Influence of feed concentration on the accuracy of permeate flux decline prediction in ultrafiltration

A theoretical model was used for permeate flux decline prediction. Ultrafiltration experiments were performed varying TMP and feed concentration. The feed solution consisted of an aqueous solution of polyethylene glycol with a molecular weight of 35,000 Da. Monotubular Al2O3–TiO2 ceramic membranes with a molecular weight cut off (MWCO) of 5 kDa (Tami Industries, France) were used in the experiments. Experimental data and model predictions were compared. The influence of feed concentration in the accuracy of permeate flux decline prediction was studied. The model predicted very accurately the reduction in the influence of transmembrane pressure (TMP) on permeate flux as the feed concentration increased. The best predictions were obtained for the highest feed concentration tested. For this feed concentration the influence of TMP on experimental and predicted permeate flux was very low.