’is obtained from the vapor pressures of the two components at two different temperatures (within the boiling range between that of the two components) Poa = Pas Pra Ed) an OP an (1.70) a" and b can be considered constant for a given system. (©) Caleulate a and 6 for assumed values of x and calculate a’ and 6’. From Eq, (1.68) evaluate P, corresponding to the different values of x. From the vapor pressure data determine the temperatures corresponding to the different values of Py. From the limit equations amy (1.2) A’ and B" can be determined also by extrapolation of the log y versus curves to x, = 0 and x)= 0, This method is subject to possible error because the experimental points at low concentration are difficult to evaluate accurately Exomple 14 For a lal presure of 100 mun Hig ealeulate tx, 1y, yx, yx eurves for the rystem m= odecane-buiylearbilol using the Catlion:Clbuen modified van Laar equations, The teolrope lerperature was found {0 be [428°C and the composition iy t= 0.65 mole fraction dodecane.[ASE EQUILIPRIA-PINARY S¥SETMS 27 Solution ‘Vapor pressures are conveniently shown as dats calculated by the Antoine equation: Boiling Polnts at Several Pressures, ‘Component 1, ‘Component 2, ‘medodecane butyl entbite! Pressure, mm Ig 30 1714 ia 13990 133 146.18 198 138.10 m3 teria i 100 100 eee =n e87 Raa gee logy = 005461 logy = 027577 ‘The constants A’ and B’ of the Caclton-Colbutn modifeation ofthe var Lanr equations are soled for by Eqs. (1.61) and (162) (oasyoans7)]" (Gsyo.as46H), = 0.75598 or 0.785 ; top) 0.08461 Bee tog ya 1 + 282). o.arsnal 1 4 620.0516 n( * ee) ons rf OIN0TITT = 051590 or 0si6 Caleulted| y28 vaPOR-LIQUIO EQUILIBRIA 7 LA wo \- a | gas r leat vas | Foe “Saas as es oO 369570 fe ig amet satin EXAMPLE 14 mDodecane-butyl carbitol system at 100 mm Hg. Equilibrium boiling point diagram. (From Houser and Yan Winkle, J. Chem. Eng. Data (35754 Many equations utilizing basic thermodynamic-relations and based upon various simplifying assumptions have been developed to describe the inter- relationship between activity coefficient and composition of nonideal binary systems for the purpose of extending partial data or correlating data. Scatchard and Hamer 39} assumed the effective molal volume of the component was equal to its molal volume, ie, gi = Vi and gs Margules (32) assumed the effective molal volumes of the components to be equal, ie., gal, = 1.0, resulting in the following frequently used equations log yy = xe[A + 2x8 — A) 1.73) Jog ys = 110 + 2x44 — 5) «7 Hala etal. 19} assumed ideal-vapor-phase behavior and derived from the Redlich and Kister (37) equations a set of equations involving vapor pressures of the pure components. White [48}, Yu and Coull (49), Li and Coull (27), Bonham (7), Black [4, 5], Black etal. [6], and others have derived equations of the van Laar type. In addition there have been many other methods proposed for prediction of vapor-liquid equilibrium data—from total pres- ‘sure measurements (35), from data on other but similar systems [48}, from selative volatilities {17, 18}, and from pure-component properties for certain classes of systems [29]. Binary azeotrope temperature and composition can be predicted at different pressures from data at one pressure (24).THASE EQUILIARIA-RIVARY SYSTEMS 29, Example 18 With the use of the Margules equations and the following experimental point determine the activity coeficient composition curves forthe ethyl alcaiol (component 1}-methy! thy! etone (component 2) binary system. ‘Azeorrope! 30.0 mete % ethanol at 165.2°F (760 mm Hig total pressure), “Antoine corstante Componente 4 a ¢ () Ethanol Boas4 13543 22.65 @ MEK 69721 12098 216.00 Solution [At the ateotrope, x1 = ys and, consequently, Y= Pal? = T0613 = 181 tog yy = 0.07225 Ye TH0/609.2 = 1.189 Log yy = COTSIE “The Margules equations are logy mad + 2048 — 0) lopys = t+ 2e4 — mm) Substituting the values of composition and activity coefiient at the azeotrope results in 0.07225 = 0.25(4 + (8 — A] 0.07518 = 02510 + (4 — BY Solution for A and Bis direct atthe mid A= 0307 9 ~02490 Repeated solution of the Margules equations gives the fllowing set of data 0.00. 1.000) oor 4029 04899 rns 407s ras 0.18796 isa 0.28900 194s30° VAROR-LIQUID EQUILIBRIA 20) 0 02 0a 08 EXAMPLE 1.5 Activity coefficient—tig- vid composition relation for the methyl ethyl ketone—ethyl alcohol system at 760 mm Hg. 1.3. PREDICTION OF BINARY VAPOR-LIQUID EQUILIBRIUM DATA FROM PURE-COMPONENT PROPERTIES ‘The prediction of mixture properties from properties of the components composing the mixture is the most desirable method because extremely large quantities of pure-component data are available compared to mixture prop- erty data, For very gross approximations, the values of the a and b constants of the van Laar equation can be estimated from DIRT, oe (1.75) Rr, One (1.76) Critical data can be estimated by the method of Lyderson [30]. When the data are substituted into the van Laar equations for 4 and &, the following results are obtained: am i} a At best these approximations could be applicable only to binary mixtures, showing no chemical interaction effects or hydrogen bondingPHASE EQUIIDRIS-RINSRY SysteMS 31 Finch and Van Winkle [14], using Scatchard's and Hamer’s (39, 40] modification of the van Laar equations, Hildetrand’s [21] approximation for internal pressure and postulates by van Arkel [42] and Lundon [28], developed a method for prediction of vapor-liquid equilibria for binary systems (rom pure-component properties. For nanpolat-nonpolar systems, satisfactory agreement with experimental data was ottained, For pola nonpolar systems the agreement was less satisfactory, and for polar-polar systems the method failed to predict data within required engineering 1.4, CONSISTENCY TESTS FOR BINARY YAPOR-LIQUID EQUILIBRIUM. DATA Distillation process design generally involves the use of some exjerimental vapor-liquid equilibrium data. These data may be reported in the literatuce as binary, ternary, or multicomponent data and may be complete, patially complete, or consist of only a few points, Recause the sources of such data vary widely with regard to methods used, techniques involved, and even in skill of the individuals determining them, it may be necessary for the engineer, who proposes to use the data, o apply some kind of test t0 determine, approximately at least, theit quality. Although this testing procedure may indicate only whether or not the data are internally consistent, such an indication is of considerable value in establishing the probable accuracy of calculations in which the data are involved. In addition, there ray be more than one set of data reported on one system which differ in numerical values of temperature or composition. Consistency tests will generally enable the selection of the data which are probably more correc. Most of the consistency tests have been derived from thermodynamic relations for binary data, and most of the tests are based upon the Gibbs- Dutem (15, 13} reintion Rearrangement of the Gibbs-Duhem expression written fora binary system results in (rw --2 w This states that the ratio of the slopes of the activity co -Mficient curves versus liquid composition is equal to the negative ratios of the mole fractions at any point, This is called a thermodynamic consistency: test since the Gibbs-Duhem equation was developed from basic thermodynamic defini- tions. ‘To check the consistency of any set of data, the curves are plotted as indicated and the slopes at selected compositions ere evaluated and com- pared with the ratios of the liquid compositions.32 VaPOR-LiQUIO EQUILIBRIA Te ° D to ° : 70 Nicure 14 Redlich-Kister tet, __FigURE 1. Broughton-Drearley test. Redlich and Kister (38) recommended plotting the ratio of log (yy) versus the liquid composition x,, and determining the areas above and below the line where log (ys/y) = 0, and extending to the curve, If the areas are equal, the data are considered to be consistent, The extent of deviation from equality is a measure of the inconsistency. Figure 14 illustrates the method. Broughton and Brearley [8] developed a consistency test based upon the assumption that the term T'log y is independent of temperature. They recommended plotting T log (ys/ys) versus x, and measuring the areas above and below the line T log (71/94) = 0 and the plotted line, This test is shown in Fig. 1.5. A different type of consistency test based on the van Lar equations was proposed by Black (4] wherein the values of (log y,)** are plotted against the values of (log ¥4)"4, as shown in Fig. 1.6. Ia straight line results, the data are said to be consistent with the van Laar relations. ‘The Norrish and Twigg [33] consistency test used the following relations: sain(2%4) cu) whete & is the ratio of the molal latent heat of vaporization of the lower boiling component to that of the higher boiling component. They claim that the data are consistent when a straight line results from plotting z against the liquid composition x,, as indicated in Fig. 1.7. Thus: remetC (1.81) where mand C are constants for a particular binaiy system. Adler et al, {1} proposed a thermodynamic consistency test for binary systems when the more volatile component is above its critical temperaturePHASE EqUULinRrs. ISARY SYSTENS 33 & g oa»? ° 5 © RGURE 1.6 Diack test FHoune 17 Norvsh and Twig test. “The following equations are presented: adhe) +%dIn(yPirnp) = ZedinP, (1.82) where Fabs nae (183) where My, = volume of liquid, eu f/mote sidin Porn) + adn (KP ph, (13 where Kat (1.83) xidin Ky din Ky = 2, + Zig (gp ainP; (1.86) Exomple 1.6 ‘Two sts of experimental equilibrium data, A and B, are available om the binary sytem methyl propyl ketone-ethyl alcohol. Py = 760 mm He. 4 2 ae ee 1 099 0995 13 ost 0765 178 035 035s 180 033 036 185 os 035varon-tiguio equitlalh HOH NK Temperature, °C 80 260 4 os ma ‘Vapor pressure, mm Hg| Temperature, °C | Vapor pressure, mm Hg. 20 285 20 o a3 6 200 no 200 40 a8 40 760 1033 160 Determine which set of data is the more internally consistent by () the Redlich and Kister test and (6) the Broughion and Brearley test. Sotution ‘Activity coeficiens are caleulated by why not (ethanol is component 1) System A aD my ae ee ka oses| zest 7654 | 2008 | 09m | 3067 | —oasor| —1a5 0369] 7434 | 7603 | 2787 | 0998 | 290) | Coase] ian osx] rsa9 | soso | 2962 | sare | 176 | -0203 | —2136 0301 sia | 304 | tas | 1526 | Sooms | “toa oxo] weit | soaso | 3es7 | tes | tase | ons | aso System B a me jon |e nm | | tepra] Theg ovira 099 | 7s1 | 7654 | 208 | oe | 132 | —o1s| —a639 ost | wer | 996 | 2923 | as | rser | Sosa | —a209 035 | sis | eon0 | si67 | 1200 | 1307 | -oomo | Sat 033 | sao | eso7 | ssza | reas | soz | Moors | 502 as | sav | sees | aaa | ire | sss | “oosie | 2200-o2|—-tf ff -03; A sparem coa{_|ssrrem || es lalerie| ssystem 4 | ssjten @—[ | oerezososescearomasio "a nrozosarasoearon O10 te te EXAMPLE 1.6 (a) Redlich and Kister thermodynamie consistency test. (8) Hrough= ton and Brearley thermodynamic consistency test. 7 ‘The reults are plotted on the accompanying figures. The ratio of aren above the abeciaa to area (absolute value) below are: ‘Sytem Redlich and Kister | Broughton and Brearley 4 020 031 a 040 030 [Neither et of data enn be regarded as consistent by thet two tests although system is the better of the two, The data of system A are noted {0 be the smoother ofthe two sets, Example 1.7 ‘The following data on the bionry system a-octane-Cellorolve ae tobe tested fr thermo- dynamie consatency by (2) the Black tes (8) the Noth and Twigg text “The vaporliqvid equiliium data forthe -octane-Cellooive system a 760 mm Iigare: fen Jee ‘010 0.064 056 0.261 0.108 0386 0.302 0.340 ost 0.620 2.80 0.46 sone36 -VarOR-LIQUID rqUILIBRIA ‘The Antoine constants for octane are: 4 a © 6027 135.13 209.52 e-vapor-presure data for Cellosolve are Te P. mim ig 1169) ans i227 ‘05 a or i) 707 Solution (e) The activity coeficients are solved for a in the previous example. Component Iie rm | len [Copvo} togre | clogs 0997 | 071773 | 08a | —000130 0.993 | 0.66237 | 0140 | —0.00303 rot | oseia | 07626 | coats | o.osss tise 1is0 | 03416 | osss | o.c6oro | ost unt 1942 | 009237 | 03039 | o2aa2s | 05369 liso sexr | ooosee | oss | assess | 0.7476 m6 | ose 3163 | 0.00000 | 0.0000 | as7sss | 0.7546 ‘The data are not isothermal and, consequently, & sttgighi-line curve would not be “anticipated, ‘The smoothness of the curve joining the points indicates, however, that the boiling points were prectely measured (}) ‘The Norrish-Twigg method requites the evaluation of (ade Cie ‘The Clausiat-Clapeyeon equation it used forthe evaluation of latent bea of vaporization a Ole = TO = ¥ el be suticiently accurate to ats materials, Hence, that the catio(H, — VOM, — Poy = Lor these tte, _ (éPldid, Bit,” @Pidin.ase rouuioniacnmany evsres 37 From the Antoine quater, a r OT an aryee ay + Hee) where the compariton is made at fined temperature ‘A temperate inthe middle of the range, 12°C, is chosen, A sel of Antoine constants for Cellosoive is found for # narrow region arcund 128°C; namely, log P= 7.6412 — 1958)(¢ 4 213). Then ( a y 1 ~Tcsseneony Lars) 78" Soorozesoaosoeoiasosio | ~°"o o1ozasoats OEOTOB ODIO ry o FXAMPLE 1.7 (a) Dlack thermodynamic consistency test. (6) Norrish-Twigg thermodynamic consistency tet,38 varoR-LuQUID EQUILIBRIA ‘A very good siraight ynamie consistency. ean be drawn through the plot of 2 verus x in King thermo Equations (1.82) to (1.86) are graphically integrated with the use of available data or generalized correlations if the data are not at hand. Figures 1.8 and 1.9 schematically illustrate the integration aula ouRe 1.8 Integration of equation. riaure 1,9 Integration of equation. Herington [20] devised an empirical test which claims to be a consistency test for binary data, He recommends the following procedure. 1, Calculate the value of log (ys/74) for a series of compositions and plot these values against the values of x, 2, Evaluate the integral Jas follows: 1m J (loeen- toe?) ae (187) From Fig. 1.10, (os (188) Also, Eeaae (1.39) 3. Calculate the percentage deviation D by 100111 D (1.90) The value of is taken with a positive sign.PASE EQUILINRA- Inany systes 39 igURE 1.10 Areas used in Herington test 4. Determine the lowest boiling point in the binary system f,, which is ither the boiling point of the lowest boiling component of that of the mini- mum boiling azeotrope and the maximum Boiling range 9. From Eq, (1.91) determine J, 150 [01 J 1.91 7 “sn, The data are considered consistent when De (1.92) PROBLEMS 1. Experimental vapor pressure deta forfour alcohols are ws follow Methanat Eihanol | 1Prepanol | 2Tropanal Pome Hg 1G ALP, mm Hg © 30 190 364 De 100 212 Ms 28 395 400 99 os no ons 2am 0 313: 179 113 Sam ms 1260 190 1302 (@) Determine the Antoine constant forthe alcohols covering the range of peesnes from 40 to 400 mm Tg. (6) Determine the Antoine con 0 3800 men Ig. (c) Draw a Cox chatt Including on the chart all ofthe alcohols listed (2 Predict the vapor pressure temperature line for I-butanol and draw ion the chart ni forthe alcohols forthe pressure range of 400 mm40 YAPOR-LIQUID EQUILIBRIA Determine the relative volatility of methanol relative to each of the other three alcohols sted at 45°C and at 120°C, Determine the vaporliguid equilibrium data for x = 20, 40, 0, 80 mole % forthe binary syste ety aleahol-benzene at Coal pressure of 700 min Hg feom fugacity fang ecivitycoefclentealeulations. A’ea = 0.6, B'gx = 0.699 (Calculate the (o)van Lane constants; (0) Margules constants; (e) Carlson and Colburn modified van Laat constants forthe Binary system tolvene-2-methl heptane which shows an azcotrope at pq) = 0.82 and 110.3°C, Determine the tx, ty, 7 dala for the system methanol-ethyl aceite at 760 mm presure ‘Antoine constants Components > S B ce ‘MeOH tar | noo BiOAe nant | 21700 (One experimental data point is available zyx = 04975, 9 asc. ‘AL 400 mim Hg prestureealeulate the tx, tp. xy, andy curves forthe binary system ethyl eyeloherane-elhilbenzene from the following date son = 05970, 1 = suey = 050 r= OC Yacn = LOS ye = 103 Components = EB 693722 124.255 213.206 ECH sa70u1 Haas 016 nisiza (o) Assuoving the ehlocoforntacetone syste to bean “ideal” syste (a0 mPsoteore) Cnleulte the data for the tx, #y diagram. Apply the Redlich Kiser est to Out tnlevlated data, (2) Using the following experimental data apply the same txt ‘Acetone-Chloroform System at 760 men Hg4 The vapor presture datas *CatP, mm Hg Components 20 | «0 | 70 | 1520 ‘Aestone nr | ass | ses | 796 Chiorotorm | 259 | 7 | as | me 3. Apply the Black tet tothe following data on the {.3-butadiene-chloroprene system, Chloreprene 205 9. Apply the Norrish and Twigg consistency tet to the data in Prob, & 10, Apply he tests indicated to the following data on the eyelohexane-tolvene system. 342 0368 0.596 908 0.504 0.702 as 076 once a7 osm 0.92642 VaPOR-LIQUID EQUILIBRIA ‘Antoine constants ‘Components 4 Gyeloberane | 64s Toluene 69533 (2) Apply the Redlich and Kiser test (@) Apply the Herington test. Nomenclature @ = constant, van der Waals equ: ‘constant A. = Antoine constant, or constant defined in activity coefficient equa- tions A. = van Laar constant A’ = Carlson and Colburn modified van Laar constant constant, van der Waals equation = Antoine constant or constant defined in activity coefficient equations = van Laar constant Carlson and Colburn modified van Lar constant = Antoine constant or number of phase-rule components = differential jefined quantity, Eq, (1.90) superscript) denotes excess property rugacity = Gibbs molal free energy Gibbs partial molal free energy GF = excess Gibbs free energy H_ = enthalpy of mixture LH? = molar enthalpy of component of state, or molecular force aot mys A Rees 1 = defined quantity, Eqs, (1.87) and (1.88) i = subscript, any component J = defined quantity, Eq, (1.91) J = subscript, any component k ‘= ratio of molal heats of vaporization of lower to that of higher boiling component K = constant, m= constant n= subscript, nth component p= absolute pressure P= number of phasesZ,’ = critical compressibility factor, Z, = PHASE EQUILIORIA-BINARY SySTEMS 43 critical pressure vapor pressure of component / total pressure partial pressure of component f effective molal volume of component ideal-gas-law proportionality factor molal entropy excess molal entropy, Sieea critical temperature absolute temperature = temperature ~ Saccost 7 ~ normal boiling point molal internal energy = internal energy at zero pressure partial molal volume = molal volume degrces of freedom = molal volume of pure component ~ mole fraction of component in liquid mole fraction of component in vapor compressibility factor effective volume fraction; defined quantity, Eq. (1.89) PY, Greek leters = relative volatility = activity coefficient A. = difference = differential operator + = fugacity coefficient 0 = constant E_ = summation or defined by Eq. (1.89) Subseripes k= gas if = higher boiling component L_ = lower boiling component 1 = tiquia REFERENCES 1 2 ‘Adler, 8. D., Leo Friend, RL, Pigford, and C. M, Rosse: ACHE J, 6:104 (1960), Antoine, Cu: Comp Rend. dead. Sel, Pals, 107-681, 836, 1143 (1888),44° varoR-Liquip EQUILIBRIA 3. APL Projet Report No, 4: "Seated Values of Phsal and Themodjeami Prope o Hfrocarbot and else Compounds Camepc Pes, Piorgh ne 4. ek, ined, a. Chem, 3:03 (1958, 5 Blick Cle: AChE 2, 8309 (1959) 6 nck, Cline, ©. Dev ad MN. Pipadopouts: Ind. Eg. Chem, 8040-48, 1963: 030-47, 196, 2. Bonam, M8. MS, Ths in Chena! Engineing HT, 94 © toughen, D. Brand ©. 8. Bray Id ng. Chem, 4738195) % Carlson, 11. D.,and A. P. Colburn: Ind. Eng. Chem. 34:581 (19425. 0. CoE Ind. vg. Chm, 1:92 (1923), I. Dato, 1*hew Sater of Chemie! Phieiophy.” ok si 1, Manes, Engen 08 12, Dreath, KR. Phys ropes of Chenical Compounds 1 and Hh, dn Chen St an, Chm Ser os 15 (959, 25 (99). 3 UE 13, Duh Be Compe Rend Acad, Sel Par 11:09 (108) 14. Fin ec and Mc Van Winks AcLCRE. J. 845$-€0 (1960 1B. Gis.) We Trane Com. dad Se, 3192 (100, 16 Gibb) Ws"Caeed Work Longmans 1926 1%, Giimon, RD Zudheih and D-. Othe Eng Chen, 33:29 (96, TR. Gilman, EA, Weinnan, Keane, Mile Pasa and 0. F- Othner Sati Chest aae 85 19, Ht, Esai ek Fed, and. Vitis Vapor Lgl guia” Tes Ind by ©. Stand; eeamon Pre, New Vrds 938 20. Mesnglon EP. Gen Pera, 31487 (951) 2H. eb, Hy and RL Seat “Sosy of Non Eee” 3 eA {hum Soe Monopap Sr, Reid, New Yer, 1835" Dove 96 22. Hoven, 0, Ac Ke M. Walon, nd RA, Rage! “Chea! Pcs iri Themodjaumie Bes Way, Nearer, 13. 23. Ihe. Vand BF Dodge: Chem, ag Se, 2120 (1983 2. foe. Ia Bg. Chem 47:28) 198) 2. Kat D Uy DCm R-Kobayai FM Poetnann C.F Weinaug.3. A. Vay th J. A Fenn and of Natu Cas Enincring MeGrss-t New You. 3, 26. Lev, 1 Myand M Randal J Am, Chm Soe, 9:29 (1921 2. Un Ys Mand Col J st, Peel, 3402 (98, 28 Lenion, ts Tie, Frey See, 39:8 (93 D. ber C-¥samd W.F Grydon nd Eg, Che 491058 (95) 30. Lpdenon, Ac: teal Can of Pate Compounds, Une. Micon Ey ap SR. Cap 8 31, Lydenon, Ar LH A, ieenkor, and O.A, Hougn: Geveaied Thermodynani oper of Poe Fa, Uni Wcontn By ape Sia. Rpna 4 (cto 3) 22. Magu, ME Staber ded Wl Wen, Math, Mare 1, 41249 (193), 33, Now, RS and GIL Togg Ind. Bg. Chem 620 (1934 34. ey 3 "Chea Engineer Handbook” 3 ed, McGrew il 190 vol. 1, New Yor inciple—3. 36. a a ». 4, 4 2 4a. a 4s “6 "7, 8, PHASE FOUILIMRIA-RINARY SYSTEMS 45 Prengle, H. W.,and M.A, Pike, Je: J, Chem. Eng. Data, 6:400 (1961), Raoult, F. M.: Z. Physik. Chemie, 22359 (1888), Redlich, O., and A. T.Kister: Ind. Eng. Chem. 40:341 (1944), Redlich, O., and A. T. Kister: Chem. Ree, 10:148 (1982), Scatchard, G., and W. J. Hamer: J. Am, Chem. $3, 57:1805 (1938), Seatchard, G., Chem. Ree, 2321 (1931, Schechter, R.S..and M. Van Winkle: Petrol. Reiner, p. 301 (September, 1957) Van Ackel, A. E.: Trans, Faraday Soe, 420:84 (1948). van der Waal, J, D.: “Over deontinuet van den gasten vocisofteentand,” Leiden, 193, Yan Lane, 3 Je: 2. Phys. Chem, 188235 (1929) White, RR: Trans, AJLCKE., 413846 (1985), Wohl, Kurt: Tron, A1.OhE, 422215 (1945), Wehl, Kutt: Chem, Eng. Pregr.49:218 (1983), Wood, §. Et Ind. Eng. Chem, 42:660 (1988), Yo, KT. and J. Coull: Chem. Eng. Progr. Symp. Ser. 48(2):38 (1982),chapter Multicomponent vapor- liquid equiiibria A. multicomponent system is defined here as a system composed of more than two identifiable actual compounds or pseudocompounds or materials to which physical propertics can be reasonably assigned, As in the case of binary systems, multicomponent systems are encountered which essentially behave ideally under conditions of low pressure and normal disiling tem peratures, Those composed of mixtures of hydrocarbons, mixtures of iso- meric compounds, or mixtures of homologous compounds may approximate ideal behavior in both vapor and liquid states, Where mixtures of materials of highly dissimilar nature are encountered oF the conditions of pressure and temperature are severe, extreme nonideal behavior is evidenced 2.1, IDEAL MULTICOMPONENT SYSTEMS Multicomponent vapor-liquid equilibrium data for those systems which approximate ideal behavior may be calculated by Raoult's and Dalton's laws in the same manner as those for binary systems aay 22) ea 46MULTICOMrONENT varoR-LiQuiD rauitinnia 47 For a five-component system with two phases in equilibrium according to Gibbs phase rule, there are five degrees of freedom, four inécpendent composition variables, and temperature or’ pressure. If the liquid com- position is known and the temperature selected, the vapor composition can be calculated from Eq. (2.4). (If four liquid compositions are known, the fifth is also known because 3 x, = 1.0.) xP, YO SR Par Pauly Pe e and the total pressure is evaluated by means of Pea Py bP t mPa t Pet mle es If the temperature and vapor composition is known, the equilibrium com- position of the liquid is also fixed but must be determined by trial-and-error calculation. This may be accamplished by assuming a total pressure and calculating a total pressure by Syren, eo This may alzo be done by a dew-point pressure caleuation, By definition the dew-point pressure is that presure at which the fist drop of liquid is formed a8 the presute on the vapor mixture i increased at constant tem perature, Mathematically, an for an ideal mixture. In either calculation, plotting the calculated total pressures versus the assumed total pressures for three or four trials usually will enable the correct pressure to be evaluated from the plot with sufficient accuracy. Then the liquid compositions can be calculated by using the correctly determined pressure in wh xe (28) Considering a five:component system of a fixed constant pressure with coexisting vapor and liquid phases,:there are four remaining degrees of freedom, four independent compositions, and the temperature is fixed. If the liquid compositions are known, a bubb/e-point temperature calculation usually is made to determine the temperature, By definition, the bubble- point temperature is that temperature at which the first bubble of vapor is48 varoniquip rquiLinnin formed on heating a liquid at constant pressure. Thus, quantitatively, (2.9) ‘A minimum of thiee temperatures,is assumed, the corresponding vapor pressures are obtained, and Eq, (2.9) is used to calculate the sunimation of yy. The sum is plotted versus temperature and, at the point where the sum equals unity, the correct bubble-point temperature is established. By ricans of this temperature the individual values of y ae ealeulated, using the correspondingly correct vapor pressures. ‘When the vapor compositions and total pressure are known, the tempera- ture and the composition of the liquid in equilibrium may be calculated by a dew-point temperature calculation, ‘The dew-point temperature is defined as the temperature at which the fist drop of liquid is formed upon cooling A vapor mixture at constant pressure. Mathematically itis expressed by Eq, (2.7). Atleast three values of temperature are selected, the correct vapor pressures are obtained, and the summation is calculated by Eq. (2.7) ‘The ealeulated sums are plotted versus the assumed temperatures, and where Ex, = 1.0 occuts the correct equilibrium (dew-peiat) temperatuce is estab lished. The liquid compositions may then be calculated by means of Eq 28) Relative volatilities may be used where the temperature range (the dilfer ence in boiling points ofthe lightest and heaviest components) of the raulti- component equilibrium data is not great. Usually the relative volatilities are based on the heaviest component in the mixture. Arithmetic average relative volatilities (atthe highest and lowest temperatures) may be used or, if the variation in relative volatility with temperature is slight, they may be evaluated at one temperature and assumed to be constant over the tem perature range. ‘Thus, bh sug oma te ay ry for the ideal five-component mixture, Divide numerator and denominator of Eq. (2.4) by Py. The composition of component 1 in the vapor is, (2.10) and this in terms of relative volatility is 2 Kite a8 tn + Seta ~ Quy natMULTICOMPONENT varOR-LIQUID EQUILINRIA 49 Simiacly, ' Flint laa + Jalan tm Fe oa tn tems of any number of components, @3) eu) 2.2. IDEAL TERNARY SYSTEMS ‘Ternary systems represent a somewhat special case since many distillation problems can be reduced to consideration of essentially a ternary system, nd the calculations are somewhat simplitied as the result. From the phase role there are three degrees of freedom for a ternary system in two co- existing phases. Raoult’s and Dalton’s laws are followed, i.e., Eqs. (2.1) to (2.3) apply and bubble-point and dew-point calculations are maule to deter rmine the phase compositions, temperature, and pressure as needed. ecause xe tas Qs) and dota (2.16) 17) and (2.18) In many eases, particularly in graphical studies, it is useful to determine isothermal vapor and liquid composition lines. This involves fixing # and P, and enables direct calculation of the compositions (fixing one and cal- culating the others) from the following relations au (2.20) ean l=) 222)50 varon-UiquiD EQUILIBRIA ‘Thus, with temperature and pressure fixed, the vapor pressures are fixed, and the variation of x, as a function of x, can be calculated. ternary systems this isa steaight Example 2.1 Caleulate the iethermal vapor a For ideal fine on an x, versus x; plot. id liquid fines for the teraary ayttem tolvene- ehylbenzene-praylene, ata tolal pressure of 800 mm Hg and for a composition ea = 030, Fyn = O48. Components (D Toluene @) Ethylbenzene @) pXpene “The syslem may be assumed to behave ideally fo this elevation 6.95038 69519 6.99082 1990 1424255 453.40 Solution ‘The liquid composition line is found from m a ‘The temperature whieh x, = 0.3, Vapor presure (©) pXylene 500 sus 600 ar 700 122 (2) Ethythenzene sy 0s at, rom Hig ™ Hoiting point (D Toluene we 077 146 270 17 199 so 0.953 ons ‘unitMULTICOMFONENTT VArOR-LIQUIO MAUILIWRIA ST With the us ofa graph of x (calculated) versus 7, the final value of Px» $87 i ncated, At this pressure, = 129°C, Py = 1244 mm Hg, Py = 623 min Hg, y= 0.3242 = 0.05515 Fro Fg. (229, em 05041 = 0.11074 Similarly, the isothermal vapor composition line ean be calculated by (2.23) (This represents the equation of the solid line in Fig. 2.1.) 10 oo 002 0406 08 10 re ta rele traction, ethyibesine 2. Isothermal vapor and liquid lines for Example 2.1 When one composition is selected, the remaining vapor and liquid com. positions -d oF obtained from the isothermal vapor oF iso- thermal liquid composition plots at the fixed total pressure, In terms of relative volatilities, the following equations for ideal ternary mixtures ean be used: 1 he caer V+ (a= Day + (an cy “Tan tes De a = 2.26 08 yelllas = DEC — OFT n — 227) (Tess = D+ Mens — FT52 vAPOR-LIQUIO EQUILIBRIA Qualitative Graphical Study Visualizing the variance of composition, temperature, and pressure in a binary system is relatively easy since the number of variabes is such that the variation can be shown on a two-dimensional plot. In the case of ternary systems there is difcity in’ visualizing the variation of composition with temperature at constant pressure, unless solid figures ofthe type shown in Figs. 22, 23, and 24 are constructed of plastic sheet models such a8 shown in Figs. 25 to.2.7 are available. These ace extremely useful and informative, although it may not be feasible to spend the necessary time and effort on this kind of thre-dimensional plot. On many occasions the engineer is interested in estimating qualitatively the temperature-composition relationship before the actual ternary data are determined or calculated. In such cases a fairly good representation of the ‘x and 1y surfaces can be estimated from the characteristics of the binary systems alone. To do this, diagrams of the type shown in Figs. 2.8 to 2.17 riguae 22 Solid model of a ternary HOURE 2.3. Solid model of a ternary temperaturecomposition phase dia: temperature-composition phase di {gram showing liquid. and ‘vapor sur- gram showing liquid and vapor sur- faces—thinee “normal” binary systems. faces—two binary azeotcopes.MULTICOMPONENT VAFOR-LIQUID EQUILIBRIA 53 "S| el] Zz [= ‘| a Fe =| _ SET Fioune 2.4 Solid model of a ternary rigune 2.5 Plastic shest medel of lemperature-composition phase dia ternary temperature-composition dine gram showing vapor and liquid sur gram—three “normal” binary system, Faces—three binary azeotropes Each sheet shows the two-phase region at that temperature and representative tie Fines. are drawn up. These consist of conventional ternary plots using triangular coordinates or right angle, triangular plots using rectangular coordinates. Extensions are drawn from the corners of the triangle at right angles to the sides, giving three two-dimensional plots—one on each side of the triangle ‘The scale on the lines perpendicular to the sides of the triangle can be either temperature of pressure. Temperature is usually selected as the varinble since practically al distillations are carried out at constant pressure ‘On each of the binary plots the rx and 1y curves are plotted for the binary system indicated. Then a series of lines at a number of selected temperatures are drawn on the binary dingrams. The temperatures are selected such that isotherms intersect the binary diagrams at intervals ranging from the lowest, temperature at which the two-phase vapor-liquid region occurs to the highest boiling temperature on the diagram, The intersections of the vapor and id lines bound the isothermal two-phase region at that temperatuce, ‘The points, vapor-vapor and liquid-liquid, are connected with lines, and the$4 VaROR-LIQUIO EQUILIBRIA reat belwean te Tne Ar the apport iocherna,tecmay, two-phase ‘The te ines (lines connecting vapor liquid equiibium compositions) for the binaries ein the sides ofthe teary plot and in the ternary Bagram al ts lins generally tend fo point {o a corner of the dngram. If binary azeotropes are present, the Ue ine end fo converge a the bir azeotrope compositions Such dngres ave weful a characerising the vapor and quid suetnoes of a trnay aster, Sine the Ue linet sketched infor each itothermal tworphase epi represent (a ther intersections withthe liquid-phase and vapor-phase boundaries) the liquid and vapor compositions in equim that temperature, continuous lines can be drawn from one composition oe part of the diagram to another composition in another part ofthe disgrat. ‘Ths ie represents the lcis ofthe langeney points ofthe citve with the equilib Ge lines throughout the whole composton-empeature ‘ange. Such lines, as we sal se later, represent posible tiation paths, i T= = —— = FIGURE 26 Plastic sheet model of riGURE 2.7 Plastic sheet model of a emary temperature-composition dia- ternary temperature-composition dit- gram—two “normal” binary systems gram—iwo binary azeotropes and one ‘and one binary azeotrope. Exch sheet normal binary" system, Each sheet shows the two-phase region at that shows the (wophase region at that temperature and representative ti ines, temperature and represeniative tie lines,MULTICOMPONENT ¥/ FORLIQUID TaUILIMRIA 55 NauRe 28 Isothermal two-phase ternary diagrams from binaries—normal ternary system, Also diagrams of this sort enable reasonable prediction of the existence of valleys and hills or ridges on the surfaces which aid in selecting distillation conditions to obtain betier advantages with regard to maximum changes in composition from plate to plate. 2.3. NONIDEAL VAPOR-LIQUID EQUILIBRIA— MULTICOMPONENT SYSTEMS In the discussion of nonideal vapor-liquid equilibria for multicomponent systems, the subject naturally falls into two categories: nonideal ternary- system vapor-liquid equilibria and nonideal vapor-liquid equilibria of systems composed of more than three components. The reason for this is princix pally the degree of complexity of the equations. There have been many correlations proposed for the enleulation of ternary vapor-liquid equilibria relating activity coefficients and composition, whereas relatively only a few Ihave been proposed for systems having a greater nuinber of components ‘Therefore, the less complex ternary material will be considered first56 VAROR-LIQUID EQUILIBRIA No attempt is made to give a thorough discussion of correlations for multicomponent data because of space considerations. The reader is in- vited to refer to other books on this subject, e.g., Hala et al. (24). Nonideat Ternary Systems ‘The nonideal ternary data needed for ternary disilation calculations may be experimentally determined or calculated from other information. Experi- mental evaluation of ternary vapordiquid data is time consuming and expensive, because approximately # minimum of 100 points of experimental data are necessary (0 characterize completely an extremely nonideal ternary vaporliquid system. If coreelations are available by means of which the necessary data may be predicted, and sufiient information for use in the correlations is at hand or can be reasonably obtained, the ternary data are fegyne a _lipinaemil |YOphagy rn) diagrams {tom binasics~twO normal binaries, one binary azeotrope.MULTICOMPONENT VATOR-LIQUID EQUILINRIA $7 HGURF 2.10 Isothermal iwo:phase ternary diagran's from binaries —one nora! binary, eo binary azeatrapes. usually predicted even at the loss of some accurney. The ideal situation would be one in which ternary vapor-liquid equilibrium data could be pre- dicted from some correlation oF correlations from pure-component data, baccause most of the available information is pure-component data whereas nly a comparatively small amount of data are available on mixture properties involving two or more components. This cannot be done at present with the necessary precision, Calculation of Nonideat Ternary Vapor-Liquid Equilibrium Dara There have been a number of methods proposed for the calculation of ternary vapor-liquid equilibrium data with the use of equations derived From thermodynantic consideration of excess free energy of mixing, These58 VaroR-Liquip EQUILIBRIA FlouRE 2.11 Isothermal ternary diagrams from binaries—three binary azeo- tropes. equations range from purely empirical relationships to semitheoretical re- lations which differ only because of certain assumptions made in their derivation, The various methods may require the use of only binary data, fon one hand, to relatively large amounts of ternary data on the other. van Laar (ype of equations The original van Laar equations (50) were developed only for binary systems. However, several investigators have extended the van Laar type of equations to ternary systems, basing their ‘equations on essentially the same assumptions as those of van Laar but using constants of different form. Wohl [57, 58] through his expansion of the molal excess free-energy equations developed van Laar type of equations of the second, third, and fourth order for ternary systems.MULTICOMFONENT VATOR-LIQUID EQUILIMRIA 59 His second-order or two-sulfix equs Laar constants and composition is for log 4 in terms of binary van sat i nt a Da (Andis) + Aa Aaad?? (2.28) where dy, = Carlson-Colburn modification, van Laar A for systems 1-2 Ison-Colburn modification, van Laar B for system 1-2 ‘arlson-Colburn modification, van Laar A for system 2-3, ete. (Equations for the other components, 2 and 3, ean be written by rotating the subscripts as follows: i an 32 Thus, wherever 1 appears, 2 is substituted, 3 is substituted for 2, and 1 is substituted for 3.) ‘These binary constants are defined as follows: salog ys Ayn d= ogy, (1-4 ERED) @2) wi log yi a= B= ogy (1-4 BED) ean The equations will describe many “normal” ternary systems using only the constants derived from component binary data. However, they do not adequately describe highly nonideal systems. For such systems the higher order, four- and five-suffix equations, should be used along with sufficient multicomponent data to enable evaluation of the interaction terms. The Li and Coull [35] ternary equations of the van Laat type ace om : (ue om Uta + (ulbadx, + (balbadss? ms we tet) (beet Bon Tilda, Fale60 vaPoR-tiquiD EQUILIBRIA Esample 2.2 (e) Determine the binary van Laar constants fromm pure-component data forthe system -cetane-elhylbenaene-Cellosolve at 760 mn Hig pressure. (@) Determine the ternary activity coeficients using Eqs. (2.36), (2.37), and (2.38) nd the caleulated values trom pact (2) for zxo = 025, xan = 0.82, T= 125°C Solution ‘The van der Waals o and 8a Calculated from the following: “The critical properties of moclane and cthylbenrene are reported jn Drcsbach (15. ‘The eitieal properties of Cellosolve ethylene glycal~monoethyl ether were ealeslated from group contributions by the method of Lydetsen.t (Creal properties Components oe — Pam tig | Palm Hg (1) mOctane 168s 2) Ethylbenzene 67.69 0) Celirotve 179.47 The Ay and VB, ate eeaily caleulsted Aw = 13800 9.8) Ay = 17180 16st Awe taue Tye 2321 Ayn omss Vii ~ bas Ayn 05a VT ~ 10419 Ay = 0803 Fy = 2819 1 As reporied in Reid and Sherwood {4}MULTICOMPONENT VAPOR-LIQUID EQUILIBRIA 61 Calculations are to be carted out at 2 = 025, 24 = 052, x) —023, T= 3082K. pny, w [2S2=9.8) + 0.29K0.80099—13.657)]" = [Cait aon) + O52 + HOON) Tiny = 53275 ya had Tiny 482 y= 101 Tin yy = 23526 yy = 1.06, ‘The constants h and & are evaluated from the binary data with the use of the binary equations: ogy = F(Z) a m4) Criog raves = BCP) 34 (fA) 235 ‘The constants may be evaluated by plotting (T log y,)-* versus xy/xy. The slope of the like is (by/Dy)(bylkya)"* and the intercept is (bylk,.)"* The constants may be determined for the other binary systems in a similar manner by plotting the correct terms. These equations correlate the data on a wide range of systems including hydrocarbons, alcohols, Ketones, and water. White [32] presented equations similar in nature to those of Li and Coull, uttizing the same type of binary system constants for evaluation of ternary data. The Bonham (9, 45] ternary equations are of the van Laae type: Vin + tAnVEa an (dia Fab 4a) Tiny (::duVig + seAnV! on PO da tt Bd (xAyVIg + VT 238) TIO daa EF A) The general form for these equations is as follows [Exedy (29) Ee AoP (2.39) Tiny62 _varoR-Liquio equitiaRiA where k = component whose activity coefficient is being evaluated J = reference component, usually the heavy key i= any component from I tom (2.40) (24le) (241) where a and 6 are the van der Waals constants (see Chap. 1). The Bonham and Li and Coull equations are identical when compared on the basis of the same assumptions (see appendix to Chap. 2). They require only binary data for solution, and describe many of the types of nonideal ternary systems. Margules type of equations Wohl [57, 58] in his work on the thermo- dynamics of vapor-liquid equilibria also developed equations of the two-, three-, and four-suffix Margules type. ‘The three-sulfix equation relating the activity coefficient for component I with composition in ternary systems that are symmetrical and require binary and some ternary data for the evaluation of the constants is og ys = xa"ldun + 2ey(duy — Aua)] + 23'lAiy + 2es(Ans — Au) + xpath + dat Ann + ds — Aas — And An = Aunt Ant = Aud) + (= (daa = An) = (1 = 28964) (242) Equations for the activity coefficients for the other two components in terms of composition and binary and ternary constants may be written by using the rotation principle for subscripts Ct =H (arse + can + aian + Oras + Ones + Oss) — Sain (2.43) In simple, normal systems the value of C* = 0 and the ternary data can be found by using only the binary constants derived from the binary data Normal systems are those in which the component characteristics are similar ‘and there is no chemical interaction. When the system is extremely nontideal, ternary data are needed for evaluation of the ternary constant C*, SevernsMULTICOMFONENT VArOR-LIQUID EQUILINRIA 63 et al, [47] proposed a method whereby a minimum of ternary data are required for evaluation of the ternary constant C* which simplifies the complete evaluation of the ternary data The defining equations for the Margules binary constants are 1m 2log ys Ft log ys + EZ (244) A= (2.45) Benedict et al. (6) used a somewhat different expansion of the excess free-energy equation which led to equations essentially of the Margules type. These equations which relate activity coefficient with liquid com- position may he expanded to the desiced order. The two-sulix equations are as follows: RT Mn yy = 2x — Xa) + Ay — HAs — Weataty (2.46) RT In yy 2x —8189)Aiy — Beitady + My — ANIA (2.47) AT In yy = —2e Aig) + 24 — HVA, + Ua — HLVAL, (248) The authors indicate that only a few systems are represented by the Wo- sullix equations. Usually four-suffix equations are needed, and these require evaluation of a number of points of ternary data in order to establish the value of the constants. Empirical Prediction of Ternary Vapor-Liquid Equilibria Edwards otal, [17] sed the correlating equations developed by Gilmont ct a. [20] to predict ternary and molticomponent data from relative volatility data related to activity coeicient, He used an empirical form of equation to evaluate the integral form of equation developed. Schcibel and Friedland (46) developed an empirical method for preiting ternary vaporcliquid equilibria which classed al ternary systems into three groupe: (1) Ternary systems made up of three binary aystems all having the fame sign of deviation from Raoult's law; (2) ternary systems made Up of fone idea! and two nonideal binary systems all showing the same sigh of deviation; and (3) ternary systems miade wp of one nonideal binary system having the opposite sign deviation from those of the other two binaries A diferent graphical method was proposed for exch type of system. For the systems tested, m high degree of accuracy was claimed.rane 2 Calculation of 7, from Maryules Theee-sufix Equation for Example 2.3, Gol 9 = xt + Belly = Aad Dele) a |e | | ® | « @ a) ep OT oe Lm fA | S| Of wee | Pr | ake tas | om | oxo | oms | aos | oo | 02s | oa cos | oone 231 |e | som’ | cas | os | Som | oa | come | cast | ooze 3t2 | ou | omas | om | ons | -oo | om | cows | orm | dota cae ye pom | wm | | we [os 0 | an ref ve | oy | | gone | nies | er oe on | cosa ons 0.700 ols 025 0.00750 0.70750 | 0.03743 231 [om | oom | oms | cos | oo) | ts | og ows | sonst Ha | om | oa | tae | ous | ooo | om | nome | cam | one Cah Wa FA = A= dada k ay) (20) ay a ay | ea eo | e6) a Au Au Ay da da ~ Wee | wx2s 123 | oots | ooss ans | e700 | ones | —oass ous | oss 13 | ooms | oss | cms | toes | Cons | Como | oles | ms iia | oro | ons | oss | ows | moos | Teoes | “toes | “tovas9 C139 m= yb d= paral cn eons | = calgon on @ | | 09 ] o | oo Mae en eet cae |an eto eee lear teres | vewe |e ce 123 | oom | -oms| ons | ono | cos | oas | ooms | on | om | ose 231 | oass | “oss | cass | —oors |- oom | 032 | ome | om | oz | cosso 312 | oro | sans | oss | “oes | —aoes | -oas | -ooinss | 02s | os | o1se0 Seer eet) Colas = = pe 7) ] oo | oo ] @ a @)] o | | | pe | OD | | see | GD an wa | xxa | Sn fala] -oxce 72a | om | on | a2 | oss] 05s | -oow | acm | oz | os | oo 231 | 02 | as | -o | ans | oro | “oois | oem | 32 | -om | —cooi2 3i2 | bas | os | oz | ooas | oats | oo | om | oz7 | ove | oss Gal. 30 = foro 0 1 Cok $= y,for c= 00) peer ® @ oo | op ro oD oo Jel Sea] aden fositea |S? | ade | oan | oetsen |S 123 0.35595, 0.04257 0.10298 1.268 | 0.37095 0.04437 0.10478, 1273 231 | -o1eo | -coosr | come | tae | -oiemo | -oomee | corm | 1007 Siz | ios | eno | oansr | iss | noses | orem | omar | toes66 —varoRLiQuiD QUILiDRIA Example 23 Experimental Margules binaty constants were deterined at 760 mas Hg fr the systern octane-ethylbenzene-Cellosolve as follows ‘Margoles constants a system ‘-Octane-ethylbenzene ‘O85 mOeiane-Cellosolve Eihylbenzene-Cellosolve (@) Compute the ternary activity coeficients at xo = 0.28, sey — 082, using the Wohi form of the Margules equation. (@) Compute the ternary activity coeticiens if the ternary constant C* = ~0.0. Solution (9) In order to apply Eq, (242) andthe corresponding equations for activity coeeients for components 2 and 3, the components are designated 1, 2,3 (octane, ethylbenzene, CCellosolve) and the Matgules constants are sulixed according Ay 2008S Ay = 0700 4, An TOOK Ay O71S Ay = 0455 HO ye 0S2 y= 023 oes In order to cary out the work in a systematic mane, a table i set up. The rotation of subscripts is ullied to setup column headings applicable to each componeat. Result Kcr = 0) Hem 1268 yy LOT yyw 1957 (0) For c* = =003, LAT ye LOT yy = 1.965 Example 24 Assuming the constants given in Example 2.8 were wan Lage constants, determine the ternary activity coefcients for the comporition yun ~ O28, aun = O52 at Py 700 mo Hg, using the Wel ternary twoosulin van Laar equations, Ateume ¢ ~ 125°C Solution Equation (2.24) and correponing equations can be expressed by meant of the following several equation, witha specie form Io be obtained hy rotation of suberipts - 1.42) Tey ATA) sta ATF “The results ofthe tabulated calenations are W126 oy 109 y= 1.9681s ‘ante 22. Calculation of 7, from van Laar Two-suftx Equation for Example 24 Co. = 2h te Tones ol o © @ @ % ea) cere tl rane cee] vee) sus. | ezner = 2 123 | os | ozoo | oms | oo | 10 19 come | aan | aon 231 | om | oosm | oass | aes | iies2 | 13960 | comes | 02s | oosas 312 | eas | aocas | arco | ans | oom: | aos | oom | os | ozs Col = atu paras |e cee ete co em) , 09) ry a pe) | G2 [owen | axis foxtnen | ats | nix | seine | ae 123 | ons | ovo | toe | ioe | oom | oom | oaases | assess | oniseo 231 | eos | cos | io fio” | msn | 10 Tazo | -020° | oanse 312 | ests | o4ss | oases | amuse | omen | tinea | osoous | ~rososs | 213000 Col, 20 Aidt (ty + dn te ma GEE (ta dt rje | (20) en a | a [ao] an eo) | a | an / Sxi2x18 | 17x19 | T+ I4¢20] xs Ox 12 x [2+24246| 2x25 | 2226] y, 123 | ainaie | oacea | osoam | asa | oasoa | ozs | iaoos | sams | orem | i208 231 | ooems | -oorses | ooina | oanez | 025” | ose | oan | ie | coro | to 312 | atom | onsie | aaesio | o2uns | oasom | 023 | osisre | assem | azoao | 136868 —varoR-tiquID eqUiLiARTA Esample 2.5 Using the equations of Benedict eta, dternine the ternary ativiy coeicents forthe moctane-ethylbentene-Cellosolve stem, sg = 028, rn = 0.52 Solution Since Eqs. (2.46) theough (2.48) should reduce to the symmetrical Margules equation foe the binsties (x4 = 0), 2a — 58) 4 Iny = MS were the extreme rghtshand expression is the symmetical Margules equation for ln Then ‘The data given forthe systenn ave unsymimettcal in dhe sense Chat dys Fay AS An approximation it wil be necessary to use an average, a8 Equations (2.46) dirougs (248) then reduce 19 ayy = (er xt (te = sande ated ‘The data are: Ay 0088 Ay 01018 dy = 0.420 mn 025 ye O52 y= 023 Solving In ys = (052 ~ 0.190085) + (023 — 0.0578%0.7078) ~ (0.11960.420) Inyy = 010896 yy = 123 0.01589 yy = 1007 lay 02981 yy = 1968 Chao and Hougen [12] applied the Th and Dodge [28] correction to the Gibbs-Duhem equation for isobaric conditions ancl combined this with the expression for excess free energy. This resulted in equations of the form Pa in? ¥ fl bale — a) + cul28y — Oy — Fafa, — baa — Gas(284 = 89) + celts — 25) — bulsy — = elses 5) 403 A] = lq = 9) baby = ll Feasts = (4 =" ] 2.49)MULTICOMPONENT VAPOR-LIQUIO EQUILIBRIA 69) The rotation principle ean be used to get other ratios of In(ydy.). The Redlich and Kister [43] fourth-order equations for ternary systems are similar in nature: Jog ttm busy — 2) ~ eles — 39 — Davee + ls — 5) % (54 — ay) = Aye) + albus # enl2my — 2) + die, — WIN — 2) = bay = enone = a9) — daly — IGN — 4) + oO — 4) + Dyxi2xy — 3) + Dyri(y — 25)] (2.50) For correlation of ternary data, the Wohl three- and four-suffix equations utilizing some ternary data to evaluate the ternary constants are reasonably accurate, Where no ternary data are available, the Li-Coull and Bonhain equations have proved satisfactory and reliable for most ternary systems, using only binary data, The Redlich-Kister and Chao-Hougen equations require evaluation of some ternary constants to obtain accurate reproduction of experimental data. Finally, the Bonham type of equations are good for very rough approximation only, using pute-component data to evaluate the binary constants. Presentation of Data Ternary data may be presented in tabular form or in graphical form. The tabular data are exact, but this advantage is lost when inter- polation is required, and therefore graphical presentation is generally more useful Such graphs may take the form of activity coefficient of a component in solution versus the ternary compositions. Three graphs are necessary such 2.12. as shown in Fi rioune 2.12 Activity-coeticient composition curves.70 VaROR-LIQUID EQUILIORIA S we Figuae 2.13 Relative volatility composition curves. ‘The relative volatility of two components based on the third may be plotted for constant values of composition af the secand co Asa third alternative the actual y: versus x diagram may be plotted on a ternary diagram with parameters of xy, x4 versus yy and yy (Fig. 2.14), pment (see Fig. 2.13) Nonideal Multicomponent Systems Black Equations Black {7,8} utilized modifications of the van Laar equations to relate activity coefficients and composition for multicomponent systems. He FioURE 2.14 Ternary vapor-liquid compositionsMULTICOMPONENT VAFOR-LIQUID EOQUILIRRIA 71 distinguished between classes of components (“class™ constituting a homo- ogous series made up of a number of components), class variables, and com- ponent variables. The general equation for activity coefficient in terms ‘of composition and class designation is ZluReeyt +E Bintee) eee toe, aa) +E, ast) where Fu SNe — KZ eed) + 2E [os = xn eaxe)] 33 [on tar(Zenrrn)] 252 and where = 7 Bern = RraRyalldy — a)* — ante) (2.83) The terms R and Af refer to two classes of compounds, and r and m are variables denoting compounds within classes R and M.S refers to a par- ticular class and i,j, and k refer to specific compounds within classes. The compositions x, and x, refer to mole fractions of the individual components in classes R and Mf, respectively. The sums of compositions within the classes are Xp and Xy,, respectively. Black established a number of rules for determining the values of the variables and coefficients in the equations for handling the different situations by use of his equations. For the various systems considered which included both ternary and multicomponent systems and hydrocarbons, alcohols, ketones and water, the method gave very good results when the calculated data are compared with the experimental data. For complete details concerning the method, see Black [7, 8). Witson Equation A method of eateulating multicomponent vapor-Jiquid equilibria by means of the Wilson [56] equation, which extended the theoretical equation of Flory [18] and Huggins (27), has been proposed by Orye and Prausnitz (39) “They utili the excess iee-energy coricept to express solution nonideality and present a generalized equation (2.54) for multicomponent systems tayym —in (Saudy) 1S ae ast) a72 varor-tiquip eQuiLiuNA where Ye etity-soney Aum feet 1 (2.55) eetaeaninry (2.56) Only binary interaction terms are required in the generalized equation, Therefore, activity coefficients for the compounds in multicomponent systems can be calculated from the binary activity coefficients determined experi- mentally ABLE 2.3 Summary of Actvity-Coeficient Equations for Multicomponent System Reference arenes Application Dats needed Woh (57.38) van Lane “Ternary | Binary, some aT very Bonkarn {9,45} ee Teena. = xT | snslacorspanent | scout 5) van base | Ternary Binaty rer | | Wont (7, $8 sarguler | Ternary inary, some ‘nulicomponent Redieh-Kistr 9) | Mangler Temary inary, some ee ternary Benedict eal (6) | Margues Ternary Binary, some Jone ‘ernaty Dick 1,61 san Laar Tennay Binary, some pel ‘aulicomponent | multicomponent Witon (56,39 Flory-tuggins | ery 1 nary ye | aticonponen | white (82 cantons} Feary inary | Weimerseausnitg (51) | inary. ternary, {Pure component, | mulcogtee|Se edsaraset att) | Empieat Tema. inary, some Selwiteltieuland M6] Fmpiccal | Teraty | inaty imurncouronel vioRuguio ravi 73 Although theoretically only one point of x, yy 7, and P data is required to evaluate A,, and Ay, the proponents of the method state that a series of data points gives more reliable result. For a binary solution the activity-coeffcient equations are An Aw tay —In(s, + Ane) + a(t — 4) sn A An In yy —In (xy + Ane) — a(n - mh) (28) Using the values of Ay, and Ay, determined from binary data, the activity coefficients for the components in a multicomponent mixture can be cal- culated from the component binaries. ‘The authors predicted the vapor-liquid data on some-65 systems using the Wilson parameters and found that the predicted and experimental data agreed very well. Comparison with the predicted data from the van Laar ‘and Margules equations showed that those derived from the Wilson equation ‘were as good in most cases and betler in many others. ‘The disadvantage of this type of correlation is that it does not apply to partially miscible systems, whereas the Black method is applicable. How ever, the number of constants needed for this correlation is less than required for the Black correlation, and thus fewer data are needed. Equilibrium Vaporization Ratios Equilibrium vaporization ratios or X values have proved to be extremely useful in hydrocarbon distillation, absorption, and stripping calculations. Kis defined as Kyat! 2.59) In terms of vapor pressure, acti Kw DE (2.60) When the system can be considered ideal, y, = 1.0 and », = 1.0, and the equilibrium vaporization ratio reduces to Kod 261) K-values for a particular compound are a function of temperature, pressure, composition of the mixture, and the amount of the compound in the mixture,74 VAROR-LIQUID EQUILIBRIA ‘There are in general two principal methods for obtaining K values—the thermodynamic method and the empirical method. The thermodynamic method involves the use of equations which describe all ofthe thermodynamic properties of the phases including the fugacity, while the empirical approach involves the prediction of K values without particular regard to thermo- dynamic properties which represent the best fit to the experimental éata available for derivation of the correlation, Thermodynamic Methods for K The thermodynamic method normally requires an equation of state to describe phase behavior of mixtures. Benedict et al. [2-5] developed a practical equation describing the behavior of light hydrocarbons and theit mixtures, This equation is similar in form to the Beattie-Bridgeman (1] equation of state, The equation includes eight constants evaluated from vapor pressure, critical constants, and pressure, volume, and temperature properties of the pure component. The coefficients describing the mixture behavior were calculated from the individual pure-component constants, ‘The equation, while accurate to a high degree, must be solved by a com plicated trial-and-error solution for each point and requires the use of a high-speed digital computer. ‘A practical correlation for primarily paraffinic hydrocarbon X values was developed by Benedict et al. by using molal average boiling point as a bulk phase parameter, Fugacity coefficients and activity coefficients were Getermined over wide ranges for the different variables, and were plotted on usable charts as functions of temperature, pressure, and miolal average boiling point of the two phases. These comprise the Kellogg [30] charts De Priester [14] presented the same fugacity functions on pressure, temperature, and composition charts for a pressure range of 0 to 1000 psia, He also constructed nomographs to show average A's for first trai calcula ions, The De Priester chatts were an improvement because only two charts were needed for ench hydrocarbon and the entire pressure range was ine cluded on each chart. However, use of the charts for exact determination of K values requires trial-and-error calculations Edmister and Ruby (16) developed generalized correlations of activity coeflicients based on the fugacity functions developed by Benedict et al ‘The activity coefficients for both liquid and vapor phases were correlated with reduced temperature, reduced pressure, and ratio of component boiling point to the molal average boiling point of the phases. This is a further improventent in determining values but still trial-and-error caleulations are required. The Edmister-Ruby correlations cover a pressure range 10 3600 psiaMULTICOMPONENT VATOR-LIQUID EQUILIORIA 75 Chao and Seader {13} developed a generalized correlation for K values in hydrocarbon mixtures using three factors: gy, the fugacity coeficient of component fin the vapor mixtures r, the fugacity coelTeient of pure lquid at system conditions; and y, the activity eoeficient of component fin liquid solution. mm K (2.62) % Methods and equations for evaluation of the three factors in Eq. (2.62) are given in the original reference and are suited to solution using a high-speed digital computer. The authors claim an over-all average deviation of calculated from experimental values of 8.7%. This method is claimed to apply to mixtures of paraffin, aromatic, naphthenic, and olefinic hydro- carbons and inert gases such as hydrogen. Hoffman et al. 25} presented a method for evaluation of fugacitics of the components in a multicomponent liquid mixture to aid in predicting vaporization equilibrium ratios for multicomponent mixtures. Prausnitz et al, [42] used the Hildebrand solubility parameter to calculate X values for light hydrocarbons in paraffinic, naphthenic, and aromatic absorption oils with reasonably good results. The foregoing methods represent only a few of the many which utilize some of the various equations of state to evaluate individual and mixture properties from which K values for various materials can be calculated, Readers interested in this aspect of K-value determination should refer to modern thermodynamics books and the copious literature on the subject. Extended discussion of this phase of the subject is considered beyond the scope of this book Empirical Methods for K ‘The empirical methods vary greatly in detail but generally are based upon ting curves to experimental data points to get the best agreement and (0 devise suitable means of extrapolating and interpolating for other conditions while maintaining the same relative accuracy, Essentially these correlations involve modifying the pure-component fugacity relations by means of a “convergence pressure” parameter which corrects for the behavior of the components in a mixture and thus includes composition effects, A number of such correlations have been devised (10, 11, 21-23, 29, 36-28, 40, 41, 48, 49, 54, 55}. Table 2.4 lists some of the general equilibrium vaporization ratio, or K, correlations. In most enses these are for paraffinic and olefinic hydro- carbons, but may present K values for hydrogen, hydrogen sulfide, sulfur dioxide, carbon dioxide, nitrogen, water, as well as aromatic and saturated76 varon-tiquip equitionin TAOLE 24 Equilibrlum Vaporization Ratios pecewe | PEE Ten | reaurevnge Winn 41 Nomographs | HC |“wo-se0 | aim=i0.0 pia vim 3) Nomoprois | He foam | Aim-io000 fsa Madden

You might also like

• The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  

  From Everand

  The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  Mark Manson

  Rating: 4 out of 5 stars

  4/5 (5814)
• The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  

  From Everand

  The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  Brené Brown

  Rating: 4 out of 5 stars

  4/5 (1092)
• Never Split the Difference: Negotiating As If Your Life Depended On It

  

  From Everand

  Never Split the Difference: Negotiating As If Your Life Depended On It

  Chris Voss

  Rating: 4.5 out of 5 stars

  4.5/5 (845)
• • Magazines
  • Podcasts
  • Sheet music
• Principles: Life and Work

  

  From Everand

  Principles: Life and Work

  Ray Dalio

  Rating: 4 out of 5 stars

  4/5 (609)
• The Glass Castle: A Memoir

  

  From Everand

  The Glass Castle: A Memoir

  Jeannette Walls

  Rating: 4.5 out of 5 stars

  4.5/5 (1717)
• Grit: The Power of Passion and Perseverance

  

  From Everand

  Grit: The Power of Passion and Perseverance

  Angela Duckworth

  Rating: 4 out of 5 stars

  4/5 (590)
• Sing, Unburied, Sing: A Novel

  

  From Everand

  Sing, Unburied, Sing: A Novel

  Jesmyn Ward

  Rating: 4 out of 5 stars

  4/5 (1104)
• Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  

  From Everand

  Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  Margot Lee Shetterly

  Rating: 4 out of 5 stars

  4/5 (897)
• Shoe Dog: A Memoir by the Creator of Nike

  

  From Everand

  Shoe Dog: A Memoir by the Creator of Nike

  Phil Knight

  Rating: 4.5 out of 5 stars

  4.5/5 (540)
• The Perks of Being a Wallflower

  

  From Everand

  The Perks of Being a Wallflower

  Stephen Chbosky

  Rating: 4.5 out of 5 stars

  4.5/5 (2104)
• The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  

  From Everand

  The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  Ben Horowitz

  Rating: 4.5 out of 5 stars

  4.5/5 (348)
• Bad Feminist: Essays

  

  From Everand

  Bad Feminist: Essays

  Roxane Gay

  Rating: 4 out of 5 stars

  4/5 (1018)
• Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  

  From Everand

  Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  Ashlee Vance

  Rating: 4.5 out of 5 stars

  4.5/5 (474)
• Her Body and Other Parties: Stories

  

  From Everand

  Her Body and Other Parties: Stories

  Carmen Maria Machado

  Rating: 4 out of 5 stars

  4/5 (822)
• The Outsider: A Novel

  

  From Everand

  The Outsider: A Novel

  Stephen King

  Rating: 4 out of 5 stars

  4/5 (1866)
• The Emperor of All Maladies: A Biography of Cancer

  

  From Everand

  The Emperor of All Maladies: A Biography of Cancer

  Siddhartha Mukherjee

  Rating: 4.5 out of 5 stars

  4.5/5 (271)
• The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  

  From Everand

  The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  Viet Thanh Nguyen

  Rating: 4.5 out of 5 stars

  4.5/5 (122)
• Angela's Ashes: A Memoir

  

  From Everand

  Angela's Ashes: A Memoir

  Frank McCourt

  Rating: 4.5 out of 5 stars

  4.5/5 (441)
• Brooklyn: A Novel

  

  From Everand

  Brooklyn: A Novel

  Colm Tóibín

  Rating: 3.5 out of 5 stars

  3.5/5 (1947)
• The Little Book of Hygge: Danish Secrets to Happy Living

  

  From Everand

  The Little Book of Hygge: Danish Secrets to Happy Living

  Meik Wiking

  Rating: 3.5 out of 5 stars

  3.5/5 (401)
• A Man Called Ove: A Novel

  

  From Everand

  A Man Called Ove: A Novel

  Fredrik Backman

  Rating: 4.5 out of 5 stars

  4.5/5 (4770)
• The World Is Flat 3.0: A Brief History of the Twenty-first Century

  

  From Everand

  The World Is Flat 3.0: A Brief History of the Twenty-first Century

  Thomas L. Friedman

  Rating: 3.5 out of 5 stars

  3.5/5 (2259)
• Steve Jobs

  

  From Everand

  Steve Jobs

  Walter Isaacson

  Rating: 4.5 out of 5 stars

  4.5/5 (808)
• The Yellow House: A Memoir (2019 National Book Award Winner)

  

  From Everand

  The Yellow House: A Memoir (2019 National Book Award Winner)

  Sarah M. Broom

  Rating: 4 out of 5 stars

  4/5 (98)
• Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  

  From Everand

  Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  Gilbert King

  Rating: 4.5 out of 5 stars

  4.5/5 (266)
• The Art of Racing in the Rain: A Novel

  

  From Everand

  The Art of Racing in the Rain: A Novel

  Garth Stein

  Rating: 4 out of 5 stars

  4/5 (4203)
• A Tree Grows in Brooklyn

  

  From Everand

  A Tree Grows in Brooklyn

  Betty Smith

  Rating: 4.5 out of 5 stars

  4.5/5 (1929)
• A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  

  From Everand

  A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  Dave Eggers

  Rating: 3.5 out of 5 stars

  3.5/5 (231)
• Yes Please

  

  From Everand

  Yes Please

  Amy Poehler

  Rating: 4 out of 5 stars

  4/5 (1898)
• Team of Rivals: The Political Genius of Abraham Lincoln

  

  From Everand

  Team of Rivals: The Political Genius of Abraham Lincoln

  Doris Kearns Goodwin

  Rating: 4.5 out of 5 stars

  4.5/5 (234)
• The Woman in Cabin 10

  

  From Everand

  The Woman in Cabin 10

  Ruth Ware

  Rating: 3.5 out of 5 stars

  3.5/5 (2522)
• Fear: Trump in the White House

  

  From Everand

  Fear: Trump in the White House

  Bob Woodward

  Rating: 3.5 out of 5 stars

  3.5/5 (738)
• Wolf Hall: A Novel

  

  From Everand

  Wolf Hall: A Novel

  Hilary Mantel

  Rating: 4 out of 5 stars

  4/5 (3811)
• John Adams

  

  From Everand

  John Adams

  David McCullough

  Rating: 4.5 out of 5 stars

  4.5/5 (2409)
• On Fire: The (Burning) Case for a Green New Deal

  

  From Everand

  On Fire: The (Burning) Case for a Green New Deal

  Naomi Klein

  Rating: 4 out of 5 stars

  4/5 (74)
• The Light Between Oceans: A Novel

  

  From Everand

  The Light Between Oceans: A Novel

  M.L. Stedman

  Rating: 4.5 out of 5 stars

  4.5/5 (789)
• Manhattan Beach: A Novel

  

  From Everand

  Manhattan Beach: A Novel

  Jennifer Egan

  Rating: 3.5 out of 5 stars

  3.5/5 (792)
• The Constant Gardener: A Novel

  

  From Everand

  The Constant Gardener: A Novel

  John le Carré

  Rating: 3.5 out of 5 stars

  3.5/5 (104)
• The Unwinding: An Inner History of the New America

  

  From Everand

  The Unwinding: An Inner History of the New America

  George Packer

  Rating: 4 out of 5 stars

  4/5 (45)
• Rise of ISIS: A Threat We Can't Ignore

  

  From Everand

  Rise of ISIS: A Threat We Can't Ignore

  Jay Sekulow

  Rating: 3.5 out of 5 stars

  3.5/5 (137)
• Little Women

  

  From Everand

  Little Women

  Louisa May Alcott

  Rating: 4 out of 5 stars

  4/5 (104)