(327c) Optimization of Ethanol Dehydration By PSA With 3A Zeolite

Increasing demand for renewable fuels such as biodiesel and bioethanol had promoted research in separation processes. High water content in alcohol used for blending with gasoline could cause problems in engines. Anhydrous ethanol used as fuel must have purity 99.5%. Due to azeotrope formation, traditional separation process cannot be used. Adsorption processes has been widely used for industrial production of anhydrous ethanol.

Extractive distillation, heteroazeotropic distillation and pervaporation are alternative technology for ethanol-water separation. Comparison between alternative technologies and adsorption must be made with economical analysis.

In the present study, dehydration of ethanol by PSA using zeolite 3A was optimized with a sequential simplex algorithm. Adsorption processes are attractive alternative for its low energy consumption but it has high capital cost by equipment acquisition. Several studies had optimized adsorption process with recovery and product purity as objective function; however profits and cost are real comparative variables.

Objective function considers product sale, raw material cost, utilities and capital cost. Adsorption/desorption pressure, purge/feed ratio, bed dimensions and cycle time were chosen as independent variables for optimization. Auxiliary equipment such as compressor, condenser, vacuum system and utilities consumptions were estimated using algebraic equations but adsorption was modeled with a partial differential equation set. In all cases information for optimization was in the cyclic steady state (CSS). Start-up generated hot spot that could cause damage in adsorbent particles.

Differential equations were solved with method of lines, with upwind discretization in axial dimension and integrator ODE 15S of Matlab® for stiff problem. Model considers component balance for water, overall mass balance, “one-phase” energy balance, kinetic rate and equilibrium relation. Linear driving force model was used for kinetics and Langmuir isotherm for equilibrium. Ethanol was considered as an inert.

Numerical solution considers the following limits for optimization variable: Adsorption pressure; 1-4 bar, Desortion pressure; 0.1-1bar, purge/feed ratio; 0.05-0.2, diameter: 1-3m, length; 3-8m and time adsorption/desorption: 150-500s. Some fixed conditions were 393K for inlet temperature, 0.1 molar fraction of water in feed and constant feed of 200.000 l/day.

Increasing energy for compression or vacuum system allow to separate ethanol-water mixtures with a small bed on the other hand a largest column can achieve the separation with less energy for compression. Optimization process analyzed the effects and chose a better condition.

Separation with molecular sieves needs a previous process to concentrate ethanol near to azeotropic point.  Azeotropic ethanol price was considered as a constant fraction of anhydrous ethanol price. Fraction of cost for raw material depends of several variables: sugar source cost, utilities in fermentative and concentration process and capital cost for reactor and distillation column. Optimization process considered several fractions of cost.