Separation of isobutyl alcohol and isobutyl acetate by extractive distillation and pressure-swing distillation: Simulation and optimization

Separation of isobutyl alcohol and isobutyl acetate by extractive distillation and pressure-swing distillation: Simulation and optimization

Separation and Purification Technology 50 (2006) 175–183 Separation of isobutyl alcohol and isobutyl acetate by extractive distillation and pressure-...

503KB Sizes 327 Downloads 268 Views

Separation and Purification Technology 50 (2006) 175–183

Separation of isobutyl alcohol and isobutyl acetate by extractive distillation and pressure-swing distillation: Simulation and optimization R. Mu˜noz ∗ , J.B. Mont´on, M.C. Burguet, J. de la Torre Departamento de Ingenier´ıa Qu´ımica, Escuela T´ecnica Superior de Ingenier´ıa, Universitat de Val`encia, Dr. Moliner, 50, 46100 Burjassot, Valencia, Spain Received 21 September 2005; received in revised form 22 November 2005; accepted 22 November 2005

Abstract We have studied, simulated and evaluated economically two separation alternatives of a mixture made up of 52 mole% of isobutyl alcohol and 48 mole% of isobutyl acetate by means of a practical case of a plant to treat 12,000 Tm/year of the original mixture. The simulation has been carried out satisfactorily by means of a package of commercial software (Aspen HYSYS® ) using the thermodynamic model UNIQUAC with binary parameters obtained experimentally by us. The two processes evaluated (extractive distillation using n-butyl propionate as a solvent and pressure-swing distillation) have been optimized independently from each other and the best configurations have been evaluated economically. The simulation and economic evaluation of the two separation alternatives that we have considered allow us to conclude that, for a 12,000 Tm/year plant, the pressure-swing distillation is more attractive than the extractive distillation using n-butyl propionate as an entrainer. © 2005 Elsevier B.V. All rights reserved. Keywords: Extractive distillation; Pressure-swing distillation; Simulation; Isobutyl alcohol; Isobutyl acetate; n-Butyl propionate

1. Introduction Isobutyl acetate (IBAc) is a solvent widely used in Chemical Industry. It is used alone or in solvent blends in applications including coatings, inks, adhesives, industrial cleaners and degreasers. The IBAc is produced by estherification of acetic acid with isobutyl alcohol (IBA). Final purification of acetate by traditional technologies is a relatively complex procedure due to the existence of a minimum boiling point azeotrope in the IBA + IBAc mixture at atmospheric pressure. Azeotropes are non-ideal mixtures whose components are very difficult and, hence, expensive to separate. This can be

Abbreviations: BUP, butyl propionate; DMF, dimethylformamide; EC, extractive column; ED, extractive distillation; FCI, fixed capital investment (D 103 ); HPC, high pressure column; IBA, isobutyl alcohol; IBAc, isobutyl acetate; LPC, low pressure column; PSD, pressure-swing distillation; RHD, reboiler heat duty (MJ/h); SRC, solvent recovery column; TAC, total annual costs (D 103 /year); VLE, vapour–liquid equilibrium ∗ Corresponding author. Tel.: +34 963544319; fax: +34 963544898. E-mail address: [email protected] (R. Mu˜noz). 1383-5866/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.seppur.2005.11.022

overcome by several techniques including azeotropic and extractive distillation [1–3], reactive distillation [4,5], liquid–liquid extraction [6], adsorption [7], membrane pervaporation [8], salt addition [9] and pressure-swing distillation [10]. In this work, only extractive distillation (ED) and pressure-swing distillation (PSD) will be considered. ED can be used to separate the components of an azeotropic mixture adding an agent (entrainer) that modifies the relative volatility of the mixture. Also, PSD can be used to recover pure components with a simple change in pressure, a fact which results in a change of the azeotropic composition, provided that it is pressure-sensitive. Laboratory experiments in either extractive distillation or pressure-swing distillation are time-consuming and expensive because of the large number of parameters involved. It would be desirable to predict the experimental data with the help of available simulation programs. Computer simulations using commercial process simulators have been used with success as an aid for process development. They were used to set up the guidelines for further pilot experiments and moreover, to optimize the operating parameters governing the process at steady-state.

176

R. Mu˜noz et al. / Separation and Purification Technology 50 (2006) 175–183

Nomenclature A C im ir K

UNIQUAC binary interaction parameters (cal/mol) costs (D 103 /year) minimum acceptable rate of return fixed capital recovery rate (depreciation rate) vapour–liquid equilibrium constant

Greek letter α relative volatility Subscripts f fixed i, j primitive mixture components S solvent (entrainer) v variable Superscript ∞ infinite dilution

The synthesis and design of extractive distillation processes take place in two steps [11]. The first one involves the selection of one or more candidate solvents (which facilitate the separation by changing the relative volatilities in the mixture through physical or chemical interactions with the original components), and the choice of one or more column configurations. The second step, process design, involves the search for optimal process parameter values. The success of the second step depends on the solutions obtained for the first one because efficiency in extractive distillation is largely determined by the choice of a suitable entrainer. In this work, based on the guidelines for the solvent screening, we have chosen three solvents: N,N-dimethylformamide (DMF), 1-hexanol and butyl propionate (BUP). DMF was recommended as a potential entrainer for alcohol–acetate azeotropic mixtures because of its high polarity [12] and 1-hexanol and BUP have been chosen because they are, respectively, in the same homologous series with one of the key-components [13]. Therefore, in order to be able to select the best solvent among them, we have carried out simulations with Aspen HYSYS® v3.2 of Aspen Technology Inc., using the binary interaction parameters correlated from experimental data obtained for all binaries involved [14–17]. According to the results obtained, the best solvent seems to be butyl propionate. Once the solvent has been selected, we have designed the separation sequence and optimized the operating parameters. On the other hand, to investigate how the pressure-swing distillation works with the IBA + IBAc azeotropic system, we have done a simulation of the vapour–liquid equilibrium using DISTIL v5.0 of Hyprotech Ltd. at different pressures with the interaction parameters obtained from experimental VLE data [14]. Based on these results we have decided to carry out the design and optimization of the pressure-swing distillation process. The aim of this work is to study the influence of the operation variable values and column configuration on the performance of

the IBA + IBAc separation by extractive distillation with BUP as entrainer and by swing-pressure distillation, with the help of a commercial simulator (Aspen HYSYS® v3.2 of Aspen Technology Inc.). Finally, we have chosen the best alternative for the separation of the azeotropic mixture under study from the economic point of view. 2. Simulation 2.1. Problem definition The two alternatives considered in this study (ED and PSD) were simulated with the same basic data. The feed is a mixture made up of 52 mole% of isobutyl alcohol and 48 mole% of isobutyl acetate, with a flow rate of 12,000 Tm/year; we took 8000 working hours per year, that is a mass flow of 1500 kg/h. 2.2. Property package Computer simulation using commercial process simulators is a useful tool to predict qualitatively the influence of the operating variables on the column performance, provided that the interaction binary parameters for the studied mixture are available in their own data-bank. The accuracy of the simulated results is strongly dependent on the quality of the binary parameters for the liquid-phase activity coefficient models. In this paper, the simulation was undertaken with HYSYS and DISTIL. UNIQUAC activity model was chosen because it was the most suitable, but unfortunately, no data exist in their own library to cover all possible interactions between the components studied in this work, so we have used the binary interaction parameters published by us in previous papers [14–17]. The parameters used are listed in Table 1. 2.3. Extractive distillation 2.3.1. Solvent selection Since the solvent is the core of extractive distillation, more attention should be paid on the selection of potential solvents. Of all possible entrainers that can be used for the separation of IBA and IBAc azeotrope mixture we have chosen three: N,N-dimethylformamide, 1-hexanol and butyl propionate. DMF was recommended as a potential entrainer for alcohol–acetate azeotropic mixtures because of its high polarity and 1-hexanol and BUP have been chosen because they are, respectively, in the same homologous series with one of the key-components. Table 1 UNIQUAC binary interaction parameters Component i

Component j

Aij (cal/mol)

Aji (cal/mol)

IBA IBA IBA IBA IBAc IBAc IBAc

IBAc DMF 1-Hexanol BUP DMF 1-Hexanol BUP

116.17–0.53Ta 56.603 461.869 35.346 551.320 316.172 131.014

361.90–0.45Ta −155.920 −340.476 49.224 −281.210 −227.902 −128.309

a

T is the temperature in Kelvin.

R. Mu˜noz et al. / Separation and Purification Technology 50 (2006) 175–183

177

tion and the relative volatility. The greater the relative volatility value, the easier the separation. In Table 2, it can be seen that DMF reverses the volatility of the primitive mixture (αSi,j < 1). 1-Hexanol alters only slightly the activity coefficients of the components to be separated, BUP being the best entrainer since it shows a higher relative volatility and therefore it will allow a separation sequence not too expensive.

Fig. 1. VLE data plotted on a solvent free basis for the system IBA (1) + IBAc (2) + solvent (3) at 101.3 kPa. Continuous line for x3 = 0; dashed lines simulated with UNIQUAC model with parameters given in Table 1 for x3 = 0.7: (- · · -) BUP, (- - -) 1-hexanol and (- · -) DMF.

In order to verify the effect of the aforementioned solvents, the VLE of the ternary mixtures, on a solvent free basis, was simulated and plotted in Fig. 1. It may be observed that DMF reverses the volatility of the original mixture, that is, enhances the relative volatility of IBAc with regard to IBA in such a way that IBAc would be obtained as the overhead product in the extractive column, IBA and DMF being the bottom products. On the contrary, 1-hexanol and BUP enhance the relative volatility of IBA with regard to IBAc, both in the natural way, that is, IBA would be obtained as the overhead product and IBAc together with the solvent (1-hexanol or BUP) would be obtained as the bottom product. This figure confirms that BUP brings about a larger enhancement of the relative volatility, so that BUP is the best promising entrainer for the separation of IBA and IBAc azeotropic mixture by extractive distillation. Another criterion for solvent selection is through the relative volatility in presence of a solvent [18–20], defined as: αSi,j =

∞ Ki,S

(1)

∞ Kj,S

∞ is the infinite dilution K-value for trace of species i in where Ki,S ∞ is the infinite dilution K-value for trace of the solvent and Kj,S species j in the solvent. Table 2 lists the K-values at infinite dilu-

Table 2 K-values at infinite dilution and relative volatility in presence of a solventa Solvent

∞ Ki,S

∞ Kj,S

αSi,j

DMF 1-Hexanol BUP

2.8 5.2 4.4

6.0 3.7 2.1

0.46 1.40 2.10

a

Simulated values with DISTIL using UNIQUAC model.

2.3.2. Sequencing of the extractive distillation process After the entrainer has been selected, in this work we have chosen the butyl propionate, attention is directed to the sequence of the distillation towers. The process configuration is shown in Fig. 2, in which the solvent is added at the top trays of the extractive column (EC). In this column, BUP increases the volatility of IBA with respect to IBAc and thus makes the separation easier. Since BUP is much less volatile than either IBA or IBAc, it flows down the column to leave with the bottom product. The solvent recovery column (SRC) removes IBAc from BUP. This is an easy separation because the solvent is much less volatile than IBAc. The lean solvent is then cooled and recycled back to the extractive column. A low impurity solvent is needed for the recycling, so as not to adversely affect the performance of the extractive column. If the recovery of the solvent is high a very small amount of solvent make-up is required to maintain the solvent-to-feed ratio constant. 2.3.3. Optimization 2.3.3.1. Partial optimization based on the total reboiler heat duty as a reference variable. In order to select the best conditions to carry out the global economic optimization, we start with a partial optimization specifying some variables and using the total reboiler duty as a reference variable. The variables chosen to be specified can be characterized as either design variables or optimization variables. Design variables are those whose values are set by market demands or physical conditions. In our study we specified the temperature, pressure, flow rate and composition of binary feed. We also specified distillate purity and recovery of IBA in the first column, and distillate purity of IBAc and bottom purity of solvent in the second column. Finally, solvent make-up was chosen as a pure component at ambient conditions. All the previously mentioned variables are taken as design variables in this problem and the selected values are shown in Table 3. Once design variables have been specified, their values remain constant throughout the optimization procedure. On the contrary, optimization variables are those that must arbitrarily be assigned a value. The values are subject to change as we proceed from the base case to the optimal design. In this problem, the solvent-to-feed ratio in the extractive column and the number of trays in each column are optimization variables. The optimization procedure using HYSYS requires setting the number of trays in both columns. With the short-cut design facility of HYSYS we did a case study analyzing the variation of stage number and reboiler heat duty (RHD) as a function of the reflux ratio, in order to set the number of ideal trays and feed position of the solvent recovery column. Fig. 3 shows the results

178

R. Mu˜noz et al. / Separation and Purification Technology 50 (2006) 175–183

Fig. 2. HYSYS process flow diagram (PFD) of extractive distillation.

and it looks like a good decision to set the number of ideal trays for this column at 30 (feed entry at stage number 13). In the extractive column we cannot use the short-cut design facility, so we have studied five cases varying the number of ideal trays from 35 to 70 (fewer than 35 do not converge whatever the solvent-to-feed ratio is; more than 70 is not reasonable), selecting in each case the best solvent and feed entry stage. In Table 4 the five cases are specified. With the specifications showed in Table 3 and for the cases specified in Table 4, the columns were then established with the rigorous steady-state column facility of HYSYS. From previous simulation runs, it was found that the solvent-to-feed ratio is

Table 3 Specifications of design variables in the extractive distillation process Streams Feed streams Binary feed

Solvent make-up Extractive column Distillate

Variable

Specification

Temperature (◦ C) Pressure (kPa) Molar flow (kmol/h) Molar composition

100 110 15.91 (1500 kg/h) 48% IBAc

Temperature (◦ C)

25 Pure component

Purity of IBA Recovery of IBA

98.5% 99.75%

Solvent recovery column Distillate Purity of IBAc Bottom Purity of recycled solvent

99.5% 99.5%

Fig. 3. Solvent recovery column case study in the extractive distillation process. Solid line: stage number vs. reflux ratio. Dashed line: reboiler heat duty vs. reflux ratio.

Table 4 Number of ideal trays in the extractive column for each case Case

Ideal trays number Solvent entry stage (optimum) Feed entry stage (optimum)

EC-1

EC-2

EC-3

EC-4

EC-5

35 7 24

40 7 30

50 8 35

60 9 42

70 10 50

R. Mu˜noz et al. / Separation and Purification Technology 50 (2006) 175–183

179

Table 6 European utility prices [23] (Chemical Engineering Plant Cost Index (CEPCI2004) ≡ 444.2 and project life 10 years) Capital cost Utility Low-pressure steam (D /t) Cooling water (D /m3 ) Electricity (D /kWh) Solvent (BUP) (D /kg)

Price 17 0.04 0.041 1.40

In this optimization Cf was assumed to be 10% of FCI, and ir + im was supposed to be 20% of FCI; hence, Eq. (2) could be rewritten as: TAC (D 103 /year) = Cv + 0.30 · FCI

Fig. 4. Total reboiler heat duty vs. solvent-to-feed ratio in the extractive distillation process (case EC-3).

very critical, having a significant effect on the reflux flow rate in both columns, and therefore on the total reboiler heat duty, so the partial optimization was made over this variable. In each case, the solvent-to-feed ratio was then adjusted to minimize the total RHD needed for the process. As an example (EC-3), Fig. 4 shows this effect. This procedure was applied successfully for all five cases. Table 5 shows a summary of the partial optimization based on total reboiler heat duty as a reference variable. 2.3.3.2. Global economic optimization. As it can be seen in Table 5 the minor reboiler heat duty corresponds to EC-5. However, this case requires a big number of stages in the extractive column; so, it would be an expensive column. Therefore, to determine the best of these five alternatives it is necessary to carry out an economic evaluation based on the minimum total annual costs (TAC), using the following objective function [21]: TAC (D 103 /year) = Cv + Cf + (ir + im ) · FCI

(3)

Fixed capital investment (FCI) was estimated using the costs estimation program CAPCOST of Turton et al. [22] with an updated Chemical Engineering Plant Cost Index [23]. A 10-year project life is selected (ir = 10%). Variable costs were obtained applying the unitary costs from Table 6 [24]. Fig. 5 shows the calculated TAC data for the five cases at different solvent-to-feed ratios (Table 4). As it can be observed, there is a minimum for case EC-3 (solvent-to-feed ratio = 1.33). Table 7 lists details of extractive and solvent recovery columns needed to meet the design objectives together with the total annual cost for this optimum. In Appendix A are listed the estimated capital investment of each individual unit. 2.4. Pressure-swing distillation It is well-known that, in some cases, changing the system pressure can affect the vapour–liquid equilibrium (VLE) of a mixture. This effect can be exploited to separate a binary mixture containing a minimum boiling azeotrope [2], provided that

(2)

where Cv is the process variable cost, mostly annual utility consumption (steam, cooling water and electrical power) and BUP make-up cost; Cf the annual fixed costs, i.e. maintenance and wages; FCI the fixed capital investment; ir the fixed capital recovery rate applied to FCI (depreciation rate); im is the minimum acceptable rate of return on FCI. Table 5 Optimum solvent-to-feed ratio in the extractive column for each case Case

Solvent-to-feed ratio

Total reboiler heat duty (RHD) (MJ/h)

EC-1 EC-2 EC-3 EC-4 EC-5

3.00 2.00 1.33 1.00 0.90

5646 4562 3350 3094 2823

Fig. 5. Total annual cost (TAC) vs. solvent-to-feed ratio in the extractive distillation process.

180

R. Mu˜noz et al. / Separation and Purification Technology 50 (2006) 175–183

Table 7 Global economic optimum for extractive distillation process (EC-3) Design parameters

Extractive column (EC)

Solvent recovery column (SRC)

Number of stages

50

30

Feed (top down stage number)

8 (solvent) 35 (binary feed)

13

Assumed tray efficiency (%) Number of trays Reflux ratio Reflux rate (kg/h) RHD (MJ/h)

70 72 3.91 2453.0 2102

70 43 3.67 3221.5 1248

Total annual costs (TAC) ≡ D 1.33 million per year

this mixture significantly changes composition over a moderate pressure range. 2.4.1. Operating pressures selection To investigate the pressure sensitivity of IBA + IBAc azeotropic mixture, we begin with a simulation of the VLE by DISTILL at different pressures, using UNIQUAC thermodynamic model with the parameters listed in Table 1. In Fig. 6, the IBAc mole fraction and temperature of the azeotrope are plotted as a function of pressure. We can observe a significant pressure influence on the azeotropic composition. The operating pressure should be chosen within such a range that water can be used as coolant for the overhead condenser and steam can be used as a heating medium for the reboiler. According to that, the low pressure column (LPC) will work at 20 kPa and the high pressure column (HPC) at atmospheric pressure (101.3 kPa). 2.4.2. Sequencing of the pressure-swing distillation process For a binary mixture presenting a pressure-sensitive minimum boiling point azeotrope, the separation sequence is formed by two columns operating at different pressures [2].

Fig. 6. IBAc mole fraction and temperature of the azeotrope as a function of pressure. Solid line: azeotrope composition (IBAc) vs. pressure. Dashed line: azeotrope temperature vs. pressure.

Table 8 Specifications of the design variables in the pressure-swing distillation process Variable

Specification (◦ C)

Binary feed

Temperature Molar flow (kmol/h) Molar composition

100 15.91 (1500 kg/h) 48% IBAc

High pressure column (HPC)

Pressure (kPa) Purity of IBAc (bottom)

101.3 99.5%

Low pressure column (LPC)

Pressure (kPa) Purity of IBA (bottom)

20 98.5%

Fig. 7 shows the pressure-swing sequence for the separation of IBA from IBAc. The feed enters the high pressure distillation column at 101.3 kPa and the distillate of this column has a composition that approaches the high pressure azeotrope. This distillate is the feed stream to the low pressure column at 20 kPa, and distillate has a composition that approaches the low pressure azeotrope. This distillate has a composition that is similar to the feed composition and it is recycled to mix with the feed to HPC. High purity IBAc (99.5 mole%) is produced as a bottom stream from the HPC and IBA (98.5 mole%) is produced as a bottom stream from the LPC that will be recycled to the estherification reactor. In order that the comparison of this process with the ED may be effective, we have considered the same degrees of purity for both processes. 2.4.3. Optimization We begin the procedure with a local optimization, specifying some design variables and using the total reboiler heat duty as an objective function, following a procedure similar to that used in ED. The design variables selected in this study are the flow rate, composition and temperature of binary feed, the operating column pressures and the purities of the bottom streams in both columns. Table 8 shows the specification chosen for all the previously mentioned variables. On the other hand, the optimization variables are the number of trays in each column, the recycle flow rate and the HPC distillate composition. Like in extractive distillation, the rigorous steady-state column facility of HYSYS requires to set the number of trays in both columns. Therefore, with the short-cut column design facility we did a preliminary optimization (case study) of the number of trays and feed position for both columns. From the results of these case studies we selected 30 ideal trays (feed entry at stage 8) for the HPC and 16 ideal trays (feed entry at stage 4) for the LPC. With the design variable specifications showed in Table 8 and once the number of trays was fixed, the system then converged successfully in the rigorous facility of HYSYS. To carry out the optimization, we studied six cases varying the HPC distillate composition from 0.675 to 0.800, selecting in each case the best recycle flow rate that minimizes the total reboiler heat duty needed for the PSD process. The variation range of this optimization variable was established to produce reasonable results. This variable presents a maximum interval of variation which

R. Mu˜noz et al. / Separation and Purification Technology 50 (2006) 175–183

181

Fig. 7. HYSYS process flow diagram (PFD) of pressure-swing distillation.

corresponds to the values of the azeotropic compositions at both pressures. But when the value approaches either of both compositions, the design is critical, or the system does not converge or it requires very high reflux ratios in one of the two columns. In any case, the design becomes not feasible technically or economically. Table 9 shows the minimum reboiler heat duty for each specified case. As it can be seen in Fig. 8 and Table 9 the global optimum for the PSD process corresponds to PSD-4. This optimum is based on the minimum reboiler heat duty but it coincides with the global economic optimum, since, having previously fixed the number of stages in each column, the operation costs (mainly the total reboiler heat duty) dominate the total costs. In order to compare this alternative with the other alternative of separation (ED), which was previously studied, it will be necessary to evaluate economically the PSD optimum, using an objective function similar to the one considered previously (Eqs. (2) and (3)), with the utility costs shown in Table 6. Details of

the columns for PSD process together with the total annual costs for this case are listed in Table 10. In Appendix B are listed the estimated capital investment of each individual unit. As can be seen in Tables 7 and 10, the annual costs estimated for the PSD process are quite lower than the ones corresponding to the ED process (D 0.99 million against D 1.33 million) both obtained using the same criteria and evaluation procedures. This result may seem surprising, since usually the ED is a separation process more attractive than PSD process, provided that an

Table 9 Optimum values of the optimization variables in the pressure-swing distillation (PSD) process Case

HPCa distillate composition (IBA)

Recycle flow rate (kg/h)

Total RHDb (MJ/h)

PSD-1 PSD-2 PSD-3 PSD-4 PSD-5 PSD-6

0.675 0.700 0.725 0.750 0.775 0.800

3300 2400 1900 1500 1100 850

5629 5079 4729 4594 4791 5871

a b

High pressure column. Reboiler heat duty.

Fig. 8. Total reboiler heat duty vs. recycle flow rate in the pressure-swing distillation process: () PSD-1; () PSD-2; (䊉) PSD-3; () PSD-4; () PSD-5; () PSD-6. Solid line: minimum envelope.

182

R. Mu˜noz et al. / Separation and Purification Technology 50 (2006) 175–183

Table 10 Global economic optimum for PSD process (PSD-4) Design parameters

High pressure column (HPC)

Low pressure column (LPC)

Number of stages Feed (top down stage number) Assumed tray efficiency (%) Number of trays Packed height (m) Reflux ratio Reflux rate (kg/h) RHD (MJ/h)

30 8 70 43 – 1.86 3952.5 3119

16 4 70 – 9.29 1.33 1995.0 1475

Total annual costs (TAC) ≡ D 0.99 million per year

appropriate solvent is used in the extractive column and that it becomes easy to recover. We will analyze the structure of the total annual costs more carefully. In Table 11 we can see more precisely the details of the annual costs for each separation alternative. The first thing that calls our attention in Table 11 is the big difference of the capital investment value between both alternatives, because of the big size of the columns in the ED process (see Table 7) needed to obtain the desired purity of the products and recover the solvent efficiently. This big difference of FCI leads to an important difference of the costs associated with the capital investment (mainly fixed capital recovery and minimum acceptable rate of return). In the end, these costs determine the difference between both alternatives. By lowering these costs the difference between them would be less although the PSD process would always be favourable. On the other hand, the cost of the steam for the reboilers is, by far, the most important of the utility costs, which fully justifies the decisions previous to carrying out the local optimization (in those cases in which there is no FCI variation) only based on these costs. As it can be seen, the cost of steam is clearly higher in the PSD process than in the ED process, because the reflux rates are higher in the first case (Tables 7 and 10). All these results have been obtained for a practical case based on a plant for the treatment of 12,000 Tm/year of an IBA + IBAc mixture. If we consider larger plants, the difference between both alternatives becomes less and less, precisely because the steam costs grow almost proportionally to the flow rate, while

Table 11 Summary of economic results Process

Pressure-swing distillation

(D 103 )

3460.83

2197.50

(D 103 /year)a

Cost proportional to FCI Steam (D 103 /year) Cooling water (D 103 /year) Electric power (D 103 /year) Solvent make-up (D 103 /year)

1038.25 219.08 16.85 9.70 46.07

659.25 300.38 23.62 7.56 –

Total annual costs (D 103 /year)

1329.95

990.81

a

3. Conclusions The simulation of processes with a commercial software program (HYSYS) used appropriately is a very powerful tool to analyze the separation alternatives. When applied to the study of the separation of IBA + IBAc azeotropic mixture, using UNIQUAC model with the values of the binary parameters obtained by us, it has produced satisfactory results. The simulation and economic evaluation of the two separation alternatives that we have considered allow us to conclude that for a 12,000 Tm/year plant, the PSD is more attractive than the ED using BUP as an entrainer. Probably for larger plants (100,000 Tm/year), the ED option will be more attractive, but by a narrow margin in any case, since with this entrainer the distillation columns necessary to carry out the ED and the further recovery of solvent are very large and therefore the necessary investment is much higher than in the PSD process. Acknowledgement The authors acknowledge the financial support from the Ministerio de Ciencia y Tecnolog´ıa of Spain, through Project No. CTQ2004-04477/PPQ. Appendix A Estimated capital investment of each individual unit in the extractive distillation (ED) process D 103 Extractive column (EC) Tower + trays Reboiler Condenser Reflux pump Reflux vessel Total

Extractive distillation

Fixed capital investment

the cots associated with the investment grow much more slowly, and therefore the total costs of both alternatives are much closer. We have to keep in mind, however, that when we increase the size of the plant significantly the costs of the make-up solvent grow in the same proportion, if we maintain the specification of purity and recovery.

Cost proportional to FCI = Cf + (ir + im ) · FCI.

Solvent recovery column (SRC) Tower + trays Reboiler Condenser Reflux pump Reflux vessel Total Cooler Recycled pump Fixed capital investment for ED process

1995.44 108.71 31.95 42.21 37.68 2215.99 1016.65 92.98 25.14 39.83 43.17 1217.77 6.60 20.47 3640.83

R. Mu˜noz et al. / Separation and Purification Technology 50 (2006) 175–183

Appendix B Estimated capital investment of each individual unit in the pressure-swing distillation (SPD) process D 103 High pressure column (HPC) Tower + trays Reboiler Condenser Reflux pump Reflux vessel Total Low pressure column (LPC) Tower + packing Reboiler Condenser Reflux pump Reflux vessel Vacuum system Total Recycled pump Fixed capital investment for PSD process

1146.52 105.61 47.92 45.05 54.39 1399.49 475.46 61.37 58.72 30.19 38.47 115.08 779.29 18.72 2197.50

References [1] P.C. Wankat, Equilibrium Staged Separations, Prentice Hall, Englewood Cliffs, NJ, 1984. [2] J.D. Seader, E.J. Henley, Separation Process Principles, Wiley, 1998.

183

[3] M.F. Doherty, M.F. Malone, Conceptual Design of Distillation Systems, McGraw-Hill, New York, 2001. [4] M.F. Doherty, G. Buzad, Chem. Eng. Res. Des. 70 (1992) 448–458. [5] R.W. Maier, J.F. Brennecke, M.A. Stadtherr, Comput. Chem. Eng. 24 (2000) 1851–1858. [6] J.W. Drew, Chem. Eng. Prog. 71 (2) (1975) 91–99. [7] D.R. Garg, J.P. Ausikaitis, Chem. Eng. Prog. 79 (4) (1983) 60–65. [8] H.L. Fleming, Chem. Eng. Prog. 88 (7) (1992) 46–52. [9] W.F. Furter, Can. J. Chem. Eng. 55 (1977) 229–239. [10] T.C. Frank, Chem. Eng. Prog. (April) (1997) 52–63. [11] I. Rodriguez-Donis, V. Gerbaud, X. Joulia, Ing. Eng. Chem. Res. 40 (2001) 2729–2741. [12] L. Berg, A. Yeh, AIChE J. 30 (1984) 871–874. [13] E.G. Scheibel, Chem. Eng. Prog. 44 (1948) 927–936. [14] J.B. Monton, R. Munoz, M.C. Burguet, J. de la Torre, Fluid Phase Equilib. 227 (2005) 19–25. [15] R. Munoz, J.B. Monton, M.C. Burguet, J. de la Torre, Fluid Phase Equilib. 232 (2005) 62–69. [16] R. Munoz, J.B. Monton, M.C. Burguet, J. de la Torre, Fluid Phase Equilib. 235 (2005) 64–71. [17] R. Munoz, J.B. Monton, M.C. Burguet, J. de la Torre, Fluid Phase Equilib. 238 (2005) 65–71. [18] L.T. Biegler, I.E. Grossmann, A.W. Westerberg, Systematic Methods of Chemical Process Design, Prentice Hall, New Jersey, 1997. [19] I. Sucksmith, Chem. Eng. 89 (13) (1982) 91–95. [20] C.J. King, Separation Processes, second ed., McGraw Hill, 1980. [21] P. Langston, N. Hilal, S. Shingfield, S. Webb, Chem. Eng. Proc. 44 (2005) 345–351. [22] R. Turton, R.C. Bailie, W.B. Whiting, J.A. Shaeiwitz, Analysis, Synthesis and Design of Chemical Processes, Prentice Hall, New Jersey, 1998. [23] Plant Cost Index, Chem. Eng. (May) (2005) 78. [24] A. Szanyi, P. Mizsey, Z. Fonyo, Ing. Eng. Chem. Res. 43 (2004) 8269–8274.