Separation of a Chloroform–Acetone–Toluene Mixture by Pressure

May 20, 2011 - Department of Building Services and Process Engineering, Budapest University of Technology and Economics, Muegyetem rkp. 3−5, H-1521 ...
0 downloads 0 Views 3MB Size
ARTICLE pubs.acs.org/IECR

Separation of a ChloroformAcetoneToluene Mixture by Pressure-Swing Batch Distillation in Different Column Configurations Gabor Modla* Department of Building Services and Process Engineering, Budapest University of Technology and Economics, Muegyetem rkp. 35, H-1521 Budapest, Hungary ABSTRACT: In this paper, a new, triple-column configuration application is introduced to separate ternary mixtures by pressureswing batch distillation. This study investigates the separation of a chloroformacetonetoluene ternary mixture in different column configurations (a combination of the batch stripper and the double-column batch rectifier, the double-column batch rectifier, and the triple-column configuration) by a feasibility study and by rigorous simulation and compares these processes at two different charge compositions by its overall specific energy consumption. The influences of the most important operational parameters are studied for all processes, and the performances of the column configurations are compared. On the basis of the results, we state that the triple-column configuration is more efficient than the other column configurations investigated.

1. INTRODUCTION Pressure-swing distillation is an efficient method for the separation of pressure-sensitive azeotropic mixtures. Many mixtures form an azeotrope whose position can be shifted substantially by changing the system pressure, that is, a pressure-sensitive azeotrope. (At some pressures, the azeotrope may even disappear.) This effect can be exploited to separate azeotropic mixtures without application of a separating agent by the so-called pressure-swing distillation. A considerable amount of literature has been published on pressure-swing distillation in a continuous system.19 In general, these studies focused on the separation of binary mixtures, but Knapp and Doherty5 presented some ternary systems as well. In recent years, there has been an increasing amount of literature1018 on the separation of binary azeotropic mixtures by pressure-swing batch and semicontinuous distillation in a single column or in double-column systems but only few studies13,15 investigated the separation of ternary mixtures, the study of which does not present any rigorous calculation or experimental results. Modla and Lang12 studied the feasibility of pressure-swing batch distillation (PSBD) of binary mixtures (forming a minimum or maximum azeotrope) in different column configurations assuming maximal separation. They suggested two novel configurations containing two rectifying (double-column batch rectifier, DCBR) or two stripping sections (double-column batch stripper, DCBS). They made rigorous simulation calculations for different column configurations, as well. The different configurations were compared for a given set of operational parameters. The best results were obtained with the two new double-column configurations. For the separation of minimum azeotropes, they suggested the application of DCBS or a batch stripper (BS) and for maximum azeotropes DCBR or a batch rectifier, respectively. They stated that the middle-vessel column (MVC) is not suitable for the separation of pressure-sensitive binary azeotropes. The main advantage of the suggested new double-column systems (DCBR and DCBS) is that the pressure change is realized in r 2011 American Chemical Society

different column sections instead of in different process steps (as is done at the batch rectifier and BS); hence, there are not unproductive time periods between the low- and high-pressure process steps. Moreover, at the double-column systems, the two columns can be thermally integrated in order to save energy.16 Thanks to these advantages, the double-column systems can be operated at lower cost. Modla et al.15 studied the feasibility of PSBD separation of ternary homoazeotropic mixtures in different single- and doublecolumn configurations. In that paper, separation of the most frequent types of ternary mixtures was investigated. They extended the classification of the ternary system for pressuresensitive azeotropic mixtures. Our long-term mission is to improve the batch processes in order to reduce the energy demand for the separation of different azeotropic mixtures. In previous works,1218 we introduced and studied new double-column configurations for the separation of binary pressure-sensitive azeotropic mixtures. In this paper, we introduce a new triple-column configuration (TCC) for the separation of a ternary mixture, chloroformacetonetoluene, by PSBD and investigate and compare the separation in different column configurations.

2. FEASIBILITY STUDY The aim of this section is to determine feasible processes and column configurations by analysis of the residue curve maps (RCMs) and binary vaporliquid equilibrium diagrams. The method is based on the determination of feasible compositions of products (continuously withdrawn) and those of residues (remaining in the vessel) by analysis of the vessel paths in the RCMs at two different pressures (PLP and þPHP). By the Received: July 23, 2010 Accepted: May 20, 2011 Revised: May 1, 2011 Published: May 20, 2011 8204

dx.doi.org/10.1021/ie101578j | Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research

ARTICLE

Figure 1. RCMs: (A) P = 1.01 bar; (B) P = 10 bar.

feasibility studies, the separation steps are determined for the different column configurations. The feasible region (FR) of the separation is defined as follows:12 All feed compositions from where all three components (or two components at the binary system) can be purely recovered by maximal separation (in at least one appropriate one- or two-column configuration) at the given pressure or by the application of pressure swing. (The number of separation steps is not limited.)

The operating region of pressure-swing distillation is defined as follows: All feed compositions from which, through pressure changes, the different components can be purely recovered by maximal separation. 2.1. VaporLiquid Equilibrium Data. On the basis of Modla and Lang,13 the RCM classification of a chloroform (I)acetone (L)toluene (H) mixture is 4P-0-0 (and 1.0-2 by Serafimov19). The acetonechloroform mixture forms a maximum boilingpoint pressure-sensitive azeotrope that is a saddle point (AzLI) in 8205

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research

ARTICLE

Figure 2. Sketch of the separation process: (A) BS process step; (B) DCBR process step.

Figure 3. Sketch of the separation process in DCBR. Figure 6. Vessel paths: (A) separation by the BSDCBR system; (B) separation by DCBR.

Figure 4. Vessel paths and x profile: (A) separation by BSDCBR; (B) separation by DCBR.

Figure 7. Vessel paths: separation by TCC.

Figure 5. Sketch of the separation process in a modified TCC.

the RCM, as shown in Figure 1. The vertex toluene is the stable node, while vertices acetone and chloroform are unstable nodes. The RCMs contain two simple distillation, batch rectification,

and stripping regions that coincide. The location of the boundary (between AzLI and H) considerably varies with the pressure; hence, pressure-swing distillation is a feasible method for separating the mixture. 2.2. Column Configurations. 2.2.1. Batch Stripper. The BS (inverted batch column; Figure 2a) is a well-known column configuration in the literature. The charge is fed into a vessel located at the top of the column. The stream (LBS) leaving the top vessel is fed onto the first stage of the column. The top vapor is condensed and fed into the top vessel. The bottom stream (LN) is divided into two parts: one part is the product (W) withdrawn continuously and collected in a tank, and the other part is evaporated (VN) by a heat exchanger and fed into the last stage of the column. 2.2.2. Double Column Batch Stripper. The DCBR (Figure 3) is introduced by Modla and Lang12 for separation of the maximum boiling-point pressure-sensitive azeotropic mixture. The charge is 8206

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research fed into the common vessel located at the bottom of the columns. The liquid leaving the common vessel (Ltotal = LLP þ LHP) is divided into two parts [φL = LLP/(LLP þ LHP)]. In both branches, there is a heat exchanger operated at PLP and PHP, respectively. In the branch of low pressure, it is enough to partially vaporize the liquid because, after the valve, the remaining liquid can be vaporized as a result of a decrease of the pressure. In the branch of high pressure, there is a total reboiler. The liquid streams flowing down from the two rectifying sections are mixed together in the common bottom vessel operated at the higher pressure (PHP). The pressure of the liquid arriving from the column of lower pressure (PLP) must be increased with a pump. The distillates (DLP and DHP) are collected in tanks. 2.2.3. New Triple Column Configuration. The original DCBR contains a common bottom vessel and two rectifier column sections operated at different pressures. The new TCC (Figure 5) is completed with a stripping section. The TCC is the combination of a DCBR and a MVC. Two liquid streams leave the middle vessel. The first stream (LLP) is fed into the low-pressure rectifying column after pressure reduction and evaporation. The second stream (LHP) is fed into the first stage of the stripping column operated at higher pressure. The top vapor of the stripping column is fed into the high-pressure rectifying column. At the top of the rectifying column, the product is continuously withdrawn into a tank. At the bottom of the stripping column, the product is continuously withdrawn into a tank. The main advantage of this column configuration is that no reboiler is needed for the rectifier section of the high-pressure column because its vapor flow is ensured by the top vapor of the stripping column section; hence, energy savings can be realized. 2.3. Feasible Separation Processes. Two different cases are investigated: first the charge composition (xch) is located in the operating region of pressure-swing distillation (AzLILP AzLIHPH triangle), and then it is out of this interval. By analysis of the RCMs, we stated that, at both initial charge compositions, three different processes (column configurations) are feasible. 2.3.1. Charge Composition in the Operating Region of Pressure-Swing Distillation. Separation by BS and DCBR Systems. In the first separation step, the toluene (H) component is removed by a BS operated at low or high pressure (Figure 2a). The vessel composition (S1, Figure 4a) tends to the LI edge, and it arrives between the two azeotropic points (AzLILPAzLIHP, the operating region of pressure-swing distillation). The residue contains acetone and chloroform, which form a maximum boiling-point pressure-sensitive azeotrope so the next feasible bottom product is an azeotrope. Because of this fact, the application of DCBR is suggested to separate the acetone chloroform mixture. The residue of the first step is fed into the common vessel of the DCBR (Figure 2b). The second step is the production of acetone (L) and chloroform (I) by pressure-swing distillation; both columns are operated. The low-pressure column (LP) produces chloroform (I), and the high-pressure column (HP) produces acetone (L). The vessel composition (S2, Figure 4a) does not change during this step. At the end of this step, the common vessel is empty. Separation by DCBR. The initial charge is fed into the common vessel of the DCBR (Figure 3). The low-pressure column (LP) produces chloroform (I), and the high-pressure column (HP) produces acetone (L). The vessel composition tends to the toluene (H) vertex (Figure 4b). At the end of the

ARTICLE

separation, the vessel contains pure toluene. The main advantage of this process is that separate equipment (a BS) is not needed in order to produce toluene. Separation by the New TCC. The initial charge is fed into the common vessel of the TCC (Figure 5). The low-pressure rectifying column (LP) produces chloroform (I), and the highpressure column (HP) produces acetone (L). The stripping column produces toluene (H). At the end of the process, the vessel is empty. During the process, the vessel composition does not change. 2.3.2. Charge Composition out of the Operating Region of Pressure-Swing Distillation. Separation by BS and DCBR Systems. In the first separation step, the toluene (H) component is removed by the BS operated at low or high pressure (Figure 2a). The vessel composition (S1 in Figure 6a) tends to the LI edge, and it arrives out of the two azeotrope points (AzLILPAzLIHP). The residue of the first step is fed into the common vessel of the DCBR (Figure 2b). Because the initial vessel composition is out of the operating region of pressure-swing distillation, a prepurification process step is needed. During the prepurification process step, at least one component is removed from the system in order for the vessel composition to arrive in the operating region of pressure-swing distillation (AzLILPAzLIHPH). In this case, the prepurification process step means the production of acetone (L)—only the high-pressure column is operated— until the vessel composition arrives (S2 in Figure 6a) to the operating region of pressure-swing distillation (AzLILPAzLIHPH). During the third process step, both columns are operated, producing chloroform (low-pressure column) and acetone (high-pressure column). The vessel composition (S3 in Figure 6a) does not change during this step. At the end of this step, the vessel is empty. Separation by DCBR. The initial charge is fed into the common vessel of the DCBR. The first step is a prepurification process step. In this case, the prepurification process step means the production of acetone (L)—only the high-pressure column is operated—until the vessel composition arrives (S1 in Figure 6b) in the operating region of pressure-swing distillation (AzLILP AzLIHPH). In the second process step, both columns are operated, producing chloroform (by a low-pressure column) and acetone (by a high-pressure column). The vessel composition (S2 in Figure 6b) tends to the toluene (H) edge. At the end of this step, the vessel contains toluene. The main advantage of this process is that separate equipment (a BS) is not needed in order to produce toluene. Separation by the New TCC. The initial charge is fed into the common vessel of the TCC. The first step is a prepurification process step, production of acetone (L)—only the high-pressure column is operated—and production of toluene by the stripping column section until the vessel composition arrives (S1 in Figure 7) in the operating region of pressure-swing distillation (AzLILPAzLIHPH). In the second process step, all columns are operated, producing chloroform (by a low-pressure rectifying column), acetone (by a high-pressure rectifying column), and toluene (by a stripping column). The vessel composition (S2 in Figure 7) does not change during this step. At the end of the process, the vessel is empty.

3. RIGOROUS SIMULATION RESULTS However, the feasibility study is a simple, widely used method to determine a feasible process, it has a number of limitations, and 8207

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research

ARTICLE

The liquid stream leaving the common vessel (Ltotal = LLP þ L ) is divided into two parts [φL = LLP/(LLP þ LHP)]. The variable operational parameters (at the given conditions) are the total flow rate of the liquid (Ltotal) leaving the common vessel and the liquid division ratio [φL = LLP/(LLP þ LHP)]. Because the total flow rate has no influence on the product quality, it will not be investigated. 3.1.3. New Triple Column Configuration. The initial charge is fed into the common vessel of the TCC (Figure 5). At the top of the rectifying columns, the products are continuously withdrawn into the tanks (distillate flow rates DLP and DHP and specified product compositions xLPspec and xHPspec). The reflux ratios (RLP and RHP) are changed by PID controllers by manipulating the distillate flow rates DLP and DHP. At the bottom of the stripping column, the product is continuously withdrawn into a tank. The reboil ratio (Rs) is changed by a PID controller by manipulating the bottom flow rate (W). The variable operational parameters (at the given conditions) are the total flow rate of the liquid leaving the common vessel and the liquid division ratio (φL). Because the total flow rate has no influence on the product quality, it will not be investigated. 3.2. Input Data. At the BS, the column works at PBS = 1.01 bar. At the DCBR, the columns work at PLP = 1.01 bar and PHP = 10 bar. At the TCC, the rectifying column sections work at PLP = 1.01 bar and PHP = 10 bar and the stripping column section works at PBS = 10 bar. The number of theoretical stages (N) for each column section is 40. (The total condenser and total reboiler do not provide a theoretical stage.) The liquid holdup is 2 dm3/plate. In each case, the quantity of the charge is 0.785 m3. At the start, the columns are filled with boiling point liquid at the operation pressure of the column (“wet start-up”). Such as at the feasible study, two different cases are investigated: first the charge composition is located in the operating region of pressure-swing distillation (AzLILPAzLIHPH triangle; xch = 0.5, 0.3, and 0.2), and then it is out of this interval (xch = 0.3, 0.3, and 0.4). The specified purities are 98 mol % for all components. 3.3. Charge Composition in the Operating Region of Pressure-Swing Distillation. 3.3.1. Separation by BS and DCBR Systems. Batch Stripping Process Step: Production of Toluene. The first separation step is done in the BS (Figure 2a). During this step, toluene is withdrawn as the bottom product. The total flow rate of liquid leaving the top vessel is 1.00 m3/h. The influence of the toluene remaining in the top vessel on the overall specific energy consumption (SQ/SPrT) is investigated. The specified criteria for the concentration of toluene in the top vessel (xv) at the end of the process are less than 2.0/1.0/0.5/ 0.2 mol %. The calculation results are presented in Table 1. In Figure 8, the evolution of the reboil ratio (Rs) is presented. It can be seen that from 145 min the reboil ratio increases sharply, which causes the overall energy consumption to rise considerably as well. The results show that there is a significant correlation between the toluene concentration, reflux ratio, and overall specific energy consumption; hence, to define the end point of the process of key importance in the industrial practice. Double-Column Batch Rectifying Process Step: Production of Chloroform and Acetone. At the beginning of the process step, the common vessel is filled with the vessel residue of the BS step. Its quantity is 0.785 m3. (It is collected from more BS step HP

Figure 8. Evolution of the reboil ratio (BS).

hence confirmation of the results is essential. For verifying the results, we perform a rigorous simulation calculation. Furthermore, in this section, the influence of the most important operational parameters of the processes on the overall specific energy consumption (SQ/SPr) is investigated and the performances of the different column configurations are compared. The following simplifying assumptions are applied: theoretical trays, constant volumetric liquid holdup on the trays, and negligible vapor holdup. The model equations to be solved are well-known: (a) nonlinear differential equations (material balances, heat balances, etc.) and (b) algebraic equations [vaporliquid equilibrium (VLE) relationships, summation equations, holdup, and physical property models]. For a description of phase equilibria, the UNIQUAC model is applied (VLE and UNIQUAC parameters can be found in the Appendix). For the solution of the above equations, the dynamic simulator of ChemCad 6.220 (program CC-DCOLUMN) is applied. Column sections are modeled by the module DYNAMIC COLUMN and the common vessel and product tanks by DYNAMIC VESSEL. Besides these units, the flowsheet still contains HEAT EXCHANGERs (condensers), MIXERs, DIVIDERs (stream splitters), PUMPs, VALVEs, CONTROLLERs, and CONTROL VALVEs. 3.1. Operation of the Column Configurations. 3.1.1. Batch Stripper. The BS (Figure 2a) can be operated under a fixed reboil ratio (the product quality changes during the process) or a fixed bottom product quality (the reboil ratio changes during the process). In this paper, the product quality is specified and fixed; hence, the reboil ratio (Rs) is changed by a PID controller by manipulating the bottom flow rate (W) in order to keep the constant toluene concentration in the bottom flow. The variable operational parameter (at the given conditions) is the flow rate of the stream (LBS) leaving the top vessel. Because the flow rate of the stream has no influence on the product quality, it will not be investigated. Moreover, it is worth studying the end point of the process, which is determined as the residue quality, because it has an influence on the overall specific energy consumption; hence, it will be investigated. 3.1.2. Double Column Batch Stripper. The DCBR (Figure 3) can be operated under fixed reflux ratios (RLP and RHP) or fixed distillate qualities (distillate flow rates DLP and DHP; specified product compositions xLPspec and xHPspec). In this paper, the product qualities are specified (xLPspec and HP x spec); hence, the reflux ratios (RLP and RHP) are changed by PID controllers by manipulating the distillate flow rates (DLP and DHP).

8208

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research

ARTICLE

Figure 9. Overall specific energy consumption: (A) influence of the initial toluene concentration; (B) influence of the liquid division ratio.

Figure 10. Calculated vessel paths (BS and DCBR).

Figure 11. Influence of the liquid ratio division on the overall specific energy consumption (DCBR).

residue so that the same equipment size can be applied.) The lowpressure column produces chloroform, and the high-pressure column produces acetone (Figure 3). The whole process is finished when the amount of liquid in the common vessel decreases to 5 vol % of the initial quantity of the vessel liquid. The influence of the liquid division ratio (φL) as an operational parameter on the performance of the process is studied for different initial toluene concentrations (xv,T). The optimum value yielding the minimal overall specific energy consumption (SQ/SPr) is determined (Figure 9b). The liquid division ratio is varied in the region 0.10.9. When the toluene concentration (xv,T) in the vessel residue of the BS process step is decreased, the overall energy consumption of the BS step increases considerably, while the overall energy consumption of the DCBR step slightly decreases; hence, the total (BS þ DCBR) overall specific energy consumption increases (Figure 9a). The best result, 536 MJ/kmol, is obtained at φL = 0.57 and xv,T = 0.02. The specific overall energy consumption of the BS and DCBR process steps is 815 MJ/kmol. The calculated vessel path is presented in Figure 10. As was mentioned in section 2.3.1, during the batch stripping process, the vessel path (S1) tends from charge composition (xch) to the chloroformacetone edge, but on the basis of the simulation, it cannot reach this edge because of the toluene content remaining. Thanks to the remaining toluene, the vessel composition is not constant (Figure 10, vessel path S2) during the next process step (producing chloroform and acetone in DCBR).

3.3.2. Separation by DCBR. The total flow rate of liquid leaving the common vessel (LLP þ LHP; Figure 3) is 8.78 m3/h. The low-pressure column produces chloroform, and the highpressure column produces acetone. The whole process is finished when the toluene concentration reaches 98 mol % in the vessel liquid. The influence of the liquid division ratio as an operational parameter on the performance of the process is studied. The optimum value yielding the minimal overall specific energy consumption (SQ/SPr) is determined (Figure 11). The liquid division ratio is varied in the region 0.10.9. The best result, 696 MJ/kmol, is obtained at φL = 0.57. The calculated vessel path and x profiles (process time: 60 min) are presented in Figure 12. As was mentioned in section 2.3.1 (the feasibility section), during the process, the vessel path S1 tends from charge composition (xch) to the toluene vertex. The column profile presents that the two columns operate in different distillation regions thanks to the different operation pressures. 3.3.3. Separation by the New TCC. At the beginning of the process, the common vessel is filled with charge (Figure 5). The quantity of the charge is 0.785 m3. The low-pressure rectifying column produces chloroform, and the high-pressure column produces acetone. The stripping column produces toluene. The whole process is finished when the amount of the liquid in the common vessel decreases to 5 vol % of the charge. At different liquid division ratios (φL), the minimum value of the overall energy consumption is determined (Figure 13). 8209

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research

ARTICLE

Figure 14. Calculated vessel path (TCC). Figure 12. Calculated vessel path and x profiles (DCBR).

Table 2. Comparison of the Efficiencies of the Different Column Configurations SQ/SPr [MJ/kmol]

savings [%]

BSDCBR

815

DCBR

696

15

TCC

345

58

Table 3. Calculated Results of the Batch Stripping Process Step

Figure 13. Influence of φL on the overall specific energy consumption (TCC).

xv of

SQ/SPr [MJ/

process time

Rs

chloroformtolueneacetone

kmol]

[min]

final

0.3949, 0.0199, 0.5852

289

148.5

48

0.3993, 0.0099, 0.5908 0.4015, 0.0049, 0.5936

341 395

177.0 205.0

97 195

0.4029, 0.0019, 0.5952

468

241.5

486

Table 1. Calculated Results of Batch Stripping Process Step Table 4. Comparison of the Efficiencies of the Different Column Configurations

SQ/SPrT

process time

Rs

xv of chloroformtolueneacetone

[MJ/kmol]

[min]

final

0.7122, 0.0198, 0.2680

278

145.0

47

0.7205, 0.0099, 0.2696

328

172.5

BSDCBR

783

0.7247, 0.0049, 0.2704

380

200

188

DCBR

617

21

0.7273, 0.0019, 0.2708

448

235

463

TCC

293

63

SQ/SPr [MJ/kmol]

94

From the graph (Figure 13), we can see that the curve has the same minimum point as that observed in the DCBR process. The best result, 345 MJ/kmol, is obtained at φL = 0.4. The calculated vessel path is presented in Figure 14. This result is in contrast to the result of the feasible study (section 2.3.1). The calculated vessel path is not constant; it tends to the chloroformacetone edge because the speed of toluene removal is higher than those of chloroform and acetone. In Table 2, the calculation results are summarized. As Table 2 shows, there is a significant difference between the overall energy consumptions of the different processes. It can be seen that the most efficient configuration is the TCC, and the least one is the BSDCBR. On the basis of this result, we can state that with application of the TCC more than 50% energy savings can be realized compared with the BSDCBR.

savings [%]

3.4. Charge Composition out of the Operating Region of Pressure-Swing Distillation. 3.4.1. Separation by BS and DCBR Systems. Batch Stripping Process Step: Production of Toluene.

The first separation step is done in the BS. During this step, toluene is withdrawn as the bottom product. The total flow rate of liquid leaving the top vessel is again 1.00 m3/h. The influence of the remaining toluene in the top vessel on the overall specific energy consumption (SQ/SPrT) is again investigated. The specified criteria for the concentration of toluene in the top vessel (xv) at the end of the process are less than 2.0/1.0/0.5/ 0.2 mol %. The calculation results are presented in Table 3. In Figure 15, the evolution of the reboil ratio (Rs) is presented. We get the same result as that in section 3.3.1. The reboil ratio increases sharply from 148.5 min, which causes the overall energy consumption to rise considerably as well (Table 3). 8210

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research

ARTICLE

The results show that there is a significant correlation between the toluene concentration, reflux ratio, and overall specific energy consumption. Double-Column Batch Rectifying Process Step: Production of Chloroform and Acetone. The initial quantity of the vessel liquid is 0.785 m3. (It is collected from more BS step residue so that the same equipment size can be applied.) The initial vessel composition equals that of the BS process step residue (0.3949, 0.0199, and 0.5852). Because the initial composition of this process step is out of the operating region of pressure-swing distillation, a prepurification step is needed. During this step, only the highpressure column is operated, producing acetone. The end point of this step is an operational parameter, which is determined on the basis of the acetone concentration in the vessel (xAc). The whole process is finished when the toluene concentration reaches 98 mol % in the vessel liquid. The influence of the liquid division ratio and the end point of the prepurification process step as the operational parameters on the performance of the process are studied. The optimum values of xAc and φL yielding the minimal overall specific energy consumption (SQ/SPr) are determined (Figure 16). By varyiation of the end point of the prepurification process step (xAc; in the region 0.160.41) at different liquid division ratios (φL = 0.34, 0.45, 0.57, 0.68, and 0.74), the

minimum overall specific energy consumptions are determined (Figure 16). The best result, 783 MJ/kmol, is obtained at xAc = 0.3 and φL = 0.57. On the basis of the results, we can state that the end point of the prepurification process step (Figure 16a) and the division ratio (Figure 16b) have a significant influence on the overall specific energy consumption. The calculated vessel path is presented in Figure 17. During the batch stripping process, the vessel path S1 tends from charge composition (xch) to the chloroformacetone edge, but it can not reach it. Thanks to the remaining toluene, the vessel composition is getting farther from the edge (Figure 18, vessel paths S2 and S3) during the next process steps. 3.4.2. Separation by DCBR. The total flow rate of liquid leaving the common vessel (Ltotal) is again 8.78 m3/h. The whole process is finished when the toluene concentration reaches 98 mol % in the vessel. Because the initial charge composition is out of the operating region of pressure-swing distillation, a prepurification step is needed when only the high-pressure column is operated, producing acetone until the vessel composition arrives (S1 in Figure 18) into the operating region of pressure-swing distillation. The end point of the prepurification step is an operational parameter that is determined based on the acetone concentration in the vessel (xAc). During the next step, both columns are operated, producing chloroform (low-pressure column) and acetone (high-pressure

Figure 15. Evolution of the reboil ratio (BS).

Figure 17. Calculated vessel composition (BSDCBR).

Figure 16. Overall specific energy consumptions: (A) influence of the end point of the prepurification step (xAc); (B) influence of φL (DCBR process step). 8211

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research

ARTICLE

column). The vessel composition (S2 in Figure 18) tends to the toluene edge. At the end of the whole process, the vessel contains toluene in the prescribed purity. The influence of the liquid division ratio (φL) and the end point of the preliminary step (xAc) as operational parameters on the performance of the process are studied. The optimum value yielding the minimal overall specific energy consumption (SQ/SPr) is determined (Figure 19). By variation of the end point of the preliminary step (xAc; in the region 0.030.12) at different liquid division ratios (φL = 0.34, 0.45, 0.51, 0.54, 0.57, 0.68, and 0.74), the minimum overall specific energy consumptions are determined (Figure 19). The best result, 783 MJ/kmol, is obtained at xAc = 0.06 and φL = 0.54. We can state that the end point of the preliminary step (between xAc = 0.040.12; Figure 19a) has a slight influence on the overall specific energy consumption, while the division ratio (Figure 19b) has a considerable influence on it. 3.4.3. Separation by TCC. At the beginning of the process, the common vessel is filled with charge. The quantity of the charge is 0.785 m3. The low-pressure rectifying column produces chloroform, and the high-pressure rectifying column produces acetone. The stripping column produces toluene. The whole process is finished when the amount of the liquid in the common vessel decreases to 5% of the charge. Because the initial charge composition is out of the operating region of pressure-swing distillation, a prepurification process step is needed when only the high-pressure column sections are operated, producing acetone (rectifying section) and

toluene (stripping section) until the vessel composition arrives (S1 in Figure 20) in the operating region of pressureswing distillation. The end point of the prepurification process step is an operational parameter that is based on the chloroform concentration in the vessel (xChl). The minimum value of the overall energy consumption is determined at different division ratios of the liquid leaving the common vessel (φL) by variation of the end point of the prepurification step (xChl = 0.40.68; Figure 21). The best result, 293 MJ/kmol, is obtained at φL = 0.57 and xChl = 0.71. In Table 4, the calculation results are summarized. As Table 4 shows, there again is a significant difference between the overall energy consumptions of the different processes. It can be seen that the most efficient configuration is the TCC, and the least one is the BSDCBR. On the basis of the result, we can state that with application of the TCC more than 60% energy savings can be realized thanks to the fact that the top vapor of the stripping section heats the high-pressure rectifying section compared with the BSDCBR. 3.5. Thermal Integration. In our work, we focus on the energy reduction in the different batch processes. Thermal integration can be realized at the DCBR and TCC as well because the boiling point of the acetone product (top stream temperature of the high-pressure column; 142.9 °C at 10 bar) is higher than the boiling point of the mixture (bottom stream temperature of the low-pressure column;

Figure 18. Calculated vessel composition (DCBR).

Figure 20. Calculated vessel composition (TCC).

Figure 19. Overall specific energy consumption: (A) influence of the end point of the preliminary step (xAc); (B) influence of φL. 8212

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research

ARTICLE

Figure 21. Overall energy consumption (TCC): (A) influence of xChl; (B) influence of φL.

less than 110 °C at 1.01 bar), so heat integration could be attractive in terms of energy consumption. Figure 22 illustrates thermal integration at the DCBR. The same solution was applied at the TCC. The top vapor of the high-pressure rectifying section (VHP,top) is fed into an economizer. The economizer is used to recover energy from the top vapor by partial vaporization of the reboil stream of the low-pressure rectifying section (LLP). The operational parameter of the economizer is the temperature of the hotstream outlet (the recycled top vapor from the high-pressure column is higher by 11 °C than that of the cold-stream outlet). The economizer is completed with an auxiliary reboiler in order to totally vaporise the reboil stream. At the top of the highpressure column there is a auxiliary condenser in order to complete the condensation. At both column configurations, significant energy reduction can be realized with thermal integration. Figure 23 shows that at the DCBR the best result is 457 MJ/ kmol (at φL = 0.45), while at the TCC, it is 213 MJ/kmol (at φL = 0.45), which mean 35% (DCBR) and 39% (TCC) energy savings. 3.6. Cost Analysis. In the previous section, we found that separation with the TCC column is more cost-effective than that with the DCBR. In this section, a simplified payback period (PBP) is calculated in order to analyze the investment of the third column that is needed for the TCC column configuration. PBP ¼ investment cost ðICÞ=annual cost savings ðACSÞ The investment cost of this column is calculated based on work by Douglas.21 The column configuration capacity is 7200 kmol/year, and the unit cost of energy is 6$/GJ. The effect of the plate number is investigated (Figure 24). The results show that if the plate number is fewer than 10, the annual cost savings decreases sharply (Figure 24a) and the PBP is less than 15 years, which is acceptable (Figure 24b); hence, we can state that application of the TCC is economical.

4. SUMMARY This study set out to introduce a new TCC for separation of a ternary chloroformacetonetoluene mixture by pressure-

Figure 22. Scheme of a DCBR with thermal integration.

swing distillation. The TCC is the combination of a DCBR and a MVC. For separation of this ternary mixture, three different column configurations (separation processes) were applied, investigated, and compared at two different charge compositions. The studied column configurations were the BSDCBR, DCBR, and TCC. All columns were operated in an open mode. The influence of the most important operational parameters was studied for all processes, and the performances of the different column configurations were compared. This study has shown the main advantages of the DCBR: compared with BSDCBR, separate equipment (BS) is not needed to produce toluene and hence energy savings can be realized, and compared with the TCC, there is no need for a stripping section and hence the investment cost is lower. On the other hand, it was also shown that the main advantage of the TCC is that no reboiler is needed for the rectifier section of the high-pressure column because its vapor flow is ensured by the top vapor of the stripping column section; hence, energy savings can be realized. This energy reduction was presented by a rigorous simulation calculation. Moreover, we found that in the BSDCBR system the lower toluene concentration in the vessel at the end of the BS step does not improve the performance of the DCBR process step to such 8213

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research

ARTICLE

Figure 23. Effect of thermal integration: (A) DCBR; (B) TCC.

Figure 24. Effect of the plate number of the third column: (A) on the costs; (B) on the PBP.

an extent that the increase of energy consumption of the BS step is compensated for. Results of the rigorous simulations showed that the acetone concentration in the vessel at the end of the prepurification process step of the DCBR has a significant influence on the overall specific energy consumption. Furthermore, on the basis of the results obtained, we stated that the DCBR is more efficient by almost 20% and the TCC is more efficient by at least 40% compared with the BSDCBR. The other fact is that thermal integration can be realized at the DCBR and TCC as well by which energy reduction can be realized. This study has found that in general, although the operational cost of the TCC is the lowest, its investment cost is the highest and its operation is the most complicated because more columns are operated. The cost analysis shows that the PBP of the installation of a 10-plate column is 15 years, so application of the TCC is economical. These findings suggest, in general, that a significant energy reduction can be realized with modification, improving the column configuration (batch process) in order for it to be more suitable for the separation problem.

a Antoine parameters: lnðpÞ ¼ A 

B TþC

where p is the vapor pressure (mmHg) and T the temperature (K). component A B C chloroform toluene acetone

16.516 16.266 16.732

2938.6 3242.4 2975.9

36.997 47.181 34.523

b UNIQUAC Parameters for Chloroform (A)Toluene (B)Acetone (C) i j uij  ujj [cal/mol] uji  uii [cal/mol] A A B

B C C

554.887 726.9578 555.74

860.821 1147.662 315.28

Appendix 2: Values of the PID Parameters Used for Rigorous Calculations

column PB [%] TI [min] LP 0.95 0.9 HP 0.97 0.9 SC 0.99 0.9

’ APPENDIX Appendix 1: Values of the Parameters Used for the Phase Equilibrium Calculations 8214

TD [min] 13 13 13

set point [mol/mol] 0.98 methanol 0.98 acetone 0.98 toluene

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215

Industrial & Engineering Chemistry Research

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT This work was financially supported by the Hungarian Scientific Research Fund (OTKA) (Grant K-82070) and by the Janos Bolyai Research Scholarship of the HAS. The author is grateful to Prof. Peter Lang for his valuable help. ’ NOTATION Az azeotrope BS batch stripper D distillate flow rate, m3/h DCBR double-column batch rectifier L liquid flow rate, m3/h L, I, H lightest, intermediate, and heaviest components N plate number P pressure, bar reboil ratio Rs R reflux ratio S vessel composition path SQ/SPr overall specific energy consumption, MJ/kmol t time, min TCC triple-column configuration V vapor flow rate, m3/h W bottom flow rate, m3/h x liquid mole fraction y vapor mole fraction Greek Letters

O

division

Subscripts

Ac ch Chl spec e L T v 1N

acetone charge chloroform specified value end liquid toluene vessel plate index

ARTICLE

(5) Knapp, J. P.; Doherty, M. F. A new pressure-swing-distillation process for separating homogeneous azeotropic mixtures. Ind. Eng. Chem. Res. 1992, 31, 346–357. (6) Luyben, W. L. Comparison of Pressure-Swing and Extractive Distillation Methods for Methanol Recovery Systems. Ind. Eng. Chem. Res. 2005, 44 (15), 5715–25. (7) Luyben, W. L. Control of the Maximum-Boiling Acetone/ Chloroform Azeotropic Distillation System. Ind. Eng. Chem. Res. 2008, 47 (8), 6140–6149. (8) Luyben, W. L. Comparison of Extractive Distillation and Pressure-Swing Distillation for AcetoneMethanol Separation. Ind. Eng. Chem. Res. 2008, 47 (8), 2696–2707. (9) Luyben, W. L. Design and Control of a Fully Heat-Integrated Pressure-Swing Azeotropic Distillation System. Ind. Eng. Chem. Res. 2008, 47, 2681–2695. (10) Phimister, J. R.; Seider, W. D. Semicontinuous, Pressure-Swing Distillation. Ind. Eng. Chem. Res. 2000, 39, 122–130. (11) Repke, J. U.; Klein, A.; Bogle, D.; Wozny, G. Pressure-Swing Batch Distillation for Homogenous Azeotropic Separation. Chem. Eng. Res. Des. 2007, 85 (4), 152. 492–501. (12) Modla, G.; Lang, P. Feasibility of new pressure-swing batch distillation methods. Chem. Eng. Sci. 2008, 63 (11), 2856–2874. (13) Modla, G; Lang, P. Separation of ternary homoazeotropic mixture by pressure swing distillation. Hung. J. Ind. Chem. 2008, 36 (12), 89–94. (14) Kopasz, A.; Modla, G.; Lang, P. Operation and Control of a New Pressure-Swing Batch Distillation System. Compu.-Aided Chem. Eng. 2009, 27, 1503–1508. (15) Modla, G.; Lang, P.; Denes, F. Feasibility of separation of ternary mixtures by pressure-swing batch distillation. Chem. Eng. Sci. 2010, 65 (2), 870–881. (16) Modla, G.; Lang, P. Separation of an AcetoneMethanol Mixture by Pressure-Swing Batch Distillation in a Double-Column System with and without Thermal Integration. Ind. Eng. Chem. Res. 2010, 49 (8), 3785–3793. (17) Modla, G. Pressure-swing batch distillation by double column systems in closed mode. Comput. Chem. Eng. 2010, 34, 1640–1654. (18) Modla G. Reactive pressure swing batch distillation by a new double column system. Comput. Chem. Eng. 2011, DOI: 10.1016/j. compchemeng.2011.01.002. (19) Serafimov, L.A. The Azeotropic Rule and the Classification of Multicomponent Mixtures VII. Diagrams for Ternary Mixtures. Russ. J. Phys. Chem. 1970, 44 (4), 567–571. (20) CHEMCAD Dynamic Column Calculation User’s Guide;Chemstations Inc.: Houston, TX, 2007. (21) Douglas, J. M Conceptual Design of Chemical Processes; McGrawHill: New York, 1989.

Superscripts

LP HP BS L, I, H

low-pressure (column index) high-pressure (column index) batch stripping (column index) lightest, intermediate, and heaviest component indices

’ REFERENCES (1) Lewis W. K. Dehydrating Alcohol and the Like. U.S. Patent 1,676,700, July 10, 1928. (2) Black, C. Distillation Modelling of Ethanol Recovery and Dehydration Processes for Ethanol and Gasahol. Chem. Eng. Prog. 1980, 76, 78–85. (3) Abu-Eishah, S. I.; Luyben, W. L. Design and Control of TwoColumn Azeotropic Column Azeotropic Distillation System. Ind. Eng. Chem. Process Des. Dev. 1985, 24, 132–140. (4) Chang, T.; Shih, T. T. Development of an Azeotropic Distillation Scheme for Purification of Tetrahydrofuran. Fluid Phase Equilib. 1989, 52, 161–168. 8215

dx.doi.org/10.1021/ie101578j |Ind. Eng. Chem. Res. 2011, 50, 8204–8215