CYCLIC DISTILLATION FOR ENERGY CONSERVATION IN SPIRIT PRODUCTION By Nicole Elizabeth Shriner A DISSERTATION Submitted to Michigan State University i n partial fulfillment of the requirements for the degree of Chemical Engineering Doctor of Philos ophy 2018 ABSTRACT CYCLIC DISTILLATION FOR ENERGY CONSERVATION IN SPIRIT PRODUCTION By Nicole Elizabeth Shriner Distillation technology has been used worldwide throughout the chemical industry for many years, originally being invented in China aroun d 800 BC. There are currently over 40,000 columns in operation worldwide [29]. Although there are many variations in column types, sizes, tray configurations, etc., the underlying princip le is the same. Distillation is the phy sical process of heating a liq uid mixture to separate components based the relative volatilit ies of each component. Distilleries consume about 40% of the total energy used to operate plants in the chemical industry, and over 95% of th at energy is used in s eparation processes [31]. Ther efore, improving distillation technology is one area that chemical engineers are working to transform into a more efficient process by use of process intensification. The purpose of this research was to investigate the possibl e economic and energy impacts of a form of process intensification Cyclic distillation is an alternate mode of operation first proposed and studied by Cannon and Mc flow and liquid flow. During the vapor flow period heat is supplied to the column, vapor flow upwards and distillate is collected as in normal operation. During the liqui d flow period, the heat source is ceased and the liquid on each tray is transferred to the tray below. The cycle of alternating between the two periods is continued for the entire distillation. Cyclic distillation has been applied to many different systems and configurations and has be en demonstrated experimentally and theoretically to increase column throughput, lower energy requirements and achieve higher separation performance. In this study, a 150 L Carl © still was used to distill fermented apple cider and apple brandy low wines. C yclic distillation and conventional operation were compared using the same system. Distillation samples were collected and analyzed using a gas chromatograph. The system was also simulated using MAT LAB . In summary, cyclic distillation on the batch column u sed was able to show a decrease in energy (steam) requirements for finishing runs but not in stripping runs. Cyclic distillation trends included ethanol concentration decreased at a slower rate compared to conventional operation and as a result temperature profiles mimicked this phenomenon. It was shown that the volume of hearts (product) was increased and volum e of tails and heads (unwanted by - products) was decreased. This research has shown that the application of cyclic distillation on spirit production is a viable option for distilleries large and small. In extension of this work, it is suggested that cyclic distillation is applied to other types of spirits and to columns with a larger number of trays, and to trays with true plug flow capability. iv This dissertation is dedicated to my mom , Dr. Tamara Shriner . Thank you for pushing me t o achieve my highest potential even when I was slightly opposed to continue my academic path. Thank you for instilling in me that I can do anything I set my min d to. Thank you for showing me how a hardworking and independent woman carries herse lf. I am forever indebted and grateful to you. I love you. v ACKNOWLEDGEMENTS I would like to acknowledge and thank my advisor, Dr. Kris Berglund. Without him I would no t have found my passion in the fermented beverage industry or have the connections and opportunities that I have and will continue to receive. His confidence in me since un der grad has helped to instill confidence in myself and in my work. I would also like to thank my colleagues Jacob Rochte, Yasheen Jadidi, and John Szfransk i. Without their help I would likely still be googling how to fix things in the lab, getting a lot less hours of sleep, and still working on errors in my mat lab code. THANK YOU! vi TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ....................... viii LIST OF FIGURES ................................ ................................ ................................ ....................... ix KEY TO SYMBOLS ................................ ................................ ................................ .................. xvii CHAPTER 1: Introduction ................................ ................................ ................................ ............. 1 1.1 Distillation Theory and Background ................................ ................................ ..................... 1 1.2 Distillatio n Improvement and Cyclic Distillation ................................ ................................ . 2 1.3 Cyclic Distillation Application to Spirits Industry ................................ ................................ 3 CHAPTER 2: Literature Review ................................ ................................ ................................ .... 7 CHAPTER 3: Modeling Overview ................................ ................................ ............................... 30 3.1 Batch Distillatio n ................................ ................................ ................................ ................. 30 3.1.1 Alembic Distillation The Rayleigh Equation ................................ ............................ 30 3.1.2 Multistage Batch Distillation ................................ ................................ ........................ 31 3.1.3 Stage - By - Stage M ethods for Batch Rectification ................................ ........................ 34 3.1.4 Governing Equations and Thermodynamic Model for System in Study ..................... 38 3.2 Cyclic Continuous Distillation Modeling Approach ................................ ........................... 40 3.2.1 Assumptions ................................ ................................ ................................ ................. 40 3.2.2 Operationa l Constraints ................................ ................................ ................................ 40 3.2.3 Model of Vapor Flow Period ................................ ................................ ........................ 41 3.2.4 Model of Liquid Flow Period ................................ ................................ ....................... 41 3.2.5 Solution Method ................................ ................................ ................................ ........... 42 3.3 Design of Cyclic Distillation ................................ ................................ ............................... 43 3.3.1 Design Methodology ................................ ................................ ................................ .... 43 3.4 Batch Cyclic Distillation Dynamic Analysis ................................ ................................ ...... 45 3.4.2 Vapor Flow Period ................................ ................................ ................................ ....... 46 3.4.3 Liquid Flow Period ................................ ................................ ................................ ....... 48 3.4.4 Solution Method ................................ ................................ ................................ ........... 51 CHAPTER 4: Methods ................................ ................................ ................................ ................. 52 4.1 Materials and Equipment ................................ ................................ ................................ .... 52 4.1.1 Manufacturing Equipment ................................ ................................ ............................ 52 4.2 Measurement Methods ................................ ................................ ................................ ........ 53 4.2.1 GC Method ................................ ................................ ................................ ................... 53 4.3 Research Design ................................ ................................ ................................ .................. 55 4.4 Procedures ................................ ................................ ................................ ........................... 56 4.5 Data Collection and Analysis ................................ ................................ .............................. 56 CHAPTER 5: Results ................................ ................................ ................................ ................... 58 vii 5.1 Results ................................ ................................ ................................ ................................ . 58 5.1.1 All Distillation Results ................................ ................................ ................................ . 58 5.1.2 Component Concentration Results ................................ ................................ ............... 61 5.1.3 R esults by Type of Distillation ................................ ................................ ..................... 65 5.1.4 Results o f Cuts Taken ................................ ................................ ................................ ... 68 5.1.5 Temperature Profile Results ................................ ................................ ......................... 70 5.2 Simulation Results ................................ ................................ ................................ ............... 71 5.3 Reproducibility ................................ ................................ ................................ .................... 80 CHAPTER 6: Discussion ................................ ................................ ................................ .............. 81 6.1 Summary ................................ ................................ ................................ ............................. 81 6.2 Conclusions ................................ ................................ ................................ ......................... 81 6.2.1 All Distillation Discussion ................................ ................................ ........................... 81 6.2.2 Component Concentration Trends ................................ ................................ ................ 82 6.2.3 Low Wines v s Finishing Tends ................................ ................................ .................... 83 6.2.4 Effects on Cuts ................................ ................................ ................................ ............. 84 6.2.6 Reflux Observations and Discussion ................................ ................................ ............ 85 6.2.7 Simulation Discussion ................................ ................................ ................................ .. 85 6.3 Limitations and Sources of Error ................................ ................................ ........................ 86 6.4 Recommendations for Future Research and Application ................................ .................... 87 APPENDICES ................................ ................................ ................................ .............................. 88 APPENDIX A: Graphical Distillation Results ................................ ................................ .......... 89 APPENDIX B: MATLAB Codes ................................ ................................ ........................... 133 APPENDIX C: Miscellaneous Figures/Tables ................................ ................................ ....... 142 B IBLIOGRAPHY ... . 143 viii LIST OF TABLE S Table 2.1 Definition of Va rious Efficiencie s [5] ................................ ................................ .......... 10 Table 2.2 Effects of Mixing During the Liquid Flow Period on the Separating Ability of a Controlled C ycle Rectification Still at Total Reflux [5] ................................ ............................... 13 Table 2.3 Effect of Relative Volatility on the Overall Column Efficiency of a Con trolled Cycling Rectification Still at Total Reflux [5] ................................ ................................ ........................... 13 Table 3.1 Variables an d Total Number of Equations for Multicomponent Batch Distillation ..... 37 Table 4.1 Detected Compounds [44] ................................ ................................ ............................ 54 Table 4.2 Distillation Conditions ................................ ................................ ................................ .. 55 Table 5.1 Distillat ion Efficiency Results ................................ ................................ ...................... 58 Table 5.2 Distillations Separated by Type ................................ ................................ .................... 65 Table 5.3 Cut Results for Distillations 11, 12, 13, 14 & 15 ................................ .......................... 68 Table 5.4 Parameters for Batch Distillation Simulation ................................ ............................... 72 Table 5.5 Statistical Analysis of Repro ducibility ................................ ................................ ......... 80 Table B.1 Cyclic Distillation Simulation Data ................................ ................................ ........... 141 ix LIS T OF FIGURES Figure 2.1 Cycling Increases Column Capacity for Type - Two Plates [2] ................................ ...... 7 Figure 2.2 Compositions as a Function of Fraction of Plate Holdup Dropped ( ) [ 6 ] .......... 9 Figure 2.3 Effective Plate Efficiency as a Function of Plate Number [ 6 ] .............................. 10 Figure 2.4 Effect of Fraction of Plate Holdup Dropped During the Liquid Flow Period on the Over - all Efficiency of a Controlled Cycling Rectification Still at Total Reflux [5] .................... 12 Figure 2.5 E ffective Plate Efficiencies (E 0 ) in a Controlled Cycling Rectification Column Still at Total Reflux [5] ................................ ................................ ................................ ............................. 14 Figure 2.6 (a) Dependence of Column Separation Efficiency E on Mean Vapor Velocity (b) Dependence of Column Separation Efficiency E on Vapor Feed Time and on Liquid - feed Time (c) Dependence of Column Separation Efficiency E on the Ratio of Feed Periods of Phases for Various Velocities [13] ................................ ................................ ..................... 17 Figure 2.7 Schematic of Column Used for Stepwise Period Distillation [18] .............................. 19 Figure 2.8 New Tray Design for Period Cycling Distill ation [23] ................................ ............... 21 Figure 2.9 The Three Characteristic Periods in the Cyclic Operation of a Regu lar Batch Distillation Column [39] ................................ ................................ ................................ ............... 22 Figure 2.10 L ab Batch Column Used to Apply a Batch Cyclic Operating Theory [39] ............... 23 Figure 2.11 Batch Rectification Control Diagram [25] ................................ ................................ 26 Figure 2.12 Batch Stripping Control Diagram [25] ................................ ................................ ...... 26 Figure 2.13 Mass Exchange Contact Device and Column [28] ................................ .................... 29 Figure 3.1 Sim ple Batch Distillation Schematic ................................ ................................ ........... 30 Figure 3.2 Multistage Batch Distillation ................................ ................................ ....................... 32 Figure 3.3 McCabe - Thiele Diagr am for Multistag e Batch Distillation of Ethanol/Water with Varibale Reflux ................................ ................................ ................................ ............................. 33 Figure 3.4 Schematic of Cyclic Distillation on Carl Still ................................ ............................. 38 x Fig ure 3.5 Comparison of Energy Requirements in Cyclic Distillation versus Conventional Distillation [26] ................................ ................................ ................................ ............................. 43 Figure 3.6 Cyclic Distillation Design Approach ................................ ................................ ........... 45 Figure 4.1 Manufactured Sampling Apparatus ................................ ................................ ............. 53 Figure 5.1 Ethanol Concentration in Distillate vs Time for All Distillations ............................... 59 Figure 5.2 Ethanol Concentration vs Volume for Distillations D7 D15 ................................ .... 60 Figure 5.3 Normalized Ethanol Concentration vs Volume for Distillations D7 - D15 .................. 60 Figure 5.4 Normalized A cetaldehyde Concentration vs Volume for Distillations D7 - D15 ......... 61 Figure 5.5 Normalized Aceton e Concentration vs Volume for Distillations D7 - D15 ................. 61 Figure 5.6 Normalized Ethyl Acetate Concentration vs Volume for Distillations D7 - D15 ......... 62 Figure 5.7 Normalized Methanol Concentration vs Volume for Distillations D7 - D15 ............... 62 Figure 5.8 Normalized Propanol Concentration vs Volume for Distillations D7 - D15 ................ 63 Figure 5.9 Normalized Isobutanol Concentration vs Volume for Distillations D7 - D15 .............. 63 Figure 5.10 Normalized Butanol Concentration vs Vo lume for Distill ations D7 - D14 ................ 64 Figure 5.11 Normalized Isoamyl Alcohol Concentration vs Volume for Distillati ons D7 - D15 .. 64 Fi gure 5.12 Hydrom eter ABV Reading vs Fraction of Volume Distilled for Low Wine Distillations ................................ ................................ ................................ ................................ ... 65 Figure 5.13 Steam Efficiency for All Low Wine Distillations ................................ ..................... 66 Figure 5.14 Percent Recovery for All Low Wine Distillations ................................ .................... 66 Figure 5.15 Hydrometer ABV Reading vs Fraction of Volume Distilled for Finishing Distillations ................................ ................................ ................................ ................................ ... 67 Figure 5.16 Steam Efficiency for All Finishing Distillations ................................ ....................... 67 Figure 5.17 Percent Recovery for All Finishing Distillation s ................................ ...................... 68 Figure 5.18 Normalized Proof Gallon of Heads Cut by Distillation ................................ ............ 69 xi Figure 5.19 Normailized Proof Gallon of Hearts Cut by Distillation ................................ ........... 70 Figure 5.20 Normailized Proof Gallon of Tails Cut by Distillation ................................ ............. 70 Figure 5.21 Temperature vs Time Profile for Conventional Operation (D11) ............................. 71 Figure 5.22 Temperature vs Time Profile for Cyclic Operation (D13) ................................ ........ 71 Figure 5.23 MATLAB S imulation of Con ventional Distillation: Composition vs Time ............. 73 Figure 5.24 MATLAB Simulation of Conventional Distillatio n: Temperature vs Time ............. 74 Figure 5.25 MATLAB Simulation of Conventional Distillation: Moles in Pot vs Time ............. 75 Figure 5.26 MATLAB Simulation of Conventional Distillation: Moles in Distillate vs Time .... 76 Figure 5.27 Stage and Reboiler Composition vs Time for First Vapor Flow Period ................... 77 Figure 5.28 Stage and Reboiler Temperatu re vs Time for F irst Vapor Flow Period .................... 78 Figure 5.29 Cyclic Batch Distillation Tray and Still Composition vs Time for 25 Cycles .......... 79 Figure 5.3 0 Cyclic Batch Distil lation Tray and Still Temperature vs Time for 25 Cycles ......... 79 Figure A.1 Normal Operation 1, Tray 1 Graphical Results ................................ .......................... 89 Figure A.2 Normal Operation 1, Tray 2 Graphical Results ................................ .......................... 89 Figure A.3 Normal Operation 1, Tray 3 Graphical Results ................................ .......................... 90 Figure A.4 Normal Operation 2 , Tray 1 Graphical Results ................................ ......................... 90 Figure A.5 Normal Operation 2 , Tray 2 Graphical Results ................................ ......................... 91 Figure A.6 Normal Operation 2 , Tray 3 Graphical Results ................................ ......................... 91 Figure A.7 Normal Operation 3 , Tray 1 Graphical Results ................................ ......................... 92 Figure A.8 Normal Operation 3 , Tray 2 Graphical Results ................................ ......................... 92 Figure A.9 Normal Operation 3 , Tray 3 Graphical Results ................................ ......................... 93 Figure A.10 Cyclic Distillation 1 (Vapor 9 min, Liquid 3 min) , Tray 1 Graphical Results ........ 93 Figure A.11 Cyclic Distillation 1 (Vapor 9 min, Liquid 3 min) , Tray 2 Graphical Results ........ 94 xii Figure A.12 Cyclic Distillation 1 (Vapor 9 min, Liquid 3 min) , Tray 3 Graphical Results ........ 94 Figure A.13 Cyclic Distillation 2 (Va por 6 min, Liqui d 3 min) , Tray 1 Graphical Results ........ 95 Figure A.14 Cyclic Distillation 2 (Vapor 6 min, Liquid 3 min) , Tray 2 Graphical Results ........ 95 Figure A.15 Cyclic Distillation 2 (Vapor 6 min, Liquid 3 min) , Tray 3 Graphical Results ........ 96 Figure A. 16 Cyclic Distillation 3 (Vapor 4 min, Liquid 3 min) , Tray 1 Graphic al Results ........ 96 Figure A.17 Cyclic Distillation 3 (Vapor 4 min, Liquid 3 min) , Tray 2 Graphical Results ........ 97 Figure A.18 Cyclic D istillation 3 (V apor 4 min, Liquid 3 min) , Tray 3 Graphical Results ........ 97 Figure A.19 Normal Operation Low Wines , Tray 1 Graphical Results ................................ ... 98 Figure A.20 Normal Operation Low Wines , Tray 2 Graphical Results ................................ ... 98 Figure A.21 Normal Operation Low Wines , Tray 3 Graphical Results ................................ ... 99 Figure A.22 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 9 min, Liquid 3 min) Tray 1 Graphical Results ................................ ................................ ................................ .............. 99 Figure A.23 Cyclic Operation Low Wines , Cyclic Distill ation (Vapor 9 min, Liquid 3 min) Tray 2 Graphical Results ................................ ................................ ................................ ............ 100 Figure A.24 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 9 min, Liquid 3 min) Tray 3 Graphical Resul ts ................................ ................................ ................................ ............ 100 Figure A.25 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 6 min, Liquid 3 min) Tray 1 Graphical Results ................................ ................................ ................................ ............ 101 Figure A.26 Cycl ic Operation Low Wines, Cyclic Distillation (Vapor 6 min, Liquid 3 min) Tray 2 Graphical Results ................................ ................................ ................................ ............ 101 Figure A.27 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 6 m in, Liquid 3 min ) Tray 3 Graphical Results ................................ ................................ ................................ ............ 102 Figure A.28 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 4 min, Liquid 3 min) Tray 1 Graphical Results ................................ ................................ ................................ ............ 102 Figure A.29 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 4 min, Liquid 3 min) Tray 2 Graphical Results ................................ ................................ ................................ ............ 103 xiii Figure A.30 Cyclic Operation L ow Wines, Cyclic Distillation (Vapor 4 min, Liquid 3 min) Tray 3 Graphical Results ................................ ................................ ................................ ............ 103 Figure A.31 Normal Operation Brandy Finishing Run w/ reflux - Tray 1 Graphical Results 104 Figure A.32 Normal Operation Brandy Finishing Run w/ reflux - Tray 2 Graphical Results 104 Figure A.33 Normal Operation Brandy Finish ing Run w/ reflu x - Tray 3 Graphical Results 105 Figure A.34 Normal Operation Brandy Finishing Run w/o reflux - Tray 1 Graphical Results ................................ ................................ ................................ ................................ ..................... 105 Figure A.35 Normal Operation Brandy Finishing Run w/o reflux - Tray 2 Graphical Results ................................ ................................ ................................ ................................ ..................... 106 Figure A.36 Normal Operation Brandy Finishing Run w/o reflux - Tray 3 Graphica l Results ................................ ................................ ................................ ................................ ..................... 106 Fig ure A.37 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 1 Graphical Results ................................ ................................ ................................ ................................ ......... 107 Figure A.3 8 Cyclic Operati on Brandy Finishing Tray by Tray w/ Reflux - Tray 2 Graphical Results ................................ ................................ ................................ ................................ ......... 107 Figure A.39 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 3 Graphical Results ................................ ................................ ................................ ................................ ......... 108 Figure A.40 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 1 Graphical Results ................................ ................................ ................................ ................................ ......... 108 Figure A.41 Cyclic Op eration Brandy Finishing Tray by Tray w/ Reflux - Tray 2 Graphical Results ................................ ................................ ................................ ................................ ......... 109 Figure A.42 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 3 Graphical Results ................................ ................................ ................................ ................................ ......... 109 Figur e A.43 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 1 Graphical Results ................................ ................................ ................................ ................................ ......... 110 Figure A.44 Cyclic Operation B randy Finishing Tray by Tray w/ Reflux - Tray 2 Graphical Results ................................ ................................ ................................ ................................ ......... 110 Figure A.45 Cyclic Operation Brandy Finishi ng Tray by Tray w/ Reflux - Tray 3 Graphical Results ................................ ................................ ................................ ................................ ......... 111 xiv Figure A.46 Normal Operation 1 Distillate Graphical Results ................................ ................ 112 Figure A.47 Normal Operation 2 Distillate Graphical Results ................................ ................ 112 Figure A.48 Normal Operation 3 Distillate Graphical Results ................................ ................ 113 Figure A.49 Cyclic Distillation 1 (Vapor 9 min, Liquid 3 min) Distilla te Graphical Res ults 113 Figure A.50 Cyclic Distillation 1 (Vapor 6 min, Liquid 3 min) Distillate Graphical Results 114 Fig ure A.51 Cyclic Distillati on 1 (Vapor 4 min, Liquid 3 min) Distillate Graphical Results 114 Figure A.52 Normal Operation Low Wines Distillate Graph ical Results ............................ 115 Figure A.53 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 9 min, Liquid 3 min) Distillate Graphical Results ................................ ................................ ................................ ........ 115 Figure A.54 Cyclic Oper ation Low Wine s, Cyclic Distillation (Vapor 6 min, Liquid 3 min) Distillate Graphical Result s ................................ ................................ ................................ ........ 116 Figure A.55 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 4 min, Liquid 3 min) D istillate Graphi cal Results ................................ ................................ ................................ ........ 116 Figure A.56 Normal Operation Brandy Finishing Run w/o reflux - Distillate Graphical Results ................................ ................................ ................................ ................................ ..................... 117 Figure A.57 Normal Operation Brandy Finishing Run w/ reflux - Distillate Graphical Results ................................ ................................ ................................ ................................ ..................... 117 Figure A.58 Cyclic Operation Tray by Tray Finishing Run w/ reflux Distillate Graphical Results ................................ ................................ ................................ ................................ ......... 118 Figure A.59 Cyclic Operation Tray by Tray Finishing Run w/ reflux Distillate Graphical Results ................................ ................................ ................................ ................................ ......... 118 Figure A.6 0 Cyclic Operati on Tray by Tray Finishing Run w/ reflux Distillate Graphical Results ................................ ................................ ................................ ................................ ......... 119 Figure A.61 Normal Operation 1 Bottom Column Graphical Results ................................ ..... 12 0 Figure A.62 Normal Operation 2 Bottom Column Graphical Results ................................ ..... 120 Figure A.63 Normal Op eration 3 Bottom Column Graphical Results ................................ ..... 121 xv Figure A.64 Cyclic Distillation 1 (Vapor 9 min, Liquid 3 min) Cycled Volume Gr aphical Results ................................ ................................ ................................ ................................ ......... 121 Figure A.65 Cyclic Distillation 1 (Vapor 6 min, Liquid 3 min) Cycled Volume Graphical Results ................................ ................................ ................................ ................................ ......... 122 Figure A.66 Cyclic Distillation 1 (Vapor 4 min, Liquid 3 min) Cycled Volume Graphical Results ................................ ................................ ................................ ................................ ......... 122 Figure A.67 Normal Operation Low Wines Bottom Column Graphical Results ................. 123 Figure A.68 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 9 min, Li quid 3 min) Cycl ed Volume Graphical Results ................................ ................................ .............................. 123 Figure A.69 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 6 min, Liquid 3 min) Cycled Volume Graphical Results ................................ ................................ .............................. 124 Figure A.70 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 4 min, Liquid 3 min) Cycled Volume Graphical Results ................................ ................................ .............................. 124 Figure A.71 Norma l Operation Br andy Finishing Run w/o reflux - Bottom Column Graphical Results ................................ ................................ ................................ ................................ ......... 125 Figure A.72 Normal Operation Brandy Finishing Run w/ reflux - Bottom Column Graphical Results ................................ ................................ ................................ ................................ ......... 125 Figure A.73 Cyclic Operation Tray by Tray Finishing Run w/ reflux Cycled Volume Graphical Results ................................ ................................ ................................ ........................ 126 Figure A.74 Cyclic Operatio n Tray by Tray Finishing Run w/ reflux Cycled Volume Graphical Results ................................ ................................ ................................ ........................ 126 Figure A.75 Cyclic Operation Tray by Tray Finishing Run w/ reflux Cycled Volume Graphical Results ................................ ................................ ................................ ........................ 127 Figure A.76 Acetaldehyde Concentration vs Time for All Distillations ................................ .... 128 Figure A.77 Acetone Concentration vs Time for All Distillat ions ................................ ............. 128 Figure A.78 Ethyl Acetate Concentration vs Time for All Distillations ................................ .... 129 Figure A.79 Methanol Concentration vs Time for All Distilla tions ................................ ........... 129 Figure A.80 Ethanol Concentration vs Time for All Distillations ................................ .............. 130 xvi Figure A.81 Propanol Concentrati on vs Time for A ll Distillations ................................ ............ 130 Figure A.82 Isobutanol Concentration vs Time for All Distillations ................................ ......... 131 Figure A.83 Butano l Concentration vs Time for All Distillations ................................ .............. 131 Figure A.84 Isoamyl Alcohol Concentration vs Time for All Distillations ............................... 132 Figure C1. Examp le Chromatograph Results ................................ ................................ ............. 142 xvii KEY TO SYMBOLS Nomenclature xviii Super Scripts Subscripts 1 C HAPTER 1 : Introduction 1.1 Distillation Theory and Background Distillation technology h as been used wor ldwide throughout the chemical industry for many years, originally being invented in China around 800 BC. There are currently over 40,000 columns in operation worldwide [29]. Although there are many variations in column types, sizes, tray configurations, etc., the main princip le is the same. Distil lation is the physical process of heating a liquid mixture to separate components based on the relative volatil iti es of each component. There are primarily two main type s of distillation, contin uous distillatio n and batch distillation. Continuous distilla tion requires a continuous feed into the column at an optimal tray , bottoms product and distillate product are continuously collected. This process can theoretically continue forever at steady st ate if there is feed to the column. Multiple feeds can be fed at different locations as well as multiple product distillates may be taken off at different stages in the column. In batch distillation a still pot is used to hold a certain initial volume of l iquid to be sepa rated , known as the feed . The mixture is heated, and distillate is collected until processing of the feed is completed. in the distillate change over ti me. In both type s of distillation, heat is supplied by di rect fire, steam, or electric heating. As the mixture is heated, the most volatile component, the one with the lowest boiling point, will vaporize first and travel upward through the column. The colu mn can have any number of trays ranging from 3 up to 40 o r more depending on the size of production and products being made. Trays allow for rectification of the feed throughout the column. There are many different tray types, but each serve the purpose of allowing some o f the vapor to condense to liquid 2 on the tray and some of the vapor to continue up the column. Each additional tray increases rectification and causes the vapor at the top of the column to have the highest concentration of the most volatile component at an y time during the distillation. Finally, the vapor may travel through either a partial or total condenser. Partial condensers partially condense the vapor sending some liquid back into the column, known as reflux. Total condensers totally c ondense the vapo r into liquid and the product exits the s ystem, known as the distillate. In batch distillation there are two operation modes that are commonly discussed in academia, constant reflux or constant distillate composition. When the reflux is he ld constant the distillate will initially start with a hi gh concentration of the most volatile component in the system and gradually decrease over time. If a constant distillate composition is desired, then the reflux ratio can be adjusted throughout the d istillation. In general, increasing the reflux increases rectification, so as the distillate composition begins to decrease, the reflux will be increased to keep the distillate at a constant composition. 1.2 Distillation Improvement and Cyclic Distillatio n Distillation i s the most widely used separation method. Consequently, distilleries consume about 40% of the total energy used to operate plants in the chemical industry, and over 95% of the energy used in separation processes [31]. Therefore, improving distillation te chnology is one area that chemical engine ers are working to transform into a more efficient process by use of process intensification. Process intensification (PI) is a design philosophy that aims to improve the efficiency of an already e xisting machine or process. In general, it can be a chan ge in equipment or a change in a method to increase efficiency. In distillation, PI aims to improve product quality, lower energy requirements, increase capacity, reduce time from raw 3 material to m arket and increa se safety by design. Overall, PI aims to safely increase the speed and quality of production while decreasing capital and operating costs. The four domains that PI should make use of include: spatial (structure), thermodynamic (energy), f unctional (syner gy), and temporal (time) [30]. There are many PI distillation technologies that have been suggested including heat pump assisted distillation, membrane distillation with its various designs, HiGee technology, cyclic distillation, dividing w all column, and reactive distillation [29]. The purpose of this research was to investigate the possible economic and energy impacts of cyclic distillation as well as its application to the distilled spirits industry. Cyclic distillation is an alternate m ode of operation first proposed and studied by Cannon and vapor flow period heat is supplied to the column, vapor flow upwards and dist illate is collec ted as in normal operation. During the li quid flow period, the heat source is ceased and the liquid on each tray is transferred to the tray below. The cycle of alternating between the two periods is continued for the entire distillation. Cy clic distillatio n has been applied to many different syst ems and configurations and has been demonstrated experimentally and theoretically to increase column throughput, lower energy requirements and achieve higher separation performance. Previous studies on cyclic distil lation and current usage will be discusse d in the proceeding chapter. 1.3 Cyclic Distillation Application to Spirits Industry The distilled spirits industry is a rapidly growing industry . The US Distilled Spirits Council reported an 8 th co nsecutive year o f market share gains in 2017 , s upplied s ales were up 4 percent, rising $1 billion to a total of $26.2 billion, while volumes rose 2.6 percent to 226 million cases, up 5.8 million cases from the prior year. These results reflect adult consu ste for 4 higher - end distilled products acr oss most categories [ 40 - end products causes manufactures to look for technologies that can increase efficiency of their process and profit for their company . Cyclic di stillation can h elp meet and exceed these higher demands. Distilled spirits are different from other distilled products in that they are eventually going to be consumed as a beverage by the consumer . In many processes the end goal is to concentrate one or more products w ith as high purity as possible, in distil led spirits this is not necessarily the case. To make a distill ed spirit product , first the grain is milled into a flour like constituency. The grain is then mixed with water to form a slurry , known as the mash, in a large vat called a mashtun. During the mashing process , the mash is heated, and alpha and beta amylase enzymes are added at the appropriate time, temperature and pH to enable the breakdown of starch into simple sugars, like glucose, that yeast can consum e. The mash is then cooled and transferred to a fermenter to ferment for 1 - 2 weeks. During fermentation of the mash, yeast produce mainly ethanol and carbon dioxide but also by products including other alcohols as well as esters, aldehydes, and ketones . Th e distilling industry refer s to the other compounds as congeners. The concentration of the fermentation at the end is roug hly 8 - 12% alcohol by volume. The next step in the process is stripping the mash into low wines. The fermented mash is pumped into a de signated stripping still which may or may not have trays. The purpose of stripping is to separate and concentrate the vola tile components formed during fermentation from the solid and non - volatile components. The distillate called low wines is collected en tirely and is usually 25 - 40% alcohol by volume . The next step is the finishing run in another distillation column. In larg e operations, th is is done on a continuous column and the product is pulled off the column at the appropriate ethanol concentration fo r the product being made. 5 Batch distillation for both the stripping and finishing run is used widely throughout the craft spirits industry. Many craft distilleries pride themselves on their beautiful copper stills that look like pieces of art in their dis tilleries . In batch finishing , a certain volume of low wines is added to the still and distillate is collected. During bat ch distillation of spirits, fractions of the distillate are separated into different volumes known as cuts. The first distillate produ ct is known as the heads cut, which is heavily concentrated with ethanol as well as the most volatile components mainly ac etone, acetaldehyde, methanol and ethyl acetate. Once those compounds are out of the system, detected by the distiller by the flavor a nd aroma of the distillate, the actual product is collected and known as hearts. The hearts cut has primarily ethanol, but other compounds also come through in low concentrations such as methanol , isoamyl alcohol and other higher alcohols. Towards the end of the distillation the higher alcohols start to increase in concentration in the distillate and the flavor and aroma of t he distillate become undesirable ; this is when the tails cut is made. The tails cut still contains a high amount of et hanol, so it is oftentimes re - distilled to collect as much hearts as possible. Since the ethanol concentration is high and has other hazar dous compounds , it cannot go down the drain and must be disposed of properly . Tails storage and disposal is a big prob lem from smaller distilleries with small facilities while trying to balance the energy requirement of re - distillation with the actual prof it it lends. While ethanol is the primary contributor to the average spirit flavor and aroma profile , concentrations near the ppm lev el of the other compounds mentioned are found and desired to make a balanced and flavorful product. Therefore, it is important to investigate the trends of the congeners during cyclic distillation. Thus far cyclic distillation has not been applied to disti lling spirits. The next few chapters intend to investigate a thorough literature review of cyclic distillation , to model the proposed batch cycli c 6 distillation method , as well as to present the experimental methods and data collected while cyclically disti lling apple brandy on an industry sized batch column still. 7 CHAPTER 2 : Literature Review consisted of two periods, a vapor flow period, and a liquid flow period. Cannon u sed a cycle timer with an automated valve to allow the periods to be specifically timed. Wh ile the valve was open, the vapor flowed through the column, and did not allow liquid to travel down through the trays. When the valve was closed, t he lack of vapor thrust allowed the liquid to fall to the preceding tray. This removed the need for downcome rs on plates because the liquid and vapor flowed through the same area at different times. Cannon concluded the advantages of controlled cycling in distillation we re higher capacity, simpler and cheaper plate design, and high flexibility due to a choice o f operating conditions dependent on cycle times [1]. Cannon and others applied controlled cycling to sieve and screen plate towers as well as pac ked - plate column s [2,3]. Gaska and Cannon reported experimental data on a distillation of a mixture of benz ene and toluene using two test towers, one with 9 plates spaced 18 ½ inches apart, and the other with 17 plates spaced 9 ½ inches apart. The data s howed a 48% incr ease in total vapor load at a fixed column pressure drop. They reported a significant incre ase in column capacity based on the average vapor velocity achieved which is shown graphically below. Figure 2 .1 Cycling I ncreases C olumn C apa city for T ype - T w o P late s [2] 8 Gaska and Cannon concluded that the maximum rate of phase flow was not dictat ed by solely the equipment size and properties of the system, but was also dependent on the mode of operation of the column [2]. McWhirter and Cann on also found th at it was possible to produce maximum tower efficiency, maximum tower capacity or anywhere in between these via controlled cycling of a packed plate tower [3]. McWhirter later demonstrated in his Ph.D. thesis that it was possible to double the stage effici ency a nd triple the throughput of a column simultaneously through theoretical and experimental investigations. McWhirter also demonstrated that while over multiple cycles the column appeared to be in steady - state operation, during a single cycle it was in unstea dy - state. At any instance during the vapor flow period, the composition of the liquid on a tray varied with time, and thus the composition of the vapor leaving the stage also varied with time. He also showed that the average driving f orces for mass t ransfe r in a cyclic column were considerably greater than those in a conventional column. McWhirter believed that the high stage efficiencies of a cyclic column were difficult, if not impossible, to rationalize in terms of conventional oper ation. Lastly, M cWhirt er used these ideas to develop computer simulations of cycled columns by solving the unsteady - state material balance equations describing compositions as a function of time by using the finite - difference method [3]. Schrodt was the ne xt to study cont rolled cyclic distillation and implement this operation mode on a plant - scale [4,5,6,7]. First Schrodt and others re - introduced a simple model to describe the controlled cycle column [6]. They constructed figures relating vapor compositions as a fun ction of the fraction of plate holdup dropped ( ). 9 Figure 2.2 Compositions as a F unction of F raction of P late H oldup D ropped ( ) [6] Figure 2 .2 illustrates the concept that with the closer to true plug flow and more cycles, t he purer the re sultin g vapor composition, which is not dependent on the number of plates. Schrodt defined the effective plate efficiency as: Where, , is a term used to calculate the equilibrium relationship. In addition, Figure 2 .3 was generated which illustrated the effective plate efficiency as a function of plate number. 10 Figure 2 .3 Effective P late E fficiency as a F unction of P late N umber [6 ] Next, Schrodt and others developed a co mputational method for obtaining the theoretical number of stages similar to the McCabe - Thiele. The current method will be presented later in this paper. The next paper published by Schrodt and others reported the results of computer simul ations of contr olled cyclic distillation to investigate the theoretical effects of various p arameters on the separating ability of controlled cycling [5]. At this point is important to remind the reader of the definition of va rious efficiencies below in Ta ble 2 .1. Tabl e 2 .1 Definition of Various Efficiencies [5] Efficiency Symbol Definition Murphree (vapor) point efficiency Local vapor efficiency at a point in time and space Instantaneous plate efficiency - Over - all vapor efficiency of a plate at a point in time Effective plate efficiency Over - all efficiency of a plate computed on the basis of liquid - phase plate compositions Over - all column efficiency Number of theoretical stages corresp onding to a given separation di vided by the nu mber of actual plates in the column 11 The Murphree point efficiency is assumed to be the same and equal for all plates and is defined as: Where is the composition of the vapor tha t would be in e quilibrium with the liquid on the nth stage and is given by the vapor - liquid equilibrium relationship assumed. For computer simulations, Schrodt used the previously developed equations to see how the plate compositions varied with time as well as how the column approached its pseudo - steady - state conditions [5]. T he pseudo steady state was defined as the composition profile remaining constant at the end of successive cycles. For the simulation, a binary system with a constant volatility of 1 .2 was assumed. It ran in total reflux and assumed a boil - up rate and liqui d hold up of 0.1 g mole/sec and 0.2 g mole, respectively. In addition, the initial pot was charged with 10 g moles, the Murphree efficiency was assumed to be 1, the fraction of plat e holdup dumped to the plate below was assumed to be 1, and N was assumed t o be 5. Lastly, the initial composition of the still pot was 0.5. The results from the computer simulation gave graphical insight to how a controlled cycled column operates. It w as found that t he pseudo steady state condition was achieved after roughly 150 cycles. The effect of fraction of plate holdup dropped was stud ied for values of ranging from 0 to 3.0. The effective numbers of theoretical plates in the column ( ) was calculated from the values of the still - pot and condenser compositions at the end of the vapor flow period once the pseudo - steady - state con dition was achieved, using the Fenske equation below: 12 The overall column efficiency ( 0) was then determined by dividing by the number of actual plates. The resulting data can be seen below in Figure 2 .4. Figure 2 .4 Effect of F raction of P late H oldup D ropped D uring the L iquid F low P er iod on the O ver - a ll E fficiency of a C ontrolled C ycling R ectification S till at T otal R eflux [5] studies that maxima occur at integral values of , with t he maximum of t hese occurring at = 1. It is important to note that = 0 corresponds to conventional column o peration. The next parameter studied was the effect of mixing during the liquid flow period. Values of equal to 0.5, 1.0 and 2.0 seconds we re investigated which correspond to = 0.25, 0.5 and 1.0, respectively. Results of these simulations can be see n in Table 1.2 below. 13 Table 2 .2 Effects of M ixing D uring the L iquid F low P eriod on the S eparating A bility of a C ontrolled C ycle R ectification St ill at T otal R eflux [5] Duration of the vapor - flow period Effective number of theoretical plates, Over - all column efficiency, 0.5 4.96 99.1 1.0 4.93 98.7 2.0 4.83 96.6 The data presented in Table 2 .2 illust rates that comp lete mixing during the liquid flow period decreased the separation advantages that were gained in cyclic operation. The next effect studied was the effect of relative volatility. The results are shown below in Table 2 .3. Table 2 . 3 Effect of Relative Vol atility on the Overall Column Efficiency of a C ontrolled C ycling R ectification S till at T otal R eflux [5] Relative volatility, Effective number of theoretical plates, Over - all column efficiency, 1.05 9.56 191.1 1.10 9.54 190.8 1.20 9.46 189.2 1.30 9.35 187.0 1.40 9.23 184.6 1.50 9.12 182.4 1.50 a 9.26 185.3 a The results in Table 2. 3 sh ow that the overall column efficiency decreased with an increase in the relative volatility of the distilled mixture. Next, the effect of Murphree plate efficiency was studied and it was concluded that the over - all column efficiency of a controlled cyclic distillation column increased rapidly as the individual plate efficiencies were increased. This conclusion was also in agreement with computer simulations. In this case, a straight - line equilibrium relationship was assu med. From 14 these simulations, it was found that the over - all column efficiency of a controlled cycling column increased to an asymptotic value as the number of actual plates in the still increased. The se findings are illustrated in Figure 2 .5 below. Figur e 2 .5 Effective P late E fficiencies (E 0 ) in a C ontrolled C ycling R ectification C olumn S till at T otal R eflux [5] To study plant - scale operation, a 20 - plate column was designed to separate a mixture of a cetone and wat er [7]. Schrodt and others found that the use of a conventional column in a cyclic operational mode resulted in substantial capacity improvements, if the number of trays was less than 12. This was assumed to be due to pressure drops throughou t the column which ca used trays to mix with each other during the liquid flow period. They suggested that an 100% efficiency increase, and 2 to 3 times capacity increases were still possible if true plug flow was observed. 15 True plug flow would be achieved if the volume from tr ay 4 only dropped to tray 3, and none of the liquid from tray 5 would mix with the liquid of tray 3 [7]. In 1967, Robinson and Engel presented a theoretical analysis that demonstrated the advantages of cycled mass transfer operations [9]. Considering a cl assic paper by Lewis on the effect of liquid phase composition profiles applicable to cyclic operation, Robinson and Engel developed time - axis concentration gradients on each stage, analogous to those developed for conventional columns with lateral concent paper considering three cases. Case 1, vapor flowing to the tray, liquid flows at uniform composition at all points. Case 2, vapor laterally unmixed flowing between stages, liquid flows in same lateral direction on all stages. Case 3, vapor laterally unmixed flowing between stages, liquid flows in opposite directions on alternate stages. Robinson and Engel decided that the plate directly above the reboiler was analogous to a Lewis C ase 1 with constant c omposition vapor flowing to it over short times of one cycle. The rest of the plates corresponded to Lewis Case 2 plates. Robinson and Engel then used the mathematical analyses developed by Lewis to derive a vapor flow period material balance and solutions to case 1 and case 2. The plate efficiency was found to be dependent on the fraction of liquid on the plate that drained to the plate below during cycling. The results of this analysis were like the results of McWhirter and Schrodt, t herefore the equation s and figures have not been presented [9]. In 1967, Horn published a paper on periodic countercurrent processes. The report discussed how the overall stage efficiency of a periodically operated distillation column is complicated and de pends on the number o f stages as well as the equilibrium and transport parameters. Horn developed accurate asymptotic formulas for the stage efficiency that proved the idea that the performance of a distillation column can be drastically improved by period ic operation [8]. 16 In 1968, Horn and May developed a simple asymptotic relation which described the stage efficiency of a periodically operated distillation column. The relation developed used parameters which are a function of Murphree efficiency, transp ort number, and separ ation factor. Horn and May defined an important number, the transport number as the ratio of the quantity of liquid transported per cycle to the liquid hold up of a stage. They also investigated the effect of mixing on stage efficiency of a periodically op erated countercurrent process for the case of difficult separations. They did so by modeling each stage, during the time of liquid transport, by a series of well stirred tank s [10,11]. h a bubble - cap fracti onating column in a cyclic regime [13]. The column contained six sections having an internal diameter of 80 mm and a height of 250 mm, each with five single cap plates in between the sections. The experiment allowed for varying total c ycle time from 5 to 6 0 seconds, operated at total reflux and atmospheric pressure. To incorporate the cyclic regime, electromagnetic solenoid valves were used in the vapor feed and reflux lines and controlled by two - time relays through the rectifier. From the resulting data, t hey determined an optimum total cycle time and ratio of vapor to liquid feed periods. The experiments confirmed the presence of a beneficial effect from the use of the cycle regime with bubble - cap plates as seen in Figure 1.5 below. 17 (a) ( b) (c) Figure 2 .6 (a) Dependence of C olumn S eparation E fficiency E on M ean V apor V elocity (b) Dependence of C olumn S eparation E fficiency E on V apor F eed T ime and on L iquid - feed T ime (c) Dependence of C olumn S eparation E fficiency E on the R atio of F eed P eriods of P hases for V arious V elocities [13] In the Figure 1.6(a) line 1 is the stationary regime, line 2 is cyclic regi me with and , and line 3 is cyclic regime with and . The cyclic regime was clearly more efficient than the stationary mode of operation. The maxima in the cyclic regime lines indicates a ma ximum vapor flow rate correspondi ng to a maximum efficiency. In Figure 1.6(b) line 1 is for , line 2 is for , line 3 if for and line 4 is for . This figure allowed for the assumption of the optimal vapor feed time of 8 seconds and the optimal liquid feed time to be 5 seconds. It is important to note that the optimal duration of liquid feed to the column was determined by the time to replace the liquid on each plate and was a function of plate construction and properties of the liquid. I n Figure 1.6(c) line 1 corresponds to a mean vapor velocity and line 2 corresponds to . From the data it is evident that for this column, the optimum total cycle time is about 13 sec, and the ratio of the feed periods f or vapor and liquid is in the range from 1 to 3 [13]. 18 U p to this point, most of the research done was comparing cyclic columns to conventional operation. In 1977, Rivas developed a short cut method for preliminary design of ideal cyclic distillation column s. Material balance differential equations on each plat e of a cyclic distillation column were used to develop analytical equations to calculate the ideal number of trays for cyclic countercurrent processes like distillation, absorption, and stripping [14]. One year later, Furzer studied fluid flow in a distill ation column by taking measurements of the discrete residence time distribution [16 ]. The column had five sieve plates spaced 635 mm apart. A microprocessor was used to obtain periodic control. The dat a yielded the parameters in the (2S) model which descri bed the fluid flow. Furzer concluded that in order to achieve maximum separation, modifications to the column internals were required. A few years later in 1980, Furzer and Goss studied fluid mixing an d mass transfer separations of mixtures of methylcycloh exane and n - heptane under periodically cycled conditions [17]. The theoretical improvements of 200% reported by McWhirter and Lloyd could not be reproduced. Goss and Furzer concluded that liquid bypass ing the plates caused the reduction to 107 - 126% improve ment. However, the (2S) model was successful in modeling both the fluid mixing and the mass transfer separations in the periodically cycled column. Baron, Wajc and Lavie developed a theory of stepwise periodic distillation [18,19]. The major difference in stepwise periodic v ersus controlled cycling is in stepwise the liquid flow is controlled directly instead of through pulsations of the vapor flow rate. The researchers developed this based on the idea that preventing axial mixing could improve the separat ion. Their research goal was to compare the two operating models for total reflux and for continuous distillation, as well as construct a lab - scale stepwise periodic column for data collection. The col umn analyzed 19 consisted of a boiler, a total condenser, N trays numbered starting at the bottom and N reservoirs specific to each tray. The schematic can be seen below in Figure 2 .7. Figure 2 .7 Schematic of C olumn U sed for S tepwise P eriod D istill ation [1 8] Each tray is fitted with an inlet and an outlet with on - off valves. The trays had no downcomers but could be completely emptied via the outlet to the reboiler. The reservoirs collected the condensate from each tray via the inlet. One periodic cycle star ted with vapor alone flowing through the column exchang ing mass with the stationary liquid on each tray. The condensate was then collected in the reservoirs corresponding to each tray. When the bottom tray (tray N) was full, all outlet valves were opened t o empty all plates into the boiler. Next, the outlet va lues were closed, and inlet valves were opened to all the reservoirs to empty to their corresponding plates. Lastly, all valves were closed again, and the cycle starts again. Baron and others compared controlled cycling and stepwise periodic distillation. They showed that the two processes have the same asymptotic efficiencies for large values of N while periodic distillation is 20 10 to 30% more efficient for a finite number of trays [18]. In a follow - up study, they described separating a binary mixture using stepwise periodic operation mode. They proposed a model and simulation algorithm for both controlled and periodic cycling. Through an extensive parametric study, it was concluded again that stepwise p eriodic operation was superior to ideal controlled cycl ing [19]. In 1984, Matsubara, Watanbe, and Kurimoto analyzed the work of Cannon and Baron - Wajc in hopes to combine the advantages of both schemes [20]. They successfully developed a composite scheme which could vary from the ideal cannon scheme to the Bar on - Wajc scheme by changing various parameters. In a second paper published in 1985, a laboratory - scale periodic distillation column using the idealized Cannon scheme was constructed and experimented with to view energy conservation performanc e [21]. They u sed a five - stage column to separate water and methanol. The experimental results showed an 84.6% increase in energy efficiency due to an average vapor f low rate roughly 50 - 20% lower than in conventional continuous column operation and achieving the same se paration. Around the same time, Thompson and Furzer developed hydrodynamic modeling for liquid holdup in periodically cycled plate columns [22]. Their objective was to predict the holdup distribution in a periodically - cycled column, for a specific set of design and operating conditions. They confirmed their models were satisfactory by experimenting with a 600 mm diameter Perspex column containing four s ieve plates. The column was fitted with Time Delay Plates to improve plug flow pressure equalization at the start of the liquid flow period [22]. Szonyi and Furzer then developed a new tray design to increase column performance using a periodic cycle opera ting method [23]. They ran computer simulations which predicted 200% greater column performance for sys tems with nonlinear equilibrium curves. They also conducted experiments distilling 21 methanol - water mixtures to verify the computer predictions. Using a s ingle - plate the simulations expectations were confirmed. A new tray design was then implemented in a fiv e - plate insulated mild steel column of 610mm I.D. The new trays consisted of a sieve tray plus inclined surfaces and a vessel in between each sieve tray . The new tray design can be seen in Figure 2 .8 below. Figure 2 .8 New T ray D esign for P eriod C ycling D istillation [ 23] The trays collected liquid from the sieve plate above and directed it via the inclined surface to the vessel. In doing so, the liquid in the vessel experienced a time delay before flowing to the sieve plate below. The trays were experimen tally effective at obtaining true plug flow but were limited by lack of ventilation and consequently were only able to achieve 140% greater column efficiencies compared to conventional distillation. Szonyi and Furzer recommend further improvements be made by hardware modifi cation [23]. 22 In 1989, Toftegard and Jorgensen developed an integration method for the dynamic simulation of cycled processes. Application of the method was demonstrated on a simulation of controlled cycling distillation. The time to comp ute the transient was reduced by 90% compared to using a conventional integration method [37]. Finally, in 1997 Sorensen and Prenzler applied cyclic distillation theory to batch distillation [38]. The cycle applied to a batch column was a repeated filling and dumping of th e reflux drum. In the beginning of the cycle the reflux drum was filled. Then the column ran under total reflux, and the maximum attainable separation in the column was achieved. During total reflux the light component accumulated in the reflux drum until the column reached equilibrium or steady state. At the end of the cycle, the drum was dumped, and the product was withdrawn. An illustration of each period can be seen below in Figure 2 .9. Figure 2 .9 The T hree C haracteristic P eriods in the C yclic O perat ion of a R egular B atch D istillation C olumn [39] Sorensen concluded that the best operating procedure was to adjust the reflux drum holdup online based on measurements of distillate composition as a function of time. If the purity is too l ow the 23 drum holdup could be reduced and vice versa. A laboratory batch column with 8 sieve trays was used to implement this method [39]. A schematic of the column can be seen below in Figure 2 .10. Figure 2 .10 Lab B atch C olumn U sed to A pply a B atch C ycli c O perating T heory [39] The still consisted of a 12 L reboiler, a column with 8 sieve trays, inner diameter of 50 mm, a condenser, a reflux drum with a maximum holdup of 4 L, a reflux valve, and a top product receiver. The reflux valve was a three - way sole noi d valve. Samples could be taken through a manual valve as the condensate was returned through the column. In addition, samples of the reboiler and top product were taken. The feed consisted of 10 L of a 25 mol% methanol and 75 mol% water. The 24 required a mou nt of product was 4 L with recovery of methanol at least 75 vol%, which corresponded to 65 mol%. The column started with a cold feed which was heated to the boiling point at roughly 78 C. The first cycle started when the mixture began to boil, the refl ux drum began to fill, and temperatures in the column increased. Once the drum was filled to a desired level, reflux back to the column started, and the temperatures in the column started to decrease. This was the beginning of the second total reflux perio d. Once the temperatures in the column stabilized, steady state was achieved which happened after about 95 minutes of operation. The reflux drum was then dumped to the product receiver for a duration of 1 min. A new cycle was then started, and the previous pr ocedure was repeated for another two cycles. At the end of 305 minutes of operation, 68mol% methanol was achieved in the final product which was within the specifications. The policy was found to be easy to implement with the only interaction being the dum ping of the reflux drum. The main difficulty was determining the end of the total reflux period which was chosen to be when the temperatures in the column did not change more than 0.1 C within in a 5 min period [39]. In 2000, Bausa and Tsatsaronis studie d the optimal profiles for all important control variables in continuous cyclic distillation systems [24]. These variables included the flow rates of feed, products, reflux and vapor. They studied two examples. In the first, an ideal ternary mixture was se parated into two fractions. In the second, the same mixture was separated into three fractions using a column with a side stream. The optimal control problem was formulated and solved using the software Optimal Control Code Generator for Maple (OCOMA). In the first example, the energy demand of a column with 16 trays for separating a three - component mixture operating at 22.7% higher than minimum energy demand, was reduced by 4.2%. The s tudy showed that the possible energy savings were higher for columns ope rating at considerable higher energy demands. 25 The second example was the first study on a column with a side stream. The study showed that the potential energy - savings can be reduced b y approximately 48%. Lastly, the authors concluded that the inclusion of oscillations in the optimal control profiles can be extended for the batch distillation and could lead to large energy saving s [24]. In 2012, Flodman and Timm studied batch distillat ion employing cyclic rectification and stripping operations. The authors experimented with three modes of operation: (1) a conventional batch still with a fractionation column configured for rectification, (2) an inverted batch still with the column config ured for stripping and (3) an operating policy where the column is seque ntially used for batch rectification and stripping. The aim for the third mode was to demonstrate a mode that potentially could reduce operating costs by maintaining high product rates with fluids having modest differences in volatilities. The policy was i mplemented on an educational batch rectifying/stripping still equipped with a total condenser, partial reboiler, and six, 3 - inch diameter sieve trays. The control diagram for batch rec tification is illustrated below in Figure 2 .11 . 26 Figure 2.11 Batch R ec tification C ontrol D iagram [25] The control diagram for batch stripping is illustrated below in Figure 1.12. Figure 2 .12 Batch S tripping C ontrol D iagram [25] 27 Conventional batch recti fication and inverted batch stripping was used to promote high product flow rates for a binary fractionation. While rectifying, the light component was removed as distillate, concentrating the heavy component in the reboiler. As the distillate decreased wi th time, the still was then switch ed to stripping mode. The hea vy component was removed as bottoms product, concentrating the light component in the distillate drum. For startup and for liquid transfer, tubing allowed the bottoms to be pumped to the distil late drum, or the distillate to be transferred to the reboiler. The mode was not optimized; however, they demonstrated energy and time savings compared to conventional batch or inverted batch distillation alone. The advantages of this cyclic batch operatio n were the simultaneous concentration for both the light and he avy components, as well as the capability of fractionating nearly the entire contents of the initial charge. Lastly, the authors suggested that a minimal capital investment would be required to convert a conventional batch still to be capable of this type of cyclic operation [25]. Up to this point, most of the modeling done for cyclic distillation involved an assumption of a linear equilibrium relationship. Lita, Bilde, and Kiss filled this gap in 2012 by modeling the design and control of cyclic distillat ion systems for the general case of nonlinear equilibrium [26]. The authors developed a complete model and an innovative graphical method similar to the McCabe - Thiele diagram, as well as demons trated the controllability of cyclic distillation columns. Thro ugh their study, they concluded that the energy requirements for cyclic distillation are greatly reduced especially for high purity products. Moreover, the number of trays required is reduced b y nearly 50% when the same purity is obtained with the same vap or flow rate. The authors also discovered that a minimum vapor - flow rate exists corresponding to an infinite number of trays, as well as the opposite. A minimum number of trays exists with a co rresponding infinite vapor flow 28 rate. Lastly, the authors sugge sted that columns could be easily controlled by adjusting the reflux and the vapor flow rate to sustain required product puritie s [26]. Finally, in 2012 Maleta and Maleta patented the first of ficial tray to be used for cyclic distillation [27]. They calle device works as follows. During the vapor period, the gas phase lifts a movable valve in such a way that the upper plate clos es the bottom, opening of the above tray. The vapor getting und er the upper plate is allowed through small orifices and enters a barbotage unit while passing through the liquid layer therein. At the end of the vapor period, the valve moves down due to the weight of the liquid on the above tray. The liquid from the bar botage zone on tray 1 passes through the windows and into the liquid (which is limited by a casing) on the lower tray. The overall design of the trays is a combination of bubble trays and sluic e chambers under each tray. The liquid can move from one tray t o another without mixing of liquids from adjacent trays. A figure of the trays and a complete column set up can be viewed below in Figure 2 .13. 29 Figure 2 .13 Mass Exchange Contact Device and Column [28] The most recent study on cyclic distillation was published in 2015 [28]. Maleta, Shevchenko, Bedryk, and Kiss reported on the performance of a pilot - scale distillation column for ethanol - water separation operated in cyclic mode. The column had a diameter o f 310 mm, height 5500 mm, and had 10 Maleta trays spaced 500 mm apart. The study was set up to feed two separate columns at the same time using a splitter to allow for optimal performance comparison of both operation modes. The study resulted i n higher thr oughput and equipment productivity in the cyclic operated column. With a beer mash feed of 4 - 12% ethanol, cyclic distillation had a steam usage of 1.4 times less while using 2.6 times fewer trays as compared with classical distillation. The max imum efficie ncy of the process corresponded to the minimum steam consumption due to a finite number of trays within a given column. The pilot scale distillation column was 100 - 160% per the perfect mixing theoretical stage model (classis distillation), or 4 0 - 90% per th e perfect displacement model. The authors concluded that potential applications include biofuel production, organic synthesis, specialty chemicals, gas processing, petrochemicals, and pharmaceuticals. 30 CHAPTER 3 : Modeling Overview 3.1 Batch Di stillation 3.1.1 Alembic Distillation The Rayleigh Equation The simplest type of distillation is binary distillation using no trays and consequently no rectification , also known as Rayleigh distillation. In the spirits industry, Rayleigh distillation is also known as alembic distillation. The pot is charged with an initial feed which is connected directly to a condenser. As the feed is boiled, the resulting initial vapor will be rich in the most volatile component. As the distillation continues, the conc entration of the most volatile component will decrease in the pot. Figure 3.1 is a schematic of simple batch distillation or Rayleigh distillation. Figure 3.1 Simple Batch Distillation S chematic M ass balances can be written around the entire system: 31 The feed variables , are known, as well as the desired final distillate or pot concentration either or . Given there are 3 unkno wns and 2 equations, a third equation was derived known as the Rayleigh equation. An assumption is applied that the holdup in the accumulator and column are negligible. Using this assumption, a differential mass ba lance can be applied to the system. The di fferential amount of a component - dW with concentration xD is removed from the system, resulting in the following differential mass balance: Rearrangi ng and integration gives: I n simple batch distillations, the vapor and liquid are in equilibrium. When a total condenser is used the substitution of is used, resulting in: He re x and y are in equilibrium which can be expressed as This is the common form of the Rayleigh equ ation which shows the relationship between total moles remaining in the still and the mole fraction of the mor e volatile component in the still at any given time. 3.1.2 Multistage Batch Distillation Multistage distillation refers to distillation involving one or more trays. Within these systems the mole fraction in the distillate, and the mole fraction in th e pot, are no longer in equilibrium. Therefore, a relationship between and must be found using stage by stage calculations. Below is a schematic of multistage distillation. 32 Figure 3.2 Multistage Ba tch Distillation An assumption is made that the liquid hold up on each tray, condenser and pot is negligible. Therefore, mass , material an d energy balances can be written around stage j and at the top of the column at any time t as follows: 33 When constant molal overflow for vapor, liq uid and distillate is assumed, the energy balance is no longer needed, and the compone nt balance also known as the operating line becomes: The previous equation is a straight line on an equilibrium x - y diagram, with the slope L/V and the intercept with the y=x line at . During batc h distillation, either or L/V will need to vary to satisfy the equation, and therefore the operating lin e. The McCabe - Thiele method for multistage distillation with variable reflux can be seen in Figure 3.3 below. Figure 3.3 McCabe - Thiele Diagram f or Multistage Batch Distillation of Ethanol/Water with Varibale Reflux 34 3.1.3 Stage - B y - S tage M ethods for B atch R ectification To design or simulate a multicomponent batch distillation system, stage by stage temperature, flow rates and composition profiles as a function of time are required. The calculations are in depth but can be solved with either of two types of computer - based methods using differential - algebraic equations which can be solved in two ways. The model presented by Boston, is based on the mu lticomponent batch - rectification operation [ 42 ]. The system consists of a partial reboiler (still - pot), a column with N equilibrium stages and a total condenser with a reflux drum. To start the distillation, the feed is charged to the pot and heat is suppl ied. As the mixture in the pot beings to vaporize it travels upwards through the column. When the vapor leaves s tage 1 at the top of the column it is condensed and passes to the reflux drum. At first, a total - reflux condition is established for a steady - st ate, fixed - overhead vapor flow rate. Starting at t=0, distillate is removed from the reflux drum and accumulated in the receiving tank at a constant molar flow rate, and a reflux ratio is established. The heat - transfer rate to the reboiler is adjusted to m aintain the overhead - vapor molar flow rate. The model is broken down into three separate sections and model equa tions are derived for component material balances, a total material balance and an energy balance. These can be seen respectively below. For se ction 1: 35 The derivative terms are accumulations du e to holdup, which is assumed to be perfectly mixed. To relate the vapor and liquid mole fractions at Stage 1, we have the phase equilibrium equation: By combining the co mponent material balance with the previous equat ion, a revised component material balance is derived in terms of liquid - phase compositions. By combining the total material and energy balances, a revised energy balance equation is derived that does not incl ude Equations for sections II and III are derived similarly. Below is the resulting model for , where refers to the component, refers to the stage, and is the molar liquid holdup. 1. Component mole balances for the overhead - condensi ng system, column stages, and reboiler, re spectively: Where 2. Total mol e balances for overhead - condensing system and column stages, respectively: 36 3. Enthalpy balances around overhead - condensing system, adiab atic column stages, and reboiler, respectively: 4. Phase equilibrium on the stages and in the reboiler: 5. Mole - fractions sums at column stages and in the reboiler: 6. Molar holdups in the condenser system and on the column stages, based on constant - volume holdups, . 37 Wh ere is liquid molar density. 7. Variation of molar holdup in the reboiler, where is the initial charge to the reboiler. The previous described equations make up an initial - value problem for a system of ordinary differential and algebraic equations. The total number of equations is If variables , and all are specified, and if correlations are available for computing liquid densities, vapor and liquid enthalpies, and K - values, the number of unknown variables, distributed as follows, is equal to the number of equations . Table 3.1 Variables and Total Number of Equations for Multicomponent Batch Distillation Initial values at for al l the above variables are obtained from steady - state, total reflux calculations which depend on the values of and . The previous equations in section 1, 2 and 3 include first derivatives of a nd With the exception of , derivatives of the other two variables can be approximated by incremental changes over the 38 previous time step. If the reflux ratio is high, may also be approximate d. This reduces the equations for the component material balances to be integrated in terms of the dependent variables. 3.1.4 Governing Equations and Thermodynamic Model for System in Study Applying the previous modeling met hod provides differential and alge braic equations which will be used to model the system in study. Below is schematic of the column and flows in study. Figure 3.4 Schematic of Cyclic Distillation on Carl Still The partial condensers are assumed to act li ke stages and will be modeled as s uch. The material balance equations for the stages (N=1 - 5) in the column is: The material balance for the condenser is: 39 The material balance for the reboiler is: A solution of water and ethanol is considered as a non - ideal mixture because it presents an azeotrope at 89% mole fraction of ethanol. Therefore, m odified R det er mine the vapor mole fractio n given a temperature and liquid mole fraction. Where is an activity coefficient that can be calculated by using a specific thermodynamic model. In this study, the NRTL method was the thermodynamic model chos e n based on a study which reported that the NRTL method fit experimental data accurately to an RMS va lue of 0. 403 % [ 41 ]. It was important to pick a model that accurately modeled the azeotrope between water and ethanol. The NRTL activity model is calculated using parameters and : Where The parameters and were found within literature to be - 633 and 5823.1 , respectively [ 41 ]. To determine the vapor pressure the Antoine e quation was us ed: 40 Where pressure is in mmHg and temperature is in is in ° C. Finally, the total pressure was calculated as the sum of the partial pressures using the follow equation: 3.2 Cyclic Continuous Distillation Modeling Approach As previously stated, cyclic column operation consists of two parts, a vapor flow period, and a liquid flow period. Since the operation is two separate parts, the mathematical analy sis must be done in two parts as well. S ections 3.2.1 3.2.5 describe t he method first published by Lita, Bilde and Kiss for the modeling of continuous cyclic distillation systems and comparison against conventional distillation. 3.2.1 Assumptions The c urrent accepted m odeling approach is derived using the following assumptions: Binary (mixture) distillation Ideal stages (vapor - liquid equilibrium is reached) Equal heat of vaporization (constant molar holdup and vapor flow rate) Perfect mixing on each st age Negligible va por holdup Saturated liquid feed 3.2.2 Operational Constraints F rom the condenser and reboiler mass balance, for one operating cycle : 41 t he condition follows that: 3.2.3 Model of Vapor Flow Period The model for the vapor flow period involves using equations to describe the evolution in time of stage holdup and composition. Integration of these equations from to gives t he state of the system at the end of the vapor flow period, [26]. 3.2.4 Model of Liquid Flow Period The model for the liquid flow period uses the following equations: 42 3.2 .5 Solution Method For the solution, the previous equations are written in condensed form, where and are mappings which relate the state at the beginning and at the end of the liquid and vapor flow periods , respectively: Which implies the following condition: A solution to the above equation can be found by considering an initial state and applying relationships for and until the difference betwee n two iterations becomes small. Alternatively, the convergence of the iterations can also be done by applying numerical methods and can be solved in MATLAB © [26]. Below is a graphically representation of the energy requirements in cyclic distillation comp ared to classic distillation. 43 Figure 3. 5 Comparison of E nergy R equirements in C yclic D istillation versus C onventional D istillation [26] The vapor/feed ratio is comparable to the energy requirements of a distillation column. The higher the ratio, the more energy required. represents the product purity. The energy requirements are greatly reduced and is reduced even greater when the p roduct purity is higher. 3.3 Design of Cyclic Distillation 3.3.1 Design Methodology First given the feed and the requ ired performance the mass balance over one operating cycle is used to find the product flow rates Next the vapor flow rate and the duration of the vapor - flow period, , are specified . The liquid transferred from the condenser to the top tray can be calculated. 44 For tray holdups in rectifying section and for the stripping section we have It is important t o check hydrodynamics of the column including, column diameter, vapor velocity and pressure drop. To determine the state of the reboiler at the end of the vapor - flow period the following needs to be specified: - Holdup, . Al though the results of this design procedure are independent of this value it is still necessary to specify. - Composition, . At the end of the vapor - flow period, the reboiler is richer in the heavy component and therefore is the time to r emove the bottom product if possible. To find the holdup and reboiler composition at the beginning of the vapor - flow period, equations (3.32) are integrated from to . Next determine the state of the last tray (stage NT - 1) at the end of the vapor flow period, using the mass balance for the liquid flow period: Next find the state of the reboiler and last tray at the begi nning of the vapor flow period by integrating equations 3.30 and 3.31 from . Next add one more tray, whose state at the end of the vapor - flow period is 45 And integrate the resu lting set of equations from . Repeat until the feed composition is reached for the tray NF+1. Then find the state of the feed tray a t the end of the vapor flow period. And integrate from . Similarly, repeat addition of one tray, finding its state at the end of the vapor - flow period and integration of the resulting equations until the distillate composition is reached. The previously discu ssed process is illustrated in Figure 3.5 below. Figure 3. 6 Cyclic D istillation D esign A pproach 3.4 Batch Cy clic Distillation Dynamic Analysis dissertation from the equations developed were written for a continuously fed batch system. Here the equations are re - writt en to apply to batch distillation where necessary. B elow is a schematic of the distillation column used for cyclic operation in the lab . As see n in Figure 3.6 below, in terms of dynamic modeling partial condensers act similarly to stages therefore the part ial 46 condensers are modeled as such. The following sections will provide the equations written to model the system in study. 3.4.2 Vapor Flow P eriod During the vapor flow period, only vapor flows up through the partial condensers and the column and ideally no liquid flows down the column. The column liquid hold up is trapped and held on the stages in the column during this vapor flow period. A de tailed analysis will be made for stage N. The material balance for the most volatile component on stage N is writ ten as: Where, The energy balance relationsh ip is then: Where, 47 The average enthalpies are a function of the liquid mole fraction and temperature on stage N. For simplicity, the difference between enthalpies is assumed independent of composition. This allows for the assumption of c onstant value of on each stage. Also, the vapor hold up is assumed negligible in comparison to liquid hold up. With these assumptions the energy balance becomes: The liquid holdup per stage during the vapor flow period is constant and equal for all stages. Using this and the previous assumptions, the material balance becomes: The previous equation completely describe s the materia l balance and must be written for each st age in the column of study and all must be satisfied. The reflux to the column is assumed to be cycled in phase with the liquid flow period. Also, the condenser is assumed to have a holdup equal to the amount of v apor introduced into it during the vapor flow period. Consequently, the entire contents of the condenser will be returned as reflux or withdrawn as distillate product during each cycle. The overall material balance expression for the reboiler is: 48 Where, The contents of the reboiler is assumed to be perfectly mixed. The material balance on the most volatile component is: Si milarly, the vapor hold up is co nsidered negligible in comparison to the liquid hold up. Using this assumption and substituting the material balance becomes: Adding equations for the reboiler to the previous stage equation s complet ely models the vapor flow period for the system in study. 3.4.3 Liquid Flow Period During the liquid flow period ideally only liquid flows down the column to the preceding tray and there is no vapor traveling up the column. Since the relative velo city bet ween the phases will be almost zer o and the vapor hold up is essentially negligible, no mass transfer will be assumed to take place during the liquid flow period. Consequently, time is not a factor in any calculations for the liquid flow period. Pl ug flow of the liquid from stage to stage down the column is assumed. 49 The volume of liquid that flows from stage to stage is assumed equal for all stages and constant for all liquid flow periods. The following equation gives a material balance for stage N: Where, The material balance calculates the value of the concentration of the liquid on stage N at the start of a new vapor flow period from the values at the end of the preceding vapor flow period. To simplify Equation 3.60, a new quantity is i ntroduced, , which is the ratio of the quantity of liquid which flows during the liq uid flow period to the liquid holdup per stage. Using this variable, Equation 3.60 becomes: The previous equation is valid for values of from zero to one, but when values are larger than one the following will apply: F or For 50 During the liquid flow period, no vapor leaves the reboiler, but reflux from the partial condenser is re - introduced back into the reboiler. The average composition of the reflux from the partial condenser depends on the value of For Where For For The material balance for the most volatile component for the reboiler during the liquid flow period is given below: Where 51 Equations 3.53 through 3.65 describe the operation of the column throughout one entire cycle. The equations with correct sub scripts can be found Appendix B. 3.4.4 Solution Method Like how the continuous equations are solved , e ach cycle requires the end point of the previous cycle. Therefore, to start the vapor flow period equations are used to solve with initial conditions for the state of the system at the end of vapor flow period 1. Those values are then used to solve for the state of the system at the end of liquid flow period 1. Those numbers are then used to solve for the state of the system at the end of vapor flow period 2. This process continues until the final condenser ABV drops below 10%. The MATLAB script written for the solution and modeling of cyclic batch distillation can be found in Appendix B. The results can be found in Appendix B and Chapter 5, Results. 52 CHA PTER 4: Methods 4.1 Materials and Equipment Raw materials used for this research included: fermented ap ple cider, previously distilled apple brandy, standard solutions, and cleaning agents. Equipment used for this research included: 165 L Carl batch still, leur lock syringes, copper tubing, centrifugal pump, GC - 2010 gas chromatograph, plastic GC vials, stor age tanks, hydrometer, and miscellaneous glassware. 4.1.1 Manufacturing Equipment To monitor the tray composition during distillation, four site glasse s on the distillation column where replaced with a new piece of glass made of borosilicate, which had a precut hole to allow for a sampling apparatus to be installed. Borosilicate was chosen due to its high melting point and non - reactivity with the compoun ds being distilled. The sampling apparatus consisted of a l ue r - lock system with copper tubing going to the column and resting on the tray. Outside of the column a syringe allowed a 1 mL sample to be extracted from the tray. A schematic of the self - manufact ured sampling equipment can be seen below. 53 Figure 4.1 Manufactured Sampling Apparatus In addition to the fabrication of sample ports, an extra valve between the steam condensate outlet to the drain was installed to allow for the collect ion of the conden sate. This allowed for the condensate to be collected into buckets, transported to a larger container and measured. The volume was then converted to pounds of steam and reported as pounds of steam in the data. 4.2 Measurement Methods 4.2.1 GC Method A gas chrom ato graph was used to analyze the samples withdrawn from the still. Below is a list of compounds that were of importance and their respective boiling points. 54 Table 4.1 Detected Compounds [44] Compound Boiling Point ( ° C) Retention TIME (MIN) Acetald ehyde 20.8 1.532 Acetone 56.2 2.194 Ethyl Acetate 77.0 2.913 Methanol 64.7 3.091 Isopropanol 82.6 3.509 Ethanol 78.0 3.615 Propanol 97.0 5.184 Isobutanol 108 6.132 Butanol 117.7 6.944 Isoamyl Alcohol 132.0 7.887 The above compounds were selected based on previous knowledge of the compounds metabolized by common yeast strains used in fermented beverage production and are volatile enough to come through to the final product in distilled alcoholic beverages . Although there are many other compounds t hat yeast produce including aldehydes, ketones, and esters, the compounds chosen to have the most impact o n the flavor and aroma profile of sprits. Sample chromatograph results can be seen in Figure C.1 in Appendix C. Standards with four different levels of concentrations ranging from 0.1 10 g/L were made to form a method for the GC - 2010 gas chromatograph. The column used was a 30 m Stabliwax column with 0.32 mm ID and film thickness of 0.50 um. The oven temperature s tarted at 35 ° C and ramping up to 100 ° C over a period of 8.5 minutes. The carrier gas used was helium with a linear velocity of 45 cm/sec and a split ratio of 50. The auto injector AOC - 20i was paired with the GC to analyze 12 samples in one batch. 55 4.3 Re search Design Multiple distillations were carried out to understand how cyclic distillation compares to conventional distillation in the 165 L batch still. Table 4.2 shows all distillations performed and the operating conditions associated with each disti llation including : the duration of th e liquid and vapor flow periods during cyclic operation , the initial volume and ABV, and whether reflux to the partial condensers was being utilized or not. Table 4.2 Distillation Conditions Distillation # Operation Ty pe (Vapor, Liquid) Initial Volume (ga ls) Initial ABV Reflux to Partial Condensers 1 Conventional 47 35% Off 2 Conventional 48 25% Off 3 Conventional 49 40% Off 4 Cyclic (9 min, 3 min) 47 36% Off 5 Cyclic ( 6 min, 3 min) 45 38% Off 6 Cyclic ( 3 min, 3 min) 39 36% Off 7 Conventional 25 8% Off 8 Cyclic (9 min, 3 min) 25 8% Off 9 Cyclic ( 6 min, 3 min) 25 7.4% Off 10 Cyclic ( 4 min, 3 min) 25 6.3% Off 11 Conventional 36 19% On 12 Conventional 36 27% Off 13 Cyclic ( 6 min, 3 min) Tray by Tray 36 28% On 14 Cyclic (6 min, 3min) Tray by Tray 33 27% On 15 Cyclic (6 min, 3min) Tray by Tray 33 30% On 56 4.4 Procedures To begin a distillation, the pot was filled with the intended initial volume. The door was then shut tightly, and steam val v e s open ed . The condensate valve to the steam jacket was th en opened to release any water remaining in the jacket from the previous d istillation and then closed. In the beginning of every distillation , steam was supplied to the steam jacket until the first volume of distillate was collected. The distillate flowrat e and pressure reading were monitored throughout the distillation throughout each distillation. If the flowrate dropped below average levels, the steam was increased using the gate valve to keep a relatively constant distillate flowr ate. During convention al operation, steam was supplied to the column for the entire duration of the distillation , until the hydrometer read below 10% ABV . During cyclic distillation, the first vapor flow period started when the first distillate was collec ted. After the approp riate vapor flow time, the main steam valve was closed. This started the liquid flow period. During the liquid flow period, the tray valves were opened to allow the liquid to flow to the tray below. The bottom tray was first opened fo r around 10 - 15 seconds , until the st r eam coming down from the tray nearly stopped. Next, the valve was closed, and the middle tray was opened. Lastly , the top tray valve was opened to allow the contents to fall to the middle tray and then closed again. The distillation continue d until the hydrometer read under 10% ABV at the beginning of a vapor flow period . 4.5 Data Collection and Analysis Throughout each distillation temperature, pressure, distillate flow rate, and total volume collected were taken at ti me increments roughly 10 minutes apart. Samples from each tray, the bottom of the column , and of the distillate were taken roughly 10 minutes apart and taken during 57 each vapor period for cyclic operation. The samples were placed directly into GC vials and labeled. Steam condens ate was collected into buckets and stored in a large storage tank to record overall volume after the distillation was complete. The distillate flow rate was measured using a stopwatch and graduated cylinder, collected distillate over a 10 second period and recorded as mL/s. The distillate was collected into volume incremented containers. After every distillation was complete samples were taken of the final contents of the pot and of the overall distillate, as well as of cuts if they we re taken. After the distillation was complete, each sample was r u n through the GC - 2010 gas chromatograph to analy ze the concentration of each component in each sample. The results can be seen in Chapter 5 . 58 C HAPTER 5: Results 5.1 Results A complete set of graphical results for every distillation and every compound studied can be found in Appendix A. Within this section are key results which will be discussed and referred to in the discussion section 5.1.1 All Distillation Results Table 5.1 is a summary table of all the dis tillations, included in it is the initial paramete rs, % recovered, and steam and time efficiencies. Table 5.1 Distillation Efficiency Results Figure 5.1 shows the hydrometer reading versus time throughout every distillation. D1 is not shown because not enough data was collected. 59 Figure 5.1 Ethanol C oncentration in Distillate vs Time for A ll Distillations Figure 5.2 shows the hydrometer reading vs volume for distillations D7 through D15. These distillations are the only distillation s in which volume d istilled data was collected , so they were selected for comparison. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 50 100 150 200 250 300 350 400 Alcohol by Volume (ABV) Time (min) D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 60 Figure 5.2 Ethanol Concentration vs Volume for Distillations D7 D1 5 As you can see in Figure 5.2, not all distillations collected the same volume of distillate, therefore to compare t he data, the x - axis was normalized by dividing the volume at a given point by the overall volume collected. This can be seen graphically in Figure 5.3 below. Figure 5.3 Normalized Ethanol Concentration vs Volume for Distillations D7 - D1 5 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 2 4 6 8 10 12 14 16 Alcohol by Volume (ABV) Volume (gals) D7 D8 D9 D10 D11 D12 D13 D14 D15 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Alcohol by Volume Fraction of Volume Distilled D7 D8 D9 D10 D11 D12 D13 D14 D15 61 5.1.2 Component Concentration Results Figures 5.4 5.11 show the trend in compound concentration versus fraction of volume distilled for distillation D7 D15. Figure 5.4 Normalized Acetaldehyde Concentration vs Volume for Distillations D7 - D1 5 Figure 5.5 Normaliz ed Acetone Concentration vs Volume for Distillations D7 - D1 5 0 1 2 3 4 5 6 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Concentration (g/L) Fraction of Volume Distilled D7 D8 D9 D10 D11 D12 D13 D14 D15 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Concentration (g/L) Fraction of Volume Distilled D7 D8 D9 D10 D11 D12 D13 D14 D15 62 Figure 5.6 Normalized Ethyl Acetate Concentration vs Volume for Distillations D7 - D1 5 Figure 5.7 Normalized Methanol Concentration vs Volume for Distillations D7 - D1 5 0 5 10 15 20 25 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Concentration (g/L) Fraction of Volume Distilled D7 D8 D9 D10 D11 D12 D13 D14 D15 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Concentration (g/L) Fraction of Volume Distilled D7 D8 D9 D10 D11 D12 D13 D14 D15 63 Figure 5.8 Normalized Propanol Co ncentration vs Volume for Distillations D7 - D1 5 Figure 5.9 Normalized Isobutanol Concentration vs Volume for Distillations D7 - D1 5 0 0.05 0.1 0.15 0.2 0.25 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Concentration (g/L) Fraction of Volume Distilled D7 D8 D9 D10 D11 D12 D13 D14 D15 0 0.2 0.4 0.6 0.8 1 1.2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Concentration (g/L) Fraction of Volume Distilled D7 D8 D9 D10 D11 D12 D13 D14 D15 64 Figure 5.10 Normalized Butanol Concentration vs Volume for Distillations D7 - D14 Figure 5.11 Normalized Isoa myl Alcohol Concentration vs Volume for Distillations D7 - D1 5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Concentration (g/L) Fraction of Volume Distilled D7 D8 D9 D10 D11 D12 D13 D14 D15 0 1 2 3 4 5 6 7 8 9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Concentration (g/L) Fraction of Volume Distilled D7 D8 D9 D10 D11 D12 D13 D14 D15 65 5.1.3 Results by Type of Distillation Overall, there were multiple types of distillations with and without reflux and with varying initial feed to the pot . Cyclic versus conventional operation , as w ell as , low wines distillation versus finishing run distillation. Below is a comparison table of operating parameters which led to the separation of the distillations graphically into four different categories. Table 5. 2 Distillations Separated by Type Operation Type Conventional Operation Conventional Operation Cyclic Operation Cyclic Operation Distillation Type Low Wines Finishing Low Wines Finishing Dist illation # D7 D3,D11,D12 D8,D9,D10 D4,D5,D6,D13,D14,D15 Low Wines Distillation Results Figure 5.12 below illustrates the low wines distillate results for cyclic versus conventional operation. Figure 5.12 Hydrometer ABV R eading vs Fraction of Volume Distilled for Low Wine Distillations 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Hydrometer ABV Fraction of Volume Distilled D7 - Conventional D8 - Cyclic (9,3) D9 - Cyclic (6,3) D10 - Cyclic (4,3) 66 Another important parameter to look at is the steam efficie ncy of each distillation. Steam efficiency was calculated as the lbs. of steam used per proof gallon distilled. For low wines, this can be seen in Figure 5.13 below. Figure 5.13 Steam Efficiency for All Low Wine Distillations The last measu re of efficie ncy was percent recovery which was calculated by dividing the total proof gallons distilled by the total initial proof gallons fed into the pot. For low wines this is illustrated in Figure 5.14 below. Figure 5.14 Percent Recovery for All Low Wine Distill ations 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0 lbs Steam / PG distilled D7 - Conventional D8 - Cyclic (9,3) D9 - Cyclic (6,3) D10 - Cyclic (4,3) 0.80 0.82 0.84 0.86 0.88 0.90 0.92 0.94 0.96 0.98 1.00 1.02 % Recovered D7 - Conventional D8 - Cyclic (9,3) D9 - Cyclic (6,3) D10 - Cyclic (4,3) 67 Finishing Distillation Results Figure 5.1 5 Hydrometer ABV R eading vs Fraction of Volume Distilled for Finishing Distillations Figure 5.1 6 Steam Efficiency for All Finishing Distillations 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Alcohol by Volume Fraction of Volume Distilled D11 - Coventional D12 - Conventional D13 - Cyclic (6,3) D14 - Cyclic (6,3) D15 - Cyclic (6,3) 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 lbs Steam / PG distilled D11 - Conventional D12 - Conventional D13 - Cyclic (6,3) D14 - Cyclic (6,3) D15 - Cyclic (6,3) 68 Figure 5.17 Percent Recovery for All Finishing Dist illations 5.1.4 Results of Cuts Taken Table 5.3 is a summary table of all the distillations in which cuts were taken. It shows the cuts by volume, and the concentrations of the key components initially and in each cut. Table 5.3 Cut Results for Distillati ons 11, 12, 13, 14 & 15 0.905 0.910 0.915 0.920 0.925 0.930 0.935 % Recovered D11 - Conventional D12 - Conventional D13 - Cyclic (6,3) D14 - Cyclic (6,3) D15 - Cyclic (6,3) 69 To best compare these values, the volumes were converted into proof gallons, which as a reminder is a gallon of al cohol at 50% alcohol by volume. The values were then normalized by dividing the proof gallons of a given cut by the overall proof gallons fed into the still. This shows are more direct comparisons of the amount of each cut produced as a percent of the init ial volume. F igures 5.18 5.20 show quantitatively the normalized proof gallon of each cut in comparison by distill ation. Figure 5.18 Normalized Proof Gallon of Heads Cut by Distillation 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 % PG of Initial PG D11 - Conventional D12 - Conventional D13 - Cyclic (6,3) D14 - Cyclic (6,3) D15 - Cyclic (6,3) 70 Figure 5.19 Normailized Proof Gallon of Hearts Cut by Distilla tion Figure 5.20 Normailized Proof Gallon of Tails Cut by Distillation 5.1.5 Temperature Profile Results Temper atures throughout the still and column were recorded during each distillation. Figures 5.21 - 22 show conventional and cyclic temperature prof iles for D11 and D 13. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 % PG of Initial PG D11 - Conventional D12 - Conventional D13 - Cyclic (6,3) D14 - Cyclic (6,3) D15 - Cyclic (6,3) 0 0.05 0.1 0.15 0.2 0.25 % PG of Initial PG D11 - Conventional D12 - Conventional D13 - Cyclic (6,3) D14 - Cyclic (6,3) D15 - Cyclic (6,3) 71 Figure 5.21 Temperature vs Time Profile for Conve n tional Operation (D11) Figure 5.22 Te mperature vs Time Profile for Cyclic Operation (D13) 5.2 Simulation Results The previous modeling methods were applied to the system studied using MATLAB. The MATLAB codes can be viewed in Appendix B. The initial conditions and other known parameters used for the simulations can be seen below in Table 5.4. 0 10 20 30 40 50 60 70 80 90 100 0 20 40 60 80 100 120 140 160 180 200 220 240 Time (min) T1 T2 T3 T4 T5 T6 0 10 20 30 40 50 60 70 80 90 100 0 20 40 60 80 100 120 140 160 180 200 220 240 Time (min) T1 T2 T3 T4 T5 T6 72 Table 5.4 Parameters for Batch Distillation Simulation Vapor Flow Rate (mole/min) 15 Reflux Ratio 1 Reboiler Hold Up (moles) 7000 Vapor Flow Period (min) 6 Total Operation Time (min) 185 Stage/Tray hold - up (moles) 56 Initial Condenser Mole fraction and Temperature ( ° C) (X1,T1) 0.73,55 Initial Reboiler mole fraction and Temperature ( ° C) (X2,T2) 0.64,55 Bottom tray mole fraction and Temperature ( ° C) (X3,T3) 0.61,55 Middle tray mole fraction an d Temperature ( ° C) (X4,T4) 0.56,55 Top Tray mole fraction and Temperature ( ° C) (X5,T5) 0.49,55 Partial condenser 2 mole fraction and Temperature ( ° C) (X6,T6) 0.36,55 Initial Reboiler mole fraction and Temperature ( ° C) (X7,T7) 0.10,55 For conven tional operation, the partial differential and algebraic equations were set up in the MATLA B composition and temperature over time. The resulting figures can be seen below in Figures 5.23 5.26. 73 Figure 5.23 MAT LAB S imulation of Conventional Distillation: Composition vs Time 74 Figure 5.24 MATLAB S imulation of Conventional Distillation: Tempera ture vs Time 75 Figure 5.25 MATLAB S imulation of Conventional Distillation: Moles in Pot vs Time 76 Figure 5.26 MATLAB S imulation of Conventional Distillation: Moles in Distillate vs Time first ran to determine the state of the system at the end of the vapor flow period. Plots of composi tion and temperature on the stages as a function of time were generated for each vapor flow period. The results of the first vapor flow period can be seen in Figures 5.27 - 28 below. 77 Figure 5.27 Stage and Reboiler Composition vs Time for First Vapor Flow P eriod 78 Figure 5.28 Stage and Reboiler Temperature vs Time for First Vapor Flow Period to determine the state of the system at the end of the liquid flow period. The values obtained were t hen entered back into the script process continued for 25 cycles. The data was entered in an excel spreadsheet as it was collected. The tables can be found in Appendix B. The graphical representation of the results can be seen in Figures 5.29 - 30 below. 79 Figure 5.29 Cyclic Batch Distillation Tray and Still Comp osition vs Time for 25 Cycles Figure 5.30 Cyclic Batch Distillation Tray and Still Temperature v s Time for 25 Cycles 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Ethanol Mole Frac. Cycle Total Con. Part. Cond. 2 Top Tray Middle Tray Bottom Tray Part. Cond. 1 Reboiler 50 55 60 65 70 75 80 85 90 95 100 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Temperature (deg C) Cycle Total Con. Part. Cond. 2 Top Tray Middle Tray Bottom Tray Part. Cond. 1 Reboiler 80 5. 3 Reproducibility Due to the le ngth of each distillation to run as well as to analyze the samples, one triplication experiment was r u n under the same conditions D13, D14 and D15 to prove reproducibility. The normalized heads , hearts , and tails cuts were used to compare to the normalized cuts taken in D11 and D12. First the heads, hearts and tails percentages were averaged. Next, the standard deviation was calculated and used to determine a range of values which would be accepted as within one standard deviation and therefore statisticall y the same. Then the values from D11 and D12 were determined to be within or outside the respected ranges. The values can be seen in Table 5. 5 below. Table 5.5 Statistical Analysis of Reproducibility As seen in Table 5.4, the standard deviation of the t riplicate is very small averaging around 0.017, deeming the distillation results as the same and therefore reproducible. The distillation results of D11 and D12 were determ ined to be statistically different from D13, D14 and D15 because the values were all out of the range of 1 standard deviation. 81 C HAPTER 6: Discussion 6.1 Summary In summary, cyclic distillation on the batch column used was able to show a decrease in en ergy (steam) requirements for finishing runs but not in stripping runs. Cyclic disti llation trends included the following: ethanol concentration decreased at a slower rate compared to conventional operation and as a result temperature profiles mimicked th is phenomenon. I t was shown that the volume of hearts (product) was increased and vo lume of tails and heads (unwanted by - product s ) was decreased . This research has shown that the application of cyclic distillation on spirit production is a viable option for distilleries large and small. It is suggested that cyclic distillation is applied to other types of spirits, to columns with a larger number of trays , and to trays with true plug flow capability. 6.2 Conclusions 6. 2.1 All Distillation Discussion There are a few key differences when broadly comparing cyclic to conventional operation. Fir st, it was observed that for nearly the same initial volume and ABV as feed into the still, cyclic operation took much longer to process than conventional operatio n. This trend makes sense because during conventional operation, distillate is always being c ollected . During cyclic operation distillate is only being collected during the vapor flow period, and during the beginning of the liquid flow period when the cond enser is draining. Secondly, it was noticed that over time and over volume collected the hyd rometer reading stayed higher for longer. This is illustrated in Figure 5.3. Rather than start high and decrease gradually throughout the distillation like in conv entional operation, during cyclic operation the 82 distillate concentration stayed high and then dramatically decreased towards the end of the distillation. Additionally , it was observed during each distillation that during conventional operation the hydrome ter reading started high, near 90% ABV, and then gradually decreased over time. However , duri ng cyclic operation, the hydrometer started high, near 90% ABV and gradually decreased during the vapor flow period. However, it was also observed that in the begi nning of a vapor flow period, the hydrometer reading would increase about 3 - 5% from what it w as at the end of the liquid flow period. Both phenomena can be explained by the collected and essentially re - distillation of the column and piping vapor and liquid contents. Rather than allowing the feed to vaporize and condense and increasingly become les s concentrated as the distillation continues, some of the ethanol and other alcohols return to the pot and column, therefore causing a disruption in this trend. L astly, Figures 5.21 - 22 display that the temperature in the column and top of the still helmet increase at a slower rate during cyclic distillation compared to during conventional operation. The temperature profiles were as expected because as the ethanol concentration decreases, the water concentration increases. Since water has a higher boiling point than ethanol, when there is a lower mole fraction of ethanol, the boiling point of the mixture is expected to increase . As the ethanol concertation decreases at a slower rate throughout the pot and column, the temperatures will increase at a slower r ate as well. 6.2.2 Component Concentration Trends Some compou nds were not affected by cyclic operation including acetaldehyde, acetone, and ethyl aetate. The trends can be seen in Figures 5.4 - 5.6. This makes sense because of the relative boiling points o f these compounds compared to ethanol. The boiling point of th ese 83 compounds are all lower than ethanol, therefore they are concentrated in the distillate in the beginning of the distillation and are not subject to cyclic operation. Methanol also has a lo wer boiling point than ethanol (64.7 ° C) however it is well kno wn in industry and literature to start at a high concentration in the heads, decrease throughout the hearts cut and then increase again in the tails cut. The trend observed in Figure 5.7 is like the trend Claus and Berglund found in distilling a cherry dis tillate [ 43]. Higher alcohols including propanol, isobutanol, butanol and isoamyl alcohol with higher boiling points than ethanol all exhibited similar trends and trend changes. A normal trend seen in the previous report is a gradual increase from the beg inning to the end of the distillation and a big increase in the tails fraction [43]. This trend was inverted for the distillations when the reflux was not engaged to the partial condensers . When the partial condensers are active, the vapor the heavier and less volatile components of the vapor that travels through them are more likely to condense in turn holding them back from entering the final condenser and into the distillate. 6.2.3 Low Wines v s Finishing Tends Cyclic distillation of low wines did not seem to have any noticeable changes when compared to conventional distillation of low wines. The efficiency measures compared in Figure 5.13 and 5.14 show conventional operation to be more efficien t than cyclic operation. This is attributed to the fact tha t the initial ABV of the feed is very low, roughly 10% ABV and that there are only three trays which can allow cyclic distillation to occur. In distilled sprit products this is one of the most ener gy intensive operations , because it takes more energy to he at a mixture that is 90% water than a mixture that is 70 - 75% water, and thus would be an ideal application of the 84 benefits of cyclic distillation. It is believed that with more trays and trays that allow true plug flow, the energy saving advantage of cycli c distillation would be observed. I ncrease s in efficienc y measures were seen in the finishing runs observed in Figures 5.15 - 1 7 . In cyclic distillations D13 - D15 the ABV reading on the hydrometer re mained higher for longer compared to the identical distilla steam usage per proof gallon distilled was the lowe r for D13 - D15 compared to the identical convent ional operation D11 , proving an increasing in energy effici ency. Lastly, the percent recovery values were too spread out to be able to confirm the percent recovery was better in cyclic versus conventional operation. 6.2.4 Effects on Cuts The last f ive d istillations were used to identify what effect cyclic disti llation ha d on the heads, hearts and tails cuts. The results can be referred to in Table 5.3 and Figures 5.18 - 20. The heads cut was statistically smaller during cyclic operation than in conventio nal operation. The hearts cut was larger for cyclic operati on than in conventional operation and lastly the tails cut was drastically smaller for cyclic operation than in conventional operation. This is best explained by the earlier mentioned phenomenon that the ethanol concentration stays higher for longer thro ughout the distillation during cyclic operation. If the ethanol concentration remains high and the fusel oils/ congeners do not start coming through, the distillate remains clean and pleasant to th e distiller and the hearts cut increases in proof gals. Thi s consequence allows most of the ethanol to remain in the hearts cut. Since the proof gallons collected in the hearts cut increased, the proof gallons in the tails cut decreased. 85 Based on the previ ous observations it can be concluded that cyclic distillati on causes an increase in proof gallon of hearts. This in turn increases the production efficiency and decreases energy requirement per proof gal of product. In addition, decreasing the volume of by products produced reduces the storage and processing requi rements , decreasing cost per proof gal produced. 6.2.6 Reflux Observations and Discussion It was observed that regardless of mode of operation, when reflux to the partial condenser was turned on a smaller volume of hearts was collected, however it was hig her in alcohol by volume. This is because when the partial condensers are active, they act as an additional stage by sending some condensed vapor to the preceding tray or to the reboiler/pot, depen ding on the location of the partial condenser. When the par tial condensers are not active, vapor travels past therefore it is suitable to assume they wi ll be used when making most types of spirits. 6.2.7 Simula tion Discussion Results from the simulation and modeling of conventional and cyclic operation can be seen in Figures 5.23 - 5.26 and Figures 5.27 - 5.30, respectively. The trends seen in the data were also seen in the simulation s . Temperatures in conventional operation increased at a faster rate th a n in cyclic operation. Distillate ethanol concentration decreased over time at a faster rate than in cyclic operation. During cyclic distillation modeling, some of the ethanol is transferred to the tray below, causi ng the ethanol to be held back in the column and thus the tray composition to increase and decrease slightly from cycle to cycle. This trend of cyclic tray composition is seen graphically in Figure 5.29. The model calculates the temperature based on the et hanol composition, therefore temperature and concentration are directly proportional and can be seen in Figure 5.30. 86 Lastly, given the same processing time cyclic operation was able to process less moles in the pot than in conventional operation. This was also seen in the data and can be explained on the basis that during the liquid flow period, no distillate is collected, leading to a longer distillation time. 6.3 Limitations and Sources of Error One key limitation is the lack of separation between liquid s drained from tray to tray. The trays in the Carl still have small holes that allow the liquid to fall to the previous tray regardless of whether it is during the vapor flow or the liquid flow per iod. Therefore, it was not possible to separate the liquid hold up traveling from tray to tray during each liquid flow period. Since the still used was not specifically designed for cyclic distillation, the still was not able to achieve the 200% efficiency increase predicted by theoretical models. Is it important to note that cuts are taken by the distiller and are predominantly made by sensory analysis. In industry, it is common for distillers to go through specific sensory training to become more attune with the flavor and aroma of the compounds in t he spirit. The distiller of this research has two - year part - time experience distilling which is more than the average research er , but less than a full - time trained distiller. Many factors play a role in senso ry analysis including what and what time the di stiller has last ate or drank can affect the sensory perception, therefore cuts cannot be 100% consistent. Another important note is the inconsistency of initial volume and proof gallons in the feed. When lar ge volume and large equipment is used, it becom es increasingly hard to duplicate an exact volume and initial ABV due to the necessary pumping into the still, and losses associated with the hoses. 87 6.4 Recommendations for Future Research and Application It is strongly suggested that cyclic distillation be applied to other types of spirits. This research was solely based on b r andy made from fermented apple cider, both the stripping and the finishing runs. While the energy requirements are mostly like ly the s ame, different spirits are fermented with diffe rent yeast strains and ultimately have very different components and concentrations of key flavor and aroma attributes. Rum, whiskey, gin and vodka production are all sprits that can be further investigated. L egally, vodka must be distilled to 95% alcohol by volume, so its production requires more trays than all other spirits. Therefore, the application of cyclic distillation for vodka production is a promising addition to this research for both energy conserva tion and efficiency purposes. In addition to ot her spirits, implementing the already designed Maleta © trays or a new tray design which maximizes true plug flow would increase the positive effects of cyclic distillatio n. 88 APPENDICES 89 APPENDIX A: Graphi cal Distillation Results Tray by Tray Results : Distillation Figure A.1 Normal Operation 1, Tray 1 Graphical Results Figure A.2 Normal Operation 1, Tray 2 Graphical Results 0 100 200 300 400 500 600 700 0 1 2 3 4 5 6 35 54 72 104 119 137 154 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 0 1 2 3 4 5 6 1 2 3 4 5 6 7 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 90 Figure A.3 Normal Operation 1, Tray 3 Graphical Results Distillation 2 Fi gure A.4 Normal Operation 2 , Tray 1 Graphical Results 0 100 200 300 400 500 600 700 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1 2 3 4 5 6 7 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 50 100 150 200 250 300 0 0.5 1 1.5 2 2.5 3 3.5 4 67 78 89 103 117 135 155 168 183 200 213 229 248 266 284 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 91 Figure A.5 Normal Operation 2 , Tray 2 Graphical Results Figure A.6 Normal Operation 2 , Tray 3 Graphical Results 0 50 100 150 200 250 300 350 400 450 0 1 2 3 4 5 6 7 68 79 90 103 118 135 155 169 183 201 213 229 248 267 285 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 50 100 150 200 250 300 350 400 450 500 0 1 2 3 4 5 6 7 8 69 80 90 104 118 137 156 170 184 202 214 250 268 286 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 92 Distillation 3 Figure A.7 Normal Operation 3 , Tray 1 Graphical Results Fig ure A.8 Normal Operation 3 , Tray 2 Graphical Results 0 100 200 300 400 500 600 0 0.5 1 1.5 2 2.5 3 3.5 4 55 65 77 86 96 107 116 126 137 146 157 168 180 191 202 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 0 1 2 3 4 5 6 55 66 77 86 97 108 117 127 137 147 157 169 181 191 202 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 93 Figure A.9 Normal Operation 3 , Tray 3 Graphical Results Distillation 4 Figure A.10 Cyclic Distillation 1 (Vapor 9 min, Liquid 3 min) , Tray 1 Graphical Results 0 100 200 300 400 500 600 0 1 2 3 4 5 6 57 66 78 87 97 108 118 127 138 147 158 169 182 192 203 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol 0 100 200 300 400 500 600 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 31 44 56 67 80 92 104 116 130 142 155 167 178 192 204 216 230 Ethanol Concentration(g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol 94 Figure A.11 Cyclic Distillatio n 1 (Vapor 9 min, Liquid 3 min) , Tray 2 Graphical Results Figure A.12 Cyclic Distillation 1 (Vapor 9 min, Liquid 3 min) , Tray 3 Graphical Results 0 100 200 300 400 500 600 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 32 45 56 67 80 93 104 116 130 142 155 168 179 192 205 217 231 Ethanol Concentration(g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol 0 100 200 300 400 500 600 0 1 2 3 4 5 6 33 45 57 68 81 93 104 116 130 143 156 168 179 192 205 217 231 Ethanol Concentration(g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol 95 Distillation 5 Figure A.13 Cyclic Distillation 2 (Vapor 6 min, Liquid 3 min) , Tray 1 Graphical Result s Figure A.14 Cyclic Distillation 2 (Vapor 6 min, Liquid 3 min) , Tray 2 Graphical Results 0 100 200 300 400 500 600 700 0 1 2 3 4 5 6 7 26 55 64 73 82 92 101 110 120 131 139 148 158 168 1787 186 196 205 214 224 233 242 251 261 270 280 290 299 309 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol 0 100 200 300 400 500 600 700 0 1 2 3 4 5 6 7 8 27 55 65 75 84 93 102 111 122 132 141 149 159 168 178 186 196 206 215 224 233 242 252 261 270 281 290 299 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 96 Figure A.15 Cyclic Distillation 2 (Vapor 6 min, Liquid 3 min) , Tray 3 Graphical Results Distillation 6 Figure A.16 C yclic Distillation 3 (Vapor 4 min, Liqu id 3 min) , Tray 1 Graphical Results 0 100 200 300 400 500 600 700 800 0 1 2 3 4 5 6 28 56 66 75 84 94 103 112 123 133 141 150 160 169 178 187 197 206 215 224 233 243 252 261 271 281 291 300 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 0 1 2 3 4 5 6 7 8 39 59 73 88 103 117 132 147 161 177 191 206 221 236 250 268 282 298 313 327 342 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 97 Figure A.17 Cyclic Distillation 3 (Vapor 4 min, Liquid 3 min) , Tray 2 Graphical Results Figure A.18 Cyclic Distillation 3 (Vapor 4 min, Liquid 3 min) , Tray 3 Graphical Res ults 0 100 200 300 400 500 600 700 0 1 2 3 4 5 6 7 8 9 10 40 59 74 88 103 117 131 148 162 177 191 206 221 236 251 268 283 298 313 327 343 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 800 0 1 2 3 4 5 6 7 8 40 60 74 89 103 118 132 148 162 178 191 207 222 236 251 268 283 298 313 328 343 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 98 Distillation 7 Figure A.19 Nor mal Operation Low Wines , Tray 1 Graphical Results Figure A.20 Normal Operation Low Wines , Tray 2 Graphical Results 0 10 20 30 40 50 60 70 80 90 0 0.05 0.1 0.15 0.2 0.25 30 38 46 53 60 67 70 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 50 100 150 200 250 300 0 0.5 1 1.5 2 2.5 3 3.5 31 38 46 53 61 68 70 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 99 Figure A.21 Normal Operation Low Wines , Tray 3 Graphical Results Distillation 8 Figure A.22 Cyclic Operation Low Wines, C yclic Distillation (Vapor 9 min, Liquid 3 min) Tray 1 Graphical Results 0 50 100 150 200 250 300 350 0 1 2 3 4 5 6 31 39 46 53 61 68 71 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 50 100 150 200 250 0 0.5 1 1.5 2 2.5 3 43 53 76 89 104 115 126 138 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 100 Figure A.23 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 9 min, Liquid 3 min) Tray 2 Graphical Results Fig ure A.24 Cyclic Operation Low Wines, Cyclic Distillation ( Vapor 9 min, Liquid 3 min) Tray 3 Graphical Results 0 100 200 300 400 500 600 0 1 2 3 4 5 6 7 8 9 10 43 54 77 90 104 115 126 139 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 50 100 150 200 250 300 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 43 54 77 90 104 116 127 139 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 101 Distillation 9 Figure A.25 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 6 min, Liquid 3 min) Tray 1 Graphical Results Figure A.26 Cyclic Operation Low Wines, Cyclic Distillation (Vap or 6 min, Liquid 3 min) Tray 2 Graphical Results 0 50 100 150 200 250 0 0.5 1 1.5 2 2.5 40 52 60 71 78 89 99 107 116 125 134 143 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 50 100 150 200 250 300 350 0 0.5 1 1.5 2 2.5 3 40 53 61 71 79 90 99 107 117 125 134 143 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 102 Figure A.27 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 6 min, Liquid 3 min) Tray 3 Graphical Results Distillation 10 Figure A.28 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 4 min, Liquid 3 min) Tray 1 Graphical Results 0 50 100 150 200 250 300 350 0 1 2 3 4 5 6 41 53 61 72 79 90 99 108 117 126 135 144 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 50 100 150 200 250 300 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 64 74 81 88 96 103 111 119 126 134 141 149 156 163 171 178 186 193 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 103 Figure A.29 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 4 min, Liquid 3 min) Tray 2 Graphical Results F igure A.30 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 4 min, Liquid 3 min) Tray 3 Graphical Results 0 50 100 150 200 250 300 350 0 1 2 3 4 5 6 7 64 74 82 88 96 103 111 119 126 134 142 149 156 164 172 179 186 193 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 50 100 150 200 250 300 350 400 0 1 2 3 4 5 6 65 75 82 89 97 105 112 120 127 135 143 150 157 165 172 179 186 194 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 104 Distillation 11 Figure A.31 Normal Operation Brandy Finishing Run w/ reflux - Tray 1 Graphical Results Figure A.32 Normal Operation Brandy Finishing Run w/ reflux - Tray 2 Graphical Results 0 50 100 150 200 250 300 350 400 450 500 0 1 2 3 4 5 6 7 8 9 31 40 50 59 66 74 83 93 101 109 116 126 Ethanol Concentration (g/L) Concentration (g/L) Time(min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 0 2 4 6 8 10 12 32 41 50 59 67 75 84 93 101 109 117 126 Ethanol Concentration (g/L) Concentration (g/L) Time(min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 105 Figure A.33 Normal Operati on Brandy Finishing Run w/ reflux - Tray 3 Graphical Results Distillation 12 Figure A.34 Normal Operation Brandy Finishing Run w/o reflux - Tray 1 Graphical Results 0 100 200 300 400 500 600 700 0 2 4 6 8 10 12 32 41 49 59 67 75 84 93 102 109 117 128 Ethanol Concentration (g/L) Concentration (g/L) Time(min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 50 100 150 200 250 300 350 400 450 500 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 36 43 51 57 65 72 79 85 92 99 106 114 120 127 134 141 148 155 Time (min) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 106 Figure A.35 Normal Operation Brandy Finishing Run w/o reflux - Tray 2 Graph ical Results Figure A.36 Normal Operation Brandy Finishing Run w/o reflux - Tray 3 Graphical Results 0 100 200 300 400 500 600 0 1 2 3 4 5 6 36 44 51 58 65 72 79 86 93 99 107 114 121 128 135 142 148 155 Time (min) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 0 1 2 3 4 5 6 37 44 51 58 65 72 79 86 93 100 107 114 121 128 135 142 149 155 Time (min) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 107 Distillation 13 Figure A.37 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 1 Graphical Results Figure A.38 Cyclic Operatio n Brandy Finishing Tray by Tray w/ Reflux - Tray 2 Graphical Results 0 100 200 300 400 500 600 700 0 1 2 3 4 5 6 7 8 37 48 57 64 73 80 88 95 103 109 118 124 133 140 148 155 163 170 177 184 192 200 207 215 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 800 0 2 4 6 8 10 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 108 Figure A.39 Cyclic Operation Brandy Finishing Tray by Tray w/ R eflux - Tray 3 Graphical Results Distillation 14 Figure A.40 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 1 Graphical Results 0 100 200 300 400 500 600 700 800 0 2 4 6 8 10 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 0 2 4 6 8 10 12 34 43 50 56 63 71 78 85 92 100 107 114 122 129 136 143 148 158 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 109 Figure A.41 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 2 Graphical Results Figure A.42 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 3 Graphical Results 0 100 200 300 400 500 600 700 0 2 4 6 8 10 12 14 34 43 50 56 64 71 78 86 92 100 107 115 122 129 136 143 150 158 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 0 2 4 6 8 10 12 14 34 43 50 56 64 71 78 86 93 100 108 115 122 129 137 144 151 158 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 110 Distillation 15 Figure A.43 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 1 Graphical Results Figure A.44 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 2 Graphical Results 0 100 200 300 400 500 600 0 1 2 3 4 5 6 7 8 9 28 38 45 52 59 67 74 81 89 96 104 111 118 125 134 141 149 157 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 0 2 4 6 8 10 12 28 38 46 52 59 67 74 81 89 96 104 112 118 126 134 142 149 157 164 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 111 Figure A.45 Cyclic Operation Brandy Finishing Tray by Tray w/ Reflux - Tray 3 Graphical Results 0 100 200 300 400 500 600 700 0 2 4 6 8 10 12 29 38 46 52 59 67 75 82 89 96 105 112 118 126 135 142 149 157 164 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 112 Distillate Results Distillation 1 Figure A.46 Normal Operation 1 Distillat e Graphical Results Distillation 2 Figure A.47 Normal Operation 2 Distillate Graphical Results 480 500 520 540 560 580 600 620 640 660 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1 2 3 4 5 6 7 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 62 77 88 101 116 133 152 166 180 197 211 226 245 264 282 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol 113 Distillation 3 Figure A.48 Normal Operation 3 Distillate Graphical Results Distillation 4 Figure A.49 Cyclic Distillation 1 (Vapor 9 min, Liquid 3 min) Distillate Graphical Results 0 100 200 300 400 500 600 700 0 0.5 1 1.5 2 2.5 3 3.5 4 56 67 78 87 97 109 118 127 138 148 158 169 182 193 203 217 230 246 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 31 43 57 68 81 94 105 117 131 143 156 169 180 193 206 218 232 242 255 269 Ethanol Concentration(g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 114 Distillation 5 Figure A.50 Cyclic Distillation 1 (Vapor 6 min, Liquid 3 min) Disti llate Graphical Results Distillation 6 Figure A.51 Cyclic Distillation 1 (Vapor 4 min, Liquid 3 min) Distillate Graphi cal Results 0 100 200 300 400 500 600 700 800 0 1 2 3 4 5 6 29 56 65 74 84 93 102 112 122 132 141 150 159 169 179 187 197 206 215 228 234 243 253 262 271 282 291 300 309 318 327 337 346 354 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 800 0 1 2 3 4 5 6 38 59 73 87 102 116 132 147 161 176 190 206 220 236 252 267 282 297 312 326 342 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 115 Distillation 7 Figure A.52 Normal Operation Low Wines Distillate Graphical Results Distillation 8 Figure A.53 Cy clic Operation Low Wines, Cyclic Distillation (Vapor 9 min, Liquid 3 min) Distillate Graphical Results 0 100 200 300 400 500 600 700 800 0 1 2 3 4 5 6 7 8 29 37 44 52 59 67 73 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 800 0 2 4 6 8 10 12 42 52 76 87 100 112 125 137 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 116 Distillation 9 Figure A.54 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 6 min, Liquid 3 min) Distillate Graphical Results Distillation 1 0 Figure A.55 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 4 min, Liquid 3 min) Distillate Graphical Results 0 100 200 300 400 500 600 700 800 900 0 2 4 6 8 10 12 39 52 60 70 78 87 97 106 116 125 133 142 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 800 0 2 4 6 8 10 12 14 16 63 74 81 87 96 103 111 118 125 133 141 148 156 163 171 178 185 192 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 117 Distillation 11 Figure A.56 Normal Operation Brandy Finishing Run w/o reflux - Distillate Graphical Results Distillation 12 Fi gure A.57 Normal Operation Brandy Finishing Run w/ reflux - Distillate Graphical Results 0 100 200 300 400 500 600 700 800 0 5 10 15 20 25 31 40 50 57 65 74 83 92 100 108 116 126 Ethanol Concentration (g/L) Concentration (g/L) Time(min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 800 0 1 2 3 4 5 6 35 43 50 57 64 71 78 85 92 99 106 113 120 127 134 141 148 155 Time (min) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 118 D istillation 13 Figure A.58 Cyclic Operation Tray by Tray Finishing Run w/ reflux Distillate Graphical Results Distillation 14 Figure A.59 Cyclic Operation Tray by Tray Finishing Run w/ reflux Distillate Graphical Results 0 100 200 300 400 500 600 700 800 0 2 4 6 8 10 12 14 36 47 56 64 72 80 87 95 102 109 118 124 133 140 147 155 162 169 177 184 191 199 207 215 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 800 900 0 2 4 6 8 10 12 14 16 33 42 49 56 63 70 77 85 92 99 107 114 121 128 135 143 150 157 164 Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 119 Distillation 15 Figure A.60 Cyclic Operation Tray by Tray Finishing Run w/ reflux Distillate Graphical Results 0 100 200 300 400 500 600 700 800 0 2 4 6 8 10 12 28 37 45 51 58 66 74 80 88 95 104 111 118 125 134 141 149 156 163 Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 120 Bottom Column and Cycled Volume Results Distillation 1 Figu re A.61 Normal Operation 1 Bottom Column Graphical Results Distillation 2 Figure A.62 Normal Operation 2 Bottom Column Graphical Results 0 100 200 300 400 500 600 0 1 2 3 4 5 6 53 71 103 118 136 153 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 50 100 150 200 250 300 350 400 450 500 0 0.5 1 1.5 2 2.5 3 3.5 67 78 89 102 117 133 154 167 182 199 212 228 247 265 283 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 121 Distillation 3 Figure A.63 Normal Operation 3 Bottom Column Graphical Results Distillation 4 Figure A.64 Cyclic Distillation 1 (Vapor 9 min, Liquid 3 min) Cycled Volume Graphical Results 0 100 200 300 400 500 600 700 0 0.5 1 1.5 2 2.5 3 3.5 4 55 65 76 85 95 107 115 125 136 146 156 167 180 190 201 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Ethanol Concentration(g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 122 Distillation 5 Figure A.65 Cyclic Distillation 1 (Vapor 6 min, Liquid 3 min) Cycled Volume Graphical Results Distillation 6 Figure A.66 Cyclic Distillation 1 (Vapo r 4 min, Liquid 3 min) Cycled Volume Graphical Res ults 0 100 200 300 400 500 600 700 0 1 2 3 4 5 6 50 60 70 79 88 97 107 116 126 135 144 154 164 173 182 191 201 210 219 228 237 246 256 265 276 285 294 304 313 322 331 341 350 359 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 0 1 2 3 4 5 6 7 54 69 84 99 114 129 144 159 174 188 202 217 232 257 272 287 301 316 331 346 360 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 123 Distillation 7 Figure A.67 Normal Operation Low Wines Bottom Column Graphical Results Distillation 8 Figure A.68 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 9 min, Liquid 3 min ) Cycled Volume Graphical Results 0 50 100 150 200 250 300 350 400 450 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 30 37 45 52 60 67 70 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 10 20 30 40 50 60 70 80 90 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 48 71 83 96 109 120 133 145 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 124 Distillation 9 Figure A.69 Cyclic Operation Low Wines, Cyclic Distillation (Vapor 6 min, Liquid 3 min) Cycled Volume Graphical Results Distillation 10 Figure A.70 Cyclic Operation Low Wines, Cyclic Distillation (V apor 4 min, Liquid 3 min) Cycled Volume Graphical Re sults 0 20 40 60 80 100 120 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 45 55 65 74 84 94 103 112 121 130 139 148 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 10 20 30 40 50 60 70 80 90 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 68 77 81 87 96 103 111 118 125 133 141 148 160 163 171 178 189 196 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 125 Distillation 11 Figure A.71 Normal Operation Brandy Finishing Run w/o reflux - Bottom Column Graphical Results Distillation 12 Figure A.72 Normal Operation Brandy Finishing Run w/ reflux - Bottom Column Graphical Results 0 100 200 300 400 500 600 0 1 2 3 4 5 6 33 42 52 60 67 76 85 94 101 110 117 127 Ethanol Concentration (g/L) Concentration (g/L) Time(min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 0 0.5 1 1.5 2 2.5 3 3.5 4 37 45 52 58 66 73 79 86 93 100 108 115 122 128 135 142 149 156 Time (min) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 126 Distillation 13 Figure A.73 Cyclic Operation Tray by Tray Finishing Run w/ reflux Cycled Volume Graphical Results Distillation 14 Figure A.74 Cyclic Operation Tray by Tray Finishing Run w/ reflux Cycled Volume Graphical Results 0 100 200 300 400 500 600 700 800 0 2 4 6 8 10 12 43 52 61 69 76 84 90 99 106 114 121 129 137 144 150 159 166 173 181 189 196 204 212 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 0 100 200 300 400 500 600 700 0 1 2 3 4 5 6 7 8 9 40 46 53 60 68 76 83 89 96 104 111 119 126 133 141 148 155 163 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 127 Distillation 15 Figure A.75 Cyclic Operation Tray by Tray Finishing Run w/ reflux Cycled Volume Graphical Results 0 100 200 300 400 500 600 0 1 2 3 4 5 6 7 8 9 10 34 42 49 56 63 70 79 86 92 100 107 115 122 130 139 146 153 161 Ethanol Concentration (g/L) Concentration (g/L) Time (min) Acetaldehyde Acetone Ethyl Acetate Methanol Propanol Isobutanol Butanol Isoamyl Alcohol Ethanol 128 Component Results Figure A.76 Acetaldehyde Concentration vs T ime for All Distillations Figure A.77 Acetone Co ncentration vs T ime for All Distillations 0.000 1.000 2.000 3.000 4.000 5.000 6.000 0 50 100 150 200 250 300 350 400 Concentration (g/L) Time (min) D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 0.000 0.010 0.020 0.030 0.040 0.050 0.060 0.070 0 50 100 150 200 250 300 350 400 Concentration (g/L( Time(min) D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 129 Figure A.78 Ethyl Acetate Concentration vs T ime for All Distillations Figure A.79 Methanol Concentration vs T ime for All Distillations 0.000 5.000 10.000 15.000 20.000 25.000 0 50 100 150 200 250 300 350 400 Concentration (g/L) Time (min) D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 0 50 100 150 200 250 300 350 400 Concentration (g/L) Time (min) D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 130 Figure A.80 Ethanol Concentration vs T ime for All Distillations F igure A.81 Propanol Concentration vs T ime for All Distillations 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 50 100 150 200 250 300 350 400 Hydrometer ABV Time (min) D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0 50 100 150 200 250 300 350 400 Concentration (g/L) Time (min) D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 131 Figure A.82 Isobutanol Concentration vs T ime for All Distillations Figure A.83 Butanol Concentration vs T ime for All Distillations 0.000 0.200 0.400 0.600 0.800 1.000 1.200 0 50 100 150 200 250 300 350 400 Concentration (g/L) Time (min) D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 0 50 100 150 200 250 300 350 400 Concentration (g/L) Time (min) D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 132 Figure A.84 Isoamyl Alcohol Concentration vs T ime fo r All Distillations -1.000 0.000 1.000 2.000 3.000 4.000 5.000 6.000 7.000 8.000 9.000 0 50 100 150 200 250 300 350 400 Concentration (g/L) Time (min) D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 133 A PPENDIX B : MATLAB Codes Conventional Simulation MATLAB Codes 134 135 136 137 Cyclic Simulation MATLAB Codes 138 139 140 141 MATLAB Cyclic Data Table B.1 Cyclic Distillation Simulation Data 142 A PPENDIX C : Miscellaneous Figures/Tables Figure C1. Example Chromatograph Results 143 BIBLIOGRAPHY 144 BIBLIOGRAPHY 1) Cannon MR. Controlled Cycling Improves Various Processes. Ind Eng Chem . 1961;53:629 - 629. 2) G aska RA, Cannon MR. Controlled Cycling Distillation in Sieve and Screen Plate Towers. I nd Eng Chem . 1961;52:630 - 631. 3) McWhirter JR, Cannon MR. Controlled Cycling Distillation in a Packed - Plate Column. Ind Eng Chem . 1961;52:632 - 634. 4) S chrodt VN. Controlled Cycle Operations . Ind. Engng Chem. Fundls. 1965;4:108 - 110. 5) Sommferfeld JT, Schrodt VN, Parisot PE, Chien HH. Studies of Controlled Cyclic Distillation: I. Computer Simulations and the Analogy with Conventional Operation. Separat. Sci. 1966;1:245 - 279. 6) Ch ien HH, Sommferfeld JT, Schrodt VN, Parisot PE. Studies of Controlled Cyclic Distillatio n: II. Analytical Transient Solution and Asymptotic Plate Efficiencies. Separat. Sci. 1966;1:281 - 317. 7) Schrodt VN, Sommerfeld JT, Martin OP, Parisot PE, Chien HH. Plant - scale study of controlled cyclic distillation. Chem Eng Sci . 1966;22:759 - 767. 8) Horn FJM . Periodic Countercurrent Processes. I&EC Process Design and Develop. 1966;6:30 - 35. 9) Robinson RG, Engel AJ. An Analysis of Controlled Cycling Mass Transfer Operations. Ind Eng Chem . 1967;59:22 - 29. 10) May RA, Horn FJM. Stage Efficiency of a Periodically Opera ted Distillation Column. I&EC Process Design and Develop. 1968;7:61 - 64 11) May RA, Horn FJM. Effect of Mix ing on Periodic Countercurrent Processes. I&EC Process Design and Develop. 1968;7:349 - 354. 12) Furzer IA. Periodic Cycling of Plate Columns. Chem Eng Sci. 1 973;28:296 - 299. 13) G - Cap Fractionating Colum n in a Cyclic Regime. Chem Petro Eng . 1976;11:707 - 709. 14) Rivas OR. An Analytical Solution of Cyclic Mass Transfer Operations. Ind Eng Chem, Process Des De v. 1977;16:400 - 402. 145 15) Duffy GJ, Furzer IA. Periodic Cycling of Plate Columns . Chem Eng Sci . 1978;33:897 - 904 16) Furzer IA. The Discrete Residence Time Distribution of a Distillation Column Operated with Microprocessor Controlled Periodic Cycling. Canadian Jour nal of Chem Eng. 1978;56:747 - 750. 17) Goss DW, Furzer IA. Mass Transfer in Periodically Cycled Plate Colum ns Containing Multiple Sieve Plates. AIChe Journa l. 1980;26:663 - 669. 18) Baron G, Wajc S, Lavie R. Stepwise Periodic Distillation I: Total Reflux Operatio n. Chem Eng Sci. 1980.;35:859 - 865 19) Baron G, Wajc S, Lavie R. Stepwise Periodic Distillation II: Separation of a Binary Mixture. Chem Eng Sci. 1981;36:1819 - 1827. 20) Matsubara M, Watanabe N, Kurimoto H. Binary Periodic Distillation Scheme with Enhanced Energ y Conservation. Chem Eng Sci . 1985; 40:715 - 721. 21) Matsubara M, Watanabe N, Kurimoto H, Shimizu K. Binary Periodic Distillation Scheme with Enhanced Energy Conservation - II. Chem Eng Sci . 1985;40:755 - 758. 22) Thompson MF, Furzer IA. Hydrodynamic Modelling for L iquid Holdups in Periodically Cycle d Plate Columns. AIChe Journal. 1984;30:496 - 500. 23) Szonyi L, Furzer IA. Periodic Cycling of Distillation Columns Using a New Tray Design. AIChe Journal. 1985;31:1707 - 1713. 24) Bausa J, Tsatsaronis G. Reducing the Energy Deman d of Continuous Distillation Proces ses by Optimal Controlled Forced Periodic Operation. Comp Chem Eng. 2001;25:359 - 370. 25) Flodman HR, Timm DC. Batch Distillation Employing Cyclic Rectification and Stripping Operations. ISA Trans . 2012;51:454 - 460. 26) Lita I, B ilde CS, Kiss AA. Modeling, Design and Control of Cyclic Distillation Systems. Procedia Eng . 2012;42:1202 - 1213. 27) Maleta B, MaletaO. Mass Exchange Contact Device. US Patent 8,158,073, 2012. 28) Maleta BV, Shevchenko A, Bedryk O, Kiss AA. Pilot - Scale Studies of Process Intensification by Cyclic Distillation . 29) Kiss AA. Distillation technology still young and full of breakthrough opportunities. J Chem Technol Biotechnol. 2014; 89:479 - 498. 146 30) Gerven TV, Stankiewicz A. Structure, Energy, Synergy, Time The Fundamen tals of P rocess Intensification. Ind.Eng. Chem. Res. 2009;48:2465 - 2474. 31) White DC. Optimize Energy use in Distillation. Chem Eng. Prog . 2012;35 - 41. 32) Bildea CS, Patrut C, Jorgensen SB, Abildskov J, Kiss AA. Cyclic distillation technology a mini - review. J Chem Tech nol Biotechnol. 2016;91:1215 - 1223. 33) Kiss AA. Cyclic Distillation. Advanced Distillation Technologies: Design, Control and Applications . Chichester, West Sussex: Wiley, a John Wiley & Sons. 2013:311 - 351. 34) Waterbury M. 1,000 Craft Distilleries - and Counting ! Now operating in the USA. Thewhiskeywash.com . 2016. 35) Distilled Spirits Council of the United States. Economic Contributions of Distilled Spirits Industry. Discus.org/economics. 2016. 36) Boes RE, Berglund KA. Real Time Spirit Distillation Analy sis using Gas Chromatography with a Flow Vial Sampler. 37) Toftegard B, Jorgensen SB. An Integration Method for Dynamic Simulation of Cycled Processes. Computers chem. Eng. 1989;13:927 - 930. 38) Sorensen E, Prenzler M. A Cyclic Operating Policy For Batch Distill ation Theory and Practice. Computers chem. Eng. 1997;21:S1215 - S1220. 39) S orensen E. A Cyclic Operating Policy for Batch Distillation Theory and Practice. Computers chem. Eng. 1999.;23:53 3 - 542. 40) DistiledSpirits.org. February 1, 2018. https://www.distilledspirits.org/news/distilled - spirits - council - reports - record - spirits - sales/ 41) Hadrich B, Kechaou N. Identif ication of Best Model for Equilibrium Data of Ethanol - Water Mixture . Journal of Chemistry and Chemical Engineering. 2010;4:46 - 48. 42) Seader J ,Henley E, RoperD. Separation Process Principals: Chemical and Biochemical Operations.2011;3:481 - 483 . 43) Claus M, Berg lund K. Fruit Brandy Production By Batch Column Distillation With R eflux. Journal of Food Processing Engineering. 2004;28:53 - 67. 44) National Institute of Standards and Technology. NIST Chemistry WebBook. 2018. https://webbook.nist.gov/chemistry/