Packed Tower Modeling with Least Squares, Genetic Algorithm and Design of Experiments Methods

Document Type : Research paper

Authors

Department of Chemical Engineering, Faculty of Engineering, Arak University, , Arak, Iran

Abstract

Objective: Design of Experiments is a useful tool for reducing the number of experiments and, consequently, lowering costs; moreover, it provides a model that is valid within the experimental range. Additionally, modeling is used as a method for predicting system behavior and saving costs. The aim of this research is the modeling of experiments conducted in a packed distillation column and the prediction of the Height Equivalent to a Theoretical Plate (HETP).
Materials and Methods: This research is based on available experimental data for a packed distillation column. The test materials were cyclohexane and n-heptane. The column was packed with one-inch Pall ring packing, and the experiments were conducted at two pressures of 0.35 bar and 1.65 bar. Among various methods, three widely-used approaches Genetic Algorithm, Least Squares, and Design of Experiments were selected for modeling.
Results: The models obtained by the Genetic Algorithm and Least Squares methods, although having different coefficients, yielded similar results. Regarding the Design of Experiments method, analysis of variance showed that the data were not suitable for analysis; therefore, favorable results were not obtained, because the experiments performed had not been designed on this basis. Modeling with the three methods was expected to produce similar results. The reason for the difference in the Design of Experiments model was the unsuitability of the experimental data.
Conclusion: With using the models provided by the Least Squares and Genetic Algorithm methods, the performance of the packed column can be predicted with acceptable accuracy within the experimental range, and the HETP can be calculated without the need for costly and time-consuming experiments. Furthermore, this study showed that the Design of Experiments method yields reliable results only when the data are collected from the outset according to its principles, such as randomization, appropriate dispersion, reproducibility, etc.

Keywords

Main Subjects

References

[1] Wang, X., U. Kruger, and B.J.C.E.P. Lennox, Recursive partial least squares algorithms for monitoring complex industrial processes. 2003. 11(6): p. 613-632. DOI: https://doi.org/10.1016/S0967-0661(02)00096-5   
 
[2] Leboreiro, J., J.J.C. Acevedo, and C. Engineering, Processes synthesis and design of distillation sequences using modular simulators: a genetic algorithm framework. 2004. 28(8): p. 1223-1236. DOI:10.1016/j.compchemeng.2003.06.003
 
 [3] Motlaghi, S., F. Jalali, and M.N.J.E.s.w.a. Ahmadabadi, An expert system design for a crude oil distillation column with the neural networks model and the process optimization using genetic algorithm framework. 2008. 35(4): p. 1540-1545. DOI:10.1016/j.eswa.2007.08.105
 
[4] Ghoreishi S.A., et al., Application of genetic algorithm for solving multi-objective optimization problems in robust control of distillation column. 2011. 3(1). DOI:10.4156/ijact.vol3.issue1.4
 
[5] Tavakolmoghadam, M. and M.J.P.E. Safavi, An optimized neural network model of desalination by vacuum membrane distillation using genetic algorithm. 2012. 42: p. 106-112. DOI: 10.1016/j.proeng.2012.07.400
 
[6] Müller, A.L.H., et al., Total sulfur determination in residues of crude oil distillation using FT-IR/ATR and variable selection methods. 2012. 89: p. 82-87. DOI:10.1016/j.saa.2011.12.001
 
[7] Shahandeh H.,. Ivakpour J, and Kasiri N., Internal and external HIDiCs (heat-integrated distillation columns) optimization by genetic algorithm. 2014. 64: p. 875-886. DOI :http://dx.doi.org/10.1016/j.energy.2013.10.042
 
[8] Fu, G., Sanchez Y., and Mahalec V., Hybrid model for optimization of crude oil distillation units. 2016. 62(4): p. 1065-1078. DOI: 10.1002/aic.15086
 
[9] Wang, S., et al., Optimization Investigation on Operating Parameters of a Scrubbing Tower Using a Genetic Algorithm. 2020. 43(10): p. 2109-2117. DOI: 10.1002/ceat.201900524
 
[10] Shariat, A., Kunesh,J.G. (1995). "Packing efficiency testing on a commercial scale with good (and not so good) reflux distribution."  34(4): 1273-1279. DOI: https://doi.org/10.1021/ie00043a031
 
[11] Hanley, B. and Chen C.C., New mass‐transfer correlations for packed towers. 2012. 58(1): p. 132-152. DOI:  10.1002/aic.12574
 
[12] Churchill SW Critique of the classical algebraic analogies between heat, mass and momentum transfer, 1997; 36:p.3866-3878. DOI: https://doi.org/10.1021/ie960750a
 
[13] Potnis SV, Lenz TG. Dimensionless mass-transfer correlations for packed-bed liquid-desiccant contactors. 1996. 35: p.4185-4193. DOI:  https://doi.org/10.1021/ie960212y
 
[14] Shetty S, Cerro RL. Fundamental liquid flow corrilations for the computation of design parameters for ordered packings. 1997. 36: p.771-783. DOI: https://doi.org/10.1021/ie960627j
 
[15] Fuller, E.N., et al., New method for prediction of binary gas-phase diffusion coefficients. 1966. 58(5): p. 18-27. DOI: https://doi.org/10.1021/ie50677a007
 
 
Journal of Separation Science and Engineering
Volume 18, Issue 1
June 2026
Pages 31-43

Files

Share

How to cite

Statistics