KINGDOM OF SAUDI ARABIA KING SAUD UNIVERSITY COLLEGE OF ENGINEERING DEPARTMENT OF CHEMICAL ENGINEERING
MODELING AND SIMULATION OF A REACTIVE DISTILLATION UNIT FOR PRODUCTION OF MTBE ﻣﻴﺜﻞ ﺛﺎﻟﺜﻲ ﺑﻴﻮﺗﻴﻞ اﻹﻳﺜﺮ ﻧﻤﺬﺟﺔ وﻣﺤﺎآﺎة وﺣﺪة ﺗﻘﻄﻴﺮ ﻣﺘﻔﺎﻋﻞ ﻹﻧﺘﺎج ﻣﺎدة SUBMITTED BY
FAHAD S. AL-HARTHI 424121504 SUPERVISED BY
PROF. DR. AHMED E. ABASAEED PROF. DR. IBRAHIM S. AL-MUTAZ
A Thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Chemical Engineering 2nd Semester 1428-1429 AH
JUNE 2008
KINGDOM OF SAUDI ARABIA KING SAUD UNIVERSITY COLLEGE OF ENGINEERING DEPARTMENT OF CHEMICAL ENGINEERING
MODELING AND SIMULATION OF A REACTIVE DISTILLATION UNIT FOR PRODUCTION OF MTBE Submitted by: Eng. Fahad Al-Harthi 424121504 Supervised by: Prof. Dr. Ahmed E. Abasaeed Prof. Dr. Ibrahim S. Al-Mutaz
Examination Committee Members Prof. Ahmed E. Abasaeed Prof. Ibrahim S. Al-Mutaz Prof. Kamil M. Wagialla Prof. Mohammed Asif Prof. Khalid I. Al-Humaizi A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Chemical Engineering 2nd Semester 1428-1429 AH
JUNE 2008
ﺍﳌﻤﻠﻜﺔ ﺍﻟﻌﺮﺑﻴﺔ ﺍﻟﺴﻌﻮﺩﻳﺔ ﺟﺎﻣﻌﺔ ﺍﳌﻠﻚ ﺳﻌﻮﺩ ﻛﻠﻴﺔ ﺍﳍﻨﺪﺳﺔ – ﻗﺴﻢ ﺍﳍﻨﺪﺳﺔ ﺍﻟﻜﻴﻤﻴﺎﺋﻴﺔ
ﻧﻤﺬﺟﺔ وﻣﺤﺎآﺎة وﺣﺪة ﺗﻘﻄﻴﺮ ﻣﺘﻔﺎﻋﻞ ﻹﻧﺘﺎج ﻣﺎدة ﻣﻴﺜﻞ ﺛﺎﻟﺜﻲ ﺑﻴﻮﺗﻴﻞ اﻹﻳﺜﺮ )(MTBE إﻋﺪاد : اﻟﻤﻬﻨﺪس /ﻓﻬﺪ ﺑﻦ ﺳﻔﺮ اﻟﺤﺎرﺛﻲ
إﺷﺮاف آﻞ ﻣﻦ : اﻷﺳﺘﺎذ اﻟﺪآﺘﻮر :أﺣﻤﺪ أﺑﺎ ﺳﻌﻴﺪ اﻟﺤﺎج و اﻷﺳﺘﺎذ اﻟﺪآﺘﻮر :إﺑﺮاهﻴﻢ ﺑﻦ ﺻﺎﻟﺢ اﻟﻤﻌﺘﺎز
ﺭﺴﺎﻟﺔ ﻤﻘﺩﻤﺔ ﻻﺴﺘﻜﻤﺎل ﻤﺘﻁﻠﺒﺎﺕ ﺩﺭﺠﺔ ﺍﻟﻤﺎﺠﺴﺘﻴﺭ ﻓﻲ ﺍﻟﻬﻨﺩﺴﺔ ﺍﻟﻜﻴﻤﻴﺎﺌﻴﺔ
ﺟﻤﺎدى اﻟﺜﺎﻧﻴﺔ ١٤٢٩هـ June 2008
TABLE OF CONTENTS Table of contents …………………………………………………………………………………….. i List of Figures…………………………………………………………………………………………. iii List of Tables……………………………………………………………………………………….…. v Nomenclature ………………………………………………………………………………………… vi Dedication …………………………………………………………………………………………… viii Acknowledgments……………………………………………………………………………………. ix Abstract …….…………………………………………………………………………………………. xi CHAPTER 1 : INTRODUCTION ………………………………………………………………….. 1 1.1
Properties of MTBE……………………………………………………………………… 2
1.2
Chemistry of MTBE……………………………………………………………………… 3
1.3
Demand of MTBE……………………………………………………………………….. 3
1.4
Commercial Technology of MTBE…………………………………………………….. 5 (a) lsobutylene from Steam Crackers…………………………………………………. 6 (b) Isobutylene from Refineries………………………………………………………… 6 (c) Isobutylene by dehydrogenation lsobutane……………………………………….. 7 (d) Isobutylene from TBA……………………………………………………………….. 7
1.5
Introduction to Conventional Process…………….………………………………..…. 8
1.6
Concept of Reactive Distillation Process…………………………………………….. 9
1.7
Production of MTBE by Reactive Distillation……….………………………………… 10
1.8
1.7.1
ABB LUMMUS Process...……………………………..………………. 11
1.7.2
UOP Process ……………………………………………..………….... 12
1.7.3
Huntsman TBA/PO Process……………………………..……………. 14
Thesis outline and Work Methodology……………………………………………….. 15 1.8.1
Research Outline…………………………………………………………………15
1.8.2
Research Methodology……………..………...…………………………………15
CHAPTER 2 : REACTIVE DISTILLATION 2.1
Introduction to Reactive Distillation……………………………………………………. 17
2.2
Basics and Advantages of Reactive Distillation…………………………………..…. 17 2.2.1
Importance of RD……………………………………………………………… 18
2.2.2
Constraints and Difficulties in RD implementation……………………….... 23
2.2.3
The complexity of RD…………………………………………………………. 24
2.2.4
Model Available for design of RD columns……………………………….... 25
i
CHAPTER 3 : LITERATURE SURVEY 3.1 Overall Review……………………………..………………………………….……….... 28 3.2 Computational Methods…………………………….….……………………….……... 30 CHAPTER 4 : MODELING & SIMULATION OF MTBE PROCESS 4.1
MTBE Production by PRO/II …………………………………………………………… 40 4.1.1 Introduction ……………………………………………………………………… 40 4.1.2 Process description …………………………………………………………….. 40 •
Catalyst used……………………………………………………………………43
•
MTBE Recovery section……………………………………………………….43
•
Methanol Recovery section……………………………………………………44
4.2
Modeling and Simulation of MTBE RD unit………….……………………………...... 45
4.3
Thermodynamics Data….………………………..…….…………………………........ 46
4.4
MTBE Reaction Kinetics……………..……………………………………………........ 48
4.5
Modeling Procedure………..………………………………………………………........ 51 4.5.1 Model Equations for a single Stage…..……..…………………………......... 53
4.6
Modeling and simulation framework………………………….………………..…........ 54
CHAPTER 5 : RESULTS AND DISCUSSION 5.1 Results of the simulation work…………….…………...……………..…….................. 57 5.2 Results discussion………………………….…………...……………..……................... 62 5.3 Cases studies (Optimization)…..…..….…………...……………..…………….…........ 62 CHAPTER 6 : CONCLUSIONS AND RECOMMENDATIONS 6.1
Conclusions…………………………………………………………………………..… 69
6.2
Recommendations …..…..…………….……………………………………………… 71
REFERENCES……………………………………..…………………………………………..….. 73 APPENDIXES APP (A) : Reaction model code. …………………………………………………………..…… 76 APP (B) : MTBE Process Simulation Procedure using PRO/II software. ………………….. 80 APP (C) : Process Input-File used in PRO/II software. ………………………………..…….. 98
ii
LIST OF FIGURES Figure No
Page
Figure 1.1 : Publications and Patents on Reactive Distillation (1973-2003)……………………….….... 1 Figure 1.2 : MTBE Global Demand Distribution……………………………………………………………. 5 Figure 1.3 : MTBE Production Processes……………...…………………………………….…………….. 6 Figure 1.4 : MTBE Production via conventional Processes …………………………………….............. 9 Figure 1.5 : CDTECH Process Flow Diagram……………………………………………….……............. 12 Figure 1.6 : ETHERMAX MTBE Process Flow Diagram…………………………………........…………. 13 Figure: 2.1 Simple Reactive Distillation sketch…………………………….........................……............ 18 Figure 2.2: Processing schemes for a reaction sequence A +B <==> C+ D …….………....…………. 19 Figure 2.3: Processing schemes for the esterification reaction MeOH + AcOH<==>MeOAc + H2O… 20 Figure 2.4: (a) Reactive distillation concept for synthesis of MTBE from the acid catalysed reaction between MeOH and iso-butene. The butene feed is a mixture of reactive iso- butene and nonreactive n-butene. (b) Reactive distillation concept for the hydration of ethylene oxide to ethylene glycol. (c) Reactive distillation concept for reaction between benzene and propene to form cumene. (d) Reactive distillation concept for reaction production of propylene oxide from propylene chlorohydrin and lime. The reactive sections are indicated by grid line...……………………………………………… 21 Figure 2.5: Hydrodesulphurisation of gas oil carried out in (a) co-current trickle bed reactor and (b) counter-current RD unit…………………………..…………………………………………...………………. 22 Figure 2.6: Transport processes in RD. (a) homogeneous liquid phase reaction, and (b) heterogeneous catalyzed reactions………………………………………………………...…..………..…………….……..… 25 Figure 2.7 : Length and time scales in RD……………..……………...………..……………………..……. 25 Figure 3.1: Ternary LLE diagram for Acetone-water-MIBK system at 5atm. …………………………..…33 Figure 3.2: RD Column Configuration……………..…..………..………………………………………...... 34 Figure 3.3: Liquid Composition profile in RD Column..………..……………………...……...…………….. 34 Figure 3.4 Nonreactive residue curve map for the system IBTE+ MeOH ↔ MTBE at 101.32 kPa…… 36 Figure 3.5 Reactive residue curve map for the system IBTE+ MeOH ↔ MTBE at 101.32 kPa………. 37 Figure 3.6 Reactive phase diagram for the system IBTE/ MeOH /MTBE at 101.32 kPa………………. 37 Figure 3.7 Reactive phase diagram for the system IBTE/ MeOH /MTBE at 810.56 kPa………………. 38 Figure 4.1: MTBE Plant Flowsheet.………………………………..……………...…..……......……………..41 Figure 4.2: Thermodynamic Data of PRO/II software…………..………………………..…...……………..47 Figure 4.3: A schematic representation of a catalytic distillation column………..………....……………..52
iii
Figure 4.4: Representation of a generic equilibrium stage………...……………..……….....……………..52 Figure 4.5: MTBE Process Flowsheet PRO/II………………………………..…..………….....…………….55 Figure 5.1: Feed and products concentration profile among column trays...……......……………57 Figure 5.2: Temp Profile of column T-1………………………………..…………..…....…......…………….58 Figure 5.3: Rates Profile of column T-1……………………………………………..………………………. 58 Figure 5.4: Density Profile of column T-1…………………………………………..………….…………….59 Figure 5.5: Viscosity Profile of column T-1………………………………………..………..………………..59 Figure 5.6: Temp Profile of column T-2…………………………………………..………..…..…………….59 Figure 5.7: Rates Profile of column T-2……………………………………..……………………………….59 Figure 5.8: Density Profile of column T-2…………………………………..………………….……………..60 Figure 5.9: Viscosity Profile of column T-2………………………………..………………..………………. 60 Figure 5.10: Temp Profile of column T-3…………………………………..……………..………………….60 Figure 5.11: Rates Profile of column T-3…………………………………..……………….………………..60 Figure 5.12: Density Profile of column T-3………………………………..…………………………..……..61 Figure 5.13: Viscosity Profile of column T-3………………………………..…………...…..……………….61 Figure 5.14: Results of changing the MeOH/IBTE ratio……………..……...…..…………………………63 Figure 5.15: Results of changing the number of trays of RD column………...…..………………………64 Figure 5.16: Results of changing the location of feed stream into RD column …………………………65 Figure 5.17: Results of changing the Reflux Ratio…………..…………...…..…………………………….66 Figure 5.18 Results of changing the location of reaction zone .…………...…..…………………………..67
iv
LIST OF TABLES Table
page
Table 1.1: Typical Properties of Oxygenates……………………………………………………............... 2 Table 1.2 : MTBE competitive strengths and weaknesses……………………………………................ 3 Table 1.3 : Global MTBE demand (thousand tons per year) …………………………………................ 5 Table 1.4: Typical isomer distribution in C4 product stream (volume %)……………………................. 7 Table 4.1: Reactor Feed Data………..……………….…………………………………………................. 42 Table 4.2: Reaction Stoichiometry………………………………………………………..…….................. 43 Table 4.3: Approximate Catalyst Properties of Amberlyst 15……………………….….……................ 43 Table 4.4: Binary Interaction Parameters for SRKM_VLE…………………………………..…............... 46 Table 5.1: Summary Table…………………………………………………………………..……............... 57 Table 5.2: Summary of MTBE column T-1…………………………………….………….….…................. 58 Table 5.3: Summary of water wash column T-2………………………………………….….…................. 59 Table 5.4: Summary of Methanol Recovery column T-3…………………………..….……….................. 60 Table 5.5: Overall Report…………………………………..…………………………………..….................. 61
v
NOMENCLATURE Notations a,b,c
Order of reaction of species A,B and C respectively
∆HA
Heat of adsorption of methanol, J/mol
∆HC
Heat of adsorption of MTBE, J/mol
aij
The mixing rule
Xi
Mole fraction of component i
Xj
Mole fraction of component j
Kij
Binary interaction parameter
Ks
Surface reaction rate constant, (gmole/g catalyst)
KA
Equilibrium adsorption constant for A, g cat/gmole
KC
Equilibrium adsorption constant for C, g cat/gmole
K
Equilibrium constant for the overall reaction
KAO , KCO, KSO
Preexponential factor
rs
Rate of surface reaction, (mol/gcat)/h
CA
IBTE concentration, mole/l
CB
MEOH concentration, mole/l
CC
MTBE concentration, mole/l
R
Gas constant, 8.314 J/mol.K
T
Temperature, K
E1
Activation Energy of the forward reaction, J/mol
KX
Mole fraction Equilibrium ratio
Kγ
Ratio of activity coeffiecients at equilibrium
Xk,i
The mole fraction of liquid phase for species j in stage k.
Yk,i
The mole fraction of vapor phase for species i leaving stage k
Lk
The molar liquid flow rate leaving stage k.
Vk
The molar vapor flow rate leaving stage k.
Fk
The feed flow rate to stage k.
hk
The molar enthalpy of liquid phase in stage k.
HK
The molar enthalpy of vapor phase in stage k.
Hjk
The heat of reaction in stage k.
Qk
The heat duty in stage k.
γk,i
The activity coefficient of component i in stage k.
Kk,i
The vapor-liquid equilibrium constant.
ρk
The molar density of the liquid on stage k.
vk
The volumetric hold-up of liquid on stage k
P
Pressure, bar
NC
Number of components
NT
Number of trays
vi
u,w Subscripts: i j Superscripts: F
Constants, typically integers
Tray index Component index refers to a feed
D
refers to a draw
L
refers to a liquid property
V
refers to a vapor property
vii
DEDICATION
To my mother, who will be always my inspiration. This work is also dedicated to the ones that encourage me, my wife , my daughter Lujain, and my brothers.
viii
ACKNOWLEDGMENT Firstly, I thank Allah almighty for His blessings and providing me with the strength and power to succeed. This research compiles the upmost results of a 2-year period of intense and challenging work. Independently on the nature of their contribution (intellectual or/and emotional), many people have pushed toward the completion of this work. To all of them I am deeply and sincerely indebted. My first words of thanks go, sincerely, to my promoter, Prof. Ibrahim Al-Mutaz. I thank him for his continuous guidance and scientific insight during these 2+ years and, on top of this, for his admirable kindness and politeness. I fully enjoyed the long hours we spent together, discussing a wide range of topics. Also. I would like to thank my advisor Prof. Ahmed Abasaeed for his support and I am very much thankful for allocating time in your tight agenda to discuss and repeatedly correct all the thesis chapters. Really, Without all your comments and suggestions this thesis probably would have remained as a bunch of unconnected manuscripts. Additionally, I would like to thank all the people who supported my research and me during the past two years. Foremost have been my wife who has always encouraged me to work hard and to continue this work. Moreover, my thanks also extended to my family members, relatives and colleagues for supporting me throughout my research period.
ix
א ﻣﻘﺪﻣﺔ: ﺗﻌﺘﺒﺮ ﻣﺎدة )ﻣﺜﻴﻞ ﺛﺎﻟﺜﻲ ﺑﻴﻮﺗﻴﻞ اﻹﻳﺜﺮ( اﻟﻤﻌﺮوﻓﺔ ﻋﺎﻟﻤﻴًﺎ ﺑﺎﻻﺳﻢ اﻟﻤﺨﺘﺼﺮ ) (MTBEﻣﻦ أﺑﺮز اﻟﻤﻮاد اﻷآﺴﺠﻴﻨﻴﺔ اﻟﺘﻲ ﺗﺘﻤﺘﻊ ﺑﺨﺼﺎﺋﺺ ﻓﻴﺰﻳﺎﺋﻴﺔ ﻣﻤـﺘﺎزة ﻋﻨﺪ إﺿﺎﻓﺘﻬﺎ إﻟﻰ اﻟﻮﻗـﻮد ﻟﺰﻳﺎدة اﻟﻌﺪد اﻷوآﺘﻴﻨﻲ واﻟﺤﺪ ﻣﻦ اﻟﺘﻠﻮث اﻟﺒﻴﺌﻲ )ﻣﻘﺎرﻧﺔ ﺑﺎﻟﺮﺻﺎص( ودورهﺎ آﺒﻴﺮ ﺟﺪًا ﻓﻲ اﻟﺤﺪ ﻣﻦ اﻻﻧﺒﻌﺎﺛﺎت اﻟﻀﺎرة اﻟﻤﻠﻮﺛﺔ ﻟﻠﺒﻴﺌﺔ ﻣﻦ ﻏﺎزات ﻋﺎدم اﻟﺴﻴﺎرات وﺗﺆدي ﻓﻲ ﻧﻔﺲ اﻟﻮﻗﺖ إﻟﻰ اﺣﺘﺮاق اﻟﻮﻗﻮد ﺑﺸﻜﻞ أﻓﻀﻞ وأوﻓﺮ ﻣﻦ ﻣﺎدة اﻟﺮﺻﺎص وإﻋﻄﺎء ﻃﺎﻗﺔ أﻋﻠﻰ .وﺗﻌﺪ أﻓﻀﻞ ﻃﺮﻳﻘﺔ ﻹﻧﺘﺎﺟﻪ ﺑﺎﺳﺘﺨﺪام ﺗﻘﻨﻴﺔ ﺗﺤﻮﻳﻞ ﻟﻘﻴﻢ اﻟﺒﻴﻮﺗﺎن واﻟﻤﻴﺜﺎﻧﻮل إﻟﻰ ﻣﺎدة MTBEﺑﺎﺳﺘﺨﺪام ﺛﻼث ﻋﻤﻠﻴﺎت وهﻲ ﺗﻐﻴﻴﺮ اﻟﺒﻨﻴﺔ اﻟﺠﺰﺋﻴﺔ ﻟﻠﺒﻴﻮﺗﺎن وإزاﻟﺔ اﻟﻬﻴﺪروﺟﻴﻦ ﻣﻦ اﻻﻳﺰوﺑﻴﻮﺗﺎن وﺗﻔﺎﻋﻞ اﻻﻳﺰوﺑﻴﻮﺗﺎن ﻣﻊ اﻟﻤﻴﺜﺎﻧﻮل. وﻗﺪ ﺗﻨﺎﻣﺖ اﻟﻄﺎﻗﻪ اﻹﻧﺘﺎﺟﻴﺔ ﻟﻤﺎدة MTBEﻋﺒﺮ اﻟﺴﻨﻮات اﻷﺧﻴﺮة ﻟﻠﺘﻮاﻓﻖ ﻣﻊ ﺗﺼﺎﻋﺪ اﺳﺘﺨﺪاﻣﻬﺎ ﻋﺎﻟﻤﻴ ًﺎ اﺳﺘﺠﺎﺑﺔ ﻟﻸﻧﻈﻤﺔ اﻟﻤﻨﺎدﻳﺔ ﺑﺎﻟﻤﺤﺎﻓﻈﺔ ﻋﻠﻰ ﺳﻼﻣﺔ وﺻﺤﺔ اﻟﺒﻴﺌﺔ ،وﻗﺪ أﺛﺒﺘﺖ اﻟﻤﺎدة ﻧﺠﺎﺣﻬﺎ ﻓﻲ هﺬا اﻟﻤﺠﺎل ،ﺣﻴﺚ ﻳﺠﺮي اﺳﺘﺨﺪاﻣﻬﺎ ﻓﻲ أوروﺑﺎ واﻟﻌﺪﻳﺪ ﻣﻦ اﻷﺳﻮاق اﻟﻌﺎﻟﻤﻴﺔ اﻷﺧﺮى ﺑﺼﻮر ٍة ﻃﺒﻴﻌﻴﺔ وﺁﻣﻨﺔ. ﻳﺘﻢ إﻧﺘﺎج ﻣﺎدة MTBEﻓﻲ اﻟﻄﺮﻳﻘﺔ اﻟﺘﻘﻠﻴﺪﻳﺔ ﻋﺒﺮ إﺿﺎﻓﺔ ﻣﺎدة اﻟﻤﻴﺜﺎﻧﻮل إﻟﻰ اﻻﻳﺰو ﺑﻴﻮﺗﻴﻦ آﻤﻮاد ﻣﺘﻔﺎﻋﻠﺔ ﻓﻲ ﻣﻔﺎﻋﻞ ﺑﻮﺟﻮد ﻣﺎدة ﺣﻔﺎزة ﻋﻨﺪ درﺟﺔ ﺣﺮارة ﺗﺘﺮاوح ﺑﻴﻦ 60-45درﺟﺔ ﻣﺌﻮﻳﺔ .ﺛﻢ ﻳﺘﻢ إﻣـﺮار ﻧﺎﺗﺞ اﻟﺘﻔﺎﻋـﻞ إﻟﻰ ﻋـﺪد ﻣﻦ أﺑﺮاج اﻟﺘﻘﻄﻴـﺮ ﻟﻴﺘﻢ ﻓﺼﻞ اﻟﻤﺎدة اﻟﻤﻨﺘﺠـﺔ ) (MTBEواﺳﺘﻌﺎدة اﻟﻤﻮاد اﻟﺘﻲ ﻟﻢ ﺗﺘﻔﺎﻋﻞ ﻟﻴﺘﻢ اﺳﺘﺨﺪاﻣﻬﺎ ﻣﺮﻩ أﺧﺮى. هﻨﺎك ﻃﺮﻳﻘﺔ أﺧﺮى ﻹﻧﺘﺎج اﻟـ MTBEوهﻲ اﻟﻤﺘﺒﻌﺔ ﺣﺪﻳﺜًﺎ وﻳﻌﻮد ﺗﺎرﻳﺦ هﺬﻩ اﻟﻄﺮﻳﻘﺔ إﻟﻰ ﻋﺎم 1980م ﺣﻴﺚ ﺳﺠﻞ اﻟﻌﺎﻟﻢ ﺳﻤﻴﺚ أول ﺑﺮاءة اﺧﺘﺮاع ﻹﻧﺘﺎج اﻟـ MTBEﻋﺒﺮ هﺬﻩ اﻟﻄﺮﻳﻘﺔ وﺗﺘﻠﺨﺺ ﻓﻲ إﻧﺘﺎج وﻓﺼﻞ اﻟـ MTBEﺁﻧﻴﺎ ً ،وﺗﺴﻤﻰ هﺬﻩ اﻟﻄﺮﻳﻘﺔ ﺑـﻌﻤﻠﻴﺔ اﻟﺘﻘﻄﻴﺮ اﻟﻤﺘﻔﺎﻋﻞ ، Reactive Distillation Processوﺗﻤﺘﺎز هﺬﻩ اﻟﻄﺮﻳﻘﺔ ﺑﻌﺪة ﻧﻮاﺣﻲ ﺗﻘﻨﻴﺔ واﻗﺘﺼﺎدﻳﺔ ﻣﺜﻞ :ﺧﻔﺾ ﺗﻜﺎﻟﻴﻒ إﻧﺸﺎء اﻟﻤﺼﻨﻊ وذﻟﻚ ﺑﺘﻘﻠﻴﻞ ﻋﺪد اﻟﻤﻌﺪات واﻷﺟﻬﺰة، ﺗﻘﻠﻴﻞ ﻧﺴﺒﺔ إﻧﺘﺎج اﻟﻤﻮاد اﻟﺠﺎﻧﺒﻴﺔ ﻏﻴﺮ اﻟﻤﺮﻏﻮﺑﺔ ،زﻳﺎدة آﻔﺎءﻩ اﻟﺘﻔﺎﻋﻞ واﻟﻔﺼﻞ ﻣﻤﺎ ﻳﺆدي إﻟﻰ زﻳﺎدة ﻣﻌﺪل اﻹﻧﺘﺎج واﻻﺳﺘﻔﺎدة ﻣﻦ اﻟﺤﺮارة اﻟﻨﺎﺗﺠﺔ ﻣﻦ اﻟﺘﻔﺎﻋﻞ ﻓﻲ ﺑﺮج اﻟﺘﻘﻄﻴﺮ ﺣﻴﺚ ﺗﺴﺎﻋﺪ ﻓﻲ ﻋﻤﻠﻴﺔ اﻟﻔﺼﻞ. ﺳﻮف ﻳﺘﻢ ﻓﻲ هﺬا اﻟﺒﺤﺚ ﻋﻤﻞ ﻣﺴﺢ ﻣﺮﺟﻌﻲ ﻣﻜﺜﻒ ﻟﻸﺑﺤﺎث اﻟﺴﺎﺑﻘﺔ ﻓﻲ هﺬا اﻟﻤﺠﺎل وذﻟﻚ ﺑﻬﺪف دراﺳﺔ هﺬﻩ اﻟﻄﺮﻳﻘﺔ وﻣﻦ ﺛﻢ ﺑﻨﺎء ﻧﻤﻮذج رﻳﺎﺿﻲ ﻳﺤﺎآﻲ إﻧﺘﺎج ﻣﺎدة MTBEﺑﻄﺮﻳﻘﺔ اﻟﺘﻘﻄﻴﺮ اﻟﻤﺘﻔﺎﻋﻞ وﺳﻴﺴﺘﺨﺪم هﺬا اﻟﻨﻤﻮذج ﻓﻲ دراﺳﺔ ﺗﺄﺛﻴﺮ اﻟﻌﻮاﻣﻞ اﻟﺘﺸﻐﻴﻠﻴﺔ واﻟﺘﺼﻤﻴﻤﻴﺔ اﻟﻤﺨﺘﻠﻔﺔ. ﻧﺘﺎﺋﺞ ﻋﻤﻠﻴﺔ اﻟﻤﺤﺎآﺎة اﻟﺘﻲ ﺗﻢ اﻟﺤﺼﻮل ﻋﻠﻴﻬﺎ ﺑﺎﺳﺘﺨﺪام ﺑﺮﻧﺎﻣﺞ PRO/IIأوﺿﺤﺖ ﺑﺎن اﺳﺘﺨﺪام ﻋﻤﻠﻴﺔ اﻟﺘﻘﻄﻴﺮ اﻟﻤﺘﻔﺎﻋﻞ ﻳﺰﻳﺪ ﻣﻦ ﻧﺴﺒﺔ ﺗﺤﻮﻳﻞ اﻻﻳﺰو ﺑﻴﻮﺗﻴﻦ اﻟﻰ MTBEﺑﻨﺴﺒﺔ ﻋﺎﻟﻴﺔ ﺟﺪا ً ﺗﺼﻞ اﻟﻰ %99.2وﺑﻨﻘﺎوة ﻋﺎﻟﻴﺔ اﻳﻀﺎ ً ﺗﺼﻞ اﻟﻰ .%99.7
x
ABSTRACT Methyl tertiary butyl ether (MTBE) is primarily used in gasoline blending as an octane enhancer to improve hydrocarbon combustion efficiency. Of all the oxygenates, MTBE is attractive for a variety of technical reasons. It has a low vapor pressure. It can be blended with other fuels without phase separation. It has the desirable octane characteristics. MTBE is produced via direct addition of methanol to isobutylene using sulphonated ion exchange resin as catalysts. There are two ways to produce MTBE, one is the conventional process which is mainly a reactor and separate distillation column with conversion range 8792%. Another method for the production of MTBE is newly established and date back to the way in 1980 as the scientist Smith recorded the first patent for the production of MTBE through this method, this method called Reactive Distillation Process, and there are a lot of features that makes this process attractive and practical with a conversion reached 99.2%. The main objectives from this research are developing a mathematical model for the MTBE production via reactive distillation column at steady state, coding and embedding the model into simulation software (PRO/II) and using the developed model to assess the effect of some critical design and operating parameters on column performance and doing some case studies. PRO/II software has the capabilities of solving Reactive Distillation Processes utilizing Chemdist Algorithm provided by the software, but it requires knowledge of Material and Energy balances equations as well as Thermodynamics and kinetics of reaction. In this work, we have made an extensive literature survey for the previous research done in this area to study this technology and then we have developed a mathematical model that simulates the production of MTBE in a reactive distillation unit and that model was used to study the impact of operational factors and different design parameters by doing several case studies "Optimization". The main conclusion of this work is that the higher conversion of IBTE to MTBE can be obtained by applying Reactive Distillation approach, we have obtained 99.2 % IBTE conversion and high selectivity for MTBE with 99.7%. Moreover, the optimum Methanol/Isobutylene ratio was found equal to 1.0 and the optimum number of trays for Reactive Distillation column is 30 trays.
xi
CHAPTER -1-
INTRODUCTION
CHAPTER -1-
INTRODUCTION Methyl tertiary butyl ether (MTBE) is primarily used in gasoline blending as an octane enhancer to improve hydrocarbon combustion efficiency. Of all the oxygenates, MTBE is attractive for a variety of technical reasons. It has a low vapor pressure. It can be blended with other fuels without phase separation. It has the desirable octane characteristics and is becoming increasingly important as stricter air pollution control measures are implemented. MTBE can be produced by addition of methanol to isobutylene in the liquid phase over an acidic catalyst consisting of sulfonated macroporous ion exchange resins. In 1980, Smith registered a patent to process MTBE for Chemical Research and Licensing Company. He used a reactive distillation system, containing catalytic packing. The pilot plant was 3 inches in diameter and it was used to predict the operation variables of a large commercial plant [1]. Due to its potential for improved process design and the success of its commercial applications, reactive distillation gained the interest of both academics and industry toward its use and application in commercial processes. Figures 1.1 show the growing interest in reactive distillation in recent years. These data were compiled from the ACS databases CAPLUS, CHEMCATS and CHEMLIST. Some 400 publications were taken into consideration for the period of 1973 to May 2003 [2].
Figure 1.1 Publications and Patents on RD -Historical Trends (1973-2003).
-1-
1.1
PROPERTIES OF MTBE Oxygenates are hydrocarbons that contain one or more oxygen atoms. The primary oxygenates are alcohols and ethers, including: fuel ethanol, methyl tertiary butyl ether (MTBE), ethyl tertiary butyl ether (ETBE), and tertiary amyl methyl ether (TAME). The physical properties for these components are shown in Table 1.1. The 1977 Clean Air Act amendments set requirements for "substantially similar gasoline," which requires that oxygenates be approved by the U.S. EPA before they are allowed to be used in gasoline. In 1981 the EPA allowed the blending of MTBE up to 11 volume %, and extended the limit to 15 volume % in 1988. Oxygenates are added to motor vehicle fuels to make them burn more cleanly, thereby reducing toxic tailpipe pollution, particularly carbon monoxide. Oxygenates are favored not only for their vehicle emission benefits but also their blending properties in motor gasoline (e.g., octane). Table 1.1: Typical Properties of Oxygenates.
Chemical formula Oxygen content, % by weight Octane, (R+M)/2 Blending vapor pressure, RVP
Ethanol
MTBE
ETBE
TAME
CH3CH2OH
CH3OC(CH3)3
CH3CH2OC(CH3)3
(CH)3CCH2OCH3
34.73
18.15
15.66
15.66
115
110
111
105
18
8
4
1.5
Source: National Petroleum Council, U.S. Petroleum Refining: Meeting Requirements for Cleaner Fuels and Refineries (Washington, DC, August 1993) Appendix L.
There is an increasing demand for MTBE as a gasoline additive. Currently, the worldwide consumption of MTBE reached 6.6 billion gallons of which 65% is consumed in the United States [3]. Methyl tertiary-butyl ether is an excellent gasoline blending component because it is a high octane enhancer and has low volatility. It has blending properties similar to gasoline’s, but do not exhibit undesirable properties such as azeoptrope formation, water pick-up, or phase separation. Gasoline
-2-
containing MTBE is accepted for transport on common carrier pipelines in the United States and by major and independent refiners. Table 1.2 summarizes the relative strengths and weaknesses of MTBE compared to other octane enhancers. MTBE is a widely accepted octane enhancer and enjoys several significant strengths compared to competing octane enhancers. Its major weakness is that one of its key feedstocks, isobutylene, is limited by petrochemical and refinery operations. Table 1.2 : MTBE competitive strengths and weaknesses
Strengths -
Weaknesses
High octane Low volatility Blending characteristics similar to gasoline Widely accepted in marketplace by consumers and refiners Reduces carbon monoxide and exhaust hydrocarbon emissions
-
Availability of economical isobutylene feedstocks is limited Possible methanol supply constraints Health hazard
Although MTBE has a density less than that of water, when dissolved in water, MTBE responds to localized groundwater gradients and aquifer recharge. A growing number of studies have detected MTBE in ground water in United State; in some instances these contaminated waters are sources of drinking water. Low levels of MTBE can make drinking water supplies undrinkable due to its offensive taste and odor. Therefore, the jurisdiction of the U.S. Department of Transportation (DOT) has set some restricted Regulation to minimize releases from gasoline pipelines or another potential source of leaks. MTBE is believed to have entered the water supply from leaking underground storage tank systems including underground lines that contained gasoline and/or from surface spills at gas stations. Once in the ground, MTBE behaves differently from other gasoline constituents such as benzene . Unlike petroleum hydrocarbons, it is highly water soluble, not easily absorbed into soil, and is more resistant to biodegradation. Thus, with widespread use, MTBE has the potential to occur in high concentrations in groundwater, travel far from leak sources, and accumulate to become a concern for the entire region.
-3-
1.2
CHEMISTRY OF MTBE
Methyl tertiary-butyl ether (MTBE) can be formed by the addition of methyl alcohol to the highly reactive double bond in isobutylene, as shown in the following equation:
CH 3 − OH
+
(CH 3 )2 C = CH 2
Methanol
⇔
Isobutylene
(CH 3 )3 C − O − CH 3 MTBE
MTBE synthesis occurs in the liquid phase at 40°C-100°C and 100-150 psig as an exothermic reaction (∆HR= -16,060 Btu/lb-mole). In the presence of a small amount of acidic cation exchange resin catalyst the reaction proceeds quantitatively. Indeed, there are few reactions in industrial chemistry that demonstrate such high selectivity. Several processes, as will be discussed later, have been devised for carrying out this reaction efficiently on the isobutylene contained in the so-called raffinate-I that results when butadiene is removed from the C4 fraction obtained either from steam or catalytic cracking. The C4 fraction from the former source is preferred because of its higher isobutylene concentration. This fraction contains, in addition to isobutylene, butene-I, and butane-2. The removal of isobutylene from raffinate-I provides a mixture of the normal butenes known as raffinate-Il [2].
1.3
DEMAND OF MTBE
The importance of MTBE is increasing from year to year since the global demand for MTBE is expected to grow at 4.0% annually from 1994 to 2010, as shown in Table I.3. In the 2000-2010 period, however, growth will slow to 1.7 % annually from 8.1 % from 1994-2000 as shown in Fig 1.3 [4].
-4-
Table 1.3 : Global MTBE demand (thousand tons per year). Adapted from CHEM SYSTEM.
Figure 1.2 : MTBE Global Demand Distribution
1-4
COMMERCIAL TECHNOLOGIES OF MTBE Methyl tertiary-butyl ether (MTBE) is produced through the reaction of isobutylene with methanol. The MTBE reaction is equilibrium limited. Higher temperatures increase the reaction rate, but the conversion level is lower. Lower temperatures shift the equilibrium toward ether production, but more catalyst inventory is required. Therefore, conventional MTBE units are designed with two reactors in series. Most of the etherification reaction is achieved at an elevated temperature in the first reactor and then finished at a thermodynamically favorable lower temperature in the second reactor [2]. Figure 1.3 shows the different routes to obtain IC4 for MTBE production, which vary largely due to the different isobutylene feedstock sources. As shown in the figure, isobutylene can be produced as a: •
Byproduct of ethylene manufacture in steam cracker units -5-
•
Byproduct of fluid catalytic cracking operations.
•
Main product of an isobutane dehydrogenation unit
•
Coproduct (via tertiary-butyl alcohol-TBA) of propylene oxide manufacture, as operated by ARCO Chemical and Texaco Chemical Company
Figure 1.3 : MTBE Production Processes
These various supply sources for isobutylene are briefly reviewed below. (a) lsobutylene from Steam Crackers The mixed butylene stream produced from ethylene plants has a relatively high isobutylene content. Depending on the feedstock, operating conditions, plant configuration, and other factors, the isobutylene content can vary from 35 to 50 vol%, with about 44 % considered typical. One benefit of MTBE manufacture from steam cracker sourced mixed butylenes is that it provides an easy way to separate butylene isomers [2].
(b) Isobutylene from Refineries Isobutylene is also produced as a byproduct of fluid catalytic cracking (FCC) operations in refineries. The isobutylene concentration is much lower in this mixed butylene stream compared to the steam cracker mixed C4s stream. Table 1.4 shows the relative butane/butylene isomer distribution for typical FCC and steam cracker C4 streams [2]. The higher isobutylene concentration in steam cracker mixed butylene streams reduces the capital and operating costs compared to feeding FCC mixed butylenes. -6-
Table 1.4: Typical isomer distribution in C4 product stream (volume %)
(c) Isobutylene by dehydrogenation lsobutane One possibility in alleviating the potential problem of insufficient isobutylene is to dehydrogenate isobutane. lsobutane may, in turn, be made by the isomerization of n-butane, a reaction well established, for isobutane is prepared in large volumes for alkylation reactions in the refinery. Another approach involves the skeletal isomerization of butene-1. These reactions are shown in the following equations:
(d) Isobutylene from TBA High capital costs are involved in the isomerization and dehydrogenation reactions. tertiary-Butyl alcohol (TBA), made by ARCO and Texaco as a coproduct of propylene oxide production, is another source of isobutylene. This involves dehydration to isobutylene, as shown in the following equation [2].
-7-
A number of companies have technology for MTBE manufacturing: all the routes are based on liquid phase etherification of isobutylene with methanol. Selected technology holders and licensers are: ARCO, CDTECH (ABB Lummus Crest and Chemical Research & Licensing), Snamprogetti, Phillips Petroleum, HuelslUOP, Sumitomo Chemical, Erdoelchemie, IFP. Description of some commercial conventional Processes are given below:
1.5
A TYPICAL CONVENTIONAL PROCESS As mentioned in the previous section, MTBE is produced via direct addition of methanol to isobutylene using sulphonated ion exchange resin as catalysts. The technology features a two-stage reactor system of which the first reactor is operated in a recycle mode. With this method, a slight expansion of the catalyst bed is achieved which ensures very uniform concentration profiles within the reactor and, most importantly, avoids hot spot formation. Undesired side reactions, such as the formation of dimethyl ether (DME), are minimized. The reactor inlet temperature ranges from 45°C at start-of-run to about 60°C at end-of-run conditions. The catalyst used in this process is a cation-exchange resin. Isobutylene conversions of 97% are typical for FCC feedstocks. Higher conversions are attainable when processing steam-cracker C4 cuts that contain isobutylene concentrations of 25% as illustrated in figure 1.4 [2]. MTBE is recovered as the bottoms product of a distillation unit. The methanol-rich C4 distillate is sent to the methanol-recovery section. Water is used to extract excess methanol and recycle it back to process. The isobutylenedepleted C4 stream may be sent to a raffinate stripper or to a molsieve-based unit to remove other oxygenates such as DME, MTBE, methanol and tertbutanol.
-8-
Figure 1.4 : MTBE Production via Conventional Processes
An alternative to the conventional process is Reactive Distillation process.
1.6
CONCEPT OF REACTIVE DISTILLATION PROCESS Reactive distillation is a process where simultaneous chemical reaction and vapor-liquid phase separation take place in the presence of a heterogeneous catalyst. This represents an exciting alternative to traditional liquid phase chemical reaction processing. In reactive distillation separation of product(s) from unconverted reactants allows for greater conversion, because product removal displaces equilibrium and forces the reaction to completion [4]. Both reaction and distillation acting simultaneously offers certain advantages that cannot be matched by conventional processing. The advantages of reactive distillation are specific to each system. However, the followings are some of the significant advantages of reactive distillation:[4]
-9-
•
Effecting distillation and reaction simultaneously reduces the capital costs and includes benefits such as reduction of recycle, optimization of separation, lower requirements of pumps, instrumentation and piping.
•
An equilibrium reaction can be driven to completion by separation of the products from the reacting mixture.
•
Elimination of possible side reactions by removal of the products from the reaction zone. This can serve to increase selectivity.
•
Savings associated with energy costs, through use of the energy released by exothermic reactions for vaporization. This reduces the reboiler heat duty for boil-up that is supplied normally by steam.
•
Non-reactive azeotropes may disappear under reactive distillation conditions.
•
Improved materials use. With reactive distillation, material proper usage is attained through removal of the reaction product(s) from the reaction phase. Also, elimination of by-products formation may allow use of lesser quantities of reactants.
•
Reduction of hot spots, because the liquid vaporization provides a sink for thermal energy. Synthesis of chemicals through reactive distillation has been mainly
applied to processes such as esterifications, hydrolysis reactions, transesterifications, and etherifications. A detailed list of reactions where reactive distillation is advantageous is cited by Doherty and Malone in 2001 [5]. While the use of reactive distillation is well established in the chemical process industry, very little is known regarding the importance of mass transfer resistances in limiting catalyst effectiveness in MTBE synthesis [6].
1.7
PRODUCTION OF MTBE BY REACTIVE DISTILLATION UNIT Most of the new commercial MTBE plants utilize reactive distillation technology in which chemical reaction and fractionation of products are combined into a single unit operation.
- 10 -
In the production of MTBE in reactive distillation process, a pre-reactor and a reactive distillation column replaced the conventional process. The reactive column completes the exothermic reaction and simultaneously conducts the separation. It is also possible to incorporate the pre-reactor into the reactive column thereby lowering capital costs and improving energy integration [7]. The reaction between methanol and isobutylene to form MTBE is equilibrium limited and the feed to a conventional reactor requires a relatively low methanol/isobutylene feed ratio in the feed. Maintaining the methanol/isobutylene ratio below 1.05 allows recovery of the product MTBE as the bottoms product from a distillation tower with the overhead product being at the azeotropic composition of methanol and isobutylene. Feeding a higher methanol/isobutylene ratio results in recovery of the MTBE from the bottoms at the methanol/MTBE azeotropic composition [8,9]. Use of reactive distillation allows more economical recovery of the MTBE product. Typical Reactive Distillation processes are:
1.7.1 ABB LUMMUS Process The process of production of MTBE using ABB LUMMUS process consists of three basic steps as listed below [2]: •
Isomerization of n-butanes to isobutane.
•
Dehydrogenation of isobutane to isobutylene (CATOFIN Process).
•
Etherification of isobutylene with methanol to MTBE (CDTech Process).
MTBE is formed by the catalytic etherification of isobutylene with methanol, as shown in Figure 1.5 Isobutylene-rich feedstock from the dehydrogenation unit is combined with fresh methanol from offsite storage and recycle methanol from the methanol recovery of the MTBE unit and fed to the primary fixed bed reactor. The CDTech process is based on a two step reactor design, consisting of a fixed bed reaction followed by final conversion in a catalytic distillation column [2]. After reaction in the primary reactor, the partly reacted mixture is cooled to the bubble point and pumped to the catalytic distillation reaction column that - 11 -
combines reaction and fractionation sections. The reaction is continued in the column and the MTBE product is separated from the un-reacted C4s. The column allows a high conversion (99 %) of isobutylene. Methanol and C4s form a minimum boiling azeotrope, so that the methanol is carried into the catalyst packing where the reaction proceeds. As MTBE is formed, it is removed from the reaction zone by distillation, which allows the reaction to proceed well beyond the limit set by chemical equilibrium for a conventional reactor system. MTBE leaves the reaction column as the bottoms product together with a small amount of reaction byproducts. All byproducts formed in the MTBE reaction are gasoline compatible components. They include TBA (formed by the reaction of isobutylene and water) and diisobutylene (formed by the dimerization of isobutylene). Any residual methanol distills overhead with the unreacted C4s and is routed to the methanol extraction column. The recovered methanol (methanol column overhead) is recycled to the MTBE reactor feed. The methanol column bottoms product (water) is recycled to the methanol extraction column.
Figure 1.5 : CDTECH Process Flow Diagram
1.7.2 UOP Process The process of production of MTBE using UOP Process consists of three basic steps as listed below [2]: •
Butane isomerization (Butamer Process).
•
Isobutane dehydrogenation (Oleflex Process). - 12 -
•
Isobutylene Etherification (Ethermax Process).
In the Ethermax process isobutylene reacts selectively with methanol to yield MTBE. The reaction proceeds in the liquid phase at mild conditions in the presence of a sulfonic acid ion exchange resin. A simplified flow diagram of the process is shown in Figure 1.6. The reaction of isobutylene with methanol is conducted in the presence of a small excess of methanol relative to that required for the stoichiometric reaction of the isobutylene contained in the C4 feed.
Figure 1.6 : ETHERMAX MTBE Process Flow Diagram
The reaction is virtually 100 % selective except for minor side reactions owing to the presence of certain impurities in the feed. Thus, feed water will give rise to equivalent amounts of tertiary-butyl alcohol (TBA) in the product, whereas feed isoamylene will yield tertiary-amyl methyl ether (TAME). These byproducts are usually unimportant, since they can be used equally as gasoline blending agents and need not be separated from the MTBE product. The Ethermax process combines the fixed bed reactor technology (originally developed by Huels) with Koch Engineering’s RWD reaction with distillation
technology
to
overcome
equilibrium
limitations
found
with
conventional fixed bed technologies. The methanol and C4 streams are combined, heated, and charged to a fixed bed reactor. In the reactor,
- 13 -
condensation of isobutylene and methanol takes place under relatively mild conditions. Although a tubular reactor can be used, UOP has opted for an adiabatic packed bed reactor for the large flow rates associated with Oleflexbased units. The catalytic section of the RWD column uses a structured packing that contains catalyst to overcome reaction equilibrium constraints by continuously fractionating the product from unreacted components. The key to success with this type of technology is proper distribution of liquid and vapor in the reaction zone; efficient contact of reactants with catalyst; and instantaneous fractionation of products. Thus, the KataMax structured packing, developed by Koch Engineering provides an efficient mass and heat transfer between vapor, liquid, and solid catalyst, and exhibits a high hydraulic capacity. MTBE is taken off as RWD bottoms and C4 raffinate is recovered in the overhead fraction. Residual methanol in the raffinate stream is recovered and recycled with a water wash system to reduce methanol consumption to about stoichiometric levels.
1.7.3 Huntsman TBA/PO Process Huntsman has a 181,000 metric ton per year propylene oxide plant in Port Neches, Texas, which started up in 1994. The plant, originally built by Texaco was designed with Texaco technology. The byproduct TBA is converted to 14,000 barrels per day of MTBE (560,000 metric tons). The Texaco/Huntsman process is believed to be essentially similar to Lyondell's propylene oxide production process. However, MTBE is produced directly from TBA with no intermediate production of isobutylene as in the Lyondell process, according to patent information. The Texaco/Huntsman process route consists of the well-known steps of peroxidation of isobutane to TBHP and TBA; epoxidation of propylene with TBHP to propylene oxide and TBA; and the reaction of TBA with methanol to form MTBE, although this has not been previously practiced as a one step reaction.
- 14 -
1.8
THESIS OUTLINE AND WORK METHODOLOGY
1.8-1
Research Objectives The main aim of this work is to develop a mathematical model for the production of MTBE in a reactive distillation unit at steady state operation. The model will incorporate rate data, thermodynamics properties and vapor-liquid non-idealities. A suitable coding language (FORTRAN) will be used to code the relevant model equations into used simulation software. The coded equations will be embedded into PROII software using UAS
feature "User Added
Subroutine". The simulation package will be used to assess the effects of some important design and operating variables on RD column performance. 1.8-2
Research Methodology The methodology that will be approached to achieve the objectives of this research topic can be summarized as follows: ¾
Collect and choose the required data for the RD unit (thermodynamics, kinetics, operating parameters…etc).
¾
Formulate a mathematical model for the RD unit.
¾
Code the relevant model equations using a suitable coding language.
¾
Embed the coded equations using a suitable embedding technique such as User Added Subroutines (UAS), Excel Visual Basic for Applications (VBA) or CapeOpen interface into a simulation package whenever it is possible.
¾
Use a suitable simulation software (PROII, Aspen Plus) to solve the model equations.
¾
Perform sensitivity analysis on the effect of some important design, operating and kinetics parameters on column performance.
- 15 -
CHAPTER -2REACTIVE DISTILLATION
- 16 -
CHAPTER -22.1
INTRODUCTION TO REACTIVE DISTILLATION Currently there is considerable academic and industrial interest in multifunctional reactors, involving in-situ separation of products from the reactants [10]. Reactive distillation is one of the most common means of in-situ product removal and has been receiving increasing attention in recent years as an alternative to the conventional reaction-followed-by-distillation processes [11]. Doherty and Buzad in 1992 have placed this subject in historical perspective and list references to show that the advantages of reactive distillation were recognized as early as in 1921. [12] Reactive distillation is potentially attractive whenever a liquid phase reaction must be carried out with a large excess of one reactant. Under such circumstances, conventional processes incur large recycle costs for excess reactant. Reactive distillation, on the other hand can be carried out closer to stoichiometric feed conditions, thereby eliminating recycle costs. Both homogeneous and heterogeneous catalysed chemical reactions can be carried out in a reactive distillation column. Process development, design and operation of RD processes are highly complex tasks. The potential benefits of this intensified process come with significant complexity in process development and design. The nonlinear coupling of reactions, transport phenomena and phase equilibrium can give rise to highly system-dependent features, possibly leading to the presence of reactive azeotropes and/or the occurrence of steady-state multiplicities Furthermore, the number of design decision variables for such an integrated unit is much higher than the overall design degrees of freedom of separate reaction and separation units [13].
2.2
BASICS AND ADVANTAGES OF REACTIVE DISTILLATION Reactive distillation is an old concept that combines chemical reaction and physical separation in the same unit. It is a unit operation that combines a reactor as an integral part of the distillation column as depicted in Figure 2.1. It - 17 -
can be utilized for either equilibrium reactions or non-equilibrium (irreversible) reactions. In the first case, the withdrawal of products as they are formed results in an increase in the conversion that can be achieved. This increase is achieved through a shift in the equilibrium, based on Le Chatelier’s principle. In the second case, it is generally applied to systems where products may react with reactants, causing a decrease in product yield in conventional reactors. The term catalytic distillation is also used for such systems where a catalyst (homogeneous or heterogeneous) is used to accelerate the reaction. In this thesis we use the generic name reactive distillation, with the acronym RD, to cover both catalyzed or unanalyzed reactions systems.
Figure 2.1 Simple Reactive Distillation sketch.
2.2-1 Importance of RD Consider the reversible reaction scheme A+ B ⇔C+ D where the boiling points of the components follow the sequence A, C, D and B. The traditional flow-sheet for this process consists of a reactor followed by a sequence of distillation columns; see Figure 2.2 (a). The mixture of A and B is fed to the reactor, where the reaction takes place in the presence of a catalyst and reaches equilibrium. A distillation train is required to produce pure products C and D. The unreacted components, A and B, are recycled back to the reactor. - 18 -
In practice the distillation train could be much more complex than the one portrayed in Figure 2.2 (a) if one or more
azeotropes are formed in the
mixture. The alternative RD configuration is shown in Figure 2.2 (b). The RD column consists of a reactive section in the middle with non-reactive rectifying and stripping sections at the top and bottom. The task of the rectifying section is to recover reactant B from the product stream C. In the stripping section, the reactant A is stripped from the product stream D. In the reactive section the products are separated in-situ, driving the equilibrium to the right and preventing any undesired side reactions between the reactants A (or B) with the product C (or D). For a properly designed RD column, virtually 100% conversion can be achieved.[14]
Figure 2.2 Processing schemes for a reaction sequence A+ B ⇔ C+ D where C and D are both desired products. (a) Typical configuration of a conventional process consisting of a reactor followed by a distillation train. (b) The reactive distillation configuration. The components A, C, D and B have increasing boiling points. The reactive sections are indicated by grid lines.
The most spectacular example of the benefits of RD is in the production of methyl acetate. The acid catalyzed reaction MeOH + AcOH ⇔ MeOAc + H2O was traditionally carried out using the processing scheme shown in Figure 2.3 (a), which consists of one reactor and a train of nine distillation columns. In the RD implementation (see Figure 2.3 (b)) only one column is required and nearly 100 % conversion of the reactant is achieved. The capital and operating costs are significantly reduced.
- 19 -
Figure 2.3 Processing schemes for the esterification reaction MeOH + AcOH ⇔ MeOAc + H2O (a) Conventional processing scheme consisting of one reactor followed by nine distillation columns. (b) The reactive distillation configuration. The reactive sections are indicated by grid lines.
For the acid catalyzed reaction between iso-butene and methanol to form methyl tert-butyl ether, isobutylene + MeOH ⇔ MTBE , the traditional reactor-followed-by-distillation concept is particularly complex because the reaction mixture leaving the reactor forms three minimum boiling azeotropes. The RD implementation requires only one column to which the butenes feed (consisting of a mixture of n-butene, which is non-reactive, and iso- butene which is reactive) and methanol are fed near the bottom of the reactive section.[13] The RD concept shown in Figure 2.4 (a) is capable of achieving close to 100% conversion of iso-butene and methanol, along with suppression of the formation of the unwanted dimethyl ether. Also, some of the azeotropes in the mixture are “reacted away”.[11] For the hydration of ethylene oxide to mono-ethylene glycol: EO + H2OÆ EG, the RD concept, shown in Figure 2.4 (b) is advantageous for two reasons [15]. Firstly, the side reaction EO+EGÆ DEG is suppressed because the concentration of EO in the liquid phase is kept low because of its high volatility. - 20 -
Secondly, the high heat of reaction is utilized to vaporize the liquid phase mixtures on the trays. To achieve the same selectivity to EG in a conventional liquid phase plug flow reactor would require the use of 60% excess water Similar benefits are also realized for the hydration of iso-butene to
tert-butanol [16] and hydration of 2-methyl-2-butene to tert-amyl alcohol
[17].
Figure 2.4 (a) Reactive distillation concept for synthesis of MTBE from the acid catalysed reaction between MeOH and iso-butene. The butene feed is a mixture of reactive iso- butene and non-reactive n-butene. (b) Reactive distillation concept for the hydration of ethylene oxide to ethylene glycol. (c) Reactive distillation concept for reaction between benzene and propene to form cumene. (d) Reactive distillation concept for reaction production of propylene oxide from propylene chlorohydrin and lime. The reactive sections are indicated by grid lines.
Several alkylation reactions, Aromatic + Olefin ⇔ Alkylaromatic, are best carried out using the RD concept not only because of the shift in the reaction equilibrium due to in-situ Separation but also due to the fact that the undesirable side reaction, Alkyl Aromatic+Olefin ⇔ dialkyl aromatic , is suppressed. The reaction of propene with benzene to form cumene, Benzene+Propene ⇔ Cumene ([18]; see Figure 2.4 (c)), is advantageously carried out in a RD column because not only is the formation of the undesirable di-isopropylbenzene suppressed, but also the problems posed by high exothermicity of the reaction for operation in a conventional packed bed reactor are avoided. Hot spots and runaway problems are alleviated in the RD concept where liquid vaporisation acts as a thermal flywheel. The alkylation of isobutane to isooctane, isobutene+ n-butane ⇔ isooctane , is another reaction that benefits from a RD implementation because in-situ separation of the product prevents further alkylation: isooctane + n-butene ⇔ C12H24 [18]. - 21 -
The reaction between propylene chlorohydrin (PCH) and Ca(OH)2 to produce propylene oxide (PO) is best implemented in an RD column, see Figure 2.4 (d). Here the desired product PO is stripped from the liquid phase by use of live steam, suppressing hydrolysis to propylene glycol [19]. Co-current gas-liquid down flow trickle bed reactors are widely applied for hydroprocessing of heavy oils. This co-current mode of operation is disadvantageous in most hydroprocesses [11], and counter-current flow of gas and liquid would be much more desirable as shown in Figure 2.5. The countercurrent reactor shown in Figure 2.5 (b) is essentially a RD column wherein the H2S is stripped from the liquid phase at the bottom and carried to the top. The quantitative advantages of the RD implementation for hydroprocessing are brought out in a design study carried out by Van Hasselt in 1999. For a 20,000 bbl/day hydrodesulphurisation unit with a target conversion of 98% conversion of sulphur compounds, the catalyst volume required for a conventional trickle bed reactor is about 600 m3. For counter-current RD implementation the catalyst volume is reduced to about 450 m3.
Figure 2.5 Hydrodesulphurisation of gas oil carried out in (a) cocurrent trickle bed reactor and (b) counter-current RD unit
From the foregoing examples, the benefits of RD can be summarized as follows: a) Simplification or elimination of the separation system can lead to significant capital savings. b) Improved conversion of reactant approaching 100 %. This increase in conversion gives a benefit in reduced recycle costs.
- 22 -
c) Improved selectivity. Removing one of the products from the reaction mixture or maintaining a low concentration of one of the reagents can lead to reduction of the rates of side reactions and hence improved selectivity for the desired products. d) Significantly reduced catalyst requirement for the same degree of conversion. e) Avoidance of azeotropes. RD is particularly advantageous when the reactor product is a mixture of species that can form several azeotropes with each other. RD conditions can allow the azeotropes to be “reacted away” in a single vessel. f) Reduced by-product formation. g) Heat integration benefits. If the reaction is exothermic, the heat of reaction can be used to provide the heat of vaporization and reduce the reboiler duty. h) Avoidance of hot spots and runaways using liquid vaporization as thermal fly wheel. 2.2-2
Constraints and Difficulties in RD implementation Against the above mentioned advantages of RD, there are several constraints and foreseen difficulties [20]. 1. Volatility constraints. The reagents and products must have suitable volatility to maintain high concentrations of reactants and low concentrations of products in the reaction zone. 2. Residence time requirement. If the residence time for the reaction is long, a large column size and large tray hold-ups will be needed and it may be more economic to use a reactor-separator arrangement. 3. Scale up to large flows. It is difficult to design RD processes for very large flow rates because of liquid distribution problems in packed RD columns. 4. Process conditions mismatch. In some processes the optimum conditions of temperature and pressure for distillation may be far from optimal for reaction and vice versa.
- 23 -
2.2-3 The complexity of RD The design and operation issues for RD systems are considerably more complex than those involved for either conventional reactors or conventional distillation columns. The introduction of an in-situ separation function within the reaction zone leads to complex interactions between vapour-liquid equilibrium, vapour-liquid mass transfer, intra-catalyst diffusion (for heterogeneously catalysed processes) and chemical kinetics. Figure 2.6 shows the various transfer processes in homogeneous and heterogeneous RD. In heterogeneous RD the problem is exacerbated by the fact that these transfer processes occur at length scales varying from 1 nm (pore diameter in gels, say) to say a few meters (column dimensions); see Figure 2.7. The time scales vary from 1 ms (diffusion within gels) to say a few hours (column dynamics). The phenomena at different scales interact with each other. Such interactions, along with the strong non-linearities introduced by the coupling between diffusion and chemical kinetics in counter-current contacting, have been shown to lead to the phenomenon of multiple steady states and complex dynamics, which have been verified in experimental laboratory and pilot plant units [36]. Successful commercialization of RD technology requires careful attention to the modelling aspects, including column dynamics, even at the conceptual design stage [12]. In some cases the reactor and distillation paradigms do not translate easily to RD. The potential advantages of RD could be nullified by improper choice of feed stage, reflux, amount of catalyst, boilup rate, etc. Thus, it is possible to decrease conversion by increasing the amount of catalyst under certain circumstances. Increased separation capability could decrease process performance [21].
- 24 -
Figure 2.6 Transport processes in RD. (a) homogeneous liquid phase reaction, and (b) heterogeneous catalyzed reactions.
Figure 2.7 Length and time scales in RD.
2.2-4
Models available for design of RD columns A variety of models exist in the literature for design of RD columns. They can be classified in the following manner. 1. Steady-state equilibrium (EQ) stage model 2. Dynamic EQ stage model 3. Steady-state EQ stage model with stage efficiencies
- 25 -
4. Dynamic EQ stage model with stage efficiencies 5. Steady-state nonequilibrium (NEQ) stage model, where the interphase mass transfer is described by rigorous Maxwell-Stefan diffusion equations 6. Dynamic NEQ stage model 7. Steady-state NEQ cell model, developed by Higler in 1999, in order to account for staging of the vapour and liquid phases during cross-current contacting on a distillation tray [22]. The equilibrium stage model assumes that the vapor and liquid streams leaving a given stage are in thermodynamic equilibrium with one another. These models can be coupled with the assumption of chemical equilibrium at each stage or the kinetics can be described using an nth order kinetic reaction model. The column is described by a group of equations that model the equilibrium stages in a column configuration. These are known as the MESH equations. MESH stands for: •
M: Material balance equations for each component and total mass.
•
E: Equilibrium equations.
•
S: Summation equations or composition constraints.
•
H: Heat or energy balance equations.
- 26 -
CHAPTER -3-
LITERATURE SURVEY
- 27 -
CHAPTER -33.1
OVERALL REVIEW The combination of chemical reaction with distillation in only one unit is called reactive distillation. The performance of reaction with separation in one piece of equipment offers distinct advantages over the conventional, sequential approach [7]. In reactive distillation (RD) chemical reactions occur within the distillation column to achieve specific goals, such as to obtain high conversions and high purity products as well to minimize side reactions. Currently RD has various applications, such as the production of methyl tertiary -butyl ether (MTBE) and ethylene oxide (EO) and others. Reactive distillation is being used in industrial applications with more frequency because of increasing research and development of this technology, a result of commercial and academic experience and success. Examples of commercialized technologies are the Ethermax process from UOP-Huls and Koch Engineering that uses the KataMax structured packing from Koch and Catacol, a low cost reactive distillation technology for etherification from IFP’s Industrial Division [22]. So far, there is no generally accepted method for the design of distillation with reaction. Most of the systematic methods available possess limitations because of their simplified assumptions. Moreover, these methods have rarely been proven with a variety of reactive distillation processes and they do not consider the design in detail [12]. In spite of the advances in separation with reaction processes, reactive distillation still relies on intuition and expertise. A reactive distillation problem can be studied using different approaches including: feasibility, simulation, modeling, design and experimental studies in the laboratory and the pilot plant. A combination of all of these methods gives rise to the most accurate solution to the problem. One very important aspect of predicting the behavior in these systems is the model used to design and simulate the reactive distillation process. In the literature, the most common models that have been developed and proven are the equilibrium stage model and the non-equilibrium stage - 28 -
model. The equilibrium stage model is based on the conventional equilibriumstage model of a distillation column with the addition of the reaction terms in the mass and energy balances. The non-equilibrium stage model for reactive distillation, also known as the rate-based model, is an extension of the conventional rate based model for distillation. In this chapter, a discussion of the important aspects of modeling, simulation, design and analysis of reactive distillation is provided. Matthias et al, [24] presented a conceptual design methodology for the reactive distillation columns, the method assesses feasibility of a proposed reactive distillation, designs the column and allows evaluation of the design for both fully reactive and "hybrid" column configurations. Stage composition lines are used to represent all possible liquid compositions in a column section for specified product compositions and for all reflux or reboil ratios. Reaction equilibrium is assumed on each reactive stage, and vapor-liquid equilibrium is assumed on all stages. The methodology is illustrated by application to an ideal reactive system and for MTBE production. They developed a new graphical design methodology to assess feasibility and design columns for proposed reactive distillation processes. Also, they concluded that the methodology is restricted to systems with a single-feed two-product columns. It is assumed that equilibrium reaction take place in the liquid phase only. A simulation study of the impact of various process configuration catalyst types and catalyst loading on the production rate and separation of the MTBE product was carried by Shah et al, [25]. They mentioned in their paper the advantages and disadvantages of each configurations. For reactive distillation, the catalyst distribution on a range of trays and the feed location is considered for the optimal production/separation scenario. They also studied an approach, based purely on steady state analyses, for synthesizing effective control structures for reactive distillation (RD) columns [26]. The main idea was to analyze the steady-state relationships between the manipulated (input) variables and potential controlled (output) variables to identify input-output (IO) pairings that are sensitive and avoid steady state multiplicities providing a large range of nearly linear operating region around the base case design. The MTBE case study shows that input and output multiplicity in the IO relation - 29 -
occurs in the steady-state relationships. It also shows that the occurrence of multiplicity depends on the control strategy implemented. Rivera and Johan [27] considered the residue curve mapping technique (RCM) a powerful tool for the flow-sheet development and preliminary design of conventional multi-component separation processes. An RCM-based feasibility analysis has been applied to the homogeneous RD synthesis of MTBE at 11 atm from methanol and isobutylene and in the presence of n-butane. The reaction space, defined in terms of transformed composition variables, has been divided into sub-regions characterized by separation boundaries. A feasibility analysis of the RD process has been performed based upon the location of the reacting mixture, defined initial separation sequences have been generated according to the feed transformed-composition. In the generated sequence, high purity MTBE has been obtained as a product, due to the appearance of a pseudo reactive azeotrope, which imposes limitation to the separation task. The progress in chemical engineering unit operations which laid out the catalytic distillation column is a significant advancement in chemical engineering technology. Some of the equilibrium-limited reactions show an improving yield due to the continuous removal of the products from the reaction zone. A substantial research work, mainly experimental has been done for this combined unit operation, most of which have appeared in patents [28 & 29]. However, essential research aspects of the reactive distillation process mathematical modeling, design, optimization and control have been covered yet. Also, little attention has been paid to plate efficiency when reaction occurs. Research and development work on reactive distillation has focused on: •
Computational methods to solve the simulations chemical reaction and vapor-liquid equilibrium equations.
•
Chemical equilibrium aspects in reactive systems (e.g. thermodynamic models).
3.2
COMPUTATIONAL METHODS The Computational methods used for reactive distillation are extensions to the algorithms developed for the solution of the equations for conventional - 30 -
distillation. The first attempts to model reactive distillation were by using the simplified plate-to-plate calculations. Rigorous mathematical models for computer simulation were not developed until the 1970's. Since that time, various techniques have been developed that allow the rigorous solution of the equations. These techniques include equation partitioning methods, and Newton-Raphson based methods. Nelson in 1971 modified the Tierny-Bruno algorithm based on tray-bytray calculation, by taking into consideration the non-ideal vapor liquid equilibrium. Which represented in material balance equations no longer linear in composition [30]. Suzuki et al. in 1971 applied the successive iteration method to reactive distillation problems, and concluded that it converges rapidly and it is stable. However, difficulties with convergence arise for systems with non-ideal solutions, because of non linearity of the equations. In addition, successive iteration methods have the disadvantage that as the solution is approached, the progress of iteration calculations decelerates [31]. Murthy in 1984 considered an extension of the Newton-Raphson algorithm to columns in which chemical reactions occur, while Venkataraman et al. in 1990 were developed an extension of the " inside-out" approach proposed by Boston that combines the advantages of Newton's method with those of the inside-out strategy very effectively [32]. Jelinek and Hlavacek in 1976 applied the relaxation method to solve steady state countercurrent equilibrium stage separation with chemical reaction problems. They confirmed the suitability of this method where an azeotrpoe exists [33]. Teirney and Riquelme in 1982 proposed a correction algorithm which gives quadratic convergence near the solution. Its use was demonstrated on a sample problem of the separation of meta-and para-xylene using experimental results from an earlier study done by Satio et al. in 1971. They found that their solution agree well with the numerical solution obtained by Satio et al, but not with Saito's experimental data. Teirney and Riquelme in 1982 claimed the equilibrium in the reaction is not satisfied in each stage. This contradicts the experimental results of Satio et al. in 1971, showing that the equilibrium - 31 -
constant is almost constant in the range of operating temperature between (50100 ˚C) [34]. The reactive distillation column has no reboiler, although it has to be used in conjunction with the gas/liquid polymer reactor. The model was based on the material balance equations describing the system, and the simulation was carried out by using a fourth-order Runge-Kutta method. The usefulness of the model was limited because of the number of assumptions and simplifications that were made in development of the model, such as: neglecting enthalpy balances, neglecting tray hydraulics [34]. Reactive Distillation (RD) is state-of-the-art multifunctional reactor concept that integrates reaction and distillation in a single process unit. Some recent reactions that have been proposed to utilize RD technology involve noncondensable species like hydrogen which proposed by Kamath et al in 2005. Reactions involving liquid phase splitting have also been examined in RD . Very little research has been done for such complex cases and the potential of RD for such reactive systems has not been thoroughly investigated [36]. This work focuses on the presence of both non-condensables and liquid phase splitting in RD using the one-step synthesis of methyl isobutyl ketone (MIBK) from acetone as a case study. Traditionally, MIBK is produced in three steps by the condensation of acetone to diacetone alcohol followed by its dehydration to mesityl oxide (MO) and then its hydrogenation to MIBK. The first two reactions are reversible and limited by thermodynamic equilibrium. But the third reaction (MO to MIBK) is fast and irreversible and hence the overall reaction (acetone to MIBK) is expected to be an irreversible reaction. However, the presence of water as a product inhibits the reaction rate considerably leading to pseudo-equilibrium [36]. The work employs the RD technology to minimize the amount of water in the reactive zone by simultaneous distillation and thereby increase the overall conversion of acetone. The hydrogenation reaction results in the appearance of a non-condensable species in the RD column while the ternary acetone-waterMIBK system results in liquid-liquid phase splitting (Figure 3.1).
- 32 -
Figure 3.1 Ternary LLE diagram for Acetone-water-MIBK system at 5atm.
A potential RD configuration for the synthesis of MIBK was developed as shown in Figure 3.2. The column configuration is capable of achieving a very high conversion of around 95% by employing a very high reflux ratio (defined as ratio of the liquid refluxed to the liquid distillate). The liquid phase composition profile in the column is shown in Figure 3.3. The simultaneous distillation in the reactive zone keeps the water concentration below 35% which would not have been possible in a conventional reactor. Because of the presence of a binary acetone-water azeotrope at higher pressures, a large number of stages may be required to reduce the water concentration in the top. As expected, phase-splitting is observed in the stripping section. However, the presence of sufficiently large amount of acetone in the reactive zone ensures a homogenous liquid phase [35].
- 33 -
Figure 3.2 RD Column Configuration.
Figure 3.3 Liquid Composition profile in RD Column.
It was concluded from the above work that there is a further scope for improvement of the proposed RD configuration. The column performance is very sensitive to the feed locations and their flow rates. The presence of noncondensable
hydrogen
poses
additional
complexities
in
the
VLLE
computations. A successful RD process is not only capable of achieving a much higher conversion as compared to the conventional route but also leads to process intensification and compactness. - 34 -
An approach, based purely on steady-state analyses, for synthesizing effective control structures for reactive distillation (RD) columns . The main idea is to analyze the steady-state relationships between the manipulated (input) variables and the potential controlled (output) variables to identify input–output (IO) pairings that are sensitive and avoid steady-state multiplicities providing a large range of nearly linear operating region around the base case design. Traditional SISO control loops are then implemented using these IO pairings to obtain control structures that maintain the column near the design product purity and conversion for the anticipated primary disturbances. The Niederlinski Index is used to eliminate dynamically unstable pairings in control structures with multiple loops. The approach is demonstrated on an example MTBE RD column. The impact of steady-state multiplicities on control structure design is highlighted. [25] This work illustrates the systematic analysis of the steady-state IO relationships for synthesizing effective control structures for RD columns. Such control structures are necessary to provide robust, stable, safe and economical column operation to tide over disturbances entering the column. Effective control structures can be obtained by identifying IO pairings that avoid multiplicities and allow a sizeable nearly linear operating region around the base case. The MTBE case study shows that input and output multiplicity in the IO relations occurs in the steady-state relationships. It also shows that the occurrence of multiplicity depends on the control strategy implemented. For the MTBE example, a control structure that uses the reboiler duty to control Tray 11 temperature in the stripping section and the butene feed to control Tray 10 isobutylene composition in the reactive section is found suitable. Proper choice of the manipulated variables and the tray locations for the measurements in the two loops is especially critical for effective column regulation. At a more general level, the results show that maintaining the stoichiometric balance of the fresh feeds is needed for the effective control of double feed RD columns. M. H. M. Reis et al in 2006, presented a novel approach for establishing the route for process intensification through the application of two developed - 35 -
softwares to characterize reactive mixtures is presented. A robust algorithm was developed to build up reactive phase diagrams and to predict the existence and the location of reactive azeotropes. The proposed algorithm does not depend on initial estimates and is able to compute all reactive azeotropes present in the mixture. It also allows verifying if there are no azeotropes, which are the major troubles in this kind of programming. An additional software was developed in order to calculate reactive residue curve maps. Results obtained with the developed program were compared with the published in the literature for several mixtures, showing the efficiency and robustness of the developed software's. [40] Figure 3.4 shows the nonreactive residue curve map for the system under consideration. The binary azeotrope between methanol and tert-butyl methyl ether (MTBE) is a saddle point and between isobutylene and methanol an unstable node appears, leading to the occurrence of a distillation boundary. This distillation boundary splits the diagram in two different regions of distillation, making the separation of the ternary mixture impossible. [40]
Figure 3.4 Nonreactive residue curve map for the system IBTE+ MeOH ↔ MTBE at 101.32 kPa.
Figure 3.5 shows the reactive residue curve map for this system, where it can be seen that, under atmospheric pressure, this reactive mixture does not present any azeotrope, showing how the chemical equilibrium influences the phase equilibrium. This conclusion is in agreement with the results of Barbosa
- 36 -
and Doherty (1988b). Figure 3.6 shows the reactive phase diagram for isobutylene/methanol/MTBE at 101.3 kPa, confirming that this reactive system is zeotropic.
Figure 3.5 Reactive residue curve map for the system IBTE+ MeOH ↔ MTBE at 101.32 kPa.
Figure 3.6 Reactive phase diagram for the system IBTE/ MeOH /MTBE at 101.32 kPa.
Maier et al. (2000) studied this same system at 810.56 kPa (8 atm) and with different constant values for the equilibrium constant (Kr). Similar results were obtained in this work. When Kr is equal to 49.0, the mixture forms two azeotropes at different compositions. Figure 3.7 shows the reactive phase equilibrium diagram for this case. [40] - 37 -
Figure 3.7 Reactive phase diagram for the system IBTE/ MeOH /MTBE at 810.56 kPa.
In actual processes to produce MTBE through reactive distillation, the source of iC4 consists of C4 cuts available from steam or catalytic crackers. In order to analyze this multicomponent system, it is considered that C4 stream consists, basically, of isobutylene and n-butenes, being n-butenes inert components in the reaction. For this quaternary system (ethanol/isobutylene/MTBE/n-butene) the search of azeotropes were carried out at pressures of 1013.2, 2026.4 and 4052.8 kPa (10, 20, and 40 atm). Although in this high pressure it is not correct to consider the vapor as an ideal mixture, this assumption will be still considered, in order to compare the obtained results with those published by Maier et al. (2000). [40]
- 38 -
CHAPTER -4MODELING AND SIMULATION OF MTBE PROCESS
- 39 -
CHAPTER -44.1
MTBE PRODUCTION BY PROII 4.1-1 Introduction This section will demonstrate the use of PRO/II in the simulation of the synthesis of methyl tert-butyl ether (MTBE), also the simulated process for MTBE production will be depicted and each unit in the process will be elaborated. A PRO/II simulation model of a typical MTBE plant is presented here. The process plant includes a reactor along with an azeotropic distillation column for separation of the MTBE product. A reactive distillation section is added to the MTBE azeotropic column in order to increase the overall conversion to MTBE. This is followed by the methanol recovery section which includes a liquid-liquid extractor. The thermodynamics successfully predicts the azeotropic removal of methanol from the MTBE product stream. 4.1-2 Process Description There are several variations on MTBE plant designs. In general, an MTBE plant is comprised of the three sections, a reactor section, a MTBE recovery section, and a methanol recovery section. For the Ethermax process, the MTBE recovery section includes a second reaction zone in the distillation column. The complete process flow diagram for the MTBE reactive distillation plant model used in this simulation is given in Figure 4.1.
- 40 -
Figure 4.1 MTBE Plant Flowsheet.
MTBE is manufactured by catalytically reacting isobutylene and methanol in a fixed-bed reactor at a moderate temperature and pressure. The reaction is exothermic and reversible, and is carried out in the liquid phase over a fixed bed of ion-exchange resin-type catalyst. It is highly selective since methanol reacts preferentially with the tertiary olefin. In the Ethermax MTBE process, modeled here, an isobutylene-rich mixed C4 stream is mixed with fresh methanol along with a small amount of recycle methanol and fed to the reactor section. The reactors are cooled to under 200 °F to prolong catalyst life and to minimize the undesirable side reactions such as dimerization of isobutylene. The methanol-to-isobutylene ratio in the reactor feed is kept low to minimize the costs of recovering unreacted methanol, and to facilitate the operation of the MTBE column which will be discussed later. Generally, this ratio is maintained close to 1:1. Table 4.1 contains the reactor feed composition used in this model. - 41 -
Table 4.1 Reactor Feed data
Stream Name C4 Feed Methanol Feed Methanol Recycle Stream Number 2 1 20 850 277.5 4.3 Flowrate (kg-mol/hr) 16 16 44 Temperature (°C ) 1620 1620 1724 Pressure (kPa) Mole % Component N-butane 9.0 0.0 0.0 Isobutane 41.0 0.0 0.0 1-butene 7.0 0.0 0.0 Cis 2-butene 4.0 0.0 0.0 Trans 2-butene 6.0 0.0 0.0 Isobutylene (IBTE) 33.0 0.0 0.0 MTBE 0.0 0.0 0.0 Tert-butanol (TBA) 0.0 0.0 0.0 Water 0.19 0.0 6.98 Di-isobutylene (DIB) 0.0 0.0 0.0 Methanol (MEOH) 0.0 100 93.02
An isobutylene conversion to MTBE of 90 to 93% is easily achieved in the reactor. Overall isobutylene conversions higher than those obtained in the standard process can be achieved by either recycling a portion of the MTBE column overhead product, or by providing a second reactor unit and MTBE column downstream of the first MTBE column. The cost-effectiveness of these options vary from plant to plant, but both require greater capital expenditure. In the reactive distillation process, no major increase in capital expenditure is required and overall isobutylene conversions of over 99% are easily obtained. Any water in the reactor feed (from recycle methanol) is instantly converted to t-butanol. Another impurity, di-isobutylene, is formed by the dimerization of isobutylene. While the formation of di-isobutylene and t-butanol should be minimized, their presence in small concentrations in the MTBE product is acceptable since these byproducts also have very high octane numbers. Table 4.2 shows the three main reactions used in the stoichiometric
- 42 -
reactor model. The base component and the fraction converted are also shown. Table 4.2 Reaction Stoichiometry
Reaction 2 (IBTE) = DIB H O + IBTE = TBA 2
Base Component Conversion % IBTE 0.25 100.00 H0
IBTE + MEOH = MTBE
2
MEOH
93.0
Catalyst Used A common catalyst for the MTBE synthesis process is the Amberlyst 15 polymeric catalyst developed by Rohm and Haas. Approximate properties of the commercial form of this catalyst, along with suggested operating conditions are provided below in Table 4.3. Table 4.3 Approximate Catalyst Properties of Amberlyst 15
Physical form Ionic form Acid site concentration Moisture content Apparent density Particle size Shrinkage: Wet to methanol Wet to MTBE Porosity Average pore diameter Surface area
Spherical beads Hydrogen 1.8 meg/ml (4.9 meg/g) 53% 770 g/l 0.35-1.2 mm 4% 12% 0.30 cc/g 250 A 2
45 m /g 3
Bulk density
48 lb/ft Maximum operating temperature 120 °C Minimum bed depth 0.61 m -1 Flowrate, LHSV 1-5 hr MTBE Recovery Section
In the Hüls process, the reactor products are processed in the MTBE column where MTBE, along with t-butanol, dimerized butylene, and a trace amount of methanol, are removed as the bottoms product. In the Ethermax process, further reaction of the isobutylene to MTBE takes place in a section of the distillation column containing the catalyst resin in tower packing. The MTBE is - 43 -
removed as the bottoms product in a manner similar to the standard process. The MTBE product is greater than 99.5% pure and requires no further purification. The key to operating the MTBE column is to have a sufficient amount of C4s in the column feed to form azeotropes with the methanol in the feed. Conversely, if a proportionately large amount of methanol is present in the column feed, it may result in breakthrough of methanol with the MTBE bottoms product. Therefore, suitable azeotrope formation is possible only when a limited excess of methanol is used in the reactor feed. In this manner, unreacted methanol, which has a higher boiling point than MTBE, is fractionated away from the MTBE bottoms. The overhead product containing non-reactive linear butenes, iso and normal butanes, and unreacted methanol and isobutylene, is sent to the methanol recovery section. Methanol Recovery Section In the methanol recovery section, the MTBE column overhead product is water washed to extract methanol. This unit is simulated as a liquid-liquid extraction column. The raffinate, which contains less than 10 ppm methanol, is suitable for recovering high purity C4 isomers, or as a feed to an alkylation unit. The extract phase which contains water, methanol and small amounts of dissolved hydrocarbons is warmed and flashed to remove the hydrocarbons. The resultant methanol-water mixture is fractionated to recover methanol as the overhead product. The methanol (with a trace of water) is recycled to the MTBE reactor. The wash water stream from the bottoms, along with a small amount of makeup water, is returned to the water wash column.
- 44 -
4.2
MODELING AND SIMULATION OF MTBE PROCESS USING PRO/II
PRO/II software has the capabilities of solving Reactive Distillation Processes utilizing Chemdist Algorithm provided by the software, but it requires knowledge of Material and Energy balances equations as well as Thermodynamics and kinetics of reaction. The Chemdist algorithm in PRO/II is a Newton based method which is suited to solving non-ideal distillation problems involving a smaller number (10 vs. 100) of chemical species. These conditions are generally encountered in chemical distillations as opposed to crude fractionation where the I/O algorithm would be a better choice. Chemdist is designed to handle both vapor-liquid and vapor-liquid-liquid equilibrium problems as well as chemical reactions. All derivatives for the Jacobian matrix are calculated analytically. User-added thermodynamic options that are used by Chemdist must provide partial derivatives with respect to component mole fractions and temperature. Chemdist uses the chain rule to convert these to the needed form. PRO/II automatically generates numerical estimates for the reaction rate derivatives. In many cases, this is sufficient. However, certain reactions require the use of more accurate analytical derivatives that provide better solutions. Reactions requiring these more accurate analytical derivatives include: reversible reactions, exothermic reactions, and/or reactions where the equilibrium is sensitive to temperature. There are some assumption will be considered to simplify the simulation work which can be listed as : 1. The reactor is operated under steady state conditions. 2. The reactor is isothermal operating condition (55 C), it is maintained by circulating a coolant. 3. The pressure drop across reactor is specified with 69 kpa. 4. One dimensional homogeneous model is considered. 5. Dispersion of mass and energy terms are negligible and a plug flow reactor model is assumed. 6. Complete mixing of gases entering the collection chamber before the catalyst beds. No reaction takes place inside the chamber, the reaction
- 45 -
is only in the liquid phase. 7. Physical properties are considered as a function of temperature and pressure, and SRKM thermodynamic method is used. 4.3
THERMODYNAMIC DATA The VLE fractionators are simulated well with PRO/II's modified SoaveRedlich-Kwong (SRKM) equation of state method. For this method, PRO/II contains extensive, built-in databanks that encompass binary interaction parameter data for the majority of component pairs present in this simulation. However, binary interaction data (Kijs) are directly supplied for 8 component pairs to improve the accuracy of the separations in the columns as demonstrated in Figure 4.2. The binary interaction parameters are listed below in Table 4.4. The thermodynamic set used for VLE thermodynamics is referred to as SRKM_VLE. The used components and their abbreviation are as follows: Table 4.4 Binary Interaction Parameters for SRKM_VLE Comp I
NC4
1BUT
BTC2
BTT2
IBTE
MEOH
TBA
MTBE
n-butane
1-butene
Cis2-butene
Trans2butene
Isobutylene
Methanol
TertButanol
Comp j
TBA
MEOH
MEOH
MEOH
MEOH
TBA
H20
DIB
kija
0.0469
0.136
0.136
0.136
0.13553
-
-0.145
0.05785
-0.253
-0.0093
0.07397 kjia
0.1260
-0.0323
-0.0323
-0.0323
-0.0322
0.05522
kijb
0.0
0.0
0.0
0.0
0.0
0.0
0.0
-10.144
kjib
0.0
0.0
0.0
0.0
0.0
0.0
0.0
6.17
UOM
K
K
K
K
K
K
K
K
- 46 -
Figure 4.2 Thermodynamic Data of PRO/II software.
Transport properties are needed in order to use the rigorous heat exchanger model in the MEOH recovery section. Transport property calculations are set to pure-component averages by default and can be modified by modifying the thermodynamic method of choice. The liquid extraction unit is simulated using the SRKM method for VLLE thermodynamics with binary interaction data again supplied as part of the input. The thermodynamic set used for VLLE thermodynamics is referred to as SRKM_VLLE. The L1KEY component (i.e., the predominant component in the L1 liquid phase) is specified as n-butane. The L2KEY component is specified to be water. Explicitly specifying the key components eliminates the need for PRO/II to find an appropriate immiscible pair, reducing the computation time. All the azeotropes are properly predicted. The form of SRKM is presented as: 1 2
aij = (ai a j ) [(1 − kij ) + (kij − k ji )( X i /( X i + X j )) ij ] c
Where : Xi= Mole fraction of component i Xj= Mole fraction of component j Kij= Kji = Binary interaction parameter aij= The mixing rule, Cij= constant - 47 -
………..………(4.1)
The accuracy of correlating vapor-liquid equilibrium data using a cubic equation of state can be improved further by choosing an appropriate mixing rule. The original mixing rule was derived from the van der Waals one-fluid approximation: P = RT /(v − b) − a (T ) /(v 2 + ubv + wb 2 ) …………………………………....……..(4.2)
where: P = the pressure T = the absolute temperature v = the molar volume u,w = constants, typically integers
a = ∑∑ xi x j aij …………………………………………………………....……..(4.3) i
j
b = ∑ xi bi …………………………………...............................................…….(4.4) i
where: Xi = mole fraction of component i. The binary interaction parameter, kij, is introduced into the mixing rule to correct the geometric mean rule of parameter a in the general cubic equation of state (4.2): The original mixing rule is capable of representing vapor-liquid equilibria for nonpolar and/or slightly polar systems using only one (possibly temperature-dependent) binary interaction parameter. 4.4
MTBE REACTION KINETICS The algorithm used for the reactive distillation column model is a Newtonian-based algorithm. Therefore, in order to accurately model the MTBE reaction in the distillation column, we need to determine not only the reaction rate of the reaction, but also the temperature and composition derivatives of the rate. These derivatives may be generated numerically by an estimation method, or analytically by an expression based on the reaction rate equation. The MTBE reactions fit all three of the reaction types, requiring the use of analytical derivatives. The reaction rate expression and its analytical derivatives can be easily and readily entered by the user in the Procedure Data - 48 -
category of input. The MTBE reaction rate expression used in this simulation model is based on the rate expression described in a paper by Al-Jarallah [38]. First, in the Reaction Data category of input, the stoichiometry of the forward reaction is given (IBTE + MEOH = MTBE). The kinetic data will be provided later on in the Procedure Data category of input using FORTRAN-like language as the procedure named ALJD. The procedure data used in the reactive distillation column model is entered in the Kinetic Procedure data entry window shown in App (A). The MTBE synthesis reaction can be represented by: A + B ↔ C ………………………………………………..…..………………….(4.5) Where, A, B and C denote methanol, isobutylene and MTBE respectively, In general, the forward reaction is order a in A and order b in B, and the reverse reaction is order c in C. The rate of surface reaction, rs, is assumed to be the rate controlling step, as there were no mass transfer limitations. There are two possible mechanisms by which this surface reaction takes place: 1. Reaction between adsorbed molecules of both A and B on adjacent active centers, and 2. Reaction between one adsorbed reactant and the other reactant in solution [37]. The first mechanism is the Langmuir-Hinshelwood mechanism and the second one is the Rideal-Eley mechanism as discussed by Smith in 1981 and Satterfield in 1980. In these references the reaction is assumed to be a simple reaction, that is, the reaction is first order in all species. The following rate equations were derived for general orders of reaction a, b and c. For a Langmuir-Hinshelwood model, the rate of reaction can be represented by the following equation [37]:
⎡ ⎤ C Aa C Bb − C Cc / k rs = k s k k × ⎢ …………..……….…………….(4.6) a +b ⎥ ⎣⎢ (1 + K A C A + K B C B + K C C C ) ⎦⎥ a A
b B
For the case of the Rideal-Eley mechanism, there are two possibilities in which either one of the two reactants is adsorbed on the catalyst and then reacts with - 49 -
the other reactant in solution. For the case when the methanol (A) is adsorbed and reacted with the isobutylene (B) in solution, the final rate equation is:
⎡ C Aa C Bb − C Cc / k ⎤ rs = k s k Aa × ⎢ …………………………..………………..….(4.7) a ⎥ ⎢⎣ (1 + K A C A + K C C C ) ⎥⎦ For the case when isobutylene is absorbed and reacted with methanol in solution, the final rate equation is:
⎡ C Aa C Bb − C Cc / k ⎤ rs = k s k Bb × ⎢ …………………………..…………………...(4.8) b ⎥ ⎢⎣ (1 + K B C B + K C C C ) ⎥⎦ For a given set of a, b and c the unknown parameters in Equation (4.6), (4.7) and (4.8) are the surface reaction rate constant, Ks the equilibrium adsorption constants KA, KB and KC and the thermodynamic equilibrium constant, K. This equilibrium constant can be calculated from experimental concentration data in which concentration equilibrium has been reached [37]. Since K = Kx Kγ …………………………………………………………………………..(4.9) The dependence of the rate constant, Ks, on temperature was determined from the Arrhenius equation, Ks = Kso exp (-E1 / RT)………………………………………………..………....(4.10)
The values of Kso and E1 were found from the least squares fit of Equation (4.10) Thus: 13
Ks = 1.2 x 10
exp (-87,900/RT)……………………………………………..(4.11)
The dependence of the adsorption constants, KA and KC, was determined from the van’t Hoff equation, KA = KAo exp (∆HA / RT)………………………………………………..…........(4.12)
- 50 -
KC = KCo exp (∆HC / RT)………………………………………………..…...…..(4.13)
The values of KAO, KCO, ∆HA and ∆HC where obtained from the least squares fit of the above two equations. Thus: KA = 5.1 x 10-13 exp (97,500/RT)……………………………………………….(4.14) and KC = 1.6 x10-16 exp (119,000/RT)…………………………………….………..(4.15) The reaction rate equation described by Al-Jarallah takes into account the forward and the reverse reaction. We have modified Al-Jarallah's rate equation for MTBE process to simulate the effect of catalyst loading on the reaction rate. This was achieved by removing the catalyst terms from the concentration terms. The modified reaction rate is given by:
⎡ ⎤ C Aa C Bb − C Cc / k rs = k s k A × ⎢ …………………………..………………(4.16) a +b ⎥ ⎣⎢ (1 + K A C A + K B C B ) ⎦⎥ where:
4.5
•
Ks: surface reaction rate constant= 1.2x1013 exp(-87900/RT) in (gmole/g catalyst)
•
KA: equilibrium adsorption constant = 5.1x10-13 exp(97500/RT) in g catalyst/gmole
•
KC : equilibrium adsorption constant = 1.6x10-16 exp(119000/RT) in g catalyst/gmole
•
K : equilibrium constant
•
CA : IBTE concentration in mole/l
•
CB : MEOH concentration in mole/l
•
CC : MTBE concentration in mole/l
MODELING PROCEDURE In modeling this process, the Steady-state equilibrium stage model, will all its
underlined assumptions has been used. The main feature of the model is the assumption that the vapor and liquid streams leaving a given stage are in a thermodynamic equilibrium. The reaction kinetics in the reaction zone was assumed to follow Al-Jarallah kinetics [37]. A typical schematic diagram of a catalytic distillation column is shown in Figure 4.3.
- 51 -
Figure 4.3 A schematic representation of a catalytic distillation column.
A schematic
diagram of an equilibrium tray is shown in Figure 4.4. It is
assumed that the tray has one feed stream Fk one vapor side streams SV one liquid side stream SL and coolings Q occurs.
Feed
V k
Lk+1
SVk
Stage K
Fk
Qk
HEAT DUTY
T k , Pk
∆Rk, Mk
SLk V
Lk
k-1
Figure 4.4 Representation of a generic equilibrium stage.
- 52 -
4.5-1 Model equations for a single stage The model equations for a generic stage k and component i are presented based on the commonly used distillation column, with incorporation of the reaction terms, they may be expressed as follows: 1. Component material balance on stage k: r
Lk+1 X k+1,i + Vk −1Yk−1,i − (Lk + SLk ) X k ,i − (Vk + SVk )Yk ,i + Fk Zk ,i ± ∑ Rjk,i = 0 …………..…(4.17) j =1
Where; Xk,j Yk,i (Lk+SLk) (Vk+SVk) Fk ∑R jk,i
is the mole fraction of liquid phase for species i in stage k. is the mole fraction of vapor phase for species i leaving stage k is the molar liquid flow rate leaving stage k. is the molar vapor flow rate leaving stage k. is the feed flow rate to stage k. is the rate of disappearance of component i due to reaction j on stage k (i.e. when the sign is negative).
2. Total material balance on stage k:
Lk +1 + Vk −1 − ( Lk + SLk ) − (Vk + SVk ) + Fk + Rk = 0 …………….…………….(4.18) 3. Energy balance on stage k: r
Lk +1hk +1 + Vk −1 H k −1 − ( Lk + SL k ) hk − (Vk + SV k ) H k + Fk h fk − Q k + ∑ R jk H jk = 0 ………..(4.19) j =1
Where; hk HK Hj,k Qk
denotes the molar enthalpy of liquid phase in stage k. denoted the molar enthalpy of vapor phase in stage k. is the heat of reaction in stage k. is the heat duty in stage k.
For stages without feed or side streams or no reaction, these terms are assigned a value of zero in Equations (4.17-4.19). 4. Component equilibrium relationship on stage k:
Yk,i = γk,i Kk,i X k,i ……………………………….....…….………..(4.20) Where; γk,i Kk,i
denotes the activity coefficient of component i in stage k. denotes the vapor-liquid equilibrium constant.
- 53 -
5. Constraint equations:
∑Y
k ,i
− 1 = 0 …………………………………………………….………..(4.21)
i
∑X
k ,i
− 1 = 0 …………………………………………………….…..…..(4.22)
i
The unknowns, alternatively referred to as iteration or primitive variables: (X, Y, L, V)i , where i= 1, NT are solved for directly using an augmented NewtonRaphson method. Specification equations involving the iteration variables are added directly to the above equations and solved simultaneously. 4.6
MODELING AND SIMULATION FRAMEWORK It has been assumed in this work that the production of MTBE using PRO/II is a steady-state process. Therefore, the necessary process flowsheet for production of MTBE by Reactive Distillation has been built and developed using PRO/II as shown in Figure 4.5 , the detailed procedure is described step by step in App. (B). All required files (Model file and process input file) are given in App. (A) and App. (C), respectively.
- 54 -
Figure 4.5 MTBE Process Flowsheet PRO/II .
55
CHAPTER -5RESULTS AND DISCUSSION
56
CHAPTER -5RESULTS OF THE SIMULATION WORK
5.1
In this section, the outcome of simulation work will be presented for each equipment with some demonstrated graphs. The used software which is PRO/II can extract the results on excel file after running the model as shown below. Table 5.1 Summary Table Stream Name Stream Description
1 MEOH FEED
2 OLEFINS
3
5
6 T-1 OVHD
7
Liquid
Liquid
Liquid
Liquid
Liquid
Liquid
Temperature (C?) Pressure, bar
16.0000000 16.2000008
16.0000000 16.2000008
43.5000000 15.8549995
72.0000000 14.8199997
43.5000000 6.2100000
129.0753174 6.9749999
Flowrate, Kgmol/hr
277.5000000
850.0000000 1131.8957520 869.3738403
575.9934692
278.0000000
0.1328171 0.6050468 0.1032997 0.0590279 0.0885424 0.0039012 0.0000071 0.0073578 0.0000000 0.0000000 0.0000000
0.0000005 0.0000001 0.0000002 0.0000014 0.0000007 0.0000000 0.9976990 0.0003628 0.0006739 0.0000000 0.0012612
Phase
Composition NC4 IC4 1BUTENE BTC2 BTT2 IBTE MTBE MEOH TBA H2O DIB
0.0000000 0.0000000 0.0000000 0.0000000 0.0000000 0.0000000 0.0000000 1.0000000 0.0000000 0.0000000 0.0000000
0.0900000 0.4100000 0.0700000 0.0400000 0.0600000 0.3300000 0.0000000 0.0000000 0.0000000 0.0000000 0.0000000
0.0675875 0.3078933 0.0525667 0.0300382 0.0450572 0.2478143 0.0000000 0.2488773 0.0000000 0.0001655 0.0000000
0.0879966 0.4008669 0.0684401 0.0391087 0.0586630 0.0202760 0.3013479 0.0226821 0.0002155 0.0000000 0.0004033
COLUMN T-1 1.00
Liquid Fraction of IBTE Liquid Fraction of MEOH Liquid Fraction of MTBE
0.80
Liquid Fraction of NC4
Fraction
0.60
0.40
0.20
0 0
6.0
12.0
18.0
24.0
Tray Number
Figure 5.1 Feed and products concentration profile among column trays.
57
30.0
Table 5.2 Summary of MTBE column T-1 Column T-1 Summary Tray 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 26 27 28 29 30
0.0
20.0
Temp. Pressure C BAR 43.5 6.21 50.6 6.21 51.3 6.24 51.7 6.26 52.1 6.29 52.4 6.32 52.8 6.35 53.5 6.37 54.5 6.40 56.2 6.43 59.1 6.46 63.5 6.48 68.7 6.51 73.4 6.54 76.9 6.57 91.0 6.59 105.5 6.62 115.4 6.65 121.0 6.67 123.9 6.70 125.5 6.73 126.5 6.76 127.1 6.78 127.5 6.81 127.9 6.84 128.2 6.87 128.4 6.89 128.6 6.92 128.8 6.95 129.1 6.97
40.0
Tem perature, C 60.0 80.0
100.0
Liquid
Net Flow Rates Vapor Feed KG-MOL/HR
633.6 661.6 659.7 658.5 657.3 655.6 651.6 636.0 614.4 582.6 539.7 494.3 462.8 445.4 1206.9 1180.5 1202.3 1237.5 1264.7 1281.9 1292.5 1299.2 1303.8 1307.2 1309.8 1311.9 1313.8 1315.4 1316.9
120.0
1209.5 1237.6 1235.7 1234.4 1233.3 1231.5 1227.6 1214.5 1195.5 1166.4 1126.8 1084.6 1054.2 1036.7 928.9 902.5 924.3 959.5 986.7 1003.9 1014.5 1021.2 1025.8 1029.2 1031.8 1033.9 1035.8 1037.4 1038.9
140.0
5 7 9
Tray
Tray
11
17 19 21 23 25 27
23.7889
0.0
200.0
400.0
600.0
800.0
1000.0
1200.0
1400.0
0.0
200.0
400.0
600.0
800.0
1000.0
1200.0
1400.0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
Vapor
29
Figure 5.2 Temp Profile of column T-1
278.0
Vapor Rate, KG-MOL/HR
3
15
576.0
Duties M*KJ/HR -23.6655
869.4
1
13
Product
Liquid Rate, KG-MOL/HR
Liquid
Figure 5.3 Rates Profile of column T-1
58
Vapor Density, KG/M3 5.000
10.000
15.000
Vapor Viscosity, PAS 20.000
25.000
0.0000
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
Tray
Tray
0.000
520.000
540.000
Vapor
560.000
580.000
600.000
620.000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0001
0.0001
0.0001
0.0002
0.0002
0.0002
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 0.0001 Vapor
Liquid Density, KG/M3
Liquid
Liquid Viscosity, PAS
Liquid
Figure 5.4 Density Profile of column T-1
Figure 5.5 Viscosity Profile of column T-1
Table 5.3 Summary of water wash column T-2 Column T-2 Summary Tray 1 2 3 4 5
38.2
Temp. C 38.2 38.2 38.2 38.2 38.2
38.2
Pressure BAR 7.92 7.92 7.92 7.92 7.92 Tem perature, C 38.2 38.2
Liquid 947.2 947.2 947.3 948.1
38.2
Net Flow Rates Vapor Feed KG-MOL/HR 375.0 0.0 0.0 0.0 0.0 576.0
Product
Duties M*KJ/HR
572.1
378.8
38.3
Vapor Rate, KG-MOL/HR
1
0.0
0.2
0.4
0.6
0.8
1.0
1
2
Tray
Tray
2
3
3
4 4
5 947.0 Vapor
5
Figure 5.6 Temp Profile of column T-2
947.2
947.4
947.6
947.8
948.0
Liquid Rate, KG-MOL/HR
Liquid
Figure 5.7 Rates Profile of column T-2
59
948.2
Vapor Density, KG/M3 0.200
0.400
0.600
Vapor Viscosity, PAS 0.800
1.000
0.0000
1
1
2
2
Tray
Tray
0.000
3
0.2000
0.4000
0.6000
0.8000
1.0000
0.0003
0.0003
0.0003
0.0003
3
4
4
5 603.700 603.800 603.900 604.000 604.100 604.200 604.300 604.400 Vapor
5 0.0003
0.0003
Vapor
Liquid Density, KG/M3
Liquid
Liquid Viscosity, PAS
Liquid
Figure 5.8 Density Profile of column T-2
Figure 5.9 Viscosity Profile of column T-2
Table 5.4 Summary of Methanol Recovery column T-3 Column T-3 Summary Tray 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
0.0
Temp. Pressure C BAR 30.0 1.03 74.1 1.38 76.5 1.40 81.7 1.42 92.8 1.44 104.6 1.46 108.8 1.48 109.9 1.50 110.4 1.51 110.8 1.53 111.6 1.55 112.2 1.57 112.8 1.59 113.3 1.61 113.7 1.63 114.2 1.65 114.6 1.67 114.9 1.69 115.3 1.71 115.7 1.73
20.0
40.0
Liquid
Tem perature, C 60.0 80.0
Net Flow Rates Vapor Feed KG-MOL/HR
96.9 105.9 103.7 99.4 95.8 95.9 96.2 96.3 96.4 483.1 483.6 484.0 484.4 484.8 485.1 485.5 485.8 486.1 486.4
100.0
101.3 110.3 108.1 103.8 100.2 100.3 100.6 100.7 100.7 108.7 109.2 109.6 110.0 110.4 110.7 111.0 111.3 111.6 111.9
120.0
Product 4.4
Duties M*KJ/HR -4.0730
374.4
4.5077
378.8
140.0
Vapor Rate, KG-MOL/HR
1
98.0
100.0
102.0
104.0
106.0
108.0
110.0
112.0
114.0
1
3
3 5
7
7
Tray
5
Tray
9 11
9 11 13 15
13
17
15
19
17
0.0
19
Vapor
100.0
200.0
300.0
400.0
500.0
600.0
Liquid Rate, KG-MOL/HR
Liquid
Figure 5.10 Temp Profile of column T-3
Figure 5.11 Rates Profile of column T-3
60
Vapor Density, KG/M3 0.500
1.000
Vapor Viscosity, PAS
1.500
2.000
0.0000
1
1
3
3
5
5
7
7
9 11
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
11
13
13
15
15
17
17
19 0.000
0.0000
9
Tray
Tray
0.000
19 200.000
Vapor
400.000
600.000
800.000
1000.000
0.0000
0.0001
Vapor
Liquid Density, KG/M3
Liquid
0.0002
0.0003
0.0004
0.0005
0.0006
Liquid Viscosity, PAS
Liquid
Figure 5.12 Density Profile of column T-3
Figure 5.13 Viscosity Profile of column T-3
Table 5.5 Overall Report Stream (Summary)
Name Description Phase Thermodynamic System Total Molar Rate Total Mass Rate Temperature Pressure Total Molecular Weight Total Specific Enthalpy Total Cp Total Molar Component Rates
UOM
kg-mol / hr kg / hr C bar kJ / kg J/kg-K kg-mol / hr
NC4 IC4 1BUTENE BTC2 BTT2 IBTE MTBE MEOH TBA H2O DIB Total Molar Component Fractions
NC4 IC4 1BUTENE BTC2 BTT2 IBTE MTBE MEOH TBA H2O DIB Vapor Molar Rate Vapor Mass Rate Vapor Molecular Weight Vapor Mole Fraction Vapor Specific Enthalpy Vapor CP Vapor Composition
NC4 IC4 1BUTENE BTC2 BTT2 IBTE MTBE MEOH TBA H2O DIB Liquid Molar Rate Liquid Mass Rate Liquid Molecular Weight Liquid Mole Fraction Liquid Specific Enthalpy Liquid CP Liquid Composition
NC4
10
11
12
13
14
10 10 Liquid S1 375.0209471 6756.40571 37.9999939 7.93 18.01607554 169.3191479 4366.688375
11 C4'S Liquid S1 572.136588 32942.70972 38.1693034 7.92 57.57840072 92.15106329 2503.383034
12 12 Liquid S1 378.87777 6886.158242 38.1787967 7.92 18.17514456 166.9490502 4341.893717
13 13 Mixed S1 378.87777 6886.158242 98.9999939 7.575 18.17514456 422.8609698 0
14 14 Mixed S1 378.87777 6886.158242 99.0980516 2.41 18.17514456 422.8609697 0
0 0 0 0 0 0 0 277.5 0 0 0
0 0 0 0 0 0 1.68576E-18 0.021275489 3.14002E-36 374.9996714 0
76.4962078 348.4874116 59.49979477 33.99945592 50.99964646 2.247034765 0.004041775 0.003515716 2.8833E-10 0.399479196 1.03273E-11
0.005569965 0.015603494 0.000171456 0.000241614 0.000192594 5.202E-06 2.16251E-05 4.255771777 2.73157E-13 374.6001924 1.28821E-23
0.005569965 0.015603494 0.000171456 0.000241614 0.000192594 5.202E-06 2.16251E-05 4.255771777 2.73157E-13 374.6001924 1.28821E-23
0.005569965 0.015603494 0.000171456 0.000241614 0.000192594 5.202E-06 2.16251E-05 4.255771777 2.73157E-13 374.6001924 1.28821E-23
0 0 0 0 0 0 0 1 0 0 0 N/A N/A N/A N/A N/A N/A
0 0 0 0 0 0 4.49511E-21 5.67315E-05 8.37292E-39 0.999943269 0 N/A N/A N/A N/A N/A N/A
0.133702702 0.60909828 0.103995787 0.059425418 0.089138936 0.003927445 7.06435E-06 6.14489E-06 5.03954E-13 0.000698223 1.80503E-14 N/A N/A N/A N/A N/A N/A
1.47012E-05 4.11835E-05 4.52538E-07 6.3771E-07 5.08327E-07 1.373E-08 5.70768E-08 0.011232572 7.20964E-16 0.988709874 3.40006E-26 N/A N/A N/A N/A N/A N/A
1.47012E-05 4.11835E-05 4.52538E-07 6.3771E-07 5.08327E-07 1.373E-08 5.70768E-08 0.011232572 7.20964E-16 0.988709874 3.40006E-26 0.000577937 0.030431601 52.65556269 1.52539E-06 629.8510681 2096.037884
1.47012E-05 4.11835E-05 4.52538E-07 6.3771E-07 5.08327E-07 1.373E-08 5.70768E-08 0.011232572 7.20964E-16 0.988709874 3.40006E-26 0.029560113 1.20838919 40.87904483 7.80202E-05 950.3195163 2012.784279
N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
0.13081499 0.711658135 0.007379891 0.004576501 0.006164664 0.000268759 8.803E-07 0.010672162 7.31467E-14 0.128464018 2.18059E-20
0.122759833 0.422089572 0.004578206 0.005051828 0.004781536 0.000144091 2.36786E-06 0.032527184 1.93873E-13 0.408065382 4.35717E-22
N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 277.5 8891.699411 32.04216003 1 23.45157451 2454.101878
N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 375.0209471 6756.40571 18.01607554 1 169.3191479 4366.688375
N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 572.136588 32942.70972 57.57840072 1 92.15106329 2503.383034
N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 378.87777 6886.158242 18.17514456 1 166.9490502 4341.893717
7.56028E-05 0.000411294 4.26511E-06 2.64493E-06 3.56279E-06 1.55326E-07 5.08758E-10 6.16784E-06 4.22742E-17 7.42441E-05 1.26024E-23 378.8771922 6886.127812 18.17509196 0.999998475 422.860055 4133.847462
0.003628795 0.012477015 0.000135332 0.000149333 0.000141343 4.25934E-06 6.99943E-08 0.000961507 5.7309E-15 0.012062459 1.28798E-23 378.84821 6884.949852 18.17337306 0.99992198 422.7683945 4134.720639
0
0
0.133702702
1.47012E-05
1.45017E-05
5.12388E-06
fraction
kg-mol / hr kg / hr fraction kJ / kg J/kg-K fraction
NC4 IC4 1BUTENE BTC2 BTT2 IBTE MTBE MEOH TBA H2O DIB Vapor Component Rate
1
1 MEOH FEED Liquid S1 277.5 8891.699411 15.9999939 16.2 32.04216003 23.45157451 2454.101878
kg-mol / hr
kg-mol / hr kg / hr fraction kJ / kg J/kg-K fraction
61
5.2
RESULTS DISCUSSION The results of this simulation shown in previous section indicate that the overall conversion of IBTE is 99.2%, the conversion can be calculated as following:
Conv. (%) =
(IBTE) inlet - (IBTE) outlet × 100 (IBTE) inlet
280.5 - 2.32 × 100 280.5 Conv. (%) = 99.2 % Conv. (%) =
Also with a selectivity to MTBE of 99.7%. In the reactive distillation column itself, 87.2% of the IBTE fed to the column is converted to MTBE. The MTBE product is 99.77% pure and needs no further purification. There are a number of factors that affect the overall conversion rate of IBTE. Some of these are:
•
MeOH to IBTE ratio.
•
Number of reaction trays.
•
Type of catalyst used.
•
Reflux ratio
Note, however, that while the IBTE conversion in the conversion reactors increases as the MEOH:IBTE ratio is increased, the overall IBTE conversion reaches a maximum, then decreases as the MEOH:IBTE ratio is increased. This is due to the fact that more MTBE product is carried upward through the column stripping section into the reaction trays. This promotes the reverse reaction of MTBE to methanol and IBTE, thus reducing the overall conversion of IBTE. 5.3
CASE STUDIES The validated MTBE primary reactor and reactive distillation column models were used to develop new process flow diagrams for the MTBE unit. Concerning the evaluation of plant performance, there are some parameters were taken into account for the plant optimization as follows: 62
o
MeOH to IBTE ratio.
o
Number of trays.
o
Location of feeds into RD column.
o
Reflux ratio of RD column.
o
Number of reaction trays.
Case Title: Case-1: Changing the Methanol/ Isobutylene Ratio Case Clarifications:
• • • • •
R1 = 0.90 ==> (MeOH)Feed= 252.5 kgmol/hr R2 = 0.95 ==> (MeOH)Feed= 266.5 kgmol/hr R3 = 0.98 ==> (MeOH)Feed= 277.5 kgmol/hr R4 = 1.0 ==> (MeOH)Feed= 280.5 kgmol/hr (Base case) R5 = 1.05 ==> (MeOH)Feed= 294.5 kgmol/hr
MeOH/IBTE Ratio Conversion(%) MTBE Purity(wt%)
0.900 0.950 87.500 92.300 87.900 92.700
0.980 96.970 99.600
1.000 99.200 99.700
1.050 96.970 99.600
Case Observations and Comments: - The overall conversion of IBTE into MTBE is increased when the Methanol/Isobutylene ratio increased but it dropped when the ratio exceed 1.0 as shown in the figure 5.14. - The purity of MTBE is increased when the Methanol/Isobutylene ratio increased as shown in the figure 5.14. The last two runs are showing almost stable purity. - The optimum Methanol/Isobutylene ratio is 1.0 as demonstrated in figure 5.14.
Figure 5.14 Results of changing the MeOH/IBTE ratio
63
Case Title: Case-2: Changing the Number of trays for RD column Case Clarifications:
• • • • • •
R1 = The RD column consists of 20 trays R2 = The RD column consists of 25 trays R3 = The RD column consists of 30 trays (Base case) R4 = The RD column consists of 32 trays R5 = The RD column consists of 35 trays R6 = The RD column consists of 40 trays
RD Trays Conversion(%) MTBE Purity(wt%)
20 97.00 97.20
25 98.00 99.20
30 99.20 99.70
32 99.20 99.70
35 99.20 99.80
40 99.20 99.80
Case Observations and Comments: - The overall conversion of IBTE to MTBE is increased when the number of trays increased as shown in the figure 5.15. - The purity of MTBE is increased when the number of trays increased as shown in the figure 5.15. The last three runs are showing almost stable purity. - The last two runs are almost comparable which indicates that the both parameters will get stable. - The optimum number of trays are 30 (The max limit to have the desired results).
Figure 5.15 Results of changing the number of trays of RD column
64
Case Title: Case-3: Changing the Location of feed stream into RD column Case Clarifications:
• • • • •
R1 = The feed stream entered the RD column at tray # 8 R2 = The feed stream entered the RD column at tray # 10 R3 = The feed stream entered the RD column at tray #12 R4 = The feed stream entered the RD column at tray #15 (Base case) R5 = The feed stream entered the RD column at tray #17
Note:- The Reaction zone applied in this case is from 11-13 trays of RD column Feed stream to tray Conversion(%) MTBE Purity(wt%)
8 94.50 94.50
10 95.30 95.80
12 97.91 98.40
15 99.20 99.70
17 99.20 99.70
Case Observations and Comments: - The overall conversion of IBTE to MTBE is increased when the location of feed stream changed to the next tray as shown in the figure 5.16. - The purity of MTBE is increased when the location of feed stream changed to the next tray as shown in the figure 5.16. - The last two runs are almost comparable which indicates that the both parameters will get stable. - The optimum location for the feed stream is tray 15 (The middle of the column).
Figure 5.16 Results of changing the location of feed stream into RD column
65
Case Title: Case-4: Changing the Reflux Ratio for RD column Case Clarifications:
• • • • •
R1 = The Reflux Ratio is equal to 0.8 R2 = The Reflux Ratio is equal to 0.9 R3 = The Reflux Ratio is equal to 1.0 R4 = The Reflux Ratio is equal to 1.1 (Base case) R5 = The Reflux Ratio is equal to 1.2
Reflux Ratio Conversion(%) MTBE Purity(wt%)
0.8 98.33 98.80
0.9 98.75 99.30
1.0 99.16 99.60
1.1 99.20 99.70
1.2 99.20 99.80
Case Observations and Comments: - The overall conversion of IBTE to MTBE is increased when the Reflux Ratio increased as shown in the figure 5.17. - The purity of MTBE is increased when the Reflux Ratio increased as shown in the figure 5.17. - The overall conversion of IBTE into MTBE is getting stable for the last three runs at the value 99.2 %.
Case-4: Changing the Reflux Ratio for RD column Conversion(%)
MTBE Puirty(wt%)
100.00 99.80 99.60 99.40 99.20 99.00 98.80 98.60 98.40 98.20 0.6
0.7
0.7
0.8
0.8
0.9
0.9
1.0
1.0
1.1
1.1
1.2
Reflux Ratio
Figure 5.17 Results of changing the Reflux Ratio
66
1.2
1.3
Case Title: Case-5: Changing the Location of reaction zone in RD column Case Clarifications:
• • • • •
R1 = The Reaction zone from 9-13 trays R2 = The Reaction zone from 10-13 trays R3 = The Reaction zone from 11-13 trays (Base case) R4= The Reaction zone from 12-13 trays R5 = The Reaction zone only tray # 13
Reaction zone Conversion(%) MTBE Purity(wt%)
9-13 99.20 99.80
10-13 99.20 99.80
11-13 99.20 99.70
12-13 99.17 99.70
13 98.50 99.20
Case Observations and Comments: - The overall conversion of IBTE to MTBE is decreased when the location of reaction zone get changed as shown in the figure 5.18. - The purity of MTBE is stable then decreased when the reaction zone get small (only one tray) as shown in the figure 5.18. - The optimum location for the reaction zone is tray 11-13 (The middle of the packing bed). Case-5: Changing the Location of reaction zone in RD column Conversion(%)
MTBE Puirty(wt%)
100.00 99.80 99.60 99.40 99.20 99.00 98.80 98.60 98.40 0
1
2
3
4
5
Reaction Zone
Figure 5.18 Results of changing the location of reaction zone
67
6
CHAPTER -6-
CONCLUSIONS AND RECOMMENDATIONS
68
6.1
CONCLUSIONS •
The combination of reaction and distillation helps in achieving products of higher purity and higher conversion of reactants as compared to old conventional processes.
•
As shown in this study it is possible to obtain an overall conversion of IBTE around 99.2% with purity of MTBE almost 99.7%.
•
The mathematical model developed has shown satisfactory results in simulating a reactive distillation column for the etherification of methanol and isobutylene to form MTBE.
•
The model and computational technique has also been successful in determining the effect of various design variables namely reflux ratio, feed plate location, reaction zone, Methanol/Isobutylene molar ratio.
•
The selectivity of IBTE to MTBE is 99.7%. In the reactive distillation column itself, 87.2% of the IBTE fed to the column is converted to MTBE. The MTBE product is 99.77% pure and needs no further purification.
•
The
overall
conversion
of
IBTE
to
MTBE
is
increased
when
the
Methanol/Isobutylene ratio increased but it dropped when the ratio exceed 1.0 as shown in the figure 5.14.
•
The purity of MTBE is increased when the Methanol/Isobutylene ratio increased as shown in the figure 5.14. The last two runs are showing almost stable purity.
•
The overall conversion of IBTE to MTBE is increased when the number of trays increased as shown in the figure 5.15.
•
The purity of MTBE is increased when the number of trays increased as shown in the figure 5.15. The last three runs are showing almost stable purity.
•
The optimum number of trays are 30 (The max limit to have the desired results).
•
The overall conversion of IBTE to MTBE is increased when the location of feed stream changed to the next tray as shown in the figure 5.16.
•
The purity of MTBE is increased when the location of feed stream changed to the next tray as shown in the figure 5.16. but the last two runs are almost comparable which indicates that the both parameters will get stable.
69
•
The optimum location for the feed stream is tray 15 (The middle of the column).
•
The overall conversion of IBTE to MTBE is increased when the Reflux Ratio increased as shown in the figure 5.17.
•
The purity of MTBE is increased when the Reflux Ratio increased as shown in the figure 5.17.
•
The overall conversion of IBTE to MTBE is decreased when the location of reaction zone get changed as shown in the figure 5.18.
•
The purity of MTBE is stable then decreased when the reaction zone get small (only one tray) as shown in the figure 5.18.
•
The optimum location for the reaction zone is tray 11-13 (The middle of the packing bed).
•
It is not recommended to exceed the MEOH/IBTE ratio more than 1.0 to achieve the desired conversion as well as the MTBE purity, this due to limitation observed from the simulation results.
•
It is not recommended to increase the number of trays more than 30 trays since the conversion of IBTE and purity of MTBE were not increased much when the trays increased up to 40. this will also maintain the capital cost.
•
It is recommended to keep the feed stream into RD column at stage no. 15 which represents the middle of the column.
•
It is recommended to keep the reflux ration equal to 1.1 since the conversion and purity of MTBE were no effected when it increased more than 1.1.
•
It is recommended to keep the location of the reaction zone in the space of tray 11-13 (The middle of the packing bed).
•
It is recommended to keep close monitoring system on the recycled methanol to the reactor to monitor the moisture content in the stream to avoid forming impurities such as t-butanol.
•
It is recommended to maintain a sufficient amount of C4s in the column feed to have smooth operation for the MTBE RD column. Also to form azeotropes with the methanol in the feed.
70
6.2
RECOMMENDATIONS •
It is recommended to perform an Economic Evaluation study for RD process against the conventional process.
•
It is recommended also, to perform a dynamic study.
•
It is recommended to validate the results by using another data or another simulation software.
•
It is recommended to perform an effective control study on control structures for RD column to have effective system to control the critical parameters.
•
It is recommended to study the possibility of having quality performance system for the feed and recycled streams to avoid forming undesired products.
•
It is recommended to use this model as a tool for design, optimization and control of Reactive Distillation processes.
71
BIBLIOGRAPHY
72
REFERENCES 1-
J. H. Gregor, S.T. Bakas, and M. A. Ulowetz, "Converting field butanes into MTBE", Seventy-First Annual Convention Gas Processors Association, Anaheim, California, 1992, sku P1992.36.
2-
Sandra Viviana, " Computer Simulation Studies for the production of 7-Tertadecene by Reactive Distillation", Mississippi State University , USA, 2 August 2003.
3-
D. N. Nakamura, “HP in Processing,” Hydrocarbon Processing, Vol. 77(1), 15, 1998.
4-
CHEM System Inc, "Methyl tertiary-Butyl Ether (MTBE)", PERP 94/95-4, 1996.
5-
M. F. Doherty, and M. F. Malone, "Conceptual Design of Distillation Systems", McGraw Hill, New York, 2001
6-
E. J. Chang and S. M. Leiby, “Ethers Help Gasoline Quality,” Hydrocarbon Processing, Vol. 71(2), 41–44, 1992.
7-
J. D. Seader, "Separation Process Principles", John Wiley and Sons, Inc., New York, 1998.
8-
J. L. DeGmo, V. N. Pdekar, and V. Pinjala, "Consider Reactive Distillation," Chem Eng. Prog. 88, 43-50, 1992.
9-
R. K. Toghiani, H. Toghiani, and G. Venkateswarlu, "Vapor-liquid Equilibria for methyl ter-butyl ether+methanol and ter-amyl methyl ether+methanol", Fluid Phase Equilibria 122, 157-168, 1996.
10- A. Tuchlenski, A. Beckmann, D. Reusch, R. Dussel, U. Weidlich, and R. Janowsky, "Reactive Distillation-Industrial Applications, Process Design and Scale-up", Chemical Engineering Science 56, 387-394, 2001. 11- Krishna, R. and Sie, S.T. "Strategies for Multiphase Reactor Selection", Chem.Engng Sci., 49, 4029-4065, 1994. 12- Agreda, V.H., Partin, L.R. and Heise, W.H. "High-purity methyl acetate via reactive distillation", Chem. Eng. Progress, 2, 40 – 46, 1990. 13- Doherty, M.F. and G. Buzad, "Reactive Distillation by Design", Chem. Eng. Res& . Design., Trans I.Chem.E., Part A, 70, 448-458, 1992. 14- Siirola, J.J., "An Industrial Perspective on Process Synthesis", A.I.Ch.E. Symposium Series No. 304, 91,222 – 233, 1995. 15- Stichlmair, J.G. and Frey, T., "Reactive distillation processes", Chemical and Engineering Technology, 22, 95-103, 1999. 16- Ciric, A.R., and D. Gu., "Synthesis of nonequilibrium reactive distillation by MINLP optimization", A.I.Ch.E.J., 40, 1479 – 1487,1994
73
17- Gonzalez, J. C., & Fair, J. R., "Preparation of tertiary amyl alcohol in a reactive distillation column.1 Reaction kinetics, chemical equilibrium, and mass transfer issues". Industrial and Engineering chemistry Research, 36, 3833-3844,1997. 18- Velo, E., Puigianer, L. & Recasens, F., "Inhibition by product in the liquid-phase hydration of iso-butane to tert-butyl alcohol – kinetics & equilibrium studies", Industrial and Engineering chemistry Research, 27, 2224-2231, 1988. 19- Shoemaker, J.D. and E.M. Jones, "Cumene by catalytic distillation",
Hydrocarbon
Processing, June 1987, 57 – 58, 1987. 20- Subawalla, H., Gonzalez, J.C., Seibert, A.F., and Fair, J.R., "Capacity and Efficiency of Reactive Distillation Bale Packing: Modeling and Experimental Validation", Ind. Eng. Chem. Res., 36, 3821-3832, 1997 21- Van Gulijk, C. , "Using computational fluid dynamics to calculate transversal dispersion in a structured packed bed", Comput. Chem. Eng., 22, S767-S770, 1998. 22- Higler, A., Krishna, R., Ellenberger, J. and Taylor, R., "Counter-current operation of a structured Catalytically Packed Bed Reactor: Liquid phase mixing and mass transfer", Chem. Eng. Sci., in press, 1999c. 23- Frey, S. J., Ozmen, S. M., Hamm, D. A., Pinjala, V., & DeGarmo, J. L. , "Advanced ether technology commercialized". Paper presented at the world conference on clean fuels and air quality control, Washington, DC, 1993. 24- G. Matthias, M. Ramona, Dragomir and J. Megan, "Conceptual design of reactive distillation columns using stage composition lines", Chemical Engineering Science 60, 5049-5068, 2005. 25- V. B. Shah, D. Bluck, J. W. Kovach III, "The sensitivity of the design and operability of the MTBE processes with respect to changes in reaction parameters and process configurations", Refining LNG and Petrochmasia 1994. 26- P. S. Bhanu, S. Ram, M.V Pavan, and K. Nitin, "Steady state analysis for reactive distillation control: An MTBE case study", Journal of Loss Prevention in the process Industries 18, 283-292, 2005. 27- C. A. Rivera and G. Johan, "Feasibility of Equilibrium controlled reactive distillation process: application of residue curve mapping", Computers and Chemical Engineering 28, 17-25, 2004. 28- Smith Jr., L.A. " U.S. Patent 4443559", Catalytic distillation structure, 1984. 29- Johnson, K.H. , "U.S. Patent 5189001", Catalytic distillation structure,1993. 30- Nelson, P. A., "Countercurrent equilibrium stage separation with reaction". American Institute of chemical Engineers Jornal, 17, 1043-1049, 1971.
74
31- Suzuki, I., Yagi, H., Komatsu, H., and Hirata, M., " Calculation of multicomponent distillation accompanied by chemical reaction", J. Chem. Eng. Jap., 4, 26-33, 1971. 32- Venkataraman, S., Chan, W. K., & Boston, J. F., "Reactive distillation using ASPEN PLUS". Chemical Engineering Process, 86(8), 45-54, 1990. 33- Jelink, J., & Halvacek, V., "Steady state Countercurrent equilibrium stage separation with
chemical
reaction
by
relaxation
method".
Chemical
Engineering
communications,2, 79-85, 1976. 34- Saito, S., Michishita, T., & Maeda, S. , " Separation of meta- and para-xylene mixture by distillation accompanied by chemical reaction". Journal of Chemical Engineering Japan, 4, 37-43, 1971. 35- Grosser, J.H., Doherty, M.F. and Malone, M.F., "Modeling of reactive distillation systems", Ind. Chem. Eng. Res., 26, 983-989, 1987. 36- Kamath R., Qi Z., Sundmacher K., Aghalayam P., Mahajani S., "Modelling and Optimization for Energy Saving and Pollution Reduction", 8th conference on Process Integration, The Italian Association of Chemical Engineering (AIDIC) 309, 2005. 37- Al-Jarallah, A. M., M. A. B. Siddiqui and A. K. K. Lee, “Kinetics of Methyl Tertiary Butyl Ether Synthesis Catalyzed by Ion Exchange Resin”, Canadian J. Chem. Eng. , 66, 802-7, 1988. 38- A. Abufares, and P.L. Douglas, "Mathematical Modeling and Simulation of an MTBE Catalytic Distillation Process using Speedup and Aspenplus", Chemical Engineering Research and Design 79 (A1), 3-12, 1995. 39- J. M. Bader, S. Guesneux, Axens, "Increase MTBE plant productivity", Hydrocarbon Processing, October 2005. 40- M. H. M. Reis; L. F. S. Mascolo; M. R. Wolf-Maciel, " Development of a robust algorithm to compute reactive azeotropes ", Brazilian Journal of Chemical Engineering, vol.23 no.3 , July/Sept. 2006.
75
APPENDIX (A) REACTION MODEL CODE
76
APPENDIX (A) : MTBE Reaction Model Code In this section the entire code of reaction kinetics will be shown below along with a description of purpose. The first portion of the defines real and integer variables REAL KS , KA , KC , KALJ , KREH1 , KREH2 , KIZQ , KEQREF INTEGER IBTE , MEOH , MTBE
Next, the indices for the components are set, a value is given for the gas constant in J/gm-mole K, and the basis for the temperature values in the procedure is set to an absolute basis. In addition, the temperature and composition rate derivatives are initially set equal to zero. $ $ $ $ $ $ $ $ $ $ $ $
INITIALIZE DATA: SET INDEXES FOR COMPONENTS DEFINE GAS CONSTANT IN JOULES/GM-MOLE K NOTE: R COULD HAVE BEEN RETRIEVED IN INPUT UNITS BY R=RGAS. HOWEVER, SINCE THE REACTION BASIS WON'T CHANGE, AND RGAS WILL CHANGE WITH THE DEFAULT UNITS, THIS ELIMINATES ONE POSSIBLE SOURCE OF ERROR. INITIALIZE THE LOCAL VARIABLE TK TO THE ABSOLUTE TEMPERATURE. NOTE: THE TEMPERATURE BASIS FOR THE FLOWSHET MUST BE C OR K. SET TEMPERATURE AND COMPOSITION DERIVATIVES TO ZERO. IBTE MTBE MEOH
= 6 = 7 = 8
R
= 8.314
TK
= RTABS
$ $ $
1000
DO 1000 I1 = 1,NOR DRDT(I1) = 0.0 DO 1000 I2 = 1,NOC DRDX(I2,I1) = 0.0
The surface reaction rate constant, ks, and the equilibrium adsorption constants, K , A
and K , are calculated using the expressions given previously as (1a), (1b), and (1c). B
$ $ $ $ $ $ $
CALCULATE THE SURFACE REACTION RATE CONSTANT, KS, AND THE EQUILIBRIUM ADSORPTION CONSTANTS KA AND KB. THE ACTIVATION ENERGY IS IN J/GM-MOLE. UNITS: KS - (GM-MOLE / GM CATALYST)**1.5 /HOUR KA - GM-CATALYST / MOLE KC - GM-CATALYST / MOLE KS KA KC
= 1.2E+13*EXP(-87900.0/(R*TK)) = 5.1E-13*EXP( 97500.0/(R*TK)) = 1.6E-16*EXP(119000.0/(R*TK))
77
Next, the derivatives of these constants are computed and are used later on in calculating the rate derivatives. $ DKSDT = DKADT = DKCDT =
KS KA KC
* 87900.0 / R / (TK*TK) * (-1.0) * 97500.0 / R / (TK*TK) * (-1.0) * 119000.0 / R / (TK*TK)
Then the bulk concentration of components A, B, and °C per gram of catalyst (RHOA, RHOB, and RHOC) are determined from the liquid mole fractions of the components (XLIQ), the density of the liquid, and the catalyst loading (GCAT) in g/l. Note that the liquid density, DENS, obtained directly from PRO/II using the predefined 3
variables, RLMRAT and RLVRAT, is in the user-specified units of kg-moles/m (SI units). Our basis for calculations is gm-moles/l and the conversion factor between 3
these kg-moles/m and gm-moles/l is 1.0. Also, note that the value of GCAT used here is 12.4 g/l. This value is used because it is the catalyst loading at which data for the AlJarallah rate equation was collected. $ ---$ $ $ $ $
CALCULATE THE EQUILIBRIUM CONSTANT. UNITS - (GM-MOLES/GM-CATALYST)/HOUR PHASE - LIQUID PHASE REACTION CATALYST - ION EXCHANGE RESIN AMBERLYST 15, THE EQUILIBRIUM SHOULD BE INDEPENDANT OF THE CATALYST
Expressions for the equilibrium constant and its derivative as functions of temperature are provided based on equilibrium data published by Al-Jarallah et al. $ KALJ
= EXP(-17.31715+(7196.776/TK))
$ DKALJDT =
-
KALJ
* 7196.776 / (TK*TK)
Then the reaction rate and rate derivatives with respect to temperature and composition are determined. $ $ $ $ $ $ $ $ $ $ $
BULK CONCENRATIONS OF COMPONENTS PER GRAM OF CATALYST, XLCONC IS IN MOLES/FLOW VOLUME. XLCONC WILL BE PASSED TO THE PROCEDURE IN USER INPUT UNITS. INTERNALLY TO PRO/II, IT IS IN SI UNITS (KG-MOLE/CUBIC METER). THE BASIS FOR THESE REACTION EQUATIONS IS GM-MOLES/LITER. THE CONVERSION FACTOR FROM INPUT UNITS OF KG-MOLES/CUBIC METER TO THE REACTION BASIS OF GM-MOLES/LITER IS ONE. THEREFORE, XLCONC CAN BE USED WITH NO CONVERSION. RHOA=(XLCONC(MEOH)/GCAT) |-THIS SHOULD BE EQUIVALENT TO BELOW. RHOB=(XLCONC(IBTE)/GCAT) | IT HAS BEEN WRITTEN EXPLICITLY BELOW RHOC=(XLCONC(MTBE)/GCAT) | TO MAKE IT OBVIOUS HOW TO DO THE ANALYTICAL DERIVATIVES.
78
$ $
CALCULATE DENSITY IN MOLES / VOLUME
$ ---$ ---$ $ $
GCATX = 12.4 DENS=RLMRAT/RLVRAT RHOA=(XLIQ(MEOH)*DENS/GCATX) RHOB=(XLIQ(IBTE)*DENS/GCATX) RHOC=(XLIQ(MTBE)*DENS/GCATX) CALCULATE REACTION RATE AND DERIVATIVES BY TERMS UNITS - RATE - GRAM-MOLE / GRAM CATALYST / HR. DENOMINATOR & DERIVATIVES. RDEN DRDDT DRDDME DRDDIB DRDDMT
$ $
0.0 0.0
+ ( KC*RHOC ) + RHOC*DKCDT
0.0 KC/GCATX*DENS
FIRST FACTOR IN RATE EQUATION. FACT1 = KS *KA/RDEN DFAC1DT = DKSDT*KA/RDEN + KS*DKADT/RDEN - KS*KA/RDEN**2 * DRDDT
$ $
$ $
--
$
--
$
= 1.0 + ( KA*RHOA ) + = RHOA*DKADT + = KA/GCATX*DENS = =
SECOND FACTOR IN RATE EQUATION. FACT2 = RHOA*RHOB**0.5 - RHOC**1.5/KALJ DFAC2DT = 0.0 + RHOC**1.5/KALJ**2 * DKALJDT COMBINE TERMS TO CALCULATE RATE AND DERIVATIVES. RATE EQUATION (RATE PER ONE GRAM OF CATALYSIS). RATE = FACT1 * FACT2 RATE TEMPERATURE DERIVATIVE. DRDT(1) = DFAC1DT * FACT2 + FACT1 * DFAC2DT
--
RATE COMPOSITION DERIVATIVES. DRDX(MEOH,1) = -KS*KA/RDEN**2 + FACT1 DRDX(IBTE,1) = -KS*KA/RDEN**2 + FACT1 DRDX(MTBE,1) = -KS*KA/RDEN**2 FACT1
&
* DRDDME * FACT2 & * (RHOB**0.5/GCATX*DENS) * DRDDIB * FACT2 & * (RHOA/2.0/RHOB**0.5/GCATX*DENS) * DRDDMT * FACT2 & * (1.5* RHOC**0.5/GCATX/KALJ*DENS)
It is important to note, however, that the rate and rate derivatives calculated above are computed on a basis of 1 gram of catalyst. The reactive distillation algorithm requires that these values (RRATES, DRDT, and DRDX) be supplied on a unit reaction volume basis. Therefore, the rate and rate derivatives are multiplied by the grams of catalyst per unit volume, GCAT. $ ---$ ---$ ---$
CONVERT RATE EQUATION AND DERIVATIVES TO A STRAIGHT VOLUME BASIS BY MULTIPLING THE BASE RATE BY THE GRAMS OF CATALYST/UNIT VOLUME. THE RATE IS RETURNED IN INPUT UNITS, KG-MOLES/CUBIC METER/HOUR. RRATES(1) DRDT(1) DRDX(MEOH,1) DRDX(IBTE,1) DRDX(MTBE,1)
= = = = =
GCAT GCAT GCAT GCAT GCAT
* * * * *
RXFACT RXFACT RXFACT RXFACT RXFACT
* * * * *
RETURN
79
RATE DRDT(1) DRDX(MEOH,1) DRDX(IBTE,1) DRDX(MTBE,1)
APPENDIX (B) MTBE PROCESS SIMULATION PROCEDURE USING PRO/II
80
APPENDIX (B) : MTBE Process Simulation Procedure using PRO/II
Process Simulation General Data SI units are used in this simulation. The total calculation sequence is specified. The calculator CAL0 is processed before the MTBE column in order to set the reaction factor equal to 1.0 on the first pass through the flowsheet.
In order to check the overall material balance, PRO/II is instructed to print out an overall flowsheet mass balance in the Miscellaneous Report Options menu.
81
Component Data All the components in the simulation are available in the PRO/II databank.
Thermodynamic Data The VLE fractionators are simulated well with PRO/II's modified Soave-Redlich-Kwong (SRKM) equation of state method. For this method, PRO/II contains extensive, built-in databanks that encompass binary interaction parameter data for the majority of component pairs present in this simulation. In this case, however, binary interaction data (kijs) are directly supplied for 8 component pairs to improve the accuracy of the separations in the columns. The binary interaction parameters are listed below in Table 1. The thermodynamic set used for VLE thermodynamics is referred to as SRKM_VLE. Table 1. Binary Interaction Parameters for SRKM_VLE Comp I
NC4
1BUTENE
BTC2
BTT2
IBTE
MEOH
TBA
MTBE
Comp j
TBA
MEOH
MEOH
MEOH
MEOH
TBA
H20
DIB
kija
0.0469
0.136
0.136
0.136
0.13553
-0.07397
-0.145
0.05785
kjia
0.1260
-0.0323
-0.0323
-0.0323
-0.0322
-0.05522
-0.253
-0.0093
kijb
0.0
0.0
0.0
0.0
0.0
0.0
0.0
-10.144
kjib
0.0
0.0
0.0
0.0
0.0
0.0
0.0
6.17
UOM
K
K
K
K
K
K
K
K
Transport properties are needed in order to use the rigorous heat exchanger model in the MEOH recovery section. Transport property calculations are set to pure-component
82
averages by default and can be modified by modifying the thermodynamic method of choice. Modification is not required in this simulation, as pure-component averages should be used. The liquid extraction unit is simulated using the SRKM method for VLLE thermodynamics with binary interaction data again supplied as part of the input. The thermodynamic set used for VLE thermodynamics is referred to as SRKM_VLLE.
Note that the L1KEY component (i.e.,
the predominant component in the L1 liquid phase) is specified as n-butane. The L2KEY component is specified to be water. Explicitly specifying the key components eliminates the need for PRO/II to find an appropriate immiscible pair, reducing the computation time. All the azeotropes are properly predicted.
Stream Data Feed Stream The mixed C4 feed stream, and the methanol feed stream are specified in the normal manner, using the compositions and stream conditions given in Table 2.
Recycle Stream The composition of the recycle methanol-water stream from the MEOH recovery section is estimated initially for the first run through the flowsheet (see Table 2).
83
The amount of wash water in stream 10 (the feed to column T-2) is provided. The temperature and pressure of the cooling water stream (CW) for the condenser for column T-3 is provided, along with an estimate of the flowrate. An estimated value is given for the flowrate of the make-up water stream, MKUP.
Unit Operations MTBE Reaction The MTBE reaction section of the plant is shown in Figure 2 below.
84
Figure 2. MTBE Reaction Section With reference to the previous figure, mixed C4s (stream 2) are combined with fresh methanol (stream 1) and recycle methanol (stream 20) and pre-heated in a heat exchanger (HX-1) to 43.5 C. The heated feed (stream 3) is then sent to a conversion reactor (RX-1) which is maintained at 55 °C by circulating a coolant. A pressure drop of 69 KPa through the reactor is also specified.
Since this is a conversion reactor. The three reactions defined in Table 3 take place in this reactor at the specified conversion levels. The extent of reaction are defined based on the values in Table 3.
85
The stoichiometries of the major and minor reactions in the MTBE process are provided in the reaction definitions data entry window.
MTBE Distillation and Recovery The MTBE distillation and recovery section of the plant is shown in Figure 3 below.
FIGURE 3. MTBE Distillation and Recovery Section
86
The reactor product (stream 4) exchanges heat with the MTBE column bottoms product in exchanger HX-2. Normally, this would create a thermal calculation loop. However, since the temperature of stream 5 is known, this process is modeled by two separate heat exchangers, HX-2A and HX-2B as shown in Figure 3. Stream 4 from the reactors is heated to 72 °C in HX-2A to produce stream 5. The product of column T-1, stream 7, is cooled in exchanger HX2-B to produce the MTBE product stream 8. The duty of exchanger HX-2B is defined to be equal to the duty in HX-2A. This approach avoids an unnecessary calculation loop since the temperature of stream 5 is fixed at 72 °C.
The heated stream 5 is fed to tray 15 of the 30 tray MTBE column (T-1). The MTBE column is simulated with the CHEMDIST algorithm using the SIMPLE initial estimate generator (IEG). A top pressure of 621 KPa and a column pressure drop of 76.5 KPa are given. The condenser is operated at a fixed temperature of 43.5 °C and pressure of 621 KPa. The control specifications are a bottoms flowrate of 278 kgmoles/hr and a reflux ratio of 1.1. The condenser and reboiler duties are varied to achieve these specifications.
87
The next step is to provide all the information required for specifying the reaction trays in the distillation column.
Implementing Reactive Distillation You can visualize the reaction zone of a distillation column as a series of boiling pot reactors. On each reaction tray sits a bed of solid catalyst. Each tray is connected to the next in the forward direction (down the column) by the flow of liquid from one tray to the next, and in the reverse direction by the vapor flow moving up from one tray to the previous tray. For the reactive distillation process, the reaction zone (trays 8 through 13) is specified in the column data entry window.
88
Note that the liquid volume of each of the reaction trays is also specified using the Reaction Volumes button in the Options section of the data entry window, and that the concentration of the dry catalyst (GCAT, in g/l) is specified using a DEFINE statement specified using the Subroutine/Procedure Data button in the Options section of the data entry window.
89
A value of 360 g/l is given for GCAT to represent commercial catalyst loadings (corresponding to a wet catalyst density of 770 g/l at 53% moisture content — see Table 1). The reaction factor, RXFACT, is used to demonstrate how the reaction rate in the simulation model can be varied to match data from an actual plant. For this case, RXFACT is set equal to 1.0, indicating that the reaction rate has not been adjusted.
MTBE Kinetic Model The algorithm used for the reactive distillation column model is a Newtonian-based algorithm. Therefore, in order to accurately model the MTBE reaction in the distillation column, we need to determine not only the reaction rate of the reaction, but also the temperature and composition derivatives of the rate. These derivatives may be generated numerically by an estimation method, or analytically by an expression based on the reaction rate equation. PRO/II automatically generates numerical estimates for the reaction rate derivatives. In many cases, this is sufficient. However, certain reactions require the use of more accurate analytical derivatives that provide better solutions. Reactions requiring these more accutate analytical derivatives include: reversible reactions, exothermic reactions, and/or reactions where the equilibrium is sensitive to temperature . The MTBE reactions fit all three of the reaction types, requiring the use of analytical derivatives. The reaction rate expression and its analytical derivatives can be easily and readily entered by the user in the Procedure Data category of input. The MTBE reaction rate expression used in this simulation model is based on the rate expression described in a paper by Al-Jarallah. In this case, we will detail how to enter the reaction rate and the associated
90
analytical derivatives for the MTBE reaction. First, in the Reaction Data category of input, the stoichiometry of the forward reaction is given (IBTE + MEOH = MTBE). The kinetic data will be provided later on in the Procedure Data category of input using FORTRAN-like language as the procedure named ALJD.
The reaction rate equation described by Al-Jarallah takes into account the forward and the reverse reaction. We have modified Al-Jarallah's rate equation for this case to simulate the effect of catalyst loading on the reaction rate. This was achieved by removing the catalyst terms from the concentration terms. The modified reaction rate is given by:
⎡ ⎤ C Aa C Bb − C Cc / k rs = k s k A × ⎢ a +b ⎥ ⎢⎣ (1 + K A C A + K B C B ) ⎥⎦ where: 13
k = surface reaction rate constant = 1.2x10 exp(- 87900/RT) in (gmole/g catalyst) s
(1a) K = equilibrium adsorption constant = 5.1x10 a
(1b) K
C
-13
-16
= equilibrium adsorption constant = 1.6x10
exp(97500/RT) in g catalyst/gmole exp(119000/RT) in g catalyst/gmole
(1c) K = equilibrium constant eq
C = IBTE concentration in mole/l A
C = MEOH concentration in mole/l and C = MTBE concentration in mole/l B
C
Kinetic Data The procedure data used in the reactive distillation column model is entered in the Kinetic Procedure data entry window shown below.
91
Once the column is converged, the top and bottom product compositions are known. Exchanger (HX-2B) is now simulated for heat exchange between the column feed (see HX2A) and the bottom product (stream 7). The duty in this exchanger is set equal to the duty in exchanger HX-2A. The cooled hot side fluid is the MTBE product (stream8). Pump P-1 pumps the liquid distillate (stream 6) at a pressure of 827 KPa to the methanol recovery section. A calculator (CONVERSION) is set up to compute the conversions of IBTE and MEOH to MTBE in the reactive distillation column itself.
92
Methanol Recovery The methanol recovery section of the process is shown in Figure 4.
FIGURE 4. MTBE Distillation and Recovery Section The methanol-C4s azeotrope (stream 6P) is delivered by pump P-1 to heat exchanger HX-3 where it is cooled to 38 °C against cooling water (CW). The exchanger also calculates the utility (CW) requirement given a CW delivery temperature of 16 °C and a return temperature of 32 °C. The cooled process stream is fed to the bottom of the water wash column (T-2). Column T-2 is simulated as a liquid-liquid extractor with 5 theoretical trays. Recirculating wash water is fed to the top of the column. A top pressure specification of 792 KPa is given. This column uses the VLLE SRK thermodynamic set (SRKM_VLLE) defined previously in the Thermodynamic Data Category of the input file. The raffinate leaves the top of the column (stream 11) and contains the unreacted and non- reactive C4s. The extract phase (stream 12) exits at the bottom. It enters the cold side (HX4A) of the feed-bottoms heat exchanger where it is warmed to 99 °C against the recycle
93
wash water (stream 21) which in turn is cooled (in unit HX-4B described later on).
Valve V-1 drops the pressure of the heated methanol-water stream (13) to 241 KPa generating a mixed phase stream (14) which is adiabatically flashed in unit D-1. The vapor phase (stream 15) containing the dissolved hydrocarbons which have been released is vented as a flare gas; the liquid phase (stream 16) is pumped (P-2) to the methanol column to recover methanol. The methanol column (T-3) is simulated with 20 theoretical trays. The feed (stream 17) enters on tray 10. The column top pressure is 138 KPa; the pressure drop through the column is 34.5 KPa. A Subcooled, Fixed Temperature condenser type operating at 30 °C and 103.5 KPa is specified. The separation of methanol from water is readily solved using the I/O algorithm and conventional IEG. The performance specifications are 99.5% recovery of methanol in the overhead product and 99.95% recovery of water in the bottoms product. Tray rating calculations are done for this column for 610 mm diameter sieve trays throughout the column.
94
A calculator (CAL1) computes the total loss of water as a result of carry over with the C4s (stream 11), the vent gas (stream 15) and by consumption in the reactor. This total quantity is the amount of make-up water required. The flowrate of the make-up water stream (MKUP) is established through a procedure call to the PRO/II stream function SRXSTR.
95
Pump P-4 pumps the recovered wash water from the methanol column bottoms combined with make-up water as stream 21 to heat exchanger HX4B. This unit represents the hot side of the exchanger HX-4 (see HX-4A described previously) and calculates the exit temperature for stream 22. Trim cooler (HX-5) further cools the wash water (stream 22) to the desired temperature of 38 °C before it (stream 10) goes back to the water wash column. At this stage, the first recycle loop between unit T-2 (water wash column) and HX-5 (trim cooler) is closed. The second recycle loop between unit HX-1 (feed heater) and P-3 (recycle pump) is closed when the pump P-3 recycles the overhead (stream 19) from the top of the methanol column (T-3) to the reactor section. Then, as an illustrative example, a rigorous heat exchanger (RC-1) is used to rigorously rate the methanol column condenser. This rigorous heat exchanger is modeled as an attached heat exchanger to column T-3. This unit takes as its input the exchanger's mechanical data such as shell and tube dimensions, tube layout pattern, the baffle cut and 2
shell and tube side nozzle sizes. A fouling factor of 0.00035 m -K/kW is used for the condenser cooling water side. The ZONES option is selected to determine where phase changes occur in the exchanger. An extended data sheet is printed in the output.
96
97
APPENDIX (C) PROCESS INPUT FILE USED IN PRO/II
98
Appendix (C): Process Input-File used in PRO/II $ Generated by PRO/II Keyword Generation System $ Generated on: Thu Jan 24 01:49:16 2008 TITLE PROJECT=MasterThises, PROBLEM=MTBE PLANT, USER=Fahad Harthi, & DATE=Dec 2007 PRINT INPUT=ALL, STREAM=ALL, RATE=M, MBALANCE DIMENSION SI, TEMP=C, PRES=BAR SEQUENCE DEFINED=CONVERSION,HX-1,RX-1,HX-2A,CAL0,T-1,HX-2B,P-1, & HX-3,T-2,HX4A,V-1,D-1,P-2,T-3,CAL1,P-4,HX4B,HX-5,P-3,RC-1 COMPONENT DATA LIBID 1,NC4/2,IC4/3,1BUTENE/4,BTC2/5,BTT2/6,IBTE/7,MTBE/8,MEOH/ & 9,TBA/10,H2O/11,244TM1P,,DIB THERMODYNAMIC DATA METHOD SYSTEM=SRKM, SET=S1, DEFAULT KVAL(VLE) SRKM 1,9,0.046973,0.126027,0,0,0,0,1,1 SRKM 3,8,0.136,-0.0323,0,0,0,0,1,1 SRKM 4,8,0.136,-0.0323,0,0,0,0,1,1 SRKM 5,8,0.136,-0.0323,0,0,0,0,1,1 SRKM 6,8,0.135525,-0.032271,0,0,0,0,1,1 SRKM 8,9,-0.073971,-0.055222,0,0,0,0,1,1 SRKM 9,10,-0.145,-0.253,0,0,0,0,1,1 SRKM 7,11,0.05785,-0.0093,-10.144,6.17,0,0,1,1 METHOD SYSTEM(VLLE)=SRKM, L1KEY=1, L2KEY=10, SET=S2 KVAL(VLE) SRKM 1,9,0.046973,0.126027,0,0,0,0,1,1 SRKM 3,8,0.136,-0.0323,0,0,0,0,1,1 SRKM 4,8,0.136,-0.0323,0,0,0,0,1,1 SRKM 5,8,0.136,-0.0323,0,0,0,0,1,1 SRKM 6,8,0.135525,-0.032271,0,0,0,0,1,1 SRKM 8,9,-0.073971,-0.055222,0,0,0,0,1,1 SRKM 9,10,-0.145,-0.253,0,0,0,0,1,1 SRKM 7,11,0.05785,-0.0093,-10.144,6.17,0,0,1,1 STREAM DATA PROPERTY STREAM=1, TEMPERATURE=16, PRESSURE=16.2, PHASE=M, & COMPOSITION(M,KGM/H)=8,277.5 PROPERTY STREAM=2, TEMPERATURE=16, PRESSURE=16.2, PHASE=M, & RATE(M)=850, COMPOSITION(M)=1,9/2,41/3,7/4,4/5,6/6,33, & NORMALIZE PROPERTY STREAM=10, TEMPERATURE=38, PRESSURE=7.93, PHASE=M, & COMPOSITION(M,KGM/H)=10,375 PROPERTY STREAM=20, TEMPERATURE=44, PRESSURE=17.24, PHASE=M, & COMPOSITION(M,KGM/H)=8,4/10,0.3 PROPERTY STREAM=CW, TEMPERATURE=21, PRESSURE=6.9, PHASE=M, & RATE(LV)=175, COMPOSITION(M)=10,100 PROPERTY STREAM=MKUP, TEMPERATURE=38, PRESSURE=3.5, PHASE=M, & COMPOSITION(M,KGM/H)=10,500 NAME 1,MEOH FEED/2,OLEFINS/20,MEOH RECYC/MKUP,MKUP WATER/ & 6,T-1 OVHD/8,MTBE/11,C4'S/15,FLARE GAS RXDATA RXSET ID=ST1
99
REACTION ID=1 STOICHIOMETRY 6,-2/11,1 REACTION ID=2 STOICHIOMETRY 6,-1/9,1/10,-1 REACTION ID=3 STOICHIOMETRY 6,-1/7,1/8,-1 RXSET ID=ALJX REACTION ID=ALJ0 STOICHIOMETRY 6,-1/7,1/8,-1 PROCEDURE DATA PROCEDURE(KINETIC) ID=ALJD, NAME=MTBE Process PDATA GCAT,RXFACT CODE REAL KS , KA , KC , KALJ , KREH1 , KREH2 , KIZQ , KEQREF INTEGER IBTE , MEOH , MTBE $ $ INITIALIZE DATA: $ SET INDEXES FOR COMPONENTS $ DEFINE GAS CONSTANT IN JOULES/GM-MOLE K $ NOTE: R COULD HAVE BEEN RETRIEVED IN INPUT UNITS BY R=RGAS. $ HOWEVER, SINCE THE REACTION BASIS WON'T CHANGE, AND $ RGAS WILL CHANGE WITH THE DEFAULT UNITS, THIS $ ELIMINATES ONE POSSIBLE SOURCE OF ERROR.. $ INITIALIZE THE LOCAL VARIABLE TK TO THE ABSOLUTE TEMPERATURE. $ NOTE: THE TEMPERATURE BASIS FOR THE FLOWSHET MUST BE CENTIGRADE $ OR KELVIN. $ SET TEMPERATURE AND COMPOSITION DERIVATIVES TO ZERO. $ IBTE = 6 MTBE = 7 MEOH = 8 $ R = 8.314 $ TK = RTABS $ DO 1000 I1 = 1,NOR DRDT(I1) = 0.0 DO 1000 I2 = 1,NOC 1000 DRDX(I2,I1) = 0.0 $ $ CALCULATE THE SURFACE REACTION RATE CONSTANT, KS, AND THE $ EQUILIBRIUM ADSORPTION CONSTANTS KA AND KB. THE ACTIVATION $ ENERGY IS IN J/GM-MOLE. $ UNITS: KS - (GM-MOLE / GM CATALYST)**1.5 /HOUR $ KA - GM-CATALYST / MOLE $ KC - GM-CATALYST / MOLE $ KS = 1.2E+13*EXP(-87900.0/(R*TK)) KA = 5.1E-13*EXP( 97500.0/(R*TK)) KC = 1.6E-16*EXP(119000.0/(R*TK)) $
100
DKSDT = DKADT = DKCDT = $ $ ---$ $ $ $ $ $ $
KS KA KC
* 87900.0 / R / (TK*TK) * (-1.0) * 97500.0 / R / (TK*TK) * (-1.0) * 119000.0 / R / (TK*TK)
CALCULATE THE EQUILIBRIUM CONSTANT. UNITS - (GM-MOLES/GM-CATALYST)/HOUR PHASE - LIQUID PHASE REACTION CATALYST - ION EXCHANGE RESIN AMBERLYST 15, THE EQUILIBRIUM SHOULD BE INDEPENDANT OF THE CATALYST
KALJ
= EXP(-17.31715+(7196.776/TK))
$ DKALJDT = $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $
-
KALJ
* 7196.776 / (TK*TK)
BULK CONCENRATIONS OF COMPONENTS PER GRAM OF CATALYST, XLCONC IS IN MOLES/FLOW VOLUME. XLCONC WILL BE PASSED TO THE PROCEDURE IN USER INPUT UNITS. INTERNALLY TO PRO/II, IT IS IN SI UNITS (KG-MOLE/CUBIC METER). THE BASIS FOR THESE REACTION EQUATIONS IS GM-MOLES/LITER. THE CONVERSION FACTOR FROM INPUT UNITS OF KG-MOLES/CUBIC METER TO THE REACTION BASIS OF GM-MOLES/LITER IS ONE. THEREFORE, XLCONC CAN BE USED WITH NO CONVERSION.
RHOA=(XLCONC(MEOH)/GCAT) |-THIS SHOULD BE EQUIVALENT TO BELOW. RHOB=(XLCONC(IBTE)/GCAT) | IT HAS BEEN WRITTEN EXPLICITLY BELOW RHOC=(XLCONC(MTBE)/GCAT) | TO MAKE IT OBVIOUS HOW TO DO THE ANALYTICAL DERIVATIVES. CALCULATE DENSITY IN MOLES / VOLUME GCATX = 12.4 DENS=RLMRAT/RLVRAT RHOA=(XLIQ(MEOH)*DENS/GCATX) RHOB=(XLIQ(IBTE)*DENS/GCATX) RHOC=(XLIQ(MTBE)*DENS/GCATX)
$ $ ---$ ---$ $ $
CALCULATE REACTION RATE AND DERIVATIVES BY TERMS UNITS - RATE - GRAM-MOLE / GRAM CATALYST / HR. DENOMINATOR & DERIVATIVES. RDEN DRDDT DRDDME DRDDIB DRDDMT
$ $
= 1.0 + ( KA*RHOA ) + = RHOA*DKADT + = KA/GCATX*DENS = =
FIRST FACTOR IN RATE EQUATION.
101
0.0 0.0
+ ( KC*RHOC ) + RHOC*DKCDT
0.0 KC/GCATX*DENS
FACT1 = KS *KA/RDEN DFAC1DT = DKSDT*KA/RDEN + KS*DKADT/RDEN - KS*KA/RDEN**2 * DRDDT $ $
SECOND FACTOR IN RATE EQUATION. FACT2 = RHOA*RHOB**0.5 - RHOC**1.5/KALJ DFAC2DT = 0.0 + RHOC**1.5/KALJ**2 * DKALJDT
$ $ $
--
$ $
--
$
--
$ $ ---$ ---$ ---$
COMBINE TERMS TO CALCULATE RATE AND DERIVATIVES. RATE EQUATION (RATE PER ONE GRAM OF CATALYSIS). RATE = FACT1 * FACT2 RATE TEMPERATURE DERIVATIVE. DRDT(1) = DFAC1DT * + FACT1 * RATE COMPOSITION DERIVATIVES. DRDX(MEOH,1) = -KS*KA/RDEN**2 + FACT1 DRDX(IBTE,1) = -KS*KA/RDEN**2 + FACT1 DRDX(MTBE,1) = -KS*KA/RDEN**2 FACT1
FACT2 DFAC2DT
&
* DRDDME * FACT2 & * (RHOB**0.5/GCATX*DENS) * DRDDIB * FACT2 & * (RHOA/2.0/RHOB**0.5/GCATX*DENS) * DRDDMT * FACT2 & * (1.5* RHOC**0.5/GCATX/KALJ*DENS)
CONVERT RATE EQUATION AND DERIVATIVES TO A STRAIGHT VOLUME BASIS BY MULTIPLING THE BASE RATE BY THE GRAMS OF CATALYST/UNIT VOLUME. THE RATE IS RETURNED IN INPUT UNITS, KG-MOLES/CUBIC METER/HOUR. RRATES(1)
= GCAT * RXFACT * RATE
DRDT(1)
= GCAT * RXFACT * DRDT(1)
$ $ DRDX(MEOH,1) = GCAT * RXFACT * DRDX(MEOH,1) DRDX(IBTE,1) = GCAT * RXFACT * DRDX(IBTE,1) DRDX(MTBE,1) = GCAT * RXFACT * DRDX(MTBE,1) RETURN UNIT OPERATIONS CALCULATOR UID=CONVERSION, NAME=CONVERSION OF IBTE-MEOH TO MTBE RESULT 1,IN - MEOH/2,IN - IBTE/3,IN - MTBE/4,OUT - MEOH/ & 5,OUT - IBTE/6,OUT - MTBE/20,IBTE CONV/21,MEOH CONV DEFINE P(1) AS STREAM=4, RATE(KGM/H), COMP=8,WET DEFINE P(2) AS STREAM=4, RATE(KGM/H), COMP=6,WET DEFINE P(3) AS STREAM=4, RATE(KGM/H), COMP=7,WET DEFINE P(4) AS STREAM=6, RATE(KGM/H), COMP=8,WET DEFINE P(5) AS STREAM=6, RATE(KGM/H), COMP=6,WET DEFINE P(6) AS STREAM=6, RATE(KGM/H), COMP=7,WET DEFINE P(7) AS STREAM=7, RATE(KGM/H), COMP=8,WET DEFINE P(8) AS STREAM=7, RATE(KGM/H), COMP=6,WET DEFINE P(9) AS STREAM=7, RATE(KGM/H), COMP=7,WET PROCEDURE $ --LOAD RATES R( 1) = P( 1)
102
R( 2) = P( 2) R( 3) = P( 3) R( 4) = P( 4) + P( 7) R( 5) = P( 5) + P( 8) R( 6) = P( 6) + P( 9) $ --CALCULATE CONVERSION R(20) = ( R(2) - R(5) ) / R(2) R(21) = ( R(1) - R(4) ) / R(1) $ --DISPLAY RESULTS DISPLAY R( 1: 9 ) DISPLAY R( 20:21 ) RETURN HX UID=HX-1, NAME=FEED HEAT COLD FEED=1,2,20, L=3, DP=0.345 OPER CTEMP=43.5 CONREACTOR UID=RX-1, NAME=REACTORS FEED 3 PRODUCT L=4 OPERATION ISOTHERMAL, TEMPERATURE=55, DP=0.69 RXCALCULATION MODEL=STOIC RXSTOIC RXSET=ST1 REACTION 1 BASE COMPONENT=6 CONVERSION 0.0025 REACTION 2 BASE COMPONENT=10 CONVERSION 1 REACTION 3 BASE COMPONENT=8 CONVERSION 0.93 HX UID=HX-2A, NAME=FEED-BTMS-A COLD FEED=4, L=5, DP=0.345 OPER CTEMP=72 CALCULATOR UID=CAL0, NAME=COPY RXFACT PROCEDURE IF (R(2) .NE. 1.0) R(1) = 1.0 $Set RXFACT TO 1 on first call. R(2) = 1.0 RETURN COLUMN UID=T-1, NAME=MTBE COLUMN PARAMETER TRAY=30,CHEMDIST=35 FEED 5,15, SEPARATE PRODUCT OVHD(M)=6, BTMS(M)=7,280, SUPERSEDE=ON CONDENSER TYPE=TFIX, PRESSURE=6.21, TEMPERATURE=43.5 DUTY 1,1,,CONDENSER DUTY 2,30,,SIDEHC2 PSPEC PTOP=6.21, DPCOLUMN=0.765 PRINT COMPOSITION=M, PROPTABLE=ALL ESTIMATE MODEL=SIMPLE, RRATIO(L)=1.1 SPEC ID=COL1SPEC1, STREAM=7, RATE(KGM/H),TOTAL,WET, VALUE=278 SPEC ID=COL1SPEC2, RRATIO, PHASE=L, VALUE=1.1 DEFINE GCAT AS 360 DEFINE RXFACT AS 1
103
VARY DNAME=CONDENSER,SIDEHC2 PLOT LOG, XCOMPONENT=6,6/8,8/7,7/1,1 TSIZE SECTION(1)=2,29,VALVE, DMIN=381, FF=80, DPCALC=0 VLLECHECK CHECK=OFF LVOL 8,5/9,5/10,5/11,5/12,5/13,5 RXTRAY REFERENCE=ALJX, LOCAL=L_ALJX, KPROCEDURE=ALJD, 13 RXSET LOCAL=L_ALJX REACTION ID=ALJ0, COPTION=KINETICS KINETICS PEXP(C,KG,M3,KPA,HR) HX UID=HX-2B, NAME=FEED-BTMS-B HOT FEED=7, L=8, DP=0.345 DEFINE DUTY(KJ/HR) AS HX=HX-2A, DUTY(KJ/HR) PUMP UID=P-1, NAME=T-1 OVHD FEED 6 PRODUCT L=6P OPERATION EFF=65, PRESSURE=8.27 HX UID=HX-3, NAME=COOLER HOT FEED=6P, L=9, DP=0.345 UTILITY WATER, TIN=16, TEMPERATURE=32 CONFIGURE COUNTER OPER HTEMP=38 COLUMN UID=T-2, NAME=WATER WASH PARAMETER TRAY=5,LLEX=25 FEED 9,5/10,1 PRODUCT OVHD(L1,M)=11, BTMS(M)=12,185, SUPERSEDE=ON PSPEC PTOP=7.92 PRINT PROPTABLE=PART ESTIMATE MODEL=SIMPLE METHOD SET=S2 HX UID=HX4A, NAME=FEED-BTMS COLD FEED=12, L=13, DP=0.345 OPER CTEMP=99 VALVE UID=V-1, NAME=VALVE FEED 13 PRODUCT M=14 OPERATION PRESSURE=2.41 FLASH UID=D-1, NAME=SEPARATOR FEED 14 PRODUCT V=15, L=16 ADIABATIC PUMP UID=P-2, NAME=FEED PUMP FEED 16 PRODUCT L=17 OPERATION EFF=65, PRESSURE=6.9 COLUMN UID=T-3, NAME=MEOH COLUMN PARAMETER TRAY=20,IO=10 FEED 17,10 PRODUCT OVHD(M)=19, BTMS(M)=18,182, SUPERSEDE=ON CONDENSER TYPE=TFIX, PRESSURE=1.035, TEMPERATURE=30 DUTY 1,1,,CONDENSER DUTY 2,20,,SIDEHC2
104
TRAY=8, &
PSPEC PTOP=1.38, DPCOLUMN=0.345 PRINT PROPTABLE=PART ESTIMATE MODEL=CONVENTIONAL, RRATIO=10 SPEC ID=COL3SPEC1, STREAM=19, RATE(KGM/H), COMP=8,WET, DIVIDE, & STREAM=17, RATE(KGM/H), COMP=8,WET, VALUE=0.995 SPEC ID=COL3SPEC2, STREAM=18, RATE(KGM/H), COMP=10,WET, DIVIDE, & STREAM=17, RATE(KGM/H), COMP=10,WET, VALUE=0.9995 VARY DNAME=CONDENSER,SIDEHC2 TRATE SECTION(1)=2,19,SIEVE, PASSES=1, DIAMETER(TRAY)=610, & DIAMETER(SIEVEHOLE,IN)=0.5, WEIR=50.8, DCC=38.1 CALCULATOR UID=CAL1, NAME=MAKEUP SEQUENCE STREAM=MKUP DEFINE P(1) AS STREAM=11, RATE(KGM/H), COMP=10,WET DEFINE P(2) AS STREAM=15, RATE(KGM/H), COMP=10,WET DEFINE P(3) AS STREAM=19, RATE(KGM/H), COMP=10,WET PROCEDURE R(1) = P(1) + P(2) + P(3) CALL SRXSTR(SMR,R(1),MKUP) RETURN PUMP UID=P-4, NAME=WATER PUMP FEED 18,MKUP PRODUCT L=21 OPERATION EFF=65, PRESSURE=8.62 HX UID=HX4B, NAME=FEED-BOTS HOT FEED=21, L=22, DP=0.345 DEFINE DUTY(KJ/HR) AS HX=HX4A, DUTY(KJ/HR) HX UID=HX-5, NAME=COOLER HOT FEED=22, L=10, DP=0.345 OPER HTEMP=38 PUMP UID=P-3, NAME=RECYCLE PUMP FEED 19 PRODUCT L=20 OPERATION EFF=65, PRESSURE=17.24 HXRIG UID=RC-1, NAME=T-3 COND TYPE TEMA=AES TUBES FEED=CW, L=WOUT, LENGTH=5.75, OD=19, BWG=14, PASS=2, & PATTERN=90, PITCH=25.4, FOUL=0.00035, METHOD=S1 SHELL DPUNIT=0.49244, METHOD=S1, ID=381 BAFFLE CUT=0.18 SNOZZLE TYPE=CONV, ID=152,102 TNOZZLE ID=102,102 PRINT EXTENDED, ZONE ATTACH COLUMN=T-3, TYPE=CONDENSER RECYCLE DATA ACCELERATION TYPE=WEGSTEIN LOOP NUMBER=1, START=T-2, END=HX-5,WEGSTEIN LOOP NUMBER=2, START=HX-1, END=P-3,WEGSTEIN CASESTUDY OLDCASE=BASECASE, NEWCASE=CS1 PARAMETER ID=MTBE, COLUMN=T-1, SPEC(1),STEP(ACT)=0, CYCLES =1,1 RESULT ID=RESULT1, STREAM=15,FRACTION, COMP=1,WET END
105
