UNDERSTANDING FATE AND TRANSPORT OF PFAS

Transcription

1 UNDERSTANDING FATE AND TRANSPORT OF PFAS Erika Houtz, PhD May 8-1, 217 NEWMOA PFAS Workshops Presentation Overview Major environmental sources of PFAS Atmospheric Fate and Transport Subsurface Fate and Transport Idealized Conceptual Site Model of a Fire Training Area Case studies Arcadis May

2 Major Sources of PFAS Contamination *Most cannot be detected by laboratory analysis Arcadis May PFAS Atmospheric Fate & Transport Shorter transport potential ~3 5 days atmospheric lifetime (PM 2.5 ) Longer transport potential ~2 days atmospheric lifetime (8:2 FtOH) Dry Deposition of Particle Associated PFAS Wet Deposition of Particle Associated PFAS Particle phase/ Aerosol transport PFAAs and Precursors Vapor Phase Transport Mainly neutral precursors Atmospheric transformation of precursors to other PFAS by reaction with: NO x, OH, O 3, O 2 Wet Deposition of Vapor Phase PFAS PFOA associated with small particles (<.14 m) PFOS associated with larger particles (1.38 to 3.81 m) A. Dreyer et al. Chemosphere 215 Arcadis May

3 Air Deposition of PFAS to Remote Lakes - Grenoble, France Air deposition is sole source of PFAS to mountainous lakes outside of Grenoble Concentrations of PFOS and PFOS precursors in fish were similar between reference lake and lakes near Grenoble Concentrations of PFCAs and PFCA precursors in fish were dependent on proximity to local industrial sources Arcadis 216 Ahrens et al., Environ. Chem May Perfluorinated compounds (PFCs) Perfluorinated Compounds (PFCs) generally are the Perfluoroalkyl acids (PFAAs) PFAAs include: Perfluoralkyl carboxylates (PFCAs) e.g. PFOA Perfluoroalkyl sulfonates (PFSAs) e.g. PFOS Perfluoroalkyl phosphinic and phosphonic (minor) There are many PFAAs with differing chain lengths, PFOS and PFOA have 8 carbons (C8) octanoates Anionic at environmental ph Recalcitrant to biotransformation C1 Methane C2 Ethane C3 Propane C4 Butane C5 Pentane C6 Hexane C7 Heptane C8 Octane C9 Nonane C1 Decane C11 Unodecane C12 Dodecane C13 Tridecane C14 Tetradecane Arcadis May

4 PFAA Groundwater Fate & Transport: Chemical Properties and Implications PFAA plumes are generally longer High solubility Low log K OC Recalcitrant Mostly anionic PFAA sorption increases with perfluorinated chain length, e.g. PFOS (C8) is more sorptive than PFBS (C4) Chemical Properties PCB (Arochlor 126) PFOA PFOS TCE Benzene Molecular Weight Solubility (@2-25 C), mg/l Vapor Pressure (@25 C), mmhg 4.1x x Henry s Constant, atm-m 3 /mol 4.6x x x Log Koc Arcadis 216 Persistence and mobility can lead to large plumes 4 May Polyfluorinated Compounds - PFAA Precursors More diverse transport Volatile Relatively mobile in groundwater - Likely to sorb strongly to negatively- charged soils 6:2 fluorotelomer sulfonate Multiple charges behavior more difficult to predict Arcadis 216 4

5 Subsurface Retardation of PFAS in Groundwater Hydrophobic interaction Predominant sorption mechanism for long chain PFAS ~.5 log Koc increase for each CF 2 group (Higgins & Luthy 26, ES&T) Organic rich soils retard movement of PFAS f oc increases -> K d increases Oil and other organics may also increase sorption Electrostatic effects Positively charged PFAS (i.e., some precursors) sorb to negatively charged minerals Negatively charged PFAS sorb to positively charged minerals Electrostatic repulsion can decrease PFAS sorption High ionic strength dulls electrostatic repulsion and attraction Arcadis May Abiotic Transformation of PFAA Precursors Both Sulfonamido and Fluorotelomer Precursors predominately oxidize to form PFCAs 8:2 FTOH C 8 F 17 OH Atmospheric Oxidation, Aqueous Indirect Photolysis PFOA C 7 F 15 O O + shorter PFCAs Ethyl- FOSA C 8 F 17 O S O N H PFOA C 7 F 15 O O Arcadis 216 5

6 Aerobic Biotransformation of Fluorotelomer Precursors Forms PFCAs Example of Soil Microcosms with Ansul AFFF (Harding-Marjanovic et al. ES&T 215) [6:2 FtTAoS] M AFFF AFFF DGBE Day Live Similar results with fluorotelomer compounds seen in: Dinglasan et al. 24, Wang et al. 25, Lee et al. 21, Liu et al 21, Dasu et al. 212, Zhang et al. 213 [FtCA] M [PFCA] µm :3 FtCA Live 6:2 FtUCA Live 5:3 FtCA Autoclaved 6:2 FtUCA Autoclaved Day PFBA Live PFPeA Live PFHxA Live PFBA Autoclaved PFPeA Autoclaved PFHxA Autoclaved Day 6:2 FtUCA 5:3 FtCA PFHxA PFPeA PFBA Arcadis 216 Slow transformation of Sulfonamido Precursors to PFOS Aerobic Soil Microcosms from Mejia-Avendano et al. ES&T 216 Arcadis May

7 Biotransformation of Fluorotelomer Precursor under Sulfate- Reducing Conditions: No PFCAs observed Shan Yi American Chemical Society Philadelphia % 3% 6:2 FtTAoS 6:2 FtTPA 36% 3% Related study - Zhang et al. 216, Chemosphere: no transformation of 6:2 fluorotelomer sulfonate observed in anaerobic sediment, transformation of 6:2 FtOH did not yield significant PFCA products Arcadis May Precursors form dead end PFAAs such as PFOS and PFOA through aerobic biotransformation Arcadis 216 PFAAs persist indefinitely 4 May

8 n C8F17 C8F17 C8F17 Class B Foam F FC F C 8F17 S N H N + H3C H CF CH S C1H9 CH Source Zone - Hidden Cationic and Zwitterionic Dark Matter Cationic and zwitterionic PFAS are bound via ion exchange to negatively charged soils (e.g. silts & clays) in the source zone. Precursor biotransformation is slow under anaerobic source conditions. Short hydrocarbon plume Hydrocarbon LNAPL Direction of groundwater flow Anionic precursor biotransformation increases as aerobic conditions develop CH CH H N NH + F S F C F n CH C 6F13 S N C 4H9 H 3C F N + FC H F F CH 3 CH 3 CH 3 Hidden anionic mobile PFAA precursors - Dark Matter C8 C7 C6 C4 C5 C3? C2? H F N S F C F n CH N C 5F11 C4H9 H 3C S C6F17 C 4H9 H 3C CH C F H 3C S C 4H9 CH 3 H 3C S C8F17 C 4H9 C S 8F17 CF CH 3 C 6F13 S S C S 8F17 C 6F13 S S S Anionic PFAA dead end daughters C8 C7 C6 C4 C5 C3? C2? C S C6F13 C8F17 6F13 S S S Increasing mobility of shorter perfluoroalkyl chain PFAS -3mV -2mV REDOX ZONATION -1mV mv 1mV 2mV 4 May Expanding Analytical Tool Box for PFAA Precursors Total oxidizable precursor (TOP) Assay Initial LC-MS/MS analysis with re-analysis following oxidative digest Detection limits to ~ 2 ng/l (ppt) Commercially available in UK, Australia, under development in US & Canada Particle-induced gamma emission (PIGE) Spectroscopy Isolates organofluorine compounds on solid phase extraction, measures total fluorine Detection limits to ~ 15 ug/l (ppb) F Commercially available in US Adsorbable organofluorine (AOF) Isolates organofluorine compounds with activated carbon and measures F by combustion ion chromatography Detection limits to ~ 1 ug/l (ppb) F Commercially available in Germany, Australia LC-QTOF (Quantitative Time of Flight) Tentative identification of PFAS through exact mass measurement Commercially available in U.S. Arcadis May

9 TOP Assay Applied to AFFF Formulations: Many formulations appear PFAS-free until precursors are revealed by TOP Assay Concentration, ng/ml AFFF Formulations, Pre-TOPA ECF, 21 FT 1, 25 FT 2, 22 FT 3, 29 FT 4, 23 PFOS PFHpS PFHxS PFBS PFNA PFOA PFHpA PFHxA PFPA PFBA Concentration, g/l AFFF Formulations, Post-TOPA ECF, 21 FT 1, 25 FT 2, 22 FT 3, 29 FT 4, 23 PFOS PFHpS PFHxS PFBS PFNA PFOA PFHpA PFHxA PFPA PFBA FT = Fluorotelomer-based AFFF ECF = Electrochemical Fluorination- based AFFF Arcadis 216 TOP Assay Applied to Groundwater Concentration, micrograms/l Sample 1 Key use of TOP Assay and Expanded Analyte List both indicate significant PFAS mass that conventional analysis does not capture Sample 1 Pre-TOPA Sample 1 Post-TOPA Sample 2 Key use of TOP Assay indicates significant PFAS mass that conventional analysis does not capture Sample 2 Pre-TOPA Sample 2 Post-TOPA Sample 3: TOP Assay provides additional information on PFAS contamination scale Sample 3 Pre-TOPA Sample 3 Post-TOPA Sample 4: TOP Assay Confirms no Additional PFAS Sample 4 Pre-TOPA Sample 4 Post-TOPA Sample 5: Key use of TOP Assay confirms no PFAS present Sample 5 Pre-TOPA Sample 5 Post-TOPA PFNA PFOA PFHpA PFHxA PFPA PFBA 6:2 FtS PFOS PFHxS PFBS Arcadis May

10 Case Study 1: PFAS-Impacted Industrial Sludges used as Agricultural Fertilizer, Southwest Germany ~199 s compost blended with industrial sludge used as agricultural fertilizer in SW Germany near Baden Baden Sludge contained Polyfluoroalkyl Phosphates (PAPs) and fluorinated polymers Additional AFFF source from fire event Largest PFAS Site in Germany (3.7 Km 2 ); 3 Million m 3 of affected soil. Underlying alluvial sandy aquifer used for drinking water Arcadis May PFAS Fingerprint: Average Concentrations C 8 > 1 individual samples of soil and groundwater Soil fingerprint dominated by longer chain compounds; groundwater shows predominance of shorter chain compounds & PFOA possibly reflective of a mixture of sources Lab scale K D determinations (compounds with arrows) to evaluate adsorption Possible Influencing Factors: f oc, carbon chain length, functional groups, anion exchange capacity, ph, ionic strength, PFAS concentration Arcadis May

11 Adsorption strongly correlated with f oc K D did not show strong relationship to Anion Exchange Capacity ph Grain size Clay content K D did show strong relationship to Total organic carbon (TOC) PFAS chain length Arcadis May Precursor Analysis via TOP Assay: Most precursors found on soils TOP Assay on soils: ~3-7% increase in PFAAs C4 to C9 increases observed TOP Assay on groundwater: ~5-1% increase in PFAAs Minimal evidence of precursors Soil 1 Soil 1 Soil 3 Soil 3 Pre TOP Post TOP Pre TOP Post TOP Soil 2 Soil 2 Soil 4 Soil 4 Pre TOP Post TOP Pre TOP Post TOP GW 1 GW 1 GW 3 GW 3 Pre TOP Post TOP Pre TOP Post TOP GW 2 GW 2 GW 4 GW 4 Pre TOP Post TOP Pre TOP Post TOP Arcadis May

12 PFAAs sorbed better to anionic exchange resins (AIX) PFAA Precursors sorbed better to GAC t1 ~ 2 weeks t2 ~ 4 weeks Faster breakthrough of PFAAs with GAC than AIX Inflow Effluent Post GAC Effluent Post Reactivated GAC Effluent Post AIX Total organofluorine (i.e. PFAAs + PFAA precursors) show that total PFAS has faster breakthrough with AIX Arcadis 216 Inflow Effluent Post GAC Effluent Post Reactivated GAC Effluent Post AIX 4 May Case study 2: Fate of AFFF in a Wastewater Treatment Plant during Annual AFFF Testing ~ 15 days of testing performance of AFFF equipment and AFFF specs AFFF waters conveyed via wash racks to industrial treatment system Project undertaken by California DTSC Environmental Chemistry Lab Concentration µg/l 11/19 11/23 12/3 12/15 11/2 11/25 11/3 12/2 12/6 12/9 12/14 11/19 First AFFF Addition /25 End of 1 st Round of AFFF 12/1 More AFFF 12/3 End of AFFF 11/19/215 11/24/215 11/29/215 12/4/215 12/9/215 12/14/215 Sample collected Analyzed by LC MS/MS for PFAS Suite Analyzed by TOP Assay Analyzed by LC QTOF MS 6:2 FtS PFOS Arcadis May

13 Schematic of SFO Industrial Treatment Plant Discharge to SF Bay Equalization Tank Wash water >5%, dry weather Effluent Sample Disinfection/Dechlor Clarifiers Rapid Mix Basin 1 Influent Sample AFFF Stormwater, Detention basins <5%, dry weather Flocculation Tank ~.5.6 MGD <3 hrs residence time Trickling Filter Rapid Mix Basin 2 Dissolved Air Flotation Tanks Arcadis 216 Midpoint Sample 4 May Directly Measured Analytes vs. Post-TOP Assay Total PFAS Mass Concentration, µg/l Composite Influent Composite Midpoint Composite Effluent Concentration, µg/l ~95 98% of PFAS mass is not directly measured by target analyte list FtS 62 FtS 82 FtS PFBA PFPeA PFHxA PFHpA PFOA PFNA PFDA PFBS PFHxS PFOS PFHxPA Sum PFAS Pre TOP Sum PFAS Post TOP Arcadis May

14 Influent to Effluent: >99% loss of 6:2 FTSAoS, ~5% loss of Total PFAS 2.4E+8 2.E+8 Substantial disappearance after trickling filter sorption or transformation? 7 6 QTOF Inst. Response 1.6E+8 1.2E+8 8.E+7 4.E+7 Minimal effect of treatment processes Influent 12/3/15 Midpoint 12/3/15 Effluent 12/3/15 Concentration, µg/l E+ FTAB FTSAS Sum PFAS Post TOP Arcadis May Biotransformation Pathway in WWTP Resembles Aerobic Soil Microcosm Pathway Biotransformation of Ansul 6:2 FtTAoS in Aerobic Soil Microcosms (Harding-Marjanovic et al. 215) This WWTP PFHpA Arcadis May

15 Plant Clearance of PFAS Arcadis May Case Study 3: Ellsworth AFB Fire Training Area Active FTA from 1942 to 199 Site received treatment for VOCs, SVOCs until 211: Soil vapor extraction Pump and treat (watch discharged back to onsite pond): oil water separation, air sparging, GAC Dual phase extraction trench Several wells received bioventing and oxygen infusion Surface soil, aquifer solids, and groundwater collected in 211 for PFAS analysis Data summarized in two studies: Houtz et al. ES&T 213, McGuire et al. ES&T 214 Arcadis May

16 Case Study: Ellsworth AFB Fire Training Area Surface Soil PFAS profile resembled original benzene plume 1+ years after active remediation PFHxA PFOA PFOS Arcadis 216 McGuire et al. 214, ES&T 4 May Case Study: Ellsworth AFB Fire Training Area PFHxA in GW PFOA in GW PFOS in GW Total Precursors (TOP Assay) in GW Arcadis 216 Why the spatial disparity between PFCAs and PFOS/Precursors? McGuire et al. 214, ES&T 4 May

17 Case Study: Ellsworth AFB Fire Training Area PFHxA in GW Historical benzene plume received significant treatment PFHxA signal (and other PFCAs) likely the result of precursor transformation In Eastern source area, no treatment Precursor signal is more dominant, suggesting significant transformation to terminal products has not occurred Total Precursors (TOP Assay) in GW McGuire et al. 214, ES&T Arcadis May Case Study: Ellsworth AFB Fire Training Area Ratio of PFHxS to PFOS measured in some AFFF formulations is 1:1 (Houtz et al. 213) Significant presence of PFHxS precursors documented in some AFFF formulations (Houtz et al. 213) Ratio of PFHxS to PFOS is up to 5:1 in the historical benzene plume Significant formation of PFHxS from precursors is apparent Ratio of PFHxS to PFOS in GW McGuire et al. 214, ES&T Arcadis May

18 Arcadis PFAS Management Remediation is not always necessary: Site A: Site-specific characterization of PFOS aquifer retardation factors justified an MNA management approach to protect a drinking water supply well Site B: Site-specific modelling of soil leaching, groundwater flow, and reservoir hydraulics determined acceptable residual PFOS concentrations in impacted soils Arcadis is developing new technologies where remediation or treatment of PFAS may be required: Water treatment reactor systems which destroy PFAS via sonolysis Regenerable, cost effective sorbent materials which remove both short and long chain PFAS Stabilization solutions using multiple reagents for source areas Oxidative and reductive technologies showing promise for PFAS treatment Arcadis 216 May 4, Conclusions Despite broad background PFAS contamination, there are a finite number of PFAS point sources Transport depends on chemical structure: Precursors have many different kinds of functional groups PFAAs are generally non-volatile and mobile in groundwater PFAAs are non-reactive PFAA precursors biotransform more rapidly under aerobic than anaerobic conditions Similar transformation pathways seen in lab studies and full-scale wastewater treatment plants Arcadis May

19 Conclusions Atmospheric deposition of PFAS can occur tens of miles away from the release location Long chain PFAA retardation in subsurface is dominated by hydrophobic sorption Electrostatic effects may be more significant for cationic precursors and short chain PFAAs Multiple sources and local hydrogeology contribute to PFAS distribution at specific sites Subsurface fate and transport concepts have implications on performance of GAC, AIX, and other sorptive treatment technologies Treatment solutions are available and constantly being innovated Arcadis May Discover the innovations that will change the remediation industry. Download your free copy of our new Advances in Remediation book today! Go to: arcad.is/air216 Arcadis Authored Guidance on PFAS Available at: ublications/concawereports Remediation Engineering: Design Concepts Purchase your copy of this fully updated 2 nd edition of the industry s premier textbook that covers current remediation technologies as well as future trends. Receive a 2% discount by using AQQ8 when you order online at (Regular price: $169.95) Arcadis May