Smoldering combustion is a flameless, self-sustaining process that occurs on the surface of a condensed (i.e., solid or liquid‐phase) fuel, converting organic material into primarily heat, carbon dioxide, and water. Smoldering has been well documented for solid porous materials and burning charcoal in a barbeque is a familiar example. Smoldering of an organic liquid (i.e., non-aqueous phase liquid [NAPL]) embedded within an inert porous matrix is also possible and the reaction continues in a self‐sustaining manner (i.e., continue in the absence of external energy input following a onetime, local ignition) and would destroy the NAPL if an oxidant (e.g., oxygen in air) and fuel (NAPL) were in sufficient quantity. The self‐sustaining nature makes smoldering very energy efficient and therefore cost effective as energy is only added at the start – unlike incineration approaches that require continual energy input.

The temperatures obtained through smoldering can be significant, and exceed temperatures needed to destroy per- and polyfluoroalkyl substance (PFAS). However, unlike solid or liquid fuels, PFAS are not contaminants that can support smoldering combustion in of themselves. Therefore, the objective of this study was to evaluate the use of a surrogate fuel that can support the smoldering process that achieves temperatures (greater than 900°C) sufficient to destroy PFAS. This study was conducted in two phases. 

Phase 1 results are available in the Phase 1 Final Report.

Summary of Phase I Efforts

Technical Approach

The first phase (Phase I) evaluated whether granular activated carbon (GAC) and ground rubber could smolder at required temperatures (greater than 900°C) when mixed with sand. Successful smoldering was assessed through successful ignition, the consumption of oxygen and production of combustion gasses (CO2 and CO), measuring the average peak temperature and smoldering front velocity, and whether self-sustaining smoldering was observed through either steady or increasing peak temperatures after the heater was turned off.

The second phase (Phase II) examined the treatment of: (1) a PFAS-impacted liquid by absorbing the PFAS to a solid surrogate fuel (e.g., GAC) and subsequently used as the fuel surrogate; and (2) a simulated PFAS‐impacted waste soil or drilling waste amended with a surrogate fuel; the degree of PFAS destruction was assessed by measuring: (1) PFAS concentration in soil before and PFAS concentration in soil/ash after treatment; (2) PFAS in emissions; (3) and hydrofluoric acid (HF) concentrations as a measure of total mineralization of PFAS.

All testing was conducted in specialty designed columns used to evaluate smoldering processes.


Phase I Results

A total of eight column tests were conducted in Phase I. GAC was found to be the best fuel surrogate and produced temperatures greater than 900°C when mixed with sand or a surrogate soil mixture between 40 and 60 g/kg soil. Crumb rubber also worked but produced undesirable residues. Generally, higher GAC concentration yielded higher average peak temperatures, and higher air flow increased the smoldering front velocity.

The first two Phase II tests (II-1 and II-2) examined treating three PFAS, perflurooctanoic acid (PFOA), perflurooctanesulfonic acid (PFOS), and perflurohexanesulfonic acid (PFHxS), absorbed to GAC. The GAC was mixed with a sand to a target ratio of 40 g GAC/kg sand. The test was completed twice and showed that the average peak temperature exceeded 900°C, with an average smoldering front velocity of 0.7 cm/min. Prior to smoldering, the calculated concentration in the GAC/sand mixture for PFOA, PFOS, and PFHxS was 590, 140 and 240 mg/kg in the first test, and 510, 120 and 220 mg/kg in the second test for each respective compound. After smoldering all compounds were not detected (ND) at a detection limit of 0.4 μg/kg.

Test II-3 and II-4 used a surrogate soil mixture with a known organic fraction on which the PFAS were absorbed. Test II-3 examined the treatment of the same three PFAS used in tests II-1 and II-2, and Test II-4 used six PFAS (PFOA, PFOS, PFHxS, perflurononanoic acid [PFNA], perflurobutanesulfonic acid [PFBS], perfluroheptanoic acid [PFHpA]). For both of these tests, a virgin GAC was used as the fuel source which was mixed into the soil mixture.

In test II-3, the PFAS concentration was measured in soil samples before and after smoldering, and HF was measured in the emissions captured from the column. The analysis indicated that all PFAS was removed from the soil (ND at a detection limit of 0.5 μg/kg), and that 82% of the available fluorine was captured as HF.

In test II-4 the PFAS concentration was measured in soil samples before and after smoldering. HF was measured in the emissions captured from the column, and sorbent tube (XAD) tubes containing GAC in series were used to capture any PFAS breakdown products via a separate collection line from the column. Unfortunately, a line leak prevented accurate capture of HF in this test. All of the PFAS added to the soil mixture was not detected in the soil after treatment. Small amounts of PFAS were extracted from the GAC from the XAD tubes; however, the total amount of PFAS emitted cannot be calculated, as we estimated that the efficiency of PFAS extraction from GAC was approximately 50%. However, given the adsorption capacity of PFAS onto GAC, and the measured concentrations, it appears that the amount of PFAS emitted during smoldering is small. The observed PFAS compounds in the emission suggests that there may be some conversion of the sulfonates to their carboxylate-versions of PFAS (PFOS > PFOA, PFHxS > perfluorohexanoic (PFHxA), PFBS > perflurobutanoic acid [PFBA]). Although PFAS will completely breakdown at high temperatures with sufficient residency time, it is possible that during that heating, there is sufficient energy to break off the sulfonate headgroup and volatilize a small fraction of PFAS. From a practical consideration, the emitted PFAS could be captured in an off-gas GAC treatment system, and the GAC subsequently used/treated by smoldering. The successful treatment of PFAS by smoldering was demonstrated. Additional work is recommended to further improve the mass balance of the PFAS destruction, as well as address impacts of heterogeneity on the smoldering process.


The results from this study combined with commercially available smoldering remediation technologies suggests two real-world applications: (1) ex situ treatment of PFAS contaminated investigation derived wastes or excavated soils or spent GAC containing PFAS as shown in Figure 1; and (2) in situ application to treat PFAS source area or GAC used to create permeable sorbent walls to treat groundwater plumes of PFAS as shown in Figure 2.

Ex situ treatment would involve placing a soil/GAC mixture on a commercially available engineered system (HottPad). The HottPad is a low-profile platform that supplies the heat and air used to initiate and sustain the smoldering combustion reaction. Soil mixing could simply be done using backhoes or other commercially available methods. The high temperatures that are achieved during smoldering also allows destruction of other co-contaminants and allows treatment of wet soils. Off-gases would be collected and treated by appropriate off the shelf methods.

Figure 1: Ex Situ Treatment of PFAS-Impacted Materials

In situ treatment requires the mixing of the GAC into the soil using traditional soil mixing technologies (e.g., 10- to 12-foot augers). The GAC could be mixed into a source zone that is located above or below the water table or could also be mixed into soils to form sorbent permeable walls to intersect a PFAS plume. Ignition points (IP) would be installed post soil mixing along with vacuum extraction points and off-gas treatment system. A downhole portable heater inserted into the IP is used to initiate the smoldering process which is maintained by the addition of air through the IP. Application of in situ smoldering is commercially available.

Figure 2: In Situ Treatment of PFAS

Regardless of the in or ex situ application, and although mixing will homogenize soils, the impacts of soil heterogeneity on the performance of smoldering combustion requires further evaluation. The completeness of PFAS combustion also requires further evaluation as it may impact off gas treatment (e.g., removal of partially decomposed PFAS, and the amount of HF produced). (Anticipated Phase II Completion - 2023)


Duchesne, A.L., J.K. Brown, D.J. Patch, D. Major, K.P. Weber, and J.I. Gerhard, 2020. Remediation of PFAS-Contaminated Soil and Granular Activated Carbon by Smoldering Combustion, Environmental Science and Technology, 54(19):12631-12640. doi.org/10.1021/acs.est.0c03058

Wang, J., M. Song, I. Abusallout, and D.Hanigan. 2023. Thermal Decomposition of Two Gaseous Perfluorocarboxylic Acids: Products and Mechanisms. Environmental Science and Technology, 57(15):6179-6187. doi.org/10.1021/acs.est.2c08210.


Harrison, B.G. 2022. Self-Sustaining Smouldering Combustion of Granular Activated Carbon Contaminated with Per- and Polyfluoroalkyl Substances – Fluorine Mass Balance and Enhanced Methods of Mineralization. Electronic Thesis and Dissertation Repository (Master’s Thesis). Western University, London, Ontario.