Appendix B. Case Study for Integrating a Distributional Approach to Using
Percent Crop Area (PCA) and Percent Crop Treated (PCT) into Drinking Water
Assessment
1.	Executive Summary
Drinking water assessments (DWA) follow a tiered process that is used to distinguish pesticides
that that do not pose a potential risk from pesticides that may require a detailed and more in-
depth analysis. The Percent Cropped Area (PCA)/Percent Crop Treated (PCT) project provides
an approach to apply use and usage data to refine estimated drinking water concentrations
(EDWCs) in higher-tier assessments for agricultural and non-agricultural uses individually or in
combinations. The goal of the PCA and PCT refinements are to generate EDWCs that are
appropriate for human health risk assessment that reduce the magnitude of overestimation
due to variability in crops and actual pesticide usage. The background and concepts of the
PCA/PCT project are presented in the PCA/PCT White Paper while this case study provides an
example of employing these new methods in a highly refined DWA on a Hydrologic Unit Code
(HUC-2) basis.
This case study first provides a high-level summary of the scoping and problem formulation
process for a pesticide called "pestl", including the fate properties and use patterns, as well as
the results of the Tier 2 DWA for HUC-2 Regions 03, 04 and 05. The subsequent sections of the
case study describe the application of the new PCA and PCT methods as a Tier 3 refinement. For
this case study, after applying the new PCA method, the EDWCs in HUC-04 are all expected to
be below the level of concern, while the results for HUC-03 and HUC-05 indicate the need for
further PCT refinements. After applying the new PCT method, in HUC-03, PCT refinement had a
minimal impact on the EDWCs due to the lack of PCT data for non-agricultural uses. PCT
refinements in HUC-05 reduced the number of watersheds with Drinking Water Level of
Comparison (DWLOC) exceedances by 37%.
This case study focuses specifically on applying this new approach to utilize PCA and PCT in Tier
3 DWA refinements to the EDWC in surface water for pestl, and not on the DWA process up to
that point which has been already completed. A description of the entire tiered DWA process
can be found in the Draft Framework for Conducting Pesticide Drinking Water Assessments for
Surface Water (DWA Framework) (USEPA, 2019).
2.	Scoping and Problem Formulation
The DWA Framework describes the problem formulation and scoping process prior to, and
during, the conduct of a DWA. This involves a holistic look at the pesticide to be assessed, what
is known about the potential and actual use, the underlying environmental fate and human
health hazard data (i.e., DWLOC), and the results of previous risk assessments.
1

-------
a. Use Characterization
Pestl is a nationally registered insecticide with 80 uses in terrestrial food and feed crops,
terrestrial non-food crops, greenhouse food/non-food, and non-agricultural indoor and outdoor
crops. The spatial distribution of the average agricultural usage data in lbs pestl/agricultural
acre between 2008-2012 are presented in Figure 1.
Average lbs pestl per agricultural acre
Not surveyed
No use reported
I. ! 0.01 -0.10
m °11020
^ 0.21 - 0.30
0 31 - 0 40
> 0.40
Figure 1. Spatial Distribution of Pestl Agricultural Use (2008-2012)
Based on yearly average usage data from 2004 to 2013 provided by the Biological and Economic
Analysis Division (BEAD), approximately 7.2 million pounds of pestl are used each year for
agricultural purposes in the United States. Approximately 21% and 19% of the total volume of
pestl used in the United States each year is applied to soybeans (1.5 million lbs) and corn (1.4
million lbs), respectively. On average only 5% of total soybean acreage and about 2,5% of total
corn acreage is treated with pestl each year. Other crops with relatively high usage of pestl (at
least 100,000 lbs/year) include alfalfa, almonds, apples, apricots, cotton, grapes, oranges,
peanuts, pecans, sugar beets, walnuts and wheat. A large fraction, at least 40%, of the total
acreage planted with apples, asparagus, broccoli, onions, and walnuts, is treated with pestl.
Agricultural usage has declined every year from 1992 - 2012 as shown in Figure 2.
2

-------
' j'lhli
Nn^mwsomot-Nn^iocoNesmo'-N
fflOiffimOOimmOOOOOOOOOOrrr
mcnmuimuicimooooooooooooo
CVI(NJ(\Jl>JCMr^Jl>J
-------
c. Previous Assessments
As described above, this case study builds on a DWA completed up to Tier 2 consistent with the
DWA Framework (USEPA, 2019). The previous Tier 2 DWA for pestl produced EDWCs that
covered maximum label rates for all uses using standard modeling scenarios. It included
consideration of typical rates and regional (HUC-2 scale) PCA adjustment factors.
At the Tier 2 assessment level, many use scenarios resulted in concentrations above the DWLOC
even when considering regional use patterns (i.e., maximum regional percent cropped area)
and when assuming non-agricultural uses had minimal influence on the overall exposure
profile. The range of Tier 2 EDWC by use site are shown graphically on Figure 3.
The results of the Tier 2 DWA for pestl indicates that further refinements consistent with Tier 3
of the DWA Framework (USEPA, 2019) are needed.
140
rr 120
"O
a>
¦*->
to
E
a>
"O
o
u
5
a.
100
80
60
40
20
NATIONAL SUMMARY
11-day ¦ 21-day
DWLOC=5|Ig/L
H+h-
III""'
3 <0*

er

Use Site
/•
* v ¦
Figure 3. National Screening Level Estimated Concentrations for the Case Study Pestl
Resulting from Maximum Labeled (single and 21-day rolling average) Rates and Minimum
Retreatment Intervals for Uses on Agricultural Sites and Non-Agricultural Sites
d. Current Assessment
This case study lays out an example of how PCA and PCT refinements as described in the
PCA/PCT White Paper can be incorporated into a Tier 3 DWA using pestl as an example. In
order to demonstrate the proposed approach for incorporating the full distribution of CWS
watershed PCAs and PCTs, this case study begins with the results from the regional Tier 2 level
assessment and focuses on three, HUC-2 Regions 03, 04 and 05 as summarized in the sub-
sections below. Simplifications of the use profile and other factors have been made for each
region for brevity of the case study. In addition, for the purposes of this case study only the
4
-------
4838 quality assured delineated drinking water intake watersheds are considered1. This method
will ultimately be extended to all delineated CWS watersheds and the HUC-12 surrogates
described in the Development of Community Water System Drinking Water Intake Percent
Cropped Area Adjustment Factors for use in Drinking Water Exposure Assessments: 2014
Update (USEPA, 2014). Also, this case study uses a 21-day average DWLOC of 5 |ag/L.
3. Tier 2 Analysis
a.	Introduction
Consistent with the DWA Framework (USEPA, 2019) Tier 2 considers maximum use rate data
and regional PCAs as described in the DWA Framework. This section describes the results for
the three representative HUC-2 Regions chosen for this case study which inform the transition
to Tier 3 and the application of the distributional PCA and PCT refinements to EDWC values for
pestl in three representative HUC-2 Regions: regions 03, 04, and 05.
b.	PCA and Watershed Descriptions
i.	South Atlantic-Gulf (HUC-2 Region 03)
The South Atlantic-Gulf region (HUC-2 Region 03) encompasses several states including Florida,
Georgia, North Carolina, South Carolina, and Alabama, parts of Mississippi, Louisiana and
Virginia. There are roughly 470 known surface water intakes used to supply drinking water in
the region. In the South Atlantic-Gulf region, pestl may be applied to orchard crops including
apples and cherries as well as non-agricultural setting (i.e., wide area use) but not in residential
areas (i.e., turf). This region includes use sites with high labeled use rates (4 lbs ai/A per year).
The maximum regional all-agricultural and non-agricultural land PCA is 1 for all CWS
watersheds. Based on Tier 2 modeling and maximum label rates, the l-in-10 year 21-day
average concentrations for the South Atlantic-Gulf region are 17.5, 49.6, and 55.6 ju.g/L for
apple, citrus, and non-agricultural, respectively. Therefore, EDWCs for pestl resulting from
maximum label rates are above the DWLOC of 5 |ag/L in the South Atlantic-Gulf region and
more refinements need to be considered.
ii.	Great Lakes (HUC-2 Region 04)
The Great Lakes region (HUC-2 Region 04) encompasses primarily Michigan with overlap with
small portions of Ohio, Indiana, Pennsylvania, and New York. There are 150 known surface
water intakes used to supply drinking water in the region. The Great Lakes region has a diverse
range of crops to which pestl may be applied. This region includes use sites with high labeled
use rates (4 lb a.i./A/yr on orchards). The maximum regional all-agricultural and non-
agricultural land PCA is 0.92 for all CWS watersheds. Based on Tier 2 modeling and maximum
label rates, the l-in-10 year 21-day average concentrations for region 04 are 90.9, 16.1, 42.4,
9.8, and 7.8 ju.g/L for cherries, apples, corn, soybeans, and sugar beets, respectively. Therefore,
1 Details on development of the CWS PCA and their use as a DWA refinement may be found in the PCA/PCT White
Paper and Development of Community Water System Drinking Water Intake Percent Cropped Area Adjustment
Factors for use in Drinking Water Exposure Assessments: 2014 Update (USEPA, 2014)
5
-------
EDWCs for pestl resulting from maximum label rates are above the DWLOC of 5 |ag/L in the
Great Lakes region and more refinements need to be considered.
iii.	Ohio (HUC-2 Region 05)
The Ohio Region (HUC-2 Region 05) encompasses Ohio, and parts of Pennsylvania, West
Virginia, Indiana and Kentucky. There are over 600 known surface water intakes used to supply
drinking water in this region. HUC-2 05 has a diverse range of crops to which pestl may be
applied. This area contains use sites with high labeled use rates (4 lbs ai/A per year on
orchards). The maximum regional all-agricultural and non-agricultural land PCA is 0.89 for all
CWS watersheds. Based on Tier 2 modeling and maximum label rates, the l-in-10 year 21-day
average concentrations for region 05 are 87.9,15.6, 41.0, 9.5, and 22.1 ju.g/L for cherries,
apples, corn, soybeans, and alfalfa, respectively. Therefore, EDWC for pestl resulting from
maximum label rates are above the DWLOC of 5 |ag/L in the Ohio Region and more refinements
need to be considered.
iv.	Tier 2 Summary
Results of the Tier 2 DWA for pestl in HUC-2 Regions 03, 04, and 05 demonstrate that each
HUC-2 EDWCs are above the DWLOC. While not proposed for Tier 2, the application of the full
suite of PCA's for each HUC-2 indicates most of the watersheds require additional refinement.
Table 2. Summary of Tier 2 Results by HUC-2 Region
Region
Use Sites3
Highest Estimated 21-
day Average
Total Number of
CWSb
South Atlantic-Gulf (HUC-2
Region 03)
non-agricultural, citrus, and apple
55.6
468
Great Lakes (HUC-2 Region
04)
cherries, apples, corn, soybean, and
alfalfa
90.9
151
Ohio (HUC-2 Region 05)
cherries, apples, corn, soybeans, and
alfalfa
87.9
626
a Bold use site indicates use pattern with maximum EDWC
b based on the 4838 delineated CWS watersheds only
4. Tier 3 Analysis
As described in the PCA/PCT White Paper and consistent with the DWA Framework (USEPA,
2019), after considering standard Tier 2 refinements, our proposed Tier 3 DWA considers
additional data to further refine EDWCs including using the full distribution of individual CWS
watershed PCAs where the EDWCs exceed the DWLOC followed by pesticide specific usage data
(i.e., Percent Crop Treated, or PCT data). It is important to note that, because the PCA data are
readily available within EFED and the data are generic to all pesticides, application of PCA
refinements is often easier than PCT refinements. Usage data provided by BEAD is pesticide
specific and must be generated on a case by case basis. For pestl, pesticide usage data
including PCT is available from previous assessments which included information on typical
application rates and dates of application. Usage information including typical application rates
6
-------
and application timing is integrated in this case study with incorporation of regionally specific
modeling scenarios for HUC-2 Regions 03, 04, and 05.
a. Percent Cropped Area Analysis
There are several potential Tier 3 PCA refinement options to better understand the exposure
potential of pestl as described in the White Paper. The PCA refinements applied in this case
study are summarized below and the results are presented by HUC-2 Region.
1.	APPLY USE PATTERN SPECIFIC PCA. The use pattern specific PCA is the PCA value for the
combination of crops/groups of crops specific to the registered uses of a pesticide. This
captures the area of the watershed allocated to proposed or registered use sites, rather
than using the default all-agricultural land PCA as is typically applied at Tier 2 when
there are multiple crop uses or Tier 1 when non-agricultural uses are under
consideration, understanding that both options can overestimate the use footprint. An
example is the orchard PCA being applied to cherries or apples, but not both.
2.	EXAMINE FULL DISTRIBUTION OF WATERSHED PCA VALUES. Instead of only using the
maximum national/regional PCA value, the Tier 3 PCA analysis considers all watershed
PCAs within each HUC-2 Region to identify the percentage of watersheds that have PCA
adjusted EDWCs that exceed the DWLOC and whether the pestl specific use sites (e.g.,
cherries) occur in the respective watersheds.
3.	CALCULATE THE CRITICAL PCA AND PERCENT OF WATERSHEDS WITH PCA VALUES
LARGER THAN THE CRITICAL PCA. The Critical PCA, the ratio between the unrefined
EDWC and the DWLOC, is the PCA value that would generate a refined EDWC equal to
the DWLOC. The Critical PCA quickly identifies the percentage of watersheds within each
region with exposure concerns and is applied to each use alone and in the aggregate.
4.	COMPARE OVERLAP OF WATERSHEDS WITH PCAS LARGER THAN THE CRITICAL PCA
WITH USE SITE FOOTPRINT. PCA values for groups of crops (i.e., orchards, vegetables)
are derived from generalized crop data layers based on the National Land Cover
Database (NLCD) and Census of Agriculture (Ag Census). This approach has the potential
to overestimate the percent of a given watershed with the noted use site (e.g., planted
with a single crop). For instance, an individual CWS watershed with an orchard PCA of
20% may very well have little or no cherries grown within the watershed. Spatial overlap
helps further identify CWS watersheds with potential exposure concerns.
5.	DEVELOP AGGREGATED ESTIMATED DRINKING WATER CONCENTRATIONS. Prior to this
step, EDWCs are based on the highest EDWC of all uses determined using modeling
scenarios for individual uses or generalized crop groups, however, the relative
contributions of each modeled use site can be determined by adding the contribution
concentrations (i.e., EDWCs*PCA) based on relative contribution within each CWS
watershed. This is the sum of the crop-specific PCA adjusted EDWC values for each
registered crop/group of crops within each watershed. This aggregation step is actually
7
-------
a two-step process. The first step is to aggregate individual 1 in 10 year EDWCs for each
use site in a region without regard to timing (calendar day). For CWS watersheds that
continue to exceed the DWLOC a second step can be employed where individual time-
series (e.g., chemographs) from each modeled use (e.g., cherries and turf) can be added
together on a calendar day basis. This process of aggregating chemographs can be
performed manually or can be automated (in this case study the process was performed
manually). This second step captures the temporal variability across uses with different
application timing and weather.
i. South Atlantic-Gulf (HUC-2 Region 03)
In the South Atlantic-Gulf region, non-agricultural uses are permitted, including wide-area
mosquito adulticide use. As such, in the Tier 2 DWA, a PCA of 1 for all CWS watersheds is used
when there are agricultural and non-agricultural (e.g., rights of way, perimeter treatment) uses.
As a first step, a use pattern (i.e., non-agricultural and orchard) specific PCA was calculated by
subtracting out the cropped area that does not correlate to uses (i.e., corn, wheat, soybean,
cotton, and vegetable) permitted in the HUC-2 Region from the maximum all-agriculture/non-
agricultural PCA of 1 used in Tier 2. Using this approach for HUC-2 Region 03 the maximum
refined PCA for the registered uses for pestl is 0.65. This PCA includes non-agricultural uses
sites including institutional turf (e.g. sports fields) and wide-area uses. The distribution of
EDWCs for the South Atlantic-Gulf region are shown in Figure 4 for all CWS watersheds.
Applying this use-site specific PCA, for pestl use on apples, 92% of CWS watersheds have
EDWCs below the DWLOC while for non-agricultural and citrus uses of pestl less than 18% of
watersheds have EDWCs below the DWLOC. EDWCs for pestl resulting from maximum label
rates for all three uses remain above the DWLOC of 5 |ag/L in the South Atlantic-Gulf region
when the use-site specific PCA is applied. As such, more refinements needed to be considered
for all uses.
8
-------
HUC-03
0.4 0.6
Percentile
• Non-ag • Citrus • Apple • DWLOC
Figure 4. Estimated l-in-10 Year 21-Day Average Concentrations Adjusted for the Individual
Watershed PCAsforthe pestl Resulting from Maximum Labeled (single and yearly) Rates and
Minimum Retreatment Intervals for Uses in HUC-2 Region 03
Usage data were examined, and EDWC were updated as a next step in the refinement process
for those watersheds and uses that have EDWCs above the DWLOC. Table 3 lists the EDWCs of
pestl in HUC-2 Region 03 based on the typical use rates provided by BEAD. Note that no typical
use information is available for the non-agricultural uses permitted in HUC-2 Region 03;
therefore, the maximum label rates continue to be used for non-agricultural uses.
Table 3. Summary of Estimated Pestl Concentrations Considering Maximum Regional (HUC-2
Region 03) Percent Cropped Area Adjustment Factors
Use
Maximum
Regional PCA
Estimated 21-
day Average
Concentration
Refined
Regional Use
Pattern Specific
Maximum PCAs
Typical Use
Number of
Applications;
Rate (lb/A)
l-in-10 Year
Estimated 21-
day Average
Concentration
(Typical Rate)a b
l-in-10 Year
Refined Regional
Adjusted 21-day
Average
Concentration13
Tier 2
Tier 3
Apple
1.0
17.5
0.65b
i; 4
7.0
4.55 (NR)
Citrus
49.6
1; 2.8
17.5
11.14
Non-ag
55.6
Unknown
55.6
36.01
NR No more Refinement needed, EDWC
-------
concentrations of pestl for applications to apples is below the DWLOC. Examination of the full
distribution PCAs identifies what percentage of watersheds within a HUC-2 Region have a use
pattern-specific PCA-refined EDWC value that still exceeds the DWLOC. In this refinement step,
the Critical PCA, which is defined as the PCA value which when applied to a refined EDWC will
be below the DWLOC, is identified. The Critical PCA value for the pesticide is calculated from
the DWLOC and EDWC (PCA unadjusted value) using Equation 1.
Equation 1	Critical PCA = DWLOC/EDWC^^
Where:
Critical PCA =	PCA value below which the PCA-adjusted EDWC is less than the DWLOC
EDWCmax =	Maximum EDWC for a given crop or use pattern
DWLOC =	Drinking water level of concern
Based on this analysis in HUC-2 Region 03, the critical PCA is 0.09 and 0.29 for non-agricultural
use and citrus crops, respectively. Applying these PCAs to each individual CWS watershed in this
region indicates that roughly 14% (non-agricultural) and 92% (citrus) of watersheds have
refined EDWC below the DWLOC for this region. This is illustrated in Figure 5.
HUC-03
0.4 0.6
Percentile
• Non-ag • Citrus * Apple • DWLOC
Figure 5. Distribution of Estimated l-in-10 Year 21-Day Average Concentrations Adjusted for
the Individual Watershed PCAs for the pestl Resulting from Typical Applications in HUC-2
Region 03
The next refinement is to compare the overlap of the CWS watersheds area and use site specific
land cover area where the PCAs are above the Critical PCAs. In the previous step an assumption
is made using generalized PCA data layers (e.g., orchards) which were derived using the
National Land Cover Database (NLCD) All-Ag data layer and generalized acreage data from the
Census of Agriculture (Ag Census) as described in the original CWS PCA documentation (USEPA,
10
-------
2014). For example, specific orchard crops (e.g., citrus) are represented by a generalized
orchard data layer.
The previous analysis has the potential to overestimate the percent of a given watershed that is
planted with a specific orchard crop. Therefore, an additional step is employed where an
overlap analysis is conducted which compares individual CWS watersheds with crop specific
information from the Ag Census with the goal to exclude CWS watersheds where the specific
crop is not likely to be present. For CWS watersheds with no acreage being reported in Ag
Census as being present for a specific crop [e.g., apples), it is assumed that no use would be
present and thus no exposure from that use would be possible and the CWS watershed can be
excluded from further refinements. Conversely, a county where Ag Census suggests a crop may
be present, even in cases where no explicit acreage is reported, the county is assumed to have
the use and the EDWCs are considered present. This is done because it is not uncommon for
data to be censored in order to protect grower confidentiality. The results of this analysis are
shown in Figure 6. This figure shows that citrus growing counties overlap with watersheds with
PCAs above the Critical PCA in central Florida. These areas may also have non-agricultural uses
of pestl.
Figure 6. Community Water System Watersheds where Watershed PCA > Critical PCA of 9%
for HUC-2 Region 03 Overlaid with Orchard crop footprint (shown in pink and gray) and
county level Citrus acreage (shown in yellow).
For the watersheds and uses that have EDWCs above the DWLOC, the next step in the
refinement process is aggregation of relevant uses to generate a more realistic EDWC. This
moves away from the assumption that the entire area of the watershed is entirely planted with
Watersheds
PCA value
Counties with Citrus
0-0.09
>0.09
No acreage reported
Acreage reported
11
-------
the crop that leads to the highest EDWC by weighting the EDWCs using watershed landcover
data (i.e., PCA) on a crop-specific basis. Instead aggregation allows for the relative contributions
of each use within the watershed to be considered in the EDWC as shown in Equation 2. The
White Paper outlines a two-step approach for developing an aggregated EDWCs. The first
approach uses the l-in-10 year concentration while the second approach uses the individual
time series for each simulation to develop a new time series of data for each CWS watershed
which a l-in-10 year concentrations can be estimated. The first step is expected to be a
conservative approach as the l-in-10 year concentrations are not expected to occur on the
same days for all uses under consideration. The second approach adds the individual PCA
adjusted chemographs together to create a new chemograph. Note that only one chemograph
for each available PCA (i.e., use category) must be selected for the aggregation process (e.g.,
apples or cherries but not both). The l-in-10 year concentration is calculated from the new
aggregated chemograph and compared to the DWLOC. This second approach allows for
consideration of the temporal aspect of exposure which can differ due to timing or application
among other factors occuring within a watershed.
Equation 2	(CWS crop specific (n)PCAx crop specific(n) EDWC) +
(CWS crop specific (n + l)PCAx crop specific (n + 1)EDWC) = aggregated EDWC
Where:
CWS crop specific PCA =	Community water system specific percent cropped area adjustment factor
Crop specific EDWC =	Crop or crop group specific estimated concentration
Aggregated EDWC	Aggregated estimated drinking water concentration
In HUC-03, n represents non-ag use while n+1 represents orchard use in the above equation. As
a first step, the l-in-10 year concentration for each use site (i.e., non-agricultural and orchard)
can be adjusted by the CWS specific PCA for each crop or crop group and then be added
together to generate an aggregated EDWC. This approach suggests that roughly 67 CWS
watersheds of the total 468 CWS watersheds have pestl concentrations below the DWLOC.
Refining further using the entire time series of estimated concentrations suggests that many of
the roughly 67 CWS watersheds still have concentrations below the DWLOC and additional
refinements need to be considered. This is because non-ag use of pestl is driving the aggregate
exposure estimates in this example.
Table 4. Summary of Estimated Pestl EDWC using Maximum Crop-Watershed Specific PCAs
Relevant to High Orchard Production in HUC-2 Region 03	
Use
Typical Use
Number of
Applications;
Rate (lb/A)
Estimated 21-day
Average
Concentration
(Hg/L)
Maximum Crop-
Watershed
Specific PCA
Refined PCA Adjusted
Individual
Concentrations
(w>/l)
Aggregated
Estimated
Concentration
(Hg/L)
HUC-03
Non-ag
Unknown
55.6
0.65
36.4
38.2
Apple
i; 4
17.5a
0.12
2.1a
Citrus
1; 2.8
7.0
NR
NR
a. Used maximum concentration from all orchard uses.
12
-------
b. NR in this table to show aggregation process but citrus is considered in appropriate watersheds.
ii. Great Lakes (HUC-2 Region 04)
As was done for HUC-2 Region 03, a use pattern specific PCA that accounts for orchards, corn,
soybean, and sugar beets was calculated by subtracting out the cropped area that does not
correlate to uses (i.e., wheat, cotton, rice and vegetable) permitted in HUC-04 from the
maximum all-agriculture for each respective watershed. This approach functionally develops a
miscellaneous agricultural PCA for crops for which PCAs have not been specifically developed,
in this case, sugar beets. The maximum PCA for miscellaneous agriculture for HUC-04 is 0.50.
The highest use site (i.e., orchard, corn, soybean, and miscellaneous) specific PCA calculated is
0.85. The distribution of EDWCs for the Great Lakes region are shown in Figure 7 for all CWS
watersheds. EDWCs for pestl resulting from maximum label rates are above the DWLOC of 5
Hg/L in the Great Lakes region and more refinements needed to be considered for all uses.
HUC-04
c
o
4-"
ro
4->
c
(LI
O
c
o
u
"D
(L)
4-"
ro
E
100
50
&o
IL
1.00
0
0.00 0.20 0.40 0.60 0.80
Percentile
« • Cherry	• Apple	• Corn
« Soybean	• Sugarbeet • DWLOC
Figure 7. Estimated l-in-10 Year 21-Day Average Concentrations Adjusted for the Individual
Watershed PCAs for the pestl Resulting from Maximum Labeled (single and yearly) Rates and
Minimum Retreatment Intervals for Uses in HUC-2 Region 04
Usage data were integrated, and model estimates were updated including use of regionally
specific PCAs (e.g., cherry represented by HUC-2 Region 04 orchard PCA) as a next step in the
refinement process. Table 5 lists the EDWCs of pestl in HUC-2 Region 04 based on the typical
use rates provided by BEAD in combination with the use site specific PCA adjusted
concentrations. These refined EDWCs are below the DWLOC for pestl use on corn, soybean,
and sugar beet in HUC-2 Region 04. Additional refinements are necessary for cherry and apple.
13
-------
Table 5. Summary of Estimated Pestl Concentrations Considering Maximum Regional (HUC-
04) Percent Cropped Area Adjustment Factors	
Use
Estimated 21-
day Average
Concentration
Maximum
Regional PCA
Refined
Regional Use
Pattern Specific
Maximum PCAs
Typical Use
Number of
Applications;
Rate (lb/A)
l-in-10 Year
Estimated 21-
day Average
Concentration
(Typical Rate)3
l-in-10 Year Refined
Regional Adjusted 21-
day Average
Concentration13^
Tier 2
Tier 3
Cherry
90.9
0.92
0.84
i; 2
31.4
26.4
Apple
42.4
1; 2.8
44.0
37.0
Corn, Field
16.1
i; 1
3.8 (NR)
NR
Soybean
9.8
1; 1.1
4.2 (NR)
NR
Sugar beet
7.8
1; 1.2
4.7 (NR)
NR
NR No more Refinement needed, EDWC	Percentile
c
2 • Cherry • Apple • DWLOC
Figure 8. Distribution of Estimated l-in-10 Year 21-Day Average Concentrations Adjusted for
the Individual Watershed PCAs for the pestl Resulting from Typical Applications in HUC-2
Region 4
14
-------
As done for HUC-2 Region 3, a comparison of the watersheds with PCAs equal to or greater
than the Critical PCAs with Ag Census acreage data for cherries and apple is shown in Figure 6.
Overlap analysis for HUC-2 Region 04 shows that CWS watersheds with orchard PCAs greater
than the critical PCA for a given crop/group of crops overlap with cherry and apple acreage.
Figure 9. Community Water System Watersheds where Watershed PCA > Critical PCA OF 11%
for HUC-2 Region 04 Overlaid with Orchard crop footprint
As a next step, aggregated EDWCs were developed which considers contributions from all crop
uses for the overlapping watersheds. The l-in-10 year 21-day average concentration from
manually aggregated chemographs2 (i.e., sum of PCA-adjusted chemographs and recalculation
of l-in-10 year 21-day average concentration) are not expected to be higher than the DWLOC.
As such, no additional refinements are necessary for pestl uses in HUC-2 Region 4.
iii. Ohio (HUC-2 Region 05)
Like done for HUC-2 Region 4, a use pattern specific PCA for orchard, corn, soybean, and
miscellaneous (for alfalfa) was calculated by subtracting out the cropped area that does not
correlate to uses (i.e., wheat, cotton, rice and vegetable) permitted in HUC-04 from the
maximum all-agriculture PCA for each respective watershed. Like sugar beets in HUC-03, the
2 When the PCA/PCT project is fully implemented the process of aggregating modeled chemographs will be
automated.
Counties with Cherries
No acreage reported
Acreage reported
Counties with Apples
No acreage reported
Acreage reported
Watersheds
PCA value
15
-------
approach functionally develops a miscellaneous agricultural PCA for crops for which PCAs have
not been specifically developed. In this case, for alfalfa. The maximum PCA for miscellaneous
agriculture for HUC-2 Region 4 is 0.78. The highest use site (i.e., orchard, corn, soybean, and
miscellaneous) specific PCA calculated is 0.88. The distribution of individual watershed PCA
adjusted EDWCs for the Ohio region are shown in Figure 10 for all CWS watersheds. EDWCs for
pestl resulting from maximum label rates are above the DWLOC of 5 |ag/L in the Great Lakes
region and more refinements needed to be considered for all uses.
HUC-05
100
"O
Percentile
(TJ
£
t]	• Cherries	• Apples	• Corn
Soybean	• Alfalfa	• DWLOC
Figure 10. Estimated l-in-10 Year 21-Day Average Concentrations Adjusted for the Individual
Watershed PCAs for the Case Study Pesticide Resulting from Maximum Labeled (single and
yearly) Rates and Minimum Retreatment Intervals for Uses in HUC-2 Region 05
Usage data (e.g., typical rates and application timing) were incorporated into modeling along
with regionally specific modeling scenarios for both individual crops (e.g., cherries) and generic
crop groups (e.g., orchard) as a next step in the refinement process. Table 6 lists the EDWCs of
pestl in HUC-2 Region 05 based on the typical use rates provided by BEAD. In combination with
the use site specific PCA, concentrations are expected to be below the DWLOC for pestl use on
alfalfa; however, all the other uses result in pestl concentrations above the DWLOC.
16
-------
Table 6. Summary of Estimated Pestl Concentrations Considering Maximum Regional (HUC-
05) Percent Cropped Area Adjustment Factors	
Use
Estimated 21-
day Average
Concentration
Maximum
Regional PCA
Refined
Regional Use
Pattern Specific
Maximum PCAs
Typical Use
Number of
Applications;
Rate (lb/A)
l-in-10 Year
Model Estimated
21-day Average
Concentration
(Typical Rate)3
Refined Regional
PCA Adjusted l-in-
10 Year 21-day
Average
Concentration13^
Tier 2
Tier 3
Cherry
87.9
0.92
0.85°
i; 3
15.3
13.5
Apple
15.6
1; 2.8
14.3
12.6
Corn
41.0
1; 4a
17.4
15.3
Soybean
9.5
1; 1.1
12.8
11.2
Alfalfa
22.1
i; 1
4.5 (NR)
NR
NR No more Refinement needed, EDWC
-------
A comparison of the watersheds with PCAs equal to or greater than the Critical PCAs with
landcover data is shown in Figure 12 which indicates wide spread overlap and that additional
refinements are necessary.
Counties
Watersheds
PCA Value
0-0.29
> 0.29
with Cherries
No acreage reported
| Acreage reported
Counties with Apples
| No acreage reported
J Acreage reported
Counties with Corn
No acreage reported
Acreage reported
Counties with Soybeans
No acreage reported
¦ Acreage reported
	z	
Figure 12. Community Water System Watersheds where Watershed PCA > Critical PCA of 29%
for HUC-2 Region 05 Overlaid with Orchard Crop Footprint
Aggregated EDWCs suggest there are still a number of watersheds with concentrations above
the DWLOC in HUC-2 Region 05. The l-in-10 year 21-day average aggregated concentrations
are shown in Table 7 for simplicity; however, similar results were obtained when aggregated
chemographs were considered and the l-in-10 year 21-day average was calculated. The
similarity between aggregated individual EDWC and aggregated chemograph derived EDWC is
likely due to overlap in rainfall events across scenarios driving runoff events.
Table 7. Summary of Estimated Pestl EDWC using Maximum Crop-Watershed Specific PCAs
Relevant to Use Sites in HU<
1-2 Region 05
Use
Typical Use
Number of
Applications;
Rate (lb/A)
Estimated 21-day
Average
Concentration
(Hg/L)
Maximum Crop-
Watershed
Specific PCA
Refined PCA Adjusted
Individual 21-day
Average
Concentrations
(Hg/L)
Aggregated 21-day
Average Estimated
Concentration
(Hg/L)
HUC-05
Cherry
i; 3
15.3
0.034
0.52
13.3
Apple
1; 2.8
14.3
-
-
Corn
1; 4a
17.4
0.51
8.9
18
-------
Use
Typical Use
Number of
Applications;
Rate (lb/A)
Estimated 21-day
Average
Concentration
(Hg/L)
Maximum Crop-
Watershed
Specific PCA
Refined PCA Adjusted
Individual 21-day
Average
Concentrations
(Hg/L)
Aggregated 21-day
Average Estimated
Concentration
(Hg/L)
Soybean
1; 1.1
12.8
0.58
7.4

Alfalfa
i; 1
4.5
0.76
3.4
-- Only the maximum EDWC orchard crop was included in the aggregate EDWC calculation
a.	Used maximum concentration from all orchard uses.
b.	Since no PCA is available for sugar beet a surrogate PCA is calculated by subtracting all the individual
crops/crop groups from the all-Ag PCA.
i. Summary of PCA Analysis
Considering PCA refinements along with typical use and regionally representative scenarios,
there are more than 100 watersheds in HUC-2 Region 03 and HUC-2 Region 05 where
concentrations may still exceed the DWLOC. For HUC-2 Region 03, pestl use in non-agricultural
areas is driving the exposure conclusions while in HUC-2 Region 05 contributions from multiple
crop uses are resulting in EDWCs greater than the DWLOC. Overlays of the CWS watersheds
with the land cover data and aggregate EDWC values suggest the EDWCs in HUC-2 Region 04
are below the DWLOC for all CWS within the region.
b. Percent Cropped Treated Analysis
As described in the PCA/PCT White Paper there are several ways to integrate PCT data and
allocate or distribute the acres treated across a CWS watershed. Regardless of the method
chosen, the same series of steps are used to calculate the total treated acreage and distribute it
to the individual watersheds. This case study focused on examples using maximum PCT and the
upper and uniform distribution methods to demonstrate how the process would work. The PCT
refinements steps in this case study are summarized below and the results are presented for
HUC-2 Regions 03 and 05.
1. CALCULATE THE MAXIMUM STATE-LEVEL TREATED ACREAGE FOR EACH CROP BASED
ON THE MAXIMUM PCT VALUES AND CAG. USEPA Biological and Economic Analysis
Division (BEAD) provided the PCT values for each crop based on 5 years of survey data.
From the 5 years of data BEAD will provide a maximum PCT, minimum PCT, and an
average of the 5 years of PCT for each state covered by the labeled uses for pestl. The
surface water EDWC for CWS watersheds exceeding the DWLOC after consideration of
PCA refinements consistent with the PCA/PCT White Paper will be further refined by
considering the Base Acres Treated (BAT) within those watersheds. Consistent with the
White Paper, the selected PCT value (e.g., maximum) will be used to calculate the BAT
for each registered use by state.
19
-------
2.	ALLOCATE TREATED ACREAGE TO EACH CWS WATERSHED. Once a BAT is derived, there
are three proposed methods to distribute the treated acres across each CWS using the
upper, uniform, and lower distribution methods, as appropriate. This case study only
considers the upper and uniform distribution methods, as the lower distribution
resulted in no treated acres allocated to the watersheds.
3.	CALCULATE THE TREATED AREA SCALING PERCENTAGES (TASP) FOR EACH
WATERSHED/PCT/DISTRIBUTION METHOD COMBINATION. The TASP, the ratio of the
treated acreage within the watershed to the total area of the watershed, is the
multiplicative scaling factor used to adjust the unmodified EDWC values based on the
treated acreage within the watershed. It is similar to the PCA, but factors in both
percent cropped area and percent crop treated into a single value. Every watershed/
crop/PCT/distribution method will have a unique TASP.
4.	CALCULATE THE AGGREGATE TASP-ADJUSTED EDWC FOR EACH WATERSHED. Similar to
the aggregate PCT-adjusted EDWC, the relative contributions of each use site can be
determined by adding the contribution concentrations (i.e., EDWCs*TASP) based on the
relative contribution for each CWS watershed. This is the sum of the TASP adjusted
EDWC values for each registered crop/group of crops within each watershed.
5.	COMPARE ADJUSTED EDWC TO DWLOC. The new aggregate EDWC values for each
watershed are compared to the DWLOC to determine how many watersheds have
EDWCs that are expected to exceed the DWLOC.
i.	South Atlantic-Gulf (HUC-2 Region 03)
There is very limited PCT data available for non-agricultural pesticide use, which limits the
applicability of PCT refinements to Pestl. In the absence of reliable PCT data, the method
assumes 100% crop treated. Since the non-agricultural use pattern of pestl is the major
contributor to the EDWC, PCT refinements will have a minimal impact on the EDWC values.
While non-agricultural PCT data are not available for pestl, a non-agricultural PCT value of 9%
would be needed for concentrations to be below the DWLOC. This analysis points out the
importance of usage data for non-agricultural uses and the development of means to
approximate non-agricultural use when usage data is not available.
ii.	Ohio (HUC-2 Region 05)
For purposes of this case study, the Ohio PCT is assumed to represent the entire HUC-2 Region
05. For this example, this assumption is reasonable given that the bulk of HUC-2 Region 05 is
within Ohio. However, when fully implemented the PCT data for all states spanning each HUC-2
would be allocated proportionally relative to the percentage of each state overlap with the
entire HUC-2 Region using an automated process.
20
-------
Table 8 below presents the statistics for average annual acres grown and PCT for the uses of
pestl within HUC-2 Region 05 (as represented by Ohio) compiled by BEAD. The treated acreage
for each crop within the state is the product of the average annual acres grown and the
maximum PCT value. Between the years of 2010 to 2014, the crop with the largest treated
acreage was soybeans (55,400 acres treated), followed by field corn (37,100 acres treated).
These represent the maximum number of treated acres that can be allocated to a watershed
for each crop.
Table 8. Average Annual Acres Grown and Percent Crop Treated in Ohio.
Crop or Use
Ohio (HUC-2 Region 05)
Average Annual Acres
Grown
Maximum PCT
Maximum BAT
Alfalfa
264,000
7
18,480
Apples
4,709
43
2,024
Corn, field
3,710,002
1
37,100
Corn, sweet
16,020
48
7,690
Soybean
2,769,996
2
55,400
To account for the uncertainty in the location of the treated acreage throughout the state, this
case study considers two of the three treated acreage allocation methods described in the
White Paper for PCA and PCT Refinements: upper and uniform distribution. These examples will
illustrate how the acreage was allocated to the representative example watershed, described in
the following sections. However, this methodology can be applied to all watersheds with EDWC
that remain above the DWLOC at any point after the Tier 3 distributional PCA approach has
been utilized. Similar to the aggregation of PCA adjusted EDWC in previous sections, the
consideration of PCT methods in regional and national DWA will require automated processes
to efficiently implement. In this case study the consideration of PCT methods has been done
manually.
The sample watershed is a 310,245-acre watershed located within HUC-2 Region 05. The
watershed-scale PCA values and cropped area for the target crops of pestl are shown in Table
9. The PCA value for orchards in the example watershed is zero, therefore orchards will not
contribute to the EDWCs in the watershed and are excluded from the subsequent analysis. The
aggregate PCA-refined EDWC of pestl in the example watershed is 5.26 |ag/L, which exceeds
the DWLOC of 5 jig/L
Table 9. Example Watershed Characteristics
Watershed area
(A)
Watershed PCA Value
Cropped Area Within Watershed (A)
Alfalfa
Corn
Soybeans
Orchard
Alfalfa
Corn
Soybeans
Orchard
310,245
0.045
0.27
0.22
0.0
13,961
84,061
67,879
0
21
-------
The upper distribution assumes that all the treated acres for each individual crop are allocated
to the example watershed, up to a maximum of the PCA adjusted area of the watershed for
each crop. Therefore, the treated acreage within the watershed is the lower value of either the
state-level treated acreage or the cropped area within the watershed (Table 10). In the
example watershed, the maximum PCT state-level treated acreage was greater than the
cropped area for alfalfa and thus no adjustment to the EDWC is warranted for alfalfa.
Conversely, for corn and soybeans, the max PCT state-level treated acreage is less than the
number of acres grown in the watershed, therefore the PCT refinement will reduce the EDWCs
for both crops. If the cropped area of the watershed was less than PCT treated acreage for all
crops, then the PCT refinement would not alter the EDWCs and further refinement would need
to be considered.
Table 10. Comparison of State-Level Treated Acreage and Watershed Cropped Area to
Determine Upper Distribution Treated Acreage	
Crop or Use
Ohio (HUC-2 Region 05)
Maximum PCT
Treated Acreage
Cropped Area of
Watershed
(A)
Upper Distribution Treated
Acreage
Alfalfa
18,480
13,961
13,961
Corn, (field+sweet)
44,790
84,061
44,790
Soybean
55,400
67,879
55,400
The uniform distribution assumes the treated acres are distributed evenly through the state
where the crop and watershed are located. For each crop, the treated area within a watershed=
(acres within the watershed that overlaps with the land cover class)*(state-level PCT), up to a
maximum of the total treated acres within the state. The acres within a watershed that overlap
with the land cover class = (watershed area)*(land cover class PCA). Combining the two
equations gives the formula for calculating the uniform distribution treated acreage within a
watershed (Equation 3).
Equation 3
Treated Area
= (watershed area)x(land cover class PCA) x (land cover class PCT)
The PCA, PCT and uniform distribution treated acreage for each relevant landcover are given in
Table 11. The calculated uniform distribution treated acreages are 977 A, 1005 A, and 1365 A
for alfalfa, corn, and soybeans, respectively. These values are all less than the total treated
acres within the state from Table 10, thus, represent the total treated acreage within the
example watershed based on the uniform distribution method.
22
-------
Table 11 Uniform Distribution Treated Acreage Values
or the Example Watershed
Crop or Use
Watershed Area
(A)
Land Cover Class
PCA
State-level Land
Cover Class
Maximum PCT
Uniform
Distribution Treated
Acreage
Alfalfa
310,245
0.045
0.07
977
Corn,
(field+sweet)
0.27
0.012
1005
Soybean
0.22
0.02
1365
The Treated Area Scaling Percentage (TASP) is the multiplicative scaling factor for each crop
that is used to refine the EDWC values based on the treated area of the watershed. It is equal to
the (watershed treated area)/(total watershed area). Separate TASP values are calculated for
every combination of watershed, PCT value, and treated acreage distribution method. The TASP
values for the maximum and uniform treated acreage distribution methods are shown in Table
12. For the upper distribution of treated acreage in the example watershed, the TASP for alfalfa
is equal to the alfalfa PCA because the total treated acreage allocated to the watershed was
equal to the total area of the land cover class in the watershed (i.e., 100% PCT). The other TASP
values are all lower than the PCA values for their respective crops, indicating that the PCT
refined EDWC values will be lower than the PCA refined values. The uniform distribution TASP
values are 1-2 orders of magnitude lower than the upper distribution.
Table 12. Calculated TASP Values for the Example Watershed Based on the Maximum PCT and
Upper and Uniform Treated Acreage Distributions	
Crop or Use
Watershed Area
(A)
Land Cover
Class PCA
Upper Distribution
Method
Uniform Distribution
Method
T reated
Area
(A)
TASP
Treated
Area
(A)
TASP
Alfalfa
310,245
0.045
13,961
0.045
977
0.0032
Corn,
(field+sweet)
0.27
44,790
0.14
1005
0.0032
Soybean
0.22
55,400
0.18
1365
0.0044
Treated Area Scaling Percentage (TASP)
The refined EDWC values for each land cover class are calculated by multiplying the unmodified
EDWCs and the TASP for each PCT/treated acreage distribution combination. The aggregate
TASP-adjusted EDWCs are the sum of the refined individual crop EDWCs (Table 13 and Table
14). The TASP-refined EDWCs for the watershed are 3.39 and 0.12 |ag/L for the upper and
uniform treated acreage distributions, respectively. The refined EDWC using both the upper
and uniform distribution methods are below the DWLOC of 5 |ag/L, therefore, no further
refinements are necessary for the watershed.
23
-------
Table 13. Summary of Estimated Upper Distribution TASP-Refined Pestl Concentrations in the
Example Watershed
Use
Estimated 21-day
Average Concentration
Upper Distribution
TASP
Refined TASP Adjusted
Individual
Concentrations
Aggregated Estimated
Concentration
Alfalfa
17.4
0.045
0.78

Corn
12.8
0.14
1.79
3.39
Soybean
4.5
0.18
0.81

Treated Area Scaling Percentage (TASP)
Table 14. Summary of Estimated Uniform Distribution TASP-Refined Pestl Concentrations in
the Example Watershed	
Use
Estimated 21-day
Average Concentration
Uniform
Distribution TASP
Refined TASP Adjusted
Individual
Concentrations
Aggregated Estimated
Concentration
Alfalfa
17.4
0.0032
0.056
0.12
Corn
12.8
0.0032
0.041
Soybean
4.5
0.0044
0.020
Treated Area Scaling Percentage (TASP)
TASP-refined EDWCs were calculated for the full set of CWS watersheds in HUC-2 Region 05 for
the maximum PCT and both the upper and uniform treated acreage distribution methods.
Based on a maximum state-level PCT and the upper treated acreage distributions, pestl EDWCs
are above the DWLOC in 66 of 626 watersheds (Figure 13). This represents a 37% reduction in
the number of watersheds where the EDWC would exceed the DWLOC from the number
focused only on the distributional PCA options.
24
-------
14
12
c 10
o
4-» „
2 8
u 4
2
0
0	0.2	0.4	0.6	0.8	1
Percentile
Individual PCA-PCT Upper-Max 	DWLOC
Figure 13. Distribution of Pestl Concentrations for Aggregated Use Scenarios in Assuming a
Maximum PCT with Maximum Distribution of Treated Crops within Each Individual
Community Water System Watershed
HUC-05




0
— /
§
©
1
o—
Jf















Using a maximum state-level PCT and the uniform treated acreage distributions none of 626
watersheds are expected to have pestl EDWC above the DWLOC (Figure 14). Based on the
maximum PCT/uniform distribution, no further refinements would be needed.
6
5
I 4
4-»
03
£3
Q)
U
c
O 2
u
1
0
0	0.2	0.4	0.6	0.8	1
Percentile
Individual PCA-PCT Uni-Max 	DWLOC
Figure 14. Distribution of Pestl Concentrations for Aggregated Use Scenarios in HUC-05
Assuming a Maximum PCT with a Uniform Distribution
HUC-05
25
-------
iii. Summary of PCT Analysis
After applying PCT refinements, several watersheds in HUC-2 Region 03 still have
concentrations that may be above the DWLOC. However, this conclusion is uncertain due to the
lack of available usage data for non-agricultural use of pestl. Concentrations in roughly 10% of
CWS watersheds in HUC-2 Region 05 exceed the DWLOC when an upper-max PCT analysis is
considered; however, when a uniform max PCT analysis is considered no CWS are predicted to
exceed the DWLOC.
When interpreting application of PCT using the methods describe above consideration must be
given to the strengths, weaknesses and uncertainty associated with the available usage data
and methods to distribute state and national level usage data to watersheds within a state.
Similarly, for watersheds that span multiple states PCT data will be allocated proportionally
from multiple states to the watershed.
For pestl in this case study a clear distinction is seen between the Upper and Uniform
distribution methods for PCT where the difference in EDWC is between 1 and 2 orders of
magnitude. The differences between results in this case study point to a significant source of
uncertainty. While it is unlikely that all pesticide use will occur in all watersheds simultaneously
without specific information on where pest pressure is occurring, it cannot be ruled out that
some concentration of use is possible across all watersheds in a region. The likely scenario is
that some concentration of pesticide use is going to occur but not across all watersheds at the
same time. This points to the potential that some intermediate method of distributing usage
(e.g. only 90% of treated acreage is in a given watershed) may provide a more realistic
representation of where EDWC are likely to occur above the DWLOC.
Ultimately, the decision whether a specific set of CWS watersheds remain above the DWLOC
after consideration of the full distribution of PCA and the matrix of options for considering PCT
will have to weigh the strengths, weaknesses, and uncertainty associated with the underlying
data sets.
5. Conclusion
This case study illustrates the application of full distribution PCA refinements and PCT
refinements to EDWCs for the Tier 3 drinking water assessments of pestl in HUC-2 Regions 03,
04, and 05 consistent with USEPAS OPP's DWA Framework (USEPA, 2019) and the methods
described in the PCA/PCT White Paper.
The scoping exercise for pestl indicated that after completing a Tier 2 DWA analyses, there
were still DWLOC exceedances for several uses for the three regions considered in this case
study. After completing a Tier 3 level of assessment including integration of additional PCA and
PCT refinements the following results were obtained:
• For HUC-2 Region 03, there are roughly 15% of CWS watersheds where concentrations
may be above the DWLOC after considering PCA refinements. There is not enough PCT
data on non-agricultural uses to apply PCT refinements to the pestl EDWC values.
26
-------
However, it was possible to determine what amount of usage would be needed to
exceed the DWLOC. Based on the aggregate EDWC value, non-agricultural PCT value of
<9% would lead to pestl EDWC values less than the DWLOC. Additional non-agricultural
usage data would help further refine the pestl EDWCs in the region.
•	For HUC-2 Region 04, there are no CWS where concentrations are expected to be above
the DWLOC after considering PCA refinements along with cropland overlap analysis and
development of aggregated EDWCs (no PCT analysis was needed).
•	For HUC-2 Region 05, roughly 10% of CWS have EDWC values that exceed the DWLOC
after considering PCA refinements and where using the maximum PCT with maximum
distribution of treated acres. Based on the maximum PCT with a uniform distribution of
treated acres there were no CWS watersheds where the EDWC values exceeded the
DWLOC.
The approaches outlined above represent a high level of refinement consistent with Tier 3 of
USEPA's DWA Framework (USEPA, 2019). This case study provides representative examples of
how a step wise approach to successive refinements using the full suite of CWS watershed PCA
and state level PCT data (national level for non-ag uses) can be utilized to focus DWA
refinements on groups of CWS watersheds most likely to be of concern. For those locations
where there are differences in exceedances of the DWLOC depending on the state level PCT
value (e.g., maximum vs average) chosen and the PCT distribution method (e.g., upper vs.
uniform) a Weight of Evidence (WoE) approach that considers the strengths, weaknesses and
uncertainty in the usage data should be considered for determining further refinement options
and/or mitigation options. Subsequent refinements at Tier 4 can be conducted consistent with
the DWA Framework and should focus on factors relevant to the groups of CWS watersheds
that continue to have EDWC greater than the DWLOC after consideration of the methods
presented in the PCA/PCT White Paper illustrated in this case study.
6. References
USEPA. 2014 Development of Community Water System Drinking Water Intake Percent Cropped
Area Adjustment Factors for use in Drinking Water Exposure Assessments: 2014 Update.
9/9/14. Environmental Fate and Effects Division. Office of Chemical Safety and Pollution
Prevention. U.S. Environmental Protection Agency. Available at
https://www.epa.gov/pesticide-science-and-assessing-pesticide-risks/development-
community-water-system-drin king-water.
USEPA, 2019. Draft Framework for Conducting Pesticide Drinking Water Assessments for
Surface Water. Office of Pesticides Programs, U.S. Environmental Protection Agency.
August 22, 2019. Available at htt ps://www. regulations.gov/docu me nt?D=EPA-HQ-OPP-
2019-0417-0006
27
-------