Hydrodynamic modeling of bubble column reactors with vertical heat exchanging internals

Bubble column reactors are frequently used apparatures regarding multiphase flows and chemical reactions such as the Fischer-Tropsch synthesis and the Methanol synthesis etc. Since most of these reactions are of exothermic nature, the emerging reaction heat has to be suficiently removed from the reactor. Therefore, longitudinal flow heat exchanger tube bundles are immersed into the column, which have the advantage of immediate heat removal at the source and they offer a large amount of specific surface area for good heat removal properties. On the downside, those internals cover a large portion of the reactor’s cross-sectional area, which leads to a strong influence on the hydrodynamics in bubble columns.
The main aim of this thesis was to develop and validate a hydrodynamic model for bubble column reactors with vertical heat exchanging internals. To fulfill this task, an existing phenomenological cell model by Schilling (2014) has been improved and further developed. The model combines several modeling ideas, such as the vertical compartment approach by Shimizu et al. (2000), the horizontal compartment model to capture up- and downflow regions by Gupta et al. (2001) and a two-bubble class model to account for polydispersity (large and small gas bubbles). Furthermore, sub-models for the prediction of flow Patterns (Vitankar and Joshi, 2002) as well as for breakup and coalescence (Liao, 2013) have been implemented. To incorporate the influence of heat exchanger internals, the developed model is based on the idea of multiple up- and downflow regions, as it has been proven by local radial holdup profiles for bubble columns with internals. These holdup profiles show strong fluctuations and follow a polynomial behavior. Therefore, the reactor is compartmentalized even further in horizontal direction according to the number of gas hold-up peaks detected from previous experimentally obtained results by ultrafast X-ray computer tomography.
The model was evaluated for superficial gas velocities ranging from 2 cm s-1 to 12 cm s-1 to cover homogeneous as well as heterogeneous flow regimes. Five internal configurations, namely, two different tube bundle patterns (triangular and square pitch) and two different tube sizes (8 and 13mm) as well as the empty bubble column reactor as counterpart were examined. Furthermore, the influence of a bended bottom structure (cf. u-tube heat exchanger) has been investigated with the developed model. The obtained parameters have been compared to experimental data provided by Seiler (2016) and empirical correlations from the literature (Shah et al. (1982), Akita and Yoshida (1973)). The results for the overall gas hold-up and the bubble size distribution are generally in good agreement with the experimental data and represent a significant improvement to the original phenomelogical cell model. On the other hand, the model predicts an increasing Sauter mean diameter, which is contrary to experimental findings for empty bubble column reactors. Those results also influence the specific gas-liquid interfacial area and the volumetric mass transfer coefficient significantly. Furthermore, the model is not able to represent the characteristics of the different internal configurations.
Finally, a sensitivity analysis after Plackett and Burman (1946) was carried out to determine the model’s stability and robustness against parameter changes.