AIChE Journal, Vol.51, No.1, 162-177, 2005
Performance efficiency of tubular inorganic membrane modules for pervaporation
Because the separation performance improves the rate of transport in inorganic high flux pervaporation membranes, resistances in the boundary layer become the performance bottleneck in the dehydration of solvents. By operating the membranes at high temperature levels, permeate water fluxes of more than 10 kg m(-2) h(-1) can be obtained. Therefore, the design of efficient industrial-scale modules becomes significant. In this study, a set of generic rules for the construction of technical and economically feasible modules is derived and different design versions are compared, including tubular, hollow-fiber, spiral-wound, and multichannel monolith types. In particular, the performance efficiency and applicational limitations of the isothermal module concept Pervap(R) SMS (Silica Membrane System)from Sulzer Chemtech are evaluated. Steam condensation or a thermal liquid ensures the constant operation temperature for any permeate loading. For assessing the module efficiency the impact of concentration and temperature polarization is modeled with empirical correlations for the mass and heat transfer on the feed side. For these high-flux membranes the overall flux reduction by polarization effects in the technical module is around 30 to 50%. At lower permeate fluxes the losses in membrane performance are more the result of mass transfer limitations, whereas at higher permeation rates heat transfer becomes more significant. The pressure drop in the porous support was calculated by Knudsen diffusion, whereas the pressure loss in the permeate channel was determined with laminar flow according to Hagen-Poiseuille. At a permeate flux of 10 kg/m(2)h the friction loss in the support can reach up to 100 mbar, whereas the pressure drop in the permeate tube is negligible. (C) 2004 American Institute of Chemical Engineers.
Keywords:inorganic membranes;vapor permeation;pervaporation;module design;performance efficiency;optimization