Energy Fuels 2010, 24, 6099–6102 Published on Web 10/15/2010
: DOI:10.1021/ef100671w
New Design Methodology Based on Self-Heat Recuperation for Production by Azeotropic Distillation Yasuki Kansha, Naoki Tsuru, Chihiro Fushimi, and Atsushi Tsutsumi* Collaborative Research Center for Energy Engineering, Institute of Industrial Science, The University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8505, Japan Received May 31, 2010. Revised Manuscript Received September 18, 2010
In this paper, an innovative design methodology is proposed for production by azeotropic distillation using self-heat recuperation technology to reduce energy consumption. Based on this design methodology, the heat of the distillate and condenser in each distillation column is recovered by compressors and exchanged with the heat of the corresponding feed and reboiler. Hence, a larger amount of heat, which consists of the sensible heat and latent heat of the process streams, is circulated within the process than that in a conventional azeotropic distillation process by the heat exchanger, leading to a significant reduction in the process energy required. Process simulation results for bioethanol production show that the azeotropic distillation process resulting from the proposed design methodology achieves a large reduction in energy compared to a conventional azeotropic distillation process.
integration processes based on conventional heat recovery technologies.7-9 As a result, the minimum energy requirement of the overall process has not been reduced because changes to the condition of the process stream are constrained in conventional heat recovery technologies. Moreover, most cost minimization analyses for bioethanol plants have been conducted based on these conventional processes and technologies.10-12 On the other hand, many novel heat-integrated distillation columns have been developed using pressure difference to reduce energy consumption, such as the vapor recompression distillation column (VRC),13-15 and the heat-integrated distillation column (HIDiC).16-21 However, the design and operation of such distillation columns are complex, resulting in increased capital costs, as compared to a conventional distillation column. Additionally, although the reboiling heat in the distillation column is recognized as a target for using recovered heat, preheating of the feed stream to the distillation
Introduction Recently, bioethanol has attracted increased interest in many countries as a substitute for petroleum to suppress global warming.1 To produce bioethanol, ethanol must be separated from the ethanol-water mixture after fermentation. In practice, distillation is widely used for the separation of this mixture. However, conventional distillation is energyconsuming because ethanol and water form an azeotropic mixture, and it is believed that about half of the heat value of bioethanol is required to distill the ethanol from the mixture. To separate pure ethanol from ethanol-water mixtures by distillation, it is necessary to use an entrainer (azeotroping agent) because the azeotropic mixture is one that vaporizes without any change in composition in azeotropic points.2,3 Benzene, cyclohexane, or isopropyl alcohol can be used as entrainers for the ethanol-water mixture. Therefore, at least two separation units are required to produce pure ethanol, leading to increased energy consumption. To overcome this problem, many researchers have proposed membrane separations4,5 or pressure swing adsorption (PSA)6 as alternatives to azeotropic distillation, often successfully developing appropriate membranes or sorbents to achieve an efficient separation. However, in many cases, they have paid little attention to the overall process scheme or have developed heat
(9) Ebrahim, M.; Kawari, A. Appl. Energy 2000, 65, 45–49. (10) Fernando, S.; Adhikari, S.; Chandrapal, C.; Murali, N. Energy Fuels 2006, 20, 1727–1737. (11) Zamboni, A.; Shah, N.; Bezzo, F. Energy Fuels 2009, 23, 5121– 5133. (12) Zamboni, A.; Bezzo, F.; Shah, N. Energy Fuels 2009, 23, 5134– 5143. (13) Brousse, B.; Claudel, B.; Jallut, C. Chem. Eng. Sci. 1985, 40, 2073–2078. (14) Annakou, O.; Mizsey, P. Heat Recovery Syst. CHP 1995, 15, 241–247. (15) Gros, H. P.; Diaz, S.; Brgnole, E. A. J. Supercrit. Fluids 1998, 12, 69–84. (16) Huang, K.; Nakaiwa, M.; Akiya, T.; Aso, K.; Takamatsu, T. J. Chem. Eng. Jpn. 1996, 29, 344–351. (17) Huang, K. J.; Nakaiwa, M.; Akiya, T.; Owa, M.; Aso, K.; Takamatsu, T. J. Chem. Eng. Jpn. 1996, 29, 656–661. (18) Nakaiwa, M.; Huang, K.; Naito, K.; Endo, A.; Owa, M.; Akiya, T.; Nakane, T.; Takamatsu, T. Comput. Chem. Eng. 2000, 24, 239–245. (19) Olujic, Z.; Fakhri, F.; de Rijke, A.; de Graauw, J.; Jansens, P. J. J. Chem. Technol. Biotechnol. 2003, 78, 241–248. (20) Huang, K.; Matsuda, K.; Takamatsu, T.; Nakaiwa, M. J. Chem. Eng. Jpn. 2006, 39, 652–660. (21) Huang, K.; Nakaiwa, M.; Wang, S.-J.; Tsutsumi, A. AIChE J. 2006, 52, 2518–2534.
*To whom correspondence should be addressed. Telephone: þ81-35452-6727. Fax: þ81-3-5452-6728. E-mail:
[email protected]. (1) Petrou, E. C.; Pappis, C. P. Energy Fuels 2009, 23, 1055–1066. (2) Mussati, M. C.; Aguirre, P. A.; Espinosa, J.; Iribarren, O. A. AIChE J. 2006, 52, 968–985. (3) Von Halle, E.; Shacter, J. In Distillation, and Distillation, Azeotropic and Extractive, Kirk-Othmer Separation Technology, 2nd ed.; Seidal, A., Ed.; John Wiley and Sons: Hoboken, NJ, 2008; pp 871-984. (4) Gomez, M. T. D. P.; Klein, A.; Reple, J. U.; Wozny, G. Desalination 2008, 224, 28–33. (5) Vane, L. M.; Alvarez, F. R. J. Chem. Technol. Biotechnol. 2008, 83, 1275–1287. (6) Modla, G.; Lang, P. Chem. Eng. Sci. 2008, 63, 2856–2874. (7) Linnhoff, B.; Hindmarsh, E. Chem. Eng. Sci. 1983, 38, 745–763. (8) Linnhoff, B.; Eastwood, A. R. Chem. Eng. Res. Des. 1997, 75, S138–144. r 2010 American Chemical Society
6099
pubs.acs.org/EF
Energy Fuels 2010, 24, 6099–6102
: DOI:10.1021/ef100671w
Kansha et al. Table 1. Feed Conditions composition (mol %) flow rate (kg mol/h) temperature (°C) pressure (kg/cm2)
ethanol, 80; water, 20 10 77 1 (98.07 kPa)
Table 2. Specifications of the Bottom Product Composition DC1 DC2
benzene < 1.0 10-4 mol fraction ethanol < 1.0 10-3 mol fraction
Table 3. Conditions for the Distillation Columns first distillation column (DC1) number of stages pressure (kg/cm2) feed stage of azeotropic mixture feed stage of benzene feed stage of recycled EtOH
Figure 1. Proposed integrated azeotropic distillation process module.
second distillation column (DC2) 20 1 4 1 4
number of stages pressure (kg/cm2) feed stage reflux ratio
10 1 5 2
first distillation column (DC1). The vapor stream from DC1 is compressed adiabatically by a compressor (C1) [4 f 5]. Subsequently, stream 5 is cooled in a heat exchanger (HX1) [5 f 6], and the pressure and temperature of stream 6 are adjusted by a valve (V1) and a cooler (L1) [6 f 7 f 8]. The liquid stream [8] is divided into two streams [9 and 10] in a decanter (D). Stream 9 consists mainly of the entrainer, which is recycled to the feed benzene [3]. The bottom [11] of DC1 is divided into two streams [12 and 14]. Stream 14 becomes a product stream (pure ethanol). Stream 12 is heated in a heat exchanger (HX1) and fed into DC1. In M2, the effluent stream [10] from M1 is heated in a heat exchanger (HX2) and fed to the second distillation column (DC2). At the same time, the recycled stream, which is the distillate stream from DC2, is adiabatically compressed by a compressor (C3) [18 f 27] and cooled by exchanging heat in HX2 [27 f 28]. The pressure and temperature of stream 28 are adjusted by a valve and cooler (V2 and L2) [28 f 29 f 30], and stream 30 is fed into DC1 as the recycled stream. Next, in M3, the feed stream [15] is separated into the distillate [16] and the bottoms [17] by the distillation column (DC2). The vapor distillate [16] is divided into two streams [18 and 19] by a separator. Stream 18 is recycled to M2, while stream 19 is adiabatically compressed [19 f 20] and exchanged with the heat in a heat exchanger (HX3) [20 f 21]. The temperature and pressure of stream 21 are adjusted by a valve (V3) and a cooler (L3) [21 f 22 f 23], and then the effluent stream is fed into DC2. Subsequently, the bottoms [17] from DC2 is divided into two streams [24 and 25]. Stream 25 is the product water. The other stream [24] is vaporized in HX3 and fed into DC2 [26].
column is not or less recognized as a possibility for heat recovery.22 Therefore, the energy balance for heating and cooling duties is not well-considered; it requires additional heat sources, and there is potential to decrease energy requirements. Recently, by incorporating compressors and heat exchangers, the authors have developed an attractive technology for reducing energy consumption and applied this to petroleum refinery processes.23-25 In this approach, each process unit is divided into functions to analyze the process energy balance and heat is recirculated internally in the process. As a result, the energy consumption of a process can be greatly reduced. Using this technology, an innovative design methodology for azeotropic distillation processes is described in this paper. The application of this methodology to the design of a dehydration process in a bioethanol plant is expected to reduce energy consumption. Process Design Methodology and Process Configuration Based on Internal Heat Recovery Process Design Methodology. Conventional azeotropic distillation processes, which have one distillation column to separate ethanol and another to separate water from their mixture, are divided into three modules. The sum of the feed enthalpy is made equal to that of the effluent stream enthalpy in each module to analyze the heating and cooling loads of all process streams. According to this analysis, the recovery streams are selected and the internal heat of the process stream in each module can be recovered and recirculated using a compressor and heat exchanger using self-heat recuperation technology. Process Configuration. Figure 1 shows the structure of the proposed integrated process module, consisting of three modules, two distillation modules, namely, the first distillation module (M1), the second distillation module (M3), and the heat circulation module (M2). In this integrated process module, stream 1 represents a feed stream of the ethanol-water azeotropic mixture and stream 2 represents an entrainer (benzene and cyclohexane) feed stream. These streams are fed into the
Simulation Results and Discussion We calculated the energy consumption for the proposed integrated process module for the azeotropic distillation column and compared it to the energy consumption for a benchmark azeotropic distillation column. The process simulation was conducted using a commercial simulator, PRO/II, version 8.1 (Invensys). Ethanol-Water Mixture. Considering a distillation process that separates ethanol from a mixture of ethanol (80 mol%) and water (20 mol%) at a temperature of 77 °C and pressure of 1 kg/cm2 (98.07 kPa) (Table 1), we assumed that the feed flow rate was 10 kg-mol/h, the mole fraction of benzene (entrainer) was less than 1.0 10-4 in the product ethanol from the first distillation column (DC1), the mole fraction of ethanol was less than 1.0 10-3 in the product water from the second distillation column (DC2) (Table 2), and the pressure and heat losses of the processes were negligible.
(22) Horiuchi, K.; Nakaiwa, M.; Iwakabe, K.; Matsuda, K.; Toda, M. Kagaku Kogaku Ronbunshu 2008, 34, 70–75. (23) Kansha, Y.; Tsuru, N.; Sato, K.; Fushimi, C.; Tsutsumi, A. Ind. Eng. Chem. Res. 2009, 48, 7682–7686. (24) Kansha, Y.; Tsuru, N.; Fushimi, C.; Shimogawara, K.; Tsutsumi, A. Chem. Eng. Sci. 2010, 65, 330–334. (25) Kansha, Y.; Tsuru, N.; Fushimi, C.; Tsutsumi, A. J. Chem. Eng. Jpn. 2010, 43, 502–507.
6100
Energy Fuels 2010, 24, 6099–6102
: DOI:10.1021/ef100671w
Kansha et al.
Figure 3. Simulation results for the conventional azeotropic distillation column.
Figure 2. Simulation results for the proposed integrated azeotropic distillation process module.
The energy consumption of the proposed integrated process module was elucidated by comparing it to that of a conventional azeotropic distillation system. Other conditions of the distillation column are shown in Table 3. In all heat exchange systems, the minimum temperature difference was kept constant at 10 K. The non-random two liquid (NRTL) model was applied to the stream state equation. We assumed that the adiabatic efficiency in the compressor was 100%. The work required for changing pressure (WC) is expressed as follows: WC ¼ Hout - Hin
ð1Þ
where the enthalpy of the stream changes from Hin to Hout in an adiabatic, irreversible process. In the proposed system, the net work (Wnet) is represented by the following equation: Wnet ¼ WC
Figure 4. Temperature-heat diagram for the proposed integrated azeotropic distillation process module.
ð2Þ
was discarded into coolers in each module (28.0, 2.3, and 17.1 kW), because the sum of enthalpy in the feed streams was equal to that of the effluent streams in each module when using internal heat recovery. As this relationship indicates, the compression work was used for inducing heat recovery and circulation in each module and exhausted as low exergy heat. The total energy consumption for the case in which the adiabatic efficiency in the compressor was changed to 80%, as more representative of a real industrial process, was 49.3 kW (=18.0 þ 28.3 þ 3.0). It can be seen that the total energy consumption was almost equal to the case where the adiabatic efficiency was 100%, because part of the input work for the adiabatic pressure change was transformed into heat. Thus, the proposed integrated module dramatically reduces the total energy consumption, indicating that the integrated module is very promising for an azeotropic distillation process. In addition, to follow the proposed design methodology perfectly, three additional compressors are required instead of boilers in the optimal energy point of view. The capital costs would increase compared to the conventional distillation process. However, these compressors are possibly cumulated to a compressor or two compressors using a conventional hot charge method to sacrifice the energy reduction. Thus, there is the trade-off between capital costs and operation costs. As mentioned above, the conventional distillation and dehydration process consumes a large amount of energy
where WC represents the work of the compressor. The distillates from the columns were condensed, and the latent heat exchanged to the bottoms and feed, as shown in Figure 2. The total internal heat exchange duty in HX1, HX2, and HX3 was increased to 453.2 kW, as compared to the conventional azeotropic distillation. It can be seen that the total heating duty was covered by internal heat recovery, resulting in a considerable reduction in energy consumption. The total work of the proposed module with internal heat recovery could thus be reduced [48.7 kW (=28.9 þ 3.3 þ 16.5 kW); Figure 2] compared to the external heat load of the benchmark process [395.0 kW (=282.4 þ 112.6 kW); Figure 3]. At the same time, the cooling duties of L1, L2, and L3 of the proposed module are 28.0, 2.3, and 17.1 kW. Figure 4 shows a temperature-heat diagram of the proposed integrated process module for azeotropic distillation. Note that numbers beside the composite curve mean stream numbers in Figure 1. It can be seen that the latent heats of the effluent streams are exchanged with those of the feed streams, as well as the sensible heats in each module, leading to minimization of the exergy loss in the heat exchangers. Thus, the cooling duties of L1, L2, and L3 are almost equal to the compression duties of C1, C2, and C3, respectively. From this figure, it can be understood that all of the process heat is recirculated without any heat addition and the total heating duty was covered by internal heat recovery. All of the compression work in each module (28.9 kW in M1, 3.3 kW in M2, and 16.5 kW in M3) 6101
Energy Fuels 2010, 24, 6099–6102
: DOI:10.1021/ef100671w
Kansha et al.
for pure bioethanol production, leading to an increase in the bioethanol price. However, when the integrated azeotropic distillation modules were installed based on the proposed design methodology, the energy cost of pure bioethanol production can be dramatically reduced, even though the capital costs to renew the process is required. This design methodology and the proposed integrated module may be helpful in rapidly advancing the availability of bioethanol to society. Conclusion
based on the proposed design methodology requires only the energy necessary to drive heat circulation and results in the reduction of total energy consumption. Simulation studies based on a feed mixture of ethanol (80 mol%) and water (20 mol%) show that the proposed modules can reduce the energy required for azeotropic distillation to 12.3% of that of benchmark distillation systems. As a consequence, this design methodology and the resultant module show promise not only for energy saving but also cost minimization of bioethanol plants.
An innovative design methodology and integrated module based on self-heat recuperation for the distillation of azeotropic mixtures were described in this paper. This module
Acknowledgment. This study was supported by the New Energy and Industrial Technology Development Organization (NEDO), Japan.
6102
