(384e) Comparative Study of An Extractive Distillation System for the Production of Fuel Grade Ethanol Optimized From the Equilibrium Model and Its Rate-Based Representation

This research presents a comparison between two representations of the same extractive distillation system, one based on the equilibrium model and the other based on the rate of mass and heat transfer in the separation stages. The system is design to separate ethanol ? water mixtures using ethylene glylcol as entrainer, achieving purity higher than 99.5% molar that comply with the legal parameters established in Colombia for its use as gasoline oxygenate [1].

The system is made up by an extractive distillation column and a recovery column, each one with total condenser and reboiler. The extractive distillation column is fed with an ethanol ? water mixture with composition close to the azeotrope and a high purity stream of ethylene-glycol; this column produces as distillate ethanol with purity higher than 99.5% molar and as bottom product a mixture composed mainly by water and ethylene-glycol. The recovery column is fed with the bottoms from the extractive distillation column; water with a small content of ethanol is obtained as distillate and high purity ethylene-glycol as bottom product.

The operating conditions of the distillation columns are established through an optimization process based on the equilibrium model, which maximizes the profitability of the fuel grade ethanol production. The thermodynamic equilibrium is calculated with the use of the Non-Random Two Liquids (NRTL) model for the liquid phase and supposes ideal gas behavior for the vapor phase. The validity of the thermodynamic representation used in the simulations is tested by a thermodynamic consistency test [2].

The equilibrium model of the distillation columns is developed from the system of non-linear equations known as MESH equations, which allows representing the separation stages rigorously through the mass balances, energy balances, mole fraction summations and phase equilibria [3]. Furthermore, Murphree tray efficiencies [4] are ussed to allow a realistic representation of the process.

The strategy followed to find the optimal operating conditions includes the specification of the degrees of freedom of the system and the application of an Interior Point [5] deterministic algorithm that obtains the values of the optimization variables: solvent rate to the extraction column, reflux ratios, reboiler heat duties and solvent recirculating percentege.

The configuration and the operating conditions found for the extractive distillation system through the equilibrium model are reproduced using the rate-based model. This model is based on the researches made by Krishnamurthy & Taylor [6; 7; 8] in which the mass and energy flows at the interphase are calculated from an integral non-equilibrium model.

The mass transfer coefficients for multicomponents are established from the matrix of inverted binary mass transfer coefficients [6]; the binary coefficients are obtained for the liquid and vapour phases according to the correlations recommended by AIChE for sieve trays [6]. In absence of methods to determine the heat transfer coefficients, they are calculated from the Chilton ? Coulburn analogy [6].

The results obtained from the rate-based model are analyzed in order to prove the consistency of the thermodynamic state of the phases coming out of the separation stages. It is found that as a consequence of the structure of the model, the compositions of the phases leaving the separation stages present a minimum deviation that place them above the bubble point temperature for the liquid phase and below the dew point for the vapour phase; this is an evidence of the disagreement between the conditions of the streams and their thermodynamic state. This disagreement is analyzed and discuss considering the information previously reported by other researchers.

Comparing the results obtained by both models, differences are found concerning flow of products, composition profiles and temperature profiles. As an attempt to establish a comparison framework obtaining the greatest resemblance between the two models, Murphree tray efficiencies used for the equilibrium model are varied.

This process is accomplished through an optimization of the equilibrium model, in which the objective function minimizes the quadratic error between the results of both models and the optimization variables are the Murphree tray efficiencies of the equilibrium model. This procedure finds the conditions of the equilibrium model that better fit the processes of mass and heat transfer described by the rate-based model. As a result, some of the Murphree tray efficiency values found are outside the usual ranges due to the complex interactions involved in multicomponent mass transfer [6].

The state of the art in extractive distillation does not include researches in which optimized systems have been represented by the equilibrium and the rate-based models in order to analyze the divergences in their results. Furthermore, this research exposes the implications of each model in the design of extractive distillation systems.

Acknowledgements

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

References

[1] 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.

[2] Liebermann, E. & Vojtech, F. (1972). Thermodynamic consistency test methods. Industrial & Engineering Chemistry Fundamentals, 11, 280.

[3] Russell, R.A. (1983). A simple and reliable method solves single tower and crude-distillation-column problems. Chemical Engineering, 90, 53.

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

[5] Biegler, L. & Grossmann, I. (2004). Retrospective on optimization. Computers & Chemical Engineering, 28, 1169.

[6] Krishnamurthy, R. & Taylor, R. (1985). A non equilibrium stage model of multicomponent separation processes. Part I: model description and method of solution. AIChE Journal, 31, 449.

[7] Krishnamurthy, R. & Taylor, R. (1985). A non equilibrium stage model of multicomponent separation processes. Part II: comparison with experiment. AIChE Journal, 31, 456.

[8] Krishnamurthy, R. & Taylor, R. (1985). A non equilibrium stage model of multicomponent separation processes. Part II: the influence of unequal component ? efficiencies in process design problems. AIChE Journal, 31, 1973.