Untitled

The aim of this work was to compare the data on flux decline with time obtained with ultrafiltration experiments of polyethylene glycols (PEGs) performed under different experimental conditions (transmembrane pressures and crossflow velocities) and the theoretical predictions for these experimental conditions obtained with the model proposed by Ho and Zydney (2000). This model considers that two main fouling mechanisms, pore blocking and cake formation, can occur when macromolecules are ultrafiltered. In this work, flux decline predictions were achieved without using empirical parameters dependent on operating conditions. The fouling experiments were performed using monotubular ZrO2–TiO2 ceramic membranes with a molecular weight cut off (MWCO) of 15 kDa (Orelis, France). A 5 g/L PEG (35,000 Da molecular weight) aqueous solution was used as feed. The model does not accurately describe the fouling dynamics under every experimental condition tested. However, the best predictions were obtained for high transmembrane pressures (TMPs) and low crossflow velocities when the cake layer is more likely to form.

The model does not accurately describe the fouling dynamics under the experimental conditions tested. However, the lowest discrepancies between experimental results and those predicted by the model were obtained for the experimental conditions that favour the formation of a cake layer: high TMPs and low crossflow velocities. A comparison of the predicted values and the experimental ones shows that predictions are not good enough for any experimental condition tested. The inexistency of the pore blocking phenomenon under certain experimental conditions and the possibility of gel layer formation occurring for very short time scales may be the reason for the discrepancies between predicted and experimental results. In this way, these discrepancies could also be due to model hypothesis that may not be correct for the experimental conditions tested in this paper. For example, some of the membrane pores could be totally blocked. Moreover, the gel layer could form all over the membrane surface and not only over the membrane surface whose pores have been previously blocked.