1128_An approach to determine the fouling index for a RO/NF



Title: A systematic approach to determine the fouling index for a RO/NF membrane process 

Journal: Desalination (2009)
Authors: June-Seok Choia, Tae-Mun Hwanga, Sangho Leea, Seungkwan Hongb 
aDepartment of Environmental Research, Korea Institute of Construction Technology, Korea
Tel. +82 (31) 910-0320; Fax +82 (31) 910-0291; email: s-lee@kict.re.kr
bDepartment of Civil Engineering, Korea University, Seoul, Korea 

The original and creativity of paper: This study presented the novel method to determine fouling index of nanofiltration (NF) and reverse osmosis (RO) membrane. The paper aimed to provide the better prediction of RO/NF fouling potential. 


A novel method for better prediction of RO/NF fouling prediction was developed by utilizing the different set of membrane filters. Experiments were done by testing fouling via crossflow nanofiltration (NF) unit. Moreover, the characteristics of pretreated waters were examined by applying five pretreatment methods including flocculation, PAC adsorption, ozonation, microfiltration (MF) and ultrafiltration (UF). Since, prediction of fouling potential by SDI and MFI seems to be difficult when there is more than one type of fouling mechanisms were considered. Therefore, a new approach to evaluate fouling potential for various fouling mechanisms is attempted by combining MFI values measured using multiple test membranes. This method was designated to combine fouling indices and also allowed considering the contribution of particles, hydrophobic matters, colloids, and organics to RO/NF fouling at the same time. 

CFI of feed waters can be examined the efficiency of pretreatment by employing a graphical representation in a radial plot. According to these plots, the characteristics of each pretreatment can be naturally compared. For example, Figure 6(d) and 6(f) clearly present that MFI-HP (fouling due to hydrophobic matters) could effectively decrease by ozonation, whereas all fouling potentials could decrease by UF. Figure 6 shows the radial plots of MFI test using a hydrophilic filter (MFI-HL), MFI test using a hydrophilic filter (MFI-HP), and MFI test using an UF membrane (MFI-UF) for various feed waters.

Fig. 6. Graphical representation of combined fouling index (CFI) for various feed waters. The axis was normalized by the
MFI values of raw water. (a) Raw water (without pretreatment). (b) Flocculation pretreatment. (c) PAC pretreatment.
(d) Ozone pretreatment. (e) MF pretreatment. (f) UF pretreatments.

CFI was simply evaluated by assuming a linear combination of each fouling index term:

CFI = w1 M 1+ w2 M2 + w3 M3+ w4

where M1 is the MFI?HL value, M2 is the MFI? HP value, and M3 is the MFI?UF value. The weighting factors, w1, w2, w3, and w4, depend on the characteristics of RO/NF membrane due to those factor closely related to fouling mechanisms. 

The results concluded that:

1. MF and UF pretreatment showed high efficient to remove particles and colloidal.
2. Colloidal foulants could not effectively remove during pretreatment by using ozonation and PAC adsorption.
3. SDI and MFI tests depended on both characteristics of feed waters as well as pore size and hydrophilicity of test membrane.
4. The combined fouling index (CFI) using multiple test membranes was found to be a promising tool for measuring the potential of RO/NF fouling by various mechanisms. 
5. CFI can be used to compare the efficiency of various pretreatment processes. 

Application & further study: The proposed approach will be useful for researcher who interested about fouling index for NF/RO membrane process. 

By Monruedee Moonkhum
Email: moon@gist.ac.kr

첨부 (1)
131.3KB / Download 2