Title: Chemical cleaning of biofouling in reverse osmosis membranes evaluated using magnetic resonance imaging 
Journal: Journal of Membrane Science
Authors: S.A. Crebera, J.S. Vrouwenvelderb, c, M.C.M. van Loosdrechtc and M.L. Johnsa, 
Corresponding author: M.L. Johnsa,

Institute: 
a Department of Chemical Engineering, University of Cambridge, Pembroke Street, Cambridge CB2 3RA, UK
b Wetsus, Centre of Excellence for Sustainable Water Technology, Agora 1, P.O. Box 1113, 8900 CC Leeuwarden, The Netherlands
c Department of Biotechnology, Delft University of Technology, Julianalaan 67, 2628 BC Delft, The Netherlands

The original and creativity of paper: This paper applied magnetic resonance imaging (MRI) for direct studies of chemical cleaning of biofouled RO membrane as well as extract out quantitative measure of cleaning effectiveness.

Summary

In this study, Membrane Fouling Simulator (MFS) was used to study biofouling of spiral wound membrane system (UF, NF, and RO) on lab-scale. Two sizes of MFSs were used for the experiments: the larger flow without spacer (MSF) and smaller flow with spacer (S-MFS). 


Fouling experiments

Fouling experiments were done by feeding stream of dechlorinated tap water supplemented with nutrient solution through the cells. The nutrient solution was composed of sodium acetate (CH3COONa), sodium phosphate (NaH2PO4.2H2O) at concentrations of 1000?g C/L, 200?g N/L, and 100?g P/L respectively. After that, the systems were cleaned with different flow rates and temperatures. For MFS system, the cleaning solutions, i.e., NaOH at 12 pH, 10 mM SDS, and NaOH at 12 pH plus10 mM SDS, were applied at 45 oC and pumped through RO cell for 90 min at flowrage of 100 ml/min. In case of S-MFS cells, the cleaning solution is NaOH at pH 12 heated to 45 oC at 100 mL/min for 90 min. Then, based on these experiments, cleaned and fouled membranes were characterized using MRI.

Fouling and image analysis

Fig. 1 shows structure images for an unfouled and fouled cell. The bright regions are those of strongest signal intensity, while the dark grid pattern is membrane spacer (Fig. 1 a (i)). On the other hand, image of fouled membrane revealed several dark regions which referred to biofouling (Fig. 1 a (ii)). Moreover, velocity images for the unfouled and fouled cell are also shown in Fig. 1(b). The unfouled image (i) presents a relatively homogeneous flow distribution, influenced by the presence of the membrane spacer. The fouled cell (ii) however shows significant different pattern due to a result of the biofilm growth.



Fig. 1. (a) MRI structural images of (i) an unfouled and (ii) a fouled S-MFS. (b) MRI velocity images of (i) an unfouled and (ii) a fouled S-MFS.

Then, an effective surface area (ESA) of cell, which defined as the percentage of the total membrane surface area with significant flow across it, was calculated. For ESA values calculated based on structural image analysis represent as ESAimg; whereas, ESA values calculated based on velocity image analysis represent as ESAvel.

Fig. 2(a) shows the ESAimg images for the unfouled and fouled S-MFS, respectively (extracted from Fig. 1(a)). The result showed that ESAimg reduced from 74% to 70% because of fouling. Fig. 2(b) shows the corresponding images for ESAvel for the unfouled and fouled S-MFS respectively. The ESAvel of unfouled cell is 81%, and reduces to 35% for the fouled cell. This is affected by biofilm development. 



Fig. 2. (a) Gated structural images of (i) an unfouled and (ii) a fouled S-MFS. (b) Gated velocity images of (i) an unfouled and (ii) a fouled S-MFS. 
Multiple cleaning solution tests

Fig. 3(a) shows the effect of NaOH cleaning. The result showed that most of biofilm was removed after cleaning process. ESAimg was further calculated which found to increase from 25% to 44% after cleaning. On the other hand, when system was cleaned using SDS plus SDS and NaOH, images of cleaned and fouled membrane are almost same (Fig. 3(b)). 

Fig. 3. (a) Structural images of MFS (i) fouled and (ii) cleaned using NaOH. (b) Structural images of MFS (i) fouled and cleaned using (ii) SDS, and (iii) SDS + NaOH.

The effect of cleaning solutions on velocity are displayed in Fig. 4(a) and Fig. 4(b), respectively. When the system was cleaned using NaOH, the result shows a significant recovery of flow in previously stagnant regions following cleaning (Fig. 4(a)). Quantitatively, ESAvel was increased from 3% to 23% after cell cleaning; this increasing is proved to be more sensitive to biofilm accumulation. 

Whereas, the system which cleaned using SDS and SDS + NaOH did not show a significant recovery of flow channels as a result of cleaning (Fig. 4(b)). Furthermore, the ESAvel did not change significantly after cleaning as well. 

Fig. 4. (a) Velocity images of MFS (i) fouled and (ii) cleaned using NaOH. (b) Velocity images of MFS (i) fouled and cleaned using (ii) SDS, and (iii) SDS + NaOH. (c) Pressure drop measurements recorded during cleaning of the MFS with (i) NaOH and (ii) a series of cleaning solutions.


This paper showed successful visualization of effectiveness of cleaning process using MRI. Based this valuable methodology, it can be applied to examined fouling potential as fouling coverage ratio.

Application & further study: MRI can be an option tool for identification of fouling accumulation on membrane.

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