- Email: [email protected]
New configuration for hetero-azeotropic batch distillation: I. Feasibility studies
18th European Symposium on Computer Aided Process Engineering – ESCAPE 18 Bertrand Braunschweig and Xavier Joulia (Editors) © 2008 Elsevier B.V. All rights reserved.
115
New Configuration For Hetero-Azeotropic Batch Distillation: I. Feasibility Studies Peter Langa , Ferenc Denesa and Xavier Jouliab a
BUTE Dept. of Building Services & Process Engineering, H-1521 Budapest, Muegyetem rkp. 3-5, bLGC-ENSIACET-INPT, 118 route de Narbonne, 31077 Toulouse, France
Abstract For heterogeneous batch distillation a new double column configuration operating in closed system is suggested. This configuration is investigated by feasibility studies based on the assumption of maximal separation and is compared with the traditional batch rectifier. The calculations are performed for a binary (n-butanol – water) and for a ternary heteroazeotropic mixture (isopropanol – water + benzene as entrainer). Keywords: heteroazeotrope, batch distillation, feasibility studies.
1. Introduction If components of a mixture form a heteroazeotrope or by the addition of an entrainer (E) a heteroazeotrope can be formed, the azeotropic composition can be crossed by decantation. In the pharmaceutical and fine chemical industries batch processes including the batch heteroazeotropic distillation (BHD) are widely applied. As far as we know the BHD was exclusively applied in the industry in batch rectifiers (equipped with a decanter) in open operation mode (with continuous top product withdrawal). The batch rectifier (BR) was investigated with variable decanter holdup by Rodriguez-Donis et al. (2002) and with continuous entrainer feeding by Modla et al. (2001, 2003) and Rodriguez-Donis et al. (2003), respectively. Recently the BHD was extensively studied for the BR and multivessel columns by Skouras et al. (2005a,b). The objectives of this paper are - to suggest a new double-column system (DCS) for the BHD, - to investigate this configuration by feasibility studies, - to compare its performance with that of the traditional BR. Calculations are performed for a binary (n-butanol – water) and for a ternary heteroazeotropic mixture (isopropanol – water + benzene).
2. The column configurations studied First the BR then the new DCS is studied by assuming maximal separation. 2.1. Batch rectifier (Fig. 1.) First the separation of the binary then that of the ternary mixture is presented. Separation of a binary mixture If the charge (feed) composition (xch,A (mole fraction of component A)) is in the Br Ar heterogeneous region ( x AZ, ) it is worth to separate it by decantation A < x ch,A < x AZ,A into an A-rich ( x Ar ) and a B-rich ( x Br ) phase before the start of the distillation. AZ,A AZ,A One production cycle consists of two distillation steps. In the first step we select the phase to be distilled so that the overall quantity of the two products in the first cycle be
P. Lang et al.
116
Fig. 1. BR producing A from mixture A-B
Fig. 2. DCS for binary mixture
maximal. It can be derived (for pure products) that we have to distil the A-rich phase Br Ar Br first if x ch ,A > x AZ ,A /[1 − ( x AZ,A − x AZ,A )] . Step 1: Production of A: The A-rich phase ( x Ar ) of the heteroazeotrope (xAZ,A) is reAZ,A Br fluxed and the B-rich one ( x AZ, ) is withdrawn as distillate. The bottoms is product A. A
Step 2: Production of B: The B-rich phase(s) is (are) distilled. The B-rich phase of the azeotrope is refluxed and the A-rich one is withdrawn as distillate. The bottom residue is product B. The main disadvantages of the BR are that in one step only one component can be produced (in the residue) and that the recovery is limited since the other component in the distillate is saturated with this component. Separation of the ternary mixture The separation of a homoazeotropic isopropanol (A) – water (B) mixture is considered. Addition of an entrainer, in a small amount, is needed. The steps of a production cycle are: Er
Step 1: Production of A: The E-rich phase ( x TAZ ) of the ternary azeotrope ( x TAZ ) is Br
refluxed and the B-rich phase ( x TAZ ) is withdrawn as distillate, which is distilled in Step 2. The bottom residue is product A. Step 2: Removal of E: The B-rich phase of the azeotrope is refluxed and the E-rich phase is withdrawn as distillate. The bottom residue still contains some A. Step 3: Purification of B from A: A is removed (from the bottom residue of Step 2) in the form of binary A-B homoazeotrope ( x BAZ ) in the distillate and the bottom residue is product B. 2.2. The new double column system (Fig. 2.) The two column system is operated in closed mode (without continuous product withdrawal) with a single decanter. The two components are simultaneously produced as bottom residues. Separation of a binary mixture A heterogeneous charge is separated by decantation. The A-rich phase is filled in the reboiler of the column α (producing A) and a B-rich one to the other reboiler β. A homogeneous charge can be divided between the two reboilers. The top vapour of both columns is of azeotropic composition. The A-rich phase is sent to the top of column α and the B-rich one is fed to the top of column β.
New Configuration for Hetero-Azeotropic Batch Distillation: I. Feasibility Studies
117
Separation of the ternary mixture The homogeneous charge can be arbitrarily divided between the two reboilers.The entrainer, which is filled at the start only in the reboiler of column α, circulates in the system. The top vapour of the column α is ternary azeotrope and that of column β is AB binary azeotrope. The E-rich phase is sent to the top of column α and the B-rich one (containing negligible amount of E) is fed to the top of column β.
3. Feasibility method Our aim is to estimate the duration of the processes and the amount of products. A simplified model was applied based on the following assumptions: maximal separation, negligible hold-up on the trays and in the decanter, constant molar overflow, the flow rates do not vary with the time, one-phase liquid streams leave the decanter, negligible duration of pumping between the operation steps (BR), no entrainer loss (in the case of the ternary mixture). The total and component material balances for one column and the decanter are analytically solved. For the DCS we assume that both products reach the prescribed purity at the same time, that is, the duration is minimal. The process time (τ) for both configurations and for the DCS the optimal division (vα) of total vapour flow rate (V) between the two reboilers and that of the charge (Ubα/Uch) are shown. 3.1. Equations for the BR Separation of a binary mixture jr ( x irAZ,i − x AZ U ,i )( x e ,i − x b ,i ) Duration of a step: τ = ir ⋅ b jr V ( x AZ,i − x AZ,i )( x e,i − x AZ ) ,i
where i, j: components (i is produced in the given step); x b,i , x e, i : mole fraction of i in the reboiler at the beginning and end of the step; U: molar holdup in the reboiler. Distillation of the ternary mixture Step 1: We suppose that product A does not contain E (it is polluted only by B). Er Br ( x TAZ U , A − x TAZ, A )( x spec, A − x ch , A ) Duration of this step: τ(1) = Er ⋅ ch , Br V ( x TAZ, A − x TAZ, A )( x spec, A − x TAZ ) ,A where x spec,A is the specified purity of product A. Step 2: The top vapour has ternary azeotropic composition. Duration of this step: Br Br x Er ,A − x TAZ ,A x TAZ,E U b τ ( 2) = TAZ ⋅ ⋅ Er Er V x TAZ ,A − x TAZ,A x TAZ,E Step 3: In this step only A and B are present, the top vapour is the homoazeotrope. There is no need for a decanter. Duration of this step: x b,A − (1 − x spec,B ) U τ (3) = ⋅ (1 + R ) ⋅ b , where R is the reflux ratio. x BAZ,A − (1 − x spec,B ) V 3.2. Equations for the double column system Separation of the binary mixture Ar β Br For a heterogeneous charge: x αb,A = x AZ ,A and x b ,A = x AZ,A . For a homogeneous one:
x αb, A = x βb, A = x ch , A , the number of independent equations is less by one than in the previous case, hence one of the unknowns ( U αb , U βb , v α ) must be specified. The value of the main operational parameters:
P. Lang et al.
118
τ=
α α α Ar Br Br Δ( U α x αA ) − ΔU α x AZ, A x Ar AZ, A − x AZ, A α Δ ( U x A ) − ΔU x AZ, A x AZ, A − x AZ, A , v ⋅ = ⋅ Ar Br Br Δ( U α x αA ) − ΔU α x AZ, A x AZ x Ar x AZ, A − x AZ AZ, A − x AZ, A ,A , A − x AZ, A
where ΔU α = U eα − U αb , Δ( U α ⋅ x αA ) = U eα ⋅ x spec,A − U αb ⋅ x αb,A , U αb =
x ch ,A − x βb,A x αb,A − x βb,A
⋅ U ch
Separation of a ternary mixture Initially only the reboiler α contains E. We neglect the content of E of the B-rich phase. Hence there is no E in column β whose top vapour is A-B binary azeotrope. The number of unknowns is more than the number of independent equations by one. Hence one of the unknowns must be specified. The composition of the phases is a function of vα but this function can be mathematically expressed with difficulty which would make the solution of the set of equations difficult. Hence we specify vα. The values of the main parameters of the operation: U αe ⋅ ( x spec,A − x ch ,A ) x Er E − x TAZ, E α α τ = Er , = + ⋅ vα ⋅ V ⋅ τ , U U b e Er α x L ⋅ ( x Er − x ) + v ⋅ V ⋅ ( x − x ) E A ch ,A ch ,A TAZ,A U eα =
x ch ,A − (1 − x spec,B ) x spec,A − (1 − x spec,B )
⋅ U ch
4. Calculation results The total vapour flow rate of the DCS was taken equal to that of the BR ( V = 20 kmol/h ). (The heat duty is proportional to the vapour flow rate.) For the DCS we determine the optimal division of the charge between the two reboilers (and the division of the total vapour flow rates belonging to it). In all cases the amount of charge is 100 kmol and the specified purity (xspec,i) is 99.5 mol% for both products. 4.1. Separation of binary mixtures (n-butanol(A) – water(B)) The composition of the heteroazeotrope and that of the A-rich and B-rich phases, Ar
Br
respectively: x AZ = [0.2562, 0.7438] , x AZ = [0.568, 0.432] , x AZ = [0.012, 0.988] All possible cases are studied: two homogeneous charges (one rich in A and the other rich in B) and a heterogeneous one. 4.1.1. Homogeneous charge rich in A a.Batch rectifier: x ch = [0.9, 0.1] . In Step 1 A is produced (Table 1). b. Double Column System We determine τ and vα for different ratios Ubα/Uch (Figs. 3. & 4.) The best operational policy (Table 1) is when the total amount of the charge is fed into reboiler α ( U αb / U ch = 1 ).The duration of the cycle is nearly equal for the two configuBR Step 1 Step 2 Ubα,β/Uch Div. of vap. fl.rate Duration [h] 0.862 0.014 Prod. A [kmol] 90.336 0.000 Prod. B [kmol] 0.000 9.544 Byprods. [kmol] 0.120 Byps. comp. [mol%] 56.80
DCS Col. α Col. β 1.00 0.00 0.9844 0.0156 0.880 90.404 0.000 0.000 9.596 0.000 -
BR Step 1 Step 2 Ubα,β/Uch Div. of vap. fl.rate Duration [h] 0.101 0.034 Prod. A [kmol] 0.502 0.000 Prod. B [kmol] 0.000 99.112 Byprods. [kmol] 0.386 Byps. comp. [mol%] 1.12
DCS Col. α Col. β 0.00 1.00 0.2538 0.7462 0.136 0.505 0.000 0.000 99.495 0.000 -
Table 1. Results (binary mixture rich in A) Table 2. Results (binary mixture rich in B)
New Configuration for Hetero-Azeotropic Batch Distillation: I. Feasibility Studies BR Step 1 Step 2 α, Ub β/Uch Div. of vap. fl.rate Duration [h] 2.006 0.100 Prod. A [kmol] 29.298 0.000 Prod. B [kmol] 0.000 69.823 Byprods. [kmol] 0.879 Byps. comp. [mol%] 56.80
DCS Col. α Col. β 0.5180 0.4820 0.9530 0.0470 2.141 29.798 0.000 0.000 70.202 0.000 -
Step 3 0.261 0.000 30.723 5.224 BAZ
DCS Col. α Col. β 0.9981 0.0019 0.9650 0.0350 8.494 67.576 0.000 0.000 32.424 0.000 -
Table 4. Results (ternary mixture)
25
1.0
20
0.8
15
0.6
10
0.4
5
0.2 0.0
0 0.0
0.2
0.4
0.6
α
0.8
0.0
1.0
0.2
U b /U ch
0.4
0.6
U bα /U ch
0.8
1.0
Fig. 4. Rel. vap. flow rate of column α (binary mixture rich in A)
Fig. 3. Duration of the process (binary mixture rich in A)
1.0
16
0.8
U bα /U ch
12
τ [h]
BR Step 2 0.010 0.000 0.000 0.160 TAZ
vα
τ [h]
Table 3. Results (bin. heterogeneous mix.)
Step 1 Ubα,β/Uch Div. of vap. fl.rate Duration [h] 8.055 Prod. A [kmol] 64.001 Prod. B [kmol] 0.000 Byprods. [kmol] Byprods. compn. -
119
8 4
0.6 0.4 0.2
0
0.0 0.5
0.6
0.7
v
α
0.8
0.9
Fig. 5. Duration of the process (ternary mixture)
1.0
0.5
0.6
0.7
vα
0.8
0.9
1.0
Fig. 6. Division of the charge (ternary mixture)
rations. In the case of DCS by the best policy the whole amount of A is already in the reboiler α at the start and only B must be eliminated from it. The reason of the small value of vβ is that the B-rich phase flowing from the decanter into column β has very high B-content ( x Br AZ,B = 0.988 ). Hence only a small amount of A must be removed in the form of azeotrope for the purification of B. The main advantage of the DCS is that there is no residue at all. 4.1.2. Homogeneous charge rich in B a. Batch rectifier: x ch = [0.01, 0.99] . In Step 1 B is produced (Table 2). b. DCS: We determined τ and vα for different divisions of the charge. The best operational policy (Table 2) is when the total amount of the charge is fed into reboiler β. The duration of the cycle is nearly equal in the two cases. The ratio of the duration of the two steps for the BR: τ (1) / τ ( 2) = 2.971 The ratio of vapour flow rates of the two columns for the DCS: v β / v α = 2.940
P. Lang et al.
120
The values of these two ratios show that energy demand of the production of each component is nearly the same for the two configurations. The division of the charge can be explained similarly as in the case of the previous charge composition. 4.1.3. Heterogeneous charge Before the distillation the charge of composition x ch = [0.3, 0.7] is separated by decantation into an A-rich ( U Ar = 51.8 kmol ) and a B-rich ( U Br = 48.2 kmol ) phases. a. Batch rectifier: In Step 1 the A-rich phase is distilled and A is produced (Table 3). b. DCS: The preliminary decantation provides the division of the charge which determines the value of vA. Hence only one solution exists (Table 3). The duration of the cycle is nearly equal in the two cases. 4.2. Separation of the ternary mixture (isopropanol (A) – water (B) + benzene (E))
Binary azeotropic charge ( x BAZ = [0.674, 0.326,0] ) is separated by the application of an entrainer (E). The composition of the ternary IPA – water – benzene heteroazeotrope and those of its E-rich and B-rich phases: Er
Br
x TAZ = [0.238, 0.239, 0.523] , x TAZ = [0.277, 0.048, 0.675] , x TAZ = [0.103, 0.894, 0.003] a.Batch rectifier. Calculation results are shown in Table 4. b.DCS: We determine τ and U αb / U ch for different rel. vapour flow rates of column α (Figs. 5-6). Calculation results for the best operational policy are shown in Table 4. The duration of cycle is nearly equal in the two cases. The amount of the final residue is more than 5 % of the charge for the BR, whilst there is no residue at all by the DCR.
5. Conclusion We suggest using a new double column system (DCS) for heterogeneous batch distillation. It is operated in closed mode without continuous product withdrawal. This configuration was investigated by feasibility studies based on a simplified model (maximal separation, negligible holdup) and was compared with the traditional batch rectifier (BR). The calculations were performed for the mixtures n-butanol – water and isopropanol – water + benzene (entrainer). The performance of the DCS was compared with that of the BR. The main benefit of the DCS is that it produces no residue to be separated later. The DCS proved to be feasible and in the cases studied competitive with the BR. In comparison with the BR it gave for the ternary mixture better and for the binary one similar results, respectively. Feasibility studies were completed by rigorous simulations. The results of these calculations based on much less simplifying assumptions are published in a separate paper.
References Modla G., P. Lang and K. Molnar, (2001). Batch Heteroazeotropic Rectification … 6th WCCE, Melbourne, Australia, (10 pages on CD),. Modla G., P. Lang , B. Kotai and K. Molnar, (2003). AIChE J, 49 (10), 2533. Rodriguez-Donis I, V. Gerbaud, X. Joulia, (2002). AIChE J, 48 (6), 1168. Rodriguez-Donis Y., J. Equijarosa, V. Gerbaud, X. Joulia, (2003). AIChE J, 49, 3074. Skouras S., V. Kiva , S. Skogestad, (2005a). Chem. Eng. Sci., 60, 2895. Skouras S., S. Skogestad, V. Kiva, (2005b). AIChE Journal, 51 (4), 1144-1157.
115
New Configuration For Hetero-Azeotropic Batch Distillation: I. Feasibility Studies Peter Langa , Ferenc Denesa and Xavier Jouliab a
BUTE Dept. of Building Services & Process Engineering, H-1521 Budapest, Muegyetem rkp. 3-5, bLGC-ENSIACET-INPT, 118 route de Narbonne, 31077 Toulouse, France
Abstract For heterogeneous batch distillation a new double column configuration operating in closed system is suggested. This configuration is investigated by feasibility studies based on the assumption of maximal separation and is compared with the traditional batch rectifier. The calculations are performed for a binary (n-butanol – water) and for a ternary heteroazeotropic mixture (isopropanol – water + benzene as entrainer). Keywords: heteroazeotrope, batch distillation, feasibility studies.
1. Introduction If components of a mixture form a heteroazeotrope or by the addition of an entrainer (E) a heteroazeotrope can be formed, the azeotropic composition can be crossed by decantation. In the pharmaceutical and fine chemical industries batch processes including the batch heteroazeotropic distillation (BHD) are widely applied. As far as we know the BHD was exclusively applied in the industry in batch rectifiers (equipped with a decanter) in open operation mode (with continuous top product withdrawal). The batch rectifier (BR) was investigated with variable decanter holdup by Rodriguez-Donis et al. (2002) and with continuous entrainer feeding by Modla et al. (2001, 2003) and Rodriguez-Donis et al. (2003), respectively. Recently the BHD was extensively studied for the BR and multivessel columns by Skouras et al. (2005a,b). The objectives of this paper are - to suggest a new double-column system (DCS) for the BHD, - to investigate this configuration by feasibility studies, - to compare its performance with that of the traditional BR. Calculations are performed for a binary (n-butanol – water) and for a ternary heteroazeotropic mixture (isopropanol – water + benzene).
2. The column configurations studied First the BR then the new DCS is studied by assuming maximal separation. 2.1. Batch rectifier (Fig. 1.) First the separation of the binary then that of the ternary mixture is presented. Separation of a binary mixture If the charge (feed) composition (xch,A (mole fraction of component A)) is in the Br Ar heterogeneous region ( x AZ, ) it is worth to separate it by decantation A < x ch,A < x AZ,A into an A-rich ( x Ar ) and a B-rich ( x Br ) phase before the start of the distillation. AZ,A AZ,A One production cycle consists of two distillation steps. In the first step we select the phase to be distilled so that the overall quantity of the two products in the first cycle be
P. Lang et al.
116
Fig. 1. BR producing A from mixture A-B
Fig. 2. DCS for binary mixture
maximal. It can be derived (for pure products) that we have to distil the A-rich phase Br Ar Br first if x ch ,A > x AZ ,A /[1 − ( x AZ,A − x AZ,A )] . Step 1: Production of A: The A-rich phase ( x Ar ) of the heteroazeotrope (xAZ,A) is reAZ,A Br fluxed and the B-rich one ( x AZ, ) is withdrawn as distillate. The bottoms is product A. A
Step 2: Production of B: The B-rich phase(s) is (are) distilled. The B-rich phase of the azeotrope is refluxed and the A-rich one is withdrawn as distillate. The bottom residue is product B. The main disadvantages of the BR are that in one step only one component can be produced (in the residue) and that the recovery is limited since the other component in the distillate is saturated with this component. Separation of the ternary mixture The separation of a homoazeotropic isopropanol (A) – water (B) mixture is considered. Addition of an entrainer, in a small amount, is needed. The steps of a production cycle are: Er
Step 1: Production of A: The E-rich phase ( x TAZ ) of the ternary azeotrope ( x TAZ ) is Br
refluxed and the B-rich phase ( x TAZ ) is withdrawn as distillate, which is distilled in Step 2. The bottom residue is product A. Step 2: Removal of E: The B-rich phase of the azeotrope is refluxed and the E-rich phase is withdrawn as distillate. The bottom residue still contains some A. Step 3: Purification of B from A: A is removed (from the bottom residue of Step 2) in the form of binary A-B homoazeotrope ( x BAZ ) in the distillate and the bottom residue is product B. 2.2. The new double column system (Fig. 2.) The two column system is operated in closed mode (without continuous product withdrawal) with a single decanter. The two components are simultaneously produced as bottom residues. Separation of a binary mixture A heterogeneous charge is separated by decantation. The A-rich phase is filled in the reboiler of the column α (producing A) and a B-rich one to the other reboiler β. A homogeneous charge can be divided between the two reboilers. The top vapour of both columns is of azeotropic composition. The A-rich phase is sent to the top of column α and the B-rich one is fed to the top of column β.
New Configuration for Hetero-Azeotropic Batch Distillation: I. Feasibility Studies
117
Separation of the ternary mixture The homogeneous charge can be arbitrarily divided between the two reboilers.The entrainer, which is filled at the start only in the reboiler of column α, circulates in the system. The top vapour of the column α is ternary azeotrope and that of column β is AB binary azeotrope. The E-rich phase is sent to the top of column α and the B-rich one (containing negligible amount of E) is fed to the top of column β.
3. Feasibility method Our aim is to estimate the duration of the processes and the amount of products. A simplified model was applied based on the following assumptions: maximal separation, negligible hold-up on the trays and in the decanter, constant molar overflow, the flow rates do not vary with the time, one-phase liquid streams leave the decanter, negligible duration of pumping between the operation steps (BR), no entrainer loss (in the case of the ternary mixture). The total and component material balances for one column and the decanter are analytically solved. For the DCS we assume that both products reach the prescribed purity at the same time, that is, the duration is minimal. The process time (τ) for both configurations and for the DCS the optimal division (vα) of total vapour flow rate (V) between the two reboilers and that of the charge (Ubα/Uch) are shown. 3.1. Equations for the BR Separation of a binary mixture jr ( x irAZ,i − x AZ U ,i )( x e ,i − x b ,i ) Duration of a step: τ = ir ⋅ b jr V ( x AZ,i − x AZ,i )( x e,i − x AZ ) ,i
where i, j: components (i is produced in the given step); x b,i , x e, i : mole fraction of i in the reboiler at the beginning and end of the step; U: molar holdup in the reboiler. Distillation of the ternary mixture Step 1: We suppose that product A does not contain E (it is polluted only by B). Er Br ( x TAZ U , A − x TAZ, A )( x spec, A − x ch , A ) Duration of this step: τ(1) = Er ⋅ ch , Br V ( x TAZ, A − x TAZ, A )( x spec, A − x TAZ ) ,A where x spec,A is the specified purity of product A. Step 2: The top vapour has ternary azeotropic composition. Duration of this step: Br Br x Er ,A − x TAZ ,A x TAZ,E U b τ ( 2) = TAZ ⋅ ⋅ Er Er V x TAZ ,A − x TAZ,A x TAZ,E Step 3: In this step only A and B are present, the top vapour is the homoazeotrope. There is no need for a decanter. Duration of this step: x b,A − (1 − x spec,B ) U τ (3) = ⋅ (1 + R ) ⋅ b , where R is the reflux ratio. x BAZ,A − (1 − x spec,B ) V 3.2. Equations for the double column system Separation of the binary mixture Ar β Br For a heterogeneous charge: x αb,A = x AZ ,A and x b ,A = x AZ,A . For a homogeneous one:
x αb, A = x βb, A = x ch , A , the number of independent equations is less by one than in the previous case, hence one of the unknowns ( U αb , U βb , v α ) must be specified. The value of the main operational parameters:
P. Lang et al.
118
τ=
α α α Ar Br Br Δ( U α x αA ) − ΔU α x AZ, A x Ar AZ, A − x AZ, A α Δ ( U x A ) − ΔU x AZ, A x AZ, A − x AZ, A , v ⋅ = ⋅ Ar Br Br Δ( U α x αA ) − ΔU α x AZ, A x AZ x Ar x AZ, A − x AZ AZ, A − x AZ, A ,A , A − x AZ, A
where ΔU α = U eα − U αb , Δ( U α ⋅ x αA ) = U eα ⋅ x spec,A − U αb ⋅ x αb,A , U αb =
x ch ,A − x βb,A x αb,A − x βb,A
⋅ U ch
Separation of a ternary mixture Initially only the reboiler α contains E. We neglect the content of E of the B-rich phase. Hence there is no E in column β whose top vapour is A-B binary azeotrope. The number of unknowns is more than the number of independent equations by one. Hence one of the unknowns must be specified. The composition of the phases is a function of vα but this function can be mathematically expressed with difficulty which would make the solution of the set of equations difficult. Hence we specify vα. The values of the main parameters of the operation: U αe ⋅ ( x spec,A − x ch ,A ) x Er E − x TAZ, E α α τ = Er , = + ⋅ vα ⋅ V ⋅ τ , U U b e Er α x L ⋅ ( x Er − x ) + v ⋅ V ⋅ ( x − x ) E A ch ,A ch ,A TAZ,A U eα =
x ch ,A − (1 − x spec,B ) x spec,A − (1 − x spec,B )
⋅ U ch
4. Calculation results The total vapour flow rate of the DCS was taken equal to that of the BR ( V = 20 kmol/h ). (The heat duty is proportional to the vapour flow rate.) For the DCS we determine the optimal division of the charge between the two reboilers (and the division of the total vapour flow rates belonging to it). In all cases the amount of charge is 100 kmol and the specified purity (xspec,i) is 99.5 mol% for both products. 4.1. Separation of binary mixtures (n-butanol(A) – water(B)) The composition of the heteroazeotrope and that of the A-rich and B-rich phases, Ar
Br
respectively: x AZ = [0.2562, 0.7438] , x AZ = [0.568, 0.432] , x AZ = [0.012, 0.988] All possible cases are studied: two homogeneous charges (one rich in A and the other rich in B) and a heterogeneous one. 4.1.1. Homogeneous charge rich in A a.Batch rectifier: x ch = [0.9, 0.1] . In Step 1 A is produced (Table 1). b. Double Column System We determine τ and vα for different ratios Ubα/Uch (Figs. 3. & 4.) The best operational policy (Table 1) is when the total amount of the charge is fed into reboiler α ( U αb / U ch = 1 ).The duration of the cycle is nearly equal for the two configuBR Step 1 Step 2 Ubα,β/Uch Div. of vap. fl.rate Duration [h] 0.862 0.014 Prod. A [kmol] 90.336 0.000 Prod. B [kmol] 0.000 9.544 Byprods. [kmol] 0.120 Byps. comp. [mol%] 56.80
DCS Col. α Col. β 1.00 0.00 0.9844 0.0156 0.880 90.404 0.000 0.000 9.596 0.000 -
BR Step 1 Step 2 Ubα,β/Uch Div. of vap. fl.rate Duration [h] 0.101 0.034 Prod. A [kmol] 0.502 0.000 Prod. B [kmol] 0.000 99.112 Byprods. [kmol] 0.386 Byps. comp. [mol%] 1.12
DCS Col. α Col. β 0.00 1.00 0.2538 0.7462 0.136 0.505 0.000 0.000 99.495 0.000 -
Table 1. Results (binary mixture rich in A) Table 2. Results (binary mixture rich in B)
New Configuration for Hetero-Azeotropic Batch Distillation: I. Feasibility Studies BR Step 1 Step 2 α, Ub β/Uch Div. of vap. fl.rate Duration [h] 2.006 0.100 Prod. A [kmol] 29.298 0.000 Prod. B [kmol] 0.000 69.823 Byprods. [kmol] 0.879 Byps. comp. [mol%] 56.80
DCS Col. α Col. β 0.5180 0.4820 0.9530 0.0470 2.141 29.798 0.000 0.000 70.202 0.000 -
Step 3 0.261 0.000 30.723 5.224 BAZ
DCS Col. α Col. β 0.9981 0.0019 0.9650 0.0350 8.494 67.576 0.000 0.000 32.424 0.000 -
Table 4. Results (ternary mixture)
25
1.0
20
0.8
15
0.6
10
0.4
5
0.2 0.0
0 0.0
0.2
0.4
0.6
α
0.8
0.0
1.0
0.2
U b /U ch
0.4
0.6
U bα /U ch
0.8
1.0
Fig. 4. Rel. vap. flow rate of column α (binary mixture rich in A)
Fig. 3. Duration of the process (binary mixture rich in A)
1.0
16
0.8
U bα /U ch
12
τ [h]
BR Step 2 0.010 0.000 0.000 0.160 TAZ
vα
τ [h]
Table 3. Results (bin. heterogeneous mix.)
Step 1 Ubα,β/Uch Div. of vap. fl.rate Duration [h] 8.055 Prod. A [kmol] 64.001 Prod. B [kmol] 0.000 Byprods. [kmol] Byprods. compn. -
119
8 4
0.6 0.4 0.2
0
0.0 0.5
0.6
0.7
v
α
0.8
0.9
Fig. 5. Duration of the process (ternary mixture)
1.0
0.5
0.6
0.7
vα
0.8
0.9
1.0
Fig. 6. Division of the charge (ternary mixture)
rations. In the case of DCS by the best policy the whole amount of A is already in the reboiler α at the start and only B must be eliminated from it. The reason of the small value of vβ is that the B-rich phase flowing from the decanter into column β has very high B-content ( x Br AZ,B = 0.988 ). Hence only a small amount of A must be removed in the form of azeotrope for the purification of B. The main advantage of the DCS is that there is no residue at all. 4.1.2. Homogeneous charge rich in B a. Batch rectifier: x ch = [0.01, 0.99] . In Step 1 B is produced (Table 2). b. DCS: We determined τ and vα for different divisions of the charge. The best operational policy (Table 2) is when the total amount of the charge is fed into reboiler β. The duration of the cycle is nearly equal in the two cases. The ratio of the duration of the two steps for the BR: τ (1) / τ ( 2) = 2.971 The ratio of vapour flow rates of the two columns for the DCS: v β / v α = 2.940
P. Lang et al.
120
The values of these two ratios show that energy demand of the production of each component is nearly the same for the two configurations. The division of the charge can be explained similarly as in the case of the previous charge composition. 4.1.3. Heterogeneous charge Before the distillation the charge of composition x ch = [0.3, 0.7] is separated by decantation into an A-rich ( U Ar = 51.8 kmol ) and a B-rich ( U Br = 48.2 kmol ) phases. a. Batch rectifier: In Step 1 the A-rich phase is distilled and A is produced (Table 3). b. DCS: The preliminary decantation provides the division of the charge which determines the value of vA. Hence only one solution exists (Table 3). The duration of the cycle is nearly equal in the two cases. 4.2. Separation of the ternary mixture (isopropanol (A) – water (B) + benzene (E))
Binary azeotropic charge ( x BAZ = [0.674, 0.326,0] ) is separated by the application of an entrainer (E). The composition of the ternary IPA – water – benzene heteroazeotrope and those of its E-rich and B-rich phases: Er
Br
x TAZ = [0.238, 0.239, 0.523] , x TAZ = [0.277, 0.048, 0.675] , x TAZ = [0.103, 0.894, 0.003] a.Batch rectifier. Calculation results are shown in Table 4. b.DCS: We determine τ and U αb / U ch for different rel. vapour flow rates of column α (Figs. 5-6). Calculation results for the best operational policy are shown in Table 4. The duration of cycle is nearly equal in the two cases. The amount of the final residue is more than 5 % of the charge for the BR, whilst there is no residue at all by the DCR.
5. Conclusion We suggest using a new double column system (DCS) for heterogeneous batch distillation. It is operated in closed mode without continuous product withdrawal. This configuration was investigated by feasibility studies based on a simplified model (maximal separation, negligible holdup) and was compared with the traditional batch rectifier (BR). The calculations were performed for the mixtures n-butanol – water and isopropanol – water + benzene (entrainer). The performance of the DCS was compared with that of the BR. The main benefit of the DCS is that it produces no residue to be separated later. The DCS proved to be feasible and in the cases studied competitive with the BR. In comparison with the BR it gave for the ternary mixture better and for the binary one similar results, respectively. Feasibility studies were completed by rigorous simulations. The results of these calculations based on much less simplifying assumptions are published in a separate paper.
References Modla G., P. Lang and K. Molnar, (2001). Batch Heteroazeotropic Rectification … 6th WCCE, Melbourne, Australia, (10 pages on CD),. Modla G., P. Lang , B. Kotai and K. Molnar, (2003). AIChE J, 49 (10), 2533. Rodriguez-Donis I, V. Gerbaud, X. Joulia, (2002). AIChE J, 48 (6), 1168. Rodriguez-Donis Y., J. Equijarosa, V. Gerbaud, X. Joulia, (2003). AIChE J, 49, 3074. Skouras S., V. Kiva , S. Skogestad, (2005a). Chem. Eng. Sci., 60, 2895. Skouras S., S. Skogestad, V. Kiva, (2005b). AIChE Journal, 51 (4), 1144-1157.