For direct flow sterilizing-grade filtration, a linear scale-up between the performance of discs and pleated filter cartridges has traditionally been assumed. Linear scale-up assumes that the filtration performances, defined here as filter flux and capacity, scale linearly with the membrane area and remains independent of the selected device formats. However, experimental results show that the later assumption does not hold in all cases. In this article, we investigated the effect of solution properties and operating parameters on scale-up with both fouling and non-fouling feeds. For non-fouling solutions, such as buffers, the flux ratio, defined as a, between pleated filter cartridges and disc filters range from 0.5 to 0.85. For complex fouling feeds, such as protein or cell culture media solutions, the ratio of initial flux between pleated filter cartridges and discs was the same as the flux ratio, a. For fouling solutions, the ratio of filtration capacity between pleated cartridges and discs, referred to as capacity ratio, (3, was variable. We found that (3 was sensitive to the particle size distribution of the challenge solution and the mode of filtration operation (constant pressure or constant flux), whereas it was less sensitive to the magnitude of the operating pressure or flux and concentration of the fouling species. For most conditions tested, (3 among pleated cartridges and discs was within ±20% variation of unity. At the end, we present a modified standard model that accounts for both variations in flux ratio, a, as well as capacity ratios, (3, for estimating the requirement for membrane area at manufacturing scale with proteinacious fouling and non-proteinon-fouling feeds. The data show that for cases where filtration is capacity controlled, flux ratios between the pleated filter and disc are not critical. For such cases, the use of a high-area laid-over pleated cartridge construction allows for reducing the number of 10 inch pleated filter cartridges required to process the batch volume.
展开▼
