Fouling of ultrafiltration and nanofiltration membranes by dissolved organic matter

Fouling is the drop in permeate flux per unit of applied pressure due to the accumulation of materials in the pores and on the surface of the membrane. Dissolved organic matter (DOM) in natural waters is often an important factor in irreversible fouling of membranes. This work characterizes fouling of nanofiltration and ultrafiltration membranes by compounds selected as models for the polysaccharide, polyhydroxyaromatic: and proteinaceous fractions of DOM commonly obtained by pyrolysis-GC/MS. Polygalacturonic acid, rosolic acid and bovine serum albumin were selected as model compounds. Although proteins have been observed to foul membranes, pyrolysis-GC/MS analysis alone of DOM extracted from foulant cakes on membranes yields relatively little proteinaceous material. It was hypothesized that proteins are more significant foulants than is indicated by pyrolysis-GC/MS analysis of the foulant cakes. It was also hypothesized that mixtures of organic molecules would not foul the membrane to the same extent as they will separately and that these flux-reducing foulants were both in the pores and on the surface.; Membranes were employed in a dead-end filtration configuration. Surface topology was characterized using atomic force microscopy, the data from which was used to calculate surface roughness by a multifractal analysis algorithm. Foulant location was investigated using infrared spectroscopy with attenuated total reflectance. These laboratory data were compared to pilot plant data collected at the San Patricio Municipal Water District (SPMWD).; Long-chain PgA molecules and BSA had the strongest influence on fouling. Mixing the different DOM fractions enhanced fouling. Combinations of the different organic molecules produced more fouling than would be predicted from the sum of each compound's fouling potential. The degree to which the organic molecules fouled the membranes was influenced by both the nature of the membrane materials and by their MWCOs. The fouling effect per millimole foulant was higher in the UF range, likely due to greater pore fouling. Results from the laboratory experiments and from the pilot plant at the SPMWD were not strictly comparable, but do indicate that the presence of a substantial deposit on the membrane surface does not necessarily account for an observed reduction in permeate flux.