Monday, November 8, 2010

NOM Fouling of Low-Pressure Membranes: Demystifying the Research

A Decision Tree Analysis helps unveil the guilty fraction of NOM.

In my previous post I wrote about my Holy Grail like quest to find what fraction of natural organic matter (NOM) was responsible for low-pressure membrane fouling. After reading numerous research papers, each looking at the membrane fouling potential of only one or two of the many NOM characteristics at once, I was starting to think that perhaps the Holy Grail would be easier to find.

I therefore decided to pull together as many of the findings as possible in a decision tree analysis to see if I could piece together a consistent trend. If so, this decision tree could be used as a guide to determine pretreatment requirements for low-pressure membranes. Gaps on the decision tree could help direct where further research is required.

In developing the decision tree, I focused on the research reported in what I think are four very good papers that collectively evaluate the impacts of a wide range of NOM and membrane characteristics. The result? I was pleasantly surprised to find a consistent trend in the conclusions as can be seen in Figure 1.

Figure 1: Decision Tree for Identifying Low-Pressure Membrane Foulants

The Conclusions

From the decision tree, it is fairly clear that the NOM responsible for fouling low-pressure membranes comes from the high molecular weight fraction. Within this fraction, it is specifically the hydrophilic material consisting of carbohydrates (polysaccharides and proteins) and colloids that cause the membrane fouling. While the four research projects did not all look at the same set of NOM characteristics (Howe just looked at size, Yamamura looked at functionality and charge, while Lee and Humbert looked at size and functionality) when plotted on the decision tree, it can be seen that the research findings for each parameter are consistent.

Howe’s assumption that the NOM foulants removed by coagulation are more hydrophobic and acidic conflicted with the other research results, but his findings that the large Molecular Weight (MWt) colloidal material is responsible for fouling is consistent with the other research. Howe and Lee also found that pore size is important where fouling of microfiltration (MF) membranes is more likely to occur via pore blockage, which is more difficult to reverse, compared to ultrafiltration (UF) fouling which occurs by gel layer formation (UF pores are an order of magnitude smaller than MF pores). Most of the research concluded that negatively charged humics are not a major membrane foulant, probably due to electrostatic repulsion with the membrane materials which are also negatively charged.

In addition to the impact of NOM fractions, Yamamura found that specific membrane materials can impact the degree of fouling where the greater electronegativity of PVDF membranes results in stronger hydrogen bonding of hydrophilic NOM compared to PE membranes which have a lower electronegativity.

Having pulled together this decision tree, at least for me, the mystery of what fraction of NOM is responsible for low-pressure membrane fouling is beginning to unravel. All I need now is for a well funded research team to conduct a study looking at all of the NOM and membrane characteristics together and the Holy Grail will be found – maybe….

References
1. Howe, K; Clark, M; “Effect of coagulation pretreatment on membrane performance”, Journal AWWA, April 2006.
2. 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.
3. Humbert, H; Gallard, H; Jacquemet, V; Croue, J; “Combination of coagulation and ion exchange for the reduction of UF fouling properties of a high DOC content surface water”, Water Research 41(2007) 3803-3811.
4. 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.