Formatted Title
Removing Short- and Long-Chain Per- and Polyfluoroalkyl Substances from Landfill Leachate: Comparing Two Pilot Studies
Background/Objectives
The presence of per- and polyfluoroalkyl substances (PFAS) in landfill leachate is an increasing concern in solid waste management. Leachate resultant from a municipal solid waste landfill in the Upper Midwest of the United States is currently treated using aerated pond stabilization followed by final treatment and disposal spray irrigation. However, the local regulatory authority indicated that spray irrigation will not be a long-term option for leachate disposal without additional treatment, specifically due to concerns over PFAS. While future PFAS permit limits are not known, based on measured leachate PFAS concentrations the landfill anticipates treatment will need to achieve additional removal of up to >50% PFBA, >90% PFHxA, >99% PFHxS and PFOS, and >99.9% PFOA. Additionally, leachate boron (B), nitrate (NO3-), and total nitrogen (TN) concentrations have been of compliance concern. Improved removal of B, NO3-, TN, and manganese (Mn) is necessary to continue on-site leachate disposal and accommodate potential landfill expansion. Furthermore, removal of total organic carbon (TOC) from leachate was determined necessary for adequate functionality of multiple evaluated leachate treatment methods. Limited options exist for removal of short-chain PFAS, such as PFBA and PFHxA, due to their lower hydrophobicity than long-chain PFAS. These compounds posed a challenge to certain technologies that have been used to remove PFAS from leachate, such as foam fractionation, as was confirmed by bench-scale evaluation. Following an alternatives analysis, two potential leachate treatment trains were identified. The present study was carried out to evaluate these trains for their efficacy to consistently achieve anticipated permit limits for all targeted constituents of concern, and for operational stability and treatment costs.
Approach/Activities
The first leachate treatment train was a constructed wetland treatment system (CWTS) followed by mechanical treatment. Outflow from the CWTS was first treated by an advanced oxidation process (AOP) using a combination of ozone, hydrogen peroxide, and ultraviolet light (UV) targeting TOC. Additionally, media filtration was used for Mn removal; granular activated carbon (GAC) and ion exchange (IX) for PFAS removal; and a second IX for B removal. The hybrid CWTS was evaluated during the summer of 2022. Approximately 500,000 gallons of leachate were treated, with the mechanical treatment train operated between 0.5 and 3.0 gallons per minute (gpm). Pilot testing included varying the ozone and hydrogen peroxide doses to evaluate TOC removal, varying filtration media to evaluate Mn removal, and supplemental bench-scale tests.
The second leachate treatment train was membrane treatment using two stages of reverse osmosis (RO) filtration followed by GAC polishing for PFAS and IX polishing for B. The membrane treatment pilot was evaluated during the summer of 2023, over a series of 55 filtration batch tests. This was a less complex train due to the breadth of contaminants removed by RO filtration. However, the high level of TOC and TDS of the raw leachate presented concern regarding potential membrane fouling and reduced flux. The membrane pilot operation included testing of multiple pH and anti-scalant amendment configurations to determine how to better maximize recovery and decrease fouling.
Results/Lessons Learned
The proposed presentation will share treatment results for B, Mn, NO3-, TN, TOC, and specific PFAS through both pilot treatment trains. Mechanical treatment following the CWTS was able to meet all desired effluent limits, but operational efficacy and efficiency was decreased by insufficient TOC and Mn removal. AOP processes under pilot conditions only achieved approximately 35% TOC removal, corresponding to an AOP effluent TOC of approximately 100 mg/L. Batch testing showed that to achieve the target effluent TOC of <20 mg/L an ozone dose 4,000 mg/L was required, approximately four-fold higher than piloted AOP capacity. Three media were evaluated for Mn removal, though only Pro-OX media was able to achieve targeted Mn removal efficacy.
Due to the time spent optimizing TOC and Mn treatment, downstream PFAS removal unit processes were only operated for nine days. However, this was sufficient to begin to see PFAS breakthrough. GAC breakthrough occurred rapidly, with PFBA reaching 100% breakthrough and PFHxA reaching >50% breakthrough after only six days. By the end of the nine-day evaluation, PFBA was approaching the anticipated effluent limit in the IX effluent, indicating frequent media replacement or media regeneration would be necessary to maintain long-term compliance using this treatment train. While long-chain PFAS remained below anticipated effluent limits, the short media life associated with PFBA and PFHxA highlights challenges associated with removing short-chain PFAS from landfill leachate. Although this treatment train achieved all required effluent limits for a short period of time, operational challenges specific to TOC and Mn removal, rapid media exhaustion associated with short-chain PFAS removal, and general treatment train complexity represent substantial limitations of the tested treatment train to consistently achieve the anticipated leachate effluent limits over long-term operation.
The RO treatment train consistently achieved all anticipated effluent limits. During pilot design GAC polishing for PFAS was not anticipated to be necessary, but early in pilot operation RO effluent concentrations of PFOA exceeded the anticipated effluent limit. Addition of GAC polishing effectively decreased effluent PFOA concentrations below the anticipated limit. Importantly, since RO treatment removed short-chain PFAS to below anticipated limits, GAC polishing media replacement would not be driven by short-chain PFAS, resulting in less frequent media replacement than that required under the hybrid CWTS. Additionally, water recovery demonstrated during RO pilot testing was better than anticipated. Approximately 90% water recovery through the first stage RO was achieved, while limiting time between membrane cleaning to four days. Although consistent operation was a concern under these conditions, data and observations obtained during evaluation of this RO leachate treatment train support continued evaluation of RO as an overall leachate treatment method.