Advancing Pb Exposure and Biokinetic Modeling for U.S. EPA Regulatory Decisions and Site Assesstnents using Bunker Hill Mining and Metallurgical Cotnplex Superfund Site Data


nvEPA
United States
Environmental
Protection Agency
EPA 600/R-21/017F | April 2021 | www.epa.gov/research
Advancing Pb Exposure and Biokinetic
Modeling for U.S. EPA Regulatory Decisions
and Site Assessments using Bunker Hill
Mining and Metallurgical Complex
Superfund Site Data
Office of Research and Development
Center for Public Health and Environmental Assessment
Health and Environmental Effects Assessment Division

-------
&EPA
United States
Environmental
Protection Agency
EPA60Q/R-21/017F | April 2021 | www.epa.gov/research
Advancing Pb Exposure and Biokinetic
Modeling for U.S. EPA Regulatory Decisions
and Site Assessments using Bunker Hill
Mining and Metallurgical Complex
Superfund Site Data
Office of Research and Development
Center for Public Health and Environmental Assessment
Health and Environmental Effects Assessment Division

-------
Advancing Pb Exposure and Biokinetic Modeling
Disclaimer
The views expressed in this report are those of the author(s) and do not necessarily represent
the views or policies of the U.S. Environmental Protection Agency. This document does not
represent and should not be construed to represent any Agency determination or policy. Any
mention of trade names, products, or services does not imply an endorsement by the U.S.
Government or the U.S. Environmental Protection Agency. The EPA does not endorse any
commercial products, services, or enterprises.
Acknowledgments
Cover photograph of Coeur d'Alene River shows Black Bridge and the Trail of the Coeur
d'Alene's bike path in Enaville. Photograph from Coeur d'Alene River Road on July 1, 2014
provided courtesy of Jill Dorsey of Alta Science & Engineering, Inc. The authors thank Dana
Swift and Andy Helkey at the Idaho Department of Environmental Quality and the local Kellogg
Idaho Panhandle Health District for allowing use of these data for this study. The authors
acknowledge the thousands of Silver Valley residents that participated in the lead health and
remediation activities over the years. This work was funded, in part, under U.S. EPA contract
EP-C-17-015.

-------
Advancing Pb Exposure and Biokinetic Modeling
Authors and Contributors
Dr. James S. Brown—Center for Public Health and Environmental Assessment, Office of
Research and Development, U.S. EPA, Research Triangle Park, NC
Dr. Gary L. Diamond—SRC, Inc., North Syracuse, NY
Dr. Mark H. Follansbee—SRC, Inc., Scarborough, ME
Dr. William C. Thayer—SRC, Inc., North Syracuse, NY
Susan M. Spalinger—Alta Science and Engineering, Inc., Moscow, ID
Mara Thorhaug—Alta Science and Engineering, Inc., Kellogg, ID
Sarah G. Weppner—Alta Science and Engineering, Inc., Boise, ID
Kynan J. Witters Hicks—Alta Science and Engineering, Inc., Moscow, ID
Reviewers
Dr. Philip Goodrum—GSI Environmental, Inc., Fayetteville, NY
Dr. Rosalind A. Schoof—Ramboll, Environment and Health, Seattle, WA
Dr. Kathleen L. Vork—Office of Environmental Health Hazard Assessment, California EPA,
Oakland, CA

-------
Advancing Pb Exposure and Biokinetic Modeling
Contents
Executive Summary	xvi
Section 1 Purpose and Background	1
1.1	Quality Assurance and Peer Review	1
1.2	Report Structure	2
1.3	IEUBK Model Overview	2
1.4	Bunker Hill Superfund Site Background	4
Section 2 BHSS Data Description	8
2.1	Blood Lead Data	8
2.2	House Dust Data	9
2.2.1	Box House Dust	9
2.2.2	Basin House Dust	10
2.3	Soil Data	10
2.3.1	Box Soil	10
2.3.2	Basin Soil	11
2.4	Bioavailability Data	12
2.5	Drinking Water Data	13
2.5.1	Box Drinking Water	13
2.5.2	Basin Drinking Water	13
2.6	Air Data	14
2.7	Additional Site Data	14
Section 3 Preparation and Selection of BHSS Paired Dataset	14
3.1	Blood Lead Data	15
3.2	House Dust Data	15
3.3	Soil Data	16
3.4	Drinking Water Data	17
3.5	BHSS Data Selected for Model Evaluation	17
Section 4 Model Evaluation Approach	20
4.1	IEUBK Inputs	20
4.1.1	Air	20
4.1.2	Dietary Intake	20
4.1.3	Drinking Water	20
4.1.4	Soil/Dust Ingestion Rates	20
4.1.5	Exposure Media Bioavailability	21
4.1.6	Dust and Soil Partitions	21
4.1.7	Weighted Media Concentrations	21
4.1.8	Lead in Paint	21
4.2	Model Evaluation Methods	22
4.2.1 Geometric Mean BLL Comparisons	23
iv

-------
Advancing Pb Exposure and Biokinetic Modeling
4.2.2	Probability of Exceeding 5 |jg/dL Comparisons	23
4.2.3	Evaluation of Individual Child BLL Distributions	24
4.2.4	Additional Supporting Information of Model Predictions	26
4.2.5	Sensitivity Analyses	27
Section 5 IEUBK Model Evaluation Results and Discussion	29
5.1	BDLBLLs	30
5.2	Prediction of Population GM BLLs	33
5.2.1	Performance of Default IEUBK v2.0 Model	33
5.2.2	Alternative IRsd and Soil and Dust Partitions	33
5.3	Prediction of Probability of Exceeding 5 |jg/dL	44
5.3.1	Performance of Default IEUBK v2.0 Model	44
5.3.2	Alternative IRSd and Soil and Dust Partitions	44
5.4	Prediction of Distribution of Individual Child BLLs	57
5.4.1	Prediction Intervals for IEUBK Predicted GM BLLs	57
5.4.2	Cumulative Distributions of Individual Child BLLs	63
5.5	Sensitivity Analyses	67
5.5.1	Alternative Assumptions for Clean Backfill Soil Pb Concentration	67
5.5.2	Alternative Assumptions for GSDi	69
5.5.3	Alternative Assumptions for Dietary Pb Intake	69
5.5.4	Comparison of IEUBK 2.0 and v1.1 Model IRsd	72
5.6	Discussion	73
Section 6 Conclusions	75
6.1	Performance of Default IEUBK v2.0 Model	76
6.2	Alternative IRsd and Soil/Dust Partitions	77
6.3	Model Sensitivity to Clean Backfill Soil Pb Concentration, GSD,, Dietary Pb
Intake, and IEUBK v1.1 Model Default IRsd	77
References	79
Appendices
Appendix A Default Inputs to IEUBK (v1.1 and v2.0)	A-1
Appendix B Additional Sampling/Monitoring and Data for the BHSS	B-1
Appendix C LHIP Blood Lead Screening Questionnaire	C-1
Appendix D Floor Mat Questionnaire	D-1
Appendix E Paired Dataset Summary Tables and Figures	E-1
Appendix F Censor Level 1 (<30.5 |jg/dL) Tables and Figures	F-1
Appendix G Censor Level 2 (<30.5 |jg/dL and No BDL) Additional Tables and Figures	G-1
Appendix H Censor Level 3 (<10.5 |jg/dL and No BDL) Tables and Figures	H-1
Appendix I Not Human Subjects Research Determination	1-1
Appendix J Biographical Summaries of Peer Reviewers	J-1
Appendix K SRC Quality Assurance Project Plan	K-1
v

-------
Advancing Pb Exposure and Biokinetic Modeling
Figures
Figure 1. Bunker Hill Superfund Site, Box
Figure 2. Coeur d'Alene Basin
Figure 3. Soil Pb concentration boxplot for paired records (mg/kg) in the Box, 1995-2018
Figure 4. Soil Pb concentration boxplot for paired records (mg/kg) in the Basin, 2002-2018
Figure 5. Summaries of mean observed and predicted BLLs with 95% CIs and lines of unity for
geographic areas, censor level 2 (omits predicted and observed BLLs >30.5 |jg/dL
and observed BDL BLLs)
Figure 6. Summary of mean observed and predicted BLLs with 95% CIs for geographic areas,
censor level 2 (omits predicted and observed BLLs >30.5 |jg/dL and observed BDL
BLLs)
Figure 7. Summaries of mean observed and predicted BLLs with 95% CIs and lines of unity for
site-wide age groups, censor level 2 (omits predicted and observed BLLs >30.5
|jg/dL and observed BDL BLLs)
Figure 8. Summary of mean observed and predicted BLLs with 95% CIs for site-wide age
groups, censor level 2 (omits predicted and observed BLLs >30.5 |jg/dL and
observed BDL BLLs)
Figure 9. Weighted linear regression model for observed and predicted age group GM BLLs
with 95% CIs, censor level 2 (omits predicted and observed BLLs >30.5 |jg/dL and
observed BDL BLLs)
Figure 10. Average yearly media concentrations (mg/kg) in the Box used as inputs for the
IEUBK model, censor level 2 (omits predicted and observed BLLs >30.5 |jg/dL and
observed BDL BLLs), 1995-2018
Figure 11. Average yearly media concentrations (mg/kg) in the Basin used as inputs for the
IEUBK model, censor level 2 (omits predicted and observed BLLs >30.5 |jg/dL and
observed BDL BLLs), 2002-2018
Figure 12. Summaries of observed and predicted average probability of exceeding 5 |jg/dL with
95% CIs and lines of unity for the geographic areas, censor level 2 (all predicted and
observed BLLs <30.5 |jg/dL and no BDLs)
Figure 13. Summary of observed and predicted average probability of exceeding 5 |jg/dL with
95% CIs for the geographic areas, censor level 2 (all predicted and observed BLLs
<30.5 |jg/dL and no BDLs)
Figure 14. Summaries of observed and predicted average probability of exceeding 5 |jg/dL with
95% CIs and lines of unity for site-wide age groups, censor level 2 (all predicted and
observed BLLs <30.5 |jg/dL and no BDLs)
Figure 15. Summary of observed and predicted average probability of exceeding 5 |jg/dL with
95% CIs for site-wide age groups, censor level 2 (all predicted and observed BLLs
<30.5 |jg/dL and no BDLs)
Figure 16. Weighted linear regression model for observed and predicted age group P5 with
95% CIs, censor level 2 (omits predicted and observed BLLs >30.5 |jg/dL and
observed BDL BLLs)
Figure 17. Plot of predicted and observed BLLs from an ideal lognormal distribution of BLLs
with 95% Pis
VI

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 18. Percent of observed BLLs that fall outside the 95% PI of the predicted BLL GMs for
the geographic areas, censor level 2 (all predicted and observed BLLs <30.5 |jg/dL
and no BDLs)
Figure 19. Percent of observed BLLs that fall outside the 95% PI of the predicted BLL GMs for
site-wide age groups, censor level 2 (all predicted and observed BLLs <30.5 |jg/dL
and no BDLs)
Figure 20. Cumulative distribution functions for observed and predicted BLLs for site-wide,
censor level 2 (all predicted and observed BLLs <30.5 |jg/dL and no BDLs)
Figure 21. Cumulative distribution functions for observed and predicted BLLs in the Box, censor
level 2 (all predicted and observed BLLs <30.5 |jg/dL and no BDLs)
Figure 22. Cumulative distribution functions for observed and predicted BLLs in the Basin,
censor level 2 (all predicted and observed BLLs <30.5 |jg/dL and no BDLs)
Figure 23. Comparison of predicted geometric mean BLLs (|jg/dL) using the IEUBK v2.0 default
diet Pb intake estimates and diet Pb intake estimated by Zartarian et al. (2017)
Figure 24. Comparison of mean observed and predicted BLLs (|jg/dL) with 95% CIs using three
combinations of model inputs and ingestion rates
Figure E-1. Blood lead level histograms for paired records (|jg/dL), 1995-2018 - Box
Figure E-2. Soil lead concentration histograms for paired records (mg/kg), 1995-2018 - Box
Figure E-3. Vacuum lead concentration histograms for paired records (mg/kg), 1995-2018 - Box
Figure E-4. Blood lead level histograms for paired records (|jg/dL), 2002-2018 - Basin
Figure E-5. Soil lead concentration histograms for paired records (mg/kg), 2002-2018 - Basin
Figure E-6. Vacuum lead concentration histograms for paired records (mg/kg), 2002-2018 -
Basin
Figure F-1. Summary of mean observed and predicted blood lead levels for geographic areas
for censor level 1 (dataset omits predicted and observed BLLs >30.5 |jg/dL)
Figure F-2. Summary of observed and predicted blood lead geometric means for site-wide age
groups for censor level 1 (dataset omits predicted and observed BLLs >30.5 |jg/dL)
Figure F-3. Summary of observed and predicted average probability of exceeding 5 |jg/dL by
geographic areas for censor level 1 (all predicted and observed BLLs <30.5 |jg/dL)
Figure F-4. Summary of observed and predicted average probability of exceeding 5 |jg/dL for
site-wide age groups for censor level 1 (all predicted and observed BLLs <30.5
|jg/dL)
Figure F-5. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Site-wide
Figure F-6. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Box
Figure F-7. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Basin
Figure F-8. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Site-wide, age <1 year
Figure F-9. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Site-wide, age 1 year
Figure F-10. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Site-wide, age 2 years

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure F-11. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Site-wide, age 3 years
Figure F-12. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Site-wide, age 4 years
Figure F-13. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Site-wide, age 5 years
Figure F-14. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Site-wide, age 6 years
Figure F-15. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Site-wide Ages 2-6 years
Figure G-1. Summary of mean observed and predicted blood lead levels for geographic areas
for censor level 2 (omits predicted and observed BLLs >30.5 |jg/dL and observed
BDL BLLs)
Figure G-2. Summary of mean observed and predicted blood lead levels for site-wide age
groups for censor level 2 (omits predicted and observed BLLs >30.5 |jg/dL and
observed BDL BLLs)
Figure G-3. Summary of observed and predicted average probability of exceeding 5 |jg/dL for
the geographic areas for censor level 2 (all predicted and observed BLLs <30.5
|jg/dL and no BDLs)
Figure G-4. Summary of observed and predicted average probability of exceeding 5 |jg/dL for
the site-wide age groups for censor level 2 (all predicted and observed BLLs <30.5
|jg/dL and no BDLs)
Figure G-5. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 1 - Site-wide
Figure G-6. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Site-wide
Figure G-7. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Box
Figure G-8. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Basin
Figure G-9. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Site-wide, Age <1 year
Figure G-10. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Site-wide, Age 1 year
Figure G-11. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Site-wide, Age 2 years
Figure G-12. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Site-wide, Age 3 years
Figure G-13. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Site-wide, Age 4 years
Figure G-14. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Site-wide, Age 5 years
Figure G-15. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Site-wide, Age 6 years
Figure G-16. Scatterplots of observed/predicted blood lead pairs that fall outside prediction
intervals for censor level 2 - Site-wide ages 2-6 years

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure H-1. Summary of mean observed and predicted blood lead levels for geographic areas
for censor level 3 (omits predicted and observed BLLs >10.5 |jg/dL and observed
BDL BLLs)
Figure H-2. Summary of mean observed and predicted blood lead levels for site-wide age
groups for censor level 3 (omits predicted and observed BLLs >10.5 |jg/dL and
observed BDL BLLs)
Figure H-3. Summary of observed and predicted average probability of exceeding 5 |jg/dL for
geographic areas for censor level 3 (all predicted and observed BLLs <10.5 |jg/dL
and no BDLs)
Figure H-4. Summary of observed and predicted average probability of exceeding 5 |jg/dL for
site-wide age groups for censor level 3 (all predicted and observed BLLs <10.5 |jg/dL
and no BDLs)
ix

-------
Advancing Pb Exposure and Biokinetic Modeling
Tables
Table 1. Bunker Hill Superfund Site average bioavailability of Pb in soil and dust
Table 2. Number of individual children and homes in the paired dataset used for model
evaluation
Table 3. Summary of IEUBK model configurations and batch mode files
Table 4. Summary of paired BLL and Pb exposures for study population, censor level 2
(dataset omits paired records where an observed or predicted BLL is >30.5 |jg/dL or
an observed BLL is BDL)
Table 5. IEUBK model output summary of observed BLLs recorded as BDL
Table 6. Summary of mean observed and predicted BLLs for IEUBK v2.0 model, censor level
2 (dataset omits predicted and observed BLLs >30.5 |jg/dL and observed BDL BLLs)
Table 7. Summary of the difference between observed and predicted GM BLLs for IEUBK
v2.0 model, censor level 2 (dataset omits predicted and observed BLLs >30.5 |jg/dL
and observed BDL BLLs)
Table 8. Summary of observed and predicted average probability of exceeding 5 |jg/dL for
IEUBK v2.0 model, censor level 2 (dataset omits predicted and observed BLLs >30.5
|jg/dL and observed BDL BLLs)
Table 9. Summary of the difference between observed and predicted average probability of
exceeding 5 |jg/dL for IEUBK v2.0 model, censor level 2 (dataset omits predicted
and observed BLLs >30.5 |jg/dL and observed BDL BLLs)
Table 10. Summary of observed/predicted PbB pairs that fall outside Pis for IEUBK v2.0
model, censor level 2 (dataset omits predicted and observed BLLs >30.5 |jg/dL and
observed BDL BLLs)
Table 11. Summary of the K-S 2-sample comparison for IEUBK v2.0 model, censor level 2
(dataset omits predicted and observed BLLs >30.5 |jg/dL and observed BDL BLLs)
Table 12. Comparison of soil summary statistics using different clean soil Pb concentrations in
the Box for IEUBK v2.0 model, default IRsd, censor level 2 (dataset omits predicted
and observed BLLs >30.5 |jg/dL and observed BDL BLLs)
Table 13. Comparison of predicted BLL summary statistics using different clean soil Pb
concentrations in the Box for IEUBK v2.0 model, default IRsd, censor level 2 (dataset
omits predicted and observed BLLs >30.5 |jg/dL and observed BDL BLLs)
Table 14. GSD sensitivity analysis for IEUBK v2.0 model, default IRsd, 55/45 partition, censor
level 2 (dataset omits predicted and observed BLLs >30.5 |jg/dL and observed BDL
BLLs)
Table 15. Diet sensitivity analysis for IEUBK v2.0 model
Table 16. Comparison of GM PbB concentrations predicted by IEUBK v2.0 model with default
and IEUBK v1.1 model IRsd
Table A-1. IEUBK input parameters based on TRW Lead Committee recommendation
Table A-2. Age-dependent inputs to the IEUBK v1.1win model Build 11
Table A-3. Age-dependent inputs to the IEUBK v2.0 model Build 1.6
Table B-1. Average annual Bunker Hill Superfund Site lead concentration in air (|jg/m3)
presented by station
Table B-2. Summary of lead concentration in air (|jg/m3) for Bunker Hill Support Zone
monitoring (1995-1999)
Table B-3. Number of additional paired records if using floor mat dust data
x

-------
Advancing Pb Exposure and Biokinetic Modeling
Table E-1. Box paired blood, yard soil, and vacuum lead concentrations for LHIP participants
ages 0-6 years old, 1988-2018
Table E-2. Box - Age (in years) of LHIP participants with paired soil and dust data, 1988-2018
Table E-3a. Basin paired blood, soil, and vacuum lead concentrations for LHIP participants ages
0-6 years old, 1996-2018
Table E-3b. Basin paired blood and drinking water lead concentrations for LHIP participants
ages 0-6 years old, 1996-2018
Table E-4. Basin - age (in years) of LHIP participants with paired soil and dust data, 1996-2018
Table E-5. Box - community soil lead concentrations (mg/kg) by year
Table E-6. Basin - community soil Pb concentration (mg/kg) by year
Table F-1. Summary of observed and predicted BLLs for censor level 1 (dataset omits predicted
and observed BLLs >30.5 |jg/dL)
Table F-2. Summary of observed and predicted average probability of exceeding 5 |jg/dL for
censor level 1 (all predicted and observed BLLs <30.5 |jg/dL)
Table F-3. Summary of observed/predicted blood Pb pairs that fall outside prediction intervals
for censor level 1 (all predicted and observed BLLs <30.5 |jg/dL)
Table F-4. Summary of sum of squared differences for censor level 1 (dataset omits predicted
and observed BLLs >30.5 |jg/dL)
Table F-5. Percent differences <1 |jg/dL between observed and predicted BLLs for censor level
1	(dataset omits predicted and observed BLLs >30.5 |jg/dL)
Table F-6. Summary of the difference between observed and predicted GM BLLs for IEUBK
v2.0 model, censor level 1 (dataset omits predicted and observed BLLs >30.5 |jg/dL)
Table F-7. Summary of the difference between observed and predicted average probability of
exceeding 5 |jg/dL for IEUBK v2.0 model, censor level 1 (dataset omits predicted
and observed BLLs >30.5 |jg/dL)
Table G-1. Summary of mean observed and predicted BLLs for IEUBK v1.1 build 11 and censor
level 2 (dataset omits predicted and observed BLLs >30.5 |jg/dL and observed BDL
BLLs)
Table G-2. Summary of observed and predicted average probability of exceeding 5 |jg/dL for
IEUBK v1.1 build 11 and censor level 2 (dataset omits predicted and observed BLLs
>30.5 |jg/dL and observed BDL BLLs)
Table G-3. Summary of observed/predicted blood lead pairs that fall outside prediction intervals
for censor level 2 (dataset omits predicted and observed BLLs >30.5 |jg/dL and
observed BDL BLLs)
Table G-4. Summary of sum of squared differences for censor level 2 (dataset omits predicted
and observed BLLs >30.5 |jg/dL and observed BDL BLLs)
Table G-5. Percent differences <1 |jg/dL between observed and predicted BLLs for censor level
2	(dataset omits predicted and observed BLLs >30.5 |jg/dL and observed BDL BLLs)
Table G-6. Summary of goodness-of-fit between observed and predicted BLLs for IEUBK v2.0
model, censor level 2 (dataset omits predicted and observed BLLs >30.5 |jg/dL and
observed BDL BLLs)
Table G-7. IEUBK v1.1 model output summary of observed BLLs recorded as below detection
limits
Table H-1a. Summary of mean observed and predicted BLLs for IEUBK model version 2.0 and
censor level 3 (dataset omits predicted and observed BLLs >10.5 |jg/dL and
observed BDL BLLs)
xi

-------
Advancing Pb Exposure and Biokinetic Modeling
Table H-1b. Summary of mean observed and predicted BLLs for IEUBK model version 1.1 build
11 and censor level 3 (dataset omits predicted and observed BLLs >10.5 |jg/dL and
observed BDL BLLs)
Table H-2a. Summary of observed and predicted average probability of exceeding 5 |jg/dL for
IEUBK model version 2.0 censor level 3 (dataset omits predicted and observed BLLs
>10.5 |jg/dL and observed BDL BLLs)
Table H-2b. Summary of observed and predicted average probability of exceeding 5 |jg/dL for
IEUBK model version 1.1 build 11 and censor level 3 (dataset omits predicted and
observed BLLs >10.5 |jg/dL and observed BDL BLLs)
Table H-3. Summary of observed/predicted blood lead pairs that fall outside prediction intervals
for censor level 3 (dataset omits predicted and observed BLLs >10.5 |jg/dL and
observed BDL BLLs)
Table H-4. Summary of sum of squared differences for censor level 3 (dataset omits predicted
and observed BLLs >10.5 |jg/dL and observed BDL BLLs)
Table H-5. Percent differences <1 |jg/dL between observed and predicted BLLs for censor level
3 (dataset omits predicted and observed BLLs >10.5 |jg/dL and observed BDL BLLs)
xii

-------
Advancing Pb Exposure and Biokinetic Modeling
Acronyms and Abbreviations
ABA	absolute bioavailability
AFP	absorption fraction percent
Alta	Alta Science & Engineering, Inc.
BDL	below detection limit
BHSS	Bunker Hill Mining and Metallurgical Complex Superfund Site
BLL	blood lead level or blood lead concentration
BPRP	Basin Property Remediation Program
CDA Trust Coeur d'Alene Custodial and Work Trust
CDC	Centers for Disease Control and Prevention
CERCLA Comprehensive Environmental Response, Compensation and Liability Act
CERCLIS Comprehensive Environmental Response, Compensation and Liability
Information System
CFR	Code of Federal Regulations
CI	confidence interval
df	degree(s) of freedom
EFH	Exposure Factors Handbook
FDA	Food and Drug Administration
GM	geometric mean
GSD	geometric standard deviation
GSDi	individual geometric standard deviation (1.6)
HHRA	Human Health Risk Assessment
HHRE	Human Health Remedial Evaluation
HUD	U.S. Department of Housing and Urban Development
ICP	Institutional Controls Program
IDEQ	Idaho Department of Environmental Quality
IDHW	Idaho Department of Health and Welfare
IEUBK	Integrated Exposure Uptake Biokinetic (model)
IRsd	soil/dust ingestion rate
ISI	Influential Science Information
IVBA	in vitro bioaccessibility
K-S	Kolmogorov-Smirnov
LHIP	Lead Health Intervention Program
MCL	maximum contaminant level
xiii

-------
Advancing Pb Exposure and Biokinetic Modeling
NAAQS
U.S. National Ambient Air Quality Standards
NAS
National Academy of Sciences
NHANES
National Health and Nutrition Examination Survey
OMB
Office of Management and Budget
OU
Operable Unit
Pb
lead
PbB
blood lead
PHD
Panhandle Health District
PI
prediction interval
PRP
Potentially Responsible Party
Px
probability of exceeding the blood lead level decision point, x
QA
quality assurance
QAPP
Quality Assurance Project Plan
QC
quality control
RAO
Remedial Action Objective
RBA
relative bioavailability
RCRA
Resource Conservation and Recovery Act
RI/FS
Remedial Investigation/Feasibility Study
ROD
Record of Decision
ROW
right-of-way
SRC
SRC, Inc.
SSD
sum(s) of squared differences
SSE
sum of squares error
SST
sum of squares total
TDS
Total Diet Study
TRW
Technical Review Workgroup
TSP
total suspended particulate
USACE
U.S. Army Corps of Engineers
USEPA
U.S. Environmental Protection Agency
WSE
weighting standard error
WWEIA
What We Eat In America
XRF
x-ray fluorescence
xiv

-------
Advancing Pb Exposure and Biokinetic Modeling
Units

%
percent
dL
deciliter
L
liter
kg
kilogram
m
meter
mg
milligram
M9
microgram

-------
Advancing Pb Exposure and Biokinetic Modeling
Executive Summary
The Integrated Exposure Uptake Biokinetic (IEUBK) Model for Lead in Children was developed
by the U.S. Environmental Protection Agency (USEPA) to support assessments of health risks
from exposures of children to lead (Pb). This report documents the performance evaluation of
the IEUBK version 2.0 (v2.0) model as it would be typically applied at Superfund sites to predict
blood lead levels (BLLs) in populations of children. The IEUBK v2.0 model was evaluated by
comparing model predictions of BLLs to approximately 1200 observed BLLs in a population of
children at the Bunker Hill Mining and Metallurgical Complex Superfund Site (BHSS) for which
there were paired estimates of environmental Pb concentrations. Predicted site-wide population
geometric mean (GM) BLLs (GM: 3.4 |jg/dL, 95% confidence interval [CI]: 3.3, 3.5) were within
0.26 |jg/dL of the observed GM (3.6 |jg/dL, 95% CI: 3.5, 3.8). The model predicted the observed
age trend in GM BLLs and explained approximately 90% of the variance in the observed age-
stratified GM BLLs. Differences (predicted minus observed) between area-stratified GM BLLs
ranged from -0.44 to +0.36 |jg/dL (mean difference: -0.1 |jg/dL). The predicted site-wide
probability of exceeding 5 |jg/dL (Ps) was 26.7% (95% CI: 24.0, 29.1) and within 5.5% of
observed Ps(32.1%, CI: 29.4, 34.8). Differences between area-stratified Ps values ranged
from -10.8 to +13.8%. Although the more general applicability of these findings remains to be
determined in future studies of other populations, these results support applications of the
IEUBK v2.0 model for informing risk-based discussions regarding remediation of soils and
mitigation of exposures at Comprehensive Environmental Response, Compensation and
Liability Act (CERCLA) sites where the majority of the exposure unit GM BLLs are expected to
be <5 |jg/dL.
xvi

-------
Advancing Pb Exposure and Biokinetic Modeling
Section 1 Purpose and Background
The purpose of this report is to present an evaluation of a new draft version of the Integrated
Exposure Uptake Biokinetic (IEUBK) Model for Lead (Pb) in Children (version 2.0 build 1.6;
v2.0) using available data from the Bunker Hill Mining and Metallurgical Complex Superfund
Site (BHSS) in northern Idaho. Subsequent builds of the IEUBK v2.0 model were minor changes
to the graphical user interface and the help file. These minor changes did not alter the numerical
results of the model and did not affect the validity of this evaluation. This evaluation is
comprised of the following three parts:
• description of the available BHSS blood lead (PbB) and environmental exposure data
and selection of the data subset for model evaluation,
• evaluation of IEUBK v2.0 conformance with the selected BHSS data subset using three
performance metrics, and
• analysis of the sensitivity in model predictions to the following variables: soil/dust
ingestion rates (IRsd), clean soil Pb concentration assumption, geometric standard
deviation (GSD), and dietary Pb intake.
The primary purpose was to evaluate the new IEUBK v2.0 model. Modeling results using the
current available software at the time of this evaluation, i.e., lEUBKwin version 1.1, build 11
(v1.1), are presented in appendices to this report.
SRC, Inc. (SRC) collaborated with Alta Science and Engineering, Inc. (Alta) to support the U.S.
Environmental Protection Agency's (USEPA) Office of Research and Development, Center for
Public Health and Environmental Assessment (SRC-USEPA Contract No. EP-C-17-015,
68HERC19F0099, Task Order No. 0011).
1.1 Quality Assurance and Peer Review
The use of quality assurance (QA) and peer review helps ensure that USEPA conducts high-
quality science that can be used to inform policymakers, industry, and the public. QA activities
performed by USEPA ensure that the Agency's environmental data are of sufficient quantity and
quality to support the Agency's intended use. Detailed QA Project Plans (QAPPs) that were
developed as a requirement for contracted technical support during this evaluation of the IEUBK
v2.0 model and development of this report are shown in Appendix K. The IEUBK model is
classified as providing Influential Scientific Information (ISI), which is defined by the Office of
Management and Budget (OMB) as scientific information that the agency reasonably can
determine will have or does have a clear and substantial impact on important public policies or
private sector decisions (OMB. 2004). OMB requires the Agency to subject ISI to be peer review
prior to dissemination. To meet this requirement, EMS, Inc., under contract EP-W-13-016 with
USEPA's Office of Superfund Remediation and Technology Innovation, conducted an external
peer review of a March 2020 draft IEUBK evaluation report. Independent of EPA, reviewers
were chosen to create a balanced review panel based on factors such as technical expertise,
knowledge, experience, and absence of any real or perceived conflicts of interest. In April-May
2020, letter peer reviews were received from three doctoral level toxicologists with expertise
related to human health risks associated with Pb exposures and the use of modeling to assess
exposures and/or biokinetics. Biographical summaries of the reviewers appear in Appendix J.
Peer review comments were considered in the development of this final report.
1

-------
Advancing Pb Exposure and Biokinetic Modeling
1.2 Report Structure
The report is organized as follows:
Section 1 provides an overview of the IEUBK model and uncertainty in the model, and a brief
background of the BHSS and past applications of the IEUBK model at the BHSS.
Section 2 describes the available BHSS PbB and Pb exposure data including the past data
collection efforts.
Section 3 summarizes the preparation of the paired dataset and the data subset selected for
this evaluation.
Section 4 describes the methods used to validate the IEUBK v2.0 model build 1.6.
Section 5 presents results and discusses the model evaluation outcomes.
Section 6 presents conclusions and recommendations.
1.3 IEUBK Model Overview
The IEUBK model was developed by the USEPA to support the assessment of health risks to
children caused by exposure to Pb (U.S. EPA, 2010a, 1994a).1 The IEUBK model is used
currently in human health risk assessment (HHRA) at Comprehensive Environmental
Response, Compensation, and Liability Act (CERCLA) hazardous waste sites and Resource
Conservation and Recovery Act (RCRA) corrective action facilities where Pb is a chemical of
potential concern (U.S. EPA, 1998, 1994a, b).
In the USEPA CERCLA program, the IEUBK model is used to support decisions regarding the
need for remediation of soils and sediments at Pb-contaminated sites (U.S. EPA. 1998, 1994b)
Central to the application of the IEUBK model in risk assessment is the prediction of the
probability of an exposed child population to exceed a specific blood lead level (BLL) decision
point. Site cleanup levels are selected to achieve a probability of exceeding the BLL decision
point, x (Px), that is no more than 5% (Pxs5%). A widely applied BLL decision point has been 10
|jg/dL, which was the Centers for Disease Control and Prevention's (CDC) previous action level
(U.S. EPA. 1994b).
The model predicts Pxfrom a simulation of the relationships between multimedia exposures to
Pb and quasi-steady state BLLs in children across an age range of 1-6 years (U.S. EPA,
2017b). This is achieved in four components to the model: exposure, uptake, biokinetic, and
variability. The exposure component of the model calculates daily intakes of Pb (|jg/day)
averaged over each year of a child's life. Daily intake is based on the combined exposures to Pb
in air, diet, indoor dust, soil, drinking water, or other user-defined sources of exposure. Media-
specific time-averaged Pb absorption rates (|jg/day) are calculated from Pb intake rates. The
uptake component models the process by which Pb intake (Pb that has entered the child's body
through ingestion or inhalation) is transferred to the blood plasma. Uptake (|jg/day) is the
quantity of Pb absorbed per unit time from portals of entry (gut, lung) into the systemic
circulation of blood. The biokinetic component simulates transfer of absorbed Pb between blood
and other body tissues, as well as elimination of Pb from the body (via urine, feces, skin, hair,
and nails), and predicts a geometric mean (GM) BLL for each of 84 months. In the variability
component, a lognormal probability model is used to calculate the probability of occurrence of a
specified BLL over a user-specified age range in a hypothetical child or population of similarly
1 IEUBK v2.0 software will be available for download at https://www.epa.gov/superfund/lead-superfund-
sites-software-and-users-manuals.
2

-------
Advancing Pb Exposure and Biokinetic Modeling
exposed children, based on the GM BLL and the GSD expected among similarly exposed
children (GSD,).
Confidence in risk-based decisions regarding remediation of Pb in soil at CERCLA sites
depends on confidence in the IEUBK model predictions of the Px. Over-prediction of the Px could
lead to unneeded, expensive, and disruptive remediation activities, whereas under-prediction of
the Px could result in leaving contaminated soil in place and under-estimating residual post-
remediation risk at the site. Numerous model-attributed factors contribute to uncertainty in the
Px, including uncertainties and/or errors about model parameter values. Model uncertainty can
be grouped into two broad categories: (1) uncertainties that affect confidence in the prediction of
the GM BLL of the child population, including uncertainties related to media Pb concentrations,
intakes, bioavailability, and Pb biokinetics; and (2) uncertainty in the interindividual variance in
BLLs among similarly exposed children such as biological and behavioral differences that are
encompassed by the GSDi.
Several studies have evaluated these two major categories of uncertainties in the IEUBK model
by comparing predicted GM BLL and BLL distributions with observations made in populations of
exposed children (Zartarian et al., 2017; Li et al., 2016; von Lindern et al., 2003a; Bowers and
Mattuck, 2001; Griffin et al.. 1999; Hoqan et al.. 1998) and by quantifying the effects of
parameter uncertainty on uncertainty in model predictions (Griffin et al.. 1999).
The Hoqan et al. (1998) study was particularly important to regulatory application of the IEUBK
model as a risk assessment tool at CERCLA sites (U.S. EPA, 1998), specifically for the purpose
of predicting the probability of exceeding a BLL of 10 |jg/dL (P10). Hoqan et al. (1998) compared
observed population GM BLLs and BLL probability distributions with predictions made by the
IEUBK model in a sample population of 478 children who resided at four CERCLA sites. Hoqan
et al. (1998) reported several important statistical evaluations of model performance. The
following conclusions from Hoqan et al. (1998) are relevant to the objectives of the model
evaluations described in this report: (1) predicted site GM BLLs were not statistically different
from observed site GM BLLs and were within <1 |jg/dL of observations, (2) predicted site mean
P10 values were not statistically different from observed site mean P10 values and were within
4% of observations, and (3) >80% of the observed individual child BLLs at each site were within
the 95% prediction interval (PI) of the model (defined by the predicted GM BLL and GSDi).
Since the Hoqan et al. (1998) evaluation, several changes were made to the IEUBK model
prompting the need for the model evaluation described in this report. Input parameter values for
the IEUBK model were revised to reflect more recent data. These included adjustments to
parameters that govern exposures to, and intakes of, Pb in air, diet, drinking water, soil, and
soil-derived dust.
In addition, newer guidance for application of the IEUBK model recommends that the model be
used to estimate the Px based on the average GM BLL predicted for the age range of
1-<6 years, rather than 0.5 months to <7 years, or other intermediate age ranges (U.S. EPA.
2017b). Laboratory methods to estimate relative bioavailability (RBA) of Pb in soils/dusts also
have since been validated, and RBA can be incorporated into the calculation of Pb uptake (U.S.
EPA. 2009). Newer epidemiological findings and national trends in BLLs prompt consideration
of BLL decision points below 10 |jg/dL (Paulson and Brown. 2019; U.S. EPA. 2013; CDC.
2012). Foreseeing potential applications of the model to predict the probability of exceeding
lower BLLs, this evaluation considered the probability of exceeding 5 |jg/dL (Ps). This is the
lowest action level at which environmental investigations are recommended by eight U.S. states
(U.S. EPA. 2020). Five states have no recommendation and the remaining 37 states
recommend action levels between 10 and 25 |jg/dL.
3

-------
Advancing Pb Exposure and Biokinetic Modeling
Lowering the BLL decision point has several implications for IEUBK model applications. Site GM
BLLs in the Hogan et al. (1998) study ranged from 5 to 7 |jg/dL. Using a GSD of 1.6, the Ps
corresponds to a population GM BLL of 2.3 |jg/dL. IEUBK model performance for predicting
lower range BLLs (<5 |jg/dL) in child populations has not been evaluated. A further complicating
factor in the performance of the IEUBK model at this lower BLL range is the substantially larger
contribution of dietary Pb to total Pb uptake. As soil Pb levels decrease, agreement between
observed and predicted BLLs becomes increasingly more dependent on assumptions about
dietary Pb intake in the receptor population, a variable that is usually not measured at CERCLA
sites. Dietary Pb intakes in the receptor population are typically represented by national-level
estimates (e.g., see Zartarian et al. (2017)).
The changes noted above to the IEUBK model (especially revision of the IRsd) and the need to
apply the model in assessing risks at lower BLLs prompted this evaluation of the most recent
draft model version, IEUBK v2.0, described in this report. This study was designed to evaluate
performance of the IEUBK v2.0 model as it would be typically applied at CERCLA sites to
predict BLLs in populations of children. To accomplish this objective, data on child BLLs paired
with soil and indoor dust (house dust) Pb exposure concentrations were needed from a
population of children whose BLL distribution included a substantial number of children with
BLLs <5 |jg/dL.
1.4 Bunker Hill Superfund Site Background
The BHSS has one of the most extensive assemblages of paired PbB and environmental
concentration data in the nation due to the long-term Pb health response efforts and CERCLA
remediation actions conducted at the BHSS from the 1980s to the present day. In addition,
more recent BHSS data have a significant number of children exhibiting BLLs <5 |jg/dL because
site cleanup efforts are largely complete.
The BHSS is located in the Coeur d'Alene River basin and extends across northern Idaho from
the Montana border on the east to the Washington border on the west. It was added to the
National Priorities List in 1983 and has been assigned the Comprehensive Environmental
Response, Compensation and Liability Information System (CERCLIS) identification number
IDD048340921. The BHSS includes mining-contaminated areas in the Coeur d'Alene River
corridor, adjacent floodplains, downstream water bodies, tributaries, and fill areas, as well as the
21-square-mile area referred to as the "Box" that surrounds the historic smelting operations at
the Bunker Hill complex.
USEPA identified three Operable Units (OU) within the BHSS: OU1 is the populated areas of
the Box; OU2 is the non-populated areas of the Box; and OU3, or "the Basin," is the areas of
mining-related contamination outside the Box in the broader river basin. OU1 includes the cities
of Kellogg, Wardner, Smelterville, and Pinehurst, and several smaller unincorporated
communities such as Page (Figure 1). OU3 is divided into two geographic areas split by the
Box: (1) the Upper Basin, which consists of the following communities: Burke/Ninemile, Mullan,
Osburn, Side Gulches, Silverton, and Wallace, and the (2) Lower Basin geographic area, which
consists of two community areas including Kingston and the Lower Basin (Figure 2). These
eight Basin community areas were originally delineated in the HHRA (TerraGraphics. 2001c)
based on identified routes of potential human exposure and public use patterns in each area at
that time.
4

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 1. Bunker Hill Superfund Site, Box
This map was proceed using information attained
from several different sources thai have not been
ndependentfy venfied. These sources have also
not provided information on the precision and accuracy
of the data. information on thts map is not a substitute
for sirvey data
Bunker Hill "Box"
and OU2)
	 Major Streets
Interstate 90
Streams and Tributaries
jB ¦ City Limit Boundary
J "Box" Boundary

Print Date
November 14, 2019
project Manager
M. Thorhaug
1:48.000 a
Project name
Figure 1
Bunker Hill

Requestor
1 inch = 1 miles j Ik
Advancing Pb Exposure
and Biokinetic Modeling

Project Number
M. Thorhaug
0 0.5 1 Miles /W\
Superfund Site, Box

18128
Cartographer
B. Bailey
	
5

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 2. Coeur d'Alene Basin
Area 4
Kingston
Area 2
Burke Ninemile
Bunker Hill "Box1
Area 1
Mullan
s.Area 8
.Wallace
Area 3
'.Side
Gulches
Area 4
Kingston
Kootenai C ounty
Benewah Count}-
Upper Basin
Lower Basinui
Figure 2
Bunker Hill Mining and Metallurgical
Complex Superfund Site
Advancing Pb Exposure
aid Biokinetic Modeling
6

-------
Advancing Pb Exposure and Biokinetic Modeling
The BHSS was the first CERCLA site where regulatory agencies employed the IEUBK model to
develop site-specific soil and house dust Pb cleanup levels (U.S. EPA, 2002; TerraGraphics,
2001c; U.S. EPA. 1992; CH2M Hill. 1991; U.S. EPA. 1991; TerraGraphics. 1990). Due to the Pb
health response efforts undertaken in the 1980s, property-specific paired data for children's
blood, soil, and house dust were available for validation of the original IEUBK and for use in the
BHSS site-specific analysis conducted in 1990.
The 1990 analysis found that the observed site-specific, dose-response relationship between
soil, dust, and BLLs was consistently lower than default IEUBK model predictions, using the
default bioavailability parameter of 30%, default IRsd, and default soil/dust partition
(TerraGraphics, 1990). This was attributed to lower soil and dust bioavailability, although it was
acknowledged that the reduced dose-response was likely a combination of lower bioavailability
and ingestion rates (von Lindern et al., 2003b).
A variation of the IEUBK model was developed for the Box, which used site-specific data and
inputs that accounted for the lower observed dose-response relationship at the BHSS. It
became known as the "Box Model." The Box Model applied an assumed soil and dust
bioavailability of 18% and a house dust to yard soil to community soil partition of 40/30/30 as
compared to the defaults of 30% bioavailability and a partition of 55/45 for dust/soil. The
40/30/30 dust/soil partition assumes that 40% of a child's total dust/soil ingestion derives from
house dust, 30% derives from residential yard soils, and 30% derives from community-wide
soils (based on a GM) (von Lindern et al., 2003a; von Lindern et al., 2003b; TerraGraphics,
2001a, 2000). Annual monitoring and evaluation of the dose-response relationship in the Box
occurred until the Remedial Action Objective (RAO) was met in 2002 (TerraGraphics. 2004).
The Box Model was used in the Basin HHRA and the 2002 Basin Records of Decision (ROD) to
select site-specific remedial objectives for the Basin. The IEUBK analyses conducted as part of
the Basin HHRA used paired soil, dust, drinking water, and PbB results from 1996 to 1999 from
participants in the Basin and indicated that potentially different exposures were occurring in the
Basin community areas. In areas east of the Box, the Box Model was a better predictor of
observed BLLs, especially immediately east of the Box (Osburn, the Side Gulches, and
Silverton). In the Lower Basin (west of the Box), the Box Model under-predicted mean BLLs and
P10. The "default" IEUBK model (55/45 dust/soil partition) described mean BLLs fairly well in the
Lower Basin but failed to capture the predicted P10. The poor fit of both models in the Lower
Basin was partly attributed to potential recreational exposures to Pb in soils and sediment in the
Lower Basin (TerraGraphics, 2001c). Monitoring and evaluation of the dose-response
relationship in the Basin did not occur as consistently as it did in the Box.
The "Box Model" continues to be used to periodically evaluate achievement of the RAOs in both
the Box and the Basin (TerraGraphics, 2015; U.S. EPA, 2015, 2010a; TerraGraphics, 2008;
U.S. EPA, 2005a; TerraGraphics, 2004).
The von Lindern et al. (2016) study re-evaluated the Box dose-response relationship using
bioavailability information based on in vitro bioaccessibility (IVBA) data from the site (as
described in Section 2.4). That re-evaluation used Box paired blood, soil, and house dust data
from 1988 to 2002. Structural equations modeling resulted in a new recommended dust and soil
partition scenario of 50/25/10/15 (dust/yard soil/neighborhood soil mean/community soil mean).
Ultimately, child IRsd values were derived from re-evaluation of the data in that study.
Comprehensive information about site history, risk assessment/management, and sampling and
remediation activities can be found in the following documents: the OU1, OU2, and OU3 RODs
(U.S. EPA, 2002, 1992, 1991), the Risk Assessment Data Evaluation Report (TerraGraphics,
1990), the HHRA (TerraGraphics, 2001c), the Human Health Remedial Evaluation (HHRE)
7

-------
Advancing Pb Exposure and Biokinetic Modeling
(TerraGraphics. 2004), the National Academy of Sciences (NAS) review of the Basin (NRC,
2005). and the 2015 Five-Year Review (U.S. EPA. 2015).
Section 2 BHSS Data Description
This section describes BHSS site-specific PbB and environmental concentration data collection
efforts for available data that were collected through 2018 and considered for use in this
evaluation of the IEUBK v2.0 model. More recent site data from 2019 are not discussed
because these data were not available at the time the dataset was compiled for this model
evaluation.
Information on additional BHSS data not used in this evaluation is included in Appendix B of this
report.
2.1 Blood Lead Data
The BHSS Lead Health Intervention Program (LHIP) is administered by the Panhandle Health
District (PHD) as a public health service and offers activities designed to reduce children's Pb
exposures through biological monitoring, follow-up consultations, parental awareness,
counseling, and education. Through the LHIP, annual voluntary PbB screening has been offered
since 1985 in the Box and since 1996 in the Basin. Annual PbB screening occurs in July or
August to capture anticipated peak BLLs in young children. In the Box, a thorough door-to-door
solicitation approach was conducted from 1985 to 2002 and again in 2013 in order to maximize
the identification and monitoring of eligible children. The door-to-door solicitation approach
employed in the Box was also employed in the Basin in 1996, with fixed-site voluntary blood
draws in all other years.
To encourage participation, a monetary incentive was offered to families with participating
children in specific age ranges during the annual July/August PbB screening. In the Box,
monetary incentives were offered from 1988 through 2002 to children between 6 months and
9 years of age. The monetary incentive was reinstated in 2013 and again in 2016 through the
present to families of children between 6 months and 6 years of age in response to low
participation rates in the Box. In the Basin, monetary incentives were offered from 1996 through
2000 to families of children between 6 months and 9 years of age, and beginning in 2001,
children between 6 months and 6 years of age.
Venous PbB screening of children was employed at the BHSS from 19882 to 2002, when
capillary PbB testing was adopted. From 2002 through 2011, confirmatory venous samples
were collected when a capillary result was >8 |jg/dL; in 2012, this confirmatory venous threshold
was lowered to 5 |jg/dL in response to recommendations by the CDC (2012). The detection limit
for venous samples is <1.0 |jg/dL.
From 2002 to 2017, capillary blood testing equipment with a detection limit of <1.4 |jg/dL was
used in the Box and Basin. In 2017, new capillary blood testing equipment with a detection limit
of <1.9 |jg/dL was used along with the existing capillary blood testing equipment with a
detection limit of <1.4 |jg/dL. In 2018, only the new capillary equipment was used for PbB
screening. All PbB tests were reported to a whole number or to one decimal place (tenth).
During each annual PbB survey, the child's residence, age, and sex were recorded, and
participants completed a questionnaire with multiple-choice and fill-in-the-blank questions at the
2 Erythrocyte protoporphyrin levels were monitored in the Box prior to 1988 (from 1985 to 1987).
8

-------
Advancing Pb Exposure and Biokinetic Modeling
time of the blood draw. The questionnaire used in 2017 and 2018 is included in Appendix C for
reference.
Annual PbB monitoring data are confidential, and individual information is not released by PHD
or the State of Idaho to anyone other than the participant's legal guardian. This project,
evaluating the IEUBK model using BHSS site-specific PbB and environmental data, was
reviewed according to the requirements of USEPA Order 1000.17A (Policy and Procedures on
Protection of Human Research Subjects) and USEPA Regulation 40 CFR (Code of Federal
Regulations) 26 (Protection of Human Subjects), and determined as not human subjects
research. The determination letter is provided in Appendix I.
2.2 House Dust Data
House dust samples were collected regularly in the Box and the Basin from personal home
vacuum cleaner bags and from carpeted floor mats that were placed by a trained sampling
crew.
Vacuum bag Pb concentrations were used in health risk assessment/management evaluations
for the BHSS, and floor mat results were used to evaluate the effectiveness of the soil remedy in
both the Box and the Basin. The floor mat technique was first used in BHSS homes in the 1996
Coeur d'Alene River Basin Environmental Health Exposure Assessment (Idaho Department of
Health and Welfare. 2000). House dust samples were generally collected between July and
September and were sieved through a No. 80 mesh (177 |jm) prior to laboratory analysis. The
80 mesh sieve (for both soil and house dust) was initiated at the BHSS in 1974 and focused
analyses on particle sizes more likely to adhere to hands and other surfaces and be ingested by
children (Panhandle Health District, 1986). Minimum detection limits for Pb in house dust varied
over the years, ranging up to 50 mg/kg.
During the annual house dust sampling program, participants completed a questionnaire with
multiple-choice and fill-in-the-blank questions at the time of floor retrieval. An example floor mat
questionnaire is included in Appendix D for reference.
The floor mats employed by the dust mat sampling methodology changed twice over the years
(in 2002 and again in 2013) because the mat manufacturer discontinued production of the prior
mat model and a new mat model was introduced. In 2013, there were a few remaining
discontinued dust mats in stock, which were used in the Box, and the new dust mat model was
used in the Basin. In addition, a different dust extraction method was applied in 2017 and 2018.
2.2.1 Box House Dust
Box house dust samples were collected from home vacuum cleaner bags since 1988 and from
carpeted floor mats since 1996.3 Annual house dust monitoring in the Box ceased after 2005;
however, dust mats and vacuum samples were collected in 2008, 2013, and 2018. In years
when annual monitoring efforts did not occur, dust samples were collected by PHD or Idaho
Department of Environmental Quality (IDEQ) through the LHIP follow-up and intervention
services when needed or requested.
3 Box residential dust data are described in detail in the HHRE (TerraGraphics, 2004). the 2013 House
Dust and Blood Lead Data and Risk Management Evaluation (TerraGraphics, 2015). the 2005, 2010, and
2015 Five-Year Reviews (U.S. EPA. 2015, 2010c, 2005b). and the 2018 Box House Dust Data Summary
Report (Alta, 2019b).
9

-------
Advancing Pb Exposure and Biokinetic Modeling
2.2.2 Basin House Dust
Dust mat and vacuum cleaner samples were collected in the Basin in 1996-1999, 2004-2011,
2013, 2015, 2017, and 2018.4 Since 2002 when the ROD was signed, dust sampling efforts
typically targeted Upper Basin homes. Dust samples from homes in the Lower Basin were
limited to the opportunistic collection of vacuum samples through the Basin Property
Remediation Program (BPRP) and targeted sampling of homes where a child participated in the
LHIP (typically five homes or less per year), with the exception of 2018, when a concerted effort
to monitor dust in Lower Basin homes was completed for the upcoming Five-Year Review.
2.3 Soil Data
Residential soil sampling has been conducted at the BHSS beginning in 1986 in the Box and in
1996 in the Basin. This section summarizes the residential property soil data from the Box and
Basin and how these data were applied in remediation decisions, risk assessment, and
evaluation of dose-response relationships. Differences in the transport and distribution of Pb
contamination to residential soils and the application of the Selected Remedies in the Box and
the Basin are also discussed. As with house dust samples, composite soil samples collected
from properties in the Box and the Basin were sieved through a No. 80 mesh (177 |jm) prior to
laboratory analysis. Minimum detection limits for Pb in soil varied over the years, ranging up to
50 mg/kg.
2.3.1 Box Soil
Various entities conducted property soil sampling in the Box. In 1986, residential property soil
samples were collected by PHD and the State of Idaho. Between 1991 and 2006, the majority of
the residential property soil sampling was conducted by the Potentially Responsible Parties
(PRPs), collectively referred to as the Upstream Mining Group. The State of Idaho also collected
supplemental soil samples on properties outside the PRP's responsibility during those years.
The Box residential soil cleanup was largely complete by 2004. As of the 2015 Five-Year
Review, 14 properties had not been remediated due to owner refusal (U.S. EPA. 2015). For all
other properties in the Box, yard soils with Pb levels >1000 mg/kg were removed to a depth of
6 or 12 inches and replaced with clean soil (e.g., soils containing less than 100 mg/kg Pb; with
no replacement soil exceeding 150 mg/kg Pb (Idaho Legislature, 2007)). Physical barriers (such
as fabric or a visual liner layer) were also installed at properties where soil Pb concentrations
were >1000 mg/kg at depths >12 inches. Vegetable gardens were remediated to a depth of
24 inches when soil samples were >1000 mg/kg Pb.
Prior to remediation, backfill materials were sampled to ensure that Pb concentrations did not
exceed 100 mg/kg. Although sampling results of actual replacement soils generally averaged
<55 mg/kg throughout the cleanup (McCulley, 1997), no confirmatory soil sampling was or has
been conducted at remediated properties after the placement of clean backfill. Therefore, a
conservative value of 100 mg/kg was used to represent a remediated soil concentration.
Because of the ubiquitous nature of contamination in the Box, when a property's yard required
remediation, other discrete areas of the property (such as gardens and driveways) were also
remediated, resulting in complete remediation of most properties with yard soils >1000 mg/kg
Pb. A few discrete area remediations were completed in the Box, in which soils were replaced in
4 Basin residential dust data are described in detail in the HHRA (TerraGraphics, 20010); the 2005, 2010,
and 2015 Five-Year Reviews (U.S. EPA, 2015, 2010a, 2005b); the 2015, 2017, and 2018 Data Summary
Reports (CPA Trust, 2019, 2018; TerraGraphics, 2016); and the House Dust Evaluation (Alta, 2020).
10

-------
Advancing Pb Exposure and Biokinetic Modeling
only select portions of the property such as a driveway or garden. More discrete area
remediations were completed in Pinehurst compared to the other Box communities.
Given the application of the remedy in the Box, the yard soil Pb concentration was considered
representative of the property's overall soil Pb concentration and was used to evaluate property
soil Pb contribution to house dust and remedy effectiveness in the Box. Consequently, it was
reasonable to assume that the yard soil Pb concentration is a representative surrogate of a
child's typical surface soil exposure at a residential property (e.g., the residential soil exposure
index) in the Box. Similarly, the yard soil concentration for all of the properties in a community
was considered representative of the central tendency of the overall distribution of soil Pb for
each Box community.
Throughout the years, Box community mean soil Pb concentrations were estimated annually in
the following manner. All surface soil (0-1 inch depth interval) yard samples and the remediation
year were paired by residential property. As properties were sampled and remediated, yard soil
data from those properties were added to the existing dataset and were assumed to apply to all
years until the year that the property was remediated. Once remediated, the yard soil Pb
concentration was replaced with the clean soil value of 100 mg/kg. Community soil means were
updated annually through 2004 using new yard soil sampling results and remediation dates
(U.S. EPA. 2015').
2.3.2 Basin Soil
As with the Box, various entities conducted property soil sampling in the Basin throughout the
years. Surface soil sampling (0-1 inches) was initially conducted by the State of Idaho and PHD
in 1996 as part of the Coeur d'Alene River Basin Environmental Health Exposure Assessment
(Idaho Department of Health and Welfare, 2000). Between 1998 and 2000, USEPA and the
State conducted soil sampling at select properties as part of the Remedial
Investigation/Feasibility Study (RI/FS). The U.S. Army Corps of Engineers (USACE) also
collected soil data for soils 0-6 inches deep using x-ray fluorescence (XRF). However, the
majority of the Basin property soil data were collected in 2002 through 2018 by the State after
the ROD was signed (U.S. EPA. 2002). The State managed the BPRP from 2002 through 2015,
at which time the Coeur d'Alene Custodial and Work Trust (CDA Trust) took over funding and
managing the BPRP to present day.
The residential soil remedy was largely complete in the Basin, although as of 2015, owners of
approximately 200 properties had refused sampling and/or remediation (U.S. EPA, 2015).
Sampling and remediation of these refusal properties continues to occur as ownership changes
or owners consent to sampling/remediation.
The Selected Remedy for residential soils in the Basin was similar to the Box in that it included:
(1) installation of a 6- or 12-inch clean soil cap for properties where soil samples were
>1000 mg/kg Pb or >60 mg/kg arsenic, (2) an additional physical barrier below the soil cap for
properties where soils deeper than 12 inches exceeded those concentrations, and
(3) remediation of vegetable gardens to a depth of 24 inches when soil samples were
>1000 mg/kg Pb or 60 mg/kg arsenic. The Selected Remedy for the Basin included an
additional component: "greening" (or barrier enhancement) for properties where soil samples
were >700 mg/kg Pb.
In the Basin, contamination was transported through the region by flooding, railroad spillage, or
use of contaminated materials for fill or gravel cover. These activities resulted in a patchwork of
contamination between geographic areas, within communities, and within individual properties.
As a result, the yard soil Pb concentration was not assumed to be representative of an entire
property during sampling and remediation activities (as it was in the Box where aerial deposition
11

-------
Advancing Pb Exposure and Biokinetic Modeling
was a significant source of contamination). Instead, discrete areas (including driveways, parking
areas, play areas, flower gardens, vegetable gardens, right-of-way [ROW], and other areas)
were sampled at each property at the same time the yard was sampled. The remedial actions
described above were then applied to each discrete sampled area based on Pb and arsenic
concentrations and the depth of contamination.
Due to the large number of discrete area remediations in the Basin, use of the yard soil Pb
concentration as the residential soil exposure index (as done in the Box) may not be
representative of the residential and community soil exposures. Instead, all discrete and yard
soil results from a property were combined and weighted spatially. This spatially-weighted mean
approach was used to account for size variation between discrete sample locations in order to
better represent potential exposure concentrations (TerraGraphics. 2010).
Similar to the Box, the Basin community mean soil concentrations were estimated periodically
based on new soil sampling results and remediation dates. All surface soil (0-1 inch depth
interval) yard and discrete area samples (from Basin residential properties) and the remediation
year(s) were included in the dataset. As properties were sampled and remediated, soil data
from those properties were added to the existing dataset and were assumed to apply to all
years until the year of property remediation. The soil Pb concentration for each remediated yard
and discrete area was replaced with the clean soil value of 100 mg/kg. Average square footage
for each type of discrete area in each community was calculated in 2010 based on a subset of
properties. Those average square-foot values were used to calculate a spatially-weighted soil
mean for each property and for each community.
Currently, PHD manages soils in the Box and the Basin through the Institutional Controls
Program (ICP) in a manner consistent with the Rules of Idaho Health District #1 (IDAPA
41.01.01) (Idaho Legislature, 2007) to continue to reduce Pb levels in community soils and
maintain the community mean of <350 mg/kg. The ICP monitors and permits projects
throughout the BHSS that may disturb clean and/or contaminated soils. Under the ICP,
disturbed soils with Pb concentrations >350 mg/kg are disposed of in one of the designated
repositories or placed under a cap. New property development and future modifications to
existing properties will continue to create clean barriers under the ICP.
2.4 Bioavailability Data
Subsequent to the cleanup in the Box, USEPA adopted an in vitro methodology to estimate site-
specific bioavailability of Pb in soil and dust (U.S. EPA, 2012, 2007a). In 2012, this methodology
was applied to a subset of archived soil and dust samples from the Box. The objective of the
2012 study was to estimate age-specific IRsd through reanalysis of the dose-response
relationship using new soil and house dust Pb bioavailability data (von Lindern et al., 2016).
Archived soil and dust samples collected in the Box between 1986 and 2002 were retrieved
from storage. Those with intact seals, legible identification numbers, and sufficient mass for
analysis were then checked to ensure that PbB data and information on child age and sex,
home location, and property remediation status were available. In total, 271 samples
(193 house dust samples, 73 yard soil samples, and 5 quality control [QC] samples) were
sieved to a No. 80 mesh (177 |jm) and then analyzed for total Pb (Method 6010B) and IVBA
(TerraGraphics, 2012; U.S. EPA, 2012, 2007a). Soil bioavailability averaged 33% and dust
bioavailability averaged 28% (Table 1; (von Lindern et al., 2016)).
12

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 1. Bunker Hill Superfund Site average bioavailability of Pb in soil and dust
Kellogg
Page
Pinehurst
Smelterville
Wardner
24
7
33
8
34±3%
33±4%
32±4%
39±2%
*
66
12
75
36
4
28±6%
27±5%
28±6%
30±4%
27±5%
Total
73
33±4%
193
28±6%
Basint
65
33±3%
157
28±6%
Notes:
*Kellogg soil ABA used for Wardner because n=1.
"'"Derived from the average of ABA for Box communities, except for Smelterville.
ABA = absolute bioavailability
2.5 Drinking Water Data
The following subsections describe available individual drinking water data for residential
properties and do not address community drinking water supplies.
2.5.1	Box Drinking Water
As part of the Box Selected Remedy, the PRPs were tasked with sampling and closing domestic
drinking water wells that exceeded the federal drinking water standards for total arsenic,
cadmium, Pb, or zinc, and connecting the affected residences to a municipal water system. The
PRPs reported the well test results and well status in remedial action certification reports
developed as part of USEPA certification process. The Box properties that had a private water
source were connected to city water systems as part of the Selected Remedy for the Box, and
there was little information available on private drinking water data at individual property wells in
the Box for use in this evaluation. Of the thousands of properties in the Box, <100 wells were
closed, were not found, were inaccessible, did not require closure, or refused closure (U.S.
EPA. 2015V
The Fourth Five-Year Review for the Box reported that 60 private wells did not require testing or
closure and 13 private wells were not closed due to owner refusal at the time (U.S. EPA, 2015).
In 2020, IDEQ investigated the status of those 13 wells and determined that all were dedicated
to non-potable uses. In addition, IDEQ confirmed that each of the 13 refusals were either
connected to city water or had new wells installed in compliance with the "area of drilling
concern" guidelines (Idaho DEQ, 2020).
2.5.2	Basin Drinking Water
The Selected Remedy for the Basin included testing of private drinking water sources and
provision of safe drinking water for homes that exceed federal drinking water standards for
arsenic, cadmium, or Pb. Through the BPRP, drinking water samples were collected from
properties with private wells. In general, a first-draw (initial) sample was collected from a source
that had not been used for at least 6 hours. A second sample (purged) was collected after a
10-minute flush. Samples were collected at the kitchen sink when accessible. Because owners
were not always home at the time of sampling, water samples may have been collected from an
exterior spigot or a spigot located in a garage or shop. If a filter was used, a second set of
13

-------
Advancing Pb Exposure and Biokinetic Modeling
samples was collected at the source prior to any holding tanks or filters, if possible. In addition,
at some homes, water samples were collected from sources that were used for irrigation.
Occasionally, follow-up water sampling was conducted when the initial sample results were not
consistent with the purged sample results, or when exterior water samples were not consistent
with interior water samples (e.g., one sample exceeded maximum contaminant levels [MCLs]
and the other does not). If a drinking water source exceeded the MCL for arsenic, cadmium, or
Pb, filtration or an alternative drinking water source was provided, and the well was capped, if
necessary.
2.6 Air Data
Air monitoring for Pb was conducted in the Box to determine if goals for fugitive dust control
identified in the ROD were being met. Air quality monitoring using total suspended particulate
(TSP) monitors was performed from 1995 through 2002, predominantly by USACE. During this
time, average annual Pb concentrations in air did not exceed the U.S. National Ambient Air
Quality Standards (NAAQS) Pb risk standard of 1.5 |jg/m3 in any year (U.S. EPA. 2005b).
USEPA and the State decided to discontinue air quality monitoring after 2002 because all major
source removal actions had been completed by 2000 and no exceedances of NAAQS standards
(for particulate matter or Pb) had occurred in the final 2 years of air monitoring (U.S. EPA.
2005b).
USACE collected particulate air Pb concentration samples in seven locations in the Box.
Additional data were collected at the Kellogg Medical Clinic from 1995 through 2002 by the
State of Idaho (Idaho DEQ, 2005; TerraGraphics. 2001a, 2000). Average annual Pb
concentrations for all locations from 1995 to 1998, and the additional Kellogg Medical Clinic
samples through 2002, ranged from 0.00 to 0.10 |jg/m3 (Appendix B). However, these data were
suspect due to contradictory results from the Bunker Hill Remedial Action Project Closure
Report, discussed in the supporting documentation to the 2000 Five-Year Review
(TerraGraphics. 2000; Morrison-Knudsen, 1999). The Remedial Action Project Closure Report
suggested that average air Pb concentrations at each monitoring station were higher than those
collected by the State (Appendix B, compare Tables B-1 and B-2). As an example, Table B-2
shows an average Pb concentration of 0.334 |jg/m3 at the East Gate monitoring station for the
years 1995-1999; however, Table B-1 shows that yearly averages at East Gate were between
0.04 and 0.08 |jg/m3 from 1995 through 1998. Air Pb concentration would have needed to be
1.45 |jg/m3 in 1999 to match the average reported in Table B-2.
2.7 Additional Site Data
Additional BHSS data were collected throughout the years in both the Box and the Basin (e.g.,
questionnaire information on child behavior). However, these data may be limited in scope or
size and may therefore have limited application to this model evaluation (e.g., dust samples
were collected in area schools in one year, Pb-based paint assessments were completed at
10 homes). These additional BHSS data collection efforts and studies are briefly described in
Appendix B.
Section 3 Preparation and Selection of BHSS Paired Dataset
This section describes how children's BLL and Pb exposure concentration data for the Box and
the Basin were paired for use in this IEUBK v2.0 model evaluation. In general, the data used in
14

-------
Advancing Pb Exposure and Biokinetic Modeling
this model evaluation were prepared in a manner consistent with previous BHSS risk
assessments, and risk management and dose-response evaluations, except community soil Pb
concentrations.5
Appendix E includes summary tables of all of the available paired blood, yard soil, and vacuum
Pb concentrations for 6-month-old to 6-year-old children residing in the Box and in the Basin
between 1988 and 2018. These data were carefully considered for use in this model evaluation;
however, only a subset was selected as described in Section 3.5.
Sections 3.1 through 3.4 describe important information about how the site-specific data were
prepared in order to pair the environmental exposure variables with BLLs.
3.1 Blood Lead Data
The BLL data consisted of a mix of venous samples analyzed at a detection limit of <1 |jg/dL
and capillary blood analyzed at detection limits of <1.4 or <1.9 |jg/dL, depending on the type of
capillary blood testing equipment used (see Section 2.1). BLLs were recorded to either the
nearest tenth (capillary) or whole number (venous). If a confirmatory venous sample result was
available, the venous sample result was used as the BLL for that child; otherwise, the capillary
result was used as the primary BLL. All BLLs were restricted to children <7 years old as that is
the range simulated in the IEUBK model.
If a child participated in the LHIP in more than one year, the BLL for each year (and the age of
the child at that time) was retained in the dataset. If multiple children participated in the LHIP
from the same home, each child's BLL was retained in the dataset.
The BLLs were then paired contemporaneously with house dust data (described in Section 3.2)
and any available property soil and private drinking water data (described in Sections 3.3 and
3.4, respectively) by the address in which the child resided at the time he/she participated in the
LHIP.
3.2 House Dust Data
All estimates of house dust Pb concentrations were based on vacuum samples. Dust mat data
were not included in the analysis. All available vacuum data in the Box and the Basin, as
described in Section 2.2, were paired contemporaneously with available PbB results. For
example, a vacuum sample collected from a home in 2006 was paired with PbB results for
children residing in that home in 2006.
When a dust sample result was non-detect, the sample was assigned a concentration of half the
detection limit. When a split dust sample was collected for QA/QC purposes, the lower
concentration of either the original or split was removed from the data, and the higher Pb
concentration was retained.
For the Box, if multiple vacuum samples were collected from the same home in a given year,
the highest sample result was retained. This approach may have introduced upward bias in the
vacuum dust concentrations in the Box; however, it was likely to be of minor consequence since
the number of homes for which multiple vacuum samples were collected in the same year was
low. For the Basin, an arithmetic average was calculated for this evaluation when there were
multiple vacuum samples collected from a home in the same year. Similarly, the number of
homes with multiple vacuum samples collected in the same year was low.
5 This model evaluation used arithmetic mean community soil Pb concentrations, whereas previous BHSS
analyses used geometric mean community soil Pb concentrations.
15

-------
Advancing Pb Exposure and Biokinetic Modeling
3.3 Soil Data
Soil sample results were paired with a blood and dust Pb result, regardless of the year in which
they were collected, until the year the property was remediated. A soil value of 100 mg/kg was
applied to remediated properties to estimate post-remedial property soil Pb concentrations and
community mean Pb soil concentrations.
Surface (0-1 inch) yard soil data collected from Box properties between 1991 and 2006 were
used to pair with blood and dust data; however, when no other soil data for a property existed,
soil samples collected by PHD and the State in 1986 were used.
For the Basin, surface (0-1 inch) soil data collected from 1998 to 2000 and from 2002 to 2018
were used to pair with blood and dust Pb data. Although a large yard soil sampling effort was
undertaken in 1996 for the Coeur d'Alene River Basin Environmental Health Exposure
Assessment (Idaho Department of Health and Welfare, 2000), those data were not used to
make remediation decisions for the Basin and were not included in community-wide soil mean
calculations. However, the 1996 soil data were retained and paired with a contemporaneous
dust and blood data for this evaluation when no other soil sampling results existed.
Residential soil sampling conducted by USEPA from 1998 through 2000 for the Basin RI/FS
involved the collection of multiple discrete soil samples from each sample area.6 However,
residential soil sampling conducted in 1999 and from 2002 through 2018 involved the collection
of composite soil samples from various locations within a property (yard, ROW, garden, etc.).
Consequently, the 1998-2002 Pb results from discrete sample areas were averaged to yield a
single concentration per area. The average soil Pb result per area was comparable to
composite sample results generated by the 1999 and 2002-2009 Basin sampling efforts.
In some cases, a Basin residence may have been sampled in multiple years and/or more than
one yard, ROW, or other discrete area may have been sampled at the property. When this
occurred, all sample results were retained and used in the calculation of arithmetic community-
wide and property averages. The spatially-weighted approach described in Section 2.3.2 was
applied to the calculation of community-wide and property soil averages for this evaluation.
In the Basin, when a property required soil resampling (i.e., the Pb concentration was in the
range of 900-999 mg/kg while the arsenic result was <100 mg/kg, or the arsenic concentration
was in the range of 60-99 mg/kg and the Pb concentration was <1000 mg/kg), the higher of the
two results was retained in the dataset.
In some cases, a Basin residence may have been remediated in multiple years. In these
instances, the most recent remediation year was used for soil Pb concentration calculation
purposes. For example, if a property was remediated in 2007 and again in 2010, the property
was assigned the remediation year of 2010, and a clean soil value of 100 mg/kg was applied to
remediated areas of the property in all post-2010 soil Pb concentration estimations.
In both the Box and the Basin, when soil sample results were below detection limits (BDL), the
sample was given a concentration of half the detection limit prior to data summary. When a
duplicate or split soil sample was collected for QA/QC purposes, the lower concentration of
either the original or duplicate/split was removed from the data, and the higher Pb concentration
was retained. Although the higher soil Pb concentrations in these cases may have biased soil
6 XRF soil Pb concentration data collected in 2001 by USACE were inconsistent with other BHSS soil
sampling approaches, were not used in risk assessment or evaluation of dose-response relationships in
the Basin, and were therefore not included in the dataset for this evaluation.
16

-------
Advancing Pb Exposure and Biokinetic Modeling
exposures high, annual data validation summaries generally indicated low relative percent
differences and these data were unlikely to substantially influence the results of this evaluation.
3.4	Drinking Water Data
There were no private drinking water data for children residing in the Box. Consequently,
drinking water Pb concentrations were assigned the IEUBK model default values, as presented
in Appendix A, for all child PbB records in the Box and for all Basin children residing at homes
connected to a public drinking water source (i.e., no private drinking water data, as described in
the following paragraphs).
Drinking water data were only available for use in this evaluation from a limited number of Basin
homes (where a private water source exists). Water sample results were paired with a PbB
result regardless of the year in which they were collected (for example, water samples collected
from a particular home in 2006 were paired with PbB results from a child residing in that home
in any year).
The hierarchy for selecting Basin drinking water data to pair with BLLs was as follows:
•	Samples from a kitchen sink
•	Samples from a different location within the house (when no sample from a kitchen sink
was available)
•	Samples at the well source (when no sample from within the house was available)
The IEUBK guidance recommended using a weighted average from home initial and purged
and school/daycare fountain water (U.S. EPA, 2007b). However, drinking water data from BHSS
elementary schools and/or daycares were not available to pair with PbB observations (and it
was not possible to know what school a child attended). Consequently, the child's drinking water
intake was assumed to be the average of the initial and purged residential water sample results
for this model evaluation, as agreed upon by Alta, SRC, and USEPA. Because the PbB sample
was collected in late summer when children were not attending school, lack of school drinking
water information was unlikely to influence the results of this evaluation.
When a water sample Pb concentration was BDL, the sample was given a concentration of half
the method detection limit.
3.5	BHSS Data Selected for Model Evaluation
The selected subset of BHSS data used for this IEUBK v2.0 model evaluation included Box data
from 1995-2018 and Basin data from 2002-2018. This subset was selected to incorporate
variability in soil Pb concentrations (Figures 3 and 4) and maximize the number of recorded
BLLs <5 |jg/dL. The paired data subset included a total of 1283 records, with BLLs collected
from 958 individual children (Table 2).
17

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 3. Soil Pb concentration boxplot for paired records (mg/kg) in the Box,
1995-2018
1995 1996 1997 1998 1999 2000 2001 2002 2003-2008 2013 2018
Year
Note: 35 soil concentration records >2,000 rng/kg were removed on this plot.
18

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 4. Soil Pb concentration boxplot for paired records (mg/kg) in the Basin, 2002-
2018
4000
2002-2005 2006 2007 2008 2009 2010 2011 2013 2015 2017 2018
Year
Note: two soil concentration records >4,000 mg/kg were removed on this plot.
Table 2. Number of individual children and homes in the paired dataset used for
model evaluation

Number of Children*
Number of Homes
Site-wide
958
601
Box
708
448
Basin
279
153
Notes:
*29 children were in both the Box and Basin, presumably because they moved from one region to the other in the
timeframe when the data were collected.
19

-------
Advancing Pb Exposure and Biokinetic Modeling
Section 4 Model Evaluation Approach
This section presents the IEUBK inputs and the methods used to evaluate the IEUBK v2.0
model. Although the primary focus of this model evaluation is on IEUBK v2.0; inputs for IEUBK
v1.1 are also described for reference. Two distinct IEUBK versions were employed:
•	lEUBKwin v2.0, build 1.67 (v2.0; which is the new draft version and focus of this model
evaluation)
•	lEUBKwin v1.1, build 11 (v1.1; which is the current available model)
4.1 IEUBK Inputs
Site-specific inputs were used for soil, dust, soil/dust bioavailability, and drinking water where
available. All other inputs (air, dietary intake, IRsd, and bioavailability for exposure media other
than soil/dust) used IEUBK default parameters recommended by the Technical Review
Workgroup (TRW) Lead Committee for IEUBK v2.0 and v1.1 models (Appendix A). The
parameter inputs used in the model runs for this evaluation are described below for each
exposure medium. The batch-mode application of the IEUBK model was applied.
4.1.1	Air
The IEUBK model default outdoor air concentration of 0.1 |jg/m3 was used in place of a site-
specific value due to unreliable site-specific air concentration data (described in Section 2.6). In
addition, default indoor air concentration and ventilation rates were used.
4.1.2	Dietary Intake
IEUBK model default assumptions were applied (see Tables A-2 and A-3). No site-specific data
were available.
4.1.3	Drinking Water
Drinking water concentrations for 27 individual properties in the Basin were used, as described
in Section 3.4; otherwise, the default drinking water concentration for Pb was used
(Appendix A).
4.1.4	Soil/Dust Ingestion Rates
Two separate sets of age-specific IRsd were used for the IEUBK v2.0 model evaluation (see
Table A-3):
•	default values in IEUBK v2.0, from (von Lindern et al. (2016): hereinafter referred to as
v2.0 IRsd), and
•	the USEPA Exposure Factors Handbook ((U.S. EPA. 2017a); hereinafter referred to as
EFH ingestion rates).
For the IEUBK v1.1 model, two IRsd values were used (see Table A-2):
•	v2.0 IRsd from von Lindern et al. (2016).
7 As noted earlier, differences in build number of the IEUBK v2.0 model software had no impact on model
predictions and did not affect the validity of the findings of this report.
20

-------
Advancing Pb Exposure and Biokinetic Modeling
• default values in IEUBK v1.1 (U.S. EPA. 2007b).
4.1.5 Exposure Media Bioavailability
Site-specific soil and dust bioavailabilities summarized in Table 1 were used to calculate total
bioavailable Pb concentrations in soil and dust for each property with PbB data for a resident
child. Total bioavailable Pb concentrations were calculated by multiplying bioavailabilities in
Table 1 with measured soil or dust Pb concentrations. Consequently, the soil and dust exposure
variables in the batch mode files accounted for site-specific soil/dust bioavailability and,
therefore, the parameters for soil and dust bioavailability were changed in the model's user
interface to 100% for the model runs.
All other exposure media bioavailability inputs remained at default values (see Table A-1).
4.1.6 Dust and Soil Partitions
This model validation used two dust/soil ingestion weighting factors, or partitions: (1) IEUBK
default 55% dust / 45% property soil (hereinafter referred to as the 55/45 partition), and (2) the
Box Model 40% dust / 30% property soil / 30% community soil (hereinafter referred to as the
40/30/30 partition). The 40/30/30 partition incorporates the concept that, in addition to residential
soils, children were exposed to Pb in soils in their neighborhood or community (von Lindern et
al.. 2016; U.S. EPA. 2005b; von Lindern et al.. 2003a). The child's soil exposure variable was
calculated prior to input into the IEUBK model by averaging the property-specific bioavailable Pb
with the arithmetic average community-wide bioavailable Pb (30% property soil and 30%
community soil was equivalent to averaging those two variables as input to the model).8 9
4.1.7 Weighted Media Concentrations
Apart from the 40/30/30 dust/soil partition and property-specific drinking water data, no other
weighted media concentrations were applied.10
4.1.8 Lead in Paint
Data on Pb paint in the BHSS were more limited (briefly described in Appendix B), and although
questionnaire data existed regarding the participant's knowledge of Pb paint in the home, these
8 In the Basin, property-specific bioavailable Pb soil values were spatially weighted arithmetic averages.
9 A third dust/soil partition (50/25/10/15) may be applied in future model evaluations. This partition
scenario was derived from Box paired data from 1988-2002 (von Lindern et al.. 2016) and required a level
of property sampling that may not always be available at other sites. The 50/25/10/15 dust/soil partition
represented the following contribution to soil Pb concentrations: 25% from individual property soil, 10%
from neighborhood soil (e.g., the arithmetic mean soil concentration from all properties within 500 feet of
the child's home), and 15% from community soil (e.g., the arithmetic mean soil concentrations from all
properties within a specified community).
10 If a substantial fraction of the child's activity occurred outside the home and data were available, media
concentrations should be weighted based on primary residence and secondary residence or daycare/
school. The LHIP questionnaires (Appendix B) provided data for some children regarding the length of
time spent at their residence versus other locations, the addresses of the other locations, and the amount
of time spent outdoors (although specific locations were not necessarily documented). Weighted media
concentrations based on where a child spent his/her time were not applied in this model validation due to
time and budget constraints. However, this information could be reviewed in the future.
21

-------
Advancing Pb Exposure and Biokinetic Modeling
data were not used in this IEUBK model evaluation. The IEBUK default input value of 0 |jg/dL
was used for Pb paint.
4.2 Model Evaluation Methods
The intended application of the IEUBK model is to predict a plausible distribution of PbB
concentrations centered on the GM BLL for a typical child living in a residence or for a
population of similarly exposed children (U.S. EPA, 1994a). A lognormal probability model was
used to calculate the probability of occurrence of a specified BLL (e.g., 5 |jg/dL) based on the
parameters GM and GSD,.
Evaluation of IEUBK v2.0 model performance included comparisons of model predictions to
empirical BHSS population data using the following three metrics:
• population GM BLLs,
• population probabilities of the BLL exceeding 5 |jg/dL, and
• the individual child BLL distributions.
Table 3 presents the main differences in the four IEUBK v2.0 model configurations using
alternative parameter assumptions of IRsd and soil/dust partitions. In general, the model
evaluation followed similar approaches to those described in Hogan et al. (1998) and von
Lindern et al. (2016).
Table 3. Summary of IEUBK model configurations and batch mode files
Parameter
IEUBK v2.0 software
IEUBK v1.1 software*
Dust/soil partition
I
55/45
40/30/30
55/45
40/30/30
IRsd
v2.0t
EFH (U.S.
EPA. 2017a)t
v2.0t
EFH (U.S.
EPA. 2017a)t
v2.0t
V1.1*
v2.0t
V1.1*
Soil/dust bioavailability
site-specific
site-specific
BLL cutoff value (pg/dL)
5.0
5.0
Strata (categories in
each batch file)
Box, Basin, Upper Basin, Lower Basin Box and Basin geographic areas
further broken down by calendar year and age
Model run
1
2
3
4
5
6
7
8
Notes:
*IEUBKv1.1 outcomes presented in Appendices F through H.
"•"Appendix A, Table A-3 presents age-specific IRsd values for IEUBK v2.0.
^Appendix A, Table A-2 presents age-specific IRsd values for IEUBK v1.1.
In addition, model prediction sensitivity was evaluated using alternative values for the following
parameters or assumptions: GSD,, dietary intake, clean soil backfill, and IEUBK v1.1 IRsd.
The predicted BLLs were rounded to one decimal place (tenth) to more closely correspond to
the significant figures reported for observed BLLs prior to conducting the analyses described
hereinafter.
Statistical analyses were performed using SAS/STAT® software, Version 9.4 of the SAS System
for Windows.
22

-------
Advancing Pb Exposure and Biokinetic Modeling
4.2.1	Geometric Mean BLL Comparisons
Observed and lEUBK-predicted GM BLLs were compared in order to evaluate model
conformity. The following steps were completed for this performance metric:
•	Observed and predicted GM BLLs and corresponding 95% confidence intervals (CIs)
were calculated.
o Observed and predicted GM BLLs and corresponding 95% CIs were calculated on
the log-transformed BLLs (SAS PROC MEANS) and outcomes were then
exponentiated.
o Summary tables present the GM BLLs and CIs for all model configurations by
multiple data categories (e.g., geographic area, age, calendar year).
o Figures show observed and predicted BLLs, CIs, and a line of unity (similar to Figure
1 in Hogan et al., 1998) for all model runs by multiple categories (e.g., age).
•	The differences (predicted minus observed) of GM BLLs were then calculated.
o Summary tables presented differences by various categories (e.g., geographic area,
age, calendar year), with the smallest difference in each category highlighted.
•	Criteria used to indicate model performance and best fit models include identification of
the smallest differences (in particular those differences that were <±1 |jg/dL) between
observed and predicted GM BLLs and/or overlapping observed and predicted CIs.
•	Weighted regression analysis was performed to evaluate correlation between observed
and predicted GM BLLs for the default IEUBK v2.0 model (55/45 partition with v2.0
IRsd). Regression output, such as the r2, was used to evaluate model performance (e.g.,
the high r2 values near one indicate more variance explained by the regression model).
Weighted regression analysis of BLLs (site-wide by age) was performed using ordinary
least squares regression or weighted linear regression (SAS PROC NLIN) with
uncorrelated weights assigned to each sample (Wi) calculated from Equation 1:
w _ w(Xj) x w(Yj)
1 w(Xi)+w(Yi) x (|32)
where w(Xj) and w(Yi) are the weights (1/SEi2) for the independent and dependent
variables, respectively, and p is the slope of the linear regression line fit by minimizing
the weighted sum of squared residuals (York et al., 2004). Studentized residuals (>2.5 or
<-2.5) were used to identify any statistical outliers. Weighting standard errors (WSEs)
were calculated from the 95% CIs on the mean (Equation 2):
WSE =	[2]
t(a,df)
where CI is the 95% CI and t (a, ^ is the t-value at a=0.05 and degrees of freedom
(df) = n-1.
4.2.2	Probability of Exceeding 5 fjg/dL Comparisons
lEUBK-predicted probabilities of a BLL >5 |jg/dL (Ps) were compared with the percentage of
observed BLLs >5 |jg/dL. The following steps were completed for this performance metric:
•	The percentage of observed BLLs >5 |jg/dL were used to estimate the Ps for the
observed BLLs. The Wald method (Snedecorand Cochran, 1989) using the 'Wald'
(Normal) interval was used to calculate the 95% CIs (Equation 3). When this method
23

-------
Advancing Pb Exposure and Biokinetic Modeling
resulted in a lower CI <0% or an upper CI >100%, due to a small sample size or an
average exceedance near zero, the Wilson score (Newcombe, 1998) interval method
was used instead (Equation 4).
Wald interval = p ±z,
P(l-P)
a/U
[3]
Wilson score interval (w ,w+)
[4]
where p is the observed percentage or predicted probability to exceed 5 |jg/dL, 2 is
the Z-score for the standard normal distribution at percentile a/2, a is the significance
level of 0.05, and n is the number of observations.
•	Individual lEUBK-predicted P5 values were averaged and 95% CIs were calculated using
the Wald (Normal) and Wlson score interval methods described above. The average
exceedance probability was treated as a binomial probability for the purpose of
estimating 95% CIs (as in Hogan et al., 1998).
o Summary tables present the average percentage, with 95% CIs, of observed BLLs
>5 |jg/dL and average lEUBK-predicted exceedance probabilities, with 95% CIs, for
all model configurations by category.
•	The differences (predicted minus observed) of the P5 and observed percentages were
then calculated.
o Summary tables present the differences by category, with the smallest difference
highlighted for each category.
•	Criteria used to indicate model performance and best fit models include identification of
the smallest differences between observed and predicted P5 and/or overlapping
observed and predicted CIs.
•	Weighted regression analysis of average observed percentages and average predicted
exceedance probabilities (P5) by age, using site-wide data, was also performed as
described in Section 4.2.1.
4.2.3 Evaluation of Individual Child BLL Distributions
For this performance metric, two analyses were conducted to evaluate child BLLs and their
distributions including: (1) the number of observed BLLs that fall outside 95% prediction limits of
IEUBK model predicted GM BLLs, and (2) differences between the empirical distributions of
observed and IEUBK predicted GM BLLs.
24

-------
Advancing Pb Exposure and Biokinetic Modeling
4.2.3.1	Prediction Intervals for IEUBK Predicted Mean BLLs
The IEUBK model was not intended to predict individual child BLLs; however, when applied to
predicting BLLs at an individual property (as is the case when using the batch mode), the
expectation was that BLLs from "hypothetical" children living in that home would fall within the
prediction limits (or the PI) of the model. For this evaluation, lower and upper 95% Pis for
predicted GM BLLs were defined as the 2.5th and 97.5th percentiles of the lognormal distribution
centered around the mean predicted BLL and the GSD/, where the GSD, represents the
expected variance in BLLs among similarly exposed children (e.g., children exposed at the
same home environment; (Hogan et al., 1998)). The following steps were completed:
•	95% Pis were calculated for each predicted GM BLL from the IEUBK batch mode output
using Equations 5a and 5b.
o Summary tables present the percentage of observed BLLs that fall outside the 95%
Pis of the lEUBK-predicted GM BLLs for all model runs by category, with the lowest
percentages highlighted to indicate better model performance relative to observed
BLLs. Ideally, <5% of the observed BLLs would fall outside Pis.
o Scatter plots show observed and predicted BLLs with the 95% Pis (similar to Figures
2-4 in Hogan et al., 1998) for all model runs and select categories.
95% lower PI =	(G5£,0xZo.o25)	[5a]
95% upper PI = g(ln(GM)+ln (G5D0xZo.97s)	[5b]
Where:
GM = geometric mean PbB concentration
GSDj= individual geometric standard deviation (1.6)
Z0.025, Z0.975 = Z-score for the standard normal distribution at the 2.5th and
97.5th percentiles
4.2.3.2	Differences Between Empirical BLL Distributions
The differences between the empirical distributions of observed and predicted BLLs were also
evaluated. The following steps were completed:
•	The nonparametric two-sample Kolmogorov-Smirnov (denoted K-S) test was performed
(SAS PROC NPAR1WAY). The K-S test statistic (D) quantifies distance between the
empirical cumulative distributions of observed and lEUBK-predicted GM BLLs, focusing
on the largest vertical distance between the two distributions. The null hypothesis
assumes that the cumulative distributions of observed and predicted GM BLLs are from
the same population. The assumptions for the K-S two-sample test are that observed
and predicted GM BLLs are random and mutually independent; however, in this case,
both were, in part, dependent upon soil and residential dust Pb exposures (for example,
an increase in dust concentration at a home could have an impact on both an observed
and a predicted BLL in the same way). Since both the observed and predicted BLL
depend on soil and dust, they are not necessarily mutually independent. Nonetheless,
the two-sample K-S test was used to test for differences in the observed and predicted
BLL under the null hypothesis of no difference between their cumulative distributions,
F(x). The K-S test was applied using Equation 6 to determine the K-S test statistic (D)
and Equations 7a and 7b to estimate the critical deviation (Dcriticai) between distributions
25

-------
Advancing Pb Exposure and Biokinetic Modeling
required to reach statistical significance at the level of a and the probability of a
significant difference between distributions, respectively:
Maximum deviation: D = max|F] J — F2 (x, )|
Where j = l, 2	n
Where ni and ni are the number of observed and predicted BLL, respectively,
n = rh + ri2, and a is the significance level of 0.05
00
Pvalue: Prob (D > d) =	y~1e(~2i2zZ)
i-1
Where:
¦	/' indexes the number of terms used to approximate the probability of a
maximum difference (D) greater than the maximum difference observed
(d).11
¦	z = d^nin2/n
o Summary tables present K-S test results by geographic area for all IEUBK v2.0
model configurations.
o Figures show the cumulative distributions for observed and predicted GM BLL for all
model runs as specified above.
o Criteria used to indicate model performance include whether the K-S test rejects the
null hypothesis and the smallest maximum difference (d) between distributions. Due
to the large number of samples used for this evaluation, the critical value (Dcriticai) was
relatively small (Equation 7a) leading to rejections of the null hypothesis for values of
d that may not be meaningful for Pb risk assessment. Subsequently, visual
comparisons of the cumulative distributions were also discussed.
4.2.4 Additional Supporting Information of Model Predictions
Although the following analyses were not used to evaluate the IEUBK v2.0 model because they
were completed for each individual paired record, they are included in appendices as supporting
information.
• A difference was calculated for each observed and predicted BLL pair (log transformed).
The difference was then squared and summed for all model runs by category. This is
referred to as the sum of squared differences (SSD). Tables F-4, G-4, and H-4 present
the SSDs, with highlights indicating the lowest SSD.
11 It is unclear from SAS documentation how many terms (i.e., summations) are used in the
approximation, but 10 was sufficient to reproduce the probability (to 1E 04) reported by SAS; for example
2 (arthritis data) in the NPAR1 WAY documentation (see
https://support.sas.com/documentation/onlinedoc/stat/141/npar1way.pdf).
26

-------
Advancing Pb Exposure and Biokinetic Modeling
• A proportion was calculated for the number of records where the absolute difference
between a pair of observed and predicted BLLs was <1 |jg/dL. Tables F-5, G-5, and H-5
present these percentages for all model runs by category.
• Goodness-of-fit between the observed and IEUBK model predicted GM BLLs was
assessed by calculating the following (this analysis was completed using censored data,
more specifically, censor level 2; see Section 5 for more information on censored data).
o The ratio of the sum of squares error (SSE) to the sum of squares total (SST) was
calculated (SSE/SST). Equations 8 and 9 were used for the calculation of SSE to
SST:
SSE = 2J.GMi - BLLO2 [8]
i=i
SST - - BLL)2 [9]
f=i
Where:
¦ GMi = geometric mean predicted by the IEUBK model for child /'
¦ BLL= measured PbB concentration for child /'
¦ BLL = arithmetic average of measured PbB concentrations
Note: SST + SSR (i.e., sum of squares regression) + SSE for the IEUBK evaluation
because SSE may exceed SST since SEE was not minimized by fitting model
predictions to the observed PbB.
o Summary Table G-6 presents SSE/SST ratios for all model runs and by geographic
area, with the lowest ratio highlighted indicating better fit of observed and model
predicted GM BLLs.
4.2.5 Sensitivity Analyses
The sensitivity of certain parameters and assumptions was evaluated including the
concentration of clean soil, GSD,, diet intake, and IEUBK v1.1 IRsd.
4.2.5.1 Clean Soil Values
In the Box, the PRPs reported annual backfill soil averages as less than 55 mg/kg (McCullev.
1997). Therefore, a value of 50 mg/kg for remediated soil was also evaluated. Sensitivity
analysis using this lower clean soil value was accomplished by comparing IEUBK v2.0 model
results from two sets of Box batch mode files (one data set used 50 mg/kg, and the other used
100 mg/kg). As described previously, remediation in the Box generally consisted of all or most
of the parcel (as opposed to partial property remediation in the Basin and the use of average
property soil exposure values). Consequently, lowering the clean soil value in the Box data set
was expected to result in the largest change to the soil exposure variable and subsequent
predicted BLLs. Prior to this analysis, the Box soil data set was reworked by assigning the value
of 50 mg/kg as the remediated soil exposure concentration. Annual community averages were
also calculated using 50 mg/kg to prepare the soil input values for the 40/30/30 partition. The
27

-------
Advancing Pb Exposure and Biokinetic Modeling
Box batch mode input files were updated with these new soil exposure values. All other
exposure inputs and model parameters remained the same as described previously.
• Summary tables present predicted GM BLLs, P5 values, and CIs for both data sets
where one applies a clean soil value of 100 mg/kg and the other applies a clean soil
values of 50 mg/kg. Methods are similar to those described in Sections 4.2.1 and 4.2.2.
4.2.5.2 Geometric Standard Deviation
The GSDi sensitivity analysis followed a two-step process. First, a new GSDi value was
estimated such that 95% of the individual BLLs were located within the resulting model Pis (see
Equations 5a and 5b). As GSD, increased, Pis widened and included a larger number of
observations. The P5 was then recalculated using the adjusted GSD,.
• Summary tables present the adjusted GSD, that results in 95% of the observations
falling within the model Pis, as well as the newly estimated P5 and CIs. These outcomes
were compared to the original model results.
4.2.5.3 Diet
In addition to changes in the IRsd in IEUBK v2.0, other adjustments were made to model
parameters that govern exposures and intakes of Pb in air, diet, and drinking water.
The diet sensitivity analysis used default IEUBK v2.0 model values for intake of Pb in market
basket foods. These values were derived from data from the What We Eat In America (WWEIA)
dietary interview component of the National Health and Nutrition Examination Survey (NHANES;
2003-2006) (CDC, 2010a) (CDC, 2010b) and the Food and Drug Administration (FDA) Total
Diet Study (TDS) food contaminants data for 1995-2005 (FDA. 2010). Zartarian et al. (2017)
used dietary Pb intakes based on the FDA's TDS data 2007-2013. The effect on predicted GM
BLLs and P5 values was evaluated by using different values for dietary Pb intake, including
IEUBK v2.0 default dietary Pb intakes and dietary Pb intakes reported by Zartarian et al. (2017).
To isolate the effect of dietary Pb intake on predicted GM BLLs, Pb intakes due to soil, dust,
water, and air were set to zero.
• Tables and figures present predicted mean BLLs using these different dietary Pb intake
values.
4.2.5.4 Effect of Updated IRSD
The effect of changing the IRSd in IEUBK v2.0 was evaluated by comparing two sets of ingestion
rates (IEUBK v1.1 compared to v2.0 IRsd) using the mean site soil and dust concentration
values of 413 and 598 mg/kg, respectively, in the IEUBK v2.0 model software. Results are
summarized in a table and figure.
In addition, the original scope of this evaluation included IEUBK v1.1 model runs using both v1.1
and v2.0 IRsd. The mean BLLs of the IEUBK v1.1 software using v1.1 and v2.0 IRsd values
were plotted with v2.0 software using v2.0 IRsd and observed mean BLLs and 95% CIs. Means
and CIs that were nearest to each other can be visually examined.
28

-------
Advancing Pb Exposure and Biokinetic Modeling
Section 5 IEUBK Model Evaluation Results and Discussion
The results and evaluation metrics were summarized by various categories including child age,
geographic area, and calendar year. A child's age factors into the IEUBK exposure and
biokinetic components of the model and was therefore an important category by which to
summarize results and review model performance. Model results were also summarized by
geographic area because exposures at the BHSS have previously been shown to vary by
location (e.g., residents living in the Box were more impacted by smelter emissions, whereas
residents in the Lower Basin were more impacted by sediment deposition or mechanical
tracking/movement of soils), and by calendar year due to differences in Pb exposure as property
soil remediation progressed.
This performance assessment focused on the default IEUBK v2.0 model that assumes the
55/45 soil and dust partition and v2.0 IRsd derived from von Lindern et al. (2016). At the BHSS,
alternative partitions and IRsd were derived from the paired Box data (1988-2002) to provide
more realistic exposure scenarios for a given child based on soil exposures beyond the
residential property boundary and within the community (von Lindern et al.. 2016; von Lindern et
al.. 2003a). These alternative partitions and IRsd showed improved conformance between
predicted and observed mean BLLs in past BHSS evaluations. The partition used in the Box
model, described in Section 1.4 of this report, assumed that 40% of a child's total dust/soil
ingestion was derived from house dust, 30% was derived from residential yard soils, and 30%
was derived from community-wide soils (referred to as a 40/30/30 partition). The USEPA also
derived children's IRsd based on averaging multiple studies in the EFH (U.S. EPA, 2017a).
Although the focus of this report is on how well the default conditions in the IEUBK v2.0 model
perform, both partition scenarios (55/45 and 40/30/30) and both IRsd values (v2.0 and EFH) are
also presented and discussed.
The batch mode output data were first reviewed and censored for certain criteria. Only the
second level of censored data are discussed in this section for reasons described hereinafter.
• Censor level 1: any predicted or observed BLLs >30.5 |jg/dL were removed from this
data subset prior to model evaluation. Two records were removed because predicted
BLLs exceed model performance limits (U.S. EPA. 1994c).
o This censor level resulted in a site-wide total of 1281 paired records, and results are
presented in Appendix F.
• Censor level 2: In addition to censor level 1, observed BLLs that were BDL of the
method applied to the sample were also excluded from this data subset. These
137 records were removed because of the potential bias associated with comparisons of
model predictions to BDL BLLs (see Section 5.1 of this report for further discussion of
the BDL BLLs).
o This censor level resulted in a site-wide total of 1144 paired records, and these
results are presented and discussed throughout this section (and Appendix G).
• Censor level 3: In addition to censor level 2, observed BLLs that were >10.5 |jg/dL were
also removed from this data subset. These 53 records were omitted in order to focus
evaluation of model predictions at the lower range of observed BLLs.
o This censor level resulted in a site-wide total of 1091 paired records, and results are
presented in Appendix H.
29

-------
Advancing Pb Exposure and Biokinetic Modeling
The resulting data set from censor level 2 consisted of 1144 BLL observations derived from a
total of 853 individual children. The arithmetic mean BLL for the evaluation data set (n=1144)
was 4.4 |jg/dL (95% CI: 1.5-0 |jg/dL) and 75% of the BLLs were <5 |jg/dL (Table 4). Summary
statistics of the environmental exposure data associated with censor level 2 are also presented
in Table 4. Approximately 23% of the 853 children provided more than one BLL during different
years of the LHIP. Typically, this was two BLLs per child (n=136); a few children provided three
BLLs (n=36) or up to six BLLs (n=25). Repeated measurements from these individual children
and their corresponding soil and dust Pb concentrations were treated as independent
observations. In this approach, intra- and inter-individual variance in BLLs were not
distinguished. This was assumed acceptable given that the objective of the evaluation was to
determine how well the model predicted observed GM BLLs and Ps based on soil and dust Pb
concentrations measured near the time of each BLL screening. Multiple BLL measurements in
individual children also spanned periods in which the soil and dust Pb concentrations in the
child's local environment changed as a result of remediation at the BHSS; therefore, it was of
interest to include these observations in the performance evaluation dataset.
5.1 BDL BLLs
As described earlier, BDL BLLs were excluded from the data at censor levels 2 and 3 due to the
bias associated with model predictions of BDL BLLs. Restricting the analysis to values above
the detection limit made it unnecessary to assign values below the detection limit, which would
have introduced error of varying degrees into the observed BLL estimates for individual children.
Exclusion of the BLLs that were BDL would not appreciably affect evaluation of model
performance for predicting GM BLLs and Ps of the study population; however, it would introduce
some level of bias into inferences made about the BHSS population as a whole, which was not
the objective of this evaluation. Regardless, model evaluation results that included BLLs that
were BDL are presented in Appendix F of this report.
Of the paired records selected for model evaluation, 137 BLLs were BDL. In order to understand
the impact that BDL results may have on model outputs, these 137 records were assigned a
value of the capillary test detection limit (1.4 or 1.9 |jg/dL) and input into the IEUBK v2.0
model.12 The predicted GM BLLs and Ps values for these 137 records are summarized by
detection limit in Table 5.
Of the IEUBK v2.0 model configurations, the default model using the 55/45 partition and v2.0
IRsd resulted in the lowest GM predicted BLL and average Ps value (as did the 55/45 partition
using the EFH IRsd). Mean predicted BLLs of the default IEUBK v2.0 model were 2.5 and 2.4
|jg/dL relative to observed BLLs with detection limits of 1.4 and 1.9 |jg/dL, respectively. Overall,
the four IEUBK v2.0 model configurations (see Table 3) resulted in mean predicted BLLs
ranging from 2.3 to 3.0 |jg/dL (Table 5). Table G-7 in Appendix G presents IEUBK v1.1 model
configuration results for the same 137 BDL BLLs. The lower predicted GMs and average Ps
from IEUBK v1.1 using v2.0 IRsd, compared to v2.0 model configurations, are likely due to the
lower default dietary intakes in IEUBK v1.1 than v2.0.
12 No venous draw samples were BDL.
30

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 4. Summary of paired BLL and Pb exposures for study population, censor level 2 (dataset omits paired records
where an observed or predicted BLL is >30.5 pg/dL or an observed BLL is BDL)


Arithmetic
Geometric




Percentile





Standard

Standard







Paired Data by Category
N
Mean
Deviation
Mean
Deviation
Minimum
Maximum
5th
25th
50th
75th
95th
Site-Wide












BLL (Mg/dL)
1144
4.4
3.0
3.6
1.8
1.0
22.0
1.5
2.4
3.5
5.0
10.0
Property soil* (mg/kg)
1144
413
830
198
2.7
31.1
9180
100
100
100
383
1611
Soil 30/30* (mg/kg)
1144
608
557
454
2.1
125
5573
126
287
462
805
1438
House dust (mg/kg)
1144
598
1141
359
2.5
11.3
15300
82.3
213
345
610
1570
Tap water* (|jg/L)
55
2.3
2.0
1.8
1.8
0.9
8.8
1.0
1.5
1.5
1.5
7.5
Box (Total)












BLL (|jg/dL)
875
4.7
3.2
3.9
1.8
1.0
22.0
1.5
2.5
4.0
6.0
11.0
Property soil* (mg/kg)
875
389
836
179
2.7
31.1
9180
100
100
100
345
1620
Soil 30/30* (mg/kg)
875
628
572
463
2.2
125
5573
126
303
462
819
1535
House dust (mg/kg)
875
659
1240
405
2.4
17.1
15300
102
245
400
687
1600
Tap water* (|jg/L)
0
-
-
-
-
-
-
-
-
-
-
-
Basin (Total)












BLL (|jg/dL)
269
3.3
2.0
3.0
1.6
1.0
14.0
1.6
2.1
2.8
4.0
7.0
Property soil* (mg/kg)
269
493
809
280
2.6
44.4
6814
100
113
237
530
1611
Soil 30/30* (mg/kg)
269
544
500
427
1.9
133
3997
161
285
406
580
1267
House dust (mg/kg)
269
403
703
242
2.5
11.3
6000
52.0
153
241
351
1170
Tap water* (|jg/L)
55
2.3
2.0
1.8
1.8
0.9
8.8
1.0
1.5
1.5
1.5
7.5
Site-Wide by Age (years)












0.5 to <1 year












BLL (|jg/dL)
70
4.5
2.9
3.8
1.7
1.5
14.0
2.0
2.6
3.6
5.0
10.0
Property soil* (mg/kg)
70
530
940
237
3.1
94
5363
100
100
108
520
2779
Soil 30/30* (mg/kg)
70
614
578
450
2.2
126
3122
126
266
450
655
2013
House dust (mg/kg)
70
432
509
283
2.5
21
3470
67.0
150
277
454
1370
Tap water* (|jg/L)
1
-
-
-
-
-
-
-
-
-
-
-
1 to <2 years












BLL (|jg/dL)
138
5.4
3.4
4.6
1.7
1.0
20.0
2.0
3.0
4.0
6.9
12.0
Property soil* (mg/kg)
138
492
1199
193
2.9
52.7
9180
100
100
100
345
2568
Soil 30/30* (mg/kg)
138
675
744
474
2.3
126
5573
130
296
473
819
2188
House dust (mg/kg)
138
586
776
384
2.4
48.4
6800
62.0
220
360
670
1600
Tap water* (|jg/L)
6
2.9
2.4
2.3
2.0
1.5
7.5
1.5
1.5
1.5
4.0
7.5
31

-------
Advancing Pb Exposure and Biokinetic Modeling


Arithmetic
Geometric




Percentile





Standard

Standard







Paired Data by Category
N
Mean
Deviation
Mean
Deviation
Minimum
Maximum
5th
25th
50th
75th
95th
2 to <3 years












BLL (Mg/dL)
202
4.8
3.2
4.0
1.8
1.0
18.0
1.6
2.8
4.0
6.0
11.0
Property soil* (mg/kg)
202
390
668
198
2.7
31.1
4140
100
100
100
435
1650
Soil 30/30* (mg/kg)
202
559
482
414
2.2
125
3015
126
252
450
701
1258
House dust (mg/kg)
202
474
589
313
2.5
11.3
6000
82.0
177
310
570
1380
Tap water* (|jg/L)
8
2.0
1.3
1.7
1.7
0.9
4.0
0.9
1.3
1.5
2.8
4.0
3 to <4 years












BLL (Mg/dL)
176
4.5
3.1
3.8
1.8
1.0
20.0
1.5
2.6
3.8
5.1
12.0
Property soil* (mg/kg)
176
437
861
199
2.8
66.6
5370
100
100
100
329
1760
Soil 30/30f (mg/kg)
176
636
589
466
2.2
125
3630
126
290
459
819
1729
House dust (mg/kg)
176
657
1336
371
2.6
22.0
15300
80.0
212
346
656
1880
Tap waterj (Mg/L)
13
2.5
2.6
1.8
2.1
0.9
8.8
0.9
1.1
1.5
1.5
8.8
4 to <5 years












BLL (Mg/dL)
187
4.2
3.0
3.5
1.9
1.0
22.0
1.5
2.1
3.1
5.0
10.0
Property soil* (mg/kg)
187
389
805
192
2.7
31.1
6814
100
100
100
416
1180
Soil 30/30* (mg/kg)
187
582
534
435
2.1
125
3997
126
258
462
713
1209
House dust (mg/kg)
187
613
1105
343
2.7
17.1
11200
62.0
210
328
590
1800
Tap water* (Mg/L)
12
1.9
1.8
1.6
1.7
1.0
7.5
1.0
1.3
1.5
1.5
7.5
5 to <6 years












BLL (Mg/dL)
185
3.9
2.6
3.3
1.8
1.0
17.0
1.4
2.0
3.0
5.0
8.0
Property soil* (mg/kg)
185
452
893
212
2.9
34.0
7370
100
100
100
431
1623
Soil 30/30* (mg/kg)
185
673
591
515
2.1
126
4565
139
367
491
930
1535
House dust (mg/kg)
185
644
1231
405
2.4
11.3
15300
111
250
406
652
1700
Tap water* (Mg/L)
7
2.8
2.7
2.2
2.0
1.5
8.8
1.5
1.5
1.5
3.3
8.8
6 to <7 years












BLL (Mg/dL)
186
3.6
2.4
3.1
1.7
1.0
17.0
1.4
2.0
3.0
4.0
8.0
Property soil* (mg/kg)
186
298
454
183
2.3
100
4140
100
100
100
345
1128
Soil 30/30* (mg/kg)
186
547
387
441
2.0
125
3015
126
326
462
656
1184
House dust (mg/kg)
186
690
1632
390
2.4
19.9
15000
94.3
250
390
590
1200
Tap water* (Mg/L)
8
1.9
1.1
1.7
1.6
1.0
4.0
1.0
1.3
1.5
2.4
4.0
Notes:
*Soil concentration input to IEUBK model for the 55/45 partition (property soil concentrations).
*Soil concentration input to IEUBK model for the 40/30/30 partition (average of the community-wide and property soil concentrations, per property).
*Site-specific tap water samples were included where available.
BDL: capillary detection limits were 1.4 and 1.9 |jg/dL and the venous detection limit was 1.0 |jg/dL.
- No data were available.
32

-------
Advancing Pb Exposure and Biokinetic Modeling
5.2 Prediction of Population GM BLLs
Table 6 and Figures 5 through 8 present observed and predicted GM BLLs with 95% CIs for
censor level 2 (similar tables and figures for censor levels 1 and 3 are presented in
Appendices F and H, respectively).
5.2.1	Performance of Default IEUBK v2.0 Model
The site-wide predicted GM BLL from the default IEUBK v2.0 model was 3.4 |jg/dL (95% CI:
3.3, 3.5) and was within 0.26 |jg/dL of the observed GM BLL of 3.6 |jg/dL (95% CI: 3.5, 3.8)
(Tables 6 and 7). Observed GM BLLs showed a distinct trend with age, which was captured by
the IEUBK model (Figures 7 through 9). As illustrated in Figure 9, the model explained
approximately 90% of the variance in the age-stratified GM BLLs (intercepts.32, slope=0.69,
r2=0.90). The open circle in Figure 9 represents the GM for children <7 years old and was not
included in the regression model since it is an aggregate of the other age strata.
When stratified by geographic area, the model performed quite well at predicting the observed
GM BLLs in the Upper and Lower Basin areas (Tables 6 and 7; Figures 5 and 6). CIs for
observed and predicted area GM BLLs overlapped for all areas except the Box (Figures 5 and
6). Differences between predicted and observed GM BLLs by geographic area (not age specific)
ranged from -0.44 |jg/dL in the Box to +0.36 |jg/dL in the Upper Basin (Table 7). Collectively,
these results suggest that population GM BLLs predicted by the default IEUBK v2.0 model
agreed well with observations; differences in observed and predicted GM BLLs were within
1 |jg/dL in all geographic areas and site-wide by age, and the model did not systematically over-
or under-predict observed GM BLLs.
5.2.2	Alternative IRsd and Soil and Dust Partitions
Predicted GM BLLs for alternative values for IRsd and soil and dust partitions are presented in
Tables 6 and Figures 5 through 8. When the IEUBK v2.0 default values for IRsd were replaced
with values recommended in the EFH IRsd, the predicted site-wide GM BLL was 3.3 |jg/dL (95%
CI: 3.2, 3.4), within 0.34 |jg/dL from observed (Table 7). The 40/30/30 partition models resulted
in larger differences between site-wide predicted and observed GM BLLs; however, the
differences for all four IEUBK v2.0 model configurations (see Table 3) were within 1 |jg/dL
(Table 7). These outcomes were similar for censor levels 1 and 3, site-wide (see Appendices F
and H, respectively).
Predicted and observed GM BLLs for child age groups are presented in Tables 6 and Figures 7
and 8. Site-wide, models that assumed the 55/45 partition provided better predictions of the GM
BLLs for ages <2 years compared to versions that assumed the 40/30/30 partition. Models that
assumed the 40/30/30 partition typically provided closer predictions of the observed GM BLLs
for children 2-<7 years old, with the exception of children 5-<6 years old. However, the
differences between predicted and observed GM BLLs in children 2 years of age and older were
generally within 1 |jg/dL regardless of the partition or IRsd applied (Table 7). Results for other
censor levels are presented in Appendices F and H.
33

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 5. IEUBK model output summary of observed BLLs recorded as BDL




Number of
Percent of


Minimum Maximum Average Predicted 95% CI around
IEUBK Model and


Number of
Records where
Records where
Mean (GM)
95% CI
Predicted
Predicted
Probability of
Average
Dust/Soil

Detection
Observed
Predicted BLL
Predicted BLL
of Predicted
around
BLL
BLL
Exceeding 5 |jg/dL
Probabilities
Partition
IRsd
Limit
BLL BDL
BDL
BDL
BLL (|jg/dL)
GM
(Mfl/dL)
(Mg/dL)
(%)
(%)
IEUBK v2.0
\/0 n
1.4
109
0
0%
2.5
2.3, 2.7
1.4
14.7
12.8
6.6, 19.1

vz.u
1.9
28
9
32%
2.4
2.1, 2.7
1.4
7.4
10.5
3.6, 26.9f
55/45 dust/soil









ingestion weighting
EFH (U.S.
1.4
109
0
0%
2.5
2.3, 2.7
1.5
13.7
12.2
6.1, 18.4
factor
EPA. 2017a)
1.9
28
9
32%
2.3
2, 2.6
1.5
6.3
9.6
3.1, 25.7f
IEUBK v2.0
\/i n
1.4
109
0
0%
3.0
2.8, 3.3
1.6
12
20.4
12.8, 27.9

vz.u
1.9
28
5
18%
2.6
2.3, 2.9
1.5
6.2
12.3
0.2, 24.6
40/30/30 dust/soil*









ingestion weighting
EFH (U.S.
1.4
109
0
0%
3.0
2.8, 3.2
1.6
11.2
19.2
11.8, 26.5
factor
EPA. 2017a)
1.9
28
6
21%
2.5
2.3, 2.8
1.6
5.3
10.6
3.7, 27.1f
Notes:
*40% dust, 30% property soil, 30% community soil.
tWilson's method used to approximate CIs.
34

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 6. Summary of mean observed and predicted BLLs for IEUBK v2.0 model, censor level 2 (dataset omits
predicted and observed BLLs >30.5 pg/dL and observed BDL BLLs)










Predicted Blood Lead






Observed Blood Leadt


55/45 Dust/Soil Partition


40/30/30 Dust/Soil*
Partition







v2.0 IRsd
EFH (U.S. EPA. 2017a) IRSD

V2.0 IRsd
EFH (U.S. EPA. 2017a) IRSD



95% CI of



95% CI of


95% CI of


95% CI of


95% CI of



the
Geometric


the


the


the


the


Geometric Geometric
Standard

Geometric Geometric

Geometric Geometric

Geometric Geometric

Geometric Geometric


Mean
Mean
Deviation

Mean
Mean

Mean
Mean

Mean
Mean

Mean
Mean

N
(ug/dL)
(ug/dL)
(ug/dL)
N
(ug/dL)
(ug/dL)
N
(ug/dL)
(Ug/dL)
N
(ug/dL)
(ug/dL)
N
(ug/dL)
(ug/dL)
Site-Wide (Total)
1144
3.6
3.5, 3.8
1.8
1144
3.4
3.3, 3.5
1144
3.3
3.2, 3.4
1144
4.3
4.1, 4.4
1144
4.1
4.0, 4.3
Box (Total)
875
3.9
3.7, 4.0
1.8
875
3.4
3.3, 3.6
875
3.3
3.2, 3.5
875
4.4
4.2, 4.6
875
4.3
4.1, 4.4
Basin (Total)
269
3.0
2.8, 3.1
1.6
269
3.2
3.0, 3.4
269
3.1
2.9, 3.3
269
3.8
3.6, 4.0
269
3.7
3.5, 3.9
Upper Basin (Total)
209
2.9
2.8, 3.1
1.5
209
3.3
3.0, 3.5
209
3.2
3.0, 3.5
209
4.0
3.8, 4.3
209
3.9
3.7, 4.1
Lower Basin (Total)
60
3.13
2.7, 3.6
1.7
60
2.9
2.5, 3.4
60
2.8
2.4, 3.2
60
3.20
2.8, 3.6
60
3.05
2.7, 3.4
Site-Wide By Age (years)
0.5 to <1
70
3.8
3.4, 4.4
1.7
70
4.8
4.2, 5.6
70
4.2
3.7, 4.9
70
6.3
5.6, 7.0
70
5.4
4.8, 6.1
1 to <2
138
4.6
4.2, 5.1
1.7
138
5.0
4.5, 5.5
138
4.8
4.4, 5.3
138
6.5
6.0, 7.1
138
6.3
5.8, 6.8
2 to <3
202
4.0
3.6, 4.3
1.8
202
3.5
3.2, 3.7
202
3.3
3.1, 3.5
202
4.3
4.1, 4.6
202
4.1
3.8, 4.3
3 to <4
176
3.8
3.4, 4.1
1.8
176
3.3
3.0, 3.6
176
3.2
2.9, 3.5
176
4.2
3.9, 4.5
176
4.0
3.7, 4.3
4 to <5
187
3.5
3.2, 3.8
1.9
187
3.2
3.0, 3.5
187
3.0
2.8, 3.3
187
4.0
3.7, 4.3
187
3.7
3.5, 4.0
5 to <6
185
3.3
3.0, 3.6
1.8
185
3.0
2.8, 3.3
185
3.1
2.9, 3.4
185
3.8
3.5, 4.0
185
3.9
3.7, 4.2
6 to <7
186
3.1
2.8, 3.3
1.7
186
2.6
2.4, 2.7
186
2.7
2.5, 2.9
186
3.2
3.0, 3.4
186
3.4
3.2, 3.6
Box (Total) by Age
(years)
0.5 to <1
47
3.9
3.3, 4.6
1.7
47
4.7
3.9, 5.6
47
4.1
3.4, 4.9
47
6.4
5.6, 7.3
47
5.5
4.8, 6.4
1 to <2
112
4.9
4.4, 5.4
1.7
112
5.2
4.7, 5.8
112
5.0
4.5, 5.6
112
6.8
6.2, 7.5
112
6.5
6.0, 7.2
2 to <3
149
4.3
3.9, 4.8
1.8
149
3.6
3.3, 3.9
149
3.4
3.1, 3.7
149
4.5
4.2, 4.9
149
4.2
3.9, 4.6
3 to <4
137
4.0
3.6, 4.4
1.8
137
3.4
3.1, 3.7
137
3.2
2.9, 3.6
137
4.3
3.9, 4.7
137
4.1
3.8, 4.5
4 to <5
144
3.7
3.3, 4.2
1.9
144
3.2
2.9, 3.4
144
3.0
2.7, 3.2
144
4.1
3.8, 4.4
144
3.8
3.5, 4.1
5 to <6
144
3.4
3.1, 3.8
1.8
144
3.0
2.8, 3.3
144
3.1
2.9, 3.4
144
3.9
3.6, 4.2
144
4.1
3.8, 4.4
6 to <7
142
3.3
3.0, 3.6
1.8
142
2.6
2.4, 2.9
142
2.8
2.6, 3.0
142
3.3
3.1, 3.6
142
3.5
3.3, 3.8
Basin (Total) by Age
(years)
0.5 to <1
23
3.8
3.0, 4.7
1.7
23
5.1
3.9, 6.7
23
4.4
3.4, 5.8
23
6.0
4.8, 7.4
23
5.2
4.2, 6.4
1 to <2
26
3.6
2.9, 4.4
1.7
26
4.1
3.3, 5.1
26
3.9
3.2, 4.9
26
5.4
4.6, 6.4
26
5.2
4.4, 6.1
2 to <3
53
3.1
2.7, 3.6
1.7
53
3.1
2.7, 3.6
53
3.0
2.6, 3.4
53
3.8
3.4, 4.2
53
3.6
3.2, 4.0
3 to <4
39
3.0
2.5, 3.5
1.7
39
3.1
2.6, 3.7
39
3.0
2.5, 3.5
39
3.8
3.4, 4.4
39
3.7
3.2, 4.2
4 to <5
43
2.7
2.4, 3.0
1.4
43
3.3
2.8, 4.0
43
3.2
2.6, 3.8
43
3.7
3.2, 4.3
43
3.5
3.0, 4.1
5 to <6
41
2.8
2.5, 3.2
1.4
41
3.0
2.5, 3.5
41
3.1
2.6, 3.7
41
3.4
2.9, 3.9
41
3.5
3.0, 4.0
6 to <7
44
2.5
2.3, 2.8
1.4
44
2.4
2.1, 2.6
44
2.5
2.2, 2.8
44
2.8
2.5, 3.1
44
3.0
2.7, 3.3
35

-------
Advancing Pb Exposure and Biokinetic Modeling










Predicted Blood Lead






Observed Blood Leadt


55/45 Dust/Soil Partition


40/30/30 Dust/Soil*
Partition







V2.0 IRsd
EFH (U.S. EPA. 2017a) IRSD

V2.0 IRsd
EFH (U.S. EPA. 2017a) IRSD



95% CI of



95% CI of


95% CI of


95% CI of


95% CI of



the
Geometric


the


the


the


the


Geometric Geometric
Standard

Geometric Geometric

Geometric Geometric

Geometric Geometric

Geometric Geometric


Mean
Mean
Deviation

Mean
Mean

Mean
Mean

Mean
Mean

Mean
Mean


(ug/dL)
(ug/dL)
(ug/dL)
N
(ug/dL)
(ug/dL)
N
(ug/dL)
(ug/dL)
N
(ug/dL)
(ug/dL)
N
(ug/dL)
(ug/dL)
Box (Total) by Calendar
Year
















1995
99
4.9
4.4, 5.5
1.8
99
4.4
3.8, 5.1
99
4.3
3.8, 5.0
99
6.8
6.1, 7.6
99
6.7
6.0, 7.4
1996
102
4.9
4.3, 5.6
1.9
102
3.9
3.4, 4.4
102
3.8
3.4, 4.3
102
6.0
5.5, 6.6
102
5.9
5.4, 6.5
1997
69
4.6
4.0, 5.3
1.8
69
3.5
3.1, 4.1
69
3.4
3.0, 3.9
69
5.1
4.5, 5.7
69
4.9
4.4, 5.5
1998
100
4.3
3.8, 4.9
1.9
100
3.9
3.5, 4.3
100
3.8
3.5, 4.2
100
5.0
4.6, 5.4
100
4.9
4.5, 5.3
1999
115
4.2
3.7, 4.7
1.8
115
3.8
3.4, 4.2
115
3.7
3.3, 4.0
115
4.6
4.2, 5.0
115
4.4
4.1, 4.8
2000
90
4.0
3.5, 4.5
1.8
90
3.4
3.0, 3.8
90
3.3
3.0, 3.7
90
4.1
3.7, 4.5
90
4.0
3.7, 4.4
2001
81
3.2
2.8, 3.7
1.8
81
3.0
2.7, 3.3
81
2.9
2.6, 3.2
81
4.0
3.7, 4.4
81
3.9
3.6, 4.2
2002
78
2.6
2.4, 2.9
1.5
78
2.8
2.5, 3.1
78
2.7
2.4, 3.0
78
3.5
3.2, 3.8
78
3.4
3.1, 3.7
2003-2008
10
4.1
2.5, 6.9
2.0
10
3.6
2.1, 6.4
10
3.5
2.0, 6.2
10
3.8
2.2, 6.4
10
3.6
2.2, 6.1
2013
107
2.8
2.6, 3.1
1.5
107
2.7
2.5, 3.0
107
2.6
2.4, 2.8
107
2.7
2.6, 2.9
107
2.6
2.5, 2.8
2018
24
3.5
2.8, 4.4
1.7
24
2.2
1.9, 2.5
24
2.1
1.9, 2.4
24
2.2
2.0, 2.5
24
2.2
2.0, 2.4
Basin (Total) by
Calendar Year
















2002-2005
24
4.0
3.2, 5.0
1.7
24
5.1
3.7, 7.0
24
4.9
3.7, 6.7
24
6.1
4.8, 7.8
24
5.9
4.7, 7.4
2006
25
3.2
2.6, 3.8
1.6
25
4.5
3.4, 6.1
25
4.4
3.3, 5.8
25
5.4
4.4, 6.6
25
5.2
4.2, 6.3
2007
25
3.1
2.5, 3.9
1.7
25
3.3
2.6, 4.2
25
3.2
2.5, 4.1
25
4.4
3.7, 5.3
25
4.3
3.5, 5.1
2008
21
2.3
2.0, 2.7
1.3
21
2.7
2.3, 3.3
21
2.7
2.3, 3.2
21
3.4
2.9, 3.9
21
3.3
2.9, 3.8
2009
57
3.0
2.8, 3.3
1.4
57
2.7
2.5, 3.0
57
2.7
2.4, 3.0
57
3.6
3.3, 3.9
57
3.5
3.2, 3.8
2010
32
2.2
1.9, 2.5
1.5
32
2.9
2.6, 3.2
32
2.8
2.5, 3.1
32
3.4
3.0, 3.8
32
3.3
3.0, 3.7
2011
9
3.7
2.8, 4.8
1.4
9
3.4
2.5, 4.5
9
3.3
2.5, 4.3
9
4.1
3.3, 4.9
9
3.9
3.2, 4.8
2013
29
2.8
2.4, 3.2
1.5
29
2.8
2.3, 3.5
29
2.8
2.3, 3.4
29
3.1
2.6, 3.6
29
3.0
2.6, 3.5
2015
27
2.6
2.2, 3.2
1.6
27
3.4
2.6, 4.3
27
3.2
2.5, 4.2
27
3.5
2.8, 4.3
27
3.3
2.7, 4.1
2017
16
4.2
3.1, 5.7
1.7
16
2.3
2.0, 2.7
16
2.3
2.0, 2.6
16
2.6
2.3, 2.9
16
2.4
2.2, 2.7
2018
4
3.2
1.6, 6.6
1.6
4
4.4
1.7, 11.2
4
4.4
1.8, 10.8
4
4.2
2.1, 8.5
4
4.1
2.1, 8.3
Notes:
* 40% dust, 30% property soil, 30% arithmetic average community soil.
IRsd = soil/dust ingestion rate.
CI = confidence interval.
BDL = below detection limit (capillary detection limits are 1.4 and 1.9 |jg/dL, and the venous detection limit is 1.0 |jg/dL).
BLL = blood lead level.
|jg/dL = microgram per deciliter.
36

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 7. Summary of the difference between observed and predicted GM BLLs for IEUBK v2.0 model, censor level 2
(dataset omits predicted and observed BLLs >30.5 pg/dL and observed BDL BLLs)


Predicted PbB (IEUBK Model v2.0 build 1.6)


55/45 Dust/Soil Partition
40/30/30 Dust/Soil* Partition


EFH m.S. EPA. 2017a)

EFH m.S. EPA. 2017a)

v2.0 IRsd
IRsd
v2.0 IRsd
IRsd

Difference between GMs
Difference between GMs
Difference between GMs
Difference between GMs

(MQ/dL)
(MQ/dL)
(MQ/dL)
(MQ/dL)
Site-Wide (Total)
-0.26
-0.34
0.61
0.50
Box (Total)
-0.44
-0.52
0.53
0.42
Basin (Total)
0.23
0.15
0.84
0.73
Upper Basin (Total)
0.36
0.29
1.08
0.98
Lower Basin (Total)
-0.20
-0.32
0.07
-0.07
Site-Wide By Age (years)




0.5 to <1
0.99
0.37
2.41
1.56
1 to <2
0.39
0.19
1.94
1.66
2 to <3
-0.49
-0.68
0.36
0.10
3 to <4
-0.45
-0.58
0.42
0.25
4 to <5
-0.27
-0.44
0.52
0.28
5 to <6
-0.26
-0.16
0.52
0.65
6 to <7
-0.51
-0.37
0.12
0.32
Box (Total) by Age (years)




0.5 to <1
0.83
0.23
2.52
1.65
1 to <2
0.35
0.14
1.95
1.66
2 to <3
-0.71
-0.92
0.19
-0.08
3 to <4
-0.64
-0.77
0.27
0.10
4 to <5
-0.58
-0.75
0.34
0.10
5 to <6
-0.38
-0.29
0.51
0.65
6 to <7
-0.64
-0.49
0.05
0.26
Basin (Total) by Age (years)




0.5 to <1
1.32
0.67
2.19
1.39
1 to <2
0.51
0.36
1.85
1.63
2 to <3

-0.13
0.71
0.48
3 to <4
0.10
-0.01
0.85
0.69
4 to <5
0.63
0.47
1.02
0.78
5 to <6

0.23
0.54
0.64
37

-------
Advancing Pb Exposure and Biokinetic Modeling


Predicted PbB (IEUBK Model v2.0 build 1.6)


55/45 Dust/Soil Partition
40/30/30 Dust/Soil* Partition


EFH m.S. EPA. 2017a)

EFH m.S. EPA. 2017a)

v2.0 IRsd
IRsd
v2.0 IRsd
IRsd

Difference between GMs
Difference between GMs
Difference between GMs
Difference between GMs

(MQ/dL)
(MQ/dL)
(MQ/dL)
(MQ/dL)
6 to <7
-0.15
-0.03
0.27
0.44
Box (Total) by Calendar Year




1995
-0.50
-0.57
1.89
1.76
1996
-1.02
-1.08
1.08
1.00
1997
-1.07
-1.18
0.45
0.30
1998
-0.39
-0.47
0.67
0.56
1999
-0.41
-0.51
0.40
0.26
2000
-0.63
-0.66
0.08
0.03
2001
-0.20
-0.30
0.82
0.68
2002
0.15
0.09
0.87
0.79
2003-2008
-0.52
-0.62
-0.38
-0.51
2013
-0.08
-0.19
-0.10
-0.19
2018
-1.33
-1.36
-1.27
-1.31
Basin (Total) by Calendar Year




2002-2005
1.11
0.95
2.12
1.92
2006
1.36
1.20
2.19
1.98
2007
0.17
0.07
1.33
1.14
2008
0.41
0.38
1.02
0.99
2009
-0.30
-0.36
0.54
0.46
2010
0.67
0.59
1.19
1.08
2011
-0.33
-0.41
0.36
0.24
2013
0.07
0.03
0.30
0.25
2015
0.72
0.60
0.84
0.71
2017
-1.88
-1.97
-1.66
-1.78
2018
1.19
1.16
1.01
0.93
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
Differences are calculated as predicted GM minus observed GM.
BDL: capillary detection limits were 1.4 and 1.9 |jg/dL and the venous detection limit was 1.0 |jg/dL.
Grey shading and bold text indicate the smallest difference.
38

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 5. Summaries of mean observed and predicted BLLs with 95% CIs and lines
of unity for geographic areas, censor level 2 (omits predicted and observed
BLLs >30.5 pg/dL and observed BDL BLLs)
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Predicted Blood Lead GM (ng/dL)
A) IEUBK v2.0, v2.0 IRs, 55/45 partition
, Box
'Site-wide
Upper
Basin
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Predicted Blood Lead GM (ng/dL)
C) IEUBK v2.0, v2.0 IRs. 40/30/30 partition
5.0 -
4.5 -
4.0 -
3.5 -
3.0 -
2.5
2.0 -
1.5 -
1.0 -
0.5 -
0.0 -
' Upper
Basin
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Predicted Blood Lead GM (ng/dL)
B) IEUBK v2.0, EFH IRs, 55/45 partition
D) IEUBK v2.0, EFH IRs, 40/30/30 partition
Upper
Basin
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Predicted Blood Lead GM (ng/dL)
39

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 6. Summary of mean observed and predicted BLLs with 95% CIs for
geographic areas, censor level 2 (omits predicted and observed BLLs
>30.5 pg/dL and observed BDL BLLs)
Site-wide
Box
Basin
Observed » 55/45, v2.0 IR •- 55/45, EFH IR ~ 40/30/30, v2.0 IR 40/30/30, EFH IR
40

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 7. Summaries of mean observed and predicted BLLs with 95% CIs and lines
of unity for site-wide age groups, censor level 2 (omits predicted and
observed BLLs >30.5 pg/dL and observed BDL BLLs)
Predicted Blood Lead GM (ng/dL)
A) IEUBK v2.0, v2.0 IRs, 55/45 partition - Site-wide by
Age1
Predicted Blood Lead GM (|j.g/dL)
B) IEUBK v2.0, EFH IRs, 55/45 partition - Site-wide by
Age1
C) IEUBK v2.0, v2.0 IRs, 40/30/30 partition - Site-wide	D) IEUBK v2.0, EFH IRs, 40/30/30 partition - Site-wide
by Age1	by Age1
1 Numbers in the figure represent age ranges as follows: 0.5 - 0.5 to <1 years; 1 -1 to <2 years; 2 - 2 to <3 years;
3 - 3 to <4 years; 4 - 4 to <5 years; 5 - 5 to <6 years; 6 - 6 to <7 years.
41

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 8. Summary of mean observed and predicted BLLs with 95% CIs for site-wide
age groups, censor level 2 (omits predicted and observed BLLs >30.5 pg/dL
and observed BDL BLLs)
0.5 to <1 1 to <2	2 to <3	3 to <4	4 to <5	5 to <6	6 to <7
Age (Years)
¦ Observed » 55/45, v2.0 IR ~ 55/45, EFH IR ~ 40/30/30, v2.0 IR 40/30/30, EFH IR
42

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 9. Weighted linear regression model for observed and predicted age group
GM BLLs with 95% CIs, censor level 2 (omits predicted and observed BLLs
>30.5 pg/dL and observed BDL BLLs)
6-<7 yr
2.0 H	1	1	1	1	1	1	1	
2.0	3.0	4.0	5.0	6.0
Predicted GM BLL (pg/dL)
Points are GMs. Flags are 95% CIs. The open circle represents the GM for children
<7 years and was not included in the regression model because it is an aggregate of
all ages. The dotted lines show the 95% CIs on the weighted regression model.
Observed and predicted BLLs were weighted by the inverse of the squared variances
(1/SE2). Parameter values are as follows: intercepts .32, slope=0.69, r2=0.90.
43

-------
Advancing Pb Exposure and Biokinetic Modeling
Predicted GM BLLs are also presented by calendar year in Table 6. In the Box, no specific
partition or IRsd consistently better predicted observed GM BLLs than another by calendar year,
and between 1998 and 2013, the differences between predicted and observed GM BLLs for all
four IEUBK v2.0 model configurations were within ±1 |jg/dL (Table 7). In the Basin, the
55/45 partition using EFH IRsd more consistently predicted observed mean BLLs by calendar
year, typically with the smallest differences between 2002 and 2015, with the exception of 2009
and 2011 (Table 7). In more recent years when remediation was largely complete (Basin:
2015-2018, and Box: 2003-2018), the two partition scenarios predicted similar GM BLLs and the
differences between partitions were negligible (Tables 6 and 7). This outcome was likely due to
the fact that community soil averages were very near the property soil Pb concentrations
(Figures 10 and 11).
5.3 Prediction of Probability of Exceeding 5 |jg/dL
Table 8 and Figures 12 through 15 present the average observed and predicted Ps values (with
95% CIs) for each category for censor level 2. Table 9 presents the differences (predicted minus
observed) between the average Psfor each category (similar tables and figures for censor
levels 1 and 3 are presented in Appendices F and H, respectively).
5.3.1 Performance of Default IEUBK v2.0 Model
The mean site-wide predicted Ps using the default IEUBK v2.0 model was 26.6% (95% CI: 24.0,
29.1) and approximately 5.5% less than the observed Ps of 32.1% (95% CI: 29.4, 34.8).
Differences between the predicted and observed Ps values by geographic area (Box, Upper
Basin, and Lower Basin) ranged from -10.8 to +13.8% (Table 9). In general, the Box under-
predicted Ps averages while the Basin over-predicted P5. The 95% CIs on the observed and
predicted mean Ps did not overlap in most areas, the exception being the Lower Basin, which
comprised 5% of the total observations (n=60) and had the widest confidence limits (Figures 12
and 13).
The default IEUBK model generally performed well in predicting observed site-wide age Ps
means, with exceptions at the youngest and oldest ages (Figures 14 and 15). The model under-
predicted average Ps by approximately 5-10% for children 2 years and older (Table 9). Overall,
the model predicted an age trend in the Ps well (intercepts4.4, slope=0.71, r2= 0.91;
Figure 16).
5.3.2 Alternative IRsd and Soil and Dust Partitions
Predicted Ps averages for alternative values of IRsd and soil/dust partitions are presented in
Table 8 and Figures 12 through 15. As expected, because of the similarity in the values for IRsd,
performance of the model was essentially unchanged when the EFH values for IRsd were used
in the model (Figures 12 and 13). The average predicted site-wide Ps was 25.2% (95% CI: 22.7,
27.7; Table 8) using the 55/45 partition with the EFH IRsd. Using the IEUBK v2.0 IRsd, the
predicted site-wide Ps was higher when the 40/30/30 partition was assumed (39.0%; 95% CI:
36.2, 41.8) compared to the 55/45 partition (26.6%; 95% CI: 24.0, 29.1). The 55/45 partition
model with IEUBK v2.0 IRsd and the 40/30/30 partition with EFH IRsd corresponded best with
site-wide observed Ps values (Table 9). However, relative to the observed site-wide Ps, the
55/45 partition under-predicted, whereas the 40/30/30 partition over-predicted the Ps as
illustrated in Figures 12 and 13. Differences between site-wide predicted and observed Ps
averages for all four IEUBK v2.0 model configurations were within 7%.
44

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 8. Summary of observed and predicted average probability of exceeding 5 pg/dL for IEUBK v2.0 model, censor
level 2 (dataset omits predicted and observed BLLs >30.5 pg/dL and observed BDL BLLs)









Predicted Blood Lead





Observed Blood Lead1,


55/45 Dust/Soil Partition



40/30/30 Dust/Soil*
Partition






v2.0 IRsd
EFH (U.S. EPA. 2017a) IRSD

V2.0 IRsd
EFH (U.S. EPA. 2017a) IRSD


95% CI for
Percent of Percent

Average
Predicted
95% CI for

Average
Predicted
95% CI for

Average
Predicted
95% CI for

Average
Predicted
95% CI for


BLLs >5
BLLs >5

Probabilities Probabilities

Probabilities Probabilities

Probabilities Probabilities

Probabilities Probabilities


ug/dL (%) ug/dL (%)
N
>5 |jg/dL (%) >5 |jg/dL (%)
N
>5 |jg/dL (%) >5 |jg/dL (%)
N
>5 |jg/dL (%) >5 |jg/dL (%)
N
>5 |jg/dL (%) >5 |jg/dL(%)
Site-Wide (Total)
1144
32.1
29.4 - 34.8
1144
26.6
24.0-29.1
1144
25.2
22.7 - 27.7
1144
39.0
36.2 - 41.8
1144
37.4
34.6 - 40.2
Box (Total)
875
37.9
34.7 - 41.2
875
27.2
24.2-30.1
875
25.8
22.9 - 28.7
875
41.1
37.8 - 44.3
875
39.5
36.3 - 42.8
Basin (Total)
269
13.0
9.0-17.0
269
24.6
19.5-29.8
269
23.2
18.2-28.2
269
32.3
26.7 - 37.9
269
30.4
24.9 - 35.9
Upper Basin (Total)
209
12.0
7.6-16.4
209
25.8
19.8-31.7
209
24.5
18.7-30.3
209
34.8
28.3 - 41.2
209
33.1
26.7 - 39.5
Lower Basin (Total)
60
16.7
7.2-26.1
60
20.6
10.4-30.8
60
18.6
8.8 - 28.5
60
23.5
12.8-34.2
60
21.0
10.7-31.3
Site-Wide By Age (years)
0.5 to <1
70
30.0
19.3-40.7
70
44.6
32.9 - 56.2
70
37.4
26.1 -48.8
70
61.5
50.1 -72.9
70
52.8
41.1 -64.4
1 to <2
138
45.7
37.3 - 54.0
138
47.0
38.7 - 55.3
138
44.6
36.3 - 52.9
138
64.8
56.8 - 72.7
138
62.2
54.1 -70.3
2 to <3
202
36.6
30.0 - 43.3
202
28.0
21.8-34.2
202
25.2
19.2-31.2
202
40.0
33.2 - 46.7
202
36.2
29.6 - 42.8
3 to <4
176
33.5
26.5 - 40.5
176
25.0
18.6-31.4
176
23.5
17.2-29.7
176
37.4
30.3 - 44.6
176
35.1
28.1 -42.2
4 to <5
187
30.5
23.9-37.1
187
23.0
17.0-29.1
187
20.7
14.9-26.5
187
35.2
28.3 - 42.0
187
31.7
25.0 - 38.4
5 to <6
185
26.5
20.1 -32.8
185
21.1
15.3-27.0
185
22.4
16.4-28.5
185
32.4
25.6-39.1
185
34.3
27.4-41.1
6 to <7
186
23.7
17.5-29.8
186
13.4
8.5-18.3
186
15.2
10.1 -20.4
186
22.3
16.4-28.3
186
25.4
19.1 -31.6
Box (Total) by Age (years)
0.5 to <1
47
34.0
20.5 - 47.6
47
43.3
29.1 -57.5
47
36.1
22.4 - 49.9
47
63.0
49.2 - 76.8
47
54.3
40.0 - 68.5
1 to <2
112
50.0
40.7 - 59.3
112
49.5
40.2 - 58.8
112
47.1
37.8 - 56.3
112
67.1
58.4 - 75.8
112
64.7
55.8 - 73.5
2 to <3
149
43.0
35.0 - 50.9
149
30.1
22.7 - 37.5
149
27.1
20.0 - 34.3
149
43.4
35.4-51.3
149
39.5
31.7 - 47.4
3 to <4
137
37.2
29.1 -45.3
137
25.7
18.4-33.0
137
24.1
16.9-31.2
137
38.8
30.7 - 47.0
137
36.5
28.4 - 44.5
4 to <5
144
38.9
30.9 - 46.9
144
22.2
15.4-28.9
144
19.8
13.3-26.4
144
36.7
28.8 - 44.6
144
33.1
25.4 - 40.8
5 to <6
144
32.6
25.0 - 40.3
144
20.7
14.1 -27.4
144
22.1
15.3-28.9
144
34.1
26.4-41.9
144
36.1
28.3 - 44.0
6 to <7
142
29.6
22.1 -37.1
142
14.1
8.4-19.8
142
15.9
9.9-21.9
142
24.5
17.4-31.6
142
27.7
20.3 - 35.0
Basin (Total) by Age
(years)
0.5 to <1
23
21.7
4.9 - 38.6
23
47.2
26.8 - 67.6
23
40.0
20.0-60.1
23
58.3
38.2 - 78.5
23
49.7
29.2-70.1
1 to <2
26
26.9
9.9 - 44.0
26
36.2
17.8-54.7
26
34.0
15.7-52.2
26
54.6
35.4 - 73.7
26
51.6
32.4 - 70.8
2 to <3
53
18.9
8.3 - 29.4
53
22.1
11.0-33.3
53
19.7
9.0 - 30.4
53
30.4
18.0-42.8
53
26.9
14.9-38.8
3 to <4
39
20.5
7.8 - 33.2
39
22.7
9.6 - 35.9
39
21.3
8.4-34.1
39
32.6
17.9-47.3
39
30.3
15.9-44.7
4 to <5
43
2.3
0.4-12.1
43
26.0
12.9-39.1
43
23.7
11.0-36.4
43
30.0
16.3-43.7
43
27.0
13.7-40.3
5 to <6
41
4.9
1.3-16.1
41
22.5
9.7 - 35.3
41
23.7
10.7-36.7
41
26.2
12.7-39.6
41
27.7
14.0-41.4
6 to <7
44
4.5
1.3-15.1
44
11.2
1.9-20.5
44
13.0
3.1 -23.0
44
15.4
4.7 - 26.0
44
18.0
6.6 - 29.3
45

-------
Advancing Pb Exposure and Biokinetic Modeling










Predicted Blood Lead






Observed Blood Leadt


55/45 Dust/Soil Partition



40/30/30 Dust/Soil*
Partition







V2.0 IRsd
EFH (U.S. EPA. 2017a) IRSD

V2.0 IRsd
EFH (U.S. EPA. 2017a) IRSD




95% CI for

Average


Average


Average


Average




Percent of Percent

Predicted
95% CI for

Predicted
95% CI for

Predicted
95% CI for

Predicted
95% CI for



BLLs >5
BLLs >5

Probabilities Probabilities

Probabilities Probabilities

Probabilities Probabilities

Probabilities Probabilities



ug/dL (%) ug/dL (%)
N
>5 |jg/dL (%) >5 |jg/dL (%)
N
>5 |jg/dL (%) >5 |jg/dL (%)
N
>5 |jg/dL (%) >5 |jg/dL (%)
N
>5 |jg/dL (%) >5 |jg/dL(%)
Box (Total) by Calendar















Year

















1995
99
56.6
46.8 - 66.3
99
38.9
29.3 - 48.5
99
38.1
28.6 - 47.7
99
66.5
57.2 - 75.8
99
65.9
56.6 - 75.2

1996
102
50.0
40.3 - 59.7
102
33.1
24.0 - 42.3
102
32.4
23.3 - 41.5
102
60.1
50.5 - 69.6
102
59.7
50.1 -69.2

1997
69
55.1
43.3 - 66.8
69
29.0
18.3-39.7
69
27.5
16.9-38.0
69
50.3
38.5-62.1
69
48.1
36.3 - 59.9

1998
100
52.0
42.2 - 61.8
100
33.7
24.5 - 43.0
100
32.3
23.1 -41.4
100
48.0
38.2 - 57.8
100
46.5
36.7 - 56.3

1999
115
43.5
34.4 - 52.5
115
30.9
22.5 - 39.4
115
29.0
20.7 - 37.3
115
42.0
33.0-51.1
115
39.8
30.9 - 48.8

2000
90
46.7
36.4 - 57.0
90
25.6
16.6-34.6
90
25.0
16.1 -34.0
90
34.5
24.7 - 44.3
90
33.5
23.7 - 43.2

2001
81
23.5
14.2-32.7
81
20.9
12.1 -29.8
81
19.2
10.6-27.7
81
35.3
24.9 - 45.7
81
32.4
22.2 - 42.6

2002
78
10.3
3.5-17.0
78
16.9
8.6 - 25.2
78
15.7
7.6 - 23.7
78
26.3
16.5-36.0
78
24.6
15.0-34.2

2003-2008
10
30.0
1.6-58.4
10
31.6
2.8 - 60.5
10
30.6
2.0-59.1
10
35.2
5.6 - 64.9
10
33.7
4.4 - 63.0

2013
107
6.5
1.9-11.2
107
16.7
9.7 - 23.8
107
14.8
8.0-21.5
107
15.0
8.3-21.8
107
13.0
6.6-19.4

2018
24
25.0
7.7 - 42.3
24
7.4
1.9-24.6
24
6.6
1.6-23.6
24
6.6
1.6-23.6
24
5.8
1.3-22.5
Basin (Total) by Calendar















Year

















2002-2005
24
33.3
14.5-52.2
24
47.9
27.9 - 67.9
24
46.5
26.6 - 66.5
24
58.8
39.2 - 78.5
24
57.4
37.6 - 77.2

2006
25
20.0
4.3 - 35.7
25
45.3
25.8 - 64.8
25
44.5
25.0 - 63.9
25
53.2
33.7 - 72.8
25
51.6
32.0-71.1

2007
25
24.0
7.3 - 40.7
25
29.0
11.3-46.8
25
27.4
9.9 - 44.9
25
43.4
24.0 - 62.9
25
40.5
21.2-59.7

2008
21
0.0
o
o
o
o
21
16.9
0.9 - 33.0
21
15.9
0.2-31.5
21
24.0
5.7 - 42.3
21
22.7
4.8 - 40.6

2009
57
10.5
2.6-18.5
57
16.3
6.7 - 25.9
57
14.9
5.6-24.1
57
28.0
16.4-39.7
57
26.1
14.7-37.5

2010
32
3.1
0.6-15.7
32
16.5
3.6 - 29.3
32
15.1
2.7 - 27.4
32
24.4
9.5 - 39.3
32
22.5
8.1 -37.0

2011
9
11.1
2.0 - 43.5
9
25.9
8.1 -58.1
9
24.0
7.1 -56.4
9
34.7
3.6 - 65.8
9
32.8
2.1 -63.5

2013
29
6.9
1.9-22.0
29
18.1
4.1 -32.1
29
17.0
3.3 - 30.7
29
20.6
5.9 - 35.3
29
19.0
4.8 - 33.3

2015
27
3.7
0.7-18.3
27
27.0
10.3-43.8
27
24.9
8.6-41.2
27
27.4
10.6-44.2
27
25.2
8.8-41.6

2017
16
25.0
3.8 - 46.2
16
7.9
1.6-30.4
16
6.2
1.1 -28.2
16
9.9
2.4 - 32.9
16
7.8
1.6-30.4

2018
4
25.0
4.6 - 69.9
4
46.5
13.2-83.2
4
46.2
13.1 -83.0
4
40.5
10.5-79.8
4
39.8
10.2-79.4
Notes:
* 40% dust, 30% property soil, 30% arithmetic average community soil
IRsd = soil/dust ingestion rate
CI = confidence interval
BLL = blood lead level
|jg/dL = microgram per deciliter
46

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 9. Summary of the difference between observed and predicted average probability of exceeding 5 pg/dL for
IEUBK v2.0 model, censor level 2 (dataset omits predicted and observed BLLs >30.5 pg/dL and observed BDL
BLLs)


Predicted PbB (IEUBK Model v2.0 build 1.6)


55/45 Dust/Soil Partition
40/30/30 Dust/Soil* Partition


EFH (U.S. EPA. 2017a)

EFH m.S. EPA. 2017a)

v2.0 IRsd
IRsd
v2.0 IRsd
IRsd

Difference Between
Difference Between
Difference Between
Difference Between

Average Percent BLLs
Average Percent BLLs
Average Percent BLLs
Average Percent BLLs

>5 ng/dL (%)
>5 ng/dL (%)
>5 ng/dL (%)
>5 ng/dL (%)
Site-Wide (Total)
-5.5
-6.9
6.9
5.3
Box (Total)
-10.8
-12.1
3.1
1.6
Basin (Total)
11.6
10.2
19.3
17.4
Upper Basin (Total)
13.8
12.5
22.8
21.1
Lower Basin (Total)
3.9
2.0
6.8
4.3
Site-Wide By Age (years)




0.5 to <1
14.6
7.4
31.5
22.8
1 to <2
1.3
-1.1
19.1
16.5
2 to <3
-8.6
-11.4
3.3
-0.4
3 to <4
-8.5
-10.1
3.9
1.6
4 to <5
-7.4
-9.8
4.7
1.2
5 to <6
-5.3
-4.0
5.9
7.8
6 to <7
-10.3
-8.4
-1.3
1.7
Box (Total) by Age (years)




0.5 to <1
9.3
2.1
29.0
20.2
1 to <2
-0.5
-2.9
17.1
14.7
2 to <3
-12.9
-15.8
0.4
-3.4
3 to <4
-11.5
-13.1
1.6
-0.8
4 to <5
-16.7
-19.0
-2.2
-5.8
5 to <6
-11.9
-10.5
1.5
3.5
6 to <7
-15.5
-13.7
5.1
-1.9
47

-------
Advancing Pb Exposure and Biokinetic Modeling


Predicted PbB (IEUBK Model v2.0 build 1.6)


55/45 Dust/Soil Partition
40/30/30 Dust/Soil* Partition


EFH m.S. EPA. 2017a)

EFH m.S. EPA. 2017a)

v2.0 IRsd
IRsd
v2.0 IRsd
IRsd

Difference Between
Average Percent BLLs
>5 ng/dL (%)
Difference Between
Average Percent BLLs
>5 ng/dL (%)
Difference Between
Average Percent BLLs
>5 ng/dL (%)
Difference Between
Average Percent BLLs
>5 ng/dL (%)
Basin (Total) by Age (years)




0.5 to <1
25.4
18.3
36.6
27.9
1 to <2
9.3
7.0
27.7
24.6
2 to <3
3.3
0.8
11.5
8.0
3 to <4
2.2
0.7
12.1
9.8
4 to <5
23.7
21.4
27.7
24.7
5 to <6
17.6
18.8
21.3
22.8
6 to <7
6.6
8.5
10.8
13.4
Box (Total) by Calendar Year




1995
-17.7
-18.4
9.9
9.3
1996
-16.9
-17.6
10.1
9.7
1997
-26.0
-27.6
-4.8
-7.0
1998
-18.3
-19.7
-4.0
-5.5
1999
-12.5
-14.4
-1.4
-3.6
2000
-21.0
-21.6
-12.2
-13.2
2001
-2.6
-4.3
11.8
9.0
2002
6.6
5.4
16.0
14.4
2003-2008
1.6
0.6
5.2
3.7
2013
10.2
8.2
8.5
6.5
2018
-17.6
-18.4
-18.4
-19.2
48

-------
Advancing Pb Exposure and Biokinetic Modeling

Predicted PbB (IEUBK Model v2.0 build 1.6)

55/45 Dust/Soil Partition
40/30/30 Dust/Soil* Partition


EFH m.S. EPA. 2017a)

EFH m.S. EPA. 2017a)

v2.0 IRsd
IRsd
v2.0 IRsd
IRsd

Difference Between
Average Percent BLLs
>5 ng/dL (%)
Difference Between
Average Percent BLLs
>5 ng/dL (%)
Difference Between
Average Percent BLLs
>5 ng/dL (%)
Difference Between
Average Percent BLLs
>5 ng/dL (%)
Basin (Total) by Calendar Year




2002-2005
14.6
13.2
25.5
24.1
2006
25.3
24.5
33.2
31.6
2007
5.0
3.4
19.4
16.5
2008
16.9
15.9
24.0
22.7
2009
5.8
4.3
17.5
15.6
2010
13.3
11.9
21.3
19.4
2011
14.8
12.9
23.6
21.7
2013
11.2
10.1
13.7
12.1
2015
23.3
21.2
23.7
21.5
2017
-17.1
-18.8
-15.1
-17.2
2018
21.5
21.2
15.5
14.8
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
Differences calculated as predicted Ps minus observed Ps.
Grey shading and bold text indicate the smallest difference.
49

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 10. Average yearly media concentrations (mg/kg) in the Box used as inputs for
the IEUBK model, censor level 2 (omits predicted and observed BLLs
>30.5 pg/dL and observed BDL BLLs), 1995-2018
1600
3 1400
iC
D>
E
1200
1000
800
600
400
200
—Property Soil	Community Soil —*—40/30/30 Soil — Property Dust
a The soil input for the 40/30/30 parition (40% dust, 30% property soil, 30% community soil) is an
average of property soil and community soil.
50

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 11. Average yearly media concentrations (mg/kg) in the Basin used as inputs
for the IEUBK model, censor level 2 (omits predicted and observed BLLs
>30.5 pg/dL and observed BDL BLLs), 2002-2018
b	c
—•—Property Soil —•—Community Soil —•—40/30/30 Soil —Property Dust
3 Community soil concentration shown is from 2005. Community soil concentration in 2002 through
2005 ranged from 1,092 to 1,178 mg/kg.
b Spatially weighted average
c The soil input for the 40/30/30 parition (40% dust, 30% property soil, 30% community soil) is an
average of property soil and community soil.
51

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 12. Summaries of observed and predicted average probability of exceeding
5 pg/dL with 95% CIs and lines of unity for the geographic areas, censor
level 2 (all predicted and observed BLLs <30.5 pg/dL and no BDLs)
Predicted P5 (%)
A) IEUBK v2.0, v2.0 IRs, 55/45 partition
Predicted P5 (%)
B) IEUBK v2.0, EFH IRs, 55/45 partition
10 15 20 25 30 35 40 45
Predicted P5 (%)
Lower
Basin
5 10 15 20 25 30 35 40 45
Predicted P5 (%)
Basin
	T1 T
Basirr-
Upper
Basin
C) IEUBK v2.0, v2.0 IRs, 40/30/30 partition	D) IEUBK v2.0, EFH IRs, 40/30/30 partition
52

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 13. Summary of observed and predicted average probability of exceeding
5 pg/dL with 95% CIs for the geographic areas, censor level 2 (all predicted
and observed BLLs <30.5 pg/dL and no BDLs)
Site-wide
Box
Basin
Observed * 55/45, v2.0 IR » 55/45, EFH IR ~ 40/30/30, v2.0 IR 40/30/30, EFH IR
53

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 14. Summaries of observed and predicted average probability of exceeding
5 pg/dL with 95% CIs and lines of unity for site-wide age groups, censor
level 2 (all predicted and observed BLLs <30.5 pg/dL and no BDLs)
A) IEUBK v2.0, v2.0 IRs, 55/45 partition - Site-wide
by Age1
B) IEUBK v2.0, EFH IRs, 55/45 partition - Site-wide by
Age1
1 Numbers in the figure represent age ranges as follows: 0.5 - 0.5 to <1 years; 1 -1 to <2 years; 2 - 2 to <3 years;
3 - 3 to <4 years; 4 - 4 to <5 years; 5 - 5 to <6 years; 6 - 6 to <7 years.
C) IEUBK v2.0, v2.0 IRs, 40/30/30 partition - Site-wide
by Age1
D) IEUBK v2.0, EFH IRs, 40/30/30 partition - Site-wide
by Age1
30 40 50
Predicted P5 (%)
10 20 30 40 50 60 70 80
Predicted P5 (%)
— 50
30 40 50
Predicted P5 (%)
70 -
60 -
S"
— 50 -
a.
"S 40 -
>
01
» 30 -
o
20 -
30 40 50
Predicted P5 (%)
60 -
^50 -
Ql
U 40 -
>
fll
» 30 -
O
20 -
60
£
— 50
a.
U 40
>
U
» 30
o
20
54

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 15. Summary of observed and predicted average probability of exceeding
5 pg/dL with 95% CIs for site-wide age groups, censor level 2 (all predicted
and observed BLLs <30.5 pg/dL and no BDLs)
0.5 to <1	1 to <2	2 to <3	3 to <4	4 to <5	5 to <6	6 to <7
Age (Years)
~ Observed • 55/45, v2.0 IR ~ 55/45, EFH IR ~ 40/30/30, v2.0 IR 40/30/30, EFH IR
55

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 16. Weighted linear regression model for observed and predicted age group P5
with 95% CIs, censor level 2 (omits predicted and observed BLLs
>30.5 pg/dL and observed BDL BLLs)
60
50
SP
ov
1/1
a.
T3

0J
in
_Q
o
l-<2 yr \
_ 40
30
20
6-<7 Y.r-f
<7 yr
5-<6 yr
10
10
20	30 40
Predicted P5(%)
50
60
Points are mean P5. Flags are 95% CIs. The open circle represents the mean P5
for children <7 years and was not included in the regression model because it is an
aggregate of all ages. The dotted lines show the 95% CIs on the weighted
regression model. Observed and predicted P5 were weighted by the inverse of the
squared variances (1/SE2). Parameter values are as follows: intercepts4.4,
slope=0.71, r2=0.91.
56

-------
Advancing Pb Exposure and Biokinetic Modeling
In the Box, the average P5 predicted by the 40/30/30 partition model with EFH IRsd (39.5%; 95%
CI: 36.3, 42.8) best corresponds to the average observed Ps(37.9%; 95% CI: 34.7, 41.2) with
the smallest difference of 1.6%. The 55/45 partition under-predicts the observed P5, regardless
of IRsd by 10.8% using IEUBK v2.0 IRsd and 12.1% using EFH IRsd. This outcome is not
surprising given that the 40/30/30 partition model was originally developed using Box blood, soil,
and dust data.
In the Basin, the predicted P5average using the 55/45 partition model with EFH IRsd (23.2%;
95% CI: 18.2, 28.2) was closest to the average observed Ps(13.0%; 95% CI: 9.0, 17.0), with the
smallest difference of 10.2% (Table 9). All four IEUBK v2.0 model configurations (see Table 3)
over-predicted P5 compared to observed Basin values by 10.2-19.3% (Table 9 and Figure 13).
For the small number of children in the Lower Basin, the predicted P5 using the 55/45 partition
compared well to observed values by within 2.0-3.9% (Table 9 and Figure 12).
Site-wide P5 results by age showed similarities to the GM evaluation outcome. For the youngest
children <2 years old, Figures 14 and 15 show that using the 55/45 partition provided better
predictions of the P5 than using the 40/30/30 partition. For older children (2-<5 and 6-<7 years
old), the 40/30/30 partition best corresponded with observed P5 values (with differences ranging
from -1.3 to +4.7%) compared to using the 55/45 partition. In general, across all age ranges, the
40/30/30 partition model tended to over-predict P5 compared to observed, whereas the
55/45 partition model tended to under-predict P5 (except for the youngest children under 2 years
old).
5.4 Prediction of Distribution of Individual Child BLLs
Two evaluations were completed for this third performance metric.
5.4.1 Prediction Intervals for IEUBK Predicted GM BLLs
The intended application of the IEUBK model is to predict the probabilities of the BLL
distribution for populations of similarly exposed children and is not intended to predict individual
BLLs. However, it was expected that observed BLLs for a population of children living on each
property should distribute within the prediction limits (or the PI) of the model. The Pis are
dependent on the assumed GSD,. Theoretically, if a representative sample from a population of
children living at an individual property was available and their environmental exposures were
correctly assigned, it is reasonable to assume that 95% of the observed BLLs for children at that
property would fall within the 95% Pis of the model (Figure 17). However, several factors could
interfere with achieving this outcome, including: (1) the misclassification of children's exposures
to Pb; (2) a small sample size of BLLs that were unlikely to represent the child population at the
property; and (3) a GSD, that does not represent the actual variance in BLLs of children living at
the property. Based on a recent evaluation of 2017 and 2018 BHSS BLLs, it is likely that for
some children in the BHSS dataset used in this model evaluation, the misclassification of
exposures exists (Alta, 2019a). Additionally, the BHSS population GSD will exceed the GSD,
since it includes exposure variability between individual properties; therefore, it was expected
that more than 5% of the observed BLLs would fall outside the 95% Pis.
Table 10 presents the percent of observed BLLs for censor level 2 that fall outside the 95% Pis,
as well as the percent that were below the lower PI and above the upper PI. Figures 18 and 19
present the percent that fall outside the 95% Pis by geographic area (i.e., site-wide, Box, Basin)
and site-wide by age, respectively. Scatter plots with lower and upper prediction limits for censor
level 2 are presented in Appendix G for geographic regions (Figures G-6 to G-8) and site-wide
age groups (Figures G-9 to G-16).
57

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 17. Plot of predicted and observed BLLs from an ideal lognormal distribution
of BLLs with 95% Pis
100
Predicted BLL (|jg/dL)
Lines are the lower and upper 95% prediction limits defined as the 2,5th and
97.5th percentiles of the distribution of predicted BLLs. Predicted BLLs (BLLp)
are 1000 random draws from a distribution BLLp = lognormal (GMp, GSD,) with
GMp set to the GM of the predicted BLLs for ages <6 years (3.3 pg/dL) and
GSD, set to the IEUBK model default value, 1.6. Simulated observed BLLs
(BLLs) are random draws from a distribution BLLs = lognormal (BLLp, GSD,).
58

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 10. Summary of observed/predicted PbB pairs that fall outside Pis for IEUBK v2.0 model, censor level 2 (dataset
omits predicted and observed BLLs >30.5 pg/dL and observed BDL BLLs)




55/45 Dust/Soil Partition



40/30/30 Dust/Soil* Partition






IEUBK v2.0




IEUBK v2.0














EFH EFH (U.S. EPA. 2017a)



v2.0 IRsd

EFH (U.S. EPA. 2017a) IRsn

/2.0 IRsd


IRsd




%

%>
%


%

%>
%
%<
%>



Outside
%<
Upper
Outside
%<
%>
Outside
%<
Upper
Outside
Lower
Upper


N
PI
Lower PI
PI
PI
Lower PI
Upper PI
PI
Lower PI
PI
PI
PI
PI
Site-Wide (Total)
1144
14%
6%
8%
15%
6%
9%
13%
9%
4%
12%
8%
4%
Box (Total)
875
15%
6%
9%
15%
5%
10%
13%
9%
4%
13%
8%
4%
Basin (Total)
269
11%
7%
4%
12%
7%
5%
12%
9%
3%
11%
8%
3%
Upper Basin (Total)
209
10%
7%
3%
11%
7%
4%
11%
10%
1%
10%
9%
1%
Lower Basin (Total)
60
15%
7%
8%
17%
7%
10%
13%
5%
8%
13%
5%
8%
Site-Wide by Age (years)














0.5 to <1
70
17%
10%
7%
17%
9%
9%
24%
20%
4%
16%
11%
4%
1 to <2
138
12%
9%
4%
12%
7%
4%
13%
12%
1%
12%
11%
1%
2 to <3
202
16%
4%
11%
16%
3%
12%
13%
6%
7%
13%
5%
8%
3 to <4
176
11%
3%
7%
11%
3%
9%
9%
7%
2%
9%
7%
2%
4 to <5
187
19%
9%
10%
20%
7%
12%
14%
10%
4%
15%
9%
6%
5 to <6
185
12%
7%
5%
13%
8%
5%
10%
9%
1%
10%
9%
1%
6 to <7
186
15%
4%
10%
14%
5%
9%
12%
6%
6%
13%
8%
5%
Box (Total) by Age (years)














0.5 to <1
47
15%
9%
6%
13%
6%
6%
23%
21%
2%
13%
11%
2%
1 to <2
112
13%
9%
4%
11%
7%
4%
11%
9%
2%
11%
9%
2%
2 to <3
149
17%
5%
12%
17%
3%
13%
13%
7%
7%
15%
7%
8%
3 to <4
137
13%
4%
9%
14%
3%
11%
9%
7%
2%
9%
7%
2%
4 to <5
144
19%
8%
12%
21%
6%
15%
14%
9%
5%
15%
8%
7%
5 to <6
144
13%
7%
6%
14%
8%
6%
10%
9%
1%
10%
9%
1%
6 to <7
142
17%
5%
12%
16%
6%
10%
15%
8%
7%
16%
10%
6%
Basin (Total) by Age
(years)














0.5 to <1
23
22%
13%
9%
26%
13%
13%
26%
17%
9%
22%
13%
9%
1 to <2
26
12%
8%
4%
15%
8%
8%
23%
23%
0%
19%
19%
0%
2 to <3
53
13%
4%
9%
13%
4%
9%
11%
4%
8%
9%
2%
8%
3 to <4
39
3%
3%
0%
3%
3%
0%
5%
5%
0%
5%
5%
0%
4 to <5
43
16%
14%
2%
16%
14%
2%
14%
12%
2%
14%
12%
2%
5 to <6
41
10%
7%
2%
10%
7%
2%
10%
10%
0%
10%
10%
0%
6 to <7
44
7%
2%
5%
7%
2%
5%
5%
2%
2%
5%
2%
2%
59

-------
Advancing Pb Exposure and Biokinetic Modeling




55/45 Dust/Soil Partition



40/30/30 Dust/Soil* Partition






IEUBK v2.0




IEUBK v2.0














EFH EFH (U.S. EPA. 2017a)



v2.0 IRsd

EFH (U.S. EPA. 2017a) IRsn

/2.0 IRsd


IRsd




%

%>
%


%

%>
%
%<
%>



Outside
%<
Upper
Outside
%<
%>
Outside
%<
Upper
Outside
Lower
Upper


N
PI
Lower PI
PI
PI
Lower PI
Upper PI
PI
Lower PI
PI
PI
PI
PI
Box (Total) by Calendar
Year














1995
99
19%
9%
10%
18%
9%
9%
17%
16%
1%
18%
17%
1%

1996
102
22%
7%
15%
21%
6%
15%
15%
12%
3%
15%
11%
4%

1997
69
22%
7%
14%
25%
6%
19%
17%
12%
6%
14%
10%
4%

1998
100
13%
5%
8%
13%
5%
8%
14%
8%
6%
15%
9%
6%

1999
115
13%
5%
8%
14%
4%
10%
11%
5%
6%
11%
5%
6%

2000
90
23%
9%
14%
22%
9%
13%
16%
10%
6%
16%
10%
6%

2001
81
15%
6%
9%
16%
5%
11%
12%
10%
2%
14%
9%
5%

2002
78
10%
8%
3%
9%
6%
3%
12%
10%
1%
10%
9%
1%

2003-2008
10
10%
0%
10%
10%
0%
10%
0%
0%
0%
0%
0%
0%

2013
107
4%
3%
1%
4%
2%
2%
3%
2%
1%
2%
1%
1%

2018
24
21%
0%
21%
21%
0%
21%
21%
0%
21%
25%
0%
25%
Basin (Total) by Calendar
Year














2002-2005
24
21%
17%
4%
22%
17%
4%
21%
17%
4%
21%
17%
4%

2006
25
20%
20%
0%
20%
20%
0%
24%
24%
0%
24%
24%
0%

2007
25
16%
12%
4%
20%
12%
8%
16%
16%
0%
12%
12%
0%

2008
21
5%
5%
0%
5%
5%
0%
10%
10%
0%
5%
5%
0%

2009
57
2%
0%
2%
4%
0%
4%
4%
4%
0%
2%
2%
0%

2010
32
0%
0%
0%
0%
0%
0%
6%
6%
0%
6%
6%
0%

2011
9
11%
0%
11%
11%
0%
11%
0%
0%
0%
0%
0%
0%

2013
29
14%
7%
7%
14%
7%
7%
14%
7%
7%
14%
7%
7%

2015
27
15%
11%
4%
15%
11%
4%
7%
7%
0%
7%
7%
0%

2017
16
31%
0%
31%
31%
0%
31%
31%
0%
31%
31%
0%
31%

2018
4
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Notes:
* 40% dust, 30% property soil, 30% arithmetic average community soil.
Grey shaded cells with bold text represent the lowest percentage of all 8 model runs, for each data category. Corresponding percentages above and below PLs are bold.
The number of records (N) for each model run corresponds to the same Ns shown in Tables 6 and 8.
Percentages were rounded to the nearest whole percent. For this reason, the percent of BLLs < Lower PI and > Upper PI may not add up to the total percent of BLLs outside Pis.
60

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 18. Percent of observed BLLs that fail outside the 95% Pi of the predicted BLL
GMs for the geographic areas, censor level 2 (all predicted and observed
BLLs <30.5 pg/dL and no BDLs)
Site-wide
Box
Basin
55/45, v2.0 IR ¦ 55/45, EFH IR ¦ 40/30/30, v2.0 IR 40/30/30, EFH IR
61

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 19. Percent of observed BLLs that fall outside the 95% PI of the predicted BLL
GMs for site-wide age groups, censor level 2 (all predicted and observed
BLLs <30.5 pg/dL and no BDLs)
25%
0%
6 to <7
0.5 to <1 1 to <2	2 to <3	3 to <4	4 to <5	5 to <6
Age (Years)
¦ 55/45, v2.0 IR ¦ 55/45, EFH IR ¦ 40/30/30, v2.0 IR 40/30/30, EFH IR
62

-------
Advancing Pb Exposure and Biokinetic Modeling
5.4.1.2	Performance of default IEUBK v2.0 model
Of the 1144 site-wide observed and predicted BLLs, 14% fell outside the model Pis; 6% were
below the 95% lower PI and 8% were above the upper PI (Table 10). When stratified by
geographical area, the percentage outside of the Pis ranged from 10% for Upper Basin
(209 observations) to 15% for both the Box (875 observations) and Lower Basin
(60 observations).
5.4.1.3	Alternative IRsd and soil and dust partitions
Site-wide, the alternative models resulted in 12-15% of observed BLLs outside the model Pis
(Figure 18). The IRsd does not greatly change the outcome as differences in the percentage of
observed BLLs that fall outside the model Pis (same partition, different IRsd) were typically <1%
with the exception of the youngest age group with the 40/30/30 partition (Figure 19). The
40/30/30 partition with EFH IRsd had the lowest site-wide percent (12%) of observed BLLs
outside the model Pis. The 40/30/30 partition also had the lowest percentage of BLLs outside
the Pis in both the Box (13%) and Basin (11%), although the 55/45 IEUBK v2.0 IRsd in the
Basin performed equally well with only 11% outside the PI (Figure 18). Data for censor levels 1
and 3 are presented in Appendices F and H, respectively.
The Pis for each predicted GM BLL depend completely on the value selected for the GSD,
assumed in the model, 1.6. If the GSD, was assigned a higher value, the Pis would widen and
encompass a larger proportion of BLL observations. It was expected that the percent of
observed BLLs in our sample that fall outside of the model Pis would be greater than the 5%.
Additionally, not all exposures can be classified correctly without review of additional site-
specific data (such as paired questionnaire data, which was not completed as part of this
evaluation). Nevertheless, the site-wide and geographic area model runs resulted in 10-17% of
observed BLLs falling outside the 95% Pis.
5.4.2 Cumulative Distributions of Individual Child BLLs
Table 11 summarizes site-wide results of the K-S 2-sample comparison test used to evaluate
the differences between the empirical distributions of the observed and predicted GM BLLs. The
probabilities presented in Table 11 are for two-sided tests of the null hypothesis. The null
hypothesis was that the observed and predicted BLLs were sampled from the same distribution.
Figures 20 through 22 present the cumulative distributions of observed and predicted GM BLLs.
Table 11. Summary of the K-S 2-sample comparison for IEUBK v2.0 model, censor
level 2 (dataset omits predicted and observed BLLs >30.5 |jg/dL and
observed BDL BLLs)

55/45 Dust/Soil Partition

40/30/30 Dust/Soil*
Partition



EFH EFH (U.S. EPA.


EFH EFH (U.S. EPA.

v2.0 IRsd
2017a) IRsd
v2.0 IRsd
2017a) IRsd

D+
P
value*
D+
p value*
D+
P
value*
D+
p value*
Site-Wide (Total)
0.15
<0.01
0.18
<0.01
0.17
<0.01
0.17
<0.01
Box (Total)
0.21
<0.01
0.23
<0.01
0.15
<0.01
0.15
<0.01
Basin (Total)
0.07
0.45
0.07
0.58
0.26
<0.01
0.22
<0.01
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
fThe D statistic indicates the maximum vertical distance between observed and IEUBK model predicted GM BLL distributions.
*The null hypothesis that the samples were from the same distribution is rejected if the p-value was <0.05.
63

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 20. Cumulative distribution functions for observed and predicted BLLs for site-wide, censor level 2 (all predicted
and observed BLLs <30.5 pg/dL and no BDLs)
A) v2.0 IRs, 55/45 partition - Site-wide
Note: Probabilities shown are two-tailed
B) EFH IRs, 55/45 partition - Site-wide
Note: Probabilities shown are two-tailed
C) v2.0 IRs. 40/30/30 partition - Site-wide
Note: Probabilities shown are two-tailed
D) EFH IRs. 40/30/30 partition - Site-wide
Note: Probabilities shown are two-tailed
Predicted BLLs	— Observed BLLs
Predicted BLLs	Observed BLLs
¦ Predicted BLLs		 Observed BLLs
Predicted BLLs	Observed BLLs
64

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 21. Cumulative distribution functions for observed and predicted BLLs in the Box, censor level 2 (all predicted
and observed BLLs <30.5 pg/dL and no BDLs)
• Predicted BLLs
¦ Observed BLLs
¦ Predicted BLLs
¦ Observed BLLs
A) v2.0 IRs. 55/45 partition - Box
Note: Probabilities shown are two-tailed
B) EFH IRs. 55/45 partition - Box
Note: Probabilities shown are two-tailed
C) v2.0 IRs. 40/30/30 partition - Box
Note: Probabilities shown are two-tailed
D) EFH IRs. 40/30/30 partition - Box
Note: Probabilities shown are two-tailed
65

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 22. Cumulative distribution functions for observed and predicted BLLs in the Basin, censor level 2 (all predicted
and observed BLLs <30.5 pg/dL and no BDLs)
£ 0«
o
CL
a>
.>
0 4
3
E
3
o
0.2
10	15
BLL
Predicted BLLs
A) v2.0 IRs. 55/45 partition - Basin
Note: Probabilities shown are two-tailed
Pr > KSa 0.4469
Observed BLLs
Predicted BLLs
¦ Observed BLLs
B) EFH IRs, 55/45 partition - Basin
Note: Probabilities shown are two-tailed
Predicted BLLs
Observed BLLs
10
BLL
¦ Predicted BLLs
- Observed BLLs
Pr > KSa < 0001
C) v2.0 IRs, 40/30/30 partition - Basin
Note: Probabilities shown are two-tailed
D) EFH IRs. 40/30/30 partition - Basin
Note: Probabilities shown are two-tailed
66

-------
Advancing Pb Exposure and Biokinetic Modeling
5.4.2.2 Performance of default IEUBK v2.0 model
The K-S test rejected the null hypothesis for site-wide and Box comparisons, indicating that the
difference between the distributions of observed and predicted GM BLLs was statistically
significant. However, this was likely due to the large sample size (see Equation 7a) but a visual
review of the site-wide cumulative percentile plots in Figure 20 generally showed close
agreement between the observed and predicted BLLs, particularly in the upper percentiles. Site-
wide, the maximum difference between the two distributions was 0.15 and occurred at a BLL of
3.9 |jg/dL (Table 11 and Figure 20). When stratified by geographic area, the maximum
difference ranged from 0.07 in the Basin to 0.21 in the Box. The distributions of observed and
predicted BLL in the Basin were not significantly different for the 55/45 dust/soil partition.
5.4.2.3 Alternative IRsd and soil and dust partitions
Site-wide, the K-S test rejected the null hypothesis for the alternative IEUBK v2.0 models,
indicating that the difference between the two distributions was statistically significant. The
maximum deviation between the two distributions varied slightly from 0.17 (40/30/30 partition,
both IRsd) to 0.18 (55/45 partition, EFH IRsd; Table 11). The maximum deviation appears close
to the median (within 0.5 |jg/dL) of the observed BLLs in all four IEUBK v2.0 model
configurations (Figure 20).
When evaluating K-S tests and cumulative percentiles by Box and Basin (Table 11), the
smallest maximum deviation occurred in the Box using the 40/30/30 partition (both IRsd: 0.15)
and in the Basin using the 55/45 partition (both IRsd: 0.07). The K-S tests for the Basin
55/45 partition models failed to reject the null hypothesis, indicating that observed and predicted
BLLs were from the same distribution (regardless of IRsd); however, upon review of Figure 22,
the two distributions deviate above approximately 5 |jg/dL in the upper-most percentiles. In
contrast, the upper percentiles above approximately 5 |jg/dL in the Box models using the
40/30/30 partition appear to closely match (Figure 21). The distributions of the 55/45 partition
models, regardless of IRsd, appeared to closely match in the lowest BLLs of about <2.5 |jg/dL
for site-wide and the Box and Basin (see Figures 20-22).
5.5 Sensitivity Analyses
The following assumptions and parameters were modified in order to understand the impact on
model results.
5.5.1 Alternative Assumptions for Clean Backfill Soil Pb Concentration
The average of all soil backfill pile samples in any given year remediation occurred in the Box
was <100 mg/kg Pb, with no individual backfill sample exceeding 150 mg/kg (Idaho Legislature,
2007). Consequently, all BHSS data evaluations assumed that clean property soil was
100 mg/kg after remediation was complete. However, the potential for recontamination of the
remediated property soils in the past 3-30 years was unknown. No present-day property soil
sampling results were available to estimate more recent property soil Pb exposures. Therefore,
a value of 100 mg/kg continues to be used to represent remediated soil at a property in all
BHSS data evaluations, including in this model evaluation.
To evaluate the effect of the clean fill soil Pb concentration on the GM and Ps, the use of
100 mg/kg to represent soil Pb concentrations at a remediated property was replaced with a
clean soil value of 50 mg/kg. Table 12 presents summary statistics of the input soil
concentrations using the two different clean soil values. On average, soil exposure
concentrations differed by about 30 mg/kg when using 100 versus 50 mg/kg.
67

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 12. Comparison of soil summary statistics using different clean soil Pb concentrations in the Box for IEUBK v2.0
model, default IRsd, censor level 2 (dataset omits predicted and observed BLLs >30.5 pg/dL and observed BDL
BLLs)


55/45 Soil/Dust Partition

40/30/30 Soil/Dust Partition*





95% CI of




95% CI of






the




the





Arithmetic
Arithmetic
Standard


Arithmetic
Arithmetic
Standard
Clean Soil

Minimum
Maximum
Mean
Mean
Deviation
Minimum
Maximum
Mean
Mean
Deviation
(mg/kg)
N
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
100
875
31
9180
389
333-444
836
125
5573
628
590-666
572
50
875
31
9180
355
299-411
847
78
5566
600
562-639
578
Difference

34

28
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
68

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 13 presents predicted GM BLLs, average predicted P5 values, and 95% CIs. As expected,
the predicted GM BLLs were lower when the clean soil value of 50 mg/kg was used. The
predicted GM BLLs were about 0.1 |jg/dL lower using 50 versus 100 mg/kg regardless of
partitioning. Similarly, the difference between the average P5 values was also about 1% lower
using 50 versus 100 mg/kg regardless of partitioning (Table 13). Overall, using clean soil
concentrations between 50 and 100 mg/kg appeared to have minimal effect on the predicted
GM BLLs and P5 values.
5.5.2 Alternative Assumptions for GSDi
The IEUBK model's GSD, of 1.6 is an important parameter used in the estimation of P5.
Consequently, P5 was sensitive to changes in the GSD,. Consequently, differences in P5 were
evaluated from adjustments to GSD,.
Table 14 presents model results using alternative GSDs (calculated to capture 95% of observed
BLLs within the 95% Pis). For censor level 2, a calculated GSD, value of 1.97 encompassed
95% of the 1144 observed individual child BLLs (site-wide) compared to 86% when the default
GSD, of 1.6 was assumed (Table 14). The calculated GSD, value of 1.97 resulted in a site-wide
average Psof 31% (95% CIs: 28, 34) compared to an observed P5 of 32% (95% CIs: 29, 35)
and the original P5 of 27% (95% CIs: 24, 29) when the default GSD, was assumed. Site-wide by
age group, the P5 based on the calculated GSD, over-estimated the observed P5 for children
<2 years and under-estimated the observed P5 for children 2-<7 years of age.
The population GSD was expected to exceed the GSD/, since the population GSD included
variability in soil and dust Pb exposures. The calculated GSD, values that include 95% of
observations were always >1.6 and were greater than the observed GSDs, with the exception of
children 3-<4 years old.
The differences between average P5 values (P5 from calculated GSD, - P5 from default GSD/)
were about 4% site-wide and for the Box and the Basin (Table 14). The absolute difference
between average P5 values using calculated and default GSD, values ranged from about 1 to
7% by age (Table 14).
5.5.3 Alternative Assumptions for Dietary Pb Intake
Predicted GM BLLs using the IEUBK v2.0 default dietary Pb intakes were compared to
predictions made using dietary Pb intakes reported by Zartarian et al. (2017). To isolate the
effect of dietary Pb intake on predicted BLLs, Pb intakes due to soil, dust, water, and air were
set equal to zero.
Table 15 shows the dietary Pb intakes and predicted GM BLLs by age. Figure 23 shows the
predicted GM BLLs by age for both dietary Pb intake values. Differences in the predicted GM
BLLs ranged from approximately 0.5 |jg/dL for children <2 years old to approximately 0.3 |jg/dL
for the 3- to <7-year-old age range. For the 1- to <6-year-old age range used for site Pb risk
assessment, the GM BLLs estimated using the IEUBK v2.0 model default dietary Pb intakes
were approximately 0.4 |jg/dL greater than the GM BLLs estimated using the Zartarian et al.
(2017) dietary Pb. This magnitude of difference would not appreciably affect model performance
for predicting GM BLLs; however, the effect on the P5 depends on the exposure. For a soil
exposure of 200 mg/kg with all other parameter values set to default, the IEUBK v2.0 model
predicted a GM BLL of 2.3 |jg/dL and P5 of 5.0%; if the dietary estimates from Zartarian et al.
(2017) were used in place of default values, the resulting predicted GM BLL was 2.0 |jg/dL and
Ps was 2.3%.
69

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 13. Comparison of predicted BLL summary statistics using different clean soil Pb concentrations in the Box for
IEUBK v2.0 model, default IRsd, censor level 2 (dataset omits predicted and observed BLLs >30.5 pg/dL and
observed BDL BLLs)


55/45 Soil/Dust Partition
40/30/30 Soil/Dust Partition*
Clean Soil
(mg/kg)
N
GM (|Jg/dL)
95% CI of the
GM (|Jg/dL)
GSD (Hg/dL)
Average
Predicted
Probabilities
>5 |jg/dL (%)
95% CI for
Probabilities
>5 |jg/dL (%)
GM (|Jg/dL)
95% CI of the
GM (|Jg/dL)
GSD (|Jg/dL)
Average
Predicted
Probabilities
>5 |jg/dL (%)
95% CI for
Probabilities
>5 |jg/dL (%)
100
875
3.4
3.3-3.6
1.78
27.2
24.2-30.1
4.4
4.2-4.6
1.69
41.1
37.8-44.3
50
875
3.3
3.2-3.4
1.81
26.0
23.1-28.9
4.3
4.1-4.4
1.72
39.7
36.4-42.9
Difference

0.1

-0.04
1.20

0.1

-0.03
1.4

Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
Table 14. GSD sensitivity analysis for IEUBK v2.0 model, default IRsd, 55/45 partition, censor level 2 (dataset omits
predicted and observed BLLs >30.5 pg/dL and observed BDL BLLs)




Excel Calculated
IEUBK Output
Difference of





Average of





Average P5





Calculated P5

95% CI for P5

Original
95% CI for P5
(IEUBKP5

¦

Observed GSD
Calculated GSD*
(%)f
95% UCI
(%)
IEUBK GSD
Average P5 (%)
(%)
CalculatedP5)
Site-Wide (Total)
1144
1.80
1.97
30.8
33.5%
28.2 - 33.5
1.6
26.6
24.0-29.1
-4.3
Box (Total)
875
1.84
1.96
31.3
34.4%
28.3 - 34.4
1.6
27.2
24.2-30.1
-4.2
Basin (Total)
269
1.58
1.99
29.1
34.6%
23.7 - 34.6
1.6
24.6
19.5-29.8
-4.5
Site-Wide by Age (years)

n









0.5 to <1
70
1.71
2.22
47.2
58.9%
35.5 - 58.9
1.6
44.6
32.9 - 56.2
-2.6
1 to <2
138
1.74
1.81
47.8
56.1%
39.4-56.1
1.6
47.0
38.7 - 55.3
-0.8
2 to <3
202
1.82
2.01
32.6
39.0%
26.1 -39.0
1.6
28.0
21.8 - 34.2
-4.6
3 to <4
176
1.79
1.75
27.1
33.7%
20.6 - 33.7
1.6
25.0
18.6-31.4
-2.1
4 to <5
187
1.85
1.99
28.3
34.8%
21.9 - 34.8
1.6
23.0
17.0-29.1
-5.3
5 to <6
185
1.75
1.81
24.2
30.3%
18.0-30.3
1.6
21.1
15.3-27.0
-3.0
6 to <7
186
1.73
2.09
20.7
26.5%
14.9-26.5
1.6
13.4
8.5-18.3
-7.3
Notes:
*GSD based on 95% of observed BLLs fall within the 95% prediction intervals
fCalculated probability to exceed 5 ug/dL based on calculated GSD. Microsoft Excel Equation: (1-LOGNORM.DIST(5, LN(PredictedBLL), LN(CalculatedGSD),TRUE))*100
CI = 95% Confidence Interval
GSD = Geometric Standard Deviation
GSD: = Individual Geometric Standard Deviation
P5 = probability to exceed 5 |jg/dL
IRsd = soil/dust ingestion rate
70

-------
Advancing Pb Exposure and Biokinetic Modeling
Table 15. Diet sensitivity analysis for IEUBK v2.0 model
Age
(years)
Diet Intake (pg/day)
Zartarian et al.
v2.0 default (2017)
Predicted BLL (pg/dL)
Zartarian et al. Difference
v2.0 default (2017) (v2.0 - Zartarian)
0.5 to <1
2.66
0.70
0.71
0.20 0.52
1 to <2
5.03
2.58
0.97
0.47 0.50
2 to <3
5.21
3.44
0.96
0.60 0.36
3 to <4
5.38
3.54
0.92
0.61 0.31
4 to <5
5.64
3.57
0.91
0.59 0.32
5 to <6
6.04
3.85
0.90
0.58 0.32
6 to <7
5.95
3.80
0.85
0.54 0.30
Notes:
*Zartarian et al. (2017) did not provide a diet Pb intake for the 0.5- to <1-year age range. The diet intake for the 0.5- to
<1-year age range shown is the value Zartarian et al. (2017) used for the 0-to 0.5-year age range.
All IEUBK v2.0 model inputs were set to zero except for the dietary inputs listed.
Figure 23.
Comparison of predicted geometric mean BLLs (pg/dL) using the IEUBK
v2.0 default diet Pb intake estimates and diet Pb intake estimated by
Zartarian et al. (2017)
1.2
3 l.o
o
0.8
S 0.6 ¦
0.4
o
&>
0£>
- 0.2
o
s
c
£
o.o
—•—Version 2 default diet
- Version 2, Zartarian, et al. (2017) diet
6 11	12 23	24 35	36 47
Age (mouths)
48 59
60 71
72 84
71

-------
Advancing Pb Exposure and Biokinetic Modeling
5.5.4 Comparison of IE UBK 2.0 and v1.1 Model IRsd
The effect of the von Lindern et al. (2016) IRsd (IEUBK v2.0 IRsd) were compared to the IEUBK
v1.1 default IRsd. The IEUBK v1.1 IRsd values were based on several studies that used trace
metals and mass balance and were intended to be central tendency values (U.S. EPA. 1999).
The effect of these two IRsd values was evaluated using average, site-wide soil and dust
concentrations of 413 and 598 mg/kg from censor level 2 (Table 4), respectively. Default IEUBK
v2.0 values were used for all other model inputs. Table 16 compares the model predictions
using the IEUBK v2.0 IRsd and v1.1 IRsd. As expected, because of the difference in the values
for IRsd for all age groups other than 0.5-<1 years, the predicted GM BLLs using IEUBK v2.0
default IRsd were much lower than those using the IEUBK v1.1 values (Table 16). The predicted
GM BLLs for CERCLA recommended ages of 1- to <6-year-old children using the IEUBK v2.0
default and v1.1 IRsd were 4.6 and 6.7 |jg/dL, respectively, and the Ps values were 43 and 74%,
respectively.
Table 16. Comparison of GM PbB concentrations predicted by IEUBK v2.0 model
with default and IEUBK v1.1 model IRsd

IRsD(mg/day)
Predicted GM PbB Concentration
(|jg/dL)
Age (years)
v2.0
(von Lindern
et al.. 2016)
V1.1
(U.S. EPA. 1999)
IEUBK v2.0 with IEUBK v2.0 with
v2.0 IRsd v1.1 IRsd
0.5 to <1
86
85
h-
(D
h-
(D
1 to <2
94
135
6.4 7.9
2 to <3
67
135
4.8 7.5
3 to <4
63
135
4.2 7.2
4 to <5
67
100
o
CD
5 to <6
52
90
3.6 5.1
6 to <7
55
85
3.2 4.5
Notes:
IEUBK v2.0 model runs used default input values except for soil and dust concentrations, which were set to 413 and
598 mg/kg, respectively. See Appendix A and Table 3 for more information on model parameters and configurations.
In addition, the original scope of this evaluation included IEUBK v1.1 model runs using both
IEUBK v1.1 and v2.0 IRsd (Appendices F, G, and H present all IEUBK v1.1 results). Figure 24
presents predicted GM BLLs from three model runs (IEUBK v1.1 with v1.1 and v2.0 IRsd, and
v2.0 using v2.0 IRsd) relative to the observed site-wide GM BLLs by age. Comparing the two
model versions using the same IEUBK v2.0 IRsd also allowed for a more direct comparison of
other IEUBK v2.0 software changes. With the exception of the youngest and oldest age groups
(<1- and 6-year-old children), IEUBK v2.0 using v2.0 IRsd conformed to observed BLLs and
improves model fit relative to IEUBK v1.1 with either IRsd (Figure 24). This finding supports the
recommendation to focus on the 12- to 72-month age group for risk assessment using the
IEUBK model (U.S. EPA. 2017b).
Overall, the IRsd modifications to the IEUBK v2.0 model software improved model performance.
72

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure 24. Comparison of mean observed and predicted BLLs (pg/dL) with 95% CIs
using three combinations of model inputs and ingestion rates
0.5 to <1	1 to <2	2 to <3	3 to <4	4 to <5	5 to <6
Age (Years)
~ Observed • v2.0, v2.0 !R • v1.1, v2.0 IR v1.1, v1.1 IR
6 to <7
IEUBK v1.1 model runs used default parameters except for IRsd as shown in legend. IEUBK v2.0
model used default parameters including default IRsd. See Appendix A and Table 3 for more
information on model parameters and configurations.
5.6 Discussion
This report summarizes the first empirical evaluation of the IEUBK v2.0 model performance,
which includes several revisions to previously released versions of the model (e.g., adjustments
to parameters that govern exposures and intakes of Pb in air, diet, drinking water, and soil-
derived dust) and includes many children with observed BLLs <5 jjg/dL. The evaluation of the
IEUBK v2.0 model focused on three performance metrics: (1) prediction of population GM BLLs;
(2) prediction of probability of the BLL exceeding 5 [jg/dL (Ps); and (3) prediction of distribution
of individual child BLL. These performance metrics were selected for the below reasons.
Although USEPA recommended application of the model in the CERCLA program for predicting
the P10 (probability of the BLL exceeding 10 pg/dL) at residential exposure units (U.S. EPA.
1998, 1994a), more recent assessments by the Agency (U.S. EPA. 2013) suggested the need
to consider lower decision levels of Ps. Predicting the Ps depends on a prediction of the mean
73

-------
Advancing Pb Exposure and Biokinetic Modeling
BLL for each residence. Although predicting BLLs of individual children was not a recommended
application of the model, evaluation of the predicted mean BLL and P5 has relied, in this study
and in all previous studies, on BLL data of individual children. Consequently, the third
performance metric supported our understanding of how well the model predicted distributions
of individual child BLLs relative to an evaluation dataset such as the BHSS.
The BHSS data provided several important features that contributed to the strengths of this
study. These included:
• a relatively large number of child BLLs (1144 BLLs measured above detection or
reporting limits and <30.5 |jg/dL) that could be geographically matched to property soil
and house dust Pb concentrations and, for approximately 5% of children, measured Pb
in drinking water;
• the distribution of BLLs included a substantial number of BLLs (approximately 70%) that
were <5 |jg/dL;
• measurements of Pb RBA in BHSS soils and house dust that improved prediction of Pb
absorption and, thereby, mean BLLs at the site (von Lindern et al. 2016);
• estimates of IRsd at the BHSS supported the values for the IRsd parameter used in the
new IEUBK v2.0 model (von Lindern et al., 2016); and
• all PbB measurement were collected in July or August to capture peak seasonal PbB.
For this model evaluation, individual child BLLs were matched with measured concentrations of
Pb from house dust and property/community soils, and were assumed to be the major source of
exposures to Pb from soil and dust.13 The alternative partitioning of soil exposures was
considered and included contributions from nearby (neighborhood) community-wide soil, which
resulted in better model performance for older, likely more mobile, children.
Numerous factors other than measured soil and dust Pb concentrations could affect model
performance. Concentrations of Pb in drinking water were known from direct measurement of
tap water in approximately 5% of the BLLs (i.e., Basin properties not hooked to a public
system); for all other children, no tap water data were available. As a result, the IEUBK model
default concentration of 0.9 |jg/L was used for these children. Mean tap water concentrations
measured over a 17-year period in the Basin area tended to range between 0.9 and 1.5 |jg/L,
with one exception, when the mean was 4.6 |jg/L in 2009. This high mean tap water
concentration was a result of five children from two homes with water Pb concentrations
>7 |jg/L. An under-estimate of drinking water Pb concentration of 0.6 |jg/L would have lowered
the predicted GM BLL by <0.1 |jg/dL. In addition, all children were assigned the IEUBK model
default value for Pb in air of 0.1 |jg/m3. Annual averages in the Box for the period 1995-1998
reportedly ranged from 0.04 to 0.07 |jg/m3. Although these values were extremely uncertain, this
suggested that the model may have over-estimated exposures to air Pb. However, air makes a
negligible contribution to BLL relative to ingestion of Pb in soil, dust, water, and diet
(<0.1 |jg/dL). Consequently, using default air values was unlikely to have introduced substantial
error into the predictions of mean BLLs.
Difference in model results between the Box and Basin may be explained by: (1) how the
contamination came to be on each property, (2) the different spatial characteristics of properties
13 This assumption was not verified using questionnaire data, although it might be possible to do so from
analysis of data collected in home interviews that provided information on how much time the child
spends away from home, as was done in the Hoqan et al. (1998^) study. For example, a recent BHSS BLL
evaluation using questionnaire data and LHIP follow up information indicated that children with elevated
BLLs in 2017-2018 are likely exposed from recreational soils/sediments or other exposures (Alta, 20193).
74

-------
Advancing Pb Exposure and Biokinetic Modeling
and relative size of the communities, and (3) differences in the way that property soil and
community average Pb concentrations were calculated. Basin communities encompassed a
larger spatial extent than communities in the Box and many of the properties in the Basin were
larger than those in the Box. It was unknown if children living on large, rural properties (in
certain areas of the Basin) were as frequently exposed to neighboring or community soils as
often as a child living in a more typical town setting such as Kellogg or Smelterville (e.g., ease of
riding bikes or walking a few blocks to play with neighbors). If children in the Basin encountered
Pb exposures more often on their property than in neighboring or community areas, the
40/30/30 partition may not be as representative.
All children were assigned IEUBK model default values for dietary intake of Pb in market basket
foods. These values were derived from the WWEIA dietary interview component of the
NHANES 2003-2006 (CDC, 2010a, b) and the FDA TDS food contaminants data for 1995-2005
(FDA. 2010). The BHSS data used in this evaluation included data from before this timeframe.
Uncertainty in the dietary Pb exposures became increasingly important at lower population
mean BLLs because the contribution of food to total Pb uptake and BLL increased (Zartarian et
al., 2017). For example, the IEUBK v2.0 model predicted a mean BLL concentration of 2.3
|jg/dL (P5=5%) when the soil Pb level was 200 mg/kg (and air, water, and diet exposures were
set to national default values). At a soil Pb level of 200 mg/kg, the contribution of soil (and soil-
derived house dust) to total Pb uptake was predicted to be approximately 53%, while the
contribution of dietary Pb was approximately 42%.
The above limitations of information on individual child exposures would be potentially
significant if this study was intended to quantify variance in individual child BLL explained by the
model. However, this report focuses on the evaluation of model performance to predict mean
BLLs and Pswhen the model was applied as it would in a CERCLA HHRA. For these types of
assessments, data on dietary Pb, air Pb, exposures to Pb in settings outside of the child's
residence or estimates of IRsd are rarely available for use in the IEUBK model.
The findings from this model evaluation suggest that even in the absence of complete exposure
information (and therefore reliance on national default values for some contributing Pb
exposures), the model performed well at predicting the population mean BLL, Ps, and the overall
distribution of individual child BLLs in the BHSS population. Although the more general
applicability of these findings remains to be determined in future studies of other populations,
these results support the application of the IEUBK v2.0 model for informing risk-based
discussions regarding remediation of soils and mitigation of exposures at CERCLA sites where
the majority of the population BLLs are expected to be <5 |jg/dL.
Section 6 Conclusions
This section summarizes the conclusions of the IEUBK v2.0 model evaluation. The evaluation
used 1144 paired records (i.e., BLLs and environmental data) from an original dataset of 1283
paired records that were censored to exclude observed BLLs that were below the limit of
detection and to exclude predicted or observed BLLs that were >30.5 |jg/dL, thereby outside the
range of values that were used in the calibration and empirical validation of the model by Hogan
et al. (1998). Within this report these exclusion criteria are referred to as censor level 2.
Although analyses are provided that consider alternative ingestion rates and partitioning among
sources of ingested Pb from soil and dust, the performance assessment focused on the default
IEUBK v2.0 model IRsd derived from von Lindern et al. (2016) and the default 55/45 partition.
The objective was to evaluate the performance of IEUBK v2.0 using its defaults that were
recommended by the TRW for Pb Committee. The changes to the IEUBK model in updating
75

-------
Advancing Pb Exposure and Biokinetic Modeling
v1.1 to v2.0 (especially the TRW recommended changes to highly influential soil/dust ingestion
rates) mandated this evaluation of model.
6.1 Performance of Default IEUBK v2.0 Model
As discussed in Section 4.2, the evaluation of IEUBK v2.0 model performance was primarily
based on comparisons of model predictions to empirical BHSS population data using three
metrics: (1) population GM BLLs; (2) population probabilities of the BLL exceeding 5 |jg/dL; and
(3) the individual child BLL distributions.
The IEUBK model evaluation results based on these three metrics can be compared with the
prior evaluation by Hogan et al. (1998). Figure 5 shows observed and predicted BLLs, CIs, and
a line of unity (similar to Figure 1 in Hogan etal., 1998) for all model runs by multiple categories
(i.e., geographic area). Table 7 (no BDL or >30.5 |jg/dL) shows the model under predicted site-
wide GM BLL by 0.26 |jg/dL and was within ±0.4 |jg/dL among the three geographical areas.
Table F-6 (no >30.5 |jg/dL) shows the model under predicted GM BLL by 0.04 |jg/dL and was
within -0.3 to 0.6 |jg/dL among the three geographical areas. Most appropriately compared with
Table F-6, Table 3 of Hogan et al. (1998) shows the model under predicted the GM BLL by 0.10
|jg/dL and was within -0.6 to 0.7 |jg/dL among the three geographical areas. In our evaluation,
the site-wide Ps was underestimated by 3.6% (Table F-7), whereas Hogan et al. (1998) over
predicted P10 by 2.5% (based on Table 4 of their paper). It is unclear how to interpret this
difference due to the large difference in the BLL evaluated for exceedances, i.e., 5 vs. 10 |jg/dL.
Scatter plots (Figures F-5 through F-15) show observed and predicted BLLs with the 95% Pis
similar to Figures 2 through 4 in Hogan et al. (1998) for all model runs and select categories.
We found 86% of the observed individual child BLLs were located within the 95% PI of the
model (Table 10) as compared to 80% in the Hogan et al. (1998) study. Overall, we found better
agreement between predicted and observed BLL for IEUBK v2.0 than Hogan et al. (1998) found
for IEUBK v0.99d.
In general, the IEUBK v2.0 model performed well at predicting overall population GM BLLs
using the selected BHSS dataset. The site-wide predicted mean was within 0.26 |jg/dL of the
observed mean (Table 7) and resulted in a strong correlation between means by age groups
(r2=0.90; Figure 9). An error in predicted mean BLLs <1 |jg/dL is tolerable for applications to site
risk assessment, given the relatively large uncertainties in estimating Pb exposure and intake
(von Lindern et al.. 2003a; Bowers and Mattuck, 2001; Griffin et al.. 1999; Hogan et al.. 1998).
The default IEUBK v2.0 model predicted a mean Ps within 5.5% of the observed and showed
relatively low negative and positive bias in the prediction of the Pswhen applied to geographic
area (Table 9). There was generally close agreement between the site-wide empirical
distributions of observed individual child BLLs and predicted GM BLLs, particularly in the upper
percentiles (Figure 20). Although the distributions were statistically different using
nonparametric two-sample Kolmogorov-Smirnov test, the critical value for this test was relatively
small due to the large sample size, which led to rejections of the null hypothesis for small
differences that may not be meaningful for Pb risk assessment. The maximum difference
between the empirical distributions of observed individual child BLLs and predicted GM BLLs
appeared near the median. Prediction of individual child BLLs is beyond the application of the
IEUBK model; nevertheless, 86% of the observed individual child BLLs were located within the
95% PI of the model (Table 10). Collectively, these results support the application of the IEUBK
v2.0 model as it would be typically used in CERCLA-related HHRAs.
76

-------
Advancing Pb Exposure and Biokinetic Modeling
6.2	Alternative IRsd and Soil/Dust Partitions
Some additional analyses, described in Sections 4.2.4 and 4.2.5, were performed to explore
alternative values for some model parameters and in response to peer review comments
received on an early draft of this evaluation report. For example, since the IEUBK v2.0 default
IRsd differs from the EPA recommendations in the EFH (U.S. EPA. 2017a), model performance
was assessed and compared using both of these IRsd. This evaluation found that for some
categories of data, alternate soil/dust partitions and IRsd resulted in better conformance with
the selected BHSS dataset, although model predictions of GM BLL and Ps and the percent of
observations outside model Pis were more sensitive to changes in soil and dust partitions than
to IRsd.
The 40/30/30 partition was developed for the BHSS Box area. Unsurprisingly, the model
runs using the 40/30/30 partition (the Box Model) resulted in better predictions of GM BLL
and Ps in the Box (EFH IRsd resulted in the best predictions) and model predictions of BLLs
using the 40/30/30 partition also had a better goodness-of-fit with observed BLLs (lowest
SSE/SST). This finding shows that partitions other than the IEUBK v2.0 default may be
advised to improve model performance at some sites. The goodness-of-fit results were not
relied on as an index of model performance because it was observed that error in the upper
tails of the distribution can greatly increase SSE and thereby decrease the apparent
goodness-of-fit of model predictions with observed BLLs. Indeed, the goodness-of-fit results
in Table G-6 provide nearly the opposite conclusions related to model performance relative
to differences in population GM BLLs in Table 7, which were a primary focus of this
evaluation as discussed in Section 6.1. The 40/30/30 partition with EFH IRsd also had the
site-wide lowest percent (12%) of observed BLLs outside the model Pis (Table 10), which
was only marginally better than the 14% outside the Pis using the default 55/45 partition and
the IEUBK v2.0 IRsd. The 40/30/30 partition also had the lowest percentage of BLLs outside
the Pis in both the Box (13%) and Basin, although, the 55/45 IEUBK v2.0 IRsd in the Basin
performed equally well with only 11% outside the Pis. However, for all geographic areas
(site-wide, Box, and Basin) the IEUBK v2.0 default partition and IRsd had a more balanced
number of samples falling above and below the Pis than was observed for the alternative
partition and IRsd. When evaluating K-S tests and cumulative percentiles, the smallest
maximum deviation occurred in the Basin using the 55/45 partition (Figure 22).
These findings indicate that the IEUBK v2.0 model is robust to changes in partitions and
IRsd and allows for the application of alternative parameters when site-specific data were
available to support these adjustments (as is the case for the BHSS). Robust site data are
not typically available to support such adjustments to the model. Thus, these results support
the application of the IEUBK v2.0 model using its default parameter values as it would be
typically used in CERCLA-related HHRAs.
6.3	Model Sensitivity to Clean Backfill Soil Pb Concentration, GSD/, Dietary Pb
Intake, and IEUBK v1.1 Model Default IRsd
Sensitivity analyses for plausible alternative assumptions of clean soil backfill Pb
concentration, GSD/, and dietary intake in Section 5.5 showed that model predictions of GM
varied by <1 |jg/dL (results from clean soil concentration and dietary intake sensitivity
analyses) and predictions of P5 varied by <3% with changes to dietary intake and by 1-7%
with changes to GSD,. Additionally, the predicted GM BLLs using IEUBK v2.0 default IRsd
were much lower than those using the IEUBK v1.1 values for all age groups other than 0.5-<1
years.
77

-------
Advancing Pb Exposure and Biokinetic Modeling
The results of the sensitivity analyses indicate that the default IEUBK v2.0 model was robust
to slight changes in model parameters and that the updated IRsd used in the IEUBK v2.0
model resulted in vast improvements in model predictions relative to v1.1 IRsd. Overall,
these additional sensitivity analyses also support the use of the IEUBK v2.0 model as it
would be typically used in CERCLA-related HHRAs.
78

-------
Advancing Pb Exposure and Biokinetic Modeling
References
Alta (Alta Science Engineering, Inc.). (2019a). 2017 and 2018 blood lead data evaluation -
Bunker Hill mining and metallurgical complex superfund site. Kellogg, ID.
Alta (Alta Science Engineering, Inc.). (2019b). 2018 box house dust data summary report -Final.
Kellogg, ID.
Alta (Alta Science Engineering, Inc.). (2020). House dust evaluation for OU3 of the Bunker Hill
mining and metallurgical complex superfund site. Kellogg, ID.
Bowers. TS; Mattuck, RL. (2001). Further comparisons of empirical and epidemiological data
with predictions of the Integrated Exposure Uptake Biokinetic Model for lead in children.
Hum Ecol Risk Assess 7: 1699-1713. http://dx.doi.org/10.1080/20018091095311.
CPA Trust (Successor Coeur d'Alene Custodial and Work Trust). (2018). 2017 basin property
remediation program (BPRP) and house dust sampling data summary report (DSR).
Kellogg, ID.
CPA Trust (Successor Coeur d'Alene Custodial and Work Trust). (2019). 2018 basin property
remediation program (BPRP) and house dust sampling data summary report (PSR).
Kellogg, IP.
CPC (Centers for Pisease Control and Prevention). (2010a). National health and nutrition
examination survey. 2003-2004 Examination, dietary, and demographics files. Atlanta,
GA.
https://wwwn.cdc.gov/nchs/nhanes/ContinuousNhanes/Pefault.aspx?BeginYear=2003.
CPC (Centers for Pisease Control and Prevention). (2010b). National health and nutrition
examination survey. 2005-2006 Examination, dietary, and demographics files. Atlanta,
GA.
https://wwwn.cdc.gov/nchs/nhanes/ContinuousNhanes/Pefault.aspx? BeginYear=2005.
CPC (Centers for Pisease Control and Prevention). (2012). CPC Response to Advisory
Committee on Childhood Lead Poisoning Prevention Recommendation in "Low level
lead exposure harms children: A renewed call for primary prevention." Atlanta, GA.
https://www.cdc.gov/nceh/lead/docs/cdc_response_lead_exposure_recs.pdf.
CH2M Hill. (1991). House dust remediation report for the Bunker Hill CERCLA site populated
areas RI/FS. (BHPA-HPR-F-R0-052091). Boise, IP: Idaho Pepartment of Health and
Welfare.
FPA (U.S. Food and Prug Administration). (2010). Total Piet Study. Silver Spring, MP.
https://www.fda.gov/food/science-research-food/total-diet-studv.
Griffin, S: Marcus, A: Schulz, T: Walker, S. (1999). Calculating the interindividual geometric
standard deviation for use in the integrated exposure uptake biokinetic model for lead in
children. Environ Health Perspect 107: 481-487. http://dx.doi.org/10.2307/3434631.
Hogan, K; Marcus, A: Smith, R; White, P. (1998). Integrated exposure uptake biokinetic model
for lead in children: empirical comparisons with epidemiologic data. Environ Health
Perspect 106: 1557-1567. http://dx.doi.oro/10.1289/ehp.98106s61557.
Idaho Pepartment of Health and Welfare. (2000). Coeur d'Alene River Basin environmental
health exposure assessment. Atlanta, GA: Agency for Toxic Substances and Pisease
Registry.
https://ntrl.ntis.gOv/NTRL/dashboard/searchResults/titlePetail/PB2000107701.xhtml#.
79

-------
Advancing Pb Exposure and Biokinetic Modeling
Idaho DEQ (Idaho Department of Environmental Quality). (2005). 2004 Air Quality Data
Summary. Boise, ID.
Idaho DEQ (Idaho Department of Environmental Quality). (2020). OU1 -Water Wells. Boise, ID.
Idaho Legislature. I PAPA 41.01.01: Rules of Idaho public health district#! (2007).
https://adminrules.idaho.gov/rules/2011/41/0101.pdf
Li. Y: Hu. J: Wu. W: Liu. S: Li. M: Yao. N: Chen. J: Ye. L: Wang. Q: Zhou. Y. (2016). Application
of IEUBK model in lead risk assessment of children aged 61-84months old in central
China. Sci Total Environ 541 (January 2016): 673-682.
McCulley, Frick,, and Gilman, (McCulley, Frick, and Gilman Inc.). (1997). Completion of
remedial action-Certification report for Smelterville reasonable segregable area. Osburn,
ID.
Morrison-Knudsen (Morrison-Knudsen Corporation). (1999). Bunker Hill remedial action project
closure reports for contract DACW68-94-D-0005. (MK Project No. 4415). Boise, ID.
Newcombe, RG. (1998). Two-sided confidence intervals for the single proportion: Comparison
of seven methods. Stat Med 17: 857-872. http://dx.doi.Org/10.1002/(SICI)1097-
0258(19980430)17:8<857::AID-SIM777>3.0.CO:2-E.
NRC (National Research Council). (2005). Superfund and mining megasites: Lessons from the
Coeur d'Alene River basin. Washington, DC: National Academies Press.
http://dx.doi.org/10.17226/11359.
OMB (U.S. Office of Management and Budget). (2004). Final information quality bulletin for peer
review. (Memorandum M-05-03). Washington, DC.
https://obamawhitehouse.archives.gov/sites/default/files/omb/assets/omb/memoranda/fv
2005/m05-03.pdf.
Panhandle Health District. (1986). Kellogg revisited-1983: Childhood blood lead and
environmental status
report, http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockev=9101QTYB.TXT.
Parsons, R; Munuo, SS: Buckman, DW; Tooze, JA; Dodd, KW. (2009). User's guide for analysis
of usual intakes: For use with versions 1.1 of the Mixtran, Distrib, and Indivint SAS
macros. Rockville, MD: National Cancer Institute.
https://epi.grants.cancer.gov/diet/usualintakes/Users Guide Mixtran Distrib Indivint 1.
1.pdf.
Paulson. JA: Brown. MJ. (2019). The CDC blood lead reference value for children: Time for a
change [Letter], Environ Health 18: 16. http://dx.doi.Org/10.1186/s12940-019-0457-7.
Petrosvan, V: von Braun, MC: Spalinger. SM: von Lindern, IH. (2006). Seasonal variations of
lead concentration and loading rates in residential house dust in northern Idaho. J
Hazard Mater 132: 68-79. http://dx.doi.Org/10.1016/i.ihazmat.2005.11.093.
Snedecor, GW; Cochran, WG. (1989). Statistical methods (8th ed.). Ames, IA: Iowa State
University Press.
Spalinger, SM: von Braun, MC: Petrosvan, V; von Lindern, IH. (2007). Northern Idaho house
dust and soil lead levels compared to the Bunker Hill superfund site. Environ Monit
Assess 130: 57-72. http://dx.doi.Org/10.1007/s10661-006-9450-z.
TerraGraphics (TerraGraphics Environmental Engineering). (1990). Risk assessment data
evaluation report (RADER) for the populated areas of the Bunker Hill Superfund Site.
80

-------
Advancing Pb Exposure and Biokinetic Modeling
TerraGraphics (TerraGraphics Environmental Engineering). (2000). Final 1999 five year review
report: Bunker Hill Superfund Site. Moscow, ID.
TerraGraphics (TerraGraphics Environmental Engineering). (2001a). 1999 Five year review
report Bunker Hill Superfund Site populated areas.
TerraGraphics (TerraGraphics Environmental Engineering). (2001b). 2000 Interior school dust
data summary report.
TerraGraphics (TerraGraphics Environmental Engineering). (2004). Final human health
remedial evaluation report for the Bunker Hill Superfund Site box. Moscow, ID.
TerraGraphics (TerraGraphics Environmental Engineering). (2006). Final Coeur d'Alene basin
blood lead absorption and exposure report revision 3.
TerraGraphics (TerraGraphics Environmental Engineering). (2008). Background and rationale
for proposed modifications to the Bunker Hill Superfund Site OU 1 house dust remedial
action objectives.
TerraGraphics (TerraGraphics Environmental Engineering). (2010). Assessment of residential
and community mean soil exposure indices in Operable Unit 3, Bunker Hill Superfund
Site: Revision 1.
TerraGraphics (TerraGraphics Environmental Engineering). (2012). Work plan for: "Estimating
soil ingestion through the blood lead: soil/dust lead relationship at the Bunker Hill
Superfund Site in the Silver Valley of Idaho".
TerraGraphics (TerraGraphics Environmental Engineering). (2015). Bunker Hill Mining and
Metallurgical Complex Superfund Site OU1 2013 house dust and blood lead data and
risk management evaluation.
TerraGraphics (TerraGraphics Environmental Engineering). (2016). 2015 Basin house dust data
summary report for the Bunker Hill Mining and Metallurgical Complex Superfund Site.
TerraGraphics. URS Greiner.. and CH2M Hill. (TerraGraphics Environmental Engineering, URS
Greiner, and CH2M Hill). (2001c). Final human health risk assessment for the Coeur
d'Alene Basin extending from Harrison to Mullan on the Coeur d'Alene River and
tributaries: Remedial investigation/feasibility study. Boise, ID: Idaho Department of
Environmental Quality.
TerraGraphics and PHD (TerraGraphics Environmental Engineering and Panhandle Health
District). (2005). Comparison of US Department of Housing and Urban Development
(HUD) risk assessment methodology to methods used at the Bunker Hill Superfund site
in Idaho for the identification of the risk from lead in house dust. Washington, DC: U.S.
Department of Housing and Urban Development.
TerraGraphics and USACE (TerraGraphics Environmental Engineering and U.S. Army Corps of
Engineers). (2002). Bunker Hill house dust pilot final remedial effectiveness report.
Tooze, JA; Midthune, D; Dodd, KW; Freedman, LS: Krebs-Smith, SM; Subar, AF; Guenther,
PM: Carroll. RJ: Kipnis, V. (2006). A new statistical method for estimating the usual
intake of episodically consumed foods with application to their distribution. J Am Diet
Assoc 106: 1575-1587. http://dx.doi.orq/10.1016/i.iada.2006.07.003.
U.S. EPA (U.S. Environmental Protection Agency). (1991). EPA Superfund Record of Decision,
Bunker Hill Mining and Metallurgical Complex Residential Soils Operable Unit,
Shoshone County, Idaho. Washington, DC.
81

-------
Advancing Pb Exposure and Biokinetic Modeling
U.S. EPA (U.S. Environmental Protection Agency). (1992). Superfund record of decision:
Bunker Hill Mining and Metallurgical Complex Residential Soils Operable Unit Shoshone
County, Idaho. (EPA/ROD/R10-92/041). Washington, DC.
https://nepis.epa.qov/Exe/ZvPURL.cqi?Dockev=91000XXB.txt.
U.S. EPA (U.S. Environmental Protection Agency). (1994a). Guidance manual for the integrated
exposure uptake biokinetic model for lead in children [EPA Report], (EPA/540/R-93/081).
Washington, DC. http://nepis.epa.qov/Exe/ZvPURL.cqi?Dockev=2000WN4R.txt.
U.S. EPA (U.S. Environmental Protection Agency). (1994b). Memorandum: OSWER directive:
Revised interim soil lead guidance for CERCLA sites and RCRA corrective action
facilities. (OSWER Directive #9355.4-12; EPA/540/F-94/043). Washington, DC.
https://nepis.epa.qov/Exe/ZvPURL.cqi?Dockev=P100PJUV.txt.
U.S. EPA (U.S. Environmental Protection Agency). (1994c). Technical support document:
Parameters and equations used in integrated exposure uptake biokinetic model for lead
in children (v 0.99d) [EPA Report], (EPA/540/R-94/040). Washington, DC.
https://ntrl.ntis.gov/NTRI-/dashboard/searchResults/titleDetail/PB94963505.xhtml.
U.S. EPA (U.S. Environmental Protection Agency). (1998). Clarification to the 1994 revised
interim soil lead guidance for CERCI-A Sites and RCRA corrective action facilities.
(EPA/540/F-98/030). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (1999). Short Sheet: IEUBK model soil/dust
ingestion rates. (EPA 540-F-00-007). Washington, DC.
https://nepis.epa.qov/Exe/ZvPURL.cqi?Dockev=90181206.txt.
U.S. EPA (U.S. Environmental Protection Agency). (2002). The Bunker Hill Mining and
Metallurgical Complex Operable Unit 3 Record of Decision. (EPA/ROD/R10-02/032).
Washington, DC. https://nepis.epa.qov/Exe/ZvPURL.cqi?Dockev=P100SHZQ.txt.
U.S. EPA (U.S. Environmental Protection Agency). (2005a). Basin Environmental Monitoring
Plan (BEMP). Washington, DC.
U.S. EPA (U.S. Environmental Protection Agency). (2005b). Second five-year review for the
Bunker Hill Mining and Metallurgical Complex Superfund Site Operable Units 1, 2, and 3,
Idaho and Washington. (EPA 910-R-05-006). Seattle, WA.
https://semspub.epa.qov/work/10/100029538.pdf.
U.S. EPA (U.S. Environmental Protection Agency). (2007c). Guidance for evaluating the oral
bioavailability of metals in soils for use in human health risk assessment.
(OSWER 9285.7-80). Washington, DC.
https://nepis.epa.gov/Exe/ZyPDF. eg i/93001C3I.PDF?Dockey=93001C3l. PDF.
U.S. EPA (U.S. Environmental Protection Agency). (2007a). Estimation of relative bioavailability
of lead in soil and soil-like materials using in vivo and in vitro methods [EPA Report],
(OSWER 9285.7-77). Washington, DC.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=93001 C2U.txt.
U.S. EPA (U.S. Environmental Protection Agency). (2007b). User's guide for the Integrated
Exposure Uptake Biokinetic Model for Lead in Children (IEUBK) Windows® [EPA
Report], (EPA 9285.7-42). Washington, DC.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P1002RKA.txt.
U.S. EPA (U.S. Environmental Protection Agency). (2009). Validation assessment of in vitro
lead bioaccessibility assay for predicting relative bioavailability of lead in soils and soil-
82

-------
Advancing Pb Exposure and Biokinetic Modeling
like materials atsuperfund sites. (OSWER 9200.3-51). Washington, DC.
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P100GELG.txt.
U.S. EPA (U.S. Environmental Protection Agency). (2010a). 2010 five-year review for the
bunker hill mining and metallurgical complex superfund site operable units 1, 2, and 3
Idaho and Washington. Seattle, WA: U.S. Environmental Protection Agency, Region 10.
U.S. EPA (U.S. Environmental Protection Agency). (2010b). The analysis of regulated
contaminant occurrence data from public water systems in support of the second six-
year review of national primary drinking water regulations. (EPA/815/B09/006).
Washington, DC: U.S. Environmental Protection Agency, Office of Ground Water and
Drinking Water.
https://ntrl.ntis.gov/NTRI-/dashboard/searchResults/titleDetail/PB2010108286.xhtml.
U.S. EPA (U.S. Environmental Protection Agency). (2010c). Final six-year review of national
primary drinking water regulations: "Final_6Yr_Lead_12.23.10.accdb.". Washington, DC:
Office of Groundwater and Drinking Water.
U.S. EPA (U.S. Environmental Protection Agency). (2012). Standard operating procedure for an
in vitro bioaccessibility assay for lead in soil. (EPA/9200.2-86). Washington, DC.
https://nepis.epa.gov/Exe/ZyPDF.cgi/P100GESL.PDF?Dockey=P100GESL.PDF.
U.S. EPA (U.S. Environmental Protection Agency). (2013). Integrated science assessment for
lead [EPA Report], (EPA/600/R-10/075F). Research Triangle Park, NC: U.S.
Environmental Protection Agency, National Center for Environmental Assessment.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=255721.
U.S. EPA (U.S. Environmental Protection Agency). (2015). Fourth five-year review report for
bunker hill superfund site, Shoshone and Kootenai counties, Idaho. Seattle, WA.
U.S. EPA (U.S. Environmental Protection Agency). (2016). OLEM policy directive: Updated
scientific considerations for lead in soil cleanups. (OLEM Directive 9200.2-167).
Washington, DC: U.S. Environmental Protection Agency, Office of Land and Emergency
Management, https://semspub.epa.gov/work/08/1884204.pdf.
U.S. EPA (U.S. Environmental Protection Agency). (2017a). Exposure factors handbook chapter
5 (update): Soil and dust ingestion. (EPA/600/R17/384F). U.S. Environmental Protection
Agency, Office of Research and Development.
https://cfpub.epa.gov/ncea/efp/recordisplav.cfm?deid=337521.
U.S. EPA (U.S. Environmental Protection Agency). (2017b). Recommendations for default age
range in the IEUBK model. (OLEM Directive 9200.2-177). Washington, DC: U.S.
Environmental Protection Agency, Office of Land and Emergency Management.
https://semspub.epa.gov/work/HQ/10000Q689.pdf.
U.S. EPA (U.S. Environmental Protection Agency). (2017c). Update of the adult lead
methodology's default baseline blood lead concentration and geometric standard
deviation parameters and the integrated exposure uptake biokinetic model's default
maternal blood lead concentration at birth variable. (OLEM Directive 9285.6-56).
Washington, DC: U.S. Environmental Protection Agency, Office of Land and Emergency
Management. https://semspub.epa.gov/work/HQ/196766.pdf. 10/18/2019.
U.S. EPA (U.S. Environmental Protection Agency). (2020). Review of dust-lead post-abatement
clearance levels, proposed rule. Fed Reg 85: 37810-37819.
von Lindern, I: Spalinger, S: Petroysan, V; von Braun, M. (2003a). Assessing remedial
effectiveness through the blood lead:soil/dust lead relationship at the Bunker Hill
83

-------
Advancing Pb Exposure and Biokinetic Modeling
Superfund site in the Silver Valley of Idaho. Sci Total Environ 303: 139-170.
http://dx.doi.orq/10.1016/S0048-9697(02)00352-2.
von Lindern, I: Spalinger, S: Stifelman, ML: Stanek, LW; Bartrem, C. (2016). Estimating
children's soil/dust ingestion rates through retrospective analyses of blood lead
biomonitoring from the Bunker Hill Superfund Site in Idaho. Environ Health Perspect
124: 1462-1470. http://dx.doi.org/10.1289/ehp.1510144.
von Lindern. IH: Spalinger. SM: Bero, BN: Petrosvan, V: von Braun, MC. (2003b). The influence
of soil remediation on lead in house dust. Sci Total Environ 303: 59-78.
http://dx.doi.orq/10.1016/S0048-9697(02)00356-X.
York, D; Evensen, NM; Martinez, ML: Delgado, JD. (2004). Unified equations for the slope,
intercept, and standard errors of the best straight line. American Journal of Physics 72:
367-375. http://dx.doi.orci/10.1119/1.1632486.
Zartarian, V; Xue, J: Tornero-Velez, R; Brown, J. (2017). Children's lead exposure: A multimedia
modeling analysis to guide public health decision-making. Environ Health Perspect 125:
097009. http://dx.doi.org/10.1289/EHP1605.
84

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix A
Default Inputs to IEUBK (v1.1 and v2.0)
A-1

-------
Advancing Pb Exposure and Biokinetic Modeling
Table A-1. IEUBK input parameters based on TRW Lead Committee recommendation

lEUBKwin v1.1 build 111
IEUBK v2.0
Parameter
Value
Basis
Value
Basis
Soil
concentration
(mg/kg)
Csoil
EPC soil concentration
for the DU or residential
yard
Csoil
EPC soil concentration
for the DU or residential
yard
Dust
concentration
(mg/kg)
Cdust = 0.7 • Csoil(weighted)
+
(air cone • 100)
OR
Cdust
Indoor dust lead is
derived from residential
soil data using Msd
(default shown)
or
using site-specific data
Cdust = 0.7 • Csoil(weighted)
+
(air cone • 100)
OR
Cdust
Indoor dust lead is
derived from residential
soil data using Msd
(default shown)
or
using site-specific data
Outdoor air
concentration
(jjg per cubic
meter [m3])
0.1
IEUBK Default
0.1
IEUBK Default
Indoor air
concentration
(|jg/m3)
30% of outdoor air
concentration
IEUBK Default
30% of outdoor air
concentration
IEUBK Default
Drinking water
concentration
(|jg per liter [L])
4
IEUBK Default
0.9
EPA (U.S. EPA. 2010b.
c)
Maternal PbB at
birth (jjg/dL)*
1
IEUBK Default
0.6
Based on National
Health and Nutrition
Examination Survey
(NHANES) update
(2009-2014); see U.S.
EPA (2017c)
Absorption
Fractions (AFP*)
at low intakes:
Air
Diet
Water
Soil/dust
Sediment and
disturbed surface
water
50%
50%
30%
Site-specific or 30%
Site-specific or 30%
IEUBK Default
or
site-specific based on
U.S. EPA (2007tf
50%
50%
30%
Site-specific or 30%
Site-specific or 30%
IEUBK Default
or
site-specific based on
U.S. EPA (2016")
Fraction soil
45%

45%
IEUBK Default
GSDi (individual
geometric
standard
deviation)
1.6
IEUBK Default
1.6
IEUBK Default
Target PbB
5 jjg/dL

5 jjg/dL

1U.S. EPA (1994a, c).
*Maternal PbB at birth does not impact results for the 6- to 72-month age range.
tAFP = RBA% * 0.5.

-------
Advancing Pb Exposure and Biokinetic Modeling
Table A-2. Age-dependent inputs to the IEUBK v1.1win model Build 11

Air2
Diet2
Water2
Soil-Dust2
Age1
(months)
Time
Outdoors
(hours)
Ventilation
Rate
(m3/day)
Dietary
Intake
(Ijg/day)
Intake
(L/day)
v1.1Intake
IRsd
(mg/day)
6 to <12
1
2
2.26
0.2
85
12 to <24
2
3
1.96
0.5
135
24 to <36
3
5
2.13
0.52
135
36 to <48
4
5
2.04
0.53
135
48 to <60
4
5
1.95
0.55
100
60 to <72
4
7
2.05
0.58
90
72 to <84
4
7
2.22
0.59
85
1The age range of the IEUBK v1.1 model is 6-84 months (U.S. EPA. 1994a).
2The values shown are the midpoint of the age range. More information related to default input parameters are
available in U.S. EPA (2007b).
Table A-3. Age-dependent inputs to the IEUBK v2.0 model Build 1.6

Air
Diet3
Water4
Soil-Dust5
Age1
(months)
Time
Outdoors
(hours)
Ventilation
Rate2
(m3/day)
Dietary
Intake
(Ijg/day)
Intake
(L/day)
v2.0
IRsd
(mg/day)6
EFH
IRsd
(mg/day)7
6 to <12
1
3.22
2.66
0.40
86
70
12 to <24
2
4.97
5.03
0.43
94
90
24 to <36
3
6.09
5.21
0.51
67
60
36 to <48
4
6.95
5.38
0.54
63
60
48 to <60
4
7.68
5.64
0.57
67
60
60 to <72
4
8.32
6.04
0.60
52
60
72 to <84
4
8.89
5.95
0.63
55
60
1The age range of the IEUBK v2.0 model is 12-72 months, which is also the recommended age range for Superfund
sites using the IEUBK model (U.S. EPA. 2017b).
2The values shown are the midpoint of the age range. IEUBK (v2) model uses a regression equation to calculate
inhalation rate as a continuous non-linear function of age (U.S. EPA. 2016).
3TRW Lead Committee analysis (U.S. EPA. 2016) of Dietary Pb Concentration 1995-2005 TDS (FDA. 2010): Dietary
Intake 2003-06 NHANES WWEIA (CDC. 2010a. b); Methodology NCI Method (Parsons et al.. 2009; Tooze etal..
2006).
4TRW Lead Committee Analysis (U.S. EPA. 2016) of Office of Water Six-Year data (U.S. EPA. 2010b).
5Two soil-dust ingestion rates evaluated using the IEUBK v2.0 model build 1.6.
6TRW Lead Committee analysis (U.S. EPA. 2016) of age-specific values derived from structural equations modeling
of children from Bunker Hill Superfund site assuming 50% dust, 25% yard soil, 10% neighborhood soil, and 15%
community soil (von Lindern et al.. 2016).
7General population central tendency values as described in Table 5-1 of Exposure Factors Handbook Chapter 5
(U.S. EPA. 2017a).

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix B
Supplemental BHSS Sampling/Monitoring Data
B-1

-------
Advancing Pb Exposure and Biokinetic Modeling
Additional Sampling/Monitoring and Data
This appendix provides summaries of additional BHSS data collected during sampling/
monitoring events that occurred for specific reasons. Multiple events have been conducted in
both the Box and the Basin. These data could also be useful for this or future model evaluation
if comparisons of observed BLL to predicted IEUBK modelling appear disparate.
House Dust Remediation Pilot -1990
An interior remediation investigation pilot project was first conducted in 1990 to compare the
efficiency of cleaning and removal of in-home Pb. The study was conducted at six houses in the
populated areas of the Box following the first year of residential soil cleanup in 1990. Main living
area carpets, rugs, and one piece of upholstered furniture were vacuumed for 1.5 minutes per
square yard, cleaned with shampoo, and then removed. Floors were then wet washed after
removal of the carpet. Carpet vacuum samples were taken before and after the shampoo
cleaning, and pieces of the carpet and pad were ground into a powder for laboratory analysis.
Water and sludge from the shampoo cleaning were also retained as samples. Additional
vacuum dust and shampoo effluent samples were collected from furniture and rugs. One year
later, subsequent sampling of interior house dust was conducted to measure long-term
effectiveness of the one-time interior remediation (CH2M Hill. 1991).
Basin Exposure Survey -1996
In 1996, the Idaho Department of Health and Welfare (IDHW), in partnership with PHD,
evaluated exposures to individuals living in the Coeur d'Alene River Basin (outside the Box) to
Pb and cadmium in the environment. This study included individuals within up to 1.5 miles
beyond the 100-year flood plain of the South Fork and the main stem of the Coeur d'Alene
River. Of the estimated 1643 households in the study area, 1513 participated in the census.
Household and personal questionnaires were collected from 836 households to evaluate risk
exposure factors. The individual questionnaire included questions about personal behaviors,
occupations, hobbies, tobacco use, and garden vegetable consumption. Additionally, PbB and
urine cadmium samples were collected from participants (765 and 752 samples, respectively)
as well as up to five environmental source exposure samples (i.e., soil, water, house dust from
vacuums and floors mats, and house paint) (Idaho Department of Health and Welfare. 2000). Of
the 765 individuals who participated in the PbB screening, 98 were between the ages of 0 and
9.
Northern Idaho House Dust and Soil Lead Levels Compared to the Bunker Hill
Superfund Site -1999
To better understand the contribution of Pb contamination in homes within the BHSS, a study
was conducted to evaluate dust Pb concentration and dust and Pb loading rates in towns
comparable to BHSS communities but unaffected by the mining industry. Ninety-six dust and
50 yard soil samples collected in five Idaho towns (Bovill, Coeur d'Alene, Moscow, Post Falls,
and Potlatch) demographically similar and dissimilar to the BHSS were compared to samples
collected in three Box communities (Kellogg, Pinehurst, and Smelterville). Dust samples were
collected from vacuum bags (49 samples) and dust mats (47 samples) from March to May 1999.
Data from a household questionnaire was also collected from each household sampled and
included information such as the number of people residing in the home, age of the house,
interior remodeling of the house, and interior and exterior paint conditions were also collected
(Spalinqer et al.. 2007).

-------
Advancing Pb Exposure and Biokinetic Modeling
Seasonal Variations of Lead Concentration and Loading Rates in Residential
House Dust in northern Idaho -1999
In this study, seasonality of dust Pb concentrations and Pb and dust loading rates in residential
homes were evaluated in five towns (Bovill, Coeur d'Alene, Moscow, Post Falls, and Potlatch) in
northern Idaho. Dust and yard soil samples were collected from 34 homes outside of Pb-
contaminated areas. Soil samples were collected in March 1999 while vacuum and floor mat
dust samples were collected every 2 months from March to November 1999. A questionnaire
was completed for each participating house with information on house age, general condition of
house, number of people living in the house, resident habits, activities and occupations,
presence of pets, household income, and education (Petrosvan et al.. 2006).
House Dust Remediation Pilot - 2000
Another house dust interior remediation investigation was conducted in 2000 to assess the
effectiveness of commercial cleaning services and removal of permanent reservoirs of Pb dust
in the Box. Eighteen homes in Smelterville with an additional five control homes participated in
the study. A questionnaire addressing basic housing and resident characteristics was obtained
from each participating household. Dust samples were collected from vacuum bags, dust mats,
and dust wipes. Additional floor samples and occasional attic and/or basement samples were
collected using the Baltimore Repair and Maintenance method; a composite dust sample was
only collected from an attic and/or basement if the attic and/or basement was not used for living
space and was accessible. Filters from air duct cleaning equipment were also weighed and
sampled for Pb concentration. Indoor air was monitored for dust during the U.S. Department of
Housing and Urban Development (HUD) and commercial cleaning treatments. Even though Pb-
based paint was not included in the remediation process, assessment of interior paint was
completed using an XRF. Samples were taken several times during the project including pre-
and post-cleaning, during cleaning, and 6-12 months following cleaning (TerraGraphics and
USACE. 2002).
Interior School Dust Data Summary - 2000
Dust in nine BHSS schools (in Silverton, Wallace, Mullan, and Osburn) was sampled by the
State in November 2000 as a result of a court order from the Fourth Judicial District Court of the
State of Idaho over concern of Pb contamination. Four sampling methods were used to collect
dust: vacuums, dust mats, Baltimore Repair and Maintenance, and dust wipes. Fountain
drinking water and indoor air samples were also collected (TerraGraphics. 2001b).
HUD Risk Assessment Grant - 2004
PHD was awarded a HUD grant in 2004 to evaluate the relationship between risk assessment
approaches for estimating exposure to Pb from soil and paint in house dust. Baseline data were
collected for house dust Pb levels and Pb paint conditions in the BHSS and rural Idaho
communities outside the BHSS. The study compared two methods for interior risk assessment:
(1) HUD's methodology of using XRF for Pb paint concentration and dust wipes for dust
collection and (2) the BHSS technique of collecting dust samples from vacuums and floor mats.
For comparison, the study was performed at 75 residential units (primarily units built prior to
1960) in the Box, Basin, and communities outside the mining district with comparable home-age
and socio-economic status. A questionnaire was used to screen homes based on home age,
families with children, and the presence of a vacuum cleaner. In addition to dust and paint, soil
samples were taken using both HUD and BHSS protocols (TerraGraphics and PHD. 2005).

-------
Advancing Pb Exposure and Biokinetic Modeling
Door-to-Door Survey - 2004
In 2004, a door-to-door survey was conducted for 8-10 weeks in the Coeur d'Alene River Basin
to identify high-risk properties, encourage residents to participate in the PbB program, and
solicit consent for the BPRP. Consent forms and questionnaires about whether children or
pregnant women resided at a home, and the age of the children was completed.
Mine and Mill Sites
Remediation at several mine and mill sites is being implemented to minimize exposure to
recreational users and reduce Pb and sediment loading to surface water. Additional mine and
mill sites were identified in the 2012 ROD Amendment and are currently being prioritized by
USEPA for remedial action by the CDA Trust (U.S. EPA. 2015). A number of mine and mill sites
are located near residential properties in the BHSS. Data for these sites include analytical
metals concentrations in soil samples, XRF readings in soils, spatial coordinates, and physical
descriptors.
Floor Mat House Dust Data
In addition to vacuum samples, floor mat data were also considered for this IEUBK evaluation
as a potential house dust data source to pair with blood and soil data. Before adding floor mat
Pb concentrations to the paired dataset, the comparability and representativeness of these two
sampling approaches was evaluated. Paired vacuum and floor mat Pb concentrations obtained
from the same home were evaluated to determine if mat Pb concentrations are generally similar
to vacuum concentrations. Prior comparisons of mat and vacuum samples conducted using Box
data indicated that before 2003, mat Pb levels in the Box were generally higher than vacuum
bag Pb levels, and from 2003 to 2013, paired data were no longer significantly different
(TerraGraphics. 2006). Dust data from the Box in 2018 have not been compared.
Paired vacuum and mat samples from the Basin have not been compared in recent years. A
cursory comparison using Basin dust data collected between 2010 and 2018 indicates
significant differences between the two sampling methods. Given these findings, time and
budget limitations for this project, and the impacts of changes in mat models and dust extraction
techniques described in Section 2.2, floor mat data from the Box and Basin are not included in
the paired dataset for this evaluation. However, if floor mat concentrations were to be included
in the dataset for those homes where no vacuum data are available, 54 paired records from the
Box and 187 paired records from the Basin would be added to the dataset (Table B-3).

-------
Advancing Pb Exposure and Biokinetic Modeling
Table B-1. Average annual Bunker Hill Superfund Site lead concentration in air (|jg/m3)
presented by station
Location*
1995
1996
1997
1998
1999
2000
2001
2002
Bunker Avenue
0.00
0.00
0.01
0.05
—
—
—
—
Kellogg Medical Clinic
0.10
0.07
0.07
0.08
0.04
0.05
0.03
0.03
East Gate
0.08
0.05
0.05
0.04
—
—
—
—
Eastgate Collected
0.09
0.10
0.10
0.03
—
—
—
—
Multiplate
0.00
0.02
0.02
0.04
—
—
—
—
Pinehurst
0.00
0.03
0.03
0.02
—
—
—
—
Smelterville
0.04
0.06
0.06
0.00
—
—
—
—
West Gate
0.03
0.04
0.04
0.02
—
—
—
—
Average
0.07
0.05
0.05
0.04
—
—
—
—
*Data from 1995-1998 for all locations (TerraGraphics. 2001a. 2000): data from 1999-2002 at Kellogg Medical Clinic
(Idaho DEQ. 2005).
Note: — indicates No Data

-------
Advancing Pb Exposure and Biokinetic Modeling
Table B-2. Summary of lead concentration in air (|jg/m3) for Bunker Hill Support Zone
monitoring (1995-1999)
Location**
Lead Data (pg/m3)
Total

Count of Pb
45
Bunker Ave.
Max of Pb
Min of Pb
1.489
0.000

Average of Pb
0.264

Count of Pb
166
East Gate
Max of Pb
3.271
Min of Pb
0.000

Average of Pb
0.334

Count of Pb
165
East Gate Collocated
Max of Pb
Min of Pb
3.167
0.000

Average of Pb
0.355

Count of Pb
54
Multiplate
Max of Pb
Min of Pb
1.322
0.000

Average of Pb
0.195

Count of Pb
46
Pinehurst
Max of Pb
2.799
Min of Pb
0.000

Average of Pb
0.186

Count of Pb
135
Smelterville Gate
Max of Pb
Min of Pb
4.224
0.000

Average of Pb
0.305

Count of Pb
179
West Gate
Max of Pb
Min of Pb
1.564
0.000

Average of Pb
0.150

Total Count of Pb
790
Totals
Total Max of Pb
4.224
Total Min of Pb
0.000

Total Ave. of Pb
0.270
*Data from Table 3.1 in (TerraGraphics, 2000).
"TSP equipment used for monitoring.

-------
Advancing Pb Exposure and Biokinetic Modeling
Table B-3. Number of additional paired records if using floor mat dust data
Year
Number of Paired Observations - Box
Number of Paired Observations - Basin
2003
2

2004
—
12
2005
1
10
2006
3
7
2007
—
6
2008
1
9
2009
—
26
2010
—
17
2011
—
12
2013
27
33
2015
—
19
2017
—
23
2018
20
13
Total
54
187
Note:
— indicates no data

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix C
LHIP Blood Lead Screening Questionnaire
C-1

-------
Public Health
Prevent. Promote. Protect.
Panhandle Health District
Panhandle Health District
Healthy People in Healthy Communities
Institutional Controls Program
35 Wildcat Way, Suite A
Kellogg, ID 83837
Phone: 208-783-0707
Fax: 208-783-4242
www.panhandlehealthdistrict.ora
ID#
SILVER VALLEY LEAD HEALTH INTERVENTION
Box ~ Basin ~
Child Questionnaire
1. How long has this family been living at the current address: (circle one)
1 = Less than 1 month	5 = 6 months to 1 year
2=1 month or more but less than 2 months	6 = More than 1 year but less than 5 years
3	= 2 to 3 months	7 = 5 years or more
4	= More than 3 months but less than 6 months	8 = Do not know or unknown
2. What was the prior residence of this family: Use codes from question #3 to answer "How Long".
City/State	Address	How long?
City/State	Address	How long?
City/State	Address	How long?
3.	What is your telephone number:	HOME/ CELL	/	/	
WORK	/	/	MESSAGE	/	/	
4.	Does your child spend time at another location? Daycare	Relative	No	
Address:	 Amount of time/day	hours.
5.	What is the main source of drinking water for the home?
1	= well water	4 = other	
2	= city water	9 = don't know
3	= bottled water
6.	Has the yard at this residence been remediated? 	No 	Yes Year	Don't Know_
7.	Has the immediate yard or home ever flooded? 	No 	Yes Year	Don't Know_

-------
8. Do you eat vegetables from a local garden?
Type of produce	
Yes
No
How often
9. Do you eat wild game?.
Yes
No How often
10.	Do you own or rent your home?
11.	Occupation: You	
Spouse:	
Rent
Own
12.	Do you plan to do any remodeling or landscaping in the near future? 	
13.	Do you have paint peeling inside your current residence? 	Yes
14.	Do you have paint peeling outside your current residence? 	Yes
15.	Do you have dogs, cats or any other pets that go in and out of the house?
1 = yes 2 = no	9 = don't know
16.	Does anyone smoke cigarettes inside your house? 1 = yes 2 = no
17.	Where (child's name) plays?
1 = Mostly grass,	2 = Mostly dirt or 3 = Other
	 concrete, asphalt,	sand
or wood
1 = Mostly grass,	2 = Mostly dirt or 3 = Other
	 concrete, asphalt,	sand
or wood
1 = Mostly grass,	2 = Mostly dirt or 3 = Other
	 concrete, asphalt,	sand
or wood
1 = Mostly grass,	2 = Mostly dirt or 3 = Other
	 concrete, asphalt,	sand
or wood
1 = Mostly grass,	2 = Mostly dirt or 3 = Other
	 concrete, asphalt,	sand
or wood
_Yes	No
	No
No
9 = don't know
9 = don't know
9 = don't know
9 = don't know
9 = don't know
9 = don't know

-------
18. How many hours a day on the average does (child's name) play outdoors? (99 = don't know)
Name	Hours	Name	Hours
19. Were (child's name) hands almost always, sometimes or almost never washed before bed?
= always
2 = sometimes
3 = almost never
9 = don't know
= always
2 = sometimes
3 = almost never
9 = don't know
= always
2 = sometimes
3 = almost never
9 = don't know
= always
2 = sometimes
3 = almost never
9 = don't know
= always
2 = sometimes
3 = almost never
9 = don't know
20. Do any members of the household (including you) do the following activities? [First check Yes, No, or
Don't Know for each activity. Complete the remaining columns only for activities checked Yes. Be
sure to circle the appropriate unit (# times or # days) for all activities checked Yes.]
Activitv
No
Don't
Know
Yes
# Times or Davs in
the Past Three
Months
Pottery



times / days
Ceramics



times / days
Jewelry Making



times / days
Stained Glass



times / days
Target Shooting



times / days
Bullet
Manufacture



times / days
Lead Soldering



times / days
Auto Repair



times / days

-------
21. Do any members of the household (including you) do the following activities? [First check Yes, No, or
Don't Know for each activity. Complete the remaining columns only for activities checked Yes. Be
sure to circle the appropriate unit (# times or # days) for all activities checked Yes.]
Activitv
No
Don't
Know
Yes
# Times or Davs in
the Past Three
Months
Location(s)
Dirt biking/4-
wheeling



times / days

Mountain biking



times / days

Mudding



times / days

Camping



times / days

Boating



times / days

Swimming



times / days

Hunting



times / days

Fishing



times / days

22. What is your total gross household i
1	= less than $10,000 per year
2	= $10,000 to $14,999 per year
3	= $15,000 to $19,999 per year
4	= $20,000 to $24,999 per year
before taxes?
5	= $25,000 to $29,000 per year
6	= $30,000 to $39,999 per year
7	= more than $40,000 per year
8	= refused	9 = don't know

-------
23. What is the highest year of education that was completed by the head of this household?
No schooling
000







Elementary School
001
002
003
004
005
006
007
008
High School (GED-012)
009
010
011
012




Technical or Trade School
T13
T14






Junior or Community College
J13
J14






4-year College or University
013
014
015
016




Graduate School (or higher)
017







Refused to answer
088







Don't know
099







24. Race (check all that apply)
American Indian


Asian


Black or African American


Native Hawaiian


Pacific Islander


White


Refused to answer


Unknown


25. Ethnicity
Hispanic


Non-Hispanic


Unknown

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix D
Floor Mat Questionnaire
D-1

-------
House Id#
GISID:
Parcel #
2015 Household Questionnaire
[Interviewer: Complete this questionnaire on all households in which a mat is picked up. Let the
participant know that it may take a few minutes to complete this questionnaire.]
Mat Retrieved?	 Date		Vacuum Bag?	
House ID #:
Date:
Interviewer Initials:
Name:
Street Address:
Mailing Address:
Home Phone Number:
(208)
Work Phone Number:
( )
Cell Phone Number:
( )
* Must have both street and mailing address. If mailing address is the same, write "same." Please include zip code.
1.	Did you vacuum the mat?
1) yes	2) no	9) don't know
2.	How many times did you pick up or shake the mat?
1) only once	2) 2-4 times	3) More than 4 times
4) None	9) don't know
3.	Did you move it from its original location to a new location?
1) yes	2) no	9) don't know
4.	Did the mat get wet?
1) yes	2) no	9) don't know
5.	Did the mat get physically damaged by animals or otherwise?
1) yes	2) no	9) don't know
6.	How many days did you go on vacation/ stay away from the house since the mat was
placed? 	 days
7.	How many people regularly live in the home?
Adults	Children	Pregnant Women	
If children live in the home, go to Question 8. If not, go to Question 10.
If children live at the home, what are their ages? [Circle appropriate unit (years or
months) for each child's age.]
Child 1	years / months Child 2	years / months
Child 3	years / months Child 4	years / months
2015DUSTMATQUESTIONAIRE.DOC	A-l

-------
House Id#	GISID:	Parcel #
9.	How many hours per day does the most active or oldest child spend outside?
During the summer:	hours
During the winter: 	hours
10.	How many dogs, cats, or other pets regularly go in and out of the house?
1) 1 animal	2) 2 or more animals
4) none	9) Don't know
11.	Do you have a forced air heating or cooling system in your home (i.e., air ducts)?
1) yes	2) no	9) don't know
12.	How many rooms are in the home (count all regularly accessed rooms including
bedrooms, bathrooms, kitchen, living rooms, family rooms and others)?	
13. How many of those rooms are carpeted?
[Based on Questions 12 and 13, calculate the percentage of the total rooms that is carpeted]
1) <50% of the rooms	2) >50% of the rooms
14.	From the following choices, how old is the oldest carpet in your home?
1) less than 1 year old	2) 1-5 years old	3) 6-10 years old
4) older than 10 years	9) don't know
15.	Are there throw rugs/entrance mats at the entrances to this home?
1) Yes	2) No
16.	Do people generally remove their shoes before entering the home?
1) yes	2) no
17.	What year was this home built? (oldest part) Year:	
1) before 1960	2) 1960 - 1978	3) 1979 or later
9) don't know
18.	Do you own or rent your home?
1) rent	2) own
19.	How long have you lived in this home?
1) <1 year	2) 1-5 years	3) >5 years
20.	Does your home contain lead-based paints?
l)yes	2) no	9) don't know
21.	Has the yard (or ground immediately surrounding this residence) or the inside of this
home been flooded?
1) yes When?	 2) no	9) don't know
22.	Is there a daycare run out of this home?
1) yes	2) no
2015DUSTMATQUESTIONAIRE.DOC
A-2

-------
House Id#
GISID:
Parcel #
23. If children live in the home, describe the condition of children's play areas using the
following condition codes. [Use 9 if not applicable.]
Condition codes:
1	= grassy, vegetated, no bare soil
2	= some bare soil, partly grassy
3	= moderate amount of bare soils, gravel, dust
4	= area is mostly or totally bare soil, garden, gravel, riverbed
9 = N/A
Location
Condition
Yard

Play area

Day care

Neighbors

Vacant lot

Hillsides

Relatives

Other

24. Do any members of the household (including you) do the following activities? [First
check Yes, No, or Don't Know for each activity. Complete the remaining columns only
for activities checked Yes. Be sure to circle the appropriate unit (# times or # days) for all
Activity
No
Don't
Know
Yes
# Times or Days
in the Past
Three Months
Location(s)
Dirt biking/4-wheeling



times / days

Mountain biking



times / days

Mudding



times / days

Camping



times / days

Boating



times / days

Swimming



times / days

Hunting - upland game



times / days

Hunting - waterfowl



times / days

Fishing



times / days

Other (such as
picnicking, geocaching,
hiking, bird/wildlife
viewing, horseback
riding, scenic driving,
and foraging (morels,
huckleberries)



times / days

2015DUSTMATQUESTIONAIRE.DOC
A-3

-------
House Id#
GISID:
Parcel #
[Total number of activities:
J (from chart above).
25. In the last 3 months, has any member of this household (including you) been employed in
the following jobs?
Occupation	Yes	No
Milling or concentrating ore	1	2
Carpentry or remodeling work	1	2
Foundry work	1	2
Professional plumbing/plumber	1	2
Mining	1	2
Landscaping/excavation	1	2
Construction Work in the Silver Valley	1	2
Don't Know
9
9
9
9
9
9
9
Don't Know
9
9
9
26. Within the last 3 months, has any member of this household (including you) done any of
the following activities in this home more than once?
Activity	Yes No
Painted pictures with artist's paints	1 2
Worked with stained glass or made metal jewelry	1 2
Cast lead into fishing sinkers, bullets or anything	1 2
else
Worked with soldering in electronics or plumbing	12 9
Worked in a vegetable or flower garden around the	1 2 9
home
Made pottery	12 9
Made tole paintings	12 9
Painted cars or bicycles	12 9
Reloaded bullets	12 9
[The following blanks are to be filled out by the interviewer upon inspection of the home.]
1. Condition of paint:
Inside:	1) good condition
Outside: 1) good condition
2) chipping, chalking, peeling or bite marks
2) chipping, chalking, peeling or bite marks
2. Rate the grass coverage in yard:
1) mostly soil/dirt
3) mostly grass
2) half bare/half covered
3. Rate general household hygiene
1) poor: a lot of noticeable dust/odor/dirt 2) good
Comments:
2015DUSTMATQUESTIONAIRE.DOC
A-4

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix E
Paired Dataset Summary Tables and Figures
E-1

-------
Table E-1. Box paired blood, yard soil, and vacuum Pb concentrations for LHIP participants ages 0-6 years old, 1988-2018


Number of

Blood Lead Level (ng/dL) *

Vacuum Dust Lead Exposures (mg/kg) t
c
Soil Lead Exposures (mg/kg)
t


Paired
# Below


Arithmetic
Geometric


Arithmetic
Geometric


Arithmetic
Geometric
Year
Community
Observations
Detection
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
1988
Kellogg
Page
28
1
0
0
4.0
28.0
10.3
5.7
8.8
1.8
94.0
17,200
2,286
3,579
1,359
2.6
136
7,510
2,773
1,661
2220
2.2

Smelterville
13
0
10.0
37.0
16.8
7.3
15.6
1.4
209
4,640
1,955
1,648
1,256
3.0
369
10,700
2,639
2,867
1674
2.7

Wardner
3
0
9.0
18.0
13.3
4.5
12.8
1.4
427
1,480
778
608
646
2.0
915
1,930
1,253
586
1173
1.5

Box-wide
45
0
4.0
37.0
12.5
6.7
10.8
1.7
94.0
17,200
2,065
2,964
1,259
2.6
136
10,700
2,591
2,040
1921
2.3
1989
Kellogg
Page
14
2
0
0
4.0
29.0
15.5
6.9
13.9
1.7
552
3,370
1,452
850
1,252
1.8
136
5,320
2,358
1,554
1784
2.5

Smelterville
9
0
7.0
28.0
16.4
6.7
15.2
1.5
209
4,640
1,446
1,320
993
2.7
369
3,150
1,073
925
815
2.1

Wardner
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Box-wide
25
0
4.0
29.0
16.2
6.5
14.8
1.6
209
4,640
1,427
989
1,145
2.1
136
5,320
1,777
1,430
1271
2.5
1990
Kellogg
43
0
4.0
19.0
9.7
4.1
8.8
1.6
117
5,380
1,580
1,143
1,186
2.4
100
6,790
1,750
1,915
622
5.5

Page
5
0
10.0
21.0
15.8
4.3
15.3
1.3
898
2,070
1,221
487
1,159
1.4
100
3,480
1,212
1,380
548
5.0

Pinehurst
23
0
4.0
13.0
7.0
3.0
6.4
1.5
363
7,990
1,678
2,153
1,041
2.4
176
1,530
532
342
450
1.8

Smelterville
10
0
5.0
16.0
8.2
3.0
7.8
1.4
1,000
4,210
2,079
1,269
1,786
1.8
100
6,790
2,795
2,223
1260
5.9

Wardner
4
0
4.0
8.0
6.8
1.9
6.5
1.4
691
2,220
1,073
765
925
1.8
847
13,200
3,935
6,177
1683
3.9

Box-wide
85
0
4.0
21.0
9.0
4.2
8.1
1.6
117
7,990
1,620
1,457
1,186
2.2
100
13,200
1,614
2,147
644
4.4
1991
Kellogg
32
0
4.0
31.0
8.0
6.6
6.4
1.9
274
3,910
1,550
786
1,365
1.7
100
3,220
901
1,133
314
4.6

Page
4
0
4.0
14.0
7.5
4.7
6.5
1.8
545
1,430
1,186
430
1,105
1.6
100
811
278
356
169
2.8

Pinehurst
30
0
4.0
26.0
6.1
4.9
5.2
1.6
65.0
13,500
1,097
2,362
645
2.3
117
3,060
661
714
457
2.3

Smelterville
17
0
4.0
13.0
6.9
3.1
6.3
1.6
790
2,700
1,426
557
1,335
1.4
100
5,460
1,299
1,787
346
5.7

Wardner
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Box-wide
85
0
4.0
31.0
7.0
5.3
5.9
1.7
65.0
13,500
1,326
1,508
1,010
2.1
100
5,460
848
1,160
345
3.8
1992
Kellogg
80
0
4.0
25.0
8.9
5.3
7.6
1.7
104
3,230
1,173
792
909
2.2
100
6,930
1,547
1,825
518
5.3

Page
5
0
4.0
8.0
5.6
1.8
5.4
1.4
473
1,500
792
421
719
1.6
100
708
222
272
148
2.4

Pinehurst
42
0
4.0
13.0
6.3
2.7
5.8
1.5
298
3,470
804
687
620
2.0
94.0
3,060
604
623
437
2.2

Smelterville
17
0
4.0
30.0
9.4
7.0
7.8
1.8
140
3,790
982
817
773
2.0
100
2,350
318
630
141
2.7

Wardner
7
0
4.0
10.0
6.7
2.3
6.4
1.4
322
5,240
1,543
1,731
955
2.9
100
100
100
0.0
100
1.0

Box-wide
151
0
4.0
30.0
8.0
4.9
6.9
1.7
104
5,240
1,053
833
798
2.1
94.0
6,930
1,035
1,489
380
4.2
1993
Kellogg
77
0
1.0
24.0
7.2
4.9
5.8
2.0
111
2,350
946
538
789
1.9
100
4,300
772
1,263
246
4.1

Page
4
0
5.0
12.0
7.8
3.1
7.3
1.5
139
794
527
282
440
2.2
100
956
314
428
176
3.1

Pinehurst
40
0
1.0
13.0
3.7
3.1
2.7
2.3
111
3,460
732
859
485
2.4
100
3,060
688
813
411
2.7

Smelterville
13
0
2.0
11.0
6.3
2.8
5.6
1.7
201
3,350
1,044
781
827
2.1
100
4,250
958
1,415
279
5.0

Wardner
4
0
3.0
6.0
4.8
1.3
4.6
1.3
588
1,290
1,025
303
983
1.4
100
1,850
1,183
754
773
3.9

Box-wide
138
0
1.0
24.0
6.0
4.4
4.6
2.2
111
3,460
883
666
681
2.1
100
4,300
764
1,132
296
3.8
1994
Kellogg
75
0
1.0
41.0
7.0
5.4
5.8
1.8
88.0
2,160
771
497
607
2.1
100
4,884
737
1,431
202
3.9

Page
3
0
5.0
12.0
8.3
3.5
7.8
1.5
90.0
605
366
259
280
2.7
499
999
721
255
692
1.4

Pinehurst
28
0
1.0
10.0
4.9
2.4
4.3
1.7
95.0
1,490
498
339
409
1.9
79.0
931
285
229
221
2.0

Smelterville
25
0
2.0
13.0
5.9
3.2
5.3
1.6
228
2,310
1,017
714
779
2.2
100
6,970
728
1,672
177
3.9

Wardner
7
0
2.0
11.0
5.3
3.1
4.5
1.9
211
2,270
1,049
907
708
2.7
100
2,020
374
726
154
3.1

Box-wide
138
0
1.0
41.0
6.3
4.5
5.3
1.8
88.0
2,310
766
562
581
2.2
79.0
6,970
625
1,291
203
3.4

-------
Table E-1. Box paired blood, yard soil, and vacuum Pb concentrations for LHIP participants ages 0-6 years old, 1988-2018

Number of

Blood Lead Level (ng/dL) *

Vacuum Dust Lead Exposures (mg/kg) t
c
Soil Lead Exposures (mg/kg)
t

Paired
# Below


Arithmetic
Geometric


Arithmetic
Geometric


Arithmetic
Geometric
Year
Community
Observations
Detection
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
1995
Kellogg
60
0
1.0
16.0
5.6
3.4
4.7
1.8
123
4,400
959
940
676
2.3
100
9,180
916
1,697
282
4.1

Page
3
0
4.0
10.0
7.3
3.1
6.8
1.6
239
1,430
791
600
622
2.5
100
499
233
230
171
2.5

Pinehurst
19
0
1.0
12.0
4.8
2.3
4.3
1.7
22.0
1,720
447
434
280
3.1
100
2,670
348
577
214
2.3

Smelterville
15
0
3.0
15.0
7.5
4.1
6.6
1.7
297
3,470
1,165
1,217
782
2.4
100
7,370
1,321
2,254
275
5.7

Wardner
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Box-wide
99
0
1.0
16.0
5.8
3.4
4.9
1.8
22.0
4,400
873
920
570
2.6
100
9,180
831
1,622
257
3.9
1996
Kellogg
71
0
1.0
18.0
6.4
4.1
5.2
1.9
85.0
1,700
653
352
549
1.9
100
5,370
720
1,307
223
3.9

Page
Pinehurst
2
21
0
0
2.0
7.0
4.0
1.8
3.7
1.6
100
2,100
468
415
367
2.0
100
1,210
438
378
257
2.7

Smelterville
5
0
4.0
8.0
5.6
1.8
5.4
1.4
320
11,300
2,732
4,793
1,019
4.0
100
468
174
165
136
2.0

Wardner
3
0
7.0
15.0
11.0
4.0
10.5
1.5
130
890
637
439
469
3.0
614
2,568
1,917
1,128
870
6.5

Box-wide
102
0
1.0
18.0
5.9
3.8
4.9
1.9
85.0
11,300
706
1,121
505
2.1
100
5,370
658
1,147
230
3.6
1997
Kellogg
35
0
1.0
22.0
6.1
4.1
4.9
2.0
110
6,800
907
1,155
601
2.4
100
3,800
327
713
145
2.6

Page
Pinehurst
2
16
0
0
2.0
11.0
4.6
2.7
3.9
1.7
140
15,000
1,251
3,673
371
3.0
100
940
320
275
236
2.2

Smelterville
10
0
3.0
5.0
4.1
0.7
4.0
1.2
110
515
291
142
259
1.7
100
307
140
84.4
125
1.6

Wardner
6
0
2.0
10.0
6.2
3.3
5.3
1.9
220
1,100
668
473
509
2.3
100
100
100
0.0
100
1.0

Box-wide
69
0
1.0
22.0
5.5
3.4
4.6
1.8
110
15,000
864
1,938
463
2.5
100
3,800
286
531
160
2.4
1998
Kellogg
50
0
1.0
15.0
4.6
2.6
3.9
1.9
140
3,600
815
701
630
2.0
100
3,860
267
590
138
2.3

Page
3
0
3.0
5.0
4.0
1.0
3.9
1.3
550
1,500
927
505
844
1.7
572
1,650
962
598
856
1.8

Pinehurst
23
0
1.0
17.0
4.4
3.3
3.7
1.8
71.0
2,000
395
415
290
2.1
100
874
404
209
354
1.8

Smelterville
17
0
4.0
20.0
8.2
3.7
7.6
1.5
480
1,100
655
181
636
1.3
100
588
157
137
128
1.8

Wardner
7
0
1.0
13.0
5.1
4.0
3.9
2.3
270
6,000
2,004
2,735
826
4.0
100
1,040
234
355
140
2.4

Box-wide
100
0
1.0
20.0
5.2
3.3
4.3
1.9
71.0
6,000
778
943
543
2.2
100
3,860
299
469
179
2.4
1999
Kellogg
63
0
1.0
14.0
4.9
2.9
4.1
1.8
199
15,300
1,090
2,621
597
2.2
100
5,363
397
967
153
2.8

Page
Pinehurst
2
38
0
0
1.0
17.0
5.4
3.5
4.5
1.8
45.0
4,010
441
619
312
2.2
100
1,540
404
294
319
2.0

Smelterville
10
0
2.0
8.0
3.8
1.8
3.5
1.5
259
823
401
194
368
1.5
100
550
257
193
197
2.2

Wardner
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Box-wide
115
0
1.0
17.0
4.9
3.0
4.2
1.8
45.0
15,300
803
1,998
457
2.3
100
5,363
389
737
206
2.6
2000
Kellogg
49
0
1.0
16.0
5.0
2.6
4.4
1.7
49.0
11,200
934
2,164
446
2.7
100
4,500
263
700
125
2.2

Page
Pinehurst
2
22
0
0
1.0
15.0
4.2
3.5
3.2
2.1
150
2,300
595
640
402
2.3
100
1,540
520
369
394
2.3

Smelterville
16
0
1.0
11.0
4.9
3.4
3.8
2.1
150
820
403
174
371
1.5
100
467
169
148
134
1.9

Wardner
1
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Box-wide
90
0
1.0
16.0
4.8
2.9
4.0
1.8
49.0
11,200
744
1,640
411
2.5
100
4,500
320
564
176
2.5
2001
Kellogg
51
0
1.4
17.0
3.9
3.0
3.2
1.8
64.0
1,900
449
386
346
2.0
100
3,560
299
611
143
2.5

Page
Pinehurst
2
22
0
0
1.4
11.0
3.3
2.2
2.9
1.7
57.0
1,200
346
326
249
2.2
34.0
517
235
182
173
2.3

Smelterville
3
0
1.4
5.4
3.0
2.1
2.6
2.0
220
420
293
110
281
1.4
100
393
198
169
158
2.2

Wardner
3
0
1.4
6.8
4.3
2.7
3.5
2.3
180
960
670
427
532
2.6
100
727
309
362
194
3.1

Box-wide
81
0
1.4
17.0
3.9
2.9
3.2
1.8
57.0
1,900
415
365
310
2.1
34.0
3,560
282
497
157
2.4

-------
Table E-1. Box paired blood, yard soil, and vacuum Pb concentrations for LHIP participants ages 0-6 years old, 1988-2018


Number of

Blood Lead Level (ug/dL) *

Vacuum Dust Lead Exposures (mg/kg) t
c
Soil Lead Exposures (mg/kg)
t


Paired
# Below


Arithmetic
Geometric


Arithmetic
Geometric


Arithmetic
Geometric
Year
Community
Observations
Detection
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
2002
Kellogg
35
0
1.4
7.7
2.9
1.4
2.7
1.5
62.0
3,500
615
810
385
2.5
100
3,560
347
851
135
2.6

Page
3
0
3.3
5.1
4.5
1.0
4.4
1.3
250
270
263
11.5
263
1.0
425
1,160
670
424
594
1.8

Pinehurst
26
1
1.4
5.5
2.3
1.0
2.1
1.5
51.0
1,200
206
233
151
2.1
31.1
687
225
170
168
2.3

Smelterville
13
0
1.5
6.3
3.4
1.5
3.1
1.6
54.0
535
332
193
250
2.5
100
467
128
102
113
1.5

Wardner
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Box-wide
79
1
1.4
7.7
2.8
1.4
2.6
1.5
51.0
3,500
418
587
261
2.5
31.1
3,560
277
587
148
2.4
2003
Box-wide
3
0
3.3
10.8
8.0
4.1
7.1
1.9
306
3,890
1,556
2,023
825
3.9
100
100
100
NA
100
NA
2005
Box-wide
2
1
—
—
— —
— —
—
—
— —
— —
—
—
— —
— —
2006
Box-wide
2
0
—
—
— —
— —
—
—
— —
— —
—
—
— —
— —
2008
Box-wide
4
0
1.5
4.8
2.8
1.4
2.5
1.6
155
336
227
88.3
214
1.5
100
100
100
NA
100
NA
2013
Kellogg
72
17
1.4
20.0
2.9
2.6
2.4
1.7
17.1
1,310
348
255
262
2.3
100
698
134
130
113
1.6

Pinehurst
37
15
1.4
6.0
2.4
1.2
2.1
1.6
20.5
1,940
387
410
244
2.8
100
1,270
302
292
205
2.4

Smelterville
30
7
1.4
6.0
2.6
1.1
2.4
1.5
67.0
859
276
189
230
1.8
100
608
163
142
131
1.8

Wardner
7
0
1.5
4.6
2.7
1.1
2.6
1.5
268
454
322
89.9
313
1.3
100
100
100
NA
100
NA

Box-wide
146
39
1.4
20.0
2.7
2.0
2.3
1.6
17.1
1,940
342
287
253
2.3
100
1,270
181
197
135
1.9
2018
Kellogg
18
5
1.9
9.0
2.9
1.7
2.6
1.5
37.6
757
256
246
179
2.3
100
100
100
NA
100
NA

Pinehurst
14
10
1.9
2.8
2.0
0.3
2.0
1.1
90.4
523
213
176
164
2.0
100
785
149
183
116
1.7

Smelterville
8
1
1.9
10.0
6.0
3.0
5.3
1.8
177
333
242
60.6
236
1.3
100
689
247
273
162
2.4

Wardner
3
3
1.9
1.9
1.9
NA
1.9
NA
54.5
105
71.3
29.2
67.8
1.5
242
979
488
426
386
2.2

Box-wide
43
19
1.9
10.0
3.1
2.2
2.7
1.6
37.6
757
227
192
171
2.1
100
979
170
205
126
1.9
Notes:
*When a PbB result was below detection, the detection limit was used for data summary purposes (1.0, 1.4, or 1.9 ug/dL).
+When a vacuum dust result was below detection, a value of half the detection limit was used for data summary purposes. If more than one vacuum sample was collected from a home in the same year, the higher of the two results
was used.
JSoil lead exposure for the Box is the yard lead concentration for each property, with an assumed soil concentration of 100 mg/kg lead for remediated properties.
Min = minimum
Max = maximum
S.D. = Standard Deviation
— No paired observations are available for this community or geographic area in this year.
NA = not applicable

-------
Table E-2. Box - age (in years) of LHIP participants with paired soil and dust data, 1988-2018
Year
Number of Paired
Observations
<1 Year
1 to <3 Years
3 to <5 Years
5 to <7 Years
1988
45
1
12
12
20
1989
25
1
10
7
7
1990
85
3
22
30
30
1991
85
2
29
28
26
1992
151
4
56
46
45
1993
138
6
36
49
47
1994
138
6
41
44
47
1995
99
3
26
34
36
1996
102
3
26
29
44
1997
69
2
23
27
17
1998
100
2
38
28
32
1999
115
10
32
36
37
2000
90
2
28
24
36
2001
81
8
25
24
24
2002
79
7
15
31
26
2003
3
0
2
1
0
2005
2
0
0
2
0
2006
2
0
1
0
1
2008
4
0
2
1
1
2013
146
11
44
49
42
2018
43
4
13
9
17

-------
Table E-3a. Basin paired blood, soil, and vacuum Pb concentrations for LHIP participants ages 0-6 years old, 1996-2018


Number of

Blood Lead Level (ng/dL) *


Vacuum Dust Lead Exposures (mg/kg) t

Soil Lead Exposures (mg/kg)
t


Paired
# Below


Arithmetic
Geometric


Arithmetic
Geometric


Arithmetic
Geometric
Year
Community
Observations
Detection
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
1996
Burke/Ninemile
Kingston
Lower Basin
Mullan
Osburn
Side Gulches
Silverton
Wallace
1
2
1
0
0
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
4
0
2.0
7.0
4.0
2.2
3.6
1.7
578
1181
893
247
865
1.3
978
1278
1135
165
1126
1.2
1998
Burke/Ninemile
Kingston
Lower Basin
Mullan
Osburn
Side Gulches
Silverton
Wallace
1
2
1
2
2
0
0
0
0
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
8
0
4.0
16.0
7.8
4.4
6.9
1.7
154
1000
716
339
610
2.0
101
1636
946
501
759
2.4
1999
Burke/Ninemile
4
0
3.0
18.0
8.8
6.5
7.1
2.1
773
1070
875
140
867
1.2
106
36000
9147
17902
687
14.4

Kingston
11
0
1.0
16.0
6.1
4.7
4.7
2.2
102
1750
461
472
323
2.4
54.2
1255
412
383
259
3.0

Lower Basin
5
0
6.0
18.0
13.2
4.6
12.4
1.5
68.0
72.0
70.4
2.2
70.4
1.0
29.7
30.9
30.4
0.7
30.4
1.0

Mullan
5
0
3.0
12.0
7.6
3.8
6.8
1.8
632
4060
1557
1433
1194
2.1
547
2153
1428
734
1254
1.8

Osburn
16
0
1.0
11.0
4.6
2.7
3.9
1.9
82.0
1340
310
322
216
2.3
199
1832
639
432
515
2.0

Side Gulches
6
0
2.0
5.0
3.5
1.0
3.4
1.4
162
507
366
162
328
1.7
36.2
290
145
117
105
2.5

Silverton
9
0
2.0
15.0
6.2
4.4
5.1
1.9
75.0
3390
1132
1324
551
4.0
128
1436
460
443
323
2.4

Wallace
4
0
4.0
8.0
5.5
1.7
5.3
1.3
83.0
681
382
345
238
3.4
1110
1112
1111
1.0
1111
1.0

Basin-wide
60
0
1.0
18.0
6.3
4.4
5.1
2.0
68.0
4060
593
797
323
3.0
29.7
36000
1135
4607
330
3.9
2002
Burke/Ninemile
Kingston
Lower Basin
Mullan
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Osburn
4
0
2.2
13.0
6.0
4.8
4.8
2.1
283
662
490
156
468
1.4
1514
6814
4440
2778
3683
2.1

Side Gulches
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Silverton
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Wallace
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
6
0
2.2
13.0
5.6
4.0
4.7
1.9
283
662
482
163
456
1.5
272
6814
3292
2829
2102
3.3

-------
Table E-3a. Basin paired blood, soil, and vacuum Pb concentrations for LHIP participants ages 0-6 years old, 1996-2018


Number of

Blood Lead Level (ng/dL) *


Vacuum Dust Lead Exposures (mg/kg) t

Soil Lead Exposures (mg/kg)
t


Paired
# Below


Arithmetic
Geometric


Arithmetic
Geometric


Arithmetic
Geometric
Year
Community
Observations
Detection
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
2003
Burke/Ninemile
Kingston
Lower Basin
Mullan
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Osburn
3
0
2.0
5.7
3.8
1.9
3.5
1.7
170
170
170
NA
170
NA
247
247
247
NA
247
NA

Side Gulches
1
1
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Silverton

—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Wallace

—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
4
1
1.4
5.7
3.2
1.9
2.8
1.9
170
224
184
27.0
182
1.1
247
642
345
198
313
1.6
2004
Burke/Ninemile
Kingston
Lower Basin
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Mullan
3
0
3.5
5.0
4.3
0.8
4.3
1.2
945
945
945
NA
945
NA
2779
2779
2779
NA
2779
NA

Osburn
1
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Side Gulches
4
1
1.4
5.6
2.8
1.9
2.4
1.8
105
273
231
84.0
215
1.6
97
1128
870
515
611
3.4

Silverton

—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Wallace

—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
10
1
1.4
6.1
3.8
1.6
3.4
1.6
105
945
486
326
393
2.0
97
2779
1366
1026
963
2.8
2005
Burke/Ninemile
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Kingston
4
1
1.4
5.2
3.5
1.6
3.1
1.8
211
1350
496
570
336
2.5
100
3549
962
1724
244
6.0

Lower Basin
2
1
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Mullan
2
2
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Osburn
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Side Gulches
3
2
1.4
1.4
1.4
NA
1.4
NA
37.1
307
127
156
75.0
3.4
100
110
107
NA
107
NA

Silverton
1
1
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Wallace
2
1
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
14
8
1.4
12.0
2.9
2.9
2.2
2.0
37.1
1350
314
330
214
2.5
100
3549
475
922
204
3.1
2006
Burke/Ninemile
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Kingston
6
4
1.4
8.0
2.8
2.6
2.1
2.0
166
869
417
352
312
2.3
437
3549
1565
1551
1006
2.8

Lower Basin
8
4
1.4
3.5
2.0
0.8
1.8
1.4
48.4
300
160
122
116
2.4
52.7
1611
277
539
120
3.0

Mullan
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Osburn
8
0
2.0
7.2
3.3
1.7
3.0
1.5
189
1010
510
344
408
2.1
100
1623
804
536
556
3.0

Side Gulches
7
1
1.4
5.5
3.2
1.4
2.9
1.6
53.0
1370
578
575
302
4.0
110
1421
736
539
503
3.0

Silverton
4
2
1.4
6.0
2.8
2.2
2.3
2.0
131
368
277
102
259
1.6
100
193
153
46.9
148
1.4

Wallace
5
2
1.4
6.7
2.8
2.2
2.3
1.9
332
2500
785
959
525
2.4
100
1176
827
440
628
2.8

Basin-wide
38
13
1.4
8.0
2.8
1.8
2.4
1.7
48.4
2500
446
486
280
2.7
52.7
3549
735
836
384
3.4

-------
Table E-3a. Basin paired blood, soil, and vacuum Pb concentrations for LHIP participants ages 0-6 years old, 1996-2018

Number of

Blood Lead Level (ng/dL) *


Vacuum Dust Lead Exposures (mg/kg) t

Soil Lead Exposures (mg/kg)
t

Paired
# Below


Arithmetic
Geometric


Arithmetic
Geometric


Arithmetic
Geometric
Year
Community
Observations
Detection
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
2007
Burke/Ninemile
1
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Kingston
6
2
1.4
2.6
1.8
0.5
1.7
1.3
77.0
424
169
129
140
1.9
93.7
416
264
167
213
2.1

Lower Basin
4
1
1.4
4.7
2.6
1.5
2.3
1.7
52.8
240
160
94.9
135
2.1
44
111
89
30.2
84
1.5

Mullan
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Osburn
8
0
2.1
9.0
4.4
2.1
4.0
1.5
213
332
269
53.0
264
1.2
247
1069
631
395
518
2.0

Side Gulches
2
1
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Silverton
3
1
1.4
7.1
3.7
3.0
2.9
2.3
136
222
165
49.7
160
1.3
100
1662
1142
902
651
5.1

Wallace
6
0
1.5
6.2
3.9
2.1
3.3
1.9
30.5
301
193
128
131
3.1
100
3844
1707
1724
768
5.3

Basin-wide
30
5
1.4
9.0
3.2
2.0
2.7
1.8
30.5
424
192
103
158
2.0
44.4
3844
705
981
333
3.4
2008
Burke/Ninemile
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Kingston
15
8
1.4
3.3
1.8
0.6
1.7
1.3
115
409
222
93.5
205
1.5
100
615
366
176
320
1.8

Lower Basin
3
1
1.4
4.9
2.6
2.0
2.2
2.0
71.5
387
190
172
145
2.4
50.3
111
87.2
32.4
82.4
1.5

Mullan
3
1
1.4
2.0
1.7
0.3
1.6
1.2
51.8
51.8
51.8
NA
51.8
NA
155
155
155
NA
155
NA

Osburn
5
3
1.4
3.2
1.9
0.8
1.8
1.5
215
394
299
87.3
290
1.3
100
330
221
116
193
1.8

Side Gulches
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Silverton
4
0
2.4
3.2
2.8
0.4
2.8
1.1
446
1170
627
362
568
1.6
128
233
154
52.7
149
1.4

Wallace
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
34
13
1.4
4.9
2.0
0.8
1.9
1.4
51.8
1170
273
200
219
2.0
50.3
616
260
166
212
1.9
2009
Burke/Ninemile
5
1
1.4
3.1
2.4
0.7
2.4
1.4
112
544
219
187
175
2.0
100
158
113
25.2
111
1.2

Kingston
5
2
1.4
10.0
4.3
3.5
3.3
2.3
156
223
196
36.7
193
1.2
200
259
224
32.2
222
1.2

Lower Basin
3
0
2.7
3.1
2.9
0.2
2.9
1.1
86.5
157
134
40.7
129
1.4
66.6
80.5
71.3
8.0
71.0
1.1

Mullan
4
0
1.6
4.0
2.8
1.0
2.7
1.5
224
492
314
121
299
1.4
210
385
258
85.3
249
1.3

Osburn
31
5
1.4
7.2
2.7
1.2
2.5
1.5
11.3
3860
378
672
202
3.2
100
1094
374
270
293
2.0

Side Gulches
10
3
1.4
5.3
3.3
1.5
2.9
1.7
198
422
251
72.9
243
1.3
100
480
350
146
307
1.8

Silverton
6
0
2.2
5.1
3.0
1.1
2.8
1.4
34.3
269
129
88.4
105
2.1
100
555
268
167
224
2.0

Wallace
4
0
2.2
3.9
3.2
0.7
3.2
1.3
260
484
373
128
356
1.4
100
648
396
293
298
2.5

Basin-wide
68
11
1.4
10.0
2.9
1.4
2.7
1.5
11.3
3860
297
464
200
2.4
66.6
1094
312
225
245
2.0
2010
Burke/Ninemile
4
2
1.4
1.8
1.5
0.2
1.5
1.1
116
511
215
198
168
2.1
102
263
143
80.2
130
1.6

Kingston
5
1
1.4
2.2
1.9
0.3
1.8
1.2
144
149
147
2.7
147
1.0
181
193
186
6.4
186
1.0

Lower Basin
4
0
1.6
2.0
1.8
0.2
1.7
1.1
46.1
69.1
57.6
13.3
56.4
1.3
171
320
246
86.0
234
1.4

Mullan
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Osburn
15
2
1.4
4.1
2.0
0.8
1.9
1.4
55.1
349
181
117
146
2.0
100
739
439
182
384
1.9

Side Gulches
10
3
1.4
6.9
3.0
1.7
2.6
1.7
61.6
552
347
179
293
2.0
110
648
287
201
238
1.8

Silverton
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Wallace
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
40
8
1.4
6.9
2.2
1.1
2.0
1.5
46.1
853
232
187
172
2.2
100
739
303
191
248
1.9

-------
Table E-3a. Basin paired blood, soil, and vacuum Pb concentrations for LHIP participants ages 0-6 years old, 1996-2018


Number of

Blood Lead Level (ng/dL) *


Vacuum Dust Lead Exposures (mg/kg) t

Soil Lead Exposures (mg/kg)
t


Paired
# Below


Arithmetic
Geometric


Arithmetic
Geometric


Arithmetic
Geometric
Year
Community
Observations
Detection
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
Min.
Max.
Mean
S. D.
Mean
S. D.
2011
Burke/Ninemile
Kingston
Lower Basin
Mullan
3
2
1
0
0
1
4.0
4.4
4.2
0.2
4.2
1.1
160
461
360
173
324
1.8
141
422
328
162
293
1.9

Osburn
4
1
1.4
4.8
3.1
1.4
2.8
1.7
72.6
318
221
120
189
2.0
100
585
342
280
242
2.8

Side Gulches
1
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Silverton

—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Wallace

—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
11
2
1.4
6.5
3.4
1.6
3.1
1.7
72.6
461
250
124
222
1.7
66.3
797
437
296
313
2.6
2013
Burke/Ninemile
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Kingston
7
0
2.7
10.0
4.9
2.6
4.5
1.6
237
284
257
25.1
256
1.1
107
173
135
35.2
131
1.3

Lower Basin
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Mullan
1
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Osburn
12
1
1.4
4.0
2.5
0.7
2.4
1.3
88.4
345
183
88.5
165
1.6
100
739
271
168
234
1.8

Side Gulches
6
2
1.4
2.7
2.0
0.5
1.9
1.3
40.3
524
268
206
195
2.6
103
495
335
156
296
1.8

Silverton
5
1
1.4
2.5
2.1
0.5
2.1
1.3
80.1
4540
1913
2400
532
7.7
187
235
216
25.9
214
1.1

Wallace
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
33
4
1.4
10.0
2.8
1.7
2.6
1.5
40.3
4540
513
1052
248
2.7
100
739
233
140
201
1.7
2015
Burke/Ninemile
1
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Kingston
4
0
2.8
4.1
3.3
0.6
3.2
1.2
445
2760
1603
1337
1108
2.9
103
182
142
45.9
137
1.4

Lower Basin
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Mullan
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Osburn
4
0
1.0
4.1
2.6
1.3
2.3
1.8
97.0
6000
3052
3404
787
10.4
100
739
342
276
268
2.3

Side Gulches
12
1
1.4
10.0
2.8
2.5
2.3
1.8
52.0
275
164
82.8
143
1.8
110
209
187
37.5
182
1.3

Silverton
6
1
1.4
4.0
2.6
1.0
2.4
1.5
153
236
175
35.6
172
1.2
155
439
318
91.2
305
1.4

Wallace
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
29
2
1.0
10.0
2.9
1.7
2.5
1.6
52.0
6000
838
1591
295
3.6
100
739
231
135
203
1.7
2017
Burke/Ninemile
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Kingston
5
0
3.2
14.0
6.1
4.5
5.2
1.8
177
177
177
NA
177
NA
113
113
113
NA
113
NA

Lower Basin
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Mullan
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Osburn
7
0
2.5
8.0
3.5
2.0
3.2
1.5
73.9
263
200
82.6
180
1.7
126
444
288
117
267
1.5

Side Gulches
1
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Silverton
1
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Wallace
2
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—

Basin-wide
16
0
2.5
14.0
5.0
3.6
4.2
1.7
25.0
267
188
71.6
166
1.9
81.5
444
193
116
168
1.7

-------
Table E-3a. Basin paired blood, soil, and vacuum Pb concentrations for LHIP participants ages 0-6 years old, 1996-2018


Number of
Blood Lead Level (ng/dL) *
Vacuum Dust Lead Exposures (mg/kg) t
Soil Lead Exposures (mg/kg)1


Paired
# Below

Arithmetic
Geometric

Arithmetic
Geometric


Arithmetic
Geometric
Year
Community
Observations
Detection
Min. Max.
Mean S. D.
Mean S. D.
Min. Max.
Mean S. D.
Mean S. D.
Min.
Max.
Mean S. D.
Mean S. D.
2018
Burke/Ninemile
Kingston
Lower Basin
Mullan
2
2
— —
— —
— —
— —
— —
— —
—
—
— —
— —

Osburn
9
6
1.9 6.0
2.4 1.3
2.3 1.5
187 1880
632 574
459 2.3
110
440
252 119
224 1.7

Side Gulches
2
0
— —
— —
— —
— —
— —
— —
—
—
— —
— —

Silverton
—
—
— —
— —
— —
— —
— —
— —
—
—
— —
— —

Wallace
2
1
— —
— —
— —
— —
— —
— —
—
—
— —
— —

Basin-wide
15
9
1.9 8.0
3.0 2.0
2.6 1.7
168 32500
4825 11246
746 5.5
66.9
440
196 118
164 1.9
Notes:
*When a PbB result was below detection, the detection limit was used for data summary purposes (1.0, 1.4, and 1.9 ug/dL).
t When a vacuum dust result was below detection, a value of half the detection limit was used for data summary purposes. If more than one vacuum sample was collected from a home in the same year, the higher of the two results was
used.
tSoil Pb exposure for the Basin was calculated using a weighted arithmetic average lead concentration for each property based on sample areas, with an assumed soil concentration of 100 mg/kg lead for remediated areas.
§Number of water samples that paired with blood, dust, and soil data'
Min = minimum
Max = maximum
S.D. = Standard Deviation
— No paired observations are available for this community or geographic area in this year.
NA - not applicable

-------
Table E-3b. Basin paired blood and drinking water Pb concentrations for LHIP participants ages 0-6 years old, 1996-2018
Geographic
Area
Number of
Paired
Observations
Blood Lead Level (ia.g/dL) *
# Below	Arithmetic Geometric
Detection Min. Max. Mean S. D. Mean S. D.
Tap Water Lead Exposures (|jg/L)	Well Water Lead Exposures (ng/L)
# Below	# Below
Detection Arithmetic	Geometric Detection Arithmetic Geometric
Initial Purged Min. Max. Mean S. D.	Mean S. D. N t Initial Purged Min. Max. Mean S. D. Mean S. D.
Notes:
*When a PbB result was below detection, the detection limit was used for data summary purposes. If the detection limit was 1.9 ug/dL, a value of 1.4 ug/dL was used.
tNumber of property-specific drinking water samples that paired with blood, dust, and soil data.
Min = minimum
Max = maximum
S.D. = Standard Deviation
— No paired observations are available for this geographic area in this year.
NA - not applicable
#= number

-------
Table E-4. Basin - age (in years) of LHIP participants with paired soil and dust
data, 1996-2018
Year
Number of
Paired
Observations
<1 Year*
1 to <3
Years
3 to <5
Years
5 to <7
Years
1996
4

2
0
2
1997
—
—
—
—
—
1998
8

4
2
2
1999
60

21
17
22
2000
—
—
—
—
—
2001
—
—
—
—
—
2002
6
—
1
2
3
2003
4
—
2
1
1
2004
10
3
4
1
2
2005
14
1
6
3
4
2006
38
5
11
13
9
2007
30
3
9
9
9
2008
34
1
9
10
14
2009
68
5
18
23
22
2010
40
2
10
15
13
2011
11
0
4
4
3
2012
—
—
—
—
—
2013
33
3
9
6
15
2014
—
—
—
—
—
2015
29
2
8
12
7
2016
—
—
—
—
—
2017
16
2
8
3
3
2018
15
1
4
7
3
* Children less than 1 year of age are included in the 1 to <3 year category for years 1996 through 1999.

-------
Table E-5. Box - community soil lead concentrations (mg/kg) by year



Arithmetic

Geometric



Standard

Standard
Year
Community
Mean
Deviation
Mean
Deviation

Kellogg
1965
2474
856
4.0

Page
959
1216
485
4.0
1995
Pinehurst
566
516
415
2.0

Smelterville
1759
2736
540
5.0

Wardner
1807
3205
786
4.0

Kellogg
1890
2463
786
5.0

Page
940
1215
470
4.0
1996
Pinehurst
563
516
412
2.0

Smelterville
381
1161
155
3.0

Wardner
1807
3205
786
4.0

Kellogg
1538
2264
550
5.0

Page
940
1215
470
4.0
1997
Pinehurst
561
517
409
2.0

Smelterville
177
256
131
2.0

Wardner
1797
3206
776
4.0

Kellogg
1210
2205
360
5.0

Page
940
1215
470
4.0
1998
Pinehurst
556
510
406
2.0

Smelterville
177
256
131
2.0

Wardner
1789
3209
760
4.0

Kellogg
881
2002
246
4.0

Page
907
1221
437
4.0
1999
Pinehurst
552
509
402
2.0

Smelterville
159
146
127
2.0

Wardner
1789
3209
760
4.0

Kellogg
824
1889
235
4.0

Page
907
1221
437
4.0
2000
Pinehurst
506
482
364
2.0

Smelterville
159
146
127
2.0

Wardner
1751
3218
707
4.0

Kellogg
800
1870
230
4.0

Page
907
1221
437
4.0
2001
Pinehurst
449
451
320
2.0

Smelterville
159
146
127
2.0

Wardner
1751
3218
707
4.0

Kellogg
784
1861
225
4.0

Page
907
1221
437
4.0
2002
Pinehurst
404
384
293
2.0

Smelterville
159
146
127
2.0

Wardner
1629
3072
651
4.0

-------
Table E-5. Box - community soil lead concentrations (mg/kg) by year



Arithmetic

Geometric



Standard

Standard
Year
Community
Mean
Deviation
Mean
Deviation

Kellogg
462
1073
173
3.0

Page
907
1221
437
4.0
2003
Pinehurst
393
333
288
2.0

Smelterville
159
146
127
2.0

Wardner
1581
3071
612
4.0

Kellogg
272
710
139
2.0

Page
907
1221
437
4.0
2004
Pinehurst
393
333
288
2.0

Smelterville
159
146
127
2.0

Wardner
1045
1525
433
4.0

Kellogg
205
453
128
2.0

Page
862
1222
402
4.0
2005
Pinehurst
388
306
287
2.0

Smelterville
159
146
127
2.0

Wardner
213
306
139
2.0

Kellogg
171
366
121
2.0

Page
309
499
182
3.0
2006
Pinehurst
372
274
278
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
164
345
120
2.0

Page
265
264
175
2.0
2007
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
163
341
119
2.0

Page
265
264
175
2.0
2008
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
161
331
119
2.0

Page
265
264
175
2.0
2009
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
161
331
119
2.0

Page
265
264
175
2.0
2010
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

-------
Table E-5. Box - community soil lead concentrations (mg/kg) by year



Arithmetic

Geometric



Standard

Standard
Year
Community
Mean
Deviation
Mean
Deviation

Kellogg
161
331
119
2.0

Page
265
264
175
2.0
2011
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
160
329
119
2.0

Page
265
264
175
2.0
2012
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
152
194
118
2.0

Page
265
264
175
2.0
2013
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
151
190
118
2.0

Page
265
264
175
2.0
2014
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
151
190
118
2.0

Page
265
264
175
2.0
2015
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
151
190
118
2.0

Page
265
264
175
2.0
2016
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
149
186
118
2.0

Page
265
264
175
2.0
2017
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

Kellogg
149
186
118
2.0

Page
265
264
175
2.0
2018
Pinehurst
349
245
263
2.0

Smelterville
159
146
127
2.0

Wardner
199
274
135
2.0

-------
Table E-6. Basin - community soil lead concentration (mg/kg) by year



Arithmetic

Geometric



Standard

Standard
Year
Community
Mean
Deviation
Mean
Deviation

Burke/Ninemile
2605
6108
958
4.1

Kingston
501
1436
170
3.9

Lower Basin
227
681
63
3.7
or\r\o
Mullan
1616
3258
764
3.4

Osburn
1181
2132
589
3.0

Side Gulches
905
2642
392
3.7

Silverton
956
1906
465
3.3

Wallace
1441
1741
893
2.9

Burke/Ninemile
2605
6108
958
4.1

Kingston
501
1436
170
3.9

Lower Basin
227
681
63
3.7

Mullan
1613
3258
760
3.4
ZUUj
Osburn
1156
2097
574
3.0

Side Gulches
902
2641
390
3.7

Silverton
956
1906
465
3.3

Wallace
1430
1744
877
2.9

Burke/Ninemile
2604
6108
956
4.1

Kingston
501
1436
170
3.9

Lower Basin
227
681
63
3.7
or\r\A
Mullan
1596
3260
740
3.5
ZUU4
Osburn
1099
2066
530
3.1

Side Gulches
900
2642
389
3.7

Silverton
952
1906
462
3.3

Wallace
1426
1746
869
2.9

Burke/Ninemile
2536
6024
914
4.2

Kingston
492
1424
168
3.9

Lower Basin
225
677
63
3.7

Mullan
1263
2746
548
3.6
zuuo
Osburn
1010
2004
476
3.1

Side Gulches
870
2631
370
3.7

Silverton
939
1904
452
3.3

Wallace
1402
1751
831
3.0

Burke/Ninemile
2433
5872
846
4.3

Kingston
478
1411
164
3.8

Lower Basin
224
677
63
3.7

Mullan
992
2370
397
3.6
zuuo
Osburn
911
1853
431
3.1

Side Gulches
799
2599
335
3.6

Silverton
842
1528
412
3.3

Wallace
1338
1751
752
3.2

-------
Table E-6. Basin - community soil lead concentration (mg/kg) by year



Arithmetic

Geometric



Standard

Standard
Year
Community
Mean
Deviation
Mean
Deviation

Burke/Ninemile
2329
5868
744
4.5

Kingston
458
1390
159
3.7

Lower Basin
221
673
63
3.6
onn 7
Mullan
766
1659
306
3.5
ZUU I
Osburn
807
1797
371
3.1

Side Gulches
680
1303
297
3.5

Silverton
660
1375
317
3.2

Wallace
1191
1665
606
3.5

Burke/Ninemile
2131
5829
588
4.7

Kingston
450
1384
157
3.7

Lower Basin
217
662
62
3.6

Mullan
622
1407
261
3.2
zuuo
Osburn
715
1740
324
3.0

Side Gulches
644
1287
279
3.5

Silverton
564
1287
274
3.0

Wallace
1029
1580
474
3.7

Burke/Ninemile
1866
5666
467
4.7

Kingston
434
1365
154
3.6

Lower Basin
216
661
62
3.6
onna
Mullan
527
1270
233
3.0
zuuy
Osburn
629
1632
288
2.9

Side Gulches
577
1259
246
3.3

Silverton
479
1224
236
2.9

Wallace
767
1355
341
3.5

Burke/Ninemile
1413
5229
320
4.3

Kingston
395
1334
144
3.5

Lower Basin
213
659
62
3.6
oni n
Mullan
442
1123
209
2.8
ZU I u
Osburn
495
1360
246
2.7

Side Gulches
469
1032
213
3.1

Silverton
395
700
213
2.7

Wallace
534
856
250
3.2

Burke/Ninemile
1226
5134
269
4.0

Kingston
361
1263
137
3.3

Lower Basin
208
643
61
3.5

Mullan
416
1100
200
2.7
ZU I I
Osburn
469
1319
237
2.6

Side Gulches
432
1000
199
3.0

Silverton
375
680
205
2.7

Wallace
429
732
211
2.9

-------
Table E-6. Basin - community soil lead concentration (mg/kg) by year


Arithmetic

Geometric


Standard

Standard
Year
Community
Mean
Deviation
Mean
Deviation

Burke/Ninemile
992
4992
217
3.5

Kingston
238
721
118
2.8

Lower Basin
173
539
58
3.3

Mullan
343
978
182
2.5
2013






Osburn
426
1241
226
2.5

Side Gulches
350
783
179
2.8

Silverton
255
355
172
2.3

Wallace
305
506
175
2.5

Burke/Ninemile
878
4840
198
3.3

Kingston
185
386
110
2.6

Lower Basin
140
431
55
3.0

Mullan
307
914
174
2.4
Z\j I 0
Osburn
371
963
215
2.4

Side Gulches
307
629
169
2.6

Silverton
244
328
168
2.2

Wallace
262
396
164
2.3

Burke/Ninemile
428
2382
166
2.6

Kingston
180
361
110
2.6

Lower Basin
131
406
55
2.9

Mullan
249
316
168
2.2
2017






Osburn
333
804
208
2.3

Side Gulches
258
451
158
2.5

Silverton
232
270
166
2.2

Wallace
255
356
162
2.3

Burke/Ninemile
326
1523
158
2.4

Kingston
177
354
109
2.6

Lower Basin
130
397
55
2.9

Mullan
246
308
167
2.2
ZU I o
Osburn
329
802
207
2.3

Side Gulches
257
451
158
2.5

Silverton
224
252
163
2.1

Wallace
253
350
161
2.3

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-1. Blood lead level histograms for paired records (|jg/dL), 1995-2018 - Box
(2 pages)
1995
(N=99)
Blood Lead Level (pg/dL)
100%
80%
60%
40%
20%
0%
1996
(N=102)
Blood Lead Level (pg/dL)
100%
80%
60%
40%
20%
0%
1997
(N=69)
100%
80%
1998
(N=100)
\.%



Blood Lead Level (ng/dL)
Blood Lead Level (pg/dL)
100%
80%
60%
40%
20%
0%
1999
(N=115)
100%
80%
60%
2000
(N=90)
\.1

vA

Blood Lead Level (pg/dL)
Blood Lead Level (pg/dL)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-1. Blood lead level histograms for paired records (|jg/dL), 1995-2018 - Box
(2 pages)
100%
80%
60%
40%
20%
0%
kSI
2001
(N=81)

v*

Blood Lead Level (jjg/dL)
100%
80%
60%
40%
20%
0%
2002
(N=79)
v*

Blood Lead Level (pg/dL)
100%
80%
60%
40%
20%
0%
\Sl
2003-2008
(N=11)
2013
(N=146)
100%
80%
60%
40%
20%
0%




K.&
Blood Lead Level (jjg/dL)
Blood Lead Level (|jg/dL)
100%
80%
60%
40%
20%
0%
2018
(N=43)
Blood Lead Level (jjg/dL)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-2. Soil lead concentration histograms for paired records (mg/kg), 1995-2018
Box (2 pages)
1995
(N=99)
Property Soil Lead Concentration
(mg/kg)
1996
(N=102)
Property Soil Lead Concentration
(mg/kg)
1997
(N=69)
1998
(N=100)
Property Soil Lead Concentration
(mg/kg)
Property Soil Lead Concentration
(mg/kg)
1999
(N=115)
100%
80%
60%
2000
(N=90)
40%
20%
»a0<>	^ ^
Property Soil Lead Concentration
(mg/kg)
Property Soil Lead Concentration
(mg/kg)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-2. Soil lead concentration histograms for paired records (mg/kg), 1995-2018
Box (2 pages)
100%
80%
60%
2001
(N=81)
40%
20%
Property Soil Lead Concentration
(mg/kg)
100%
80%
60%
2002
(N=79)
40%
20%
Property Soil Lead Concentration
(mg/kg)
100%
80%
60%
2003-2008
(N=11)
2013
(N=146)
40%
20%
Property Soil Lead Concentration
(mg/kg)
0-^	^ ^
Property Soil Lead Concentration
(mg/kg)
2018
(N=43)
*** ^ ^ ^
Property Soil Lead Concentration
(mg/kg)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-3. Vacuum lead concentration histograms for paired records (mg/kg), 1995-
2018 - Box (2 pages)
100%
80%
60%
40%
20%
0%
1995
(N=99)
^
House Dust Vacuum Bag Lead
Concentration (mg/kg)
1996
(N=102)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
100%
80%
60%
40%
20%
0%
1997
(N=69)
1998
(N=100)
^ ^ ^
House Dust Vacuum Bag Lead
Concentration (mg/kg)
^ ^ ^
House Dust Vacuum Bag Lead
Concentration (mg/kg)
100%
80%
60%
40%
20%
0%
1999
(N=115)
100%
80%
60%
40%
20%
0%
2000
(N=90)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
House Dust Vacuum Bag Lead
Concentration (mg/kg)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-3. Vacuum lead concentration histograms for paired records (mg/kg), 1995-
2018 - Box (2 pages)
100%
80%
60%
40%
20%
0%
2001
(N=81)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
100%
80%
60%
40%
20%
0%
2002
(N=79)
^ ^ ^ J*
House Dust Vacuum Bag Lead
Concentration (mg/kg)
2003-2008
(N=11)
2013
(N=146)
100%
80%
60%
40%
20%
0%
100%
80%
60%
40%
20%
0%
House Dust Vacuum Bag Lead
Concentration (mg/kg)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
2018
(N=43)
100%
80%
60%
40%
20%
0%
*** ^ ^ ^
House Dust Vacuum Bag Lead
Concentration (mg/kg)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-4. Blood lead level histograms for paired records (|jg/dL), 2002-2018 - Basin
(2 pages)
100%
80%
60%
40%
20%
0%
2002-2005
(N=34)


Blood Lead Level (pg/dL)
2006
(N=38)
Blood Lead Level (|jg/dL)
100%
80%
60%
40%
20%
0%
kSI
2007
(N=30)
vi


Blood Lead Level (|jg/dL)
2008
(N=34)
Blood Lead Level (pg/dL)
100%
80%
60%
40%
20%
0%
2009
(N=68)
Blood Lead Level (pg/dL)
2010
(N=40)
100%
80%
60%
40%
20%
Blood Lead Level (pg/dL)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-4. Blood lead level histograms for paired records (|jg/dL), 2002-2018 - Basin
(2 pages)
100%
80%
60%
40%
20%
0%

2011
(N=11)
vl

Blood Lead Level (pg/dL)
2013
(N=33)
Blood Lead Level (pg/dL)
100%
80%
60%
40%
20%
0%
2015
(N=29)
Blood Lead Level (pg/dL)
100%
80%
60%
40%
20%
0%
\fl
2017
(N=16)
v3

Blood Lead Level (pg/dL)
2018
(N=15)
100%
80%
60%
40%
20%
0%
Blood Lead Level (pg/dL)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-5. Soil lead concentration histograms for paired records (mg/kg), 2002-2018
Basin (2 pages)
100%
80%
60%
40%
20%
0%
2002-2005
(N=34)
1
^ ^ ^
Property Soil Lead Concentration
(mg/kg)
100%
80%
60%
40%
20%
0%
2006
(N=38)
J
Property Soil Lead Concentration
(mg/kg)
100%
80%
60%
40%
20%
0%
2007
(N=30)
Property Soil Lead Concentration
(mg/kg)
100%
80%
60%
40%
20%
0%
2008
(N=34)
Property Soil Lead Concentration
(mg/kg)
100%
80%
60%
40%
20%
0%
2009
(N=68)
. \C>0 . oCp . O,o0	6oO 6oO
o ^	o,0v ^ 7
Property Soil Lead Concentration
(mg/kg)
100%
80%
60%
40%
20%
0%
2010
(N=40)
Property Soil Lead Concentration
(mg/kg)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-5. Soil lead concentration histograms for paired records (mg/kg), 2002-2018
Basin (2 pages)
100%
80%
60%
2011
(N=11)
40%
20%
Property Soil Lead Concentration
(mg/kg)
100%
80%
60%
40%
20%
0%
2013
(N=33)
Property Soil Lead Concentration
(mg/kg)
100%
80%
60%
40%
20%
0%
2015
(N=29)
Property Soil Lead Concentration
(mg/kg)
100%
80%
60%
40%
20%
0%
2017
(N=16)
Property Soil Lead Concentration
(mg/kg)
100%
2018
(N=15)
80%

60%

40%

20%

0%
0

Property Soil Lead Concentration
(mg/kg)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-6. Vacuum lead concentration histograms for paired records (mg/kg), 2002-
2018 - Basin (2 pages)
100%
80%
60%
40%
20%
0%
2002-2005
(N=34)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
100%
80%
60%
40%
20%
0%
2006
(N=38)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
100%
80%
60%
40%
20%
0%
2007
(N=30)

House Dust Vacuum Bag Lead
Concentration (mg/kg)
2008
(N=34)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
100%
80%
60%
40%
20%
0%
2009
(N=68)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
100%
80%
60%
40%
20%
0%
2010
(N=40)

House Dust Vacuum Bag Lead
Concentration (mg/kg)

-------
Advancing Pb Exposure and Biokinetic Modeling
Figure E-6. Vacuum lead concentration histograms for paired records (mg/kg), 2002-
2018 - Basin (2 pages)
100%
80%
60%
40%
20%
0%
2011
(N=11)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
100%
80%
60%
40%
20%
0%
2013
(N=33)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
100%
80%
60%
40%
20%
0%
2015
(N=29)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
2017
(N=16)
House Dust Vacuum Bag Lead
Concentration (mg/kg)
100%

80%

60%

40%

20%

0%

2018
(N=15)
-I

House Dust Vacuum Bag Lead
Concentration (mg/kg)

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix F
Censor Level 1 (<30.5 |jg/dL) Tables and Figures
F-1

-------
Site-Wide (Total)







	











	



|H




Box (Total)
935
3.7
3.5,3.8
1.9
935
3.3
3.2,3.5
935
3.3
3.2, 3.4
935
3.0
2.9,3.2
929
4.3
4.1,4.4
935
4.2
4.1,4.4
935
4.1
4.0,4.3
935
3.9
3.8,4.1
933
5.7
5.5, 5.9
Basin (Total)
346
2.5
2.4, 2.7
1.7
346
3.1
2.9,3.2
346
3.0
2.8,3.2
346
2.7
2.5,2.9
346
3.8
3.6,4.1
346
3.7
3.5,3.9
346
3.6
3.4, 3.8
346
3.4
3.2, 3.5
346
4.9
4.6,5.1
Upper Basin (Total)
258
2.6
2.4, 2.7
1.6
258
3.2
3.0, 3.4
258
3.1
2.9, 3.3
258
2.8
2.6, 3.0
258
4.0
3.7,4.3
258
3.9
3.7,4.2
258
3.8
3.6,4.0
258
3.6
3.4, 3.8
258
5.3
5.0, 5.6
Lower Basin (Total)
88
2.4
2.2, 2.8
1.8
88
2.8
2.5,3.1
88
2.7
2.4, 3.0
88
2.4
2.1,2.7
88
3.3
2.9, 3.7
88
3.1
2.8, 3.4
88
3.0
2.7, 3.3
88
2.7
2.4, 3.1
88
3.8
3.4,4.2
Site-Wide by Age (years)
0.5 to <1

3.4
3.0,3.9
1.8

4.6
4.0,5.2

4.0
3.5,4.5

4.6
4.1, 5.3

4.6
4.0,5.3

6.0
5.4, 6.7

5.2
4.7, 5.8

6.1
5.5,6.8

6.1
5.5,6.8
1 to <2
153
4.1
3.7,4.5
1.9
153
4.8
4.4, 5.3
153
4.6
4.2,5.1
153
4.5
4.1, 5.0
151
5.6
5.1, 6.2
153
6.3
5.8, 6.8
153
6.0
5.6, 6.5
153
6.1
5.6, 6.6
152
7.7
7.1, 8.3
2 to <3
224
3.6
3.3, 3.9
1.9
224
3.3
3.1, 3.6
224
3.2
3.0, 3.4
224
3.0
2.8, 3.2
224
4.7
4.3,5.1
224
4.2
3.9,4.4
224
3.9
3.7,4.1
224
3.8
3.6,4.1
224
6.1
5.7, 6.6
3 to <4
202
3.3
3.1, 3.6
1.9
202
3.2
2.9, 3.4
202
3.1
2.8, 3.3
202
2.8
2.6,3.1
201
4.7
4.3, 5.2
202
4.0
3.7,4.3
202
3.8
3.6,4.1
202
3.7
3.4,4.0
201
6.3
5.9, 6.8
4 to <5
201
3.3
3.0, 3.6
1.9
201
3.1
2.9, 3.4
201
3.0
2.8, 3.2
201
2.8
2.6, 3.0
200
3.9
3.6,4.3
201
3.9
3.6,4.2
201
3.7
3.4, 3.9
201
3.6
3.3,3.8
201
5.2
4.8,5.6
5 to <6
206
3.0
2.8,3.3
1.8
206
2.9
2.7, 3.1
206
3.0
2.8, 3.2
206
2.6
2.4, 2.8
205
3.6
3.3, 3.9
206
3.6
3.4, 3.9
206
3.8
3.5,4.0
206
3.3
3.0, 3.5
206
4.8
4.4, 5.1
6 to <7
215
2.8
2.6, 3.0
1.8
215
2.5
2.4, 2.7
215
2.7
2.5, 2.8
215
2.2
2.0, 2.3
214
3.0
2.8, 3.2
215
3.1
2.9, 3.3
215
3.3
3.1,3.5
215
2.8
2.6, 2.9
215
3.9
3.6,4.1





























0.5 to <1
52
3.6
3.0,4.2
1.8
52
4.5
3.8,5.3
52
3.9
3.3,4.7
52
4.6
3.9,5.5
52
4.6
3.9, 5.4
52
6.1
5.3,7.0
52
5.3
4.6,6.0
52
6.2
5.4,7.1
52
6.2
5.4,7.1
1 to <2
117
4.6
4.2, 5.2
1.8
117
5.1
4.6, 5.7
117
4.9
4.4, 5.4
117
4.9
4.3, 5.4
115
6.0
5.4, 6.7
117
6.6
6.0, 7.3
117
6.3
5.8, 6.9
117
6.4
5.8,7.1
116
8.1
7.4, 8.9
2 to <3
158
4.1
3.7,4.5
1.9
158
3.5
3.2, 3.8
158
3.3
3.1, 3.6
158
3.2
2.9, 3.5
158
5.0
4.6, 5.5
158
4.4
4.1,4.7
158
4.1
3.8,4.4
158
4.1
3.7,4.4
158
6.5
6.0,7.1
3 to <4
147
3.8
3.4,4.2
1.9
147
3.3
3.0, 3.6
147
3.2
2.9, 3.5
147
3.0
2.7, 3.3
146
4.9
4.4, 5.5
147
4.1
3.8,4.5
147
4.0
3.6,4.3
147
3.8
3.5,4.2
146
6.6
6.0, 7.2
4 to <5
148
3.6
3.3,4.0
2.0
148
3.1
2.9, 3.4
148
3.0
2.7, 3.2
148
2.8
2.6,3.1
147
3.9
3.6,4.3
148
4.0
3.7,4.3
148
3.8
3.5,4.1
148
3.7
3.4,4.0
148
5.4
4.9,5.9
5 to <6
155
3.2
2.9,3.6
1.8
155
3.0
2.7, 3.2
155
3.0
2.8, 3.3
155
2.6
2.4, 2.8
154
3.7
3.3,4.0
155
3.8
3.5,4.0
155
3.9
3.6,4.2
155
3.4
3.1, 3.7
155
5.0
4.6, 5.4
6 to <7
158
3.0
2.8, 3.3
1.8
158
2.6
2.4, 2.8
158
2.7
2.5, 3.0
158
2.3
2.1,2.5
157
3.1
2.8, 3.3
158
3.2
3.0, 3.4
158
3.4
3.1,3.6
158
2.9
2.6,3.1
158
4.0
3.7,4.4
Basin (Total) by Age (years)




























0.5 to <1
28
3.2
2.5,4.0
1.8
28
4.7
3.7, 5.9
28
4.1
3.2,5.1
28
4.7
3.7, 6.0
28
4.7
3.7, 5.9
28
5.9
4.9,7.0
28
5.1
4.2,6.1
28
6.0
5.0,7.1
28
5.9
5.0,7.1
1 to <2
36
2.8
2.2, 3.4
1.8
36
3.9
3.3,4.7
36
3.8
3.2,4.5
36
3.6
3.0,4.4
36
4.6
3.8, 5.5
36
5.3
4.7, 5.9
36
5.1
4.5, 5.7
36
5.0
4.4, 5.7
36
6.4
5.6, 7.3
2 to <3
66
2.7
2.3,3.1
1.8
66
3.0
2.6, 3.4
66
2.8
2.5, 3.2
66
2.6
2.3, 3.0
66
4.0
3.5,4.7
66
3.7
3.3,4.1
66
3.5
3.1, 3.8
66
3.3
3.0, 3.7
66
5.3
4.7, 6.0
3 to <4
55
2.4
2.1, 2.8
1.7
55
2.9
2.6, 3.4
55
2.8
2.5, 3.2
55
2.5
2.2, 3.0
55
4.3
3.6,5.1
55
3.7
3.3,4.1
55
3.5
3.1, 3.9
55
3.3
2.9, 3.7
55
5.8
5.1, 6.5
4 to <5
53
2.4
2.2, 2.7
1.5
53
3.2
2.7, 3.8
53
3.0
2.6, 3.6
53
2.8
2.3, 3.3
53
3.9
3.2,4.8
53
3.6
3.2,4.1
53
3.4
3.0,3.8
53
3.2
2.7, 3.7
53
4.6
4.0, 5.4
5 to <6
51
2.5
2.2,2.8
1.5
51
2.8
2.5,3.3
51
2.9
2.5, 3.4
51
2.4
2.0,2.8
51
3.4
2.9,4.1
51
3.3
2.9,3.7
51
3.4
3.0, 3.8
51
2.8
2.5, 3.2
51
4.1
3.6,4.8
6 to <7
57
2.2
2.0, 2.4
1.4
57
2.3
2.1,2.6
57
2.5
2.2, 2.7
57
2.0
1.8,2.2
57
2.7
2.4, 3.1
57
2.8
2.6,3.1
57
3.0
2.7, 3.2
57
2.5
2.2, 2.7
57
3.5
3.1,3.8
box (Total) by Calendar Year
99
4.9
4.4, 5.5
1.8
99
4.4
3.8,5.1
99
4.3
3.8,5.0
99

3.5,4.8
97
5.7
4.9,6.6
99
6.8
6.1, 7.6
99
6.7
6.0, 7.4
99
6.5
5.9,7.3
98
9.5
8.6, 10.5
1996
102
4.9
4.3,5.6
1.9
102
3.9
3.4,4.4
102
3.8
3.4,4.3
102
3.6
3.1,4.1
102
5.1
4.5, 5.9
102
6.0
5.5, 6.6
102
5.9
5.4, 6.5
102
5.7
5.2, 6.3
102
8.4
7.6, 9.3
1997
69
4.6
4.0, 5.3
1.8
69
3.5
3.1,4.1
69
3.4
3.0, 3.9
69
3.2
2.8, 3.8
68
4.6
4.0, 5.3
69
5.1
4.5, 5.7
69
4.9
4.4, 5.5
69
4.8
4.2, 5.4
69
7.2
6.4, 8.2
1998
100
4.3
3.8,4.9
1.9
100
3.9
3.5,4.3
100
3.8
3.5,4.2
100
3.6
3.2,4.0
100
5.3
4.7, 5.9
100
5.0
4.6, 5.4
100
4.9
4.5, 5.3
100
4.7
4.3,5.1
100
7.0
6.4, 7.6
1999
115
4.2
3.7,4.7
1.8
115
3.8
3.4,4.2
115
3.7
3.3,4.0
115
3.5
3.1, 3.9
113
4.8
4.4, 5.3
115
4.6
4.2,5.0
115
4.4
4.1,4.8
115
4.3
3.9,4.7
114
6.2
5.7, 6.7
2000
90
4.0
3.5,4.5
1.8
90
3.4
3.0,3.8
90
3.3
3.0,3.7
90
3.1
2.7, 3.5
89
4.2
3.8,4.8
90
4.1
3.7,4.5
90
4.0
3.7,4.4
90
3.8
3.4,4.2
90
5.5
5.0,6.1
2001
81
3.2
2.8, 3.7
1.8
81
3.0
2.7, 3.3
81
2.9
2.6, 3.2
81
2.7
2.4, 3.0
81
3.9
3.4,4.4
81
4.0
3.7,4.4
81
3.9
3.6,4.2
81
3.8
3.4,4.1
81
5.5
5.0, 6.0
2002
79
2.6
2.3, 2.8
1.5
79
2.7
2.5,3.1
79
2.7
2.4, 3.0
79
2.4
2.2, 2.8
79
3.4
3.0, 3.8
79
3.5
3.2, 3.8
79
3.4
3.1, 3.7
79
3.2
2.9, 3.5
79
4.6
4.1, 5.1
2003-2008
11
3.8
2.3, 6.2
2.1
11
3.8
2.3, 6.3
11
3.7
2.2,6.1
11
3.5
2.0, 6.0
11
5.0
2.8, 8.8
11
3.8
2.4, 6.2
11
3.7
2.3, 5.9
11
3.5
2.1, 5.9
11
5.1
3.1, 8.7
2013
146
2.3
2.2, 2.5
1.6
146
2.6
2.5,2.8
146
2.5
2.4, 2.7
146
2.3
2.2,2.5
146
3.3
3.0,3.5
146
2.6
2.5,2.8
146
2.6
2.4, 2.7
146
2.4
2.2,2.5
146
3.3
3.1, 3.5

43
2.7
2.3,3.1
1.6
43
2.2
2.0, 2.4
43
2.2
1.9,2.4
43
1.9
1.7,2.2
43
2.6
2.3, 3.0
43
2.3
2.1,2.6
43
2.3
2.1,2.5
43
2.1
1.8,2.3
43
2.9
2.6, 3.2
Basin (Total) by Calendar Year

























34

6.1, 9.0
2006
38
2.4
2.0, 2.9
1.7
38
4.0
3.2,4.9
38
3.8
3.1,4.8
38
3.5
2.7,4.5
38
4.9
3.8, 6.3
38
4.8
4.0, 5.6
38
4.6
3.9, 5.4
38
4.3
3.6, 5.3
38
6.2
5.2, 7.5
2007
30
2.7
2.2, 3.4
1.8
30
3.1
2.5, 3.9
30
3.1
2.5, 3.8
30
2.7
2.0, 3.5
30
3.8
2.8, 5.0
30
4.1
3.5,4.9
30
4.0
3.4,4.7
30
3.7
3.0,4.6
30
5.4
4.4, 6.7
2008
34
1.9
1.7, 2.2
1.4
34
2.6
2.3, 3.0
34
2.6
2.3, 2.9
34
2.3
1.9, 2.6
34
3.3
2.8, 3.8
34
3.3
2.9, 3.7
34
3.2
2.9, 3.6
34
2.9
2.6, 3.3
34
4.3
3.8,4.9
2009
68
2.7
2.4, 3.0
1.5
68
2.8
2.5,3.1
68
2.7
2.5, 3.0
68
2.4
2.2, 2.7
68
3.5
3.1, 3.9
68
3.7
3.4,4.0
68
3.6
3.3,3.9
68
3.3
3.0,3.7
68
4.9
4.4, 5.4
2010
40
2.0
1.8,2.3
1.5
40
2.7
2.4, 3.0
40
2.6
2.4, 2.9
40
2.3
2.0,2.7
40
3.4
3.0,3.9
40
3.3
3.0,3.7
40
3.2
3.0, 3.6
40
3.0
2.6,3.3
40
4.4
4.0, 5.0
2011

3.1
2.2,4.3
1.7

3.0
2.3,4.0

2.9
2.2, 3.9

2.7
1.9, 3.7

4.0
2.8, 5.6

3.7
3.0,4.6

3.6
2.9,4.4

3.4
2.7,4.3

5.2
4.1, 6.5
2013
33
2.6
2.2, 3.0
1.5
33
2.8
2.3, 3.3
33
2.7
2.3, 3.3
33
2.4
2.0, 3.0
33
3.4
2.7,4.1
33
3.0
2.6, 3.5
33
3.0
2.6, 3.4
33
2.7
2.3, 3.2
33
3.8
3.2,4.5
2015
29
2.5
2.1, 3.0
1.6
29
3.2
2.5,4.1
29
3.1
2.5,4.0
29
2.9
2.2, 3.8
29
4.1
3.2, 5.4
29
3.4
2.7,4.1
29
3.2
2.7,4.0
29
3.0
2.4, 3.8
29
4.3
3.5, 5.4
2017
16
4.2
3.1, 5.7
1.7
16
2.3
2.0,2.7
16
2.3
2.0,2.6
16
2.0
1.7, 2.5
16
2.8
2.3,3.5
16
2.6
2.3,2.9
16
2.4
2.2,2.7
16
2.3
1.9,2.7
16
3.2
2.7, 3.8
2018
13
2.2
1.8,2.7
1.4
13
3.2
2.4,4.1
13
3.1
2.4,4.0
13
2.9
2.2,3.8
13
4.2
3.0,5.8
13
3.2
2.6,4.0
13
3.1
2.5,3.8
13
3.0
2.4,3.7
13
4.3
3.3,5.6

-------
Box (Total)
Basin (Total)
Upper Basin (Total)
	Lower Basin fTotal)
Site-Wide by Age (years)
0.5 to <1
1	to <2
2	to <3
3	to <4
Box (Total) by Age (years)
0.5 to <1
1	to <2
2	to <3
3	to <4
Basin (Total) by Age (years)
0.5 to <1
1	to <2
2	to <3
3	to <4
Box (Total) by Calendar Year
Average	95% C for
Percent	Average %
BLLs >5	BLLs >5 pg'dL
|ig?dL (%)	(%)
16.6-35.9
33.4-49.0
26.9-39.2
22.9-35.5
22.1 - 34.6
) - 29.6
3.7 - 32.0
6.5 - 32.4
6.5 - 23.8
5.2 - 23.9
0.3-9.9
-13.2
J - 66.3
40.3-59.7
43.3-66.8
42.2-61.8
34.4-52.5
36.4-57.0
14.2-32.7
3.5-16.8
1.0-53.6
Average
Pred cted 95% C for
Probabilities >5 Probabilities 26
20.5-32.1
17.5-29.2
16.5-28.0
21.6-35.7
17.6-31.5
15.1-28.4
13.6-26.1
29.3-48.5
24.0-42.3
18.3-39.7
24.5-43.0
22.5-39.4
16.6-34.6
12.1-29.8
22.8-55.6
22.9-53.8
10.6-42.1
2.4 - 25.9
Average
Pred cted 95% C for
Probabilities >5 Probabilities 25
22.0-27.5
16.7-25.3
17.3-27.5
24.2-45.0
34.5-50.2
18.1-29.2
16.2-27.6
14.5-25.5
15.5-26.7
9.3 - 19.2
9.1 - 27.8
8.4 - 29.1
10.5-32.6
9.7 - 32.1
23.1-41.4
20.7-37.3
16.1-34.0
10.6-27.7
7.5 - 23.5
5.1 - 60.6
7.6-18.6
I-15.3
6.7 - 23.8
2.6 - 23.4
5.9 - 49.4
20.7 - 26.1
15.4-23.8
15.7-25.6
36.5 - 54.6
18.7-32.2
15.3-28.7
12.6-25.2
9.8-31.5
6.4-26.9
0.9-15.0
0.6-21.8
5.4-21.7
1.2-20.6
49.2 - 67.2
40.5 - 56.0
37.5 - 53.7
24.2 - 60.9
27.7 - 60.1
24.7 - 47.9
25.8-51.6
32.6	- 66.2
34.7	- 66.5
20.6 - 55.4
10.2-39.1
25.8-35.5
27.9-39.4
13.3-30.6
46.6-73.2
56.4-73.7
33.7-49.0
29.0-44.6
28.2 - 43.6
24.8-39.5
16.2-40.4
12.0-35.4
6.1 - 24.9
57.2-75.8
50.5-69.6
38.5-62.1
38.2-57.8
33.0-51.1
24.7-44.3
24.9-45.7
16.5-35.9
36.5-70.1
31.3-63.0
21.3-56.2
8.4 - 36.5
18.4-40.0
10.0-36.1
2.6 - 56.5
6.2 - 33.4
34.5 - 40.7
24.1	- 33.6
26.2	- 37.6
11.5-28.1
30.1 - 67.1
33.1 - 65.7
15.1 -36.2
56.6	- 75.2
50.1	- 69.2
36.3 - 59.9
36.7	- 56.3
30.9 - 48.8
23.7 - 43.2
22.2	- 42.6
15.0-33.9
6.4-62.6
35.0 - 68.5
29.4-61.1
19.2-53.7
33.0	- 39.2
22.4-31.7
24.1	- 35.3
11.0-27.5
52.0 - 67.6
27.7 - 40.1
24.9 - 37.7
25.1 - 40.2
21.2-35.3
13.2-25.5
40.5 - 77.0
32.7 - 65.4
14.2-35.0
13.4-36.2
12.8-35.9
7 - 30.5
I - 66.6
35.4 - 59.0
34.7 - 54.2
29.7 - 47.5
21.3-40.3
21.7-42.0
13.9-32.5
6.1-62.2
6.0-16.2
2.4-18.6
33.1 - 66.7
27.9 - 59.5
18.2-52.4
5.2-31.1
14.8-35.4
6.7-31.0
8.5-54.4
53.1	- 59.5
41.9-52.4
46.2	- 58.4
22.5 - 42.1
47.5	- 74.0
67.6	- 83.2
56.6-71.5
50.8-81.7
40.1 - 64.2
43.9 - 70.1
53.4-71.2
43.7 - 64.3
45.2 - 66.8
31.5-53.3
27.1 - 57.7
23.7 - 82.7
15.6-47.3
19.7-54.9
is, both partition node I runs. By censoring the data sets, the lEUBKvl .1,
.1 IRs for both paitmons resun in fewer BLL records due to predicted BLLs >30 iigfdL.

-------
Table F-3. Summary of observed/predicted blood Pb pairs that fall outside prediction intervals for censor level 1 (dataset omits predicted and observed BLLs S30.5 M#dL)

IEUBK v2.0
IEUBK v1.1


IEUBK v2.0
IEUBK vl.1

v2.0 IRso
EFH (USEPA 2017b) IRs, v2.0 IR^

V1.1 IRsp
v2.0 IRso
EFH (USEPA 2017b) IRsD v2.0 IRgp

vl.1 IRsd
% Outside % < Lower % > Upper % Outside % < Lower % > Upper % Outside % < Lower % > Upper % Outside % < Lower % > Upper
% Outside %< Lower %> Upper % Outside %< Lower %> Upper % Outside %< Lower %> Upper % Outside %< Lower %> Upper
Site Wide (Total)
Box (Total)
Basin (Total)
Upper Basin (Total)
Lower Basin (Total)
13%
12%
16%
12%
22%
9351
14%
17%
16%
18%
26%
23%
33%
35%
26%
22%
31%
34%
22%
Site-Wide by Age (years)
0.5 to <1
1	to <2
2	to <3
3	to <4
4	to <5
5	to <6
6	to <7
22%
20%
20%
20%
18%
14%
14%
10%
12%
31%
32%
28%
31%
23%
22%
28%
32%
26%
30%
22%
21%
Box (Total) by Age (years)
0.5 to <1
1	to <2
2	to <3
3	to <4
4	to <5
5	to <6
6	to <7
14%
20%
14%
11%
13%
10%
14%
10%
14%
29%
25%
25%
25%
22%
23%
27%
25%
23%
24%
20%
23%
14%
Basin (Total) by Age (years)
0.5 to <1
1	to <2
2	to <3
3	to <4
4	to <5
5	to <6
6	to <7
21%
22%
14%
25%
25%
14%
31%
26%
29%
21%
14%
31%
21%
29%
36%
44%
14%
36%
53%
35%
29%
53%
32%
*40% dust, 30% property soil, 30% arithmetic average community soil.
Grey shaded/bolded text cells represent the lowest percentage of all 8 model runs, for each data categoiy.
The number of records (N) for each model run corresponds to the s<
i Tables F-1 and F-2.

-------
Table F-4. Summary of sum of squared differences for censor level 1 (dataset omits predicted and observed BLLs >30.5 ng/dL)


55/45 Dust/Soil Partition


40/30/30 Dust/Soil* Partition


IEUBK v2.0
IEUBK V1.1
IEUBK V2.0
IEUBK V1.1


EFH(USEPA



EFH(USEPA



v2.0 IRsd
2017b) IRsd
v2.0 IRsd
v1.1 IRsd
v2.0 IRsd
2017b) IRsd
v2.0 IRsd
v1.1 IRsd
Site-Wide (Total)
583
577
653
681
562
541
563
854
Box (Total)
434
435
491
457
389
379
397
567
Basin (Total)
149
143
162
225
173
162
165
287
Upper Basin (Total)
104
99
111
165
128
120
119
224
Lower Basin (Total)
45
43
51
59
45
42
46
63
Site-Wide by Age (years)








0.5 to <1
41
36
42
41
56
44
58
57
1 to <2
64
63
66
78
79
73
76
113
2 to <3
109
111
122
135
101
98
102
166
3 to <4
78
78
88
107
75
72
73
151
4 to <5
107
107
120
120
92
87
92
136
5 to <6
83
83
96
93
76
79
75
118
6 to <7
100
100
119
106
84
88
87
114
Box (Total) by Age (years)








0.5 to <1
22
20
23
23
31
24
32
32
1 to <2
44
43
45
51
49
45
47
71
2 to <3
74
76
83
83
64
63
66
101
3 to <4
61
62
70
70
52
51
54
95
4 to <5
82
85
93
80
67
65
68
93
5 to <6
66
65
76
66
59
61
59
89
6 to <7
86
84
101
85
68
70
72
86
Basin (Total) by Age (years)








0.5 to <1
19
16
19
19
25
21
26
25
1 to <2
21
19
21
27
30
28
29
41
2 to <3
36
34
39
52
37
35
36
64
3 to <4
17
16
18
37
23
21
20
56
4 to <5
24
22
28
40
25
22
24
43
5 to <6
18
18
20
27
17
18
16
29
6 to <7
15
15
18
21
16
18
15
27
Notes:
* 40% dust, 30% property soil, 30% arithmetic average community soil
Grey shaded/boIded text cells represent the lowest value of all 8 model runs, for each data category.
< = less than
IRsd = soil/dust ingestion rate
BDL = below detection limit
BLL = blood lead level
(jg/dL = microgram per deciliter
The number of records (N) for each model run corresponds to the same Ns shown in Tables F-1 and F-2.

-------
Table F-5. Percent differences <1 ng/dL between observed and predicted BLLs for censor level 1 (dataset omits predicted and observed BLLs >30.5 ng/dL)


55/45 Dust/Soil Partition


40/30/30 Dust/Soil* Partition


IEUBK v2.0
IEUBK V1.1
IEUBK V2.0
IEUBK V1.1


EFH (USEPA



EFH (USEPA



v2.0 IRsd
2017b) IRsd
v2.0 IRsd
v1.1 IRsd
v2.0 IRsd
2017b) IRsd
v2.0 SIRsd
v1.1 IRsd
Site-Wide (Total)
38%
39%
38%
32%
35%
35%
36%
22%
Box (Total)
35%
36%
35%
31%
34%
34%
33%
21%
Basin (Total)
47%
47%
45%
35%
39%
40%
42%
22%
Upper Basin
48%
48%
44%
34%
40%
41%
44%
18%
Lower Basin
43%
44%
45%
35%
36%
36%
39%
33%
Site-Wide By Age (in years)








0.5 to <1
29%
38%
30%
30%
16%
15%
13%
13%
1 to <2
33%
32%
31%
21%
18%
19%
20%
14%
2 to <3
37%
37%
35%
29%
37%
38%
37%
16%
3 to <4
38%
39%
36%
28%
38%
37%
36%
16%
4 to <5
37%
39%
39%
28%
34%
38%
37%
22%
5 to <6
41%
40%
41%
40%
39%
38%
39%
31%
6 to <7
47%
45%
45%
43%
48%
46%
50%
32%
Box (Total) by Age








0.5 to <1
35%
42%
37%
37%
19%
19%
15%
15%
1 to <2
33%
32%
33%
23%
18%
19%
21%
14%
2 to <3
30%
31%
29%
28%
35%
34%
32%
16%
3 to <4
35%
35%
34%
26%
38%
37%
34%
17%
4 to <5
31%
32%
32%
26%
32%
36%
35%
23%
5 to <6
39%
39%
38%
38%
36%
36%
35%
30%
6 to <7
42%
41%
42%
39%
45%
43%
46%
28%
Basin (Total) by Age








0.5 to <1
18%
29%
18%
18%
11%
7%
7%
7%
1 to <2
31%
31%
25%
17%
19%
19%
19%
14%
2 to <3
53%
50%
48%
32%
41%
47%
50%
15%
3 to <4
47%
47%
42%
33%
38%
38%
40%
15%
4 to <5
53%
58%
57%
34%
40%
43%
43%
19%
5 to <6
45%
45%
51%
45%
47%
45%
51%
35%
6 to <7
58%
54%
53%
51%
56%
54%
60%
40%
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
Grey shaded/boIded text cells represent the highest value of all 8 model runs, for each data category.
The number of records (N) for each model run corresponds to the same Ns shown in Tables F-1 and F-2.

-------
Advancing Pb Exposure and Biokinetic Modeling
Table F-6. Summary of the difference between observed and predicted GM BLLsfor IEUBK v2.0 model, censor level 1 (dataset omits predicted and observed BLLs£30.5 uq/dL)

Observed


Predicted Blood Lead (IEUBK Model v2.0 build 1.6)



Blood Lead*

55/45 Dust/Soil Partition


40/30/30 Dust/Soil* Partition



v2.0 IRsd
EFH (USEPA 2017b) IRSD
v2.0 IRsD
EFH (USEPA 2017b) IRSD

Geometric
Geometric
Difference between
Geometric
Difference between
Geometric
Difference between
Geometric
Difference between

Mean (pg/dL)
Mean (pg/dL)
geomeans (pg/dL)
Mean (pg/dL)
geomeans (pg/dL)
Mean (pg/dL)
geomeans (pg/dL)
Mean (pg/dL)
geomeans (pg/dL)
Site-Wide (Total)
3.31
3.26
-0.04
3.19
-0.12
4.09
0.78
3.98
0.68
Box (Total)
3.65
3.34
-0.31
3.27
-0.38
4.24
0.59
4.14
0.49
Basin (Total)
2.53
3.06
0.53
2.98
0.46
3.71
1.18
3.60
1.07
Upper Basin (Total)
2.56
3.16
0.59
3.09
0.53
3.94
1.38
3.84
1.28
Lower Basin (Total)
2.44
2.80
0.37
2.70
0.26
3.11
0.67
2.98
0.54
Site-Wide By Age (years)









0.5 to <1
3.44
4.56
1.12
3.98
0.54
6.01
2.57
5.20
1.76
1 to <2
4.11
4.80
0.70
4.61
0.51
6.27
2.16
6.01
1.90
2 to <3
3.61
3.34
-0.27
3.16
-0.45
4.15
0.55
3.91
0.30
3 to <4
3.34
3.19
-0.15
3.07
-0.27
3.98
0.65
3.83
0.49
4 to <5
3.26
3.14
-0.12
2.98
-0.28
3.90
0.64
3.66
0.40
5 to <6
3.02
2.93
-0.10
3.01
-0.01
3.63
0.60
3.75
0.73
6 to <7
2.79
2.52
-0.27
2.66
-0.12
3.08
0.30
3.27
0.49
Box (Total) by Age (years)









0.5 to <1
3.58
4.51
0.93
3.93
0.35
6.09
2.51
5.27
1.69
1 to <2
4.64
5.10
0.46
4.90
0.26
6.62
1.98
6.34
1.70
2 to <3
4.09
3.51
-0.59
3.31
-0.79
4.37
0.28
4.11
0.01
3 to <4
3.76
3.28
-0.48
3.16
-0.60
4.12
0.36
3.96
0.20
4 to <5
3.63
3.13
-0.50
2.95
-0.68
4.01
0.38
3.77
0.14
5 to <6
3.22
2.95
-0.27
3.05
-0.18
3.76
0.53
3.89
0.67
6 to <7
3.03
2.59
-0.44
2.73
-0.30
3.19
0.16
3.38
0.35
Basin (Total) by Age (years)









0.5 to <1
3.20
4.65
1.45
4.07
0.87
5.86
2.66
5.07
1.87
1 to <2
2.76
3.95
1.19
3.80
1.03
5.27
2.50
5.06
2.29
2 to <3
2.67
2.98
0.31
2.83
0.16
3.68
1.01
3.46
0.80
3 to <4
2.43
2.95
0.52
2.84
0.41
3.65
1.23
3.50
1.08
4 to <5
2.42
3.20
0.78
3.04
0.62
3.60
1.18
3.37
0.95
5 to <6
2.49
2.84
0.35
2.91
0.42
3.26
0.77
3.35
0.86
6 to <7
2.21
2.34
0.13
2.47
0.26
2.81
0.60
2.98
0.77
Box (Total) by Calendar Year









1995
4.91
4.41
-0.50
4.34
-0.57
6.81
1.89
6.68
1.76
1996
4.91
3.89
-1.02
3.83
-1.08
5.99
1.08
5.91
1.00
1997
4.62
3.55
-1.07
3.45
-1.18
5.08
0.45
4.92
0.30
1998
4.30
3.91
-0.39
3.82
-0.47
4.97
0.67
4.86
0.56
1999
4.18
3.77
-0.41
3.66
-0.51
4.57
0.40
4.44
0.26
2000
3.98
3.36
-0.63
3.32
-0.66
4.07
0.08
4.02
0.03
2001
3.20
3.00
-0.20
2.90
-0.30
4.03
0.82
3.88
0.68
2002
2.58
2.74
0.16
2.68
0.10
3.47
0.89
3.38
0.80
2003-2008
3.76
3.77
0.02
3.65
-0.10
3.85
0.09
3.71
-0.04
2013
2.34
2.62
0.27
2.54
0.20
2.64
0.29
2.56
0.22
2018
2.68
2.20
-0.48
2.16
-0.52
2.34
-0.34
2.29
-0.38
Basin (Total) by Calendar Year









2002-2005
2.93
4.28
1.35
4.16
1.22
5.55
2.62
5.38
2.44
2006
2.40
3.98
1.58
3.84
1.45
4.76
2.36
4.57
2.17
2007
2.73
3.13
0.40
3.07
0.34
4.13
1.40
4.00
1.27
2008
1.92
2.64
0.72
2.61
0.70
3.27
1.36
3.24
1.32
2009
2.68
2.79
0.10
2.72
0.03
3.67
0.98
3.56
0.87
2010
2.02
2.69
0.67
2.63
0.61
3.33
1.31
3.24
1.22
2011
3.10
3.00
-0.10
2.93
-0.17
3.68
0.58
3.55
0.45
2013
2.55
2.78
0.23
2.74
0.19
3.04
0.49
2.99
0.44
2015
2.53
3.23
0.71
3.13
0.60
3.37
0.84
3.24
0.71
2017
4.23
2.35
-1.88
2.25
-1.97
2.56
-1.66
2.44
-1.78
2018
2.23
3.17
0.94
3.08
0.85
3.23
1.00
3.12
0.89
*40% dust, 30% property soil, 30% arithmetic average community soil
Differences are calculated as predicted GM minus observed GM
IRbd = soil/dust ingestion rate
BLL = blood lead level
GM = geometric mean
pg/dL = microgram per deciliter

-------
Advancing Pb Exposure and Biokinetic Modeling




Predicted Blood Lead (IEUBK Model v2.0 build 1.6)





55/45 Dust/Soil Partition


40/30/30 Dust/Soil* Partition



V2.0 IRS„
EFH (USEPA 2017b) IRSD
v2.0 IRsd
EFH (USEPA 2017b) IRSD

Average Percent
Average Predicted
Difference between
Average Predicted
Difference between
Average Predicted
Difference between
Average Predicted
Difference between

of BLLs >5 pg/dL
Probabilities >5 pg/dL
average percent
Probabilities >5 pg/dL
average percent
Probabilities >5 pg/dL
average percent
Probabilities >5 pg/dL
average percent

(%>
(%)
BLLs >5 ng/dL (%)
(%)
BLLs £5 ng/dL (%)
(%)
BLLs >5 ng/dL (%)
(%)
BLLs £5 ng/dL (%)
Site-Wide (Total)
28.6
25.0
-3.6
23.8
-4.9
36.8
8.2
35.2
6.6
Box (Total)
35.5
26.0
-9.5
24.8
-10.7
39.1
3.6
37.6
2.1
Basin (Total)
10.1
22.4
12.2
21.0
10.9
30.7
20.5
28.8
18.7
Upper Basin (Total)
9.7
23.6
13.9
22.4
12.7
33.6
24.0
31.9
22.2
Lower Basin (Total)
11.4
18.7
7.3
17.0
5.6
21.9
10.5
19.8
8.4
Site-Wide By Age (years)









0.5 to <1
26.3
41.5
15.2
34.6
8.3
59.1
32.8
50.3
24.0
1 to <2
41.2
44.7
3.5
42.3
1.1
62.1
20.9
59.5
18.3
2 to <3
33.0
26.3
-6.7
23.6
-9.4
37.7
4.7
34.1
1.1
3 to <4
29.2
23.4
-5.9
21.9
-7.3
34.7
5.5
32.5
3.3
4 to <5
28.4
22.3
-6.1
20.0
-8.4
33.9
5.5
30.5
2.1
5 to <6
23.8
19.8
-3.9
21.1
-2.7
30.1
6.3
31.9
8.1
6 to <7
20.5
12.8
-7.7
14.5
-5.9
20.8
0.4
23.7
3.3
Box (Total) by Age (years)
0.0
0.0







0.5 to <1
30.8
41.2
10.4
34.3
3.5
59.9
29.2
51.2
20.4
1 to <2
47.9
48.0
0.2
45.6
-2.2
65.1
17.2
62.6
14.8
2 to <3
40.5
28.7
-11.8
25.8
-14.7
41.4
0.9
37.7
-2.8
3 to <4
34.7
24.6
-10.1
23.0
-11.7
36.8
2.1
34.6
-0.1
4 to <5
37.8
21.7
-16.1
19.4
-18.4
35.9
-2.0
32.4
-5.5
5 to <6
30.3
19.9
-10.5
21.2
-9.1
32.2
1.9
34.1
3.8
6 to <7
26.6
13.6
-13.0
15.3
-11.3
22.8
3.8
25.7
-0.8
Basin (Total) by Age (years)









0.5 to <1
17.9
42.1
24.2
35.2
17.3
57.6
39.7
48.6
30.7
1 to <2
19.4
33.9
14.4
31.6
12.1
52.5
33.1
49.4
29.9
2 to <3
15.2
20.6
5.4
18.4
3.3
29.0
13.8
25.7
10.5
3 to <4
14.5
20.1
5.5
18.7
4.2
29.1
14.6
27.0
12.4
4 to <5
1.9
23.8
21.9
21.5
19.6
28.3
26.4
25.3
23.4
5 to <6
3.9
19.8
15.9
20.9
17.0
23.7
19.8
25.2
21.3
6 to <7
3.5
10.6
7.1
12.4
8.9
15.5
12.0
18.2
14.7
Box (Total) by Calendar Year









1995
56.6
38.9
-17.7
38.1
-18.4
66.5
9.9
65.9
9.3
1996
50.0
33.1
-16.9
32.4
-17.6
60.1
10.1
59.7
9.7
1997
55.1
29.0
-26.0
27.5
-27.6
50.3
-4.8
48.1
-7.0
1998
52.0
33.7
-18.3
32.3
-19.7
48.0
-4.0
46.5
-5.5
1999
43.5
30.9
-12.5
29.0
-14.4
42.0
-1.4
39.8
-3.6
2000
46.7
25.6
-21.0
25.0
-21.6
34.5
-12.2
33.5
-13.2
2001
23.5
20.9
-2.6
19.2
-4.3
35.3
11.8
32.4
9.0
2002
10.1
16.7
6.6
15.5
5.3
26.2
16.1
24.5
14.3
2003-2008
27.3
34.1
6.9
32.8
5.6
36.3
9.0
34.5
7.3
2013
4.8
14.6
9.8
13.1
8.3
13.4
8.6
11.9
7.1
2018
14.0
8.3
-5.7
7.5
-6.5
8.8
-5.2
7.6
-6.4
Basin (Total) by Calendar Year









2002-2005
23.5
39.2
15.7
38.1
14.6
53.3
29.8
51.8
28.2
2006
13.2
38.4
25.2
37.1
23.9
47.2
34.0
45.2
32.1
2007
20.0
26.4
6.4
25.2
5.2
38.7
18.7
36.5
16.5
2008
0.0
14.2
14.2
13.2
13.2
22.4
22.4
21.1
21.1
2009
8.8
16.7
7.9
15.2
6.4
29.2
20.4
27.1
18.3
2010
2.5
14.2
11.7
13.0
10.5
23.0
20.5
21.4
18.9
2011
9.1
21.5
12.4
19.9
10.8
29.5
20.4
27.7
18.7
2013
6.1
17.3
11.2
16.2
10.1
19.8
13.7
18.3
12.2
2015
3.4
25.3
21.9
23.4
19.9
25.8
22.3
23.7
20.3
2017
25.0
7.9
-17.1
6.2
-18.8
9.9
-15.1
7.8
-17.2
2018
7.7
23.1
15.4
22.2
14.5
22.2
14.5
20.8
13.1
Notes:
* 40% dust, 30% property soil, 30% arithmetic average community soil
Differences are calculated as predicted P5 minus observed P5
IRsd = soil/dust ingestion rate
BLL = biood lead level

-------
ure F-1. Summary of mean observed and predicted blood lead levels for geographic areas for censor level 1 (dataset omits predicted and observed BLLs > 30.5 pg/dL)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Predicted Blood Lead GM (ng/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Predicted Blood Lead GM (ng/dL)
B) IEUBK v2.0 EFH IRs, 55/45 partition
4.5
_ 4.0
	I
"O
^ 3.5
^ 3.0
"? 2.5
"D
O
O
"O
CD
>
_Q
o 0.5
0.0
BOX
Site-wide ^
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Predicted Blood Lead GM (|ag/dL)
C) IEUBK v1.1 v2.0 IRs, 55/45 partition
4.5
* B0X
Site-wide
X, *^Upper
ower± Basin
Basin
Basin
0.0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Predicted Blood Lead GM (ng/dL)
00
=L
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Predicted Blood Lead GM (ng/dL)
*
Site-wide
Lower
Basin
erl „... L
Basin
Upper
Basin
~i	1	1	1	1	1	1	1	1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Predicted Blood Lead GM (ng/dL)
Box
>#HSite-wide
Basin
Lower
Basin
Upper
Basin
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition
E) IEUBK v2.0 EFH IRs, 40/30/30 partition
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition

-------
Figure F-2. Summary of observed and predicted blood lead geometric means for site-wide age groups for censor level 1 (dataset omits predicted and observed BLLs > 30.5 pg/dL)
l	l	l	l	l	l
1.0 2.0 3.0 4.0 5.0 6.0
Predicted Blood Lead GM (ng/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide by Age	B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide by Age	C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide by Age
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Predicted Blood Lead GM (ng/dL)
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
Predicted Blood Lead GM (ng/dL)
0)
4.0
1.0 2.0 3.0 4.0 5.0 6.0
Predicted Blood Lead GM (ng/dL)
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
Predicted Blood Lead GM (ng/dL)
I	I	I	I	I	I
1.0 2.0 3.0 4.0 5.0 6.0
Predicted Blood Lead GM (ng/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide by Age
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide by Age
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide by Age

-------
Figure F-3. Summary of observed and predicted average probability of exceeding 5 pg/dL by geographic areas for censor level 1 (all predicted and observed BLLs <30.5 pg/dL)
50
45
40
35
^ 30
§ 25
I—
CD
£ 20
o
15
Site-wide -L/

i 11

^Lower f!
Basin
4
t=L
Upper
Basin
Basin
Predicted (%)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition
50
45
40
35
^ 30
§ 25
I—
CD
£ 20
o
15
Site-wide x
Lower
Basin """Basin
^ Upper
Basin
Predicted (%)
B) IEUBK v2.0 EFH IRs, 55/45 partition
0 5 10 15 20 25 30 35 40 45 50
Predicted (%)
C) IEUBK v1.1 v2.0 IRs, 55/45 partition
Box
l—I~H
Site-wide
Basin"1" 1 Upper
Basin
0 5 10 15 20 25 30 35 40 45 50
Predicted (%)
0 5 10 15 20 25 30 35 40 45 50
Predicted (%)
Predicted (%)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition
E) IEUBK v2.0 EFH IRs, 40/30/30 partition
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition

-------
Figure F-4. Summary of observed and predicted average probability of exceeding 5 pg/dL for site-wide age groups for censor level 1 (all predicted and observed BLLs <30.5 pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide by Age
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide by Age
10 20 30 40 50 60 70 80
Predicted (%)
S? 50
30
£ 50
¦O
> 40
CD
CO
8 30
20
30 40 50 60 70 80
Predicted (%)
s? 50
30
20 30 40 50 60 70 80
Predicted (%)
£ 50
"O
> 40
O)
CO
8 30
20
30 40 50
Predicted {%)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide by Age	E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide by Age
Predicted (%)
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide by Age
Predicted (%)
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide by Age

-------
Figure F-5. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Site-wide
CD
"O
©
2
ft
8
-C
o
100
® 10
m
¦a
03
t
03
tfl
-Q
o
100
100
¦D
o
o
m
~o
Hi
£
lEUBK-Predicted Mean Blood Lead (ijg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (iig'dL)
100
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide
100
TJ
O
o
m
"O
©
£
o
(A
J2
o
100
10
£

M
JQ
O
o /
o /
o o o/Qd
OQ / O
o sf o o
ts o o /
o o So oo
O O / O Ol
o o o or oop Qo o
3
o o o o
o
o o o
1 o
o o
o o ow coo o
oofidSxo ooo ooooo aocc' a
o oa/o coa^fi (togfeao o*
o o /
0 <*X> 0 0/ 0
/o° «°0 ° •'
/f O c CO % g" o o
DOO /
/ o
/ y
°W°0 /
0
o e**™ "b1 / o °
00 O® 00 00 <¥ O O 0
/
0	o
1	9 0 0
a
100
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
t
©
jfl
£
o
10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide

-------
Figure F-6. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Box
100
"O
o
o
m
-o
©
£
©
1/1
o
100
100
10
m
-a
©
£
©
¥1
JD
o
100
lEUBK-Predicted Mean Blood Lead (pg/dL}
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
* 10
m
¦o
©
£
03
M
_Q
o
lEUBK-Predicted Mean Blood Lead (|jg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Box
B) IEUBK v2.0 EFH IRs, 55/45 partition - Box
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Box
100
m
TJ
©
£
©
V?
-O
O
100
100
© 10
¦O
o
o
m
Xf
OJ
£
©
(A
ja
O
O /
0 /
/
o o o/£> <
oo y o
0 / oo
6 O /
o o
00 0 0
0
00 0
O O O Or O® qO O
0 0
e o y tibci oao ooa
o omoaao ooo aoooo nomo  0 0/ 0
O O a/oODHKDaMOB «ni»» o<
O 0 iL00 °
-
/J«"<» " U ° «w
•ami 0 /
So 00 0
• % •>« "/ °
0 0
OO O OO oo /
OO 0
O V O OO OO • /
OO OOD O O O
— in re00 0 cd

lEUBK-Predicted Mean Blood Lead ((jg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
100
"D
6
o
¦O
©
£
100
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Box
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Box
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Box

-------
Figure F-7. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Basin
100
100
100
100
100
lEUBK-Predicted Mean Blood Lead (|jg/dL)
lEUBK-Predicted Mean Blood Lead (|jg/cll_)
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Basin
B) IEUBK v2.0 EFH IRs, 55/45 partition - Basin
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Basin
100
100
10
0 ©AO
oo° og- cF°go a® oo?
e0OOo«f £> & O wO O O O'
V. "o®" S0080 100 v 00
08 o 08 o cfbo o°
~ -0-00 0S0
lEUBK-Predicted Mean Blood Lead {\igldL\
lEUBK-Predicted Mean Blood Lead (|jg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
100
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Basin
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Basin
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Basin

-------
Figure F-8. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Site-wide, age <1 year
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (iig/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age
<1 year
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age
<1 year
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age
<1 year
10
lEUBK-Predicted Mean Blood Lead (MU'dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
IEUBK-P re dieted Mean Blood Lead (|ig'dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
<1 year
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
<1 year
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age
<1 year

-------
Figure F-9. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Site-wide, age 1 year
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age
1 year
100
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age
1 year
100
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age
1 year
100
10
lEUBK-Predicted Mean Blood Lead (Mg/dL)
lEUBK-Predicted Mean Blood Lead (Mg'dL)
10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
D)
10
lEUBK-Predicted Mean Blood Lead (gg/dL)
IEUBK v2.0 v2.0 IRs, 40/30/30 partition
1 year
- Site-wide Age
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
1year
1	10
lEUBK-Predicted Mean Blood Lead (Mg/dL)
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition
1 year
- Site-wide Age

-------
Figure F-10. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Site-wide, age 2 years
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age	B) IEUBKv2.0 EFH IRs, 55/45 partition - Site-wide Age 2 years	C) IEUBK vl.l v2.0 IRs, 55/45 partition - Site-wide Age 2 years
2 years
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (|jg/0oooflp& ®o 
°o°a 0 0
' • B %oa
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition
years
- Site-wide Age 2

-------
Figure F-11. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Site-wide, age 3 years
100
m
•a
13
2
o
tf)
A
o
10
O /

o /

° y
O O
/ 0
o
o
/ o
/
/ ® o e

o o o jr o o
o /o
O oat(So ° O o ° oo
0
/ o
/
o / a oo oo o o
jf o o
o /
/a cafaxao oocmb o
f ° 0 °oo° 0 °
88o o c8§° dBoo o o /
°°oo /
a o° 00 ° o ° ° /
0 o 0 /
o o
0°00 O 8 O oo

o /

Q O O DO /
ooo s/
a
CD
"O
©
2
Hi
W
o
100.0
m
T3
05
t
03
o
10.0
10
lEUBK-Predicted Mean Blood Lead (pg/'dL)
100
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
100
10.00
lEUBK-Predicted Mean Blood Lead (Mg'dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age
3 years
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age
3 years
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age
3 years
100
100
ffl
¦o
©
£
8
-C
O
o°°p °
o° ° °o o oo° 8 3 o« §
» » 0
O.O
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
100
® 10
"D
o
o
m
TJ
0)
t
5
M
-C
o
0 /
o/
o / o
s °
/ 0
o o
o
o
/ o o oo
/
/o o oo o
o o/
S o oo o®oo % o
0 /
/ o
/
/ 0 0 0 oo°o<» OO 0
0 /
A ° °
/
O • ^b8o® ow oo .
' o
• ° °°a /

o°° % o m°^8 o« 8 /oo
0
00° ° /

% o oo y

° ° 8 /

\> °° • OOD /# o

O O O CD /

o o /

OO 0 / 0

•
1
\i
o
o
0

100
10
lEUBK-Predicted Mean Blood Lead (|jg/dL)
100
10
o
O
CD
2
qj
0
-O
O

/
/
O

0 0
0 0
/ 9
/ o
o
o
/ o o oo

j6 o oo o
y% o o »oo % o
/ o
A> o oco o a o
°A 0
,/ o OO ® fcb° WD OUJO oo ,
°° ° 0°°° /
8 ° S O OO (PS o® 8 / o o
° /
O °o° ° 8° 7
o o/
o /
a /
o
o
O O
o o° fi^0® y© o o

o o oo o /
OO O / 0
o o oooo emxmo A a a a

10
lEUBK-Predicted Mean Blood Lead (pg/dL)
100
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age 3
years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
3years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age 3
years

-------
Figure F-12. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Site-wide, age 4 years
100
100
® 10
12
o
100
100
lEUBK-Predicted Mean Blood Lead (MS'dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age
4 years
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age
4 years
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age
4 years
100
too
o
to
100
100
lEUBK-Predicted Mean Blood Lead ((jg/dL)
lEUBK-Predicted Mean Blood Lead (}jg/cJL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
4 years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
4years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age
4 years

-------
Figure F-13. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Site-wide, age 5 years
100
m
T3
©
£

£
(A
J2
o
100
m
T3
05
£
03
in
Si
o
lEUBK-Predicted Mean Blood Lead (Mg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
100
lEUBK-Predicted Mean Blood Lead (Mg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age
5 years
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age
5 years
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age
5 years
100
m
¦o

£
\A
M
o
100
£
0>
i
M
O
100
o
o
m
J2
o
100
lEUBK-Predicted Mean Blood Lead ((jg/dL)
lEUBK-Predicted Mean Blood Lead (|jg/dL)
lEUBK-Predicted Mean Blood Lead {yig/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
5 years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
5 years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age
5 years

-------
Figure F-14. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Site-wide, age 6 years
100
100
100
100
lEUBK-Predicted Mean Blood Lead (|jg/dL)
lEUBK-Predicted Mean Blood Lead (|jg/cll_)
lEUBK-Predicted Mean Blood Lead (Mg'dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age
6 years
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age
6 years
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age
6 years
100

100
100
100
100
lEUBK-Predicted Mean Blood Lead (|Jt|/dL)
lEUBK-Predicted Mean Blood Lead (|ig/dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
6 years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
6years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age
6 years

-------
Figure F-15. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Site-wide Ages 2-6 years
lEUBK-Predicted Mean Blood Lead (pg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Ages 2-6
years
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Ages 2-6 years C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Ages 2-6 years
10


jf
oo
o
o
/o
O 0/ o
/ o
/ O0o
oo o
1 o
o oo/ooooax*
o%o • oo ooo
oo ooo ooo ooo
0t&m txfp
o o /
0 /
oO o/

(L0 O ©
SpaABUJlUBDOOOO o
V"*- °j


° o°° /



»o °
ooo o oooo co co ao
0° A°W
o /
/ o
o o
0


10
lEUBK-Predicted Mean Blood Lead (pg/dL)
o /
OO
/
O O /
O O O/ o
O / O
o O Or ®00
0
o o o ay o°a> oo oo 0®
£6o0 aao oo o co ooo ood
o OO ^WOOCIDq 11^*111 o
0
00 /
0 /
0% y
° ° g0°» O •
0°» /
/ 008 8o0g8gxi/tfCjL 0
oo °oo°8 9oo
00 00
° 80080880<5^8 S»o° o/°
0 0 00 7L 	/ ©
000000 000 0
® O OO
a


lEUBK-Predicted Mean Blood Lead (|jg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Ages 2-6
years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Ages 2-6
years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Ages 2-6 years

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix G
Censor Level 2 (<30.5 |jg/dL and No BDL) Additional Tables and Figures
G-1

-------
Table G-1. Summary of mean observed and predicted BLLs for IEUBK model v1.1 build 11 and censor level 2 (dataset omits predicted and observed BLLs >30.5 pg/dL and observed BDL BLLs)





Predicted Blood Lead




55/45 Dust/Soil Partition

40/30/30 Dust/Soil* Partition

Observed Blood Lead


IEUBK v1.1


IEUBK *1.1



v2.0 IR
SD
*1.1 IR
SD
v2.0 IRsd *1.1 IR
SD
Geometric 95% CI of the
Mean Geometric
N (|jg/dL) Mean (pg/dL)
Geometric
Standard
Deviation
(pg/dL)
Geometric
Mean
N (pg/dL)
95% CI of the
Geometric
Mean (pg/dL) Nf
Geometric
Mean
(pg/dL)
95% CI of the
Geometric
Mean (pg/dL) N
Geometric 95% CI of the Geometric
Mean Geometric Mean
(pg/dL) Mean (pg/dL) N+ (pg/dL)
95% CI of the
Geometric
Mean (|jg/dL)
Site-Wide (Total)
1144
3.6
3.5, 3.8
1.80
1144
3.0
2.9, 3.2
1138
4.3
4.2, 4.5
1144
3.9
3.8, 4.1
1142
5.7
5.5, 5.9
Box (Total)
875
3.9
3.7, 4.0
1.84
875
3.1
3.0, 3.3
869
4.4
4.2, 4.6
875
4.1
4.0, 4.3
873
6.0
5.8, 6.2
Basin (Total)
269
3.0
2.8, 3.1
1.58
269
2.8
2.6, 3.0
269
4.0
3.7, 4.3
269
3.5
3.2, 3.7
269
5.0
4.7, 5.3
Upper Basin (Total)
209
2.9
2.8, 3.1
1.54
209
2.9
2.7, 3.2
209
4.2
3.8, 4.6
209
3.7
3.4, 3.9
209
5.4
5.1, 5.8
Lower Basin (Total)
60
3.1
2.7, 3.6
1.70
60
2.5
2.1, 3.0
60
3.4
2.9, 4.0
60
2.8
2.4, 3.3
60
3.8
3.4, 4.4
Site-Wide By Age (years)
















0.5 to <1
70
3.8
3.4, 4.4
1.71
70
4.9
4.3, 5.7
70
4.9
4.2, 5.7
70
6.4
5.7, 7.2
70
6.3
5.7, 7.1
1 to <2
138
4.6
4.2, 5.1
1.74
138
4.7
4.3, 5.2
136
5.9
5.3, 6.5
138
6.3
5.8, 6.9
137
8.0
7.4, 8.7
2 to <3
202
4.0
3.6, 4.3
1.82
202
3.2
2.9, 3.4
202
4.9
4.6, 5.4
202
4.0
3.7, 4.3
202
6.4
6.0, 6.9
3 to <4
176
3.8
3.4, 4.1
1.79
176
3.0
2.7, 3.3
175
5.0
4.5, 5.5
176
3.9
3.6, 4.2
175
6.7
6.2, 7.2
4 to <5
187
3.5
3.2, 3.8
1.85
187
2.9
2.6, 3.1
186
4.0
3.7, 4.4
187
3.7
3.4, 4.0
187
5.3
4.9, 5.8
5 to <6
185
3.3
3.0, 3.6
1.75
185
2.6
2.4, 2.9
184
3.7
3.4, 4.1
185
3.4
3.2, 3.7
185
5.0
4.7, 5.4
6 to <7
186
3.1
2.8, 3.3
1.73
186
2.2
2.1, 2.4
185
3.0
2.8, 3.3
186
2.9
2.7, 3.1
186
4.0
3.8, 4.3
Box (Total) by Age (years)
















0.5 to <1
47
3.9
3.3, 4.6
1.73
47
4.8
4.0, 5.8
47
4.8
4.0, 5.7
47
6.5
5.7, 7.5
47
6.5
5.6, 7.5
1 to <2
112
4.9
4.4, 5.4
1.73
112
5.0
4.5, 5.6
110
6.1
5.5, 6.8
112
6.6
6.0, 7.3
111
8.4
7.6, 9.2
2 to <3
149
4.3
3.9, 4.8
1.82
149
3.3
3.0, 3.6
149
5.2
4.7, 5.7
149
4.2
3.9, 4.6
149
6.8
6.2, 7.4
3 to <4
137
4.0
3.6, 4.4
1.80
137
3.1
2.8, 3.4
136
5.1
4.6, 5.7
137
4.0
3.6, 4.4
136
6.9
6.3, 7.5
4 to <5
144
3.7
3.3, 4.2
1.93
144
2.8
2.6, 3.1
143
4.0
3.6, 4.4
144
3.8
3.5, 4.1
144
5.5
5.0, 6.0
5 to <6
144
3.4
3.1, 3.8
1.82
144
2.7
2.4, 2.9
143
3.8
3.4, 4.1
144
3.6
3.3, 3.9
144
5.2
4.8, 5.7
6 to <7
142
3.3
3.0, 3.6
1.80
142
2.3
2.1, 2.5
141
3.1
2.9, 3.4
142
3.0
2.8, 3.3
142
4.3
3.9, 4.6
Basin (Total) by Age (years)
















0.5 to <1
23
3.8
3.0, 4.7
1.69
23
5.2
4.0, 6.8
23
5.2
3.9, 6.7
23
6.1
4.9, 7.5
23
6.1
4.9, 7.5
1 to <2
26
3.6
2.9, 4.4
1.67
26
3.8
3.0, 4.8
26
4.8
3.7, 6.1
26
5.2
4.3, 6.2
26
6.6
5.6, 7.9
2 to <3
53
3.1
2.7, 3.6
1.70
53
2.8
2.4, 3.2
53
4.3
3.7, 5.1
53
3.5
3.1, 3.9
53
5.6
5.0, 6.3
3 to <4
39
3.0
2.5, 3.5
1.65
39
2.7
2.2, 3.2
39
4.5
3.7, 5.5
39
3.5
3.0, 4.0
39
6.1
5.2, 7.1
4 to <5
43
2.7
2.4, 3.0
1.42
43
2.9
2.3, 3.6
43
4.1
3.3, 5.2
43
3.3
2.8, 3.9
43
4.8
4.0, 5.7
5 to <6
41
2.8
2.5, 3.2
1.44
41
2.6
2.1, 3.1
41
3.7
3.0, 4.5
41
3.0
2.6, 3.5
41
4.3
3.7, 5.1
6 to <7
44
2.5
2.3, 2.8
1.38
44
2.0
1.7, 2.3
44
2.7
2.3, 3.2
44
2.4
2.2, 2.7
44
3.4
3.0, 3.9
Box (Total) by Calendar Year
















1995
99
4.9
4.4, 5.5
1.78
99
4.1
3.5, 4.8
97
5.7
4.9, 6.6
99
6.5
5.9, 7.3
98
9.5
8.6, 10.5
1996
102
4.9
4.3, 5.6
1.88
102
3.6
3.1,4.1
102
5.1
4.5, 5.9
102
5.7
5.2, 6.3
102
8.4
7.6, 9.3
1997
69
4.6
4.0, 5.3
1.82
69
3.2
2.8, 3.8
68
4.6
4.0, 5.3
69
4.8
4.2, 5.4
69
7.2
6.4, 8.2
1998
100
4.3
3.8, 4.9
1.90
100
3.6
3.2, 4.0
100
5.3
4.7, 5.9
100
4.7
4.3, 5.1
100
7.0
6.4, 7.6
1999
115
4.2
3.7, 4.7
1.79
115
3.5
3.1, 3.9
113
4.8
4.4, 5.3
115
4.3
3.9, 4.7
114
6.2
5.7, 6.7
2000
90
4.0
3.5, 4.5
1.85
90
3.1
2.7, 3.5
89
4.2
3.8, 4.8
90
3.8
3.4, 4.2
90
5.5
5.0, 6.1
2001
81
3.2
2.8, 3.7
1.81
81
2.7
2.4, 3.0
81
3.9
3.4, 4.4
81
3.8
3.4, 4.1
81
5.5
5.0, 6.0
2002
78
2.6
2.4, 2.9
1.54
78
2.5
2.2, 2.8
78
3.4
3.0, 3.9
78
3.2
2.9, 3.5
78
4.6
4.2, 5.1
2003-2008
10
4.1
2.5, 6.9
2.04
10
3.3
1.8, 6.1
10
4.7
2.5, 8.6
10
3.4
1.9, 6.1
10
4.9
2.8, 8.7
2013
107
2.8
2.6, 3.1
1.52
107
2.5
2.2, 2.7
107
3.5
3.2, 3.8
107
2.5
2.3, 2.7
107
3.5
3.2, 3.7
2018
24
3.5
2.8, 4.4
1.69
24
1.9
1.6, 2.2
24
2.6
2.2, 3.2
24
2.0
1.7, 2.2
24
2.8
2.4, 3.2
Basin (Total) by Calendar Year
















2002-2005
24
4.0
3.2, 5.0
1.71
24
4.8
3.4, 6.7
24
6.6
4.8, 9.2
24
5.8
4.5, 7.5
24
8.2
6.5, 10.5
2006
25
3.2
2.6, 3.8
1.56
25
4.1
2.9, 5.7
25
5.7
4.1, 7.9
25
5.0
4.0, 6.3
25
7.2
5.9, 8.9
2007
25
3.1
2.5, 3.9
1.68
25
2.9
2.1, 3.9
25
4.0
2.9, 5.5
25
4.1
3.3, 5.1
25
5.9
4.7, 7.4
2008
21
2.3
2.0, 2.7
1.34
21
2.3
1.8, 2.9
21
3.3
2.6, 4.1
21
2.9
2.4, 3.5
21
4.3
3.6, 5.2
2009
57
3.0
2.8, 3.3
1.40
57
2.4
2.1, 2.7
57
3.4
3.0, 3.9
57
3.2
2.9, 3.6
57
4.8
4.3, 5.3
2010
32
2.2
1.9, 2.5
1.47
32
2.5
2.2, 2.9
32
3.7
3.2, 4.2
32
3.0
2.7, 3.5
32
4.5
3.9, 5.1
2011
9
3.7
2.8, 4.8
1.42
9
3.0
2.2, 4.3
9
4.5
3.1, 6.5
9
3.8
3.0, 4.7
9
5.7
4.5, 7.1
2013
29
2.8
2.4, 3.2
1.48
29
2.5
2.0, 3.1
29
3.4
2.7, 4.3
29
2.8
2.3, 3.3
29
3.9
3.2, 4.6
2015
27
2.6
2.2, 3.2
1.62
27
3.0
2.3, 4.0
27
4.3
3.2, 5.7
27
3.1
2.5, 4.0
27
4.5
3.6, 5.7
2017
16
4.2
3.1, 5.7
1.74
16
2.0
1.7, 2.5
16
2.8
2.3, 3.5
16
2.3
1.9, 2.7
16
3.2
2.7, 3.8
2018
4
3.2
1.6, 6.6
1.58
4
4.1
1.4, 11.5
4
6.0
1.9, 19.0
4
3.9
1.8, 8.5
4
5.8
2.4, 13.9
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
+ The observed BLL summary statistics correspond to six model runs and not the IEUBK v1.1, v1.1 IRs, both partition model runs. By censoring the data sets, the IEUBK v1.1, v1.1 IRs for both partitions result in fewer BLL records due to predicted BLLs >30 ng/dL.
IEUBK v1.1 model runs used default parameters except for IRSd as shown. See Appendix A and Table 3 for more information on model parameters and configurations.

-------
Table G-2. Summary of observed and predicted average probability of exceeding 5 ng/dL for IEUBK model v1.1 build 11 and censor level 2 (dataset omits predicted and observed BLLs >30.5 pg/dL and observed BDL BLLs)














Predicted Blood Lead















55/45 Dust/Soil Partition







40/30/30 Dust/Soil* Partition




UDservea Biooa i_eaa





IEUBK v1.1




IEUBK v2.0




IEUBK v1.1










v2.0 IRsd


v1.1 IRsd

ft 2017b) So


v2.0 IRsd



v1.1 IRsd


Average
Percent of
95% CI for

Average
Predicted
95% CI for


Average
Predicted
95% CI for


Average
Predicted
95% CI for


Average
Predicted
95% CI for


BLLs >5 pg/dL Percent BLLs

Probabilities
Probabilities
Standard

Probabilities
Probabilities
Standard

Probabilities
Probabilities
Standard

Probabilities
Probabilities

N
(%)
>5 pg/dL (%)
N
>5 pg/dL (%)
>5 pg/dL (%)
Error (%)
N+
>5 pg/dL (%)
>5 pg/dL (%)
Error (%)
N
>5 pg/dL (%)
>5 pg/dL (%)
Error (%)
N+
>5 pg/dL (%)
>5 pg/dL (%)
Site-Wide (Total)
1144
32.1
29.4
34.8
1144
23.9
21.4
26.3
0.87
1138
39.6
36.8
42.5
0.86
1144
35.8
33.0
38.6
0.91
1142
56.5
53.6
59.4
Box (Total)
875
37.9
34.7
41.2
875
24.5
21.7
27.4
1.00
869
40.6
37.3
43.9
1.00
875
38.0
34.8
41.2
1.05
873
58.8
55.6
62.1
Basin (Total)
269
13.0
9.0-
17.0
269
21.8
16.9
26.7
1.74
269
36.5
30.8
42.3
1.63
269
28.6
23.2
34.0
1.73
269
48.9
42.9
54.8
Upper Basin (Total)
209
12.0
7.6-
16.4
209
22.7
17.1
28.4
1.97
209
39.1
32.5
45.7
1.85
209
30.9
24.6
37.1
1.97
209
53.3
46.6
60.1
Lower Basin (Total)
60
16.7
7.2-
26.1
60
18.5
8.7-
28.4
3.69
60
27.6
16.3
38.9
3.18
60
20.8
10.6
31.1
3.45
60
33.3
21.3
45.2
Site-Wide By Age (years)
0.5 to <1
70
30.0
19.3
40.7
70
45.7
34.0
57.4
3.90
70
45.3
33.6
56.9
3.46
70
62.8
51.5
74.1
3.23
70
62.3
50.9
73.6
1 to <2
138
45.7
37.3
54.0
138
44.3
36.0
52.6
2.65
136
57.1
48.8
65.5
2.28
138
62.6
54.5
70.6
2.35
137
75.5
68.3
82.7
2 to <3
202
36.6
30.0
43.3
202
24.7
18.7
30.6
1.97
202
47.5
40.6
54.4
1.95
202
36.1
29.5
42.7
2.04
202
63.5
56.9
70.2
3 to <4
176
33.5
26.5
40.5
176
22.2
16.1
28.4
2.21
175
46.1
38.8
53.5
2.16
176
34.0
27.0
41.0
2.24
175
64.7
57.6
71.8
4 to <5
187
30.5
23.9
37.1
187
20.1
14.4
25.9
2.01
186
35.4
28.5
42.2
2.00
187
31.7
25.1
38.4
2.10
187
52.6
45.4
59.7
5 to <6
185
26.5
20.1
32.8
185
17.9
12.4
23.4
1.87
184
32.7
25.9
39.5
2.01
185
28.3
21.8
34.8
1.98
185
49.3
42.1
56.5
6 to <7
186
23.7
17.5
29.8
186
10.9
64-
15.3
1.54
185
21.1
15.2
26.9
1.76
186
18.7
13.1
24.4
1.66
186
36.0
29.1
42.9
Box (Total) by Age (years)























0.5 to <1
47
34.0
20.5
47.6
47
44.5
30.3
58.7
4.73
47
44.0
29.8
58.2
4.16
47
64.4
50.7
78.1
3.87
47
63.9
50.1
77.6
1 to <2
112
50.0
40.7
59.3
112
47.0
37.8
56.3
2.96
110
59.8
50.6
68.9
2.55
112
65.2
56.4
74.1
2.62
111
77.5
69.7
85.2
2 to <3
149
43.0
35.0
50.9
149
26.8
19.7
33.9
2.35
149
50.3
42.3
58.4
2.34
149
39.8
31.9
47.6
2.44
149
66.6
59.0
74.1
3 to <4
137
37.2
29.1
45.3
137
23.0
16.0
30.1
2.53
136
47.3
38.9
55.7
2.50
137
35.7
27.7
43.7
2.59
136
66.0
58.0
73.9
4 to <5
144
38.9
30.9
46.9
144
19.3
12.9
25.7
2.21
143
34.6
26.8
42.4
2.26
144
33.4
25.7
41.2
2.38
144
54.6
46.5
62.8
5 to <6
144
32.6
25.0
40.3
144
17.5
11.3
23.7
2.03
143
32.8
25.1
40.5
2.25
144
30.1
22.6
37.6
2.22
144
52.0
43.8
60.1
6 to <7
142
29.6
22.1
37.1
142
11.6
6.3-
16.9
1.90
141
21.7
14.9
28.5
2.11
142
20.9
14.2
27.6
2.02
142
38.9
30.9
46.9
Basin (Total) by Age (years)























0.5 to <1
23
21.7
4.9
38.6
23
48.2
27.8
68.6
7.00
23
47.8
27.3
68.2
6.28
23
59.5
39.5
79.6
5.91
23
59.0
38.9
79.1
1 to <2
26
26.9
9.9
44.0
26
32.8
14.7
50.8
5.56
26
46.0
26.9
65.2
4.55
26
51.1
31.8
70.3
4.83
26
67.3
49.3
85.4
2 to <3
53
18.9
8.3
29.4
53
18.8
8.3-
29.3
3.42
53
39.7
26.5
52.8
3.12
53
25.8
14.0
37.6
3.28
53
55.0
41.6
68.4
3 to <4
39
20.5
7.8
33.2
39
19.4
7.0-
31.8
4.55
39
42.0
26.5
57.5
4.16
39
28.0
13.9
42.1
4.38
39
60.3
45.0
75.7
4 to <5
43
2.3
0.4
12.1
43
22.9
10.3
35.5
4.68
43
38.0
23.5
52.5
4.20
43
26.0
12.9
39.1
4.38
43
45.5
30.7
60.4
5 to <6
41
4.9
1.3
16.1
41
19.3
7.2-
31.4
4.56
41
32.3
18.0
46.6
4.36
41
22.1
9.4-
34.8
4.25
41
39.8
24.8
54.8
6 to <7
44
4.5
13
15.1
44
8.5
0.3-
16.7
2.18
44
19.0
7.4-
30.6
2.72
44
11.7
2.2-
21.2
2.32
44
26.7
13.6
39.7
Box (Total) by Calendar Year
1995
99
56.6
46.8
66.3
99
36.1
26.7
45.6
3.59
97
51.9
41.9
61.8
2.69
99
63.9
54.4
73.3
2.87
98
82.9
75.4
90.3
1996
102
50.0
40.3
59.7
102
30.3
21.3
39.2
3.26
102
47.7
38.0
57.4
2.65
102
57.0
47.4
66.6
2.92
102
77.5
69.3
85.6
1997
69
55.1
43.3
66.8
69
26.2
15.9
36.6
3.58
68
43.5
31.7
55.3
3.45
69
47.2
35.4
59.0
3.62
69
70.0
59.2
80.8
1998
100
52.0
42.2
61.8
100
30.6
21.5
39.6
3.06
100
51.5
41.7
61.3
2.69
100
44.5
34.7
54.2
2.88
100
69.5
60.5
78.6
1999
115
43.5
34.4
52.5
115
28.1
19.9
36.3
2.72
113
45.7
36.6
54.9
2.31
115
38.6
29.7
47.5
2.57
114
62.3
53.4
71.2
2000
90
46.7
36.4
57.0
90
22.8
14.2
31.5
3.11
89
38.0
27.9
48.0
2.64
90
30.8
21.3
40.3
2.84
90
54.0
43.7
64.3
2001
81
23.5
14.2
32.7
81
18.3
9.9
26.8
2.63
81
34.8
24.4
45.1
2.41
81
31.8
21.7
42.0
2.72
81
56.0
45.2
66.8
2002
78
10.3
3.5-
17.0
78
14.8
6.9
22.7
2.83
78
26.6
16.8
36.4
2.74
78
23.2
13.8
32.6
2.97
78
42.6
31.7
53.6
2003-2008
10
30.0
1.6-
58.4
10
29.9
1.5
58.3
13.35
10
39.9
9.5-
70.2
12.39
10
33.4
4.1
62.6
12.78
10
44.9
14.1
75.7
2013
107
6.5
1.9-
11.2
107
14.5
7.8
21.2
1.94
107
28.6
20.1
37.2
1.49
107
12.8
6.4-
19.1
1.64
107
27.3
18.8
35.7
2018
24
25.0
7.7-
42.3
24
5.6
1.2
22.3
1.98
24
15.3
0.9-
29.7
1.29
24
4.7
0.9-
21.0
1.41
24
15.2
0.9-
29.6
Basin (Total) by Calendar Year
2002-2005
24
33.3
14.5
-52.2
24
45.7
25.7
-65.6
7.87
24
59.8
40.2
79.5
6.13
24
56.1
36.3
-76.0
6.63
24
74.3
56.8
91.8
2006
25
20.0
4.3
35.7
25
42.5
23.2
-61.9
7.40
25
57.4
38.1
76.8
6.08
25
49.4
29.8
-69.0
6.42
25
70.4
52.6
88.3
2007
25
24.0
7.3
40.7
25
26.9
9.5
44.2
6.24
25
40.6
21.3
59.8
5.50
25
40.1
20.9
-59.3
5.78
25
61.0
41.9
80.1
2008
21
0.0
0.0
-0.0
21
13.7
4.7
34.0
4.89
21
26.4
7.6-
45.3
4.12
21
19.3
2.4-
36.1
4.77
21
40.9
19.9
61.9
2009
57
10.5
2.6
18.5
57
13.0
4.3
21.7
2.17
57
29.5
17.7
41.4
2.63
57
23.8
12.7 ¦
-34.8
2.90
57
47.6
34.6
60.5
2010
32
3.1
0.6
15.7
32
12.8
1.2
24.4
2.71
32
31.3
15.2
47.3
3.05
32
20.1
6.2-
34.0
3.45
32
43.2
26.1
60.4
2011
9
11.1
2.0
43.5
9
22.0
6.2
54.6
6.50
9
46.3
13.7
78.9
5.26
9
29.7
10.1
-61.5
5.33
9
60.6
28.7
92.6
2013
29
6.9
1.9
22.0
29
16.0
2.7
29.4
4.85
29
26.9
10.7
43.0
4.36
29
17.9
4.0-
31.9
4.71
29
31.9
15.0
48.9
2015
27
3.7
0.7
18.3
27
24.5
8.2
40.7
6.43
27
36.3
18.1
54.4
5.78
27
24.2
8.1 -
40.4
6.05
27
39.5
21.0
57.9
2017
16
25.0
3.8
46.2
16
6.0
1.0
28.0
1.66
16
17.3
5.8-
41.5
1.40
16
7.7
1.6-
30.1
1.94
16
21.5
1.3-
41.6
2018
4
25.0
4.6
69.9
4
43.7
11.9
-81.7
16.29
4
62.1
21.7-
•90.7
13.02
4
37.1
9.0-
77.8
13.84
4
59.3
20.1 •
•89.4
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
"The observed BLL summary statistics correspond to six model runs and not the IEUBK v1.1, v1.1 IRs, both partition model runs. By censoring the data sets, the IEUBK v1.1, v1.1 IRs for both partitions result in fewer BLL records due to predicted BLLs >30 |ig/dL.
IEUBK v1.1 model runs used default parameters except for IRSd as shown. See Appendix A and Table 3 for more information on model parameters and configurations.

-------
Table G-3. Summary of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 (dataset omits predicted and observed BLLs >30.5 [iqldL and observed BDL BLLs)




55/45 Dust/Soil Partition




40/30/30 Dust/Soil* Partition






IEUBK v1.1






IEUBK v1.1





v2.0 IRsd


v1.1 IRsd


v2.0 IRsd


v1.1 IRsd



% Outside
% < Lower
% > Upper

% Outside
% < Lower
%> Upper

% Outside
% < Lower
% > Upper

% Outside
% < Lower
%> Upper

N
PI
PI
PI
N
PI
PI
PI
N
PI
PI
PI
N
PI
PI
PI
Site-Wide (Total)
1144
19%
6%
13%
1138
17%
13%
4%
1144
14%
8%
5%
1142
23%
22%
2%
Box (Total)
876
20%
5%
15%
869
17%
12%
4%
875
14%
9%
5%
873
23%
21%
1%
Basin (Total)
269
14%
7%
7%
269
19%
16%
3%
269
12%
8%
4%
269
25%
22%
2%
Upper Basin (Total)
209
11%
7%
5%
209
19%
16%
2%
209
11%
9%
2%
209
26%
25%
1%
Lower Basin (Total)
60
23%
7%
17%
60
20%
13%
7%
60
17%
5%
12%
60
18%
12%
7%
Site-Wide by Age (years)
















0.5 to <1
70
19%
11%
7%
70
19%
11%
7%
70
27%
23%
4%
70
27%
23%
4%
1 to <2
138
14%
8%
6%
136
13%
12%
1%
138
13%
11%
2%
137
26%
26%
0%
2 to <3
202
19%
3%
15%
202
22%
17%
5%
202
13%
5%
8%
202
26%
24%
2%
3 to <4
176
14%
3%
11%
175
18%
16%
2%
176
9%
6%
2%
175
26%
25%
1%
4 to <5
187
25%
9%
17%
186
19%
14%
5%
187
16%
9%
7%
187
23%
21%
2%
5 to <6
185
17%
6%
11%
184
14%
12%
2%
185
11%
8%
3%
185
22%
21%
1%
6 to <7
186
22%
4%
18%
185
15%
8%
7%
186
13%
6%
7%
186
16%
13%
3%
Box (Total) by Age (years)
















0.5 to <1
47
17%
11%
6%
47
17%
11%
6%
47
28%
26%
2%
47
28%
26%
2%
1 to <2
112
13%
8%
5%
110
13%
11%
2%
112
11%
9%
2%
111
23%
23%
0%
2 to <3
149
20%
3%
17%
149
21%
16%
5%
149
15%
7%
8%
149
25%
23%
2%
3 to <4
137
16%
3%
13%
136
17%
15%
2%
137
9%
7%
3%
136
24%
24%
1%
4 to <5
144
26%
7%
19%
143
20%
13%
6%
144
16%
8%
8%
144
22%
20%
1%
5 to <6
144
19%
6%
13%
143
13%
11%
2%
144
12%
8%
3%
144
24%
23%
1%
6 to <7
142
26%
5%
21%
141
16%
8%
8%
142
16%
8%
8%
142
18%
15%
3%
Basin (Total) by Age (years)
















0.5 to <1
23
22%
13%
9%
23
22%
13%
9%
23
26%
17%
9%
23
26%
17%
9%
1 to <2
26
15%
8%
8%
26
15%
15%
0%
26
23%
19%
4%
26
35%
35%
0%
2 to <3
53
15%
4%
11%
53
25%
19%
6%
53
9%
2%
8%
53
30%
26%
4%
3 to <4
39
8%
3%
5%
39
21%
21%
0%
39
5%
5%
0%
39
31%
31%
0%
4 to <5
43
21%
14%
7%
43
19%
16%
2%
43
16%
12%
5%
43
28%
26%
2%
5 to <6
41
12%
7%
5%
41
17%
15%
2%
41
10%
7%
2%
41
15%
15%
0%
6 to <7
44
9%
2%
7%
44
14%
9%
5%
44
5%
2%
2%
44
11%
9%
2%
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
Grey shaded/bolded text cells represent the lowest percentage of all 8 model runs, for each data category. Corresponding percentages above and below PLs are bold.
IEUBKvl .1 model runs used default parameters except for IRSd as shown. See Appendix A and Table 3 for more information on model parameters and configurations.
The number of records (N) for each model run corresponds to the same Ns shown in Tables G-1 and G-2.

-------
Table G-4. Summary of sum of squared differences for censor level 2 (dataset omits predicted and
observed BLLs >30.5 pg/dL and
observed BDL BLLs)

55/45 Dust/Soil Partition
40/30/30 Dust/Soil* Partition

IEUBK v1.1
IEUBK v1.1

v2.0 IRsd
v1.1 IRsd
v2.0 IRsd
v1.1 IRsd
Site-wide (Total)
601
577
487
706
Box (Total)
475
423
381
530
Basin (Total)
126
154
106
175
Upper Basin (Total)
91
121
77
145
Lower Basin (Total)
36
33
29
30
Site-Wide by Age (years)




0.5 to <1
36
36
45
44
1 to <2
55
61
58
87
2 to <3
114
118
90
140
3 to <4
79
80
61
115
4 to <5
116
109
86
122
5 to <6
91
82
70
105
6 to <7
111
91
78
93
Box (Total) by Age (years)




0.5 to <1
20
20
28
28
1 to <2
43
47
44
66
2 to <3
82
80
65
96
3 to <4
68
62
51
87
4 to <5
92
77
67
90
5 to <6
73
61
57
84
6 to <7
96
76
68
79
Basin (Total) by Age (years)




0.5 to <1
16
16
17
16
1 to <2
12
14
14
20
2 to <3
31
38
26
44
3 to <4
11
18
9
28
4 to <5
24
32
18
31
5 to <6
18
22
13
21
6 to <7
15
14
9
14
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
IEUBK v1.1 model runs used default parameters except for IRSD as shown. See Appendix A and Table 3 for more information on
model parameters and configurations.
Grey shaded/bolded text cells represent the lowest value of all 8 model runs, for each data category.
The number of records (N) for each model run corresponds to the same Ns shown in Tables G-1 and G-2.

-------
Table G-5. Percent differences <1 pg/dl_ between observed and predicted BLLs for censor level 2 (dataset
omits predicted and observed BLLs >30.5 |jg/dl_ and observed BDL BLLs)	

55/45 Dust/Soil Partition
40/30/30 Dust/Soil* Partition

IEUBK v1.1
IEUBK v1.1

v2.0 IRsd
v1.1 IRsd
v2.0 IRsd
v1.1 IRsd
Site-Wide (Total)
34%
31%
34%
22%
Box (Total)
32%
30%
31%
21%
Basin (Total)
41%
36%
45%
25%
Upper Basin (Total)
43%
36%
49%
22%
Lower Basin (Total)
33%
35%
33%
35%
Site-Wide By Age (years)




0.5 to <1
29%
29%
14%
14%
1 to <2
32%
23%
22%
15%
2 to <3
30%
27%
35%
15%
3 to <4
32%
29%
35%
18%
4 to <5
37%
28%
37%
22%
5 to <6
37%
38%
36%
30%
6 to <7
39%
43%
46%
32%
Box (Total) by Age (years)




0.5 to <1
34%
34%
17%
17%
1 to <2
33%
23%
21%
14%
2 to <3
26%
27%
30%
15%
3 to <4
30%
26%
31%
18%
4 to <5
31%
26%
34%
23%
5 to <6
36%
36%
33%
28%
6 to <7
38%
39%
41%
27%
Basin (Total) by Age (years)




0.5 to <1
17%
17%
9%
9%
1 to <2
27%
23%
27%
19%
2 to <3
43%
28%
51%
13%
3 to <4
38%
38%
46%
18%
4 to <5
56%
35%
47%
19%
5 to <6
41%
44%
49%
39%
6 to <7
43%
55%
64%
48%
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
IEUBK v1.1 model runs used default parameters except for IRSd as shown. See Appendix A and Table 3 for more information on model
parameters and configurations.
Grey shaded/bolded text cells represent the highest value of all 8 model runs, for each data category.
The number of records (N) for each model run corresponds to the same N's shown in Tables G-1 and G-2.

-------
Advancing Pb Exposure and Biokinetic Modeling
Table G-6. Summary of goodness-of-fit between observed and predicted BLLs for IEUBK v2.0 model, censor level 2 (dataset omits
predicted and observed BLLs >30.5 pg/dL and observed BDL BLLs)	


55/45 Dust/Soil Partition
40/30/30 Dust/Soil* Partition

v2.0 IRSd
EFH (USEPA 2017b) IRSD
v2.0 IRSd
EFH (USEPA 2017b) IRSD|

SSE/SST*
SSE/SST*
SSE/SST*
SSE/SST*
Site-Wide (Total)
1.79
1.72
1.49
1.40
Box (Total)
1.72
1.68
1.43
1.36
Basin (Total)
3.05
2.73
2.55
2.25
Upper Basin (Total)
3.71
3.32
3.17
2.78
Lower Basin (Total)
2.01
1.82
1.56
1.42
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
fSum of Squares Error (SSE) and Sum of Squares Total (SST)
Grey shading and bold text is the smallest SSE/SST ratio.

-------
Advancing Pb Exposure and Biokinetic Modeling
Table G-7. IEUBK v1.1 model output summary of observed BLLs recorded as BDLs










Average











Predicted
95% CI around




No. of records
Percent of


Minimum
Maximum
Probability of
Average
IEUBK Model and


No. of PBBs
where PRED
Records where
Mean (GM) of
95% CI
PRED
PRED
Exceeding 5
Probabilities
Dust/Soil Partition
IRsd
DL
BDL
BDL
PRED BDL
PRED (|jg/dL)
around GM
(ng/dL)
(Mg/dL)
ng/dL (%)
(%)
IEUBK v1.1
v2.0
1.4
109
15
14%
2.2
2.0, 2.4
0.9
14.4
10.6
4.8, 16.4


1.9
28
12
43%
2.1
1.8, 2.5
1.1
7.5
8.5
2.6, 24.5 t
55/45 dust/soil ingestion
V1.1
1.4
109
3
3%
3.0
2.8, 3.3
1.2
22.1
21.4
13.7, 29.1
weighting factor
1.9
28
5
18%
2.9
2.4, 3.5
1.4
7.5
20.7
5.6, 35.5
IEUBK v1.1
v2.0
1.4
109
4
4%
2.7
2.5, 2.9
1.2
11.7
17.4
10.2, 24.4


1.9
28
8
29%
2.3
2.0, 2.7
1.2
6.3
10.4
3.5, 26.6 t
40/30/30 dust/soil* ingestion
V1.1
1.4
109
0
0%
3.9
3.6, 4.2
1.6
18.4
34.0
25.1, 42.8
weighting factor
1.9
28
2
7%
3.2
2.8, 3.8
1.5
6.3
24.3
8.3, 40.0
Notes:
*40% dust, 30% property soil, 30% community soil.
"^Wilson's method used to approximate confidence intervals.
PRED = Predicted BLL
PBB = Observed BLL
DL = Detection Limit

-------
Figure G-1. Summary of mean observed and predicted blood lead levels for geographic areas for censor level 2
(omits predicted and observed BLLs > 30.5 pg/dL and observed BDL BLLs)
A) IEUBKv1.1, v2.0 IRs, 55/45 partition	B) IEUBK v1.1, v2.0 IRs, 40/30/30 partition
IEUBK v1.1 model runs used default parameters except for IRsd as shown. See Appendix A and Table 3 for more
information on model parameters and configurations.
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
Lower
Basin Site-wide
Upper
Basin
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Predicted Blood Lead GM (ng/dL)

-------
Figure G-2. Summary of mean observed and predicted blood lead levels for site-wide age groups for censor level 2
(omits predicted and observed BLLs > 30.5 pg/dL and observed BDL BLLs)
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
Predicted Blood Lead GM (ng/dL)
A) IEUBKv1.1, v2.0 IRs, 55/45 partition - Site-wide by Age
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
Predicted Blood Lead GM (ng/dL)
B) IEUBK v1.1, v2.0 IRs, 40/30/30 partition - Site-
wide by Age
IEUBK v1.1 model runs used default parameters except for IRsd as shown. See Appendix A and Table 3 for more
information on model parameters and configurations.

-------
Figure G-3. Summary of observed and predicted average probability of exceeding 5 pg/dL for the geographic areas for censor
level 2 (all predicted and observed BLLs <30.5 pg/dL and no BDLs)
Predicted P5 (%)
Predicted P5(%)
A) IEUBKv1.1, v2.0 IRs, 55/45 partition	B) IEUBK v1.1, v2.0 IRs, 40/30/30 partition
IEUBK v1.1 model runs used default parameters except for IRsd as shown. See Appendix A and Table 3 for more information
on model parameters and configurations.

-------
Figure G-4. Summary of observed and predicted average probability of exceeding 5 pg/dL for the site-wide age groups for
censor level 2 (all predicted and observed BLLs <30.5 pg/dL and no BDLs)
Predicted P5(%)
Predicted P5 (%)
A) IEUBK v1.1, v2.0 IRs, 55/45 partition - Site-wide by Age	B) IEUBK v 1.1, v2.0 IRs, 40/30/30 partition - Site-wide by
Age
IEUBK v1.1 model runs used default parameters except for IRsd as shown. See Appendix A and Table 3 for more information
on model parameters and configurations.

-------
Figure G-5. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 1 - Site-wide
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide
Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide
100 	
lEUBK-Predicted Mean Blood Lead (pg/dL)
lEUBK-Predicted Mean Blood Lead (Mg/dL)
10
lEUBK-Predicted Mean Blood Lead (Mg/dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (Mg'dL)
1C
lEUBK-Predicted Mean
D) IEUBK v2.0v2.0 IRs, 40/30/30 partition-Site-wide	E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide	F)
IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide

-------
Figure G-6. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 = Site-wide
A) IEUBKv2.0v2.0IRs, 55/45 partition-Site-wide	B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide	C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide	E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide	F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide

-------
Figure G-7. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 - Box
« /
o f
O O O OO/o

9 CO
O 00 O O O
O 9
0
O o o °oc*> o yo ®oooo 0
00 /
0 o a o mmGamDS D oocoxd o o
0 0 /
o oo oooorogmo opo® o a> o
yC 0
oooooceftooaoA ®oa>®o® ® o ®o oo
OOOO /
® ° O 0

o o o
OO >6
Pi ft ^

XJjV " . ft- y
S o o00 af D ° ®/
/ p, 4P COG' tfu O

o° 88080°„ a? n° „o y 0
OOO O 00
' •»•/ 0
O
0 o80alii8aSrffliiTPcman ootmSaa o ° e o
OO 0
# O 06 08 ® /

0 OO « « o 0 /



0 /
0 /
0 _ 0 0 _ 0 aSo
O
00 jr
00 /
0 a 00
O O o OO 0
O
O O OO 0
0 0 ,
0 yo
000 00 %c0 y3 arct> 0 co 00 y
0 O O OODC/OO fttOOOODOOO O OO 0 /
00 0 03 irmftiiiocyo co Qq 0 ao 0 0 &i A> 0
oao occo AMifb obco 0® 00 aooooo/
		Z
.o0£^£o_	°___ A °o...

/a00nflftic,wo O O n
8° 8K>*0o9S °°a° X 0
0° a>
00 0 0 OO 0
0
0 0
O O OO
og 0 0 00 00/
00 00 a> Voo
OOOO CO CD OD OO

lEUBK-Predicted Mean Blood Lead (pg'dL)
1	10
lEUBK-Predicted Mean Blood Lead (pg'dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Box
B) IEUBK v2.0 EFH IRs, 55/45 partition - Box
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Box
0 /
0 s

0 0 o/oa
00 7i
0 jf 00
oaf
0
0 y 0 o<
0 0 o n/ os Oo 0
O O
0 0 Jrafta * mo two
0 ggca» o oo * «do 

lEUBK-Predicted Mean Blood Lead {pg'dL}
lEUBK-Predicted Mean Blood Lead (pg/dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Box
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Box
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Box

-------
Figure G-8. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 - Basin
a BotPo oo
»o°OT§> o°
~ ¦ "oo . o. ° " an
® OqOo o " o%o ooo
f:8§s>»
a 8a«»ooo8°Vd®<>A8
o /

o X o

0 /

/ o

0 /
O-
o
a
¦\
0
/

o
o
o
o o
o
o oo'o cffc n> 8jo8c8oo / o

8° 0 oo 8 00 /o

o oo^ooaoo o o^o / o

O CO /

lEUBK-Predicted Mean Blood Lead (|jg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Basin
B) IEUBK v2.0 EFH IRs, 55/45 partition - Basin
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Basin
, /

/ 0

•/ *>
/ ° ..
/• o»* w
/ ° • °° o °
. . "»• y
o o o y ® ° B o 0 /
o ° o° o~t> ^"ogoo ooY ° Vl
0°S°® ^ ®o8°o • o o
o °H° 8 A /«,
o o «B o 5 o/' o
6 o oocoo ta as 00/0 e
o° oi° °s° / 0 0
0 e©»0 0000/ 0
*/
y
/ 00
0
0
0 00 /

10
lEUBK-Predicted Mean Blood Lead (pg/dL)
100
o° / 0

0 /
O / O °oo
/ • • %
o/y 0>#
/ O OO O gO 0 O
Q O gfWiOS 0O® 0 ,
°oc ^80° 0 0 /
O 0 °°ff? "jg *
8 <**<*" # Qo
00 ooi^afl^d 0 o'oo
0 0000 °/ O
000 0 / 0
0 00 / 0 0
0 /
/ 0
f 0 0
0
0
10
lEUBK-Predicted Mean Blood Lead (Mg/dL)
100
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Basin
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Basin
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Basin

-------
Figure G-9. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 - Site-wide, Age <1 year
lEUBK-Predicted Mean Blood Lead (jjg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (iig/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age <1 year
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age <1 year
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age <1 year
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (|jg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
<1 year
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
<1 year
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age
<1 year

-------
Figure G-10. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 - Site-wide, Age 1 year
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)


o/o
o

o
0° o
o /o
/ o o

/ o o oo

Of OO cP O O O
o / o
o ar o o o
o /
- / 00 oo ^cmxEo oo
/ o
yb oo o oo o „ tfi0 000 >
BO O
\
o
o
o
o
o

/ o o o /

' o o O 0D o oooo xSo
o O O rP /
oo ° o° /

o° o ° /
o o ° y o

o o /
o /

1	10
lEUBK-Predicted Mean Blood Lead (iig/dL)
1	10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age 1 year
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age 1 year
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age 1 year
10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
1 year
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
1year
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age 1
year

-------
Figure G-11. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 - Site-wide, Age 2 years
® 10


° , o /

° 0 °o°

° o°o • ° d>o //
0 ooooc&^p gco cm o a? o /
r.-i 0 00 • /0

e oo o ° o S o



® 10
lEUBK-Predicted Mean Blood Lead (pg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10	100
lEUBK-Predicted Mean Blood Lead (ijg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age 2 years
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age 2 years
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age 2 years
® 10
® 10
"O
c
o
m
¦o
£
•
n
o


o o /oo oo
O jS 00 OO «B0
/o . o ooooo* oo o
/ °° „°
0 o O OOB B^CO CO o
o^o a>cg>8o^m ^ — 00/
0 • % %o° /
° A
X
3 10
10	100
lEUBK-Predicted Mean Blood Lead ((jg/dL)
1	10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
1	10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
2 years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
2years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age
2 years

-------
Figure G-12. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 - Site-wide, Age 3 years
® 10
® 10
® 10
o /
° /
o q/
/ o
/ O
o o
o
o
/
o
o
: 8
\

(
i
o
o
o
o A
* n v n
o aaccar o o° oo
o /
o Q/6)aDO°o§b o o
o /
8arf8 qdood com o ° ,
y® a °<%c o ° «° o/
/ o
/QBo o to® o9oo o o /
/ OOno /
/ Oog OO O oo o ° /
o o
n 8 O ° /
0°00 CCT O O QE) Or

0 0° /

° O 8 ° °° O s

O 00 /

lEUBK-Predicted Mean Blood Lead (jjg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (iig/dL)
10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age 3 years
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age 3 years
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age 3 years
® 10
® 10
10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
3 years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
3years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age
3 years

-------
Figure G-13. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 - Site-wide, Age 4 years
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age 4 years	B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age 4 years	C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age 4 years
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10
lEUBK-Predicted Mean Blood Lead (|jg/dL)
10
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
4 years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
4years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age 4
years

-------
Figure G-14. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 - Site-wide, Age 5 years
® 10


° /
O
/ °

o / o o
/
o/ o
/
o y£a>0 oo ° o
/

V
/ o 0

0 jr o ocP oo oo o o o
o /
a6 °
/
o ogoofi 0oao oo o ° y
' o
/ o 0 /

oo o° oo Sboao oo%d o o y'

° 0° o ° o °0 /

8o°0o°0o ° // °
0
8o o^o oo o8oo8gpo o ®/4o
0

o /

. ...v ^

® 10
lEUBK-Predicted Mean Blood Lead (pg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age 5 years
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age 5 years
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age 5 years
® 10
"O
c
o
m
¦o
«
£
0!
¦a
-C
O
10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)

/



5
/
0
B
a
/ O

"O
a
¦
/• 000

*a
o
/ O OO °«D O
/
o
m
' "" „
/
"O
•
£
©
V)
°°°'*Z*' °°"° ' V

O
o o o afco8>o ooo oo ao /
0 0000 o 0 /
'

8 008 8s <&oo acamp/co o


o / o


/ ""


5 10




/ °

o o

/ 0 o

A oo °«o o

o /oo o » • .0
o /
o«/o o°rfb o o ooo o

j6 o °

/o oo °axxx$> CO MD ®o % o

/

ooo a£b afro o o ®o /

° ooo° °% o y

° 8g°8 ° ° 0 / °

8 008 8o <&o ® mx>c6 o
O
O 0 /

0 / o

o oo oo / oo o o

lEUBK-Predicted Mean Blood Lead (pg/dL)
1	10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
5 years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
5 years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age
5 years

-------
Figure G-15. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 - Site-wide, Age 6 years
® 10
° /

° /

o /

o o /

o
o

O 00 /
o
obud o
o /
>6 o

g oooo/ O OO 000(0 o

/ o

ooa/o ocamnoo o ,
/ o
o/o o o /

A> o /

o/o ooooo o o o o o o o/
o
/o° 00 ° o /

c£o ^ X

o
s
o
0
b
1
0
o
o
o
\
o
o
o

o oo o o /

oo o /

lEUBK-Predicted Mean Blood Lead (jjg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (iig/dL)
10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Age 6 years
B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Age 6 years
C) IEUBK v1.1 v2.0 IRs, 55/45 partition - Site-wide Age 6 years
® 10
• 10
10	100
lEUBK-Predicted Mean Blood Lead ((jg/dL)
1	10	100
lEUBK-Predicted Mean Blood Lead (jjg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Age
6 years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Age
6 years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Age
6 years

-------
Figure G-16. Scatterplots of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 2 - Site-wide ages 2-6 years
® 10
/
/
o o 00 OCDy^° ° oa»° °
oo o oo txapbxa too o A o o
ooo oaaoafocDO OOJOO to ° oaooo
o  sb
°° ° °
g 0oo$OB^bodxxaP  ow o m
o«/
8>o0o §o8°|o0 °o° o o- 0 /
/ Wo
o /c aBoeocoagSfcafawnqm amt> a omo/
^°o00oo8800 o°o° 00 y
o ° o °
o8ooPO§i8cLjA8a?-,of^o 0 a.
O O 00
o ° oo°o 00 0 ° X



lEUBK-Predicted Mean Blood Lead (pg/dL)
1	10
lEUBK-Predicted Mean Blood Lead (pg/dL)
10	100
lEUBK-Predicted Mean Blood Lead (|jg/dL)
A) IEUBK v2.0 v2.0 IRs, 55/45 partition - Site-wide Ages 2-6 years B) IEUBK v2.0 EFH IRs, 55/45 partition - Site-wide Ages 2-6 years C) IEUBK v1,1 v2.0 IRs, 55/45 partition - Site-wide Ages 2-6 years


O O O JK cfibo ® OO ooo
cos/coac oo oco ooo ooo
o 00/60 oooood0 maSiias
OO 0/
'O ° ^O O

oo / o
o o /» 0
0 O
o o a/a oanyoBcftm^ o
o o a/aPocamc nAuo %aotn> o
/ 0° e o
ttoo o/


X o"5 4bo°ol^ oo° % °o /

°°o ^o*4,0 r" ° 00



lEUBK-Predicted Mean Blood Lead (pg/dL)
1	10	100
lEUBK-Predicted Mean Blood Lead (pg/dL)
D) IEUBK v2.0 v2.0 IRs, 40/30/30 partition - Site-wide Ages
2-6 years
E) IEUBK v2.0 EFH IRs, 40/30/30 partition - Site-wide Ages
2-6 years
F) IEUBK v1.1 v2.0 IRs, 40/30/30 partition - Site-wide Ages 2-6
years

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix H
Censor Level 3 (<10.5 |jg/dL and No BDL) Tables and Figures
H-1

-------
Table H-1a. Summary of mean observed and predicted BLLs for IEUBK model version 2.0 and censor level 3 (dataset omits predicted and observed BLLs >10.5 pg/dL and observed BDL BLLs)












Predicted Blood Lead









ed Bloc




55/45 default dust/soil partition



40/30/30 Dust/Soil* Partition




ser\

ea




IEUBK v2.0





IEUBK v2.0










v2.0 IRsd


EFH (USEPA 2017b) IRsd

v2.0 IRsd


EFH (USEPA 2017b) IR
D





Geometric


















95% CI of the
Standard


95% CI of the


95% CI of the


95% CI of the


95% CI of the


Geometric
Geometric
Deviation

Geometric
Geometric

Geometric
Geometric

Geometric
Geometric

Geometric
Geometric

N
Mean (pg/dL)
Mean (pg/dL)
(pg/dL)
N
Mean (pg/dL)
Mean (pg/dL)
N
Mean (pg/dL)
Mean (pg/dL)
N
Mean (pg/dL)
Mean (pg/dL)
N
Mean (pg/dL)
Mean (pg/dL)
Site-Wide (Total)
1091
3.4
3.3
3.5
1.68
1091
3.3
3.2
3.4
1091
3.2
3.1, 3.3
1091
4.1
4.0
4.3
1091
4.0
3.9,
4.2
Box (Total)
826
3.6
3.5
3.7
1.71
826
3.3
3.2
3.4
826
3.2
3.1, 3.3
826
4.3
4.1
4.4
826
4.2
4.0,
4.3
Basin (Total)
265
2.9
2.8
3.1
1.53
265
3.2
3.0
3.4
265
3.1
2.9, 3.3
265
3.8
3.6
4.0
265
3.7
3.5,
3.9
Upper Basin (Total)
207
2.9
2.7
3.0
1.51
207
3.3
3.0
3.5
207
3.2
3.0, 3.4
207
4.0
3.8
4.3
207
3.9
3.7,
4.1
Lower Basin (Total)
58
3.0
2.6
34
1.60
58
2.9
2.5
3.4
58
2.8
2.4, 3.2
58
3.2
2.8
3.7
58
3.1
2.7,
3.5
Site-Wide by Age (years)




















0.5 to <1
67
3.6
3.2
4.1
1.61
67
4.7
4.1
5.4
67
4.1
3.6, 4.7
67
6.1
5.5
6.8
67
5.3
4.8,
5.9
1 to <2
124
4.1
3.8
4.4
1.56
124
4.7
4.3
5.1
124
4.5
4.1,4.9
124
6.2
5.8
6.8
124
6.0
5.5,
6.5
2 to <3
190
3.7
3.4
4.0
1.69
190
3.5
3.2
3.7
190
3.3
3.1, 3.5
190
4.3
4.0
4.6
190
4.0
3.8,
4.3
3 to <4
166
3.5
3.2
3.7
1.63
166
3.1
2.9
3.4
166
3.0
2.8, 3.3
166
4.0
3.7
4.3
166
3.8
3.6,
4.1
4 to <5
181
3.3
3.0
3.6
1.77
181
3.1
2.9
3.4
181
2.9
2.7, 3.2
181
3.9
3.6
4.2
181
3.7
3.4,
3.9
5 to <6
180
3.1
2.9
3.4
1.67
180
3.0
2.7
3.2
180
3.0
2.8, 3.3
180
3.7
3.5
4.0
180
3.8
3.6,
4.1
6 to <7
183
3.0
2.8
3.2
1.68
183
2.6
24
2.7
183
2.7
2.5, 2.9
183
3.2
3.0
3.4
183
3.4
3.2,
3.6
Box (Total) by Age (years)




















0.5 to <1
45
3.7
3.2
4.3
1.64
45
4.4
3.8
5.2
45
3.9
3.3, 4.5
45
6.1
5.4
7.0
45
5.3
4.7,
6.0
1 to <2
98
4.2
3.9
4.6
1.52
98
4.8
4.4
5.4
98
4.7
4.2, 5.2
98
6.5
5.9
7.1
98
6.2
5.7,
6.8
2 to <3
139
4.0
3.6
4.4
1.69
139
3.6
3.3
3.9
139
3.4
3.1, 3.7
139
4.5
4.1
4.8
139
4.2
3.9,
4.5
3 to <4
128
3.7
3.4
4.0
1.63
128
3.2
2.9
3.5
128
3.1
2.8, 3.3
128
4.1
3.7
4.4
128
3.9
3.6,
4.3
4 to <5
138
3.5
3.2
3.9
1.84
138
3.0
2.8
3.3
138
2.9
2.7, 3.1
138
4.0
3.7
4.3
138
3.7
3.4,
4.0
5 to <6
139
3.2
3.0
3.5
1.72
139
2.9
2.7
3.2
139
3.0
2.8, 3.3
139
3.8
3.5
4.1
139
4.0
3.7,
4.3
6 to <7
139
3.2
2.9
3.5
1.74
139
2.6
24
2.9
139
2.8
2.6, 3.0
139
3.3
3.1
3.6
139
3.5
3.3,
3.8
Basin (Total) by Age (years)




















0.5 to <1
22
3.6
2.9
4.4
1.57
22
5.2
4.0
6.9
22
4.6
3.5, 6.0
22
6.1
5.0
7.6
22
5.3
4.3,
6.6
1 to <2
26
3.6
2.9
4.4
1.67
26
4.1
3.3
5.1
26
3.9
3.2, 4.9
26
5.4
4.6
6.4
26
5.2
4.4,
6.1
2 to <3
51
2.9
2.6
3.3
1.59
51
3.2
2.7
3.6
51
3.0
2.6, 3.4
51
3.9
3.5
4.3
51
3.6
3.3,
4.0
3 to <4
38
2.9
2.5
3.3
1.56
38
3.0
2.6
3.5
38
2.9
2.5, 3.4
38
3.8
3.3
4.3
38
3.6
3.2,
4.1
4 to <5
43
2.7
2.4
3.0
1.42
43
3.3
2.8
4.0
43
3.2
2.6, 3.8
43
3.7
3.2
4.3
43
3.5
3.0,
4.1
5 to <6
41
2.8
2.5
3.2
1.44
41
3.0
2.5
3.5
41
3.1
2.6, 3.7
41
3.4
2.9
3.9
41
3.5
3.0,
4.0
6 to <7
44
2.5
2.3
2.8
1.38
44
2.4
2.1
2.6
44
2.5
2.2, 2.8
44
2.8
2.5
3.1	
44
3.0
2.7,
3.3
Box (Total) by Calendar Year




















1995
89
4.4
4.0
4.9
1.64
89
3.9
3.4
4.5
89
3.9
3.4, 4.4
89
6.3
5.7
7.0
89
6.2
5.7,
6.8
1996
91
4.3
3.9
4.8
1.71
91
3.6
3.2
4.1
91
3.6
3.2, 4.0
91
5.7
5.2
6.2
91
5.6
5.1,
6.1
1997
65
4.3
3.8
4.9
1.71
65
3.4
3.0
3.9
65
3.3
2.9, 3.8
65
5.0
4.4
5.6
65
4.8
4.3,
5.4
1998
94
4.0
3.5
4.5
1.79
94
3.8
3.4
4.2
94
3.7
3.4, 4.1
94
4.9
4.5
5.4
94
4.8
4.4,
5.2
1999
109
3.9
3.5
4.3
1.70
109
3.7
3.3
4.1
109
3.6
3.3, 4.0
109
4.5
4.2
4.9
109
4.4
4.1,
4.8
2000
85
3.7
3.3
4.2
1.74
85
3.3
2.9
3.7
85
3.3
2.9, 3.7
85
4.0
3.6
4.4
85
4.0
3.6,
4.3
2001
78
3.0
2.7
3.4
1.69
78
3.0
2.7
3.3
78
2.9
2.6, 3.2
78
4.0
3.7
4.4
78
3.9
3.6,
4.2
2002
78
2.6
2.4
2.9
1.54
78
2.8
2.5
3.1
78
2.7
2.4, 3.0
78
3.5
3.2
3.8
78
3.4
3.1,
3.7
2003-2008
8
3.3
2.0
5.2
1.76
8
3.3
1.7
6.6
8
3.2
1.6, 6.3
8
3.2
1.7
6.0
8
3.1
1.7,
5.7
2013
105
2.7
2.6
2.9
1.42
105
2.7
2.5
2.9
105
2.6
2.4, 2.8
105
2.7
2.5
2.9
105
2.6
2.4,
2.7
2018
24
3.5
2.8
4.4
1.69
24
2.2
1.9
2.5
24
2.1
1.9, 2.4
24
2.2
2.0
2.5
24
2.2
2.0,
2.4
Basin (Total) by Calendar Year




















2002-2005
22
3.6
3.0
4.3
1.53
22
5.1
3.7
7.2
22
5.0
3.6, 6.9
22
6.3
4.9
8.0
22
6.1
4.8,
7.6
2006
25
3.2
2.6,
3.8
1.56
25
4.5
3.4,
6.1
25
4.4
3.3, 5.8
25
5.4
4.4,
6.6
25
5.2
4.2,
6.3
2007
25
3.1
2.5,
3.9
1.68
25
3.3
2.6,
4.2
25
3.2
2.5, 4.1
25
4.4
3.7,
5.3
25
4.3
3.5,
5.1
2008
21
2.3
2.0,
2.7
1.34
21
2.7
2.3,
3.3
21
2.7
2.3, 3.2
21
3.4
2.9,
3.9
21
3.3
2.9,
3.8
2009
57
3.0
2.8,
3.3
1.40
57
2.7
2.5,
3.0
57
2.7
2.4, 3.0
57
3.6
3.3,
3.9
57
3.5
3.2,
3.8
2010
32
2.2
1.9,
2.5
1.47
32
2.9
2.6,
3.2
32
2.8
2.5, 3.1
32
3.4
3.0,
3.8
32
3.3
3.0,
3.7
2011
9
3.7
2.8,
4.8
1.42
9
3.4
2.5,
4.5
9
3.3
2.5, 4.3
9
4.1
3.3,
4.9
9
3.9
3.2,
4.8
2013
29
2.8
2.4,
3.2
1.48
29
2.8
2.3,
3.5
29
2.8
2.3, 3.4
29
3.1
2.6,
3.6
29
3.0
2.6,
3.5
2015
27
2.6
2.2,
3.2
1.62
27
3.4
2.6,
4.3
27
3.2
2.5, 4.2
27
3.5
2.8,
4.3
27
3.3
2.7,
4.1
2017
14
3.6
2.9,
4.4
1.41
14
2.3
1.9,
2.7
14
2.2
1.9, 2.6
14
2.5
2.2,
2.9
14
2.4
2.1,
2.7
2018
4
3.2
1.6,
6.6
1.58
4
4.4
1.7,11.2 |
4
4.4
1.8, 10.8
4
4.2
2.1,
8.5
4
4.1
2.1,
8.3
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.

-------
Tabh^-lb^ummar^jnTieai^bsewec^incyanKlictec^LL^oMEUB^










Predicted Blood Lead










55/45 Default Dust/Soil Partition



40/30/30 Dust/Soil*
Partition



UDservea uiooa Leaa



IEUBK v1.1




IEUBK v1.1








v2.0IRsd

v1.1 IR
D

v2.0IRsd

v1.1 IR
D




Geometric















95% CI of the
Standard


95% CI of the


95% CI of the


95% CI of the


95% CI of the


Geometric
Geometric
Deviation

Geometric
Geometric

Geometric
Geometric

Geometric
Geometric

Geometric
Geometric

N
Mean (ug/dL)
Mean (ug/dL)
(ug/dL)
N
Mean (ug/dL)

N+
Mean (ug/dL)
Mean (ug/dL)
N
Mean (ug/dL)
Mean (ug/dL)
N+
Mean (ug/dL)
Mean (ug/dL)
Site-Wide (Total)
1091
3.4
3.3, 3.5
1.68
1091
3.0
2.8, 3.1
1086
4.2
4.0, 4.3
1091
3.8
3.7, 4.0
1090
5.6
5.4, 5.8
Box (Total)
826
3.6
3.5, 3.7
1.71
826
3.0
2.9, 3.1
821
4.2
4.1, 4.4
826
4.0
3.8, 4.1
825
5.8
5.6, 6.0
Basin (Total)
265
2.9
2.8, 3.1
1.53
265
2.8
2.6, 3.0
265
4.0
3.7, 4.3
265
3.5
3.2, 3.7
265
5.0
4.7, 5.3
Upper Basin (Total)
207
2.9
2.7, 3.0
1.51
207
2.9
2.7, 3.2
207
4.2
3.8, 4.6
207
3.7
3.4, 3.9
207
5.4
5.0, 5.8
Lower Basin (Total)
58
3.0
2.6, 3.4
1.60
58
2.5
2.1, 3.0
58
3.4
2.9, 4.0
58
2.8
2.4, 3.3
58
3.9
3.4, 4.4
Site-Wide by Age (years)
















0.5 to <1
67
3.6
3.2, 4.1
1.61
67
4.8
4.2, 5.5
67
4.8
4.1, 5.5
67
6.3
5.6, 7.0
67
6.2
5.6, 6.9
1 to <2
124
4.1
3.8, 4.4
1.56
124
4.4
4.0, 4.9
123
5.5
5.0, 6.1
124
6.0
5.5, 6.5
124
7.7
7.1, 8.4
2 to <3
190
3.7
3.4, 4.0
1.69
190
3.1
2.9, 3.4
190
4.9
4.5, 5.4
190
4.0
3.7, 4.3
190
6.4
6.0, 6.9
3 to <4
166
3.5
3.2, 3.7
1.63
166
2.8
2.6, 3.1
165
4.7
4.3, 5.1
166
3.7
3.4, 4.0
165
6.4
5.9, 6.9
4 to <5
181
3.3
3.0, 3.6
1.77
181
2.8
2.6, 3.0
180
3.9
3.6, 4.3
181
3.6
3.3, 3.9
181
5.2
4.8, 5.6
5 to <6
180
3.1
2.9, 3.4
1.67
180
2.6
2.4, 2.8
179
3.6
3.4, 4.0
180
3.4
3.1, 3.6
180
4.9
4.6, 5.3
6 to <7
183
3.0
2.8, 3.2
1.68
183
2.2
2.1, 2.4
182
3.0
2.8, 3.3
183
2.9
2.7, 3.1
183
4.0
3.7, 4.3
Box (Total) by Age (years)
















0.5 to <1
45
3.7
3.2, 4.3
1.64
45
4.5
3.9, 5.3
45
4.5
3.8, 5.3
45
6.3
5.5, 7.1
45
6.2
5.5, 7.1
1 to <2
98
4.2
3.9, 4.6
1.52
98
4.6
4.1, 5.1
97
5.7
5.2, 6.4
98
6.3
5.7, 6.9
98
8.0
7.3, 8.8
2 to <3
139
4.0
3.6, 4.4
1.69
139
3.3
3.0, 3.6
139
5.2
4.7, 5.7
139
4.2
3.8, 4.5
139
6.7
6.1, 7.3
3 to <4
128
3.7
3.4, 4.0
1.63
128
2.9
2.6, 3.2
127
4.8
4.3, 5.3
128
3.8
3.5, 4.2
127
6.5
6.0, 7.1
4 to <5
138
3.5
3.2, 3.9
1.84
138
2.7
2.5, 3.0
137
3.8
3.5, 4.2
138
3.7
3.4, 4.0
138
5.3
4.9, 5.8
5 to <6
139
3.2
3.0, 3.5
1.72
139
2.6
2.4, 2.8
138
3.6
3.3, 4.0
139
3.5
3.2, 3.8
139
5.1
4.7, 5.5
6 to <7
139
3.2
2.9, 3.5
1.74
139
2.3
2.1, 2.5
138
3.1
2.9, 3.4
139
3.0
2.8, 3.3
139
4.3
3.9, 4.6
Basin (Total) by Age (years)
















0.5 to <1
22
3.6
2.9, 4.4
1.57
22
5.3
4.1, 7.0
22
5.3
4.0, 7.0
22
6.3
5.1, 7.8
22
6.2
5.0, 7.7
1 to <2
26
3.6
2.9, 4.4
1.67
26
3.8
3.0, 4.8
26
4.8
3.7, 6.1
26
5.2
4.3, 6.2
26
6.6
5.6, 7.9
2 to <3
51
2.9
2.6, 3.3
1.59
51
2.8
2.4, 3.3
51
4.4
3.7, 5.2
51
3.5
3.1, 4.0
51
5.7
5.0, 6.4
3 to <4
38
2.9
2.5, 3.3
1.56
38
2.6
2.2, 3.1
38
4.4
3.6, 5.3
38
3.4
2.9, 3.9
38
5.9
5.1, 6.9
4 to <5
43
2.7
2.4, 3.0
1.42
43
2.9
2.3, 3.6
43
4.1
3.3, 5.2
43
3.3
2.8, 3.9
43
4.8
4.0, 5.7
5 to <6
41
2.8
2.5, 3.2
1.44
41
2.6
2.1, 3.1
41
3.7
3.0, 4.5
41
3.0
2.6, 3.5
41
4.3
3.7, 5.1
6 to <7
44
2.5
2.3, 2.8
1.38
44
2.0
1.7, 2.3
44
2.7
2.3, 3.2
44
2.4
2.2, 2.7
44
3.4
3.0, 3.9
Box (Total) by Calendar Year
















1995
89
4.4
4.0, 4.9
1.64
89
3.6
3.1, 4.1
88
5.1
4.5, 5.8
89
6.1
5.5, 6.7
89
9.0
8.1, 9.9
1996
91
4.3
3.9, 4.8
1.71
91
3.3
2.9, 3.8
91
4.8
4.1, 5.5
91
5.4
4.9, 5.9
91
8.0
7.2, 8.8
1997
65
4.3
3.8, 4.9
1.71
65
3.1
2.7, 3.6
64
4.5
3.9, 5.1
65
4.7
4.1, 5.3
65
7.1
6.2, 8.0
1998
94
4.0
3.5, 4.5
1.79
94
3.5
3.1, 3.9
94
5.1
4.6, 5.7
94
4.6
4.2, 5.1
94
6.9
6.3, 7.5
1999
109
3.9
3.5, 4.3
1.70
109
3.4
3.1, 3.8
107
4.7
4.2, 5.2
109
4.3
3.9, 4.7
108
6.1
5.6, 6.6
2000
85
3.7
3.3, 4.2
1.74
85
3.0
2.6, 3.4
84
4.1
3.7, 4.7
85
3.7
3.4, 4.1
85
5.4
4.9, 6.0
2001
78
3.0
2.7, 3.4
1.69
78
2.7
2.4, 3.0
78
3.8
3.4, 4.4
78
3.8
3.4, 4.1
78
5.5
5.0, 6.0
2002
78
2.6
2.4, 2.9
1.54
78
2.5
2.2, 2.8
78
3.4
3.0, 3.9
78
3.2
2.9, 3.5
78
4.6
4.2, 5.1
2003-2008
8
3.3
2.0, 5.2
1.76
8
3.0
1.4, 6.2
8
4.2
2.0, 8.9
8
2.9
1.5, 5.6
8
4.1
2.1, 8.1
2013
105
2.7
2.6, 2.9
1.42
105
2.4
2.2, 2.6
105
3.4
3.1, 3.7
105
2.4
2.2, 2.6
105
3.4
3.2, 3.7
2018
24
3.5
2.8, 4.4
1.69
24
1.9
1.6, 2.2
24
2.6
2.2, 3.2
24
2.0
1.7, 2.2
24
2.8
2.4, 3.2
Basin (Total) by Calendar Year
















2002-2005
22
3.6
3.0, 4.3
1.53
22
4.9
3.4, 6.9
22
6.6
4.7, 9.3
22
6.0
4.6, 7.8
22
8.4
6.6, 10.6
2006
25
3.2
2.6, 3.8
1.56
25
4.1
2.9, 5.7
25
5.7
4.1, 7.9
25
5.0
4.0, 6.3
25
7.2
5.9, 8.9
2007
25
3.1
2.5, 3.9
1.68
25
2.9
2.1, 3.9
25
4.0
2.9, 5.5
25
4.1
3.3, 5.1
25
5.9
4.7, 7.4
2008
21
2.3
2.0, 2.7
1.34
21
2.3
1.8, 2.9
21
3.3
2.6, 4.1
21
2.9
2.4, 3.5
21
4.3
3.6, 5.2
2009
57
3.0
2.8, 3.3
1.40
57
2.4
2.1, 2.7
57
3.4
3.0, 3.9
57
3.2
2.9, 3.6
57
4.8
4.3, 5.3
2010
32
2.2
1.9, 2.5
1.47
32
2.5
2.2, 2.9
32
3.7
3.2, 4.2
32
3.0
2.7, 3.5
32
4.5
3.9, 5.1
2011
9
3.7
2.8, 4.8
1.42
9
3.0
2.2, 4.3
9
4.5
3.1, 6.5
9
3.8
3.0, 4.7
9
5.7
4.5, 7.1
2013
29
2.8
2.4, 3.2
1.48
29
2.5
2.0, 3.1
29
3.4
2.7, 4.3
29
2.8
2.3, 3.3
29
3.9
3.2, 4.6
2015
27
2.6
2.2, 3.2
1.62
27
3.0
2.3, 4.0
27
4.3
3.2, 5.7
27
3.1
2.5, 4.0
27
4.5
3.6, 5.7
2017
14
3.6
2.9, 4.4
1.41
14
2.0
1.6, 2.5
14
2.8
2.2, 3.6
14
2.2
1.8, 2.7
14
3.2
2.6, 3.9
2018
4
3.2
1.6, 6.6
1.58
4
4.1
1.4, 11.5
4
6.0
1.9, 19.0
4
3.9
1.8, 8.5
4
5.8
2.4, 13.9
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
+The censored data sets for the 1EUBK v1.1, v1.1 IRs, both partitions result in fewer observed BLL records than presented in these columns. The observed BLL summary statistics correspond to the other six

-------
Table H-2a. Summary of observed and predicted average probability of exceeding 5 pg/dL for IEUBK model version 2.0 censor level 3 (dataset omits predicted and observed BLLs >10.5 pg/dL and
observed BDL BLLs)









Predicted Blood Lead







t

55/45 default dust/soil partition



40/30/30 Dust/Soil* Partition


Ob
Ob
served Bio
od Lead


IEUBK v2.0




IEUBK v2.0







v2.0 IRsd
EFH(USEPA 2017b) IRsd

v2.0 IRsd
EFH(USEPA 2017b) IRsd


Average
95% CI for

Average


Average


Average


Average



Percent
Average %

Predicted
95% CI for

Predicted
95% CI for

Predicted
95% CI for

Predicted
95% CI for


BLLs >5
BLLs >5

Probabilities
Probabilities

Probabilities
Probabilities

Probabilities
Probabilities

Probabilities
Probabilities

N
ng/dL (%)
pg/dL (%)
N
>5 pg/dL (%)
>5 pg/dL (%)
N
>5 pg/dL (%)
>5 pg/dL (%)
N
>5 pg/dL (%)
>5 pg/dL (%)
N
>5 pg/dL (%)
>5 pg/dL (%)
Site-Wide (Total)
1091
28.8
26.1-31.5
1091
25.1
22.5 - 27.7
1091
23.8
21.2-26.3
1091
37.7
34.8 - 40.5
1091
36.0
33.2 - 38.9
Box (Total)
826
34.3
31.0-37.5
826
25.3
22.3 - 28.2
826
24.0
21.0-26.9
826
39.4
36.0 - 42.7
826
37.8
34.5 - 41.2
Basin (Total)
265
11.7
7.8-15.6
265
24.5
19.4-29.7
265
23.1
18.1 -28.2
265
32.3
26.7 - 37.9
265
30.4
24.9 - 36.0
Upper Basin (Total)
207
11.1
6.8-15.4
207
25.5
19.6-31.5
207
24.3
18.4-30.1
207
34.7
28.2-41.1
207
33.0
26.6 - 39.4
Lower Basin (Total)
	58_
13.8
4.9 - 22.7
58
21.1
10.6-31.5
58
19.1
9.0-29.2
58
23.9
12.9-34.9
58
21.4
10.9-32.0
Site-Wide by Age (years)















0.5 to <1
67
26.9
16.3-37.5
67
43.5
31.6-55.3
67
36.1
24.6-47.6
67
61.1
49.4-72.8
67
52.1
40.2-64.1
1 to <2
124
39.5
30.9-48.1
124
43.9
35.2-52.6
124
41.5
32.8-50.1
124
62.8
54.3-71.3
124
60.2
51.6-68.8
2 to <3
190
32.6
26.0 - 39.3
190
28.0
21.6-34.4
190
25.2
19.0-31.3
190
39.7
32.7-46.6
190
36.0
29.1 -42.8
3 to <4
166
29.5
22.6 - 36.5
166
22.2
15.9-28.5
166
20.6
14.5-26.8
166
35.1
27.8 - 42.4
166
32.7
25.6-39.9
4 to <5
181
28.2
21.6 - 34.7
181
21.8
15.7-27.8
181
19.5
13.7-25.2
181
33.9
27.0-40.8
181
30.5
23.8-37.2
5 to <6
180
24.4
18.2-30.7
180
20.1
14.2-25.9
180
21.3
15.4-27.3
180
31.2
24.4 - 37.9
180
33.1
26.2-40.0
6 to <7
183
22.4
16.4-28.4
183
13.4
8.5-18.4
183
15.3
10.1-20.5
183
22.3
16.3-28.3
183
25.3
19.0-31.6
Box (Total) by Age (years)















0.5 to <1
45
31.1
17.6-44.6
45
40.8
26.5-55.2
45
33.4
19.6-47.2
45
61.5
47.3-75.7
45
52.4
37.8-67.0
1 to <2
98
42.9
33.1 - 52.7
98
45.9
36.1 -55.8
98
43.5
33.6-53.3
98
65.0
55.6-74.5
98
62.5
52.9-72.0
2 to <3
139
38.8
30.7 - 47.0
139
29.9
22.3-37.5
139
27.0
19.6-34.3
139
42.8
34.6-51.0
139
39.0
30.9-47.1
3 to <4
128
32.8
24.7 - 40.9
128
22.5
15.3-29.8
128
21.0
13.9-28.0
128
36.3
28.0 - 44.6
128
33.9
25.7-42.1
4 to <5
138
36.2
28.2-44.3
138
20.4
13.7-27.2
138
18.2
11.7 - 24.6
138
35.1
27.2-43.1
138
31.6
23.8 - 39.3
5 to <6
139
30.2
22.6-37.8
139
19.3
12.8-25.9
139
20.6
13.9-27.4
139
32.7
24.9-40.5
139
34.7
26.8-42.6
6 to <7
139
28.1
20.6 - 35.5
139
14.2
8.4 - 20.0
139
16.0
9.9 - 22.1
139
24.5
17.3-31.6
139
27.6
20.2-35.1
Basin (Total) by Age (years)















0.5 to <1
22
18.2
2.1-34.3
22
48.8
27.9-69.7
22
41.5
20.9-62.1
22
60.3
39.9-80.8
22
51.5
30.6-72.4
1 to <2
26
26.9
9.9-44.0
26
36.2
17.8-54.7
26
34.0
15.7-52.2
26
54.6
35.4-73.7
26
51.6
32.4-70.8
2 to <3
51
15.7
5.7-25.7
51
22.8
11.3-34.3
51
20.3
9.3-31.3
51
31.3
18.6-44.0
51
27.7
15.4-40.0
3 to <4
38
18.4
6.1-30.7
38
20.9
8.0 - 33.8
38
19.4
6.8 - 32.0
38
31.1
16.4-45.8
38
28.8
14.4-43.2
4 to <5
43
2.3
0.4-12.1
43
26.0
12.9-39.1
43
23.7
11.0-36.4
43
30.0
16.3-43.7
43
27.0
13.7-40.3
5 to <6
41
4.9
1.3-16.1
41
22.5
9.7 - 35.3
41
23.7
10.7-36.7
41
26.2
12.7-39.6
41
27.7
14.0-41.4
6 to <7
44
4.5
1.3-15.1
44
11.2
1.9-20.5
44
13.0
3.1-23.0
44
15.4
4.7-26.0
44
18.0
6.6-29.3
Box (Total) by Calendar Year















1995
89
51.7
41.3-62.1
89
34.0
24.1 -43.8
89
33.2
23.4-42.9
89
63.9
53.9-73.9
89
63.3
53.3-73.3
1996
91
44.0
33.8 - 54.2
91
29.5
20.1 -38.8
91
28.8
19.5-38.2
91
56.9
46.7-67.1
91
56.6
46.4-66.7
1997
65
52.3
40.2 - 64.5
65
27.9
17.0-38.8
65
26.3
15.6-37.0
65
49.5
37.3-61.6
65
47.2
35.1 -59.4
1998
94
48.9
38.8 - 59.0
94
32.3
22.9-41.8
94
30.8
21.4-40.1
94
47.3
37.2-57.4
94
45.7
35.7-55.8
1999
109
40.4
31.2-49.6
109
30.1
21.5-38.7
109
28.4
19.9-36.8
109
41.3
32.0-50.5
109
39.2
30.1 -48.4
2000
85
43.5
33.0 - 54.1
85
24.2
15.1-33.3
85
23.8
14.7-32.8
85
33.6
23.6-43.6
85
32.7
22.8-42.7
2001
78
20.5
11.6 - 29.5
78
21.0
11.9-30.0
78
19.2
10.5-28.0
78
35.5
24.9-46.1
78
32.6
22.2 - 43.0
2002
78
10.3
3.5-17.0
78
16.9
8.6-25.2
78
15.7
7.6-23.7
78
26.3
16.5-36.0
78
24.6
15.0-34.2
2003-2008
8
12.5
2.2-47.1
8
26.3
7.7 - 60.2
8
25.6
7.4 - 59.6
8
25.1
7.2 - 59.2
8
24.2
6.8 - 58.4
2013
105
4.8
0.7 - 8.8
105
15.4
8.5 - 22.4
105
13.5
7.0 - 20.0
105
13.9
7.3 - 20.6
105
11.9
5.7-18.2
2018
24
25.0
7.7-42.3
24
7.4
1.9 - 24.6
24
6.6
1.6 - 23.6
24
6.6
1.6 - 23.6
24
5.8
1.3 - 22.5
Basin (Total) by Calendar Year















2002-2005
22
27.3
8.7-45.9
22
47.9
27.0-68.8
22
46.5
25.6-67.3
22
59.8
39.3-80.3
22
58.4
37.8-79.0
2006
25
20.0
4.3-35.7
25
45.3
25.8-64.8
25
44.5
25.0-63.9
25
53.2
33.7-72.8
25
51.6
32.0-71.1
2007
25
24.0
7.3-40.7
25
29.0
11.3-46.8
25
27.4
9.9 - 44.9
25
43.4
24.0-62.9
25
40.5
21.2-59.7
2008
21
0.0
0.0-15.5
21
16.9
0.9 - 33.0
21
15.9
0.2-31.5
21
24.0
5.7-42.3
21
22.7
4.8-40.6
2009
57
10.5
2.6-18.5
57
16.3
6.7-25.9
57
14.9
5.6-24.1
57
28.0
16.4-39.7
57
26.1
14.7-37.5
2010
32
3.1
0.6-15.7
32
16.5
3.6 - 29.3
32
15.1
2.7 - 27.4
32
24.4
9.5 - 39.3
32
22.5
8.1 - 37.0
2011
9
11.1
2.0-43.5
9
25.9
8.1 -58.1
9
24.0
7.1 - 56.4
9
34.7
3.6 - 65.8
9
32.8
2.1 - 63.5
2013
29
6.9
1.9-22.0
29
18.1
4.1 -32.1
29
17.0
3.3 - 30.7
29
20.6
5.9 - 35.3
29
19.0
4.8 - 33.3
2015
27
3.7
0.7-18.3
27
27.0
10.3-43.8
27
24.9
8.6-41.2
27
27.4
10.6-44.2
27
25.2
8.8-41.6
2017
14
14.3
4.0-39.9
14
7.7
1.5 - 32.2
14
6.2
1.0 - 30.3
14
9.6
2.1 - 34.5
14
7.8
1.5 - 32.3
2018
4
25.0
4.6-69.9
4
46.5
13.2-83.2
4
46.2
13.1-83.0
4
40.5
10.5-79.8
4
39.8
10.2-79.4
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.

-------
Table H-2b. Summary of observed and predicted average probability of exceeding 5 pg/dL for IEUBK model version 1.1 build 11 and censor level 3 (dataset omits predicted and observed BLLs £10.5











Predicted Blood Lead







..


.

55/45 default dust/soil partition




40/30/30 Dust/Soil* Partition



UDservea mooa Leacr



IEUBK v1.1






IEUBK v1.1









v2.0 IRsd


v1.1 IRsd


v2.0 IRsd


v1.1 IRsd



Average
95% CI for

Average



Average



Average



Average




Percent
Average %

Predicted
95% CI for

Predicted
95% CI for

Predicted
95% CI for

Predicted
95% CI for


BLLs >5
BLLs>5

Probabilities
Probabilities

Probabilities
Probabilities

Probabilities
Probabilities

Probabilities
Probabilities

N
W/dL (%)
pg/dL (%)
N
>5 pg/dL (%)
>5 pg/dL (%)
N
>5 pg/dL (%)
>5 pg/dL (%)
N
>5 pg/dL (%)
>5 pg/dL (%)
N
>5 pg/dL (%)
>5 pg/dL (%)
Site-Wide (Total)
1091
28.8
26.1
31.5
1091
22.4
19.9
24.9
1086
38.1
35.2
41.0
1091
34.4
31.6
37.2
1090
55.3
52.4
58.3
Box (Total)
826
34.3
31.0
37.5
826
22.6
19.7
25.4
821
38.6
35.3
42.0
826
36.3
33.0
39.5
825
57.3
54.0
60.7
Basin (Total)
265
11.7
7.8-
15.6
265
21.7
16.8
26.7
265
36.6
30.8
42.4
265
28.6
23.2
34.1
265
49.0
43.0
55.0
Upper Basin (Total)
207
11.1
6.8-
15.4
207
22.5
16.8
28.2
207
38.9
32.3
45.5
207
30.7
24.4
37.0
207
53.2
46.4
60.0
Lower Basin (Total)
58
13.8
4.9-
22.7
58
18.9
8.9-
29.0
58
28.2
16.6
39.8
58
21.2
10.7
31.8
58
33.9
21.7
46.1
Site-Wide by Age (years)




















0.5 to <1
67
26.9
16.3
37.5
67
44.6
32.7
56.5
67
44.2
32.3
56.0
67
62.4
50.8
74.0
67
61.9
50.3
73.6
1 to <2
124
39.5
30.9
48.1
124
41.1
32.5
49.8
123
54.5
45.7
63.3
124
60.5
51.9
69.1
124
74.2
66.5
81.9
2 to <3
190
32.6
26.0
39.3
190
24.7
18.5
30.8
190
47.4
40.3
54.5
190
35.8
29.0
42.7
190
63.2
56.3
70.0
3 to <4
166
29.5
22.6
36.5
166
19.3
13.3
25.3
165
43.4
35.9
51.0
166
31.5
24.5
38.6
165
63.0
55.7
70.4
4 to <5
181
28.2
21.6
34.7
181
18.8
13.1
24.5
180
34.0
27.1
40.9
181
30.4
23.7
37.1
181
51.4
44.1
58.7
5 to <6
180
24.4
18.2
30.7
180
16.8
11.4
22.3
179
31.4
24.6
38.2
180
27.1
20.6
33.6
180
48.1
40.8
55.4
6 to <7
183
22.4
16.4
28.4
183
10.9
6.4-
15.5
182
21.0
15.1
27.0
183
18.7
13.1
24.3
183
35.9
28.9
42.8
Box (Total) by Age (years)




















0.5 to <1
45
31.1
17.6
44.6
45
42.1
27.6
56.5
45
41.6
27.2
56.0
45
62.9
48.8
77.0
45
62.4
48.2
76.5
1 to <2
98
42.9
33.1
52.7
98
43.3
33.5
53.1
97
56.8
47.0
66.7
98
63.0
53.5
72.6
98
76.1
67.6
84.5
2 to <3
139
38.8
30.7
47.0
139
26.6
19.3
34.0
139
49.9
41.6
58.2
139
39.2
31.1
47.3
139
65.7
57.8
73.6
3 to <4
128
32.8
24.7
40.9
128
19.8
12.9
26.7
127
44.3
35.7
53.0
128
33.1
24.9
41.2
127
64.2
55.8
72.5
4 to <5
138
36.2
28.2
44.3
138
17.5
11.2
23.9
137
32.7
24.9
40.6
138
31.8
24.0
39.6
138
53.2
44.9
61.5
5 to <6
139
30.2
22.6
37.8
139
16.1
10.0
22.2
138
31.1
23.3
38.8
139
28.6
21.1
36.1
139
50.5
42.2
58.9
6 to <7
139
28.1
20.6
35.5
139
11.7
6.4-
17.1
138
21.7
14.8
28.6
139
20.9
14.1
27.7
139
38.8
30.7
46.9
Basin (Total) by Age (years)




















0.5 to <1
22
18.2
2.1
34.3
22
49.8
28.9
70.7
22
49.4
28.5
70.3
22
61.5
41.2
81.9
22
61.0
40.6
81.4
1 to <2
26
26.9
9.9
44.0
26
32.8
14.7
50.8
26
46.0
26.9
65.2
26
51.1
31.8
70.3
26
67.3
49.3
85.4
2 to <3
51
15.7
5.7
25.7
51
19.4
8.6-
30.3
51
40.7
27.2
54.2
51
26.7
14.5
38.8
51
56.4
42.8
70.0
3 to <4
38
18.4
6.1
30.7
38
17.5
5.4-
29.6
38
40.5
24.9
56.1
38
26.4
12.4
40.4
38
59.3
43.7
74.9
4 to <5
43
2.3
0.4
12.1
43
22.9
10.3
35.5
43
38.0
23.5
52.5
43
26.0
12.9
39.1
43
45.5
30.7
60.4
5 to <6
41
4.9
1.3
16.1
41
19.3
7.2-
31.4
41
32.3
18.0
46.6
41
22.1
9.4-
34.8
41
39.8
24.8
54.8
6 to <7
44
4.5
1.3
15.1
44
8.5
0.3-
16.7
44
19.0
7.4-
30.6
44
11.7
2.2-
21.2
44
26.7
13.6
39.7
Box (Total) by Calendar Year




















1995
89
51.7
41.3
62.1
89
31.0
21.4
40.6
88
48.1
37.7
58.6
89
61.1
51.0
71.2
89
81.7
73.7
89.8
1996
91
44.0
33.8
54.2
91
26.5
17.5
35.6
91
43.7
33.6
53.9
91
53.7
43.5
64.0
91
75.3
66.4
84.1
1997
65
52.3
40.2
64.5
65
25.1
14.6
35.7
64
42.2
30.1
54.3
65
46.4
34.3
58.5
65
69.3
58.1
80.5
1998
94
48.9
38.8
59.0
94
29.1
19.9
38.3
94
50.2
40.1
60.3
94
43.7
33.7
53.7
94
69.3
60.0
78.6
1999
109
40.4
31.2
49.6
109
27.3
19.0
35.7
107
44.4
34.9
53.8
109
37.9
28.8
47.0
108
61.1
51.9
70.3
2000
85
43.5
33.0
54.1
85
21.4
12.7
30.1
84
36.4
26.1
46.7
85
29.9
20.1
39.6
85
53.0
42.4
63.6
2001
78
20.5
11.6
29.5
78
18.4
9.8
27.0
78
34.6
24.1
45.2
78
32.1
21.8
42.5
78
56.0
45.0
67.0
2002
78
10.3
3.5
17.0
78
14.8
6.9
22.7
78
26.6
16.8
36.4
78
23.2
13.8
32.6
78
42.6
31.7
53.6
2003-2008
8
12.5
2.2
47.1
8
24.7
7.0
58.8
8
32.5
0.0-
64.9
8
23.3
6.4-
57.6
8
32.9
0.4-
65.5
2013
105
4.8
0.7
8.8
105
13.2
6.7
19.7
105
27.3
18.8
35.9
105
11.7
5.5-
17.8
105
26.0
17.6-
¦ 34.4
2018
24
25.0
7.7
42.3
24
5.6
1.2
22.3
24
15.3
0.9-
29.7
24
4.7
0.9-
21.0
24
15.2
0.9-
29.6
Basin (Total) by Calendar Year




















2002-2005
22
27.3
8.7
45.9
22
45.6
24.8
- 66.4
22
60.4
40.0
80.8
22
57.1
36.4
- 77.7
22
75.9
58.1
93.8
2006
25
20.0
4.3
35.7
25
42.5
23.2
- 61.9
25
57.4
38.1
76.8
25
49.4
29.8
- 69.0
25
70.4
52.6
88.3
2007
25
24.0
7.3
40.7
25
26.9
9.5
44.2
25
40.6
21.3
59.8
25
40.1
20.9
- 59.3
25
61.0
41.9
80.1
2008
21
0.0
0.0
15.5
21
13.7
4.7
34.0
21
26.4
7.6-
45.3
21
19.3
2.4-
36.1
21
40.9
19.9
61.9
2009
57
10.5
2.6
18.5
57
13.0
4.3
21.7
57
29.5
17.7
41.4
57
23.8
12.7 ¦
- 34.8
57
47.6
34.6
60.5
2010
32
3.1
0.6
15.7
32
12.8
1.2
24.4
32
31.3
15.2
47.3
32
20.1
6.2-
34.0
32
43.2
26.1
60.4
2011
9
11.1
2.0
43.5
9
22.0
6.2
54.6
9
46.3
13.7
78.9
9
29.7
10.1
- 61.5
9
60.6
28.7
92.6
2013
29
6.9
1.9
22.0
29
16.0
2.7
29.4
29
26.9
10.7
43.0
29
17.9
4.0-
31.9
29
31.9
15.0
48.9
2015
27
3.7
0.7
18.3
27
24.5
8.2
40.7
27
36.3
18.1
54.4
27
24.2
8.1 -
40.4
27
39.5
21.0
57.9
2017
14
14.3
4.0
39.9
14
5.7
0.9
29.6
14
17.6
5.6-
43.6
14
7.2
1.3-
31.6
14
21.5
0.0-
43.1
2018
4
25.0
4.6
69.9
4
43.7
11.9
- 81.7
4
62.1
21.7 ¦
- 90.7
4
37.1
9.0-
77.8
4
59.3
20.1 •
• 89.4
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
The observed BLL summary statistics correspond to six model n
s and not the IEUBK v 1.1
1 IRs, both partition model runs. By censoring the data sets, the IEUBK v 1.1
1 IRs for both partitions result in fewer BLL records due to predicted BLLs >30 |ig/dL.

-------
Table H-3. Summary of observed/predicted blood lead pairs that fall outside prediction intervals for censor level 3 (dataset omits predicted and observed BLLs >10.5 pg/dL and
observed BDL BLLs)	



55/45 Dust/Soil Partition



40/30/30 Dust/Soil* Partition




IEUBK v2.0




IEUBK V2.0




v2.0 IRsd

EFH(USEPA 2017b) IRSD

v2.0 IRsd

EFH(USEPA 2017b) IRSD

% Outside
% < Lower
% > Upper
% Outside
% < Lower
% > Upper
% Outside
% < Lower
% > Upper
% Outside
% < Lower
% > Upper

Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction

Limit
Limit
Limit
Limit
Limit
Limit
Limit
Limit
Limit
Limit
Limit
Limit
Site Wide (Total)
13%
7%
6%
13%
6%
7%
11%
9%
2%
11%
9%
2%
Box (Total)
13%
6%
7%
13%
6%
8%
12%
9%
2%
12%
9%
3%
Basin (Total)
10%
7%
3%
11%
7%
4%
11%
9%
2%
10%
8%
2%
Upper Basin
10%
7%
3%
10%
7%
3%
11%
10%
1%
10%
9%
1%
Lower Basin
12%
7%
5%
14%
7%
7%
10%
5%
5%
10%
5%
5%
Site-Wide by Age (years)












0.5 to <1
16%
10%
6%
16%
9%
7%
24%
21%
3%
15%
12%
3%
1 to <2
10%
9%
2%
10%
7%
2%
13%
13%
0%
12%
12%
0%
2 to <3
12%
5%
7%
11%
4%
7%
9%
6%
3%
9%
6%
3%
3 to <4
9%
4%
5%
10%
3%
7%
7%
7%
0%
7%
7%
0%
4 to <5
18%
9%
9%
19%
8%
12%
13%
10%
3%
14%
9%
5%
5 to <6
11%
7%
4%
12%
8%
4%
10%
9%
1%
10%
9%
1%
6 to <7
13%
4%
9%
13%
5%
7%
11%
7%
4%
13%
8%
4%
Box (Total) by Age (years)












0.5 to <1
16%
9%
7%
13%
7%
7%
24%
22%
2%
13%
11%
2%
1 to <2
10%
9%
1%
8%
7%
1%
10%
10%
0%
10%
10%
0%
2 to <3
12%
5%
7%
12%
4%
8%
10%
7%
3%
10%
7%
3%
3 to <4
11%
4%
7%
12%
3%
9%
8%
8%
0%
8%
8%
0%
4 to <5
19%
8%
11%
20%
6%
14%
13%
9%
4%
14%
9%
6%
5 to <6
12%
7%
4%
12%
8%
4%
10%
9%
1%
10%
9%
1%
6 to <7
15%
5%
10%
14%
6%
8%
13%
8%
5%
15%
10%
5%
Basin (Total) by Age (years)












0.5 to <1
18%
14%
5%
23%
14%
9%
23%
18%
5%
18%
14%
5%
1 to <2
12%
8%
4%
15%
8%
8%
23%
23%
0%
19%
19%
0%
2 to <3
10%
4%
6%
10%
4%
6%
8%
4%
4%
6%
2%
4%
3 to <4
3%
3%
0%
3%
3%
0%
5%
5%
0%
5%
5%
0%
4 to <5
16%
14%
2%
16%
14%
2%
14%
12%
2%
14%
12%
2%
5 to <6
10%
7%
2%
10%
7%
2%
10%
10%
0%
10%
10%
0%
6 to <7
7%
2%
5%
7%
2%
5%
5%
2%
2%
5%
2%
2%
Box (Total) by Calendar Year












1995
17%
9%
8%
16%
9%
7%
18%
18%
0%
18%
17%
1%
1996
19%
8%
11%
18%
7%
11%
14%
13%
1%
15%
11%
4%
1997
18%
8%
11%
22%
6%
15%
14%
12%
2%
14%
10%
4%
1998
12%
5%
6%
12%
5%
6%
13%
9%
4%
15%
9%
6%
1999
10%
6%
5%
11%
5%
6%
8%
6%
3%
11%
5%
6%
2000
21%
9%
12%
20%
9%
11%
13%
11%
2%
16%
10%
6%
2001
13%
6%
6%
13%
5%
8%
10%
10%
0%
14%
9%
5%
2002
10%
8%
3%
9%
6%
3%
12%
10%
1%
10%
9%
1%
2003-2008
0%
0%
0%
0%
0%
0%
0%
0%
0%
9%
9%
0%
2013
4%
3%
1%
4%
2%
2%
3%
2%
1%
3%
3%
1%
2018
21%
0%
21%
21%
0%
21%
21%
0%
21%
16%
2%
14%
Basin (Total) by Calendar Year












2002-2005
18%
18%
0%
18%
18%
0%
18%
18%
0%
38%
35%
3%
2006
20%
20%
0%
20%
20%
0%
24%
24%
0%
37%
37%
0%
2007
16%
12%
4%
20%
12%
8%
16%
16%
0%
13%
13%
0%
2008
5%
5%
0%
5%
5%
0%
10%
10%
0%
18%
18%
0%
2009
2%
0%
2%
4%
0%
4%
4%
4%
0%
9%
9%
0%
2010
0%
0%
0%
0%
0%
0%
6%
6%
0%
10%
10%
0%
2011
11%
0%
11%
11%
0%
11%
0%
0%
0%
0%
0%
0%
2013
14%
7%
7%
14%
7%
7%
14%
7%
7%
15%
9%
6%
2015
15%
11%
4%
15%
11%
4%
7%
7%
0%
7%
7%
0%
2017
21%
0%
21%
21%
0%
21%
21%
0%
21%
31%
0%
31%
2018
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
0%
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
Grey shaded/bolded text cells represent the lowest percentage of all 8 model runs, for each data category.
The number of records (N) for each model run corresponds to the same Ns shown in Tables H-1a and H-2a.

-------
Table H-4. Summary of sum of squared differences for censor level 3 (dataset omits predicted and observed BLLs >10.5 ng/dL and observed BDL BLLs)


55/45 Dust/Soil Partition


40/30/30 Dust/Soil* Partition


IEUBK v2.0
IEUBK V1.1
IEUBK V2.0
IEUBK V1.1


EFH(USEPA



EFH(USEPA



v2.0 IRsd
2017b) IRsd
v2.0 IRsd
v1.1 IRsd
v2.0 IRsd
2017b) IRsd
v2.0 IRsd
v1.1 IRsd
Site-Wide (Total)
459
454
531
536
430
412
439
680
Box (Total)
361
360
414
389
333
322
341
511
Basin (Total)
98
94
116
147
97
90
97
169
Upper Basin (Total)
76
73
87
119
77
72
74
144
Lower Basin (Total)
23
21
29
28
20
18
23
26
Site-Wide by Age (years)








0.5 to <1
32
28
33
33
40
32
42
41
1 to <2
43
41
45
55
54
49
52
83
2 to <3
76
76
86
103
72
68
73
132
3 to <4
58
59
70
76
52
51
53
111
4 to <5
94
95
109
105
79
75
81
119
5 to <6
72
72
86
80
66
68
66
104
6 to <7
83
82
102
85
67
69
71
90
Box (Total) by Age (years)








0.5 to <1
19
17
20
19
26
21
28
27
1 to <2
32
31
34
41
39
36
38
63
2 to <3
55
56
62
69
52
50
53
92
3 to <4
50
51
59
57
42
41
44
83
4 to <5
75
76
85
73
62
59
63
88
5 to <6
58
57
68
58
53
56
54
83
6 to <7
73
71
88
70
58
60
62
75
Basin (Total) by Age (years)








0.5 to <1
13
11
13
13
14
11
14
14
1 to <2
11
10
12
14
14
13
14
20
2 to <3
21
21
24
34
20
19
19
40
3 to <4
9
9
11
18
10
9
9
28
4 to <5
20
18
24
32
18
16
18
31
5 to <6
15
15
18
22
12
13
13
21
6 to <7
11
11
15
14
9
9
9
14
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
Grey shaded/boIded text cells represent the lowest value of all 8 model runs, for each data category.
The number of records (N) for each model run corresponds to the same Ns shown in Tables H-1 and H-2.

-------
Table H-5. Percent differences <1 ng/dL between observed and predicted BLLs for censor level 3 (dataset omits predicted and observed BLLs >10.5 ng/dL and observed BDL BLLs)


55/45 Default Dust/Soil Partition


40/30A/30 A Default Dust/Soil Partition


IEUBK v2.0
IEUBK V1.1
IEUBK V2.0
IEUBK V1.1


EFH (USEPA



EFH (USEPA



v2.0 IRsd
2017b) IRsd
v2.0 IRsd
v1.1 IRsd
v2.0 IRsd
2017b) IRsd
v2.0 IRsd
v1.1 IRsd
Site-Wide (Total)
37%
38%
36%
33%
37%
37%
36%
22%
Box (Total)
35%
36%
34%
31%
34%
34%
33%
22%
Basin (Total)
46%
47%
41%
37%
45%
46%
46%
25%
Upper Basin (Total)
48%
49%
43%
37%
47%
48%
49%
22%
Lower Basin (Total)
36%
40%
34%
36%
38%
38%
34%
36%
Site-Wide by Age (years)








0.5 to <1
27%
37%
28%
28%
19%
16%
15%
15%
1 to <2
37%
36%
35%
24%
23%
23%
24%
15%
2 to <3
35%
34%
32%
29%
37%
38%
37%
16%
3 to <4
37%
38%
34%
30%
41%
39%
37%
19%
4 to <5
36%
39%
38%
29%
36%
39%
38%
22%
5 to <6
40%
39%
38%
39%
39%
39%
37%
31%
6 to <7
44%
43%
40%
43%
48%
48%
47%
32%
Box (Total) by Age (years)








0.5 to <1
31%
42%
33%
33%
22%
20%
18%
18%
1 to <2
38%
37%
38%
25%
21%
22%
23%
14%
2 to <3
29%
29%
27%
29%
35%
34%
32%
17%
3 to <4
34%
34%
32%
27%
38%
36%
34%
19%
4 to <5
31%
33%
32%
27%
33%
36%
36%
23%
5 to <6
39%
38%
37%
37%
35%
35%
34%
29%
6 to <7
41%
40%
39%
40%
42%
42%
42%
27%
Basin (Total) by Age (years)








0.5 to <1
18%
27%
18%
18%
14%
9%
9%
9%
1 to <2
35%
35%
27%
23%
27%
27%
27%
19%
2 to <3
51%
47%
45%
29%
41%
49%
53%
14%
3 to <4
50%
50%
39%
39%
50%
50%
47%
18%
4 to <5
51%
58%
56%
35%
44%
47%
47%
19%
5 to <6
44%
44%
41%
44%
51%
51%
49%
39%
6 to <7
52%
55%
43%
55%
66%
64%
64%
48%
Notes:
*40% dust, 30% property soil, 30% arithmetic average community soil.
Grey shaded/boIded text cells represent the highest value of all 8 model runs, for each data category.
The number of records (N) for each model run corresponds to the same Ns shown in Tables H-1 and H-2.

-------
Figure H-1. Summary of mean observed and predicted blood lead levels for geographic areas for censor level 3 (omits predicted and observed BLLs > 10.5 pg/dL and observed BDL BLLs)
m 3.5
1 3'°
7 2.5
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Predicted Blood Lead GM (ng/dL)
Upper
Basin
Basin
Predicted Blood Lead GM (ng/dL)
A) IEUBK v2.0, v2.0 IRs, 55/45 partition	B) IEUBK v2.0, EFH IRs, 55/45 partition	C) IEUBK v1.1, v2.0 IRs, 55/45 partition
Site-wide
Upper
Basin
Basin
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Predicted Blood Lead GM (ng/dL)
00
=L
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Predicted Blood Lead GM (ng/dL)
Jf
Basin Upper
Basin
m 3.5
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Predicted Blood Lead GM (ng/dL)
pper
n Basin
Basin
D) IEUBK v2.0, v2.0 IRs, 40/30/30 partition
E) IEUBK v2.0, EFH IRs, 40/30/30 partition
F) IEUBK v1.1, v2.0 IRs, 40/30/30 partition

-------
Figure H-2. Summary of mean observed and predicted blood lead levels for site-wide age groups for censor level 3 (omits predicted and observed BLLs > 10.5 pg/dL and observed BDL BLLs)
A) IEUBK v2.0, v2.0 IRs, 55/45 partition - Site-wide by Age	B) IEUBK v2.0, EFH IRs, 55/45 partition - Site-wide by Age	C) IEUBK v1.1, v2.0 IRs, 55/45 partition - Site-wide by Age
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Predicted Blood Lead GM (ng/dL)
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
Predicted Blood Lead GM (ng/dL)
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
Predicted Blood Lead GM (ng/dL)
M
=L
(J
"O
ro
tu
5.0
(J
"~
ro
0)
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
I	I	I	I	I	I
1.0 2.0 3.0 4.0 5.0 6.0
Predicted Blood Lead GM (ng/dL)
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
Predicted Blood Lead GM (ng/dL)
ID
"O
03
0)
4.0
i	i	i	i	i	i
1.0 2.0 3.0 4.0 5.0 6.0
Predicted Blood Lead GM (ng/dL)
D) IEUBK v2.0, v2.0 IRs, 40/30/30 partition - Site-wide by Age
E) IEUBK v2.0, EFH IRs, 40/30/30 partition - Site-wide by Age
F) IEUBK v1.1, v2.0 IRs, 40/30/30 partition - Site-wide by Age

-------
Figure H-3. Summary of observed and predicted average probability of exceeding 5 pg/dL for geographic areas for censor level 3 (all predicted and observed BLLs <10.5 pg/dL and no BDLs)
0 5 10 15 20 25 30 35 40 45 50
Predicted {%)
A) IEUBKv2.0, v2.0 IRs, 55/45 partition
0 5 10 15 20 25 30 35 40 45 50
Predicted {%)
B) IEUBKv2.0, EFH IRs, 55/45 partition
0 5 10 15 20 25 30 35 40 45 50
Predicted {%)
C) IEUBK v1.1, v2.0 IRs, 55/45 partition

50

45

40

35

30
"~

CD
>
25
(1)

>S)
O
20
O


15

10

5

0
0 5 10 15 20 25 30 35 40 45 50
Predicted (%)
0 5 10 15 20 25 30 35 40 45 50
Predicted (%)
0 5 10 15 20 25 30 35 40 45 50
Predicted (%)
D) IEUBK v2.0, v2.0 IRs, 40/30/30 partition
E) IEUBK v2.0, EFH IRs, 40/30/30 partition
F) IEUBK v1.1, v2.0 IRs, 40/30/30 partition

-------
Figure H-4. Summary of observed and predicted average probability of exceeding 5 pg/dL for site-wide age groups for censor level 3 (all predicted and observed BLLs <10.5 pg/dL and no BDLs)
A) IEUBK v2.0, v2.0 IRs, 55/45 partition - Site-wide by Age
B) IEUBK v2.0, EFH IRs, 55/45 partition - Site-wide by Age
C) IEUBK v 1.1, v2.0 IRs, 55/45 partition - Site-wide by Age
o
0 10 20 30 40 50 60 70 80
Predicted (%)
50
30
0
0 10 20 30 40 50 60 70 80
Predicted (%)
20 30 40 50 60 70 80
Predicted (%)
£ 50
1 40
-Q
O
30
50
30
20 30 40 50 60 70 80
Predicted (%)
£ 50
"O
> 40 ¦
CD
CO
8 30
20
30 40 50 60 70 80
Predicted {%)
I	I	I	I	I	I
30 40 50 60 70 80
Predicted {%)
50
30
£ 50
"O
> 40 ¦
CD
CO
8 30
20
10 ¦
D) IEUBK v2.0, v2.0 IRs, 40/30/30 partition - Site-wide by Age
E) IEUBK v2.0, EFH IRs, 40/30/30 partition - Site-wide by Age
F) IEUBK v1.1, v2.0 IRs, 40/30/30 partition - Site-wide by Age

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix I
Not Human Subjects Research Determination
1-1

-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D C. 20460
OFFICE OF THE
SCIENCE ADVISOR
MEMORANDUM
SUBJECT:
REQUEST No.
DATE:
Not Human Subjects Research Determination
HSR-001081
May 9, 2019
FROM:
Daniel Nelson
Human Subjects Officer
NHEERL/ORD/EP A
TO:
James Brown
National Center for Environmental Assessment (RTP)/ORD
• PURPOSE: The goal of this project evaluate the predictive ability of EPA's Integrated Exposure
Uptake Biokinetic (IEUBK) model for children using blood lead level (BLL) matched with other lead
in other media (especially soil lead). This work is being conducted under contract with SRC (Contract
No. EP-C-17-015, Task Order No. 11).
• PARTICIPANTS: Research subjects are children as described above. Research investigators include:
James Brown (Contracting Officer's Representative, TOCOR), Gary Diamond (SRC Program
Manager), Mark Follansbee (SRC Task Order Manager), and Susan Spalinger (Alta Science &
Engineering Project Manager)
• PROCEDURES: The data have been collected during routine BLL monitoring at the Bunker Hill
Superfund site by the Idaho Department of Environmental Quality (IDEQ). IDEQ contracts with Alta
Science and Engineering, Inc. for technical, scientific, and engineering services at the Bunker Hill
Superfund Site, including maintaining the blood lead and environmental databases. Alta maintains
these data in a confidential (coded) format that prevents the identification of the participants and their
home locations. An informed consent example is provided. This project is discussed under Task 4a-c
of the attached Quality Assurance Project Plan (QAPP). As described in on p. 12 of the QAPP for
Contract No. EP-C-17-015, Task Order No. 11: "SRC Project Team members from Alta will use paired
BLL and environmental media concentration data from the Bunker Hill Superfund Site in Idaho to
evaluate the IEUBK model (v. 2.0) predictions. Alta will conduct the IEUBK model evaluation and QC
using de-identified data sets that Alta maintains. No individual or confidential data will be transmitted
to SRC, ICF, or EPA.
• NOTE: This review is provided by D. Nelson on behalf of ORD/EP A, since the submitter for this
project (J. Brown) is also the desigated reviewer for NCEA.

-------
I have reviewed the study described above according to the requirements of EPA Order 1000.17A (Policy
and Procedures on Protection of Human Research Subjects) and EPA Regulation 40 CFR 26 (Protection
of Human Subjects), and have determined that it is not human subjects research. This determination is based
on the fact that the study will not obtain data about people through intervention or interaction with
individuals or obtain identifiable private information (40 CFR 26.102(f)).
Should the plan and/or protocol change as it is developed to include the participation of human subjects, it is
expected that the investigators will submit an application to an appropriate IRB and a request for approval to
the HSRRO. In addition, all researchers are expected to conduct their work with integrity and with attention
to best practices in research ethics regardless of regulatory oversight.
Please refer to the HSR Request Number (HSR-001081) for all future communication.
Principal Investigator:
EPA Contact:
Application/Grant/Award
Number:
Application Title:
Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory
Decisions and Site Assessments
James Brown
James Brown
N/A

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix J
Biographical Summaries of Peer Reviewers
J-1

-------
Appendix J - Biographical Summaries of Peer Reviewers
PHILIP E. GOODRUM
Dr. Philip Goodrum is a Senior Consultant with Cardno ENTRIX with more than 20 years of
experience in environmental modeling and applications of probability and statistics to human
health and ecological risk assessment, compliance monitoring, and natural resources damages
assessment. He received a Ph.D. in Environmental Engineering from the State University of
New York (SUNY) College of Environmental Science and Forestry (ESF) in 1999; an M.S. in
Environmental Engineering from SUNY ESF in 1995; and a B.S. in Environmental Technology
from Cornell University in 1989. Dr. Goodrum developed and demonstrated applications of the
Integrated Stochastic Exposure Model for lead, which uses Monte Carlo simulation to quantify
variability and uncertainty in childhood blood lead concentrations based on variability and
uncertainty in exposures. Dr. Goodrum specializes in quantitative uncertainty analysis and lead
risk assessment, having served for approximately 10 years as a consultant for USEPA's
Technical Review Workgroup for Lead. As a senior project manager for Syracuse Research
Corporation from 1996 to 2006, he conducted and reviewed numerous lead risk assessments,
managed EPA's "Lead Hotline" which assisted the public with applications of both the Integrated
Exposure Uptake Biokinetic (IEUBK) model and the interim Adult Lead models, co-authored
numerous platform presentations, technical white papers and guidance documents, and actively
participated in the research and development of EPA's All Ages Model for lead. Dr. Goodrum
has been an active member of community outreach and professional peer review panels. In
1998-1999, he served as the chair of the Syracuse Regional Lead Task Force, responsible for
coordinating public outreach and educational programs for the Syracuse community on
childhood lead exposure. Dr. Goodrum served on a peer review panel for U.S. EPA National
Center for Exposure Assessment (NCEA) for the All-Ages Risk Model in 2000. He was an
invited speaker by NCEA for the National Air Quality Criteria for Lead Workshop held in Chapel
Hill, NC, Feb. 1-3, 2005. In 2006-2007, Dr. Goodrum served on the Clean Air Scientific Advisory
Committee Panel as a member of EPA's Science Advisory Board charged with reviewing the
Lead Renovation, Repair, and Painting (LRRP) report and Office of Pollution Prevention and
Toxics Dust study. Currently he is a member of the Interstate Technology and Regulatory
Council's technical workgroup on Incremental Sampling Methodology, charged with developing
guidance and training on new sampling methodologies for use in risk assessment. Dr.
Goodrum's current research focuses on evaluating performance of environmental models based
on empirical data, as well as developing sampling designs, assessing data usability for
assessments, and conducting exploratory data analysis, regression and correlation analyses,
multivariate analyses, hypothesis testing, trend analysis, outlier analysis, geospatial analysis,
and hotspot identification (cluster analysis).

-------
ROSALIND A. SCHOOF
Dr. Rosalind Schoof is a board certified toxicologist with more than 35 years' experience
assessing human health effects and exposures from chemical substances in a variety of
settings, such as contaminated sites, commercial/ industrial/agricultural/residential projects,
product uses, dietary exposures and general home and community exposures. Her projects
have included numerous formal health risk assessments conducted under various US and
international regulatory settings, as well as regulatory, research and litigation projects. Dr.
Schoof has directed evaluations of chemical toxicity, derivation of risk-based exposure levels,
health risk assessments for cancer and noncancer end points and multimedia exposure
assessments. Dr. Schoof is an internationally recognized expert on evaluation of arsenic and
metals in the environment and in the diet, and on the bioavailability of metals from soil with over
35 peer-reviewed publications. She has served on numerous peer review panels for US
agencies and Canadian ministries, and on several National Research Council committees. She
is currently a member of the US Department of Defense Strategic Environmental Research and
Development Program (SERDP) Science Advisory Board. Prior to her consulting career, Dr.
Schoof worked for a pharmaceutical company conducting safety assessments for new drugs,
and designing and directing toxicity studies. She also worked in the Office of Toxic Substances
at USEPA.

-------
KATHLEEN YORK
Dr. Kathleen Vork is a Staff Toxicologist for the Office of Environmental Health Hazard
Assessment (OEHHA) at the California Environmental Protection Agency. She received her
Ph.D. in Environmental Health Sciences from the University of California at Berkeley in 2003
and her MPH degree in Occupational and Environmental Health from the University of
Minnesota School of Public Health in 1988. Prior to her position at OEHHA, Dr. Vork worked for
the California Childhood Lead Poisoning Prevention Program. Dr. Vork has extensive
experience and expertise relating to exposure pathways and the pharmacokinetics (PBPK) of
lead in workers and the general population. Dr. Vork has implemented various statistical and
mathematical modeling methods to estimate, adjust, and check the accuracy and consistency of
predictions from models combining exposure pathways with physiologically based
pharmacokinetic and bio-kinetic models. She is the primary author of the report entitled
"Estimating Workplace Air and Worker Blood Lead Concentration using an Updated
Physiologically-based Pharmacokinetic (PBPK) Model" (2013). She has conducted work
involving the derivation of human lactation transfer coefficients for various chemicals including
lead for the "Risk Assessment Guidelines Technical Support Documents for Exposure
Assessment and Stochastic Analysis" (2012), and contributed to "The Derivation of Non-cancer
Reference Exposure Levels" (2007) for the California Air Toxics Hot Spots program. Dr. Vork
has worked collaboratively with multiple agencies and the public. She has recently served on
USEPA peer-review consult panels involving complex modeling of lead exposure and
pharmacokinetics (2015, 2016, and 2017). She has also served on the California Advisory
Committee for Training Regulations for Lead Paint Abatement while working for the California
Lead Poisoning Prevention Program, and she chaired the Lead Training Course Planning
Committee while working for the Alameda County Lead Poisoning Prevention Program. She
attended the University of California at Berkeley (Ph.D. in Environmental Health Sciences
(2003)) and the University of Minnesota (MPH in Occupational and Environmental Health).

-------
Advancing Pb Exposure and Biokinetic Modeling
Appendix K
SRC Quality Assurance Project Plan
K-1

-------
QUALITY ASSURANCE PROJECT PLAN
ADVANCING PB EXPOSURE AND BIOKINETIC MODELING FOR EPA
REGULATORY DECISIONS AND SITE ASSESSMENTS
CONTRACT NO. EP-C-17-015
68HERC19F0099
TASK ORDER NO. 0011
EPA QAPP ID: B-RTP-0031816-QP-1-0
Submitted to:
ATTN: Matthew Growney, Contracting Officer
Renita Tyus, Contracting Officer
U.S. Environmental Protection Agency
National Center for Environmental Assessment
26 W. Martin Luther King Drive
Cincinnati, Ohio 45268
Submitted by:
SRC, Inc.
Technical Services Division
7502 Round Pond Road
North Syracuse, NY 13212
Revision No. 2
April 18, 2019

-------
I
QAPP Tor Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18.2019
Page 2 of 22
A. Project Management
A.l	Title and Approval Sheet
QUALITY ASSURANCE PROJECT PLAN
for
ADVANCING PB EXPOSURE AND BIOKINETIC MODELING FOR EPA
REGULATORY DECISIONS AND SITE ASSESSMENTS
(CONTRACT NO. EP-C-17-015, 68HERC19F0099)
(TASK ORDER NO. 0011)
	tlii/ii	
Susan Spalinger, 7	Date	/ /
At (^Science & pioneering Project Manager
T4A^.s	-: ;
James Brown,	Date
EPA Task Order Contracting Officer's Representative (TOCOR)
Date
' /

«w s-wm
Date

Date



feat _	4/22/2019
Cheryl Itkin, {7	Date
NCEA QA Manager

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 3 of 22
A.2 Table of Contents
A.	Project Management	2
A.i Title and Approval Sheet	2
A. 2 Table of Contents	3
A3 Distribution List	4
A.4 Project Task/Organization	4
A.4.1 Management Staff	5
A.4.2 Technical Staff (Information Scientists, Toxicologists, Chemists) 7
A.4.3 Organization	9
A.5 Problem Definition/Background	11
A.6 Project Task/Description	11
A7 Quality Objectives and Criteria	15
A. 8 Special Training/Certification	16
A.9	Documents and Records	16
B.	Data Generation and Acquisition	17
B.i	Sampling Process Design (Experimental Design)	17
B.2 Sampling Methods	17
B.3 Sample Handling and Custody	18
B.4 Analytical Methods (Laboratory)	18
B.5 Quality Control	18
B.6 Instrument/Equipment Testing, Inspection, and Maintenance	18
B.7 Instrument/Equipment Calibration and Frequency	18
B.8 Inspection/Acceptance of Supplies and Consumables	18
B.9 Non-direct Measurements	19
B.10	Data Management	19
C.	Assessment and Oversight	19
C.i	Assessments and Response Actions	19
C.2	Reports to Management	20
D.	Data Validation and Usability	20
D.i	Data Review, Verification, and Validation	20
D.2 Verification and Validation Methods	20
D.3 Reconciliation with User Requirements	21
Appendix A. SRC, Inc. Quality Control (QC) and Quality Assurance (QA) SOP and
Documentation	22

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 4 of 22
A.3 Distribution List
Ruth Corn, EPA COR
James Brown, EPA TOCOR
Cheryl Itkin, NCEA QA Manager
Gary Diamond, SRC Program Manager
Mark Follansbee, SRC Task Order Manager
David Linton, SRC Contracts
Dylan Heh, SRC Quality Engineer
All SRC project staff
All ICF project staff
All Alta Science & Engineering project staff
EPA project staff
Electronic copies of the approved QAPP will be distributed to all SRC project staff by Dr.
Follansbee, who will be responsible for maintaining the current version.
A.4 Project Task/Organization
This Quality Assurance Project Plan (QAPP) describes how SRC, Inc. (SRC) and its
subcontractors (Alta Science & Engineering and ICF) will provide support to the U.S.
Environmental Protection Agency's (EPA) Office of Research and Development (ORD),
National Center for Environmental Assessment (NCEA), in the development and evaluation of
computational models to estimate the disposition of lead (Pb) in humans in support of the EPA's
Superfund Program and other program offices relying on biokinetic models for lead.
The project will be executed in steps as follows:
Task la. Project
Task 2. Kickoff teleconference
Task 3. Prepare QAPP (this document)
Task 4a. Obtain access to blood Pb and environmental data necessary for IEUBK model
evaluation, develop data files necessary for proposed analyses
Task 4b. Perform evaluation on data from 4a; draft report describing the results of model
evaluation; perform additional analyses and revise report
Task 4c. Prepare manuscript describing the data, analysis, and results from 4a & 4b
Task 5. AALM interface for batch runs with instructions describing how to use the batch
run module
Task 6a. Adapt AALM batch run for diverse outputs with instructions describing how to
use the batch run module
Task 6b. Link AALM to SHEDS with instructions describing how to use the batch run
module linking SHEDS and the AALM
Task 7a. Support preparation for SAB evaluation of AALM
Task 7b. Participate in AALM peer-review meeting

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 5 of 22
The quality assurance (QA) requirements in this QAPP relate to data handling and analysis to
verify and track associated quality control (QC) activities that were conducted during the
performance of this task order. The QAPP to support this project is consistent with SRC's
Quality Management Plan submitted to EPA for this contract (EP-C-17-015).
Dr. Mark Follansbee, the SRC Task Order Manager, is responsible for maintaining the most
current version of the QAPP and is also responsible for assuring that all personnel assigned to the
project have the latest version. The most up-to-date electronic version will be stored on a project
folder on an SRC server, which may be accessed by all project personnel.
A.4.1 Management Staff
SRC Staff
Dr. Gary Diamond is the Program Manager for this NCEA support contract. He will ensure
compliance with all technical and contractual requirements of the project and implementation of
quality evaluation and improvement processes, including all QC/QA procedures specified under
the QAPP.
Dr. Diamond has more than 30 years of experience in experimental research, applications of
toxicological and epidemiological research to human health risk assessment, and project
management. Dr. Diamond has served as SRC Project Leader and/or Technical Project Manager
on 60 SRC contracts that have supported EPA and other Federal Government programs
dedicated to chemical and biological hazard and risk assessment. He has led and/or contributed
to numerous research projects to develop improved methods for assessing human health risks
associated with exposures to hazardous chemicals. Dr. Diamond's scientific specialty areas are
metals toxicology and risk assessment, toxicokinetics, and physiologically based
pharmacokinetic (PBPK) modeling. He led the SRC program to assist EPA ORD and Office of
Solid Waste and Emergency Response (OSWER) in developing and maintaining the Integrated
Exposure Uptake Biokinetic (IEUBK) Model for Lead in Children and the Adult Lead
Methodology. He also led the SRC program supporting the EPA ORD Computation Toxicology
program (All Ages Cadmium Model, All Ages Mercury Model) and the program supporting
OPPT in their assessments of risk from exposure to lead resulting from renovation, repair, and
painting activities in public and commercial buildings (All Ages Lead Model, AALM). Dr.
Diamond has developed, researched, and applied PBPK models extensively in support of
chemical toxicology assessments conducted for the EPA NCEA Computational Toxicology,
Integrated Risk Information System (IRIS), and PTV programs. He has served on and/or
contributed to numerous national advisory committees in his field of expertise, including the
EPA Science Advisory Board (SAB), Environmental Health Committee, EPA Risk Assessment
Forum Metals Framework Committee, and National Research Council. He has co-authored more
than 200 reports and publications on various topics of toxicology, chemical and microbial hazard
assessment, and risk assessment.

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 6 of 22
Dr. Mark Follansbee will serve as the Technical Lead and Task Order Manager for this project.
He will be the principal point of contact with the EPA Task Order Contracting Officer's
Representative (TO COR) and will ensure that communication lines are always open. Dr.
Follansbee will provide overall project direction and will ensure the availability and proper
application of SRC resources required to meet project objectives. He will monitor technical
performance and expenditures and will provide QC review of all deliverables to EPA.
Dr. Follansbee has more than 20 years of applied human health risk assessment and project
management experience. He has authored or contributed to a variety of assessments of human
health risks associated with exposures to hazardous chemicals (specialty areas are lead risk
assessment and asbestos exposure and risk assessment). He has provided technical support to
EPA in the preparation or review of risk assessment documents (e.g., baseline risk assessment,
RODs, five-year reviews, ATSDR health consultations) for lead or asbestos at more than 100
Superfund, emergency response, and other hazardous waste sites, including Libby, Montana;
Bunker Hill/Coeur d'Alene, Idaho; Herculaneum/Doe Run, Missouri; Upper Columbia River,
Washington; Omaha Lead Site, Nebraska; Southwest Jefferson County Mine Site in Missouri;
Former United Zinc Site in Iola, Kansas; Cherokee County Site in Tri-State Mining District of
Kansas; Exide Smelter, Pennsylvania; St. Joe Park, Missouri; East Helena, Montana; Coffeyville,
Kansas; Neihart and Tenmile Creek, Montana; Vermont Asbestos Group, Vermont; BoRit in
Ambler, Pennsylvania; Hanford in Richland, Washington; Paducah, Kentucky. He has also
supported EPA emergency response efforts such as the Gold King Mine Spill (Silverton,
Colorado). In addition, he has contributed to the development of national guidance on the
application of bioavailability data to human health risk assessment and the risk-based
methodology for assessing exposure at Superfund sites. Dr. Follansbee has more than 20 years of
experience with the Integrated Exposure Uptake Biokinetic Model for Lead in Children (IEUBK
model), including experience in developing training materials (including videos) and conducting
on-site training in the U.S. and overseas.
Mr. Dylan Heh, SRC's Quality Engineer, has reviewed this QAPP and will provide periodic
inspections to ensure that the processes outlined in the QAPP are followed.
Mr. Heh has over 7 years of experience in the development of processes that are compliant with
industry standards and best practices, as well as the development and delivery of process-related
training. During this time, Mr. Heh has also ensured compliance to internal and external
requirements through process audits, supplier assessments, and product reviews. In addition, Mr.
Heh has implemented corrective and preventive action plans and worked on continuous
improvement initiatives. Mr. Heh is a member of SRC's Quality Department. SRC's Quality
Department is separate and independent from all operating divisions of SRC, including the
Environmental Health Analysis Group of the Technical Services Division, which will be
responsible for the performance of the work. Mr. Heh will not be involved in any day-to-day
technical activities for this Task Order.
Subcontract Staff
A Ita Science & Engineering Staff

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 7 of 22
Susan Spalinger, M.S., will serve as Alta's Technical Lead and Project Manager for this project.
She will oversee Alta's technical work and will provide QC and review. She will assist with
IEUBK modeling, statistical data analysis, QC activities, and project documentation.
Ms. Spalinger is a Principal and co-owner of Alta Science and Engineering, Inc. (Alta). She has
more than 20 years of experience as an environmental scientist, project manager, and program
manager. Her experience spans from small environmental site assessments to some of the
nation's largest CERCLA sites. Ms. Spalinger managed and was the Principal-in-charge of
Alta's multi-million dollar, multi-year sampling and remediation program at the Bunker Hill
Mining and Metallurgical Complex Superfund Site (BHSS) and has managed or contributed to a
number of environmental site investigations and risk evaluations at abandoned mine/smelter,
Department of Energy, and petroleum sites. She specializes in human health risk and exposure
assessment, metals contaminated sites, site characterization and sampling, statistical data
analysis, data management, data verification and validation, and remedy effectiveness
monitoring and evaluation. Much of Ms. Spalinger's career has focused on the human health
remedy at the BHSS, and is currently the Principal-in-Charge and contract manager of Alta's
Technical, Scientific and Engineering Services Contract with the Idaho Department of
Environmental Quality (IDEQ) and oversees the remedial effectiveness monitoring, risk
assessment/management, data management, and GIS activities at the BHSS. Ms. Spalinger has
also co-authored on a number of peer-reviewed papers related to the BHSS and soil/dust
ingestion rates.
ICF Staff
Dr. Cara Henning will serve as the ICF project manager and technical lead.
Dr. Henning has 12 years of experience conducing lead exposure and risk assessments. She has
implemented the IEUBK, Adult Lead Methodology (ALM), and the Leggett model throughout
the support of EPA/OPPT rulemaking, including the Residential Renovation, Repair and
Painting (RRP) rule, the draft Public and Commercial Building RRP, and the draft lead hazard
standard. She oversaw the incorporation of both the IEUBK and the Leggett model in an overall
python modeling system to simulate population exposure to lead dust generated during
renovation activities. In recent years, she and ICF colleagues have collaborated with SRC
scientist Gary Diamond to incorporate a Fortran version of the Leggett AALM into the full
AALM suite. This work included supporting model evaluation using newly available data sets
and building the Excel user interface to complement the existing acslX Excel interface.
A.4.2 Technical Staff (Information Scientists, Toxicologists, Chemists)
SRC Staff
Dr. William Thayer has 31 years of experience. Dr. Thayer will design and conduct the
statistical analyses and develop the predictive model.

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 8 of 22
Dr. Thayer is an expert in statistical analysis of data generated from complex sampling designs
such as the NHANES, American Housing Survey and Lead and Allergens in U.S. Housing
Survey; examples include an analysis of NHANES blood lead concentration data to recommend
new default values for several variables in the IEUBK model and Adult Lead Model (ALM), and
estimation of the distribution of urinary cadmium concentration data from NHANES to support
the calibration of a dietary exposure model developed at SRC. Dr. Thayer serves as the project
manager for SRC's support of EPA Region 10's HHRA of the Upper Columbia River (UCR).
During this time, SRC has prepared data analysis reports for two large surveys: a tribal dietary
consumption and resource use survey and a recreational use survey. Both surveys were designed
to provide data for estimating exposure factors for the UCR HHRA. As part of the data analysis
for the tribal survey, Dr. Thayer estimated fish consumption using a mixed-effects, two-part non-
linear regression model developed by the National Cancer Institute (NCI). Dr. Thayer also
supported NCEA in using the Omaha Lead Site blood lead data.
Mr. John Adams will support Dr. Thayer in the statistical analyses and provide technical
database support.
Mr. Adams has 16 years of experience in developing information systems with Visual Basic, C#,
ASP.NET, multiple SQL based query languages through Visual Studio and Microsoft Visual
Basic IDEs, HTML, XML, and Javascript. He has additional experience in the fields of server-
side controls including Microsoft IIS, SQL Server and MySQL administration. He has a high
proficiency with Microsoft office applications, including Access, Excel, Word, and PowerPoint,
as well as experience with ARC GIS and Google API mapping systems and SAS statistical
software. Mr. Adams also supported NCEA using the Omaha Lead Site blood lead data.
Mr. Fred Ramseyer will support technical writing for the reports and be responsible for
managing the QA log for project deliverables.
Mr. Ramseyer has over 15 years of experience in preparation of technical documents for a
variety of government-sponsored projects, including Section 508 compliant documents, and
assisting in document formatting, editing, and distribution. He also has over 5 years of QA log
experience for governmental project deliverables.
Subcontract Staff
Altci Science & Engineering Staff
Mara Thorhaug, M.P.H., will manage and prepare Bunker Hill data, conduct QC and review of
data sets, assist with IEUBK batch mode modeling, support statistical analyses, and draft project
documents.
Ms. Thorhaug is an environmental health scientist who has managed and provided technical
support on applied public health projects with Alta for more than 11 years. Her experience
ranges from small international environmental site assessments to the nation's largest Superfund
sites. Her experience includes human health risk assessments, remedial effectiveness monitoring

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 9 of 22
and assessment at the Bunker Hill Mining and Metallurgical Complex Superfund Site (BHSS),
quality assurance/quality control (QA/QC), and management of the BHSS environmental health
database. More specifically, Ms. Thorhaug has extensive experience as a Project Manager and
Quality Assurance Manager for Alta at the BHSS. She was the project manager and oversaw the
collection of nearly 100,000 soil, house dust, and drinking water samples from more than 3,000
residential properties in the BHSS.
Sarah Weppner, M.S., will manage the IEUBK batch mode modeling and summaries, assist with
Bunker Hill data set preparation and QC, support statistical analyses, and draft project
documents.
Ms. Weppner is an environmental health scientist with nearly 10 years of experience as a human
health risk assessor and remediation project manager. Ms. Weppner is an experienced
communicator and educator who has conveyed technical information to a wide range of
audiences. She coordinated public involvement activities and developed public information
documents for Idaho communities impacted by hazardous waste sites. She managed a state-wide
environmental health assessment and education program. She managed a number of projects that
included environmental and biological sampling and monitoring for heavy metals and risk
analysis for heavy metals, petroleum, and radionuclides. She has broad experience in human
health risk assessment, environmental health research, and remedial effectiveness monitoring at
the BHSS. She has Comprehensive Environmental Response, Compensation, and Liability Act
(CERCLA) experience at sites in Idaho, Washington, and Montana, and she has conducted risk
assessments in Idaho, Oregon, and Washington. She also has worked in Nicaragua and Nigeria
on the lead poisoning outbreak.
ICF Staff
Dr. Graham Glen will serve as the ICF quality assurance and testing reviewer.
Dr. Glen has over 20 years of experience in exposure modeling and implementation. On the
contracting side, he is the technical lead and primary coder for 1) the Air Pollutants Exposure
(APEX) model used by EPA/OAR for the National Ambient Air Quality rules and 2) the
Stochastic Human Exposure and Dose Simulation (SHEDS) model used by EPA/NERL for
chemical assessment and prioritization. He is currently serving as the quality assurance and
testing reviewer for the Consumer Exposure Model (CEM) for EPA/OPPT and he served as the
quality assurance and testing reviewer for the Excel file used for the EPA/OPPT lead hazard
standard.
A.4.3 Organization
The organizational chart for this project is provided in Figure 1. Mr. Dylan Heh will serve as
SRC's Quality Engineer on this project. Figure 1 shows the lines of communication between all
project members.

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 10 of 22
Figure 1. Project Organization for Advancing Lead Exposure and Biokinetic Modeling
EPA COR
R. Corn
EPA TO COR
J. Brown
SRC Technical
Lead/Task Order
Manager
M. Follansbee
SRC
Technical Staff
B. Thayer
J. Adams
F. Ramseyer
Alta Science &
Engineering
S. Spalinger
NCEA QA Officer
C. Itkin
SRC Quality Engineer
D. Heh
SRC Program
Manager
G. Diamond

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 11 of 22
A.5 Problem Definition/Background
The specific objectives of this task order are to provide services to assist the U.S. Environmental
Protection Agency's (EPA) National Center for Environmental Assessment (NCEA), Office of
Research and Development (ORD), in the development and evaluation of computational models
to estimate the disposition of lead (Pb) in humans in support of the EPA's Superfund Program
and other program offices relying on biokinetic models for Pb. The Integrated Exposure Uptake
Biokinetic (IEUBK) Model for Pb in Children (v. 0.99d) was released in 1994 and has been
widely accepted in the risk assessment community as a tool for site-specific risk assessments
related to childhood Pb exposure. The IEUBK model was designed to assess changes in blood Pb
of children under seven years of age over periods of a month or more. Prior evaluations of the
IEUBK model were completed using datasets where children's blood Pb levels (BLL) exceeded
those currently observed in NHANES surveys. A new draft evaluation version of the IEUBK
model (v. 2.0) has been developed and contains proposed updated default values for some model
variables (e.g., soil/dust ingestion rates) that differ from those used during prior model
evaluations. Decreasing childhood BLL in the U.S. and a change in influential default values in
the IEUBK model (v. 2.0) both mandate an evaluation of the predictive ability of the model prior
to public release. Recognizing the need to model changes in BLL for intermittent Pb exposures
in both children and adults, EPA's Office of Research and Development (ORD), in collaboration
with Office of Chemical Safety and Pollution Prevention (OCSPP), developed an All-Ages Lead
Model (AALM). The AALM is currently implemented in Fortran with an Excel user interface
that provides a tool for rapidly evaluating the impact of possible sources of Pb on Pb
concentrations in blood and other tissues of humans from birth to 90 years of age. The AALM
will allow users to assess the effects of both intermittent Pb exposures and stable exposure
conditions. Efforts are underway by NCEA to have the AALM reviewed by EPA's Science
Advisory Board (SAB). The tasks associated with the project objectives are described below.
A.6 Project Task/Description
Task 1 - Project Management
Subtask la. A Technical Cost Proposal and Staffing Plan was submitted and approved by EPA.
The relevant portions are included below.
Subtask lb. SRC will prepare monthly project status reports that will describe financial status as
well as technical progress made during the month for all tasks. The MPR will also include a
discussion of problems encountered or anticipated that may delay successful completion of the
task, deliverable and/or entire delivery order as well as schedule impacts; corrective actions
taken with regard to the identified problem; planned activities for the succeeding month and
estimated completion dates of all significant milestones, tasks, and/or deliverables.
Task 2. - Kickoff Teleconference
SRC will schedule a 60-minute teleconference call with the EPA TOCOR and the SRC Project

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 12 of 22
Team (relevant staff from SRC, Alta Science & Engineering, and ICF) to discuss all tasks and to
clarify any specific issues. SRC will prepare and distribute meeting minutes of the topics discussed
and any action items agreed to following the conference call. Notes from the call will be QC
reviewed prior to delivery to EPA.
Task 3 - Quality Assurance and Privacy
A project-specific Quality Assurance Project Plan (QAPP) was developed (this document) to
describe the quality assurance (QA) procedures for this project. Specifically, the QAPP describes
data handling and analysis of data from the Bunker Hill site to verify and track associated QC
activities that were conducted during task order performed as well as model development
activities. The QAPP to support this project will be consistent with SRC's Quality Management
Plan submitted to EPA for this contract (EP-C-17-015).
SRC Project Team members (SRC staff and subcontractors) will abide by the EPA privacy policy
governing confidentiality and protection of personal identifiable information (Policy 2151.0:
Privacy Policy https://www.epa.gov/privacy/epa-policy-21510-privacy-policy) as well as any
project-specific privacy and confidentiality requirements associated with the BLL data for the
Bunker Hill Site.
Task 4a - Obtain access to blood Pb and environmental data necessary for IEUBK model
evaluation, develop data files necessary for proposed analyses, and draft report on the
cohort to be used for the proposed analysis
SRC Project Team members from Alta will use paired BLL and environmental media
concentration data from the Bunker Hill Superfund Site in Idaho to evaluate the IEUBK model
(v. 2.0) predictions. Alta will conduct the IEUBK model evaluation and QC using de-identified
data sets that Alta maintains. No individual or confidential data will be transmitted to SRC, ICF,
or EPA. The BLL data used for the analysis will include data for children within the age ranges
in the IEUBK model (i.e., children age 0-84 months). The environmental media concentration
data will include lead information for air, water, soil, and indoor dust (as well as bioavailability
information from soil and dust). All blood lead and environmental data will be QC reviewed by
Alta prior to analysis. A cohort of children from the period 2000 to present (with BLL
measurements linked with residential addresses) will be selected for IEUBK model evaluation.
The evaluation will consider BLLs and Pb concentrations in relevant environment media (soil,
dust, water and air) over time and cross-sectional estimates of Pb relative bioavailability (RBA)
in soil and dust (von Lindern et al., 2016).
After linking BLL with environmental media data for the Bunker Hill Site, SRC will deliver a
brief report summarizing the available data proposed for inclusion in the model evaluation. The
report will provide summaries of children's ages, sex, BLL, and paired environmental media Pb
concentrations grouped by year and basic geographical location (e.g., community level). The
report will include summary statistics on the BLL of the children in the cohort and will compare
average age-group BLLs to those reported in a previous evaluation of the model (Hogan et al.
Environmental Health Perspectives, 106(S6): 1557-1567, 1998). The report will not contain any

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 13 of 22
individual or confidential data. If available data indicate that children were exposed at multiple
locations, SRC Project Team will propose a methodology to time-weight the media
concentrations for input into the IEUBK model. Other summary data (e.g., the dates/season of
blood samples and whether venous or capillary samples) will be provided in the report to assess
the utility of the data for the IEUBK evaluation. In addition, an outline of the overall approach
for conducting the IEUBK model evaluation will be described in the draft report. To the extent
possible, the SRC Project Team will propose approaches similar to those employed by Hogan et
al. (1998) study and von Lindern et al. (2003, 2016), in which comparisons of predicted and
observed BLL were made between communities and at the individual level. All deliverables will
be QC reviewed prior to delivery to EPA.
SRC will deliver a draft report describing all data to be used and approach for IEUBK model
evaluation. As directed, the SRC Project Team will respond to EPA review comments.
Task 4b. - Perform evaluation on data from 4a; draft report describing the results of
model evaluation; perform additional analyses and revise report that includes results of the
analyses
SRC Project Team will use the dataset identified and the evaluation approaches described in
Task 4a to evaluate the IEUBK model's predictive ability for a cohort of children from the
Bunker Hill Site. Batch files for IEUBK will be developed for the comparisons between
observed and predicted BLL according to the approaches utilized by Hogan et al. (1998) and von
Lindern et al. (2003, 2016). All other available site-specific data (e.g., soil bioavailability)
appropriate for the model evaluation will be used for the analysis. Where site-specific data are
not available, default model values will be used. SRC has assumed that no additional data (aside
from the Bunker Hill site data) will be provided. SRC will deliver the batch mode file summaries
used in the IEUBK model analysis to the EPA TOCOR. Any and all data provided by Alta will
maintain confidentiality of the subjects.
SRC has assumed the evaluation of IEUBK model performance relative to the Bunker Hill
exposure and BLL data will include three soil-dust ingestion rate alternatives:
(1) the IEUBK (v. 2.0) default from von Lindern et al. (2016),
(2) the IEUBK (v. 2.0) with soil/dust ingestion rates specified by the EPA Exposure Factor
Handbook (2017), and
(3) the IEUBK (v. 1.1 Build 11) with defaults.
Because of the way soil-dust is partitioned in the risk assessment for the Bunker Hill site, soil-
dust partitions will be handled outside of the IEUBK model (version 2.0) to accommodate the
analyses. One or two additional soil-dust partitions as described in von Lindern et al. (2016) will
be included in the evaluation, such as 40/30/30 (dust/yard soil/community soil) or 50/25/10/15
(dust/yard soil/neighborhood soil/community soil). This will be in addition, to the IEUBK
default assumption of indoor dust lead concentration being 70% of the outdoor residential soil
lead concentration (plus a small contribution from outdoor air lead) and ingestion rate
partitioning using the 55/45 dust-soil ingestion weighting factor.

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 14 of 22
An update to the draft report completed under Task 4a will be developed that describes the
IEUBK model performance using the 3 alternative approaches (as well as discussion of the
potential impact of elimination rates if such information is available). The report will document
other site-specific sources of Pb exposure, such as Pb-based paint, age of the housing stock, and
food (where such information is available) and the extent to which each are quantitatively
accounted for in the model evaluation. Where available, behavioral and demographic data for
children, such as time spent outside or away from home, as well as time-weighted media
concentration (in the event there are not enough data to address this quantitatively) of will be
addressed in the report qualitatively (as a source of uncertainty in the analysis). The report will
include summary tables and figures describing site-specific trends in BLL data over the period
that the data span and comparisons for IEUBK model predictions and observations for the
Bunker Hill data set under the 3 alternative run conditions. All deliverables will be QC reviewed
prior to delivery to EPA.
Task 4c. - Prepare manuscript describing the data, analysis, and results from 4a & 4b
Upon EPA acceptance of the Task 4b deliverable, SRC Project Team will deliver a manuscript
describing the IEUBK model evaluation that conforms to the submission guidelines of
Environment Health Perspectives (or another journal agreed upon between SRC Project Team
and the TOCOR). All deliverables will be QC reviewed prior to delivery to EPA.
Task 5. - AALM interface for batch runs with instructions describing how to use the batch
run module
SRC Project Team will develop a user-friendly interface for batch runs of the AALM using
Microsoft Excel. The batch run interface will support model evaluations of the AALM such as
those described above under Task 4b for the IEUBK model and will be designed to be flexible in
the future to accommodate the complex exposure scenarios possible in the AALM (including
SHEDS distributional inputs and outputs) and output compartmental lead concentration data
described in Task 6a. The output interface of the batch results is required to provide the age and
predicted blood lead of an individual and for quality assurance, a text string providing reference
to the input data such as filename and date of run. SRC will deliver instructions to users
describing how to use the batch run module (not including other technical or scientific aspects of
the AALM). All deliverables will be QC reviewed prior to delivery to EPA.
Task 6a. - Adapt AALM batch run for diverse outputs with instructions describing how to
use the batch run module
SRC Project Team will adapt the batch run module for the AALM developed under Task 5a for
more diverse outputs such as (1) to receive output Pb concentrations in multiple body
compartments (e.g., blood and bone) for a given exposure scenario; (2) to receive output for
multiple body compartments at several discrete ages specified by the user, and (3) to receive
output of average Pb concentration in compartments over some age period (e.g., soft tissue

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 15 of 22
compartments from age 1 to the 6th birthday). The batch run module will be implemented in
Microsoft Excel and include a brief set of instructions describing how to use the batch run
module with these features (not including other technical or scientific aspects of the AALM). All
deliverables will be QC reviewed prior to delivery to EPA.
Task 6b. - Link AALM to SHEDS with instructions describing how to use the batch run
module linking SHEDS and the AALM
SRC Project Team will adapt the batch run module developed under Task 6a to accept input
media concentrations directly from Stochastic Human Exposure and Dose Simulation (SHEDS)-
Multimedia Model (as described in Zartarian et al., 2017 for the IEUBK model). The SRC
Project Team will also extend the batch run capabilities to accept or simulate distributional
inputs and distribution outputs that may be accessed for additional statistical analyses (either in
Microsoft Excel or SAS). SRC Project Team will prepare a brief set of instructions describing
how to use the batch run module linking SHEDS and the AALM (not including other technical
or scientific aspects of the AALM or SHEDS). All deliverables will be QC reviewed prior to
delivery to EPA.
Task 7a. -Support preparation for SAB evaluation of AALM
SRC Project Team will support the AALM peer-review by the SAB, including preparation
and/or review of briefing materials and other documents (if necessary) for the in-person SAB
peer-review meeting of the AALM. All briefing materials will be QC reviewed prior to delivery
to EPA.
Task 7b. -Participate in AALM peer-review meeting
SRC will send one (1) staff member to participate in the SAB peer-review meeting of the
AALM. For the purpose of this cost estimate, the trip will consist of two (2) days and two (2)
nights of travel to the two-day meeting is expected to take place in Washington, DC.
A.7 Quality Objectives and Criteria
SRC Project Team will follow the procedures outlined in the EPA's Performance Work
Statement (PWS; as presented in Section A.6 of this QAPP, which describe how existing data
(i.e., data for the Bunker Hill site maintained by Alta Science & Engineering) relevant to
evaluating the relationship between lead-contaminated environmental media and blood lead
concentration in young children will be used. The methods used for statistical analyses and
modeling/model development will be developed in collaboration with EPA.
Technical QC for this task order pertains to the handling and use of site blood lead concentration
and environmental data. Technical QC will include ensuring confidentiality and privacy are
maintained, as well as ensuring data integrity of information in the database used by Alta Science
& Engineering. Furthermore, SRC Project Team will perform QC to ensure accuracy and
completeness of data analyses and verification of results using spot checks in reviews by senior

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 16 of 22
scientists. All deliverables are subject to QC review (in general, a minimum of 10% of the
mathematically-derived data checked and if errors are identified then all data are reviewed). A
senior toxicologist will review the deliverables and analyses (including obtained from software
programs) to verify accuracy, completeness with respect to project goals and documentation to
ensure reproducibility and adherence to established format.
A.8 Special Training/Certification
All senior SRC project staff have more than 20 years of experience in the use and application of
the IEUBK model for lead risk assessment. SRC project staff also supported the development
and the site-specific application of the IEUBK model software, as well as other lead risk
assessment tools. SRC and Alta Science & Engineering project staff were involved in the Bunker
Hill site risk assessment and are familiar with the unique issues associated with lead risk
assessment modeling and handling of blood lead data. Alta project staff have more than 20 years
of experience managing and maintaining the Bunker Hill data following privacy, confidentiality,
and other security policies and procedures. The Alta Science & Engineering project staff
involved in this project are the same staff that have worked under contract to maintain the project
data. Alta's staff are trained to follow Alta's Data Confidentiality and Privacy Policy and
procedures. These procedures include securing folder access, de-identifying data, as indicated on
the informed participant consent agreements:
I also understand that records will be provided to the Idaho Department of Environmental Quality
(IDEQ) through its contractor Alta Science and Engineering, Inc. as directed by IDEQ. Alta will
analyze the information and de-identify it before sending it on to IDEQ in compliance with HIPAA.
No information containing personal identifiers will be provided directly by Alta to IDEQ, any other
agency, entity or the public.
In addition, SRC and ICF staff have experience with the AALM software.
Senior scientific staff members hold advanced degrees in toxicology or fields relevant to their
role in this project. Oversight will be provided by a senior toxicologist. No additional specialized
training or certification is expected for SRC personnel in order to successfully complete the
outlined tasks under the project.
A.9 Documents and Records
Dr. Mark Follansbee, the SRC Task Order Manager, is responsible for ensuring that all project
personnel are provided with the most current approved version of the QAPP. The QAPP will be
delivered electronically by email after approval and also stored in a public folder on a project
directory of SRC's local area network (LAN) that is accessible to all project personnel. A full
back up of SRC's LAN is performed every Friday evening and differential backups of new data
are run each Monday through Thursday evening.
The following records will be maintained on file indefinitely:
• EPA Task Order PWS

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 17 of 22
• QAPP
• Technical proposal
• Technical monthly progress reports
The project data files (including the database analyses, data summaries, and reports) will be
maintained for a period of three years from project closeout. Alta will ensure that confidentiality
is maintained with project files as directed by the Idaho Department of Environmental Quality.
The Bunker Hill database and data files containing individual data are maintained in a
confidential manner by Alta Science and Engineering on behalf of the Idaho Department of
Environmental Quality. Alta maintains the data in a secure and confidential manner by
employing the following methods:
• Alta's data servers are secured and remain behind a firewall.
• Alta's data folders containing the data are regulated such that only Alta staff with a
business need have access to the data folders. This is regulated by Alta's Corporate
Compliance Officer, Susan Spalinger.
• Project files are coded and de-identified.
• Data are not transmitted within or outside of Alta. Alta maintains project files on the
secured folder locations after Alta receives the data from the local Panhandle Health
District (PHD) via physical hand off of drives. The external drive is maintained in
custody by Alta's trained staff (from the PHD office in Kellogg, Idaho to the Alta office
in Kellogg, Idaho), saved to the secured folder location, then removed from the external
drive.
Adherence to the QA procedures described herein, as well as any problems or issues identified,
will be described in the technical monthly progress reports. These records will be stored on the
limited-access project management folders on SRC's LAN. Periodic assessments of adherence to
QA procedures will be performed by the SRC Program Manager or SRC QA Representative (for
example, see Appendix A). Assessments will be based on the review of project tracking files and
deliverables and, as needed, communication logs and conferences with the EPA TO COR.
B. Data Generation and Acquisition
B.l Sampling Process Design (Experimental Design)
The work products developed and provided in this project will be based on data maintained by
Alta Science & Engineering for the Bunker Hill site. Because the data already exist, sampling
process design and methods do not apply to the collection of these data, so this requirement is
not applicable.
B.2 Sampling Methods
The sampling methods will be developed in conjunction with EPA based on study questions and
objectives specified in the statement of work for the task order. Because the data already exist,

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 18 of 22
sampling methods do not apply to the collection of these data, so this requirement is not
applicable.
B.3 Sample Handling and Custody
The work products developed and provided in this project are based on data maintained by Alta
Science & Engineering for the Bunker Hill site. Sample Handling and Custody do not apply to
the collection of these data, so this requirement is not applicable. SRC Project Team will,
however, ensure that the integrity of the database is maintained by verifying that the number of
records and information in those records does not change by periodically comparing the working
version of the database with the original master maintained by Alta Science and Engineering and
ensuring that only experienced database project staff have access to the database.
B.4 Analytical Methods (Laboratory)
The work products developed and provided in this project are based on data maintained by Alta
Science & Engineering for the Bunker Hill site. Laboratory analytical methods do not apply to
the collection of these data, so this requirement is not applicable.
B.5 Quality Control
As indicated in the EPA Requirements for Quality Assurance Project Plans (EPA QA/R-5), this
requirement addresses QC activities for each sampling, analysis, or measurement technique. No
sampling, analysis, or measurements will be obtained or made in this project, so this requirement
is not applicable.
B.6 Instrument/Equipment Testing, Inspection, and Maintenance
The work products developed and provided in this project are based on data maintained by Alta
Science & Engineering for the Bunker Hill site. No instrumentation or equipment will be used,
so this requirement is not applicable.
B.7 Instrument/Equipment Calibration and Frequency
The work products developed and provided in this project are based on based on data maintained
by Alta Science & Engineering for the Bunker Hill site. No instrumentation or equipment will be
used, so this requirement is not applicable.
B.8 Inspection/Acceptance of Supplies and Consumables
The work products developed in this project may include the generation of hard copy reports.
Only recycled paper will be obtained and used for the hard copy of deliverables generated during
this project.

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 19 of 22
B.9 Non-direct Measurements
This project is to utilize existing data (i.e., environmental media data and blood lead data
maintained by Alta Science & Engineering for the Bunker Hill site) that would be useful for
understanding the relationship between blood lead concentration and sources of lead exposure.
Prior to beginning the analysis, the SRC Project Team will collaborate with EPA to develop data
adequacy evaluation criteria. These criteria will be used to evaluate the data and organize the
analyses to ensure the data quality, quantity and purpose (that is, the intended analyses) are
clearly understood and agreed upon prior to undertaking the work. The analyses will be
conducted in Microsoft Excel or SAS as appropriate. The results of the analyses and the
analytical tools used will be provided to EPA as described in Section A.6 and/or as directed by
EPA.
B.10 Data Management
The work products developed in this project are based on data obtained from the Bunker Hill
site. Experimental data will not be generated in the field or laboratory; therefore, tracking and
archiving of new data will not be needed and this requirement is not entirely applicable.
Documentation of the original version of the database that is used by Alta Science &
Engineering will be maintained to ensure database integrity. SRC will maintain draft and final
deliverables in electronic format that is accessible to all SRC personnel via the SRC intranet. The
SRC Task Order Manager will oversee all components of the retrieval, storage, and maintenance
of retrieved data to assure technical quality.
C. Assessment and Oversight
C.l Assessments and Response Actions
The SRC Task Order Manager (Mark Follansbee), a Ph.D. toxicologist with over 20 years of
experience, will oversee all components of this project to assure technical quality. Technical QC
for this project pertains to review and verification of the process and outcome of existing data
and analyses to answer the study questions as well as review of the results in the report
associated with those results.
The SRC Task Order Manager will ensure that the approved QA/QC procedures, as documented
in this QAPP, are followed. The SRC Task Order Manager will notify the EPA TO COR via
email of any significant difficulties in accomplishing the tasks listed in the PWS. If performance
objectives are not being met, the SRC Task Order Manager will notify SRC's Quality Engineer,
who will identify and oversee implementation of any corrective actions that may be needed. The
SRC Task Order Manager will make every effort to immediately correct any problems to ensure
customer satisfaction. If a problem persists, the SRC Task Order Manager will submit a plan of
corrective action to the EPA TO COR.
The SRC Quality Engineer for EPA contract number EP-C-17-015 (Mr. Heh) is responsible for
conducting periodic audits of activities conducted under this task order to identify areas of risk,

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 20 of 22
as well as to evaluate the overall level of compliance to QA requirements. Quality audits will
include the review of documentation of QC activities. Reports as shown in Appendix A are
generated quarterly for the SRC Quality Engineer for verification and tracking of quality control
activities. The Contract Quality Engineer (or delegate) will contact the Program Manager or Task
Order Manager to determine if they are complying with the requirements of the QAPP. In
addition, the Contract Quality Engineer (or delegate) may review the SharePoint quality tracking
database at any time or other contractor performance reports provided by EPA and contact
appropriate individuals within EPA to determine if the quality of work is meeting expectations or
if problems exist. Based on the results of the audit, SRC will modify its procedures, as necessary,
to increase efficiency of processes in place for this Task Order. Any changes made as a result of
audit findings will also be considered, if applicable, for similar procedures followed under other
Task Orders of EP-C-17-015. If any deficiencies in quality control procedures are identified
during the audit, the Contract Quality Engineer will identify and oversee implementation of any
corrective actions that may be needed. QA audits reports are available to EPA upon request.
C.2 Reports to Management
The SRC Task Order Manager (Mark Follansbee) will have frequent contact with the EPA TO
COR to report on the progress of efforts on this project and any issues that may arise. In addition,
the Task Order Manager will prepare a written progress report each month that summarizes the
work accomplished during the month, including a summary of the deliverables submitted and the
QA/QC performed. Monthly progress reports will be stored indefinitely on the limited- access
project management folders on SRC's LAN (system backups are performed daily).
D. Data Validation and Usability
D.l Data Review, Verification, and Validation
Electronic information obtained by SRC, as described in Section A.6, above, will be accepted
provided the analyses adhere to the analytical plan as outlined in EPA's PWS (as described in
Section A.6 of this QAPP).
D.2 Verification and Validation Methods
The QA Audit Reporting Log that will be used for documenting technical QC and QA
procedures is provided in Appendix A. Technical QC reviews for this project will include the
following elements:
• A senior SRC toxicologist will review all deliverables and analyses to ensure that the
project goals are being met. While quality procedures are tailored to the nature of the
deliverable, all deliverables are subject to QC review (in general, a minimum of 10%
of the mathematically-derived data checked and if errors are identified then all data
are reviewed).
• The SRC Task Order Manager will confirm that QC activities are followed in
accordance with the project objectives outlined in EPA's PWS (as described in

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 21 of 22
Section A.6 of this QAPP). All deliverables will be reviewed by the SRC Task Order
Manager to ensure consistency with the formatting requirements established in
collaboration with EPA.
• SRC staff will log QC activities using a SharePoint tracking tool to document quality
control activities to facilitate review by SRC or EPA. SRC's SOP and an example of
the Project QA Tracking Log are provided in Appendix A.
D.3	Reconciliation with User Requirements
SRC will work in close collaboration with EPA and will adjust procedures based on specific
technical direction from the EPA TO COR.

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 22 of 22
Appendix A. SRC, Inc. Quality Control (QC) and Quality Assurance (QA) SOP and
Documentation
The Quality Control Gate
A senior-level scientist is assigned to review work products prior to delivery. External reviews may
also be conducted depending on the requirements of the contract. This acts as a peer
review, a quality gate through which all work must pass. Any necessary editing,
formatting, or other design activities can occur at this point or as a separate editing
step depending on the type of task. Strict control and preservation (configuration
management) is maintained for revisions made to such products as documents and
databases.
Delivery
Prior to the delivery of a work product, the Work Assignment or Task Order Manager and/or senior
scientist verify that the product is complete and correct. This ensures that the delivered
work product accurately reflects the completed work. Any checklists that are applicable
to the type of product are completed as needed. The final form of the product is
preserved and the product is delivered. Source materials are archived as needed or
as required by the contract.
Quality Assurance
A Quality Auditor checks an EHA project at a minimum of two points during the development
process, depending on the size of the project, in order to verify compliance with the
EHA Process. The first required checkpoint occurs after the creation of the work plan.
The audit verifies that the work plan exists and that it contains the required information.
The audit also verifies that resources/personnel have been allocated to execute the
plan. The second required checkpoint occurs at the end of the project to verify that the
quality control activities have occurred. The following provides an example QA
Activities Report.
QA Audit Reporting Log for Period ():
Project Plan and QAPP are on file
(Yes/No)
Data Management consistent with the QAPP
(Yes/No/Not Applicable)
Technical QC has occurred by senior personnel in
accordance with the QAPP
(Yes/No)
Technical Editing has occurred in accordance with
the QAPP
(Yes/No)
Interim and Final deliverables have been
transmitted and archived in
accordance with the QAPP
(Yes/No)
SRC has adhered to the QAPP during period
(Yes/No)
Signature and Date

-------
QAPP for Advancing Lead Exposure and Biokinetic Modeling for EPA Regulatory Decisions and Site Assessments
Revision No. 2
April 18, 2019
Page 23 of 22
SRC SharePoint QA Tracking System (example)
Itwi 	
Oiw ¦T rss


C«t ^«cv«nK
i i ccu«l *.3 C*-- o - .,
iw i^Nr
Oan ConoMcd
QC toiani
QC Con*f
«K WC Mrrd Dr«3

f&V

1 :*3i 1 1111" *1 Ti "¦—~r~i'Tlir~ TV *f TIbb.iP Tin I'i pnrjhfii i » Cnl
fi»

V-IJWl
!»»•
»
imw»»
V

^5MiA3^]^?4v-«»" in-j3irr?r ±*-zz< «te
ur

ias:i
*r»w
MT

Mt
iimii
7C41
v»

IfiZK
niyii
t*
law
t&
»Hpc:-

k»

Hit*
«Hi MM
M¥
! 145W
iM
HiM<

ftf


"1" JH1*
HF
' WJfW
k#
" V«-X>

k«

" »»+*
nUK*
1*

V#
*nvu




vtHM
if

MP


Iff

tUXil
s :«xvi
>X

if
»'U*M

iff

Hr»3Bi
ta!»Mw
MF




4#

iL .11 3H«
U IIH- 4
yr
li-Miri*
Uf
I2H3H4

UV

Silt*?'
HH5MM
1*
ti-MSif
«r
«i»»u

V

tiBtzrt
IHVU
w

»or
ISWVU

k«

>±31211
» »3H?
»r
IVHKi
s#
• 1PJW.1'

MP


* MM4?
u1
4»«U
!>?




«f«tJ
¦n Mil
kr
i!« Ml
»




± K»i)
«lhXK*
hV
tHTOI

m Iks



~sn^il
-if we

*»*CI
»





ixX7
tar
9.U SSI
u>
»M»S



lUSil



fc*




i»SI»

a*
HM3B1*
«p
i^acM
¦g-'i .' l- r e^J^w'Si'.'i mk3






















mt

*r

•!-[ f»J, t*'! '•- '*


¦«%w
553K*
L»

par

c4in ^vtw K>wSlW6ti|iiPI)iMi)ii^'y#
hTsX.4
~qtftffCTlIN CidltJCaflt*?t-'2j;tafrlg»4ffAg*w>MJJulOtlgBfcV |07SQfl#C-AJKZ.MCBC-
iff

J X 3I>

l*

>?
H»9HI
*2^5-C»* IV1L3J5S ? C 4^
W

iWLifcil

fef

a


ur








va

+WMI*
>»»

IflM
NT

-pjtfr--1 I1
ur
















fi*w<'ini-l>«3)U'56<:^i.1M!i5ilSl9l*r*«4i*ie>elJii2Ai5#3CSfcV>»*-H i >»«.r.•-¦^r^'~








if #• ^-4j1ib>.'rJ *J


¦ XXA
* nscift
«
*1 JIM
w*
Ltim



¦»Si»
!+»»

' k i-
hS

'^a-* >.f: •~ %?.fgs.-INr^r itb»k
w







?:cc*«r»5aiiK'i«^*a**vr*-!iMaMociCi>v*»» lOf.^iii^nCAKs


• 'M
' uku
m

k>
lWH»
• ^..'¦Iry-y4 —--..r-X;C I^s»i r^r-«"
k»

•«

wr
"fa »H
UP
'U Ml*
»»__r-frTat4«|r»*,*»«,+-2hllkjrfV.iT.:5H4.j-'-H>r
ur


05:1
ar
9 £ SS4
k>

f^-V-cT.L-CtC 1; '1^«-~£»-. -
3J

taxi)
v:: ;:h
ft#
rU.*x
&

•ir'-i-e+iVfeiTfTf"' U* i
Iff

lUlik
tftlM
k*
*3 HOI
»
Mk^MSD

t»

::sii

I^A»
;;xh
»*

*ra*_frhit«Dpw a^.*ur ' |. i jjLC


t3XM

Wr*
1 tr-to
A
¦ li .»!•
U.>wm.®>xifjhrii*rC%V!r «.S*K>V,-«.3:-vi
ur



fcfj*

IV
tVXto
ra>ftycng*affl Lx»x-&<7ipg2t3&mw*2AMr>»SMi*K?«ir,Ttv» ¦ [DTUmiE -solmcjsd
AL


•JtMl*
UP*
-Uio«
Ml
vlr JH*
'MH^lfSRU l*C»l
S#


"11 Wi-
MU1

00


V


ll i*;*
w.a
w

raw*

k»



s^r.a
¦ i-»ii



wr



«*
tttWM
VFflO
is*M»*

hTA

. i	; ••jr^n t* i'i IP*
«Ttn3r~fg rttrti
Htjai:	->-i,ca^%rrL*/rS
» M l	>,_0»	^
»-Tir«.	ire.05m.TS li
¦ JSri TO.'^=o^T-^»-»^ir7*.A SJ^B>W^3G3
rs»5«rC^TC^iT»:i3t003^W3»^t«-n'MASLe»,li
httn: < iVnrani e.- r=c /sirr. Ta .v t»
Pckter I IT I •*	j; jjaj;
~Wi {*:««.. •» • *c.onyn a,
•1 .**;NTCaaa*sfcar ' ^n"iVi"fc2DQGC
V,*t '"jaiUw* f iiOar>»33jr^oWVwr •>*"al-»» *
.HAntaMk *U»M
¦ -T»1H

-------
vvEPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
rcecyciea/Kecyciaoie Crimea on paper tnat contains a minimum ot
50% postconsumer fiber content processed chlorine free

-------