Organic Fouling of Membranes: Could the guilty NOM fraction please come forward!

It is a well known fact that natural organic matter (NOM) is one of the major contributors to low-pressure membrane fouling but can anyone really pinpoint what specific fraction of NOM is responsible? Based on the variety of research conclusions I have seen, it is likely there is no 'one-size-fits-all' answer.

I am surprised that low-pressure membrane giants such as Siemens/Memcor and GE/Zenon have not put more research effort into solving the organic fouling riddle so that they can better optimize the performance of their products. Perhaps they have and are keeping the solution to themselves for competitive advantage…
There are many ways that NOM can be characterized and typically the research has only looked at one or two of these characteristics simultaneously to identify what fraction of NOM is responsible for membrane fouling.

NOM Ain’t NOM…

NOM characteristics can vary significantly from water source to water source depending on the organic matter in the catchment, geological conditions, climate, etc. The most common variables used to characterize organic matter are size, structure and functionality.

These variables are further defined as follows:

Size: Size fractions can be defined as particulate (>0.45 µm), dissolved (<0.45 µm) and colloidal (~1 nm to 1 µm) where the colloidal fraction crosses over the dissolved and particulate definitions. Molecular weight (MW) distribution is also used to characterize NOM size fractions.

Structure: NOM fractions can also be described in terms of hydrophobic (aromatic) or hydrophilic (aliphatic) structure which indicates whether NOM is water repelling (hydrophobic) or water adsorbing (hydrophilic).

Functionality/Charge: NOM can also be characterized by functional group content (humic and fulvic acids) which impacts the NOM’s (negative) charge density.

To further complicate the characterization, hydrophobic through to hydrophilic NOM can be low to high molecular weight in size, although as a rule of thumb, hydrophobic NOM fractions generally have higher MW.

In addition to NOM characteristics, the low-pressure membrane properties themselves are also likely to play a role as to whether NOM will result in fouling or not. Variables such as membrane pore size, surface charge and surface roughness may impact whether NOM fouling occurs by pore blockage, adsorption or gel layer formation.

So What Causes the Fouling??

With all of this knowledge about NOM and membranes, we must to able to pinpoint which characteristics lead to membrane fouling, right? There have been many studies conducted, from laboratory to full-scale, to investigate the impacts on membrane performance of different NOM fractions. Here is a very brief summary of the findings of some of these:

Howe and Clark (2006) concluded that it is the dissolved and colloidal fraction of NOM in the 100 kDa to 1 µm size range that is responsible for most of the membrane fouling potential. This higher MW fraction of NOM is removed by coagulation and is assumed to be more hydrophobic and more acidic.

Yamamura et al. (2008) and Lee et al. (2004) identified that the hydrophilic fraction of NOM provides a larger contribution to membrane fouling than hydrophobic humic substances. Lee et al. also concluded that macromolecules of a hydrophilic character (e.g. polysaccharides) and/or colloidal organic matter in the hydrophilic NOM fraction may foul low-pressure membranes. It is interesting to note that for the technique used by Lee to isolate the polar fractions of NOM, organic colloids also end up in the hydrophilic fraction. This may be why some generalizations are made that hydrophilic NOM is responsible for membrane fouling when this isolation technique is used while the colloidal material may actually be responsible for the fouling.

Greater flux decline was seen by Lee et al. during microfiltration (MF) compared to ultrafiltration (UF) where MF fouling was caused by pore blockage by the larger macromolecules and organic colloids, while for UF (which has smaller pores), flux decline was caused by gel layer formation which is more easily reversed than pore blockage. Interestingly, hydrophobic/hydrophilic properties of the membranes were found to have minimal impact on membrane fouling.

The proof is in the pilot….

Why are there so many different conclusions as to what fractions of NOM are responsible for membrane fouling? Is it because membrane fouling is due to a combination of NOM and membrane characteristics that have not been looked at simultaneously or is there another characteristic that has not yet been identified?

My conclusion is that all of the critical variables have been identified, but nobody has yet (at least as far as I have found) looked at all of these in the one study. The work conducted by Lee et al. comes closest. Different fractions of NOM have different impacts on membranes with different properties. That sets up a pretty complex research matrix. If you are looking at membrane treatment for a particular water source, at least you just have one set of NOM variables to evaluate.

Unfortunately, I don’t think we are at the point where we can adequately fingerprint the NOM in a water source to be able to predict the fouling potential for a specific membrane type. This will only be determined by trial and error on a continuously operated pilot plant. You will be misguided to aim for a total NOM removal target as the treatment goal when choosing a pretreatment process without also looking at the fouling characteristics of the specific NOM in the membrane feed water.

References

1. Howe, K; Clark, M; “Effect of coagulation pretreatment on membrane performance”, Journal AWWA, April 2006.

2. Yamamura, H; Kimura, K; Okajima, T; Tokumoto, H; Watanabe, Y; “Affinity of Functional Groups for Membrane Surfaces: Implications for Physically Irreversible Fouling”, Environmental Science & Technology, Vol. 42, No. 14, 2008.

3. Lee, N; Amy, G; Croue, J; Buisson, H; “Identification and understanding of fouling in low-pressure membrane (MF/UF) filtration by natural organic matter (NOM)”, Water Research 38 (2004), 4511-4523.