Effect of MLSS on membrane performance

   In early days of MBR technology, MLSS had been considered a controlling parameter of membrane fouling. But, later studies revealed MLSS itself was only weakly correlated with membrane fouling in its common ranges such as 6-15 g/L as long as membrane scouring was performed reasonably well. In fact, as discussed here, SRT and F/M ratio has replaced MLSS as dominant factors affecting membrane fouling.  

   According to the critical flux concept discussed here, only very small sub-micron sized particles can deposit on membrane surface due to the particle back transport phenomena under crossflow condition. Since sub-micron sized particles or macromolecules take only a very small portion of biosolids, MLSS itself should not be strongly correlated with membrane fouling. 

   Unfortunately there are lots of literature contradicting among others regarding MLSS effect. The possible causes are as follow.

1. SRT effect

   When MLSS effect is explored empirically, all other process parameters must be identical including SRT, pH, temperature, origin of sludge, etc. However, many observations on MLSS effect appear not from rigorously controlled environment under the solid understanding on process dynamics. 

  •    In some studies, membrane fouling rates are measured while biosolids are allowed to concentrate gradually by reducing excess biosolids removal rate from the system. In this condition, SRT is gradually increasing while biosolids never have sufficient time to reach equilibrium condition at each SRT condition. As a result, the measured membrane fouling rates must be affected not only by MLSS, but also by SRT in this case. Under this condition, the likely conclusion is that membrane fouling increases as MLSS increases, but the real cause of the rising membrane fouling is the increasing young microorganisms, which decreases effective SRT.
  •    Conversely, if MLSS is allowed to decrease gradually by increasing biosolids removal, the effective SRT is declining while MLSS is also declining. Under this condition, membrane fouling rates likely increases as MLSS declines. But, the real cause of increasing membrane fouling is the declining SRT.
  •    Any experiment performed without rigorously controlling SRT would conclude anywhere between above two.

   For example, if the experimental results in Yong et al. (2006) is analyzed with respect to MLSS, one may conclude membrane fouling decreases as MLSS increases based on Fig. 1 below. This is not a right interpretation because it neglects the varying SRT effect as MLSS increases. In this case the actual SRT of four different MLSS were 3, 5, 10, 20 days, respectively. (Note: This is only a thought experiment, which is not what the authors intend to do in the original paper)

   On the contrary to the above example, MLSS were raised by not discharging excess sludge before the experiment to obtain the target MLSS in other example (Damayanti, 2011). Once the target MLSS was reached, filtration tests were performed to estimate membrane fouling at the MLSS. Though apparent SRT increases as excess sludge removal decreases, biosolids are not fully adapted with the longer SRT. As a result, membrane fouling rate increases quite dramatically as MLSS increases as shown in Fig. 2.

   In other example, the MBRs based on attached growth and suspended growth were compared, where attached growth reactor had far less MLSS than suspended growth reactor despite the identical volumetric organic loading rates (Lee, 2011). Authors suggested that the lower the MLSS was the higher the membrane fouling rate was, but the greater fouling rates in attached growth MBR could be due to much shorter SRT or higher effective F/M ratio (Lee, 2002).

2. DO effect

   If MLSS is raised to investigate its effect on membrane fouling, oxygen demand increases due to the increased endogenous respiration rate by more dense microorganisms. Meanwhile, oxygen dissolution rate can decrease as a result of lowering oxygen transfer efficiency (OTE) as discussed here. Unless aeration rate is raised, DO tends to decrease, which in turn cause faster membrane fouling as discussed here. The same can occur when F/M ratio is raised, if oxygen dissolution is not sufficient.

Fig. 1. Effect of MLSS on membrane fouling rate (Modified from Yong, 2006).

MLSS.h38Fig. 2. Trend of membrane fouling rate (dTMP/dt) as a function of the flux at various MLSS (Damayanti, 2011).


© Seong Hoon Yoon