Identification of transport processes in column experiments using a frequency domain approach

Area of Science:

• Environmental Science
• Chemical Engineering
• Physical Chemistry

Background:

• Solute transport processes can be analyzed using system identification techniques with input-output data.
• Traditional rectangular pulse input signals in column experiments are suboptimal for system identification due to limited frequency content.

Purpose of the Study:

• To propose and evaluate stepped sine signals as an optimal input for solute transport system identification.
• To compare the performance of stepped sine signals against rectangular pulses in identifying parameters for the convection-dispersion equation (CDE) and mobile-immobile model (MIM).

Main Methods:

• Generated stepped sine concentration signals using High Performance Liquid Chromatography (HPLC).
• Applied stepped sine and rectangular pulse inputs in column experiments for system identification.
• Conducted numerical experiments to assess the robustness of identification under white noise.

Main Results:

• Stepped sine signals allow independent selection of amplitude and passband, enhancing excitation of transport processes.
• Numerical experiments demonstrated that stepped sine signals yield more robust and accurate identification of the convection-dispersion equation compared to rectangular pulses, even under noisy conditions.

Conclusions:

• Stepped sine signals are a more effective input for system identification of solute transport processes than rectangular pulses.
• The proposed method improves the accuracy and reliability of parameter estimation for transport models like CDE and MIM.