Elsevier

Water Research

Volume 44, Issue 12, June 2010, Pages 3545-3554
Water Research

Systematic analysis of micromixers to minimize biofouling on reverse osmosis membranes

https://doi.org/10.1016/j.watres.2010.03.038Get rights and content

Abstract

Micromixers, UV-curable epoxy traces printed on the surface of a reverse osmosis membrane, were tested on a cross-flow system to determine their success at reducing biofouling. Biofouling was quantified by measuring the rate of permeate flux decline and the median bacteria concentration on the surface of the membrane (as determined by fluorescence intensity counts due to nucleic acid stains as measured by hyperspectral imaging). The micromixers do not appear to significantly increase the pressure needed to maintain the same initial permeate flux and salt rejection. Chevrons helped prevent biofouling of the membranes in comparison with blank membranes. The chevron design controlled where the bacteria adhered to the membrane surface. However, blank membranes with spacers had a lower rate of permeate flux decline than the membranes with chevrons despite having greater bacteria concentrations on their surfaces. With better optimization of the micromixer design, the micromixers could be used to control where the bacteria will adhere to the surface and create a more biofouling resistant membrane that will help to drive down the cost of water treatment.

Introduction

Biofouling of water-treatment membranes has a negative economic impact in desalination and water treatment. With biofouling there is a decrease in permeate production and an increase in energy expenditure due to increased cross-flow pressure needed. Biofouling leads to increased cleaning expenditures. It also accelerates the degradation of the membranes increasing membrane replacement costs. Finally, due to the concentration polarization caused by biofouling the water quality of the permeate degrades. Even when wastewater is being pre-treated by microfiltration, biofilms grow on the downstream reverse osmosis (RO) membranes (Sadr Ghayeni et al., 1998). Flemming (1997) estimates that approximately 30% of the total operating costs for Water Factory 21, a RO plant in Orange County, CA, used for wastewater reuse, is for controlling biofouling.

Biofouling is the unwanted growth of biofilms on a surface. In the case of water-treatment membranes, if microorganisms exist in the feed water it is likely that they will adhere to the membrane. Once adhesion has occurred, colonies can grow with the associated extracellular polymeric substance (EPS) forming biofilms thus fouling the membranes. The intention of this work is to minimize the chance for this initial cell adhesion.

To date, a universal successful and cost-effect method for controlling biofouling has not been implemented. Methods for controlling biofouling include minimizing microorganisms and potential nutrients from the feedwater, disinfection of the system, and cleaning of the membranes. Pre-treatment methods for controlling biofouling are site-specific and time and cost intensive (Flemming, 1997). Oxidating biocides can degrade the polyamide RO membranes (da Silva et al., 2006). Cleaning has been found to temporarily improve the flux through the membrane, but biofouling continues to degrade the membrane performance and with subsequent cleanings, the amount of improvement decreases (Vrouwenvelder et al., 1998). Use of biofouling-resistant membranes is a cost-savings alternative to minimize biofouling.

Studies on the use of corrugated membranes and spacers show promise as a method to improve membrane performance. Corrugated membranes are said to enhance flux by promoting turbulence near the membrane wall region. This, in turn, reduces concentration polarization and improves performance (Scott et al., 2000, Vanderwaal et al., 1989, Racz et al., 1986). Scott et al. (2000) illustrate that corrugations can lead to energy savings up to 88% and the savings is dependent on the angle of the corrugations relative to the direction of flow with the largest savings occurring when corrugations were 90° to the direction of flow. Likewise, spacer thickness and configuration can influence the formation of vortices and shear stress, and turbulence at the membrane surface, thus decreasing concentration polarization and increasing flux (Lipnizki and Jonsson, 2002, Sablani et al., 2002, Schwinge et al., 2003). Li et al., 2002a, Li et al., 2002b used computational fluid dynamic (CFD) modeling to show vortices leading to enhanced mass transfer, presumably due to reduced concentration polarization. They indicate that there is an optimal spacer geometry to maximize mass transfer and minimize power consumption.

Control of critical flux, the flux condition at which particles irreversibly adhere to a membrane, can also assist in minimizing biofouling. Wang et al. (2005) did not observe pressure losses or fouling below the critical flux. Neal et al. (2003) demonstrate that critical flux can, in part, be controlled by spacer orientation.

The goal of our studies is to determine whether micromixers can be used to minimize biofouling on the surface of a membrane. We define micromixers as features on the surface of a water-treatment membrane that promote chaotic mixing. By maximizing unsteady flow and the critical flux, we intend to minimize the amount of initial bacterial adhesion to a membrane. The hope is that bacterial adhesion can be reduced to the extent that biofilms will not form or only minimally form, thus minimizing biofouling. Initial studies showed some promise that micromixers could reduce biofouling (Ho et al., 2008). However, the mixed results in that study suggested that a more systematic and controlled study, presented here, is needed.

Section snippets

Cultivation, preparation and enumeration of microorganisms

For each experiment, a new stock of Pseudomonas fluorescens (ATCC 700830) was prepared using a Cryobank™ bead (Copan Diagnostics Inc) with the P. fluorescens culture. The beads were stored in a −20 °C freezer prior to use. For each test, two vials of inoculum were incubated for 24 hours at 30 °C, each with one bead and 9 ml of Trypticase Soy Broth (TSB). Two of these samples were then vortexed and centrifuged and the pellets combined and reconstituted in 9 ml of sterile DI water. 500 μL of this

Flux decline

Comparisons of the testing results for the different membranes are presented in Fig. 3. The values of the different parameters used to generate the box plots are included in Table 2. The lowest rates of flux decline (the lower the rate of flux decline, the less fouling) were for the membranes without chevrons (blanks) tested with spacers (Fig. 3A). The membranes with printed chevrons had the next lowest rate of flux decline and the untreated blank membranes had the highest rate of flux decline.

Discussion

It is clear that membranes with printed chevrons perform better than untreated or blank membranes when they are tested without spacers. This is indicated by both the slower rates in flux decline (and therefore less fouling) and the lower measured median bacteria concentration on the membrane surface. This observation helps to confirm our hypothesis that micromixers can work to decrease bacterial adhesion and thus biofilm growth on the surface of a water treatment membrane.

However, the flux

Conclusions

The micromixer designs tested in this study show promise as a method to control where bacteria will adhere to a water treatment membrane, and therefore controlling biofouling. This is seen by both the slower rate of flux decline and the lower bacteria concentration on the membrane surface as compared to an untreated membrane. However, for micromixers to work the design needs to be further optimized in order to outperform the rate of flux decline on untreated membranes with spacers. The

Acknowledgements

This research was funded under the Sandia National Laboratories Laboratory Directed Research and Development (LDRD) program. Menachem Elimelech, Moshe Herzberg, and Atar Adout from Yale University are thanked for their advice in designing and conducting these experiments. Laura Halbleib assisted with the Design of Experiments Analysis. Sandia National Laboratories is a multi-program laboratory operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin company, for the U.S.

References (27)

Cited by (9)

  • Reverse osmosis desalination modules containing corrugated membranes – Computational study

    2017, Desalination
    Citation Excerpt :

    The patterns on the membrane caused a chaotic flow in the feed channel [9]. The membrane with patterns had less bio-fouling than the membrane with spacers, but the membrane with spacers had greater water flux [10]. Kang and Chang [11] performed numerical simulations and flow visualizations to study flow and mass transfer in a channel with ribs attached to electrodialyzer membranes in different configurations.

  • A novel reverse osmosis membrane with regenerable anti-biofouling and chlorine resistant properties

    2013, Journal of Membrane Science
    Citation Excerpt :

    These significant changes in the salt rejection and water flux of the virgin membrane result from the deterioration of the functional structure of the crosslinked aromatic polyamide active layer (as mentioned in Section 3.4) caused by chlorination. Moreover, the metal ions such as iron and aluminum presenting in tap water could act as a catalyst in the membrane chlorination process [53,54]. For the IU-modified membrane, the slight change of salt rejection results from the grafted IU on the membrane surface that protects the aromatic polyamide functional layer against free chlorine attack.

View all citing articles on Scopus
View full text