Process Intensification through Multifunctional reactors Sanjay Mahajani Sanjay Mahajani Professor of Chemical Engineering Indian Institute of Technology Bombay 20 th December, 2011.
Process Intensification through Multifunctional reactors
Sanjay MahajaniSanjay MahajaniProfessor of Chemical Engineering
Indian Institute of Technology Bombay
20th December, 2011.
What are multifunctional reactors ?What are multifunctional reactors ?
The term multifunctional reactor can be defined f fas a reaction equipment in which performance of reaction is synergetically enhanced by means of integrating one or more additional process functions
This additional function can be a separation step
RDResearchGroup
Multifunctional reactors:Combine reaction with separation
• Reactive Distillation• Reactive Chromatography• Membrane reactors•• …
Concept of Reactive DistillationConcept of Reactive Distillation
We can improve equilibrium conversionWe can improve equilibrium conversionby removing product in the reactorby removing product in the reactorby removing product in the reactorby removing product in the reactor..
(Le Châtelier’s principle)(Le Châtelier’s principle)
A + B ↔ C + DA + B ↔ C + D
If we remove C the equilibriumIf we remove C the equilibriumconversion to product D increasesconversion to product D increases
CD
xxxxK BAxxRD
ResearchGroup
If the volatilities are right, If the volatilities are right, we can remove a component by vaporizationwe can remove a component by vaporization
C (A, B, D)
A
B A + B ↔ C + DA + B ↔ C + D
A, B, D
Improved yieldImproved yieldRD
ResearchGroup Concept of Reactive Distillation
P id fl l f f d iP id fl l f f d iProvide reflux to prevent loss of feed in vaporProvide reflux to prevent loss of feed in vapor
CC
A, B, D
A
BB
A, B, D
We can strip the product and recycle the feeds We can strip the product and recycle the feeds p p y fp p y f(if the volatilities are right)(if the volatilities are right)
CC
A A, B, A
B
, ,D
A, BA, B
DD
… this is starting to look like a column !
DReactive distillation
RD Process for methyl acetate
methanol + Acetic Acid methyl acetate + water
AcOH
MeOAc
1
AcOHMeOHH2SO4
H2SO4
AcOHTo impurity removal columns
2MeoAc
MeOH
Return form impurityremoval columns
3 Water
MeOH
H2O
4Heavies
Water + H2SO4
2
RD performs difficult separations RD
ResearchGroup
Way back in 1869…
Cooler
NH3+CO2 for absorber
Cooler
Filter liquorLiquor preheater
Reactions takes place close to boiling
Strippper (NH4)2CO3 = 2NH3 + CO2 + H2O
close to boiling conditions
Prelimer
Lime milk
Design is based on the Concept of Reactive Distillation !!
Distillation Column
(NH4)2SO4 + CaO= 2NH3 + CaSO4 + H2O2NH4Cl+ CaO= 2NH3 + CaCl2 + H2O
Low Pressure Steam
Ammonia Recovery in Solvay ProcessRD
ResearchGroup Solvay Process
Publications on Reactive Distillation90
Publications on Reactive Distillation
70
80 Papers
Patents
50
60
70
40
50
20
30
0
10
01970 1975 1980 1985 1990 1995 2000
RDResearchGroup (Hiwale et al., Int. J. Chem React. Engg. 2004)
R ti Di till ti ibl b fitReactive Distillation: possible benefits
• Capital Savings• Improve Selectivity
• Reduce Raw Materials Usage • Reduce Byproducts Prevent Pollution
• Reduce Energy Use • Handle Difficult Separations• Handle Difficult Separations
– Avoid Separating Reactants – Eliminate/Reduce Solvents /
• “Beat” Low Equilibrium Constants • Heat integration benefits• Longer catalyst life
Constraints: Volatilities are not right; g ;Complexity in design and operation
Butyl acetate in RD
H+
Butyl acetate in RD
CH3COOH + C4H9OH C4H9COOCH3 + H2OH+
Formation of heterogeneous Minimum Boiling ternary Azeotrope
butanolrecyclerecycle
Acetic acid +
Butanol Pre-Reactor
water
waterbutanol
Acetic acid +Butanol
butanol+ water
Butyl acetate Butyl acetateButyl acetate
The energy consumption in the RD process is expected to be 40% less
Work Elements of RD Process Development
E i t f E i t VLE d tExperiments for reaction kinetics and regression
Experiments on VLE data generation and regression
Conceptual Design and Synthesis
Hardware selection
Steady State and Dynamic Simulation
Column Experiments
Optimizationand Control
Case StudyCase Study Synthesis of 2-Ethylhexyl Acetate
Reaction Scheme:
Acetic Acid + 2-ethyl hexanol 2 ethyl hexyl acetate + water
Applications:
Good solvent power Coatings additives in cleaners improves flow and filmGood solvent power, Coatings, additives in cleaners, improves flow and film formation
Alcohol is heavy and boiling temperatures would exceed thermal stability limit of catalyst
Conventional Process
Entrainer Based Reactive DistillationEntrainer Based Reactive Distillation EBRD
Entrainer is separated in l b kdecanter and recycled back.
AdvantagesE h d ti t Enhanced reaction rates
Overcomes limitations due to distillation boundaries
Helps in lowering down reboiler and column temperaturetemperature
Better protection against catalyst deactivation
Reduced catalyst loading
Th d i (Ph ilib i )Thermodynamics (Phase equilibrium)
UNIQUAC‐HOC selected forQthermodynamic model
Parameters for the thermodynamic model taken from Aspen Data Bank.from Aspen Data Bank.
VLE experiments conducted to pdetermine binary interaction parameters of component pairs for which information not pavailable in Aspen Data Bank.
Generated T-xy dataGenerated T xy data
Toluene ‐ 2‐ethylhexyl acetate
Acetic Acid – 2‐ethylhexyl acetate
Txy data continued…Txy data continued…
• toluene – 2‐ethylhexanol 2‐ethylhexanol – 2‐ethylhexyl acetateacetate
Regressed binary interaction parametersRegressed binary interaction parameters
Component i 2-EHOH TOLUENE AA TOLUENE
Component j 2-EHAC 2-EHOH 2-EHAC 2-EHAC
Temperature units K K K K
aij 1 81989793 2 90232722 4 03069111 -4 92152717aij 1.81989793 2.90232722 4.03069111 4.92152717
aji -0.617807244 -4.67110786 -9.88579185 5.38982778
bij -1236.45922 -956.324435 -1594.93966 2249.1795
bji 548.659258 1490.66902 3828.60687 -2461.82773
Azeotropes in presence of entrainer*Azeotropes in presence of entrainer
Components Water Toluene AA 2EHOH 2EHAc Boiling Pt. Azeotrope
t% (°C)wt% (°C)
Water-Toluene 19.55 80.45 --- --- --- 84.29 Heterogeneous
Water-2EHOH-2EHAc 75.22 --- --- 8.99 15.79 98.97 Heterogeneous
Water-2EHAc 75 75 --- --- --- 24 25 99 09 HeterogeneousWater 2EHAc 75.75 24.25 99.09 Heterogeneous
Water-2EHOH 81.84 --- --- 18.16 --- 99.17 Heterogeneous
T l AA 66 44 33 56 103 99 HToluene-AA --- 66.44 33.56 --- --- 103.99 Homogeneous
2EHOH-2EHAc --- --- --- 78.85 21.15 183.96 Homogeneous
* from Aspen Data Bank
Residue Curve Map (RCM) for the system: (a) 2-ethylhexanol toluene and water; (b) 2 ethyhexylethylhexanol, toluene and water; (b) 2-ethyhexyl
acetate, 2-ethylhexanol and water.
(a) (b)
Reaction kinetics
Conc. based modified LHHW model
2
2
2 22
( )(1 )
i f AcH EHOH b EHAc H Oii
cat m AcH w H O
k C C k C CdCVrM dt K C K C
22 2( )i f AcH EHOH b EHAc H Oik a a k a adxnr
Activity based modified LHHW model
)/exp( 00 RTEkk fff
2
2(1 )icat m AcH w H O
rM dt K a K a
Where,)/exp( 00 RTEkk bbb
Optimization problemActivity – UNIQUAC model
22
1 1 1
minijMNDyn Nmeans
ijk j ijk ikji j k
SSE W z t z
p pAspen custom modeler
Column Experimentsp
Total height = 3.18m Reactive section 1.38 m Reactive Packing :
K t k S fill d ithKatapak‐S filled with Amberlyst‐15 IER (NTSM = 3)
Non‐reactive Packing: HYFLUX (NTSM = 8)El i ll h d Electrically heated Reboiler
Comparison of experiment and simulation results for one of the representative run : Reboiler Duty=120W; Molar Feed Ratio = 1:1;representative run : Reboiler Duty 120W; Molar Feed Ratio 1:1;
AA +Tol at Stage No.9;2EHOH at Stage No. 12
19 19 19
15
17 AA-Sim
AA-Expt15
17Toluene-Sim
Toluene-Expt15
172EHOH-Sim
2EHOH-Expt
11
13
No.
11
13
No.
11
13
No.
7
9Sta
ge
7
9Sta
ge
7
9Sta
ge
5
7
5
7
5
7
1
3
0 0 2 0 4 0 6 0 8 1
1
3
0 0 2 0 4 0 6 0 8 11
3
0 0 2 0 4 0 6 0 8 1
0 0.2 0.4 0.6 0.8 1
Mole Fraction 0 0.2 0.4 0.6 0.8 1
Mole Fraction0 0.2 0.4 0.6 0.8 1
Mole Fraction
Comparison of experiment and simulation ltresults
19 19 19
15
172EHAc-Sim
2EHAc-Expt 15
17Water-Sim
Water-Expt 15
17 Simulation
Experiment
11
13
No.
11
13
No.
11
13
No.
7
9Sta
ge
7
9Sta
ge
7
9Sta
ge
5
7
5
7
5
7
1
3
0 0.2 0.4 0.6 0.8 11
3
0 0 2 0 4 0 6 0 8 11
3
80 110 140 170 200
0 0.2 0.4 0.6 0.8 1
Mole Fraction 0 0.2 0.4 0.6 0.8 1
Mole Fraction80 110 140 170 200
Temperature (C)
Effect of different parameters on conversion
EBRD configurationg
mol fraction 2EHOH negligible AA negligible
Decanter aqueous stream
organic stream as
fl
2EHAc negligibleWater 0.7003 Toluene 0.2997 2.1401 kmol/hr
streamreflux
2-ethylhexanol 1.5 kmol/hr
2EHOH negligible AA negligible 2EHAc negligible Water 0 9999
2EHOH negligible AA negligible 2EHAc negligible Water 0 0029
acetic acid 1.5 kmol/hr
Water 0.9999Toluene negligible 1.4971 kmol/hr
Water 0.0029Toluene 0.9971 0.6430 kmol/hr
entrainer makeup0.001 kmol/hr
2EHOH 2.16e-3 AA 2.08e-3
final product
2EHAc 0.9952 Water negligible Toluene negligible 1.5039 kmol/hr
Conventional vs Proposed RD processConventional vs Proposed RD process
Conventional Process
Comparison of proposed process with ti lconventional process
Proposed RD Process Conventional Process
• Steam consumption per kg product =0.43
• Steam consumption per kg product =1.46
• Single RD column is sufficient
• Additional separation columns are needed
• Easy operation • Catalyst separation required
Diisobutene in RD
Parameter Value
Column diameter, ID, cm 5
IB + IBDIBDIB + IBTIB
•Conventional reactors need selectivity enhancer•RD works without enhancers•RD performs much better than
Column height, m 3
Height of reactive section, m 1
Total catalyst loading, gm 30
Height of non-reactive section, m 2
Number of theoretical stages in the 3 DIB + IBTIBTIB + IBTEBIB + 2-Buc-DIM1 Bu2 Bu
pconventional reacting systems (selectivity: 80% in lab)
Talwalkar et al., IECR, 2005, 2007Kamath et al. IECR, 2005, 2006
Number of theoretical stages in the reactive zone
3
Number of non reactive stages 16
Operating pressure, atm 10
Feed pump rating, lit/hrs 0 to 3
Reboiler holdup lit 2 1-Bu2-BuReboiler holdup, lit 2
Aldol Condensation of Acetone in RDAldol Condensation of Acetone in RD
7
8
0.8
1
ty
0.02 kg/m of catalyst conc.0.04 Kg/m of catalyst conc.Batch reactor
5
6
ber Reactive
stages0.4
0.6
DAA
sele
ctiv
it
Simulation
4
5
Stag
e nu
m
AcetoneDAA
0
0.2
0 0.1 0.2 0.3
C i
D
5m
2
3
S DAAMOWaterSimulation
Conversion
It is for the first time selectivity of
10 0.2 0.4 0.6 0.8 1
Mole fraction
the order of 50-60% with acid catalyst is reported for this reaction.
Mole fraction
3kDAA MO + H O 3 DAAk ar r
2Ac 1 Ac 2 DAAr k a k a
Acetone DAA
Concentrations of DAA and MO are very low on reactive stages
2DAA MO + H O DAA Ac 2
w w
r r1 k a
Reactions studied in IIT-B
• Esterifications:Butyl acetate, 2-ehtyl hexyl acetate, cellosolve acetate, triacetin, ethylene glycol diacetate, methyl acetate (recovery of acetic y g y , y ( yacid), lactate esters
• Acetalization:Acetalization: 2-methyl pentenone, glycerol acetals
• Aldol condensation:Diacetone alcohol, aldol of cyclohexanone, y
• DImerization: isooctene
Reactive chromatographyg p y
Process development steps in Reactive chromatography
Adsorption isotherms
R ti Ki ti
Pilot scale FBCR experiments
Reaction Kinetics
Conceptual Design :Triangle theory
SMBR simulations
Process Optimisation and further integration
Fixed Bed Chromatographic Reactor (FBCR)g p ( )
ConstructionC t l t l Catalyst also as adsorbent
Residence time as a critical parameterp
Wave phenomena within the reactorwithin the reactor
Use in non reactive binary adsorptionbinary adsorption isotherm experiments
What is reactive chromatography?What is reactive chromatography?
A is taken in large excess
Isotherm measurement Binary adsorption (breakthrough) experimentsBinary adsorption (breakthrough) experiments
Acetic acid- 2-ethylhexyl acetate, T = 800CAcetic acid- water, T = 800C
2-ethylhexanol 2-ethylhexyl acetate
Flow rate = 2ml/min, T = 800C
2-ethylhexanol 2-ethylhexyl acetate
Flow rate = 4ml/min, T = 800C
Adsorption parameter estimation
Non-reacting Binary mixtures possible areAcetic acid – water, Acetic acid – 2-ethylhexyl acetate2-ethylhexanol – 2-ethylhexyl acetate
Objective function LSQNONLIN Function was2
,exp ,1 1
( ) min ( )N m
ij ij mi j
F x x x
LSQNONLIN Function was used using MATLAB Solver
Parameters of Multicomponent Langmuir isotherm (8-parameter)
Components K (cm3/mol) (m mol/gm)Components K (cm /mol) (m mol/gm)Acetic acid 413.20 22.352-ethylhexanol 284.52 7.42-ethylhexyl acetate 196.12 2.55Water 1232.02 45.3
Affi it O dAffinity Order:Water > Acetic acid > 2-ethylhexanol > 2-ethylhexyl acetate
Parity plots for mole fraction of acetic acidy p
0 80.9
1
0 80.9
1
0.50.60.70.8
odel
ed
0.50.60.70.8
odel
ed
0.20.30.4M
o
0.10.20.30.4M
o
00.1
0 0.5 1observed
00.1
0 0.5 1
Observedobserved
Conc. based LHHW model Activity based LHHW model
• Choice of model is normally governed by how well the kinetic data is explained
• Modified LHHW model explains the data very well
Modeling approach for RCModeling approach for RC
Assumptions: Langmuir adsorption isotherm governs the
multicomponent adsorption phenomenamulticomponent adsorption phenomena. Isothermal operation. Axial dispersion is accounted for.
Resin phase concentration is in equilibrium Resin phase concentration is in equilibrium with the bulk phase liquid concentration.
Superficial liquid velocity is constant.Swelling effect of the resin is neglected Swelling effect of the resin is neglected.
FBCR: Experimental setup and resultsp p
1
0.6
0.8
actio
n
AcH
2EHOH
2EHAc
0.2
0.4
Mol
e fra Water
00 100 200 300 400
Time(mins)
Separation zone
Reactive Separation
Steady state
(R t h i ht 70 t di t 1 5 t l t 39 46
zone pzone Reactive
zone
(Reactor height = 70 cm, reactor diameter = 1.5 cm, catalyst = 39.46 gms,feed flow rate = 1 ml/min, Mole ratio (2EHOH: AcH = 1:1), T = 800C)
Simulated moving bed reactor (SMBR)Simulated moving bed reactor (SMBR)
Construction and workingworking
Catalyst and d b tadsorbent
Switching time gand cycle of an SMBRSMBR
Virtual solids velocityvelocity
Separation within the reactor
SMBR : Working principleg p p
Desorbent (AA)Raffinate (A+D)
3 4
Section IVDesorbent (AA)Raffinate (A+D)
Section IV
2 5
Direction of Port switching and fluid flow
Section III
Feed (A+B)
Raffinate (A+D)
Soldir
Fluid dire c
1 6Section II
( )
Extract (A+C)
lid flow
rection
flow
ction
8 7
Feed (A+B) Section IIExtract (A+C)
Section I
Desorbent (A) Extract (A C)
True moving Bed Simulated moving Bed reactorTrue moving Bed g
SMBR : Working principleSMBR : Working principle
Eluent (A) Recycling
Feed (A+B)
II III IVI
Feed (A+B)
Adsorbent Recycling
Extract (A+C) Raffinate (A+B)
The Triangle Theoryg y
* net fluid flow rate(1 ) net solid flow rate
j colj
Q t Vm
V
(1 ) net solid flow ratecolV
Concentration profile at the end of switch atConcentration profile at the end of switch at cyclic steady state
Raffinate Port
Extract Port
•X-axis depicts the position along the column
Feed
p p g
• Length is relative the nodes which are physically moving in every switch
•Conc. profile remains unchanged due to cyclic steady state in successive switches
SMBR downstream processingSMBR downstream processing
Section IVDesorbent (A)Raffinate (A+D) A
3 4
2 5D
1 6
Direction of Port switching and fluid flow
8 7 C
Feed (A+B) Section IIExtract (A+C)
C
makeup
AMuch of the cost lies in distillative separations
Comparison RC and RD in terms of energyComparison RC and RD in terms of energy consumption (kg of steam / kg of product)
RC + Conventional post-processing
RD
RemarksSolvent
Acetic Acid Alcohol
Methyl Acetate
Methyl Acetate-Acetic Acid 0.29
Not feasible 0.48 RD. is favorable
Water-Acetic Acid 2.06
Total Steam consumption 2.35
Ethyl Acetate-Acetic Acid 0.74
RC. and RD are
Water-Acetic Acid 1.74
Total Steam consumption 2.48
Ethyl Acetate Not feasible 1.97 comparableAlternative 1.74
Iso-Propyl Acetate-Acetic Acid 0.82
Water-Acetic Acid 1.59
Iso-Propyl Acetate Not feasible 1.95RC and RD are comparable
Total Steam consumption 2.41
Alternative 2.05
Butyl Acetate-Acetic Acid 2.9 3.85
Water-Acetic Acid 0 74 0 84
Butyl Acetate 0.68 RD is favorable
Water-Acetic Acid 0.74 0.84
Total Steam consumption 3.64 4.69
Alternative 3.16 -----
2-Ethylhexyl Acetate-Acetic
2-Ethylhexyl Acetate
y yAcid 0.39
Not feasible 0.43 RD is favorable
Water-Acetic Acid 0.92
Total Steam consumption 1.31
Acknowledgementg
Research Scholars:
Sandip Talwalkar
Suman Thotla
Amit Hasabnis
Prafulla Patidar
Rahul Bhat
Amit Agarwalg
Bhoja Reddy
Funding Agencies:
DST, Schenctady Herdillia, VW foundationDST, Schenctady Herdillia, VW foundation