Modeling of Carbon Dioxide Absorption Using Hydroxyethyl Piperazine Solution: Application of Response Surface Methodology and Artificial Neural Networks

Bahmanzadegan, Fatemeh; Ghaemi, Ahad

doi:10.22075/jme.2026.37443.2839

Modeling of Carbon Dioxide Absorption Using Hydroxyethyl Piperazine Solution: Application of Response Surface Methodology and Artificial Neural Networks

Articles in Press

Document Type : Research Paper

Authors

Fatemeh Bahmanzadegan ¹

Ahad Ghaemi ²

¹ Iran University of Science and Technology

² Chemical Engineering, Iran University of Science and Technology

10.22075/jme.2026.37443.2839

Abstract

In this study, the modeling of carbon dioxide (CO2) absorption in a wet wall column were conducted using response surface methodology (RSM) and artificial neural networks (ANN). The input variables included the molality of the absorbent, inlet CO2 pressure, solvent loading (%), total system pressure, and gas flow rate, while the output variable was the outlet pressure of the wet wall column. The RSM approach was first applied to evaluate and optimize the system response, followed by prediction using two neural network architectures: Multi-Layer Perceptron (MLP) and Radial Basis Function (RBF) networks. Based on the outlet pressure, the CO2 flux through the column could be calculated. In the MLP model, a mean squared error (MSE) of 3.2244 × 10⁻6 was achieved using two hidden layers with 3 and 7 neurons over 219 epochs. The R² value obtained from the RSM model using a quadratic function was 0.9998. For the neural networks, the R² values were 0.99999 and 0.99997 for the MLP (with Trainbr training function) and RBF models, respectively. The results demonstrate that both ANN and RSM approaches can effectively predict the outlet pressure of the wet wall column, offering reliable tools for modeling reactive CO2 absorption processes.

Keywords

CO₂ Absorption

Wet Wall Column

Artificial Neural Network

Multilayer Perceptron

Radial Basis Functions

Subjects