EPA-540-R-03-008
OSWER # 9285.7-76
ASSESSING INTERMITTENT OR VARIABLE
EXPOSURES AT LEAD SITES
Office of Solid Waste and Emergency Response
U.S. Environmental Protection Agency
Washington, DC 20460
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
TECHNICAL REVIEW WORKGROUP FOR LEAD
The Technical Review Workgroup for Lead (TRW) is an interoffice workgroup convened by the
U.S. EPA Office of Solid Waste and Emergency Response/Office of Superfund Remediation and
Technology Innovation (OSWER/OSRTI).
CO-CHAIRPERSONS
Region 8 NCEA/Washington
Jim Luey Paul White
Denver, CO
MEMBERS
Region 1 NCEA/Washington
Mary Ballew Karen Hogan
Boston, MA
NCEA/Cincinnati
Region 2 Harlal Choudhury
Mark Maddaloni
New York, NY NCEA/Research Triangle Park
Robert Elias
Region 4
Kevin Koporec OSRTI Mentor
Atlanta, GA Larry Zaragoza
Office of Superfund Remediation and Technology
Region 5 Innovation
Patricia VanLeeuwen Washington, DC
Chicago, IL
Executive Director
Region* Richard Troast
Ghassan Khoury office Qf Superfimd Remediation and Technology
Dallas> T* Innovation
Washington, DC
Region 7
Michael Beringer
Kansas City, KS Associate
Scott Everett
R . 1ft Department of Environmental Quality
Mafc Stifelman Salt Lake City, UT
Seattle, WA
-------
TABLE OF CONTENTS
1.0 SUMMARY 5
2.0 INTRODUCTION 6
2.1 APPROPRIATE USES OF THE TIME-WEIGHTED APPROACH 9
3.0 TIME WEIGHTING EXPOSURE 10
3.1 SIMPLE TIME WEIGHTING 11
3.2 VARYING INTENSITY OF EXPOSURE 12
3.3 MATRIX APPROACH FOR EVALUATING EXPOSURE ASSUMPTIONS 15
3.4 CALCULATING A PRELIMINARY REMEDIATION GOAL (PRG) 16
4.0 APPLICATIONS OF THE APPROACH WITH THE IEUBK 19
4.1 CONTRIBUTION OF TRACKED-IN SOIL 20
4.2 APPLICABILITY OF THE APPROACH TO VARIOUS AGE GROUPS 21
4.2.1 CHILDREN 0-84 MONTHS 21
4.2.2 ADOLESCENTS 22
4.2.3 ADULTS 23
4.3 SEASONAL VARIABILITY IN LEAD EXPOSURE AND PsB 23
4.4 INCREASED SOIL INGESTION (INCREMENTAL APPROACH) 24
5.0 AIR PATHWAY 26
6.0 UNCERTAINTIES IN THE APPROACH 27
6.1 UNCERTAINTIES IN ASSUMPTIONS REGARDING SOIL INTAKE AND ABSORPTION .... 28
6.2 UNDERESTIMATION OF PEAK PaB 28
6.3 UNCERTAINTY IN HEALTH EFFECTS FROM ACUTE, HIGH-LEVEL EXPOSURES 30
6.4 SEASONAL vs. ANNUAL EXPOSURE 30
APPENDIX A 31
EXAMPLE 1: RECREATION EXPOSURE SCENARIO FOR PARK 32
EXAMPLE 2: RECREATIONAL & TRESPASSING EXPOSURE SCENARIO 34
EXAMPLE 3: DAYCARE EXPOSURE SCENARIO FOR INDUSTRIAL PARK DAYCARE 37
EXAMPLE 4: INTERMITTENT NON-RESIDENTIAL EXPOSURE FOR ADULTS 39
EXAMPLE 5: SEASONAL EXPOSURE: CHILDREN VISITING A SITE 4 MONTHS OF THE YEAR
41
EXAMPLE 6: LAWN MAINTENANCE NEAR A RIVER 43
REFERENCES 46
-------
1.0 SUMMARY
The methodology presented in this guidance may be appropriate for the assessment of
lead risks when exposures are not continuous and chronic, such as:
• Exposure at secondary locations where media concentrations differ from the residential
scenario (e.g., daycare or other caregiver, agricultural work [farming on contaminated
land, as well as farming with agrichemicals that contain lead]). In this context, location
means any area where media concentrations differ.
Exposure at neighboring parks or play areas.
• Less frequent exposure connected with recreational or other site visits (more relevant, but
perhaps not limited, to older children). These may include:
• Recreational activity on and around lead sources: contaminated sediments, waste
piles, etc.
• Visiting and trespassing scenarios on contaminated site properties.
This methodology is not intended to replace the approaches recommended for assessing
standard residential or continuous non-residential exposure scenarios, which are the most
common applications for the Integrated Exposure Uptake Biokinetic (IEUBK) model (U.S. EPA,
1994) and the Adult Lead Methodology (AIM) (U.S. EPA, 1996b), respectively. Instead, this
methodology is intended to be used when certain criteria are satisfied. These criteria are
illustrated in Figure 1. Because this approach is supplemental to the typical residential approach
for the IEUBK model (children 0-84 months) and non-residential approach for the ALM (adults),
users are cautioned that the discussion herein assumes familiarity with the IEUBK model and the
ALM and guidance.
This document presents general guidance for many typical scenarios involving
intermittent, non-residential exposure to lead sites. This guidance is not a regulation itself, nor
does not it change or substitute for any regulations. Thus, it does not impose legally binding
requirements on EPA, States, or the regulated community. This guidance does not confer legal
rights or impose legal obligations upon any member of the public. Interested parties are free to
raise questions and objections about the substance of this guidance and the appropriateness of the
application of this guidance to a particular situation. EPA and other decision makers retain the
discretion to adopt approaches on a case-by-case basis that differ from those described in this
-------
guidance. Additional complexity may be appropriate for certain sites. Contact the Technical
Review Workgroup (TRW) for more information
nittp:/Avww.epa.gov/superfund/programs/leadX
-------
2.0 INTRODUCTION
EPA's lead models simulate soil lead exposures at a single location of concern (e.g., the
residence for the IEUBK model, and a single non-residential location for the AIM). This
guidance addresses how to use the IEUBK model and ALM to assess a wider variety of exposure
scenarios, including exposure from more than one location, varying intensities of exposure,
track-in of soil from another location, and intermittent air exposures). This document describes
the methods, assumptions, limitations, and uncertainties associated with time weighting of
exposures to account for intermittent or highly varying exposure levels, and several examples of
how the methodology can be applied at sites. This guidance accounts for cumulative exposures
when contact with lead-contaminated media at a second defined source in the community is
likely (in addition to exposures to contaminated media at residences). For children or youths,
secondary lead-contaminated locations can include playgrounds, recreational areas, daycare
centers at industrial areas, or traversing contaminated sites on the way to school or play. For
adults, secondary locations can include repeated exposure to work areas with different levels of
lead contamination, or exposure to contaminated recreational areas.
The time-weighting approach, described in this report, can be applied to the IEUBK
model or the ALM. Because children are the most sensitive receptors, this guidance recommends
the IEUBK model be used when exposures occur both at the primary residence and at a second
location accessible to young children. Exposure to soil at the secondary location will result in an
increase in blood lead (PbB) concentration above the "baseline" PbB concentration attributed to
the residential sources of lead, if the exposure level or soil ingestion rate at the secondary
location is higher than that at the residence. The magnitude and duration of the increase in PbB
concentration will vary depending on the temporal pattern of exposure at the secondary location.
The increase will be greatest if exposure at the secondary location occurs every day in succession
over an extended period (e.g., over the summer); in comparison, intermittent exposures at the
secondary locations (e.g., once every 7 days) would give rise to a smaller PbB increase. The
TRW has recommended that the IEUBK model and the ALM be applied to exposures that exceed
a minimum frequency of one day per week and duration of 3 consecutive months (U.S. EPA,
-------
1994). Three months is considered to be the minimum exposure to produce a quasi-steady-state
PbB concentration. The reliability of the models for predicting PbB concentrations for exposure
durations shorter than 3 months has not been assessed.
The approaches described herein are consistent with the conceptual structure of the
IEUBK model. The IEUBK model (White et al., 1998) was validated using central tendency
exposure assumptions to predict a geometric mean blood lead (GM PbB) concentration (Hogan et
al., 1998). The following approaches (i.e., time-weighted averaging and incremental approach)
tend to be inherently protective to the extent that observed PbB concentrations integrate all
exposures: residential and secondary locations.
Several criteria should be satisfied when considering the assessment of cumulative risks
from exposures at a primary residence and at a secondary location using the time-weighted
approach. A decision tree (Figure 1) is provided to determine whether this approach is suitable to
your site. If suitable, this methodology may be used in conjunction with the IEUBK model and
the ALM to assess a variety of scenarios where activities may result in additional exposure to
contaminated media. Further characterization of plausible site-specific exposure patterns (e.g.,
exposure duration and likely subpopulations at risk) are also described herein. The TRW
recommends considering several possible alternative scenarios to characterize worst case or
upper bound estimates, as well as central tendency risk estimates. Additional complexity can be
added to site-exposure scenarios if needed; contact the TRW for more information.
-------
Minimum exposure frequency and
duration: 1 day per week for 34 months
I
The non-residential site of concern has a
soil lead concentration >400 mg/kg
Child exposed is 6-84
months in age.
(Section 4.2)
Adult Exposure
(Section 4.2)
Does the residential scenario
adequately cower all exposure
scenarios?
Does one non-residential
scenario adequately cover all
exposure scenarios?
Yes
No
No
Yes
\BJBK Model
Simple TJrnB-weighted Average
Approach
1
ALM
Figure 1. Decision tree for determining the appropriate approach to assess cumulative lead risk
from one or more locations.
-------
2.1 Appropriate Uses of the Time-Weighted Approach
The TRW does not recommend time weighting unless the criteria shown in Figure 1 are
satisfied. If the planned site use is residential, then it is generally not necessary to assess
additional exposures as long as the other areas are not contaminated. For most residential
exposure scenarios, the IEUBK model can be used with the residence as the only source of
exposure. Similarly, because the default ALM is based on assessment of non-residential exposure
and includes a baseline for residential exposures, most applications will not require time
weighting to assess exposure to assess a non-residential exposure in combination with the
residence. Time weighing approaches should only be used in the ALM to assess exposures to two
or more non-residential locations.
The IEUBK model and the ALM were designed to simulate PbB concentrations
associated with exposures of sufficient duration to result in a quasi-steady state (U.S. EPA, 1994,
1996b). The TRW has recommended 3 months as the minimum duration of exposure that is
appropriate for modeling exposures that occur no less often than once every 7 days (U.S. EPA,
1994, 1996b, 1999a). The reliability of the models for predicting PbB concentrations in children
exposed to lead for durations shorter than 3 months has not been assessed. Because the IEUBK
model assumes constant exposures during each age-year, it can provide only an approximation of
quasi-steady-state PbB concentrations during non-continuous exposure scenarios of less than a
year. As a result of this limitation, short-term fluctuations in PbB concentrations that might occur
in response to intermittent exposures cannot be explicitly represented in the model and may be
underestimated if short-term exposures are time averaged over the entire year. For public health
purposes, it would be reasonable to consider the possibility of adverse health effects from acutely
elevated PbB concentrations that could occur over a period of a few months. Therefore, it is
generally recommended that time-weighted exposure inputs for the IEUBK model and ALM not
be annualized and instead, be calculated only for the duration of the shorter-term exposure (with
the realization that the IEUBK model will treat such weighted values as applying to a full year
exposure duration). For example, for an intermittent exposure that occurs each year over a period
10
-------
of 3 months, the time-weighted exposure inputs would not be further adjusted to account for the
exposure period of 90 days per year (see Example 6 in Appendix).
Accordingly, the predicted quasi-steady-state PbB concentrations corresponding to site
exposure (not annualized across 1 year) will tend to be higher than the annual average PbB
concentrations corresponding to more limited exposure durations. The TRW recommends either
the IEUBK model or the ALM for assessing risks associated with short-term exposures of 3
months or longer in duration for the following reasons:
• An extensive body of research has demonstrated an association between chronic health
effects of lead and elevated steady-state PbB levels. Currently, the health effects (acute or
chronic) of peak PbB levels that occur after acute exposures are not well understood.
• Pharmacokinetic studies of humans (adults) exposed to lead in the diet and of swine and
other animals exposed to lead in soil indicate that PbB concentrations will achieve a
pseudo-steady state within 1 to 3 months of repeated daily exposure (U.S. EPA, 1994).
• Evaluating the exposure over a 3-month or longer time period is consistent with the time
frame for a time-critical removal action, which is typically defined as a few weeks to 6
months.
When using the IEUBK model to evaluate short-term continuous exposure of no less than
3 months, it is recommended that:
• When data for individual children or populations at the site are not available, default
inputs to the model should be used rather than maximum values (e.g., use the default soil
intake rather than estimates of "high normal" soil intakes or estimates of pica). The
IEUBK model is intended to provide a plausible distribution of PbB levels that may be
expected to occur at a site based on site-specific exposure inputs to the model.
The model should only be used to predict the quasi-steady-state PbB concentration that
will be achieved within approximately 3 months of exposure to a given level of lead. It
has not been validated for predicting the rate at which the PbB concentration will
decrease after exposure sources are removed or reduced, or how long it will take to reach
a new quasi-steady state.
3.0 TIME WEIGHTING EXPOSURE
The input menus of the IEUBK model and the ALM are somewhat limited for scenarios
in which exposures to soil lead from multiple sites occur. The IEUBK input menus for "school",
"dust", and "other" assume that exposure to soil from these sources is continuous, and do not
11
-------
permit the user to assess intermittent exposures from multiple locations. The TRW recommends
that, for reasons of feasibility and maximum clarity, separate calculations be made up front (i.e.,
outside the model) to obtain appropriately weighted average concentrations of soil lead. These
average values can then be entered directly into the model as fixed media concentrations.
Calculation of the time-averaged values are described in the next sections.
3.1 Simple Time Weighting
There are no "default" recommendations for the relative weights to be used in calculating
time-weighted media concentrations; rather, the assumptions should be stated clearly and reflect
plausible estimates of the typical exposure scenarios. The TRW recommends time-weighted
exposure calculations be applied to derive an average value for the two (or more) locations. In
this approach, a weighted value is assigned to a medium (e.g., soil) that reflects the fraction of
outdoor exposure to residential or site soil. The soil concentrations are weighted based on the
estimated fraction of total soil ingestion that occurs at the residence and at the site. Equation 1
shows the fundamental equation for time-weighting exposures to soil from the residence and a
secondary location.
n
Weighted PbCmedlum = E Ci-EFi Equation 1
<=i
where:
Weighted PbCm(.dium = Weighted lead concentration in medium (ppm).
Cj = Media concentration at location i (in this case, /= residential yard or secondary
location) (e.g., ppm).
EFj = Exposure frequency at location i (in this case, /= residential yard or secondary
location) (e.g., days/week). The sum of the days/week spent at the residential yard
and secondary location is 7.
The time-weighting factor should be based on the smallest time period in which the
exposures repeat (the exposure event period). For example, in an exposure scenario in which one
expects exposures 3 days per week for 210 days, the exposure event period is 7 days since 3
exposure events occur every 7 days; therefore, the time weighting should be 3 days/7 days NOT
90 days/210 days (i.e., 3 days/week x 4.3 weeks/month x 7 months) (see Example 6 in
12
-------
Appendix). Although the differences in predicted PbB, in this case, are small (<10%), larger
differences could arise in more complex time-weighting adjustments that are based on the typical
calendar units of time, rather than in units that best reflect the exposure event period. As a rule
of thumb, the latter will be achieved if the time-weighting factor is the number of exposure
events within the event period cycle:
TWA = C* N + P Equation 2
p
where:
C = Exposure concentration.
Np = Number of exposure events within the exposure event period.
P = Event period.
While Equation 1 may be appropriate for time-weighting exposure media at some sites,
the intensity and time of contact with contaminated media may vary with the type of activity for
the different locations. This situation is discussed in the following section.
3.2 Varying Intensity of Exposure
The TRW expects that soil ingestion rates will generally tend to be higher for time spent
outdoors in comparison with time spent indoors. For home daycare scenarios, however, the TRW
does not generally support the use of different ingestion rates for children's activities at daycare
as compared to activities at home, since indoor and outdoor play activities will occur at both
locations, and there will be a comparable mix of other activities such as meals and "quiet times"
at both locations. For alternate residence or daycare scenarios, the TRW recommends that
exposure be apportioned according to waking hours to derive weighted estimates for media
concentration. For example, in a 24-hour period, a child's activities might include 12 hours of
sleep during the night, and an 8- to 10-hour stay at a daycare facility, with the remaining hours
spent awake at home. In this example, roughly 45% (8 to 9 hours per day x 5 days/week = 40 to
45 hours/week at daycare and 12 hours/day x 7 days per week = 84 hours/week) of the child's
total waking hours may be spent at daycare. This time-estimate of 45% of a child's waking hours
at daycare as compared to waking hours spent at home would then be used to derive a weighted
13
-------
soil lead concentration based on the child's exposure to daycare soils and home soils. The
fraction of soil dust exposure at each location may be calculated as shown in Equation 3 .
f = Wakhig hours at location Equation 3
' Total waking hows
where:
fj = Fraction of total dust and soil exposure that occurs at location/.
Barring additional site-specific considerations, indoor and outdoor soil ingestion would
still be applied according to the IEUBK model default Soil/Dust Ingestion Weighting Factor of
45% soil and 55% dust (U.S. EPA, 1994). To derive an indoor dust concentration when only
outdoor soil data are available for multiple exposure locations, apply the soil-to-dust mass
transfer parameter (M^ to weighted outdoor soil lead concentration. Consideration should be
given to whether indoor dust sources (particularly lead-based paint) are likely to be present (see
U.S. EPA, 1994, 1999b). Equations 4, 5, and 6 show how fractional exposure at each location
can be used to derive time-weighted estimates for soil and dust
The TRW recommends that dust samples be collected at both locations for such
calculations. If this is not possible (e.g., if the assessment is for a proposed future use), an
estimate of the composite residential indoor soil-derived dust (PbD) concentration may be
derived using the default soil-to-dust mass transfer parameter if it is reasonable to assume that
the site-specific conditions permit using the default mass transfer rate (MS(1) to apply to the
situation. In the absence of further information upon which to evaluate the site-specific mass
transfer of soil into dust, the TRW recommends using the default M^ value of 0.70 to estimate
PbD levels for this application (U.S. EPA, 1998). This is reasonable if soil lead is the major
source of indoor PbD and no enrichment of indoor dust is expected (such as by lead-based paint)
(U.S. EPA, 1994, 1998).
14
-------
The weighted medium concentration is the sum of the fractional concentrations:
n
Weighted PbCmetltim = E Ct-ft Equation 4
1=1
where:
Weighted PbCmcdium = Weighted lead concentration across all exposure locations (ppm).
C, = Lead concentration for the medium at each location (ppm).
f, = Fraction of time spent at each location (hours/day or days/week).
Example for weighted soil concentration from home and daycare:
PbSw = (PbS, x / ) t (PbS x
where:
PbSw = Weighted soil lead concentration across all exposure locations (ppm).
PbSj = Soil lead concentration for each location (i = home; j = daycare) (ppm).
f; = Fraction of time spent at each location (i =home; j = daycare) (hours/day or
hours/week).
Example for weighted dust concentration from home and daycare using multi-source analysis:
PbQw = (J*S x M+tS x
where:
PbDw = Weighted dust lead concentration (ppm).
PbSwdaycarc = Weighted soil lead concentration from daycare (ppm).
Hsddaycarc = Soil to dust mass transfer parameter (unitless).
PbSWhomc = Weighted soil lead concentration from residence (ppm).
Msdhomc = Soil to dust mass transfer parameter from residence (unitless).
These time-weighted estimates for PbSw and PbDw can be entered directly into the
IEUBK model soil and dust media concentration parameter data windows to calculate risk.
Similarly, weighted soil estimates for two or more non-residential locations can be entered into
the ALM spreadsheet.
15
-------
Note that this approach does not require separate estimates of the amount of time spent
outdoors and indoors at both locations. A more elaborate analysis could be constructed that
attempts to apportion children's time spent outdoors and indoors among multiple sites (e.g.,
daycare and their homes); however, the TRW believes that any plausible estimates based on this
approach would depend upon considerable data on the children's specific activity patterns at each
site.
This approach is not appropriate for scenarios involving outdoor areas where ingestion is
expected to be higher than IEUBK and ALM default values, due to increased soil contact or
adhesion (e.g., lake or beach). Such scenarios are discussed m Increased Contact with Soil
(Incremental Approach) (Section 4.4).
3.3 Matrix Approach for Evaluating Exposure Assumptions
The TRW suggests that a matrix approach be used for evaluating different exposure
assumptions. For example, different alternatives may be plausible within a range of waking hours
spent at each location. The matrix approach permits an evaluation of how activity patterns, using
proposed soil cleanup levels, impact estimated risks of elevated PbB and proposed cleanup goals.
By conducting several model runs using the alternate values, the implications of alternative
assumptions can be evaluated. This approach can also be useful when presenting options to risk
managers in cases where there are no data to suggest that one exposure scenario is more plausible
than another. An example of the matrix approach is provided in Table 1.
16
-------
Table 1. Matrix Showing Impact of Various Exposure Assumptions (Number of Site Visits per Week) on Model
Predicted Blood Lead Concentration (PbB) and Probability of Exceeding 10 Hg/dL (Pip).
Exposure scenario
Zero visits per week (residential only scenario)
1 site visit per week
2 site visits per week
3 site visits per week
4 site visits per week
5 site visits per week
6 site visits per week
7 site visits per week
PbSw'
100
171
242
314
386
458
528
600
PbDw2
70
120
169
220
270
321
370
420
GM PbB (ng/dL)3
2.5
3.1
3.7
4.3
4.9
5.5
6.1
6.6
P,o(%)3
0.1
0.6
1.8
3.8
6.6.
10.2
14.3
18.9
1 PbSw= weighted soil lead concentration; calculated using Equation 1 . Residential PbS=l 00 ppm; site PbS=600 ppm.
2 PbDw= weighted dust lead concentration; calculated using Equation 5.
3 Results from 1EUBK. model. GM= geometric mean. All other parameters were set to 1EUBK default values. All runs using
0-84 months in 1EUBK model. (Exposure continued throughout the 84-month period.)
The example in Table 1 shows that 3-4 site visits per week result in PbB concentrations
and P10 values near the EPA goal. The next section explains how the time-weighted approach can
be used to develop a preliminary remediation goal (PRO) for the risk assessment.
3.4 Calculating a Preliminary Remediation Goal (PRG)
The current version of the IEUBK model does not automatically back-calculate
environmental lead levels. Risk-based target soil concentrations should be determined through
several runs of the model by varying the media concentrations until the appropriate risk level is
reached (the iterative approach). However, a risk-based target site concentration can be back-
calculated from a site-specific model estimate of the overall soil lead concentration associated
with a 5% individual risk of elevated PbB (Equation 7). An alternate approach would be to assess
the secondary location alone using a continuous exposure scenario. In cases where the residence
(IEUBK) or primary non-residential site (ALM) are less contaminated than the intermittent
exposure site, assessing the secondary location as a continuous exposure is expected to be more
conservative than time weighting the exposures.
17
-------
Equations 7 and 8 illustrate how the weighted values can be used to derive a cleanup goal
based on average soil lead concentration at the site. The time-weighting equation (Equation 1)
can be expanded to explore site soil concentrations that are risk protective:
w = EFiiteX [(file* PbSsiu')* (fyardX PbSyard)]* (EFyard X PbSyanl) Equation 7
where:
PbSsitc = Average soil lead concentration at an exposure unit on the site (ppm).
PbSw = Weighted soil lead concentration (ppm).
PbSyard = Average soil lead concentration near home (ppm).
fyard = Fraction of daily outdoor time at local background soil lead concentration (usually near
home) = l-fsitc (dimensionless).
EFsitc = Exposure frequency expressed as fraction of the days/week child visits the secondary
location during the exposure period.
EFyard = Exposure frequency expressed as fraction of the days/week child does not visit the
secondary location during the exposure period = l-EFsitc.
fsitc = Fraction of daily outdoor time spent at the secondary location on days when the site is
visited (dimensionless).
Equation 7 may be rearranged to solve for average soil lead concentration at the site. Starting
with the Office of Solid Waste and Emergency Response (OSWER) soil lead screening
concentration of 400 ppm (associated with a 5% individual risk of elevated PbB when the model
defaults are relevant) for PbSw, a soil lead concentration for the site can be derived that is
protective of human health:
pbs . w Equations
It is important to recognize that apportioning cleanup across two or more locations is a
risk management decision and not solely a risk assessment decision. Consequently, the TRW
suggests that the matrix approach be used to present the range of cleanup options that are health
18
-------
protective. Table 2 illustrates how various cleanup options and exposure assumptions may be
presented to risk managers.
Table 2. Example of Various Exposure Assumptions and Associated Risk Estimates for Presentation to Risk
Managers.
Options
1: SitePbS=700
Residence PbS=100
2: SitePbS=400
Residence PbS=100
3: SitePbS=300
Residence PbS=100
Exposure
assumption
3 site visits
per week
5 site visits
per week
6 site visits
per week
PbSw'
357
315
271
PbDw2
250
220
189
GM PbB3 (jig/dL)
4.7
4.3
4.0
P,o(%)3
5.4
3.8
2.4
Site PRG
(ppm)4
676
445
388
1 PbSw^ weighted soil lead concentration; calculated using Equation 1 .
2 PbD^weighted dust lead concentration; calculated using Equation 5.
3 Results from 1EUBK model. GM=geometric mean. All other parameters were set to 1EUBK default values. All
runs using 0-84 months in 1EUBK model. (Exposure continued throughout the 84-month period.)
4 Preliminary remediation goal (PRG) for the site. Note that PRGs are typically rounded to the nearest 50 or 100
ppm. The residential soil lead concentration of 100 ppm was unchanged.
A matrix approach is a useful way to demonstrate to risk managers the health
protectiveness of PRGs under various exposure assumptions (Table 2). The OSWER soil lead
guidance limits the individual risk of elevated PbB for a typical child to less than 5%, which is
not the same as limiting the population risk to less than 5% (see U.S. EPA, 1994).
4.0 APPLICATIONS OF THE APPROACH WITH THE IEUBK
In certain cases, additional assumptions may be appropriate for the risk assessment to
reflect the added contribution of site soils to interior house dust lead when older siblings trespass
on the site and track soils into the home, thus exposing younger siblings (Section 4.1). The
approach can be applied to various age groups, including children, teens, and adults (Section
4.2). Although modeling of seasonal variability in lead exposures is difficult and usually
unnecessary for characterizing maximum seasonal exposures, the temporal pattern of exposure
should be considered in assessments (Section 4.3). In addition, the incremental approach can be
used to assess playground or trespasser scenarios where activities may result in more intense
19
-------
contact with contaminated soils than at home, daycare, or other residential sites, hi such cases, a
higher fraction of ingestion would appropriately be attributed to the site than would be suggested
by a calculation based on just apportionment of total waking hours (Section 4.4).
4.1 Contribution of Tracked-In Soil
Developing a modeling approach for exposures occurring via soil and dust ingestion from
multiple locations also has applications to the trespasser exposure scenario. For instances where
there is a strong possibility of trespassing on the site (generally non-residential), one should
consider the potential for older children tracking site soils into the home, thereby increasing
interior dust lead levels and increasing residential exposure for younger children. The IEUBK
model should only be used to assess risks to children from 0 to 84 months of age. When older
children (>84 months) are expected to be exposed, the ALM should be used with appropriate
consideration given to the inputs (see Section 4.2.2). Pets may also track soil into the house,
which would contribute to dust lead. Contribution of tracked-in, contaminated soil to indoor dust
is expected to affect default dust concentration if no site-specific data are available.
The IEUBK model default assumption for the transfer of residential PbS to PbD was not
developed for a situation where a significant source of lead in soil is distant from the house.
Some track-in from the site is likely, but all other things being equal, track-in may be less than if
the soil source is the residential yard. There would likely be fewer incidents of track-in per day
per person visiting the site in comparison with a residential yard. On the other hand, more intense
or sustained play and sporting activities at the site could result in larger "loading" of soil on the
children (or adults) that could be tracked into the home. Activities at the site, such as organized
sports, could contribute to a greater than usual accumulation of soil to bring back to the
residence. The extent to which this soil is actually transferred into the residence would depend on
a variety of site- and individual-specific factors. For example, soil adhering to outerwear has
20
-------
more time to drop off clothing the more distant the site is from the residence. On the other hand,
if weather conditions are damp, then the maximum mass of soil picked up is more likely to be
tracked back to the residence. For more information on track-in of contaminated outdoor soil, see
Bornschein et al. (1985) and Matte et al. (1991).
Without some actual measurements of house PbD concentrations under these conditions,
estimates of PbD concentrations are uncertain. Given this uncertainty, the TRW recommends that
the fraction of interior dust attributable to the non-residential site should not exceed the fraction
of the trespassing child's total soil exposure thought to come from the site. For example, if the
•
end assumption is that 20% of the trespassing child's combined soil exposure (home + site) is
attributed to soil coming from the site, then it is probably appropriate to assume that no more
than 20% of the interior dust comes from the site.
4.2 Applicability of the Approach to Various Age Groups
4.2.1 Children 0-84 months
Younger children may not be expected to visit a site; however, in some cases, it may be
appropriate to consider exposures to these children to assess increased exposure in these
situations. As described in the preceding section on track-in, children who do not visit the site
can have exposure to soil brought home from the site by older children and adults. Also, cases
have been documented where older children brought younger children to visit areas where adult
supervision would be desired. To assess such scenarios, the IEUBK model (as previously
described) can be run using the increased media concentrations of lead at the residence.
Before choosing a cleanup goal, it is useful to consider both the entire population in
general and the most highly-exposed individual in developing a set of use patterns. In the context
of IEUBK model runs, exposure to lead-contaminated media differs by age. When running the
21
-------
IEUBK model, one can model six exposure patterns, each with a different 1-year age range of
exposure to the composite PbSw, and each having all other age ranges at the default residential
levels. For example, exposure at the site during 1 year can be considered for each of six different
age groups: 12-24, 24-36, 36-48, 48-60, 60-72, and 72-84 months. An example of this approach
is shown in Table 3.
Table 3. IEUBK. Model Risk Estimates for Various Age Groups. Estimated Geometric Mean (GM) PbB and
Probability of PbB >10 Hg/dL (Plo) Based on 1EU BK Model Simulations of Weighted Average Exposures to Site
and Residential Soils'
Age range
(months)
0-84
0-12
12-24
24-36
36-48
48-60
60-72
72-84
1 site visit/week
PbSw2 = 460 ppm
PbDw3 = 322 ppm
GM PbB4
(Hg/dL)
5.5
5.7
6.9
6.5
6.2
5.1
4.4
3.9
PIO
(%)
10.3
11.4
21.5
17.6
15.1
7.8
4.0
2.4
2 site visits/week
PbSw = 9 10 ppm
PbDw= 637 ppm
GM PbB
(Hg/dL)
8.8
8.6
11.0
10.4
10.0
8.3
7.1
6.3
PIO
(%)
39.4
37.8
58.3
53.1
49.8
34.9
23.0
16.1
4 site visits/week
PbSw= 1809 ppm
PbDw= 1267 ppm
GM PbB
(Hg/dL)
14.2
13.3
17.5
16.5
16.1
13.7
11.7
10.4
P,o
(%)
77.0
72.5
88.2
85.8
84.5
74.8
63.2
53.5
1 PbSsitc = 3 159 ppm, PhS,^,,^ = 10 ppm. Estimates apply to quasi-steady-state elevations in PbB
concentrations during the period of exposure (i.e., no "annualization" of PbSw or PbDw).
2 PbSw= weighted soil lead concentration; calculated using Equation 1 .
3 PbD^p weighted dust lead concentration; calculated using Equation 5 .
4 Results from IEUBK. model. GM= geometric mean. All other parameters were set to IEUBK default
values. All runs using 0-84 months in IEUBK model. (Exposure continued throughout the 84-month period.)
Table 3 shows how this approach can illustrate the impact of exposure on a most highly
exposed individual, which is typically the toddler. Depending on the conditions of the site,
bounding the exposures in this way may be appropriate to illustrate the possible risk to the likely
receptor population.
22
-------
4.2.2 Adolescents
In general, the TRW expects that cleanup goals designed to be protective for children less
than 84 months old, the most sensitive subpopulation for chronic health effects, will be at least as
protective for older children. Although the IEUBK model is limited to 0-84 months, the ALM
could be used to assess older children. When using the ALM to assess older children, however, it
may be necessary to adjust default ALM values for ingestion rate and bioavailability (which are
defined for adults) to appropriate values for the exposed population. Users should refer to the
Frequently Asked Questions on the ALM for more discussion on the evaluation of the adolescent
scenario (http://www.epa.gov/superfund/programs/lead/adult.htm). While the IEUBK and the
ALM results could be considered bounds for the risk of elevated PbBs and for cleanup goals for
adolescents with direct exposure to site soil, the toxicokinetics of adolescents are not well
understood, so that any scaling, such as linear interpolation, between the predictions of the two
models cannot be supported. Contact the TRW for guidance concerning use of the ALM in such
instances.
4.2.3 Adults
To estimate PbBs for adult populations exposed to a single non-residential scenario, the
default (i.e., not time weighted) ALM is recommended. As with the IEUBK model, it is
necessary to perform calculations outside the model to derive weighted soil lead concentrations
for use in the ALM if contact with contaminated media occurs at more than one non-residential
location. Note that if the site scenario includes a residence, then the IEUBK model should be
used to assess that location; the ALM should only be used in this context when the two sites are
both non-residential (e.g., adult exposures to contamination in a warehouse and a factory). As
noted earlier, consideration should be given to track-in of soil from these areas to the home.
Example 4 in the appendix to this report provides an example of how the ALM can be used to
assess intermittent exposures.
23
-------
4.3 Seasonal Variability in Lead Exposure and PbB
The IEUBK model was designed to consider routine seasonal variability in media
exposures for children. Although the model was calibrated with environmental data that were
taken to represent sustained daily exposures, the seasonal fluctuation of PbB concentrations are
suspected to represent seasonal variability in both exposure and physiological factors. In some
geographic regions of the U.S., children may have less direct exposure to soil in the colder
months, and their decreased outdoor activity also corresponds to a lower contribution of soil to
indoor dust lead. During the winter months in some regions of the U.S., exposures to exterior soil
may be greatly reduced because the ground is frozen and covered with snow. Interior dust lead
and PbB concentrations were as much as 50% lower in the coldest months in Boston (U.S. EPA,
1995). Nevertheless, exposure to soil may not be negligible during the winter months, occurring
outdoors or from soil tracked into the home.
The calibration and validation data sets that have been used with the IEUBK model were
generated cross-sectionally, including children with at least 3 months of residency at the sampled
locations, at a time of year (late summer) when soil exposure and PbB concentrations were
expected to be at annual maximum (Hogan et ai, 1998). IEUBK predictions are therefore
expected to approximate the PbB concentrations related to the higher lead exposure levels in an
annual cycle of lead exposure, where measured lead concentrations in soil are expected to remain
relatively constant, hi addition, from a public health perspective, it may be more appropriate to
focus on the seasonal maximum exposures than to try to quantify variability in seasonal
exposures. Consequently, users should focus analyses on plausible exposure estimates during
seasons when PbB concentrations are likely to peak.
4.4 Increased Soil Ingestion (Incremental Approach)
Soil ingestion may be greater than default levels in connection with at least some contact-
intensive activities. For example, soil contact and ingestion may be increased at sites where
increased soil adherence would be expected, such as at contaminated waterfront areas or when
dirt biking on contaminated areas. Because of the potential for higher contact rates with soil at
24
-------
the site (e.g., when children dirt bike on a slag pile), scenarios entailing additional soil ingestion
may be warranted. In the absence of site-specific data, risk assessors may want to explore the
impact of a variety of reasonable soil ingestion rates. A recommended approach would be to
bound risk estimates using several reasonable soil ingestion rates. Additional guidance on soil
ingestion is available from U.S. EPA's Exposure Factors Handbook (U.S. EPA, 1997).
A hypothetical example would be to set the default soil ingestion rates as a lower bound.
A reasonable medium exposure scenario of 145% of default rates could also be assumed to occur
at the site (i.e., the default, plus an additional 45% to account for outdoor activities). For a
Reasonable Maximum Exposure (RME) scenario, the 200 mg/day value that has been used in
Superfund assessments as a high average daily soil ingestion rate could be added to the model's
default total dirt ingestion rates.
Just as model limitations require external calculations to achieve the composited PbSw
and PbDw input values for multiple-site scenarios, so must composite ingestion rates be
calculated external to the model when non-residential ingestion rates are reasonably expected to
exceed model default values (Table 4). The methodology suggested herein is a somewhat
simplified and conservative approach, since it may in fact overstate the child's total daily
exposure time (because time spent at the secondary location would be time that is not spent at the
residence).
25
-------
Table 4. Examples of a Range of Hypothetical Dirt (soil/dust) Ingestion Rates Associated with Exposure at Non-
Residential Sites.
Age
group
(months)
0-1 11
12-23
24-35
36-47
48-59
60-71
72-84
Total dirt ingestion rate (g/day)
Low scenario
ingestion = 1EUBK. default
0.085
0.135
0.135
0.135
0.100
0.090
0.085
Medium scenario
ingestion= 1.45 * 1EUBK. default
0.085
0.196
0.195
0.195
0.145
0.131
0.123
High scenario
ingestion = 200 mg/day -t- 1EUBK.
default
0.085
0.335
0.335
0.335
0.300
0.290
0.285
1 Additional soil contact is not applicable to children <1 year, since they are not likely to have significant
additional exposure to site soil.
2 The high exposure scenario is shown for consistency with OSWER guidance on assessing risk under the
Reasonable Maximum Exposure (RME) scenario.
Table 4 provides an example of a plausible range of hypothetical total dirt ingestion rates
associated with greater soil ingestion rates. An example of how the additional soil ingestion rate
values are incorporated in an assessment is shown in Example 2 of the Appendix.
5.0 AIR PATHWAY
In some instances, receptors do not need to be at different locations to have intermittent
exposures. The IEUBK model was designed to predict PbB concentrations associated with
relatively stable, long-term exposures that result in quasi-steady-state PbB concentrations (e.g.,
relatively constant exposures of at least 3 months in duration). The model has not been evaluated
for predicting PbB concentrations that might occur with rapidly varying exposures, such as those
that often result from air emissions from remediation activities at contaminated sites where lead
is a major contaminant. On the other hand, varying air lead exposures of an episodic nature may
be assessed using a time-weighted approach. In this case, continually changing concentrations
that occur each week or within a day could be assessed using the IEUBK model or the ALM. The
26
-------
potential for recontamination of soil and dust by ongoing deposition of airborne lead should also
be considered.
To simulate intermittent exposures, a variation of Equation 1 using air instead of soil can
be used to derive time-averaged air lead concentrations. In the case of air, the extra exposure
would be time averaged and added to exposure to the baseline air concentration which can be set
as IEUBK default (0.1 |ig/m3) or based on upwind site sampling. An example of this approach
for adults (using the ALM) is shown in Equation 9.
DkA pbA** •EF+ pbA*°* • (AT - EF) Equation 9
PbAw = AT
where:
PbAw = weighted air lead concentration ((ig/m3).
PbAsitc = observed or expected air lead concentration from the site (ng/m3).
PbAbasc = baseline air lead exposure concentration that would be expected to occur in the absence
of site exposure (0.1 ng/m3, or based site-specific data).
EF = the exposure frequency (days/week).
AT = the averaging time (days/week).
The intermittent exposure scenario and time-weighting equations are also applicable to
fugitive emission scenarios. At some sites air data may be lacking, yet it would be helpful to take
fugitive emissions from lead-contaminated soil into account. One way to do this is to estimate the
fugitive emissions, then use that equation to modify the cleanup goal for lead. Additional
information concerning the fugitive emission pathway including equations specific to site
activities are available from the Soil Screening Guidance (U.S. EPA, 2001) in which soil- and
site-specific Particulate Emission Factors (PEFs) modify cleanup goals. For example, the
Preliminary Remediation Goal (PRG) for a site can be derived when direct soil contact and
fugitive emissions come from two different sources of lead-contaminated material. Li this
example, the residence is located near a pile of lead-contaminated fines (subject to fugitive dust
contamination) and the child's playground (located in an area apart from the residence) has been
contaminated with fill material containing lead, hi this example, time-weighted averages would
be used in the IEUBK (because the residence is used) to account for soil exposure in combination
with the fugitive dust equation from the Soil Screening Guidance.
27
-------
6.0 UNCERTAINTIES IN THE APPROACH
Various factors could contribute to either an overestimate or an underestimate of the PbB
concentration when weighted exposures to media concentrations are used as inputs to the IEUBK
model and the ALM. These factors need to be considered in interpreting model predictions that
are based on such an approach. Several areas of uncertainty should be considered, including
absorption assumptions (Section 6.1), peak blood lead for successive exposure scenarios (Section
6.2), health effects from acute high-term exposures (Section 6.3), and seasonal versus annual
exposure (washout) (Section 6.4).
6.1 Uncertainties in Assumptions Regarding Soil Intake and Absorption
Estimates for soil and dust ingestion rates used in the model are intended as average daily
rates for typical children. Depending on play routines, sports activities, and soil exposure while at
the site, actual ingestion may exceed typical average values. The TRW recommends that users
consider the potential for alternative higher ingestion rates that may occur during soil contact-
intensive activities, and include risk calculations using these rates in the assessment to bound the
results (see Section 4.4 and table 4). Also, the IEUBK model predicts that the relative absorption
fractions will decrease when higher quantities of lead are ingested. Thus, time averaging may
result in a higher predicted absorption fraction than would be predicted for periods when actual
intakes are higher than the time-weighted average intake.
6.2 Underestimation of Peak PbB for Successive Exposure Scenarios
If exposures to contaminated media from a secondary location were to occur over a
number of days in succession, the cumulative effect would be a temporary elevation of the PbB
concentration during and after this period of exposure. This elevation may be greater in
magnitude (though of shorter duration) than that estimated using a time-weighted average
approach, because the IEUBK model can provide only a quasi-steady-state approximation to PbB
concentrations during non-continuous exposure scenarios (the IEUBK model only allows for
changing exposure variables annually). A hypothetical example of the difference in predicted
PbB when using different approaches to deriving a time-weighted average to annualize an
intermittent exposure is shown in Figure 2.
28
-------
6.3 Uncertainty in Health Effects from Acute, High-Level Exposures
The time-weighted approach assumes that the adverse health effects of lead are related to
long-term average PbB concentrations. While this has been established for chronic effects of
lead, the health effects (acute or chronic) of elevated PbB levels that occur after acute exposures
resulting in short-term PbB concentrations less than 20 ng/dL are not well understood.
6.4 Seasonal vs. Annual Exposure
For seasonal exposures that are restricted to only a fraction of a year (e.g., summer
months), some of the lead burden accumulated during the exposure season will be eliminated
during the intervening months between seasonal exposures. However, the IEUBK model cannot
simulate this loss of lead; model predictions correspond to a full year of exposure to a constant
exposure level regardless of the actual exposure period. For seasonal exposures that occur in
successive years, the TRW recommends that exposures be simulated for individual age-years and
predicted blood lead concentrations for each age-year of exposure be averaged.
For risk assessment purposes, the impact of repeated shorter-term site exposure on an
annual basis is important to consider. This can be approached by first considering the case where
exposure occurs only once and is not repeated annually. Such an exposure estimate would also
characterize children who return to the site for a period each year, and whose added blood lead
burden is eliminated during the intervening months between successive annual exposures.
Illustrations of this point are presented in Examples 5 (one-year exposure) and 6 (multi-year
exposure) of the Appendix. Example 6 shows how different risk management decisions or site-
specific conditions can affect the risk calculation approach.
Both examples highlight the importance of closely examining the exposure assumptions
for the site and how those exposure assumptions are used as parameter estimates for the IEUBK
model and ALM. The TRW recommends running the models with several reasonable sets of
assumptions (plausible combinations) to present a range of possible risks or cleanup options for
the site.
30
-------
APPENDIX A
Case Studies
This appendix provides case studies that illustrate the considerations inherent in assessing
risks posed by lead from a variety of intermittent exposure scenarios using both the ALM and the
IEUBK model. The terms used herein are defined in the body of the report.
31
-------
EXAMPLE 1: RECREATION EXPOSURE SCENARIO FOR PARK
The site uses in this scenario are assumed to be primarily recreation. A proposal is being
considered whereby the site would be developed as recreational area. The goal of the cleanup is
to minimize lead exposure for children who would visit the site during the warmer months for
recreation.
Goals:
1 . Calculate the PbB and P10 risk estimates for children up to age 84 months.
2. Estimate the lead concentration in site soil that would result in a 5% probability of
exceeding a PbB concentration of 10 |ag/dL (i.e., P10
Assumptions for the scenario:
1 . Children have exposure to site soil each day the site is visited, for a total of 52
days spread evenly over 1 year, 6 months, or 3 months (i.e., 1, 2, or 4 days per
week, respectively); exposure during the remaining waking hours of the day is
indoors at the residence.
2. The lead concentrations of site and residential soil are 2000 and 50 ppm,
respectively.
The above assumptions yielded PbSw and PbDw concentrations for 1, 2, or 4 visits/week
to the site (the default M^ of 0.70 to estimate PbD levels for this application). These were used
as inputs to the IEUBK model, along with default values for all other model variables. In
particular, residential dust concentrations were calculated using the weighted mean soil
concentration to which the child was assumed to be exposed and the model default assumptions
for the mass transfer of soil into house dust. In all cases, all other inputs were kept at default
values, including soil ingestion rates. Note that for simplicity, these calculations assume that total
soil ingestion occurs at the default rate. The predicted geometric mean PbB concentrations and
estimates of the probability (%) of exceeding 10 ng/dL (P10) for children 0-84 months old are
shown in Table A- 1 .
32
-------
Table A-l. Matrix Showing Presentation of Various Exposure Assumptions for Evaluation by Risk Managers.
Exposure scenario
Zero visits per week
1 site visit per week
2 site visits per week
4 site visits per week
PbSw' (ppm)
50 _
329
607
1164
PbDw2 (ppm)
35
230
425
815
GM PbB (jig/dL)3
1.9
4.5
6.7
10.5
P,o(%)3
0.03
4.3
19.3
53.8
1 PbSw weighted soil lead concentration; calculated using Equation 1 .
2 PbDw weighted dust lead concentration; calculated using Equation 6.
3 Results from the 1EUBK model. GM = geometric mean. Residential PbS=50 ppm; Site PbS=2000 ppm. All runs
using 0-84 months in 1EUBK model.
As shown in Table A-l, scenarios having site exposures that occur 2 or more times per
week produce risk estimates that exceed the 5% goal. Multiple iterations of the IEUBK model
were run using all model defaults to identify the weighted PbS concentrations corresponding to
the P10 of no more than 5% for children 0-84 months of age (residential PbS was held at 50
ppm). Equation 7 was then used to calculate cleanup goals corresponding to the three use
patterns. These cleanup goals are summarized in Table A-2. For exposure scenarios in which site
visits occurred on 1, 2, or 4 days per week, the risk-based soil goals were 2000, 1050, and 550
ppm, respectively as calculated by Equation 8.
Table A-2. Matrix Showing Possible Site Cleanup Goals Based on Various Exposure Assumptions.
Exposure Scenario
1 site visit per week
PbS site = 2000 ppm
2 site visits per week
PbS site = 1050 ppm
4 site visits per week
PbS site = 550 ppm
PbSw' (ppm)
329
336
335
PbDw2 (ppm)
230
235
235
GM PbB (ng/dL)3
4.5
4.5
4.5
P,o(%)3
4.3
4.5
4.5
1 PbSw weighted soil lead concentration; calculated using Equation 1.
2 PbDw weighted dust lead concentration calculated using Equation 5.
3 Results from the IEUBK. model. GM = geometric mean. Residential PbS held at 50 ppm. All runs using 0-84
months in IEUBK model.
33
-------
EXAMPLE 2: RECREATIONAL & TRESPASSING EXPOSURE SCENARIO
The site in this scenario is a slag pile in an area where the use is assumed to be primarily
industrial and commercial. Although the site does not contain a developed recreational area, the
slag pile is an attractive nuisance and children have been observed dirt biking on the hill. The
goal of the soil cleanup level is to minimize lead exposure for children who would visit the site
during the warmer months for recreation or possibly trespassing.
The State standard for non-residential areas is 1000 ppm for lead. This soil lead
concentration was used to evaluate the possible impacts of child exposure. It was averaged with a
default residential soil lead concentration of 100 ppm, based on the assumption that 30% of soil
ingested would be from the site and 70% from the home, yielding a PbSw of 370 ppm. This
weighted concentration was not averaged over the entire year, since exposures were expected to
occur for only part of the year (4 continuous months). In all cases, other model inputs were kept
at default values, including soil ingestion rates (PbDw = 0.7 * PbSw). Note that for simplicity,
these calculations assume that total soil ingestion occurs at the default rate.
In addition to typical residential exposure to lead in soil, it was expected that dirt biking
would result in additional soil ingestion. The assumption that 10% of waking hours (1.2 hours/12
hours) could be spent at the site was incorporated into the calculations to imply that an additional
10% of typical total dirt ingestion would occur at the site.
The following assumptions were made in running the IEUBK model:
1. Site exposure would include an additional 10% of typical total dirt ingestion.
2. Daily lead intake over the 4 months was averaged over 12 months for input to the IEUBK
model.
3. An exposure period of 112 days/year was selected (7 days/week for 16 weeks).
4. Residential exposure was characterized by the IEUBK default exposure levels, since no
site-specific data for the residence were available. This is appropriate for 0-11 month
children, because they are not expected to have significant contact with site soil based on
the likely exposure pathway.
34
-------
To evaluate the possible impact of soil ingestion assumptions and in the absence of site-
specific information concerning soil ingestion, various soil ingestion assumptions were explored
to bound the results. For the purposes of this assessment, the default total soil ingestion rates
were used to bound the low-exposure scenario. Because of the potential for higher contact rates
with soil at the site, additional contact-intensive scenarios are also warranted. A low-exposure
scenario using the IEUBK model defaults was chosen. For a medium-exposure scenario, the total
dirt ingestion rates would be 145% of default rates. For a high-exposure scenario, an additional
200 mg/day was used. The various exposure assumptions are shown in Table A-3.
Table A-3. Calculation of a Plausible Range of Dirt (soil/dust) Ingestion Rates for Site to Bound Risk Estimates
When Site-Specific Soil Ingestion is Unknown.
Age group
(months)
0-1 11
12-23
24-35
36-47
48-59
60-71
72-84
Total dirt ingestion rates (g/day)
Low exposure scenario
Total=default
0.085
0.135
0.135
0.135
0.100
0.090
0.085
Medium exposure scenario
Total=145% of default total
0.085
0.195
0.195
0.195
0.145
0.131
0.125
High exposure scenario2
Total=0.200 g/day+ default
0.085
0.335
0.335
0.335
0.300
0.290
0.285
1 Additional soil contact is not applicable to children <1 year, since they are not likely to have significant
additional exposure to site soil.
2 The high exposure scenario is based on adding 200 mg/day to the default ingestion rates, consistent with
OSWER guidance on assessing risk for Reasonably Maximally Exposed (RME) individuals.
Since daily exposure to lead in soil for 4 months is expected to produce a pseudo-steady-
state PbB concentration, the daily average soil concentration over the 4-month period for this
application is the most relevant input to the IEUBK model. Use of an annualized daily
concentration does not allow for estimating the body burden that results during sustained shorter
periods of relatively higher exposure. Thus, the TRW expects a serious underestimation of the
actual PbB distribution to result from averaging the site exposure over the entire year. However,
35
-------
IEUBK predictions using the average daily exposure level for a 4-month period might be
expected to be somewhat of an overestimate, since the model was designed to project PbB
concentrations from sustained daily exposure over the first 84 months of childhood. This built-in
assumption of sustained, chronic daily exposure does not allow for a wash-out period between
the annual 4-month exposures, which would be associated with the exposure scenario for this
site. EPA generally anticipates that there will be some seasonal fluctuation of exposure
conditions. IEUBK predictions are therefore expected to approximate the PbB concentrations
related to the higher lead exposure levels in an annual cycle of lead exposure, where measured
lead concentrations in soil are expected to remain relatively constant.
Table A-4. Risk Estimates for the Various Alternate Soil Ingestion Scenarios Using Each Age Group in the IEUBK
Model Using Daily Average Soil Concentration over the 4-Month Period.
Age group (months)
0-11
12-23
24-35
36-47
48-59
60-71
72-84
Risk estimates for each ingestion scenario1
Default (low)
GM PbB
(Hg/dL)
5.0
6.0
5.6
5.3
4.4
3.8
3.4
PIO
(%)
7.0
13.6
10.8
8.9
4.2
2.0
1.6
Medium
GM PbB
(|ig/dL)
5.0
7.4
7.2
6.8
5.7
4.9
4.4
PIO
(%)
7.0
26.0
23.8
20.9
11.6
6.3
3.9
High
GM PbB
(Hg/dL)
5.0
10.3
10.4
10.0
9.2
8.4
7.8
PIO
<%)
7.0
52.8
53.5
50.3
42.8
35.6
29.5
1 IEUBK. inputs were PbSw= 370 ppm and PbDw = 259 (using Equation 5). All others were default. GM =
geometric mean. P,0 = probability of exceeding 10 Hg/dL.
This matrix demonstrates to the risk manager that the State Applicable or Relevant and
Appropriate Requirements (ARAR) is not protective for these exposure scenarios.
36
-------
EXAMPLE 3: DAYCARE EXPOSURE SCENARIO FOR INDUSTRIAL PARK DAYCARE
The site in this scenario is a proposed daycare facility in an area that is zoned for
industrial and commercial land use. The goal is to determine whether the proposed soil cleanup
for the site (700 ppm) is protective of children exposed at the proposed daycare facility.
The following assumptions were considered plausible:
1. Children may be exposed to lead in exterior soil and interior dust both at the
daycare facility as well as at home (located outside the site).
2. The concentration of lead in exterior soil at the daycare facility is 700 ppm, the
proposed cleanup level for the site.
3. A child visits the daycare facility 5 days per week and stays home 2 days per
week.
4. Site sampling indicates a mean residential soil concentration of 100 ppm.
After estimating media concentrations for each location, the fraction of waking hours that
a child spends in each location should be estimated to determine a reasonable estimate of the
time-weighted average concentration across all locations. The fraction of waking hours for each
location then can be used to calculate the time-weighted average soil and dust concentrations that
can be entered directly into the IEUBK model.
Apportioning exposure across locations according to hours awake:
„ _ 8hows/day • 5days/week _ 40 Q .g
daycart \2hourslday- 7days/week 84
Note that the equation needs to be modified to account for the hours that a child spends
on different days at the location. This would be true for the home (weekday versus weekend
activities); however, the home fraction can be more easily calculated by subtracting the fraction
of hours spent at other locations from 1.0; thus, the remaining time spent awake at home:
- (1-0-0.48) = 0.52
37
-------
Deriving a weighted soil concentration from home and daycare (Equation 5):
PbSw = (PbS x F+ PbS x
PbSw = (WQpptn x 0.52) + (700 ppm x 0.48)
PbSw= 52 ppm + 336 ppm
PbSw= 388 ppm
The estimated ratios of indoor dust lead concentration are applied to soil lead
concentration (IEUBK default for Msd is 0.7). Because the residences were relatively new homes,
lead-based paint was not expected to contribute to indoor dust.
Example for weighted dust concentration from home and daycare:
PbDw = PbSw x Mid
PbDw =3BZppmxQl
PbDw =21 2 ppm
According to the IEUBK model, these weighted concentrations result in a GM PbB of 4.9
Hg/dL and a P10 of 6.7%. These results suggest that the proposed soil lead concentration at the
daycare facility (700 ppm) would not be protective of children.
38
-------
EXAMPLE 4: INTERMITTENT NON -RESIDENTIAL EXPOSURE FOR ADULTS
The following example shows how the ALM can be used to assess risk posed by lead
contamination at a non-residential site under two different exposure scenarios. The site soil lead
concentration is 500 ppm indoors and 1000 ppm outdoors. In this example, a utility worker is
laying new line at a contaminated site. The new line is expected to take 3 days/week for 13 weeks
during the year. The worker is otherwise employed indoors at the site for the other 2 days of the
week. Site data suggest that the worker is involved in contact-intensive jobs both indoors
(sweeping) and outdoors (digging); thus, a site-specific soil central tendency ingestion rate of 100
mg/day is appropriate for both indoors and outdoors.
To calculate the time-weighted soil concentration to which the worker is exposed, the
following equation applies (Equation 1):
PbSw = (PbS,ndoors x EFIndoors) + (PbS0utdoors x EF0utdoors)
PbSw = (500 ppm x 2 days/7 days) + (1000 ppm x 3 days/7 days)
PbSw = 142.8 ppm + 428.6 ppm
PbSw = 571 ppm
This PbSw can then be entered into the ALM with the following changes:
Averaging time (AT) = 91 days (13 weeks x 7 days/week).
Exposure frequency (EF) = 65 days (13 weeks x 5 days/week).
Site-specific soil ingestion rate of 100 mg/day for both indoors and outdoors.
Baseline PbB0 and GSD for the range of inputs were selected from NHANES in
analysis to span the range from all races/ethnic groups in the U.S. to the Mexican
American group in the U.S.
Note that residential exposure to lead is not reflected in this time weighting calculation,
since the residential contribution is reflected in baseline blood lead. To assess residential lead
contamination, the IEUBK model should be used. The ALM would be run as shown in Table A-
5. The results suggest a range of P10 values for the fetus depending on the site conditions (GSD
for homogeneous or heterogeneous site exposure histories of the population at the site). Because
the site workers are anticipated to reflect people from throughout a varied community, a
heterogeneous GSD (2.3) would be appropriate to protect the most sensitive population. The P10
39
-------
is 4.4% for this group, which is less than the 5% EPA goal. This suggests that the outdoor soil
lead concentration of 1000 ppm at this site would be protective under the conditions described.
Table A-5. ALM Inputs and Results for the Utility Worker Exposure Scenario Using Time-Weighted Average Soil
Lead Concentration.
Exposure variable
PbS
Rfetal/matemal
BK.SF
GSDi
PbBO
IRS
AFS,D
EPS, D
ATS, D
Description of exposure variable
Soil lead concentration
Fetal/maternal PbB ratio
Biokinetic slope factor
Geometric standard deviation PbB
Baseline PbB
Soil ingestion rate (including soil-derived
indoor dust)
Absorption fraction (same for soil and dust)
Exposure frequency (same for soil and dust)
Averaging time (same for soil and dust)
Units
|lg/g or ppm
-
Hg/dL per (ig/day
—
|ig/dL
g/day
-
days/year
days/year
Inputs
571
0.9
0.4
2.1
1.5
0.1
0.12
65
91
571
0.9
0.4
2.3
1.7
0.1
0.12
65
91
RESULTS
PbBadult
PbBfetal, 0.95
PbBt
P(PbBfetal >
PbBt)
PbB of adult worker, geometric mean
95th percentile PbB among fetuses of adult
workers
Target PbB level of concern (e.g., 10 Hg/dL)
Probability that fetal PbB > PbBt, assuming
lognormal distribution
Hg/dL
Hg/dL
Hg/dL
%
2.5
7.6
10
2.2%
2.7
9.5
10
4.4%
40
-------
EXAMPLE 5: SEASONAL EXPOSURE: CHILDREN VISITING A SITE 4 MONTHS OF THE YEAR
One cannot simulate exposures as discrete 4-month exposure events using the IEUBK
model and ALM. In these models, the exposure must be time-weighted to calculate risk. Table
A-6 summarizes the predicted mean PbB concentrations across age groups for an example
scenario in which children experience a single, non-recurring 4-month exposure at a site
contaminated with a PbS of 2850 ppm. Residential sampling showed 200 ppm for the residential
PbS, and a PbSw of approximately 1000 ppm is assumed (based upon 70% of soil ingestion
occurring at the residence [200 ppm] and 30% at the site due to site-specific conditions [2850
ppm]; this weighted concentration is not averaged over the entire year). Dirt (soil + dust)
ingestion rates are also kept at model default values. For comparative purposes, the left portion of
the table indicates P10 values calculated by the model under the more typical constant, or
cumulative, type of scenario. This was done for PbS concentrations of 200 ppm (residential
exposure only), 430 ppm (the PbS at which P10=5% when all other model input parameters are at
default levels), and 1000 ppm (the weighted average). The right portion of the table provides the
model predictions resulting from the one-time exposures occurring singly in each of the six
analyzed age ranges (no site exposure was assumed for the 6-11 month group). For each age
group, the interval associated with the site exposure is underlined.
41
-------
Table A-6. 1EUBK. Risk Estimates (P,0; Percent of Popuktion Exceeding 10 Hg/dL) Corresponding to Varying Soil
Exposures by Age. All the values have been updated with lEUBKwin (build 253).
Age
group
(months)
6-114
12-23
24-35
36-47
48-59
60-71
72-84
12-84
Constant (cumulative) exposure to
lead in soil
200 ppm1
2.9
4.6
3.4
2.6
1.1
0.5
0.3
1.6
430 ppm2
8.6
13.1
10.3
8.5
4.0
1.9
1.1
5.4
1000 ppm3
30.0
40.5
35.3
32.0
19.6
11.4
7.4
23.5
One yea
to 1000
soil leat
2.9
34.5
8.0
2.7
1.1
0.5
0.3
r exposure during underlined age range (in months)
ppm site soil lead, otherwise 200 ppm residential
1 was used as the media concentration5
2.9
4.6
26.7
7.4
1.2
0.5
0.3
2.9
4.6
3.4
22.1
4.8
0.6
0.3
2.9
4.6
3.4
2.6
9.9
2.3
0.4
2.9
4.6
3.4
2.6
1.1
5.0
1.4
2.9
4.6
3.4
2.6
1.1
0.5
Z£
16.9
1 200 ppm is the residential PbS based on sampling.
2 430 ppm in soil is the concentration associated with a 5% risk of elevated PbB for children 12-84 months old,
when all other 1EUBK. inputs are at default (dust lead=200 ppm, geometric mean PbB=4.7 |J.g/dL, GSD = 1.6).
This level of precision is reported only for completeness; in general, back-calculated 1EUBK estimates of soil and
dust lead concentrations should be rounded to the nearest 100 ppm, or 400 ppm in this case.
3 Weighted average of 70% at 200 ppm and 30% at 2850 ppm; default dirt ingestion rates were assumed.
4 Age group shown for baseline; no site exposure assumed; PbB concentrations were not included in overall
means.
5 The predictions in italics correspond nominally to the children having no further access
to the site.
The predictions in italics in Table A-6, which correspond nominally to the children
having no further access to the site, suggest that it takes more than a year for the predicted annual
average PbB concentrations to return to baseline (200 ppm column). Note: the IEUBK model
was not validated to estimate elimination kinetics following a sharp change in exposure levels.
Nevertheless, this example indicates the importance of evaluating the available information about
exposure patterns at the site to assess how well the assessment assumptions are likely to
approximate the actual exposure patterns that may be occurring at the site. Where there is
uncertainty in choosing the most appropriate concentrations for these variables, the TRW
generally recommends considering several possible alternative scenarios to address a range of
plausible possibilities.
42
-------
EXAMPLE 6: LAWN MAINTENANCE NEAR A RIVER
For a lead-contaminated site (mean soil lead concentration at the site is 2000 ppm)
located along a river, the most likely future use of the property was lawn mowing and other
minor groundskeeping activities. This scenario was not envisioned as including soil-intensive
activities due to the extensive ground cover at the site, so it assumes a central tendency soil
ingestion of 50 mg/day (U.S. EPA, 1997). For the ALM, the central tendency value would be
appropriate. The goal is to develop PRG for the site based on the most likely receptor, the lawn
maintenance worker. The PRG spreadsheet of the ALM model may be used with the following
changes.
Based on current activities at this site, it was assumed that the lawn would be mowed for
three days out of the week for seven months of the year. Because of vagaries in the Gregorian
calendar and for consistency with lead biokinetic models (as explained in section 3.1 of the text),
risk can be assessed as a time-weighted average soil concentration based on 3 days of exposure
out of 7 days. Alternately, the exposure could be expressed as an exposure frequency (EF) of 90
(3 days/week x 4.3 weeks/month x 7 months) days and an averaging time (AT) of 211 (7
days/week x 4.3 weeks/month x 7 months) days. In the EF/AT relationship, the factors of 4.3
weeks/month and 7 months drop out in the calculation, resulting in an EF of 3 and an AT of 7 for
the spreadsheet.
The exposure scenario specified at this site, using the ALM, results in 7 months of
exposure (3 times per week) and 5 months of "washout" when no excess site-related lead
exposure occurs (see Figure 2 of the text). In determining whether the "washout" period should
be considered in the risk calculation, a determination must be made whether the duration of site
exposure could reasonably produce a body burden of lead that results in an adverse health effect.
In this example, 7 months of exposure would satisfy the minimum exposure duration to achieve a
quasi-steady state PbB concentration (3 months). Moreover, this exposure duration would also
likely be sufficient time for a body burden of lead to develop that would be associated with
adverse health effects. Therefore, a plausible risk calculation for this site would be based on 3
43
-------
days of exposure out of 7 days as if the exposure occurred for the entire year and ignores the
effect of the 5 months of the year when site exposure does not occur. This can also be interpreted
as follows: the increase in blood lead concentration during the exposure season is the basis for
the risk calculation, and the "washout" period is not considered in the calculation of the PRO.
To calculate the time-weighted soil concentration to which the lawn worker is exposed,
the PRO spreadsheet of the ALM is used with the following changes (see Table A-7):
Averaging time (AT) = 7.
Exposure frequency (EF) = 3.
Site-specific soil ingestion rate (IRS) of 50 mg/day for both indoors and outdoors.
Other parameters set as specified in the ALM Guidance.
Baseline PbB0 and GSD for the range of inputs were selected fromNHANES IE
analysis to span the range from all races/ethnic groups in the U.S. to the Mexican
American group in the U.S.
Note that residential exposure to lead is not reflected in this time weighting calculation,
since the residential contribution is reflected in baseline blood lead. To assess residential lead
contamination, the IEUBK model should be used. This information can be entered into the
spreadsheets provided for the calculation of blood lead or preliminary remediation goals (PRGs).
For a site where it is assumed that the population has a high baseline blood lead concentration
and a high geometric standard deviation of the blood lead, the PRG values ranged from 1729
ppm and for U.S. data for all races combined to 1092 ppm for Mexican-American population
from all regions in the U.S. (see Table A-7). In this case, a decision was made to average
exposure during the quasi-steady state period (exposure season) and consider this as if it occurred
throughout the year, ignoring the "washout" period. This is a reasonably conservative approach
for the site.
44
-------
Table A-7. ALM Inputs and Results for the Lawn Maintenance Worker Exposure Scenario (#1) Using ALM PRO
Spreadsheet.
Exposure variable
PbBfetal, 0.95
Rfetal/maternal
BK.SF
GSDi
PbB0
1RS
AFS D
EFS>D
ATS,D
PRO
Description of exposure variable
95th percentile PbB among fetuses of adult
workers
Fetal/maternal PbB ratio
Biokinetic slope factor
Geometric standard deviation PbB
Baseline PbB
Soil ingestion rate (including soil-derived
indoor dust)
Absorption fraction (same for soil and dust)
Exposure frequency (same for soil and dust)
Averaging time (same for soil and dust)
Preliminary Remediation Goal
Units
Hg/dL
--
Hg/dL per Jig/day
--
Hg/dL
g/day
--
days/year
days/year
ppm
Inputs
10
0.9
0.4
2.1
1.5
0.05
0.12
3
7
1729
10
0.9
0.4
2.3
1.7
0.05
0.12
3
7
1092
1RS = Intake rate of soil, and outdoor soil derived dust.
EFSD = Exposure frequency for contact with assessed soils and/or dust derived in part from these soils. In this
example, based on 3 days/week x 4.3 weeks/month x 7 months.
ATSD = Averaging time; the total period during which soil contact may occur. Based on , based on 7 days/week x
4.3 weeks/month x 7 months.
45
-------
REFERENCES
Bornschein R.L., P. Succop, K..N. Dietrich, C.S. Clark, S. Que Hee, and P.B. Hammond. 1985. The
influence of social and environmental factors on dust lead, hand lead, and blood lead levels in young
children. Environ. Res. 38(1): 108-18.
Hogan, K., A. Marcus, R. Smith, and P. White.. 1998. Integrated Exposure, Uptake, Biokinetic Model for
Lead in Children: Empirical Comparison with Epidemiologic Data. Environ. Health Perspect. 106 (S6):
1557-67.
Matte, T.D., J.P. Figueroa, S. Ostrowski, G. Burr, L. Jackson-Hunt, and E.L. Baker. 1991. Lead Exposure
from Conventional and Cottage Lead Smelting in Jamaica. Arch. Environ. Contam. Toxicol. 21: 65-71.
U.S. EPA. 1994. Guidance Manual for the Integrated Exposure Uptake Biokinetic Model for Lead in
Children. Officeof Emergency and Remedial Response. Washington, D.C., U.S. EPA 540/R-93/081,
PB93-963510.
U.S. EPA. 1995. Seasonal Rhythms of'Blood-Lead Levels: Boston, 1979-1983. Washington, D.C., U.S.
EPA 747-R-94-003.
U.S. EPA. 1996a. Urban Soil Lead Abatement Demonstration Project, Volume I: EPA Integrated Report.
Washington, DC, U.S. EPA 600/P-93/001aF.
U.S. EPA. 1996b. Recommendations ofthe Technical Review Workgroup for Lead for an Interim Approach
to Assessing Risks Associated with Adult Exposures to Lead in Soil. Washington, DC, U.S. EPA Technical
Review Workgroup for Lead (TRW) http://www.epa.gov/superfund/programs/lead/products.htmtfgTiidance.
U.S. EPA. 1997. Exposure Factors Handbook (Final) Volumes I, II, III. Washington, DC, U.S. National
Center for Environmental Assessment, http://www.epa.gov/ncea/exposfac.htm.
U.S. EPA. 1998. Short Sheet: IEUBK Model Mass Fraction of Soil in Indoor Dust (MSD) Variable.
Washington, DC, U.S. EPA Technical Review Workgroup for Lead (TRW). EPA #540-F-00-008, OSWER
#9285.7-34. http://www.epa.gov/superfund/programs/lead/products.htmtfshort.
U.S. EPA. 1999a. Frequently Asked Questions on the IEUBK Model: Description of Acute Exposures;
Washington, DC, U.S. EPA Technical Review Workgroup for Lead (TRW).
http://www.epa.gov/superfiind/programs/leadAeubkfaq.htni.
U.S. EPA. 1999b. Short Sheet: IEUBK Model SoiUDust Ingestion Rates. Washington, DC, U.S. EPA
Technical Review Workgroup for Lead (TRW). EPA 540-F-00-007, OSWER 9285.7-33.
http://www.epa.gov/superfund/program s/lead/prod\icts.htm#short.
U.S. EPA. 2001. Supplemental Guidance for Developing Soil Screening Levels for Superfund Sites (Peer
Review Draft). Washington, DC, U.S. EPA Office of Solid Waste and Emergency Response. EPA 9355.4-
24. http://www.epa.gOV/.superfund/resource.s/soiV.ssgmarch01.pdf.
White, P., P. Van Leeuwen, B. Davis, M. Maddaloni, K.. Hogan, A. Marcus, and R. Elias . 1998. The
Conceptual Structure of the Integrated Exposure, Uptake, Biokinetic Model for Lead in Children. Environ.
Health Perspect. 106 (S6): 1513-30.
46
------- |