Changing product specification in extractive distillation process using intelligent control system
- PDF / 2,105,243 Bytes
- 12 Pages / 595.276 x 790.866 pts Page_size
- 27 Downloads / 194 Views
(0123456789().,-volV)(0123456789(). ,- volV)
ORIGINAL ARTICLE
Changing product specification in extractive distillation process using intelligent control system A. P. Arau´jo Neto1 • G. W. Farias Neto1 • T. G. Neves1 • W. B. Ramos1 • K. D. Brito1 • R. P. Brito1 Received: 11 September 2019 / Accepted: 3 December 2019 Springer-Verlag London Ltd., part of Springer Nature 2019
Abstract Obtaining anhydrous ethanol by extractive distillation has already been the object of several studies in the control literature. However, despite the presence of varying degrees of purity of anhydrous ethanol owing to its applications in industrial and commercial sectors, little attention has been given to dynamic and control for changing the operating conditions to provide anhydrous ethanol with different specifications. Using a soft sensor based on artificial neural network, this work aimed to develop an intelligent control system to contemplate the changes in the specification of anhydrous ethanol, considering the whole process (extractive and recovery columns), and keeping the process operating at an optimal point. Using the developed intelligent control system, the only necessary modification is the new specification and all new set-points values for controllers (temperature and solvent to azeotropic feed ratio) are updated automatically, without human interference, while with a conventional control system, all the new set-points values must be modified manually. The results showed that for the studied anhydrous ethanol specification range (99.1–99.9% mole), the new optimum operating conditions (new steady-state) were reached in a short time (between 1 and 2 h), with no evidence of overflow or emptying of sumps and reflux vessels of the columns. In addition to the easy implementation of the intelligent control system, the existing control structure remains unchanged, not requiring the investment for new instrumentation. Keywords Extractive distillation process Intelligent control system Soft sensor Artificial neural networks
1 Introduction and problem definition Unconventional distillation techniques, such as extractive distillation, azeotropic distillation, pressure swing distillation, pervaporation, and other hybrid method, are used for mixtures that have azeotrope or are formed by species with near boiling point to perform separation. However, extractive distillation is still the most attractive method owing to its applicability in industrial scale and in terms of energy consumption [5, 13]. This separation method is based on the addition of a solvent that is responsible for modifying the relative volatility of the original components. The implementation of a suitable control structure is as important as process optimization, directly impacting the & R. P. Brito [email protected] 1
Department of Chemical Engineering, Federal University of Campina Grande, Campina Grande, PB 58109-970, Brazil
performance of the extractive distillation process and, consequently, the associated economic costs [12]. However, t
Data Loading...
