As expected intuitively, membrane fouling rate can be mitigated by increasing scouring air flow rate in immersed membrane process. Air bubbles not only contact directly with membranes, but also increases bubble induced water flow on membrane surface. In addition, in the case of hollow fiber membrane, rising bubbles also increase random fiber movement that causes acceleration and deceleration of fibers in liquid, which greatly increases the anti-fouling effect.
Fig. 1 shows membrane fouling rates as a function of flux at various specific air flow rates under a typical MBR condition. As can be seen, membrane fouling rate decreases as specific air flow rate increases in general. It is noticeable that high specific air flow rate has more significant impact on membrane fouling at high fluxes.
Fig. 1. Effect of scouring air flow rate on membrane fouling rate. Vertically mounted hollow fiber membranes from GE were used to filter MBR mixed liquor. (Ben Aim, 2004)
Higher air flow rate generally means higher upflow velocity, but it is well known that there is a maximum plateau. When horizontally mounted hollow fiber membranes made of polyethylene (Mitsubishi Rayon, Japan) were used in MBR, upflow velocity quickly reached its maximum at 40 cm/s at an air flow rate of 0.4 m3/min as shown in Fig. 2a. There is no mentioning about diffuser pore size in the reference, but it is likely 6 mm, which was a standard for the membrane system at the time of the experiment. No further increase of average upflow velocity was observed with higher air flow rates than 0.4 m3/min, but, interestingly, the standard deviation of upflow velocity continued to increase as air flow rate increases. However, both upflow velocity and its standard deviation decreased when air flow rate was raised from 0.75 m3/min to 0.95 m3/min (Ueda, 1997).
The existence of the maximum upflow velocity has been confirmed by many other studies performed later (Liu, 2000; Cui, 2003; Bérubé, 2006b; Yeo, 2007), although the maximum upflow velocities reported vary widely between 0.4 and 0.8 m/s depending on literature.
As air flow rate increases, the TMP required to obtain a constant flux decreases as a result of the better membrane scouring by faster upflow (Fig. 2b). It is noteworthy that TMP of 15 LMH case was significantly lower at an air flow rate at 0.5 m3/min than at 0.35 m3/min although there was little difference in the upflow velocities. Perhaps the higher standard deviation of upflow velocity at 0.5 m3/min (Fig. 2a) played a role, which indicates the extent of the random flow acceleration and declaration that do not necessarily increase average upflow velocity.
Fig. 2. Effect of air flow rate on upflow velocity and TMP in MBR. Horizontally mounted hollow fiber membranes (Mitsubishi Rayon Co., Japan) was used. TMP was measured 8 minutes after the start of suction in 280st day (Ueda, 1997)
When 0.65 g/L bentonite solution was filtered using vertically mounted hollow fiber membranes with 1 mm inner diameter, membrane fouling rate measured by TMP increasing rate decreased rapidly as superficial air velocity increased as shown in Fig. 3. However, above a certain superficial velocity (0.8 mm/s in this case), only a marginal decline of membrane fouling rate was observed (Xia et al., 2013). This observation is in line with the observation on the existence of maximum upflow velocity regardless of the air flow rates as discussed in above paragraph.
Fig. 3. Fouling rate of vertically mounted PVDF hollow fiber membranes (Memstar) in 0.65 g/L bentonite solution, where superficial air velocity is measured by dividing air flow rate (mm3/s) by cross-sectional area of tank (mm2) (Xia et al., 2013)
As superficial velocity on membrane surface increases, total filtration resistance calculated by the equation here tends to decreases, but it is only to a certain extent. According to Chang (2011), bubbling can reduce mainly reversible resistances caused by loosely bound cake layer that can be removed physically (Fig. 4). The irreversible resistance caused by the adsorption of small molecules such as soluble microbial products (SMP) or extracellular polymeric substances (EPS) are not significantly affected by superficial velocity. It is perhaps because small molecules have very low back transport velocity and/or they strongly interact with membrane surface. At very high superficial velocity, irreversible resistance can even increase as shown in Fig. 3. Exact reason is not known, but it is possible that the lack of reversible resistance caused by larger particles provides an easy access for SMP and EPS to the membrane surface. In conclusion, membrane performance can be improved only to a certain extent by increasing superficial velocity (or bubbling). Excessive superficial velocity can diminish membrane performance by increasing irreversible resistance.
Fig. 4. Effect of superficial velocity on reversible and irreversible resistance of hollow fiber membranes installed in a plate and frame module with liquid and bubble circulation (Chang, 2011)
© Seong Hoon Yoon