£% United States
Environmental Protection
™ Agency
Office of Water
4304T
EPA-820P25001
January 2025
DRAFT SEWAGE SLUDGE RISK
ASSESSMENT FOR
PERFLUOROOCTANOIC ACID
(PFOA) CASRN 335-67-1 AND
PERFLUOROOCTANE SULFONIC
ACID (PFOS) CASRN 1763-23-1
January 2025
U.S. Environmental Protection Agency Office of Water, Office of
Science and Technology, Health and Ecological Criteria Division
Washington, D.C.
-------
PFOA/PFOS Risk Assessment
Acknowledgements
Technical Analysis Leads:
Sophie Greene, Office of Water, Office of Science and Technology, Health and Ecological
Criteria Division
Todd Phillips, Standards and Water Quality Branch, Water Division, US EPA Region 7
Tess Richman, Office of Water, Office of Science and Technology, Health and Ecological
Criteria Division
Lisa Weber, Office of Water, Office of Science and Technology, Health and Ecological Criteria
Division
David Tobias, Office of Water, Office of Science and Technology, Health and Ecological Criteria
Division
Modelers:
Ted Lilly, Anne Lutes, Rohit Warrier, and Donna Womack RTI subcontracted through GLEC
Contract No.: 68HERC20D0019
Reviewers:
Brittany Jacobs, Amanda Jarvis, James Justice, Casey Lindberg, Jacques Oliver, and Colleen
Flaherty, Office of Science and Technology, Health and Ecological Criteria Division
EPA Peer Reviewers
Michael Brooks, Marina Evich, Ronald Hermann, and John Washington. Office of Research and
Development
Norman Birchfield and Jason Mills, Office of Land and Emergency Management
Aderonke Adegbule, Marcy Card, and Dirk Young, Office of Chemical Safety and Pollution
Prevention
Inter-Agency Review
Inter-agency reviewers only evaluated the sections of this document related to plant and
livestock uptake (2.5, 2.9.3.4 & 2.9.3.5).
Sara Lupton and David Smith, Agricultural Research Service, United States Department of
Agriculture
Ian Edhlund, Barry Hooberman (retired), Lynn Post, Sara Sklenka, and Amanda Wiley, Center
for Veterinary Medicine, Food and Drug Administration
DRAFT
ii
-------
PFOA/PFOS Risk Assessment
EXECUTIVE SUMMARY
The purpose of this draft document is to characterize the potential human health and environmental
risks associated with land application, surface disposal, and incineration of sewage sludge that contains
perfluorooctanoic acid (PFOA) or perfluorooctane sulfonic acid (PFOS). The draft risk assessment focuses
on those living on or near impacted sites or those that rely primarily on their products (e.g., food crops,
animal products, drinking water); the draft risk assessment does not model risks for the general public.
This draft risk assessment will help inform whether PFOA or PFOS, based on our current understanding
of their toxicity, persistence, concentration, mobility, or potential for exposure, may be present in
sewage sludge in concentrations which may adversely affect public health or the environment (Clean
Water Act section 405(d)(2)(A), 33 U.S.C. 1345(d)(2)(A)). The United States Environmental Protection
Agency (EPA) uses the term "biosolids" to mean sewage sludge that has been treated to meet the Clean
Water Act (CWA) requirements and is intended to be applied to land as a soil amendment or fertilizer.
This draft risk assessment is not a regulation and is not EPA guidance.
All wastewater treatment plants (WWTPs) treating domestic sewage generate sewage sludge that needs
to be managed either by disposal or reuse. Based on recent data received by the EPA from certain large
publicly owned treatment works (POTWs) in the states where the EPA is the permitting authority, 3.76
million dry metric tons (DMT) of sewage sludge is generated each year.1 There are several ways in which
sewage sludge is disposed of in the U.S. In 2022, approximately 56% of sewage sludge generated by
these POTWs was land applied, 24% was landfilled, 3% was disposed of in a sewage sludge monofill, 16%
was incinerated, and 1% was disposed of using another method. Decisions about how to manage
sewage sludge are influenced by site-specific factors, including local landfill capacity, access to sewage
sludge incinerators (SSIs), demand for biosolids for use as an agricultural soil amendment, proximity to
disposal/reuse mechanisms (i.e., land suitable for application, monofills, landfills, incinerators), efforts
to reduce methane releases by diverting organics from landfills, and other economic or feasibility
considerations. In some states, POTWs primarily rely on one use or disposal method (for example,
POTWs in Rhode Island and Connecticut primarily incinerate sewage sludge; POTWs in Nebraska and
Colorado primarily rely on agricultural land application; POTWs in Louisiana and Kentucky primarily
dispose of sewage sludge in landfills). Other states have roughly equal numbers of POTWs employing
each use and disposal strategy (for example, Michigan and New Hampshire).2
PFOA and PFOS are two chemicals in a large class of synthetic chemicals called per- and polyfluoroalkyl
substances (PFAS). PFAS have been manufactured and used by a broad range of industries since the
1940s, and there are estimated to be thousands of PFAS present in the global marketplace that are used
in many consumer, commercial, and industrial products. PFOA and PFOS have been widely studied, and
they were once high production volume chemicals within the PFAS chemical class. PFAS manufacturers
voluntarily phased out domestic manufacturing of PFOS by 2002 and of PFOA by 2015, and the EPA
restricted their uses by Significant New Use Rules (SNURs) issued under section 5(a)(2) of the Toxic
Substances Control Act (TSCA), 15 U.S.C. 2604(a)(2). Although domestic manufacturing of PFOA and
1 See Biosolids Annual Reports from states where EPA is the Biosolids Program permitting authority covering 2022 submitted to
the EPA's Office of Enforcement and Compliance, https://www.epa.gov/biosolids/basic-information-about-sewage-sludge-
and-biosolids#statistics
2 See summaries of state sewage sludge use and disposal data, https://www.biosolidsdata.org/state-summaries
DRAFT
iii
-------
PFOA/PFOS Risk Assessment
PFOS have been phased out and their uses restricted, multiple activities still result in PFOA, PFOS, and
their precursors being released to WWTPs.
PFOA and PFOS were prioritized for biosolids risk assessment for several reasons. First, they are difficult
to degrade or treat in wastewater treatment plants because they are non-volatile, non-biodegradable,
and sorb to solids. Second, both PFOA and PFOS bioaccumulate in humans, plants, fish, and livestock
and are persistent in the environment. Finally, these chemicals are highly toxic to human beings; the EPA
has classified both chemicals as likely to be carcinogenic to humans, and the available human
epidemiological and animal toxicological evidence indicates that they adversely impact developmental,
cardiac, hepatic, and immune systems depending on exposure conditions.3,4,5
There are recent, well-documented examples of significantly elevated PFOA and PFOS concentrations in
U.S. sewage sludge contaminated by industrial sources to wastewater treatment plants. Statewide
surveys of sewage sludge also find that PFOA and PFOS are consistently detected at wastewater
treatment plants that do not receive wastewater from industrial users of the chemicals. This widespread
occurrence in sewage sludge is likely due to the historic or ongoing presence of PFOA, PFOS, and their
precursors in consumer, commercial, and industrial products. Following land application of sewage
sludge contaminated with PFOA or PFOS, these chemicals have been detected in soils, groundwater,
livestock, crops, surface water, and game. Limited or no data are available on environmental releases
associated with sewage sludge monofills or sewage sludge incinerators. Though data are available
regarding groundwater and leachate contamination with PFAS at landfills accepting mixed municipal
solid wastes, it is not clear the portion of this contamination that could be attributed to sewage sludge
disposal.
The goal of this risk assessment is to describe the potential human health and environmental risks
associated with the use and disposal practices regulated under CWA Section 405(d) and regulation 40
C.F.R. Part 503, Standards for the Use or Disposal of Sewage Sludge: land application, surface disposal
(e.g., disposal in sewage sludge monofills), and incineration of sewage sludge that contains PFOA or
PFOS. Not all the scenarios described in the draft risk assessment may be common practice. The draft
risk assessment does not assess human health or environmental risks associated with disposal in
municipal solid waste landfills, a common management practice for disposal of sewage sludge, because
that practice is regulated under the Resource Conservation and Recovery Act (RCRA) and the regulation
40 CFR Part 258, Criteria for Municipal Solid Waste Landfills. For the incineration scenario, the draft
assessment does not provide quantitative risk estimates given significant data gaps related to PFOA and
PFOS destruction efficiency during incineration and potential exposure to products of incomplete
combustion. The findings presented in this draft risk assessment are preliminary. The EPA expects to
publish a final risk assessment after reviewing public comments and revising the risk assessment
accordingly.
Prior to the writing of this draft refined risk assessment, the EPA performed a screening-level risk
analysis for PFOA and PFOS in sewage sludge using a high-end deterministic exposure model for a farm
3 US EPA, Office of Water Final Human Health Toxicity Assessment for PFOA (2024). 815R24006 and US EPA Office of Water
Final Human Health Toxicity Assessment for PFOS (2024). 815R24007.
4 US EPA, Guidelines for Carcinogen Risk Assessment (2005). EPA/630/P-03/001B. https://www.epaov/risk/guidelines-
carcinogen-risk-assessment.
5 US EPA, ORD staff handbook for developing IRIS assessments (2022). (EPA 600/R-22/268).
https://cfpub.epa.gov/ncea/iris_drafts/recordisplay.cfm?deid=356370
DRAFT
iv
-------
PFOA/PFOS Risk Assessment
family living on a pasture or crop farm (see Appendix E). This screening approach assumed high starting
concentrations of PFOA and PFOS in sewage sludge and high consumption rates for each exposure
pathway. The high-end screening model resulted in elevated risk levels for every human exposure
pathway (e.g., drinking water; consumption of fish, milk, beef, vegetables). Given the risk indicated in
the screening-level assessment, the EPA continued to a refined risk assessment. In this refined risk
assessment, the EPA assessed risks under median (i.e., central tendency, 50th percentile), rather than
high-end exposure conditions, to better understand the potential scope and magnitude of risks under
different sewage sludge use and disposal scenarios. To complete the central tendency deterministic
modeling steps of the refined risk assessment, the EPA (1) assessed available fate and transport models
to select the best available models for PFOA and PFOS, and (2) parameterized the models with inputs
and exposure factors to reflect median U.S. conditions and consumption behaviors.
The draft risk assessment is scoped to model risks to human populations because available data indicate
that humans are more sensitive to PFOA and PFOS exposures than aquatic or terrestrial wildlife or
livestock. For the land application scenarios, the EPA modeled potential PFOA and PFOS exposures and
estimated human health risks under three scenarios: (1) application to a farm with majority pasture-
raised dairy cows, beef cattle, or chickens (pasture farm scenario), (2) application to a farm growing
fruits or vegetables (crop farm scenario), and (3) application to reclaim damaged soils such as an
overgrazed pasture (reclamation scenario). For the surface disposal scenario, the EPA modeled potential
PFOA or PFOS exposures via groundwater to those living near a lined or unlined surface disposal site.
Due to uncertainties around PFOA and PFOS destruction when sewage sludge is incinerated, the EPA did
not quantitatively model the sewage sludge incineration scenarios for this draft risk assessment;
instead, the EPA qualitatively described potential risks to communities living near a sewage sludge
incinerator.
Based on the central tendency modeling results presented in the draft risk assessment, the EPA finds
that draft risk estimates exceed the agency's acceptable human health risk thresholds for some pasture
farm, food crop farm, and reclamation scenarios when assuming that the land-applied sewage sludge
contains 1 part per billion (ppb) of PFOA or PFOS. The EPA also finds that there may be human health
risks associated with drinking contaminated groundwater sourced near a surface disposal site when
sewage sludge containing 1 ppb of PFOA or sewage sludge containing 4 to 5 ppb of PFOS is disposed in
an unlined or clay-lined surface disposal unit.
The presence and magnitude of human health risks from sewage sludge use and disposal to those living
on or near impacted properties or primarily relying on their products is expected to vary across regions
and among properties depending on the concentration of PFOA and PFOS in sewage sludge; the number
of applications; the amount land applied; the climate, geology, and hydrology at the use or disposal site;
agronomic practices; human behavioral patterns (e.g., drinking water ingestion rates, consumption rate
of impacted products); and many other site-specific factors. Not all farms or disposal sites where sewage
sludge containing PFOA or PFOS have been used or disposed of are expected to pose a risk to human
health. For example, human health risks are expected to be lower when sewage sludge is applied to
areas with protected groundwater, sites that are distant from surface waters used for fishing or as a
drinking water source, and when applied to certain crops, such as grain, fuel, or fiber crops. However,
the EPA's modeling results from the draft risk assessment suggest that under certain scenarios and
conditions, land-applying or disposing of sewage sludge containing a detectable level (i.e., 1 ppb or
more) of PFOA or PFOS could result in human health risks exceeding the agency's acceptable thresholds
for cancer and non-cancer effects.
Modeling for land application scenarios suggests that, when the majority of the consumer's dietary
intake of a product comes from a property impacted by the land application of sewage sludge, the
DRAFT
v
-------
PFOA/PFOS Risk Assessment
highest risk pathways include (1) drinking milk from majority pasture-raised cows consuming
contaminated forage, soil, and water, (2) drinking water sourced from contaminated surface or
groundwater on or adjacent to the impacted property, (3) eating fish from a lake impacted by runoff
from the impacted property, and (4) eating beef or eggs from majority pasture-raised hens or cattle
where the pasture has received impacted sewage sludge. The risk calculations assume each of these
farm products (e.g., milk, beef, eggs) or drinking water consumed comes from the impacted property
but does not combine risks from each of these products. The EPA did not estimate risk associated with
occasionally consuming products impacted by land application of contaminated sewage sludge nor
foods that come from a variety of sources (e.g., milk from a grocery store that is sourced from many
farms and mixed together before being bottled).
Draft risk estimates are presented in the risk assessment as cancer risk levels and hazard quotients
(HQs). Cancer risk levels represent the number of expected excess lifetime cancer cases due to exposure
to the carcinogenic pollutant in a given population size (e.g., a cancer risk level of 1 in 1,000 indicates
that lifetime exposure to the carcinogenic pollutant would be expected to cause one additional case of
cancer for every one thousand people in the exposed population). Risk for noncancer effects are
expressed as HQs that represent the ratio of the potential exposure to a pollutant to the level below
which adverse noncancer effects are not expected (i.e., an HQ of less than 1 means adverse noncancer
health effects are unlikely and thus risk can be considered negligible; an HQ greater than 1 means
adverse noncancer effects are possible and thus risk is indicated).
Risk estimates for the highest risk pathways can exceed the EPA's acceptable thresholds by several
orders of magnitude. For example, for the land application scenarios, cancer risk levels associated with
drinking the modeled amount of contaminated milk (i.e., 32 ounces per day for adults) can exceed 1 in
1,000, and HQs for non-cancer effects associated with eating the modeled amount of contaminated fish
(i.e., 1 to 2 servings per week for adults) can reach up to 45. For the crop farm scenario, there are
limited scientific studies available regarding the uptake of PFOA and PFOS from sewage sludge-amended
soils into certain fruits and vegetables; however, the draft risk assessment suggests that cancer risks
from consuming the modeled amount of these contaminated foods (e.g., 1 serving per day for adults for
certain categories of fruits and vegetables) can exceed 1 in 100,000 for PFOA. Because the draft risk
assessment indicates risks associated with individual exposure pathways, there may be potential risks to
populations beyond the farm family (e.g., people living near a use or disposal site who use contaminated
groundwater as a source of drinking water or people who primarily consume produce, dairy, or meat
from a farm that has applied contaminated sewage sludge under the modeled conditions).
For the surface disposal sites, there are no exceedances of the EPA's risk thresholds for PFOA or PFOS in
down-gradient groundwater at composite-lined surface disposal sites. However, for unlined and clay-
lined surface disposal sites, there can be exceedances of the risk thresholds for the drinking water
pathway: for unlined sites, the cancer risk levels can exceed 1 in 1,000 and HQs are as high as 12; for
clay-lined sites, the cancer risk levels can exceed 1 in 1,000 and HQs are up to 9. As mentioned above,
the draft risk assessment does not include quantitative risk estimates for incineration due to data
limitations.
The draft risk calculations are not conservative estimates because they (1) model risks associated with
sludge containing 1 ppb of PFOA or PFOS, which is on the low end of measured U.S. sewage sludge
concentrations, (2) reflect median exposure conditions (e.g., 50th percentile drinking water intake rates),
(3) do not include non-sewage sludge exposures to PFOA or PFOS (e.g., consumer products, other
dietary sources), (4) do not account for the combined risk of PFOA and PFOS together, and (5) do not
account for exposures from the transformation of PFOA or PFOS precursors. As such, risk estimates that
account for multiple dietary exposures (e.g., consuming impacted milk, water, and eggs), multiple
DRAFT
vi
-------
PFOA/PFOS Risk Assessment
sources of exposure {e.g., exposure to PFOA or PFOS-containing consumer products), or exposure to
other PFAS would be greater than those presented in this draft risk assessment. Further, the EPA's draft
risk assessment relies on models where risks scale linearly with the starting concentration of PFOA or
PFOS in sewage sludge. As such, sewage sludge containing ten times more PFOA or PFOS (i.e., 10 ppb)
would yield risk estimates that are ten times greater than those presented in the draft risk assessment
(assuming all other factors are constant).
The EPA did not complete Monte Carlo probabilistic modeling because risks exceeding acceptable
thresholds were identified in multiple scenarios and pathways in the central tendency deterministic
modeling results. For example, in the EPA's draft risk assessment, when calculating risks from egg
consumption in the central tendency approach, the model assumes that an adult living on a farm
consumes, on average, 1 egg per day from the impacted property for ten years, which represents the
median egg consumption rate for households who farm.6 The model further assumes that when the
adult lives on the impacted farm, they have no sources of PFOA or PFOS exposure other than eggs and
that for the remainder of the adult's life, they have no exposure to PFOA or PFOS through any pathway.
Since risk is indicated under this central tendency scenario, Monte Carlo probabilistic modeling, which
would examine the entire distribution of potential exposures to PFOA or PFOS and report the 95th
percentile of the risk distribution, is not warranted. For this reason, the EPA is not conducting additional
modeling exercises at this time, but rather is focusing on sharing the central tendency modeling results
and identifying actions that could be taken to mitigate risks.
In summary, the results of the draft risk assessment indicate that there are potential risks to human
health to those living on or near impacted properties or primarily relying on their products from land
application and surface disposal of sewage sludge containing PFOA and PFOS and that risk is dependent
on (1) the concentration of PFOA and PFOS in sewage sludge, (2) the specific type of management
practice (e.g., type of land application or presence of a liner in a monofill), and (3) the local
environmental and geological conditions (e.g., climate and distance to groundwater). Risks are possible,
though not quantified, from the incineration of PFOA and PFOS-containing sewage sludge. Site-specific
factors should be considered when planning risk mitigation and management practices to reduce human
exposures associated with PFOA and PFOS in sewage sludge.
6 See EPA's Exposure Factors Handbook, https://www.epa.gov/expobox/about-exposure-factors-handbook, Table 13-40
DRAFT
vii
-------
PFOA/PFOS Risk Assessment
TABLE OF CONTENTS
EXECUTIVE SUMMARY Ill
1 BACKGROUND 1
1.1 Clean Water Act Section 405 Authority 1
1.2 Purpose 1
1.3 Use and Disposal of Sewage Sludge 2
1.4 History of Sewage Sludge Risk Assessment 4
2 PROBLEM FORMULATION 6
2.1 Literature Search Strategy and Information Management 6
2.2 The Nature of the Chemical Stressor 6
2.2.1 Chemical Identity 6
2.2.2 Transformation and Degradation of Precursors 9
2.2.3 Environmental Fate and Transport 9
2.3 Sources to Wastewater Treatment Plants and Biosolids 11
2.4 Occurrence in Biosolids 11
2.5 Uptake and Bioaccumulation 14
2.5.1 Animals 14
2.5.2 Plants 15
2.6 Effects on Humans and Aquatic and Terrestrial Biota 16
2.6.1 Human Health Effects 16
2.6.1.1 Oral 16
2.6.1.2 Inhalation 18
2.6.1.3 Dermal 19
2.6.2 Ecological Effects 19
2.6.2.1 Effects on Aquatic Organisms 19
2.6.2.2 Effects on Terrestrial Organisms 21
2.6.3 Scoping: Sensitive Receptors and Endpoints 22
2.7 Exposure Pathways for Humans and Aquatic and Terrestrial Biota 23
2.7.1 Considerations of Aggregate Exposures 24
2.7.2 Considerations of Cumulative Exposures 24
2.8 Conceptual Models 24
2.8.1 Farms 25
2.8.1.1 Crop Farm Scenario 25
2.8.1.2 Pasture Farm Scenario 29
2.8.2 Land Reclamation 31
2.8.3 Surface Disposal 32
2.8.4 Incineration 34
2.8.5 Other Land Application Scenarios 37
2.9 Analysis Plan 39
2.9.1 Modeling Plan 39
2.9.1.1 High End Deterministic 40
2.9.1.2 Central Tendency Deterministic 40
2.9.1.3 Probabilistic (Monte Carlo Analysis) 40
2.9.2 Model Selection 41
2.9.2.1 PFOA- and PFOS-specific Fate and Transport Considerations 41
2.9.2.2 Soil Surface Modeling 42
2.9.2.3 Surface Water Modeling 44
2.9.2.4 Groundwater Modeling 44
2.9.2.5 Air Dispersion Modeling 45
2.9.2.6 Plant and Animal Uptake Equations 45
DRAFT viii
-------
PFOA/PFOS Risk Assessment
2.9.3 Model Parameterization
2.9.3.1 Toxicity Values
2.9.3.2 Sewage Sludge PFOA, PFOS Concentration and Other Characteristics
2.9.3.3 Physical and Chemical Properties
2.9.3.4 Plant Uptake Factors
2.9.3.5 Livestock Uptake Factors
2.9.3.6 Livestock Dietary Intakes
2.9.3.7 Bioaccumulation Factors in Fish
2.9.3.8 Consumption Rates for Food and Water
2.9.3.9 Cooking and Food Preparation Loss Assumptions
2.9.3.10 Soil Ingestion Rates
2.9.3.11 Body Weight
2.9.3.12 Duration of Exposure Modeling
2.9.3.13 Location-specific Parameters
2.9.3.14 Biosolids Application Assumptions
2.9.3.15 Surface Disposal Assumptions
3 ANALYSIS
3.1 Exposure Characterization, Central Tendency Models
3.1.1 Crop Farm
3.1.2 Pasture Farm
3.1.3 Reclamation Site
3.1.4 Sewage Sludge Disposal Site
3.1.5 Implications for Home Gardening
3.1.6 Other Land Application Use Scenarios
3.1.7 Incineration
3.2 Modeled Media Concentrations over Time
3.2.1 Soil Concentrations over Time
3.2.2 Surface Water Concentrations over Time
3.2.3 Groundwater Concentrations over Time
4 RISK CHARACTERIZATION
4.1 Methods for Estimating Human Health Hazard and Cancer Risk
4.2 Crop Farm Risk Estimation
4.3 Pasture Farm Risk Estimation
4.4 Reclamation Risk Estimation
4.5 Potential Impacts beyond the Farm Family
4.6 Sewage Sludge Disposal Site Risk Estimation
4.7 Other Land Application Risk Estimation
4.8 Additional Risk Considerations for All Scenarios
4.9 Monte Carlo Analysis
5 UNCERTAINTY, VARIABILITY, AND SENSITIVITY
5.1 Variability
5.2 Uncertainty
5.2.1 Systemic Uncertainties Resulting in Underestimation of Risk
5.2.2 Systemic Uncertainties that Result in Overestimation of Risk
5.2.3 Random Uncertainties
5.3 Sensitivity of Models
6 COMPARISON OF MODELED CONCENTRATIONS AND OBSERVED CONCENTRATIONS IN RELEVANT MEDIA ...
6.1 Biosolids Investigations in Ottawa, Ontario, Canada
6.2 Biosolids Investigations in Decatur, Alabama
..47
..48
..49
..49
..51
..56
..66
..68
..70
..74
..74
..75
..75
..75
..77
..78
..79
..79
..79
..84
..88
..91
..93
..94
..95
..95
..95
..98
..99
101
101
103
105
106
108
109
110
111
112
112
112
113
113
115
115
116
116
116
119
DRAFT
ix
-------
PFOA/PFOS Risk Assessment
6.3 Biosolids Investigations in Wixom, Michigan 121
6.4 Biosolids Investigations at Various Farms in Maine 122
7 REFERENCES 122
APPENDIX A. SUMMARY OF PFOA AND PFOS OCCURRENCE IN BIOSOLIDS IN THE US A-l
APPENDIX B. MODEL INPUTS B-l
APPENDIX C. GROUNDWATER MODELING C-l
APPENDIX D. SENSITIVITY ANALYSIS D-l
APPENDIX E. SCREENING-LEVEL RESULTS FROM BST E-l
LIST OF TABLES
Table 1. Chemical and Physical Properties of PFOA 7
Table 2. Chemical and Physical Properties of PFOS 8
Table 3. Toxicity Values for PFOA 18
Table 4. Toxicity Values for PFOS 18
Table 5. Freshwater Aquatic Life AWQCs for PFOA and PFOS 21
Table 6. Toxicity Values for PFOA 49
Table 7. Toxicity Values for PFOS 49
Table 8. Koc Values for PFOA 50
Table 9. Koc Values for PFOS 50
Table 10. Fraction Organic Carbon Values by Medium 50
Table 11. Moisture Adjustment Factors by Type of Produce 52
Table 12. Plant BCFs from Yoo et al. 2011 52
Table 13. Selected Plant BCFs 56
Table 14. Selected Livestock BTFs 66
Table 15. Study Quality Criteria Used by Burkhard (2021) 70
Table 16. Fish BAFs by Trophic Level 70
Table 17. Overview of Selected Human Consumption Rates 74
Table 18. PFOA Media Concentrations for Crop Farm (ppt): Maximum 10- and 1-year Averages 79
Table 19. PFOS Media Concentrations for Crop Farm (ppt): Maximum 10- and 1-year Averages 80
Table 20. PFOA Exposures for Crop Farm (ng/kg-day): LADD and ADD 83
Table 21. PFOS Exposures for Crop Farm (ng/kg-day): LADD and ADD 83
Table 22. PFOA Media Concentrations for Pasture Farm (ppt): Maximum 10- and 1-year Averages 84
Table 23. PFOS Media Concentrations for Pasture Farm (ppt): Maximum 10- and 1-year Averages 85
Table 24. PFOA Exposures for Pasture Farm (ng/kg-day): LADD and ADD 86
Table 25. PFOS Exposures for Pasture Farm (ng/kg-day): LADD and ADD 87
Table 26. PFOA Media Concentrations for Reclamation Site (ppt): Maximum 10- and 1-year Averages 88
Table 27. PFOS Media Concentrations for Reclamation Site (ppt): Maximum 10- and 1-year Averages 88
Table 28. PFOA Exposures for Reclamation Site (ng/kg-day): LADD and ADD 90
Table 29. PFOS Exposures for Reclamation Site (ng/kg-day): LADD and ADD 90
Table 30. PFOA Groundwater Concentrations for Sludge Disposal Unit (ppt): Maximum 10- and 1-year
Averages by Liner Scenario 92
Table 31. PFOS Groundwater Concentrations for Sludge Disposal Unit (ppt): Maximum 10- and 1-year
Averages by Liner Scenario 92
Table 32. PFOA Exposures for Surface Disposal Site (ng/kg-day): LADD and ADD 93
Table 33. PFOS Exposures for Surface Disposal Site (ng/kg-day): LADD and ADD 93
Table 34. PFOA Risk Results for Crop Farm, Cancer and Non-Cancer 103
Table 35. PFOS Risk Results for Crop Farm, Cancer and Non-Cancer 103
Table 36. PFOA Risk Results for Pasture Farm, Cancer and Non-Cancer 105
Table 37. PFOS Risk Results for Pasture Farm, Cancer and Non-Cancer 105
DRAFT x
-------
PFOA/PFOS Risk Assessment
Table 38. PFOA Risk Results for Reclamation Site, Cancer and Non-Cancer 106
Table 39. PFOS Risk Results for Reclamation Site, Cancer and Non-Cancer 107
Table 40. PFOA Groundwater Risk Results for Sludge Disposal Site, Cancer and Non-Cancer 109
Table 41. PFOS Groundwater Risk Results for Sludge Disposal Site, Cancer and Non-Cancer 110
LIST OF FIGURES
Figure 1. Distribution of sewage sludge use and disposal from Biosolids Annual Reports covering 2022
submitted to the EPA's Office of Enforcement and Compliance 3
Figure 2. Conceptual visual depiction of crop farming scenario 26
Figure 3. Crop farm conceptual model 27
Figure 4. Conceptual visualization of pasture farm scenario 29
Figure 5. Pasture farm conceptual model 30
Figure 6. Conceptual model for disposal in a surface disposal site 33
Figure 7. Conceptual model for sewage sludge incineration 36
Figure 8. Conceptual model for other land application scenarios 38
Figure 9. Plot of PFOA and PFOS concentrations over time in the "moderate" climate crop farm scenario
with the low Koc setting, assuming biosolids application ceases after 40 years 97
Figure 10. Plot of PFOA and PFOS concentrations in soil over time in the "moderate" climate pasture farm
scenario with the high Koc setting, assuming biosolids application ceases after 40 years 97
Figure 11. PFOA and PFOS concentrations over time in the low Koc, pasture farm, moderate climate
setting 98
Figure 12. PFOA and PFOS concentrations over time in the high KoC, pasture farm, moderate climate
setting 99
Figure 13. PFOA and PFOS concentrations over time in the low Koc, pasture farm, moderate climate
setting 100
Figure 14. PFOA and PFOS concentrations over time in the high Koc, pasture farm, moderate climate
setting 101
DRAFT xi
-------
PFOA/PFOS Risk Assessment
LIST OF ACRONYMS
Acronym
Definition
1-D
1-dimensional
3-D
3-dimensional
3MRA
Multimedia, Multipathway, and Multireceptor Risk Assessment modeling system
ADD
average daily dose
AFFF
aqueous film forming foam
ALT
alanine transaminase
ATSDR
Agency for Toxic Substances and Disease Registry
AWI
air-water interface
AWQC
Ambient Water Quality Criteria
BAF
bioaccumulation factor
BAR
Biosolids Annual Report
BCF
bioconcentration factor
BST
Biosolids Tool
BTF
biotransfer factor
BW
body weight
CAA
Clean Air Act
CASRN
Chemical Abstracts Service registry number
CDC
Centers for Disease Control and Prevention
CDPHE
Colorado Department of Public Health & Environment
CERCLA
Comprehensive Environmental Response, Compensation, and Liability Act
CFR
Code of Federal Regulations
Class AEq
class A exceptional quality (biosolids)
COR
carry over rate
CR
consumption rate
CRL
cancer risk level
CSA
community supported agriculture
CSF
cancer slope factor
CT DEEP
Connecticut Department of Energy and Environmental Protection
CWA
Clean Water Act
diPAP
polyfluoroalkyl phosphate diesters
DMT
dry metric tons
DOC
dissolved organic carbon
DW
dry weight
ec25
25% effect concentration
ECCC
Environment and Climate Change Canada
ED
exposure duration
EFH
Exposure Factors Handbook
EFSA
European Food Safety Authority
EPA
US Environmental Protection Agency
EPACMTP
EPA's Composite Model for Leachate Migration with Transformation Products
EXAMS
Exposure Analysis Modeling System
FDA
US Food and Drug Administration
FEQG
Federal Environmental Quality Guideline (Canada)
FGD
Flue gas desulfurization
foe
fraction of organic carbon
FOSAA
perfluorooctane sulfonamidoacetic acid
FOSE
perfluorooctane sulfonamidoethanol
FR
Federal Register
FTC A
fluorotelomer carboxylic acid
FTOH
fluorotelomer alcohol
FTP
fluorotelomer-based polymer
FTS
fluorotelomer sulfonate
GIS
geographic information system
GSAF
grass soil accumulation factor
GSCM
Generic Soil Column Model
GW
groundwater
DRAFT
-------
PFOA/PFOS Risk Assessment
Acronym
Definition
HELP
Hydrologic Evaluation of Landfill Performance model
HGDB
Hydrogeologic Database
HHRAP
Human Health Risk Assessment Protocol
HLC
Henry's law constant
HQ
hazard quotient
HSDB
Hazardous Substances Data Bank
HUC
hydrologic unit code
IC25
25% inhibition concentration
IQR
interquartile range
IUR
inhalation unit risk
Kd
solid-phase adsorption coefficient
Koc
organic carbon distribution coefficient
K0w
water-octanol partitional coefficient
LADD
lifetime average daily dose
LAU
land application unit
LC
liquid chromatography
LC50
50% lethal concentration
LD50
50% lethal dose
LOD
limit of detection
LOQ
limit of quantification
LWS
Local Watershed Model
MAF
moisture adjustment factor
Maine DEP
Maine Department of Environmental Protection
MDL
method detection limit
Ml EGLE
Michigan Department of Environment, Great Lakes, and Energy
MPART
Michigan PFAS Action Response Team
MRL
maximum residue level
MS
mass spectrometry
MSW
municipal solid waste
MW
molecular weight
n.d.
non-detect
NEtFOSA
N-ethyl perfluorooctane sulfonamide
NEtFOSAA
N-ethyl perfluorooctane sulfonamidoacetic acid
NEtFOSE
N-ethyl perfluorooctane sulfonamidoethanol
NFCS
Nationwide Food Consumption Survey
NH DES
New Hampshire Department of Environmental Services
NHANES
National Health and Nutrition Examination Survey
NIST
National Institute of Standards and Technology
NLM
National Library of Medicine
NMeFOSAA
N-methylperfluorooctane sulfonamidoacetic acid
NSSS
National Sewage Sludge Survey
OC
organic carbon
OM
organic matter
OPP
Office of Pesticide Programs
PAN
plant available nitrogen
PAP
polyfluoroalkyl phosphoric acid
PBPK
physiologically based pharmacokinetic
PCBs
polychlorinated biphenyls
PEM
Particulate Emissions Model
PFAA
polyfluoralkyl acids
PFAS
per- and polyfluoroalkyl substances
PFBS
perfluorobutane sulfonic acid
PFCAs
perfluoroalkyl carboxylic acids
PFDA
Perfluorodecanoic acid
PFDoDA
Perfluorododecanoic acid
PFHpA
Perfluoroheptanoic acid
PFHxA
perfluorohexanoic acid
PFHxS
perfluorohexane sulfonic acid
DRAFT
xiii
-------
PFOA/PFOS Risk Assessment
Acronym Definition
PFHxS perfluorohexane sulfonic acid
PFI polyfluorinated iodide
PFNA Perfluorononanoic acid
PFOA perfluorooctanoic acid
PFOS perfluorooctane sulfonic acid
PFOSA perfluorooctanesulfonamide
PFSA perfluorosulfonic acid
PFUnDA Perfluoroundecanoic acid
PIC product of incomplete combustion
pKa acid dissociation constant
POM percent organic matter
POTW publicly owned treatment works
ppb parts per billion
ppt parts per trillion
RAGS Risk Assessment Guidance for Superfund
RCF root concentration factor
RCRA Resource Conservation and Recovery Act
RfC reference concentration
RfD reference dose
RSC relative source contribution
RSL regional screening level
sAmPAP perfluorooctane sulfonamidoethanol-based phosphate diester
SAMSON Solar and Meteorological Surface Observation Network
SATK saturated hydraulic conductivity
SCS Soil Conservation Service
SDU surface disposal unit
SFEI San Francisco Estuary Institute
SI surface impoundment
SNUR Significant New Use Rule
SPM suspended particulate matter
SSI sewage sludge incinerators
SSURGO Soil Survey Geographic database
STATSGO State Soil Geographic database
SW surface water
TNSSS Targeted National Sewage Sludge Survey Sampling and Analysis Technical Report
TOP total oxidizable precursors (assay)
TRI Toxics Release Inventory
TSCA Toxic Substances Control Act
TSDF treatment, storage, and disposal facility
TSS total suspended solids
USDA U.S. Department of Agriculture
USGS U.S. Geological Survey
USLE Universal Soil Loss Equation
VT DEC Vermont Department of Environmental Conservation
VVWM Variable Volume Waterbody Model
WBAN Weather Bureau-Army-Navy
WW wet weight
WWTP wastewater treatment plant
DRAFT xiv
-------
1 BACKGROUND
1.1 Clean Water Act Section 405 Authority
Section 405(d) of the Clean Water Act (CWA), 33 U.S.C. 1345(d), requires the United States
Environmental Protection Agency (EPA) to establish numerical limitations and management practices,
when appropriate, that protect public health and the environment from the reasonably anticipated
adverse effects of toxic pollutants in sewage sludge. Section 405(d) also requires the EPA to review
sewage sludge regulations at least every two years for the purpose of identifying additional pollutants
that may be present in sewage sludge and, if appropriate, to propose practices and standards for those
pollutants consistent with the requirements set forth in the CWA.
Section 405(e) of the CWA, 33 U.S.C. 1345(e), prohibits any person from disposing of sewage sludge
from a publicly owned treatment works (POTW) or other treatment works treating domestic sewage
through any use or disposal practice for which regulations have been established pursuant to Section
405 except in compliance with the Section 405 regulations at 40 CFR part 503. Section 405(g) of the
CWA, 33 U.S.C. 1345(g), authorizes the EPA to conduct public information projects and to disseminate
information pertaining to the safe use of sewage sludge.
In 1993, the EPA promulgated final regulations regarding sewage sludge, the "Standards for the Use or
Disposal of Sewage Sludge" (40 CFR Part 503). That rule contains management practices and pollutant
limits that protect public health and the environment from reasonably anticipated adverse effects often
regulated pollutants in sewage sludge when the sewage sludge is land applied, placed in a surface
disposal unit, or fired in a sewage sludge incinerator. The terms "biosolids" and "sewage sludge" are
often used interchangeably by the public; however, the EPA typically uses the term "biosolids" to mean
sewage sludge that has been treated to meet the requirements in Part 503 and is intended to be applied
to land as a soil amendment or fertilizer. The EPA's rules and the CWA only use the term "sewage
sludge."
1.2 Purpose
The goal of human health and ecological risk assessment is to estimate the nature and probability of
adverse health effects in humans or other ecological populations that may be exposed to chemicals in
contaminated environmental media, now or in the future. The risk assessment process includes 1)
planning the scope of the assessment, 2) identifying the hazards by describing how the stressor has the
potential to cause harm to humans and/or ecological systems, 3) assessing exposures to the humans
and ecological receptors and 4) characterizing the risks to those exposed human and ecological
populations. Risk assessments also include a discussion of areas of uncertainty and variability in the
assessment.
Perfluorooctanoic acid (PFOA) and perfluorosulfonic acid (PFOS) are two chemicals within the family of
fluorinated organic substances called per- and polyfluoroalkyl substances (PFAS). The purpose of this
draft risk assessment is to assess the potential human health and environmental risks associated with
land application and disposal of sewage sludge that contains PFOA or PFOS. This draft risk assessment
considers several common use and disposal scenarios for sewage sludge and the resulting exposures to
aquatic and terrestrial wildlife, in addition to impacted human populations. There are four detailed
sewage sludge modeling scenarios described in this document: reuse (land application) on a farm
growing fruits and vegetables (crop farm scenario), reuse (land application) on a farm raising livestock
(pasture farm scenario), disposal in a surface disposal site (surface disposal scenario), and reuse (land
application) to restore degraded soils (land reclamation scenario). Potentially impacted human
populations included in the modeled scenarios are farm families, those drinking water impacted by
DRAFT
1
-------
sewage sludge disposal sites or biosolids land application sites, participants in community supported
agriculture7 (CSA), those growing food in home or community gardens, and those who eat freshwater
fish.
In this document, Section 2 (Problem Formulation) describes the scope of the draft risk assessment and
assessment endpoints for PFOA and PFOS. Section 3 (Analysis) presents estimated concentrations in
relevant media for exposure, such as groundwater and soil. Section 4 (Risk Characterization) includes
risk estimation and risk description. Section 5 (Uncertainty, Variability, and Sensitivity) describes how
uncertainty may affect the draft risk assessment. Finally, Section 6 (Comparison of Modeled
Concentrations and Observed Concentrations in Relevant Media) compares modeled results from this
draft assessment to biosolids investigations in various locations.
This draft risk assessment is not a regulation and is not EPA guidance. Furthermore, the draft risk
assessment does not include a discussion of risk management options. The draft risk assessment was
externally peer reviewed through a task order with a contractor.8 A panel of five scientists reviewed the
draft risk assessment and responded to charge questions through the contractor on August 6, 2024. The
peer reviewers' comments and the EPA's responses are available in a separate document (US EPA,
2024n).
1.3 Use and Disposal of Sewage Sludge
Each year, certain large POTWs9 in the United States are required to summarize their sewage sludge
management practices and compile compliance information in Biosolids Annual Reports (BARs). The
EPA collects BARs from roughly 2,500 facilities in the 41 states where the EPA is the permitting
authority.10 These POTWs generate approximately 3.76 million dry metric tons (DMT) of sewage sludge
each year that either needs to be disposed of or reused. Disposal options include landfilling,
incineration, and other disposal methods like deep well injection. Landfilling can occur in a sewage
sludge monofill (i.e., surface disposal) which is regulated under the CWA in 40 CFR Part 503, but most
landfilling occurs at municipal solid waste (MSW) landfills, which are regulated under RCRA in 40 CFR
part 258 and will not be part of this assessment. Based on the BARs covering 2022 from facilities where
the EPA is the permitting authority, approximately 27% of all generated sewage sludge was landfilled,
16% was incinerated, and 1% was disposed of using another method. Reuse of sewage sludge is often
preferred by treatment works because it tends to be less costly, produces fewer carbon emissions,
and/or provides a benefit as a soil amendment. Reuse options include land application on agricultural
lands, at reclamation sites, or at home gardens or other sites like golf courses, often through the sale of
bulk or bagged product. As of 2022, land application at agricultural sites accounted for 31% of sewage
7 Community Supported Agriculture is an arrangement where consumers purchase a share of produce typically from one or a
small number of farmers. Commonly a variety of produce will be included in the arrangement so that purchasers receive
regular deliveries throughout the local growing season.
8 Versar under Contract No. 68HERH23A0021 Task Order 68HERH23F0320
9 BARs are required from by POTWs that 1) serve 10,000 people or more; 2) are Major POTWs (POTWs with a design flow rate
greater than or equal to one million gallons per day); 3) are Class 1 Management Facilities (POTWs with an approved
pretreatment program or facilities that have been classified as such by the EPA or state); or are otherwise required to report
by EPA or permitting authority, that land apply, surface dispose or incinerate in a sewage sludge incinerator. The EPA does
not receive data from smaller POTWs, private or federal treatment works, or those that use alternate use or disposal
practices like landfilling except on a voluntary basis.
10 There are nine states (Arizona, Idaho, Michigan, Ohio, Oklahoma, South Dakota, Texas, Utah, and Wisconsin) that are
authorized through the National Pollutant Discharge Elimination System (NPDES) Program to be the permitting authority for
biosolids. The EPA will transition to electronic reporting for the remaining authorized states as part of Phase 2
implementation of the NPDES eRule by December 2025.
DRAFT
2
-------
sludge use, land application at reclamation sites accounted for 1%, and other land application accounted
for 24% (Figure 1). Overall, about 56% of all sewage sludge generated is land applied.
Biosolids Use & Disposal from
2022 Biosolids Annual Reports
Incineration (16%)
Land Application (56%)
3 Reclamation (1%)
Agricultural (31%)
| Other (e.g., home garden,
landscaping, golf course
etc.) (24%)
Other (e.g., storage, deep-
well injection, etc.) (1%)
Landfilling (27%)
Municipal Solid Waste
Landfill (2494)
Morofill (3%)
Figure 1. Distribution of sewage sludge use and disposal from Biosolids Annual Reports covering
2022 submitted to the EPA's Office of Enforcement and Compliance.
Current regulations for the land application of sewage sludge at 40 CFR part 503 require land application
at the agronomic rate for nitrogen. 40 CFR § 503.14(d). The main exception is when the goal of land
application is reclamation of a site that has been degraded (e.g., repairing the surface of a mining site);
in such cases, sewage sludge can be applied above the agronomic rate to restore organic material and
encourage vegetative regrowth. 40 CFR § 503.14(d). Biosolids land application can also be conducted as
frequently as desired if the agronomic, pathogen, and vector attraction requirements within 40 CFR part
503 are met for the crop or farming activity (note that domestic septage, which is defined as the liquid
or solid material removed from septic tanks, cesspools, portable toilets, Type III marine sanitary devices,
or similar systems, can be similarly land applied at application rates which are based on agronomic rates
for nitrogen. 40 CFR § 503.13(c). Additionally, home gardeners who purchase or receive bulk or bagged
biosolids are not required to apply biosolids at an agronomic rate. 40 CFR §§ 503.10(b)-(g).
Surface disposal is the placement of sewage sludge onto land for final disposal in a sewage sludge unit
(e.g., sewage sludge-only landfill or "monofill"). 40 CFR § 503.21(n). Requirements for surface disposal in
Part 503 include placement restrictions, methane monitoring, and pollutant limits where applicable,
among others. Surface disposal sites may be unlined or lined with leachate collection systems. Preamble
DRAFT
3
-------
to 40 CFR § 503, 58 FR 9301, February 19, 1993. There are no chemical pollutant limits in Part 503 for
surface disposal sites with a liner and leachate collection system. Unlined surface disposal sites must
meet the applicable pollutant requirements in Part 503, 40 CFR §§ 503.23(a)-(b). Liners at surface
disposal sites would be required if the sewage sludge exceeds contamination levels for certain metals in
40 CFR part 503, 40 CFR §§ 503.23(a)-(b). The only restrictions on distance to adjacent properties from
surface disposal are based on the contamination levels of the sewage sludge with arsenic, chromium,
and nickel at an unlined surface disposal site. 40 CFR § 503.23(a)(2).
Sewage sludge incinerators (SSIs) are regulated by Part 503 under the CWA and under the Clean Air Act
(CAA). Requirements for incineration in Part 503 include stack monitoring and pollutant concentrations,
among others. 40 CFR § 503 Subpart E. Pollutant limits in sewage sludge fed into an incinerator are
based on risk specific concentrations calculated using dispersion factors and operating parameters
including stack height. 40 CFR § 503.43. Sewage sludge incineration regulations allow higher dispersion
factors for stack heights over 65 meters. 40 CFR § 503.43. The EPA's rules regarding emissions from SSIs
were updated in 2016. More information on the EPA's CAA regulations for SSIs can be found on EPA's
website for the New Source Performance Standards and Emission Guidelines (US EPA, 2023a).
1.4 History of Sewage Sludge Risk Assessment
In 1987, the US Congress passed the Water Quality Act, which amended the CWA to require the EPA to
establish a comprehensive program to reduce potential environmental risks associated with sewage
sludge management and maximize the beneficial reuse of sewage sludge. As amended, Section 405(d) of
the CWA required the EPA to establish numerical limits and management practices that protect public
health and the environment from the reasonably anticipated adverse effects of toxic pollutants in
sewage sludge. The amendment required two rounds of sewage sludge regulations and set deadlines for
the EPA to establish those regulations. In 1993, the EPA promulgated the first rule (called "Round One,"
40 CFR part 503), which set numeric limits in sewage sludge for ten metals (arsenic, cadmium,
chromium, copper, lead, mercury, molybdenum, nickel, selenium, and zinc). In that action, the EPA
further identified 31 pollutants and pollutant categories to be prioritized for the second planned rule
(called "Round Two"). On October 25, 1995 (60 FR 54763), chromium land application pollutant limits
were withdrawn, the selenium limits were modified, and the EPA narrowed the original list of 31
prioritized pollutants to two pollutant groups for the second round of rulemaking: polychlorinated
dibenzo-p-dioxins/dibenzofurans and dioxin-like co-planar polychlorinated biphenyls (PCBs) (US EPA
1996). On December 23, 1999, the EPA proposed numeric limits for dioxins, dibenzofurans, and co-
planar PCBs (also called "dioxin-like PCBs") in sewage sludge applied to land and proposed not to
regulate dioxins in sewage sludge disposed of in a surface disposal unit or fired in a sewage sludge
incinerator (64 FR 72045). On June 12, 2002, the EPA published a Notice of Data Availability containing
new information related to dioxins in land-applied sewage sludge and requested public comments (67
FR 40554). Based on these new data and revised risk assessment conclusions, on October 24, 2003, the
EPA determined that regulation of dioxins in sewage sludge was not warranted (68 FR 61084). The
supporting technical documentation for the 1993 "Round One" regulation and the 2003 "Round Two"
determination not to regulate put forward a general framework for sewage sludge risk assessment that
is used for this draft risk assessment of PFOA and PFOS.
As described above, the EPA's previous sewage sludge risk assessments have assessed uses and disposal
options for sewage sludge that potentially present risk to humans, crops, livestock, or wildlife (US EPA,
1992; US EPA, 1995a; US EPA, 2003a). In the 1992 technical support document, the EPA based numerical
limits for sewage sludge when applied to agricultural land on a modeled assessment of the potential risk
to public health and the environment through 14 pathways of exposure related to land application or
disposal. These pathways were split into two categories: pathways relevant to agricultural land and
DRAFT
4
-------
pathways relevant to non-agricultural land. Agricultural land application scenarios included use of
sewage sludge by a farmer for food or feed crops on pasture or rangeland, including large farms or
home gardeners. Non-agricultural use and disposal scenarios included use on forest land; land
reclamation sites; "public contact sites/' which may include lands like golf courses; and surface disposal
sites. When evaluating risks associated with sewage sludge that is incinerated, the EPA assessed a single
pathway of exposure - inhalation - and did not include air transport and deposition onto soils or surface
waters. The 14 pathways of exposure modeled in the 1992 assessment were as follows; the exposed
individual for each pathway with a human receptor is listed in square brackets:
• Reuse (land application)
- Sludge-soil-plant-human [consumer] (pathway 1)
- Sludge-soil-plant-human [home gardener] (pathway 2)
- Sludge-soil-human [child] (pathway 3)
- Sludge-soil-plant-animal-human [farm household] (pathway 4)
- Sludge-soil-animal - human [farm household] (pathway 5)
- Sludge soil-plant-animal (pathway 6)
- Sludge-soil-animal (pathway 7)
- Sludge-soil-plant (pathway 8)
- Sludge-soil- soil organism (pathway 9)
- Sludge-soil-soil organism-soil organism predator (pathway 10)
- Sludge-soil-airborne dusts-human [tractor operator] (pathway 11)
- Sludge-soil-surface water -human [person consuming drinking water and fish] (pathway 12)
- Sludge-soil-air-human [off-site resident] (pathway 13)
- Sludge-soil-groundwater-human [person consuming drinking water] (pathway 14)
• Surface disposal
- Sludge-soil-air-human [off-site resident] (pathway 13)
- Sludge soil groundwater-human [person consuming drinking water] (pathway 14)
• Incineration
- Sludge-incineration particulate -air-human [off-site resident] (pathway 13).
A graphical depiction of each of the pathways evaluated in this risk assessment is presented in Section
2.8.
As described in the 1992 technical support document, the farm family was considered to be the most
exposed population to land applied sewage sludge due to their potential exposures to consuming their
own crops and interacting directly with the contaminated soils. All the human-health based regulations
were protective of the incidental soil ingestion pathway for children because this pathway was
considered to be sensitive for human health across all life stages and potential exposure pathways.
Chemicals were also assessed for ecological risk including risk to crop growth and livestock that fed on
those plants.
In the second round of risk assessment, the EPA considered dioxin-like compounds to be the only
chemicals that merited a full risk assessment. The EPA performed a Monte Carlo analysis of exposure to
the farm family using national sewage sludge survey concentrations to estimate exposures across the
dietary pathways established in the 1993 regulations, with minor adjustments to allow for the
assessment of specific animal products (such as milk) relevant to dioxins. The risk assessment
aggregated ingestion exposures pathways (milk, meat, soil etc.) and included a cumulative assessment
across chemicals in the dioxin category (US EPA, 2003a;b). The EPA later concluded that the 95th
DRAFT
5
-------
percentile exposures from this assessment did not exceed the hazard-based reference doses in the
assessment (US EPA, 2003c). This conclusion justified the decision to not regulate PCBs or dioxins in any
use or disposal practice for sewage sludge based on the risk levels estimated for highly exposed
populations (US EPA, 2003c).
This assessment for PFOA and PFOS follows the general frameworks set out in the EPA's 1992 and 2003
assessments, with some modifications to account for the chemical and environmental characteristics of
PFOA and PFOS.
2 PROBLEM FORMULATION
2.1 Literature Search Strategy and Information Management
Risk assessment for land application and disposal of sewage sludge requires the synthesis of available
information from a diverse set of academic fields of research: chemical occurrence in sewage sludge,
environmental fate and transport, human toxicology, aquatic toxicology, plant toxicity, and wildlife or
ecological effects. The assessment further benefits from a background understanding of the chemical's
use profile in the U.S. economy and the uses or disposal options common for sewage sludge generated
in the U.S. To efficiently synthesize this information, the EPA takes a hierarchical approach to
information management. When possible, the EPA sources background information and risk assessment
conclusions from publicly available, peer-reviewed documents such as EPA Human Health Toxicity
Assessments, Health Effects Support Documents, Aquatic Life Ambient Water Quality Criteria, Agency
for Toxic Substances and Disease Registry (ATSDR) Toxicological Profiles, Environment and Climate
Change Canada (ECCC) Federal Environmental Quality Guidelines, European Food Safety Authority
(EFSA) Scientific Opinions, and other such assessments. When assessments are not available on a
specific topic or not up to date with current scientific findings, the EPA conducts literature reviews of
peer-reviewed journal articles and state agency "gray literature" reports. Background information
summarized in the Problem Formulation (Section 2) of this assessment is based on existing assessments.
The literature search strategies employed for the model parameters are described in Model
Parameterization (Section 2.9.3).
2.2 The Nature of the Chemical Stressor
2.2.1 Chemical Identity
PFOA and PFOS are manufactured for direct use in industry and in commerce, in addition to a range of
other chemical structures containing fluorinated carbons (Buck et al., 2011; OECD, 2021; US EPA,
2021b). Some of these other PFAS can degrade in the environment to PFOA or PFOS, which are then
stable degradation and metabolic products. The PFAS that degrade to PFOA and PFOS are called
precursors. Generally, precursors to PFOA and PFOS also contain a fluorinated carbon chain with eight
or more carbons.
PFOA and PFOS have been part of a voluntary phase out for domestic manufacture and their uses have
been restricted by Significant New Use Rules (SNURs) issued by the EPA under the Toxic Substances
Control Act (TSCA) and the US Food and Drug Administration (FDA) phase out for food packaging (FDA,
2016; US EPA, 2024a). While these actions may have reduced the presence of these chemicals in
domestic sewage, PFOA and PFOS continue to be detected in wastewater and sewage across the U.S.
due to their presence in residential, commercial, and industrial products that were manufactured or
imported before the phase-out, their presence in products or processes associated with the limited
number of ongoing allowable uses (US EPA, 2021c), their persistence in waste disposal sites like landfills,
and their pervasive existing environmental contamination (see Sections 2.2 and 2.3).
DRAFT
6
-------
PFOA: PFOA is a synthetic fluorinated organic chemical that has been manufactured and used in a
variety of industries since the 1940s (US EPA, 2018a). The chemical abstracts service registry number
(CASRN), common synonyms, chemical formula, and other basic chemical properties are described in
Table 1. PFOA repels water and oil, is chemically and thermally stable, and exhibits surfactant
properties. Based on these properties, it has been used in the manufacture of many materials, including
cosmetics, paints, polishes, and nonstick coatings on fabrics, paper, and cookware. It is very persistent in
the human body and the environment (Calafat et al., 2007; 2019). More information about PFOA's uses
and properties can be found in the EPA's 2024 Final Human Health Toxicity Assessments for PFOA (US
EPA, 2024b). In 2006, the EPA invited eight major companies to commit to working toward the
elimination of their production and use of PFOA (and chemicals that degrade to PFOA) and elimination
of these chemicals from emissions and products by the end of 2015. All eight companies have since
phased out manufacturing PFOA. Despite this commitment of these major producers, PFOA may be
produced, imported, and used by companies not participating in the PFOA Stewardship Program and
some uses of PFOA are ongoing (see 40 CFR 721.9582). PFOA is included in EPA's SNUR issued in January
2015, which ensures that the EPA will have an opportunity to review any efforts to reintroduce the
chemical into the marketplace and take action, as necessary, to address potential concerns (US EPA,
2015). Limited existing uses of PFOA-related chemicals, including as a component of anti-reflective
coatings in the production of semiconductors, were excluded from the regulations (US EPA, 2021c) and
PFOA may still be a component of articles (manufactured items) imported into the U.S.
Table 1. Chemical and Physical Properties of PFOA.
Property
PFOA, acidic form1
Source
CASRN
335-67-1
NA
Chemical Abstracts Index
2,2,3,3,4,4,5,5,6,6,7,7,8,8,8-
NA
Name
pentadecafluorooctanoic acid
Synonyms
PFOA; Pentadecafluoro-1-octanoic acid;
Pentadecafluoro-n-octanoic acid; Octanoic
acid, pentadecafluoro-; Perfluorocaprylic
acid; Pentadecafluorooctanoic acid;
Perfluoroheptanecarboxylic acid;
NA
Chemical Formula
C8HF15O2
NA
Molecular Weight (grams
per mole [g/mol])
414.07
PubChem Identifier (CID 9554) (URL:
https://oubchem.ncbi.nlm.nih.aov/comoound/9554); Lide
(2007)
Color/Physical State
White powder (ammonium salt)
PubChem Identifier (CID 9554) (URL:
https://oubchem.ncbi.nlm.nih.aov/comoound/9554)
Boiling Point
192.4 °C
Lide (2007); SRC (2016)
Melting Point
54.3 °C
Lide (2007); SRC (2016)
Vapor Pressure
0.525 mm Hg at 25 °C (measured)
0.962 mm Hg at 59.25 °C (measured)
Hekster et al. (2003); SRC (2016) ATSDR (2021); Kaiser et
al. (2005)
Kaw
0.00102 (experimentally determined,
equivalent to Henry's Law Constant of
0.000028 Pa-m3/mol at 25 °C)
Li etal. (2007)
Kow
Not measurable
UNEP (2015)
pKa
3.15 (mean measured)
Burns et al. (2008) and 3M (2003) as reported in EPA
Chemistry Dashboard (URL:
https://comotox.eoa.aov/dashboard/dsstoxdb/results
?search=D TXSID8031865#prooerties)
Solubility in Water
9,500 mg/L (estimated);
3,300 mg/L at 25 °C (measured)
Hekster etal. (2003);
ATSDR (2021)
1 PFOA is most commonly produced as an ammonium salt (CASRN 3825-26-1). Properties specific to the salt are not included.
PFOS: PFOS is a synthetic fluorinated organic chemical that has been manufactured and used in a variety
of industries since the 1940s (US EPA, 2018a). The CASRN, common synonyms, chemical formula, and
DRAFT
7
-------
other basic chemical properties are described in Table 2. Like PFOA, PFOS repels water and oil, is
chemically and thermally stable, and exhibits surfactant properties. Based on these properties, it has
been used in the manufacture of many materials, including cosmetics, paints, polishes, and nonstick
coatings on fabrics, paper, and cookware. Like PFOA, PFOS is very persistent in the human body and the
environment (Calafat et al., 2007; 2019). More information about PFOS's uses and properties can be
found in the EPA's 2024 Final Human Health Toxicity Assessments for PFOS (US EPA, 2024c). In 2000, the
principal manufacturer of PFOS agreed to a voluntary phase-out of PFOS production and use over time.
This phase-out agreement was completed in 2002 (US EPA, 2007). PFOS is included in EPA's SNUR issued
in December 2002, which ensures that the EPA will have an opportunity to review any efforts to
reintroduce PFOS into the marketplace and take action, as necessary, to address potential concerns (US
EPA, 2002). Limited existing uses of PFOS-related chemicals, including as an anti-erosion additive in fire-
resistant aviation hydraulic fluids and as a component of antireflective coating in the production of
semiconductors, were excluded from the regulation (US EPA, 2013) and articles imported into the U.S.
may have PFOS. Due to the high human health toxicity of PFOS, all environmental releases may be
significant; however, known major sources of PFOS contamination in the U.S. include past
manufacturing of PFOS, use of PFOS as a mist suppressant in chrome plating facilities, use of PFOS as an
oil and water-resistant coating for paper products, textiles, and leather, and use of PFOS-containing
firefighting foams, especially at training and testing sites.
Table 2. Chemical and Physical Properties of PFOS
Property
PFOS, acidic form1
Source
CASRN
1763-23-1
NA
Chemical Abstracts Index
Name
1,1,2,2,3,3,4,4,5,5,6,6,7,7,8,8,8- heptadecafluoro-1-octanesulfonic
acid
NA
Synonyms
Perfluorooctane sulfonic acid; heptadecafluoro-1-octane sulfonic
acid; PFOS acid; perfluorooctane sulfonate
NA
Chemical Formula
C8HF17O3S
NA
Molecular Weight (grams per
mole fg/moll)
500.13
Lewis (ed. 2004); SRC (2016)
Color/Physical State
White powder (potassium salt)
OECD (2002)
Boiling Point
258-260 °C
SRC (2016)
Melting Point
No data
Vapor Pressure
2.48 x 10"6 mm Hg at 20°C (potassium salt)
ATSDR (2021)
Henry's Law Constant
Not measurable; not expected to volatilize from aqueous solution
(< 2.0x10-6)
ATSDR (2021)
Kow
Not measurable
EFSA (2008); ATSDR (2021)
pKa (modeled)
0.14 (no empirical measurements available)
ATSDR (2021)
Solubility in Water
570-680 mg/L
OECD (2002); ATSDR (2021)
1 PFOS is commonly produced as a potassium salt (CASRN 2795-39-3). Properties specific to the salt are not included.
Tables 1 and 2 include a summary of physical and chemical properties for PFOA and PFOS. While these
values provide important context for understanding the general behaviors of the chemical, when
assessing the relevance of reported physical properties to their behavior in the soil environment, it is
important to ensure that the method for collecting the physical and chemical data is relevant to the
environmental conditions modeled in the risk assessment. For example, measurements of volatility like
vapor pressure or Henry's law constant performed on the acid at low pH (Li, 2007) may be useful for
understanding PFOA or PFOS in a laboratory or industrial setting, but farm fields tend to have pH values
closer to neutral pH where PFOA and PFOS exist as an anion. Using these physical property values
directly to estimate volatility from a farm field may be misleading. Section 2.9.3 of this document
describes the physical and chemical properties used to parameterize models used in this risk assessment
and describes how studies were selected to best capture relevant environmental conditions.
DRAFT
8
-------
2.2.2 Transformation and Degradation of Precursors
PFOA and PFOS do not undergo degradation under environmentally relevant conditions (US EPA, 2008;
OECD, 2002; Schultz et al., 2003), in part because environmental degradation pathways and processes
do not apply enough energy to break fluorine-carbon bonds (3M, 2000; Hekster et al., 2003; Schultz et
al., 2003). ATSDR Toxicological Profiles for Perfluoroalkyls (including PFOA and PFOS) conclude that
these perfluoroalkyl acids are resistant to biodegradation, direct photolysis, atmospheric
photooxidation, and hydrolysis (ATSDR, 2021; OECD, 2002; Prevedouros et al., 2006). Some researchers
are exploring the potential for degradation in soil systems that are undergoing remediation (Huang et
al., 2022).
The processing of influent and sewage sludge at wastewater treatment plants (WWTPs) provides
opportunities for fluorinated precursors to biodegrade to PFOA or PFOS, which are terminal degradants.
Examples of precursors to PFOS include perfluorooctane sulfonamidoethanol-based phosphate diester
(sAmPAPs) containing carbon-chain moieties with at least eight fluorinated carbons, A/-ethyl
perfluorooctane sulfonamidoethanol (NEtFOSE), A/-ethyl perfluorooctane sulfonamidoacetic acid
(NEtFOSAA), perfluorooctane sulfonamidoethanol (FOSE), perfluorooctane sulfonamidoacetic acid
(FOSAA), A/-ethyl perfluorooctane sulfonamide (NEtFOSA), and perfluorooctanesulfonamide (PFOSA).
Commonly detected precursors to PFOA include fluorotelomer alcohols (FTOHs), polyfluoroalkyl
phosphoric acids (PAPs) and polyfluorinated iodides (PFIs) that contain a fluorinated carbon chain
moiety with at least eight carbons in the chain {i.e., 8:2 FTOH). Sidechain fluorotelomer-based polymers
(FTPs), especially those used on textiles, could also be significant sources of PFOA to WWTPs because
they can transform when laundered or cleaned and when weathered in soils (Washington et al., 2015;
Washington & Jenkins, 2015; Liagkouridis et al., 2022; van der Veen et al., 2022). The treatment of
sewage sludge to create biosolids (Thompson et al., 2023a) and the land application of biosolids
(Schaefer et al., 2022) both provide opportunities for precursors to degrade into PFOA and PFOS.
PFOA and PFOS precursors have been used by industry and imported in consumer products. When these
chemicals enter the environment, the molar yields for their transformation to PFOA or PFOS and their
degradation rates vary. Laboratory measurements have shown that microbes common to WWTPs
(Lange, 2000) and other environmental systems can biodegrade these precursors to PFOA and PFOS.
Biosolids-amended soil in column studies have observed that the degradation of PFAS precursors may
be responsible for a significant portion of PFOA and PFOS that occur in the environment (Schaefer et al.,
2022).
Due to data gaps regarding the occurrence, environmental fate and transport, degradation pathways,
bioaccumulation, and toxicity of precursors to PFOA and PFOS, the EPA is focusing this draft risk
assessment on PFOA and PFOS. That said, the occurrence data of PFAS in biosolids indicate precursors
significantly contribute to the overall PFOA and PFOS loading to soils and disposal facilities (see Section
2.4). Future assessments could be expanded to include other chemicals including environmental
precursors to PFOA and PFOS, or other PFAS. Additionally, policy decisions regarding the treatment of
quantifiable precursors to PFOA and PFOS could be considered in the future.
2.2.3 En vironmentai Fa te and Transport
PFOA and PFOS are persistent in the environment and are commonly called "forever chemicals" due to
the lack of observed degradation pathways. They are also mobile in the environment and bioaccumulate
in organisms. The EPA and state monitoring programs have found that historic land application of
sewage sludge containing PFOA and PFOS has contaminated soil, surface water, groundwater, crops,
DRAFT
9
-------
beef, eggs, and milk and impacted farm families (Washington et al., 2010; Lindstrom et al., 2011; Yoo et
al., 2011; Moavenzadeh Ghaznavi et al., 2023).
PFOA and PFOS can undergo several transport mechanisms after release to a soil environment. These
include sorption to soils and sediments, sorption to fluid-fluid interphases, runoff, erosion, migration to
groundwater, and uptake into plants and animals. The surfactant-like properties of PFOA and PFOS
influence the way they move through natural systems. For example, because PFOA and PFOS sorb to
fluid-fluid interfaces (Sharifan et al., 2021) some modeling assumptions used for other organic chemicals
are not appropriate for PFOA and PFOS. PFOA and PFOS exhibit varying partitioning between soil and
water, air and water, or biosolids and water depending on the presence and type of organic matter
(Ebrahimi et al., 2021), oxalate-extractable grain coatings, mineral composition (Gravesen et al., 2023),
the presence of air-water interfaces (Costanza et al., 2019) and other factors (Sharifan et al., 2021).
These properties have been shown in the literature to result in a wide range of potential values of soil-
water and air-water sorption constants across different types of soils (see Appendix C). The degree of
soil-water and air-water sorption influences transport behavior from the soil to other media like
groundwater and surface water.
Several studies characterize PFAS partitioning behavior between the solid and aqueous phases in
sewage sludges (Zhang et al., 2013; Ebrahimi et al., 2021; Lewis et al., 2023; Gravesen et al., 2023).
While a correlation has been found between bulk organic matter content and PFAS partitioning
(particularly for long-chain PFAS), research has shown that protein content has the strongest correlation
to PFAS partitioning in biosolids when compared to lipids and bulk organic matter, which aligns with the
observation that PFOA and PFOS bind to proteins in animals (Zhang et al., 2013; Ebrahimi et al., 2021,
Section 2.5.1). Also, a more recent study investigated the effects of microbial weathering on PFAS
partitioning over time after biosolids land application to examine the fate and transport of PFAS leaching
from biosolids into the environment (Lewis et al., 2023). Results revealed that microbial weathering
plays a role in PFAS partitioning, contributing to the biodegradation of organic matter and leading to an
increased potential for PFAS leaching to groundwater. Another recent study examined oxalate-
extractable iron and aluminum in relation to PFAS partitioning in biosolids, finding that iron was
correlated with PFOA partitioning and aluminum was correlated with both PFOA and PFOS partitioning
(Gravesen et al., 2023). In addition, bulk organic matter was associated with PFOS partitioning, while
protein content tended to be more strongly correlated with the partitioning of shorter-chain PFAS
(Gravesen et al., 2023).
The partitioning trends described in the prior paragraph are observed in sewage sludge and are also
relevant to organic-matter rich topsoils that have been amended with biosolids; PFOA and PFOS
partitioning behavior in subsurface soils is distinct due to lower organic content, differences in the
mineral or amorphous mineraloid composition of grains and grain coatings, and the presence of air-
water interphases. Due to the low concentrations of natural organic matter in subsurface soils (0.01-
0.05%), PFAS sorption in the subsurface may have significant contributions from sorption to the surfaces
of minerals and mineraloids and sorption to the air-water interphases (Lyu et al., 2019). Most studies in
this area have been lab-based tests in well-defined media such as quartz sand or limestone, which differ
from natural soil systems. Additional study is needed on the most significant variables related to PFOA
and PFOS retention in natural subsurface soil systems.
Although volatilization of PFOA and PFOS is expected to be low from soil systems in general due to the
chemicals being ionized at typical soil pH, there may be soil systems where volatilization contributes to
atmospheric concentrations. Past research regarding soil-water environments has shown that PFAS
volatilization increases as pH decreases (Johansson et al., 2017; Sima and Jaffe, 2021). In an experiment
examining water-air transfer, highest rates of PFOA volatilization occurred at a pH of 1, while PFOA
DRAFT
10
-------
volatilization was found to be negligible at pH levels greater than 2.5 (Johansson et al., 2017).
Consequently, under natural soil-water conditions, PFOA volatilization is considered to be negligible
(Johansson et al., 2017; Sima and Jaffe, 2021). However, under natural soil-water conditions, there could
be a concern with the volatilization of precursors that can biodegrade and transform into persistent
PFAS. For example, in a past study, as much as 3% of 6:2 polyfluoroalkyl phosphate diester (diPAP, a
precursor that can transform into perfluorohexanoic acid [PFHxA], for instance) volatilized under natural
soil conditions, while 8:2 diPAP (a precursor of PFOA, for example) was found to be negligible (Liu and
Liu, 2016). More study is needed on the volatilization of other PFOA or PFOS precursors under natural
soil conditions.
2.3 Sources to Wastewater Treatment Plants and Biosolids
The EPA Chemical Data Reporting rule under TSCA requires manufacturers (including importers) to
report certain data about chemicals in commerce in the U.S., including information on PFOA and PFOS
(subject to a 2,500 pound reporting threshold at a single site). The last time PFOA and PFOS
manufacturing information was reported to the EPA pursuant to this rule was in 2013 and 2002,
respectively. However, Toxics Release Inventory (TRI) data for 2020 shows that PFOA and PFOS continue
to be released into the environment. Pursuant to TRI reporting requirements, facilities in regulated
industry sectors must report annually on releases and other waste management of certain listed toxic
chemicals that they manufacture, process, or otherwise use above certain threshold quantities
(currently 100 pounds per industrial site for PFOA and PFOS).
Despite the phase out of domestic manufacturing of PFOA and PFOS, multiple activities result in PFOA,
PFOS, and their precursors being present in WWTP influent including industrial releases (e.g.,
semiconductor manufacturing, pulp and paper plants), commercial releases (e.g., hotels, car washes,
industrial launderers), and down the drain releases from homes (e.g., laundering of coated textiles, use
of residential products). These chemicals have been used in a variety of industrial processes and
commercial and consumer products, which results in a range of potential sources to WWTPs within
communities. For example, homes may still have PFOA and PFOS-containing products in use, like after-
market water resistant sprays, floor finishes, textiles with PFOA and PFOS coatings, or ski wax. These
products could be washed down a drain or released when cleaned or laundered, or they may be
disposed of at a lined MSW landfill. The leachate from that landfill could be another ongoing source of
PFOA and PFOS to WWTPs, as the most common off-site management practice for landfill leachate is
transfer to a WWTP (US EPA, 2024g). At different WWTPs across the country, any of these release
mechanisms may play a role in PFAS entering the plant.
Sewage sludge contaminant monitoring based on typical analytical methods (e.g., EPA Method 1633, US
EPA, 2024d) can be used to test for 40 PFAS but does not include precursors such as sAmPAPs and
diPAPs. Several studies using soil columns and non-targeted analysis have found that most of the
environmental loading to biosolids will come from these precursor chemicals (Schaefer et al., 2022;
Thompson et al., 2023a;b).
2.4 Occurrence in Biosolids
Studies have shown that PFAS are frequently found in biosolids around the globe (D'eon et al., 2009;
Yoo et al., 2009; Lee et al., 2010; Washington et al., 2010; Lindstrom et al., 2011; Sepulvado et al., 2011;
Venkatesan and Halden, 2013; Lee et al., 2014; Armstrong et al., 2016; Navarro et al., 2016; Eriksson et
al., 2017; Moodie et al., 2021; Munoz et al., 2022; Fredriksson et al., 2022; Helmer et al., 2022; Johnson,
2022; Schaefer et al., 2022; Thompson et al., 2023a,b; Link et al., 2024). For a summary of PFOA and
PFOS concentrations found in biosolids in the U.S. based on studies from published peer-reviewed
literature and state reports, please see Appendix A, Tables A-l (PFOA) and A-2 (PFOS). Overall, these
DRAFT
11
-------
studies have demonstrated that PFOS is typically found more often and at higher concentrations than
PFOA in biosolids. Concentrations are reported on a dry weight basis. This review focuses on
concentrations that occur as most studies do not identify sources of PFOA or PFOS.
The PFOA and PFOS concentrations found in U.S. biosolids vary across studies (Appendix A). At the
national scale, the Venkatesan and Halden 2013 study measured 13 PFAS in composite samples
compiled from archived biosolids collected during the EPA's 2001 National Sewage Sludge Survey
(NSSS). The study authors randomly divided the 110 available samples from the 2001 NSSS (94 POTWs)
into 5 composite samples, finding average concentrations of 34±22 parts per billion (ppb) for PFOA and
403±127 ppb for PFOS (Venkatesan and Halden, 2013). PFOA concentrations ranged from 12-70 ppb and
PFOS concentrations ranged from 308-618 ppb (Venkatesan and Halden, 2013). Of the 13 PFAS analytes
measured, 10 were detected in all composited samples with PFOS found at the highest levels, surpassing
PFOA, which had the second highest concentrations (Venkatesan and Halden, 2013). A more recent U.S.
study found, on average, lower concentrations of PFOA and PFOS in biosolids: PFOA concentrations
ranged from 0.8-8.12 ppb and PFOS concentrations ranged from 0.386-150 ppb in samples collected
from multiple states (7 WWTPs with a variety of treatment processes in urban areas receiving both
industrial and domestic sources) (Schaefer et al., 2022). Another recent U.S. study analyzed samples
before and after treatment from 8 WWTPs representing the four most common biosolids treatment
processes in Florida, finding PFOA concentrations ranging from 1.7-21 ppb (before treatment) and 1.1-
7.7 ppb (after treatment), and PFOS concentrations ranging from 4-41 ppb (before treatment) and 1.4-
19 ppb (after treatment) (Thompson et al., 2023a). Though these studies with samples collected after
the PFOA and PFOS phased out in the U.S. observe lower levels of PFOA and PFOS in sewage sludge than
pre-phase out samples, sewage sludge samples with significantly elevated concentrations of PFOA and
PFOS have been identified from industrially impacted WWTPs as recently as 2022 (Link et al., 2024).
Several states also have implemented programs to monitor for PFAS in their biosolids. For instance,
Michigan's extensive sampling found that PFAS levels tended to be higher in biosolids receiving
industrial sources (Ml EGLE, 2021a,b; 2022; Helmer et al., 2022). Consequently, Michigan instigated
industrial pretreatment program best management practices to limit PFAS source contributions (Ml
EGLE, 2021a, 2022; Helmer et al., 2022), which most recently has led to a PFOS reduction in biosolids of
more than 85% at four of the six wastewater treatment plants studied (Ml EGLE, 2022). A recent study
analyzed Michigan's statewide biosolids data collected between 2018 and 2022 from 190 wastewater
treatment plants representing both industrial and domestic sources, finding mean dry weight
concentrations of 4.8±11 ppb for PFOA with a detection rate of 63% and 40±179 ppb for PFOS with a
detection rate of 95% (Link et al., 2024). Based on biosolids data in Maine's Environmental and
Geographic Analysis Database collected from 2019 to 2022, Maine's comprehensive state sampling
found mean PFOA concentrations of 9.4 ppb in 2019, 8.2 ppb in 2020, 5.3 ppb in 2021, and 6.6 ppb in
2022, and mean PFOS concentrations of 27.2 ppb in 2019, 16.6 ppb in 2020, 22.7 ppb in 2021, and 19.3
ppb in 2022 (Brown and Caldwell, 2023). New Hampshire also has performed detailed PFAS analyses of
soils, biosolids, solid/water partitioning, and groundwater leaching through a three-phase study
conducted by the US Geological Survey and the New Hampshire Department of Environmental Services
(Phase 1: Santangelo et al., 2022; Phase 2: Tokranov et al., 2023; Phase 3: Santangelo et al., 2023). Data
from the finished biosolids collected from facilities in 2021 as part of Phase 2 found PFOA and PFOS dry
weight concentrations of less than 8 ppb across samples (Tokranov et al., 2023).
The EPA is currently planning for the next NSSS in collaboration with the Effluent Guidelines Program's
POTW Influent Study, both of which will focus on testing for PFAS ([March 26, 2024] (89 FR 20962);
[October 10, 2024] (89 FR 82238)). This joint monitoring study will provide a current and comprehensive
DRAFT
12
-------
national dataset of PFAS concentrations found in the influent, effluent, and sewage sludge of POTWs, as
well as their industrial and domestic sources.
As previously described, precursors also can transform into PFOA or PFOS in primary and secondary
processes of wastewater treatment plants and in the environment. As an example, diPAPs can
biodegrade and transform into persistent perfluoroalkyl carboxylic acids (PFCAs) (e.g., 8:2 diPAP can
transform into PFOA) (D'eon and Mabury, 2007; Lee et al., 2010; 2014), potentially leading to increased
PFCA loads in wastewater effluent and land-applied biosolids that can contribute to aquatic and
terrestrial contamination (Lee et al., 2010; 2014). A recent field study in Germany demonstrated that
diPAPs stemming from paper production have the capacity to transform into PFCAs that can leach out of
soil into drinking water sources (Lammer et al., 2022). The Schaefer et al. 2022 study not only tested U.S.
WWTP biosolids, but also performed column mesocosm leaching experiments, finding that precursors to
the 18 measured polyfluoroalkyl acids (PFAAs; e.g., diPAPs, 5:3 fluorotelomer carboxylic acid [FTCA],
perfluorophosphonic acids) accounted for over 75% of the total PFAS fluorine mass in biosolids
(Schaefer et al., 2022). Notably, this study found that total oxidizable precursor assay (TOP assay) in
biosolids extracts was not able to quantify all precursors to PFOA and PFOS because the assay did not
fully oxidize precursors like diPAPs. In addition, the Thompson et al. 2023a study analyzed 92 PFAS
analytes in total, including precursors, and found that 6:2 diPAP, 6:2/8:2 diPAP, and 8:2 diPAP were the
most common diPAPs identified in biosolids (Thompson et al., 2023a). The study results also showed
that there is currently an underestimation of total PFAS concentrations due to the high potential for
precursor transformation and lack of available analytical test methods that include these precursors in
their targeted list of PFAS analytes (Thompson et al., 2023a). In another recent article analyzing toilet
paper samples from the U.S. and other countries, along with U.S. sludge samples, 6:2 diPAP was
detected at the highest concentrations in both toilet paper and sludge samples (Thompson et al.,
2023b). Though some of these precursors do not transform to PFOA and PFOS, conducting non-targeted
analysis and including more precursors in targeted methods can aid in resolving this issue of identifying
unknown PFOA and PFOS precursors. Appendix A (Table A-3) provides examples of occurrence data for
potential PFOA and PFOS precursors found in biosolids in the U.S. based on recent studies.
Despite the phase-out of long-chain PFAS (e.g., PFOA and PFOS), the most recent U.S. studies still show
that PFOS is typically found at the highest concentrations in biosolids of the traditional targeted list of
PFAS analytes measured (Helmer et al., 2022; Link et al., 2024). Recent investigations in Michigan that
include industrially impacted biosolids have shown PFOS concentrations as high as 2,150 ppb (Ml EGLE,
2022; Link et al., 2024) and 6,500 ppb (Ml EGLE, 2021a; Helmer et al., 2022); Michigan implemented
industrial pretreatment program best management practices to address these PFOS sources and these
concentrations have been reduced (Ml EGLE, 2021a; 2022; Helmer et al., 2022). Michigan did not
include PFOA and PFOS precursors in their industrial pretreatment and biosolids management strategy.
Schaefer et al. 2022 found that concentrations of 8:2 diPAP exceeded concentrations of PFOA in
biosolids. Schaefer et al. 2023 found that the sum of N-ethyl perfluorooctanesulfonamide (NEtFOSA),
PFOSA, 8:2 fluorotelomer sulfonate (8:2 FTS), 8:2 FTCA, and N-methyl perfluorooctanesulfonamide
acetic acid (NMeFOSAA) exceeded concentrations of PFOA and PFOS in biosolids. Thompson et al. 2023a
found that the sum of 8:2 diPAP, 6:2/8:2 diPAP, FOSAA, NMeFOSAA, NEtFOSAA and 8:2 FTS
concentrations also exceeded PFOA and PFOS concentrations in biosolids, with 8:2 diPAP and 6:2/8:2
diPAP being the most significant contributors to the total measured PFOA and PFOS precursor
concentration. 8:2 diPAP and 6:2/8:2 diPAP are not currently included in EPA's analytical method
recommended for sewage sludge, EPA 1633 (US EPA, 2024d).
DRAFT
13
-------
2.5 Uptake and Bioaccumulation
This section provides a brief overview of PFOA and PFOS accumulation into animals (Section 2.5.1) and
plants (Section 2.5.2). There are several characteristics of PFOA and PFOS uptake in humans, other
animals, and plants that are important to understanding the overall fate and toxicity of these chemicals
in biosolids-specific environmental release scenarios. A detailed description of how each uptake factor is
parameterized in biosolids fate and transport models can be found in Section 2.9.3. Overall, PFOA and
PFOS bioaccumulate in humans, fish, livestock, wildlife, and plants.
2.5.1 Animals
Humans: PFOA and PFOS accumulate in humans, and a detailed description of human absorption,
distribution, metabolism, and elimination for PFOA and PFOS is available in the EPA's Final Human
Health Toxicity Assessments (US EPA, 2024b;c). In contrast to many persistent organic pollutants that
tend to partition to fats, PFOA and PFOS preferentially bind to proteins (Martin et al., 2003a;b). Within
the body, PFOA and PFOS tend to bioaccumulate within protein-rich tissues, such as the blood serum
proteins, liver, kidney, and gall bladder (De Silva et al., 2009; Martin et al., 2003a;b). Half-lives in humans
differ by sex due to the elimination pathway of menstruation, lactation, and childbirth for women. PFOA
and PFOS undergo enterohepatic recirculation, in which PFOA and PFOS are excreted from the liver in
bile to the small intestine, then reabsorbed and transported back to the liver (GoeckeFlora and Reo,
1996). Reuptake also occurs through the kidneys (US EPA, 2024b;c). This reabsorption is one reason why
PFOA and PFOS are retained for long time periods in the human body, and in the bodies of some other
animals. PFOA and PFOS can be passed from mother to child in utero (through placental transfer) and in
early life through breastmilk (US EPA, 2024b;c).
The Centers for Disease Control and Prevention (CDC) National Health and Nutrition Examination Survey
(NHANES) has measured blood serum concentrations of several PFAS in the general U.S. population
since 1999. PFOA and PFOS have consistently been detected in up to 98% of serum samples collected in
biomonitoring studies that are representative of the U.S. general population. However, blood levels of
PFOA declined by more than 60% between 1999 and 2014, presumably due to restrictions on PFOA
commercial usage in the US (CDC, 2017). Blood levels of PFOS similarly declined by more than 80%
between 1999 and 2014, a decline which also coincides with restrictions on PFOS commercial usage in
the U.S. (CDC, 2017). Serum levels of people living in regions impacted by point source releases of PFOA,
PFOS, and their precursors have elevated serum levels compared to the general population (MDH, 2010;
US EPA, 2024b;c). For example, a 2024 study in Maine of 30 individuals from 19 households who have
been living on farms with PFAS contamination for an average of 23.7 years found that this group's serum
levels of PFOA, PFOS and other PFAS were significantly higher than the general population (Criswell et
al., 2024). Further, the authors found that the farm families had serum levels similar to those seen in
other highly exposed populations, including the C8 study cohort (Criswell et al., 2024; Frisbee et al.,
2009).
Other animals: PFOA and PFOS are consistently detected in aquatic and terrestrial animals across the
globe (Giesy and Kannan, 2001; US EPA, 20241,m; De Silva et al., 2021). Accumulation is observed in
game species (deer, ducks, fish) as well as other wildlife (Death et al., 2021). In wildlife, PFOS is generally
observed with a higher frequency of detection and concentration than PFOA. In several areas with point
sources of PFOS to the environment, state agencies have issued consumption restriction advisories for
fish and game (MDHHS, 2023; MDIFW, 2021; MPCA, 2023a; NCDHHS, 2023).
Just as there are sex differences in the elimination rate of PFOA and PFOS in humans, these sex
differences have also been observed in non-human animal species. For example, Lee and Schultz 2010
observed that the elimination rate of PFOA from blood plasma was ten times faster in female fathead
DRAFT
14
-------
minnows compared to males. The faster elimination rate may be related to sex hormone (i.e., androgen
and estrogen) levels, as the elimination rate in females decreased four-fold following exposure to the
androgen trenbolone (Lee and Schultz, 2010). This pattern has also been observed in rats, where the
elimination of PFOA was 70 times faster in females than males, which was attributed to sex-related
differences in the expression of organic anion transporters in kidneys (Kudo et al., 2002). The degree to
which sex-related differences in elimination rate apply to other fish species, or other taxonomic groups,
may vary.
The EPA recently published a summary of PFAS occurrence information in freshwater fish from
randomly selected sampling points in the U.S. portion of the Great Lakes (US EPA, 2024h). This study
finds that PFOS is detected in the edible filets of 100% of freshwater fish samples, while PFOA was
detected in 23% of samples. The range of PFOS concentrations found in filets is 0.366 to 49.3 ng/g. The
range of detected concentrations of PFOA in fish filets is 0.086 to 1.41 ng/g. Both dissolved PFOS and
sediment-sorbed PFOS contribute to the uptake of PFOS into freshwater fish (Balgooyen & Remucal,
2022; Barbo etal., 2023).
2.5.2 Plants
Uptake of PFOA and PFOS from biosolids-amended soils by crops is a potential pathway for entry into
the human food chain, and plant uptake generally is one of several potential pathways for wildlife and
livestock exposures. Generally, the degree of phytoaccumulation of a given chemical from soil to plants
is either assessed with greenhouse-based lab studies or field-based studies. Previous biosolids risk
assessments for metals indicated that greenhouse studies tended to result in higher measured uptake
from soil to plants than field studies (US EPA, 1992). This assessment hypothesized that the differences
in observed uptake of metals could be due to 1) increased transpiration in humid greenhouses, 2) higher
concentrations of soluble salts in greenhouse pot soil due to the application of nutrients in a limited soil
volume, which increases diffusion of metals from soil particles to roots, 3) soil acidification in
greenhouse pots due to application of certain fertilizers in a small soil volume, which results in increased
metal uptake, and 4) the soil-sludge mixture in greenhouse pots comprise the entire rooting medium,
while in the field, sludge amended soils only extend to the tillage depth, and roots can extend below this
depth. Some of these factors are also applicable to PFOA and PFOS uptake in plants, and consideration is
needed of these factors when assessing plant uptake studies.
It has long been known that PFOA and PFOS can accumulate in plants eaten by humans (D'Hollander et
al., 2015; 3M, 2001). Few studies have measured plant uptake data available for biosolids amended soil
at field sites, but some PFOA and PFOS data from these biosolids-specific field studies are available (Yoo
et al., 2011; Blaine et al., 2013). Both of these field-based studies are useful for understanding uptake
into forage and silage, which improves the strength of the assessment of exposures to pastured
livestock due to diet. Data on PFOA and PFOS accumulation into other plant species (e.g., human food
crops like fruits and vegetables) grown in biosolids-amended fields are somewhat limited; this
represents a data gap for biosolids risk assessment because these measurements are the most direct
way to understand exposures to humans who consume crops (fruits and vegetables) from biosolids-
amended soils.
In October 2023, the EPA announced a new funding opportunity for research that furthers our
understanding of PFAS uptake and bioaccumulation in plants and animals in agricultural, rural, and
Tribal communities (US EPA, 2023b). The EPA is also aware of several ongoing studies regarding PFAS
and plant uptake from biosolids-amended soils, which will likely be completed by the end of 2024.
Ideally, as more data are collected in this area and the mechanistic understanding of PFOA and PFOS
uptake into various types of plants and plant compartments improves, the uncertainty regarding
exposure modeling from plants to humans, livestock, and wildlife will decrease.
DRAFT
15
-------
PFOA and PFOS are taken up into various compartments of plants. There are many types of study design
that measure PFOA and PFOS plant uptake, including laboratory studies that grew plants in natural soils,
spiked soils, or spiked water, and field studies investigating plants grown in soil that have or have not
been amended with biosolids. Some studies focus on accumulation in plants consumed by humans or
animal feed (Yoo et al., 2011; Lechner and Knapp, 2011; Lee et al., 2014; Blaine et al., 2014;
Bizkarguenaga et al., 2016a,b; Wen et al., 2016; Liu et al., 2017; Navarro et al., 2016; Ghisi et al., 2019;
Kim et al., 2019; Li et al., 2019), others focus on how accumulation intersects with phytotoxicity (Lin et
al., 2020; Zhou et al., 2016), and finally others look for "hyperaccumulating" plants including aquatic
plants like pondweed and water-starwort, or terrestrial plants like long beech fern, sunflower, and hemp
(Li et al., 2021; Nassazzi et al., 2023).
PFOA and PFOS both accumulate in food and feed crops grown in biosolid-amended soils. These studies
generally indicate that uptake is stronger into the vegetative parts of plants (stems, leaves) than the
edible part of the plant (fruit, seeds). For example, PFOA and PFOS accumulation factors are higher in
corn silage than in corn grain (Simones et al., 2023). It is hypothesized that PFOA and PFOS accumulation
is higher in shoot or root crops due to an increasing number of biological barriers as the contaminant is
transported from roots to shoots to fruits (Blaine et al., 2014; Lesmeister et al., 2021). However, there
are other large differences between uptake factors in measurements from different types of plants.
Researchers have hypothesized that reasons for these differences may include differences in protein
content, differences in root system types and surface areas, the amount of water transpired, the
presence of precursors in soil, or the soil conditions where the plants were grown (Ghisi et al., 2019;
Lesmeister et al., 2021).
In field studies at locations where biosolids had been applied, there does not appear to be a significant
or consistent correlation between PFOA and PFOS uptake factors and soil concentration, pH, organic
matter content or cation exchange capacity (Simones et al., 2023). The use of contaminated irrigation
water increases uptake of PFOA and PFOS in plants (Gredelj et al., 2020; Blaine et al., 2014). PFAS
precursors commonly found in biosolids, such as diPAPs, result in increased perfluorocarboxylic acid
concentrations in plants, including PFOA concentrations (Lee et al., 2013; Bizkarguenaga et al., 2016b).
More information on the literature search strategy and selected studies used to parameterize PFOA and
PFOS uptake factors for the fate and transport models used in this risk assessment is found in Section
2.9.3.
2.6 Effects on Humans and Aquatic and Terrestrial Biota
2.6.1 Human Health Effects
Biosolids risk assessment can consider human health effects that occur after oral, inhalation, or dermal
exposures. Due to potential differences in toxicity across oral, inhalation, and dermal exposure
pathways, the EPA develops different toxicity values for each pathway.
2.6.1.1 Oral
Based on animal toxicology and human epidemiology studies, oral exposure to either PFOA or PFOS is
associated with numerous adverse health effects, including several types of cancer. Through conducting
a systematic review of the literature, PFOA and PFOS are relatively high potency PFAS, with very low
noncancer reference doses. A detailed description of the health effects observed at various levels of
PFOA or PFOS exposure can be found in EPA's recently published Final Human Health Toxicity
Assessments (US EPA, 2024b;c).
For PFOA, EPA's toxicity assessment concludes that overall, the available evidence indicates that PFOA
exposure is likely to cause hepatic, immunological, cardiovascular, and developmental effects in
humans, given sufficient exposure conditions (e.g., at serum levels in humans as low as 1.1 to 5.2 ng/mL
DRAFT
16
-------
and at doses in animals as low as 0.3 to 1.0 mg/kg/day)(US EPA, 2024b). These judgments are based on
data from epidemiological studies of infants, children, adolescents, pregnant individuals, and non-
pregnant adults, as well as short-term (28-day), subchronic (90-day), developmental (gestational), and
chronic (2-year) oral-exposure studies in rodents. For hepatic effects, the primary support is evidence of
increased alanine transaminase (ALT) levels in humans and coherent evidence of hepatotoxicity in
animals, including increased liver weights and hepatocellular hypertrophy accompanied by necrosis,
inflammation, or increased liver enzyme levels marking liver injury. For immunological effects, the
primary support is evidence of decreased antibody response to vaccination against tetanus, diphtheria
and rubella in children, and evidence of immunotoxicity in rodents, including decreased Immunoglobulin
M response to sheep red blood cells, reduced spleen and thymus weights, changes in immune cell
populations, and decreased splenic and thymic cellularity. For cardiovascular effects, the primary
support is evidence of increased serum lipids levels in human and alterations to lipid homeostasis in
animals. For developmental effects, the primary evidence is decreased birth weight in human infants
and decreased offspring survival, decreased fetal and pup weight, delayed time to eye opening, and
related pre- and post-natal effects in animals.
The PFOA toxicity assessment also concludes, consistent with EPA's Guidelines for Carcinogen Risk
Assessment (US EPA, 2005a), that the weight of the evidence across epidemiological, animal
toxicological, and mechanistic studies indicate PFOA is Likely to Be Carcinogenic to Humans via the oral
route of exposure. Epidemiological studies provided evidence of kidney and testicular cancer in humans
and some evidence of breast cancer in susceptible subpopulations. Chronic oral animal toxicological
studies in Sprague-Dawley rats reported Leydig cell tumors, pancreatic acinar cell tumors, and
hepatocellular tumors. PFOA exposure is associated with multiple key characteristics of carcinogenicity
(Smith, 2016). Available mechanistic data suggest that multiple human relevant modes of action could
be involved in the renal, testicular, pancreatic, and hepatic tumorigenesis associated with PFOA
exposure in humans and animal models.
For PFOS, EPA's Final Toxicity Assessment concludes the available evidence indicates that PFOS exposure
is likely to cause hepatic, immunological, cardiovascular, and developmental effects in humans, given
sufficient exposure conditions (e.g., at serum levels in humans as low as 0.57 to 5.0 ng/mL and at doses
in animals as low as 0.0017 to 0.4 mg/kg/day). These judgments are based on data from epidemiological
studies of infants, children, adolescents, pregnant individuals, and non-pregnant adults, as well as short-
term (28-day), subchronic (90-day), developmental (gestational), and chronic (2-year) oral-exposure
studies in rodents. For hepatic effects, the primary support is evidence of increased ALT levels in humans
and coherent evidence of hepatotoxicity in animals, including increased liver weights and hepatocellular
hypertrophy accompanied by necrosis, inflammation, or increased liver enzyme levels marking liver
injury. For immunological effects, the primary support is decreased antibody response to vaccination
against tetanus, diphtheria, and rubella in children, and evidence of immunotoxicity in rodents,
including decreased plaque forming cell response to sheep red blood cells, extramedullar
hematopoiesis in the spleen, reduced spleen and thymus weights, changes in immune cell populations,
and decreased splenic and thymic cellularity. For cardiovascular effects, the primary support is evidence
of increased serum lipids levels in humans and alterations to lipid homeostasis in animals. For
developmental effects, the primary support is evidence of decreased birth weight in humans and
decreased fetal and maternal weight in animals.
The PFOS Toxicity Assessment also concludes that the weight of evidence across epidemiological and
animal toxicological studies indicates that PFOS is Likely to Be Carcinogenic to Humans via the oral route
of exposure. Epidemiological studies provided evidence of bladder, prostate, liver, kidney, and breast
cancers in humans, although evidence was limited or mixed for some cancer types. Findings from
DRAFT
17
-------
chronic oral animal toxicological studies supported findings from human studies. Bioassays conducted in
rats reported hepatocellular tumors, pancreatic islet cell tumors, and thyroid follicular cell tumors. Some
studies observed multi-site tumorigenesis (liver and pancreas) in male and female rats. PFOS exposure is
associated with multiple key characteristics of carcinogenicity (Smith et al., 2016). Available mechanistic
data suggest that multiple human relevant modes of action could be involved in pancreatic and hepatic
tumorigenesis associated with PFOS exposure in animal models.
These assessments include the derivation of chronic reference doses (RfDs) and cancer slope factors
(CSFs). Chronic RfDs are defined as an estimate (with uncertainty spanning perhaps an order of
magnitude) of a daily oral exposure for a chronic duration (up to a lifetime) to the human population
(including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects
during a lifetime. CSFs are defined as an upper bound, approximating a 95% confidence limit, on the
increased cancer risk from a lifetime oral exposure to an agent. RfDs and CSFs are calculated to be
protective of the most sensitive effects with the strongest supporting evidence (i.e., those occurring in
the lower dose range, also called co-critical effects) relevant to the entire lifespan, including sensitive life
stages such as development and pregnancy. For PFOA, the noncancer co-critical effects include reduced
antibody response to vaccinations in children (diphtheria and tetanus) (Budtz-Jorgensen & Grandjean,
2018); decreased birth weight (Wikstrom et al., 2020); increased serum total cholesterol (Dong et al.,
2019) and the cancer critical effect is increased risk of renal cell carcinoma (Shearer et al., 2021). The
noncancer co-critical effects associated with oral exposure to PFOS include decreased birth weight
(Wikstrom et al., 2020); increased serum total cholesterol (Dong et al., 2019) and the cancer critical
effect is increased incidence of combined hepatocellular adenomas and carcinomas (Thomford, 2002;
Butenhoff et al., 2012).
As at least one of the co-critical effects identified for PFOA and PFOS are a developmental endpoint and
can potentially result from a short-term exposure during critical periods of development (in this case,
exposure during pregnancy and early life). The EPA concludes that the RfDs for PFOA and PFOS are
applicable to both short-term (from 1 to 30 days) and chronic (lifetime) exposure scenarios.
Table 3. Toxicity Values for PFOA
Toxicity Value Type
Value
Critical Effect(s), Critical Study/Studies
RfD (based on
epidemiological data)
3 x 10"8 mg/kg/day
Reduced antibody response to vaccinations in children
(diphtheria and tetanus) (Budtz-Jorgensen & Grandjean,
2018); decreased birth weight (Wikstrom et al., 2019);
increased serum total cholesterol (Dong et al., 2019)
CSF (based on
epidemiological data)
29,300 (mg/kg/day)1
Renal cell carcinoma (RCC) (Shearer et al., 2021)
Table 4. Toxicity Values for PFOS
Toxicity Value Type
Value
Critical Effect(s), Critical Study/Studies
RfD (based on
epidemiological data)
1 x 10"7 mg/kg/day
Decreased birth weight (Wikstrom et al., 2019); increased
serum total cholesterol (Dong et al., 2019)
CSF (based on animal
toxicological data)
39.5 (mg/kg/day)-1
Combined hepatocellular adenomas and carcinomas in female
rats (Thomford, 2002; Butenhoff et al., 2012, 1276144)
2.6.1.2 Inhalation
The EPA has not completed an assessment of health effects caused by inhalation exposure to PFOA and
PFOS. Since an inhalation toxicity value is not available from the EPA or another federal agency, any
modeled volatilization of PFOA or PFOS would lead to inhalation exposures that could not be assessed
for risk. Also, it is not clear that vapor pressure or Henry's law constants are sufficient to model
DRAFT
18
-------
volatilization of PFOA or PFOS from soil. Surfactants adhere to interfaces and parameters like Henry's
law constant are meant for chemicals that reside mainly within water. Furthermore, ionized compounds
are commonly less likely to volatilize rapidly and PFOA and PFOS will mainly be in their ionized phase in
most farm soils due to the chemicals' acid dissociation constant (pKa) values. No measured data was
found to benchmark volatility estimates of PFOA or PFOS from farm soil systems, forested soil systems,
lagoons, or sewage sludge monofills. For these reasons, inhalation of PFOA and PFOS are not included as
pathways for exposure in the biosolids assessments.
2.6.1.3 Dermal
The EPA's Final Toxicity Assessments for PFOA and PFOS include some discussion of the dermal toxicity
and dermal absorption for PFOA and PFOS in humans (US EPA, 2024b;c). ATSDR (2021) also includes
some discussion of dermal toxicity and dermal absorption in their Toxicological Profile for
Perfluoroalkyls. Neither assessment includes the derivation of a hazard value for direct-contact skin
effects or provides a conclusive estimate for dermal absorption rates of PFOA or PFOS. Animal studies of
dermal absorption for PFOA indicate that absorption rates of PFOA are impacted by the pH of the
exposure media, with highly acidic media and mostly protonated PFOA resulting in higher dermal
absorption than less acidic exposure media (ATSDR 2021). There is not expected to be significant dermal
absorption of PFOA or PFOS associated with swimming or bathing in waters at normal environmental pH
(ATSDR 2021). Though there are uncertainties regarding PFOA and PFOS absorption through dermal soil
exposure, at this time there is insufficient information to quantify risk from dermal exposures in the
biosolids assessment.
2.6.2 Ecological Effects
2.6.2.1 Effects on Aquatic Organisms
The EPA published Final Aquatic Life Ambient Water Quality Criteria (AWQC) for PFOA and PFOS in
October, 2024 (US EPA, 2024l;m). These national recommended criteria represent the highest
concentrations of PFOA and PFOS in water that are not expected to pose a risk to the majority {i.e., 95%)
of freshwater genera from acute and chronic exposures.
The EPA's final aquatic life AWQCs for PFOA finds that aquatic ecotoxicity data are readily available for
freshwater fish, aquatic invertebrates, plants, and algae. Section 3 and Section 4 in the Final Aquatic Life
Ambient Water Quality Criteria for Perfluorooctanoic Acid (PFOA) provide study summaries of individual
publicly available aquatic life studies, and Appendix A through Appendix H of that document summarize
the current PFOA aquatic life ecotoxicity data (US EPA, 2024l;m). The mechanisms underpinning the
toxicity of PFOA to aquatic organisms is an active and on-going area of research. Additional research is
still needed from a mechanistic perspective to better understand how the different modes of action
elicit specific biological responses. Molecular disturbance at the cellular and organ level resulting in
effects on reproduction, growth and development at the individual level are associated with the sex-
related endocrine system; thyroid-related endocrine system; and neuronal, lipid, and carbohydrate
metabolic systems (see Ankley et al., 2020 and Lee et al., 2020 for the latest reviews on the subject). The
underlying mechanisms of PFOA toxicity to aquatic animals, and fish in particular, appear to be related
to oxidative stress, apoptosis, thyroid disruption, and development-related gene expression (Lee et al.,
2020). The published research suggests that many of these molecular pathways interact with each other
and could be linked. For example, for several PFAS including PFOA, oxidative stress appears correlated
with effects on egg hatching and larval formation, linking reproductive toxicity, oxidative stress, and
developmental toxicity (Lee et al., 2020). The actual mechanism(s) through which PFAS induce oxidative
stress require additional study, but increased G-oxidation of fatty acids and mitochondrial toxicity are
proposed triggers (Ankley et al., 2020).
DRAFT
19
-------
Of particular importance is that PFOA exposure-related disruption of the sex-related endocrine system
(e.g., androgen and estrogen) at the molecular, tissue, and organ levels appears to have adverse
reproductive outcomes in fish and invertebrates, and likely in both freshwater and saltwater and via
multiple exposure routes, i.e., waterborne and dietary (Lee at al., 2020). The reproductive effects were
observed in the F0, Fi and F2 generations of zebrafish, Danio rerio, in the multi-generational PFOA
exposure reported by Lee et al. (2017). PFOA causes a wide range of adverse effects in aquatic
organisms, including reproductive failure, developmental toxicity, androgen, estrogen and thyroid
hormone disruption, immune system disruption, and neuronal and developmental damage.
The published Final Aquatic Life Ambient Water Quality Criteria finds that PFOS ecotoxicity studies are
readily available for fish, aquatic invertebrates, plants, and algae. Fewer studies are available for
aquatic-dependent birds, reptiles, and mammals; these taxa are not represented in Aquatic Life Ambient
Water Quality Criteria and studies on these taxa were not reviewed in EPA's most recent criteria.
Sections 3 and 4 of the Final Aquatic Life Ambient Water Quality Criteria for Perfluorooctane Sulfonate
(PFOS) provide study summaries of individual, publicly available aquatic life toxicity studies, and
Appendix A through Appendix H of that document summarize current PFOS aquatic life ecotoxicity data
( US EPA, 2024l;m). PFOS is one of the most studied PFAS in the ecotoxicity literature, with reported
adverse effects on survival, growth, and reproduction. However, additional research is needed to better
understand the modes of action of PFOS. Specifically, additional research from a mechanistic
perspective is needed to better understand how the different modes of action elicit specific biological
responses in fish, aquatic invertebrates, and amphibians. Potential effects of PFOS involving multiple
biological pathways are a research challenge for PFOS. Toxicity literature indicate that PFOS causes a
wide range of adverse effects in aquatic organisms, including reproductive effects, developmental
toxicity, and estrogen, androgen and thyroid hormone disruption (see Sections 3 and 4 and Appendices
A.l through H.l; US EPA, 2024l;m). Following exposure to PFOS, molecular level events can perturb
estrogen-, androgen- and thyroid-related endocrine systems, as well as neuronal, lipid, and
carbohydrate metabolic systems and lead to cellular- and organ-level disturbances and ultimately result
in effects on reproduction, growth, and development at the individual organism level (Ankley et al.,
2020; Lee et al., 2020). The mechanisms of PFOS toxicity to fish in particular appear to be related to
oxidative stress, apoptosis, thyroid disruption, and alterations of gene expression during development
(Lee et al., 2020). Notably, PFOS exposure appeared to be related to the disruption of the sex hormone-
related endocrine system at the molecular, tissue, and organ levels, resulting in observed adverse
reproductive outcomes in freshwater and saltwater fish and invertebrates alike. Further, these effects
have been reported after exposure via multiple exposure routes (i.e., waterborne, dietary, maternal; Lee
et al. 2020). And these reproductive effects also appeared to be trans-generational, as observed in a
multi-generational zebrafish (Danio rerio) study by Wang et al. (2011a).
The EPA established the national recommended criteria for PFOA and PFOS to be protective of most
aquatic organisms in the community (i.e., approximately 95 percent of tested aquatic organisms
representing the aquatic community). The criteria are protective of aquatic life designated uses for
freshwaters. The PFOA and PFOS criteria documents contain acute and chronic criteria for freshwaters
(see Table 5). The criteria documents also contain chronic criteria expressed as tissue-based
concentrations to protect aquatic life from PFOA and PFOS bioaccumulation. The chronic freshwater and
chronic tissue criteria are intended to be independently applicable and no one criterion takes primacy.
The criteria reflect the maximum concentrations, with associated frequency and duration specifications,
that would support protection of aquatic life from acute and chronic effects associated with PFOA and
PFOS in freshwaters.
DRAFT
20
-------
Table 5. Freshwater Aquatic Life AWQCs for PFOA and PFOS
Criteria
Component
Acute Water
Column (CMC)1
Chronic Water
Column (CCC)2
Invertebrate
Whole-Body
Fish Whole-
Body
Fish Muscle
PFOA
Magnitude
3.1 mg/L
0.1 mg/L
1.18 mg/kg ww
6.49 mg/kg ww
0.132 mg/kg ww
PFOS
Magnitude
0.071 mg/L
0.00025 mg/L
0.028 mg/kg ww
0.201 mg/kg ww
0.087 mg/kg ww
Duration
1-hour average
4-day average
Instantaneous3
Frequency
Not to be
exceeded more
than once in three
years, on average
Not to be
exceeded more
than once in
three years, on
average
Not to be exceeded
1 Criterion Maximum Concentration.
2 Criterion Continuous Concentration.
3 Tissue data provide instantaneous point measurements that reflect integrative accumulation of PFOA or PFOS over time and
space in aquatic life population(s) at a given site.
2.6.2.2 Effects on Terrestrial Organisms
Plant and terrestrial vertebrate studies are typically focused on mortality, reproduction, development,
or growth effects that would impact a large fraction of the population. Studies on these organisms that
are sub-lethal are less commonly available, especially for plants and terrestrial vertebrates. As a result,
more sensitive adverse endpoints in wildlife may not be observed, even if they do exist. These factors
may lead to hazard values that are higher (indicative of lower toxicity) than studies measuring effects at
the individual organism level.
Plants: There are no existing federal assessments that describe the phytotoxicity of PFOA and PFOS,
though there are several journal publications on the topic. Tests to find the 50% inhibition concentration
(the contaminant concentration that causes 50% of the inhibition effect in organism growth, or IC5o) of
PFOA and PFOS tend to find results ranging from the 10's to 10,000's piM, which are significantly higher
than concentrations typically found in the environment (Li et al., 2022). The phytotoxicity of direct soil
exposure to Brassica chinensis root growth after a 7-day exposure to PFOA and PFOS in six different soils
was evaluated (Zhao et al., 2011). The 50% effect concentration for root elongation (EC5o) values ranged
from 95 mg/kg to >200 mg/kg for PFOS and from 107 mg/kg to 246 mg/kg for PFOA. In a study by
Brignole et al. (2003), the effects of PFOS on the seedling emergence and growth of seven species of
plants was evaluated after a 21-day exposure. Lettuce (Lactuca sativa) was the most sensitive species
tested with a 25% effect concentration (EC25) of 6.79 mg/kg, based on seedling height. The EC25s for the
other six plant species were: 7.51 mg/kg (ryegrass), 11.7 mg/kg (tomato), 12.9 mg/kg (onion), 53.3
mg/kg (alfalfa), 81.6 mg/kg (flax), and 160 mg/kg (soybean), all based on shoot weight.
Invertebrates: Toxicity values for direct soil exposure to earthworms (Eisenia fetida) have been
determined for PFOA and PFOS. The 14-day 50% lethal concentration (LC50) values for earthworms
exposed to a loamy sandy soil spiked with varying concentrations of PFOA and PFOS were determined to
be 811 mg/kg and 541 mg/kg, respectively (Yuan et al., 2017). PFOS toxicity values are also available for
two additional invertebrate species, Folsomia Candida (springtail) and Oppia nitens (oribatid mite)
(Princz et al., 2018). Springtails and oribatid mites were exposed to PFOS in two types of soil: a coarse-
textured sandy loam and fine-textured clay loam. The 25% inhibition concentration (IC25) values, based
on juvenile reproduction, for oribatid mites were 13 mg/kg and 33 mg/kg in the fine and coarse soil,
respectively. For springtails, the IC25s were 74 mg/kg and 185 mg/kg for the fine and coarse soil,
respectively.
DRAFT
21
-------
Birds: To date, a limited number of laboratory studies have been conducted on a small number of bird
species to determine the toxicity of PFAS. The 50% lethal dose (LD5o) values for juvenile mallard ducks
(Anas platyrhynchos) and northern bobwhite quail (Colinus virginianus) fed for five days with PFOS in
their diet were determined to be 150 mg PFOS/kg bw/day and 61 mg PFOS/kg bw/day, respectively
(Newsted et al., 2006). For Japanese Quail (Coturnix japonica) fed PFOS and PFOA in their diet for five
days, the LD50s were 38 mg/kg bw/day and 68 mg/kg bw/day for PFOS and PFOA, respectively (Bursian
et al., 2021). A chronic laboratory study examined the adult health, body and liver weights, feed
consumption, gross morphology and histology of body organs, and reproduction in adult mallard ducks
and bobwhite quail exposed to PFOS in their diet for 21 weeks (Newsted et al., 2007). For bobwhite
quail and mallard ducks exposed to 50 and 100 mg PFOS/kg feed, lethality was observed within five
weeks from the onset of exposure, whereas no effects on survival were observed in the 10 mg PFOS/kg
feed treatment. In the 10 mg PFOS/kg treatment groups, no significant effects were noted in mallard
ducks. However, the lowest observable adverse effect level (LOAEL) was determined to be 10 mg/kg
PFOS in feed based on decreased survivorship of 14-day-old bobwhite quail offspring.
In 2018, ECCC published Federal Environmental Quality Guidelines (FEQGs) for PFOS (ECCC 2018). These
FEQGs are benchmarks for the quality of the ambient environment that are based solely on the
toxicological effects or hazards posed by substances. ECCC identified the quail survivorship study as a
critical study for effects in birds and calculated a bird egg FEQG of 1.9 ug/g ww. The assessment also
notes a field study compared reproductive success in tree swallows from a contaminated urban lake
versus a reference lake (Custer et al., 2012). The authors concluded that PFOS concentrations above
0.15 ng/g egg were detrimental to hatching success; however, the FEQG authors state that this study
could not be considered in FEQG development because of variability in hatch success between the two
field seasons, variations in egg PFOS concentrations within clutches, and concurrent exposure to other
PFAS. More information is needed on adverse impacts of PFOS to birds.
Livestock and game: A recent review paper (Death et al. 2021) summarizes the literature on toxic
effects of PFOA and PFOS in livestock and wild game. Studies measuring the uptake, elimination, and
distribution of PFOA and PFOS in various livestock have not reported adverse effects in the test animals
(Wilson et al. 2020, Vestegren et al. 2013, Numata et al., 2014). Death et al. (2021) similarly finds that
while there are multiple studies identifying PFOA and PFOS occurrence in wild game (ducks, deer, wild
boar), these studies have not identified adverse effects in the game associated with PFOA and PFOS
exposure.
2.6.3 Scoping: Sensitive Receptors and Endpoints
Overall, adverse effects observed in plants, invertebrates, fish, and birds occur at concentrations that
are higher than levels that would be associated with adverse effects in humans; there have been no
studies reporting adverse effects occurring in livestock. Therefore, this draft risk assessment has been
scoped to focus on human health risks. Below is a brief comparison of the inherent toxicity of PFOA and
PFOS to humans versus other organisms.
Soil. As described above in Section 2.6.2.2, adverse effects observed in plant studies occur at soil
concentrations ranging from the 10's to 100's of mg/kg PFOA and PFOS. Similar ranges of effect levels in
soils are found for terrestrial invertebrates, where the effect levels for terrestrial invertebrates are in
the 10's to 100's of mg/kg PFOA or PFOS. Concentrations of PFOA or PFOS in soil that are protective of
human health through incidental ingestion are expected to be significantly lower than the effect levels
observed for plants and terrestrial invertebrates. For example, risk-based thresholds for PFOA and PFOS
in soils protecting against non-cancer effects in children are on the order of 0.001-0.010 mg/kg PFOA or
PFOS in soil (US EPA, 2024i).
DRAFT
22
-------
Surface water. The thresholds established in EPA's Final Aquatic Life AWQCs protective offish are higher
(less stringent) than would be expected to be protective of human consumers of home-caught fish. The
EPA is developing national recommended human health criteria for PFOA and PFOS, based on the
agency's final toxicity assessments (US EPA 2024b,c) which would take into account exposures via
drinking water, fish consumption, and other sources (e.g., other dietary sources, consumer products,
etc.). The most stringent EPA national recommended (chronic) aquatic life criteria for PFOA and PFOS
are 1.0 x 10 1 mg/L (100 ug/Lj for PFOA and 2.5 x 10"4 mg/L (0.25 ug/Lj for PFOS; State surface water
standards to protect human health due to fish consumption have established values that are lower than
the EPA's Aquatic Life AWQC for PFOA and PFOS. For example, the state of Minnesota has established
surface water criteria protective of non-cancer effects in fish consumers that are 8.8 x 10"5 mg/L (88
ng/Lj for PFOA (MPCA, 2023b) and 5 x 10"8 mg/L (0.05 ng/L) for PFOS (MPCA, 2020). Fish tissue
thresholds protective against cancer effects in human fish consumers would be lower (more protective)
than those developed in Minnesota. Further, some surface waters are used as a source of drinking
water. Risk-based thresholds for PFOA and PFOS in drinking water are also lower than the aquatic life
criteria (less than 5 x 10 s mg/L, 5 ng/L, for PFOA and PFOS, US EPA 2024b;c).
These trends of human health-based thresholds being more stringent than ecologically protective
thresholds are evident due to the extremely potent nature of PFOA and PFOS toxicity in humans. More
study of PFOS and PFOA effects in wildlife could result in a narrowing of the gap between levels
protective of ecological endpoints and levels protective of human health. However, based on currently
available data, the EPA is focusing on human health endpoints for the biosolids draft risk assessment,
with the understanding that establishing practices protective of human health will also offer protection
to aquatic life, terrestrial wildlife, and livestock health.
If future studies indicate ecological toxicity of PFOA or PFOS at lower doses/concentrations (e.g., for
terrestrial organisms), the EPA may conduct further ecological risk assessment, as warranted. This
human health focused draft risk assessment for PFOA and PFOS does not preclude any future biosolids-
related unacceptable risk finding for aquatic life, terrestrial wildlife, or livestock.
2.7 Exposure Pathways for Humans and Aquatic and Terrestrial Biota
As described in Section 1.3, sewage sludge can be disposed of via solid waste landfill, surface disposal at
a dedicated sewage sludge disposal site, or incineration, or it can be land applied as a soil amendment to
a variety of sites (agricultural fields, public access areas, road construction, landfill cover material, soil
material in remediation efforts, and more). These disposal and land application options all result in
potential pathways for PFOA and PFOS exposure to humans and wildlife, such as drinking water
consumption, dietary intake, soil ingestion, and inhalation of particulate-bound contaminants.
Currently, there is insufficient information available to model occupational exposures for workers that
repeatedly apply biosolids at different farms throughout the year or to determine whether the farm
family or farm worker exposures will exceed the exposures of these professional biosolids applicators. If
this type of worker is repeatedly spray applying biosolids on farm fields, that could lead to airborne
exposures over many days of the year and this type of exposure is not represented within the modeled
pathways for the farm family. The EPA also does not currently have survey or other data to estimate the
behavior patterns of these types of workers with missing information including amount of biosolids
mass aerosolized during application, time spent per day applying biosolids, and number of days worked
per year. As mentioned above, there is not currently a reference concentration (RfC) or inhalation unit
risk (IUR) available for PFOA or PFOS, so the risks this type of worker may face cannot be assessed due
to an absence of exposure and toxicity values. For these reasons, this draft risk assessment does not
include receptors of professional biosolids land appliers in the conceptual models.
DRAFT
23
-------
2.7.1 Considera tions of Aggrega te Exposures
Aggregate exposure and risk assessment involve the analysis of exposure to a single chemical by
multiple pathways and routes of exposure. This draft risk assessment does not aggregate exposure and
risk, and instead presents estimated exposure and risk for each individual exposure pathway {i.e.,
consuming fish, drinking water, incidentally ingesting soil). This approach does not account for exposure
from multiple modeled pathways simultaneously, pathways that were not modeled due to data gaps
(including inhalation and dermal exposure pathways) or exposure pathways not related to sewage
sludge use and disposal (such as exposure from use of personal care products, cleaning supplies,
household dust, etc.). This decision to assess each pathway individually allows modeling results to be
interpreted as risk contributed from sewage sludge for each pathway across a variety of sewage sludge
use and disposal scenarios.
Assessing individual pathways also allows risk assessors to consider a variety of potential receptors who
may have exposure from some, but not all of the potentially relevant exposure pathways. However, in
each given scenario, a receptor may be exposed from multiple pathways at the same time and from
pathways not modeled in this risk assessment. For example, farmers who consume animal products
produced on the farm likely also consume drinking water sourced locally as many rural areas of the
country rely on groundwater as a source of drinking water.
2.7.2 Considerations of Cumulative Exposures
Cumulative exposure and risk assessment involve analysis of exposures from multiple stressors that
occur simultaneously. A receptor may be exposed to both PFOA and PFOS at the same time. PFOA and
PFOS have been shown to be dose additive (US EPA, 2024e) and are nearly always found in mixtures in
sewage sludge. It follows that the environmental media impacted by use or disposal of sewage sludge
also contains mixtures of PFOA and PFOS. The presence of mixtures and multiple pathways for exposure
will result in higher risks of adverse health effects at a population scale than are reflected in the
pathway-specific results. This draft risk assessment presents exposures and risks (hazard quotients and
cancer risk levels) associated with single chemicals (PFOA or PFOS) to provide information about which
compound is contributing most significantly to exposure and risk in each pathway. Though this draft risk
assessment is scoped narrowly to PFOA and PFOS, other PFAS are also known to be present in biosolids
(see Section 2.4), and the EPA may consider additional PFAS for risk assessment in the future.
2.8 Conceptual Models
There are a multitude of potential unique strategies and hydrogeochemical settings for biosolids
disposal and reuse across the U.S. It is not feasible to model or assess each of these environmental
release scenarios individually. Instead, the EPA has strategically selected a discrete number of common
reuse or disposal scenarios to model in detail and will use the findings from these detailed modeling
exercises to qualitatively describe other relevant scenarios. These detailed modeling scenarios were
selected because 1) they are commonly used for sewage sludge, biosolids, or septage in the U.S., 2) they
are likely to result in higher exposures for humans, or 3) they include numerous pathways that are
applicable to other reuse or disposal scenarios. In some cases, such as biosolids incineration and other
uses of biosolids in land application (silviculture, golf courses, etc.), there are data limitations that
restrict our ability to quantitatively assess exposure outcomes.
Four detailed modeling scenarios are described in this document: reuse on a farm growing fruits and
vegetables (crop farm scenario), reuse on a farm raising livestock (pasture farm scenario), disposal in a
surface disposal site (surface disposal scenario), and reuse to restore over-grazed pastureland (land
reclamation scenario). These detailed models are used quantitatively to estimate exposure and describe
potential risks to human receptors in each scenario. These models are also used to qualitatively estimate
DRAFT
24
-------
relative exposures and risks associated with other types of land application like use in silviculture or
application to golf courses, other types of land reclamation like mine reclamation or road construction,
and incineration. For each of the modeling scenarios, the EPA conducts modeling runs parameterized for
hypothetical regions in a wet climate, a dry climate, and a moderate climate. These region-specific
meteorological conditions, soil conditions, and hydrologic conditions are described in Section 2.9.3.
These models are not intended to characterize conditions at any specific site.
The following sections illustrate the conceptual models for PFAS application, transport, uptake, and
exposure in each disposal and reuse scenario; additional information on the computational models used
and the parameterization of those models can be found in Section 2.9 and Appendices B and C. As
described in Section 2.6.1, dermal exposures are not expected to meaningfully contribute to overall
exposure, and dermal exposure pathways are not included in the conceptual models for this risk
assessment. Inhalation is not expected to be a significant source of exposure for these scenarios and
there are no inhalation toxicity values (RfCs or lURs) available for PFOA and PFOS; for these reasons,
inhalation pathways are also not included in the conceptual models. Finally, data available to date
indicate that PFOA and PFOS are significantly more toxic to humans than wildlife or livestock, such that
actions taken to protect human health will also protect wildlife and livestock health. The following
conceptual models therefore only include exposure pathways relevant to humans.
2.8.1 Farms
Two types of farming scenarios are included in this assessment: a farm growing fruits and vegetables
(the crop farm) and a farm raising animals (the pasture farm).
2.8.1.1 Crop Farm Scenario
The crop farming scenario is designed to capture relevant human exposure pathways for PFOA and PFOS
following biosolids land application to fields used to grow human food. Figure 2 provides a schematic
visualization of the crop farming scenario. Figure 3 presents the conceptual model for the crop farm,
showing the different pathways evaluated.
Previous biosolids assessments (US EPA 1992, 1995a, 2003a) have assessed this scenario, and the
original exposure pathway numbers from the 1993 assessment {i.e., the one conducted to support the
1993 regulation, US EPA 1992) are included in Figure 3 for reference. While some states have regulated
the application of biosolids to fields used to grow human food, this practice is not regulated in other
states. Furthermore, because of the extreme persistence of PFOA and PFOS in soils, a property with
previous biosolids land application that has been repurposed as a farm for human food could still have
multiple relevant human exposure pathways. Class A biosolids currently have no restrictions on crop
type or harvesting delay restrictions for agricultural applications. Finally, Class A exceptional quality
(Class AEq) biosolids, which can be used by home or hobby gardeners, have no restrictions on their
application rates or use to grow food for human consumption, though they do have some restrictions on
maximum concentrations of some metals.
DRAFT
25
-------
\U.t i-
Reservoir
Figure 2. Conceptual visual depiction of crop farming scenario.
DRAFT
26
-------
Source Release Mechanism Media
Exposure Scenarios
Exposure Routes*
Receptors**
* Pathway numbers refer to those in the original 1993 biosolids risk assessment. No pathway number indicates a pathway that has been added
that original assessment.
** Receptor populations may include farm families, home gardeners, CSA participants, or nearby residents.
Figure 3. Crop farm conceptual model
DRAFT
27
-------
The crop farm model can be applied to several scenarios of biosolids use. First is the model application
of biosolids to commercial crop farm or hobby/subsistence farm, where a family lives adjacent to the
land used for crop cultivation. In this scenario, adults and children on the farm could have exposure
through consuming crops grown on the field, drinking water, and incidental soil ingestion. The crop farm
model also includes pathways that could be relevant to neighbors, those supporting the farm through
CSA arrangements, or those purchasing food at the family's farm stand. Finally, by evaluating exposure
with the non-limited application rates allowed for Class AEq biosolids, this scenario captures potential
impacts to the home gardener from applying biosolids at their personal or community gardens.
The crop farm scenario is important to model quantitatively because it includes receptors that are likely
to have higher exposure rates than receptors in other scenarios. For example, a self-sufficient farm
family that spends most of the year immediately adjacent to the farm is assumed to be exposed to
multiple transport pathways (drinking water, soil, fish, other food) and engage in behaviors that lead to
them having higher exposures than the general population (i.e., consuming a high portion of their total
produce intake from a single potentially contaminated farm). Community members that purchase large
amounts of produce from the farm via CSA or frequent farm stand purchases will also have more of their
dietary intake from a single, potentially contaminated source than the general population. A risk
assessment of these pathways is therefore also protective of produce consumers in the general U.S.
population.
The potential impacts from application of contaminated biosolids to a particular farm site (e.g., field)
can have broader implications to the farm's neighbors and the larger community. The use of the farm
family as a surrogate to represent other populations means that an assessment of the potentially
impacted populations from the land application of biosolids should not be limited to self-sufficient
farmers. For example, a farm's neighbors or an entire community might rely on the same drinking water
source as the farm family.
After land application of contaminated biosolids, there are multiple potential human exposure
pathways. Once biosolids have been applied, PFOA and PFOS will either stay in the soil column of the
farm field or garden, move with windblown soil particles, infiltrate through the soil column into
groundwater, or mobilize in the particulate or sorbed phase through runoff and erosion into surface
water. PFOA and PFOS in the soil on the farm field can be taken up into the edible or non-edible portion
of crops. PFOA and PFOS that infiltrate into groundwater will infiltrate to the uppermost aquifer and
then flow downgradient with groundwater, where they could end up in well water used for human
drinking water. The chemicals transported to a nearby lake or reservoir could be in drinking water or be
taken up into edible fish tissues. The PFOA and PFOS in soils is available for child or adult incidental soil
ingestion.
This draft risk assessment will focus on potential exposures that result from drinking water ingestion,
dietary ingestion, and incidental soil ingestion. Exposure from drinking water ingestion could result from
contamination of groundwater following leaching of PFOA and PFOS through soil and from
contamination of surface water following erosion and runoff. Exposure from dietary ingestion could
include consumption offish and produce (fruits and vegetables). Soil ingestion exposures are based on
incidental soil ingestion values for children from soil on the farm field or gardening area.
The exposure model estimates the most significant transport pathways for chemicals in biosolids, but
some less significant pathways are not included. For example, in some scenarios, farmers may use
groundwater or surface water that is contaminated by PFOA and/or PFOS as irrigation water, which
could result in additional crop uptake of these chemicals and thus potential human exposure.
Additionally, the model assumes that the farm field has no PFOA or PFOS present in soils (e.g., via
atmospheric deposition) prior to the application of biosolids.
DRAFT
28
-------
2.8.1,2 Pasture Farm Scenario
The pasture farm scenario is similar to the crop farm scenario but models a farm that only raises animals
(cows and chickens) and crops used for livestock feed, rather than fruit and vegetable crops for human
consumption. Figure 4 provides an illustrative visualization of the pasture farming scenario. Figure 5
presents the conceptual model for the pasture farm, showing the different pathways evaluated.
This scenario has also been considered in prior risk assessments (see pathway numbers in Figure 5).
While some states restrict the use of some biosolids on fields used to grow food for human
consumption, there are very few states or other jurisdictions that restrict the use of biosolids for fields
used to grow feed for animals or fields used for animal grazing (ECOS, 2023). Furthermore, because of
the extreme persistence of PFOA and PFOS in soils, a property with previous biosolids land application
that has been repurposed as a pasture for animal grazing or field for growing feed would stili have
multiple potential human exposure pathways available.
Reservoir
Figure 4. Conceptual visualization of pasture farm scenario.
DRAFT
29
-------
Source Release Mechanism Media Exposure Scenarios Exposure Routes* Receptors**
* Pathway numbers refer to those in the original 1993 biosolids risk assessment. No pathway number indicates a pathway that has been added
since that original assessment. Some pathways have been modified or combined; here, pathways 4 (cattle eating plants) and 5 (cattle eating
soil) have been combined, and consumption of chicken and eggs has been added.
** Receptor populations may include farm families, home gardeners, CSA participants, or nearby residents.
Green boxes and lines represent farm pathways that may not be present at all reclamation sites.
Figure 5. Pasture farm conceptual model.
DRAFT
30
-------
The pasture farm model can be applied to several scenarios of biosolids use, similar to the crop farm
model. First is the application of biosolids to a commercial or hobby/subsistence farm, where a family
lives adjacent to the land used for grazing cows, raising chickens, or growing feed for these animals. In
this scenario, adults and children on the farm could have exposure through meat, dairy, or egg products
they produce, and incidental soil ingestion in the yard or land near their home. Similar to the crop farm,
the pasture farm also includes pathways that could be relevant to neighbors and the larger community,
e.g., those sharing a drinking water supply, supporting the farm through CSAs or those frequently
purchasing meat, milk, or dairy at a local market or farm stand.
The pasture farm scenario is also important to model quantitatively because it includes receptors that
are likely to have higher exposure rates. For example, a self-sufficient farm family that spends most of
the year living on the farm may be exposed to multiple transport pathways (drinking water, soil, fish,
food) and may engage in behaviors that lead to them having higher exposures than the general
population (i.e., consuming a high portion of their total meat, milk, and egg intake from a single source).
Farm neighbors or an entire community could use the same drinking water source as the farm family.
Community members that purchase large amounts of food from the farm via CSAs and frequent market
or farm stand purchases will also have more of their dietary intake from a single, potentially
contaminated source, potentially resulting in higher exposures than the general population. A risk
assessment of these pathways is therefore also protective of meat, milk, and egg consumers in the
general U.S. population.
After land application of biosolids, there are multiple pathways that could cause exposures to humans in
the pasture farm model. Unlike the crop farm scenario, in the pasture farm scenario, it is not assumed
that biosolids are tilled into the soil. For this reason, once biosolids have been applied, more PFOA and
PFOS will be available to move with windblown soil particles or mobilize in the particulate or sorbed
phase through runoff and erosion into surface water. The PFOA and PFOS that remain in the soil on the
farm field could be taken up into the grass or hay used for animal feed or grazing. In the pasture farm
scenario, groundwater and surface water can be used by humans and livestock as a drinking water
source. The soil on the field can be consumed by animals foraging or grazing. Other potential pathways
relevant to the crop farm scenario (human ingestion of fish and soils) are also relevant to the pasture
farm scenario.
2.8.2 Land Red a ma tion
One known use of biosolids in the U.S. is for the purpose of increasing the organic matter content in
fields that have been over-grazed and have degraded soil quality. Biosolids have been used in these
settings as a beneficial soil amendment and may be applied at rates higher than those allowed under
traditional agricultural settings. For the purposes of this draft risk assessment, the EPA has modeled the
land reclamation scenario using the same conceptual model used for the pasture farm (Figure 5).
Though the pathways relevant to the reclamation scenario are the same as those relevant to the pasture
farm scenario, the rate of biosolids land application in the reclamation scenario is modeled as higher
than the land application rate used for the pasture farm. Additionally, in the reclamation scenario there
is only one application of biosolids, instead of ongoing annual applications modeled in the pasture
scenario. However, other than the differences in biosolids land application rate and timing, the potential
human exposure pathways in this scenario are the same as those in the pasture farm scenario. If a site is
being remediated in this fashion to improve soil quality, but is not then used as a pasture, the same
conceptual model applies, except that the pathways related to livestock are not relevant.
Land reclamation can take many forms, and no two land reclamation projects are exactly alike. Biosolids
have been used in a wide variety of land reclamation settings such as remediating closed mines,
remediating soils at clean-up sites with industrial pollution, or amending soils disturbed by new
DRAFT
31
-------
construction. These different types of sites could have a variety of hydrologic, geologic, and geochemical
conditions than influence the fate and transport of PFAS. These sites could also have a number of
different potential exposure pathways for human exposure. Previous risk assessments have focused on
mine reclamation and over-grazed farmland because these activities were judged to be most sensitive
reclamation scenarios for the chemicals being modeled (US EPA, 1992; US EPA, 2003a). The EPA is
selecting a reclamation scenario of remediating over-grazed farmland for PFOA and PFOS because the
accumulation of these chemicals into livestock is likely to represent a higher human health risk scenario
for a farmland reclamation relative to other reclamation activities. This scenario also includes some
potential pathways (such as the soil to groundwater to drinking water pathway) that are widely
applicable across many potential land reclamation settings.
2.8.3 Surface Disposal
Surface disposal is the placement of sewage sludge in an active sewage sludge unit for final disposal, not
for treatment, storage, or to condition the soil or fertilize crops grown in the soil. The surface disposal
scenario is designed to capture potential human exposure pathways for PFOA and PFOS that are
available after sewage sludge is placed in a surface disposal site. Figure 6 presents the conceptual model
for the surface disposal scenario, showing the pathway evaluated.
DRAFT
32
-------
Source Release Mechanism Media Exposure Scenarios Exposure Routes* Receptors**
* Pathway numbers refer to those in the original 1993 biosolids risk assessment.
** Receptor population is nearby residents.
Figure 6. Conceptual model for disposal in a surface disposal site.
DRAFT
33
-------
The characteristics of surface disposal sites are varied. In some cases, sewage sludge is dewatered and
disposed of in a sewage sludge-only landfill (known as a monofill) which can be lined or unlined. In other
cases, the sewage sludge is not dewatered before disposal. These surface disposal sites can also be lined
or unlined. This draft risk assessment models the potential impacts of PFOA and PFOS migration at lined
and unlined surface disposal sites. Surface disposal of sewage sludge that has not been dewatered
represents the scenario with the greatest potential for environmental releases via leaching and
infiltration, so this specific scenario is modeled in the risk assessment.
Sewage sludge may also be sent to a lagoon. The EPA considers lagoons to be waste stabilization ponds
or basins designed and built to reduce organic content, suspended solids, and pathogens in wastewater
and sewage sludge. They can be lined or unlined. From a groundwater infiltration perspective, these
lagoons are not dissimilar from sewage sludge surface disposal sites accepting materials that have not
been dewatered. Though lagoons are a treatment technology, not a disposal method, the modeling
exercises in this risk assessment can also be used to qualitatively understand potential infiltration risks
at some lagoons.
MSW landfills also receive sewage sludge for disposal along with many other waste streams, but those
facilities are outside the scope of CWA section 405 and will not be assessed here as they fall under the
regulations of Resource Conservation and Recovery Act (RCRA) Subtitle D. Similarly, the use of sewage
sludge as daily cover on MSW landfills is also relevant to RCRA. The EPA has provided information on
MSW landfills for the disposal of sewage sludge and other PFAS-containing wastes in its Interim
Guidance on Destroying and Disposing of Certain PFAS and PFAS-Containing Materials That Are Not
Consumer Products (US EPA, 2024g).
2.8.4 Incineration
The incineration model (Figure 7) illustrates PFOA and PFOS exposure pathways that are possible after
sewage sludge is incinerated in an SSI. Contaminant levels for sewage sludge entering an SSI are
regulated by the CWA via part 503, and emissions from SSIs are regulated under the CAA (US EPA,
2023b), but the CAA regulations do not currently include any requirements related to PFOA or PFOS.
SSIs are devices used for the combustion of dewatered sewage sludge. In the U.S., the two main types of
SSIs include multi-hearth furnaces and fluidized bed combustors (US EPA, 2016). In a multi-hearth
furnace, the sludge is typically dried at temperatures ranging from 425°C to 760°C (800°F to 1,400°F) (US
EPA, 1995b). The combustion of the sewage sludge is performed as the temperature is increased from
815°C to 925°C (1500°F to 1,700°F) (US EPA, 1995b). The gas residence times are usually four to five
seconds (US EPA, 1995b). Emission controls can consist of wet scrubbers, wet electrostatic precipitators,
afterburners, and regenerative thermal oxidizers (US EPA, 1995b). In a fluidized bed combustor, the
sludge is typically combusted at temperatures ranging from 750°C to 925°C (1,400°F to 1,700°F) (US EPA,
1995b). The gas residence times are usually two to five seconds (US EPA, 1995b). Emission controls can
consist of venturi scrubbers, multicyclones, fabric filters, activated carbon injection, and carbon bed
absorbers (US EPA, 1995b).
SSI unit design and operation can vary widely across the nation. Current SSI standard operating
conditions may not be effective for the treatment of PFAS. There is a concern with PFAS being emitted
as products of incomplete combustion (PICs). A recent study performed on aqueous film forming foam
(AFFF) showed that temperatures above 1100°C were necessary to promote PFAS destruction and
minimize PICs (Shields et a I., 2023). Additionally, longer residence times are recommended coupled with
the use of high-temperature thermal oxidizers to reduce emissions. While this research was performed
on a liquid-phase material and more research is still needed on semi-solid and solid-phase matrices,
these findings indicate that current temperatures used for SSIs may not be high enough and the gas
DRAFT
34
-------
residence times may not be long enough to completely destroy PFAS. Furthermore, an additional recent
study monitored PFAS fate from both a multi-hearth furnace and a fluidized bed combustor (Winchell et
al., 2024). The PFAS testing found that the stack emissions from the multi-hearth furnace contained
reportable levels of all targeted PFAS measured, representing an average of 5% of the total targeted
PFAS monitored in the feed per sample run with emissions consisting mainly of shorter-chain PFAS
(Winchell et al., 2024). Moreover, for both the multi-hearth furnace and fluidized bed combustor,
nonpolar fluorinated organics were detected in the wet scrubber water streams, which were sourced
from treated wastewater effluent (Winchell et al., 2024). Additional testing is still needed comparing
more units of multi-hearth furnaces and fluidized bed combustors, while also using newly released air
methods (e.g., OTM-50) to test SSI emissions for more volatile PICs (US EPA, 2024f). Consequently, one
issue is that volatile PFAS released as PICs may be inhaled by populations near the SSI and PICs could
have the potential to transform and degrade into more persistent PFAS (e.g., PFOA and PFOS), which
can be distributed through atmospheric deposition to soil and water.
Due to these uncertainties around PFOA and PFOS destruction in SSIs, the potential for PICs to be
released that degrade to PFOA and PFOS, and other uncertainties around thermal destruction
conditions, the SSI model will not be quantitatively modeled for this draft risk assessment. However, the
conceptual model in Figure 7 illustrates which pathways may be available for PFOA and PFOS exposure
after sewage sludge is incinerated. This conceptual model focuses on the deposition of PFOA and PFOS
to soil or surface water bodies because of the absence of an inhalation hazard value for PFOA and PFOS.
Once PFOA and PFOS are deposited on the soil surface or water surface, many of the same potential
exposure pathways are available that were described in the prior conceptual models, including exposure
through fish consumption, soil ingestion, food intake, and drinking water intake. The risk assessment will
qualitatively discuss the potential for risk in these pathways in the SSI scenario. The EPA has provided
information on incineration in its Interim Guidance on Destroying and Disposing of Certain PFAS and
PFAS-Containing Materials That Are Not Consumer Products (US EPA, 2024g).
DRAFT
35
-------
Source
Release Mechanisms
Media
Exposure Scenarios Exposure Routes Receptors*
a negligible pathway compared to soil uptake and is not shown.
* Receptor populations may include farm families, home gardeners, CSA participants, or nearby residents.
Figure 7. Conceptual model for sewage sludge incineration.
DRAFT
36
-------
2.8.5 Other Land Application Scenarios
Biosolids land application can occur at many types of sites with low or high public contact, including
forests, tree farms, road construction sites, golf courses, and more. A generic model for land application
sites with low public contact would include potential pathways like leaching to groundwater and runoff
to surface water, but it would not include pathways like ingestion of soil. The conceptual model in Figure
8 illustrates which pathways may be available for PFOA and PFOS exposure in other land application
scenarios.
DRAFT
37
-------
Source Release Mechanism Media
* Receptor population is nearby residents.
Figure 8. Conceptual model for other land application scenarios.
DRAFT
Exposure Scenarios Exposure Routes Receptors*
Ingestion of
Adult, child
drinking water
Ingestion of
drinking water
Ingestion of fish
38
-------
The EPA has not modeled biosolids fate and transport in forest settings or other sites with low public
contact in previous assessments (US EPA, 1995a; US EPA, 2003a). There is limited information available
about the biosolids application rates in other types of land application related to roadway development,
forestry, and others. Further, there are no pre-existing models for the fate and transport of biosolids
applied to forests, tree farms, or other applicable sites, and studies to parameterize new models specific
to biosolids fate and transport in silviculture are limited. Therefore, the EPA will assess these pathways
qualitatively.
Biosolids that are applied at golf courses, parks, playgrounds, schools and homes may be Class AEq,
meaning that the Class A pathogen requirements and the stricter requirements for chemicals in part 503
must both be met. Class AEq biosolids can be sold directly to the public, e.g., at hardware stores, without
any further requirements on the method, rate, or location of application. Restrictions that apply to other
classes of biosolids that may reduce exposures do not apply to Class AEq biosolids. Since they also may
be applied to residential locations, the potential for incidental soil ingestion by children becomes a more
significant concern given the likelihood that a larger number of children may be repeatedly exposed at
these types of sites.
The EPA's Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) program
generates regional screening levels (RSLs) for residential soil for CERCLA hazardous substances based on
the RfD for a chemical and a high-end incidental soil ingestion rate for children. PFOA and PFOS were
added to the CERCLA hazardous substance list in May 2024, and the EPA developed screening values
(1.9e-5 mg/kg PFOA, equivalent to 0.019 ppb; 6.3e-3 mg/kg PFOS, equivalent to 6.3 ppb, from US EPA
2024i), as starting points for determining if a chemical needs to be considered in a Superfund site's
remediation plan. The incidental soil ingestion exposure pathway evaluated for CERCLA screening values
is relevant to Class AEq biosolids that are land applied in places like parks, playgrounds, schools, and
homes.
Finally, domestic septage is sometimes managed through land application to agricultural sites or other
sites with low potential for human exposure (i.e., turf farms, forested lands, and reclamation sites).
Record keeping by the appliers is required for domestic septage land application for five years after the
application, but these records are not required to be reported to the permitting authority. As a result,
the EPA has limited data available on the types of lands used for domestic septage land application or
the rates of application used. For more information on domestic septage application to grow crops, see
the EPA's 2024 factsheets on Requirements for Application of Domestic Septage to Agricultural Land
(US EPA, 2024j;k).
2.9 Analysis Plan
2.9.1 Modeling Plan
The CWA requires the EPA to ensure that the reuse of biosolids and disposal of sewage sludge does not
adversely affect public health or the environment. To achieve this goal, the EPA conducts human health
and ecological risk assessments for contaminants known or expected to be in biosolids (US EPA, 1993;
2003a; 2023c). In such risk assessments, the EPA conducts a series of modeling exercises with increasing
refinement to estimate and characterize potential risks posed by activities associated with biosolids
disposal or reuse (US EPA, 2023c). If the potential for risk exists from a chemical/contaminant in
biosolids, the EPA typically determines the concentration of that chemical in biosolids that interfere with
each use or disposal practice. The following sections describe the overall modeling approach that the
EPA is taking for PFOA and PFOS in this draft risk assessment.
DRAFT
39
-------
2.9.1.1 High End Deterministic
The EPA first screens chemicals that have been detected in biosolids using a high-end deterministic
model for pasture and crop farming scenarios called the Biosolids Tool (BST; US EPA, 2023d). By using a
high-end deterministic approach, the EPA ensures that its initial risk screening is conservative (health
protective) in several ways. First, the screening tool uses modeling scenarios (crop and pasture farm)
that generally result in the higher potential exposure rates than other biosolids reuse or disposal
options. Second, the exposure modeling in this screening tool assumes high-end (95th percentile)
consumption rates for drinking water, fish ingestion, produce consumption, and milk and meat
consumption. Third, the exposure modeling assumes that the biosolids concentrations applied to the
farm are at the 95th percentile of concentrations that have been measured in U.S. biosolids. The high-
end deterministic model outputs estimated daily doses (mg/kg-day) that receptors are exposed to
through each pathway in the model. These estimated exposures are then assessed individually against
the available toxicity values. EPA used the BST to screen PFOA and PFOS, finding that every pathway
modeled indicated that this scenario could result in excess risk (a summary of the BST inputs and
screening results for PFOA and PFOS can be found in Appendix E). These findings motivated EPA to
further assess the fate and transport of these compounds in various biosolids use and disposal
scenarios. This high-end deterministic assessment approach is similar to the approach used in prior
sewage sludge risk assessments (US EPA, 1992; 1995a) which focused on identifying risks to someone
with a "reasonable maximum exposure."
2.9.1.2 Central Tendency Deterministic
Given the results of high-end deterministic modeling for PFOA and PFOS in the farming scenarios (see
Appendix E), the EPA decided to assess risks under median conditions rather than high-end conditions,
to better understand the potential scope and magnitude of potential risks under different use and
disposal scenarios. Given that all sewage sludge requires some type of disposal or reuse management
activity, it is also important to understand risks from biosolids used in the farm (crop, pasture) scenarios
in the context of other use and disposal scenarios, such as land reclamation, silviculture, surface disposal
or incineration. Completing a central tendency deterministic modeling exercise for multiple reuse and
disposal options provides an understanding of exposure risks associated with biosolids at conditions that
approximate average conditions for each use scenario. This intermediate step between high-end
deterministic screening and refined probabilistic risk assessment can help inform which scenarios, if any,
should be the focus of more refined risk modeling (i.e., deriving risk-based values protective of the 95th
percentile exposure scenario using Monte Carlo analysis).
To complete the central tendency deterministic modeling steps, the EPA 1) assessed available fate and
transport models to ensure that they are the best available models for assessing PFOA and PFOS and 2)
parameterized the modeling inputs to reflect an overall set of median U.S. conditions. Section 2.9.2
discusses the model selection process for refining the PFOA and PFOS fate and transport modeling.
Section 2.9.3 discusses the input parameters used for this modeling approach. At a high level, the input
parameters for this central tendency deterministic modeling exercise represent less health protective
assumptions than the EPA would typically use in a risk assessment for biosolids or other environmental
media. For example, the EPA is assuming that the drinking water intake rate is about 1 L/day for an 80-
kg adult, compared to the 90th percentile value of 2.4 L/day that is typically used for CWA purposes.
2.9.1.3 Probabilistic (Monte Carlo Analysis)
Monte Carlo simulation is a statistical technique by which a quantity is calculated repeatedly, using
randomly selected values from assigned distributions for each calculation. These results approximate
the range of possible outcomes and the likelihood of each. When Monte Carlo simulation is applied to
risk assessment, risk appears as a frequency distribution rather than a single value, which allows for the
DRAFT
40
-------
identification of risks at specific percentiles. Previous sewage sludge risk assessment has used a Monte
Carlo probabilistic modeling approach and targeted risk-based thresholds protective of 95% of the
modeled population (US EPA 2003a). This draft risk assessment does not include Monte Carlo Analysis
because the central tendency deterministic modeling indicates that risks are prevalent even when
targeting median (50th percentile) conditions in individual exposure pathways. See section 4.9 for more
discussion on the EPA's rationale for not conducting Monte Carlo Analysis in this assessment.
2.9.2 Model Selection
This draft risk assessment relies on several independent models to understand PFOA and PFOS fate and
transport across the exposure scenarios (crop farm, pasture farm, reclamation site, and surface disposal
site). In the farming and reclamation scenarios, the first step is to model how much PFOA and PFOS sorb
to soil, are moved through runoff and erosion, and move through the unsaturated zone and saturated
zone into groundwater after biosolids have been land-applied to soils. A separate model estimates the
amount of PFOA and PFOS moving through runoff and erosion; this model then estimates the
concentrations of PFOA and PFOS that enter a nearby lake or reservoir. Finally, a third model estimates
the amount of PFOA and PFOS moving through groundwater to nearby drinking water wells. In the
surface disposal scenario, a model is used to estimate how much PFOA and PFOS may leach through the
underlying soil from a lined or unlined surface disposal site. Then the same groundwater model used in
the farming and reclamation scenarios is used to understand how leached PFOA and PFOS move through
groundwater to neighboring groundwater wells. The following sections describe how and why EPA
selected various models for this assessment.
The results of the fate and transport modeling include concentrations of PFOA and PFOS in
environmental media over time, such as soil concentrations on the farm field, soil concentrations on
nearby "buffer" land, surface water concentrations in the nearby lake or reservoir, and groundwater
concentrations at wells with given depths and distances from the field. These media concentration
results are then used to calculate concentrations of PFOA and PFOS in drinking water, vegetables, fruits,
feed crops, livestock products (milk, beef, chicken, eggs), and fish using various uptake factors, such as
biotransfer factors (BTFs) and bioaccumulation factors (BAFs). Finally, the concentrations of PFOA and
PFOS in each media type are used to calculate exposure and risk to the relevant receptors in each
conceptual modeling scenario.
2.9.2.1 PFOA- and PFOS-specific Fate and Transport Considerations
The mobility of PFOA and PFOS in the environment, an active area of research, is known to be affected
by a number of factors, including:
• hydrophobic/hydrophilic-surfactant behavior {e.g., fluid-fluid or air-fluid interface retention);
• attraction to the solid phase in sediment (Higgins and Luthy, 2006), sludge (Milinovic et al.,
2016), soil (Milinovic et al., 2015), and organic carbon in general (Higgins and Luthy, 2006);
• ionic behavior as a function of pH (Place and Field, 2012; Pereira et al. 2018); and
• competition among these processes.
Methodologies for assessing the impact of PFOA and PFOS retention at the air-water interface (AWI)
have been proposed (Brusseau, 2018), modeled (Guelfo et al., 2020), and implemented in various fate
and transport simulators (Guo et al., 2020, 2022; Silva et al., 2022). The fact that AWI has been shown to
be a significant factor in PFAS fate and transport has focused modeling efforts on the vadose zone
though AWI retention is also relevant to saturated aquifer environments because some air may be
entrained in pore spaces of the saturated zone (Bumb et al., 1992). In equilibrium transport modeling, it
is assumed that sorption occurs at much faster rates than the residence time of groundwater. However,
studies have observed that solid phase sorption processes for PFOA and PFOS are not always well
DRAFT
41
-------
represented by reversable equilibrium partitioning assumptions due to rate-limited air-water interfacial
adsorption and fluid-fluid interfacial adsorption (Guelfo et al., 2020; Brusseau, 2020). Sorption of PFOA
and PFOS to non-advective domains influences the magnitude and timing of transport from the vadose
zone to groundwater.
Soil transport modeling studies that incorporate PFOA- and PFOS-specific, non-linear adsorption
processes predict that even after the source of PFOA and PFOS in the vadose zone has been
discontinued, PFOA and PFOS mass can remain in the vadose zone for decades, centuries, or longer
(Zheng & Guo, 2023). In some of this soil modeling, PFOA and PFOS appear to not break through the
vadose zone and enter groundwater aquifer for hundreds or thousands of years after they are applied to
the surface (see Section 3.2.3 and Zheng & Guo, 2023). Given that PFOA and PFOS manufacturing only
began in the 1940's, this modeling would suggest that groundwater contamination associated with land
application of PFOA and PFOS contaminated biosolids would not be observed for many years into the
future. However, instances of high groundwater concentrations of PFOA and PFOS have been
documented for both shallow and deep vadose zones (Brusseau et al., 2020; Dauchy et al., 2019) and in
various states including Maine, Michigan, and Alabama, where PFOA and PFOS contamination is
attributable to land-applied biosolids (see Section 6). For example, Brusseau et al. 2020, in their review
of PFAS concentrations at contaminated and non-contaminated soil sites, found that though PFOS
concentrations are highest in the upper portion of the soil profile (as expected), PFOS is still present at
significant (~1-10 ppb) concentrations in soil samples from 25-40 m below the surface. If PFOA and PFOS
were so successfully retained in the surface soils and upper vadose zone subsurface soils, these real-
world examples of transport deep in the vadose zone and groundwater would not be expected.
Soil heterogeneities, preferential transport pathways, and colloidal transport mechanisms are
environmental characteristics that are often omitted from modeling studies and that may be responsible
for faster migration of PFAS through the vadose zone than is expected from current modeling (Zeng and
Guo, 2021; Bierbaum et al. 2023). These factors may also result in more PFOA and PFOS mass being
transported through the soil column than is estimated using currently available models, resulting in
higher observed groundwater concentrations.
The EPA assessed fate and transport models that explicitly include retention on the AWI (such as the
Guo et al. 2022 model) and existing EPA models that can be parameterized to better reflect PFOA and
PFOS transport behavior (see appendix C). However, the ability of any model to reliably predict the
timing of PFOA and PFOS impacts to groundwater in highly characterized, non-idealized environments is
low (Zeng and Guo, 2021). In fact, available models (Guo et al. 2022 and EPA models) likely overestimate
the time required for PFOA and PFOS to reach groundwater, and this modeling of PFAS transport in soil
systems remains an active area of ongoing research. Consistent with previous sewage sludge risk
assessments, this draft risk assessment will consider the peak groundwater concentrations when
calculating risks, regardless of the timing of their occurrence, to avoid underestimating risks through this
pathway (US EPA, 1992; US EPA 2003a;b).
In addition to transport models for PFOA and PFOS movement through soil, this draft risk assessment
also requires models for understanding transport to surface waters and groundwater transport. The
following sections describe which models were assessed and selected for use in this assessment.
2.9.2.2 Soil Surface Modeling
The underlying model for the evaluation of the concentrations of PFOA and PFOS in soil is the EPA's
Multimedia, Multipathway, and Multireceptor Risk Assessment (3MRA) modeling system (US EPA,
DRAFT
42
-------
2003f;g), developed by the EPA Office of Land and Emergency Management. The 3MRA modeling
system includes a number of modules.
The Land Application Unit (LAU) module within 3MRA models the incorporation of contaminants in
biosolids into the top layer of soil and then simulates:
• The vertical movement of those contaminants through the top 20 cm of soil, estimating a
leachate mass flux that is used by the EPA's Composite Model for Leachate Migration with
Transformation Products (EPACMTP) to model transport through the vadose zone to
groundwater;
• The horizontal movement of those contaminants via erosion and runoff from the field to a
buffer area, and ultimately to the surface water body, estimating a waterbody load that is used
by a surface water model, the Variable Volume Waterbody Model (VVWM), to model transport
within the waterbody; and
• The losses of contaminant to air via wind erosion of particulates; this mass is removed from the
LAU but is not modeled further.
The mass that remains after these processes is the basis of the soil concentration in the top layer of soil
that is available for plant uptake, soil consumption by livestock, and consumption of soil by humans.
The 3MRA model has been peer reviewed and used extensively to support regulatory risk assessments
conducted for EPA's Office of Resource Conservation and Recovery and Office of Water (US EPA, 2003f).
As part of the 3MRA modeling system, the LAU source module was developed to estimate annual
average surface soil constituent concentrations and constituent mass emission rates to air, downslope
land, and groundwater. These estimates are passed to other environmental fate and transport models
and used to calculate exposure and risk. Additionally, the LAU source module incorporates a local
watershed submodule (a "local" watershed is a sheet-flow-only watershed containing the LAU and a
downslope buffer area between the LAU and the waterbody) to provide estimates of constituent mass
flux rates from runoff and erosion from the field to the downslope buffer, and then from the buffer to a
downslope water body (called the drinking water reservoir in conceptual models for the pasture and
crop farm). The LAU module also produces constituent soil concentrations on the field, as well as in the
downslope buffer area.
The LAU model conserves mass while accounting for releases from the agricultural field via leaching,
volatilization, particulate emissions, runoff, and erosion, and release from the buffer via runoff and
erosion to the waterbody. The model also accounts for deposition onto the plants on the field, but not
back onto the soil of the agricultural field or buffer, so soil concentrations in these areas may be slightly
underestimated. Though the LAU model can account for abiotic degradation, biodegradation, and
volatilization, these factors are not relevant for PFOA and PFOS (ATSDR, 2021).
The specific inputs and the data used in the LAU source model are presented in Appendix B. The LAU
model runs for 150 years, starting with the year of first application, and outputs a time series of daily
and annual average soil concentrations for the field and the buffer, daily and annual average
concentrations of contaminant mass, eroded solids, and runoff from the field and buffer, annual average
leachate concentrations, and air emission rates (particulate). This assessment assumes that land
application occurs for 40 years and then stops. Peak concentrations in the soils, runoff, and leachate are
expected to occur around the time application ceases, however, the longer simulation time allows for
confirmation that 150 years is sufficient to capture peak concentrations in these media.
The 3MRA Surface Impoundment module is used to model the amount of PFOA and PFOS that may be
released from a surface disposal site to the vadose zone under the site. The resulting leachate mass flux
DRAFT
43
-------
is used by EPACMTP to estimate transport through the vadose zone and groundwater. The inputs used
to parameterize the surface impoundment model are also included in Appendix B.
2.9.2.3 Surface Water Modeling
Erosion and runoff loadings from the downslope buffer area (calculated by the 3MRA LAU module) are
fed into VVWM (US EPA, 2019b), developed by the EPA Office of Pesticide Programs for estimating
concentrations in surface water bodies. The VVWM model estimates concentrations of PFOA and PFOS
in a drinking water reservoir; dissolved concentrations in the water column are used to calculate risks
associated with drinking water whereas total water column concentrations are used to calculate fish
tissue concentrations using BAFs.
VVWM was developed from another EPA model, the Exposure Analysis Modeling System (EXAMS; Burns,
2000) that simulates standard water bodies that receive chemicals from the standard field. VVWM
behaves much like EXAMS, simulating the US EPA standard water bodies (i.e., farm pond and index
reservoir), but with greater efficiency and flexibility. The VVWM also allows for variations in water body
volume daily due to runoff, precipitation, and evaporation. Temperature, wind speeds, and chemical
dissipation processes are also allowed to vary daily.
The VVWM consists of two regions: a water column and a benthic region (US EPA, 2019b). Each
individual region is completely mixed and at equilibrium with all phases in that region, with equilibrium
described by a linear isotherm. The two regions are coupled by a turbulent-mixing, first-order mass-
transfer process. The water volume may vary by inputs of precipitation and runoff and by outputs of
evaporation and overflow. Degradation via biodegradation, hydrolysis, and photolysis can be
parameterized for each compartment as applicable in VVWM, but PFOA and PFOS do not undergo these
degradation processes (ATSDR, 2021).
2.9.2.4 Groundwater Modeling
Modeling of the groundwater pathway is accomplished using two models: a model responsible for
releasing PFOA or PFOS into the subsurface, the 3MRA LAU Module, and a subsurface flow and
transport model, EPACMTP (US EPA, 2003d;e). The 3MRA source modules calculate the amount of PFOA
and PFOS that leave the top layer of soil for the LAU or the bottom of the surface disposal unit as part of
the leachate. The maximum mass flux of any constituent in the modeled leachate occurs in all cases
during the application period and is fully captured within the 150-year modeling timeframe of the
source modules. The subsurface model (EPACMTP) is allowed to run as long as 10,000 years if necessary
to observe the peak groundwater concentrations at simulated wells (see Section 2.9.2.1 for discussion of
the known modeling deficiencies in predicting the timing of groundwater impacts and see Section 3.2.3
for a discussion of modeled groundwater concentrations over time).
EPACMTP is then used to calculate the amount of PFOA and PFOS that travel through the remaining soil
column (the vadose zone) to the groundwater table and downgradient to a drinking water well located 5
meters from the edge of the field or surface disposal unit (in the middle of the 10-meter-wide buffer).
The modeled depth of the vadose zone varies depending on the geographical location. As described in
Section 2.9.2.1, PFOA and PFOS present challenges for calculating soil transport compared to typical
organic contaminants due to their surfactant properties. PFOA and PFOS can reside at the air-water
interface and electrostatically sorb to minerals in soils after moving into the vadose zone. Depending on
the hydrogeology and minerology of the location, this may retard the movement of the chemical into
the groundwater table. EPACMTP has been used within the EPA for decades to estimate subsurface
transport through the vadose zone to groundwater but has not traditionally been parametrized to
estimate air-water interface effects.
DRAFT
44
-------
The hypothetical drinking water well in EPACMTP is represented by four observation locations placed at
0.5, 1.0, 1.5, and 2.0 meters below the water table to ensure the maximum groundwater concentration
is observed. The highest concentration observed across the four depths is used to calculate a
proportional constant that represents the minimum cumulative reduction and attenuation of leachate
concentrations as they migrate through the subsurface to the drinking water well.
In Appendix C (groundwater modeling), models besides EPACMTP are compared for their relevance to
PFOA and PFOS vertical transport through the vadose zone. Other models can incorporate PFAS-specific
parameters like air-water interface effects and nonlinear adsorption. These factors result in lower peak
groundwater concentrations and longer delays in the transport of PFOA and PFOS to the groundwater at
the farm. EPACMTP estimates arrival times of aquifer contamination at the water table that are, in some
cases, much longer than those that have been observed at biosolids application sites in Maine and
Michigan, but closer to those observed breakthrough times than models that incorporate air-water
interface effects and nonlinear adsorption. For this reason, EPACMTP was selected as being more
appropriate for modeling vertical transport through the soil column.
The model implementation also includes some assumptions to protect groundwater resources now and
in the future. Firstly, the draft risk assessment assumes that drinking water receptors have wells that are
placed in the center of the buffer, five meters from the edge of the field or surface disposal unit and
centered around the highest concentration in the groundwater plume below the water table. If a
homeowner had a deeper well or a well located on the fringe of the plume, rather than the center of the
plume, they would have lower drinking water concentrations and lower risks. The draft risk assessment
also presents exposures that occur during the years with the highest media concentrations for soils,
surface water, and groundwater. While these assumptions may overestimate risk to a specific person at
a specific site, they are reasonable for the purpose of a national draft risk assessment seeking to
determine if levels of PFOA and PFOS in sewage sludge may adversely affect human health or the
environment. For example, it is important to protect the groundwater as a source for potential drinking
water regardless of when that peak may be reached or where a well may be placed.
2.9.2.5 Air Dispersion Modeling
Generally, the EPA uses AERMOD to parameterize the transport of most chemicals from farm fields;
however, for the PFOA and PFOS assessment, the volatilization rate has been set to zero and no
dispersion modeling is needed. The only airborne loss of PFOA and PFOS is due to wind erosion
emissions of dust from the field, and this loss is calculated within the 3MRA LAU module.
2.9.2.6 Plant and Animal Uptake Equations
The produce, meat, and milk exposures are calculated using the methodology found in the Human
Health Risk Assessment Protocol (HHRAP; US EPA, 2005b), developed for hazardous waste combustion
facilities and slightly modified to account for the available data for parameterization. The fate and
transport models generate the estimated concentrations of the contaminated media that are used to
calculate concentrations in crops and animal feed (Equation 1), animal products (Equation 2), and fish
tissue (Equation 3).
DRAFT
45
-------
Crops
Equation 1. Crop Concentrations Due to Root Uptake from Soil, Pproduce, Pfeed (mg/kg)
Produce (Aboveground Fruits and Vegetables, Root Vegetables)
Feed crops (Forage and Silage)
n fWO-MAF^
^produce Csoil * * ( 100 )
Pfeed ~ CSoil * ^r
Name
Description
Source
P produce,
Pfeed
Concentration of contaminant in crops (aboveground fruits or
vegetables, and root vegetables or animal feed (Pfeed)
Calculated
Csoil
Concentration of contaminant in soil, averaged overfilling depth
(mg/kg)
LAU model output
Br
[plant](^-dry weight)
Soil-to-plant bioconcentration factor: wrr
[soil](-^wet weight)
See model parameterization,
Section 2.9.3.4
MAF
Plant tissue-specific moisture adjustment factor to convert dry
weight concentrations into wet weight (percent)
See model parameterization,
Section 2.9.3.4
100
Conversion factor from percent to fraction (unitless)
NA
Livestock
Equation 2. Concentration in Animal Products, A (mg/kg WW)
where
A - BTF x [lsoll + Ifeed + Iwater\
1soil ~ ^soil * Qsoil * Bs
Ifeed ~ ^ ' Pi * Qi * ^i
i
Iwater * Qw *
Name
Description
Value
A
Concentration in the animal product (beef, chicken, egg, milk)
Calculated
I
Livestock intake of soil (Isoii), feed (Ifeed), and water (Iwater)
Calculated
BTF
[animal product](-^-ww)
Biotransfer factor for animal product: mr
intake rate(——)
KdayJ
See Section 2.9.3.5
Csoil
Average concentration in surficial soil (mg/kg)
LAU model output
Qsoil
Quantity of soil consumed per day (kg/day)
See Section 2.9.3.6
Bs
Bioavailability factor in soil (fraction)
1
Pi
Average concentration in plant type / (forage, silage, grain)
(mg/kg DW)
Calculated; see Equation 1 for
forage and silage. Grain assumed
to be purchased from an
uncontaminated source
Qi
Quantity of plant type / consumed per day (kg DW/day)
See Section 2.9.3.6
Fi
Fraction of plant type i grown in contaminated soil (unitless)
See Section 2.9.3.6
Cgw
Average concentration in groundwater (mg/L)
LAU model output
Qw
Quantity of water consumed per day (L/day)
See Section 2.9.3.6
F w
Fraction of water contaminated (unitless)
1
DRAFT
46
-------
Fish
Equation 3. Concentration in Fish Filet, Cfnet(mg/kg)
Cfilet = Ctot X BAF
Name
Description
Value
Cdtot
Total water column concentration (mg/L)
VVWM model output
BAF
Bioaccumulation factor for fish filet (L/kg)
See Section 2.9.3.7
2.9.3 Model Parameteriza tion
To calculate exposure and risk for the pathways depicted in the four conceptual modeling scenarios
where the EPA is quantitatively assessing outcomes (crop farm, pasture farm, surface disposal site, and
land reclamation), the EPA must parameterize hundreds of values used in fate and transport and
exposure models. This includes parameters related to the fate and transport of PFOA and PFOS in soil
columns, groundwater, surface water, and into crops and animals. These parameters also include
toxicity values for PFOA and PFOS and exposure factors for the many pathways of human exposure
depicted in the conceptual models. Finally, these parameters include characteristics of the modeled
environment, like the size of the modeled surface water reservoir or the size of the field receiving
biosolids for land application.
Establishing chemical-specific values for some of these parameters can be challenging for PFOA and
PFOS because these chemicals present different characteristics than are typical for other organic
chemicals. For example, while some environmental fate parameters for other organic compounds can
be predicted using the water-octanol partitional coefficient (Kow), this value cannot be measured for
PFOA and PFOS because of their surfactant properties; experimental data are needed for these
parameters instead. If there is an existing assessment available from the EPA or another agency that is
relevant to a chemical-specific parameter, the conclusions of those assessments are prioritized over
results from individual studies. For example, rather than compile individual human health toxicity
studies for PFOA and PFOS, this draft risk assessment will rely on the conclusions of the EPA's Final
Toxicity Assessments for PFOA and PFOS (US EPA 2024b;c) as a source of toxicity values (reference doses
and cancer slope factors). Similarly, this draft risk assessment will rely on fish uptake factors
(bioaccumulation factors) presented in EPA's Draft Human Health Criteria for PFOA and PFOS (US EPA,
2024o;p).
For many of the fate, transport, and uptake parameters, there are no relevant existing assessments that
can be used for parameterizing model inputs needed for this assessment. In these cases, the EPA
searched and reviewed the available peer-review literature. The EPA applied the following hierarchy to
the available fate and transport studies:
1. Field studies from sites with biosolids application
2. Laboratory (including greenhouse) studies using biosolids-amended soils
3. Field studies from other types of PFAS-impacted sites
4. Laboratory (including greenhouse) studies using materials with other sources of PFAS
contamination.
DRAFT
47
-------
Study quality metrics relevant to each study type are described in this section. When there are multiple
measurements or studies of sufficient quality available for the same parameter in the same data tier,
the EPA used the median value to parameterize the models. As an example, plant uptake factors were
determined by prioritizing studies where biosolids contaminated with PFOA and PFOS were applied in
the study area/field. If there were multiple acceptable field-studies available where the source of PFAS
contamination was sewage sludge, the median of these data was selected for the study parameter.
While field studies are generally preferred over laboratory studies for most parameters, field studies
with real-world contamination are likely to include potential confounders, including other PFAS, which
may or may not be precursors to PFOA and PFOS. While use of these studies may overestimate PFOA
and PFOS transport or uptake in some settings, the degree of PFAS diversity seen in these real-world
field studies is not dissimilar to the degree of PFAS diversity found in biosolids (Thompson et al., 2023a).
For this reason, the benefit of using biosolids-specific data in most cases outweighs the uncertainty
contributed from the potential for PFOA and PFOS precursors or other confounders to influence the final
parameter values.
Several conceptual models are based on agricultural sites, where a farm family's exposure is modeled.
When parameterizing human exposure factors, food consumption data specific to home-produced foods
or consumption rate amongst farmers are prioritized over general population data. This draft risk
assessment uses exposure factors presented in the most recently updated version of the EPA's Exposure
Factors Handbook (EFH; US EPA, 2011) chapter for home-produced foods (Chapter 13), when available.
If there are not data specific to home-produced food available, chapters of the EFH describing the
general population are used. Some of these chapters have been updated since 2011 and issued as
separate documents; in all cases, the most recent update is used and referenced.
The environmental fate and transport models used in this draft risk assessment also require parameters
related to the environmental setting, such as size of the field used for land application. When these
parameters are not specific to sewage sludge use and disposal (for example, porosity of benthic
sediments in the waterbody near to the field), default values provided in the peer-reviewed EPA model
or values from previous EPA sewage sludge assessments are used (US EPA, 2003a). When these
parameters are specific to a setting, regionally representative values from a wet, moderate, and dry
climate in the US are used (see Section 2.9.3.13). When the parameters are relevant to practices for
sewage sludge land application or disposal, median values from relevant US datasets are used,
consistent with prior sewage sludge risk assessments (US EPA, 1992; 1995a; 2003a). Descriptions of the
selected values for each parameter are included below. Tables of values used for each parameter are
also summarized in Appendix B.
2.9.3.1 Toxicity Values
In 2024, the EPA published final human health toxicity assessments for PFOA and PFOS (US EPA
2024b;c). These final assessments include an RfD and CSF for PFOA and PFOS. These values are relevant
to all oral ingestion pathways, including drinking water and dietary intake. While PFOA and PFOS
exposures have been associated with numerous adverse health outcomes in humans, the RfDs and CSFs
are derived based on the most sensitive adverse health outcomes; protecting against these outcomes
will also protect against the outcomes that occur after higher levels of exposure. PFOA and PFOS are
classified as likely carcinogens (L). The biosolids exposure models assesses the cancer risks and non-
cancer risks associated with each exposure pathway. As described in Section 2.6.1, the RfDs and CSFs for
PFOA and PFOS are as follows:
DRAFT
48
-------
Table 6. Toxicity Values for PFOA
Toxicity Value Type
Value
Critical Effect(s), Critical Study/Studies
RfD (based on
epidemiological data)
3 x 10"8 mg/kg/day
Reduced antibody response to vaccinations in children
(diphtheria and tetanus) (Budtz-Jorgensen & Grandjean,
2018); decreased birth weight (Wikstrom et al., 2019);
increased serum total cholesterol (Dong et al., 2019)
CSF (based on
epidemiological data)
29,300 (mg/kg/day)1
Renal cell carcinoma (RCC) (Shearer et al., 2021)
Table 7. Toxicity Values for PFOS
Toxicity Value Type
Value
Critical Effect(s), Critical Study/Studies
RfD (based on
epidemiological data)
1 x 10-7 mg/kg/day
Decreased birth weight (Wikstrom et al., 2019); increased
serum total cholesterol (Dong et al., 2019)
CSF (based on animal
toxicological data)
39.5 (mg/kg/day)-1
Combined hepatocellular adenomas and carcinomas in female
rats (Thomford, 2002; Butenhoff et al., 2012, 1276144)
2.9.3.2 Sewage Sludge PFOA, PFOS Concentration and Other Characteristics
The fate and transport models used in this assessment require a starting concentration for PFOA and
PFOS in sewage sludge. For this central tendency assessment, the EPA is using low starting
concentrations of 1 ppb for PFOA and 1 ppb for PFOS (dry weight). These values were selected because
they represent a concentration that is lower than most sewage sludge generated in the U.S., including
sewage sludge that represents only domestic sources (see Section 2.4 and Appendix A). This value is also
near the reporting limits expected in most major laboratories using EPA Method 1633 on sewage sludge
(US EPA 2024d). Notably, the models and calculations used in this draft risk assessment result in a linear
relationship between the starting concentration of PFOA or PFOS in sewage sludge and the resulting
concentrations and risks. For example, if modeling a starting concentration of 1 ppb PFOA results in a
hazard quotient of 1 or a cancer risk level of 1 in 1 million (1 x 10 s), a starting concentration of 10 ppb
PFOA would result in a hazard quotient of 10 and a cancer risk level of one in one hundred thousand (1 x
10"5). Should the EPA's draft modeling find risks in a given potential pathway with this low starting
concentration of PFOA and PFOS in sewage sludge, it is reasonably anticipated that these risks could be
prevalent across use and disposal sites in the U.S.
The fate and transport models also require additional characterization of the sewage sludge, including
the dry bulk density, the fraction organic carbon, and the silt content of the sewage sludge. The silt
content was assumed to be 10% (the median of the distribution provided in the 2003 risk assessment
documentation). The bulk dry density of biosolids was assumed to be 0.7 g/cm3, which is the median of
the range provided in Gunn et al. 2004. This value was updated from the 2003 risk assessment
documentation, which reported a bulk dry density of 1.6 g/cm3, cited from the Technical Support
Document for the Land Application of Sewage Sludge (US EPA, 1992).
2.9.3.3 Physical and Chemical Properties
PFOA and PFOS partitioning data between water and soil are needed to model the fate and transport of
these chemicals through the environment. To represent solid-phase sorption of PFOA and PFOS in
environmental media potentially affected by land-applied biosolids, the modeling framework uses the
organic carbon distribution coefficient (Koc)- Koc is then used to calculate the solid-phase adsorption
coefficient (Kd) using the fraction of organic carbon (foc) in each modeled soil type. The EPA conducted a
literature search to aggregate measured Koc and Kd values for PFOA and PFOS in biosolids field studies,
DRAFT
49
-------
other field studies, and laboratory studies. The methodology and results from this literature search are
described in Appendix C, Section C.3.2.1.
Based on this literature review, the EPA concluded that there are a range of Koc values reported under
various environmental conditions in soils. To represent the range of potentially relevant Koc values at
each site, the EPA modeled a "low-Koc" scenario and a "high-Koc" scenario, representing the 10th and 90th
percentiles of the distribution, respectively. The values for PFOA and PFOS are provided in Tables 8 and
9, respectively. See Appendix C for more information on the distribution of observed Koc values for PFOA
and PFOS.
Table 8. Koc Values for PFOA
Scenario
Value
Source
Low Koc (10th percentile)
26 cm3/g
Literature search; see text and Appendix C
High Koc (90th percentile)
1,100 cm3/g
Literature search; see text and Appendix C
Table 9. Koc Values for PFOS
Parameter
Value
Source
Low Koc (10th percentile)
250 cm3/g
Literature search; see text and Appendix C
High Koc (90th percentile)
22,000 cm3/g
Literature search; see text and Appendix C
Koc values are used in the model to estimate Kd values in four media: surface soil, subsurface soil,
sediment, and suspended sediments. The sediment and suspended sediment values are for the drinking
water reservoir. The foc is multiplied by the Kocto obtain the Kd for each medium.
As described above, Koc values vary for PFOA and PFOS in different studies and Kd will vary across sites.
Other soil parameters including protein content and oxalate-extractable iron and aluminum may also be
more relevant for a particular site for soil adsorption and Kd. Oxalate-extractable iron and aluminum
content may be particularly relevant to deep soil settings, where organic matter content is low. The goal
of this modeling activity is to provide estimates of a range of transport behaviors as this parameter is
varied to reflect the environmental variability that will occur at different locations. To achieve this the
model has been parameterized with multiple Koc values and to calculate Kd across the simulated media
in the 3MRA model including soil, biosolids, and sediments in the drinking water reservoir. Table 10
shows the values used to represent foc across each type of media.
Table 10. Fraction Organic Carbon Values by Medium
Medium
foc Value
Reference
Natural soil under the field
0.0118
STATSGO (USDA, 1994)
Bed sediments
0.04
VVWM (US EPA, 2019b)
Suspended sediments
0.04
VVWM (US EPA, 2019b)
Biosolids
0.40
Biosolids 2003 (US EPA, 2003a)
The LAU model uses both the biosolids foc and the soil foc and calculates a depth-weighted average of the
two over the total incorporation depth (20 cm for crop, 2 cm for pasture or reclamation). For the
purposes of this average, the waste depth is the application rate for a single application divided by the
biosolids bulk density, and the soil depth is the rest of the application depth. The result is a higher foc
than the soil value, but lower than the biosolids foc.
Both PFOA and PFOS are stable in air and water (UNEP, 2015; ATSDR, 2021), so half lives in air and water
were not used. Other chemical-specific property values may be found in Appendix B.
DRAFT
50
-------
2.9.3.4 Plant Uptake Factors
Bioconcentration factors (BCFs) are the uptake factors used for plants and are defined as the
concentration of the compound in the relevant compartment of the plant divided by the concentration
of the compound in the underlying soil. BCFs are unitless. Plant BCFs can be derived from studies with
various experimental designs as long as the study measures concentrations of the chemical in the plant
tissues and the soil. Surveys of plant tissue concentrations alone (e.g., market surveys) are not useful for
modeling or generating plant BCFs as they lack the corresponding soil data, though these studies can be
used for general context of what types of exposures may be occurring. Because the matrix of biosolids
and natural field conditions may impact the accumulation of PFOA and PFOS into plants, the following
data hierarchy for plant BCF studies is used in this assessment:
1. Field studies with biosolids-amended soil
2. Greenhouse studies of potted plants with biosolids-amended soil
3. Field studies with other sources of PFAS contamination impacting the soil
This data hierarchy allows the EPA to preferentially select studies with biosolids-specific contamination
sources and field conditions over other types of studies, as data are available.
The following literature search strategy was used to identify potentially relevant studies:
Database searched: PubMed
Search string: Title/Abstract search, ("PFAA"OR "PFAS" OR "PFCA" OR "PFOA" OR "PFOS") AND ("food"
OR "crop*") AND ("soil" OR "biosolid*" OR "sludge")
Date searched: 3/15/2024. No date limitations on results.
Relevant federal and state government reports are also included.
Results: 133 studies and results from recent literature surveys by Lesmester, 2023 and Li, 2022
The following criteria must be met:
• Measured PFOA or PFOS concentrations in plants and soil
• Study must relate to one of the 3 categories in the data hierarchy
• Known source of contamination
• Soil not contaminated by spiking (lab contaminated soil)
Several key findings of the following papers include that grass and leafy greens likely exhibit the highest
soil to plant uptake (or plant BCF) values amongst the plants that have been studied. Roots and tubers
that are consumed (e.g., carrots) may also have high uptake, but a field study does not exist to verify the
available greenhouse data (Wen et al., 2016) for that compartment of plants.
For PFOA and PFOS, fruits and seeds have lower uptake than the stems and leaves (vegetative parts of
the plants), likely due to the need to cross additional membranes to reach the fruit and seeds.11 Blaine
et al., 2013 collected corn stover (stalk, leaves, and cobs), corn grain, and soil from biosolids amended
fields in the Midwest; these researchers found no detectable PFOA or PFOS in corn grain from fields with
soil concentrations of 4.4 ng/g PFOA and 4.3 ng/g PFOS. Blaine et al., 2014, a greenhouse-based study
11 Note that pea pods, tomatoes, and eggplants are fruits like apples, oranges, and blueberries are the ripened ovary of a plant
and therefore "fruits" from a botanical perspective. However, some dietary surveys may create confusion based on common
usage of terms like vegetable and fruit.
DRAFT
51
-------
of plant uptake using biosolids-amended soils, similarly found that uptake into fruit was one to three
orders of magnitude less than uptake into roots or shoots for PFOA and PFOS.
Authors of most studies have estimated uptake factors based on dry weight concentrations in the plant
matter. The basis of the soil concentration is reported in most of the following articles and specified as
dry weight. The model calculates and applies these uptake factors to wet weight of soil; therefore, while
the study discussions below present the data as reported (dry weight plant concentration to dry weight
soil concentration), the final BCFs presented in Appendix B have been converted to dry weight plant
concentration to wet weight soil concentration using field capacity (water content of soil) and porosity
(water plus air content) of soil for feed crops; no further conversion is necessary as animal dietary data
are also commonly expressed on a dry weight basis. For fruits and vegetables consumed by children and
adults, the relevant consumption data are available on a wet weight basis, so a moisture adjustment
factor (MAF) is needed. The MAF used for each type of crops is shown in Table 11. These crops
groupings are defined as in the Exposure Factors Handbook, with exposed fruits and vegetables defined
as those that the edible portion grows aboveground without a protective rind or pod (e.g., leafy greens,
apples) and protected as those that the edible portion grows aboveground with a rind or pod that is not
eaten (e.g., peas, oranges). Root vegetables include tubers and roots, for which the edible portion grows
underground (e.g., carrots, potatoes).
Table 11. Moisture Adjustment Factors by Type of Produce
w
<1)
w
a>
_Q
-Q
ra
w
a>
"O
"3
LL.
<1)
O)
<1)
>
"3
LL.
a)
O)
<1)
>
_a
o
o
"O
a>
"O
a>
a>
"0
<1)
"(J
O)
£
"O
0
Q_
0
Q_
<1)
<1)
0
Description
X
LU
X
LU
Q_
Q_
en
Reference
MAF
Moisture adjustment factor
85
90
87
81
81
EFH :2011 (US EPA, 2011)
(% water)
Field Studies with Bioso/ids-Amended Soil
Yoo et al. 2011
Overview: This study collected grass samples from fields near Decatur, Alabama, that had received
applications of sludge from a WWTP contaminated by industrial releases. There was no known irrigation
at these sites. The study evaluated three grass species: Kentucky blue grass, Tall Fescue, and Bermuda
grass. Each of these grasses could be forage for animals in pastures or used for hay or silage production.
Soil and plant samples were collected at least several months after the last sludge application.
Results: The study presented soil to plant BCFs. The BCF values are labeled as grass soil accumulation
factors (GSAF) for each of the grasses across multiple plots in terms of dry weight plant concentrations
over dry weight soil concentration. The table below presents the results from 5 plots of grass, and a
mean over all the grasses with tall fescue being weighted more heavily as it was in 3 of the 5 plots.
Table 12. Plant BCFs from Yoo et al. 2011
Plant Species
PFOA Plant BCF
PFOS Plant BCF
Kentucky blue grass
0.27
0.083
Tall Fescue (average of 3 plots)
0.29
0.076
Bermuda grass
0.13
0.035
Mean over all grasses
0.25
0.07
DRAFT
52
-------
Uncertainties: Yoo et al. evaluated plants relevant to livestock consumption and evaluated the uptake
factor in a farm field that was contaminated with PFAS due to biosolids application. It is possible that
some degradation of PFOA or PFOS precursors occurred within the plant or soil that could lead to an
over-estimation of plant BCF. The study reported that all FTOHs were nondetectable; this provides some
indication that precursor conversion at this site may be a small effect. Given that uptake factors from
biosolids-amended fields are considered the most relevant to risk assessment for the farm family,
uncertainty related to the presence of precursors is unavoidable as many precursors cannot be
quantified by available lab methods. The study also does not give a clear description of the distribution
or total number of plant samples per species or soil samples from the fields taken in the study to create
the uptake factors.
Blaine et al. 2013
Overview: This study evaluated tomatoes, lettuce, and corn in a midwestern US field that had been
fertilized with biosolids at multiple rates. The study does not specify if fields were irrigated or the PFAS
profile of irrigation water.
Results: The only field that produced measurable data for PFOA and PFOS was a pilot field that had
biosolids applied at 4 times the agronomic rate. This field had PFOS soil concentrations of 13.9 ppb and
PFOA soil concentrations of 5.2 ppb. All corn grain and tomato fruit samples had concentrations of PFOA
and PFOS below the limit of quantification (LOQ), which were 0.2 ng/g for PFOA and 0.1 ng/g for PFOS.
Within that field, the BCF for soil to lettuce (phrased as a BAF in the study) was 0.10 for PFOS. BCFs for
PFOA were not quantifiable in any of the produce and PFOS data was only measurable in the lettuce.
The PFOS value from this study for lettuce likely serves as a confirmation that the vegetative parts of
plants will take up PFOS and that the BCF may be in the range from 0.07 to 0.10.
Uncertainties: The only fields with detectable amounts of PFOS in the plants received biosolids
applications above the agronomic rate. The increased application rates were necessary to raise the
contamination levels in the plants above analytical detection limits, but it isn't clear if the increased
nutrients (N and P) in the soil would increase or decrease the plant uptake factor (BCF). As with any
study that uses field applied biosolids, it is possible that there were precursors present that could
breakdown to PFOA or PFOS and which increased the BCF.
The Yoo and Blaine studies both meet the qualifications for the highest tier of data for evaluating risks
due to land application of biosolids and are particularly relevant for pasture scenarios. For pastured
livestock (e.g., cattle, chickens, pigs, sheep) the uptake of PFOA and PFOS into grasses indicate that
consumption of meat, dairy, and eggs could be a significant pathway of human exposure to these
chemicals for farm families or CSA purchasers. The above studies provide data for parameterizing forage
and silage for this pathway.
Greenhouse Studies Using Bioso/ids-Amended Soil
Uncertainties: For the following greenhouse studies, the BCFs calculated were typically higher than field
studies and may be overestimated. Plants in pots for greenhouse studies have indicated higher uptake
factors than field studies, this may be due to the roots having higher exposure to soil in the contained
pot as opposed to a field. Although the BCFs may be elevated, it is still thought that the relative
concentrations amongst the plant compartments in greenhouse studies indicate a pattern that would be
representative of plants grown under field conditions. These studies also use biosolids-amended soil
which may contain precursors.
Blaine et al. 2013
Overview: This greenhouse study (also cited above) investigated PFOA and PFOS fate in lettuce and
tomatoes raised in biosolids-amended soil.
DRAFT
53
-------
Results: The BCF values (1.34 to 2.52) for PFOA in lettuce grown under greenhouse conditions were
much larger than the value cited for grasses (above). Greenhouse values for BCFs of PFOS in lettuce
ranged from 0.32 to 1.67 and may indicate that greenhouse studies over-estimate field values for the
same crop group, although for some shorter PFAS chain lengths in this Blaine study the field and
greenhouse values are similar between the greenhouse and field studies.
Blaine et al. 2014
Overview: This study focused on biosolid-amended soil used in greenhouse studies of radishes, celery,
tomatoes, and peas.
Results: The BCF values for soil to root concentration are significantly higher than the values for plants in
the field studies mentioned above (Yoo et al., 2011 and Blaine et al., 2013). While this BCF value may be
a valid indication of increased concentration in the roots of plants, it has not been confirmed by a field
study. An important result of the greenhouse study, which is consistent with field studies, is that uptake
into fruit is much lower than into roots or shoots, indicating that the presence of PFOA and PFOS in the
edible fruit/seed portion of plants like tomatoes and peas may be of lower concern than consumption of
edible greens (celery, lettuce, spinach, etc.) or roots (carrots, radishes). That the concentration is lower
in edible portion of plants is perhaps unsurprising for long chain PFAS like PFOA and PFOS because the
chemical must be transported across more membranes to enter the seeds, grains, and fruits of a plant.
Wen et al. 2016
Overview: This study focused on the role of protein content differences between tissues on the
transport of PFAS within plants. This study illustrated the limitations of using plant uptake factors from
greenhouse studies for risk assessment of farm crops.
Results: The uptake factors for PFOS and PFOA were several times larger than those calculated in the
field studies above. For example, lettuce had a BCF for the shoot of 1.18 for PFOA and 0.396 for PFOS.
However, the positive correlation between the uptake factors and the total protein content of the
shoots and roots (higher uptake factors for tissues with higher protein content) is an interesting factor
to consider for which plants may pose the most concern for human or livestock consumption. As
measured in soil and fish, protein levels in distinct plant tissues may indicate where PFAS will
preferentially partition within a specific crop group. The data on radish from this study are part of the
range for determination of the root concentration factors for PFOA and PFOS.
Lechner and Knapp 2011
Overview: Lechner and Knapp 2011 employed greenhouse conditions to grow carrots, cucumbers, and
potatoes in biosolid-amended soils, and the highest transfer factors for PFOA and PFOS were for the
vegetative portions of each crop.
Results: The significantly larger plant uptake of PFOA and PFOS measured in greenhouse studies could
be of concern for farms that utilize greenhouses for year-round marketing or for home gardeners that
use a mixture of soil and biosolids in their greenhouse and potted plants.
Field Studies with Non-Biosoiids Sources of Contamination
Since data were available to estimate needed parameters in the first two higher tiers of the data
hierarchy, none of the studies from this tier were used in the risk assessment. Zhang et al. (2020) and Liu
et al. (2019) focused on vegetables raised in fields that were directly impacted by their proximity to a
PFAS manufacturing site. A summary of these data is presented by Li et al., 2022. The plant uptake
values from these studies are frequently much larger, sometimes by over an order of magnitude, than
the data available from the other literature sources cited above in this section. Given that there is a
possibility that air deposition and irrigation water contamination stemming from the nearby PFAS
manufacturing facility are increasing the concentrations of PFAS in the plants measured in these studies,
DRAFT
54
-------
these data are not appropriate for biosolids risk assessment. Specifically, Fig. 2.1 of Li et al. (2022)
presents a summary of the data from Zhang et al. (2020) and indicates that the BCF values for plants
ranged from 0.5 to 31 for PFOA. A plant BCF value of 31 for PFOA in zucchini is amongst the highest
reported in the literature. These high BCF values may be due to multiple exposure pathways in addition
to soil for PFAS in a field adjacent to a PFAS manufacturing site, e.g., contaminated water or air
deposition.
Scher et al. (2018) measured uptake of PFAS in garden produce at homes with contaminated irrigation
water as the source of PFAS. Uptake factors from soil to plants for PFOS were consistent across leaves,
fruits, and roots, ranging from 0.01 to 0.04. The uptake factors for PFOA were significantly higher with
values for spinach as high as 1.4, but with most uptake factors ranging from 0.1 to 0.7.
BCF Selection for Plants
All selected values were based on studies using biosolids-amended soil. However, field data were
available only for forage, silage, and above ground vegetables for PFOS. The remaining categories rely
on greenhouse studies using biosolids-amended soils to grow plants in pots. For forage and silage, this
assessment will use the mean BCF calculated across all the grasses in the Yoo et al. (2011) field study.
This mean was used to represent the plant BCF for vegetative parts of plants that are common to
forages and silages. For above ground vegetables (whether exposed or protected12), this assessment will
use the single field value (for lettuce) available for PFOS from the study of Blaine et al. (2013), and a
median of greenhouse values from Blaine et al. (2013, 2014) for PFOA, for which no field data were
available. For fruits (whether exposed or protected), this assessment will use the median of detected
greenhouse values from Blaine et al. (2013, 2014) and Lechner and Knapp (2011). Finally, for root
vegetables, this assessment will use the median of detected greenhouse values from Blaine et al. (2014),
Lechner and Knapp (2011), and Wen et al. (2016). The units for all the parameters below are dry weight
crop concentrations divided by dry weight soil concentration which results in a unitless BCF plant uptake
factor.13 These values are summarized in Table 13.
12 "Protected" means that the edible part of the plant is covered (e.g., orange) while "exposed" means that typically the
exterior of the fruit or vegetable is eaten (e.g., apple).
13 Note that these values were converted to wet weight soil concentration for use in the model, and Appendix B presents them
in those units, so the values differ somewhat. The conversion assumed a field capacity of 0.22 and a porosity of 0.43 (also
used elsewhere in the models), resulting in a dry soil mass fraction of 0.87. The values based on wet weight soil in Appendix B
were calculated by dividing the values above by that dry soil mass fraction.
DRAFT
55
-------
Table 13. Selected Plant BCFs
Plant Type
Chemical
Plant uptake
BCF (unitless)
Basis
Source
Forage
PFOA
0.25
field
Yoo et al. (2011) for grass
PFOS
0.07
field
Yoo et al. (2011) for grass
Fruit
PFOA
0.11
pot
median or geomean of tomatoes from Blaine et al., 2013,
sugar snap peas from Blaine et al. (2014), and cucumbers
from Lechner and Knapp (2011)
PFOS
0.03
pot
Sugar snap peas from Blaine et al. (2014) - only detected
value for PFOS
Root
Vegetables
PFOA
0.6
pot
median of pot carrots, potatoes, radish from Lechner and
Knapp (2011), radish from Blaine (2014), and radish from
Wen (2016)
PFOS
0.7
pot
median of pot carrots, potatoes, radish from Lechner and
Knapp (2011), radish from Blaine (2014), and radish from
Wen (2016)
Silage
PFOA
0.25
field
Yoo et al. (2011) for grass
PFOS
0.07
field
Yoo et al. (2011) for grass
Vegetables
(above
ground)
PFOA
1.3
pot
median of pot celery from Blaine et al (2014), pot lettuce
industrial biosolids, and pot lettuce municipal biosolids from
Blaine et al. (2013).
PFOS
0.1
field
field lettuce from Blaine et al. (2013) - only field study for
vegetables with a detected value
2.9.3.5 Livestock Uptake Factors
There are no existing EPA, FDA, or US Department of Agriculture (USDA) assessments that include
livestock BTFs for PFOA and PFOS. In the context of agricultural risk assessment, BTFs are defined as the
ratio of the concentration in the final product (i.e., meat, milk, eggs) to the total intake rate of that
chemical by the animal, represented in units of day per kg of food product. If a BTF is higher, this
indicates that the animal transfers or accumulates larger amounts of the chemical into the tissue that
becomes the finished food product. Because no existing source of BTFs was available for PFOA and
PFOS, the EPA reviewed the available literature, assessed the available studies, calculated BTFs from the
data provided in these published studies, and selected the most representative BTF for PFOA and PFOS
in each food product. If more than one high-quality BTF was available for a food type, the median BTF
was used. The following literature search strategy was used:
Database searched: PubMed
Search string: Title/Abstract search, ("perfluoroalkyl substance*" OR "polyfluoroalkyl substance*" OR
"PFAS" OR "PFOA" OR "perfluorooctanoic acid" OR "perfluorooctanesulfonic acid" OR "PFOS" OR
"perfluorooctane sulfonic acid") AND ("livestock" OR "chicken*" OR "hen" OR "cattle" OR "cow" OR
"cows" OR "swine" OR "pig" OR "pigs") AND ("uptake" OR "accumulation" OR "transfer" OR
"bioaccumulation" OR "biotransfer" OR "toxicokinetic*")
Date searched: 8/1/2024. No date limitations on results.
Results: 58 studies
The following criteria must be met:
• Measured PFOA or PFOS concentrations in the exposure media
• Measured PFOA or PFOS concentrations in the finished animal product (i.e., meat, milk, eggs)
DRAFT
56
-------
• Durations of exposure relevant to common agricultural practices and environmental exposures
(i.e., durations that reach steady state or replicate the typical lifespan of the livestock before
slaughter)
• Exposures in a media relevant to environmental exposures (i.e., water, feed, soil)
The following criteria are preferred:
• Known rather than estimated intake rates of contaminated media
• Larger sample sizes
The results of this BTF selection effort are described below.
Eggs and Chicken Meat
Wilson et al. 2020
Overview: This study was a controlled laboratory study that included 119 laying hens. All hens were 30-
weeks old at the beginning of the study period. The hens were divided into 5 groups (22 hens in the
control group and lowest concentration treatment group and 25 hens in remaining 3 treatment groups).
Hens were exposed to PFOA, perfluorohexane sulfonic acid (PFHxS), PFOS, and PFHxA via drinking water
at concentrations of 0, 0.3, 3, 30, and 300 ng/L for 61 days. Eggs were collected throughout the
treatment period. At the end of the treatment period, the treatment group hens were given PFAS-free
water for 30 days. Eggs were also collected and analyzed during this depuration period. No meat
samples were collected in this study.
Results: No negative health, welfare, or behavioral changes in the hens over the course of the study
were noted. A subset of eggs was sampled to analyze the relative distribution of PFAS in egg yolk and
albumen (egg white). Over 99% of the PFOA and PFOS present in eggs were distributed to the egg yolk
rather than the albumen, consistent with data of Kowalczyk et al., 2020. For all hens in the treatment
groups, egg concentrations of PFOA and PFOS increase until days 24-30. After this initial increase, PFOA
and PFOS concentrations in eggs reached apparent steady state until the cessation of treatment.
PFOA concentrations in whole eggs ranged from 500 to 400,000 ng/kg and for PFOS ranged from 800 to
1,000,000 ng/kg, with concentrations proportional to dose. Daily intake rates ranged from 40 to 47,000
ng/day for PFOS and 57 to 54,000 ng/day for PFOA, depending on the treatment group. Biotransfer
factors (BTFs) were calculated for each treatment group using the calculated average intake rate and the
average egg concentration during the steady-state period. The average BTF for PFOS was 21 day/kg
(ranges from 19-24 day/kg) and the average BTF for PFOA was 8.6 day/kg (ranges from 8.1-9.2 day/kg).
Uncertainties: This study was well controlled with limited uncertainties. Though there were several
quality control metrics included in the study, this study still includes some uncertainty in the PFAS
exposure for the treated hens. Hens were only included in this study if they had non-detectable levels of
the four studied PFAS in their eggs prior to the start of the study, ensuring that no prior exposure was
impacting the results. The feed and bedding material for the animals was tested and confirmed to be
free of the studied PFAS. Because this study used drinking water for exposure that was prepared in-lab,
it is known to not contain precursors to PFOA and PFOS. Overall, this is a high-quality study with a large
sample size.
Kowalczyk et al. 2020
Overview: This was a laboratory study which included 12 laying hens. The hens were 6 months old at the
beginning of the study and were fed a combination of highly contaminated hay (harvested from a field
that received contaminated biosolids and paper-derived compost in southern Germany) and barley for
25 days. The barley and hay were analyzed for 14 PFAS, and TOP assay (Gockener et al., 2020). TOP
assay converts oxidizable precursors of PFOA and PFOS to PFOA. Of the 14 PFAS analyzed, 12 were
DRAFT
57
-------
below the limit of quantification in the hay and barley. The PFOA concentrations were 0.8 ng/kg in
barley and 287 ng/kg in hay, for a combined average intake rate of 0.6 ng/day per hen. The PFOS
concentrations were 2.5 ng/kg in barley and 1,654 ng/kg in hay, for a combined average intake rate of
2.8 ng/day per hen. After TOP analysis, PFOA levels of the mixed feed increased 786%; again, note that
TOP analysis oxidizes PFOA and PFOS precursors to PFOA. After the 25-day feeding period, 4 hens were
slaughtered and 8 were fed a non-contaminated diet until study day 67. At this point, the remaining
hens were slaughtered. Muscle, liver, and kidney samples and egg yolks were then analyzed.
Results: This study did not record any treatment related adverse health effects in the hens. In the subset
of eggs where both yolk and albumen were tested, over 99% of PFOA and PFOS present in egg were
measured in the egg yolk. Over the duration of the exposure period, concentrations of PFOA and PFOS
increased rapidly from days 0-10, with slower increases for the remaining 15 days of the exposure
period.
At exposure day 25, the average PFOA concentration in egg yolks was 18.6 ng/kg wet weight (ww),
which corresponded to an average total egg concentration of 5.2 ng/kg ww. At exposure day 25, the
average PFOS concentration in egg yolks was 560 ng/kg ww, which corresponded to an average total egg
concentration of 157 ng/kg ww. Using the reported feed intake rates, egg BTFs were calculated for PFOA
at 8.7 day/kg and for PFOS at 56 day/kg. After TOP assay of the egg yolks, PFOA concentrations
increased 647%.
At exposure day 25, the average PFOS concentration in hen muscle was 6.2 ng/kg ww and the average
PFOA concentration was 0.3 ng/kg ww. These concentrations correspond to a muscle BTF of 2.2 day/kg
for PFOS and 0.5 day/kg for PFOA. Hen liver concentrations for PFOA and PFOS were significantly higher
than muscle concentrations (3.7 ng/kg ww for PFOA and 72.3 ng/kg ww for PFOS). TOP assay was not
performed on muscle samples.
Uncertainties: This study has several limitations that could influence how the results are interpretated.
The hens in this study were exposed for 25 days. In Wilson et al. 2020, the daily egg concentrations of
PFOA and PFOS increased during the beginning of the treatment window and did not stabilize until
treatment day 24-30, depending on the treatment group. In the Kowalczyk et al. (2020) study, egg
concentrations appeared to level-off after day 10 of exposure but continued to trend fractionally
upward until the end of the treatment phase on day 25. For this reason, the relatively shorter exposure
duration in this study could result in a slight underestimate of the BTF. The exposure media in this study
also contains significant concentrations of precursors, a fraction of which appeared to transform to
PFOA or PFOS in the hens or eggs (as previously described, the TOP analysis increases PFOA
concentrations 786% in feed, but only 647% in eggs, which indicates that though the majority of
precursors are transferred to eggs intact, a fraction appear to have transformed to their terminal
degradates of PFOA or PFOS). The presence of precursors in feed could thus result in a slight
overestimate of BTFs. Although this study has some uncertainties, the overall study quality is sufficient
for quantifying BTFs in eggs and meat.
BTF Selection for Eggs and Chicken
Eggs: Either the Wilson et al. (2020) study or the Kowalczyk et al. (2020) study would be sufficient for
estimating BTFs in eggs. The fact that PFOA BTFs from both studies are nearly identical (8.6 and 8.7)
increases confidence in these results. The PFOS BTF calculated from data of Kowalczyk et al. (2020) were
approximately three times higher than those calculated from data of Wilson et al. (2020) (21 compared
to 56). Kowalczyk et al. (2020) hypothesized that this discrepancy is due to the presence of significant
levels of PFOS precursors in the hay used in the study. It is possible that some PFOS precursors present
in the feed used in Kowalczyk et al. (2020) readily biotransformed to PFOS in the hens, while other PFOA
or PFOS precursors are passed to the egg yolk without transformation. Given that the Wilson et al.
DRAFT
58
-------
(2020) study has a significantly larger sample size than Kowalczyk et al. (2020) and that this study does
not have the compounding variable of precursors in the feed ingredients, this assessment will use the
BTFs from Wilson et al. (2020) of 21 day/kg for PFOS and 8.6 day/kg for PFOA.
Meat: Kowalczyk et al. (2020) is the only study available to quantify BTFs for PFOA and PFOS in chicken
meat. There was significantly less transfer of PFOA and PFOS to laying hen muscle compared to egg yolk,
which aligns with other studies reporting lower PFOA and PFOS concentrations in chicken meat than in
eggs (Braunig et al., 2017; EFSA CONTAM Panel, 2020; Lasters et al., 2023). Importantly, this study only
analyzes meat from laying hens, which is feasibly consumed by those with backyard flocks, but generally
used commercially only in making processed foods and canned soup. Broiler chickens, which are grown
for meat production and are generally slaughtered at 6-10 weeks after hatching, may have different
accumulation rates than laying hens. For example, the elimination pathway of egg laying is not available
to broiler chickens and all laying hens are female, while broiler chicken flocks contain both sexes.
Chickens raised for meat also have a shorter lifespan than laying hens. The USDA is currently conducting
a PFAS uptake study on broiler chickens; until these data are available, the chicken meat BTFs presented
in this assessment have uncertainties when applied to meat consumed from animals other than laying
hens. Though Kowalczyk et al. (2020) may overestimate the PFOS BTF due to known contamination of
feed with PFOS precursors, the study is nevertheless of sufficient quality to quantify a BTF for this
assessment. The BTF is 0.2 day/kg for PFOA and is 2.2 day/kg for PFOS. Again, the current BTFs,
calculated from laying hens, are appropriate for this assessment scenario where a farm family primarily
has hens for egg production, but occasionally slaughters hens for food (for example, at the end of their
laying life). This scenario is not relevant to commercial food operations raising broiler hens for meat
production. This assessment is also not considering intake and exposure from consumption of animal
livers; given the elevated levels observed in livers, this may be an important pathway of exposure for
those who consume liver. More data are needed on PFOA and PFOS uptake into breeds of chickens
more typically used for meat.
Beefand Milk
Vestergren et al. 2013
Overview. Vestergren et al. (2013) reported an observational study of milk and meat from a dairy farm
in Sweden. The farm had 92 Swedish Red dairy cows of varied ages that had consumed silage, corn, and
barley grown on the farm and drinking water from a groundwater well. The farm was not known to have
any PFAS point sources (such as contaminated sludge application) and was selected as a representative
"background" dairy farm for Sweden, meaning that PFOA and PFOS contamination is thought to be
caused only by long-range atmospheric transport and deposition. The cows at this dairy were mainly
confined to a barn but were allowed to graze on a pasture during the summer months. This study did
not quantify PFAS soil concentrations and milk was only sampled during the months that the animals
were confined to the barn. The average intake of PFOA and PFOS for the cows was estimated using feed
PFAS concentrations and consumption rates of silage, corn, barley, and water. Pooled milk samples were
collected from the milk storage tank monthly (from November to April), on the same days when
representative samples of feed were collected. Over the course of the study, five cows were
slaughtered, and muscle, liver and whole blood were analyzed for PFAS. Given that the animals had
been living on the contaminated farm for their entire lives, PFAS concentrations in the animals are
assumed to be at steady state.
Results: Daily intake rates were estimated to be 613 ng/day for PFOA and 303.6 ng/day for PFOS, based
on the measured PFOA and PFOS concentrations in feed, water, and supplements and assumed
consumption rates for each category. The supplements at this dairy were purchased from a supplier and
found to have no detectible PFOA and PFOS, but the authors assumed that supplements contained PFOA
DRAFT
59
-------
and PFOS at Vz the MDL for each analyte. PFOA and PFOS concentrations in pooled milk samples were
6.7 and 6.2 ng/L, respectively. Muscle of the five slaughtered cows contained, on average, 7 ng/kgfor
PFOA and 21 ng/kg for PFOS. The milk BTFs were 0.01 day/kg for PFOA and 0.02 day/kg for PFOS. The
muscle BTFs were 0.01 day/kg for PFOA and 0.07 day/kg for PFOS.
Uncertainties: This study site is a farm where there are no known proximal sources of PFAS
contamination (and no known history of sludge application to the pastureland), indicating that the
source of PFAS is long-range atmospheric transportation and deposition; the study included
measurements of PFOA, perfluoroheptanoic acid (PFHpA), perfluorohexanoic acid (PFHxA), PFNA, PFDA,
PFUnDA, PFDoDA, PFHxS, and PFOS. It is likely that PFOA and PFOS precursors were present in the
water, feed, and soil around this farm, but it is unknown the degree to which precursors could impact
the calculated BTFs. If precursors to PFOA and PFOS were present in the feed, water, and/or soil, this
would result in overestimation of the BTFs. This study also did not attempt to quantify the average
annual intake from soil that cows consume during the summer months when cows were pastured. Not
including soil as a potential intake pathway could result in overestimated BTFs. The intake rates in this
study are also somewhat uncertain because they are estimated from the farmer's assumptions about
the intake rates for his cows (for example, the farmer communicated that his cows consume about 50
L/day of drinking water). Though this study is large for milk (92 cows' milk was pooled and analyzed),
only 5 cows were slaughtered for the muscle analysis. Overall, this study is of sufficient quality to use
quantitatively in the assessment but includes some areas of uncertainty that would likely result in
overestimates of BTFs for dairy cows.
Kowalczyk et al. 2013
Overview: Six lactating Holstein cows housed at the German Federation for Risk Assessment were fed a
PFAS-contaminated diet for 28 days. After the exposure period, three cows were slaughtered and the
remaining three were fed a PFAS-free diet for an additional 21 days. The diet was mixed from PFAS-
contaminated grass silage and hay harvested from a contaminated farm in Lower Saxony (the same
materials used in Kowalczyk et al., 2020). The cows were housed in individual tie-stalls and their intake
of feed was quantified each day. Meat and milk samples were analyzed for PFOA, PFOS, PFHxS, and
perfluorobutanesulfonic acid (PFBS). The serum half-life of PFOS in beef cattle was estimated to be 116
days (Lupton et al., 2015); with an exposure of only 28 days, PFOS concentrations in muscle and milk in
this study are not expected to represent steady-state concentrations. In contrast, the estimated half-life
of PFOA in cattle is estimated to be 19 hours (Lupton et al., 2012) and dairy cows having lifetime
exposures to PFAS contaminated feed and water appeared to have rapidly cleared PFOA, as evidenced
by no accumulation above quantification limits in serum (Lupton et al., 2022).
Results: There were no noted adverse effects in the cows over the duration of the study. Average PFOS
concentrations measured in grass silage and hay were 200 and 1,924 ng/kg while average PFOA
concentrations in grass silage and hay were 79.3 and 333 ng/kg. Consumption of grass silage and hay
were 8.9 and 1.4 kg/day, respectively corresponding to an average intake rate of 1,172 ng PFOA/day and
4,472 ng PFOS/day. Note that the PFOA and PFOS concentrations in Table 1 of Kowalczyk et al. (2013)
are switched; this mistake has been confirmed with the study's lead author. During the exposure period,
PFOS concentrations in milk increased at a steady rate. Once the exposure period ended, the three cows
fed a PFAS-free ration had milk PFOS concentrations similar to the level reached on the last day of
exposure, indicating that depuration of PFOS in milk was slow. During the exposure period, PFOA
concentrations in milk did not exceed the LOD (0.1 ng/L) until around exposure day 10. PFOA
concentrations in milk then hovered near the LOD until the exposure period ended. In cows fed PFAS-
free feed after exposure, PFOA concentrations in milk were non-detectable. For PFOS, mean milk
concentrations were calculated using the milk from the last day of the exposure period and the
DRAFT
60
-------
depuration period (study days 29-45). Mean PFOS concentrations in milk were 32.1 ng/L. This resulted in
a BTF of 0.007 day/kg. PFOA was only detected in milk during the latter part of the exposure period, and
not in milk collected during the depuration period. Using the average detected concentration of PFOA in
milk during the exposure period, a BTF of 0.00006 (6 x 10"5) day/kg is calculated. Thus, this study
indicates that PFOA had very limited accumulation into milk over the given exposure period.
The average PFOS muscle concentration in the three animals slaughtered after the exposure period was
145 ng/kg for PFOS. After the exposure period ended, the PFOS concentration measured in muscle
tissues increased to 178 ng/kg. The average muscle concentrations of all animals (slaughtered at study
day 29 and 45) was 161.5 ng/kg. The BTF for PFOS in muscle calculated using this average value is 0.036
day/kg. The average PFOA muscle concentration measured in the three animals slaughtered after the
exposure period was 0.6 ng/kg. PFOA was not detectable in the remaining three animals slaughtered
after the depuration period. A BTF calculated using only the three animals slaughtered on study day 29
is 0.00006 (6 x 10"5) day/kg. Thus, this study also indicated that PFOA had very limited accumulation in
dairy cow meat.
Uncertainties: The 28-day PFAS exposure period in this study is not long enough for PFOS to reach
steady-state concentrations; Lupton et al. (2012) demonstrated that steady state concentrations of
perfluorosulfonic acids (PFSAs) in dairy animals were not met until after about 1.5 years of exposure.
Therefore, BTFs calculated from this study will underestimate accumulation of PFOS. PFOA
concentrations in milk and meat in this study were below, or near, the limits of detection, which
indicates that PFOA does not significantly accumulate in meat or milk of dairy cows and adds to the
uncertainties of these values. The feed used in this study is the same feed that is used in the Kowalczyk
et al. 2020 chicken study; it is known that this feed contained significant concentrations of PFOS
precursors. The presence of PFOS precursors in the feed may explain why PFOS concentrations in milk
and meat are elevated even after the exposure period ends. Another potential explanation of this
observation is that PFOS can be stored in other compartments of cattle, like skin, which could result in
ongoing excretion through milk even after exposure through feed and water has ended (Lupton et al.,
2022). The presence of PFOS precursors in feed would result in underestimations of BTFs. Finally, this
study has a relatively small sample size of six animals.
Drew et al., 2021; 2022
Overview: Drew et al. (2021) reported on the accumulation of PFAS in Belted Galloway beef cattle and
mixed breeds of sheep raised for meat on a hobby farm in Australia that had water contaminated with
AFFF from a nearby facility. This discussion will focus on accumulation results for cattle. The study was
split into two phases, each approximately one year long, with one year between phases. The first phase
occurred before remediation activities were taken to reduce PFAS levels in the livestock drinking water
by attempting to divert contaminated water from the neighboring property away from the farm. The
second phase took place after this remediation activity. The only source of feed for the cattle during the
duration of the study was forage.
During phase one, soil (n = 36) and grass (n = 5) samples were collected from the forage area. Drinking
water was measured two times, at the beginning of phase one and near the end of phase one; this
sampling only monitored for four PFAS and is not reported. Serum levels from 5 cows (9-14 years old)
and 9 cattle (2-22 months old) were collected.
During phase two, the stock water for the cattle was sampled again, and serum from 19 cattle were
collected (all adults over 1.5 years old, three were steers and the remaining were heifers and cows). The
19 cattle included in phase two were moved to a research facility 19 days before the last blood sampling
event in this study. At the research facility, five of the animals had PFAS blood monitoring for 214 days
post removal from the farm; 11 animals were euthanized on day 63 post removal from the farm and
DRAFT
61
-------
PFAS levels were measured in tissues (these results are presented in a different study, Drew et al.,
2022). Using the tissue concentration data and the elimination half-lives, PFAS levels in tissue could be
estimated for the animals at the time of transfer from the farm. The water remediation activity was
completed Fall of 2016, and the serum samples were collected in Spring of 2018; the authors assume
that PFAS concentrations in the animals are at steady state throughout phase two of blood sampling.
Results: There were no adverse effects in cattle or sheep reported in this study. Phase one soil sampling
found that PFOS had a mean concentration 0.003 mg/kg dry weight (3 ng/kg dry weight) in soil; PFOS
was not detected at quantifiable concentrations in grass (LOQ = 0.0005 mg/kg wet weight, equivalent to
0.5 ng/kg). PFOA was not present at a measurable level in soil or grass. Water results from phase one
are not presented in the publication. Because the accumulation calculations were only conducted on
animals included in phase two, the discussion of results will focus on water and serum levels for those
19 cattle.
During phase two, the mean water concentrations were 3.0 \x.g/\ (3,000 ng/L) for PFOS and 0.87 ng/L
(870 ng/L) for PFOA. Water concentrations of PFOA and PFOS did not significantly differ between the
two sampling dates. During the phase two study (while the cattle are on the contaminated farm), serum
levels of PFOS range from 275 to 455 ng/mL while PFOA was consistently non-detectable in all serum
samples. Drew et al., 2019 (the companion study reporting data collected at the research facility) found
that the serum half-life for PFOS in these cattle was 74 days. This study also found that the partitioning
coefficient from serum to muscle for PFOS was 0.072 ± 0.02 on day 62 (the transfer day to the research
facility) and 0.08 ± 0.03 on day 215. The authors of this study calculate steady-state serum
concentrations of 436.2 ± 59.0 ng/ml. Using the steady-state serum concentration, the median muscle
partitioning coefficient (0.076), the mean concentration in drinking water, and an assumed drinking
water consumption rate of 59.8 L/day (Drew et al., 2021), a biotransfer factor of 0.18 day/kg is
calculated for PFOS. This BTF assumes that all intake is derived from contaminated drinking water. No
BTF is derived for PFOA because the study finds that there is no measurable accumulation of PFOA into
cattle serum.
Uncertainties: Overall, there are some uncertainties in this study from the lack of precise information on
the amount of PFOS intake in the 19 phase two cattle used to derive BTFs. Because the pre-remediation
activity drinking water concentrations are not presented, it is not clear if residual PFOS loading from
original drinking water source could be continuing to impact the serum levels of phase two cattle. The
transfer factor calculation also assumes negligible intake of PFOS from grass and soil; this assumption
may lead to an overestimate of BTF. The BTF calculation also assumes a drinking water intake for the
cattle based on the climate and weight of the cows, rather than a measured drinking water intake.
Overall, this study is of sufficient quality to calculate BTFs.
Johnston et al. 2023 & Lupton et al., 2022
Overview: Johnston et al. (2023) and Lupton et al. (2022) measured blood, ear notch (skin), and muscle
PFAS concentrations in dairy cattle from a farm containing AFFF-contaminated drinking water in New
Mexico. Silage at this farm was also sampled and confirmed to contain PFAS. Blood and ear notch
samples were collected from 175 cattle on the farm. Thirty of these cows (10 heifers, 15 lactating, and 5
dry) were moved to an uncontaminated research facility (New Mexico State University). Two weeks
after the move, 20 of the cows were euthanized and necropsied, with blood plasma and tissues
analyzed. In the remaining 10 cows, blood samples were collected every two weeks. The two oldest
cows died during the study period. Finally, the 8 remaining cows were euthanized and necropsied at
either 137 or 153 days after arrival to the research facility. In all, paired blood and muscle data are
available from 28 cows.
DRAFT
62
-------
Results: Though two cows died during the course of the study, these were the oldest cows in the cohort
and there were no reported adverse effects in the cows. At the contaminated farm, the mean
concentration of PFOS was 818 ng/kg in water and in 3,482 ng/kg in silage. PFOA results are not
reported for the contaminated media at the farm or in the blood and muscle samples. This study does
not report an observed or estimated feed or water consumption rates for the cows on this farm. To
calculate the total intake of PFOS through silage and water, the reported silage consumption rate in
Vestergren et al. (2013) is used (38.5 kg/day) and the general estimated drinking water intake for dairy
cows is used (92 L/day, US EPA 2003g). This calculation estimates a total PFOS intake from silage and
water to be 209,313 ng/day.
Serum levels of PFOS in the cows decline over time as the animals depurate PFOS at the research facility.
There is a log-linear relationship between PFOS levels in plasma and muscle. The total PFOS
concentrations in plasma and muscle in each of the 28 cows with this data available are reported in
Supplemental Information Table B (in this table, the heifers are reported as "young"). These data are
used to calculate partitioning coefficients between plasma and muscle for each cow. The partition
coefficient from plasma to muscle ranges from 0.03 to 0.11. Dry and lactating cows have similar
partitioning coefficients, but the heifers have lower PFOS partitioning to muscle (a smaller partitioning
coefficient). Because most of the young cows have not reached a steady-state serum level, the average
of partitioning coefficient excluding the young is used for further calculations. This mean partitioning
coefficient is used to estimate the muscle concentrations on the dry and lactating cows using the plasma
concentrations at the time of removal from the farm. The mean muscle concentration in these dry and
lactating cows is 7.3 ng/g. Using this mean muscle concentration and the estimated PFOS intake rate,
the BTF for PFOS in muscle in dairy cows is 0.035 day/kg.
Uncertainties: These studies did not include any PFOA results above a limit of quantification. This study
is also focused on muscle uptake in dairy cows, which are not commonly used for beef production. As
described previously, it is expected that there are significant differences in BTFs for dairy cows and
cattle primarily used for beef consumption. Also, this study does not include information about the
amount of water and feed consumed by the animals at this farm, and the estimate of total intake does
not include any exposure from soil.
Chou et al. 2023
Overview: This study describes the development of a generic physiologically based pharmacokinetic
(PBPK) model for adult beef cattle and lactating dairy cows useful for estimating tissue and milk
distribution and depuration rates of PFAS. The generic beef cattle model consisted of four tissue
compartments including liver, kidney, muscle, and the rest of the body (notably, the model does not
include a compartment for plasma). The generic model structure for dairy cows is the same as the beef
cattle model, but also includes udder and milk compartments. Physiological parameters, including body
weight, cardiac output, fractions of blood flow to tissues, and the volume fractions of individual organs
were collected from a previous review article that summarized published experimental data in various
breeds of beef and dairy cattle (Swedish Red, Holstein Friesian, Belted Galloway, Australian Lowline,
American Angus, and Japanese Black). Chemical-specific parameters included protein binding,
absorption/elimination rate constants, partition coefficients, enterohepatic circulation, and renal
reabsorption parameters. There were not chemical specific data for PFOA and PFOS in cattle and cows;
instead, these chemical-specific values were parameterized using a previously published PBPK model for
PFOA and PFOS in rats. The model can consider intake from soil and water. The model outputs muscle
and milk concentrations over time. The final PBPK model was coded as a R-Shiny application and is
available online.
DRAFT
63
-------
Results: The authors validated their models against an independent PFOA and PFOS dataset in Chinese
beef and milk that was not used to parameterize the model. Because these datasets included final food
products and did not include information on exposure of the animals, the exposure was estimated using
data describing water and soil PFAS concentrations in China. The model was also validated using the
tissue results reported by Kowalczyk et al. (2013). The model was generally within a two-fold error range
of the observed PFOA and PFOS concentrations in all compartments except for PFOA in milk, which was
underestimated by the model (compared to PFOA concentrations observed in milk and yogurt products
in China), and PFOS in muscle, which was overestimated by the model.
This model could be used to calculate BTFs by setting a concentration of PFOA and PFOS in soil and
water, calculating intake rates using the consumption rates employed in the model, and comparing the
estimated milk and muscle concentrations after two years of exposure to the calculated intake rates.
Two years of exposure was selected as the time window to estimate BTFs because this is the age that
beef cattle are slaughtered and that dairy cows generally enter milk production. This exposure time is
also more than sufficient to reach steady state in cows and cattle (Lupton et al., 2012; Lupton et al.,
2015). To simplify this calculation, the authors assumed that water was the only source of PFOA and
PFOS exposure and set the water concentration to 2 ng/L for both chemicals.
Using these assumptions, the model estimated that PFOS in muscle after a 2-year exposure would be
12.2 ng/g in beef cattle and 65.9 ng/g in dairy cows. The modeled estimated PFOA muscle after a 2-year
exposure would be 0.0253 ng/g in beef cattle and 0.166 ng/g in dairy cows. This results in muscle BTFs
for PFOS of 0.09 and 0.30 day/kg in beef cattle and dairy cows, respectively. Muscle BTFs for PFOA are
0.0002 and 0.001 day/kg in beef cattle and dairy cows, respectively.
The model also calculated PFOA and PFOS concentrations in milk. After 2 years of exposure, the
predicted PFOA concentration in milk was 0.03 ng/mL and the predicted PFOS concentration in milk was
24.4 ng/mL, resulting in milk BTFs of 0.0001 and 0.11 day/L for PFOA and PFOS, respectively.
Uncertainties: The BTFs generated from this publication's data are based on modeled, not measured
concentrations in exposure media and animal products. The BTFs presented from this study therefore
represent estimates from a PBPK model with uncertainties in many of the parameters. There are
significant uncertainties in the results.
Xiao et al. 2024
Overview: This study measured PFAS concentrations in feed and raw milk from 92 dairy farms across 20
provinces of China. At 70 of these farms, the researchers were also able to measure PFAS in the cow's
water. Researchers calculate the "carry over rate" (COR) for PFOA and PFOS, which is defined as the
mass of the chemical eliminated through milk secretion divided by the mass of the chemical consumed
through feed and water. This COR can be converted to a BTF by dividing the value by the assumed milk
secretion rate (kg/day). This study calculated CORs using an assumed daily consumption rate of silage of
20.6 kg/d dry weight, an assumed drinking water consumption rate of 83.6 L/d, and an assumed milk
production rate of 26.5 kg/d. For the 22 farms where no drinking water data were available, the
researchers used the mean drinking water values for PFOA and PFOS in the farm's region to calculate a
COR.
Results: PFOA water concentrations ranged from non-detect to 113 ng/L (mean of 5 ng/L, detection rate
of 79%) and PFOS water concentrations ranged from non-detect to 18 ng/L (mean of 0.7 ng/L, detection
rate of 53%). Feed concentrations ranged from non-detect to 10.6 ng/g for PFOA (mean of 0.7 ng/g,
detection rate of 35%) and non-detect to 0.45 ng/g for PFOS (mean of 0.08 ng/g, detection rate of 47%).
Raw milk concentrations ranged from non-detect to 500 ng/L for PFOA (mean of 80 ng/L, detection rate
of 57%) and from non-detect to 160 ng/L for PFOS (mean of 20 ng/L, detection rate of 62%). The
DRAFT
64
-------
researchers did not provide COR results for each farm, instead reporting mean intakes, mean excretions,
and mean CORs for each chemical. The mean COR for PFOA is 15.78, which equates to a BTF of 0.006
day/kg. The mean COR for PFOS is 29.58%, which equates to a BTF of 0.01 day/kg.
Uncertainties: This study has uncertainty in the intake rates because there are assumed consumption
rates for food and water that are not specific to the farm or region and because 22 of the 92 included
farms had assumed rather than measured drinking water concentrations. Because farm-specific data
were not included in the publication, it was not possible to recalculate CORs for only farms with
measured drinking water intakes. This study finds that the majority of intake to cows at the farms
included in this study derived from feed rather than water, which reduces the impact of the
uncertainties regarding drinking water exposures to the cows. This study does not include
measurements of PFOA or PFOS precursors in feed, water, or milk. Overall, this study includes a large
number of farms and finds similar BTFs for PFOA and PFOS as are derived from Vestergren et al. (2013).
BTF Selection for Milk and Beef
Milk: Kowalczyk et al. (2013) is not suitable as a basis for BTFs for PFOS due to the short exposure time
of the study, which will result in an underestimate of PFOS accumulation in a farm scenario. This study
also has PFOA levels in milk that are below or around the detection limit, leading to uncertainty. The
Chou et al. (2023) PBPK model is parameterized using PFOA and PFOS-specific constants derived from
rat studies. There are obvious physiologic and significant differences in these values between rats and
cows. For example, the plasma half-life of PFOA in cattle is <24 hours (Lupton et al., 2012) whereas the
plasma half-life of PFOA in male rats is 16 days (DeSilva et al., 2009). This modeling study is thus too
uncertain to be used in deriving BTFs for the assessment. Both Vestergren et al. (2013) and Xiao et al.
(2024) are potential candidates for BTFs in milk. Though the Vestergren et al. (2013) study has some
uncertainties regarding the presence of precursors and potential impacts of soil ingestion from grazing,
overall, this is the best available study for deriving BTFs because Xiao et al. (2024) includes assumed
rather than measured drinking water concentrations for some of the farms that are included in the
reported summary statistics. Notably, Kowalczyk et al. (2013) and Chou et al. (2023) indicate that PFOA
accumulation in milk is close to zero, while Vestergren et al. (2013) and Xiao et al. (2024) find that PFOA
accumulation in milk is only two-fold less than PFOS accumulation. Additional studies of PFOA
accumulation into milk would improve our understanding of potential exposure risks for this pathway.
Beef Cattle: Much of the same rationale for study selection of BTFs for beef applies as did for milk. Given
the limitations of Kowalczyk et al. (2013), Johnston et al. (2023) and Lupton et al. (2022), and Chou et al.
(2023), the best studies for quantifying BTFs in beef are Vestergren et al. (2013) and Drew et al. (2021).
Note that the Vestergren et al. (2013) study measured muscle concentrations from lactating cows, not
cattle raised for beef production. As illustrated by Chou et al. (2023) and Drew et al. (2021), different
BTFs would be expected for lactating cows and beef cattle, in part due to the added excretion pathway
of milk production in lactating cows; In fact, Drew et al. (2021) finds significantly higher BTFs for PFOS in
cattle than Vestergren et al. (2013) and Johnston et al. (2023) found for meat in dairy cows. Because
Drew et al. (2021) is a high-quality study measuring uptake into breeds used for beef production, it is
selected for the PFOS BTF in beef. Drew et al. (2021) did not find that PFOA accumulates to measurable
levels in serum of cattle used for beef production. However, Vestergren et al. (2013) does find
measurable levels of PFOA in beef from culled dairy cows. While extrapolating the PFOA BTF measured
in dairy cows to more commercially relevant beef production settings introduces significant uncertainty,
the PFOA BTF calculated from Vestegren et al. (2013) represents the best available estimate for PFOA
uptake at this time. More studies are needed on the uptake of PFOA and PFOS into breeds typically used
for beef production.
DRAFT
65
-------
Pigs
Numata et al. 2014
Overview. Three groups of fattening pigs (10 gilts, 10 barrows, and 10 young boars) were housed at the
German Federation for Risk Assessment. In each group 8 were fed a PFAS-contaminated diet and 2 were
fed PFAS-free feed. Feed intake was restricted to 2 kg/day per hog with an exposure period of 21 days.
The diet was mixed with PFAS-contaminated hay harvested from a contaminated farm in Lower Saxony
(the same hay used in Kowalczyk et al. 2020). Representative feed samples were analyzed for PFAS
content on five separate exposure days. Plasma samples were taken on five days throughout the
exposure period and the day of slaughter (day 22). Urine samples were also collected sporadically
throughout the sampling period, with an average of 2.5 urine samples collected per pig. Muscle, plasma,
urine, and organs were analyzed for 12 PFAS. The results were used to parameterize a PBPK model.
Results: No adverse health impacts of test animals were reported by the study authors. Serum levels of
PFOA and PFOS increased throughout the duration of the 21-day study. Based on the plasma
measurements taken during the study, the authors estimate that the elimination half-life of PFOS is 634
days, significantly longer than the exposure timeframe of this study. The elimination half-life for PFOA is
236 days, also significantly longer than the exposure timeframe of this study. Given that the
concentrations of PFOA and PFOS in serum did not level off during the exposure duration of the study, it
is not possible to extrapolate tissue concentrations at 180 days, which is generally the time when pigs
are slaughtered.
Uncertainties: BTFs calculated from tissue concentrations at day 22 of exposure would significantly
underestimate uptake. Similarly, BTFs calculated using the PBPK model presented in this study would
represent steady-state conditions, which were not reached by the time of slaughter. BTFs calculated
using the PBPK would therefore significantly overestimate risk. Additional studies are needed to
understand BTFs exposure durations expected in the conceptional model of an agricultural setting.
BTF Selection for Pork
Only one study was available in pigs and this study was not sufficient to calculate BTFs for PFOA and
PFOS. For this reason, pigs are not included in the farming models in this assessment.
Overview of Livestock Uptake Parameters
Table 14. Selected Livestock BTFs
PFOA BTF
PFOS BTF
Livestock type
Product
(day/kg)
(day/kg)
Study
Chicken
Meat
0.2
2.2
Kowalczyk et al. 2020
Chicken
Eggs
8.6
21
Wilson et al. 2020
Cows
Beef
0.01
0.18
Vestergren et al. 2013 for PFOA; Drew et
al., 2021 for PFOS
Cows
Milk
0.01
0.02
Vestergren et al. 2013
2.9.3.6 Livestock Dietary Intakes
The produce, meat, and milk exposures will be evaluated using the methodology found in HHRAP (US
EPA, 2005), developed for hazardous waste combustion facilities. That methodology includes
recommended input values for many, but not all of the livestock diets included in this assessment. For
example, HHRAP does not evaluate water consumption by livestock, which was considered an
insignificant pathway for combustor emissions. It is known that PFOA and PFOS can be present in
groundwater and surface water, so that pathway was included in this analysis.
DRAFT
66
-------
There are no data available on PFOA and PFOS bioavailability to livestock specifically from feed, water,
or soil; this assessment assumes 100% is available when orally ingested. The studies used to derive BTFs
for livestock include a variety of exposure scenarios for the experimental animals. In some cases, the
animals are exposed through water only, in other cases the animals are exposed through feed only, and
in other cases the animals are sampled from a pasture farm where they have exposure from feed, water,
and soil. When comparing the PFOA BTFs derived for chicken eggs from the Wilson et al. (2020) study
(animals exposed only through water) and the Kowalzak et al. (2020) (animals exposed only through
contaminated feed), the calculated BTFs are nearly identical. This indicates that if there is a reduced
bioavailability of PFOA in chicken feed, that effect is likely negligible. In the case of dairy cows, the BTFs
selected for this study (from Vestegren et al., 2013) were derived by calculating the exposure from feed
and water combined. If there were a reduced bioavailability of PFOA or PFOS in feed in dairy cows, this
would already be factored into the BTF calculation. For beef cattle, the BTFs were also derived using
data from pasture-fed cows (Vestegren et al., 2013 and Drew et al., 2021), so these factors also
inherently consider differences in bioavailability between feed and water in the calculated values. Note
that part of the reason previous assessments included assumptions about reduced bioavailability in feed
compared to water is because the BTFs in these assessments were modeled, not measured. By using
BTFs derived from empirical experiments with multiple sources of livestock exposure, the uncertainty
regarding bioavailability across livestock exposure pathways is reduced or eliminated.
Chicken Dietary intake
HHRAP recommends an overall chicken consumption rate of 200 g DW/day and assumes that this is
composed entirely of grain, but this assumption is relevant to broiler chickens. It also recommends a soil
consumption rate of 22 g/day. Laying hens consume less than broiler chickens, generally 100-150 g
DW/day (Alabama A&M & Auburn Universities Extension, 2022). For this analysis, forage, drinking
water, and homegrown hay are relevant exposure sources in the pasture farm scenario, with grain
assumed to be from an uncontaminated, off-site source. Grain for chicken feed was assumed to be
purchased from an uncontaminated source because purchasing feed is more common rather than
growing it locally, likely due to the specific dietary needs of laying hens (Poultry Extension, 2024). The
assumption is further supported by the finding that the grains typically included in poultry feed (oats,
cracked corn) typically have low PFOA and PFOS accumulation (see Section 2.9.3.4). Because HHRAP
assumes that chickens only consume grain, additional data sources for characterizing chicken diets were
also sought. The EPA identified three studies or reports that included information about chicken dietary
intakes.
Kowalczyk et al. 2020
This study on PFAS uptake by chickens and distribution to eggs included 12 hens fed a combination of
highly contaminated hay (harvested from a field that received contaminated biosolids and paper-
derived compost in southern Germany) and barley for 25 days. Kowalczyk et al. provided a detailed
breakdown of the experimental chicken diet: 37% barley (grain), 8% hay, soybean meal (19%), triticale
(28%), oil (1.5%), mineral feed (3%), and calcium carbonate (3.5%). The hens in this experiment were
caged, so no opportunity for consumption of soil or insects occurred.
Dal Bosco et al. 2014
This study analyzed the impact of range enrichment (either sorghum plantings or olive trees) on
behavior and diet of 250 free-range broiler chickens in each of two seasons on two farms (1,000 birds
total). Forage intake was calculated for five subareas at increasing distance from the shelter. Total
forage intake per bird (summed across the different distances from the shelter) for the sorghum-planted
ranges are 30 g dry weight (DW)/day in summer and 18 g DW/day in winter. For the unenriched ranges,
the corresponding values reported are 15 g DW/day for both seasons. However, the authors note that
DRAFT
67
-------
forage intakes for laying hens are likely higher, due to the younger age of broiler chickens compared to
laying chickens. The authors cite several other studies that found values for laying hens in the 30-40 g
DW/day range.
RAAF Base Williamtown, Australia site investigation (AECOM 2017)
The Australian Department of Defense completed an investigation and risk assessment associated with
PFAS contamination around a base that used PFAS-containing firefighting foam. The risk assessment
included a commissioned study of PFOA and PFOS uptake into chicken eggs (Wilson et al., 2020).
Additional supplemental information from that study was published in a Department of Defense report
(AECOM, 2017). This report includes a water intake for chickens of 0.208 L/day.
Diet Selection for Chickens
Starting with the total diet of 200 g DW/day from HHRAP, the diet fractions for silage and grain from
Kowalczyk were applied to obtain intakes of 16 g DW/day of hay and 74 g DW/day of grain. For forage, a
value of 30 g DW/day from Dal Bosco was selected. For soil ingestion, the value from HHRAP was
rounded to 20 g/day and used. Finally, the water consumption rate from the Australian Department of
Defense report of 0.208 L/day was rounded to 0.21 L/day and used (AECOM, 2017). These values are
summarized in Appendix B.
Cow Dietary Intake
HHRAP recommends cattle consumption rates for forage, silage, grain, and soil for both beef and dairy
cattle. For dairy cows, the dietary intake rates are 13.2 kg DW/day forage, 4.1 kg DW/day silage, and 3.0
kg DW/day grain. Water consumption rates for cows vary according to many factors, such as breed,
body size, amount of milk produced per day, air temperature, humidity, and moisture content of feed
(Harris and Van Horn, 1992). An analysis of water intake rates done for the 3MRA modeling system (US
EPA, 2003g) was used to select a water intake of 92 L/day. That value reflects the average of data
measured by Harris and Van Horn (1992) and reflect the variability in water consumption of dairy cows
across different temperatures and milk production rates. This value falls within the water consumption
ranges reported for other cow breeds by the University of Nebraska Cooperative Extension (1998),
which for lactating Holstein cows was reported as 18 to 40 gallons/day, equivalent to 68-151 L/day. For
beef cattle, the dietary intake rates are 8.8 kg DW/day forage, 2.5 kg DW/day silage, 0.47 kg DW/day
grain, and 0.5 kg DW/day soil. The drinking water intake rate for beef cattle is 53 L/day (US EPA, 2003g).
2.9.3.7 Bioaccumulation Factors in Fish
The EPA selected fish bioaccumulation factors for this biosolids risk assessment to be consistent with
draft AWQC for the Protection of Human Health for PFOA and PFOS (US EPA, 2024o,p). The EPA
calculated draft BAFs for the PFOS and PFOA human health AWQC based on each chemical's properties
{e.g., ionization and hydrophobicity), metabolism, and biomagnification potential (US EPA, 2024o,p; US
EPA, 2000a; 2003h).The EPA's national BAFs represent the long-term, average bioaccumulation
potential of a chemical in aquatic organisms that are commonly consumed by humans throughout the
United States (US EPA, 2000a). The EPA evaluated results from field BAF and laboratory BCF studies on
aquatic organisms commonly consumed by humans in the United States for use in developing national
trophic-level BAFs.
To develop the draft BAFs for PFOA and PFOS, the EPA conducted a systematic literature search in
October 2022 of publicly available literature sources to determine whether they contained information
relevant to calculating national BAFs for human health AWQC (US EPA, 2000a; 2003h). The literature
search for reporting the bioaccumulation of PFOA and PFOS was implemented by developing a series of
chemical-based search terms, consistent to the process used in the derivation of BAFs used in the
development of the Final Aquatic Life AWQC for PFOA and PFOS (US EPA, 2024l;m) and described in
DRAFT
68
-------
Burkhard (2021). These terms included chemical names and CAS numbers, synonyms, tradenames, and
other relevant chemical forms (i.e., related compounds). Databases searched were Current Contents,
ProQuest CSA, Dissertation Abstracts, Science Direct, Agricola, TOXNET, and UNIFY (database internal to
the EPA's ECOTOX database). The literature search (including literature published through the first two
quarters of 2020) yielded >37,000 citations that were further refined by excluding citations on analytical
methods, human health, terrestrial organisms, bacteria, and where PFOA or PFOS was not a chemical of
study. The citations meeting the search criteria were reviewed for reported BAFs and/or reported
concentrations in which BAFs could be calculated. Data from papers that met the inclusion and data
quality screening criteria described below were extracted into the chemical dataset.
Specifically, studies were evaluated for inclusion in the dataset used for calculating national BAFs using
the following evaluation criteria:
• Only BAF studies that included units for tissue, water, and/or BAFs were included.
• Mesocosm, microcosm, and model ecosystem studies were not selected for use in calculating
BAFs.
• BAF studies in which concentrations in tissue and/or water were below the minimum level of
detection were excluded.
• Only studies performed using freshwater or brackish water were included; high salinity values
were excluded.
• Studies of organisms (e.g., damselfly, goby) and tissues (e.g., fish bladder) not commonly
consumed by humans or not used as surrogate species for those commonly consumed by
humans were excluded. Information on the ecology, physiology, and biology of the organism
was used to determine whether an organism is a reasonable surrogate of a commonly
consumed organisms.
• Studies in which the BAFs were not found to be at steady state were excluded.
• For pooled samples, averaging BAF data from multiple locations was only considered acceptable
if corresponding tissue and water concentrations were available from matching locations (e.g., a
BAF would not have been calculated using water and tissue samples collected from eight
separate locations with tissue concentrations collected from only six of these corresponding
locations).
In addition to the evaluation criteria listed above, PFOS bioaccumulation data were also evaluated using
five study quality criteria outlined in Burkhard (2021) and shown in Table 15.
DRAFT
69
-------
Table 15. Study Quality Criteria Used by Burkhard (2021)
Criteria
1
2
3
Number of water
samples collected
> 3 samples
2-3 samples
1 sample
Number of organism
samples collected
> 3 samples
2-3 samples
1 sample
Temporal coordination
of water and biota
samples
Concurrent collection of
samples
Collected within a
1 year time frame
Collected > 1 year time frame
Spatial coordination of
water and biota
samples
Collected from same
locations
Collected from
reasonably close
locations
(1 kilometer (km)-
2 km)
Significantly different sampling
locations
General experimental
design
Assigned a default
value of zero for studies
in which tissues from
individual species were
identified and analyzed
Assigned a value of 3 for studies
in which tissues were from
mixed species or reported as a
taxonomic group.
Note: The scores for each BAF were totaled and used to determine the overall confidence ranking for each
individual BAF. The sum of quality values for the five criteria listed in Table 2 were classified as high quality (total
score of 4 or 5), medium quality (total score of 5 or 6) or low quality (total score > 7). Only high and medium
quality data were included in final national BAFs calculations.
For the detailed derivation of PFOA and PFOS national BAFs, see US EPA 2024o and US EPA 2024p. Table
16 summarizes the draft national BAFs for PFOA and PFOS for trophic levels 3 and 4. Trophic level 2 BAFs
are not relevant to the fish consumption scenarios assessed in this document (see Section 2.9.3.8, Fish
Consumption Rate).
Table 16. Fish BAFs by Trophic Level
Trophic
PFOA
PFOS
Level
(L/kg)
(L/kg)
TL3
49
1,700
TL4
31
860
2.9.3.8 Consumption Rates for Food and Water
The exposure factors used to parameterize the central tendency approach are selected to represent
median values for the distribution of people represented by the various receptors captured in the
conceptual models. If median values are not available, a mean value is used instead. Most of the
exposure parameters are selected from tables presented in the most recent version of EPA's EFH; unless
otherwise noted, that is US EPA (2011). Note that the exposure factors used for the central tendency
modeling run are not those that would be used to calculate a risk-based regulatory threshold. A
summary of human exposure factors can be found in Appendix B, Table B.12.
Fish Consumption Rate
In this assessment, the EPA selected a fish consumption rate of 0.47 g/kg-day for adults (~1.3 ounces per
day), 0.31 g/kg-day for children 12-19 (~0.6 ounces per day), and 0.55 g/kg-day for children 6-11 (~0.6
ounces per day). These values represent the 50th percentile of Consumer-Only Intake of Home-Caught
Fish (EFH Chapter 13, table 13-20). This survey did not have sufficient sample size to calculate fish intake
rates for children aged 1-5, so the intake rate for children aged 6-11 was used for this group. A typical
DRAFT
70
-------
fish meal for adults is a 4 ounce to Vz pound serving of raw fish, which is 113-227 g; the adult fish
consumption rate used in this assessment (assuming an 80 kg adult bodyweight) amounts to consuming
an average of one to two fish servings per week. Serving sizes for children increase from 1 to 4 oz as
they age from 1 to 11 years old. These exposure factors also equate to about 1-2 meals per week for
children aged 1-5 and 6-11. Assuming a serving size of 4 oz per fish meal, the intake rate for children
aged 12-19 equates to about one meal per week.
Bioaccumulation rates for PFOA and PFOS differ by trophic level (see Section 2.9.3.7). In this assessment,
fish consumption is apportioned between trophic level 3 and trophic level 4 using data presented in EFH
Chapter 10, Table 10-74, Total Consumption of Freshwater Fish Caught by All Survey Respondents
During the 1990 Season. The species presented in this table were assigned trophic levels from the three
following sources, in order of preference: 1) Estimated Fish Consumption Rates for the US Population
and Selected Subpopulations (NHANES 2003-2010), Table 3 (US EPA, 2014); 2) The journal publication
"Comparing trophic position of freshwater fish calculated using stable nitrogen isotope ratios (6N15)
and literature dietary data," (Zanden et al., 1997); and 3) A publicly available database that catalogues
information on various fish species published in the Journal of Fish Biology, Journal of Applied
Ichthyology, and Acta Ichthyologica et Piscatoria (FishBase, 2024). The survey data presented in EFH
Table 10-74 indicate that 14% of freshwater fish consumption is of fish in trophic level 3 (for example,
lake whitefish, chub), while 86% of fish consumption is of fish in trophic level 4 (for example, brown
trout, yellow perch, smallmouth bass).
Drinking Water intake Rate
The drinking water intake rates for the central tendency modeling effort were selected from the latest
edition of EPA's EFH, chapter 3 (ingestion of water and other select liquids; US EPA, 2019c). The values
selected represent the 50th percentile of reported direct and indirect consumption of community water,
in milliliters per bodyweight per day from the NHANES 2005-2010 survey (US EPA, 2019c, Table 3-21).
The median drinking water intake is 13.4 ml/kg-day for adults, 6.5 ml/kg-day for children 12-19, 11.5
ml/kg-day for children 6-11 and 16.2 ml/kg-day for children 1-5. Assuming a bodyweight of 80 kg, this
amounts to an adult drinking water intake rate of approximately 1 L/day.
Note that this drinking water intake rate used in this central-tendency modeling run is significantly lower
than the drinking water intake rate used for other CWA purposes, such as development of national
recommended human health criteria, and for Safe Drinking Water Act purposes, such as developing
regulatory standards or setting non-regulatory health advisories.
Protected Fruits and Vegetables intake Rates
"Protected produce" is a fruit or vegetable that has an outer protective coating that is typically removed
before consumption. Examples of protected vegetables included pumpkin, corn, peas, and beans.
Examples of protected fruits include melons like watermelon and cantaloupe, citrus fruits like oranges
and grapefruit, and bananas.
The intake rates for protected fruits are the 50th percentile values in EFH chapter 13, table 13-59 and are
presented in grams wet-weight fruit per kilogram of bodyweight per day. The median consumption rate
of protected fruit is 2.1 g/kg-day for adults, 1.2 g/kg-day for children 12-19, 2.3 g/kg-day for children 6-
11 and 2.3 g/kg-day for children 1-5. Given that a typical serving of fruit is 100-200 grams, the adult
protected fruit intake equates to about one serving of protected fruit a day.
The intake rates for protected vegetables are the 50th percentile values in EFH chapter 13, table 13-61
and are presented in grams wet-weight vegetable per kilogram of bodyweight per day. The median
consumption rate of protected vegetables is 0.6 g/kg-day for adults, 0.58 g/kg-day for children 12-19,
0.79 g/kg-day for children 6-11 and 1.4 g/kg-day for children 1-5. Given that a typical serving of
DRAFT
71
-------
vegetables is about 100 grams, the adult protected vegetable intake equates to about one serving of
protected vegetables every other day.
Unprotected Fruits and Vegetables Intake Rates
"Unprotected" or "exposed" foods are those that are grown above ground and may be contaminated by
pollutants deposited on surfaces of the foods that are eaten. Examples of unprotected vegetables
include cauliflower, tomatoes, eggplant, cucumber, snap peas, herbs, and mushrooms. Examples of
unprotected fruits include fresh or dried apples, pears, peaches, grapes, and berries.
The intake rates for unprotected fruits are the 50th percentile values in EFH chapter 13, table 13-58 and
are presented in grams wet-weight fruit per kilogram of bodyweight per day. The median consumption
rate of unprotected fruit is 1.3 g/kg-day for adults, 0.61 g/kg-day for children 12-19, 1.11 g/kg-day for
children 6-11 and 1.82 g/kg-day for children 1-5. Given that a typical apple is about 240 grams, the adult
unprotected fruit intake equates to about one apple every other day.
The intake rates for unprotected vegetables are the 50th percentile values in EFH chapter 13, table 13-60
and are presented in grams wet-weight vegetable per kilogram of bodyweight per day. The median
consumption rate of unprotected vegetables is 1.4 g/kg-day for adults, 0.66 g/kg-day for children 12-19,
0.64 g/kg-day for children 6-11 and 1.5 g/kg-day for children 1-5. Given that a typical serving of
vegetables is about 100 grams, the adult unprotected vegetable intake equates to about one serving of
unprotected vegetables every day.
Root Vegetables Intake Rates
Root vegetables are vegetables where the consumed portion of the plant is the root. Root vegetables
often have different uptake rates of environmental contaminants than vegetables where other portions
(stems, leaves) of the plant are consumed. Examples of root vegetables include onions, carrots, beets,
turnips, and potatoes. The intake rates for root vegetables are the 50th percentile values in EFH chapter
13, table 13-62 and are presented in grams wet-weight fruit per kilogram of bodyweight per day. The
median consumption rate of root vegetables is 0.88 g/kg-day for adults, 0.57 g/kg-day for children 12-
19, 0.52 g/kg-day for children 6-11 and 0.69 g/kg-day for children 1-5. Given that a typical serving of
vegetables is about 100 grams (~ Vz an average-sized russet potato), the adult root vegetable intake
equates to about five servings of root vegetables a week.
Milk and Dairy Intake Rates
The milk consumption rates for the central tendency scenario models were selected from the most
recent edition of the EFH chapter 11, Meats, Dairy Products, and Fat (US EPA, 2018b) and chapter 13,
Home Produced Foods (US EPA, 2011). Although chapter 13 (Intake of Home-Produced Foods) included
some national data on intake of milk and other dairy products, there was only one age category in the
available surveys with sufficient sample size to calculate descriptive statistics (ages 20-39). The
respondents were additionally divided between families that answer yes to the question "Did anyone in
the household produce any animal products such as milk, eggs, meat, or poultry for home use in your
household?" (described as "households who farm") versus families that answer yes to the question "Did
anyone in the household operate a farm or ranch?" (described as "households who raise animals").
Because the description of "households who farm" was best aligned with the conceptual model for the
pasture farm, the 50th percentile dairy intake from this survey was used for adults (12.1 g/kg-day). For
an 80 kg adult, this amounts to approximately 34 fluid ounces of milk consumed per day, which is four,
8oz glasses.
Because there were no data available for milk consumption in children specifically from families that
produce milk at home, national milk consumption data was used for these age categories. Note that this
national data likely underestimates the amount of milk consumed by children who grow up on dairy
DRAFT
72
-------
farms. For example, the available data for milk intake in adults finds that adults who live on farms
consume about six times more dairy than in adults in national surveys. The values selected for the child
age categories represent the 50th percentile of reported dairy consumption rate, in grams wet weight
per kilogram bodyweight per day from the NHANES 2005-2010 survey (US EPA 2018b, Table 11-4). The
median milk intake is 4.3 g/kg-day for children 12-19 (amounts to ~1, 8 oz glass per day), 12 g/kg-day for
children 6-11 (amounts to 1.5, 8 oz glasses a day), and 30 g/kg-day for children 1-5 (amounts to ~2, 8 oz
glasses per day).
Beef Intake Rates
Beef consumption rates were selected from the EFH chapter 13, Home produced foods. The values
selected represent the 50th percentile of reported beef consumption rate for consumers-only, in grams
per kilogram bodyweight per day from the Nationwide Food Consumption Survey (NFCS), 1987-1988
(Table 13-33). The median beef intake is 1.6 g/kg-day for adults, which represents the median intake of
respondents in "households who farm." The beef intake rate is 1.5 g/kg-day for children 12-19 (~3
ounces per day) and 2.1 g/kg-day for children 6-11 (~2 ounces per day). This survey did not have data
available for beef intake for children 1-5. The models assume that the intake rate for this group is the
same as the intake rate for the slightly older children of 2.1 g/kg-day. This assumption is supported by
information provided in Chapter 11 of the EFH (meat, dairy, and fats; US EPA, 2018b), which reports in
Table 11-6 that the mean beef intake rate for age 2-6 ranges from 1.6 to 1.7 g/kg-day for the general
public (NHANES 2005-2010). Assuming a bodyweight of 80 kg, the adult consumption rate amounts to
an adult beef intake rate of slightly over one, three ounce serving of beef every day.
Egg Intake Rates
Egg consumption rates were selected from the EFH chapter 13, Home Produced Foods. The value
selected represents the 50th percentile of reported egg consumption rate for consumers-only in
"households who farm," reported in grams per kilogram bodyweight per day from the NFCS, 1987-1988
(Table 13-40). The median egg intake is 0.7 g/kg-day for all ages. This survey does not include age
breakdowns for children and adults. EFH chapter 11, Meat and Dairy (US EPA, 2018b) does not include a
survey specific to egg consumption. Because of this lack of age-specific intake rates, the "all ages" value
from Table 13-40 will be used to represent egg intake rates for all age groups. Given an adult
bodyweight of 80kg and a 50g average egg mass, this amounts to an intake rate of about 1 egg per day.
Chicken Intake Rates
The chicken consumption rates for the central tendency scenario models were selected from EFH
chapter 11, Meats, Dairy Products, and Fat (US EPA, 2018b) and chapter 13, Home Produced Foods (US
EPA, 2011). Although chapter 13, Intake of Home-Produced Foods, included some national data on
intake of poultry, there were limited age categories with sufficient sample size to calculate descriptive
statistics (see Table 13-52, Consumer-Only Intake of Home-Produced Poultry). As described previously,
the respondents categorized as from "households who farm" was best aligned with the conceptual
model for the pasture farm, the 50th percentile poultry intake from this survey was used for adults (1.1
g/kg-day). This survey does not include chicken-specific consumption rates, but rather consumption
rates for "poultry," which includes chicken, turkey, and other poultry. The EPA finds that this represents
the best available data for parameterizing intake rates of home-produced chickens. For an 80 kg adult,
this intake rate amounts to about one three-ounce serving of chicken every day.
Because there was no data available for chicken consumption in children specifically from families that
produce their own food, national chicken consumption data was used for these age categories (EFH,
chapter 11, Table 11-6). The survey available for this consumption category reports mean intake values
rather than median intake values. Mean intake values are likely slightly higher than median intake values
but are still appropriate for this central tendency modeling exercise. The survey also only reported
DRAFT
73
-------
intake rates for poultry, rather than chicken only. A separate survey (represented in chapter 11, Table
11-7; US EPA, 2018b) indicates that for most Americans, the majority of poultry intake is chicken. The
values selected for the child age categories represent the mean reported poultry consumption rate, in
grams wet weight per kilogram bodyweight per day from the NHANES 2005-2010 survey (US EPA 2018b,
Table 11-6). The mean intake is 1.1 g/kg-day for children 12-19 (~2 oz per day), 1.6 g/kg-day for children
6-11 (~1.6 oz per day) and 2.4 g/kg-day for children 1-5 (~1.3 oz/day).
Overview of Consumption Rates
Table 17. Overview of Selected Human Consumption Rates
Category
Adult (g/kg-day for
all except drinking
water)
Child 1-5 (g/kg-day
for all except
drinking water)
Child 6-11 (g/kg-
day for all except
drinking water)
Child 12-19 (g/kg-
day for all except
drinking water)
Fish
0.47 (1.3 oz per
day; ~1-2 servings a
week)
0.55 (0.3 oz per
day; ~1-2 servings a
week)
0.55 (0.6 oz per
day; ~1-2 servings a
week)
0.31 g (0.6 oz per
day; ~1 serving a
week)
Drinking water
13.4 ml/kg-day (1 L
per day)
16.2 ml/kg-day (240
ml per day)
11.5 ml/kg-day (330
ml per day)
6.5 ml/kg-day (300
ml/day)
Protected fruits
2.1 (6 oz per day;
~1 serving per day)
2.3 (1 oz per day)
2.3 (2.4 oz per day)
1.2 (2.6 oz per day)
Protected
vegetables
0.6 (1.7 oz per day;
~1/4 serving a day)
1.4 (0.8 oz per day)
0.79 (0.8 oz per
day)
0.58 (1.2 oz per
day)
Unprotected fruits
1.3 (3.6 oz per day;
-1/2 an apple a day)
1.82 (1 oz per day)
1.11 (1.1 oz per
day)
0.61 (1.3 oz per
day)
Unprotected
vegetables
1.4 (4 oz per day; ~
1 serving per day)
0.64 (0.3 oz per
day)
0.64 (0.65 oz per
day)
0.66 (1.4 oz per
day)
Root vegetables
0.88 (2.5 oz per
day; ~ 1/4 a small
potato a day)
0.69 (0.4 oz per
day)
0.52 (0.5 oz per
day)
0.57 (1.2 oz per
day)
Milk and dairy
12.1 (34 oz a day;
~4, 8 oz glasses)
30 (15 oz per day;
~2, 8 oz glasses per
day)
12 (12 oz per day;
-1.5, 8 oz glasses a
day)
4.3 (9 oz per day;
~1, 8 oz glass per
day)
Beef
1.6 (4.5 oz per day)
2.1 (1 oz per day)
2.1 (2.1 ounces per
day)
1.5 (3.2 ounces per
day)
Egg
0.7 (~1 egg per day)
0.7 (~1 egg every 5
days)
0.7 (~1 egg every
other day)
0.7 (~1 egg per day)
Chicken
1.1 (3.1 oz per day,
~1 serving per day)
2.4 (1.3 oz/day)
1.6 (1.6 oz per day)
1.1 (2 oz per day)
2.9.3.9 Cooking and Food Preparation Loss Assumptions
Risk assessments that include food consumption pathways often consider if a portion of the
contaminant is lost during the food prep or cooking process. EFSA conducted an assessment of ingestion
risks for PFOA and PFOS through food exposures in 2018 (EFSA CONTAM Panel, 2018). In the
assessment, the authors summarized the available literature on food loss in preparing or cooking various
types of food containing PFOA and PFOS. They find that some studies report loss of PFOA and PFOS
while other studies find PFOA and PFOS concentrations increase, perhaps due to loss of water during the
cooking process, which increases the concentration of remaining contaminant. Overall, ESFA concludes
that the limited number of studies gives an inconsistent view about whether losses or increases occur
for PFOA and PFOS across different food types and cooking strategies. The biosolids draft risk
assessment will thus assume 0% loss in fruits, vegetables, meats, eggs, and milk.
2.9.3.10 Soil Ingestion Rates
The soil ingestion rates for the central tendency modeling effort were selected from the EFH, chapter 5
(soil and dust ingestion). The values selected represent the central tendency of soil ingestion (which
DRAFT
74
-------
includes soil and outdoor dust), in mg per day (US EPA 2017, Table 5-1). The central tendency soil
ingestion rate is 10 mg/day for adults, 10 mg/day for children 12-19, 30 mg/day for children 6-11 and 40
mg/day for children 1-5. The EFH notes that soil and dust ingestion is likely higher in adults following a
"traditional rural or wilderness lifestyle." It is likely that this central tendency estimate would
underestimate soil ingestion for a farmer who frequently works weeding, harvesting, or otherwise
disturbing soils on a farm. However, the EFH does not include a dust ingestion rate specific for adults
who work on farms.
2.9.3.11 Body Weight
In this draft risk assessment, the EPA selected a bodyweight of 80 kg for adults, 61 kg for children 12-19,
29 kg for children 6-11 and 15 kg for children 1-5. These rates are based on 50th percentile American
bodyweight, Table 8-3 of US EPA (2011), NHANES 1990-2006. Note that bodyweight assumptions are
only required when bodyweight-normalized intake rates are not available.
2.9.3.12 Duration of Exposure Modeling
The exposure model does not assume that the residents spend their entire life at the relevant site;
rather, it is assumed that the residents have moved over the course of their life. For this draft risk
assessment, the EPA selected an exposure duration of 10 years for adults, corresponding to the 50th
percentile of total residence time for farms from Table 16-113 of US EPA (2011). The 50th percentile of
residential occupancy from the EFH, Table 16-109 is 9 years. Thus, 10 years is a reasonable value for
nearby residents who are not farmers as well. This residency assumption applies to the entire family,
including children. The exposure period for cancer risk and non-cancer is assumed to occur around the
time of maximum media concentrations within the modeling period (so, if the peak media concentration
occurs in model year 40, the 10-year exposure duration would run from model years 35 to 44 and the 1-
year exposure duration would be for model year 40). The cancer exposure model assumes that the
receptors are at the relevant site for 350 days per year (either their non-farm home or farm home,
depending on the conceptual model).
Because an exposure duration of 10 years is used for the entire farm family, the exposure factors for
children aged 1-5 and 6-11 were combined (using a weighted average based on sample size in each age
bin reported) into values appropriate for ages 1-11. For soil consumption, which is not based on a single
study, this assessment used the slightly higher value of 40 g/day (for children 1-5, vs 30 g/day for
children 6-11) for children 1-11. These average intake values are provided in Appendix B.
2.9.3.13 Location-specific Parameters
Models were parametrized to represent a range of climatological conditions (dry, moderate, and wet)
using datasets from three regions located near Boulder, CO; Chicago, IL; and Charleston, SC. These
locations were used as a basis for selecting, in order of preference, representative local (e.g.,
meteorological parameters), regional (e.g., soil and hydrologic parameters), and national data (e.g.,
application characteristics). Where distributions of parameter values are available at the regional or
national level, median values were selected.
Meteorological data. Daily meteorological data (precipitation, temperature) from a five-mile radius
surrounding the three locations were represented by the nearest gridded dataset developed by the EPA
primarily for pesticides modeling (Fry et al., 2016). The mean annual windspeed for each region was also
identified. Parameters describing general soil properties in the field and surrounding watershed for
overland flow and transport calculations are represented by median values selected from national
distributions developed in support of other pollutant evaluations for the EPA (see Table B-6). By
selecting the weather and soil data from the same geographic location, the models are pairing climate
and soil conditions that naturally co-occur.
DRAFT
75
-------
Application location size. Parameters describing general site characteristics applicable to crop, pasture,
and reclamation land application scenarios are also based on median national values developed as part
of various Federal agency missions (e.g., USDA national farm field sizes) or in support of other pollutant
evaluations for EPA. The 80-acre field14 where biosolids are applied, and which is used to grow crops
(crop scenario) or to pasture cows (pasture and reclamation scenarios), is assumed to be square. Though
the model allows for the site to have vehicles and corresponding particulate spread through dust, this
assessment assumes no vehicles regularly drive over the site.
Surface water size, location. A 13-acre index reservoir15 that drains the adjacent local watershed serves
as an alternative source of drinking water for the farm family (their primary drinking water source is
assumed to be groundwater). The index reservoir is based on the standard waterbody parameters for
VVWM, the waterbody model used to estimate concentrations in surface water (US EPA, 2019b; 2020).
A 10-meter wide, rectangular buffer exists between the field and the index reservoir;16 the LAU source
model estimates runoff and erosion from the field to the buffer and then from the buffer to the
reservoir. The farm family is assumed to live in the buffer.
Soil characteristics. Soil characteristics for determining regional recharge rates to groundwater and to
parameterize the unsaturated portion of the groundwater model are based on the predominant soil
mega-texture within a 5-mile radius of the field location from the same national data source as the
watershed characteristics. The EPA HELP model (Schroeder et al., 1994) was used to calculate regional
recharge rates using meteorological data assigned to each location, and HELP default values for the
following parameters corresponding to predominate soil mega-texture at each location:
• Soil Porosity: ratio of the volume of void spaces in a volume of soil.
• Field Capacity: The volume of water remaining in void spaces in a volume of soil after freely
draining from a saturated state, expressed as a percentage.
• Wilting Point: volume of water remaining in void spaces in a volume of soil at which plants wilt
and fail to recover, expressed as a percentage.
• Soil Hydraulic Conductivity: the amount of water moving vertically through a unit area of
saturated soil in unit time under unit hydraulic gradient.
The following parameters are used in EPACMTP to describe flow in the unsaturated zone in addition to
the soil hydraulic conductivity:
• Alpha and Beta: soil-specific shape parameters used in the van Genuchten (1980) model for
modeling soil-water content as a function of pressure head.
• Residual water content: the irreducible water content obtained after lowering the pressure head
in the soil.
• Saturated water content: maximum fraction of total volume of soil occupied by water in the soil
(equivalent to soil porosity).
• Percent Organic Matter: measure of amount of organic material present within the soil of the
unsaturated zone, as a weight percent.
14 The field size is based on the 50th percentile from the 2012 Census of Agriculture (USDA, 2014).
15 The index reservoir is based on the standard waterbody parameters for Variable Volume Water Model (VVWM), the
waterbody model used to estimate concentrations in surface water (US EPA, 2019; 2020); see Section A.2.3.2.
16 The Part 503 regulations state that "bulk sewage sludge shall not be applied to agricultural land, forest, or a reclamation site
that is 10 meters or less from waters of the United States." The buffer for the index reservoir has been set to 10 m in
accordance with this standard.
DRAFT
76
-------
The values for each of these parameters are based on median values specific to each mega-texture
associated with each location selected from national distributions developed in support of other
pollutant evaluations for EPA.
Aquifer characteristics. Aquifer characteristics (depth to water table, aquifer thickness, regional
hydraulic gradient, and aquifer hydraulic conductivity) were based on median values from the EPA's
Hydrogeologic Database (HGDB). The HGDB was developed by the American Petroleum Institute (Newell
et al., 1989; 1990) to specify correlated data sets of these four parameters for the 12 distinct
hydrogeologic environments described in Newell et al. (1990). The EPA first developed a national
geographic coverage of the 12 hydrogeologic environments, and then used GIS to overlay the three
simulated locations and assign each location a hydrogeologic environment. Median values were selected
for each of the four parameters from the assigned environments. One exception was at the "wet"
region, where a mean value for the hydraulic conductivity of the saturated zone was used instead of the
median. The use of the median value (315 m/yr), in conjunction with other inputs, resulted in a
mounded water table that exceeded the elevation of the ground surface, violating an underlying
assumption of EPACMTP model (Section 4.3.6 of EPA 2003e). Adjusted values for this parameter input
are also noted in Appendix B. Other flow and transport-related parameters not associated with chemical
properties are selected from national distributions developed in support of other pollutant evaluations
for the EPA (US EPA, 2003a) or where specifically noted in Appendix B. These parameters include aquifer
porosity, bulk density, dispersivity, aquifer fraction of organic content, and temperature.
2.9.3.14 Biosolids Application Assumptions
Biosolids applications of 10 MT dry weight per hectare of field area were modeled to occur once per
year on April 1 for 40 years for the crop and pasture scenarios. The EPA's prior risk assessment of dioxins
and PCBs also used a 40 year timeframe for application of sewage sludge to a field (US EPA, 2003a). The
existing sewage sludge regulations in 40 CFR part 503 assume 100 consecutive years of sewage sludge
land application when calculating cumulative and annual loading rates for metals. As there are not data
available on the longevity of sewage sludge application to a given field or location, the EPA is continuing
to model risks for scenarios with 40 years of application, in line with the prior risk assessment. To
estimate a reasonable median agronomic application rate, probabilistic plant available nitrogen (PAN)
calculations were conducted using the PAN and agronomic spreadsheet calculation tool available from
the Colorado Department of Public Health & Environment (CDPHE, 2018) and @Risk (Palisade
Corporation), a Microsoft Excel plug-in. The basic annual rate calculation is based on PAN per metric ton
of biosolids on a per hectare basis and the crop nitrogen requirement. Probabilistic simulations were
conducted assuming an absence of residual nitrogen from any sources (background or previous biosolids
or fertilizer application) and varying several parameters such as crop yield and days to incorporation.
The analysis is described in more detail in Appendix E of US EPA (2023c). This produced a range for dry
weight agronomic application rate of approximately 0.5 to 30 dry MT/ha and an overall median value of
7.6 dry MT/ha. This range is consistent with recommended ranges found elsewhere in the literature for
crop applications (US EPA, 2000b), which range from around 2 to 20 dry MT/ha. The application rate
value of 10 dry MT/ha used in this assessment is based on rounding the analysis median value to the
nearest order of magnitude to account for variability. Biosolids are assumed to be tilled (i.e., fully mixed)
into the top 20 cm of the field for the crop scenario whereas in pasture and reclamation scenarios,
biosolids are assumed to be unincorporated with field soils after application.
For the reclamation scenario, a single application of biosolids at a rate of 50 MT dry weight per hectare
of field area is modeled to occur on the April 1 of the first year of the simulation.
DRAFT
77
-------
2.9.3.15 Surface Disposal Assumptions
The surface disposal unit (SDU) is modeled as having a square footprint with an area of 3,400 m2; this
value is calculated from the median values of depth and flow from the Industrial D Screening Survey
data presented in the 3MRA modeling documentation (US EPA, 2003g) and the operating life described
below. The SDU is assumed to operate for 50 years, consistent with the 2003 sewage sludge screening
assessment (US EPA, 2003a; appendix G), during which time, liquids and dissolved chemical mass in the
liquids can pass through the bottom of the unit. Liquids in the unit are assumed to maintain a near
constant volume and are not aerated. Darcy's law is used to calculate the rate of leakage through the
base of the unit into the unsaturated zone, and the base of the unit may be unlined, clay lined,
composite lined. After 50 years, the SDU is assumed closed and no additional chemical mass is released
to the environment; however, the groundwater model assumes the long-term average volumetric rate
of liquids leaving the unit are assumed to persist to beyond 50 years. The source of groundwater for
drinking is assumed to be 5 meters down gradient of the SDU, in the middle of a 10-meter buffer area
(the same as the land application scenarios). The surface disposal unit is assumed to be "clean closed" at
the end of its 50-year economic life such that no residual PFOA or PFOS remains.
The key parameters governing the rate of leakage through the bottom of the SDU are, as organized by
liner scenario:
• Unlined and Clay Lined SDUs:
- The maximum height of liquids above the bottom of the SDU (2 m)
- Flow rate into the SDU: (4 x 10 s m3/s)
- Precipitation rate that are specific to each of the three locations representing dry, average,
and wet climates
- Material properties of the settled sediment in the SDU, including saturated hydraulic
conductivity (5 x 10"7 m/s) and soil-water retention parameters (Alpha 0.016 1/cm; Beta
1.37)
• Clay Lined SDUs
- Material properties and dimensions of the clay liner, including saturated hydraulic
conductivity (1 x 10"9 m/s), soil-water retention parameters (Alpha 0.008 1/cm; Beta 1.09),
and liner thickness (0.9144 m)
• Composite Lined SDUs:
- Specified infiltration rate through a composite liner (1.4 x 10 s m/d)17
The key processes and non-chemical specific parameters governing the concentration of chemical mass
of PFOA and PFOS in the liquids passing through the bottom of the SDU are limited to sorption and
solids generation and removal:
• Influent total suspended solids concentration (0.1 g/cm3)
• Fraction organic carbon in suspended solids (0.4 g/g)
• Solids removal rate (calculated based on flow rate, SDU area, and suspended particle sizes).
17 The approximate 90th percentile infiltration rate from Table 4.6 of US EPA, 2003d.
DRAFT
78
-------
3 ANALYSIS
3.1 Exposure Characterization, Central Tendency Models
The following sections present and discuss the modeled concentration and exposure results for
individual exposure pathways in each of the biosolids use or disposal scenarios outlined in Section 2.8.
The modeled media concentration results are presented in units of ng PFOA or PFOS per mg wet weight
of media (e.g., milk, soil, water, beef). All modeling runs assume that the starting concentrations of
PFOA and PFOS in sewage sludge are 1 ppb (1 ng/kg). This concentration is near available detection
thresholds for PFOA and PFOS in sewage sludge (US EPA, 2024d) and below levels commonly detected in
U.S. sewage sludge (see Section 2.4 and Appendix A). The models and calculations used in this
assessment have a linear relationship between the starting concentration of PFOA and PFOS in sewage
sludge and the modeled concentration in each environmental media. This means that if the starting
concentration of PFOA or PFOS in sewage sludge were to increase from 1 ppb to 10 ppb, the modeled
media concentration would increase by a factor of 10.
As described in Section 5.3, the concentration results from fate and transport modeling are highly
sensitive to the parameters associated with the climate setting and Koc. For this reason, modeled
exposures for a given pathway will be presented for each climate (dry, moderate, and wet) and for a low
Koc (10th percentile) and high Koc (90th percentile).
3.1.1 Crop Farm
The crop farm scenario models the fate and transport of PFOA and PFOS as they move from biosolids
through soil, surface water, and groundwater. The models then estimate the direct exposure to adults
and children to those media, and the uptake and exposure from those media to fruits, vegetables, and
fish. In this central tendency modeling exercise, the concentrations of PFOA and PFOS in the modeled
biosolids are low (1 ppb) for each chemical. The following tables show the modeled concentrations of
PFOA and PFOS in each media type during either a ten-year averaging time or a one-year averaging
time. These averaging windows include the maximum concentration year for each media type. The
tables include three climate scenarios: dry, moderate, and wet. These climate scenarios also represent
varied soil types, depths to groundwater, hydrological connectivity and other related hydrogeological
conditions that would be expected in these climate settings.
Table 18. PFOA Media Concentrations for Crop Farm (ppt): Maximum 10- and 1-year Averages
Pathway
Low Koc
High Koc
10-yr | 1-yr
10-yr | 1-yr
Dry Climate
Exposed Fruit
0.86
0.89
2.0
2.1
Exposed Vegetables
6.6
6.9
15
16
Fish3
260
270
48
49
Groundwater
0.40
0.45
2.1E-9
2.1E-9
Protected Fruit
0.74
0.77
1.7
1.8
Protected Vegetables
13
13
29
30
Root vegetable
6.1
6.4
14
15
Soil
34
43
92
100
Surface water
7.8
8.0
1.4
1.4
Moderate climate
Exposed Fruit
0.050
0.076
0.81
0.86
Exposed Vegetables
0.39
0.58
6.2
6.6
Fish3
14
15
48
51
Groundwater
4.5
5.5
0.12
0.12
Protected Fruit
0.044
0.066
0.70
0.75
Protected Vegetables
0.74
1.1
12
13
DRAFT
79
-------
Pathway
Low Koc
High Koc
10-yr
1-yr
10-yr
1-yr
Root vegetable
0.36
0.54
5.8
6.1
Soil
0.56
1.8
29
39
Surface water
0.42
0.46
1.4
1.5
Wet Climate
Exposed Fruit
0.046
0.088
0.62
0.64
Exposed Vegetables
0.35
0.67
4.8
4.9
Fish3
9.5
12
34
36
Groundwater
4.5
4.5
0.48
0.48
Protected Fruit
0.04
0.076
0.54
0.56
Protected Vegetables
0.67
1.3
9.0
9.4
Root vegetable
0.33
0.62
4.4
4.6
Soil
0.52
1.6
21
27
Surface water
0.28
0.36
1.0
1.1
a These values represent the weighted average fish tissue concentration by the percent consumption of trophic levels 3 and 4.
Table 19. PFOS Media Concentrations for Crop Farm (ppt): Maximum 10- and 1-year Averages
Pathway
Low Koc
High Koc
10-yr | 1-yr
10-yr | 1-yr
Dry Climate
Exposed Fruit
0.33
0.34
0.57
0.58
Exposed Vegetables
0.81
0.83
1.4
1.4
Fish3
4500
4500
40
41
Groundwater
0.054
0.059
4E-31
4E-31
Protected Fruit
0.29
0.29
0.49
0.5
Protected Vegetables
1.5
1.6
2.6
2.7
Root vegetable
11
11
19
20
Soil
60
70
120
130
Surface water
4.6
4.6
0.039
0.039
Moderate climate
Exposed Fruit
0.076
0.083
0.50
0.53
Exposed Vegetables
0.19
0.2
1.2
1.3
Fish3
1700
1800
49
52
Groundwater
0.98
0.98
1.8E-05
1.8E-05
Protected Fruit
0.066
0.072
0.44
0.46
Protected Vegetables
0.35
0.38
2.3
2.5
Root vegetable
2.6
2.8
17
18
Soil
8.4
14
83
110
Surface water
1.7
1.8
0.048
0.051
Wet Climate
Exposed Fruit
0.055
0.062
0.46
0.47
Exposed Vegetables
0.13
0.15
1.1
1.2
Fish3
1100
1100
54
57
Groundwater
2.7
2.7
0.01
0.015
Protected Fruit
0.048
0.054
0.4
0.41
Protected Vegetables
0.25
0.29
2.1
2.2
Root vegetable
1.9
2.1
15
16
Soil
4.9
8.9
84
97
Surface water
1.1
1.1
0.052
0.055
a These values represent the weighted average fish tissue concentration by the percent consumption of trophic levels 3 and 4.
The crop farm scenario outputs concentrations over time for two categories of fruits (exposed and
protected), three categories of vegetables (exposed, protected, and root), fish, surface water, soil, and
groundwater. Groundwater concentrations range from effectively 0 ng/L to 5.5 ng/L for PFOA and
effectively 0 to 2.7 ng/L for PFOS. Surface water concentrations range from 0.028 to 8.0 ng/L for PFOA
and from 0.039 to 4.6 ng/L for PFOS. Soil concentrations range from 0.52 to 100 ng/kg for PFOA and 4.9
DRAFT
80
-------
to 130 ng/kg for PFOS. Fish tissue concentrations range from 9.5 to 270 ng/kg for PFOA and 40 to 4,500
ng/kg for PFOS. Finally, fruit and vegetable concentrations range from 0.040 to 30 ng/kg for PFOA and
0.048 to 20 ng/kg for PFOS. Root, protected, and exposed vegetables have higher PFOA and PFOS
concentrations than the other produce categories.
Overall, the one-year and ten-year average concentrations for each media are similar. Ten-year average
concentrations are often the same or only slightly lower than the one-year averages. This trend reflects
the fact that for many media types, yearly average concentrations remain elevated for years at a time.
See Section 3.2 for more discussion on temporal trends in modeled concentrations.
Potential groundwater contamination associated with PFOA and PFOS leaching from biosolids-amended
soils is of high concern, in part because biosolids are often land-applied in areas where nearby residents
rely on groundwater as a source of drinking water. This modeling exercise allows us to explore the
potential impacts to groundwater at biosolids concentrations that are commonly exceeded around the
U.S. (concentrations of 1 ppb for PFOA and PFOS) when they are annually applied to land used to grow
fruits and/or vegetables. The draft modeling results show that when biosolids are applied with these low
PFOA and PFOS concentrations, groundwater concentrations of PFOA and PFOS vary depending on the
Koc and climate setting in each modeled scenario. For PFOA, groundwater concentrations range from
effectively zero in the high Koc dry climate to 4.5-5.5 ng/L in the low Koc moderate and wet climates. For
PFOS, groundwater concentrations range from effectively zero in the high Koc moderate and dry climate
to 2.7 ng/L in the low Koc wet climate setting. Overall, these groundwater results are similar to the
results seen for the pasture farm scenario (see Section 3.1.2), which also models a farm setting, but the
pasture farm scenario assumes no tilling of soil whereas the crop farm assumes annual tilling of the
field.
These groundwater outcomes can be partially explained by the sorption behavior of PFOA and PFOS in
soils. The Kd is calculated by measuring the concentration of PFOA or PFOS in soil and dividing it by the
equilibrium concentration of PFOA or PFOS in the soil pore water. This metric indicates the relative
amount of PFOA or PFOS that sorbs to soil in comparison to the amount dissolved in the surrounding
water. In EPA's models, Kd is calculated by multiplying Koc by the foc in the biosolids-amended soils for
each climate setting (see Section 2.9.3.3). This allows the models to adjust Kd based on the amount of
organic matter in the underlying soils for each climate and geological setting. However, measurements
of Kd are more common than measurements of Koc and more directly capture soil leaching potential in
field conditions. PFOS generally has higher measured Kd than PFOA in biosolids-amended soils. Though
observed Kd for both compounds in biosolids-amended soils can vary more than two orders of
magnitude across locations, within a single study site, Kd values for PFOS are higher than those for PFOA.
For example, in a recent study of PFOA and PFOS in biosolids-amended soils in New Hampshire, the
average log(Kd) for PFOS was generally between 2 and 2.5 L/kg while the log(Kd) for PFOA was between 1
and 2 L/kg (Tokranov et al., 2023). Correspondingly, the model results show that a higher proportion of
PFOS is retained in soils and a higher portion of PFOA is mobilized through the soil column to
groundwater. These trends are reflected in both the soil and groundwater concentrations generated by
modeling runs, in that when PFOA and PFOS are at the same concentration in biosolids (1 ppb), soil
concentrations are higher for PFOS than PFOA while groundwater concentrations are higher for PFOA
than PFOS. Note that when modeling the fate and transport of PFOA and PFOS from biosolids
contaminated with concentrations of 1 ppb for each compound, the resulting groundwater
concentrations are often, but not always, below the minimum reporting level (MRL) of 4 ng/L for each
compound using EPA's groundwater method EPA 533.
Another media of high concern is fish tissue, especially for PFOS, which is known to be highly
bioaccumulative in the commonly consumed portions of fish like filets. In this modeling scenario, PFOA
DRAFT
81
-------
and PFOS accumulate in fish after the chemicals leave the farm field and travel over a 10-meter soil
buffer to the nearby surface water reservoir that is 13 acres in size. This transfer occurs in the models in
the water phase through overland flow of dissolved and particle-bound mass, though PFOS or PFOA
bound to particulates transported through the air could also be source to nearby waterbodies. The
models include daily-scale meteorological data, which allows the model to capture episodic increases in
runoff and erosion from storm events. This model does not include any connection between
groundwater and the surface water reservoir. The surface water can be thought of as a source of
drinking water or only as the route of PFOA and PFOS contamination to the fish. The concentration of
PFOA and PFOS is linearly correlated to the size of the modeled surface water reservoir, such that if the
volume of water in the reservoir increases by a given percentage, the concentration in PFOA and PFOS in
surface water and fish tissue will decrease by the same percentage.
Overall, the daft modeling finds that surface water concentrations for PFOS are lower than surface
water concentrations for PFOA across each climate and Koc scenario. However, PFOS fish tissue
concentrations are consistently higher than PFOA fish tissue concentrations in each scenario. This trend
is due to the high BAFs for PFOS, which are 1,700 (trophic level 3) and 860 (trophic level 4), compared to
the BAFs of 49 (trophic level 3) and 31 (trophic level 4) for PFOA. A recent FDA study using FDA's PFAS
methods for food had a maximum residue level (MRL) of 39 ppt for PFOS and 90 ppt for PFOA (FDA,
2022). Modeled concentrations of PFOS in fish tissue are consistently above MRLs in the low Koc
scenarios, but not in the high Koc scenarios, where more PFOS is retained in soil. The modeled
concentrations of PFOA in fish tissue are consistently below MRLs in FDA's PFAS methods.
In some instances, surface water bodies are used for drinking water instead of groundwater. The results
of the modeling exercises show that the concentrations of PFOA and PFOS in surface water are
consistently higher than the concentrations in groundwater in a given modeling run. This indicates that
those using a surface water reservoir as a source of drinking water would be expected to have higher
PFOA and PFOS drinking water exposure than those using groundwater as a source of drinking water,
assuming that biosolids are applied within ten meters of the reservoir. If biosolids were applied further
from the drinking water reservoir or the reservoir were larger, the concentrations of PFOA and PFOS
would decrease.
The PFOA and PFOS concentrations in fruits and vegetables predicted in these models are primarily
dependent on 1) uptake factors for the grouping of plants, 2) modeled retention of PFOA and PFOS in
the soils, and 3) the percent moisture factor used to convert dry weight to wet weight measurements.
There are significant data limitations on the uptake factors used for each category of fruits and
vegetables included in this assessment, which results in a high degree of uncertainty in the modeled
plant concentrations. As described in Section 2.9.3.4, these limitations on available data for uptake
factors in fruits and vegetables likely indicate that the exposures from fruits and vegetables are over-
estimated. Given these limitations, there are some general trends that the modeling can show us.
Though plant uptake factors are generally higher for PFOA than PFOS, more PFOS is generally retained in
soils due to PFOS's higher Koc. As a result, PFOA or PFOS concentrations can be higher in fruits and
vegetables depending on the climate and Koc setting. Exposed vegetables, where humans tend to eat
leaves, shoots, or stalks (i.e., spinach, celery, lettuce) tend to have the higher concentrations of PFOA
and PFOS due to the higher uptake factors. Overall, the modeled concentrations for fruits and
vegetables should be seen as rough estimates, with a high variability and uncertainty. Additional data of
PFOA and PFOS uptake into fruits and vegetables, especially when these plants are grown on biosolids-
impacted soils, would help reduce this uncertainty.
For reference, the exposures for each pathway for the crop farm are presented In Tables 20 and 21 in
units of ng/kg-day. These exposures are calculated using the consumption rates described in Section
DRAFT
82
-------
2.9.3.8 as well as other factors described in Sections 2.9.3.9 through 2.9.3.12. The lifetime average daily
dose (LADD) averages the daily exposure during the exposure duration over a lifetime of 70 years and is
used for calculating cancer risk (see Section 4.1, Equation 4). The average daily dose (ADD) averages the
daily exposure over the exposure duration of one year, not the full lifetime, and is used for calculating
noncancer hazard (see Section 4.1, Equation 5).
Table 20. PFOA Exposures for Crop Farm (ng/kg-day): LADD and ADD
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
LADD | ADD
LADD | ADD
LADD | ADD
LADD | ADD
Dry Climate
Exposed fruit
0.00015
0.0012
0.00016
0.0012
0.00035
0.0027
0.00036
0.0028
Exposed vegetable
0.0013
0.0096
0.00090
0.0069
0.0029
0.022
0.0021
0.016
Fish
0.017
0.13
0.020
0.15
0.0031
0.023
0.0036
0.027
Groundwater
0.00073
0.006
0.00076
0.0062
3.8E-12
2.8E-11
3.9E-12
2.9E-11
Protected fruit
0.00021
0.0016
0.00023
0.0018
0.00049
0.0038
0.00054
0.0041
Protected vegetable
0.0010
0.0078
0.0019
0.014
0.0024
0.018
0.0043
0.033
Root vegetable
0.00073
0.0056
0.00049
0.0037
0.0017
0.013
0.0011
0.0087
Soil
5.8E-07
5.4E-06
8.8E-06
8.2E-05
1.6E-06
1.3E-05
2.4E-05
0.00019
Surface water
0.014
0.11
0.015
0.11
0.0026
0.019
0.0028
0.02
Moderate Climate
Exposed fruit
9.00E-06
9.8E-05
9.2E-06
0.0001
0.00014
0.0011
0.00015
0.0011
Exposed vegetable
7.40E-05
0.00082
5.3E-5
0.00058
0.0012
0.0093
0.00085
0.0066
Fish
0.00092
0.0072
0.0011
0.0084
0.0031
0.024
0.0036
0.028
Groundwater
0.0082
0.074
0.0086
0.077
0.00023
0.0016
0.00024
0.0017
Protected fruit
1.3E-05
0.00014
1.4E-05
0.00015
0.00020
0.0016
0.00022
0.0017
Protected vegetable
6.1E-05
0.00066
0.00011
0.0012
0.00097
0.0076
0.0018
0.014
Root vegetable
4.3E-05
0.00047
2.9E-05
0.00032
0.00069
0.0054
0.00047
0.0036
Soil
9.6E-09
2.2E-07
1.5E-07
3.4E-06
5.0E-07
4.8E-06
7.6E-06
7.4E-05
Surface water
0.00078
0.0061
0.00081
0.0064
0.0026
0.02
0.0027
0.021
Wet Climate
Exposed fruit
8.1E-06
0.00011
8.3E-06
0.00012
0.00011
0.00084
0.00011
0.00085
Exposed vegetable
6.8E-05
0.00094
4.8E-05
0.00067
0.00091
0.0069
0.00065
0.0049
Fish
0.00061
0.0056
0.00072
0.0066
0.0022
0.017
0.0026
0.02
Groundwater
0.0083
0.060
0.0086
0.063
0.00088
0.0064
0.00092
0.0067
Protected fruit
1.1E-05
0.00016
1.2E-05
0.00017
0.00015
0.0012
0.00017
0.0013
Protected vegetable
5.5E-05
0.00077
0.0001
0.0014
0.00074
0.0056
0.0014
0.01
Root vegetable
3.9E-05
0.00055
2.6E-05
0.00037
0.00053
0.004
0.00036
0.0027
Soil
8.9E-09
2.0E-07
1.4E-07
3.1E-06
3.6E-07
3.4E-06
5.4E-06
5.1E-05
Surface water
0.00052
0.0048
0.00055
0.005
0.0019
0.014
0.0019
0.015
Table 21. PFOS Exposures for Crop Farm (ng/kg-day): LADD and ADD
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
LADD | ADD
LADD | ADD
LADD | ADD
LADD | ADD
Dry Climate
Exposed fruit
5.9E-05
0.00044
6.0E-05
0.00045
0.0001
0.00076
0.00010
0.00077
Exposed vegetable
0.00015
0.0012
0.00011
0.00083
0.00027
0.0020
0.00019
0.0014
Fish
0.29
2.1
0.34
2.5
0.0026
0.019
0.0030
0.022
Groundwater
9.9E-5
0.00078
0.00010
0.00082
7.4E-34
5.4E-33
7.7E-34
5.7E-33
Protected fruit
8.2E-05
0.00061
9.0E-05
0.00067
0.00014
0.0011
0.00015
0.0012
Protected vegetable
0.00013
0.00094
0.00023
0.0017
0.00022
0.0016
0.00040
0.0030
Root vegetable
0.0013
0.01
0.00090
0.0067
0.0023
0.017
0.0015
0.012
Soil
1.0E-06
8.8E-06
1.6E-05
0.00013
2.1E-06
1.6E-05
3.2E-05
0.00024
Surface water
0.0084
0.062
0.0087
0.065
7.2E-05
0.00053
7.5E-05
0.00055
DRAFT
83
-------
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
LADD | ADD
LADD | ADD
LADD | ADD
LADD | ADD
Moderate Climate
Exposed fruit
1.4E-05
0.00011
1.4E-05
0.00011
9.0E-05
0.00069
9.2E-05
0.0007
Exposed vegetable
3.6E-05
0.00028
2.5E-05
0.0002
0.00024
0.0018
0.00017
0.0013
Fish
0.11
0.83
0.13
0.97
0.0032
0.025
0.0037
0.029
Groundwater
0.0018
0.013
0.0019
0.014
3.2E-08
2.3E-07
3.4E-08
2.5E-07
Protected fruit
1.9E-05
0.00015
2.1E-05
0.00017
0.00013
0.00096
0.00014
0.0011
Protected vegetable
2.9E-05
0.00023
5.3E-05
0.00042
0.00019
0.0015
0.00035
0.0027
Root vegetable
0.00031
0.0025
0.00021
0.0016
0.0021
0.016
0.0014
0.011
Soil
1.4E-07
1.8E-06
2.2E-06
2.7E-05
1.4E-06
1.4E-5
2.2E-05
0.00021
Surface water
0.0032
0.024
0.0033
0.025
8.8E-5
0.00068
9.2E-05
0.00071
Wet Climate
Exposed fruit
9.8E-06
8.1E-05
1.0E-05
8.3E-05
8.1E-05
0.00062
8.3E-05
0.00063
Exposed vegetable
2.6E-05
0.00021
1.8E-05
0.00015
0.00021
0.0016
0.00015
0.0012
Fish
0.068
0.52
0.079
0.61
0.0035
0.027
0.0041
0.031
Groundwater
0.0049
0.036
0.0051
0.037
1.9E-5
0.00020
2.0E-5
0.00021
Protected fruit
1.4E-05
0.00011
1.5E-05
0.00012
0.00011
0.00086
0.00012
0.00095
Protected vegetable
2.1E-05
0.00017
3.8E-05
0.00032
0.00017
0.0013
0.00032
0.0024
Root vegetable
0.00022
0.0019
0.00015
0.0012
0.0019
0.014
0.0012
0.0095
Soil
8.4E-08
1.1E-06
1.3E-06
1.7E-05
1.4E-06
1.2E-05
2.2E-05
0.00019
Surface water
0.0020
0.015
0.0021
0.016
9.6E-5
0.00074
0.0001
0.00077
3.1.2 Pasture Farm
The pasture farm scenario models the fate and transport of PFOA and PFOS as they move from biosolids
through soil, surface water, and groundwater. The models then estimate the direct exposure to adults
and children to those media, and the uptake and exposure from those media to animal feed, animal
products, and fish. The pasture farm model includes the same assumptions about time living on the farm
as the crop farm model. Notably, the pasture farm model does not include annual tilling of the farm
fields, which is included in the crop farm model.
Table 22. PFOA Media Concentrations for Pasture Farm (ppt): Maximum 10- and 1-year Averages
Pathway
Low Koc
High Koc
10-yr | 1-yr
1—
>
l
T—
1—
>
1
o
T—
Dry Climate
Beef
5.2
7.7
31
32
Eggs
27
41
220
230
Fish3
340
340
140
140
Groundwater
2.8
2.8
0.026
0.026
Milk
8.4
12
44
46
Chicken
0.64
0.96
5.2
5.4
Soil
60
100
760
790
Surface Water
10
10
4.2
4.2
Moderate Climate
Beef
1
1.3
4.3
5.7
Eggs
4.2
6.4
30
42
Fish3
60
64
49
52
Groundwater
4.3
4.3
0.27
0.27
Milk
1.7
2.1
6.2
8.3
Chicken
0.099
0.15
0.71
0.97
Soil
4.8
12
100
140
Surface Water
1.8
1.9
1.5
1.5
Wet Climate
Beef
0.7
0.88
2.9
4.2
Eggs
2.9
4.2
20
30
DRAFT
84
-------
Pathway
Low Koc
High Koc
10-yr
1-yr
10-yr
1-yr
Fish3
36
39
29
30
Groundwater
2.6
2.6
0.78
0.78
Milk
1.2
1.4
4.3
6.1
Chicken
0.067
0.098
0.47
0.69
Soil
3
7.8
65
97
Surface Water
1.1
1.2
0.86
0.88
a These values represent the weighted average fish tissue concentration by the percent consumption of trophic levels 3 and 4.
Table 23. PFOS Media Concentrations for Pasture Farm (ppt): Maximum 10- and 1-year Averages
Pathway
Low Koc
High Koc
10-yr | 1-yr
10-yr | 1-yr
Dry Climate
Beef
120
140
280
290
Eggs
160
200
550
570
Fish3
8100
8300
240
240
Groundwater
0.22
0.22
6.1E-31
6.1E-31
Milk
20
24
40
41
Chicken
17
21
57
60
Soil
280
350
1100
1100
Surface Water
8.3
8.5
0.24
0.24
Moderate Climate
Beef
29
33
170
180
Eggs
24
33
340
350
Fish3
2300
2500
230
230
Groundwater
1.1
1.1
6.8E-4
6.8E-4
Milk
5.1
5.7
25
26
Chicken
2.6
3.4
35
37
Soil
29
46
670
710
Surface Water
2.4
2.5
0.22
0.23
Wet Climate
Beef
11
21
110
120
Eggs
13
22
220
230
Fish3
1300
1400
160
170
Groundwater
2
2
0.012
0.012
Milk
1.9
3.6
16
17
Chicken
1.4
2.3
23
24
Soil
21
34
430
450
Surface Water
1.4
1.4
0.16
0.17
a These values represent the weighted average fish tissue concentration by the percent consumption of trophic levels 3 and 4.
The pasture farm scenario outputs concentrations over time for milk and beef, eggs and chicken or
poultry, fish, surface water, soil, and groundwater. Groundwater concentrations range from 0.026 to 4.3
ng/L for PFOA and effectively 0 to 2 ng/L for PFOS. Surface water concentrations range from 0.86 to 10
ng/L for PFOA and 0.16 to 8.5 ng/L for PFOS. Soil concentrations range from 3 to 790 ng/kg for PFOA and
21 to 1,100 ng/kg for PFOS. Fish tissue concentrations range from 29 to 340 for PFOA and 160 to 8,300
ng/kg for PFOS. Milk concentrations range from 1.2 to 46 ng/L for PFOA and 1.9 to 41 ng/L for PFOS.
Beef concentrations range from 0.7 to 32 ng/kg for PFOA and 11 to 290 ng/kg for PFOS. Egg
concentrations range from 2.9 to 230 ng/kg for PFOA and 13 to 570 ng/kg for PFOS. Finally, chicken
ranges from 0.67 to 5.4 ng/kg for PFOA and 1.4 to 60 ng/kg for PFOS.
The trends in soil and groundwater concentrations for PFOA and PFOS seen in the pasture farm model
are similar to those seen in the crop farm model, where PFOA concentrations are higher in groundwater
and PFOS concentrations are higher in soils; however, maximum estimated soil concentrations are
DRAFT
85
-------
higher in pasture than crop farms. The pasture model does not include tilling of biosolids into the top 20
cm of soil, which results in slightly lower groundwater concentrations and higher soil concentrations.
The higher soil concentrations result in a higher loading of runoff into surface water, which results in
higher fish tissue concentrations for PFOS. All modeled groundwater results for PFOS would fall below
the current MRL for EPA drinking water methods, but the low Koc PFOA results for some climate
scenarios would exceed the existing MRLs.
Trends in surface water and fish concentrations for PFOA and PFOS are also similar between the crop
farm model and the pasture farm model. In general, the lack of tilling in the pasture model results in
more PFOA and PFOS at the surface, available for erosion and runoff into the nearby waterbody. This
correspondingly allows for more PFOA and PFOS to be available for fish uptake.
Dairy cows can be exposed to PFOA and PFOS through their feed, forage materials, drinking water, and
soil exposure. This model uses uptake factors for lactating dairy cows when calculating both meat and
milk concentrations, and assumes that cows are eating non-contaminated grain, but contaminated
silage, forage (grass), water, and soil. Overall, high Koc settings result in higher PFOA and PFOS milk and
beef concentrations than low Koc settings. Higher Koc settings result in more PFOA and PFOS partitioning
to the soils, which in this model also allows more PFOA and PFOS to be available for uptake into forage
and silage. Compared to feed, soil is a less significant vector of exposure to cows. A 2012 FDA survey of
PFAS concentrations in commercially available milk used a method with MDLs of 120 ppt for PFOA and
130 ppt for PFOS (FDA, 2012); all modeled concentrations fall below these detection thresholds. A more
recent datasetfrom the FDA total diet study (released in 2023) had MDLs of 24 ng/kgfor PFOA and 28
ng/kg for PFOS, which was applicable for beef samples (FDA 2023). The modeled results for PFOS were
consistently above that MDL, but results for PFOA were often below the MDL. It is important to note
that the beef results for PFOA are modeling uptake from dairy cows into muscle; a different BTF would
be needed to understand PFOA accumulation into the edible tissues of cows typically raised for beef.
Additional data on PFOA and PFOS uptake into beef would help to reduce the uncertainty around these
modeled results.
Chickens can also be exposed to PFOA and PFOS through their feed, forage materials, drinking water,
and soil exposure. This model uses uptake factors for laying hens when calculating both the egg and
meat concentrations. Similar to the cow results, chicken results show that there are higher modeled egg
and meat concentrations when Koc is high and in dry climate conditions, where more PFOA and PFOS are
retained in the soil. Again, a recent FDA total diet study (FDA 2023) had MDLs of 24 ng/kg for PFOA and
28 ng/kg for PFOS, which was applicable for egg samples. Modeled egg concentrations for PFOS are
consistently above that MRL, but modeled egg concentrations for PFOA are sometimes below that MRL.
For reference, the exposures for each pathway for the pasture farm are presented In Tables 24 and 25 in
units of ng/kg-day. These exposures are calculated using the consumption rates described in Section
2.9.3.8 as well as other factors described in Sections 2.9.3.9 through 2.9.3.12. The LADD is used for
calculating cancer risk and the ADD for noncancer hazard.
Table 24. PFOA Exposures for Pasture Farm (ng/kg-day): LADD and ADD
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
LADD | ADD
LADD | ADD
LADD | ADD
LADD | ADD
Dry Climate
Beef
0.0011
0.012
0.0015
0.016
0.0067
0.051
0.0088
0.067
Eggs
0.0026
0.029
0.0026
0.029
0.022
0.16
0.022
0.16
Fish3
0.022
0.16
0.025
0.19
0.0091
0.067
0.011
0.079
Groundwater
0.0052
0.038
0.0054
0.04
4.7E-05
0.00035
5.0E-05
0.00036
DRAFT
86
-------
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
LADD
ADD
LADD
ADD
LADD
ADD
LADD
ADD
Milk
0.014
0.15
0.025
0.27
0.073
0.55
0.13
1.0
Chicken
9.6E-05
0.0011
0.00017
0.0019
0.00078
0.006
0.0014
0.011
Soil
1.0E-06
1.3E-05
1.6E-05
0.00019
1.3E-05
9.9E-05
0.0002
0.0015
Surface water
0.018
0.14
0.019
0.14
0.0077
0.057
0.008
0.06
Moderate Climate
Beef
0.00022
0.0021
0.00029
0.0028
0.00093
0.0091
0.0012
0.012
Eggs
0.00041
0.0045
0.00041
0.0045
0.0029
0.029
0.0029
0.029
Fish3
0.0038
0.030
0.0045
0.035
0.0032
0.024
0.0037
0.028
Groundwater
0.0078
0.057
0.0082
0.06
0.00049
0.0036
0.00051
0.0037
Milk
0.0028
0.026
0.0051
0.047
0.010
0.10
0.019
0.18
Chicken
1.5E-05
0.00016
2.7E-05
0.0003
0.00011
0.0011
0.00019
0.0019
Soil
8.3E-08
1.5E-06
1.3E-06
2.4E-05
1.7E-06
1.7E-05
2.6E-05
0.00027
Surface water
0.0033
0.026
0.0034
0.027
0.0027
0.021
0.0028
0.021
Wet Climate
Beef
0.00015
0.0014
0.00020
0.0018
0.00064
0.0067
0.00084
0.0088
Eggs
0.00028
0.0029
0.00028
0.0029
0.0019
0.021
0.0019
0.021
Fish3
0.0023
0.018
0.0027
0.022
0.0019
0.014
0.0022
0.016
Groundwater
0.0047
0.035
0.0049
0.036
0.0014
0.010
0.0015
0.011
Milk
0.0019
0.017
0.0036
0.031
0.0071
0.074
0.013
0.13
Chicken
1.0E-05
0.00011
1.8E-05
0.0002
7.1E-05
0.00076
0.00013
0.0014
Soil
5.1E-08
9.7E-07
7.8E-07
1.5E-05
1.1E-06
1.2E-05
1.7E-05
0.00018
Surface water
0.0020
0.016
0.0021
0.016
0.0016
0.012
0.0016
0.012
Table 25. PFOS Exposures for Pasture Farm (ng/kg-day): LADD and ADD
Pathway
Low Koc
Hiqh Koc
Adult
Child
Adult
Child
LADD | ADD
LADD | ADD
LADD | ADD
LADD | ADD
Dry Climate
Beef
0.027
0.23
0.035
0.30
0.062
0.47
0.081
0.61
Eggs
0.016
0.14
0.016
0.14
0.053
0.40
0.053
0.40
Fish3
0.52
3.9
0.61
4.5
0.016
0.11
0.018
0.13
Groundwater
0.00041
0.0030
0.00043
0.0031
1.1E-33
8.2E-33
1.2E-33
8.6E-33
Milk
0.032
0.29
0.059
0.53
0.065
0.50
0.12
0.91
Chicken
0.0026
0.023
0.0047
0.041
0.0087
0.066
0.016
0.12
Soil
4.9E-06
4.3E-05
7.4E-05
0.00066
1.9E-05
0.00014
0.00029
0.0022
Surface water
0.015
0.11
0.016
0.12
0.00043
0.0032
0.00045
0.0033
Moderate Climate
Beef
0.0063
0.052
0.0083
0.069
0.038
0.29
0.050
0.38
Eggs
0.0023
0.023
0.0023
0.023
0.032
0.25
0.032
0.25
Fish3
0.15
1.2
0.18
1.4
0.015
0.11
0.017
0.13
Groundwater
0.0021
0.015
0.0022
0.016
1.2E-06
9.1E-06
1.3E-06
9.5E-06
Milk
0.0085
0.069
0.015
0.13
0.041
0.31
0.075
0.57
Chicken
0.00039
0.0038
0.0007
0.0068
0.0053
0.041
0.0096
0.074
Soil
5.1E-07
5.7E-06
7.7E-06
8.7E-05
1.1E-05
8.9E-05
0.00017
0.0013
Surface water
0.0044
0.034
0.0046
0.035
0.00041
0.0030
0.00043
0.0032
Wet Climate
Beef
0.0024
0.034
0.0032
0.044
0.025
0.19
0.033
0.25
Eggs
0.0012
0.015
0.0012
0.015
0.021
0.16
0.021
0.16
Fish3
0.087
0.66
0.10
0.77
0.011
0.081
0.012
0.095
Groundwater
0.0036
0.026
0.0038
0.027
2.1E-05
0.00016
2.2E-05
0.00016
Milk
0.0031
0.044
0.0057
0.080
0.027
0.20
0.049
0.37
Chicken
0.00020
0.0025
0.00037
0.0045
0.0034
0.026
0.0062
0.048
Soil
3.5E-07
4.3E-06
5.4E-06
6.5E-05
7.4E-06
5.7E-05
0.00011
0.00086
Surface water
0.0025
0.019
0.0026
0.020
0.00029
0.0022
0.00030
0.0023
DRAFT
87
-------
3.1.3 Redama tion Site
The reclamation site model is similar to the pasture farm model, except the reclamation site models a
single large application of biosolids rather than ongoing applications of biosolids at an agronomic rate.
Assumptions about the duration of time a family spends living near the reclamation site are the same as
described for the crop and pasture farm models (10 years). The reclamation site model is also run in dry,
moderate, and wet climate settings. This modeling exercise assumes that a dairy farm is established at
the site, which is thought to represent a location being reclaimed from over-grazing. However, any of
the pathways related to soil, surface water, groundwater, and fish are relevant to many other
reclamation scenarios of a similar size (one application of biosolids to 80 acres of remediated land).
Table 26. PFOA Media Concentrations for Reclamation Site (ppt): Maximum 10- and 1-year Averages
Low Koc
High Koc
Pathway
10-yr
1-yr
10-yr
1-yr
Dry Climate
Beef
3.8
7.3
7.0
7.6
Eggs
26
56
52
58
Fish3
57
58
18
18
Groundwater
0.17
0.17
0.003
0.003
Milk
5.7
11
10
11
Chicken
0.61
1.3
1.2
1.3
Soil
84
200
180
200
Surface water
1.7
1.7
0.55
0.55
Moderate Climate
Beef
0.8
4.7
5.7
7.5
Eggs
5.5
36
40
56
Fish3
6.5
00
CO
14
15
Groundwater
0.054
0.42
0.023
0.023
Milk
1.2
6.8
8.3
11
Chicken
0.13
0.83
0.93
1.3
Soil
18
120
130
190
Surface water
0.19
0.26
0.42
0.45
Wet Climate
Beef
0.35
2.1
4.6
6.7
Eggs
2
15
31
49
Fish3
8.4
15
14
16
Groundwater
0.24
2.4
0.032
0.032
Milk
0.54
3
6.7
9.7
Chicken
0.046
0.34
0.72
1.1
Soil
5.1
48
97
160
Surface water
0.25
0.44
0.42
0.48
a These values represent the weighted average fish tissue concentration by the percent consumption of trophic levels 3 and 4.
Table 27. PFOS Media Concentrations for Reclamation Site (ppt): Maximum 10- and 1-year Averages
Low Koc
High Koc
Pathway
10-yr
I 1-yr
10-yr
1-yr
Dry Climate
Beef
45
51
48
51
Eggs
83
100
92
100
Fish3
1200
1300
33
33
Groundwater
0.032
0.032
2.1E-32
2.1E-32
Milk
6.6
7.2
6.8
7.2
Chicken
8.7
10
9.6
10
Soil
160
200
180
200
Surface water
1.3
1.3
0.032
0.032
DRAFT
88
-------
Pathway
Low Koc
High Koc
10-yr | 1-yr
10-yr | 1-yr
Moderate Climate
Beef
22
48
42
50
Eggs
34
89
77
98
Fish3
480
580
30
31
Groundwater
0.13
0.13
1.1 E-5
1.1 E-5
Milk
3.4
6.9
6
7.1
Chicken
3.5
9.4
8.1
10
Soil
62
180
150
200
Surface water
0.49
0.59
0.03
0.03
Wet Climate
Beef
20
40
36
46
Eggs
26
71
62
88
Fish3
660
850
28
30
Groundwater
0.12
0.12
3.1 E-5
3.1 E-5
Milk
3.2
6.1
5.2
6.6
Chicken
2.8
7.4
6.5
9.2
Soil
45
140
120
180
Surface water
0.68
0.87
0.027
0.029
a These values represent the weighted average fish tissue concentration by the percent consumption of trophic levels 3 and 4.
The reclamation scenario outputs concentrations over time for milk and beef, eggs and chicken or
poultry, fish, surface water, soil, and groundwater. Groundwater concentrations range from 0.003 to 2.4
ng/L for PFOA and effectively 0 to 0.13 ng/Lfor PFOS. Surface water concentrations range from 0.19 to
1.7 ng/L for PFOA and 0.027 to 1.3 ng/L for PFOS. Fish tissue concentrations range from 6.5 to 58 for
PFOA and 28 to 1300 ng/kg for PFOS. Soil concentrations range from 5.1 to 200 ng/kg for PFOA and 45
to 200 ng/kg for PFOS. Milk concentrations range from 0.54 to 11 ng/L for PFOA and 3.2 to 7.2 for PFOS.
Beef concentrations range from 0.35 to 7.6 ng/kg for PFOA and 20 to 51 ng/kg for PFOS. Egg
concentrations range from 2 to 58 ng/kg for PFOA and 26 to 100 ng/kg for PFOS. Finally, chicken ranges
from 0.046 to 1.3 ng/kg for PFOA and 2.8 to 10 ng/kg for PFOS.
Groundwater concentrations in the remediation scenario are lower than those modeled in the pasture
farm model. Though the remediation scenario assumed a higher rate of biosolids application than the
pasture farm scenario (50 vs 10 DMT per field hectare), the remediation scenario only included a single
application of biosolids, while the pasture farm scenario included annual applications for 40 years. This
modeling suggests that a single application of low concentration biosolids is unlikely to result in a
detectable PFOA concentration in groundwater, though this outcome is more likely in scenarios where
the underlying soils had a low Koc (meaning low soil sorption), such as sandy soils or soils damaged by
human activity in a way that results in geochemical conditions less conducive to soil sorption. One
example of a soil condition resulting in low PFOA and PFOS sorption is high soil pH; at normal soil pH
ranges, the pKa of PFOA and PFOS indicate they would be negatively charged such that lower soil pH
results in higher rates of nonspecific anion absorption (Oliver et al., 2019). Given that soil remediation
can occur in a variety of depleted or disturbed sites, it is possible that these low sorption conditions are
relevant to many biosolids reuse scenarios where the biosolids are used to remediate disturbed soils.
Soil, surface water, and fish tissue concentrations are also lower in the remediation scenario than in the
pasture farm scenario. However, there are smaller differences between the location settings and high
Koc and low Koc settings in the remediation scenario than in the pasture farm scenario. This is likely
because, when biosolids are only applied one time, the maximum PFOA and PFOS concentrations are
reached more rapidly, and differences in leaching potential from the soil over time are less impactful on
the maximum observed concentration.
DRAFT
89
-------
Beef, chicken, eggs, and milk concentrations are also lower in the remediation scenario than the pasture
farm scenario, where PFOA and PFOS concentration have time over repeated applications to accumulate
in soils. Because the media concentrations in these scenarios are lower than in the pasture farm setting,
many (but not all) of the modeled concentrations would fall below currently available MDLs. However,
given the high bioaccumulation of PFOS in fish and eggs, these media would consistently have
detectable concentrations of PFOS in this scenario.
For reference, the exposures for each pathway for the reclamation site are presented In Tables 28 and
29 in units of ng/kg-day. These exposures are calculated using the consumption rates described in
Section 2.9.3.8 as well as other factors described in Sections 2.9.3.9 through 2.9.3.12. The LADD is used
for calculating cancer risk and the ADD for noncancer hazard.
Table 28. PFOA Exposures for Reclamation Site (ng/kg-day): LADD and ADD
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
LADD | ADD
LADD | ADD
LADD | ADD
LADD | ADD
Dry Climate
Beef
0.00084
0.012
0.0011
0.015
0.0015
0.012
0.0020
0.016
Eggs
0.0025
0.039
0.0025
0.039
0.0050
0.040
0.0050
0.040
Fish3
0.0037
0.027
0.0043
0.032
0.0012
0.0087
0.0014
0.010
Groundwater
0.00032
0.0023
0.00033
0.0024
5.4E-06
4.0E-05
5.7E-06
4.1E-05
Milk
0.0094
0.13
0.017
0.23
0.017
0.13
0.031
0.24
Chicken
9.3E-05
0.0014
0.00017
0.0026
0.00018
0.0015
0.00033
0.0027
Soil
1.4E-06
2.5E-05
2.2E-05
0.00038
3.1E-06
2.5E-05
4.7E-05
0.00038
Surface water
0.0031
0.023
0.0033
0.024
0.0010
0.0074
0.0010
0.0077
Moderate Climate
Beef
0.00017
0.0076
0.00023
0.0099
0.0012
0.012
0.0016
0.016
Eggs
0.00053
0.025
0.00053
0.025
0.0038
0.039
0.0038
0.039
Fish3
0.00042
0.0041
0.00049
0.0048
0.00091
0.0072
0.0011
0.0084
Groundwater
0.00010
0.0056
0.00010
0.0059
4.1E-05
0.00030
4.3E-05
0.00032
Milk
0.0019
0.082
0.0035
0.15
0.014
0.13
0.025
0.24
Chicken
1.9E-05
0.00091
3.5E-05
0.0017
0.00014
0.0014
0.00025
0.0026
Soil
3.0E-07
1.5E-05
4.6E-06
0.00023
2.2E-06
2.4E-05
3.4E-05
0.00037
Surface water
0.00035
0.0035
0.00037
0.0037
0.00078
0.0061
0.00081
0.0064
Wet Climate
Beef
7.6E-05
0.0033
0.00010
0.0043
0.0010
0.011
0.0013
0.014
Eggs
0.00019
0.010
0.00019
0.010
0.0030
0.034
0.0030
0.034
Fish3
0.00054
0.0070
0.00063
0.0081
0.00091
0.0076
0.0011
0.0089
Groundwater
0.00045
0.032
0.00047
0.034
5.8E-05
0.00042
6.1E-05
0.00044
Milk
0.00088
0.036
0.0016
0.066
0.011
0.12
0.02
0.21
Chicken
7.0E-06
0.00037
1.3E-05
0.00068
0.00011
0.0012
0.0002
0.0023
Soil
8.8E-08
6.0E-06
1.3E-06
9.2E-05
1.7E-06
2.0E-05
2.5E-05
0.00031
Surface water
0.00046
0.0059
0.00048
0.0062
0.00077
0.0065
0.00080
0.0068
Table 29. PFOS Exposures for Reclamation Site (ng/kg-day): LADD and ADD
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
LADD | ADD
LADD | ADD
LADD | ADD
LADD | ADD
Dry Climate
Beef
0.0099
0.082
0.013
0.11
0.01
0.082
0.014
0.11
Eggs
0.0079
0.070
0.0079
0.07
0.0088
0.070
0.0088
0.070
Fish3
0.080
0.59
0.093
0.69
0.0021
0.016
0.0025
0.018
Groundwater
5.8E-05
0.00042
6.1E-05
0.00044
3.9E-35
2.8E-34
4.0E-35
2.9E-34
Milk
0.011
0.087
0.020
0.16
0.011
0.087
0.020
0.16
Chicken
0.0013
0.011
0.0024
0.021
0.0015
0.012
0.0026
0.021
DRAFT
90
-------
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
LADD
ADD
LADD
ADD
LADD
ADD
LADD
ADD
Soil
2.8E-06
2.5E-5
4.2E-05
0.00038
3.1E-06
2.5E-05
4.8E-05
0.00038
Surface water
0.0023
0.017
0.0024
0.018
5.8E-05
0.00044
6.1E-05
0.00046
Moderate Climate
Beef
0.0048
0.076
0.0063
0.10
0.0092
0.081
0.012
0.11
Eggs
0.0032
0.063
0.0032
0.063
0.0074
0.069
0.0074
0.069
Fish3
0.031
0.27
0.036
0.32
0.0020
0.015
0.0023
0.017
Groundwater
0.00024
0.0018
0.00025
0.0019
2.0E-08
1.5E-07
2.1E-08
1.5E-07
Milk
0.0056
0.082
0.010
0.15
0.0099
0.085
0.018
0.16
Chicken
0.00053
0.010
0.00097
0.019
0.0012
0.011
0.0022
0.021
Soil
1.1E-06
2.2E-05
1.6E-05
0.00034
2.6E-06
2.5E-05
4.0E-05
0.00037
Surface water
0.00090
0.0079
0.00094
0.0082
5.4E-05
0.00041
5.7E-05
0.00043
Wet Climate
Beef
0.0043
0.065
0.0056
0.085
0.0078
0.074
0.010
0.098
Eggs
0.0025
0.049
0.0025
0.049
0.0059
0.062
0.0059
0.062
Fish3
0.043
0.40
0.050
0.47
0.0018
0.014
0.0021
0.017
Groundwater
0.00021
0.0015
0.00022
0.0016
5.6E-08
4.1E-07
5.9E-08
4.3E-07
Milk
0.0052
0.073
0.0095
0.13
0.0086
0.079
0.016
0.15
Chicken
0.00041
0.0081
0.00075
0.015
0.00098
0.010
0.0018
0.018
Soil
7.7E-07
1.8E-05
1.2E-05
0.00027
2.0E-06
2.2E-05
3.1E-05
0.00033
Surface water
0.0012
0.012
0.0013
0.012
5.0E-05
0.00039
5.3E-05
0.00041
3.1.4 Sewage Sludge Disposal Site
The sewage sludge disposal site models the fate and transport of PFOA and PFOS after they are disposed
of in an unlined, lined with a composite liner, or clay-lined impoundment. This scenario assumes that the
biosolids being disposed of are not dewatered because this a common practice across the U.S. and the
practice more likely to result in groundwater infiltration risks. The model considers infiltration from the
impoundment through soil and into groundwater. The model then calculates PFOA and PFOS
concentrations in groundwater that is used for drinking water. The results in the table below report
groundwater concentrations in wet, moderate, and dry climates in a well screened up to 2 m below the
water table and 5 m distance from the impoundment site. These climate scenarios also represent the
varied soil types, depths to groundwater, hydrological conditions that would be expected in these three
climate settings. This scenario assumes that an adult's lifetime only includes 10 years living near the
impoundment. The following tables show the modeled concentrations of PFOA and PFOS in each
disposal site liner type during either a ten year or one year of averaging time. These averaging windows
include the maximum concentration year for groundwater.
DRAFT
91
-------
Table 30. PFOA Groundwater Concentrations for Sludge Disposal Unit (ppt): Maximum 10- and 1-year
Averages by Liner Scenario
Liner
Low Koc
High Koc
10-yr | 1-yr
10-yr | 1-yr
Dry Climate
No Liner
25
25
0.075
0.077
Clay liner
21
21
0.049
0.050
Composite liner
0.013
0.014
1.6E-30
1.6E-30
Moderate Climate
No Liner
00
CO
8.9
0.024
0.025
Clay liner
5.8
5.8
0.016
0.016
Composite liner
0.0011
0.0011
1.5E-30
1.5E-30
Wet Climate
No Liner
16
17
0.17
0.17
Clay liner
12
13
0.077
0.078
Composite liner
0.0041
0.0041
8.7E-13
8.9E-13
Table 31. PFOS Groundwater Concentrations for Sludge Disposal Unit (ppt): Maximum 10- and 1-year
Averages by Liner Scenario
Liner
Low Koc
High Koc
10-yr | 1-yr
10-yr | 1-yr
Dry Climate
No Liner
1.3
1.3
0.00046
0.00048
Clay liner
0.91
0.93
0.00031
0.00033
Composite liner
2.3E-06
2.3E-06
2E-32
2E-32
Moderate Climate
No Liner
0.43
0.44
0.00018
0.00018
Clay liner
0.25
0.25
0.00010
0.00011
Composite liner
4.5E-14
4.6E-14
2.2E-32
2.3E-32
Wet Climate
No Liner
2.2
2.2
0.0022
0.0023
Clay liner
1.2
1.2
0.00092
0.00097
Composite liner
1.2E-05
1.3E-05
3.2E-32
3.4E-32
The surface disposal scenario outputs groundwater concentrations over time for three types of disposal
sites: unlined, clay-lined, and lined with a composite liner. As expected, groundwater concentrations are
the highest in unlined surface disposal sites (PFOA from 0.024 to 25 ng/L; PFOS from essentially zero to
2.2 ng/L). Clay-lined surface disposal sites have slightly lower groundwater concentrations than unlined
sites. Finally, composite-lined surface disposal sites result in very low groundwater infiltration, with
essentially no infiltration of PFOS and only low breakthrough for PFOA (PFOA groundwater
concentrations from zero to 0.014 ng/L; PFOS remains essentially zero in all scenarios). Differences in
modeled groundwater concentrations between dry, moderate and wet climates reflect the differences
in depth to the water table, infiltration rate, and the amount of dilution of the disposal site material with
rainfall in each hypothetical setting.
For reference, the exposures for groundwater for the sludge disposal unit are presented In Tables 32
and 33 in units of ng/kg-day. These exposures are calculated using the consumption rates described in
Section 2.9.3.8 as well as other factors described in Sections 2.9.3.9 through 2.9.3.12. The LADD is used
for calculating cancer risk and the ADD for noncancer hazard.
DRAFT
92
-------
Table 32. PFOA Exposures for Surface Disposal Site (ng/kg-day): LADD and ADD
Liner
Low Koc
Hiqh
Koc
Adult
Child
Adult
Child
LADD | ADD
LADD | ADD
LADD | ADD
LADD | ADD
Dry Climate
No liner
2.5E-05
0.00018
2.6E-05
0.00019
2.9E-33
2.2E-32
3E-33
2.3E-32
Clay liner
0.046
0.34
0.048
0.35
0.00014
0.001
0.00014
0.0011
Composite liner
0.038
0.28
0.039
0.29
9.1E-05
0.00068
9.5E-05
0.00071
Moderate Climate
No liner
2.0E-06
1.5E-05
2.1E-06
1.6E-05
2.7E-33
2.1E-32
2.9E-33
2.2E-32
Clay liner
0.016
0.12
0.017
0.12
4.4E-05
0.00033
4.6E-05
0.00034
Composite liner
0.011
0.078
0.011
0.081
2.9E-05
0.00021
3.0E-05
0.00022
Wet Climate
No liner
7.5E-06
5.5E-05
7.8E-06
5.7E-05
1.6E-15
1.2E-14
1.7E-15
1.2E-14
Clay liner
0.029
0.22
0.03
0.23
0.00031
0.0023
0.00032
0.0024
Composite liner
0.023
0.17
0.024
0.18
0.00014
0.0011
0.00015
0.0011
Table 33. PFOS Exposures for Surface Disposal Site (ng/kg-day): LADD and ADD
Liner
Low Koc
Hiqh Koc
Adult
Child
Adult
Child
LADD | ADD
LADD | ADD
LADD | ADD
LADD | ADD
Dry Climate
No liner
0.0024
0.018
0.0025
0.019
8.4E-07
6.4E-06
8.8E-07
6.7E-06
Clay liner
0.0017
0.012
0.0018
0.013
5.7E-07
4.4E-06
6E-07
4.6E-06
Composite liner
4.1E-09
3.1E-08
4.3E-09
3.2E-08
3.6E-35
2.7E-34
3.8E-35
2.9E-34
Moderate Climate
No liner
0.00079
0.0059
0.00083
0.0062
3.2E-07
2.5E-06
3.4E-07
2.6E-06
Clay liner
0.00046
0.0034
0.00048
0.0035
1.8E-07
1.4E-06
1.9E-07
1.5E-06
Composite liner
8.3E-17
6.2E-16
8.6E-17
6.5E-16
4E-35
3.1E-34
4.2E-35
3.2E-34
Wet Climate
No liner
0.004
0.029
0.0041
0.031
4.0E-06
3.1E-05
4.1E-06
3.2E-05
Clay liner
0.0021
0.016
0.0022
0.016
1.7E-06
1.3E-05
1.8E-06
1.4E-05
Composite liner
2.3E-08
1.7E-07
2.4E-08
1.8E-07
5.8E-35
4.5E-34
6.1E-35
4.7E-34
3.1.5 Implica tions for Home Gardening
This assessment does not explicitly model how use of Class AEq biosolids in home gardens could impact
soil, fruit and vegetable, and groundwater concentrations. Class AEq biosolids have no application
requirements; they do not need to be applied at agronomic rates. Sizes of home gardens vary greatly
but are generally much smaller than a field used for growing crops at even a small commercial farm. The
smaller application areas for Class AEq biosolids at a given site likely reduces concerns over PFOA and
PFOS impacts to surface water and groundwater, though if larger amounts of biosolids were bulk
applied to a hobby farm or community garden as fertilizer, there could be potential impacts.
There is a high degree of uncertainty in the rates of PFOA and PFOS uptake to fruits and vegetables.
With the limited data available, it appears that vegetables like spinach and lettuce are the most likely to
uptake PFOA and PFOS, with PFOA exhibiting higher rates of uptake than PFOS. It is conceivable that a
home gardener using biosolids-based products in their raised beds or backyard garden could apply
enough biosolids, potentially over multiple years, to sufficiently elevate PFOA and PFOS concentrations
in soils such that detectable levels of PFOA and PFOS could be found in some fruits and vegetables. It is
also possible that a home gardener with backyard chickens could have enough PFOA and PFOS in
vegetable scraps, soil, grubs, and grass to result in measurable concentration of PFOA and PFOS in eggs.
DRAFT
93
-------
Finally, it is possible that homes have been developed on land that was previously used as farmland and
that had historic biosolids land application. Homeowners living in these developments could start a
garden with or without adding any new biosolids-based products to their garden beds. Given the long
residency times for PFOA and PFOS in soils in the crop and pasture modeling scenarios, it is possible that
a home gardener could be exposed to PFOA and PFOS in homegrown food or home raised eggs if they
are living on land that previously accepted agronomic land application, even if that homeowner does not
add any new PFOA or PFOS to their yard.
3.1.6 Other Land AppUca tion Use Scenarios
As described in the conceptual model section of this report (Section 2.8), there are numerous potential
land application scenarios that have not been explicitly modeled in this report, including biosolids
applications to forests, tree farms, road construction sites, golf courses, and more. If these locations are
considered "low public contact," potential pathways of exposure include groundwater used for drinking
water, surface water used for drinking water, and fish consumption from an impacted waterbody. If
biosolids are applied in an area with potential for soil exposure, this pathway could be relevant as well.
The scenarios modeled in this draft risk assessment are also not designed to explicitly account for
exposures that may occur where Class AEq biosolids are applied at non-agricultural sites. Soil
concentrations at sites where Class AEq biosolids application have occurred may be roughly described by
the pasture farming scenario, however, farmers are required to limit application rates for Class A and B
biosolids to the nutrient needs for the crop at the farm. Class AEq biosolids are sold to the general public
and landscapers and may be used without an understanding of matching the nutrient need of the soil to
the application rate, so over-application is possible. Therefore, it may be appropriate to consider that
soil concentrations could more rapidly rise in a Class Aeq application scenario than in a crop or pasture
farm scenario. This assessment is using a central tendency incidental soil ingestion rate (40 mg/day for
children aged 1-5), when incidental soil ingestion is evaluated for children. When creating CERCLA
screening values for residential areas an upper percentile rate is used for children (200 mg/day).
Conservatism in exposure assessment for Class AEq biosolids is warranted for children's incidental soil
ingestion exposures given that larger number of children may be exposed at homes, playgrounds, parks
or other areas where Class AEQ biosolids may be used in larger proportion than other land application
sites like farms.
The trends observed in the modeling performed for remediation sites and farms can inform the types of
concentrations expected in some other types of land application scenarios, acknowledging that each
land application scenario is unique. For example, annual application of biosolids to a golf course or turf
farm, applied at agronomic rates for fertilizing turf grass, is likely to show similar soil, surface water, fish,
and groundwater concentration trends as the pasture farm scenario, with the understanding that the
size of the biosolids-applied area will linearly scale with the final modeled media concentrations.
Similarly, annual application of biosolids to a forest or tree farm could result in similar media
concentration trends as the pasture farm scenario, with the caveat that silviculture or forested areas
likely have meaningfully different rates of runoff and erosion than a grass field. Additionally, a forested
land application scenario could have some amount of PFOA and PFOS uptake into trees, which could
result in less mass available for runoff into a nearby waterbody or infiltration into groundwater.
Using biosolids during road construction is a somewhat common practice. For example, a recent report
from the City of Juneau, Alaska, explains that there is a growing market for biosolids use as an erosion
control technique for construction projects, including road construction (City and Borough of Juneau
Wastewater Utility, 2017). The report explains that biosolids pellets can be used to enhance topsoil, to
fill void spaces and limit channelized flow of water on roadsides, provide a more permeable surface to
promote infiltration, and aid in revegetation along the road. The report adds that dried and pelletized
DRAFT
94
-------
biosolids could be used in a filter sock to prevent water from reaching storm drains as a replacement for
a silt fence or straw bale barrier for stormwater control. The modeling included in this report would not
capture the fate and transport of PFOA and PFOS when biosolids are used in this manner, though it is
possible that soils, surface water, groundwater, and fish may all be impacted from PFOA and PFOS in
these settings.
Mine reclamation is another type of biosolids land application that is not explicitly modeled in this
assessment. When biosolids are used in mine reclamation, there is generally one or a small number of
larger application of biosolids to increase the organic material and/or pH at the site. Former mining sites
can vary greatly in size and hydrogeological settings. They can also have more extreme geochemical
conditions and soil properties, including very low organic content and potentially high concentrations of
metals. These factors would need to be modeled with site-specific information to understand how they
are likely to impact the fate and transport of PFOA and PFOS at the site.
3.1.7 Incinera tion
Current SSIs may not operate at temperatures that are sufficient to completely destroy PFAS
compounds to mineralized compounds (C02, HF, F2). Therefore, incineration could result in PFOA or
PFOS emissions via either incomplete combustion of those chemicals in the sewage sludge or if other
PFAS are only partially destroyed and create PFOA and PFOS or their precursors. Given that SSIs can
destroy some proportion of PFOA and PFOS (Winchell et al., 2024), deposition of PFOA and PFOS from
an SSI to nearby soils would lead to lower exposures than the land application of equivalently
contaminated sewage sludge. However, past sewage sludge assessments (US EPA, 1992) have separately
assessed incineration. This assessment is not attempting to create an incineration exposure estimate
given the active research and investigation of PFAS destruction efficiency during incineration and
potential exposure to PICs.
3.2 Modeled Media Concentrations over Time
The fate and transport models used in this assessment calculate estimates of media concentrations over
time with daily resolution. This allows for understanding how PFOA and PFOS might be transported
throughout the modeled environment over time. For illustrative purposes, the following sections
describe the changes in PFOA and PFOS concentrations over time in the crop or pasture farm model run
at the "moderate" climate setting.
3.2.1 Soil Concentra tions over Time
The current modeling effort does not take into account the effects of PFOA and PFOS precursor
transforming to PFOA and PFOS in soil over time. Studies of biosolids land-application sites with PFAS
contamination indicate that the transformation of precursors acts as a long-term source of PFOA and
PFOS, well after land application has ceased (Washington et al., 2010; Yoo et al. 2010). That said,
modeled soil concentrations over time are still valuable in understanding how soil concentration change
as material is added, eroded, taken up into plants and animals, and leached to groundwater.
There are differences in the modeled concentration trends over time for the low and high Koc settings.
The low Koc setting at the "moderate" climate crop farm is depicted in Figure 9. In this setting, the PFOA
is quickly mobilized from the soil, such that levels do not build up with annual additions of biosolids.
Despite these low-sorption soil conditions, these models still indicate that PFOS will persist long enough
in soils to accumulate over time during the timeframe of biosolids application. However, PFOS
concentrations in the topsoil drop quickly after land-application end, and PFOS concentrations averaged
across the soil profile also have a steady declining trend.
DRAFT
95
-------
As seen in Figure 10 (the high Koc setting), the soil concentrations increase over time as annual biosolids
land applications occur. When applications of biosolids stop after 40 years, the concentrations of PFOA
and PFOS in soil begin to decline. The rates of decline are significantly faster for the top layer of soil,
while the average soil concentration in the top 20 centimeters declines more slowly. The rate of decline
is faster for PFOA than PFOS because PFOA is more mobile and thus more leachable from soils. The
variability in concentrations over time reflects ongoing natural mixing of the soil and changes in weather
over time. In this high Koc setting, PFOS concentrations remain elevated throughout the model run
duration (150 years).
In the high Koc scenario for PFOS, the soil concentrations decline along an expected trend line until year
80, when they dip dramatically and appear to rebound. This trend is a known artifact of the numerical
modeling used in 3MRA's Land Application Unit module and does not impact the risk calculations in this
assessment. In short, the numerical formulation of the LAU's Generic Soil Column Model (GSCM; US
EPA, 1999) solves the three components of the governing transport equation—diffusive transport,
advective transport, and contaminant decay—in a layered soil column. The advective process moves
mass downward through each layer of the soil column with an effective convection velocity corrected
for contaminant partitioning to the water and solid phases; this effective convection velocity is heavily
influenced by Koc. The advective component of the transport equation moves contaminant mass down
to the next layer (and ultimately, out the bottom of the LAU) at discrete time intervals equal to the time
it takes for dissolved contaminants to traverse a layer via convective transport. At large Koc (e.g., the 90th
percentile Koc for PFOS), the contaminant's effective velocity is very slow and the amount of mass sorbed
to soil is much greater, resulting a relatively large amount of sorbed mass leaving the system at once at
discrete time intervals and resulting in the sharp drops at predictable intervals visible in the media
concentration charts for PFOS with high Koc- The magnitude and frequency of these oscillations are
directly related to the magnitude of the Koc: this numerical artifact is always present, but with smaller
Koc, the oscillation is much smaller and more frequent and so not distinguishable from numerical noise.
Regardless of the size of the oscillations, they do not affect the risk results, as those are based on the 1-
year average concentrations at the peak (for noncancer) or averaged over the 10-year period that is
centered on the peak (for cancer). The peak soil concentration is always close to year 40 in the pasture
farm and crop farm scenarios, when biosolids stop being added to the field and before the oscillatory
behavior becomes apparent.
DRAFT
96
-------
Soil Concentration, Crop Scenario, Moderate Climate, Low Koc
20
PFOA Surface
40
60
¦PFOA Depth Avg
80
Year
100
¦PFOS Surface
120 140
¦PFOS Depth Avg
160
Figure 9. Plot of PFOA and PFOS concentrations over time in the "moderate" climate crop farm
scenario with the low Koc setting, assuming biosolids application ceases after 40 years.
8.E+05
7.E+05
5.E+05
-------
3.2.2 Surface Water Concentrations over Time
The trends in modeled surface water concentrations also change over time depending on if the farm is
modeled using the low or high Koc setting. In the low Koc setting (Figure 11), PFOA and PFOS
concentrations in surface water steadily increase over time up until land application stops after 40 years.
PFOA is more leachable into the aqueous phase and has a larger degree of transport to surface water in
the dissolved phase; for this reason, concentrations of PFOA in surface water are more responsive to
changes in precipitation. After land application ends, PFOA and PFOS concentrations decrease in surface
water rapidly over the next 20 to 40 years, and then more slowly from model years 80 to 150.
The PFOA and PFOS surface water trends are different in the high Koc setting (Figure 12), where PFOS
concentrations rise and fall slowly in surface water compared to PFOA concentrations. This trend likely
reflects the fact that the high sorption scenario for PFOS results in more retention in the soil column and
less mobilization into surface water.
Figure 11. PFOA and PFOS concentrations over time in the low Koc, pasture farm, moderate climate
setting.
DRAFT
98
-------
1800
Figure 12. PFOA and PFOS concentrations over time in the high Koc, pasture farm, moderate climate
setting.
3.2.3 Groundwater Concentrations over Time
Both the low Koc and high Koc settings for PFOA and PFOS show that it takes a considerable amount of
time for these chemicals to move from the soil (where they are applied in biosolids) into the vadose
zone, and through to groundwater. In the low Koc setting (Figure 13), models indicate it takes between
10 and 30 years for PFOA and 500 and 1000 years for PFOS to reach a hypothetical well five meters
away from the field. The well depth was selected after reviewing the concentration profile in
groundwater at depth increments of 0.5 meters to 2.0 meters below the water table and selecting this
highest concentration depth for this distance from the field. Assessment of the concentration of PFOA
and PFOS with depth indicated that the concentration is relatively constant down to 6 to 8 meters below
the water table, so the choice of selecting the maximum value in the top 2.0 meters of the aquifer does
not significantly impact the assessment (see Appendix C).
In the high Koc setting (Figure 14), the models indicate that it takes between 300 to 400 years for PFOA
and 6,000 to 8,000 years for PFOS to reach that hypothetical well. Empirical observations of
groundwater concentrations in monitoring wells and drinking water wells near biosolids land-application
sites indicate that these modeled timeframes for higher Koc settings are likely incorrect (too long) by
orders of magnitude (see Section 5.3 and Appendix C for more details).
The leaching potential for PFOA and PFOS at any specific site can be highly variable due to a variety of
factors, many of which are not captured in this draft risk assessment. For example, a recent study
investigated the effects of microbial weathering on PFAS partitioning over time after biosolids land
application to examine the fate and transport of PFAS leaching from biosolids into the environment
(Lewis et al., 2023). Results revealed that microbial weathering plays a role in PFAS partitioning,
contributing to the biodegradation of organic matter and leading to an increased potential for PFAS
leaching to groundwater. The weathering of the biosolids matrix is not taken into account in this
assessment's groundwater models. Additionally, another study showed that the dry-wet and freeze-
thaw cycles that are a natural occurrence in subsurface soils can lead to increased PFOA leaching
DRAFT
99
-------
(Borthakur et al., 2021). It is not entirely clear how the freeze-thaw cycles result in more leaching, and
there are no models available that incorporate this effect. A third recent study examined how colloidal
transport mechanisms may facilitate faster rates of PFAS leaching (Bierbaum et al., 2023). In general, the
existence of preferential flow pathways in soils, sometimes called "macropores," may also facilitate
faster leaching than is modeled in this assessment. Colloidal transport mechanisms and preferential flow
pathways like cracks, soil type boundaries, or worm and insect tunnels are not accounted for in the
groundwater model used in this assessment. EPA will continue evaluating the availability of
groundwater and vadose zone models as this assessment is finalized.
Figure 13. PFOA and PFOS concentrations over time in the low Koc, pasture farm, moderate climate
setting.
DRAFT
100
-------
Groundwater Concentration, Pasture Scenario, Moderate Climate, High Koc
PFOS Observation Times [years]
0 5,000 10,000 15,000 20,000 25,000 30,000 35,000
3.0E+02 . 1 t 1 1 1 3.5E+00
2.5E+02
2.0E+02
1.5E+02
5
"O
2 1.0E+02
o 5.0E+01
in
—
b
<
£ O.OE+OO
3.0E+00
2.5E+00
2.0E+00 u
1.5E+00
1.0E+00 uj
5.0E-01
0.0E+00
¦PFOA
¦PFOS
PFOA Observation Times [years]
Figure 14. PFOA and PFOS concentrations over time in the high Koc, pasture farm, moderate climate
setting.
4 RISK CHARACTERIZATION
The following sections integrate modeled media concentration results with human intake rates for each
media and human health effects thresholds to describe risks to receptors. Risks are discussed for each of
the biosolids use or disposal scenarios outlined in Section 2.8. As described in Section 5.3, the exposure
results from fate and transport modeling are sensitive to the parameters associated with the climate
and Koc. For this reason, risk results are presented in the same manner as media concentrations, with
results presented for each climate (dry, moderate, and wet) and for a low Koc and high Koc. Risks are
further disaggregated into hazard quotients (HQs) for non-cancer effects and cancer risk levels (CRLs).
4.1 Methods for Estimating Human Health Hazard and Cancer Risk
Cancer risk is characterized by calculating the lifetime excess cancer risk for the target population, which
is the increased probability that a member of that population will develop cancer over a lifetime
because of exposure to the pollutant. To evaluate oral exposures to carcinogens, the LADD is used. The
LADD is calculated by finding the modeling year with the highest average daily dose for the given media
(i.e., groundwater, surface water, soil), and calculating the average daily dose for the ten years around
the maximum concentration modeling year. The result is a lifetime average daily dose that spans a 10-
year residency on the site and is centered around the year associated with the highest dose for
groundwater, surface water or soil. The models run for 150 years and assume that a lifetime only
includes 10 years on the contaminated site (see section 2.9.3.12 for discussion of the duration of
exposure modeling), with the remainder of the 60 years taking place in a location with zero PFOA and
PFOS exposure. For example, if the highest concentration of PFOA or PFOS in groundwater used as
drinking water does not occur until forty years after biosolids application on a farm field begins, the
lifetime cancer risk is calculated by averaging the daily dose of exposure from drinking water spanning
from model year 35 to model year 44. That daily dose average is then scaled down to 350 days per year
(to account for travel time away from the residence) and normalized over a 70-year lifetime to calculate
DRAFT
101
-------
the LADD. This LADD is multiplied by the CSF to calculate the excess lifetime cancer risk from using
impacted groundwater as drinking water. This approach is used to calculate LADDs for children and
adults, and assumes that there are no exposures to PFOA or PFOS from drinking water when the
resident is not living at the contaminated residence or when the resident is traveling away from the
home during their period of residence. These assumptions about residency time and off-site exposure
could result in an underestimation of risk.
The EPA does not have a single cancer risk level that is used for risk assessments, but generally targets
cancer risk levels of 1 in 100,000 (1 x 10"5) or 1 in 1 million (1 x 10 s) for carcinogens, depending on the
size of the impacted population (e.g., US EPA, 2000a). Given that this central tendency modeling
exercise is parameterized with median values and is modeling risks for PFOA and PFOS near the
detection limit for biosolids (1 ppb), the EPA anticipates that these model scenarios may be applicable
across many biosolids use and disposal sites in the U.S.. Further, because the starting concentration of
PFOA and PFOS are linearly related to the modeled media concentrations, a scenario modeled to exceed
a 1-in-l-million cancer risk level in this draft risk assessment would exceed a l-in-100,000 cancer risk
level if the starting concentration for PFOA or PFOS were 10 ppb. Monitoring of sewage sludge in states
like Michigan indicate that biosolids with either PFOA or PFOS exceeding 10 ppb are common (see
Section 2.4). Therefore, this draft risk assessment will highlight excess cancer risks exceeding 1-in-l-
million (1 x 10 s).
Noncancer hazard is characterized by calculating an HQ based on the maximum one-year ADD for
ingestion exposures and the RfD. The ADD is used instead of the LADD for non-carcinogenic endpoints
because at least one of the co-critical effects identified for PFOA and PFOS is a developmental endpoint
and can potentially result from a short-term exposure during critical periods of development. Unlike
cancer risk estimates, HQs are risk indicators rather than risk estimates; the RfD represents a daily
exposure that is likely to be without appreciable risk of deleterious noncancer effects during a lifetime.
An HQ of 1 is used to establish a threshold of concern for a specific health effect. An HQ greater than 1
indicates risk (US EPA, 1986; 2000a; 2024e).
Equation 4. Human Cancer Risk(unitless)
Ingestion Exposures
ADD x ED x EF
LADD = ——7r-r^-.—t) Risknral = LADD x CSFnral
AT x 365 day/y mal mal
Name
Description
Value
ADD
Average daily dose (mg/kg-day)
Calculated
ED
Exposure duration (yr)
10 years
EF
Exposure frequency (day/yr)
350 days/year
AT
Averaging time (yr)
70 years
CSF oral
Oral cancer slope factor (mg/kg-day)"1
29,300 (mg/kg/day)"1for PFOA; 39.5
(mg/kg/day)-1 for PFOS
Equation 5. Human Hazard Quotient, HQ (unitless)
Ingestion Exposures
ADD
HQ oral ~ RfD
Name
Description
Value
ADD
Average daily dose (mg/kg-day)
Calculated
RfD
Noncancer reference dose (mg/kg-day)
3 x 10"8 mg/kg/day for PFOA; 1 x 10 7
mg/kg/day for PFOS
DRAFT
102
-------
The following tables include CRLs and HQs from exposure to various media for PFOA and PFOS under
each conceptual modeling scenario. All highlighted cells represent settings with risks or hazards
exceeding the acceptable threshold. Risks are presented individually per pathway; a given receptor may
have exposure from multiple pathways at one time. A given receptor may also have exposure to PFOA
and PFOS at the same time. The presented risks and hazard quotients only represent risks contributed
by contaminated biosolids use, not total risks to the receptor from that pathway, which may be larger.
4.2 Crop Farm Risk Estimation
The following table includes cancer risk levels and hazard quotients for receptors in the crop farm
scenario, disaggregated by pathway.
Table 34. PFOA Risk Results for Crop Farm, Cancer and Non-Cancer
Pathway
Low
Koc
High Koc
Adult
Child
Adult
Child
CRL | HQ
CRL* | HQ
CRL | HQ
CRL* | HQ
Dry Climate
Exposed fruit
4.5E-06
0.039
4.6E-06
0.04
1.0E-05
0.09
1.0E-05
0.092
Exposed vegetable
3.7E-05
0.32
2.6E-05
0.23
8.5E-05
0.75
6.1E-05
0.53
Fish
4.9E-04
4.2
5.8E-04
4.9
9.1E-05
0.76
1.1E-04
0.89
Groundwater
2.1E-05
0.2
2.2E-05
0.21
1.1E-13
9.2E-10
1.2E-13
9.6E-10
Protected fruit
6.3E-06
0.054
6.9E-06
0.059
1.4E-05
0.13
1.6E-05
0.14
Protected vegetable
3.0E-05
0.26
5.5E-05
0.48
6.9E-05
0.61
1.3E-04
1.1
Root vegetable
2.2E-05
0.19
1.4E-05
0.12
4.9E-05
0.43
3.3E-05
0.29
Soil
1.7E-08
0.00018
2.6E-07
0.0027
4.6E-08
0.00042
7.1E-07
0.0064
Surface water
4.2E-04
3.6
4.4E-04
3.7
7.7E-05
0.65
8.1E-05
0.68
Moderate Climate
Exposed fruit
2.6E-07
0.0033
2.7E-07
0.0034
4.2E-06
0.037
4.3E-06
0.038
Exposed vegetable
2.2E-06
0.027
1.6E-06
0.019
3.5E-05
0.31
2.5E-05
0.22
Fish
2.7E-05
0.24
3.1E-05
0.28
9.1E-05
0.79
1.1E-04
0.93
Groundwater
2.4E-04
2.5
2.5E-04
2.6
6.6E-06
0.055
6.9E-06
0.057
Protected fruit
3.7E-07
0.0046
4.0E-07
0.005
5.9E-06
0.052
6.5E-06
0.057
Protected vegetable
1.8E-06
0.022
3.3E-06
0.041
2.8E-05
0.25
5.2E-05
0.46
Root vegetable
1.3E-06
0.016
8.5E-07
0.011
2.0E-05
0.18
1.4E-05
0.12
Soil
2.8E-10
7.4E-06
4.3E-09
0.00011
1.5E-08
0.00016
2.2E-07
0.0025
Surface water
2.3E-05
0.2
2.4E-05
0.21
7.7E-05
0.67
8.1E-05
0.7
Wet Climate
Exposed fruit
2.4E-07
0.0038
2.4E-07
0.0039
3.2E-06
0.028
3.3E-06
0.028
Exposed vegetable
2.0E-06
0.031
1.4E-06
0.022
2.7E-05
0.23
1.9E-05
0.16
Fish
1.8E-05
0.19
2.1E-05
0.22
6.4E-05
0.57
7.5E-05
0.66
Groundwater
2.4E-04
2
2.5E-04
2.1
2.6E-05
0.21
2.7E-05
0.22
Protected fruit
3.3E-07
0.0053
3.7E-07
0.0058
4.5E-06
0.039
4.9E-06
0.043
Protected vegetable
1.6E-06
0.026
3.0E-06
0.047
2.2E-05
0.19
4.0E-05
0.34
Root vegetable
1.1E-06
0.018
7.7E-07
0.012
1.6E-05
0.13
1.0E-05
0.09
Soil
2.6E-10
6.7E-06
4.0E-09
0.0001
1.0E-08
0.00011
1.6E-07
0.0017
Surface water
1.5E-05
0.16
1.6E-05
0.17
5.4E-05
0.48
5.7E-05
0.5
*CRLs for children represent lifetime cancer risks stemming from 10 years of exposure during childhood. These results do not describe risks of
childhood cancer.
Table 35. PFOS Risk Results for Crop Farm, Cancer and Non-Cancer
Pathway
Low
Koc
Hiqh Koc
Adult
Child
Adult
Child
CRL | HQ
CRL* | HQ
CRL | HQ
CRL* | HQ
Dry Climate
Exposed fruit
2.3E-09
0.0044
2.4E-09
0.0045
4.0E-09
0.0076
4.1E-09
0.0077
Exposed vegetable
6.1E-09
0.012
4.4E-09
0.0083
1.0E-08
0.02
7.5E-09
0.014
DRAFT
103
-------
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
CRL
HQ
CRL*
HQ
CRL
HQ
CRL*
HQ
Fish
1.1E-05
21
1.3E-05
25
1.0E-07
0.19
1.2E-07
0.22
Groundwater
3.9E-09
0.0078
4.1E-09
0.0082
2.9E-38
5.4E-32
3.1E-38
5.7E-32
Protected fruit
3.2E-09
0.0061
3.6E-09
0.0067
5.6E-09
0.011
6.1E-09
0.012
Protected vegetable
5.0E-09
0.0094
9.1E-09
0.017
8.5E-09
0.016
1.6E-08
0.030
Root vegetable
5.3E-08
0.1
3.6E-08
0.067
9.1E-08
0.17
6.1E-08
0.12
Soil
4.0E-11
8.8E-05
6.2E-10
0.0013
8.2E-11
0.00016
1.2E-09
0.0024
Surface water
3.3E-07
0.62
3.5E-07
0.65
2.8E-09
0.0053
3.0E-09
0.0055
Moderate Climate
Exposed fruit
5.3E-10
0.0011
5.5E-10
0.0011
3.5E-09
0.0069
3.6E-09
0.007
Exposed vegetable
1.4E-09
0.0028
1.0E-09
0.002
9.3E-09
0.018
6.7E-09
0.013
Fish
4.3E-06
8.3
5.0E-06
9.7
1.3E-07
0.25
1.5E-07
0.29
Groundwater
7.1E-08
0.13
7.4E-08
0.14
1.3E-12
2.3E-06
1.3E-12
2.5E-06
Protected fruit
7.5E-10
0.0015
8.2E-10
0.0017
5.0E-09
0.0096
5.4E-09
0.011
Protected vegetable
1.1E-09
0.0023
2.1E-09
0.0042
7.6E-09
0.015
1.4E-08
0.027
Root vegetable
1.2E-08
0.025
8.2E-09
0.016
8.1E-08
0.16
5.4E-08
0.11
Soil
5.7E-12
1.8E-05
8.7E-11
0.00027
5.6E-11
0.00014
8.5E-10
0.0021
Surface water
1.3E-07
0.24
1.3E-07
0.25
3.5E-09
0.0068
3.6E-09
0.0071
Wet Climate
Exposed fruit
3.9E-10
0.00081
3.9E-10
0.00083
3.2E-09
0.0062
3.3E-09
0.0063
Exposed vegetable
1.0E-09
0.0021
7.3E-10
0.0015
8.5E-09
0.016
6.0E-09
0.012
Fish
2.7E-06
5.2
3.1E-06
6.1
1.4E-07
0.27
1.6E-07
0.31
Groundwater
1.9E-07
0.36
2.0E-07
0.37
7.5E-10
0.002
7.9E-10
0.0021
Protected fruit
5.4E-10
0.0011
5.9E-10
0.0012
4.5E-09
0.0086
4.9E-09
0.0095
Protected vegetable
8.3E-10
0.0017
1.5E-09
0.0032
6.9E-09
0.013
1.3E-08
0.024
Root vegetable
8.8E-09
0.019
5.9E-09
0.012
7.4E-08
0.14
4.9E-08
0.095
Soil
3.3E-12
1.1E-05
5.1E-11
0.00017
5.6E-11
0.00012
8.6E-10
0.0019
Surface water
7.8E-08
0.15
8.2E-08
0.16
3.8E-09
0.0074
4.0E-09
0.0077
*CRLs for children represent lifetime cancer risks stemming from 10 years of exposure during childhood. These results do not describe risks of
childhood cancer.
All highlighted cells represent hazard or cancer risk above acceptable thresholds for the crop farm
scenario. Overall, PFOA risks are higher than those posed by PFOS and occur under more settings (low
and high Koc; dry, moderate, and wet climate conditions). The pathway with the highest risk for PFOS is
fish consumption (CRL up to 1.3 x 10"5 and HQ up to 25). The pathways with the highest risks for PFOA
are groundwater used as drinking water, surface water used as drinking water, and fish consumption,
which have maximum cancer risks from 2.5 x 10"4 to 5.0 x 10"4 and maximum hazard quotients from 2.6
to 4.9. Every setting (dry, moderate, wet, low Koc, high Koc) results in at least one exceedance of cancer
or hazard thresholds for every pathway. In the model, crop exposures result in cancer risk for PFOA, but
these risks are based on greenhouse studies of pots in plants that likely over-estimate plant uptake and
the estimates for plant uptake. See section 5.2 for more discussion of the uncertainties with uptake of
PFOA and PFOS into fruits and vegetables
Soil concentrations remain below risk thresholds in all scenarios for PFOA and PFOS, but some scenarios
are within a factor of 10 of the risk threshold. Notably, the only pathway exceeding risk thresholds for
PFOS is fish consumption, and only when Koc is low. This indicates that if soil sorption conditions are high
for PFOS and only PFOS is present at low concentrations in biosolids, the material could be land applied
for growing crops for human consumption without meaningfully increasing risk in any pathway. If land
application occurs with a larger than 10-meter buffer from the closest fishable waterbody, this could
mitigate risks posed by PFOS.
DRAFT
104
-------
4.3 Pasture Farm Risk Estimation
The following table includes cancer risk levels and hazard quotients for receptors in the pasture farm
scenario, disaggregated by pathway.
Table 36. PFOA Risk Results for Pasture Farm, Cancer and Non-Cancer
Pathway
Low
Koc
High Koc
Adult
Child
Adult
Child
CRL | HQ
CRL* | HQ
CRL | HQ
CRL* | HQ
Dry Climate
Beef
3.4E-05
0.41
4.4E-05
0.54
2.0E-04
1.7
2.6E-04
2.2
Eggs
7.7E-05
0.96
7.7E-05
0.96
6.3E-04
5.5
6.3E-04
5.5
Fish
6.3E-04
5.4
7.4E-04
6.3
2.7E-04
2.2
3.1E-04
2.6
Groundwater
1.5E-04
1.3
1.6E-04
1.3
1.4E-06
0.012
1.5E-06
0.012
Milk
4.1E-04
5
7.4E-04
9.1
2.1E-03
18
3.9E-03
34
Poultry
2.8E-06
0.035
5.1E-06
0.064
2.3E-05
0.2
4.2E-05
0.36
Soil
3.0E-08
0.00043
4.6E-07
0.0065
3.8E-07
0.0033
5.8E-06
0.05
Surface water
5.4E-04
4.6
5.6E-04
4.8
2.3E-04
1.9
2.4E-04
2
Moderate Climate
Beef
6.5E-06
0.07
8.6E-06
0.092
2.7E-05
0.3
3.6E-05
0.4
Eggs
1.2E-05
0.15
1.2E-05
0.15
8.5E-05
0.97
8.5E-05
0.97
Fish
1.1E-04
1
1.3E-04
1.2
9.3E-05
0.81
1.1E-04
0.95
Groundwater
2.3E-04
1.9
2.4E-04
2
1.4E-05
0.12
1.5E-05
0.12
Milk
8.2E-05
0.86
1.5E-04
1.6
3.0E-04
3.3
5.5E-04
6.1
Poultry
4.3E-07
0.0055
7.9E-07
0.01
3.1E-06
0.035
5.7E-06
0.064
Soil
2.4E-09
5.1E-05
3.7E-08
0.00078
5.0E-08
0.00058
7.7E-07
0.0089
Surface water
9.6E-05
0.85
1.0E-04
0.89
7.9E-05
0.68
8.3E-05
0.71
Wet Climate
Beef
4.5E-06
0.047
5.9E-06
0.061
1.9E-05
0.22
2.5E-05
0.29
Eggs
8.1E-06
0.098
8.1E-06
0.098
5.7E-05
0.69
5.7E-05
0.69
Fish
6.9E-05
0.61
8.1E-05
0.72
5.4E-05
0.46
6.4E-05
0.54
Groundwater
1.4E-04
1.2
1.4E-04
1.2
4.2E-05
0.35
4.4E-05
0.36
Milk
5.7E-05
0.57
1.0E-04
1
2.1E-04
2.5
3.8E-04
4.5
Poultry
2.9E-07
0.0036
5.4E-07
0.0065
2.1E-06
0.025
3.8E-06
0.046
Soil
1.5E-09
3.2E-05
2.3E-08
0.00049
3.3E-08
0.0004
5.0E-07
0.0061
Surface water
5.9E-05
0.52
6.1E-05
0.55
4.6E-05
0.39
4.8E-05
0.41
*CRLs for children represent lifetime cancer risks stemming from 10 years of exposure during childhood. These results do not describe risks of
childhood cancer.
Table 37. PFOS Risk Results for Pasture Farm, Cancer and Non-Cancer
Pathway
Low
Koc
High Koc
Adult
Child
Adult
Child
CRL | HQ
CRL* | HQ
CRL | HQ
CRL* | HQ
Dry Climate
Beef
1.0E-06
2.3
1.4E-06
3
2.4E-06
4.7
3.2E-06
6.1
Eggs
6.2E-07
1.4
6.2E-07
1.4
2.1E-06
4
2.1E-06
4
Fish
2.1E-05
39
2.4E-05
45
6.2E-07
1.1
7.3E-07
1.3
Groundwater
1.6E-08
0.03
1.7E-08
0.031
4.4E-38
8.2E-32
4.6E-38
8.6E-32
Milk
1.3E-06
2.9
2.3E-06
5.3
2.6E-06
5
4.7E-06
9.1
Poultry
1.0E-07
0.23
1.9E-07
0.41
3.4E-07
0.66
6.2E-07
1.2
Soil
1.9E-10
0.00043
2.9E-09
0.0066
7.4E-10
0.0014
1.1E-08
0.022
Surface water
6.0E-07
1.1
6.3E-07
1.2
1.7E-08
0.032
1.8E-08
0.033
Moderate Climate
Beef
2.5E-07
0.52
3.3E-07
0.69
1.5E-06
2.9
2.0E-06
3.8
Eggs
9.3E-08
0.23
9.3E-08
0.23
1.3E-06
2.5
1.3E-06
2.5
Fish
6.0E-06
12
7.0E-06
14
5.8E-07
1.1
6.8E-07
1.3
Groundwater
8.2E-08
0.15
8.6E-08
0.16
4.9E-11
9.1E-05
5.1E-11
9.5E-05
DRAFT
105
-------
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
CRL
HQ
CRL*
HQ
CRL
HQ
CRL*
HQ
Milk
3.3E-07
0.69
6.1E-07
1.3
1.6E-06
3.1
3.0E-06
5.7
Poultry
1.5E-08
0.038
2.8E-08
0.068
2.1E-07
0.41
3.8E-07
0.74
Soil
2.0E-11
5.7E-05
3.0E-10
0.00087
4.5E-10
0.00089
6.9E-09
0.013
Surface water
1.7E-07
0.34
1.8E-07
0.35
1.6E-08
0.03
1.7E-08
0.032
Wet Climate
Beef
9.7E-08
0.34
1.3E-07
0.44
9.8E-07
1.9
1.3E-06
2.5
Eggs
4.9E-08
0.15
4.9E-08
0.15
8.2E-07
1.6
8.2E-07
1.6
Fish
3.4E-06
6.6
4.0E-06
7.7
4.2E-07
0.81
4.9E-07
0.95
Groundwater
1.4E-07
0.26
1.5E-07
0.27
8.5E-10
0.0016
8.9E-10
0.0016
Milk
1.2E-07
0.44
2.2E-07
0.8
1.0E-06
2
1.9E-06
3.7
Poultry
8.1E-09
0.025
1.5E-08
0.045
1.4E-07
0.26
2.5E-07
0.48
Soil
1.4E-11
4.3E-05
2.1E-10
0.00065
2.9E-10
0.00057
4.5E-09
0.0086
Surface water
1.0E-07
0.19
1.0E-07
0.2
1.2E-08
0.022
1.2E-08
0.023
*CRLs for children represent lifetime cancer risks stemming from 10 years of exposure during childhood. These results do not describe risks of
childhood cancer.
Modeling for the pasture farm scenario includes multiple pathways that exceed acceptable cancer risk
or hazard levels for PFOA and PFOS. As is seen in the results from the crop farm scenario, PFOA results in
more elevated risk pathways and pathways with higher risks than PFOS, due to PFOA's higher cancer
slope factor. The highest risk pathways for PFOA include milk, beef, egg, fish, and drinking water
consumption (either sourced from groundwater or surface water). The highest risk pathways for PFOS
include fish, milk, eggs, and beef.
For PFOA, all climate and Koc settings result in exceedances of acceptable risk and hazard thresholds for
levels in milk, with cancer risk levels ranging from 5.7 x 10~5 to 3.9 x 10"3 and hazard quotients ranging
from 5.7 to 34. The modeling suggests that even when modeled concentrations are below currently
available method detection limits (MDLs), estimated cancer risks associated with PFOA can exceed
acceptable thresholds. This indicates that there may be exceedances of acceptable risk thresholds due
to PFOA levels in milk from farms with biosolids land application that fall below detectable limits. In the
model, PFOS also exceeds risk thresholds in milk in most settings.
4.4 Reclamation Risk Estimation
The following table includes cancer risk levels and hazard quotients for receptors in the land reclamation
scenario, disaggregated by pathway.
Table 38. PFOA Risk Results for Reclamation Site, Cancer and Non-Cancer
Pathway
Low
Koc
Hiqh Koc
Adult
Child
Adult
Child
CRL | HQ
CRL* | HQ
CRL | HQ
CRL* | HQ
Dry Climate
Beef
2.5E-05
0.39
3.2E-05
0.51
4.5E-05
0.41
5.9E-05
0.54
Eggs
7.4E-05
1.3
7.4E-05
1.3
1.5E-04
1.3
1.5E-04
1.3
Fish
1.1E-04
0.9
1.3E-04
1.1
3.5E-05
0.29
4.0E-05
0.34
Groundwater
9.4E-06
0.078
9.8E-06
0.081
1.6E-07
0.0013
1.7E-07
0.0014
Milk
2.7E-04
4.2
5.0E-04
7.7
4.9E-04
4.4
9.0E-04
8
Poultry
2.7E-06
0.048
4.9E-06
0.087
5.4E-06
0.049
9.8E-06
0.09
Soil
4.2E-08
0.00083
6.4E-07
0.013
9.0E-08
0.00084
1.4E-06
0.013
Surface water
9.1E-05
0.77
9.5E-05
0.8
2.9E-05
0.25
3.1E-05
0.26
Moderate Climate
Beef
5.1E-06
0.25
6.7E-06
0.33
3.7E-05
0.4
4.8E-05
0.52
Eggs
1.6E-05
0.83
1.6E-05
0.83
1.1E-04
1.3
1.1E-04
1.3
Fish
1.2E-05
0.14
1.4E-05
0.16
2.7E-05
0.24
3.1E-05
0.28
DRAFT
106
-------
Pathway
Low Koc
High Koc
Adult
Child
Adult
Child
CRL
HQ
CRL*
HQ
CRL
HQ
CRL*
HQ
Groundwater
2.9E-06
0.19
3.1E-06
0.2
1.2E-06
0.01
1.3E-06
0.011
Milk
5.7E-05
2.7
1.0E-04
5
4.0E-04
4.3
7.3E-04
7.9
Poultry
5.7E-07
0.03
1.0E-06
0.055
4.1E-06
0.048
7.5E-06
0.086
Soil
8.9E-09
0.00051
1.4E-07
0.0078
6.5E-08
0.0008
9.9E-07
0.012
Surface water
1.0E-05
0.12
1.1E-05
0.12
2.3E-05
0.2
2.4E-05
0.21
Wet Climate
Beef
2.2E-06
0.11
2.9E-06
0.14
2.9E-05
0.36
3.9E-05
0.47
Eggs
5.6E-06
0.34
5.6E-06
0.34
8.7E-05
1.1
8.7E-05
1.1
Fish
1.6E-05
0.23
1.8E-05
0.27
2.7E-05
0.25
3.1E-05
0.3
Groundwater
1.3E-05
1.1
1.4E-05
1.1
1.7E-06
0.014
1.8E-06
0.015
Milk
2.6E-05
1.2
4.8E-05
2.2
3.2E-04
3.9
5.9E-04
7.1
Poultry
2.0E-07
0.012
3.7E-07
0.023
3.2E-06
0.042
5.8E-06
0.076
Soil
2.6E-09
0.0002
3.9E-08
0.0031
4.9E-08
0.00068
7.4E-07
0.01
Surface water
1.3E-05
0.2
1.4E-05
0.21
2.3E-05
0.22
2.4E-05
0.23
*CRLs for children represent lifetime cancer risks stemming from 10 years of exposure during childhood. These results do not describe risks of
childhood cancer.
Table 39. PFOS Risk Results for Reclamation Site, Cancer and Non-Cancer
Pathway
Low
Koc
High Koc
Adult
Child
Adult
Child
CRL | HQ
CRL* | HQ
CRL | HQ
CRL* | HQ
Dry Climate
Beef
3.9E-07
0.82
5.1E-07
1.1
4.1E-07
0.82
5.4E-07
1.1
Eggs
3.1E-07
0.7
3.1E-07
0.7
3.5E-07
0.7
3.5E-07
0.7
Fish
3.1E-06
5.9
3.7E-06
6.9
8.3E-08
0.16
9.7E-08
0.18
Groundwater
2.3E-09
0.0042
2.4E-09
0.0044
1.5E-39
2.8E-33
1.6E-39
2.9E-33
Milk
4.3E-07
0.87
7.8E-07
1.6
4.4E-07
0.87
8.1E-07
1.6
Poultry
5.2E-08
0.11
9.4E-08
0.21
5.7E-08
0.12
1.0E-07
0.21
Soil
1.1E-10
0.00025
1.7E-09
0.0038
1.2E-10
0.00025
1.9E-09
0.0038
Surface water
9.2E-08
0.17
9.6E-08
0.18
2.3E-09
0.0044
2.4E-09
0.0046
Moderate Climate
Beef
1.9E-07
0.76
2.5E-07
1
3.6E-07
0.81
4.8E-07
1.1
Eggs
1.3E-07
0.63
1.3E-07
0.63
2.9E-07
0.69
2.9E-07
0.69
Fish
1.2E-06
2.7
1.4E-06
3.2
7.7E-08
0.15
9.1E-08
0.17
Groundwater
9.6E-09
0.018
1.0E-08
0.019
8.0E-13
1.5E-06
8.4E-13
1.5E-06
Milk
2.2E-07
0.82
4.0E-07
1.5
3.9E-07
0.85
7.2E-07
1.6
Poultry
2.1E-08
0.1
3.8E-08
0.19
4.8E-08
0.11
8.7E-08
0.21
Soil
4.2E-11
0.00022
6.4E-10
0.0034
1.0E-10
0.00025
1.6E-09
0.0037
Surface water
3.6E-08
0.079
3.7E-08
0.082
2.1E-09
0.0041
2.2E-09
0.0043
Wet Climate
Beef
1.7E-07
0.65
2.2E-07
0.85
3.1E-07
0.74
4.1E-07
0.98
Eggs
9.9E-08
0.49
9.9E-08
0.49
2.3E-07
0.62
2.3E-07
0.62
Fish
1.7E-06
4
2.0E-06
4.7
7.2E-08
0.14
8.4E-08
0.17
Groundwater
8.3E-09
0.015
8.7E-09
0.016
2.2E-12
4.1E-06
2.3E-12
4.3E-06
Milk
2.0E-07
0.73
3.8E-07
1.3
3.4E-07
0.79
6.2E-07
1.5
Poultry
1.6E-08
0.081
3.0E-08
0.15
3.9E-08
0.1
7.0E-08
0.18
Soil
3.1E-11
0.00018
4.7E-10
0.0027
8.1E-11
0.00022
1.2E-09
0.0033
Surface water
4.9E-08
0.12
5.1E-08
0.12
2.0E-09
0.0039
2.1E-09
0.0041
*CRLs for children represent lifetime cancer risks stemming from 10 years of exposure during childhood. These results do not describe risks of
childhood cancer.
Modeling for the reclamation scenario includes multiple pathways that exceed acceptable cancer risk or
hazard levels for PFOA and PFOS. As is seen in the modeling results from the pasture farm scenario,
PFOA results in more elevated risk pathways and pathways with higher risks than PFOS owing to PFOA's
DRAFT
107
-------
higher cancer slope factor. The highest risk pathways for PFOA include milk, beef, egg, and drinking
water consumption (either sourced from groundwater or surface water). The highest risk pathways for
PFOS include fish and milk. Overall, risk levels in the reclamation scenario are lower than risks in the
pasture farm scenario due to the modeling being based on a one-time biosolids application, rather than
ongoing applications.
Though the reclamation scenario presents fewer risks than the pasture farm scenario, for PFOA, all
climate and Koc settings still result in exceedances of acceptable risk and hazard thresholds for levels in
milk, with cancer risk levels up to 9.0 x 10"4 and hazard quotients up to 8. If one assumes that the
remediation site does not include a grazing pasture for dairy cows, the modeling still suggests that there
are risks for PFOA in groundwater, surface water, and fish pathways and for PFOS in fish pathways.
Given the linear relationship between the loading of PFOA and PFOS to the field and the calculated risks,
the risk results for a scenario with a single application of sewage sludge at a rate of 10 DMT/ha (more
typical of a median farming scenario rather than a land reclamation scenario) would be 1/5 of the values
presented in tables 38 and 39. This indicates that there are scenarios and pathways that may exceed the
EPA's acceptable risk thresholds after a single application of 10 DMT/ha given the modeling conditions.
4.5 Potential Impacts beyond the Farm Family
The media concentrations modeled in the pasture and crop farm scenarios are relevant to many
potential receptors beyond the farm family. Because the modeling suggests that risk thresholds are
exceeded for individual exposure pathways, a person or population exposed through only one pathway
(like drinking water or milk consumption) could still have an increased risk of adverse health effects.
Potential impacts outside the farm family are described by pathway below:
Surface water and fish pathways. A land application site where PFOA and PFOS concentrations in
biosolids were higher than 1 ppb and further from the surface waterbody may have similar outcomes to
the modeled surface water and fish tissue concentrations. Thus, it is possible that a significant fraction
of biosolids land application sites could have elevated PFOA and PFOS concentrations in surface water
and fish tissue. These impacts could include drinking water concentrations that exceed acceptable risk
thresholds and significant exposure from eating fish. Populations with elevated fish consumption rates
could have higher exposures than the population modeled in this assessment (farmers).
Milk pathway. This assessment is focused on milk consumption by people living on dairy farms, who
have higher milk consumption rates than the general population (US EPA 2018b). General population
milk consumers are likely to consume milk blended from multiple farms with or without a history of
biosolids land application. In parts of the U.S. with active dairy farms, community members may
purchase milk and dairy products directly from local farms, either by participating in a CSA, frequenting
farm stands, or purchasing their milk and dairy from a farmers' market. Those regularly consuming
products from a farm contaminated with PFOA or PFOS would likely be at a greater risk than the general
population, which is mostly comprised of people consuming blended milk products from a diversity of
sources.
Groundwater pathway. Once PFOA and PFOS enter groundwater after leaching from soil, they will
migrate along with the path of groundwater movement. The size of a PFOA or PFOS groundwater plume
depends on the amount of the chemicals deposited on land, the rate of groundwater flow, and the time
that has passed since application of biosolids contaminated with PFOA and PFOS. Depending on site-
specific circumstances, there could be many neighboring families to a crop farm or pasture farm with
impacted groundwater wells. Additionally, should a farm field that previously accepted biosolids be
developed into housing later, there could be ongoing groundwater contamination, leading to drinking
water impacts.
DRAFT
108
-------
Beef and chicken pathways. This draft risk assessment models the concentration and risks associated
with eating meat from laying hens and, in the case of PFOA, lactating cows. These scenarios were
selected because for PFOA and PFOS, there is no data available on uptake into broiler hens, which are
more commonly raised for meat. Similarly, though there are muscle uptake data available for PFOS
accumulation into cattle used for beef production, the only PFOA data available for uptake into cow
muscle is from a study that included dairy cows. Many families and commercial farms cull (and consume
or sell for consumption) laying hens and dairy cows after they cease to produce sufficient quantities of
milk or eggs. However, most chicken and beef consumed in the U.S. is not from these types of animals;
most chicken is sourced from faster-growing broiler hens and most beef is sourced from cows like Black
Angus, Red Angus, and Herefords. These animals raised primarily for meat will have different uptake
factors for PFOA and PFOS and different dietary intakes than the laying hens and lactating dairy cows.
For this reason, there are uncertainties in PFOA and PFOS exposure for those in the general population
who do not have backyard hens and (for PFOA) dairy cows that they may slaughter for food.
Fruits and vegetables. As discussed previously, there are considerable uncertainties regarding the
concentration and risk calculations for fruit and vegetable pathways due to data limitations on the
uptake of PFOA and PFOS into these types of plants. However, there are many populations outside of
the farm family that could be impacted by contamination of fruits and vegetables. It is increasingly
popular for fruit and vegetable farms to develop CSA programs, where participants receive weekly
deliveries of produce from a single farm and use this produce as their primary fruit and vegetable
source. It is also not uncommon for families to frequent a single nearby farm stand or farmers market
stand as a primary source of produce, especially during the fall, summer, and spring seasons. Finally,
there are many home gardeners who, for a hobby or for economic reasons, grow a large portion of their
produce in their yard or at a community garden plot. Because these groups also primarily source their
produce from a single site, should there be PFOA and PFOS biosolids impacts, produce could be a
meaningful source of exposure.
4.6 Sewage Sludge Disposal Site Risk Estimation
The following table includes cancer risk levels and hazard quotients for drinking water receptors in the
surface disposal scenario.
Table 40. PFOA Groundwater Risk Results for Sludge Disposal Site, Cancer and Non-Cancer
Pathway
Low
Koc
High Koc
Adult
Child
Adult
Child
CRL | HQ
CRL* | HQ
CRL | HQ
CRL* | HQ
Dry Climate
No liner
1.3E-03
11
1.4E-03
12
4.0E-06
0.034
4.2E-06
0.036
Clay liner
1.1E-03
9.2
1.2E-03
9.6
2.7E-06
0.023
2.8E-06
0.024
Composite liner
7.3E-07
0.0061
7.6E-07
0.0063
8.5E-35
7.2E-31
8.9E-35
7.5E-31
Moderate Climate
No liner
4.8E-04
4
5.0E-04
4.2
1.3E-06
0.011
1.3E-06
0.011
Clay liner
3.1E-04
2.6
3.2E-04
2.7
8.4E-07
0.0071
8.7E-07
0.0074
Composite liner
5.9E-08
0.0005
6.2E-08
0.00052
8.0E-35
6.9E-31
8.4E-35
7.2E-31
Wet Climate
No liner
8.5E-04
7.5
8.9E-04
7.8
9.0E-06
0.076
9.4E-06
0.08
Clay liner
6.6E-04
5.6
6.9E-04
5.9
4.1E-06
0.035
4.3E-06
0.037
Composite liner
2.2E-07
0.0018
2.3E-07
0.0019
4.7E-17
4E-13
4.9E-17
4.2E-13
*CRLs for children represent lifetime cancer risks stemming from 10 years of exposure during childhood. These results do not describe risks of
childhood cancer.
DRAFT
109
-------
Table 41. PFOS Groundwater Risk Results for Sludge Disposal Site, Cancer and Non-Cancer
Pathway
Low
Koc
High Koc
Adult
Child
Adult
Child
CRL | HQ
CRL* | HQ
CRL | HQ
CRL* | HQ
Dry Climate
No liner
9.5E-08
0.18
1.0E-07
0.19
3.3E-11
6.4E-05
3.5E-11
6.7E-05
Clay liner
6.6E-08
0.12
6.9E-08
0.13
2.3E-11
4.4E-05
2.4E-11
4.6E-05
Composite liner
1.6E-13
3.1E-07
1.7E-13
3.2E-07
1.4E-39
2.7E-33
1.5E-39
2.9E-33
Moderate Climate
No liner
3.1E-08
0.059
3.3E-08
0.062
1.3E-11
2.5E-05
1.3E-11
2.6E-05
Clay liner
1.8E-08
0.034
1.9E-08
0.035
7.3E-12
1.4E-05
7.6E-12
1.5E-05
Composite liner
3.3E-21
6.2E-15
3.4E-21
6.5E-15
1.6E-39
3.1E-33
1.6E-39
3.2E-33
Wet Climate
No liner
1.6E-07
0.29
1.6E-07
0.31
1.6E-10
0.00031
1.6E-10
0.00032
Clay liner
8.4E-08
0.16
8.8E-08
0.16
6.7E-11
0.00013
7.0E-11
0.00014
Composite liner
9.0E-13
1.7E-06
9.4E-13
1.8E-06
2.3E-39
4.5E-33
2.4E-39
4.7E-33
*CRLs for children represent lifetime cancer risks stemming from 10 years of exposure during childhood. These results do not describe risks of
childhood cancer.
The surface disposal scenario models groundwater impacts with three types of liner options: no liner,
clay liner, and composite liner. The modeling runs suggest that for PFOA, cancer risk thresholds are
exceeded in all scenarios where the surface disposal site is unlined or lined with clay. For unlined surface
disposal sites, cancer risks for PFOA in groundwater range from 1.3 x 10 s to 1.4 x 10"3. The upper end of
these values represents risks three orders of magnitude higher than the acceptable threshold. Hazard
quotients in this setting for PFOA go up to 12 in dry climates for child receptors. Unlined surface disposal
sites and lagoons are common across the U.S., and thus groundwater around these sites is at high risk
for contamination. Risks are only slightly mitigated by using a clay liner, and but are significantly
mitigated by the use of a composite liner.
Unlike PFOA, PFOS appears to be less mobile in surface disposal sites and therefore poses lower risks.
PFOS also has less potent toxicity than PFOA, such that concentrations can be higher without exceeding
risk thresholds. None of the surface disposal lining options result in exceedances of PFOS risk thresholds
when the concentration of sludge is 1 ppb. Given that there is a linear relationship between the starting
concentration of PFOS in sludge and the groundwater concentration, it is anticipated that cancer risk
thresholds may be exceeded in some unlined scenarios around concentrations of 10 ppb for PFOS and
HQs may exceed 1 at concentrations around 4 ppb in some unlined scenarios and 5 ppb in some clay-
lined scenarios.
4.7 Other Land Application Risk Estimation
As described in Section 2.8, there are many biosolids land application scenarios that are not
quantitatively or qualitatively assessed in this document. Examples include land application of biosolids
or septage on turf fields, golf courses, tree farms, or natural forested areas. Based on the risk values for
pathways like groundwater, surface water, fish, and soil of the farming scenarios, it is possible that
application of biosolids or septage in these alternative land application scenarios could also lead to
exceedances of acceptable risk thresholds in these pathways. For PFOS, runoff from an 80-acre
application site to a 13-acre lake or reservoir could result in risk exceedances for fish and surface water
pathways - a typical 18-hole golf course requires 100 to 175 acres of land. For PFOA, applications of this
size could result in risk exceedances for groundwater, surface water, and fish pathways as well. This
indicates that ongoing use of biosolids to fertilize a golf course could present risks, especially if there are
DRAFT
110
-------
nearby water bodies used for fishing or drinking water or if there were downgradient residents using
groundwater as a source of drinking water.
When biosolids are used to fertilize forested lands, there may or may not be ongoing annual applications
of biosolids. However, for the remediation pathway models, one-time application of 1 ppb PFOA and
PFOS in biosolids still results in exceedances of groundwater, surface water, and fish risk thresholds in
most climate and sorption scenarios. This indicates that even one-time application of biosolids to a
forested site could present risks, depending on the concentrations of PFOA and PFOS in the biosolids,
the size of applied lands, the size of the nearby waterbody, and if there were any receptors nearby, such
as those eating fish from waterways, those hunting or gathering food, those using groundwater as a
source of drinking water, or those using surface water as a source of drinking water. Additionally, in
some parts of the U.S., forests are used for livestock grazing. That said, there are many site-specific
factors that could influence risk in forested settings, including the fate and transport behaviors in those
specific settings.
Use of biosolids in road construction projects could present risks, depending on how the biosolids were
used, the amount used, and the concentration of PFOA and PFOS in the materials. Application of
biosolids as a groundcover over small areas of roadside likely represents much less land cover area than
the 80-acre fields modeled in this assessment. However, depending on the conditions, the disturbed
land adjacent to roadwork could present higher risks of transport through runoff and erosion to a
nearby waterbody.
Application of Class AEq biosolids to residential areas (parks, schools, playgrounds, homes) could pose
risks to children from incidental soil ingestion of biosolids or soil contaminated with PFOA or PFOS. The
EPA has posted non-cancer residential soil screening levels for CERCLA site evaluation at 1.9 ppb and 6.3
ppb for PFOA and PFOS, respectively, based on non-cancer risks; soil levels based on cancer for PFOA are
lower at 0.019 ppb (more stringent) (US EPA 2024i). The CERCLA screening levels are calculated with a
soil ingestion rate (200 mg/day) based on upper percentile of behavior patterns of children presented in
the Exposure Factors Handbook. The soil ingestion rate used in developing CERCLA soil screening levels
is larger than the soil ingestion rate used in this assessment (40 mg/day), which is meant to represent
central tendency exposures. The goal of this central tendency risk assessment is to identify the potential
scope and magnitude of risks under different biosolids use and disposal scenarios; historically, EPA
biosolids assessments have used upper percentile estimates to derive risk-based values, consistent with
other EPA programs (US EPA 1992; US EPA 2003a).
4.8 Additional Risk Considerations for All Scenarios
This draft risk assessment is based on the simplification that the risk to human health from sewage
sludge use or disposal can be represented by focusing on the concentrations of PFOS or PFOA in sewage
sludge and the resulting soil or other media concentrations. However, studies of sewage sludge indicate
that precursors to both PFOS and PFOA are present (see Sections 2.2.2 and 2.4) and several studies
indicate that ongoing loading of PFOA and PFOS to soils occurs over time through the degradation of
precursors that were also present in sewage sludge (see Section 2.2.2). Several of these precursors are
also present in EPA Method 1633 and may be monitored with that method in soil, water, and sewage
sludge. Concentrations of PFOA and PFOS would increase in each medium if precursors were included in
this assessment, resulting in an increased risk finding. Precursors to PFOA and PFOS may also pose their
own hazards to human health.
The risk tables in this assessment display results for adults; these risk values represent an average risk
between women and men of adult age. The EPA's final toxicity assessments conclude that both PFOA
and PFOS are likely to cause cancer, hepatic effects and cardiovascular effects; these effects are relevant
DRAFT
111
-------
to men and women in the adult population (US EPA 2024b;c). However, the EPA's toxicity assessments
also conclude that PFOA and PFOS are likely to cause developmental effects in children when mothers
are exposed during pregnancy or when the infants are exposed during early life. The development
effects are listed as co-critical with the hepatic and cardiovascular effects, which indicates that they are
equally sensitive effects as the other critical effects. Women of childbearing age, pregnant women, and
lactating women all have elevated bodyweight-normalized drinking water intake rates compared to the
general adult population (US EPA, 2019c). The median drinking water intake rate used for the general
population in this assessment is 13.4 ml/kg-day. Though the median drinking water intake rates for
women of childbearing age, pregnant women, and lactating women are not presented in the EFH, the
mean drinking water consumption rates for those groups are as follows: 15.6 ml/kg-day for women of
childbearing age, 15.5 ml/kg-day for pregnant women, and 22.9 ml/kg-day for lactating women (EFH,
Chapter 3, Table 3-3; US EPA 2019c). These values are from 14-71% higher than the drinking water
intake rates for the general population. For this reason, there are some drinking water pathways in the
surface disposal scenario that are currently not exceeding the risk threshold for adults but would do so
for lactating women.
4.9 Monte Carlo Analysis
The central tendency deterministic modeling described in this draft risk assessment suggests that there
are unacceptable risks associated with PFOA and PFOS in multiple individual exposure pathways across
every assessed use and disposal practice, even when central tendency exposure parameters are
assumed. Further refinement of the risk assessment from the central tendency model to a probabilistic
risk assessment would result in an increased risk finding because the goal of a probabilistic assessment is
to identify the threshold protective of 95th percentile exposures, while the central tendency modeling is
modeling median (50th percentile) conditions. For this reason, the EPA is not conducting additional
modeling exercises at this time, but rather focusing on sharing the central tendency modeling results
and identifying actions that could be taken to mitigate risks. Any further refinement of the draft risk
assessment (e.g., probabilistic modeling of 95th percentile exposures) would delay future risk
management decisions.
5 UNCERTAINTY, VARIABILITY, AND SENSITIVITY
5.1 Variability
Variability describes the changes in true conditions for a parameter over time or space. Nearly every
parameter used to run the biosolids use and disposal models are variable across U.S. populations or
geography. For example, the meteorological data for each modeled climate scenario (dry, moderate
wet) is variable over time and space. Soil composition can be variable regionally but may also vary within
a single farm or site. Uptake factors for plants and livestock vary by species and location; human
consumption of these plants and animal products also vary individually and by region. Though a Monte
Carlo analysis would allow for the quantification and propagation of variability throughout the modeling
process, the median risks presented at the lowest detectable PFOA and PFOS concentrations are high
enough to ensure that modeling 95th percentile exposure scenarios - even when quantifying variability
and uncertainty - would also result in unacceptable risk scenarios. In selecting median values for most
of the input parameters, and selecting high and low values for the most sensitive parameters, the
outputs represent a set of reasonable risk or hazard values that are relevant to the diversity of biosolids
use and disposal sites in the U.S. Again, this assessment is not designed to capture site-specific
conditions or outcomes, but rather give an estimate of the range of realistic outcomes that are possible
across a variety of common scenarios that exist in the U.S. and inform potential future risk mitigation
actions.
DRAFT
112
-------
5.2 Uncertainty
There are two types of uncertainty: 1) systemic uncertainty, which are unknowns, errors, or assumptions
that produce results in one direction, and 2) random uncertainty, which generates in a spread of final
results above and below the central tendency value (i.e., median value). An example of random
uncertainty is uncertainty introduced by the modeling of the impacts of weather, where there are
random natural variations in parameters like rainfall year to year. This assessment includes both
systemic uncertainties and random uncertainties. Some systemic uncertainties produce results towards
higher-risk outcomes, and some produce results towards lower risk outcomes.
5.2.1 Systemic Uncertainties Resulting in Underestima tion of Risk
This assessment includes several assumptions that could result in an underestimate of risk at specific
sites. Perhaps most significantly, this assessment assumes that the starting concentration of PFOA and
PFOS in biosolids is only 1 ppb. The available biosolids monitoring data from the U.S. suggest that nearly
all biosolids have higher concentrations than this threshold; for example, the annual average PFOS
concentration in biosolids produced in Maine is between 16 ppb and 27 ppb from 2019-2022 and the
annual average PFOA concentration is between 5.3 ppb and 9.4 ppb during this same time window
(Maine DEP, 2023). Sampling from other states (Michigan, California) align with these trends (Link et al.,
2024; Mendez et al., 2021). Highly impacted biosolids can exceed 10 times the average concentrations
(Higgins et al., 2005; 3M, 2001). Furthermore, the modeling indicates that PFOA and PFOS incorporated
into soils from biosolids can be persistent sources of contamination to groundwater, surface water, and
human or animal food over time; concentrations of PFOA and PFOS in the past were likely higher than
currently observed due to the historically high use of PFOA and PFOS in commercial, industrial, and
consumer products. In this way, historic land application of contaminated biosolids could present
significantly more ongoing risks than current-day applications.
A second significant systemic uncertainty that underestimates risk in this assessment is that PFOA and
PFOS precursors cannot be included in the model at this time due incomplete information about which
PFOA and PFOS precursors are present in sewage sludge, the rate of transformation of each precursor to
its terminal degradate, the yield of PFOA and PFOS generation, the toxicity of the precursors, and the
environmental fate of the precursors. As discussed previously, PFOA and PFOS precursors are well
known to act as ongoing sources of PFOA and PFOS in soils, like FTOHs and diPAPs. Some precursors are
measurable using EPA Method 1633, yet others are not. Basing a risk assessment solely on the presence
of PFOA and PFOS will therefore result in modeling that underestimates the exposures and risks
resulting from reuse or disposal of biosolids because this assessment is not accounting for additional
loading of PFOA and PFOS over time as precursors transform. The EPA may consider whether the
environmental precursors for PFOA and PFOS should be included in the future.
A third systemic uncertainty that results in an underestimate of risk are assumptions in the models
related to each receptor's exposure outside a single residence. Currently the models assume that there
are zero exposures to PFOA and PFOS during the times when someone is traveling away from their
home and during the majority of the years of their life when they are not living at the impacted site (60
of their 70 years of life are assumed to have zero PFOA and PFOS exposure from any source). The EPA's
National Primary Drinking Water Regulation for PFAS (April 26, 2024; 89 FR 32532) estimates that 6-10%
of all public drinking water systems in the US contain detectable amounts of PFOA and PFOS (the
prevalence of PFOA and PFOS contamination in private groundwater wells is not known). It is also
known that there are many other pervasive sources of PFOA and PFOS exposure that are unrelated to
biosolids use and disposal (e.g., foods like fish and shellfish; consumer products; household dust). It is
likely that even if a person moved from a residence impacted by PFOA and PFOS contamination from a
biosolids-related source, they would still have ongoing sources of PFOA and PFOS exposure. This
DRAFT
113
-------
assessment also does not attempt to estimate the concentrations of these chemicals that may occur in
human breastmilk due to sewage sludge related exposures or non-sewage sludge related exposures.
Therefore, readers should interpret risk estimates for each pathway narrowly as risk stemming from the
biosolids use or disposal only, and not total risk to the receptor.
A fourth systemic uncertainty that results in underestimation of risk are the assumptions that no
"background" levels of PFOA or PFOS are present in soil from long-range atmospheric transport of PFOA,
PFOS, and their precursors or any other source of non-biosolids related contamination to the farm.
Rankin et al. 2016 sampled soils across North America and the globe that were judged to have "no
evident human impact/' meaning that they were from undeveloped locations with no known or likely
proximal point sources of PFAS. PFOA was detected in all soil samples, and PFOS was detected in all
samples except one from rural Estonia. Even the most remote samples included in this study (locations
like Lake Bonney, Antarctica; Mapunguwe National Park, South Africa; Inuvik, Canada; and Montevideo,
Uruguay) had measurable levels of PFOA from 15 to 270 ppt and PFOS from 4 to 26 ppt. Though it is
possible that there were unknown local sources of PFOA and PFOS to these soils, it is likely that some
amount of PFOA and PFOS are present ubiquitously around the globe and the US. These background soil
concentrations are within the range of modeled soil results for land application of biosolids containing
PFOA and PFOS at 1 ppb, especially in low Koc settings. If contributions of PFOA and PFOS were
considered from ongoing and historic atmospheric deposition, risks and hazards in these pathways
would increase. Again, readers should interpret the risks presented in this draft risk assessment as
added risks solely from sewage sludge use or disposal, not total risks to the receptor.
Finally, this draft risk assessment does not attempt to quantify total (aggregate) exposures or risks to a
single receptor to each chemical, nor does it account for PFOA and PFOS dose additivity. Aggregate
exposure and risk assessment involve the analysis of exposure to a single chemical by multiple pathways
and routes of exposure. This assessment does not aggregate exposure and risk, and instead presents
estimated exposure and risk for each individual exposure pathway that was modeled (i.e., consuming
fish, drinking water, incidentally ingesting soil). This approach does not account for exposure from
multiple modeled pathways simultaneously, sewage sludge-related pathways that were not modeled
due to data gaps (including inhalation and dermal exposure pathways) or exposure pathways not related
to sewage sludge use and disposal (such as exposure from use of personal care products, cleaning
supplies, household dust, etc.).
This decision to assess each pathway individually allows modeling results to be interpreted as risk
contributed from sewage sludge for each pathway across a variety of sewage sludge use and disposal
scenarios. However, in each given scenario, a receptor may be exposed from multiple pathways at the
same time as well as via pathways not modeled in this draft risk assessment. For example, farmers who
consume animal products produced on the farm likely also consume drinking water sourced locally as
many rural areas of the country rely on groundwater. That farmer may also have PFOA or PFOS exposure
that is unrelated to the land application of biosolids on his property. Other farm families with biosolids
land application on their property may be largely self-sufficient, sourcing nearly all of their produce,
animal products, and water from their property. These families would have biosolids-related exposures
from many or all the modeled pathways. Still more individuals may be impacted by a single pathway of
biosolids-related exposures, such as a person who fishes from an impacted waterbody but has no other
sources of biosolids-related exposures, or an individual whose drinking water source is impacted, but
otherwise sources food from non-impacted sources. These pathways are not summed in the assessment
and outside exposure is not accounted for using a relative source contribution (RSC) term or any other
method. There is a substantial amount of variability and uncertainty surrounding the populations who
are exposed to one or multiple pathways of biosolids-related exposure. Because single pathways of
DRAFT
114
-------
exposure may result in exceedances of acceptable risk levels and because there are unknowns regarding
the numbers of people exposed to each combination of exposure pathways, the EPA finds that
presenting pathway-specific risks is the most efficient way of presenting risks at this time.
Cumulative exposure and risk assessment involve analysis of exposures from multiple stressors that
occur simultaneously. A receptor may be exposed to both PFOA and PFOS at the same time. PFOA and
PFOS have been shown to be dose additive (US EPA, 2024e) and are nearly always found in mixtures in
biosolids, and it follows that the environmental media impacted by use or disposal of biosolids also
contains mixtures of PFOA and PFOS. The presence of mixtures and multiple pathways for exposure
would result in higher risks of adverse health effects at a population scale than are reflected in the
pathway-specific results.
5.2.2 Systemic Uncertainties that Result in Overestimation of Risk
One systemic uncertainty resulting in an overestimate of risk stems from a lack of data on PFOA and
PFOS plant uptake factors. There are limited studies of uptake into fruits and vegetables, particularly in
field conditions where biosolids are a source of contamination. Studies of biosolids-amended soils
generally find less uptake under field conditions than when the same biosolids-amended soils are moved
to a pot and plants are cultivated in a greenhouse. Though this assessment aimed to use biosolids-
specific field studies for parameterizing vegetable and fruit uptake, there were no such studies available.
Based on the observed trend that field-based studies produce uptake values that are consistently lower
than greenhouse studies, if there were more biosolids-specific field data available for the entire basket
of often-grown fruits and vegetables in the U.S., the mean uptake factors may be lower than the one
currently used in this assessment. It should be noted that the data for uptake into plants like grasses
used for forage, hay, or silage did not have these same issues related to availability of field data, so
livestock exposures are based on studies of plants in fields where biosolids were land applied.
Another systemic uncertainty resulting in an overestimate of risk is the inability to account for
precursors presence when parameterizing uptake values for food crops, feed crops, and livestock. The
available livestock studies, in particular, may be capturing contamination settings where precursors to
PFOA and PFOS are available in addition to PFOA and PFOS themselves. If these precursors transform in
the livestock to PFOA and PFOS due to metabolism in the liver or other organs, this will result in an
overestimate of PFOA and PFOS uptake. There is more discussion of this potential effect in the livestock
model parameterization section of the report (Section 2.9.3.5).
The current modeling scenario assumes that a farm will receive yearly applications of biosolids for 40
consecutive years, which is consistent with the prior EPA biosolids risk assessment for PCBs and dioxins
(US EPA, 2003a) but lower than the years assumed to calculate the annual and cumulative loading rates
for metals that support the existing sewage sludge regulations under CWA section 503. The use of 40
years may be an overestimate of the loading for some farms, but the EPA does not have data to indicate
the frequency of application at a given site across the country. The current biosolids regulations allow
land application to happen yearly or multiple times per year if the amount of biosolids land applied is
consistent with the nitrogen needs of the crops grown at the farm, and thus, a 40 consecutive years of
annual biosolids application on a farm is a reasonable assumption.
5.2.3 Random Uncertainties
Most of the random uncertainties included in this report stem from modeling parameters where there
are data limitations, resulting in an over- or underestimation of the "true" conditions. For example,
exposure factors used in this assessment (drinking water intake, fish intake, intake of various types of
foods) are based on surveys conducted at various times in the U.S. These surveys vary in sample size and
methodology and may be imperfect measurements of "true" consumption behavior. These surveys also
DRAFT
115
-------
do not capture all potentially relevant consumption behavior, like the consumption of animal livers,
which are known to have higher levels of PFOA and PFOS than muscle tissues. As a result, the mean or
median of the survey may be over- or underestimating reality. Despite these uncertainties, this
assessment relies on the best available datasets for exposure factors.
Other random uncertainties are introduced in the three sites and two Koc settings used in the fate and
transport models. The EPA selected hydrogeological and geochemical conditions at three locations,
using data from those sites to ensure that the combination of parameters at each site were as realistic
as possible. Of course, there is a large amount of variability in the U.S. in site conditions, for example,
variability in depth to groundwater. Though these three sites represent wet, moderate, and dry climates
in the US, they may not capture the full extent of important hydrogeological and geochemical
conditions. Any specific site with biosolids use or disposal may or may not be well-approximated by
these models.
5.3 Sensitivity of Models
The EPA assessed the sensitivity of each model parameter in the groundwater and surface water models
used in this report. Overall, the EPA finds that the Koc, depth from ground surface to water table,
hydraulic gradient in the aquifer, and hydraulic conductivity of saturated zone parameters are the most
sensitive in the groundwater models. Koc and foc are the most sensitive parameters in the surface water
models. The detailed results of the sensitivity analysis can be found in Appendix D.
6 COMPARISON OF MODELED CONCENTRATIONS AND OBSERVED
CONCENTRATIONS IN RELEVANT MEDIA
Though this draft risk assessment is not aiming to model risks stemming from biosolids use or disposal at
any specific site, the modeled concentrations generated in this assessment seem reasonable when
compared to "real life" observations of PFOA and PFOS in various media stemming from contamination
of biosolids. The best datasets available for ground truthing our models would include known PFOA and
PFOS composition of the land-applied biosolids, known timeframes for when the biosolids were applied
and known application rates, observed concentrations of PFOA and PFOS in all of the relevant media
(soil, groundwater, surface water, fish, produce, livestock feed, animal products), and a very detailed
understanding of the hydrogeological and soil conditions at the site. To date, such a complete study is
not available. However, there are other datasets with some of this information that can be used to
determine if the range of modeled results in this assessment are supported by real-world observations.
These datasets generally represent high-end contamination scenarios in the U.S. prior to the phase out
of PFOA and PFOS, though there is one study of a field-based experiment in Ontario, Canada with mass
loading rates of PFOA and PFOS that are more analogous to those used in this draft risk assessment. The
high-end contamination scenarios are also useful in understanding the fate and transport behaviors of
PFOA and PFOS in natural environments after land application of biosolids.
6.1 Biosolids Investigations in Ottawa, Ontario, Canada
In 2008, Canadian researchers applied dewatered municipal biosolids to a 14-hectare experimental
research field located in Ottawa that had never previously had biosolids applied (Gottschall et al., 2017).
The biosolids were applied one time at a rate of 22 Mg dry weight per hectare (equivalent to 22 MT
dw/ha). The biosolids applied to the field contained 1.6 ng/g (ppb) PFOA and 7.2 ng/g (ppb) PFOS. The
researchers then planted winter wheat and spring wheat on the field with biosolids application and a
control plot in the same research station without any history of biosolids application. Both fields were
independently tile drained (tile drains are an artificial subsurface drainage system installed to facilitate
plant growth in wetter climates; these drainage systems are also commonly used in regions of the
DRAFT
116
-------
United States). In this case, the tile drains were installed 1.2 meters below the soil surface. The
researchers then monitored groundwater, tile drainage water, soil, and wheat grain.
The details of sampling strategies for these media are described in Gottschall et al. 2017. In brief,
shallow soil samples (0-0.3 m depth) were collected in triplicate pre-application, 6 months, 9 months
and 12 months post application from each of the 8 sampling locations in the biosolids applied plot and
the reference plot. Due to the competent nature of the dewatered biosolids and the strategy used for
their incorporation (mouldboard plowing), it was also possible to identify biosolids aggregates in the top
layer of soils even a year after biosolids application. These aggregates were also sampled for PFAS
content at 1 month, 2, months, 6 months, and 12 months post biosolids application. Tile water was
sampled during rain events, with the first sample collected within 15 minutes of rain event water
appearing in the drainage system, followed by sampling at 1, 2, and 6 hour intervals. Due to cost
constraints, only the first sample was analyzed for PFAS concentrations because this sample was
expected to have the highest concentration of pollutants. Some additional tile water samples were also
collected during low flow conditions. For groundwater sampling, each field (control and experimental)
had two piezometer nests/wells. Each piezometer nest included three piezometers with intakes
centered at 2, 4, and 6 meters below the soil surface. Groundwater was sampled on a monthly basis pre-
and post-application. Pre- and post-application groundwater samples were then pooled by depth for
PFAS analysis. Grain was sampled from the harvester grain storage bin at various intervals during the
harvesting process and mixed to form a single composite grain sample for each field; the reference field
was harvested first to avoid cross contamination of grain samples. In total, the post-application
monitoring period for this study spanned from October 2008 to November 2009.
Pre-application soils in the reference field and experimental field had low or non-detectable levels of
PFOA and PFOS (PFOA of 118 ppt and non-detectable PFOS in the experimental field; ~100 ppt PFOA and
PFOS in the reference field). In the biosolids application field, post-application soil samples had
increasing levels of PFOA and PFOS throughout the study period. PFOA levels in surface soils increase
from ~100 ppt before application to ~400 ppt at 6 months and ~800 ppt at 9 and 12 months. PFOS levels
in surface soils increase from non-detectable to 200 ppt at 9 months and 400 ppt at 12 months (a 6-
month concentration is not reported for this compound). This increase in soil PFOA and PFOS
concentrations after a single biosolids application could be due to the slow release and mixing of
biosolids aggregates into soils and/or degradation of PFOA and PFOS precursors, which were not
measured in this study.
The soil concentrations in this field study are reasonably well-aligned with the modeled concentrations
of PFOA and PFOS reclamation scenario of this assessment, though the slow breakdown of biosolids
aggregates and the possible presence of PFOA and PFOS precursors are likely influencing the fate and
transport of PFOA and PFOS in the field study. The reclamation scenario modeled in this assessment
assumed an application rate of 50 Mt dry weight per hectare of biosolids containing 1 ppb PFOA and
PFOS. This amounts to an application of 50 mg/ha of PFOA and PFOS. The Ontario study applied
biosolids at a rate of 22 Mt/ha with a starting biosolids concentration of 1.6 ppb PFOA and 7.2 ppb PFOS,
which amounts to an application of 3.52 mg/ha PFOA (~14 x lower than modeled) and 158 mg/ha PFOS
(~3 x higher than modeled). Because our models assume a linear relationship between the PFOA and
PFOS mass loading and the corresponding soil concentrations, our modeling would expect soil
concentrations in this scenario to range from 0.4 to 14 ppt for PFOA and from 135-600 ppt PFOS. The
measured soil values for PFOA (~800 ppt) are higher than what was expected by ~10-80 times, but the
measured values for PFOS (~400 ppt) are within the range of expected results. The discrepancy between
measured and modeled soil concentrations for PFOA could be due to PFOA precursors present in the
DRAFT
117
-------
field study biosolids, the challenges of sampling soils with heterogeneous inclusion of biosolids
aggregates, or other site-specific factors.
The authors report PFOA and PFOS concentrations in tile drainage water and groundwater before and
after the application of biosolids in the experimental and reference plots. For groundwater, the
reference plot had "marginally detectable" (0.5-0.6 ppt) levels of PFOS at the end of the monitoring
period, but no detectable levels of PFOA. In the experimental plot, PFOA was detected in groundwater
after biosolids application, with concentrations ranging from 1.5-3 ppt over the course of the year. PFOS
was also detected in groundwater after the application of biosolids to the experimental plot (0.8 ppt),
but this detection did not occur until one year after the application of biosolids. For tile drainage
samples, the reference plot had one detection of PFOS in tile drainage (~1.2 ppt) before the biosolids
were applied to the experimental plot but had non-detectable levels of PFOS in tile drainage in all
subsequent samples. There was no PFOS detected in tile drainage water at the experimental plot prior
to biosolids application. There was also no PFOA detected in tile drainage water in any of the control
plot samples or in the experimental plot prior to biosolids land application. The PFOS concentrations in
post-application tile drainage water from the experimental plot were mostly non-detectable, but there
was a sample with ~1.2 ppt PFOS shortly after the biosolids land application and a sample with ~0.5 ppt
PFOS about six months following biosolids application. The PFOA concentrations in tile drainage water at
the experimental plot after biosolids application were also mostly non-detectable, but there were three
samples with detections that ranged from ~4 to 24 ppt.
The modeling in this assessment does not attempt to capture the potential effects of tile drainage on
surface water or groundwater fate and transport dynamics for PFOA and PFOS. The modeling in this risk
assessment is also predicting concentrations of PFOA and PFOS in a nearby surface water body (a lake or
pond), which is not analogous to concentrations in tile drainage water. Finally, the low levels of PFOA
and PFOS in biosolids applied in this study result in water media concentrations that are close to the
detection limits for these compounds, which can render results difficult to interpret. That said, the
trends in groundwater and tile drainage water concentrations observed in this study broadly align with
trends observed in this assessment's modeling of groundwater and surface water. Firstly, the
researchers found consistently elevated PFOA concentrations in groundwater and occasionally elevated
PFOA concentrations in tile drainage water. The fact that concentrations were higher and more
frequently detected in groundwater and tile drainage water than PFOS aligns with the observation in our
assessment that PFOA is more mobile in water than PFOS. Assuming a linear relationship between the
mass loading and groundwater concentration, our modeling would predict PFOA groundwater
concentrations in this scenario from essentially zero to 0.17 ppt. The observed concentration of 1.5-3
ppt are 10-20 times higher than the upper range of the modeled values, which is a similar margin of
underestimation observed for the soil media. This again suggests that the presence of PFOA precursors
may be resulting in higher than expected levels of PFOA in the field study. Compared to PFOA, which is
detected in shallow groundwater immediately after the application of biosolids, PFOS does not become
detectable in groundwater until a year after the biosolids were land applied. This also supports the
findings of the modeling that PFOS takes more time to impact groundwater than PFOA. The observed
PFOS concentration of 0.8 ppt is close to the higher range of the estimated concentration based on our
modeling (0.4 ppt). The variability of PFOA and PFOS concentrations in the tile drainage water is likely a
function of many factors, including the amount of rainfall in each rain event where tile drainage water
was sampled.
The study found that PFOA and PFOS were not detectable in grains harvested from either the
experimental or control plot in this study. This finding is in alignment with expectations based on the
DRAFT
118
-------
low soil concentration of PFOA and PFOS in this study and the low observed uptake factors from soils
into the grains of plants including wheat.
Though this study has many differences from the scenarios modeled in this risk assessment, it is
encouraging that the overall trends in transport behavior between PFOA and PFOS in the agricultural
system are similar between the study and the modeled estimates in this risk assessment. The observed
soil and groundwater concentrations in this field study are also within the ballpark of the expected
values based on extrapolation from this assessment's modeling of a single land application to a field (the
reclamation scenario). Finally, this study found that fields with a single low PFOA and PFOS
concentration biosolids application have measurably higher PFOA and PFOS soil concentrations than
those with no history of biosolids application.
6.2 Biosolids Investigations in Decatur, Alabama
From 1990 to 2008, the Decatur Utilities Dry Creek WWTP in Decatur, Alabama treated wastewater
effluent from more than one local industry producing or using PFAS, including PFOA and PFOS. Between
1995 and 2008, the utility supplied over 34,000 DMT of contaminated biosolids to local farmers on
approximately 2,000 hectares of fields across three counties (Lindstrom et al., 2011). The 3M company,
which was the main producer of PFAS at this site, conducted a study that measured PFAS in various
matrices - WWTP effluent, biosolids, input water to the drinking water treatment plant, finished
drinking water, leachate from the local MSW landfill, drinking water reservoir (where applicable), and
surface water from a small pond - at this city and five others from 1999 to 2001. This study was called
the "Multi-City study." The study included four cities with PFAS-related industry (Decatur, AL; Mobile,
AL; Columbus, GA; and Pensacola, FL) and two cities without known PFAS-related industry (Cleveland,
TN; Port St. Lucia, FL). The results of the Multi-City study show that PFOS concentrations in sludge
ranged from not-quantified (detection at a concentration between the detection limit and quantification
limit) to 3,120 ppb for PFOS and non-detect to 244 ppb for PFOA. For both PFOA and PFOS, the highest
sludge concentrations were found in Decatur (3M, 2001).
Understanding the Multi-City sampling results from the Decatur site is complicated because PFAS were
released directly from PFAS industrial facilities, wastewater effluent into the Tennessee River, landfill
leachate at regional landfills accepting industrially-impacted waste, and sewage sludge from the local
WWTP and from New York City. It is not possible to fully disaggregate impacts from each of these
secondary sources in the overall contamination setting at Decatur, especially because the report did not
provide specific sampling locations. However, 3M's sampling found that a small waterbody (it is not
stated where this waterbody was located with respect to biosolids fields or other release points) had
108 to 114 ng/L PFOS and 57 to 63 ng/L PFOA. Though 3M did not detect PFOA or PFOS in Decatur's
drinking water, subsequent analysis by the drinking water utility in 2005 and 2006 found between 30
and 155 ng/L PFOA in finished drinking water.
The 3M Multi-City Study did not include sampling of environmental conditions at any of the sewage
sludge land application sites, but EPA researchers investigated water contamination at various land
application sites used by the Decatur WWTP (Lindstrom et al., 2011). These researchers collected 51
different water samples, including drinking water wells (n = 6), wells used for other purposes (livestock,
watering gardens, washing, n = 13), and surface water (ponds and streams, n = 32). These samples were
collected from 21 separate farms that had received contaminated biosolids. In most cases, the water
sources (wells or surface water) were either on or within 500 meters of a biosolid applied field. Farms
ranged in size from 9 to 308 hectares, with a total area of more than 2000 hectares receiving WWTP
biosolids for as long as 12 years. In the well samples, PFOA was detected in four well samples at
concentrations ranging from 149 to 6,410 ng/L and PFOS was detected in three samples, with
concentrations ranging from 12 to 151 ng/L (the limit of quantification in this study was 10 ng/L for
DRAFT
119
-------
water samples). In surface water samples, PFOA was detected in 24 samples with concentrations
ranging from 13.6 to 11,000 ng/L, and PFOS was detected in 12 samples with concentrations ranging
from 11.6 to 83.9 ng/L. The size of these ponds was not reported.
Additional results from these Decatur land application sites are published in Washington et al. (2010),
which reported PFAS concentrations in soils. These researchers found that PFOA was present in all
samples at concentrations ranging from 3 to 317 ng/g (equivalent to 3,000 to 317,000 ppt) and PFOS
was present in all but one sample, with concentrations ranging from 1.78 to 325 ng/g (1,780 to 325,000
ppt). The EPA authors of these studies note that there are many unknowns about the PFOA and PFOS
content of applied biosolids at each site and the time that elapsed since application; they highlight that
the sewage sludge data available is from a period with anomalously high PFOA content in sludge from
2002 to 2006, and it is not known what the PFOA and PFOS content was in the biosolids that were
applied to each site. It is also not known what types and concentrations of PFOA and PFOS precursors
were present in the sludge that was applied to the sites. Additional data related to these study sites are
also published in Yoo et al. 2009, 2010, and 2011.
In 2009, the USDA sampled blood, tissue, and milk from animals that had grazed on fields that had
received Decatur WWTP biosolids. The results of this sampling were reported in a CDC ATSDR Health
Consultation memo (CDC, 2013). Researchers sampled blood and tissue from 7 cows that had grazed on
"high" application fields and 2 cows that had grazed on "minimally" applied fields. At the time, USDA's
"minimum proficiency level" for PFOA and PFOS in these blood and tissue samples were 20 ppb (20,000
ppt); results below this level were considered "non detections." They did not detect PFOA or PFOS in
these cow tissue or blood samples. The FDA sampled milk from a single dairy cow and milk from a bulk
milk tank that was used by regional dairy farms. The single milk sample from the cow did not result in a
detection of PFOA or PFOS, but the bulk milk tank had 160 ppt PFOS and no detection of PFOA.
Though the various studies of PFOA and PFOS impacts at and around the Decatur biosolids land
application sites do not include all the data necessary to compare this assessment's modeled results to
"real life" setting, there are many trends in the Decatur studies that are also seen in the modeled
results. First, when PFOA and PFOS-contaminated biosolids were land-applied to fields, these studies
show impacted soils, surface waters, groundwater, and dairy cows. These results confirm our modeling
that PFOA is more mobile in water than PFOS, causing more widespread impacts to groundwater and
surface water. These data also show that while PFOS does migrate to surface water and groundwater, it
is more strongly sorbed to soils. Additionally, these data show that PFOS is more likely to be detected in
milk than PFOA, which aligns with our higher uptake factors for PFOS than PFOA in dairy cow scenarios.
In this assessment's models, which are assuming PFOA and PFOS have a concentration of 1 ppb in
biosolids, groundwater concentrations for PFOA range up to 4.3 ng/L and for PFOS range up to 2 ng/L at
pasture farms. Though the exact starting concentrations of PFOA and PFOS in the Decatur sewage
sludge that was land-applied at each site is unknown, one can assume that the concentrations of PFOA
and PFOS in the sewage sludge applied at these sites were the same as the concentrations reported in
3M's Multi-City study (3,120 ppb PFOS and 244 ppb PFOA). Assuming a linear relationship between
PFOA and PFOS concentration in biosolids and their corresponding concentrations in groundwater and
that all other biosolids application settings stay constant, this assessment's model would predict ~1,050
ng/L groundwater concentrations for PFOA and ~6,240 ng/L groundwater concentration for PFOS. This
PFOA concentration is within the range of observed values in Decatur for groundwater in wells near the
land application sites, but the predicted PFOS concentration is higher than the maximum measured
PFOS concentration of 151 ng/L. Similarly extrapolating our modeled surface water samples to assume
starting conditions of 3,120 ppb PFOS and 244 ppb PFOA gives a modeled value of ~ 2,440 ng/L PFOA
and "400-26,500 ng/L PFOS, depending on the climate and Koc scenario. These surface water modeled
DRAFT
120
-------
results are within the range of observed values for PFOA, but higher than the observed values for PFOS.
Given the significant uncertainties around the actual application rate, timing, and PFOA and PFOS
concentrations of Decatur biosolids, the farming practices at the farms with the sampled cows, and the
size and location of the surface water bodies, modeled and observed values are within a reasonable
range.
This study also included samples at a background field that had not received any biosolids. The
background field was sampled in 2007 and 2009. In 2007, PFOA and PFOS were not detectable in the top
10 inches of soil. In 2009, PFOA and PFOS were detected at very low levels in the top 10 inches of soil
(less than 1 ppt for PFOA and 1 to 2 ppt for PFOS), and slightly higher levels in the deeper soils collected
between 38 and 53 inches in depth (PFOA ranging from 64 to 226 ppt and PFOS ranging from ND to 248
ppt). The background results in the top ten inches of soil are consistently below the modeled PFOA and
PFOS soil concentrations observed after land applying low concentration (1 ppb) biosolids for 40 years.
However, the deeper PFOA and PFOS soil concentrations are slightly higher than modeled in this
assessment's land application scenarios. This might reflect the fact that the models assume zero other
sources of PFAS to the field, including zero impacts of atmospheric deposition. This site was near a local
PFAS industry that may have led to localized atmospheric deposition in soils. Soil studies at remote
locations around the globe show that PFOA and PFOS loading in the atmosphere has resulted in small
amounts of atmospheric deposition to soils, especially during the time window when PFOA, PFOS, and
their precursors were actively being manufactured in large quantities (Rankin et al., 2016). Long-term
deposition of PFOA and PFOS at this background site, along with biotic and abiotic mixing of the soil
profile, may have led to a build-up of PFOA and PFOS in soil 30 to 50 inches below the surface. PFOA and
PFOS at this depth may have less ability to be taken up into grasses or other plants that are used to grow
livestock. Similarly, only the top layer of soil, which has low PFOA and PFOS concentrations in this
background site, would be relevant for livestock ingestion of soil.
6.3 Biosolids Investigations in Wixom, Michigan
In 2018, Michigan discovered that the Wixon WWTP had been receiving PFAS waste from a local auto
supplier conducting chrome plating; biosolids sampled that year were found to have 2,150 ppb PFOS
(MPART 2023). PFOA concentrations in the biosolids were much lower, between 1 and 5 ppb (Ml EGLE,
2021c). Michigan selected six historic biosolids land application sites used by this WWTP, where they
sampled drinking water for humans and livestock, soil surface water, crops, and beef (Ml EGLE 2021c).
Three of the sampled sites are owned by the same farmer; these sites are fields ranging from 20 to 35
acres. Each site received annual biosolids applications totaling from 184 to 521 DMT over 5 years.
Another site from this farmer is 120 acres and received a total of 490 DMT over five years. The last two
sites are owned by a second farmer, are 13 to 24 acres, and received from 188 to 242 DMT over 4 or 5
years. The soils at all sites are loamy sand and glacial till; soil borings show interspersed layers of clay.
Soil concentrations at these sites ranged from 2.48 to 96.7 ppb PFOS (2,480-96,700 ppt) and below
detection to 1.53 ppb PFOA (detection limit from 800-900 ppt; highest observed concentrations equate
to 1,530 ppt). Total organic content of the soils ranged from 7,800 to 12,000 mg/kg. Surface water
samples included perched water on the field, water from nearby ponds, water from nearby creeks, and
one tile drain sample. Surface water samples ranged from below detection to 533 ppb PFOS (detection
limit ~1.5 ppt; up to 533,000 ppt) and below detection to 64.4 ppb PFOA (detection limit also ~1.5 ppt;
up to 64,400 ppt). The tile drain sample had a PFOA concentration of 5.98 ppb PFOA and 17.6 ppb PFOS
(5,980 and 17,600 ppt). Groundwater monitoring wells were installed and sampled; pre-existing
livestock and home drinking water wells were also sampled. Groundwater wells all showed non-
detectable levels of PFOA and PFOS (less than 2 ppt). The report authors note that all groundwater wells
are screened below a confining clay layer. In a separate advisory, Michigan PFAS Action Response Team
DRAFT
121
-------
(MPART) shared that beef (roasts and steaks) from one of the sampled farms had between 0.98 and 2.48
ppb PFOS (980-2,480 ppt) (MPART, 2023). There is no publicly available information on the farming
practices at this farm, including if feed was sourced from the farm or what the source of drinking water
was for the animals.
All except one of the fields in this investigation are smaller than the 80-acre field modeled in this
assessment. Like the investigations in Decatur, AL, there are significant uncertainties around the
concentration of PFOA and PFOS in the biosolids that were applied at each site. The biosolids application
rate is within the range of DMT/hectare modeled in the pasture and reclamation scenarios. These sites
accepted biosolids for 5 years; our pasture model assumes annual applications every year for 40 years
and our reclamation model assumes a single application.
The modeled soil concentrations in this assessment range from 3 to 790 ppt for PFOA and 21 to 1100
ppt for PFOS in the pasture farm scenario. These sites have soil concentrations ranging from 2 to 10
times the high-end modeled PFOS concentration (2,480 to 96,700 ppt) and mostly within the modeled
range for PFOA (less than 900 ppt to 1,530 ppt). Given that our modeled PFOS scenario is for fields more
than twice the size of the sampled fields, for application timeframes that amount to 10 times the length
of application at these fields, but at concentrations likely 1/2000 of the concentrations in this setting,
the soil results in this setting are within the ballpark of what would be expected using our models. The
same ballpark agreement is true for PFOA in soils, which was likely applied at concentrations 1 to 5
times the modeled values. Our pasture model scenario found surface water concentrations range from
0.69 to 10 ppt for PFOA and 0.13 to 8.5 ppt for PFOS. In this site, surface water samples ranged from
below detection to 64,400 ppt PFOA and from below detection to 533,000 ppt for PFOS. The higher
range of these results are higher than expected for PFOA and may reflect higher PFOA concentrations in
the applied biosolids than is estimated from the modern-day sample included in the report. The higher
end of the PFOS results is also slightly higher than would be expected if the starting concentration of
biosolids were ~2,000 times what was modeled, though they are within one or two orders of magnitude.
The beef tissue PFOS results that were reported as being associated with grazing on these sites (980 to
2,480 ppt) are 20 times lower than the modeled results on the low end and 250 times lower than the
modeled results on the high end. Again, given the potential that biosolids in this setting were 2000 times
the modeled results, there are significant differences in the sizes of fields and application rates of
biosolids, and there is no information available on the livestock exposure pathways at this farm (e.g.,
feed, water, soil), the observed results are within the ballpark of what would be expected via our
models.
6.4 Biosolids Investigations at Various Farms in Maine
There have been several farms in Maine with PFOA and PFOS impacts from land applying contaminated
biosolids to fields later used for growing crops, growing feed for animals, or grazing animals. Though
investigations at these farms have sampled milk, hay, crops for human consumption, soil, surface water
and groundwater, the specific results for each impacted site have not yet been published in a journal
article or public report. Therefore, these sites cannot be used to compare against our modeling
exercises.
7 REFERENCES
3M Company. (2000). Information on perfluorooctanoic acid and supplemental information on
perfluorooctane sulfonates and related compounds. EPA/OTS Doc. No. #FYI-OTS-0500-1378: 4297
pp.
DRAFT
122
-------
3M. (2001). Analysis of PFOS, FOSA, and PFOAfrom Various Food Matrices Using HPLC
Electrospray/Mass Spectrometry. Retrieved December 2023 from
https://www.documentcloud.org/documents/6151162-3M-Food-2001.html
3M Company. (2003). Perfluorooctanoic Acid Physiochemical Properties and Environmental Fate Data.
3M Center, St. Paul, MN. EPA. Office of Pollution Prevention and Toxics (OPPT); OPPT2003-00 12.
AECOM. (2017). Off-Site Human Health Risk Assessment, December 2017. RAAF Base Williamtown.
Stage 2B Environmental Investigation. Appendix I. Retrieved from
https://web.archive.org.au/awa/20230427171827mp /https:/defence.gov.au/Environment/PFAS/d
ocs/Williamtown/Reports/HHRAReports/20171205HH RADecember2017Volume2.pdf
Alabama A&M & Auburn Universities Extension. (2022). Backyard & Small Poultry Flock Management
Series: Feeding the Laying Hen. Available online:
https://www.aces.edu/blog/topics/farming/backyard-small-poultrv-flock-management-series-
feeding-the-laying-hen/
Ankley, G.T., Cureton, P., Hoke R. A., Houde, M., Kumar, A., Kurias, J., Lanno, R., McCarthy, C., Newsted,
J., Salice, C. J., Sample, B. E., Sepulveda, M.S., Stevens J. & Valsecchi, S. (2020). Assessing the
ecological risks of per- and polyfluoroalkyl substances: Current state-of-the science and a proposed
path forward. Environmental Toxicology & Chemistry, 40(3), 564-605.
Armstrong, D. L., Lozano, N., Rice, C. P., Ramirez, M., & Torrents, A. (2016). Temporal trends of
perfluoroalkyl substances in limed biosolids from a large municipal water resource recovery facility.
Journal of Environmental Management, 165, 88-95. https://doi.Org/10.1016/i.ienvman.2015.09.023
ATSDR (Agency for Toxic Substances and Disease Registry). (2021). Toxicological Profile for
Perfluoroalkyls. C5274127-A. Retrieved from https://www.atsdr.cdc.gov/toxprofiles/tp200.pdf
Balgooyen, S., & Remucal, C.K. (2022). Tributary Loading and Sediment Desorption as Sources of PFASto
Receiving Waters. Environmental Science & Technology, Water, (3), 436-445. DOI:
10.1021/acsestwater.lc00348
Barbo, N., Stoiber, T., Naidenko O., & Andrews, D.Q. (2023). Locally caught freshwater fish across the
United States are likely a significant source of exposure to PFOS and other perfluorinated
compounds. Environmental Research, 220, 115165. https://doi.Org/10.1016/i.envres.2022.115165
Bierbaum, T., Hansen, S. K., Poudel, B., & Haslauer, C. (2023). Investigating rate-limited sorption,
sorption to air-water interfaces, and colloid-facilitated transport during PFAS
leaching. Environmental Science and Pollution Research, 30, 121529-121547
https://doi.org/10.1007/sll356-023-3Q811-2
Bizkarguenaga E., Zabaleta, I., Mijangos, L., Iparraguirre, A., Fernandez, L.A., Prieto, A., & Zuloaga, O.
(2016b). Uptake of perfluorooctanoic acid, perfluorooctane sulfonate and perfluorooctane
sulfonamide by carrot and lettuce from compost amended soil. Science of The Total Environment,
571, 444-451, https://doi.Org/10.1016/i.scitotenv.2016.07.010
Bizkarguenaga, E., Zabaleta, I., Prieto, A., Fernandez, L. A., & Zuloaga, O. (2016a). Uptake of 8:2
perfluoroalkyl phosphate diester and its degradation products by carrot and lettuce from compost-
amended soil. Chemosphere, 152, 309-317, https://doi.Org/10.1016/i.chemosphere.2016.02.130
Blaine, A.C., Rich, C.D., Hundal, L.S., Lau, C., Mills, M.A., Harris, K.M., & Higgins, C.P. (2013). Uptake of
perfluoroalkyl acids into edible crops via land applied biosolids: Field and greenhouse studies.
Environmental Science and Technology, 47 (24), 14062-14069, https://doi.org/10.1021/es403094q
DRAFT
123
-------
Blaine, A. C., C. D. Rich, E. M. Sedlacko, L. S. Hundal, K. Kumar, C. Lau, M. A. Mills, K. M. Harris & C. P.
Higgins (2014). Perfluoroalkyl acid distribution in various plant compartments of edible crops grown
in biosolids-amended soils. Environmental Science and Technology, 48 (14), 7858-7865.
https://doi.org/10.1021/es500Q16s
Borthakur, A., Olsen, P., Dooley, G. P., Cranmer, B. K., Rao, U., Hoek, E. M.V., Blotevogel, J., Mahendra,
S., & Mohanty, S. K. (2021). Dry-wet and freeze-thaw cycles enhance PFOA leaching from subsurface
soils. Journal of Hazardous Materials Letters, Volume 2, 100029.
https://doi.Org/10.1016/i.hazl.2021.100029
Braunig, J., Baduel, C., Herrernan, A., Rotander, A., Donaldson, R., & Mueller. (2017). Fate and
redistribution of perfluoroalkyl acids through AFFF-impacted groundwater. Science of the Total
Environment, 596-597, 360-369.
Brignole, A. J., Porch J. R., Krueger, H. O., & Van Hoven, R. L. (2003). PFOS: a toxicity test to determine
the effects of the test substance on seedling emergence of seven species of plants. Toxicity to
Terrestrial Plants. EPA Docket AR226-1369. Easton, MD: Ltd. Wildlife International.
Brown and Caldwell (prepared for Maine DEP). (2023). An Evaluation of Biosolids Management in Maine
and Recommendations for the Future (December 2023). Available from:
https://www.maine.gov/tools/whatsnew/attach. php?id=12198306&an=l
Brusseau, M. L. (2018). Assessing the potential contributions of additional retention processes to PFAS
retardation in the subsurface. Science of the Total Environment, 613-614, 176-185.
doi:10.1016/j.scitotenv.2017.09.065
Brusseau, M. L. (2020). Simulating PFAS transport influenced by rate-limited multi-process
retention. Water research, 168, 115179. https://doi.Org/10.1016/j.scitotenv.2017.09.065
Brusseau, M. L., Anderson, R. H., & Guo, B. (2020). PFAS concentrations in soils: Background levels
versus contaminated sites. Science of the Total environment, 740, 140017.
https://doi.Org/10.1016/j.scitotenv.2020.140017
Buck R. C., Franklin, J., Berger, U., Conder, J. M., Cousins, I. T., de Voogt, P., Jensen, A. A., Kannan, K.,
Mabury, S. A., & van Leeuwen, S.P. (2011). Perfluoroalkyl and polyfluoroalkyl substances in the
environment: terminology, classification, and origins. Integrated Environmental Assessment and
Management, 2011;7:513-541. doi: 10.1002/ieam.258.
Budtz-J0rgensen, E & Grandjean, P. (2018). Application of benchmark analysis for mixed contaminant
exposures: Mutual adjustment of perfluoroalkylate substances associated with immunotoxicity.
pLoS ONE 13: e0205388. http://dx.doi.org/10.1371/iournal.pone.0205388.
Bumb, A. C., Murphy, C. L., & Everett, L. G. (1992). A Comparison of Three Functional Forms for
Representing Soil Moisture Characteristics. Groundwater, 30, 2, 1770185. 10.1111/j.l745-
6584.1992.tbO 1789.x.
Burkhard, L. (2021). "Evaluation of published bioconcentration factor (BCF) and bioaccumulation factor
(BAF) data for per- and polyfluoroalkyl substances across aquatic species." Environmental Toxicology
and Chemistry, 40 (6), 1530-1543. https://doi.org/10.1002/etc.501Q
Burns. L. A. (2000). Exposure Analysis Modeling System (EXAMS): User Manual and System
Documentation. EPA/600/R100/081. Retrieved from:
https://www.epa.gov/sites/default/files/documents/EXAMREVG.PDF
DRAFT
124
-------
Burns, D. C., Ellis, D. A., Li, H., Mcmurdo, C. J., & Webster, E. (2008). Experimental pKa determination for
PFOA and the potential impact of pKa concentration dependence on laboratory-measured
partitioning phenonmena and environmental modeling. Environmental Science and Technology, 42,
9283-9288. https://doi.org/10.1021/es802047v
Bursian, S. J., Link, J. E., McCarty, M., & Simcik, M. F. (2021). The subacute toxicity of perfluorooctane
sulfonate and/or perfluorooctanoic acid and legacy aqueous film-forming foams to Japanese quail
(Coturnix japonica) chicks. Environmental Toxicology and Chemistry, 40: 695-710.
https://doi.org/10.10Q2/etc.4684
Butenhoff, J. L., Chang, S. C., Olsen, G. W., & Thomford, P. J. (2012). Chronic dietary toxicity and
carcinogenicity study with potassium perfluorooctanesulfonate in Sprague Dawley rats. Toxicology,
293, 1-15. http://dx.doi.Org/10.1016/i.tox.2012.01.003
Calafat, A. M., Kato, K., Hubbard, K., Jia T., Botelho, J. C., & Wong, L. Y. (2019). Legacy and alternative
per- and polyfluoroalkyl substances in the US general population: paired serum-urine data from the
2013-2014 National Health and Nutrition Examination Survey. Environment International, 131,
105048. doi:10.1016/j.envint.2019.105048.
Calafat, A. M., Wong, L. Y., Kuklenyik, Z., Reidy, J. A., & Needham, L. L. (2007). Polyfluoroalkyl chemicals
in the US population: data from the National Health and Nutrition Examination Survey (NHANES)
2003-2004 and comparisons with NHANES 1999-2000. Environmental Health Perspectives
115(11): 1596-1602. doi:10.1289/ehp,10598.
CDC (US Centers for Disease Control and Prevention). (2013). Health Consultation, Exposure
Investigation Report. Perfluorochemical Serum Sampling in the vicinity of Decatur, Alabama,
Morgan, Lawrence, and Limestone Counties. Atlanta, GA. Retrieved from
https://www.atsdr.cdc.gov/HAC/pha/Decatur/Perfluorochemical Serum%20Sampling.pdf
CDC (US Centers for Disease Control and Prevention). (2017). Fourth Report on Human Exposure to
Environmental Chemicals, Updated Tables, January 2017, Volume One. US Department of Health
and Human Services, CDC, Atlanta, GA. Accessed April 2022.
https://www.cdc.gov/biomonitoring/pdf/fourthreport updatedtables volumel jan2017.pdf; Data
summary available at https://www.atsdr.cdc.gov/pfas/docs/PFAS in People.pdf
CDPHE (Colorado Department of Public Health & Environment). (2018). Biosolids plant available nitrogen
(PAN) and agronomic rate calculation tool. Excel spreadsheet. Available at
https://www.colorado.gov/pacific/cdphe/wq-biosolids-permits.
Chou, W-C., Tell, L. A., Baynes, R. E., Davis, J. L., Cheng, Y-H., Maunsell, F. P., Riviere, J. E. & Lin, Z. (2023).
Development and application of an interactive generic physiologically based pharmacokinetic
(igPBPK) model for adult beef cattle and lactating dairy cows to estimate tissue distribution and
edible tissue and milk withdrawal intervals for per- and polyfluoroalkyl substances (PFAS). Food and
Chemical Toxicology, 181, 114062. https://doi.Org/10.1016/i.fct.2023.114062
City and Borough of Juneau Wastewater Utility (The). (2017). Biosolids End Use Analysis. Juneau, AK.
Retrieved from https://iuneau.org/wp-content/uploads/2016/10/Biosolids-End-Use-Analvsis-
Final.pdf.
Costanza, J., Arshadi, M., Abriola, L.M. & K.D. Pennell. (2019). Accumulation of PFOA and PFOS at the air-
water interface. Environmental Science & Technology Letters, 6(8), 487-491.
https://doi.org/10.1021/acs.estlett.9b00355
DRAFT
125
-------
Criswell R.L., Simones T., Chatterjee M., Waite J., Diaz S., & Smith A. (2024). Quantifying levels of per-
and polyfluoroalkyl Substances (PFAS) in water and serum after contamination from agricultural
biosolid application. Environ Int. 24;190:108850. doi: 10.1016/j.envint.2024.108850.
Custer, C.M., Custer, T.W., Schoenfuss, H.L., Poganski, B.H. & Solem, L. (2012). Exposure and effects of
perfluoroalkyl compounds on tree swallows nesting at Lake Johanna in east central Minnesota, USA.
Reproductive Toxicology, 33, 556-562. https://doi.Org/10.1016/i.reprotox.2011.01.005
D'eon, J. C., & Mabury, S.A. (2007). Production of perfluorinated carboxylic acids (PFCAs) from the
biotransformation of polyfluoroalkyl phosphate surfactants (PAPS): Exploring routes of human
contamination. Environmental Science & Technology, 41, 4799-4805.
https://doi.org/10.1021/es070126x
D'eon, J. C., P.W. Crozier, V.I. Furdui, E.J. Reiner, E.L. Libelo & S.A. Mabury. (2009). Observation of a
commercial fluorinated material, the polyfluoroalkyl phosphoric acid diesters, in human sera,
wastewater treatment plant sludge, and paper fibers. Environmental Science & Technology, 43(12),
4589-4594. https://doi.org/10.1021/es9001Q0d
Dal Bosco, A., Mugnai, C., Rosati, A., Paoletti, A., Caporali, S., & Castellini, C. (2014). Effect of range
enrichment on performance, behavior and forage intake of free-range chickens. Journal of Applied
Poultry Research, 23, 137-145. http://dx.doi.org/10.3382/iapr.2013-00814
Dauchy, X., Boiteux, V., Colin, A., Hemard, J., Bach, C., Rosin, C., & Munoz, J. F. (2019). Deep seepage of
per-and polyfluoroalkyl substances through the soil of a firefighter training site and subsequent
groundwater contamination. Chemosphere, 214, 729-737.
De Silva A.O., Armitage, J.M., Bruton, T.A., Dassuncao, C., Heiger-Bemays, W., Hu, X.C., Karrman, A.,
Kelly, B., Ng, C., Robuck, A., Sun, M., Webster, T.F., & Sunderland, E.M. (2021). PFAS exposure
pathways for humans and wildlife: a synthesis of current knowledge and key gaps in understanding.
Environmental Toxicology Chemistry. 40:631-657.
De Silva, A.O., Benskin, J. P., Martin, L.J., Arsenault, G., McCrindle, R., Riddel I, N., Martic, J.W., & Mabury,
S.A. (2009). Disposition of perfluorinated acid isomers in Sprague-Dawley rats. Part 2: Subchronic
dose. Environmental Toxicology and Chemistry, 28, 3, 555-567.
Death, C., Bell, C., Champness, D., Milne, C., Reichman, S. & Hagen, T. (2021). Per- and polyfluoroalkyl
substances (PFAS) in livestock and game species: a review. Science of the Total Environment,
774:144795.
D'Hollander, W., Herzke, D., Huber, S., Hajslova, J., Pulkrabova, J., Brambilla, G., De Filippis, S.P.,
Bervoets, L., & de Voogt, P. (2015). Occurrence of perfluorinated alkylated substances in cereals,
salt, sweets and fruit items collected in four European countries. Chemosphere, 129:179-85. doi:
10.1016/j.chemosphere.2014.10.011.
Dong, Z., Wang, H., Yu, Y.Y., Li, Y.B., Naidu, R., & Liu, Y. (2019). Using 2003-2014 US NHANES data to
determine the associations between per- and polyfluoroalkyl substances and cholesterol: Trend and
implications. Ecotoxicology and Environmental Safety, 173, 461-468.
Drew, R., Hagen, T.G., & Champness, D. (2021). Accumulation of PFAS by livestock - determination of
transfer factors from water to serum for cattle and sheep in Australia. Food additives &
Contaminants: Part A. https://doi.org/10.1080/19440049.2Q21.1942562.
DRAFT
126
-------
Drew, R., Hagen, T.G, Champness, D., & Sellier, A. (2022). Half-lives of several polyfluoroalkyl substances
(PFAS) in cattle serum and tissues. Food Additives & Contaminants: Part A^:!, 320-340.
https://doi.org/10.1080/19440049.2021.1991004.
Ebrahimi, F., Lewis, A.J., Sales, C.M., Suri, R. & McKenzie, E.R. (2021). Linking PFAS partitioning behavior
in sewage solids to the solid characteristics, solution chemistry, and treatment processes.
Chemosphere, 271, 129530. https://doi.Org/10.1016/j.chemosphere.2020.129530
ECCC (Environment and Climate Change Canada). (2018). ECCC PFOS Guidelines: Canadian
Environmental Protection Act, 1999 Federal Environmental Quality Guidelines Perfluorooctane
Sulfonate (PFOS). pp.1-22.
ECOS (Environmental Council of the States). (2023). PFAS in Biosolids: A Review of State Efforts &
Opportunities for Action. Retrieved from: https://www.ecos.org/wp-
content/uploads/2023/01/PFAS-in-Biosolids-A-Review-of-State-Efforts-and-Opportunities-for-
Action.pdf
EFSA (European Food Safety Authority). (2008). Opinion of the scientific panel on contaminants in the
food chain on perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) and their salts.
EFSA Journal. 653: 1-131.
EFSA Panel on Contaminants in the Food Chain (EFSA CONTAM Panel). (2018). Risk to human health
related to the presence of perfluorooctane sulfonic acid and perfluorooctanoic acid in food, doi:
10.2903/j.efsa.2018.5194.
EFSA Panel on Contaminants in the Food Chain (EFSA CONTAM Panel). (2020). Risk to human health
related to the presence of perfluoroalkyl substances in food, doi: 10.2903/j.efsa.2020.6223
Eriksson, U., Haglund, P.& Karrman, A. (2017). Contribution of precursor compounds to the release of
per- and polyfluoroalkyl substances (PFASs) from wastewater treatment plants (WWTPs). Journal of
Environmental Sciences, 61, 80-90. https://doi.Org/10.1016/i.ies.2017.05.004
FDA (US Food and Drug Administration). (2012). Determination of Perfluorochemicals in Cow's Milk
Using Liquid Chromatography-Tandem Mass Spectrometry. Retrieved from
https://www.fda.gov/food/process-contaminants-food/analytical-results-testing-food-pfas-
environmental-contamination
FDA (US Food and Drug Administration). (2016). Indirect Food Additives: Paper and Paperboard
Components Rule. 81 FR 83672. Retrieved from https://www.govinfo.gov/app/details/FR-2016-ll-
22/2016-28116/summary
FDA (US Food and Drug Administration). (2022). Analytical Results for PFAS in 2022 Seafood Survey
(Parts Per Trillion). Retrieved from https://www.fda.gov/media/159570/downloadPattachment
FDA (US Food and Drug Administration). (2023). Analytical Results for PFAS in 2021 Total Diet Study
Sampling (Parts Per Trillion)—Dataset—7. Retrieved from https://www.fda.gov/food/process-
contaminants-food/analytical-results-testing-food-pfas-environmental-contamination.
FishBase. (2024). Accessed May 10, 2024. Available from: https://www.fishbase.se/search.php
Fredriksson, F., Eriksson, U., Karrman, A., &Yeung, L.W.Y. (2022). Per-and polyfluoroalkyl substances
(PFAS) in sludge from wastewater treatment plants in Sweden - First findings of novel fluorinated
copolymers in Europe including temporal analysis. Science of The Total Environment, 846, 157406.
http://dx.doi.Org/10.1016/i.scitotenv.2022.157406
DRAFT
127
-------
Frisbee, S., Brooks, A., Maher, A., Flensborg, P., Arnold, S., Fletcher, T., Steenland, K., Shankar, A., Knox,
S., Pollard, C., Halverson, J., Vieira, V., Jin, C., Leyden, K., Ducatman, A., 2009. The C8 health project:
Design, methods, and participants. Environ Health Perspectives, 117 (12), 1873-1882
Fry, M. M., Rothman, G., Young, D. F., Thurman, N. (2016). Daily gridded weather for pesticide exposure
modeling. Environmental Modelling & Software, (167-173).
Ghisi, R., Vamerali, T., & Manzetti, S. (2019). Accumulation of perfluorinated alkyl substances (PFAS) in
agricultural plants: A review. Environmental Research, 169, 326-341.
https://doi.Org/10.1016/i.envres.2018.10.023
Giesy, J.P. & Kannan, K. (2001). Global distribution of perfluorooctane sulfonate in wildlife.
Environmental Science & Technology, 35(7): 1339-1342.
Gockener, B., Eichhorn, M., Lammer, Kotthoff, M., Kowalczyk, J., Numata, J., Schafft, H., Lahrssen-
Wiederholt, M., & Bucking, M. (2020). Transfer of PFAS from feed into the eggs of laying hens. Park
1: Analytical results including a modified total oxidizable precursor assay. Journal of agricultural and
food chemistry, 68, 12527-12538.
Goecke-Flora C.M. & Reo, N.V. (1996). Influence of carbon chain length on the hepatic effects of
perfluorinated fatty acids. A 19F- and 31P-NMR investigation. Chemical Research in Toxicology, 9, 4,
689-95. doi: 10.1021/tx950217k. PMID: 8831811
Gottschall., N., Topp, E., Edwards, M., Payne, M., Kleywegt, S., & Lapen, D.R. (2017). Brominated flame
retardants and perfluoroalkyl acids in groundwater, tile drainage, soil, and crop grain following a
high application of municipal biosolids to a field. Science of the Total Environment, 574, 1345-1359.
http://dx.doi.Org/10.1016/i.scitotenv.2016.08.044
Gravesen, C. R., Lee, L. S., Choi, Y. J., Silveira, M. L., & Judy, J. D. (2023). PFAS release from wastewater
residuals as a function of composition and production practices. Environmental Pollution, 322,
121167. https://doi.Org/10.1016/j.envpol.2023.121167
Gredelj, A., Nicoletto, C., Valsecchi, S., Ferrario, C., Polesello, S., Lava, R., Zanon, F., Barausse, A., Palmeri,
L., Guidolin, L., & Bonato, M. (2020). Uptake and translocation of perfluoroalkyl acids (PFAA) in red
chicory (Cichorium intybus L.) under various treatments with pre-contaminated soil and irrigation
water. Science of The Total Environment, 708, 134766,
https://doi.Org/10.1016/i.scitotenv.2019.134766.
Guelfo, J. L., Wunsch, A., McCray, J., Stults, J. F., & Higgins, C. P. (2020). Subsurface transport potential of
perfluoroalkyl acids (PFAAs): Column experiments and modeling. Journal of Contaminant
Hydrology, 233, 103661.
Guo, B., Zeng, J., Brusseau, M. L. & Zhang, Y. (2022). A screening model for quantifying PFAS leaching in
the vadose zone and mass discharge to groundwater. Advances in Water Resources, 21(160),
104102. https://doi.Org/10.1016/i.advwatres.2021.104102
Guo, B., Zeng, J., & Brusseau, M. L. (2020). A mathematical model for the release, transport, and
retention of per- and polyfluoroalkyl substances (PFAS) in the vadose zone. Water Resources
Research, 56, e2019WR026667. https://doi.org/10.1029/2019WR026667
Harris, B., Jr., & Van Horn, H. H. (1992). Water and Its Importance to Animals. Circular 1017, Dairy
Production Guide, Florida Cooperative Extension Service.
DRAFT
128
-------
Hekster, F. M., Laane, R. W., & de Voogt, P. (2003). Environmental and toxicity effects of
perfluoroalkylated substances. Reviews of Environmental Contamination and Toxicology 179:99-
121.
Helmer, R.W., Reeves, D.M. & Cassidy, D.P. (2022). Per- and polyfluorinated alkyl substances (PFAS)
cycling within Michigan: Contaminated sites, landfills and wastewater treatment plants. Water
Research, 210, 117983. https://doi.Org/10.1016/j.watres.2021.117983
Higgins, C. P., Field, J. A., Criddle, C. S., & Luthy, R. G. (2005). Quantitative determination of
perfluorochemicals in sediments and domestic sludge. Environmental Science and Technology, 39,
3946-3956. https://doi.org/10.1021/esQ48245p
Higgins, C. P., & Luthy, R. G. (2006). Sorption of Perfluorinated Surfactants on Sediments. Environmental
Science & Technology, 40, 23, 7251-7256.https://doi.org/10.1021/esQ61000n
Huang, S., Sima, M., Long, Y., Messenger, C. & Jaffe, P.R. (2022). Anaerobic degradation of
perfluorooctanoic acid (PFOA) in biosolids by Acidimicrobium sp. strain A6. Journal of Hazardous
Materials, 424(D), 127699. https://doi.Org/10.1016/j.jhazmat.2021.127699
Johansson, J. H., Yan, H., Berger, U. & Cousins, I.T. (2017). Water-to-air transfer of branched and linear
PFOA: Influence of pH, concentration and water type. Emerging Contaminants, 3, 46-53.
https://doi.Org/10.1016/j.emcon.2017.03.001
Johnson, G. R. (2022). PFAS in soil and groundwater following historical land application of biosolids.
Water Research, 211, 118035. https://doi.Org/10.1016/j.watres.2021.118035
Johnston, J. Ebel, E., Williams, M., Esteban, E., Lupton, S., Scholljegerdes, J., Ivey, S., Powell, M. & Smith,
D. (2023). Blood-based ante-mortem method for estimating PFOS in contaminated dairy cattle.
Agricultural Science & Technology, 3, 835-844.
Kaiser M. A., Larsen, B. S., Kao, C-PC., & Buck, R. C. (2005). Vapor pressures of perfluoro-octanoic, -
nonanoic, -decanoic, undecanoic, and dodecanoic acids. Journal of Chemical & Engineering Data 50,
6, 1841-1843.
Kim, H., Ekpe, O. D., Lee, J. H., Kim, D. H., & Oh, J. E. (2019). Field-scale evaluation of the uptake of
Perfluoroalkyl substances from soil by rice in paddy fields in South Korea. Science of The Total
Environment, 671, 714-721. https://doi.Org/10.1016/i.scitotenv.2019.03.240.
Kowalczyk, J., Ehlers, S., Oberhausen, A., Tischer, M., Furst, P., Schafft, H.& Lahrssen-Wiederholt, M.
(2013). Absorption, Distribution, and Milk Secretion of the Perfluoroalkyl Acids PFBS, PFHxS, PFOS,
and PFOA by Dairy Cows Fed Naturally Contaminated Feed. Journal of Agricultural and Food
Chemistry, 61, 2903-2912. Dx.doi.org/10.1021/jf304680j
Kowalczyk, J., Gockener, B., Eichhorn, M., Kotthoff, M., Bucking, M., Schafft, H., Lahrssen-Wiederholt M.
& Numata, J. (2020). Transfer of Per- and Polyfluoroalkyl Substances (PFAS) from Feed into the Eggs
of Laying Hens. Part 2: Toxicokinetic Results Including the Role of Precursors. Journal of Agricultural
and Food Chemistry, 68, 12539-12548. https://dx.doi.org/10.1021/acs.iafc.0c04485
Kudo, N., Katakura, M., Sato, Y., & Kawashima, Y. (2002). Sex hormone-regulated renal transport of
perfluorooctanoic acid. Chemico-Biological Interactions, 139, 3, 301-16.
Lammer, R., Weidemann, E., Gockener, B., Stahl, T., Breuer, J., Kowalczyk, J., Just, H., Boeddinghaus, R.S.,
Gassmann, M., Kling, H-W., & Bucking. M. (2022). Evaluation of the transformation and leaching
behavior of two polyfluoroalkyl phosphate diesters in a field lysimeter study. Journal of Agricultural
and Food Chemistry, 70(45), 14329-14338. https://doi.org/10.1021/acs.iafc.2cQ3334
DRAFT
129
-------
Lange C.C. (2000). The Aerobic Biodegradation of N-EtFOSE Alcohol by the Microbial Activity Present in
Municipal Wastewater Treatment Sludge. 3M Company; St. Paul, MN, USA.
Lasters, R., Sundert, K.V., Groffen, T., Buytaert, J., Eens, M., Bervoets, L. (2023). Prediction of
perfluoroalkyl acids (PFAAs) in homegrown eggs: Insights into abiotic and biotic factors affecting
bioavailability and derivation of potential remediation measures. Environment International, 181,
108300.
Lechner, M. & Knapp, H. (2011). Carryover of perfluorooctanoic acid (PFOA) and perfluorooctane
sulfonate (PFOS) from soil to plant and distribution to the different plant compartments studied in
cultures of carrots (Daucus carota ssp Sativus), potatoes (Solanum tuberosum), and cucumbers
[Cucumis sativus). Journal of Agriculture and Food Chemistry, 59, 20, 11011-11018.
https://doi.org/10.1021/if201355y
Lee, J.J. & Schultz, I.R. (2010). Sex differences in the uptake and disposition of perfluorooctanoic acid in
fathead minnows after oral dosing. Environmental Science and Technology. 44, 44: 491-496.
Lee, H., D'eon, J. & Mabury, S.A. (2010). Biodegradation of polyfluoroalkyl phosphates as a source of
perfluorinated acids to the environment. Environmental Science & Technology, 44(9), 3305-3310.
https://doi.org/10.1021/es9Q28183
Lee., H., Tevlin, A., Mabury, S.A., & Mabury, S.A. (2013). Fate of polyfluoroalkyl phosphate diesters and
their metabolites in biosolids-applied soil: Biodegradation and plant uptake in greenhouse and field
experiments. Environmental Science and Technology, 48, 340-349.
Lee, H., Tevlin, A.G., Mabury, S.A. & Mabury, S.A. (2014). Fate of Polyfluoroalkyl Phosphate Diesters and
Their Metabolites in Biosolids-Applied Soil: Biodegradation and Plant Uptake in Greenhouse and
Field Experiments. Environmental Science & Technology, 48,1, 340-349. DOI: 10.1021/es403949z
Lee, J.W., Lee, J.W., Kim, K., Shin, Y.J., Kim, J., Kim, S., Kim, H., Kim, P. & Park, K. (2017). PFOA-induced
metabolism disturbance and multi-generational reproductive toxicity in Oryzias latipes. Journal of
Hazardous Materials, 340, 231-240.
Lee, J.W., Choi, K., Park, K., Seong, C., Yu, S.D., & Kim, P. (2020). Adverse effects of perfluoroalkyl acids
on fish and other aquatic organisms: A review. Science of the Total Environment. 707, 135334.
https://doi.Org/10.1016/j.scitotenv.2019.135334
Lesmeister, L., Lange, F. T., Breuer, J., Biegel-Engler, A., Giese, E., & Scheurer, M. (2021). Extending the
knowledge about PFAS bioaccumulation factors for agricultural plants - A review. Science of The
Total Environment, 766, 142640. https://doi.Org/10.1016/i.scitotenv.2020.142640.
Lewis, A. J., Ebrahimi, F., McKenzie, E.R., Suri, R. & Sales, C.M. (2023). Influence of microbial weathering
on the partitioning of per- and polyfluoroalkyl substances (PFAS) in biosolids. Environmental Science:
Processes & Impacts, 25, 415-431. https://doi.org/10.1039/D2EM0035QC
Lewis, R.J., Sr., ed. 2004. Sa"s Dangerous Properties of Industrial Materials. luh ed. Wiley Interscience,
Wiley & Sons, Inc., Hoboken, NJ.
Li J., Sun, J. & Li, P. (2022). Exposure routes, bioaccumulation and toxic effects of per- and
polyfluoroalkyl substances (PFASs) on plants: A critical review. Environment International, 158,
106891. https://doi.Org/10.1016/i.envint.2021.106891
Li, H., Ellis, D. & Mackay, D. (2007). Measurement of low air-water partition coefficients of organic acids
by evaporation from a water surface. Journal of Chemical Engineering Data, 52, 1580-1584.
DRAFT
130
-------
Li, P., Oyang, X., Zhao, Y., Tu, T., Tian, X., Li, L., Zhao, Y., Li, J., Xiao, Z. (2019). Occurrence of
perfluorinated compounds in agricultural environment, vegetables, and fruits in regions influenced
by a fluorine-chemical industrial park in China. Chemosphere, 225, 650-669.
https://doi.Org/10.1016/j.chemosphere.2019.03.045
Li, Y., He, L., Lv, L., Xue, J., Wu, L., Zhang, Z. & Yang, L. (2021). Review on plant uptake of PFOS and PFOA
for environmental cleanup: potential and implications. Environmental Science and Pollution
Research 28, 30459-30470. https://doi.org/10.1007/sll356-021-14Q69-0
Liagkouridis, I., Awad, R., Schellenberger, S., Plassmann, M., Cousins, I.T., & Benskin, J.P. (2022).
Combined Use of Total Fluorine and Oxidative Fingerprinting for Quantitative Determination of Side-
Chain Fluorinated Polymers in Textiles. Environmental Science & Technology Letters 9, 30-36.
https://doi.org/10.1021/acs.estlett.lc00822
Lide, D.R. 2007. CRC Handbook of Chemistry and Physics 88TH Edition 2007-2008. CRC Press, Taylor &
Francis, Boca Raton, FL. p. 3-412.
Lin, Q., Zhou, C., Chen, L., Li, Y., Huang, X., Wang, S., Qiu, R., & Tang, C. (2020). Accumulation and
associated phytotoxicity of novel chlorinated polyfluorinated ether sulfonate in wheat seedlings.
Chemosphere, 249, 126447. https://doi.Org/10.1016/i.chemosphere.2020.126447
Lindstrom, A. B., Strynar, M. J., Delinsky, A. D., Nakayama, S. F., McMillan, L., Libelo, E. L., Neil I, M. &
Thomas, L. (2011). Application of WWTP biosolids and resulting perfluorinated compound
contamination of surface and well water in Decatur, Alabama, USA. Environmental Science &
Technology, 45,19, 8015-8021. https://doi.org/10.1021/eslQ39425
Link, G. W., Reeves, D. M., Cassidy, D. P. & Coffin, E. S. (2024). Per- and polyfluoroalkyl substances (PFAS)
in final treated solids (Biosolids) from 190 Michigan wastewater treatment plants. Journal of
Hazardous Materials, 463, 132734. https://doi.Org/10.1016/i.ihazmat.2023.132734
Liu, C., & Liu, J. (2016). Aerobic biotransformation of polyfluoroalkyl phosphate esters (PAPs) in soil.
Environmental Pollution, 212, 230-237. http://doi.Org/10.1016/i.envpol.2016.01.069
Liu, Z., Lu, Y., Shi, Y., Wang, P., Jones, K., Sweetman, A.J., Johnson, A.C., Zhang, M., Zhou, Y., Lu, X., Su, C.,
Sarvajayakesavaluc, S., & Khan, K. (2017). Crop bioaccumulation and human exposure of
perfluoroalkyl acids through multi-media transport from a mega fluorochemical industrial park,
China. Environment International, 106, 37-47, https://doi.Org/10.1016/i.envint.2017.05.014.
Liu, Z., Lu, Y., Song, X., Jones, K., Sweetman, A.J., Johnson, A.C., Zhang, M., Lu, X. & Su, C. (2019).
Multiple crop bioaccumulation and human exposure of perfluoroalkyl substances around a mega
fluorochemical industrial park, China: Implication for planting optimization and food safety.
Environment International, 127, 671-684, https://doi.Org/10.1016/i.envint.2019.04.008.
Lyu, X., Liu, X., Sun, Y, Ji, R., Gao, B., & Wu, J. (2019). Transport and retention of perfluorooctanoic acid
(PFOA) in natural soils: Importance of soil organic matter and mineral contents, and solution ionic
strength. Journal of Contaminant Hydrology, 225, 103477.
https://doi.Org/10.1016/j.jconhyd.2019.03.009
Lupton, S. J., Huwe, J. K., Smith, D. J., Dearfield, K. L., & Johnston, J. J. (2012). Absorption and excretion
of 14C-perfluorooctanoic acid (PFOA) in angus cattle (Bos taurus). J. Agric. Food Chem., 60,
1128-1134.
DRAFT
131
-------
Lupton, S. J., Dearfield, K. L., Johnston, J. J., Wagner, S., & Huwe, J. K. (2015). Perfluorooctane sulfonate
plasma half-life determination and long-term tissue distribution in beef cattle (Bos taurus). J. Agric.
Food Chem., 63, 10988-10994.
Lupton, S. J., Smith D. J., Scholljegerdes E., Ivey, S., Young, W., Genualdi, S., DeJager, L., Snyder, A.,
Esteban, E., & Johnston, J. J. (2022). Plasma and skin perfluoroalkyl substance (PFAS) levels in dairy
cattle with lifetime exposures to PFAS contaminated drinking water and feed. Journal of Agriculture
and Food Chemistry. 70:15945-15954. https://doi.org/10.1021/acs.iafc.2cQ662
Maine DEP. (2023). An evaluation of biosolids management in Maine and recommendations for the
future.
Martin, J. W., Mabury, S.A., Solomon, K.R., Muir, D.C. (2003a). Bioconcentration and tissue distribution
of perfluorinated acids in rainbow trout (Oncorhynchus mykiss). Environmental Toxicology and
Chemistry, 22, 196-204. https://doi.org/10.1002/etc.562022Q126
Martin, J.W., Mabury, S.A., Solomon, K.R., & Muir, D.C. (2003b). Dietary accumulation of perfluorinated
acids in juvenile rainbow trout (Oncorhynchus mykiss). Environmental Toxicology and Chemistry, 22,
189-195. https://doi.org/10.1002/etc.562022Q125
Mendez, M., Lin, D., Wong, A., Yee, D., & Sutton, R. (2021). Study of Per- and Polyfluoroalkyl Substances
in Bay Area POTWs: Phase 1 Memo. San Francisco Estuary Institute, Richmond, CA.
https://bacwa.org/wp-content/uploads/2023/03/Memo BACWA-PFAS-Phase-l.pdf
MDH (Minnesota Department of Health). (2010). Community Brief: East Metro PFC Biomonitoring Study
follow-up Analysis of PFC Levels in Blood and Past Levels of PFCs in Drinking Water. Retrieved from
https://www.health.state.mn.us/communities/environment/biomonitoring/proiects/pfas.html
MDHHS (Michigan Department of Health and Human Services). (2023). PFAS in Deer. Accessed
December, 2023 from
https://www.michigan.gov/pfasresponse/fishandwildlife/deer#:~:text=For%20this%20reason%2C%
20MDHHS%20recom mends.at%20800%2D648%2D6942
MDIFW (Maine Department of Inland Fisheries & Wildlife). (2021). PFAS Do Not Eat Advisory for deer
and wild turkey in portions of Fairfield and Skowhegan. Accessed online, December 2023, from
https://www.maine.gov/ifw/hunting-trapping/hunting/laws-rules/pfas-related-consumption-
advisory. html#:~:text=MDIFW%20and%20Maine%20CDC%20recommend, using%20the%20best%20
available%20data
Ml EGLE (Michigan Department of Environment, Great Lakes, and Energy). (2021a). Land application of
biosolids containing PFAS: Interim strategy (March 2021). Retrieved from
https://www.michigan.gov/egle/-
/media/Proiect/Websites/egle/Documents/Programs/WRD/Biosolids/PFAS-Biosolids-Strategy.pdf
Ml EGLE (Michigan Department of Environment, Great Lakes, and Energy). (2021b). Evaluation of PFAS
in Influent, Effluent, and Residuals of Wastewater Treatment Plants (WWTPs) in Michigan. Lansing,
Ml. Retrieved from https://www.michigan.gov/pfasresponse/-
/media/Proiect/Websites/egle/Documents/Programs/WRD/IPP/pfas-initiatives-statewide-full-
report.pdf?rev=946d5956f41b45a2a00d752149f57dc0&hash=E980FD486D131EA41E7BF55149929A
65
Ml EGLE (Michigan Department of Environment, Great Lakes, and Energy). (2021c). Evaluation of Wixom
Wastewater Treatment Plant (WWTP) Biosolids Land Application Sites 02N05E01-BC01 & BC02,
02N05E02-BC01 & AG01, 03N06E04-JW01 & JW05. Livingston County, Michigan. Lansing Ml.
DRAFT
132
-------
Retrieved from https://www.michigan.gov/egle/-
/media/Proiect/Websites/egle/Documents/Programs/WRD/Biosolids/PFAS-Biosolids-Field-Report-F-
Wixom-WWTP.pdf
Ml EGLE (Michigan Department of Environment, Great Lakes, and Energy). (2022). Land application of
biosolids containing PFAS: Interim strategy (Updated April 2022). Retrieved from
https://www.michigan.gov/egle/-
/media/Proiect/Websites/egle/Documents/Programs/WRD/Biosolids/PFAS-Biosolids-lnterim-
Strategy-2022.pdf
Milinovic J., Lacorte S., Rigol A., & Vidal M. (2016). Sorption of perfluoroalkyl substances in sewage
sludge. Environmental Science and Pollution Research International, 23, 9, 8339-48.
doi:10.1007/sll356-015-6019-9
Milinovic, J., Lacorte, S., Vidal, M., & Rigol, A. (2015). Sorption behaviour of perfluoroalkyl substances in
soils. Science of The Total Environment, 511:63-71. https://doi.Org/10.1016/i.scitotenv.2014.12.017
Moavenzadeh Ghaznavi, S., Zimmerman, C., Shea, M.E., MacRae, J.D., Peckenham, J.M., Noblet, C.L.,
Apul, O.G. & Kopec, A.D. (2023). Management of per- and polyfluoroalkyl substances (PFAS)-laden
wastewater sludge in Maine: Perspectives on a wicked problem. Biointerphases, 18, 041004.
https://doi.Org/10.1116/6.0002796
Moodie, D., Coggan, T., Berry, K., Kolobaric, A., Fernandes, M., Lee, E., Reichman, S., Nugegoda, D. &
Clarke, B.O. (2021). Legacy and emerging per- and polyfluoroalkyl substances (PFASs) in Australian
biosolids. Chemosphere, 270, 129143. https://doi.Org/10.1016/i.chemosphere.2020.129143
MPART (Michigan PFAS Action Response Team). (2023). Wixom WWTP Biosolids Fields Area of Interest
(Wixom, Livingston County). Accessed December 2023 from
https://www.michigan.gov/pfasresponse/investigations/sites-aoi/livingston-county/wixom-wwtp-
biosolids-fields-area-of-interest
MPCA (Minnesota Pollution Control Agency). (2020). Human health protective water quality criteria for
PFOS (wq-s6-61a). Retrieved from: https://www.pca.state.mn.us/business-with-us/site-specific-
water-quality-criteria
(Minnesota Pollution Control Agency). (2023a). PFAS in Fish. Accessed online, December 2023, from
https://www.pca.state.mn.us/air-water-land-climate/pfas-in-fish.
MPCA (Minnesota Pollution Control Agency). (2023b). Water Quality Standards: Human Health
Protective Water Quality Criteria for Per- and Polyfluoroalkyl Substances (PFAS). Retrieved from
https://www.pca.state.mn.us/business-with-us/site-specific-water-quality-criteria
Munoz, G., Michaud, A. M., Liu, M., Duy, S.V., Montenach, D., Resseguier, C., Watteau, F., Sappin-Didier,
V., Feder, F., Morvan, T., Houot, S., Desrosiers, M., Liu, J., & Sauve S. (2022). Target and nontarget
screening of PFAS in biosolids, composts, and other organic waste products for land application in
France. Environmental Science & Technology, 56 (10), 6056-6068.
https://doi.org/10.1021/acs.est.lcQ3697
Nassazzi, W., Wu, T-C., Jass, J., Lai, F. Y. & Ahrens, L. (2023). Phytoextraction of per- and polyfluoroalkyl
substances (PFAS) and the influence of supplements on the performance of short-rotation crops.
Environmental Pollution, 333, 122038. https://doi.Org/10.1016/i.envpol.2023.122038.
Navarro, I., de la Torre, A., Sanz, P., Pro, J., Carbonell, G.& Martinez, M. (2016). Bioaccumulation of
emerging organic compounds (perfluoroalkyl substances and halogenated flame retardants) by
DRAFT
133
-------
earth worm in biosolid amended soils. Environmental Research, 149, 32-39.
https://doi.Org/10.1016/j.envres.2016.05.004
NCDHHS (North Carolina Department of Health and Human Services). (2023). NCDHHS Recommends
Limiting Fish Consumption from the Middle and Lower Cape Fear River Due to Contamination with
"Forever Chemicals." Accessed online, December 2023, from https://www.ncdhhs.gov/news/press-
releases/2023/07/13/ncdhhs-recommends-limiting-fish-consumption-middle-and-lower-cape-fear-
river-due-contamination#:~:text=No%20more%20than%201%20meal%20per%20year%2C.-
combined%20across%20all&text=Levels%20of%20PFQS%20were%20higher.recent%20data%20from
%20the%20USEPA
Newell, C.J., Hopkins, L.P. & Bedient, P.B. (1990). A Hydrogeologic Database for Ground-Water
Modeling. Groundwater. 28, 5, 703-714.
Newell, C.J., Hopkins, L.P., & Bedient, P.B. (1989). Hydrogeologic Database for Ground Water Modeling.
API Publication No. 4476. American Petroleum Institute, Washington, DC.
Newstead, J.L., Beach, S.A., Gallagher, S.P. & Giesy, J.P. (2006). Pharmacokinetics and Acute Lethality of
Perfluorooctanesulfonate (PFOS) to Juvenile Mallard and Northern Bobwhite. Arch. Environ.
Contam. Toxicol., 50, 411-420. https://doi.org/10.1007/sQ0244-005-1137-x
Newstead, J.L., Coady, K.K., Beach, S.A., Butenhoff, J.L., Gallagher, S., & Giesy, J.P. (2007). Effects of
perfluorooctane sulfonate on mallard and northern bobwhite quail exposed chronically via the diet.
Environmental Toxicology and Pharmacology, 23, 1-9. https://doi.Org/10.1016/i.etap.2006.04.008
Newsted, J. L., Beach, S. A., Gallagher, S. P., & Giesy, J. P. (2006). Pharmacokinetics and Acute Lethality
of Perfluorooctanesulfonate (PFOS) to Juvenile Mallard and Northern Bobwhite. Archives of
Environmental Contamination and Toxicology, 50, 3, 411-420. https://doi.org/10.1007/sQ0244-005-
1137-x
Numata, J., Kowalczyk J., Adolphs, J., Ehlers, S., Schafft, H., Fuerst, P., Muller-Graf, C., Lahrssen-
Widerholt, M., & Matthias, G. (2014). Toxicokinetics of Seven Perfluoroalkyl Sulfonic and Carboxylic
Acids in Pigs Fed a Contaminated Diet. Journal of Agricultural and Food Chemistry, 62, 6861-6870.
Dx.doi.org/10.1021/jf405827u
OECD (Organisation for Economic Co-operation and Development). (2002). Hazard Assessment of
Perfluorooctane Sulfonate (PFOS) and its Salts. ENV/JM/RD(2002)17/FINAL. Report of the
Environment Directorate, Joint Meeting of the Chemicals Committee and the Working Party on
Chemicals, Pesticides and Biotechnology, Co-operation on Existing Chemicals, Paris, November 21,
2002.
OECD (Organisation for Economic Co-operation and Development). (2021). Reconciling terminology of
the universe of per- and polyfluoroalkyl substances: recommendations and practical guidance.
Retrieved from https://www.oecd.org/chemicalsafetv/portal-perfluorinated-chemicals/terminology-
per-and-polvfluoroalkyl-substances.pdf
Oliver, D.P., Li, Y., Orr, R., Nelson, P., Barnes, M., McLaughlin, M. J., & Kookana, R.S. (2019). The role of
surface charge and pH changes in tropical soils on sorption behaviour of per- and polyfluoroalkyl
substances (PFASs). Science of The Total Environment, 673, 197-206.
https://doi.Org/10.1016/i.scitotenv.2019.04.055
Pereira, H., Ullberg, M., Kleja, D.B., Gustafsson, J.P. & Ahrens, L. (2018). Sorption of perfluoroalkyl
substances (PFASs) to an organic soil horizon - Effect of cation composition and pH. Chemosphere,
207:183-191.https://doi.org/10.1016/i.chemosphere.2018.05.012
DRAFT
134
-------
Place, B.J., & Field, J.A. (2012). Identification of Novel Fluorochemicals in Aqueous Film-Forming Foams
Used by the US Military. Environmental Science & Technology, 46 (13):7120-
7127.10.1021/es301465n
Poultry Extension. (2024). Feeding chickens for egg production in small and backyard flocks. Retrieved
from: https://poultrv.extension.org/articles/feeds-and-feeding-of-poultry/feeding-chickens-for-egg-
production/
Prevedouros K, Cousins, I.T., Buck, R.C., & Korzeniowski, S.H. (2006). Sources, fate and transport of
perfluorocarboxylates. Environmental Science and Technology, 40, 1, 32-44.
Princz, J., Jatar, M., Lemieux, H. & Scroggins, R. (2018). Perfluorooctane sulfonate in surface soils: Effects
on reproduction in the collembolan, Folsomia Candida, and the oribatid mite, Oppia nitens.
Chemosphere, 208, 757-763. https://doi.Org/10.1016/i.chemosphere.2018.06.020
Rankin, K., Mabury, S.A., Jenkins, T. M., & Washington, J.W. (2016). A North American and global survey
of perfluoroalkyl substances in surface soils: Distribution patterns and mode of occurrence.
Chemosphere, 161, 333-341. https://doi.Org/10.1016/i.chemosphere.2016.06.109.
Santangelo, L. M., Tokranov A.K., Welch, S.M., Schlosser, K.E.A., Marts, J.M., Drouin, A.F., Ayotte, J.D.,
Rousseau, A.E., & Harfmann, J.L. (2022). Statewide survey of shallow soil concentrations of per- and
polyfluoroalkyl substances (PFAS) and related chemical and physical data across New Hampshire,
2021. US Geological Survey data release. https://doi.org/10.5066/P9KG38B5
Santangelo, L. M., Welch, S.M., Tokranov, A.K., Drouin, A.F., Schlosser, K.E.A., Marts, J.M., Lincoln, T.A.,
Deyette, N.A., & Perkins, K. (2023). Field-scale investigation of per- and polyfluoroalkyl substances
(PFAS) leaching from shallow soils to groundwater at two sites in New Hampshire, 2021-2022. US
Geological Survey data release. https://doi.org/10.5066/P92C21F6
Schaefer, C. E., Hooper, J., Modiri-Gharehveran, M., Drennan, D.M., Beecher, N., & Lee, L. (2022).
Release of poly- and perfluoroalkyl substances from finished biosolids in soil mesocosms. Water
Research, 217, 118405. https://doi.Org/10.1016/i.watres.2022.118405
Schaefer, C., Hopper, K.J.L., Strom, L.E., Abusallout, I., Dickenson, E.R.V., Thompson, K.A., Ram Mohan,
G., Drennan, D., Wu, K., & Guelfo, J.L. (2023). Occurrence of quantifiable and semi-quantifiable
poy- and perfluoroalkyl substances in United States wastewater treatment plants. Water
Research, 233, 119724. https://doi.Org/10.1016/i.watres.2023.119724
Scher, D.P., Kelly, J.E., Huset, C.A., Barry, K.M., Hoffbeck, R.W., Yingling, V.L. & Messing, R.B. (2018).
Occurrence of perfluoroalkyl substances (PFAS) in garden produce at homes with a history of PFAS-
contaminated drinking water. Chemosphere, 196, 548-555.
Schroeder, P.R., Aziz, N.M., Lloyd, C.M., & Zappi, P.A. (1994). The Hydrologic Evaluation of Landfill
Performance (HELP) Model. User's Guide for Version 3. EPA/600/R-94/168a. US EPA Risk Reduction
Engineering Laboratory. Cincinnati, Ohio, 84 p. and appendix.
Schultz M.M., Barofsky D.F., & Field, J.A. (2003). Fluorinated alkyl surfactants. Environ Eng Sci, 20(5):487-
501.
Sepulvado, J. G., Blaine, A.C., Hunda, L.S., & Higgins, C.P. (2011). Occurrence and fate of
perfluorochemicals in soil following the land application of municipal biosolids. Environmental
Science and Technology, 45, 8106-8112. https://doi.org/10.1021/esl03903d
DRAFT
135
-------
Sharifan, H., Bagheri, M., Wang, D., Burken, J.G., Higgins, C.P., Liang, Y., Liu, J., Schaefer, C.E., Blotevogel,
J. (2021). Fate and transport of PFAS in the vadose zone. Science of the Total Environment, 771,
145427. https://doi.Org/10.1016/j.scitotenv.2021.145427
Shearer, J.J., Callahan, C.L., Calafat, A.M., Huang, W.Y., Jones, R.R., Sabbisetti, V.S., Freedman, N.D.,
Sampson, J.N., Silverman, D.T., Purdue, M.P., & Hofmann J.N. (2021). Serum concentrations of per-
and polyfluoroalkyl substances and risk of renal cell carcinoma. Journal of the National Cancer
Institute, 113, 580-587. http://dx.doi.org/10.1093/inci/djaal43.
Shields, E. P., Krug, J.D., Roberson, W.R., Jackson, S.R., Smeltz, M.G., Allen, M.R., Burnette, P., Nash, J.T.,
Virtaranta, L., Preston, W., Liberatore, H.K., Wallace, M.A.G., Ryan, J.V., Kariher, P.H., Lemieux, P.M.
& Linak, W.P. (2023). Pilot-scale thermal destruction of per- and polyfluoroalkyl substances in a
legacy aqueous film forming foam. ACS ES&T Engineering, 9, 1308-1317
https://doi.org/10.1021/acsestengg.3cQ0098
Silva J.A.K., Guelfo, J. L., Simunek, J., & McCray, J.E. (2022). Simulated leaching of PFAS from land-applied
municipal biosolids at agricultural sites. Journal of Contaminant Hydrology, 251, 104089.
https://doi.Org/10.1016/i.iconhyd.2022.104089.
Sima, M.W. & Jaffe, P.R. (2021). A critical review of modeling poly- and perfluoroalkyl substances (PFAS)
in the soil-water environment. Science of the Total Environment, 757, 143793.
https://doi.Org/10.1016/i.scitotenv.2020.143793
Simones, T., Evans, C., Goossen, C., Kersbergen, R. & Smith, A. (2023). Summary of finds of a multiyear
study on PFAS uptake into grasses and corn plants on biosolid amended fields in Maine. Center for
PFAS Research Annual Symposium - Current Knowledge and Applications for Agricultural Production.
Poster retrieved December 2023 from https://www.canr.msu.edu/pfas-research/Annual-
Svmposium/Simones%20poster.pdf
Smith, M.T.; Guyton, K.Z.; Gibbons, C.F.; Fritz, J.M.; Portier, C.J.; Rusyn, I., DeMarini, D.M., Caldwell, J.C.,
Kavlock, R. J., Lambert, P.F., Hecht, S. S., Bucher, J. R., Stewart, B. W., Baan, R.A, Cogliano, V. J., &
Straif, K. (2016b). Key characteristics of carcinogens as a basis for organizing data on mechanisms of
carcinogenesis. Environmental Health Perspectives, 124, 713-721.
http://dx.doi.org/10.1289/ehp.1509912
SRC (Syracuse Research Corporation). 2016. PHYSPROP Database. Accessed May 2016.
http://www.srcinc.com/what-we-do/environmental/scientific-databases.html.
Thomford, PJ. (2002). 104-week dietary chronic toxicity and carcinogenicity study with perfluorooctane
sulfonic acid potassium salt (PFOS; T-6295) in rats (pp. 002148-002363). (Study No. 6329-183).
Madison, Wl: Covance Laboratories.
Thompson, J. T., Robey, N.M., Tolaymat, T.M., Bowden, J.A., Solo-Gabriele, H.M. & Townsend, T.G.
(2023a). Underestimation of per- and polyfluoroalkyl substances in biosolids: Precursor
transformation during conventional treatment. Environmental Science & Technology, 57, 9, 3825-
3832. https://doi.org/10.1021/acs.est.2cQ6189
Thompson, J.T., Chen, B., Bowden, J.A. & Townsend, T.G. (2023b). Per- and polyfluoroalkyl substances in
toilet paper and the impact on wastewater systems. Environmental Science & Technology
Letters, 10, 3, 234-239. https://doi.org/10.1021/acs.estlett.3cQ0094
Tokranov, A. K., Welch, S.M., Santangelo, L.M., Kent, D.B., Repert, D.A., Perkins, K., Bliznik, P.A., Roth,
D.A., Drouin, A.F., Lincoln, T.A., Deyette, N.A., Schlosser, K.E.A. & Marts, J.M. (2023). Solid/Water
Partitioning of per- and polyfluoroalkyl substances (PFAS) in New Hampshire soils and biosolids:
DRAFT
136
-------
Results from laboratory experiments at the US Geological Survey. US Geological Survey data release.
https://doi.org/10.5066/P9TKSM8S
UNCE (University of Nebraska Cooperative Extension). (1998). Water Quality and Requirements for Dairy
Cattle. Cooperative Extension, Institute of Agriculture and Natural Resources, University of
Nebraska-Lincoln.
UNEP (United Nations Environmental Program). (2015). Proposal to List Pentadecafluorooctanoic Acid
(CAS No: 335-67-1, PFOA, Perfluorooctanoic Acid), its Salts and PFOA-related Compounds in Annexes
A, B and/or C to the Stockholm Convention on Persistent Organic Pollutants.
UNEP/POPS/POPRC.11/5. Persistent Organic Pollutants Review Committee, Eleventh meeting Rome,
Italy, October 19-23, 2015. Accessed May 2016. http://www.bennington.edu/docs/default-
source/docs-pfoa/un-stockholm-convention-review-of-pfoa-(2016).pdf?sfvrsn=2.
USEPA (US Environmental Protection Agency). 1986. Guidelines for the Health Risk Assessment of
Chemical Mixtures. Federal Register 51(185):34014-34025.
US EPA (US Environmental Protection Agency). (1992). Technical Support Document for Land Application
of Sewage Sludge. Office of Water, Office of Science and Technology. EPA 822/R-93-001a.
US EPA (US Environmental Protection Agency). (1995a). A Guide to the Biosolids Risk Assessments for
the EPA Part 503 Rule. EPA/832-B-93-005. https://www.epa.gov/sites/default/files/2018-
ll/documents/guide-biosolids-risk-assessments-part503.pdf
US EPA (US Environmental Protection Agency). (1995b). Chapter 2: Solid waste disposal. In Compilation
of Air Pollutant Emission Factors (AP-42), Fifth Edition (Vol. 1).
https://www3.epa.gov/ttn/chief/ap42/ch02/index.html
US EPA (US Environmental Protection Agency). (1996). Technical Support Document for the Round Two
Sewage Sludge Pollutants. EPA-822-R-96-003.
US EPA (US Environmental Protection Agency). (1999). Source Modules for Non-Wastewater Waste
Management Units (Land Application Units, Waste Piles, and Landfills); Background and
Implementation for The Multimedia, Multipathway, and Multireceptor Risk Assessment (3MRA) for
HWIR99. Office of Solid Waste. Washington, DC. October. Available at
https://archive.epa.gov/epawaste/hazard/web/pdf/s0056.pdf
US EPA (US Environmental Protection Agency). (2000a). Methodology for deriving ambient water quality
criteria for the protection of human health (2000). EPA-822-B-00-004. Office of Water, Washington,
DC. https://www.epa.gov/sites/default/files/2018-10/documents/methodology-wqc-protection-hh-
2000.pdf
US EPA (US Environmental Protection Agency). (2000b). Biosolids Technology Fact Sheet: Land
Application of Biosolids. EPA 832-F-00-064. Office of Water, Washington, DC. September.
US EPA (US Environmental Protection Agency). (2002). Perfluoroalkyl Sulfonates; Significant New Use
Rule. EPA. Federal Register Dec 9, 2002, 67:72854.
US EPA (US Environmental Protection Agency). (2003a). Exposure Analysis for Dioxins, Dibenzofurans,
and Coplanar Polychlorinated Biphenyls in Sewage Sludge. Technical Background Document.
https://www.epa.gov/biosolids/exposure-analvsis-dioxins-dibenzofurans-and-coplanar-
polvchlorinated-biphenvls-sewage
US EPA (US Environmental Protection Agency). (2003b). Technical Background Document for the Sewage
Sludge Exposure and Hazard Screening Assessment. US EPA, Office of Water, Washington, DC. EPA-
DRAFT
137
-------
822-B-03-001. December 19. Retrieved from: https://www.epa.gov/sites/default/files/2018-
11/documents/tech-background-doc-sewage-sludge-hazard.pdf.
US EPA (US Environmental Protection Agency). (2003c). Final Action Not to Regulate Dioxins in Land-
Applied Sewage Sludge. EPA-822-F-03-007. October, 2003. Retrieved from:
https://www.epa.gov/sites/default/files/2018-ll/documents/final-action-regulate-dioxins-
factsheet.pdf
US EPA (US Environmental Protection Agency). (2003d). EPA's Composite Model for Leachate Migration
with Transformation Products (EPACMTP): Parameters/Data Background Document. Washington,
DC. https://www.epa.gov/sites/production/files/2016-03/documents/v2 pd.pdf
US EPA (US Environmental Protection Agency). (2003e). EPACMTP: Technical Background Document.
Washington, DC. https://www.epa.gov/sites/production/files/2016-03/documents/v2 tbd.pdf
US EPA (US Environmental Protection Agency). (2003f). Multimedia, Multipathway, and Multireceptor
Risk Assessment (3MRA) Modeling System, Volume I: Modeling System and Science. EPA530-D-03-
001a. Washington, DC. httpsV/archive.epa.gov/epawaste/hazard/web/pdf/l toc.pdf,
httpsV/archive.epa.gov/epawaste/hazard/web/pdf/l 05 land.pdf,
httpsV/archive.epa.gov/epawaste/hazard/web/pdf/l 08 srfcwtr.pdf
US EPA (US Environmental Protection Agency). (2003g). Multimedia, Multipathway, and Multireceptor
Risk Assessment (3MRA) Modeling System, Volume II: Data Collection. EPA530-D-03-001b.
Washington, DC.
US EPA (US Environmental Protection Agency). (2003h). Methodology for Deriving Ambient Water
Quality Criteria for the Protection of Human Health (2000) Technical Support Document Volume 2:
Development of National Bioaccumulation Factors. EPA-822-R-03-030. Available from:
https://www.epa.gov/sites/default/files/2018-10/documents/methodology-wqc-protection-hh-
2000-volume2.pdf
US EPA (US Environmental Protection Agency). (2005a). Guidelines for Carcinogen Risk Assessment.
EPA/630/P-03/001B. US Environmental Protection Agency, Washington, DC. Accessed May 2016.
https://www.epa.gov/risk/guidelines-carcinogen-risk-assessment.
US EPA (US Environmental Protection Agency). (2005b). Human Health Risk Assessment Protocol for
Hazardous Waste Combustion Facilities. EPA530-R-05-006. Washington, DC. https://epa-
prgs.ornl.gov/radionuclides/2005 HHRAP.pdf
US EPA (US Environmental Protection Agency). (2007). Perfluoroalkyl Sulfonates; Significant New Use
Rule. EPA. Federal Register Oct 9, 2007, 72:57222.
US EPA (US Environmental Protection Agency). (2008). Perfluorooctanoic acid (PFOA) and fluorinated
telomers. US Environmental Protection Agency, http://www.epa.gov/oppt/pfoa/. May 29, 2008
US EPA (US Environmental Protection Agency). (2011). EP"s Exposure Factors Handbook (EFH). Retrieved
from: https://www.epa.gov/expobox/about-exposure-factors-handbook
US EPA (US Environmental Protection Agency). (2013). Certain perfluoroalkyl sulfonates; Significant New
Use Rule. EPA. Federal Register Dec. 9, 2002, 67:72863, as amended Oct 9, 2007, 72:57229; Oct 22,
2013, 78:62450.
US EPA (US Environmental Protection Agency). (2014). Estimated Fish Consumption Rates for the US
Population and Selected Subpopulations (NHANES 2003-2010). EPA-820-R-14-002.
DRAFT
138
-------
US EPA (US Environmental Protection Agency). (2015). Significant New Use Rules: Long Chain
Perfluoroalkyl Carboxylate and Perfluoroalkyl Sulfonate Chemical Substances. US EPA, Office of
Water. Federal Register, Jan 21, 2015, 80:2885.
US EPA (US Environmental Protection Agency). (2016). Final emissions and unit inventory for federal
plan requirements for sewage sludge incineration units constructed on or before October 14, 2010.
https://www.regulations.gov/document?D=EPA-HQ-QAR-2012-0319-0020
US EPA (US Environmental Protection Agency). (2017). Update for Chapter 5 of the Exposure Factors
Handbook: Soil and Dust Ingestion. EPA/600/R-17/384F. National Center for Environmental
Assessment, Office of Research and Development, Washington, DC. September. Available at
https://www.epa.gov/expobox/about-exposure-factors-handbook.
US EPA (US Environmental Protection Agency). (2018a). PFAS Explained. EPA website. Accessed April
2022. https://www.epa.gov/pfas/pfas-explained.
US EPA (US Environmental Protection Agency). (2018b). Update for Chapter 11 of the Exposure Factors
Handbook: Intake of Meats, Dairy Products, and Fats. EPA/600/R-17/485F. National Center for
Environmental Assessment, Office of Research and Development, Washington, DC. April. Available
at https://www.epa.gov/expobox/about-exposure-factors-handbook.
US EPA (US Environmental Protection Agency). (2019b). The Variable Volume Water Model, Revision B,
USEPA/OPP 734S16002. Office of Pesticide Programs, Washington, DC.
https://www.epa.gov/pesticide-science-and-assessing-pesticide-risks/models-pesticide-risk-
assessment
US EPA (US Environmental Protection Agency). (2019c). Update for Chapter 3 of the Exposure Factors
Handbook: Ingestion of Water and Other Select Liquids. EPA/600/R-18/259F. National Center for
Environmental Assessment, Office of Research and Development, Washington, DC. February.
Available at https://www.epa.gov/expobox/about-exposure-factors-handbook
U.S. EPA (Environmental Protection Agency). 2020. Pesticide in Water Calculator (PWC), Version 2.00.
U.S. Environmental Protection Agency, Office of Pesticide Programs, Washington DC. September.
Available at https://www.epa.gov/pesticide-science-and-assessing-pesticide-risks/models-pesticide-
risk-assessment
US EPA (US Environmental Protection Agency). (2021a). PFAS Strategic Roadmap: EPA's Commitments to
action 2021-2024. EPA -100-K-21-002. Retrieved from
https://www.epa.gov/system/files/documents/2021-10/pfas-roadmap final-508.pdf
US EPA (US Environmental Protection Agency). (2021b). Toxic substances control act reporting and
recordkeeping requirements for perfluoroalkyl and polyfluoroalkyl substances.
https://www.regulations.gov/document/EPA-HQ-OPPT-2020-0549-0001
US EPA (US Environmental Protection Agency). (2021c). Compliance Guide for Imported Articles
Containing Surface Coatings Subject to the Long-Chain Perfluoroalkyl Carboxylate and Perfluoroalkyl
Sulfonate Chemical Substances Significant New Use Rule. Document ID No.: RIN 2070-ZA23.
Accessed April 2022. https://www.epa.gov/sites/default/files/2021-01/documents/final Icpfac-
snur surfacecoating-compliance-guide O.pdf.
US EPA. (2022). ORD staff handbook for developing IRIS assessments. (EPA 600/R-22/268). Washington,
DC: U.S. Environmental Protection Agency, Office of Research and Development, Center for Public
Health and Environmental Assessment.
https://cfpub.epa.gov/ncea/iris drafts/recordisplay.cfm?deid=356370
DRAFT
139
-------
US EPA (US Environmental Protection Agency). (2023a). Sewage Sludge Incineration Units (SSI): New
Source Performance Standards (NSPS) and Emission Guidelines (EG). Retrieved in December 2023
from https://www.epa.gov/stationary-sources-air-pollution/sewage-sludge-incineration-units-ssi-
new-source-performance
US EPA (US Environmental Protection Agency). (2023b). Research for Understanding PFAS Uptake and
Bioaccumulation in Plant and Animals in Agricultural, Rural, and Tribal Communities Request for
Applications (RFA). Retrieved December 2023 from https://www.epa.gov/research-grants/research-
understanding-pfas-uptake-and-bioaccumulation-plant-and-animals
US EPA (US Environmental Protection Agency). (2023c). A Standardized Framework for Sewage Sludge
Chemical Risk Assessment. 822D23001.
US EPA (US Environmental Protection Agency). (2023d). Biosolids Tool (BST) appendixes. Retrieved from:
https://casac.epa.g0v/0rds/sab/r/sab apex/sabl05/0?mm id=6174&request=APPLICATION PROCE
SS%3DMEETING FILE&session=5051460399579
US EPA (US Environmental Protection Agency). (2024a). Risk Management for Per- and Polyfluoroalkyl
Substances (PFAS) under TSCA (website). Accessed January, 2024 from
https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/risk-management-and-
polyfluoroalkyl-substances-pfas
US EPA (US Environmental Protection Agency). (2024b). Office of Water Final Human Health Toxicity
Assessment for Perfluorooctanoic Acid (PFOA). 815R24006.
US EPA (US Environmental Protection Agency). (2024c). Office of Water Final Human Health Toxicity
Assessment for Perfluorooctane Sulfonic Acid (PFOS). 815R24007.
US EPA (US Environmental Protection Agency). (2024d). CWA Analytical Methods for Per- and
Polyfluorinated Alkyl Substances (PFAS). https://www.epa.gov/cwa-methods/cwa-analytical-
methods-and-polyfluorinated-alkyl-substances-pfas
US EPA (US Environmental Protection Agency). (2024e). Framework for Estimating Noncancer Health
Risks Associated with Mixtures of Per- and Polyfluoroalkyl Substances (PFAS). EPA-815-R-24-003.
Retrieved from: https://www.epa.gov/system/files/documents/2024-04/final-pfas-mix-framework-
3.25.24 final-508.pdf
US EPA (US Environmental Protection Agency). (2024f). Other Test Method 50 (OTM-50) Sampling and
Analysis of Volatile Fluorinated Compounds from Stationary Sources Using Passivated Stainless-Steel
Canisters, https://www.epa.gov/system/files/documents/2024-01/otm-50-release-l O.pdf
US EPA (US Environmental Protection Agency). (2024g). Interim Guidance on Destroying and Disposing
of Certain PFAS and PFAS-Containing Materials That Are Not Consumer Products. Retrieved from:
https://www.epa.gov/system/files/documents/2024-04/2024-interim-guidance-on-pfas-
destruction-and-disposal.pdf
US EPA (US Environmental Protection Agency). (2024h). 2020 Great Lakes Human Health Fish Fillet
Tissue Study. Retrieved from: https://www.epa.gov/choose-fish-and-shellfish-wiselv/2020-great-
lakes-human-health-fish-fillet-tissue-studv
US EPA (US Environmental Protection Agency). (2024i). Regional Screening Level (RSL) Resident Soil
Table (TR=lE-06, HQ=1) May 2024. Retrieved from: https://semspub.epa.gov/work/HQ/404467.pdf
US EPA (US Environmental Protection Agency). (2024j). Requirements For Application of Domestic
Septage to Agricultural Land Utilizing Injection or Incorporation. Retrieved from:
DRAFT
140
-------
https://www.epa.gov/svstem/files/documents/2024-Q4/requirements-application-domestic-
septage-iniection-2024.pdf
US EPA (US Environmental Protection Agency). (2024k). Requirements For Application of Domestic
Septage to Agricultural Land Utilizing Alkali Addition. Retrieved from:
https://www.epa.gov/svstem/files/documents/2024-Q4/requirements-application-domestic-
septage-alkali-2024.pdf
US EPA (US Environmental Protection Agency). (20241). Final Aquatic Life Ambient Water Quality Criteria
for Perfluorooctanoic Acid (PFOA). EPA-822-R-23-001. https://www.epa.gov/wqc/aquatic-life-
criteria-perfluorooctanoic-acid-pfoa
US EPA (US Environmental Protection Agency). (2024m). Final Aquatic Life Ambient Water Quality
Criteria for Perfluorooctane Sulfonate (PFOS). EPA-822-R-23-002.
https://www.epa.gov/wqc/aquatic-life-criteria-perfluorooctane-sulfonate-pfos
US EPA (US Environmental Protection Agency). (2024n). EPA Response to External Peer Review
Comments on the Draft Sewage Sludge Risk Assessment for Perfluorooctanoic Acid (PFOA) CASRN
335-67-1 and Perfluorooctane Sulfonic Acid (PFOS) CASRN 1763-23-1. 820R25002.
US EPA (US Environmental Protection Agency). (2024o). Draft Human Health Life Ambient Water Quality
Criteria for Perfluorooctane Sulfonate (PFOS).
US EPA (US Environmental Protection Agency). (2024p). Draft Human Health Life Ambient Water Quality
Criteria for Perfluorooctanoic Acid (PFOA).
USDA (US Department of Agriculture). 1994. State Soil Geographic (STATSGO) Data Base. Data use
information. Miscellaneous Publication Number 1492. Natural Resources Conservation Service, Fort
Worth, TX. December.
USDA (US Department of Agriculture). 2014. 2012 Census of Agriculture. National Agricultural Statistics
Service ACH12-2/May 2014. Retrieved from:
https://www.nass.usda.gov/Publications/Highlights/2014/Farm Economics/Highlights Farm Econo
mics.pdf
van der Veen, I., Schellenberger, S., Hanning, A., Stare, A., de Boer, J., Weiss, J.M., & Leonards, P.E.G.
(2022). Fate of Per- and Polyfluoroalkyl Substances from Durable Water-Repellent Clothing during
Use. Environmental Science & Technology, 56, 5886-5897,
van Genuchten, M.Th. (1980). A closed-form equation for predicting the hydraulic conductivity of
unsaturated soils, Soil Sci. Soc. J., v. 44, pp. 892-898.
Venkatesan, A.K. & Halden, R.U. (2013). National inventory of perfluoroalkyl substances in archived US
biosolids from the 2001 EPA National Sewage Sludge Survey. Journal of Hazardous Materials, 252-
253, 413-418. https://doi.Org/10.1016/i.ihazmat.2013.03.016
Vestergren, R., Orata, F., Berger, U. & Cousins, I.T. (2013). Bioaccumulation of perfluoroalkyl acids in
dairy cows in a naturally contaminated environment. Environmental Science Pollution Research, 20,
7959-7969. DOI 10.1007/sll356-013-1722-x
Wang, M., Chen, J., Lin, K., Chen, Y., Hu, W., Tanguay, R. L., Huang, C.& Dong, Q. (2011a). Chronic
zebrafish PFOS exposure alters sex ratio and maternal related effects in Floffspring. Environmental
Toxicology and Chemistry. 30, 9, 2073-2080. 10.1002/etc.594
Washington, J. W., & Jenkins, T. M. (2015). Abiotic hydrolysis of fluorotelomer polymers as a source of
perfluorocarboxylates at the global scale. Environmental Science & Technology, 49, 14129-14135.
DRAFT
141
-------
Washington, J. W., Yoo, H., Ellington, J. J., Jenkins, T. M., & Libelo, L. (2010). Concentrations, distribution,
and persistence of perfluoroalkylates in sludge-applied soils near Decatur, Alabama, USA.
Environmental Science and Technology, 44, 8390-8396. https://doi.org/10.1021/esl003846
Washington, J.W., Jenkins, T.M., Rankin, K., 7 Naile, J.E. (2015). Decades-Scale Degradation of
Commercial, Side-Chain, Fluorotelomer-based Polymers in Soils & Water. Environmental Science &
Technology, 49, 915-923.
Wen, B., Wu, Y., Zhang, H., Liu, Y., Hu, X., Huang, H., & Zhang, S. (2016). The roles of protein and lipid in
the accumulation and distribution of perfluorooctane sulfonate (PFOS) and perfluorooctanoate
(PFOA) in plants grown in biosolids-amended soils. Environmental Pollution, 216, 682-688.
https://doi.Org/10.1016/i.envpol.2016.06.032.
Wikstrom, S., Lin, P.I., Lindh, C.H., Shu, H., & Bornehag, C.G. (2020). Maternal serum levels of
perfluoroalkyl substances in early pregnancy and offspring birth weight. Pediatric Research, 87,
1093-1099.
Wilson, T.B., Stevenson, G., Crough, R., de Araujo, J., Fernando, N., Anwar, A., Scott, T., Quinteros, J.A.,
Scott, P.C. & Archer, M.J.G. (2020). Evaluation of Residues in Hen Eggs After Exposure of Laying Hens
to Water Containing Per- and Polyfluoroalkyl Substances. Environmental Toxicology, 40 (3), 735-743.
DOI: 10.1002/etc.4723
Winchell, L.J., Wells, M.J.M., Ross, J.J., Kakar, F., Teymouri, A., Gonzalez, D. J., Dangtran, K., Bessler, S.M.,
Carlson, S., Almansa, X.F., Norton, J.W., & Bell, K.Y. (2024). Fate of perfluoroalkyl and polyfluoroalkyl
substances (PFAS) through two full-scale wastewater sludge Incinerators. Water Environment
Research, 96, ell009. https://doi.org/10.1002/wer.110Q9.
Xiao, K., Li, X., Xu, N., Wang, X., Hao, L., Bao, H., Zhang, L., Shi, Y., & Cai, Y. (2024). Carry-over rate of per-
and polyfluoroalkyl substances to raw milk and human exposure risks in different regions of China.
Science of the Total Environment, 944, 173902. https://doi.Org/10.1016/j.scitotenv.2024.173902.
Yoo, H., Washington, J.W., Jenkins, T.M., & Ellington, J.J. (2011). Quantitative determination of
perfluorochemicals and fluorotelomer alcohols in plants from biosolid-amended fields using
LC/MS/MS and GC/MS. Environmental Science and Technology, 45 (19), 7985-7990.
https://doi.org/10.1021/eslQ2972m
Yoo, H., Washington, J.W., Jenkins, T.M., & Libelo, E.L. (2009). Analysis of perfluorinated chemicals in
sludge: Method development and initial results. Journal of Chromatography A, 1216, 7831-7839.
https://doi.Org/10.1016/i.chroma.2009.09.051
Yoo, H., Washington, J.W., Ellington, J.J., Jenkins, T.M., & Neil, M.P. (2010). Concentrations, distribution,
and persistence of fluorotelomer alcohols in sludge-applied soils near Decatur, Alabama, USA.
Environmental Science and Technology, 44, 8397-8402.
Yuan, Z., Zhang, J., Zhao, L., Li, J. & Liu, H. (2017). Effects of perfluorooctanoic acid and perfluorooctane
sulfonate on acute toxicity, superoxide dismutase, and cellulase activity in the earthworm Eisenia
fetida. Environmental Science and Pollution Research International, 24, 22, 18188-18194.
https://doi.org/10.1007/sll356-017-9477-4.
Zanden, J.M., Cabana, G., & Rasmussen, J.B. (1997). Comparing trophic position of freshwater fish
calculated using stable nitrogen isotope ratios (61SN) nad literature dietary data. Canadian Journal of
Fish Aquatic Science, 54, 1142-1158.
DRAFT
142
-------
Zeng, J. & Guo, B. (2021). Multidimensional simulation of PFAS transport and leaching in the vadose
zone: Impact of surfactant-induced flow and subsurface heterogeneities. Advances in Water
Resources, 155, 104015. https://doi.Org/10.1016/i.advwatres.2021.104015.
Zeng, J. & Guo, B. (2023). Reduced Accessible Air-Water Interfacial Area Accelerates PFAS Leaching in
Heterogeneous Vadose Zones. Geophysical Research Letters, 50, 8, e2022GL102655.
https://doi.org/10.1029/2022GL102655.
Zhang, C., Yan, H., Li, F., Hu, X. & Zhou, Q. (2013). Sorption of short- and long-chain perfluoroalkyl
surfactants on sewage sludges. Journal of Hazardous Materials, 260, 689-699.
https://doi.Org/10.1016/j.jhazmat.2013.06.022
Zhang, M., Wang, P., Lu, Y., Lu, X., Zhang, A., Liu, Z., ... & Sarvajayakesavalu, S. (2020). Bioaccumulation
and human exposure of perfluoroalkyl acids (PFAAs) in vegetables from the largest vegetable
production base of China. Environment international, 135, 105347.
Zhao, H., Chen, C., Zhang, X., Chen, J. & Quan, X. (2011). Phytotoxicity of PFOS and PFOAto Brassica
chinensis in different Chinese soils. Ecotoxicology and Environmental Safety, 74, 5, 1343-1347.
https://doi.Org/10.1016/i.ecoenv.2011.03.007
Zhou, L., Xia, M., Wang, L., & Mao, H. (2016). Toxic effect of perfluorooctanoic acid (PFOA) on
germination and seedling growth of wheat (Triticum aestivum L.). Chemosphere, 159, 420-425.
https://doi.Org/10.1016/i.chemosphere.2016.06.045
DRAFT
143
-------
PFOA/PFOS Risk Assessment
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
APPENDIX A. SUMMARY OF PFOA AND PFOS OCCURRENCE IN BIOSOLIDS IN THE US
This appendix presents concentration data compiled from published peer-reviewed literature and state reports that were available as of January
2024. Table A-l presents occurrence data for PFOA and Table A-2 provides occurrence data for PFOS. Table A-3 highlights recent studies of
PFOA and PFOS precursor occurrence.
A.l Occurrence of PFOA and PFOS
Table A-l. PFOA Occurrence in Biosolids in the US
Reference
Geographic
Area
Years
Sampled
Sample Type
PFOA Concentration
(Dry Weight Basis):
Range (Min-Max),
Mean, and/or Median
Method
Used
Notes
3M Environmental
Laboratory, 2001
US (Multiple
states)
1999-
2001
WWTP Biosolids
Range:
<17 ppb (4 WWTPs)
<244 ppb (Decatur Utilities Plant)
Modified
Sampled 6 test cities, including Decatur, AL (3M Multi-City
Study)
Higgins et al., 2005
US (Multiple
states)
1998-
2004
WWTP Biosolids
Range: n.d.-29.4 ppb
Modified
Digested sludge samples from 8 WWTPs and primary
settled solids from 3 WWTPs
(9 WWTPs in total)
Schultz et al., 2006
US (Pacific
Northwest
Region)
2004
WWTP Biosolids
Range:
Digested sludge: <3 ppb
Modified
Analyzed wastewater and sludge samples throughout the
treatment process
Sinclair and Kannan,
2006
US (New
York)
2005
WWTP Biosolids
Range:
Plant A
Plant B
Mean:
Plant A
Plant B
Median
Plant A
Plant B
69-241 ppb
18-89 ppb
144 ppb
70 ppb
134 ppb
80 ppb
Modified
Sampled wastewater at 6 WWTPs, two of which were also
sampled for biosolids (five times each)
Loganathan et al.,
2007
US (Kentucky
and Georgia)
2005
WWTP Biosolids
Range:
Plant A: 8.3-219 ppb
Plant B: 7-130 ppb
Modified
Sampled two WWTPs: rural (Plant A, Kentucky) and urban
(Plant B, Georgia)
DRAFT
A-l
-------
PFOA/PFOS Risk Assessment
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
Reference
Geographic
Area
Years
Sampled
Sample Type
PFOA Concentration
(Dry Weight Basis):
Range (Min-Max),
Mean, and/or Median
Method
Used
Notes
Yoo etal.,2009
US (Alabama
and New
York)
2007
WWTP Sewage
Sludge/ Biosolids
Mean:
Decatur WWTP:
Sample A: 50.3±4.7 ppb
Sample B: 128±8.3 ppb
New York City WWTPs:
Sample A: 8.7±0.7 ppb
Sample B: 8.4±5 ppb
Sample C: 20±3.9 ppb
Modified
Isotopic-
Dilution
Method with
LC-MS/MS
Conducted a method development study for measuring
PFAS, using sludge samples from a WWTP in Decatur,
AL; this method was then used to assess PFAS in a NIST
sludge sample and sludge samples from New York City
WWTPs
Washington et al.,
2010
US (Alabama)
2007 and
2009
Land-applied
Biosolids
Range:
2009: <320 ppb
Modified
Conducted two sampling surveys (2007 and 2009)
Lindstrom et al., 2011
US (Alabama)
2009
Well and Surface
Water near
Land-applied
Biosolids Sites
Range:
Well Water:
-------
PFOA/PFOS Risk Assessment
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
Reference
Geographic
Area
Years
Sampled
Sample Type
PFOA Concentration
(Dry Weight Basis):
Range (Min-Max),
Mean, and/or Median
Method
Used
Notes
Schaefer etal., 2022
US (Multiple
states)
2020
WWTP Biosolids
and Column
Mesocosm
Leaching
Experiments
Range: 0.8-8.12 ppb
Modified
Sampled 7 WWTPs with a variety of treatment processes
in urban areas, receiving both industrial and domestic
sources, as well as performed column mesocosm leaching
experiments; found that PFAS precursors accounted for
over 75% of total PFAS
Thompson etal.,
2023a
US (Florida)
2021
WWTP Sewage
Sludge/
Biosolids
Range:
Sludge (Before Treatment):
1.7-21 ppb
Biosolids (After Treatment):
1.1-7.7 ppb
Modified
Interviewed 39 facilities in Florida to learn treatment
processes from 2019-2021; Then, in 2021, collected 16
samples (before and after treatment) from 8 facilities
representing the four most common treatment processes;
studied 92 PFAS analytes, including precursors
Thompson etal.,
2023b
US (Florida)
Sludge:
2021
Toilet
Paper:
2021-
2022
WWTP Sewage
Sludge/
Biosolids and
Toilet Paper
Range:
Sludge: 1.7-21 ppb
Toilet Paper:
-------
PFOA/PFOS Risk Assessment
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
Reference
Geographic
Area
Years
Sampled
Sample Type
PFOA Concentration
(Dry Weight Basis):
Range (Min-Max),
Mean, and/or Median
Method
Used
Notes
CT DEEP:
Weston & Sampson,
2023
US
(Connecticut)
2021-
2022
WWTP Biosolids
Range:
Sludge (liquid): 0-51 ppt
Sludge (solid): 0-13 ppb
Mean:
Sludge (liquid): 13 ppt
Sludge (solid): 1 ppb
Median:
Sludge (liquid): 8.6 ppt
Sludge (solid): 0 ppb
Modified
Study of PFAS in WWTPs
PFOA Detection Rate:
Sludge (liquid): 90%
Sludge (solid): 23%
VT DEC:
Weston & Sampson,
2020
US (Vermont)
2018-
2019
WWTP
Biosolids
Range:
Sludge (liquid):
Average sum of 5 VT DEC regulated
PFAS (PFHxS, PFHpA, PFOA, PFOS,
and PFNA) across WWTPs:
<80 ppt, except one facility at 505 ppt
Sludge (solid):
Average sum of 5 VT DEC regulated
PFAS (PFHxS, PFHpA, PFOA, PFOS,
and PFNA) across WWTPs:
5-50 ppb, except one facility at 85 ppb
Modified
EPA Method
537.1
Study of PFAS in landfill leachate and WWTPs
Collected 75 sludge samples: Report summarized results
as sum of 5 VT DEC regulated PFAS (PFHxS, PFHpA,
PFOA, PFOS, and PFNA)
VT DEC:
Weston & Sampson,
2022
US (Vermont)
2021
PFAS Sources to
WWTPs
PFOA commonly detected in sources
(residential, commercial, and industrial
inputs)
Modified
EPA Method
537.1
Study of PFAS sources to WWTPs
Maine DEP:
Brown and Caldwell,
2023
US (Maine)
2019-
2022
WWTP Biosolids
Rang
2019
2020
2021
2022
Mean
2019
2020
2021
2022
e:
n.d.-46 ppb
0.6-63 ppb
0.3-25 ppb
0.8-38.9 ppb
:
9.4 ppb
8.2 ppb
5.3 ppb
6.6 ppb
Modified
EPA Method
537.1
Based on biosolids data in Maine's Environmental and
Geographic Analysis Database collected from 2019-2022
LOQ = Limit of Quantification
LOD = Limit of Detection
n.d. = non-detect
* below reporting limit or limit of detection
** estimated value based on quality assurance review
DRAFT
A-4
-------
PFOA/PFOS Risk Assessment
Table A-2. PFOS Occurrence in Biosolids in the US
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
Reference
Geographic
Area
Years
Sampled
Sample Type
PFOS Concentration
(Dry Weight Basis):
Range (Min-Max),
Mean, and/or Median
Method Used
Notes
3M Environmental
Laboratory, 2001
US (Multiple
states)
1999-
2001
VWVTP
Biosolids
Range:
58-159 ppb (4 WWTPs)
<3120 ppb (Decatur Utilities Plant)
Modified
Sampled 6 test cities, including Decatur, AL (3M Multi-City Study)
Higgins et al., 2005
US (Multiple
states)
1998-
2004
VWVTP
Biosolids
Range: 14.4-2610 ppb
Mean (Post-2002):
124 ppb (n=8)
Modified
Digested sludge samples from 8 WWTPs and primary settled
solids from 3 WWTPs
(9 WWTPs in total)
Schultz et al., 2006
US (Pacific
Northwest
Region)
2004
VWVTP
Biosolids
Range:
Digested sludge:
81-160 ppb
Modified
Analyzed wastewater and sludge samples throughout the
treatment process
Sinclair and
Kannan, 2006
US (New
York)
2005
VWVTP
Biosolids
Range:
Plant A: 26-65 ppb
Plant B: <10-34 ppb
Mean:
Plant A: 37 ppb
Plant B: 25 ppb
Median:
Plant A: 28 ppb
Plant B: 32 ppb
Modified
Sampled wastewater at 6 WWTPs, two of which were also
sampled for biosolids (five times each)
Loganathan et al.,
2007
US
(Kentucky
and Georgia)
2005
VWVTP
Biosolids
Range:
Plant A: 8.2-990 ppb
Plant B: <2.5-77 ppb
Modified
Sampled two WWTPs: rural (Plant A, Kentucky) and urban (Plant
B, Georgia)
Yoo et al., 2009
US
(Alabama
and New
York)
2007
VWVTP
Sewage
Sludge/
Biosolids
Mean:
Decatur VWVTP:
Sample A: 346.3±44.4 ppb
Sample B: 417.9±57.2 ppb
New York City WWTPs:
Sample A: 76.8±27.8 ppb
Sample B: 61.1±17.1 ppb
Sample C: 32.2±0.7 ppb
Modified
Isotopic-
Dilution
Method with
LC-MS/MS
Conducted a method development study for measuring PFAS,
using sludge samples from a VWVTP in Decatur, AL; this method
was then used to assess PFAS in a NIST sludge sample and
sludge samples from New York City WWTPs
Washington et al.,
2010
US
(Alabama)
2007 and
2009
Land-applied
Biosolids
Range:
2009: <410 ppb
Modified
Conducted two sampling surveys
(2007 and 2009)
Lindstrom et al.,
2011
US
(Alabama)
2009
Well and
Surface Water
near Land-
applied
Biosolids
Sites
Range:
Well Water:
-------
PFOA/PFOS Risk Assessment
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
PFOS Concentration
Reference
Geographic
Area
Years
Sampled
Sample Type
(Dry Weight Basis):
Range (Min-Max),
Mean, and/or Median
Method Used
Notes
Sepulvado et al.,
2011
US (Illinois)
2004-
2007
Land-applied
Biosolids
Range: 80-219 ppb
Mean: 144±57 ppb
Modified
Compiled 6 composite samples
PFOS Detection Rate = 100%
Venkatesan and
Halden, 2013
US (Multiple
states)
2001
WWTP
Biosolids
Range: 308-618 ppb
Mean: 403±127 ppb
Modified EPA
Method 1694
Compiled 5 composite samples from 110 archived biosolids
samples from the US EPA 2001 NSSS (94 POTWs)
PFOS Detection Rate = 100%
Armstrong et al.,
2016
US (Mid-
Atlantic
Region)
2005-
2013
WWTP
Biosolids
Mean: 22.5 ppb
Median: 19.3 ppb
Modified
Performed temporal trend study (multiple samples collected
between 2005 and 2013 from 1 urban WWTP)
Lazcano et al., 2020
US (Multiple
states)
2014,
2016,
2018
Biosolids-
based
Products
Range:
Biosolids-based products:
2.6-88.5 ppb
Modified
Analyzed multiple types of biosolids-based and non-biosolids
organic products
Pepper et al., 2021
US (Arizona)
2020
WWTP
Biosolids and
Land-applied
Biosolids
Range:
Biosolids: 14-36 ppb
Modified EPA
Method 537.1
Collected samples in 2020 from a WWTP in Arizona and field
sites where Class B biosolids were land applied from 1984-2019
Helmer et al., 2022
US
(Michigan)
2018-
2020
WWTP
Biosolids
Range: 4-6500 ppb
For 8 of 11 samples, PFOS
dominant
Modified EPA
Method 537.1
Analyzed 11 samples from 6 industrially impacted WWTPs;
PFOS was the dominant type of PFAS measured in 8 of the 11
biosolids samples (-73%)
Johnson, 2022
US (Western
Region)
2015
Land-applied
Biosolids
Mean: 12 ppb
Modified
Collected 2 biosolids samples
Schaefer et al.,
2022
US (Multiple
states)
2020
WWTP
Biosolids and
Column
Mesocosm
Leaching
Experiments
Range: 0.386-150 ppb
Modified
Sampled 7 WWTPs with a variety of treatment processes in
urban areas, receiving both industrial and domestic sources, as
well as performed column mesocosm leaching experiments;
found that PFAS precursors accounted for over 75% of total
PFAS
Thompson et al.,
2023a
US (Florida)
2021
WWTP
Sewage
Sludge/
Biosolids
Range:
Sludge (Before Treatment):
4-41 ppb
Biosolids (After Treatment):
1.4-19 ppb
Modified
Interviewed 39 facilities in Florida to learn treatment processes
from 2019-2021; Then, in 2021, collected 16 samples (before and
after treatment) from 8 facilities representing the four most
common treatment processes; studied 92 PFAS analytes,
including precursors
Thompson et al.,
2023b
US (Florida)
Sludge:
2021
Toilet
Paper:
2021-
2022
WWTP
Sewage
Sludge/
Biosolids and
Toilet Paper
Range:
Sludge: 4-41 ppb
Modified
Focused on studying diPAPs in sludge (Florida, US) and toilet
paper samples (US and other countries)
Link et al., 2024
US
2018-
WWTP
Range:<2150 ppb
Modified EPA
Sampled 190 WWTPs
(Michigan)
2022
Biosolids
Mean: 40±179 ppb
Method 537.1
PFOS Detection Rate = 95%
DRAFT
A-6
-------
PFOA/PFOS Risk Assessment
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
PFOS Concentration
Reference
Geographic
Area
Years
Sampled
Sample Type
(Dry Weight Basis):
Range (Min-Max),
Mean, and/or Median
Method Used
Notes
Ml EGLE, 2021
US
(Michigan)
2018-
2021
VWVTP
Biosolids
Industrially Impacted:
Range: 360-6500 ppb
Not Industrially Impacted:
Mean: 18 ppb
Median: 11 ppb
Modified EPA
Method 537.1
State PFAS Survey - Interim Strategy:
Surveyed 42 WWTPs;
Industrially impacted:
6 WWTPs
Ml EGLE, 2022
US
(Michigan)
2017/
2018,
2021
VWVTP
Biosolids
Industrially Impacted:
Range:
2017/2018:160-2150 ppb
2021:33-180 ppb
Modified EPA
Method 537.1
Update to State PFAS Survey - Interim Strategy
USGS/NH DES:
Phase 1:
Santangelo et al.,
2022;
US (New
Hampshire)
2021-
2022
Soils, Land-
applied
Biosolids,
Solid/Water
Range:
Finished biosolids (collected from
facilities in 2021 as part of Phase
2):
Eurofins LC-
MS/MS and
Isotope
Dilution
Three-phase PFAS study of soils, land-applied biosolids,
solid/water partitioning, and groundwater leaching
Phase 2: Tokranov
et al., 2023;
Partitioning,
and
2.2-7.9 ppb
Phase 3:
Groundwater
Santangelo et al.,
2023
Leaching
San Francisco
Estuary Institute
(SFEI):
Phase 1: Mendez et
US
(California)
2020
VWVTP
Biosolids
Range: n.d.-49 ppb
Mean: 14 ppb
Median: 13 ppb
SGS AXYS
Method MLA-
110
PFAS Study of Bay Area WWTPs: Phase 1
al., 2021
MPCA, 2008
US
2007-
VWVTP
Range:
Modified
Monitored PFAS at WWTPs in 2007 and 2008
(Minnesota)
2008
Sewage
Sludge
2007: <0.382**-861 ppb
2008: 4.15**-442 ppb
CT DEEP:
US
2021-
VWVTP
Range:
Modified
Study of PFAS in WWTPs
Weston &
Sampson,
2023
(Connecticut
)
2022
Biosolids
Sludge (liquid): 0-21 ppt
Sludge (solid): 0-43 ppb
Mean:
Sludge (liquid): 7 ppt
Sludge (solid): 12.4 ppb
Median:
Sludge (liquid): 4.9 ppt
Sludge (solid): 10 ppb
PFOS Detection Rate:
Sludge (liquid): 70%
Sludge (solid): 85%
DRAFT
A-7
-------
PFOA/PFOS Risk Assessment
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
Reference
Geographic
Area
Years
Sampled
Sample Type
PFOS Concentration
(Dry Weight Basis):
Range (Min-Max),
Mean, and/or Median
Method Used
Notes
VT DEC:
Weston &
Sampson,
2020
US
(Vermont)
2018-
2019
WWTP
Biosolids
Range:
Sludge (liquid):
Average sum of 5 VT DEC
regulated PFAS (PFHxS, PFHpA,
PFOA, PFOS, and PFNA) across
WWTPs:
<80 ppt, except one facility at 505
ppt
Sludge (solid):
Average sum of 5 VT DEC
regulated PFAS (PFHxS, PFHpA,
PFOA, PFOS, and PFNA) across
WWTPs:
5-50 ppb, except one facility at 85
ppb
Modified EPA
Method 537.1
Study of PFAS in landfill leachate and WWTPs
Collected 75 sludge samples: Report summarized results as sum
of 5 VT DEC regulated PFAS (PFHxS, PFHpA, PFOA, PFOS,
and PFNA)
VT DEC:
Weston &
Sampson,
2022
US
(Vermont)
2021
PFAS
Sources to
WWTPs
PFOS commonly detected in
sources (residential, commercial,
and industrial inputs)
Modified EPA
Method 537.1
Study of PFAS sources to WWTPs
Maine DEP:
Brown and
Caldwell,
2023
US (Maine)
2019-
2022
WWTP
Biosolids
Rang
2019
2020
2021
2022
Mear
2019
2020
2021
2022
e:
2.2-120 ppb
2.5-51.9 ppb
2.1-111 ppb
1.2-66 ppb
:
27.2 ppb
16.6 ppb
22.7 ppb
19.3 ppb
Modified EPA
Method 537.1
Based on biosolids data in Maine's Environmental and
Geographic Analysis Database collected from 2019-2022
LOQ = Limit of Quantification
LOD = Limit of Detection
n.d. = non-detect
* below reporting limit or limit of detection
** estimated value based on quality assurance review
DRAFT
A-8
-------
PFOA/PFOS Risk Assessment Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
A.2 Occurrence of PFOA and PFOS Precursors
Table A-3. Recent Examples of PFOA and PFOS Precursor Occurrence in Biosolids in the US
Reference
Geographic
Area
Years
Sampled
Sample Type
PFOA Precursor:
8:2 diPAP Concentration
(Dry Weight Basis):
Range
(Min-Max)
PFOS Precursor:
NEtFOSAA Concentration
(Dry Weight Basis):
Range
(Min-Max)
Method
Used
Notes
Schaefer etal.,
2022
US (Multiple
states)
2020
WWTP Biosolids
and Column
Mesocosm
Leaching
Experiments
Range:
13.5-347 ppb
Range:
0.297-18 ppb
Modified
Sampled 7 WWTPs with a variety of treatment
processes in urban areas, receiving both
industrial and domestic sources, as well as
performed column mesocosm leaching
experiments; found that PFAS precursors
accounted for over 75% of total PFAS
Thompson etal.,
2023a
US (Florida)
2021
WWTP Sewage
Sludge/
Biosolids
Range:
Sludge
(Before Treatment):
21-300 ppb
Biosolids (After
Treatment):
5.9-100 ppb
Range:
Sludge
(Before Treatment):
0-7.6 ppb
Biosolids (After Treatment):
0-3.9 ppb
Modified
Interviewed 39 facilities in Florida to learn
treatment processes from 2019-2021; Then,
in 2021, collected 16 samples (before and
after treatment) from 8 facilities representing
the four most common treatment processes;
studied 92 PFAS analytes, including
precursors
Thompson etal.,
2023b
US (Florida)
Sludge:
2021
Toilet Paper:
2021-2022
WWTP Sewage
Sludge/
Biosolids and
Toilet Paper
Range:
Sludge:
21-300 ppb
Toilet Paper:
-------
PFOA/PFOS Risk Assessment
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
A.3 References
3M Environmental Laboratory. (2001). 3M Environmental Monitoring - Multi-City Study Water, Sludge,
Sediment, POTW Effluent and Landfill Leachate Samples. 3M Environmental Laboratory: St. Paul,
MN. https://static.ewg.org/files/MultiCity execsum.pdf
Armstrong, D. L., Lozano, N., Rice, C. P., Ramirez, M., & Torrents, A. (2016). Temporal trends of
perfluoroalkyl substances in limed biosolids from a large municipal water resource recovery facility.
Journal of Environmental Management, 165, 88-95. https://doi.Org/10.1016/i.ienvman.2015.09.023
Brown and Caldwell (prepared for Maine DEP). (2023). An Evaluation of Biosolids Management in Maine
and Recommendations for the Future (December 2023).
https://www.maine.gov/tools/whatsnew/attach. php?id=12198306&an=l
Helmer, R. W., Reeves, D. M., & Cassidy, D. P. (2022). Per- and polyfluorinated alkyl substances (PFAS)
cycling within Michigan: Contaminated sites, landfills and wastewater treatment plants. Water
Research, 210, 117983. https://doi.Org/10.1016/i.watres.2021.117983
Higgins, C. P., Field, J. A., Criddle, C. S., & Luthy, R. G. (2005). Quantitative determination of
perfluorochemicals in sediments and domestic sludge. Environmental Science and Technology, 39,
3946-3956. https://doi.org/10.1021/esQ48245p
Johnson, G. R. (2022). PFAS in soil and groundwater following historical land application of biosolids.
Water Research, 211, 118035. https://doi.Org/10.1016/i.watres.2021.118035
Lazcano, R. K., Choi, Y. J., Mashtare, M. L., & Lee, L. S. (2020). Characterizing and comparing per- and
polyfluoroalkyl substances in commercially available biosolid and organic non-biosolid-based
products. Environmental Science and Technology, 54, 8640-8648.
Lindstrom, A. B., Strynar, M. J., Delinsky, A. D., Nakayama, S. F., McMillan, L., Libelo, E. L., Neil I, M., &
Thomas, L. (2011). Application of WWTP biosolids and resulting perfluorinated compound
contamination of surface and well water in Decatur, Alabama, USA. Environmental Science &
Technology, 45(19), 8015-8021. https://doi.org/10.1021/eslQ39425
Link, G. W., Reeves, D. M., Cassidy, D. P., & Coffin, E. S. (2024). Per- and polyfluoroalkyl substances
(PFAS) in final treated solids (Biosolids) from 190 Michigan wastewater treatment plants. Journal of
Hazardous Materials, 463, 132734. https://doi.Org/10.1016/i.ihazmat.2023.132734
Loganathan, B. G., Sajwan, K. S., Sinclair, E., Kumar, K. S., & Kannan, K. (2007). Perfluoroalkyl sulfonates
and perfluorocarboxylates in two wastewater treatment facilities in Kentucky and Georgia. Water
Research, 41, 4611-4620. https://doi.Org/10.1016/i.watres.2007.06.045
Mendez, M., Lin, D., Wong, A., Yee, D., & Sutton, R. (2021). Study of Per- and Polyfluoroalkyl Substances
in Bay Area POTWs: Phase 1 Memo. San Francisco Estuary Institute, Richmond, CA.
https://bacwa.org/wp-content/uploads/2023/03/Memo BACWA-PFAS-Phase-l.pdf
Michigan Department of Environment, Great Lakes, and Energy (Ml EGLE). (2021). Land application of
biosolids containing PFAS: Interim strategy (March 2021). https://www.michigan.gov/egle/-
/media/Proiect/Websites/egle/Documents/Programs/WRD/Biosolids/PFAS-Biosolids-Strategy.pdf
Michigan Department of Environment, Great Lakes, and Energy (Ml EGLE). (2022). Land application of
biosolids containing PFAS: Interim strategy (Updated April 2022). https://www.michigan.gov/egle/-
/media/Proiect/Websites/egle/Documents/Programs/WRD/Biosolids/PFAS-Biosolids-lnterim-
Strategy-2022.pdf
DRAFT
A-10
-------
PFOA/PFOS Risk Assessment
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
Minnesota Pollution Control Agency (MPCA). (2008). PFCs in Minnesota's Ambient Environment: 2008
Progress Report, https://www.pca.state.mn.us/sites/default/files/c-pfcl-02.pdf
Pepper, I. L., Brusseau, M. L., Prevatt, F. J., & Escobar, B. A. (2021). Incidence of PFAS in soil following
long-term application of class B biosolids. Science of the Total Environment, 793, 148449.
https://doi.Org/10.1016/i.scitotenv.2021.148449
Santangelo, L. M., Tokranov, A. K., Welch, S. M., Schlosser, K. E. A., Marts, J. M., Drouin, A. F., Ayotte, J.
D., Rousseau, A. E., & Harfmann, J. L. (2022). Statewide survey of shallow soil concentrations of per-
and polyfluoroalkyl substances (PFAS) and related chemical and physical data across New
Hampshire, 2021: US Geological Survey data release. https://doi.org/10.5066/P9KG38B5
Santangelo, L. M., Welch, S. M., Tokranov, A. K., Drouin, A. F., Schlosser, K. E. A., Marts, J. M., Lincoln, T.
A., Deyette, N. A., & Perkins, K. (2023). Field-scale investigation of per- and polyfluoroalkyl
substances (PFAS) leaching from shallow soils to groundwater at two sites in New Hampshire, 2021-
2022: US Geological Survey data release. https://doi.org/10.5066/P92C21F6
Schaefer, C. E., Hooper, J., Modiri-Gharehveran, M., Drennan, D. M., Beecher, N., & Lee, L. (2022).
Release of poly- and perfluoroalkyl substances from finished biosolids in soil mesocosms. Water
Research, 217, 118405. https://doi.Org/10.1016/i.watres.2022.118405
Schultz, M. M., Higgins, C. P., Huset, C. A., Luthy, R. G., Barofsky, D. F., & Field, J. A. (2006).
Fluorochemical mass flows in a municipal wastewater treatment facility. Environmental Science and
Technology, 40(23), 7350-7357. https://doi.org/10.1021/es061025m
Sepulvado, J. G., Blaine, A. C., Hundal, L. S., & Higgins, C. P. (2011). Occurrence and fate of
perfluorochemicals in soil following the land application of municipal biosolids. Environmental
Science and Technology, 45, 8106-8112. https://doi.org/10.1021/esl03903d
Sinclair, E., & Kannan, K. (2006). Mass loading and fate of perfluoroalkyl surfactants in wastewater
treatment plants. Environmental Science and Technology, 40(5), 1408-1414.
https://doi.org/10.1021/es051798v
Thompson, J. T., Robey, N. M., Tolaymat, T. M., Bowden, J. A., Solo-Gabriele, H. M., & Townsend, T. G.
(2023a). Underestimation of per- and polyfluoroalkyl substances in biosolids: Precursor
transformation during conventional treatment. Environmental Science & Technology, 57(9), 3825-
3832. https://doi.org/10.1021/acs.est.2cQ6189
Thompson, J. T., Chen, B., Bowden, J. A., & Townsend, T. G. (2023b). Per- and polyfluoroalkyl substances
in toilet paper and the impact on wastewater systems. Environmental Science & Technology
Letters, 10(3), 234-239. https://doi.org/10.1021/acs.estlett.3cQ0094
Tokranov, A. K., Welch, S. M., Santangelo, L. M., Kent, D. B., Repert, D. A., Perkins, K., Bliznik, P. A., Roth,
D. A., Drouin, A. F., Lincoln, T. A., Deyette, N. A., Schlosser, K. E. A., & Marts, J. M. (2023).
Solid/water Partitioning of per- and polyfluoroalkyl substances (PFAS) in New Hampshire soils and
biosolids: Results from laboratory experiments at the US Geological Survey: US Geological Survey
data release. https://doi.org/10.5066/P9TKSM8S
Venkatesan, A. K., & Halden, R. U. (2013). National inventory of perfluoroalkyl substances in archived US
biosolids from the 2001 EPA National Sewage Sludge Survey. Journal of Hazardous Materials, 252-
253, 413-418. https://doi.Org/10.1016/i.ihazmat.2013.03.016
DRAFT
A-ll
-------
PFOA/PFOS Risk Assessment
Appendix A: PFOA and PFOS Occurrence in Biosolids in the US
Washington, J. W., Yoo, H., Ellington, J. J., Jenkins, T. M., & Libelo, L. (2010). Concentrations, distribution,
and persistence of perfluoroalkylates in sludge-applied soils near Decatur, Alabama, USA.
Environmental Science and Technology, 44, 8390-8396. https://doi.org/10.1021/esl003846
Weston & Sampson (prepared for CT DEEP). (2023). Water Pollution Control Facility PFAS Sampling
Study: Report (June 2023). https://portal.ct.gov/-
/media/DEEP/water/municipal wastewater/PFAS/CTDEEP-WPCF-PFAS-Study-06 08 23-FINAL.pdf
Weston & Sampson (prepared for VT DEC). (2020). Poly- and Perfluoroalkyl Substances at Wastewater
Treatment Facilities and Landfill Leachate: 2019 Summary Report (January 2020).
https://dec.vermont.gov/sites/dec/files/wmp/SolidWaste/Documents/02.03.20 PFAS%20in%20LF%
20and%20WWTF%20Final%20Report.pdf
Weston & Sampson (prepared for VT DEC). (2022). Poly- and Perfluoroalkyl Substances Inputs to
Wastewater Treatment Facilities: Summary Report (March 2022).
https://dec.vermont.gov/sites/dec/files/wmp/residual/2021%20VTDEC%20PFAS%20lnputs%20to%2
0WWTF%20Studv.2022March29.pdf
Yoo, H., Washington, J.W., Jenkins, T.M., & Libelo, E.L. (2009). Analysis of perfluorinated chemicals in
sludge: Method development and initial results. Journal of Chromatography A, 1216, 7831-7839.
https://doi.Org/10.1016/i.chroma.2009.09.051
DRAFT
A-12
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
APPENDIX B. MODEL INPUTS
This appendix is organized by the component models used in this assessment, as follows:
• B.l General Inputs (used by multiple models)
• B.2 Land Application Unit Source Model (3MRA LAU Source Module)
• B.3 Surface Disposal Unit Source Model (3MRA SI Module)
• B.4 Groundwater Model (EPACMTP)
• B.5 Surface Water Model (VVWM)
• B.6 Food Chain Calculations
• B.7 Exposure Calculations
• B.8 Risk Calculations.
Within any section, multiple tables may be provided if inputs vary with scenario, chemical, or climate
location. Within each table, inputs are listed alphabetically. Note that some values may be rounded for
clarity of presentation.
B.l General Inputs
Table B-l. Chemical-specific Inputs
Parameter
Description & Units
PFOA
PFOS
Reference
Comment
ChemType
Type of chemical (e.g.,
organic, metal/inorganic,
mercury, dioxin-like)
0
0
NA
This parameter is used by the
source and food chain models
to identify the appropriate
algorithms and inputs, as these
differ between organics and
inorganics. PFOA and PFOS
are both organic chemicals.
Da
Diffusivity in air (cm2/s)
NA
NA
NA
This assessment does not
include modeling transport
through air
Dw
Diffusion coefficient in
water (cm2/s)
5.52E-06
4.96E-06
US EPA (2016)
HLC
Henry's law constant
[atm-m3/mol]
NA
NA
NA
Volatilization is not expected
under environmental conditions
(see Section 2.2.2)
Koc-high
Organic carbon partition
coefficient (high end)
[mL/g]
1,100
22,000
PFOA: Campos-
Pereira et al., 2023;
PFOS: Chen et al.,
2020
90th percentile from literature
search; n = 203 for PFOA, 253
for PFOS; see Appendix C for
more details
Koc-low
Organic carbon partition
coefficient (low end)
[mL/g]
26
250
PFOA: Hubert, M„ et
al, 2023; PFOS:
Johnson et al., 2007
10th percentile from literature
search; n = 203 for PFOA, 253
for PFOS; see Appendix C for
more details
MW
Molecular weight [g/mol]
414
500
PFOA: HSDB (US NLM, 2010); PFOS: Physprop (SRC,
2016)
DRAFT
B-l
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
B.2 Regional Location-based Parameters
In addition to general chemical parameters, the assessment modeled three climates—dry, moderate,
and wet—represented by specific meteorological stations. These were chosen based on the number of
precipitation days per year, not total annual rainfall. The meteorological stations and their general
descriptive data are as follows (all from SAMSON—US DOC & US DOE, 1993):
• Dry climate:
- Location of meteorological station: Boulder, CO
- WBAN station number: 94018
- Meteorological station latitude: 40.0167°
- Long-term average annual air temperature: 10.11 °C
• Moderate Climate:
- Location of meteorological station: Chicago, IL
- WBAN station number: 94846
- Meteorological station latitude: 41.983°
- Long-term average annual air temperature: 9.69 °C
• Wet Climate:
- Location of meteorological station: Charleston, SC
- WBAN station number: 13880
- Meteorological station latitude: 32.9°
- Long-term average annual air temperature: 18.18 °C
B.3 LAU Source Model Inputs (3MRA LAU Module)
Chemical-, scenario-, and location-specific inputs are presented in Tables B-2, B-3, and B-4, respectively.
The LAU Source Module has three submodules, the Generic Soil Colum Model (GSCM), which evaluates
movement vertically through the soil column; the Local Watershed Model (LWS), which evaluates
movement horizontally onto and off the field; and the Particulate Emissions Model (PEM), which
accounts for particulate emissions to air. The PEM accounts for losses only; this assessment does not
model transport through air. The "Used in" column notes which submodel uses an input (or says "LAU"
if the input is general to all submodules).
Inputs for which there is a single value (i.e., they are not specific to a chemical, scenario, or location) are
presented in Table B-5, grouped by LAU submodel.
DRAFT
B-2
-------
PFOA/PFOS Risk Assessment Appendix B. Model Inputs
Table B-2. Chemical-Specific Inputs to the LAU Source Module
Parameter
Description & Units
Used in
PFOA
PFOS
Reference
Comment
ChemFracNeutral
Fraction of chemical concentration
in the neutral species (fraction)
GSCM
NA
NA
NA
Used to adjust properties for chemicals that ionize; not applicable to this
assessment.
ChemTemp
Temperature (°C)
GSCM
NA
NA
NA
Temperature at which degradation and volatilization rates are measured;
not applicable to this assessment
ksoil
Soil biodegradation rate (1/day)
GSCM
0
0
NA
Based on PFOA/PFOS degradation literature
Sol
Solubility [mg/L]
GSCM
9500
680
US EPA (2017a)
Used to determine if solubility is exceeded in the soil column during
model run
Table B-3. Scenario-Specific Inputs to the LAU Module
Parameter
Description & Units
Used in
Crop
Pasture
Reclamation
Reference
Comment
AppDepth
Depth of biosolids incorporation
(m)
GSCM
0.2
0.02
0.02
Assumption
For the crop scenario, biosolids are tilled into the soil to
a depth of 20 cm at application. For the pasture and
reclamation scenarios, the biosolids are not tilled in, but
are assumed to be incorporated to a depth of 2 cm by
bioturbation. This assumption is consistent with the
2003 Biosolids assessment, US EPA (2003a).
CN_wmu
SCS curve number for field
(dimensionless ratio)
LWS
81
71
71
USDA (1986)
Average across hydrologic soil groups and hydrologic
conditions for straight row crops (crop scenario) or
pasture lands (pasture, reclamation scenarios)
DryApplRate
Application rate of biosolids to
the field, dry weight per
application (MT DW/ha/appI)
LAU
10
10
50
Crop & pasture: Biosolids Tool
(BST; US EPA, 2023a; see
Appendix E); Reclamation:
Sopper (1993)
Crop & Pasture: median of agronomic rates from
probabilistic plant available nitrogen (PAN) calculations
conducted for the BST; Reclamation: 5 x agronomic
rate
fcult
Number of cultivations per
application (count)
PEM
5
1
1
TSDF Fugit. Air (US EPA,
1989b)
Impacts spreading and compacting losses
fd
Frequency of surface
disturbance per month on field
(1/mo)
PEM
0.21
0.042
0.042
Biosolids 2003 (US EPA,
2003a)
Impacts wind erosion losses
OpLife
Number of years biosolids are
applied to the field (years)
LAU
40
40
1
Assumption
Chosen for consistency with 2003 Biosolids
assessment (US EPA, 2003a) and 3MRA default value
(US EPA, 2003b)
Pwmu
USLE erosion control factor for
field (fraction)
LWS
0.5
1
1
Wanielista & Yousef (1993)
a value of 1 means no erosion control practices; these
are the 3MRA defaults.
Rappl
Application rate of biosolids to
the field, whole weight per year
(MT WW/m2-year)
LAU
0.0025
0.0025
0.0125
Calculated
[DryApplRate x Nappl x 1E-4 ha/m2]/[%solids/100]
zruf
Roughness height of the field
(cm)
PEM
1
3.7
3.7
TSDF Fugit. Air (US EPA,
1989b)
Impacts wind erosion losses
DRAFT
B-3
-------
PFOA/PFOS Risk Assessment
Table B-4. Location-Specific Inputs to the LAU Module
Appendix B. Model Inputs
Parameter
Description & Units
Used in
Dry
Moderate
Wet
Reference
R
USLE rainfall/erosivity factor (1/year)
LWS
50
155
360
Wischmeier and Smith (1978)
Uw
Mean annual wind speed (m/sec)
PEM
3.783
4.632
3.788
SAMSON (US DOC & US DOE, 1993)
Table B-5. Individual Inputs to the LAU Module
Parameter
Description & Units
Value
Reference
Comment
General LAU Module Inputs
%solid
Percent solids of biosolids applied to field (mass percent)
48
TNSSS (US EPA, 2009)
Midpoint of range (0.14-94.9%) based on 84 samples
Area field
Area of the agricultural field (m2)
323,750
USDA (2014)
80 acres
Nappl
Number of biosolids applications per year (1/year)
1
Assumption
Application is assumed to occur on April 1, at the start of the
growing season.
Ss
Silt content of soil (mass %)
42.5
STATSGO (USDA, 1994)
area weighted average for each soil texture within met region
- median value
WSpH
Soil pH (pH units)
NA
NA
Used for ionizable chemicals to adjust properties; not
applicable to this assessment
GSCM Inputs
BDwaste
Dry bulk density of biosolids applied to field (g/cm3)
0.7
Gunn etal. (2004)
foc_biosolids
Fraction organic carbon of biosolids applied to field
(fraction)
0.4
Biosolids 2003 (US EPA 2003a)
foc_soil
Fraction organic carbon for natural soil in the soil column
under the field (fraction)
0.0118
STATSGO (USDA, 1994)
Calculated using percent organic matter from STATSGO,
based on EPACMTP - median value
fwmu
Fraction of waste in LAU (fraction)
1
Assumption
Indicates that all sewage sludge is applied to field
Ksat
Saturated hydraulic conductivity of soil (cm/h)
0.45
Carsel & Parrish (1988)
based on surface soil textures - median value
WCS
Saturated volumetric water content, porosity for soil
(mL/cm3)
0.43
Carsel & Parrish (1988)
based on surface soil textures - median value
LWS Inputs
Area_buffer
Area of the buffer between the field and the waterbody (m2)
5690
Calculated
=length of source x buffer width; length is 569 m, width is 10 m
per Part 503 Biosolids rule; ~1.4 acres
C
USLE cover factor (fraction)
0.1
HHRAP (US EPA, 2005)
CN_buffer
SCS curve number (dimensionless ratio)
69
USDA (1986)
Average across hydrologic soil groups and hydrologic
conditions for good pasture and farmsteads
ConVs
Settling velocity of suspended solids in runoff from field
(m/day)
5.36
Schroeder (1977)
derived from "mineral sludge" values - median value
DRZ
Root zone depth (cm)
82.7
Dunne & Leopold (1978)
median value
K
USLE soil erodibility factor (kg/m2)
0.0716
STATSGO (USDA, 1994)
area weighted average for each soil texture within met region
- median value
LS
USLE length-slope factor (empirical)
1.5
HHRAP (US EPA, 2005)
Default assessment values from HHRAP
P_buffer
USLE erosion control factor for buffer (fraction)
1
Wanielista & Yousef (1993)
A value of 1 means no erosion control practices. These are the
3MRA defaults.
DRAFT
B-4
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
Parameter
Description & Units
Value
Reference
Comment
SMb
Soil moisture coefficient (vol %)
5.3
Clapp & Hornberger (1978)
based on surface soil textures - median value
SMFC
Soil moisture field capacity (vol %)
22.48
Carsel et al. (1988)
based on average hydrologic soil group for each soil texture -
median value
SMWP
Soil moisture wilting point (vol %)
11.48
Carsel et al. (1988)
based on average hydrologic soil group for each soil texture -
median value
Theta
Slope of watershed (degrees)
3.66
STATSGO (USDA, 1994)
area weighted average for each soil texture within met region
- median value
X
Flow length for local watershed (m)
129
Mills etal. (1985)
calculated from theta and LS using equation in cited reference
PEM Inputs
asdm
Mode value of the size of soil aggregates in an LAU (mm)
0.5
3MRA (US EPA, 2003b)
3MRA default
effdust
Dust suppression control efficiency (field) (fraction)
0
NA
no regular vehicular activity
Lc
Soil roughness ratio (dimensionless ratio)
2.31E-04
TSDF Fugit. Air (US EPA, 1989b)
mt
Distance vehicle travels on field (m)
0
NA
no regular vehicular activity
nv
Number of vehicles per day on field (1/day)
0
NA
no regular vehicular activity
nw
Number of wheels on each vehicle (count)
4
NA
no regular vehicular activity
Sw
Silt content of biosolids (mass %)
10
AP-42 (US EPA, 1995)
veg
Fraction vegetative cover for the field (fraction)
0.8
Assumption
This is the minimum of the assumed 3MRA distribution (which
is 0.8 -1, mean of 0.9, assumed normal). That's based on
"best professional judgement, assuming unit is vegetated
during operation and after closure."
vw
Vehicle weight (MT)
0
NA
no regular vehicular activity
B.4 Surface Disposal Unit Source Model Inputs (3MRA Surface Impoundment Module)
Table B-6. SDU Inputs
Parameter
Description & Units
Value
Reference
Comment
General Parameters
Area SI
Area of the SDU (m2)
3,400
calculated
=Qwmu/(dwmu * EconLife)
Bio_yield
Biomass yield of the SDU (g/g)
0.6
Tchobanoglous et al. (1979)
Median; generally ranges from 0.4 to 0.8
d wmu
Depth of the SDU (m)
2
3MRA (US EPA, 2003b)
Median of data from Industrial D Screening Survey
DBGS
Depth of SDU below ground surface (m)
0
EPACMTP (US EPA, 2003c)
EconLife
Operating life of surface disposal unit (yr)
50
3MRA (US EPA, 2003b)
3MRA default
Q wmu
Volumetric flow rate into SDU (m3/s)
4E-06
3MRA (US EPA, 2003b)
Median of data from Industrial D Screening Survey
Waste Parameters
C in
Concentration of constituent in SDU influent (mg/L)
Calculated
CTPWasteDry * TSSJn
CBOD
Biological oxygen demand of SDU influent (g/cm3)
8E-3
Tchobanoglous et al. (1979)
Tbl 3-6, typical value for untreated septage
dmeanTSS
Particle diameter of solids in SDU (cm)
0.001
Tchobanoglous et al. (1979)
Default value from the surface impoundment module of 3MRA
kbal
Biologically active solids/total solids ratio in SDU (unitless)
0.4
Tchobanoglous et al. (1979)
Tbl 11-4, typical value for digested sludge
rho_part
Density of solids in SDU (g/cm3)
2.5
Tchobanoglous et al. (1979)
Default value from the surface impoundment module of 3MRA
DRAFT
B-5
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
Parameter
Description & Units
Value
Reference
Comment
SrcPh
pH of SDU influent (pH units)
NA
NA
Used to adjust chemical properties for ionizable chemicals for
temp and pH; not applicable to PFOA/PFOS
SrcTemp
Temperature of waste in SDU (°C)
NA
NA
Used to adjust chemical properties for ionizable chemicals for
temp and pH; not applicable to PFOA/PFOS
TSS in
Total suspended solids in SDU influent (g/cm3)
0.1
Tchobanoglous et al. (1979)
Tbl 11-4, typical value for digested sludge
TSS_out
Total suspended solids in SDU effluent (g/cm3)
NA
NA
This assessment assumes that there is no effluent from the
surface disposal site
Sediment Layer Properties
d setpt
Max fraction of SDU area occupied by sediments (fraction)
0.5
3MRA (US EPA, 2003b)
Median of data from Industrial D Screening Survey
hydc ssed
Hydraulic conductivity of the SDU sediment layer (m/s)
5E-07
Tchobanoglous et al. (1979)
Median; generally ranges from 1E-9 to 1E-6
k dec
Digestion rate of sediments in the SDU (1/s)
7E-07
Tchobanoglous et al. (1979)
Median; generally ranges from 4.6E-7 to 8.7E-7
SedAlpha
Soil retention parameter alpha of SDU sediment (1/cm)
0.016
Carsel and Parrish (1988)
Mean for silt soils
SedBeta
Soil retention parameter beta of SDU sediment (unitless)
1.37
Carsel and Parrish (1988)
Mean for silt soils
Liner Properties (used to calculate leachate quantity to pass to EPACMTP)
d liner
Thickness of clay liner (m)
0.9144
EPACMTP (US EPA, 2003c)
Default
hydc liner
Saturated conductivity of clay liner (m/s)
1E-09
EPACMTP (US EPA, 2003c)
Default
Infil CompLiner
Infiltration rate through composite liner (m/d)
1.4E-06
EPACMTP (US EPA, 2003c)
90th percentile (Table 4.6)
Liner Alpha
Soil retention parameter alpha of the SDU liner (1/cm)
0.008
Carsel and Parrish (1988)
Mean for clay soils
LinerBeta
Soil retention parameter beta of the SDU liner (unitless)
1.09
Carsel and Parrish (1988)
Mean for clay soils
Vadose Zone and Aquifer Properties (used to calculate amount of infiltration to pass to EPACMTP)
AquSATK
Saturated hydraulic conductivity (m/yr)
1890
EPACMTP (US EPA, 2003c)
National median values; the SDU source model uses these to
estimate infiltration rate and does not distinguish location; the
GW modeling uses location-specific values.
AquThick
Saturated zone thickness (m)
14.3
EPACMTP (US EPA, 2003c)
VadAlpha
Soil retention parameter alpha (1/cm)
0.0152
EPACMTP (US EPA, 2003c)
VadBeta
Soil retention parameter beta (unitless)
1.37
EPACMTP (US EPA, 2003c)
VadSATK
Saturated hydraulic conductivity of vadose zone soil (cm/h)
0.0089
EPACMTP (US EPA, 2003c)
VadThick
Thickness of vadose zone (m)
6.1
EPACMTP (US EPA, 2003c)
Aerator Properties (Not Used)
d imp
Impeller diameter (cm)
0
NA
SDU modeled as quiescent SI
F aer
Fraction surface area-turbulent (fraction)
0
NA
J
Oxygen transfer factor (lb 02/h-hp)
0
NA
njmp
Number of Impellers/aerators (dimensionless)
0
NA
02eff
Oxygen transfer correction factor (dimensionless)
0
NA
Powr
Total Power for Impellers/aerators (hp)
0
NA
w_imp
Impeller speed (rad/s)
0
NA
DRAFT
B-6
-------
PFOA/PFOS Risk Assessment Appendix B. Model Inputs
B.5 Groundwater Model Inputs (EPACMTP)
Table B-7. EPACMTP Inputs
Parameter
Description & Units
Dry
Moderate
Wet
Reference
Comment
Vadose Zone Properties
ALPHA
Moisture retention parameter (Van
Genuchten) for unsaturated zone (1/cm)
0.07
0.009
0.016
FGD (US EPA, 2023b)
Median based on soil texture
(ISTYPE1)
BETA
Moisture retention parameter for
unsaturated zone (unitless)
1.885
1.236*
1.409
FGD (US EPA, 2023b),
*except silty clay loam had
no distribution in FGD, so
median from Carsel and
Parrish (1988)
Median based on soil texture
(ISTYPE1)
DISPR
Longitudinal dispersivity in unsaturated
zone (m)
0.21437
0.2884
0.10382
EPACMTP (US EPA,
2003c)
calculated from DSOIL using
Eqn.5.2 in source
DSOIL
Depth from ground surface to water table
(m)
8.835
12.2
3.81
Newell etal. (1990)
median
ISTYPE1
Soil type of vadose zone and aquifer
2 (Sandy Loam)
3 (Silty Clay Loam)
1 (Silty Loam)
SSURGO (USDA, 2016)
POM
Percent organic matter in unsaturated zone
(percent)
0.701
0.978
0.876
SSURGO (USDA, 2016)
mean within 3-mile radius;
depends on soil texture
(ISTYPE1)
RHOB
Bulk density of unsaturated zone soil
(g/cm3)
1.6
1.67
1.65
Carsel and Parrish (1988)
Depends on soil texture
(ISTYPE1)
SATK
Saturated hydraulic conductivity of the
unsaturated zone (cm/hr)
2.302
0.017
0.112
FGD (US EPA, 2023b)
Median for ash in fills;
depends on soil texture
(ISTYPE1)
WCR
Residual water content of the unsaturated
zone (unitless)
0.065
0.089*
0.068
FGD (US EPA, 2023b),
*except silty clay loam had
no distribution in FGD, so
median from Carsel and
Parrish (1988)
Median based on soil texture
(ISTYPE1)
WCS
Saturated water content (effective porosity)
of the unsaturated zone (unitless)
0.41
0.43
0.45
Carsel and Parrish (1988)
Depends on soil texture
(ISTYPE1)
Aquifer Properties
Aquifer
Type
Aquifer type
2 (Bedded Sed.
Rock)
12 (Solution
Limestone)
10 (Un- & Semi-
consolidated Shallow
Surficial Aquifers)
Newell etal. (1990)
AL
Longitudinal dispersivity in the aquifer (m)
0.4437
EPACMTP (US EPA,
2003c)
estimated using Eqn. 5.11 in
source and distance to well
(XWELL) of 30 m, alpha_ref of
1 m
AT
Horizontal transverse dispersivity in the
aquifer (m)
0.05546
EPACMTP (US EPA,
2003c)
estimated using Eqn. 5.13 in
source (AL/8)
DRAFT
B-7
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
Parameter
Description & Units
Dry
Moderate
Wet
Reference
Comment
AV
Vertical transverse dispersivity in the aquifer
(m)
0.002773
EPACMTP (US EPA,
2003c)
estimated using Eqn. 5.14 in
source (AL/160)
BULKD
Aquifer soil bulk density (g/cm3)
2.184
2.554
1.558
EPACMTP (US EPA,
2003c)
calculated from porosity using
Eqn.5.6 in source; note Eqn.
5.6 has a typo; constant
(which represents soil particle
density) should 2.65 instead
of 2.851"
FOC
Fraction of organic carbon in saturated soils
(wt fraction)
0.004029
0.005621
0.005035
SSURGO (USDA, 2016)
calculated from POM of
vadose zone (POM/174)
GRADNT
Regional hydraulic gradient in the aquifer
(m/m)
0.0135
0.006
0.005
Newell etal. (1990)
median
POR
Volume fraction of connected pore space in
the aquifer (unitless)
0.176
0.0363
0.412
Wolff (1982)
mean for aquifer type
TEMP
Ambient groundwater temperature (C)
9.6
12
19.4
Collins (1925)
XKX
Hydraulic conductivity of saturated zone
(aquifer) (m/yr)
252.5
1580
2295*
Newell etal. (1990)
median for aquifer type,
*except Charleston [shallow
surficial aquifer], where a
mean value was used to avoid
water table mounding
ZB
Thickness of saturated zone (m)
21.3
18.9
7.62
Newell etal. (1990)
median
t Eqn 5.6, as corrected for a particle density of 2.65 and using the variable names here, is BULKD = 2.65 (1-POR). Note that the porosities of some of the locations are very low due to
the aquifer material, and hence the bulk density is relatively high compared to the particle density.
DRAFT
B-8
-------
PFOA/PFOS Risk Assessment Appendix B. Model Inputs
B.6 Surface Water Model Inputs (WWM)
Table B-8. Standard Index Reservoir Parameters
Parameter
Description & Units
Value
Reference
Comment
Area reserv
Reservoir area of the reservoir (m2)
52,555
US EPA (2019a)
WWM default
BNMAS
Reservoir benthic region areal concentration of biota (g/m2)
0.006
US EPA (2019a)
WWM default
bsp
Reservoir bed sediment porosity(fraction)
0.5
US EPA (2019a)
WWM default
Bulk_density
Reservoir benthic region bulk density (g/mL)
1.85
US EPA (2019a)
WWM default
CHL
Chlorophyll concentration in water column (mg/L)
0.005
US EPA (2019a)
WWM default
D over dx reserv
Mass transfer coefficient D/Ax (index reservoir) (m/s)
6E-09
US EPA (2019a)
WWM default
db
Depth of upper benthic layer in reservoir (m)
0.05
US EPA (2019a)
WWM default
DFAC
Photolysis parameter for reservoir
1.19
US EPA (2019a)
WWM default
D0C1
Concentration of dissolved organic carbon in water column (mg/L)
5
US EPA (2019a)
WWM default
D0C2
Concentration of dissolved organic carbon in benthic region (mg/L)
5
US EPA (2019a)
WWM default
dwc reservoir
Water column depth in the reservoir (m)
2.74
US EPA (2019a)
WWM default
foc_bs (FR0C2)
Fraction organic carbon for bed sediments (fraction)
0.04
US EPA (2019a)
WWM default
foc_sw(FROC1)
Fraction organic carbon for suspended sediments (fraction)
0.04
US EPA (2019a)
WWM default
PLMAS
Concentration of suspended biota (biomass) in water column
(mg/L)
0.4
US EPA (2019a)
WWM default
SUSED
Suspended solids concentration in water column (mg/L)
30
US EPA (2019a)
WWM default
Table B-9. Other WWM Inputs
Parameter
Description & Units
Value
Reference
Comment
Baseflow
Reservoir baseflow (m3/s)
0
Assumption
burialflag
Sediment burial flag: true = burial occurring and
removing chemical
TRUE
NA
Depth_0
Depth at which the input concentrations of physics
parameters were measured for reservoir (m)
2.74
Set to the
depth of the
waterbody
(dwc_reservoir)
Depthjnax
Maximum depth in the reservoir before overflow (m)
2.74
Set to the
depth of the
waterbody
(dwc reservoir)
Flow_averaging
Number of days that are used to average the influent
water in WWM (#)
30
NA
is_add_return_
frequency
Is alternative return frequency to be used in addition to
the 10-year return default for output?
FALSE
NA
is_calc_prben
Is fraction of mass going to sediment calculated (TRUE)
or prescribed by PRBEN (FALSE)?
TRUE
NA
DRAFT
B-9
-------
PFOA/PFOS Risk Assessment
Table B-10. Unused Chemical-specific VVWM Parameters
Appendix B. Model Inputs
Parameter
Description & Units
PFOA
PFOS
Reference
Comment
Heat_of_Henry
Enthalpy of phase transformation, aqueous
to air solution fJ/moll
NA
NA
NA
This assessment does not include
modeling transport through air
Kaer
Surface water column aerobic
biodegradation rate (1/day)
0
0
NA
Based on PFOA/PFOS degradation
literature
Kanaer
Sediment anaerobic degradation rate
(1/day)
0
0
NA
kh
Surface water hydrolysis rate (1/day)
0
0
NA
Kpo
Surface water photolysis rate (1/day)
0
0
NA
temp_ref_aer_all
Reference temperature for water column
degradation (C)
NA
NA
NA
Not used as all degradation rates
are zero
temp_ref_anae_all
Reference temperature for benthic
degradation (C)
NA
NA
NA
B.7 Food Chain Calculations
Table B-ll. Plant Uptake Parameters
Parameter
Description & Units
Value
Reference
Comment
MAF_expfruit
Moisture adjustment factor for exposed
fruit (% water)
85
EFH :2011 (US EPA,
2011)
Tables 9-37 (MAFs) and 13B-1 (crops
assigned to categories). Average of
MAFs for all crops in category
MAF_exveg
Moisture adjustment factor for exposed
vegetables (% water)
90
EFH :2011 (US EPA,
2011)
Tables 9-37 (MAFs) and 13B-1 (crops
assigned to categories). Average of
MAFs for all crops in category
MAFJorage
Moisture adjustment factor for forage (%
water)
80
MSU Extension (2011)
MAF_grain
Moisture adjustment factor for grain (%
water)
NA
NA
Not used: grain assumed to be
uncontaminated; see Section 2.9.3.4
MAF_profruit
Moisture adjustment factor for protected
fruit (% water)
87
EFH :2011 (US EPA,
2011)
Tables 9-37 (MAFs) and 13B-1 (crops
assigned to categories). Average of
MAFs for all crops in category
MAF_proveg
Moisture adjustment factor for protected
vegetables (% water)
81
EFH :2011 (US EPA,
2011)
Tables 9-37 (MAFs) and 13B-1 (crops
assigned to categories). Average of
MAFs for all crops in category
MAF_root
Moisture adjustment factor for root
vegetables (% water)
81
EFH :2011 (US EPA,
2011)
Tables 9-37 (MAFs) and 13B-1 (crops
assigned to categories). Average of
MAFs for all crops in category
MAF_silage
Moisture adjustment factor for silage (%
water)
65
NDSU Extension (2021)
VG_root
Empirical correction factor (root
vegetables) (fraction)
1
HHRAP (US EPA,
2005)
Adjustment factor for high log Kow
chemicals; Kow is not applicable to
PFOA/PFOS
Unused Plant-Air Pathway Parameters
Fw
Fraction of wet deposition adhering to
plant surface (fraction)
NA
NA
The conceptual model for this
assessment assumes no deposition
or diffusion to plants
KpPar
Plant surface loss coefficient
(particulate) (1/yr)
NA
NA
RpJX]
Interception fraction (by category of
aboveground plant) (fraction)
NA
NA
td
Time period of deposition (yrs)
NA
NA
Tp_[X]
Length of plant exposure to deposition
(by category of above ground plant)
(yrs)
NA
NA
VG_[X]
Crop yield (by category of aboveground
plant) (kg DW/m2)
NA
NA
Yp_[X]
Empirical correction factor (by category
of aboveground plant) (fraction)
NA
NA
DRAFT
B-10
-------
PFOA/PFOS Risk Assessment
Table B-12. Livestock Exposure Parameters
Appendix B. Model Inputs
Parameter
Description & Units
Value
Reference
Comment
Dairy Cows
Fforage
Fraction of forage contaminated (fraction)
1
Assumption
F grain
Fraction of grain contaminated (fraction)
0
Assumption
Assumes all grain is
uncontaminated
F silage
Fraction of silage contaminated (fraction)
1
Assumption
F water
Fraction of water contaminated (fraction)
1
Assumption
Qforage
Quantity of forage consumed by livestock (kg
DW/day)
13.2
HHRAP (US EPA, 2005)
Qgrain
Quantity of grain consumed by livestock (kg
DW/day)
3
HHRAP (US EPA, 2005)
Qsilage
Quantity of silage consumed by livestock (kg
DW/day)
4.1
HHRAP (US EPA, 2005)
Qsoil
Quantity of soil consumed by livestock (kg/day)
0.4
HHRAP (US EPA, 2005)
Qwater
Quantity of water consumed by livestock (L/day)
92
3MRA (US EPA, 2003b)
3MRA default
Beef Cattle
Fforage
Fraction of forage contaminated (fraction)
1
Assumption
F grain
Fraction of grain contaminated (fraction)
0
Assumption
Assumes all grain is
uncontaminated
F silaqe
Fraction of silage contaminated (fraction)
1
Assumption
F water
Fraction of water contaminated (fraction)
1
Assumption
Qforage
Quantity of forage consumed by livestock (kg
DW/day)
8.8
HHRAP (US EPA, 2005)
Qgrain
Quantity of grain consumed by livestock (kg
DW/day)
0.47
HHRAP (US EPA, 2005)
Qsilage
Quantity of silage consumed by livestock (kg
DW/day)
2.5
HHRAP (US EPA, 2005)
Qsoil
Quantity of soil consumed by livestock (kg/day)
0.5
HHRAP (US EPA, 2005)
Qwater
Quantity of water consumed by livestock (L/day)
53
3MRA (US EPA, 2003b)
3MRA default
Chickens (Laying Hens)
Fforaqe
Fraction of forage contaminated (fraction)
1
Assumption
F grain
Fraction of grain contaminated (fraction)
0
Assumption
Assumes all grain is
uncontaminated
F silaqe
Fraction of silage contaminated (fraction)
1
Assumption
F water
Fraction of water contaminated (fraction)
1
Assumption
Qforage
Quantity of forage consumed by livestock (kg
DW/day)
0.03
Dal Bosco et al. (2014)
Qgrain
Quantity of silage consumed by livestock (kg
DW/day)
0.074
Kowalczyk et al. (2020)
Qsilage
Quantity of grain consumed by livestock (kg
DW/day)
0.016
Kowalczyk et al. (2020)
Qsoil
Quantity of soil consumed by livestock (kg/day)
0.02
HHRAP (US EPA, 2005)
Qwater
Quantity of water consumed by livestock (L/day)
0.21
AECOM (2017)
DRAFT
B-ll
-------
PFOA/PFOS Risk Assessment Appendix B. Model Inputs
Table B-13. Transfer Factors for Food Chain Pathways
Parameter
Description & Units
PFOA
PFOS
Reference
Comment
Fish
BAF_T3F
Bioaccumulation factor for trophic level 3
fish filet ([mg/kg fishl/fmg/L waterl)
49
1700
US EPA (2024a)
BAF_T4F
Bioaccumulation factor for trophic level 4
fish filet (fmg/kg fishl/fmg/L waterl)
31
860
US EPA (2024a)
Plants
Br_Exfruit
Soil to plant uptake factor for exposed
fruit ([mg/kg DW plant]/[mg/kg soil])
0.13
0.03
PFOA: Blaine et al. (2013,2014); Lechnerand Knapp
(2011); PFOS: Blaine etal. (2014)
PFOA: median of tomatoes, sugar snap peas,
cucumbers; pot studies; PFOS: sugar snap peas,
pot study
Br_ExVeg
Soil to plant uptake factor for exposed
vegetables ([mg/kg DW plant]/[mg/kg
soill)
1.5
0.11
PFOA: Blaine etal. (2013,2014); PFOS: Blaine etal.
(2013)
PFOA: median of lettuce, celery; pot studies; PFOS:
lettuce; field study
Br_Forage
Soil to plant uptake factor for forage
([mg/kg DW plantl/[mg/kg soill)
0.29
0.08
Yooetal. (2011)
grass; field study
Br_Grain
Biotransfer factor (soil to grain) (mg/kg
DW plantl/[mg/kg soil)
NA
NA
NA
Not used; all grain is assumed to be
uncontaminated, see Section 2.9.3.4
Br_Profruit
Soil to plant uptake factor for protected
fruit ([mg/kg DW plant]/[mg/kg soil])
0.13
0.03
PFOA: Blaine et al. (2013,2014); Lechnerand Knapp
(2011); PFOS: Blaine etal. (2014)
PFOA: median of tomatoes, sugar snap peas,
cucumbers; pot studies; PFOS: sugar snap peas,
pot study
Br_Proveg
Soil to plant uptake factor for protected
vegetables ([mg/kg DW plant]/[mg/kg
soill)
1.5
0.11
PFOA: Blaine etal. (2013,2014); PFOS: Blaine etal.
(2013)
PFOA: median of lettuce, celery; pot studies; PFOS:
lettuce; field study
Br_Root
Soil to plant uptake factor for root
vegetables ([mg/kg DW plant]/[mg/kg
soill)
0.73
0.8
Blaine et al. (2014); Lechner and Knapp (2011); Wen et
al. (2016)
median of carrots, potatoes, radish; pot studies
Br_Silage
Soil to plant uptake factor for silage
([mg/kg DW plantl/[mg/kg soill)
0.29
0.08
Yoo etal. (2011)
grass; field study
Bv
Biotransfer factor (vapor phase air to
plant) (ug/g DW plantl/[ug/g air)
NA
NA
NA
Volatilization is not expected under environmental
conditions (see Section 2.2.2)
Animal Products
Bs
Bioavailability of chemical in soil relative
to plants (fraction)
1
1
HHRAP (US EPA, 2005)
Reflects the efficiency of transfer of contaminants
from soil to livestock relative to transfer from plants
to livestock. HHRAP cites inadequate data to set
this to anything other than 1
BTF_beef
Biotransfer factor for beef ([mg/kg
WW]/[kg DW/day])
0.01
0.18
PFOA: Vestergren etal. (2013)
PFOS: Drew etal. (2021)
PFOA: Dairy cattle
PFOS: Beef cattle
BTF_eggs
Biotransfer factor for eggs ([mg/kg
WW]/[kg DW/day])
8.6
21
Wilson etal. (2021)
Laying hens
BTF_milk
Biotransfer factor for milk ([mg/kg
WW]/[kg DW/day])
0.01
0.02
Vestergren et al. (2013)
Dairy cattle
DRAFT
B-12
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
Parameter
Description & Units
PFOA
PFOS
Reference
Comment
BTF_poultry
Biotransfer factor for chicken ([mg/kg
WWl/[kg DW/dayl)
0.2
2.2
Kowalczyk et al. (2020)
Laying hens
B.8 Human Exposure Calculations
Table B-14. Exposure Factors
Parameter
Description & Units
Child
(1-11yrs)
Adult
Farmer
Reference
Comment
BW
Body weight (kg)
21
80
EFH:2011 (US EPA, 2011)
Table 8-1, mean (general population)
CR_beef
Daily human consumption rate of beef (g
VWV/kg BW/day)
2.1
1.6
EFH:2011 (US EPA, 2011)
Tbl 13-33,50th percentile
CR_dw
Daily human consumption rate of water
(mL/kg-day)
14
13.4
EFH:2019 drinking water update (US EPA,
2019b)
Tbl 3-21,50th percentile
CR_eggs
Daily human consumption rate of eggs (g
VWV/kg BW/day)
0.7
0.7
EFH:2011 (US EPA, 2011)
Tbl 13-40,50th percentile, households that
farm (all ages)
CR_exfruit
Daily human consumption rate of exposed
fruit (g VWV/kg BW/day)
1.33
1.3
EFH:2011 (US EPA, 2011)
Tbl 13-58,50th percentile
CR_exveg
Daily human consumption rate of exposed
vegetables (g VWV/kg BW/day)
1
1.4
EFH:2011 (US EPA, 2011)
Tbl 13-60,50th percentile
CRJish
Daily human consumption rate of fish (g
VWV/kg BW/day)
0.55
0.47
EFH:2011 (US EPA, 2011)
Tbl 13-20,50th percentile, no data for 1-5, so
based on 6-11; adult based on households that
fish (all ages)
CR_milk
Daily human consumption rate of milk (g
VWV/kg BW/day)
22
12
Children: EFH: 2018 meat & dairy update
(US EPA, 2018); Adult: EFH:2011 (US EPA,
2011)
Child: Tbl 11-4,50 th percentile; Adult: Tbl 13-
25, 50th percentile
CR_poultry
Daily human consumption rate of poultry (g
VWV/kg BW/day)
2
1.1
Children: EFH: 2018 meat & dairy update
(US EPA, 2018); Adult: EFH:2011 (US EPA,
2011)
Child: Tbl 11-6, mean; Adult: Tbl 13-52, 50th
percentile, households that farm (all ages)
CR_profruit
Daily human consumption rate of protected
fruit (g VWV/kg BW/day)
2.3
2.1
EFH:2011 (US EPA, 2011)
Tbl 13-59,50th percentile
CR_proveg
Daily human consumption rate of protected
vegetables (g VWV/kg BW/day)
1.1
0.6
EFH:2011 (US EPA, 2011)
Tbl 13-61,50th percentile
CR_root
Daily human consumption rate of below
ground vegetables (g VWV/kg BW/day)
0.59
0.88
EFH:2011 (US EPA, 2011)
Tbl 13-62,50th percentile
CRs
Daily human incidental soil ingestion rate
(mg/day)
40
10
EFH:2017 soil update (US EPA, 2017b)
Table 5-1; data for soil only, which includes
outdoor settled dust
Fi
Fraction of human diet item /'contaminated
(fraction)
1
1
Assumption
Assumes all food items in the category
contaminated
F_T3
Fraction of fish consumed that is trophic level
3 (fraction)
0.14
0.14
EFH:2011 (US EPA, 2011)
Table 10-74
DRAFT
B-13
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
Parameter
Description & Units
Child
(1-11yrs)
Adult
Farmer
Reference
Comment
F_T4
Fraction of fish consumed that is trophic level
4 (fraction)
0.86
0.86
EFH:2011 (US EPA, 2011)
Table 10-74
Li
Food preparation or cooking loss for diet item
/' (fraction)
0
0
Assumption
Assumes no cooking or prep losses
DRAFT
B-14
-------
PFOA/PFOS Risk Assessment Appendix B. Model Inputs
B.9 Risk Calculations
Table B-15. Cancer Dose Inputs
Parameter
Description & Units
Value
Reference
Comment
AT
Averaging time for cancer risk (yr)
70
RAGS Pt A (US
EPA, 1989a)
ED
Exposure duration (yr)
10
EFH:2011 (US EPA,
2011)
Based on residential mobility data, Tbl 16-113
(farmers), 50th percentile; also used for nearby
residents: the 50th percentile for general
population, all ages, from Table 16-108 is 9 yrs,
so this is a reasonable value for nearby residents
as well. Value used for children as well, assuming
whole family has same exposure duration.
EF
Exposure frequency (day/yr)
350
Policy
Table B-16. Toxicity Inputs
Parameter
Description & Units
PFOA
PFOS
Reference
Comment
CSForal
Oral cancer slope factor ([mg/kg/dayl1)
29,300
39.5
US EPA (2024b)
Final PFOA-PFOS tox values
RfD
Reference dose (mg/kg/day)
3E-08
1E-07
US EPA (2024b)
Final PFOA-PFOS tox values
DRAFT
B-15
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
B.10 References
AECOM. (2017). Plant PFAS Uptake Study: RAAF Base Williamtown Stage 2B Environmental
Investigation. Prepared for the Australian Department of Defence. 29 November. Available at
https://web.archive.org.au/awa/20230427171827mp /https:/defence.gov.au/Environment/PFAS/d
ocs/Williamtown/Reports/HHRAReports/20171205HH RADecember2017Volume2.pdf
Blaine, A. C., Rich, C. D., Hundal, L. S., Lau, C., Mills, M. A., Harris, K. M. & Higgins, C. P. (2013). Uptake of
perfluoroalkyl acids into edible crops via land applied biosolids: field and greenhouse studies.
Environmental Science & Technology, 47, 14062-14069.
Blaine, A. C., Rich, C. D., Sedlacko, E. M., Hundal, L. S., Kumar, K., Lau, C., Mills, M. A., Harris, K. M., &
Higgins, C. P. (2014). Perfluoroalkyl acid distribution in various plant compartments of edible crops
grown in biosolids-amended soils. Environmental Science & Technology, 48, 7858-7865.
Campos-Pereira, H., Kleja, D. B., Ahrens, L., Enell, A., Kikuchi, J., Pettersson, M., & Gustafsson, J. P.
(2023). Effect of pH, surface charge and soil properties on the solid-solution partitioning of
perfluoroalkyl substances (PFASs) in a wide range of temperate soils. Chemosphere, 321, 138133-
138141.
Carsel, R. F., & Parrish, R. S. (1988). Developing joint probability distributions of soil water retention
characteristics. Water Resources Research, 24(5), 755-769. May.
Carsel, R. F., Parrish, R. S., Jones, R. L., Hansen, J. L., & Lamb, R. L. (1988). Characterizing the uncertainty
of pesticide leaching in agricultural soils. Journal of Contaminant Hydrology, 2, 111-124.
Chen, X., Yu, P., Xiang, L., Zhao, H., Li, Y., Li, H., Zhang, X., Cai, Q„ Mo, C., & Wong, M. H. (2020).
Dynamics, thermodynamics, and mechanism of perfluorooctane sulfonate (PFOS) sorption to
various soil particle-size fractions of paddy soil, Ecotox. and Env. Safety, 206, 111105-111116.
Clapp, R. B., & Hornberger, G. M. (1978). Empirical equations for some soil hydraulic properties. Water
Resources Research, 14, 601-604.
Collins, W.D. (1925). Temperature of Water Available for Industrial Use in the United States. United
States Geological Survey, Water Supply Paper 520-F. Available online at
https://pubs.usgs.gov/publication/wsp520F: map converted to degrees C at
https://www3.epa.gov/ceampubl/learn2model/part-two/onsite/tempmap.html
Dal Bosco, A., Mugnai, C., Rosati, A., Paoletti, A., Caporali, S. & Castellini, C. (2014). Effect of range
enrichment on performance, behavior, and forage intake of free-range chickens. Journal of Applied
Poultry Research, 23(2), 137-145
Drew, R., Hagen, T.G., & Champness, D. (2021). Accumulation of PFAS by livestock - determination of
transfer factors from water to serum for cattle and sheep in Australia. Food additives &
Contaminants: Part A. https://doi.org/10.1080/19440049.2Q21.1942562.
Dunne, T., & Leopold, L. B. (1978). Water in Environmental Planning. New York: W.H. Freeman and
Company.
Gunn, A. P, Dewhurst, R. E., Giorgetti, A., Gillott, N. L., Wishart, S. J. W., & Pedley, S. (2004). Use of
Sewage Sludge Products in Construction. London: CIRIA.
Hubert, M., Arp, H.P.H., Hansen, M. C., Castro, G., Meyn, T., Asimakopoulos, A. G., & Hale, S. E. (2023).
Influence of grain size, organic carbon and organic matter residue content on the sorption of per-
DRAFT
B-16
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
and polyfluoroalkyl substances in aqueous film forming foam contaminated soils - Implications for
remediation using soil washing. Science of The Total Environment, 875, 162668-162680.
Johnson, R., Anschutz, A. J., Smolen, J. M., Simcik, M. F., & Penn, R. L. 2007. "The Adsorption of
Perfluorooctane Sulfonate onto Sand, Clay, and Iron Oxide Surfaces. "Journal of Chemical
Engineering Data 52: 1165-1170.
Kowalczyk, J., Gockener, B., Eichhorn, M., Kottholff, M., Bucking, M., Schafft, H., Lahrssen-Wiederholt,
M., & Numata, J. (2020). Transfer of per- and polyfluoroalkyl substances (PFAS) from feed into the
eggs of laying hens. Part 2: Toxicokinetic results including the role of precursors. J. Agric. Food
Chem., 68, 12539-12548.
Lechner, M. & Knapp, H. (2011). Carryover of perfluorooctanoic acid (PFOA) and perfluorooctane
sulfonate (PFOS) from soil to plant and distribution to the different plant compartments studied in
cultures of carrots (Daucus carota ssp. Sativus), potatoes (Solanum tuberosum), and cucumbers
[Cucumis sativus). Journal of Agricultural and Food Chemistry, 59, 11011-11018.
Mills, W. B., Porcella, D. B., Ungs, M. J., Gherini, S. A., Summers, K. V., Mok, L., Rupp, G. L., & Bowie, G. L.
(1985). Water Quality Assessment: A Screening Procedure for Toxic and Conventional Pollutants,
Part I. Revised. EPA-600/5-85/002a. Prepared by Tetra Tech, Inc., for US EPA, Office of Research and
Development, Environmental Research Laboratory, Athens, GA.
MSU Extension (Michigan State University Extension). (2011). Understanding how hay dries in the field.
https://www.canr.msu.edu/news/understanding how hay dries in the fieldl
NDSU Extension (North Dakota State University Extension). (2021). Moisture Matters When Making
Good Corn Silage, https://www.drovers.com/news/beef-production/moisture-matters-when-
making-good-corn-silage
Newell, C. J., Hopkins, L. P., & Bedient, P. B. (1990). A hydrogeologic database for ground-water
modeling. Groundwater 28(5), 703-714.
Schroeder, E. D. (1977). Water and wastewater treatment. McGraw-Hill Series in Water Resources and
Environmental Engineering. Chow, V.T., Eliassen, R., and Linsley, R.K. (eds.) New York, NY: McGraw-
Hill, Inc. 156pp.
Sopper, W. E. (1993). Municipal Sludge Use in Land Reclamation. Boca Raton, FL: Lewis Publishers.
SRC (Syracuse Research Corporation). 2016. Physprop (Physical Properties Database).
http://www.srcinc.com/what-we-do/databaseforms.aspx?id=386
Tchobanoglous, G., Cerra, F. J., and Maisel, J. W. (eds.). (1979). Wastewater Engineering: Treatment,
Disposal, Reuse. 2nd Edition. McGraw-Hill Book Company, New York, NY.
USDA (Department of Agriculture). (1986). Urban Hydrology for Small Watersheds. TR-55. Natural
Resources Conservation Service. June.
USDA (Department of Agriculture). (1994). State Soil Geographic (STATSGO) Data Base. Data use
information. Miscellaneous Publication Number 1492. Natural Resources Conservation Service, Fort
Worth, TX. December.
USDA (Department of Agriculture). (2014). 2012 Census of Agriculture. United States Summary and State
Data. Volume 1-Geographic Area Series-Part 51. May.
DRAFT
B-17
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
USDA (Department of Agriculture). (2016). Soil Survey Geographic (SSURGO) Database. Natural
Resources Conservation Service. Available online at: sdmdataaccess.sc.egov.usda.gov. Last accessed
on 11/1/16.
US DOC (Department of Commerce) and US DOE (Department of Energy) National Renewable Energy
Laboratory. (1993). Solar and Meteorological Surface Observation Network (SAMSON), 1961-1990.
Version 1.0. National Climatic Data Center, Asheville, NC.
US EPA (Environmental Protection Agency). (1989a). Risk Assessment Guidance for Superfund. Volume I:
Human Health Evaluation Manual (Part A) (Interim Final). EPA/540/1-89/002. Prepared by US EPA,
Office of Emergency and Remedial Response, Washington, DC.
US EPA (Environmental Protection Agency). (1989b). Hazardous Waste TSDF- Fugitive Particulate Matter
Air Emissions Guidance Document.
US EPA (Environmental Protection Agency). (1995). Compilation of Air Pollutant Emission Factors.
Volume I: Stationary Point and Area Sources, Fifth Edition. Office of Air Quality Planning and
Standards, Research Triangle Park, NC. January.
US EPA (Environmental Protection Agency). (2003a). Technical Background Document for the Sewage
Sludge Exposure and Hazard Screening Assessment. US EPA, Office of Water, Washington, DC. EPA-
822-B-03-001. December 19.
US EPA (Environmental Protection Agency). (2003b). Multimedia, Multipathway, and Multireceptor Risk
Assessment (3MRA) Modeling System, Volume II: Site-based, Regional, and National Data.
EPA/530/D-03/001b. US EPA, Office of Solid Waste and Emergency Response, Washington, DC.
US EPA (Environmental Protection Agency). (2003c). EPA's Composite Model for Leachate Migration
with Transformation Products (EPACMTP): Parameters/Data Background Document. Office of Solid
Waste, Washington, DC. April.
US EPA (Environmental Protection Agency). (2005). Human Health Risk Assessment Protocol (HHRAP) for
Hazardous Waste Combustion Facilities, Final. Office of Solid Waste and Emergency Response
(5305W). EPA530-R-05-006. September.
US EPA (Environmental Protection Agency). (2009). Targeted National Sewage Sludge Survey Sampling
and Analysis Technical Report. EPA-822-R-08-016. Office of Water, Office of Science and Technology,
Washington, DC. Available at http://water.epa.gov/scitech/wastetech/biosolids/biosolids index.cfm
US EPA (Environmental Protection Agency). (2011). Exposure Factors Handbook: 2011 Edition.
EPA/600/R-090/052F. National Center for Environmental Assessment, Office of Research and
Development, Washington, DC. September. Available at http://www.epa.gov/ncea/efh/pdfs/efh-
complete.pdf.
US EPA (Environmental Protection Agency). (2016). EPA On-line Tools for Site Assessment Calculation:
Estimated Diffusion Coefficients. Ecosystems Research, Athens, GA.
https://www3.epa.gov/ceampubl/learn2model/part-two/onsite/estdiffusion-ext.html
US EPA (Environmental Protection Agency). (2017a). Technical Fact Sheet—Perfluorooctane Sulfonate
(PFOS) and Perfluorooctanoic Acid (PFOA). EPA 505-F-17-001. Office of Land and Emergency
Management, Washington, DC. November.
US EPA (Environmental Protection Agency). (2017b). Update for Chapter 5 of the Exposure Factors
Handbook: Soil and Dust Ingestion. EPA/600/R-17/384F. National Center for Environmental
DRAFT
B-18
-------
PFOA/PFOS Risk Assessment
Appendix B. Model Inputs
Assessment, Office of Research and Development, Washington, DC. September. Available at
https://www.epa.gov/expobox/about-exposure-factors-handbook.
US EPA (Environmental Protection Agency). (2018). Update for Chapter 11 of the Exposure Factors
Handbook: Intake of Meats, Dairy Products, and Fats. EPA/600/R-17/485F. National Center for
Environmental Assessment, Office of Research and Development, Washington, DC. April. Available
at https://www.epa.gov/expobox/about-exposure-factors-handbook
US EPA (Environmental Protection Agency). (2019a). The Variable Volume Water Model, Revision B. US
Environmental Protection Agency, Office of Pesticide Programs, Washington DC. September.
USEPA/OPP 734S16002. Available at https://www.epa.gov/pesticide-science-and-assessing-
pesticide-risks/models-pesticide-risk-assessment
US EPA (Environmental Protection Agency). (2019b). Update for Chapter 3 of the Exposure Factors
Handbook: Ingestion of Water and Other Select Liquids. EPA/600/R-18/259F. National Center for
Environmental Assessment, Office of Research and Development, Washington, DC. February.
Available at https://www.epa.gov/expobox/about-exposure-factors-handbook
US EPA (Environmental Protection Agency). (2023a). Biosolids Tool (BST) User's Guide: Version 1. Office
of Water, Washington, DC. February.
US EPA (Environmental Protection Agency). (2023b). Beneficial Use Evaluation: Flue Gas Desulfurization
Gypsum as an Agricultural Amendment. EPA 530-R-23-004. Office of Land and Emergency
Management, Washington, DC. March.
US EPA (Environmental Protection Agency). (2024a). Draft Human Health Criteria for PFOA and PFOS.
date.
US EPA (Environmental Protection Agency). (2024b). FINAL Candidate Toxicity Values for PFOA and
PFOS. date.US NLM (National Library of Medicine). 2010. Hazardous Substances Data Bank (HSDB).
Available at http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgenPHSDB
U.S. NLM (National Library of Medicine). 2010. Hazardous Substances Data Bank (HSDB). Available at
http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgenPHSDB. van der Leeden et al., 1990, Water Enc.
Vestergren, R., Orata, F., Berger, U., & Cousins, I. T. (2013). Bioaccumulation of perfluoroalkyl acids in
dairy cows in a naturally contaminated environment. Environmental Science and Pollution Research,
20, 7959-69.
Wanielista M. P. & Yousef, Y. A. (1993). Stormwater Management. New York, NY. John Wiley & Sons, Inc.
Wen, B., Wu, Y., Zhang, H., Liu, Y., Hu, X., Huang, H., & Zhang, S. (2016). The roles of protein and lipid in
the accumulation and distribution of perfluorooctane sulfonate (PFOS) and perfluorooctanoate
(PFOA) in plants grown in biosolids-amended soils. Environmental Pollution, 216, 682-688.
Wilson, T., Stevenson, G., Crough, R., de Araujo, J., Fernando, N., Anwar, A., Scott, T., Quinteros, J., Scott,
P., & Archer, M. (2021). Evaluation of residues in hen eggs after exposure of laying hens to water
containing per-and polyfluoroalkyl substances. Env. Tox. and Chem., 40(3), 735-743
Wischmeier, W.H., & Smith, D. D. (1978). Predicting Rainfall Erosion Losses: A Guide to Conservation
Planning. USDA Agricultural Handbook No. 537. Washington, DC. Department of Agriculture. 62pp.
Wolff, R.G. (1982) Physical Properties of Rocks—Porosity, Permeability, Distribution Coefficients, and
Dispersivity. USGS Water Resources Investigation Open File Report 82-166.
DRAFT
B-19
-------
PFOA/PFOS Risk Assessment Appendix B. Model Inputs
Yoo, H., Washington, J. W., Jenkins, T. M., & Ellington, J. J. (2011). Quantitative determination of
perfluorochemicals and fluorotelomer alcohols in plants from biosolid-amended fields using
LC/MS/MS and GC/MS. Environ. Sci. Techno!., 45(19), 7985-7990.
DRAFT
B-20
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
APPENDIX C. GROUNDWATER MODELING
This section describes the refined groundwater modeling used to predict the fate and transport of
PFOA/PFOS present in land-applied biosolids and biosolids disposed in surface disposal units to
underlying soils and groundwater to determine impacts on drinking water wells that are connected to
groundwater. Sections C.l and C.2 provides a discussion on model selection, an overview of the
conceptual model, including the basic approach and assumptions. Section C.3 discusses the input
parameters and values used in this risk analysis. Section C.4 discusses the model outputs. Appendix B
provides additional information about the inputs used in modeling the groundwater pathway using
EPACMTP (US EPA, 2003a,b,d; 1997).
C.l Model Selection
The groundwater pathway was modeled for this analysis to estimate the receptor concentrations that
result from a predicted release of PFOA/PFOS from land-applied biosolids and sewage sludge disposed
in surface disposal units. The release of PFOA/PFOS occurs when these pollutants in land-applied wastes
or in sewage sludge stored in surface disposal units percolate through soils and into the subsurface. The
releases of pollutant mass and infiltrating water were determined using waste management unit-
specific models (land application unit, or LAU, and surface disposal unit, or SDU) developed for 3MRA, as
described in assessment Section 2.9. These models generate time-series of pollutant mass fluxes and
infiltrating water fluxes to the subsurface as well as releases to other exposure pathways, the latter a
capability not available in the source terms provided in the groundwater model, EPACMTP. Therefore, to
satisfy the multi-pathway analysis plan for this risk assessment, the 3MRA waste management unit
models (LAU and SDU) are used to provide mass and water fluxes to EPACMTP for fate and transport
simulations of the subsurface environment.
PFOA/PFOS in the land-applied wastes or leaching from sludge stored in surface disposal units are
transported via aqueous-phase migration through the unsaturated zone (i.e., the soil layer beneath the
area of waste application and subsurface above the groundwater table) to the underlying saturated
zone (i.e., groundwater), and then down-gradient to a hypothetical residential drinking water well
located 5 meters from the edge of the farm field (i.e., center of the buffer). For this analysis, the
exposure concentration was evaluated as the peak concentration at the intake point of the drinking
water well (hereafter referred to as the receptor well). Figure C-l shows the conceptual model of the
groundwater fate and the transport of contaminant releases from either a LAU or a SDU to a down-
gradient receptor well with associated dilution and attenuation. Details about the modeled receptor
well are provided later in this section.
DRAFT
C-l
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
LEACHATE CONCENTRATION
¦SURFACE DISPOSAL UNIT
UNSATURATED
ZONE
SATURATED
ZONE
LEACHATE PLUME
Figure C-l. Schematic diagram of groundwater modeling scenario.
C.1.1 Groundwater Model Selection
The mobility of PFAS in the environment, an active area of research, is known to be affected by their
hydrophobic/hydrophilic-surfactant behavior (e.g., fluid-fluid interface retention), attraction to the solid
phase in sediment (Higgins and Luthy, 2006; Liu et al., 2005), sludge (Milinovic et al., 2016), soil
(Milinovic et al., 2015), to organic carbon in general (Higgins and Luthy, 2006), ionic behavior as a
function of pH (Place and Field, 2012; Pereira et al. 2018), and the competition between these
processes. Methodologies for assessing the impact of PFAS retention at the air-water interface (AWI)
have been proposed (Brusseau, 2018; Zhang and Guo, 2024), modeled (Guelfo et al., 2020), and
implemented in various fate and transport simulators (Guo et al., 2020; Silva et al., 2020; Guo et al.
2022).
Three simulation models were examined to determine which is best suited to support risk assessment
objectives when assessing PFOA and PFOS:
• EPA's Composite Model for Leachate Migration with Transformation Products (EPACMTP, US
EPA, 2003a&b). EPACMTP is EPA's conventional groundwater model and has been the
traditional model used for both probabilistic and deterministic simulations of contaminant
migration through the vadose zone to groundwater.
• HYDRUS ID with HD1 Pro Module (ver.5.01; Silva et al, 2020). This is a deterministic model that
includes a new AWI retention model developed specifically to address PFAS fate and transport.
This model will be referred to as HYDRUS.
• A recently published analytical PFAS leaching model (Guo et al., 2022). This model includes
some simplifying assumptions on the processes incorporated into the HYDRUS ID PFAS module.
This model will be referred to as ANALYTICAL.
Predictions of contaminant concentrations at the water table of an unconfined aquifer generally depend
on two major processes within the vadose zone: flow and transport. For surfactants like PFOA and PFOS,
transport processes that may occur when released into the subsurface include retention at the AWI,
surfactant enhanced flow (e.g., Guo et al., 2020; Silva et al., 2020), self-assembly during sorption (e.g.,
Kalam et al., 2021), and enhanced transport of co-contaminants through emulsions (e.g., Kostarelos et
DRAFT
C-2
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
al., 2021) and micelles (e.g., An et al., 2002; Simmons and McGuffin, 2007). Except for AWI, these other
transport processes may be excluded on account of assuming relatively "small" PFAS concentrations in
biosolids (e.g., formation of micelles) and exclusion of mixed wastes (e.g., transport of co-
contaminants). Table C-l shows how each of the above models handles flow, transport and AWI
processes.
Table C-l. Evaluated Models and How Major Processes Are Handled
Major
Processes
EPACMTP
HYDRUS 1Dw/ PFAS Module
(HYDRUS)
Analytical PFAS Model
(ANALYTICAL)
Flow
Steady state variable
saturated flow
Steady state and transient
variable saturated flow
Steady state unsaturated flow
Transport
Transient transport with
linear equilibrium
partitioning
Transient transport with linear
equilibrium partitioning and AWI
retention
Transient transport with equilibrium
partitioning, AWI and kinetic solid-
phase sorption
AWI
No
Yes
Yes
C. 1.2 Approach to Model Selection Evaluation
The general approach to evaluating these models uses the land application unit (LAU) module (US EPA,
2003c) as a source term for unsaturated zone flow and transport simulations. The LAU module was
developed to estimate annual average surface soil constituent concentrations and constituent mass
release rates to the air, downslope land, and groundwater. The model simulates the vertical movement
of pollutants within the agricultural land (releases through leaching to groundwater), volatile and
particle releases to the air, and horizontal movement of pollutants (runoff and erosion from the
agricultural land across any buffer area to a nearby waterbody). The model has the ability to consider
losses from agricultural land due to hydrolysis and biodegradation, as well as leaching, volatilization, and
particle emissions due to tilling (mixing) operations and wind erosion. LAU produces the following
outputs resulting from land-applied biosolids to be used as inputs to the vadose zone models under
consideration:
• Annual leach flux (g/m2-day)
• Annual infiltration (m/day)
• Annual leachate concentrations (mg/L); these are computed as the ratio of the annual leach flux
and the annual infiltration and used as input for the upper boundary condition for the transport
models.
The LAU module can be used simulate both crop and pasture scenarios, which both reflect biosolids
applied at an agronomic rate to a field and differ only in whether biosolids are tilled into the soil (crop)
or not (pasture). This assessment uses the crop scenario for this evaluation as tilling reduces the amount
of available contaminant mass to move off the field, maximizing the amount of mass that can leach to
the subsurface.
C.1.3 Scenarios Selected for Model Evaluation
To fully evaluate differences among the models, EPA developed eight basic scenarios that reflect a broad
range of key hydrogeologic conditions. Specifically, this assessment considered two different values that
represent a range of national conditions for each of three parameters:
• Depth to water table
• Soil texture
• Meteorological setting.
DRAFT
C-3
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Depth to Water Table. This parameter, also called vadose zone thickness, defines the modeling region,
so relatively short and long values were used to capture a range. Hydrogeologic environment data for
national modeling of the groundwater pathway are sourced from the Hydrogeologic Database for
Groundwater Modeling (Newell et al., 1990). Figure C-2 summarizes those data for a variety of settings.
Based on those data, we selected depths of 1 m and 10 m to capture roughly the second and third
quartiles: this provides a range of values without including extreme values.
jf
s?
.o\*
s-S*
A*"
~ Q1
¦ Min
• Mean
A Median
X Max
X Q3
Figure C-2. Depth to water table data from HGDB.
Soil Texture. The relationship between infiltration (or pressure head) and the water content for a
particular soil is known as the soil-water characteristic curve and is a key parameter needed to solve the
governing flow equations in the unsaturated zone. These curves differ for different soil textures. Two
reference soil texture types, loam and loamy sand, were chosen for model comparisons because they
represent a broad range in saturated hydraulic conductivities that would likely result in significant
contaminant mass transport to the water table; this helps evaluate the conservatism of each model and
if the soil water characteristic curves used in the two models are similar. All three models evaluated use
the empirical function proposed by Mualem (1976) and van Genuchten (1980) to estimate unsaturated
hydraulic conductivity for both soil textures. This empirical function estimates the unsaturated hydraulic
conductivity using the residual and saturated water contents (0r, 0s) along with empirical Van
Gneuchten parameters, a and (B, that are obtained from characteristic soil-water retention curves for
each soil type. Table C-2 shows the values used for these soil properties; the same values were used for
all three models evaluated.
DRAFT
C-4
-------
PFOA/PFOS Risk Assessment Appendix C. Groundwater Modeling
Table C-2. Soil Parameter Values Used
Soil Column Property
Notation
Units
Loam
Loamy
Sand
Depth to water table
-
m
1 or 10
1 or 10
Residual water content
0r
dimensionless
0.078
0.057
Saturated water content
9s
dimensionless
0.43
0.41
Saturated hydraulic conductivity
Ks
cm/hr
1.04
14.59
Van Genuchten parameter
a
cm-1
0.036
0.124
Van Genuchten parameter
P
dimensionless
1.56
2.28
Bulk Density
P
g/cc
1.33
1.65
Dispersivity
ai_
m
0.1 or 1
0.1 or 1
Percent organic matter
%OM
dimensionless
0.174
0.174
Fraction organic carbon
foe
dimensionless
0.001
0.001
Meteorological Setting. Both models simulate the soil water content as a function of infiltration (or
pressure head) using the Van Genuchten model (1980) but using two infiltration scenarios can help
evaluate whether both models simulate long term average flow conditions similarly given varying
infiltration or recharge inputs. This assessment uses three meteorological settings: wet, moderate, and
dry. For this evaluation, we used the wet and dry settings, as they represent bounding conditions. Ten
years of meteorologic data from Charleston, SC, and Boulder, CO, are cycled 15 times to represent wet
and dry meteorology, respectively, for 150-year simulations. In summary, the eight basic scenarios are
presented Table C-3.
Table C-3. Modeling Scenarios
Depth to Water
Table
Soil Type
Meteorological
Conditions
1 m
Loam
Wet
Dry
Loamy Sand
Wet
Dry
10 m
Loam
Wet
Dry
Loamy Sand
Wet
Dry
EPACMTP can simulate only linear, instantaneous solid-phase adsorption; HYDRUS and ANALYTICAL can
be run only assuming instantaneous and kinetic effects of adsorption (no AWI effects) or also including
AWI effects. While HYDRUS can only model instantaneous linear solid-phase adsorption, ANALYTICAL
can model both instantaneous and kinetic effects of adsorption. Kinetics associated with solid-phase
adsorption were shown to be present in both batch and miscible -displacement experiments. Further,
both HYDRUS and ANALYTICAL can model AWI effects using different values of the equilibrium
distribution constant between the liquid phase and air-water interface (Kh). HYDRUS and ANALYTICAL
were run assuming linear, solid-phase adsorption mode and with and without AWI effects for three
different values of Kh, however, the input specifications required to simulate AWI effects differ between
the HYDRUS and ANALYTICAL models: HYDRUS calculates Kh using a Langmuir approach whereas Kh is
directly specified in ANALYTICAL.
Constituent Transport Parameters. The models require various constituent-specific transport factors.
Table C-4 presents the values used for PFOA and PFOS and indicates which of the three models
evaluated use them.
DRAFT
C-5
-------
PFOA/PFOS Risk Assessment Appendix C. Groundwater Modeling
Table C-4. Constituent Transport Parameters Including AWI and Sources
Used in Models
Values
Chemical Property
Notation
Units
EPACMTP
HYDRUS
ANALYTICAL
PFOA
PFOS
Reference/Notes
Diffusion coefficient in
Diffin
m2/yr
•
•
•
0
0
water
H20
Organic partition
coefficient
Koc
mL/g
•
•
•
1.99E+03
1.86E+04
Silva et al., 2020
Solid-phase
Kd
mL/g
•
•
•
1.99
18.60
Silva et al., 2020
(instantaneous)
adsorption coefficient
Langmuir adsorption
isotherm maximum
Tmax
mol/cm2
—
•
—
5.54E-07
3.50E-07
Only applicable for
HYDRUS (Silva et
interfacial adsorbed
al., 2020). Kh AWI
(AWI) concentration
directly specified in
Analytical model.
Langmuir coefficient
for AWI adsorption
KL_aw
cm3/mol
•
6.67E+03
1.37E+05
Silva et al., 2020.
Kh_AWI directly
specified in
Analytical model.
Fraction of sorbent for
Fs
fraction
—
—
•
Guo et al., 2022
which sorption is
instantaneous
(only applicable to
two-domain solid-
phase sorption
models)
First order
as
—
—
•
rate constant for
kinetic sorption
Equilibrium distribution
Kh AWI
cm
—
•
•
3.69E-03
4.79E-02
Silva et al., 2020
constant between
calculated Kh AWI
liquid phase and air-
water interface
=KL_AWrTmax
Scaling constant to
Seal AWI
-
•
•
1
1
linearly scale the
interfacial area
Langmuir air-water
interface sorption
parameter
Nu_AWI
m3/g
•
•
0
0
Set equal to zero if
Langmuir sorption to
the air-water
interface is not to be
considered
Non-linear
Beta AWI
-
•
•
1
1
Set equal to one
(Freundlich) sorption
Coefficient, (3, for
material type.
since Freundlich
sorption to the air-
water interface is
not to considered.
• = used
— = not used
Boundary Conditions. The models require different types of upper and lower boundary conditions for
flow and transport. Table C-5 presents the types of flow and transport boundary conditions used by
each of the models.
Table C-5. Boundary Conditions - Flow and Transport
Model
Upper Boundary Condition
Lower Boundary Condition
Flow
HYDRUS
Variable Pressure Head/Flux
Free Drainage/Zero Pressure Gradient
EPACMTP
Constant Flux
Constant Pressure Head
ANALYTICAL
Constant Pressure Head/Flux
Free Drainage/Zero Pressure Gradient
Transport
HYDRUS
Constant Mass Flux
Zero Concentration Gradient
DRAFT
C-6
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Model
Upper Boundary Condition
Lower Boundary Condition
EPACMTP
Constant Mass Flux
Zero Concentration Gradient
ANALYTICAL
Constant Mass Flux
Zero Concentration Gradient
Model Simulation Parameters-Time Steps, Duration, Discretization. The process by which most
physics-based simulators generate predictions of contaminant concentration in space and time is
through the partitioning of both dimensions into small, discrete segments, and then repeatedly solving
one or more equations of state or mass conservation across each spatial compartment in the model
domain for each increment of time for some duration. The selection of an incremental space (i.e.,
distance, area, volume) and time for a simulation is dependent in part on a simulator's numerical
approach to solving the physics-based equations, the modeling objective, and balancing the need for
accuracy and computational effort. In general, for porous media flow and transport, the spatial domain
is divided into computational cells or nodes that are small enough to capture the spatial variability of the
state variable (e.g., saturation, pressure, dissolved concentration) in the region of interest at any point
in time, and in small enough time increments to capture key changes in the state variable, like the peak
elevation or concentration. Finally, the number of time increments to evaluate should be sufficient to
capture the temporal variability of the process in the region of interest. For these model comparisons,
the objective is to evaluate the arrival and dissipation of the contaminant at the water table. Table C-6
presents the spatial and temporal discretization parameters used for the simulators and scenarios
conducted in this comparison.
For HYDRUS and the ANALYTICAL model, space and time discretization is prescriptive - the modeler
must select these parameters. In EPACMTP, time and space discretization are internally determined and
optimized to accurately capture water table concentrations for thousands of Monte Carlo simulations.
Spatial discretization is finer near the water table to capture the region of the unsaturated zone where
saturation changes most rapidly. The number of time steps are fixed but sufficient to capture the arrival
and dissipation of the contaminant front at the water table. As the model domains examined here are
small, computational burden is not an issue. Therefore, rectilinear grid cells for both HYDRUS and the
ANALYTICAL model were specified as 1 cubic centimeter and concentration predictions were generated
daily at the water table. Simulation durations in all models were dictated by the combination of slow
advection in the dry environment and high retardation based on Kd and the objective of capturing the
entire concentration breakthrough at the water table.
Table C-6. Spatial and Temporal Simulation Parameters
Model
EPACMTP
HYDRUS
ANALYTICAL*
Spatial
Discretization of
Unsaturated Zone
Computational points are automatically
established at 0.0,0.6,0.75,0.85,0.95,
and 1.0 x depth to water table in meters
1 cm grid cells were specified to
discretize the depth to water table
for both 1 m and 10 m scenarios
1 cm grid cells were specified to
discretize the depth to water table
for both 1 m and 10 m scenarios1
Temporal
Discretization
3000 equal timesteps are automatically
determined between an estimated arrival
and dissipation time of the concentration
front at the water table in years
Daily timesteps were specified
Daily timesteps were specified
Simulation
Duration
10,000 and 20,000 years were specified
for wet and dry scenarios, respectively
10,000 and 20,000 years were
specified for wet and dry
scenarios, respectively
10,000 and 20,000 years were
specified for wet and dry
scenarios, respectively
* Spatial and temporal discretization of the ANALYTICAL model are not used for computing numerical solution but for data
presentation purposes only.
LAU Outputs Used as Inputs to Vadose Zone Models
Crop scenario simulations conducted with the LAU module for biosolids containing PFOA and PFOS
applied in wet and dry environs were used to create time series of mass and water fluxes to represent
the leaching of these contaminants from land applied biosolids. Figure C-3 shows resulting mass fluxes
DRAFT
C-7
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
(left-hand vertical axes) and water fluxes (right-hand vertical axes) for PFOS (top) and PFOA (bottom) in
wet (orange data points) and dry (yellow data points) environments. The cyclic nature of these fluxes
reflects the cyclic meteorological datasets. Figure C-4 shows the resulting leachate concentration over
time for PFOS and PFOA in the two meteorologic settings. Leachate concentration was calculated by
dividing the mass flux by the water flux for each time point. These plots show that constant annual
concentrations in leachate are generated from surface soils during the 40-year period of biosolids
application. In the case of PFOA, concentrations drop off after the 40-year period reflecting no
additional mass and the dissolution of residual PFOA sorbed to soils. For PFOS, the leachate
concentration does not change much over time. This is attributed to the high Koc value limiting the
amount of dissolvable mass to infiltrating water and that the reservoir of sorbed mass is enough to
maintain the limited available mass for a longer time.
20 AO 60 80 100 320 140
Flwx-OiV —*••• Mass Fl
— Water Flu* Dry » Water Fit
Figure C-3, Leachate flux for PFOS (left) and PFOA (right) for crop scenario.
te Content rationDry
^achate Concentration Wet
Figure C-4. Leachate concentration for PFOS (left) and PFOA (right) for crop scenario.
For the purposes of model comparison, all three models were subject to constant water flux (infiltration
rate) at the top of the soil column equal to the average water flux over the 150-year simulation for wet
and dry scenarios. Likewise, constant leachate concentrations for each constituent-meteorology
combination during the 40-year application period were used to define the transport boundary
condition at the top of each model. Modeled values for infiltration and leachate concentrations are
presented in Table C-7.
DRAFT
C-8
-------
PFOA/PFOS Risk Assessment Appendix C. Groundwater Modeling
Table C-7. Boundary Conditions - Infiltration Rates and Leachate Concentrations
Boundary Condition
PFOA
PFOS
Infiltration [m/yr]
Wet Meteorology
9.33E-2
Dry Meteorology
1.49E-3
Leachate Concentration fmq/Ll
Wet Meteorology
3.93E-4
1.96E-3
Dry Meteorology
1.53E-3
7.85E-3
C. 1.4 Model Selection Evaluation Results
This section presents modeling results from EPACMTP, ANALYTICAL and HYDRUS for the various
unsaturated zone model scenarios described above. Results for groundwater flow within the vadose
zone are presented first to evaluate whether both models can simulate similar water content profiles
within the vadose zone for the same set of initial and boundary conditions. If both models simulate
similar water content profiles, differences in PFOA/PFOS concentration results from the transport
simulations, both with and without AWI effects can be inferred to be due to differences in how each
model handles PFOA/PFOS transport.
C. 1.4.1 Flow
Groundwater flow results from EPACMTP, ANALYTICAL and HYDRUS for the various model scenarios
were compared using the simulation of volumetric water content profiles at steady state within the
vadose zone. The volumetric water content describes the volume of water per unit volume of soil,
generally expressed as a dimensionless fraction or percentage. Comparing these profiles would
illuminate any differences between the mathematical formulations used in simulating unsaturated zone
flow. As described earlier, the governing flow equation in EPACMTP is given by Darcy's law, a steady-
state infiltration is used in ANALYTICAL model, while HYDRUS uses a modified form of the Richards
equation. Note these profiles are not constituent specific.
Figure C-5 presents steady-state water content profiles from EPACMTP/HYDRUS for a 1-m (top) and 10-
m (bottom) soil column under wet (left) and dry (right) conditions for loam and loamy sand, assuming
the same boundary conditions. Note that the ANALYTICAL model is not used in this initial comparison
since the model assumptions lend to a single computed value of water content instead of a depth-
dependent profile.
For the 1-m soil column (Figure C-5, top), there is little difference between the HYDRUS and EPACMTP
models for any given soil texture or infiltration condition, and for both models, the simulated water
content profiles are very similar wet and dry conditions given the same soil texture: the maximum
difference is approximately 2%. These similar water content profiles for varying infiltration scenarios
suggests that the default parameters in the Van Genuchten (1980) model used by both models to
simulate the soil water content as a function of infiltration (or pressure head) are compatible and can be
used to simulate steady state conditions for a given infiltration input.
Conversely, the differences between predicted water content profiles between the two soil textures (for
the same infiltration and model) are larger, 8-22% between loam and loamy sand under wet scenario
for HYDRUS (Figure C-5). This can be attributed to differences in the soil water characteristic curves for
the two soil textures. Similar water content profiles simulated by both models for each soil texture
suggests that the parameters used in the soil water characteristic curves to solve the differing governing
flow equations used in both models do not have a significant impact on the predicted water contents.
DRAFT
C-9
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
For the 10-m soil column (Figure C-5, bottom), we see similar results for infiltration scenarios:
differences between dry and wet scenarios for a specified soil texture are less than 2% for all depths
except very close to the water table, between 9 and 10 meters, where the difference between the
predicted water contents from the two models is up to 15%. At a depth of 9 meters, HYDRUS first
predicts lower water contents by up to 6% as compared to EPACMTP and then predicts higher water
contents than EPACMTP closer to the water table at 10 meters, with the maximum difference of 15% at
9.5 meters for loamy sand, dry infiltration scenario. Since these differences between the two models are
only observed close to the water table, it may suggest discrepancies in interpretation of the water table
boundary condition by the two models. This may also be an indication that for a deeper vadose zone
column, the differing mathematical formulations governing flow used in the two models (Darcy's law vs.
Richard's equation), may show an observed influence on water content profiles, particularly close to the
water table.
Wet Conditions
1 m Soil Column
E, 0.4
1-0.5
Dry Conditions
1 m Soil Column
l\
\
l\
l\
\\
\\
11
l\
\\
\\
11
»\
\\
\\
\\
VI
\\
\
\
>
\
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Water Content
CMTP:Wet-Loamy Sand CMTP:Wet-Loam
HYDRUS:Wet-Loamy Sand HYDRUS:Wet-loam
CMTP:Dry-Loamy Sand CMTP:Dry-Loam
— HYDRUS:Dry-Loamy Sand — — - HYDRUS:Dry-Loam
E, 4
5
Wet Conditions
10 m Soil Column
i
i
»
i
t
i
/
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Water Content
Dry Conditions
10 m Soil Column
i
/
V
V-
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Water Content
CMTP:Wet-Loamy Sand CMTP:Wet-Loam
HYDRUS:Wet-Loamy Sand HYDRUS:Wet-Loam
CMTP: Dry-Loamy Sand CMTP:Dry-Loam
— HYDRUS:Dry-Loamy Sand HYDRUS:Dry-Loam
Figure C-5. Water content profiles for 1-m (top) and 10-m (bottom) soil column for loam and loamy
sand under wet (left) and dry (right) scenarios simulated using EPACMTP (solid lines)
and HYDRUS (dashed lines).
Even though EPACMTP computes a variably saturated soil profile that compares well with HYDRUS
under the same boundary conditions (Figure 4), EPACMTP uses a depth-averaged water content for the
analytical transport solution. This is a very useful technique when running several thousand model runs
DRAFT
C-10
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
under a probabilistic framework. The ANALYTICAL model also uses a singular water content value for the
analytical transport solution. However, a direct comparison of the water content values between the
EPACMTP and ANALYTICAL models would not be correct since both models assume different boundary
conditions at the water table (i.e., constant pressure head vs. free drainage). However, the boundary
conditions in HYDRUS can be changed to free drainage for comparison with the ANALYTICAL model even
though HYDRUS uses the variable water content profile shown in Figure C-5 for its transport solutions.
Figure C-6 presents the steady-state water content profiles from the ANALYTICAL model (blue bar) and
HYDRUS (orange bar) for a 1-m and 10-m soil column under wet and dry conditions for loam and loamy
sand. From Figure C-6, it can be observed that the simulated water contents from the ANALYTICAL and
HYDRUS models are very similar for every scenario tested under the same boundary conditions (blue
and orange bars).
25% I
Dry, L, 1 m Dry, LS, 1 m Dry, L, 10 m Dry, LS, 10 m Wet, L, 1 m Wet, LS, 1 m Wet, L, 10 m Wet, LS, 10 m
¦ ANALYTICAL ¦ HYDRUS
Figure C-6. Water content profiles for 1-m and 10-m soil column for loam and loamy sand under
wet and dry scenarios simulated using HYDRUS and ANALYTICAL models.
Overall, the simulation results shown in Figures C-5 and C-6 confirm that there is little difference
between three models in simulating variable saturated flow regardless of soil textures, meteorological
environments, vadose zone thickness. However, for a deeper vadose zone soil column, the influence of
water table boundary conditions and governing flow equations on simulated water content profiles
should be carefully considered.
DRAFT
C-ll
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
— EPACMTP
HYDRUS
ANALYTICAL
EPACMTP
HYDRUS
ANALYTICAL
C.1.4.2 Transport
The figures in this section present breakthrough curves from EPACMTP, ANALYTICAL and HYDRUS
models for PFOA (top) and PFOS (bottom) for the different soil column depths and meteorological
conditions. For all figures, both models were run using linear adsorption (no AWI effects; solid lines) and
both the ANALYTICAL model and HYDRUS were run with AWI effects (dashed lines) using three values of
the equilibrium distribution constant between the liquid phase and air-water interface (Kaw). Note that
both the ANALYTICAL model and the most recent version of HYDRUS (version 5.01) used in this testing
can only simulate instantaneous, equilibrium sorption at the AWI. Recent column experiments have
shown that kinetics associated with AWI adsorption is minimal under steady-state flow conditions
(Brusseau, 2020; Brusseau et al., 2021).
Shallow Water Table, Wet Environment
For a 1-m soil column in a wet environment (Figure C-7), breakthrough curves simulated by the
EPACMTP, ANALYTICAL and HYDRUS models show excellent agreement for both soil types and
constituents in terms of their magnitude and peak arrival time when considering only linear, solid-phase
sorption.
300 400 500 600 700 800 0 200 400 600
Time (years) Time (years)
15
<5
u
12
c
o
PFOS
£
10
c
c
c
u
¦o
7
N
5
A3
E
o
Z
2
Figure C-7. Breakthrough curves at a 1-m water table depth in a wet environment using linear
sorption.
Note that for the same soil type and meteorological condition, the peak magnitude observed at the
water table for PFOA is much greater than that for PFOS (peak concentrations >80% of input
concentrations for PFOA vs 14-17% for PFOS). This lower peak magnitude observed at the water table
for PFOS is due to stronger solid phase adsorption of PFOS to soil organics and is reflected by the
difference in their representative Kd value chosen for the modeling effort (see Table C-4). When
comparing the simulated peak magnitudes for the same constituent but across soil types, the peaks for
Loam are slightly greater than that for Loamy Sand (approximately less than 3%). These differences are
DRAFT
C-12
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
likely due to the higher calculated dispersion coefficients for Loamy Sand as compared to Loam. The
dispersion coefficients in all three models are calculated as:
D = dispersion coefficient (L2T1 such as cm2/s)
V = pore velocity or Darcy velocity (LT1 such as cm/s)
Q = infiltration rate (LT1 such as cm/s)
N = model-calculated water content (L2/L3 such as cm2/cm3)
aL = dispersivity (L such as cm).
While the infiltration rate (q) and the dispersivity (aL) are the same for both soil types, the model
computed pore velocities (v) and water contents are different owing to the differences in their
saturated hydraulic conductivities input to the model. The pore velocities computed for Loamy Sand are
higher than that for Loam while the water content for Loam is higher than Loamy Sand. For example, in
the wet scenario, 1 m soil column, the ANALYTICAL model calculated pore velocity for Loam at 1.26 x 10"
6 cm/s corresponding to a water content of ~24.5%. For Loamy Sand, the calculated pore velocity was
2.84 x 10 s cm/s corresponding to a water content of ~10.4%. Since the dispersion coefficient is directly
proportional to the pore velocity (or inversely proportional to the water content), the higher pore
velocity for Loamy Sand may have contributed to increased dispersion and produced the slightly lower
peak observed.
The simulated PFOA/PFOS arrival time at the water table based on the breakthrough curves (Figure C-7)
are only illustrative for model comparison purposes and may appear greater than those observed in field
studies. For instance, modeled peak values of PFOS arrive at the water table in a Loamy Sand column
under a wet scenario in approximately 216 years. This might appear contradictory to field observations
of PFOS observed at the water in deeper soil columns. There may be several reasons to note regarding
differences between model results and field observations:
• In the models, the modeled peak arrival times are a direct function of the representative Kd
value chosen for the model comparison simulations (see Table C-4). That is, lower values of Kd
chosen, the earlier the modeled peak arrival time. This can be illustrated by selecting an
extremely low value of Kd=0 (i.e., considering only advection and dispersion). In this scenario,
PFOS is simulated to arrive at the water table in 4.4 years. Kd values for PFAS span a large range
generally due to the varying soil types, field or laboratory conditions under which they were
measured. A single Kd, organic partitioning coefficient (Koc) and fraction organic carbon (foc),
values were chosen for each of PFOA/PFOS (see Tables C-2 and C-4). However, the range of Koc
reported in literature is large and can often span several orders of magnitude. There is
significant ongoing research on refining the Koc values and understanding the partitioning
behavior of PFAS in the environment, which can also depend on site-specific factors that are not
considered here. When modeling site-specific observations, Kd should be carefully considered
prior to making model comparisons to field observations. For the experiments reported here,
the Kd values for each constituent was kept constant throughout the simulations for an
appropriate comparison of results from different models.
• Models tested here do not account for preferential flow paths for PFAS migration to the water
table due to soil heterogeneity that maybe present under field conditions. For example, Zeng
and Guo (2021) have shown that preferential flow pathways generated by soil
where
DRAFT
C-13
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
heterogeneities can destroy air-water interfaces that can lead to early arrival and
accelerated leaching of (especially long-chain) PFAS.
• Source leachate concentrations used in this modeling exercise are less than 0.01 mg/L (see Table
C7), which is likely several orders of magnitude lower than source concentrations often reported
in several studies (Anderson et al., 2016).
Including the effects of AW I (Figure C-8), we see that the ANALYTICAL and HYDRUS models show
excellent agreement for both soil types and constituents in terms of their magnitude and peak arrival
time. Generally, it is observed for both models that AWI decreases the magnitude of the peak
concentrations and increases the arrival time of the peak at the water table for both PFOA and PFOS
compared to scenarios considering only linear, solid-phase sorption. The higher the interfaciai
adsorption coefficient at the AWI, the more pronounced the effects on peak concentrations and arrival
times. This is consistent with increased retardation of PFOA/PFOS anticipated with retention at the AWI.
In addition to the difference in peak magnitude noted between PFOA/PFOS, the tail of the breakthrough
curve is much longer for PFOS in comparison to PFOA. The longer breakthrough curve tail observed for
PFOS is likely owing to its stronger retention to the solid phase. Another interesting observation is that
when including the effects of AWI, the peak magnitude observed for Loamy Sand is higher than that for
Loam within each constituent (PFOA or PFOS) and Kaw value. This is the opposite of what was observed
when only considering solid phase sorption, when the peak magnitude for Loam was higher than Loamy
Sand The higher peak magnitude observed for Loamy Sand as compared to Loam is likely because there
is less AWI adsorption for Loamy Sand than Loam. The reason for this is that the model computed total
air water interfaciai area is lower for Loamy Sand (~55 cm2/cm3) than Loam (~112 cm2/cm3), which leads
to reduced AWI adsorption and lower retention of PFAS within the vadose zone of a Loamy Sand column
than Loam.
LOAM LOAMY SAND
Time (years) Time (years)
Figure C-8. Breakthrough curves at a 1-m water table depth in a wet environment using linear
sorption and AWI effects for three values of Kaw (Silva et al., 2020).
DRAFT
C-14
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
However, some recent studies based on field data have shown a negative association between clay
content and PFAS migration to the water table (see e.g., Andersen et al., 2019). Using data from US Air
Force sites, these authors have shown that soils with higher clay content show statistically significant
lower soil retention (or higher groundwater concentrations) than more permeable soils. Andersen et al.
(2019) suggested three possibilities to explain their observation:
1. Relatively lower clay content soils are better drained and less prone to saturation during
precipitation events. Lower water saturation would lead to higher magnitude of air-water
interfacial area, and therefore retardation (Peng and Brusseau, 2005).
2. Electrostatic interactions between the negatively charged clay minerals and anionic PFAS may
enhance transport to the water table due to anionic repulsion (Wang et al., 2015).
3. Soils with higher clay content retain relatively larger volumetric water content following
precipitation events resulting in longer reaction time between aqueous and adsorbed PFAS and
thus, kinetic-limited PFAS sorption (Wei et al., 2017), which would promote greater partitioning
in higher clay content soils.
While the findings of Andersen et al. (2019) may seem contradictory to the modeled results presented
here, one important thing to note is that all the model soil columns are forced with the same infiltration
rate to make even comparisons for the purposes of this report. At field sites, the infiltration rate is likely
to vary by soil type owing to the differences in their water retention capacities and surface evaporation
rates.
Shallow Water Table, Dry Environment
For a 1-m soil column in a dry environment (Figure C-9), the breakthrough curves for EPACMTP,
ANALYTICAL and HYDRUS under linear, solid phase sorption suggest PFOA/PFOS mass is strongly
adsorbed to the soil and very little mass reaches the water table (<0.27% for PFOS and <2.5% for PFOA),
even though the total mass of PFAS applied are the same for both dry and wet environments. These
results suggest that in the dry environment there is much less advective/dispersive transport of
PFOA/PFOS to the water table. In the model, this is evidenced by the calculated pore velocities and
dispersion coefficients that are 2 orders of magnitude lower for the dry scenario as compared to the wet
scenario. There is also slightly stronger solid-phase adsorption calculated in the dry scenario as
compared to the wet scenario because of a higher retardation factor under the dry scenario. This is due
to solid phase retardation is inversely proportional to the soil water content, which is approximately
12.7% for a wet, Loamy Sand, lm column and 4.6% for a dry Loamy Sand, 1 m column. These
observations are consistent with studies that used field data from many sites and showed that PFAS soil
to groundwater mass transfer is strongly influenced by the degree of flushing at these sites (see e.g.,
Anderson et al. 2019). In other words, under low precipitation and deep groundwater, increased
retention of PFAS is anticipated within the soil column (or decreased PFAS discharge to groundwater
table) due to flushing limitations. Even though strong vadose zone retention is observed under the dry
scenario, all three models are in excellent agreement on the simulated breakthrough curve peak
magnitude and timing (Figure C-9). The maximum difference in peak magnitudes simulated by the three
models for PFOA is 0.27% and for PFOS is 0.02%.
Including the effects of AWI under the dry scenario (Figure C-10), we see that the ANALYTICAL and
HYDRUS models show excellent agreement for both soil types and constituents in terms of their
magnitude and peak arrival time. As with the wet scenario, it is observed for both models that AWI
decreases the magnitude of the peak concentrations and increases the arrival time of the peak at the
water table for both PFOA and PFOS than when considering only linear, solid-phase sorption. The higher
the interfacial adsorption coefficient at the AWI, the more pronounced the effects on peak
concentrations and arrival times.
DRAFT
C-15
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
LOAMY SAND
E
g £ 0.15 H
u_ ^
Q- c
U 0.10 -
TJ
0)
EPACMTP
HYDRUS
ANALYTICAL
EPACMTP
HYDRUS
ANALYTICAL
2500 5000 7500 10000 12500 15000 17500 20000
Time (years)
2500 5000 7500 10000 1250015000 17500 20000
Time (years)
Figure C-9. Breakthrough curves at a 1-m water table depth in a dry using linear sorption.
LOAMY SAND
E 1.00
2*
^ i 0.75
E
| 0.25
— 0175
g
£ 0.150
tj
c 0.125
Kaw = 3.69x10 ~3cm3lcm2
' HYDRUS
ANALYTICAL
.. Kaw = 3.69x10 _2cm3/cm2)
HYDRUS
ANALYTICAL
Kaiv = 3.69xl0-1cm3/cm2
HYDRUS
-A- ANALYTICAL
1.75
1.50
1.25
1.00
0.75
0.50
0.25
0.00
Kaw = 3.69x10 ~3cm3/cm2
HYDRUS
ANALYTICAL
Kaw = 3.69x10 ~2cm3/cm2)
HYDRUS
ANALYTICAL
i|
Kaw = 3.69x10 -10B3/cm2
HYDRUS
-jfc-
ANALYTICAL
2500 5000 7500 10000 12500 15000 17500 20000
2500 5000 7500 10000 12500 15000 17500 20000
o g
c
Q 0.075
"8
£ 0.050
I
S 0.025
Kaw = 4.79x10 ~2cm3/cm2
HYDRUS
ANALYTICAL
Kaw = 4.79x10 ~lcm3/cm2
HYDRUS
ANALYTICAL
Kaw = 4.79cm3/cm2
HYDRUS
ANALYTICAL
Kaw = 4.79x10 ~2cm3/cm2
' HYDRUS
ANALYTICAL
Kaw = 4.79x10 _1cm3/cm2
K HYDRUS
ANALYTICAL
. Kaiv = 4.7 9cm 3/cm2
HYDRUS
-M- ANALYTICAL
2500 5000 7500 10000 12500 15000 17500 20000
Time (years)
2500 5000 7500 10000 12500 15000 17500 20000
Time (years)
Figure C-10. Breakthrough curves at a 1-m water table depth in a dry environment using linear
sorption and AWI effects for three values of Kaw (Silva et al., 2020).
DRAFT
C-16
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Deeper Water Table, Wet Environment
For a 10 m soil column in a wet environment, the breakthrough curves show excellent agreement
between the models for peak magnitude and arrival time for both constituents and soil textures under
linear sorption only (Figure C-ll) and including the effects of AWI (Figure C-12). The PFOA/PFOS
concentrations at the water table for a 10 m soil column are much lower as compared to the 1 m soil
column under the wet scenario. For example, under the wet scenario for aim Loam soil column,
approximately 90% of the input PFOA concentrations were observed at the water table (Figure C-7)
while approximately 12% of the input PFOA concentrations were observed at the water table for the
same conditions in the 10 m column. This is because the same total input mass is applied to the top of
both soil columns, but the deeper soil column has larger soil volume and therefore greater sites for
solid-phase sorption of the same input mass.
LOAM LOAMY SAND
Time (years) Time (years)
Figure C-ll. Breakthrough curves at a 10-m water table depth in a wet environment using linear
sorption.
DRAFT
C-17
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
LOAMY SAND
12 -
Kaw = 3.69xl0~3cm3/cm2
12
HYDRUS
ANALYTICAL
10
10 -
Kaw = 3.69xl0"2cm3/cm2)
HYDRUS
ANALYTICAL
Kaw = 3.69x10 _1cm3/cm2
HYDRUS
6
6-
-A- ANALYTICAL
4-
4
2 -
2
0-M
0
(b) A
Kaw = 3.69x10 _3cm3/cm2
HYDRUS
ANALYTICAL
Kaw = 3.69x10 -2trm3/cm2)
HYDRUS
1 A\
ANALYTICAL
Kaitf = 3.69xl0_1cm3/cm2
HYDRUS
-A- ANALYTICAL
6000
8000
250 500 750 1000 1250 1500 1750 2000
Kaw = 4.79x10-2cm3/cm2
~t~~ HYDRUS
*- ANALYTICAL
Kaw = 4.79x10 _1cm3/cm2
HYDRUS
~ - ANALYTICAL
. Kaw - 4.7 9 cm2/cm2
HYDRUS
A- ANALYTICAL
0.8 -
0.6 -
0.4 -
0.2 -
(d)
Kaw = 4.79xl0~2cm3/cm2
i \
HYDRUS
—-
ANALYTICAL
Kaw = 4.79xl0_1cm3/cm2
HYDRUS
ANALYTICAL
Kaw - 4.79cm3/cm2
f \ \V
HYDRUS
/ \\
-A-
ANALYTICAL
{ \.
* -T~r
2500 5000
7500 10000 12500 15000 17500 20000
Time (years)
2500 5000
7500 10000 12500 15000 17500 20000
Time (years)
Figure C-12.
Breakthrough curves at a 10-m water table depth in a wet environment using linear
sorption and AWI effects for three values of Kaw (Silva et al., 2020).
Deeper Water Table, Dry Environment
For a 10 m soil column in a dry environment, the breakthrough curves for PFOA assuming solid phase
adsorption only (Figure C-13, top row) correspond well between all models tested. Under the same
scenario, all three models suggest that PFOS is not transported in the timeframe modeled to the water
table due to strong solid phase adsorption. However, the simulated breakthrough curves are shown for
completeness, but the reader will observe that the simulated concentrations are insignificantly low
(<8x10~7% of input concentrations; Figure C-13, bottom row). Finally, including the effects of AWI to
PFOA/PFOS (Figure C-14, first column) shows good agreement between HYDRUS and ANALYTICAL
model-simulated breakthrough curves. However, while the ANALYTICAL model was able to simulate the
breakthrough curves for Loamy Sand under the same scenario, the numerical solution of HYDRUS
became unstable beyond approximately 2,500 years for PFOA and 3,500 years for PFOS and the
solutions did not converge. As noted earlier, the simulated PFOA/PFOS arrival time at the water table
based on the breakthrough curves (Figure C-7) are not absolute and are only illustrative for model
comparison purposes.
DRAFT
C-18
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
LOAMY SAND
y
Cj 0.20
0 2500 5000 7500 1000012500 15000 17500 20000
le-7
0 2500 5000 7500 10000 1250015000 17500 20000
le-7
£
5 7
»-6
c
.2 5
E
Q- c
o 3
U J
T3
a; ?
N *¦
"ro
§ i
O
z
0
2500 5000 7500 1000012500 15000 17500 20000
Time (years)
2500 5000 7500 10000 1250015000 17500 20000
Time (years)
Figure C-13. Breakthrough curves at a 10-m water table depth in a dry environment using linear
sorption.
LOAMY SAND
0.175
. _) Kaw = 3.69xl0~3O7J3/C/T)2
HYDRUS
0.150
ANALYTICAL
A'
K Kaw = 3.69x10 ~2cm3/cm2)
0.125
HYDRUS
ANALYTICAL
0.100 -
Kaw = 3.69xl0~1cm3/cm2
HYDRUS
0.075
ANALYTICAL
//
0.050 •
//
//
//
0.025
a
y
0.000 -
0.175 -
lb) Kaw = 3.69xl0~3 cm3/cm2
-
HYDRUS
s
0.150 -
ANALYTICAL
/
/
/
)c Kaw = 3.69xl0-2cm3/cm2)
/
/
0.125 -
HYDRUS
/
/
ANALYTICAL
/
t
0.100 -
Kaw = 3.69xl0_1cm3/cm2
/
HYDRUS
/
0.075 -
-A- ANALYTICAL
/
/
/
0.050 -
/
/
/
x'
0.025 -
/
/
-r''
0.000 -
> ~ i
5000 7500 10000 12500 15000 17500 20000
0 2500 5000 7500 10000 12500 15000 17500 20000
le-11
Kaw = A.7 9xl0~2 cm3/cm2
HYDRUS
ANALYTICAL
Kaw = 4.79xl0-1cm3/cm2
HYDRUS
ANALYTICAL
. Kaw = 4.79cm3fcm2
HYDRUS
-A- ANALYTICAL
2500 5000 7500 10000 12500 15000 17500 20000
Time (years)
0 2500 5000 7500 10000 12500 15000 17500 20000
Time (years)
Figure C-14. Breakthrough curves at a 10-m water table depth in a dry environment using linear
sorption and AWI effects for three values of Kaw (Silva et al., 2020).
DRAFT
C-19
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
C.1.5 Model Selection Conclusions
Three simulation models—EPACMTP, ANALYTICAL, and HYDRUS—were examined to determine which is
best suited to support risk assessment objectives. The models were evaluated by comparing flow and
PFOA/PFOS transport results from eight scenarios that reflect a broad range of key hydrogeologic
conditions on a national scale including depth to water table, soil texture and meteorological conditions.
Additionally, transport simulations included comparison of model outputs assuming solid phase
adsorption only as well as solid phase adsorption with AWI effects.
Overall, the simulation results confirm that there is little difference between three models in simulating
variable saturated flow regardless of soil textures, meteorological environments, and vadose zone
thickness. Comparison of breakthrough curves at the water table when considering only linear, solid-
phase sorption from transport simulations show excellent agreement between all three models for both
soil columns (lm and 10m), soil types (loam and loamy sand), infiltration (dry and wet) and constituents
(PFOA and PFOS) in terms of their magnitude and peak arrival time. The maximum difference in peak
magnitudes for the various scenarios simulated by the three models for PFOA is less than 0.3% and for
PFOS is 0.03%. However, the magnitude and peak arrival times were observed to be different between
soil types, infiltration scenarios, constituent simulated and soil column depths. Our modeling results
show that when biosolids are land applied at the surface, the greatest mass of PFAS arriving at the water
table (~90% of input concentrations) in the shortest amount of time (~54 years) is observed for PFOA
moving through a short, 1 m vadose zone under wet conditions. In contrast, all three models suggest
that PFOS is not transported to the water table in a 10 m soil column in a dry environment due to strong
solid phase adsorption. As noted earlier, the simulated PFOA/PFOS peak magnitudes and arrival time at
the water table based on the breakthrough curves are not absolute or site-specific but only illustrative
for model comparison purposes. Nevertheless, these bounding simulations highlight the importance of
selecting appropriate values for location-specific and contaminant-specific critical factors such as Kd, Kaw,
soil texture, depth to water table and net infiltration when conducting risk assessments, a conclusion
that was also suggested by Pepper et al. (2023). Additionally, all solid phase adsorption simulations were
performed assuming instantaneous equilibrium. However, the ANALYTICAL model is capable of
simulating kinetics associated with solid phase adsorption, which maybe an important process in real
soils with organic carbon or in clayey soils (see e.g., Guelfo et al., 2020 and Schaefer et al., 2021).
Including the effects of AWI, we see that the ANALYTICAL and HYDRUS models show excellent
agreement for both soil types and constituents in terms of their magnitude and peak arrival time.
Generally, it is observed for both models that AWI decreases the magnitude of the peak concentrations
and increases the arrival time of the peak at the water table for both PFOA and PFOS compared to
models considering only linear, solid-phase sorption. The higher the interfacial adsorption coefficient at
the AWI, the more pronounced the effects on peak concentrations and arrival times. This is consistent
with increased retardation of PFOA/PFOS anticipated with retention at the AWI. While both models
were able to simulate solid-phase and AWI retention processes and showed good agreement between
model-simulated breakthrough curves, the numerical solution of HYDRUS became unstable for a 10 m
soil column in a dry environment, while the ANALYTICAL model did not have any issues.
The contribution of PFOA/PFOS mass retention at the AWI was further evaluated at Boulder (Dry
climate, deep water table) and Charleston (wet climate, shallow water table) using the ANALYTICAL
model, Koc bounds, and location-specific environmental parameters discussed in Section C.3. Model
simulation results indicate that AWI retention is not a significant mechanism for PFOA/PFOS retention
for the specific chemical- and environment-specific conditions modeled. Less than 1% of the total
PFOA/PFOS mass leached from biosolids applied to the LAU is retained at the AWI with the remaining
applied mass either sorbed to solid-phase or transported through aqueous phase under high-Koc
DRAFT
C-20
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
conditions. Under low-Koc conditions, the AWI contributes 7-9% of total leached PFOA/PFOS mass,
except at Charleston for PFOS where 18% of total leached PFOS mass is retained at the AWI. These
results suggest that PFOA/PFOS mass retained at the AWI is not a significant contributor to mass
retention for the chemical- specific and environment-specific conditions modeled.
The overall objective of the preceding analysis was to evaluate transport processes available in
unsaturated zone flow and transport simulators to predict PFOA/PFOS migration through the vadose
zone to the water table for a range of environmental settings and constituent-specific fate and transport
parameters. Overall, we observe that the vadose zone module in EPACMTP would produce higher (i.e.,
risk-conservative) PFAS concentrations at the water table because the model does not have the ability
to address PFAS-specific retention behavior at the AWI. While both HYDRUS and the ANALYTICAL
models are capable of simulating PFAS-specific retention behavior, and generally in good agreement
when simulating PFOA/PFOS leaching from surface soils resulting from the application of biosolids
through the vadose zone to groundwater, these models require site-specific inputs to model AWI that
are not available in the current risk assessment framework. Further, though the time to breakthrough
on all models are longer than existing field studies indicate are possible, incorporating AWI into the
HYDRUS and ANALYTICAL models only increases the time lag observed in the models compared to the
monitored data. Evaluation of PFOA/PFOS mass retained at the AWI was not determined to be a
significant contributor based on ANALYTICAL model simulations and the chemical-specific and
environment-specific conditions discussed in Section C.3. Therefore, EPACMTP is used to conduct
unsaturated and saturated zone flow and transport simulations to evaluate the fate and transport of
PFOA and PFOS in land applied biosolids in this risk modeling framework.
C.2 Overview of EPACMTP
The transport of leachate from the land-applied biosolids or sewage sludge managed in surface disposal
units through the unsaturated and saturated zones is evaluated quantitatively using EPACMTP (US EPA,
2003a,b,d; 1997). EPACMTP simulates the flow and transport of contaminants in the unsaturated zone
and aquifer beneath a waste management unit to yield the concentration that arrives at a specified
receptor location. The LAU and SDU source models determine the leachate concentration used as an
input to EPACMTP. As described in the Addendum to the EPACMTP Technical Background Document (US
EPA, 2003a), new functionality was added to the EPACMTP model to create a dynamic, mass-conserving
linkage between the source models and EPACMTP.
The groundwater model accounts for advection, hydrodynamic dispersion, equilibrium linear or
nonlinear sorption, and transformation processes via chemical hydrolysis. In this analysis, data were
compiled from the scientific literature to develop organic carbon partition coefficients to simulate
equilibrium linear partitioning for PFOA and PFOS. Organic carbon partition coefficient inputs are
discussed in Section C.3 and Appendix B.
EPACMTP is a composite model that consists of two coupled modules: (1) a 1-dimensional (1-D) module
that simulates vertical infiltration and dissolved contaminant transport through the unsaturated zone,
and (2) a saturated zone flow and transport module that includes three groundwater transport solution
options: (i) fully 3-D transport, (ii) quasi-3-D transport (a combination of cross-sectional and areal
solutions), and (iii) pseudo-3-D transport (hybrid analytical and numerical solution). The applicability and
appropriateness of each of the transport solution options depend on the problem considered. The
pseudo-3-D solution is the most computationally efficient of the available options. In addition, the
pseudo-3-D solution can accurately and efficiently generate full breakthrough curves at the receptor
location. For these reasons, the pseudo-3-D solution option was chosen for this analysis.
DRAFT
C-21
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
The EPACMTP algorithms assume that the soil and aquifer are uniform porous media and that flow and
transport are described by Darcy's law and the advection-dispersion equation, respectively. EPACMTP
does not account for preferential pathways, such as fractures and macropores, or facilitated transport,
which may affect the migration of PFAS. For example, Zeng and Guo (2021) have shown that preferential
flow pathways generated by soil heterogeneities can reduce the strength of retention at the air-water
interfaces that can lead to early arrival and accelerated leaching of (especially long-chain) PFAS.
EPACMTP models the advective movement in the unsaturated zone as 1-D, whereas the saturated zone
module accounts for 3-D flow and transport. EPACMTP also considers mixing due to hydrodynamic
dispersion in both the unsaturated and saturated zones. In the unsaturated zone, flow is gravity-driven
and prevails in the vertically downward direction. Therefore, the flow is modeled in the unsaturated
zone as 1-D in the vertical direction. It is also assumed that transverse dispersion (both mechanical
dispersion and molecular diffusion) is negligible in the unsaturated zone. This assumption is reasonable
given that lateral migration due to transverse dispersion is negligible compared with the horizontal
dimensions of the waste management unit. In addition, this assumption is environmentally protective
because it allows the leading front of the pollutant plume to arrive at the water table with greater peak
concentration in the case where the duration of leaching is finite.
In the saturated zone, the movement of pollutants is primarily driven by ambient groundwater flow,
which in turn is controlled by a regional hydraulic gradient and hydraulic conductivity in the aquifer
formation. The model considers the effects of infiltration from the waste source and the regional
recharge into the aquifer. The effect of infiltration from the waste source is an increase in groundwater
flow in the horizontal transverse and vertical directions underneath and in the immediate vicinity of the
waste source, as may result from groundwater mounding. This 3-D flow pattern will enhance the
horizontal and vertical spreading of the plume. Regional recharge outside of the waste source causes a
(vertically) downward movement of the plume as it travels in the (longitudinally) downgradient
groundwater flow direction. In addition to advective movement and groundwater flow, the model
simulates the mixing of contaminants with groundwater due to hydrodynamic dispersion, which acts in
the longitudinal direction (i.e., along the groundwater flow direction) and in the horizontal and vertical
transverse directions.
Leachate pollutants can be subject to complex geochemical interactions in soil and groundwater, which
can strongly affect their rate of transport in the subsurface. EPACMTP treats these interactions as
equilibrium-sorption processes. The equilibrium assumption means that the sorption process occurs
instantaneously, or at least very quickly, relative to the time scale of pollutant transport. However,
studies have observed that PFOA/PFOS solid phase sorption processes are not always well represented
by reversable equilibrium partitioning assumptions due to rate-limited sorption considerations (Guelfo
et al., 2020; Brusseau, 2020). Guo et al. (2022) implemented a linear isotherm simplification for solid
phase adsorption and compared predicted simulations for a wide range of sand-packed miscible-
displacement experiments for PFAS under water-unsaturated conditions as well as a simulation of PFAS
leaching at a model AFFF-impacted fire training area site. These authors found that their model with
linear isotherm simplification reproduced solutions identical to a full-scale numeric model that accounts
for a set of comprehensive PFAS-specific transport processes, including nonlinear solid phase
adsorption. While this is an active area of research, uncertainties in PFAS-specific, non-linear behavior in
assessing the exposures associated with land-applied biosolids on the groundwater pathway may need
consideration. Although sorption, or the attachment of leachate pollutants to solid soil or aquifer
particles, may result from multiple chemical processes, EPACMTP combines these processes into an
effective soil-water partition coefficient (Kd). The retardation factor, R, accounts for the effects of
DRAFT
C-22
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
equilibrium sorption of dissolved pollutants onto the solid phase. R, a function of the pollutant-specific
Kd and the soil or aquifer properties, is calculated as follows:
R=1+PbE±
$
where
R = retardation factor (unitless)
Pb = soil or aquifer bulk density (g/cm3)
Kd = solid-water partition coefficient (cm3/g)
(f> = water content (in unsaturated zone) or porosity (in saturated zone) (unitless).
Chemicals with low Kd values will have low retardation factors, which means that they will move at
nearly the same velocity as the groundwater. Chemicals with high Kd values will have high retardation
factors and may move many times slower than groundwater.
As modeled in EPACMTP, the Kd of an organic pollutant is assumed to be constant within each modeled
soil column and is calculated as the product of the mass fraction of organic carbon in the soil or aquifer
and a pollutant-specific organic carbon partition coefficient (Koc). Multiple literature searches were
conducted to identify field and laboratory studies reporting either measured or estimated values of Kd or
Koc, with and without associations to biosolids or land application of biosolids, for surface and
subsurface soils, aquifer materials, and settled and suspended surface water sediments. The results of
the literature survey, discussed in Section C.3, demonstrated that the spread and median values of log
Koc tend to show less variability across the various categories of field and laboratory studies, biosolids
and non-biosolids related studies, and across media than log Kd. This behavior is reasonable as log Kd
additionally reflects the effects of organic carbon (OC) variability in various matrices, as well as other
parameters (e.g., pH), whereas log Koc does not. Therefore, /Cocwas used as inputs to the model, along
with fraction of organic carbon corresponding to the dominant soil mega texture at each location
(Section C.3), model sensitivity tested at a lower and upper bound Koc values as described in
Appendix D.
EPACMTP simulates steady-state flow in both the unsaturated and saturated zones and can
accommodate either steady-state or transient contaminant transport. Steady-state transport modeling
is a protective modeling approach in which a unit continues to release contaminants indefinitely
(continuous source); eventually, the model will predict that the receptor well concentration reaches a
constant value. However, in this analysis, transient transport simulations were performed. This finite
source approach simulates the amount of time over which the land application unit is active and the
time-dependent movement of chemical pollutants in the subsurface to the receptor well.
C. 2.1 Groundwater Receptor
One of the most important inputs for EPACMTP is receptor location, which for this risk analysis included
a residential drinking water well located 5 meters from the edge of the farm field (i.e., center of the
buffer). EPACMTP can also evaluate the exposure concentration of a hypothetical residential drinking
water well at a specified depth below the water table. For this evaluation, four depths below the water
table were considered (0.5 m, 1.0 m, 1.5 m and 2.0 m) and the maximum of the peak concentrations at
all depths was considered as the exposure concentration. The well depths were limited to the top 2.0 m
below the water table (1) to be consistent with a residential well scenario (these wells are generally
shallow because of the higher cost of drilling a deeper well) and (2) to produce a conservative estimate
of risk (because the infiltration rate is generally lower than the groundwater seepage velocity,
DRAFT
C-23
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
groundwater plumes tend to be relatively shallow). Limitation of well depth is further validated by the
consistent PFOA and PFOS groundwater concentration profiles with depth modeled at the residential
drinking water well located 5 meters from the edge of the farm field (Figures C-15 and C-16). These
profiles show that contamination is roughly constant over the top 6-8 m of the aquifer at all modeled
site locations at the residential drinking water well and alleviates concerns on overpredicting modeled
risks by selecting the maximum of the peak concentrations within the top 2.0 m below the water table.
DRAFT
C-24
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Koc = Low
Koc = High
2 -
4 -
8 -
10
un
r+
fD
P3
O
C
Ct-
rl)
ft-
rd
n
n"
Oi
ua
o
r"t"
0)
n
rr
a*
fP
tri
0.2 0.4 0.6 0.8
Relative Concentration
0.2 0.4 0.6 0.8
Relative Concentration
Figure C-15. Well depth below water table (m) vs. relative PFOA concentrations for point
observations (solid circles) for CROP; low Koc (left panels) and high Koc (right panels) at a
well located 5 meters away from edge of field.
DRAFT
C-25
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Koc = Low
Koc = High
8 -
10
2 -
4 "
6 -
8 -
10
On
2 "
4
6 "
8 -
10 -
in
ri-
II
n
3"
n'
£U
VQ
O
y?.
r"t
0)
n
ZT
QJ
—s
fD
0.2 0.4 0.6
Relative Concentration
0.2 0.4 0.6
Relative Concentration
0.8
Figure C-16. Well depth below water table (m) vs. relative PFOS concentrations for point observations
(solid circles) for CROP, low Koc (left panels) and high Koc (right panels) at a well located 5
meters away from edge of field.
DRAFT
C-26
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
C. 2.2 Groundwater Path way Simulations
The leachate fluxes (g/m2/yr) and infiltration water fluxes (m/d) estimated by the land application and
surface disposal source models were used as input to the groundwater fate and transport model,
EPACMTP, to generate pollutant concentrations at receptor wells. These fluxes serve as the flow and
transport boundary conditions within the footprints of the field and disposal units, The leachate fluxes
(g/m2/yr) and infiltration water fluxes (m/d) were estimated using regional and local variables at three
geographic regions (approximated using data from Boulder, Chicago and Charleston, USA)
corresponding to dry, moderate and wet climate conditions. For the groundwater pathway simulations
in this analysis, the flux of pollutants from the LAU and the SDU were not constant. Instead, the source
models predict a time series of leachate flux, whereby the mass transfer to the groundwater pathway
varies from year to year. The maximum leachate flux and corresponding annual infiltration rate from the
profiles generated by the source models were used as inputs to EPACMTP along with various chemical-,
location-, and environment-specific variables, as discussed in Section C.3 below.
C. 2.3 Key Assumptions
This section presents key assumptions associated with the groundwater modeling approach. More
comprehensive documentation of EPACMTP and associated assumptions are available in the EPACMTP
Technical Background Document (US EPA, 2003a).
¦ The model assumes that the vertical migration is 1-D and that transverse dispersion is negligible
in the unsaturated zone.
¦ The model assumes linear equilibrium sorption for PFOA/PFOS in the unsaturated soil and
aquifer zones and homogeneous aquifer conditions.
¦ The model assumes that receptors use the uppermost (water table) aquifer, rather than a
deeper aquifer, as a source of drinking water. This assumption could overestimate risks in cases
in which the uppermost aquifer is not used.
¦ The model assumes that long-term average conditions are sufficient for exposure calculation
and that shorter frequency fluctuations (e.g., in rainfall/infiltration) are insignificant in
estimating long-term risk.
¦ Biodegradation in groundwater was excluded given the recalcitrant nature of PFOA/PFOS.
¦ Preferential flow in karst aquifers or in fractures was not considered, although such conditions
are known to exist over broad areas. Preferential flow can allow contamination to migrate faster
and at a higher concentration than in a standard porous medium. However, the contamination
typically does not spread over such a broad area. As a result, the modeling may underestimate
or overestimate the concentrations in groundwater, depending on how concentrations are
averaged spatially and temporally.
C.3 Model Inputs
Appendix B presents the input values used in modeling the groundwater pathway using EPACMTP.
Below is a description of some key EPACMTP inputs.
C.3.1 Fluxes from Source Models
The releases of PFOA/PFOS mass and infiltrating water were determined using waste management unit-
specific models (land application unit, or LAU, and surface disposal unit, or SDU) developed for 3MRA.
These models generate time-series of PFOA/PFOS mass fluxes and infiltrating water fluxes to the
subsurface as well as releases to other exposure pathways, the latter a capability not available in the
source terms provided in EPACMTP. Therefore, to satisfy the multi-pathway analysis plan for this risk
assessment, the 3MRA waste management unit models are used provide mass and water fluxes to
EPACMTP for fate and transport simulations of the subsurface environment. For example, Figure C-17
DRAFT
C-27
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
(a) through (c) shows the time series of PFOA and PFOS mass fluxes and infiltrating water fluxes leaching
to the subsurface as generated for the Pasture LAU scenario in Chicago (moderate meteorological
conditions). PFOA and PFOS concentrations in the infiltrating water resulting from the corresponding
mass and infiltrating water fluxes for PFOA and PFOS, assuming a "low-Koc" value (discussed in Section
C.3) are shown in Figure C-17 (d) and (e).
8.E-04
7.E-04
6.E-04
5.E-04
J 4.E-04
u.
| 3.E-04
$
2.E-04
l.E-04
0.E+00
a
l\
11 1
n
1 |\
ft
jj\ i
InU
V 1
\1
I 1
!
td.
(a) Water Flux
20 30
Year
_ 3.5E-06
"D
> 3.0E-06
m 2.5E-06
J 2.0E-06
u.
g 1.5E-06
1 1.0E-06
¦5 5.0E-07
re
* 0.0E+00
0
(b) PFOA - Pasture
JL
JL
JL
A
p\
\h
m
A}
i 1
i
• *
I
L
10 20 30 40 50
Year
_ 5.0E-03
CLD
E, 4.0E-03
c
% 3.0E-03
E
*-»
C
U 2.0E-03
C
o
1.0E-03
ITS
§ 0.0E+00
—J
(d) PFOA - Pasture
L
0
10 20 30 40 50
Year
5.0E-03
4.0E-03
txo
E
c
| 3.0E-03
L
+-»
C
8 2.0E-03
c
o
u
w 1.0E-03
S O.OE+OO
(c) PFOS - Pasture
150
200
(e) PFOS - Pasture
50
100
Year
150
200
Figure C-17.
Simulated time series of PFOA and PFOS mass and water fluxes generated for the
pasture LAU scenario in Chicago, representing moderate meteorological conditions, (a)
infiltrating water fluxes; (b) and (c) PFOA and PFOS mass fluxes leaching to the
subsurface; and (d) and (e) PFOA and PFOS concentrations in the infiltrating water
resulting from the corresponding mass and infiltrating water fluxes.
In Figure C-17 (a), the simulated water fluxes at Chicago, representing moderate meteorological
conditions, vary between 5.5x105 m/d to 4.2 xlCT4 m/d. EPACMTP simulations presented in this report
DRAFT
C-28
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
assume a uniform water flux corresponding to the maxmimum leachate flux value at each location and
source model, as summarized in Appendix B. Note that the water fluxes only vary by source model and
geographic location and independent of chemical constituent simulated.
Similar to the water fluxes, the leachate fluxes (Figure C-17(b) and (c)) entering the subsurface vary from
year to year for PFOA and PFOS. For the Pasture LAU scenario, leachate concentration and flux
variations are greatest for the first 40 years, the modeled duration of leaching from the source in the
pasture LAU scenario.
Maximum leachate fluxes (g/m2/d) estimated by the land application and surface disposal source
models were used as input to EPACMTP to generate PFOA and PFOS concentrations. PFOA and PFOS
concentrations (Figure C-17(d) and (e)) in the infiltrating water remains constant during the modeled
duration of leaching from the source (e.g., 40 years for pasture LAU scenario). All leachate fluxes from
the source models were applied uniformly over the footprint of the either the LAU or SDU at the top of
the unsaturated soil column. Appendix B summarizes the input maximum leachate PFOA and PFOS
concentrations for various source models and geographic locations.
C.3.2 Koc
The primary chemical-specific input parameters of concern within the groundwater pathway for PFOA
and PFOS are their organic carbon distribution coefficient (Koc) and effective diffusion coefficient in
water (D*). Under natural soil-water conditions, volatilization of PFOA and PFOS is negligible (Johansson
et al., 2017; Sima and Jaffe 2021), making inputs like diffusion coefficient in air and Henry's law constant
irrelevant. PFOA and PFOS also do not degrade, so degradation rates are also not relevant. Values for D*
are straightforward and provided in Appendix B (Table B-l). Koc, however, is highly variable and this
section describes the literature review conducted to establish input values for EPACMTP modeling.
A review of measured Koc and solid phase adsorption coefficient (Kd) values reported in literature for
PFOA and PFOS was conducted with and without associations to biosolids or land application of
biosolids, for surface and subsurface soils, aquifer materials, and settled and suspended surface water
sediments. This section describes the literature search methodology, data selection and review, data
extraction, results and conclusions.
C.3.2.1 Literature Search Methodology
The EPA conducted multiple literature searches, the most recent of which was conducted on March 18,
2024, to identify papers addressing per- and polyfluoroalkyl substances (PFAS) in general as well as
PFOA and PFOS specifically. Data were analyzed and categorized employing well defined data quality
criteria, summarized, and finally evaluated for use in modeling exercises.
This section describes the overall search methodology, including the databases searched, specific search
strings, and the abstract review strategy and article selection.
The following online databases were searched:
• PubMed
• Web of Science (includes Science Citation Index Expanded, Social Sciences Citation Index, and
Conference Proceedings Citation Indexes for Science and for Social Science and Humanities)
• Environment Complete
• CAB Abstracts
• Fish, Fisheries & Aquatic Biodiversity Worldwide
• TOXLINE
DRAFT
C-29
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
• ScienceDirect.
Results were restricted to papers in English published after 1990 up through to the date of the last
literature search.
Three sets of searches were conducted: (a) a broad search that did not specify a relationship to biosolids
and included per- and polyfluoroalkyl substances (PFAS) in general; (b) a search that looked specifically
for PFOA or PFOS and for mention of biosolids or land application and other related terms; and (c) a
search focused solely on solid-phase partitioning and sorption of PFOA and PFOS. Searches (a) and (b)
were broader than just solid-phase sorption coefficients.
Broad search (a) included PFOA or PFOS as well as additional PFAS. These search strings were
formulated in the three following parts:
Part 1:
("perfluoroalkyl substance*" OR "polyfluoroalkyl substance*" OR "PFAS" OR "PFASs" OR "PFOA" OR
"perfluorooctanoic acid" OR "PFOS" OR "perfluorooctane sulfonic acid" OR "perfluorooctanesulfonic
acid" OR "PFNA" OR "perfluorononanoic acid" OR "heptadecafluorononanoic acid" OR "perfluoro-n-
nonanoic acid "OR "PFHxS" OR "perfluorohexanesulfonic acid" OR "perfluorohexane-l-sulphonic acid"
OR "Gen X" OR "GenX" OR "stain repellent*" OR "water resistant" OR "aqueous film-forming foam" OR
"AFFF" OR "perfluoroalkyl acid" OR "PFAA" OR "surfactant")
Part 2:
AND ("fate" OR "transport")
AND ("Retention" OR "Model")
AND ("Adsorption" OR "Sorption")
AND (("Field" OR "Lab" OR "Laboratory") AND ("Data" OR "Experiment"))
Part 3:
AND ("Vadose" OR "unsaturated")
AND ("leaching").
Eight individual searches were conducted that used Part 1 plus each possible pair of the separate lines in
Parts 2 and 3.
Broad search (b) targeted just PFOA and PFOS and was limited to papers published in 2017 or later.
These search strings were formulated in four parts:
Part 1 - Constituents of Concern:
("PFOA" OR "perfluorooctanoic acid" OR "PFOS" OR "perfluorooctane sulfonic acid" OR
"perfluorooctanesulfonic acid" )
Part 2 - Properties:
AND ("uptake" OR "*transfer" OR "*accumulation" OR "BCF" OR "BAF" OR "propert*" OR "health" OR
"effect" OR "diffusiv*" OR "partition*")
Part 3 - Biosolids:
AND ("sewage sludge" OR "biosolids" OR "treated sewage" OR "sludge treatment" OR "sewage
treatment")
Part 4 - Land Application:
AND ("land application" OR "farm" OR "agriculture" OR "soil")
DRAFT
C-30
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Three individual searches were conducted using the following string combinations:
Search 1: Part 1 + Part 2 (to capture the universe of PFAS and properties).
Search 2: Part 1 + Part 3 (to capture the universe of PFAS and biosolids).
Search 3: Part 1 + Part 4 (to capture the universe of PFAS and land application).
Focused search (c) targeted keywords associated with solid-phase sorption for PFOA, PFOS, and PFAS,
without limitations on publishing date. These search strings were formulated in two parts:
Part 1 - Constituents of Concern:
("PFOA" OR "perfluorooctanoic acid" OR "PFOS" OR "perfluorooctane sulfonic acid" OR
"perfluorooctanesulfonic acid" OR "PFAS")
Part 2 - Properties:
AND ("Koc" OR "partitioning coefficient" OR "organic carbon?water partitioning coefficient" OR "Kd" OR
"soil *sorption coefficient" OR "*sorption coefficient" OR "distribution coefficient" OR "solid?liquid
partitioning coefficient" OR "soil?water partitioning coefficient" " OR "*sorption")
One search was conducted using the following string combinations:
Search 1: Part 1 + Part 2 (capture the universe of PFOA and PFOS solid-phase sorption and properties).
C.3.2.2 Review and Data Selection Process
The results of the above searches were compiled, and duplicates removed, yielding 1,864 unique
articles. We added two additional sources to those: Articles cited in the PFAS Technical and Regulatory
Guidance Document and Fact Sheets (ITRC, 2022; Table 4.1) and data from the United States Geological
Survey (USGS) for New Hampshire (Tokranov et al., 2023).
The EPA reviewed the abstracts, or if no abstract was available, the titles, and categorized them for
further review for a variety of purposes based on keywords. For this review, we identified 234 articles
that mentioned Kd or Koc and PFOA or PFOS. We obtained the full text of those papers.
Upon reviewing the full text and evaluating the data quality, The EPA classified the articles into three
types, based on whether they contained biosolids-related keywords (biosolid, wastewater treatment
plant, or sewage sludge):
Studies in biosolids (46 articles)
Studies not in biosolids (169 articles)
Type could not be determined due to lack of clarity (19 articles; the EPA evaluated these further in the
data extraction step).
C.3.2.3 Data Extraction
The EPA searched each article for partitioning data (Kd and/or Koc) and identified whether the data were
from field or laboratory experiments:
Field experiments included cases where the partitioning data was estimated directly from the original
field sample condition or from the field sample spiked with PFOA/PFOS for concentration measurement
purposes.
Laboratory experiments included cases where the partitioning coefficient was estimated through
sorption/desorption or column experiments.
The EPA then assigned data to four categories:
DRAFT
C-31
-------
PFOA/PFOS Risk Assessment Appendix C. Groundwater Modeling
Type A: biosolids-related field data
Type B: biosolids-related lab data
Type C: not biosolids field data
Type D: not biosolids lab data
C.3.2.4 Search Results
Table C-8 summarizes the data extracted from the literature review articles and other data sources. In
total, the EPA extracted about 2,000 data points from 101 articles.
Table C-8. Summary of Data Extracted from Articles Identified in the Literature Search and USGS
Constituent
Scope
General Literature Review
NH/USGS Data3
log Kd
log Koc
log Kd
Field
Studies
Lab
Studies
Field
Studies
Lab
Studies
Field
Studies
Lab
Studies
PFOA
Biosolids
3
27
0
3
30
89
Not biosolids
152
231
71
200
18
133
PFOS
Biosolids
0
14
0
11
22
86
Not Biosolids
152
307
96
234
18
133
a Data reported by Tokranov et al. (2023).
The detailed results are presented by constituent below. The figures include only the literature search
data unless otherwise specified. All values are presented as log Kd or log Koc. Note that Kd is defined as
the concentration in the solid phase divided by the concentration in the aqueous phase. Accordingly, a
negative log Kd (i.e., Kd less than 1) means that less constituent is present in the solid phase than the
aqueous phase, and thus there is low solid phase sorption.
Results for PFOA
A significant finding for the purposes of identifying representative values of these parameters for
modeling is that reported log Kd and log Koc values span more than four orders of magnitude (Figure C-
18). For log Kd, the median value from field data is greater than the median for laboratory data for PFOA
(this is including both biosolids and non biosolids studies). Somewhat more than half of the PFOA data
were from laboratory data (n = 258 for lab data, n = 155 for field data). Similarly, the median log Koc for
field data is greater than the median for laboratory data. Unlike Kd studies, however, approximately
three times as many Koc observations are from laboratory studies (n = 203) than are from field studies
(n = 71), and no biosolids-related field data were identified.
The range of log Kd values for field studies not related to biosolids (Type C) is larger than the
corresponding study type related to biosolids (Type A), and the range of the biosolids-related values is
entirely encompassed within the range of not biosolids related values (Figure C-19). This may be due to
the smaller number of biosolids-related field studies (3 reported values) than non-biosolids related field
studies (152 reported values). However, the range of log Kd values are similar for lab studies related to
biosolids (Type B) and not related to biosolids (Type D). The similar range of log Kd values for lab-studies
maybe related to the larger pool of lab studies (258 reported values) than field studies (155 reported
values). Note there were no biosolids-related log Koc data for PFOA, so no equivalent comparison to the
one shown in Figure C-19 for log Kd could be made.
DRAFT
C-32
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
PFOA Log Kd field vs lab
PFOA Log Koc field vs lab
CTi
q
Shaded portion represents the interquartile range (IQR); whisker ends are quartile ±1.5xlQR; diamonds are outliers.
FE=Field, LE=lab
Figure C-18. Boxplots of log Kd (left) and log Koc (right) values for PFOA: field (FE) vs. laboratory (LE)
studies.
5 -
4-
3 -
1? 2 -
CT«
3
l -
o -
-l -
Shaded portion represents the interquartile range (IQR); whisker ends are quartile ±1,5xlQR; diamonds are outliers.
FE=Field, LE=lab
Figure C-19. Boxplot of log Kd values for PFOA: biosolids vs. not biosolids.
We further evaluated the data by sample media: soil, sludge, sediment, or suspended particulate matter
(SPM). Most log Kd and log Koc studies have been performed in soil (n=235 for Kd, n=204for Koc) or
sediment (n=122 for Kd, n=62 for Koc), with a lesser number in sludge (n=47 for Kd, n=4 for Koc), and very
few in SPM (n=2 for Kd, n=4 for Koc).
The overall range of log Kd values in soil from field and lab studies spans approximately three orders of
magnitude (log Kd approximately -1.3 to 2.1; Figure C-20). However, this is clearly dominated by non-
biosolids related lab studies for both media (green and blue bars for Types C and D in Figure C-20). The
range for biosolids-related values for soil (there are none for sediment) cover a considerably smaller
A: Biosolids, field
B: Biosolids, lab
C: Not biosolids, field
D: Not biosolids, lab
t *
E3
FE
~
~
LE
DRAFT
C-33
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
range (less than one order of magnitude within study type, though the range across field and lab studies
is still about 2 orders of magnitude; pink and brown bars for Types A and B in Figure 3). However, as
noted earlier, there are fewer biosolids-related data points, which may account for the lesser variation.
Compared to soils, sediments span a much larger range in reported values, primarily due to the field
studies (log Kd approximately -0.7 to 4.9). Median log Kd values reported for sludge and SPM are
generally higher, though of similar variability. However, the number of studies reporting log Kd in sludge
or SPM are few
5 -
4 -
3 -
£ 2-
3
l-
o -
-l -
Shaded portion represents the interquartile range (IQR); whisker ends are quartile ±1,5xiQR; diamonds are outliers.
SPM = suspended particulate matter.
Figure C-20. Boxplots of log Kd values for PFOA by matrix for field studies (left) and lab studies
(right).
As shown in Figure C-21, overall log Koc values in soil and sediment from field and lab studies are
similarly variable as log Kd values, covering about three orders of magnitude for soil (log Koc
approximately 0 to 3.5) and nearly five orders of magnitude in sediment (log Koc approximately 0.5 to
5.2). No literature values were identified for log Koc measured in field studies of soils. However,
laboratory studies of log Koc measurements in soil were reported in both biosolids and non-biosolids
studies. The reported results for soils from both biosolids and non-biosolids are comparable (Figure C-
21, plot on right) although the range of non-biosolids reported results are much larger (0 to
approximately 3.5).
Reported log Koc values for sludge are considerably less variable (log Koc approximately 2.6 to 3). Values
for SPM are more variable than sludge and somewhat less variable than soil and sediment. The SPM
values are considerably higher than other samples of the same type (i.e., field) but are based on only
four studies (all field), so additional data may be needed to assess if there is a difference between log Koc
values in SPM compared to other media.
npared to soils or sediments.
PFOA Kd field by matrix
11: 28
T
A: Biosolids, field
C: Not biosolids, field
S?
i
PFOA Kd lab by matrix
I I B: Biosolids, lab
I I D: Not biosolids, lab
si
T
* ^ ^ *
up
DRAFT
C-34
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
PFOA Koc field by matrix
PFOA Koc lab by matrix
e
T
L7'
T
D: Not Biosolids, lab
B: Biosolids, lab
cpv
¥
Aft
Cp
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Results for PFOS
As was found for PFOA, reported log Kd and log Koc values for PFOS in literature span approximately six
and five orders of magnitude, respectively (Figure C-23). For log Kd, the median value from field data and
lab data are similar (this is including both biosolids and non biosolids studies). Unlike log Koc, the median
log Kd value for field data is somewhat more than the median for laboratory data. The overall range of
reported results for log Koc and log Kd are larger for field studies as compared to laboratory studies. This
maybe because more observations are derived from laboratory studies than field studies.
PFOS Log Kd field vs lab
PFOS Log Koc field vs lab
Ol
q
6
5
4
3
2
1
0
-1
~r
FE LE FE LE
Shaded portion represents the interquartile range (IQR); whisker ends are quartile ±1,5xiQR; diamonds are outliers.
Figure C-23. Boxplots of log Kd (left) and log Koc (right) values for PFOS: field (FE) vs. laboratory (LE)
studies.
As shown in Figure C-24, the range of log Kd values (left side) based on lab data is similar for studies
related to biosolids (Type B; brown box) and studies not related to biosolids (Type D; blue box), despite
there being more non-biosolids data. For log Koc (right side, Figure C-24), the lab data associated with
biosolids (Type B; brown box) appears to be a subset of the range of reported values for non-biosolids
studies (Type D; blue box). No field studies of biosolids (Type A) were identified for either log Kd or log
Koc. However, the non-biosolids field studies (Type C; green box, Figure C-24) spans a large range that
encompasses results for other study types (both biosolids and non biosolids lab studies, Types B and D).
We further evaluated the data by sample media: soil, sludge, sediment, or suspended particulate matter
(SPM). Most log Kd and log Koc studies have been performed in soil (n=309 for Kd, n=253 for Koc) or
sediment (n=133 for Kd, n=78 for Koc), with a lesser number in sludge (n=24 for Kd, n=4 for Koc), and very
few in SPM (n=7 for Kd, n=5 for Koc).
As shown in Figure C-25, the overall range of log Kd values in soil (approximately -0.8 to 3.9) is narrower
than the range of log Kd values in sediment (approximately -1.3 to 6.2). Note that the soil values are all
lab studies, as no field studies (in biosolids or otherwise) were identified for soil.
Compared to soils and sediments, log Kd values reported for sludge and SPM are generally less variable.
Reported log Kd values for sludge range from about 2 to 3.4; studies with SPM reported much higher log
Kd values (greater than approximately 3 to 5). However, only seven field studies reported log Kd in SPM;
additional measurements may be needed to assess if there is a difference between log Kd values in SPM
compared to other media.
DRAFT
C-36
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
s2 3
Ol
3 2
B: Biosolids, lab
C: Not biosolids, field
D: Not biosolids, lab
|^|
Dl
q
FE
LE
Shaded portion represents the interquartile range (IQR); whisker ends are quartile ±1,5xlQR; diamonds are outliers.
FE=field, LE=lab
Figure C-24. Boxplots of log Kd (left) and log Koc (right) values for PFOS: biosolids vs. not biosolids.
PFOS Kd field by matrix
PFOS Kd lab by matrix
Ol
q
6 -
5 -
4 -
3
2
1 -
0 -
-1 -
~ ~r
~
n: 2
n: 114
C: Not biosolids, field
cn
3
B: Biosolids, lab
D: Not biosolids, lab
X
r11
±
_2fi!
X
i ¦ 1Q
4^
4*
4
i^
cfr*
iff
.«
Shaded portion represents the interquartile range (IQR); whisker ends are quartile ±1,5xlQR; diamonds are outliers.
SPM = suspended particulate matter.
Figure C-25. Boxplots of log Kd values for PFOS by matrix for field studies (left) and lab studies (right).
As shown in Figure C-26, the log Koc values in soil span about four orders of magnitude, from about 1 to
5.4, while sediment log Koc values vary more, about five orders of magnitude, from about 0.7 to 5.9.
Closer inspection reveals that the larger range in sediment log Koc values is due to the very wide range
(including many outliers) in reported values from field studies that are not biosolids related (green box
for sediment, Figure C-26 left panel). Biosolids-related laboratory studies for sediments report a much
smaller range of log Kd values approximately between 1 and 3.6 but this maybe an artifact of the
number of available studies or perhaps less variability of sediment conditions in the lab compared to the
DRAFT
C-37
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
field. There are fewer biosolids-related laboratory study reported values (n=21) in comparison to field
studies (n=57).
Compared to soils and sediments, log Koc values reported for SPM, where available, are generally higher
(Figure C-26). However, the number of studies reporting log Kd in SPM (n=5) are much fewer than soils
(n=253) or sediments (n =78).
PFOS Koc field by matrix
PFOS Koc lab by matrix
I
B: Biosolids, lab
D: Not biosolids, lab
IU-2J o
V*
A*
s?
<=5?
Shaded portion represents the interquartile range (IQR); whisker ends are quartile ±1,5xlQR; diamonds are outliers.
SPM = suspended particulate matter.
Figure C-26. Boxplots of log Koc values for PFOS by matrix for field studies (left) and lab studies
(right).
Figure C-27 compares log Kd values reported in literature for soils alongside those reported by USGS for
NH (Tokranov et al., 2023). The range of log Kd values reported is broken down by study type (Type A
through D), where available, to facilitate closer inspection. Generally, the range of reported log Kd values
in literature and by the USGS compare well, but the NH data spans a smaller range (log Kd between 0.5
and approximately 3) in comparison to the literature dataset (log Kd between -0.8 and 4). Additionally, in
both datasets, while median log Kd values for biosolids laboratory studies are higher than those for non-
biosolids studies, they are similar when comparing for the same conditions and content which was not
the case for PFOA.
DRAFT
C-38
-------
PFOA/PFOS Risk Assessment
Literature
Appendix C. Groundwater Modeling
New Hampshire
tn
o
O
x
X
A: Biosolids, field
B: Biosolids, lab
C: Not biosolids, field
D: Not biosolids, lab
Soil Soil
Shaded portion represents the interquartile range (IQR); whisker ends are quartiie ±1,5xlQR; diamonds are outliers.
Figure C-27. Boxplots of log Kd values in soil for PFOS by from literature and USGS/NH,
C.3.2.5 Conclusions
Overall, the results for log Kd found in the literature remain consistent with older review papers. For
example, Li et ai. (2018) presents quartiie plots of log Kd for PFOA and PFOS based on data retrieved
from 28 peer-reviewed articles and reports spanning 2001 to 2017 (Figure C-28). These authors reported
that measured values from the field for log Kd are greater than laboratory data, and log Kd values for
PFOS are greater than PFOA. Median values from Li et al. (2018) agree very well with data presented in
this review. A comparable plot for log Koc was not identified in the literature.
Field-derived data
—| Laboratory-derived data
4 -I
O)
^3
1.11
.
' PFOS '
1 PFOA '
"T"
PFNA 1 1 PFDA 1
PFAS chemicals
Shaded portion represents the interquartile range (IQR); whisker ends are quartiie ±1,5xlQR; diamonds are outliers.
Figure C-28. Data from Li et al. (2018) showing median values of log Kd in field and laboratory
studies.
DRAFT
C-39
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
The trend of field studies yielding higher values than laboratory studies was consistent everywhere
except for log Koc for PFOS (Figure C-23, right side); several reasons may account for this, including an
artifact of the field vs. lab classification scheme used in this analysis, a difference in the number of data
points between field (n=95) and lab studies (n=245) or due to some other unidentified reason.
Although the number of biosolids-associated data points are lower than non-biosolids, the range of
biosolids-associated values are usually captured within the range of non-biosolids oriented studies. The
spread and median values of log Koc tend to show less variability across the various categories of field
and laboratory studies, biosolids and non-biosolids related studies, and across media than log Kd. This
behavior is reasonable as log Kd additionally reflects the effects of organic carbon (OC) variability in
various matrices, as well as other parameters (e.g., pH), whereas log Koc does not.
In terms of using the information gathered in this review for predictive purposes within the current risk
assessment framework, log Koc would be preferred for several reasons. First, sampling a matrix-specific
(soil, sediment, or SPM) value of Kd with an implicit organic content value would not likely be the same
as the organic carbon content in the same simulated matrix based on soil survey information (e.g.,
gSSURGO18). Media matching of Kd would also be limited to soils and sediments. Using Koc would remove
that potential inconsistency, letting the matrix organic carbon content determine the value of Kd, and
there is more data available on organic content on a national scale than media specific values of Kd.
Consequently, Koc was used as an input parameter to EPACMTP along with estimated location-specific
fraction organic carbon (discussed below) to compute location-specific solid phase adsorption
coefficient, Kd. All simulations were performed using a "low-Koc" and a "high-Koc" value for both PFOA
and PFOS (Table C-9). The "low-Koc" and "high-Koc" values represent the closest-reported literature
values corresponding to the 10th percentile and the 90th percentile values of the corresponding
distribution for the soil matrix.
Table C-9. EPACMTP input parameter values for PFOA and PFOS organic carbon partition coefficient
Scenario
PFOA Koc (cm3/g)
PFOS Koc (cm3/g)
Low-Koc
26 (Hubert, M. etal., 2023)
250 (Johnson et al., 2007)
High-Koc
1,100 (Campos-Pereira, H. et al., 2023)
22,000 (Chen, X. T. et al., 2020)
C.3.3 Environment-specific Parameters
EPACMTP requires information about soil and aquifer properties as model inputs.
C.3.3.1 Soil Properties
For soils, EPACMTP uses soil texture as a key to generate consistent hydrological properties for the
unsaturated zone model. The primary data source for soil properties was the Soil Survey Geographic
(SSURGO) database. SSURGO is a repository of nationwide soil properties collected by the National
Cooperative Soil Survey over the last century (USDA, 2017). SSURGO data were collected at scales
ranging from 1:12,000 to 1:63,360 and are linked to map unit polygons ranging between 1 and 10 acres.
These map units provide the finest spatial resolution and span most of the conterminous United States.
Soil attributes linked to these map unit polygons are stored within a relational database broken out by
soil component and soil horizon. Each map unit contains data on the prevalence of each component and
horizon within the map unit. Table C-10 shows the crosswalk used to assign the SSURGO detailed soil
textures to basic Soil Conservation Service (SCS) textures, and then to the EPACMTP mega textures.
SSURGO soils are classified into 21 texture classes, which map to 12 SCS textures. EPACMTP uses three
18 https://www.nrcs.usda.gov/resources/data-and-reports/gridded-soil-survev-geographic-gssurgo-database
DRAFT
C-40
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
soil mega textures to represent the variability of hydrologic soil properties, so each SSURGO soil texture
was cross walked to the EPACMTP mega texture with the most similar hydrogeologic properties.
The dominant soil texture was estimated by computing the percentages of the three mega-textures
(Silty Clay Loam, Silty Loam and Sandy Loam) within a 5-mile radius of each geographic location: Boulder
(lat/long: 40.037361, -105.228139), Chicago (lat/long: 41.979444, -87.904444) and Charleston (lat/long:
32.898611,-80.040833). As shown in the Unsaturated Zone section of Appendix B, site-specific soil
texture model inputs reflect the dominant mega texture of Sandy Loam at Boulder, Silty Clay Loam for
Chicago and Silty Loam for Charleston. These model inputs include saturated hydraulic conductivity
(SATK), van Genuchten soil moisture parameters (ALPHA and BETA), residual and saturated water
contents (WCR and WCS, respectively), percent organic matter (POM), and soil bulk density.
Table C-10. Soil Texture Crosswalk
Detailed SSURGO Soil
Basic SCS
EPACMTP Soil
Texture
Texture
Mega texture
Loamy Sand
Loamy Coarse Sand
Loamy Sand
Loamy Fine Sand
Loamy Very Fine Sand
Sand
Coarse Sand
Sand
Sandy Loam
Fine Sand
Very Fine Sand
Sandy Loam
Coarse Sandy Loam
Sandy Loam
Fine Sandy Loam
Very Fine Sandy Loam
Silt Loam
Silt Loam
Silt
Silt
Loam
Loam
Silt Loam
Sandy Clay Loam
Sandy Clay
Loam
Clay Loam
Clay Loam
Silty Clay Loam
Silty Clay
Loam
Sandy Clay
Sandy Clay
Silty Clay Loam
Silty Clay
Silty Clay
Clay
Clay
C.3.3.2 Hydrogeologic En vironment
Each location modeled in this analysis was assigned a hydrogeologic environment from EPA's
Hydrogeologic Database (HGDB) to characterize four subsurface parameters required by EPACMTP:
depth to ground water, aquifer thickness, hydraulic gradient, and saturated hydraulic conductivity (see
Appendix B). The HGDB was developed by the American Petroleum Institute (Newell et al., 1989; 1990)
to specify correlated empirical probability distributions of these four parameters for the 12 distinct
hydrogeologic environments described in Newell et al. (1990).
To assign appropriate aquifer conditions to each unit's geographic location, EPA first developed a
national geographic map of the 12 hydrogeologic environments (Figure C-29). The following individual
map layers were combined using GIS software to develop a single map layer for assigning the 12
hydrogeologic environments across the United States:
DRAFT
C-41
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Shallowest principal aquifers from Principal Aquifers of the Conterminous United States, Hawaii, Puerto
Rico, and the US Virgin Islands [USGS map file: aquifrp025]. 1:2,500,000 map scale, was used as the base
layer in the assessment and to delineate several of the 12 hydrogeologic environments.
Alluvial and glacial aquifers from Aquifers of Alluvial and Glacial Origin [USGS map file: alvaqfp025],
1:2,500,000 map scale, was used to represent alluvial and glacial aquifers for the 22 states north of the
southernmost line of glaciation. Note that the alluvial aquifers in this coverage are identical to those in
the Hunt (1979) surficial geology layer below.
Surficial geology of the conterminous United States was taken from:
• Surficial Geology of the Conterminous United States [map file: geol75m]. 1:7,500,000 map scale,
provided by Hunt (1979), these data were used to characterize shallow soil lithology and alluvia!
aquifers.
• The Surficial Deposits and Materials in the Eastern and Central United States (East of 102
degrees West Longitude) [map file: sfgeoep020], 1:1,000,00 map scale, includes the line of
maximum glacial advance and represents surficial materials that accumulated or formed during
the past two million years, including residual soils, alluvium, and glacial deposits.
Karst aquifers from Engineering Aspects of Karst [map file: karst0p075], l:7,500,000-map scale, showing
karst and pseudokarst [i.e., karst-like terrain produced by processes other than the dissolution of rocks)
across the United States.
Bedrock geology from Generalized Geologic Map of the United States [map file: geolgyp075], 1:7,500,00
map scale, showing the bedrock geology at or near land surface (i.e., beneath surficial soils, alluvium and
glacial deposits).
STATSGO soils, 1:250,000 map scale, from the digital map and attribute data for soils.
# Modeled locations (HGDB)
¦] Igneous and Metamorphic Rocks
1 Solution Limestone
HGDB Environments
J Outwash
1 Till and Till Over Outwash
1 Alluvial Basins, Valleys and Fans
[~ River Alluvium with Overbank Deposits
HI Till Over Sedimentary Rock
1 II Bedded Sedimentary Rock
1 1 River Alluvium without Overbank Deposits
1 1 Unconsol. & Semi-consol. Shallow Sfc Aquifers
Coastal Beaches
1 Sand and Gravel
Figure C-29. National geographic map of the 12 hydrogeologic environments developed by EPA.
DRAFT
C-42
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
To create the hydrogeologic environment layer, each individual data layer described above was obtained
as a GIS shapefile and processed, as needed, to ensure that coordinate systems matched and the layers
could be overlain. Additional details of the data used to parameterize the unsaturated zone and the
development and use of the HGDB are given in the EPACMTP Parameters/Data Background Document
(US EPA, 2003b). The national hydrogeologic environment layer developed in GIS was used for assigning
an aquifer type to each geographic location of interest: Boulder (lat/long: 40.037361, -105.228139),
Chicago (lat/long: 41.979444, -87.904444) and Charleston (lat/long: 32.898611,-80.040833). Given an
aquifer code setting for each application unit, a correlated sample of key aquifer model input
parameters (hydraulic conductivity, hydraulic gradient, depth to the water table, and saturated
thickness) was selected from a population of samples taken from similar hydrogeological settings.
C. 3.3.3 Other Calculated Environment-Specific Input Parameters
Unsaturated Zone Longitudinal Dispersivity
Dispersion is the phenomenon by which a dissolved constituent in soil or ground water is mixed with
uncontaminated water and becomes reduced in concentration at the perimeter of the plume. Not all of
a constituent is traveling at the same velocity, due to differences in pore size and flow path length and
friction along pore walls, resulting in mixing along the flow path which decreases solute concentrations.
Note that the unsaturated zone longitudinal dispersivity is measured along the path of flow in the
downward direction. For the current risk assessment, longitudinal dispersivity is calculated as a linear
function of the total depth of the unsaturated zone according to the following equation which is based
on a regression analysis of data presented by EPRI (1985) and has a correlation coefficient of 0.66:
aLu = 0.02 + 0.022 Du
where,
aLu = longitudinal dispersivity (m)
Du = total depth of the unsaturated zone (m).
Saturated Zone Longitudinal, Transverse and Vertical Dispersivity
The longitudinal dispersivity is the characteristic length that defines spatial extent of dispersion of
contaminants, measured in the longitudinal direction, that is, along the flow path or in the X-direction.
The longitudinal dispersivity is calculated using equation 5.11 of the EPACMTP Parameters/Data
Background Document (US EPA, 2003b) based on a receptor well distance of 30 meters and a reference
dispersivity corresponding to 1 meter. The horizontal transverse dispersivity is calculated as l/8th the
longitudinal dispersivity in accordance with equation 5.13 of the EPACMTP Parameters/Data
Background Document (US EPA, 2003b). Similarly, the vertical dispersivity is calculated as l/160th the
longitudinal dispersivity in accordance with equation 5.14 of the EPACMTP Parameters/Data
Background Document (US EPA, 2003b).
Fraction organic carbon
The fraction organic carbon in the soil or aquifer is estimated from the location-specific percentage of
organic matter (see Table B-9, Appendix B) by dividing it by a factor of 174 in accordance with equation
3.3 of the EPACMTP Parameters/Data Background Document (US EPA, 2003b).
Recharge
Recharge is water percolating through the soil to the aquifer outside the footprint of the unit. Typically,
EPACMTP selects a recharge rate using a meteorological station assignment (based on the geographic
location and topography of a unit setting) and by the unit's associated soil texture mentioned above.
Using the soil texture and station assignment, a recharge rate was computed using the HELP model (US
EPA, 2020) and using the nearest OPP synthetic weather data (Fry et al., 2016). Further details about
DRAFT
C-43
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
how these rates were determined and other options for determining recharge rates outside of the
EPACMTP model can be found in the EPACMTP Parameters/Data Background Document (US EPA,
2003b).
C.4 Model Outputs
The output of EPACMTP is a prediction of the contaminant concentration arriving at a downgradient
groundwater receptor location, and is a time-dependent concentration, corresponding to the finite
source scenario. The model can calculate both the peak concentration arriving at the well and maximum
time-averaged concentrations. In this analysis, peak concentrations were used to develop human risk
estimates. Because the subsurface migration of PFOA and PFOS may be very slow, it may take a long
time for the plume to reach the receptor well, and the maximum exposure may not occur until a very
long time after the land application ceases. For example, the peak arrival time for PFOA and PFOS at the
receptor well varies between 34 years and 9974 years across the modeled locations, scenarios and
chemicals. Therefore, for this analysis, maximum exposures occurred within EPACMTP's maximum
default time horizon of 10,000 years.
Table C-ll provides a summary of the simulated PFOA and PFOS exposure concentrations at a receptor
well located at 5 meters from the edge of the farm field (i.e., center of the buffer). As noted in Section
C.3, the exposure concentration for this evaluation is the maximum of the peak concentrations at the
receptor well across four well depths (0.5 m, 1.0 m, 1.5 m and 2.0 m below the water table). Exposure
concentrations reported in Table C-ll were used as inputs to calculate risks for the groundwater
pathway.
Table C-ll. Groundwater Pathway Receptor Exposure Concentrations: Maximum of the Peak
Concentrations at a Receptor Well Located at the Center of the Buffer across Four Well
Depths (mg/L)
Scenario
Low Koc
High Koc
Dry | Average | Wet
Dry | Average | Wet
PFOA
LAU (Crop)
1.7E-03
8.8E-03
7.4E-03
4.4E-34
4.1E-15
1.2E-06
LAU (Pasture)
4.1E-03
5.8E-03
3.2E-03
5.2E-34
2.9E-06
1.5E-05
LAU (Reclamation)
4.0E-03
1.2E-03
4.0E-03
1.8E-35
3.9E-11
5.9E-09
SDU (No Liner)
1.9E-02
1.3E-02
1.3E-02
4.7E-11
7.6E-15
3.4E-07
SDU (Clay Liner)
1.7E-02
8.3E-03
9.8E-03
8.7E-14
8.0E-16
1.7E-07
SDU (Composite Liner)
1.7E-05
1.6E-06
3.8E-06
2.8E-35
3.0E-35
4.5E-35
PFOS
LAU (Crop)
1.2E-14
1.8E-05
2.3E-05
8.3E-35
7.8E-35
5.7E-16
LAU (Pasture)
1.6E-16
8.2E-05
4.7E-04
1.4E-34
1.2E-15
1.3E-07
LAU (Reclamation)
1.5E-06
2.2E-06
6.5E-06
5.2E-36
1.6E-36
1.7E-08
SDU (No Liner)
1.5E-05
4.4E-06
4.4E-05
1.9E-36
1.9E-36
8.0E-13
SDU (Clay Liner)
1.1E-05
2.9E-06
2.0E-05
1.9E-36
1.9E-36
1.0E-15
SDU (Composite Liner)
6.9E-34
6.8E-34
6.6E-34
1.9E-36
2.0E-36
2.9E-36
DRAFT
C-44
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
C.5 References
An, Y. J., Carraway, E. R., & Schlautman, M. A. (2002). Solubilization of polycyclic aromatic hydrocarbons
by perfluorinated surfactant micelles. Water Research, 36(1), 300-308.
Anderson, R. H., Long, G. C., Porter, R. C., & Anderson, J. K. (2016). Occurrence of select perfluoroalkyl
substances at U.S. Air Force aqueous film-forming foam release sites other than fire-training areas:
Field-validation of critical fate and transport properties. Chemosphere,
http://dx.doi.Org/10.1016/i.chemosphere.2016.01.014
Anderson, R. H., Adamson, D. T., & Stroo, H. F. (2019). Partitioning of poly-and perfluoro- alkyl
substances from soil to groundwater within aqueous film-forming foam source zones. J. Contam.
Hydro!., 220, 59-65.
Brusseau, M. L. (2018). "Assessing the potential contributions of additional retention processes to PFAS
retardation in the subsurface." Science of The Total Environment, 613-614, 176-185.
https://doi.Org/10.1016/i.scitotenv.2017.09.065
Brusseau, M. L. (2020). Simulating PFAS transport influenced by rate-limited multi-process retention.
Water Research, 168, 115179.
Brusseau, M. L., Guo, B., Huang, D., Yan, N., & Lyu, Y. (2021). Ideal versus nonideal transport of PFAS in
unsaturated porous media. Water Res., 117405.
Campos-Pereira, H., Kleja, D. B., Ahrens, L., Enell, A., Kikuchi, J., Pettersson, M., & Gustafsson, J. P.
(2023). Effect of pH, surface charge and soil properties on the solid-solution partitioning of
perfluoroalkyl substances (PFASs) in a wide range of temperate soils. Chemosphere, 321, 138133-
138141.
Chen, X., Yu, P., Xiang, L., Zhao, H., Li, Y., Li, H., Zhang, X., Cai, Q„ Mo, C., & Wong, M. H. (2020).
Dynamics, thermodynamics, and mechanism of perfluorooctane sulfonate (PFOS) sorption to
various soil particle-size fractions of paddy soil, Ecotox. and Env. Safety, 206, 111105-111116.
EPRI. (1985). A review of field scale physical solute transport processes in saturated and unsaturated
porous media. Electric Power Res. Inst., Palo Alto, CA.
Fry, M. M., Rothman, G., Young, D. F., & Thurman, N. (2016). Daily gridded weather for pesticide
exposure modeling. Environmental Modelling & Software, 82, 167-173.
Guelfo, J. L., Wunsch, A., McCray, J., Stults, J. F., &Higgins, C.P. (2020). Subsurface transport potential of
perfluoroalkyl acids (PFAAs): Column experiments and modeling. J. Contam. Hydrol., 233, 103661
Guo, B., Zeng, J., & Brusseau, M. L. (2020). A mathematical model for the release, transport, and
retention of per-and polyfluoroalkyl substances (PFAS) in the vadose zone. Water Resour. Res., 56,
e2019WR026667.
Guo, B., Zeng, J., Brusseau, M. L., & Zhang, Y. (2022). A screening model for quantifying PFAs leaching in
the vadose zone and mass discharge to groundwater. Adv. Water Res., 160 (2022), Article 104102.
Higgins, C.P., & Luthy, R.G. (2006). Sorption of perfluorinated surfactants on sediments. Environmental
Science & Technology, 40(23), 7251-7256
Hubert, M., Arp, H.P.H., Hansen, M. C., Castro, G., Meyn, T., Asimakopoulos, A. G., & Hale, S. E. (2023).
Influence of grain size, organic carbon and organic matter residue content on the sorption of per-
and polyfluoroalkyl substances in aqueous film forming foam contaminated soils - Implications for
remediation using soil washing. Science of The Total Environment, 875, 162668-162680.
DRAFT
C-45
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Hunt, Charles D., 1979, National Atlas of the United States of America-Surficial Geology: U.S. Geological
Survey, NAC-P-0204-75M-0, [map file: geol75m].
Interstate Technology & Regulatory Council (ITRC).(2022). PFAS Technical and Regulatory Guidance
Document and Fact Sheets PFAS-1. Washington, D.C.: Interstate Technology & Regulatory Council,
PFAS Team, https://pfas-l.itrcweb.org/
Johansson, J. H., Yan, H., Berger, U., & Cousins, I. T. (2017). Water-to-air transfer of branched and linear
PFOA: Influence of pH, concentration and water type. Emerging Contaminants, 3(1), 46-53.
https://doi.Org/10.1016/i.emcon.2017.03.001
Johnson, R., Anschutz, A. J., Smolen, J. M., Simcik, M. F., & Penn, R. L. 2007. "The Adsorption of
Perfluorooctane Sulfonate onto Sand, Clay, and Iron Oxide Surfaces. "Journal of Chemical
Engineering Data 52: 1165-1170.
Kalam, S., Abu-Khamsin, S. A., Kamal, M. S., & Patil, S. (2021). Surfactant adsorption isotherms: A review.
ACS Omega, 6(48), 32342-32348 DOI: 10.1021/acsomega.lc04661
Kostarelos, K., Sharma, P., Christie, E., Wanzek, T., & Field, J. (2020). Viscous microemulsions of aqueous
film-forming foam (AFFF) and jet fuel a inhibit infiltration and subsurface transport. Environ. Sci.
Technol. Lett., 8, 2, 142-147. DOI: 10.1021/acs.estlett.0c00868
Li, Y., Oliver, D. P., & Kookana, R. S. (2018). A critical analysis of published data to discern the role of soil
and sediment properties in determining sorption of per and polyfluoroalkyl substances (PFASs). Sci
Total Env., 628-629, 110-120.
Liu, J. X., & Lee, L. S. (2005). Solubility and sorption by soils of 8:2 fluorotelomer alcohol in water and
cosolvent systems. Environ. Sci. Technol., 39, 7535-7540.
Milinovic, J., Lacorte, S., Vidal, M., Rigol, A. (2015). Sorption behaviour of perfluoroalkyl substances in
soils. Sci. Total Environ., 511, 63-71. http://dx.doi.Org/10.1016/i. scitotenv.2014.12.017
Milinovic, J., Lacorte, S., Rigol, A., & Vidal, M. (2016). Sorption of perfluoroalkyl substances in sewage
sludge. Environ. Sci. Pollut. Res., 23, 8339-8348. http://dx.doi.org/10.1007/sll356-015-6Q19-9
Mualem, Y. (1976) A new model for predicting the hydraulic conductivity of unsaturated porous media.
Water Resources Research, 12, 513-522. http://dx.doi.org/10.1029/WR012i003pQ0513
Newell, C. J., Hopkins, L. P., & Bedient, P. B. (1989). Hydrogeologic Database for Ground Water
Modeling. API Publication No. 4476. American Petroleum Institute, Washington, DC.
Newell, C. J., Hopkins, L. P., & Bedient, P. B. (1990). Hydrogeologic database for ground water modeling.
Ground Water, 28(5), 703-714.
Peng, S., & Brusseau, M.L. (2005). Impact of soil texture on air-water interfacial areas in unsaturated
sandy porous media. Water Resour. Res., 41(3).
Pepper, I., Kelley, C., & Brusseau, M. (2023). Is PFAS from land applied municipal biosolids a significant
source of human exposure via groundwater? Sci. of the Total Env., 864(161154).
Pereira, H. C., Ullberg, M., Kleja, D. B., Gustafsson, J. P., & Ahrens, L. (2018). Sorption of perfluoroalkyl
substances (PFASs) to an organic soil horizon-Effect of cation composition and pH. Chemosphere,
207, 183-191.
Place, B. J., & Field, J. A. (2012). Identification of novel fluorochemicals in aqueous film-forming foams
used by the US military. Environmental Science & Technology 46 (13):7120-7127.
https://doi.org/10.1021/es3Q1465n
DRAFT
C-46
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Schaefer, C. E., Nguyen, D., Christie, E., Shea, S., Higgins, C. P., & Field, J. A. (2021). Desorption of poly-
and perfluoroalkyl substances from soil historically impacted with aqueous film-forming foam. J.
Environ. Eng. 147, 06020006.
Silva, J. A. K., Simunek, J., & McCray, J. E. (2020). A modified HYDRUS model for simulating PFAS
transport in the vadose zone. Water, 12, 2758.
Sima, M. W., & Jaffe, P. R. (2021). A critical review of modeling poly- and perfluoroalkyl substances
(PFAS) in the soil-water environment. Science of The Total Environment, 757, 143793.
https://doi.Org/10.1016/i.scitotenv.2020.143793.
Simmons, R. N., & McGuffin, V.L. (2007). Modeling transport effects of perfluorinated and hydrocarbon
surfactants in groundwater by using micellar liquid chromatography. Analytica Chimica Acta, 603(1),
93-100.
Tokranov, A. K., Welch, S. M., Santangelo, L. M., Kent, D. B., Repert, D. A., Perkins, K., Bliznik, P. A., Roth,
D. A., Drouin, A. F., Lincoln, T. A., Deyette, N. A., Schlosser, K. E. A., & Marts, J.M. (2023). Solid/Water
Partitioning of Per- and Polyfluoroalkyl Substances (PFAS) in New Hampshire Soils and Biosolids:
Results from Laboratory Experiments at the US Geological Survey: US Geological Survey data release.
USDA (Department of Agriculture). (2017). Soil Survey Staff, Natural Resources Conservation Service,
United States Department of Agriculture. Web Soil Survey.US EPA (Environmental Protection
Agency) (1997). EPA's Composite Model for Leachate Migration with Transformation Products:
User's Guide. Office of Solid Waste. Washington, DC.
US EPA (Environmental Protection Agency). (1997). EPA's Composite Model for Leachate Migration with
Transformation Products (EPACMTP): User's Guide. US Environmental Protection Agency, Office of
Solid Waste, Washington, D.C. 20460.
US EPA (Environmental Protection Agency). (2003a). EPA's Composite Model for Leachate Migration
with Transformation Products (EPACMTP): Technical Background Document. US Environmental
Protection Agency, Office of Solid Waste, Washington, D.C. 20460. US EPA 530-R-03-002.
US EPA (Environmental Protection Agency). (2003b). EPA's Composite Model for Leachate Migration
with Transformation Products (EPACMTP): Parameter/Data Background Document. US
Environmental Protection Agency, Office of Solid Waste, Washington, D.C. 20460. US EPA 530-R-03-
003.
US EPA (Environmental Protection Agency). (2003c). Multimedia, Multipathway, and Multireceptor Risk
Assessment (3MRA) Modeling System Volume I: Modeling System and Science. US Environmental
Protection Agency, Office of Solid Waste, Washington DC. 530-D-03-001a.
US EPA (Environmental Protection Agency). (2003d). Addendum to the EPACMTP Technical Background
Document. US Environmental Protection Agency, Office of Solid Waste, Washington, D.C. 20460.
Available as document ID number EPA-HQ-RCRA-2003-0001-0166 at www.regulations.gov.
US EPA (Environmental Protection Agency). (2020). Hydrologic Evaluation of Landfill Performance, HELP
4.0 User Manual. Office of Research and Development, Cincinnati, OH. https://www.epa.gov/land-
research/hydrologic-evaluation-landfill-performance-help-model
van Genuchten, M. Th. (1980). A closed-form equation for predicting the hydraulic conductivity of
unsaturated soils, Soil Sci. Soc. J., 44, 892-898.
Wang, F., Shih, K., & Leckie, J. O. (2015).Effect of humic acid on the sorption of perfluorooctane
sulfonate (PFOS) and perfluorobutane sulfonate (PFBS) on boehmite. Chemosphere, 118, 213-218.
DRAFT
C-47
-------
PFOA/PFOS Risk Assessment
Appendix C. Groundwater Modeling
Wei, C., Song, X., Wang, Q., Hu, Z.H., 2017. Sorption kinetics, isotherms and mechanisms of PFOS on soils
with different physicochemical properties. Ecotoxicol. Environ. Saf. 142, 40-50.
Zeng, J., & Guo, B. (2021). Multidimensional simulation of PFAS transport and leaching in the vadose
zone: Impact of surfactant-induced flow and subsurface heterogeneities. Adv. Water Resour. 155
(104015).
Zhang, W., & Guo, B. (2024). Anomalous adsorption of PFAS at the thin-water-film air-water interface in
water-unsaturated porous media. Water Resources Research, 60,
https://doi.org/10.1029/2023WR035775
DRAFT
C-48
-------
PFOA/PFOS Risk Assessment
APPENDIX D. SENSITIVITY ANALYSIS
D.l Introduction
Appendix D. Sensitivity Analysis
Sensitivity analysis is the evaluation of model input parameters to see how they affect model outputs,
thereby providing a fundamental understanding of the simulated system (Reilly and Harbaugh, 2004). In
the current study, a sensitivity analysis was performed to understand the sensitivity of downstream
models and predicted risk outputs to individual constituent-specific and environment-specific
parameters. The sensitivity of predicted risk outputs from inputs to two downstream models, EPACMTP
and VVWM are discussed here.
D.2 Methods
The sensitivity analysis of the EPACMTP model was conducted such that the model input value for a
single parameter was varied at a time and the change in the ratio of initial PFOA and PFOS source
concentrations in biosolids leachate to predicted PFOA and PFOS concentrations at an observation well
located in the center of the buffer, at 5 meters from the source. The ratio of initial PFOA and PFOS
source concentrations in biosolids leachate to predicted PFOA and PFOS concentrations at an
observation well is directly proportional to the predicted risk outputs and is therefore an appropriate
metric to understand model sensitivity.
In contrast, the sensitivity analysis for VVWM was targeted by reviewing the governing fate and
transport equations within the model, ignoring irrelevant pathways for PFOA and PFOS (e.g.,
degradation) and only testing sensitivity of parameters from appropriate pathways.
Additional discussion on reducing relevant pathways and parameter selection is provided below.
D.2.1 EPACMTP
Model sensitivity was tested for environment- and chemical-specific input parameters for both PFOA
and PFOS using three source models or scenarios including crop, pasture and surface impoundment at
multiple geographic locations as described below.
D. 2.1.1 Unsaturated and Saturated Zone Parameters
For each chemical and source model, model sensitivity was tested at two locations, Boulder and
Charleston, with bounding meteorological conditions of dry and wet, respectively and assuming three
representative source models, crop, pasture and surface impoundment/disposal unit with clay liner. For
each chemical, source model and location, model sensitivity was tested one-at-a-time for 13 EPACMTP
input parameters consisting of 8 unsaturated zone and 5 saturated zone parameters. A total of 312
EPACMTP model simulations were performed for this sensitivity analysis.
The 13 EPACMTP input parameter sensitivities tested are listed in Table D-l. For each parameter,
bounding values were selected from the cumulative frequency distribution generated by performing a
representative nationwide landfill modeling analysis using the regional site-based modeling
methodology as reported in the EPACMTP Parameters/Data Background Document (US EPA, 2003).
Selection of bounding values from this cumulative frequency distribution for input parameter
sensitivities is appropriate as it reflects nationwide variability of these parameters. For each parameter,
the lower and upper bound values were selected as the 10th and 90th percentile of the cumulative
frequency distribution, respectively. However, adjustment of some parameter inputs from the 10th
percentile lower bound value were made to the crop and pasture source model scenarios because these
parameter combinations resulted in the violation of underlying assumptions of the model (e.g.,
excessive water table mounding). Adjusted values for these parameter inputs are also noted in Table D-
1.
DRAFT
D-l
-------
PFOA/PFOS Risk Assessment Appendix D. Sensitivity Analysis
Table D-l. Saturated and Unsaturated Zone Parameters Tested and Corresponding Lower and Upper
Bound Values Tested
EPACMTP
Lower
Upper
Model
bound
bound
Parameter
Code
Units
(10th %ile)
(90th %ile)
Reference3
Unsaturated Zone
Saturated hydraulic conductivity
US01
cm/hr
6.79E-03
1.93E+00
Table 5.5
Van Genuchten alpha parameter, a soil-
US02
cm-1
5.96E-03
5.90E-02
Table 5.7
specific shape parameter
Van Genuchten beta parameter, a soil-
US03
unitless
1.20E+00
1.82E+00
Table 5.8
specific shape parameter
Residual water content
US04
unitless
4.89E-02
9.37E-02
Table 5.9
Saturated water content (effective porosity)
US05
unitless
4.10E-01
4.50E-01
Table 5.10
Depth from ground surface to water table
US06
m
1.68E+00
3.96E+00b
4.27E+01
Table 5.2
Percent organic matter
US08
%
1.05E-01
2.15E-01
Table 5.12
Bulk density of unsaturated soil
US09
g/cm3
1.60E+00
1.67E+00
Table 5.11
Saturated Zone
Effective porosity of aquifer
AS02
unitless
3.00E-05
6.94E-01
Wolff (1982)
Aquifer soil bulk density
AS03
g/cm3
1.30E+00
1.70E+00
Table 5.18
Thickness of saturated zone
AS04
m
4.27E+00
1.43E+01C
9.14E+01
Table 5.20
Hydraulic conductivity of saturated zone
AS05
m/yr
1.73E+02
3.15E+04
Table 5.21
(aquifer)
Reqional hydraulic qradient in the aquifer
AS07
m/m
9.00E-04
3.10E-02
Table 5.22
a Table references correspond to the EPACMTP Parameters/Data Background Document (US EPA, 2003).
b A lower bound value at the 25th percentile for depth to water table from ground surface was used for simulations in Charleston to
account for the simulated water table rising above ground surface resulting in a mounding violation within the model,
c A lower bound value at 50th percentile for saturated zone thickness was used for simulations in Charleston to account for the
simulated water table rising above ground surface resulting in a mounding violation within the model
D.2.1.2 Chemical-specific Parameters
Model sensitivity to the PFOA and PFOS organic carbon partition coefficient parameter (Koc, cm3/g),
which is the ratio of a constituent's concentration in a theoretical soil containing only organic carbon to
its concentration in the ground water was tested at three locations, Boulder, Chicago and Charleston,
representing dry, moderate and wet meteorological condition, respectively and assuming two
representative source models, crop and pasture. For each chemical, source model and location, model
sensitivity was tested by varying the organic partition coefficient (Koc) one-at-a-time between a "low-
K0c", "representative-Koc" and "high-Koc" value as shown in Table D-2.
Table D-2. Values of Organic Carbon Partition Coefficient (koc) Tested for PFOA and PFOS
PFOA Koc
PFOS
Sensitivity Scenario
(cm3/g)
Koc (cm3/g)
Representative-Koc
114.8
371.5
Low-Koc
2.88
2207
High-Koc
19,953
108,081
PFOA and PFOS Koc values of 114.8 cm3/g and 371.5 cm3/g, respectively, were used as the
"representative" value as these Koc values were reported by EPA in the Health Effects Support
Document for PFOA (US EPA, 2016a) and PFOS (US EPA, 2016b). A review of measured Koc and solid
phase adsorption coefficient (Kd) values reported in literature for PFOA and PFOS was conducted with
and without associations to biosolids or land application of biosolids, for surface and subsurface soils,
aquifer materials, and settled and suspended surface water sediments (Appendix C). Based on the
results of this review, a "low-Koc" and a "high-Koc" bounding value was developed, representing the
DRAFT
D-2
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
upper and lower extremes {i.e., upper and lower whiskers of a box plot). Here, the lower extreme is
mathematically represented by the 1st Quartile (25th percentile; Ql) minus 1.5 times the inter-quartile
range (IQR = Q3 minus Ql) and the upper extreme is represented by the 3rd Quartile (75th percentile;
Q3) plus the IQR. Although the "low-Koc" was intended to be a bounding value, note that in the case of
PFOS, the "low-Koc" value of 2206.73 cm3/g is greater than the "representative-Koc" value of 371.5 cm3/g.
This maybe because of the large range in PFOS Koc reported in literature with an underlying skewed
distribution and several outliers identified in the lower end of the distribution (see e.g., Appendix C,
Figure C-26).
A total of 36 EPACMTP model simulations were performed for this sensitivity analysis.
D.2.2 VVWM
Model sensitivity was tested for environment- and chemical-specific input parameters for both PFOA
and PFOS using two source scenarios (crop and pasture) and two geographic locations to capture the
range of meteorological conditions (dry and wet) that govern modeled overland flow rates.
The sensitivity analysis for VVWM was targeted by reviewing the governing fate and transport equations
presented in the model documentation (US EPA, 2019) and ignoring those parameters associated with
irrelevant pathways for PFOA and PFOS (e.g., volatilization, degradation). The governing equations
contain four effective parameters (Equations (5) through (8)) that influence concentrations in the water
column and benthic region of the water body: the effective degradation rates of chemical mass in the
water column and the benthic region, the mass transfer coefficient describing mass transfer between
the water column and benthic region, and ratio of solute holding capacities of the two domains. As
PFOA and PFOS are known for being very stable in the natural environment, it was possible to ignore all
first order degradation rates in the formulation. Additionally, the dimensions of the index reservoir19
used in the risk assessment as the receiving water body are not subject to change and, therefore were
not examined as sensitive parameters. The flow of water through the reservoir was indirectly examined
by evaluating parameter sensitivity for dry and wet meteorological conditions.
Eliminating first order degradation rates from VVWM Equations (5) through (8) yield the following
effective parameters:
Hydroiogic Washout (Ti)
where
Q = volumetric flow rate of water out of the littoral compartment[m3/s]
v1 = volume of water in littoral compartment [m3].
As mentioned earlier, the volume of the reservoir will be constant, and flow into and out of the reservoir
will be examined indirectly by comparing simulation results for wet and dry meteorological conditions.
Therefore, there are no input parameters here to vary directly.
Effective Benthic Region Dissipation (fr)
j, BKsed 2
msed_2^sed_2 + mDOC_2^DOC_2 + V2
where
19 The index reservoir is based on the standard waterbody parameters for Variable Volume Water Model (VVWM), the
waterbody model used to estimate concentrations in surface water (US EPA, 2019; 2020); see Section A.2.3.2.
DRAFT D-3
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
B = burial rate of sediment [kg/s]
msed_2 = mass of sediment in benthic region [kg]
Ksed_2 = linear partitioning coefficient for benthic sediments [ml/g]
mDoc 2 = mass of dissolved organic carbon (DOC) in benthic compartment [kg]
Kdoc_2 = linear partitioning coefficient for DOC in benthic region [ml/g]
v2 = volume of water in benthic compartment [m3]
The mass of sediments and DOC in the benthic compartment are based on the concentration of each in
the volume of that compartment. Therefore, as the volume of the compartment is fixed, the sensitivity
of the concentration of sediments and DOC in the benthic compartment will be examined. Partitioning
coefficients for sediments and DOC are calculated using Koc and fraction of organic carbon on those
sediments and DOC. Sensitivity to Koc will be expressed separately by differences in low Koc values for
PFOAand PFOS.
Mass Transfer Coefficient (O)
D
n =
msed_2^sed_2 + mDOC_2^DOC_2 + v2
where
D = water column to benthic dispersion coefficient [m2/s]
The numerical formulation incorporates the dispersion coefficient into a mass transfer coefficient that
relates the overall dispersion through a boundary layer between the littoral and benthic compartments
having thickness Ax. This is expressed in the VVWM input parameter D_over_dx, and will therefore be
examined for sensitivity.
msed_2^sed_2 + mDOC_2^DOC_2 + v2
0 =
msed_l^sed_l + mDOC_l^DOC_l + V1
where
msed_i = mass of sediment in water column [kg]
KSed_i = linear partitioning coefficient in suspended sediments in water column [ml/g]
mDoc_i = mass of DOC in water column [kg]
Kdoc_i = linear partitioning coefficient for DOC in water column [ml/g].
Sensitivities of sediment and DOC mass and partitioning will be examined through varying the
concentration of each and the fraction of organic content on those components.
The VVWM input parameter sensitivities to be tested are listed in Table D-3. For each parameter,
bounding values were established by increasing and decreasing VVWM default values by an order of
magnitude, or nearly so in most cases. Koc bounding values for PFOA and PFOS are adopted from
Table D-2. Bounding values for the flow rate through the water body are dictated by overland runoff
generated by the hydrology module of the land application unit module using meteorologic data for the
dry and wet environments.
DRAFT D-4
-------
PFOA/PFOS Risk Assessment Appendix D. Sensitivity Analysis
Table D-3. VVWM Parameters Tested for Sensitivity
Domain
Model
Code
Parameter
Units
Scenario
Lower
bound
Default3
Upper
bound
Media
BNMAS
Areal Biomass in Benthic
Compartment
g/m2
NA
0.0006
0.006
0.06
D0C2
DOC in Benthic Region
mg/L
NA
1
5
20
D0C1
DOC in Water Column
mg/L
NA
1
5
20
FR0C2
Fraction OC on Benthic Sediments
fraction
NA
0.004
0.04
0.4
FR0C1
Fraction OC on Suspended
Sediments in Water Column
fraction
NA
0.004
0.04
0.4
D over dx
Mass Transfer Coefficient
m/s
NA
1E-10
6E-9
1E-8
PLMAS
Suspended Biomass
Concentration
mg/L
NA
0.04
0.4
4.0
SUSED
Suspended Sediment in Water
Column (TSS)
mg/L
NA
10
30
100
Meteorology
Q
Volumetric flow rateb
m3/d
Crop
41
NA
354
Pasture
17
NA
166
Chemical
Koc
Organic carbon partition
coefficient1
cm3/g
PFOA
2.9
114.8
19,953
PFOS
2,207
371.5
108,081
a These are VVWM defaults for all but Koc; representative Koc values for PFOA and PFOS are described in Appendix C.
b Lower and upper bound are based on long-term annual average values from transient simulations from land application,
c Lower and upper bound Koc values are those presented in Table D-2.
Each bounding parameter value was evaluated for both crop and pasture scenarios under dry and wet
conditions for both PFOA and PFOS. A one-at-a-time approach was used to evaluate a bounding value
for one parameter for all combinations of biosolids application scenarios (e.g., crop or pasture), and
meteorological environments (e.g., dry or wet). All other parameters are represented by VVWM defaults
or representative Koc values. Peak concentration values corresponding to the adult receptor, Surface
Water pathway, and noncancer benchmarks are used to calculate ratios of concentration corresponding
to the bounding value of a parameter to the VVWM default value for the same parameter. To evaluate
the sensitivity of flow through the surface water body, ratios of peak concentrations derived from dry
and wet meteorology are examined for each chemical and application scenario.
D.3 Results and Discussion
As described earlier, model sensitivity is evaluated by comparing the ratio of initial PFOA and PFOS
source concentrations in biosolids leachate to predicted PFOA and PFOS concentrations at an
observation well located in the center of the buffer for the lower and upper bound parameter input
values for each parameter. The ratio of initial PFOA and PFOS source concentrations in biosolids
leachate to predicted PFOA and PFOS concentrations at an observation well is directly proportional to
the predicted risk outputs and is therefore an appropriate metric to understand model sensitivity.
Figures D-l through D-13 present the model sensitivity results for each parameter comparing the three
source models or scenarios, chemical (PFOA and PFOS) at two locations, Boulder and Charleston, with
bounding meteorological conditions of dry and wet.
D.3.1 EPACMTP
D.3.1.1 Sensitivity to Unsaturated Zone Parameters
The saturated hydraulic conductivity of the unsaturated zone soil is a measure of the soil's ability to
transmit water under fully saturated conditions. It is used as an input to the unsaturated zone flow
module in EPACMTP and is used to calculate the moisture content in the soil under a given rate of
leachate infiltration. The difference between the ratio of initial PFOA and PFOS source concentrations in
DRAFT
D-5
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
biosolids leachate to predicted PFOA and PFOS concentrations at the observation well at the lower and
upper bound values of this input parameter is less than a factor of 1.7 across PFOA and PFOS for all
three scenarios and two locations (Figure D-l). Therefore, this input parameter is not considered to be
sensitive to the predicted model risk outputs.
1.0E+05
£ m 1.0E+04
1.0E+01
1.0E+00
¦ US01, 10 PERC
¦ US01, 90 PERC
$ in
25.0
20.0
5.0
0.0
I US01, 10 PERC
I US01, 90 PERC
CROP PASTURE SI CROP PASTURE SI
PFOA PFOS
¦¦.. JH li ii
CROP PASTURE SI CROP PASTURE SI
PFOA PFOS
Figure D-l. Sensitivity to saturated hydraulic conductivity of the unsaturated zone for Boulder (dry
climate, left) and Charleston, SC (wet climate, right).
The Van Genuchten shape parameters, alpha and beta, of the unsaturated zone are soil-specific shape
parameters that are obtained from an empirical relationship between pressure head and volumetric
water content; are one of the parameters in the van Genuchten (1980) model used for modeling soil-
water content as a function of pressure head and are used to calculate the moisture content in the soil
under a given rate of leachate infiltration. The van Genuchten parameters alpha and beta are inputs to
the unsaturated zone flow module and are used to calculate the moisture content in the soil under a
given rate of leachate infiltration. The difference between the ratio of initial PFOA and PFOS source
concentrations in biosolids leachate to predicted PFOA and PFOS concentrations at the observation well
at the lower and upper bound values of the alpha input parameter is less than a factor of 1.6 across
PFOA and PFOS for all three scenarios and two locations (Figure D-2). The difference between the ratio
of initial PFOA and PFOS source concentrations in biosolids leachate to predicted PFOA and PFOS
concentrations at the observation well at the lower and upper bound values of the beta input parameter
is less than a factor of 1.3 across PFOA and PFOS for all three scenarios and two locations (Figure D-3).
Therefore, these input parameters are not considered to be sensitive to the predicted model risk
outputs.
S lti 1.0E+04
T3 ||
^
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
1.0E+05
$ in 1.0E+04
IUS03, 10 PERC
IUS03, 90 PERC
1.0E+00
a; £
$ m
20.0
15.0
10.0
0.0
I US03, 10 PERC
I US03, 90 PERC
ll
¦¦ ¦¦
IIII .1
CROP PASTURE SI CROP PASTURE SI
PFOA PFOS
CROP PASTURE SI CROP PASTURE SI
PFOA PFOS
Figure D-3. Sensitivity to van Genuchten parameter beta for Boulder (dry climate, left) and
Charleston, SC (wet climate, right).
The residual water content is the moisture content of the soil below which a reduction in the pressure
head does not result in the loss of moisture. It is an input to the unsaturated zone flow module and is
used to calculate the moisture content in the soil under a given rate of leachate infiltration. The
difference between the ratio of initial PFOA and PFOS source concentrations in biosolids leachate to
predicted PFOA and PFOS concentrations at the observation well at the lower and upper bound values
of this input parameter is less than a factor of 1.1 across PFOA and PFOS for all three scenarios and two
locations (Figure D-4). Therefore, this input parameter is not considered to be sensitive to the predicted
model risk outputs.
£ £
QJ £
l/l TO
20.0
15.0
I US04, 10 PERC
I US04, 90 PERC
CROP PASTURE SI CROP PASTURE SI
PFOA PFOS
¦¦ .. II II II
CROP PASTURE SI CROP PASTURE SI
PFOA PFOS
Figure D-4. Sensitivity to residual water content of the unsaturated zone for Boulder (dry climate,
left) and Charleston, SC (wet climate, right).
The saturated water content represents the maximum fraction of the total volume of soil that is
occupied by the water contained in the soil at atmospheric pressure. The difference between the ratio
of initial PFOA and PFOS source concentrations in biosolids leachate to predicted PFOA and PFOS
concentrations at the observation well at the lower and upper bound values of this input parameter is
less than a factor of 1.3 across PFOA and PFOS for all three scenarios and two locations (Figure D-5).
Therefore, this input parameter is not considered to be sensitive to the predicted model risk outputs.
DRAFT
D-7
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
1.0E+05
oi E
5 1.0E+04
£ ^
1.0E+03
1.0E+00
I US05, 10 PERC
I US05, 90 PERC
^ £
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
1.0E+04
1.0E+03
1.0E+02
1.0E+01
1.0E+00
IUS08, 10 PERC
I US08, 90 PERC
$ m
£ H
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
l US08, 10 PERC
I US08, 90 PERC
II II
CROP PASTURE SI CROP PASTURE SI
PFOA PFOS
CROP PASTURE SI
PFOA
ll II
CROP PASTURE SI
PFOS
Figure D-7. Sensitivity to percent organic matter of the unsaturated zone for Boulder (dry climate,
left) and Charleston, SC (wet climate, right).
The dry bulk density of the soil is the ratio of the mass of the solid soil to its total volume. The dry soil
bulk density (mass of soil per unit volume) is used to calculate the retardation coefficient of organic
constituents and to convert soil mass to volume. The difference between the ratio of initial PFOA and
PFOS source concentrations in biosolids leachate to predicted PFOA and PFOS concentrations at the
observation well at the lower and upper bound values of this input parameter is less than a factor of 1.2
across PFOA and PFOS for all three scenarios and two locations (Figure D-8). Therefore, this input
parameter is not considered to be sensitive to the predicted model risk outputs.
cu E
£ un
£ ^
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
QJ E
5 tn
£ VJ
C
c o
1.0E+03
1.0E+00
IAS02, MIN
IAS02, MAX
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
PFOS for all three scenarios and two locations (Figure D-ll). This behavior is more pronounced for SI
results due to the shallower penetration of dissolved contaminant into the aquifer from a source area
that is more than 50 times smaller than the agricultural field - there is more attenuation at shallower
depths that tends to increase the concentration ratios. Therefore, this input parameter is sensitive to
the predicted model risk outputs.
Boulder
a> E
5 m
CO CT5
1.0E+04
1.0E+01
IAS04, 10 PERC
IAS04,90 PERC
CROP PASTURE SI CROP PASTURE SI
PFOA PFOS
1.0E+07
a;
E
1.0E+06
5
LD
¦O
II
a>
o
1.0E+05
a>
c
V)
(TJ
-Q
o
to
1.0E+04
"O
in
'E
c
o
1.0E+03
E
2
=>
c
1.0E+02
£
a;
<_>
X
c
ro
o
1.0E+01
2
<_>
1.0E+00
Charleston
¦ AS04, 10 PERC
¦ AS04, 90 PERC
1
II ..
L
CROP PASTURE SI
PFOA
CROP PASTURE SI
PFOS
Charleston (Crop and Pasture Detail)
Results could not be computed for PFOA/PFOS in Charleston
at the 10th percentile thickness of saturated zone (4.3 m)
because the simulated water table was above ground surface
resulting in a mounding violation within the model. Therefore,
the 50th percentile (14.3 m) was used as the lower bound
instead
5 LO
-o II
gj a,
2 u
v =
l/) TO
II
ro to
+- c
E .2
tl
=5 C
F ^
c o
IAS04, 10 PERC
IAS04, 90 PERC
II ii
II
CROP PASTURE CROP PASTURE
PFOA PFOS
Figure D-ll. Sensitivity to saturated zone thickness for Boulder (dry climate, left) and Charleston, SC
(wet climate, right; top row on right shows LAU crop and pasture and SI, lower right
shows detail for crop and pasture on a smaller y-axis scale).
Hydraulic conductivity is a measure of the ability to transmit water under a unit hydraulic gradient. The
aquifer hydraulic conductivity is an input to the saturated zone flow module. The hydraulic conductivity,
together with the hydraulic gradient, controls the ground-water flow rate. The difference between the
ratio of initial PFOA and PFOS source concentrations in biosolids leachate to predicted PFOA and PFOS
concentrations at the observation well at the lower and upper bound values of this input parameter is a
factor of several orders of magnitude across PFOA and PFOS, particularly for SI scenarios at both
geographic locations (Figure D-12). Therefore, this input parameter is sensitive to the predicted model
risk outputs.
DRAFT
D-ll
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
c
E
.E (_>
1.0E+18
1.0E+16
1.0E+14
1.0E+12
1.0E+10
1.0E+08
1.0E+06
1.0E+04
1.0E+02
1.0E+00
¦ m PFRf
¦ AS05, 90 PERC
¦
¦
¦
¦
1
1
1
1
5 m
to 03
-s «
1.0E+18
1.0E+16
1.0E+14
1.0E+12
1.0E+10
1.0E+08
1.0E+06
1.0E+04
1.0E+02
1.0E+00
¦ AS05, 10 PERC
¦ AS05, 90 PERC
¦
¦ ¦ -
-1 _¦ ¦
CROP PASTURE
PFOA
CROP PASTURE
PFOS
CROP PASTURE SI
PFOA
CROP PASTURE SI
PFOS
aJ E
5 ^
to ro
-£ to
E £
1,000
100
10
IAS05, 10 PERC
IAS05, 90 PERC
I
ll
CROP PASTURE CROP PASTURE
PFOA PFOS
Figure D-12. Sensitivity to hydraulic conductivity of the saturated zone for Boulder (dry climate, left)
and Charleston, SC (wet climate, right; top shows LAU crop and pasture and SI, lower
shows detail for crop and pasture on a smaller y-axis scale).
Hydraulic gradient measures the head difference between two points as a function of their distance. For
an unconfined aquifer such as that modeled with EPACMTP, the hydraulic gradient is simply the slope of
the water table in a particular direction. It is calculated as the difference in the elevation of the water
table measured at two locations divided by the distance between the two locations. The difference
between the ratio of initial PFOA and PFOS source concentrations in biosolids leachate to predicted
PFOA and PFOS concentrations at the observation well at the lower and upper bound values of this
input parameter is a factor of several orders of magnitude across PFOA and PFOS, particularly for SI
scenarios at both geographic locations (Figure D-13). Therefore, this input parameter is sensitive to the
predicted model risk outputs.
DRAFT
D-12
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
aj E
5 m
1.0E+11
1.0E+10
1.0E+09
1.0E+08
1.0E+07
1.0E+06
1.0E+05
1.0E+04
1.0E+03
1.0E+02
1.0E+01
1.0E+00
IAS07, 10PERC
IAS07, 90 PERC
aj E
5 t-o
1.0E+07
1.0E+06
1.0E+05
1.0E+04
1.0E+03
1.0E+02
1.0E+01
1.0E+00
CROP PASTURE
PFOA
CROP PASTURE SI
PFOS
¦ AS07, 10 PERC
¦ Miu/, 3U rtrci,
m
11 .. 1
CROP PASTURE SI CROP PASTURE SI
PFOA PFOS
Figure D-13. Sensitivity to regional hydraulic gradient of the saturated zone for Boulder (dry climate,
left) and Charleston, SC (wet climate, right).
D.3.1.3 Sensitivity to Chemical-specific Parameter
The organic carbon partition coefficient (cm3/g) is the ratio of a constituent's concentration in a
theoretical soil containing only organic carbon to its concentration in the ground water. Thus, koc
describes the affinity of a constituent to attach itself to organic carbon. This parameter is applicable to
organic constituents which tend to sorb onto the organic matter in soil or in an aquifer.
The figures in this section present a comparison of the ratio of initial PFOA and PFOS source
concentrations in biosolids leachate to predicted PFOA and PFOS concentrations at an observation well
located in the center of the buffer for crop, pasture and surface impoundment scenarios at three
locations: Boulder, Chicago, and Charleston representing a dry, moderate and wet meteorology. All
sensitivity simulations were performed assuming both the "low-Koc", and "high-Koc" input parameters.
At all three locations, the ratio of initial PFOA and PFOS source concentrations in biosolids leachate to
predicted PFOA and PFOS concentrations at an observation well located in the center of the buffer
increases with an increase in the assumed Koc value. As noted earlier (Section D.2.1), for PFOS, the "low-
Koc" value of 2206.73 cm3/g is greater than the "representative-Koc" value of 371.5 cm3/g- These results
are expected as constituents with high Koc values tend to move more slowly through the soil and
ground water. The effect of equilibrium sorption is expressed in EPACMTP through the retardation
coefficient, R, which is a function of the chemical-specific organic carbon partition coefficient, koc:
pb
R = 1 +
0
where,
R
Pb
0
kd
foe
koc
retardation coefficient
bulk density of unsaturated soil (g/cm3)
soil water content (dimensionless)
soil-water partition coefficient (L/kg) = foc x koc
fractional organic carbon content in the soil or aquifer (unitless)
organic carbon partition coefficient (cm3/g)
For a similar geographic location, representative scenario and Koc value tested, the ratio of initial PFOS
source concentrations in biosolids leachate to predicted PFOS concentrations at an observation well
located in the center of the buffer was observed to be greater than that for PFOA. For example,
DRAFT
D-13
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
Figure D-14 demonstrates this observation at Chicago for Crop and Pasture scenarios under low and
high assumed Koc. This is consistent with a higher Koc value for PFOS in comparison to PFOA, all
environmental parameters being the same. A higher Koc value for PFOS results in greater retardation and
thus lower concentrations in the observation well located in the center of the buffer.
Crop
Chicago, Crop Low Koc
Pasture
Chicago, Pasture Low Koc
O
o
5
o
o
0
¦C
O)
1
Figure D-14. Maximum leachate to well concentration ratio for land application unit for Chicago
(moderate climate): crop (left) and pasture (right), low Koc (top) and high Koc (bottom).
The ratio of initial PFOA and PFOS source concentrations in biosolids leachate to predicted PFOA and
PFOS concentrations at an observation well located in the center of the buffer was simulated to be
different between geographic locations and corresponding meteorological conditions. Particularly, for
the same representative scenario, chemical and Koc value tested, higher PFOA and PFOS concentration
ratios were observed at dry (Boulder) vs. wet (Charleston) conditions (Figure D-15). This suggests that
lower PFOA and PFOS concentrations are observed at the well located in the center of the buffer under
drier conditions likely due to lower mass transport on account of lesser infiltration.
DRAFT
D-14
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
Crop
o
o
5
o
o
0
¦C
g)
1
!
I!
I t 124
l.E+22
l.E+16
l.E+14
l.E+12
11E"°
8 ir+ro
£ l.tlUb
1.EI02
1.E+00
Figure D-15.
Crop High Koc
I .
1£+12
1£+11
ir+?r.
1±I24
l£+22
| £
| *5 1£+12
Pasture
Pasture Low Koc
¦ PFOA
¦ PFOS
Pasture High Koc
I
¦ PFOA
¦ PFOS
Maximum leachate to well concentration ratio for land application unit by climate: crop
(left) and pasture (right), low Koc (top) and high Koc (bottom).
The sensitivity of the ratio of initial PFOA and PFOS source concentrations in biosolids leachate to
predicted PFOA and PFOS concentrations at an observation well located in the center of the buffer was
also simulated for the surface impoundment scenario at all three geographic locations assuming a low-
and high-Koc value (Figure D-16). Based on these simulations, higher concentration ratios are observed
for a higher Koc value and more impermeable liner type (e.g., composite liner vs. clay liner), consistent
with an expected lower mass transport.
DRAFT
D-15
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
II
;
l.E+30
l.E+27
l.E+24
l.E+21
l.E+18
l.E+15
T l.E+12
% l.E+09
« l.E+06
~ l.E+03
1.E+00
II
l.E+30
l.E+27
l.E+24
l.E+21
l.E+18
l.E+15
l.E+12
| 1.E.09
« l.E+06
~ l.E+03
1.E+00
^ l.E+30
o
= 5 l.E+27
it £ wT- l.E+24
1 J= c 1E+21
£ -E £ l.E+18
o "o a.
^ 2 £ l.E+15
for l.E+12
e" ! | l.E+09
| 2 J l.E+06
jf § l.E+03
Low Koc
Boulder Low Koc
¦ No Liner
I Clay Liner
I Composite Liner
Chicago Low Koc
¦ No Liner
I Clay Liner
I Composite Liner
Charleston Low Koc
¦ No Liner
l Clay Liner
I Composite Liner
l.E+03
1.E+00
^ l.E+30
o
= £ l.E+27
j?. S -j7T l.E+24
> j= c l.E+21
£ ~ o
£ £ | l.E+18
o -a
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
D.3.2 VVWM
Each bounding parameter value in Table D-3 was evaluated for both crop and pasture scenarios under
dry and wet conditions for both PFOA and PFOS. A one-at-a-time approach was used to evaluate a
bounding value for one parameter for all combinations of biosolids application scenarios (e.g., crop or
pasture), and meteorological environments (e.g., dry or wet). All other parameters are represented by
VVWM defaults or representative Koc values. Peak concentration values corresponding to the adult
receptor, surface water pathway, and noncancer benchmarks are used to calculate ratios of
concentration corresponding to the bounding value of a parameter to the VVWM default value for the
same parameter. To evaluate the sensitivity of flow through the surface water body, ratios of peak
concentrations derived from dry and wet meteorology are examined for each chemical and application
scenario. Chemical-related sensitivity was examined by computing ratios of peak concentration based
on bounding values of Koc case relative to default Koc values for all combinations of dry and wet
meteorology and application scenario.
D. 3.2.1 Sensitivity to Media Parameters
Figure D-17 and Figure D-18 present sensitivity results of the ratio of predicted peak, dissolved PFOA
and PFOS surface water concentrations, respectively, using bounding parameter values to the same
surface water concentrations simulated using default parameter values for Crop, Pasture scenarios in (a)
Boulder [dry] and (b) Charleston [wet]. Ratios that are approximately 1.0 indicate that there was little or
no difference in peak concentrations for result using a bounding value and baseline value. Ratios less
than 1.0 indicate that the peak concentration simulated using the bounding parameter value was less
than peak surface water concentration predicted using the baseline value for the same parameter.
Likewise, rations greater than 1.0 indicate that the simulated peak concentration using the bounding
value is greater than the corresponding simulation results for the baseline value for the same
parameter. The only parameter showing any significant sensitivity is the fraction of organic carbon of
the benthic sediments. Lower levels of organic carbon result in less sorption of a chemical to the
sediments and higher dissolved concentrations and the converse for high FOC content. The difference in
behaviors exhibited by the crop and pasture scenarios reflect the impact of application practices: tilling
binds more mass to soil reducing concentrations in runoff and reducing the partitioning from eroded
sediments making it to the reservoir where FOC is high. When FOC is low, more dissolved mass is moved
off the field in runoff and released from solids reaching the reservoir.
PFOA, Dry
PFOA, Wet
8 ° E
i— Qfl -3 0.8
i c | o.6
1 ! I 0.4
¦ Crop
¦ Pasture
¦ Crop
¦ Pasture
High Low
High Low
High Low High Low
High Low
High Low
High Low
High Low
•B 3
High Low
High Low
High Low
High Low
High Low
High Low
High Low
High Low
Areal
DOC
DOC FOC
FOC
Mass
Susp.
TSSin
Areal
DOC
DOC
FOC
FOC
Mass
Susp.
TSSin
Biomass
Benthic
Water Benthic
Water
Transfer
Biomass
Water
Biomass
Benthic
Water
Benthic
Water
Transfer
Biomass
Water
Cone
Reg.
Column Sediment
Column
Column
Cone
Reg.
Column
Sediment
Column
Column
Figure D-17. Sensitivity of VVWM to media parameters for PFOA for Boulder (dry climate, left) and
Charleston, SC (wet climate, right).
DRAFT
D-17
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
PFOS, Dry
PFOS, Wet
¦ Crop
¦ Pasture
¦ Crop
¦ Pasture
High Low High Low
High Low
High Low High Low
High Low
High Low High Low
•Jj
—
High Low
High Low
High Low
High Low
High Low
High Low
High Low
High Low
Areal DOC
DOC
FOC FOC
Mass
Susp. TSS in
Areal
DOC
DOC
FOC
FOC
Mass
Susp.
TSS in
Biomass Benthic
Water
Benthic Water
Transfer
Biomass Water
Biomass
Benthic
Water
Benthic
Water
Transfer
Biomass
Water
| Cone | Reg.
Column
Sediments Column
Column
Cone
Reg.
Column
Sediments
Column
Column
Figure D-18. Sensitivity of VVWM to media parameters for PFOS for Boulder (dry climate, left) and
Charleston, SC (wet climate, right).
D. 3.2.2 Sensitivity to Meteorology Parameters
Figure D-19 presents sensitivity of surface water concentrations to meteorological data as ratios of
resulting concentrations from overland flow into and through the water body for dry conditions over
wet conditions. The amount of mass available for a given scenario is fixed. Adding more precipitation
increases the dilution of dissolved chemical in runoff, decreasing the concentration of the chemical
entering the water body. As a result, all ratios are greater than 1.0.
Figure D-19. Sensitivity of VVWM to meteorology for PFOA and PFOS for crop and pasture: ratio of
peak surface water concentrations for dry climate to wet climate.
D. 3.2.3 Sensitivity to Chemical Parameters
Figure D-20 presents sensitivity of surface water concentrations to low and high Koc values. Here,
VVWM was run in isolation using a fixed loading and varying only Koc. For PFOA and PFOS, VVWM is
entirely insensitive to climate data; the results shown here are for Charleston (wet), but those for
Boulder (dry) are identical. There is no distinction between crop and field here because that affects only
the loading, not what happens within the surface water body. As noted earlier (Section D.2.1), for PFOS,
the "low-Koc" value of 2206.73 cm3/g is greater than the "representative-Koc" value of 371.5 cm3/g-
As expected, as Koc increases, the fraction sorbed to bed sediment increases, increasing the
concentration in sediments and decreasing the concentration in the water column.
DRAFT
D-18
-------
PFOA/PFOS Risk Assessment
Appendix D. Sensitivity Analysis
PFOA: Water Column
PFOA: Sediment
PFOS: Water Column
PFOS: Sediment
II
High Low
High Low
1 0
High Low
^ High Low
High Low
High Low
Total Water
Dissolved
Sediment
Total Water
Dissolved
Sediment
Figure D-20. Sensitivity of VVWM to Koc for PFOA (left) and PFOS (right).
D.3.2.4 Summary of Most Sensitive Parameters for VVWM
The parameters to which VVWM is most sensitive are as follows:
• Fraction of organic carbon of the benthic sediments
• Climate
• Organic carbon partition coefficient (Koc).
D.4 References
Reilly, T. E., & Harbaugh, A. W. (2004). Guidelines for evaluating ground-water flow models. Scientific
Investigations Report 2004-5038. https://doi.org/10.3133/sir20045038.
US EPA (United States Environmental Protection Agency). (2003). EPA's Composite Model for Leachate
Migration with Transformation Products (EPACMTP): Parameters/Data Background Document.
Office of Solid Waste, Washington, DC. April.
US EPA (Environmental Protection Agency). (2016a). Health Effects Support Document for
Perfluorooctanoic Acid (PFOA). EPA 822-R-16-003. Office of Water: Washington DC.
US EPA (Environmental Protection Agency). (2016b). Health Effects Support Document for
Perfluorooctane Sulfonate (PFOS). EPA 822-R-16-002. Office of Water: Washington DC.
US EPA (United States Environmental Protection Agency). (2019). The Variable Volume Water Model,
Revision B, USEPA/OPP 734S16002. Office of Pesticide Programs, Washington, DC.
https://www.epa.gov/pesticide-science-and-assessing-pesticide-risks/models-pesticide-risk-
assessment
van Genuchten, M. Th. (1980). A closed-form equation for predicting the hydraulic conductivity of
unsaturated soils. SoilSci. Soc. J., 44, 892-898.
Wolff, R. G. (1982). Physical Properties of Rocks—Porosity, Permeability, Distribution Coefficients, and
Dispersivity. USGS Water Resources Investigation Open File Report 82-166.
DRAFT
D-19
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
APPENDIX E. SCREENING-LEVEL RESULTS FROM BST
E.l BST Screening Inputs
Table E-l. Scenario Inputs
Scenario
Model Code
Model Value
Units
Description
Reference
Crop
OpLife
40
[yrsl
Number of year of biosolids applications to field
Biosolids 2003 (US EPA 2003)
Crop
Nappl
1
[1/vrl
Number of biosolids applications per year
Biosolids 2003 (US EPA 2003)
Crop
Rappl
0.0025
[MTwet/m2-yrl
Biosolids application rate (wet weiqht)
Calculated
Crop
zruf
1
(cml
Rouqhness heiqht (field)
TSDF Fuqit. Air (US EPA, 1989b)
Crop
Ztillinq
0.2
[ml
Tillinq depth
Biosolids 2003 (US EPA 2003)
Pasture
OpLife
40
[yrsl
Number of year of biosolids applications to field
Biosolids 2003 (US EPA 2003)
Pasture
Nappl
1
[1/vrl
Number of biosolids applications per year
Biosolids 2003 (US EPA 2003)
Pasture
Rappl
0.0025
[MTwet/m2-yrl
Biosolids application rate (wet weiqht)
Calculated
Pasture
zruf
3.7
[cml
Rouqhness heiqht (field)
TSDF Fuqit. Air (US EPA, 1989b)
Pasture
Ztillinq
0.02
[ml
Tillinq depth
Biosolids 2003 (US EPA 2003)
Reclamation
OpLife
1
[yrsl
Number of year of biosolids applications to field
Biosolids 2003 (US EPA 2003)
Reclamation
Nappl
1
[1/yrl
Number of biosolids applications per year
Biosolids 2003 (US EPA 2003)
Reclamation
Rappl
0.0125
[MTwet/m2-yrl
Biosolids application rate (wet weiqht)
Calculated
Reclamation
zruf
3.7
[cml
Rouqhness heiqht (field)
TSDF Fuqit. Air (US EPA, 1989b)
Reclamation
Ztillinq
0.02
[ml
Tillinq depth
Biosolids 2003 (US EPA 2003)
Table E-2. Fate Inputs
Model Code
Moderate
Value
Dry
Value
Wet
Value
Units
Description
Reference
Comment
%solids
40
40
40
[mass %1
Percent solids in land applied biosolids
Biosolids 2003 (US EPA 2003)
AirTemp
9.69
10.11
18.18
[CI
Averaqe air temperature
SAMSON (US DOC & DOE, 1993)
determined by met station
Area reserv
52555
52555
52555
[m21
Area (index reservoir)
WWM
asdm
0.5
0.5
0.5
fmml
Mode of the aqqreqate size distribution
TSDF Fuqit. Air (US EPA, 1989b)
Bdwaste
0.7
0.7
0.7
[q DW/cm31
Dry bulk density (biosolids)
Gunn etal. (2004)
bsp
0.6
0.6
0.6
[fraction]
Porosity (bed sediment)
MPE/IEM (US EPA, 1998)
C
0.1
0.1
0.1
[fraction]
USLE cover manaqement factor
HHRAP (US EPA, 2005a)
db
0.05
0.05
0.05
[ml
Depth of upper benthic layer
MPE/IEM (US EPA, 1998)
chanqed for WWM
DTR
12
12
12
[m2/m31
Drainaqe area to capacity ratio (watershed)
SAB (Index Res)
dwc pond
2
2
2
[ml
Water column depth (farm pond)
WWM
dwc reservoir
2.74
2.74
2.74
[ml
Water column depth (index reservoir)
WWM
foe bedsed
0.04
0.04
0.04
[fractionl
Fraction orqanic carbon (bed sediments)
WWM
foe biosolids
0.4
0.4
0.4
[fractionl
Fraction orqanic carbon (biosolids)
Biosolids 2003 (US EPA 2003)
foe sw
0.04
0.04
0.04
[fractionl
Fraction orqanic carbon (suspended sediments)
WWM
P
1
1
1
[fractionl
USLE supportinq practice factor (watershed)
Wanielista & Yousef, 1993
DRAFT
E-l
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
Model_Code
Moderate
Value
Dry
Value
Wet
Value
Units
Description
Reference
Comment
PI field
0
0
0
(%1
Percent impervious (field)
CWP, 1998
R
155
50
360
(1/yrl
USLE rainfall/erosivity factor
Wischmeier& Smith, 1978
determined by met station
SiteLatitude
41.983
40.0167
32.9
[degrees]
Site latitude
SAMSON (US DOC & DOE, 1993)
determined by met station
Sw
10
10
10
[mass %1
Silt content of biosolids
AP-42 (US EPA, 1995)
Theta water
1.024
1.024
1.024
[empiricall
Temperature correction factor
Chapra, 1996
TwaterOI
270
273
284
[deg K1
Waterbody temperature (January)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
Twater02
267
271
282
[deg K1
Waterbody temperature (February)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
Twater03
270
274
283
[deg K1
Waterbody temperature (March)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
Twater04
276
277
287
[deg K[
Waterbody temperature (April)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
Twater05
282
282
291
[deg K1
Waterbody temperature (May)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
Twater06
289
287
295
[deg K[
Waterbody temperature (June)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
Twater07
294
293
299
[deg K[
Waterbody temperature (July)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
Twater08
297
296
300
[deg K[
Waterbody temperature (August)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
Twater09
295
295
299
[deg K1
Waterbody temperature (September)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
TwaterlO
291
290
297
[deg K1
Waterbody temperature (October)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
Twaterl 1
285
284
292
[deg K1
Waterbody temperature (November)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
Twater12
278
277
288
[deg K1
Waterbody temperature (December)
Water Enc. (van der Leeden et al., 1990)
depends on HUC Region
uw
4.632
3.783
3.788
[m/secl
Mean annual wind speed
SAMSON (US DOC & DOE, 1993)
determined by met station
zavgjower
0.2
0.2
0.2
[ml
Lower averaging depth for soil concentration
Biosolids 2003 (US EPA 2003)
Zmix tilled
20
20
20
[cml
Mixing depth of tilled soil (field)
Biosolids 2003 (US EPA 2003)
Zmix untilled
2
2
2
[cml
Mixing depth of untilled soil (field)
Biosolids 2003 (US EPA 2003)
Zmodeled
0.2
0.2
0.2
[ml
Depth of modeled soil column
Biosolids 2003 (US EPA 2003)
Table E-3. Exposure Inputs
Scenario
Receptor
Model Code
Model Value
Units
Description
Reference
Crop
Farmer
BW child1-5
15
[kgl
Body weight (child aged 1-5)
EFH:2011 (US EPA, 2011)
Crop
Farmer
BW child6-11
29
[kgl
Body weight (child aged 6-11)
EFH:2011 (US EPA, 2011)
Crop
Farmer
BW child12-19
61
[kgl
Body weight (child aged 12-19)
EFH:2011 (US EPA, 2011)
Crop
Farmer
BW adult
79
[kgl
Body weight (adult)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR dw child1-5
44
[m L/kg-dayl
Consumption rate, water (child aged 1 -5)
EFH:2019-dw
Crop
Farmer
CR dw child6-11
31
[m L/kg-dayl
Consumption rate, water (child aged 6-11)
EFH:2019-dw
Crop
Farmer
CR dw childl 2-19
25
[mL/kg-dayl
Consumption rate, water (child aged 12-19)
EFH:2019-dw
Crop
Farmer
CR dw adult
34
[mL/kg-dayl
Consumption rate, water (adult)
EFH:2019-dw
Crop
Farmer
CR exfruit childl-5
5.4
[g WW/kg BW/dayl
Consumption rate, exposed fruit (child aged 1-5)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR exfruit child6-11
7
[g WW/kg BW/dayl
Consumption rate, exposed fruit (child aged 6-11)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR exfruit childl2-19
3.4
[g WW/kg BW/dayl
Consumption rate, exposed fruit (child aged 12-19)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR exfruit adult
5
[g WW/kg BW/dayl
Consumption rate, exposed fruit (adult)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR_exveg_child1-5
6.4
[g WW/kg BW/dayl
Consumption rate, exposed vegetables (child aged 1-5)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR_exveg_child6-11
3.2
[g WW/kg BW/dayl
Consumption rate, exposed vegetables (child aged 6-11)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR_exveg_child12-19
2.4
[g WW/kg BW/dayl
Consumption rate, exposed vegetables (child aged 12-19)
EFH:2011 (US EPA, 2011)
DRAFT
E-2
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
Scenario
Receptor
Model Code
Model Value
Units
Description
Reference
Crop
Farmer
CR_exveg_adult
6
lg WW/kg BW/dayl
Consumpt
on rate, exposed vegetables (adult)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR fish child1-5
5.2
[g WW/dayl
Consumpt
on rate, fish (child aged 1-5)
NHANES:2014-fish
Crop
Farmer
CR fish child6-11
7.7
[g WW/dayl
Consumpt
on rate, fish (child aged 6-11)
NHANES:2014-fish
Crop
Farmer
CR fish childl2-19
9.6
[g WW/dayl
Consumpt
on rate, fish (child aged 12-19)
NHANES:2014-fish
Crop
Farmer
CR fish adult
22
[g WW/dayl
Consumpt
on rate, fish (adult)
NHANES:2014-fish
Crop
Farmer
CR profruit childl-5
16
[g WW/kg BW/dayl
Consumpt
on rate, protected fruit (child aged 1-5)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR profruit child6-11
16
[g WW/kg BW/dayl
Consumpt
on rate, protected fruit (child aged 6-11)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR profruit childl2-19
7.4
[g WW/kg BW/dayl
Consumpt
on rate, protected fruit (child aged 12-19)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR_profruit_adult
14
lg WW/kg BW/dayl
Consumpt
on rate, protected fruit (adult)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR_proveg_child1-5
3.1
lg WW/kg BW/dayl
Consumpt
on rate, protected vegetables (child aged 1-5)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR_proveg_child6-11
2.1
lg WW/kg BW/dayl
Consumpt
on rate, protected vegetables (child aged 6-11)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR_proveg_child12-19
1.9
lg WW/kg BW/dayl
Consumpt
on rate, protected vegetables (child aged 12-19)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR_proveg_adult
3.6
lg WW/kg BW/dayl
Consumpt
on rate, protected vegetables (adult)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR root childl-5
5.7
lg WW/kg BW/dayl
Consumpt
on rate, root vegetables (child aged 1 -5)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR root child6-11
3.8
lg WW/kg BW/dayl
Consumpt
on rate, root vegetables (child aged 6-11)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR root child12-19
2.3
lg WW/kg BW/dayl
Consumpt
on rate, root vegetables (child aged 12-19)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CR root adult
3.1
lg WW/kg BW/dayl
Consumpt
on rate, root vegetables (adult)
EFH:2011 (US EPA, 2011)
Crop
Farmer
CRs childl-5
40
Img/dayl
Consumpt
on rate, soil (child aged 1-5)
EFH:2017-soil
Crop
Farmer
CRs child6-11
30
Img/dayl
Consumpt
on rate, soil (child aged 6-11)
EFH:2017-soil
Crop
Farmer
CRs childl 2-19
10
Img/dayl
Consumpt
on rate, soil (child aged 12-19)
EFH:2017-soil
Crop
Farmer
CRs adult
10
Img/dayl
Consumpt
on rate, soil (adult)
EFH:2017-soil
Crop
Farmer
ED childl-5
13
lyrsl
Exposure duration (child aged 1-5)
EFH:2011 (US EPA, 2011)
Crop
Farmer
ED child6-11
13
lyrsl
Exposure duration (child aged 6-11)
EFH:2011 (US EPA, 2011)
Crop
Farmer
ED childl 2-19
13
lyrsl
Exposure duration (child aged 12-19)
EFH:2011 (US EPA, 2011)
Crop
Farmer
ED adult
48
lyrsl
Exposure duration (adult)
EFH:2011 (US EPA, 2011)
Crop
Farmer
ShowerTime
15
Iminl
Time in shower stall during shower
EFH:2011 (US EPA, 2011)
Crop
Farmer
T bathroom
5
fminl
Time spent in bathroom, not in shower
EFH:2011 (US EPA, 2011)
Pasture
Farmer
BW childl-5
15
Ikgl
Body weight (child aged 1-5)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
BW child6-11
29
Ikgl
Body weight (child aged 6-11)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
BW child12-19
61
Ikgl
Body weight (child aged 12-19)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
BW adult
79
Ikgl
Body weight (adult)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
CR beef childl-5
11
lg WW/kg BW/dayl
Consumption rate, beef (child aged 1-5)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
CR beef child6-11
11
lg WW/kg BW/dayl
Consumption rate, beef (child aged 6-11)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
CR beef childl 2-19
3.5
lg WW/kg BW/dayl
Consumption rate, beef (child aged 12-19)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
CR beef adult
5.4
lg WW/kg BW/dayl
Consumption rate, beef (adult)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
CR dw childl-5
44
ImL/kg-dayl
Consumption rate, water (child aged 1 -5)
EFH:2019-dw
Pasture
Farmer
CR dw child6-11
31
ImL/kg-dayl
Consumption rate, water (child aged 6-11)
EFH:2019-dw
Pasture
Farmer
CR dw childl 2-19
25
ImL/kg-dayl
Consumption rate, water (child aged 12-19)
EFH:2019-dw
Pasture
Farmer
CR dw adult
34
ImL/kg-dayl
Consumption rate, water (adult)
EFH:2019-dw
Pasture
Farmer
CR milk childl-5
59
lg WW/kg BW/dayl
Consumption rate, milk (child aged 1-5)
EFH:2018-meatdairy
Pasture
Farmer
CR milk child6-11
26
lg WW/kg BW/dayl
Consumption rate, milk (child aged 6-11)
EFH:2018-meatdairy
Pasture
Farmer
CR milk childl 2-19
12
lg WW/kg BW/dayl
Consumption rate, milk (child aged 12-19)
EFH:2018-meatdairy
DRAFT
E-3
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
Scenario
Receptor
Model Code
Model Value
Units
Description
Reference
Pasture
Farmer
CR milk adult
35
[g WW/kg BW/dayl
Consumption rate, milk (adult)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
ED child1-5
13
fyrsl
Exposure duration (child aged 1-5)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
ED child6-11
13
fyrsl
Exposure duration (child aged 6-11)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
ED childl 2-19
13
fyrsl
Exposure duration (child aged 12-19)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
ED adult
48
fyrsl
Exposure duration (adult)
EFH:2011 (US EPA, 2011)
Pasture
Farmer
ShowerTime
15
fminl
Time in shower stall during shower
EFH:2011 (US EPA, 2011)
Pasture
Farmer
T bathroom
5
fminl
Time spent in bathroom, not in shower
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
BW childl-5
15
fkgl
Body weight (child aged 1-5)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
BW child6-11
29
fkgl
Body weight (child aged 6-11)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
BW childl 2-19
61
fkgl
Body weight (child aged 12-19)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
BW adult
79
fkgl
Body weight (adult)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
CR beef childl-5
11
fg WW/kg BW/dayl
Consumption rate, beef (child aged 1-5)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
CR beef child6-11
11
fg WW/kg BW/dayl
Consumption rate, beef (child aged 6-11)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
CR beef childl 2-19
3.5
fg WW/kg BW/dayl
Consumption rate, beef (child aged 12-19)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
CR beef adult
5.4
fg WW/kg BW/dayl
Consumption rate, beef (adult)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
CR dw childl-5
44
fmL/kg-dayl
Consumption rate, water (child aged 1 -5)
EFH:2019-dw
Reclamation
Farmer
CR dw child6-11
31
fmL/kg-dayl
Consumption rate, water (child aged 6-11)
EFH:2019-dw
Reclamation
Farmer
CR dw childl 2-19
25
fmL/kg-dayl
Consumption rate, water (child aged 12-19)
EFH:2019-dw
Reclamation
Farmer
CR dw adult
34
fmL/kg-dayl
Consumption rate, water (adult)
EFH:2019-dw
Reclamation
Farmer
CR milk childl-5
59
fg WW/kg BW/dayl
Consumption rate, milk (child aged 1-5)
EFH:2018-meatdairy
Reclamation
Farmer
CR milk child6-11
26
fg WW/kg BW/dayl
Consumption rate, milk (child aged 6-11)
EFH:2018-meatdairy
Reclamation
Farmer
CR milk childl 2-19
12
fg WW/kg BW/dayl
Consumption rate, milk (child aged 12-19)
EFH:2018-meatdairy
Reclamation
Farmer
CR milk adult
35
fg WW/kg BW/dayl
Consumption rate, milk (adult)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
ED childl-5
13
fyrsl
Exposure duration (child aged 1-5)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
ED child6-11
13
fyrsl
Exposure duration (child aged 6-11)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
ED childl 2-19
13
fyrsl
Exposure duration (child aged 12-19)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
ED adult
48
fyrsl
Exposure duration (adult)
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
ShowerTime
15
fminl
Time in shower stall during shower
EFH:2011 (US EPA, 2011)
Reclamation
Farmer
T_bathroom
5
fminl
Time spent in bathroom, not in shower
EFH:2011 (US EPA, 2011)
Table E-4. Chemical-specific Inputs
Chemical
Name
Model Code
Value
Units
User
Modified
Description
Reference
UserComment
PFOA
BCF_beef
0.153
[mg/kg beef]/[mg/kg DW]
FALSE
Bioconcentration factor (beef)
Vestergren et al.
(2013)
PFOA
BCF_milk
0.233
[mg/kg milk]/[mg/kg DW]
FALSE
Bioconcentration factor (milk)
Vestergren et al.
(2013)
PFOA
BCF_T3F
8.5
[mg/kg fish]/[mg/L water]
FALSE
Bioaccumulation factor (TL3 fish, filet; used for
human)
Burkhard 2021
Table 4
PFOA
BCF_T3W
140
[mg/kg fish]/[mg/L water]
FALSE
Bioaccumulation factor (TL3 fish, whole; used for
eco)
Burkhard 2021
Table 4
DRAFT
E-4
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
Chemical
Name
Model Code
Value
Units
User
Modified
Description
Reference
UserComment
PFOA
BCF_T4F
8.5
[mg/kg fish]/[mg/L water]
FALSE
Bioaccumulation factor (TL4 fish, filet; used for
human)
Burkhard 2021
Table 4
PFOA
BCF_T4W
140
[mg/kg fish]/[mg/L water]
FALSE
Bioaccumulation factor (TL4 fish, whole; used for
eco)
Burkhard 2021
Table 4
PFOA
BrExFruit
0.11
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer (soil to exposed fruit)
Blaine etal. (2014)
PFOA
BrExVeg
1.6
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to exposed vegetables)
Blaine etal. (2013)
PFOA
BrForage
0.25
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to forage)
Blaine etal. (2013)
PFOA
BrGrain
0.25
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to grain)
Blaine etal. (2013)
PFOA
BrProFruit
0.11
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to protected fruit)
Blaine etal. (2014)
PFOA
BrProVeg
1.6
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to protected vegetables)
Blaine etal. (2013)
PFOA
BrSilage
0.25
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to silage)
Blaine etal. (2013)
PFOA
Bv
0
[ug/g DW plant]/[ug/g air]
FALSE
Biotransfer factor (vapor phase air to all plants;
organics only)
No Data
No data available
PFOA
CSFOral
29300
[per mg/kg-day]
FALSE
Oral cancer slope factor (human toxicity)
Prop PFOA-PFOS
Tox
October 2022
Candidate Tox Values
PFOA
CTPWasteDry
0.307
fug/g DWl
FALSE
Dry biosolids concentration
VT DEC
PFOA
Da
0.0515
[cm2/sl
FALSE
Diffusivity in air
EPA Estimation Tool
PFOA
Dw
5.52E-
6
[cm2/s]
FALSE
Diffusion coefficient in water
EPA Estimation Tool
PFOA
Heat_of_Henry
50000
[J/mol]
FALSE
Enthalpy of phase transformation from aqueous
solution to air solution
EPISuite (US EPA,
2010)
PFOA
HLC
0
[atm-m3/mol]
FALSE
Henry's law constant
HSDB (US NLM,
2010)
Sec 6.10, "not
expected to volatilize
from water or moist
soil"
PFOA
IUR
0
[ug/m3]A-1
FALSE
Inhalation unit risk (human toxicity, cancer)
No Data
PFOA
kaer
0
(1/day)
FALSE
Aerobic biodegradation rate (surface-water column)
No Data
No data, estimation
tools not appropriate
PFOA
Kanaer
0
[1/day]
FALSE
Anaerobic degradation rate (sediment)
No Data
No data, estimation
tools not appropriate
PFOA
kh
0
[1/day]
FALSE
Hydrolysis rate
No Data
No data, estimation
tools not appropriate
PFOA
Koc
114.8
[mL/g]
FALSE
Organic carbon partition coefficient
PFOA Health Effects
Support Document
DRAFT
E-5
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
Chemical
Name
Model Code
Value
Units
User
Modified
Description
Reference
UserComment
PFOA
Kpo
0
(1/day)
FALSE
Photolysis degradation rate in the surface of the
water column
No Data
PFOA
ksoil
0
[1/day]
FALSE
Biodegradation rate (soil)
No Data
No data, estimation
tools not appropriate
PFOA
MW
414
[g/mol]
FALSE
Molecular weight
HSDB (US NLM,
2010)
PFOA
RCF
0.02
[ug/g WW plant]/[ug/mL
soil waterl
FALSE
Root concentration factor*
Lechner and Knapp
(2011)
PFOA
Ref BMD Bird
0
[mg chem/kg BW/dayl
FALSE
Reference benchmark dose (bird)
No Data
PFOA
Ref BMD Mammal
0
[mg chem/kg BW/dayl
FALSE
Reference benchmark dose (mammal)
No Data
PFOA
Ref BW Bird
0
[kgl
FALSE
Reference body weight (bird)
No Data
PFOA
Ref BW Mammal
0
fkgl
FALSE
Reference body weight (mammal)
No Data
PFOA
RFC
0
[mg/m3]
FALSE
Reference concentration (human toxicity,
noncancer)
No Data
PFOA
RfD
3e-8
[mg/kg-day]
FALSE
Reference dose (human toxicity, noncancer)
Prop PFOA-PFOS
Tox
October 2022
Candidate Tox Values
PFOA
RFD_By_Pathway
False
NA
FALSE
True if Model run with pathway-specific RfD
System
PFOA
Sol
0.0095
[mg/Ll
FALSE
Solubility
Physprop
PFOA
temp_ref_aer_all
25
[C]
FALSE
Reference temperature for water column
degradation
Default
PFOA
temp ref anae all
25
[CI
FALSE
Reference temperature for benthic degradation
Default
PFOS
BCF_beef
0.874
[mg/kg beef]/[mg/kg DW]
FALSE
Bioconcentration factor (beef)
Vestergren et al.
(2013)
PFOS
BCFjnilk
0.44
[mg/kg milk]/[mg/kg DW]
FALSE
Bioconcentration factor (milk)
Vestergren et al.
(2013)
PFOS
BCF_T3F
1500
[mg/kg fish]/[mg/L water]
FALSE
Bioaccumulation factor (TL3 fish, filet; used for
human)
Burkhard 2021
Table 4
PFOS
BCF_T3W
3500
[mg/kg fish]/[mg/L water]
FALSE
Bioaccumulation factor (TL3 fish, whole; used for
eco)
Burkhard 2021
Table 4
PFOS
BCF_T4F
1500
[mg/kg fish]/[mg/L water]
FALSE
Bioaccumulation factor (TL4 fish, filet; used for
human)
Burkhard 2021
Table 4
PFOS
BCF_T4W
3500
[mg/kg fish]/[mg/L water]
FALSE
Bioaccumulation factor (TL4 fish, whole; used for
eco)
Burkhard 2021
Table 4
PFOS
BrExFruit
0.03
[mg/kg DW plant]/[mg/kg
soill
FALSE
Biotransfer (soil to exposed fruit)
Blaine etal. (2014)
PFOS
BrExVeg
1.5
[mg/kg DW plant]/[mg/kg
soill
FALSE
Biotransfer factor (soil to exposed vegetables)
Blaine etal. (2013)
PFOS
BrForage
0.07
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to forage)
Blaine etal. (2013)
PFOS
BrGrain
0.07
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to grain)
Blaine etal. (2013)
DRAFT
E-6
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
Chemical
Name
Model Code
Value
Units
User
Modified
Description
Reference
UserComment
PFOS
BrProFruit
0.03
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to protected fruit)
Blaine etal. (2014)
PFOS
BrProVeg
1.5
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to protected vegetables)
Blaine etal. (2013)
PFOS
BrSilage
0.07
[mg/kg DW plant]/[mg/kg
soil]
FALSE
Biotransfer factor (soil to silage)
Blaine etal. (2013)
PFOS
Bv
0
[ug/g DW plant]/[ug/g air]
FALSE
Biotransfer factor (vapor phase air to all plants;
organics only)
No Data
No data available
PFOS
CSFOral
45.2
[per mg/kg-day]
FALSE
Oral cancer slope factor (human toxicity)
Prop PFOA-PFOS
Tox
October 2022
Candidate Tox Values
PFOS
CTPWasteDry
2.15
fug/g DWl
FALSE
Dry biosolids concentration
Ml EGLE
PFOS
Da
0.0466
[cm2/sl
FALSE
Diffusivity in air
EPA Estimation Tool
PFOS
Dw
4.96E-
6
[cm2/s]
FALSE
Diffusion coefficient in water
EPA Estimation Tool
PFOS
Heat_of_Henry
37000
[J/mol]
FALSE
Enthalpy of phase transformation from aqueous
solution to air solution
EPISuite (US EPA,
2010)
PFOS
HLC
0
[atm-m3/mol]
FALSE
Henry's law constant
HSDB (US NLM,
2010)
Sec 6.10, HLC "<4.9E-
9", set to zero due to
uncertainty
PFOS
IUR
0
[ug/m3lA-1
FALSE
Inhalation unit risk (human toxicity, cancer)
No Data
PFOS
kaer
0
(1/day)
FALSE
Aerobic biodegradation rate (surface-water column)
No Data
No data, estimation
tools not appropriate
PFOS
Kanaer
0
[1/day]
FALSE
Anaerobic degradation rate (sediment)
No Data
No data, estimation
tools not appropriate
PFOS
kh
0
[1/day]
FALSE
Hydrolysis rate
No Data
No data, estimation
tools not appropriate
PFOS
Koc
371.5
[mL/gl
FALSE
Organic carbon partition coefficient
PFOS HESD
PFOS
Kpo
0
(1/day)
FALSE
Photolysis degradation rate in the surface of the
water column
No Data
PFOS
ksoil
0
[1/day]
FALSE
Biodegradation rate (soil)
No Data
No data, estimation
tools not appropriate
PFOS
MW
500
[g/mol]
FALSE
Molecular weight
Physprop
PFOS
RCF
0.08
[ug/g WW plant]/[ug/mL
soil waterl
FALSE
Root concentration factor*
Lechner and Knapp
(2011)
PFOS
Ref BMD Bird
0
[mg chem/kg BW/dayl
FALSE
Reference benchmark dose (bird)
No Data
PFOS
Ref BMD Mammal
0
[mg chem/kg BW/dayl
FALSE
Reference benchmark dose (mammal)
No Data
PFOS
Ref BW Bird
0
[kgl
FALSE
Reference body weight (bird)
No Data
PFOS
Ref BW Mammal
0
[kgl
FALSE
Reference body weight (mammal)
No Data
PFOS
RFC
0
[mg/m3]
FALSE
Reference concentration (human toxicity,
noncancer)
No Data
PFOS
RfD
1.00E-
07
[mg/kg-day]
FALSE
Reference dose (human toxicity, noncancer)
Prop PFOA-PFOS
Tox
October 2022
Candidate Tox Values
DRAFT
E-7
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
Chemical
User
Name
Model Code
Value
Units
Modified
Description
Reference
UserComment
PFOS
RFD_By_Pathway
False
NA
FALSE
True if Model run with pathway-specific RfD
System
PFOS
Sol
370
(mg/Ll
FALSE
Solubility
OECD, 2002
in fresh water
PFOS
temp_ref_aer_all
25
[C]
FALSE
Reference temperature for water column
degradation
Default
PFOS
temp_ref_anae_all
25
[CI
FALSE
Reference temperature for benthic degradation
Default
DRAFT
E-8
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
E.2 BST Screening Results
All results are for the farm family (adult farmer or child of farmer). Pathway abbreviations are as follows:
• Beef: consumption of beef from beef cattle pastured on the farm
• ExFruit: consumption of exposed fruits grown on the farm
• ExVeg: consumption of exposed vegetables grown on the farm
• Fish: consumption offish caught in farm pond
• GW: consumption of groundwater from well located on farm
• Milk: consumption of milk from dairy cows pastured on the farm
• ProFruit: consumption of protected fruits grown on the farm
• ProVeg: consumption of protected vegetables grown on the farm
• Root: consumption of root vegetables grown on the farm
• Soil: consumption of soil from the farm field
• SW: consumption of surface water from nearby reservoir.
Media concentration, dose, HQ and CRL results are presented for PFOA and PFOS in tables E-5 to E-8. In
the tables with non-cancer results, the dose column represents the ADD. In the tables with cancer
results the dose column represents the LADD. HQ s and CRLs for sensitive pathways, such as milk, beef,
fish, and drinking water are often very high (over 100 HQ or over 1 in 1000 CRL). Note that the chicken
egg consumption pathway was not included in the BST.
DRAFT
E-9
-------
PFOA/PFOS Risk Assessment
Table E-5. Noncancer Results: PFOA
Appendix E: Screening-level Results from BST
Scenario
Receptor
Pathway
Media
Cone.
Units
Dry Climate
Moderate Climate
Wet Climate
HQ
(unitless)
Dose
(mq/kq-d)
Media Cone,
(units see left)
HQ
(unitless)
Dose
(mq/kq-d)
Media Cone,
(units see left)
HQ
(unitless)
Dose
(mq/kq-
d)
Media Cone,
(units see left)
Crop
Adult
GW
mg/L
7,417
2.2E-04
6.5E-03
2,237
6.7E-05
2.0E-03
1,899
5.7E-05
1.7E-03
Crop
Adult
SW
mg/L
2,346
7.0E-05
2.1E-03
583
1.7E-05
5.1E-04
394
1.2E-05
3.5E-04
Crop
Adult
Fish
mg/kg WW
1,255
3.8E-05
1.4E-01
570
1.7E-05
6.1E-02
324
9.7E-06
3.5E-02
Crop
Adult
ExVeg
mg/kg WW
393
1.2E-05
2.3E-03
100
3.0E-06
5.9E-04
84
2.5E-06
5.0E-04
Crop
Adult
ProVeg
mg/kg WW
318
9.5E-06
3.0E-03
81
2.4E-06
7.7E-04
68
2.1E-06
6.5E-04
Crop
Adult
ProFruit
mg/kg WW
64
1.9E-06
1.9E-04
16
4.9E-07
4.9E-05
14
4.1E-07
4.2E-05
Crop
Adult
ExFruit
mg/kg WW
30
8.9E-07
2.3E-04
8
2.3E-07
5.7E-05
6
1.9E-07
4.8E-05
Crop
Adult
Root
mg/kg WW
0.2
6.3E-09
2.2E-06
0.05
1.6E-09
5.5E-07
0.05
1.4E-09
4.6E-07
Crop
Adult
Soil
mg/kg
0.06
1.7E-09
1.3E-02
0.009
2.8E-10
2.2E-03
0.007
2.0E-10
1.6E-03
Crop
Child
GW
mg/L
9,599
2.9E-04
6.5E-03
2,895
8.7E-05
2.0E-03
2,458
7.4E-05
1.7E-03
Crop
Child
SW
mg/L
3,036
9.1E-05
2.1E-03
754
2.3E-05
5.1E-04
510
1.5E-05
3.5E-04
Crop
Child
Fish
mg/kg WW
1,562
4.7E-05
1.4E-01
710
2.1E-05
6.1E-02
403
1.2E-05
3.5E-02
Crop
Child
ExVeg
mg/kg WW
419
1.3E-05
2.3E-03
106
3.2E-06
5.9E-04
90
2.7E-06
5.0E-04
Crop
Child
ProVeg
mg/kg WW
274
8.2E-06
3.0E-03
69
2.1E-06
7.7E-04
59
1.8E-06
6.5E-04
Crop
Child
ProFruit
mg/kg WW
73
2.2E-06
1.9E-04
19
5.6E-07
4.9E-05
16
4.7E-07
4.2E-05
Crop
Child
ExFruit
mg/kg WW
32
9.6E-07
2.3E-04
8
2.4E-07
5.7E-05
7
2.1E-07
4.8E-05
Crop
Child
Soil
mg/kg
1.2
3.5E-08
1.3E-02
0.2
6.0E-09
2.2E-03
0.14
4.2E-09
1.6E-03
Crop
Child
Root
mg/kg WW
0.4
1.2E-08
2.2E-06
0.10
3.0E-09
5.5E-07
0.08
2.5E-09
4.6E-07
Pasture
Adult
GW
mg/L
8,664
2.6E-04
7.6E-03
1,278
3.8E-05
1.1E-03
854
2.6E-05
7.5E-04
Pasture
Adult
SW
mg/L
3,366
1.0E-04
3.0E-03
878
2.6E-05
7.8E-04
587
1.8E-05
5.2E-04
Pasture
Adult
Milk
mg/kg WW
3,520
1.1E-04
3.0E-03
481
1.4E-05
4.1E-04
382
1.1E-05
3.3E-04
Pasture
Adult
Fish
mg/kg WW
1,333
4.0E-05
1.4E-01
358
1.1E-05
3.9E-02
169
5.1E-06
1.8E-02
Pasture
Adult
Beef
mg/kg WW
232
7.0E-06
2.3E-03
32
9.5E-07
3.2E-04
25
7.5E-07
2.5E-04
Pasture
Adult
Soil
mg/kg
0.2
7.1E-09
5.6E-02
0.03
9.7E-10
7.7E-03
0.02
7.4E-10
5.8E-03
Pasture
Child
GW
mg/L
11,213
3.4E-04
7.6E-03
1,654
5.0E-05
1.1E-03
1,106
3.3E-05
7.5E-04
Pasture
Child
SW
mg/L
4,356
1.3E-04
3.0E-03
1,137
3.4E-05
7.8E-04
760
2.3E-05
5.2E-04
Pasture
Child
Milk
mg/kg WW
5,934
1.8E-04
3.0E-03
812
2.4E-05
4.1E-04
643
1.9E-05
3.3E-04
Pasture
Child
Fish
mg/kg WW
1,660
5.0E-05
1.4E-01
446
1.3E-05
3.9E-02
210
6.3E-06
1.8E-02
Pasture
Child
Beef
mg/kg WW
473
1.4E-05
2.3E-03
65
1.9E-06
3.2E-04
51
1.5E-06
2.5E-04
Pasture
Child
Soil
mg/kg
5
1.5E-07
5.6E-02
0.7
2.1E-08
7.7E-03
0.5
1.6E-08
5.8E-03
Reclamation
Adult
Milk
mg/kg WW
3,726
1.1E-04
3.2E-03
3,220
9.7E-05
2.8E-03
2,483
7.4E-05
2.1E-03
Reclamation
Adult
Beef
mg/kg WW
246
7.4E-06
2.5E-03
211
6.3E-06
2.1E-03
164
4.9E-06
1.6E-03
Reclamation
Adult
GW
mg/L
139
4.2E-06
1.2E-04
163
4.9E-06
1.4E-04
112
3.4E-06
9.9E-05
Reclamation
Adult
Soil
mg/kg
0.2
7.5E-09
5.9E-02
0.2
5.8E-09
4.6E-02
0.2
4.9E-09
3.9E-02
Reclamation
Adult
SW
mg/L
0.005
1.5E-10
4.5E-09
6.6E-05
2.0E-12
5.8E-11
0.04
1.1E-09
3.3E-08
Reclamation
Adult
Fish
mg/kg WW
5.5E-04
1.65E-11
5.93E-08
6.4E-06
1.93E-13
6.92E-10
0.004
1.22E-10
4.38E-07
DRAFT
E-10
-------
PFOA/PFOS Risk Assessment Appendix E: Screening-level Results from BST
Dry Climate
Moderate Climate
Wet Climate
Media
Dose
Cone.
HQ
Dose
Media Cone.
HQ
Dose
Media Cone.
HQ
(mg/kg-
Media Cone.
Scenario
Receptor
Pathway
Units
(unitless)
(mg/kg-d)
(units see left)
(unitless)
(mg/kg-d)
(units see left)
(unitless)
d)
(units see left)
Reclamation
Child
Milk
mg/kg WW
6,280
1.9E-04
3.2E-03
5,428
1.6E-04
2.8E-03
4,186
1.3E-04
2.1E-03
Reclamation
Child
Beef
mg/kg WW
500
1.5E-05
2.5E-03
430
1.3E-05
2.1E-03
333
1.0E-05
1.6E-03
Reclamation
Child
GW
mg/L
180
5.4E-06
1.2E-04
211
6.3E-06
1.4E-04
145
4.4E-06
9.9E-05
Reclamation
Child
Soil
mg/kg
5
1.6E-07
5.9E-02
4
1.2E-07
4.6E-02
3
1.0E-07
3.9E-02
Reclamation
Child
SW
mg/L
0.007
2.0E-10
4.5E-09
8.5E-05
2.5E-12
5.8E-11
0.05
1.4E-09
3.3E-08
Reclamation
Child
Fish
mg/kg WW
6.9E-04
2.1E-11
5.9E-08
8.0E-06
2.4E-13
6.9E-10
0.005
1.5E-10
4.4E-07
Table E-6. Noncancer Results: PFOS
Media
Dry Climate
Moderate Climate
Wet Climate
Cone.
HQ
Dose
Media Cone.
HQ
Dose
Media Cone.
HQ
Dose
Media Cone.
Scenario
Receptor
Pathway
Units
(unitless)
(mg/kg-d)
(units see left)
(unitless)
(mg/kg-d)
(units see left)
(unitless)
(mg/kg-d)
(units see left)
Crop
Adult
Fish
mg/kg WW
210,949
2.1E-02
7.6E+01
124,481
1.2E-02
4.5E+01
57,646
5.8E-03
2.1E+01
Crop
Adult
GW
mg/L
9,281
9.3E-04
2.7E-02
2,805
2.8E-04
8.3E-03
2,314
2.3E-04
6.8E-03
Crop
Adult
SW
mg/L
3,067
3.1E-04
9.0E-03
1,499
1.5E-04
4.4E-03
993
9.9E-05
2.9E-03
Crop
Adult
ExVeg
mg/kg WW
1,171
1.2E-04
2.3E-02
389
3.9E-05
7.7E-03
301
3.0E-05
6.0E-03
Crop
Adult
ProVeg
mg/kg WW
948
9.5E-05
3.0E-02
315
3.2E-05
1.0E-02
243
2.4E-05
7.8E-03
Crop
Adult
ProFruit
mg/kg WW
56
5.6E-06
5.6E-04
18
1.8E-06
1.9E-04
14
1.4E-06
1.4E-04
Crop
Adult
ExFruit
mg/kg WW
26
2.6E-06
6.5E-04
9
8.6E-07
2.2E-04
7
6.6E-07
1.7E-04
Crop
Adult
Root
mg/kg WW
0.8
8.3E-08
2.8E-05
0.3
2.8E-08
9.4E-06
0.2
2.1E-08
7.3E-06
Crop
Adult
Soil
mg/kg
0.2
1.8E-08
1.4E-01
0.05
4.8E-09
3.8E-02
0.03
2.9E-09
2.3E-02
Crop
Child
Fish
mg/kg WW
262,600
2.6E-02
7.6E+01
154,960
1.5E-02
4.5E+01
71,760
7.2E-03
2.1E+01
Crop
Child
GW
mg/L
12,011
1.2E-03
2.7E-02
3,630
3.6E-04
8.3E-03
2,995
3.0E-04
6.8E-03
Crop
Child
SW
mg/L
3,969
4.0E-04
9.0E-03
1,940
1.9E-04
4.4E-03
1,285
1.3E-04
2.9E-03
Crop
Child
ExVeg
mg/kg WW
1,250
1.2E-04
2.3E-02
415
4.2E-05
7.7E-03
321
3.2E-05
6.0E-03
Crop
Child
ProVeg
mg/kg WW
816
8.2E-05
3.0E-02
271
2.7E-05
1.0E-02
209
2.1E-05
7.8E-03
Crop
Child
ProFruit
mg/kg WW
63
6.3E-06
5.6E-04
21
2.1E-06
1.9E-04
16
1.6E-06
1.4E-04
Crop
Child
ExFruit
mg/kg WW
28
2.8E-06
6.5E-04
9
9.2E-07
2.2E-04
7
7.1E-07
1.7E-04
Crop
Child
Soil
mg/kg
4
3.8E-07
1.4E-01
1.0
1.0E-07
3.8E-02
0.6
6.2E-08
2.3E-02
Crop
Child
Root
mg/kg WW
2
1.5E-07
2.8E-05
0.5
5.1E-08
9.4E-06
0.4
3.9E-08
7.3E-06
Pasture
Adult
Fish
mg/kg WW
279,038
2.8E-02
1.0E+02
58,899
5.9E-03
2.1E+01
26,066
2.6E-03
9.4E+00
Pasture
Adult
GW
mg/L
4,068
4.1E-04
1.2E-02
2,186
2.2E-04
6.4E-03
1,561
1.6E-04
4.6E-03
Pasture
Adult
SW
mg/L
5,882
5.9E-04
1.7E-02
1,734
1.7E-04
5.1E-03
1,017
1.0E-04
3.0E-03
Pasture
Adult
Milk
mg/kg WW
10,898
1.1E-03
3.1E-02
1,255
1.3E-04
3.6E-03
1,088
1.1E-04
3.1E-03
Pasture
Adult
Beef
mg/kg WW
2,451
2.5E-04
8.2E-02
281
2.8E-05
9.4E-03
242
2.4E-05
8.1E-03
Pasture
Adult
Soil
mg/kg
1.1
1.1E-07
8.9E-01
0.13
1.3E-08
9.9E-02
0.11
1.1E-08
8.4E-02
Pasture
Child
Fish
mg/kg WW
347,360
3.5E-02
1.0E+02
73,320
7.3E-03
2.1E+01
32,448
3.2E-03
9.4E+00
Pasture
Child
GW
mg/L
5,264
5.3E-04
1.2E-02
2,829
2.8E-04
6.4E-03
2,021
2.0E-04
4.6E-03
DRAFT
E-ll
-------
PFOA/PFOS Risk Assessment Appendix E: Screening-level Results from BST
Media
Dry Climate
Moderate Climate
Wet Climate
Cone.
HQ
Dose
Media Cone.
HQ
Dose
Media Cone.
HQ
Dose
Media Cone.
Scenario
Receptor
Pathway
Units
(unitless)
(mg/kg-d)
(units see left)
(unitless)
(mq/kq-d)
(units see left)
(unitless)
(mq/kq-d)
(units see left)
Pasture
Child
SW
mg/L
7,612
7.6E-04
1.7E-02
2,244
2.2E-04
5.1E-03
1,316
1.3E-04
3.0E-03
Pasture
Child
Milk
mg/kg WW
18,371
1.8E-03
3.1E-02
2,115
2.1E-04
3.6E-03
1,833
1.8E-04
3.1E-03
Pasture
Child
Beef
mg/kg WW
4,992
5.0E-04
8.2E-02
572
5.7E-05
9.4E-03
494
4.9E-05
8.1E-03
Pasture
Child
Soil
mg/kg
24
2.4E-06
8.9E-01
3
2.6E-07
9.9E-02
2
2.2E-07
8.4E-02
Reclamation
Adult
Milk
mg/kg WW
5,237
5.2E-04
1.5E-02
4,864
4.9E-04
1.4E-02
3,828
3.8E-04
1.1E-02
Reclamation
Adult
Beef
mg/kg WW
1,178
1.2E-04
3.9E-02
1,086
1.1E-04
3.6E-02
846
8.5E-05
2.8E-02
Reclamation
Adult
GW
mg/L
167
1.7E-05
4.9E-04
83
8.3E-06
2.4E-04
54
5.4E-06
1.6E-04
Reclamation
Adult
Soil
mg/kg
0.5
5.4E-08
4.3E-01
0.5
4.8E-08
3.8E-01
0.4
3.8E-08
3.0E-01
Reclamation
Adult
Fish
mg/kg WW
0.2
1.5E-08
5.4E-05
0.002
1.8E-10
6.4E-07
1.2
1.2E-07
4.4E-04
Reclamation
Adult
SW
mg/L
0.009
8.6E-10
2.5E-08
1.1E-04
1.1E-11
3.2E-10
0.07
6.8E-09
2.0E-07
Reclamation
Child
Milk
mg/kg WW
8,829
8.8E-04
1.5E-02
8,199
8.2E-04
1.4E-02
6,453
6.5E-04
1.1E-02
Reclamation
Child
Beef
mg/kg WW
2,400
2.4E-04
3.9E-02
2,213
2.2E-04
3.6E-02
1,723
1.7E-04
2.8E-02
Reclamation
Child
GW
mg/L
217
2.2E-05
4.9E-04
107
1.1E-05
2.4E-04
70
7.0E-06
1.6E-04
Reclamation
Child
Soil
mg/kg
11
1.1E-06
4.3E-01
10
1.0E-06
3.8E-01
8
8.0E-07
3.0E-01
Reclamation
Child
Fish
mg/kg WW
0.2
1.9E-08
5.4E-05
0.002
2.2E-10
6.4E-07
2
1.5E-07
4.4E-04
Reclamation
Child
SW
mg/L
0.011
1.1E-09
2.5E-08
1.4E-04
1.4E-11
3.2E-10
0.09
8.8E-09
2.0E-07
Table E-7. Cancer Results: PFOA
Scenario
Receptor
Pathway
Media
Cone.
Units
Dry Climate
Moderate Climate
Wet Climate
Risk
(unitless)
Dose
(mg/kg-
d)
Media Cone,
(units see
left)
Risk
(unitless)
Dose
(mg/kg-d)
Media Cone,
(units see left)
Risk
(unitless)
Dose
(mg/kg-d)
Media Cone,
(units see left)
Crop
Adult
GW
mg/L
4.2E+00
1.4E-04
6.4E-03
9.4E-01
3.2E-05
1.4E-03
9.2E-01
3.2E-05
1.4E-03
Crop
Adult
SW
mg/L
1.0E+00
3.4E-05
1.5E-03
2.4E-01
8.3E-06
3.7E-04
1.6E-01
5.3E-06
2.4E-04
Crop
Adult
Fish
mg/kg WW
5.4E-01
1.8E-05
1.0E-01
2.3E-01
7.8E-06
4.2E-02
1.3E-01
4.3E-06
2.4E-02
Crop
Adult
ExVeg
mg/kg WW
1.9E-01
6.5E-06
2.0E-03
4.1E-02
1.4E-06
4.2E-04
3.5E-02
1.2E-06
3.6E-04
Crop
Adult
ProVeg
mg/kg WW
1.5E-01
5.3E-06
2.6E-03
3.3E-02
1.1E-06
5.5E-04
2.9E-02
9.7E-07
4.7E-04
Crop
Adult
ProFruit
mg/kg WW
3.1E-02
1.1E-06
1.6E-04
6.7E-03
2.3E-07
3.5E-05
5.8E-03
2.0E-07
3.0E-05
Crop
Adult
ExFruit
mg/kg WW
1.4E-02
4.9E-07
1.9E-04
3.1E-03
1.1E-07
4.1E-05
2.7E-03
9.1E-08
3.5E-05
Crop
Adult
Root
mg/kg WW
1.0E-04
3.5E-09
1.8E-06
2.2E-05
7.5E-10
3.9E-07
1.9E-05
6.5E-10
3.4E-07
Crop
Adult
Soil
mg/kg
1.5E-05
5.1E-10
6.2E-03
2.1E-06
7.2E-11
8.7E-04
1.6E-06
5.3E-11
6.4E-04
Crop
Child
GW
mg/L
1.2E+00
4.0E-05
6.4E-03
3.6E-01
1.2E-05
2.0E-03
3.1E-01
1.0E-05
1.7E-03
Crop
Child
SW
mg/L
3.3E-01
1.1E-05
1.9E-03
8.6E-02
2.9E-06
4.7E-04
5.4E-02
1.8E-06
3.0E-04
Crop
Child
Fish
mg/kg WW
1.7E-01
5.9E-06
1.2E-01
7.2E-02
2.4E-06
5.0E-02
4.0E-02
1.4E-06
2.8E-02
Crop
Child
ExVeg
mg/kg WW
4.3E-02
1.5E-06
2.3E-03
9.9E-03
3.4E-07
5.3E-04
7.9E-03
2.7E-07
4.2E-04
Crop
Child
ProVeg
mg/kg WW
3.3E-02
1.1E-06
2.9E-03
7.6E-03
2.6E-07
6.9E-04
6.1E-03
2.1E-07
5.5E-04
Crop
Child
ProFruit
mg/kg WW
1.0E-02
3.5E-07
1.9E-04
2.4E-03
8.2E-08
4.4E-05
1.9E-03
6.5E-08
3.5E-05
Crop
Child
ExFruit
mg/kg WW
5.2E-03
1.8E-07
2.2E-04
1.2E-03
4.2E-08
5.1E-05
9.8E-04
3.4E-08
4.1E-05
DRAFT
E-12
-------
PFOA/PFOS Risk Assessment Appendix E: Screening-level Results from BST
Scenario
Receptor
Pathway
Media
Cone.
Units
Dry Climate
Moderate Climate
Wet Climate
Risk
(unitless)
Dose
(mg/kg-
d)
Media Cone,
(units see
left)
Risk
(unitless)
Dose
(mg/kg-d)
Media Cone,
(units see left)
Risk
(unitless)
Dose
(mg/kg-d)
Media Cone,
(units see left)
Crop
Child
Root
mg/kg WW
4.5E-05
1.5E-09
2.1E-06
1.0E-05
3.5E-10
4.9E-07
8.2E-06
2.8E-10
3.9E-07
Crop
Child
Soil
mg/kg
8.2E-05
2.8E-09
8.7E-03
9.3E-06
3.2E-10
1.2E-03
5.9E-06
2.0E-10
8.2E-04
Pasture
Adult
GW
mg/L
2.8E+00
9.5E-05
4.3E-03
5.7E-01
1.9E-05
8.7E-04
3.2E-01
1.1E-05
4.9E-04
Pasture
Adult
SW
mg/L
1.6E+00
5.4E-05
2.4E-03
3.7E-01
1.2E-05
5.6E-04
2.3E-01
8.0E-06
3.6E-04
Pasture
Adult
Fish
mg/kg WW
6.1E-01
2.1E-05
1.1E-01
1.5E-01
5.1E-06
2.8E-02
6.3E-02
2.2E-06
1.2E-02
Pasture
Adult
Milk
mg/kg WW
9.1E-01
3.1E-05
1.3E-03
1.1E-01
3.9E-06
1.7E-04
7.5E-02
2.5E-06
1.1E-04
Pasture
Adult
Beef
mg/kg WW
6.0E-02
2.0E-06
1.0E-03
7.5E-03
2.6E-07
1.3E-04
4.9E-03
1.7E-07
8.5E-05
Pasture
Adult
Soil
mg/kg
6.0E-05
2.0E-09
2.4E-02
7.1E-06
2.4E-10
2.9E-03
4.6E-06
1.6E-10
1.9E-03
Pasture
Child
GW
mg/L
9.3E-01
3.2E-05
4.5E-03
2.1E-01
7.0E-06
1.1E-03
1.4E-01
4.7E-06
7.5E-04
Pasture
Child
SW
mg/L
5.0E-01
1.7E-05
2.8E-03
1.2E-01
4.1E-06
6.7E-04
8.7E-02
3.0E-06
4.8E-04
Pasture
Child
Fish
mg/kg WW
2.0E-01
6.7E-06
1.3E-01
4.8E-02
1.6E-06
3.3E-02
2.4E-02
8.1E-07
1.6E-02
Pasture
Child
Milk
mg/kg WW
4.1E-01
1.4E-05
1.9E-03
4.1E-02
1.4E-06
2.2E-04
3.2E-02
1.1E-06
1.7E-04
Pasture
Child
Beef
mg/kg WW
4.6E-02
1.6E-06
1.5E-03
5.0E-03
1.7E-07
1.7E-04
3.8E-03
1.3E-07
1.3E-04
Pasture
Child
Soil
mg/kg
3.2E-04
1.1E-08
3.5E-02
3.0E-05
1.0E-09
3.9E-03
2.3E-05
8.0E-10
2.8E-03
Reclamation
Adult
Milk
mg/kg WW
5.3E-01
1.8E-05
7.8E-04
9.0E-02
3.1E-06
1.3E-04
5.9E-02
2.0E-06
8.8E-05
Reclamation
Adult
GW
mg/L
8.0E-02
2.7E-06
1.2E-04
7.6E-02
2.6E-06
1.2E-04
2.7E-02
9.2E-07
4.1E-05
Reclamation
Adult
Beef
mg/kg WW
3.4E-02
1.2E-06
6.0E-04
5.9E-03
2.0E-07
1.0E-04
3.9E-03
1.3E-07
6.7E-05
Reclamation
Adult
Soil
mg/kg
3.3E-05
1.1E-09
1.4E-02
5.2E-06
1.8E-10
2.1E-03
3.4E-06
1.2E-10
1.4E-03
Reclamation
Adult
SW
mg/L
2.9E-06
1.0E-10
4.5E-09
3.2E-08
1.1E-12
4.9E-11
8.9E-06
3.0E-10
1.4E-08
Reclamation
Adult
Fish
mg/kg WW
3.2E-07
1.1E-11
5.9E-08
3.1E-09
1.1 E-13
5.9E-10
9.7E-07
3.3E-11
1.8E-07
Reclamation
Child
Milk
mg/kg WW
4.6E-01
1.6E-05
2.1E-03
1.5E-01
5.2E-06
4.9E-04
1.0E-01
3.4E-06
3.2E-04
Reclamation
Child
GW
mg/L
2.2E-02
7.7E-07
1.2E-04
2.6E-02
9.0E-07
1.4E-04
1.8E-02
6.2E-07
9.9E-05
Reclamation
Child
Beef
mg/kg WW
4.8E-02
1.7E-06
1.6E-03
1.2E-02
4.1E-07
3.8E-04
7.9E-03
2.7E-07
2.5E-04
Reclamation
Child
Soil
mg/kg
3.4E-04
1.2E-08
3.7E-02
1.1E-04
3.8E-09
7.9E-03
7.2E-05
2.5E-09
5.2E-03
Reclamation
Child
SW
mg/L
8.1E-07
2.7E-11
4.4E-09
1.0E-08
3.5E-13
5.7E-11
5.9E-06
2.0E-10
3.2E-08
Reclamation
Child
Fish
mg/kg WW
8.5E-08
2.9E-12
5.8E-08
9.9E-10
3.4E-14
6.8E-10
6.3E-07
2.1E-11
4.3E-07
Table E-8. Cancer Results: PFOS
Media
Dry Climate
Moderate Climate
Wet Climate
Cone.
Risk
Dose
Media Cone.
Risk
Dose
Media Cone.
Risk
Dose
Media Cone.
Scenario
Receptor
Pathway
Units
(unitless)
(mg/kg-d)
(units see left)
(unitless)
(mg/kg-d)
(units see left)
(unitless)
(mg/kg-d)
(units see left)
Crop
Adult
Fish
mg/kg WW
4.5E-01
1.0E-02
5.4E+01
2.4E-01
5.4E-03
2.9E+01
1.2E-01
2.7E-03
1.5E+01
Crop
Adult
GW
mg/L
2.8E-02
6.1E-04
2.7E-02
6.4E-03
1.4E-04
6.3E-03
5.9E-03
1.3E-04
5.8E-03
Crop
Adult
SW
mg/L
6.6E-03
1.5E-04
6.5E-03
3.0E-03
6.7E-05
3.0E-03
2.0E-03
4.4E-05
2.0E-03
Crop
Adult
ExVeg
mg/kg WW
2.9E-03
6.4E-05
1.9E-02
9.2E-04
2.0E-05
6.1E-03
7.1E-04
1.6E-05
4.8E-03
Crop
Adult
ProVeg
mg/kg WW
2.3E-03
5.2E-05
2.5E-02
7.4E-04
1.6E-05
8.0E-03
5.8E-04
1.3E-05
6.2E-03
Crop
Adult
ProFruit
mg/kg WW
1.4E-04
3.0E-06
4.6E-04
4.4E-05
9.6E-07
1.5E-04
3.4E-05
7.5E-07
1.1E-04
DRAFT
E-13
-------
PFOA/PFOS Risk Assessment Appendix E: Screening-level Results from BST
Media
Dry Climate
Moderate Climate
Wet Climate
Cone.
Risk
Dose
Media Cone.
Risk
Dose
Media Cone.
Risk
Dose
Media Cone.
Scenario
Receptor
Pathway
Units
(unitless)
(mq/kq-d)
(units see left)
(unitless)
(mq/kq-d)
(units see left)
(unitless)
(mq/kq-d)
(units see left)
Crop
Adult
ExFruit
mg/kg WW
6.3E-05
1.4E-06
5.4E-04
2.0E-05
4.5E-07
1.7E-04
1.6E-05
3.5E-07
1.3E-04
Crop
Adult
Root
mg/kg WW
2.0E-06
4.5E-08
2.3E-05
6.5E-07
1.4E-08
7.5E-06
5.1E-07
1.1E-08
5.8E-06
Crop
Adult
Soil
mg/kg
2.5E-07
5.6E-09
6.8E-02
5.9E-08
1.3E-09
1.6E-02
4.3E-08
9.5E-10
1.1E-02
Crop
Child
Fish
mg/kg WW
1.5E-01
3.3E-03
6.8E+01
8.4E-02
1.9E-03
3.8E+01
4.2E-02
9.2E-04
1.9E+01
Crop
Child
GW
mg/L
7.7E-03
1.7E-04
2.7E-02
2.3E-03
5.2E-05
8.3E-03
1.9E-03
4.3E-05
6.8E-03
Crop
Child
SW
mg/L
2.2E-03
4.8E-05
7.9E-03
1.1E-03
2.4E-05
3.9E-03
6.8E-04
1.5E-05
2.5E-03
Crop
Child
ExVeg
mg/kg WW
6.4E-04
1.4E-05
2.2E-02
2.1E-04
4.7E-06
7.3E-03
1.6E-04
3.5E-06
5.5E-03
Crop
Child
ProVeg
mg/kg WW
4.9E-04
1.1E-05
2.9E-02
1.6E-04
3.6E-06
9.5E-03
1.2E-04
2.7E-06
7.2E-03
Crop
Child
ProFruit
mg/kg WW
4.5E-05
9.9E-07
5.3E-04
1.5E-05
3.3E-07
1.8E-04
1.1E-05
2.5E-07
1.3E-04
Crop
Child
ExFruit
mg/kg WW
2.3E-05
5.1E-07
6.2E-04
7.6E-06
1.7E-07
2.0E-04
5.8E-06
1.3E-07
1.6E-04
Crop
Child
Root
mg/kg WW
8.8E-07
2.0E-08
2.7E-05
2.9E-07
6.4E-09
8.9E-06
2.2E-07
4.9E-09
6.7E-06
Crop
Child
Soil
mg/kg
1.4E-06
3.1E-08
1.0E-01
2.7E-07
6.0E-09
2.3E-02
1.9E-07
4.1E-09
1.6E-02
Pasture
Adult
Fish
mg/kg WW
6.0E-01
1.3E-02
7.3E+01
1.3E-01
2.9E-03
1.6E+01
5.7E-02
1.3E-03
6.8E+00
Pasture
Adult
GW
mg/L
1.2E-02
2.7E-04
1.2E-02
5.5E-03
1.2E-04
5.4E-03
3.9E-03
8.6E-05
3.8E-03
Pasture
Adult
SW
mg/L
1.3E-02
3.0E-04
1.3E-02
3.5E-03
7.8E-05
3.5E-03
2.3E-03
5.0E-05
2.2E-03
Pasture
Adult
Milk
mg/kg WW
2.0E-02
4.3E-04
1.9E-02
2.6E-03
5.7E-05
2.5E-03
1.7E-03
3.8E-05
1.7E-03
Pasture
Adult
Beef
mg/kg WW
4.4E-03
9.7E-05
4.9E-02
5.8E-04
1.3E-05
6.5E-03
3.9E-04
8.6E-06
4.3E-03
Pasture
Adult
Soil
mg/kg
2.0E-06
4.4E-08
5.3E-01
2.6E-07
5.7E-09
6.8E-02
1.7E-07
3.8E-09
4.5E-02
Pasture
Child
Fish
mg/kg WW
2.0E-01
4.5E-03
9.1E+01
4.2E-02
9.4E-04
1.9E+01
1.9E-02
4.1E-04
8.3E+00
Pasture
Child
GW
mg/L
3.4E-03
7.5E-05
1.2E-02
1.8E-03
4.0E-05
6.4E-03
1.3E-03
2.9E-05
4.6E-03
Pasture
Child
SW
mg/L
4.2E-03
9.3E-05
1.5E-02
1.1E-03
2.4E-05
3.9E-03
6.2E-04
1.4E-05
2.3E-03
Pasture
Child
Milk
mg/kg WW
8.5E-03
1.9E-04
2.7E-02
8.4E-04
1.9E-05
2.9E-03
5.9E-04
1.3E-05
2.0E-03
Pasture
Child
Beef
mg/kg WW
3.2E-03
7.2E-05
7.2E-02
3.3E-04
7.3E-06
7.5E-03
2.4E-04
5.3E-06
5.2E-03
Pasture
Child
Soil
mg/kg
1.0E-05
2.3E-07
7.7E-01
9.7E-07
2.1E-08
7.9E-02
6.8E-07
1.5E-08
5.5E-02
Reclamation
Adult
Milk
mg/kg WW
8.0E-03
1.8E-04
7.6E-03
1.3E-03
2.9E-05
1.2E-03
1.0E-03
2.3E-05
9.9E-04
Reclamation
Adult
Beef
mg/kg WW
1.8E-03
3.9E-05
2.0E-02
2.8E-04
6.3E-06
3.2E-03
2.2E-04
5.0E-06
2.5E-03
Reclamation
Adult
GW
mg/L
5.0E-04
1.1E-05
4.9E-04
2.5E-04
5.4E-06
2.4E-04
1.6E-04
3.6E-06
1.6E-04
Reclamation
Adult
Soil
mg/kg
7.9E-07
1.7E-08
2.1E-01
1.1E-07
2.5E-09
3.0E-02
8.8E-08
1.9E-09
2.3E-02
Reclamation
Adult
Fish
mg/kg WW
4.5E-07
9.9E-09
5.4E-05
5.2E-09
1.2E-10
6.3E-07
3.6E-06
8.0E-08
4.4E-04
Reclamation
Adult
SW
mg/L
2.5E-08
5.6E-10
2.5E-08
3.2E-10
7.2E-12
3.2E-10
2.0E-07
4.5E-09
2.0E-07
Reclamation
Child
Milk
mg/kg WW
3.5E-03
7.8E-05
1.1E-02
2.0E-03
4.3E-05
4.5E-03
1.6E-03
3.6E-05
3.6E-03
Reclamation
Child
Beef
mg/kg WW
1.3E-03
2.9E-05
2.9E-02
5.7E-04
1.3E-05
1.2E-02
4.5E-04
1.0E-05
9.3E-03
Reclamation
Child
GW
mg/L
1.4E-04
3.1E-06
4.9E-04
6.9E-05
1.5E-06
2.4E-04
4.5E-05
9.9E-07
1.6E-04
Reclamation
Child
Soil
mg/kg
4.1E-06
9.0E-08
3.1E-01
2.2E-06
4.8E-08
1.1E-01
1.8E-06
4.0E-08
8.6E-02
Reclamation
Child
Fish
mg/kg WW
1.2E-07
2.7E-09
5.4E-05
1.4E-09
3.1E-11
6.3E-07
9.7E-07
2.1E-08
4.3E-04
Reclamation
Child
SW
mg/L
7.1E-09
1.6E-10
2.5E-08
9.0E-11
2.0E-12
3.2E-10
5.6E-08
1.2E-09
2.0E-07
DRAFT
E-14
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
E.3 References
Blaine, A.C., Rich, C.D., Hundal, L.S., Lau, C., Mills, M.A., Harris, K.M., & Higgins, C.P. (2013). Uptake of
perfluoroalkyl acids into edible crops via land applied biosolids: Field and greenhouse studies.
Environmental Science and Technology, 47 (24), 14062-14069, https://doi.org/10.1021/es403094q
Blaine, A. C., C. D. Rich, E. M. Sedlacko, L. S. Hundal, K. Kumar, C. Lau, M. A. Mills, K. M. Harris & C. P.
Higgins (2014). Perfluoroalkyl acid distribution in various plant compartments of edible crops grown
in biosolids-amended soils. Environmental Science and Technology, 48 (14), 7858-7865.
https://doi.org/10.1021/es500Q16s
Burkhard, L. (2021). "Evaluation of published bioconcentration factor (BCF) and bioaccumulation factor
(BAF) data for per- and polyfluoroalkyl substances across aquatic species." Environmental Toxicology
and Chemistry, 40 (6), 1530-1543. https://doi.org/10.1002/etc.501Q
Chapra, S.C. 1996. Surface Water-Quality Modeling: Preliminary Edition. New York: McGraw Hill, Inc.
CWP (Center for Watershed Protection, Inc). 1998. Rapid Watershed Planning Handbook. Center for
Watershed Protection, Inc. Ellicott City, MD. EFH:2017-soil
US EPA (US Environmental Protection Agency). (2018b). Update for Chapter 11 of the Exposure Factors
Handbook: Intake of Meats, Dairy Products, and Fats. EPA/600/R-17/485F. National Center for
Environmental Assessment, Office of Research and Development, Washington, DC. April. Available
at https://www.epa.gov/expobox/about-exposure-factors-handbook.
US EPA (US Environmental Protection Agency). (2019c). Update for Chapter 3 of the Exposure Factors
Handbook: Ingestion of Water and Other Select Liquids. EPA/600/R-18/259F. National Center for
Environmental Assessment, Office of Research and Development, Washington, DC. February.
Available at https://www.epa.gov/expobox/about-exposure-factors-handbook
Gunn, A.P, R.E. Dewhurst, A. Giorgetti, N.L. Gillott, S.J.W. Wishart, and S. Pedley. 2004. Use of Sewage
Sludge Products in Construction. London: CIRIA. Lechner and Knapp (2011)
US EPA (US Environmental Protection Agency). (2014). Estimated Fish Consumption Rates for the US
Population and Selected Subpopulations (NHANES 2003-2010). EPA-820-R-14-002.
OECD (Organisation for Economic Co-operation and Development). (2002). Hazard Assessment of
Perfluorooctane Sulfonate (PFOS) and its Salts. ENV/JM/RD(2002)17/FINAL. Report of the
Environment Directorate, Joint Meeting of the Chemicals Committee and the Working Party on
Chemicals, Pesticides and Biotechnology, Co-operation on Existing Chemicals, Paris, November 21,
2002.
U.S. DOC (Department of Commerce) and U.S. DOE (Department of Energy) National Renewable Energy
Laboratory. 1993. Solar and Meteorological Surface Observation Network (SAMSON), 1961-1990.
Version 1.0. National Climatic Data Center, Asheville, NC. US EPA 2003, Biosolids
U.S. EPA (Environmental Protection Agency). 1989b. Hazardous Waste TSDF- Fugitive Particulate Matter
Air Emissions Guidance Document. US EPA, 1995, AP-42
U.S. EPA (Environmental Protection Agency). 1998. Methodology for Assessing Health Risks Associated
with Multiple Pathways of Exposure to Combustor Emissions. Update to EPA/600/6-90/003.
Methodology for Assessing Health Risks Associated with Indirect Exposure. US EPA, 2005a, HHRAP
U.S. EPA (Environmental Protection Agency). 2010. Estimation Programs Interface Suite (EPISuite) for
Microsoft Windows, v 4.10. Washington, DC: U.S. Environmental Protection Agency. September. US
EPA, 2011, EFH:2011
DRAFT
E-15
-------
PFOA/PFOS Risk Assessment
Appendix E: Screening-level Results from BST
U.S. NLM (National Library of Medicine). 2010. Hazardous Substances Data Bank (HSDB). Available at
http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgenPHSDB. van der Leeden et al., 1990, Water Enc.
Vestergren, R., Orata, F., Berger, U. & Cousins, I.T. (2013). Bioaccumulation of perfluoroalkyl acids in
dairy cows in a naturally contaminated environment. Environmental Science Pollution Research, 20,
7959-7969. DOI 10.1007/sll356-013-1722-x
Wanielista M.P., and Y.A. Yousef. 1993. Stormwater Management. New York, NY. John Wiley & Sons, Inc.
Wischmeier, W.H., and D.D. Smith. 1978. Predicting Rainfall Erosion Losses: A Guide to Conservation
Planning. USDA Agricultural Handbook No. 537. Washington, DC. Department of Agriculture. 62pp.
DRAFT
E-16
------- |