US Department of Defense–Funded Research on Treatment of Per- and Polyfluoroalkyl Substance–Laden Materials

Abstract

The scope of the present article is the status of the research, development, and demonstrations of treatment technologies for per- and polyfluoroalkyl substance (PFAS)-laden material that are funded by the Strategic Environmental Research and Development Program (SERDP) and the Environmental Security Technology Certification Program (ESTCP). Both SERDP and ESTCP are US Department of Defense (DoD) programs and are coordinated with the US Environmental Protection Agency and the US Department of Energy. The present article is 1 of a 3-part series of Focus articles on the status of research, development, and demonstration efforts that will assist project delivery teams, within the DoD, in their efforts to manage an expansive portfolio of aqueous film-forming foam (AFFF)-impacted sites. The present article focuses on treatment technologies, and the second article in the series covers fate and transport. A third, overview, article briefly summarizes exposure pathways; analytical and environmental sampling methods; fate and transport; characterization; bioaccumulation, ecotoxicity, and ecological risk assessment; and treatment technologies. The PFAS have been recognized as being one of the most persistent categories of anthropogenic chemicals found in the environment. In contrast to chlorinated solvents, these fluorinated compounds are largely impervious to common biological degradation processes and conventional chemical oxidation processes. They are a concern for the DoD and for municipal airports, due to the use of legacy AFFFs. They are also a concern for the community at large, due to the presence of PFAS in consumer products. Ex situ groundwater treatment has become a common alternative for managing PFAS-impacted groundwater. Although existing technologies are acceptable for ex situ treatment of PFAS-impacted groundwater (e.g., relying on adsorptive media such as granular activated carbon [GAC]), operation of these pump-and-treat systems represents a considerable and growing expense, especially as more of these systems have to be installed across the United States. Technologies such as GAC and ion exchange continually generate residuals (e.g., spent media) that require off-site treatment and/or disposal. Large quantities of investigation-derived wastes (IDW) continue to be generated during characterization of PFAS-impacted sites. More cost-effective alternatives are needed for disposal of residuals, and PFAS-laden IDW materials (e.g., drill cuttings, well development water). Given the substantial number of pump-and-treat systems that are currently in operation, sizable cost savings can be realized via development of more cost-effective treatment technologies. Advances may be realized in many different forms: improved media for groundwater treatment that require much less frequent replacement; media with improved capabilities for removing short-chain PFAS constituents; destruction technologies to allow for on-site treatment of groundwater, spent media, regenerant solutions, and/or IDW; and improved destruction technologies for off-site treatment of residuals and IDW. Development of effective in situ treatment technologies for PFAS-contaminated groundwater represents another important goal. Considerable cost savings will be realized if a portion of the existing pump-and-treat systems can be replaced with passive, in situ treatment systems. The treatment technologies we discuss are broadly classified as either ex situ or in situ (Figure 1 and Tables 1–3). For the ex situ treatment technologies, the projects are primarily subclassified under the following categories: aqueous media treatment, investigation-derived wastes/soils, and residuals. Residuals includes materials such as spent treatment media (GAC and ion exchange resin), and concentrated brines derived from regeneration of ion exchange resin. Some of the treatment technologies under development do not necessarily fit within a single category (e.g., plasma-based treatment processes can be applied to aqueous media, IDW water, and brine regenerant solutions). Also, some of the ex situ treatment technologies could potentially be applied in situ. The present article is not intended to provide comprehensive coverage of each and every PFAS treatment project being funded by SERDP and ESTCP, but rather to provide an overview, and to highlight a select list of representative treatment projects. D. Call, (North Carolina State University, Raleigh, NC, USA) Proof-of-concept (POC) in progress D. Chiang (CDM Smith, New York, NY, USA) M. Crimi (Clarkson University, Potsdam, NY, USA) POC in progress M. Fuller (Aptim, Baton Rouge, LA, USA) POC follow-on effort in progress K. Ozekin (Water Research Foundation, Denver, CO, USA) F. Barranco (EA Engineering, Science, and Technology, Hunt Valley, MD, USA) Proof of concept (POC) complete T. Boving (University of Rhode Island, South Kingston, RI, USA) POC in progress E. Cates (Clemson University, Clemson, SC, USA) POC follow-on effort in progress B. Chaplin (University of Illinois at Chicago, IL, USA) POC in progress H. Cho (The University of Texas at Arlington, TX, USA) POC in progress T. Holsen (Clarkson University, Potsdam, NY, USA) POC in progress D. Jassby (University of California, Los Angeles, CA, USA) POC in progress P. Koster van Groos (Aptim, Princeton, NJ, USA) POC in progress S.D. Pillai (Texas A&M University, College Station, TX, USA) POC follow-on effort in progress J. Quinnan [B] (Arcadis, Brighton, MI, USA) C. Sales (Drexel University, Philadelphia, PA, USA) POC follow-on effort in progress T. Strathmann (Colorado School of Mines, Golden, CO, USA) POC in progress J. Wehrmann (Paragon, Anchorage, AK, USA) H. Yu (Amriton, Norristown, PA, USA) POC complete POC follow-on effort in progress The stability of the C–F bond is believed to be due to the inherent bond strength, and the short length of the bond. The C–F bond is the strongest single bond known to organic chemistry, with a bond dissociation energy as high as 544 kJ/mol (for tetrafluoromethane). Because of the close proximity of the fluorine atoms that are bound to the carbon atoms, the outer fluorine atoms shield the underlying carbon backbone of PFAS constituents from reactive species. Owing to the extraordinary stability of the C–F bonds within PFAS, a considerable amount of energy is required to defluorinate PFAS constituents. Thus, destruction technologies for PFAS-laden materials are relatively energy intensive, whereas nondestructive treatment processes are much less energy intensive. In general, the energy efficiency of destruction technologies for PFAS-laden materials is greater for moderate or highly concentrated waste streams; however, most of the destruction technologies can be applied to either dilute or concentrated waste streams. Nondestructive processes may be coupled with destructive processes to achieve a complete treatment solution (e.g., use of ion exchange resin to treat groundwater, coupled with incineration of spent media). The development of complete and energy-efficient treatment solutions is critical. Both lifecycle costs and greenhouse gas emissions will need to be taken into consideration, to compare and rank the merits of specific processes and combinations of processes. Analysis of groundwater from AFFF fire training source areas has revealed that a large fraction of the total organic fluorine remains unmeasurable via conventional analytical methods (Schaefer et al. 2019). A groundwater sample from a DoD fire training source area was tested for the standard 24 PFAS analytes, total oxidizable precursors (TOPs), and also for total organic fluorine via combustion ion chromatography (TOF–CIC). The result from the TOP assay was added to the sum-total result from the 24 measured PFAS analytes, and then compared with the result from TOF–CIC assay. The TOF-CIC assay result was much higher, and comparison of the results indicated that approximately 65% of the organic fluorine present in the sample was not accounted for by adding the result from the TOP assay to the sum-total result from the 24 measured PFAS analytes. This finding has important implications for PFAS treatment technologies. The presence of unmeasurable PFAS constituents (i.e., “dark matter”) has an important impact on the capacity of adsorbent media, and ion exchange resins. Also, the presence of unmeasurable PFAS constituents increases the energy requirements for PFAS destruction technologies and makes it more difficult to predict when change-out of adsorbent media will be required. This is one reason why site-specific column testing and/or pilot test data are critically important for design of full-scale treatment systems, for estimating lifespan for treatment media, and for estimating electrical power costs for destruction technologies. Aqueous media can be generally divided into 2 categories for the context of treatment: relatively dilute waste streams (e.g., groundwater and drinking water), and moderately concentrated waste streams (e.g., surface water impoundments that have been used to collect run-off with relatively high concentrations of PFAS). Highly concentrated brines from regeneration processes are discussed later in the Treatment of residuals section. Nondestructive treatment processes include GAC and ion exchange, which have become the default technologies for treatment of dilute aqueous media. Reverse osmosis and membrane filtration (RO/MF) processes are less widely used, but are also included in this category. Both GAC and ion exchange are generally favored over RO/MF processes because they are less energy intensive. The RO/MF processes require relatively high pressures to force the influent through nanoporous membranes. Single-use ion exchange resins are used more frequently than regenerable ion exchange resins. If the influent PFAS levels are high enough, then selection of a regenerable ion exchange resin may be justified. When the PFAS levels exceed a threshold concentration range (i.e., in the area of 10–100 μg/L total PFAS), then regenerable ion exchange resin may be more cost effective than single-use. Below this range, single-use ion exchange resin will generally be more cost-effective (N. Hagelin [Purolite, King of Prussia, PA, USA], F. Boodoo [Purolite], and S. Woodard [ECT2, Portland, ME, USA], personal communication, 2020). An ion exchange pilot study and lifecycle comparison of costs for single-use ion exchange versus regenerable ion exchange resin is being performed under an ESTCP project initiated in 2018 (N. Hagelin, Table 1). Because activated carbon has a much surface area than it also has a much greater use of an filtration is required for activated The use of activated carbon is being under an ESTCP project initiated in Table 1). that are more effective for treatment of PFAS-impacted chemical to media are also under Table 1). are also under of a adsorbent media a adsorbent Table a adsorbent Table and a Table 1). The of these projects are to that have a capacity for PFAS, and also that are in of being to short-chain are used to capacity (i.e., of chemical of or of ion exchange column also be performed to the number of of water that can be treatment via either or ion exchange, the short-chain PFAS are the to Thus, column the number of of the short-chain PFAS be performed to compare different of A study of different of ion exchange resins is being Table 1). In a of lifecycle costs of different treatment ion exchange, and as well as activated is also Table 1). For moderately concentrated aqueous waste a such as foam or may need to be used to high influent is one of a of treatment processes that of the inherent of PFAS to the to PFAS from the aqueous This is generally less effective for short-chain PFAS which are not as to the as PFAS constituents. An is being as a under a SERDP project Table 1). processes are by A treatment (e.g., GAC or ion be required following one of the of the destructive processes require electrical power to the treatment systems. is required to the and/or and in some for The amount of power required is on the influent PFAS and also on a of water (e.g., organic total and This category includes the following treatment and Although it necessarily require biological treatment is also included in this category. the treatment technologies under development will have to be compared for energy requirements (e.g., of water The energy also into the of PFAS destruction (e.g., in PFAS and processes in of the energy efficiency for of that a is more than an more than more than and more than et al. The and processes are the however, the for a treatment processes on of electrical and media. The electrical by S. and at University NY, USA) by a of water through a with a of the water The and the of at the Because the primarily at the the amount of electrical power required for treatment is believed to be relatively to water that could in of treatment processes. of are used to the PFAS constituents to the The PFAS are to the of the In of of development, the that is being under projects T. Table is than most of the ex situ treatment being funded by A plasma-based treatment is also being and has the Table 1). treatment has been to be of PFAS, and is of dilute and relatively concentrated waste streams. different of have been have been to be an effective material for conventional of and materials are also under A currently under development Table is such that the influent through a of This and the An is also being tested Table 1). of treatment is that it will not be by the high of brine regenerant A high of is for conventional systems because the of treatment is on the of the within the if is may be as an of activated processes are a of projects. The and can be used to generate and The of this is of PFAS constituents due to with SERDP projects are currently treatment C. Table 1). A is currently being for PFAS treatment Table 1). The of and The design on the influent with the which through a of The of the is by a source treatment, the is from the water, and for processes are also being in a project Table 1). are a material with an activated carbon and a of with high capacity and degradation PFAS and regeneration on use of to of in of the with high and of also during of the The of high and for destruction of PFAS constituents. The from the of PFAS constituents to the of the was used by one project for destruction of PFAS, following concentration and Table The of fate and processes for PFAS have of some PFAS, or et al. some of PFAS constituents and can be into such as and for and at University NJ, USA) have a that may be of complete of and and 2019). of electrical the on a of and the as the and as the The that has been is of and but for a portion of The study of as a PFAS treatment is in can be PFAS degradation have to be to be to that it could be performed within a In a to a could be for treatment of PFAS-impacted surface and are to on this under a SERDP project Table 1). projects are also combinations of treatment processes in a treatment A of in situ treatment is being in situ groundwater treatment via oxidation will be with ex situ treatment via regenerable ion exchange is used for treatment of the regenerant solution Table 1). project is a and Table 1). A and is being Table and another is a treatment chemical oxidation by regenerable and via Table 1). Table 2 the of funded projects to ex situ treatment of PFAS-impacted IDW and under SERDP and The IDW may include and The present will on as well as materials high levels of (e.g., and some of the treatment processes we also be for and water oxidation processes are being for treatment of IDW and/or is for relatively wastes (e.g., whereas processes are more for high are required for complete of PFAS constituents from et al. 2019). present in the be to achieve such high A of energy is required to the and processes use a of and for destruction of PFAS the need to for destruction to at from a SERDP project that destruction for treatment can be by the (i.e., et al. 2019). The treatment technologies that have been used for treatment include incineration and The is used to the PFAS constituents from the to the a of oxidation may on The PFAS-laden that are generated during can be either or through a for on-site treatment, when are in the they are concentrated by the through The GAC then be for off-site disposal. operation of a an or the requirements of an with of for off-site is to be more because it the need to an or to requirements of an A fluorine was data from a treatment for PFAS-impacted et al. 2019). The included a coupled with a for destruction of from the oxidation of the PFAS-laden the fluorine indicated that the PFAS present in the was into The 2 with each test in in fluorine as that from to with a of The measured destruction and efficiency for the treatment was whereas the was at with a of 1 to The was at with a of 2 of into PFAS-impacted to treatment is being tested Koster van Table The presence of may allow for destruction of PFAS at a and may also of emissions due to of A combustion is also being for PFAS-impacted and IDW Table combustion the presence of organic which can be added to the the organic is high of the organic a through the and/or destruction of the PFAS constituents the need for of energy into the to the for combustion may require an emissions testing of has been under a study Table The on an to generate high concentrations of which with water to atoms, and In the be the media, with a to allow the media to under the The of the media, which may to degradation to the this of may require an emissions projects are also combinations of the treatment processes. and are being in 2 one of which was in treatment processes Table A will be used to and PFAS constituents Table exposure to are from The are of PFAS constituents. The residuals includes materials such as spent treatment media ion exchange resin), concentrated brines derived from regeneration of ion exchange and from membrane filtration processes such as of the treatment processes discussed are also to treatment of residuals. strength, and are important for residuals, which can For the or and also be PFAS constituents and GAC during treatment are also because the presence of GAC can have and on the and destruction of PFAS constituents. is a of residuals water residuals will be more to processes such as and of residuals is also an important for some of treatment processes. The of the waste in some the amount of energy required to the treatment For and water and high are exchange resin has a of approximately whereas GAC has a of to during incineration of the of the combustion in the combustion and a amount of organic are For the waste material to be the amount of required to the combustion the organic through to the Thus, GAC represents a for whereas ion exchange resin represents a because the organic from ion exchange resin can to the cost of for operation of the For the of spent GAC can most incineration the combustion is are Thus, the of the due to combustion of the from spent GAC and spent ion exchange resin will have a high The high can the of the material for incineration and because will be required to the Thus, the can the and the of the material as a for the treatment of spent destruction of PFAS can on the surface of the GAC (i.e., the PFAS constituents in the or the PFAS constituents may into the destruction can be used as a to the of destruction on the surface of the GAC versus the to which to the have that destruction of PFAS constituents on the surface of the GAC can be favored by to or the et al. and an of the PFAS constituents to into the that study indicated that a much greater of PFAS destruction for PFAS that to PFAS in the of et al. also that a of was for complete destruction of PFAS constituents. This was on data for PFAS that organic fluorine was of PFAS when the was at The that a combustion to in with an of be for of PFAS-laden in that it destruction of PFAS on the surface of the and The to which the adsorbent media has been with PFAS can also that have been highly with PFAS can be for treatment processes because of the amount of gas that will be generated during and of spent GAC can be an when the PFAS influent levels are relatively if the PFAS influent concentration the area of μg/L on the of the 24 measured PFAS GAC may the spent material for and it may have to be to an that has been during treatment of concentrated brines is because the concentration of PFAS constituents is may with the treatment The electrical on to PFAS the which has the a study (N. Hagelin, Table it was found that the brine to be by a of to from with operation of the the the University by also found that improved of short-chain PFAS constituents (e.g., could be by adding a to the brine The that and the was more to the of the than have to be into concentrated PFAS residuals to incineration to the high of the may with by in concentrations in the and the Treatment of concentrated brines is being under 2 projects Table 1). Table a of funded projects to in situ treatment of PFAS-impacted aqueous media, source and under SERDP and of for of PFAS constituents is being under 2 projects D. Table activated carbon have been for for in situ treatment of groundwater, and these materials are being tested for PFAS a to as is also being under 2 projects M. Table of PFAS constituents not via the use of of these they can be effective as a for of groundwater the of of at test will be required. Although they under ESTCP for treatment of chlorinated et al. may also be applied for in situ treatment of The are installed to the of groundwater and are to a that in water from an that is into the well then through treatment adsorbent media, being to the of the The treatment can be and replaced with media. have been performed to design for and of PFAS of media Crimi Table An well is currently being for PFAS-impacted groundwater at (Colorado CO, may also be to use destructive processes (e.g., to an electrical power in with Development of a of adsorptive for PFAS treatment will be tested under a ESTCP project Table Both regenerable and single-use ion exchange resin will be tested for this in situ PFAS treatment as of the A and will be into the design of the In situ treatment for PFAS-impacted in the will also be tested Table The will with a coupled with ex situ destruction of PFAS and destruction technologies. The will be of a single to and can of in a The was to an in situ treatment that is being tested to PFAS constituents from groundwater within treatment Table The treatment of and/or in the of the of the treatment The PFAS are to the of the within the treatment well a that a portion of the groundwater the treatment treatment, a foam at the surface of the water within the treatment the foam is and for off-site is a for and it is most to be for treatment processes will need to be to be of PFAS constituents that not and Also, in capabilities for total organic fluorinated and fluorinated will have to be taken into during of treatment in treatment and disposal of media will have important The for (e.g., waste and emissions and the need to greenhouse gas emissions will technologies of these into Also, the presence of currently unmeasurable PFAS constituents will continue to present for the design and of treatment systems. A of technologies is currently under development for treatment of PFAS-laden to be and technologies be to be cost energy and with technologies, of and have a greenhouse gas is critically important that data be generated during development and demonstration of technologies. of that allow for is for and of treatment technologies. data be for nondestructive (e.g., capacity and and destructive technologies (e.g., energy and For destructive technologies, fluorine data are for the to which complete is The nondestructive technologies such as GAC and ion exchange are to continue to be the default for treatment of groundwater, for at the are to existing nondestructive technologies with of destruction technologies. the and improved nondestructive technologies are also in to in situ treatment Development of improved treatment and combinations of nondestructive and destructive technologies will to capabilities for of PFAS-laden The in this article are of the and do not necessarily the and of the US of or the US of or not or for and are from the