Process efficiency of casein separation from milk using polymeric spiral-wound microfiltration membranes

Microfiltration is largely used to separate casein micelles from milk serum proteins (SP) to produce a casein-enriched retentate for cheese making and a permeate enriched in native SP. Skim milk microfiltration is typically performed with ceramic membranes and little information is available about the efficiency of spiral-wound (SW) membranes. We determined the effect of SW membrane pore size (0.1 and 0.2 µm) on milk protein separation in total recirculation mode with a transmembrane pressure gradient to evaluate the separation efficiency of milk proteins and energy consumption after repeated concentration and diafiltration (DF). Results obtained in total recirculation mode demonstrated that pore size diameter had no effect on the permeate flux, but a drastic loss of casein was observed in permeate for the 0.2-µm SW membrane. Concentration-DF experiments (concentration factor of 3.0× with 2 sequential DF) were performed with the optimal 0.1-µm SW membrane. We compared these results to previous data we generated with the 0.1-µm graded permeability (GP) membrane. Whereas casein rejection was similar for both membranes, SP rejection was higher for the 0.1-µm SW membrane (rejection coefficient of 0.75 to 0.79 for the 0.1-µm SW membrane versus 0.46 to 0.49 for the GP membrane). The 0.1-µm SW membrane consumed less energy (0.015-0.024 kWh/kg of permeate collected) than the GP membrane (0.077-0.143 kWh/kg of permeate collected). A techno-economic evaluation led us to conclude that the 0.1-µm SW membranes may represent a better option to concentrate casein for cheese milk; however, the GP membrane has greater permeability and its longer lifetime (about 10 yr) potentially makes it an interesting option.