RISK ASSESSMENT FOR
SECTION 403 ANALYSES
Draft Report
VOLUME III
APPENDICES W-Z
Prepared by
Battelle
505 King Avenue
Columbus, Ohio 43201
for
U.S. Environmental Protection Agency
EPA Contract No. 68-D5-0008
Task 2-10
Chemical Management Division
Office of Pollution Prevention and Toxics
Office of Prevention, Pesticides, and Toxic Substances
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
September 25, 1996
PRELIMINARY DRAFT DOCUMENTATION
ON THE EPI MODEL
APPENDIX W
-------
Multi-Media Model Based on the Rochester Lead-in-Dust Study (EPI Model)
for Use in Predicting a National Distribution of Children's Blood-Lead Concentrations
for the Section 403 Risk Assessment
EXECUTIVE SUMMARY
Purpose Of The Report
This report documents development and evaluation of a multi-media model (EPI model)
relating measures of lead in a residential environment to geometric mean children's blood-lead
concentrations. The model will be used as one tool in the Section 403 risk assessment where
EPA will estimate the range of risk reductions that are expected to result from a variety of
potential Section 403 standards. EPA will estimate a national distribution of blood-lead levels
(and, ultimately, potential health effects) before enactment of the Section 403 standards, and then
rely on models which relate environmental levels of lead to children's blood-lead levels to
estimate a national distribution of blood-lead levels (and potential health effects) after enactment
of specific 403 standards. Environmental measures of lead from the HUD National Survey will
be used as inputs to the EPI model to predict the national distribution of blood-lead
concentrations. Therefore, the model development was constrained to variables in the HUD
National Survey data set. Given time and budget constraints the goal for the EPI model
development could not include construction of the best possible model based on multiple data
sources. Rather, the goal was to develop a model that could be used to give an approximation of
expected blood-lead concentrations related to residential environmental lead based on a single
source of data.
In this report the EPI model is presented and its prediction of a national distribution of
blood-lead concentrations is compared to the results of the NHANES III Survey
The EPI Model
The mathematical model used to characterize the relationship was:
In(PbB) = P0 + p, • In(PbS) + P2 • PbP + P3 • In(PbF) + P4 • In(PbW) + e
where PbB represents the blood-lead concentration, PbF corresponds to exposure from interior
floor dust, PbW represents environmental lead from window sills, PbS represents soil-lead, PbP
represents paint hazard, and e represents the residual error left unexplained by the model.
Model Development Issues
The choice and construction of variables, the mathematical form of the EPI model,
assessment of goodness-of-fit and influential points, and the treatment of measurement error in
predictor variables were all given consideration during the development of the EPI model.
Draft - Do Not Cite or Quote \ September 25, 1996
-------
One particular difficulty was the fact that the EPI model was constructed using dust lead
results collected from wipe sampling in the Rochester study, whereas dust lead results in the
HUD National Survey were collected from blue nozzle vacuum sampling. Two different
statistical methods were assessed to account for differences between the sample collection
methods. The first involved converting the National Survey blue nozzle measurements to wipe
measurements before input into the EPI model, thereby accounting for systematic differences
between the two methods. The second involved more complicated methodology to account for
both systematic differences as well as differences in error structures between the two sampling
methods. The first approach was chosen for implementation in predicting the national
distribution of blood-lead concentrations through use of the EPI model.
Results Of The Comparison With NHANES III
The predicted distribution of blood-lead concentrations obtained by applying the EPI
model to the HUD National Survey Data was compared to NHANES III to measure how well the
EPI model performed. Results of this comparison indicate:
• The national geometric mean blood-lead concentration can be adequately
predicted by the EPI model applied to the HUD National Survey.
• The variability in the national distribution of blood-lead concentrations predicted
by the EPI model using the HUD National Survey is approximately 1.80 (GSD),
in contrast to a GSD of 2.05 for NHANES IE.
• The estimated proportions of blood-lead concentrations exceeding 10,20 or 30
ug/dL using the EPI model predictions are lower than the corresponding
proportions estimated by NHANES ffl.
Differences between the Rochester study population and the national population may
explain why a multi-exposure model (such as the EPI model) developed using data from the
Rochester Study does not fully capture the variability in blood-lead concentrations nationwide.
Use Of The EPI Model
The EPI model is expected to be used in the Section 403 risk assessment and economic
analyses to predict a distribution of blood-lead concentrations related (jointly) to measures of
lead in three media at the child's primary residence: paint, dust and soil. Estimates of
environmental levels of lead after enactment of the Section 403 standards and after interventions
resulting from the standards will be used as inputs to the model. The EPI model should only be
used to predict a distribution of blood-lead levels when environmental levels for all media are
known or estimated.
Draft - Do Not Cite or Quote \\ September 25, 1996
-------
1.0 INTRODUCTION
In order to better inform risk managers as they consider various options for the Section
403 standards, EPA will estimate the range of risk reductions that are expected to result from a
variety of potential standards. In order to do this, EPA will estimate a national distribution of
blood-lead levels (and, ultimately, potential health effects) before enactment of the Section 403
standards, and then rely on models which relate environmental levels of lead to children's blood-
lead levels to estimate a national distribution of blood-lead levels (and potential health effects)
after enactment of specific 403 standards. The EPI model is expected to be used in the Section
403 risk assessment and economic analysis to predict a distribution of blood-lead concentrations
related Gointly) to measures of lead in three media at the child's primary residence: paint, dust,
and soil. Given time and budget constraints, the goal for the EPI model development could not
include construction of the best possible model based on multiple data sources. Rather, the goal
was to develop a model that could be used to give an approximation of expected blood-lead
concentrations related to residential environmental lead based on a single source of data.
In this report the EPI model is presented and its prediction of a national distribution of
blood-lead concentrations is compared to the results of the NHANES III Survey as follows:
A national distribution of housing and population characteristics
was estimated using the HUD National Survey of environmental
levels of lead in paint, dust, and soil in residential housing along
with pertinent Census information. The Census information and
the HUD National Survey measurements of environmental lead
(after appropriate conversions) were used as inputs to the model to
predict a national distribution of children's blood-lead levels
before enactment of the Section 403 standards. This pre-
rulemaking distribution was compared to the national distribution
of children's blood-lead concentrations estimated by the NHANES
III survey to assess the adequacy of the model and its applicability
on a national level.
The EPI model will later be used to estimate the national distribution of children's blood-
lead levels after enactment of the Section 403 standards. Estimates of environmental levels of
lead after enactment of the Section 403 standards and after interventions resulting from the
standards will be used as inputs to the model. Comparison of the pre- and post-rulemaking
distributions will allow estimation of the benefits associated with the rulemaking.
The EPI model is not intended to be used to estimate the effect of a single media on
blood-lead levels. The model should only be used to predict a distribution of blood-lead levels
when environmental levels for all media are known or estimated. Individual parameter estimates
should not be interpreted in isolation.
Draft - Do Not Cite or Quote 1 September 25. 1996
-------
2.0 DESCRIPTION OF THE EPI MODEL
The mathematical model used to characterize the relationship was:
In(PbB) = P0 + P, • In(PbS) + P2 • PbP + P3 • In(PbF) + P4 • In(PbW) + e
where PbB represents the blood-lead concentration, PbF corresponds to exposure from interior
floor dust, PbW represents environmental lead from window sills, PbS represents soil-lead, PbP
represents paint hazard, and e represents the residual error left unexplained by the model.
The EPI model was developed using measures of environmental lead and other factors
observed in both the Rochester Lead-in-Dust Study and the HUD National Survey. Variables
whose definition provided a convenient translation when applied to the National Survey, whose
predictive power in Rochester were high, and whose spread in the National Survey populations
covered a wide enough range of values, were used in the EPI model. These variable descriptions
are contained in Table 1.
The paint/pica hazard predictor variable was interpreted differently between the
Rochester Study and the HUD National Survey data. The primary difference was that the
paint/pica hazard input variable from the HUD National Survey data was based on the presence
of lead based paint on both interior and exterior surfaces, whereas the variable used in Rochester
for estimation of the paint/pica model parameter was based on the presence of lead based paint
on only interior surfaces. Lead based paint on exterior surfaces was not considered in the
estimation of the paint/pica model parameter based on Rochester data because approximately 84
percent of houses in the Rochester Study were built prior to 1940 and as a result virtually every
home surveyed in the Rochester Study had lead based paint on exterior surfaces. Therefore, a
paint/pica hazard variable which included presence of exterior lead based paint in Rochester lost
its statistical significance and its predictive power.
The choice and construction of variables, the mathematical form of the EPI model,
assessment of goodness of fit and influential points, and the treatment of measurement error in
predictor variables were all given consideration during the development of the EPI model.
Decisions made with respect to adjustments for measurement error in predictor variables are
documented in the following section.
Parameter estimates and associated standard errors, and measures of R-squared and the
residual standard deviation for the EPI model are provided in Table 2. Note that the parameter
estimate associated with floor dust-lead loading was only borderline statistically significant when
considered jointly with the effect of window sill dust-lead loading (and other exposure variables)
in the EPI Model.
Draft - Do Not Cite or Quote 2 September 25. 1996
-------
Table 1. Variables in the EPI Model and Corresponding Input Variables From the HUD
National Survey
Predictor
Variable
Rochester Study
HUD National Survey Input Variables
Soil
Natural log transformation of driplme soil-
lead concentration (fine soil fraction).
The natural log transformation of the average ot dnpiine an
entryway soil-lead concentrations. If both of these values
were missing, an imputed value' was used.
Floor Dust
The natural logarithm of the area weighted
arithmetic average (wipe) dust-lead loading
from carpeted and uncarpeted floors.
In the comparison with NHANES III. the HUD National
Survey blue-nozzle dust-lead loadings were converted using
the following conversion equation: Wipe = 11.40"IBIue-
Nozzle)0890. The natural log transformation of the area-
weighted arithmetic average dust-lead loadings from 3
sample locations (wet, dry and entry rooms) was then used
as the measure of lead in dust. If the dust-lead loading
from all of the 3 sample locations was missing, an imputed
value1 was used.
Window Sill
Dust
The natural logarithm of the area-weighted
arithmetic overage (wipe) dust-lead loading
from window sills.
In the comparison with NHANES III, the HUD National
Survey blue-nozzle dust-lead loadings from window sills
were converted using the following conversion equation:
Wipe = 5.79*(Blue-Nozzle)10". The natural log
transformation of the area-weighted arithmetic average
dust-lead loadings from window sills from 3 sample
locations (wet, dry and entry rooms) was then used as the
measure of lead in dust. If the window sill dust-lead loading
from all of the 3 sample locations was missing, an imputed
value1 was used.
Interior
Pica/Paint
An indicator variable which was nonzero
whenever each of the following conditions
existed in a residential unit: presence of
deteriorated or damaged interior paint in the
household; and presence of interior lead-
based paint in the household; and presence
of a child with paint pica in the household.
The paint variable had values of:
0 No LBP (XRF reading < 11, or
condition* is Good, or child does not
exhibit paint pica;
1 LBP (XRF reading i 1), condition is
Fair or Poor, and child exhibits paint
pica rarely;
2 LBP (XRF reading 2 1), condition is
Fair or Poor, and child exhibits paint
pica at least sometimes.
HUD National Survey homes were determined to have
deteriorated LBP whenever there is any deterioration in
interior or exterior lead-based paint, as measured by square
footage (that is, square footage of detenorated LBP surface
> 0). That is, the LBP indicator was defined as
1 Whenever square footage of surface exhibiting
detenored LBP (interior and exterior) > 0
0 Otherwise
The pica factor was only considered for houses with
deteriorated LBP. In these houses, it was assumed that 9%
of U.S. children aged 1-2 years have pica for paint. For the
children with pica for paint, the pica value was defined to
be 1.5".
Imputed values were based on a presence of LBP indicator variable and on a house age-specific indicator. The presence of
LBP indicator was defined as:
0 Predicted maximum XRF < 1 for both interior and exterior samples
1 Predicted maximum XRF 2 1 for either interior or exterior samples.
The house age-specific indicator had categories: Pre-1940, 1940-1960, 1960-1979, Post-1979.
The imputed values for dust and soil were constructed by taking the means for the associated subsets formed by crossing
the paint and age of house categories.
Condition of the paint is characterized as Good whenever less than 5% of the surface is deteriorated; Fair whenever 5% to
15% of the surface is deteriorated; and Poor whenever more than 15% of the surface is deteriorated.
Paint pica was collapsed in the Rochester study to have values:
0 No paint pica
1 Child exhibits paint pica rarely
2 Child exhibits paint pica at least sometimes.
A value of 1.5 was chosen as the input value for those children exhibiting pica at least rarely in applying the EPI Model to
the HUD National Survey.
Draft - Do Not Cite or Quote
September 25, 1996
-------
Table 2. Parameter Estimates and (Associated Standard Errors), Coefficient of
Determination (R2) and Residual Standard Deviation for EPI Model
Parameter
30
3,
Pz
Pa
P»
R2
0
Description
Intercept
log (PbS): Drip-line Soil-Lead Concentration (fine soil
fraction)
PbP: Indicator of Interior Paint/Pica Hazard
log (PbF): Area-Weighted Arithmetic Mean (Wipe)
Dust-Lead Loading from Any Floor (Carpeted or
Uncarpeted)
log (PbW): Area-weighted Arithmetic Mean (Wipe)
Dust-Lead Loading from Window Sills
Coefficient of Determination
Root Mean-Square Error (Residual Error)
Estimate
0.418
(0.240)
0.114
(0.035)
0.248
(0.100)
0.066
(0.040)
0.087
(0.036)
21.67%
0.56188
3.0 ADJUSTMENTS TO THE EPI MODEL TO ACCOUNT FOR DIFFERENCES IN
SAMPLING METHODS BETWEEN ROCHESTER AND THE HUD NATIONAL
SURVEY
The goal of the EPI Model is to provide a relationship between blood-lead concentration
and various environmental lead exposures as measured by and occurring in the HUD National
Survey. Unfortunately, the HUD National Survey contains no information about blood-lead
concentration. However, data from the Rochester Lead-in-Dust Study can provide a basis for the
EPI Model. At issue is how to use the Rochester data set to develop an EPI Model applicable to
the data observed in the HUD National Survey. Matters are further complicated by the fact that
the sampling methodology used to measure lead exposures in HUD is different from that used in
Rochester. Specifically, two of the lead exposure measurements in HUD are blue nozzle floor
dust lead loading and blue nozzle window sill dust lead loading, compared to floor wipe dust
lead loading and window sill wipe dust lead loading in Rochester.
The following two different statistical methods were investigated to account for
differences in sample collection methods between the Rochester Study and the HUD National
Survey when assessing the impact of 403 rulemaking on children's blood-lead levels:
[ 1 ] Use of conversion equations to convert Blue Nozzle Vacuum dust-lead loadings
measured in the HUD National Survey to Wipe dust-lead loadings as input to the EPI
Model, which was based on wipe samples collected in the Rochester Study.
[2] Establishing a relationship between blood-lead and environmental variables as
measured by methods used in the Rochester Study, and then adjusting this relationship
Draft - Do Not Cite or Quote
September 25, 1996
-------
to use environmental variables as measured in the HUD National Survey. The
adjustment takes into account both systematic differences and differences in error
structures between the Rochester predictor variables and the HUD National Survey
predictor variables. The method provides a relationship between blood-lead
concentration and floor and window sill dust lead loadings and other covariates as
observed in the HUD National Survey. An errors in variables measurement error
adjustment is applied as an intermediate step in reaching this goal. The method may
be described as follows:
The first step involves fitting an errors in variables measurement error adjusted
multimedia exposure model that assumes blood-lead concentration is a function of
true unobserved floor and window sill dust lead loadings along with other
covariates used in the model. While the dependence of blood-lead concentration
on true dust lead loadings and other covariates can not be observed, it is easy to
estimate via equations (2.2.12) and (2.2.16) in FullerfMeasurement Error Models.
1987). In order to use these equations for estimating this relationship, the
measurement error associated with each particular dust lead loading must be
obtained. This is achieved by taking individual measurements of dust lead
loadings within households and calculating their variability. The average of all
within household variances is then used as an estimate of the true measurement
error associated with each particular dust lead loading. The estimated
measurement errors are then used to calculate parameter estimates for a model
based on Rochester data that relates blood-lead concentration to true dust lead
loadings and other covariates. Keep in mind that the model must be developed
using Rochester data because there is no blood-lead concentration variable in the
HUD data set.
If the goal had been to identify the nature of the dependence of blood-lead
concentration on true floor and window sill dust lead loadings and other
covariates, then the adjustment described above would have been all that was
required. However, the relationship of interest is blood-lead concentration as a
function of floor and window sill dust lead loadings and other covariates as
observed in the HUD National Survey. Therefore, adjusting for measurement
error is only the first step toward a final solution to this problem.
The next step in this process is to define the relationship between blood-lead
concentrations, observed floor and window sill dust lead loadings in both HUD
and Rochester, true floor and window sill dust lead loadings, and any other
covariates in the multimedia exposure model. It is assumed that, in
conjunction, these random variables have a multivariate normal distribution.
Once this assumption is made, standard statistical theory allows for deriving
the distribution of blood-lead concentration conditioned on floor and window
sill dust lead loadings and other covariates as measured in HUD. Estimates of
the parameters for a multimedia exposure model that relates blood-lead
concentration to lead exposures as measured in the HUD National Survey are
then obtained from this conditional distribution.
Draft - Do Not Cite or Quote 5 September 25. 1996
-------
An assessment of the above two statistical approaches was made based on a preliminary
version of the EPI Model which included only one variable (floor dust lead loading) in which
there was a difference in sample collection methods between the Rochester Study and the HUD
National Survey. The results of this assessment demonstrated that for this preliminary version of
the EPI Model, there was not a large difference between the two statistical approaches when used
to predict the national distribution of children's blood-lead levels. Based on the results of this
assessment, and taking into consideration time constraints, and the fact that the first approach
was much easier to implement, a decision was made to implement the first approach.
4.0 APPLICATION OF THE EPI MODEL TO HUD NATIONAL SURVEY
Environmental conditions observed in the National Survey were used as input to the EPI
model for predicting blood-lead levels in children 1-2 years old. As noted in Table 1, the
variables used for prediction were drip-line soil lead concentration, dust-lead loading on floors
(carpeted or uncarpeted), dust-lead loading on window sills, and an indicator of paint/pica
hazard. The EPI model was employed to estimate an average log-transformed childhood blood-
lead concentration associated with each home in the HUD National Survey.
A population of children (aged 1-2 years) associated with environmental lead levels found
at each home in the HUD National Survey was constructed using weights from the American
Housing Survey for 1997. The distribution of blood-lead concentrations associated with each
home can be characterized by assigning a geometric mean (predicted by the EPI model) and a
geometric standard deviation. A geometric standard deviation of 1 .6 was assumed for the
distribution of blood-lead concentrations associated with each home. The default geometric
standard deviation of blood-lead concentrations for children at similar environmental-lead levels
for the IEUBK model is 1.6 and the estimated variability about the Rochester prediction from the
regression model was 1 .66 as measured by the exponentiation of the root mean square error.
Therefore, the EPI model along with an assumed GSD of 1.6 was used to predict blood-
lead concentrations associated with specific housing conditions defined by the four
environmental-lead variables given in Table 1 .
Table 3 provides the partition of the distribution of log blood-lead concentrations into the
seven intervals about the log of the geometric mean. Figure 1 graphically illustrated this
partitioning. The two tails of the distribution represent log blood-lead concentrations below or
above 2.5 standard deviations from the mean, respectively. The percentage of the distribution
assigned to each of these intervals, 0.62%, is based on the area under a standard normal curve
for z-values less than -2.5 in the lower tail or greater than 2.5 in the upper tail. The assigned log
blood-lead concentration for the lower tail is the expected value of a standard normal random
deviate lying in the interval from - °° to -2.5; the assigned log blood-lead concentration was
similarly chosen for the upper tail, and mid-points were used for the finite-length intervals. The
assigned blood-lead concentration for each interval was obtained by exponentiating the assigned
log blood-lead for the interval. For example, for the lower tail,
n-2
-2 82«o =
GSD
282
Draft - Do Not Cite or Quote
September 25, 1996
-------
#
Distribution of Loo(Blood — lead levels)
Figure 1. Distribution of Blood-Lead Levels About Geometric Mean on Logarithmic Scale.
For this lower tail, if N children were associated with the specific housing condition then 0.62
percent of the N children were assigned a blood-lead concentration of GM/GSD2 82. The
remaining 99.28 percent were similarly assigned to the other blood-lead concentrations presented
in Table 3 using the percentages given in the second column of the table. In this manner, the
distribution of blood-lead concentrations of the N children were allocated to a distribution of
blood-lead concentrations centered around the GM predicted by the EPI model with a GSD of
1.6. The predicted distributions at each housing condition were then combined to generate a
distribution of childhood blood-lead levels over all of the housing conditions present in the HUD
National Survey.
5.0 RESULTS OF THE COMPARISON WITH NHANES III
The predicted distribution of blood-lead concentrations obtained by applying the EPI
model to the HUD National Survey Data was compared to NHANES III to measure how well
the EPI model performed. Table 4 contains characteristics of the predicted blood-lead
distribution for the EPI model, including estimates of exceedance proportions (the estimated
proportion of blood-lead concentration exceeding 10, 20 or 30 ug/dL), the geometric mean, and
the geometric standard deviation. Results in Table 4 for both the NHANES III distribution and
the EPI model are presented first with exceedance proportions calculated from the emprical
distribution and second for exceedance proportions calculated assuming a lognormal distribution
with the calculated geometric mean and geometric standard deviation.
Draft - Do Not Cite or Quote
September 25. 1996
-------
Table 3. Allocation of Blood-Lead Distribution to Seven Intervals
Log Blood-Lead Concentrations
Interval for LOB Blood Lead*
[-«o, 11 - 2.5 • 0]
(0-2.5 • 0,0- 1.5 * 0]
(0-1.5 • 0,0 -0.5 *0|
(0-0.5 • 0.0 + 0.5 *0]
(0 + 0.5 '0,0 + 1.5 '0]
(0 + 1 .5 * 0, 0 + 2.5 • 0]
(0 + 2.5 ' 0, + ~]
Percentage of
Distribution in Interval
0.0062
0.0606
0.2417
0.3830
0.2417
0.0606
0.0062
Assigned Log Blood
Lead for Interval
0-2.82 *ob
0 - 2.00 * 0
0 - 1 .00 • 0
0
0 + 1 .00 * 0
0 + 2.00 * 0
0 + 2.82 •Oe
Assigned Blood-Lead
Concontration for Interval
GM/IGSD282]
GM/IGSD200]
GM/IGSD1 °°1
GM
GM"[GSDIOOJ
GM*[GSD200]
GM'IGSD2'82]
Blood-lead concentrations were assumed to have a tog-normal distribution with the geometric mean (GM) predicted by the
EPI model and a geometric standard deviation (GSD) of 1.6 (the default geometric standard deviation for the IEUBK modal).
The distribution of log blood-lead concentrations was assumed to be normal with mean a given by log(GM) and standard
deviation o given by log (GSD = 1.6).
" The expected value of a normal random deviate known to lie in the interval [—, -2.5) is -2.82.
' The expected value of a normal random deviate known to lie in the interval (2.5. + °»] is +2.82.
Table 4. Predicted Distribution Characteristics for EPI Model compared to Rochester and
NHANES III
Model Results
(Empirical
Distribution)
NHANES III
(ages 1 -2 yrs)
Rochester Study data
EPI Model
Model Results
(Lognormal Distri-
bution Assumed)
NHANES III
{ages 1-2 yrs)
EPI Model
Geometric
Mean (0g/dL)
4.05
6.36
4.10
Geometric
Mean (0fl/dL)
4.05
4.10
Geometric Standard
Deviation (09/dL)
2.06
1.85
1.81
Geometric Standard
Deviation (0g/dL)
2.06
1.81
# > 10
00,/dL1
1 1 .04%
22.90%
7.03%
# > 10
lig/dL1
10.48%
6.65%
# > 20
j/g/dL1
1.74%
2.90%
0.56%
# > 20
09/dL1
1.34%
0.38%
#> 30
00/dL1
0.42%
1.00%
0.06%
#> 30
08/dL1
0.27%
0.04%
A histogram of the NHANES III distribution is displayed in Figure 2, above the
corresponding distribution of blood-lead concentrations predicted by the EPI Model from the
HUD National Survey Data (1-2 year old children).
Draft - Do Not Cite or Quote
September 25, 1996
-------
1O.O -i
8 1O 12 14 16 18 2O 22
Blood — Lead Concentration (
24 26 28 3O 32
NHANES III (1-2 year old children)
1O 1Z
10 10 20 22 2* 20 28 3O 33
Blood — Lead Concentration (
HUD National Survey Data as Predicted from EPI Model
Figure 2. Distribution of Blood-Lead Concentrations (/vg/dL) for NHANES III and
as Predicted by the EPI Model Applied to the HUD National Survey
Data(1-2 year old children)."
8 The Percentage of children in NHANES III with blood-lead concentrations greater than 32 //g/dL
was 0.2%; for predictions from the EPI Model, the percentage of children with blood-lead
concentrations exceeding 32/vg/dL was 0.04%.
Draft - Do Not Cite or Quote
September 25. 1996
-------
6.0 DISCUSSION
The results of the comparison with NHANES III for the revised EPI model indicate:
• The national geometric mean blood-lead concentration can be adequately predicted
by the EPI model applied to the HUD National Survey. For example, the geometric
mean of 4.10 predicted by the EPI Model in Table 4 is comparable to that reported in
NHANES III.
• The variability about the predicted values in the Rochester data set (as measured by
the regression model RMSE) is 1.66 GSD, which is similar to previous estimates in
the IEUBK documentation of variability in a homogeneous population of children
with similar exposure scenarios.
• The variability in the national distribution of blood-lead concentration predicted by
the EPI model using the HUD National Survey is approximately 1.81 (GSD), in
contrast to a GSD of 2.06 for NHANES III.
• The estimated proportions of blood-lead concentrations exceeding 10,20, or 30
ug/dL using the EPI model predictions are lower than the corresponding proportions
estimated by NHANES III.
It may be the case that a multi-exposure model developed using data from the Rochester
Study does not fully capture the variability in blood-lead or environmental lead-levels that are
present nationwide. Differences between the Rochester study population and the national
population include the following:
a. Almost one-quarter (22.9%) of the Rochester children had observed blood-lead
concentrations above 10 ug/dL, whereas only 11.5% of children nationwide were
estimated to have blood-lead concentrations above 10 ug/dL by NHANES III.
b. The geometric mean blood-lead concentration in Rochester is 6.36, whereas the
geometric mean blood-lead concentration nationwide as estimated by NHANES III is
4.05. The GSD for Rochester is 1.85, compared to 2.06 for NHANES III.
c. Approximately 84 percent of the housing included in the Rochester study was built
prior to 1940, and there is a well documented relationship between age of housing
and presence of lead-based paint. Only 20% of housing nationwide was built prior to
1940.
d. Approximately 40% of the sample of children in the Rochester study were African
Americans, compared to an estimated 13% of the population of children nationwide
(from 1997 US Census Projections), and compared to approximately 7% in the HUD
National Survey.
Draft - Do Not Cite or Quote 10 September 25, 1996
-------
e. Environmental levels of lead in soil in the Rochester study were higher than would
be expected in the HUD National Survey. For example, the geometric mean drip-
line soil-lead concentration in the HUD National Survey was approximately 75 ppm
whereas the Rochester geometric mean was approximately 730 ppm.
f. Subjects recruited into the Rochester Study represent children whose primary
exposure to lead was from dust, soil and paint at the primary residence. Children
whose parents had lead exposure, who spent time away from the home, or whose
homes underwent renovation or remodeling were excluded from the study. Only 376
of 1,536 families were eligible to participate in the study after the initial telephone
screening. The selection criteria utilized in the Rochester Study may have resulted in
a biased sample of children, since children who had potential lead exposure outside
of the primary residence were excluded.
The difference in the observed blood-lead distributions between the Rochester study and
NHANES HI is illustrated in Figure 3. Although there are limitations associated with the
Rochester Study, there are also positive aspects of the study that recommend its use:
a. all media, locations, and surfaces that are being considered for Section 403 standards
were measured for lead in the Rochester study.
b. the Rochester study includes dust-lead loadings from wipe sampling and the Section
403 dust standard is expected to be based on dust-lead loading from wipe sampling.
c. the selection of homes and children in the Rochester study, although targeted, was
more random and more representative of a general population than is the case with
most recent epidemiological studies of lead exposure in non-smelter communities.
Other factors that may reflect the ability of any EPI model to match the national distribution of
blood-lead concentrations estimated by NHANES HI include:
a. The EPI model accounts only for factors related to environmental lead exposures at
the residence, and does not account for other factors that might affect childhood
blood lead. Such factors that may affect children's blood-lead concentration but may
not be able to be controlled by the Section 403 rule include:
(1) home and personal cleaning habits,
(2) diet and nutritional status,
(3) bio-availability of the lead found in residential environmental media,
(4) non-residential exposures,
(5) inhalation exposure,
(6) children's behavior,
(7) socio-economic factors,
(8) renovation and remodeling (R&R) activity,
(9) hobbies,
(10) occupation.
Draft - Do Not Cite or Quote 11 September 25, 1996
-------
b. It should also be noted that NHANES III itself is only an estimate of the true national
distribution of blood-lead concentrations, and that an "exact" match of NHANES III
does not mean an exact match of the true national distribution.
Finally, it should be noted that the EPI Model contains variables that differ from variables
created for a best-fit of the Rochester data, because the goal of the EPI Model was to provide a
basis for using measures of lead from the HUD National Survey to predict a national distribution
of childhood blood-lead concentrations. In particular, the EPI model differs from the multimedia
regression model used to characterize the dose-response relationship between environmental-lead
and blood-lead.
100.0
10.0
1.0
0.1
NHANES III
White
and Other
Rochester
White
and Other
NHANES III
Black
Rochester
Black
Figure 3. Box Plot of NHANES III versus Rochester Datasets.
Draft - Do Not Cite or Quote
12
September 25, 1996
-------
September 17, 1996
CONVERSIONS EQUATIONS FOR USE IN
THE SECTION 403 RULEMAKING
APPENDIX X
Note to EPA: Draft Report is currently undergoing peer review
-------
DISCLAIMER
The material in this document has not been subject to Agency
technical and policy review. Views expressed by the authors are their own
and do not necessarily reflect those of the U.S. Environmental Protection
Agency. Mention of trade names, products, or services does not convey,
and should not be interpreted as conveying, official EPA approval,
endorsement, or recommendation. Do not quote or cite this document.
This report is copied on recycled paper.
Draft -Do Note Cite or Quote '» September 17. 1996
-------
AUTHORS AND CONTRIBUTORS
This study was funded and managed by the U.S. Environmental Protection Agency. The
analysis was conducted by Battelle Memorial Institute under contract to the Environmental
Protection Agency. Each organization's responsibilities are listed below.
Battelle Memorial Institute (Battelle)
Battelle was responsible for identifying the relevant studies and obtaining the data, fitting
models to the data from the individual studies, developing an approach for combining data across
studies, and developing final conversion equations.
U.S. Environmental Protection Agency (EPA)
The Environmental Protection Agency was responsible for managing the review,
providing guidance on the objectives for the review and report, contributing to the development
of conclusions and recommendations, and coordinating the EPA and peer reviews of the draft
report. In addition, EPA provided access to study results not yet available in the general
literature. The EPA Work Assignment Managers were John Schwemberger, Janet Remmers, and
Bradley Schultz. The EPA Project Officer was Sineta Wooten.
Draft -Do Note Cite or Quote iii September 17, 1996
-------
EXECUTIVE SUMMARY
1 . What Does This Report Say?
This report presents equations used to carry out various conversions between wipe lead
loading and vacuum lead loading and concentrations, based on two different vacuum samplers,
the Blue Nozzle vacuum (BN), used in the HUD National Survey, and the HVS3 vacuum used in
the Baltimore Repair and Maintenance Study. The HVS3 vacuum will be referred to as a BRM
vacuum for the remainder of this report.
Three sets of equations are described using the BN vacuum sampler: 1) for converting a
wipe lead loading to a BN lead loading, 2) for converting a wipe lead loading to a BN lead
concentration, and 3) for converting a BN lead loading to a wipe lead loading. Each set contains
a separate equation for samples collected from uncarpeted floors, window sills, and window
wells.
The following equations were developed for converting a wipe lead loading (ug/ft2) to a
BN lead loading (ug/ft2):
Uncarpeted floors: BNload = 0.131 Wipe1 °3
Window sills: BNload = 0.828 Wipe0 6I3
Window wells: BNload = 5.63 Wipe0 4"
Thus, for example, a wipe lead loading of 100 ug/ft2 on an uncarpeted floor would be converted
to a BN lead loading of 14.9 ug/ft2, with an approximate 95% confidence interval of 10.6 to 21 .0
ug/ft2, by applying the first of the above equations. On the other hand, 95% of the BN lead
loadings associated with a wipe lead loading of 100 ug/ft2 are predicted to be between 2.13 and
105 ug/ft2. The difference in interpretation of a confidence interval and a prediction interval is
discussed in Section 3.3.
The following equations were developed for converting a wipe lead loading (ug/ft2) to a
BN lead concentration (jig/g):
Uncarpeted floors: BN^- 34.2 Wipe0613
Draft -Do Note Cite or Quote iv September 17, 1996
-------
Window sills: BN,^ = 115 Wipe0451
Window wells: BNconc = 23.7 Wipe0673
Thus, a wipe lead loading of 100 ug/ft2 on an uncarpeted floor would be converted to a BN lead
concentration of 576 ug/g, with an approximate 95% confidence interval of 459 to 723 ug/g, by
applying the first of the above equations. However, 95% of the BN lead concentrations
associated with a wipe lead loading of 100 jig/ft2 are predicted to be between 166 and 2000 ng/g.
For converting a BN lead loading (fig/ft2) to a wipe lead loading (fig/ft2), the following
equations were developed:
Uncarpeted floors: Wipe = 11.4 BN0690
Window sills: Wipe = 5.79 BN1 °8
Window wells: Wipe = 7.37 BN°752
Thus, a BN lead loading of 100 fig/ft2 on an uncarpeted floor would be converted to a wipe lead
loading of 273 fig/ft2, with an approximate 95% confidence interval of 191 to 388 ng/ft2, by
applying the first of these equations. Alternatively, 95% of the wipe lead loadings associated
with a BN loading of 100 jig/ft2 are predicted to be between 50.7 and 1,460 ug/ft2.
For samples collected with the BRM vacuum sampler, from uncarpeted floors, carpeted
floors, window sills, and window wells, the following equations were developed for converting a
BRM lead loading (ug/ft2) to a wipe lead loading (fig/ft2):
Uncarpeted floors: Wipe = 8.79 BRM0313.
Carpeted floors: Wipe = 2.21 BRM0271.
Window sills: Wipe = 17.0 BRM0421.
Window wells: Wipe = 14.1 BRM0571.
For example, a BRM lead loading of 100 ug/ft2 on an uncarpeted floor would be converted to a
wipe lead loading of 37.1 jig/ft2, with an approximate 95% confidence interval of 33.3 to 41.6
Draft -Do Note Cite or Quote v September 17, 1996
-------
Hg/ft2 by applying the first equation. However, 95% of the wipe loadings associated with a BRM
loading of 100 ug/ft2 are predicted to be between 4.20 and 329 ug/ft2.
2. Why Was This Report Written?
This report was written for three reasons, all related to the determination of Section 403
standards. First, it is likely that the Section 403 standards for lead-contaminated dust will be
expressed as a measured lead loading collected by a dust wipe sample. In considering different
options for this standard, it is important to evaluate the number of homes that would be affected
by the different options. The HUD National Survey of pre-1980 housing (the only national
survey of dust lead levels) is the best source for making this assessment. However, the BN
vacuum was used in the National Survey to collect dust samples. Therefore, in order to use this
data appropriately, it is necessary to convert the raw BN lead loading data to wipe lead loadings
or to convert options for standards based on wipe samples to equivalent standards for the BN
vacuum.
The second reason this report was written is that the IEUBK model may be used to
predict post-intervention blood-lead levels. To predict post-intervention blood lead using the
IEUBK model, it is required that post-intervention dust-lead concentrations, and not dust lead
loadings, be provided as input. However, Section 403 analyses (risk assessment and economic
analysis) require specification of various options of wipe lead loadings in a residence following
an intervention. Therefore to obtain IEUBK model input, an equation is presented to convert
wipe lead loadings to equivalent dust concentrations of samples collected via the Blue Nozzle
vacuum.
Thirdly, because Section 403 dust-lead standards will likely be expressed in terms of
wipe loadings, the analyses upon which these standards are based must utilize wipe loading
information. Since the Baltimore Repair and Maintenance study dust samples were collected
using a BRM vacuum, a conversion to a wipe lead loading is necessary in order for this data to
be applicable to Section 403 analyses, such as sensitivity/specificity analyses and prevalence
statistics.
Draft -Do Note Cite or Quote vi September 17, 1996
-------
3. How Large is the Uncertainty Associated With These Conversion Equations?
There is a considerable degree of uncertainty in the conversion equations based on BN
vacuum samples. For the BN vacuum to wipe conversion, there is relatively little data. For
example, on uncarpeted floors, one field study produced six pairs of side-by-side wipe and
vacuum measures, another produced seven pairs, and a third produced 24 pairs. However, a
large amount of data was available to develop the conversion equations based on BRM vacuum
samples. The Rochester Lead-in-Dust study alone provided almost 400 BRM and wipe pairs on
each housing component. Although this large amount of data allows us to fairly accurately
characterize the relationship between the geometric average wipe lead loading and an observed
BRM lead loading, the inherent variability in wipe measures makes it important to recognize the
wide range of plausible wipe lead loadings that could be associated with any observed BRM lead
loading. Confidence and prediction intervals around point estimates, which can be used in
sensitivity analyses, are presented in Tables 12,13, and 14 for selected nominal BN levels and in
Table 17 for selected nominal BRM levels. The intervals around the point estimates from the
BN relationships are quite wide.
4. How Will These Conversion Equations Be Used?
The BN to wipe conversion equations will be used to convert the BN dust-lead loadings
measured in the National Survey to equivalent lead loadings for wipe samples. The transformed
lead loadings will then be used to determine the numbers and percentages of houses that would
be affected for various options for defining what is lead-contaminated dust. Alternatively, the
BN to wipe conversion equations can be inverted to convert the proposed Section 403 standards
(based on wipe loadings in ug/ft2) to equivalent vacuum loadings (ug/ft2), thereby allowing the
original data to be used in determining the number of homes affected by the Section 403 ruling.
Note that this latter approach is not the same as using the wipe to BN lead loading conversion
equations. At present, those equations are not going to be used in the development or assessment
of Section 403.
A second exercise will be undertaken to assess the impact of the Section 403 standard on
children's blood-lead levels. Blood-lead levels will be predicted for the post-Section 403
environment. This may be done in more than one way, but it will certainly be done using the
Draft -Do Note Cite or Quote vii September 17, 1996
-------
IEUBK. model. For this exercise, it will be assumed that floor lead loadings measured by wipe
samples will be equal to, say, 40 ug/ft2 after an intervention conducted in response to the 403
rule in all units that exceed the 403 standard for floors. The baseline number of 40 fig/ft2 (or
whatever it may eventually be) will be converted to an equivalent dust-lead concentration for a
National Survey vacuum sample using the wipe loading (ug/ft2) to BN concentration (jig/g)
conversion equation. The estimated equivalent National Survey dust-lead concentration is 328
ug/g, with an approximate 95% confidence interval of 255 to 423 ug/g. It should be noted that
95% of the BN dust-lead concentrations associated with a wipe dust-lead loading of 40 ug/ft2 are
predicted to be between 94.2 and 1,140 ug/g, reflecting the wide variability in typical dust-lead
concentrations. This converted concentration (328 ug/g) will be assumed in the houses which
trigger an intervention based on their measured pre-intervention dust-lead loading, and it will be
used as input to the IEUBK model to predict blood-lead levels. The model will then be used to
determine the resulting distribution of children's blood-lead levels and inferred health effects in
the nation. Comparisons will be made to the prior distribution of children's blood-lead levels (as
determined from the NHANES III Survey) to determine the benefits of the Section 403 rule.
Similarly, the BRM to wipe conversion equations will be used to transform the BRM
vacuum lead loadings in the Baltimore Repair and Maintenance study to equivalent wipe lead
loadings for use in determining prevalence statistics and in completing a sensitivity/specificity
analysis which relates the incidence of elevated (* 10 ug/dL) children's blood-lead levels to wipe
lead loadings.
Draft-Do Note Cite or Quote viii September 17, 1996
-------
TABLE OF CONTENTS
Page
EXECUTIVE SUMMARY iv
1. What Does This Report Say? iv
2. Why Was This Report Written? vi
3. How Large is the Uncertainty Associated With These Conversion Equations? vii
4. How Will These Conversion Equations Be Used? vii
1.0 INTRODUCTION 1
2.0 DATA 3
2.1 STUDIES INCLUDED 3
2.2 DESCRIPTIVE STATISTICS 7
2.3 RANGE ISSUES 7
2.3.1 Range of Data 7
2.3.2 Abated versus Unabated 11
2.4 OTHER ISSUES 13
3.0 STATISTICAL APPROACH 17
3.1 MEASUREMENT ERROR 18
3.2 COMBINING RESULTS ACROSS STUDIES 19
3.3 CONFIDENCE AND PREDICTION INTERVALS 21
4.0 RESULTS 23
4.1 RESULTS OF STATISTICAL ANALYSES FOR THE BLUE NOZZLE VACUUM 23
4.2 RESULTS OF STATISTICAL ANALYSES FOR THE BRM VACUUM 25
5.0 CONCLUSIONS AND DISCUSSION 60
APPENDIX A A-1
Distribution of Data A-2
APPENDIX B B-1
Residual Analysis and Influential Observations for the BN-Wipe Conversion Equations . . B-2
APPENDIX C C-1
Simple Regression Results Based on Data Pooled Across Studies C-2
APPENDIX D D-1
Residual Analysis and Influential Observations for the BRM/Wipe Conversion
Equations D-2
Draft-Do Note Cite or Quote ix September 17, 1996
-------
LIST OF TABLES
Page
Table 1. Summary Table for the Studies from which Data were Used to Develop
Conversion Equations 4
Table 2. Descriptive Statistics for Vacuum Dust Lead Measures and Wipe Lead Loading
by Study 8
Table 3. Data Ranges by Housing Component and Vacuum Type for the Data Used to
Develop the Vacuum Loading to Wipe Loading Conversion Equations and for the
Data upon Which the Conversion Equations Will Be Used 11
Table 4. Wipes Used in Various Studies 15
Table 5. Information Related to Non-detection for Lead Measures in Various Studies 16
Table 6. Regression Equations for Predicting Blue Nozzle Vacuum Lead Loading
from Wipe Lead Loading 26
Table 7. Regression Equations for Predicting Blue Nozzle Vacuum Lead Concentration
From Wipe Lead Loading 27
Table 8. Regression Equations for Predicting Wipe Lead Loading from Blue Nozzle
Vacuum Lead Loading 28
Table 9. Final Conversion Equations for Uncarpeted Floors Based on Combined Data 28
Table 10. Final Conversion Equations for Window Sills Based on Combined Data 29
Table 11. Final Conversion Equations for Window Wells Based on Combined Data 29
Table 12. Predicted Vacuum Lead Loadings and Concentrations For Selected Wipe Lead
Loadings and Predicted Wipe Lead Loadings for Selected Blue Nozzle Lead
Loadings Based on Weighted Average of Regression Coefficients on Uncarpeted
Floors 30
Table 13. Predicted Vacuum Lead Loadings and Concentrations For Selected Wipe Lead
Loadings and Predicted Wipe Lead Loadings for Selected Blue Nozzle Lead
Loadings Based on Weighted Average of Regression Coefficients on Window Sills. . 31
Table 14. Predicted Vacuum Lead Loadings and Concentrations For Selected Wipe Lead
Loadings and Predicted Wipe Lead Loadings for Selected Blue Nozzle Lead Loadings
Based on Weighted Average of Regression Coefficients on Window Wells 32
Table 15. Regression Equations for Predicting Wipe Lead Loading From BRM Vacuum
Lead Loading 33
Table 16. Final Conversion Equations Based on Combined Data for Uncarpeted Floors,
Carpeted Floors, Window Sills, and Window Wells 34
Table 17. Predicted Wipe Lead Loadings Based on Final Conversion Equations For Selected
BRM Vacuum Lead Loadings 34
Table A-1. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading CAPS
Pilot Study — Uncarpeted Floors A-2
Table A-2. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading, CAPS
Pilot Study — Uncarpeted Floors and Sills A-2
Table A-3. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading, R&M
Pilot Study — Uncarpeted Floors A-3
Table A-4. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading, R&M
Pilot Study — Uncarpeted Floors, Sills, and Wells A-3
Table A-5. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading,
NCLSH/Westat Study — Uncarpeted Floors A-4
Table A-6. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading,
NCLSH/Westat Study — Uncarpeted Floors, Sills, and Wells A-4
Table A-7. Number of Samples by Wipe Lead Loading and Vacuum Lead Concentration,
CAPS Pilot Study — Uncarpeted Floors A-5
Draft -Do Note Cite or Quote
September 17, 1996
-------
LIST OF TABLES
(Continued)
Page
Table A-8. Number of Samples by Wipe Lead Loading and Vacuum Lead Concentration,
CAPS Pilot Study - Uncarpeted Floors and Sills A-5
Table A-9. Number of Samples by Wipe Lead Loading and Vacuum Lead Concentration,
R&M Pilot Study — Uncarpeted Floors A-6
Table A-10. Number of Samples by Wipe Lead Loading and Vacuum Lead Concentration,
R&M Pilot Study - Uncarpeted Floors, Sills, and Wells A-6
Table C-1. Regression Equations for Predicting Wipe Lead Loading From BRM Vacuum Lead
Loading based on Data Pooled Across Studies C-2
LIST OF FIGURES
Figure 1. Wipe Loadings (//g/ft2) as Predicted from Blue Nozzle Vacuum Loadings (//g/ft2)
versus Blue Nozzle Vacuum Loadings with 95% Confidence and Prediction
Intervals Superimposed for Uncarpeted Floors, Window Sills, and Window Wells
from the HU 9
Figure 2. Wipe Loadings (//g/ft2) as Predicted from BRM Vacuum Loadings (//g/ft2) versus BRM
Vacuum Loadings with 95% Confidence and Prediction Intervals Superimposed for
Uncarpeted Floors, Window Sills, and Window Wells from the Baltimore Repair and
Maintenance Survey 12
Figure 3. Blue Nozzle Vacuum Concentrations versus Wipe Lead Loading Displaying the
Distribution of Abated and Unabated Homes Across the CAPS Pilot and R&M Pilot
Data Used to Develop the Conversion Equations for Uncarpeted Floors, Window
Sills, and Window Wells 14
Figure 4. Blue Nozzle Vacuum Lead Loading versus Wipe Lead Loading for Uncarpeted
Floors and Window Sills from the CAPS Pilot Study 34
Figure 5. Blue Nozzle Vacuum Lead Loading versus Wipe Lead Loading for Uncarpeted
Floors, Window Sills, and Window Wells from the NCLSH/Westat Study 36
Figure 6. Blue Nozzle Vacuum Lead Loading versus Wipe Lead Loading for Uncarpeted
Floors, Window Sills, and Window Wells from the R&M Pilot Study 37
Figure 7. Wipe Lead Loading versus Blue Nozzle Lead Loading for Uncarpeted Floors and
Window Sills from the CAPS Pilot Study 38
Figure 8. Wipe Lead Loading versus Blue Nozzle Vacuum Lead Loading for Uncarpeted
Floors, Window Sills, and Window Wells from the NCLSH/Westat Study 39
Figure 9. Wipe Lead Loading versus Blue Nozzle Vacuum Lead Loading for Uncarpeted
Floors, Window Sills, and Window Wells from the R&M Pilot Study 40
Figure 10. Blue Nozzle Vacuum Lead Concentration versus Wipe Lead Loading for
Uncarpeted Floors and Window Sills from the CAPS Pilot Study 41
Figure 11. Blue Nozzle Lead Concentration versus Wipe Lead Loading for Uncarpeted
Floors, Window Sills, and Window Wells from the R&M Pilot Study 42
Figure 12. Predicted BN Lead Loading Versus Wipe Lead Loading on Uncarpeted Floors,
Individual and Combined Regressions. Confidence Bands for Combined
Regressions and Data Overlaid 43
Figure 13. Predicted BN Lead Concentration Versus Wipe Lead Loading on Uncarpeted Floors,
Individual and Combined Regressions. Confidence Bands for Combined
Regressions and Data Overlaid • 44
Figure 14. Predicted Wipe Lead Loading Versus BN Lead Loading on Uncarpeted Floors,
Individual and Combined Regressions. Confidence Bands for Combined
Regressions and Data Overlaid 45
Draft -Do Note Cite or Quote
XI
September 17, 1996
-------
LIST OF FIGURES
(Continued)
Page
Figure 15. Predicted BN Lead Loading Versus Wipe Lead Loading on Window Sills, Individual
and Combined Regressions. Confidence Bands for Combined Regressions and
Data Overlaid 46
Figure 16. Predicted BN Lead Concentration Versus Wipe Lead Loading on Window Sills,
Individual and Combined Regressions. Confidence Bands for Combined
Regressions and Data Overlaid 47
Figure 17. Predicted Wipe Lead Loading Versus BN Lead Loading on Window Sills, Individual
and Combined Regressions. Confidence Bands for Combined Regressions and
Data Overlaid 48
Figure 18. Predicted BN Lead Concentration Versus Wipe Lead Loading on Window Wells,
Individual and Combined Regressions. Confidence Bands for Combined
Regressions and Data Overlaid 49
Figure 19. Predicted BN Lead Concentration Versus Wipe Lead Loading on Window Wells,
Individual and Combined Regressions. Confidence Bands for Combined
Regressions and Data Overlaid 50
Figure 20. Predicted Wipe Lead Loading Versus BN Lead Loading on Window Wells, Individual
and Combined Regressions. Confidence Bands for Combined Regressions and
Data Overlaid 51
Figure 21. Wipe Lead Loadings versus BRM Vacuum Lead Loadings for Uncarpeted Floors,
Window Sills, and Window Wells from the R&M Mini Survey 52
Figure 22. Wipe Lead Loading versus BRM Vacuum Lead Loading for Uncarpeted Floors
and Carpeted Floors from the NCLSH 5-Method Comparison Study 53
Figure 23. Wipe Lead Loadings Versus BRM Vacuum Lead Loadings for Uncarpeted Floors,
Carpeted Floors, Window Sills, and Window Wells from the Rochester Lead-in-
Dust Survey 54
Figure B-1. Influential Observations in the Regression of Blue Nozzle Vacuum Lead Loading on
Wipe Lead Loading for Uncarpeted Floors from CAPS Pilot, NCLSH/Westat, and
R&M Pilot Studies B-5
Figure B-2. Residual (Log Scale) versus Predicted Blue Nozzle Vacuum Lead Loading from
the Regression of Blue Nozzle Lead Loading on Wipe Lead Loading for Uncarpeted
Floors from CAPS Pilot, NCLSH/Westat, and R&M Pilot Studies B-6
Figure B-3. Influential Observations in the Regression of Wipe Lead Loading on Blue Nozzle
Vacuum Lead Loading for Uncarpeted Floors from CAPS Pilot, NCLSH/Westat,
and R&M Pilot Studies B-7
Figure B-4. Residual (Log Scale) versus Predicted Blue Nozzle Vacuum Lead Loading from the
Regression of Wipe Lead Loading on Blue Nozzle Lead Loading for Uncarpeted Floors
from CAPS Pilot, NCLSH/Westat, and R&M Pilot Studies B-8
Figure B-5. Influential Observations in the Regression of Blue Nozzle Lead Concentration on
Wipe Lead Loading for Uncarpeted Floors from the R&M Pilot and CAPS Pilot
Studies B-9
Figure B-6. Residual (Log Scale) and Predicted Blue Nozzle Lead Concentration from the
Regression of Blue Nozzle Vacuum Lead Concentration on Wipe Lead Loading
for Uncarpeted Floors from the R&M Pilot and CAPS Pilot Studies B-10
Figure B-7. Influential Observations in the Regression of Blue Nozzle Vacuum Lead Loading
on Wipe Lead Loading for Window Sills from the CAPS Pilot, NCLSH/Westat,
and R&M Pilot Studies B-11
Draft-Do Note Cite or Quote xii September 17. 1996
-------
LIST OF FIGURES
(Continued)
Page
Figure B-8. Residual (Log Scale) and Predicted Blue Nozzle Vacuum Lead Loading from the
Regression of Blue Nozzle Lead Loading of Wipe Lead Loading for Window Sills
from the CAPS Pilot, NCLSH/Westat, and R&M Pilot Studies B-12
Figure B-9. Influential Observations in the Regression of Wipe Lead Loading on Blue Nozzle
Vacuum Lead Loading for Window Sills from the CAPS Pilot, NCLSH/Westat,
and R&M Pilot Studies B-13
Figure B-10. Residual (Log-Scale) and Predicted Wipe Lead Loading from the Regression of
Wipe Lead Loading on Blue Nozzle Vacuum Lead Loading for Window Sills from
the CAPS Pilot, NCLSH/Westat, and R&M Pilot Studies B-14
Figure B-11. Influential Observations in the Regression of Blue Nozzle Vacuum Lead
Concentration on Wipe Lead Loading for Window Sills from the R&M Pilot and
CAPS Pilot Studies B-15
Figure B-12. Residual (Log Scale) and Predicted Blue Nozzle Lead Concentration from the
Regression of Blue Nozzle Vacuum Lead Concentration on Wipe Lead Loading
for Window Sills from the R&M Pilot and CAPS Pilot Studies B-16
Figure B-13. Influential Observations in the Regression of Blue Nozzle Vacuum Lead Loading
on Wipe Lead Loading for Window Wells from the R&M Pilot and NCLSH/Westat
Studies B-17
Figure B-14. Residual (Log Scale) and Predicted Blue Nozzle Lead Loading from the Regression
of Blue Nozzle Vacuum Lead Loading on Wipe Lead Loading for Window Wells
from the R&M Pilot and NCLSH/Westat Studies B-18
Figure B-15. Influential Observations in the Regression of Wipe Lead Loading on Blue Nozzle
Vacuum Lead Loading for Window Wells from the R&M Pilot and NCLSH/Westat
Studies B-19
Figure B-16. Residual (Log Scale) and Predicted Wipe Lead Loading from the Regression of
Wipe Lead Loading on Blue Nozzle Vacuum Lead Loading for Window Wells from
the R&M Pilot and NCLSH/Westat Studies B-20
Figure B-17. Individual Observations in the Regression of Blue Nozzle Vacuum Lead
Concentration on Wipe Lead Loading for Window Wells from the R&M Pilot
Study B-21
Figure B-18. Residual (Log Scale) and Predicted Blue Nozzle Vacuum Lead Concentration
from the Regression of Blue Nozzle Vacuum Lead Concentration on Wipe Lead
Loading for Window Wells from the R&M Pilot Study B-22
Figure D-1. Influential Observations in the Regression of Wipe Lead Loading on BRM
Vacuum Lead Loading for Uncarpeted Floors from the R&M Mini, NCLSH
5-Method Comparison, Rochester Lead-in-Dust, and Milwaukee Low Cost
Interventions Studies D-4
Figure D-2. Residual (Log Scale) and Predicted Wipe Lead Loading from the Regression of
Wipe Lead Loading on BRM Vacuum Lead Loading for Uncarpeted Floors from
the R&M Mini, NCLSH 5-Method Comparison, Rochester Lead-in-Dust, and
Milwaukee Low Cost Interventions Studies D-5
Figure D-3. Influential Observations in the Regression of Wipe Lead Loading on BRM
Vacuum Lead Loading for Carpeted Floors from the Rochester Lead-in-Dust, and
NCLSH 5-Method Comparison Studies.. D-6
Figure D-4. Residual (Log Scale) and Predicted Wipe Lead Loading from the Regression of
Wipe Lead Loading on BRM Vacuum Lead Loading for Carpeted Floors from the
Rochester Lead-in-Dust, and NCLSH 5-Method Comparison Studies D-7
Draft -Do Note Cite or Quote xiii September 17. 1996
-------
LIST OF FIGURES
(Continued)
Page
Figure D-5. Influential Observations in the Regression of Wipe Lead Loading on BRM
Vacuum Lead Loading for Window Sills from the Rochester Lead-in-Dust, and
R&M Mini Studies D-8
Figure D-6. Residual (Log Scale) and Predicted Wipe Lead Loading from the Regression of
Wipe Lead Loading on BRM Vacuum Lead Loading for Window Sills from the
Rochester Lead-in-Dust, and R&M Mini Studies D-9
Figure D-7. Influential Observations in the Regression of Wipe Lead Loading on BRM
Vacuum Lead Loading for Window Wells from the Rochester Lead-in-Dust, and
R&M Mini Studies D-10
Figure D-8. Residual (Log Scale) and Predicted Wipe Lead Loading from the Regression of
Wipe Lead Loading on BRM Vacuum Lead Loading for Window Wells from the
Rochester Lead-in-Dust, and R&M Mini Studies D-11
Draft -Do Note Cite or Quote xiv September 17, 1996
-------
1.0 INTRODUCTION
The objective of this report is to present conversion equations for use in developing the
Section 403 Rule. Equations were developed based on two distinct vacuum samplers, the Blue
Nozzle (BN), used in the HUD National Survey, and the HVS3 (BRM) used in the Baltimore
Repair and Maintenance study. The first type of equation converts a dust-lead loading of a wipe
sample to an equivalent dust-lead loading of a vacuum sample collected via the Blue Nozzle
(BN) vacuum. The second type of equation converts a dust-lead loading of a wipe sample to an
equivalent dust-lead concentration of a BN vacuum sample. A third type of equation converts a
dust-lead loading measured by the BN sampler to a dust-lead loading as measured by a wipe.
Each of the conversion equations based on samples collected using the BN vacuum was
developed for uncarpeted floors, window sills, and window wells. A fourth set of conversion
equations transforms a BRM vacuum lead loading to an equivalent wipe lead loading for samples
collected from uncarpeted floors, carpeted floors, window sills, and window wells.
These equations were developed because the Section 403 standard for dust lead will
likely be defined in terms of a wipe lead loading. Therefore, any analysis associated with these
standards needs to be based on wipe lead loadings as well. The Economic Analysis and Risk
Assessment for Section 403 will use the HUD National Survey to estimate the numbers of
homes in the U.S. housing that have dust-lead levels above various options for the standard.
Dust samples were collected in that study using the BN sampler. Therefore, it will be necessary
to convert the National Survey Blue Nozzle lead loading data to wipe lead loadings or to convert
the different options for a standard from a wipe lead loading to an equivalent lead loading
standard for a dust sample collected via the BN vacuum.
Similarly, the Baltimore Repair and Maintenance data will be used to perform a
sensitivity/specificity analysis and to calculate prevalence statistics. Since the BRM vacuum was
used to collect data in this study, a means of converting the BRM lead loading to a wipe lead
loading is required.
Furthermore, the IEUBK model will be used to predict post-intervention blood-lead
levels. These predictions will be based on an assumed post-intervention dust-lead level from the
HUD National Survey, which will also be stated as a wipe lead loading. Because lead
concentrations are used as inputs to the IEUBK model, a conversion equation is required to
Draft -Do Note Cite or Quote 1 September 17, 1996
-------
bridge the gap between a wipe lead loading and a vacuum lead concentration. Following conduct
of a dust intervention, the Section 403 Economic Analysis and Risk Assessment will assume that
the dust lead is reduced to various options for the standard. The post-intervention level of dust
lead will be expressed in terms of a dust-lead loading of a wipe sample (e.g., 40 ug/ft2). The
second equation presented in this report will be used to convert the post-intervention dust-lead
loading based on a wipe sample to an appropriate dust-lead concentration based on a floor dust
sample collected by a BN vacuum sampler. This dust-lead concentration will then be used as
input to the IEUBK model to predict a childhood blood-lead concentration.
It may be noted that the relationship between true lead loading and true lead concentration
depends only on the amount of dust on the sampled surface. Therefore, the dust loading in the
homes used to develop the conversion equations are assumed to be similar to the dust loading in
the homes on which conversion equations will be applied.
Draft -Do Note Cite or Quote 2 September 17, 1996
-------
2.0 DATA
A wide range of samplers and sampling protocols have been used in environmental-lead
studies for the collection of dust samples. This section describes the data used for determining
relationships between lead loadings and lead concentrations as measured by the wipe and BN
methods. A description of the data used to determine the relationship between wipe and BRM
vacuum lead loadings is also presented.
2.1 STUDIES INCLUDED
Table 1 provides information on the studies used to develop the conversion equations
discussed in this report. Details regarding the design of each study and the intervention history
of the houses included in each study are presented. Comments are also provided to identify
limitations or special considerations, such as correction factors, associated with each study.
Three studies report side-by-side paired data on wipe lead loading and Blue Nozzle
vacuum lead loading and were used to develop the conversion equations for the BN vacuum:
1. CAP Pilot study
2. Baltimore Repair and Maintenance (R&M) Pilot study
3. National Center for Lead-Safe Housing (NCLSH)/Westat study
Of these, the NCLSH/Westat study did not have data on BN concentration. In the CAPS Pilot
and NCLSH/Westat studies, wipe samples were analyzed using the hot nitric acid/peroxide
digestion method typically used in HUD-related work, while the R&M Pilot study employed the
cold hydrochloric acid digestion procedure used in the State of Maryland. Therefore, it is
necessary to use an additional correction factor to convert the wipe lead measurements in the
R&M Pilot study to equivalent HUD wipe measurements. Wipe lead loadings in the R&M Pilot
study were multiplied by 1.58 to be more reflective of HUD wipe measurements. An estimate of
the correction factor was obtained from the NCLSH 5-Method Comparison study [5]. This study
reported wipe measurements that were analyzed by both chemical extraction procedures.
Draft -Do Note Cite or Quote 3 September 17. 1996
-------
Table 1. Summary Table for the Studies from which Data were Used to Develop Conversion Equations
Vacuum
Type
Study
Surface Type
IN)1"
Relevant Issues
of Study Design
Intervention History
Comments
9
R&M Pilot
Study (11
Uncarpeted Floors
Window Silts
Window Wells
(24)
(23)
(24)
Side-by-side wipe and BN
vacuum samples were
collected from uncarpeted
floors, window sills, and
window wells.
Of the 6 homes in the study, 2 homes were
occupied and had received comprehensive
abatement in 1986-87, 2 homes were
vacant, unabated, older urban homes, and
2 homes were vacant modern urban homes.
No description was provided for the type of
abatement performed on the two previously
abated homes No interventions were
performed as a part of the R&M Pilot Study
In the R&M Pilot study, the bioavailable digestion
method was used to determine the lead content ir
wipe samples. Since this report focuses on wipe
samples analyzed by the total digestion
procedure, a correction factor of 1 .SB was used
to adjust bioavailable lead to total lead.
The correction factor is based on the NCLSH S-
Method Comparison study. Wipe samples were
analyzed by both the bioavailable and total
digestion methods in that study. The correction
factor of 1.58 is estimated as the ratio of the
geometric mean wipe lead loading from the total
method to that from the bioavailable method.
Prior to calculating the ratio, each total lead
loading value was multiplied by 1.25 to account
for the lead lost to the bioavailable analysis.
Blue
Nozzle
NCLSH/
Westat [2]'
Uncarpeted Floors
Window Sills
Window Wells
(7)
(42)
(6)
Side-by-side wipe and BN
vacuum samples were
collected from uncerpeted
floors, window sills, and
window wells.
Forty homes owned by the Baltimore
Housing Authority were included in this
study. Of these, 30 were rehabilitated and
10 were not. No description of the
rehabilitation of the homes was provided.
Since several sample leed measures in these
homes were below the detection limit,
additional samples were collected from 5
homes owned by City Homes, Inc. No
information about the intervention status of
those 5 homes was provided. No
interventions were performed as a part of
the NCLSH/Westat Study.
BN concentrations were not available to
characterize the relation between wipe and BN
concentrations
t
I
(o
to
o>
CAPS Pilot
Study |3|
Uncarpeted Floors (6)(c>
Window Sills (6)(c>
Two side-by-side wipe
samples were taken adjacent
to two side-by-side BN
samples on uncarpeted
floors. Geometric average of
wipe pairs were related to
geometric average of BN
samples. One wipe half-
window sill sample was
taken adjacent to one BN
half-window sill sample.
2 methods of abatement: 2 homes were
abated by encapsulation/enclosure
methods, and 2 homes were abated
byremoval method approximately 1 year
prior to sampling.
None
-------
Table 1. Summary Table for the Studies from which Data were Used to Develop Conversion Equations (Continued)
Vacuum
Type
BRM
Study
R&M Mini
Study
Rochester
Lead-in-Dust
Study[4]
NCLSH 5
Method
Comparison
Study |5|
Milwaukee
Low Cost
Interventions
Study
Surface Type (N)M
Uncarpeted Floors (25)
Window Sills (27)
Window Wells (25)
Uncarpeted Floors(389l
Carpeted Floors (398)
Window Sills (362)
Window Wells (403)
Uncarpeted Floors (68)
Cerpeted Floors (67)
Floors" (135)
Relevant Issues
of Study Deslnn
Side-by-side wipe and BRM
vacuum samples were
collected from uncarpeted
floors, window sills, and
window wells.
Side-by-side wipe and BRM
vacuum samples were
collected from uncarpeted
floors, carpeted floors,
window sills, and window
wells.
Side-by-side wipe and BRM
vacuum samples were
collected from uncarpeted
and carpeted floors.
Side-by-side wipe and BRM
vacuum samples were
collected from uncarpeted
floors.
Intervention History
Seven pre-1940, vacant homes in Baltimore
City were included in this study. These
homes were not known to have undergone
any lead paint abatement. No interventions
were carried out as a part of the R&M Mini
Study.
The 205 homes in this study were known
to have not undergone extensive renovation
or remodeling within the year prior to the
eligibility interview. No interventions were
performed as a part of the Rochester Study.
No interventions were carried out as a part
of the NCLSH 5-Method Comparison Study.
Only pre-mtervention data was used in this
analyses. No counseling or medical
treatment was received by residents in the
included homes, and no homes had
undergone abatement prior to the study.
Wipe lead loadings were multiplied by 1 .58 to
adjust the measure of bioavailable lead to total
available lead.
None
Wipe lead loadings were multiplied by 1 .25 to
correct for only including 80% of wipe samples ir
the chemical analyses.
Duplicate vacuum samples were excluded in
determining the relationships.
i
I
Ul
(a) Number of pairs
(b) No dust-lead concentrations were obtained from the NCLSH/ Westat Study.
(c) Each member of each pair is an average of two side-by-side samples That is, two side-by-side wipe and two adjacent vacuum samples were taken from each house in the CAP Pilot Study.
Id) All floor samples from the Milwaukee Low Cost Interventions Sludy are uncarpeted per personal conversation with Brad Schultz, statistician, USEPA |6|.
-------
Four studies report side-by-side paired data on wipe lead loading and BRM vacuum lead
loading and were used to develop the BRM loading to wipe loading conversion equations:
1. R&M Mini study
2. Rochester Lead-in-Dust study
3. NCLSH 5-Method Comparison study
4. Milwaukee Low Cost Interventions study
Notice from Table 1 that the housing components sampled varied by study. The Rochester Lead-
in-Dust study was the only study for which side-by-side wipe and BRM samples were collected
on all four components. In the Rochester Lead-in-Dust, NCLSH 5-Method Comparison, and
Milwaukee Low Cost Interventions studies, wipe samples were analyzed using the hot nitric
acid/peroxide digestion method. The R&M Mini study, however, used the cold hydrochloric
acid digestion procedure mentioned above. The degree to which the cold hydrochloric acid
digestion procedure yields different lead loadings than the hot nitric acid procedure was
examined in the NCLSH 5-Method study. It was estimated that samples digested using hot nitric
acid yielded wipe lead loadings 1.58 times greater than those digested with cold hydrochloric
acid. Therefore, the wipe lead loadings in the R&M Mini study were also adjusted using the
correction factor of 1.58 obtained from the NCLSH 5-Method Comparison study. This
correction factor only affected the wipe samples in the R&M Mini study and the R&M Pilot
study.
In the NCLSH 5-Method Comparison study, only 80 percent of each sample collected by
wipe was extracted using the hot nitric acid/peroxide digestion method (total lead digestion).
The lead loadings for total lead digestion reported in the NCLSH 5-Method Comparison study
were therefore multiplied by 100/80 = 1.25 to adjust for the chemical analysis of a reduced
sample.
In the Milwaukee Low Cost Interventions study, there were some duplicate BRM
samples collected from a location adjacent to the side-by-side samples provided for this analysis.
These duplicates were not included in the development of these conversion equations.
Draft - Do Not Cite or Quote 6 September 17, 1996
-------
2.2 DESCRIPTIVE STATISTICS
Table 2 presents descriptive statistics for the variables used in developing the various
conversion equations in this report. Notice that data was available for the relationship between
wipe and BN on uncarpeted floors, window sills, and window wells. No information regarding
this relationship on carpeted floors was found. However, for exploring the relationship between
wipe loadings and BRM loadings, data was available from uncarpeted floors, carpeted floors,
window sills, and window wells.
2.3 RANGE ISSUES
This section deals with issues involving the ranges of the data used to develop the
conversion equations and the impact of these ranges on the applicability of the equations.
2.3.1 Range of Data
The BN to wipe conversion equations will be used to convert the BN dust-lead loadings
measured in the National Survey to equivalent lead loadings for wipe samples. The transformed
lead loadings will then be used to determine the numbers and percentages of houses that would
be affected for various options for defining what is lead-contaminated dust.
The data used to develop the BN lead loading to wipe lead loading conversion equation
ranges from 1 to 2164 ug/ft2 for uncarpeted floors. The BN lead loadings hi the HUD National
Survey range from 0.014 to 380 ng/ft2 for uncarpeted floors. Of the 364 non-missing BN
vacuum lead loadings from uncarpeted floors in the HUD National Survey, 190 (52%) fall below
the minimum BN vacuum lead loading used to develop the BN loading to wipe loading
conversion equation. Of the 284 homes surveyed, ISO have at least one uncarpeted floor dust-
lead loading falling below the range of the conversion data.
Figure 1 displays the predicted wipe lead loading, by housing component, associated with
each BN vacuum lead loading observed in the HUD National Survey. Also included on the plots
are the approximate 95% confidence and prediction intervals corresponding to each predicted
wipe lead loading. Note the width of the confidence intervals increases at the lower and upper
ends of the data. There is greater uncertainty in predicting wipe lead loadings from BN vacuum
Draft - Do Not Cite or Quote 7 September 17, 1996
-------
I
I
0.
I
5
Table 2. Descriptive Statistics for Vacuum Dust Lead Measures and Wipe Lead Loading by Study
Vacuum
Type
Blue
Nozzle
BRM
Study
R&M Pilot
Study |1]
NCLSH/
Westat 12)
CAPS Pilot
Study (3]
R&M Mini
Study
Rochester
Lead-in-Dust
StudyKI
NCLSH 5
Method
Comparison
Study 15]
Milwaukee
Low Cost
Interventions
Study
Surface Type
Uncarpeted
Floors
Window Sills
Window Wells
Uncarpeted
Floors
Window Sills
Window Wells
Uncarpeted
Window Sills
Uncarpeted
Floors
Window Sills
Window Wells
Uncarpeted
Floors
Carpeted Floors
Window Sills
Window Wells
Uncarpeted
Carpeted Floors
Floors'"
No. of Pairs
24
23
24
7
42
6
6""
6'"
25
27
25
389
398
362
403
68
67
135
Vacuum Loading (pg/ft2)
Geometric
Mean
16.6
48.1
6,030
11.2
14.7
171
10.7
36.0
320
4,400
217.000
12.8
180
227
11.400
26.6
413
37.2
Log
Std. Dev.
2.27
2.62
2.73
0.526
0.832
0.863
1.66
1.72
2.62
4.08
3.80
2.07
1.73
2.42
3.39
2.43
1.17
2.19
Vacuum Concentration
b/g/g)
Geometric
Mean
712
2,390
4.780
(a)
212
584
Log
Std. Dev.
1.29
1.84
2.41
(a)
1.16
0.989
Wipe Loading (jig/ft'l
Geometric
Mean
111
418
2.650
50.6
133
13.400
51.0
164
226
2,870
40.500
16.6
11.3
163
2,590
25.8
4.97
38.4
Log
Std. Dav.
1.80
3.31
2.61
0.362
1.28
2.03
1.93
1.83
1.74
3.03
2.45
1.25
1.13
1.53
2.61
1.52
079
1.18
00
(a) No dust-lead concentrations were obtained from the NCLSH/Westat Study.
(b) Each member of each pair is an average of two side-by-side samples. That is, two side-by-side wipe and two adjacent side-by-side vacuum samples were taken
from each house in the CAP Pilot Study.
(c) All floor samples from the Milwaukee Low Cost Interventions Study are uncarpeted per personal conversation with Brad Schultz, statistician. USEPA |6).
0)
-------
I
O
o
2
2
o
8
q
D
o
• Predicted Wipe Loading
— l.ower Confidence Bound
— Upper CcnJldence Bound
Lower Prediction Bound
Upper Prediction Bound
Blue Nozzle Uncarpeled Floor Dust-Lead Loadings (ug/fl2)
01 01 I 10 100 I OOO 10000
Blue Nozzle Window Sill Dust-Lead Loadings (ug/ft2)
I
*
1
Predicted Wipe Loading
Lower Confidence Bound
Upper Confidence Bound
Lower Prediction Bound
Upper Prediction Bound
01 1 10 100 l.OOO 10.000
Blue Nozzle Window Well Dust-Lead Loadings (ug/ft2)
CO
8
Figure 1. Wipe Loadings (//g/ft2) as Predicted from Blue Nozzle Vacuum Loadings (//g/ft2) versus Blue Nozzle Vacuum
Loadings with 95% Confidence and Prediction Intervals Superimposed for Uncarpeted Floors, Window Sills, and
Window Wells from the HUD National Survey.
-------
lead loadings in these ranges. For example, the greatest precision for converting a BN to a wipe
lead loading is obtained at a BN lead loading of about 10-20 ug/ft2.
For window sills, the BN lead loadings from the HUD National Survey range from 0.004
to 11,899 ug/ft2 with 136 (35%) of the 392 non-missing observations falling outside the range of
the conversion data. One BN lead loading is above the maximum of the data used to develop the
window sill conversion equation (8964 ug/ft2) and 135 BN loadings are below the minimum (1.4
ug/ft2). Of the 284 homes surveyed, 107 have at least one window sill dust-lead loading falling
below the range of the conversion data.
The range of the BN lead loadings used to develop the BN loading to wipe loading
conversion equation for window wells is 35.5 to 761,842 ^g/ft2. Forty-seven (30%) of the 158
non-missing BN vacuum lead loadings from window wells in the HUD National Survey fall
below the minimum of 35.5 ug/ft2. Forty of the 284 homes included in the survey have at least
one window well dust-lead loading falling below this minimum. Table 3 lists the ranges of the
data used to develop the BN loading to wipe loading conversion equations for each housing
component. Also presented are the ranges of the HUD National Survey BN loadings that will be
converted using these equations.
Table 3 also includes the range of the data used in developing the BRM loading to wipe
loading conversion equation for each housing component. These equations will be used to
transform the BRM vacuum lead loadings in the Baltimore Repair and Maintenance study to
equivalent wipe lead loadings for use in determining prevalence statistics and in completing a
sensitivity/specificity analysis which relates the incidence of elevated (2 10 ug/dL) children's
blood-lead levels to wipe lead loadings. The ranges of the Baltimore Repair and Maintenance
BRM loadings are provided in Table 3 as well.
Figure 2 presents the predicted wipe lead loading, by housing component, for each area-
weighted average BRM vacuum loading per home from the Baltimore Repair and Maintenance
study. Approximate 95% confidence and prediction intervals are also included on the graphs.
The width of the intervals is dependent on the range of the data used to develop the conversion
equation. That is, the variability in the parameter estimates of the combined regression model
determine the uncertainty associated with each predicted wipe loading. Thus, if data over a
different range were used to develop the conversion equation, the levels with the most precise
Draft - Do Not Cite or Quote 10 September 17, 1996
-------
conversions would be different. The BRM loadings in the Baltimore Repair and Maintenance
study fall within the range of the data used to develop the conversion equations for all housing
components.
Table 3. Data Ranges by Housing Component and Vacuum Type for the Data Used to
Develop the Vacuum Loading to Wipe Loading Conversion Equations and for the
Data upon Which the Conversion Equations Will Be Used.
Vacuum
Type
BN
BRM
Data
CAPS. R&M
Pilot, and
NCLSH/Westat
Development
HUD National
Survey
(Application)
R&M Mini.
Rochester,
NCLSH 5-
Method,
Milwaukee
(Development)
Baltimore R&M"1
(Application)
Housing Component
Uncarpeted Floors
Min.
1.00
0.014
0.080
3.60
Max.
2,164
380
74.100
23,416
Carpeted
Floors
Min.
NA
NA
1.42
NA
Max.
NA
NA
141.000
NA
Window
Sills
Min.
1.40
0.004
0.250
2.04
Max.
8,964
11,899
4.169,649
115.321
Window
Wells
Min.
35.5
0.042
1.88
36.2
Max.
761.842
40,457
6.610.797
2.496,630
(a) Statistics are calculated from area-weighted averages of all composite samples per housing component per home.
2.3.2 Abated versus Unabated
A second exercise will be undertaken to assess the impact of the Section 403 standard on
children's blood-lead levels. Blood-lead levels will be predicted for the post-Section 403
environment. This may be done in more than one way, but it will certainly be done using the
IEUBK model. For this exercise, it will be assumed that floor lead loadings measured by wipe
samples will be equal to, say, 40 ug/ft2 after an intervention conducted in response to the 403
rule in all units that exceed the 403 standard for floors. The baseline number of 40 ug/ft2 (or
whatever it may eventually be) will be converted to an equivalent dust-lead concentration for a
National Survey vacuum sample using the wipe loading (ug/ft2) to BN concentration (fig/g)
conversion equation. The equivalent National Survey dust-lead concentration is 328 ug/g, with
Draft - Do Not Cite or Quote
11
September 17, 1996
-------
3
D
O
n>
o
D
o
I'irJt' 1nJ »i|><- Loading
• L..-rr Confidence Bound
- Upper Confidence Bound
Lo.er Prediction Bound
BRM Uncarpeted Flo
100 I 000 10.000
• Dusl-Lead Loadings (ug/fl2)
ID 100 1.000 10 000
BRM Carpeted Floor Dust-Lead Loadings (ug/ft2)
I
Pred.c
led *,pe Lo.d,n,
i.o-r, Confidence Bound
Upper Confidence Bound
Lower Prediction Bound
Upper Prediction Bound
100 I 000 10 000 100000
Sill Dust - Lead Loadings (ug ft2)
Predicted *ipr Loading
Lower Confidence Bo.in.
• Upper Confidence Boun
- Upper Prediction Bound
100 1 000 10000 100000 1 000000
BRM Wmdow Well Dust-Lead Loadings (ug/fl2)
NJ Figure 2. Wipe Loadings (/sg/ft2) as Predicted from BRM Vacuum Loadings (j/g/ft2) versus BRM Vacuum Loadings with 95%
- Confidence and Prediction Intervals Superimposed for Uncarpeted Floors, Window Sills, and Window Wells from
£ the Baltimore Repair and Maintenance Survey.
-------
an approximate 95% confidence interval of 255 to 423 ug/g. It should be noted that 95% of the
BN dust-lead concentrations associated with a wipe dust-lead loading of 40 ug/ft2 are predicted
to be between 94.2 and 1,140 ug/g. This concentration (328 ug/g) will be assumed for the
houses which trigger an intervention based on their measured pre-intervention dust-lead loading,
and it will be used as input to the IEUBK model to predict blood-lead levels. The model will
then be used to determine the resulting distribution of children's blood-lead levels and inferred
health effects in the nation. Comparisons will be made to the prior distribution of children's
blood-lead levels (as determined from the NHANES III Survey) to determine the benefits of the
Section 403 rule.
Figure 3 displays all of the data used to develop the wipe loading to BN concentration
conversion equation (i.e., CAPS Pilot and R&M Pilot) for each housing component. Abated and
unabated homes are plotted with different symbols. It can be seen that for each housing
component, both types of homes span the range of the data. This supports the use of all the data
to develop the wipe loading to BN concentration conversion equation to predict post-intervention
blood-lead concentration.
2.4 OTHER ISSUES
Seasonal rhythms in blood lead levels have been observed in several studies. In addition,
seasonal rhythms in environmental levels were observed in a report [7] conducted in Boston
during 1979 to 1983. In that study, floor dust-lead loadings were on average 50 percent higher in
July compared to December. Environmental samples in the HUD National Survey were
collected during the winter months, and may have been higher if they were collected during the
summer. Because seasonal variations in dust lead are most likely a function of geographic
location, the seasonal variations estimated in Boston may not be applicable to the HUD National
Survey. Therefore, an adjustment to the floor dust-lead loadings in the HUD National Survey to
account for seasonal variations was not investigated.
The types of wipes used in the various studies is identified in Table 4, with an indication
of whether there is known to be background lead in these wipes. The wipes used in the CAPS
Draft - Do Not Cite or Quote 13 September 17. 1996
-------
o
r*
n
S
o
Uncarpeted Kloor Wipe Lead Loading (Ug/Tt2)
Abated Homes
Unoboled Homt
100 I 000 10000
Window Sill Wipe Lead Loading (ug/flZ)
100 000 1 OOO 000
Sf
(o
O)
100 I OOO 10 000 100 000
Window Well Wipe Lead Loading (ug/ftS)
Abated Homes
Unabated Homes
Figure 3. Blue Nozzle Vacuum Concentrations versus Wipe Lead Loading Displaying the Distribution of Abated and
Unabated Homes Across the CAPS Pilot and R&M Pilot Data Used to Develop the Conversion Equations for
Uncarpeted Floors, Window Sills, and Window Wells.
-------
Table 4. Wipes Used in Various Studies.
Vacuum Type
Blue Nozzle
BRM
Study
CAPS
R&M Pilot
NCLSH/Westat
R&M Mini
Rochester Lead-in-Dust
NCLSH 5-Method Comparison
Milwaukee Low Cost
Interventions
Wipes Used
Chubbs
WetOnes
iai
WetOnes
Little Ones Baby
Little Ones Baby
Wash-a-bye Baby
Background Lead
Yes
None known
iai
None known
None known
None known
1 -2 //g lead per
wipe
(a) Could not be determined from report.
and the Milwaukee studies had trace amounts of lead. These trace amounts would be expected to
only slightly increase the variability, and slightly bias the relationships.
In Table 5, information related to the non-detection of lead loadings and lead
concentrations in the sets of data being used to develop the conversion equations is presented.
This includes the instrument detection limit (IDL) in each study, the number of samples below
the IDL, and a description of how these samples were handled in the analysis. For the BRM
conversion equations, a very small percentage of the data was below the IDL, so their influence
is probably small. For the BN conversions, the vast majority of the samples collected in the
NCLSH/Westat study were below the IDL, and these were excluded. For the CAPS and R&M
Pilot studies, either IDL or IDL/2* was used in place. None of these points were found to be
influential observations. (Appendix B presents and analysis of the influential data points.)
Draft - Do Not Cite or Quote
15
September 17, 1996
-------
Table 5. Information Related to Non-detection for Lead Measures in Various Studies.
Vacuum
Type
Blue Nozzle
BRM
Study
CAPS
R&M Pilot
NCLSH/
Westat
R&M Mini
Rochester
Lead-in-Dust
NCLSH 5-
Method
Comparison
Milwaukee
Low Cost
Interventions
Instniment Detection
Limit (IDL)
13.77 pg/sample
An approximate value of 7 /sg for
IDL of wipe samples was
determined from the data set
5 fig for blue nozzle
25 j/g for wipe
-
FAA: < 1 0 fjg/samp\e
GFAA: wipe < 0.25 //g/sample
BRM < 0.1 5 //g/sample
(b)
IDLS varied from 1 .69 to 1 .89
/ig/ft2 for wipe and from 1 .08 to
48.6 //g/ft2 for BRM.
Number of Samples
Below IDL
1'"
3 wipe measurements (2
sills and 1 well)
For 292 of 351 the side-
by- side pairs, one or
both measurements
were below IDL
None
wipe: 1 uncarpeted floor
1 window sill
BRM: 3 uncarpeted floor
3 window sill
1 window well
(b)
1 wipe
9 BRM
How were Below IDL
samples handled in
this report
IDL was used to
compute lead loading
Measurements were
replaced by IDL/2"2
Only paired samples
with both members of
the pair above the IDL
were used in analysis
Samples below the
limit of detection for
flame AA were
reanalyzed by graphite
furnace AA, and for
these, the GFAA
response was used in
this report. The
reported values for the
7 samples below the
GFAA detection limit
were included in the
analysis with no
changes.
Samples below the
limit of detection for
flame AA were
reanalyzed by graphite
furnace AA, and for
these, the GFAA
response was used in
this report.
Wipe and BRM
loadings below and
above the IDL were
handled the same
way. The detection
limit was ignored in
reporting of instrument
responses.
(a) Both side-by-side wipes were below IDL. Their average was used for comparison with average of side-by-side vacuum
samples.
(b) Could not be determined from the report.
Draft - Do Not Cite or Quote
16
September 17, 1996
-------
3.0 STATISTICAL APPROACH
Battelle recommends a log-linear model to characterize the relationship between lead
loadings for two different samplers on all housing components. For instance, the model relating
wipe lead loading to Blue Nozzle lead loading may be written as
log(BN) = log(a) + p log(W) + log(E),
or equivalently as,
BN = aW|JE,
where BN=Blue Nozzle lead loading and W=wipe lead loading. E represents a random error
term. This form of model has been used in many studies for relating one lead measurement
method to another (e.g., [1], [2], [3], [4], [6]). Including a variable, P, to represent the exponent
on the independent variable permits the ratio to depend on the level of the predictor. This
exponent is usually tested for significance and is often rejected, leading to the use of a simple
scale factor to characterize the relationship. The more general model was used in this analysis
because results are being combined across studies.
An analogous model is recommended for characterizing the relationship between wipe
lead loading and BN lead concentration, with BN representing BN lead concentration.
A third, similar model is recommended for converting a lead loading from either vacuum
sampler to a wipe lead loading:
log(W) = log(cc) + p log(V) + log(E),
or equivalently as,
W = aVpE,
where V=Blue Nozzle or BRM vacuum lead loading and W=wipe lead loading, with E
representing a random error term.
Note that the models relate lead loadings from side-by-side samples. Therefore, the
models are most appropriate for converting one measure to another at an adjacent location.
However, the last model will also be used to convert a composite BRM lead loading to a
composite wipe lead loading, since the Baltimore Repair and Maintenance Study is comprised of
composite dust samples.
Draft - Do Not Cite or Quote 17 September 17, 1996
-------
For all of the models developed in this report, the least squares method was used to obtain
point estimators of the model parameters, namely, a and P. To estimate confidence or prediction
intervals, however, it is important to make an assumption about the functional form of the error
term, E. In practice, the distribution of E can often be reasonably described by a lognormal
distribution. We adopt this assumption in this report.
While the approach is direct and relatively simple to address the conversion, a serious
limitation is the relatively weak relationships observed between measured lead loading and
measured lead concentration. For example, it is possible that two greatly differing lead loadings
will convert to overlapping distributions of lead concentrations. This will happen if there is high
variability in lead concentrations corresponding to a lead loading. In this case, it is quite possible
for a particular lead concentration to be measured in two samples with very different lead
loadings. However, the method used in this report will predict a unique lead concentration
corresponding to each lead loading. As noted above, the relationship between true lead loading
and true lead concentration depends only on the amount of dust on the sampled surface.
Therefore, the dust loading in the homes used to develop the conversion equations are assumed
to be similar to the dust loading in the homes on which conversion equations will be applied.
3.1 MEASUREMENT ERROR
We have not adjusted the conversion equations presented in this report for the effects of
measurement error. Specifically, we have not used an errors-in-variables approach in developing
these equations. While both wipe and vacuum lead measurements are subject to error, the
purpose of the equations is for prediction of measured levels that would be obtained by another
sampling method, based on the measured level of a given sampling method. In these cases, it is
not appropriate to use an errors-in-variables correction.
For example, in predicting the BN lead concentration based on information obtained with
a wipe sample, the goal is not to determine what the true lead concentration is at that site.
Rather, the goal is to determine what the geometric mean measured concentration would be
across all BN samples collected at that site.
Draft - Do Not Cite or Quote 18 September 17. 1996
-------
3.2 COMBINING RESULTS ACROSS STUDIES
Having fitted separate regression equations from the appropriate studies to predict a wipe
lead loading from a vacuum lead loading for each surface and vacuum type, an approach is
needed to obtain a single conversion equation for each surface and vacuum type. Because no
single study was determined to provide data of superior quality to the other studies, it was
preferable to combine information obtained in different studies to characterize the relationship.
We assumed in this analysis that a linear model on a log scale is appropriate to describe
the relationship between wipe (W) and vacuum (V) measures within each study as well as across
all studies. That is
log(W) = log(o) + p log(V) + log(E).
Separate estimates of a and P were obtained from each of the studies. In determining a final
estimate of the intercept, we used a weighted average of the intercept estimates obtained from
each of the studies. The weighting used was the inverse of the variance of the intercept obtained
in each study (that is, the squared inverse of the standard error). An estimate of the slope was
obtained similarly as the average of the slope estimates obtained in the studies, each weighted by
the inverse of its estimated variance.
It can be shown in general that if a single estimate is to be obtained based on two or more
independent, unbiased estimates of the same parameter using a weighted average of the
individual estimates, then weighting each individual estimate by the inverse of its variance
minimizes the variance of the final estimate.
As the model is stated above, the intercept is the predicted log lead loading of a sample
collected by a vacuum sampler with a lead loading of 1 ng/ft2. A lead loading of 1 ug/ft2 (which
has a log of 0) is near the lower bound of the range of vacuum lead loadings in each of the
studies used to develop these equations. Thus, an estimate of the intercept at a lead loading of
1 ug/ft2 would have greater uncertainty than if the intercept was estimated near the "center" of
the observed vacuum lead loadings. In fact, the uncertainty of the intercept is minimized if it is
estimated at the average, or the "center," of the values taken on by the independent variable.
Therefore, the log lead loading of each vacuum sample in each study was centered by subtracting
Draft - Do Not Cite or Quote 19 September 17, 1996
-------
the average log lead loading obtained by vacuum samples across the studies for which data was
available for that housing component and vacuum type. For each study, the regression model
was then fined to the centered data.
In summary, the following seven steps describe the method used for computing the
weighted average of the conversion equations for the particular case of converting vacuum lead
loadings to wipe lead loadings on uncarpeted floors. The steps for the other housing components
and conversions are analogous.
1. The mean log lead loading of vacuum samples over all samples in the appropriate
studies was computed.
2. The overall mean was subtracted from each log vacuum lead loading to center the
data.
3. A regression model was fitted to the centered data from each study to estimate the
slope, the centered intercept, and their respective standard errors.
4. The overall slope was computed as a weighted average of the individual slopes using
the inverse of the estimated slope variances as the weights.
5. The overall centered intercept was computed as a weighted average of the individual
centered intercepts using the inverse of the estimated intercept variances as weights.
6. The overall (uncentered) intercept was computed using the following formula:
aw = of"""* d'tt) - (mean log lead loading) Pw,
where a£nteni datt) is the weighted average of the estimated intercept for the
centered data, pw is the weighted average of the estimated slopes.
7. Thus, to convert a vacuum lead loading, V, to a wipe lead loading, W, the formula is:
log(W) =
-------
loadings, for the surface types with available data. Separate centering constants were determined
for each sample type (i.e., floors, window sills or window wells), and predictor method (i.e.,
BRM, BN, or wipe). This technique of centering the data is discussed on page 260 of Draper and
Smith [8]. (Note that the centering has no effect on the estimated slopes and their variances.)
3.3 CONFIDENCE AND PREDICTION INTERVALS
When one of the equations discussed in this report is used, the outcome is a prediction
(i.e., an estimate). In the context of Section 3.2 using the equation log(W) = log(cc)+p log(V), the
result is an estimate of a wipe lead loading given a particular vacuum lead loading. Confidence
intervals and prediction intervals are used to assess how "good" the estimate is.
The following formulae were used to calculate an approximate 95% confidence interval
for a predicted wipe lead loading from the above model:
Lower Confidence Bound = e(B * P<""«V) ' v> - 2y^)
Upper Confidence Bound = e<" + P<"*'7) *2^ >
where a = the estimated intercept of the combined regression model, p = the estimated slope of
the combined regression model, V = mean log lead loading from vacuum samples from all
studies, and s2 = variance associated with the estimate of the wipe lead loading and is calculated
as:
s2 = s.e.(o)2 + (log(V) - V)2 * s.e.(P)2
with s.e.(cc) and s.e.(P) being the standard errors of the parameter estimates from the combined
regression model.
The calculation of approximate 95% prediction intervals utilizes the same formulae, with
one exception. Since the desired intervals are to encompass an estimate of a single, future wipe
lead loading, an extra component must be added to the variability associated with this prediction.
Thus, for an approximate 95% prediction interval around a predicted wipe lead loading from the
above model, the following equations are used:
Draft - Do Not Cite or Quote 21 September 17. 1996
-------
Lower Prediction Bound =e
Upper Prediction Bound = e" * P("'(V)
where a, P, and V are as before and s.2 is calculated as:
s2 = s.e.(a)2 + (log(V) - V)2 * s.e.(P)2 + MSB
with MSE being the weighted average of the mean square errors from the separate study
regressions, weighted by their respective degrees of freedom. Tables 12, 13, and 14 display
predicted values and approximate 95% confidence and prediction intervals for selected input
values for the conversion equations based on BN vacuum measures. Table 1 7 presents similar
results for the conversion equations based on BRM vacuum loadings.
Draft - Do Not Cite or Quote 22 September 17, 1996
-------
4.0 RESULTS
This section presents the results of the analyses on the conversion equations based on the
studies identified as relevant for each vacuum type.
4.1 RESULTS OF STATISTICAL ANALYSES FOR THE BLUE NOZZLE VACUUM
The estimated conversion equations for predicting a Blue Nozzle vacuum lead loading
from a wipe lead loading for the CAPS Pilot, NCLSH/Westat, and R&M Pilot studies are
displayed in Table 6. Table 7 presents the equations for predicting a Blue Nozzle concentration
from a wipe lead loading and Table 8 presents the equations for predicting a wipe lead loading
from a Blue Nozzle lead loading. Results for uncarpeted floors, window sills, and window wells
are presented separately in each table.
The results presented for the R&M Pilot study in Tables 6, 7, and 8 include an
adjustment for the chemical extraction procedure employed for the HUD wipe samples. (As
described in Section 2 an estimate of 1.58 for the ratio of the average HUD wipe lead loading to
bioavailable wipe lead loading was used to adjust the bioavailable wipe loadings. This value was
derived based on uncarpeted floors and was also applied to window sills and wells.)
Figures 4,5, and 6 display the modeled relationships for wipe to vacuum lead loadings
for the CAPS Pilot, NCLSH/Westat, and R&M Pilot studies, respectively. In each of these
figures, the relationships for floors, sills, and wells are displayed simultaneously. The
relationships from BN lead loading to wipe lead loading are presented in Figures 7, 8, and 9 for
the same three studies. In Figures 10 and 11, the relationships between wipe lead loading and
Blue Nozzle vacuum lead concentration are presented for the CAPS Pilot and R&M Pilot studies.
Table 9 presents the three final conversion equations for uncarpeted floors. (See Section
3.2 regarding the combining of results across studies.) Total sample sizes used to develop each
conversion equation are also provided. Tables 10 and 11 present the corresponding results for
window sills and window wells, respectively.
Table 12 presents the predicted levels associated with nominal levels for each conversion
equation for uncarpeted floors at nominal loadings of 10,40,100,200,500, and 1000 ug/ft2.
The top portion of the table provides conversions from wipe lead loading to BN lead loading or
concentration. The bottom portion of the table provides conversions from BN lead loading to
Draft - Do Not Cite or Quote 23 September 17. 1996
-------
wipe lead loading. For example, the BN loading of 200 ug/ft2 gets converted to a wipe lead
loading of 440 ug/ft2.
The second line in each cell represents an approximate 95% confidence interval on the
conversion. These represent confidence bounds on the estimate of the geometric mean level
(either BN loading, BN concentration, or wipe loading), associated with the nominal levels that
they are converted from. Thus, for a wipe lead loading of 100 ng/ft2, the geometric mean BN
lead loading is 14.9 fig/ft2, and we have 95% confidence that the geometric mean BN lead
loadings taken at the same location would be between 10.6 and 21.0 fig/ft2.
The third line in each cell provides an approximate 95% prediction interval for the
conversions. These are bounds that are expected to contain 95% of the individual observations
associated with the specified nominal levels of the prediction variables. Thus, for a wipe lead
loading of 100 jig/ft2, the geometric mean point estimate of the BN lead loading is 14.9 jig/ft2,
and we could expect 95% of BN lead loadings measured at the same location to be between 2.13
and 105 ug/ft2. Tables 13 and 14 present the corresponding results for window sills and window
wells, respectively.
Figure 12 displays, in a concise fashion, the BN lead loading versus wipe lead loading
results observed for uncarpeted floors. Data from each of the three studies is plotted with
different symbols, along with regression lines from the individual studies and combined across
studies, and confidence bounds on the combined regression line. Figure 13 displays the same
type information regarding the BN lead concentration versus wipe lead loading relationship for
uncarpeted floors', and Figure 14 is a similar graph for the wipe lead loading versus BN lead
loading for uncarpeted floors.
Figures 15,16, and 17 present similar information on window sills for the three
relationships studied in this report, and results for window wells are displayed in Figures 18,19,
and 20. Appendix A provides cross-tabulation information on the distribution of the wipe and
BN vacuum pairs of data in the CAPS Pilot, NCLSH/Westat, and R&M Pilot studies used to
develop the BN conversion equations. Appendix B presents a validation of assumptions made in
'Note that data was available from only two studies for predicting BN lead concentration from wipe lead loadings on
uncarpeted floors.
Draft - Do Not Cite or Quote 24 September 17, 1996
-------
conducting this analysis. This includes a residual analysis and plots highlighting the data values
in each study that have an impact on the parameter estimates from the regression.
4.2 RESULTS OF STATISTICAL ANALYSES FOR THE BRM VACUUM
The estimated conversion equations for predicting a wipe lead loading from a BRM lead
loading for the R&M Mini, NCLSH S-Method Comparison, Rochester Lead-in-Dust, and
Milwaukee Low Cost Interventions studies are displayed in Table 15.
As described in Section 2, the results presented for the R&M Mini study in Table IS
include an adjustment for the chemical extraction procedure employed for the HUD wipe
samples. An estimate of 1.58 for the ratio of the average HUD wipe lead loading to bioavailable
wipe lead loading was used to adjust the bioavailable wipe loadings. This value is based on
uncarpeted floors and was also applied to window sills and wells.
Figures 21 to 24 display the modeled relationships for BRM to wipe lead loadings for the
R&M Mini, NCLSH 5-Method Comparison, Rochester Lead-in-Dust, and Milwaukee Low Cost
Interventions studies, respectively. In each of these figures, the relationships for all housing
components sampled are displayed individually.
Table 16 summarizes, for each of the four housing components, the final conversion
equations and the total sample sizes used in developing these equations. The coefficients in this
table are based on the centered predictor variables for purposes of combining results across
studies. (See Section 3.2 regarding the combining of results across studies.) As an alternative
approach to combining the results across studies, it was requested that a simple log-linear
regression be performed on the data pooled across studies, without regard to possible differences
in variability across studies. Appendix C presents those results.
Table 17 presents the predicted wipe lead loadings for each BRM conversion equation
presented in this report at nominal BRM lead loadings of 10,40,100,200,500, and 1000 ug/ft2.
For example, a BRM loading of 200 ug/ft2 on an uncarpeted floor gets converted to a wipe lead
loading of 46.1 ug/ft2 with an approximate 95% confidence interval of 40.6 to 52.6 ug/ft2.
The second entry in each cell of Table 17 provides 95% confidence intervals on the
conversions for all housing components. These represent confidence bounds on the geometric
Draft - Do Not Cite or Quote 25 September 17. 1996
-------
mean wipe lead loading associated with the nominal levels that they are converted from. Thus,
for a BRM lead loading of 100 ng/ft2 from an uncarpeted floor, the estimated geometric average
of wipe lead loading is 37.1 ug/ft2, and we have 95% confidence that the geometric average wipe
lead loadings taken at the same location would be between 33.3 and 41.6 fig/ft2.
The third entry in each cell of Table 17 lists 95% prediction intervals for bounds expected
to contain 95% of the individual observations associated with the specified nominal levels of the
prediction variables. Thus, for a wipe lead loading of 100 ug/ft2, from an uncarpeted floor, the
point estimate of wipe lead loading is 37.1 ug/ft2, and we could expect 95% of wipe lead
loadings measured at the same location to be between 4.20 and 329 ug/ft2.
Figure 25 displays, in a concise fashion, the wipe lead loading versus BRM lead loading
results observed for uncarpeted floors. Data from each of the four studies is plotted with
different symbols, along with regression lines from the individual studies and combined across
studies, and confidence bounds on the combined regression line. Figures 26,27, and 28 display
the same type of information for carpeted floors, window sills, and window wells, respectively;
information was available from less than four studies for these three housing components.
Appendix D presents a validation of assumptions made in conducting this analysis. This includes
a residual analysis and plots highlighting the data values in each study that have an impact on the
parameter estimates from the regression.
Draft - Do Not Cite or Quote 26 September 17. 1996
-------
Table 6. Regression Equations for Predicting Blue Nozzle Vacuum Lead Loading from Wipe
Lead Loading.
Surface
Floors
(Uncarpeted)
Window Sills
Window Wells
Estimated Regression Model for . .
Wipe Pb Loading to Blue Nozzle Vacuum Pb Loading1 '
CAPS Pilot
logla'(BN)= -
0.413 + 0.708llog(W)
n = 6 R2= 0.683
Range W '?: 13.8-2498.5
Range BN: 1.9-149.1
log(BN) =-0.941 +0.887*log(W)
n = 6 R2 =0.927
Range W: 24.4-4216.9
Range BN: 6.3 - 600.3
(b)
NCLSH/Westat
log(BN) = 2.29 + 0.034*log(W)
n = 7 R2=0.001
Range W: 33.5-81.4
Range BN: 5.4 - 20.9
log(BN) = 0.504 +
0.447 'log(W)
n = 42 R2'0.476
Range W: 26.7-6196.8
Range BN: 4.2-139.5
log(BN) = 1.51 + 0.383*log(W)
n = 6 R2= 0.806
Range W: 229.3-47616.4
Range BN: 35.5 - 478.5
R&M Pilot
log(BN) = -2.52+1. 13'log(W)
n=24 R2=0.810
Range W: 7.6 - 6755
Range BN: 1-2164
log(BN) = -0.343 + 0.699 * log(W)
n=23 R2 =0.777
Range W: 3 - 424560
Range BN: 1 .4 - 8964
log(BN) = 3.25 + 0.692 *log(W|
n=24 R2 =0.438
Range W:' 7.1-335297
Range BN: 78-761842
(a) Log refers to natural logarithm
(b) Equation was not fitted because data were not available or insufficient.
(c) Units are pg/ft2 for vacuum and wipe dust-lead loadings.
Table 7. Regression Equations for Predicting Blue Nozzle Vacuum Lead Concentration
From Wipe Lead Loading.
Surface
Floors
I Uncarpeted)
Window Sills
Window Wells
Estimated Regression Model for . ..
Wipe Pb Loading to Blue Nozzle Vacuum Pb Concentration
CAPS Pilot
log""(BN) = 3.28+0.528'log
-------
Table 8. Regression Equations for Predicting Wipe Lead Loading from Blue Nozzle
Vacuum Lead Loading.
Surface
Floors
(Uncarpeted)
Window Sills
Window Wells
Estimstod RsQroraion Moctol for . .
Blue Nozzle Vacuum Pb Loading to Wipe Pb Loading10'
CAPS Pilot
log'a)(W) = 1. 64 + 0.965 'loglBN)
n = 6 R2= 0.683
Range Wlcl: 13.8-2498.5
Range BN: 1.9-149.1
log(W) = 1.36 + 1.04'loglBN)
n = 6 R2 =0.927
Range W: 24.4-4216.9
Range BN: 6.3 - 600.3
(b)
NCLSH/Westat
log(W) = 3.89+0.016*log(BN)
n = 7 R2 = 0.001
Range W: 33.5-81.4
Range BN: 5.4 - 20.9
loglW) = 2.03 + 1.07*log(BN)
n=42 R2'0.476
Range W: 26.7-6196.8
Range BN: 4.2-139.5
log(W) = -1.34+2.11*log(BN)
n = 6 R2 = 0.806
Range W: 229.3-47616.4
Range BN: 35.5 - 478.5
R&M Pilot
log(W) = 2.69+0.715"log(BN)
n = 24 R2 =0.810
Range W: 7.6 - 6755
Range BN: 1 - 2164
log(W) = 1.73 + 1.11*log(BN)
n = 23 R1 =0.777
Range W: 3 - 424560
Range BN: 1 .4 - 8964
log(W) = 2.37 + 0.63*log(BN)
n = 24 R2 =0.438
Range W: 7.1-335297
Range BN: 78-761842
(a) Log refers to natural logarithm
(b) Equation was not fitted because data were not available or insufficient.
(c) Units are //g/ft2 for vacuum and wipe dust-lead loadings.
Table 9. Final Conversion Equations for Uncarpeted Floors Based on Combined Data.
Conversion
Lead loading to lead loading""
Wipe to BN
BN to wipe
Lead loading to lead concentration""
Wipe to BN
n
37
37
30
Conversion Equation
BN = 0.131 W103
W = 11.4BN0690
BN = 34.2 W0-613
(a) Units are //g/ft2 for vacuum and wipe dust-lead loadings.
(b) Units are j/g/g for vacuum dust-lead concentrations
Draft - Do Not Cite or Quote
28
September 17, 1996
-------
Table 10. Final Conversion Equations for Window Sills Based on Combined Data.
Conversion
Lead loading to lead loading'"
Wipe to BN
BN to wipe
Lead loading to lead concentration""
Wipe to BN
n
71
71
29
Conversion Equation
BN = 0.828 W0-613
W = 5.79 BN1-08
BN = 115 W0451
(a) Units are ;/g/ft2 for vacuum and wipe dust-lead loadings.
(b) Units are //g/g for vacuum dust-lead concentrations.
Table 11. Final Conversion Equations for Window Wells Based on Combined Data.
Conversion
Lead loading to lead loading""
Wipe to BN
BN to wipe
Lead loading to lead concentration11"
Wipe to BN
n
30
30
24
Conversion Equation
BN = 5.63 W0457
W = 7.37 BN0-752
BN = 23.7 W0-673
(a) Units are pg/ft2 for vacuum and wipe dust-lead loadings.
(b) Units are /jg/g for vacuum dust-lead concentrations.
Draft - Do Not Cite or Quote
29
September 17, 1996
-------
Table 12. Predicted Vacuum Lead Loadings and Concentrations For Selected Wipe Lead
Loadings and Predicted Wipe Lead Loadings for Selected Blue Nozzle Lead
Loadings Based on Weighted Average of Regression Coefficients on Uncarpeted
Floors.
Wipe Pb Loading
fe/g/ft2)
10
40
100
200
500
1000
Blue Nozzle
Vacuum Pb
Loading
hig/ft2)
10
40
100
200
500
1000
Blue NozzleVacuum Loading (jig/ft2)
(95% Confidence Interval)
(95% Prediction Interval)
1.40
(0.802 , 2.45)
(0.190. 10.3)
5.80
(4.00 . 8.46)
(0.825,41.1)
14.9
(10.6,21.0)
(2.13, 105)
30.3
(20.7 , 44.7)
(4.30, 215)
77.6
(47.0, 129)
(10.7 , 565)
158
(85.5 , 294)
(12.1 . 1190)
Wipe Lead Loading (jig/ft2)
(95% Confidence Interval)
(95% Prediction Interval)
55.7
(43.9 , 70.2)
(10.6 , 292)
145
(110, 189)
(27.4 , 763)
273
(191 , 388)
(50.7 , 1460)
440
(284 , 677)
(80.3 , 2400)
828
(477 , 1430)
(146,4660)
1340
(704 , 2520)
(229 . 7750)
Blue NozzleVacuum Concentration (//g/g)
(95% Confidence Interval)
(95% Prediction Interval)
140
(97.6 , 202)
(39.2 , 503)
328
(255 , 423)
(94.2, 1140)
576
(459 , 723)
(166, 2000)
881
(689, 1130)
(253 , 3070)
1540
(1130,2110)
(438 , 5460)
2360
(1620, 3440)
(658 . 8500)
Draft - Do Not Cite or Quote
30
September 17, 1996
-------
Table 13. Predicted Vacuum Lead Loadings and Concentrations For Selected Wipe Lead
Loadings and Predicted Wipe Lead Loadings for Selected Blue Nozzle Lead
Loadings Based on Weighted Average of Regression Coefficients on Window
Sills.
Wipe Pb Loading
J/ig/ft2)
10
40
100
200
500
1000
Blue Nozzle
Vacuum Pb
Loading
(Aig/ft2)
10
40
100
200
500
1000
Blue NozzleVacuum Loading big/ft2)
(95% Confidence Interval)
(95% Prediction interval)
3.40
(2.41 , 4.79)
(0.572 , 20.2)
7.95
(6.29, 10.0)
(1.36,46.3)
13.9
(11.6, 16.7)
(2.40 , 80.7)
21.3
(18.0,25.2)
(3.68, 123)
37.4
(30.8 , 45.4)
(6.44, 217)
57.2
(45.2 , 72.3)
(9.80 . 333)
Wipe Lead Loading (pg/ft2)
(95% Confidence Interval)
(95% Prediction Interval)
69.4
(51.8,92.9)
(6.45 , 746)
310
(236 , 405)
(28.8 , 3320)
832
(583. 1190)
(76.6 . 9010)
1760
(1120,2740)
(159, 19300)
4720
(2640 , 8430)
(416, 53400)
9970
(5000, 19800)
(854, 116000)
Blue NozzleVacuum Concentration (/ig/g)
(95% Confidence Interval)
(95% Prediction Interval)
326
(182,585)
(39.7 , 2680)
610
(385 , 967)
(76.4 , 4870)
922
(617, 1380)
(117,7270)
1260
(866, 1840)
(161 ,9900)
1910
(1310,2770)
(243, 15000)
2610
(1760, 3870)
(331 . 20500)
Draft - Do Not Cite or Quote
31
September 17, 1996
-------
Table 14. Predicted Vacuum Lead Loadings and Concentrations For Selected Wipe Lead
Loadings and Predicted Wipe Lead Loadings for Selected Blue Nozzle Lead
Loadings Based on Weighted Average of Regression Coefficients on Window
Wells.
Wipe Pb Loading
(//g/ft2)
10
40
100
200
500
1000
Blue Nozzle
Vacuum Pb
Loading
(in/ft*)
10
40
100
200
500
1000
Blue NozzleVacuum Loading (//g/ft2)
(95% Confidence Interval)
(95% Prediction Interval)
16.1
(5.70 . 45.6)
(0.294 , 884)
30.4
(13.2,69.9)
(0.582, 1590)
46.3
(22.9 , 93.3)
(0.907 , 2350)
63.5
(34.5, 117)
(1.26, 3180)
96.6
(58.4, 159)
(1.95,4760)
133
(85.6 , 205)
(2.70 . 6490)
Wipe Lead Loading (//g/ft2)
(95% Confidence Interval)
(95% Prediction Interval)
41.7
(6.52 , 267)
(0.626 , 2780)
118
(26.4 , 532)
(2.05 , 6830)
235
(65.3 , 852)
(4.41 , 12600)
397
(128, 1240)
(7.78 , 20300)
790
(301 , 2080)
(16.2, 38600)
1330
(556 . 3200)
(27.9 . 63600)
Blue NozzleVacuum Concentration (//g/g)
(95% Confidence Interval)
(95% Prediction Interval)
112
(21.6, 579)
(2.66 , 4700)
284
(76.3 , 1060)
(7.72 , 10500)
526
(173, 1610)
(15.3, 18100)
838
(317,2230)
(25.5 , 27700)
1550
(687 , 3530)
(49.1 , 49300)
2480
(1190, 5170)
(79.9 . 77100)
Draft - Do Not Cite or Quote
32
September 17, 1996
-------
Table 15. Regression Equations for Predicting Wipe Lead Loading From BRM Vacuum Lead Loading.
Surface
Floors
(Uncarpeted)
Floors
(Carpeted)
Window Sills
Window Wells
Estimated Regression Model for
BRM Vacuum Pb Loading to Wipe Pb Loading
R&M Mini1"1
log""(W) = 2.24 +0.551 'log(BRM)
n = 25 R2= 0.689
Range W(cl: 13.0-8826
Range BRM: 1.1 -8337
(b)
log(W) = 2.51 -I- 0.650'loglBRM)
n = 27 R'=0.763
Range W : 7.6 - 87885
Range BRM -08- 41 69645
log(W) = 3.05 + 0.606 'log(BRM)
n = 25 R2 = 0.884
Range W : 145.1 - 572440
Range BRM : 2.4 - 4539847
NCLSH 5-Method"1
log(W) = 1.81 + 0.438Mog(BRM|
n= 68 R'= 0.493
Range W: 1.0-918
Range BRM : 0.5 - 7770
loglWI = -0.311 + 0.318'loglBRM)
n= 67 Rl=0.218
Range W : 0.8 - 60
Range BRM : 47.0 - 5640
(b)
IW
Rochester Lead-in-Dust
log(W) = 2.13 + 0.267*log(BRM)
n = 389 R2 = 0.197
Range W: 0.1 -18130
Range BRM : 0.1 - 74100
log(W) = 1.06 + 0.263'log(BRM)
n = 398 R2 = 0.162
Range W : 0.5 - 34600
Range BRM = 1.4-141000
log(W) = 2.97 -f 0.391 *log(BRM)
n = 362 R2 = 0.382
Range W : 0.4 - 42052
Range BRM : 0.3 - 230853
log(W) = 2.63 -f 0.560'loglBRM)
n = 403 RJ = 0.532
Range W: 2.7- 1361787
Range BRM: 1.9 -6610798
MQwaukee Low Cost Interventions'11
log(W) = 2.63 + 0.283Mog(BRM)
n = 135 R2 = 0.274
Range W : 0.4 - 636
Range BRM : 0.2 - 22616
(b)
Ib)
(b)
u
u
(a) Log refers to natural logarithm
(bl Equation was not fitted because data were not available or insufficient.
(c) Units are /ig/ft2 for vacuum and wipe dust-lead loadings.
(d) Wipe lead loadings in the R&M Mini study were multiplied by 1.58 to adjust the measure of bioavailable lead to total available lead.
(e) Wipe lead loadings in the NCLSH 5-Method study were multiplied by 1.25 to correct for only Including 80% of wipe samples in chemical analyses.
(f) Duplicate vacuum samples were excluded in determining the relationship.
8
0>
-------
Table 16. Final Conversion Equations Based on Combined Data for Uncarpeted Floors,
Carpeted Floors, Window Sills, and Window Wells.
Dwelling
Uncarpeted Floors
Carpeted Floors
Window Sills
Window Wells
n
617
465
389
428
Conversion Equation'8'
W = 8.79 BRM0313
W = 2.21 BRM0271
W = 1 7.0 BRM0421
W = 14.1 BRM0671
(a) Units are pg/ft2 for vacuum and wipe dust-lead loadings.
Table 17. Predicted Wipe Lead Loadings Based on Final Conversion Equations For
Selected BRM Vacuum Lead Loadings.
BRM Vacuum
Lead Loading
vvg/ft2)
10
40
100
200
500
1000
Predicted Wipe Lead Loading U/g/ft2) by Dwelling Component
(95% Confidence Interval)
(95% Prediction Interval)
Uncarpeted
Floors
18.1
(16.5, 19.9)
(2.05, 160)
27.9
(25.4 , 30.7)
(3.16,247)
37.1
(33.3,41.6)
(4.20 , 329)
46.1
(40.6 , 52.6)
(5.22 , 409)
61.4
(52.5 , 72.0)
(6.93 , 546)
76.3
(63.7, 91.6)
(8.60 . 679)
Carpeted
Floors
4.13
(3.41 , 5.00)
(0.557 , 30.6)
6.01
(5.28 , 6.84)
(0.814,44.4)
7.70
(6.96 , 8.52)
(2.05 , 56.8)
9.29
(8.47 , 10.2)
(1.26,68.4)
11.9
(10.7, 13.2)
(1.62,87.8)
14.4
(12.6, 16.3)
(1.95 . 105)
Window
Sills
44.8
(36.5, 55.1)
(3.83 , 525)
80.3
(68.7, 94.1)
(6.89 , 938)
118
(103, 135)
(10.1 , 1380)
158
(139, 180)
(13.6, 1840)
233
(205 , 265)
(20.0, 2710)
311
(271 , 359)
(26.7 . 3630)
Window
Wells
52.6
(36.5 , 75.9)
(1.57, 1760)
116
(85.1 , 159)
(3.48 , 3880)
196
(149 , 259)
(5.89 , 6520)
291
(226 . 375)
(8.77 . 9670)
491
(394 , 614)
(14.8, 16300)
730
(596 , 895)
(22.1 , 24200)
Draft - Do Not Cite or Quote
34
September 17, 1996
-------
o
o
I
O
'
•§.
D
I
I
3
IOO I 000
Uncarpeled Floor Lead Loading (ug/9q fl). Wipe Samples
CO
01
I
I
(o
3
i
E
1
IOO ) 000
Window Sill Lead Loading (ug/sq ft). Wipe Samples
Figure 4. Blue Nozzle Vacuum Lead Loading versus Wipe Lead Loading for Uncarpeted Floors and
Window Sills from the CAPS Pilot Study.
-------
o
o
O
8
10 100 J 000 10000
Uncarpeled Floor Lead Loading (ug/aq ft). Wipe Samples
10 100 1.000 10000
Window Sill Lead Loading (ug/sq HI Wipe Samples
U)
OJ
100 I 000 10000 100000
• Well Lead Loading (ug. aq ft) Wipe Samples
2 Figure 5. Blue Nozzle Vacuum Lead Loading versus Wipe Lead Loading for Uncarpeted Floors, Window Sills, and Window
to
Ol
Wells from the NCLSH/Westat Study.
-------
I
D
o
I
O
8
10 1OO I 000
Uncorpcltd Floor Und bonding (ug/sq ft). Wipe Samples
1
E
I
3
I
*
I 100 I.OOO 10000 IOOOOO 10000OO
»in,lr,w Sill Lead Loading (ug/sq ft). Wipe Samples
CJ
0 100 1000 10000 IOOOOO I 000 000
Window Well Lead Loading (ug/sq ft). Wipe Samples
- Figure 6. Blue Nozzle Vacuum Lead Loading versus Wipe Lead Loading for Uncarpeted Floors, Window Sills, and Window
Wells from the R&M Pilot Study.
0>
-------
o
o
o
(•V
CD
O
Uncarpeted Floor Lead Loading (ug/sq ft), BN Vacuum Samples
I
E
Window Siill Lead Loading dig 'sq fl), BN Vacuum Samples
<0
to
O)
Figure 7. Wipe Lead Loading versus Blue Nozzle Lead Loading for Uncarpeted Floors and Window
Sills from the CAPS Pilot Study.
-------
o
o
o
o
(S
O
-i
O
o
o- i ooo
Uncarpelsd floor Lead Loading (ug/.q (l| BN Vocuum Somplcs
Window Sill Lead Loading (ug/aq ft > BN v,,<-unm Samples
CO
CD
•^ 100 000
E
Wmdoo Well Lead Loading (ug'iq (I) BN Vacuum Samples
^ Figure 8. Wipe Lead Loading versus Blue Nozzle Vacuum Lead Loading for Uncarpeted Floors, Window Sills, and Window
$ Wells from the NCLSH/Westat Study.
-------
o
o
z
o
r-f
O
f*
n>
o
O
c
o
•; "*
to ]OO I 000 IOOOO
Uncarpeled Floor Lead Loading (ug/sq ft), BN Vacuum Samples
Window Sill Lead Loading (ug sq II). BN Vacuum Samples
I
•£_ 100 000
100 I OOO 10 000
Window Well Lead Loading (ug sq ft) BN Vacu
100 000 1 OOO OOO
i Samples
^ Figure 9. Wipe Lead Loading versus Blue Nozzle Vacuum Lead Loading for Uncarpeted Floors, Window Sills, and Window
£ Wells from the R&M Pilot Study.
-------
O
o
2
n
ft
o
-1
D
o
c
+
100 I 000
Uncerpeted Floor Lead Loading (ug/sq ft), Wipe Samples
I
~-»
N
to
O)
1
E
\
3
Window Sill Lead Loading (ug/sq ft) Wipe Samples
Figure 10. Blue Nozzle Vacuum Lead Concentration versus Wipe Lead Loading for Uncarpeted
Floors and Window Sills from the CAPS Pilot Study.
-------
s
D
S
I
Uricarpeted Floor Lead Loading (ug/aq 11). Wipe Samples
) IOO I.OOO 10 OOO 100000
Window Sill Lead Loading (UK sq ft) Wipe Samples
£ 100.000
I
0 100 1 DOO IOOOO 100 OOO I OOO 000
Window Well Lead Loading (ug 'aq flj Wipe Samples
2 Figure 11. Blue Nozzle Lead Concentration versus Wipe Lead Loading for Uncarpeted Floors, Window Sills, and Window
£ Wells from the R&M Pilot Study.
-------
o
o
2
O
I-*
O
S
o
0
c
o
1.000,000
I
m 10.000
a
E
o
O
01
-—
0!
1 ,000
10
+ + +
Combined Regression Line
CAPS Pilot, dato
NCLSH/Westot, Predicted
CAPS Pilot. Predicted
A A A R&M Pilot, dato
000
Upper Confidence Bound
NCLSH/Westat. data
R&M Pilot, Predicted
100 1.000 10.000
Lead Loading (ug/sq.ft), Wipe Samples
1,000.000 10.000,000
2 Figure *\2. Predicted BN Lead Loading Versus Wipe Lead Loading on Uncarpeted Floors, Individual and Combined
£> Regressions. Confidence Bands for Combined Regressions and Data Overlaid.
-------
O
o
f
n
ro
g
D
o
1.000,000
00
Cr
<6
Q.
E
D
I/I
E
D
D
O
D
01
01
H
O
~
O
L.
~E.
V
01
Combined Regression Line
+ +• + CAPS Pilot, data
R4M Pilot, Predicted
Lower Confidence Bound
CAPS Pilot, Predicted
Upper Confidence Bound
A A A R&M Pilot, data
100 1.000 10.000
Lead Loading (ug/sq.ft). Wipe Samples
100.000
1.000.000 10.000,000
to
to
O)
Figure 13. Predicted BN Lead Concentration Versus Wipe Lead Loading on Uncarpeted Floors, Individual and Combined
Regressions. Confidence Bands for Combined Regressions and Data Overlaid.
-------
o
o
o
r+
O
rV
0)
o
D
o
1.000.000
I
i
I
a
E
a
a
i/i
cr.
o
5
T3
0
a.
10.000
1,000
100
10
f 4- 4-
IM i
1
Combined Regression Line
CAPS Pilot, data
NCLSH/Westat, Predicted
. , , . n| i i— r-i-r-m
10 1
CAPS Pilot, Predicted C> O O
A A A R&M Pilot, data
)0 1.000 10,000
Upper Confidence Bound
NCLSH/Westat. data
R&M Pilot. Predicted
100,000 1.000.000
T —r T • r IT rr
10.000.000
Lead Loading (ug/sq.ft), Vacuum Samples
CT)
Figure 14. Predicted Wipe Lead Loading Versus BN Lead Loading on Uncarpeted Floors, Individual and Combined
Regressions. Confidence Bands for Combined Regressions and Data Overlaid.
-------
I
o
o
q
S
O
-1
i>
QJ
Q.
E
D
C
T>
O
-0
1.000.000
1.000 -
f + +
Combined Regression Line
CAPS Pilot, dota
NCLSH/Weslal. Predicled
CAPS Pilot. Predicted
A A A R&M Pilot, dota
Upper Confidence Bound
O O O NCLSH/Westol. doto
R&M Pilot. Predicted
"T"
in
1.000.000 10.000.000
Lead Loading (ug/sq.fl). Wipe Samples
- Figure 15. Predicted BN Lead Loading Versus Wipe Lead Loading on Window Sills. Individual and Combined Regressions.
£ Confidence Bands for Combined Regressions and Data Overlaid.
-------
D
o
o
rT
ro
o
D
o
1.000.000
i 10.000
Q.
E
o
CO
E
§ 1.000
tn
3
U
c
o
o
-f -t-
t CAPS Pilot, data
R4M Pilot. Predicted
CAPS Pilot. Predicted A A A R4M Pilot, data
1 ' ' ' 1
1
10
100 1.000 10.000 100,000 1.000.000 10
Lead Loading (ug/sq.ft), Wipe Samples
^ Figure 16. Predicted BN Lead Concentration Versus Wipe Lead Loading on Window Sills, Individual and Combined
!£ Regressions. Confidence Bands for Combined Regressions and Data Overlaid.
-------
o
o
o
r*
O
if
g
O
o
oo
D
I/)
HI
O.
D
D
100.000
- Combined Regression Line
+ + + CAPS Pilot, doto
- NCLSH/Westol, Predicted
Lower Confidence Bound
CAPS Pilot. Predicted
A A A R4M Pilot, dota
- Upper Confidence Bound
COO NCLSH/Westol. data
R&U Pilot. Predicted
T
100
1.000 10.000
Lead Loading (ug/sq.fl), Vacuum Samples
100,000
1.000.000 10.000.000
2 Figure 17. Predicted Wipe Lead Loading Versus BN Lead Loading on Window Sills, Individual and Combined Regressions.
£ Confidence Bands for Combined Regressions and Data Overlaid.
-------
o
o
z
o
r+
O
r*
CD
O
-s
D
o
!
V
u
E
0
o
en
c
'D
o
3
•o
o
Q)
1.000.000
100.000
10.000
1,000
0.01
- Combined Regression Line
O O O NCLSH/Westal. dolo
R&U Pilot. Prsdicled
Lower Confidence Bound
NCLSH/Weslol. Predicted
Upper Confidence Bound
A A A RiM Pilot, dala
TT1—
10
' ""I I ' I r-
100 1,000 10.000
Lead Loading (ug/sq.ft), Wipe Samples
100.000
1.000.000 10.000.000
2 Figure 18. Predicted BN Lead Concentration Versus Wipe Lead Loading on Window Wells, Individual and Combined
$ Regressions. Confidence Bands for Combined Regressions and Data Overlaid.
-------
O
o
o
o
rV
ra
o
o
r-f
03
01
o
o
(SI
E
3
O)
3
O
o
-o
10.000
Combined Regression Line
A A A RAM Pilot, dolo
Lower Confidence Bound
R&u Pilot. Predicted
Upper Confidence Bound
100 1,000 10.000
Lead Loading (ug/sq.ft). Wipe Samples
100.000
i.ooo.ooo to.ooo.ooo
2 Figure 19. Predicted BN Lead Concentration Versus Wipe Lead Loading on Window Wells, Individual and Combined
g Regressions. Confidence Bands for Combined Regressions and Data Overlaid.
-------
I
I
o
o
re
o
-i
D
o
Co
I
(6
10,000,000 -
1.000,000
1/1
Q.
CT
0>
CD
C
TJ
O
0.1
GOO
Combined Regression Line
NCLSH/Weslot. dolo
R&M Pilol. Predicted
- NCLSH/Westol. Predicted
Upper Confidence Bound
A A A R*U Pilot, doto
100 1.000 10.000 100.000
Lead Loading (ug/sq.ft), Vacuum Samples
i.ooo.ooo 10.000,000
2 Figure 20. Predicted Wipe Lead Loading Versus BN Lead Loading on Window Wells, Individual and Combined Regressions.
oj Confidence Bands for Combined Regressions and Data Overlaid.
-------
f
o
n
o
§
M 100
10 100 1.000 10 000
Uncarpeted Floor Lead Loading {ug/sq ft). BRM Vacuum Samples
I 10 100 1 OOO IO.OOD 100000 1 OOO OOO IGOOOOOQ
Window Sill Lead Loading (ug/sq ft) BRM Vacuum Samples
01
10 100 1,000 10OOO IOOOOO I OOO OOO 10 OOO 000
Window Well Lead Loading (ug/sq ft) BRM Vacuum Samples
Figure 2"\. Wipe Lead Loadings versus BRM Vacuum Lead Loadings for Uncarpeted Floors, Window Sills, and Window Wells
from the R&M Mini Survey.
-------
D
O
O
ri'
(D
O
D
O
a
on 100
§
1
0 1 10 100 1 000 10000
Uncarpeled Floor Lead Loading (ug/sq ft). BRM Vacuum Samples
Ul
CO
I
to
10
O)
0.
10 100 1.000 10000
Carpeted Floor Lead Loading (ug/sq ft). BRM Vacuum Samples
Figure 22. Wipe Lead Loading versus BRM Vacuum Lead Loading for Uncarpeted Floors and
Carpeted Floors from the NCLSH 5-Method Comparison Study.
-------
0>
O
o
Z
O
r-»
o
l-»
ro
g
D
o
0 I 10 100 1,000 10.000 100000
Uncarpeled Floor Lead Loading (ug/sq.fl), BRM Vacuum Samples
j^^vv ***; *: *
t , +» *
+ i **** **>*-> +
*;*$ **:/*'? >:
I 10 100 I OOO 10.000 IOOODO 1 OdO 000
Carpeted Floor Lead Loading (ug/sq ft), BRM Vacuum Samples
I
I 10 100 1000 10000 IOOOOO I 000 OOO
Window Sill Lead Loading (ug/sq fl). BRM Vacuum Samples
10 100 I OOO 10000 IOOOOO 1 OOO OOO IOOOOOOO
Window Well Lead Loading (ug/sq It). BRM Vacuum Samples
CO
Figure 23. Wipe Lead Loadings Versus BRM Vacuum Lead Loadings for Uncarpeted Floors, Carpeted Floors, Window Sills,
and Window Wells from the Rochester Lead-in-Dust Survey.
-------
u
o
z
o
o
rV
O
-1
D
c
0
F-*
o>
01
a\
%
Ts
1
er
-A
><
to
to
0)
1 000
m
a
£
(0
C/)
d>
°- 100
»
-•->
<*-
cr
-•W
g
1 *
on
| 10-
co
o
J
T3
CO
-** \ + + + ++
+ + ^^ + + +
+ -b^"*" "*" *"+*" + +
+ -W++ + + *
++ +
^ + +
+ ++ + +
+
+
+
+
-i-
0 1 10 100 1.000 10,000 100.000
Uncarpeted Floor Lead Loading (ug/sq.ft). BRM Vacuum Samples
Figure 24. Wipe Lead Loading versus BRM Vacuum Lead Loading for Uncarpeted Floors from the Milwaukee
Interventions Study.
-------
I
o
o
o
o
S
o
c?
o
01
O)
1,000.000
0)
a
E
o
01
a
o
10.000
A
i
^^
A
* fl * VW^;£MSV^~
A " A a a
rf ' a
A A ^
R&M Mini, Predicted O O O NCLSH 5-Method, doto
- NCLSH 5-M«1hod. Predicted A A A Rochester Lead-in-Dust, data
Rochester Leod-ln-Dust. Predicted O v O Milwaukee Low Cost Interventions, doto
— Milwaukee Low Cost Interventions, Predicted
' ' '"1 1 '"1 ' l''""| ' 1 1 -! TTITJ - ri | 1 .11(1111 , , ,
0-1 1 10 1 00 1 .000 1 0.000 1 00.000 1 .000.000 1 (
Lead Loading (ug/sq.ft). Vacuum Samples
^ Figure 25. Predicted Wipe Lead Loading versus BRM Lead Loading, Individual Regressions and Combined Results for
5£ Uncarpeted Floors. Confidence Intervals on Combined Results and Data Overlaid.
-------
I
D
O
O
r-t
CD
O
D
o
Go
O
(n
CT>
D
O)
C
10.000
100
- Combined Regression Line
+ + -I NCLSH 5-Melhod, data
- Rochester Leod-ln-Dust. Predicted
Lower Confidence Bound
NCLSH 5-Method, Predicled
Upper Confidence Bound
Rochester Lead-in-Dust, data
T~"
0.01
10 100 1,000 10.000
Lead Loading (ug/sq.ft), Vacuum Samples
100.000 i.ooo.ooo to.ooo.ooo
^ Figure 2G. Predicted Wipe Lead Loading versus BRM Lead Loading, Individual Regressions and Combined Results for
g Carpeted Floors. Confidence Intervals on Combined Results and Data Overlaid.
-------
1
0
0
2
s
i
t?
°
o
sr
CO
CO
ft
•8
~*
SJ
<0
to
0)
1.000,000
100.000
10.000
tfl
0)
"a
E
D
l/l
41 1.000 -
a
5
^5
cr
g 100
\
a>
3
en
1 10,
5
-0
3
1 :
0.1 -
0.01 -
+ +.
+ 4, '
+ / ' "*' '
A + 6 A A
-- ^^^~^ ^^^"^
& A 4 £1 ^^--^^^ ^r^-~~^' ^^-^~
A A" ^*~^ ^""^ ^~^"
A A a ^ A -^*~Ir~-*^\, '*~**'^
a * V'"" "a ^l^^-:^^^
& A * *' aAA& ^^-'^^^'^^^^^ A
A *A'h \ ^^^^"
A« 4*4 V ''i&iCiS^' * * » *
i a «f a A -fl a5,AiC-^«*^ " ^ ^ a a
• 4*^SiJ^^^ 1 "
4 ^A/A^yg^P%, "^ ^ a
^ • J*£li&F* ^ a "> 4 .
- =^^^V^ 1*-A AAa
b^-^^^Z^**'^ " A A A A ^
--^^^ ^+ A A ^ A ^ ^
-^•^ Aa A ?A^
a i "a
-w a A a
A
A
Combined Regression Line Lower Confidence Sound Upper Confidence Bound
4- -f + RAM Mini, doto RAM Mini. Predicted A A A Rochester Lead-in-Dust, doto
Rochester Lead-in-Oust. Predicted
^ 1 1 l| 1 | ' 1 1 1 MM| 1 ' 1 1 Ml!| 1 1 1 1 MM, 1 1 1 1 1 I.I! 1 1 1 1 1 Ml, , 1 1 1 1 1 M] 1 Try*
0.01 0.1 1 10 100 1,000 10.000 100.000 1.000,000 10.000,000
Lead Loading (ug/sq.ft), Vacuum Samples
Figure 27. Predicted Wipe Lead Loading versus BRM Lead Loading, Individual Regressions and Combined Results
Window Sills. Confidence Intervals on Combined Results and Data Overlaid.
-------
a
I
o
o
z
o
r-»
O
S
o
D
o
Ul
CO
to
10.000,000 -
1 .000,000
100.000
10,000
Ul
V
Q
^. 1 ,000
01
3
T>
o
o
_J
-D
u
Combined Regression Line
+ + + R&M Mini, doio
Rochester Leod- In-Oust. Predicted
Lower Confidence Bound
R&u Mini. Predicted
" Upper Confidence Bound
A A A Rochester Leod-ln-Dusf. data
T
0.01
rrr,
100
r,
10 100 1.000 10.000
Lead Loading (ug/sq.ft), Vacuum Samples
100,000 1.000.000 10.000.000
2 Figure 28. Predicted Wipe Lead Loading versus BRM Lead Loading, Individual Regressions and Combined Results for
£ Window Wells. Confidence Intervals on Combined Results and Data Overlaid.
-------
5.0 CONCLUSIONS AND DISCUSSION
The CAPS Pilot, R&M Pilot, and NCLSH/Westat studies provided the data used to determine
the relationship between wipe lead loading and Blue Nozzle vacuum lead loading and lead
concentration. Relationships were developed for uncarpeted floors, window sills and window
wells. No information was found for the relationship on carpeted floors.
The following equations were developed for converting a wipe lead loading (ug/ft2) to a BN
lead loading (fig/ft2):
Uncarpeted floors: BNload = 0.131 Wipe'-03
Window sills: BNload = 0.828 Wipe0613
Window wells: BNload = 5.63 Wipe0 4"
Thus, for example, a wipe lead loading of 100 fig/ft2 on an uncarpeted floor would be converted
to a BN lead loading of 14.9 ug/ft2, with an approximate 95% confidence interval of 10.6 to 21.0
fig/ft2, by applying the first of the above equations. Alternatively, 95% of the BN loadings
associated with a wipe loading of 100 ug/ft2 are predicted to be between 2.13 and 105 ug/ft2.
The following equations were developed for converting a wipe lead loading (fig/ft2) to a
BN lead concentration (ug/g):
Uncarpeted floors: BN^,. = 34.2 Wipe0 6I3
Window sills: BN^ =115 Wipe04SI
Window wells: BN,^ = 23.7 Wipe0673
Thus, a wipe lead loading of 100 ug/ft2 on an uncarpeted floor would be converted to a BN lead
concentration of 576 ug/g, with an approximate 95% confidence of 459 to 723 ug/g, by applying
the first of the above equations. On the other hand, 95% of the BN concentrations associated
with a wipe loading of 100 ug/ft2 are predicted to be between 166 and 2000 ug/g.
For converting a BN lead loading (ug/ft2) to a wipe lead loading (ug/ft2), the following
equations were developed:
Draft - Do Not Cite or Quote 60 September 17. 1996
-------
Uncarpeted floors: Wipe = 11.4 BN° m
Window sills: Wipe = 5.79 BN1 °8
Window wells: Wipe = 7.37 BN0752
Thus, a BN lead loading of 100 ug/ft2 on an uncarpeted floor would be converted to a wipe lead
loading of 273 fig/ft2, with an approximate 95% confidence interval of 191 to 388 ug/ft2, by
applying the first of these equations. However, 95% of the wipe loadings associated with a BN
loading of 100 ug/ft2 are predicted to be between 50.7 and 1,460 ug/ft2.
The third set of equations are preferred for converting a Blue Nozzle lead loading to a
wipe loading. The first set of equations will not be used in Section 403 rulemaking.
The R&M Mini, NCLSH 5-Method Comparison, Rochester Lead-in-Dust, and
Milwaukee Low Cost Interventions studies provided the data used to determine the relationship
between wipe lead loading and BRM vacuum lead loading. The following conversion equations
were developed based on data combined across the relevant studies:
Uncarpeted floors: Wipe = 8.79 BRM0313.
Carpeted floors: Wipe = 2.21 BRM0271.
Window sills: Wipe = 17.0 BRM0421.
Window wells: Wipe = 14.1 BRM°S71.
Table 17 summarizes the mean predicted wipe lead loadings, along with approximate 95%
confidence and prediction intervals, associated with measured BRM lead loadings of 10,40,100,
200,500, and 1000 ug/ft2 based on these equations.
Draft - Do Not Cite or Quote 61 September 17. 1996
-------
REFERENCES
[1] "Lead-Based Paint Abatement and Repair and Maintenance Pilot Study." July, 1992. Draft
Final Report prepared by Battelle for the U.S. Environmental Protection Agency under
Contract No. 68-DO-0126.
[2] NCLSH Westat Blue Nozzle Study. Prepared by Westat Inc. Revised January, 1995.
[3] "Comprehensive Abatement Performance Pilot Study." Volumes 1 and 2. February, 1995.
EPA 747-R-93-007.
[4] "The Relation of Lead-Contaminated House Dust and Blood Lead Levels Among Urban
Children." Volumes I And II. June, 1995. Final Report to the U.S. Department of Housing
and Urban Development Grant from The University of Rochester School of Medicine,
Rochester, New York, and The National Center for Lead-Safe Housing, Columbia
Maryland.
[5] "Comparison of Five Sampling Methods for Settled Lead Dust: a Pilot Study." June, 1993.
Preliminary Draft Report prepared by The National Center for Lead-Safe Housing,
Columbia Maryland.
[6] All floor samples from the Milwaukee Low Cost Interventions study are uncarpeted per
personal conversation with Brad Schultz, statistician, USEPA.
[7] "Seasonal Rhythms of Blood-Lead Levels: Boston, 1979-1983," EPA Report No. 747-R-
94-003,1995.
[8] Draper, N.R., and Smith, H., Applied Regression Analysis. 1981, Wiley.
Draft - Do Not Cite or Quote 62 September 17, 1996
-------
APPENDIX A
Distribution of the Data used to Develop
the BN Vacuum Conversion Equations
Draft - Do Not Cite or Quote A-1 September 17, 1996
-------
DISTRIBUTION OF DATA
Table A-l presents the number of uncarpeted floor samples by groupings of wipe lead
loading and vacuum lead loading for the CAPS Pilot study. Similarly, Table A-2 provides
combined information for uncarpeted and carpeted floors, window sills, and window wells. In
Tables A-3 through A-6, results are presented in a similar manner for the R&M Pilot and
NCLSH/ Westat studies. As seen in the tables, the bulk of the data for uncarpeted floors are
available for vacuum lead loadings less than 50 ug/ft2 and wipe lead loadings less than 100
ug/ft2. Data are present at the higher ends of the ranges for window sills and window wells.
Table A-1. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading CAPS
Pilot Study — Uncarpeted Floors.
Wipe Lead Loading
(//g/ft2)
0-50
150-200
200 +
Total
Blue Nozzle Vacuum Lead Loading U/g/ft2)
0-50
5
0
0
5
50-100
0
0
0
0
100-150
0
0
1
1
200 +
0
0
0
0
Total
5
0
1
6
Table A-2. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading, CAPS
Pilot Study — Uncarpeted Floors and Sills.
Wipe Lead Loading
(Jig/ft2)
0-50
1 50-200
200 +
Total
Blue Nozzle Vacuum Lead Loading (//g/ft2)
0-50
7
1
1
9
50-100
0
1
0
1
100-150
0
0
1
1
200 +
0
0
1
1
Total
7
2
3
12
Draft - Do Not Cite or Quote
A-2
September 17. 1936
-------
Table A-3. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading, R&M
Pilot Study — Uncarpeted Floors.
Wipe Lead
Loading
Uvg/ft2)
0-25
25-50
50-75
75-100
100-150
150-200
200 - 400
400 +
Total
Blue Nozzle Vacuum Lead Loading (/ig/ft2)
0-25
9
0
4
0
2
1
0
0
16
25-50
0
0
0
0
0
0
1
0
1
50-75
0
0
0
0
0
0
1
0
1
75-100
0
0
0
0
0
0
0
1
1
100-150
0
0
0
0
0
1
0
0
1
150-200
0
0
0
0
0
0
0
0
0
200-400
0
0
0
0
0
0
0
0
0
400 +
0
0
0
0
0
1
0
3
4
Total
9
0
4
0
2
3
2
4
24
Table A-4. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading, R&M
Pilot Study — Uncarpeted Floors, Sills, and Wells.
Wipe Lead
Loading
-------
Table A-5. Number of Samples by Wipe Lead Loading and Vacuum Lead Loading,
NCLSH/Westat Study - Uncarpeted Floors.
Wipe Lead
Loading
-------
Tables A-7 through A-10 display the distribution of data by groupings of wipe lead loading
and BN vacuum lead concentration for the CAPS Pilot and R&M Pilot studies. Tables A-7 and
A-8 show that most of the data from the CAPS Pilot study have wipe lead loadings less than 50
ug/ft2, while Tables A-9 and A-10 indicate that a large portion of the R&M Pilot study has
vacuum lead concentrations greater than 400 ug/g.
Table A-7. Number of Samples by Wipe Lead Loading and Vacuum Lead Concentration,
CAPS Pilot Study — Uncarpeted Floors.
Wipe Lead
Loading
fo/g/ft2)
0-50
50-200
200 +
Total
Blue Nozzle Vacuum Lead Concentration (/ug/g)
50-100
2
0
0
2
100-200
1
0
0
1
200-300
1
0
0
1
300-400
1
0
0
1
400 +
0
0
1
1
Total
5
0
1
6
Table A-8. Number of Samples by Wipe Lead Loading and Vacuum Lead Concentration,
CAPS Pilot Study — Uncarpeted Floors and Sills.
Wipe Lead
Loading
(/ig/ft2)
0-50
50-200
200 +
Total
Blue Nozzle Vacuum Lead Concentration (//g/g)
50-100
2
0
0
2
100-200
1
1
0
2
200-300
1
0
0
1
300-400
2
0
0
2
400 +
1
1
3
5
Total
7
2
3
12
Draft - Do Not Cite or Quote
A-5
September 17, 1996
-------
Table A-9. Number of Samples by Wipe Lead Loading and Vacuum Lead Concentration,
R&M Pilot Study — Uncarpeted Floors.
Wipe Lead
Loading
fcig/ft2)
0-25
25-50
50-75
100-150
150-200
200-400
400 +
Total
Blue Nozzle Vacuum Lead Concentration (//g/g)
75-100
1
0
0
0
0
0
0
1
100-150
1
0
0
0
0
0
0
1
150-200
3
0
0
0
0
0
0
3
200-400
3
0
2
0
0
0
0
5
400 +
1
0
2
2
3
2
4
14
Total
9
0
4
2
3
2
4
24
Table A-10. Number of Samples by Wipe Lead Loading and Vacuum Lead Concentration,
R&M Pilot Study - Uncarpeted Floors, Sills, and Wells.
Wipe Lead
Loading
(fig/ft2)
0-25
25-50
50-75
100-150
1 50-200
200-400
400 +
Total
Blue Nozzle Vacuum Lead Concentration (f/g/g)
75-100
1
0
0
0
0
0
0
1
100-150
2
0
0
0
0
0
0
2
150-200
4
0
0
0
0
0
0
4
200-400
6
0
2
0
0
2
1
11
400 +
5
0
4
3
5
6
30
53
Total
18
0
6
3
5
8
31
71
Draft - Do Not Cite or Quote
A-6
September 17, 1996
-------
APPENDIX B
Residual Analyses and Influential Observations
for Blue Nozzle Vacuum Conversion Equations
Draft - Do Not Cite or Quote B-1 September 17, 1996
-------
Residual Analysis and Influential Observations
for the BN-Wipe Conversion Equations
When conducting a regression analysis, it is important to note any points that are
influential. There are several approaches to determining influential data points. One measure of
influence is known as the DFBETA statistic, which is a scaled measure of the influence of any
one data value upon the separate parameter estimates. For a simple linear regression, DFBETA
is calculated for the intercept (a) and for the slope (P), respectively. The statistic is a measure of
the difference between the parameter estimate (a or P) as calculated by including all data values
and as calculated by excluding the i* data value. The threshold value for determining which
points are most influential is recommended by Belsley, Kuh, and Welsch to be 2'. Thus, for each
regression performed in this analysis, DFBETA was calculated for a and P, and compared to the
threshold value of 2. In the figures that follow, the data for the individual regression is plotted
with triangles indicating the observations with significant influence on the intercept, squares
indicating the observations with significant influence on the slope, and circles indicating the
observations with significant influence on both the intercept and the slope.
There are also underlying assumptions that are made when conducting a regression analysis
that must be validated. One such assumption, is that the variability in the data remains constant
across the range of the data. Residual plots (i.e., residual values (log scale) versus predicted
values) have been provided for all BN regressions performed. A random scatter of the points on
the graph around the reference line at zero with similar range across predicted values is an
indication that the assumption of the homogeneity of variance is satisfied. However, if there are
more data points at one predicted value than at another, the range should be wider because there
is greater likelihood of observing more extreme values with more observations.
Figure B-l displays the data, with indicated influential observations, for predicting a BN
lead loading from a wipe lead loading from uncarpeted floors for CAPS Pilot, NCLSH/Westat,
and R&M Pilot. Figure B-2 presents the corresponding residuals versus the predicted BN
loadings from the regression of BN loading on wipe loading for uncarpeted floor samples from
1 SAS/STAT User's Guide. Version 6, Fourth Edition, Volume 2,1990, SAS Institute
Inc., pg 1419.
Draft - Do Not Cite or Quote B-2 September 17, 1996
-------
CAPS Pilot, NCLSH/Westat, and R&M Pilot. Figures B-3 and B-4 present similar information
for predicting a wipe loading from a BN loading from uncarpeted floors, while Figures B-5 and
B-6 provide the information for predicting a BN concentration from a wipe loading from
uncarpeted floors. Analogous information is displayed in Figures B-7 through B-12 for window
sills and Figures B-13 through B-18 for window wells.
It should be noted that only 5 observations are influential in any of the BN regressions. In
all cases, the data point is influential to the estimate of both the intercept and the slope. All of
the influential observations are in the CAPS Pilot or NCLSH/Westat study. Each of the
influential observations had predicted value significantly distant from the other observations in
the dataset, most often on the nigh side. This is not surprising given the limited amount of data
available in these two studies. Each of the BN regressions involving uncarpeted floor samples
has an influential observation from the CAPS Pilot study (See Figures B-l, B-3, and B-S). Note
that this is only one value from the data set. The regression of BN concentration on wipe loading
on samples from window sills in the CAPS Pilot study (Figure B-l 1) also has one influential
observation. Finally, the NCLSH/Westat regression of wipe lead loading to BN lead loading on
window wells has one observation that is influential to the intercept and the slope as well (See
Figure B-l 5). Note that if an influential observation is removed, and the regression repeated, the
parameter estimates from that study may change substantially. However, the methodology
employed in this analysis to combine the parameter estimates from the individual studies,
weights each estimate according to the inverse of the uncertainty associated with it. Therefore, if
a study has little data or much variability in its parameter estimates, removing an influential point
from that study will not significantly affect the combined parameter estimates.
For instance, consider the regression of wipe lead loading on BN vacuum lead loading for
uncarpeted floors shown in Figure B-3. When the influential point in the CAPS Pilot Study is
removed, the estimates of a and P change dramatically for this study. The estimate of a
including the point is 4.22 with a standard error of 0.51. By excluding the observation, the
estimate of a becomes 3.28 with a standard error of 0.19. Likewise, the estimate of P is 0.97
with a standard error of 0.33 with the influential observation and 0.17 with a standard error of
0.15 without the influential observation. In addition, the estimates of a in the other two studies
Draft - Do NOT Cite or Quote B-3 September 17, 1996
-------
change minimally due to the centering of the data by the overall mean log BN vacuum lead
loading which was altered slightly by the removal of the data value. However, the combined
conversion equation is not significantly affected by this deletion. The equation based on all data
points is W = 11.4 BN° 69°. With the removal of the influential observation, the equation
becomes W = 11.8 BN°579. The final conversion equations are similar because the uncertainty
associated with the slope in the CAPS Pilot Study (using all observations) is high, thereby
lessening its contribution to the combined slope estimate.
It must be noted, however, that there is no basis for eliminating this point (or any of the
other influential observations) from this analysis. The design of the analysis minimizes the effect
of such observations.
Regarding the residual plots, no patterns or trends can be seen, and variability appears
constant across predicted values for each of the regressions performed.
Draft - Do Not Cite or Quote B-4 September 17, 1996
-------
I
\
I
5
01 1 10 100 1 0(
CAPS Uncarpeted Floors Wipe Lead Loading (ug/ftS)
100 1000
NCLSH Uncarpeled Floors Wipe Lead Loading (ug/ft2)
01
• • • Non-Influent, al Points - . . Non - Influential Points
• ' • Points Influencing Alpha . . . po,nls Influencing Alpha
- • • Points Influencing Beta . „ » Points Influencing Beta
0 • • Points Influencing Alpha and Beta - - - points Influencing Alpha and Beta
~ 10.000
3
BO
c
| 1.000
1
E
3
a
a IDO
z
ID
u
a
\L
•a 10
i
|
1
D
S i
£
-*.
+ *-*
-* -»-
*
+ + 4
+
*4+
+
+
+
*** 4*
RM Pilot Uncarpeted Floors Wipe Lead Loading (ug/ft2)
Non-InTupnlinl Pnmts
Points Influencing Alpha
Points Influencing B
-------
CD
a
i
CD
o
I
CD
O)
§
I
(6
NCLSH Uncarpeted Floors Predicted BN Vacuu
I.OOO 10 OOO
i Lead Loading (ug-'ft2)
10 1 DO I OOO 10 000
RM Pilot Uncarpeled floors Predicted BN Vacuum Lend Loading (ug/fl2)
>i Figure B-2. Residual (Log Scale) versus Predicted Blue Nozzle Vacuum Lead Loading from the Regression of Blue Nozzle
- Lead Loading on Wipe Lead Loading for Uncarpeted Floors from CAPS Pilot, NCLSH/Westat, and R&M Pilot
to
O)
Studies.
-------
I
o
o
o
5
10 too i ooo
CAPS Uncerpeted Floors BN Vacuum Lead Loading {ug/U2)
Non-Influent IB] Points
• Points Influencing Alpha
1 Points Influencing Beta
1 Points Influencing Alpha and Beta
NCLSH Uncarpeled Floors BN Vacuum Lead Loading (ug/fl2)
Non-Influential Points
Points Influencing Alpha
1 Points Influencing Beta
Points Influencing Alpha and Beta
1OO I.OOO 10OOO
RM Pilot Uncarpeled Floors BN Vacuum Lead Loading (ug/rtZ)
• Non-Influential Points
* * ' Points Influencing Alpha
' ° Points Influencing Beta
*_ ° Points Influencing Alpha and Beta
^ Figure B 3. Influential Observations in the Regression of Wipe Lead Loading on Blue Nozzle Vacuum Lead Loading for
§ Uncarpeted Floors from CAPS Pilot, NCLSH/Westat, and R&M Pilot Studies.
-------
^
I
I
O
'
o
I
CAPS Uncarpeted Floors Predicted Wipe Lead Loading (ug/ft2)
NCLSH Uncarpeted Floors Predicted Wipe Lead Loading (ug/ft2)
00
I
•-*
^
10 100 I 000 10.000
RM Pilol Uncarpeled Floors Predicted Wipe Lead Loading (ug/fl2)
Figure B-4. Residual (Log Scale) versus Predicted Blue Nozzle Vacuum Lead Loading from the Regression of Wipe Lead
Loading on Blue Nozzle Lead Loading for Uncarpeted Floors from CAPS Pilot, NCLSH/Westat, and R&M Pilot
Studies.
-------
O
5'
O
§
10 \00 1.000
RM Pilot Uncarpeted Floors Wipe Lead Loading (ug/ft2)
Non - Influential Points
Points Influencing Alpha
Points Influencing Beta
Points Influencing Alpha and Beta
DO
cb
I
CAPS Uncarpet-ed Floors Wipe Lead Loading {ug ft2)
Non-Influential Points
Points Influencing Alpha
Points Influencing Beta
Points Influencing Alpha and Beta
Figure B-5. Influential Observations in the Regression of Blue Nozzle Lead Concentration on Wipe
Lead Loading for Uncarpeted Floors from the R&M Pilot and CAPS Pilot Studies.
-------
o
3"
o
§
10 100 1 000 10 000
RM Pilol Uncarpeled Floors Predicted BN Vacuum Lead Cone (ug/g)
cp
O
I
I
to
CO
0>
10 100 1.000 10 000
CAPS Uncnrpeted Floora Predicted BN Vacuum Lead Cone (ug/g)
Figure B-6. Residual (Log Scale) and Predicted Blue Nozzle Lead Concentration from the
Regression of Blue Nozzle Vacuum Lead Concentration on Wipe Lead Loading for
Uncarpeted Floors from the R&M Pilot and CAPS Pilot Studies.
-------
I
D
o
100 1.000 10.000 100.000
CAPS Window Sills Wipe Lead Loading (ug/ftS)
Non-Influential Points
Points Influencing Alpha
Points Influencing Beta
Points Influencing Alpha and Beta
10 100 1.000 10.000 10
NCLSH Window Sills Wipe Lead Loading (ug/fl2)
Non-Influential Points
Points Influencing Alpha
Points Influencing Beta
joints Influencing Alpha and Beta
-------
CD
o
CAPS Window Sills Predicted BN Vacuum Lead Loading (ug/fl2)
*
a '
10 100 i ooo 10.000
NCLSH Window Sills Predicted BN Vacuum Lead Loading (ug/fl2)
RM Pilot Window Sills Predicted BM Va
I 000 ] 0 000
uum I,f«ri Loading (ug/f(2)
2 Figure B-8. Residual (Log Scale) and Predicted Blue Nozzle Vacuum Lead Loading from the Regression of Blue Nozzle Lead
£ Loading of Wipe Lead Loading for Window Sills from the CAPS Pilot, NCLSH/Westat, and R&M Pilot Studies.
-------
CD
3
^
i
o
O
8'
D
§
a
s. 10
CAPS Window Sills BN Vacuum Lead Loading (ug/ft2)
Non-Influential Points
Points Influencing Aipha
r on.Is Influencing Brio
Points Influencing Alpha and Beta
NCLSH Window Sills BN Vacuum Lead Loading (ug/f(2)
• • • Non-Influential Points
• • • Points Influencing Alpha
B ° ° Points Influencing Alpha and Beta |
!
10 too 1000 10000
RM Pilot Window Sills BN Vacuum Lead Loading (ug/ft2)
Non-Influential Points
Points Influencing Alpha
Points Influencing Beta
Points Influencing Alpha and Beta
^ Figure B 9. Influential Observations in the Regression of Wipe Lead Loading on Blue Nozzle Vacuum Lead Loading for
^ Window Sills from the CAPS Pilot, NCLSH/Westat, and R&M Pilot Studies.
-------
8
4
3
— 2
3
-o
in
CC
" 0
o
•o
c
* '
OT
Cu
<
0 2
-3
+
^ *
1 ID 100 1 000 IOOOO 100000 1 OODOOO
4
3
"ffl 3
3
TJ
! '
to
* 0
c
"o
a!
Q£
-3
-4
4
3
"5 z
^
"H
b
o; i
CO
o
"U
^ - 1
£
U
2: -3
-3
4
+
+
*
4- *
•>* ^,
* t »4 «*
+ + .
+ ^
t » 4-
•"»
1 10 100 t 000 10.000 100 OOO 1,000 OOO
NCLSH Window Sills Predicted Wipe Lead Loading (ug/ftZ)
*
4 + +
* +
+
+ +
*
*
+
* +
1 10 100 1 OOO IOOOO 100.000 1 OOO 000
^ Figure B-10. Residual (Log-Scale) and Predicted Wipe Lead Loading from the Regression of Wipe Lead Loading on Blue
§ Nozzle Vacuum Lead Loading for Window Sills from the CAPS Pilot, NCLSH/Westat, and R&M Pilot Studies.
-------
I
o
-1
0 IOQ I.OOO IOOOO IOO.OOO \ OOO COO
RM Pilot Window Sills Wipe Lead Loading (ug/fl3)
• • Non-Influential Points
* * Points Influencing Alpha
" ° Points Influencing Beta
0 ° Points Influencing Alpha and
CD
cn
s
£
f
100 1000 10000
CAPS Window Sills Wipe Lead Loading (ug
100.000 1 OOO 000
Points Influencing Alpha
Points Influencing Beto
Points Influencing Alpha and Beta
Figure B-11. Influential Observations in the Regression of Blue Nozzle Vacuum Lead
Concentration on Wipe Lead Loading for Window Sills from the R&M Pilot and CAPS
Pilot Studies.
-------
CD
o
I
I
10 100 1000 10.000
RM Pilot Wmdow Sills Predicted BN Vacuum Lead Cone (ug/g)
CO
cn
I
I
"•*
N
ID 100 1.000 10 000 100 000
CAPS Window Sills Predicted BN Vacuum Lead Cone (ug/g)
Figure B-12. Residual (Log Scale) and Predicted Blue Nozzle Lead Concentration from the
Regression of Blue Nozzle Vacuum Lead Concentration on Wipe Lead Loading for
Window Sills from the R&M Pilot and CAPS Pilot Studies.
-------
~ I OOO 000
o
o
I
~*
o
I
10 OOO
1 00 OOO
U 1 DO I OOO 10 OOO 10O
RM Pilot Window Wells Wipe Lead Loading
-------
IE) 100 I OOO
RM Pilot Window Wella Predicted BN Vacu
10 ooo i oo ooo i ooo ouo
i Lead Loading d.« ft.')
00
00
NCLSH Window Wells Predicted BN Vacu
10 000 100.000 1.000000
um I cad Loading (ug/fte)
to
Co
01
Figure B-14. Residual (Log Scale) and Predicted Blue Nozzle Lead Loading from the Regression of
Blue Nozzle Vacuum Lead Loading on Wipe Lead Loading for Window Wells from the
R&M Pilot and NCLSH/Westat Studies.
-------
I
_3 10 ooo
RM Pilot Window Well
000 10000 100.000
BN Vacuum Lead Loading (ug fi2)
Non-lnlluential Points
Points Influencing Alpha
Points Influencing Beta
Points Influencing Alpha and Beta
i ooo 10 ooo 100.000
NCLSH Window Wells BN Vacuum Lead Loading (ug fL2)
(• • • Non^ influential Points
• - - Pomls Influencing Alpha
000 Points Influencing Beta
0 Points Influencing Alpha and Hela
Figure B-15. Influential Observations in the Regression of Wipe Lead Loading on Blue Nozzle
Vacuum Lead Loading for Window Wells from the R&M Pilot and NCLSH/Westat
Studies.
-------
I
o
o
0
§
10 100 I 000 10 000 IOO 000 ) 000 000
RM Pilot Window Wells Predicted Wipe Lead Loading fug/ft?)
CD
NJ
O
-.
*
NCLSH Window Wells Predicted Wipe Lend Loading tug/!t2)
(o
CO
O)
Figure B 16. Residual (Log Scale) and Predicted Wipe Lead Loading from the Regression of Wipe
Lead Loading on Blue Nozzle Vacuum Lead Loading for Window Wells from the R&M
Pilot and NCLSH/Westat Studies.
-------
^
I
-------
I
o
o
5
O
I
td
3
"U
'55
OJ
3 -
2
OD
NJ
o
T3
c
+ +
o
ex
s
K
— 3
*
I
-5-
10 100 1.000 10.000 100.000
RM Pilot Window Wells Predicted BN Vacuum Lead Cone (ug/g)
1.000.000
- Figure B 18, Residual (Log Scale) and Predicted Blue Nozzle Vacuum Lead Concentration from the Regression of Blue Nozzle
£ Vacuum Lead Concentration on Wipe Lead Loading for Window Wells from the R&M Pilot Study.
-------
APPENDIX C
BRM Vacuum Conversion Equations Based
on Data Pooled Across Studies
Draft - Do Not Cite or Quote C-1 September 17, 1996
-------
Simple Regression Results Based on Data Pooled Across Studies
As an alternative approach to combining the results across studies, it was requested that a
simple log-linear regression be performed on the data pooled across studies, without regard to
possible differences across studies. Table C-l presents the conversion equations based on this
analysis. However, we recommend use of the equations presented in Table 16 in the main body of
this report.
Table C-1. Regression Equations for Predicting Wipe Lead Loading From BRM Vacuum Lead
Loading based on Data Pooled Across Studies.
Surface
Floors
(Uncarpeted)
Floors
(Carpeted)
Window Sills
Window Wells
Estimated Regression Model for
BRM Vacuum Pb Loading to
Wipe Pb Loading1"
Pooled Estimates
n = 617
n=465
n = 389
n=428
W = 7.92 BRM"60
R2 = 0.338
W = 2.97 BRM0229
R2=0.118
W= 13.6 BRM0-477
R2= 0.488
W = 12.6 BRM0-577
R2= 0.569
(a) Units are //g/ft2 for vacuum and wipe dust-lead loadings.
Draft - Do Not Cite or Quote
C-2
September 17, 1996
-------
APPENDIX D
Residual Analyses and Influential Observations
for BRM Vacuum Conversion Equations
Draft - Do Not Cite or Quote D-1 September 17. 1996
-------
Residual Analysis and Influential Observations
for the BRM/Wipe Conversion Equations
When conducting a regression analysis, it is important to note any points that are influential.
There are several approaches to determining influential data points. One measure of influence is
known as the DFBETA statistic, which is a scaled measure of the influence of any one data value
upon the separate parameter estimates. For a simple linear regression, DFBETA is calculated for the
intercept (a) and for the slope (P), respectively. The statistic is a measure of the difference between
the parameter estimate (a or P) as calculated by including all data values and as calculated by
excluding the i* data value. The threshold value for determining which points are most influential is
recommended by Belsley, Kuh, and Welsch to be 2'. Thus, for each regression performed in this
analysis, DFBETA was calculated for a and P, and compared to the threshold value of 2. In the
figures that follow, the data for the individual regression is plotted with triangles indicating the
observations with significant influence on the intercept, squares indicating the observations with
significant influence on the slope, and circles indicating the observations with significant influence on
both the intercept and the slope.
There are also underlying assumptions that are made when conducting a regression analysis
that must be validated. One such assumption, is that the variability in the data remains constant
across the range of the data. Residual plots (i.e., residual values (log scale) versus predicted values)
have been provided for all BRM regressions performed. A random scatter of the points on the graph
around the reference line at zero with similar range across predicted values is an indication that the
assumption of the homogeneity of variance is satisfied. However, if there are more data points at one
predicted value than at another, the range should be wider because there is greater likelihood of
observing more extreme values with more observations.
Figure D-l displays the data, with indicated influential observations, for predicting a wipe
lead loading from a BRM vacuum lead loading from uncarpeted floors for the R&M Mini, NCLSH
5-Method Comparison, Rochester Lead-in-Dust, and Milwaukee Low Cost Interventions studies.
Figure D-2 presents the corresponding residuals versus the predicted wipe loadings from the
1 SAS/STAT User's Guide. Version 6, Fourth Edition, Volume 2,1990, SAS Institute
Inc., pg 1419.
Draft - Do Not Cite or Quote D-2 September 17. 1996
-------
regression of wipe loading on BRM loading for uncarpeted floor samples from the four studies
included in the analysis. Figures D-3 and D-4 present similar information for carpeted floors.
Analogous information is displayed in Figures D-5 and D-6 for window sills and Figures D-7 and D-
8 for window wells. Note that there are no influential data values, as defined by DFBETA * 2, for
any of the parameter estimates of the regressions done on the BRM studies. Notice also that no
patterns or trends can be seen in the residual plots from the BRM regressions. Thus, the analysis of
influential observations and residuals did not reveal any problems with the approach used to develop
the BRM conversion equations.
Draft - Do Not Cite or Quote D-3 September 17, 1996
-------
CD
5
i
O
o
~»
O
o
D
o
— 100 ooo
R&U Mini Uncarpeled Floors BRM Vacuum Lead Loading (ug/112)
Non -Influential Points
Points Induem mg Beta
Points Influencing Alpha and Beta
NCLSH Uncarpeted Floors BRM Vacuum Lead Loading (ug/f(2)
Points Influencing Beta
Points Influencing Alpha and Beta
Co
100000
10000 100000
Rochester Uncarpeted Floors BRM Vacuum Lead Loading (ug/ft2)
:::
Non-Influential Points
Points Influencing Alpha
Points Influencing Beta
Beta
01 1 tO 100 I 000 10000 100000
Milwaukee Uncarpeled Floors BRM Vacuum Lead Loading (ug-'ftZ)
Non-Influential Points
' Points Influencing Alpha
1 Points Influencing Beta
to
Figure D-1. Influential Observations in the Regression of Wipe Lead Loading on BRM Vacuum Lead Loading for Uncarpeted
Floors from the R&M Mini, NCLSH 5-Method Comparison, Rochester Lead-in-Dust, and Milwaukee Low Cost
Interventions Studies.
-------
Q
o
o
3f
o
O
§
R&M Mini Uncarpeled Floors Predicted Wipe Lead Loading lug/112)
»
.*•>; .
NCLSH Uncarpeted Floors Predicted Wipe Lead Loading (ug, ftl')
Rochester Uncarpeted Floors Predicted Wipe Lead Loading (ug/ft2)
Milwaukee Uncarpeted Floors Predicted Wipe Lead Loading (ug/fl2)
|
I
^ Figure D-2. Residual (Log Scale) and Predicted Wipe Lead Loading from the Regression of Wipe Lead Loading on BRM
^ Vacuum Lead Loading for Uncarpeted Floors from the R&M Mini, NCLSH 5-Method Comparison, Rochester
j£ Lead-in-Dust, and Milwaukee Low Cost Interventions Studies.
-------
I
\
D
§
-C- 10000
9
•a
a
-J 1 000
E
3
i
* .v Vw **t>****** *
RochesLer Carpeted Doors BRM Vacuum Leed Loading (ug/ft2)
Non-lnfluenlial Points
Points Influencing Alpha
Points Influencing Beta
Points Influencing Alpha and Beta
9
O)
I
ID 100 1(100 IO.OOO 100000
NCLSH Carpeted Floors BRM Vacuum Lead Loading (ug 'ft2)
Points Influencing Alpha
Points Influencing Beta
Points Influencing Alpha and Bela
to
(o
O)
Figure D-3. Influential Observations in the Regression of Wipe Lead Loading on BRM Vacuum
Lead Loading for Carpeted Floors from the Rochester Lead-in-Dust, and NCLSH 5-
Method Comparison Studies.
-------
I
o
o
I
D
§
Rochester Carpeted Floors Predicted Wipe Lead Loading (ug/ft2)
!
O)
g. -
« -2
NCLSH Carpeted Floors Predicted Wipe Lead Loading (ug/fl2)
Figure D-4. Residual (Log Scale) and Predicted Wipe Lead Loading from the Regression of Wipe
Lead Loading on BRM Vacuum Lead Loading for Carpeted Floors from the Rochester
Lead-in-Dust, and NCLSH 5-Method Comparison Studies.
-------
CD
O
O
3
O
§
.,;•:«? "••.'••
I 10 100 I 000 10 000 100 000 I MO OOi
Rochester Window Sills BRM Vacuum Lead Loading (ug/fl2)
Non-Influential Points
Points Influencing Alpha
Points Influencing Beta
Points Influencing Alpha and Beta
o
do
s
1
O)
0000 10 000 OOO
R&M Mini Window Sills BRM Vacuum Lead Loading (ug/ft2)
• Non-Influential Points
4 Points Influencing Alpha
0 Points Influencing Beta
0 Points Influencing Alpha and Beta
Figure D-B. Influential Observations in the Regression of Wipe Lead Loading on BRM Vacuum
Lead Loading for Window Sills from the Rochester Lead-in-Dust, and R&M Mini
Studies.
-------
o
5
!
, 4/"* . t
•***. • "• *'•.'••
* i,:.'.
to 130 i ouo IOQOO 100 ooo t ooo ooo
Rochester Window Sills Predicted Wipe Lead Loading (ug/fl2)
9
(O
I
!
10 I 000 10000 100 000 I 000 OOO
R&M Mini Window Sills Predicted Wipe Lead Loading (ug/fl2)
Figure D-6. Residual (Log Scale) and Predicted Wipe Lead Loading from the Regression of Wipe
Lead Loading on BRM Vacuum Lead Loading for Window Sills from the Rochester
Lead-in-Dust, and R&M Mini Studies.
-------
— 10 000 000
i
o
o
_ I OOC OOO
O
§
10 100 1.000 10000 100000 1 000 000
Rochester Window Wells BRM Vacuum Lead Loading (ug/ft2)
Non-Influential Points
Points Influencing Alpha
Points Influencing Beta
Points Influencing Alpha end Beta
o
o
I
o-
3;
(o
to
10 IOO I 000 10000 100000 I OOO 000 10000000
R&M Mini Window Wells BRM Vacuum Lead Loading (ug, ft2)
Non-Influential Points
Points Influencing Alpha
Points Influencing Beta
Points Influencing Alpha and Beta
Figure D-7. Influential Observations in the Regression of Wipe Lead Loading on BRM Vacuum
Lead Loading for Window Wells from the Rochester Lead-in-Dust, and R&M Mini
Studies.
-------
o
5?'
I
(0 4
3
•o
M 3
- */ **7V*
tooo toovs 100000 i oooooo
Rochester Window Wells Predicted Wjpe Lead Loading (ug/fl2)
1
.N
to
R&rM Mini Window Wells Predicted Wipe Lead Loading (ug/ft2)
Figure D-8. Residual (Log Scale) and Predicted Wipe Lead Loading from the Regression of Wipe
Lead Loading on BRM Vacuum Lead Loading for Window Wells from the Rochester
Lead-in-Dust, and R&M Mini Studies.
-------
September 19, 1996
OPTIONS FOR SECTION 403 STANDARDS
FOR UNCARPETED FLOOR DUST, WINDOW SILL DUST,
DRIPLINE AND PLAY AREA SOIL, AND DETERIORATED PAINT
APPENDIX Y
Note to EPA: Draft Report is currently undergoing peer review
-------
DISCLAIMER
The material in this document has not been subject to Agency
technical and policy review. Views expressed by the authors are their own
and do not necessarily reflect those of the U.S. Environmental Protection
Agency. Mention of trade names, products, or services does not convey,
and should not be interpreted as conveying, official EPA approval,
endorsement, or recommendation. Do not quote or cite this document.
This report is copied on recycled paper.
Draft -Do Note Cite or Quote u September 19, 1996
-------
AUTHORS AND CONTRIBUTORS
This study was funded and managed by the U.S. Environmental Protection Agency. The
analysis was conducted by Battelle Memorial Institute under contract to the Environmental
Protection Agency. Each organization's responsibilities are listed below.
Battelle Memorial Institute (Battelle)
Battelle was responsible for evaluating candidate studies and analyses, for data
management and implementation of statistical analyses, and for overall production of the report.
Battelle authors were John Menkedick, Warren Strauss, Pam Hartford, Shawna Decker, and
Halsey Boyd.
U.S. Environmental Protection Agency (EPA)
The Environmental Protection Agency was responsible for managing the review,
providing guidance on the objectives for the review and report, contributing to the development
of conclusions and recommendations, and coordinating the EPA and peer reviews of the draft
report. In addition, EPA provided access to study results not yet available in the general
literature. The EPA Work Assignment Managers were Janet Remmers, John Schwemberger, and
Bradley Schultz. The EPA Project Officer was Sineta Wooten.
Draft -Do Note Cite or Quote iii September 19, 1996
-------
TABLE OF CONTENTS
EXECUTIVE SUMMARY vii
1.0 INTRODUCTION 1
2.0 STUDIES INCLUDED 2
2.1 ROCHESTER LEAD-IN-DUST STUDY 2
2.2 BALTIMORE REPAIR AND MAINTENANCE STUDY 3
3.0 VARIABLES INCLUDED 4
4.0 ANALYSIS METHODS 10
4.1 REGRESSION MODELS 10
4.2 SENSITIVITY/SPECIFICITY ANALYSES 10
5.0 RESULTS OF SINGLE MEDIA REGRESSION ANALYSES 15
6.0 RESULTS OF MULTI-MEDIA REGRESSION ANALYSES 17
7.0 RESULTS OF SINGLE MEDIA SENSITIVITY/SPECIFICITY ANALYSES 22
8.0 RESULTS OF MULTI-MEDIA SENSITIVITY/SPECIFICITY ANALYSES 23
8.1 DISCUSSION 27
9.0 CONCLUSION 27
Appendix A. Figures for the Single Media Regression Analyses A-1
Appendix B. Tables and Figures for the Multi-Media Regression Analyses B-1
Appendix C. Tables and Figures for the Single Media Sensitivity/Specificity
Analyses C-1
Appendix D. Tables for the Two-Way Multi-Media Sensitivity/Specificity Analyses . . D-1
Appendix E. Tables for the Three-Way Multi-Media Sensitivity/Specificity Analyses . . E-1
Appendix F. Results of Adding Window Sill Dust-Lead to the Multi-Media
Sensitivity/Specificity Analyses F-1
Appendix G. Tables for the Four-Way and Five-Way Multi-Media Sensitivity/
Specificity Analyses G-1
Appendix H. Summary of Estimated Standards Achieving Current Target Health
Effects H-1
Draft -Do Note Cite or Quote iv September 19, 1996
-------
TABLE OF CONTENTS
(Continued)
LIST OF TABLES
Table 1. Proposed Options for Section 403 Standards To Be Evaluated in the Risk
Assessment and Economic Analysis viii
Table 2. Estimated Percentage of 1-2 Year Old Children Above Thresholds of 10 and
15 //g/dL for Different House Age Categories in the NHANES III and
Rochester Studies 3
Table 3. Description of the Environmental Variables Used from the Rochester Lead-
in-Dust Study 5
Table 4. Description of the Calculated Lead-Based Paint Hazard Variables from the
Rochester-Lead-ln-Dust Study 7
Table 5. Description of the Variables Used in the Sensitivity/Specificity Analyses
of the Data from the Baltimore Repair and Maintenance Study 8
Table 6. Description of the Calculated Lead-Based Paint Hazard Variables for the
Data from the Baltimore Repair and Maintenance Study 9
Table 7. Definitions of Performance Characteristics Used to Characterize the
Empirical Data from Lead Exposure Studies to Aid in Defining the Likely
Performance of Different Health Based Standards 10
Table 8. Sample 95% Confidence Intervals for Selected Sensitivity/Specificity
Performance Measures and Varying Sample Sizes 14
Table 9. Estimated Environmental Lead Level at Which 80 and 95 Percent of
Children's Blood-Lead Concentrations will be Below Target Levels (10 and
15 //g/dL) Based on Single-Media Log-Linear Regression Models of the
Rochester Lead-in-Dust Study Data 15
Table 10. Parameter Estimates and Associated Standard Errors for Single Media Log-
Linear Models of the Relationship Between Blood-Lead Concentration and
Measures of Environmental-Lead from the Rochester Lead-in-Dust Study. ... 16
Table 11. Parameter Estimates and Associated Standard Errors from Fitting the Multi-
Media Regression Models to the Rochester Lead-in-Dust Study Data 18
Table 12. Proportion of the Distribution of Childhood Blood-Lead in Rochester That
Falls Below Target Level (10//g/dL), Assuming a Residential Unit with No
Deteriorated Lead-Based Paint, and Specified levels of Dripline Soil-Lead and
Uncarpeted Floor Dust-Lead, and Using Model Based On Indication of
Paint/Pica Hazard as Paint Variable8 20
Table 13. Proportion of the Distribution of Childhood Blood-Lead in Rochester That
Falls Below Target Level (10//g/dL ), Assuming a Residential Unit with No
Deteriorated Lead-Based Paint, and Specified levels of Play Area Soil-Lead
and Uncarpeted Floor Dust-Lead, and Using Model Based On Indication of
Paint/Pica Hazard as Paint Variable8 21
Table 14. Summary of Estimated Standards Which Achieve a Negative Predictive
Value of 95% or an Estimated 95% Probability of a Child's Blood-Lead
Concentration Below 10 //g/dL in a Dwelling that is at or Below the
Standard 29
Draft -Do Note Cite or Quote
September 19, 1996
-------
TABLE OF CONTENTS
(Continued)
Table 15. Proposed Options for Section 403 Standards To Be Evaluated in the Risk
Assessment and Economic Analysis 30
LIST OF FIGURES
Figure 1. Example of an Ideal Situation for Establishing Potential Dripline Soil-Lead
Standards 12
Figure 2. Example of a Situation Where the Negative Predictive Value and Sensitivity
Equal 100%, but the Positive Predictive Value and Specificity are Less than
100% 12
Draft -Do Note Cite or Quote vi September 19, 1996
-------
EXECUTIVE SUMMARY
The purpose of this report is to provide a range of options for the Section 403 standards
which can be farther evaluated in the risk assessment and economic analysis. The risk
assessment and economic analysis will characterize the risk reduction and determine net benefits
associated with each option.
The report focuses on options for the standards developed from models and analyses of
two epidemiological studies, the Baltimore Repair and Maintenance Study (R&M) and the
Rochester Lead-in-Dust Study (Rochester). The Rochester study was the primary source for
defining the options for standards in this document.
Two main analysis methods were used to estimate options for the standards from the
epidemiological data: regression models and sensitivity/specificity analyses. Each method was
applied to individual standards separately (single media analysis) and to a combination of
standards jointly (multi-media analysis).
It is assumed that individual standards will be set for each media, location, or surface
addressed in the rulemaking, and action will be recommended if any individual standard is
exceeded. Therefore, ranges are provided for each media, location, or surface separately. It is
also assumed that dependencies between media will be addressed in guidance to risk assessors
and in recommended actions associated with each standard.
Ranges for options of standards presented in this report are based on estimated health
effects. The high and low values for each range result from different analyses. The low values
are generally taken from the conservative single media analyses and the high values from the less
conservative joint analyses. Yet all options proposed are estimated to provide a minimum 95%
probability that a child living in a home with environmental levels below each of the Section 403
standards will have a blood-lead concentration less than 10 ng/dL.
Table 1 presents a range of standard levels which achieved (for homes that meet the
standard) either:
1. A Negative Predictive Value of 95% from a single media or multi-media
sensitivity/specificity analysis, or
2. An estimated 95% probability that a child's blood-lead concentration is below 10
(ig/dL from the single media or multi-media regression analyses.
The above two criteria were the target health effects for use in informing the choice of standards.
For each standard the upper limit of the range is specified as the maximum estimated level in one
of the multi-media analyses.
Draft -Do Note Cite or Quote vii September 19. 1996
-------
The lower limit of the standard is more problematic. Very low levels of standards (e.g.
less than 25 ng/ft2 for dust and less than 50 ug/g for soil) were estimated by the single media
analyses to be associated with the target health effect. From a practical standpoint, these
standards are not necessarily achievable. Therefore, the lower limits presented in Table 1 for the
range of options for a standard are based on practicality.
A formal analysis of the "net benefits" associated with different options for the standards
will be conducted in the risk assessment and economic analysis. The final results of all analyses
will be used, along with policy input, by EPA's risk management team to define standards that
achieve Title X's purposes: "to implement, on a priority basis, a broad program to evaluate and
reduce lead-based paint hazards in the nation's housing" and "to encourage effective action to
prevent childhood lead poisoning by establishing a workable framework for lead-based paint
hazard evaluation..." (Title X, Section 1003).
Table 1. Proposed Options for Section 403 Standards To Be Evaluated in the Risk
Assessment and Economic Analysis.
Standard
Uncarpeted Floor Dust (/sg/ft2 )
Window Sill Oust (/t/g/ft2 )
Dripline Soil (//g/g)
Play Area Soil (fjg/g)
Maximum of Percent of Interior
Components with Deteriorated Lead-
Based Paint and Percent of Exterior
Components with Deteriorated Lead-
Based Paint
Range
Low Limit
25
25
50
50
0%
High Limit
400
800
1500
1000
20%
Draft -Do Note Cite or Quote
VIII
September 19, 1996
-------
Options for Section 403 Standards
for Uncarpeted Floor Dust, Window Sill Dust,
Dripline and Play Area Soil, and Deteriorated Paint
1.0 INTRODUCTION
The purpose of the Section 403 rulemaking is to identify what constitutes a lead-based
paint hazard, i.e., a condition that causes exposure to lead from lead-contaminated dust, lead-
contaminated soil, or lead-contaminated paint that would result in adverse human health effects.
In particular, the objective for Section 403 is to set standards (condition and location of paint,
and levels of lead in dust and soil) against which to compare a residential environment when
evaluating the presence, scope, and magnitude of lead-based paint hazards.
Different options for the standards will be evaluated in a documented risk assessment and
economic analysis to characterize the risk reduction and to determine net benefits associated with
each option. The purpose of this document is to provide a range of options for the standards
which can be further evaluated in the risk assessment and economic analysis.
There are two primary methods/tools that have been considered for evaluating options for
the Section 403 standards: 1) models or analyses from epidemiological studies that have taken
measurements of children's blood-lead and levels of lead at their primary residence; and 2) the
Integrated Exposure Uptake Biokinetic Model (IEUBK) that implements a mathematical
structure designed to mimic a child's exposure to environmental lead, the uptake of that lead into
his or her body, and the biokinetic passages that lead follows after its uptake.
This report in its current form focuses on options for the standards developed from two
epidemiological studies, the Baltimore Repair and Maintenance Study (R&M) and the Rochester
Lead-in-Dust Study (Rochester). However, the Rochester study is the primary source for
defining the options for standards in this report.
Two main analysis methods were used to estimate options for the standards from the
epidemiological data: regression models and sensitivity/specificity analyses. Each method was
applied to individual standards separately (single media analysis) and to a combination of
standards jointly (multi-media analysis). It should be noted that blood-lead concentration is
significantly related to many different measures of potential lead exposure. These measures of
lead exposure are also correlated with each other and with other factors related to lead exposure
such as housing characteristics (e.g., age of house) and socio-economic status characteristics
(e.g., income). Therefore, the degree of association between any single environmental lead
variable and blood-lead concentration, is not necessarily the degree to which that single variable
causes a change in blood-lead concentrations.
We assume in this document that individual standards will be set for each media,
location, or surface addressed in the rulemaking, and action will be recommended if any
Draft - Do Not Cite or Quote 1 September 19, 1996
-------
individual standard is exceeded. Therefore, ranges are provided for each media, location, or
surface separately. It is assumed that dependencies between media will be addressed in guidance
to risk assessors and in recommended actions associated with each standard.
Ranges for options of standards presented in this report are based on estimated health
effects. The high and low values for each range result from different analyses. The low values
are generally taken from the conservative single media analyses and the high values from the less
conservative joint analyses. Yet all options proposed are estimated to provide a minimum 95%
probability that a child living in a home with environmental levels below each of the Section 403
standards will have a blood-lead concentration less than 10 ug/dL.
2.0 STUDIES INCLUDED
Two epidemiological studies, the Rochester Lead-in-Dust Study (Rochester) and the
Baltimore Repair and Maintenance Study (R&M), were used to develop options for the Section
403 standards. This section describes advantages and limitations of each study.
2.1 ROCHESTER LEAD-IN-DUST STUDY
There are both positive and negative aspects to the use of the Rochester Study. The
positive aspects are:
1. All media, locations, and surfaces that are being considered for Section 403
standards were measured for lead in the Rochester study.
2. The Rochester study includes dust-lead loadings and the Section 403 dust standard
is expected to be based on dust-lead loading from wipe sampling.
3. The selection of homes and children in the Rochester study, although targeted, was
more random and more representative of a general population than is the case with
most epidemiological studies of lead exposure.
4. The Rochester Study is recent.
The primary limitation associated with the Rochester Study is concern over the degree to
which the Rochester Study may be considered representative of the nation as a whole. The
Rochester Lead-in-Dust Study was a targeted sample which was limited to a single geographic
area. The sample consisted of children who had moderate exposure to lead in their home
environment, and did not necessarily include children with very high or very low exposure to
lead. This statement is based on the fact that:
Draft - Do Not Cite or Quote 2 September 19. 1996
-------
a. Almost one-quarter (22.9%) of the Rochester Study children (aged 1 -2'/2 years)
had blood-lead concentrations above 10 ng/dL, whereas only 11.5% of children
(aged 1-2 years) nationwide were estimated to have blood-lead concentrations
above 10 ug/dL by NHANES HI.
b. There were two children (out of 205 children, 0.09%) with a blood-lead
concentration above 30 ug/dL. Nationwide, an estimated 0.44% of children are
estimated to have blood-lead concentrations above 30 ug/dL by NHANES ffl.
c. Approximately 84% of the housing included in the survey was built prior to
1940, compared to an estimated 20% of housing nationwide that was built prior
to 1940 (estimated for 1997 from the American Housing Survey). There is a
well documented relationship between age of housing and presence of lead-
based paint.
d. Environmental levels of lead in soil in the Rochester study were higher than
would be expected by the HUD National Survey. For example, the geometric
mean dripline soil-lead concentration in the HUD National Survey was
approximately 75 jag/g while the Rochester geometric mean soil concentration
was approximately 730 ug/g.
e. In characterizing the health basis for standards in terms of the expected
probability that a child's blood-lead concentration is below 10 ug/dL, it should
be noted that the most recent NHANES in data estimates that 91% of children
aged 1-5 and 88.5% of children aged 1-2 currently have blood-lead
concentrations below 10 ng/dL (as of 1991). Table 2 below presents estimates
of the percentage of children below 10 ug/dL for different housing categories
from the NHANES III study along with the observed percentages of children
below 10 ug/dL for the same housing categories in the Rochester Study.
Table 2. Estimated Percentage of 1-2 Year Old Children Above
Thresholds of 10 and 15 //g/dL for Different House Age
Categories in the NHANES III and Rochester Studies.
Housing
Category
Pre-1946
Pre-1973
Post-1973
All Subjects
Percent of 1-2 Year Olds Above Indicated Target (10 or 15//g/dU'
NHANES III
* 10/ig/dL
25%
17%
4%
11.5%
z 15/jg/dL
8%
5%
1%
3.5%
Rochester
2 10 //g/dL
26%
15%
0%
23%
i 15 //g/dL
9%
8%
0%
8%
1 The percentages for NHANES III are based on national population using NHANES III weights. The
percentages for Rochester are the observed percentages of children in the Rochester study with
180 children in pre-1946 homes. 195 in pre-1973 homes, and 205 children overall. Whereas 95%
of children in the Rochester study lived in pre-1973 homes, NHANES III estimates of the number of
children represented by its study that live in pre-1973 homes was only 60%.
Draft - Do Not Cite or Quote
September 19, 1996
-------
f. Other limitations include the fact that the effect of race on blood-lead is very
influential in the Rochester Study, and may not be transferable nationwide.
Approximately 40% of the children in the Rochester Study were African-
American, whereas only approximately 13% of the national population is
African-American (U.S. Census).
In spite of these limitations, the Rochester Study may be considered one of the best
sources of data for evaluating national standards.
2.2 BALTIMORE REPAIR AND MAINTENANCE STUDY
The primary advantages of the R&M Study are that it includes 1) a targeted population of
residences with a known lead problem, 2) homes with higher dust-lead levels than observed in
the Rochester study, and 3) a wider spread in children's blood-lead concentration than observed
in the Rochester study. The Baltimore Repair and Maintenance Study (R&M) data has the
following limitations:
a. All dust samples were collected by the BRM vacuum method rather than wipe
sampling.
b. The sample size of the R&M study homes used in the analyses ranges from 27 to 87
homes dependent on the combination of media assessed.
c. The representativeness of the R&M study relative to the nation as a whole is
unknown.
d. The R&M study did not include the same requirements concerning the length of time
at the primary residence and limitation of exposures from non-residential sources
that were present in the Rochester Study.
3.0 VARIABLES INCLUDED
Various measures of environmental lead in soil, dust, and paint were included in the
regression and sensitivity/specificity analyses. Following is a list of the variable definitions.
Soil 1. Play-Area Soil-Lead Concentration (Fine Soil Fraction)
2. Dripline Soil-Lead Concentration (Fine Soil Fraction)
Floor 3. Area Weighted Arithmetic Mean (Wipe) Dust-Lead Loading
from Uncarpeted Floors
4. Area Weighted Arithmetic Mean (Wipe) Dust-Lead Loading
from Carpeted Floors
5. Area Weighted Arithmetic Mean (Wipe) Dust-Lead Loading
from Any Floor (Carpeted or Uncarpeted)
Draft - Do Not Cite or Quote 4 September 19, 1996
-------
Window Sill
6. Area Weighted Arithmetic Mean (Wipe) Dust-Lead Loading
from Window Sills
Window Trough 7.
Paint
Area Weighted Arithmetic Mean (Wipe) Dust-Lead Loading
from Window Troughs
8. Percentage of Interior Components with Deteriorated LBP:
The percent of interior components with deteriorated LBP
within a home is calculated as the number of interior
components within a home with deteriorated LBP (XRF
reading 2 1 and the paint condition is rated as Fair or Poor)
divided by the number of interior components sampled within
the home.
9. Percentage of Exterior Components with Deteriorated LBP:
The percent of exterior of components with deteriorated LBP
on a home is calculated as the number of exterior components
on a home with deteriorated LBP (XRF reading st 1 and the
paint condition is rated as Fair or Poor) divided by the number
of exterior components sampled on the home.
10. Average Percent of Deteriorated LBP on Interior Components:
The average percent of deteriorated LBP per interior
component within a home, was based on the presence of LBP
and the condition of the paint. If no LBP was present on a
component (XRF reading < 1) then the percent of deteriorated
LBP for a component is 0%. Otherwise, if LBP was present
(XRF reading * 1), then the percent of deteriorated LBP for a
component is 2.5% if the condition of the paint is Good, 10%
if the condition is Fair, and 20% if the condition is Poor. The
average percent of deteriorated LBP is then the sum of the
percent of deteriorated LBP over all interior components
within a home divided by the number of interior components
within a home.
11. Average Percent of Deteriorated LBP on Exterior
Components: The average percent of deteriorated LBP per
exterior component on a home, was based on the presence of
LBP and the condition of the paint. If no LBP was present on
a component (XRF reading < 1) then the percent of
deteriorated LBP for a component is 0%. Otherwise, if LBP
was present (XRF reading * 1), then the percent of
deteriorated LBP for a component is 2.5% if the condition of
the paint is Good, 10% if the condition is Fair, and 20% if the
condition is Poor. The average percent of deteriorated LBP is
then the sum of the percent of deteriorated LBP over all
exterior components on a home divided by the number of
exterior components on a home.
Draft - Do Not Cite or Quote
September 19. 1996
-------
12. Indicator of interior paint/pica hazard: An indicator variable
representing a nonzero value whenever each of the following
conditions existed in a residential unit:
a. Presence of deteriorated or damaged interior paint in the
household; and
b. Presence of interior lead-based paint in the household; and
c. Presence of a child with paint pica in the household.
13. Indicator of exterior paint/pica hazard: An indicator variable
representing a nonzero value whenever each of the following
conditions existed in a residential unit:
a. Presence of deteriorated or damaged exterior paint on the
house; and
b. Presence of exterior lead-based paint on the house; and
c. Presence of a child with paint pica in the household.
All of the above variables were calculated for the Rochester data. The final two variables
in the list, the paint/pica hazard indicators, were only used in the regression analyses. Table 3
lists the environmental media from the Rochester study included in the regression and
sensitivity/specificity analyses, along with the type of sampling method employed and the
number of samples available for analysis from each media. Table 4 provides a detailed
description of the paint hazard variables calculated for the Rochester Study.
Table 3. Description of the Environmental Variables Used from the Rochester
Lead-in-Dust Study
Study
Rochester Lead-in-
Dust
Location of Environmental Sample
Dripline Soil-Lead Concentration
Play Area Soil-Lead Concentration
Uncarpeted Floor Dust-Lead Loading
Carpeted Floor Dust-Lead Loading
Combined Uncarpeted/Carpeted Floor
Dust-Lead Loading
Window Sill Dust-Lead Loading
Window Trough Dust-Lead Loading
Sampling
Method
Core
Core
Wipe
Wipe
Wipe
Wipe
Wipe
Number of Samples
(Household
Averages) Available
for Analysis
186
87
197
179
205
196
189
Draft - Do Not Cite or Quote
September 19, 1996
-------
i
Table 4. Description of the Calculated Lead-Based Paint Hazard Variables from the Rochester-Lead-ln-Dust
Study.
Lead-Based Paint
Hazard Variable
Indication of
paint/pica hazard
Percent of
Components with
Deteriorated LBP
Average Percent
of Deteriorated
LBP per
Component8
Definition or trie
Condition of Paint
Good: Less than 5% of the
surface is deteriorated
Fair: Between 5% and
1 5% of the surface is
deteriorated
Poor: More than 15% of
the surface is deteriorated
Deteriorated LBP Indicator
Value
0
1
2
0
1
0%
2.5%
10%
20%
Definition
No LBP (XRF reading < 1) on the
component or condition of the paint is
Good, or child exhibits no paint pica
LBP (XRF reading zl) present on the
component, and condition of the paint is Fair
or Poor, and child rarely exhibits paint pica
LBP (XRF reading 2 1 ) present on the
component, condition of the paint is Fair or
Poor, and child exhibits paint pica at least
sometimes
No LBP (XRF reading < 1) on the
component or condition of the paint is Good
LBP (XRF reading * 1 ) present on the
component surface and the condition of the
paint is characterized as Fair or Poor
No LBP (XRF reading < 1) on the
component
LBP (XRF reading 2 1) and the condition of
the paint is categorized as Good
LBP (XRF reading 2 1 ) and the condition of
the paint is categorized as Fair
LBP (XRF reading 2 1) and the condition of
the paint is categorized as Poor
Calculation of the Lead-Based Paint
Hazard Variable
MAXIMUM of Deterioreted LBP Indicator
across all components sampled.
E Deteriorated LBP Indicator for
each component within a home - IQQ
9 components sampled within a home
E Deteriorated LBP Indicator for
each component within a home
ff components sampled within a home
1
For Average Percent of Deteriorated LBP per Component, the values of the Deteriorated LBP Indicator were chosen to be the midpoints of the corresponding paint
condition variable. For example, for paint characterized as Fair, 10% represents the average deterioration if deterioration is assumed to be uniform from 5% to
15%.
-------
In the Rochester data, the dust wipe samples were analyzed using either flame atomic
absorption or graphite furnace atomic absorption (GFAA) spectroscopy. If a sample was found
to be below the detection limit using the flame atomic absorption method, then GFAA was used
to analyze the sample. As a result, the detection limit for the dust wipe samples (variables 3
through 7 above) was < 0.25 ug/sample. Flame atomic absorption was used to analyze the soil
samples resulting in a detection limit of 10 ug/g. For the Rochester data, one floor dust sample
(out of 817), and one window sill dust sample (out of 363) were below the detection limit of 0.25
fig/sample. These two sample results were still included in the statistical analyses.
The R&M study data were only used in the sensitivity/specificity analyses. The data used
in the analyses included only the pre-intervention environmental levels for occupied homes
having a child with a reported blood-lead concentration. When multiple children within a home
had blood-lead samples reported, the blood-lead concentration for the youngest child in the home
was used. Note that only a subset of the variables listed above were included in the analyses.
Soil samples were taken at the dripline, property line, and "other" locations within the yard.
Only one sample was taken at the property line and the location of the "other" samples was
unclear. Therefore, only those samples taken at the dripline were included in the analysis
(variable 2 above). The dust samples in the R&M study represent composite samples, making it
impossible to distinguish uncarpeted floor samples from carpeted floor samples. For this reason,
variables 5,6, and 7 above were the only dust variables investigated. Notice, however, that each
of these variables is defined as a wipe dust-lead loading. The dust loading samples in the R&M
study were collected using the BRM vacuum. For this analysis, these BRM vacuum loadings
were converted to wipe loadings using the conversion equations listed in Table 5. Table 5 also
gives a more detailed description of the environmental variables used from the R&M study.
In the R&M data, the condition of the paint was coded differently than in the Rochester
study. It could only be determined if the paint was Intact or Non-Intact. Therefore, only
variations of variables 8 and 9 listed above could be used in the sensitivity/specificity analyses of
the R&M data. Only components with a paint condition of Intact or Non-Intact and available
XRF readings were included in the analysis. Table 6 presents the calculation of this paint hazard
variable.
In the R&M data, dust BRM vacuum samples were first analyzed using inductively
coupled plasma-atomic emission spectrometry (ICP). If the lead concentrations were below the
limit of quantification for ICP, then the samples were analyzed by GFAA. The soil samples were
analyzed using GFAA. The limit of detection (LOD) for the methods was not reported directly,
but was reported as three times the standard deviation of the instrument detection limit. Three
floor dust samples were found to be below their LOD while no soil samples were below the
LOD. For the dust samples below the detection limit, the reported value was replaced by
LOD/fi.
Draft - Do Not Cite or Quote 8 September 19. 1996
-------
Table 5. Description of the Variables Used in the Sensitivity/Specificity Analyses of the
Data from the Baltimore Repair and Maintenance Study.
Study
Baltimore Repair and
Maintenance
Location of Environmental Sample
Dripline Soil-Lead Concentration
Floor Dust-Lead Loading"
Window Sill Dust-Lead Loading
Window Trough Dust-Lead Loading
Sampling
Method
Core
Wipe"
Wipe"
Wipe"
Number of Samples
(House Averages)
Available for Analysis
27
87
87
86
8 The Baltimore R&M Study floor dust-lead loading was a composite of BRM vacuum" samples taken from all
carpeted floors, all uncarpeted floors, or a combination of uncarpeted and carpeted floors.
b Original BRM vacuum dust-lead loadings were converted to wipe dust-lead loadings using the following
equations:
Wipe Floor Dust-Lead Loading =
Proportion of Uncarpeted Floor Subsamples * 8.79'BRM Combined Floor Dust-Lead Loading0313 +
Proportion of Carpeted Floor Subsamples * 2.21*BRM Combined Floor Dust-Lead Loading0271
Wipe Window Sill Dust-Lead Loading = 17.0*BRM Window Sill Dust-Lead Loading0421
Wipe Window Trough Dust-Lead Loading = 14.1*BRM Window Trough Dust-Lead Loading0571
Table 6. Description of the Calculated Lead-Based Paint Hazard Variables for the Data
from the Baltimore Repair and Maintenance Study.
Study
Baltimore
Repair and
Maintenance
Paint Hazard
Variable
Percent of
Components
with
Deteriorated
LBP
Definition of
the
Condition of
Paint
Intact:
Paint that is
intact
Non-Intact:
Paint that is
non-intact
Deteriorated LBP Indicator
Value
0
1
Definition
No LBP (XRF
reading < 1) on
the component or
condition of the
paint is Intact
LBP (XRF i 1)
present on the
component
surface and the
condition of the
paint is
characterized as
Non-Intact
Calculation of the Lead-Based Paint
Hazard Variable
E Deteriorated LBP Indicator for
each component within a home . ^g
tt components sampled within a home
Draft - Do Not Cite or Quote
September 19, 1996
-------
4.0 ANALYSIS METHODS
Two main analysis methods were used to estimate options for the standards from the
Rochester and R&M studies: regression models and sensitivity/specificity analyses. Each
method was applied to individual standards separately (single media analysis) and to a
combination of standards jointly (multi-media analysis). Following is a general description of
each analysis method.
4.1 REGRESSION MODELS
Both single and multi-media regression models were explored for estimating options for
the Section 403 standards based on data from the Rochester Lead-in-Dust Study. Note that the
regression analyses were not done on data from the Baltimore R&M Study. The single media
models related childhood blood-lead concentrations to measures of lead from each media (floor
dust, window sill dust, window trough, soil, and paint) separately as follows:
ln(PbB.) = Po + p, * PbE, + e,, where
PbB, represents the blood-lead level of the child living in the ith home, PbE, represents a measure
of environmental lead (either on the original scale or log transformed) from the ith home, PO and
P, are intercept and slope parameters which describe the modeled relationship, and e, is the
residual error in ln(PbB,) left unexplained by the model. For each model, an estimate of the
environmental lead level at which 95 percent of the population of children would be expected to
be below 10 ug/dL was provided as input for Section 403 Standards.
The multi-media regression model related childhood blood-lead concentrations to
measures of lead from each media (dust, soil and paint) simultaneously as follows:
ln(PbB,) = po +p,*ln(Dust,) + p2*ln(Soil.) +p3*Paint, + e,
For this model, a joint estimate of environmental lead levels in paint, dust, and soil at which 95
percent of the population of children would be expected to be below 10 ug/dL was provided as
input for Section 403 Standards.
4.2 SENSITIVITY/SPECIFICITY ANALYSES
Table 7 describes the performance characteristics that were estimated as part of the
sensitivity/specificity analyses. The single media analysis focused on Media Standards that each
corresponded to a single measure of lead at the primary residence, whereas in the multi-media
analysis, the Media Standard corresponded to measures of environmental lead in several media
(dust, soil and paint). Thus, if any single measure of environmental lead was above a standard in
the multi-media analysis, the residence was categorized as being above the standard.
Draft - Do Not Cite or Quote 10 September 19. 1996
-------
Table 7. Definitions of Performance Characteristics Used to Characterize the
Empirical Data from Lead Exposure Studies to Aid in Defining the Likely
Performance of Different Health Based Standards.
Blood Lead
Concentration
Standard
In the above table, the lett
above a given blood-lead s
standard for that environm
counts the following perfo
Performance
Characteristic
Sensitivity
(or True Positive Rate)
Specificity
(or True Negative Rate)
Positive Predictive Value
(PPV)
Negative Predictive
Value (NPV)
Above
Below
Media Standard
Below
a
c
Above
b
d
er 'a' represents the number of children which have a blood lead concentration
tandard and who live in a residence with an environmental lead level below a
ental medium. Letters 'b', 'c', and 'd' represent similar counts. From these
rmance characteristics are calculated
Definition
Probability of a dwelling being above the soil lead
standard given that there is a resident child with an
elevated blood concentration.
Probability of a dwelling being below the soil lead
standard given that a resident child has a low blood
lead concentration.
Probability of a resident child having an elevated
blood lead concentration given that the observed soil
lead in the dwelling is above the standard.
Probability of a resident child having a low blood
lead concentration given that the observed soil lead
in the dwelling is below the standard.
Calculation
b/(a + b)
c/(c+d)
b/(b + d)
c/(a + c)
EPA has expressed initial interest in focusing on the NPV as the primary measure of the
performance of the potential standards. In particular, EPA has expressed interest in an NPV of
95% (i.e., a goal of 95% probability that a child will have a blood-lead concentration below 10
ug/dL given that the residence a child lives in meets the media standard). It should be noted,
however, that according to 1997 population estimates, achieving an NPV of 95% by the Section
403 rulemaking would still leave an estimated 1 million children younger than six years of age
with blood-lead concentrations above 10 ug/dL.
Although the NPV may be of primary interest, the other three statistics also offer valuable
information. From the single media analyses, we observed that when the NPV is at its highest
level, sensitivity is generally also high, but specificity and PPV are low. For instance, in the
single media analysis of the Rochester dripline soil-lead, an NPV of 94% and sensitivity of 98%
is attained at 100 ug/g, but specificity is only 11% and PPV is 26%. However, in the same
analysis, it can be seen that when the potential dripline soil-lead standard is raised to 900 ug/g,
the NPV remains high at 90%, sensitivity 80%, and specificity now rises to 57% and PPV to
37%. By allowing the NPV to be only slightly lower, the false positive rate (1 - specificity) was
Draft - Do Not Cite or Quote
11
September 19, 1996
-------
decreased significantly and the potential dripline soil-lead standard could be increased from 100
to 900 ug/g.
As indicated in the above example, incorporating all performance characteristics into the
evaluation of the potential media standards may be very informative. As displayed in Figure 1, in
an ideal situation, a standard could be found where each performance characteristic equals 1 (or
100%). In a less than ideal situation (i.e. typical data) the intent is to maximize each
characteristic. As a way to maximize all four characteristics simultaneously, we have developed
a simple statistic called the Unweighted Summary Measure (USM). This summary measure adds
all four performance characteristics together, equally weighting each characteristic.
Reference to Figure 1 reveals that in an ideal situation the maximum unweighted
summary measure would reach a value of 4 (or 400%). With the Rochester and Baltimore R&M
data, the USM is always less than 4. Figure 2 illustrates a situation that can be attained with the
Rochester and Baltimore R&M data, however, where both the NPV and sensitivity equal 100%
but the PPV and specificity are less than 100%. Criteria for choosing a recommended standard
could be based on NPV and sensitivity alone as illustrated in Figure 2, or on all performance
characteristics using the estimates of the USM.
3O -
4O -
•a
c
.2 3O-
§
0
0
1 2°-
i
o
53
10-
n -
• '••'- - ."*
"•••• V.-.". ^
•
. • •
- •
•
•
*
". •
.
• •
•
4OO 8OO 1.2OO 1.6OO
Dripline Soil —Lead Concentration
S.OOO
2.400
Figure 1. Example of an Ideal Situation for Establishing Potential Dripline Soil-Lead
Standards.
The four performance characteristics as well as the unweighted summary measure are
used in the subsequent multi-media analyses.
Draft - Do Not Cite or Quote
12
September 19, 1996
-------
5O
4O-
.2 3D
•§
o
5
10
.
V
4OO BOO 1.2OO 1.6OO
Drlpllne Soil-Lead Concentration (M«/B)
2.OOO
2.400
Figure 2. Example of a Situation Where the Negative Predictive Value and Sensitivity
Equal 100%, but the Positive Predictive Value and Specificity are Less than
100%.
One caution that should be noted is that only point estimates of each statistic are
presented. The uncertainty in these estimates is primarily dependent on sample size, and to a
lesser degree on measurement error and conversion factors. The uncertainty in the individual
calculated statistics (in particular, as affected by sample size) may be considered in evaluating
options for the standards. For example, the two-way multi-media sensitivity/specificity analysis
for uncarpeted floor dust and dripline soil for Rochester indicates that a dust-lead loading of 50
^ig/ft2 combined with a soil lead concentration of 300 ug/g will achieve an NPV of 90%. The
same analysis for the R&M study estimates an NPV of 90% can be associated with a combined
floor dust-lead loading of 50 ug/ft2 and a dripline soil concentration of only 200 ug/g. However,
the Rochester estimate is based on 34 children whereas the R&M estimate is based on only 10
children and includes an applied conversion factor. A simple confidence interval based on an
F-distribution approximation to a binomial proportion would estimate a confidence interval of
76% to 98% for the Rochester NPV and of 56% to 100% for the R&M NPV.
Confidence interval calculations for the different sensitivity/specificity performance
measures may be based on an F-distribution approximation to a binomial proportion as discussed
in Hollander and Wolfle, Nonparametric Statistical Methods. 1973. The confidence limits are
calculated as follows:
Draft - Do Not Cite or Quote
13
September 19, 1996
-------
lower confidence limit =
B
B+(n-B+l)*fa/22(n_B+1)2B
upper confidence limit = 1 -lower confidence limit,
where B is the number of successes, n is the number of trials and/Yn, n2 is the upper yth
percentile for the F distribution with rj, numerator degrees of freedom and r|2 denominator
degrees of freedom.
In the context of the sensitivity and specificity analyses, the parameters B and n can be
more specifically defined. For example, in the case of sensitivity, B is the number of homes that
have a child with an elevated blood-lead concentration and fail above the given environmental
media standard, while n is the number of homes that have a child with an elevated blood-lead
concentration. Table 8 presents some sample 95% confidence intervals for selected observed
sensitivity/specificity performance measures and related sample sizes.
Table 8. Sample 95% Confidence Intervals for Selected Sensitivity/Specificity
Performance Measures and Varying Sample Sizes.
Performance Measure
(Sample Size)
100% (5/5)
100% (10/10)
90% (9/10)
100% (20/20)
90% (18/20)
100% (30/30)
87% (26/30)
100% (40/40)
90% (36/40)
91% (31/34)
Lower 95% Confidence
Limit
48%
69%
56%
83%
68%
88%
69%
91%
76%
76%
Upper 95% Confidence
Limit
100%
100%
100%
100%
99%
100%
96%
100%
97%
98%
Note that these confidence intervals do not account for uncertainty introduced by use of
conversion factors in the R&M analyses.
Draft - Do Not Cite or Quote
14
September 19, 1996
-------
5.0 RESULTS OF SINGLE MEDIA REGRESSION ANALYSES
In an effort to provide information appropriate for setting health based standards for low
levels of lead in dust and soil, point estimates of the 80th and 95th percentiles of the distribution
of observed childhood blood-lead concentrations were estimated as a function of observed levels
of environmental lead. A potential health based standard for environmental-lead in this analysis
is the lead level at which one of the percentiles intersects some target blood-lead concentration.
Table 9 provides, for each of the measures of environmental-lead except the paint/pica
hazard indicator, the levels at which the 80th and 95th percentiles intersect target blood-lead
concentrations of 10 and 15 ug/dL based on the log-linear regression model. Results for
paint/pica hazard variables are not included in Table 9 because the variable measure is not
continuous.
Table 9. Estimated Environmental Lead Level at Which 80 and 95 Percent of Children's
Blood-Lead Concentrations will be Below Target Levels (10 and 15//g/dL) Based
on Single-Media Log-Linear Regression Models of the Rochester Lead-in-Dust
Study Data.
Variable
Play-Area Soil-Lead Concentration
Dripline Soil-Lead Concentration
Uncarpeted Floor Dust-Lead Loading
(Wipe Sample)
Carpeted Floor Dust-Lead Loading
(Wipe Sample)
Floor Dust-Lead Loading: Carpeted and
Uncarpeted (Wipe Sample)
Window Sill Dust-Lead Loading (Wipe
Sample)
Window Trough Dust-Lead Loading
(Wipe Sample)
Percentage of Interior Components with
Deteriorated LBP
Average Percent of LBP Deterioration
for an Interior Component
Percentage of Exterior Components
with Detenorated LBP
Average Percent of LBP Deterioration
for an Exterior Component
Unit of
Measure
P9/9
"9/9
A/9/ft2
09/rf
jig/ft2
OT/ff
PQ/ft'
%
%
%
%
Target
Blood-Lead
Concentration
10/yg/dL
15j/g/dL
10/yg/dL
1S//g/dL
10jig/dL
15/fg/dL
lOjig/dL
15//g/dL
10pg/dL
15//g/dL
lOj/g/dL
1 5 //g/dL
10j/g/dL
15/ig/dL
lOpg/dL
1 5 //g/dL
lOj/g/dL
1 5 //g/dL
10 //g/dL
15/ig/dL
IQjigWL
1 5j/g/dL
Estimated Environm
Below the Target Bloo
Estimated 95th Percentile
30
350
37
381
•
9
•
7
•
8
8
106
8
1134
•
6%
•
1%
•
•
•
0.03%
antal-Lead at Which
:hndren's Blood-Leads Fall
d-Lead Concentration
Estimated 80th Percentile
331
3969
554
5861
14
151
13
706
13
219
161
2309
2814
443737
11%
47%
2%
6%
7%
77%
2%
10%
Denotes that the Estimated Percentile is Always Above the Target Blood-Lead Concentration
Draft - Do Not Cite or Quote
15
September 19, 1996
-------
Figures A-l through A-l 1 in Appendix A display the estimated relationship between
blood-lead and each measure of environmental lead (except for the paint/pica indicator variable),
as well as the estimated 80th and 95th percentile of the distribution of observed childhood blood-
lead concentration.
Table 10 provides parameter estimates and other model information for each of thirteen
single media models. It should be noted that blood-lead concentration is significantly related (as
illustrated in Table 10) to many different measures of potential lead exposure. These measures
of lead exposure are also correlated with each other and with other factors related to lead
exposure such as housing characteristics (e.g., age of house) and socio-economic status
characteristics (e.g., income). Therefore, the degree of association between any single
environmental lead variable and blood-lead concentration, as measured by the single media
regression, is not necessarily the degree to which that single variable causes a change in blood-
lead concentrations.
Table 10. Parameter Estimates and Associated Standard Errors for Single Media Log-
Linear Models of the Relationship Between Blood-Lead Concentration and
Measures of Environmental-Lead from the Rochester Lead-in-Dust Study.
Model: InfPbB,) = P0 + (3,-PbE, + e,
Environmental Lead-Exposure Variable
Play-Area Soil-Pb Concentration
Dripline Soil-Pb Concentration
Uncarpeted Floor Dust-Pb Loading
Carpeted Floor Dust-Pb Loading
Floor Dust-Pb Loading
Window Sill Dust-Pb Loading
Window Trough Dust-Pb Loading
Percentage Deteriorated Interior
Average Deteriorated Interior
Indicator of Interior Paint/Pica Hazard
Percentage Deteriorated Exterior
Average Deteriorated Exterior
Indicator of Exterior Paint/Pica Hazard
Unit of
Measure
uglg
09/9
Ml/ft2
P9/ftJ
OT/ft2
09/ft2
M/ft2
%
%
-
%
%
-
n
87
186
197
179
205
196
189
205
205
205
205
205
205
H1
10.4%
12.9%
9.6%
3.7%
7.2%
11.4%
9.3%
9.7%
11.8%
4.6%
4.7%
6.1%
3.0%
Parameter Estimates
Po
se(Po)
0.945
(0.294)
0.727
(0.222)
1.384
(0.113)
1.564
(0.107)
1.447
(0.111)
1.041
(0.166)
1.167
(0.160)
1.679
(0.055)
1.635
(0.058)
1.817
(0.044)
1.756
(0.052)
1.722
(0.055)
1.832
(0.043)
P,
selO,)
0.163
(0.052)
0.172
(0.033)
0.166
(0.037)
0.100
(0.039)
0.142
(0.036)
0.152
(0.030)
0.080
(0.018)
1.130
(0.242)
9.088
(1.747)
0.332
(0.106)
0.579
(0.182)
4.609
(1.267)
0.368
(0.146)
O2
0.242
0.341
0.343
0.336
0.353
0.341
0.353
0.344
0.336
0.363
0.362
0.357
0.369
Draft - Do Not Cite or Quote
16
September 19, 1996
-------
6.0 RESULTS OF MULTI-MEDIA REGRESSION ANALYSES
As described in Section 4.1, the following multi-media regression model was fitted,
relating childhood blood-lead concentrations to measures of lead from dust, soil, and paint
simultaneously:
ln(Pbp.) = Po + p, * ln(Dust.) + P2 * ln(Soil.) + P3 * Paint, + e,
The above multi-media exposure model was fitted separately for the two different
measures of soil exposure in the Rochester Study because including measures of lead in play-area
soil significantly reduces the number of homes available for the analysis. Table 9 provides
parameter estimates and associated standard errors from a least squares fit of the multi-media
exposure model while varying the paint variable used in the analysis. Note that for models fit
with play area soil, the parameter estimates associated with paint were not statistically significant
and in one case, the parameter estimate was negative. As mentioned in the single-media
regression discussion, it should be noted that although blood-lead concentration is significantly
related (as illustrated in Table 1 1) to many different measures of potential lead exposure, the
degree of association between any single environmental lead variable and blood-lead
concentration, as measured by the multi-media regression, is not necessarily the degree to which
that single variable causes a change in blood-lead concentrations.
In an effort to provide information appropriate for setting health based standards for
levels of lead in dust and soil, point estimates for the 95th percentile of the distribution of
observed childhood blood-lead concentration were constructed as a function of observed levels
of environmental lead. A potential health based standard for environmental lead in this analysis
is the lead level at which the 95th percentile intersects some target blood-lead concentration.
In the joint media analysis, an important issue is the choice of environmental lead-levels
at which the 95th percentile estimates are constructed. For example, EPA may be interested in
an estimate of the dripline soil-lead concentration at which 95% of the distribution of children's
blood-lead concentration is below 10 ug/dL, assuming that the child lives in a residential unit
where other Section 403 Standards have been applied (e.g. no deteriorating lead-based paint, and
average floor dust-lead loading of 100 (ig/ft2). This situation would correspond to the following
formula for estimating a child's blood-lead concentration:
ln(PbB) = p0 + pI-ln(PbS) + p2-ln(PbF=100ug/ft2) + p3
The goal of this analysis is to find the value of environmental lead in soil (dripline fine-soil
fraction), if there is such a value, at which 95% of the distribution of children's blood-lead is
below 10 ng/dL when the other environmental media are held at the specific levels. Since this
type of analysis requires the values of several lead exposure predictor variables to be held
constant, while one target lead-exposure predictor variable is varied for the construction of the
Draft - Do Not Cite or Quote 17 September 19. 1996
-------
!
s
I
3
Table 11. Parameter Estimates and Associated Standard Errors from Fitting the Multi-Media Regression
Models to the Rochester Lead-in-Dust Study Data.
Variable
Intercept
Soil U/g/g)
Uncarpeted Floor
Dust-Pb Loading
U/Q/ft2)
Interior Paint
Parameter
Po
P,
Pa
P3
Number of Observations (n)
Coefficient of Determination (R2)
Mean Squared Error (o2Frrnr)
DRIPLINE SOIL PB CONCENTRATION
U/g/g)
Paint variable used in the analysis
Percentage of
Interior
Components
with
Deteriorated LBP
Estimate
(Std. Error)
0.752
(0.228)
0.113
(0.035)
0.086
(0.039)
0.759
(0.262)
181
0.197
0.315
Average Percent
of Deteriorated
LBP in Interior
Components
Estimate
(Std. Error)
0.781
(0.229)
0.107
(0.035)
0.080
(0.040)
5.89
(1.97)
181
0.199
0.313
Indication of
Paint/Pica
Hazard
Estimate
(Std. Error)
0.730
(0.228)
0.119
(0.035)
0.110
(0.038)
0.289
(0.103)
181
0.194
0.316
PLAY-AREA SOIL PB CONCENTRATION
(pg/s)
Paint variable used in the analysis
Percentage of
Interior
Components
with
Deteriorated
LBP
Estimate
(Std. Error)
0.680
(0.316)
0.130
(0.056)
0.162
(0.054)
-0.109
(0.353)
84
0.194
0.226
Average
Percent of
Deteriorated
LBP in
Interior
Components
Estimate
(Std. Error)
0.674
(0.315)
0.132
(0.055)
0.141
(0.055)
1.648
(2.523)
84
0.197
0.225
Indication of
Paint/Pica
Hazard
Estimate
(Std. Error)
0.643
(0.309)
0.124
(0.054)
0.170
(0.050)
0.236
(0.119)
84
0.230
0.216
00
I
I
I
-------
95th percentile band, the following levels of lead associated with each predictor variable were
explored as "constant" values:
Dripline Soil-Lead Concentration 10, 15, 20. 25, 50, 75, 100, 150, 200, 300, 400, 500 and 1000 //g/g
Play-Area Soil-Lead Concentration 10, 15, 20, 25, 50, 75, 150, 200, 300, 400 and 500 //g/g
Floor Dust-Lead Loading 1-5, 10, 20, 30, 40, 50, 75, 100, 150 and 200/ig/ft1
Deteriorated Interior Lead Based Paint None
Tables 12 and 13 present the estimated proportion of the distribution of children's blood
lead concentrations that falls below 10 ng/dL, assuming no deteriorated lead-based paint, for
various combinations of dripline soil and uncarpeted floor dust, and play area soil and uncarpeted
floor dust, respectively.
Appendix B provides the lead-levels at which the estimated 95th percentile of children's
blood-lead concentration intersects the target level of 10 ug/dL based on the multi-media
regression model. Figures are also provided in Appendix B which display a plot of blood-lead
versus a specific measure of environmental lead, as well as the estimated 95th percentile of
childhood blood-lead calculated by varying the specific measure of environmental lead, while
holding other measures of environmental lead constant. For example, Figure B-l provides a plot
of blood-lead versus uncarpeted floor dust-lead, as well as five lines depicting the estimated 95th
percentile of childhood blood-lead concentration for play area soil concentration of 200,300,
400, 500, and 100 ug/g respectively (each level of soil concentration is represented by a different
line).
The Tables and Figures provided in Appendix B demonstrate that setting health-based
standards for lead in environmental media (paint, dust and soil) results in very low standards
when based on the Rochester multi-media regression model and a goal of having 95 percent of
the distribution of children's blood-lead concentration below a target 10 ug/dL. These inferences
are based solely on the Rochester Lead-in-Dust Study, and may not be representative of the
nation as a whole.
Draft - Do Not Cite or Quote 19 September 19, 1996
-------
Table 12. Proportion of the Distribution of Childhood Blood-Lead in Rochester That Falls Below Target Level (10 ug/dL),
Assuming a Residential Unit with No Deteriorated Lead-Based Paint, and Specified levels of Dripline Soil-Lead
and Uncarpeted Floor Dust-Lead, and Using Model Based On Indication of Paint/Pica Hazard as Paint Variable"
M
O
Uncarpeted
Floor Dust-
Lead
Loading
(ira/ft2)
Dripline Soil-Lead Concentration (/ig/g)
0.90 0.89 0.87
a Bold line separates values greater than or equal to 95%, and represents levels at which 95% of distribution falls below 10//g/dL.
-------
I
I
Table 13. Proportion of the Distribution of Childhood Blood-Lead in Rochester That Falls Below Target Level (10 //g/dL ),
Assuming a Residential Unit with No Deteriorated Lead-Based Paint, and Specified levels of Play Area Soil-
Lead and Uncarpeted Floor Dust-Lead, and Using Model Based On Indication of Paint/Pica Hazard as Paint
Variable"
!
t
Uncarpeted
Floor Dust-
Lead
Loading
(jig/ft2)
Play Area Soil-Lead Concentration 0/g/g)
0.85 0.82 0.80
Bold line separates values greater than or equal to 95%, and represents levels at which 95% of distribution falls below 10//g/dL.
-------
7.0 RESULTS OF SINGLE MEDIA SENSITIVITY/SPECIFICITY ANALYSES
This section provides a discussion of the results of the single media sensitivity/specificity
analyses performed on various lead measures from dust, soil, and paint. Tables C-la through C-
13a in Appendix C present the single-media performance characteristics for the Rochester
dripline and play area soil-lead, uncarpeted, carpeted and combined floor dust-lead, window sill
and window trough dust-lead, the percentage of interior components with deteriorated LBP, the
percentage of exterior components with deteriorated LBP, the maximum of the percentage of
interior components with deteriorated LBP and the percentage of exterior components with
deteriorated LBP, the average percent of deteriorated LBP on interior components, the average
percent of deteriorated LBP on exterior components, and the maximum of the average percent of
deteriorated LBP on interior components and the average percent of deteriorated LBP on exterior
components, based on a blood-lead standard of 10 ug/dL. Tables C-14a through C-20a in
Appendix C present the single media performance characteristics for the R&M dripline soil-lead,
combined floor dust-lead, window sill and window trough dust-lead, the percentage of interior
components with deteriorated LBP, the percentage of exterior components with deteriorated
LBP, and the maximum of the percentage of interior components with deteriorated LBP and the
percentage of exterior components with deteriorated LBP, respectively.
Similar results are presented in Tables C-lb through C-20b in Appendix C for a blood-
lead standard of 15 ug/dL. However, it should be noted that there are only 16 children in the
Rochester study with blood-lead concentrations above IS ug/dL on which these performance
characteristic results are based.
Figures C-la through C-20a and C-lb through C-20b in Appendix C present four graphs
which summarize the relationship between the performance characteristics and the potential
single media standard. For easy comparison, the figure numbers correspond exactly to the
numbers of the tables described above.
In each figure, the first graph (upper left-hand corner) presents the blood-lead
concentration versus the media with a vertical reference line at the current media guidance level
(e.g. current soil-lead guidance level at 400 ug/g). The figure is useful for interpreting the results
seen hi the tables. For example, sliding the vertical reference line from left to right can provide a
visual image of the change in the negative predictive value (i.e. as the line moves from left to
right one can see the change in the percentage of data points with blood-lead concentrations
below 10 ug/dL and below the media guidance level).
Note that in the plots for the calculated paint hazard variable, the maximum of the
calculated interior and exterior hazards, a "star" indicates that the interior represented the
maximum paint hazard, a "circle" indicates that the exterior represented the maximum paint
hazard, and a "square" indicates that the same paint hazard was calculated for both the interior
and exterior.
Draft - Do Not Cite or Quote 22 September 19. 1996
-------
The second graph (upper right-hand corner) presents all four performance characteristics
as a function of the media. The graph allows for comparison of the four characteristics as the
potential standard for the media changes. Optimally, all four statistics should be maximized.
Using Figure C-la as an example, if the vertical line (representing the choice of a standard) were
slid along the horizontal axis toward 800 ug/g, one would see that NPV and sensitivity do not
decrease that much, but there is a marked increase in specificity and some increase in PPV.
Additionally, this graph also illustrates plateaus that occur in the performance characteristics, i.e.
where no additional information may be gained if the standard were to be increased or decreased.
Figure C-3a provides a good example of such plateaus. For instance between 125 and 175 ug/ft2,
the four characteristics are generally plateaued and there is no detectable change in performance
characteristics in that range.
The third graph (lower left-hand corner) illustrates the trade-off between positively
identifying a home with a lead hazard given that the child in the residence has an elevated blood-
lead (EBL) concentration (true positive or sensitivity) and falsely identifying a home as having a
lead hazard when the resident child does not have an EBL concentration (false positive or 1 -
specificity). The graph in Figure C-14a suggests approximately 300 ug/g as a candidate dripline
soil standard based on the single media analysis of the R&M dripline data. Note that in the same
graph the false positive rate becomes greater than the true positive rate at approximately 640 |ig/g
indicating a point at which the test has no value as a discriminator of a possible lead hazard.
The fourth graph (lower right-hand comer) shows the sum of the four performance
characteristics as the media standard changes. The point at which this sum is maximized may
indicate a potential standard.
Using the NPV alone from the single media analyses as the basis for establishing a
standard may be very difficult. As shown in the graphs and tables: (1) an NPV of 95% was only
attained at very low environmental levels or was not attained at all, (2) many standards associated
with a high NPV had very low specificity rates (i.e., high false positive rates), and (3) changing
the standard in a way that only slightly decreased the single media NPV in the neighborhood of
95% often significantly increased the other performance characteristics. The second, third, and
fourth graphs have been provided to allow easier consideration of all performance characteristics
for establishing standards.
8.0 RESULTS OF MULTI-MEDIA SENSITIVITY/SPECIFICITY ANALYSES
Two-Way Multi-Media Analyses (Uncarpeted Floor Dust and Soil)
The two-way multi-media sensitivity/specificity analysis was conducted on uncarpeted
floor dust-lead loadings jointly with either dripline soil-lead concentrations or play area soil-lead
concentrations found in the Rochester Lead-in-Dust Study (Rochester), and on combined
(carpeted and uncarpeted) floor dust-lead loadings and dripline soil-lead concentrations found in
the Baltimore Repair and Maintenance Study (R&M).
Draft - Do Not Cite or Quote 23 September 19, 1996
-------
For the two-way multi-media analyses, all four performance characteristics were
calculated for all possible combinations of the soil and dust pairs where uncarpeted floor dust-
lead and combined floor dust-lead were assessed from 50 to 200 ng/ft2, dripline soil-lead from
200 to 2,000 ug/g, and play area soil-lead from 200 to 1,000 ug/g. After generating the
performance characteristics for all combinations of soil- and dust-lead, two approaches were
taken for choosing optimal combinations of the media.
The first approach to identifying optimal combinations was to identify all combinations
that met certain targeted performance criteria. These criteria were:
1) NPV * 95%
2) NPV * 90%
For each of the criteria listed above, it was also required that at least 10 children were used in the
calculation of the performance characteristic (NPV).
Tables D-l through D-3, in Appendix D, present the performance characteristics for all of
the soil/dust combinations that met the above criteria. Note that no combination of potential
standards yielded an NPV greater than or equal to 95% for either the Rochester or the R&M data
for the range of standards considered.
Tables D-4, D-5, and D-6 present the dust/soil combinations that met the criteria of NPV
2 90%, for the Rochester uncarpeted floor dust and dripline soil, Rochester uncarpeted floor dust
and play area soil, and R&M floor dust and dripline soil, respectively. Each table presents four
selections of the combinations that met the additional criteria:
1) the combination(s) that include the highest dust level,
2) the combination(s) that include the highest soil level,
3) the combination(s) that include the current Interim Guidance level of 100 fig/ft2 for
dust, and
4) the combination(s) that include the current Interim Guidance level of 400 pg/g for soil.
It should be noted that lowering the range of potential dust-lead standards to 25 fig/ft2 in
the two-way analyses does identify two instances where a dust/soil combination would result in
an NPV of 95%. Both instances occurred in the Rochester data. Table D-7 presents the analyses
where an NPV of 2 95% was achieved with a wipe dust-lead loading of 25 ug/ft2. If wipe dust-
lead loadings are set lower than 25 ug/ft2, an NPV of 95% can also be achieved. For instance,
using the Rochester data, an NPV greater than 95% is achieved with soil-lead concentrations as
high as 2000 ug/g if the wipe dust-lead loading standard is lowered to 5 ug/ft2.
Draft - Do Not Cite or Quote 24 September 19, 1996
-------
These low levels for the standards were not included in the summary tables and primary
analyses. Including very low levels of soil or dust makes presentation of selected results (from
the universe of all possible combinations) difficult. For example, in Table D-4 the maximum
dripline soil standard which achieved an NPV of 90% is given as 1,000 ug/g associated with a
floor standard between 50 and 175 ug/ft2. If dust lead were allowed to be as low as 5 ug/ft2, the
maximum dripline soil concentration would rise to 2000 ug/g. The decision to set lower limits
on the standards considered for each media allows for presentation of combinations that include
relatively higher values for both media, rather than a combination that includes a very low
standard for one media and a very high standard for another.
The second approach to identifying optimal combinations of standards was to calculate
the summary measure (USM) described above and to rank potential standard combinations
according to the USM. Tables D-8 through D-10 present a summary of the two-way analyses
using the unweighted summary measure (USM). The tables are sorted by the USM, listing the
twenty-five combinations with the maximum USM.
Table D-8 indicates that the optimal combination (based on USM and the Rochester data)
would include a floor dust lead standard close to the Interim Guidance level of 100 ug/ft2 (50 -
75 ug/ft2), but would allow the dripline soil standard to be raised to 1,000 ng/g. Analysis of the
R&M data presented in Table D-10 suggests an optimal combination of 50 ug/ft2 for floors and
2,000 ug/g for dripline soil.
Examination of Table D-9 reveals that a standard of 100-200 ug/ft2 for floors and 900
ug/g for play area soil ranks first for the unweighted summary measure for the Rochester data.
Three-Way Multi-Media Analyses (Uncarpeted Floor Dust, Soil, and Paint)
For the three-way multi-media analyses, uncarpeted floor dust-lead and combined floor
dust-lead was assessed from 50 to 400 ug/ft2, dripline soil-lead from 200 to 1,500 ug/g, play area
soil-lead from 200 to 1,000 ug/g, and the calculated paint hazard variable of percent of
components with deteriorated LBP from 5 to 20%. (The ranges on floor dust-lead and dripline
soil-lead are slightly different from those used in the two-way analyses. The differences do not
significantly impact the results in either the two-way or the three-way analyses.)
In Appendix E, performance characteristics for three-way combinations of uncarpeted
floor dust-lead, soil-lead, and paint are presented for the Rochester data. Tables E-l through E-3
present the analyses for uncarpeted floor dust-lead, dripline soil-lead, and the percent of interior
components with deteriorated LBP, percent of exterior components with deteriorated LBP, and
the maximum of the interior and exterior deteriorated LBP percentages, respectively. Tables E-4
through E-6 are similar except the average percent of deteriorated LBP per component replaces
the percent of components with deteriorated LBP. Tables E-7 through E-l2 are similar to Tables
E-l through E-6 except dripline soil-lead is replaced by play area soil-lead. For the R&M data,
only 8 homes had values for all three media necessary for the analysis: floor dust-lead sample,
Draft - Do Not Cite or Quote 25 September 19, 1996
-------
dripline soil-lead sample, and a paint hazard value. Therefore, no three-way multi-media
analyses were run using the R&M data.
Table E-l indicates that an NPV of 100% can be attained with an uncarpeted floor dust-
lead standard between 50 and 400 ug/ft2, dripline soil-lead standard between 200 and 400 ug/g,
and the percent of interior components with deteriorated LBP standard between 5% and 20%.
Table E-3 indicates that an NPV of 95% can be attained with a floor dust-lead standard of 50 and
400 ug/ft2, a dripline soil-lead standard of 1000 ug/g and the maximum of the percent of interior
or exterior components with deteriorated lead-based paint at 20%.
As shown in Table E-4 through E-6, when using the Rochester data, no three-way multi-
media standards for uncarpeted floor dust-lead, dripline soil-lead, and the average percent of
deteriorated LBP per component had an NPV 2 95%.
Value of Adding Window Sill Dust-Lead to the Multi-Media Analyses
The multi-media standards presented up to this point have not included window sill dust-
lead. Table F-l, in Appendix F, presents selected two-way, three-way, and four-way multi-media
standards, for the Rochester data that demonstrate the value of adding window sill dust-lead to
the multi-media standards. For instance, when dripline soil-lead is held at 400 ug/g and
uncarpeted floor dust-lead is at 100 ug/ft2, 39 out of 43 (91% NPV) elevated blood-leads (EBL)
are identified. Adding window sill dust-lead at a standard of 100 ug/ft2 identifies an additional 3
EBL's (resulting in a 98% NPV).
A three-way analysis of the floor dust-lead, dripline soil-lead, and window sill dust-lead
was run on the R&M data. No combination yielded an NPV greater than or equal to 95%.
Four-Way and Five-Way Multi-Media Analyses (Uncarpeted Floor Dust, Window Sill Dust,
Soil (Dripline and/or Play Area), and Paint)
Seeing the value of adding window sill dust-lead into the multi-media analyses, four- and
five-way multi-media analyses were conducted on uncarpeted floor dust-lead, window sill dust-
lead, dripline soil-lead and/or play area soil-lead, and the maximum of the percent of interior
components with deteriorated LBP and the percent of exterior components with deteriorated LBP
for the Rochester data. Uncarpeted floor dust-lead was assessed from 50 to 400 ug/ft2, window
sill dust-lead from 100 to 800 ug/g, dripline soil-lead from 300 to 1,500 ug/g, play area soil-lead
from 200 to 1,000 ug/g, and all paint hazard variables from 5 to 20%. These ranges are the same
as ranges used in the three-way media analyses with the exception of the lower limit for the
dripline soil.
Table G-l in Appendix G presents all the four-way multi-media standards (dripline soil,
uncarpeted floor dust, window sill dust and paint) for which an NPV and a sensitivity of at least
95% have been attained.
Draft - Do Not Cite or Quote 26 September 19. 1996
-------
Table G-2 presents the four-way multi-media standards for play area soil, uncarpeted
floor dust, window sill dust, and percent of components with LBP for which an NPV and
sensitivity of at least 95% was attained using the Rochester data.
Table G-3 presents for the Rochester data, selected five-way multi-media standards
(dripline soil, play area soil, uncarpeted floor dust, window sill dust and paint) which achieved an
NPV and sensitivity of 100%.
8.1 DISCUSSION
Examination of the four-way and five-way multi-media results leads to the following
observations:
1. Whereas in the two-way multi-media sensitivity/specificity analyses no combination
produced an NPV greater than or equal to 95% (see Tables D-l through D-3), many
such combinations were found (at relatively high media levels) when using all the
media that are being considered for inclusion in the 403 standard (based on results in
Table G-3).
2. In general, in the Rochester data, the soil standard is a better discriminator for
identifying a lead hazard than the floor dust standard. In Table F-l, a floor dust
standard of 50 jig/ft2 is shown to identify only 9 of 45 homes with a child having an
elevated blood lead while a dripline soil standard of 400 ug/g identifies 38 out of 42
homes having a child with an elevated blood lead.
3. The uncarpeted floor dust standard does little for identifying a lead hazard in the
Rochester data based on the four-way analysis in Table G-l and the five-way
analysis in Table G-3. The standard for floor dust can take on any value in die range
evaluated (50 ug/g to 400 ug/g).
4. When a joint floor-dust/soil standard does not identify a lead hazard, the addition of
a window sill standard or the paint lead hazard identifies additional homes if a
hazard exists. The additional value of the window sill standard is illustrated in the
third joint standard presented in Table G-3 that assess soil, floor dust, and window
sill dust. The additional value of paint can be seen by comparing the maximum
NPV's from Table D-l with the NPV's shown in Table E-l.
9.0 CONCLUSION
Proposed ranges of standards for levels of lead in dust, soil, and paint are based on the
results of the preceding single and multi-media regression and sensitivity/specificity analyses.
Draft - Do Not Cite or Quote 27 September 19. 1996
-------
Table 14 presents a summary of estimates of standard levels which achieved (for homes
that meet the standard) either:
1. A Negative Predictive Value of 95% from a single media or multi-media
sensitivity/specificity analysis, or
2. An estimated 95% probability that a child's blood-lead concentration is below 10
|ig/dL from the single media or multi-media regression analyses.
The above two criteria are the current target health effects for use in informing the choice of
standards.
As seen in Table 14, based on the single media analyses, i.e., each Section 403 standard
considered separately, the standards must be set very low to meet the target health effects.
However, Table 14 also indicates that if the target health effects are associated with meeting all
Section 403 standards simultaneously, then standard levels can rise significantly, even well
above the levels proposed in the Interim Guidance. In the Rochester study, for any single
media/surface/location standard alone, even at very low levels, there were homes with children
with blood-lead concentrations above 10 ug/dL. However, when all media/surface/location
standards were considered jointly, even at relatively high levels, all homes with a child with a
blood-lead concentration above 10 ug/dL were identified. In other words, any home with a child
with a blood-lead concentration above 10 ug/dL exhibited a relatively high lead level in at least
one media/surface/location being considered for a Section 403 standard. This is a significant
finding related to the Section 403 rulemaking as it illustrates the effectiveness of using multiple
standards applied in a single risk assessment to identify a home with a lead-based paint hazard.
It should be noted that the above conclusion is based upon the multi-media
sensitivity/specificity analyses from Rochester alone. The multi-media regression analyses do
not show the same effect. There could be a number of reasons for this, such as the fact that the
form of the regression model fitted is not sensitive to interactions between the effect of the
different environmental media, or that many blood-lead concentrations are close to the
sensitivity/specificity target value of 10 ug/dL.
Based on Table 14, a range of options for a standard for each media/surface/location can
be specified. Table 15 below lists the proposed range of options for each standard. For each
standard the upper limit of the range is specified as the maximum estimated level in one of the
multi-media analyses.
The lower limit of the standard is more problematic, as is shown in Table H-l of
Appendix H. Very low levels of standards (e.g. less than 25 ug/ft2 for dust and less than 50 ug/g
for soil) were estimated by the single media analyses to be associated with the target health
effect. From a practical standpoint, these standards are not necessarily achievable. Therefore,
the lower limits presented in Table 15 for the range of options for a standard are based on
practicality.
Draft - Do Not Cite or Quote 28 September 19, 1996
-------
I1
Table 14. Summary of Estimated Standards Which Achieve a Negative Predictive Value of 95% or an Estimated 95%
Probability of a Child's Blood-Lead Concentration Below 10 /sg/dL in a Dwelling that is at or Below the
Standard.
Media
Dnpline Soil b/g/g)
Ray Area Soil b/g/g)
Uncarpeted Floor (fig/ft2)
Carpeted Floor U/g/ft'l
Carpeted and Uncarpeted
Floor b/o/ft1)
Window Sills big/ft')
Window Troughs big/ft1)
Percent of Interior Components
with Deteriorated LBP
Average Percent of Deteriorated
LBP per Interior Component
Percent of Exterior Components
with Deteriorated LBP
Average Percent of Deteriorated
LBP per Exterior Component
Maximum of Interior/ Exterior
Percent of Components with
Deteriorated LBP
Maximum of Interior/ Extenor
Average Percent of Deteriorated
LBP per Component
Single Media Analyses1
Regression
Models
Rochester
<50
<60
<25
<25
<2S
<25
<25
0%
0%
0%
0%
Sensitivity/Specificity
RAM
-------
Table 15. Proposed Options for Section 403 Standards To Be Evaluated in the Risk
Assessment and Economic Analysis.
Standard
Uncarpeted Floor Dust U/g/ft2)
Window Sill Dust U/g/ft2)
Dripline Soil U/g/g)
Play Area Soil (//g/g)
Maximum of Percent of Interior
Components with Deteriorated Lead-
Based Paint and Percent of Exterior
Components with Deteriorated Lead-
Based Paint
Range
Low Limit
25
25
50
50
0%
High Limit
400
800
1500
1000
20%
For dust standards, this lower limit is defined as 25 ug/ft2 because a dust standard below
25 ug/ft2 may be problematic due to laboratory analysis issues. Many currently accredited
laboratories have levels of quantification between 10 ug/ft2 and 25 ug/ft2.
The lower limit for a soil standard(s) is defined as 50 ug/g. This level was chosen since
national background levels of lead in soil have been estimated in the neighborhood of 20 ug/g to
35 ug/g.
The lower limit for a paint standard was chosen as zero percent deteriorated lead-based
paint, the level suggested by the single media analyses. Single media analyses indicate that no
single standard for paint can achieve the target health criteria of 95% probability that a child's
blood-lead concentration will be below 10 ug/dL. There are no practical difficulties associated
with setting a low limit of zero percent for deteriorated lead-based paint.
A formal analysis of the "net benefits" associated with different options for the standards
will be conducted in the risk assessment and economic analysis. The final results of all analyses
will be used, along with policy input, by EPA's risk management team to define standards that
achieve Title X's purposes: "to implement, on a priority basis, a broad program to evaluate and
reduce lead-based paint hazards in the nation's housing" and "to encourage effective action to
prevent childhood lead poisoning by establishing a workable framework for lead-based paint
hazard evaluation..." (Title X, Section 1003).
Draft - Do Not Cite or Quote
30
September 19, 1996
-------
REFERENCES
"Draft Report on Conversions Equations for Use in the Section 403 Rulemaking," August 30,
1996.
"Multi-Media Model Based on the Rochester Lead-in-Dust Study (Epi Model) For Use in
Predicting a National Distribution of Children's Blood-Lead Concentrations for the Section 403
Risk Assessment," July 30,1996.
"The Relation of Lead-Contaminated House Dust and Blood Lead Levels Among Urban
Children." Volumes I and II. June, 1995. Final Report to the U.S. Department of Housing and
Urban Development Grant from The University of Rochester School of Medicine, Rochester,
New York, and The National Center for Lead-Safe Housing, Columbia Maryland.
"Lead Paint Abatement and Repair and Maintenance Study in Baltimore." Preliminary 12-Month
Report. July, 1995. Prepared by Kennedy Krieger Research Institute for U.S. Environmental
Protection Agency.
Draft - Do Not Cite or Quote 31 September 19, 1996
-------
Appendix A
Figures for the Single Media Regression Analyses
Draft - Do Not Cite or Quote A-1 September 19. 1996
-------
50
40
30
20
E
_g
CD
10
' Observations
- Predicted (OLS)
- Estimated 95th Percentjle
Estimated BOtfi Percentile
10 100 1000 10000
Play-Area Soil-Pb Concentration ( //g/g )
100000
Figure A-1. Estimated Relationship and 80th and 95th Percentiles of Observed Childhood
Blood-Lead Concentrations as a Function of Play-Area Soil-Lead Concentration
Based on Single-Media Log-Linear Models of the Rochester Lead-in-Dust Study
Data
Draft - Do Not Cite or Quote
A-2
September 19, 1996
-------
50
40
30
20
o
_o
CO
10
' • Observations
Predicted (OLS)
- Estimated 95th Percentile
- Estimated 80th Percentile
10 100 1000 10000
Dripline Soil-Pb Concentration ( /,g/g )
100000
Figure A-2. Estimated Relationship and 80th and 95th Percentiles of Observed Childhood
Blood-Lead Concentrations as a Function of Dripline Soil-Lead Concentration
Based on Single-Media Log-Linear Models of the Rochester Lead-in-Dust Study
Data
Draft - Do Not Cite or Quote
A-3
September 19, 1996
-------
50
40
30
20
CD
10
' • Observations
- Predicted (OLS)
- Estimated 95th Percentile
- Estimated 80th Percentle
10 100 1000 10000
Uncarpeted Floor Dust-Pb Loading ( A
-------
50
40
30
20
o
o
CD
10
• ' ' Observations
- Predicted (OLS)
- Estimated 95th Percentile
Estimated 80th Percentile
10
100
1000
10000
100000
Carpeted Floor Dust-Pb Loading (
Figure A-4. Estimated Relationship and 80th and 95th Percentiles of Observed Childhood
Blood-Lead Concentrations as a Function of Carpeted Wipe Floor Dust-Lead
Loading Based on Single-Media Log-Linear Models of the Rochester Lead-in-
Dust Study Data
Draft - Do Not Cite or Quote
A-5
September 19, 1996
-------
50
40
30
O)
.a
Q- 20
•o
8
CD
10
' Observations
- Predicted (OLS)
- Estimated 95th Percentile
Estimated 80th Percentile
10 100 1000
Floor Dust-Pb Loading ( //g/ft2 )
10000
100000
Figure A-5. Estimated Relationship and 80th and 95th Percentiles of Observed Childhood
Blood-Lead Concentrations as a Function of Wipe Floor Dust-Lead Loading
(Carpeted and Uncarpeted Surfaces Combined) Based on Single-Media Log-
Linear Models of the Rochester Lead-in-Dust Study Data
Draft - Do Not Cite or Quote
A-6
September 19, 1996
-------
50
40
30
20
E
.g
CD
' ' ' Observations
Predicted (OLS)
- Estimated 95th Percentile
Estimated 80th Percentile
- "***,
10 100 1000 10000
Window Sill Dust-Pb Loading ( ug/ft.2 )
100000
Figure A-6. Estimated Relationship and 80th and 95th Percentiles of Observed Childhood
Blood-Lead Concentrations as a Function of Window Sill Wipe Dust-Lead
Loading Based on Single-Media Log-Linear Models of the Rochester Lead-in-
Dust Study Data
Draft - Do Not Cite or Quote
A-7
September 19, 1996
-------
50
40
30
20
•a
o
a
CO
10
' ' ObseivatKxis
Predicted (OLS)
- Estimated 95th Percentile
Estimated SOtti Percentile
10 100 1000 10000
Window Trough Dust-Pb Loading (
100000 1000000
Figure A-7. Estimated Relationship and 80th and 95th Percentiles of Observed Childhood
Blood-Lead Concentrations as a Function of Window Trough Wipe Dust-Lead
Loading Based on Single-Media Log-Linear Models of the Rochester Lead-in-
Dust Study Data
Draft - Do Not Cite or Quote
A-8
September 19, 1996
-------
50
40
30
Observations
Predicted (OLS)
Estimated 95th Percentite
Estimated BOtfi Percentile
10 20 30 40 50 60 70 80 90
Percentage of Interior Components with Deteriorated LBP
100
Figure A-8. Estimated Relationship and 80th and 95th Percentiles of Observed Childhood
Blood-Lead Concentrations as a Function of Percentage of Deteriorated Lead-
Based Interior Components Based on Single-Media Log-Linear Models of the
Rochester Lead-in-Dust Study Data
Draft - Do Not Cite or Quote
A-9
September 19, 1996
-------
50
40
30
20
CD
10
• • Obseivations
- Predicted (OLS)
- Estimated 95th Percentile
- Estimated 80th Percentile
10
20
Average Percent LBP Deterioration for Interior Components
Figure A-9. Estimated Relationship and 80th and 95th Percentiles of Observed Childhood
Blood-Lead Concentrations as a Function of Average Percentage Deterioration
in Lead-Based Interior Components Based on Single-Media Log-Linear Models
of the Rochester Lead-in-Dust Study Data
Draft - Do Not Cite or Quote
A-10
September 19, 1996
-------
50
40
30
20
CD
Observations
Predicted (OLS)
Estimated 95th Percentile
Estimated 80th Percentile
10 20 30 40 50 60 70 80 90
Percentage of Exterior Components with Deteriorated LBP
100
Figure A-10. Estimated Relationship and 80th and 95th Percentiles of Observed Childhood
Blood-Lead Concentrations as a Function of Percentage of Deteriorated Lead-
Based Exterior Components Based on Single-Media Log-Linear Models of the
Rochester Lead-in-Dust Study Data
Draft - Do Not Cite or Quote
A-11
September 19, 1996
-------
50
40
30
20
8
CD
10
• ' * Observations
- Predicted (OLS)
- Estimated 95th Percentile
• Estimated 80th Percentile
10
20
Average Percent LBP Deterioration for Exterior Components
Figure A-11. Estimated Relationship and 80th and 95th Percentiles of Observed Childhood
Blood-Lead Concentrations as a Function of Average Percentage Deterioration
in Lead-Based Exterior Components Based on Single-Media Log-Linear Models
of the Rochester Lead-in-Dust Study Data
Draft - Do Not Cite or Quote
A-12
September 19, 1996
-------
Appendix B
Tables and Figures for the Multi-Media Regression Analyses
Draft - Do Not Cite or Quote B-1 September 19, 1996
-------
Table B-1. Estimated Uncarpeted Floor Dust-Lead Loading at which 95 Percent of the
Distribution of Childhood Blood Lead Concentrations Falls Below 10//g/dL,
Assuming Target Lead Levels in Play Area Soil Concentration, and No
Deteriorated Lead Based Paint in the Primary Residence, Based on Paint/Pica
Variable.
Fixed Play Area
Soil Concentration
(//g/g)
10
15
20
25
30
40
50
75
100
150
200
300
400
500
1000
Estimated Uncarpeted Floor
Dust Lead Loading U/g/ft2).
Associated with 95 Percent of
Children Below 10//g/dL
37
28
22
19
16
13
11
8
7
5
4
3
2
2
*
Draft - Do Not Cite or Quote
B-2
September 19. 1996
-------
50-
40
Observations
Estimated 95th Percentile
Estimated 95th Percentile
Estimated 95tti Percentile
Estimated 95th Percentile
Estimated 95th Percentile
Play Area Soil = 200
Play Area SoB - 300
Play Area Soil = 400
Play Area Soil = 500
Play Area Soil - 1000
CO
20
10
-* t +
oi
1 10 100 1000 10000
Uncarpeted Floor Dust-Pb Loading ( u
100000
Figure B-1. Estimated 95th Percentile of Children's Blood-Lead Concentration as a
Function of Dust Lead-Loading from Uncarpeted Floors, Assuming a
Residence with Play-Area Soil-Lead Concentration of 200, 300, 400, 500
and 1000//g/g, No Deteriorated Lead-Based Paint, and Paint/Pica Hazard
Variable Used in the Model.
Draft - Do Not Cite or Quote
B-3
September 19, 1996
-------
Table B-2. Estimated Play Area Soil-Lead Concentration at which 95 Percent of the
Distribution of Childhood Blood Lead Concentrations Falls Below 10//g/dL,
Assuming Target Lead Levels in Uncarpeted Floor Dust-Lead Loading, and No
Deteriorated Lead Based Paint in the Primary Residence, Based on Paint/Pica
Variable.
Fixed Uncarpeted Floor
Dust-Lead Loading
(/ig/ft2)
1
2
3
4
5
10
20
30
40
50
75
100
150
200
Estimated Play Area Soil
Concentration U/g/g),
Associated with 95 Percent of
Children Below 10pg/dL
1444
559
321
216
159
61
24
13
9
6
3
2
*
*
Draft - Do Not Cite or Quote
B-4
September 19, 1996
-------
50^
Observations
Estimated 95th Percenfite Boor Dust = 50
Estimated 95th Percentile Floor Dust - 75
Estimated 95th Percentile Floor Dust - 100
Estimated 95th Percentile Floor Dust - 150
Estimated 95th Percentile Floor Dust - 200
10 100 1000
Play -Area Soil-Pb Concentration (
10000
100000
Figure B-2. Estimated 95th Percentile of Children's Blood-Lead Concentration as a
Function of Play-Area Soil-Lead Concentration, Assuming a Residence with
Uncarpeted Floor Dust-Lead Loading of 50, 75, 100, 150, and 200//g/ft2, No
Deteriorated Lead-Based Paint, and Paint/Pica Hazard Variable Used in the
Model.
Draft - Do Not Cite or Quote
B-5
September 19, 1996
-------
Table B-3. Estimated Uncarpeted Floor Dust-Lead Loading at which 95 Percent of the
Distribution of Childhood Blood Lead Concentrations Falls Below 10//g/dL,
Assuming Target Lead Levels in Drip Line Soil Concentration, and No
Deteriorated Lead Based Paint in the Primary Residence, Based on Paint/Pica
Variable.
Fixed Drip Line
Soil Concentration
fc/g/g)
10
15
20
25
30
40
50
75
100
150
200
300
400
500
1000
Estimated Uncarpeted Floor
Dust Lead Loading U/g/ft2),
Associated with 95 Percent of
Children Below 10 //g/dL
31
20
14
11
9
6
5
3
2
*
*
*
*
*
*
Draft - Do Not Cite or Quote
B-6
September 19, 1996
-------
50-
.4Q '.
* Observations
- Estimated 95th Percentile Dripline Soil = 200
- Estimated 95th Percentile Dripline Soil - 300
- Estimated 95th Percentile Dripline Soil = 400
- Estimated 95th Percentile Dripline Soil = 500
- Estimated 95th Percentile Dripline Soil - 1000
LJ. (-7
"8
o
CD
10
1 10 100 1000 10000
Uncarpeted Floor Dust-Pb Loading ( ug/ft.2 )
100000
Figure B-3. Estimated 95th Percentile of Children's Blood-Lead Concentration as a
Function of Dust Lead-Loading from Uncarpeted Floors, Assuming a
Residence with Dripline Soil-Lead Concentration of 200, 300, 400, 500 and
1000//g/g, No Deteriorated Lead-Based Paint, and Paint/Pica Hazard Variable
Used in the Model.
Draft - Do Not Cite or Quote
B-7
September 19, 1996
-------
Table B-4. Estimated Drip Line Soil-Lead Concentration at which 95 Percent of the
Distribution of Childhood Blood Lead Concentrations Falls Below 10//g/dL,
Assuming Target Lead Levels in Uncarpeted Floor Dust-Lead Loading, and No
Deteriorated Lead Based Paint in the Primary Residence, Based on Paint/Pica
Variable.
Fixed Uncarpeted Floor
Dust-Lead Loading
U/g/ft2}
1
2
3
4
5
10
20
30
40
50
75
100
150
200
Estimated Drip Line Soil
Concentration Cvg/g),
Associated with 95 Percent of
Children Below 10 j/g/dL
240
126
86
66
54
28
15
10
7
6
4
3
2
*
Draft - Do Not Cite or Quote
B-8
September 19, 1996
-------
501 ,
Observations
Estimated 95th Percentile Roor Dust = 50
Estimated 95th Percentile ROOT Dust - 75
Estimated 95th Percentile Roor Dust = 100
Estimated 95th Percentile Floor Dust = 150
Estimated 95th Percentile Roor Dust - 200
40
_J
f) 3°
1 *>
CD
10
.•» •> *. *
10 100 1000 10000
Dripline Soil-Pb Concentration ( /,g/g )
100000
Figure B-4. Estimated 95th Percentile of Children's Blood-Lead Concentration as a
Function of Dripline Soil-Lead Concentration, Assuming a Residence with
Uncarpeted Floor Dust-Lead Loading of 50, 75, 100, 150, and 200//g/ft2. No
Deteriorated Lead-Based Paint, and Paint/Pica Hazard Variable Used in the
Model.
Draft - Do Not Cite or Quote
B-9
September 19, 1996
-------
Table B-5. Estimated Uncarpeted Floor Dust-Lead Loading at which 95 Percent of the
Distribution of Childhood Blood Lead Concentrations Falls Below 10//g/dL,
Assuming Target Lead Levels in Drip Line Soil Concentration, and No
Deteriorated Lead Based Paint in the Primary Residence, Based on 10 Percent
of Components Exhibiting Deterioration and LBP.
Fixed Drip Line
Soil Concentration
U*9/9)
10
15
20
25
30
40
50
75
100
150
200
300
400
500
1000
Estimated Uncarpeted Floor
Dust Lead Loading (/ig/ft2).
Associated with 95 Percent of
Children Below 10 //g/dL
31
18
12
9
7
5
3
2
*
*
*
*
*
*
*
Draft - Do Not Cite or Quote
B-10
September 19, 1996
-------
50- .
40
Observations
Estimated 95th Percentite Dripline Soil = 200
Estimated 95th Percenlile Dripline Soil - 300
Estimated 95tn PercenBle Dripline Soil - 400
Estimated 95th Percantile Dripline Soil - 500
Estimated 95th Percentite Dripline Soil - 1000
10 100 1000 10000 100000
Uncarpeted Floor Dust-Pb Loading ( /t/g/ft.2 )
Figure B-5. Estimated 95th Percentile of Children's Blood-Lead Concentration as a
Function of Dust Lead-Loading from Uncarpeted Floors, Assuming a
Residence with Dripline Soil-Lead Concentration of 200, 300, 400, 500 and
1000//g/g, and 10% of Components Exhibiting Deterioration and LBP.
Draft - Do Not Cite or Quote
B-11
September 19, 1996
-------
Table B-6. Estimated Drip Line Soil-Lead Concentration at which 95 Percent of the
Distribution of Childhood Blood Lead Concentrations Falls Below 10 jig/dL,
Assuming Target Lead Levels in Uncarpeted Floor Dust-Lead Loading, and No
Deteriorated Lead Based Paint in the Primary Residence, Based on 10 Percent
of Components Exhibiting Deterioration and LBP.
Fixed Uncarpeted Floor
Dust-Lead Loading
U/g/ft2)
1
2
3
4
5
10
20
30
40
50
75
100
150
200
Estimated Drip Line Soil
Concentration (//g/g).
Associated with 95 Percent of
Children Below 100g/dL
136
80
59
47
40
23
14
10
8
7
5
4
3
2
Draft - Do Not Cite or Quote
B-12
September 19, 1996
-------
50
40
Observations
Estimated 95th Percenbte Floor Dust - 50
Estimated 95th Percentile ROOT Dust - 75
Estimated 95tti Percentile ROOT Dust = 100
Estimated 95th Percentile ROOT Dust = 150
Estimated 95th Percentile Roor Dust - 200
10 100 1000 10000
Dripline Soil-Pb Concentration ( //g/g )
100000
Figure B-6. Estimated 95th Percentile of Children's Blood-Lead Concentration as a
Function of Dripline Soil-Lead Concentration, Assuming a Residence with
Uncarpeted Floor Dust-Lead Loading of 50, 75, 100, 150, and 200//g/ft2,
and 10% of Components Exhibiting Deterioration and LBP.
Draft - Do Not Cite or Quote
B-13
September 19, 1996
-------
Table B-7. Estimated Uncarpeted Floor Dust-Lead Loading at which 95 Percent of the
Distribution of Childhood Blood Lead Concentrations Falls Below 10//g/dL,
Assuming Target Lead Levels in Drip Line Soil Concentration, and No
Deteriorated Lead Based Paint in the Primary Residence, Based on an
Average of 2.5% Deteriorated LBP on Interior Components.
Fixed Drip Line
Soil Concentration
(09/9)
10
15
20
25
30
40
50
75
100
150
200
300
400
500
1000
Estimated Uncarpeted Floor
Dust Lead Loading (//g/ft2).
Associated with 95 Percent of
Children Below 10//g/dL
13
7
5
4
3
2
*
*
*
*
*
*
*
*
*
Draft - Do Not Cite or Quote
B-14
September 19, 1996
-------
50
40
Observations
Estimated 95tti Percentite Dripline Soil - 200
Estimated 95tti Percentite Dripline Soil - 300
Estimated 9Stti Percentite Dripline Soil - 400
Estimated 95th Percentite Dripline Son - 500
Estimated 95th Percentite Dripline Soil - 1000
10 100 1000 10000
Uncarpeted Floor Dust-Pb Loading (
100000
Figure B-7. Estimated 95th Percentile of Children's Blood-Lead Concentration as a
Function of Dust Lead-Loading from Uncarpeted Floors, Assuming a
Residence with Dripline Soil-Lead Concentration of 200, 300, 400, 500 and
1000 //g/g, and an Average of 2.5% Deteriorated LBP on Interior
Components.
Draft - Do Not Cite or Quote
B-15
September 19, 1996
-------
Table B-8. Estimated Drip Line Soil-Lead Concentration at which 95 Percent of the
Distribution of Childhood Blood Lead Concentrations Falls Below 10 //g/dL,
Assuming Target Lead Levels in Uncarpeted Floor Dust-Lead Loading, and No
Deteriorated Lead Based Paint in the Primary Residence, Based on an
Average of 2.5% Deteriorated LBP on Interior Components.
Fixed Uncarpeted Floor
Dust-Lead Loading
Uig/ft2)
1
2
3
4
5
10
20
30
40
50
75
100
150
200
Estimated Drip Line Soil
Concentration (//g/g),
Associated with 95 Percent of
Children Below 10 //g/dL
70
42
31
25
21
12
7
5
4
3
2
2
*
*
Draft - Do Not Cite or Quote
B-16
September 19, 1996
-------
.Q
Q.
so
40
Observations
Estimated 95th Percentite Floor Dust = 50
Estimated 95th Percentile Ftoor Dust - 75
Estimated 95th Percentile floor Dust - 100
Estimated 95th Percentile Floor Dust = 150
Estimated 95th Percentile Floor Dust - 200
10 100 1000
Dripline Soil-Pb Concentration (
10000
100000
Figure B-8. Estimated 95th Percentile of Children's Blood-Lead Concentration as a
Function of Dripline Soil-Lead Concentration, Assuming a Residence with
Uncarpeted Floor Dust-Lead Loading of 50, 75, 100, 150, and 200 j/g/ft2,
and an Average of 2.5% Deteriorated LBP on Interior Components.
Draft - Do Not Cite or Quote
B-17
September 19, 1996
-------
Appendix C
Tables and Figures for the Single
Media Sensitivity/Specificity Analyses
Draft - Do Not Cite or Quote C-l September J9, 1996
-------
I
o
»'
O
s
1OO 1 .000 10 000
Driplme Soil Lead Concentration (*,g g)
BOO t 600 2 400 3 :*OO 4 000 4 MOO S.(KM> 6 4OO 7 ZOO B OOO
Driplme Soil Lend Concentration ( -g g)
£
c
a
£ 04 -
O
i
BOO 1 6OO 2 400 T ^00 4 OOC 4 BOO S 6OO 6 400 7 ZOO B OOO
Driphne Soil-Lead Concentration (ug g)
Poaitiv* (I -Spcciflcilvl
BOO 1 800 2 4JO 3 ?CID 4 000 4 BOO S (U)O
Driphne i>oil Lend Concent rat ion (..j; g)
400 7 2OO 0 OOO
Figure C-1a. Summary Graphs of the Performance Characteristics of Potential Dripline Soil-Lead Standards Based on Data
£ from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 10/ig/dL.
-------
Table C-la. Performance Characteristics of Potential Dripline Soil-Lead Standards Based on
Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of
10//g/dL.
Dripline
Soil-Lead
Standard
{pglg}
25
50
75
100
125
150
175
200
250
300
350
400
450
500
600
700
800
900
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
5,000
8,000
10,000
% Homes
Falling the
Standard
(# Homes Faffing/
# Homes)
99% (184/186)
95% (177/186)
93% (172/186)
91% (169/186)
88% (163/186)
86% (159/186)
83% (155/186)
82% (153/1 86)
81% (150/186)
80% (148/186)
79% (146/186)
76% (142/186)
74% (137/186)
72% (133/186)
69% (128/186)
60% (112/186)
57% (105/1 86)
52% (97/186)
47% (88/186)
29% (53/186)
18% (34/186)
13% (24/186)
9% (17/186)
7% (12/186)
7% (12/186)
5% (10/186)
4% (7/186)
1% (2/186)
1% (2/186)
Blood-Lead Standard of 10//g/dL
Sensitivity
(b/a+b)
100% (45/45)
100% (45/45)
98% (44/45)
98% (44/45)
96% (43/45)
96% (43/45)
96% (43/45)
91% (41/45)
91% (41/45)
91% (41/45)
91% (41/45)
89% (40/45)
89% (40/45)
89% (40/45)
87% (39/45)
82% (37/45)
80% (36/45)
80% (36/45)
78% (35/45)
49% (22/45)
29% (13/45)
22% (10/45)
11% (5/45)
9% (4/45)
9% (4/45)
9% (4/45)
7% (3/45)
2% (1/45)
2% (1/45)
Specificity
(c/c+d)
1% (2/1 41)
6% (9/141)
9% (13/141)
11% (16/141)
15% (21/141)
18% (25/1 41 )
21% (29/141)
21% (29/141 )
23% (32/141 )
24% (34/1 41)
26% (36/141)
28% (39/141 )
31% (44/141 )
34% (48/1 41)
37% (52/141 1
47% (66/141 )
51% (72/1411
57% (80/141)
62% (88/141)
78% (110/141)
85% (120/1 41 )
90% (127/1 41)
92% (129/141)
94% (133/1 41)
94% (133/1 41)
96% (135/141)
97% (137/141)
99% (140/141)
99% (140/141)
Dflk «!*!*• A
rosiuVe
Predictivs
Value
(b/b+d)
25% (45/184)
25% (45/1 77)
26% (44/1 72)
26% (44/169)
26% (43/163)
27% (43/1 59)
28% (43/1 55)
27% (41/1 53)
27% (41 /1 50)
28% (41/148)
28% (41/146)
28% (40/142)
29% (40/1 37)
30% (40/1 33)
31% (39/1 28)
33% (37/1 12)
34% (36/105)
37% (36/97)
40% (35/88)
42% (22/53)
38% 11 3/34)
42% (10/24)
29% (5/1 7)
33% (4/1 2)
33% (4/1 2)
40% (4/10)
43% (3/7)
50% (54)
50% (%)
Negative
Predictive
Value
(c/a+c)
100% (2/2)
100% (9/9)
93% 1 13/1 4)
94% ( 16/1 7)
91% (21/23)
93% (25/27)
94% (29/31)
88% (29/33)
89% (32/36)
90% (34/38)
90% (36/40)
89% (39/44)
90% (44/49)
91% (48/53)
90% (52/58)
89% (66/74)
89% (72/81)
90% (80/89)
90% (88/98)
83% (11 0/1 33)
79% (120/1 52)
78% (127/1 62)
76% (129/1 69)
76% (133/1 74)
76% (133/1 74)
77% (135/1 76)
77% (137/1 79)
76% (140/1 84)
76% (140/1 84)
#
Homes
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
• 186
186
186
186
186
186
186
186
186
186
Draft - Do Not Cite or Quote
C-3
September 19, 1996
-------
I
D
i
I OOO
ion (ug g)
••"•
o a -
01 -
I
I
_o
fc 03 -
O Z -
01-;
40O BOO \ .ZOO 1 6OO 2 OOO Z.4OO Z BOO 3 2OD tl 6OO 4 OOO
Play-Area Soil-Lead Concentration Ug 8>
I
o-
^
-*k
50
40O 800 1 200 1 60O 2.OOO 2.4OO 2 BOO 3 200 3,600 4 OOO
Play-Area Soil-Lead Concentration U
-------
Table C-2a. Performance Characteristics of Potential Play-Area Soil-Lead Standards Based
on Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of
10/ig/dL.
Play-Area
Soil-Lead
Standard
(ifQ/O)
25
50
75
100
125
150
175
200
250
300
400
450
500
600
700
800
900
1,000
1,500
2,000
2,500
4 000
% Homes
Failing the
Standard
(# Homes Failing/
ft Homes)
100% (87/87)
94% (82/87)
89% (77/87)
83% (72/87)
76% (66/87)
70% (61/87)
66% (57/87)
61% (53/87)
55% (48/87)
49% (43/87)
41% (36/87)
36% (31/87)
26% (23/87)
23% 120/87)
21% (18/87)
17% (15/87)
14% (12/87)
9% (8/87)
8% (7/87)
5% (4/87)
2% (2/87)
1%(1/87)
1% (1/87)
Blood-Lead Standard of 10//g/dL
Sensitivity
(b/a+b)
100% (16/1 6)
94% (15/1 6)
94% (15/1 6)
94% (15/16)
81% (13/16)
81% (13/1 6)
81% (13/1 6)
81% (13/1 6)
69% (11/1 6)
69% (11 /1 6)
56% (9/16)
44% (7/1 6)
38% (6/1 6)
38% (6/16)
38% (6/1 6)
31% (5/1 6)
31% (5/1 6)
31% (5/1 6)
25% (4/1 6)
13% (2/1 6)
6% (1/16)
0%(0/16)
0% (0/1 6)
Specificity
(c/c+d)
0%(0/71)
6% (4/71 )
13% (9/71)
20% (14/71 )
25% (18/71)
32% (23/71)
38% (27/71)
44% (31/71)
48% (34/71)
55% (39/71)
62% (44/71)
66% (47/71)
76% (54/71)
80% (57/71)
83% (59/71)
86% (61/71)
90% (64/71)
96% (68/71)
96% (68/71)
97% (69/71)
99% (70/71)
99% (70/71)
99% (70/71)
Positive
Predictive
Value
(b/b+d)
18% (16/87)
18% (15/82)
20% (15/77)
21% (15/72)
20% (13/66)
21% (13/61)
23% (13/57)
25% (13/53)
23% (11/48)
26% (11/43)
25% (9/36)
23% (7/31)
26% (6/23)
30% (6/20)
33% (6/1 8)
33% (5/1 5)
42% (5/1 2)
63% (5/8)
57% (4/7)
50% (2/4)
50% (54)
0% (0/1)
0% (0/1)
Negative
PredictivB
Value
(c/a+c)
.% (0/0)
80% (4/5)
90% (9/10)
93% (14/15)
86% (18/21)
89% (23/26)
90% (27/30)
91% (31/34)
87% (34/39)
89% (39/44)
86% (44/51)
84% (47/56)
84% (54/64)
85% (57/67)
86% (59/69)
85% (61/72)
85% (64/75)
86% (68/79)
85% (68/80)
83% (69/83)
82% (70/85)
81% (70/86)
81% (70/86)
#
Homes
87
87
87
87
87
87
87
87
87
87
87
87
87
87
87
87
87
87
87
87
87
87
87
Draft - Do Not Cite or Quote
C-5
September 19. 1996
-------
I
I
o
I
e
O
*^-
10 IOO 1.000 10000
Uncarpeted Floor Dual-Lead Loading tug sq fl)
V
S OS
IOO 1 50 200 260
O
en
s
IOO 150 2OO
Uncarpetrcl Door Dust-Lead Loading (UK sq ft)
Perft.rni.n- « Ch«
3
2 -
O SO IOO 150 2OO
L'ncarpeted Floor Dust-Lead Loading (wg sq fl
CO
CO
O)
Figure C 3a. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Uncarpeted Floors
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 10 /ig/dL.
-------
Table C-3a. Performance Characteristics of Potential Dust-Lead Standards for Uncarpeted
Floors Based on Data from the Rochester Lead-in-Dust Study and a Blood-
Lead Standard of 10//g/dL.
Uncarpeted
Floor Dust-
Lead
Standard
(pg/ft*)
1
2
3
4
5
10
15
20
25
50
75
100
125
150
175
200
300
400
% Homes
Failing the
Standard
(# Homes Faffing/
# Homes)
100% < 196/1 97)
98% (193/1 97)
94% (185/1 97)
93% (184/197)
91% (180/1 97)
76% (150/1 97)
59% (11 6/1 97)
43% (84/1 97)
29% (58/1 97)
10% (19/197)
6% (11/197)
5% (9/1 97)
3% (6/197)
3% (6/197)
3% (6/197)
3% (5/197)
2% (4/1 97)
2% (3/1 97)
Blood-Lead Standard of 10 pg/dL
Sensitivity
(b/a+b)
100% (47/47)
100% (47/47)
98% (46/47)
98% (46/47)
96% (45/47)
94% (44/47)
77% (36/47)
66% (31/47)
55% (26/47)
19% (9/47)
13% (6/47)
11% (5/47)
9% (4/47)
9% (4/47)
9% (4/47)
6% (3/47)
6% (3/47)
4% (2/47)
Specificity
(c/c+d)
1% (1/1 50)
' 3% (4/150)
7% (11 /1 50)
8% (12/1 50)
10% (15/1 50)
29% (44/1 50)
47% (70/1 50)
65% (97/1 50)
79% (11 8/1 50)
93% (140/1 50)
97% (145/1 50)
97% (146/1 50)
99% (148/1 50)
99% (148/1 50)
99% (148/1 50)
99% (148/1 50)
99% (149/1 50)
99% (149/1 50)
Positive
Predictive
Value
(b/b+d)
24% (47/196)
24% (47/193)
25% (46/185)
25% (46/184)
25% (45/180)
29% (44/1 50)
31% (36/1 16)
37% (31/84)
45% (26/58)
47% (9/19)
55% (6/11)
56% (5/9)
67% (4/6)
67% (4/6)
67% (4/6)
60% (3/5)
75% (3/4)
67% (2/3)
Negative
Predictive
Value
(c/a+c)
100% (1/1)
100% (4/4)
92% (11/1 2)
92% (12/1 3)
88% (15/1 7)
94% (44/47)
86% (70/81 1
86% (97/1 13)
85% (11 8/1 39)
79% (140/1 78)
78% (145/186)
78% (146/188)
78% (148/191)
78% (148/191)
78% (148/191)
77% ( 148/1 92)
77% (149/1 93)
77% (149/1 94)
9
Homes
197
197
197
197
197
197
197
197
197
197
197
197
197
197
197
197
197
197
Draft - Do Not Cite or Quote
C-7
September 19. 1996
-------
!
o
1O IOO I OOO 1O.OOO IOO.OOO
Carpeted Floor Dust-Lrod Loading <^g 'sq fl)
u OB-
E o.
I ..J
>.„
9
00
100 ISO 2OO Z5O
Carpeted Floor Dust-Lead Loading (ng sq ft)
itic ^ True Poilllve (Sensitivity)
F*U* Po«lliv. (l-Spectflrityt
Carpeted Floor Dust-Lead Loading (..g sq ft)
10
(0
Figure C-4a. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Carpeted Floors
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 10 //g/dL.
-------
Table C-4a. Performance Characteristics of Potential Dust-Lead Standards for Carpeted
Floors Based on Data from the Rochester Lead-in-Dust Study and a Blood-
Lead Standard of 10//g/dL.
Carpeted
Floor Dust-
Lead
Standard
Ovg/ft2)
1
2
3
4
5
10
15
20
25
50
75
100
125
150
175
200
300
400
% Homes
Falling the
Standard
(# Homes Failing/
# Homes)
98% (176/1 79)
93% (167/1 79)
91% (163/1 79)
89% (159/1 79)
87% (155/1 79)
67% (11 9/1 79)
40% (7 1/1 79)
23% (41 /1 79)
16% (29/1 79)
5% (8/1 79)
2% (4/179)
2% (3/1 79)
2% (3/1 79)
1% (2/1 79)
1% (2/179)
1% (2/1 79)
1% (2/1 79)
1% (2/1 79)
Blood-Lead Standard of 10/ig/dL
Sensitivity
(b/a+b)
100% (38/38)
100% (38/38)
100% (38/38)
100% (38/38)
100% (38/38)
87% (33/38)
66% (25/38)
45% (17/38)
32% (12/38)
8% (3/38)
3% (1/38)
0% (0/38)
0% (0/38)
0% (0/38)
0% (0/3B)
0% (0/38)
0% (0/38)
0% (0/38)
Specificity
(c/c+d)
2% (3/141)
9% (12/141)
11% (16/141)
14% (20/1 41)
17% (24/1 41)
39% (55/1 41)
67% (95/1 41)
83% (117/141)
88% (124/1 41)
97% (136/141)
98% (138/1 41)
98% (138/1 41)
98% (138/1 41)
99% (139/1 41)
99% (139/1 41)
99% (139/1 41)
99% (139/1 41)
99% (139/1 41)
Positive
Predictivo
Value
(b/b+d)
22% (38/1 76)
23% (38/167)
23% (38/163)
24% (38/1 59)
25% (38/1 55)
28% (33/1 19)
35% (25/71 1
42% (17/41)
41% (12/29)
38% (3/8)
25% (1/4)
0% (0/3)
0% (0/3)
0% (0/2)
0% (0/2)
0% (0/2)
0% (0/2)
0% (0/2)
Negative
Predictive
Value
(c/a+c)
100% (3/3)
100% ( 12/1 2)
100% (16/1 6)
100% (20/20)
100% (24/24)
92% (55/60)
88% (95/1 08)
85% (11 7/1 38)
83% (124/1 50)
80% (136/1 71)
79% (138/175)
78% (138/1 76)
78% (138/1 76)
79% (139/1 77)
79% (139/1 77)
79% (139/1 77)
79% (139/1 77)
79% (139/1 77)
»
Homes
179
179
179
179
179
179
179
179
179
179
179
179
179
179
179
179
179
179
Draft - Do Not Cite or Quote
C-9
September 19, 1996
-------
I
\
-J" — ^fc
*** •- •*•
- <&"--'" ^"-^
z "• _-^_, - ,*" — •*
Combined floor D
100 i.ooo
st -Lead Loading fug scl '*•)
u o a -
10O ISO 8OO
ln^d Floor Dual-Lead Loading (,/g aq TO
l>( fc.l iimni • < till nrt PI i«l |i
9
o
100 ISO 200
Combined Floor Dust -Lead Loading {.«g sq fl)
. (I -5pecin)
100 150 ^00
nbin'-d Floor Dust -Lend Loading (ug sq ft)
s
Figure C 5a. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Combined Floors
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 10 /sg/dL.
-------
Table C-5a. Performance Characteristics of Potential Dust-Lead Standards for Combined
Carpeted-Uncarpeted Floors Based on Data from the Rochester Lead-in-Dust
Study and a Blood-Lead Standard of 10 jig/dL.
Combinod
Floor
Dust-Lead
Standard
big/ft*)
1
2
3
4
5
10
15
20
25
50
75
100
125
150
175
200
300
400
% Homes
Falling the
Standard
(# Homes Failing/
it Homes)
100% (205/205)
98% (201/205)
96% (196/205)
94% (192/205)
91% (186/205)
76% (156/205)
53% (109/205)
36% (74/205)
27% (56/205)
9% (19/205)
6% (12/205)
5% (10/205)
4% (8/205)
4% (8/205)
4% (8/205)
3% (7/205)
2% (5/205)
2% (5/205)
Blood-Lead Standard of 10 jig/dL
Sensitivity
(b/a+b)
100% (48/48)
100% (48/48)
98% (47/48)
98% (47/48)
98% (47/48)
94% (45/48)
81% (39/48)
67% (32/48)
52% (25/48)
17% (8/48)
13% (6/48)
10% (5/48)
8% (4/48)
8% (4/48)
8% (4/48)
6% (3/48)
4% (2/48)
4% (2/48)
Specificity
(c/c+d)
0% (0/1 57)
3% (4/1 57)
5% (8/1 57)
8% (12/157)
12% (18/1 57)
29% (46/157)
55% (87/1 57)
73% (11 5/1 57)
80% (126/1 57)
93% 1 146/1 57)
96% (151/1 57)
97% (152/157)
98% (153/1 57)
98% (153/1 57)
98% (153/1 57)
98% (153/1 57)
98% (154/1 57)
98% (154/1 57)
Positive
Predictive
Value
(b/b+d)
23% (48/205)
24% (48/201)
24% (47/1 96)
25% 147/1 92)
25% (47/186)
29% 145/1 56)
36% 139/109)
43% (32/74)
45% (25/56)
42% (8/1 9)
50% (6/1 2)
50% (5/10)
50% (4/8)
50% (4/8I
50% (4/8I
43% (3/7)
40% (2/5I
40% (2/5)
Negative
Predictive
Value
(c/a+c)
.% (0/0)
100% (4/4)
89% (8/9)
92% (12/1 3)
95% (18/1 9)
94% (46/49)
91% (87/96)
88% (11 5/1 31)
85% (126/1 49)
79% (146/186)
78% (151/1 93)
78% (152/1 95)
78% (153/1 97)
78% (153/1 97)
78% (153/1 97)
77% (153/1 98)
77% ( 154/200)
77% (154/200)
#
Homes
205
205
205
205
205
205
205
205
205
205
205
205
205
205
205
205
205
205
Draft - Do Not Cite or Quote
C-11
September 19, 1996
-------
S
100 1 OOO 10 OOO
Window Sill Dusk-Lead Loading (,-K sq tM
5OO 1.0OO I GOO 2 000 2.&OO 3 ODD
) 5OO 4 OOO
O
K)
5OO I OOO
e ooo e soo
SOD i ooo i aoo
Figure C-6a. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Window Sills
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 10//g/dL.
-------
Table C-6a. Performance Characteristics of Potential Dust-Lead Standards for Window
Sills Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 10 /vg/dL.
Window
Sill Dust-
Lead
Standard
b/g/ft*)
5
10
15
20
25
50
75
100
125
150
175
200
250
300
350
400
450
500
600
700
800
900
1000
1500
2000
2500
3000
3500
4000
% Homes
Failing the
Standard
(It Homes Failing/
# Homes)
98% (192/1 96)
98% (192/196)
98% (192/196)
97% (190/196)
96% (188/196)
88% (172/196)
77% (150/196)
67% (131/196)
62% (121/196)
57% (111/196)
51% (100/196)
47% (92/196)
37% (73/1 96)
35% (68/196)
28% (55/1 96)
26% (50/196)
24% (46/196)
21% (42/1 96)
18% (36/196)
15% (30/196)
12% (24/196)
12% (24/196)
12% (23/1 96)
8% (16/196)
6% 1 12/1 96)
6% (11/196)
4% (8/196)
3% (6/196)
2% (4/196)
Blood-Lead Standard of 10 pg/dL
Sensitivity
(b/a+b)
100% (45/45)
100% (45/45)
100% (45/45)
100% (45/45)
100% (45/45)
93% (42/45)
87% (39/45)
80% (36/45)
78% (35/45)
71% (32/45)
64% (29/45)
62% (28/45)
56% (25/45)
53% (24/45)
42% (19/45)
40% (18/45)
36% (16/45)
33% 1 15/45)
27% (12/45)
24% (11/45)
20% (9/45)
20% (9/45)
20% (9/45)
18% (8/45)
16% (7/45)
13% (6/45)
9% (4/45)
9% (4/45)
4% (2/45)
Specificity
(c/c+d)
3% (4/1 51)
3% (4/1 51)
3% (4/1 51)
4% (6/151)
5% (8/1 51)
14% (21/151)
27% (40/151)
37% (56/1 51)
43% (65/1 51)
48% (72/1 51)
53% (80/1 51)
58% (87/1 51)
68% (103/151)
71% (107/151)
76% (115/151)
79% (119/1 51)
80% (121/151)
82% (124/151)
84% (127/151)
87% (132/151)
90% (136/151)
90% (136/1 51)
91% (137/151)
95% (143/1 51)
97% (146/151)
97% (146/151)
97% (147/151)
99% (149/151)
99% (149/1 51)
Positive
Predictive
Value
(b/b+d)
23% (45/192)
23% (45/192)
23% (45/192)
24% (45/190)
24% (45/188)
24% (42/172)
26% (39/150)
28% (36/131)
29% (35/121)
29% (32/1 11)
29% (29/100)
30% (28/92)
34% (25/73)
35% (24/68)
35% (19/55)
36% (18/50)
35% (16/46)
36% (15/42)
33% (12/36)
37% (11/30)
38% (9/24)
38% (9/24)
39% 19/23)
50% 18/16)
58% (7/1 2)
55% (6/11)
50% (4/8)
67% (4/6)
50% (2/4)
Negative
Predictive
Value
(c/a+c)
100% (4/4)
100% (4/4)
100% (4/4)
100% 16/6)
100% (8/8)
88% (21/24)
87% (40/46)
86% (56/65)
87% (65/75)
85% (72/85)
83% (80/96)
84% (87/104)
84% (103/1 23)
84% ( 107/1 28)
82% (115/141)
82% (119/1 46)
81% (121/1 50)
81% (124/1 54)
79% (127/1 60)
80% 1 132/1 66)
79% ( 136/1 72)
79% (136/1 72)
79% (137/1 73)
79% (143/1 80)
79% (146/184)
79% (146/1 85)
78% (147/188)
78% (149/1 90)
78% 1 149/1 92)
#
Homes
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
Draft - Do Not Cite or Quote
C-13
September 19, 1996
-------
10O l.OOO IO.OOO 100.OOO
Window Trough Dual-Lead Loading {ug/sq ft)
o e -
o 7 -
o e -
6
04 -
4 000 B OOO 12 OOO \g OOO ZO.OOO
Window Trough Dusl-Lead Loading (ug sq ft)
O
o-
3!
4 OOO 3000 12 OOO 16 OOO
Window Trough Dust-Lead Loading (Mg/sq ft)
Tru« Posilive (Spniil.vily)
4 OOO a OOO 12000 i oooo
Window Trough Dust-Lead Loading (wg/sq ft)
e Ch«r»cterlBUc - Sent * Sprc * PPV * NPV
to
CO
O)
Figure C-7a. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Window Troughs
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 10 //g/dL.
-------
Table C-7a. Performance Characteristics of Potential Dust-Lead Standards for Window
Troughs Based on Data from the Rochester Lead-in-Dust Study and on a
Blood-Lead Standard of 10/fg/dL.
Window
Trough
Dust-Lead
Standard
U/g/ft2)
SO
100
200
300
400
500
600
700
800
900
1000
1500
2000
2500
3000
3500
4000
4500
5000
10000
15000
20000
30000
40000
50000
100000
% Homes
Falling the
St&ndsrd
(» Homes
Failing/
# Homes)
98% (185/1 89)
94% (178/1 89)
89% (168/1 89)
84% (159/1 89)
83% 1 156/1 89)
80% (151/1 89)
78% (148/189)
76% (144/189)
73% (138/1 89)
73% (137/1 89)
70% (133/1 89)
64% (121/1 89)
62% (118/1 89)
61% (11 5/1 89)
59% (11 1/1 89)
56% (106/189)
55% (103/189)
51% (96/1 89)
48% (9 1/1 89)
40% (75/1 89)
34% (65/1 89)
30% (57/1 89)
23% (44/1 89)
20% (38/1 89)
16% (31 /1 89)
10% (18/1 89)
Blood-Lead Standard of 10 pg/dL
Sensitivity
(b/a+b)
95% (41/43)
95% (41/43)
95% (41/43)
88% (38/43)
86% (37/43)
84% (36/43)
81% (35/43)
81% (35/43)
77% (33/43)
77% (33/43)
77% (33/43)
70% (30/43)
70% (30/43)
70% (30/43)
70% (30/43)
70% (30/43)
70% (30/43)
67% (29/43)
63% (27/43)
56% (24/43)
47% (20/43)
40% (17/43)
37% (16/43)
37% (16/43)
30% (13/43)
19% (8/43)
Specificity
(c/c+d)
1% (2/1 46)
6% (9/146)
13% (19/146)
17% (25/146)
19% (27/146)
21% (31/146)
23% (33/146)
25% (37/146)
28% (41/146)
29% (42/146)
32% (46/146)
38% (55/146)
40% (58/146)
42% (61/146)
45% (65/146)
48% (70/146)
50% (73/146)
54% (79/146)
56% (82/146)
65% (95/146)
69% (101/146)
73% (106/146)
81% (11 8/1 46)
85% (124/1 46)
88% (128/1 46)
93% (136/1 46)
Positive
Predictive
Value
(b/b+d)
22% (41/185)
23% (41/1 78)
24% (41/168)
24% (38/1 59)
24% (37/1 56)
24% (36/1 51)
24% (35/148)
24% (35/144)
24% (33/1 38)
24% (33/1 37)
25% (33/1 33)
25% 130/121)
25% (30/1 18)
26% (30/1 15)
27% (30/1 11)
28% (30/106)
29% (30/103)
30% (29/96)
30% (27/91)
32% (24/75)
31% (20/65)
30% (17/57)
36% (16/44)
42% (16/38)
42% (13/31)
44% (8/18)
Negative
Predictive
Value
(c/a+c)
50% (2/4)
82% (9/11)
91% (19/21)
83% (25/30)
82% (27/33)
82% (31/38)
81% (33/41)
82% (37/45)
80% 141/51)
81% (42/52)
82% (46/56)
81% (55/68)
82% (58/71)
82% (61/74)
83% (65/78)
84% (70/83)
85% (73/86)
85% (79/93)
84% (82/98)
83% (95/1 14)
82% (101/1 24)
80% (106/1 32)
81% (118/1 45)
82% (124/151)
81% (128/1 58)
80% (136/1 71 )
#
Homes
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
Draft - Do Not Cite or Quote
C-15
September 19, 1996
-------
JO -i
Percent of Compo
BO IOO
Percent of Cornpo
. Home with Deteriorated Ji.lr. lor LOP
12 14
F.I.. POBiLlv
P*rceril of Components within B Home with Delenoralrd Interior LBP
Figure C-8a. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Percent of
Interior Components Within a Home with Deteriorated LBP, Based on Data from the Rochester Lead-in-Dust
Study and a Blood-Lead Standard of 10 //g/dL.
-------
Table C-8a. Performance Characteristics of Potential Paint-Lead Standards for the Percent
of Interior Components within a Home with Deteriorated Lead-Based Paint
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 10//g/dL.
Percent of
Interior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homos
Failing the
Standard
(tt Homes Failing/
# Homes)
64% (132/205)
64% (132/205)
64% (132/205)
64% ( 132/205)
63% (129/205)
59% (121/205)
50% (103/205)
46% (94/205)
43% (89/205)
Blood-Lead Standard of 10//g/dL
Sensitivity
(b/a+b)
73% (35/48)
73% (35/48)
73% (35/48)
73% (35/48)
73% (35/48)
71% (34/48)
65% (31/48)
58% (28/48)
56% (27/48)
Specificity
(c/c+d)
38% (60/1 57)
38% (60/1 57)
38% (60/1 57)
38% (60/1 57)
40% (63/1 57)
45% (70/1 57)
54% (85/1 57)
58% (91/1 57)
61% (95/1 57)
Positive
Pradictivo
Value
(b/b+d)
27% (35/1 32)
27% (35/1 32)
27% (35/1 32)
27% (35/1 32)
27% (35/1 29)
28% (34/1 21 )
30% (31/103)
30% (28/94)
30% (27/89)
Negative
Predictive
Value
(c/a+c)'
82% (60/73)
82% (60/73)
82% (60/73)
82% (60/73)
83% (63/76)
83% (70/84)
83% (85/102)
82% (91/1 11)
82% (95 /1 16)
#
Homes
205
205
205
205
205
205
205
205
205
8 Note, there were 73 homes where the percent of interior components with deteriorated LBP was zero. Of the 73 homes,
13 homes had a child with a blood-lead concentration above 10;/g/dL.
Draft - Do Not Cite or Quote
C-17
September 19, 1996
-------
§•
3 1
0 SO 30 40 50 BO 70 BO B
Percent, of Cornpo
o
CD
A a s 10 12 M ie 10 20
Percent of Components within B Home with Deteriorated Exterior LBP
Percent of Compo
Figure C-9a. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Percent of
Exterior Components Within a Home with Deteriorated LBP, Based on Data from the Rochester Lead-in-Dust
Study and a Blood-Lead Standard of 10 //g/dL.
-------
Table C-9a. Performance Characteristics of Potential Paint-Lead Standards for the Percent
of Exterior Components within a Home with Deteriorated Lead-Based Paint
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 10 /ig/dL.
Percent of
Exterior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20 0%
% Homes
Failing the
Standard
(# Homes
Failing/
# Homes)
47% (96/204)
47% (96/204)
47% (96/204)
47% (96/204)
47% (96/204)
47% (96/204)
47% (95/204)
47% (95/204)
47% (95/204)
Blood-Lead Standard of 10/rg/dL
Sensitivity
(b/a + b)
56% (27/48)
56% (27/48)
56% (27/48)
56% (27/48)
56% (27/48)
56% (27/48)
56% (27/48)
56% (27/48)
56% (27/48)
Specificity
(c/c+d)
56% (87/1 56)
56% (87/1 56)
56% (87/1 56)
56% (87/1 56)
56% (87/1 56)
56% (87/1 56)
56% (88/1 56)
56% (88/1 56)
56% (88/156)
Positive
Predictive
Value
(b/b+d)
28% (27/96)
28% (27/96)
28% (27/96)
28% (27/96)
28% (27/96)
28% (27/96)
28% (27/95)
28% (27/95)
28% (27/95)
Negative
Predictive
Value
(c/a+c)a
81% (87/1 08)
81% (87/1 08)
81% (87/1 08)
81% (87/108)
81% (87/1 08)
81% (87/108)
81% (88/1 09)
81% (88/109)
81% (88/1 09)
#
Homes
204
204
204
204
204
204
204
204
204
Note, there were 108 homes where the percent of exterior components with deteriorated LBP was zero. Of the 108
homes, 19 homes had a child with a blood-lead concentration above 10//g/dL.
Draft - Do Not Cite or Quote
C-19
September 19, 1996
-------
~r
i
|
o
ft
0
1
m
n
o
Co
1
3
0-
1
I
1
J 20 -
1
1 !
1 '
1 '° 1
1
i
0 9 •
o a
«
E
2 °e M
5 05 ^
u
8
o
£ 03 -
O 1 -
m
c
-
J _^
M*
— —
a
i o -2 o : o
5 ° 1 D
1 **-* S"*3 »
-^*:-=0-| „- * 3 «•-
1 - = - o
) 10 20 30 40 50 60 70 8O DO 1C
P ,oen f T,P
0 O O exterior
~~~— — _
**
•) 2 4 0 A 1O IB 14 IB )B i
D 0
O B
1 °7
E
a '
1
^ 05
|
«
€
°-
08
O
^
I
I
6 3-
S
g
1
D
. .
•— — — — .__
_ _ , - '
— — • • — —
f
O 2 4 a a 10 12 14 IB IB 2
Sp«r.nc.ly ^
2 4 6 fi IO IZ 1-1 16 IB 2
Percent of Components w.thin a Home w.U, D^ter.orsled Interior or Exter.or LBP
_ _ . . _ .
Figure C-10a. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Maximum of
the Percent of Interior Components with Deteriorated LBP and the Percent of Exterior Components with
Deteriorated LBP, Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 10
//g/dL.
-------
Table C-10a. Performance Characteristics of Potential Paint-Lead Standards for the
Maximum of the Percent of Interior Components with Deteriorated LBP and
the Percent of Exterior Components with Deteriorated LBP Based on Data
from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of
10//g/dL.
Maximum of the
Percent of
Interior/Exterior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
200%
% Homes
Failing the
Standard
(# Homes Failing/
# Homes)
74% (150/204)
74% (150/204)
74% 1 150/2041
74% (150/204)
73% (149/204)
71% (145/204)
66% (135/204)
64% (131/204)
63% (129/204)
Blood-Lead Standard of 10//g/dL
Sensitivity
(b/a+b)
81% (39/48)
81% (39/48)
81% (39/48)
81% (39/48)
81% (39/48)
79% (38/48)
77% (37/48)
75% (36/48)
75% (36/48)
Specificity
(c/c+d)
29% (45/1 56)
29% (45/1 56)
29% (45/1 56)
29% (45/1 56)
30% (46/1 56)
31% (49/1 56)
37% 158/1 56)
39% (61 /1 56)
40% (63/1 56)
Positive
Predictive
Value
(b/b+d)
26% (39/1 50)
26% (39/1 50)
26% (39/1 50)
26% (39/1 50)
26% (39/149)
26% (38/1 45)
27% (37/1 35)
28% (36/1 31)
28% (36/1 29)
Negative
Predictive
Value
(C/B + C)"
83% (45/54)
83% (45/54)
83% (45/54)
83% (45/54)
84% (46/55)
83% (49/59)
84% (58/69)
84% (61/73)
84% (63/75)
#
Homes
204
204
204
204
204
204
204
204
204
Note, there were 54 homes where the maximum percent of the mtenor components with deteriorated LBP and the percent
of exterior components with detenorated LBP was zero. Of the 54 homes, 9 homes had a child with a blood-lead
concentration above 10j/g/dL.
Draft - Do Not Cite or Quote
C-21
September 19, 1996
-------
s
1
5
l
«?d LDP pe. In
- i . ,r. .' .
I
I °'
I 03
Stn* * fif.tr . Ppv . NPV
to
Figure C 1 la. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Average
Percent of Deteriorated LBP per Interior Component, Based on Data from the Rochester Lead-in-Dust Study
and a Blood-Lead Standard of 10//g/dL.
-------
Table C-11a. Performance Characteristics of Potential Paint-Lead Standards for the
Average Percent of Deteriorated Lead-Based Paint per Interior Component
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 10/sg/dL.
Average
Percent of
Deteriorated
LBPper
Interior
Component
1.0%
2.546
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homes
Failing the
Standard
itt Homes Failing/
# Homes)
79% (162/205)
53% (108/205)
28% (58/205)
10% (21/205)
3% (7/205)
0% (0/205)
0% (0/205)
0% (0/205)
0% (0/205)
Blood-Lead Standard of 10 /ig/dL
Sensitivity
(b/a+b)
85% (41/48)
65% (31/48)
38% (18/48)
19% (9/48)
10% (5/48)
0% (0/48)
0% (0/48)
0% (0/48)
0% (0/48)
Specificity
(c/c+d)
23% (36/1 57)
51% (80/1 57)
75% (11 7/1 57)
92% (145/1 57)
99% (155/1 57)
100% (157/1 57)
100% (157/1 57)
100% (157/1 57)
100% (157/1 57)
Positive
Predictive
Value
(b/b+d)
25% (41/1 62)
29% (31/108)
31% (18/58)
43% (9/21)
71% (5/7)
.% (0/0)
.% (0/0)
.% (0/0)
.% (0/0)
Negative
Predictive
Value
(c/a+c)8
84% (36/43)
83% (80/97)
80% (11 7/1 47)
79% (145/184)
78% (155/1 98)
77% (157/205)
77% ( 157/205)
77% (157/205)
77% (157/205)
9
Homes
205
205
205
205
205
205
205
205
205
Note, there were 43 homes where the average percent of deteriorated LBP per intenor component was zero. Of the 43
homes. 7 homes had a child with a blood-lead concentration above 10//g/dL.
Draft - Do Not Cite or Quote
C-23
September 19. 1996
-------
o
3'
I
= S
A **** "**•
IB 20
o
f IScnlltlvlly)
CO
CO
01
Figure C-12a. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Average
Percent of Deteriorated LBP per Exterior Component, Based on Data from the Rochester Lead-in-Dust Study
and a Blood-Lead Standard of 10 //g/dL.
-------
Table C-12a. Performance Characteristics of Potential Paint-Lead Standards for the
Average Percent of Deteriorated Lead-Based Paint per Exterior Component
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 10 /ig/dL.
Average
Percent of
Deteriorated
LBPper
Exterior
Component
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homes
Failing the
Standard
(# Homes
Failing/
n Homes)
79% (161/204)
52% (106/204)
31% (64/204)
19% (38/204)
1 2% (25/204)
5% (10/204)
2% (3/204)
1% (1/204)
0% (0/204)
Blood-Lead Standard of lO/igML
Sensitivity
(b/a+b)
85% (41/48)
58% (28/48)
40% (19/48)
31% (15/48)
27% 1 13/48)
8% (4/48)
6% (3/48)
2% (1/48)
0% (0/48)
OMAMM!MI*U
apacinCiiy
(c/c-t-d)
23% (36/1 56)
50% (78/1 56)
71% (111/156)
85% (133/1 56)
92% 1 144/1 56)
96% (150/1 56)
100% (156/1 56)
100% 1 156/1 56)
100% (156/1 56)
Positive
Predictive
Value
(b/b+d)
26% (41 /1 61)
26% (28/106)
30% (19/64)
40% 11 5/38)
52% (13/25)
40% (4/10)
100% (3/3)
100% (1/1)
.% (0/0)
Negative
Pradictivs
Value
Jc/a+c)*
84% (36/43)
80% (78/98)
79% (11 1/140)
80% (133/1 66)
80% (144/1 79)
77% (150/1 94)
78% (156/201)
77% (156/203)
77% (156/204)
#
Homos
204
204
204
204
204
204
204
204
204
a Note, there were 43 homes where the average percent of deteriorated LBP per extenor component was zero. Of the 43
homes, 7 homes had a child with a blood-lead concentration above 10 jig/dL.
Draft - Do Not Cite or Quote
C-25
September 19, 1996
-------
o
I
I
3 8
2
r--«—
W ** H * **
Is §;s
- • •
<-> O 5 -
10 IB
Sp.cific.lv
— Poilllvr Pifdir
' ^ • ^ ' ^ i ^ . ^ Nv|alK-F Predlc
O
ro
O)
I
Figure C-13a. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Maximum of
the Average Percent of Deteriorated LBP per Interior Component and Average Percent of Deteriorated LBP
per Exterior Component, Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard
of 10/ig/dL
-------
Table C-13a. Performance Characteristics of Potential Paint-Lead Standards for the
Maximum of the Average Percent of Deteriorated LBP per Interior
Component and the Average Percent of Deteriorated LBP per Exterior
Component Based on Data from the Rochester Lead-in-Dust Study and a
Blood-Lead Standard of 10//g/dL.
Maximum of
Average
Percent of
Interior/Exterior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homes
Failing the
Standard
|# Homes
Failing/
# Homes)
88% (179/204)
71% (144/204)
48% (97/204)
25% (50/204)
13% (27/204)
5% 110/204)
2% (3/204)
1% (1/204)
0% (0/204)
Blood-Lead Standard of 10 pg/dL
Sensitivity
(b/a+b)
94% (45/48)
77% (37/48)
60% (29/48)
44% (21/48)
31% (15/48)
8% (4/48)
6% (3/48)
2% (1/48)
0% (0/48)
Specificity
(c/c+d)
14% (22/1 56)
31% (49/1 56)
56% (88/1 56)
81% (127/1 56)
92% (144/1 56)
96% (150/1 56)
100% (156/156)
100% (156/1 56)
100% (156/1 56)
Posftivo
Predictive
Value
(b/b+d)
25% (45/1 79)
26% (37/144)
30% (29/97)
42% (21/50)
56% (15/27)
40% (4/10)
100% (3/3)
100% (1/1)
.% (0/0)
Negative
Predictive
Value
(c/e+c)8
88% (22/25)
82% (49/60)
82% (88/107)
83% (127/1 54)
81% (144/1 77)
77% (150/1 94)
78% (156/201)
77% ( 156/203)
77% (156/204)
9
Homes
204
204
204
204
204
204
204
204
204
a Note, there were 25 homes where the maximum of the average percent of deteriorated LBP per interior component and the
average percent of deteriorated LBP per exterior component was zero. Of the 25 homes. 3 homes had a child with a
blood-lead concentration above 10 /yg/dL.
Draft - Do Not Cite or Quote
C-27
September 19. 1996
-------
I
o
I
5
I ,0
10O
Dnpline :
E
"o 06
I 200 I 6OO ?- 000 2 4OO 2 BOO 3 SOD 3 BOO J OOO
Driphne Soil-L*-»d Concentration (i.g g)
9
N)
CD
<£ o-j
40O 801
1 2OO 1 8OO 2.OOO 2 4OO 2 BOO 3 200 3.800 -« 000
Dnpline Soil -Lead Cone en I rat ion (ug g)
Dnpline
3 OOO Z 40O 8 BOO 3 ZOO
Concentration («g g)
S*ni • Sprc * PPV »
3 6OO -J OOO
Figure C-14a. Summary Graphs of the Performance Characteristics of Potential Dripline Soil-Lead Standards Based on
Data from the Baltimore Repair and Maintenance Study and a Blood-Lead Standard of 10 /ig/dL.
-------
Table C-14a. Performance Characteristics of Potential Dripline Soil-Lead Standards Based
on Data from the Baltimore Repair and Maintenance Study and a Blood-Lead
Standard of 10 /ig/dL.
Dripline
Soil-Lead
Standard
(//fl/fl)
25
50
75
100
125
150
175
200
300
400
500
600
700
800
900
1,000
1,500
2,000
2,500
3,000
3,500
4000
% Homes
Failing the
Standard
(tt Homes Failing/
# Homes)
100% (27/27)
85% (23/27)
74% (20/27)
70% 1 19/27)
67% (18/27)
67% (18/27)
63% (17/27)
63% (17/27)
63% (17/27)
59% (16/27)
56% (15/27)
56% (15/27)
48% (13/271
44% (12/271
44% (12/27)
44% (12/27)
41% (11/27)
7% (2/27)
7% (2/27)
7% (2/27)
4% (1/27)
0% (0/27)
Blood-Lead Standard of 10 pg/dL
Sensitivity
(b/a+bl
100% (8/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
75% (6/8)
63% (5/8)
63% (5/8)
38% (3/8)
38% (3/8)
38% (3/8)
38% (3/8)
38% (3/8)
0% (0/8)
0% (0/8)
0% (0/8)
0% (0/8)
0% (0/8)
Specificity
(c/c+d)
0%(0/19)
16% (3/1 9)
32% (6/19)
37% (7/1 9)
42% (8/1 9)
42% (8/1 9)
47% (9/19)
47% (9/1 9)
47% (9/19)
47% (9/1 9)
47% (9/1 9)
47% (9/19)
47% (9/1 9)
53% (10/19)
53% (10/19)
53% (10/19)
58% (11/1 9)
90% (17/19)
90% (17/1 9)
90% (17/1 9)
95% (18/1 9)
100% (19/1 9)
Positive
Predictive
Value
(b/b+d)
30% (8/27)
30% (7/23)
35% (7/20)
37% (7/19)
39% (7/18)
39% (7/18)
41% (7/1 7)
41% (7/1 7)
41% (7/1 7)
38% (6/16)
33% (5/1 5)
33% (5/1 5)
23% (3/1 3)
25% (3/1 2)
25% (3/1 2)
25% (3/1 2)
27% (3/11)
0% (0/2)
0% (0/2)
0% (0/2)
0%(0/1)
.% (0/0)
Negative
Predictive
Value
(c/a+c)
.% (0/0)
75% (3/4)
86% (6/7)
88% (7/8)
89% (8/9)
89% (8/9)
90% (9/10)
90% (9/10)
90% (9/10)
82% (9/11)
75% (9/1 2)
75% (9/1 2)
64% (9/14)
67% (10/1 5)
67% (10/1 5)
67% (10/15)
69% (11/1 6)
68% (17/25)
68% (17/25)
68% (17/25)
69% (18/26)
70% (19/27)
#
Homes
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
Draft - Do Not Cite or Quote
C-29
September 19. 1996
-------
I
0
'
o
I
g
I "
'x'"^.
n
s
Moor Dual-Lend Loading (t.g sq ft)
- [• .-,-).' - (-,,1.1.1.11 , i
F»lst Poiitlve ( 1 -Specificity)
O)
Figure C-15a. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Combined
Carpeted-Uncarpeted Floors, Based on Data from the Baltimore Repair and Maintenance Study and a Blood-
Lead Standard of 10/sg/dL.
-------
Table C-15a. Performance Characteristics of Potential Dust-Lead Standards for Combined
Carpeted-Uncarpeted Floors Based on Data from the Baltimore Repair and
Maintenance Study and a Blood-Lead Standard of 10/ig/dL.
Floor*
Dust-Lead
Stundard
big/ft2)
5
10
15
20
25
50
75
100
125
150
175
% Homes
Failing the
Standard
(» Homeis
Failing/)
# Homes))
98% (85/87)
91% (79/87)
83% (72/87)
78% (68/87)
74% (64/87)
25% (22/87)
7% (6/87)
2% (2/87)
2% (2/87)
1%(1/87)
0% (0/87)
Sensitivity
(b/a+b))
100% (39/39)
97% (38/39)
97% (38/39)
95% (37/39)
90% (35/39)
36% (14/39)
10% (4/39)
5% (2/39)
5% (2/39)
3% (1/39)
0% (0/39)
IBIood-Laad IStan
Specificity
(e/c+d))
4% (2/48)
15% (7/48)
29% (14/48)
35% (17/48)
40% (19/48)
83% (40/48)
96% (46/48)
100% (48/48)
100% (48/48)
100% (48/48)
100% (48/48)
dard of 10 //g/dL
Positive
Predictiv)e
Value)
(b/b+d))
46% (39/85)
48% (38/79)
53% (38/72)
54% (37/68)
55% (35/64)
64% (14/22)
67% (4/6)
100% (2/2)
100% (2/2)
100% (1/1)
.% (0/0)
Negative
Predietivle
Value)
(c/a+c))
100% (2/2)
88% (7/8)
93% (14/1 5)
90% (17/1 9)
83% (19/23)
62% (40/65)
57% (46/81)
57% (48/85)
57% (48/85)
56% (48/86)
55% (48/87)
#
Homes
87
87
87
87
87
87
87
87
87
87
87
Includes floor samples taken on the floor et the interior entryway.
Draft - Do Not Cite or Quote
C-31
September 19, 1996
-------
I
o
»•
-<
I
100 l.OOO
Wmdow Sill Dusl-Lead Loading [Mg, sq ft)
u
'\
bOO 1 OOO 1 500
Wmdow Sill Dust -Lead Loading d.g sq (t)
Sprrlflcily
Nrgotlv? Predic
n
w
NJ
5 o,
t
ADO I OOO I
Window Sitl Dust-Lead Loading
-------
Table C-16a. Performance Characteristics of Potential Dust-Lead Standards for Window
Sills Based on Data from the Baltimore Repair and Maintenance Study and a
Blood-Lead Standard of 10//g/dL.
Window Sill
Dust-Lead
Standard
(fig/ft1)
25
SO
75
100
125
150
175
200
300
400
500
1,000
1.500
2 000
% Homes
Failing the
Standard
(# Homes
Failing/
» Homes)
98% (85/87)
91% (79/87)
82% (71/87)
75% (65/87)
74% (64/87)
72% (63/87)
71% (62/87)
70% (61/87)
66% (57/87)
56% (49/87)
51% (44/87)
13% (11/87)
6% (5/87)
0% (0/87)
Blood-Lead Standard of 10 j/g/dL
Sensitivity
(b/a+b)
100% (39/39)
97% (38/39)
97% (38/39)
90% (35/39)
87% (34/39)
85% (33/39)
82% (32/39)
80% (31/39)
77% (30/39)
67% (26/39)
64% (25/39)
21% (8/39)
5% (2/39)
0% (0/39)
Specificity
(c/c+d)
4% (2/48)
1 5% (7/48)
31% (15/48)
38% (18/48)
38% (18/48)
38% (18/48)
38% (18/48)
38% (18/48)
44% (21/48)
52% (25/48)
60% (29/48)
94% (45/48)
94% (45/48)
100% (48/48)
Positive
Predictive
Value
(b/b+d)
46% (39/85)
48% (38/79)
54% (38/71)
54% (35/65)
53% (34/64)
52% (33/63)
52% (32/62)
51% (31/61)
53% (30/57)
53% (26/49)
57% (25/44)
73% (8/11)
40% (2/5)
.% (0/0)
Negative
Predictive
Value
(c/a+c)
100% (2/2)
88% (7/8)
94% ( 15/1 6)
82% (18/22)
78% (18/23)
75% (18/24)
72% (18/25)
69% (18/26)
70% (21/30)
66% (25/38)
67% (29/43)
59% (45/76)
55% (45/82)
55% (48/87)
»
Homes
87
87
87
87
87
87
87
87
87
87
87
87
87
87
Draft - Do Not Cite or Quote
C-33
September 19, 1996
-------
o
o
O
c
o
o -
100
I 000 10.0OO
Window Trough liu-,1 L< ml Loading (ng sq n !
T
I,
3 000 4 OOO
imdmg (..« -.q ft )
c>
CO
2 ODO 3 OOO
Trough Dust -Lead Loading (itg. aq t()
i ooo z ooo a ooo
Window Trough Dual - Lead leading (ug sq ft)
(o
8
Figure C-17a. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Window
Troughs, Based on Data from the Baltimore Repair and Maintenance Study and a Blood-Lead Standard of
10//g/dL.
-------
Table C-17a. Performance Characteristics of Potential Dust-Lead Standards for Window
Troughs Based on Data from the Baltimore Repair and Maintenance Study
and a Blood-Lead Standard of 10//g/dL.
Window
Trough
Dust-Lead
Standard
(fig/ft2)
200
300
400
500
600
700
800
900
1.000
1.500
2.000
2.500
3,000
3,500
4,000
4.500
S.OOO
10.000
15.000
20.000
30,000
40.000
50,000
60.000
% Homes
Failing the
Standard
(# Homes
Failing/
# Homes)
97% (83/86)
88% (76/86)
87% (75/86)
86% (74/86)
85% (73/86)
84% (72/86)
81% (70/86)
78% (67/86)
77% (66/86)
71% (61/86)
70% (60/86)
69% (59/86)
67% (58/86)
65% (56/86)
65% (56/86)
63% (54/86)
63% (54/86)
54% (46/86)
38% (33/86)
29% (25/86)
12% (10/86)
5% (4/86)
1%(1/86)
1%(1/86)
Blood-Lead Standard of 10 /ig/dL
Sensitivity
(b/a+b)
100% (38/38)
92% (35/38)
92% (35/38)
92% (35/38)
92% (35/38)
92% (35/38)
92% (35/38)
92% (35/38)
92% (35/38)
84% (32/38)
84% (32/38)
84% (32/38)
82% (31/38)
76% (29/38)
76% (29/38)
71% (27/38)
71% (27/38)
61% (23/38)
42% (16/38)
34% (13/38)
16% (6/38)
8% (3/38)
3% (1/38)
3% (1/38)
Specificity
(c/c+d)
6% (3/48)
15% (7/48)
17% (8/48)
19% (9/48)
21% (10/48)
23% (11/48)
27% (13/48)
33% (16/48)
35% (17/48)
40% 1 19/48)
42% (20/48)
44% (21/48)
44% (21/48)
44% (21/48)
44% (21/48)
44% (21/48)
44% (21/48)
52% (25/48)
65% (31/48)
75% (36/48)
92% (44/48)
98% (47/48)
100% (48/48)
100% (48/48)
Positive
Predictive
Value
(b/b+d)
46% (38/83)
46% (35/76)
47% (35/75)
47% (35/74)
48% (35/73)
49% (35/72)
50% (35/70)
52% (35/67)
53% (35/66)
53% (32/61)
53% (32/60)
54% (32/59)
53% (31/58)
52% (29/56)
52% (29/56)
50% (27/54)
50% (27/54)
50% (23/46
49% (16/33)
52% (13/25)
60% (6/10)
75% (3/4)
100% (1/1)
100% (1/1)
Negative
Predictive
Value
(c/a-fc)
100% (3/3)
70% (7/10)
73% 18/11)
75% (9/12)
77% (10/1 3)
79% (11/14)
81% (13/16)
84% (16/1 9)
85% (17/20)
76% (19/25)
77% (20/26)
78% (21/27)
75% (21/28)
70% (21/30)
70% (2 1/30)
66% (21/32)
66% (21/32)
63% (25/40)
59% (31/53)
59% (36/61)
58% (44/76)
57% (47/82)
57% (48/85)
57% (48/85)
»
Homes
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
Draft - Do Not Cite or Quote
C-35
September 19. 1996
-------
I
^
I
-
i
I
I
50 60
I Inle-i tor i.BJ'
CO
0>
if
is in
Figure C-18a. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Percent of
Interior Components Within a Home with Deteriorated LBP, Based on Data from the Baltimore Repair and
Maintenance Study and a Blood-Lead Standard of 10 //g/dL.
-------
Table C-18a. Performance Characteristics of Potential Paint-Lead Standards for the
Percent of Interior Components within a Home with Deteriorated Lead-Based
Paint Based on Data from the Baltimore Repair and Maintenance Study and a
Blood-Lead Standard of 10//g/dL.
Percent of
Interior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homes
Falling the
Standard
(# Homes
Failing/
# Homes)
97% (35/36)
94% (34/36)
89% (32/36)
89% (32/36)
89% (32/36)
83% (30/36)
72% (26/36)
64% (23/36)
58% (21/36)
Blood-Lead Standard of 10 //g/dL
Sensitivity
(b/a+b)
94% (15/1 6)
94% (15/1 6)
88% (14/1 6)
88% (14/1 6)
88% 1 14/1 6)
88% (14/1 6)
75% (12/1 6)
63% (10/16)
56% (9/1 6)
Specificity
(c/c+d)
0% (0/20)
5% (1/20)
10% (2/20)
10% (2/20)
10% (2/20)
20% (4/20)
30% (6/20)
35% (7/20)
40% (8/20)
Positive
Predictive
Value
(b/b+d)
43% (15/35)
44% (15/34)
44% (14/32)
44% (14/32)
44% (14/32)
47% (14/30)
46% (12/26)
44% (10/23)
43% (9/21)
Negative
Predictive
Value
(c/a+c)a
0%(0/1)
50% (1/2)
50% (2/4)
50% (2/4)
50% (2/4)
67% (4/6)
60% (6/10)
54% (7/1 3)
53% (8/1 5)
#
Homes
36
36
36
36
36
36
36
36
36
Note, there was 1 home where the percent of interior components with deteriorated LBP was zero. Of the 1 home, 1
home had a child with a blood-lead concentration above 10 figldl.
Draft - Do Not Cite or Quote
C-37
September 19, 1996
-------
I
\
I
o
5'
o
O OS-
S „,
•«N _ _
20 30
50 80
90 IOO
9
OJ
00
Tru. Po.itiv* (
IB 30
Percent of Components
8
O)
Figure C-19a. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Percent of
Exterior Components Within a Home with Deteriorated LBP, Based on Data from the Baltimore Repair and
Maintenance Study and a Blood-Lead Standard of 10 /ig/dL.
-------
Table C-19a. Performance Characteristics of Potential Paint-Lead Standards for the
Percent of Exterior Components within a Home with Deteriorated Lead-
Based Paint Based on Data from the Baltimore Repair and Maintenance
Study and a Blood-Lead Standard of 10/ig/dL.
PorcBRt of
Exterior
Componsnts
with
DotBrior&tsd
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homes
Failing the
Standard
<# Homes
Failing/
# Homes)
77% (27/35)
77% (27/35)
77% (27/35)
77% (27/35)
77% (27/35)
77% (27/35)
77% (27/35)
74% (26/35)
74% (26/35)
Blood-Lead Standard of 10 pg/dL
Sensitivity
(b/a+b)
94% (15/1 6)
94% (15/1 6)
94% (15/16)
94% (15/1 6)
94% (15/1 6)
94% (15/1 6)
94% (15/1 6)
88% (14/16)
88% (14/16)
Specificity
(c/c+d)
37% (7/19)
37% (7/1 9)
37% (7/1 9)
37% (7/19)
37% (7/1 9)
37% (7/1 9)
37% (7/1 9)
37% (7/1 9)
37% (7/1 9)
Positive
B-^-ll—Al,.-
rraoicnvo
Value
(b/b+d)
56% (15/27)
56% (15/27)
56% (15/27)
56% (15/27)
56% (15/27)
56% (15/27)
56% (15/27)
54% (14/26)
54% (14/26)
Negative
Predictive
Value
(c/a+c)8
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
78% (7/9)
78% (7/9)
*
Homes
35
35
35
35
35
35
35
35
35
Note, there were 8 homes where the percent of exterior components with detenorated LBP was zero. Of the 8 homes. 1
home had a child with a blood-lead concentration above 10 pg/dL
Draft - Do Not Cite or Quote
C-39
September 19, 1996
-------
o
60 70 BO
7
Percent of Co
n
±
o
I
-^
to
£. 03 H
Percent of Co
' (l-Sp.clllclty)
Percen
Perform
Figure C 20a. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Maximum of
the Percent of Interior Components with Deteriorated LBP and the Percent of Exterior Components with
Deteriorated LBP, Based on Data from the Baltimore Repair and Maintenance Study and a Blood-Lead
Standard of 10/yg/dL.
-------
Table C-20a. Performance Characteristics of Potential Paint-Lead Standards for the
Maximum of the Percent of Interior Components with Deteriorated LBP and
the Percent of Exterior Components with Deteriorated LBP Based on Data
from the Baltimore Repair and Maintenance Study and a Blood-Lead
Standard of 10//g/dL.
Maximum of
PfifCQflt Of
Interior/Exterior
ComponBVtts
with
Deteriorated LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homes
Failing the
Standard
(ff Homes
Failing/
# Homes)
97% (34 /35 )
97% (34 /35 )
97% (34 /35 )
97% (34 /35 )
97% (34 /35 )
94% (33 /35 )
91% (32/35)
91% (32/35)
91% (32/35)
Blood-Lead Standard of 10 //g/dL
SBnsitivity
(b/a+b)
94% (15/16)
94% (15/16)
94% (15/16)
94% (15/16)
94% (15/16)
94% (15/16)
94% (15/16)
94% (15/16)
94% (15/16)
Specificity
(c/c+d)
0%(0/19)
0% (0/19 I
0%(0/19)
0%(0/19)
0%(0/19)
5%(1 /19)
11% (2/19)
11% (2/19)
11% (2/19)
Positive
Predictive
Value
(b/b+d)
44% (15/34)
44% (15/341
44% (15 /34 I
44% (15/34)
44% (15/34)
46% (15 /33 )
47% (15 132 )
47% (15/32)
47% (15/32)
Negative
Predictive
Value
(c/a+c)a
0% (0 /1 )
0% (0 /I )
0% (0 /1 )
0% (0 /I )
0% (0 /I )
50% (1 12 )
67% (2 13 )
67% (2 13 )
67% (2 /3 )
#
Homes
35
35
35
35
35
35
35
35
35
Note, there was 1 home where the maximum of the percent of interior components with deteriorated LBP and the percent
of exterior components with detenorated LBP was zero. Of the 1 home, 1 home had a child with a blood-lead
concentration above 10 //g/dL.
Draft - Do Not Cite or Quote
C-41
September 19, 1996
-------
1
!
•i
0
o
rl-*- "
\OQ 1 .000 1O OOO
Dripline Soil-Lead Concentration (^.g g)
O 6
I °7
0 6
U O b -
BOO I 600 Z 400 3 2OO 4 C
Di-iplnitr Soil-Lead C
4.BOO f> 600 6 4OO 7 2OO 8.C
ration (t-g g )
n
i.
ro
If
1 61)0 2 400
Dripli
BOO 6 400 -?,300 B OOO
8)
p Poiitivr [ I - Specificity)
BOO 1 6CO 2 -tOO 3 ZOO 4 OOO 4 BOO 5 BOO 6 4OO 7 20O B OOO
erformonce Ch«ro< lenatic ~ '•••"» * Spec * !-i - • '. J • •
CO
(o
01
Figure (Mb. Summary Graphs of the Performance Characteristics of Potential Dripline of Soil-Lead Standards Based on
Data From the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 15 x/q/dL.
-------
Table C-1b. Performance Characteristics of Potential Dripline Soil-Lead Standards Based on
Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of
15/ig/dL.
Dripline
Soil-Lead
Standard
0/9/9)
25
50
75
100
125
150
175
200
250
300
350
400
450
500
600
700
800
900
1.000
1,500
2.000
2.500
3.000
3,500
4,000
4,500
5,000
8,000
10,000
% Homos
Failing the
Standard
(# Homes Failing/
# Homes)
99% (184/186)
95% (177/1 86)
93% (172/186)
91% (169/186)
88% (163/186)
86% (159/1 86)
83% (155/1 86)
82% (153/186)
81% (150/1 86)
80% (148/1 86)
79% (146/186)
76% (142/186)
74% (137/1 86)
72% (133/1 86)
69% (128/1 86)
60% (11 2/1 86)
57% (105/1 86)
52% (97/186)
47% (88/1 86)
29% (53/186)
18% (34/186)
13% 124/186)
9% (17/1 86)
7% (12/186)
7% (12/186)
5% (10/186)
4% (7/1 86)
1% (2/1 86)
1% (2/186)
Blood-Lead Standard of 15 pg/dL
Sensitivity
(b/a + b)
100% (15/1 5)
100% (15/1 5)
100% ( 15/1 5)
100% (15/15)
93% (14/1 5)
93% (14/1 5)
93% (14/1 5)
87% (13/1 5)
87% (13/1 5)
87% 1 13/1 5)
87% (13/15)
87% ( 13/1 5)
87% (13/1 5)
87% (13/1 5)
87% (13/1 5)
87% (13/1 5)
87% (13/1 5)
87% (13/1 5)
80% (12/1 5)
47% (7/15)
33% (5/1 51
27% (4/1 5)
13% (2/1 5)
13% (2/1 5)
13% (2/1 5)
13% (2/1 5)
13% (2/1 5)
0%(0/15)
0% (0/1 5)
Specificity
(c/c+d)
1% (2/171)
5% (9/1 71)
8% (14/171)
10% (17/171)
13% (22/1 71 )
15% (26/1 71)
18% (30/1 71)
18% (31/171)
20% (34/1 71)
21% (36/1 71 1
22% (38/171)
25% 142/1 71)
28% (47/1 71)
30% 151/171)
33% (56/1 71)
42% 172/1 71)
46% (79/1 71 1
51% (87/171)
56% (95/1 71)
73% (125/171)
83% (142/171)
88% (151/171)
91% (156/1 71)
94% (161/171)
94% (161/171)
95% (163/171)
97% (166/1 71)
99% (169/1 71)
99% (169/1 71)
Positive
Predictive
Value
(b/b+d)
8% (15/1 84)
9% (15/1 77)
9% (15/1 72)
9% (15/169)
9% (14/163)
9% (14/1 59)
9% (14/1 55)
9% (13/1 53)
9% (13/1 50)
9% 1 13/1 48)
9% (13/146)
9% (13/1 42)
10% (13/1 37)
10% (13/1 33)
10% (13/1 28)
12% (13/1 12)
12% (13/105)
13% (13/97)
14% (12/88)
13% (7/53)
1 5% (5/34)
17% (4/24)
12% (2/1 7)
17% (2/1 2)
17% (2/1 2)
20% (2/10)
29% (2/7)
0% (0/2)
0% (0/2)
Negative
Predictive
Value
(c/a+c)
100% (2/2)
100% (9/9)
100% (14/1 4)
100% (17/1 7)
96% (22/23)
96% (26/27)
97% (30/31)
94% (31/33)
94% (34/36)
95% (36/38)
95% (38/40)
96% (42/44)
96% (47/49)
96% (51/53)
97% (56/58)
97% (72/74)
98% (79/81)
98% (87/89)
97% (95/98)
94% (125/1 33)
93% (142/1 52)
93% (151/1 62)
92% (156/1 69)
93% (161/174)
93% (161/174)
93% (163/1 76)
93% (166/1 79)
92% (169/1 84)
92% (169/1 84)
#
Homes
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
186
Draft - Do Not Cite or Quote
C-43
September 19. 1996
-------
§•
£
I
100 1 .OOO
Play -Area Soil-Lead Concent ration (i.g g)
U 0 6 -
§
~*
I
—*
«o
—*
10
0 0 - -. —' r-f -, r
eoo \ aoo \ son z ooo a 400
Play Area Soil-Lead Concentration (ug g)
aoo i aoo i BOO 2 ooo 2 400 2 aoo
Play-Area Soil -Lead Concentration (,.g g)
Figure C-2b. Summary Graphs of the Performance Characteristics of Potential Play Area of Soil-Lead Standards Based on
Data From the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 15 /sg/dL.
-------
Table C-2b. Performance Characteristics of Potential Play-Area Soil-Lead Standards Based
on Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of
15//g/dL.
Play-Area
Son-Lead
Standard
(//g/g)
25
50
75
100
125
150
175
200
250
300
350
400
450
500
600
700
800
900
1,000
1,500
2,000
2.500
4.000
% Homes
Failing the
Standard
(# Homes Failing/
# Homes)
100% (87/87)
94% (82/87)
89% (77/87)
83% (72/87)
76% (66/87)
70% (61/87)
66% (57/87)
61% (53/87)
55% (48/87)
49% (43/87)
41% (36/87)
36% (31/87)
26% (23/87)
23% (20/87)
21% (18/87)
17% (15/87)
14% (12/87)
9% (8/87)
8% (7/87)
5% (4/87)
2% (2/87)
1% (1/87)
1% (1/87)
Blood-Lead Standard of 15 jrg/dL
Sensitivity
-------
o
§
\O \OO 1 OOO 10.OOO I OO.OOO
Uncat peterl Floor Dust Lead Loading (..g sq rt)
O
100 150 200 250
Unrarpeted Floor Dust - Lead Loading (*
-------
Table C-3b. Performance Characteristics of Potential Dust-Lead Standards for Uncarpeted
Floors Based on Data from the Rochester Lead-in-Dust Study and a Blood-
Lead Standard of 15 //g/dL.
Uncarpeted
Floor Dust-
Lead
Standard
Uig/ft2)
1
2
3
4
5
10
15
20
25
50
75
100
125
150
175
200
300
400
% Homes
Faffing the
Standard
(# Homes Faffing/
# Homes)
100% (196/1 97)
98% (193/1 97)
94% (185/197)
93% (184/197)
91% 1180/197)
76% (150/1 97)
59% (116/197)
43% (84/1 97)
29% (58/197)
10% (19/197)
6% (11/197)
5% (9/197)
3% (6/197)
3% (6/197)
3% (6/197)
3% (5/1 97)
2% (4/197)
2% (3/197)
Blood-Lead Standard of 15 //g/dL
Sensitivity
(b/a+b)
100% (16/1 6)
100% 1 16/1 6)
100% (16/1 6)
100% 1 16/1 6)
94% (15/1 6)
94% (15/1 6)
88% (14/1 6)
88% (14/1 6)
75% (12/1 6)
38% (6/16)
31% (5/1 6)
25% (4/1 6)
19% (3/1 6)
19% (3/1 6)
19% (3/1 6)
13% (2/1 6)
13% (2/1 6)
13% (2/1 6)
Specificity
(c/c+d)
1% (1/1811
2% (4/1 81 1
7% (12/181)
7% (13/181)
9% (16/181)
25% (46/1 81)
44% (79/1 81)
61% (111/181)
75% (135/181)
93% (168/1 81)
97% (175/181)
97% (176/181)
98% (178/1 81)
98% (178/1 81)
98% (178/1 81)
98% (178/181)
99% (179/1 81)
99% (180/181 1
a— .. tat.-..
positnra
Predicts vo
Value
(b/b+d)
8% (16/196)
8% (16/193)
9% (16/185)
9% (16/184)
8% (15/1 80)
10% (15/1 50)
12% (14/1 16)
17% (14/84)
21% (12/58)
32% (6/1 9)
46% (5/11)
44% (4/9)
50% (3/6)
50% (3/6)
50% (3/6)
40% (2/5)
50% (2/4)
67% (2/3)
Negative
Predictive
Value
(c/a+c)
100% (1/1)
100% (4/4)
100% 1 12/1 2)
100% 113/13)
94% (16/1 7)
98% (46/47)
98% (79/81)
98% (111/113)
97% (135/1 39)
94% ( 168/1 78)
94% (175/1 86)
94% ( 176/1 88)
93% (178/1 91)
93% (178/191)
93% (178/1 91)
93% ( 178/1 92)
93% (179/1 93)
93% 1 180/1 94)
tt
Homos
197
197
197
197
197
197
197
197
197
197
197
197
197
197
197
197
197
197
Draft - Do Not Cite or Quota
C-47
September 19. 1996
-------
X 20
IO 1 OO 1 .OOO 10 OOO
Carpeted Floor Dust-Lead Loading (Mg, sq ft}
Si "• -
Carpeted Floor Dust-Lead Loading (jig sq 11}
o
^
CX3
• Illvr (Senvltlvity)
. PpV . NPV
50
Figure C-4b. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Carpeted Floors
Based on Data From the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 15 /ig/dL.
-------
Table C-4b. Performance Characteristics of Potential Dust-Lead Standards for Carpeted
Floors Based on Data from the Rochester Lead-in-Dust Study and a Blood-
Lead Standard of 15 //g/dL.
Carpeted
Floor Dust-
Lead
Standard
(pg/ft2)
1
2
3
4
5
10
15
20
25
50
75
100
125
150
175
200
300
400
% Homes
Failing the
Standard
(# Homes Falling/
# Homes)
98% (176/1 79)
93% (167/179)
91% (163/1 79)
89% (159/179)
87% (155/179)
67% (11 9/1 79)
40% (71/179)
23% (41/179)
16% (29/179)
5% (8/1 79)
2% (4/179)
2% (3/1 79)
2% (3/1 79)
1% (2/1 79)
1% (2/1 79)
1% (2/1 79)
1% (2/1 79)
1% (2/179)
Blood-Lead Standard of 16 00/dL
Sensitivity
(b/a+b)
100% (11/11)
100% (11/11)
100% (11/11)
100% (11/11)
100% (11/11)
91% (10/11)
82% (9/11 )
55% (6/11)
36% (4/11)
9% (1/11)
0% (0/1 11
0% (0/11 )
0%(0/11)
0%(0/11)
0%(0/11)
0% (0/11 1
0%(0/11)
0%(0/11)
Specificity
(c/c+d)
2% (3/1 68)
7% (12/1 68)
10% (16/1 68)
12% (20/168)
14% (24/168)
35% (59/168)
63% (106/168)
79% (133/1 68)
85% (143/168)
96% (161/168)
98% (164/168)
98% (165/1 68)
98% (165/1 68)
99% (166/1 68)
99% (166/1 68)
99% (166/1 68)
99% (166/1 68)
99% (166/1 68)
Positive
Predictive
Value
(b/b+d)
6% (11/176)
7% (11/167)
7% (11/163)
7% (11/159)
7% (11/1 55)
8% (10/1 19)
13% (9/71 1
15% (6/41)
14% (4/29)
13% (1/8)
0% (0/4)
0% (0/3)
0% (0/3)
0% (0/2)
0% (0/2)
0% (0/2)
0% (0/2)
0% (0/2)
Negative
Predictive
Value
(c/a+c)
100% (3/3)
100% (12/12)
100% 1 16/1 6)
100% (20/20)
100% (24/24)
98% (59/60)
98% (106/108)
96% (133/1 38)
95% 1 143/1 50)
94% (161/171)
94% (164/1 75)
94% (165/1 76)
94% (165/1 76)
94% (166/1 77)
94% ( 166/1 77)
94% (166/1 77)
94% (166/1 77)
94% (166/177)
#
Homes
179
179
179
179
179
179
179
179
179
179
179
179
179
179
179
179
179
179
Draft - Do Not Cite or Quote
C-49
September 19, 1996
-------
I
r
10 100 i.ooo
Combined Floor Dust- Lend Loading (ug sq ft)
Combined Floor Dust - Lend Loading lug sq tt)
9
§
I
-*
50
*-*
to
Telie i •).- -- (1 f - ' '
g
s .
Figure C-5b. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Combined Floors
Based on Data From the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 15 //g/dL.
-------
Table C-5b. Performance Characteristics of Potential Dust-Lead Standards for Combined
Carpeted-Uncarpeted Floors Based on Data from the Rochester Lead-in-Dust
Study and a Blood-Lead Standard of 15 //g/dL.
Combined
Floor
Dust-Lead
Standard
brg/g)
1
2
3
4
5
10
15
20
25
50
75
100
125
150
175
200
300
400
% Homes
Failing the
Standard
(# Homes Failing/
ff Homes)
100% (205/205)
98% (201/205)
96% (196/205)
94% (192/205)
91% (186/205)
76% (156/205)
53% (109/205)
36% (74/205)
27% (56/205)
9% (19/205)
6% 1 12/205)
5% (10/205)
4% (8/205)
4% (8/205)
4% (8/205)
3% (7/205)
2% (5/205)
2% (5/205)
Blood-Lead Standard of 1S//g/dL
Sensitivity
(b/a+b)
100% (16/16)
100% (16/16)
100% (16/16)
100% (16/16)
100% (16/16)
94% (15/1 6)
94% (15/1 6)
94% (15/16)
81% (13/16)
44% (7/1 6)
31% (5/16)
25% (4/16)
19% (3/1 6)
19% (3/1 6)
19% (3/1 6)
13% (2/1 6)
13% (2/1 6)
13% (2/16)
Specificity
(c/c+d)
0% (0/1 89)
2% (4/189)
5% (9/189)
7% (13/1891
10% (19/1 89)
25% (48/1 89)
50% (95/189)
69% (130/189)
77% (146/1 89)
94% (177/189)
96% (182/189)
97% (183/189)
97% (184/189)
97% (184/189)
97% (184/189)
97% (184/189)
98% (186/189)
98% (186/189)
Positive
Predicttvo
Value
(b/b+d)
8% (16/205)
8% (16/201 1
8% (16/196)
8% (16/1 92)
9% (16/186)
10% (15/1 56)
14% (15/109)
20% (15/74)
23% (13/56)
37% (7/19)
42% (5/1 2)
40% (4/10)
38% (3/8)
38% (3/8)
38% (3/8)
29% (2/7)
40% (2/5)
40% (2/5)
Negative
Predictive
Value
(c/a+c)
.% (0/0)
100% (4/4)
100% (9/9)
100% (13/1 3)
100% (19/19)
98% (48/49)
99% (95/96)
99% (130/1 31)
98% (146/149)
95% (177/1 86)
94% (182/1 93)
94% (183/1 95)
93% (184/197)
93% (184/1 97)
93% (184/1 97)
93% (184/198)
93% (186/200)
93% (186/200)
#
Homes
205
205
205
205
205
205
205
205
205
205
205
205
205
205
205
205
205
205
Draft - Do Not Cite or Quote
C-51
September 19. 1996
-------
O
t?
100 i.ooo 10.000
Window Sill Dust-Lead Loading (MB sq ft)
10 I SOD 2 OOO 2 SOO 3 OOC
Window Sill Dust-Lead Loading (..% sq ft)
O
tn
I 000 I SOO 2 OOQ Z iOO 3 OOO 3 SOO 4 OOO
1 OOO I 500 2 OOO ? 3OO
Wmdow Sill Dust-Lead Loading
-------
Table C-6b. Performance Characteristics of Potential Dust-Lead Standards for Window
Sills Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 15 /ig/dL.
Window
Sill Dust-
Lead
Standard
U/g/ft2)
5
10
15
20
25
50
75
100
125
150
175
200
250
300
350
400
450
500
600
700
800
900
1000
1500
2000
2500
3000
3500
4000
% Homes
Falling the
Standard
{# Homes Failing/
9 Homes)
98% (192/1 96)
98% (192/1 96)
98% (192/1 96)
97% (190/1 96)
96% (188/196)
88% (172/1 96)
77% (150/1 96)
67% (131/1 96)
62% (121/1 96)
57% (11 1/1 96)
51% (100/196)
47% (92/1 96)
37% (73/196)
35% (68/1 96)
28% (55/1 96)
26% (50/196)
24% (46/1 96)
21% (42/1 96)
18% (36/1 96)
15% (30/1 96)
12% (24/1 96)
12% (24/196)
12% (23/1 96)
8% (16/196)
6% (12/1 96)
6% (11/196)
4% (8/196)
3% (6/196)
2% (4/196)
Blood-Lead Standard of 15 /ig/dL
Sensitivity
(b/a+b)
100% (16/1 6)
100% 1 16/1 6)
100% (16/1 6)
100% (16/16)
100% 1 16/1 6)
100% 1 16/1 6)
100% (16/1 6)
100% (16/1 6)
100% (16/1 6)
100% (16/16)
100% (16/1 6)
100% 1 16/1 6)
88% (14/1 6)
81%U3M6)
75% (12/1 6)
75% (12/1 6)
69% (11/1 6)
63% (10/1 6)
56% (9/1 6)
50% (8/1 6)
44% (7/16)
44% (7/1 6)
44% (7/1 6)
38% (6/16)
31% (5/1 6)
25% (4/1 6)
13% (2/1 6)
13% (2/1 6)
6% (1/16)
Specificity
(c/c+d)
2% (4/1 80)
2% (4/180)
2% (4/180)
3% (6/1 80)
4% (8/1 80)
13% (24/1 80)
26% (46/180)
36% (65/180)
42% (75/180)
47% (85/180)
53% (96/180)
58% (104/180)
67% (121/1 80)
69% (125/1 80)
76% 1 137/1 80)
79% (142/1 80)
81% (145/1 80)
82% (148/180)
85% (153/1 80)
88% (158/1 80)
91% (163/1 80)
91% (163/1 80)
91% (164/1 80)
94% (170/1 80)
96% (173/1 80)
96% (173/1 80)
97% (174/1 80)
98% (176/1 80)
98% (177/180)
Positive
PfBdlCtfVft
Value
(b/b+d)
8% (16/1 92)
8% (16/1 92)
8% (16/192)
8% (16/1 90)
9% 1 16/1 88)
9% (16/1 72)
11% (16/150)
12% (16/1 31)
13% (16/121)
14% (16/1 11)
16% (16/1 00)
17% (16/92)
19% (14/73)
19% (13/68)
22% (12/55)
24% ( 12/50)
24% (11/46)
24% (10/42)
25% (9/36)
27% (8/30)
29% (7/24)
29% (7/24)
30% (7/23)
38% (6/1 6)
42% (5/1 2)
36% (4/11)
25% (2/8)
33% (2/6)
25% (1/4)
Negative
Prodictivo
Value
(c/a+c)
100% (4/4)
100% (4/4)
100% (4/4)
100% (6/6)
100% (8/8)
100% (24/24)
100% (46/46)
100% (65/65)
100% (75/75)
100% (85/85)
100% (96/96)
100% (104/104)
98% (121/1 23)
98% (125/128)
97% (137/141)
97% (142/146)
97% (145/1 50)
96% (148/1 54)
96% (153/1 60)
95% (158/1 66)
95% (163/1 72)
95% (163/1 72)
95% (164/1 73)
94% (170/1 80)
94% (173/184)
94% (173/1 85)
93% (174/1 88)
93% 11 76/1 90)
92% (177/1 92)
tt
Homes
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
196
Draft - Do Not Cite or Quote
C-53
September 19. 1996
-------
I
o
O
o
« •*•- •
10O.OOO 1 OOO OOO
a ooo i zooo
Window Trough Dust - Lead Loading (ug sq fl)
O
co
8 OOO 12 OOO
Wmdow Trough Dust Lead Loading <<-* sq ft)
Jo
Figure C-7b. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Window Troughs
<| Based on Data From the Rochester Lead-in-Dust Study and a Blood-Lead Standard of 15 /ig/dL.
-------
Table C-7b. Performance Characteristics of Potential Dust-Lead Standards for Window
Troughs Based on Data from the Rochester Lead-in-Dust Study and a Blood-
Lead Standard of 15//g/dL.
Window
Trough
Dust-Lead
Standard
big/ft2)
50
100
200
300
400
500
600
700
800
900
1000
1500
2000
2500
3000
3500
4000
4500
5000
10000
15000
20000
30000
40000
50000
1 00000
% Homes
Failing the
Standard
(# Homes
Failing/
ft Homes)
98% (185/1 89)
94% (178/189)
89% (168/1 89)
84% (159/189)
83% (156/1 89)
80% (151/1 89)
78% (148/1 89)
76% (144/189)
73% (138/189)
73% (137/189)
70S6 (133/1 89)
64% (121/1 89)
62% (118/189)
61% (115/189)
59% (111/189)
56% (106/189)
55% (103/189)
51% (96/1 89)
48% (91/1 89)
40% (75/189)
34% (65/189)
30% (57/189)
23% (44/1 89)
20% (38/1 89)
16% (31/1 89)
10% (18/189)
Blood-Lead Standard of 15 pg/dL
Sensitivity
(b/a+b)
100% (15/1 5)
100% 1 15/1 5)
100% (15/1 5)
100% (15/15)
100% (15/1 5)
100% 1 15/1 5)
93% (14/1 5)
93% (14/15)
93% (14/15)
93% (14/1 5)
93% (14/15)
93% (14/1 5)
93% (14/15)
93% (14/1 5)
93% (14/15)
93% (14/1 5)
93% (14/15)
93% (14/1 5)
93% (14/15)
87% (13/15)
73% (11 /1 5)
67% (10/15)
60% (9/1 5)
60% (9/1 5)
53% (8/1 5)
40% (6/1 5)
Specificity
(c/c+d)
2% (4/1 74)
6% (11 /1 74)
12% (2 1/1 74)
17% (30/174)
19% (33/1 74)
22% (38/174)
23% (40/174)
25% (44/174)
29% (50/174)
29% (51/174)
32% (55/174)
39% (67/174)
40% (70/174)
42% (73/174)
44% (77/174)
47% (82/174)
49% (85/174)
53% (92/174)
56% (97/174)
64% (11 2/1 74)
69% (120/1 74)
73% (127/1 74)
80% (139/1 74)
83% (145/1 74)
87% (151/174)
93% (162/174)
Positive
Predictive
Value
(b/b+d)
8% (15/1 85)
8% (15/1 78)
9% (15/1 68)
9% (15/1 59)
10% (15/1 56)
10% (15/151)
10% (14/148)
10% (14/144)
10% 1 14/1 38)
10% |1 4/1 37)
11% (14/1 33)
12% (14/121)
12% (14/1 18)
12% (14/115)
13% (14/1 11)
13% (14/1 06)
14% (14/103)
15% (14/96)
15% (14/91)
17% (13/75)
17% (11/65)
18% (10/57)
21% (9/44)
24% (9/38)
26% (8/31)
33% (6/18)
Negative
Predictive
Value
(c/a+c)
100% (4/4)
100% (11/11)
100% (21/21)
100% (30/30)
100% (33/33)
100% (38/38)
98% (40/41)
98% (44/45)
98% (50/51)
98% (51/52)
98% (55/56)
99% (67/68)
99% (70/71)
99% (73/74)
99% (77/78)
99% (82/83)
99% (85/86)
99% (92/93)
99% (97/98)
98% (11 2/1 14)
97% (120/1 24)
96% (127/1 32)
96% (139/1 45)
96% (145/151)
96% (151/1 58)
95% (162/171)
#
Homes
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
189
Draft - Do Not Cite or Quote
C-55
September 19, 1996
-------
=t
I
o
o
I
o
c?
80 90
«-d hilenor I.BP
o
en
en
Figure C-8b. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Percent of
Interior Components Within a Home with Deteriorated LBP, Based on Data From the Rochester Lead-in-Dust
Study and a Blood-Lead Standard of 15 //g/dL.
-------
Table C-8b. Performance Characteristics of Potential Paint-Lead Standards for the Percent
of Interior Components within a Home with Deteriorated Lead-Based Paint
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 15 //g/dL.
Percent of
Interior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homes
Failing the
Standard
(* Homes
Failing/
# Homes)
64% (132/205)
64% (132/205)
64% (132/205)
64% (132/205)
63% 1 129/205)
59% (121/205)
50% (103/205)
46% (94/205)
43% (89/205)
Blood-Lead Standard of 16 //g/dL
Sensitivity
(b/a+b)
100% (16/16)
100% (16/16)
100% (16/16)
100% 1 16/1 6)
100% 1 16/1 6)
100% (16/1 6)
100% (16/1 6)
94% (15/1 6)
94% (15/1 6)
Specificity
(c/c+d)
39% (73/1 89)
39% (73/1 89)
39% (73/189)
39% (73/189)
40% (76/189)
44% (84/189)
54% (102/1 89)
58% (110/1 89)
61% (115/189)
Positive
PrediCuVB
Value
(b/b+d)
12% 1 16/1 32)
12% (16/1 32)
12% (16/132)
12% (16/132)
12% (16/1291
13% (16/121)
16% (16/1 03)
16% (15/94)
17% (15/89)
Negative
Predictive
Value
(c/e + c)a
100% (73/73)
100% (73/73)
100% (73/73)
100% (73/73)
100% (76/76)
100% (84/84)
100% (102/102)
99% (110/1 11)
99% (11 5/1 16)
#
Homos
205
205
205
205
205
205
205
205
205
8 Note, there were 73 homes where the percent of interior components with deteriorated LBP was zero. Of the 73 homes,
no homes had a child with a blood-lead concentration above 15 //g/dL.
Draft - Do Not Cite or Quote
C-57
September 19, 1996
-------
I
C3
S i
10 20 10
5O 80 7O 80 BO 100
IS 14
16 ia
00
'Tin. !• .,ll,r (S*nllllvlly>
Percent of To
to
0>
Figure C-9b. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Percent of
Exterior Components Within a Home with Deteriorated LBP, Based on Data From the Rochester Lead-in-Dust
Study and a Blood-Lead Standard of 15 //g/dL.
-------
Table C-9b. Performance Characteristics of Potential Paint-Lead Standards for the Percent
of Exterior Components within a Home with Deteriorated Lead-Based Paint
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 15 //g/dL.
Percent of
Exterior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homes
Failing the
Standard
(# Homes
Failing/
# Homes)
47% (96/204)
47% (96/204)
47% (96/204)
47% (96/204)
47% (96/204)
47% (96/204)
47% (95/204)
47% (95/204)
47% (95/204)
Blood-Lead Standard of 15 pg/dL
Sensitivity
(b/a+b)
69% (11/1 6)
69% (11/1 6)
69% (11/1 6)
69% (11/1 6)
69% (11/16)
69% (11/16)
69% (11/16)
69% (11/1 6)
69% (11/1 6)
Specificity
(c/c+d)
55% (103/1 88)
55% (103/1 88)
55% (103/1 88)
55% (103/1 88)
55% (103/1 88)
55% (103/188)
55% (104/188)
55% (104/1 88)
55% (104/1 88)
Positive
Predictive
Value
(b/b+d)
12% (11/96)
12% (11/96)
12% (11/96)
12% (11/96)
12% (11/96)
12% (11/96)
12% (11/95)
12% (11/95)
12% (11/95)
Negative
Predictive
Value
(c/a+c)a
95% (103/108)
95% (103/108)
95% (103/108)
95% (103/108)
95% (103/108)
95% (103/108)
95% (104/109)
95% (104/109)
95% (104/109)
#
Homes
204
204
204
204
204
204
204
204
204
Note, there were 108 homes where the percent of exterior components with deteriorated LBP was zero. Of the 108
homes, 5 homes had a child with a blood-lead concentration above 15 //g/dL.
Draft - Do Not Cite or Quote
C-59
September 19, 1996
-------
I
" i -* -"-- !
ZO 30 40 SO
eo 00
Percent of Couipo
o
o>
o
if
Sg
ill of Componen
Percent of Components
Figure C-10b. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Maximum of
the Percent of Interior Components with Deteriorated LBP and the Percent of Exterior Components with
Deteriorated LBP, Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of
15 //g/dL.
-------
Table C-10b. Performance Characteristics of Potential Paint-Lead Standards for the
Maximum of the Percent of Interior Components with Deteriorated LBP and
the Percent of Exterior Components with Deteriorated LBP Based on Data
from the Rochester Lead-in-Dust Study and a Blood-Lead Standard of
15/ig/dL.
Maximum of the
Percent of
Interior/Exterior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homes
Failing the
Standard
(# Homes
Failing/
# Homes)
74% (150/204)
74% (150/204)
74% (150/204)
74% (150/204)
73% (149/204)
71% (145/204)
66% (135/204)
64% (131/204)
63% (129/204)
Blood-Lead Standard of 16 /ig/dL
Sensitivity
(b/a+b)
100% (16/1 6)
100% (16/1 6)
100% (16/16)
100% (16/1 6)
100% (16/1 6)
100% 1 16/1 6)
100% (16/1 6)
94% (15/1 6)
94% (15/1 6)
Specificity
(c/c+d)
29% (54/1 88)
29% (54/1 88)
29% (54/1 88)
29% (54/188)
29% (55/1 88)
31% (59/1 88)
37% (69/1 88)
38% (72/188)
39% (74/1 88)
Positivs
Predictive
Value
(b/b+d)
11% (16/1 50)
11% (16/1 50)
11% (16/150)
11% (16/1 50)
11% (16/149)
11% (16/1 45)
12% (16/1 35)
12% (15/131)
12% (15/1 29)
Negative
Predictive
Value
(c/a+c)"
100% (54/54)
100% (54/54)
100% (54/54)
100% (54/54)
100% (55/55)
100% (59/59)
100% (69/69)
99% (72/73)
99% (74/75)
»
Homos
204
204
204
204
204
204
204
204
204
Note, there were 54 homes where the maximum of the percent of interior components with deteriorated LBP and the
percent of exterior components with LBP was zero. Of the 54 homes, no homes had a child with a blood-lead
concentration above 15 //g/dL.
Draft - Do Not Cite or Quote
C-61
September 19, 1996
-------
I
\
s
s
I
ii 1 -
1 r
o
F>lle Pontlve
-------
Table C-11b. Performance Characteristics of Potential Paint-Lead Standards for the
Average Percent of Deteriorated Lead-Based Paint per Interior Component
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 15/ig/dL.
Average
Percent of
Deteriorated
LBPper
Interior
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20 0%
% Homes
Faffing the
Standard
(ff Homes Faffing/
# Homes)
79% (162/205)
53% (108/205)
28% (58/205)
10% (21/205)
3% (7/205)
0% (0/205)
0% (0/205)
0% (0/205)
0% (0/205)
Blood-Lead Standard of 15 /ig/dL
Sensitivity
(b/a+b)
100% (16/16)
94% (15/1 6)
69% (11/16)
38% (6/1 6)
19% (3/16)
0%(0/16)
0%(0/16)
0%(0/16)
0% (0/16)
Specificity
(c/c+d)
23% (43/189)
51% (96/1 89)
75% (142/189)
92% (174/189)
98% 1 185/1 89)
100% (189/189)
100% 1 189/1 89)
100% (189/1 89)
100% (189/189)
Positive
Predictive
Value
(b/b+d)
10% (16/1 62)
14% (15/1 08)
19% (11/58)
29% (6/21)
43% (3/7)
.% (0/0)
.% (0/0)
.% (0/0)
.% (0/0)
Negative
Predictive
Value
«c/a+c)a
100% (43/43)
99% (96/97)
97% (142/147)
95% (174/184)
93% (185/1 98)
92% (189/205)
92% (189/205)
92% (189/205)
92% (189/205)
8
Homes
205
205
205
205
205
205
205
205
205
a Note, there were 98 homes where the average percent of deteriorated LBP per interior component was zero. Of the 98
homes, 1 home had a child with a blood-lead concentration above
Draft - Do Not Cite or Quote
C-63
September 19, 1996
-------
n
'
o
§
:: S
I
I 0 -i
E OJ
I
O O +-r ^ r--, n
18 20
Figure C-12b. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Average
Percent of Deteriorated LBP Per Exterior Component, Based on Data From the Rochester Lead-in-Dust
Study and a Blood-Lead Standard of 15 //g/dL.
-------
Table C-12b. Performance Characteristics of Potential Paint-Lead Standards for the
Average Percent of Deteriorated Lead-Based Paint per Exterior Component
Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 15 //g/dL.
Average
Percent of
Deteriorated
LBP per
Exterior
Component
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
Homes
Failing the
Stsnd&rd
(# Homes
Failing/
# Homes)
79% (161/204)
52% (106/204)
31 % (64/204)
19% (38/204)
12% (25/204)
5% (10/204)
2% (3/204)
1% (1/204)
0% (0/204)
Blood-Lead Standard of 15 //g/dL
Sensitivity
(b/a+b)
81% (13/16)
69% 11 1/1 6)
50% (8/16)
44% (7/1 6)
38% (6/16)
13% (2/1 6)
13% (2/1 6)
0%(0/16)
0% (0/16)
Specificity
(e/c+d)
21% (40/1 88)
50% (93/188)
70% (132/1 88)
84% (157/1 88)
90% (169/1 88)
96% (180/1 88)
100% (187/188)
100% (187/1 88)
100% (188/188)
PositivB
Predictive
Value
(b/b+d)
8% (13/161)
10% (11/106)
13% (8/64)
18% (7/38)
24% (6/25)
20% (2/10)
67% (2/3)
0%(0/1)
.% (0/0)
Negative
Predictive
Value
(c/a+c)a
93% (40/43)
95% (93/98)
94% (132/140)
95% (157/1 66)
94% (169/179)
93% 1 180/1 94)
93% 1 187/201)
92% (187/203)
92% (188/204)
# Homes
Included En
Analysis
204
204
204
204
204
204
204
204
204
8 Note, there were 43 homes where the average percent of deteriorated LBP per exterior component was zero. Of the 43
homes, 3 homes had a child with a blood-lead concentration above 15 //g/dL.
Draft - Do Not Cite or Quote
C-65
September 19, 1996
-------
3"
o
O
§
I a I I ! ! 8 i '
J_S f *^2_
^: 07-
OS-
ID 12
9
S
I
5*n> - Sprc • PPV « NPV
I
Figure C-13b. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Maximum of
the Average Percent of Deteriorated LBP per Interior Component and the Average Percent of Deteriorated
LBP per Exterior Component, Based on Data from the Rochester Lead-in-Dust Study and a Blood-Lead
Standard of 15//g/dL.
-------
Table C-13b. Performance Characteristics of Potential Paint-Lead Standards for the
Maximum of the Average Percent of Deteriorated LBP per Interior
Component and the Average Percent of Deteriorated LBP per Exterior
Component Based on Data from the Rochester Lead-in-Dust Study and a
Blood-Lead Standard of 15 //g/dL.
Maximum of
Average
Percent of
Interior/Exterior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20 0%
% Homes
Failing the
Standard
(# Homes
Failing/
# Homes)
88% (179/204)
71% (144/204)
48% (97/204)
25% (50/204)
13% (27/204)
5% (10/204)
2% 13/204)
1% (1/204)
0% (0/204)
Blood-Lead Standard of 15 //g/dL
Sensitivity
(b/a+b)
100% 1 16/1 6)
94% (15/1 6)
88% (14/1 6)
69% (11 /1 6)
44% (7/16)
13% (2/1 6)
13% (2/1 6)
0%(0/16)
0% (0/16)
Specificity
(c/c+d)
13% (25/1 88)
31% (59/1 88)
56% (105/188)
79% (149/1 88)
89% (168/1 88)
96% 1 180/1 88)
100% (187/1881
100% (187/1 88)
100% (188/188)
Positive
PfBdictivo
Value
(b/b+d)
9% (16/1 79)
10% (15/1 44)
14% (14/97)
22% (11 /SO)
26% (7/27)
20% (2/10)
67% (2/3)
0%(0/1)
.% (0/0)
Negative
ProdictivB
Value
(c/a+c)a
100% (25/25)
98% (59/60)
98% (105/107)
97% (149/1 54)
95% (168/1 77)
93% (180/194)
93% (187/201)
92% (187/203)
92% (188/204)
#
Homes
204
204
204
204
204
204
204
204
204
Note, there were 25 homes where the maximum of the average percent of deteriorated LBP per interior component and the
averege percent of deteriorated LBP per exterior component was zero. Of the 25 homes, no homes had a child with a
blood-lead concentration above 15 //g/dL.
Draft - Do Not Cite or Quote
C-67
September 19, 1996
-------
§
100 I .OOO
Driplmr Soil - Lead Cone en I roll on (ug g)
-100 BOO I 200 1 eOO Z OOO 2 4OO 2 BOO 3.2OO 3 BOO 4 OOO
Dripline Soil-Lead Concentrotion O-g g}
9
O)
00
400 eoo i zoo i soo a.ooo 2.400 2 soo 3 zoo 3«oo 4 ooo
Dripline Soil-Lead Concentration (ug g)
3 BOO 4 000
Driphne Soil-Lead Concentration («g g)
Figure C-14b. Summary Graphs of the Performance Characteristics of Potential Dripline Soil-Lead Standards Based on
Data from the Baltimore Repair and Maintenance Study and a Blood-Lead Standard of 15 //g/dL.
-------
Table C-14b. Performance Characteristics of Potential Dripline Soil-Lead Standards Based
on Data from the Baltimore Repair and Maintenance Study and a Blood-Lead
Standard of 15j/g/dL.
Dripline
Soil-Lead
Standard
(fffj/g)
25
50
75
100
125
150
175
200
300
400
500
600
700
800
900
1.000
1.500
2,000
2.500
3.000
3.500
4.000
% Homes
Failing the
Standard
(# Homes Failing/
# Homes)
100% (27/27)
85% (23/27)
74% (20/27)
70% (19/27)
67% (18/27)
67% (18/27)
63% (17/27)
63% (17/27)
63% (17/27)
59% (16/27)
56% (15/27)
56% (15/27)
48% (13/27)
44% 1 12/27)
44% ( 12/27)
44% (12/27)
41% (11/27)
7% (2/27)
7% (2/27)
7% (2/27)
4% (1/27)
0% (0/27)
Blood-Lead Standard of 15 pg/dL
Sensitivity
(b/a+b)
100% (4/4)
100% (4/4)
100% (4/4)
100% (4/4)
100% (4/4)
100% (4/4)
100% (4/4)
100% (4/4)
100% (4/4)
75% (3/4)
75% (3/4)
75% (3/4)
50% (2/4)
50% (2/4)
50% (2/4)
50% (2/4)
50% (2/4)
0% (0/4)
0% (0/4)
0% (0/4)
0% (0/4)
0% (0/4)
Sp6cificity
(c/c+d)
0% (0/23)
17% (4/23)
30% (7/23)
35% (8/23)
39% (9/23)
39% (9/23)
44% (10/23)
44% (10/23)
44% (10/23)
44% (10/23)
48% (11/23)
48% (11/23)
52% (12/23)
57% (13/23)
57% (13/23)
57% (13/23)
61% (14/23)
91% (21/23)
91% (21/23)
91% (21/23)
96% (22/23)
100% (23/23)
Positive
Predictive
Value
(b/b+d)
15% (4/27)
17% (4/23)
20% (4/20)
21% (4/19)
22% (4/18)
22% (4/18)
24% (4/17)
24% (4/17)
24% (4/17)
19% (3/1 6)
20% (3/15)
20% (3/1 5)
15% (2/1 3)
17% (2/1 2)
17% (2/1 2)
17% (2/1 2)
18% (2/11)
0% (0/2)
0% (0/2)
0% (0/2)
0%(0/1)
.% (0/0)
Negative
Predictive
Value
(c/a+c)
.% (0/0)
100% (4/4)
100% 17/7)
100% (8/8)
100% (9/9)
100% (9/9)
100% (10/10)
100% (10/10)
100% (10/10)
91% (10/11)
92% (11/12)
92% (11/1 2)
86% (12/14)
87% (13/1 5)
87% (13/1 5)
87% (13/1 5)
88% (14/1 6)
84% (21/25)
84% (2 1/25)
84% (21/25)
85% (22/26)
85% (23/27)
«
Homes
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
Draft - Do Not Cite or Quote
C-69
September 19, 1996
-------
s
o
IOO
aq ft)
p£!"""p
N««atlv« I
9
^j
o
I
—*
<°
^-*
<0
Figure C-15b. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Combined
Carpeted-Uncarpeted Floors, Based on Data from the Baltimore Repair and Maintenance Study and a Blood-
Lead Standard of 15 x/g/dL.
-------
Table C-15b. Performance Characteristics of Potential Dust-Lead Standards for Floors
Based on Data from the Baltimore Repair and Maintenance Study and a
Blood-Lead Standard of 15 //g/dL
Floor*
Dust-Lead
Standard
bvg/ft2)
5
10
15
20
25
50
75
100
125
150
175
% Homes
Failing the
Standard
(8 Homes
Failing/
# Homes)
98% (85/87)
91 % (79/87)
83% (72/87)
78% (68/87)
74% (64/87)
25% (22/87)
7% (6/87)
2% (2/87)
2% (2/87)
1%(1/87)
0% (0/87)
Blood-Lead Standard of 15 j/g/dL
Sensitivity
(b/a+b)
100% 1 19/1 9)
100% (19/1 9)
100% (19/1 9)
100% (19/19)
100% (19/1 9)
37% (7/19)
21% (4/1 9)
11% (2/19)
11% (2/19)
5% (1/1 9)
0%(0/19)
Specificity
(c/c+d)
3% (2/68)
12% (8/68)
22% (15/68)
28% (19/68)
34% (23/68)
78% (53/68)
97% (66/68)
100% (68/68)
100% (68/68)
100% (68/68)
100% (68/68)
Positive
Predictive
Value
(b/b+d)
22% (19/85)
24% (19/79)
26% (19/72)
28% (19/68)
30% (19/64)
32% (7/22)
67% (4/6)
100% (2/2)
100% (2/2)
100% (1/1)
.% (0/0)
Negative
Predictive
Value
(c/a+c)
100% (2/2)
100% (8/8)
100% (15/1 5)
100% (19/1 9)
100% (23/23)
82% (53/65)
82% (66/81)
80% (68/85)
80% (68/85)
79% (68/86)
78% (68/87)
»
Homes
87
87
87
87
87
87
87
87
87
87
87
Includes floor samples taken on the floor at the interior entryway.
Draft - Do Not Cite or Quote
C-71
September 19, 1996
-------
I
\
I
10O 1.000
Window Sill Dust-Lead Loading
-------
Table C-16b. Performance Characteristics of Potential Dust-Lead Standards for Window
Sills Based on Data from the Baltimore Repair and Maintenance Study and a
Blood-Lead Standard of 15 //g/dL.
Window Sill
Dust-Lead
Standard
big/ft2)
25
50
75
100
125
150
175
200
300
400
500
1,000
1.500
2 000
% Homes
Failing the
Standard
(tf Hornos
Failing/
# Homes)
98% (85/87)
91% (79/87)
82% (71/87)
75% (65/87)
74% (64/87)
72% (63/87)
71% (62/87)
70% (61/87)
66% (57/87)
56% (49/87)
51% (44/87)
13% (11/87)
6% (5/87)
0% (0/87)
Blood-Lead Standard of 15 //g/dL
Sensitivity
(b/a+b)
100% (19/1 9)
100% (19/19)
100% (19/1 9)
90% (17/1 9)
84% (16/1 9)
84% (16/1 9)
84% (16/19)
84% (16/1 9)
79% (15/1 9)
74% (14/1 9)
68% (13/1 9)
11% (2/1 9)
0% (0/19)
0% (0/1 9)
Specificity
(c/c+d)
3% (2/68)
12% (8/68)
24% (16/68)
29% (20/68)
29% (20/68)
31% (21/68)
32% (22/68)
34% (23/68)
38% (26/68)
49% (33/68)
54% (37/68)
87% (59/68)
93% (63/68)
100% (68/68)
Positive
Predictive
Value
(b/b+d)
22% (19/85)
24% (19/79)
27% (19/71)
26% 1 17/65)
25% (16/64)
25% (16/63)
26% (16/62)
26% (16/61 1
26% (15/57)
29% (14/49)
30% (13/44)
18% (2/11)
0% (0/5)
.% (0/0)
Negative
Predictive
Value
(c/a+c)
100% (2/2)
100% (8/8)
100% (16/1 6)
91% (20/22)
87% (20/23)
88% (21/24)
88% (22/25)
89% (23/26)
87% (26/30)
87% (33/38)
86% (37/43)
78% (59/76)
77% (63/82)
78% 168/87)
*
Homes
87
87
87
87
87
87
87
87
87
87
87
87
87
87
Draft - Do Not Cite or Quote
C-73
September 19, 1996
-------
o
£
»
I
CD
o
|
O
3?'
0
(f
_
V —•
.' Til ilM-
o
to
to
01
Figure C-17b. Summary Graphs of the Performance Characteristics of Potential Dust-Lead Standards for Window
Troughs, Based on Data from the Baltimore Repair and Maintenance Study and a Blood-Lead Standard of
15 pg/dL.
-------
Table C-17b. Performance Characteristics of Potential Dust-Lead Standards for Window
Troughs Based on Data from the Baltimore Repair and Maintenance Study
and a Blood-Lead Standard of 15 /ig/dL.
Window
Trough
Dust-Lead
Standard
u/g/ft*)
200
300
400
500
600
700
800
900
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
5,000
10,000
1 5,000
20,000
30,000
40,000
50,000
60.000
% Homes
Failing the
Standard
(9 Homes
Failing/
# Homes)
97% (83/86)
88% (76/86)
87% (75/86)
86% (74/86)
85% (73/86)
84% (72/86)
81% (70/86)
78% (67/86)
77% (66/86)
71% (61/86)
70% (60/86)
69% (59/86)
67% (58/86)
65% (56/86)
65% (56/86)
63% (54/86)
63% (54/86)
54% (46/86)
38% (33/86)
29% (25/86)
12% (10/86)
5% (4/86)
1% (1/86)
1% (1/86)
Blood-Lead Standard of 15 /ig/dL
Sensitivity
(b/a+b)
100% (19/1 9)
95% (18/1 9)
95% (18/1 9)
95% (18/19)
95% (18/1 9)
95% (18/1 9)
95% (18/19)
95% (18/19)
95% (18/1 9)
90% ( 17/1 9)
90% (17/1 9)
90% (17/1 9)
84% (16/1 9)
79% (15/1 9)
79% (15/1 9)
74% (14/19)
74% (14/19)
63% 1 12/1 9)
47% (9/19)
37% (7/19)
21% (4/1 9)
11% (2/1 9)
5% (1/19)
5% (1/19)
Specificity
(c/c+d)
5% (3/67)
13% (9/67)
15% (10/67)
16% (11/67)
18% (12/67)
19% (13/67)
22% (15/67)
27% (18/67)
28% 1 19/67)
34% (23/67)
36% (24/67)
37% (25/67)
37% (25/67)
39% (26/67)
39% (26/67)
40% (27/67)
40% (27/67)
49% (33/67)
64% (43/67)
73% (49/67)
91% (61/67)
97% (65/67)
100% (67/67)
100% (67/67)
Positive
Predictive
Value
(b/b+d)
23% (19/83)
24% (18/76)
24% (18/75)
24% ( 18/74)
25% (18/73)
25% (18/72)
26% (18/70)
27% (18/67)
27% (18/66)
28% (17/61)
28% (17/60)
29% (17/59)
28% (16/58)
27% (15/56)
27% (15/56)
26% (14/54)
26% (14/54)
26% (12/46)
27% (9/33)
28% (7/25)
40% (4/10)
50% (2/4)
100% (1/1)
100% (1/1)
Negative
Predictive
Value
(c/a+c)
100% (3/3)
90% (9/10)
91% (10/11)
92% (11/1 2)
92% (12/1 3)
93% (13/1 4)
94% (15/1 6)
95% (18/1 9)
95% (19/20)
92% (23/25)
92% (24/26)
93% (25/27)
89% (25/28)
87% (26/30)
87% (26/30)
84% (27/32)
84% (27/32)
83% (33/40)
81% (43/53)
80% (49/61)
80% (61/76)
79% (65/82)
79% (67/85)
79% (67/85)
#
Homes
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
86
Draft - Do Not Cite or Quote
C-75
September 19. 1996
-------
I
D
55'
o
O
§
50 60
e-d Interior LBP
n
-Li
o>
IB 20
5 2
Figure C-18b. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Percent of
Interior Components Within a Home with Deteriorated LBP, Based on Data from the Baltimore Repair and
Maintenance Study and a Blood-Lead Standard of 15/sg/dL.
-------
Table C-18b. Performance Characteristics of Potential Paint-Lead Standards for the
Percent of Interior Components within a Home with Deteriorated Lead-Based
Paint Based on Data from the Baltimore Repair and Maintenance Study and a
Blood-Lead Standard of 15 jig/dL.
Percent of
Interior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20 0%
% Homes
Foiling the
Standard
(tt Homes
Failing/
# Homes)
97% (35/36)
94% (34/36)
89% (32/36)
89% (32/36)
89% (32/36)
83% (30/36)
72% (26/36)
64% (23/36)
58% (21/36)
Blood-Lead Standard of 15 //g/dL
Sensitivity
(b/a + b)
100% (7/7)
100% (7/7)
86% (6/7)
86% (6/7)
86% (6/7)
86% (6/7)
86% (6/7)
86% (6/7)
71% (5/7)
Specificity
(c/c+d)
3% (1/29)
7% (2/29)
10% (3/29)
10% (3/29)
10% (3/29)
17% (5/29)
31% (9/29)
41% (12/29)
45% (13/29)
Positive
Predictive
Value
(b/b+d)
20% (7/35)
21% (7/34)
19% (6/32)
19% (6/32)
19% (6/32)
20% (6/30)
23% (6/26)
26% (6/23)
24% (5/21)
Negative
Predictive
Value
(c/a+c)'
100% (1/1)
100% (2/2)
75% (3/4)
75% (3/4)
75% (3/4)
83% (5/6)
90% (9/10)
92% (12/1 3)
87% (13/1 5)
#
Homes
36
36
36
36
36
36
36
36
36
Note, there were 2 homes where the percent of interior components with deteriorated LBP was zero. Of the 2 homes, no
homes had a child with a blood-lead concentration above 15 //g/dL.
Draft - Do Not Cite or Quote
C-77
September 19, 1996
-------
I
§
I
jf 20
i
Percent of Compo
Percent of Coin pa
•-J
oo
IB 2O
CO
<0
Figure C-19b. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Percent of
Exterior Components Within a Home with Deteriorated LBP, Based on Data from the Baltimore Repair and
Maintenance Study and a Blood-Lead Standard of 15 //g/dL.
-------
Table C-19b. Performance Characteristics of Potential Paint-Lead Standards for the
Percent of Exterior Components within a Home with Deteriorated Lead-
Based Paint Based on Data from the Baltimore Repair and Maintenance
Study and a Blood-Lead Standard of 15 //g/dL.
Percent of
Exterior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
150%
17.5%
200%
% Homes
Failing the
Standard
(# Homes
Failing/
it Homes)
77% (27/35)
77% (27/35)
77% (27/35)
77% (27/35)
77% (27/35)
77% (27/35)
77% (27/35)
74% (26/35)
74% (26/35)
Blood-Lead Standard of 15 //g/dL
Sensitivity
(b/a+b)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
Specificity
(c/c+d)
29% (8/28)
29% (8/28)
29% (8/28)
29% (8/28)
29% (8/28)
29% (8/28)
29% (8/28)
32% (9/28)
32% (9/28)
PositivB
Predictive
Value
(b/b+d)
26% (7/27)
26% (7/27)
26% (7/27)
26% (7/27)
26% (7/27)
26% (7/27)
26% (7/27)
27% (7/26)
27% (7/26)
Negative
Predictive
Value
(c/e+c)"
100% (8/8)
100% (8/8)
100% (8/8)
100% (8/8)
100% (8/8)
100% (8/8)
100% (8/8)
100% (9/9)
100% (9/9)
*
Homes
35
35
35
35
35
35
35
35
35
Note, there were 8 homes where the percent of exterior components with deteriorated LBP was zero. Of the 8 homes, no
homes had a child with a blood-lead concentration above 15 //g/dL.
Draft - Do Not Cite or Quote
C-79
September 19, 1996
-------
o
o
o
-I
O
o
30 40 SO 80 70 BO
CD
O
t
Cr
CD
5"
•-*
1
Figure C 20b. Summary Graphs of the Performance Characteristics of Potential Paint-Lead Standards for the Maximum of
the Percent of Interior Component with Deteriorated LBP and the Percent of Exterior Components with
Deteriorated LBP, Based on Data from the Baltimore Repair and Maintenance Study and a Blood-Lead
Standard of 15pg/dL.
-------
Table C-20b. Performance Characteristics of Potential Paint-Lead Standards for the
Maximum of the Percent of Interior Components with Deteriorated LBP and
the Percent of Exterior Components with Deteriorated LBP Based on Data
from the Baltimore Repair and Maintenance Study and a Blood-Lead
Standard of 15//g/dL.
Maximum of
Percent of
Interior/Exterior
Components
with
Deteriorated
LBP
1.0%
2.5%
5.0%
7.5%
10.0%
12.5%
15.0%
17.5%
20.0%
% Homes
Failing the
Standard
(9 Homes
Failing/
# Homes)
97% (34/35)
97% (34/35)
97% (34/35)
97% (34/35)
97% (34/35)
94% (33/35)
91% (32/35)
91% (32/35)
91% (32/35)
Blood-Lead Standard of 15 /ig/dL
Sensitivity
(b/a+b)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
100% (7/7)
Specificity
(c/c+d)
4% (1/28)
4% (1/28)
4% (1/28)
4% (1/28)
4% 1 1/28)
7% (2/28)
11% (3/28)
11% (3/28)
1 1 % (3/28)
Positive
Predictive
Value
(b/b+d)
21% (7/34)
21% (7/34)
21% (7/34)
21% (7/34)
21% (7/34)
21% (7/33)
22% (7/32)
22% (7/32)
22% (7/32)
Negative
Predictive
Value
(cta+c)a
100% (1/1)
100% (1/1)
100% (1/1)
100% (1/1)
100% (1/1)
100% (2/2)
100% (3/3)
100% (3/3)
100% (3/3)
»
Homes
35
35
35
35
35
35
35
35
35
8 Note, there were 1 home where the maximum of the percent of interior components with deteriorated LBP and the percent
of exterior components with deteriorated LBP was zero. Of the 1 home, no homes had a child with a blood-lead
concentration above 15 //g/dL.
Draft - Do Not Cite or Quote
C-81
September 19, 1996
-------
Appendix D
Tables for the Two-Way Multi-Media Analysis Results
Draft - Do NOT Cite or Quote D-1 September 19, 1996
-------
Table D-1. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead and Dripline Soil-Lead Based on Data From the Rochester Lead-in-Dust
Study Where the Negative Predictive Value 2 90%.
Uncarpeted
Floor Dust-
Lead
Standard
Uig/ft2)
50
75
100
125
Dripfine
Soil-Lead
Standard
(/ug/g)
300
400
500
600
700
800
900
1000
300
400
500
600
700
900
1000
300
400
500
600
700
900
1000
200
300
400
500
600
700
900
1000
% Dwellings
Failing the
Standsrd
81% (147/180)
78% (141/181)
74% (134/1 81)
71% (129/181)
64% (11 5/1 81)
60% (108/181)
56% (102/181)
52% (94/181)
81% (147/181)
78% (141/181)
73% (132/181)
70% (127/1 81)
61% (111/181)
54% (98/1 81)
50% (90/181)
81%(147/181)
78% (141/181)
73% (132/181)
70% (127/181)
61% (111/181)
54% (98/1 81)
49% (89/181)
83%(151/181)
81% (146/181)
77% (140/181)
72% (131/181)
70% (126/1 81)
61% (110/181)
54% (97/181)
49% (88/1 81)
Blood-Load Standard of 10 pg/dL
Sensitivity
(b/a+b)
93% (41/44)
91% (40/44)
91% (40/44)
89% (39/44)
86% (38/44)
84% (37/44)
84% (37/44)
84% (37/44)
93% (4 1/44)
91% (40/44)
91% (40/44)
89% (39/44)
84% (37/44)
82% (36/44)
82% (36/44)
93% (41/44)
91% (40/44)
91% (40/44)
89% (39/44)
84% (37/44)
82% (36/44)
80% (35/44)
93% (41/44)
93% (41/44)
91% (40/44)
91% (40/44)
89% (39/44)
84% (37/44)
82% (36/44)
80% (35/44)
Specificity
(c/c+d)
23% (31/1 37)
26% (36/1 37)
31% (43/1 37]
34% (47/1 37)
44% (60/1 37)
48% (66/1 37)
53% (72/1 37)
58% (80/1 37)
23% (31 /1 37)
26% (36/1 37)
33% (45/1 37)
36% (49/1 37)
46% (63/1 37)
55% (75/1 37)
61% (83/137)
23% (31/1 37)
26% (36/1 37)
33% (45/1 37)
36% (49/1 37)
46% (63/1 37)
55% (75/1 37)
61% (83/137)
20% (27/1 37)
23% (32/1 37)
27% (37/1 37)
34% (46/1 37)
37% (50/1 37)
47% (64/1 37)
56% (76/1 37)
61% (84/1 37)
Positive
Predictive
Value
(b/b+d)
28% (41/1 47)
28% (40/1 41 1
30% (40/1 34)
30% (39/1 29)
33% (38/1 15)
34% (37/108)
36% (37/102)
39% (37/94)
28% (41/147)
28% (40/1 41)
30% (40/1 32)
31% (39/1 27)
33% (37/1 1 1 )
37% (36/98)
40% (36/90)
28% (41/1 47)
28% (40/1 41)
30% (40/1 32)
31% (39/1 27)
33% (37/1 11)
37% (36/98)
39% (35/89)
27% (41/1 51)
28% (41 /1 46)
29% (40/140)
31% (40/1 31)
31% (39/1 26)
34% (37/1 10)
37% (36/97)
40% (35/88)
Negative
Predictive
Value
(c/a+c)
91% (31/34)
90% (36/40)
92% (43/47)
90% (47/52)
91% (60/66)
90% (66/73)
91% (72/79)
92% (80/87)
91% (31/34)
90% (36/40)
92% (45/49)
91% (49/54)
90% (63/70)
90% (75/83)
91% (83/91)
91% (31/34)
90% (36/40)
92% (45/49)
91% (49/54)
90% (63/70)
90% (75/83)
90% (83/92)
90% (27/30)
91% (32/35)
90% (37/41 1
92% (46/50)
91% (50/55)
90% (64/71)
91% (76/84)
90% (84/93)
»
Homes
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
Draft - Do Not Cite or Quote
D-2
September 19, 1996
-------
Table D-1. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead and Dripline Soil-Lead Based on Data From the Rochester Lead-in-Dust
Study Where the Negative Predictive Value * 90% (Continued).
Uncarpeted
Floor Dust-
Lead
Standard
(//g/ft2)
ISO
175
200
Dripline
Soil-Lead
Standard
(l/g/g)
200
300
400
500
600
700
900
1000
200
300
400
500
600
700
900
1000
500
% Dwellings
Failing the
Standard
83% (151/181)
81% (146/181)
77% (140/181)
72% (131/181)
70% (127/1 81)
61% (110/181)
54% (97/1 81)
49% (88/1 81 1
83% (151/181)
81% (146/181)
77% (140/181)
72% (131/181)
70% (126/1 81)
61% (110/181)
54% (97/181 )
49% (88/1 81)
72% (130/181)
Blood-Lead Standard of 10//g/dL
Sensitivity
(bta+b)
93% (41/44)
93% (41/44)
91% (40/44)
91 % (40/44)
89% (39/44)
84% (37/44)
82% (36/44)
80% (35/44)
93% (41/44)
93% (41/44)
91% (40/44)
91% (40/44)
89% (39/44)
84% (37/44)
82% (36/44)
80% (35/44)
89% (39/44)
Specificity
(c/c+d)
20% (27/1 37)
23% (32/1 37)
27% (37/1 37)
34% (46/1 37)
37% (50/1 37)
47% (64/1 37)
56% (76/1 37)
61% (84/1 37)
20% (27/1 37)
23% (32/1 37)
27% (37/137)
34% (46/1 37)
37% (50/1 37)
47% (64/1 37)
56% (76/1 37)
61% (84/1 37)
34% (46/1 37)
Positive
Predictive
Value
(b/b+d)
27% (41/151)
28% (41/146)
29% (40/140)
31% (40/1 31)
31% (39/1 26)
34% (37/1 10)
37% (36/97)
40% (35/88)
27% (41/1 51)
28% (41/146)
29% (40/140)
31% (40/1 31)
31% (39/1 26)
34% (37/1 101
37% (36/97)
40% (35/88)
30% (39/1 30)
Negative
Predictive
Value
(c/a+c)
90% (27/30)
91% (32/35)
90% (37/41)
92% (46/50)
91% (50/55)
90% (64/71)
91% (76/84)
90% (84/93)
90% (27/30)
91% (32/35)
90% (37/41)
92% (46/50)
91% (50/55)
90% (64/71)
91% (76/84)
90% (84/93)
90% (46/51)
ff
Homes
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
Draft - Do Not Cite or Quote
D-3
September 19, 1996
-------
Table D-2. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead and Play Area Soil-Lead Based on Data From the Rochester Lead-in-Dust
Study Where the Negative Predictive Value ;> 90%.
Uncarpeted
Floor Dust-
Lead
Stondora
U/g/ft2)
50
75
100
125
150
175
200
Play Area
Soil-Lead
Standard
U/8/9)
200
300
200
200
200
200
200
200
% Dwellings
Failing the
Standard
64% (54/84)
52% (44/84)
61% (51/84)
61% (51/84)
61% (51/84)
61% (51/84)
61% (51/84)
61% (51/84)
Blood-Lead Standard of 10 /ig/dL
Sensitivity
(b/a+b)
88% 1 14/1 6)
75% (12/1 6)
81% (13/1 6)
81% (13/1 6)
81% (13/1 6)
81% (13/1 6)
81% (13/1 6)
81% (13/1 6)
Specificity
(c/c+d)
41% (28/68)
53% (36/68)
44% (30/68)
44% (30/68)
44% (30/68)
44% (30/68)
44% (30/68)
44% (30/68)
Positive
Predictive Value
(b/b+d)
26% (14/54)
27% (12/44)
26% (13/51)
26% (13/51)
26% (13/51)
26% (13/51)
26% (13/51)
26% (13/51)
Negative
Predictive
Value
(c/a+c)
93% (28/30)
90% (36/40)
91% (30/33)
91% (30/33)
91% (30/33)
91% (30/33)
91% (30/33)
91% (30/33)
»
Homes
84
84
84
84
84
84
84
84
Draft - Do Not Cite or Quote
D-4
September 19, 1996
-------
Table D-3. Performance Characteristics of Potential Standards for Floor Dust-Lead and
Dripline Soil-Lead Based on Data From the Baltimore Repair and Maintenance
Study Where the Negative Predictive Value * 90%.
Floor Dust-
Lead
Standard
(l/g/ft*)
50
75
100
125
150
175
Dripline
Soil-Lead
Standard
l/ifl/fl)
200
300
200
300
200
300
200
300
200
300
200
300
% Dwellings
Failing the
Standard
63% (17/27)
63% 1 17/27)
63% (17/27)
63% (17/27)
63% (17/27)
63% (17/27)
63% (17/27)
63% (17/27)
63% (17/27)
63% (17/27)
63% (17/27)
63% (17/27)
Blood-Lead Standard of 10 j/g/dL
Sensitivity
(b/a+b)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
88% (7/8)
Specificity
(c/c+d)
47% (9/19)
47% (9/19)
47% (9/19)
47% (9/19)
47% (9/19)
47% (9/19)
47% (9/1 9)
47% (9/19)
47% (9/1 9)
47% (9/19)
47% (9/19)
47% (9/1 9)
Positivo
Pradictiv0
Value
(b/b+d)
41% (7/17)
41% (7/1 7)
41% (7/1 7)
41% (7/17)
41% (7/1 7)
41% (7/1 7)
41% (7/1 7)
41% (7/1 7)
41% (7/1 7)
41% (7/1 7)
41% (7/1 7)
41% (7/1 7)
Negative
Prodlctivo
Value
(c/a+c)
90% (9/10)
90% (9/10)
90% (9/10)
90% (9/10)
90% (9/10)
90% (9/10)
90% (9/10)
90% (9/10)
90% (9/10)
90% (9/10)
90% (9/10)
90% (9/10)
tt
Homes
27
27
27
27
27
27
27
27
27
27
27
27
Draft - Do Not Cite or Quote
D-5
September 19, 1996
-------
Table D-4. Summary of the Two-Way Multi-Media Analysis of the Uncarpeted Floor Dust-
Lead Loadings (/ig/ft2) and Dripline Soil-Lead Concentrations (//g/g) From the
Rochester Lead-in-Dust Study When Negative Predictive Value (NPV) is Greater
Than or Equal to 90%.
Blood-Lead Standard
of 10/ig/dL
Negative Predictive
Value
2 90%'
Summary Measure
Maximum Uncarpeted
Floor Dust-Lead Standard
Maximum Dripline Soil-
Lead Standard
Current Uncarpeted Floor
Dust-Lead Guidance
Level
Current Dripline Soil-Lead
Guidance Level
Uncarpeted Floor
Dust-Lead Standard
(//g/ft2)
200
2 50 and s 175
100
2 50 and s 175
Dripline
Soil-Lead Standard
(P9/9>
500
1,000
2 300 and
s 1,000
400
B No negative predictive value (NPV) was calculated to be greater than or equal to 95%.
Table D-5. Summary of the Two-Way Multi-Media Analysis of the Uncarpeted Floor Dust-
Lead Loadings (//g/ft2) and Play Area Soil-Lead Concentrations (//g/g) From the
Rochester Lead-in-Dust Study When Negative Predictive Value (NPV) is Greater
Than or Equal to 90%.
Blood-Lead
Standard of 10
//g/dL
Negative
Predictive Value
2 90%'
Summary Measure
Maximum Uncarpeted Floor Dust-Lead Standard
Maximum Play Area Soil-Lead Standard
Current Uncarpeted Floor Dust-Lead Guidance
Level
Current Play Area Soil-Lead Guidance Level
Uncarpeted
Floor Dust-
Lead
Standard
-------
Table D-6. Summary of the Two-Way Multi-Media Analysis of the Floor Dust-Lead
Loadings (//g/ft2) and Dripline Soil-Lead Concentrations U/g/g) From the
Baltimore Repair and Maintenance Study When Negative Predictive Value
(NPV) is Greater Than or Equal to 90%.
Blood-Lead
Standard of 10
/ig/dL
Negative
Predictive Value
* 90% a
Summary Measure
Maximum Floor Dust-Lead Standard
Maximum Dripline Soil-Lead Standard
Current Floor Dust-Lead Guidance Level
Current Dripline Soil-Lead Guidance Level
Floor Dust-
Lead
Standard
(/ig/ft2)
175
* 50 and
& 175
100
b
Dripline
Soil-Lead
Standard
(fjglg)
*. 200 and
& 300
300
* 200 and
* 300
400
8 No negative predictive value (NPV) was calculated to be greater than or equal to 95%.
b No NPV was calculated to be greater than or equal to 90%.
Draft - Do Not Cite or Quote
D-7
September 19, 1996
-------
9
Table D-7. Summary of the Two-Way Multi-Media Standards for Uncarpeted Floor Dust-Lead, and Dripline and Play
Area Soil-Lead for the Rochester Lead-in-Dust Study and Floor Dust-Lead and Dripline Soil-Lead for the
Baltimore Repair and Maintenance Study for Which a Negative Predictive Value of 95% or Greater was
Achieved with Floor Dust-Lead at 25 //g/f t2.
Study
Rochester Lead-in-
Dust
Baltimore Repair and
Maintenance
Uncarpeted Floor
Dust-Lead (/ig/ft2)
25
25
Soil-Lead
Location
Dripline
Play Area
Standard
(//g/g)
500
200
Blood-Lead Standard of 10 /ig/dL
Sensitivity
(b/a + bl
96% (42/44)
94% (15/16)
Negative Predictive Value
(c/a + c)
96% (42/44)
96% (26/27)
a
9
CO
No NPV of 95% or greater was attained.
-------
Table D-8. Potential Standards for Uncarpeted Floor Dust-Lead and Dripline Soil-Lead From the Rochester Lead-in-Dust
Study, Maximizing the Unweighted Summary Measure.
Uncarpeted
Floor Dust-Lead
Standard
(tfB/ft*)
1 50
75
2125 and s175
100
200
2125 and $175
50
75
200
50
100
2125 and s175
50
275 and si 00
2125 and si 75
Dripline
Soil-Lead
Standard
(ffl/d)
1,000
1.000
1.000
1,000
1.000
900
900
900
1.500
900
1,500
1.500
800
1 500
% Homes
Failing the
Combined
Standard
52% (94/1 81)
50% (90/1 811
49% (88/1 81)
49% (89/1 81)
48% (87/1 81)
54% (97/1 81)
56% (102/181)
54% (98/1 81)
32% (57/1 81)
53% (96/1 81)
34% (62/1 81)
31% (56/181)
57% (103/181)
60% 1109/181)
58% (105/181)
30% (54/181)
Unweighted
Summary
Measure
(USM)'
274
274
271
270
267
265
264
264
262
261
260
258
258
257
256
256
Blood-Lead Standard of 10//g/dL
Sensitivity
84% (37/44)
82% (36/44)
80% (35/44)
80% (35/44)
77% (34/44)
82% (36/44)
84% (37/44)
82% (36/44)
57% (25/44)
80% (35/44)
59% (26/44)
55% (24/44)
82% (36/44)
84% (37/44)
82% (36/44)
52% (23/44)
Specificity
58% (80/1 37)
61% (83/1 37)
61% (84/1 37)
61% (83/1 37)
61% (84/1 37)
56% (76/137)
53% (72/1 37)
55% (75/137)
77% (105/1 37)
56% (76/137)
74% (101/1 37)
77% (105/1 37)
51% (70/137)
48% (66/137)
50% (69/1 37)
77% 1 106/1 37)
Positive
Predictive
Velue
39% (37/94)
40% (36/90)
40% (35/88)
39% (35/89)
39% (34/87)
37% (36/97)
36% (37/102)
37% (36/98)
44% (25/57)
37% (35/96)
42% (26/62)
43% (24/56)
35% (36/103)
34% (37/108)
35% (36/104)
43% (23/54)
Negative
Predictive
Value
92% (80/87)
91% (83/91)
90% (84/93)
90% (83/92)
89% (84/94)
91% (76/84)
91% (72/79)
90% (75/83)
85% (105/124)
89% (76/85)
85% (101/119)
84% (105/125)
90% (70/78)
90% (66/73)
90% (69/77)
84% (106/127)
# Homes
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
181
9
o
9
to
I
I
••*
5°
1 Unweighted Summary Measure = Sensitivity + Specificity + PPV + NPV
-------
Table D-9. Potential Standards for Uncarpeted Floor Dust-Lead and Play Area Soil-Lead From the Rochester Lead-in-Dust
Study, Maximizing the Unweighted Summary Measure.
Uncarpeted
Floor Dust-Lead
Standard
(pg/ft2)
2100 and i200
75
75
50
50
50
i75 and s200
Play Area
Soil-Lead
Standard (fig/fl)
900
900
1.000
1.000
800
900
800
1.000
600
% Homes
Failing the
Combined
Standard
10% (8/84)
11% (9/84)
8% (7/84)
10% (8/84)
13% (11/84)
17% (14/84)
18% (15/84)
16% (13/84)
20% (17/84)
Unweighted
Summary
Measure
(USM)'
296
287
286
275
271
271
266
259
257
Blood-Lead Standard of 10 j/a/dL
Sensitivity
38% (6/1 6)
38% (6/1 6)
31% 15/1 6)
31% (5/16)
38% (6/1 6)
44% (7/16)
44% (7/16)
38% (6/1 6)
44% (7/16)
Specificity
97% (66/68)
96% (65/68)
97% (66/68)
96% (65/68)
93% (63/68)
90% (61/68)
88% (60/68)
90% (61/68)
85% (58/68)
Positive
Predictive
Value
75% (6/8)
67% (6/9)
71% (5/7)
63% (5/8)
55% (6/11)
50% (7/1 4)
47% (7/1 5)
46% (6/1 3)
41% (7/1 7)
Negative
Predictive
Value
87% (66/76)
87% (65/75)
86% (66/77)
86% (65/76)
86% (63/73)
87% (61/70)
87% (60/69)
86% (61/71)
87% (58/67)
9 Homes
84
84
84
84
84
84
84
84
84
Unweighted Summary Measure = Sensitivity + Specificity + PPV + NPV
-------
L?
s
i
3
I
Table D-10. Potential Standards for Floor Dust-Lead and Dripline Soil-Lead From the Baltimore Repair and Maintenance
Study, Maximizing the Unweighted Summary Measure.
Floor Dust-
Lead Standard
big/ft*)
50
250 and s175
250 and s175
50
50
50
275 and £ 175
50
275 and s175
275 and s175
Dripline
Soil-Lead
Standard
0/g/a)
2.000
2200 and «300
400
2500 and $600
1.500
2800 and si 000
2500 and s600
700
1,500
2800 and $1000
700
2.000
% Homes
Failing the
Combined
Standard
26% (7/27)
63% (17/27)
59% (16/27)
59% (16/27)
48% (13/27)
52% (14/27)
56% (15/27)
56% (15/27)
41% (11/27)
44% (12/27)
48% (13/27)
7% (2/27)
Unweighted
Summary
Measure
(USM)*
271
266
242
242
238
228
218
218
192
182
172
158
Blood-Lead Standard of 10 //g/dL
Sensitivity
(b/a+b)
50% (4/8)
88% (7/8)
75% (6/8)
75% (6/8)
63% (5/8)
63% (5/8)
63% (5/8)
63% (5/8)
38% (3/8)
38% (3/8)
38% (3/8)
0% (0/8)
Specificity
(c/c+d)
84% (16/1 9)
47% (9/19)
47% (9/19)
47% (9/19)
58% (11/1 9)
53% (10/1 9)
47% (9/19)
47% (9/1 9)
58% (11/1 9)
53% (10/19)
47% (9/1 9)
90% (17/1 9)
Positive
Predictive
Value
(b/b+d)
57% (4/7)
41% (7/1 7)
38% (6/16)
38% (6/1 6)
39% (5/1 3)
36% (5/14)
33% (5/1 5)
33% (5/15)
27% (3/11)
25% (3/1 2)
23% (3/13)
0% (0/2)
Negative
Predictive
Value
(c/a+c)
80% (16/20)
90% (9/10)
82% (9/11)
82% (9/11)
79% (11/14)
77% (10/1 3)
75% (9/12)
75% (9/1 2)
69% (11/16)
67% (10/15)
64% (9/1 4)
68% (17/25)
# Homes
27
27
27
27
27
27
27
27
27
27
27
27
t
I
• Unweighted Summary Measure = Sensitivity + Specificity + PPV + NPV
-------
Appendix E
Tables for the Three-Way Multi-Media Analysis Results
Draft - Do Not Cite or Quote E-1 September 19, 1996
-------
Table E-1. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead, Dripline Soil-Lead, and Percent of Interior Components with Deteriorated
Lead-Based Paint Which Achieved an NPV of 95% or Greater Based on the
Rochester Lead-in-Dust Study.
Uncarpeted Floor
Dust (//g/ft2)
2 50 and $400
2 50 and $400
2 50 and $400
250 and $400
50
275 and $400
275 and $400
Dripline Soil
U/g/g)
200
400
400
400
1,000
700
1,000
Percent of
Interior
Components
with
Deteriorated
LBP
25 and $20
26 and $10
15
20
25 and $10
15
25 and $10
NPV
100% (23/23)
100% (25/25)
100% (28/28)
97% (29/30)
95% (38/40)
95% (40/42)
95% (39/41)
Sensitivity
100% (44/44)
100% (44/44)
100% (44/44)
98% (43/44)
96% (42/44)
96% (42/44)
96% (42/44)
#of
Homes
181
181
181
181
181
181
181
Table E-2. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead, Dripline Soil-Lead, and Percent of Exterior Components with Deteriorated
Lead-Based Paint Which Achieved an NPV of 95% or Greater Based on the
Rochester Lead-in-Dust Study.
Uncarpeted Floor
Dust l/ig/ft2)
250 and $400
2 50 and $400
Dripline Soil
0/g/g)
200
400
Percent of
Exterior
Components
with
Deteriorated
LBP
25 and $20
26 and $20
NPV
96% (24/25)
97% (28/29)
Sensitivity
98% (43/44)
98% (43/44)
#of
Homes
180
180
Draft - Do Not Cite or Quote
E-2
September 19, 1996
-------
Table E-3. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead, Dripline Soil-Lead, and Maximum Percent of Interior/Exterior Components
with Deteriorated Lead-Based Paint Which Achieved an NPV of 95% or Greater
Based on the Rochester Lead-in-Dust Study.
Uncarpeted Floor
Dust (/ig/ft2)
2 50 and $400
2 50 and $400
250 and $400
2 75 and $400
50
Dripline Soil
(09/9)
200
400
400
1,000
1,000
Maximum Percent
of Interior/Exterior
Components with
Deteriorated LBP8
25 and $20
26 and $10
215 and $20
20
20
NPV
100% (23/23)
100% (23/23)
100% (25/25)
95% (40/42)
95% (39/41)
Sensitivity
100% (44/44)
100% (44/44)
100% (44/44)
96% (42/44)
96% (42/44)
#of
Homes
180
180
180
180
180
8 Maximum of the percent of interior components and the percent of exterior components with deteriorated LBP.
Table E-4. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead, Dripline Soil-Lead, and Average Percent of Deteriorated Lead-Based Paint
per Interior Component Which Achieved an NPV of 95% or Greater Based on
the Rochester Lead-in-Dust Study.
Uncarpeted Floor
Dust (//g/ft2)
Dripline Soil
(//g/g)
Average Percent
of Deteriorated
LBP per Interior
Component
NPV
Sensitivity
b
#of
Homes
181
b An NPV i 95% was not achieved for any combination of potential media standards.
Draft - Do Not Cite or Quote
E-3
September 19, 1996
-------
Table E-5. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead, Dripline Soil-Lead, and Average Percent of Deteriorated Lead-Based Paint
per Exterior Component Which Achieved an NPV of 95% or Greater Based on
the Rochester Lead-in-Dust Study.
Uncarpeted Floor
Dust 1/ig/ft2)
Dripline Soil
(//a/9)
Average Percent
of Deteriorated
LBP per Exterior
Component
NPV
Sensitivity
b
#of
Homes
180
An NPV t 95% was not achieved for any combination of potential media standards.
Table E-6. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead, Dripline Soil-Lead, and Maximum Average Percent of Deteriorated Lead-
Based Paint per Interior/Exterior Component Which Achieved an NPV of 95% or
Greater Based on the Rochester Lead-in-Dust Study.
Uncarpeted Floor
Dust (/ig/ftz)
Dripline Soil
(A/g/g)
Maximum
Average Percent
of Deteriorated
LBP per
Interior/Exterior
Component*
NPV
Sensitivity
b
#of
Homes
180
8 Maximum of the percent of interior components and the percent of exterior components with deteriorated LBP.
An NPV 2 95% was not achieved for any combination of potential media standards.
Draft - Do Not Cite or Quote
E-4
September 19, 1996
-------
Table E-7. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead, Play-Area Soil-Lead, and Percent of Interior Components with
Deteriorated Lead-Based Paint Which Achieved an NPV of 95% or Greater
Based on the Rochester Lead-in-Dust Study.
Uncarpeted Floor
Dust U/g/ft2)
Play-Area Soil
U/g/g)
Percent of
Interior
Components
with Deteriorated
LBP
NPV
Sensitivity
b
#of
Homes
84
An NPV 2 95% was not achieved for any combination of potential media standards.
Table E-8. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead, Play-Area Soil-Lead, and Percent of Exterior Components with
Deteriorated Lead-Based Paint Which Achieved an NPV of 95% or Greater
Based on the Rochester Lead-in-Dust Study.
Uncarpeted
Floor Dust
(Jig/ft2)
50
Play-Area Soil
(//g/g)
200
Percent of
Exterior
Components with
Deteriorated LBP
26 and s20
NPV
100% (14/14)
Sensitivity
100% (16/16)
#of
Homes
84
Draft - Do Not Cite or Quote
E-5
September 19, 1996
-------
Table E-9. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead, Play-Area Soil-Lead, and Maximum Percent of Interior/Exterior
Components with Deteriorated Lead-Based Paint Which Achieved an NPV of
95% or Greater Based on the Rochester Lead-in-Dust Study.
Uncarpeted
Floor Dust
O^/ft2)
50
50
50
50
50
Play-Area Soil
(pg/g)
200
200
200
200
600
Maximum
Percent of
Interior/Exterior
Components with
Deteriorated LBP*
5
10
15
20
20
NPV
100% (8/8)
100% (9/9)
100% (10/10)
100% (11/11)
95% (19/20)
Sensitivity
100% (16/16)
100% (16/16)
100% (16/16)
100% (16/16)
94% (15/16)
#of
Homes
84
84
84
84
84
8 Maximum of the percent of interior components and the percent of exterior components with deteriorated LBP.
Table E-10. Performance Characteristics of Potential Standards for Uncarpeted Floor Dust-
Lead, Play-Area Soil-Lead, and Average Percent of Deteriorated Lead-Based
Paint per Interior Component Which Achieved an NPV of 95% or Greater
Based on the Rochester Lead-in-Dust Study.
Uncarpeted Floor
Dust 0/g/ft2)
Play-Area Soil
(pg/g>
Average Percent
of Deteriorated
LBP per Interior
Component
NPV
Sensitivity
b
#of
Homes
84
b An NPV 2 95% was not achieved for any combination of potential media standards.
Draft - Do Not Cite or Quote
E-6
September 19, 1996
-------
Table E-11. Performance Characteristics of Potential Standards for Uncarpeted Floor
Dust-Lead, Play-Area Soil-Lead, and Average Percent of Deteriorated Lead-
Based Paint per Exterior Component Which Achieved an NPV of 95% or
Greater Based on the Rochester Lead-in-Dust Study.
Uncarpeted
Floor Dust
(/ig/ft2)
50
50
s75 and *400
Play-Area Soil
(pg/g)
200
200
200
Average Percent
of Deteriorated
LBP per Exterior
Component
5
10
10
NPV
100% (16/16)
100% (21 721)
96% (23/24)
Sensitivity
100% (16/16)
100% (16/16)
94% (15/16)
#of
Homes
84
84
84
Table E-12. Performance Characteristics of Potential Standards for Uncarpeted Floor
Dust-Lead. Play-Area Soil-Lead, and Maximum Average Percent of
Deteriorated Lead-Based Paint per Interior/Exterior Component Which
Achieved an NPV of 95% or Greater Based on the Rochester Lead-in-Dust
Study.
Uncarpeted
Floor Dust
-------
Appendix F
Results of Adding Window Sill Dust-Lead
to the Multi-Media Analysis
Draft - Do Not Cite or Quote F-1 September 19, 1996
-------
I
\
e?
I
r>
3'
I
Table F-1. Value of Window Sill Standard in Identifying EBL's U 10 //g/dL) Using the Rochester Lead-in-Dust Study Data.
Media: EBL's are identified first for the first media listed, next for the second media, etc.
Uncarpeted Floor Dust-Lead
Standard
50
100
SO
100
# EBL's
Identified
9
5
9
5
Uncarpeted Floor Dust-Lead
Standard
SO
100
# EBL's
•«J A M*£41 «%<4
IO6ntltl80
8
5
Dripline Soil-Lead
Standard
400
400
# EBL's
Identified
38
38
Dripline Soil-Lead
Standard
400
400
* EBL's
IdontifioQ
38
38
Window Sill Dust-Lead
Standard
100
100
500
500
Additional #
EBL's
Identified
28
31
12
13
Window Well Dust-Lead
Standfifd
800
800
Additional #
EBL's
Identified
26
28
Uncarpeted Floor Dust-Lead
Standard
100
100
Additional »
EBL's
Identified
1
1
Uncarpeted Floor Dust-Lead
Standard
100
100
Additional #
EBL's
Identified
1
1
Window Sill Dust-Lead
Standard
100
500
Additional S
EBL's
IdBntifffod
3
2
% of Interior or Exterior
Components with LBP
Standard
20
20
Additional #
EBL's
Identified
4
4
Window Sill Dust-Lead
Standard
100
500
Additional #
EBL's
Identified
O
O
Total » of
EBL's
Identified
37
36
21
18
Total tt of
EBL's
Identified
34
33
Total # of
EBL's
Identified
42
41
Total » of
EBL's
Identified
43
43
Total # of
EBL's
45
45
45
45
Total # of
EBL's
42
42
Total # of
EBL's
43
43
Total # of
EBL's
43
43
I
0>
-------
Appendix G
Tables for the Four-Way and Five-Way Multi-Media
Sensitivity/Specificity Analyses
Draft - Do Not Cite or Quote G-1 September 19. 1996
-------
Table G-1. Four-Way Multi-Media Standards which Achieved NPV and Sensitivity of 95%
or Greater for Dripline Soil, Uncarpeted Floor Dust, Window Sill Dust, and the
Maximum of the Percent of Interior Components and Percent of Exterior
Components with Deteriorated LBP for Data From the Rochester Lead-in-Dust
Study.
Uncarpeted Roar
Dust-Lead
Standard (pg/ft2)
400.200. 175,
150. 125. 100.
75.50
Dripline Soil-Lead
Standard
(pg/g)
600
500. 400, 300
1,500
600
1,000
600
600
900
600
700
1,500
1,000
Window Sill Dust-
Lead Standard
(pg/ft2)
100
800, 500, 300, 100
100
800, 500
100
800, 500
300
100
300
100
100
800, 500
Maximum of
Percant of
Interior/Exterior
Componojits
witn Detanorawo
LBP
20, 10, 5
20, 10, 5
20
20
20
10,5
20
20
10,5
20
10,5
20
NPV
100%
100%
96%
96%
96%
96%
96%
96%
96%
96%
95%
95%
Sensitivity
100%
100%
98%
98%
98%
98%
98%
98%
98%
98%
98%
95%
#of
Homes
174
Table G-2. Four-Way Multi-Media Standards which Achieved NPV and Sensitivity of 95%
or Greater for Play Area Soil, Uncarpeted Floor Dust, Window Sill Dust, and the
Maximum of the Percent of Interior Components and Percent of Exterior
Components with Deteriorated LBP for Data From the Rochester Lead-in-Dust
Study.
Uncarpeted Floor
Dust-Lead
Standard 0/g/ft2)
50
Play Area Soil-
Lead Standard
(figlg)
300, 200
Window Sill Dust-Lead
Standard
O/g'ff)
800, 500. 300. 100
Maximum of
Percent of
Interior/Exterior
Components
LBP
20, 10, 5
NPV
100%
Sensitivity
100%
# of
Homes
80
Draft - Do Not Cite or Quote
G-2
September 19, 1996
-------
Table G-3. Selected Five-Way Multi-Media Standards which Achieved NPV and Sensitivity
of 100% for Dripline Soil, Play Area Soil, Uncarpeted Floor Dust, Window SiOl
Dust, and the Maximum of the Percent of Interior Components with
Deteriorated LBP and Percent of Exterior Components with Deteriorated LBP
for Data From the Rochester Lead-in-Dust Study.
Uncarpeted
Floor Dust-Lead
Standard
(pg/ft2)
400,200. 175.
150. 125, 100,
75 50
Maximum of
Porcont of
Interior/Exterior
Components with
Deteriorated LBP
20. 10, 5
Dripline Soil-
Lead Standard
U/g/g)
1,500
1,500
1.000. 900,
700, 600, 500,
400, 300
Play Area Soil-
Lead Standard
1,000,900,
700, 600, 5,00,
400, 300. 200
600, 500, 400.
300, 200
1,000,900,
700, 600, 500,
400, 300, 200
Window Sill
Dust-Lead
Standard
Uig/ft2)
500 300
100
800
ROD BOO
300 100
NPV
100%
100%
100%
Sensitivity
100%
100%
100%
#of
Homes
77
Draft - Do Not Cite or Quote
G-3
September 19. 1996
-------
Appendix H
Summary of Estimated Standards Achieving
Current Target Health Effects
Draft - Do Not Cite or Quote H-1 September 19, 1996
-------
I
\
Table H-1. Summary ol Estimated Standards Which Achieve a Negative Predictive Value of 95% or an Estimated 95%
Probability of a Child's Blood-Lead Concentration Below 10//g/dL in a Dwelling that is at or Below the
Standard.
Madia
Dnpline Soil b/g/g)
Play Area Soil v/g/g)
Uncarpeted Floor 0/g/ft'l
Carpeted Floor 0/g/ft'l
Carpeted and Uncarpeted
Floor b/g/ft*)
Window Sills U/g/fl1)
Window Troughs big/ft')
Percent of Interior Components
with Deteriorated LBP
Average Percent of Deteriorated
LBP per Intenor Component
Percent of Exterior Components
with Detenorated LBP
Average Percent of Deteriorated
LBP per Exterior Component
Maximum of Intenor/ Exterior
Percent of Components with
Deteriorated LBP
Maximum of Interior/ Exterior
Average Percent of Deteriorated
LBP per Component
Single Media Analyses
Regression
•irwtnln
HIOOBI9
Rochester
37
30
•
•
•
8
8
•
•
•
•
Sensitivity/Specificity
R&M
<50t
5
25
200
•
•
*
nOCnOStOf
BO
40
2
5
2
25
<50t
•
•
•
•
•
•
Multi-Media Analyses
nBQression
Models*
Selected
Combinations
300
1
10
40
1000
1
10
40
Sensitivity /Specificity
Joint Sofl and Dust
D rtftmm ta J&
fflOCIlBSlB*
Joint
Standard
1
<200
<50
Joint
Standard
2
<200
<50
Joint Soil. Oust and Paint
Rochester*
Joint
St&ndsnf
1
1000
400
20%
Joint
Standard
2
600
50
20%
i
Joint Soil. Dust. Paint and Sills
RoctiBster
Joint
Stsnd&nf
i
500
400
800
20%
Joint
Standard
2
1500
1000
400
500
20%
X
ro
I
• No values for a standard resulted in an associated 95% probability of a child's blood-lead concentration below 10ig/dL.
t Values below 50 were not assessed.
' Multi-media sensitivity specificity results for the R&M study were not included because 1) they were not conducted for the three-way (soil, dust, paint) or four-way (soil, dust, paint, sills)
analyses since only 8 homes were available for analysis, and 2) they do not change the results given for the two-way (soil, dust) analysis.
1 The multi-media regression model included a variable that indicated a combination paint/pica hazard which was set to zero for these analyses.
1 The lowest dust lead standard assessed in the two-way joint soil and dust sensitivity/specificity analyses was 50 j/g/ft*. The lowest soil lead standard assessed was 200 jig/g. No
combination of soil and dust for either the Rochester or R&M data achieved an NPV of 95% with those lower limits on soil and dust standards.
4 In the three-way (soil, dust, paint) multi-media analyses and higher (four-way and five-way), the combination of standards with the highest values that achieve the target health criteria were
chosen for presentation in this table.
-------
June 12, 1996
ADJUSTMENTS TO NATIONAL SURVEY DATA
FOR SECTION 403 ANALYSES
APPENDIX Z
Note to EPA: Draft Report is currently undergoing peer review
-------
This is a report of research performed for the United States
Government by Battelle. Because of the uncertainties inherent in
experimental research work, Battelle assumes no responsibility or
liability for any consequence of use, misuse, inability to use, or
reliance upon the information contained herein, beyond any express
obligations embodied in the governing written agreement between
Battelle and the United States Government.
-------
EXECUTIVE SUMMARY
The Risk Assessment, Economic Analyses and Regulatory Impact Analyses being
conducted to support the rulemaking for Section 403 of Title IV all require information on
environmental-lead levels in the national housing stock. The HUD National Survey is the primary
source of data on environmental-lead levels in the national housing stock. During the course of the
rulemaking, two issues were raised concerning the quality of the dust samples collected in the
HUD National Survey: (1) the effect of low tap weights on measured dust-lead concentrations
and-(2) the use of wipe sampling for collecting samples from floors, window sills and window
wells.
Most of the dust samples in the HUD National Survey were collected using the Blue
Nozzle vacuum. The Blue Nozzle vacuum collects dust in a plastic cassette containing a piece of
filter paper to capture the dust. In the HUD National Survey, only the weight of the dust that was
tapped out of the cassette, referred to herein as tap weight, was determined. However, the
amount of lead was determined for both the dust tapped out of the cassette and residual dust on
the filter cassette. Dust-lead concentration was then computed as the total amount of lead
divided by the tap weight. Reported dust-lead concentrations will be biased if the tap weight
underestimates the combined weight of the dust tapped out of the filter cassette and residual dust
remaining on the filter cassette. This bias, if present, will result in higher lead concentrations for
samples with lower tap weights compared to samples with higher tap weights.
An analysis of the dust samples collected from the 284 privately-owned residences in the
HUD National Survey was performed to investigate possible differences in dust-lead
concentrations as a function of the tap weight. Descriptive analyses of the lead concentrations
indicated that the dust-lead concentrations reported in the HUD National Survey increased with
decreased tap weight. The trend may be due to a bias.
Laboratory experiments were conducted to generate a database for understanding and
addressing the bias in dust-lead concentration computed via tap weight in the HUD National
Survey. A regression model, developed for correcting the bias, was fitted to the laboratory data.
Dust-lead concentrations reported in the HUD National Survey were then corrected using
-------
corrections predicted by the regression model. The regression model removed the trend of dust-
lead concentrations increasing with decreased tap weight.
For purposes of analyses conducted to support Section 403, the correction will be applied
to dust samples possessing tap weights greater than or equal to 0.7 mg. Tap weights in the
dataset used to develop the regression model were all greater than or equal to 0.7 mg. The
correction factor predicted by the regression model may over-correct dust-lead concentrations for
samples with tap weights less than 0.7 mg. For the purposes of analyses conducted to support
Section 403, dust-lead concentrations of dust samples with a tap weight less than 0.7 mg will be
excluded from the analysis of dust-lead concentrations. This decision applies only to the lead
concentrations and not the lead loadings of these samples.
Dust samples collected via wipe sampling in the HUD National Survey are identified in
this report. Only two of the samples are being utilized in the analyses for Section 403. The dust-
lead concentration is missing for one sample and the lead concentration of the other sample will
be excluded from Section 403 analyses. There is no reason to suspect the quality of the dust-lead
loadings of the two samples. Therefore, they will be utilized in the analyses conducted for
Section 403.
-------
TABLE OF CONTENTS
EXECUTIVE SUMMARY
1.0 INTRODUCTION 1
2.0 EFFECT OF LOW TAP WEIGHT ON DUST-LEAD CONCENTRATION 1
2.1 Why Dust-lead Concentrations May Be Biased 2
2.2 Statistical Assessment of Low Tap Weight Samples 3
2.3 Laboratory Assessment of Low Tap Weight Samples 11
2.4 Correcting the Bias in Dust-lead Concentration 13
3.0
WIPE SAMPLES IN HUD NATIONAL SURVEY ............................. 14
4.0 DISCUSSION, SUMMARY AND CONCLUSIONS ............................ 21
APPENDIX A Laboratory Assessment of the Amount of Residual Dust Remaining in Blue Nozzle
Filter Cassettes
APPENDIX B Data Tables for Dust-Lead concentrations for Floor, Window, and Window Well
Samples in the 284 Privately-Owned Houses in the HUD National Survey
LIST OF TABLES
Table 1. Floor Dust-Lead Concentrations by Tap Weight for the 284 Privately-Owned
Houses in the HUD National Survey 5
Table 2. Window Sill Dust-Lead Concentrations by Tap Weight for the 284 Privately-
Owned Houses in the HUD National Survey 6
Table 3. Window Well Dust-Lead Concentrations by Tap Weight for the 284 Privately-
Owned Houses in the HUD National Survey 7
Table 4. Results of Outlier Analysis 11
Table 5. Test Matrix for Laboratory Experiments 12
Table 6. Floor Dust-Lead Concentrations Corrected for Bias by Tap Weight for the 284
Privately-Owned Houses in the HUD National Survey 15
Table 7. Window Sill Dust-Lead Concentrations Corrected for Bias by Tap Weight for
the 284 Privately-Owned Houses in the HUD National Survey 16
Table 8. Window Well Dust-Lead Concentrations Corrected for Bias by Tap Weight for
the 284 Privately-Owned Houses in the HUD National Survey 17
Table 9. Distribution of Wipe Samples Taken in National Survey 18
LIST OF FIGURES
Figure 1. Floor Dust-Lead Concentration Plotted Against Tap Weight for Samples with
Tap Weight Equal to 0.1 mg 9
Figure 2. Floor Dust-Lead Concentration Plotted Against Tap Weight for Samples with
Tap Weights Greater than 0.1 mg and Less than 5 mg. Two Samples with
Dust-Lead Concentrations Greater than 100,000//g/g are not Shown in the Plot. ... 9
iii
-------
TABLE OF CONTENTS
(Continued)
Page
Figure 3. Floor Dust-Lead Concentration Plotted Against Tap Weight for Samples with
Tap Weights Greater than or Equal to 5 mg and Less than 10 mg. One Sample
with a Dust-Lead Concentration Greater than 100,000 //g/g is not Shown in
the Plot 10
Figure 4. Floor Dust-Lead Concentration Plotted Against Tap Weight for Samples with
Tap Weights Greater than or Equal 10 mg 10
Figure 5. Ratio of Total Weight of Dust Divided by Tap Weight Plotted Against Tap
Weight and Overlaid with Fitted Regression Model 13
Figure 6. Dust-Lead Concentration Corrected for Bias Plotted Against Tap Weight for
Samples with Tap Weight Equal to 0.1 mg 19
Figure 7. Floor Dust-Lead Concentration Corrected for Bias Plotted Against Tap Weight
for Samples with Tap Weights Greater than 0.1 mg and Less than 5 mg. Two
Samples with Dust-Lead Concentrations Greater than 29,000 //g/g are not
Shown in the Plot 19
Figure 8. Floor Dust-Lead Concentration Corrected for Bias Plotted Against Tap Weight for
Samples with Tap Weights Greater than or Equal to 5 mg and less than 10 mg.
One Sample with a Dust-Lead Concentration Greater than 29,000 //g/g is not
Shown in the Plot 20
Figure 9. Floor Dust-Lead Concentration Corrected for Bias Plotted Against Tap Weight
for Samples with Tap Weights Greater than or Equal to 10 mg 20
-------
1.0 INTRODUCTION
Section 403 of Title IV of the Toxic Substances Control Act (TSCA), as amended in
Title X, the Residential Lead-Based Paint Hazard Reduction Act, will set standards (condition
and location of lead-based paint, levels of lead in dust and soil) against which to compare a
residential environment when evaluating the presence and magnitude of lead-based paint
hazards. The Risk Assessment, Economic Analyses and Regulatory Impact Analyses being
conducted for this rule all require information on environmental-lead levels in the national
housing stock. The HUD National Survey is the primary source of data on environmental-lead
s
levels in the national housing stock. Conducted in 1989-1990, the HUD National Survey
collected data on lead levels in paint, dust, and soil from 284 privately-owned occupied housing
units. The units were selected via a statistically-based sampling design to represent the national
housing stock built prior to 1980.
During the course of the analyses to support Section 403 rulemaking, two issues were
raised concerning the quality of the dust samples collected in the HUD National Survey: (1) the
effect of low tap weights on measured dust-lead concentrations and (2) the use of wipe sampling
for collecting samples from floors, window sills and window wells. This report documents an
assessment of these two issues. The assessment involved only dust samples collected from
entryways, dry rooms, and wet rooms in the 284 privately-owned residences sampled in the HUD
National Survey.
Section 2 discusses the effect of low tap weight on lead concentration for dust samples
collected in the HUD National Survey. The identification of dust samples collected with wipes is
addressed in Section 3. An evaluation of these two issues on the Section 403 Risk Assessment
and recommendations for resolving these two issues are presented in Section 4.
2.0 EFFECT OF LOW TAP WEIGHT ON DUST-LEAD CONCENTRATION
Section 2.1 discusses why dust-lead concentrations measured in the HUD National
Survey may be biased. Furthermore, it is indicated there that the bias is likely to increase with
decreased tap weight and that the bias may be expressed as a function of the ratio of total dust
weight to tap weight. Descriptive analyses presented in Section 2.2 show that the dust-lead
-------
concentrations reported in the HUD National Survey increased with decreased tap weight. The
trend may be due to a bias.
Laboratory experiments were conducted to measure weight of residual dust on filter
cassettes, tap weight, total dust weight, and ratio of total dust weight to tap weight for dust
samples collected via the Blue Nozzle vacuum. The experiments, summarized in Section 2.3,
provided a database for understanding and addressing the bias in dust-lead concentration
computed via tap weight in the HUD National Survey. A regression model, developed for
correcting the bias, was fitted to the laboratory data. In Section 2.4, the regression model is
applied to the dust-lead concentrations collected in the HUD National Survey. Furthermore, it is
shown in Section 2.4 that the regression model removes the trend of dust-lead concentrations
increasing with decreased tap weight.
2.1 WHY DUST-LEAD CONCENTRATIONS MAY BE BIASED
Dust samples were collected from interior window sills, window wells and floors in the
HUD National Survey using the Blue Nozzle vacuum device. The Blue Nozzle vacuum collects
dust in a plastic cassette containing a piece of filter paper to capture the dust. In the HUD
National Survey, the filter cassettes were not preweighed before sampling, and only the weight of
the dust that was tapped out of the cassette, referred to herein as tap weight, was determined.
However, the amount of lead was determined for each sample by chemical analysis of both the
dust tapped out of the cassette and the filter cassette, combined. Dust-lead concentration was
then computed as the total amount of lead divided by the tap weight.
It is believed that some dust may have remained on the cassette filter. In which case, the
dust-lead concentrations reported in the HUD National Survey may be biased. Dust-lead
concentration is defined as weight of lead in sampled dust divided by weight of sampled dust.
Dust-lead concentration reported in the HUD National Survey will be biased if the tap weight
underestimates the combined weight of the dust tapped out of the filter cassette and residual dust
remaining on the filter cassette. Therefore, the measured dust-lead concentration is equal to
-------
(ug of Pb on filter cassette) * (ug of Pb for tapped-out dusf)
(g of tapped-out dusf)
instead of
(ug of Pb on filter cassette) + (ug of Pb for tapped-out dust)
(g of dust on filter cassette) + (g of tapped-out dust)
The ratio of the two terms is
(g of dust on filter cassette) + (g of tapped-out dust) _ total dust weight
(g of tapped-out dusf) tap weight
where total dust weight is the total weight of the dust tapped out of the filter cassette and residual
dust remaining on the filter cassette. Therefore, the bias in the reported dust-lead concentrations
is a function of the ratio of the total dust weight to the tap weight. Because the weight of the
residual dust on the filter cassette is likely to comprise a larger percentage of the total dust weight
for low tap weight samples, the bias is expected to increase with decreased tap weight. This bias
in dust-lead concentration, if present in the HUD National Survey, would result in higher dust-
lead lead concentrations for samples with lower tap weights compared to samples with higher tap
weights.
2.2 STATISTICAL ASSESSMENT OF LOW TAP WEIGHT SAMPLES
An analysis of the dust samples collected from the 284 private houses was performed to
investigate possible differences in dust-lead concentrations as a function of the tap weight. Only
window and floor samples from dry-rooms, wet-rooms and entranceways were included in the
analysis. (All samples collected from common areas in the HUD National Survey were
excluded).
-------
To evaluate the impact of low tap weight on dust-lead concentrations, dust-lead
concentrations were examined for nine levels of tap weight spanning from 0.1 mg to greater than
25 mg. The first level of tap weight, tap weight of 0.1 mg, was called out separately because 0.1
mg may represent the limit of detection for the gravimetric analysis. Furthermore, dust-lead
loadings and/or dust-lead concentrations were missing for 599 of the 718 samples with a tap
weight of 0.1 mg.
Table 1 presents the number, geometric mean, and range of the dust-lead concentrations
for each of the nine levels of tap weight. The second column in the table displays the number of
interior window sill, window well and floor samples with a tap weight in the given category. The
third column displays the number of floor dust samples with a tap weight in the given category
and the fourth column provides the number of these samples with a nonmissing dust-lead
concentration. The geometric mean and range of the dust-lead concentrations for the floor dust
samples are shown in columns five to seven. Column 8 provides the number of floor dust
samples with a dust-lead concentration greater than 1000 ug/g. The number of non-missing dust-
lead loadings and the geometric mean dust-lead loadings of the floor dust samples are presented
in the last two columns of the table. Tables 2 and 3 present similar results for window sill and
window well samples, respectively.
An inverse relationship between dust-lead concentration and tap weight is apparent in
Table 1; dust-lead concentration decreased with increased tap weight. A similar relation is seen
in Table 2 for window sill dust-lead concentration. As shown in Table 3, there is no apparent
trend between dust-lead concentration and tap weight for samples collected from window wells.
However, only 25 out of the 149 dust samples collected from window wells had tap weights less
than 10 mg and window well dust-lead concentrations tend to be more variable than those
collected from floors and window sills.
-------
Table 1. Floor Dust-Lead Concentrations by Tap Weight for the 284 Privately-Owned Houses in the HUD
National Survey
Tap Weight
Level
0.1 mg
* 0.2 to < 1 mg
t 1 to < 2 mg
fc 2 to •< 3 mg
t 3 to < 4 mg
2: 4 to < 5 mg
t 5 to -< 10 mg
t 10 to -c25mg
>• = 25 mg
Number of
Dust
Samples
718"
81
56
55
52
31
179
199
617
Number of
Floor Dust
Samples
40
37
30
35
35
15
119
137
404
Dust-Lead Concentration (M9/9)
Number
5
37
30
35
35
15
118
137
402
Geometric
Mean
6543.04
1286.78
751.99
427.40
310.37
326.80
279.69
209.66
189.76
Minimum
4656.28
41.39
75.01
24.93
3.45
43.08
24.55
1.13
0.05
Maximum
13023.23
18563.00
354774.13
4327.69
114281.43
3772.20
132887.36
3756.41
11287.21
Number
>1000
5
23
12
5
7
3
11
12
42
Dust-Lead Loading
tug/ft2)
Number
26
35
30
35
35
15
117
137
402
Geometric
Mean
0.12
0.14
0.27
0.26
0.27
0.38
0.50
0.85
4.36
599 samples have missing dust-lead loadings and concentrations for this tap weight; 119 samples have either dust-
lead loading or concentration.
-------
Table 2. Window Sill Dust-Lead Concentrations by Tap Weight for the 284 Privately-Owned Houses in the HUD
National Survey
Tap Weight
Level
0.1 mg
t 0.2 to < 1 mg
t 1 to < 2 mg
t 2 to < 3 mg
t 3 to t 4 mg
t 4 to t 5 mg
t 5 to < 10 mg
fc 10to<25mg
>• = 25 rug
Number of
Dust
Samples
718
81
56
55
52
31
179
199
617
Number of
Window Sill
Dust Samples
255
41
26
19
13
13
51
44
106
Dust-Lead Concentration (ug/g)
Number
27
41
26
19
13
13
51
44
106
Geometric
Mean
3798.27
2030.71
1571 84
988.93
940.06
450.74
657.95
476.17
631.88
Minimum
124.16
15.52
37.68
55.87
15578
4286
3.06
1.74
0.79
Maximum
55871.70
5845835.5
104367 71
7501.29
9635.28
4526 01
96491.66
9651.69
101958.90
Number
>1000
22
25
16
9
7
4
20
14
41
Dust-Lead Loading
Jug/ft2)
Number
79
41
26
19
13
13
51
44
106
Geometric
Mean
032
1.03
215
1 97
393
1.87
392
4.99
44.84
-------
Table 3. Window Well Dust-Lead Concentrations by Tap Weight for the 284 Privately-Owned Houses in the HUD
National Survey
Tap Weight
Level
0.1 mg
t 0.2 to •< 1 mg
& 1 to t 2 mg
fc 2 to •< 3 mg
t 3 to < 4 mg
1 4 to •< 5 mg
t 5 to 1 10 mg
fc 10 to < 25 mg
>• = 25 mg
Number of
Dust
Samples
718
81
56
55
52
31
179
199
617
Number of
Window Well
Dust Samples
423
3
1
4
3
9
18
107
Dust-Lead Concentration (ug/g)
Number
5
3
1
4
3
9
18
106
Geometric
Mean
10680.66
1830.51
122.09
2449.07
4783.72
4920.01
1678.82
1770.24
Minimum
2431.64
30465
.
122.09
473.19
2815.24
1006.31
33.74
5.17
Maximum
457178.46
4980.61
122.09
7796.16
13335.63
63101.24
83633.26
109165.57
Number
>1000
5
2
.
0
3
3
9
10
69
Dust-Lead Loading
(ug'ff)
Number
14
3
1
4
3
9
18
106
Geometric
Mean
0.55
0.70
•
0.37
9.88
25.69
34.54
33.29
242.80
-------
Floor dust-lead concentrations are plotted against tap weights in Figures 1-4, with each of the
plots corresponding to one of four tap weight classifications. To better show the data, the y-axis
was truncated at 20,000 ug/g and floor dust samples with dust-lead concentrations greater than
20,000 ug/g were omitted from each of the plots. A total of three samples were excluded, two in
Figure 2 and one in Figure 3. All three samples had dust-lead concentrations greater than
100,000 ug/g. Dust-lead concentrations for the 0.1 mg tap weight samples are all greater than
4000 ug/g. Although the trend of increased dust-lead concentration with decreased tap weight is
quite visible in Figure 2, the relationship is not obvious in Figures 3 and 4.
Additional analyses were conducted to determine if any of the floor dust-lead
concentrations for the 284 privately-owned homes in the HUD National Survey were statistical
outliers. An outlier is a data point that is abnormally large or small in value. Because large and
small are relative terms, externally deleted studentized residuals were employed to determine
how large (or small) each dust-lead concentration was relative to the mean dust-lead
concentration for the floor samples in the same tap weight category for four categories: equal to
0.1 mg, greater than 0.1 mg and less than 5 mg, greater than or equal to 5 mg and less than 10
mg, and greater than or equal to 10 mg. First, a one-way analysis of variance was conducted
using the category of tap weight (four levels, as described above) as the explanatory factor.
Second, externally deleted studentized residuals were computed for each sample. Third, each
externally deleted residual was compared to the extreme value likely to occur if the data were
sampled from a normal distribution. The extreme value was defined to be the 0.025/(number of
floor dust samples) lower or upper percentile of the normal distribution. This definition utilizes a
Bonferoni adjustment to account for the simultaneous comparisons. Based on this analysis, three
floor dust-lead concentrations were determined to be unusually large and six were determined to
be unusually small. The sampling location, dwelling identification number, dust tap weight,
dust-lead concentration, and externally deleted studentized residual for these nine samples are
shown in Table 4. The window sill and well dust data and paint data were examined for the two
homes (2751402 and 0131102) with the large floor dust-lead concentrations. Dust-lead
concentrations for window sill and well samples were not unusual. In fact, the maximum
observed XRF reading was less than 2.0 mg/cm2 for both homes and neither home contained any
damaged lead-based paint.
8
-------
c 10000 <
I
I
0.1
Dual Tap w«lght (mg)
Figure 1. Floor Dust-Lead Concentration Plotted Against Tap Weight for
Samples with Tap Weight Equal to 0.1 mg.
o 4_
Du»t Tap weight (mg)
Figure 2. Floor Dust-Lead Concentration Plotted Against Tap Weight for Samples with Tap
Weights Greater than 0.1 mg and Less than 5 mg. Two Samples with Dust-Lead
Concentrations Greater than 100,000 pg/g are not Shown in the Plot.
-------
15OOO -t
= 1OOOO
7 e
Duat Tap W«lght (mo)
Figure 3. Floor Dust-Lead Concentration Plotted Against Tap Weight for Samples with Tap
Weights Greater than or Equal to 5 mg and Less than 10 mg. One Sample with
a Dust-Lead Concentration Greater than 100,000 /sg/g is not Shown in the Plot.
= 10000
o
5
3OOO
Dual Tap weight (mg)
Figure 4. Floor Dust-Lead Concentration Plotted Against Tap Weight for Samples with Tap
Weights Greater than or Equal 10 mg.
10
-------
Table 4. Results of Outlier Analysis
Sampling
Location
Dry-room floor
Wet-room floor
Wet-room floor
Dry-room floor
Dry-room floor
Wet-room floor
Wet-room floor
Entryway floor
Entryway floor
Dwelling ID
2751402
2751402
0131102
0310201
0430306
2622603
1041607
1731603
0340505
Dust Tap Weight
(mg)
8.2
3.5
1.2
3.6
3580.1
681.1
4999.8
11
2041.7
Dust-Lead Concentration
0/g/g)
132,900
114,300
354,900
3.45
0.58
0.84
0.05
1.13
1.01
Externally Deleted
Studentized Residual
4.62
3.97
4.83
-3.81
-4.34
-4.06
•6.21
-3.83
-3.92
2.3 LABORATORY ASSESSMENT OF LOW TAP WEIGHT SAMPLES
As discussed in Section 2.1, dust-lead concentration reported in the HUD National
Survey may be biased due to use of tap weight in calculation of dust-lead concentration. The
potential bias in the dust-lead concentration is a function of the ratio of the total dust weight to
the tap weight. Laboratory experiments were conducted using the Blue Nozzle vacuum to
measure weight of residual dust on filter cassettes, tap weight, total dust weight, and ratio of total
dust weight to tap weight. This section summarizes the results of those experiments and how the
results may be used to estimate the bias in the HUD National Survey dust-lead concentrations. A
more detailed discussion of the laboratory experiments, including tables of the collected data, is
provided in Appendix A.
Different weights of dust, of various particle sizes, were placed on a 12-inch tile
substrate, and the Blue Nozzle sampler was used to vacuum the dust off the substrate. Vacuum
sampling was done in accordance with the protocol specified in Appendix H of the EPA report
"Laboratory Evaluation of Dust and Dust Lead Recoveries for Samplers and Vacuum Cleaners".
After each experiment, tiles were cleaned with Kimwipes using distilled water and allowed to dry
overnight. The filter cassette was preweighed so that the weight of dust remaining on the filter
cassette, after dust was tapped out of the cassette, could be determined. In this way, both the
11
-------
weight of the dust tapped out of the cassette, as well as the weight of dust remaining on the filter
cassette, were determined.
As shown in Table 5, experiments were conducted for a total of nine combinations of
dust particle sizes and amount of dust applied to the sampled surface. Each combination of
particle size and amount of dust was repeated at least five times. House dust from a previously
conducted EPA study was used in these experiments.
Table 5. Test Matrix for Laboratory Experiments
Target amount of dust
tapped out
0.2-5.0 mg
5.0-10.0 mg
10.0-20.0 mg
Particle sizes of dust
< 53 //m
8
6
8
150-21 2 pm
5
6
7
250-2000 fim
5
5
9
The ratio of the total dust weight to the tap weight was computed for each experiment.
Ratio data are plotted against tap weight in Figure 5. The trend noted in Section 2.2 between
dust-lead concentration and tap weight is also apparent in Figure 5; the observed ratios decrease
with increased tap weight.
A regression model was fitted to the data to estimate the relationship between the
observed ratios and tap weights. Details of the regression model are provided in Appendix A.
The equation of the fitted model, shown as the solid line in Figure 5, is
Ratio =
31 - 19*Tap Weight + 3.3*(Tap Weight)2
2.2 - 0.049*Tap Weight
1
Tap Weight < 2.94 mg
2.94 mg * Tap Weight < 24.5 mg
24.5 mg £ Tap Weight
The last segment (tap weight ;> 24.5 mg) was added so that the estimated ratio will always be
greater than or equal to one.
12
-------
10 15 20
Tap Weight (mg)
25
30
Figure 5. Ratio of Total Weight of Dust Divided by Tap Weight Plotted Against Tap
Weight and Overlaid with Fitted Regression Model.
2.4 CORRECTING THE BIAS IN DUST-LEAD CONCENTRATION
The regression model presented in Section 2.3 was used to estimate the ratio of the total
dust weight to tap weight for dust samples collected in the HUD National Survey. The ratios
were predicted only for the floor and window samples collected from dry-rooms, wet-rooms and
entranceways in the 284 privately-owned houses in the HUD National Survey. Dust-lead
concentrations reported in the HUD National Survey were then corrected by dividing the reported
dust-lead concentration by the predicted ratio of the total dust weight to tap weight. Table B-l in
Appendix B displays the corrected dust-lead concentrations alongside those originally reported in
the HUD National Survey.
Descriptive statistics for the corrected floor, window sill and window well dust-lead
concentrations are presented in Tables 6 to 8, respectively. Tables 1 to 3 provided analogous
13
-------
information for the uncorrected dust-lead concentrations. The inverse relationship between dust-
lead concentration and tap weight noted in Tables 1 and 2 are no longer present in Tables 6 and
7, thereby providing evidence that the correction factor may have removed the bias.
The corrected floor dust-lead concentrations are plotted against tap weights in
Figures 6-9, with each of the plots corresponding to one of four tap weight classifications. To
better show the data, the y-axis was truncated 12,000 ug/g and floor dust samples with dust-lead
concentrations greater than 12,000 fig/g were omitted from each of the plots. A total of three
samples were excluded, two in Figure 2 and one in Figure 8. All three samples had dust-lead
concentrations greater than 29,000 u,g/g. Analogous plots for the uncorrected data were given in
Figures 1-4. Once again, the trend of increased dust-lead concentration with decreased tap
weight seen in Figure 2 is no longer apparent in the analogous figure, Figure 7.
3.0 WIPE SAMPLES IN HUD NATIONAL SURVEY
The EPA report on the HUD National Survey1 stated that a few dust samples were
collected with wet wipes in homes where vacuuming was impossible (page 3-39). However,
there are no identifiers in the database that distinguish between the wipe and vacuum samples.
The only record of these samples is Table C-l in Appendix C of that EPA report, which is not
legible. Westat reviewed the laboratory records from the HUD National Survey to determine
which dust samples were collected by wipe sampling. Table 9 presents the dwelling
identification numbers and sampling locations of the wipe samples. Analyses conducted for
Section 403 are utilizing the dust samples collected from floors, window sills and window wells
in either the entryway, wet room or dry room in the 284 privately-owned residences samples in
the HUD National Survey. Dust samples collected from the common areas of privately-owned
residences and samples collected from publicly-owned residences are not being utilized. Only
two of the samples listed in Table 9 meet this criteria: a window well sample from the dry room
of housing unit 1820802 and a window sill sample from the dry room of housing unit 0440602.
1 U.S. Environmental ProtectionAgency. Report on the National Survey of Lead-Based
Paint in Housing. Appendix II: Analysis. EPA Report No. R747-R95-005. June 1995.
14
-------
Table 6. Floor Dust-Lead Concentrations Corrected for Bias by Tap Weight for the 284 Privately-Owned
Houses in the HUD National Survey
Tap Weight
Level
0.1 mg
* 0.2 to < 1 mg
fc 1 to t 2 mg
1 2 to t 3 mg
t 3 to -c 4 mg
t 4 to -< 5 mg
t 5 to •< 10 mg
t 10 to < 25 mg
> = 25 mg
Number of
Dust Samples
718
81
56
55
52
31
179
199
617
Number of
Floor Dust
Samples
40
37
30
35
35
15
119
137
404
Dust-Lead Concentration (\iglg)
Number
5
37
30
35
35
15
118
137
402
Geometric
Mean
227.78
61.67
81.21
137.38
153.09
165.62
151.96
153.41
189.76
Minimum
162.10
1.65
6.19
6.36
1.70
21.98
12.59
0.68
0.05
Maximum
453.38
1077.26
29293.22
1743.53
56337.86
1919.90
73900.56
2625.57
11287.21
Number
>1000
0
1
1
2
1
1
6
5
42
Dust-Lead Loading
(M9'ff)
Number
26
35
30
35
35
15
117
137
402
Geometric
Mean
0.12
0.14
0.27
0.26
0.27
0.38
0.50
0.85
4.36
-------
Table 7. Window Sill Dust-Lead Concentrations Corrected for Bias by Tap Weight for the 284 Privately-
Owned Houses in the HUD National Survey
Tap Weight
Level
0.1 mg
* 0.2 to •< 1 mg
fc 1 to < 2 mg
t 2 to < 3 mg
t 3 to < 4 mg
t 4 to < 5 mg
t 5 to < 10 mg
t 10 to < 25 mg
>• = 25 mg
Number of
Dust
Samples
718
81
56
55
52
31
179
199
617
Number of
Window Sill
Dust Samples
255
41
26
19
13
13
51
44
106
Dust-Lead Concentration (ug/g)
Number
27
41
26
19
13
13
51
44
106
Geometric
Mean
132.23
97.15
171.10
272.75
464.73
22788
35464
326.05
631.88
Minimum
4.32
0.90
4.20
20.37
76.98
2187
1.73
1.44
0.79
Maximum
1945.07
217433.50
14512.20
2084.21
4715.78
2292.12
5115086
6253.53
101958.90
Number
>1000
3
6
3
4
5
2
13
12
41
Dust-Lead Loading
fog/ft2)
Number
79
41
26
19
13
13
51
44
106
Geometric
Mean
0.32
1.03
215
1.97
3.93
1.87
3.92
499
44.84
-------
Table 8. Window Well Dust-Lead Concentrations Corrected for Bias by Tap Weight for the 284 Privately-
Owned Houses in the HUD National Survey
Tap Weight
Level
0.1 mg
* 0.2 to < 1 mg
1 1 to < 2 mg
t 2 to •< 3 mg
t 3 to •< 4 mg
1 4 to « 5 mg
t 5 to -< 10 mg
t 10 to « 25 mg
>- = 25 mg
Number of
Dust
Samples
718
81
56
55
52
31
179
199
617
Number of
Window Well
Dust Samples
423
3
1
4
3
9
18
107
Dust-Lead Concentration (ug/g)
Number
5
3
.
1
4
3
9
18
106
Geometric
Mean
371.83
91.68
24.27
1208.81
2410.66
2608.08
1229.38
1770.24
Minimum
8465
19.22
.
24.27
233.27
1415.20
546.23
32.33
5.17
Maximum
15915.80
266.69
24.27
3824.84
6736.85
32521.60
60743.81
109165.57
Number
>1000
1
0
0
2
3
8
9
69
Dust-Lead Loading
(Mg/ft2)
Number
14
3
.
1
4
3
9
18
106
Geometric
Mean
0.55
0.70
.
0.37
9.88
25.69
34.54
33.29
242.80
-------
Table 9. Distribution of Wipe Samples Taken in National Survey
LBPJD
0720300
1931906
1820802
1952506
0440602
1830801
1932300
1860204
1861202
1861400
1860501
1860600
1861004
1861509
1860709
1160308
1160506
1160704
1161108
1161207
1660604
1760503
1660703
1760305
1660406
1760206
1660505
Location (Sample ID)
EW
(61)
1
1
1
1
OF
(62)
1
1
1
1
1
WF
(63)
1
1
1
1
WS
164)
1
1
1
1
1
ww
(65)
1
1
1
DS
(66)
1
1
1
1
1
1
DW
(67)
1
1
1
1
CH
(68)
CE
(69)
1
1
1
1
1
1
1
1
1
1
1
1
CF
(70)
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
cs
(71)
1
1
1
1
1
1
1
1
1
cw
(72)
1
1
Number of
Wipe Samples
2
1
3
2
1
4
1
3
3
3
1
2
2
1
3
6
10
1
8
8
1
2
2
3
5
1
2
Public
vs Private
Private
Private
Private
Private
Private
Private
Private
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Total Number of Wipe Samples = 81
EW = Entryway floor sample (there are no window samples from entryways)
WF = Wet Room Floor, WS = Wet Room Window Sill, WW = Wet Room Window Well
DF = Dry Room Floor, DS = Dry Room Window Sill, DW = Dry Room
CH = Common Area Hall, CE = Common Area Entryway
CF = Common Area Floor, CS = Common Room Window Sill, CF = Common Room Floor
-------
HUD National Survey Data - -Private Homes
Plot of Floor Dust Concentration vs Floor Dust Tap Weight
12000;
•3 10000J
1
§ SOOOi
ra
o 6000J
I 4000
to
Q 2000
0.1
Dust Tap Weight (mg)
Figure 6. Dust-Lead Concentration Corrected for Bias Plotted Against Tap Weight for
Samples with Tap Weight Equal to 0.1 mg.
HUD National Survey Data - - Private Homes
Plot of Floor Dust Concentration vs Roor Dust Tap Weight
12000
3 10000
1
§ 8000
p 6000
I 4000
O 2000
OL
2 3
Dust Tap Weight (mg)
Figure 7. Floor Dust-Lead Concentration Corrected for Bias Plotted Against Tap Weight for
Samples with Tap Weights Greater than 0.1 mg and Less than 5 mg. Two
Samples with Dust-Lead Concentrations Greater than 29,000 //g/g are not
Shown in the Plot.
-------
HUD National Survey Data - - Private Homes
Plot of Floor Dust Concentration vs Poor Dust Tap Weight
12000 J
-3 100001
I
8000;
6000
4000
20001
0
7 8
Dust Tap Weight (mg)
10
Figure 8. Floor Dust-Lead Concentration Corrected for Bias Plotted Against Tap Weight for
Samples with Tap Weights Greater than or Equal to 5 mg and less than 10 mg.
One Sample with a Dust-Lead Concentration Greater than 29,000 pg/g is not
Shown in the Plot.
HUD National Survey Data --Private Homes
Plot of Floor Dust Concentration vs Floor Dust Tap Weight
12000
B 10000
6000J
400QJ S*
w '
Q 2000'
1000 2000 3000 4000
Dust Tap Weight (mg)
5000
6000
Figure 9. Floor Dust-Lead Concentration Corrected for Bias Plotted Against Tap Weight for
Samples with Tap Weights Greater than or Equal to 10 mg.
20
-------
The impact of these two samples being wipe samples rather than vacuum dust samples is
negligible on the Section 403 Risk Assessment. Of the two wipe samples, the lead concentration
is missing for one sample and non-missing for the other sample. For the purposes of analyses
conducted to support Section 403, the dust-lead concentration of the one non-missing sample
will be excluded.
There is no reason to suspect the quality of the dust-lead loadings of these two samples.
Therefore, they will be utilized in the analyses conducted for Section 403. However, it may be
necessary to handle the lead loadings of these two samples differently if a conversion factor is
implemented in Section 403 analyses. If the observed lead loadings collected via a Blue Nozzle
vacuum are converted to wipe equivalent lead loadings then the conversion will not be required
for these two samples. Likewise, if the observed lead loadings are compared to a Blue Nozzle
standard equivalent to a wipe standard, then the wipe dust-lead loadings of these two samples
should be compared directly against the standard defined for a wipe sample.
4.0 DISCUSSION, SUMMARY, AND CONCLUSIONS
Dust Samples with a Tap Weight of 0.1 mg. Dust samples with a tap weight of
0.1 mg are a concern. Only 5 of the 40 floor dust samples with a tap weight of 0.1 mg had a non-
missing dust-lead concentration. Both Westat and MRI were contacted in an attempt to
determine what was the basis of the 0.1 mg number and what is the impact of a tap weight of 0.1
mg on the precision and accuracy of the measured dust-lead concentration. While it is generally
believed, both by Westat and MRI, that 0.1 mg represents the limit of detection for the
gravimetric analysis, no one at either Westat or MRI was able to confirm this. MRI stated that
the balances typically employed in their metals laboratory measure sample weights in gram units
to four decimal places. That is, the smallest nonzero weight possible for the balance commonly
used is 0.1 mg. MRI stated that detectable dust-lead amounts can be measured for dust samples
this small. However, if the dust weight is suspect then the dust-lead concentration is suspect.
Due to the large number of missing dust-lead concentrations for samples with a tap weight of 0.1
mg, and the lack of confidence in this tap weight, dust-lead concentrations of the 0.1 mg tap
weight samples will be excluded analyses conducted for Section 403. This decision applies only
21
-------
to the dust-lead concentrations of the 0.1 mg tap weight samples and not to the dust-lead loadings
of these samples.
Adjustments to Dust-Lead Concentrations. Dust-lead concentrations reported in
the HUD National Survey may be biased. Because tap weights were not used to compute dust-
lead loadings, there is no reason to suspect that the dust-lead loadings are biased. It was shown
in Section 2.2 that the HUD National Survey dust-lead concentrations tended to increase with
decreased tap weight. This trend may be due to a bias. Laboratory experiments were conducted
to generate a database for understanding and addressing the bias in dust-lead concentration
computed via tap weight in the HUD National Survey. A regression model, developed for
correcting the bias, was fitted to the laboratory data. It was shown in Section 2.4 that dust-lead
concentrations corrected for the bias via the regression model no longer exhibited the trend of
increased dust-lead concentration with decreased tap weight.
Tap weights generated in the laboratory experiments and used to develop the regression
model ranged from 0.7 mg to 27 mg. Tap weights in the HUD National Survey ranged from 0.1
mg to over 5000 mg. The regression model applied to the HUD National Survey was modified
to predict a ratio of one for tap weights greater than 24.5 mg. However, use of the model to
correct dust-lead concentrations for tap weights less than 0.7 mg may still be of concern. The
correction factor for lower tap weight samples is severe. The dust-lead concentration of a sample
with a tap weight of 0.7 mg is reduced by a factor of 18.7; the factor ranges from approximately 5
for samples with a tap weight of 2 mg to approximately 2 for samples with a tap weight of 3 mg.
Because the regression model was developed from data with tap weights all 2 0.7 mg, its
application might best be limited to samples with tap weights * 0.7 mg.
The correction factor predicted by the regression equation may have over-corrected for
lower tap weight samples. As shown in Table 6, the geometric mean of the corrected dust-lead
concentrations of floor dust samples with a tap weight * 0.2 mg and < 1 mg was 62 ug/g. The
geometric means for groups with tap weights * 2 mg ranged from 137 to 190 ug/g. There are 37
floor dust samples in the HUD National Survey with a tap weight * 0.2 mg and < 1 mg, of which
25 have tap weights < 0.7 mg and 12 have tap weights 2 0.7 mg. The geometric means of the
corrected dust-lead concentrations are 49 and 102 ug/g for the 25 and 12 samples, respectively.
22
-------
Therefore, the tendency for over-correction by the regression model may be strongest for dust
samples with a tap weight less than 0.7 mg.
For the purposes of analyses conducted to support Section 403, dust samples with a tap
weight less than 0.7 mg will be excluded from the analysis, and the analysis will be conducted on
the corrected dust-lead concentrations for all samples with a tap weight * 0.7 mg. There is no
reason to suspect the quality of the dust-lead loadings for the 25 samples with a tap weight * 0.2
mg and < 0.7 mg, and therefore they will be included in Section 403 analyses.
Outliers. A statistical analysis was performed to determine if any of the floor dust-lead
concentrations for the 284 privately-owned residences in the HUD National Survey were
statistical outliers. Three of the 814 floor dust-lead concentrations were determined to be
unusually large. All three samples possessed dust-lead concentrations greater than 100,000 ug/g.
The corrected dust-lead concentrations of the three samples were all greater than 29,000 ug/g.
No basis was uncovered for suspecting the quality of these three samples. Therefore, these
samples will be included in the analyses conducted for Section 403.
Wipe Samples. Westat provided a table identifying the dwellings and locations of wipe
samples. Only two of these samples are being used in analyses conducted for Section 403. The
dust-lead concentration is missing for one sample and will be excluded from Section 403
analyses for the other sample. There is no reason to suspect the quality of the lead loadings of
the two samples. It will not be necessary to convert the lead loadings of these two samples to
wipe equivalent loadings in the analyses conducted for Section 403.
Summary. The following decisions were made regarding the quality of dust samples
collected in the HUD National Survey for the purposes of the Section 403 Risk Assessment:
1. Dust-lead concentrations of dust samples with a tap weight of 0.1 mg will be
excluded from the analysis of dust-lead concentrations. (Five out of 814 floor dust
samples, 27 out of 340 window sill samples, and 5 out of 149 window well samples.)
2. Dust-lead concentrations of dust samples with a tap weight £ 0.2 mg and -< 0.7 mg
will be excluded from the analysis of dust-lead concentrations. (25 out of 814 floor
dust samples, 27 out of 340 window sill samples, and 1 out of 149 window well
samples.)
23
-------
3. Dust-lead concentrations of dust samples with a tap weight greater than or equal to
0.7 mg will be corrected using the regression model presented in Section 2.3. The
corrected dust-lead concentrations are presented in Appendix B-l.
4. The outliers identified in Section 2.2 will be included in the analysis. (Three samples
were determined to be unusually large.)
5. The dust-lead concentrations of wipe samples collected from the privately-owned
residences in the HUD National Survey from floors, window sills, and window wells
will be excluded from Section 403 analyses (1 window well sample and 1 window sill
sample). There is no reason to suspect the quality of the dust-lead loadings of the two
samples. Therefore, they will be utilized in analyses conducted for Section 403.
24
-------
APPENDIX A
LABORATORY ASSESSMENT OF THE AMOUNT
OF RESIDUAL DUST
REMAINING IN BLUE NOZZLE FILTER CASSETTES
May 28, 1996
Prepared by
Midwest Research Institute
425 Volker Boulevard
Kansas City, MO 64110-2299
Battelle Subcontract No. 103639-G002421
MRI Project No. 5021-10
-------
Contents
1 Introduction 1
2 Test Procedures 3
3 Test Results 7
4 Conclusions ^
-------
Section 1
Introduction
The Department of Housing and Urban Development (HUD) National Survey collected dust
samples for Pb analysis using the Blue Nozzle sampler device. This device collects dust in a
plastic cassette containing a piece of filter paper to capture the dust. However, in the HUD
National Survey, the filter cassette was not preweighed before sampling, and only the weight of
dust that could be tapped out of the cassette was determined. The weight of dust remaining on the
filter was not determined. Nevertheless, the total amount of Pb was determined by analysis of
both the "tap" dust and filter cassette, combined. Subsequently, the concentration of Pb in the
dust was calculated from the total amount of Pb found, divided by the weight of tap dust. Since
that calculation necessarily neglected the weight of any dust retained in the filter cassette,
the concentration results were probably biased high, but the extent of the bias is unknown.
However, the bias in the Pb concentration measured by the weight of tap dust does not affect the
measured dust-lead loadings in the HUD National Survey (i.e., ug/ft2).
The purpose of this project was to carry out laboratory experiments using the Blue Nozzle to
determine the weight of dust "tapped out" and the weight of dust remaining in the filter
cassette, and then to use those data to estimate the bias in the Pb concentrations that may have
been present in the HUD National Survey. Subsequent sections of this report describe the
procedures used in the experiments and the results of these experiments, along with the
conclusions from the project.
-------
Section 2
Test Procedures
Different weights of dust, of various particle sizes, were placed on a 12-inch * 12-inch
tile substrate, and the Blue Nozzle sampler was used to vacuum the dust off the substrate. The
filter cassette was preweighed so that the weight of dust remaining in the filter cassette could
be determined, after dust had been tapped out of the cassette. In this way, the weight of dust
tapped out of the cassette was determined, as well as the weight of dust remaining in the filter
cassette. These data were used to calculate the percentage of dust remaining in the filter
cassette, as shown by the equation below.
Wt of dust in filter cassette
Remaining in filter cassette =
(Wt of dust in filter cassette) + (Wt of dust tapped c
The matrix of experiments covered three different particle sizes, and the amounts of dust
tapped out covered three weight ranges. Each combination of particle size/amount was repeated
at least five times (i.e., replicates) because of variability in the results for each
combination, due to the low weight of dust tapped out, as discussed below. Also, the weight of
dust tapped out varied because both the weight of dust applied to the tile substrate and the
amount vacuumed off the substrate varied.
The specific particle sizes and the target amounts of dust tapped out were as follows:
Particle sizes of dust
<53
150-212 pm
250-2000 pm
Target amount of dust tapped out
0.2-5.0 mg
5.0-1 0.0 mg
1 0.0-20.0 mg
-------
The three particle sizes shown above represent the range of sizes used by MRI in a previous
study of house dust, where dust in vacuum cleaner bags was sieved to obtain different size
ranges. These sieved dust samples were still available and were used in this study. Some
preliminary experiments were carried out to determine how much dust must be applied in order to
obtain the target "tapped" amounts shown.
The substrate to which the dust was applied, for vacuuming with the Blue Nozzle, was 12-inch
x 12-inch smooth tile, like that used in the previous MRI study. After each use, these tiles were
wiped off with Kimwipes using distilled water and allowed to dry overnight.
In the initial experiments, it was found that the weight of the filter cassette decreased by
about 10 mg after vacuuming a clean tile (no dust), but after a second vacuuming of the clean
tile, the weight decreased by less than 0.2 mg. Thus the procedure was to first use each cassette
to vacuum a clean tile and then weigh the cassette to determine the tare weight.
It should also be noted that after any vacuuming, the weight of the cassette decreases as it
sits on the balance, but the rate of decrease slows over time. The weight recorded on the data
sheets (i.e., final weight) was taken when the weight did not change more than 0.1 mg over 1 min.
After the initial experiments, the procedure used in all tests was as given below, using the
data entry form shown in Table 2-1.
1. .Tare weigh an appropriate piece of weighing paper.
2. Add desired weight of dust onto weighing paper.
3. Transfer dust from weighing paper onto tile substrate.
4. Reweigh the paper to determine weight of dust actually transferred to tile.
5. Obtain cassette and use it to vacuum a clean tile.
6. Weigh the cassette (not including the top plastic part of the cassette), using this
weight as the tare weight of the cassette.
7. Place the cassette back in the Blue Nozzle sampler and vacuum dust off the tile that was
prepared in Step 3.
8. Tare weigh an appropriate size piece of weighing paper.
9. Tap dust out of the cassette onto tared weighing paper and determine weight of dust
"tapped out."
10. Reweigh the cassette after tapping out dust.
-------
Vacuuming with the Blue Nozzle in each experiment was done in accordance with the protocol
specified in Appendix H of the previous report titled "Laboratory Evaluation of Dust and Dust
Lead Recoveries for Samplers and Vacuum Cleaners," except as given above.
-------
Table 2-1. Data Entry Sheet
Particle Sizes: < 53, 150-212, 250-2000
Target Weight Tapped Out: 0.2-5 mg. 5-10 mg. 10-20 mg
Replicate No.: 1, 2. 3. 4, or 5
Test No..
Date
Time
Operator
A. Dust Applied
Wt of paper plus
dust
9
Wt of paper
after apply dust
Wt of dust
applied
B. Dust Tapped Out (WT)
Wt of paper plus
dust
Tare wt of paper
C. Dust Remaining en Filter (WF)
Wt of-Wter^with ~\fijj,.TTl Tare wt of filter
dust (after -before vacuum
tapping dust out
of cassette)
g - 9
D. Calculation of Percent Remaining jyi Filter
% Remaining =
W
! x loo
g)
Wt of dust
tapped out
g
(WT)
Wt of dust
remaining 0n
filter
( g) + (-
(WF)
x 100 =
g)
Reviewed by
Date
-------
Section 3
Test Results
All of the test results are tabulated in Tables 3-1, 3-2, and 3-3. Each table contains the
results for one of the three target amounts for the weight of dust tapped out. Each of the
tables shows the results for the three particle sizes of dust used in the tests. These tables
express the results in terms of the percent of dust remaining on the filter, and the ratio of
the total weight (weight of dust tapped out plus weight of dust remaining in the cassette) to
the weight of dust tapped out. This ratio is, in effect, a correction factor that might be
applied to a lead concentration that had been calculated using only the weight of Pb tapped
out. That is, the lead concentration should be divided by the ratio value to obtain a
corrected Pb concentration.
It had been anticipated that this ratio might be relatively large when the weight of dust
tapped out was small (i.e., 0.2 to 5.0 mg), but that the ratio would decrease toward a ratio
of 1.0 as the weight of dust tapped out increased. That is, for larger amounts of dust tapped
out, the percent remaining in the filter cassette would be small. To investigate this
expectation, the data in Tables 3-1,3-2, and 3-3 were plotted as shown in Figures 3-1, 3-2,
and 3-3 and were combined in Figure 3-4. A dashed reference line, at a ratio of 1.0, has
been included in Figure 3-4.
Examination of the results given in the figures shows ratios in the range of 1.1 to 2.7
for the two larger amounts of dust tapped out (i.e., 5 to 10 mg and 10 to 27 mg). On the
average, the ratio appears to increase slightly as the amount tapped out decreases toward a
value of about 3 mg. In addition, over this range the ratio does not appear to be
significantly different for the three different particle size ranges.
For an amount tapped out below about 3 mg, the ratio dramatically increases as the
amount tapped out decreases. This certainly provides evidence of a bias in the Pb
concentrations used in the HUD National Survey, where the Pb concentration was
calculated using only the weight of dust tapped out, especially for those instances where
the weight of dust tapped out was low (e.g., 0.2 mg).
-------
Table 3-1. Summa
the ranc
Wt of dust
tapped out
ry of Test Results for Weight of Dust Tapped Out (in
e of 0.2 to 5.0 mg)
Wt of dust
remaining in
filter
cassette
Img)
Particle Size < S3 um
8
10
11
12
13
16
20
24
1.4
0.7
2.4
1.4
3.5
1.3
1.3
3.9
10.8
11.6
2.7
10.0
5.9
15.8
19.4
19.7
Total wt
(mg)
Percent of
dust in filter
cassette
(%)
12.2
12.3
5.1
11.4
9.4
17.1
20.7
23.6
88.5
94.3
52.9
87.7
62.8
92.4
93.7
83.5
Particle Size 150 to 212 um
29
30
31
38
46
4.5
3.2
1.9
0.8
3.2
2.5
4.9
11.4
12.2
2.0
7.0
8.1
13.3
13.0
5.2
35.7
60.5
85.7
93.8
38.5
Ratio of
total wt to
wt tapped
out
8.7
17.6
2.1
8.1
2.7
13.2
15.9
6.1
1.6
2.5
7.0
16.3
1.6
Particle Size 250 to 2000 i/m
47
51
59
60
63
3.6
3.3
4.2
3.4
2.0
2.0
8.0
2.4
4.9
0.9
5.6
11.3
6.6
8.3
2.9
35.7
70.8
36.4
59.0
31.0
1.6
3.4
1.6
2.4
1.5
-------
Table 3-2. Summary of Test Results for Weight of Dust Tapped Out (in
the range of 5.0 to 10.0 mg)
Test No.
Wt of dust
tapped out
(mg)
Particle Size < 53 //m
7
9
15
19
22
27
7.8
5.2
5.7
8.9
7.4
9.9
Wt of dust
remaining in
filter
cassette
(mg)
Total wt
(mg)
5.4
3.4
9.9
13.1
7.3
3.6
13.2
8.6
15.6
22.0
14.7
13.5
Percent of
dust in filter
cassette
(%)
40.9
39.5
63.5
59.5
49.7
26.7
Ratio of
total wt to
wt tapped
out
1.7
1.7
2.7
2.5
2.0
1.4
Particle Size 150 to 212 urn
34
35
36
39
42
43
6.5
8.8
9.5
6.1
8.7
8.8
2.1
2.7
4.0
6.8
2.3
9.7
8.6
11.5
13.5
12.9
11.0
18.5
24.4
23.5
29.6
52.7
20.9
52.4
1.3
1.3
1.4
2.1
1.3
2.1
Particle Size 250 to 2000 /im
48
49
62
64
65
8.0
7.8
6.6
6.3
5.1
4.2
2.5
1.6
8.6
3.0
12.2
10.3
8.2
14.9
8.1
34.4
24.3
19.5
57.7
37.0
1.5
1.3
1.2
2.4
1.6
-------
Table 3-3. Summa
the ran<
Test No.
Wt of dust
tapped out
(mg)
ry of Test Results for Weight of Dust Tapped Out (in
eof 10.0 to 30.0 mg)
Wt of dust
remaining in
filter
cassette
(mg)
Total wt
(mg)
Percent of
dust in filter
cassette
{%)
Particle Size < 53 urn
14
17
18
21
23
25
26
28
13.3
11.5
24.2
12.0
24.3
10.1
24.7
14.1
2.7
2.8
8.1
6.6
8.7
14.0
8.6
6.8
16.0
14.3
32.3
18.6
33.0
24.1
33.3
20.9
16.9
19.6
25.1
35.5
26.4
58.1
25.8
32.5
Ratio of
total wt to
wt tapped
out
1.2
1.2
1.3
1.6
1.4
2.4
1.3
1.5
Particle Size 150 to 212 urn
32
33
37
40
41
44
45
13.3
13.1
22.2
13.9
21.8
17.2
16.2
3.1
4.8
3.8
2.8
3.9
4.8
8.7
16.4
17.9
26.0
16.7
25.7
22.0
24.9
18.9
26.8
14.6
16.8
15.2
21.8
34.9
1.2
1.4
1.2
1.2
1.2
1.3
1.5
Particle Size 250 to 2000 //m
50 '
52
53
54
55
56
57
58
61
16.7
24.3
13.1
27.2
27.2
15.4
18.9
16.4
11.5
15.2
9.5
11.1
9.5
11.8
3.2
8.5
10.2
0.7
31.9
33.8
24.2
36.7
39.0
18.6
27.4
26.6
12.2
47.6
28.1
45.9
25.9
30.3
17.2
31.0
38.3
5.7
1.9
1.4
1.8
1.3
1.4
1.2
1.4
1.6
1.1
10
-------
20
CD
u 15
3
0
T3
0)
Q.
Q.
ro
O
r 10
jr
en
0)
5
ra
o
M —
O
•° 5
ro
0^
0
I
•
s
•
—
V
• T
T • V T •
i i i i
31234
Weight of dust tapped out (mg)
• Particle size <53 pm « Particle size 150 um to 212 pm v Particle size 250 pm to 2,000 pm
j
5
Figure 3-1. Ratio of Total Wt of Dust to Wt of Dust Tapped Out as a Function of the Wt of Dust Tapped Out over
the Range of 0.2 to 5.0 mg
-------
5
en
0)
o
~a
a. 3
D-
ro
o
en
5 2
ro
o
'o
o
To
QC 1
0
-
"
mm
T
•
•T ' •* '
1 1 1 1
5 6 7 8 9 10
Weight of dust tapped out (mg)
• Particle size <53 pm • Particle size 150 pm to 212 pm v Particle size 250 pm to 2.000 pm
Figure 3-2. Ratio of Total Wl of Dust to Wt of Dust Tapped Out as a Function of the VVt of Dust Tapped Out over
the Range of 5.0 to 10.0 mg
-------
5
£ 4
en
0)
o
X)
a. 3
Q.
to
o
JC
O)
5 2
ro
o
o
o
(O
cr 1
i
i
1
I
—
v
T
• * Y _
*Jk • £ •
T ••
0 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Weight of dust tapped out (mg)
• Particle size <53 urn * Particle size 150 urn to 212 urn y Particle size 250 urn to 2,000 urn
Figure 3-3. Ratio of Total Wt of Dust to Wt of Dust Tapped Out as a Function of the Wt of Dust Tapped Out over
the Range of 10.0 to 30.0 mg
-------
20
O)
I
"3
o
Q)
Q.
Q.
TO
O
0)
'cu
2
'o
o
.o 5
ro
cr
10
Fitted regression:
- IfWt,.,,,^^ < 2.94 mg:
Ratio = 30.63 - 19.38*Wluw-daill + 3.29*Wt2,.pp^ „„,
— Otherwise:
Ratio = 2.20-0.049*Wll.pp^ou,
I I I
J L
0
10 15 20
Weight of dust tapped out (mg)
25
• Particle size <53 \im • Particle size 150 pm to 212 pm
T Particle size 250 pm to 2.000 |jm — Fitted regression
Figure 3-4. Ratio of Total Wt of Dust to Wt of Dust Tapped Out as a Function of the Wt of Dust Tapped Out over
the Entire Range of 0.2 to 30.0 mg
-------
If an attempt were to be made to correct the Pb concentrations from the HUD National Survey.
the results shown in the figures could be used for that purpose. That is, the reported Pb
concentration could be divided by the value of the ratio that corresponds to the specific weight
of dust tapped out. This calculation is illustrated by the equation below:
. „ Original Pb Concentration
Corrected Pb Cone = =
Value of Ratio
where: Value of Ratio is determined based on the weight of dust tapped out, which was used in the
original calculation of the Pb concentration, as discussed below.
To provide a method for determination of the ratio value that could be used to correct Pb
concentrations calculated in the HUD National Survey, a regression analysis was performed on the
data from these experiments (i.e., Tables 3-1,3-2, and 3-3). All but one data point were used in
the analysis. The result from Test No. 24 (Table 3-1) was excluded based on visual inspection of
the plot. A segmented linear model was fit to all the other data using PROC NLIN of SAS, a
statistical software package. The parameters of a second-order model followed by a first-order
model were estimated as well as the value of weight of dust tapped out at which the regression
line changes shape. Table 3-4 summarizes the regression results. The following two equations
provide the best fit to the data:
For Weight of Dust Tapped Out (Wt in mg)) below 2.94 mg:
Ratio = 30.63-19.38 (Wt) + 3.29 (Wt)2
For Weight of Dust Tapped Out (Wt in mg) above 2.94 mg:
Ratio = 2.20-0.049 (Wt)
15
-------
Tap dust range
0 to 2.94 mg
2.95 to 24 5 mg
Model parameters*
Intercept (se)
95% Cl
30.63(2.01)
[26.61,3465]
2.20 (NC)e
[1.76, 2.36]d
Linear term (se)
95% Cl
-19.38(2.29)
[-23 98, -14.79]
-0.049(0.021)
[-0.091, -0.006]
Quadratic term (se)
95% Cl
3.29 (0.56)
[2.16.441]
NA
Model summary"
R2 = 87.8%
MSE = 1.13
se = standard error of parameter estimate
95% Cl = 95% confidence interval of parameter estimate
Revalue = percent variance in data explained by model (adjusted for number of
parameters)
MSE = Mean Square Error of estimated ratio
NC: not calculated
95% Cl for ratio of 2.06 calculated at Weight,,^ of 2.94
For any given weight of dust tapped out in the HUD National Survey data, the appropriate
equation above could be used to calculate the ratio value that should be used to correct the Pb
concentration that was calculated in the HUD National Survey using that weight of dust tapped
out. Use of the equations for that purpose does, of course, involve some uncertainty. This is
because the equations are based on the data from these laboratory experiments, which utilized
sieved dust that may not duplicate the characteristics of the actual dust sampled in the HUD
National Survey. Although differences in dust characteristics could affect the ratio values, it
seems unlikely that any differences would have much effect on the equations developed from the
laboratory experiments.
Based on the above equations, the corrected Pb concentrations may decrease by a factor
ranging from 2.1 to 1.0 for the larger weights of dust tapped out (above 2.94 mg). But for
weights of dust tapped out below 2.94 mg, the corrected Pb concentrations would decrease by a
factor that might range from 2.1 (for 2.94 mg tapped out) up to 21.7 (for 0.5 mg tapped out).
It should be noted that in these experiments, the lowest amount of dust tapped out was
0.7 mg, and it is very difficult to obtain such low amounts of dust tapped out. In this report,
the above equation yielded a ratio of 18.7 for 0.7 mg tapped out and a maximum ratio of 30.6 for
0.0 mg tapped out. Obviously, it would be possible theoretically for no dust to be tapped out
(0.0 mg) when there was some weight of dust in the cassettes, so the ratio could be infinitely
high. Thus, the use of the second equation above is only a best fit of the data down to a weight
16
-------
tapped out of 0.7 mg. Use of the equation for lesser amounts (e.g., < 0.7 mg tapped out) is
uncertain, but it should provide conservatively low ratios for the lesser amounts.
17
-------
18
-------
Section 4
Conclusions
From the results discussed in the previous section, it can be concluded that the total
weight of dust collected in Blue Nozzle cassettes is higher than the weight of dust that can be
tapped out the cassette, especially for smaller weights of dust tapped out. It can, therefore,
also be concluded that Pb concentrations in dust, determined using only the weight of dust tapped
out, are biased high and that the extent of the bias is likely quite high for very small weights
of dust tapped out.
19
-------
APPENDIX B
DATA TABLES FOR DUST-LEAD CONCENTRATIONS
FOR FLOOR, WINDOW SILL, AND WINDOW WELL SAMPLES IN THE
284 PRIVATELY-OWNED HOUSES IN THE HUD NATIONAL SURVEY
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
0120105
0130708
0130906
0131003
0131102
0131201
0150102
0150201
0211102
0221101
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(Rig)
30.4
26.7
252.2
36.2
14.9
0.1
0.1
17.4
73.0
9.3
70.1
0.4
0.1
0.1
51.5
181.5
41.9
0.1
16.1
0.1
0.1
64.1
130.7
21.0
0.1
10.2
0.1
0.1
40.2
20.5
1.2
9.2
0.1
0.1
0.1
96.2
94.8
3.2
7.4
31.0
0.1
0.1
66.8
38.5
68.5
13.6
0.1
0.1
0.1
769.4
139.2
135.0
258.1
0.1
0.1
0.1
44.9
14.8
8.7
0.1
0.1
5.9
0.1
9.1
1.0
2.5
0.1
Surface
Sampled
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(pg/g)
251.05
153.26
95.77
192.39
826.13
166.77
84.74
91.00
203.31
908.15
128.87
53.04
59.09
138.13
142.65
98.04
104.70
289.38
300.83
175.51
354774.13
1042.42
210.23
77.69
262.35
2990.49
341.62
69.36
97.95
354.32
133.65
82.43
39.80
21.31
16.62
1119.92
55.93
524.47
11048.08
341.10
210.04
219.35
Revised Dust-Lead
Concentration
(ug/g)
251.05
153.26
95.77
192.39
562.03
123.77
84.74
52.17
203.31
128.87
53.04
59.09
97.89
142.65
98.04
89.41
170.20
300.83
146.81
29293.22
595.93
210.23
77.69
128.40
162-7.56
341.62
69.36
97.95
354.32
87.15
82.43
39.80
21.31
16.62
1119.92
37.92
295.69
5781.61
194.46
14.44
79.97
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
- ID
0221101
0221507
0250902
0251900
0252404
0310102
0310201
031060-7
0310706
0311100
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
0.1
0.1
0.1
34,
34,
12,
64,
0,
•71.
88.
4.
202.
10.
0.
261,
47,
2,
9.0
0.1
5.5
41.2
2.6
4.2
0.1
0.1
0.1
8.6
19.4
6.7
60.5
7.0
0.1
0.1
0.1
55.3
26.0
43.3
5.4
6.1
0.1
0.1
74.6
3.6
33.7
29.1
2.5
0.1
0.1
44.1
28.5
8.3
91.1
72.5
339.2
0.1
7.1
4.9
1.5
9.2
5.8
193.7
170.8
2.6
Surface
Sampled
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
2855.66
1118.72
973.46
971.47
611.46
1937.84
201.06
219.30
585.62
1065.19
2237.48
924.06
305.06
127.31
200.28
254.19
372.70
112.49
160.99
494.75
367.40
510.78
138.49
119.24
193.52
218.81
166.43
3.45
30.70
248.89
1655.46
92.44
26.50
77.29
1067.60
543.73
2287.72
702.40
43.08
393.17
2001.84
2693.69
2174.02
6421.20
1880.72
135.39
1136.15
1048.76
Revised Dust-Lead
Concentration
I ug/g)
1118.72
973.46
608.18
347.62
1937.84
104.15
219.30
235.93
534.14
519.54
244.16
68.02
200.28
136.88
372.70
112.49
80.73
494.75
217.97
510.78
138.49
119.24
99.99
115.09
166.43
1.70
30.70
248.89
603.57
92.44
26.50
43.10
1067.60
543.73
2287.72
379.24
21.98
43.87
1144.42
1406.04
2174.02
6421.20
1860.72
135.39
370.26
422.52
0.1
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
0311209
0320101
0320408
0320507
0320705
0321307
0330308
0331009
0340406
0340505
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
25.8
22.0
11.3
101.9
0.1
0.1
0.1
12.4
28.1
8.0
2.5
0.1
0.1
0.1
29.9
8.6
23.3
0.1
0.1
0.1
97.0
11.9
64.0
57.9
9.4
0.1
0.1
72.2
23.9
6.1
4.1
0.1
0.1
0.1
0.1
61.1
5.1
3.9
0.1
0.1
206.9
0.1
52.4
21.2
28.8
14.4
0.1
0.1
0.1
1.9
5.9
280.5
20.5
0.1
0.1
0.1
27.6
57.8
0.6
10.2
4.9
0.1
0.1
2041.7
18.6
25.4
6.4
Surface
Sampled
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
482.69
236.25
351.22
2465.96
917.84
294.57
258.67
827.73
456.61
214.32
223.06
13762.54
1046.50
62.01
585.23
1324.82
780.42
521.06
239.02
355.62
3928.96
1905.00
179.50
24.55
151.22
250.04
98.73
12.10
7.33
71.85
544.56
175.37
25.82
201.89
149.95
17.90
2069.32
101.44
42.86
1.01
55.63
1181.31
97.00
Revised Dust-Lead
Concentration
(wg/g)
482.69
210.56
213.34
2465.96
576.38
294.57
143.07
301.79
456.61
120.50
210.78
13762.54
647.23
62.01
585.23
761.67
780.42
506.41
125.73
177.89
179.50
12.59
75.27
250.04
98.73
10.42
7.33
48.08
95.79
91.77
25.82
168.88
149.95
17.90
59.66
21.87
1.01
43.17
1181.31
51.42
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
0340505
0340802
0341107
0341404
0350306
0350B01
0351205
0410100
0410605
0411207
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(rag)
1.0
0.9
0.1
89.3
20.2
6.3
4.7
1.4
0.1
0.1
124,
39,
58,
10,
3,
0,
0.1
224.9
2.5
6.0
97.6
0.1
0.1
249.2
0.1
98.9
33.4
36.4
29.5
94.2
0.1
0.1
139.2
2077.6
118.9
57.8
2.6
0.1
0.1
Surface
Sampled
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
1.3
21.0
9.1
12.6
0.1
0.1
0.1
28.7
6.8
0.1
5.7
2.0
0.1
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
218.9
162.2
18.4
229.7
120.8
0.1
0.1
606.2
163.4
12.7
0.7
0.1
111.2
0.1
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(pg/g)
115.88
304.65
829.00
72.18
69.32
553.50
1041.46
31.49
18.38
90.67
80.36
2312.77
32.20
248.32
172.44
137.81
203.44
1121.57
69.43
436.13
412.49
2497.34
943.77
2184.60
247.14
784.26
801.40
407.98
26521.10
271.54
1S6.09
4.48
104.42
45646.82
5173.31
70.90
44.65
112.46
400.89
745.16
2133.50
721.86
814.69
7390.44
2605.26
Revised Dust-Lead
Concentration
(ug/g)
7.97
19.22
829.00
59.64
36.65
281.01
104.70
31.49
18.38
90.67
47.40
1137.39
32.20
90.54
90.47
137.81
203.44
102.00
59.29
248.63
260.64
2497.34
505.56
1137.50
49.13
784.26
801.40
407.98
26521.10
271.54
156.09
4.48
104.42
45646.82
2084.21
70.90
44.65
86.62
400.89
745.16
2133.50
721.86
516.37
395.72
2605.26
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 264 Privately-Owned
Houses in the HUD National Survey
House
ID
0411306
0411603
0420901
0421206
0430108
0430207
0430306
0430702
0440107
0440305
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
426.2
403.5
86.3
175.4
12.5
0.1
0.1
58.5
54.2
3.5
16.2
0.1
0.1
0.1
135.4
1131.0
831.1
960.4
0.1
0.1
0.1
215.5
327.0
48.6
198.5
0.1
0.1
0.1
69.5
18.3
95.6
34.6
8.7
0.1
0.1
324.2
200.5
91.4
81.3
0.1
0.1
0.1
4210.2
3580.1
3820.2
5269.8
0.1
0.1
0.1
111.4
60.4
73.5
0.1
0.1
0.1
0.1
139.7
53.2
102.5
60.9
47.3
0.1
0.1
125.2
121.8
18.1
5.3
Surface
Sampled
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Original Dust-Lead
Concentration
(ug/g)
1767.13
846.25
447.53
7011.68
513.41
300.67
133.63
93.71
682.19
213.96
4.57
9.96
286.57
1399.77
67.31
1159.38
331.39
20.98
79.67
105.11
150.21
456.18
62.98
82.57
113.20
101.81
2.21
0.58
18.42
0.79
631.57
5.69
98.54
525.85
272.28
60.57
407.75
153.12
347.09
169.90
57.16
195.22
Revised Dust-Lead
Concentration
(ug/g)
1767.13
846.25
447.53
7011.68
323.41
300.67
133.63
46.20
521.49
213.96
4.57
9.96
286.57
1399.77
67.31
1159.38
331.39
20.98
61.13
105.11
150.21
257.20
82.98
82.57
113.20
101.81
2.21
0.58
18.42
0.79
631.57
5.69
98.54
525.85
272.28
60.57
407.75
153.12
347.09
169.90
43.53
100.61
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
0440305
0440602
0441105
0441204
0520106
0520403
0520700
0520809
0520908
0530105
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
2.0
0.1
0.1
60.9
3.8
75.2
5289.9
1.5
0.1
0.1
86.6
16.3
75.7
13.5
0.1
0.1
0.1
301.2
132.1
210.2
6.0
591.4
0.1
0.1
25.3
3.1
77.9
2.0
0.1
0.1
0.1
0.1
96.8
214.5
35.9
0.2
1.2
38.6
79.3
371.9
25.1
145.3
47.4
0.4
41.1
65.4
4.8
22.6
27.0
66.5
0.1
0.1
0.1
26.6
55.3
32.5
29.8
0.9
0.1
0.1
Surface
Sampled
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
1023.7
1001.1
51.8
177.4
114.8
0.1
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
517.33
220.86
272.28
82.55
1379.55
102.10
167.29
36.02
293.98
144.28
125.32
255.96
172.44
230.94
257.64
330.42
28.69
770.82
303.21
301.79
171.78
332.44
214.50
716.14
1152.84
256.50
5845835.49
12674.60
33237.82
26355.81
734.47
158.29
118.92
19863.75
419.04
5286.59
7245.79
269.44
148.33
1130.46
318.96
291.73
114.13
432.97
503.44
1655.46
Revised Dust-Lead
Concentration
(ug/g)
102.85
220.86
135.21
82.55
153.92
102.10
119.38
36.02
191.08
144.28
125.32
255.96
90.47
230.94
257.64
161.33
28.69
153.25
303.21
301.79
171.78
332.44
214.50
716.14
1152.84
256.50
1046.52
33237.82
26355.81
734.47
158.29
118.92
19863.75
5286.59
7245.79
137.14
135.76
1130.46
318.96
291.73
114.13
432.97
503.44
104.43
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
OS30600
0531301
0531400
0540203
0541201
0541300
0612002
0621607
0631408
0651901
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
27.2
18.6
0.1
0.1
0.1
0.1
350.7
355.1
109.5
125.1
39.6
0.1
23.0
64.1
80.6
173.0
5.8
0.1
0.1
13.3
2.8
0.3
19.2
0.5
0.1
0.1
15.9
55.2
6.5
0.1
0.1
0.1
0.1
25.8
19.4
1.5
46.6
0.1
7.9
5.9
0.2
68.0
Surface
Sampled
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
242.1
1035.5
398.2
85.2
71.5
0.1
0.1
212.5
82.7
367.6
0.1
0.1
0.1
0.1
3.4
120.3
0.2
0.1
5.1
0.1
0.1
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
285.29
75.65
22.42
201.05
68.03
504.51
182.89
154.28
142.88
137.71
459.04
196.80
143.64
41.04
15.76
636.01
14.02
662.18
124.16
1095.52
899.71
74.90
73.17
406.69
185.53
1244.71
942.28
896.71
307.17
715.99
135.35
56.23
101.87
1355.41
320.83
129.60
244.18
273.10
151.92
161.34
931.20
488.42
Revised Dust-Lead
Concentration
285.29
58.71
22.42
201.05
68.03
504.51
182.89
154.28
142.88
137.71
459.04
196.80
143.64
41.04
15.76
312.78
14.02
561.78
838.46
74.90
73.17
406.69
96.84
803.91
436.34
243.94
95.26
56.23
54.14
320.83
103.73
27.24
273.10
83.80
84.43
488.42
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
0651901
0710103
0711002
0720300
0720706
0721001
0730606
0750406
0820506
0821009
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Hell
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
26.5
0.1
0.1
10.2
16.0
12.4
0.1
0.1
0.1
0.1
74.2
19.5
63.6
9.6
0.
222.
0.1
10.3
54.9
6.4
41.4
25.4
0.1
0.1
193.1
87.8
4.8
0.1
0.1
0.1
0.1
61.4
247.9
26.8
1948.0
116.7
2061.8
257.3
48.6
1.4
358.5
0.1
7.6
94.4
0.1
23.2
41.9
55.7
8.1
1.2
0.1
57.5
1.3
14.9
2.5
1.7
0.1
10.8
0.1
Surface
Sampled
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Original Dust-Lead
Concentration
45.7
53.7
7.1
0.1
47.0
0.1
0.1
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
234.26
695.91
247.28
310.11
1017.93
134.24
107.37
418.18
6693.23
492.05
243.08
181.59
8709!si
1209.06
517.84
745.08
3208.34
843.12
359.00
174.41
3772.20
559.46
371.46
1891.73
4674.27
101958.90
19620.37
27746.45
374.42
4331.75
707.46
690^31
1593.61
459.07
63.88
45.13
574.31
2105.13
1615.03
1654.26
150.10
179.33
4373.86
Revised Dust-Lead
Concentration
(ug/g)
234.26
409.31
174.63
194.74
1017.93
107.87
107.37
241.78
6693.23
492.05
243.08
98.04
8709.51
713.19
517.84
394.97
3208.34
843.12
359.00
174.41
1919.90
559.46
371.46
1891.73
4674.27
101958.90
19620.37
27746.45
374.42
435.50
707.46
377.71
1593.61
431.79
63.88
45.13
318.51
173.82
1615.03
150.44
102.12
65.38
608.18
525.12
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
0840702
0911404
0911503
0911800
0920801
0920900
0921304
0930701
0940700
0940809
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Hell
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
61.9
26.9
23.4
0.1
0.2
0.1
0.1
1.9
6.4
1.9
0.1
0.2
0.1
0.1
99.7
0.1
322.0
5.3
0.1
0.1
0.1
6.9
10.4
15.5
121.7
0.1
0.1
0.1
0.1
0.4
38.9
0.1
0.1
0.1
0.1
1.9
5.1
32.9
0.1
0.1
0.1
0.1
0.2
8.0
1.6
0.1
0.1
0.1
0.1
0.6
2.6
16.3
0.1
1.6
0.1
0.1
0.1
12.8
46.1
0.1
0.1
0.1
0.1
10.9
8.9
9.3
22.6
Surface
Sampled
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
82.74
116.45
42.28
1552.11
519.33
18*7.16
740.14
687.57
230.32
26-)! 93
1918.65
169.80
183.35
132.04
111.20
5463.70
1070.87
80.06
1303.10
211.66
197.65
3879.98
247.03
1183.39
2224.52
306.42
603.02
472.06
175.71
30.50
443.11
1311.54
714.92
582.88
Revised Dust-Lead
Concentration
(ug/g)
82.74
116.45
40.14
91.35
99.21
130.19
230.32
267.93
988.85
91.20
108.47
91.66
111.20
80.06
229.22
108.54
197.65
136.63
147.10
123.45
430.33
58.68
111.72
30.50
265.99
743.55
409.86
533.49
-------
Table B-l. Original and Revised Dust-Lead Concentrations Cor All Floor,
Window Sill, and Window Well Samples in the 2B4 Privately-Owned
Houses in the HUD National Survey
House
ID
0940809
0941005
0950402
0951004
1010503
1010909
1011303
1011501
1011600
1011709
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
0.1
0.1
0.1
0.1
19.6
6.9
27.3
0.1
0.1
0.1
34.9
31.7
8.6
90.9
60.6
201.5
0.1
0.3
25.3
5.9
35.9
1.3
0.1
0.1
306.4
210.9
29.8
4.6
16.8
0.1
0.1
53.5
5.1
1.7
2.3
0.1
0.1
0.1
60.8
59.1
19.0
0.1
0.1
0.1
6.9
0.
41.
13.
0.
9.
0.1
0.1
0.4
5.0
0.8
0.1
0.1
0.1
0.1
1.8
0.3
Surface
Sampled
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Original Dust-Lead
Concentration
(pg/gl
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
354.72
241.17
4650.27
950.53
370.12
726.42
5429.96
502.42
1599.94
568.04
672.68
1950.53
3185.73
1767.42
1352.56
858.35
4526.01
2510.18
1339.70
1019.10
1603.98
3896.81
27267.66
1735.78
367.65
1688.13
2249.26
18563.00
587.71
1068.37
1862.14
5173.31
1034.66
3879.98
3448.87
3448.87
985.39
Revised Dust-Lead
Concentration
(ug/g)
286.16
129.53
4650.27
950.53
370.12
408.42
5429.96
502.42
568.04
352.02
1950.53
289.71
1767.42
1352.56
858.35
2292.12
1823.17
1339.70
522.59
223.03
1126.36
1735.78
367.65
1330.30
1208.05
1077.26
587.71
703.38
1085.87
529.24
225.17
538.38
221.90
0.1
-------
House
ID
1020205
1020304
1020403
1020502
1020700
1020809
1021005
1030204
1040500
1041607
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
0.5
21.4
1.4
5.0
0.1
0.1
0.1
6.3
2.5
10.5
0.1
0.1
0.1
0.1
10.4
10.4
9.0
0.1
0.1
0.1
0.1
15.
20.
1.0
3.5
0.1
0.4
2.2
0.6
0.1
0.1
0.1
0.1
6.5
0.1
3.4
0.3
0.1
0.1
0.1
0.5
0.1
0.8
0.3
0.1
0.1
0.1
6.8
14.5
19.2
0.1
0.1
0.1
0.1
38.0
2.8
2.0
10.3
0.1
0.1
0.1
0.1
0.9
4999.8
0.1
Surface
Sampled
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
2916.08
202.73 -
216.25
793.76
231.44
264.96
330.86
262.19
140.20
162.00
2237.48
326.88
132.37
1458.04
2526.25
718.27
205.10
630.20
610.58
428.84
9089.22
1543.73
4656.28
3408.46
4860.14
583.65
100.55
75.94
200.84
973.84
254.26
264.73
3448.87
0.05
Revised Dust-Lead
Concentration
(ug/g)
176.07
21.74
406.02
122.37
96.60
196.30
155.11
82.94
92.10
230.06
111.17
100.27
1245.38
52.35
324.52
210.90
197.80
312.65
67.51
60.31
200.84
450.96
50.55
156.16
217.56
0.05
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
1041607
1050509
1050608
1051200
1051408
1120401
1121300
1130806
1140508
1150200
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
0.1
0.1
0.1
0.5
4.1
0.1
6.2
2.2
0.2
0.1
0.1
0.1
0.1
2.2
3.2
2.4
0.1
3043.8
0.1
0.1
2.6
7.8
0.1
0.1
0.1
0.1
0.1
230.5
464.5
32.2
6.5
8.8
38.8
110.0
6.8
15.8
22.6
5.4
0.1
4.4
0.1
1.0
52.8
73.2
11.7
0.8
0.1
2.0
74.4
9.4
2.2
3.9
0.7
4.5
5.4
34.6
3.5
9.1
33.6
Surface
Sampled
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
2069.32
108.01
83.89
333.76
66.78
631.14
470.30
323.33
194.43
1.02
795.89
132.65
260.35
242.79
234.57
581.00
3.06
1653.32
434.56
483.86
65.48
17.31
118.79
2915.86
351.78
272.38
250.18
106.12
15.52
122.09
890.03
77.05
24.93
2918.28
235.02
13335.63
7185.15
319.97
31.93
54.80
705.17
Revised Dust-Lead
Concentration
(ug/g)
54.03
41.56
176.02
17.05
120.05
158.25
63.36
l!o2
320.65
72.97
260.35
242.79
234.57
308.79
1.73
1653.32
434.56
259.19
45.93
15.84
61.38
1469.39
24.19
272.38
250.18
65.24
0.90
24.27
890.03
44.30
6.36
1452.67
12.58
6736.85
3712.46
319.97
15.74
31.24
705.17
0.1
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
1150705
1210806
1221902
1241801
1250406
125110-7
1251404
1311505
1312701
1312800
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
17.5
22.1
9.5
57.2
0.1
0.1
0.1
0.1
16.3
8.6
3.7
1.4
0.1
75.5
57.0
0.3
2.1
0.1
12.8
0.1
1.8
0.2
0.4
0.1
0.6
0.1
0.1
3.0
302.3
8.8
0.3
Surface
Sampled
Carpeted
Uncarpeted
53.3
13.0
3.8
0.1
2.4
0.1
0.1
36.7
40.7
0.1
89.4
16.9
524.0
0.1
2.0
13.5
14.8
0.1
47.8
103.8
0.1
15.7
5.1
7.8
2.0
2.0
34.6
18.1
141.0
2.3
25.2
4.8
0.1
0.1
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
lug/g)
253.64
266.86
77.33
1646.05
1069.53
1291.93
422.01
1895.25
820.40
11287.21
67125.67
759.16
13031.99
615.62
137.95
206.23
3000.52
18398.79
1624.76
3756.41
3002.55
4337.62
2043.46
7501.29
16446.93
27038.38
601.72
1228.10
8375.83
3621.31
1174.31
878.26
559.28
6947.01
37454.73
25626.71
15973.72
7320.24
367.56
188.32
2431.64
398.52
282.59
141.29
23437.42
94.20
988.06
254.63
69.76
188.39
Revised Dust-Lead
Concentration
(ug/g)
188.93
238.88
44.58
1646.05
1069.53
826.57
209.56
617.64
820.40
11287.21
67125.67
553.38
13031.99
122.39
89.67
139.84
18398.79
1624.76
2625.57
1539.69
2386.21
406.26
1491.34
16446.93
20591.86
601.72
354.98
8375.83
1843.10
838.01
493.79
277.05
698.43
25626.71
15973.72
82.77
119.74
62.21
481.28
254.63
39.44
-------
Table B-l. Original and Revised Dust-Lead Concentrations Cor All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
1312800
1322601
1323609
1332402
1333806
1352608
1352806
1353309
1353-705
1410406
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
19.7
11.9
29.4
3.8
0.1
18.6
0.1
0.1
Surface
Sampled
2.0
48.0
0.7
0.1
2.3
0.1
0.1
0.1
0.1
0.1
13.1
5.7
0.1
0.1
0.1
0.1
0.1
2.4
84.2
3.1
4.4
4.0
3.3
0.1
1.1
232.1
7.9
1.1
0.3
0.1
0.1
101.4
204.8
147.2
0.1
0.1
0.1
0.1
13.9
43.9
17.3
40.6
7.7
145.0
23.6
5.6
3.8
28.7
1.5
1.3
0.1
0.1
5.2
0.1
4.9
0.1
0.1
0.1
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
268.43
635.54
4980.61
13023.23
359.14
5221.16
4656.28
218.84
61.11
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
698.11
274.13
770.27
833.19
236.70
7796.16
1457.93
271.93
3313.54
2596.06
2379.72
833.14
4008.50
218.72
253.94
85.90
85.18
56.49
7.34
615.29
33.74
147.02
14.87
6.34
37.68
238.79
1048.89
6899.22
860.28
17808.45
419.95
93.96
1045.55
235.30
Revised Dust-Lead
Concentration
(ug/g)
53.37
635.54
266.69
103.81
140.46
31.81
227.70
274.13
376.09
419.87
118.11
3824.84
109.68
271.93
1827.75
195.31
833.14
4008.50
218.72
167.19
85.90
62.99
56.49
4.03
615.29
32.33
76.35
7.39
6.34
4.20
21.72
539.22
438.94
14423.71
259.73
93.96
519.19
182.60
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
1411909
1440205
1441005
1441302
1450907
1510403
1510908
1520204
1521400
1521509
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well '
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
236.7
17.7
29.0
5.9
2.4
76.2
165.4
214.8
5.7
54.0
13.2
0.1
32.2
0.1
6.3
6.1
57.3
6.3
0.1
0.1
0.1
3.8
8.0
0.3
0.1
0.1
0.1
0.1
1298.7
144.1
356.2
4.2
0.1
0.1
0.1
21.4
5.8
20.3
0.1
0.1
177.4
0.1
163.0
0.1
15.9
4.8
0.8
24.6
0.1
65.8
6.5
3.6
0.1
0.4
131.9
152.3
75.2
48.8
461.7
81.3
105.8
20.2
0.1
15.3
0.1
117.8
0.1
Surface
Sampled
Carpeted
Carpeted
Carpeted
Unearpeted
Uncarpeted
Unearpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Original Dust-Lead
Concentration
(ug/g)
5156.01
159.58
196.23
1806.2-7
698.40
13985.58
1163.52
250.48
203.30
191.60
1199.27
848.29
231.57
111.95
• 38.82
54.20
402.97
75.01
41.39
506.98
82.05
31.74
58.97
83.02
264.55
228.93
361.22
63.80
200.02
88.15
153.39
456.46
90.48
252.51
108.02
183.15
2917.36
2030.78
1581.45
245.15
205.91
157.79
387.63
320.27
557.03
892.33
368.03
Revised Dust-Lead
Concentration
Iug/gI
5158.01
119.74
196.23
945.25
227.60
13985.58
1163.52
250.48
105.85
191.60
772.14
848.29
122.44
58.89
38.82
28.66
200.11
41.49
82.05
31.74
58.97
41.63
229.76
119.50
299.70
63.80
200.02
62.04
78.07
26.49
90.48
252.51
57.41
90.51
2030.78
1581.45
245.15
205.91
157.79
387.63
320.27
460.27
615.28
368.03
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
1521509
1530104
1530302
1530SOO
1530807
1531201
1531300
1531607
1531706
1540202
Sampling
Location
Wet Sill
Dry Well
wet well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
0.1
155.3
0.1
55.
32.
53.2
66,
0,
0,
0.1
16,
29,
20,
28.0
230,
456.
0.1
8.4
19.6
2.6
72.8
0.1
0.1
155.5
Surface
Sampled
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
102.1
107.8
11.5
45.6
0.1
155.5
4.0
369.7
0.1
144.5
56.9
0.1
899.3
0.1
89.7
18.1
59.3
0.1
1.3
17.8
328.1
54.9
9.1
141.1
6.4
0.1
65.3
0.1
108.3
23.5
14.6
445.8
0.4
825.9
269.1
9.1
10.3
3.0
0.1
0.1
244.8
650.2
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Ca rpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g>
1919.75
200.93
160.41
234.09
169.71
396.15
295.71
490.16
1421.61
786.03
1043.71
3180.58
487.16
139.12
560.79
1081.60
2271.17
273.79
215.22
237.11
404.28
14580.42
627.45
4906.98
1005.94
381.77
1093.92
2003.84
477.12
185.71
89.37
5288.92
3852.22
1550.90
574.02
705.23
172.86
96491.66
17724.80
166.86
104.11
294.28
180.25
3409.81
388.63
2141.82
461.89
111.98
171.73
290.72
133.74
Revised Dust-Lead
Concentration
(ug/g)
1919.75
200.93
160.41
234.09
169.71
281.72
295.71
410.02
1421.61
786.03
1043.71
272.40
112.23
225.93
1081.60
2271.17
273.79
215.22
144.89
404.28
627.45
2448.58
1005.94
381.77
1093.92
2003.84
477.12
141.43
89.37
480.98
2901.12
1550.90
574.02
402.05
172.86
51150.86
17724.80
166.86
99.29
198.23
180.25
388.63
2141.82
263.32
66.06
83.65
290.72
133.74
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
1540400
1540806
1541200
1550102
1550607
1551704
1631209
1631308
1722206
1730407
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Drv Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
165.0
121.6
7.3
49.5
0.1
0.1
1011.2
324.7
5.7
74.9
0.1
0.1
152.2
0.1
15.0
11.8
7.4
0.1
0.1
0.1
1632.5
58.5
0.1
17.7
7.0
0.1
0.1
165.0
82.9
30.0
38.7
0.1
0.1
0.1
0.1
250.6
37.8
79.2
0.5
0.1
0.1
13.4
69.9
100.6
1.2
42.0
0.1
0.1
0.1
8.3
20.5
65.6
190.4
0.1
0.1
0.1
0.1
166.3
0.1
15.9
0.1
0.1
0.1
113.2
116.7
37.8
232.1
Surface
Sampled
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
162.61
132.61
87.32
76.68
129.88
255.80
373.22
113.28
457.31
149.17
156.94
150.41
355.89
82.38
208.29
1557.25
13044.11
121.21
1326.84
352.78
333.00
392.87
217.70
44021.12
457178.46
21637.24
380.41
102.85
75.01
1177.54
445.03
164.54
102.52
701.00
373.30
260.29
310.76
62.06
26.82
503.73
Revised Dust-Lead
Concentration
(ug/g)
162.61
87.32
78.68
129.88
133.18
373.22
113.28
312.16
91.98
85.42
150.41
355.89
61.81
112.17
13044.11
121.21
1326.84
352.78
333.00
392.87
217.70
14019.21
380.41
102.85
6.19
1177.54
248.16
137.63
102.52
701.00
373.30
183.19
310.76
62.06
26.82
503.73
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
1730407
1730704
1730803
1731603
1740901
1741701
1741800
1743103
1750108
1751304
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
0.1
0.1
0.1
9.7
7.8
13.0
0.1
0.3
0.1
0.1
32.
32.
138.
4.
0.
0.
68
27
0.1
11.
0.
48,
0.
0.
0.
0.1
111.
211.
64.
0.
0.
0.
0.1
341.5
2111.1
37.8
0.1
0.1
15.8
134.0
0.1
162.8
68.5
0.1
0.1
302.9
0.1
30.5
21.9
9.2
0.1
0.1
0.1
0.1
29.1
8.0
3.9
4.9
0.1
119.3
20.2
0.1
Surface
Sampled
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Original Dust-Lead
Concentration
(pg/g)
661.61
508.82
112.16
4860.14
123.25
348.36
19.69
1842.06
1.13
21.25
1639.22
534.34
581.34
118.16
146.06
54.74
261.94
185.31
164.42
93.69
111.87
33.92
10.20
112.46
734.61
355.55
633.47
470.30
114.12
60.71
1.74
Revised Dust-Lead
Concentration
(pg/g)
383.61
279.91
71.76
123.25
348.36
19.69
937.53
0.68
21.25
1639.22
534.34
581.34
118.16
146.06
54.74
183.72
185.31
164.42
93.69
111.87
33.92
9.06
64.29
355.55
24.17
323.21
470.30
114.12
60.71
1.44
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 2B4 Privately-Owned
Houses in the HUD National Survey
House
ID
1820802
1830801
1830900
1831106
1831304
1840305
1840503
1841105
1851104
1921709
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
0.6
53.1
11.7
0.1
0.6
0.1
0.1
4.8
14.6
40.3
0.1
1.4
0.1
10.3
239.4
345.3
56.7
0.6
0.1
43.5
6.1
15.6
112.8
5.6
10.7
0.1
0.1
0.1
48.3
60.2
4.5
61.4
39.8
0.1
0.1
64.4
37.3
94.7
59.9
0.1
0.1
0.1
116.4
165.8
6.0
9.4
0.9
0.1
8.2
2.6
8.9
184.5
0.1
5.1
0.1
0.1
245.5
128.7
13.8
3.8
0.1
59.2
0.1
24.6
69.3
63.0
1.3
Surface
Sampled
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
1724.44
116.91
88.43
710.47
1293.33
637.80
179.72
739.04
3716.74
6356.80
2208.23
743.76
2430.07
1701.84
2434.49
798.29
906.68
486.92
2945.23
554.18
365.62
661.48
1933.69
4457.92
526.16
139.34
93.37
1039.14
4195.53
474.27
172.44
990.63
1149.62
2397.39
4327.69
445.93
267.01
2206.28
450.11
87.43
105.66
5792.25
6202.95
883.25
492.70
229.92
6367.15
Revised Dust-Lead
Concentration
Iug/g)
116.91
54.36
658.25
429.62
179.72
74.30
2192.39
6356.80
2208.23
743.76
1701.84
1280.56
556.06
906.68
252.87
1757.62
554.18
365.62
334.16
1933.69
4457.92
526.16
139.34
93.37
1039.14
4195.53
474.27
90.47
569.53
72.52
1333.22
1743.53
252.81
267.01
1131.37
450.11
87.43
69.34
2876.27
6202.95
883.25
492.70
229.92
579.03
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
1921709
1931906
1932300
1942606
1951904
1952506
1953009
202250-7
2022705
2030302
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
0.1
•713.3
0.1
24.0
43.5
192.8
37.6
0.1
0.1
285.
307.
165.
233.
44.
0.1
573.3
88.0
0.1
611.7
0.1
34.0
0.1
567.8
0.1
16.9
147.9
70.2
3.9
42.8
76.5
15.0
25.0
18.7
51.0
0.1
1.7
0.1
0.1
0.3
4.7
0.1
2.0
0.1
0.1
0.1
506.3
46.0
139.1
19.2
0.1
81.9
209.0
100.1
78.9
295.5
11.4
5.8
146.3
492.3
62.9
0.5
35.9
0.1
0.1
16.8
0.1
Surface
Sampled
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
7687.79
216.56
175.45
194.18
5841.46
108.64
363.87
588.35
1881.20
3315.25
495.08
102.00
81.20
10863.94
3772.01
2145.20
532.39
600.73
1645.54
2197.21
3177.50
2337.97
744.96
497.97
101.44
287.27
1369.20
220.14
517.33
501.27
395.58
971.77
1109.17
3082.45
6555.04
73.82
311.01
43.56
276.41
1287.01
330.67
3530.03
131.73
1374.35
48.85
371.86
Revised Dust-Lead
Concentration
(ug/g)
7687.79
211.48
175.45
194.18
5841.46
108.64
363.87
588.35
1881.20
3315.25
495.08
102.00
81.20
10863.94
3772.01
1563.71
532.39
600.73
819.13
2197.21
3177.50
1595.89
744.96
387.92
101.44
39.94
111.76
102.85
501.27
395.58
971.77
880.89
3082.45
6555.04
73.82
311.01
43.56
168.40
671.79
330.67
3530.03
131.73
48.85
270.08
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
2040301
2110906
2121507
2122000
2130706
2131902
2141505
2141604
2142107
2151207
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
23.8
65.2
71.9
1.9
0.6
0.1
0.1
57.8
69.2
516.7
32.9
8.1
0.1
0.1
204.3
95.8
28.3
0.1
0.1
0.1
0.1
252.0
172.2
6.1
89.7
0.1
0.1
0.1
17.3
196.8
96.4
19.5
0.1
0.1
0.1
20.0
51.3
4.7
50.5
0.1
0.1
0.1
0.1
2.2
1.2
0.1
0.1
0.1
0.1
262.6
150.8
100.3
0.4
0.1
13.2
15.6
46.0
112.7
12.0
0.1
0.1
0.1
0.1
13.0
6.0
98.8
0.1
Surface
Sampled
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/gi
333.90
99.55
BO.11
2683.94
3029.38
1807.97
671.33
1802.20
5346.2*7
1749.98
349.45
187.92
96.89
521.44
220.51
240.86
493.68
15312.99
532.28
74.66
186.75
185.71
211.07
39.13
140.89
290.93
183.42
77.60
1448.53
98.50
74.10
76.34
2017.59
1309.00
6499.79
186.69
52.43
50.87
421.82
456.98
339.30
10346.61
Revised Dust-Lead
Concentration
(ug/g)
322.98
99.55
80.11
472.12
1807.97
671.33
1802.20
5346.27
970.54
349.45
187.92
96.89
521.44
220.51
126.69
493.68
393.62
74.66
186.75
149.23
173.01
39.13
71.53
290.93
46.82
6.41
98.50
74.10
76.34
842.79
4527.54
186.69
62.43
31.56
269.88
239.76
339.30
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Hell Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
2151207
2211308
2211902
2220507
2230100
2230209
2230506
2240406
2311108
2332005
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
516.0
0.1
0.1
70.7
16.7
84.0
99.4
72.2
0.1
0.1
574.9
29.7
480.8
0.1
5.8
0.1
514.9
39.1
47.7
94.
0.
0.
0.1
0.1
0.1
657.1
388.0
456.4
38.9
305.5
83.0
0.1
135.1
58.0
54.5
0.3
1055.2
0.1
35.9
24.1
17.9
95.4
1.9
0.9
0.1
275.2
Surface
Sampled
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
378.0
18.5
54.6
1.9
3.6
291.6
150.6
17.0
9.1
19.4
1.2
3.0
36.1
0.1
55.6
10.5
28.9
3.7
0.4
0.1
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
lug/91
59.35
146.35
110.28
43.73
309.15
1662.34
67.67
60.27
54.66
470.95
426.00
71.98
336.21
35.19
32.79
67.72
21.58
258.64
160.07
199.81
158.20
150.33
156.90
33.60
310.40
334.54
5015.67
7449.56
65.16
70.66
114.41
1135.30
2582.96
1078.14
472.40
616.03
3739.01
9089.15
4064.62
1439.79
1855.25
279.07
32419.38
500.07
21327.28
2093.09
896.46
442.04
4016.81
38966.11
15538.53
Revised Dust-Lead
Concentration
(ug/g)
59.35
146.35
79.82
43.73
309.15
1662.34
67.67
60.27
54.66
245.82
426.00
71.98
336.21
35.19
32.79
52.35
21.58
45.50
79.10
199.81
158.20
109.97
89.45
26.89
25.63
162.95
5015.67
65.16
41.92
114.41
562.39
1078.14
472.40
616.03
3739.01
9089.15
4064.62
1439.79
1855.25
279.07
500.07
21327.28
2053.97
677.63
442.04
706.58
2458.01
15538.53
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Hell Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
2343002
2343606
2351500
2352201
2410801
2421709
2430403
2431807
2441509
2441608
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
2.4
9.0
10.2
3.2
0.1
39.7
53.6
8.3
8.1
31.1
2.2
0.1
85.1
0.1
65.5
59.3
8.4
7.8
2.5
7.3
280.1
47.2
6.1
1.5
0.1
0.1
0.1
0.1
3.8
1.2
15.0
1.3
0.1
100.5
0.1
118.4
30.1
2.2
637.0
4.2
0.1
16.2
13.6
6.5
2.8
1.0
0.1
0.1
311.0
70.5
17.2
5.6
4.1
5.6
47.3
61.9
54.9
14.8
25.1
6.1
2.2
0.1
48.2
6.7
7.0
3.3
13.4
Surface
Sampled
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
3319.54
160.95
479.80
9635.28
10502.98
1718.00
164.55
945.25
58.22
578.47
5.17
173.76
97.71
189.69
31305.13
55.87
1006.31
70.18
199.48
130.60
253.15
1742.59
1500.26
917.40
14485.26
11942.36
185.26
114.81
113.81
159.18
123.17
402.37
1049.88
269.01
580.15
155.20
6953.19
274.44
48.73
31.78
239.74
267.90
715.29
670.27
2186.17
796.97
1306.72
3002.21
5173.31
41429.38
1420.73
271.97
109.42
9651.69
Revised Dust-Lead
Concentration
(pg/g)
1081.81
91.50
282.20
4715.78
10502.98
1718.00
91.76
524.23
58.22
147.66
5.17
173.76
97.71
106.07
17221.55
20.37
546.23
70.18
199.48
68.70
28.24
865.32
123.87
626.21
1317.29
11942.36
185.26
114.81
29.05
159.18
61.7T
286.15
684.58
142.98
268.65
10.67
6953.19
274.44
35.90
16.50
119.92
139.13
715.29
670.27
2186.17
540.40
1306.72
1579.19
1320.52
41429.38
759.06
146.46
53.68
6253.53
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
2441608
2451805
2452605
2511806
2520609
2520906
2521102
2521201
2521300
2531804
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
0.1
0.1
101.1
245.5
144.6
154.4
62.3
41.3
0.1
0.1
Surface
Sampled
Uncarpeted
Uncarpeted
5.6
16.7
1.0
11.3
5.2
49.8
34.7
26.6
8.6
9.8
0.1
0.1
0.1
0.1
55.5
1.7
12.1
1.2
0.1
0.1
0.1
17.4
5.3
4.7
0.4
0.1
0.1
0.1
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
26.0
96.2
15.8
322.7
0.1
0.1
0.1
124.4
218.9
19.7
0.7
0.1
0.1
0.1
31.1
80.0
169.7
7.5
0.1
0.1
0.1
0.1
1.5
19.6
0.1
0.1
0.1
0.1
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
55871.70
46086.68
545.04
254.64
259.70
6592.53
3004.50
17659.88
4084.97
423.98
484.85
134.08
49.13
121.93
120.50
315.10
1674.79
145.64
158.44
559.37
363.89
242.53
326.71
276.75
301.53
90.39
141.92
13.12
200.19
224.71
160.21
201.46
92120.76
206.35
140.65
143.72
142.17
565.17
505.51
74.39
5085.22
Revised Dust-Lead
Concentration
(ug/g)
46086.68
283.05
184.30
17.86
4004.39
1544.57
17659.88
4084.97
423.98
272.60
77.96
49.13
16.95
74.98
26.02
1242.97
75.06
80.44
363.89
242.53
326.71
276.75
301.53
90.39
141.92
9.20
200.19
224.71
160.21
163.17
4932.63
206.35
140.65
143.72
77.58
56.40
60.01
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
2540102
2540201
2541209
2541407
2541506
2541902
2542009
2550309
2551000
2551802
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wpt Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
1403.4
156.8
46.1
63.6
4653.9
17.5
6.3
25.7
9.1
4.7
137.8
0.8
262.4
3.4
629.0
146.3
7.6
21.1
12.2
0.1
0.1
797.8
33.2
5.7
1.0
1.6
189.7
126.1
52.8
35.3
53.4
65.5
10.2
31.5
0.1
8.6
5.6
1.1
11.7
8.5
3.5
0.1
38.3
36.2
20.2
148.7
11.7
43.8
25.8
9.8
3.7
2.3
9.2
2.2
0.1
0.1
6.4
32.5
3.8
10.8
0.1
0.1
0.1
250.8
16.1
0.3
124.1
Surface
Sampled
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
126.07
1121.76
1189.52
1008.63
18.23
15312.99
9525.45
302.48
418.80
315.86
178.79
885.97
277.32
1987.34
186.89
55.48
76.19
3005.85
1804.14
373.18
543.96
285.87
1465.59
4898.71
1007.99
574.71
258.57
7284.40
767.57
1471.65
389.10
1413.84
121.67
128.91
457.07
1681.04
2284.44
473.19
167.56
344.02
437.72
1697.73
1881.24
2630.22
2960.68
81.49
211.82
181.06
826.28
552.97
73.19
57.27
76.52
107.47
149.18
41.59
825.16
61.47
Revised Dust-Lead
Concentration
(ug/g)
126.07
1121.76
1189.52
1008.63
18.23
11405.99
5036.46
302.48
238.75
160.36
178.79
51.42
277.32
977.35
186.89
55.48
41.69
2577.65
1126.05
373.18
543.96
148.84
100.79
608.91
1007.99
574.71
258.57
7284.40
767.57
1471.65
228.85
1413.84
68.41
66.95
34.39
1033.39
1280.87
233.27
167.56
344.02
361.68
1697.73
1156.47
2630.22
2960.68
47.38
104.93
52.33
472.37
141.15
38.80
57.27
38.00
64.32
149.18
29.47
61.47
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
2551802
2552107
2610103
2611101
2620508
2621704
2622603
2623007
2650208
2651206
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet: Well
Dust Tap
Weight
(mg)
1.6
0.1
454.8
1.2
21.7
1.2
3.6
0.1
0.1
0.1
36.8
5.8
33.3
6.9
0.1
0.1
0.1
0.1
6.1
0.9
4.0
0.1
0.1
0.2
11.8
4.6
5.2
0.1
0.1
0.1
0.1
49.7
31.6
0.1
7.5
0.1
0.1
0.1
5.5
0.2
681.1
20.1
0.1
0.1
0.1
2.6
3.8
0.1
4.3
0.4
0.1
0.1
17.5
32.1
8.7
0.3
3.4
0.1
0.1
191.5
0.9
65.8
33.7
13.2
4.3
6.1
Surface
Sampled
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
254.14
189.79
1418.81
40.81
344.81
155.T8
2012. -n
336.38
218.92
255.36
635.01
117.15
411.89
676.06
4042.32
239.03
162.41
357.39
206.10
166.99
189.00
308.52
1448.53
0.84
67.43
314.38
302.23
209.34
931.20
319.27
31.27
68.98
45525.09
188.67
1.83
8967.06
54.09
383.78
365.27
2815.24
2679.94
Revised Dust-Lead
Concentration
(ug/g)
31.59
189.79
117.15
35.90
28.48
76.98
336.38
114.27
255.36
341.06
61.62
25.98
337.35
147.38
82.28
183.73
206.10
166.99
103.14
159.81
0.84
55.50
126.66
150.08
105.23
237.81
31.27
38.89
92.79
1.83
565.65
54.09
383.78
235.17
1415.20
1409.67
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
2651800
2652303
2710101
2711109
2711505
2721009
2730703
2731503
2731800
2751402
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
61.4
103.2
29.8
1.1
0.1
0.1
0.1
2.9
7.0
28.0
0.1
0.1
0.1
218.5
10.8
45.7
1.0
0.1
0.1
46.8
0.1
134.4
100.9
11.5
44.1
0.1
0.1
0.1
180.3
123.7
16.8
0.1
18.5
0.1
0.1
82.6
33.7
32.8
71.7
0.1
26.3
53.1
7.1
31.8
0.1
0.8
5.7
25.4
118.3
85.9
16.3
65.6
21.4
20.6
0.1
0.1
39.0
651.5
8.4
115.6
0.1
0.1
0.1
3.1
8.2
3.5
108.3
Surface
Sampled
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
362.30
425.09
381.92
1006.44
536.15
152.TO
325.49
369.13
662.87
236.21
2375.24
198.73
137.17
120.09
138.00
170.86
224.05
187.56
310.83
5178.64
260.71
419.35
351.11
1983.53
2689.40
32617.62
329.22
66.75
2335.81
289.69
294.74
167.93
136.87
188.73
137.55
1937.45
465.55
334.70
100.79
207.26
509.88
705.15
132887.36
114281.43
89.55
Revised Dust-Lead
Concentration
Iug/g)
362.30
425.09
381.92
75.72
255.69
82.23
325.49
369.13
396.74
236.21
163.35
198.73
137.17
120.09
84.33
170.86
224.05
187.56
225.76
4003.55
260.71
419.35
351.11
1983.53
2689.40
32617.62
177.75
66.75
135.55
150.83
294.74
167.93
136.87
134.68
137.55
1682.63
391.01
334.70
100.79
115.89
509.88
344.29
73900.56
56337.86
89.55
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
2751402
2810307
2812105
2812204
2822005
2830602
2831006
2831709
2832004
2832103
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(Rig)
0.4
0.1
0.1
3.8
3.8
29.5
0.1
0.1
281.8
0.1
9.9
9.9
30.9
0.1
0.1
155.1
0.1
63.3
8.6
1.3
5.1
T.I
232.7
255.6
3.8
51.7
40.8
9.9
31.8
0.1
0.1
56.0
22.
42.
2.
0,
0.
0.1
16.3
3.5
30.2
0.1
0.1
16.8
0.1
11.8
9.5
98.0
0.8
0.1
135.7
0.1
8.
56.
3.2
26.4
6.4
6.4
8.2
0.1
Surface
Sampled
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Original Dust-Lead
Concentration
32.8
2928.99
1044.42
1044.42
49.43
821.01
275.43
275.43
128.44
1090.99
535.30
317.07
267.92
534.66
732.03
287.75
354.11
302.23
74.05
48.69
2435.11
732.07
319.64
67.42
28.48
1088.18
71.09
112.04
23.30
50.50
7-7.16
162.28
64.40
2315.05
1555.42
515.01
273.34
1063.76
434.56
1258.05
227.82
227.82
95.79
1609.91
Revised Dust-Lead
Concentration
(ug/g)
518.63
518.63
49.43
821.01
160.61
160.61
128.44
1090.99
535.30
178.27
24.37
274.17
395.24
287.75
354.11
150.08
74.05
48.69
1419.98
732.07
319.64
60.35
28.48
518.96
50.73
55.23
23.30
36.68
47.58
93.56
64.40
134.35
1555.42
291.97
273.34
520.63
1258.05
120.77
120.77
53.27
1609.91
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
2840106
2840205
2840403
2841203
2841401
2841500
2910107
2931202
2931608
2940401
Sampling
Location
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Dust Tap
Weight
(mg)
101.
80.
12.6
26.
0.
0.
0.1
6.7
61.2
384.8
56.1
27.2
0.1
0.1
1.5
1.4
1.4
0.1
Surface
Sampled
Original Dust-Lead
Concentration
(ug/g)
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
3B.O
18.5
2.7
7.1
0.1
0.1
0.1
13.7
88.2
39.2
0.1
111.5
177.1
0.1
75.7
36.2
6.1
59.8
0.5
0.1
55.6
83.3
6.6
23.6
0.3
0.1
0.1
0.1
15.2
77.8
30.7
0.1
0.1
0.1
0.1
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Carpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
1633.68
62.08
172.44
268.14
118.57
308.52
3853.58
2756.00
3154.81
71.07
41.44
57.67
1193.84
879.46
1079.32
144.08
152.06
37.70
1062.25
5276.77
747.38
240.02
70.38
36722.14
5509.65
392.95
248.99
199.54
697.70
2412.86
225.04
899.54
1672.11
2851.59
920.22
570.71
363.23
13413.62
Revised Dust-Lead
Concentration
(ug/g)
1633.68
47.99
75.37
144.77
77.56
308.52
3853.58
2756.00
3154.81
71.07
41.44
30.33
1193.84
1079.32
144.08
81.03
36.13
513.58
240.02
70.38
392.95
248.99
126.08
697.70
1289.14
225.04
899.54
1672.11
2851.59
102.67
57.38
36.52
-------
Table B-l. Original and Revised Dust-Lead Concentrations for All Floor,
Window Sill, and Window Well Samples in the 284 Privately-Owned
Houses in the HUD National Survey
House
ID
2940401
2940*708
3011103
3011509
3011905
3020401
3040706
3050101
3051000
Sampling
Location
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Entryway
Dry Floor
Wet Floor
Dry Sill
Wet Sill
Dry Well
Wet Well
Dust Tap
Weight
(mg)
0.1
0.1
0.1
4.9
13.9
10.6
10.9
1.8
0.1
0.1
0.1
0.1
Surface
Sampled
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
178.6
49.4
25.6
6.2
1.7
50.5
56.4
3.3
3.1
9.0
6.7
0.1
0.1
0.1
6.1
7.5
0.1
12.0
26.5
150.7
5.3
74.9
24.8
14.5
17.2
0.1
123.4
16.8
5.1
2.5
2.1
Uncarpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
6.5
2.3
0.7
0.1
0.1
0.1
100.4
1.1
0.1
0.1
0.1
0.1
0.1
0.1
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Carpeted
Carpeted
Uncarpeted
Uncarpeted
Original Dust-Lead
Concentration
(ug/g)
1030.71
287.30
254.88
471.21
845.89
1308.53
923.07
2104.20
14294.90
104367.71
109165.57
44369.80
1195.19
992.38
406.65
643.86
2511.49
371.91
2309.75
18493.37
3934.92
63101.24
410.52
286.47
3680.55
8830.44
83633.26
348.88
655.36
829.02
929.24
2046.60
1438.09
666.93
2021.70
Revised Dust-Lead
Concentration
(ug/g)
525.90
189.15
151.66
282.86
132.05
1308.53
923.07
2104.20
7538.72
14512.20
109165.57
44369.80
586.37
484.53
231.18
344.00
1392.87
202.95
1432.87
18493.37
3934.92
32521.60
410.52
192.32
2711.95
8830.44
60743.81
178.91
238.94
186.69
493.88
591.56
77.00
666.93
152.10
------- |