Effect of fluctuating organic loading (or F/M) on membrane fouling

   Fluctuation of organic loading often compromises dissolved oxygen (DO) concentration especially at peak loading rates. Insufficient air blowing capacity can be a cause of low DO, but oxygen transfer efficiency (OTE) itself can decrease when organic loading rate increases as discussed here. Even if air blowing capacity is sufficient, air flow adjustment only occurs after detecting low DO and hence the lagging air flow adjustment can cause temporarily low DO condition. As a consequence of the fluctuating DO with a varying organic loading rate in the field, it is hard to tell which is more significant factor in membrane fouling out of fluctuating organic loading rates and low DO. 

   A lab scale experiments were performed simulating the diurnal organic loading pattern commonly observed in municipal MBR (Zhang, 2010). As shown in Fig. 1, feed TOC was raised up 100% twice a day in treatment reactor (“feed variable”) while it was maintained nearly constant in control reactor (“feed constant”). By reducing feed TOC later in a day, the average organic loading for the two reactors were maintained same. DO was maintained above 2 ppm regardless of the organic loading in both reactors.   

   It was observed that soluble TOC and polysaccharides of supernatant, which typically indicate membrane fouling tendency with some exceptions, were slightly lower with constant organic loading in the first 80 days (phase I). However, the trends were reversed after that (phase II). Simultaneously, initially smaller particles in the treatment reactor in phase I became larger in phase II as shown in Fig. 2. In the meantime, membrane fouling rate measured by TMP increasing rate (dTMP/dt) was greater with a variable organic loading rate in phase I, but it became smaller in phase II as summarized in Fig. 3.

   It must be noted that this experiment was performed under sufficient DO conditions regardless of the organic loading rate. If the oxygenation capacities were not sufficient during the peak loading just like in full scale MBR, the result might have turned out differently.

FM_flu6Fig. 1. The hourly variations in TOC of feed and DO in aeration tank in one day (Zhang, 2010).

FM_flu8Fig. 2. The median particle diameter as a volume % at constant and variable organic loading rates (Zhang, 2010)

FM_flu9Fig. 3. TMP profile over time at 20 LMH for constant loading and variable loading (Zhang, 2010).


© Seong Hoon Yoon