Experimental validation of the process simulation a typical industrial bioethanol unit was conducted, comparing the obtained results with the information collected in an industrial plant. A standard solution containing water, ethanol and 17 congeners was chosen to represent the fermented must, whose composition was selected according to analyses of samples of industrial wines. A careful study of the vapour-liquid equilibrium of the wine components was performed. An attempt to optimise the industrial plant was conducted considering two optimising approaches: the central composite design (CCD) and the Sequential Quadratic Programming (SQP). The process was investigated in terms of bioethanol alcoholic graduation, ethanol recovery, energy consumption and ethanol loss. The results showed that the simulation approach was capable of correctly reproducing a real plant of bioethanol distillation and that the optimal conditions are slightly different from those used at the industrial plant investigated. Substantial fluctuations in wine composition were easily controlled for the two loop controls preventing an off-specification product. The optimised conditions indicate a distillation process able to produce bioethanol according to the legislation requirements and with appropriate steam consumption and loss of ethanol. However, for the production of alcohols with superior qualities, substantial changes in the production system may be required.
Supervision of an industrial plant subject to a maximal duration constraint
