Evaluation and Life Cycle Comparison of Ex-Situ Treatment Technologies for Poly- and Perfluoroalkyl Substances (PFAS) in Groundwater

All three technologies (GAC, IX, and high-pressure membranes) studied in this research showed great ability to remove PFAS from different source waters. However, the removal efficiency was impacted by various factors.

Overall, TOC concentration had the most dramatic effect on GAC use rate. In groundwater with a TOC concentration of 1.3 mg/L, the GAC use rate was 10 times that in groundwater with a TOC concentration of <0.3 mg/L. EBCT and GAC type were also important parameters that affected GAC use rates.Doubling EBCT from 10 minutes to 20 minutes decreased GAC use rates by a factor of about two for the tested GAC and water matrix combination. Out of four major inorganic anions investigated (chloride, bicarbonate, sulfate, and nitrate), nitrate had the most pronounced impact, negatively impacting the adsorption of short-chain PFAS. Fresh and weathered AFFF constituents had a negligible impact on the removal of targeted PFAS. GAC use rates were also strongly impacted by different regulatory scenarios; both lower concentration targets and inclusion of shorter-chain PFAS into treatment criteria led to substantial increases in GAC use rates.

Initial PFAS concentration and competition among co-occurring PFAS had a negligible effect on PFAS breakthrough curves for the selected IX resin. Doubling EBCT from 1.5 minutes to 3 minutes had a negligible impact on PFAS removal. TOC concentration negatively impacted IX performance, but not as strongly as GAC performance. Long-chain PFAS were more impacted by TOC than short-chain PFAS. Out of four major ions investigated (chloride, bicarbonate, sulfate, and nitrate), nitrate had the most pronounced impact, negatively impacting PFAS removal by IX. Short-chain PFAS were more strongly affected by nitrate than long-chain PFAS. Similar to GAC, IX use rates were strongly impacted by different regulatory scenarios; both lower concentration targets and inclusion of shorter-chain PFAS into treatment criteria led to substantial increases in IX use rates.

For the PFAAs investigated and over the range of operating conditions evaluated, tight NF and RO membranes provided higher separation or rejection than the loose NF membrane. For a DI water matrix, RO membrane provided greater than 99% rejection for all PFAAs, with permeate concentrations being near or below the limit of quantification. For waters with higher ionic strength, loose NF and to a lesser extent tight-NF and RO exhibited lower PFAA separation. In particular, PFAA separation by loose NF was observed to be significantly and detrimentally impacted by the presence of salts and subsequentially high product water recoveries. The addition of sodium sulfate negatively impacted the rejection of PFBS, PFOA, and PFOS by loose NF. Compared to the addition of sodium sulfate, calcium chloride had a minimal impact on the rejection of PFAAs with separation being marginally (5–10%) lower than baseline conditions across the recovery setpoints evaluated.

As part of the scope, an Excel^®-based decision support tool was developed. The tool provides a pathway to comprehensively evaluate PFAS treatment technology options consisting of GAC, IX, and high-pressure membrane systems for groundwater remediation. Treatability data generated during this project was used to evaluate the impact of various factors on costs associated with PFAS treatment using GAC, IX or high-pressure membranes. For GAC and IX treatment, the highest treatment costs were associated with treating high TOC groundwater to the proposed USEPA MCL for PFOA. High-pressure membrane treatment was found to be significantly more expensive than GAC and IX due to higher capital and O&M costs, with concentrate management costs being substantial depending on the disposal option.