No abstract

A B S T R A C TAn ultrafiltration (UF) membrane process integrating ozone oxidation prior to a coagulation, flocculation, and sedimentation (CFS) pretreatment configuration processing surface water has been evaluated at the pilot-scale. Unlike prior research limited to short-term bench-scale evaluations, this current study provides information regarding the application of ozone oxidation prior to a CFS-UF pilot process operating over a four-month period (2,800 pilot runtime hours). In this work, changes in the long-term fouling behavior of the UF membrane process in response to the application of ozone prior to CFS pretreatment were characterized using fouling indices. When an average of 2.5 mg L −1 of ozone was applied prior to coagulation requiring 27 mg L −1 of polyaluminum chloride and a UF operating water flux of 85 L h −1 m −2 , the chemically reversible and hydraulically irreversible fouling indices increased by 59 and 40%, respectively. A reduction of chemically irreversible fouling concomitant with a continuous improvement of normalized specific flux was observed over 1,240 pilot runtime hours of ozone application. The total fouling index decreased by 41% as compared to the baseline CFS-UF configuration. This research indicates that the use of ozone oxidation prior to a CFS-UF configuration can reduce membrane fouling when integrated with conventional surface water treatment.

When found in excess, phosphorus (P) has been linked to surface water eutrophication. As a result, adsorbents are now used in P remediation efforts. However, possible secondary toxicological impacts on the use of new materials for P removal from surface water have not been reported. This study evaluated the toxicity of adsorbent materials used in the removal of P from surface water including: fly ash, bottom ash, alum sludge, a proprietary mix of adsorbents, and a proprietary engineered material. Toxicity screening was conducted by performing solid-liquid extractions (SLEs) followed by the bacterial bioluminescence inhibition test with a Microtox® M500. Of the materials tested, the samples extracted at lower pH levels demonstrated higher toxicity. The material exhibiting the most toxic response was the iron and aluminum oxide coated engineered material registering a 66-67% 15-min EC50 level for pH 4 and 5 SLEs, respectively. However, for SLEs prepared at pH 7, toxic effects were not detected for this engineered material. Fly ash and bottom ash demonstrated between 82 and 84% 15-min EC50 level, respectively, for pH 4 SLE conditions. Dried alum sludge and the proprietary mix of adsorbents were classified as having little to no toxicity.

The hazard analysis and critical control point (HACCP) process is being widely adopted as a design, control, and operational methodology to maintain the reliability of delivering water quality that is safe for public health. This study applied the HACCP methodology to identify and assess the reliability of critical control points and critical monitors to manage acute and chronic health risks in potable reuse treatment trains. Specifically, a failure analysis was performed for full‐scale ultrafiltration and reverse osmosis critical control points to determine the reliability of critical monitors and their associated impacts on finished water quality. The results supported the use of ultrafiltration and reverse osmosis membranes as critical control points in potable reuse and identified the sensitivities of both current and emerging critical monitoring parameters, including turbidity, total organic carbon, ultraviolet absorbance at 254 nm, conductivity, fluorescence, calcium, sulfate, sucralose, and pressure decay tests.

In this research, we investigated the influence of feedwater ionic strength on diffusion of divalent ions through a hollow-fiber nanofiltration membrane. The results indicated that solute flux of magnesium was increased as a result of elevating the ionic strength in the feedwater. Specifically, the feedwater ionic strength was observed to have a nonlinear impact on the diffusion of magnesium during the nanofiltration process, which was under-predicted by the homogeneous solution diffusion (HSD) model. This result suggested that elevating the feedwater ionic strength had reduced the strength of the electrostatic double layer at the membrane surface. We then developed a modification of the HSD model (referred to as the HSD-IS model) which incorporated an empirical term related to the effect of feedwater ionic strength (IS) on diffusion of magnesium. The root mean squared error of the HSD-IS model was improved by 77% as compared to the HSD model, which did not incorporate a term related to feedwater ionic strength. This improvement suggested that feedwater ionic strength should be considered when modeling hardness removal during nanofiltration.