(410f) The Impact of a Pre-Existing Biofilm On Mineral Salt Scaling of RO Membranes

Thompson, J. - Presenter, University of California, Los Angeles
Lin, N. - Presenter, University of California, Los Angeles

Two of the most severe types of RO membrane fouling are biofouling and mineral scaling. Biofouling and mineral scaling are of particular concern in the desalination of tertiary wastewater. Reuse of treated wastewater for inland irrigation or aquifer recharge is driven by the need to attain high product water recoveries in the RO desalting step. The high concentration of microorganisms in wastewater can also cause significant membrane biofouling during the RO desalting step. Studies have shown that even by removing 99% of microorganisms through MF/UF pretreatment processes, microorganisms can still colonize on RO membranes to cause biofouling. At high RO product water recoveries, solution supersaturation conditions can be reached in the membrane channel (with respect to sparingly water soluble mineral salts). As a consequence, mineral salt crystallization in the bulk and deposition onto the membrane surface, as well as mineral crystallization on the membrane surface can lead to severe flux decline and shortening of the useful membrane life. Monitoring techniques are currently lacking in determining the specific type of fouling causing the observed permeate flux decline. Additionally, previous studies have suggested that biofouling could have a significant impact on membrane mineral scaling. A number of recent studies have shown that biofouling can result in enhancement of concentration polarization which in turn could exacerbate the problem of mineral scaling. Therefore, it is important to study the relationship between biofouling and mineral scaling, since they may affect each other. The present study focuses on direct observation of biofouling, followed by scaling by a model scalant gypsum, and evaluates the impact of a pre-existing biofilm, arising from RO of tertiary wastewater effluent (1100 mg/L TDS), on the nucleation and crystal growth patterns of calcium sulfate dihydrate (gypsum) on reverse osmosis membranes. The study was carried out on-site at a municipal wastewater MF/RO treatment plant, utilizing the tertiary wastewater effluent for generating the biofilm. Real-time monitoring of biofouling and mineral scaling was accomplished in-situ via an optical mineral scale detection system previously developed and demonstrated by Uchymiak et al. Studies of mineral scaling in the presence of biofilm were carried out using feed solution in which the bulk saturation index of gypsum was 0.94. Over the period of biofilm growth (~3 days) a significant flux decline was observed, with about 60% of the total flux decline occurring within the first 12 hours of biofilm growth. RO desalting of model scaling solution under operating conditions of supersaturation of gypsum at the membrane surface (initial surface SI = 1.6) resulted in additional flux decline relative to that associated with the biofilm. For the same level of initial gypsum supersaturation at the membrane surface, permeate flux decline experienced with the biofouled membrane was significantly greater than in the absence of the biofilm. This study demonstrates that monitoring flux decline alone may not indicate which type of fouling is occurring, and therefore which membrane cleaning procedure to initiate. The current direct observation method was able to distinguish whether mineral scaling or biofouling is responsible for the observed flux decline. Optical images of the membrane surface also demonstrated, consistent with the flux decline tests, that the biofouled membrane had higher scale coverage. Elemental analysis of the mineral scale on the biofouled membrane revealed the presence of calcium and phosphorous, suggesting that crystallization of calcium phosphate could have occurred at the biofilm growth stage, due to supersaturation of calcium phosphate in the tertiary treated wastewater feed. The crystal number density (on the membrane surface) was ~2-3 times higher on the biofouled membrane suggesting a higher level of crystal nucleation on the biofouled membrane attributed to the presence of the biofilm and partially to the pre-existing calcium phosphate crystals. SEM analysis of the membrane surface confirmed extensive gypsum scaling occurred primarily underneath the biofilm. In some cases, crystal rods grew to sufficient length to puncture and protrude from the biofilm layer. Furthermore, it is unlikely that any significant biofilm growth continued to occur after mineral scaling, since the tertiary wastewater effluent source was removed from the system, and the nutrient source was limited to 3 ppm of ammonium chloride and dextran during mineral scaling, in order to maintain biofilm health and prevent detachment. Therefore, it is postulated that feed solution permeated the biofilm layer where nucleation and growth of gypsum crystals occurred, and to a greater extent than in the absence of biofilm.