Precision in Practice: Water Quality Optimization for UF Plant Efficiency

Applicability A 52 MGD Water Treatment Plant was upgraded to use submerged ultrafiltration (UF) membrane to enhance energy efficiency. The plant consists of 12 trains of Veolia ZeeWeed 1000 membranes. The plant utilizes a pre-treatment system before membrane application, incorporates a Biological Treatment Unit (BTU), coagulation with alum (below 200 mg/L) and lime for pH adjustment, and Powered Activated Carbon (PAC) addition as necessary for taste and odor control. Additionally, flocculation and sedimentation precede the membrane system, with polymer addition to aid settling post-rapid mix. Clarified water undergoes two sets of four travelling band screens, directing flow to UF Trains 01-06 and Trains 07-12, respectively. The facility identified suboptimal performance in UF Trains 1-6 (fed from basins A1 and A2) compared to Trains 7-12 (fed from basin B). To address this, Carollo Engineers conducted a bench-scale study investigating the potential stabilization of UF hydraulic performance across the entire facility through adjustments in water quality. This objective was subdivided into three specific goals: enhancing the filterability of settled water from Pretreatment Basins A1 and A2, formulating an optimal dosing strategy for pretreatment chemicals, and validating proposed changes through benchtop membrane filtration studies. Our presentation will cover the results and learnings from the study and provide a practical and applicable lessons for water quality optimization. Methodology The testing methodology encompassed four distinct assessments. Firstly, Jar Testing simulated rapid mix conditions in a 6 x 2 L gator jar-style unit, generating supernatant aligned with operational conditions in Basins A1 and A2 for subsequent membrane filterability tests. Zeta Potential Testing and Titration measured zeta potential using the Malvern Zeta Sizer, incorporating incremental coagulant doses in a 2L beaker of test water, with coagulant demand calculated at zero zeta potential. Benchtop Pencil Membrane Filtration utilized water from post-treatment basins and jar test supernatant, applying preset standardized filtration and backwash conditions to the pencil module, with resistance determined through pressure transducers and flow monitoring. Finally, Water Quality Parameters, detailed in Table 1, outlined the parameters and testing methods for both plant water and test samples, utilizing water sourced from basins A1, A2, and B for jar testing and benchtop membrane apparatus. Results The pencil module filtration tests comprised 15 samples (7 from the plant, 7 from jar tests, 1 raw water). Out of the 7 plant samples, 3 were used to establish baseline filterability, 2 were used to study the effect of pH and 2 for the effect of free chlorine. Jar test samples investigated polymer dosing effects (4 samples) and optimal parameter combinations (3 samples). Each test included 30 runs, with backwashing after each 4-minute run, collecting filtrate volume (m3/m2) and resistance (m-1) data every 4 seconds. Data from each of the 30 runs per test underwent Inter Quartile Range (IQR) outlier removal. The Resistance vs. Filtrate Volume graphs for the 15 tests revealed two fouling types. Total Fouling, representing the slope of Resistance vs. Filtrate Volume for a single run, was determined by taking the median of total fouling values (see Figure 2) from all runs within a test. This value signifies the fouling removable through backwash or chemical cleans. Hydraulically Irreversible fouling, calculated as the slope of Resistance vs. Filtrate Volume across all runs for a test, denotes the fouling requiring chemical cleans for removal. A summary of the results is available in the table provided. The Resistance vs Filtrate volume graph (see Figure 1) can help visualize how the two parameters are calculated. Comparing the fouling results with the dosing strategy provided various insights. It showed that pH elevation has limited relevance to fouling. Increasing pH requires a proportional rise in coagulant to counter adverse effects on TOC removal. Chlorine, up to 1.5 mg/L, effectively reduces total fouling, but surpassing this threshold has minimal additional impact. The optimal polymer dose is around 0.05 mg/L; exceeding it leads to a marked increase in total fouling, highlighting the delicate balance needed for effective water treatment. Conclusion The research underscores the importance of tailored water quality optimization in membrane systems. It reveals that each system may behave differently depending on hydraulics, feed water quality, and end goal of treatment. The study's pivotal takeaway lies in recognizing the intricate balance required for effective water treatment, shaping a roadmap for enhanced efficiency in membrane-based water treatment plants. Relevance to Audience This presentation is tailored for operators and design engineers invested in advancing membrane filtration technologies, particularly within large-scale water treatment plants. Attendees will gain valuable insights into practical strategies for enhancing filterability, formulating optimal dosing strategies, and validating changes through membrane filtration studies. The relevance lies in the application of these innovative approaches, applicable not only to resolving hydraulic differentials but also to the broader context of optimizing membrane systems in water treatment plants.