ABSTRACT
The development of biofilm on the surface of filtration membranes is the main fouling component of water filtration systems. Chemical cleaning is only partially effective in removing biofilm components from the membrane surface. In order to identify opportunities to improve the efficiency of commercial cleaning solutions used in nanofiltration, we compared the in vitro efficacy of different commercial treatments, with or without the addition of polysaccharidases, to clean fouled membrane samples. The treatments were tested at two stages of biofilm development corresponding to 80 (D80) and 475 (D475) days of filtration in an industrial plant. The cleaning efficiency was evaluated by comparing the ATR-FTIR spectra before and after cleaning. At D80 and D475, all cleaning solutions led to a reduction of infrared signals from the biofilm. At D80, enzymatic alkaline detergent (AEDT) treatment was significantly more effective than alkaline detergent (ADT) treatment in removing proteins, but no significant difference in efficacy between the two treatments was observed for polysaccharides. The addition of polysaccharidases to AEDT did not bring any significant efficiency gain. At D475, ADT and AEDT treatments had the same efficacy, but the addition of polysaccharidases to the AEDT treatment significantly increased the removal of polysaccharides and proteins from the membrane surface. In conclusion, polysaccharidases can increase the in vitro efficacy of a commercially available alkaline enzymatic detergent cleaning solution against sufficiently developed biofilms. These results pave the way for the development of new cleaning solutions containing polysaccharide degrading enzymes for the cleaning of membranes used in the production of drinking water. Further experiments are needed to characterize the mechanism of this polysaccharidase effect and to confirm this increase in cleaning efficiency in an industrial context.
ABSTRACT
The kinetic formation of biofilms developing on nanofiltration (NF) membranes was studied for 2 years in the water production unit of Méry-sur-Oise, France. New membranes were set up in a pilot train integrated to the plant and autopsied after operation for 7, 80, 475 and 717 days. The biofouling layer was studied by confocal laser scanning microscope after 4',6-diamidino-2-phenyindole dihydrochloride and lectin staining, and by attenuated total reflectance-Fourier transform infrared spectroscopy and rheology experiments. Three stages of biofilm growth were discriminated: (1) the presence of sessile microcolonies embedded in an exopolymeric matrix (after filtration for seven days); (2) membrane coverage expansion through microcolony development and biofilm growth in three dimensions (up to 80 days filtration); and (3) biofilm maturation by densification (after filtration for 80-717 days). Biofilm maturation resulted in total coverage of the membrane surface and matrix residue diversification, development of the polysaccharide network, and the strengthening of matrix cohesion through viscosity and elasticity increases. The wettability and permeability of the fouled NF membranes decreased quickly and continuously throughout the biofilm development process. The longitudinal pressure drop (LPD) increased only after the biofilm reached a quantitative threshold. The decline in membrane permeability may be the result of contributions from many fouling mechanisms but the LPD was more substantially influenced by biofilm development.
