(133a) Optimal Design of Extractive Distillation Systems Using a Two-Level Approach

A design methodology is developed for extractive distillation systems. These systems are being used to produce bioethanol which is considered as a replacement of motor fuel. Designing highly efficient distillation processes is decisive in the competitiveness of bioehtanol against fossil fuels. This competitiveness should not depend on international price of oil and government subventions.

Extractive distillation systems are generally made up by two distillation columns, one in charge of the extraction process and the other of the solvent recovery. A solvent and a mixture to be separated are fed into the first column to obtain a highly concentrated distillate. The second column recovers the solvent that is reused in the extractive distillation column [1].

The distillation process is represented by the equilibrium model that uses MESH equations [2] to describe the separation stages. The thermodynamic equilibrium considers the non-ideality of the liquid phase through the NRTL model [3]. In addition, it uses the Murphree Tray Efficiency [4] concept to reproduce the separation level achieved in each stage.

The rigorous design of these systems implies establishing: diameter of the columns, exchange areas of condensers and reboilers, height of the columns according to the number of stages, and feed stage locations. Both columns are divided into sections separated by the feed streams. The number of stages in each section establishes the feed streams locations and assures that the distinctive configuration of the extractive distillation systems is maintained.

The development of the optimal design is formulated as a Mixed Integer Non Linear Programming (MINLP) problem. The discrete variables determine the number of stages of the columns sections. The continuous variables include the variables of the equilibrium model and other variables such as feed rate of solvent, feed temperature of solvent, and recovered solvent recirculation ratio. Other design aspects such as dimensions of condensers, dimensions of reboilers, and column diameters are calculated from the variables of the equilibrium model through empirical relations.

The solution of the optimization problem is achieved through a two-level strategy. The discrete variables are considered on the outer level that proposes different configurations whose continuous variables are optimally solved in the inner level. The inner level is a Non-Linear Programming (NLP) problem that is handled through a deterministic algorithm [5] in a simultaneous strategy for both columns that compose the system. The solution to the outer problem is reached through the use of a Simulated Annealing algorithm [6, 7]; it increases the chances of finding a global optimum. The interaction between both levels allows defining the optimal design of the equipment and their operation conditions under any objective function as quantitative criterion.

This methodology is applied to the design of the extractive distillation of the ethanol - water azeotropic mixture using glycerol as solvent. Constrains on the purity of ethanol [8] and degradation temperature for glycerol [9] are considered on the optimization process. The possible designs are evaluated through an economic objective function that considers cost of building the columns, operation costs, raw material cost, and value of products. These elements are organized in a function that calculates the annual profit of the process, and whose objective is to be maximized. The solution of the MINLP problem finds the design of the extractive distillation process and the operation conditions that offer the greater profitability in the production of bioethanol.

A thorough bibliographic review did not find any research that proposes a methodology for the design of extractive distillation systems considering discrete and continuous variables. However, the results of other researches concerning the separation of the ethanol ? water mixture by distillation [10; 11; 12] are considered for comparison in order to confirm the advantages of the design attained.

Acknowledgements: This work was supported financially by research grants from Colciencias, by financial support of research project code: 1101-452-21113.

References

[1] Zhigang Lei; Chengyue Li & Biaohua Chen (2003). Extractive Distillation: A Review. Separation & Purification Reviews, 32 pp. 121 ? 213.

[2] Seader, J. D., & Henley, E. J. (1998). Separation process principles. New York: Wiley.

[3] Renon, H. and Prausnitz, J.M. (1968). Local Compositions in Thermodynamic Excess Functions for Liquid Mixtures. AIChE Journal, 14, pp. 135-144.

[4] Murphree, E. V. (1925). Rectifying Column Calculations ? With Particular Reference to N Component Mixtures. Industrial and Engineering Chemistry, 17, 747.

[5] Nocedal, J. & Wright, S.J. (2006). Numerical Optimization. Springer series in operations research and financial engineering.

[6] Floquet, P.; Pibouleau, L. & Domenech, S (1994). Sequence Synthesis: How to use Simulated Annealing Procedure. Computers and Chemical Engineering, pp. 785?795.

[7] Cardoso, M.F.; Salcedo, R.L.; de Azevedo, S.F. & Barbosa, D. (1997). A simulated annealing approach to the solution of MINLP problems. Computers & Chemical Engineering, 21, 1349.

[8] Ministerio de Ambiente y Desarrollo Territorial y Ministerio de Minas y Energía. (2003) Resolución No. 0447 de abril 14 de 2003. Ministerio de Ambiente y Desarrollo Territorial y Ministerio de Minas y Energía. República de Colombia.

[9] Dias, M.; Junqueira, T.L.; Maciel R.; Maciel, M. & Vaz, C.E. (2009). Anhydrous Bioethanol Production using Bioglycerol ? Simulation of Extractive Distillation Processes. 19th European Symposium on Comuter Aidded Process Engineering ? ESCAPE19.

[10] Knapp, J.P. & Doherty M.F. (1990). Thermal integration of homogeneous azeotropic distillation sequences. AIChE Journal, 37, 969.

[11] Meirelles, A.; Weiss, S. & Herfurth, H. (1992). Ethanol dehydration by extractive distillation. Journal of Chemical Technology and Biotechnology, 53, 181.

[12] Gil, I.D.; Uyazán, A.M.; Aguilar, J.L.; Rodríguez, G. & Caicedo, L.A. (2008). Simulation of ethanol extractive distillation with a glycols mixture as entrainer. 2nd Mercosur Congress on Chemical Engineering & 4th Mercosur Congress on Process Systems Engineering.