This proof-of-concept project evaluated infrared heating in a small, mobile thermal desorption unit (TDU) to separate per- and polyfluoroalkyl substances (PFAS) from soil. Indirect infrared heating relies on radiant heating and does not require the production or circulation of a carrier gas to heat the soil. The main objectives of this project were to:
Heated gasses were piped through impacted soil. Once gasses cooled, the PFAS were treated on vapor-phase activated carbon (VGAC). The VGAC was returned to the vender for regeneration and PFAS in the VGAC was destroyed during regeneration. Vapors were monitored using C-18 vapor adsorption cartridges (C-18 cartridges).
The technology was applied in a small, mobile unit designed to be transported to and used at remote locations. Propane fuel was used to heat vapors in the TDU that were then cooled in a condenser, treated using VGAC, and returned to the TDU. Two soil batches were treated, a Low Concentration Test [PFAS 516 μg/kg, PFOS 429 μg/kg, and PFOA 5.55 μg/kg] and a High-Concentration Test [Total PFAS 2,244 μg/kg, PFOS 1,800 μg/kg and PFOA 22.6 μg/kg]. The soil was heated to over 350 degrees Celsius (°C) to mobilize the PFAS. PFAS concentrations in soil were tracked using daily composite samples. The amounts of materials used and concentrations of PFAS were used to construct a PFAS mass balance to determine capture efficiency and understand the fate of PFAS in the process.
The Low-Concentration Test met temperature requirements (greater than 350°C) and virtually all PFAS was removed by Day 5, with traces remaining at Day 8 (maximum detection of 2.96 μg/kg PFOS, total PFAS of 3.53 μg/kg, 99.6 percent removal of total PFAS). The High-Concentration Test did not meet the temperature requirements (sustained temperatures ranged from 307°C to 346°C) and removal of PFAS was poor. PFOA, a minor constituent, was well-removed (greater than 98.1 percent), but PFOS was only removed to 68.6 percent and the total PFAS removal was 78.2 percent. Capture on VGAC was less than 1 percent of the total mass for either test, with similar masses of PFAS captured in the condensate and VGAC. C-18 cartridge samples suggest most of the PFAS was not present in the vapor stream before the VGAC in either test, with much smaller than expected amounts detected and only the most volatile compound (perfluorobutanoic acid) detected. The mass balance only accounted for two percent of the PFAS in the Low-Concentration Test and 36 percent of the High-Concentration Test. However, the vast majority of the PFAS captured in the High-Concentration Test was the 22 percent retained in the test soil. Based on the layout of the system and where PFAS was and was not observed, we believe the PFAS traveled to the condenser and was retained on the condenser surfaces. This is not yet confirmed by sampling.
Recirculating the vapors appeared to work well. No technical problems were reported. Tracking the moisture in the soil indicated the research team was able to account for 59 to 83 percent of the soil moisture, suggesting there were not large losses from the system due to fugitive emissions. The loss of PFAS mass is likely due to cooling the vapor stream.
The test had a number of positive outcomes. First, the Low-Concentration Test demonstrated that thermal desorption at a temperature above 350°C is a viable technology for removing PFAS from soil. The removal process is temperature dependent and does not work if the temperature is too low. The test also indicated that VGAC is not suited for treating the air stream from this process. To use VGAC, the vapors must be cooled and dried, and the PFAS appears to drop out the vapor phase in this process. This means that a treatment process for the PFAS needs to maintain the temperature of the gasses throughout the process to keep the PFAS mobile. However, under the test conditions, a wet scrubber process is recommended for PFAS removal. (Project Completion - 2020)