ISSN print edition: 0366-6352 ISSN electronic edition: 1336-9075 Registr. No.: MK SR 9/7 Published monthly

Optimization of distillation column in phenol production process for increasing the isopropyl benzene concentration using response surface methodology and radial basis function (RBF) coupled with leave-one-out validation method

Hani Vaziri, Amin Hedayati Moghaddam, and Seyed Amin Mirmohammadi

Department of Chemical Engineering, Central Tehran Branch, Islamic Azad University, Tehran, Iran

E-mail: ami.hedayati_moghaddam@iauctb.ac.ir

Received: 6 January 2020  Accepted: 11 April 2020

Abstract:

In this study, phenol production process was simulated. Further, the performance of distillation column was optimized through maximizing the mole fraction of cumene in upstream flow. Response surface methodology was applied for design of experiment, modelling, and optimizing the cumene mole fraction in upstream flow of separation column. The analysis of variance was performed for finding the important operative parameters as well as their effects. In this experiment, the effects of three parameters on separation performance were investigated, including number of tray (A), column temperature (B), and reflux ratio (C). Further, radial basis function (RBF) was applied to model the separation column. To develop the neural network model, leave-one-out method was used. This robust model was used for optimizing the performance of separation column. The statistical and artificial intelligence system were capable of predicting mole fraction in upstream flow of distillation column in different conditions with R² of 0.99 and 0.93, respectively. According to statistical and RBF models, the optimized values of cumene mole fraction are 0.45 and 0.44, respectively.

Keywords: Design of experiment; Cumene; Radial basis function; Leave-one-out; Phenol production

Full paper is available at www.springerlink.com.

DOI: 10.1007/s11696-020-01162-w

Chemical Papers 74 (10) 3311–3324 (2020)