The PFAS-KB Project: Washing PFAS Impacted Soil using Ethanol

Pedro Balbachevsky, Principal Environmental Engineer at EI Australia, Sydney NSW

If you are reading an ALGA newsletter article, you probably already know that PFAS are toxic, ubiquitous and hard to remediate. Particularly, when it comes to soil remediation, it is possible to separate the available treatment technologies into three main categories, each presenting its own different challenges:

Table 1. PFAS Soil Remediation Technologies.

But Why Ethanol?

The reasons ethanol was selected as the most appropriate solvent for the PFAS-KB Project are:

• Traditional soil washing (using water) generates large volumes of liquid waste, potentially rendering the process impractical or economically non-viable. By replacing water with a recoverable solvent (ethanol), these liquid waste volumes are greatly reduced.
• Ethanol is very soluble in water and can be easily recovered by distillation, to be reused many times over in the treatment process (Figure 1).
• Ethanol is environmentally friendly, biodegradable, and inexpensive. Its n-octanol/water partition coefficient is very low (log Kow = -0.31) indicating no bioaccumulative effects.
• Ethanol has already been proven efficient in in situ soil flushing (ISSF) case studies (Senevirathna et al., 2021), with PFOS removal rates as high as 98%. ISSF involves the irrigation of impacted areas using a 50% ethanol solution, followed by the recovery of the produced leachate volumes downgradient of the targeted area, via groundwater extraction wells. The solvent can be regenerated during the cycle and reused in the process.
• Ethanol solutions have been successfully used to regenerate saturated adsorption media used in PFAS water treatment processes (SLEPE, 2021). The process is conducted under controlled temperatures (35-45°C), lasts between 1 and 3 hours, and uses a solution comprising 95% ethanol and 3% sodium chloride.
• Some PFAS (PFOS, in particular) are highly soluble in ethanol (Graphs 1a and 1b), which facilitates partitioning control during the soil washing process. PFOS is even more soluble in methanol; however, the use of methanol presents its own human toxicity challenges.

Figure 1. The PFAS-KB treatment process.

(Meng et al., 2017)

(Senevirathna et al, 2021)

Graphs 1a and 1b. PFOS solubility in various alcohols.

The Process

As depicted in Figure 1, the PFAS-KB process comprises the following steps:

1. Screening: The impacted soil entering the process needs to be initially screened to remove larger rocks and soil clusters that would reduce process efficiency. If applicable, these larger elements could be crushed and re-introduced into the process.
2. Homogenisation: After screening, the soil must be homogenised so that the solvent contact surface is increased during mixing (Step 3). This homogenisation step would also further break down soil clusters that made it through the sieving process (Step 1) into even smaller elements.
3. Mixing: During this process step, the homogenised soil is mixed with ethanol volumes that will vary according to the batch-specific sand/clay ratio. The higher the sand/clay ratio, the more ethanol can be added to the soil without compromising the dewatering step (Step 5).
4. Soaking: After mixing, the soil-ethanol mass is left to soak for a period of 30 minutes so that the partitioning changes have time to promote the soil-to-solvent mass transfer. The soaking period should be longer if ethanol solutions at lower concentrations are used. The optimum soaking times are still to be determined based on pilot trials underway at the time of writing this article.
5. Dewatering: After the soaking period, the soil-ethanol mixture can be dewatered, producing two separate streams: the treated soils and the extracted liquid phase. The treated soils can be reintroduced into the process (Step 3) if PFAS concentrations are found to be still unacceptable. The extracted liquid phase is pumped through particle filters and then reaches the boiler, where distillation (Step 6) is completed.
6. Distillation: During this step, the ethanol used in the process is recovered and transferred back to the appropriate storage tank. The distillation of ethanol starts when the temperature in the boiler stabilises at 78ºC. As soon as the temperature starts rising above that level, the boiler is automatically switched off. Whatever is left in the boiler comprises the ‘brine’, which is a very concentrated PFAS aqueous solution. The volume of brine generated during each batch accumulates until the high-level sensor in the boiler is activated. The contents of the boiler are then transferred to the brine tank, completing the soil washing process. The brine, however, can be later:
   • Disposed offsite, at a licensed liquid waste facility; or
   • Destroyed, using electrochemical oxidation technologies; or
   • Enriched, using foam fractionation technologies and then disposed offsite; or
   • Enriched, using foam fractionation technologies and then destroyed.

Potential Applications

It is envisioned that the PFAS-KB System will serve two main final users:

• Land developers: seeking a sustainable and cost-beneficial fate for their impacted soils. Particularly, if the impacted soils are already stockpiled, the PFAS-KB System would treat and greatly reduce waste management costs when compared to the alternatives described in Table 1.
• Landfills and waste management facilities: receivers of PFAS-impacted waste. Disposal costs for PFAS-impacted wastes are currently high because costly control measures need to be implemented to ensure a safe fate for these materials. If, instead, the wastes could be treated upon their receipt and managed under less strict classifications, more competitive fees could be practiced.

Bench Tests

Bench tests were completed using PFAS-impacted sediments collected from a contaminated pond located in NSW, as follows:

• Ethanol solutions at different concentrations were mixed with the collected impacted soils using a small tumbler, over a 30-minute period.
• Both vacuum and gravity filtering were used to dewater the soil-solvent mixture.
• The ethanol used was distilled out of the filtered liquid phase.
• Samples collected from the various process stages and batches enabled mass balances to be completed, showing a final PFOS removal as high as 99.6%.
• The positive outcomes of the bench tests allowed the project to advance to its next stage: the construction of the Pilot Trial Unit.

The Pilot Trial Unit

The Pilot Trial Unit was designed and built in-house, with the purpose of demonstrating the efficiency of the process under site-specific conditions. It follows the process depicted in Figure 1, except that a screw press (Figure 2) is responsible for mixing, soaking, and dewatering the impacted soil entering the process. The unit is fully containerised (Figures 3 and 4), allowing for quick field deployment. A control panel (Figure 5), also developed in-house, enables:

• The automatic triggering of the brine and solvent pumps;
• The manual control of the screw press;
• The visualisation of the status of all sensors, motors, and alarms;
• The monitoring of:
  • Temperature in the boiler;
  • Volume of solvent used;
  • Batch elapsed time.

Figure 2. The Pilot Trial Unit.

Figure 3. The Pilot Trial Unit.

Figure 4. The Pilot Trial Unit (rendered drawing – final lay-out).

Figure 5. The Control Panel.

Next Steps

Soil washing is really about controlling partitioning, i.e., the equilibrium ratio between solute concentrations in soil and in water phase, under saturated (or partially saturated) conditions. It involves complex mass transfer mechanisms, especially when PFAS are involved, as these are a mixture of many compounds with different water affinity behaviours.

Many parameters can affect partitioning, however the pilot trials will focus on optimising the following:

• Permeability, which in the case of the process depicted in Figure 1 can be translated to:
  • Clay/Sand ratio; and
  • Homogeneity.
• Solvent concentration, which will impact:
  • The amount of brine generated at the end of the process; and
  • The volumes of solvent lost via volatilisation along the process.
• Electro-conductivity, which can be controlled by the addition of ions to the ethanol solution; and
• pH, which can be controlled by the addition of buffering solutions to the ethanol solution.

References

• Bolan et al. (2021) Remediation of poly- and perfluoroalkyl substances (PFAS) contaminated soils — To mobilize or to immobilize or to degrade? J Hazard Mater. 2021 January 05; 401: 123892. doi:10.1016/j.jhazmat.2020.123892.
• Chen et al. (2012) PFOS and PFOA in influents, effluents, and biosolids of Chinese wastewater treatment plants and effluent receiving marine environments. Environ. Pollut. 170, 26–31. 10.1016/j.envpol.2012.06.016. [PubMed: 22763327]
• Darlington et al. (2018) The challenges of PFAS remediation. The Military Engineer 110, 58–60. 10.1007/978-1-4419-1157-5_1. [PubMed: 29780177]
• ITRC (2009) Evaluating LNAPL Remedial Technologies for Achieving Project Goals. Interstate Technology Regulatory Council, LNAPLs Team, Washington, D.C.
• Meng at al. (2017) Effect of hydro-oleophobic perfluorocarbon chain on interfacial behavior and mechanism of perfluorooctane sulfonate in oil-water mixture. Scientific Reports | 7:44694 | DOI: 10.1038/srep44694.
• Ross et al. (2018) A review of emerging technologies for remediation of PFASs. Remediat. J. 28, 101–126. 10.1002/rem.21553.
• Schröder (2003) Determination of fluorinated surfactants and their metabolites in sewage sludge samples by liquid chromatography with mass spectrometry and tandem mass spectrometry after pressurised liquid extraction and separation on fluorine-modified reversed-phase sorbents. J. Chromatogr. A 1020, 131–151. 10.1016/S0021-9673(03)00936-1. [PubMed: 14661764]
• Senevirathna et al. (2021) In situ soil flushing to remediate confined soil contaminated with PFOS- an innovative solution for emerging environmental issue, Chemosphere 262.
• Zhang et al. (2019) Adsorption of perfluoroalkyl and polyfluoroalkyl substances (PFASs) from aqueous solution-A review. Sci. Total Environ. 694, 133606 10.1016/j.scitotenv.2019.133606. [PubMed: 31401505]