Previous studies have shown that thermal technology can reduce levels of per- and polyfluoroalkyl substances (PFAS) in soil, combust PFAS contained in commercial products, or mineralize PFAS-based industrial chemicals. However, uncertainties remain regarding indirect thermal desorption (ITD) parameterization for application to treatment of specific PFAS-impacted waste as well as the destruction and removal efficiency (DRE) that can be achieved through off-gas treatment by thermal oxidation (TO). The overall objective of this study was to advance the current understanding of ITD/TO’s effectiveness for the treatment of soil containing a typical suite of PFAS found in, but not limited to, aqueous film forming foam (AFFF) formulations manufactured and heavily utilized prior to 2002.
A series of ten tests were conducted on sand spiked with PFAS at concentrations ranging from approximately 6,000 to 19,000 micrograms per kilogram (μg/kg). Spiked feed soil samples and thermally treated soil samples were collected for each test run. Soil samples were collected in duplicate for each condition. Over 60 soil samples were collected. All soil samples were analyzed for PFAS using U.S. Environmental Protection Agency (EPA) Method 537 Liquid Chromatography/Tandem Mass Spectrometry. For the test runs spiked with AFFF, soil samples were also analyzed for Total Oxidizable Precursor Assay.
On four of the ten tests, the off gas from the thermal desorption unit was passed through a TO, and air samples were collected from the exhaust gas. The TO was operated at 1,000 degrees Celsius (°C) with a nominal residence time of 2.0 seconds for the DRE demonstration test runs. Four comparative test runs were conducted with two test runs analyzing the TO exhaust gas for PFAS via EPA Method 0010/Method 537 (to evaluate removal and destruction) and two test runs analyzing the exhaust gas for hydrogen fluoride via EPA Method 26A (to assess mass balance of fluorine).
The demonstration tests showed conclusively that at 650°C, the ITD technology will remediate (i.e., removal by desorption) PFAS in soil to a concentration of less than 1-10 μg/kg, and that a TO can achieve a DRE of greater than 99.9997% for off-gas emissions from ITD-treated PFAS feed material. Test results also demonstrated that ITD technology effectively desorbs PFAS precursors when treating soils that were recently spiked with AFFF. The knowledge of how PFAS behaves in combustion or thermal processes is scarce. One of the more important aspects of this study was to advance the understanding of ITD/TO’s (at the temperatures tested) capability to achieve “irreversible destruction” of PFAS-containing materials (including impacted soils, spent Granular Activated Carbon (GAC), and investigation-derived waste) and if other unintentional PFAS degradation products may have been formed during the process. During the tests where the TO was employed and the exhaust gas was analyzed separately for PFAS and hydrogen fluoride, a mass balance was performed on fluorine to confirm the destruction of PFAS. Within the treated exhaust, the absence of PFAS (under extremely low detection limits at part per trillion levels) coupled with the recovery of the molar equivalent of fluorine (and associated mass balance) spiked within the feed material provided complimentary lines of evidence for PFAS removal and destruction.
Full-scale treatment of PFAS-impacted soil and spent sorbents from groundwater cleanup is currently limited to costly measures including: 1) transport to/disposal at a permitted landfill, 2) transport to/destruction at an incineration facility, and 3) high-temperature regenerative methods of sorbent media. Anticipating that future demand will only increase for PFAS-impacted soil and spent sorbent media, this study has direct benefits on advancing the state-of-the-industry by employing a mature technology (ITD/TO) to provide for “novel” and cost efficient treatment of a class of emerging chemicals. As an ancillary benefit to this study, significant improvements were made relating to laboratory methodology with sample extraction and recovery procedures for analytical EPA Method 0010/EPA Method 537. These improvements will ensure in the future that PFAS can be adequately extracted and recovered from an air sampling train for TO exhaust with Method 0010, thereby improving DRE estimates.