Identification of Best Model for Equilibrium Data of Ethanol-Water Mixture

Abstract: Four empirical models are tested for fitting the T-y-x equilibrium data of ethanol-water mixture by minimizing the Root Mean Square (RMS) between equilibrium data and theoretical points. The total pressure of the correspondent data is 101.3 kPa. Allmodels parameters are also identified. The study suggests that NRTL model fits the equilibrium data best with RMS = 0.4 %. Key words: Equilibrium datas, models, ethanol-water mixture. 1. Introduction  The bioethanol is considered as a very important renewable energy in this decade. It could be produced from natural sources (wheat, sugar cane, sugar beet, corn, etc.) by fermentation and purified by distillation. The calculation of distillation column (continuous and batch processes) is based on determination of its equilibrium-stage. This step could be obtained by graphic methods (McCabe and Thiele or Ponchon- Savarait) basing on equilibrium curve and operating lines. Usually, the equilibrium curve is presented in literature as equilibrium point. Its fitting by mathematical models is a very important step for programming the graphic methods. The equilibrium liquid-vapour of ethanol-water mixture presents some particularity. In fact, it’s considered as a non-ideal mixture and it presents an azeotrope in mole fraction of 89% [1]. In this case, the identification of the activity coefficients of ethanol and water is a very important step in fitting equilibrium data. Wilson, NRTL, Van Laar and Margules are among the most empiric models