Heat and Mass Transfer in Tubular Inorganic Membranes

In this work heat and mass transfer in tubular asymmetric ceramic membranes suitable for applications in membrane reactors have been investigated. An experimental matrix has been employed to quantify the heat and mass transfer in the membrane, which leads to the identification and validation of respective parameters in a comprehensive and consistent manner. Two types of membranes have been investigated for the characterisation of their transport parameters. However, the main part of experiments have been carried out with the larger membranes (inner diameter of 21 mm), which were closer to realistic dimensions for application on industrial scale. Thermal conductivity of the membrane has been identified and validated by the steady state and dynamic heat transfer experiments. The structural parameters of the composite membrane (mass transfer parameters) are identified by single gas permeation experiments and validated by isobaric steady state and transient diffusion experiments. The mentioned single gas permeation experiments have been conducted for every composite, precursor and intermediate, starting from the support. The application of dusty gas model enables to understand and predict the influences of temperature, pressure and molar mass of the gas. It has been further shown that the characterisation of every single layer of the composite membrane is important. A simulation analysis has been carried out to see the influence of flow direction and top layer on the mass transfer through the membrane, which reveals that the choice of flow direction may be significant, especially when employing the membrane for the selective dosing of educts in a catalytic reactor. Also the choice of the material of permselective layer is substantial in terms of fluxes and pressure drop in every individual layer. A non-dimensional analysis of isobaric diffusion, based on simulations, shows the influences of axial dispersion, volumetric flow rate and temperature on the isobaric diffusion process in terms of mole fraction and gas flow rates. The consideration of axial dispersion may be substantial for reactions, where controlled dosing of educts is desired. While identification and validation of membrane transport parameters are one important aspect, the work also shows that membrane reactor configurations can be reliably modelled in the limiting case without chemical reaction. Even in this case, thermal effects and the interrelation between heat and mass transfer should be accounted for.