EPA
United Stales
Environmental Protection
Technical Support Document: National-Scale
Mercury Risk Assessment Supporting the
Appropriate and Necessary Finding for Coal-
and Oil-Fired Electric Generating Units -
DRAFT
-------
EPA-452/D-11-002
March 2011
Technical Support Document: National-scale Mercury Risk Assessment
Supporting the Appropriate and Necessary Finding for Coal- and Oil-
fired Electric Generating Units - DRAFT
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Health and Environmental Impacts Division
Research Triangle Park, North Carolina
11
-------
DISCLAIMER
This document has been prepared by staff from the Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency. Any opinions, findings,
conclusions, or recommendations are those of the authors and do not necessarily reflect
the views of the EPA. Questions related to this document should be addressed to Dr.
Zachary Pekar, U.S. Environmental Protection Agency, Office of Air Quality Planning
and Standards, C504-06, Research Triangle Park, North Carolina 27711 (email:
pekar.zachary@epa.gov).
in
-------
Table of Contents:
Executive Summary 1
1. Review of Analysis Approach 13
1.1. Purpose and Scope of Analysis 13
1.2. Overview of Risk Metrics and the Risk Characterization Framework 17
1.3. Overview of Analytical Approach 19
1.4. Discussion of key sources of uncertainty and variability 26
1.5. Differences between the 2005 Section 112(n) Revision Rule analysis and the current
analysis in support of the Propose U.S. EGU Toxics Rule 26
2. Discussion of Analytical Results 29
2.1. Key design elements to consider when reviewing the risk assessment results 29
2.2. Mercury Emissions from U.S. EGUs 30
2.3. Mercury Deposition from U.S. EGUs as Modeled Using CMAQ 31
2.4. Fish Tissue MeHg Concentrations 35
2.5. Comparing Patterns of Hg Deposition with Hg Fish Tissue Data for the 2,461
Watersheds Included in the Risk Assessment 44
2.6. Overview of Risk Estimates 49
2.6.1. Overview of percentile risk estimates 50
2.6.2. Overview of number (and frequency) of watersheds with populations potentially at-
risk due to U.S. EGU mercury emissions 55
2.7. Sensitivity Analyses 58
2.8. Summary of Key Observations 62
Appendices: Additional Technical Detail on Modeling Elements 69
List of Tables
Table 2-1. Comparison of total and U.S. EGU-attributable mercury deposition (ug/m2) for the
2005 and 2016 scenarios.* 34
Table 2-2. Comparison of percent of total mercury deposition attributable to U.S. EGUs for
2005 and 2016.* 34
Table 2-3. Comparison of percent reduction of total mercury deposition, and U.S. EGU-
attributable deposition, based on comparing the 2016 scenario against the 2005 scenario.* 34
Table 2-4. Comparison of total and U.S. EGU-attributable Hg fish tissue concentrations
(including % change) for the 2005 and 2016 scenarios 43
Table 2-5. Comparison of U.S. EGU fraction of total Hg deposition (used to apportion Hg fish
tissue concentrations and risk) between the 2005 and 2016 scenarios. Note, that these values are
specifically for the 2,461 watersheds included in the risk assessment 43
Table 2-6. Percentile risk estimates for the high-end female consumer population assessed
nationally (2005 scenario) (for both total and U.S. EGU incremental risk, including IQ loss and
MeHg RfD-based HQ estimates) 51
IV
-------
Table 2-7. Percentile risk estimates for the high-end female consumer population assessed
nationally (2016 Scenario) (for both total and U.S. EGU incremental risk, including IQ loss and
RfD-based HQ estimates) 51
Table 2-8. Percentile risk estimates for the full set of fishing populations included in the analysis
(2005 scenario) (for both total and U.S. EGU incremental risk, only for RfD-based HQ
estimates) 52
Table 2-9. Watersheds with potentially at-risk populations based on consideration for various
degrees of U.S. EGU contribution to total risk (for IQ loss andHQ) 56
Table 2-10. Watersheds with potentially at-risk populations based on consideration for risk (both
IQ loss and HQ) based on U.S. EGU mercury deposition and resulting exposure before
considering other sources of mercury deposition 57
Table 2-11. Combination of watersheds with potentially at-risk populations based on either
consideration for (a) U.S. EGU percent contribution to total risk OR (b) risk when U.S. EGU
mercury deposition is considered alone 57
Table 2-12. Sensitivity analysis results based on constraining analysis to (a) watersheds in the
top 25th percentile with regard to total Hg deposition and (b) exclude watersheds located in MN,
LA, SE or ME) - Results for watersheds with potentially at risk populations based on U.S. EGUs
making a specified contribution to total risk (Stage 1 of the 3 stage framework) 59
Table 2-13. Sensitivity analysis results based on constraining analysis to (a) watersheds in the
top 25th percentile with regard to total Hg deposition and (b) exclude watersheds located in MN,
LA, SE or ME) - Results for watersheds with potentially at risk populations based on U.S. EGU-
incremental contribution to total risk (Stage 2 of the 3 stage framework) 60
Table 2-14. Sensitivity analysis results based on constraining analysis to (a) watersheds in the
top 25th percentile with regard to total Hg deposition and (b) exclude watersheds located in MN,
LA, SE or ME) - Results for watersheds with potentially at risk populations based on combining
both Stage 1 and Stage 2 results (Stage 3 of the 3 stage framework) 60
Table 2-15. Sensitivity analysis results based on estimating risk for watersheds with Hg fish
tissue levels based only on stationary waterbodies (i.e., excluding samples taken from flowing
waterbodies) - Results for both total risk and U.S. EGU-attributable risk 61
Table C-l. Fisher Populations Included in the Analysis for Hg Exposure and Risk 75
Table F-l. Key sources of variability associated with the analysis and degree to which they are
reflected in the design of the analysis 80
Table F-2. Key sources of uncertainty associated with the analysis, the nature of their potential
impact on risk estimates, and degree to which they are characterized 81
List of Figures
Figure 1-1 Flow Diagram of Risk Analysis Including Major Analytical Steps and Associated
Modeling Elements 21
Figure 2-1. Total mercury deposition by watershed (2005) 32
Figure 2-2. Total mercury deposition by watershed (2016) 32
Figure 2-3. U.S. EGU-attributable mercury deposition by watershed (2005) 33
Figure 2-4. U.S. EGU-attributable mercury deposition by watershed (2016) 33
Figure 2-5. Location of 2,461 watersheds with mercury fish tissue data included in the risk
assessment 38
-------
Figure 2-6. Subset of 2,170 watersheds (from the larger set of 2,461 included in the risk
assessment) located in the eastern half of the country, (this maps also illustrates limitations with
using color-coding at the watershed-level to explore trends in Hg fish tissue concentrations - see
text) 38
Figure 2-7. Total Hg fish tissue concentrations (for the 2005 Base Case) for the subset of
watersheds included in the risk assessment located in the eastern U.S 39
Figure 2-8. Total Hg fish tissue concentrations (for the 2016 Base Case) for the subset of
watersheds included in the risk assessment located in the eastern U.S 39
Figure 2-9. U.S. EGU-attributable Hg fish tissue concentrations (for the 2005 Base Case) for the
subset of watersheds included in the risk assessment located in the eastern U.S 40
Figure 2-10. U.S. EGU-attributable Hg fish tissue concentrations (for the 2016 Base Case) for
the subset of watersheds included in the risk assessment located in the eastern U.S. (Note, this
map uses the same scale as Figure 2-7, thereby supporting direct comparison between these two
time periods) 40
Figure 2-11. Top 10th percentile of watersheds based on total Hg fish tissue concentrations (for
the 2005 simulation), (ranking is based on full national set of watersheds included in the risk
assessment, but map focuses on locations in the eastern U.S.) 41
Figure 2-12. Top 10th percentile of watersheds based on total Hg fish tissue concentrations (for
the 2016 simulation), (ranking is based on full national set of watersheds included in the risk
assessment, but map focuses on locations in the eastern U.S.) 41
Figure 2-13. Top 10th percentile of watersheds based on U.S. EGU-attributable Hg fish tissue
concentrations (for the 2005 simulation), (ranking is based on full national set of watersheds
included in the risk assessment, but map focuses on locations in the eastern U.S.) 42
Figure 2-14. Top 10th percentile of watersheds based on U.S. EGU-attributable Hg fish tissue
concentrations (for the 2016 simulation), (ranking is based on full national set of watersheds
included in the risk assessment, but map focuses on locations in the eastern U.S.) 42
Figure 2-15. For the 2005 scenario, comparison of coverage of watersheds with Hg fish tissue
data (used in the risk assessment) for areas in the eastern U.S. with relatively elevated U.S. EGU-
attributable Hg deposition 46
Figure 2-16. For the 2016 Scenario, comparison of coverage of watersheds with Hg fish tissue
data (used in the risk assessment) for areas in the eastern U.S. with relatively elevated U.S. EGU-
attributable Hg deposition 46
Figure 2-17. For the 2005 scenario, plot of total Hg fish tissue concentrations versus total Hg
deposition for the 2,366 watersheds included in the risk assessment for the high-end female
consumer population 47
Figure 2-18. Cumulative distribution plots of U.S. EGU-attributable Hg deposition over the
2,366 watersheds used in modeling the high-end female consumer population as contrasted with
all 88,000 watersheds (plots provided both for the 2005 and 2016 Scenarios) 47
VI
-------
Executive Summary
The EPA has completed a national-scale risk assessment for mercury to inform the
appropriate and necessary determination for electric utility steam generating units in the United
States (U.S. EGUs), pursuant to Section 112(n)(l)(A) of the Clean Air Act (CAA). See section
III of the Preamble to the proposed U.S. EGU Toxics Rule.
This document (the "Mercury Risk TSD") describes this national-scale mercury risk
analysis. This executive summary provides an overview of the risk assessment including the
design of the risk assessment and the risk estimates that were generated. Additional technical
detail on the risk assessment design as well as more in-depth presentation and interpretation of
the risk estimates generated are provided in the body of this document.
Scope of the analysis
The national-scale risk assessment for Hg focuses on risk associated with Hg released
from U.S. EGUs that deposits to watersheds within the continental U.S., bioaccumulates in fish
and then is consumed as MeHg in fish eaten by subsistence fishers and other freshwater self-
caught fish consumers. The risk assessment considered the nature and magnitude of the risk to
public health posed by current U.S. EGU Hg emissions and the remaining risk posed by U.S.
EGU Hg emissions once CAA requirements potentially reducing Hg from U.S. EGUs are in
place. In both cases, we also assess the contribution of U.S. EGUs to potential risks from MeHg
exposure relative to total MeHg risk associated with Hg deposited by other sources both
domestic and international.
The overall design and scope of the risk assessment reflect the following factors related
to exposure to air emissions of Hg: (a) the dominant pathway associated with ambient air Hg
releases is through the consumption offish that have bioaccumulated MeHg originally deposited
to watersheds following atmospheric release and transport; (b) the primary focus in quantifying
risk associated with consumption offish containing MeHg is risk to children born to mothers
who were exposed to MeHg during pregnancy through fish consumption; (c) because U.S. EGU
Hg is likely to make a very small contribution to Hg in non-U.S. sourced bought fish consumed
in the U.S. and in bought fish sourced from further off the U.S. coast, it is not useful to assess
risks due to consumption of those bought fish;1 and (d) the type offish consumption likely to
experience the greatest contributions from U.S. EGU-sourced Hg is associated with fishing
activity at inland freshwater rivers and lakes located in regions with elevated U.S. EGU Hg
deposition.
Current conditions with regard to U.S. EGU emissions based on the 2010 ICR show HAP
emissions are closer to the 2016 emissions than to the 2005 emissions (due in part to Hg
reduction co-benefits of existing state and Federal actions).2 For this reason, in discussing risk
estimates, we focus on the 2016 results.
1 Furthermore, although commercial fish sourced closer to the U.S. coast (including estuarine areas) may have
greater U.S. EGU impacts in some cases, relative to the average U.S. EGU impact nationally, because of uncertainty
in modeling the linkage between U.S. EGU deposition and the apportionment of mercury in these fish, this
commercial consumption pathway is not included in the quantitative risk assessment.
2 A number of air quality modeling runs were completed in support of this rule. For this risk assessment, we
modeled risk for a 2005 and 2016 scenario, reflecting emissions of 52.9 and 29 tons of total mercury from U.S.
EGUs, respectively. We also developed a current estimate of Hg emissions from EGUs based on the 2010 ICR data
and that estimate was 29 tons of mercury from U.S. EGUs.
1
-------
The risk assessment calculated both the magnitude of the U.S. EGU incremental
contribution to total potential exposure and risk and the percent of total Hg exposures and risk
contributed by U.S. EGUs (i.e., the fraction of total risk associated with U.S. EGUs) to
individual watersheds for which we have fish tissue MeHg data.
Given the goal of determining whether a public health hazard is associated with U.S.
EGU emissions, we have focused this analysis on those populations likely to experience the
greatest risk when fishing at inland (freshwater) locations including subsistence-fishers.3 In
defining these high-end populations, we have included behavior that places them at greater risk
(e.g., focusing on somewhat larger fish in supplementing their diet and focusing their fishing
activity at individual watersheds - see Section 1.1 for additional detail). While these attributes
define a subset of subsistence fishers, we think that they are reasonable and that fishing
populations with these attributes are likely to exist and be active to some extent at the watersheds
included in this risk assessment. Because the general recreational angler population is likely to
experience individual risk levels well-below these high-consuming fisher populations, we have
not quantified risk for this more generalized population, although the consumption rates for
subsistence fishers are likely to be similar to consumption rates for high-end recreational anglers.
Consumption rates for the high-end fishing populations included in the risk assessment are based
on studies in the published literature, and are documented in Section 1.3 and Appendix C.
Although Hg-related risks associated with commercial fish consumption are a public
health concern, the relatively low contribution of U.S. EGU Hg to this source of dietary fish
(relative to non-U.S. Hg emissions), and the high levels of uncertainty in mapping U.S. EGU Hg
emissions to concentrations of MeHg in ocean-going fish, precludes our assessment of this
consumption pathway as part of the risk assessment. In the specific case of commercial fish
sourced from near the U.S. coast and the Great Lakes, although there is the potential for U.S.
EGUs to have a greater role in affecting Hg levels in these fish, uncertainty associated with
modeling the linkage between U.S. EGU Hg deposition and Hg exposure and risk for this dietary
pathway precludes us from including it in the risk assessment. However, it is likely that the
range of potential exposures to U.S. EGU Hg deposition across inland watersheds captures the
types of potential exposures that occur in near-coastal environments.
Risk Characterization Framework
EPA assessed risk from potential exposure to MeHg through fish consumption at a subset
of watersheds across the country for which we have measured fish tissue MeHg data. This risk
assessment uses estimates of potential exposure for subsistence fisher populations to generate
risk metrics based on comparisons of MeHg exposure to the reference dose.4 Because of
3 As noted in Section 1.2, subsistence fish consumption is defined by the EPA as individuals who rely on
noncommercial fish as a major source of their protein (USEPA, 2006). This definition does reflect a degree of
subjectivity in terms of what is meant by "major source of their protein". For this risk assessment, we consider our
high-end consumption rates (i.e., a meal every 1-2 days) as clearly subsistence.
4 We also generated estimates of the loss in intelligence quotient (IQ) points for the same populations, however, we
are focusing on the reference dose because it represents a more sensitive risk metric that potentially captures a wider
range of neurobehavioral health effects (see Section 1.2 for additional discussion). .
-------
limitations in quantifying the number of high-consumption fishers active across the set of
modeled watersheds, it is not possible to generate population-weighted distributions of risk.
For the analysis, we have developed a risk characterization framework for integrating two
types of U.S. EGU-attributable risk estimates. This framework estimates the total percent of
watersheds with fish tissue MeHg samples (approximately 2,400 out of 88,000 watersheds) that
are at risk due to potential exposures to MeHg attributable to U.S. EGU. This total percent of
watersheds where populations may be at risk from EGU-attributable Hg includes watersheds that
either have deposition of Hg from U.S. EGUs that is sufficient to lead to potential exposures that
exceed the reference dose even without considering the contributions from other U.S. and non-
U.S. sources, or have deposition of Hg from U.S. EGUs that represents a fraction (e.g., 5, 10, 15,
or 20 percent) of total Hg deposition from all sources, in watersheds where potential exposures to
MeHg from all sources (U.S. EGU, U.S. non-EGU, and non-U.S.) exceed the RfD.5
The results of the analysis include the total percent of watersheds where populations may
be at risk from EGU-attributable Hg, as well as the two component estimates. This framework
allows us to consider whether U.S. EGUs alone or in combination with other sources of Hg, pose
a potential public health hazard. The framework also allows us to evaluate the impacts of the
proposed regulation on this potential public health hazard.
Analytical Approach
Watersheds can be defined at varying levels of spatial resolution. For the purposes of this
risk analysis, we have selected to use watersheds classified using 12-digit Hydrologic Unit Codes
(HUC12) (USGS, 2009), representing a fairly refined level of spatial resolution with watersheds
generally 5 to 10 km on a side, which is consistent with research on the relationship between
changes in Hg deposition and changes in MeHg levels in aquatic biota.
After estimating total MeHg risk based on modeling consumption offish at each of these
watersheds, the ratio of U.S. EGU to total Hg deposition over each watershed (estimated using
Community Multi-scale Air Quality, CMAQ, modeling) is used to estimate the U.S. EGU-
attributable fraction of total MeHg risk. This apportionment of total risk between the U.S. EGU
fraction and the fraction associated with all other sources of Hg deposition is based on the EPA's
Office of Water's Mercury Maps (MMaps) approach that establishes a proportional relationship
between Hg deposition over a watershed and resulting fish tissue Hg levels, assuming a number
of criteria are met (USEPA, 2001). Each of the steps in the analysis is briefly described below.
Methodology for Assessing MeHg Levels in Fish Tissues
The fish tissue dataset for the risk assessment includes fish tissue Hg samples from years
1995 to 2009, with approximately 50,000 unique samples from 4,115 HUC12s out of
approximately 88,000 HUC12s in the continental U.S. The samples are more heavily focused on
locations east of the Mississippi River. For this risk assessment, a subset of the data from 2000
and later was selected, with samples distributed across 2,461 HUC12s, which provided samples
more representative of current conditions with regard to patterns of mercury deposition.
5 Any contribution of Hg emissions from EGUs to watersheds where potential exposures from total Hg deposition
exceed the RfD is a hazard to public health, but for purposes of our analyses we evaluated only those watersheds
where we determined EGUs contributed 5 percent or more to deposition to the watershed. EPA believes this is a
conservative approach given the increasing risks associated with incremental exposures above the RfD.
-------
The fish tissue samples in the master dataset come primarily from three sources: the
National Listing of Fish Advisory (NLFA) database is managed by EPA;6the U.S. Geologic
Survey (USGS), which manages a compilation of Hg datasets as part of its Environmental
Mercury Mapping and Analysis (EMMA) program, and compiles Hg fish tissue sample data
from a wide variety of sources (including the NLFA) and posts these data at
http://emmma.usgs.gov/datasets.aspx; and EPA's National River and Stream Assessment
(NRSA) study data, which includes nearly 600 fish tissue Hg samples collected at randomly
selected freshwater sites across the U.S. during the period 2008 to 2009. Additional detail on
these fish tissue MeHg data are provided in Section 1.3 and Appendix B.
Data from these three datasets were combined into a single master fish tissue dataset
covering the period 1995 to 2009. As noted above, only a fraction (2,461) of the approximately
88,000 watersheds in the continental U.S. had fish tissue concentration data and, therefore, could
be included in risk characterization. Most of the watersheds with measured fish tissue MeHg
data had multiple measurements (the average number offish tissue measurements for the period
2000 to 2009 for each of the 2,461 watersheds is 10, although some watersheds had up to 270
measurements). The assessment used the 75th percentile fish tissue value at each watershed as
the basis for exposure and risk characterization, based on the assumption that subsistence fishers
would favor larger fish which have the potential for higher bioaccumulation (i.e., use of a median
or mean value could bias low the likely MeHg levels in typically consumed fish). The 75th
percentile represents the upper bound of the interquartile range, which is generally seen as a
reasonable limit on the central tendency of a distribution. Selection of the 75th percentile
represents a reasonable assumption that acknowledges the median or mean fish may give too
much weight to smaller, less likely to be eaten fish, while avoiding assumptions that consumers
would always be able to catch and eat the largest fish with the highest MeHg levels.
Air Quality Modeling of Hg Deposition over Watersheds
Deposition of Hg was estimated using the CMAQ model v4.7.1 (www.cmaq-model.org).
The CMAQ v4.7.1 is a state of the science three-dimensional Eularian "one-atmosphere"
photochemical transport model used to estimate air quality (Byun et al., 2006, Appel et al., 2007,
Appel et al., 2008). The CMAQ simulates the formation and fate of photochemical oxidants,
ozone, primary and secondary PM concentrations, and air toxics at a 12 km gridded spatial
resolution over regional and urban spatial scales for given input sets of meteorological conditions
and emissions. Mercury oxidation pathways are represented for both the gas and aqueous phases
in addition to aqueous phase reduction reactions (Bullock et al., 2002). Because measurements
for the dry deposition of Hg do not currently exist, the modeled dry deposition performance
could not be evaluated. In EPA's view, CMAQ model wet deposition estimates agree well with
the Mercury Deposition Network (MDN) monitoring sites with a minimal seasonal bias.
Additional information on the CMAQ modeling is provided in Section 1.3 and Appendix E and
Appendix F.
CMAQ modeling at a 12 km resolution was used to estimate total annual Hg deposition
from U.S. and non-U.S. anthropogenic and natural sources over each watershed. In addition,
CMAQ simulations were conducted where U.S. EGU Hg emissions were set to zero to determine
the contribution of U.S. EGU Hg emissions to total Hg deposition. U.S. EGU-related Hg
deposition characterized at the watershed-level for the two scenarios assessed (2005 and 2016) is
6 http://water.epa.gov/scitech/swguidance/fishshellfish/fishadvisories/
-------
summarized in Table ES-1 for the complete set of 88,000 HUC12 watersheds, while Table ES-2
summarizes the percent of total Hg deposition attributable to U.S. EGUs (by percentile).
Table ES-1. Comparison of Total and U.S. EGU-Attributable Hg Deposition (ug/m2) for
the 2005 and 2016 Scenarios.*
Statistic
Mean
Median
75th percentile
90th percentile
95th percentile
99th percentile
2005 Scenario
Total Hg
Deposition
19.41
17.25
23.69
30.78
36.85
58.32
U.S. EGU-
attributable Hg
Deposition
0.89
0.24
1.07
2.38
3.60
7.77
2016 Scenario**
Total Hg
Deposition
18.66
16.59
22.83
29.90
35.16
56.23
U.S. EGU-
attributable Hg
Deposition
0.34
0.15
0.46
0.85
1.18
2.41
Statistics are based on CMAQ results interpolated to the watershed -level and are calculated using all -88,000
watersheds in the U.S.
Table ES-2. Comparison of Percent of Total Hg Deposition Attributable To U.S. EGUs for
2005 and 2016.*
Statistic
Mean
Median
75th percentile
90th percentile
95th percentile
99th percentile
2005 Scenario
5%
1%
6%
13%
18%
30%
2016 Scenario
2%
1%
3%
5%
6%
11%
* Values are based on CMAQ results interpolated to the watershed -level and reflect trends across all -88,000
watersheds in the U.S.
We note the following observations regarding estimated deposition based on information
presented in Table ES-2: (a) U.S. EGUs are estimated to contribute up to 30 percent of total Hg
deposition for the 2005 scenario and up to 11 percent for the 2016 scenario (99th percentile
values - see Table ES-2); (b) on average, U.S. EGUs contribute a substantially smaller fraction
of total Hg deposition (2 to 5 percent for the 2016 and 2005 scenarios, respectively - see Table
ES-2), this reflecting contributions made by other U.S. air emissions sources and more
importantly, by non-U.S. sources; and (c) U.S. EGU-related deposition is predicted to decrease
substantially between the 2005 and 2016 scenarios.7
7 The estimated decrease in U.S. EGU Hg emissions between 2005 and 2016 is due, in part, to decreases in SO2 and
other criteria pollutant emissions pursuant to Federal requirements and enforcement actions. If those controls were
not maintained, and Hg emissions were to increase from current levels, the U.S. EGU attributable deposition, and
fraction of deposition would be somewhere between the 2005 scenario and 2016 scenario.
-------
EPA also evaluated the potential for "hot spot" deposition near U.S. EGU emission
sources on a national scale, based on the CMAQ modeled Hg deposition for the 2005 and 2016
scenarios. We calculated the excess deposition within 50 km of U.S. EGU sources by first
calculating the average U.S. EGU attributable Hg deposition within a 500 km radius around the
U.S. EGU source. This deposition represents the likely regional contribution around the EGU.
We then calculated the average U.S. EGU attributable Hg deposition within 50 km of the U.S.
EGUs to characterize local deposition plus regional deposition near the EGU. Excess local
deposition is then the 50 km radius average deposition minus the 500 km radius average
deposition. Summary statistics for the excess local deposition are provided in Table ES-3. Table
ES-3 shows both the mean excess deposition around all U.S. EGUs, and the mean excess
deposition around just the top 10 percent of Hg emitting U.S. EGUs. Table ES-3 also shows the
excess Hg deposition as a percent of the average regional deposition to provide context for the
magnitude of the local excess deposition. In 2005, for all U.S. EGU, the excess was around 120
percent of the average deposition, while for the top 10 percent of Hg emitting U.S. EGU, local
deposition was around 3.5 times the regional average. By 2016, the absolute excess deposition
falls, however, the local excess still remains around 3 times the regional average for the highest
10 percent of Hg emitting U.S. EGUs (note, additional detail on this hot spot assessment is not
provided in the main body of the TSD, since it is clearly laid out here).
Table ES-3. Excess Local Deposition of Hg Based on 2005 CMAQ Modeled Hg Deposition
Category of Results
All U.S. EGU sites with Hg
emissions >0
(672 sites)
Top ten percent U.S. EGU in Hg
emissions
(67 sites)
50km-Radius-Average Excess Local Deposition values
fag/m2)
Mean Across EGUs (percent of regional average deposition)
2005 Scenario
1.65(119%)
4.89 (352%)
2016 Scenario
0.36 (93%)
1.18(302%)
This analysis shows that there is excess deposition of Hg in the local areas around EGUs,
especially those with high Hg emissions. Although this is not necessarily indicative of higher
risk of adverse effects from consumption of MeHg contaminated fish from watersheds around
the U.S. EGUs, it does indicate an increased chance that Hg from U.S. EGUs will impact local
watersheds.
Estimating the Proportion of Total Hg Exposure Associated with U.S. EGUs and
Projecting Changes in Fish Hg concentrations
The MMaps approach specifies that, under certain conditions (e.g., Hg deposition is the
primary loading to a watershed and near steady-state conditions have been reached), a fractional
change in Hg deposition to a watershed will ultimately be reflected in a matching proportional
change in the levels of MeHg in fish. This proportionality assumption between deposition
changes and fish tissue MeHg concentrations can be used to both estimate the portion of total Hg
exposure that is associated with U.S. EGUs and project changes in fish Hg concentrations (and
-------
consequently total exposure) associated with changes in total Hg deposition in the future.8 This
assumption holds in watersheds where air deposition is the primary source of Hg loadings, and as
a result, watersheds where this is not the case are removed from the risk analysis. MMaps is
discussed in greater detail in Section 1.3 and in Appendix E and Appendix F.
For the 2005 analysis, CMAQ modeling results for a particular watershed allow U.S. to
estimate the proportion of total exposure (estimated for that watershed) that is associated with
U.S. EGU deposition (i.e., based on the ratio of U.S. EGU Hg deposition to total Hg deposition
over the watershed as specified by the MMaps approach). In the case of the 2016 future
simulation, we can first project changes in total fish tissue Hg levels (for that watershed) by
comparing estimates of total Hg deposition in 2005 to estimates for 2016 generated by CMAQ
and then again, apportion that adjusted total risk between U.S. EGUs and all other sources, based
on comparing U.S. EGU Hg deposition to total Hg deposition in 2016.
Patterns of U.S. EGU-attributable fish tissue MeHg concentrations are summarized in
Tables ES-4 and ES-5. Table ES-4 compares total and U.S. EGU-attributable fish tissue MeHg
concentrations for the 2016 and 2005 scenarios by watershed percentile (including the percent
reduction between 2005 and 2016). Table ES-5 summarizes the percent of total fish tissue
MeHg concentrations attributable to U.S. EGUs (also by watershed percentile).
Table ES-4. Comparison of Total and U.S. EGU-Attributable Fish Tissue MeHg
Concentrations for the 2005 and 2016 Scenarios
Statistic
mean
50th Percentile
75th Percentile
90th Percentile
95th Percentile
99th Percentile
Fish tissue MeHg concentration (ppm)
2005 Scenario
Total
0.31
0.23
0.39
0.67
0.91
1.34
U.S. EGU-
attributable
0.024
0.014
0.032
0.056
0.079
0.150
2016 Scenario
Total
0.29
0.20
0.36
0.63
0.87
1.29
U.S. EGU-
attributable
0.008
0.005
0.011
0.019
0.026
0.047
% change (2016
versus 2005) in fish
tissue MeHg
concentration
Total
-6.7%
-10.0%
-6.4%
-5.9%
-4.7%
-3.7%
U.S. EGU-
attributable
-65%
-61%
-67%
-66%
-67%
-68%
Table ES-5. Comparison of U.S. EGU Fraction of Total Fish Tissue MeHg Levels*
Statistic
Mean
50th Percentile
75th Percentile
U.S. EGU-attributable percent of total fish tissue MeHg levels
2005 Scenario
9%
6%
14%
2016 Scenario
4%
3%
5%
The risk assessment estimates risk for future points in time once near steady state conditions have been reached
following simulated changes in mercury deposition and does not attempt to simulate the temporal profile of that
response. As noted in Section 1.3 and Appendix E, the amount of time required for MeHg levels in fish to fully
respond following a change in mercury deposition can range from years to decades depending on the nature of the
watershed involved (e.g., methylation potential, importance of watershed runoff to load mercury into the
waterbody).
7
-------
Statistic
90th Percentile
95th Percentile
99th Percentile
U.S. EGU-attributable percent of total fish tissue MeHg levels
2005 Scenario
20%
26%
40%
2016 Scenario
7%
9%
18%
* These values are specifically for the 2,461 watersheds included in the risk assessment.
We note the following observations regarding fish tissue MeHg levels based on
information presented in Tables ES-4 and ES-5: (a) U.S. EGUs can contribute up to 18 percent
of total fish tissue MeHg levels for a subset of the watersheds with fish tissue data (99th
percentile watershed for the 2016 Scenario - see Table ES-4); (b) on average, U.S. EGUs
contribute 4 percent of total fish tissue MeHg levels (for the 2016 Scenario); and (c) reflecting
the pattern seen with Hg deposition, the U.S. EGU-attributable fraction offish tissue MeHg
levels is estimated to decrease significantly between the 2005 and 2016 scenarios.
Defining Subsistence Fisher Scenarios
As discussed above, this analysis focuses on higher-consumption self-caught fisher
populations active at inland freshwater locations, because these populations are expected to
experience the greatest U.S. EGU-attributable risks. Therefore, in reviewing studies of fishing
populations, emphasis was placed on identifying surveys of higher consumption fishing
populations active at inland freshwater rivers and lakes within the continental U.S. Information
on the studies used to develop the high end fish consumption scenarios for the risk analysis is
provided in Section 1.3 and in Appendix C.
Based on EPA's review of the fish consumption literature, EPA defined consumption
rates for the subsistence fisher populations modeled across the 2,461 watersheds included in the
risk assessment. We used the studies referenced above as a guide to characterize high-end
consumption behavior for a scenario that could be assessed broadly across the 2,461 watersheds.
Generally all of the studies identified high-end percentile consumption rates (90th to 99th
percentiles for the populations surveyed) ranging from approximately one fish meal every few
days to a larger fish meal very day (i.e., 120 grams per day (g/day) to greater than 500 g/day fish
consumption). We used this trend across the studies to support application of a generalized
female high-end fish consumption scenario (high-end female consumer scenario) across most of
the 2,461 watersheds.9
Consumption rates for this high-end female consumer were based on values presented for
female fishers in South Carolina. Values from the South Carolina study were used because they
specifically covered high-end consumption by women, which is the population of interest in this
risk assessment. Furthermore, the consumption rates identified in the South Carolina study (123
g/day to 373 g/day for the 90th and 99th percentiles, respectively) are in the range of values seen
across the other studies reviewed in designing this analysis and therefore are considered to be
generally representative of subsistence consumption (these consumption rates translate into
approximately one 8oz fish meal every other day).
9 Reflecting the fact that higher levels of serf-caught fish consumption (approaching subsistence) have been
associated with poorer populations, we only assessed this generalized high-end female consumer scenario at those
watersheds located in U.S. Census tracts with at least 25 individuals living below the poverty line (this included the
vast majority of the 2,461 watersheds and only a handful were excluded due to this criterion).
-------
In addition to the studies used to define the high-end fish consuming populations
modeled in the risk assessment, we also reviewed a large number of additional studies
characterizing higher-level self-caught fish consumption in the U.S. While these studies had
limitations that prevented their use as the basis for defining high-end fisher scenarios to include
in the analysis, in several instances, they did support the levels of self-caught fish consumption
modeled in the analysis. These studies are described in Appendix C.
Risk Related to Exposure to MeHg in Fish and Assessment of Contribution of U.S. EGUs to
MeHg Exposure and Risk
This section provides an overview of risk estimates generated for the 2,461 watersheds
included in the risk assessment.10 As noted above, we focus on risk estimates generated for the
high-end female consumer assessed at the national level, since this population, as defined for this
analysis, provides the most comprehensive coverage for watersheds with fish tissue MeHg data
across the U.S. and because the consumption rates used to model this population represent
subsistence levels that would characterize populations most likely to experience high levels of
exposure to MeHg and thus experience higher risk.
We estimated total exposure to MeHg at each of the 2,461 watersheds.11 Estimates of
total Hg exposure were generated by combining 75th percentile fish tissue values with the
consumption rates for female subsistence fishers. A cooking loss factor (actually reflecting the
fact that the preparation offish can result in increased Hg concentrations) was also included in
exposure calculations (Morgan et al., 2007).
We estimate the fraction of total potential exposure that is associated with U.S. EGU Hg
deposition at each watershed using the proportionality assumption supported by MMaps (see
above). Once total potential exposure to MeHg has been estimated and the U.S. EGU-
attributable portion of that exposure has been estimated, we then estimate risk based on
exposures above the RfD, both due to total MeHg, and MeHg attributable to U.S. EGU without
consideration of other U.S. andnon-U.S. sources.
A summary of risk estimates is presented here and detailed summaries of the risk
estimates are presented in Section 2.6. Our estimates of total percent of watersheds where
populations may be at risk from EGU-attributable Hg are as high as 28 percent.12 The upper end
estimate of 28 percent of watersheds reflects the 99th percentile fish consumption rate for that
population modeled, and a benchmark of 5 percent U.S. EGU contribution to total Hg deposition
in the watershed. Any contribution of Hg emissions from EGUs to watersheds where potential
10 Each fish consumption rate scenario was assessed for a subset of the 2,461 watersheds based on consideration for
where "source populations" for each scenario were located (i.e., a watershed was modeled if it fell within a U.S.
Census tract containing the source population for a particular fisher scenario). The high-end female consumption
rate scenario is the scenario with the broadest spatial coverage since it was applied to all watersheds intersecting
U.S. Census tracts with at least 25 poor white individuals (i.e., the "source population" for high-end female self-
caught fish consumers). This meant that this scenario was assessed for 2,366 of the 2,461 watersheds with fish tissue
MeHg data.
11 As noted earlier, each high-end fish consuming population included in the analysis was assessed for a subset of
these watersheds, depending on which of those watersheds intersected a U.S. Census tract containing a "source
population" for that fish consuming population. Of the populations assessed, the high-end female consumer scenario
was assessed for the largest portion (2,366) of the 2,461 watersheds.
12 As noted earlier, the determination of whether U.S. EGUs make a significant contribution to total Hg deposition
is only considered for watersheds where total risk is considered to represent a potential hazard to public health (e.g.,
Potential MeHg exposure from all U.S. and non-U.S. sources exceeds the RfD).
-------
exposures from total Hg deposition exceed the RfD is a hazard to public health, but for purposes
of our analyses we evaluated only those watersheds where we determined EGUs contributed 5
percent or more to deposition to the watershed. EPA believes this is a conservative approach
given the increasing risks associated with incremental exposures above the MeHg RfD. Scenario
The percent of where populations may be at risk from EGU-attributable Hg based on the 2010
ICR would be somewhat higher due to the greater level of Hg emissions in that case (35 tons in
2010 compared with 29 tons in 2016).
Of the total percent of watersheds where populations may be at risk from EGU-
attributable Hg, we estimate that up to 22 percent of watersheds included in this analysis could
be potentially at risk based on consideration of the U.S. EGU attributable fraction (e.g., 5, 10, 15,
or 20 percent) of total Hg deposition over watersheds with total risk judged to represent a public
health hazard (MeHg total exposure greater than the RfD). The 22 percent estimate is also
generated for the 2016 scenario and is based on the same assumptions used in the estimate of
total risk provided earlier (i.e., 99th percentile fish consumption rate, and a 5 percent U.S. EGU
contribution to total Hg deposition over a given watershed). We do note however, that,
specifically with regard to the HQ estimates, any contribution of mercury from EGUs to
watersheds with exposures exceeding the MeHg RfD represents a potential hazard to public
health, but for purposes of this analysis we have focused on those waterbodies where we
determined EGUs contributed 5% or more to the hazard. We think this is a conservative
approach given the increasing risks associated with incremental exposures above the MeHg RfD.
Of the total percent of watersheds where populations may be at risk from EGU-
attributable Hg, we estimate that up to 12 percent of the watersheds could potentially be at risk
based on watersheds where the U.S. EGU incremental contribution to exposure exceed the
MeHg RfD, without consideration of contributions to exposures from U.S. non-EGU and non-
U.S. sources. The upper end estimate of 12 percent of watersheds is based on the 2016 Scenario
and reflects a scenario using the 99th percentile fish consumption rate.
The two component estimates of percent of watersheds where populations may be at risk
from EGU-attributable Hg do not sum to the total percent watersheds where populations may be
at risk from EGU-attributable Hg of 28 percent due to overlap in the risk estimates - some
watersheds where U.S. EGUs contribute greater than 5 percent to total Hg deposition also have
U.S. EGU attributable exposures that exceed the RfD without consideration of exposures from
other U.S. and non-U.S. Hg sources.
The percentage of watersheds where U.S. EGUs contribute to exposures of concern
increases dramatically as we consider higher fish consumption scenarios. Exposures based on
the 99th percentile consumption rate represent close to maximum potential individual risk
estimates. These consumption rates are based on data reported by fishers in surveys, and, thus,
represent actual consumption rates in U.S. populations. However, EPA does not have data on
the locations where these high self-caught fish consuming populations reside and fish, and as a
result, there is also increased uncertainty about the percent of watersheds that might experience
potential exposures at the highest levels.
With regard to the other fisher populations included in the full risk assessment
(Vietnamese, Laotians, Hispanics, blacks and whites in the southeast, and tribes in the vicinity of
the Great Lakes), our risk estimates suggests that the high-end female consumer assessed at the
10
-------
national-level generally provides coverage (in terms of magnitude of risk) for all of these fisher
populations except blacks and whites in the southeast.13
Variability and Uncertainty (Including Discussion of Sensitivity Analyses)
The risk assessment has been designed to reflect consideration for key sources of
variability associated with the exposure scenario (e.g., spatial pattern of total and U.S. EGU-
related Hg deposition, spatial variation in fish tissue MeHg levels, variation in the location and
behavior of high-consuming fisher populations). The degree to which critical sources of
variability have been reflected in the design of the analysis is discussed in Appendix F, Table F-
1.
Key sources of uncertainty potentially impacting the risk analysis include: (a)
uncertainty in predicting Hg deposition over watersheds using CMAQ; (b) uncertainty in
predicting which watersheds will be subject to high-end fishing activity an the nature of that
activity (e.g., frequency of repeated activity at a given watershed and the types/sizes offish
caught); and (c) uncertainty in using MMaps to apportion exposure and risk between different
sources including U.S. EGUs and predict changes in fish tissue MeHg levels for future scenarios.
We describe key sources of uncertainty impacting the risk analysis, including their potential
impact on the risk estimates and the degree to which their potential impact is characterized as
part of the analysis in Appendix F, Table F-2.
As part of the risk assessment, we have also completed a number of sensitivity analyses
focused on exploring the impact of uncertainty related to the application of the MMaps approach
in apportioning exposure and risk estimates between sources (U.S. EGU and total) and in
predicting changes in fish tissue MeHg levels. These sensitivity analyses have explored: (a)
concerns over including watersheds that may be disproportionately impacted by non-air Hg
sources;14 and (b) concern that the MMaps approach may be more representative when applied to
stationary waterbodies (in the analysis, the MMaps was applied to watersheds including a
mixture of flowing and stationary waterbodies). The results of the sensitivity analyses, when
considered in aggregate, suggest that uncertainties due to application of MMaps are unlikely to
have a substantial effect on the risk estimates discussed here.
Key Observations
The following key observations result from consideration of the risk estimates generated
(additional detail on these observations is presented in Section 2.8):
• Reflecting current emissions, U.S. EGUs can contribute up to 11 percent of total Hg
deposition (for the 99th percentile watershed in the 2016 Scenario). U.S. EGUs (for the 2016
scenario) contribute on average, about 2 percent of total Hg deposition across the country.
13 Specifically, upper percentile risk estimates for the high-end female consumer assessed at the national level were
notably higher than matching percentile estimates for the Hmong, Vietnamese, Hispanic and Tribal populations. By
contrast, risk estimates for whites in the southeast were somewhat higher than the high-end female consumer, while
risk estimates for blacks in the southeast were notably higher (see summary of risk estimates in the TSD supporting
the A&N Determination (see section 2.6.1 for additional detail on these risk estimates).
14 In addition to non-air Hg sources of loadings, some regions of concern may also have longer lag period
associated with the linkage between Hg deposition such that the fish tissue MeHg levels we are using are actually
associated with older historical Hg deposition patterns.
11
-------
The average U.S. EGU deposition decreased from approximately 5 percent of total to
approximately 2 percent of total for the 2005 and 2016 Scenarios, respectively.
• Although U.S. EGUs contribute on average, about 2 percent of total Hg deposition for the
2016 Scenario, they contribute about 4 percent of the fish tissue MeHg levels at watersheds
included in this analysis. This discrepancy reflects the fact that fish tissue MeHg sampling is
focused in the eastern half of the country which has higher U.S. EGU Hg deposition levels,
compared with the national average, and, therefore, the fraction of MeHg in fish tissue
attributable to U.S. EGUs will also be larger given that it is driven by estimates of U.S. EGU
Hg deposition over these watersheds. U.S. EGUs are estimated to contribute up to 18 percent
offish tissue MeHg levels in the 2016 scenario (for the 99th percentile watershed).
• Comparing the pattern of U.S. EGU-attributable Hg deposition with watersheds containing
fish tissue MeHg data results in our concluding that, while we have some degree of coverage
for high U.S. EGU impact areas, this coverage is limited. For this reason, we believe that the
actual number of where populations may be at risk from EGU-attributable Hg (i.e.,
watersheds where U.S. EGUs could contribute to a public health hazard) could be
substantially larger than estimated.
• We estimate that up to 28 percent of the watersheds included in this risk assessment could
have populations potentially at-risk under the 2016 scenario. This total risk estimate reflects
a combination of watersheds where the U.S. EGU incremental contribution alone is
considered to represent a potential public health hazard or where U.S. EGUs make at least a
5% contribution to total Hg deposition over watersheds where total risk is considered to pose
a public health hazard. The 28 percent total risk estimate is also based on application of the
99th percentile consumption rate for the high-end female consumer.
• Reductions in U.S. EGU-attributable Hg will reduce the magnitude of US EGU-attributable
risk, although substantial risk from Hg deposition will likely remain as a results of these
sources.
• Sensitivity analyses were conducted primarily to examine uncertainty in applying the MMaps
approach for linking Hg deposition to fish tissue MeHg levels. These analyses suggest that
uncertainty related to the MMaps approach is unlikely to substantially affect our assessment
of the public health hazard posed by Hg emissions from U.S. EGUs.
12
-------
1. Review of Analysis Approach
1.1. Purpose and Scope of Analysis
This document (the "Mercury Risk TSD") describes the national-scale risk assessment
for mercury completed to inform the appropriate and necessary determination for electric utility
steam generating units in the United States (U.S. EGUs), pursuant to Section 112(n)(l)(A) of the
Clean Air Act (CAA). See Section III of the preamble to the proposed U.S. EGU Toxics Rule.
This risk assessment focuses on risk associated with mercury released from U.S. EGUs that has
deposited to freshwater watersheds within the continental U.S., bioaccumulated in fish and is
consumed as methylmercury (MeHg) in dietary fish by the public.
The following policy-related questions were developed to help guide the design of the
risk assessment: (a) what is the nature and magnitude of the potential risk to public health posed
by current U.S. EGU mercury emissions, (b) what is the nature and magnitude of the potential
risk posed by U.S. EGU mercury emissions in 2016 considering potential reductions in EGU Hg
emissions attributable to CAA requirements,15 and (c) how is risk estimated for both the current
and future scenario apportioned between the incremental contribution from U.S. EGUs and other
sources of mercury? The last policy-related question reflects the fact that mercury emitted from
U.S. EGUs does not result in a distinct and isolated exposure pathway, but rather is combined
with mercury emitted from other sources (domestic and international) in contaminating fish.
Therefore, to consider U.S. EGU contributions to exposure and risk associated with the
consumption offish containing MeHg, we determine what share of total exposure is attributable
to U.S. EGUs.
In addition to the above policy-related questions, the overall design and scope of the risk
assessment reflects consideration of important technical factors related to air-sourced mercury,
including, in particular, mercury released from U.S. EGUs (Note, a number of these technical
factors are discussed in greater length, including provision of relevant citations, later in Section
1.3):
• While mercury exposure and risk can occur through a variety of pathways, the dominant
pathway associated with ambient air releases is through the consumption offish that have
bioaccumulated mercury originally deposited to watersheds following atmospheric
release and transport. Deposition of mercury to watersheds includes mercury originating
from local/regional sources, combined with mercury that has been transported over
greater distances, including mercury released outside of the U.S.. Generally oxidized
(divalent) and particle-bound mercury will deposit relatively closer to the release source,
while elemental mercury will travel further, often becoming part of the global pool,
before being deposited.16
15 For purposes of this analysis, we focus on 2016 as this is the first year after compliance would be required to
occur.
16 Mercury is a persistent, bioaccumulative toxic metal that is emitted from power plants in three forms: Gaseous
elemental Hg (Hg°), oxidized Hg compounds (Hg+2), and particle-bound Hg (HgP). Elemental Hg does not quickly
deposit or chemically react in the atmosphere, resulting in residence times that are long enough to contribute to
13
-------
• Available information supporting the quantification of mercury-related health effects
provides the strongest support for modeling neurological deficits in children who were
exposed to MeHg during pregnancy through maternal fish consumption.17
• U.S. EGU emitted mercury is likely to make a very small contribution to MeHg in
foreign-sourced commercial fish consumed in the U.S. and in commercial fish sourced
form further off the U.S. coast. Therefore, the risk assessment, while acknowledging
these sources of exposure to U.S. EGU-sourced mercury, does not quantify these risks
since the U.S. EGU-attributable portion of these risks is likely to be very small and any
quantitative estimates of U.S. EGU-attributable risk would be highly uncertain.18
• While areas closer to the US coast (including estuarine areas) and from the Great Lakes
may have elevated U.S. EGU impacts in some cases, because of uncertainty in modeling
the linkage between U.S. EGU deposition and the apportionment of mercury in fish, we
have not included this commercial consumption pathway in the quantitative risk
assessment.19
• The type offish consumption likely to experience the greatest contributions from U.S.
EGU-sourced mercury is associated with fishing activity at inland freshwater rivers and
lakes located in regions experiencing relatively elevated U.S. EGU mercury deposition.
While the average U.S. EGU percent contribution to total mercury deposition in the U.S.
in 2005 was estimated at -5% and in 2016 at -2%, some watersheds had U.S. EGU
contributions ranging up to 30% and higher in 2005 (see Section 2.3). Therefore, efforts
to identify areas with likely high U.S. EGU attributable MeHg exposures and risk are
focused on assessing risk for those areas which have (a) relatively elevated fish tissue
MeHg levels and (b) relatively elevated levels of U.S. EGU mercury deposition (with the
assumption that the elevated U.S. EGU deposition in these regions translates into a larger
global scale deposition. Hg(2+) and Hg(p) deposit quickly from the atmosphere impacting local and regional areas
in proximity to sources.
17 The EPA's health benchmark for methylmercury exposure (the reference dose or RfD) is based on three
epidemiological studies. These studies relate hair mercury levels in mothers (a surrogate for exposure during
pregnancy) or mercury in cord blood (a direct measure of fetal exposure) to deficits in children's performance on a
range of neuro-cognitive tests (see section 1.3).
18 While mercury released from U.S. EGUs does contribute to contamination of foreign-sourced commercial fish,
the fraction contributed by U.S. EGUs is extremely small. Current estimates of U.S. EGU mercury emissions are
-29 tons per year (see section 2.2), compared with global anthropogenic mercury emissions (for 2005), excluding
biomass burning, estimated at approximately 2,120 tons with a range of 1,347 to 3,255 tons/year (Pirrone et al.,
2010, UNEP, 2010). Based on these estimates, we would expect U.S. EGUs to contribute less than 1% of the
mercury in commercially (foreign) sourced fish. Therefore, particularly in the context of estimating individual risk,
U.S. EGU contributions to risk that residents in the US experience through consumption of foreign-sourced
commercial fish is expected to be to small to characterize. This observation would also likely hold for the U.S. EGU
contribution to commercial fish sourced from further off the U.S. coast, where total mercury loading is likely to also
be dominated by non-US anthropogenic emissions which are globally transported.
19 While air quality modeling does suggest that some near coastal areas (e.g. the Chesapeake Bay) and portions of
the Great Lakes may have elevated U.S. EGU deposition relative to the average levels in the continental U.S., a
number of factors make the simulation of the linkage between mercury deposition and fish tissue MeHg levels in
these near-coastal areas and the Great Lakes challenging and uncertain. Specifically, the size of the waterbody
involved (i.e., inner coastal waterways, near coastal areas and the Great Lakes) combined with the potential for fish
to have larger habitats in these locations, relative to inland lakes and rivers, means we cannot adequately quantify
the EGU contribution to fish tissue MeHg levels. Given the greater uncertainty associated with simulating the
linkage between near coastal U.S. EGU deposition and fish tissue MeHg levels, we have elected not to simulate this
pathway in the risk assessment.
14
-------
fraction of the MeHg levels being attributed to this source category). Furthermore, while
recreational angler activity is likely to occur broadly across the U.S., high self-caught fish
consumers (i.e., the subsistence populations) will experience the greatest degree of U.S.
EGU-attributable risk if they are active at high U.S. EGU-impact watersheds.
In consideration of the policy questions and technical factors discussed above, the
national-scale Hg risk assessment was designed as follows.
• Evaluate risk for two scenarios - a 2005 Scenario and a 2016 Scenario. Risk is estimated
both for a 2005 Scenario and a 2016 Scenario, the latter reflecting consideration of
potential HAP emission reductions from CAA requirements. The latest emissions data
(see Section 2.2) suggest that current 2010 U.S. EGU emissions are closer to levels
reflected in the 2016 Scenario and substantially lower than levels reflected in the 2005
Scenario. As a result, the 2016 Scenario analysis is most relevant for this rulemaking.
Further modeling of future emissions indicates that in the absence of binding federal
regulations U.S. EGU emissions are not likely to be substantially reduced between 2010
and 2016, as the CAA directs the Agency to consider only Federal CAA requirements in
estimating future HAP emissions (and attendant risks) associated with EGUs. Thus, we
conclude that if we find there exists a public health hazard from current U.S. EGU
mercury emissions based on the 2016 Scenario, we will also find that a public health
hazard will continue to exist in 2016.
• Include estimates of total (all mercury deposition sources) risk as well as the U.S. EGU
incremental contribution to total risk. As discussed below (Section 1.2), we focus on two
aspects of MeHg-related risk: (a) total mercury risk with an estimate of the percent of that
total risk contributed by U.S. EGUs (i.e., the fraction of total risk associated with U.S.
EGUs) and (b) risk when deposition from U.S. EGUs is considered before taking into
account deposition and exposures resulting from other sources of Hg. These two risk
metrics reflect the cumulative burden of mercury exposures and incremental contribution
that the U.S. EGU attributable deposition makes to the overall exposures to MeHg.20
• Focus on assessing risk to subsistence fishers active at inland watersheds. Given the goal
of determining whether a public health hazard is associated with U.S. EGU emissions, we
have assessed risk for a set of subsistence populations active at inland (freshwater)
watersheds. By focusing on inland watersheds, we are focusing on those locations with
the greatest U.S. EGU-attributable mercury deposition and consequently the greatest U.S.
EGU-attributable fish tissue MeHg levels. Furthermore, by focusing on subsistence fisher
scenarios, we focus on those self-caught fish consumers with the highest intake rates and
therefore, those who will experience the greatest MeHg exposures at a given watershed.
In defining the high-end fisher populations to include in the analysis, we have used peer-
reviewed study data characterizing behavior for a variety of high-end fisher populations
20 When exposures are to be compared to the EPA's reference dose (RfD) for MeHg in order to generate a hazard
quotient (HQ), we must first consider total MeHg exposure given the definition of the RfD, which is intended to be
compared against total exposure to a given hazardous air pollutant. Once an HQ reflecting total exposure is
calculated, we can then consider the U.S. EGU incremental contribution to that total risk. However, U.S. EGU
incremental risk in the form of an HQ should not be considered in isolation without placing it in context with regard
to risk associated with total MeHg in the fish being consumed.
15
-------
active in different regions of the country (e.g., Laotians, Great Lakes Tribal populations,
Black and White anglers active in the Southeast - see Section 1.3).
Generate watershed-level estimates of risk for a representative fisher for each fisher
population (these estimates are not population-weighted): In modeling risk for these
populations, we generate watershed-level estimates based on the subset of watersheds in
the U.S. where we have fish tissue MeHg data and where we believe a given high-end
fisher population could be active.21 Because it is not possible to enumerate these high-end
fisher populations, we cannot develop population-weighted risk distributions.22
Therefore, in modeling risk, we generate a risk estimate for each high-end fisher
population for each watershed where we believe that population could be active. We are
then able to generate percentile risk estimates, based on the set of watershed-level risk
estimates generated for each fisher population (i.e., assuming each watershed gets equal
weight in deriving that risk distribution).
Exclude commercial fish consumption from the quantitative risk analysis. Although risk
associated with commercial fish consumption may be a potential public health concern
under certain circumstances, the relatively low contribution of U.S. EGU mercury to this
source of dietary fish (relative to non-US mercury emissions), leads us to exclude this
consumption pathway from the risk assessment. In the specific case of commercial fish
sourced from near the US coast (e.g. Chesapeake Bay) and the Great Lakes, while there is
the potential for U.S. EGUs to have a greater role in effecting mercury levels in these
fish, as noted earlier, uncertainty associated with modeling the linkage between U.S.
EGU mercury deposition and mercury exposure and risk for this dietary pathway
precludes us from including this pathway in the risk assessment.
21 The potential for a high-end fisher population to be active at a given watershed is based on consideration for
whether members of the demographic group from which that fishing population originates are located in the US
Census tract(s) intersecting that watershed. For example, if we are considering Hispanic high-end fishers, we would
only assess that scenario at watersheds located in U.S. Census tracts with at least 25 poor Hispanics (in this case,
poor Hispanics represent a "source population" for this category of high-end fisher- see Section 1.3 for additional
detail).
22 In order to enumerate risk estimates generated for the female high-end consumer scenario used in this risk
assessment, we would need to have the following types of specific information: (a) the fraction of anglers who
consume at the subsistence-levels modeled for this population specifically at inland freshwater waterbodies, (b) for
this population, the fraction that focus their activity at individual watersheds, and target somewhat larger fish to
supplement their diet, and (c) for this subgroup, the fraction of consumers of childbearing age who either fish
themselves and consume at this level, or are associated with male fishers who fish at this level (with that female in
turn consuming at a subsistence rates). However, currently available information does not allow us to estimate each
of these subgroups of high-end fishers. Specifically, while we have data on frequency of recreational angling within
the U.S., this covers general recreational fishing and not subsistence fishing. Furthermore, there are concerns as to
whether surveys of recreational activity would effectively capture subsistence fishers who include poorer individuals
who traditionally have lower survey response rates. We do have surveys like the Burger et al., 2002 study which
provide fish consumption rates for percentiles of the survey populations, with the upper percentiles (i.e., 95th and
99th percentiles) approaching subsistence levels (we could interpret this as suggesting that 1 to 5% of the surveyed
fishing population at these shows consumes at a subsistence level). However, we would still be concerned as to the
degree that this type of fishing show population accurately captures rates of subsistence fishing by poorer
individuals who may not frequent a show like this in proportion to their prevalence in the general population. And
finally, we have focused on a subset of female subsistence consumers that we believe (a) are reasonably likely to
exist at a subset of our watersheds and (b) are likely to experience higher risk due to their behavior (i.e., favor larger
fish as a dietary source, focus their activity at individual watersheds). While we believe it is reasonable to assume
that a subset of high-end fishers would have these attributes, this introduces additional uncertainty into any effort to
enumerate this female high-end consuming population.
16
-------
1.2. Overview of Risk Metrics and the Risk Characterization Framework
The risk assessment uses estimates of exposure for subsistence fisher populations23 to
generate two categories of risk metrics including (a) IQ loss in children born to mothers from
these high-consuming fisher populations and (b) hazard quotient (HQ) estimates generated by
comparing exposure estimates for these populations to the MeHg RfD. As discussed in greater
detail in Section 1.3, risk estimates (for these populations) are generated for the subset of
watersheds in the US where we have sampled fish tissue data. Because of limitations in
quantifying the number of high-consumption fishers active across the set of modeled watersheds,
it is not possible to generate population-weighted distributions of risk (see previous section).
However, we use the watershed-level estimates of risk to consider the percentage of watersheds
modeled that fall within specific risk ranges.
Both the IQ loss and HQ risk metrics are further stratified to consider both total risk as
well as U.S. EGU-attributable risk. In considering U.S. EGU-attributable risk we generate two
types of risk estimates:
• The percent or fraction of total risk at a given watershed that is associated with U.S.
EGU's. We consider the magnitude of total risk (IQ loss or HQ) and then estimate the
fraction (or percent) of that total risk estimates that is attributable to U.S. EGUs.
• Risk when deposition from U.S. EGUs is considered before taking into account
deposition and exposures resulting from other sources: Here, we estimate risk based on
the U.S. EGU incremental contribution to total exposure. Specifically, for IQ loss, we are
using the U.S. EGU-portion of exposure at each watershed to generate an estimate of
U.S. EGU attributable IQ loss and for HQ, we are comparing U.S. EGU-attributable
exposure against the MeHg RfD.
In assessing the potential public health significance of the IQ loss risk estimates, based on
recommendations provided by the Clean Air Science Advisory Committee (CASAC) in the
context of the last National Ambient Air Quality Standard (NAAQS) review for lead completed
in 2008 (US EPA, 2007a), we interpreted IQ loss estimates of 1-2 points as being clearly of
public health significance. With regard to HQ estimates based on the MeHg RfD, we considered
exposures above the RfD to represent a potential public health hazard.24 Because HQ estimates,
by convention, have been reported to one significant figure, reflecting precision in the underlying
23 Subsistence fishers are individuals who rely on noncommercial fish as a major source of protein (US EPA, 2000).
For purposes of this risk assessment, we have interpreted this as representing self-caught fish consumption ranging
from a fish meal (8 ounce) every few days to a large fish meal (12 ounces or more) every day.
24 EPA's interpretation for this assessments is that any exposures to MeHg above the RfD are of concern given the
nature of the data available for mercury that is not available for many other chemicals, where exposures have often
had to be significantly above the RfD before they might be considered as causing a hazard to public health. The
scientific basis for the mercury RfD includes extensive human data and extensive data on sensitive subpopulations
including pregnant mothers; therefore, the RfD does not include extrapolations from animals to humans, and from
the general population to sensitive subpopulations. In addition, there is no evidence for a threshold observed for
critical effect of neurological deficits in children studied in the prinicipal studies of the IRIS assessment for MeHg.
This additional confidence in the basis for the RfD suggests that all exposures above the RfD can be interpreted with
more confidence as causing a potential hazard to public health.
17
-------
RfD, mathematically, we considered exposures that were at least 1.5 times the RfD (i.e., an HQ >
1.5), to represent a potential public health hazard since these would round to an HQ of 2.
Risk Characterization Framework
We have developed a 3-stage framework for integrating the risk metrics described above
(throughout this document, this will be referred to as the "risk characterization framework"):
Stage 1 - consider various degrees of U.S. EGU contribution to total risk at watersheds
where total risk is considered to pose a potential public health hazard: Here we
identify watersheds with populations potentially at-risk due to U.S. EGU mercury by:
(a) identifying those watersheds where total risk meets or exceeds levels considered
to represent a potential public health hazard (i.e., HQ > 2 or IQ loss estimates of 1 to
2 points or greater) and (b) U.S. EGUs contribute to total risk at this subset of
watersheds with elevated risk (we have considered various increments of U.S. EGU
contribution ranging from 5 to 15% [20%?]). We note that, any contribution of
mercury emissions from U.S. EGUs to watersheds where potential exposures from
total mercury deposition exceed the RfD is a hazard to public health, but for purposes
of our analyses we evaluated only those watersheds where we determined U.S. EGUs
contributed 5 percent or more to deposition to the watershed. EPA believes this is a
conservative approach given the increasing risks associated with incremental
exposures above the RfD..
Stage 2 - identify watersheds where risk based on considering deposition from U.S. EGUs
before taking into account deposition and exposures resulting from other sources of
Hg represents a potential public health hazard? Here we identify watersheds with
populations potentially at-risk due to U.S. EGU-attributable risk (prior to considering
mercury contributed by other sources). Although this stage focuses on U.S. EGU
exposure, it is important to keep this incremental exposure in perspective with regard
to total MeHg exposure which typically dominates the U.S. EGU increment across
watersheds.
Stage 3 — what is the combined total number of watersheds (and percentage) where
populations may be at risk from U.S. EGU-attributable Hg? Here we combine
estimates from Stages 1 and 2 to consider watersheds where populations may be at
risk due to (a) U.S. EGUs contributing to exposures at watersheds where total risk
potentially poses a potential public health hazard or (b) U.S. EGUs making an
incremental contribution to total Hg exposure which, when considered alone,
represents a potential public health hazard.
This framework allows us to consider whether U.S. EGU-related exposure when
considered alone, or as a portion of total risk, represents a potential public health hazard. More
specifically, it allows us to estimate the number and percentage of watersheds where populations
may be at risk due to U.S. EGU-related mercury emissions.
Note, that while we present both MeHg RfD-based HQ and IQ loss-based risk metrics in
section 2.6, in discussing risk estimates in the context of determining whether a potential public
18
-------
health hazard exists due to U.S. EGU mercury emissions, we focus on the HQ estimates. This
reflects concerns that the IQ loss endpoint may not capture all of the neurodevelopmental effects
associated with MeHg exposure. Specifically, concerns have been raised in the literature that if
mercury affects a set of specific neurological functions, then use of full-scale IQ as the modeled
health endpoint, could underestimate the neurdevelopmental impacts on other targeted functions
(Axelrad et al., 2007). In addition, two of the most sensitive endpoints in the Faroe Islands study
were the Boston Naming Test and California Verbal Learning Test, both of which can represent
a significant educational risk depending on severity, and those tests are not directly assessed as
part of measuring IQ in children. In addition IQ does not cover other neurologic domains such as
motor skills and attention/behavior and therefore, risk estimates based on IQ will not cover these
additional endpoints and therefore could further underestimate overall neurodevelopmental
impacts (Axelrad et al., 2007).
1.3. Overview of Analytical Approach
This section describes the analytical approach used in conducting the national-scale
mercury risk assessment (note, additional detail on specific modeling elements can be found in
the appendices).
Figure 1-1 provides a flow diagram of the risk analysis identifying the major analytical
steps and associated modeling elements. The risk assessment is based on estimating a set of
subsistence fisher scenarios at watersheds across the U.S. where we have measured fish MeHg
concentration data. After we have estimated total MeHg risk based on modeling consumption of
fish at each of these watersheds, we use the ratio of U.S. EGU to total Hg deposition over each
watershed (estimated using Community Multi-scale Air Quality (CMAQ) modeling) to estimate
the U.S. EGU incremental contribution to total Hg risk. This apportionment of total risk between
the U.S. EGU fraction and the fraction associated with all other sources of Hg deposition is
based on the EPA's Office of Water's Mercury Maps approach (MMaps) that establishes a
proportional relationship between Hg deposition over a watershed and resulting fish tissue Hg
levels, assuming a number of criteria are met. Each of the steps in the analysis is briefly
described below.
Specifying the spatial scale of watersheds
The fist step in designing the analysis was to specify the spatial scale of the watersheds to
use as the basis for risk characterization. As noted above, this risk assessment is based on
estimating risk at watersheds for which we have measured Hg fish tissue data. A number of
studies (Knights et al., 2009, Harris et al., 2007), examining the response of aquatic freshwater
ecosystems to changes in Hg deposition focused on watersheds with dimensions closest to 12-
digit Hydrologic Unit Code classifications (HUC- 12's) (representing a fairly refined level of
watersheds approximately 5-10 km on a side). This suggests that, at least in the context of these
studies, researchers believed that the relationship between changes in mercury deposition and
changes in MeHg levels in aquatic biota could be effectively explored at the level of these more
spatially refined watersheds. In addition, use of a more refined spatial scale (i.e., use of HUC12s
rather than a coarser scale of watershed) in linking changes in mercury deposition to changes in
fish tissue Hg levels also reduces the potential for averaging out areas of high Hg deposition. The
HUC 12 represents the most refined scale of watershed currently available at the national level
19
-------
and therefore was chosen as the basis for linking changes in Hg deposition to changes in fish
tissue MeHg levels. As discussed later in this section, this linkage is central to generating risk
estimates and determining the fraction of total risk associated with U.S. EGUs. Note, the term
"watershed" when used in this document refers to HUC12s unless otherwise noted (see
Appendix A for additional detail on the rationale for selecting HUC12s as the watershed spatial
scale to use in the analysis).
20
-------
Exposure modeling
Specifying
spatial scale of
watersheds
Characterizing
measured fish tissue
Hg concentrations at
the watershed-level
Estimating total fish
consumption-related
Hg exposure forfishers
active at each
watershed
CMAQ
modeling
results fortotal
and US ECU Hg
deposition at
each watershed
1
Apportioning total
exposure between
total and US EGU-
attributable at each
watershed
Defining near-subsistence and
subsistence fisher scenarios
US Demographic data
characterizing source
populations for higFi-
end fishers at the US
Census tract-level
Defining near-
subsistence and
subsistence fisher
populations by
watershed
Studies characterizing fish
consumption rates for
high-consumption fisher
populations
Risk modeling
Mercury RfD for
HQ estimation
Estimation of HQ and
IQ loss at each
watershed including total
riskand US EGU-
attributable fraction
IV
1O(
Jeling IQ loss
IQ loss concentration-
response function
Factortranslating Hg
ingestion dose into Hg hair
concentrations (in ppm)
KEY:
decision
data input
analysis step
Figure 1-1 Flow Diagram of Risk Analysis Including Major Analytical Steps and
Associated Modeling Elements
21
-------
Characterizing measured fish tissue Hg concentrations
The next step was to identify which of the approximately 88,000 HUC12s in the
continental U.S. had fish tissue concentration data and therefore, could be included in the risk
characterization. Although we had compiled fish tissue Hg sampling data for the period 1990 to
2009 from a variety of sources (see Appendix B), we decided to use a subset of these data from
the period 2000 to 2009 in the risk assessment in order to exclude fish tissue samples that likely
reflected Hg deposition levels from the 1990's when anthropogenic emissions in the U.S. were
higher than for period after 2000. We recognizes the complex spatial and temporal nature of the
response offish tissue Hg levels to changes in Hg deposition and loading and acknowledge that a
portion of the sampling data from 2000 to 2009 could still reflect higher Hg loading rates from
earlier periods (see Appendix B). Inclusion offish tissue Hg sampling data collected between
2000 and 2009 resulted in the ability to characterize fish tissue Hg levels for 2,461 of the 88,000
HUC12's in the continental U.S. These watersheds were not randomly sampled across the 88,000
HUC12s, and watersheds in the eastern U.S. are more heavily represented. In addition, because
the samples were often based on state sampling, the samples are not evenly distributed across
states, and some states have very few samples, while others have a large number of samples.
Uultimately, the risk assessment includes estimates of risk for up to 2,461 watersheds depending
on the fisher population being considered (fisher populations that are less ubiquitous such as the
Vietnamese or Great Lakes Tribal populations will not be active across the entire set of 2,461
watersheds and would therefore, only be assessed for a subset of the watersheds- see discussion
below).25
Most of the watersheds with measured Hg fish tissue data had multiple fish tissue MeHg
measurements. This necessitated selecting a fish tissue MeHg statistic to use as the basis for
generating risk estimates. This analysis uses the 75th percentile fish tissue value (computed
separately for each watershed) to model risk. The selection of this statistic reflected the potential
for subsistence-like fishers to favor larger fish, which would likely have relatively higher
mercury levels due to greater bioaccumulation (see Appendix B for additional detail).
Defining subsistence fisher scenarios
Next, we identified the suite of subsistence fishing populations to evaluate in the risk
assessment. As discussed in the introduction, this analysis focuses on populations with higher
consumption rates of self-caught fish who have the potential to fish at inland freshwater
locations, since these populations are expected to experience the greatest U.S. EGU-attributable
risks. Therefore, in reviewing studies of fishing populations, emphasis was placed on identifying
surveys of higher consumption fishing populations active at inland freshwater rivers and lakes
within the continental U.S.
A number of studies were identified that characterized activity for a selection of high-end
fishing populations that met our criteria. These populations included: (a) white and black
populations (including female and poor strata) surveyed in South Carolina (Burger et al., 2002),
(b) Hispanic, Vietnamese and Laotian populations surveyed in California (Shilling et al., 2010)
25 As discussed in Appendix B, in identifying the 2,461 watersheds to include in the risk assessment, we excluded
those that contained gold mines or non-EGU sources of mercury emissions meeting specific criteria we identified as
potentially representing a significant contribution to mercury loading within a watershed.
22
-------
and (c) Great Lakes Tribal populations (Chippewa and Ojibwe) active around the Great Lakes
(Bellinger et al., 2004). For most of these fisher populations (with the exception of the Tribal
populations near the Great Lakes) we assumed that high-end fisher population could be
generalized beyond the specific areas covered in a particular study. This type of generalization
was necessary to provide greater coverage for the continental U.S. and in particular, the eastern
part of the U.S. where U.S. EGU deposition is higher and where we have more measured Hg fish
tissue data. In deciding how to extend coverage for each fisher population, we have considered
several factors including (a) the degree to which high-end fishing activity might be culturally
related and therefore more likely to be followed by populations (e.g., of a given ethnicity) living
across the U.S. and (b) the degree to which high-end fishing activity might be driven by
economic need (i.e., to truly supplement diet) and therefore linked to groups of individuals living
below the poverty line. In each case, we have described our rationale for the regional extension
that has been applied for each of the high-end fisher populations included in the analysis (see
Appendix C and Table C-l for additional detail on the regions used in modeling each fisher
population). Note, that in identifying the specific subset of the 2,461 watersheds to model for
each fisher population, we considered whether there was a "source population" for that fishing
population within the US Census tract that intersected each watershed. For example with
Vietnamese fishers, we would require that at least 25 Vietnamese were located in a US Census
tract intersecting a watershed in order for that watershed to be included in the set modeled for
this fisher population (additional detail on "source populations" is presented in Appendix C).
The studies used to characterize high-end fishing behavior for these populations also
included either (a) high-end percentile self-caught fish consumption rates (90th to 99th percentile
values generally on the order of lOOg/day to -400 g/day or more) or (b) the statistical parameters
necessary to calculate those high-end percentiles (e.g., median and standard deviations). These
subsistence-level consumption rates were used in modeling risk for these populations (see
Appendix C and specifically, Table C-l for specific consumption rates for each of the fisher
populations included in the analysis.
As part of the analysis, in addition to the high-end fisher populations listed above, we
also included a high-end female consumer scenario that was applied more broadly to most of the
watersheds included in the risk assessment.26 This fisher scenario was based on fish consumption
rate data provided in the study by Burger et al., 2002 (see Appendix C, Table C-l) and the
consumption rates involved are generally supported by a number of the studies reviewed (see
Appendix C for additional detail). Because this high-end female consumer population (a) covers
the population of greatest concern from a MeHg exposure standpoint (women of child-bearing
age) and (b) was fairly widely applied (increasing the potential for including high U.S. EGU-
impacted watersheds), we have focused the discussion of risk results provided later in Section
2.6 on this fisher population. With regard to this fisher population, as well as the other
populations considered in the analysis, we would point out that the high-end fish consumption
rates considered, while representing high-end (near bounding) levels, are still reasonable in terms
of subsistence consumption. Most of the rates used (see Appendix C, Table C-l), translate into
26 We applied this high-end fish consuming population to watersheds located in U.S. Census tracts which had at
least 25 people living below the poverty line (this results in this scenario being applied to most of the watersheds
where we have fish tissue MeHg data, since this is a less stringent criterion for inclusion). This requirement of
having at least 25 people below the poverty line reflects the assumption that near-subsistence levels of fishing
activity is more likely among individuals who are economically disadvantaged.
23
-------
between one fish meal every few days to a large fish meal every day. Viewed from the
perspective of the type of subsistence fishing activity that is the focus of this analysis, these
consumption rates are reasonable (i.e., they represent consumption rates in line with subsistence
behavior).
Estimating total fish consumption-related Hg exposure
The next step in completing the analysis was to estimate total exposure to Hg at each of
the 2,461 watersheds for the set of high-end fish consumption populations identified for each
watershed. Estimates of total mercury exposure were generated by combining 75th percentile fish
tissue value with the consumption rates for a particular fisher populations. Note, that a cooking
loss factor (actually reflecting the fact that the preparation offish can result in increased mercury
concentrations) was also included in exposure calculations (see Appendix D for additional detail
on the calculation).
Apportioning total MeHg exposure between total and U.S. EGU-attributable exposure
Next we needed to estimate the fraction of total exposure that is associated with U.S.
EGU Hg deposition at each watershed. U.S. EGU apportionment of total Hg exposure is based
on application of the MMaps assumption (see Appendix E). Essentially this approach assumes
that under near steady state conditions, a fractional change in mercury deposition to a watershed
will be reflected in a matching proportional change in the levels of MeHg in fish. We have
extended this proportionality assumption to allow us to apportion MeHg levels in fish between
mercury sources based on the associated apportionment of mercury deposition within a given
watershed between these sources. Of course, the process of mercury loading and impacts on
mercury bioaccumulated in fish is complex and involves varying temporal profiles depending on
a variety of factors (e.g., methylation potential of the waterbody, role of watershed sediment
erosion and runoff in mercury loading to the watershed etc). However, for purposes of this
analysis, we make the assumption that given sufficient time to achieve near steady-state
conditions, a given ratio of total Hg deposition to U.S. EGU deposition will ultimately be
reflected in the fish tissue MeHg levels. In addition, we note that the MMaps assumption does
require that certain criteria be met (e.g., atmospheric deposition is the primary source of mercury
loading to the watershed and that factors related to methylation potential in watersheds be held
constant for a sufficient time to allow near-steady state conditions to be reached). These criteria
and the degree to which they were considered in our analysis are further described in Appendix
E. For example, as discussed in greater detail in Appendix E, in identifying the 2,461 watersheds
to include in the risk assessment, we excluded those that contained gold mines or had significant
non-aerial sources of mercury loading.
CMAQ modeling completed at the 12km grid cell resolution was used to estimate total
annual mercury deposition from US and foreign anthropogenic and natural sources over each
watershed, including the fraction of deposition contributed by U.S. EGUs. As noted in the
discussion of the scope of the analysis presented in Section 1.1, we are modeling two temporal
period in the analysis: (a) a 2005 scenario representing 2005 conditions as reflected in the 2005
NEI mercury emissions inventory and (b) an 2016 based case scenario after CAA-related
regulations potentially reducing Hg emissions from U.S. EGUs (e.g., the Transport Rule) are in
place. In the context of these two temporal scenarios, the CMAQ modeling results together with
24
-------
the MMaps assumption can be used to both estimate the portion of total Hg exposure that is
associated with U.S. EGUs and project changes in fish Hg concentrations (and consequently total
exposure) associated with changes in total Hg deposition in the future. Specifically, for the 2005
analysis, CMAQ modeling results for a particular watershed allow us to estimate the proportion
of total exposure (estimated for that watershed) that is associated with U.S. EGU deposition (i.e.,
based on the ratio of U.S. EGU Hg deposition to total Hg deposition over the watershed). In the
case of the 2016 simulation, we can first project changes in total fish tissue Hg levels (for that
watershed) by comparing estimates of total Hg deposition in 2005 to estimates for 2016
generated by CMAQ and then again, apportion that adjusted total risk between U.S. EGUs and
all other sources, based on comparing U.S. EGU Hg deposition to total Hg deposition in 2016.
See Appendix E for additional detail on CMAQ modeling.
Estimate risk including HO and 10 loss
Once both total exposure and the U.S. EGU incremental contribution to that total
exposure have been estimated, we can then estimate risk, including both HQ and IQ points lost.
The HQ estimate is generated by comparing total Hg exposure (as annual-average bodyweight-
adjusted ingested MeHg dose) to the MeHg RfD.27 Similarly, the U.S. EGU incremental
contribution to total HQ is estimated by comparing the U.S. EGU increment of exposure to the
MeHg RfD. As noted earlier, an HQ > 1.5 (which rounds to an HQ of 2) is considered to
represent a potential public health hazard, since it signifies that exposure has been assessed to
exceed the MeHg RfD.28
In the case of IQ loss, we first covert annual-average ingested dose estimates for MeHg
into equivalent maternal hair mercury levels, since the CR function for IQ loss is based on
estimated exposure characterized as maternal hair mercury levels. This conversion is
accomplished using a factor based on a one compartment toxicokinetic model used for deriving
the MeHg RfD by Swartout and Rice (2000). Then a CR function relating hair mercury levels to
IQ points lost in children born to mothers whose exposure is modeled in this analysis is used to
predict IQ points lost for those children. This CR function was published in Axelrad et al.,
2007and is based on application of a Bayesian hierarchical model which integrates data from the
three key epidemiological studies (Seychelles, New Zealand and Faroe Islands).29
Since the CR function was published in the Axelrad et al., 2007 study, a number of
authors have raised the possibility that neurological deficits related to Hg exposure through fish
consumption could be masked to some degree by the neurologically-beneficial effects offish oil
consumption. Some authors have suggested that the IQ loss factor should be adjusted upward to
compensate for this masking effect (see Rice et al., 2010 and Oken, 2008). However, no rigorous
27 The MeHg RfD is 0.0001 ug/kg-day (equivalent to 0.1 ug/kg-day) and was published by EPA in the Integrated
Risk Information System in 2001 (http://www.epa.gov/iris/subst/0073.htm)
28 Note, that for the U.S. EGU incremental contribution analysis, anHQ of less than 1.5 does not necessarily
indicate there is no public health hazard related to U.S. EGU emissions. Rather, it suggests that, for those specific
watersheds, we need to also consider whether total risk (i.e., the HQ reflecting total MeHg exposure) exceeds 1.5
and therefore represents a potential public health hazard. If that is the case, then we would consider the degree to
which U.S. EGUs contribute to that total exposure because incremental exposure above the RfD increase the risk.
29 The IQ loss model uses a linear slope of 0.18 IQ points per ppm hair Hg concentration (Axelrad et al., 2007).
25
-------
basis for a specific adjusted estimate has been provided to-date and therefore, we address this
potential for low-bias as part of our qualitative uncertainty discussion.
1.4. Discussion of key sources of uncertainty and variability
The risk assessment has been designed to reflect critical sources of variability to the
extent allowed by available methods and data and given the resources and time available. The
key sources of variability associated with the analysis include: (a) variation in the pattern of total
and U.S. EGU-attributable mercury deposition across watersheds in the US, (b) variation in the
patterns offish tissue MeHg levels across the US, and (c) variation of the types of high-end
fishing activity likely to occur in different parts of the country. These sources of variability and
the degree to which they are reflected in the design of the analysis are identified and described in
Appendix F, Table F-l.
Regarding uncertainty, in Appendix F, Table F-2, we have identified sources of
uncertainty impacting the analysis and attempted to characterize (a) the nature of the impact of
each source on risk estimates and (b) the degree to which the potential impact of the source of
uncertainty is characterized as part of the analysis (including whether sensitivity analyses
completed for the risk assessment address a particular source of uncertainty).
In addition to the sources of uncertainty addressed in Table G-2, which focus on factors
related directly to the exposure scenarios modeled in the risk assessment, our decision not to
model risk associated with consumption of Great Lakes fish and fish sourced from near U.S.
coastal locations (including estuarine areas) also adds additional uncertainty into the analysis. As
explained above in Section 1.1, these sources of self-caught and commercial fish were not
modeled primarily due to challenges associated with linking specific areas of U.S. EGU mercury
deposition to fish in these waterbodies. A related concern is that the greater dilution and potential
mobility offish in these larger waterbodies, could reduce the impact of elevated U.S. EGU
mercury deposition of portions of these waterbodies. Despite these considerations, the risk
assessment may have overlooked elevated U.S. EGU-attributable risks for high-consuming fisher
populations active in these locations. However, we would point out that we still expect the
greatest U.S. EGU-attributable risk to occur at inland freshwater bodies with (a) relatively
elevated fish tissue Hg levels and (b) relatively elevated levels of U.S. EGU mercury deposition.
1.5. Differences between the 2005 Section 112(n) Revision Rule analysis and the current
analysis in support of the Propose U.S. EGU Toxics Rule
In 2005, EPA conducted a set of technical analyses to support revision of the 2000
appropriate and necessary finding.30 This section identifies key differences between the
watershed-level risk assessment completed in support of the 2005 revision rule and the current
risk assessment. These differences include both technical factors related to the design of the
assessments, as well as differences in the interpretation of potential public health significance of
the risk estimates generated. Key differences between the two analyses include:
U.S. EPA. 2005. Technical Support Document: Methodology Used to Generate Deposition, Fish Tissue
Methylmercury Concentrations, and Exposure for Determining Effectiveness of Utility Emission Controls.
26
-------
Higher spatial resolution through use of CMAQ 12km grid cells; We are now using 12 km grid
cells in estimating mercury deposition using CMAQ, whereas in the 2005 analysis, we used 36
km grid cell modeling. The more refined grid cells used in the current analysis are more
appropriate for representing areas of elevated U.S. EGU deposition (and total Hg deposition in
general) compared with the 36 km grid cells used in the 2005 analysis. The 12 km grid cell also
matches up with the more refined HUC12 watersheds now being used in the analysis, thereby
allowing a more refined treatment of the intersection of aerial mercury deposition and measured
fish tissue concentrations at the watershed level.
Application of more refined HUC12 watersheds: The current analysis uses HUC12 watersheds as
the basis for risk estimation (these watersheds typically are 5-10 km on a side). By contrast, the
the 2005 analysis used HUCSs which are much larger (averaging 40km on a side). The use of
more spatially refined watersheds increases the potential for capturing areas of elevated aerial Hg
deposition (combined with measured fish tissue levels).
Inclusion of updated fish tissue data: For this analysis, we included measured fish tissue data
collected between 2000 and 2009. By contrast, the the 2005 analysis used data collected
between 1999 and 2003 (in that case to support an analysis completed in 2005).
Subsistence fisher activity better defined and considered more ubiquitous: Based on an extensive
review of available literature, we have identified studies characterizing high-end self-caught fish
consumption for a wide variety of source populations (e.g., Hispanic, Vietnamese, Whites and
Blacks in the southeast, Tribal populations). Although it was necessary to extrapolate high-end
fishing activity to regions beyond those covered in the underlying studies, we do believe that the
literature generally supports the plausibility of high-end subsistence-like fishing activity existing
across to some extent across the watersheds included in the analysis. Additionally, the variety of
studies identifying self-caught fishing activity at subsistence levels (i.e., a meal every few days
to a meal every day) for a variety of diverse populations in different regions of the country, adds
support to assessing this type of fishing behavior across the modeled watersheds (for additional
detail on the fishing populations included in this risk assessment, including consumption rates
see Section 1.4 and Appendix C).
By contrast, in the 2005 analysis, we concluded that the study data characterizing fishing activity
available at that time was limited in its ability to support modeling of subsistence fisher activity
for the following reasons: (a) it characterized regional or local activity that could not be readily
extrapolated more broadly, (b) fishing activity queried included consumption of saltwater
species, or (c) specific high-end percentiles were not identified (or if they were, they only
applied during specific harvesting periods - e.g., spearfishing months for Great Lakes Tribes).
Therefore, in the 2005 analysis, we ended up applying a high-end self-caught percentile values
(95th and 99th percentiles) based on Tribal fishing practices in the Northwest to watersheds across
the country.31 The updated literature review we have done for the current analysis, has lead us to
31 Note, that these NW Tribal fishing estimates are subject to considerable uncertainty when they, in particular, are
extrapolated to cover other areas in the U.S. These specific high-end fish consumption rates were derived for Tribes
active in the Northwest who engage in specific cultural practices focused around salmon fishing. There is significant
uncertainty in extrapolating this type of highly-specific cultural-based fishing activity to other Tribes, let alone to
27
-------
revise several of our earlier conclusions regarding high-end fishing activity. Specifically, while
many of the studies of subsistence-like activity are regional in nature, when considered together,
we now conclude that they support modeling subsistence-like fishing activity more broadly
across the entire study area. Additionally, while some of the studies may include saltwater
fishing in addition to freshwater (e.g., Burger, 2002), when those studies clearly covered both
saltwater and freshwater self-caught fish consumption, we concluded that it was reasonable to
assume that subsistence-like fishing activity could occur both at the coast and inland at
freshwater bodies.32
For the current analysis, we are also using the 75th percentile fish tissue MeHg level reflecting
the potential for high-end subsistence fishers to target larger fish which would have greater
bioaccumulation potential relative to the average fish. By contrast, in the 2005 risk assessment,
we used the maximum fish tissue MeHg level across species offish in a given HUC8, which is a
more conservative approach (i.e., resulting in higher risk, other factors equal).
Calculation of RfD-based HQ estimates including total and U.S. EGU-attributable risk and
calculation of IQ loss: For this analysis, we have compared total exposure to the MeHg RfD to
generate an HQ estimate based on total mercury exposure for fishers at a given watershed.
Furthermore, to focus on the U.S. EGU component of that total risk, we have generated two
related risk metrics: (a) U.S. EGU incremental contribution to total risk which essentially
considers the magnitude of the HQ when deposition from U.S. EGUs is considered before taking
into account deposition and exposures resulting from other sources of Hg and (b) the percent of
total HQ risk attributable to U.S. EGUs. The calculation of U.S. EGU incremental contribution
to total HQ is identical to the IDI (index of daily intake) metric used in the 2005 analysis.
However an important distinction is that in the current analysis, we highlight the fact that this
U.S. EGU-related risk is always associated with a total HQ which is generally substantially
larger (i.e., the US-EGU-attributable HQ should not be considered in isolation as was done in the
2005 analysis with the IDI). By contrast, for the 2005 analysis both of the risk metrics used (i.e.,
the IDI and the comparison of U.S. EGU-related fish tissue concentrations against EPA's water
quality criterion expressed as a mercury fish tissue value) essentially considered the U.S. EGU
portion of risk in isolation. These risk metrics in the 2005 analysis were not contrasted with the
much larger fraction of total mercury-related risk associated with the non-U.S. EGU portion of
risk.
For the current analysis, we also generated estimates of IQ loss (these were not generated for the
the 2005 analysis). Estimates of a specific health endpoint associated with the U.S. EGU-
attributable fraction of mercury in fish provides a risk metric that can be more appropriately
considered in isolation (i.e., it is more reasonable to consider the U.S. EGU-attributable IQ loss
other fish consumers in the U.S. By contrast, extrapolation of more generalized high-consumption rates (for ethnic
groups and whites and blacks) to cover portions of the U. S. as was done in the current analysis is subject to far less
uncertainty.
32 Note, that particularly in the situation where a study specifically characterized poor high-end fishing populations,
as is done in the Burger 2010 study of activity in SC, we considered it reasonable to assume that poor individuals
would likely conduct their frequent fishing activity near home. In that case, some of these high-end fishers would
likely be located near the coast and some inland. In the case of subsistence-like fishing activity in the southeast,
other studies from rivers in that area also showed subsistence-like fish consumption rates when only freshwater
rivers were considered (e.g., Burger et al., 1999 focusing on fishing activity on the Savannah river in GA).
28
-------
than to focus on the U.S. EGU-attributable fraction of an HQ). This reflects the fact that IQ loss
quantifies a discrete increment of a public health effect, while with the RfD, it is more difficult to
characterize what a "fraction" of an HQ value actually represents in terms of potential health
significance. However, as discussed in Section 1.2, there is concern that the IQ endpoint may not
fully capture all of the neurodevelomental effects associated with MeHg exposure and for this
reason, in presenting risk estimates in the context of determining whether there is the potential
for a public health hazard associated with exposure to U.S. EGU-sourced mercury, we focus on
the MeHg RfD-based HQ estimates and not on the IQ loss estimates, although both are
presented.
2. Discussion of Analytical Results
This section provides a discussion of the results from the various analyses completed as
part of the risk assessment for the 2005 simulation and the 2016 simulation. Prior to discussing
these results, a brief overview of critical design elements of the risk analysis that the reader
should keep in mind when reviewing the results (Section 2.1). The specific sets of analyses
described in this section include: (a) mercury emissions from U.S. EGUs (Section 2.2), (b)
mercury deposition from U.S. EGUs as modeled using CMAQ (Section 2.3), (c) fish tissue
MeHg concentrations (Section 2.4), (d) relationship between mercury deposition and
methlymercury fish tissue concentrations (Section 2.5) and (d) risk assessment results, including
MeHg RfD-based HQ estimates and IQ loss estimates (Section 2.6). In discussing each of these
category of results, emphasis is placed on identifying key policy-relevant observations. In
Section 2.7, we discuss the results of several sensitivity analyses conducted to characterize the
potential impact of specific sources of uncertainty on the risk estimates. In Section 2.8, we
provide a summary of critical observations from the analysis.
2.1. Key design elements to consider when reviewing the risk assessment results
The following design elements of the analysis should be considered when reviewing the
results discussed in the following sections (note, that this only highlights portions of the design
of the analysis - the reader is referenced to Section 1.3 and associated Appendices for a more in-
depth discussion of the analysis design):
• The analysis focuses on subsistence-like fishing activity at inland freshwater bodies. The
analysis is not intended to capture more generalized recreational fishing activity or to
reflect self-caught fisher exposure associated with saltwater fishing or fishing in the Great
Lakes. In comparing any risk profiles generated in this analysis to risks estimated in other
contexts, the specific focus on this analysis on these high-end populations needs to be
considered (e.g., risks in this analysis will generally be substantially higher than those
estimated for recreational fishers).
• The analysis is watershed-focused and risks are generated for subsets of the 2,461
watersheds for which we have measured fish tissue MeHg data. This watershed coverage
(which is only about 4% of the watersheds in the U.S.), leaves much of the country not
covered by the analysis, including a substantial number of watersheds with relatively
elevated levels of U.S. EGU-related mercury deposition. Further, we note that the
watersheds with fish tissue MeHg data are concentrated in the eastern part of the country
29
-------
and therefore, this portion of the continental U.S. is more heavily represented in the
watershed-level estimates of risk that are generated. Given that U.S. EGU mercury
deposition is generally higher in the eastern part of the U.S., the fact that the risk
assessment is focused on this part of the country is considered a strength of the analysis.
• The analysis uses the MMaps approach to relate changes in mercury deposition over
particular watersheds to resulting changes in mercury fish tissue concentrations.
Similarly, this approach allows us to reflect source-apportionment of mercury deposition
(i.e., between U.S. EGU and all other sources) in the underlying fish tissue levels. It is
then possible to translate those changes in fish tissue levels (or source-apportioned fish
tissue levels) into equivalent changes in exposure and risk. This approach assumes that
near steady state conditions are met in the fish tissue MeHg concentrations, which may
take years to decades at a given watershed following changes in mercury deposition.
• The analysis generates a series of risk metrics based on (a) estimates of MeHg RfD-based
HQ and (b) estimates of IQ loss in children born to mothers exposed to MeHg through
high-end fish consumption. We do consider the U.S. EGU-related contribution to both
types of risk. However, particularly in the context of HQ, U.S. EGU incremental
contributions to total risk should always be considered in the context of total HQ which is
typically substantially larger than the U.S. EGU incremental contribution when
considered in isolation.
• Because it is not feasible to enumerate the subsistence-like fisher populations modeled in
this analysis, we could not generate distributions of population-weighted risk for specific
fishing populations assessed (e.g., poor Hispanic fishers, or Tribal fishers in the vicinity
of the Great Lakes). To reiterate, this reflects the fact that we are focusing on fishers
engaging in high fish consumption and it is difficult to count individuals in this subset of
each fishing population (and assign them to specific watersheds). While we can not
enumerate these populations, we do believe that, based on surveys of their behavior, that
this type of subsistence-like activity could reasonably be expected to occur across some
fraction of the 2,461 watersheds included in the analysis.33 Therefore, we have assessed
high-end fisher risk for each watershed. We then consider the fraction of watersheds with
simulated high-end risk within specific categories of interest. While not a population-
representative characterization of risk, this approach does allow us to consider percentiles
of watersheds based on a reasonable assumption that this kind of high-end fishing activity
could occur across the watersheds modeled.
2.2. Mercury Emissions from U.S. EGUs
The most recent data on U.S. EGU emissions based on information collected from
industry through the Information Collection Request (ICR) show total mercury emissions of 29
tons in 2010. This shows a significant reduction in U.S. EGU mercury emissions from 2005,
when mercury emissions were estimated to be 52.9 tons. The reductions between 2005 and 2010
33 As discussed in section 1.3 and in additional detail Appendix C, we only considered specific high-end populations
for those watersheds located in US Census tracts with a "source population" greater than 25 for the fishing
population being assessed.
30
-------
are largely due to state mercury regulations and federal enforcement actions that achieve
mercury reductions as a co-benefit of controls for NOx and SO2 emissions. The EPA projection
of total mercury emissions from U.S. EGUs in the 2016 (once a number of the CAA-related
regulations are fully in effect) is 29 tons. Given these estimates of total mercury emissions,
characterization of "current conditions" would be better represented by our 2016 Scenario than
the 2005 Base Case, since total emission for the former (at 29 tons) is closer to our projection of
current 2010 emissions (at 29 tons). By contrast, the 2005 analysis reflects total mercury
emissions (52.9 tons) which are significantly higher than our estimate of current emissions in
2010. For this reason, as mentioned earlier, we emphasis risk estimates for the 2016 Scenario in
presenting and interpreting risk estimates.
2.3. Mercury Deposition from U.S. EGUs as Modeled Using CMAQ
This section characterizes patterns of U.S. EGU-related mercury deposition for the two
scenarios assessed (2005 and 2016 Scenario) using CMAQ. In presenting and discussing these
results, we contrast U.S. EGU-attributable deposition with deposition from all sources combined.
This discussion is based around a series of figures and tables conveying relevant information,
which are described below. After presenting these figures, a set of bulleted observations is
presented at the end of the section that draws on information conveyed in the figures and tables.
The set of figures and tables presented include:
• Figure 2-1 and 2-2: Maps presenting CMAQ modeling results for total mercury
deposition (ug/m2) at the watershed-level, for the 2005 and 2016 scenarios respectively.
• Figures 2-3 and 2-4: Maps presenting CMAQ modeling results for U.S. EGU-attributable
mercury deposition (ug/m2) at the watershed-level, again for the 2005 and 2016
scenarios, respectively.
• Table 2-1: Summary of statistics (mean, 50th, 75th, 90th, 95th and 99th percentiles) for total
mercury deposition and U.S. EGU-attributable deposition for the 2005 and 2016
scenarios.
• Table 2-2: Summary of statistics (mean, 50th, 75th, 90th, 95th and 99th percentiles) for U. S.
EGU-deposition as a percent of total deposition for the 2005 and 2016 scenarios.
• Table 2-3: Summary of statistics (mean, 50th, 75th, 90th, 95th and 99th percentiles) for
percent reduction of (a) total mercury deposition, and (b) U.S. EGU-attributable
deposition, based on comparison of the 2016 scenario against the 2005 scenario.
-------
2005 Hg total deposition (ug/m2)
16-26
26-43
43-99
99 - >500
Figure 2-1. Total mercury deposition by watershed (2005)
2016 total Hg
deposition (ug/m2)
26-43
43-98
99-=-5ljri
Figure 2-2. Total mercury deposition by watershed (2016)
32
-------
US EGU-attributable
mercury deposition (2005)
(ug/m2)
| 10-54
2.0-10
0.6102.0
0.18- 0.6
I «0.18
Figure 2-3. U.S. EGU-attributable mercury deposition by watershed (2005)
US EGU-attributable
mercury deposiiton (2016)
^H 10-21
^H 2.0- 10
O.B-2.0
0.1B-0.6
^B <0 18
Figure 2-4. U.S. EGU-attributable mercury deposition by watershed (2016)
33
-------
Table 2-1. Comparison of total and U.S. EGU-attributable mercury deposition (ug/m2) for
the 2005 and 2016 scenarios *
Statistic
Mean
Median
75th percentile
90th percentile
95th percentile
99th percentile
2005 scenario
Total Hg
Deposition
19.41
17.25
23.69
30.78
36.85
58.32
U.S. EGU-attributable
Hg Deposition
0.89
0.24
1.07
2.38
3.60
7.77
2016 scenario
Total Hg
Deposition
18.66
16.59
22.83
29.90
35.16
56.23
U.S. EGU-attributable
Hg Deposition
0.34
0.15
0.46
0.85
1.18
2.41
* Values are based on CMAQ results interpolated to the watershed -level and reflect trends across all -88,000
watersheds in the U.S.
Table 2-2. Comparison of percent of total mercury deposition attributable to U.S. EGUs
for 2005 and 2016.*
Statistic
Mean
Median
75th percentile
90th percentile
95th percentile
99th percentile
2005 scenario
5%
1%
6%
13%
18%
30%
2016 scenario
2%
1%
3%
5%
6%
11%
* Values are based on CMAQ results interpolated to the watershed -level and reflect trends across all -88,000
watersheds in the U.S.
Table 2-3. Comparison of percent reduction of total mercury deposition, and U.S. EGU-
attributable deposition, based on comparing the 2016 scenario against the 2005
scenario.*
Statistics
Mean
Median
75th percentile
90th percentile
95th percentile
99th percentile
Percent Change in Total
Hg Deposition
-4%
-1%
-5%
-12%
-16%
-27%
Percent Change in U.S.
EGU-attributable Hg
Deposition
NC**
-41%
-70%
-80%
-85%
-91%
* Values are based on CMAQ results interpolated to the watershed -level and reflect trends across all ~i
watersheds in the U.S.
* * A mean value was not calculated for this category due to presence of a number of watersheds with very small
U.S. EGU-attributable deposition values which skewed this distribution.
34
-------
Consideration of information presented above in Figures 2-1 through 2-4 and in Tables 2-
1 through 2-3 resulted in the following observations regarding estimates of total and U.S. EGU-
attributable mercury deposition for the 2005 and 2016 scenarios (note, all observations
referencing the U.S. are for the continental U.S.):
• Patterns of total and U.S. EGU-related Hg deposition differ considerably: The pattern of
total Hg deposition across the country is different from the pattern of U.S. EGU
deposition. There are areas of elevated total Hg deposition distributed around the country
(e.g., west coast, areas in Nevada, southern Mississippi, West Virginia, southeastern
Georgia) (see Figures 2-1 and 2-2). By contrast, U.S. EGU Hg deposition is concentrated
in the eastern half of the country with one of the main regions of elevated deposition
being in the Ohio River Valley (see Figures 2-3 and 2-4). Figures 2-3 and 2-4 also
illustrate that while some near-coastal areas and portions of the Great Lakes do have
elevated U.S. EGU mercury deposition, many of the highest areas (and largest expanses)
of U.S. EGU deposition occur inland (e.g., Ohio River Valley, areas in northeast Texas
and along the Mississippi River).
• US Hg deposition is generally dominated by sources other than U.S. EGUs (with the
contribution from U.S. EGUs decreasing between the 2005 and 2016 scenarios): On
average across the U.S., U.S. EGUs contribute 5% of total Hg deposition under the 2005
scenario with this level decreasing to 2% under the 2016 scenario (see Table 2-3). The
remaining Hg deposition (i.e., -95% and -98%, respectively for the two scenarios)
originates from other U.S. sources of mercury emissions and from foreign sources (both
anthropogenic and natural). There is a considerably decrease in U.S. EGU Hg deposition
between the 2005 and 2016 scenarios, with this resulting primarily from implementation
of the Transport Rule, state mercury regulations and Federal enforcement actions.34 The
median reduction in U.S. EGU Hg deposition was 41% with reductions ranging up to
85% for the 95th% watershed (when ranked according to magnitude of reduction in U.S.
EGU Hg deposition) (see Tables 2-2 and 2-3).
• The contribution of U.S. EGU deposition to total deposition does vary across watersheds
and can represent a relatively large fraction in some (more limited) instances: In the
2005 scenario, while on average, U.S. EGUs only represented 5% of total Hg deposition
in the U.S., values ranged up to 30% for the 99th% watershed (see Table 2-2). While
overall U.S. EGU Hg deposition decreased substantially for the 2016 scenario, still, U.S.
EGUs contributed 11% of total Hg deposition for the 99th% watershed (ranked according
to U.S. EGU deposition) (see Table 2-2).
2.4. Fish Tissue MeHg Concentrations
34 Controls on PM precursors, including directly emitted PM and SO2, can have significant secondary reductions on
divalent and particle-bound mercury, both of which produce much of the local and regional deposition.
35
-------
This section characterizes the set of 2,461 watershed-level fish tissue MeHg samples used
in the analysis. As noted earlier in Section 1.3, the 75th percentile of the samples falling within a
given watershed is used as the basis for the risk estimates generated for that watershed.
Consequently we have used the 75th percentile statistic (at the watershed level) as the basis for
summarizing the fish tissue MeHg data presented in this section. Recall also, that as discussed in
Section 1.3, the MMaps approach was used to estimate the U.S. EGU-attributable portion of each
75th percentile fish tissue value within each watershed (based on the fraction of total mercury
deposition associated with U.S. EGUs for each watershed). Similarly, for the 2016 scenario,
baseline fish tissue sampling data used in 2005 was first adjusted to reflect changes in total
deposition (between the 2005 and 2016 Base Cases) for a given watershed (also using the
MMaps approach), and then the U.S. EGU-attributable fraction was estimated, again based on
the fraction of total mercury deposition over that watershed that is associated with U.S. EGUs.
The summary offish tissue MeHg data is based around a series of figures and tables
conveying relevant information. The figures and tables used in summarizing the fish tissue data
are described below. Note, that, as discussed in Section 2.3, most of the areas experiencing
elevated U.S. EGU-attributable mercury deposition are located in the eastern half of the country.
We also have greater coverage with mercury fish tissue data in the eastern part of the country. In
addition, because the Transport Rule primarily affects U.S. EGUs in the eastern half of the
country, we mainly see reductions in U.S. EGU-attributable risk (in comparing the 2005 to 2016
scenarios) for the eastern portion of the country. For these reasons, in illustrating spatial
trends/patterns in fish tissue data in this section through figures, we focus primarily on the
eastern half of the country (note, however that data presented in tables are for the whole
continental U.S.). After presenting these figures and tables, a set of bulleted observations is
presented at the end of the section that draws on information conveyed in the figures and tables.
The set of figures and tables presented include:
• Figure 2-5: Map of 2,461 watersheds with fish tissue sampling data used in the risk
assessment. This map not only illustrates general coverage of the fish tissue data for
different regions of the country, it also illustrates the relatively small size of the HUC12
watersheds used in the analysis.
• Figure 2-6: Map of the subset (approximately 2,170) of the 2,461 watersheds falling in
the eastern half of the U.S.. This map uses color gradients to illustrate spatial variation in
the total mercury fish tissue concentrations (for the 2005 Base Case) across the
watersheds and as such, clearly illustrates how difficult it is to identify any discernable
patterns with this approach given the small size of the watersheds. Because of this, we
decided instead, to use graduated circles (with circle size tracking 75th percentile fish
tissue concentrations at each watershed) in the remainder of maps presented in this
section, since this approach allows spatial patterns to be more readily discerned.
• Figure 2-7 and 2-8: Maps presenting CMAQ modeling results for total mercury
deposition (ug/m2) at the watershed-level in the eastern U.S., for the 2005 and 2016
scenarios respectively.
36
-------
• Figures 2-9 and 2-10: Maps presenting CMAQ modeling results for U.S. EGU-
attributable mercury deposition (ug/m2) at the watershed-level in the eastern U.S., again
for 2005 and 2016 scenarios, respectively.
• Figures 2-11 and 2-12: Maps of the upper 10th percentile of watersheds based on total
mercury fish tissue levels for the 2005 and 2016 scenarios, respectively. These maps
allow us to consider how spatial patterns (and overall magnitude) of watersheds with the
highest fish tissue levels change between the 2005 and 2016 scenarios.
• Figures 2-13 and 2-14: Maps of the upper 10th percentile of watersheds based on U.S.
EGU-attributable mercury fish tissue levels for the 2005 and 2016 scenarios,
respectively. These maps allow us to consider how spatial patterns (and overall
magnitude) of watersheds with the highest U.S. EGU-attributable fish tissue levels
change between the 2005 and 2016 Base Cases. Note, that these maps are particular
relevant to consideration of changes in the patterns of U.S. EGU-attributable risk between
the 2005 and 2016 Base Cases.
• Table 2-4: Summary of statistics (min, max, mean, 50th, 75th, 90th, 95th and 99th
percentiles) for both total and U.S. EGU-attributable Hg fish tissue levels (for the 2005
and 2016 scenarios). These statistics are based on watershed-level data. In addition, this
table also presents the percent reduction (between the 2005 and 2016 scenarios) for both
total and U.S. EGU-attributable Hg fish tissue levels.
• Table 2-5: Summary of statistics (min, max, mean, 50th, 75th, 90th, 95th and 99th
percentiles) for U.S. EGU-attributable fraction of total Hg fish tissue (these results are
reflected directly in total and U.S. EGU-attributable risk calculations).
37
-------
Watersheds
(2,461 HUC12S our of -88,000 in the U.S.)
containing fish tissue
samples used in the
risk assessment
Figure 2-5. Location of 2,461 watersheds with mercury fish tissue data included in the risk
assessment
Mercury fish tissue
data (2005) (ppm)
Figure 2-6. Subset of 2,170 watersheds (from the larger set of 2,461 included in the risk
assessment) located in the eastern half of the country, (this maps also illustrates
limitations with using color-coding at the watershed-level to explore trends in Hg fish
tissue concentrations - see text)
38
-------
Mercury fish
tissue data (2005) (ppm)
Figure 2-7. Total Hg fish tissue concentrations (for the 2005 Base Case) for the subset of
watersheds included in the risk assessment located in the eastern U.S.
Mercury fish
tissue data (2016) (ppm)
<• 0.0-0.2
• 0.2-0.3
0.4-0.5
0.6-3.0
3.1-7.0
Figure 2-8. Total Hg fish tissue concentrations (for the 2016 Base Case) for the subset of
watersheds included in the risk assessment located in the eastern U.S.
39
-------
US EGU-attributable
mercury fish
tissue data (2005) (ppm)
» 000-0.02
O 003-0.05
O 006-0.10
O 0 11-015
O 0.16-0.30
Figure 2-9. U.S. EGU-attributable Hg fish tissue concentrations (for the 2005 Base Case)
for the subset of watersheds included in the risk assessment located in the eastern
U.S.
US EGU-attributable
mercury fish
tissue data (2016) (ppm)
« 0.00-0.02
O 0.02-0.05
O 005-0.10
O 011-0.15
O 0 15-030
Figure 2-10. U.S. EGU-attributable Hg fish tissue concentrations (for the 2016 Base Case)
for the subset of watersheds included in the risk assessment located in the eastern
U.S. (Note, this map uses the same scale as Figure 2-7, thereby supporting direct
comparison between these two time periods)
40
-------
Top 10th percentile
mercury fish tissue
concentrations (2005) (ppm)
• 0.67-0.85
* 0.66-1.00
• 1.01 - 1.50
• 1.51-200
^2.01 -7.00
Figure 2-11. Top 10 percentile of watersheds based on total Hg fish tissue concentrations
(for the 2005 simulation), (ranking is based on full national set of watersheds included
in the risk assessment, but map focuses on locations in the eastern U.S.)
Top 10th percentile
mercury fish tissue
concentrations (2016} {ppm)
Figure 2-12. Top 10 percentile of watersheds based on total Hg fish tissue concentrations
(for the 2016 simulation), (ranking is based on full national set of watersheds included
in the risk assessment, but map focuses on locations in the eastern U.S.)
41
-------
Top 10th percentile
US EGU-attributable
mercury fish tissue
concentrations (2005} (ppm)
• 0.056-0.060
• 0.061-0.080
• 0.081-0.100
• 0.101-0.200
A 0201-0.500
Figure 2-13. Top 10 percentile of watersheds based on U.S. EGU-attributable Hg fish
tissue concentrations (for the 2005 simulation), (ranking is based on full national set of
watersheds included in the risk assessment, but map focuses on locations in the eastern
U.S.)
Top 10th percentile
US EGU-attributable
mercury fish tissue
concentrations (2016) (ppm)
• 0.019-0.060
• 0.061-O.OBO
• 0.081-0.100
0 0.101-0.200
A 0201-0.500
Figure 2-14. Top 10* percentile of watersheds based on U.S. EGU-attributable Hg fish
tissue concentrations (for the 2016 simulation), (ranking is based on full national set of
watersheds included in the risk assessment, but map focuses on locations in the eastern
U.S.)
42
-------
Table 2-4. Comparison of total and U.S. EGU-attributable Hg fish tissue concentrations
(including % change) for the 2005 and 2016 scenarios.
Statistic
Mean
Median
75th %
90th %
95th %
99th %
Hg fish tissue concentration (ppm)
2005 scenario
Total
0.31
0.23
0.39
0.67
0.91
1.34
U.S. EGU-
attributable
0.024
0.014
0.032
0.056
0.079
0.150
U.S. ECU
as percent
of total
7.7%
6.2%
8.3%
8.3%
8.7%
11.2%
2016 scenario
Total
0.29
0.20
0.36
0.63
0.87
1.29
U.S. EGU-
attributable
0.008
0.005
0.011
0.019
0.026
0.047
U.S. ECU
as percent
of total
2.9%
2.7%
2.9%
3.0%
3.0%
3.7%
% change (2016 versus
2005) in Hg fish tissue
concentration
Total
-6.7%
-10.0%
-6.4%
-5.9%
-4.7%
-3.7%
U.S. EGU-
attributable
-65%
-61%
-67%
-66%
-67%
-68%
Table 2-5. Comparison of U.S. EGU fraction of total Hg deposition (used to apportion Hg
fish tissue concentrations and risk) between the 2005 and 2016 scenarios. Note, that
these values are specifically for the 2,461 watersheds included in the risk assessment.
statistic
Mean
Median
75th %
90th %
95th %
99th %
U.S. EGU-attributable fraction of
total Hg fish tissue levels
2005 scenario
0.09
0.06
0.14
0.20
0.26
0.40
2016 scenario
0.04
0.03
0.05
0.07
0.09
0.18
Consideration of information presented above in Figures 2-5 through 2-14 and in Tables
2-4 through 2-5 resulted in the following observations regarding estimates of total and U.S.
EGU-attributable fish tissue MeHg concentrations across the 2,461 watersheds included in the
risk assessment:
• Focus on U.S. EGU-attributable Hgfish tissue levels is in the eastern half of the U.S.:
Given (a) that the number of watersheds with measured fish tissue MeHg data is
substantially greater in the east (see Figure 2-5) and (b) more importantly, that the levels
of U.S. EGU Hg deposition (that largely drives U.S. EGU-attributable Hg fish tissue
levels) are much higher in the east (see Figures 2-3 and 2-4), trends in U.S. EGU-
attributable Hg fish tissue levels discussed here are driven by data in the eastern half of
the U.S.
• U.S. EGUs contribute a larger fraction to total Hgfish tissue levels in the U.S. than they
do to total Hg deposition (in terms of percent), this reflects the fact that Hgfish tissue
samples are focused in the eastwhere U.S. EGU deposition is greater. While U.S. EGUs
43
-------
contribute -5% of total Hg deposition in the U.S. (for the 2005 scenario - see Table 2-2),
their contribution to Hg fish tissue levels (summarized at the watershed-level) for the
2005 scenario is larger at -9% (see Table 2-5). This reflects the fact that Hg fish tissue
samples are heavily weighted in the eastern portion of the U.S. where U.S. EGU Hg
deposition is typical higher than in the west.35 By providing greater coverage for the
eastern half of the country, the Hg fish tissue sampling data generally provides greater
coverage for regions with potentially greater U.S. EGU-attributable risk.
• Relative to the combined impact of other sources, U.S. EGUs represent a smaller, but
still potentially important contributor to total fish tissue MeHg levels: U.S. EGUs
contribute -9% of Hg fish tissue levels on average under the 2005 scenario (see Table 2-
5). Under the 2016 scenario, the U.S. EGU contribution decreases to ~ 4% on average
(see Table 2-5). While U.S. EGU-attributable Hg fish tissue decreases notably between
the 2005 and 2016 scenarios, the impact on total Hg fish tissue levels is not that
noticeable given that U.S. EGUs contribute a relatively small fraction on total Hg fish
tissue levels in general (contrast the pattern of reduction seen in Figures 2-9 and 2-10 for
U.S. EGU-attributable Hg fish tissue levels with the relatively smaller changes seen in
Figures 2-7 and 2-8 for total Hg fish tissue levels).
• Despite the relatively small fraction of total fish tissue MeHg associated with U.S. EGUs
on average, for a subset of water sheds, they can make a substantially larger contribution:
Under the 2005 scenario, U.S. EGUs can range up to 40% of total Hg fish tissue levels
(for the 99th% watershed). Under the 2016 Scenario, this pattern is reduced, but U.S.
EGUs can still contribute up to 18% of total Hg fis tissue levels (again, for the 99th%
watershed) (see Table 2-5).
2.5. Comparing Patterns of Hg Deposition with Hg Fish Tissue Data for the 2,461
Watersheds Included in the Risk Assessment
In addition to the observations provided in the last two sections based on consideration of
the CMAQ-based Hg deposition estimates and Hg fish tissue data separately, it is also possible to
directly compare spatial patterns between these two sets of data. This comparison provides
information that can help in interpreting risk estimates discussed in Section 2.6. Specifically we
can consider: (a) whether the watershed-level Hg fish tissue levels are positively correlated with
total Hg deposition, (b) how patterns of Hg deposition for the 2,461 watersheds where we have
Hg fish tissue data compare with patterns for the full set of 88,000 watersheds in the U.S. and (c)
to what extent the watersheds for which we have Hg sampling data provide coverage for areas of
elevated U.S. EGU deposition. To address these questions, we have presented a series of figures
and tables below including:
35 As discussed in section 1.3, U.S. EGU-attributable Hg fish tissue levels are directly based on U.S. EGU Hg
deposition (at the watershed-level) together with application of the MMaps approach.
44
-------
• Figures 2-15 and 2-16: For the 2005 and 2016 scenarios respectively, maps showing
areas of elevated U.S. EGU-related Hg deposition36 and the degree to which the 2,461
watersheds with fish tissue sampling data used in the risk assessment provide coverage
for these areas.
• Figure 2-17: Provides plot for the 2005 Scenario of Hg fish tissue levels versus total Hg
deposition by watershed. This plot allows consideration for whether there appears to be a
correlation between these two factors at the watershed level.
• Figure 2-18: Presents cumulative distribution plots comparing U.S. EGU-attributable
deposition for the 2,461 watersheds used in the risk assessment with U.S. EGU-
attributable deposition of the entire set of-88,000 watersheds in the U.S. Separate sets
of plots are provided for the 2005 and 2016 Scenarios, allowing trends for each scenario
to be compared to each other. These plots allow us to consider whether the watersheds
with fish tissue MeHg data tended to fall in regions with higher U.S. EGU-attributable
Hg deposition and the degree to which this subset of watersheds provided coverage for
areas with relatively elevated U.S. EGU mercury deposition across the country.
36 Areas of "elevated U.S. EGU-related Hg deposition" refer to areas that are at or above the average deposition
level seen in watersheds with U.S. EGU-attributable exposures above the MeHg RfD. Specifically, we used
exposure estimates based on the 95th percentile fish consumption rate (for the female high consumer scenario
assessed nation-wide) to identify watersheds with U.S. EGU-attributable exposures above the MeHg RfD and then
queried for the average U.S. EGU-related Hg deposition across that subset of watersheds. This average deposition
rate differed for the 2005 and 2016 Scenarios (i.e., 3.79 and 1.28 ug/m2, respectively). These values were used as the
basis for identifying watersheds with levels of U.S. EGU-related Hg deposition for the 2005 and 2016 Scenarios
presented in Figures 2-13 and 2-14.
45
-------
:omparison of watersheds
with Hg fish tissue data and
watersheds with elevated US EGU-attritutable
Hg depositon (2005)
watersheds with relatively elevated US ECU Hg dep
Figure 2-15. For the 2005 scenario, comparison of coverage of watersheds with Hg fish
tissue data (used in the risk assessment) for areas in the eastern U.S. with relatively
elevated U.S. EGU-attributable Hg deposition.
Comparison of watersheds
with Hg fish tissue data and
watersheds with elevated US EGU-attritutable
Hg depositor? (2016)
^^| waterbodies included in the risk assessment
'prli wth relatively elevated US EGU Hg dep
Figure 2-16. For the 2016 Scenario, comparison of coverage of watersheds with Hg fish
tissue data (used in the risk assessment) for areas in the eastern U.S. with relatively
elevated U.S. EGU-attributable Hg deposition.
46
-------
• *
..
*«•/*•• *•«•* •". * % * * *
. ••*: .»: *.*. ».* * * . .•••
••* **.•.•"."•"•••.. " \$\ •*;•**..' '.- *
2005 Hg Deposition
Figure 2-17. For the 2005 scenario, plot of total Hg fish tissue concentrations versus total
Hg deposition for the 2,366 watersheds included in the risk assessment for the high-
end female consumer population.
Cumulative Distribution of Mercury Deposition by Watershed
Hg deposition {ug/m )
-2005 EGU fish tissue waterheds
"2016 EGU fish tissue watersheds
-2005 EGU all watersheds
-2016 EGU all watersheds
Figure 2-18. Cumulative distribution plots of U.S. EGU-attributable Hg deposition over
the 2,366 watersheds used in modeling the high-end female consumer population as
contrasted with all 88,000 watersheds (plots provided both for the 2005 and 2016
Scenarios).
47
-------
Consideration of information presented above in Figures 2-15 through 2-18 resulted in
the following observations regarding how estimates of Hg deposition estimates relate to
measured fish tissue MeHg levels, when considered at the watershed-level:
• The fish tissue MeHg sampling data (summarized at the watershed-level) provides limited
coverage for areas with elevated U.S. EGUHg deposition. Therefore, the number of "at
risk" watersheds as characterized in this risk assessment may be substantially higher
than estimated: As depicted in Figures 2-15 and 2-16 (for the 2005 and 2016 Scenarios,
respectively), while the 2,461 watersheds used in the risk assessment to fall into regions
in the east with elevated U.S. EGU Hg deposition, the degree of coverage is limited. This
can be seen by noting in these figures the wide expanses of areas of elevated U.S. EGU
Hg deposition (shown in red) and therefore, not covered by watersheds modeled for risk
in the analysis.37
• Hgfish tissue levels are not correlated with total Hg deposition (the relationship is highly
dependent on methylation potential of individual water bodies): As shown in Figure 2-17,
total Hg fish tissue levels (summarized at the watershed-level) are not correlated with
levels of total Hg deposition when looking across watersheds (i.e., the highest total
mercury deposition watersheds does not always have the highest fish tissue MeHg
levels). This is not unexpected given that the relationship between total Hg deposition
and total Hg fish tissue levels is highly dependent on the methylation potential at the
waterbody-level. As discussed above in Appendix E, a variety of factors that display
spatial variability are associated with methylation potential (e.g., pH, sulfate deposition,
turbidity etc). Therefore, we would anticipate that there would not be a direct correlation
between total Hg deposition and Hg fish tissue levels, again looking across watersheds.
The MMaps approach and underlying analyses (see Section 1.3 and Appendix E), support
a proportional relationship between mercury deposition and fish tissue MeHg levels
within a given watershed, such that changes in deposition will be reflected in changes in
fish tissue levels. In other words, the correlation between Hg deposition and fish tissue
MeHg concentrations does not appear to hold between watersheds, but is espected tos
hold within a given watershed.
• Hgfish tissue samples were generally collected in regions with elevated total Hg
deposition: As demonstrated in Figure 2-18, Hg fish tissue sampling appears to have
37 We completed a follow-on assessment to help interpret the significance of the group of watersheds with "elevated
U.S. EGU deposition" that were not covered in the risk assessment (i.e., areas shown in red in Figures 2-15 and 2-
16). Specifically, we were interested in knowing, for the subset of 2,366 watersheds we did assess for risk for the
female high-end consumer, what is the percentage that (a) had Hg deposition above the threshold identified here for
"elevated deposition" (i.e., 3.79 and 1.28 ug/m2, respectively for the 2005 and 2016 Scenarios) and (b) had U.S.
EGU incremental risk of an HQ =1.5. This percentage lets us know, for the watersheds we modeled, what fraction
of watersheds with elevated U.S. EGU Hg deposition also had U.S. EGU-incremental risk representing a potential
public health hazard. The results of the assessment are 37% and 9%, respectively for the 2005 and 2016 Scenarios.
We can consider use these estimates to help interpret the areas in Figures 2-15 and 2-16 that have elevated U.S.
EGU Hg deposition and are NOT included in the risk assessment. If these watersheds tracked the pattern seen in the
watersheds we modeled, then we would expect to see -40% of the red highlighted areas in Figure 2-15 translate into
relatively elevated U.S. EGU incremental risk watersheds and -10% of the red highlighted watersheds in Figure 2-
16. Note, however, that there is substantial uncertainty in extrapolating trends seen across our modeled watersheds
to the non-modeled watersheds with elevated U.S. EGU Hg deposition..
48
-------
favored areas with relatively higher total Hg deposition. This can be seen by comparing
cumulative plots of modeled watersheds (where we have fish tissue MeHg data) against
plots for the entire set of 88,000 watersheds. This comparison suggests that watersheds
where fish tissue MeHg data were collected tended to have higher total Hg deposition,
then the full set of watersheds. This likely reflects to some extent, the fact that fish tissue
sample are focused in the eastern half of the country, which does have elevated total Hg
deposition compared to the broad central region (see Figure 2-1 and 2-2).
2.6. Overview of Risk Estimates
This section provides an overview of risk estimates generated for the 2,461 watersheds
included in the risk assessment. As noted earlier in Section 1.2, presentation of risk estimates
will focus on the high-end female consumer population assessed at the national-level, since this
population provides the most comprehensive coverage for watersheds with Hg fish tissue data
across the U.S. and because the consumption rates used to model this population represent
subsistence levels and are supported by a number of studies (see Section 1.3 and Appendix D).
While this fisher population is emphasized in summarizing risk estimates, we do provide risk
estimates generated for the other populations covered in the analysis (e.g., blacks in the
southeast, Tribal populations near the Great Lakes, Hispanics). In summarizing risk estimates,
we will provide estimates for both the 2005 and 2016 Scenarios, while placing the most
emphasis on the 2016 estimates. The remainder of this section is organized as follows:
• Overview of percentile risk estimates generated for the different fisher populations
evaluated (Section 2.6.1): In this section, we provide percentile risk estimates (for HQ
and IQ loss risk) for the high-end female consumer population assessed at the national
level. We then summarize percentile risk estimates (HQ only) for the broader set of fisher
populations assessed in the analysis. The percentile risk estimates provided in this section
allow us to (a) consider Stage 2 of the 3-Stage framework developed to support the
interpretation of risk estimates (i.e., consider the U.S. EGU-related increment of total risk
- see Section 1.2) and (b) consider the magnitude of risk across the set of fisher
populations assessed in the risk assessment.
• Overview of the number (and frequency) of watersheds with populations potentially at
risk due to U.S. EGU-sourced mercury (Section 2.6.2): This set of risk estimates provides
the main input to the risk characterization framework (see Section 1.2). Specifically,
watersheds with populations potentially at risk comprise:
o Watersheds where total risk is considered to represent a public health concern and
where U.S. EGUs contribute to that total risk (in the analysis, we considered
various increments of U.S. EGU contribution including 5%, 10%, 15% and 20%,
although as noted in Section 1.2, we focus on 5%). This represents Stage 1 of the
rick characterization framework AND/OR,
o Watersheds where risk when considering U.S. EGUs mercury emissions before
considering other sources of mercury represents a potential public health hazard
(this is Stage 2 in the risk characterization framework)
49
-------
To support the discussion of risk estimates, a series of tables summarizing those estimates
are presented in the subsections below. A list of observations based on consideration for the risk
estimates summarized is presented at the end of each subsection.
2.6.1. Overview of percentile risk estimates
In presenting percentile risk estimates in this section, we have sorted risk estimates by
U.S. EGU-attributable risk in order to track trends in the magnitude of this category of risk.38
The percentile estimates themselves, are based on risk bands around the percentile value (rather
than being based on the specific watershed at that percentile). Specifically, we have taken the
average of the 5% of watershed values surrounding the specific percentile estimate in the table.
So, for example, for a 50th% risk value presented in the table, we have actually taken the average
of the 47.5th through 52.5th percentile watershed-level risk estimates (after ranking them by U.S.
EGU-attributable risk, as mentioned earlier). We used this risk band approach, rather than using
the estimates from the singe watershed located specifically at that percentile, because we wanted
to capture general trends in the patterns of U.S. EGU-attributable and total risk for watersheds
around that percentile.39 Note also, that while this risk assessment does focus on subsistence
levels offish consumption, the risk tables summarized here do include risk estimates based on
mean fish consumption rates for these higher consuming populations, which are relatively high
compared to general recreational angler rates. The following tables are used to summarize risk
estimates for the fisher populations included in the analysis.
• Tables 2-6 and 2-7: Presents risk percentiles for both IQ loss and RfD-based HQs for the
high-end female consumer assessed at the national-level (for 2,366 watersheds) for the
2005 and 2016 Scenarios, respectively. We do note that overall confidence in IQ loss
estimates above approximately 7 points decreases because we begin to apply the
underlying IQ loss function at exposure levels (ppm hair levels) above those reflected in
epidemiological studies used to derive those functions.40 We have flagged IQ loss
estimates above 7 as subject to greater uncertainty.
• Table 2-8: Provides risk percentiles for RfD-based HQs for the remaining fisher
populations assessed in this analysis specifically for the 2005 scenario. We did not
evaluate these populations for the 2016 Scenario, since the relative magnitude of risks for
these additional populations can be inferred by comparing risks for these populations
against the risk generated for the high-end female consumer population (for the 2005
38 This means that when reviewing the risk estimates presented in tables in this section, the trend in risk across
watershed percentiles (i.e., higher percentiles will have higher risk) will be seen for the U.S. EGU-attributable
portion of the risk and not necessarily for total risk, since the watersheds were ranked on U.S. EGU-attributable risk
and on total risk, prior to generating the percentile summaries provided in the tables.
39 Note, that watersheds can display considerable variation in the relationship between total and U.S. EGU-
attributable risk. Therefore, if we had selected the specific watershed associated with a given percentile, that
watershed could misrepresent the general trend (in terms of the relationship between total and U.S. EGU-attributable
risk) for watersheds in the vicinity of that percentile. By taking average values for these "risk bands" around each
percentile, we get more stable and meaningful results in terms of capturing trends in risk estimates across the
percentiles presented in the risk tables.
40 The 39.1 ppm was the highest measured ppm level in the Faroes Island study, while ~86 was the highest value in
the New Zealand study (USEPA, 2005) (a 7 IQ points loss is approximately associated with a 40 ppm hair level
given the concentration-response function we are using).
50
-------
scenario) as summarized in Table 2-8. Similarly, we did not calculate IQ loss estimates,
since these are linearly related to HQ loss and can be inferred from the estimates
presented in Table 2-8 (based on comparison of magnitude of HQ estimates between
different fisher populations).
Table 2-6. Percentile risk estimates for the high-end female consumer population assessed
nationally (2005 scenario) (for both total and U.S. EGU incremental risk, including IQ
loss and MeHg RfD-based HQ estimates)
IQ
Fisher
consumption
rate percentile
and rate (g/day)
Watershed percentile
50th percentile
total
U.S.
EGU
75th percentile
total
U.S.
EGU
90th percentile
total
U.S.
EGU
95th percentile
total
U.S.
EGU
99th percentile
total
U.S.
EGU
IQ loss (points)
mean (3 9)
90th (123)
95th (173)
99th (373)
0.8
2.5
3.5
7.6
-
0.2
0.3
0.7
0.7
2.3
3.2
6.9
-
0.3
0.4
0.9
1.0
3.0
4.2
9.1*
0.1
0.4
0.5
1.1
1.2
3.9
5.4
11.7*
0.2
0.5
0.7
1.6
1.6
5.1
7.2*
15.5*
0.3
1.0
1.4
2.9
RfD-based HQ
mean (3 9)
90th (123)
95th (173)
99th (373)
3.5
11.1
15.6
33.6
0.3
1.1
1.5
3.3
3.2
10.1
14.2
30.7
0.4
1.3
1.8
3.9
4.2
13.3
18.7
40.3
0.5
1.6
2.3
4.9
5.4
17.1
24.1
51.9
0.7
2.3
3.2
6.9
7.2
22.8
32.0
69.0
1.4
4.3
6.0
13.0
loss is <0.1 point
: IQ loss estimate subject to greater uncertainty due to application of the underlying concentration-response function
for IQ loss at levels of exposure above those in the underlying epidemiological studies (see text)
Table 2-7. Percentile risk estimates for the high-end female consumer population assessed
nationally (2016 Scenario) (for both total and U.S. EGU incremental risk, including IQ
loss and RfD-based HQ estimates)
Fisher
consumption
rate percentile
and rate (g/day)
Watershed percentile
50th percentile
total
U.S.
EGU
75th percentile
total
U.S.
EGU
90th percentile
total
U.S.
EGU
95th percentile
total
U.S.
EGU
99th percentile
total
U.S.
EGU
IQ loss (points)
mean (3 9)
90th (123)
95th (173)
99th (373)
0.6
1.7
2.4
5.3
-
-
-
0.1
0.6
2.0
2.8
6.0
-
-
-
0.2
1.0
3.2
4.5
9.8*
-
0.1
0.2
0.4
1.3
4.0
5.6
12.1*
-
0.2
0.2
0.5
1.8
5.6
7.9*
16.9*
0.1
0.4
0.5
1.1
RfD-based HQ
mean (3 9)
90th (123)
2.5
7.7
-
0.2
2.8
8.8
-
0.3
4.6
14.4
0.2
0.5
5.6
17.7
0.2
0.8
7.9
24.8
0.5
1.6
51
-------
95th (173)
99th (373)
10.9
23.4
0.2
0.5
12.3
26.6
0.4
0.9
20.2
43.6
0.8
1.7
24.9
53.6
1.1
2.3
34.9
75.3
2.2
4.8
-IQ loss is <0.1 point
* IQ loss estimate subject to greater uncertainty due to application of the underlying concentration-response function
for IQ loss at levels of exposure above those in the underlying epidemiological studies (see text)
Table 2-8. Percentile risk estimates for the full set of fishing populations included in the
analysis (2005 scenario) (for both total and U.S. EGU incremental risk, only for
RfD-based HQ estimates)
Fisher
consumption
rate percentile
and rate (g/day)
Watershed percentile
50th percentile
total
U.S.
EGU
75th percentile
total
U.S.
EGU
90th percentile
total
U.S.
EGU
95th percentile
total
U.S.
EGU
99th percentile
total
U.S.
EGU
High-end female consumer assessed nationally
mean (3 9)
90th (123)
95th (173)
99th (373)
3.5
11.1
15.6
33.6
0.3
1.1
1.5
3.3
3.2
10.1
14.2
30.7
0.4
1.3
1.8
3.9
4.2
13.3
18.7
40.3
0.5
1.6
2.3
4.9
5.4
17.1
24.1
51.9
0.7
2.3
3.2
6.9
7.2
22.8
32.0
69.0
1.4
4.3
6.0
13.0
Poor white fishers in the Southeast
mean (3 9)
90th (93)
95th (129)
99th (286)
5.0
12.0
16.6
36.9
0.5
1.1
1.6
3.5
4.9
11.6
16.1
35.8
0.6
1.4
1.9
4.2
7.3
17.6
24.4
54.1
0.7
1.7
2.4
5.3
7.3
17.5
24.3
53.9
1.0
2.5
3.4
7.5
8.8
21.1
29.2
64.7
1.6
3.9
5.5
12.1
Poor black fishers in the Southeast
mean (171)
90th (446)
95th (557)
99th
19.0
49.6
61.9
1.9
4.9
6.1
24.7
64.4
80.5
2.3
5.9
7.4
26.2
68.2
85.2
2.9
7.6
9.5
39.6
103.3
129.0
4.2
10.9
13.6
40.0
104.3
130.3
7.2
18.8
23.4
NC*
Poor Hispanic nationally
mean (26)
90th (98)
95th (156)
99th
1.5
5.5
8.8
-
0.3
0.5
2.3
8.6
13.6
0.2
0.7
1.1
3.0
11.6
18.4
0.3
1.2
2.0
3.6
13.6
21.7
0.5
1.8
2.8
6.2
23.4
37.2
0.9
3.5
5.5
NC*
Vietnamese
mean (27)
90th (99)
95th (152)
99th
1.4
5.2
8.0
-
0.3
0.5
3.0
10.9
16.8
0.2
0.6
1.0
1.7
6.2
9.5
0.3
1.0
1.6
2.8
10.4
15.9
0.4
1.4
2.2
4.6
16.8
25.8
0.9
3.1
4.8
NC*
Laotians
mean (47)
90th (145)
95th (266)
99th
2.2
6.6
12.2
0.1
0.4
0.7
3.3
10.2
18.8
0.3
0.9
1.6
2.0
6.0
11.1
0.4
1.3
2.4
3.9
12.0
22.0
0.5
1.5
2.7
6.5
20.0
36.7
1.5
4.6
8.5
NC*
Tribal (near Great Lakes)
mean (62)
90th (136)
4.1
8.9
0.1
0.2
7.4
16.2
0.2
0.4
9.4
20.6
0.4
0.9
8.7
19.1
0.6
1.3
6.9
15.0
0.9
2.0
52
-------
Fisher
consumption
rate percentile
and rate (g/day)
95th (213)
99th (493)
Watershed percentile
50th percentile
total
13.9
32.2
U.S.
ECU
0.4
0.9
75th percentile
total
25.4
58.6
U.S.
ECU
0.6
1.4
90th percentile
total
32.2
74.5
U.S.
ECU
1.4
3.2
95th percentile
total
29.9
69.2
U.S.
ECU
2.0
4.5
99th percentile
total
23.5
54.4
U.S.
ECU
3.2
7.4
- IQ loss orHQ estimates <0.1.
NA*: It was not possible to derive a 99th% consumption rate for this population due to insufficient sample size in the
underlying study. Consequently, risk estimates for the 99*% consumption rates were not generated.
Observations regarding the percentile risk estimates presented here reflect our
interpretation of the potential health significance associated with both MeHg RfD-based HQ and
IQ loss estimates (see Section 1.2). Consideration of the risk estimates summarized above in
Tables 2-6 through 2-8 results in the following observations regarding percentile risk estimates
generated for the fisher populations assessed:
• For the high-end female consumer assessed at the national-level, total IQ loss and total
HQ estimates do not change in a systematic way between the 2005 and 2016 Scenarios
with these levels often being of potential health concern across a wide variety of
consumption rates and watershedpercentiles: While there are some differences in total
IQ loss and HQ estimates generated for the high-end female consumer assessed
nationally between the 2005 and 2016 Scenarios, there is no systematic trend between the
scenarios (see Tables 2-6 and 2-7). Furthermore, for the high-end female consumer
assessed at the national-level, these estimates of total risk often exceed one IQ point loss
and an HQ of 1.5 (i.e., levels of potential health concern) across most of the combinations
of consumption rates and watersheds (see Tables 2-6 and 2-7). The absence of a
substantial change in total risk between the two simulation years is not surprising given
the relatively small fraction of total mercury deposition contributed by U.S. EGUs on
average across the modeled watersheds. This means that even substantial reductions in
U.S. EGU deposition between the simulation years is unlikely to substantially affect total
risk (although, as noted elsewhere, it can have a substantial impact on risk at the subset of
watersheds where U.S. EGUs do contribute a larger fraction of total deposition).
• By contrast (again focusing on the high-end female consumer assessed nationally), both
U.S. EGU-incremental IQ loss and the U.S. EGU increment-based HQ display notable
reductions between the 2005 and 2016 Scenarios, but U.S. EGU-attributable risk still
exceeds potential levels of concern for a over a quarter of watersheds: Comparison of
the U.S. EGU-attributable risk estimates presented in Tables 2-6 and 2-7, suggest that
these categories of risk decrease significantly between the 2005 and 2016 Scenarios. As
noted earlier, this reduction largely reflects the implementation of PM controls which
have the co-benefit of reducing divalent and particle-bound mercury together with state
regulations targeting mercury emissions directly. As noted in Section 2.2, because current
(2010) emissions of mercury are likely closer to levels used in modeling the 2016
Scenario (with the 2005 scenario reflecting emission levels that are substantially larger
than current conditions), we focus here on presenting observations based on the 2016
53
-------
Scenario. For the high-end female consumer assessed at the national-level, under the
2016 Scenario (see Table 2-7), U.S. EGU-attributable IQ loss only meets or exceeds one
point for the highest consumption range at the upper 1% of the watersheds. And, given
the relatively high total exposure associated with this simulation, our calculation of IQ
(including the U.S. EGU-attributable portion) for this combination of consumption rate
and watershed percentile is subject to increased uncertainty.41 By contrast, for the 2016
Scenario, estimates of U.S. EGU increment-based HQ for this fisher population exceed
1.5, although these exceedences are still limited to combinations of higher percentiles of
watersheds and consumption rates (e.g., 95th and 99th % consumption rates paired with
the 95th and 99th % watersheds - see Table 2-7). These HQ-based risk estimates can be
used to inform Stage 2 of the 3-Stage risk characterization framework. Specifically, for
the 2016 Scenario, a fraction of watersheds (top 5 to 10%) have U.S. EGU increment-
based HQ's that exceed 1.5 based on modeling subsistence-level fish consumption for the
high-end female consumer assessed at the national-level (see Table 2-7).
Estimates of risks generated for the high-end female consumer population (assessed at
the national-level) are generally higher than risks estimated for the other high-end fisher
populations, with the exception of white and black fisher populations assessed in the
southeast: Risk estimates generated for the 2005 scenario for the set of high-end fisher
populations assessed in this analysis suggest that risks (across all combinations of
consumption rates and watersheds) are generally higher for the high-end female
consumer population assessed at the national-level with the exception of black and white
fishers assessed in the southeast (contrast estimates presented in Table 2-8). For example,
high-end female consumer risk estimates assessed at the national-level are approximately
twice as high (in terms of both total and EGU-attributable) as estimates generated for
Hispanics and Vietnamese also assessed nationally (see Table 2-8). Risk estimates for the
high-end female consumer are approximately 50% higher than estimates generated for
Laotians assessed nationally and Tribal populations in the vicinity of the Great Lakes.
However, risks (both total and U.S. EGU-attributable) for white fishers in the southeast
are somewhat higher that risk estimates generated for the high-end female consumer at
the national-level, while estimates for black populations assessed in the southeast are
notably higher. Risk estimates for these two southeastern fisher populations are likely
higher due to: (a) the fact that the southeast has relatively higher total and U.S. EGU-
attributable fish tissue concentrations compared with the full set of watersheds with fish
tissue levels nationally (see Figures 2-1 through 2-4) (this means that upper-end
percentiles fish tissue values will be higher in the southeast) and (b) in the case of the
black fisher populations, percentile consumption rates are substantially higher than
consumption rates for the high-end female consumer population (assessed at the national-
level) see Table 2-8, "Fisher consumption rate percentile rate (g/day)" column. While
risk estimates are higher for the two populations assessed in the southeast, we decided to
focus the discussion of risk estimates in both Section 1.6.1 and 1.6.2, on the high-end
female consumer assessed at the national-level. As noted earlier, this reflects several
41 As noted in section 2.6.1, estimates of total IQ loss above 7 points, involve simulation of mercury hair levels that
exceed those used in the epidemiological studies underlying the function used in modeling of IQ loss and therefore
are subject to greater uncertainty (not in the potential for IQ loss to occur, but rather in our ability to quantify
degrees of loss above 7 points).
54
-------
factors: (a) focusing risk estimates on national-level analysis allows us to consider a
larger number of watersheds reflecting greater regional variability in factors related to
fish consumption exposure and risk (e.g., methylation potential, fish species), (b) risk
estimates in the southeast are driven in part by watersheds in SC and we have concerns
over the potential for these estimates having non-air Hg contributions (from gold mining
- see Appendix E), and (c) fish consumption rates for both of the southeastern-focused
populations (white and black) are based on smaller sample sizes compared with the
estimate for the high-end female consumers and therefore, we have greater confidence in
the consumption estimates generated for the high-end female consumer population.42
2.6.2. Overview of number (and frequency) of watersheds with populations
potentially at-risk due to U.S. EGU mercury emissions
This section discusses risk estimates based on identifying the number of watersheds with
populations potentially at risk due to mercury released from U.S. EGUs. As noted in Section 1.2,
the "at risk population" classification is based on identifying watersheds where: (a) U.S. EGUs
contribute to total risk at watersheds where that total risk is considered to represent a potential
public health hazard and/or (b) risk at the watershed-level represents a potential public health
hazard when deposition from U.S. EGUs is considered before taking into account deposition and
exposures resulting from other sources of mercury. The estimates of watersheds with at-risk
populations discussed in this section are used in the 3-Stage risk characterization framework
described in Section 1.2 for interpreting risk estimates. Specifically, the first category of at risk
populations described above maps to Stage 1 of the 3-Stage approach, while the second category
maps to Stage 2. The combination (i.e., mathematical union) of these two groups of watersheds
with at risk populations comprises the set of watersheds represented in Stage 3 of the framework.
We note that, specifically with regard to the HQ estimates, any contribution of mercury
from EGUs to watersheds with exposures exceeding the MeHg RfD represents a potential hazard
to public health, but for purposes of this analysis we have focused on those waterbodies where
we determined EGUs contributed 5% or more to the hazard. We think this is a conservative
approach given the increasing risks associated with incremental exposures above the MeHg RfD.
The estimates of watersheds with potentially at-risk populations discussed in this section
are all based on the underlying risk estimates generated for the high-end female consumer
population. The decision to focus on this fisher population in discussing risk estimates is
discussed in the previous section. The estimates of watersheds with potentially at-risk
populations are summarized in tables described below (note, observations based on consideration
of these risk estimates are presented in bullets following the tables).
• Tables 2-9: Identifies watersheds with potentially at-risk populations based on
consideration for different degrees of U.S. EGU contribution (i.e., 5, 10, 15 and 20%) at
watersheds where total risk is considered to represent a potential public health hazard
(i.e., meet or exceed 1 IQ point or an HQ of 1.5 or higher). For reference purposes, the
table also identifies the total number of watersheds (out of the 2,366 assessed for the
42 Sample size used in computing consumption rate percentiles for the female population is 149, as contrasted with
39 and 98 for poor blacks and whites, respectively (Burger et al., 2010).
55
-------
high-end female consumer population) with total risk exceeding the two thresholds,
regardless of the U.S. EGU percent contribution (see the "> 0%" row of results in the
table). In presenting results, the tables include both the number of watersheds meeting
specific criteria as well as the percent (of the 2,366 watersheds assessed) that this
represents. (Stage 1 of the 3-Stage framework)
• Tables 2-10: Identifies watersheds with potentially at-risk populations based on
consideration for the magnitude of risk when deposition from U.S. EGUs is considered
before taking into account deposition and exposures resulting from other sources of
mercury. Risks are presented both for U.S. EGU-attributable IQ loss and for U.S. EGU
increment-based HQ. (Stage 2 of the 3-Stage framework)
• Table 2-11: Presents the union of the two categories of watersheds with potentially at-risk
populations (i.e., mathematical union of the Stage 1 and 2 estimates presented in Tables
2-9 and 2-10). As noted earlier, this analysis focuses on MeHg RfD-based HQ estimates
rather than IQ loss estimates, since the HQ-based risk estimates generate a larger
percentage of watersheds with populations potentially at-risk compared with the IQ loss
estimates. Consequently, Table 2-11 considers the number and percent of watersheds that
have (a) U.S. EGUs contributing to total risk of an HQ > 1.5, OR (b) an HQ > 1.5 based
on considering U.S. EGU mercury deposition alone, before factoring in other sources of
mercury deposition. This represents Stage 3 of the risk characterization framework.
Table 2-9. Watersheds with potentially at-risk populations based on consideration for
various degrees of U.S. EGU contribution to total risk (for IQ loss and HQ)
EGU risk
threshold
Number and percentage of HUCs meeting risk threshold criteria
2005 analysis
90th fish
consumption
95th% fish
consumption
99th% fish
consumption
2016 analysis
90th fish
consumption
95th% fish
consumption
99th% fish
consumption
Total IQ points lost = 1
>0%
>5%
> 10%
> 15%
>20%
1667
948
609
345
167
(70%)
(40%)
(26%)
(15%)
(7%)
1948
1124
726
420
202
(82%)
(48%)
(31%)
(18%)
(9%)
2224
1287
829
473
225
(94%)
(54%)
(35%)
(20%)
(10%)
1558
288
43
15
7
(66%)
(12%)
(2%)
*
*
1858
394
69
27
16
(79%)
(17%)
(3%)
*
*
2202
495
94
37
20
(93%)
(21%)
(4%)
(2%)
*
Total RfD-based HQ = 1.5*
>0%
>5%
> 10%
> 15%
>20%
2191
1268
816
471
223
(93%)
(54%)
(34%)
(20%)
(9%)
2266
1305
834
475
226
(96%)
(55%)
(35%)
(20%)
(10%)
2348
1345
853
480
228
(99%)
(57%)
(36%)
(20%)
(10%)
2162
482
91
37
20
(91%)
(20%)
(4%)
(2%)
*
2248
505
95
38
20
(95%)
(21%)
(4%)
(2%)
*
2346
521
97
38
20
(99%)
(22%)
(4%)
(2%)
*
* Following convention for reporting HQ estimates to one significant digit, although this query is based on
HQ > 1.5, this translates from a science policy standpoint into an HQ > 2.
56
-------
Table 2-10. Watersheds with potentially at-risk populations based on consideration for
risk (both IQ loss and HQ) based on U.S. EGU mercury deposition and resulting
exposure before considering other sources of mercury deposition
EGU
risk
threshold
Number and percentage of 2,366 HUCs meeting risk threshold criteria*
2005
90th fish
consumption
analysis
95th% fish
consumption
99th% fish
consumption
2016 analysis
90th fish
consumption
95th% fish
consumption
99th% fish
consumption
IQ loss
>lpt
>2pts
28
1
*
*
67
5
(3%)
*
353
69
(15%)
(3%)
3
0
*
*
7
0
*
*
27
7
*
*
RfD-based HQ
>1.5**
277
(12%)
495
(21%)
1064
(45%)
17
*
48
(2%)
284
(12%)
* 2,366 watersheds reflected in this summary are those watersheds out of the 2,461 assessed for risk for the
high-end female consumer (at the national-level)
** Following convention for reporting HQ estimates to one significant digit, although this query is based
on HQ > 1.5, this translates from a science policy standpoint into an HQ > 2.
Table 2-11. Combination of watersheds with potentially at-risk populations based on
either consideration for (a) U.S. EGU percent contribution to total risk OR (b) risk
when U.S. EGU mercury deposition is considered alone
EGU risk
threshold
Number and percentage of HUCs meeting risk threshold criteria
2005 analysis
90th fish
consumption
95th% fish
consumption
99th% fish
consumption
2016 analysis
90*
consu
"fish
mption
95th% fish
consumption
99th% fish
consumption
V. S. EGU-attributable risk >1.5*HQOR total risk >1.5*HQandU. S. EGU contribution of:
>5%
> 10%
> 15%
> 20%
1271
879
584
391
(54%)
(37%)
(25%)
(17%)
1321
946
682
550
(56%)
(40%)
(29%)
(23%)
1477
1221
1109
1073
(62%)
(52%)
(47%)
(45%)
484
96
49
34
(20%)
(4%)
(2%)
*
524
121
76
61
(22%)
(5%)
(3%)
(3%)
672
325
292
286
(28%)
(14%)
(12%)
(12%)
* Following convention for reporting HQ estimates to one significant digit, although this query is based on
HQ > 1.5, this translates from a science policy standpoint into an HQ > 2.
Observations regarding watersheds with potentially at-risk populations due to U.S. EGU
sourced mercury are presented below. Note, that these observations focus primarily on results
generated for the 2016 Scenario, as explained earlier. In addition, this set of observations is
oriented around the 3-Stage risk characterization framework.
• Less than 1% of the watersheds have an IQ loss of 1 point when deposition from U.S.
EGUs is considered before taking into account deposition and exposures resulting from
other sources ofHg: While total IQ loss estimates (as described earlier) can extend above
2 IQ points for a substantial fraction of the watersheds modeled, less than 1% of the
57
-------
watersheds have U.S. EGU incremental IQ loss estimates extending into the 1 to 2 point
range. However, as noted in section 1.2, IQ does not fully cover all of the neurologic
domains such as motor skills and attention/behavior associated with MeHg exposure and
therefore, there are concerns that risk estimates based on IQ could underestimate overall
neurodevelopmental impacts (Axelrad et al., 2007). For this reason, in considering
watersheds with potentially at-risk populations due to U.S. EGU-attributable mercury, we
focus on MeHg RfD-based estimates of risk rather than IQ. Between 2 and 22% of those
watersheds with total risk HQs > 1.5 have U.S. EGUs contributing at least 5% of total
mercury deposition: With this risk metric, we consider the degree to which U.S. EGUs
contribute to total risk at watersheds where total risk represents a potential public health
hazard (i.e., total risk of an HQs > 1.5). Considering a 5% U.S. EGU contribution at
watersheds where total risk is considered a potential public health hazard, we have up to
22% of the watersheds falling into this category (with the 22% value reflecting risk
modeled using the 99th percentile fish consumption rate for the high-end female
consumer - see Table 2-9). It is important when considering this risk metric to reiterate
that any exposure above the MeHg RfD represents a potential public health hazard. .
• Between 2 and 12% of the watersheds have HQs > 1.5, based on U.S. EGU mercury
deposition before factoring in any other sources of mercury: Our analysis suggests that
between 2 and 12% of the 2,366 watersheds modeled in the risk assessment for high-end
female consumers could have an HQ > 1.5 when U.S. EGU mercury deposition and
resulting exposure are considered before other sources of mercury deposition. This range
reflects the range offish consumption rates considered for the high-end female consumer
(i.e., 95th or 99l percentile consumption rate, respectively - see Table 2-10).
• Combining (mathematical union) the two sets of watersheds with at-risk populations due
to U.S. EGU mercury emissions: Combining the two categories of watersheds with
populations at-risk due to U.S. EGU mercury emissions summarized in the last two
bullets, we get a total estimate ranging from 2 to 28% of watersheds, with this range
reflecting in part the U.S. EGU percent contribution that is considered (e.g., 5, 10, 15 or
20% - see Table 2-11). Note, that this range also reflects the different fish consumption
rates considered for the high-end female consumer (i.e., 90th, 95th and 99th percentile fish
consumption rates). The results summarized here for total "at risk" watersheds map to
Stage 3 of the 3-stage risk characterization framework.
2.7. Sensitivity Analyses
This section discusses several sensitivity analyses conducted to assess the potential
impact of key sources of uncertainty (all related to the application of the MMaps approach in our
analysis) on risk estimates. We note that, in designing the sensitivity analysis, we focused on
application of the MMaps approach because it represents a critical element of the analysis and is
acknowledged as representing a potentially important source of uncertainty. The sensitivity
analyses address two specific uncertainties related to application of the MMaps approach: (a)
concerns over including watersheds that may be disproportionately impacted by non-air mercury
sources and (b) application of the MMaps approach to both flowing and stationary freshwater
bodies (if, in reality, the approach is better at predicting source-apportioning Hg fish tissue levels
for stationary waterbodies).
58
-------
The sensitivity analyses addressing the first area of uncertainty (potential inclusion of
watersheds from regions with substantial Hg contributions from non-air deposition) included two
analyses: (a) constraining the risk analysis to only include those watersheds in the upper 25th
percentile with regards to total Hg deposition (i.e., watersheds with relatively elevated levels of
total Hg deposition so we knew this source of loading played a larger role) and (b) excluding
four states where we have concerns over the potential for non-air mercury playing a greater role
(ME, MN, SC and LA).43 The results of both sensitivity analyses are presented in terms of their
impact on the "at risk" metrics described in Section 2.6.2. Specifically, the presentation of the
sensitivity analysis results parallel the "at risk" summary table layout used in Section 2.6.2 and
as such, present risk estimates for the high-end female consumer (2005 scenario) (see Tables 2-
12 through 2-14).
The other area of uncertainty (application of the MMaps approach to both stationary and
flowing waterbodies) was assessed by running the risk assessment only for watersheds with fish
tissue MeHg values taken from stationary watersheds (i.e., ponds and lakes). Specifically, we
generated a set of percentile risk estimates, again for the high-end female consumer assessed
nationally, for the 2005 scenario (see Table 2-15). In presenting the results, we also include the
core risk estimates generated for this scenario using fish tissue MeHg samples taken from both
stationary and flowing waterbodies. Observations resulting from considering the results of the
sensitivity analyses are presented at the end of this section.
Table 2-12. Sensitivity analysis results based on constraining analysis to (a) watersheds in
the top 25th percentile with regard to total Hg deposition and (b) exclude watersheds
located in MN, LA, SE or ME) - Results for watersheds with potentially at risk
populations based on U.S. EGUs making a specified contribution to total risk (Stage 1 of
the 3 stage framework)
ECU risk
threshold
Number and percentage of HUCs meeting risk threshold criteria (2005 scenario)
2005 analysis
90th fish
consumption
95th% fish
consumption
99th% fish
consumption
2016 analysis
90th fish
consumption
95th% fish
consumption
99th% fish
consumption
Total HQ > 1.5 and U.S. ECU contribution:
ALL WATERSHEDS (2,366)
>0%
>5%
> 10%
> 15%
> 20%
2191
1268
816
471
223
(93%)
(54%)
(34%)
(20%)
(9%)
2266
1305
834
475
226
(96%)
(55%)
(35%)
(20%)
(10%)
2348
1345
853
480
228
(99%)
(57%)
(36%)
(20%)
(10%)
2162
482
91
37
20
(91%)
(20%)
(4%)
(2%)
*
2248
505
95
38
20
(95%)
(21%)
(4%)
(2%)
*
2346
521
97
38
20
(99%)
(22%)
(4%)
(2%)
*
43 The rational for excluding the four states in this sensitivity analysis examining the MMaps approach is as follows.
ME was excluded because Hg fish tissue levels there are fairly high, while Hg deposition is not relatively elevated
(compared to other eastern states) - this raising the concern that some other factor may be in play (e.g., other non-air
sources, or perhaps substantially increased methylation potential). MN was excluded for the same reason as ME
with additional concern for taconite mining which could provide non-air Hg loading. SC was also excluded due to
higher fish Hg levels and Hg air deposition that (while elevated in some locations) is not uniformly higher than other
states in the region (in addition, there is a history of gold mining in SC). Finally LA was excluded as part of the
sensitivity analysis due to concerns for the substantial industrial activity which could result in non-air Hg impacts.
59
-------
ECU risk
threshold
Number and percentage of HUCs meeting risk threshold criteria (2005 scenario)
2005 analysis
90th fish
consumption
95th% fish
consumption
99th% fish
consumption
2016 analysis
90th fish
consumption
95th% fish
consumption
99th% fish
consumption
SENSITIVITY ANALYSIS 1 - Watersheds in top 25th % with regard to total Hg deposition (591)
>0%
>5%
> 10%
> 15%
> 20%
554
323
182
122
64
(94%)
(55%)
(31%)
(21%)
(11%)
570
332
185
124
65
(96%)
(56%)
(31%)
(21%)
(11%)
584
341
192
126
67
(99%)
(58%)
(32%)
(21%)
(11%)
548
89
32
19
10
(93%)
(15%)
(5%)
(3%)
(2%)
564
90
32
19
10
(95%)
(15%)
(5%)
(3%)
(2%)
585
94
33
19
10
(99%)
(16%)
(6%)
(3%)
(2%)
SENSITIVITY ANALYSIS 2 - Watersheds excluding those in MN, LA, SC and ME (1,844)
>0%
>5%
> 10%
> 15%
> 20%
1673
1133
791
460
220
(91%)
(61%)
(43%)
(25%)
(12%)
1745
1170
809
464
223
(95%)
(63%)
(44%)
(25%)
(12%)
1827
1210
828
469
225
(99%)
(66%)
(45%)
(25%)
(12%)
1645
449
77
33
19
(89%)
(24%)
(4%)
(2%)
*
1727
472
81
34
19
(94%)
(26%)
(4%)
(2%)
*
1825
488
83
34
19
(99%)
(26%)
(5%)
(2%)
*
Table 2-13. Sensitivity analysis results based on constraining analysis to (a) watersheds in
the top 25th percentile with regard to total Hg deposition and (b) exclude
watersheds located in MN, LA, SE or ME) - Results for watersheds with potentially at
risk populations based on U.S. EGU-incremental contribution to total risk (Stage 2 of the
3 stage framework)
ECU
risk
threshold
Number and
2005
90th fish
consumption
percentage of HUCs meeting risk threshold criteria (2005 scenario)
analysis
95th% fish
consumption
99th% fish
consumption
2016 analysis
90th fish
consumption
95th% fish
consumption
99th% fish
consumption
RfD-based HQ)
ALL WATERSHEDS - core analysis (2,366)
HQ>1.5
277
(12%)
495
(21%)
1064
SENSITIVITY ANALYSIS 1 - Watersheds in top
HQ>1.5
90
(15%)
140
(24%)
274
(45%)
17
*
48
(2%)
284
(12%)
25th % with regard to total Hg deposition (591)
(46%)
8
*
25
(4%)
90
(15%)
SENSITIVITY ANALYSIS 2 - Watersheds excluding those in MN, LA, SC and ME (1,844)
HQ>1.5
225
(10%)
423
(18%)
879
(37%)
5
*
17
*
167
(7%)
Table 2-14. Sensitivity analysis results based on constraining analysis to (a) watersheds in
the top 25th percentile with regard to total Hg deposition and (b) exclude watersheds
located in MN, LA, SE or ME) - Results for watersheds with potentially at risk
populations based on combining both Stage 1 and Stage 2 results (Stage 3 of the 3 stage
framework)
ECU risk
threshold
Number and percentage of HUCs meeting risk threshold criteria (2005 scenario)
2005 analysis
90th fish
consumption
95th% fish
consumption
99th% fish
consumption
2016 analysis
90th fish
consumption
V. S. EGV -attributable risk >1.5HQ OR total risk >1.5HQand V. S.
95th% fish
consumption
99th% fish
consumption
EGU contribution of:
ALL WATERSHEDS (2,366)
60
-------
ECU risk
threshold
>5%
> 10%
> 15%
> 20%
Number and percentage of HUCs meeting risk threshold criteria (2005 scenario)
2005 analysis
90th fish
consumption
1271
879
584
391
(54%)
(37%)
(25%)
(17%)
95th% fish
consumption
1321
946
682
550
(56%)
(40%)
(29%)
(23%)
99th% fish
consumption
1477
1221
1109
1073
(62%)
(52%)
(47%)
(45%)
2016 analysis
90th fish
consumption
484
96
49
34
(20%)
(4%)
(2%)
*
95th% fish
consumption
524
121
76
61
(22%)
(5%)
(3%)
(3%)
99th% fish
consumption
672
325
292
286
(28%)
(14%)
(12%)
(12%)
SENSITIVITY ANALYSIS 1 - Watersheds in top 25th % with regard to total Hg deposition (591)
>5%
> 10%
> 15%
> 20%
323
202
152
115
(55%)
(34%)
(26%)
(19%)
334
219
179
153
(56%)
(37%)
(30%)
(26%)
377
306
286
278
(64%)
(52%)
(48%)
(47%)
89
34
24
16
(15%)
(6%)
(4%)
(3%)
99
44
37
30
(17%)
(7%)
(6%)
(5%)
146
97
92
90
(25%)
(16%)
(16%)
(15%)
SENSITIVITY ANALYSIS 2 - Watersheds excluding those in MN, LA, SC and ME (1,844)
>5%
> 10%
> 15%
> 20%
1135
811
529
339
(62%)
(44%)
(29%)
(18%)
1179
860
609
478
(64%)
(47%)
(33%)
(26%)
1275
1035
923
888
(69%)
(56%)
(50%)
(48%)
449
77
35
22
(24%)
(4%)
(2%)
*
476
87
44
30
(26%)
(5%)
(2%)
(2%)
543
208
175
169
(29%)
(11%)
(9%)
(9%)
Table 2-15. Sensitivity analysis results based on estimating risk for watersheds with Hg
fish tissue levels based only on stationary waterbodies (i.e., excluding samples taken
from flowing waterbodies) - Results for both total risk and U.S. EGU-attributable risk
Fisher
consumption
rate percentile
and rate (g/day)
Watershed percentile
50th percentile
total
U.S.
ECU
75th percentile
total
U.S.
ECU
90th percentile
total
U.S.
ECU
95th percentile
total
U.S.
ECU
99th percentile
total
U.S.
ECU
Primary analysis (all watersheds - lakes and rivers - 2005 scenario)
mean (3 9)
90th (123)
95th (173)
99th (373)
3.5
11.1
15.6
33.6
0.3
1.1
1.5
3.3
3.2
10.1
14.2
30.7
0.4
1.3
1.8
3.9
4.2
13.3
18.7
40.3
0.5
1.6
2.3
4.9
5.4
17.1
24.1
51.9
0.7
2.3
3.2
6.9
7.2
22.8
32.0
69.0
1.4
4.3
6.0
13.0
SA (lakes only - 2005 scenario)
mean (3 9)
90th (123)
95th (173)
99th (373)
2.9
9.1
12.8
27.6
0.1
0.3
0.5
1.0
3.5
11.0
15.5
33.5
0.3
0.8
1.1
2.4
3.8
12.1
17.0
36.6
0.4
1.3
1.8
4.0
4.6
14.5
20.4
44.0
0.6
2.0
2.7
5.9
6.5
20.5
28.8
62.1
1.4
4.4
6.2
13.5
Observations regarding the sensitivity analyses completed for this analysis are presented
below.
n-th
Generating risk estimates including only those watersheds falling in the top 25
percentile with regard to total Hg deposition: The percent watersheds with potentially at
risk populations ranged from just slightly higher to as much as 50% higher compared
with the core analysis which did not exclude watersheds based on the magnitude or
61
-------
ranking with regard to total Hg deposition (see Tables 2-12 through 2-14). These results
suggest that focusing on those watersheds with relatively greater total Hg deposition
would result in a slightly larger fraction of "at risk" watersheds.
• Generating risk estimates excluding watersheds located in four states (AL, SC, ME and
MN): This sensitivity analysis resulted in different effects on the Stage 1 and Stage 2 risk
estimates. Estimates of the number of watersheds with at risk populations due to U.S.
EGUs contributing at least 5% (note, other percentage contributions are also presented) to
total risk (Stage 1) demonstrated a mixture of moderate increases and decreases
compared with the core analysis, depending on the "percent contribution" category
considered (see Table 2-12). However, estimates of the percentage of watersheds with at
risk populations when considering U.S. EGU mercury deposition (without other sources
- Stage 2) based on dropping out the four states was 7% as compared with 12% when
considering all states (for the 99th percentile fish consumption rate - see Table 2-13).
While this sensitivity analysis does suggest that excluding these 4 states does reduce the
percent of watersheds with potentially at risk populations included in Stage 2, there is
still a notable fraction (i.e., 7% under the 99th% consumption rate based on an HQ > 1.5
representing a potential health concern - see Table 2-14).
• Focusing only on stationary waterbodies (lakes and ponds) and excluding flowing
waterbodies did result in notably lower U.S. EGU-incremental risk on average for the
waterbodies, however risk estimates for upper end water sheds were not substantially
effected: As can be seen in Table 2-15, running the risk model using only fish tissue
MeHg estimates from stationary waterbodies did result in lower U.S. EGU incremental
risk for the average watershed. However, importantly, high-end estimates - estimates for
the 90th percentile watershed and above - of U.S. EGU incremental risk were not
significantly different with that difference disappearing and actually reversing at the
highest level (i.e., the risk estimates for the 99l percentile watershed actually tended to
be higher for the simulation focusing on stationary waterbodies). Given that the primary
focus of the risk assessment is on high-end percentile risk, the fact that this sensitivity
analysis only showed difference in risk at mean watersheds with this difference
diminishing and actually reversing at the highest watersheds, argues that this source of
uncertainty (i.e., that the MMaps approach is most appropriate for stationary waterbodies)
does not substantially affect the analysis.
2.8. Summary of Key Observations
Key policy-relevant observations drawn from discussions presented in Sections 2.3
through 2.7 (and including Section 1.4 - the discussion of uncertainty and variability) are
presented below:
• Estimates of U.S. EGU mercury emissions suggest that the 2016 Scenario is likely closer
to current (2010) emissions compared with the 2005 Base Case (which has substantially
higher total mercury emissions for this sector). Therefore, risk estimates generated for the
62
-------
2016 Scenario have received greater emphasis and the 2005 Base Case estimates are de-
emphasized (since they are likely substantially higher than current conditions).
U.S. EGUs can contribute up to 11% of total Hg emissions over a subset of watersheds
(for the 2016 Scenario - value cited is for the 99th% watershed). However, in general,
other sources besides U.S. EGUs dominate Hg deposition, with U.S. EGUs contributing
on average (again for the 2016 Base Case), about 2% of total Hg deposition across the
country. U.S. EGU-related Hg deposition is higher in the eastern part of the country with
specific hot spots in a number of areas, including most notably, the Ohio River valley.
U.S. EGU-related Hg deposition estimates show a significant reduction between 2005
and 2016 Base Cases, reflecting mainly implementation of PM controls with the average
U.S. EGU deposition decreasing from -5% of total to -2% for the 2005 and 2016
scenarios, respectively.
U.S. EGUs can contribute up to 18% of MeHg in fish tissue (99th percentile watershed
value for the 2016 Scenario). However, generally, U.S. EGUs contribute a much smaller
fraction averaging 4% for the 2016 Scenario.
Comparing the magnitude of Hg fish tissue levels with total Hg deposition (as
characterized at the watershed-level) suggests that there is not a strong correlation. This is
not surprising given the variety of factors which effect methylation potential; factors
which can demonstrate substantial spatial variation.
Comparing the pattern of U.S. EGU-attributable Hg deposition with watersheds
containing Hg fish tissue data (i.e., the watersheds reflected in the risk assessment) results
in our concluding that, while we have some degree of coverage for high U.S. EGU
impact areas, this coverage is limited. For this reason, we believe that the actual number
of "at risk" watersheds (i.e., watersheds where U.S. EGUs could contribute to a public
health concern) could be substantially larger than estimated.
Estimates of total risk from all sources of mercury using the RfD based metrics do not
show a substantial reduction between the 2005 and 2016 Scenarios, which is expected
given that sources other than U.S. EGUs dominate Hg deposition over the vast majority
of watersheds and emissions for these sources remain largely unchanged in the
simulation. However, U.S. EGU-attributable risk does demonstrate a notable reduction
between the 2005 and 2016 Scenarios, primarily reflecting implementation of PM
precursor emissions controls, as note earlier.
Under the 2016 Base Case, IQ loss when we consider U.S. EGU mercury deposition
without including other sources of mercury is below the 1 to 2 IQ point range for over
99% of the watersheds included in the risk assessment (based on the high-end female
consumer scenario). The 1-2 IQ point loss range was identified by the SAB as
representing a level of clear public health significance, not that IQ loss below that is of no
health significance. It is important to note, that for a substantial fraction of these
watersheds, total IQ loss (reflecting all mercury sources) does reach or exceed this range
of IQ loss. And in those instances, even a relatively small incremental reduction in IQ
loss related to reducing U.S. EGU emissions would be considered beneficial from a
63
-------
public health standpoint. In addition, as noted in Section 1.2, we do not think the IQ
endpoint fully captures the neurodevelopmental risk associated with MeHg exposure and
for this reason, we have focused on MeHg RfD-based risk estimates in considering the
potential public health hazard associated with U.S. EGU-attributable mercury exposure.
Risks estimated for the high-end female consumer population at the national-level are
generally higher than those assessed for a number of the other populations covered (e.g,
Vietnamese, Hispanic, and Great Lakes Tribal fishers) and therefore provide coverage for
those additional fisher populations. However, risk estimates for both white and black
fishers in the southeast were higher than estimates for the high-end female consumers
assessed nationally. It is also important to note the uncertainty and limitations in the data
used for the southeast fisher populations (see Section 2.6.1). Given our desire to provide
broader coverage for the U.S. and concerns over Hg fish tissue levels in South Carolina
(SC) potentially reflecting non-U.S. EGU loading (fish tissue values in SC play an
important role in driving risk for the two southeastern fishing populations), we have
placed greater emphasis in discussing risk estimates on the female high-end consumers
assessed at the national-level.44
• Based on application of the 3-stage risk characterization framework described in Section
1.2, we estimate that from 3 to 28% of the watersheds included in this risk assessment
could be classified as potentially having at risk populations under the 2016 Scenario. This
percent range is based on a 95th to 99th percentile consumption rates for the high-end
female consumer assessed at the national-level and assumes further, that an HQ of > 1.5
(i.e., an exposure just above the RfD - see above) represents a potential public health
concern. These stage-3 results reflect the aggregation of results from Stages 1 and 2 of
the 3-Stage Risk Characterization Framework (i.e., watersheds where the U.S. EGU
increment-based HQ > 1.5 and watersheds where total risk is > 1.5 HQ and U.S. EGUs
make at least at 5% contribution to that risk, respectively).
• If U.S. EGU impacts to watersheds included in the risk assessment were zeroed-out, for a
significant majority of those watersheds, total exposure would still exceed (and in most
cases, significantly exceed) the RfD. Reductions in EGU attributable Hg will reduce the
magnitude of the risk, although substantial total exposure and risk from Hg deposition
will remain.
• Sensitivity analyses conducted primarily to examine uncertainty in applying the MMaps
approach for linking Hg deposition to Hg fish tissue levels, suggest that uncertainty
related to the MMaps approach is unlikely to substantially effect an assessment of
whether Hg emissions from U.S. EGUs constitute a public health concern.
44 We note that the female high-end consumer population benefits from (a) being assessed at the national-level
(which both increases the number of watersheds assessed for risk and more fully reflects spatial patterns of U.S.
EGU-attributable mercury impacts and fish tissue MeHg levels across the country) and (b) has consumption rates
based on a larger sample size compared with the more focused southeastern black and white populations (see
Appendix C, Table C-l). For these reasons, we believe that risk estimates for the female high-end consumer
population have higher overall confidence (in addition, they better cover on the population of concern for the
endpoints modeled - women of childbearing age who consume relatively large amounts of self-caught fish).
64
-------
65
-------
Citations:
Appel, K. W., A. B. Gilliland, et al. (2007). "Evaluation of the Community Multiscale Air
Quality (CMAQ) model version 4.5: Sensitivities impacting model performance Part I -
Ozone." Atmospheric Environment 41(40): 9603-9615.
Appel, K. W., P. V. Bhave, et al. (2008). "Evaluation of the community multiscale air quality
(CMAQ) model version 4.5: Sensitivities impacting model performance; Part II -
particulate matter." Atmospheric Environment 42(24): 6057-6066.
Axelrad, D. A.; Bellinger, D. C.; Ryan, L. M.; Woodruff, T. J. (2007). Dose-response
relationship of prenatal mercury exposure and IQ: an integrative analysis of
epidemiologic data. Environmental Health Perspectives. 2007, 115, 609-615.
Bullock, O. R. and K. A. Brehme (2002). "Atmospheric mercury simulation using the CMAQ
model: formulation description and analysis of wet deposition results." Atmospheric
Environment 36(13): 2135-2146.
Burger, J. (2002). Daily consumption of wild fish and game: Exposures of high end
recreationalists, International Journal of Environmental Health Research, 12:4, p. 343-
354, July, 2010.
Burger, J., Stephens, W. L., Boring, C. S., Kuklinski, M., Gibbons, J. W., Gochfeld M. Factors in
Exposure Assessment: Ethnic and Socioeconomic Differences in Fishing and
Soncumption of Fish Caught along the Savannah River. Risk Analysis, Vol. 19, No. 3,
1999, p. 427.
Byun, D. and K. L. Schere (2005). "Review of the governing equations, computational
algorithms, and other components of the models-3 Community Multiscale Air Quality
(CMAQ) modeling system." Applied Mechanics Reviews 59(1-6): 51-77'.
Harris., R. C., John W. M. Rudd, Marc Amyot, Christopher L. Babiarz, Ken G. Beaty, Paul J.
Blanchfield, R. A. Bodaly, Brian A. Branfireun, Cynthia C. Gilmour, Jennifer A.
Graydon, Andrew Heyes, Holger Hintelmann, James P. Hurley, Carol A. Kelly, David P.
Krabbenhoft, Steve E. Lindberg, Robert P. Mason, Michael J. Paterson, Cheryl L.
Podemski, Art Robinson, Ken A. Sandilands, George R. Southworth, Vincent L. St.
Louis, and Michael T. TateRudd, J. W. M., Amyot M., et al., Whole-Ecosystem study
Shows Rapid Fish-Mercury Response to Changes in Mercury Deposition. Proceedings of
the National Academy of Sciences Early Edition, PNAS 2007 104 (42) pp. 16586-16591;
(published ahead of print September 27, 2007).
Knights, D. C, Sunderland, E. M., Barber, M. C., Johnston J. M., and Ambrose, R. B.,
Application of Ecosystem-Scale Fate and Bioaccumulation Models to Predict Fish
Mercury Response Times to Changes in Atmospheric Deposition. Environmental
Toxicology and Chemistry, Vol 28, No. 4., pp. 881-893.
Morgan, J.N., M.R. Berry, and R.L. Graves. 1997. "Effects of Commonly Used Cooking
Practices on Total Mercury Concentration in Fish and Their Impact on Exposure
66
-------
Assessments." Journal of Exposure Analysis and Environmental Epidemiology 7(1): 119-
133.
Moya. J., Itkiin, C., Selevan, S.G., Rogers, J.W., Clickner, R. P., Estimates offish consumption
rates for consumers of bought and self-caught fish in Connecticut, Florida, Minnesota,
and North Dacota. Science of the Total Environment (2008),
doi:10.1016/j.scitotenv.2008.05.023.
NMFS, 2003a. National Marine Fisheries Service, Fisheries Statistics Division, National Oceanic
and Atmospheric Administration. Statistical Highlights: Fisheries of the United States,
2003,2003.
Oken e., Radesky, J.S., Wright, R. O., Bellinger D. C., Amarasiriwardena C. UJ., Kleinman, K.
P., Hu, H., Gillman, M.W., Maternal Fish Intake during Pregnancy, Blood Mercury
Levels, and Child Cognition at Age 3 Years in a US Cohort. American Journal of
Epmidemiology, Vol. 167, No. 10, p. 1171.
Pirrone N, Cinnirella S, Feng X, Finkelman RB, Friedli HR, Leaner J, et al., 2010. Global
mercury emissions to the atmosphere from anthropogenic and natural sources.
Atmospheric Chemistry and Physics Discussions, 10(2), 4719-4752.
Rice GE, Hammitt JK, Evans JS. (2010). A probabilistic characterization of the health benefits of
reducing methyl mercury intake in the United States. Environmental Science Technology.
2010 Jul l;44(13):5216-24
Shilling, Fraser, Aubrey White, Lucas Lippert, Mark Lubell (2010). Contaminated fish
consumption in California's Central Valley Delta. Environmental Research 110, p. 334-
344.
Swartout, J., and G. Rice. 2000. "Uncertainty Analysis of the Estimated Ingestion Rates Used to
Derive the Methylmercury Reference Dose." Drug and Chemical Toxicology 23(1):293-
306. 11-41
UNEP's Division of Technology, Industry and Economics. 2010. Study on mercury sources and
emissions and analysis of cost and effectiveness of control measures "UNEP Paragraph
29 study".
USEPA, U.S. Environmental Protection Agency. (2000). Guidance for Assessing Chemical
Contaminant Data for Use in Fish Advisories, Volume 3: Overview of Risk Management.
Office of Science and Technology, Office of Water, U.S. Environmental Protection
Agency, Washington, DC. EPA 823-B-00-007.
USEPA, U.S. Environmental Protection Agency, (2001). Mercury Maps - A Quantitative Spatial
Link Between Air Deposition and Fish Tissue Peer Reviewed Final Report. U.S. EPA,
Office of Water, EPA-823-R-01-009, September, 2001.
67
-------
USEPA, U.S. Environmental Protection Agency, (2007). Review of the National Ambient Air
Quality Standards for Lead: Policy Assessment of Scientific and Technical Information
OAQPS Staff Paper. Office of Air Quality Planning and Standards, Research Triangle
Park, NC. EPA-452/R-07-013 November 2007.
USEPA, U.S. Environmental Protection Agency. (2005), Neurobehavioral Assessments
Conducted in the New Zealand, Faroe Islands, and Seychelles Islands Studies of
Methylmercury Neurotoxicity in Children. David C. Bellinger, Professor of Neurology,
Harvard Medical School, March 2005. Report to the U.S. Environmental Protection
Agency EPA-HQ-OAR-2002-0056-6045[ 1 ]
USEPA, U.S. Environmental Protection Agency. (2007a) Review of the National Ambient Air
Quality Standards for Lead: Policy Assessment of Scientific and Technical Information
OAQPS Staff Paper. Office of Air Quality Planning and Standards, Research Triangle
Park, NC. EPA-452/R-07-013 November 2007.
USEPA, U.S. Environmental Protection Agency. (2007b) Lead: Human Exposure and Health
Risk Assessments for Selected Case Studies, Volume I. Human Exposure and Health
Risk Assessments—Full-Scale and Volume II. Appendices. Office of Air Quality
Planning and Standards, Research Triangle Park, NC. EPA-452/R-07-014a and EPA-
452/R-07-014b.
USEPA, U.S. Environmental Protection Agency. (2011) Regulatory Impact Assessment:
Benefits Associated with Reduced Mercury Air Emissions in the U.S. Utility Sector,
DRAFT Report. Office of Air Quality Planning and Standards, Research Triangle Park,
NC.
U.S. Geological Survey and U.S. Department of Agriculture, Natural Resources Conservation
Service, 2009, Federal guidelines, requirements, and procedures for the national
Watershed Boundary Dataset: U.S. Geological Survey Techniques and Methods 11-A3,
55 p.
68
-------
Appendices: Additional Technical Detail on Modeling Elements
(Note, citations for the appendices are provided in the citation list for the main document
see above)
A. Specifying spatial scale of watersheds
As mentioned earlier in Section 1.3, the specification of the spatial scale for watersheds
to be used as the basis for the risk assessment (i.e., HUC12's) was based, in part, on
consideration for the size of watersheds included in two studies examining the relationship
between mercury deposition and changes in mercury concentrations in aquatic media and biota.
The Knights et al., 2009 study considered a number of modeling frameworks for predicting
changes in fish mercury levels, following changes in aerial mercury deposition. The study
included simulation of five different types of waterbodies ranging from a seepage lake (with little
watershed loading) in Florida to a stratified drainage lake in NH. Response times for changes in
mercury fish tissue levels following a 50% reduction in aerial mercury deposition were simulated
for the different watersheds. The simulations showed that all five locations had a two-phase
response in fish tissue mercury concentrations including (a) an initial 1-3 year response linked to
immediate reductions in aerial deposition directly to the waterbody itself and (b) a longer-term
(decades) timeframe for a full system response that would include such factors as changes in
watershed erosion and loading to the waterbody. The study also showed that deeper lakes with
larger watersheds could have longer response times. The article reported that the initial "faster"
response (taking 1-3 years) could account for 40-60% of the total steady state response in some
instances.
The results of this study can be interpreted as suggesting that, if we are considering the
more immediate change in fish tissue mercury levels (occurring within a few years following a
change in mercury loading), we should focus on assessing the fractional change in mercury
deposition to the waterbody itself. However, if we can assume that near steady-state conditions
are met and we are interested in a more complete simulation of changes in fish MeHg levels then
we would want to consider changes in the level of mercury deposition over the entire watershed
(and not just the waterbody). These observations support using the watershed (and not the
individual waterbody) as the basis for linking Hg deposition with fish tissue MeHg levels. The
scale of the five watersheds included in the Knights et al., 2009 study range from 20 by 100 km
(for the coastal plain river location in GA) to 5 by 10km (for the Lake Waccamaw NC site).
Three of the five locations had watersheds in the 10 by 10km range (see Figure 2 in the article).
Given that the majority of locations in the study had smaller watersheds (i.e., in the 10 by 10km
range), we conclude that this would represent a reasonable watershed spatial scale to use in
linking changes in aerial deposition to changes in fish tissue levels (i.e., as the basis for risk
characterization in the analysis).
An article by Harris et al., 2007, which is based on the METALLICUS study (specifically
lake 658 catchment in northwestern Ontario Canada), also examined the temporal profile
associated with changes in media and biota mercury levels following a change in mercury
deposition. In this study, a 3yr loading of labeled mercury to the waterbody and watershed
(separate labeled mercury applied to each location) was followed by measurement of mercury in
69
-------
various media and biota to see how long it took for the loaded mercury to impact different
compartments. This analysis showed that impacts on aquatic biota from watershed mercury
could be very slow (on the order of decades). While mercury deposited directly to the waterbody
could have a more immediate effect on aquatic biota, this effect was buffered by the methylation
of older mercury which resulted in a lag in the impact (after 3-4 years, there had been a 30-40%
increase in biota mercury concentrations, however the actual change in deposition to the lake
over that same time period had been -120%, suggesting that this buffering was occurring). The
study concludes that lakes which receive most of their mercury deposition loading directly from
the atmosphere (e.g., catchment lakes) could see an effect on biota within a decade while lakes
with more complex watershed loading conditions could see a two-phase temporal change with a
longer lag. The single watershed involved in this study is relatively small (only a few km on a
side). Therefore, the spatial scale of the watershed involved in this study also support use of a
more refined spatial scale for watersheds in the risk assessment.
In addition as mentioned earlier in Section 1.3, use of watersheds that are more spatially
refined increases the potential for capturing spatial gradients in the deposition of mercury and
resulting variation in the impact of those loadings on aquatic freshwater biota. Conversely, use of
larger watersheds, while allowing us to model more of the country in the risk assessment, could
result in the dilution of areas of elevated mercury deposition from U.S. EGUs (and therefore, by
association, the dilution of U.S. EGU-attributable mercury exposure and risk).
Consideration for the information presented above, resulted in us identifying HUC12s as
the optimal watershed size for the risk analysis. HUC12's vary based on topography but
generally are from 5-1 Okm on a side.
B. Characterizing measured fish tissue concentrations at the watershed-level
In developing the fish tissue dataset for the risk assessment, we began with the master
dataset that has been developed by our team to support the Regulatory Impact Analysis (RIA).
This master datasets included fish tissue Hg samples from years 1995-2009, with approximately
50,000 unique samples from 4,115 HUC12s across the US, although the samples are more
heavily focused on locations east of the Mississippi. Note, that for the risk assessment, we used a
subset of the data from 2000 and later, with samples distributed across 2,461 HUC12s. The use
offish tissue samples from this later time frame was intended to focus on samples more
representative of current conditions and less likely to reflect Hg deposition levels prior to 2000
when substantial reductions in Hg emissions and hence deposition were taking place. The fish
tissue samples in the master dataset come primarily from three sources:
• National Listing of Fish Advisory (NLFA) database. The NLFA, managed by EPA
(http://water.epa.gov/scitech/swguidance/fishshellfish/fishadvisories/), collects and
compiles fish tissue sample data from all 50 states and from tribes across the United
States. In particular, it contains data for over 43,000 mercury fish tissue samples collected
from 1995 to 2007.
70
-------
• U.S. Geologic Survey (USGS) compilation of mercury datasets. As part of its
Environmental Mercury Mapping and Analysis (EMMA) program, USGS compiled
mercury fish tissue sample data from a wide variety of sources (including the NLFA) and
has posted these data at http://emmma.usgs.gov/datasets.aspx. The compilation includes
(1) state-agency collected and reported data (including Delaware, Iowa, Indiana,
Louisiana, Minnesota, Ohio, South Carolina, Virginia, Wisconsin, and West Virginia)
from over 40,000 fish tissue samples, covering the period 1995 to 2007 and (2) over
10,000 fish tissue samples from several other sources, including the National Fish Tissue
Survey, the National Pesticide Monitoring Program (NPMP), the National Contaminant
Biomonitoring Program (NCBP), the Biomonitoring of Environmental Status and Trends
(BEST) datasets of the USFWS and USGS (http://www.cerc.cr.usgs.gov/data/data.htm),
and the Environmental Monitoring and Analysis Program (EMAP)
(http://www.epa.gov/emap/).
• EPA's National River and Stream Assessment (NRSA) study data. These data include
nearly 600 fish tissue mercury samples collected at randomly selected freshwater sites
across the United States during the period 2008 to 2009.
Data from these three datasets were combined into a single master fish tissue dataset
covering the period 1995 to 2009. One problem encountered in combining these datasets is the
potential duplication of samples in the NLFA and USGS state-collected data. Unfortunately,
these two datasets do not contain directly comparable and unique identifiers that allow duplicate
samples to be easily identified and removed. Therefore, as an alternative, the samples from these
two datasets were subdivided into data groups according to the year and state in which they were
collected. If both datasets contained a data group for the same year and the same state, then the
data group with the fewer number of observations was excluded from the master data.
In finalizing the master datasets a number of criteria were used to screen the fish tissue
samples (e.g., include only freshwater fish species, exclude estuarine locations, exclude fish less
than 7 inches in length). In addition, we assigned "river" and "lake" identifiers to each fish tissue
sample (additional detail on the process used to develop the master fish tissue dataset can be
found in USEPA, 2011, Section 5.2.2).
As note earlier, for the risk assessment we used a subset of the Master dataset including
fish tissue samples collected 2000 and later. This risk assessment dataset comprised a total of
23,878 samples (12,500 from lakes and 11,478 from rivers) covering a total of 2,461 HUC12s.
Subsets of this dataset were used in modeling risk for each of the fisher populations included in
the analysis (i.e., those HUCs intersecting US Census tracts containing "source populations"
associated with a particular fisher scenario - see discussion in Section 1.3).
Each watershed with measured Hg fish tissue data, tended to have multiple values (the
average number offish tissue measurements for the period 2000 to 2009 for the 2,461
watersheds is 10, although some watersheds had up to 270 measurements). Therefore, we also
needed to identify a summary statistic to use for each watershed to represent fish tissue levels in
estimating exposure and risk. As noted earlier in Section 1.3, we selected the 75th percentile fish
tissue value at each watershed as the basis for exposure and risk characterization. Selection of
the 75th percentile value was based on the assumption that subsistence fishers would favor larger
71
-------
fish which have the potential for higher bioaccumulation (i.e., use of a median or mean value
could low-bias likely catch-related mercury levels). There is uncertainty associated with this
assumption and should fishers at a particular watershed favor fish that are either larger or smaller
than the type offish reflected in the 75th% sample, then risk estimates could be biased
accordingly. In deriving the 75th percentile value for a given watershed, we first generated a set
of "river" and "lake" fish tissue percentiles (e.g., a "lake" 75th percentile representing the 75th
percentile fish tissue Hg level among lake samples located within that HUC12). We then
identified the higher of the lake-75th percentile and river-75th percentile, if both existed, and
used that higher value in modeling risk.
C. Defining subsistence fisher scenarios to model
A number of criteria had to be met for a study to be used in providing explicit
consumption rates for the high-end fisher populations of interest in this analysis. For example,
studies had to provide estimates of self-caught fish consumption and not conflate these estimates
with consumption of commercially purchased fish. Furthermore, these studies had to focus on
freshwater fishing activity, or at least have the potential to reflect significant contributions from
that category, such that fish consumption rates provided in a study could be reasonably applied in
assessing freshwater fishing activity. As noted earlier, given our interest in higher-end
consumption rates, the studies also had to either provide upper percentile estimates, or support
the derivation of those estimates (e.g., provide medians and a standard deviations). Studies of
activity at specific waterbodies (e.g., creel surveys), while informative in supporting the presence
of higher-end consumption rates, could not be used as the basis for defining our high-end
consumption rates since there would be uncertainty in extrapolating activity at a specific river or
lake more broadly to fishing populations in a region. Therefore, we focused on studies
characterizing fishing activity more broadly for specific regions or states. Application of these
criteria resulted in the selection of three studies as the basis for characterizing high-end fish
consumption rates for the fisher populations included in the analysis. These studies, together
with the fisher populations characterized and a description of the regional coverage assumed for
each fisher population (see below) are presented in Table C-l.
As noted earlier in Section 1.3, with the exception of the Tribal fisher population assessed
for the Great Lakes (which was restricted to activity on lands ceded to the Ojibwa or Chippewa),
we did extend coverage for the other fisher populations beyond the specific areas covered in the
surveys.45 For example, while the Vietnamese and Laotian survey data were collected in
California, given the ethnic/cultural nature of these high fish consumption rates, we assumed that
this type of high-end fish consumption behavior could be associated with members of these
ethnic groups living elsewhere in the U.S. Therefore, the high-end consumption rates referenced
in the California study for these ethnic groups were used to model risk at watersheds elsewhere
in the U.S. In deciding which of the 2,461 watersheds included in the risk assessment might be
subject to fishing activity by a given fisher population, we used U.S. Census data to determine if
45 The decision to restrict activity for the Tribal fishers to the ceded territories reflects the fact that fish consumption
rates are particular to Tribal practices and can vary considerably across Tribes, arguing against extrapolation offish
consumption rates across Tribes.
72
-------
a "source population" for that fisher population was located in the tract(s) intersecting each
watershed. For example we modeled the Vietnamese high-end fisher population only at those
watersheds associated with U.S. Census tracts containing at least 25 Vietnamese (i.e., a "source
population" for that fisher population). This approach was similarly done for each of the other
fisher populations.46
Looking beyond these specific ethnic groups, we also need to establish a more
generalized high-end (subsistence) scenario that could be assessed broadly across the 2,461
watersheds. Generally all of the studies identified high-end percentile consumption rates (90th to
99th percentiles for the populations surveyed) ranging from approximately one fish meal every
few days to a fish meal a day (i.e., 120 g/day to greater than 500 g/day fish consumption). We
used this general trend across the studies to support application of a generalized high-end female
consumer scenario across most of the 2,461 watersheds.47
While we believe that the approach of extending coverage for these fisher populations
beyond the regions reflected in the underlying surveys is reasonable, we do acknowledge
uncertainty associated with this extrapolation (see Section 1.4 and Appendix F for a discussion
of key uncertainties in the analysis).
In addition to the studies cited in Table C-l used to define the high-end fish consuming
populations modeled in the risk assessment, we also reviewed a large number of additional
studies characterizing higher-level self-caught fish consumption in the U.S. While these studies
had limitations that prevented their use as the basis for defining high-end fisher scenarios to
include in the analysis, they do generally support the levels of self-caught fish consumption
modeled in the analysis. Several of these "supporting" studies are briefly described below:
• A study by Burger et al., 1999, examining recreational and subsistence fishing activity
along the Savannah river in Georgia to determine the role played by socio-economic
status (SES) factors (including race, education and income) and determining levels of
self-caught fish consumption in this study area. The study suggested that all three factors
are associated with levels of fishing activity. Specifically, in the case of race, the study
showed that Blacks tend to have much higher rates offish consumption than whites.
However for both groups, the study did suggest that upper end percentile consumption
rates could be high and certainly approach subsistence levels. For example, a -200 g/day
fish consumption rate represented the 98th percentile for Whites, but only the 92nd
percentile for Blacks. This study does support the presence of high-end consumption
rates for both Blacks and Whites that approach or meet subsistence levels in this area of
the country.
46 In the case of black and white fisher populations in the southeast, we further assumed that "source populations"
for each of these fisher groups would comprise at least 25 members (of that race) below the poverty line at the tract
level. This reflected the potential for greater subsistence fishing activity among economically disadvantaged
individuals.
47 Similar to the white and black fisher populations in the southeast, we only applied this fisher populations at
watersheds located in tracts with at least 25 individuals below the poverty line, which meant that 2,366 of the 2,461
watersheds with Hg fish tissue data were assessed for risk for this more generalized female high-consumer
population.
73
-------
• A study by Moya et al., 2008 examined factors associated with regional differences in
patterns offish consumption, including age, ethnicity (including Tribal affiliation),
socioeconomic status (e.g., income, education), and type/source offish consumed
(freshwater, marine, and estuarine obtained from commercial sources versus self-caught).
The study examined fishing activity in four states (CT, FL, MN, and ND). The study did
provide estimates of high-end self-caught fish consumption for populations in the four
states. Higher, subsistence-level consumption rates were identified for fishing
populations in FL, MN (specifically for Tribes) and CT (for Asian populations, although
it is not clear whether the rates for Asians hold for self-caught fish consumption in
particular). Higher-end rates reported for ND and for general fishers in CT and MN did
not approach the range of subsistence levels of consumption. However, we would point
out that the study designs used in these surveys may not effectively capture the relatively
small fraction of the overall population likely engaging in high-end subsistence levels of
self-caught fishing behavior. This study does provide added support for the existence of
subsistence fishing populations, at least within FL and for Tribes within MN. However,
failure of the surveys reviewed in the study to capture similar behavior in ND and CT
does not necessarily suggest that this type of behavior is non-existent, although it may
suggest that it is less prevalent than in FL.
74
-------
Table C-l. Fisher Populations Included in the Analysis for Hg Exposure and Risk
Fish consuming populations
covered by study (and
reference information)
Overview of study
Self-caught fish consumption rates
(mean, 90*", 95*", 99a) g/day
Observations from the study relevant to the
risk assessment
Extrapolation of study populations in
the risk assessment
Higher self-caught fish
consuming populations
(white, black and female)
surveyed in South Carolina
Citation: Daily consumption
of wild fish and game:
Exposures of high end
recreationalists, Burger et al.,
International Journal of
Environmental Health
Research, 12:4, p. 343-354,
July, 2002
Random survey of
participants in the
Palmetto Sportsmen's
Classic in Columbia SC
(1998). Population
interested in
fishing/hunting (not
general population -
represents outdoor
enthusiasts in SC)
- black: 171, 446, 557, NC *
- white: 38.8, 93, 129, 286
-female: 39.1, 123, 173,373
* n for this population is only 39,
reducing overall confidence in a 99th
consumption rate (therefore, this high-
end percentile was not included in the
risk assessment)
Sample size is variable - out of 458 respondents,
39 are blacks, 149 are female and 98 are poor -
black n is relatively smaller than the other
groups, which increases uncertainty in higher
percentile values provided for this group.
The authors point out that these results highlight
the considerable spread between high-end
consumers and more typical behavior (95th% is
more than 10X greater than the mean or median
intake rate for wild-caught fish).
Results are also provided for poor (0-20KS
annual income). These consumption rates are
relatively high particularly for the higher
percentiles (90th, 95th and 99th rates are: 285, 429
and 590 g/day). This observation forms the basis
for our decision to assess a number of the
subsistence populations only for watersheds
located in US Census tracts containing members
of source populations below the poverty line for
the white and black populations.
- the black and white fisher populations
were extrapolated to cover all watersheds
modeled for risk in the Southeastern
states.8 The rationale for this was that
fishing activity by these two groups
could be generalized in this region of the
country. Note, however that these
scenarios were only assessed for
watersheds in the Southeast located
within US Census tracts with at least 25
individuals from that ethnic group below
the poverty line.
- given the focus of the risk assessment
on consumption by women (in
considering risk to pregnant women in
particular), we extrapolated the female
consumption rates to all watersheds in
the continental US with at least 25
individuals below the poverty line (this is
the high-end female consumer
population referenced in the risk
assessment).
Higher self-caught fish
consuming ethnic populations
including Hispanics, Laotians
and Vietnamese surveyed in
California
Citation: Contaminated fish
consumption in California's
Central Valley Delta (Shilling
et al., Environmental
Research 110, p. 334-344
(2010)
Study looks at subsistence
fishing activity among
ethnic groups associated
with more urbanized areas
near the Sacramento and
San Joaquin rivers in the
Central Valley in CA.
- Hispanic: 25.8, 98*, 155.9, NC**
- Lao: 47.2, 144.8*, 265.8, NA*
- Vietnamese: 27.1, 99.1*, 152.4, NA*
* 95th percentile values were provided
in the study, however 90th percentile
values were not provided and were
calculated using Crystal Ball (based on
the median and standard deviations
provided) assuming a log-normality of
the consumption rate distributions.
** 99th percentile consumption rates
were not provided (or derived) for any
of these populations due to small
sample sizes of the study populations.
The authors note that many of these ethnic
groups relied on fishing in origin countries and
bring that practice here (e.g., Cambodian,
Vietnamese and Mexican). The authors also
note that fish consumption rates reported here
for specific ethnic groups (specifically Southeast
Asian) are generally in-line with rates seen in
WA and OR studies.
- the Hispanic fishing scenario was
extrapolated to cover watersheds located
in US Census tracts with at least 25 poor
members of the ethnic populations (e.g.,
the Hispanic consumption rates would be
applied to the subset of the 2,461
watersheds located in US Census tracts
with at least 25 poor Hispanic
individuals).
- the Laotian and Vietnamese fishing
scenarios were extrapolated to cover
watersheds located in US Census tracts
with at least 25 members of the
underlying ethnic group.
75
-------
Fish consuming populations
covered by study (and
reference information)
Overview of study
Self-caught fish consumption rates
(mean, 90*", 95*", 99a) g/day
Observations from the study relevant to the
risk assessment
Extrapolation of study populations in
the risk assessment
High-end self-caught fish
consuming Chipewa and
Ojibwa Tribal populations
active in the vicinity of the
Great Lakes.
Citation: Exposure
assessment and initial
intervention regarding fish
consumption of tribal
members in the Upper Great
Lakes Region in the United
States. Bellinger,
Environmental Research 95
(2004) p. 325-340
This study contrasted self-
reported fish consumption
rates by Tribes in the
Great Lakes area with
"actual" fish consumption
rates collected for a subset
of the original study
population (147 of 822
from 4 Tribal
population/location
combinations). The study
found that actual fish
consumption rates were
lower than reported
values.
- reported value for all Tribal areas (in
the study) combined: 62, 136.2,213.1,
492.8
Note, that all higher percentiles (90th -
99th) were derived using Crytal Ball
(based on median and standard
deviations and an assumption of log-
normally distributed variability in
consumption rates)
While the "actual" consumption rates collected
for a subset of the families were far lower than
the reported values (often an order of magnitude
smaller), a number of factors resulted in a
decision to use the reported values rather than
the actual values in the risk assessment. First,
and most importantly, the sample size is very
small for the "actual" analysis with n's ranging
from 12 to 54 individuals (representing a smaller
number of associated families) for the different
survey groups. These small sampling rates
reduce the probability of capturing individuals
with higher consumption rates in the broader
population. It also appears that the actual values
may cover Walleye specifically and not include
all fish., which could bias these values
downward. There is concern that, even if
consumption rates have decreased, actual
heritage cultural practices could still exist (or
there could be a desire to return to those rates),
in which case, risks levels associated with those
higher historical consumption rates could be
important to assess. And finally, the high-end
percentile consumption rates derived based on
reported mean consumption rates (and standard
deviations) are in-line with subsistence
consumption rates seen for other populations in
the U.S. Therefore, these Tribal high-end fish
consumption rates would general comport with
subsistence fish consumption activity and
therefore are considered reasonable to include in
the risk assessment.
Activity only assumed to occur in areas
ceded to the Tribes covered in the study
(regions in the vicinity of the Great
Lakes). Because fishing activity is
highly variable across Tribes (and
closely associated with heritage cultural
practices) we have not extrapolated
fishing behavior for these Tribes outside
of the specific populations and regions
covered.
southeast for purposes of this analysis comprises: Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Virginia and West Virginia
76
-------
D. Estimating total mercury-related exposure at the watershed-level
The following equation is used to estimate total mercury exposure at a particular
watershed as an annual average of the daily methylmecury intake per kg body weight:
IR = FTC * FIR * CAF
BW
IR: daily MeHg intake rate (ug/kg-day). This ingestion dose estimate can be directly
compared with the merthylmercury RfD to generate a total HQ.
FTC: mercury fish tissue concentration (ug/g or ppm): The 75*% value provided for each
of the 2,461 watersheds (see Section 1.3).
FIR: fish ingestion rate (g/day). These values are specific to a given population (see
Table 1-1 in Section 1.3)
CAF: cooking adjustment factor (unitless): Because MeHg is not volatile and is contained
primarily in the muscle, this translates into a factor increase of 1.5 for
concentration of mercury in the cooked fish (Morgan, Berry, and Graves, 1997).
The IR value described above can be directly compared with the MeHg RfD to generate
HQ estimates. However, in order to estimates IQ loss, we need to convert this dose estimate into
an equivalent maternal hair concentration since the IQ loss function uses hair mercury as the
dose measure. To do that, we use a dose-to-hair conversion factor (DHCV) of 12.5 (units ppm
per unit ug/kg-day) that converts ingested dose (IR) to hair mercury concentration in ppm. As
noted earlier in 1.3, the DHCV factor is based on a one compartment toxicokinetic model used
for deriving the MeHg RfD by Swartout and Rice (2000).
E. Establishing the U.S. EGU-attributable fraction of total exposure
As noted earlier in 1.3, establishing the increment of total exposure estimated for a given
fishing population active at a given watershed that is attributable to U.S. EGUs requires to key
elements: (a) application of the MMaps assumption linking source contribution to Hg deposition
over that watershed to source apportionment offish tissue Hg levels and consequently exposure
estimates and (b) use of CMAQ modeling results characterizing the fraction of mercury
deposition over a given watershed that is attributable to U.S. EGUs (as contrasted with the
fraction attributable to all other US and foreign-sourced anthropogenic and natural mercury
emissions). In this section, we provide additional detail on each of these elements.
Mercury maps assumption
To analyze the relationship between Hg deposition and MeHg concentrations in fish
across the 2,461 watersheds included in the risk assessment, as discussed in Section 1.3, we
applied the EPA's Office of Water's Mercury Maps approach. MMaps implements a simplified
form of the IEM-2M model applied in EPA's Mercury Study Report to Congress (USEPA,
1997). By simplifying the assumptions inherent in the freshwater ecosystem models that were
described in the Report to Congress, the MMaps model showed that these models converge at a
77
-------
steady-state solution for MeHg concentrations in fish that are proportional to changes in Hg
inputs from atmospheric deposition (e.g., over the long term fish concentrations are expected to
decline proportionally to declines in atmospheric loading to a waterbody). This solution only
applies to situations where air deposition is the only significant source of Hg to a water body,
and the physical, chemical, and biological characteristics of the ecosystem remain constant over
time. EPA recognizes that concentrations of MeHg in fish across all ecosystems may not reach
steady state and that ecosystem conditions affecting mercury dynamics are unlikely to remain
constant over time. EPA further recognizes that many waterbodies, particularly in areas of
historic gold and Hg mining in western states, contain significant non-air sources of Hg (note,
however, that as described below, we have excluded those watersheds containing gold mines or
with other non-EGU related anthropogenic mercury releases exceeding specified thresholds).
Finally, EPA recognizes that MMaps does not provide for a calculation of the time lag between a
reduction in Hg deposition and a reduction in the MeHg concentrations in fish. Despite these
limitations, EPA is unaware of any other tool for performing a national-scale assessment of the
change in fish MeHg concentrations resulting from reductions in atmospheric deposition of Hg.
Given that the MMaps approach only applies in those situations where aerial deposition
is the dominant source of mercury loading to a watershed, in identifying the 2,461 watersheds to
include in the risk assessment, we excluded those watersheds that either contained active gold
mines or had other substantial non-U.S. EGU anthropogenic releases of mercury. Identification
of watersheds with gold mines was based on a 2005 USGS data set characterizing mineral and
metal operations in the United States. The data represent commodities monitored by the National
Minerals Information Center of the USGS, and the operations included are those considered
active in 2003 (online link: . The identification of watersheds
with substantial non-EGU anthropogenic emissions was based on a TRI-net query for 2008 or
non-EGU mercury sources with total annual on-site Hg emissions (all media) of 39.7 pounds or
more. This threshold value corresponds to the 25th percentile annual US-EGU mercury emission
value as characterized in the 2005 NATA. The EPA team considered the 25th percentile US-EGU
emission level to be a reasonable screen for additional substantial non-U.S. EGU releases to a
given watershed.
There are a number of limitations and uncertainties associated with the application of the
MMaps approach in the context of this risk assessment. Several of these limitations are briefly
discussed here, but a more complete discussion is presented in Section 5.3.2 or the RIA TSD
supporting this regulatory review (USEP, 2011). The application of MMaps in apportioning fish
tissue mercury levels and consequently exposure and risk between U.S. EGUs and all other
sources of mercury at the watershed-level, assumes that the relationship between fish tissue
levels and mercury deposition has remained fairly consistent such that near steady-state
conditions have been reached. However, in reality, patterns of mercury deposition for the period
during which the fish tissue samples were collected (2000 to 2009) have not remained constant.
In addition, those fish tissue concentrations may actually reflect patterns of mercury deposition
from earlier time periods (e.g., the 1990s) when mercury emissions from US sources were
experiencing substantial decreases. In addition other factors that can impact rates of mercury
methylation (e.g., sulfur deposition to waterbodies, pH and eutrification for those watebodies)
have also likely not remained constant over the past 1-2 decades for most watersheds. The fact
that many of these factors related to methylation in fish have not remained constant does
introduce uncertainty into the application of the MMaps approach. However, we believe that the
78
-------
MMaps approach for apportioning fish tissue mercury levels is still appropriate to use,
particularly if we are not attempting to characterize the temporal profile for apportionment and
instead, can assume that sufficient time has passed for near steady state conditions to be
reached.48 Uncertainty related to the application of the MMaps model in the context of this risk
assessment is further discussed in Appendix F.
CMAQ modeling
The Community Multi-scale Air Quality (CMAQ) model v4.7.1 (www.cmaq-model.org) is a
state of the science three-dimensional Eularian "one-atmosphere" photochemical transport model
used to estimate air quality (Byun and Schere 2005; Appel, Gilliland et al. 2007; Appel, Bhave et
al. 2008). CMAQ simulates the formation and fate of photochemical oxidants, ozone, primary
and secondary PM concentrations, and air toxics over regional and urban spatial scales for given
input sets of meteorological conditions and emissions. Mercury oxidation pathways are
represented for both the gas and aqueous phases in addition to aqueous phase reduction reactions
(Bullock and Brehme 2002). The emissions data used in the base year and future reference and
future emissions adjustment case are based on the 2005 v4.1 platform. Emissions are processed
to photochemical model inputs with the SMOKE emissions modeling system (Houyoux et al.,
2000). The 2016 reference case is intended to represent the emissions associated with growth and
controls in that year projected from the 2005 simulation year. Other North American emissions
of criteria and toxic pollutants (including mercury) are based on a 2005 Canadian inventory and
1999 Mexican inventory. Global emissions of criteria and toxic pollutants (including mercury)
are included in the modeling system through boundary condition inflow. The lateral boundary
and initial species concentrations are provided by a three-dimensional global atmospheric
chemistry model, the GEOS-CHEM model (standard version 7-04-11).
48 We recognize that these predictions of U.S. EGU apportionment offish tissue Hg levels may not be realized for
several years or even decades, and, for that reason, the current risk from Hg from EGUs may be considerably higher
because EGU mercury emissions were substantially higher prior to 2005. Furthermore, we recognize that these
MMaps-based apportionments offish tissue Hg levels assume that conditions (in terms of patterns of mercury
deposition and factors related to methylation such as sulfate deposition) also hold relatively constant for years to
decades such that near steady state conditions in the fish tissue mercury concentrations are realized.
79
-------
F. Variability and Uncertainty
Table F-l. Key sources of variability associated with the analysis and degree to which they are reflected in the design of the
analysis
Source of variability
Description
Degree to which source is reflected in design of risk analysis
Variation in the
patterns of total and
U.S. EGU-attributable
mercury deposition
across the U.S.
Patterns of annual deposition of mercury
including total (all source) and estimates of
the U.S. ECU fraction (by watershed)
displays considerable spatial variability
across the U.S. based on the results of 12 km
grid cell CMAQ modeling (see Section 2.3)
By extrapolating CMAQ grid cell results at the more spatially refined HUC12
watershed level, we retain the greater degree of spatial resolution in characterizing
mercury deposition obtained through the use of the 12k CMAQ grid cell simulations.
Variation in the
patterns of mercury fish
tissue levels across the
U.S.
Mercury fish tissue measurements can
display considerable spatial variability
across watersheds.
Variable response of
mercury fish tissue
levels to changes in
patterns of mercury
deposition (MMaps
approach)
We have fish tissue measurements for roughly 4% of the watersheds in the U.S. (i.e.,
2,461 watersheds with measured values out of 88,000 based on data from 2000-
2009). While our measured fish tissue levels do generally provide some degree of
coverage for areas with elevated mercury deposition (in terms of both total and U.S.
EGU-attributable), this coverage is limited and there are a large number of
watersheds with high total and U.S. EGU-attributable mercury deposition for which
we do not have fish tissue measurements (see Section 2.5).
The impact of changes in mercury deposition
on mercury concentrations in fish within a
given watershed can vary greatly depending
on a number of factors (e.g., role of
watershed in loading to waterbody,
methylation potential of the waterbody, rates
of sulfate deposition, nature of aquatic biotic
foodweb including mix of upper-level
trophic fish etc). Not only do these factors
contribute to variation in the degree to which
fish tissue mercury levels will respond to
changes in mercury deposition, they also
affect the temporal profile of that response.
Variation in the methylation potential across waterbodies is minimized as a factor in
our analysis to a certain extent given our use of the MMaps approach combined with
measured mercury fish tissue levels. Specifically, variation in methylation potential
should be implicitly reflected in the measured fish tissue levels used in the analysis
(i.e., if watershed "a" has a much greater methylation potential than watershed "b",
then measured fish tissue levels for "a" should be higher reflecting that difference in
methylation potential). In other words, we are applying the MMaps proportionality
assumption to measured fish tissue levels that should reflect underlying differences in
methylation potential across waterbodies. Furthermore, because we are not predicting
temporal trends in fish tissue levels and instead consider a future point in time (once
near-steady state conditions are reached), variation in the temporal profile of changes
in fish tissue levels related to differences in methylation potential of different
watersheds is also minimized as a factor in the analysis.
Variation in the types
of subsistence fisher
populations active in
different regions
Studies reviewed in developing the approach
for this analysis suggests that there can be
considerable variation in the nature of high-
end self-caught fisher populations across
regions of the country. This variation reflects
ethnic and cultural practices and can also be
Surveys of near-subsistence and subsistence fishing populations allow us to clearly
define this type of activity for specific areas covered by those surveys (e.g., Hispanic,
Vietnamese and Laotian fishing activity in specific regions of California, high-end
fishing activity by blacks and whites in South Carolina and Tribal activity near the
Great Lakes). However, available studies on this type of high-end fishing activity at
inland freshwater bodies do not provide comprehensive coverage for all regions in
80
-------
Source of variability
Description
Degree to which source is reflected in design of risk analysis
driven be socio-economic status (SES)
related factors (e.g., poverty, degrees of
income). In addition, we would expect that
access to freshwater fishing locations would
also play a role.
the U.S. Therefore, we did extrapolate coverage from areas reflected in these studies
to other portions of the U.S. There is uncertainty associated with this process. As
discussed in Section 1.3 a lack of comprehensive survey information on near-
subsistence and subsistence fishing activity for all regions in the U.S. is a limitation
of the analysis.
For a given fisher
population, variability
in fishing activity (e.g.,
species harvested,
fishing activity focused
at one or more
waterbodies)
High-end fisher populations could display
considerable variability both in terms of the
degree to which they frequent specific
waterbodies or watersheds and the degree to
which they target specific types offish (or at
least sizes offish). Both of these factors can
impact estimates of exposure. If a fisher
population distributes their activity across a
range of waterbodies and harvests a variety
of fish species (and sizes) than the
distribution of exposure and risk across that
population will be smaller compared with a
population that focuses activity at individual
waterbodies and tends to focus on larger fish
and/or higher trophic level fish (which will
tend to have higher mercury concentrations,
other factors equal).
We do not have comprehensive information characterizing the nature of high-end
fisher behavior in terms of the factors listed here (i.e., degree to which fishing
activity targets individual waterbodies and fish species). Given that this analysis was
aimed at assessing whether a public health hazard exists due to U.S. EGU mercury
emissions we did consider the situation in which high-end fishers would engage in
fishing activity that could result in somewhat higher exposure and risk, other factors
equal. Specifically, we modeled exposure and risk assuming that (a) fishing activity
is focused within a given watershed (i.e., at waterbodies within that watershed) and
(b) that the fishers modeled would tend to favor somewhat larger fish given that they
are engaged in subsistence activity and therefore supplementing their diet with fish
(recall that we used the 75th percentile fish tissue value in exposure and risk
simulations - see Section 1.3). Note, that if a portion of a fisher population actually
distributes their activity between watersheds and/or consumes a mixture of fish
species and sizes (reflecting a fish tissue level closer to the median or mean for a
watershed), then risks would be lower than those estimated here.
Table F-2. Key sources of uncertainty associated with the analysis, the nature of their potential impact on risk estimates, and
degree to which they are characterized
Source of
uncertainty
Description
Nature of potential impact on the exposure and
risk estimates
Degree to which the potential impact of the
source of uncertainty is characterized as
part of the analysis
Factors relating to
the estimation of
mercury
deposition over
watersheds using
the CMAQ model
(e.g, estimating
Emissions are quantified from
anthropogenic and natural
sources, but re-emission of
historical emissions (pre-2005)
are not well characterized by
the modeling system.
Generally all of the sources of uncertainty reflect the
fact that mercury deposition estimated over specific
watersheds may be over or under-estimated.
The one source of uncertainty for which we can
differentiate the potential direction of impact (i.e.,
bias) is mercury wet deposition where we believe that
The analysis did not include any specific
quantitative analyses aimed at characterizing
uncertainty associated with these three sources
of uncertainty, with the exception of qualitative
consideration for the potential seasonally-
differentiated bias in wet deposition discussed
here.
81
-------
Source of
uncertainty
Description
Nature of potential impact on the exposure and
risk estimates
Degree to which the potential impact of the
source of uncertainty is characterized as
part of the analysis
mercury
emissions from
U.S. EGUs,
chemistry
associated with
mercury fate and
transport,
prediction of wet
and dry
deposition, and
global inflow of
mercury into the
U.S.)
The complete set of mercury
oxidation and reduction
reactions has not been
identified by the scientific
community.
Uncertainty in a wide variety
of model inputs (e.g.,
emissions, meteorology, global
inflow to the modeling
domain, and chemistry)
impacts estimates of mercury
wet and dry deposition.
There is considerable
uncertainty in the global
emissions inventory for
mercury and given the long
residence time of elemental
mercury it is possible that
inflow into the modeling
domain may reflect
deficiencies in the global
emissions inventory.
estimates may be slightly low in the fall and slightly
high in the winter. This assessment of potential bias
is based on comparisons of weekly estimates
(generated by CMAQ) against measurements of wet
mercury deposition collected as part of the Mercury
Deposition Network, which operates under the
National Atmospheric Deposition Program
(http ://nadp. sws.uiuc.edu/MDN/).
Characterizing
subsistence
fishing activity
within areas of
high U.S. ECU
mercury
deposition
There is uncertainty associated
with predicting high-end
fishing activity at specific
watershed and at watersheds
located within specific regions.
Furthermore, there is
uncertainty associated with
characterizing the nature of
that fishing activity (including
the frequency of activity at
different waterbodies, types of
fish targeted).
If subsistence fishing activity is assumed at a given
watershed or at group of watersheds within a region
and in reality, factors preclude that type of fishing
activity (e.g., lack of ready access, poor fishing stock
etc), then those risk estimates are not representative
and actual risk (assessed over the set of watersheds
modeled) would be reduced, since these point
estimates would be removed. Similarly, if high-end
fishing activity by a given fisher or fishing family
tends to be distributed across watersheds and not
focused at a single watershed, as assumed here, then
the upper-end risks estimated across the watersheds
will be reduced, since watersheds with the highest
We did not explore these sources of
uncertainty. Given that the focus of this
analysis is on assessing risk for populations
likely to experience the highest reasonable U.S.
EGU-attributable risk, we concluded that
because it is reasonable to assume that some
fraction of high-end fishing populations could
focus their activity on a single watershed (and
could favor larger fish), we would model this
behavior in our analysis. If these assumptions
are relaxed, then risk will be reduced
(specifically for the watersheds with the highest
U.S. EGU-attributable risk). However, we did
82
-------
Source of
uncertainty
Description
Nature of potential impact on the exposure and
risk estimates
Degree to which the potential impact of the
source of uncertainty is characterized as
part of the analysis
U.S. EGU-attributable risks will be combined with
lesser-impacted watersheds in generating risk
estimates (and not assessed independently for risk
assuming activity focused specifically on them).
Similarly, for the types of fish taken, if high-end
fishers do not favor larger fish, then risk could be
overstated by our use of the 75th percentile fish.
not explicitly model these alternative
behavioral profiles, because of our goal of
capturing reasonable estimates of high-end risk.
Application of
MMaps
assumption in
generating
estimates of the
U.S. EGU-
attributable
fraction of risk
Uncertainty associated with
Applying the MMaps
assumption in the context of
this analysis results from a
number of factors: (a) fish
tissue measurement data used
in the analysis may still reflect
earlier historical patterns of
mercury deposition (from the
1990's), (b) watersheds display
substantially different
methylation potentials,
resulting in differences in the
impact of a unit change in
mercury deposition on fish
tissue levels, (c) factors related
to methylation in watersheds
(e.g., sulfate deposition, pH,
eutrafication) have not
remained constant over time,
resulting in variation in the
methylation potential of
watersheds over time, (d)
despite efforts to exclude
watersheds with substantial
non-air sources of mercury
loading, some watersheds with
substantial non-air impacts
may have been retained in the
analysis, and (e) the potential
If air deposition patterns from the 1990s are reflected
in some of the mercury fish tissue measurements we
are using the implications can vary depending on the
nature of that difference. If fish tissue levels for a
watershed still reflect higher 1990 deposition values,
then we may overstate U.S. EGU-attributable risk,
since in reality we would expect the underlying fish
tissue levels to decrease as the impact of those earlier
higher deposition values dissipates. Conversely, if
total deposition remains the same, but only the source
distribution has changed since the 1990's, then the
effect on our risk assessment may not be that
significant, since we are making projections based on
the current source-mix (or future source-mix)
assuming near steady-state assumptions are reached
given those specific source mixes.
Differences in methylation potential across
watersheds (as discussed in table C-l), are likely to be
reflected in the underlying fish tissue levels
themselves. Therefore, these methylation differences
are likely not to have a substantial impact on our
analysis (note, that the time sequence of changes in
fish tissue levels will depend on differences in
methylation potentials, but we are not attempting to
predict these temporal profiles).
The potential that we may have failed to exclude
watersheds with significant non-air mercury loadings
could introduce high-bias into our estimates of U.S.
As part of the sensitivity analyses completed
for the analysis, we did consider the issue of
potentially having included watersheds with
substantial non-air impacts. Although as
described in Appendix E, we did exclude
watersheds with active gold mines and/or non-
EGU anthropogenic sources of Hg release
meeting specified criteria, there still is the
potential that we may included some
watersheds that should have been excluded.
Specifically, this concern exists in LA, SC,
MN, and ME, where there are either broader
concerns over non-air mercury sources (e.g.,
taconite mining in MN and gold mining in SC)
or where there appear to be relatively higher
mercury fish tissue levels in the absence of
elevated mercury deposition (again raising the
concern over non-air deposition sources). To
examine the potential impact of locations with
elevated non-air mercury sources, we
completed two sensitivity analyses including:
(a) an analysis of risks when watersheds falling
in these four states are excluded and (b) an
analysis for risk only for the subset of 2,366
watersheds falling in the upper 25th percentile
with regard to total mercury deposition (i.e.,
those watersheds having relatively elevated
mercury deposition) (see Section 2.7 for
additional detail on these sensitivity analyses).
As part of our sensitivity analyses, we also
83
-------
Source of
uncertainty
Description
Nature of potential impact on the exposure and
risk estimates
Degree to which the potential impact of the
source of uncertainty is characterized as
part of the analysis
that the MMaps approach is
more applicable in the context
of stationary waterbodies
(lakes and ponds) than flowing
waterbodies (streams and
rivers).
EGU-attributable risk, since we would overstate the
role of U.S. EGUs in contributing to risk, by
overlooking the other non-air sources.
If the MMaps approach is more effective at linking
areal Hg deposition to Hg fish tissue levels for
stationary waterbodies, then our application of the
MMaps approach to both stationary and flowing
could result in a distribution of watershed-level risks
that could be incorrect.
examined the issue of applying MMaps to
stationary and flowing waterbodies.
Specifically, we ran the model for a lakes-only
simulation and compared these results with the
baseline run including both lakes and rivers
(see Section 2.7).
Estimating IQ
loss for children
born to high-fish
consuming
mothers modeled
in the analysis
There are a number of sources
of uncertainty associated with
modeling IQ loss in this
analysis: (a) IQ may not fully
capture the most sensitivity
cognitive endpoints associated
with mercury exposure , (b)
potential confounding from
long-chained polyunsaturated
fatty acids (LCPFAs) found in
fish, and (c) including potential
outliers from the
epidemiological datasets used
in deriving the IQ loss
functions.
IQ may not represent the most sensitive cognitive
endpoint for mercury exposure (Axelrad et al., 2007 -
see Section 1.2). In addition deficits in some
categories of cognitive functioning are not captured
by IQ. Together, these sources of uncertainty suggest
that we could be under-estimating the extent of
cognitive impacts associated with mercury exposure
(when we focus on modeling the IQ loss endpoint
alone). The potentially beneficial effects of LCPFAs
(found in fish) on neurological development can
confound the effects of mercury by potentially
masking those adverse effects. This would result in
IQ loss slopes that are biased low, since the IQ loss
they are representing is counteracted to some extent
by the LCPFAs exposure. Regarding outliers, when
an outlier datapoint from the Seychelles study was
included in the integrated derivation of the IQ loss
slope factor, the factor was reduced by 25 percent
(from -0.18 IQ points per unit ppm hair mercury, to -
0.125). If in reality, this outlier actually reflects the
true response for a subset of the populations, then
risks (as modeled) could be biased high specifically
for this subpopulation.
Because we do not have readily available data
to support quantitative analyses of the first two
sources of uncertainty (IQ loss not capturing all
of the cognitive effects and potential
confounding by LCPFAs), we could only
address these factors qualitatively. We note,
that in both cases, the potential effect on the
risk assessment would be to potentially down-
bias our estimate of cognitive endpoint-related
risk for children. In the case of excluding the
outlier from the Seychelles study, we note that
the effect (given the linear nature of the IQ loss
slope) would be to simply result in a 25%
reduction in risk, if we were to include the
outlier in derivation of the slope function (i.e., a
formal rerun of the model with this alternative
slope is not required - we can just consider this
magnitude of impact on the primary risk
estimates we generate for the analysis).
84
-------
G. Potential for Deposition "Hotspots" in Areas Near U.S. EGUs
EPA also evaluated the potential for "hot spot" deposition near U.S. EGU emission
sources on a national scale, based on the CMAQ modeled Hg deposition for 2005 and 2016. We
calculated the excess deposition within 50 km of U.S. EGU sources by first calculating the
average U.S. EGU attributable Hg deposition within a 500 km radius around the U.S. EGU
source. This deposition represents the likely regional contribution around the EGU. We then
calculated the average U.S. EGU attributable Hg deposition within 50 km of the U.S. EGUs to
characterize local deposition plus regional deposition near the EGU. Excess local deposition is
then the 50 km radius average deposition minus the 500 km radius average deposition. Figure 1
shows a map of the excess local deposition based on the 2005 CMAQ modeling. Figure 2 shows
excess local deposition based on the 2016 Base Case.
The maps in Figures 1 and 2 were generated by applying an averaging kernel to the 12km
EGU attributable mercury deposition estimates from CMAQ. Averaging kernels assign a mean
value to each grid based on the averages of all neighboring grids within a predefined window. In
this case, window or kernel sizes were 50 km and 500 km-radiuses. Then 50km-radius average
values were subtracted from SOOkm-radius averages to create the final hot spot image.
Summary statistics for the excess local deposition are provided in Table 1. Table 1
shows both the mean excess deposition around all U.S. EGUs, and the mean excess deposition
around just the top 10 percent of Hg emitting U.S. EGUs. Table 1 also shows the excess Hg
deposition as a percent of the average regional deposition to provide context for the magnitude of
the local excess deposition. In 2005, for all U.S. EGU, the excess was around 120 percent of the
average deposition, while for the top 10 percent of Hg emitting U.S. EGU, local deposition was
around 3.5 times the regional average. By 2016, the absolute excess deposition falls, however,
the local excess still remains around 3 times the regional average for the highest 10 percent of
Hg emitting U.S. EGUs.
This analysis shows that there is excess deposition of Hg in the local areas around EGUs,
especially those with high Hg emissions. Although this is not necessarily indicative of higher
risk of adverse effects from consumption of MeHg contaminated fish from waterbodies around
the U.S. EGUs, it does indicate an increased chance that Hg from U.S. EGUs will impact local
waterbodies around the EGU sources, and not just impact regional deposition.
85
-------
Figure G-l. Excess Local Deposition in 2005
EGU plant locations
2005
Total Hg Emissions (tons/year)
• 0.000-0.013
• 0.014-0.037
• 0.038 - 0.068
o 0.069-0.108
o 0.109-0.159
o 0.160-0.226
0 0.227 - 0.309
o 0.310-0.424
o 0 425 - 0.639
• 0.640 - 1.096
2005 Hg Deposition Hot Spots
Excess Local Deposition Values (M9/m2)
86
-------
Figure G-2. Excess Local Deposition in 2016 (Base Case)
ECU plant locations
2016 Base case
Total Hg Emissions (tons/year)
• 0.000-0.013
• 0.014-0.037
e 0.038 - 0.068
o 0.069-0.108
o 0.109-0.159
o 0160-0.226
o 0.227 - 0.309
o 0.310-0.424
o 0.425 - 0.639
• 0.640 - 1.096
*•„«, ,/"»«&
'fto ££ 9 o-
» o
o o o
Ct'
o . °»,
O O o . •
o o oo -< .o
PW^^^^Bt. ". o
^FT ^fc w
\
2016cr Hg Deposition Hot Spots
Excess Local Deposition Values (jjg/m2)
V ^ ^ A*
-------
Table G-l. Excess local deposition of hg based on 2005 CMAQ modeled Hg deposition
All U.S. ECU sites with Hg emissions >0
(672 sites)
Top ten percent U.S. ECU in Hg
emissions
(67 sites)
50km-Radius-Average Excess Local Deposition values
(ug/m2)
Mean Across EGUs (percent of regional average deposition)
2005
1.65(119%)
4.89 (352%)
2016
0.36 (93%)
1.18(302%)
88
-------
United States Office of Air Quality Planning and Standards EPA-452/D-11-002
Environmental Protection Health and Environmental Impacts Division March 2011
Agency Research Triangle Park, NC
89
------- |