Electrical impedance tomography (EIT) is increasingly applied as a noninvasive technique for monitoring multiphase reactors. Conventional image reconstruction, however, is ill-posed and depends on regularisation, and it may even be redundant in applications where the aim is to identify the operating state rather than to reconstruct the spatial conductivity field explicitly.
In this work, the authors present a reconstruction-free framework that works in the measurement domain and maps raw complex boundary impedance data directly onto two tasks that matter for the reactor, namely the classification of the gas injection pattern and the regression of superficial gas velocities. Together, the spatial injection distribution and the total volumetric flow constitute the primary process-state information from which gas holdup can subsequently be inferred, which is why they are reported as proxies for local gas holdup monitoring.
The experiments were conducted in an acrylic bubble column of 600 millimetres height and 104 millimetres inner diameter, equipped with a 256-electrode array distributed over eight axial rings and operated at four excitation frequencies between 1 kilohertz and 1 megahertz.
Gas flow rates between 1.0 and 6.5 litres per minute were covered, corresponding to superficial gas velocities of 1.96 to 12.75 millimetres per second. Within this range, near-perfect classification of the gas injection pattern was achieved, with accuracies of 93 to 100 percent for excitation frequencies between 1 and 100 kilohertz using the full 256-electrode configuration. For the quantitative estimation of the superficial gas velocity, increasing the calibration density reduced the mean absolute error from 0.388 to 0.105 litres per minute for localised injection, which corresponds to a reduction from 7.0 to 1.9 percent of the operating range, and from 0.298 to 0.157 litres per minute for distributed injection conditions, corresponding to a reduction from 5.4 to 2.8 percent.
The results demonstrate that direct inference from raw EIT boundary measurements enables accurate, real-time monitoring of bubble-column operation without tomographic reconstruction, and they provide quantitative guidance on the selection of the excitation frequency, the axial placement of the sensing rings and the calibration resolution.
The study was carried out at Hamburg University of Technology within the Collaborative Research Centre CRC 1615 SMART Reactors.
Moritz Hollenberg, Hossein Ostovar, Zahra Sharafian, Tom Liebing, Oliver Korup, Thorsten A. Kern, Raimund Horn (2026). Reconstruction-Free EIT for Injection-Pattern Classification and Superficial Gas Velocity Regression as Proxies for Local Gas Holdup in Bubble Columns. Ind. Eng. Chem. Res.
