You are here:
Impacts of Water Relevant Factors on GAC Uptake of Per-and-Polyfluoroalkyl Substances: Implications for RSSCT Scaling and Modeling
Citation:
Abulikemu, G., J. Pressman, D. Wahman, G. Sorial, T. Sanan, L. Haupert, J. Burkhardt, S. Smith, E. Kleiner, E. Stebel, B. Gray, C. Gastaldo, E. Hughes, AND S. Pedigo. Impacts of Water Relevant Factors on GAC Uptake of Per-and-Polyfluoroalkyl Substances: Implications for RSSCT Scaling and Modeling. Presented at WQTC, Dallas, TX, November 05 - 09, 2023.
Impact/Purpose:
Granular activated carbon (GAC) adsorption is well studied for per-and-polyfluoroalkyl substances (PFAS) treatment. However, there is a lack of systematic and controlled studies conducted at drinking water relevant conditions. This presentation will cover USEPA research to evaluate the effect of multiple factors, including the properties of GAC, PFAS, and various water quality conditions. Research findings offer a clearer understanding of PFAS adsorption across drinking water matrices, providing a quantitative basis for generating more accurate RSSCT and mathematical modeling predictions of full-scale GAC treatment performance.
Description:
Granular activated carbon (GAC) adsorption of per-and-polyfluoroalkyl substances (PFAS) is an established and effective water treatment technology for PFAS removal. However, due to the multifaceted nature of adsorption, GAC treatment system efficacy is site specific. Therefore, adsorption tests are conducted under representative conditions for process design and optimization. To save time and resources, rapid small-scale column tests (RSSCTs) are used to simulate full-scale system performance. However, RSSCT scaling approach assumptions have not been validated for PFAS. Furthermore, pilot-scale column tests are often needed to calibrate RSSCTs due to affecting factors that do not scale precisely according to RSSCT assumptions. Fundamental adsorption parameters acquired under relevant conditions are the basis for determining column dimensions, and once established, enable mathematical simulations of breakthrough curves for designing and conducting RSSCTs effectively and efficiently. Previous research established baseline single-solute (S-S) batch adsorption kinetic and equilibrium parameters for nine PFAS with each of three commonly used GACs (mean particle diameter, dp, mean = 59 µm) in buffered milli-Q water. To investigate RSSCT scaling assumptions on adsorption kinetics, batch adsorption kinetics were tested with a subset of four PFAS and one GAC with two additional dp, mean (110 µm and 200 µm). Building upon previous research, the current research conducted S-S isotherm tests (IST) with dp, mean of 110 µm and 200 µm to evaluate the RSSCT assumption of constant adsorption capacity with varying dp, mean. The Freundlich equation well described the isotherm data. Although varying initial concentrations (C0) should not impact estimated Freundlich isotherm parameters for a given equilibrium concentration (Ce) range, preliminary IST with varying C0 yielded different isotherm parameter estimates. Therefore, additional S-S IST investigated the effect of C0 on adsorption isotherm parameters. Subsequently, competitive adsorption among PFAS and between PFAS and other drinking water constituents were evaluated with: (1) multi-solute (M-S) IST with one GAC and a nine-PFAS mixture in milli-Q water; (2) with controlled additions of inorganic ions (Ca2+ and SO42-); and (3) with natural organic matter co-loading and pre-loading. Finally, pH and ionic strength effects were investigated with additional M-S IST. In general, Freundlich kf increased with increasing C0 and decreasing dp, mean, while 1/n was constant over different C0 but decreased with decreasing dp, mean. The controlled experiments isolated and quantified the effects of multiple factors on Freundlich parameters. With such information, site specific corrective factors can be incorporated into RSSCT scale-up approaches and mathematical models to generate more accurate full-scale predictions of PFAS breakthrough in GAC columns.