Effect of shear stress on membrane fouling

1. MBR with immersed membrane (iMBR)

   In the MBR with immersed membrane, vigorous aeration is performed to scour membrane surface. The scouring air not only generates shear stress on membrane surface, but also generates strong turbulence in mixed liquor, which affects microbial floc formation and perhaps the physiology of microorganisms as well. Not many researches have been performed to date on the shear effect on microbiology in long-term operation, but it will be interesting to review the available data in literature. 

   Side-by-side tests were performed to elucidate the effect of high scouring air flow using two MBRs with one Kubota membrane module each (0.48 m2 surface area and 0.2 micron pore size) (Menniti, 2010). Experimental conditions are summarized in Table 1. The aeration intensity of low shear MBR was in the high end of the typical flat sheet membranes. The aeration of high shear MBR was at least three times more vigorous than the typical level. Therefore, it must be noted that the distinctive characteristics of microbial community found from the high shear MBR do not necessarily represent those of the practical MBR.

   The following were found from the experiment, which were in line with the data published by the same group as a separate paper (Menniti, 2009).

  • The chord length measured by FBRM® technique (Mettler Toledo International Inc., USA) was shorter in high shear MBR along the experiment (Fig. 1), which is acceptable intuitively.
  • Higher life forms were observed only in low shear MBR (Fig. 1).
  • Extracellular polymeric substances (EPS) that represent bound biopolymers were higher in low shear MBR in general (Fig. 2).
  • Soluble EPS, which was the COD of filtered mixed liquor by 0.2 micron filter, became higher in low shear MBR after 40 days of operation (Fig. 3).
  • TMP increasing rate was much slower in high shear MBR in spite of the higher soluble EPS (Fig. 4). This suggests that the high level membrane foulant was overcome by the high shear stress generated by high scouring air flow.

                           Table 1. Experimental condition (Menniti, 2010)

ShearS28Fig. 1. Chord length in the high and low shear MBRs and low shear MBR arbitrary worm index for experiment 2.

ShearS24Fig. 2. Floc-associated EPS concentration in high and low shear MBR.

ShearS25Fig. 3. Soluble EPS (or SMP) concentration in high and low shear MBR.

ShearS29Fig. 4. TMP with time in a low and high shear MBR (Menniti, 2010).

2. MBR with side-stream membrane (sMBR)

   In sMBR, membrane units are placed outside the aeration tank and mixed liquor is circulated through the membrane units. The very high shear stress acting on the mixed liquor in the circulation pump can cause dramatic changes in biology depending on how often mixed liquor is exposed to the high shear. In this application, only a very small fraction of the mixed liquor circulated through the membrane unit is recovered as permeate, e.g. <5%, as discussed here. The recovery becomes even lower as experimental scale decreases.

   In lab scale experiments recovery can be only a fraction of 1% in one pass. The low recovery automatically means that mixed liquor must circulate through pump so many times until it passes through membrane. (for example, mixed liquor must circulate through the pump head 200 times on average before it passes through the membrane, if recovery is 0.5 % in lab scale experiment) 

   The effect of rotary vane pump and centrifugal pump on flux and biosolids yield were compared in a side-by-side lab scale experiment (Kim, 2001). In rotary vane pump, impellers (or vanes) physically touch with the casing that perhaps generates extra shear stress in addition to the hydraulic shear. As shown in Fig. 5a, mean particle size was clearly smaller with rotary vane pump than with centrifugal pump. Due to the smaller particle sizes, much lower flux was obtained with rotary vane pump in tubular membrane system (Fig. 5b).

   The soluble COD of mixed liquor and permeate are compared in Fig. 6, where SMP is included in the soluble COD by definition. It is apparent that the rotary vane pump breaks up floc and microorganisms much more than centrifugal pump so that soluble COD is much higher in both mixed liquor and permeate.

   The biodegradable portion of disintegrated microorganisms by pump is consumed by other live microorganisms and new microorganisms are produced. The amount of new microorganisms should be less than the amount of disintegrated microorganisms. As a consequence, the shear stress of pump reduces biosolids yield as summarized in Table 2. While biosolids yield ranges 0.4-0.5 g MLVSS/g COD in a typical activated sludge process without membrane, the biosolids yield of the MBR with rotary vane pump and centrifugal pump were 0.2 g MLVSS/g COD and 0.3 g MLVSS/g COD, respectively.

   The above observation on sludge yield suggests the following.

  • Care must be taken when projecting a biosolids yield of a full-scale MBR based on a smaller scale pilot. As scale increases, shear effect on biosolids yield decreases. Therefore, there is always a risk of underestimating sludge yield when projecting bigger scale system based on smaller scale system.
  • Types of circulation pump must be factored in when biosolids yield is projected.


a) Mean floc sizeb) Flux

Fig. 5.  Effect of pump shear on floc particle size and membrane fouling in lab scale experiment (Kim, 2001).                                                 


ShearS33Fig. 6.  Variations of soluble COD concentrations in an activated sludge reactor and in the membrane permeate during the operation of a crossflow MBR (Kim, 2001).

                                      Table 2. Effect of pump type on sludge yield in MBR (Kim, 2001)


© Seong Hoon Yoon