Soluble aluminum (Al 3+ ) may react with both ambient silica and antiscalant components to form colloidal foulants during reverse osmosis (RO) treatment. Whereas conventional treatment (coagulation/filtration/sedimentation/dual-media filtration) was being used prior to RO, aluminum sulfate (alum) and polyaluminum chloride (PACl) coagulants were evaluated at ambient pH (pH 7.8 to 7.9) and suppressed pH (pH 6.7) in an effort to lower the total aluminum to below 50 µg/L-a level previously observed to prevent RO membrane fouling. Additional tests were conducted with 5 mg/L citric acid added to the RO influent to chelate the soluble aluminum fraction. All tests were conducted with 1.5 to 2.5 mg/L chloramines present. Testing of a RO process fed with optimized alum-or PACl-coagulated water showed that PACl outperformed alum regardless of pH. Alum coagulation at ambient pH resulted in 184 to 273 µg/L total aluminum passing through the filtration process. Only by lowering the mean influent water pH to 6.7 was the mean soluble aluminum residual (45 µg/L) for alum coagulation reduced to below the 50 µg/L aluminum goal. Regardless of pH, for alum-coagulated waters, the higher aluminum carryover resulted in severe RO membrane fouling within 500 hours of operation. Only when a chelating agent (citric acid) was added to the RO feed was the loss in productivity and selectivity arrested. However, PACl consistently met the 50-µg/L goal for both total and soluble aluminum for all pH levels tested, which resulted in more stable membrane performance over time. Further research on the compatibility of PACl and polyamide membranes in the presence of chloramines is needed as data from this project suggest PACl coagulation may facilitate membrane oxidation.
The potential impact of coagulants on mineral scaling in reverse osmosis (RO) feed treatment of brackish water was assessed experimentally, with respect to calcium sulfate and barium sulfate scaling potential, via a bulk crystallization and membrane scaling tests. The scale suppression effectiveness of six commercial antiscalants was first ranked based on measured observed bulk crystallization induction times. Bulk crystallization tests with three standard coagulants (ACH, FeCl 3 , and polyDADMAC), when used independently, demonstrated retardation of the observed crystallization induction time. However, along with an antiscalant dosing, antiscalant scale suppression was significantly reduced and scaling occurred at the same or greater severity relative to the additive-free feed solution. RO membrane scaling tests also indicated that, when the coagulants were present, antiscalant effectiveness was significantly reduced, consistent with the expectation based on bulk crystallization tests. Although the coagulants demonstrated a slight scale suppression quality in bulk crystallization tests at low dosages (<10 mg/L), membrane tests, even at a low dose of ∼3 mg/L, demonstrated scale formation that was at the same or greater level relative to the additive free solution. Optical imaging of the membrane surface demonstrated that, in essentially all cases, greater surface scaling was linked to a higher surface number density of mineral crystals. The percent flux decline increased with increased percent of scaled membrane area. Percent flux decline that was lower than the percent area scaled was encountered for scaling tests without additives and for the case of ACH along with added antiscalant, suggesting partial permeation through the gypsum rosettes. Scaling tests with ACH and with the antiscalant along with FeCl 3 or polyDADMAC revealed higher percent flux decline than the percent scaled membrane area. This behavior was linked to scaling by barium sulfate as indicated by SEM imaging and elemental surface analysis and possibly the complexation and/or adsorption of the coagulant onto the membrane surface. The results of the present study suggest that, when feed treatment includes the use of coagulants, excess antiscalant may be needed in order to offset the adverse impact of the coagulants on antiscalant effectiveness.
The Metropolitan Water District of Southern California carried out a study to evaluate manganese (Mn) desorption from filter media during biological filtration startup. Mn contamination of the filter media was a by‐product of prior use of ferric chloride (FeCl3), which contains Mn, for coagulation and free chlorine exposure. During pilot‐ and full‐scale testing at pH 6.5, FeCl3 coagulation (without chlorine) released soluble Mn at nearly four times the level released by aluminum sulfate (alum; 105 μg/L for FeCl3 and 27 μg/L for alum). Pilot‐scale experiments demonstrated that Mn was present in the filter column in three valence states [Mn(II), Mn(III), and Mn(IV)]. Mn(II) in the filter influent was from the FeCl3. Mn on the anthracite media was hypothesized to be Mn(II)/Mn(IV) and first physically displaced by ferric iron, aluminum, and hydrogen, and then catalytically oxidized to Mn(IV) by silica sand.
The changes in polarity of natural organic matter (NOM) across a pilot-scale conventional water treatment plant with pre-ozonation and biofiltration were analyzed using the polarity rapid assessment method (PRAM). PRAM is a novel method developed for the characterization of the polarity of NOM in environmental samples. Characterization was accomplished by monitoring UV absorbance (UVA) at 254 nm. Under ambient conditions, the NOM was characterized as highly anionic (.50%), with limited hydrophobic (10 -25%) and hydrophilic (0-10%) domains. Results showed temporal variations in the polarity of NOM entering the pilot plant. These differences were ascribed to changes in water blend and variability in NOM characteristics. Changes in NOM chemistry were also observed as the organics passed through the treatment process. Ozonation decreased the hydrophobicity and increased the polarity of NOM, confirming results from previous research utilizing resin fractionation methods. Coagulation, flocculation and biofiltration resulted in the decrease of the hydrophobic and hydrophilic character of the chromophoric NOM.However, these polarity changes varied during the pilot plant run, suggesting a more precise evaluation of NOM through unit operations is warranted.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.