Vol. 55 No. 5 2022 201 Copyright © 2022 The Society of Chemical Engineers, Japan Journal of Chemical Engineering of Japan, Vol. 55, No. 5, pp. 201–216, 2022 Novel Hybrid Methods for Identifying the Main Transition Velocities in Various Bubble Columns † Stoyan Nedel tchev, Jakub Katerla and Ewelina Basiak Institute of Chemical Engineering, Polish Academy of Sciences, Baltycka Street No. 5, 41-100 Gliwice, Poland Keywords: Flow Regime Identification, Various Bubble Columns, Signal Reconstruction, New Hybrid Index, Informa- tion Index New, reliable, and innovative methods (the new hybrid index (NHI) and the information index (II)) for flow regime (FR) identification in bubble columns (BCs) operated with air–deionized water or nitrogen–tap water systems (at ambient conditions), as well as nitrogen–ethanol systems (at pressures of 0.1 and 0.3 MPa) systems have been successfully devel- oped. New parameters extract useful hidden information from time series measurements by division of various signals (gas holdup, differential pressure (DP) and gauge pressure fluctuations) into different equal parts. Salient advantage of the new parameters is that their definitions are not based on any assumption. Based on the local minima in the NHI profile extracted from gas holdup time series, two transition gas velocities U trans at 0.032 and 0.046 m/s have been identified in an air–deionized water BC (0.1 m in ID). The NHI profile based on DP fluctuations in a nitrogen–tap water BC (0.102 m in ID) distinguishes similar U trans values at 0.029 m/s and 0.047 m/s. In a nitrogen–ethanol system at ambient pressure, the NHI identifies only the first U trans value at U g =0.028 m/s. At elevated pressure (0.3 MPa), it has been found that this first U trans value shifts to a higher U g value (0.045 m/s). In the an- nularly aerated BC (operated with an air-deionized water system), the main transition velocities have been identified at U g =0.026 m/s and 0.073–0.079 m/s (depending on the clear liquid height), respectively, based on the newly defined II profile. In summary, this work presents novel and reliable methods for FR identification in several BCs and reports new, use- ful experimental results about the FR boundaries based on them. Introduction Bubble columns (BCs) are the most important and widely used gas–liquid reactors in the chemical, biochemical, pe- troleum, mining (flotation), and metallurgical industries (Leonard et al., 2015). Various useful processes such as oxidations, chlorinations, oxychlorinations, carboxylations, carbonylations, sulphonations, dehydrosulphonations, hy- drogenations, alkylations, polymerisations, esterifications, and hydrogenations can be performed in BCs (usually op- erating at high pressure and high temperature). Absorp- tion processes accompanied by slow chemical reactions are also frequently performed in BCs (Schumpe et al., 1979). Fischer–Tropsch, methanol, and dimethyl ether syntheses, as well as aerobic fermentation and biological wastewater treatment, could also be implemented in BCs. BCs are simple but very effective gas–liquid contactors. eir main characteristics include their uncomplicated con- struction (built without moving or rotating parts), ease of maintenance, and excellent mass transfer and heat transfer properties (Shiea et al., 2013; Rzehak et al., 2017). In ad- dition, BCs exhibit large contact areas between liquid and gas phases and good mixing within the liquid phase. ese reactors further provide uniform temperature distributions, which brings them very close to isothermal operation. is important advantage of BCs leads to improved selectivity (Shetty et al., 1992). Due to their very good heat transfer performance, BCs are preferred for exothermic reactions. Coherent structures (e.g., large-scale liquid circulation and small-scale bubble wake) (Mudde et al., 1997; Joshi et al., 2002) in the heterogeneous FR (churn-turbulent flow) en- hance the heat transfer rate and lead to the uniform tem- perature distribution inside BCs (Shu et al., 2019). Such co- herent structures are related to the turbulence in the bubble bed, where turbulence is generally defined as the fluctua- tions around the mean flow. ese fluctuations are the result of both passage of the deterministic organized flow struc- tures and random disorganized irrotational motions, which together constitute the turbulent flows (Joshi et al., 2009). e organized deterministic patterns are known as eddies or turbulent flow structures, the latter of which are oſten hid- den among the incoherent turbulent motions. Back-mixing of both gas and liquid phases and scale-up issues are among the main limitations of BCs (Youssef et al., 2014). e structure of the multi-phase dispersion flows in BCs is complex, and difficult to predict (especially at high gas flow rates); most of the hydrodynamic parameters are sys- tem-dependent (Leonard et al., 2015). BC performance can also change significantly as a result of FR transition. Further, Received on August 16, 2021; accepted on March 11, 2022 DOI: 10.1252/jcej.21we082 Correspondence concerning this article should be addressed to S. Nedeltchev (E-mail address: sned@iich.gliwice.pl). † Presented at the 13th European Congress of Chemical Engineering, 20–23 Sept. 2021, Virtual Event Research Paper