| The rejection of twelve pesticides from portable water sources was investigated using reverse osmosis and nanofiltration membranes. These twelve pesticides are simazine, atrazine, cyanazine, bentazone, diuron, DNOC, pirimicarb, metamitron, metribuzin, MCPA, mecoprop, vinclozolin. A flat sheet membrane unit, a 2540 single element unit, a 4040 single element unit and a two-stage pilot plant were constructed to conduct laboratory, bench and pilot scale experiments. The membrane materials utilized in this research were cellulose acetate, polyamide, polyamideurea, and polyvinyl alcohol membranes.; The results in the laboratory study showed pesticide rejection was not directly related to molecular weight for the pesticide studies. Pesticide rejection decreased as mass transfer coefficients for solvent (water) increased. The ranked order of film for pesticide rejection in the laboratory flat sheet tests was polyamideurea > polyvinyl alcohol > polyamide > cellulose acetate. Water matrix did not have any impact on pesticide rejection.; Pesticide rejection by reverse osmosis membranes (MWC < 200) in the bench study was very high (>98%). Pesticides were rejected to less than the detection limit (<0.1 mug/L) in most observations in the bench studies. Pesticide rejection in the bench studies using nanofiltration membrane was diffusion controlled and was significantly affected by flux, recovery and pesticide solubility, charge, diffusion coefficient, and molecular weight.; Pesticide rejections in the pilot study were dependent on membrane system, stage, flux and recovery. Pesticide rejection varied significantly by element size in the pilot tests and was lower for the 2540 single element than either the 4040 single element or 2-1 array pilot system. Pesticide rejection in the pilot study was controlled by diffusion for all three system. Pesticide rejections for the 2540 single element, the 4040 single element, stage 1, stage 2 and 2-1 stage system were always highest for the high flux-low recovery and always lowest for the low flux-high recovery operating condition.; In addition, pesticide rejection by RO membranes were statistically verified using homogenous solution diffusion model, film theory model, modified film theory model and size exclusion model. The results showed the solute mass transfer coefficients were not constant but dependent on flux and recovery. |
