Abstract #656
Section: Dairy Foods
Session: Dairy Foods: Processing & chemistry
Format: Oral
Day/Time: Wednesday 11:00 AM–11:15 AM
Location: Wekiwa 7/8
Session: Dairy Foods: Processing & chemistry
Format: Oral
Day/Time: Wednesday 11:00 AM–11:15 AM
Location: Wekiwa 7/8
# 656
Effect of membrane channel geometry on limiting flux and serum protein removal during skim milk microfiltration.
Michael C. Adams1, Emily E. Hurt1, David M. Barbano*1, 1Cornell University, Ithaca, NY.
Key Words: microfiltration, channel geometry, limiting flux
Effect of membrane channel geometry on limiting flux and serum protein removal during skim milk microfiltration.
Michael C. Adams1, Emily E. Hurt1, David M. Barbano*1, 1Cornell University, Ithaca, NY.
Our objectives were to determine the limiting fluxes (LF) and serum protein (SP) removal factors (SPR) of 2 100-nm ceramic microfiltration (MF) membranes: one with 4-mm round retentate flow channels (RFC) and one with 4-mm equivalent-diameter diamond-shaped retentate flow channels (DFC). Retentate and permeate were continuously recycled to the feed tank and a uniform transmembrane pressure (TMP) (TMP at the inlet minus TMP at the outlet) was maintained at 25 ± 3 kPa using a permeate recirculation pump. The LF for each membrane was determined by increasing flux once per h from 45 kg·m−2·h−1 until flux did not increase with increasing TMP. Temperature, average cross-flow velocity, and protein concentration in the retentate recirculation loop were maintained at 50°C, 7 m·s−1, and 8.5 ± 0.03%, respectively. Experiments were replicated 3 times and the Proc GLM procedure of SAS was used for statistical analysis. The LF of the DFC membrane (71 kg·m−2·h−1) was lower (P < 0.05) than the LF of RFC membrane (88 kg·m−2·h−1). Reynolds numbers based on the hydrodynamic diameter of the retentate flow channels were calculated for each membrane. Differences in Reynolds numbers between the membranes were proportional to the differences in LF. Permeate produced using the DFC membrane contained more (P < 0.05) protein than the permeate produced using the RFC membrane due to additional casein passage through the DFC membrane. SPR was calculated by dividing true protein in the permeate by SP in the permeate portion of the feed to describe the ease of SP passage through each membrane. Higher SPR indicate higher rates of SP passage. After accounting for casein contamination in each of the permeates with SDS-PAGE, the DFC membrane SPR remained higher (P < 0.05) than the RFC membrane SPR. Though DFC membrane LF was lower, DFC permeate removal on a modular basis was higher due to the increased module membrane surface area of the DFC membrane relative to the RFC membrane (2.07 m2 vs. 1.41 m2). Depending on the size of the system, using DFC membranes could reduce the capital cost of a MF system due to a reduction in the number of stainless steel modules.
Key Words: microfiltration, channel geometry, limiting flux