As discussed here, strong form critical flux does not exist in MBR, where various particles and macromolecules are filtered under poorly defined hydrodynamic conditions. Membrane undergoes initial fouling by macromolecules immediately after contacting with mixed liquor before the membrane permeability stabilizes (Le-Clech, 2006), which might be in line with the weak form critical flux concept (Field, 1995). In practical MBR conditions, however, particles and macromolecule deposition occurs continuously even after the permeability largely stabilizes. Therefore, true critical flux does not exist in MBR whether it is a strong form or a weak form.
Though membranes are fouled continuously at a low rate in any flux, the imposed flux can sustain relatively long-term before TMP reaches an unacceptable level (Zhang, 2006). Therefore, the highest flux that can sustain extended period of time experimentally is also called “sustainable flux” instead of critical flux. Since there is no absolute consensus on how long the sustainable flux should sustain under which TMP, measured critical (or sustainable) flux is somewhat subjective.
The figure below shows a critical flux test result in a full-scale municipal MBR plant with an average daily flow (ADF) of 2,300 m3/day. TMP was monitored at a flux for 12 minutes before pause the filtration for 3 minutes. Flux was increased stepwise until TMP increased distinctively faster than in the previous cycle. As can be seen in the figure, TMP increases slowly at a rate of 2.0 kPa/hr at 25 LMH, but the TMP increasing rate does not change much in the next high flux, i.e. 34 LMH. This trend continues until flux becomes 51 LMH, where TMP increasing rate is 4.2 kPa/hr. But, when flux became 60 LMH, TMP increasing rate suddenly jumped to 8.7 kPa/hr. In this case, critical flux can be determined between 51 LMH and 60 LMH although it is somewhat subjective.
Since the extent of TMP increase in each flux level can vary depending on filtration duration, relaxation time, air flow rate, mixed liquor condition, the initial membrane etc., the measured critical fluxes are somewhat arbitrary. In addition, there is no standardized protocol in judging the threshold TMP rise that divides sub- and over-critical fluxes. As a consequence, comparing two critical fluxes from two different data sources is not straightforward.
Fig. 1. A critical flux test with flat sheet membranes (Kubota Co., Japan), influent BOD = 200 ppm, design flow rate = 2300 m3/day, MLSS = 12,000 mg/L, water temperature = 13oC, operational mode = 12 minutes suction and 3 minutes relaxation (reproduced from Yoon, 2006).
© Seong Hoon Yoon