Influence of Residual Nonaqueous-Phase Liquids (NAPLs) on the Transport and Retention of Perfluoroalkyl Substances.

Per- and polyfluoralkyl substances (PFAS) are known to accumulate at interfaces, and the presence of nonaqueous-phase liquids (NAPLs) could influence the PFAS fate in the subsurface. Experimental and mathematical modeling studies were conducted to investigate the effect of a representative NAPL, tetrachloroethene (PCE), on the transport behavior of PFAS in a quartz sand. Perfluorooctanesulfonate (PFOS), perfluorononanoic acid (PFNA), a 1:1 mixture of PFOS and PFNA, and a mixture of six PFAS (PFOS, PFNA, perfluorooctanoic acid (PFOA), perfluoroheptanoic acid (PFHpA), perfluorohexanesulfonate (PFHxS), and perfluorobutanesulfonate (PFBS)) were used to assess PFAS interactions with PCE-NAPL. Batch studies indicated that PFAS partitioning into PCE-NAPL (Knw < 0.1) and adsorption on 60-80 mesh Ottawa sand (Kd < 6 × 10-5 L/g) were minimal. Column studies demonstrated that the presence of residual PCE-NAPL (∼16% saturation) delayed the breakthrough of PFOS and PFNA, with minimal effects on the mobility of PFBS, PFHpA, PFHxS, and PFOA. Breakthrough curves (BTCs) obtained for PFNA and PFOS alone and in mixtures were nearly identical, indicating the absence of competitive adsorption effects. A mathematical model that accounts for NAPL-water interfacial sorption accurately reproduced PFAS BTCs, providing a tool to predict PFAS fate and transport in co-contaminated subsurface environments.

Influence of aqueous film forming foams on the solubility and mobilization of non-aqueous phase liquid contaminants in quartz sands.

At sites where aqueous film forming foams (AFFFs) are used for fire suppression or training activities, interactions between dissolved foam constituents and organic liquids could alter contaminant migration in the subsurface. In this study, batch reactor and column experiments were conducted to investigate the potential for AFFF solutions to enhance the solubility and mobility of three representative non-aqueous phase liquid (NAPLs), JP-4 jet fuel, trichloroethene (TCE), and tetrachloroethene (PCE). For AFFF concentrations up to 5% wt. (50,000 mg/L), aqueous solubilities of TCE and PCE increased by less than 50%, indicating the absence of micellar solubilization. However, NAPL-water interfacial tensions were reduced to less than 1.5 mN/m and resulted in accumulation of up to 2.25 mg/m2 of AFFF at the NAPL-water interface. To assess the potential for AFFF to mobilize residual (entrapped) NAPL at a field application rate of 3% wt. (30,000 mg/L), columns were packed with two size fractions of Ottawa sands (20-30 mesh and 60-80 mesh) that yielded residual NAPL saturations ranging from 11.7 to 17.6%. Following injection of 3 pore volumes of the 3% wt. AFFF solution, partial mobilization of residual NAPL was observed for PCE, TCE, and JP-4, with saturation reductions of 0.7 to 2% in 20-30 mesh and 0.3% to 1.3% in 60-80 mesh Ottawa sand. The columns were then flushed with an ultralow-IFT surfactant solution consisting of 4% wt. 1:1 Aerosol AY/OT, which resulted in nearly complete mobilization of the remaining residual NAPL. When NAPL desaturation curves were expressed in terms of the total trapping number (NT), the threshold value of NT required for NAPL mobilization by 3% wt. AFFF was approximately 2 × 10-5, consistent with previous studies. These findings demonstrate that AFFF solutions can lead to partial mobilization of residual NAPL, and that the total trapping number concept can be used to predict such behavior a priori. In addition, the observed IFT reductions are sufficient to alter NAPL accumulation and redistribution behavior in the subsurface through enhanced spreading above the water table (e.g., JP-4) or penetration into lower-permeability media (e.g., TCE, PCE).

Development and Validation of a Two-Stage Kinetic Sorption Model for Polymer and Surfactant Transport in Porous Media.

Understanding the sorption processes is critical to the successful design and implementation of a variety of technologies in subsurface application. Most transport models assume minimal interactions between adsorbed species and, thus, are unable to accurately describe the formation of adsorbed bilayers. To address this limitation, a two-stage kinetic sorption model is developed and incorporated into a one-dimensional advective-dispersive-reactive transport simulator. The model is evaluated using data obtained from column experiments conducted with a representative polymer [gum arabic (GA)] and a nonionic surfactant [Witconol 2722 (WT)] under a range of experimental conditions. Model simulations demonstrate that the first-stage polymer/surfactant-surface sorption rate is at least 1 order of magnitude greater than the second-stage rate, associated with bilayer formation, indicating that the first-stage reaction is more favorable. The reversibility of the second-stage sorption process is found to be compound-specific, with irreversible sorption observed for GA and prolonged tailing observed for WT. This study demonstrates that the developed two-stage kinetic model is superior to a two-stage equilibrium-based model in its replication of two-leg breakthrough curves observed in core flood experiments; the normalized root-mean-square error between measurement and regressed model simulations was reduced by an average of 41% with the kinetic approach.