Development and Use of the Biowanze Distillation Model Using Vali Software

Abstract

The ever-increasing price of raw materials and energy is increasingly felt by industries, which need to keep track of the optimal values of their measured data. Data validation and reconciliation (DVR) usually comes after simulation, as a method for its user to see and locate deviations of measurements from the optimal plant operation, as well as to define their origin. Previous to the core of this work, a chapter on the state-of-the-art has been made. It contains mainly the description of the Biowanze plant, as well as a comparison of similar modeled processes in terms of purity, energy consumption, and yield of Bioethanol. It also contains a comparison of alternative processes, such as replacing the Biowanze dehydration membrane with an azeotropic distillation column, extractive column, or via the use of pressure swing distillation. The objective of this work is to model the distillation section of Biowanze. On the one hand, in terms of simulation to propose possible improvements to the current state of the process. On the other hand, in terms of validation to compare the reconciled modeled data with the measured data over 2 months. Throughout the work carried out until its completion, the main difficulty was to create a simulation model as representative as possible of the Biowanze process, thanks to the use of the DISVAL columns of the Vali software. The second difficulty of this work was to be able to retrieve the measured data from Biowanze, and to associate all these measurements to the correct modeled tags. In the end, numerous plots comparing the reconciled values obtained by the model with the measured values from Biowanze are made. The origin of the discrepancies between the reconciled and measured data is not easy to explain because it depends on almost everything in the process. The main results obtained are that the feed stage of the rectification column is located too high in this column and that the reflux of the rectification column, heating the mash column, could be decreased while keeping the same ethanol production. Both of these lead to too much steam being fed into the process and higher costs. Finally, a chapter on perspectives has been written to mention the next steps for improvement. Taking into account, the efficiency of the modeled columns should decrease the differences between measured and reconciled values. Also, adding in the model the dehydration section, the fermenter, and the Evapo2 section to then make a pinch analysis of the model close to the real process could help to discover opportunities for process improvement.