Draft Final Report
VOLUME H: DETAILED STATISTICAL RESULTS
COMPREHENSIVE
ABATEMENT
PERFORMANCE
STUDY
To
U.S. Environmental
Protection Agency
March. 1993
-------
DRAFT FINAL REPORT
for the
COMPREHENSIVE ABATEMENT PERFORMANCE 8TDDY
VOLUME II: DETAILED STATISTICAL RESULTS
Battelle Project Team
Bruce Buxton, Steve Rust, Tamara Collins, Fred Todt,
John Kinateder, Dave Burgoon, Nick Basso,
Casey Boudreau, Ben Pierce, Joan Cwi
MRI Project Team
Gary Dewalt, Paul Constant,
Jim McHugh, Jack Balsinger
Battelle
505 King Avenue
Columbus, OH 43201-2693
Midwest Research Institute
425 volker Blvd.
Kansas City, MO 64110
Battelle Contract No. 68-D2-0139
MRI contract No. 68-DO-0137
John Schvemberger and Ben Lim, EPA Task Managers
Gary Grindstaff and Janet Remmers, EPA Project Officers
Technical Programs Branch
Chemical Management Division
Office of Pollution Prevention and Toxics
U.S. Environmental Protection Agency
Washington, DC 20460
-------
ABSTRACT
In 1989 the U.S. Department of Housing and Urban
Development (HUD) initiated the Lead-Based Paint Abatement
Demonstration Study in seven urban areas across the U.S. The
objectives of this study were to assess the cost, worker hazards,
and short-term efficacy of various lead-based paint abatement
methods. Among other conclusions, the FHA portion of this study
estimated that abatement costs for a single-family dwelling could
range from $2000 to $12,000.
One question which was not answered by the HUD
Abatement Demonstration was that of the long-term efficacy of the
abatement methods. Therefore, in 1990 the U.S. Environmental
Protection Agency (EPA) initiated the Comprehensive Abatement
Performance (CAP) Study to address this question. The CAP Study
was a follow up to HUD Abatement Demonstration activities
performed in Denver, Colorado. There were three primary
objectives of the CAP Study: (1) characterize lead levels in
household dust and exterior soil at abated houses and control
houses, (2) assess the long-term efficacy of two primary
abatement methods, and (3) investigate relationships between lead
levels measured in different media and sampling locations. To
address these objectives the CAP Study collected approximately 30
dust and soil samples at each of 53 HUD Demonstration houses in
Denver, approximately two years after the abatements had been
completed. These samples were analyzed for their lead content,
and these lead measurements were then used in a detailed
statistical analysis addressing the three study questions.
This report presents the conclusions from the CAP
Study. There were four primary findings which can be summarized
as follows: (1) Lead levels at abated houses were typically
higher than those at control houses which had always been
relatively free of lead-based paint contamination. Lead levels
were also higher in abated rooms than in control rooms of the
same houses. Therefore, the HUD Demonstration abatement methods
Volume II - Page i
-------
appear to have been unable to reduce lead levels at abated houses
to the same levels found at houses which never had a lead-based
paint problem. (2) However, the HUD Demonstration abatement
procedures were, in many cases, able to reduce lead levels at
abated houses to below HUD interim abatement guidelines at 200
Mg/ft2 for floors, 500 /ig/ft2 for window stools, and 800 nq/ft2
for window channels. (3) Lead Levels were typically higher in
houses abated by encapsulation/enclosure methods than in houses
abated by removal methods. These results may suggest that
encapsulation/enclosure methods are less effective than removal
methods, although houses in the HUD Demonstration Study which
were abated by encapsulation/enclosure methods typically had
greater amounts of lead-based paint requiring abatement than in
houses abated by removal methods. (4) The accuracy of
environmental lead determinations can be seriously affected by
the sampling and analysis protocols used, and by variations in
lead levels among similar housing environments and among
different sampling locations within a single housing environment.
Therefore, assessments of abatement performance and potential
lead hazards must be carefully designed to control for these
complicating sampling variations.
Volume II - Page ii
-------
TABLE OF CONTENTS
Page
1.0 INTRODUCTION AND SUMMARY 1
1.1 STUDY OBJECTIVES 1
1.2 STUDY DESIGN 4
1.3 SUMMARY OF CONCLUSIONS 11
2.0 DESCRIPTIVE STATISTICS 16
2.1 DUST COLLECTED 16
2.2 AREA SAMPLED 20
2.3 LEAD LOADING, LEAD CONCENTRATION, AND
DUST LOADING 22
2.4 DESCRIPTIVE PLOTS 28
2.5 ESTIMATED LEVEL OF DETECTION AND LEVEL
OF QUANTIFICATION 34
3.0 STATISTICAL MODELS 37
3.1 MIXED RANDOM AND FIXED EFFECTS MODEL 37
3.2 CENTERING AND SCALING OF COVARIATES 42
3.3 MODEL SELECTION 46
3.3.1 Phase 1: Abatement Effects
(Stepwise Regression) 48
3.3.2 Phase 2: Non-Abatement Factors
(Stepwise Regression) 48
3.3.3 Phase 3: Mixed Model Screening
(Backward Elimination) 49
3.4 HYPOTHESIS TESTS 50
4.0 MODELING RESULTS 51
4.1 MODELING RESULTS SUMMARY 51
4.2 DETAILED MODELING RESULTS 53
4.2.1 Analysis of Abatement
and Random Effects 53
4.2.2 Analyses of Abatement and
Random Effects by Sample Type 71
4.2.3 Analysis of Non-Abatement Factors 76
4.2.4 Non-Abatement Effects by Sample Type .... 83
5.0 CORRELATIONS 92
5.1 UNIT-TO-UNIT CORRELATIONS 92
5.2 ROOM-TO-ROOM CORRELATIONS 101
Volume II - Page iii
-------
TABLE OF CONTENTS (Continued!
Page
6.0 WIPE VERSUS VACUUM COMPARISON 110
6.1 ALL SUBSTRATES COMBINED 113
6.2 ADJUSTING FOR SUBSTRATE EFFECTS 114
7.0 COMPARISONS WITH OTHER STUDIES 116
7.1 COMPARISON OF CAP STUDY DATA AND CAP
PILOT STUDY DATA 116
7.2 COMPARISON OF CAP STUDY DATA AND HUD
ABATEMENT DEMONSTRATION DATA 120
7.3 COMPARISON OF DUST LEAD LOADINGS BETWEEN
THE CAP STUDY AND OTHER STUDIES 129
8.0 OUTLIER ANALYSES 138
8.1 APPROACH 138
8.2 DATA GROUPS 138
8.3 THE OUTLIER TEST 139
8.4 RESOLUTION OF OUTLIER QUESTIONS 144
9.0 STATISTICAL ANALYSIS OF QUALITY CONTROL DATA 145
9.1 BLANK SAMPLES 146
9.1.1 Field Quality Control 147
9.1.2 Sample Prep Quality Control 152
9.1.3 Instrumental Analysis Quality Control . . . 154
9.2 RECOVERY SAMPLES 155
9.2.1 Sample Preparation Quality Control 155
9.2.2 Instrumental Analysis Quality Control . . . 161
9.3 DUPLICATE SAMPLES 162
9.3.1 Field Quality Control 162
9.3.2 Sample Preparation Quality Control 163
9.4 TIME TREND ANALYSES 165
10. REFERENCES 168
APPENDIX A. CONDENSED DATA LISTING A-l
APPENDIX B. ADDITIONAL EXPLANATORY VARIABLES CONSIDERED
FOR INCLUSION IN THE STATISTICAL MODELS .... B-l
Volume II - Page iv
-------
TABLE OF CONTENTS (Continued)
Page
APPENDIX C. CONCISE LISTING OF STATISTICAL MODELING
RESULTS BY COMPONENT AND MEASURMENT TYPE . . . . C-l
APPENDIX D. LABORATORY CONTROL CHARTS FOR
QUALITY CONTROL SAMPLES ............ D-l
LIST OF TABLES
Table 1-1. Number of Houses Abated in the HUD
Demonstration ................... 5
Table 1-2. Number of Unabated Houses Tested by
XRF in the HUD Demonstration ........... 8
Table 1-3. Summary of Environmental Sampling Planned
for the CAP Study ................ 10
Table 2-1. Symbols Used to Denote Sample Types in
Tables and Figures ...... ......... 17
Table 2-2. Descriptive Statistics for Amounts of Dust
Collected (mg) and Area Sampled (ft2) by
Sample Type ................... 18
Table 2-3. Descriptive Statistics for Lead Loading
(jig/ ft2) , Lead Concentration (pg/g) , and
Dust Loading (mg/ft2) by Sample Type ...... 23
Table 2-4. Correlations of Lead Loading Versus Lead
Concentration for Dust Samples ......... 28
Table 2-5. Estimated Level of Detection by Instrument
Batch ...................... 36
Table 2-6. Potential Instrumental Measurement Error:
Calculated Results ............... 36
Table 3-1. Explanatory Variables That are Significant
for at Least One Sample Type .......... 41
Table 3-2. Average Percent Abated by E/E. Methods,
by Abatement Method Classification for
Interior, Exterior and Room Level
Abatement .................... 44
Volume II - Page v
-------
TABLE OF CONTENTS (Continued)
Page
Table 3-3. Centering and Scaling Parameters for
Model Covariates 44
Table 3-4. Parameter Interpretation After Centering
and Scaling 47
Table 4-1. Estimates of Effects of Primary Abatement Factors
on Lead Loading; Controlling for Significant
Covariates 54
Table 4-2. Estimates of Effects of Primary Abatement Factors
on Lead Concentration; Controlling for
Significant Covariates 55
Table 4-3. Estimates of Effects of Primary Abatement Factors
on Dust Loading; Controlling for Significant
Covariates 56
Table 4-4. Multiplicative Effects of Secondary
Abatement Factors 57
Table 4-5. Multiplicative Effects of Non-Abatement
Factors 78
Table 5-1. Correlations Among Sample Types for Unit-
Level Random Effects: Lead Loading 93
Table 5-2. Correlations Among Sample Types for Unit-
Level Random Effects: Lead Concentration .... 97
Table 5-3. Correlations Among Sample Types for Unit-
Level Random Effects: Dust Loading 98
Table 5-4. Correlations Among Sample Types for Room-
Level Random Effects: Lead Loading 103
Table 5-5. Correlations Among Sample Types for Room-
Level Random Effects: Lead Concentration .... 104
Table 5-6. Correlations Among Sample Types for Unit-
Level Random Effects: Dust Loading 105
Table 6-1. Vacuum versus Wipe Comparison Data: Room
Geometric Mean Floor Lead Loadings (/xg/ft2) . . . ill
Table 6-2. Vacuum/Wipe Multiplicative Bias Estimates .... 115
Volume II - Page vi
-------
TABLE OP CONTENTS (Continued!
Page
Table 7-1. Descriptive Statistics for Floor Dust Lead
Loadings (/tig/ft2) by Abatement Efficacy
Field Study 133
Table 7-2. Descriptive Statistics for Window Stool
Dust Lead Loadings (jjg/ft2) by Abatement
Efficacy Field Study 134
Table 7-3. Descriptive Statistics for Window Channel
Dust Lead Loadings (jig/ft2) by Abatement
Efficacy Field Study 135
Table 8-1. CAP Study Outliers - Field Samples 140
Table 8-2. CAP Study Outliers - Laboratory QC Samples . . . 142
Table 9-1. QC Sample Categorization Matrix 146
Table 9-2. Net Weight Results for Trip and Field Blanks . . 148
Table 9-3. Results of Quality Control Measures Analyses . . 151
LIST OF FIGURES
Figure 2-1. Amount of dust collected (mg) by sample type . . 19
Figure 2-2. Area sampled (ft2) by sample type 21
Figure 2-3. Lead loading (^g/ft2) by sample type 24
Figure 2-4. Lead concentration (jitg/g) by sample type .... 25
Figure 2-5. Dust loading (ing/ft2) by sample type 26
Figure 2-6a. Geometric mean lead loading (jig/ft2), lead
concentration (jig/g), and dust loading
(mg/ft2) by sample type (control units) .... 29
Figure 2-6b. Geometric mean lead loading (/ig/ft2), lead
concentration (mg/ft2), by sample type
(abated units) 30
Figure 2-7. Lead loading (jug/ft2) by sample type and
method of abatement 31
Volume II - Page vii
-------
TABLE OF CONTENTS (Continued)
Page
Figure 2-8. Lead concentration (M9/9) by sample type
and method of abatement 32
Figure 2-9. Dust loading (/xg/ft2) by sample type and
method of abatement 33
Figure 2-10. Floor dust lead concentration vs. substrate ....
Figure 3-1. Total square feet abated indoors vs.
percent encapsulated/enclosed outdoors
(for abated houses) 45
Figure 4-1. Geometric mean lead loading (jug/ft2),
lead concentration (M9/9), and dust loading
(mg/ft2) in control units after controlling
for effects of significant factors 59
Figure 4-2. Estimated multiplicative effects of
abatement from mixed model ANOVA: lead
loading (/ug/ft2) 61
Figure 4-3. Estimated multiplicative effects of
abatement from mixed model ANOVA: lead
concentration (/xg/g) 62
Figure 4-4. Estimated multiplicative effects of
abatement from mixed model ANOVA: dust
loading (mg/ft2) 63
Figure 4-5. Variance component estimates from mixed
model ANOVA: lead loading (ug/ft2) 65
Figure 4-6. Variance component estimates from mixed
model ANOVA: lead concentration (/xg/g) .... 66
Figure 4-7. Variance component estimates from mixed
model ANOVA: dust loading (mg/ft2) 67
Figure 4-8. Foundation soil lead concentration vs.
HUD Demonstration XRF/AAS levels 77
Figure 4-9. Floor dust lead concentration vs.
substrate 80
Figure 4-10. Boundary soil lead concentration vs.
age of house 82
Volume II - Page viii
-------
TABLE OF CONTENTS (Continued)
Page
Figure 4-11. Exterior entryway dust lead concentration
vs. frequency of vacuuming uncarpeted
floors 84
Figure 4-12. Exterior entryway dust loading vs. frequency
of vacuuming uncarpeted floors 85
Figure 4-13. Entryway lead concentrations vs.
proximity to lead smelter 90
Figure 5-1. Scatterplot matrix of unit-level
random effects for different sample
types: lead loading (/ig/ft2) 95
Figure 5-2. Scatterplot matrix of unit-level random
effects for different sample types:
lead concentration (pg/g) 99
Figure 5-3. Scatterplot matrix of unit-level random
effects for different sample types:
dust loadings (/*g/g) 100
Figure 5-4. Scatterplot matrix of room-level random
effects for different sample types:
lead loading (/ig/ft2) 106
Figure 5-5. Scatterplot matrix of room-level random
effects for different sample types:
lead concentration (Mg/g) 107
Figure 5-6. Scatterplot matrix of room-level random
effects for different sample types:
dust loadings (0g/g) 108
Figure 6-1. Vacuum versus wipe comparison: geometric
means of side-by-side floor lead loading
(/ig/ft2) measures 112
Figure 7-1. Comparison of CAP Pilot and CAP Study
results: unit geometric mean lead
loading (/xg/ft2) by sample type 117
Figure 7-2. Comparison of CAP Pilot and CAP Study
results: unit geometric mean lead
concentration (/*g/g) by sample type 118
Volume II - Page ix
-------
TABLE OF CONTENTS (Continued)
Page
Figure 7-3. Comparison of CAP Pilot Study and CAP Study
results: unit geometric mean dust
loading (mg/ft2) by sample type 119
Figure 7-4. Comparison of CAP Pilot Study and CAP Study
results: component geometric mean dust
loadings (mg/ft2) by sample type 121
Figure 7-5. CAP vacuum and CAP wipe vs HUD
Demonstration wipe results: geometric
mean floor lead loadings by room 123
Figure 7-6. CAP vacuum versus HUD Demonstration
wipe results: geometric mean window
lead loadings by room 124
Figure 7-7. CAP vacuum versus HUD Demonstration
wipe results: geometric mean window
channel lead loading by room 125
Figure 7-8. CAP wipe, vacuum, and HUD Demonstration wipe
versus HUD Demonstration XRF/AAS results:
geometric mean floor lead loading
(Mg/ft2) by room 126
Figure 7-9. CAP vacuum and HUD Demonstration wipe versus
HUD Demonstration XRF/AAS results:
geometric mean window stool lead loading
(Mg/ft2) by room 127
Figure 7-10. CAP vacuum and HUD Demonstration wipe versus
HUD Demonstration XRF/AAS results:
geometric mean window channel loading
(Mg/ft2) by room 128
Figure 7-11. CAP versus HUD Demonstration results:
geometric mean foundation soil lead
concentration (Mg/g) by side of unit 130
Figure 7-12. CAP soil concentration (Mg/g) and HUD
Demonstration soil concentration (Mg/g)
versus HUD Demonstration XRF/AAS results:
geometric mean by side of unit 131
Figure 9-1. Individual measurements and tolerance bounds
for M9 lead/sample in blank samples 153
Volume II - Page x
-------
TABLE OF CONTENTS (Continued)
Page
Figure 9-2. Individual measurements and tolerance bounds
for percent recovery in recovery samples .... 158
Figure 9-3. Individual measurements and tolerance bounds
for the ratio of duplicate samples 164
Figure 9-4. Time trend analyses in instrumental detection
level by instrument batch 167
Volume II - Page xi
-------
1.0 INTRODUCTION AND SUMMARY
This report presents detailed results from the
statistical analysis of data collected during the Comprehensive
Abatement Performance (CAP) Study. The objectives and design of
this study are well documented in previous reports (Battelle and
MRI, 1992a,b). In addition, the conclusions from the study are
summarized in a companion document to this report (Battelle and
MRI, 1993). For completeness, the CAP Study objectives, design,
and results are also summarized in this introductory section.
1.1 STUDY OBJECTIVES
The CAP Study is a U.S. Environmental Protection Agency
(EPA) project designed to follow up on the FHA Lead-Based Paint
Abatement Demonstration Study conducted by the U.S. Department of
Housing and Urban Development (HUD). The HUD Abatement
Demonstration is an ongoing research program in ten cities which
is assessing the costs and short-term efficacy of alternative
methods of lead-based paint abatement. A variety of abatement
methods are currently being tested in approximately 120 multi-
family public housing units in three cities — Omaha, Cambridge,
and Albany — and similar methods have already been tested in 172
single-family housing units in the FHA inventory in seven
metropolitan areas — Baltimore, Birmingham, Denver,
Indianapolis, Seattle, Tacoma, and Washington (HUD, 1991). The
FHA demonstration evaluated two classes of abatement methods,
encapsulation and enclosure methods, versus removal methods. The
study found that the cost of encapsulation and enclosure
abatements ranged from about $2000 to $8000 per housing unit,
while the cost of removal abatements ranged from about $2000 to
$12,000 per housing unit (HUD, 1990).
Whereas the HUD Demonstration was intended to focus on
short-term cost-effectiveness of abatement methods, the CAP Study
provides important information about the longer-term effec-
tiveness of these same methods. Although clearance testing of
lead levels in dust was done immediately after abatement in the
Volume II - Page I
-------
HUD Demonstration, the longer-term performance of the abatement
methods after these houses have been reoccupied is not known.
The CAP Study was therefore necessary to preclude spending large
sums of money abating lead-based paint using methods that may
prove in the long term to be ineffective at maintaining low lead
levels in household dust.
High levels of lead in household dust pose serious
health risks to occupants regardless of the source. Therefore
the CAP Study also collected important information as to how
household dust may become contaminated with lead from other
household locations. It is possible that even those houses in
which the lead-based paint hazard has been removed can become
recontaminated after the house is reoccupied. Either prior to
abatement or during the abatement process itself, lead-contami-
nated dust may have been deposited in the ventilation system or
other parts of the house which, when reoccupied by new residents,
could spread contaminated dust throughout the house. Also,
activity patterns of the occupants may re-introduce lead from
contaminated exterior soils.
Therefore, to help address these concerns, the specific
objectives of the CAP Study were as follows:
1. Characterize levels of lead in household dust and
exterior soil for HUD Demonstration and control
homes.
2. Compare abatement methods or combination of meth-
ods relative to performance. Assess whether there
are differences in performance.
3. Investigate the relationship between lead in
household dust and lead from other sources, in
particular, exterior soil and air ducts.
These objectives were intended to address at least
three important concerns presented in the HUD Comprehensive and
Workable Plan (HUD, 1990): the durability of various abatement
methods over time, the importance of adequate dust control during
Volume II - Page 2
-------
the abatement process, and the possible recontamination of
housing units from a variety of locations, such as exterior soil
and air ducts.
The HUD Demonstration intended to eliminate the lead-
based paint hazard from housing environments either by containing
the lead-based paint with encapsulation or enclosure methods, or
by eliminating the lead-based paint with removal methods.
Encapsulation and enclosure methods attempt to chemically bond or
mechanically affix durable materials over painted surfaces, while
removal methods attempt to either scrape or chemically strip
lead-based paint from painted surfaces, or to completely remove
and replace painted components (e.g., windows, doors,
baseboards).
There are two primary performance concerns with these
abatement methods. First, conducting the abatement methods
themselves might generate large amounts of leaded dust that could
significantly contaminate the housing environment. And second,
the performance of the abatement measures might degrade over
several months or years following abatement, allowing the lead
hazard to be reintroduced to the housing environment. Encapsula-
tion and enclosure methods do not attempt to remove lead-based
paint from housing surfaces and therefore may have a lesser
potential to contaminate the housing environment during the
actual abatement process. However, these two types of abatement
may be more susceptible to degradation over time. In contrast,
removal methods do attempt to remove lead-based paint from
housing surfaces and therefore may have a greater potential to
contaminate the housing environment during abatement. This
contamination might be seen shortly after abatement, or its
effects might be seen more gradually over time.
The approach in the CAP Study was to collect
environmental samples several months after abatement both inside
and outside housing units abated by HUD, and at the same time at
control houses known to be relatively free of lead-based paint,
and measure the concentration of lead in these samples. The
Volume II - Page 3
-------
performance of the abatement methods was assessed by comparing
the lead concentrations at abated houses with those at control
houses. Sampling at control houses provided a measure of the
amount of lead introduced to the housing environment from sources
other than lead-based paint abatement. If the environmental lead
concentrations at abated houses were found to be significantly
higher than those at control houses, this was taken as an
indication that abatement failed to completely eliminate the lead
hazard because lead was introduced to these environments either
immediately through inadequate dust control during abatement, or
more gradually through recontamination over time.
1.2 STUDY DESIGN
In 1989 and 1990 HUD conducted abatement demonstrations
in 172 single-family dwellings from the inventory of FHA repos-
sessed houses in seven urban areas. Three of these houses had
only pilot abatements performed, while the other 169 were com-
pletely abated. The distribution by city of these 169 houses is
presented in Table 1-1. The specific units for abatement were
selected by first identifying older housing likely to contain
lead-based paint and then testing painted surfaces for lead using
portable x-ray fluorescence (XRF). Units included in the Demon-
stration were those found to have a large number of structural
components covered by paint with a high concentration of lead.
When surveying houses for lead-based paint, HUD considered all
painted surfaces both on the interior and exterior of the house.
The HUD Demonstration originally considered six differ-
ent abatement methods: encapsulation, enclosure, and four
removal methods (i.e., chemical stripping, abrasive stripping,
heat-gun stripping, and complete removal or replacement of
painted components). Because of the diversity of housing compo-
nents containing lead-based paint, it was generally true that no
single abatement method could be used uniformly throughout a
given housing unit. One important consideration in the CAP Study
was the appropriate way in which to summarize and classify the
Volume II - Page 4
-------
Table 1-1. Number of Houses Abated in the HUD Demonstration
Interior
Exterior
Abatement Category*
<
o
|
0
H
1
•0
Ql
(D
Ui
City
Baltimore
Birmingham
Denver
Indianapolis
Seattle/Tacoma
Washington
TOTAL
*Each house is
square footage
Encap/Enclos
11
8
33
17
12
6
87~
classified according to
of interior abatement.
Removal
9
12
18
10
10
3
62
Abatement Only
Encap/Enclos Removal
—
2
5
3
1
—
11
the abatement category
—
1
1
4
3
— -
~9
accounting for the
TOTAL
20
23
57
34
26
9
169
largest
-------
abatement activities conducted at each house. Detailed
information was collected by HUD which lists each type of
interior and exterior structural component abated in the
Demonstration, along with the linear or square footage abated and
the abatement method used. For the CAP Study, each house was
primarily classified according to the abatement category (i.e.,
encapsulation/enclosure versus removal methods) accounting for
the largest square footage of interior abatement. However, at
many HUD Demonstration houses, a great deal of exterior abatement
was also performed, and another objective of the CAP Study was to
assess the performance of this exterior abatement. Therefore,
the data interpretation also considered which specific methods
were used on both the interior and exterior of the house.
Initial plans for the CAP Study included selection of
housing units from all seven urban areas in the FHA portion of
the HUD Demonstration. However, after conducting a pilot sam-
pling and analysis program (Battelle and MRI, 1991), and
subsequently developing a cost estimate for the CAP Study, it was
decided that the CAP Study would only be conducted in Denver,
where 57 of the 169 abated units are located (Table 1-1).
Because the number of abated houses in Denver was limited, all
reoccupied houses were included for recruitment in the CAP Study.
Results from a preliminary statistical power analysis indicated
that approximately 40 abated houses would be sufficient to detect
meaningful differences between the dust lead levels in abated and
control houses. Given the initial set of 57 abated houses in
Denver, 70% of these units had to be successfully recruited into
the study.
In order to use the levels of lead measured in dust and
soil samples at abated houses as a measure of the performance of
abatement at those houses, it must be demonstrated that the lead
found is not associated with other environmental sources.
Therefore, in addition to abated houses, dust and soil samples
were collected from control houses that were previously tested by
XRF in the HUD Demonstration and found to be relatively free of
Volume II - Page 6
-------
lead-based paint both inside and outside the house. The
objective in measuring lead levels at control houses was to
determine whether lead levels observed at abated houses were in
fact greater than those found at houses having very few
components covered with lead-based paint (and therefore affected
by only non-paint sources of lead).
In the FHA portion of the HUD Demonstration, a total of
304 houses of similar age and construction were tested by XRF for
lead-based paint, and 172 were abated. Of those houses not
abated, 37 houses were found to be free of structural components
covered with lead-based paint (Table 1-2). When performing the
XRF tests, HUD took three replicate XRF readings at each sampling
location and based their decisions at each location on the
average of those three readings. When interpreting the results,
HUD considered an average reading greater than 1.0 mg/cm2 to be a
positive indication that lead-based paint was covering the tested
component. While only a single round of XRF testing was per-
formed at unabated houses, in some cases a second round of XRF
and/or AAS testing was performed at abated houses to confirm
inconclusive XRF results.
Control houses for the CAP Study were recruited from
the set of unabated houses in Denver tested by XRF in the HUD
Demonstration. For the purposes of identifying control houses,
the detailed XRF results supplied by HUD were used under the
assumption that they provide an accurate and current assessment
of these houses. Results from a statistical power analysis
indicated that approximately 20 control houses were required to
detect two-fold multiplicative differences between the dust lead
levels in abated and control houses. Using a criterion that
equally weights (l) the percentage of housing components testing
positive by XRF for lead-based paint, and (2) the average XRF
testing result, the 20 unabated houses (out of 40) in Denver with
the lowest XRF results were prioritized and selected as control
houses for the CAP Study.
Volume II - Page 7
-------
Table 1-2. Number of Unabated Houses Tested
by ZRF in the HUD Demonstration
City
Baltimore
Birmingham
Denver
Indianapolis
Seattle/Tacoma
Washington
TOTAL
0
1
4
13
5
10
4
37
Number
1-2
6
5
10
9
3
2
35
Of
3-9
3
—
14
5
2
4
2?
LBP Components*
10 or more
10
5
3
—
5
9
32~
TOTAL
20
14
40
19
20
19
I3~2
*Number of structural components for which XRF testing identified
the presence of lead-based paint.
Volume II - Page 8
-------
During the CAP Study a variety of environmental samples
were collected along with questionnaire and field inspection
information to help assess the performance of abatement methods
used in the HUD Demonstration. The environmental samples that
were collected are summarized in Table 1-3. All samples were
chemically analyzed to measure the amount of lead present. The
results for vacuum dust samples are presented on both a
concentration basis (i.e., micrograms of lead per gram of dust,
Mg/g) and a loading basis (i.e., micrograms of lead per unit area
sampled, /jg/ft2). The results for wipe dust samples are
presented only on a loading basis, and the results for soil core
samples are presented only on a concentration basis.
The environmental sampling planned for the study,
included both regular samples (vacuum dust and soil cores) and
field quality control samples (wipe versus vacuum dust, blanks,
and side-by-side samples) intended to assess sampling variability
and potential sample contamination. The role of each type of
sample listed in Table 1-3 for meeting these objectives is as
follows:
Vacuum dust from floor perimeter and window stools
— Provides primary measure of performance for
interior abatement;
Vacuum dust from window channels — Provides
measure of performance for interior abatement,
possible measure of performance for exterior
abatement, and possible transport of exterior soil
from outside to inside the house;
Air ducts — Provides measure of source contribu-
tion to interior dust lead levels;
Interior and exterior entryway floor — Provides
measure of possible transport of exterior soil
from outside to inside the house;
Soil cores — Provides primary measure of perfor-
mance of exterior abatement, and measure of possi-
ble transport of exterior soil lead into the
house.
Volume II - Page 9
-------
Table 1-3. Summary of Environmental Sampling Planned
for the Cap Study
Number of Samples Planned
For 18 For 22 For 13
Sample Type Control Units Abated Units Abated Units
Regular Samples
1. Vacuum dust
a. Perimeter floor 223
b. Window channel 223
c. Window stool 223
d. Air ducts 223
e. Int. entryway floor 222
f. Ext. entryway concrete 222
2. Soil cores
a. Near foundation 222
b. Property boundary 222
c. Entryway 222
Quality Control Samples
3. Wipe vs. vacuum
4.
5.
a. Floor wipe dust
b. Floor vacuum dust
Blanks
a. Vacuum dust field blank
b. Vacuum dust trip blank
c. Soil core field blank
d. Wipe dust field blank
Side-by-side samples
a. Vacuum dust floor
b. Soil cores
Total samples
0
0
1
1
1
0
1
1
23~
2
2
1
1
1
1
1
1
~28
2
2
1
1
1
1
1
1
32
(a) 22 units where sampling was conducted in two rooms.
(b) 13 units where sampling was conducted in three rooms.
Volume II - Page 10
-------
Wipe versus vacuum dust from floors — Provides
consistency check against earlier results from HUD
Demonstration and other studies;
Blanks — Provides assessment of potential sample
contamination and uncertainty in sample weighing;
and
Side-by-side samples — Provides assessment of
sampling variability.
1.3 SUMMARY OP CONCLUSIONS
The detailed results of the CAP Study are discussed in
Sections 2 through 9 of this report. This section presents an
overall summary of the most important conclusions. These
conclusions are listed below according to the study objective to
which they pertain.
The first objective of the CAP Study was to
characterize lead levels in household dust and exterior soil for
HUD Demonstration and control houses. The lead levels found in
the CAP Study are summarized in Section 2, as well as in the
following points:
Geometric mean dust lead loadings varied from a
low of 11 /xg/ft2 for floor wipe samples to a high
of 2516 /zg/ft2 for window channel vacuum samples.
Individual dust lead loadings varied greatly.
Minimum lead loadings for all sample types were
typically only a few ng/ft.2. Maximum lead
loadings were lowest for floor wipe samples (334
/ug/ft2) and highest for window channel vacuum
samples (244,581 j*g/ft2) .
Geometric mean lead concentrations varied from a
low of 125 nq/q for boundary soil samples to a
high of 1439 ng/g for window channel vacuum
samples.
Individual lead concentrations also varied
greatly. Minimum lead concentrations for most
sample types were typically on the order of 10
Maximum lead concentrations were lowest for
Volume II - Page 11
-------
boundary and entryway soil samples (on the order
of 1000 /ig/g) and highest for window stool and
window channel vacuum samples (on the order of
50,000 /xg/g).
The second objective of the CAP Study was to compare
the performance of different abatement methods used in the HUD
Demonstration Study. Abatement performance is discussed in
Section 4, and can be summarized with the following points:
Lead levels were often found to be higher in
abated houses than in control houses. The most
significant differences in dust lead loadings were
found for air duct vacuum samples (4.7 times
higher in abated houses than control houses),
exterior entryway vacuum samples (2.2 times
higher), and floor vacuum samples (1.8 times
higher). The most significant differences in lead
concentrations were found for air duct vacuum
samples (2.1 times higher).
For every sample type where lead levels were
higher in abated houses than in control houses,
the lead levels were also higher in houses abated
by encapsulation/enclosure methods than in houses
abated by removal methods. The most significant
differences in dust lead loadings were found for
window stool vacuum samples (4.6 times higher in
encapsulation/enclosure houses than removal
houses), air duct vacuum samples (4.0 times
higher), and floor vacuum samples (1.9 times
higher). The most significant differences in lead
concentrations were found for window stool vacuum
samples (2.7 times higher) and air duct vacuum
samples (2.0 times higher). When interpreting
these results it should be noted that
encapsulation/enclosure houses typically had
larger amounts of abatement performed than removal
houses. Therefore, the differences in lead levels
noted above may be largely a result of the more
severe initial conditions in encapsulation/
enclosure houses, that is, the greater amount of
abatement required in encapsulation/enclosure
houses.
Lead levels were often lower in control rooms of
abated houses (i.e., rooms that did not require
abatement) than in abated rooms of these same
Volume II - Page 12
-------
houses, although the differences observed were
only of marginal significance. The most
significant differences in dust lead loadings were
found for window channel vacuum samples (0.53
times lower in control rooms than abated rooms),
window stool vacuum samples (0.56 times lower),
and floor vacuum samples (0.57 times lower). The
most significant differences in lead
concentrations were found for window stool vacuum
samples (0.65 times lower) and window channel
vacuum samples (0.66 times lower).
A number of secondary abatement factors were found
to be important for differentiating lead levels in
some particular sample types. For example, both
lead loadings and lead concentrations in window
stool samples differed significantly depending on
the phase of the HUD Demonstration Study during
which abatement was performed. Other factors
considered were the amount of interior and
exterior abatement performed, the specific
abatement methods used, the abatement contractor,
the time period between HUD Demonstration
clearance testing and CAP Study sampling, and the
HUD Demonstration XRF readings.
A number of non-abatement factors were also found
to be important for differentiating lead levels in
some sample types. For example, lead loadings in
floor vacuum samples differed significantly
depending on the substrate type. Other non-
abatement factors considered were the frequency
with which surfaces were typically cleaned, the
resident's occupation and activities, the age and
location of the housing unit, the number of
children living at home, and the number of months
the current residents had been living at the
house.
Soil lead concentrations were 18 percent higher at
abated houses than at control houses, but this
difference was not found to be significant.
However, foundation soil lead concentrations were
found to be significantly higher than entryway and
boundary soil lead concentrations, although this
finding was shared by both abated and control
houses. Interestingly, foundation soil lead
concentrations were significantly lower at houses
where more than an average amount of exterior
abatement was performed. Houses where twice as
much exterior abatement was performed had 35%
lower foundation soil lead concentrations. In
Volume II - Page 13
-------
addition, there was a strong positive correlation
found between foundation soil lead concentrations
and exterior XRF readings made during the HUD
Demonstration.
The third objective of the CAP Study was to investigate
the relationship between lead levels in different media (i.e.,
dust and soil) and different sampling locations (e.g., floors,
window channels, foundation soil). Correlations among observed
lead levels are discussed in Section 5, and can be summarized
with the following points:
At the house level, significant correlations in
dust lead loadings were found for two pairs of
sample types. The most significant correlations
were found between lead loadings in window
channels and window stools (correlation
coefficient of 0.64), and in air ducts and
exterior entryways (0.41).
Significant correlations in lead concentrations at
the house level were found for three pairs of
sample types. Significant correlations were found
between lead concentrations in window channels and
window stools (0.44), in entryway soil and
boundary soil (0.53), and in entryway soil and
foundation soil (0.37).
At the room level, significant correlations in
dust lead loadings was found for window channel
and window stool samples (0.33). In addition,
marginally significant correlation in lead
loadings was observed for window stool and floor
vacuum samples (0.15).
Significant correlations in lead concentrations at
the room level were found for three pairs of
sample types. Significant correlations were
observed between lead concentrations in air duct
and window stool samples (0.43), in interior and
exterior entryway samples (0.37), and in window
channel and window stool samples (0.32).
Volume II - Page 14
-------
A final objective of the CAP Study, which was not
originally stated at the study design stage but which evolved
during the course of the study, was to compare the performance of
two dust sampling protocols: cyclone vacuum sampling and wipe
sampling. The results of this comparison are discussed in
Section 6, and can be summarized as follows:
For comparative tests where side-by-side samples
of dust were collected from various substrates,
lead loadings measured by cyclone vacuum sampling
were 42% higher than those measured by wipe
samples. This difference was not found to be
significant.
The performance of these two sampling protocols
was found to be different for different
substrates. On linoleum surfaces cyclone vacuum
lead loadings were found to be equivalent to wipe
lead loadings. Cyclone lead loadings were higher
than wipe lead loadings on wood surfaces (4.2
times higher) and concrete surface (12.2 times
higher). On tile surfaces cyclone lead loadings
were 0.69 times lower than wipe lead loadings.
Volume II - Page 15
-------
2.0 DESCRIPTIVE STATISTICS
As is noted in the previous chapter, ten types of
samples were collected at each housing unit in this study.
Vacuum dust samples were collected from air ducts, entryways,
floors, window stools, and window channels within each unit. For
purposes of comparison, wipe dust samples were also collected
from floors. Core soil samples were obtained at the boundary of
the property, the foundation of the unit, and an entryway to the
unit. In the analyses that follow, abbreviations are employed in
identifying these various sample types. The abbreviations we
selected are displayed in Table 2-1.
2.1 DUST COLLECTED
When interpreting results of a field study which
included vacuum dust samples in a residential setting,
information about the amount of dust collected is important. The
detection limit for dust lead concentration is a direct function
of the amount of dust collected. In Table 2-2, descriptive
statistics are reported by sample type for the amount of dust
collected (mg) by the vacuum sampling method. The statistics
presented are the number of samples, geometric mean, logarithmic
standard deviation, minimum, and maximum. The amount of dust by
sample type is illustrated graphically in Figure 2-1. In this
figure, box and whisker plots are displayed for the logarithm of
dust collected by sample type.
Box and whisker plots illustrate the center, scatter,
and skewness of a dataset. The lower and upper guartiles of the
data are represented by the bottom and top of the box,
respectively. The distance embodied by the box is termed the
interquartile range, the range from the 25th to 75th percentile.
The bar within the box portrays the median of the data. The
lower and upper tails of the distribution of the amount of dust
are represented by the whiskers extending from the bottom and top
of the box. Extreme data points are classified as either minor
Volume II - Page 16
-------
Table 2-1. Symbols Used to Denote Sample Types in Tables and Figures
Sample Type
Air Duct Dust
Window Channel Dust
Window Stool Dust
Floor Dust
Entryway Dust
(Interior)
Entryway Dust
(Exterior)
Soil
Symbol
ARD
WCH
WST
FLW
FLR
EWI
EWO
EWY
FDN
BDY
Description
Vacuum dust samples collected from an air duct within
the unit
Vacuum dust samples collected from a window channel
within the unit
Vacuum dust samples collected from a window stool within
the unit
Wipe dust samples collected from a floor within the unit
Vacuum dust samples collected from a floor within the
unit
Vacuum dust samples collected from inside an entrvwav to
the unit
Vacuum dust samples collected from outside an entryway
to the unit
Core soil samples collected adjacent to an entrvway to
the unit
Core soil samples collected at the foundation of the
unit
Core soil samples collected at the boundary of the
property
(D
H
H
I
TJ
0>
iQ
(D
-------
Table 2-2. Descriptive Statistics for Amounts of Dust Collected (ing)
Area Sampled (ft2) by Sample Type
and
STATTYPE
Air Duct
(Vacuum)
Window
Channel
(Vacuum)
Window
Stool
(Vacuum)
Floor
(Wipe)
Floor
Entryway
Interior
Awunt of Dust fan)
Umber of Samples
Arithmetic Mean
Geometric Mean
LN Standard Deviation
Minimun
Max i nun
Number of Samples
Arithmetic Mean
Standard Deviation
Minimun
Maximum
Entryway
Exterior
109
355.42
95.49
1.68
2.20
4215.10
109
0.43
0.26
0.03
1.44
98
1324.36
617.08
1.43
0.50
13285.80
98
0.52
0.41
0.05
1.83
113
174.11
89.22
1.18
2.30
2299.40
•-•"» »°-r'^ (ft2
113
0.90
0.63
0.11
4.73
0
.
I
67
1.00
0.01
0.96
1.00
238
572.12
180.81
1.65
40.60
14426.00
238
1.00
0.03
0.96
1.40
100
2880.35
1112.18
1.66
8.50
20857.40
100
1.00
0.03
0.67
1 00
97
3081.30
1583.29
1.30
40.60
22170.30
97
0 98
0.07
0.50
1 00
§
(D
H
H
TJ
D)
vQ
(D
09
-------
10000O.O
10000.0
1000.0
(D
I
1OO.O
Q>
iQ
(0
I-1
ID
10.0
1.0
0.1
Sample Type
Figure 2-1. Amount of dust collected (mg) by sample type,
-------
(pluses) or extreme (stars) outliers based on their distance from
the quartiles relative to the interquartile range. The
arithmetic mean amount of dust is displayed as a diamond.
The amount of dust collected by the vacuum sampler was
seldom less than 10 mg (the amount targeted by the laboratory
chemists in the study plans), and never exceeded 25 grams (25000
mg). The geometric mean amount of dust for each sample type was
at least 90 mg. Problems in collecting air duct samples resulted
in their surprisingly small amount of dust. The large amount of
dust collected from window channels is due to a very high dust
loading which compensates for the very small area available for
sampling (less than for window stool samples).
2.2 AREA SAMPLED
The square footage sampled when collecting vacuum and
wipe dust samples is useful for interpreting the resulting lead
loadings and concentrations. In Table 2-2, descriptive
statistics are reported by sample type for the area sampled (ft2)
by both the vacuum and wipe sampling methods. The number of
samples, arithmetic mean, standard deviation, minimum and maximum
are reported. These results are illustrated in Figure 2-2 via
the box and whisker plots of area sampled for each sample type.
With only a few exceptions, one square foot of dust was
sampled when the entryway-interior, entryway-exterior, floor
(vacuum), and floor (wipe) measures were collected. The area
sampled during the collection of air duct, window stool and
window channel measures, however, varied considerably. In the
case of window stools, as little as 0.1 ft2 to nearly as much as
5 ft2 were sampled. Since the sampling protocol called for
collecting dust from the entire window stool or channel, the
variation is mostly a function of differences in the construction
of the units. For example, a window stool in unit 44 was 47
inches long and 14.5 inches wide, while a window stool in unit 95
was 63.5" x 7.9". The average area sampled for air duct and
window channel measures was approximately 0.4 ft2 while an
Volume II - Page 20
-------
QJ
•8
ro
Sample Type
Figure 2-2. Area sampled (ft2) by sample type.
-------
average of approximately 0.9 ft2 was sampled for window stool
measures.
2.3 LEAD LOADING. LEAD CONCENTRATION. AND DUST LOADING
Three types of measurements were collected for the dust
and soil samples. They are:
Lead Loading: Amount of lead (/*g) in household dust
per square foot (ft2) of surface area sampled.
Lead Concentration: Amount of lead (/ng) per gram (g)
of household dust sampled or amount of lead (/xg) per
gram (g) of soil sampled.
Dust Loading: Amount of household dust (mg) per square
foot (ft2) of surface area sampled.
Vacuum dust samples produce all three measurements. Wipe dust
samples produce only lead loading measurements since the amount
of dust collected cannot be determined. For soil samples, only
lead concentration may be determined because only a point, not a
surface, is sampled.
Descriptive statistics for all units combined are
presented by sample type in Table 2-3 for all three measurement
types. The descriptive statistics reported include the number of
samples collected, geometric mean, arithmetic mean, logarithmic
standard deviation, minimum and maximum. Figure 2-3 displays box
and whisker plots for lead loading across all units plotted
versus sample type. Comparable plots for lead concentration and
dust loading are presented in Figures 2-4 and 2-5, respectively.
Log-transformed responses (lead loadings, lead
concentrations, and dust loadings) were used in all of the
statistical analyses. Using log-transformed environmental lead
measures is common and supported in the literature. Reeves et al
(1982) found that the normal distribution was statistically
rejected for each of the environmental measures they studied
(lead in paint, soil, and house dust). Further, the data were
found to be closer in form to the lognormal distribution than
Volume II - Page 22
-------
Table 2-3. Descriptive Statistics for Lead Loading (fig/ft2), Lead Concentration (M9/9)
and Dust Loading (mg/ft2) by Sample Type
Measurement
Lead Loading
Oig/ft2)
Lead
Concentration
(M9/9)
Dust Loading
-------
S
H
13
PI
4
-------
100000
10000
1000
I
*
*
<
M
I
ro
NJ
Ul
100:
10
1
Sample Type
Figure 2-4. Lead concentration (M9/9) by sample type
-------
13
o>
ua
(D
to
100000
10000
1000
100
10
1-
Sample "P/pe
Figure 2-5. Dust loading (mg/ft2) by sample type.
-------
the normal distribution. The data obtained in this study
illustrate another important reason for using log-transformed
data - the responses range over four to five orders of magnitude.
In addition, the geometric means are often much closer to the
medians than the arithmetic means (illustrated in Figures 2-3, 2-
4 and 2-5). This is evidence that the distributions are more
symmetric on a log scale than a linear scale.
The geometric mean and logarithmic standard deviation
are natural summary parameters for lognormally distributed data.
The geometric mean is calculated by taking the natural logarithm
of the data values, calculating their arithmetic mean, and
exponentiating (taking the antilog). The logarithmic standard
deviation, in turn, is determined by taking the natural logarithm
of the original data values and then calculating their standard
deviation.
The correlations among lead loadings, lead
concentrations, and dust loadings were assessed for the six types
of vacuum dust samples collected. Table 2-4 displays these
estimated correlations for each type of sample. These estimates
are based on the log-transformed data. For all six sample types
the estimated correlations between lead loadings and lead
concentrations, and lead loadings and dust loadings were
significantly different from zero. In contrast, the estimated
correlations between lead concentrations and dust loadings were
not significantly different from zero for any of the sample
types. The estimated correlations between lead and dust loadings
were higher than those between lead loadings and lead
concentrations, except for window stool and channel samples.
When the samples are pooled across sample types all the average
correlations are significantly different from zero. The average
estimated correlation among lead concentrations and dust loadings
(0.12), however, is smaller than those among lead and dust
loadings (0.82), and lead loadings and lead concentrations
(0.67).
Volume II - Page 27
-------
Table 2-4. Correlations of Lead Loading versus Lead
Concentration for Dust Samples
Sample Type
Air Ducts
Window Channel
Window Stool
Floor
Entryway Interior
Entryway Exterior
Across Sample Types
Number
of
Samples
109
98
113
238
100
97
755
Estimated Correlation
Pb Load
vs
Pb Cone
0.50*
0.76*
0.84*
0.58*
0.56*
0.66*
0.67*
Pb Load
vs
Dust Load
0.92*
0.66*
0.70*
0.83*
0.86*
0.79*
0.82*
Pb Cone
vs
Dust Load
0.12
0.002
0.19
0.02
0.05
0.07
0.12*
* Significant at the 0.01 level.
2.4 DESCRIPTIVE PLOTS
Figures 2-6a and 2-6b present the geometric mean lead
loading, lead concentration, and dust loading results by sample
type for control units and abated units, respectively. These
plots facilitate assessment of the three types of measurements
and their relationship across sample and unit types. With a
single exception (entryway outside dust loading in abated units),
the highest loadings and concentrations were obtained from window
channel samples. Similar geometric mean lead concentrations were
measured for the three soil sample types, though foundation
samples from abated units were highest.
The environmental efficacy of the abatement procedures
may be roughly assessed by examining Figures 2-7, 2-8, and 2-9.
In Figure 2-7, the geometric mean lead loading for control,
predominantly encapsulated/enclosed, and predominantly removal
units are displayed by sample type. Figures 2-8 and 2-9 present
similar bar charts for lead concentration and dust loading,
respectively. For all sample types, the predominantly
Volume II - Page 28
-------
10000
I
(D
1OOO
• 100
^
^ ^
ID Lead Loading 1N\\1 Lead Concen.
Diwt Loading
Figure 2-6a. Geometric mean lead loading (/xg/ft2), lead concentration (M9/9) / and
dust loading (mg/ft2) by sample type (control units) .
-------
10000
H
H
2?
•8
OJ
o
1000
100
Y/////////////////////////////X
—
—
Y//////////////////////X
i
Y//////////////////////A
V////////////////////////S
V/////////////////////A
X
Lead Loading i\\\i Lead Concert.
Duat LxMdlng
Figure 2-6b. Geometric mean lead loading (/ig/ft2), lead concentration (Mg/g) » and
dust loading (mg/ft2) by sample type (abated units).
-------
10000
1OOO
<
o
n>
H
H
•O
ft)
^Q
0>
100
Sample iype
Figure 2-7. Lead loading (/xg/ft2) by sample type and method of abatement.
-------
10000
1000
100
0>
iQ
(D
to
to
Sample Type
Control
LVsVI Encap/Enctos
Removal
Figure 2-8. Lead concentration (/xg/g) by sample type and method of abatement.
-------
10000
o
M
H
I
V
0>
u>
U)
1000
100
10
I
Samp)» TVP«
Control
Enosp/Enotov
Removal
Figure 2-9. Dust loading (jig/ft2) by sample type and method of abatement.
-------
encapsulated/enclosed units exhibited the highest geometric mean
lead concentrations. The geometric mean lead concentration for
predominantly removal units are usually higher than for control
units, with the exception of air ducts and entryway exterior
samples. This pattern is not duplicated in either the lead or
dust loading results. The results for control units, however,
were usually lowest. A striking exception is evident for window
channel samples. The geometric mean lead and dust loading
results for window channels are higher for control units than for
predominantly encapsulated/enclosed units.
2.5 ESTIMATED LEVEL OF DETECTION AND LEVEL OF QUANTIFICATION
In order to assess the validity of the lead
concentration and lead loading results reported, it is
appropriate to discuss the sensitivity of the laboratory
procedures employed. This assessment may be performed by
considering two parameters of sensitivity, the estimated level of
detection (ELOD) and the level of quantification (LOQ). Both
parameters are stated in terms of the instrument response
concentration, which is the amount of lead (jig) per dilution
volume (mL) in instrument samples. The ELOD is a practical upper
bound on the estimated concentration (/ig/mL) that would result
from the analysis of samples which contain no lead. The LOQ, in
turn, is the smallest concentration which will consistently
produce estimated concentrations that are within 30% of the true
concentration.
Table 2-5 contains the ELODs for the 24 instrument
batches performed in analyzing the regular field samples. Three
percent (35 out of 1169) of the regular samples had instrument
response concentrations below the ELOD for their instrument
batch. These samples are detailed in Table A-3 of the Appendix.
The LOQ was determined from information outlined in the
memorandum, "Potential Instrumental Measurement Error for Lead
Analysis," dated September 21, 1992. This memo, portions of
which are excerpted in Table 2-6, documented the instrumental
Volume II - Page 34
-------
measurement error for a series of known lead concentrations
ranging from (0.02 to 0.50 pq/roiL). The results suggested an LOQ
of 0.208 ng/riL.
Approximately 19% (226 out of 1169) of the regular
field samples had instrument response concentrations below the
LOQ. To examine the potential impact of how these samples are
handled in the statistical analysis, two sets of statistical
analyses were performed. In the first set of analyses, we set
the instrument response concentrations of those samples below
their ELOD equal to their ELOD. No modifications were made to
samples with concentrations above their ELOD but below the LOQ.
In the second set of analyses, we set the instrument response
concentrations of all samples below the LOQ equal to the LOQ. We
fit the mixed model described in Section 4 separately to each set
of data. Since the second set of analyses agree appreciably with
the first, only the results of the first set of analyses are
presented in this report. The only notable disagreement between
the two sets of analyses was that the difference in lead
concentrations in air ducts between abated and control homes was
not as great by the second analysis.
Volume II - Page 35
-------
Table 2-5. Estimated Level of Detection by Instrument Batch
Instrument
Batch
E04272A
E04292A
E05042A
E05072B
E05122B
E05132A
E05192A
E05262A
E05272A
E06022A
E06042A
E06112A
ELOD
Mg/mL
0.0298
0.0138
0.0383
0.0324
0.0308
0.0255
0.0293
0.0461
0.0634
0.0400
0.0465
0.0553
Instrument
Batch
E06122A
E06152A
E06242A
E06262A
E06292A
E07142A
E07212A
E07242A
E07302A
E08032A
E08062A
E08242A
ELOD
/jg/mL
0.0370
0.0254
0.0263
0.0655
0.0527
0.0300
0.0593
0.0354
0.0514
0.0272
0.0349
0.0240
Table 2-6. Potential instrumental Measurement Error:
Calculated Results
Lead
Concentration
(Mg/mL)
0.02
0.03
0.05
0.07
0.10
0.30
0.50
Average
Response
(Mg/mL)
0.03303
0.04253
0.06625
0.08816
0.11709
0.31963
0.52871
n-1
Standard
Deviation
0.01682
0.01893
0.02012
0.01891
0.02000
0.02643
0.02155
% Relative
Standard
Deviation
50.91%
44.50%
30.36%
21.45%
17.08%
8.27%
4.08%
Volume II - Page 36
-------
3.0 STATISTICAL MODELS
In this section we present the statistical models that
were fitted to the lead loading, lead concentration, and dust
loading data. We also discuss centering and scaling of design
variables which were used to produce easily interpretable model
parameters. The stepwise regression and mixed model procedures
used to arrive at final models are defined and model parameters
are related to specific hypotheses of interest.
3.1 MIXED RANDOM AND FIXED EFFECTS MODEL
In this section we describe the statistical models that
were fitted to the observed lead loadings, lead concentrations,
and dust loadings. These models are the basis for the
statistical analyses described in Sections 4 and 5.
The following model contains all of the design factors
considered in the study, random effects for housing unit-to-unit
and room-to-room variation, and additional explanatory variables
or covariates. This model was fitted separately to the data for
air duct, interior entryway, window channel, and window stool
dust samples.
= In (a) + UL + R +
Oi + IndSso^SOi PODi +
ln(jSPRD)PRDi:, + IndSgRjSRij + ln(0SRD)SRijPRDij
ln(7)X
for
i = 1, 2, ... , # units
j = 1, 2 rooms
where
measured lead concentration, lead loading, or
dust loading in the jth room in the ith unit,
Volume II - Page 37
-------
a = overall geometric average lead concentration
in control units for nominal values of
covariates,
UL = random effect for the ith unit; assumed to
follow a normal distribution with mean zero
and standard deviation av,
RAj = random effect for the jth room in the ith
unit; assumed to follow a normal distribution
with mean zero and standard deviation aR,
/3PZ = fixed multiplicative effect associated with a
unit that has undergone abatement; /3PR is
similarly defined for room-level abatement,
Pli = 1 if abatement was performed in the ith unit
and zero otherwise; PR^ is similarly defined
for room-level abatement,
0piD = fixed multiplicative effect of interior
abatement by E/E methods rather than removal
methods; 0POD and j8FRO are similarly defined
for outside abatement and room-level
abatement,
PIDi = the percentage of interior abatement that was
performed by E/E methods; PODL and PRDLi are
similarly defined for exterior abatement: and
room-level abatement,
j8SI = multiplicative effect of increasing the log-
square footage of abatement; 0SO and j3SR are
similarly defined for outside abatement and
room-level abatement,
SIL = log-square footage of interior abatement in
the ith unit or ln(l+SFIi) where SFIL is the
square footage of interior abatement in the
ith unit; SOL and SR^ are similarly defined
for outside abatement and room-level
abatement,
= ratio of the multiplicative effect of
increasing the log-square footage of interior
abatement by E/E methods to the
multiplicative effect of the same increase in
the log-square footage of interior abatement
by removal methods; j8SOD and 0SRD are
similarly defined for outside abatement and
room-level abatement.,
Volume II - Page 38
-------
X . = vector of additional covariates, and
7 = vector of multiplicative effects associated
with increases in the corresponding
covariates in the vector X.
The unit-level random effect terms for the different
sample types within a unit are assumed to be correlated.
Similarly, the room-level random effect terms for the different
sample types within a room are assumed to be correlated.
In the case of floor dust vacuum samples, an additional
within-room random error term was added to the model,
fiijfc = random effect for the kth sample in the jth
room of the ith unit.
Floor dust wipe samples were taken from only one
location in each of the abated houses. Therefore, no room level
effects were included in the model, nor can differences between
abated and control houses be estimated. The following model was
used for these samples:
= In (a) + UL + R^
Ii + ln(0SID)SIL
SOi + IntfsouJSOi
+ ln(7)X.
The model fitted to the data for exterior entryway dust
samples is
ln(a) + UL + Si +
where
CL = measured lead concentration at ith unit,
Volume II - Page 39
-------
S^ = random effect for the jth side of ith unit;
assumed to follow a normal distribution with
mean zero and standard deviation as,
and all other terms are defined as above. The side-level random
effect terms for the three different sample types within a side
of a unit are assumed to be correlated.
For foundation soil, boundary soil, and entryway soil,
an additional within-side of house component of variation is
added to this model:
Eijk = random effect for the kth sample on the jth
side of ith unit; assumed to follow a normal
distribution with mean zero and standard
deviation aE,
The additional explanatory variables (covariates) that
were considered for inclusion in the model are listed in Appendix
B. The variables considered included questionnaire responses,
field inspection variables, and measurements taken during the HUD
Demonstration. Explanatory variables that were found to be
significant for at least one of the sample types are listed by
category in the second column of Table 3-1. Nominal values of
these covariates and the sample types for which the covariates
are significant are listed in the third and fourth columns.
In the model, the a term represents the geometric
average lead level that can be expected in houses where no
abatement was necessary and none was performed (control houses)
for nominal values of the covariates included in the model. The
random effect term for units (UL) allows each housing unit to
have its own average lead level. The random effect terms for
rooms (Rij), entryways (E^) , and sides (S^) allow each room,
entryway, or side within the unit to have its own average lead
level.
The terms PIL and PIDL and the corresponding
coefficients, 0PI and j8PID, allow estimation of the effect of
Volume II - Page 40
-------
Table 3-1. Explanatory Variables that are Significant for at Least One Sample Type
Explanatory
Variable Category
Abatement
Substrate
Cleanliness
Occupation
Activities
Age/
location of house
Other resident
factors
Sampling deviations
Explanatory Variable
Abatement contractor
Months between clearance and sampling
Phase of HUD Demonstration (of three) in which residence was abated
HUD XRF or AAS measure of lead loading
Specific removal method used on the interior
- chemical stripping
- remove and replace
- heat gun
- removal
Substrate type
Substrate condition
Frequency of wet mopping uncarpeted floors
Frequency of window sill dusting
Frequency of vacuuming uncarpeted floors
Wearing home work clothes from an occupation with potential lead contamination
Resident employed in welding occupation
Resident employed in paint removal occupation
Frequency of removing paint at home
Frequency of pipe or electrical component soldering
Proximity to supposed lead smelter
Year house was built
Number of children (between ages of 7 and 17)
Months at residence
Ownership of home
Air duct samples taken from cover of air duct
Window channel samples taken with small nozzle
Nominal Value
24
C:0.38 (mg/cm2)
A:1.10 (mg/cm2)
30X
15X
15X
40%
Carpet
Carpet
Wood
Good
12
1
0
0
0
0
0
0
> 3 mi
1932
0
12
Owner
No
No
Sample Types for
Which Explanatory
Variable is
Significant
ARD, FLU
WST
VST
FDN
WCH
FLR
EUI
WCH
ARD, WCH
ARD
ARD
EUI
FDN. FLR
BDY
EUI, FDN
BDY
EUY
BDY
EUI
FDN
FLW, FDN
ARD
WCH
-------
abatement and a distinction of effects between methods. 0PI
characterizes the abatement effect without distinguishing between
E/E methods and removal methods. 0PID characterizes the
difference in the interior abatement effects for E/E methods
versus removal methods. Outside abatement and room-level
abatement effects are handled similarly in the model.
The term SIL and the corresponding coefficients, j3SI
and |3SID, allow the effect of the amount of interior abatement,
on a per log-square foot abated basis, to be estimated by the
model. 0SI characterizes the interior abatement effect per log-
square foot abated without distinguishing between E/E methods and
removal methods. /?SID characterizes the difference in the
interior abatement effects per log-square foot abated for E/E
methods versus removal methods. Exterior abatement and room-
level abatement effects are handled similarly in the model.
3.2 CENTERING AND SCALING OF COVARIATE8
Several covariates included in the models were centered
and scaled so that the model parameters would have more
meaningful interpretations. In order to determine the
appropriate centering and scaling parameters, three classes of
abated units were identified: (1) predominantly E/E, (2)
predominantly removal, and (3) abated. The third class is the
combination of the first and second classes. If the percentage
of abatement performed by E/E methods is more than 50%, then an
abated unit is classified as predominantly E/E. Otherwise, it is
classified as predominantly removal. Each unit is classified
separately for interior and exterior abatement.
For each of the three classes of abated units two
quantities were determined:
Typical percentage abated by E/E methods, and
Typical square footage abated.
Volume II - Page 42
-------
These values are reported in Table 3-2 for interior, exterior,
and room-level abatement. The typical percentage abated by E/E
methods was determined by taking an average across all units in
the class.
A correlation was observed between total square feet
abated in a house and the method used to perform the abatement.
Typically, significantly more square feet were abated when E/E
methods were used than when removal methods were used. This
occurred both indoors and outdoors. Therefore, the typical
square footage abated was treated as a function and allowed to
vary with the percentage abated by E/E methods. To accomplish
this, a simple linear regression of log-square feet abated versus
percent abated by E/E methods was fit to the data for all abated
units. Figure 3-1 displays the regression relationship for
interior abatement. Similar regression relationships were
developed for exterior and room level abatement.
The typical square footage abated values reported in
Table 3-2 are taken from the regression relationship for the
typical percentage abated by E/E methods; Taking interior
abatement for example, a predominantly E/E unit with 93% E/E
abatement is predicted to have 282 total square feet of interior
abatement. Similarly, a predominantly removal unit with 4% E/E
abatement is predicted to have only 61 total square feet of
interior abatement. Finally, an abated unit with 67% E/E
abatement is predicted to have 180 total square feet of interior
abatement. The typical square footage abated values in Table 3-2
for exterior and room level abatement were determined in a
similar fashion.
The values in Table 3-2 were used to center and scale
corresponding model covariates as indicated in Table 3-3.
Control unit values were never centered. For abated units, PID,
POD, and PRO values were centered by subtracting off the typical
percent abated by E/E methods for an "abated" unit. ' These values
were then scaled by dividing the centered variable by the
difference between the typical percent abated by E/E methods for
Volume II - Page 43
-------
Table 3-2. Average Percent Abated by E/E Methods, by Abatement
Method Classification for Interior, Exterior and
Room Level Abatement
Level
Interior
Exterior
Room
Typical % Abated by
E/E Methods
E/E
93
92
96
Removal
4
27
3
Abated
67
78
69
Typical Square Footage
Abated
E/E
282
628
70
Removal
61
260
36
Abated
180
519
58
Table 3-3. Centering and Scaling Parameters
for Model Covariates
Covariates
PID
POD
PRD
SI
so
SR
PR
Value Subtracted
Control
0
0
0
0
0
0
0
Abated
67%
78%
69%
ln(57)+0.0172*(E/E%)
ln(180)+0.0136*(E/E%)
ln(35)+0.0072*(E/E%)
1
Value
Divided By
89%
65%
93%
ln(2)
ln(2)
ln(2)
-1
Volume II - Page 44
-------
10000
o
H-
H
M
2?
Ifl
(D
1000
100
10
'' I ' ' i i I • ' i ' 1—'—i—i—i—I—i—i—i—i—I—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r
O.O 0.1 O.2 0.3 O.4 O.5 O.6 O.7 O.8 O.Q
Percent Abated by E/E Method*
1.O
Figure 3-1. Total square feet abated indoors vs. percent encapsulated/enclosed
indoors (for abated houses).
-------
a typical "E/E" unit minus a typical "removal" unit. SI, SO,
and SR values were centered by subtracting off the logarithm of
the predicted square footage abated based on the regressions
versus E/E percentage discussed above. These values were then
scaled by dividing by ln(2). Finally for abated units, PR was
centered by subtracting one (making abated rooms the default for
abated units). This value was then scaled by dividing by -1.
The result of the data transformations detailed in
Table 3-3 is that the primary factors in the statistical models
can be interpreted as indicated in Table 3-4. These
interpretations are consistent with the hypotheses we wish to
test, as will be discussed in Section 3.4.
The other explanatory factors previously described were
also centered and scaled where appropriate. Centering was
accomplished by subtracting off the nominal value reported in
Table 3-1. These nominal levels are reported again in Section 4
in each table where estimates are given, along with the scaling
factor used. The selection of nominal values is also discussed
in more detail in Section 4.
3.3 MODEL SELECTION
The procedure used to select models to fit to the data
was developed in concert with the study objectives. Specific
terms corresponding to the primary design factors were included
in the model to test the hypotheses presented in Section 4.5 of
the Detailed Design Document (Battelle and MRI, 1992a).
Every model used in this study included the following
primary design factors:
A term to distinguish between control units and
abated units (PI), and
A term to distinguish between abatement methods
(PID for interior samples, POD for exterior
samples).
Volume II - Page 46
-------
Table 3-4. Parameter Interpretation After Centering and Sealing
Parameter
PPI
PPO
PPR
Ppn>
PpOD
PpRD
Ps,
Pso
PSR
Psm
PSOD
PSRD
Interpretation
Ratio of the expected lead level in a typical abated room in a
typical abated unit(l) to the expected lead level in a control
unit
Ratio of the expected soil lead level for a typical abated
unitw to the expected soil lead level for a control unit
Ratio of the expected lead level in a control room in a typical
abated unitw to the expected lead level in a typical abated
room in the same abated unit
Ratio of the expected lead level in a typical abated room in a
typical E/E unit(1> to the expected lead level in a typical
abated room in a typical removal unit
Ratio of expected soil lead level for typical E/E unitw to
expected soil lead level for typical removal unit
Ratio of the expected lead level in a typical E/E room in an
abated unit to the expected lead level in a typical removal
room in the same abated unit
Multiplicative effect of doubling the square footage of
interior abatement in a typical abated unit04
Multiplicative effect of doubling the square footage of
exterior abatement in a typical abated unit04
Multiplicative effect of doubling the square footage of room-
level abatement in a typical abated room04 while holding the
square footage and mix of unit level abatement constant
Ratio of the multiplicative effect of doubling the square
footage of interior abatement in a typical E/E unit0*' to the
multiplicative effect of doubling the square footage of
interior abatement in a typical removal unit04
Ratio of the multiplicative effect of doubling the square
footage of exterior abatement in a typical E/E unit04 to the
multiplicative effect of doubling the square footage of
exterior abatement in a typical E/E unit04
Ratio of the multiplicative effect of doubling the square
footage of room-level abatement in a typical E/E room04 to the
multiplicative effect of doubling the square footage of room-
level abatement in a typical removal room*) while holding the
square footage and mix of unit level abatement constant
(a) Typical with respect to both E/E% and square footage abated as indicated
in Table 3-2.
(b) Typical with respect to E/E% as indicated in Table 3-2 but with varying
square footage abated.
Volume II - page 47
-------
Models for interior dust measurements also contained:
A term to distinguish between control rooms and
abated rooms in abated units (PR).
In addition to the three primary design factors, many
additional factors (questionnaire data, field observations) were
included to estimate other effects which may affect lead levels.
The additional factors included in each model were selected using
a phased stepwise regression approach.
3.3.1 Phase 1; Abatement Effects (Steowise Regression)
First, stepwise regression was used to select
additional abatement design factors which were significant above
and beyond the effects of the three primary design factors
described above. The additional abatement factors included
square-footage abated by room, as well as a breakdown of square-
footage by abatement method.
In the stepwise regression, factors were retained only
if they were significant at the 5 percent level. Any factor
found to be significantly associated with either lead
concentration or lead loading was automatically forced to be
retained in the model for the next selection phase.
3.3.2 Phase 2; Non-Abatement Factors (Stepvise Regression)
In a second phase of factor selection, all remaining
factors, including questionnaire and visual observation data, HUD
Demonstration Data, and other practical measures (see Table 3-1),
were considered as candidate factors in addition to the design
factors discussed above. Stepwise regression was used again to
select significant factors. Any factors found to be significant
at the 5 percent level were retained for the next selection
phase.
Volume II - Page 48
-------
Appendix B presents a list of all the nonabatement
factors considered for inclusion in the models. To avoid
confounding, a preliminary correlation analysis was performed to
screen any factors which were strongly correlated with others.
For example, for 15 of the 16 homes in which a resident wore work
clothes home from their occupation, their clothes were also
washed at home. Therefore, only the former was included as a
candidate factor in the stepwise regression. Specifically, if
any factor was more than 80 percent correlated with another, one
of the factors was excluded from the models. The factor with the
most complete data was used in fitting the models.
3.3.3 Phase 3; Mixed Model Screening (Backward Elimination)
Phase 1 and Phase 2 identified a subset of factors with
some association with lead levels. However, due to software
limitations, the stepwise regressions were based on fixed effect
models whereas it is proper to use a mixed model with random
effects in the factor selection process described above.
Therefore a mixed model was fit with random unit and random
room/side of house effects where appropriate. Any factors not
found to be significant by the mixed model analysis at the 10%
level were removed from the model (aside from the three design
factors described at the beginning of Section 3.3). This process
was repeated, refitting the model each time and removing one
factor at a time, until all factors remaining were observed as
significant covariates for either lead loading or lead
concentration.
The final models varied by sample type. Appendix C
displays the selected factors and their estimated effects by
sample type and response (lead concentration, dust loading, lead
loading). This table is explained in more detail in Section 4.
In particular, all wipe samples were taken on floors in only one
room of each of abated unit. Therefore, although for 4 of the 34
houses where these were taken a control room was selected, any
room-to-room differences in lead loadings are confounded by
Volume II - Page 49
-------
house-to-house variation. Therefore, room-level abatement
effects were not estimated from the data collected by wipe
sampling. In particular, differences between control rooms and
abated rooms and substrate effects were ignored for wipe samples.
By the same reasoning, we did not distinguish between substrates
since samples in each house were taken side-by-side.
3.4 HYPOTHESIS TESTS
As stated in the Detailed Design Document (Battelle and
MRI, 1992a), data were collected to test the following
hypotheses:
H01: Average lead levels in a typical abated room in a
typical abated unit are equivalent to average lead
levels in a control unit.
H02: Average lead levels in a typical abated room in a
typical E/E unit are equivalent to average lead
levels in a typical abated room in a typical
removal unit.
H03: Average lead levels in a typical abated room in a
typical abated unit are equivalent to average lead
levels in a control room in a typical abated unit.
H04: Unit to unit differences above and beyond those
explained by the models are uncorrelated.
Hypothesis H01 is equivalent to the hypothesis that 0PI=0,
hypothesis H02 is equivalent to the hypothesis that 0PID=0, and
hypothesis H03 is equivalent to the hypothesis that j8PR=0. Thus,
the model parameters align perfectly with the hypotheses to be
tested. Hypothesis H04 will be tested via extensive correlation
analyses in Section 5.
Volume II - Page 50
-------
4.0 MODELING RESULTS
4.1 MODELING RESULTS SUMMARY
For most of the components sampled, lead levels were
found to be higher in abated units than in control units. The
most notable differences were observed for air ducts, floors, and
exterior entryway (dust) samples. For each of these components,
the differences in lead loadings were statistically significant.
For floors and exterior entryways, this difference was marginal.
In air ducts lead concentrations were significantly higher as
well.
Differences were also observed in houses abated by the
different methods. In every case where lead levels were higher
in abated houses than in control houses, levels were also higher
in houses abated by E/E methods than in those abated by removal
methods. The most notable of these differences were observed in
air ducts, on window stools, and on floors. For air ducts and
window stools there were statistically significant differences in
both lead loadings and lead concentrations. On floors the
difference was marginally significant for loadings.
In all cases except interior entryway samples, lead
levels were lower in the control rooms of abated units than in
the abated rooms of abated units. In no case was the difference
statistically significant.
There were some exceptions to the above
generalizations. Lead loadings and lead concentration in window
channels, and lead concentrations in dust at interior entryways
of abated houses were slightly lower than in the control house
samples. Lead loadings for window channels and lead
concentrations for interior entryways were also slightly lower in
E/E houses than in removal houses. In none of these cases was
the result statistically significant.
One cautionary note should be mentioned concerning the
interpretation of the differences observed in houses abated by
the different methods. Due to abatement cost, most of the units
which had very large amounts of abatement performed were abated
Volume II - Page 51
-------
by E/E methods. This means that the lead problem was often
greater in the houses selected for abatement by E/E methods. In
other words, the results presented here indicating that lead
levels were higher after abatement by E/E methods may simply be
an artifact of the more severe initial condition in these houses.
There were some other factors which were found to be
significantly correlated with differences in levels. Some of
these were associated with house-to-house differences. These
include:
Specific abatement method used
Abatement contractor
HUD Demonstration abatement phase (early,
intermediate, or late)
Year the house was built
Time since the house was cleared
Questionnaire answers
- occupations of residents
ages of occupants
- measures of cleanliness
activities of occupants
ownership
Proximity to a lead smelter.
Some factors were associated with differences at the sample
level. These include:
Substrate type and condition
XRF measures taken prior to abatement
Sampling deviations.
Volume II - Page 52
-------
4.2 DETAILED MODELING RESULTS
4.2.1 Analysis of Abatement and Random Effects
This section presents estimated effects of the various
abatement factors considered in the study on lead loading, lead
concentration, and dust loading for each sample type collected.
These estimates are to be interpreted as having been corrected
for other practical effects found to be significant (e.g.,
ownership, XRF measurements, cleanliness, substrate, etc.). Also
described in this section is uncontrolled and unexplained random
variation from unit to unit, room to room (or side to side), and
within room/side for each sample type.
Effects of Primary Abatement Factors
Table 4-1 displays estimates of the effects of the
primary abatement factors on lead loadings. The first column
provides the number of samples included in the model for each
sample type. In many cases these numbers are lower than the
total number of samples because of missing values of significant
covariates. For instance, in some cases, the housing unit
resident interviewed did not know the answers to some of the
questionnaire items (e.g., ownership, cleanliness measures,
etc.). The number of missing values were fewer than 20 for most
sample types. However, for foundation soil samples, 30
observations were excluded. The HUD Demonstration XRF measures
were found to be a significant factor and there were several
observations in the CAP Study for which there was no
corresponding XRF measure available from the HUD Demonstration.
Table 4-2 displays the estimated effects of the primary
abatement factors for lead concentrations. Table 4-3 provides
the corresponding results for dust loadings.
The second column in these tables contains the
estimated geometric mean in units which were not abated. The
estimate is to be interpreted as the average lead loading in
control houses when the covariates included in the model are
fixed at the nominal levels described in Table 4-4. The log
Volume II - Page 53
-------
Table 4-1.
Estimates1 of Effects of Primary Abatement Factors on Lead Loading;
Controlling for Significant Covariates
Sample
Type
Air Duct
(Vacuum)
Window
Channel
(Vacuum)
Window
Stool
(Vacuum)
Floor
(Wipe)
Floor
(Vacuum)
Entryway
(Interior
Vacuum)
Entryway
(Exterior
Vacuum)
No. of
Samples/
Denominator
Degrees of
Freedom
86
(35)
86
(33)
111
(59)
65
(32)
233
(105)
90
(34)
97
(46)
Geometric Mean in
Control Units
After Controlling
for Effects of
Significant
Factors
76
(0.52)
3909
(0.73)
37.9
(0.38)
X
trcifs.
35.3
(0.30)
531
(0.35)
220
(0.37)
Ratio of Levels
in Abated Rooms
of Abated Units
to those in
Control Units
4.70
(0.61)
.02
0.73
(0.78)
.68
1.95
(0.49)
.18
1.83
(0.34)
.08
1.05
(0.38)
.90
2.24
(0.44)
.07
Ratio of Levels
in E/E Units to
those in R/R
Units
3.99
(0.68)
.05
0.60
(0.89)
.57
4.64
(0.62)
.02
1.48
(0.38)
.31
1.87
(0.36)
.08
1.15
(0.44)
.75
1.06
(0.50)
.91
Ratio of Levels
in Control
Rooms of Abated
Units to those
in Abated Rooms
of Abated Units
0.73
(0.39)
.43
0.53
(0.53)
.23
0.56
(0.42)
.18
-•
0.57
(0.33)
.10
1.63
(0.41)
.24
--
Standard Deviation Estimates
Unit-to-
wn! t Log
Standard
Deviation
1.52
(0.86)
.00
1.09
(0.87)
.12
0.90
(0.74)
.15
0.73
(0.45)
.01
0.00
0.00
0.91
(0.69)
.08
Residual
Room-to-Room Log
Log Standard Standard
1.18
1.66
1.76
0.56
1.27 0.94
(0.53)
.00
1.48
1.47
e
(D
Ui
1 Top value is multiplicative estimate, middle value is logarithmic standard error of estimate, and bottom value is observed significance level.
-------
Table 4-2.
Estimates1 of Effects of Primary Abatement Factors on Lead Concentration;
Controlling for Significant Covariates
Sample
Type
Air Duct
(Vacuum)
Window
Channel
(Vacutm)
Uindow Stool
(Vacuum)
Floor
(Vacuum)
Entryway
(Interior
Vacuum)
Entryway
(Exterior
Vacuum)
Entryway
(Soil)
Foundation
(Soil)
Boundary
(Soil)
No. of
Sanples/
Denominator
Degrees of
Freedom
86
(35)
83
(29)
111
(59)
233
(105)
90
(34)
97
(46)
109
(12)
88
(14)
120
(20)
Geometric Mean in
Control units
After Controlling
for Effects of
Significant
Factors
332
(0.19)
1572
(0.52)
413
(0.29)
122
(0.19)
182
(0.24)
184
(0.22)
111
(0.18)
J90
7719)
109
(0.14)
Ratio of Levels
in Abated Rooms
of Abated Units
to those in
Control Units
1.59
(0.23)
.05
0.81
(0.55)
.71
1.70
(0.37)
.16
1.07
(0.22)
.76
0.85
(0.27)
.56
1.19
(0.26)
.51
1.18
(0.22)
.46
1.17
.21
.46
1.18
(0.17)
.35
Ratio of
Levels in E/E
Units to those
in R/R Units
2.01
(0.24)
.01
1.86
(0.63)
.34
2.71
(0.48)
.04
1.20
(0.23)
.42
0.95
(0.31)
.88
1.00
(0.29)
.99
1.18
(0.24)
.52
0.72
(0.33)
.34
1.27
(0.18)
.21
Ratio of Levels
in Control
Rooms of Abated
Units to those
in Abated Rooms
of Abated Units
0.79
(0.23)
.30
0.66
(0.36)
.25
0.65
(0.31)
.16
0.84
(0.22)
.44
1.28
(0.26)
.34
-
-
-
-
Standard Deviation Estimates
Unit-to-
Unit Log
Standard
Deviation
0.00
(.)
0.84
(0.62)
.06
0.74
(0.56)
.08
0.00
0.49
(0.41)
.15
0.52
(0.41)
.10
0.37
(0.35)
.27
.21
.24
.44
0.38
(0.25)
.01
Room-to-Room
Log Standard
Deviation
Residual
Log
Standard
Deviation
0.79
1.12
1.29
0.73
(0.35)
.00
0.77
0.84
0.89
0.71
(0.38)
.00
.44
.26
.00
0.44
(0.22)
.00
0.40
0.27
0.21
Top value is multiplicative estimate, middle value is logarithmic standard error of estimate, and bottom value is observed significance level.
-------
Table 4-3.
Estimates1 of Effects of Primary Abatement Factors on Dust Loading;
Controlling for Significant Covariates
Sample
Type
Air Duct
(Vacuum)
Window
Channel
(Vacuum)
Window Stool
(Vacuum)
Floor
(Vacuum)
Entryway
(Interior
Vacuum)
Entryway
(Exterior
Vacuum)
No. of
Samples/
Denominator
Degrees of
Freedom
86
(35)
86
(33)
111
(59)
233
(105)
90
(34)
97
(46)
Geometric Nean
in Control
Units After
Controlling for
Effects of
Significant
Factors
202
(.48)
2451
(0.51)
92
(0.22)
348
(0.26)
3053
(0.25)
1151
(0.25)
Ratio of Levels
in Abated Rooms
of Abated Units
to those in
Control Units
3.11
(.57)
.05
0.90
(0.54)
.84
1.16
(0.28)
.60
1.69
(0.29)
.08
1.19
(0.28)
.54
1.95
(0.30)
.03
Ratio of Levels
in E/E Units to
those in R/R
Units
1.80
(0.63)
.36
0.33
(0.61)
.08
1.72
(0.36)
.14
1.52
(0.32)
.19
1.24
(0.31)
.49
1.05
(0.33)
.88
Ratio of Levels
in Control
Rooms of Abated
Units to those
in Abated Rooms
of Abated Units
0.91
(0.34)
.78
0.80
(0.37)
.56
0.86
(0.26)
.55
0.67
(0.25)
.12
1.31
(0.29)
.36
Standard Deviation Estimates
Unit-to-Unit
Log Standard
1.45
(0.79)
.00
0.72
(0.69)
.28
0.43
(0.43)
.31
0.49
(0.44)
.22
0.00
-
0.40
(0.50)
.52
Room- to- Room
Log Standard
Residual
Log
Standard
Deviation
1 00
1.18
1.08
0.84
(0.45)
.00
0 86
1.06
1 19
•o
0)
iQ
(D
U>
0\
Top value is multiplicative estimate, middle value is logarithmic standard error of estimate, and bottom value is observed significance level.
-------
Table 4-4. Multiplicative Effects of Secondary Abatement Factors
Factor
Total Interior Square
Feet Abated
Total Exterior Square
Feet Abated
Percent Exterior
Abatement by E/E Methods
Interior Removal Method
• Chemical Stripping
• Removal/Replace
• Heat Gun
• Removal
Abatement Contractor
• A (3 units)
• B (15 units)
• C (13 units)
• D (4 units)
Phase
• 1 (13 units)
• 2 (13 units)
• 3 (9 units)
Months between clearance
and sampling
Last XRF measure at
sample location during
HUD demonstration (log)
Sample Type
Floor
(Vacuum)
Window
Channel
Window
Channel
Foundation
Floor (Wipe)
Window
Channel
Air Ducts
Floor (Wipe)
Window Stool
Window Stool
Foundation
Nominal
282 for Typical E/E
61 for Typical Removal
282 for Typical E/E
61 for Typical Removal
628 for Typical E/E
260 for Typical Removal
628 for Typical E/E
260 for Typical Removal
Typical Abated
30X
15X
15X
40X
NA
NA
NA
24
-0.964 for control
0.094 for abated
Deviation
Double square feet abated
Double square feet abated
Double square feet abated
Double square feet abated
(Effect equals ratio of typical E/E to
typical removal)
+10X
+1 OX
+10X
+10X
NA
NA
NA
1 month
Multiplicative Effect
Lead
1.03
1.34
0.49*
NA
NA
0.44"
0.90
1.00
1.06
1.06
0.55
1.01
0.78
3.35
•
4.70
0.68
1.58
0.12
*
12.43
1.04
0.33
*
0.62
NA
Lead
1.14*
1.31*
0.59*
0.65*
NA
*
1.05
1.03
1.33
0.70
*
2.34
0.77
0.91
1.81
NA
*
6.73
1.09
0.44
*
0.67
1.28*
Dust
0.91
1.02
0.83
NA
NA
NA
0.85
0.97
0.79
1.52
0.24
1.36
0.83
1.87
NA
1.84
0.97
0.77
0.93
NA
Significant at the 10X level.
-------
standard error of these estimates appears in parentheses below
each estimate.
Figure 4-1 displays estimated geometric means in
control units by sample type for lead loading, lead concentration
and dust loading. Some interesting points to note regarding
these geometric means are as follows:
The highest lead loadings were observed in the
window channels, and the lowest were observed on
floors.
Floor dust loadings and interior entryway dust
loadings were higher than those in the air ducts.
Foundation soil lead concentration appears to be
much higher than boundary or entryway
concentrations in these houses.
Similar relationships exist among the three
measures for interior and exterior entryway
dust, with higher levels indoors.
One thing to keep in mind when observing dust levels on floors
(and interior entryways) is that substrate was an important
differentiating factor. The geometric means are based on samples
taken from carpet, because it was the most frequent substrate
sampled, but carpet had much higher levels of dust than any of
the other substrates. Therefore, geometric mean lead loadings
appear higher than if substrate was not controlled.
The third column in Tables 4-1, 4-2, and 4-3 displays
the estimated ratio of levels in abated rooms of abated houses to
levels in control houses. The fourth column contains the
estimated impact of abatement method, which should be interpreted
as the ratio of levels in typical E/E units to levels in typical
removal units (see Section 3.2). The fifth column in these
tables gives an estimate of the ratio of levels in control rooms
of abated houses to levels in abated rooms of abated houses. The
log standard error and significance level of these estimates
Volume II - Page 58
-------
10000
<
o
ID
•a
o>
«Q
(D
Ui
\D
V/////////////A
X
V
Y/////////////A
Y/////////////A
f
SampteTVpe
Lead Loading
Lead Goncen.
Dust Loading
Figure 4-1.
Geometric mean lead loading (fig/ft2), lead concentration (Mg/9)» and dust
loading (mg/ft2) in control units after controlling for effects of
significant factors.
-------
appear beneath each estimate. The latter represents the observed
significance of a test that the ratio equals 1.
The following are the significant results for the
estimated effects of primary abatement factors:
Air Ducts — Lead loadings, lead concentrations,
and dust loadings were higher in abated units than
in control units. Lead loadings and lead
concentrations were higher in E/E units than
removal units.
Window Channels — Dust loadings were lower in
abated units.
Window Stools — Lead loadings and lead
concentrations were higher in E/E units than in
removal units.
Floors (Vacuum) — Lead loadings and dust loadings
were higher in abated units than in control units
(marginally). Lead loadings were higher in E/E
units than in removal units (marginally). Lead
loadings were lower in control rooms of abated
units than in abated rooms (marginally).
Exterior Entryway — Lead loadings and dust
loadings were higher in abated units than in
control units. (For lead loadings, the difference
was marginal.)
The estimates from columns 3, 4, and 5 of Tables 4-1,
4-2, and 4-3 are displayed graphically in Figures 4-2, 4-3, and
4-4 for lead loading, lead concentration, and dust loading,
respectively. Reference lines are provided on these plots at a
level of one. A bar which rises above the reference line for the
*Abatement' factor indicates that for this sample type levels
were higher in abated houses than in control houses. A bar which
rises above the reference line for the 'Method (E/R)' factor
indicates that the levels in E/E units were higher than those in
removal units. If the 'Ctrl Room' effect is greater than one,
then levels in unabated rooms of abated houses were higher than
in abated rooms.
Volume II - Page 60
-------
10.0
<
o
(D
(D
Oi
o
ARD WCH
W8T FLW FLA
EWI EWO Sample Type
Figure 4-2. Estimated multiplicative effects of abatement from mixed model ANOVA:
Lead loading (/ig/ft2).
-------
1O.O
c
s
(D
ON
NJ
ft V J3 0 JSU
s ijs ifs ii§
AFD WCH
WST
FLR
EW1 EWO EWY FDN BOY Sampla Type
Figure 4-3. Estimated multiplicative effects of abatement
from mixed model ANOVA: lead concentration
-------
1O.O
s
M
§
(D
•O
01
«Q
ID
Ot
CJ
Y////////////////A
i
ARD
WCH
W6T
FLR
EW1
EWO Sample Type
Figure 4-4. Estimated multiplicative effects of abatement from
mixed model ANOVA: dust loading (mg/ft2).
-------
The most significant difference between abated and
control houses was observed in the air ducts for lead loadings
and lead concentrations. Perhaps more striking in these figures
is the frequency with which the «Method (E/R)' bar rises above
the reference line. As mentioned above, this indication that E/E
houses have higher lead levels than removal houses could simply
be a reflection of a more serious initial lead problem in the E/E
houses.
Similarly, the figures portray lower levels in the
control rooms of abated houses than in abated rooms of the same
houses. This indicates that abatement performed in the rooms
that needed it did not reduce lead levels to the baseline levels
found in control rooms that did not require abatement.
Analysis of Random Effects
The last three columns of Tables 4-1 through 4-3
provide estimates of the unit-level, room/side-level, and
residual error-level variance components. Only in the case of
vacuum floor samples and soil samples were the room/side of
house-level variance components estimable. The values presented
are given as standard deviations of the log-transformed
responses. Except in the case of residual standard deviation,
each estimate is followed by its standard error estimate and a
test of significance that the log standard deviation equals zero.
Figures 4-5, 4-6, and 4-7 display the estimates of these variance
components. The variances are summed and stacked in these plots
providing an estimate of overall uncontrolled variance in the
measures. Interesting points to note regarding the variance
estimates are the following:
There is much more variation from sample to sample
on the window channel and window stool
measurements than floor measurements.
In air ducts, although there is large unit-to-unit
variation in lead loading, it is negligible for
lead concentration.
Volume II - Page 64
-------
0>
•tf
01
a\
Ul
Variance Component
Sample Type
Unit KXVl Location
Figure 4-5. Variance component estimates from mixed model AMOVA:
lead loading (jig/ft2).
-------
a
c
I
H
0*
iQ
(D
Wanee Component
Sampteiype
UnK r?WI Location
Error
Figure 4-6. Variance component estimates from mixed model ANOVA:
lead concentration (M9/g)•
-------
I
•0
0>
iQ
(D
Vartanoe Component
Sample type
\Jrtt ivwi Locttdon
Ennor
Figure 4-7. Variance component estimates from mixed model ANOVA: dust loading (mg/ft2)
-------
Among soil samples, uncontrolled (error) variation
was smallest for foundation samples, largest for
entryways.
One of the considerations in interpreting these variance
components is that different models were fit to each sample type.
Therefore, for some sample types, more factors are controlled.
For example, more factors were controlled in the case of
foundation soil samples than any of the other soil samples; in
particular, this was the only sample type for which XRF measures
from the HUD Demonstration were included.
Effects of Secondary Abatement Factors
Table 4-4 displays estimates of the effects of
secondary abatement factors found to be significantly associated
with lead levels for at least one of the sample types. Some
important items to note regarding the effects of these secondary
abatement factors are:
Units with large amounts of interior abatement
were associated with higher lead concentrations
and loadings on floors and window channels.
Units with large amounts of exterior abatement
were found to have lower lead loadings and
concentrations in window channels, and lower lead
concentrations in foundation soil samples.
Lead concentrations in window channels were
impacted by the specific removal method used
(chemical stripping, removal and replacement, heat
gun stripping, or other removal).
Different contractors used to perform the
abatements in the HUD Demonstration were found to
significantly effect the lead levels in air ducts
and wipe floor samples.
Houses abated during the Demonstration were abated
in three different phases according to the
magnitude of the abatement required; units with
the most severe lead problems were abated first.
Abatement phase was found to be significant in
Volume II - Page 68
-------
distinguishing between lead levels in window stool
samples.
The length of time since clearance testing during
the HUD Demonstration was found to be significant,
with lower lead levels found at houses cleared
more than 24 months before CAPS sampling.
The average of the (log) XRF or AAS measures taken
during the HUD Demonstration was found to be
significant, with higher lead concentrations found
at houses with higher XRF/AAS readings.
The second column of Table 4-4 displays the sample
types for which each of these factors was significant. This is
followed by a description of the nominal level of the factor.
The geometric means displayed in Table 4-1 through 4-3 should be
interpreted as though levels of these factors were fixed at the
nominal levels. The fourth column of Table 4-4 describes the
deviation from nominal with which the multiplicative effects in
the last three columns are associated. An asterisk is placed in
the multiplicative effect box for each response where the
association was significant at the 10 percent level.
For example, the estimated geometric mean lead
concentration on window channels in control units (Table 4-2) was
1572. The time between clearance and CAPS sampling, and the
specific removal method used were found to be significant.
Nominally, time since clearance was 24 months. To obtain an
estimate of the level for a control house which was cleared by
XRF 27 months before sampling - holding all other factors
constant - multiply the geometric mean from Table 4-2 by the
estimated effect (0.67) of this factor 3 times. That is
1572 * (0.67)3 = 473.
To evaluate the effect of deviating from the nominal levels of
abatement by specific removal methods, each of the deviations
needs to be accounted for. For example, the multiplicative
Volume II - Page 69
-------
adjustment to lead concentration necessary to describe levels in
an abated unit in which 50 percent of the removal was done with a
heat gun and 50 percent was done by chemical stripping, would be
(1.05)2 (1.03)'1-5 (1.33)'1-5 (0.70)1 = 0.48. The factors in this
equation come from Table 4-4, and relate to the interior removal
abatement method (e.g., factor 1.05 for chemical stripping), as
well as the percentage of each method used as it deviates from
the nominal level (e.g., the exponent for factor 1.05 represents
two "deviations" from the nominal percentage of 30%).
By the method of variable screening, every factor
represented in this table is significant for either lead loading
or lead concentration. It is interesting to note that every one
of these factors had a significant association with lead
concentration.
Appendix C contains the detailed model fitting results
listed by sample type and response ('CONC' = lead concentration,
'DUST' = dust loading, 'LOAD' = lead loading). Each factor
included in the final model is listed under 'Parameter'. This is
followed by the degrees of freedom for estimating the error term
to which the effect was compared (except in the row displaying
the number of observations). This is followed by the column
containing the multiplicative effect of the factor (simply the
geometric mean for the intercept). The logarithm of this effect
appears in the next column (except in the case of the lines for
variance components). The log standard error is then presented,
followed by the significance of the result. Standard deviation
estimates for variance components' ('STUDYID' = house, 'LOCATION
(STUDYID)' = room or side, 'Standard Reviation Estimate') appear
in the 'Estimate' column.
Volume II - Page 70
-------
4.2.2 Analyses of Abatement and Random Effects bv Sample TYPO
The previous section summarized modeling results across
all sample types collected. This section breaks down these
modeling results into more detailed discussions for each sample
type separately.
4.2.2.1 Dust Samples
This subsection presents modeling results for all
locations at which dust samples were collected.
Air Ducts. There were significantly higher levels of
lead in air ducts of abated houses than in control houses, and
levels were significantly higher in houses abated by the E/E
methods than by the removal methods. Lead loadings were 5 times
higher and lead concentrations were 60 percent higher in abated
homes. Lead loadings in typical E/E units were approximately 4
times higher than in typical removal units. Concentrations were
only about twice as high. Control rooms in the abated houses did
not have lead levels significantly different than those in abated
rooms of the same houses.
Unit-to-unit variation was highest in air ducts for
lead loadings and dust loadings. However, unit-to-unit variation
in air duct lead concentration was negligible. This indicates
that for air ducts, most house-to-house variation in air duct
lead loading is due to the differences in dust levels in these
houses.
Window Channels. There was no significant difference
in lead levels observed in the window channels of abated and
control houses. Nor were there differences between levels in
units abated by different methods.
However, total square feet abated - both interior and
exterior - was a significant covariate. Doubling exterior square
feet abated was associated with a reduction of lead loadings by
half, and lead concentrations by 40 percent. Doubling interior
square feet abated was associated with a 31 percent increase in
lead concentration.
Volume II - Page 71
-------
Units abated by E/E methods typically had much more
abatement performed than the units abated primarily by removal
methods. The estimates provided are adjusted for this potential
confounding factor. The baseline square footage abated in a
typical removal house is 61 square feet; the baseline in a
typical E/E house is 282 square feet. These numbers are based on
a regression of (log) square feet abated on the percent abated by
E/E methods.
There were marginally significant differences in lead
concentration associated with use of the specific removal methods
on the interior. Of the four different methods, heat gun use was
associated with the highest concentrations. Pure removal was
associated with the lowest concentrations.
Window channels and window stools were associated with
the greatest total variation in lead levels. The differences
were particularly notable for lead concentrations. On window
channels there was large house-to-house variation in lead levels.
It was statistically significant for lead concentrations.
Window Stools. Although lead loadings were about twice
as high on window stools of abated units, there was no
significant difference in lead loadings or concentrations
observed between abated and control houses. Lead loadings were
4.6 times as high on window stools in the average E/E house than
in the average removal house. Lead concentrations were about 2.7
times as high in these houses. Each of these results was
significant. There were no significant differences in dust
loadings between these houses, so it appears that the differences
in lead levels are primarily related to lead concentration.
Although lead levels were lower in control rooms of abated houses
(almost 50% lower for lead loadings), the differences were not
statistically significant.
As mentioned above, variation in window stool measures
is very high compared to other sample types. Thus, it is even
more notable that the differences between houses abated by the
different methods are statistically significant.
Volume II - Page 72
-------
Floor (Wipe). Method was the only abatement effect
which was estimated for floor lead loadings from wipe samples.
This was not a significant factor.
House-to-house variation was significant for this
sample type, but it was moderate in magnitude. The estimated
residual log standard deviation was smallest for this sample
type, but this requires some explanation. By design, the floor
wipe samples were taken to compare with the floor vacuum samples
(see Section 6). Two side-by-side samples were taken per abated
house. Thus, the residual log standard deviation is really a
measure of side-by-side sample variability. This is in contrast
with the other dust sample types for which the two samples per
house were often taken from different rooms.
Floor (Vacuum Samples). About twice as many floor
(vacuum) samples were taken than any other sample type in the
study. For lead loadings, there were marginally significant
differences in each of the following relationships:
Levels were twice as high in abated units as in
control units (p = .08),
Levels were twice as high in units abated by E/E
methods as in units abated by removal methods (p
= .08), and
Levels were almost twice as high in abated rooms
of abated units as in control rooms of abated
units (p = .10).
Similar but less extreme differences were seen in the dust
loadings. Any differences in lead concentration were not
significant. Thus, differences in lead loadings seem mainly due
to differences in dust present on floors.
However, there was a significant relationship between
the total square feet abated indoors and lead concentration.
Houses where large amounts of abatement were performed were
associated with higher levels. Doubling square feet abated
indoors was associated with about 30% higher concentrations.
Volume II - Page 73
-------
There were negligible house-to-house differences in
both lead loadings and lead concentrations. Although not
significant, there were differences present in dust loadings.
There were significant room-to-room differences within houses for
all three measures. It is interesting to note that in Figure
4-3, the room-to-room variance component alone for vacuum floor
samples is greater than the estimated total variance for the
corresponding wipe samples. Another practical note is that the
residual log standard deviation estimate (the within-room
component) is larger than that for wipe floor samples. In some
cases, repeated samples taken within the same room were taken
from different locations within the room. Thus, this standard
deviation includes side-to-side within-room variation, whereas,
the floor wipe residual standard deviation does not. Section 6
compares the results of these two samples types in detail.
Interior Entryvay. There was no significant difference
observed in lead levels among the three categories of homes. Nor
was there a significant control room effect in abated homes.
Perhaps the most interesting thing to note about these
samples is the corrected geometric mean lead loading. The
estimated lead loading for interior entryways is 15 times higher
than that for regular floor (vacuum) samples. This difference is
due to only a 50% difference in lead concentration, but a 10 fold
difference in dust loading. As mentioned earlier, the baseline
substrate to which these estimates pertain is carpet. Carpet
retains much more dust than other substrates, and since people
tend to brush off dirt from their shoes at the entryways of
homes, this may be the cause of the difference.
There was significant house-to-house variation in lead
concentration, but not in lead loading or dust loading. Residual
log standard deviation was relatively large for lead loading.
Exterior Entryvay (Dust). There were marginally
significant differences in lead loading (p = .07) in the dust
outside the entryways sampled. These differences were due to
significantly higher dust levels at the abated units (p = .03),
Volume II - Page 74
-------
not to higher concentrations of lead in this dust. There was no
difference observed in levels abated by different methods.
House-to-house variation in lead loading at exterior
entryways was marginally significant. This was due to variations
in lead concentration (which were also marginally significant),
not to dust loading variations. Residual log standard deviation
was very large for lead loading (as for the interior entryways).
4.2.2.2 Soil Samples
Lead concentrations were highest in the soil near
foundations, followed by entryway soil and boundary soil (which
had similar levels). Concentrations were about 18 percent higher
in abated houses than in control houses for each of these sample
types, but in no case was this difference significant. There was
significant side-to-side variation for each of the measures.
Side-by-side variation was largest at the entryways, smallest at
the boundaries.
Although abatement was not a significant factor
distinguishing between lead concentration, several other factors
were significantly associated with lead concentration. These are
discussed in the next two subsections.
Entryway (Soil). There was no significant difference
observed between soil concentrations at the entryways of abated
and control units, and there was no difference observed between
units abated by different methods on the exterior.
There was no significant house-to-house difference in
entryway soil lead concentrations, but there were significant
differences between levels observed at different entryways to the
same houses.
Boundary Soil. There was no difference observed
between soil concentrations at the boundaries of abated and
control units, nor were there differences observed between levels
at houses abated by different methods.
There was significant random house-to-house variation,
and significant side-to-side within-house variation.
Volume II - Page 75
-------
Foundation Soil. There was no significant difference
found between lead concentrations in abated units and control
units. Also, differences observed between levels in houses
abated by different methods were not significant. Lead
concentrations were significantly lower in the foundation soil of
houses with more than average abatement performed on the
exterior. Houses where twice as much abatement was performed
outside were found to have 35% lower lead concentrations.
House-to-house differences were not significant, but
side-to-side variation was significant. There was a strong
correlation between the foundation soil lead concentrations
observed in the CAP Study and the XRF/AAS measures taken during
the HUD Demonstration. This relationship is displayed in Figure
4-8. In this figure, lines of best fit are drawn separately for
control and abated houses. Although lead concentrations are
higher on average in abated houses than in control houses, there
is evidently a similar relationship in both groups of houses.
4.2.3 Analysis of Non-Abatement Factors
Table 4-5 displays the effects of non-abatement factors
found to be significantly associated with lead levels. These
included substrate, questionnaire responses, proximity of the
house to a lead smelter, age of the house, etc. The format of
the table is exactly the same as Table 4-4 with an initial column
added to distinguish between classes of related factors. These
classes include substrate, cleanliness, occupation, activities,
age/location of home, ownership, and sampling deviations.
None of these factors was found to be significant for
more than three sample types. For every sample type, lead
loading or lead concentration was observed to be significantly
associated with at least one of these factors.
The substrate from which samples were collected was a
significant factor for window channels and floors. This is
displayed in Figure 4-9 for floors with a box and whisker plot.
(The same format is used in this plot as was used in Section 2
Volume II - Page 76
-------
10000
c
o>
i
ti
0>
5 iocx)
100
10
0.1
X
X'
1.0 10.0
HUD Demonstration XRF/AAS (mgA^m2)
Legend: +, Solid line for Control Units
x, Dashed line for Abated Units
100.0
Figure 4-8. Foundation soil lead concentration vs. HUD Demonstration XRF/AAS levels,
-------
Table 4-5. Multiplicative Effects of Non-Abatement Factors
Type of
Expanatory
Variable
Substrate
Cleanliness
Factor
Substrate Type
Substrate Condition
Frequency of wet mopping uncarpeted
floors
Frequency of window sill dusting
Frequency of vacuum ng uncarpeted floors
Sample
Type
Window
Channel
Floor
(Vacuum)
Entryway
(Interior)
Air Duct
Window
Channel
Air Duct
Air Duct
Floor
(Vacuum)
Entryway
(Interior)
Entryway
(Exterior)
Nominal
Wood (44)**
Carpet (84)
Carpet (47)
Good (82)
Good (48)
12
1
12
12
12
Deviation
Concrete (1)
Metal (33)
Plastic (5)
Concrete (4)
Linoleum (85)
Tile (20)
Wood (40)
Linoleum (26)
Plastic (2)
Tile (7)
Wood (8)
Damaged (1)
Peeling (3)
Chalking (2)
Peeling (33)
6
1
6
6
6
Multiplicative Effect
Lead
Loading
*
0.15
0.37
0.03
*
13.76
0.14
0.12
1.45
*
0.15
0.01
0.03
0.35
*
41
28
2.84
2.25
0.97
0.99
1.02
1.06'
1.00
Lead
Concentration
0.58
0.41
0.23
*
3.56
1.06
1.15
2.70
0.94
0.76
1.08
1.34
*
1.5
6.7
*
4.07
2.35
0.98*
1.03*
1.03'
1.06'
1.05*
Dust
Loading
•
0.28
0.94
0.13
*
4.24
0.14
0.11
0.57
*
0.15
0.01
0.03
0.27
*
28
2.5
0.71
0.95
0.98
0.96
0.99
0.99
0.96*
n
0)
«
(D
^J
00
* Significant at the 10X level.
** Nuifaers in parentheses following substrate types and substrate conditions represent the number of substrates of this type or this condition for this sample
type.
-------
Table 4-5. (Continued)
Type of
Expanatory
Variable
Occupation
Activities
Age/Location
of Unit
Other
Resident
Factors
Sampling
Deviations
Factor
Wearing home work clothes from an
occupation with potential lead
contamination
Resident employed in welding occupation
Resident employed in paint removal
occupation
Frequency of removing paint at home
Frequency of pipe or electrical component
soldering
Proximity to supposed lead smelter
Year house was built
Number of children (7-17)
Ownership of home
Number of months at residence
Substrate Location
Sampling Device
Sample
Type
Window
Stool
Floor
(Vacuum)
Foundation
Boundary
Entryway
(Interior)
Foundation
Boundary
Entryway
(Soil)
Foundation
Boundary
Entryway
(Interior)
Foundation
Foundation
Air Duct
Window
Channel
Nominal
No
No
No
No
0
0
0
> 3 miles
1932
1932
0
Owner
18
Inside Air
Duct (48)***
Large Nozzle (60)
Deviation
Yes
Yes
Yes
Yes
1
1
1
1 mile closer
10 years
10 years
1 child
Renter
1 month
Cover of Air
Duct (38)***
Small nozzle (26)
Multiplicative Effect
Lead
Loading
3.04*
1.49
NA
NA
1.06
NA
NA
NA
NA
NA
0.64*
NA
NA
0.18*
2.61
Lead
Concentration
1.53
1.26°
1.13'
0.45'
1.10'
0.84*
1.46'
1.37'
0.73'
0.84*
0.81*
0.34*
0.94"
0.78
1.18
Dust
Loading
1.96*
1.19
NA
NA
0.97
NA
NA
NA
NA
NA
0.78*
NA
NA
0.26*
2.26*
* Significant at the 10% level.
** See discussion in Section 4.2.4.
*** Number in parentheses following sampling deviation levels
represents the number of samples collected in this manner.
-------
c
3
(D
0)
«
CD
O
100000
5
j
< i
O
10000
7 1000
1OO
10
1
I
Substrate
Figure 4-9. Floor dust lead concentration vs. substrate,
-------
plots.) The corrected geometric means presented in Tables 4-1
through 4-3 are to be interpreted as the mean of the nominal
substrate. Table 4-4 presents the nominal substrate and the
ratio of levels observed for other substrates relative to those
for the nominal substrate. Nominal substrates were chosen
because they were sampled most frequently in the study.
In general, on the floors (including interior
entryways), carpet had higher dust loadings than any of the other
sample types. (There were only four samples taken on concrete,
but dust loading there was higher.) Lead concentrations were
typically highest on wood for all of the sample types where
substrate was found to be significant. Lead loadings were higher
on wood than on carpet for regular floor samples, but the
opposite was true at the entryways. This is because the ratios
of lead concentration and dust loading on wood to those on carpet
for regular floor samples are both twice as high as for
entryways.
The condition of the substrate was also significant,
with damaged, peeling, and chalking substrates noted for higher
lead concentrations. Sampling deviations were also significant
factors. On some air ducts, the cover was not removable and so a
sample was taken from the cover. These samples had one quarter
of the dust loading and lower lead concentrations as compared
with regular samples taken from inside the air ducts. Tor some
window stool and some window channel samples a small nozzle was
used on the end of the vacuum sampler. Lead concentrations and
dust loadings were greater for these samples than for those
collected with the large nozzle. Older homes had significantly
higher soil lead concentrations than newer homes. Lead
concentrations were significantly higher at the foundations and
boundaries of older homes. This is demonstrated for boundary
soil in Figure 4-10.
Other significant factors were less intuitive. For
instance, the frequency of removing paint at home was associated
with houses with higher lead concentrations at the interior
entryways, and lower lead concentrations in soil near the
Volume II - Page 81
-------
M
H
(D
CD
to
1000
100
1
10-
4-
+^
-*--
.-•*•-—+
+
2O
3O
-T"-
40
1 i ' ' ' ' ' ' ' ' ' i ' i '
50 60 70
Age of House (years)
1
80
90
100
Figure 4-10. Boundary soil lead concentration vs. age of house.
-------
foundations. Houses where pipes or electronic parts were
soldered had substantially higher lead concentrations.
There were two houses in which the resident interviewed
stated that the uncarpeted floors were vacuumed every day. In
these houses, lead concentrations were particularly high, and
dust loadings were particularly low for interior entryway,
exterior entryway, and regular floor samples (these relationships
are portrayed in Figures 4-11 and 4-12). Whereas, the frequency
of vacuuming uncarpeted floors was found to be very significantly
associated with lead concentrations and dust loadings for the
houses in the study, when the two houses discussed above were
excluded, the factor was not observed as significant. Except for
a marginal relationship with lead concentration for interior
entryway samples, this factor was not at all significant.
However, for the analyses, data from these two houses are
included.
Finally, it is interesting to note that these non-
abatement factors were primarily significant for lead
concentrations (as were the abatement factors) - as opposed to
lead loadings. This excludes substrate type, sampling
deviations, and an indicator of whether a resident wears work
clothes home from an occupation which involves potential lead
contamination.
4.2.4 Non-Abatement Effects by Sample Type
4.2.4.1 Dust Samples
Air Ducts. One hundred nine (109) air duct samples
were taken. Two of the 109 air duct samples were taken from
baseboard-type heating elements and two others were taken from
cold-air returns. There were differences between results on
these and other types of samples. To avoid making
unsubstantiated conclusions about the impact of these deviations
and to simplify interpretation, these four samples were deleted
from the analyses. Due to common difficulties in removing covers
from air ducts, 46 of the 109 samples were taken from the
Volume II - Page 83
-------
(0
00
100000
10000
1
1000
100
10
1
T <-
o.oooo
10.0000 2O.OOOO
Frequency Uncarpetted HOOTB Vacuumed
—'—i—
30.0000
Figure 4-11 Exterior entryway dust lead concentration vi
frequency of vacuuming uncarpeted floors.
-------
c
3
ro
(D
CO
100000
10000
1000
100
10-
0.0000
—I—'—•
10.0000 20.0000
Frequency Uncarpetted HOOT* Vacuumed
I I I
30.0000
Figure 4-12
Exterior entryway dust loading vs. frequency
of vacuuming uncarpeted floors.
-------
exterior fins or grates covering the air ducts. The remaining 59
samples were taken from inside the air ducts. This had a
significant impact on the results. The substrate condition was
also observed to have a significant effect. Table 4-5 presents
estimates of these effects.
Lead loadings were substantially lower in samples taken
from the exterior grates. This was mainly due to significantly
lower dust loadings, but concentrations were also slightly lower
(though not significantly lower). One air duct was damaged and
three air ducts had peeling substrates. Lead levels were
significantly higher on the damaged and peeling substrates.
There were significant differences in lead
concentrations among houses abated by different contractors.
Four contractors (denoted A, B, C, and D) were employed to abate
the units in this study. There was as much as a three-fold
difference in concentrations in houses abated by different
contractors. The differences are presented in Table 4-5 as
ratios to levels in control houses. For the two contractors (B
and C) which abated most of houses, lead loadings and lead
concentrations were very similar.
Lead concentrations were significantly lower in houses
where there was frequent wet-mopping of uncarpeted floors (p =
0.4). In houses where the window stools were frequently dusted,
there were higher concentrations in the air ducts.
Window Channels. Substrate and condition of substrate
were important factors associated with lead levels in window
channels. Thirty-six (36) of the channels were made of metal; 44
were made of wood. Differences in lead loadings on these were
significant. For concentrations, the differences were not
significant. Lead loadings were 38% lower on metal than on wood.
Conditions of these substrates were primarily either good or
peeling. These differences were shown to have an association
with lead concentrations. On peeling surfaces, concentrations
were 2.36 times as high as on channels which were intact.
Volume II - Page 86
-------
Twenty-seven (27) percent of the window channel samples
were taken with the small nozzle attached to the vacuum. Lead
loadings were estimated to be 2.6 times higher in these samples,
but this was not estimated to be a significant effect due to the
large variation.
Window Stools. The two most significant practical
effects associated with window stool levels were related to
abatement and clearance times. The houses abated in the HUD
Demonstration in Denver were abated in three different phases
according to the magnitude of abatement required. The worst
units were abated first. Table 4-4 indicates higher lead levels
were found in homes abated earlier. Table 4-4 presents these
estimated effects for lead loadings, lead concentrations, and
dust loadings.
There was also a significant relationship between lead
levels and the number of months between clearance and CAPS
sampling for the unit. Longer times since clearance were
associated with lower lead levels.
Finally, higher lead loadings were observed in houses
where a resident wears work clothes home from an occupation with
potential lead exposure. This was not due to higher
concentrations of lead in these houses, but there were higher
dust loadings found.
Interior Entryway. The most influential variable for
lead loading appeared to be substrate, with highest loadings
observed in samples taken from carpets. Most of the samples were
taken on carpet and linoleum with smaller numbers taken on tile
and wood floors. Lead loadings were about six times higher on
carpet than on linoleum; three times higher on carpet than on
wood; and about 30 times higher on carpet than on tile. This
difference is attributed to greater levels of dust retained by
the carpet, since there were not significant differences in
concentrations among these substrates.
There were significantly higher lead loadings in homes
where there was frequent vacuuming of uncarpeted floors. Higher
Volume II - Page 87
-------
concentrations were observed in houses where paint removal was
recently done. Lower loadings and concentrations were observed
in houses where there were more children between the ages of 7
and 17.
Exterior Entryway (Dust). Aside from abatement, only
frequency of vacuuming uncarpeted floors was found to be
significantly related to levels of lead in the dust outside the
entryways to these homes. Lead concentrations were found to be
higher in houses where vacuuming of uncarpeted floors was more
frequent. Dust levels were lower in these houses. These two
relationships combined to yield no association between the factor
and lead loading.
Floor (Wipe). Rented homes had lead loadings on floors
62 percent lower than those in owner-occupied homes.
Floor (Vacuum). Perhaps the most significant factor
associated with floor lead levels was substrate. Most of the
samples were taken on carpet (84), linoleum (85), tile (20), and
wood (40). Of these, dust loading was, by far, greatest on
carpets. Lead concentrations were similar on carpet, linoleum,
and tile, but on wood they were almost three times as large.
This contributed to the highest lead loadings (excluding four
samples taken on concrete) on wood. They were about 50 percent
higher on wood than on carpet - much lower on linoleum and tile.
In houses where uncarpeted floors were vacuumed
frequently, there were significantly higher lead concentrations.
Dust loadings were slightly lower in these houses, so lead
loadings were not significantly higher. Homes in which a
resident was employed in welding had lead concentrations four
times as large as homes which did not. In those same houses,
dust loadings were more than twice as high, contributing to lead
loadings more than 10 times as great.
4.2.4.2 Soil Samples
Entryway Soil. The only non-abatement factor found to
be significantly associated with entryway soil lead levels was a
Volume II - Page 88
-------
crude measure of proximity to an establishment described as a
lead smelter. Houses were classified as being either less than 1
mile away, 1 to 2 miles away, 2 to 3 miles away, or greater than
3 miles away from the smelter. The geometric mean in Table 4-2
is to be interpreted as being representative of houses more than
four miles from the smelter. Figure 4-13 displays lead
concentrations versus these classifications. Viewing this
measure as roughly continuous, lead concentrations were estimated
to increase by 37 percent with each mile closer to the smelter.
Foundation Soil. Several factors were very
significantly associated with lead concentrations in foundation
soil. Age of the house, ownership, and length of residence were
all very significant factors. Older houses had much higher
levels. A ten-year difference in age was associated with a
difference of 27 percent in lead concentrations near the
foundation. However, lead levels were lower in houses where the
residents have lived longer since abatement. About a 6 percent
drop per month was estimated. Controlling for the other factors,
lead concentrations around homes rented by their residents were
34 percent of those in homes owned by their residents. Another
factor found to be significantly associated with lower lead
concentrations was recent paint removal at the house. Lead
concentrations were twice as high around houses where a resident
was employed in a welding occupation.
Boundary Soil. Lead concentrations in boundary soil
were significantly associated with the age of the house. The
average age of houses in the study was about 60 years at the time
of sampling. An increment in age of 10 years was associated with
an increment in lead concentration of about 22 percent. From
Figure 4-10, we see that log concentrations increased fairly
linearly with age of house. Three homes were observed in which a
resident was employed in an occupation involving paint removal.
In these houses, lead concentration was significantly (p = .04)
lower (55 percent lower). Table 4-4 indicates there was also a
significant association found between lead concentration in
Volume II - Page 89
-------
10000
1000
c
3
(D
100
vo
o
1
10
>3 miles
2-3 miles
Proximity to
1-2 miles
Smelter
O-1
Figure 4-13. Entryway lead concentrations vs. proximity to lead smelter.
-------
boundary soil and the frequency with which pipes or electronic
parts were soldered in the last 6 months. Levels were
significantly higher in such-houses (p=0.02).
Volume II - Page 91
-------
5.0 CORRELATIONS
Section 4 summarized the relationship between lead
levels and various abatement, sampling and other factors by
sample type. Here we discuss correlations of lead levels between
the various sample types after correction for the estimated
effects of the factors discussed in Section 4. Thus, these
correlations should be interpreted as relationships between
different sample types above and beyond that which are explained
by things like abatement, age of house, cleanliness measures, and
other factors included in the models.
This analysis involves examining correlation matrices
and scatterplot matrices. The primary data used to examine these
relationships are the estimated random unit (house) effects and
the estimated random location-within-house effects. Both of
these random effects are estimated after controlling for the
estimated fixed effects in the model for each sample type.
5.1 UNIT-TO-UNIT CORRELATIONS
The correlation matrix of random unit-to-unit
differences in lead loading is presented in Table 5-1. To locate
a correlation of interest, locate the row corresponding to the
first sample type and the column corresponding to the second
sample type. Correlation information for the two sample types is
presented in the corresponding box. Within each box, three
values are presented:
Top value: Correlation coefficient between the
logarithms of the geometric unit means,
Middle value: Degrees of freedom used in
calculating the correlation coefficient, and
Bottom value: Observed significance level of the
test of the hypothesis of no correlation
(correlation coefficient equal to zero).
Volume II - Page 92
-------
Table 5-1.
Correlations* Among Sample Types for Unit-Level
Random Effects: Lead Loading
Vacuum
Air Duct
Window
Channel
(Vacuum)
Window Stool
(Vacuum)
Floor (Wipe)
Entryway
Exterior
(Vacuum)
Vacuum
Air Duct
Window Channel
(Vacuum)
.178
33
.31
Window Stool
(Vacuum)
.052
36
.76
.641
37
.00
•
Floor
(Wipe)
.323
22
.12
.110
23
.60
.075
25
.71
Entryway
Exterior
(Vacuum)
.409
36
.01
.169
37
.30
.081
42
.60
.333
29
.07
(D
•0
01
vfl
(D
VO
U)
* Top number is estimated correlation; middle number is degrees of freedom; and bottom
number is significance level.
-------
Only the upper right-hand half of the matrix, above the shaded
diagonal, is filled in since the lover left-hand half of the
matrix would contain redundant information.
When controlling for the fixed effects, degrees of
freedom for the estimation of correlation are specified to
estimate the fixed effects. This was accounted for in the
significance levels and the degrees of freedom displayed in the
correlation tables.
The following method was used to calculate degrees of
freedom for estimating the unit-level correlation of two sample
types, A and B:
1. Let mA/B denote the number of houses from which
samples of both types were taken, and
2. Let fL denote the number of unit-level fixed
effects in the model fit for sample type i
(i=A,B).
3. dfA/B = mA/B - max (fft, fB) - 2.
In most cases we had at least 30 degrees of freedom. Estimates
of correlations with floor wipe samples had fewer degrees of
freedom because the samples were only taken in the abated houses.
Some sample types are not represented in the unit-level
correlation analysis. This is because in some cases the
restricted maximum likelihood (REML) estimates of the random
unit-to-unit differences were negligible after controlling for
the fixed effects. This happened in the case of interior
entryway lead loadings and floor lead loadings and
concentrations.
The lead loading random unit effect estimates are
presented graphically in Figure 5-1. This figure is a
scatterplot matrix, or a collection of bivariate plots organized
into matrix form. As with the correlation matrix, to locate a
Volume II - Page 94
-------
ARD
WCH
WST
FLW
EWO
Figure 5-1. Scatterplot matrix of unit-level random
effects for different sample types: lead
loading (jig/ft2).
Volume II - Page 95
-------
plot of interest, identify the row associated with one sample
type and the column associated with the other sample type. The
plot is presented in the corresponding box. Within each box, the
horizontal axis represents increasing values of the column
variable on a logarithmic scale. Similarly, the vertical axis
represents increasing values of the row variable on a logarithmic
scale. The abbreviations employed on the diagonal to identify
the different sample types are defined in Table 2-1.
The ellipse plotted in each box of Figure 5-1 is the
ellipse that contains 95% of the probability associated with the
estimated bivariate normal distribution for the plotted data.
The narrower the ellipse, the stronger the correlation between
the two sample types. If the ellipse is oriented from the lower
left-hand corner of the box to the upper right-hand corner of the
box, the sample types are positively correlated. If, on the
other hand, the ellipse is oriented from the upper left-hand
corner of the box to the lower right-hand corner of the box, the
sample types are negatively correlated.
Table 5-2 contains unit-to-unit correlation estimates
for lead concentrations; Table 5-3 provides the same for dust
loading. Figure 5-2 is the analog to Figure 5-1 for lead
concentrations; Figure 5-3 provides the same information about
dust loadings.
There were several indications of a positive unit-level
correlation between different sample types. No significant
negative correlations were observed. Thus, unexplained (not
accounted for by the models) differences between lead and dust
levels in different houses appear to be similar for certain pairs
of sample types.
The strongest correlation in lead loadings was observed
between window channels and window stools. The estimated
correlation was 0.64 with 37 degrees of freedom. This was highly
significant.
Examining Figures 5-2 and 5-3 reveals that this
relationship is due to positive correlations in both lead
Volume II - Page 96
-------
Table 5-2.
Correlations* Among Sample Types for Unit-
Level Random Effects: Lead Concentration
Vacuum
Soil
Window
Channel
(Vacuum)
Window
Stool
(Vacuum)
Entryway
Interior
(Vacuum)
Entryway
Exterior
(Vacuum)
Entryway
(Soil)
Foundation
(Soil)
Boundary
(Soil)
Vacuum
Window
channel
(Vacuum)
Window
Stool
(Vacuum)
.444
37
.00
Entryway
Interior
(Vacuum)
.208
38
.20
.099
42
.52
Entryway
Exterior
(Vacuum)
.264
37
.10
-.006
42
.97
.253
43
.09
Soil
Entryway
(Soil)
.276
38
.08
.245
43
.11
.255
44
.09
.238
46
.10
Foundation
(Soil)
.158
24
.44
.199
28
.29
.114
28
.55
.213
28
.259
.367
29
.04
Boundary
(Soil)
.135
38
.41
.266
43
.077
.121
44
.42
-.148
44
.327
.531
45
.00
.034
29
.86
o
(D
H
•tJ
0)
|Q
(D
\O
Top number is estimated correlation; middle number is degrees of freedom; and bottom number is
significance level.
-------
Table 5-3.
Correlations* Among Sample Types for Unit-Level
Random Effects: Dust Loading
Vacuum
Air Duct
Window
Channel
(Vacuum)
Window
Stool
(Vacuum)
Floor
(Vacuum)
Entryway
Exterior
(Vacuum)
Vacuum
Air Duct
Window Channel
(Vacuum)
-.297
33
.08
Window Stool
(Vacuum)
.009
36
.96
.388
37
.01
Floor
(Vacuum)
.134
37
.42
.097
38
.55
.278
43
.06
Entryway
Exterior
(Vacuum)
.328
36
.04
.013
37
.94
.111
42
.47
.333
44
.02
o
•0
0>
05
(D
\O
00
Top number is estimated correlation; middle number is degree of freedom; and bottom
number is significance level.
-------
WCH
W8T
EWI
EWO
BDY
EWY
FDN
Figure 5-2.
Scatterplot matrix of unit-level random effects for
different sample types: lead concentration (/ig/g) .
Volume II - Page 99
-------
ARD
WCH
WST
FLR
EWO
Figure 5-3. Scatterplot matrix of unit-level random effects for
different sample types: dust loadings (/ig/g).
Volume II - Page 100
-------
concentrations and dust loading. The other significant
correlation observed for lead loadings was that between air duct
and exterior entryway lead loadings.
The unit-to-unit variation in air duct lead
concentrations was negligible (refer to Table 4-2). However,
there was significant correlation observed in dust loadings for
these two sample types. That is, at houses where much dust was
found at the exterior entryways, there was also much dust found
in the air ducts.
There were also significant correlations observed in
soil lead concentrations at different property locations.
Entryway soil lead concentrations were significantly correlated
with both boundary concentrations (53%, p < .005) and foundation
concentrations (37%, p = .04). The correlation between boundary
and foundation lead concentrations was not significant.
Although unit-to-unit variation in lead loading and
lead concentration was negligible in vacuum floor samples, dust
loading was not. There was significant correlation observed
between dust loading on interior floors and exterior entryways
(33%, p = .02). There was also a marginal correlation observed
between dust loading on floors and window stools (28%, p = .06).
5.2 ROOM-TO-ROOM CORRELATIONS
Whereas the previous section discussed house-to-house
variations in lead and dust levels, this section discusses
within-unit correlations among sample types. Thus, the purpose
of this analysis is to determine if there is significant co-
variation in lead levels as one moves from room to room or side
to side at a house.
For interior dust samples (except floor samples), there
was typically only one sample taken per room. For these sample
types, it was impossible to estimate random room effects apart
from within-room variation. Residuals from the fit of the full
model were used in the correlation calculations.
Volume II - Page 101
-------
For floor and soil samples, the model included a
room/side level random effect term for each room/side sampled.
Therefore for these sample types, residuals from the full model
were added to the estimates of the room/side levels random
effect. There were many rooms/sides where more than one vacuum
floor sample was taken. In these cases, (for purposes of
correlation estimation) the residual estimates were averaged.
We used the following method to calculate degrees of
freedom for estimating the within-unit correlation of two sample
types, A and B:
1. Let hAfB denote the number of houses from which
samples of both types were taken, and
2. Let !AfB denote the number of locations from which
both sample types were taken, and
3. Let ff denote the number of room-level fixed
effects in the model fit for sample type (i=A,B).
4- dfA,B = ^B-VB-^^A'fB)-2-
Estimates of correlations with floor wipe samples had fewer
degrees of freedom because the samples were only taken in the
abated houses. Therefore the correlations presented in this
section are really those of room-to-room plus within-room
variation among the different dust sample types.
Table 5-4 presents these correlations for lead loading;
Table 5-5 presents the correlations for lead concentrations; and
Table 5-6 presents the correlations for dust loading. The format
used in these tables is the same as that of Tables 5-1, 5-2, and
5-3. Figure 5-4 displays scatterplot matrices of room-level
differences in lead loadings; Figures 5-5 and 5-6 provide the
same for lead concentrations and dust loadings.
All significant room-level correlations were positive.
The strongest correlations were observed between window stools
and window channels. The estimated correlation between room-to-
room variations in lead loading was 33% (p=.00), for lead
Volume II - Page 102
-------
Table 5-4.
Correlations* Among Sample Types for Rooi
Level Random Effects: Lead Loading
Vacuum
Soil
Air Duct
Window
Channel
(Vacuum)
Window
Stool
(Vacuum)
Floor
(Wipe)
Floor
(Vacuum)
Entryway
Interior
(Vacuum)
Entryway
Exterior
(Vacuum)
Air
Duct
Window
Channel
(Vacuum)
.052
>5- !?
.73 .ft1
Window
Stool
(Vacuum)
.174
70-^-5
-14 .?/
.327 ,
9* tf
.00 ,/6
Floor
(Wipe)
X*7Tx--vV.-r •—
T
_/»^JV-5>«
*$w
'~83ffl>
f
Floor
(Vacuum)
.044
80 #?
.69 ?JL
.086 .
98 Z1
•39 ,,/>
.150
134 V?
.08 .SO
IT
Entryway
Interior
(Vacuum)
s-r«s- r^'-rt-'J
*$&?
/&
y^STN
^r OO
^r • O&
*r-*-vK.-r*r
?&&(/
/85s
•JJ7
Entryway
Exterior
(Vacuum)
•
/^MW
-.033
K ^
• 90 .f7
T$P?
A
IT
.139
77 3*
.22 .fr
i
n>
H
•d
01
03
(D
o
w
Top number is estimated correlation; middle number is degrees of freedom; and bottom number is
significance level.
-------
Table 5-5.
Correlations* Among Sample Types for Room-
Level Random Effects: Lead Concentration
Vacuum
Soil
Air Duct
Window
Channel
(Vacuum)
Window
Stool
(Vacuum)
Floor
(Vacuum)
Entryway
Interior
(Vacuum)
Entryway
Exterior
(Vacuum)
Entryway
(Soil)
.yjctnntT
Air
Duct
Window
Channel
(Vacuum)
.084 „
& 5
• 5-7 ?:-
Window
Stool
(Vacuum)
.434
x >-
•JBO .S.
.324
33 ;s
.00 ,
Floor
(Vacuum)
.100 .,
*1 /'
•>7 i,
.103
99 ;: •
.31 /;.
-.011
13X 02^
/^V
-.o&r'
>1 *•
X8S
^4)98 /
*^*XL
^^X
•il^
x^V *
/.19X
Entryway
Exterior
(Vacuum)
•
?'
-.062
>5 „•
.8i '•
X-
•374 „,.
76 ^>'
.06 0:-
Entryway
(Soil)
•
.17^
y^74
.330
^ ^
.^ . //
N^ss/
-.031
87 5>
.yi .*±
-.074
^4 f/
.«o .^/
o
(-•
i
(D
•d
01
iQ
(D
Top number is estimated correlation; middle number is degrees of freedom; and bottom number is
significance level.
-------
Table 5-6.
Correlations* Among Sample Types for Unit-
Level Random Effects: Dust Loading
Vacuum
Soil
Air Duct
Window
Channel
(Vacuum)
Window
Stool
(Vacuum)
Floor
(Vacuum)
Entryway
Interior
(Vacuum)
Entryway
Exterior
(Vacuum)
JE
Air
Duct
Window
Channel
(Vacuum)
.244
vr 1
.ja j'»
3n»v
Window
Stool
(Vacuum)
-.133
s*Q 2 ./
ytf . 5
.207
92
.05
Floor
(Vacuum)
.114
^»0 '^'/'
**1 .£,:•
-.055
<* 2'f
•&* ,1V
.236 „
«"4 / ~
.01 /'
SSfi.
Entryway
Interior
(Vacuum)
X-
/57^
V4
/H '
/.iV,
5T'
'J-
Entryway
Exterior
(Vacuum)
•
V087"
\r *
X5a
.145
)A --
v&9 - V-f
13/6
VQ *
XQS
.035
JT 3'
<>6 ,;r
o
ID
H
0*
iQ
A
o
ui
Top number is estimated correlation; middle number is degrees of freedom; and bottom number is
significance level.
-------
ARD
WCH
WST
FLW
FLR
EWI
EWO
Figure 5-4. Scatterplot matrix of room-level random
effects for different sample types: lead
loading (/ig/f t2).
Volume II - Page 106
-------
ARD
WCH
WST
FLR
EWI
EWO
EWY
Figure 5-5. Scatterplot matrix of room-level random effects for
different sample types: lead concentration (jxg/g) .
Volume II - Page 107
-------
ARO
WCH
WST
FLR
EWI
EWO
Figure 5-6. Scatterplot matrix of room-level random effects for
different sample types: dust loadings (/xg/g) .
Volume II - Page 108
-------
concentration it was 32% (p=.00), and for dust loading it was 21%
(p=.05). Each of these results was statistically significant.
There was marginal correlation observed between the lead loadings
for window stools and floor (vacuum) samples (15%, p = .08).
The correlation between air duct and window stool lead
concentrations was highly significant (43%, p=.00), but the
corresponding estimated correlation in dust loading was negative
(-13%, p=.27), negating any correlation in lead loading.
The only other significant room-to-room correlation in
lead concentration was between interior and exterior entryway
dust samples (37%, p=.00). These were not at all correlated with
lead concentrations in entryway soil samples, despite the fact
that these estimates are based on many degrees of freedom.
Floor dust loadings were significantly correlated with
interior entryway dust loadings (24%, p=.oi), but no association
was found for lead levels.
Volume II - Page 109
-------
6.0 WIPE VERSUS VACUUM COMPARISON
The two major HUD programs investigating levels of lead
in household dust utilized different sampling methods. In the
Demonstration Study, dust was wipe sampled. In the National
Survey, dust was vacuum sampled. As part of the CAP Study,
several side-by-side dust samples were taken by the wipe and
vacuum sampling methods.
To investigate the relationship between lead loading
determinations made by the two methods, four side-by-side samples
were taken from a selected room in each abated unit. Two of the
samples were taken by the vacuum method and two by the wipe
method. Samples were collected in 34 of the 35 abated houses
sampled. In one house (589), all floors were carpeted so no
wipe/vacuum comparison samples were taken. Of the 34 houses
sampled, one of the comparison samples in unit 585 was lost
during analysis. This also happened to be the only unit in which
the wipe/vacuum comparison samples were taken from a concrete
floor. The three observed loadings were substantially higher
than corresponding measures in all the other houses. The
analysis was performed both with and without the data from this
house. Since the results are only slightly different when this
unit is excluded, it is included in the results provided.
The geometric means of the paired floor lead loadings
are listed in Table 6-1 and plotted in Figure 6-1. In the
figure, lead loadings from vacuum samples are plotted versus lead
loadings from wipe samples. A solid reference line which
represents complete agreement between the two sampling methods is
also plotted along with the best fit regression line. A
statistical analysis was performed to quantify this relationship.
This is discussed in Section 6.1. Samples taken on different
substrates (linoleum, wood, concrete, or tile) are distinguished
by different plotting symbols in Figure 6-1. Since the
relationships between vacuum and wipe responses were different
for each substrate, the analysis was also performed adjusting for
substrate. This analysis is discussed in Section 6.2.
Volume II - Page 110
-------
Table 6-1.
Vacuum versus Wipe Comparison Data: Room
Geometric Mean Floor Lead Loadings (ug/ft2)
Substrate
Concrete
Linoleum
Tile
Hood
Unit
585
502
517
525
532
537
540
543
554
569
570
571
574
577
578
580
581
586
587
590
548
563
572*
579
588
510
511
512
515
528
544
562
566
584
Location
LDY
KIT
HAL
KIT
BAT
BAT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
BAT
BSM
KIT
KIT
HAL
KIT
KIT
BA2
KIT
KIT
KIT
KIT
B02
KIT
KIT
LDY
LVG
LVG
DIN
KIT
DIN
Vacuum
Loadino
4075.33
6.07
3.89
2.84
38.93
0.85
5.63
26.77
34.81
51.23
1.03
980.96
11.83
1.03
4.57
21.35
3.47
87.02
2.17
1.55
1.14
3.19
552.54
2.06
5.24
48.26
195.17
27.06
206.14
10.53
104.66
175.91
11.24
183.66
Wipe
Loadincr
333.56
3.96
3.84
3.56
10.41
18.07
6.85
7.34
5.82
4.00
5.18
21.10
7.37
4.83
5.57
23.31
39.70
52.69
7.30
6.94
2.86
13.37
69.37
3.64
13.05
26.91
14.76
26.92
4.24
10.56
6.26
24.71
26.61
28.97
An outlier analysis of the vacuum-wipe ratios revealed this ratio as
significantly different. It was excluded in the estimation of
multiplicative biases by substrate.
Volume II - Page 111
-------
i
n>
•n
0
*a
ro
10000.0
1000.0
100.0-
10.0
1.0
0.1-
0.1
1.0
a n a
10.0
100.0
1000.0
10000.0
Wipe Pb Laadng (u»H2)
Unoteum xxx Wood * * * Tile
Concrete y-x Line Beat-m Una
Figure 6-1. Vacuum versus wipe comparison: geometric means of side-by-side
floor lead loading (/tg/ft2) measures. (Estimate of vacuum/wipe ratio
is 1.42; confidence interval is (0.78, 2.60).)
-------
The effect of room type on the wipe/vacuum relationship
was also investigated. Categories such as wet versus dry and
eating versus non-eating were considered. No significant
differences were observed.
6.1 ALL SUBSTRATES COMBINED
It is assumed that the relationship between vacuum and
wipe measures is log-linear:
log(V) = log(a) + j8 log(W) (1)
where V and W represent the true expected loadings by the vacuum
and wipe methods. Restating the model in terms of the
untransformed loadings we have
V = a W*. (2)
If 0 is not equal to one, the multiplicative bias between the two
sampling methods changes with the magnitude of the measurements.
However, if 0=1, there is a fixed multiplicative bias (a) between
the sampling methods which does not change with the magnitude of
the measurements. Also, for j8=l, the model of Equations (1) and
(2) simplifies to the assumption that the ratio W/V follows a
lognormal distribution with geometric mean a.
Since the vacuum and wipe determinations are both
measured with error, a simple linear regression for (1) is
inappropriate. An errors-in-variables approach was used.
Specifically, we do not observe V and W in (1), but rather we
have V* and W* where
log(V*) = log(V) + log(6), and
log(W*) = log(W) + log(5),
Volume II - Page 113
-------
with 6, S independent, and lognormally distributed. We also
assume that W is lognormally distributed. Using simple linear
regression produces biased estimates of a and 0. However,
formulas to correct for these biases are well known, and were
used in the results that follow.
The first step was to test the hypothesis of a fixed
multiplicative bias (H0:0=l). The estimate of 0 was 1.40 with a
standard error of 0.37. Since the hypothesis could not be
rejected at any reasonable significance level (p=0.29), the model
was then refitted with the 0 parameter set to one. The estimate
of the multiplicative bias (a) of vacuum over wipe measurements
is 1.42 with a 95% confidence interval of (0.78, 2.60). This
result implies that, on the average, vacuum lead loadings are
1.42 times larger than matching wipe lead loadings on floors.
The precision of the vacuum and wipe measurements is
also a relevant quantity. On average, side-by-side vacuum
measures were significantly more variable than wipe measures.
The estimated log standard deviation for vacuum samples was 0.95
with a 95 percent confidence interval of (0.77, 1.25) whereas for
wipe samples it was 0.55 with a 95 percent confidence interval of
(0.45, 0.73).
6.2 ADJUSTING FOR SUBSTRATE EFFECTS
The above approach was used to investigate the
vacuum/wipe relationship separately for each of the substrate
categories sampled. For each of the substrates, the hypothesis
of a fixed multiplicative bias (0=1) could not be rejected at any
reasonable level. There was only one set of side-by-side
comparison samples taken on concrete, so this hypothesis was not
tested for concrete. We only remind the reader that the highest
lead loadings observed for wipe/vacuum comparison samples
occurred on concrete by both methods. The estimated biases vary
according to substrate. There appears to be a relationship
between the smoothness of the substrate and these biases. Table
6-2 displays the estimated multiplicative bias for each substrate
Volume II - Page 114
-------
along with confidence bounds. The ratio observed on wood was
significantly different from the ratios observed on both linoleum
and tile. The bias appears to increase with coarseness of the
substrate. If the wipe method fails to extract dust particles
embedded in recesses on the substrate surface then this
relationship would be expected.
Table 6-2. Vacuum/Wipe Multiplicative Bias Estimates
Substrate
Tile
Linoleum
Wood
Concrete
Observations
4
19
9
1
Estimated
Vacuum/Wipe
Multiplicative
Bias
0.69
1.01
4.15
12.22
Lower
Confidence
Bound
0.1S
0.47
1.36
0.43
Upper
Confidence
Bound
3.06
2.17
12.65
344.87
Volume II - Page 115
-------
7.0 COMPARISONS WITH OTHER STUDIES
The environmental sampling results of the CAP Study may
be compared to those from other studies. In particular,
comparisons to the earlier CAP Pilot Study, the HUD Abatement
Demonstration Project, and other studies assessing the efficacy
of an abatement procedure seem most applicable. Chapter 7.1
compares the results from the pilot and full CAP studies. A
comparison to the HUD Demonstration results is presented in
Chapter 7.2. Finally, CAP results are compared to the results of
other abatement efficacy studies.
7.1 COMPARISON OF CAP STUDY DATA AND CAP PILOT STUDY DATA
The CAP Pilot Study investigated field, laboratory, and
statistical analysis procedures planned for the CAP Study. As
such, it employed comparable sampling techniques and locations.
A complete discussion on the Pilot Study is available in another
report (Battelle and MRI, 1991).
Of the six residential units surveyed in the Pilot
Study, five were revisited in the CAP Study. Figure 7-1 displays
the differences for those five units between the CAP Pilot and
full CAP studies in geometric mean lead loading results, by
sample type. Similar plots for lead concentration and dust
loading are portrayed in Figures 7-2 and 7-3, respectively. Each
line segment in the figures represents the change in lead loading
for a particular unit and sample type. For example, the floor
lead loading results were higher in the full Study than in the
Pilot for all units except Unit 51. In the figure, this is
evidenced by the appropriate line segments rising from left to
right.
As the figures suggest, there is no single pattern of
change across the various sample types. For example, a
particular unit may have higher air duct lead loadings in the CAP
Study, but lower window channel lead loadings. The floor and
entryway interior changes are the most similar unit to unit,
especially for the lead loading and dust loading results.
Volume II - Page 116
-------
10000
1000
(D
7
»«
to
100
*
1 -
I /
v
• \
ARO
Unit
WCH
17
\ /
• \ /
\\ /
\ '
V
Xs
WST FLW FUR
Sample Type
19 33 61 80
'/
1 /
» '
'
'
EW1
Figure 7-1. Comparison of CAP Pilot and CAP Study results: unit geometric mean
lead loading (Mg/ft2) by sample type.
-------
1OOOO-
3 iooo
< 5
o *|
I I
H &
I
(D
H
H
00
10
\
\
V '
^ ^ \
. v \ N v!
\ ~ V — . \ v / \. > / '
\ ^ \^ 'A » '
^C \ xV x /\/ \-
\-\ \ / \ '< xx
Ni / ^i\
^ \^
o
\
A W W F E E
R C 8 L W W
D H T R I Y
Sample Type
I lr-v»+- 4^ H(1 - AA «•-«
— •
^=^
F B
D D
M Y
dhtffc
Figure 7-2. Comparison of CAP Pilot and CAP Study results: unit geometric mean
lead concentration (M9/9) by sample type.
-------
9
o
o
T-
Vol
e II
>U
5
ro
9
9
o
^
9
o
.
K /
Vs /
ARD
Unit:
- >••
WCH WST
Sample "Type
17 10 33
FLR
80
/ /
1 i
/ i '
r, i
' ! !
EWI
Figure 7-3.
Comparison of CAP Pilot Study and CAP Study results: unit geometric mean
dust loading (mg/ft2) by sample type.
-------
The window channel and stool results, in turn, were the least
consistent. Not surprisingly, the soil lead concentration
measurements did not change significantly in the time between the
two studies. Despite the greater efficiency of the dust sampler
in the CAP Study, the dust loading unit geometric means did not
all increase. In fact, the dust loading results for Unit 51 were
usually lower in the CAP Study than the Pilot Study. Only the
air ducts had on average a decrease in dust loadings. The
greatest geometric mean increase in dust loading, 9.5 times,
occurred for floor samples. Since the CAP Study only collected
wipe samples from abated units, only three units had two sets of
wipe samples. The lead loading results in those units decreased.
It is noted above that a more efficient dust vacuum
sampler was utilized in the CAP Study. When revisiting the Pilot
units in the CAP Study, an attempt was made to collect dust
samples from the same room and component. Figure 7-4 presents a
comparison of the dust loading results from these two studies.
The dust loading results for the Pilot Study are plotted versus
those for the CAP Study. The different sample types are
indicated by individual plotting symbols. The cloud of points
and its location are somewhat surprising. Given the greater
efficiency of the sampler used in the CAP Study, one might have
expected the CAP Study dust loadings to be consistently higher.
Evidently, other factors such as time, occupancy, and sample-to-
sample variation overwhelm any obvious difference in sampling
efficiency.
7.2 COMPARISON OF CAP STUDY DATA AND HUD
ABATEMENT DEMONSTRATION DATA
The HUD Abatement Demonstration project included the
collection of detailed environmental data at all units, both
abated and control. Each unit underwent extensive XRF/AAS
testing to assess the amount of lead-based paint. Units found to
contain lead-based paint had additional environmental samples
Volume II - Page 120
-------
100OOO
I
0)
Ifi
n>
10000
1000
100
10
10
Component: + + + ARD o o o EWI
100 1000
GAP Pilot Study
ODD FLR xxx VVCH
10000
n n n VVST
I
100000
Figure 7-4. Comparison of CAP Pilot Study and CAP Study results: component geometric
mean dust loadings (mg/ft2) by sample type.
-------
collected. Those samples were from individual components within
a room and soil core samples (both before and after the
abatement) collected on all four sides of the unit. The HUD
Demonstration dust and post-abatement soil samples were collected
between November 1989 and July 1990. The pre-abatement soil
samples were collected between August and December 1989. The CAP
Study results, in turn, were obtained in March and April 1992.
Though a seasonal effort may be influencing the comparisons that
follow, it cannot be separated from other differences between the
projects such as sampling protocols.
Figure 7-5 contrasts the CAP Study floor dust lead
loading (jig/ft2) results to those from the HUD Demonstration.
For the CAP Study, geometric mean dust lead loadings are
calculated for all floor dust vacuum and wipe samples collected
within a room and unit. Since the dust samples collected in the
HUD Demonstration project were part of the clearance procedures,
only the final floor dust wipe sample collected in a room was
retained. Figures 7-6 and 7-7 present similar comparisons for
window stools and window channels, respectively. Recall that in
the CAP Study, dust wipe samples were collected only on the
floors of abated units. As is evidenced in the figures, there
appears to be little agreement between the CAP Study results and
those from the HUD Demonstration. The higher dust lead loadings
from the CAP Study, most apparent for the window channel samples,
may be due to increased lead concentration in the dust or the
greater efficiency of the vacuum sampler employed.
For purposes of comparison, a geometric mean XRF/AAS
result (mg/ft2) was calculated by room and unit from the
extensive HUD Demonstration XRF/AAS measurements within the room.
Figure 7-8 compares the CAP Study floor dust lead loading results
(both wipe and vacuum) and the HUD Demonstration dust wipe lead
loadings to these room geometric mean XRF/AAS results. Similar
comparisons are portrayed for window stools (Figure 7-9) and
window channels (Figure 7-10). The resulting clouds of points
suggest little or no correlation between dust lead loading and
Volume II - Page 122
-------
100000-1
<
o
Id
(D
M
u>
10000
1
I
1000
100
10
1-
1.O
10.0 100.0 1OOO.O 1000O,O
HUD Demo Floor Duet Lead Loading (uo/ft2)
Sample Origin: + + + CAP Study Vacuum o o o CAP Study Wipe
100000
Figure 7-5. CAP vacuum and CAP wipe vs HUD Demonstration wipe results:
geometric mean floor lead loading by room.
-------
100000
<
o
01
vQ
(D
N)
10000
I
1000
100
1 ' ' ' 'I— —' '—'—' ' ' '' I
1O.O 100.0 1000.0 10000.0
HUD Demo Window Stool Duct Lead Loading (uo/ft2)
100000
Figure 7-6. CAP vacuum versus HUD Demonstration wipe results:
geometric mean window stool lead loadings by room.
-------
100000
M
M
(Jl
10000
I
I
1000
100
10
1-
1.0
I I I I I I I— —I I I I I I I 1 I I I 1 I I I I I [
10.0 100.0 1000.0 10000.0
HUD Demo Window Channel Duet Lead Loading (ug/ft2)
100000
Figure 7-7,
CAP vacuum versus HUD Demonstration wipe results:
geometric mean window channel lead loading by room.
-------
H
£
(D
h->
N;
o\
100000.0
10000.0
1000.0
100.0
10.0
1.0
0.1
"
000
o o *
f *+oV^*+ +0+ +
+ -HOI-
~T~
0.1
Sample Origin:
—I 1—I—I—r-|— —i— —i 1 1 1—i—i—i—|— 1 1
1.0 10.0
HUD Demo XRF (mg/cm2)
+ + + HUD Demo Wipe o o a CAP Study Wipe
ooo CAP Study Vacuum
100.0
Figure 7-8. CAP wipe, vacuum, and HUD Demonstration wipe versus HUD Demonstration ZRF/AAS
results: geometric mean floor lead loading (jig/ft2) by room.
-------
100000.0
10000.0
1000.0
<
M !
(D
H
H
wm
•0
0> L
1
100.0
10.0
++ + +
1^$^^ ++<> + ++ *
Iftf^^" ^ + + o
^ ~*~ + +
+1 1 II 1 III
^TTI
1 • r - T — I— T- ¥ T^ 1 1 1 1 1 1 1 1— | 1 III 1 1 1 F— |
O.1 1.0 10.0 100.O
HUD Demo XRF (mg/cm2)
Sample Origin: + + + HUD Demo Wipe o o o CAP Study Vacuum
Figure 7-9. CAP vacuum and HUD Demostration wipe versus HUD Demostration XRF/AAS results:
geometric mean window stool lead loading (jzg/ft2) by room.
-------
1000000.0
100000.0-
g 10000.0
? 1000.0
M
§ "2 100.0
_J
1 fi 10.0
0?
«Q
(0
M 1.0
to
00
0.1
o
o
o o
0° o° 0
O O rt ^
i||flv+l/+ + +
PS^/^V**** +++
+ +
+ + -H- + + +
0.1 1.0 10.0 100.0
Sample Origin:
HUD Demo XRF (mg/tem2)
HUD Demo Wipe o o o CAP Study Vacuum
Figure 7-10. CAP vacuum and HUD Demonstration wipe versus HUD Demonstration XRF/AAS
results: geometric mean window channel Loading (/ig/ft2) by room.
-------
the XRF/AAS results for both the HUD Demonstration and the CAP
Study projects. The scatter is somewhat more pronounced at lower
paint-lead loadings. Higher dust lead loadings are at times
evident for the CAP dust vacuum samples and again particularly so
for the window channel results.
Figure 7-11 compares the HUD Demonstration and CAP
studies relative to soil lead concentrations (pg/g), collected on
the same side of the unit. The pre-abatement soil samples are
also included as a basis of comparison. The HUD Demonstration
pre- and post-abatement results appear positively correlated.
The CAP soil lead concentrations, in contrast, exhibit a high
degree of scatter. The six measurements in the lower right
quadrant of the figure are from three units (16, 78, and 95) with
low concentrations in the CAP Study and high post-abatement soil
lead levels in the HUD Demonstration. There is some suggestion
of a positive association between the levels of the CAP and HUD
Demonstration when these six measurements are ignored.
In Figure 7-12, soil lead concentrations (pg/g) are
plotted versus the HUD Demonstration XRF/AAS loadings (mg/ft2),
measured for the adjacent exterior wall. The CAP soil samples
and the HUD Demonstration soil results, both pre- and post-
abatement, do not appear to exhibit any trend with increasing
paint-lead loading.
7.3 COMPARISON OF DUST LEAD LOADINGS BETWEEN
THE CAP STUDY AND OTHER STUDIES
It is worthwhile to contrast the CAP Study dust lead
loading results with those from other comparable studies,
including the HUD Demonstration Study. Though considerable
differences exist in the sampling frames, collection procedures,
and instrumental analyses used in each study, the respective lead
loading results may still provide insight on the range of
environmental lead levels which exist in U.S. housing. The
following four field studies were examined:
Volume II - Page 129
-------
10000
3
1000
100
10
10
Sample Origin:
+ £
o o
100 1000
HUD Demo Post-Abatement Soil Lead Concentration (uo/0)
+ + HUD Demo Pr» o o o CAP Study
10000
Figure 7-11. CAP versus HUD Demonstration results: geometric mean foundation
soil lead concentration (M9/9) by side of unit.
-------
4
!
<
i
i
i
Volume II
<5
•0
(D
H
10000
I
' 1000
I
I
I
100
10
+
D
% ~ *
m D v ^Q +
^ ffl o ch D °"1"
B D a a o j£to
$ §S
* 9 D 9 °g °
o * o
S ® o o
o
0.1
1.0 10.0
HUD Demo XRF
Sample Origin:
+ + + HUD Demo Pre a a a HUD Demo Post
ooo GAP Study
100.0
Figure 7-12. CAP soil concentration (/*g/g) and HUD Demonstration soil concentration
(/xg/g) versus HUD Demonstration XRF/AAS results: geometric mean by side
of unit.
-------
HUD Abatement Demonstration Study,
HUD National Survey of Lead-Based Paint,
Traditional versus Modified Practices Study, and
Experimental Abatement Practices Pilot Study.
The dust lead loading results for these studies were either
calculated from available datasets or extracted from reported
results in the scientific literature.
The comparison produced two primary results. First,
the floor and window stool lead loading levels measured in the
CAP Study were lower than those in the other studies. Second,
the CAP Study window channel lead loadings were higher than the
clearance levels measured in the HUD Demonstration and the post-
abatement levels collected in the Experimental Practices Pilot.
Table 7-1 compares the CAP Study floor dust lead
loading results for control and abated units to those measured in
the four studies listed above. For each study, the number of
samples, log standard deviation, geometric mean, and 10th, 25th,
75th, and 90th percentiles are presented. Only the number of
samples and geometric means were available for two of the studies
reported in the literature. Tables 7-2 and 7-3 provide similar
comparisons for window stool and window channel dust lead
loadings, respectively.
The HUD Demonstration intended to eliminate the lead-
based paint hazard from housing environments either by containing
the lead-based paint with encapsulation or enclosure methods, or
by eliminating the lead-based paint with removal methods (HUD,
1991). Because of the diversity of housing components containing
lead-based paint, it was generally true that no single abatement
method could be used uniformly throughout a given housing unit.
The housing units selected for complete abatement included 169
single-family dwellings from the inventory of FHA repossessed
houses in seven urban areas. The clearance (immediately post-
Volume II - Page 132
-------
Table 7-1.
Descriptive Statistics for Floor Dust Lead Loadings (/ig/ft2)
by Abatement Efficacy Field study
Study
CAPS Full
HUD Demo
National
Survey
Kennedy- Pre-Abate.
Kreiger1
Post
Post
(6 months)
Kennedy- Pre-Abate.
Kreiger2 Post
Post (6 m)
Unit Type
Control
Abated
High XRF
Low XRF
Traditional
Modified
Traditional
Modified
Traditional
Modified
Sample
Size
51
187
1026
234
304
280
82
271
50
234
57
70
70
63
Log
St.Dev.
2.12
2.00
1.53
1.82
1.61
na
na
na
na
na
na
na
na
na
P10
1.09
1.69
9.31
0.23
0.08
na
na
na
na
na
na
na
na
na
P25
5.71
6.73
23.55
0.70
0.22
na
na
na
na
na
na
na
na
na
Geom.
Mean
21.38
28.97
66.01
2.40
0.64
250.84
288.00
1440.00
650.32
315.87
315.87
520.26
130.06
55.74
P75
64.99
104.34
185.06
8.23
1.91
na
na
na
na
na
na
na
na
na
P90
289.23
408.58
467.99
24.90
5.08
na
na
na
na
na
na
na
na
na
o
(D
01
vQ
ID
1 Farfel and Chisolm (1990)
2 Farfel and Chisolm (1991)
-------
Table 7-2.
Descriptive statistics for window stool Dust Lead Loadings (/*g/ft2)
by Abatement Efficacy Field Study
Study
CAPS
Full
HUD Demo
National
Survey
Kennedy- Pre-Abate.
Kreiger1
Post
Post
(6 months)
Kennedy- Pre
Kreiger2 Post
Post (6 m)
Unit
Type
Control
Abated
High XRF
Low XRF
Traditional
Modified
Traditional
Modified
Traditional
Modified
Sample
Size
35
78
783
123
126
280
82
271
50
234
57
70
70
63
Log
St.Dev.
1.93
2.18
1.79
2.64
2.13
na
na
na
na
na
na
na
na
na
P10
3.79
7.02
9.03
0.29
0.10
na
na
na
na
na
na
na
na
na
P25
9.85
15.43
26.70
1.42
0.37
na
na
na
na
na
na
na
na
na
Geom.
Mean
46.90
91.57
89.06
8.40
1.57
1337.80
1802.32
3595.35
603.87
1542.19
1635.09
4607.99
325.16
408.77
P75
224.68
467.23
297.09
49.70
6.59
na
na
na
na
na
na
na
na
na
P90
571.47
1315.08
878.56
246.22
24.06
na
na
na
na
na
na
na
na
na
i
(D
0>
vfl
(D
U>
1 Farfel and Chisolm (1990)
2 Farfel and Chisolm (1991)
-------
Table 7-3.
Descriptive Statistics for Window Channel Dust Lead Loadings (fig/ft2)
by Abatement Efficacy Field Study
Study
CAPS
Full
HUD Demo
National
Survey
Kennedy- Pre-Abate.
Kreiger1
Post
Post
(6 months)
Kennedy- Pre-Abate.
Kreiger2 Post
Post (6 m)
Unit
Type
Control
Abated
High XRF
Low XRF
Traditional
Modified
Traditional
Modified
Traditional
Modified
Sample
Size
27
71
756
56
38
280
82
271
50
234
57
70
70
63
Log
St.Dev.
2.02
2.33
1.93
2.28
2.46
na
na
na
na
na
na
na
na
na
P10
84.16
51.74
42.90
11.91
0.73
na
na
na
na
na
na
na
na
na
P25
738.00
510.51
138.10
47.40
3.27
na
na
na
na
na
na
na
na
na
Geom.
Mean
2330.21
2589.90
506.21
220.00
17.18
15496.22
18274.03
14353.52
8082.56
12467.59
24879.43
29422.39
938.32
1003.35
P75
12427.41
18883.56
1855.57
1021.11
90.35
na
na
na
na
na
na
na
na
na
P90
20517.84
39308.26
5973.47
4065.07
402.41
na
na
na
na
na
na
na
na
na
c.
3
(D
0)
iQ
(D
LJ
Ul
1 Farfel and Chisolm (1990)
2 Farfel and Chisolm (1991)
-------
abatement) dust wipe lead loading results from these units were
considered in this instance. The geometric mean floor and window
stool lead loading levels measured in the HUD Demonstration were
higher than those collected in the CAP Study. In contrast, the
geometric mean window channel lead loading results were lower in
the HUD Demonstration than the CAP Study. There are a variety of
possible explanations, including long-term recontamination of the
dust by the external soil lead which was unabated.
The HUD National Survey was conducted to examine on a
national basis the relationships among soil, dust, and paint lead
levels (HUD, 1990). No abatement procedures were performed. In
seeking to represent the pre-1980 housing stock in the U.S., a
total of 381 housing units were sampled: 284 privately-owned
residences and 97 public housing units. The dust vacuum lead
loading results from a subset, 182 units, of the privately-owned
residences sampled were examined in Tables 7-1, 7-2, and 7-3.
The units were partitioned into two groups: the high XRF group
having units with at least one interior and exterior XRF result
exceeding 1.0 mg/cm2, and the low XRF group having units with all
interior and exterior XRF readings less than or equal to 1.0
mg/cm2. There were 102 units in the high XRF group and 80 units
in the low XRF group. The unusually low lead loadings measured
in the National Survey may be misleading, due in part to the
sampling apparatus employed. The so called "blue nozzle" vacuum
sampler used in the National Survey has subsequently been
determined to be approximately 5 times less efficient than the
wipe and vacuum sampling techniques employed in the CAP Study
(MRI, 1991).
The Traditional versus Modified Practices Study was
performed by Kennedy Institute (Farfel and Chisolm, 1990).
Serial dust wipe lead loading measurements were collected from 71
dwellings in Baltimore, Maryland. Samples were collected before,
immediately after, and six months after abatement of lead-based
paint within the dwellings. Local abatement requirements
addressed deteriorated paint on surfaces up to four feet from the
Volume II - Page 136
-------
floor and all paint on easily accessible "biting" surfaces where
lead content of the paint was greater than 0.7 rag/cm2 by XRF or
0.5 percent by weight. Traditional practices involved only
cursory clean-up following the abatement, and allowed a variety
of abatement methods to be used. The modified practices called
for more substantial efforts to clean the dwelling following
abatement, and excluded the use of open-flame burning and sanding
techniques. Most of the study dwellings were low-income row
houses constructed before 1940. The geometric mean floor, window
stool, and window channel dust lead loadings in the CAP Study
were at least an order of magnitude lower than the geometric mean
post-abatement values for both the traditional and modified
practices procedures. The incomplete nature of the traditional
and modified abatement procedures may explain the resulting high
dust lead loadings. Window channels, for example, were not
abated as part of these procedures.
The Experimental Practices Pilot Study was also
performed by Kennedy Institute (Farfel and Chiso1m, 1991). The
experimental practices are described as abatement procedures
which included, (1) treatment of lead-painted surfaces above and
below 4 ft from the floor; (2) sealing and covering of wooden
floors; (3) procedures for containment of dust during abatement;
and (4) a final cleanup using a high-efficiency particle air
(HEPA) vacuum. Dust wipe lead loading samples were collected in
six two-story, six-room low income row houses constructed in the
1920's. Measurements were taken before, immediately following,
and six months after the abatement procedures occurred. The CAP
Study geometric mean lead loading levels measured on floors and
window stools were lower than those measured following the
experimental abatement procedures. Interestingly, the geometric
mean window channel lead loadings were higher in the CAP Study
than the Experimental Practices Pilot. The CAP Study samples were
collected more than six months post-abatement, so perhaps
recontamination occurs over longer periods of time.
Volume II - Page 137
-------
8.0 OUTLIER ANALYSIS
In this section, the outlier analysis is discussed.
First, we provide the general approach to the analysis, and then
we provide details on how the data were grouped, the outlier
analysis procedure used, and how the outliers found were handled
in our statistical analysis. Data from Unit 08 (which were
excluded from the full statistical analysis) were included in
this outlier analysis.
8.1 APPROACH
Formal statistical outlier tests were performed on the
field sample data and the laboratory QC sample data. Data were
placed into groups of comparable values, and a maximum absolute
studentized residual procedure was used to identify potential
outliers. When a potential outlier was identified, that value
was excluded from the group, and the outlier test was performed
again. This procedure was repeated until no additional outliers
were detected. After all potential outliers were identified, a
list of these samples was sent to the laboratory for rechecking.
The following sections further explain this procedure.
8.2 DATA GROUPS
Samples collected from inside the houses were grouped
according to the predominant interior abatement method, sample
medium (cassette or wipe) and component (air duct, floor, window
channel, field blank, trip blank, etc.). Soil samples and
exterior entryway cassette samples were grouped according to the
predominant exterior abatement method. In addition, interior
floor samples were split into two groups, those taken from
carpeted floors and those taken from uncarpeted floors. Separate
outlier analyses were then performed for each group on the
natural logarithm of lead loading values, the natural logarithm
of lead concentration values, sample concentration values (field
blanks only) and net weight values (trip blanks only).
Volume II - Page 138
-------
Normally, soil samples were collected along the
foundation of each house. In one case, however, pavement along
the foundation required the use of a vacuum cassette to collect
two dust samples rather than the usual two soil samples.
Additional outlier tests were performed (1) grouping these two
samples with foundation soil samples and (2) grouping these two
samples with exterior entryway cassette samples.
Laboratory QC values were grouped according to type of
sample and sample medium. Outlier analyses were than performed
on the natural logarithms of the appropriate measurement for each
type of sample (spike recovery for spiked samples, amount of lead
for method blanks, calibration blanks, and unspiked samples,
percent recovery for interference check samples, calibration
standards, calibration verification samples and blind reference
material samples, and range of spike recovery for duplicate
spiked samples).
8.3 THE OUTLIER TEST
The SAS procedure GLM (SAS PC, ver. 6.04) was used to
compute the studentized residual for each data value in a group
by fitting a "constant" model (i.e., mean value plus error term)
to the log-transformed data in each group. The absolute values
of the studentized residuals were then compared to the upper
.10/n guantile of a t distribution with n-2 degrees of freedom,
where n is the number of data values in the group. If the
maximum absolute studentized residual was greater than or equal
to the .10/n quantile, the corresponding data value was flagged
as a potential outlier. The outlier test was then repeated,
excluding additional potential outliers, until no more outliers
were detected. Table 8-1 lists the field sample outliers found
as a result of this test. Table 8-2 lists the laboratory QC
outliers.
Often, the minimum and/or maximum data values in a
group were flagged as outliers by the test described above. If
(ft
the minimum and maximum values in a group were not flagged, they
Volume II - Page 139
-------
Table 8-1. CAP Study Outliers - Field Samples
Lead Loading Outliers
Instrument
Batch
E04292A
EOS072B
EOS072B
EOS132A
E06022A
E07272A
E07272A
E08032A
E08032A
Sample
Preparation
Batch
WIO
WIR
WJB
WJC
WIG
WIZ
WIZ
WKF
WKG
MRIID
902924
903347
903556
903116
902546
903392
903769
905079
905143
Sample
Medium
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Wipe
Dust-Wipe
Study ID/
Sample ID
28/01
96/02
19/01
96/01
45/07
19/02
21/25
21/26
57/27
Location
Kitchen
Hall
Living Room
Hall
Kitchen
Living Room
Laundry Room
Laundry Room
Bathroom #2
Component
Floor
Floor
Floor
Floor
Floor
Floor
Floor
Floor
Floor
Lead
Loading
(ug/ft2)
< 0.34
2365.43
1102.35
11641.25
1765.38
6745.20
7046.70
333.56
< 2.72
Lead Concentration Outliers
Instrument
Batch
E04272A
E04292A
E04292A
EOS072B
E05072B
E05072B
E05072B
EOS122B
EOS122B
EOS132A
EOS192A
EOS262A
E06022A
E06042A
E06112A
E06122A
E06152A
E06292A
E06292A
E06292A
E07212A
E07212A
E08242A
E08242A
Sample
Preparation
Batch
WIL
WIL
WIO
WIR
WIR
WJD
WJD
WJE
WJF
WJC
WIQ
WIT
WJG
WJP
WIW
WJR
WJV
WKB
WKB
WKB
WJG
WJR
WJA
WJX
MRIID
902564
902761
903673
902605
903347
902142
903487
902126
902220
903116
904271
904054
902546
902380
904433
903291
903089
902955
903020
903163
902953
902169
904397
902275
Sample
Medium
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Soil
Soil
Dust-Vacuum
Dust-Vacuum
Soil
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Soil
Dust-Vacuum
Study ID/
Sample ID
17/13
94/12
46/05
79/12
96/02
49/02
60/01
79/14
51/02
96/01
81/17
79/16
45/07
68/10
51/18
72/11
68/12
80/11
03/04
31/07
51/01
19/12
53/19
10/12
Location
Front
Hall
Bathroom
Kitchen
Hall
Kitchen
Bedroom H\
Back
Bathroom
Hall
Back
Back
Kitchen
Dining Room
Back
Hall
Kitchen
Living Room
Bathroom
Bathroom #2
Bathroom
Kitchen
Left
Kitchen
Component
Outside Entryway
Inside Entryway
Air Duct
Inside Entryway
Floor
Floor
Floor
Outside Entryway
Floor
Floor
Foundation
Entryway
Floor
Air Duct
Foundation
Inside Entryway
Inside Entryway
Inside Entryway
Window Stool
Floor
Floor
Inside Entryway
Boundary
Inside^Entryway
Lead
Concentration
(ug/g)
8.84
21.67
462343
2723.16
1724 32
< 4.56
< 11.00
16335.45
13567.76
6217.62
3351.12
< 4.55
639860
5644.54
< 5.491
965
1200.39
533200
48271 93
1.71
1218630
2293.62
1074.242
9.24
Volume II - Page 140
-------
Table 8-1. CAP Study Outliers - Field Samples (Contd)
"Amount of Lead" Outliers
Instrument
Batch
E04292A
E05272A
E06112A
E06152A
E08032A
E08242A
Sample
Preparation
Batch
WIO
WIV
WIW
WJU
WKG
WIT
MR! ID
902825
904161
904333
903654
905133
904183
Sample
Medium
Dust-Vacuum
Soil
Soil
Dust-Vacuum
Dust-Wipe
Soil
Study ID/
Sample ID
18/06
70/22
50/22
07/06
94/28
99/22
Location
Kitchen
Front
Right
Living Room
Kitchen
Front
Component
Field Blank
Field Blank
Field Blank
Field Blank
Field Blank
Field Blank
Amount
of Lead
(ug/sample)
< 0.344
35.638
271.6253
2.682
35.445
< 1.197
Net Weight (Trip Blank) Outliers
Instrument
Batch
TRIPBLNK
TRIPBLNK
TRIPBLNK
TRIPBLNK
TRIPBLNK
TRIPBLNK
MRIID
902217
902516
902964
903144
903146
903722
Sample
Medium
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Dust-Vacuum
Study ID/
Sample ID
19/23
90/23
40/23
07/23
65/23
55/23
Location
Bedroom #1
In Van
Living Room
Living Room
Living Room
Living Room
Component
Trip Blank
Trip Blank
Trip Blank
Trip Blank
Trip Blank
Trip Blank
Sample
Weight
(g)
-0.0052
0.0051
0.0002
• 0.0007
0.0009
0.0015
'Value subsequently corrected to 271.625 jtg/g - no longer an outlier.
tyalue subsequently corrected to 1072.76 /ig/g - still an outlier.
'Value subsequently corrected to <5.49 - no longer an outlier.
Volume II - Page 141
-------
Table 8-2. CAP study Outliers - Laboratory QC samples
Spike Recovery Outliers
Instrument
Batch
E04272A
E04272A
E05042A
E05042A
E05072B
E05072B
E05072B
E05072B
E05122B
E05122B
E05192A
E05272A
E05272A
E06042A
E06042A
E07142A
E07212A
E07272A
E07272A
Inatrument
Batch
E07272A
E07272A
E07272A
E07272A
Sample
Preparation
Batch
NIL
WIL
WIR
WIR
WJB
WJB
WJD
WJD
HJE
HJE
WIP
HJO
HJO
WJP
HJP
WKF
WJC
WKJ
WKJ
Sample
Preparation
Batch
WIZ
WIZ
WKJ
WKJ
Sample
ID
903695
903701
903551
903555
903604
903597
903584
903753
903454
903484
904266SPD
903360
903628
903320
903321
905240
903546
903303
903079
Method Blank
Sample
ID
MB1
MB2
MB1
HB2
Run
Number
102
104
31
33
34
42
116
118
110
112
33
115
116
29
30
45
234
148
149
Outliers
Run
Number
38
39
116
142
Sample .
Type
Flag
2
3
2
3
2
3
2
3
2
3
3
2
3
2
3
2
3
2
3
Sample
Type
Flag
4
4
4
4
Spike
%
Recovery
128.5
134.0
104.1
104.0
101.5
101.5
97.8
97.9
101.2
101.2
130.9
_ 98.5
98.4
100.6
100.3
99.2
113.7
108.5
109.0
Amount of
Lead
(pg/ sample)
<4.0202
<4.0202
4.0380
20.6810
Volume II - Page 142
-------
Table 8-2. CAP Study Outliers - Laboratory QC Samples (Contd)
Reference Material Recovery Outliers
Instrument
Batch
E06292A
E07302A
E08212A
E08212A
Sample
Preparation
Batch
WZX
WKJ
WKJ
WIZ
• Sample
ID
904326
902699
902699
902731
Run
Number
181
156
28
29
Sample
Type
Flag
5
5
5
5
Reference
Material
% Recovery
114.8
34.4
22.9
27.0
Continuing Calibration Blank Outliers
Instrument
Batch
E05152A
E05152A
E08182A
Sample
Preparation
Batch
WIK
WIK
REF
Sample
ID
CCB
CCB
CCB
Run
Number
44
93
55
Sample
Type
Flag
9
9
9
Amount
of Lead
(jjg/ml)
0.0130
0.0111
0.0004
Volume II - Page 143
-------
were nevertheless included in Tables 8-1 and 8-2 as being
potential outliers. Of the 838 lead loading values reported,
nine (1%) are listed as potential outliers. This includes 7 out
of 770 cassette samples and 2 out of 68 wipe samples. Of the
1124 lead concentrations reported, 24 (2%) are listed as
potential outliers. This includes 20 out of 770 cassette samples
and 4 out of 354 soil samples. Of the 139 field blanks, six (4%)
are listed as potential outliers, and of the 53 trip blanks, six
(11%) are listed as potential outliers.
8.4 RESOLUTION OF OUTLIER QUESTIONS
Tables 8-1 and 8-2 were sent to the laboratory for
review. This review resulted in corrections to three of the
identified field sample outliers (as indicated in footnotes to
Table 8-1) and two other values which had not been identified as
outliers. No errors were found in the reporting of the
laboratory QC sample data.
Volume II - Page 144
-------
9.0 STATISTICAL ANALYSIS OF QUALITY CONTROL DATA
To assure that the sampling and analytical protocols
employed in the CAP Study were producing data of sufficient
quality, a number of different guality control (QC) samples were
included in the study design. The intended purpose of each QC
sample varied, but each sample type belonged to one of three
categories:
1. Field QC-Samples, originating in the field,
intended to assess the quality of the sample
collection procedures;
2. Sample Preparation QC-Samples, originating in the
sample preparation laboratory, which examine the
preparation of field samples for analysis, and;
3. Instrumental Analysis QC-Samples, produced in the
instrument analysis laboratory, that evaluate the
quantitative analysis of the samples.
These individual categories reflect distinct goals of the QC
analysis, and separate steps in the collection and analysis of a
sample. From a statistical analysis perspective, however, the QC
samples may be partitioned somewhat differently. This
partitioning reflects the nature of the parameter considered when
assessing a particular QC measure. Specifically, the QC samples
are partitioned analytically into three groups: (1) blank
samples, (2) recovery samples, and (3) duplicate samples. Table
9-1 below is helpful in considering these two approaches to
categorizing the QC results. Each type of QC sample employed in
the CAP Study is identified within a particular cell of the
table. For example, spiked samples were analyzed as recovery
samples, but their results address the quality of the sample
preparation procedures. A total of ten QC measures were
employed. Detailed results of the statistical analyses performed
on these QC measures are reported in the sections that follow.
To facilitate understanding of the analytical results, our report
sections are organized by analysis category. Within each
Volume II - Page 145
-------
Table 9-1. QC sample Categorization Matrix
Blank
Samples
Recovery
Samples
Duplicate
Samples
Field
QC
• trip blanks
• field blanks
• side-by-sides
Sample Preparation
QC
- method blanks
• spikes
• blind reference
materials
• spiked duplicates
Instrument Analysis
QC
• calibration
blanks
• interferant
check standards
• calibration
verifications
category, however, the implications of the results to each
procedure step are discussed.
As an overall summary, the following conclusions may be
drawn regarding the QC samples:
Analysis of the blank samples suggests little if
any procedural contamination. The majority of
blanks were measured to have lead content below
the instrumental level of detection.
Despite some procedural problems in their creation
and analysis, the results for the recovery samples
indicate very good method performance.
Spiked duplicate samples created in the laboratory
exhibited very good agreement. Side-by-side field
samples, on the other hand, suggest significant
variability in field sampling. Greater inherent
variation was seen in dust samples than in soil
samples.
There is no significant evidence of a time-based
trend in any of the QC samples.
9.1 BLANK SAMPLES
Blank samples are expected, by the nature of their
collection or preparation, to contain no lead or only a very
small amount of lead. In the CAP Study, four types of blank
samples were analyzed: trip blanks, field blanks, method blanks,
and calibration blanks. The parameter of interest was the amount
Volume II - Page 146
-------
of lead (ng) measured for the sample (lead content). In the
case of trip and field blanks, the net weight (g) of the sample
was also examined. Evidence of a significant amount of lead in a
blank sample would suggest a bias in the results for the regular
field samples. As was the case for the regular field data, the
lead content of the blanks was assumed to follow a lognormal
distribution. The amounts, therefore, were log-transformed
during their analysis.
9.1.1 Field Quality Control
Trip blanks are vacuum dust cassettes that are weighed
in the gravimetric laboratory before and after being transported
to the field. They are similar to field blanks, except they are
not exposed to the field environment. Trip blanks provide
information on the sample weight variability resulting from
gravimetric laboratory activities in the absence of field
handling. Used in combination with the field blank net weight
data, they provide a means of determining the weight error
contribution from the gravimetric laboratory should the net
weight data from the field blanks show an unusual result.
Accordingly, no lead analysis was performed on trip blanks. One
trip blank was generated for each housing unit sampled by
selecting, at random, one vacuum dust cassette from all unused
vacuum dust cassettes transported to the field.
Descriptive statistics for the net weights measured for
both trip and field blanks from the CAP Pilot and CAP Studies are
presented in Table 9-2. The number of samples, arithmetic mean,
standard deviation, minimum and maximum net weights are
presented. Net weight data from trip blanks indicate that
gravimetric laboratory processing resulted in a net mean weight
gain of 3.5 mg. This gain is about twice as large as that
observed during execution of the pilot study (net mean weight
gain of 1.8 mg). The weight difference between the CAP Study and
CAP Pilot Study can be attributed, in part, to protocol changes
made in gravimetric processing. The clearance criterion for the
Volume II - Page 147
-------
Table 9-2. Net Weight Results for Trip and Field Blanks
Statistic
Number of
Samples
Net Weight Mean
(ing)
Net Weight
Standard
Deviation (ing)
Minimum Net
Weight (mg)
Maximum Net
Weight (mg)
CAP Pilot Study
Trip
Blanks
54
1.8
0.3
1.1
2.6
Field
Blanks
9
2.4
0.5
1.4
3.0
CAP Study
Trip
Blanks
51*
3.5
1.2
0.2
5.1
Field
Blanks
52
0.4
3.0
-6.3
5.2
* Excluding one sample identified as an outlier.
determination of cassette stability was increased from ± 1 mg to
± 2 ing. This change was made to reduce the excessive
equilibration time required during the pilot study. It was
anticipated that the resulting losses in accuracy at low sample
weights would be offset by the increased collection efficiency of
the sampling system used for dust sample collection. Indeed, the
summary in Table 2-2 of the amount of dust collected suggests
that the lack of sufficient sample was not usually a problem.
The mean weights of the collected dust were sufficiently large to
override the mean weight gain resulting from gravimetric
laboratory processing.
Field blanks are identical to regular field samples,
except that no sample is actually collected. Field blanks
provide information on the extent of lead contamination
experienced by field samples resulting from a combination of
laboratory processing and field handling. In addition, field
blanks for cassettes provide information on the sample weight
variability resulting from the combination of gravimetric
Volume II - Page 148
-------
laboratory activities and field handling. Field blanks for
vacuum dust, wipe dust (abated houses only), and core soil were
collected for each housing unit.
Field blanks, as opposed to trip blanks, better
represent the handling experienced by field samples. Any
adjustments to weight data, if required, are best based on field
blank net weight data. As shown in Table 2-2, the mean weights
of collected dust for field samples are considerably larger than
the net mean weight of 0.4 mg measured for the field blanks as
shown in Table 9-2. No adjustments were made, therefore, to
field sample weights of vacuum dust cassettes for the calculation
of lead concentration (ng/g) values.
Mean net weights between the trip and field blanks for
the CAP Pilot (1.8 mg compared to 2.4 mg) were relatively close
as indicated in Table 9-2. Mean net weights between the trip and
field blanks for the CAP Study, however, differ more considerably
(3.5 mg compared to 0.4 mg). The CAP Study data imply that field
handling produces weight reduction of the vacuum dust cassettes.
The change between the CAP Pilot and CAP Study data is suspected
to be related to a combination of two factors: the protocol
changes made in gravimetric processing (discussed earlier), and
the humidity at the sampling site.
Handling of field blanks exposes the cassettes to the
atmosphere at the field site. The procedure of collecting field
blanks included the following steps: remove the cassette from
the sealed plastic bags, open the cassette casing, insert it into
the cyclone sampler, remove it from the sampler, close the
cassette casing, and replace the cassette into the sealed plastic
bags used for transport. Trip blanks were not removed from their
sealed plastic bags in the field. The collection site was in an
area known for low humidity (Denver has a dry climate). When
opened in a low humidity environment, field blanks would be
expected to lose water (and weight) absorbed during equilibration
in the gravimetric laboratory. It is suspected that the change
in gravimetric clearance criterion did not permit sufficient
Volume II - Page 149
-------
equilibration in the gravimetric laboratory to allow the
cassettes to gain back all the weight lost during their exposure
to the low humidity field environment. This would account for
the observed net weight difference between the field and trip
blanks. Gravimetric records were reviewed for data to support
this supposition. However, no weights were recorded for the
first 72 hours after vacuum dust cassettes were placed into the
gravimetric laboratory (standard equilibration) and there exist
no field humidity data. There are insufficient data available,
as a result, to either discount or support the protocol change
and humidity effect explanation.
Field blank samples also were measured for lead
content. A summary of the field blank lead content results (and
in fact, of all the QC results) is presented in Table 9-3. The
descriptive statistics reported include the number of samples,
number above the instrumental detection limit (IDL), minimum and
maximum. When possible, the geometric mean and logarithmic
standard deviation for the amount of lead per sample are
presented. A 95% upper confidence bound on the 95th percentile
for lead content is also provided. For the sake of simplicity,
this bound will be referred to as the estimated 95% tolerance
bound. These calculations were possible only when a sufficient
number of results were above the IDL.
If all results were above the IDL, calculation of the
geometric mean and logarithmic standard deviation was routine,
and the estimated 95% tolerance bound was determined using an
exact procedure for lognormal distributions. In cases where a
portion of the results were below the IDL, statistical procedures
which recognize these data as censored values were used to
estimate the geometric mean and logarithmic standard deviation.
A lognormal model was fitted to the data and its parameters
estimated. The SAS procedure LIFEREG was utilized in obtaining
these estimates. LIFEREG maximizes the log-likelihood function
via a ridge stabilized Newton-Raphson algorithm, thereby
providing maximum likelihood estimates of the log mean and log
Volume II - Page 150
-------
Table 9-3. Results of Quality Control Measures Analyses
Quality Control
Measure
Field Vacuun
Blanks wipe
Soil
Method Vacuun
Blanks wipe
Soil
Calibration Blanks
Blind I
References 1 1
ICS
Calibration Verifications
Spikes Vacuun
Wipe
Soil
Spi ked Vacuun
Duplicates Wipe
Soil
Side-by-Sides Vacuun
Vacuun
Soil
Parameter
Considered
Amount
Amount
Amount
X Recovery
X Recovery
X Recovery
X Recovery
Ratio
Ratio (fig/ft2)
Ratio (jtg/g)
#of
Samples
52 (6)
34 (1)
51 (4)
48 (13)
6 (1)
22 (1)
431 (33)
75
37
144
274
96
12
44
48
6
22
52
52
51
Minimum
0.344
2.723
1.198
0.468
2.723
1.276
0.0004
0.344
0.229
0.997
0.962
0.930
0.862
0.733
1.000
1.001
1.001
1.027
1.022
1.004
Maximum
2.682
35.445
35.638
20.681
3.975
3.297
0.068
1.749
1.131
1.2112
1.058
1.428
1.000
1.309
1.094
1.151
1.308
40.381
81.101
4.569
Geometric
Mean
0.228
na
0.067
0.414
na
na
0.007
1.061
0.881
1.060
1.014
1.030
0.926
0.981
1.031
1.063
1.081
2.334
2.071
1.296
Log Standard
Deviation
1.059
na
2.387
1.135
na
na
0.956
0.206
0.316
0.035
0.016
0.068
0.044
0.098
0.039
0.080
0.109
1.110
1.129
0.399
Lower 95X
Tolerance
Bound
0.706
0.396
0.993
0.986
0.904
0.820
0.799
Upper 95X
Tolerance
Bound
2.006
na
9.162
4.369
na
na
0.041
1.594
1.957
1.131
1.043
.174
.044
.205
.068
.238
.227
6.403
6.605
1.951
i
(D
H
I
•a
(D
Censored Analysis
^ The number of samples measured above the instrumental detection limit (IDL) is enclosed in parentheses.
This value represents an extra ICS analyzed in the middle of an analysis run from an instrument analysis batch containing no field samples. This
batch contained only re-runs of SRM 1646 under the conditions described in Section 9.2.1. The next highest ICS, 1.182, was also measured in the
same analysis batch.
na - The statistic could not be calculated due to the large number of censored samples.
-------
standard deviation. Further, an approximate procedure was -used
to calculate the estimated 95% tolerance bound. The
"approximate" nature of this statistical procedure was in
employing the 'censor* estimates for log mean and log standard
deviation in calculating a traditional 95% tolerance bound.
Since this procedure did not include an adjustment to the bounds
reflecting censored data, the estimated tolerance bound is
approximate.
The data for field blank samples are illustrated in
Figure 9-1. The amount of lead (jig) found in each blank sample
is plotted by sample type. Different plotting symbols are used
to indicate whether the result was above the IDL or below, in
which case the detection limit is plotted. In those instances
where an estimated tolerance could be calculated, the estimated
95% tolerance bound is illustrated in the figure by a bar which
has the bound as its upper value.
Most of the field blanks generated for each sample type
were below the IDL: more than 88% of the vacuum dust samples
were, as well as more than 97% of the wipe dust samples, and more
than 92% of the soil samples. No field blank result exceeded
five times the average IDL measured during the analysis
activities (0.037 pg of lead per mL) . Geometric means for all
three sample types are less than this IDL mean. These data
suggest that no lead contamination occurred during field sample
activities.
9.1.2 Sample Prep Quality Control
Method blanks are blank samples generated in the
laboratory during sample preparation activities. They are
processed in a manner identical to field samples except that no
sample material or sample medium is present in the container used
for sample digestion. Method blanks provide information on the
potential lead contamination experienced by field samples
resulting solely from laboratory processing. Method blanks were
Volume II - Page 152
-------
0s
ro
ro
I
1OO.OOOO
10.0000
1.0000
0.1000
0.0100
O.O010
0.0001
8
w
8
Field Blanks
ooo Censored
OB
W
Method Blanks
+ + + Non Censored
Figure 9-1. Individual measurements and tolerance bounds for
Mg lead/sample in blank samples.
-------
generated at a frequency of two samples per batch of
approximately 40 field samples.
A summary of the method blank results is presented in
Table 9-3 and presented graphically in Figure 9-1. These results
were obtained using the same procedures outlined for field
blanks. All method blank data met the data quality objective of
lead levels less than 10 times the IDL. Most of the method
blanks generated for each sample type were below the IDL: 72% of
the vacuum dust samples, 83% of the wipe dust samples, and 95% of
the soil samples. In fact, a geometric mean, log standard
deviation, and approximate 95% tolerance bound could only be
calculated for the vacuum cassettes. Only one method blank
result exceeded five times the average IDL measured during the
analysis activities (0.037 jig of lead per mL). This method blank
was one of two in a sample preparation batch which contained only
high sample weight vacuum dust samples (minimum field sample
weight of 4 grams each). This method blank, with a measured lead
level near six times the instrumental detection limit, was
insignificant with respect to the lead levels within the batch.
The other method blank in this high sample weight batch was less
than the IDL. These data indicate no lead contamination occurred
during laboratory processing of field samples.
9.1.3 Instrumental Analysis Quality Control
Calibration blanks were analyzed along with field
samples to assure adequate instrument performance during lead
determinations. They are useful in assessing any changes in
instrument performance which may affect the estimated lead
concentrations reported for regular field samples. Descriptive
statistics summarizing the results for calibration blanks are
presented in Table 9-3. The individual results and their
approximate 95% tolerance bound are portrayed in Figure 9-1. As
with the field blank results, the geometric mean, log standard
deviation, and approximate 95% tolerance bound are adjusted to
reflect the censored nature of many of the results. Greater than
Volume II - Page 154
-------
92% of the calibration blanks, which included both initial and
continuing calibration blanks, were below the IDL. The maximum
lead concentration measured for any calibration blank was less
than two times the average IDL for all instrumental analysis runs
(0.037 ng of lead per mL). Their geometric mean was well below
the average IDL. These results suggest that the field sample
results appear to be free from any significant bias caused by
carryover.
9.2 RECOVERY SAMPLES
Recovery samples are expected to contain a known amount
of lead or to have had a known amount of lead added (spiked).
Four types of recovery samples were incorporated into the design
of the CAP Study: blind reference material samples, spiked
samples, calibration verification samples, and interferant check
standards (ICS). The parameter of interest was the ratio of the
amount of lead measured for the sample (lead content) to the
known amount of lead in the sample. This ratio should be
approximately one, and when multiplied by 100 is commonly
referred to as the percent recovery. Percent recovery values
over 100% indicate a measured value exceeding the known amount of
lead in the sample and values under 100% indicate a measured
value below the known amount. Spiked soil samples were slightly
different in that the spike was added to a sample already
containing a measured amount of lead. The percent recovery value
is assumed to follow a lognormal distribution. If the geometric
mean of the lognormal distribution is 100%, this is an indication
that lead is over-recovered half the time and under-recovered
half the time.
9.2.1 Sample Preparation Quality Control
Spiked samples were blank samples (or regular field
soil samples) fortified with known levels of lead prior to sample
preparation activities, and were processed in a manner identical
Volume II - Page 155
-------
to field samples. They provided lead recovery information for
use as an assessment of the accuracy and precision of field
sample data through sample preparation and analysis activities.
Spiked samples were generated at a frequency of four (two spikes
and two spiked duplicates) per batch of approximately 40 field
samples.
As is noted earlier, spiked soil samples were prepared
and analyzed somewhat differently compared to vacuum and wipe
dust spikes. Whereas spiked cassette and wipe samples involved
spiking a known amount of lead into a blank, spiked soil samples
were created by spiking a regular soil sample with a known amount
of lead. For cassette and wipe spikes, the ratio of measured
amount to known spiking amount was considered (percent recovery).
Since a soil spike sample already contained some lead, a
different calculation of percent recovery was required.
Specifically, the spiked soil percent recovery was determined as,
f measured \ig lead] _ \ measured \ng lead
[for spiked samplel \for unspiked sample
\ig lead for spike
Use of QC sample data as an assessment of the accuracy
and precision achieved for field samples is partially dependent
on the chemical constituent matching (matrix matching) between
the QC sample and field sample. This is because data generated
from a given analytical processing scheme are generally matrix
sensitive. In the case of soil samples, the matrix matching was
very good, because unspiked and spiked samples were generated
from splits of homogenized soil samples from field sample
locations. Spiked sample data for soils, therefore, were
expected to closely mimic that of the field samples. Blank
cassettes and wipes, however, were used for the generation of the
unspiked and spiked samples for dust. Dust field samples (vacuum
and wipe) could not be split in a homogeneous manner. As a
result, the spiked sample QC data for dust samples may be less
Volume II - Page 156
-------
useful than the spiked sample QC data generated for soils.
Still, the spiked sample QC data do provide an adequate measure
of the degree of successful execution of the analytical
methodology. The methodology (sample preparation and analysis)
is procedurally very similar to methods commonly used and
verified successfully for many different types of environmental
samples. The spiked sample QC data for dust samples generated
during this project are still useful in estimating of precision
and accuracy for field samples.
A summary of the spiked sample results is presented in
Table 9-3. Descriptive statistics presented include the number
of samples, minimum, maximum, geometric mean, and log standard
deviation. In addition, an estimated central 90% tolerance
interval was calculated using an exact procedure for lognormal
data. This interval was derived from a 95% upper confidence
bound on the 95th percentile and a 95% lower confidence bound on
the 5th percentile. Performance-Control charts showing
individual spiked sample recovery data are shown for each sample
type in Figures D-l, D-2, and D-3 of Appendix D.
The data for all recovery samples, including the spiked
samples, are illustrated in Figure 9-2. The individual percent
recovery results for each type of recovery sample are plotted.
The estimated central 90% tolerance interval is presented in the
figure by a bar extending from the lower confidence bound on the
5th percentile to the upper confidence bound on the 95th
percentile.
Spiked sample recoveries for all but four data points
met the data quality objectives of accuracy of ±30% from the true
spiked value. Three of these four points were the result of a
spiking error (samples were spiked 10 times less than planned).
This error produced measurements approaching both the IDL and
background lead levels detected in blank cassettes used in the
generation of the spiked samples. Accurate determination of
spike recoveries under such conditions is difficult and is not
Volume II - Page 157
-------
1000
o
(D
M
"0
0>
100
i
I
i
(J\
co
10
I
II
Blind Reference
ICS
cv
w
8
Spiked
Figure 9-2. Individual measurements and tolerance bounds for
percent recovery in recovery samples.
-------
anticipated to be reflective of performance related to field
samples. The other data point (soil sample) was only slightly
outside the data quality objectives (130.9%). Geometric means
for all three sample types are within ±10% of the true spiked
amount. The estimated tolerance intervals for all three media
contain 100% or complete recovery. These data infer that
accuracy for field samples was good and well within data quality
objectives.
Blind reference material samples were generated by
placing known quantities of NIST standard certified reference
materials (SRMs) into blank samples and inserting them into the
sample batches in a blind manner prior to sample preparation
activities. These reference materials were processed by the
laboratory in the same way as the field samples. Their results
provide lead recovery information that can be used as an
assessment of accuracy of field sample data as determined by
sample preparation and analysis activities. The blind nature of
the insertion into the sample processing stream helped provide QC
data unbiased by laboratory activities. Blind reference
materials were generated at a frequency of two (one each of two
different materials) per batch of approximately 40 field samples.
As was discussed for the spiked QC samples earlier,
matrix matching is an important determinant of the usefulness of
QC samples in assessing the accuracy achieved for regular field
samples. In general, reference materials are included in an
analysis scheme to help provide higher confidence in the accuracy
of field sample data than can be obtained using only spiked
samples. Unfortunately when this study initiated, no suitable
dust or soil SRMs were available. Two SRMs were chosen as the
best readily available approximations to matrix matching the
field samples (matching with respect to general matrix components
and anticipated lead levels). These were NIST SRM No. 2704
Buffalo River Sediment and NIST SRM No. 1646 Estuarine Sediment.
Given the limitations of the matrix match, some caution is
appropriate in extending the accuracy results of these reference
Volume II - Page 159
-------
materials. These data, combined with the spiked results, still
do provide reasonable confidence that analytical methodologies
were carried out as planned.
Performance-Control charts, showing the percent
recovery of lead from the two blind reference materials, are
shown for each sample type in Figures D-4, D-5, and D-6. Blind
reference material recoveries for NIST SRM No. 2704 met the data
quality objectives for accuracy of ±30% from the true spiked
value. Recoveries for NIST SRM No. 1646, however, were sporadic.
Eight of 37 data points were outside data quality objectives.
Investigation into these recovery problems suggested they were
related to corrections for spectral interferences during instru-
mental analysis measurements. SRM No. 1646 has a low lead
concentration (28.2 jug/g) combined with high levels of other
metals such as iron. The iron-to-lead ratio is over 1000 to 1.
In order to correct for potential iron interferences, the analyst
conducting the instrumental measurement efforts must perform
serial dilution of all digests to get iron levels within the
calibration range of the ICP instrument. For field samples,
extra dilutions were rarely needed (indicating limits in the
ability of SRM No. 1646 to mimic field sample matrices). For the
blind SRM No. 1646 reference materials, extra dilution was always
required. This extra dilution pushed the measurable lead level
down to within a few multiples of the instrumental detection
limit where measurement variance increases relative to higher
lead level digests. (A detailed discussion of the affect of
decreasing lead levels and measurement variance is presented in
Volume III under laboratory experiences.) The result of the extra
dilutions were the sporadically poor recoveries seen for SRM No.
1646.
The dilution theory for explaining sporadic recoveries
for SRM No. 1646 was verified by reanalyzing the original digests
using the ICP-AES reconfigured to extend the linear range of the
instrument for detecting iron. The extra dilution requirement
was avoided. The results of the measurements are plotted as the
Volume II - Page 160
-------
DF=1 data points in the Performance-Control charts shown. Using
the reconfigured instrument, all but two blind reference material
recoveries for NIST SRM No. 1646 met the data quality objectives
of accuracy of ±30% from the true spiked value. The remaining
two points were associated with extra high weight sample batches
that required a sample preparation protocol change. The change
resulted in a four fold increase in final digestion volume. The
increase, in turn, reduced lead levels beyond that which could be
made-up by elimination of any extra dilution volumes.
Blind reference material results, shown in Table 9-3
(identified as group I), reflect summarized data from the
originally analyzed SRM No. 1646. The results of the reanalysis
of SRM No. 1646 at no extra dilution volumes (data points plotted
in the figures as DF=1) are identified as group II. These
results are illustrated in Figure 9-2. Both geometric means were
within ±12% of the NIST certified value. The estimated central
90% tolerance intervals both contain 100% recovery. Even with
the matrix match limitations for these SRMs, these data infer
that accuracy for field samples was good and well within data
quality objectives.
9.2.2 Tnafcrumental Analysis Quality Control
Calibration verification samples were analyzed along
with field samples during instrumental measurement activities to
verify calibration standard levels and monitor drift of
instrument response. A summary of lead results for calibration
verification samples is shown in Table 9-3 and Figure 9-2. These
statistics are calculated using the same procedures described for
spiked samples. All calibration verification results met design
specifications. In addition, the estimated central 90% tolerance
interval is narrow and contains 100%. It seems reasonable to
conclude that the field samples results are free from any
significant bias caused by instrumental drift.
Interference check standards (ICS) were used to verify
accurate analyte response in the presence of possible spectral
Volume II - Page 161
-------
interferences from other analytes present in the sample. A
summary of lead results for ICS is available in Table 9-3 and
Figure 9-2. As with the calibration verifications, the estimated
central 90% tolerance interval is remarkably narrow and contains
100%. There is no evidence of any significant bias in the
regular field sample results caused by commonly encountered
interferences.
9.3 DUPLICATE SAMPLES
Duplicate samples are expected to be have similar lead
content either because they were collected side-by-side in the
field or because they were created to be comparable in the
laboratory. In both cases, such samples are analyzed one after
the other in the same analytical batch. The analytical result of
interest for each pair of duplicate samples was the ratio of the
larger measured lead result to the smaller measured lead result.
This ratio has a minimum value of one. The log of this ratio was
assumed to follow the absolute value of a normal distribution
with mean zero and standard deviation a. In the CAP Study, two
types of duplicate samples were examined: side-by-side samples
collected in the field, and spiked duplicate samples created in
the sample preparation laboratory.
9.3.1 Field Quality Control
Side-by-sides were included to determine variability
due to the sample collection process; however, this source of
variability will also be confounded with short-scale variations
attributable to nearby sampling locations within a room. Side-
by-sides were collected for dust vacuum and soil core samples,
but not for wipe samples. A pair of dust and soil duplicates
were collected at each housing unit surveyed.
Table 9-3 reports descriptive statistics for the side-
by-side samples. The statistics presented are the number of
samples collected, minimum ratio, maximum ratio, geometric mean
ratio, and log standard deviation. An estimated 95% tolerance
Volume II - Page 162
-------
bound was also calculated, using an exact procedure for the
distribution assumed for the log transformed ratio.
The side-by-side results are illustrated in Figure 9-3.
The ratio for each pair of samples is plotted by sample type.
The estimated 95% tolerance bound is portrayed in the figure by a
bar extending from a value of one up to the tolerance bound.
The soil side-by-sides exhibit better agreement than
the vacuum dust pairs. Their geometric mean was approximately
40% smaller than that for the paired dust vacuum lead
concentrations. The inherent variability between field samples,
however, is evident in these results. Despite being collected
side-by-side, a number of the pairs were measured to have
distinct lead contents. This disparity is reflected in the
higher ratios and relatively large estimated tolerance bounds.
9.3.2 Sample Preparation Quality Control
Spiked duplicate samples originate in the sample
preparation laboratory and are developed with identical lead
content. Each pair is derived from two identical spiked samples.
The spiked sample results are presented in section 9.2.1 where a
more detailed presentation of their development is available.
Spiked duplicates were generated at a frequency of two pair
(two spikes and two spiked duplicates) per batch of approximately
40 field samples.
A summary of the spiked duplicate sample results is
presented in Table 9-3. This summary is portrayed graphically in
Figure 9-3. The descriptive statistics are the same as those
developed for the field side-by-side samples. Performance-
Control charts showing the range of spiked sample and spiked
sample duplicate pairs are shown for each sample type in Figures
D-7, D-8, and D-9.
The range of spiked duplicate percent recoveries were
tighter for dust samples than for soil samples. This is not
surprising given the sampling protocol. Recall that spiked
blanks were employed for dusts (cassettes and wipes could not be
Volume II - Page 163
-------
100
s
0>
0)
vfl
ro
ON
1°
1
8
W
(Mg/g)
Side-by-Sides
Spiked Duplicates
Figure 9-3. Individual measurements and tolerance bounds for
the ratio of duplicate samples.
-------
split homogeneously) and regular field sample splits were
utilized for soils (see section 9.2.1). The slightly wider range
observed for dust wipes compared to vacuum dust may be a result
of an increased variability in the native lead background
commonly observed in baby wipes. The ranges observed for soils
imply that the 0.5 grain nominal sample weight used for sample
preparation may not be sufficient to overcome some heterogeneity
apparently still present in the dried, sieved, and homogenized
soil samples used for analysis. Still, the geometric means are
close to one and the estimated 95% tolerance bounds are not
unreasonably large. The results do suggest good agreement
between the spiked duplicate samples.
9.4 TIME TREND ANALYSES
The extensive samples collected in the CAP Study
required laboratory analyses which spanned several months. One
natural question, therefore, was whether any trend across time
was apparent in the samples. Specifically, is there a time-based
bias in the sampling results? The QC samples, expected to
demonstrate consistent sampling results, are ideal for this
examination.
The individual results for each of the QC measures
outlined above were plotted using a common frame of reference.
Each QC sample was plotted according to the instrument analysis
batch it was included in, and its run number within that batch.
The instrument batches were ordered based on the time they were
processed. For each QC sample type, the appropriate parameter
was displayed for the individual results. The measured amount of
lead (jig), for example, was displayed for the 52 vacuum dust
field blank results.
An examination of these plots suggested no evidence of
time trends, excepting the soil field and method blank results.
The plots are available in Volume III. Recall that more than 92%
of the soil field blank results were censored, as were 83% of the
soil method blanks. In the results, censored samples are set
Volume II - Page 165
-------
equal to the instrumental detection limit. Furthermore, these
blanks were all analyzed using the same dilution factor (50 mL).
Their apparent time trends were determined, therefore, to be a
function of the IDLs for the instrument batches containing the
soil samples. Figure 9-4 presents the available IDL results for
each instrument batch. Those batches which included soil samples
are identified as circles. Note that they do exhibit an apparent
quadratic trend across time. The IDLs considered as a whole, in
contrast, show no evidence of a trend. To assess the
significance of the apparent trend in the soil IDLs, quadratic
equations were fit to all the IDLs and only to those including
soil samples. The two resulting fits were not significantly
different (p=0.13). Given the apparent randomness exhibited by
the IDLs, there is no evidence of a time trend in the soil field
or method blank results.
Volume II - Page 166
-------
(D
H
H
I
•0
0)
•8
O.07
o.oe
O.06
O.O4
O.03
O.O2
O.O1
—i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 r
EEEEEEEEEEEEEEEEEEEEEEEE
oooooooooooooooooooooooo
446666666666666667777888
22OO11122OO1112221223OO2
7947239672412646941 4O364
222222222222222222222222
AAABBAAAAAAAAAAAAAAAAAAA
Instrument Batch
Batch Contained Soil? + + + No o o o
Figure 9-4. Time trend analyses in instrumental detection level by instrument batch,
-------
10. REFERENCES
Battelle Memorial Institute and Midwest Research Institute, 1991,
"Draft Final Report Comprehensive Abatement Performance Pilot
Study", report to U.S. EPA Office of Pollution Prevention and
Toxics, prepared under Contract No. 68-DO-0126 and Contract No.
68-DO-0137.
Battelle Memorial Institute and Midwest Research Institute,
1992a, "Detailed Design Document for the Comprehensive Abatement
Performance Study", report to U.S. EPA Office of Pollution
Prevention and Toxics, prepared under Contract No. 68-DO-0126 and
Contract No. 68-DO-0137.
Battelle Memorial Institute and Midwest Research Institute,
1992b, "Quality Assurance Project Plan for the Comprehensive
Abatement Performance Study", report to U.S. EPA Office of
Pollution Prevention and Toxics, prepared under Contract No. 68-
DO-0126 and Contract No. 68-DO-0137.
Battelle Memorial Institute and Midwest Research Institute, 1993,
"Draft Final Report, Comprehensive Abatement Performance Study,
Volume I: Results", report to U.S. EPA Office of Pollution
Prevention and Toxics, prepared under Contract No. 68-D2-0139 and
Contract No. 68-DO-0137.
Farfel, M. R., Chisolm, J. J., Jr. (1990) "Health and
environmental outcomes of traditional and modified practices for
abatement of residential lead-based paint." American Journal of
Public Health.
Farfel, M. R., Chisolm, J. J., Jr. (1991) "An evaluation of
experimental practices for abatement of residential lead-based
paint: report on a pilot project." Environmental Research.
55:199-212. 80(10):1240-1245.
Midwest Research Institute, 1991, "Engineering Study to Explore
Improvements in Vacuum Dust Collection", report to U.S. EPA
Office of Pollution Prevention and Toxics, prepared under
Contract No. 68-DO-0137.
Reeves, R., Kjellstron, T., Dallow, M., Mullins, P. (1982)
"Analysis of Lead in Blood, Paint, Soil, and House Dust for the
Assessment of Human Lead Exposure in Auckland". New Zealand
Journal of Science, 25:221-227.
Richardson and Wu (1970), "Least Squares and Grouping Method
Estimators in the Errors in Variables Model", JASA, p.724-749.
Volume II - Page 168
-------
U.S. Department of Housing and Urban Development, 1990,
Comprehensive and Workable Plan for the Abatement of Lead-Based
Paint in Privately Owned Housing, report to Congress, Washington,
DC, December 7, 1990.
U.S. Department of Housing and Urban Development, 1991, The HUD
Lead-Based Paint Abatement Demonstration (FHA). Office of Policy
Development and Research, Washington, DC.
Volume II - Page 169
-------
APPENDIX A
CONDENSED DATA LISTING
-------
o
(D
H
H
I
•d
(D
I
Unit Sample
ID ID
03 20
18
05
01
02
04
13
16
22
11
10
07
06
09
07 20
14
16
18
21
10
07
08
09
13
15
22
12
24
25
26
27
28
19
17
05
11
01
02
06
03
04
Room or
Yard
General
Sample
Location Location
BAC
BAC
BAT
BAT
BAT
BAT
FRO
FRO
FRO
KIT
LVG
LVG
LVG
LVG
BAC
BAC
BAC
BAC
BAC
DIN
DIN
DIN
DIN
EXT
FRO
FRO
FRO
HAL
KIT
KIT
KIT
KIT
KIT
KIT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
LVG
1
3
3
1
1
2
6
5
N/A
6
3
1
N/A
2
1
5
5
3
3
3
1
2
2
6
6
N/A
5
4
4
4
4
N/A
2
4
3
6
1
1
N/A
2
2
Lead
Date Concentration
Component
BDY
FDN
ARD
FLR
FLR
UST
EUO
EUY
N/A
EUI
ARD
FLR
N/A
UST
BDY
EUO
EUY
FON
FDN
ARD
FLR
UCH
UST
EUO
EUY
N/A
EUI
FLR
FLR
FLR
FLR
N/A
BDY
FDN
ARD
EUI
FLR
FLR
N/A
UCH
UST
Sample
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
"
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Collected
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
(ug/g)
129.80
111.99
74.82
32.73
269.71
48271.93
208.07
160.89
.
55.57
341.73
206.53
346.43
763.49
63.74
577.42
55.92
63.23
97.80
140.37
149.45
168.61
196.05
198^42
76.58
m
TsisS
soirs
187.38
.
.
m
52.53
105.83
768.11
176.59
99.93
72.79
893.83
120.29
168.80
Lead
Loading
-------
Unit
ID
09
O
M
e
3
(D
•o
ta
(D
>
to
Room or General
Sample Yard Sample
ID Location Location Component Sample Medium Sample Type
Lead Lead
Date Concentration Loading
Collected (ug/g)
-------
Room or General
Unit Sample Yard Sample
ID ID Location Location Component
10
c
&
(D
H
•0
fu
u
Sample Medium Sample Type
Lead Lead
Date Concentration Loading
Collected (ug/g) (ug/ft2)
Dust
Loading Below
(ug/ft2) Detection
20
14
16
18
05
01
02
03
04
19
21
13
15
17
22
11
32
12
29
24
25
26
27
28
31
10
07
06
09
BAG
BAG
BAG
BAG
BD1
BD2
BD2
BD2
BD2
BD2
BSM
EXT
FRO
FRO
FRO
FRO
FRO
FRO
HAL
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LVG
LVG
LVG
LVG
1
5
5
3
3'
1
1
2
2
2
2
6
6
4
N/A
6
3
5
1
4
4
4
4
N/A
2
3
1
N/A
2
BDY
EUO
EUY
FDN
ARO
FLR
FLR
UCH
UST
BDY
BDY
EUO
EUY
FDN
N/A
EUI
ARD
EUI
FLR
FLR
FLR
FLR
FLR
N/A
UST
ARD
FLR
N/A
UST
Soil
Dust
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Soil
Soil
Dust
Soil
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
- Vacuum
• Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
• Vacuum
• Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Uipe
- Uipe
• Uipe
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Field Blank
Regular
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
f
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
226.15
275.54
393.71
331.75
442!42
126.36
121.85
386.85
231 .32
25o!52
222.47
181.71
461.89
434.17
•
229.14
108.44
9.24
27.56
191.95
84.19
m
•
•
122.32
1775.20
98.04
253.40
6029.25
•
1066.25
.
73io6
56.65
45.78
3681.30
73.34
•
m
1933.10
•
•
•
642.65
14.03
25.50
1.31
5.11
27.41
4.16
13.05
•
12.86
982.57
35.52
•
1223.81
•
3869.70
•
16s!l4
448.30
375.70
9516.00
317.06
•
•
10638.30
•
2804i60
129.42
2758.80
47.70
26.60
325.60
.
m
m
105.11
553.50
362.30
•
202.98
-------
o
M
i
0)
•0
ftl
Unit Sample
ID ID
11 14
21
16
05
01
02
24
25
26
27
28
03
04
19
13
15
17
22
12
20
18
10
11
07
06
08
09
Room or
Yard
Location
BAG
BAG
BAG
BAT
BD1
BD2
DIN
DIN
DIN
DIN
DIN
DIN
DIN
DIN
DIN
DIN
EXT
FRO
FRO
FRO
FRO
FRO
KIT
LDY
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
General
Sample
Location
5
5
3
1
1
4
4
4
4
N/A
2
2
2
6
6
4
N/A
5
1
3
3
6
1
N/A
2
2
Lead
Date Concentration
Component
EWO
EUY
EUY
ARD
FLR
FLR
FLR
FLR
FLR
FLR
N/A
UCH
UST
BDY
EWO
EUY
FDN
N/A
EUI
BDY
FDN
ARD
EUI
FLR
N/A
UCH
UST
Sample
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Soil
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Side-by-Side
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Collected
03/25/92
03/25/92
03/25/92
_
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
f
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
(ug/g)
260.47
106.18
106.59
164o!o9
625.56
207.17
2027.74
429.51
B
B
.
804.04
754.99
10s!46
1653.27
116.04
117.03
m
392 !62
96.76
138.90
1019.72
1690.93
638.45
834.29
735.25
147.27
Lead
Loading
(ug/ft2)
270.00
m
m
661 !29
65.56
48.21
340.86
98.96
47.94
17.51
,
111S!39
119.21
|
4596! 25
_
\
l/olsA
.
m
S72. 71
137.98
26.11
.
21067.64
10.68
Dust
Loading Below
(ug/ft2) Detection
1036.60
403.20
104.80
232.70
168.10
230.40
.
_
<
1387123
157.90
278o!lO
| <
455 ISO
m
855 !83
81.60
40.90
<
28653 !82
72.54
-------
•o
o*
•8
Ul
Unit Sample
ID ID
14 05
01
02
03
04
13
15
12
19
17
10
11
07
06
08
09
20
14
16
21
18
22
16 05
01
02
03
04
10
07
06
09
12
19
13
15
17
22
20
21
18
11
14
16
Room or
Yard
Location
BD2
BD2
BD2
BD2
BD2
FRO
FRO
LDY
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
RGT
RGT
RGT
RGT
RGT
RGT
BD2
BD2
BD2
BD2
BD2
BD3
BD3
BD3
BD3
DIN
FRO
FRO
FRO
FRO
FRO
LFT
LFT
LFT
LVG
RGT
RGT
General
Sample
Location
3
1
1
2
2
6
6
5
2
4
3
6
1
N/A
2
2
1
5
5
5
3
N/A
3
1
1
2
2
3
1
N/A
2
5
2
6
6
4
N/A
1
1
3
6
5
5
Lead
Date Concentration
Component
ARD
FLR
FLR
WCH
UST
EUO
EUY
EUI
BDY
FDN
ARD
EUI
FLR
N/A
WCH
UST
BDY
EUO
EUY
EUY
FDN
N/A
ARD
FLR
FLR
WCH
UST
ARD
FLR
N/A
UST
EUI
BDY
EUO
EUY
FDN
N/A
BDY
BDY
FDN
EUI
EUO
EUY
Sample
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Oust -
Soil
Soil
Soil
Soil
Soil
Soil
Dust -
Dust -
Soil
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Field Blank
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Collected
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
04/06/92
177.92
203.24
151.76
147.08
181 .67
162.33
171.63
63.22
89.07
152.51
141.41
98.79
66.39
227.18
221.08
10.15
137.27
97.69
180.72
237.06
135.38
•
454.74
102.50
216.43
5330.04
4170.78
130.96
86.16
263.53
778.07
157.01
183.64
143.76
129.49
64.19
_
101 !38
87.86
64.71
138.23
316.23
215.83
Lead
Loading
(ug/ft2)
158.98
79.16
27.61
46.40
9.85
688.15
19!07
.
35.10
14.33
21.83
m
TZ'.OO
1.65
,
s!97
,
f
B
.
7.37
15.12
49.00
7481.60
571.47
7.60
21.39
t
146.42
1414.36
m
457.25
f
_
.
,
^
250.98
352.78
.
Dust
Loading Below
(ug/ft2) Detection
893.54
389.50
181.90
315.47
54.22
4239.10
301 !oO
m
248.23
145.10
328.80
<
325 .66
162.72
40!60
m
^
m
<
16.20 <
147.50
226.40
1403.67
137.02
58.06 <
248.20
<
188*. 18
9008.30
3180!70
m
<
*m
m
m
1815.60
1115.60
-------
Room or General
Unit Sample Yard Sample
ID ID Location Location Component Sample Mediun Sample Type
17
cf
t->
I
(D
•U
01
«Q
(D
Cft
Lead Lead
Date Concentration Loading
Collected (ug/g) (ug/ft2)
Dust
Loading Below
(ug/ft2) Detection
20
14
16
18
22
10
07
09
13
15
05
12
01
02
06
03
04
21
19
17
32
11
29
24
25
26
27
28
30
31
BAG
BAG
BAG
BAG
BAG
BD1
BD1
BD1
EXT
FRO
FRO
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LFT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
LVG
LVG
LVG
LVG
1
5
5
3
N/A
3
1
2
6
6
3
5
1
1
N/A
2
2
2
4
3
6
1
4
4
4
4
N/A
2
2
BDY
EWO
EUY
FDN
N/A
ARD
FLR
WST
EWO
EWY
ARD
EWI
FLR
FLR
N/A
WCH
WST
BDY
BDY
FDN
ARD
EWI
FLR
FLR
FLR
FLR
FLR
N/A
WCH
WST
Soil
Dust
Soil
Soil
Soil
Dust
Dust
Dust
Dust
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Uipe
- Uipe
- Uipe
- Vacuum
- Vacuum
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Field Blank
Regular
Regular
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
m
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
62
131
62
59
1067
577
410
8
71
464
239
38
73
1351
303
71
62
46
79
870
237
589
156
740
649
385
.15
.80
.03
.36
.
.92
.69
.47
m
.84
.29
.37
.86
.07
.04
.31
.96
.49
.36
.58
.34
.45
.94
.07
.08
.27
.
m
m
.19
.66
m
74.54
m
.
m
46.13
321.14
168.63
_
46J91
-
135.83
9.79
0.49
1.29
.
567.89
1.43
.
.
m
67.25
119.87
104.38
58.65
186.77
4.61
8.49
B
831.02
25.11
m
565.50
m
,
m
43.20
555.90
410.82
m
5305 !00
m
292.50
40.80
12.80
17.60
.
1868.27
20.00
m
m
m
77.26
503.80
177.20
375.80
252.30
.
.
m
1280.09
65.12
-------
6
(D
Di
«Q
(D
Unit Sample
ID ID
18 20
14
16
21
18
22
10
07
08
09
13
15
05
12
01
02
24
25
26
27
06
28
03
04
19
17
11
19 14
16
10
07
08
09
19
13
15
21
17
12
20
18
22
05
11
01
02
06
04
Room or
Yard
Location
BAG
BAG
BAG
BAG
BAG
BAG
BD1
BD1
BD1
BD1
EXT
FRO
FRO
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LFT
LFT
LVG
BAG
BAG
BD1
BD1
BD1
BD1
FRO
FRO
FRO
FRO
FRO
KIT
LFT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
General
Sample
Location
1
5
5
5
3
N/A
3
1
2
2
6
6
3
5
1
1
4
4
4
4
N/A
N/A
2
2
2
4
6
5
5
3
1
2
2
2
6
6
6
4
5
1
3
N/A
3
6
1
1
N/A
2
Lead
Date Concentration
Component
BDY
EWO
EWY
EWY
FDN
N/A
ARD
FLR
WCH
WST
EWO
EWY
ARD
EWI
FLR
FLR
FLR
FLR
FLR
FLR
N/A
N/A
WCH
WST
BDY
FDN
EWI
EWO
EWY
ARD
FLR
WCH
WST
BDY
EWO
EWY
EWY
FDN
EWI
BDY
FDN
N/A
ARD
EWI
FLR
FLR
N/A
WST
Sample
Soil
Dust -
Soil
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust •
Dust -
Dust -
Soil
Soil
Dust -
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Wipe
Wipe
Vacuum
Wipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Field Side-by-Side
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Field Blank
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Collected
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
(ug/g)
101.01
146.75
79.42
86.93
84.38
.
1552.50
94.43
9540.68
1991.10
145 '.89
59.58
1988.82
266.13
104.83
121.56
100.67
432.56
m
m
-491.83
m
799.14
403.44
62.03
189.45
67.87
566.74
99.65
316.17
197.42
176.30
109.00
77.48
259.01
38.00
38.29
39.05
2293.62
43.22
237.06
.
475.66
327.54
444.55
467.57
•300.29
272.15
Lead
Loading
(ug/ft2>
438.97
.
m
m
U36.t(>
50.50
19792.59
406.69
20l! 58
329! 07
82.13
11.86
11.55
16.61
72.97
4.15
8.15
m
.
1229.73
14.89
\
1415.50
988.85
38.45
61.71
186.71
3.40
202.65
.
.
,
261 l! 75
_
.
m
2384 .*72
2259.57
1102.35
6745.20
.
224.68
Dust
Loading Below
(ug/ft2) Detection
2991 .20
m
<
925.26
534.80
2074.55
204.26
1381.' 70
165 !46
308.60
113.10
95.00
165.00
168.70
.
a
<
\ <
1538.81
36 !90
20857i40
1744.80
121 !60
312.60
1059.02
31.20
782 i40
.
1138!70
"
«
5013^50
6898 ! 60
2479.70
14426.00
<
825^60
-------
c
ID
TJ
0)
iQ
(D
00
Unit Sample
ID ID
21 19
14
16
17
22
01
02
06
04
32
29
30
31
20
13
15
21
18
12
11
05
07
24
25
33
26
27
28
08
09
Room or
Yard
Location
BAG
BAG
BAG
BAG
BAG
BAT
BAT
BAT
BAT
BD3
BD3
BD3
B03
EXT
FRO
FRO
FRO
FRO
FRO
CAM
HAL
KIT
LOT
LDY
LOT
LDY
LDY
LDY
LDY
LDY
LDY
POR
General
Sample
Location
2
5
5
4
N/A
1
1
N/A
2
3
1
2
2
1
6
6
6
3
5
6
3
1
4
4
4
4
4
N/A
2
2
Component
BDY
EWO
EUY
FDN
N/A
FLR
FLR
N/A
WST
ARD
FLR
UGH
VST
BDY
EWO
EWY
EUY
FDN
EUI
EUI
ARD
FLR
FLR
FLR
FLR
FLR
FLR
N/A
UGH
UST
Sample
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust •
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Oust -
Dust -
Dust -
Dust -
Dust -
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vaciun
Vacuum
Vacuun
Vacuun
Vacuun
Vacuum
Vacuum
Vacuum
Vacuum
Vacuun
Wipe
Wipe
Wipe
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Regular
Field Blank
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Vacuum Wipe Comparison
Vacuun Wipe Comparison
Vacuum Wipe Comparison
Vacuun Wipe Comparison
Vacuum Wipe Comparison
Field Blank
Regular
Regular
Date
Collect*
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
Lead
Concentration
(ug/g)
119.61
251.32
130.00
272.29
172!15
282.75
-810.78
190.11
1905.20
263.00
25342.38
6824.51
276J93
1143.43
641.01
413.57
190.37
199.70
281.43
360.86
238.66
1033!44
734.95
2456.61
6197.57
Lead Dust
Loading Loading Below
(ug/ft2) (ug/ft2) Detection
1165.25 4636.50
60.82
161.96
36.61
181.07
121.09
12281.48
1299.66
497.33
230.71
740.78
420.25
7046)70
2356.89
333.56
26137.63
6494.53
353.30
572.80
95.04
460.40
484.62
190.44
2588.15 2263.50
2490.40
819.80
2052.80
1760.90
3206.85
10639.72
1047.92
-------
i
ID
TJ
OJ
*Q
(D
vo
Unit Sample
ID ID
22 01
02
03
04
10
07
08
09
19
15
21
17
22
12
13
11
06
24
OS
01
02
03
04
10
12
06
19
21
13
15
17
07
24
25
26
27
28
08
09
14
11
Room or
Yard
Location
BAT
BAT
BAT
BAT
BD1
BD1
BD1
BD1
FRO
FRO
FRO
FRO
FRO
KIT
LFT
LVG
LVG
BD2
BD3
BD3
BD3
BD3
BD3
DIN
DIN
DIN
EXT
FRO
FRO
FRO
FRO
FRO
KIT
LDY
LDY
LDY
LDY
LDY
LDY
LDV
LFT
LVG
General
Sample
Location
1
1
2
2
3
1
2
2
2
6
6
4
N/A
5
6
6
N/A
3
1
1
2
2
3
5
N/A
2
2
6
6
4
1
4
4
4
4
N/A
2
2
5
6
Lead
Date Concentration
Component
FLR
FLR
UCH
UST
ARD
FLR
UCH
UST
BDY
EUY
EUY
FDN
N/A
EUI
EWO
EUI
N/A
ARD
FLR
FLR
UCH
UST
ARD
EUI
N/A
BDY
BDY
EWO
EUY
FDN
FLR
FLR
FLR
FLR
FLR
N/A
UCH
UST
EWO
EUI
Sample
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Oust -
Dust -
Dust -
Dust -
Dust -
Dust -
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Field Blank
Regular
Regular
Regular
Field Blank
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Field Blank
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Collected
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
03/28/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
04/03/92
(ug/g)
89.89
64.95
412.95
235.93
228.91
118.54
436.34
164.52
53.39
56.97
116.24
59.99
.
83.47
58.41
77.14
3091.50
574!46
10o!&6
73.61
172.02
66.33
339.00
219.48
219.41
15K01
157.18
459.18
89.11
95.63
81.52
361.17
177.94
.
m
m
209.97
41.04
211.57
175.05
Lead
Loading
(ug/ft2)
16.26
24.29
5028.69
36.36
15.14
33.10
2033.28
26.41
.
.
_
m
TZ! 16
115.88
175.10
2is!25
36.90
33.97
413.73
5.22
259.11
1059.40
m
m
907.75
|
25.25
262!&3
161.67
5.07
3.54
m
2239.25
16.94
4690.50
482.93
Dust
Loading Below
(ug/ft2) Detection
180.90
373.90
12177.46
154.13
66.15
279.20
4659.84
160.54
m
m
m
<
864! 50
1983.90
2269.90
379.93
365.80
461.50
2405.12
78.75
764.34
4826.90
|
1976.90
|
309!&0
727.70
908.60
m
<
<
10664i61
412.83
22170.30
2758.80
-------
o
M
i
(D
•0
0>
(D
>
Unit Sample
ID ID
25 20
14
16
21
18
22
05
01
02
03
04
32
29
30
31
19
13
15
17
10
12
07
24
25
26
27
06
28
08
09
11
Room or
Yard
Location
BAG
BAG
BAG
BAG
BAG
BAG
BAT
BAT
BAT
BAT
BAT
DIN
DIN
DIN
DIN
EXT
FRO
FRO
FRO
FRO
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LVG
General
Sample
Location
1
5
5
5
3
N/A
3
1
1
2
2
3
1
2
2
2
6
6
4
3
5
1
4
4
4
4
N/A
N/A
2
2
6
Lead
Date Concentration
Component
BOY
EUO
EWY
EWY
FDN
N/A
ARD
FLR
FLR
UCH
UST
ARD
FLR
UCH
UST
BOY
EUO
EWY
FDN
ARD
EWI
FLR
FLR
FLR
FLR
FLR
N/A
N/A
UCH
UST
EUI
Sample
Soil
Dust -
Soil
Soil
Soil
Soil
Dust •
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Vacuum
Uipe
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Field Side-by-Side
Regular
Field Blank
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Vacuum Uipe Conparison
Vacuum Uipe Conparison
Vacuum Uipe Conparison
Vacuum Uipe Conparison
Field Blank
Field Blank
Regular
Regular
Regular
Collected
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
m
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
03/31/92
(ug/g)
60.83
519.22
124.26
117.49
156.93
.
264.26
210.19
158.92
9306.87
1372.86
94.49
36.36
17291.82
1648.67
,
54.61
141.19
340.88
128.17
130.13
131.99
71.82
22.88
25.57
.
.
274.26
M
9101.28
1657.85
93.35
Lead
Loading
(ug/ft2)
1034! 49
.
B
.
.
3.20
21.90
18.20
15567.63
69.36
6.18
54.96
19735.20
50.16
m
m
84.15
m
B
5.60
149.83
35.01
2.47
3.27
3.49
3.63
m
m
35787.70
291.87
306.40
Dust
Loading Below
(ug/ft2) Detection
1992! 40
m
u
m
<
12!l3
104.20
114.50
1672.70
50.52
65.37
1511.40
1141.30
30.42
m
m
596 ! 00
.
.
43.05
1135.10
487.50
108.00
128.00
<
m
<
<
3932.16
176.05
3282.40
26
BAT
EXT
KIT
-------
o
t->
§
(D
H
I
•O
0>
vQ
(D
Unit Sample
ID ID
27 10
07
05
01
02
06
03
04
12
13
15
21
08
09
19
14
16
17
11
20
18
22
28 19
14
16
17
10
07
09
13
15
12
05
01
02
06
03
04
11
20
21
18
22
Room or
Yard
Location
BA2
BA2
BD1
BD1
BD1
BD1
BD1
BD1
DIN
FRO
FRO
FRO
CAN
GAM
LFT
LFT
LFT
LFT
LVG
RGT
RGT
RGT
BAG
BAG
BAG
BAG
BAT
BAT
BAT
FRO
FRO
HAL
KIT
KIT
KIT
KIT
KIT
KIT
LVG
RGT
RGT
RGT
RGT
General
Sample
Location
3
1
3
1
1
N/A
2
2
5
6
6
6
2
2
2
5
5
4
6
1
3
N/A
2
5
5
4
3
1
2
6
6
5
3
1
1
N/A
2
2
6
1
1
3
N/A
Lead
Date Concentration
Component
ARD
FLR
ARO
FLR
FLR
N/A
UCH
UST
EWI
EWO
EWY
EWY
UCH
UST
BDY
EWO
EWY
FDN
EWI
BDY
FDN
N/A
BDY
EWO
EWY
FDN
ARD
FLR
UST
EWO
EWY
EWI
ARD
FLR
FLR
N/A
UCH
UST
EWI
BDY
BDY
FDN
N/A
Sample
Dust -
Dust -
Oust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Dust -
Dust -
Soil
Dust -
Soil
Soil
Dust -
Soil
Soil
Soil
Soil
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Soil
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuun
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Field Blank
Collected
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
(ug/g)
281.26
99.03
211.52
149.77
125.06
495.59
442.87
384.68
183.24
219.37
181.74
180.89
961.17
1136.78
129.80
117.43
127.23
148.89
345.74
256.67
1015.56
.
44.64
282.36
61.12
72.20
675.43
196.13
243.54
56.88
50.44
93.11
437.07
41.99
1351.31
-292.14
1325.94
121.97
98.70
55.22
57.74
125.04
.
Lead
Loading
(ug/ft2)
26.04
9.77
8.60
319.70
224.56
m
20517.83
84.70
378.98
11.85
m
u
17639.13
1367.46
88.27
m
m
487i25
.
.
246.93
.
128!63
5.71
9.82
25.42
104.35
98.91
0.34
0.58
m
2521 !44
3.79
207.31
m
a
_
m
Dust
Loading Below
(ug/ft2) Detection
92.58
98.70
40.64
2134.60
1795.60
<
46328 ! 75
220.18
2068.20
54.00
t
\
18351.71
1202.92
751^70
f
1409!30
.
.
874! 50
190!44
29.10
40.32
447.00
1120.70
226.30
8.20 <
0.43 <
<
1901 !62
31.06
2100.40
\
f
<
-------
o
I-1
i
(D
•0
01
(D
>
H
to
Unit Sample
ID ID
31 10
07
06
09
14
16
13
15
21
01
02
24
25
26
27
28
12
19
17
11
20
18
22
33 20
21
15
18
22
05
01
02
06
12
04
19
17
10
11
07
08
09
24
16
Room or
Yard
Location
BA2
BA2
BA2
BA2
BAC
BAG
EXT
FRO
FRO
FRO
HAL
HAL
HAL
HAL
HAL
HAL
HAL
HL2
LFT
LFT
LVG
RGT
RGT
RGT
FRO
FRO
FRO
FRO
FRO
KIT
KIT
KIT
KIT
LDY
LDY
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
RGT
General
Sample
Location
3
1
N/A
2
5
5
6
6
6
1
1
4
4
4
4
N/A
5
2
4
6
1
3
N/A
1
6
3
N/A
3
1
1
N/A
5
2
2
4
3
6
1
2
2
2
5
Component Sample Medium
ARD
FLR
N/A
UST
EWO
EUY
EWO
EWY
EWY
FLR
FLR
FLR
FLR
FLR
FLR
N/A
EWI
BDY
FDN
EWI
BDY
FDN
N/A
BDY
EUY
EUY
FDN
N/A
ARD
FLR
FLR
N/A
EUI
UST
BDY
FDN
ARD
EUI
FLR
UCH
UST
WST
EWY
Dust
Dust
Dust
Dust
Dust
Soil
Dust
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Soil
Dust
Soil
Soil
Soil
Soil
Soil
Soil
Soil
Soil
Dust
Dust
Dust
Dust
Oust
Dust
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Soil
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Uipe
- Uipe
- Uipe
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
Sample Type
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Field Side-by-Side
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Field Side-by-Side
Regular
Regular
Field Blank
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Lead
Date Concentration
Collected (ug/g)
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
04/04/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
516.49
1.71
182.84
442.48
696.04
154.23
555 ! 00
979.62
773.37
105.64
181.41
75.75
123.01
.
.
_
350.96
174.22
404.17
67.45
184.33
226.49
•
52.44
39.02
8.54
134.24
f
136.13
158.21
148.44
-245.69
25.55
3339.94
78.72
128.98
191.20
128.56
61.42
3067.23
340.90
439.10
193.32
Lead
Loading
(ug/ft2)
3686.76
2.21
7.82
4368 ! 40
478 ! 08
1U23
72.93
127.98
59.17
31.59
87.90
512)86
7.76
.
4)25
1.28
1.89
2.56
3007.05
\
1l!90
1665.96
63.70
2929.08
45.07
74.85
Dust
Loading Below
(ug/ft2) Detection
7138.13
1287.20
17)68
6276.10
861 !40
106.30
402)00
1689.50
481.00
t
1461 '.30
115.00
! <
| f
31)20
8.10
12.70
10o!30
900.33
62.24
12959.00
1037.10
954.96
132.21
170.46
-------
c7
M
I
(D
•0
0>
(D
>
Room or
Unit Sample Yard
10 ID
39 U
16
05
01
02
29
06
03
04
10
07
08
09
_
19
13
15
21
17
12
24
25
26
27
28
20
18
22
11
Location
BAG
BAG
BAT
BAT
BAT
BAT
BAT
BAT
BAT
BD1
BD3
BD3
B03
BD3
DIN
EXT
FRO
FRO
FRO
FRO
FRO
HAL
KIT
KIT
KIT
KIT
KIT
KIT
LFT
LFT
LFT
LVG
PAN
POR
General
Sample
Location
5
5
3
1
1
1
N/A
2
2
3
1
2
2
2
6
6
4
5
4
4
4
4
N/A
1
3
N/A
6
Component
EWO
EUY
ARD
FLR
FLR
FLR
N/A
WCH
UST
ARD
FLR
UGH
UST
BDY
EWO
EUY
FDN
FDN
EWI
FLR
FLR
FLR
FLR
N/A
BDY
FDN
N/A
EUI
Sample
Oust -
Soil
Dust •
Dust •
Dust -
Dust -
Oust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust •
Soil
Soil
Soil
Dust -
Dust •
Dust •
Dust •
Dust -
Dust •
Soil
Soil
Soil
Dust •
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuun
Vacuum
Vacuum
Wipe
Wipe
Wipe
Vacuum
Sample Type
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Vacuun Wipe Comparison
Vacuum Wipe Comparison
Vacuun Wipe Comparison
Vacuun Wipe Comparison
Field Blank
Regular
Regular
Field Blank
Regular
Date
Collect*
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
03/09/92
Lead
:entration
(ug/g)
161.29
183.67
712.39
349.90
200.88
62.65
•769.52
5156.14
5442.72
395)41
110.54
1573.52
207.37
184)46
62.98
545.76
454.29
370.01
123)45
106.41
505.51
.
•
73)85
37U98
Lead
Loading
(ug/ft2)
15.45
m
12011.20
25.37
8.42
3.27
^
7620.12
750.79
22l)75
34.39
90.16
20.37
•
18)94
w
.
•
14)lO
5.59
22.60
27.77
25.49
.
715)50
Dust
Loading
(ug/ft2)
95.80
_
16860.40
72.50
41.90
52.20
m
1477.87
137.94
560)&0
311.10
57.30
98.24
•
300)80
e
_
•
114)20
52.50
44.70
m
•
m
1923)50
Below
Detection
-------
cf
(D
•ti
0>
(D
>
Unit Sample
ID ID
40 19
14
16
21
17
05
24
25
26
27
28
03
04
10
07
06
OB
09
13
15
01
02
12
11
20
18
22
Room or
Yard
Location
BAG
BAG
BAG
BAG
BAG
BAT
BAT
BAT
BAT
BAT
BAT
BAT
BAT
B02
BD2
BD2
BD2
BD2
EXT
FRO
FRO
HAL
HAL
HL2
KIT
LVG
RGT
RGT
RGT
General
Sample
Location
2
5
5
5
4
3
4
4
4
4
N/A
2
2
3
1
N/A
2
2
6
6
1
1
5
6
1
3
N/A
Lead
Date Concentration
Component
BDY
EUO
EWY
EWY
FDN
ARD
FLR
FLR
FLR
FLR
N/A
WCH
WST
ARD
FLR
N/A
WCH
WST
EUO
EWY
FLR
FLR
EWI
EWI
BDY
FDN
N/A
Sample
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust •
Dust •
Dust -
Dust -
Dust -
Dust •
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Wipe
Wipe
Wipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Field Blank
Collected
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
>
03/10/92
03/10/92
03/10/92
03/10/92
m
03/10/92
03/10/92
03/10/92
03/10/92
03/10/92
(ug/g)
163.85
119.06
66.62
71.09
114.29
1588.70
209.59
152.26
.
.
m
255.06
2208.04
113.73
206.92
871 .43
235.73
160.31
230 !47
148.50
119.78
125.19
m
84.83
200.96
77.33
129.05
.
Lead
Loading
(ug/ft2)
15l!55
m
B
.
1176.93
1.66
0.64
3.54
6.60
B
44.25
10.23
1.90
128.41
B
239.30
18.86
41l!s3
.
230.55
577.51
siss
2920.20
_
m
m
Dust
Loading Below
(ug/ft2) Detection
1272!90
m
.
_
74o!si
7.90
4.20
<
m
<
173.49
4.63
16.70
620.60
<
1015.13
117.66
USsiftO
.
1924.70
4613.10
42 1 20
14531 !lO
_
m
<
-------
i
>
H
U1
Unit Sample
ID ID
41 14
IB
10
07
08
09
05
01
02
06
03
04
32
29
30
31
19
21
13
IS
17
22
11
24
25
26
27
28
12
Room or
Yard
Location
BAG
BAG
BAT
BD1
BD1
BD1
BD1
BD3
BD3
BD3
BD3
BD3
BD3
BD4
DIN
DIN
DIN
DIN
EXT
FRO
FRO
FRO
FRO
FRO
FRO
HAL
KIT
KIT
KIT
KIT
KIT
LVG
P02
General
Sample
Location
5
3
3
1
2
2
3
1
1
N/A
2
2
3
1
2
2
2
2
6
6
4
N/A
6
4
4
4
4
N/A
5
Lead
Date Concentration
Component
EWO
FDN
ARD
FLR
UCH
WST
ARD
FLR
FLR
N/A
WCH
WST
ARD
FLR
WCH
WST
BDY
BDY
EWO
EUY
FDN
N/A
EWI
FLR
FLR
FLR
FLR
N/A
•
EUI
Sample
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Wipe
Wipe
Wipe
Vacuum
Sample Type
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Blank
Regular
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Field Blank
Regular
Collected
03/30/92
03/30/92
•
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
>
03/30/92
03/30/92
03/30/92
03/30/92
>
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
<
03/30/92
(ug/g)
1668.88
1334.51
m
676.37
242.69
11398.99
3037.50
929.47
363.88
222.72
-893.45
3622.45
2379.10
_
60oi27
2398.76
6066.85
1653.03
,
179.22
186.61
320.32
147.30
182.74
•
235.42
127.94
60.40
•
.
.
m
vn.v>
Lead
Loading
(ug/ft2)
615.57
.
m
65.40
118.12
11413.91
430.03
13213.87
479.93
323.25
•
3342.36
131.96
m
102!54
43.42
15336.58
966.96
_
.
m
111.31
•
_
•
1982.30
3.24
1.45
6.59
8.08
.
>
45o!70
Dust
Loading Below
(ug/ft2) Detection
368.85
•
m
96.69
486.70
1001.31
141.57
14216.53
1318.90
1451.40
<
922.68
55.47
m
17o!83
18.10
2527.93
584.96
^
•
.
347.50
•
•
<
8420.20
25.30
24.00
•
•
<
m
9a'.90
-------
Room or General
Unit Sample Yard Sample
ID ID Location Location Component Sample Medium Sample Type
44
O
M
i
(D
0>
(D
>
H
Ot
Lead Lead
Date Concentration Loading
Collected (ug/g) (ug/ft2)
Dust
Loading Below
(ug/ft2) Detection
14
16
19
21
13
IS
17
22
10
07
08
09
12
01
02
24
25
26
27
06
28
03
04
OS
11
20
18
BAC
BAC
BAT
BD2
BSM
EXT
FRO
FRO
FRO
FRO
FRO
FRO
GAM
GAM
GAM
GAM
HAL
HAL
HAL
HAL
HAL
HAL
HAL
HAL
HAL
HAL
HAL
KIT
LVG
RGT
RGT
5
5
2
2
6
6
4
N/A
3
1
2
2
5
1
1
4
4
4
4
N/A
N/A
2
2
3
6
1
3
EWO
EWY
BDY
BDY
EWO
EWY
FDN
N/A
ARD
FLR
WCH
WST
EWI
FLR
FLR
FLR
FLR
FLR
FLR
N/A
N/A
WCH
WST
ARD
EUI
BDY
FDN
Dust
Soil
Soil
Soil
Dust
Soil
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Soil
- Vacuum
- Vacuun
- Vacuun
- Vacuun
- Vacuun
- Vacuun
- Vacuun
- Vacuun
- Vacuun
- Vacuun
- Vacuun
• Wipe
- Wipe
- Vacuun
- Wipe
- Vacuun
- Vacuun
- Vacuun
- Vacuun
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Vacuun Wipe Comparison
Vacuun Wipe Comparison
Vacuun Wipe Comparison
Vacuun Wipe Comparison
Field Blank
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
03/16/92
03/16/92
m
.
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
834.07
104.21
m
\
137.13
96.75
133.06
132.78
1211.55
.
202.46
95.78
72.90
69.56
95.05
199.06
58.63
39.25
16.67
.
m
172.14
m
298.51
97.07
118.71
256.45
120.23
113.41
2911.75
B
j
m
275.23
.
t
39.77
14.34
759.14
1.35
6.66
14.63
1.69
3.37
4.48
4.17
3.54
f
.
969.02
8.25
187.10
796.70
.
3491 .00
m
\
2068.50
|
_
196 !45
149.70
10413.80
19.38
70.10
73.50
28.90
85.90
268.60
f
m
B
f
3246.17
85.00
1576.13
3106.60
.
-------
Unit Sample
ID ID
45 14
05
01
02
04
19
13
15
17
22
10
12
07
08
09
11
06
20
18
21
Room or
Yard
Location
BAC
BD2
BD2
BD2
BD2
FRO
FRO
FRO
FRO
FRO
KIT
KIT
KIT
KIT
KIT
LVG
LVG
RGT
RGT
RGT
General
Sample
Location
5
3
1
1
2
2
6
6
4
N/A
3
5
1
2
2
6
N/A
1
3
3
Lead
Date Concentration
Component
EWO
ARD
FLR
FLR
WST
BOY
EWO
EWY
FDN
N/A
ARD
EWI
FLR
UCH
UST
EUI
N/A
BDY
FDN
FDN
Sample
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Nediun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Sample Type
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Field Side-by-Side
Collected
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
03/19/92
(ug/g)
247.32
292.32
55.13
107.73
703.50
131.35
49.51
72.91
55.80
,
475.26
167.61
6398.60
7044.19
617.04
51.74
-1594.76
200.54
121.47
129.79
Lead
Loading
(ug/ft2)
110.23
19.95
29.51
12.88
104.97
e
200.08
m
m
m
88.35
32.95
1765.38
12427.41
151.21
14.90
e
m
m
.
Dust
Loading Below
(ug/ft2) Detection
445.70
68.25
535.30
119.60
149.22
4041 !oO
m
m
<
185.89
196.60
275.90
1764.21
245.06
288.00
<
m
m
.
•0
(D
T
-------
Component Sample Hediun Sample Type
Lead Lead
Date Concentration Loading
Collected (ug/g) (ug/ft2)
Dust
Loading Below
(ug/ft2) Detection
<
O
H
I
•d
(D
>
00
14
16
05
01
02
24
25
26
27
28
03
04
19
21
13
15
17
12
20
18
10
11
07
06
08
09
22
21
19
13
15
17
05
12
01
02
03
04
20
18
22
10
11
07
06
08
09
14
16
BAC
BAC
BAT
BAT
BAT
BAT
BAT
BAT
BAT
BAT
BAT
BAT
BD1
FRO
FRO
FRO
FRO
FRO
CAN
KIT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
RGT
FRO
FRO
FRO
FRO
FRO
KIT
KIT
KIT
KIT
KIT
KIT
LFT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
RGT
RGT
5
5
3
1
1
4
4
4
4
N/A
2
2
2
2
6
6
4
5
1
3
3
6
1
N/A
2
2
N/A
2
6
6
4
3
5
1
1
2
2
1
3
N/A
3
6
1
N/A
2
2
5
5
EUO
EUY
ARD
FLR
FLR
FLR
FLR
FLR
FLR
N/A
UCH
UST
BDY
BDY
EUO
EUY
FDN
EUI
BDY
FDN
ARD
EUI
FLR
N/A
UCH
UST
N/A
BDY
BDY
EUO
EUY
FDN
ARD
EUI
FLR
FLR
UCH
UST
BDY
FDN
N/A
ARD
EUI
FLR
N/A
UCH
UST
EUO
EUY
Oust
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Soil
Dust
Soil
Soil
Dust
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Soil
Soil
Dust
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Uipe
- Uipe
- Wipe
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
Regular
Regular
Regular
Regular
Field Side-by-Side
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Field Blank
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
m
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
m
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
03/17/92
983.74
296.50
4623.43
67.73
95.30
109.85
75.77
.
.
m
5385.66
218.97
161 .57
228.79
580.03
66.47
318.96
m
55o!75
59.79
555.49
205.24
702.58
487.79
-214.24
2842.09
108.38
•
104.94
102.62
166.78
549.80
185.28
368.54
173.94
369.82
4.56
841.47
68.45
119.77
177.51
.
1807.51
144.18
77.11
-768.10
278.90
296.58
286.03
252.78
1107.40
m
555.87
10.24
2.99
0.89
0.81
7.20
45.37
m
9458.11
10.22
[
.
2672.25
a
553i 78
m
m
33.77
1256.35
31.37
_
859.69
1.01
•
m
30.84
m
f
1.85
2.09
0.92
0.81
2402.05
10.25
B
.
a
108.12
1.23
16.85
f
1486.47
6.13
516.75
.
1125.70
120.23
151 !20
31.40
8.10
10.70
a
.
p
1756.17
46.69
|
.
4607.10
-
1 DOS! 50
.
164.52
1788.20
64.30
_
302.49
9.36
.
184.90
.
m
s!oi
12.00
2.50
177.80
2854.59
149.76
m
m
m
59.82
8.50
218.50
m
5329.75
20.67
1806.60
.
-------
c?
(D
M
H
I
•0
(D
Unit Sample
ID ID
50
19
14
16
17
21
.
01
02
24
25
26
27
06
28
04
13
15
12
10
07
09
11
20
18
22
Room or
Yard
Location
BA2
BAG
BAG
BAG
BAG
BAG
BAT
BD2
BSN
BSM
BSH
BSM
BSN
BSN
BSN
BSH
BSN
EXT
FRO
FRO
CAN
HAL
KIT
KIT
KIT
LVG
RGT
RGT
RGT
General
Sample
Location
2
5
5
4
4
1
1
4
4
4
4
N/A
N/A
2
6
6
5
3
1
2
6
1
3
N/A
Lead
Date Concentration
Component
BDY
EWO
EWY
FDN
FDN
FLR
FLR
FLR
FLR
FLR
FLR
N/A
N/A
UST
EWO
EWY
EUI
ARD
FLR
UST
EUI
BDY
FDN
N/A
Sample
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuun
Uipe
Uipe
Vacuun
Uipe
Vacuum
Vacuun
Vacuum
Vacuun
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Field Side-by-Side
Vacuun Uipe Comparison
Vacuun Uipe Comparison
Vacuum Uipe Comparison
Vacuun Uipe Comparison
Field Blank
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Collected
•
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
m
•
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
m
03/27/92
03/27/92
m
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
(ug/g)
•
60.03
151.82
79.35
200.32
285.35
m
•
106.55
38.85
57.00
84.07
•
•
153.08
•
2058.54
m
75.27
50.14
m
138!42
182.55
47.18
807.44
127.62
28.72
192.16
•
Lead Dust
Loading Loading Below
(ug/ft2) (ug/ft2) Detection
899.50 5924.70
23.74
1.99
7.05
2.96
3.54
8.76
525.10
20! 10
111.58
4.26
1.26
114.13
269.55
222.80
51.10
123.60
35.20
255.09
267!10
806.10
23.36
26.70
141.35
2112.10
M
VO
-------
Room or General
Unit Sample Yard Sample
ID ID Location Location Component Sample Median Sample Type
Lead Lead
Date Concentration Loading
Collected
-------
0
H
Unit Sample
ID ID
53 20
14
16
18
21
05
01
02
06
03
04
10
07
OB
09
12
13
15
19
17
22
11
Room or
Yard
Location
BAG
BAG
BAG
BAG
BAG
BAT
BAT
BAT
BAT
BAT
BAT
BD3
BD3
BD3
BD3
FAN
FRO
FRO
LFT
LFT
LFT
LVG
General
Sample
Location
1
5
5
3
3
3
1
1
N/A
2
2
3
1
2
2
5
6
6
2
4
N/A
6
Lead
Date Concentration
Component
BOY
EUO
EUY
FDN
FDN
ARD
FLR
FLR
N/A
UCH
UST
ARD
FLR
UCH
UST
EUI
EUO
EUY
BOY
FDN
N/A
EUI
Sample
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Soil
Dust -
Mediun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Sample Type
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Collected
03/25/92
03/24/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
03/25/92
(ug/g)
131.30
77.88
137.32
128.59
137.04
358.59
78.23
42.38
556.19
656.25
50.63
130.68
164.24
360.86
290.39
217.01
54.42
242.06
1072.76
100.91
f
96.89
Lead
Loading
(ug/ft2)
13!60
m
t
.
13.54
2.30
1.09
t
573s!31
0.86
3.53
26.74
207.27
6.72
301.43
60.40
a
.
_
.
36.55
Dust
Loading Below
(ug/ft2) Detection
174i60
p
m
37.76
29.40
25.80
<
8736i47
16.90 <
27.04
162.80
574.37
23.13 <
1389.00
1109.80
,
m
<
377.20
•a
to
to
-------
Room or General
Unit Sample Yard Sample
ID ID Location Location
55
O
M
c
9
(D
0)
(D
>
to
NJ
14
16
10
07
08
09
19
13
15
17
12
20
21
22
05
11
01
02
24
25
26
27
06
28
03
04
18
BAG
BAG
BD1
BD1
BD1
BD1
EXT
FRO
FRO
FRO
FRO
GAM
HAL
KIT
LDY
LFT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
LVG
LVG
LVG
LVG
LVG
LVG
RGT
5
5
3
1
2
2
2
6
6
4
1
1
N/A
3
6
1
1
4
4
4
4
N/A
N/A
2
2
3
nponent Sample Medium
EUO
EUY
ARD
FLR
UGH
UST
BOY
EUO
EUY
FDN
EUI
BDY
BOY
N/A
ARD
EWI
FLR
FLR
FLR
FLR
FLR
FLR
N/A
N/A
UCH
WST
FDN
Dust
Soil
Dust
Dust
Dust
Dust
Soil
Dust
Soil
Soil
Dust
Soil
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
- Vacuum
- Vacuum
• Vacuum
- Vacuum
• Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Wipe
- Wipe
- Vacuum
- Wipe
- Vacuum
- Vacuum
Sample Type
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Field Side-by-Side
Vacuum Wipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Field Blank
Regular
Regular
Regular
Date
Collected
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
Lead
Concentration
(ug/g)
368.59
53.50
180.87
321.71
2781.74
1071.56
281 !?5
212.40
45.70
265.78
321 !99
174^50
194.68
.
1021 .74
196.34
965.96
372.07
392.40
251 .47
f
.
-154.82
m
9585.86
12798.85
208.07
Lead
Loading
(ug/ft2)
449.35
60s! 04
24.39
6038.22
200.71
•
•
486.03
•
45.40
•
•
_
3059.10
42.14
302.93
193.55
8.16
13.58
8.37
13.33
.
51335.38
732.01
•
Dust
Loading
(ug/ft2)
1219.10
•
3361.75
75.80
2170.67
187.31
"
2288.20
•
141 !oO
•
•
•
2994.00
214.60
313.60
520.20
20.80
54.00
•
"
5355 !32
57.19
Be I OH
Detection
-------
(D
0>
UQ
(D
Unit Sample
ID 10
57 24
25
26
27
28
20
21
14
16
18
22
10
07
08
09
05
01
02
06
03
04
13
15
12
19
17
11
Room or
Yard
Location
BA2
BA2
BA2
BA2
BA2
BAC
• BAC
BAC
BAC
BAC
BAC
BAT
BAT
BAT
BAT
BD1
B01
BD1
BD1
BD1
BD1
EXT
FRO
FRO
GAM
LFT
LFT
LVG
General
Sample
Location
4
4
4
4
N/A
1
1
5
5
3
N/A
3
1
2
2
3
1
1
N/A
2
2
6
6
5
2
4
6
Lead
Date Concentration
Component
FLR
FLR
FLR
FLR
N/A
BOY
BOY
EUO
EUY
FDN
N/A
ARD
FLR
UCH
WST
ARD
FLR
FLR
N/A
UCH
UST
EUO
EUY
EUI
BOY
FDN
EUI
Sample
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Dust -
Medium
Vacuum
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Collected
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
.
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
(ug/g)
38.81
44.79
.
.
.
55.49
56.90
35.03
42.85
73.34
.
487.64
1136.51
252.36
317.53
198.28
37.34
35.42
-183.22
1302.19
827.06
.
125.24
63.88
189.08
70.21
51.35
211.52
Lead
Loading
(ug/ft2)
0.96
1.36
3.00
2.72
.
.
.
16.29
m
m
m
369.83
12.39
21.47
12.22
780.17
34.40
30.19
.
854.87
171.44
.
667.35
.
78.17
.
m
690.48
Dust
Loading Below
-------
O
!-•
i
(D
•0
0>
ID
T
(O
Unit Sample
ID ID
60 21
20
14
16
18
22
05
01
02
03
04
13
15
12
24
25
26
27
28
19
17
10
11
07
06
08
09
Room or
Yard
Location
BAG
SAC
BAG
BAC
BAG
BAC
BD1
BD1
BD1
BD1
BD1
BD2
EXT
FRO
FRO
GAM
HAL
KIT
KIT
KIT
KIT
KIT
KIT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
General
Sample
Location
1
5
5
3
N/A
3
1
1
2
2
6
6
5
4
4
4
4
N/A
2
4
3
6
1
N/A
2
2
Component
BDY
BOY
EWO
EUY
FDH
N/A
ARD
FLR
FLR
UCH
UST
EWO
EWY
EWI
FLR
FLR
FLR
FLR
N/A
BDY
FDN
ARD
EUI
FLR
N/A
WCH
WST
Sample
Soil
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Oust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust •
Dust -
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Wipe
Wipe
Wipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Field Side-by-Side
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Date
Col lecte
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
03/15/92
Lead
Concentration
(ug/g)
322.87
150.19
208.90
109.32
114.80
11.00
44.15
635.15
324.21
92.67
124.57
65.70
131.67
81.95
97.18
937.98
384.85
163.83
155.93
-236.25
121.87
599.61
Lead Dust
Loading Loading Below
(ug/ft2) (ug/ft2) Detection
1477.20 7071.20
31.71
4.05
12.32
7244.55
13.71
5.31
1.38
3.08
3.54
4.71
1248.96
90.11
15.70
215 '.S3
27.45
542.29
368.60
279.10
11406.00
42.27
266.38 2874.60
80.80
10.50
37.60
3245.28
550.00
100.70
1768!52
45.78
-------
General
Sample
Location Component Sample Medium Sample Type
Lead Lead
Date Concentration Loading
Collected (ug/g) (ug/ft2)
Dust
Loading Below
(ug/ft2) Detection
O
M
§
•a
0)
(D
>
M
U1
19
U
16
17
22
05
01
02
03
04
20
21
13
15
18
11
32
12
29
30
31
10
07
06
08
09
14
16
05
01
02
06
03
04
19
13
15
17
12
10
11
07
08
09
21
20
18
22
BAC
BAC
BAC
BAC
BAC
BAT
BAT
BAT
BAT
BAT
EXT
FRO
FRO
FRO
FRO
FRO
GAM
HAL
KIT
KIT
KIT
KIT
KIT
LVG
LVG
LVG
LVG
LVG
BAC
BAC
BD4
BD4
BD4
BD4
BD4
604
FRO
FRO
FRO
FRO
KIT
LVG
LVG
LVG
LVG
LVG
RGT
RGT
RGT
RGT
2
5
5
4
N/A
3
1
1
2
2
1
1
6
6
3
6
3
5
1
2
2
3
1
N/A
2
2
5
5
3
1
1
N/A
2
2
2
6
6
4
5
3
6
1
2
2
1
3
N/A
BDY
EUO
EWY
FDN
N/A
ARD
FLR
FLR
WCH
WST
BDY
BDY
EUO
EWY
FDN
EUI
ARD
EUI
FLR
UCH
UST
ARO
FLR
N/A
UCH
UST
EUO
EWY
ARD
FLR
FLR
N/A
UCH
UST
BDY
EUO
EWY
FDN
EUI
ARD
EUI
FLR
UCH
UST
BDY
BDY
FDN
N/A
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust •
Dust -
Dust -
Soil
Dust •
Dust •
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Soil
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Field Blank
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
m
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
f
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
03/27/92
361.24
226.21
297.93
420.62
_
438 ! 89
145.43
128.92
17053.46
1055.32
137!57
132.39
188.49
183.32
312.00
>
12a!z4
1178.23
243.68
69.61
19169.39
7649.41
290.13
139.85
-2903.72
4849.87
605.82
85.00
205.76
709.18
170.21
249.36
372.06
1190.49
346.48
229.22
691.82
262.09
225.26
240.18
269.40
430.33
842.02
586.83
376.21
208.86
181.96
356.71
.
•
325.73
•
•
•
146.10
49.59
86.41
112917.92
153.50
|
B
772.32
B
.
m
128?!00
468.60
1153.00
151.57
39308.39
3481.40
22.56
156.76
•
24453.52
112.37
816.70
•
27.63
64.99
123.76
.
855.25
47.62
•
7127.30
_
•
69.97
4.04
1149.45
441.39
738.00
32.82
m
m
m
.
•
1439.90
•
•
•
332.89
341.00
670.20
6621.41
145.46
°
m
4097.40
.
m
10051 !30
397.71
4731.60
2177.40
2050.58
455.12
77.76
1120.90
5042! 10
185.48
9607.70
•
38.96
381.80
496.30
_
7ia!40
137.44
•
10302.20
•
291.30
15.01
2671.10
524.20
1257.60
87.25
m
m
m
m
-------
Room or General
Unit Sample Yard Sample
10 ID Location Location Component Sample Medium Sample Type
68
3
(D
Lead Lead
Date Concentration Loading
Collected (ug/g) (ug/ft2)
Dust
Loading Below
(ug/ft2) Detection
19
21
H
16
17
10
07
06
08
09
20
13
15
18
22
05
12
01
02
03
04
11
BAC
BAC
BAC
BAC
BAC
DIN
DIN
DIN
DIN
DIN
FRO
FRO
FRO
FRO
FRO
KIT
KIT
KIT
KIT
KIT
KIT
LVG
2
2
5
5
4
3
1
N/A
2
2
1
6
6
3
N/A
3
5
1
1
2
2
6
BOY
BDY
EWO
EUY
FDN
ARD
FLR
N/A
UCH
WST
BDY
EUO
EUY
FDN
N/A
ARD
EWI
FLR
FLR
UCH
UST
EUI
Soil
Soil
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
03/30/92
66.56
58.92
515.91
631.73
8.51
5644.54
115.60
-1935.81
10116.70
5062.83
212.58
190.49
655.14
648.06
m
254.25
1200.39
245.28
200.66
14271.12
3234.31
237.05
•
275.19
•
.
4652.72
31.96
m
33021.38
597.57
•
234.63
_
•
.
161.27
1190.43
79.55
1.97
244581.21
84.83
1108.40
.
533.40
•
•
824.29
276.50
•
3264.05
118.03
.
1231.70
m
•
m
634! 28
991.70
324.30
9.80
17138.19
26.23
4675.90
TJ
D>
(D
>
NJ
O\
-------
(D
H
H
•a
01
(D
>
Unit Sample
ID ID
69 14
16
32
19
13
15
17
22
29
24
25
26
27
28
31
05
12
01
02
03
04
20
21
18
10
11
07
06
09
Room or
Yard
Location
BAG
BAG
BAT
BD1
BSN
EXT
FRO
FRO
FRO
FRO
FRO
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LOT
LDY
LOY
LDY
LDY
LDY
LFT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
General
Sample
Location
5
5
3
2
6
6
4
N/A
1
4
4
4
4
N/A
2
3
5
1
1
2
2
1
3
3
6
1
N/A
2
Lead
Date Concentration
Conponent
EWO
EUY
ARD
BDY
EWO
EUY
FDN
N/A
FLR
FLR
FLR
FLR
FLR
N/A
WST
ARD
EUI
FLR
FLR
UGH
UST
BDY
FDN
FDN
ARD
EUI
FLR
N/A
UST
Sample
Dust -
Soil
Dust -
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Mediun
Vacuum
Vacuum
Vecuun
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Field Blank
Regular
Collected
04/01/92
04/01/92
04/01/92
•
•
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
(ug/g)
169.95
133.53
709.30
|
137i61
334.09
205.84
306.72
.
47.03
153.56
105.29
•
.
m
419.69
106.61
192.03
248.24
62.59
209.83
92.03
84.92
401.90
248.50
1012.70
142.71
88.60
1659.27
27058.49
Lead
Loading
(ug/ft2)
73.64
m
148.74
•
•
717.66
•
•
_
6i73
94.96
27.64
3.54
4.51
m
9.67
35.64
1010.65
47.24
31.03
232.06
2.86
.
.
f
283.02
57.20
26.63
p
6883.68
Dust
Loading
(ug/ft2)
433.30
•
209.70
|
|
2148!lO
.
>
KsilO
618.40
262.50
B
,
B
23.04
334.33
5263.10
190.30
495.80
1105.92
31.12
•
•
279.47
400.80
300.50
•
254.40
Below
-------
<
O
ID
D)
(0
>
03
Unit Sample
10 ID
70 20
21
18
10
07
09
19
13
15
17
22
11
05
12
01
02
24
25
26
27
28
03
04
14
16
06
Roan or
Yard
Location
BAC
BAG
BAC
BAT
B01
BD2
BD2
BD2
DIN
EXT
FRO
FRO
FRO
FRO
FRO
HAL
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LFT
LFT
LVG
PAN
General
Sample
Location
1
3
3
1
2
2
6
6
4
H/A
6
3
5
1
1
4
4
4
4
N/A
2
2
5
5
N/A
Lead
Date Concentration
Component
BDY
FDN
FDN
ARD
FLR
UST
BDY
EUO
EWY
FDN
N/A
EUI
ARD
EUI
FLR
FLR
FLR
FLR
FLR
FLR
N/A
UCH
UST
EUO
EWY
N/A
Sample
Soil
Soil
Soil
Dust -
Dust -
Dust •
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Oust -
Dust -
Dust •
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Field Side-by-Side
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Field Blank
Collected
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
(ug/g)
182.04
380.03
437.66
550 ! 54
106.40
4733.33
444! 76
324.49
265.20
99.70
1124!68
848.19
852.90
94.43
150.21
74.89
514.58
B
m
•
10407.95
5276.03
153.30
358.19
260.19
Lead
Loading
(ug/ft2)
•
•
48!l8
28.46
413.26
2036!80
B8!06
559.26
1807.90
1.16
1.61
0.69
1.54
5.60
4.80
68986167
2934.53
16.65
B
Dust
Loading
(ug/ft2>
•
87!s2
267.50
87.31
6277ioO
7s!30
659.36
2119.70
12.30
10.70
9.20
3.00
.
.
662&I27
556.20
108.60
[
Be I on
-------
Component Sample Mediun Sample Type
Lead
Date Concentration
Collected (ug/g)
Lead Dust
Loading Loading Below
(ug/ft2) (ug/ftZ) Detection
O
(D
•O
01
(O
vo
14
16
10
07
08
09
19
13
IS
17
21
01
02
24
25
26
27
28
03
04
12
32
11
29
06
30
31
20
18
22
BAC
BAG
BAT
BD1
BD1
BD1
BD1
BD2
BSN
EXT
FRO
FRO
FRO
FRO
FRO
GAM
HAL
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LDY
LVG
LVG
LVG
LVG
LVG
LVG
RGT
RGT
RGT
5
5
3
1
2
2
2
6
6
4
4
1
1
4
4
4
4
N/A
2
2
5
3
6
1
N/A
2
2
1
3
N/A
EWO
EWY
ARD
FLR
UCH
UST
BOY
EWO
EWY
FDN
FDN
FLR
FLR
FLR
FLR
FLR
FLR
N/A
UCH
UST
EUI
ARD
EUI
FLR
N/A
UCH
UST
BDY
FDN
N/A
Oust
Soil
Dust
Dust
Dust
Dust
Soil
Dust
Soil
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Soil
Soil
• Vacuum
- Vacuum
• Vacuum
- Vacuum
- Vacuum
• Vacuum
• Vacuum
• Vacuum
• Vacuum
• Vacuum
- Uipe
• Uipe
• Uipe
• Vacuum
- Vacuum
- Vacuum
- Vacuum
• Vacuum
• Vacuum
- Vacuum
- Vacuum
• Vacuum
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Field Side-by-Side
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Field Blank
03/23/92
03/23/92
m
03/23/92
03/23/92
03/23/92
03/23/92
m
•
•
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
•
.
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
164.06
345.99
a
1370.52
143.47
7714.68
6078.61
•
•
.
326.19
161.95
92.94
78.99
92.28
•
.
58.30
390.70
109.38
158.21
•
•
•
18311.15
13469.26
370.39
400.72
250.55
156.67
-769.52
2649.43
1990.75
138.02
218.86
.
678.10
.
m
15703.01
139.42
2335.39
686.94
•
•
•
.
910.07
•
•
.
•
.
6.50
31.61
7.18
63.51
35.29
15.40
•
73338.35
1760.65
1094.63
1689.80
1285.80
66.13
•
12387.93
553.83
•
m
.
4133.20
•
m
11457.73
971.80
302.72
113.01
•
_
_
.
5619.60
•
•
m
•
_
111.50
80.90
65.60
401.43
•
,
•
4005.12
130.72
2955.30
4216.93
5132.00
422.10
m
4675.70
278.20
•
m
•
-------
(D
M
M
I
01
U)
O
Unit Sample
ID ID
72 10
07
06
08
09
05
01
02
03
04
19
13
15
17
22
11
32
24
25
26
27
28
29
31
30
20
18
21
Room or
Yard
Location
BAT
BAT
BAT
BAT
BAT
BD1
BD1
BD1
BD1
BD1
BD3
BSM
EXT
FRO
FRO
FRO
FRO
FRO
HAL
KIT
KIT
KIT
KIT
KIT
KIT
LVG
LVG
LVG
RGT
RGT
RGT
General
Sample
Location
3
1
N/A
2
2
3
1
1
2
2
2
6
6
4
N/A
6
3
4
4
4
4
N/A
1
2
2
1
3
3
Lead
Date Concentration
Component
ARD
FLR
N/A
UCH
UST
ARD
FLR
FLR
UCH
UST
BOY
EWO
EUY
FDN
N/A
EUI
ARD
FLR
FLR
FLR
FLR
N/A
FLR
UCH
UST
BDY
FDN
FDN
Sample
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust •
Dust -
Dust -
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Collected
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
B
.
,
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
03/14/92
(ug/g)
379.81
225.80
506.74
143.21
1310.53
1098.45
594.85
322.67
1622.06
1054.74
.
.
_
447! 77
92.73
156.76
319.40
t
9.65
257.78
19.46
78.19
.
.
.
200.45
9421.10
2903.62
192.92
310.60
211.16
Lead
Loading
(ug/ft2)
1004.87
119.41
.
45.66
3319.39
6101.28
261 .68
109.22
2214.40
1007.37
m
m
B
441.99
m
m
m
27.20
1229.31
1.70
1.41
6.23
7.74
m
58.01
49932.75
445.13
.
m
.
Dust
Loading Below
(ug/ft2) Detection
2645.73
528.80
<
318.85
2532.86
5554.44
439.90
338.50
1365.18
955.10
m
m
f
4766.50
B
.
<
2820.20
4768.92
87.30
18.00
.
m
m *
289.40
5300.10
153.30
m
m
m
-------
Unit
ID
74
c7
t->
i
(D
•fl
0>
«Q
(D
U)
Room or General
Sample Yard Sample
ID Location Location Component Sample Medium Sample Type
Lead
Date Concentration
Collected (ug/g)
Lead Dust
Loading Loading Below
(ug/ft2) (ug/ft2) Detection
19
14 •
16
17
21
05
01
02
24
25
26
27
27
06
28
03
04
13
15
10
07
08
09
12
11
20
18
22
BAG
BAG
BAG
BAG
BAG
BAT
BD1
BD2
BD2
BD2
BD2
BD2
BD2
BD2
BD2
B02
B02
802
B02
EXT
FRO
FRO
KIT
KIT
KIT
KIT
LOT
LVG
RGT
RGT
RGT
2
S
S
4
4
3
1
1
4
4
4
4
4
N/A
N/A
2
2
6
6
3
1
2
2
5
6
1
3
N/A
BDV
EWO
EUY
FDN
FDN
ARD
FLR
FLR
FLR
FLR
FLR
FLR
FLR
N/A
N/A
UGH
UST
EWO
EUY
ARD
FLR
UGH
UST
EUI
EUI
BOY
FDN
N/A
Soil
Oust
Soil
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Soil
Soil
- Vacuum
Regular
Regular
Regular
Regular
Field Side-by-Side
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Wipe
• Wipe
- Wipe
- Vacuum
- Wipe
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
Regular
Regular
Field Side-by-Side
Vacuum Wipe
Vacuum Wipe
Vacuum Wipe
Vacuum Wipe
Vacuum Uipe
Field Blank
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Comparison
Genoa risen
Comparison
Comparison
Comparison
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
m
•
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
m
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
144.24
186.61
637.89
1678.61
654.01
m
•
509.22
115.10
132.71
313.32
104.09
•
•
.
16370.53
•
30077.53
679.25
m
228.52
555.52
467.33
328.05
3474.09
3135.61
196.74
233.37
167.86
1125.26
•
m
1094.70
.
•
m
m
4s!38
24.78
62.93
57.09
40.80
15.30
10.39
122.63
m
m
28664.44
164.48
m
2353.80
m
74.11
592.63
1519.71
155.21
251.65
1585.00
m
m
m
,
5866.10
m
•
.
m
95!oO
215.30
474.20
182.20
392.00
m
m
m
m
m
953.02
242.15
m
10300.40
m
158.59
1806.50
437.44
49.50
1279.10
6791.90
m
•
.
-------
Room or General
Unit Sample Yard Sample
ID ID Location Location Component Sample Medium Sample Type
77
(D
H
Lead Lead
Date Concentration Loading
Collected (ug/g) (ug/ft2)
Dust
Loading Below
(ug/ft2) Detection
20
14
16
18
21
13
15
12
01
02
24
25
26
27
28
03
04
19
17
22
10
11
07
06
08
09
BAC
BAC
BAC
BAC
BAC
EXT
FRO
FRO
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LFT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
1
5
5
3
3
6
6
5
1
1
4
4
4
4
N/A
2
2
2
4
N/A
3
6
1
N/A
2
2
BDY
EUO
BUY
FDN
FDN
EWO
BUY
EUI
FLR
FLR
FLR
FLR
FLR
FLR
N/A
UCH
UST
BDY
FDN
N/A
ARD
EUI
FLR
N/A
UCH
UST
Soil
Dust -
Soil
Soil
Soil
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Field Side-by-Side
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Field Blank
Regular
Regular
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
m
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
03/24/92
71.
94.
107.
201.
208.
15l!
203!
91.
76.
83.
42.
89.
m
.
_
149.
193.
91.
111.
m
464.
207.
190.
-1762.
1945.
1782.
27
76
98
14
84
17
70
59
63
29
48
23
80
30
09
33
35
22
76
11
34
75
235 ! 77
.
.
1426184
.
38.49
13.86
9.83
2.33
13.61
4.03
11.65
m
118.33
4.58
_
u
n
5.89
744.73
97.40
,
2303.87
467.23
2488!lO
_
9438!50
m
420.*20
180.90
118.00
54.90
152.50
.
B
.
789.94
23.69
\
m
i2i68
3593.80
510.60
1184130
262.08
(D
>
N>
-------
o
(D
H
H
I
•a
(D
>
Unit Sample
ID ID
78 20
13
15
21
18
10
07
08
09
19
17
22
05
11
01
02
06
03
04
79 20
14
16
18
05
01
02
06
03
04
13
15
12
19
10
11
21
07
08
09
22
Room or
Yard
Location
FRO
FRO
FRO
FRO
FRO
KIT
KIT
KIT
KIT
LFT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
LVG
BAG
BAG
BAG
BAG
BD1
BD1
BD1
BD1
BD1
BD1
FRO
FRO
KIT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
RGT
General
Sample
Location
1
6
6
3
3
1
2
2
2
4
N/A
3
6
1
1
N/A
2
2
1
5
5
3
3
1
1
N/A
2
2
6
6
5
2
3
6
6
1
2
2
N/A
Lead
Date Concentration
Compone
BDY
EUO
EUY
FDN
FDN
ARD
FLR
UCH
UST
BDY
FDN
N/A
ARD
EUI
FLR
FLR
N/A
UCH
UST
BDY
EUO
EUY
FDN
ARD
FLR
FLR
N/A
UCH
UST
EUO
EUY
EUI
BDY
ARD
EUI
EUY
FLR
UCH
UST
N/A
nt Sample
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Dust -
Dust -
Dust •
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Dust -
Dust -
Soil
Dust -
Dust -
Dust -
Soil
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Blank
Collected
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
03/11/92
(ug/g>
46.01
98.96
37.69
70.36
66.29
1321.97
66.41
314.55
209.93
70.35
67.26
.
217.78
69.82
44.16
56.54
143.09
160.25
98.67
55.88
16335.45
4.55
11.00
1476.59
217.68
125.31
386.09
6852.68
1849.83
300.02
514.93
2723.16
116.40
1398.66
318.28
459.07
134.28
10323.76
1586.51
.
Lead
Loading
(ug/ft2)
.
86e!oi
m
.
.
359.49
19.11
12877.92
30.13
.
m
m
5.43
235.28
33.09
57.94
.
867.96
11.31
3608!50
m
m
3059.73
225.95
289.23
.
2357.76
341.85
355.91
.
7349.00
^
2650.52
2381.40
i
18.32
13867.20
283.29
.
Dust
Loading Below
(ug/ft2) Detection
.
8771 !«0
_
.
^
271.94
287.80
40941 .36
143.54
m
B
<
24.92
3369.75
749.40
1024.70
. <
5416.20
114.62
22o!90
. <
m
2072.16
1038.00
2308.10
. <
344.06
184.80
1186.30
.
2698.70
m
1895.04
7482.00
e
136.40
1343.23
178.56
<
-------
Unit Sample
ID ID
80
O
(D
D)
(D
>
u
Room or General
Yard Sample
Location Location Component Sample Medium Sample Type
Lead Lead
Date Concentration Loading
Collected (ug/g) (ug/ft2)
Dust
Loading Below
(ug/ft2) Detection
20
14
16
21
18
22
05
01
02
03
04
29
30
31
10
07
06
08
09
13
IS
32
24
25
26
27
28
19
17
11
12
BAC
BAG
BAC
BAC
BAC
BAC
BAT
BAT
BAT
BAT
BAT
BD1
BD1
BD1
BD2
BD3
BD3
BD3
BD3
BD3
EXT
FRO
FRO
KIT
KIT
KIT
KIT
KIT
KIT
LFT
LFT
LVG
PAN
1
5
5
5
3
N/A
3
1
1
2
2
1
2
2
3
1
N/A
2
2
6
6
3
4
4
4
4
N/A
2
4
6
5
BOY
EWO
EUY
EUY
FDN
N/A
ARD
FLR
FLR
WCH
VST
FLR
UCH
UST
ARD
FLR
N/A
UCH
UST
EWO
EUY
ARD
FLR
FLR
FLR
FLR
N/A
BDY
FDN
EUI
EUI
Soil
Dust
Soil
Soil
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Oust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Soil
Dust
Dust
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Wipe
- Wipe
- Wipe
- Vacuum
- Vacuum
Regular
Regular
Regular
Field Side-by-Side
Regular
Field Blank
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Field Blank
Regular
Regular
Regular
Regular
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
,
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
p
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
377.27
569.03
1068.07
416.89
293.89
•
207.83
816.65
144.04
4872.21
472.75
443.88
45229.26
133.62
m
595.59
275.17
442.45
513.36
219.70
m
3395.09
417.62
373.86
1707.55
1241.52
.
m
m
248.46
659.74
5332.00
1380.58
•
170.37
•
B
•
•
123.04
33.97
24.24
6552.82
65.66
27.21
29102.05
15.43
p
29!48
85.63
m
2666.68
132.84
m
3495!25
.
304.48
1884.45
510.64
24.84
17.92
•
•
•
6512.50
186.24
299!40
.
-
•
.
592.00
41.60
168.30
1344.94
138.90
61.30
643.43
115.45
m
49!50
311.20
•
5194.59
604.64
m
1029.'50
•
814.40
1103.60
411.30
•
B
•
•
•
1221.40
134.90
-------
c7
(0
«Q
(D
Unit Sample
ID ID
81 19
14
16
17
10
07
06
08
09
13
15
01
02
24
25
26
27
28
03
04
12
20
21
18
11
22
Room or
Yard
Location
BAG
BAG
BAG
BAG
BAT
BD1
BD1
BD1
BD1
BD1
BD4
EXT
FRO
FRO
HAL
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LDY
LFT
LFT
LFT
LVG
RGT
General
Sample
Location
2
5
5
4
3
1
N/A
2
2
6
6
1
1
4
4
4
4
N/A
2
2
5
1
1
3
6
N/A
Lead
Date Concentration
Component
BOY
EWO
EUY
FDN
ARD
FLR
N/A
WCH
UST
EWO
EUY
FLR
FLR
FLR
FLR
FLR
FLR
N/A
UCH
WST
EUI
BOY
BDY
FDN
EUI
N/A
Sample
Soil
Dust -
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Dust -
Soil
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Field Blank
Collected
03/23/92
03/23/92
03/23/92
03/23/92
m
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
m
03/23/92
03/23/92
m
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
03/23/92
(ug/g)
447.05
250.23
852.78
3351.12
>
153!26
88.17
-146.75
3975.77
2378.37
e
206I&4
177.61
m
424.39
566.41
168.14
172.72
.
.
a
12816.04
2444.70
363.85
235.15
278.95
329.85
225.05
.
Lead
Loading
(ug/ft2)
149&!36
m
.
9J02
45.62
_
10809.87
189.90
m
165.60
.
m
6.66
7.82
3.08
3.92
26.22
60.12
m
16534.40
830.87
5501.00
_
.
B
1298.00
.
Dust
Loading
(ug/ft2)
5987)89
m
ss'.sa
517.40
_
2718.93
79.84
>
800'.60
.
m
IS!TO
13.80
18.30
22.70
m
m
^
1290.13
339.86
15118.90
.
.
m
5767.70
.
Below
Detection
<
<
<
U)
Ul
-------
c
m
•0
pi
(D
>
CJ
0\
Unit Sample
ID ID
84 14
16
10
07
08
09
12
19
21
13
15
17
22
24
25
26
27
28
05
11
01
02
06
03
04
20
18
88 19
13
15
17
22
10
11
07
08
09
21
05
01
02
06
03
04
20
18
Room or
Yard
Location
BAC
BAC
BAT
BAT
BAT
BAT
DIN
EXT
FRO
FRO
FRO
FRO
FRO
FRO
KIT
KIT
KIT
KIT
KIT
LDY
LVG
LVG
LVG
LVG
LVG
LVG
LVG
RGT
RGT
FRO
FRO
FRO
FRO
FRO
KIT
KIT
KIT
KIT
KIT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
RGT
RGT
General
Sample
Location
5
5
3
1
2
2
5
2
2
6
6
4
N/A
4
4
4
4
N/A
3
6
1
1
N/A
2
2
1
3
2
6
6
4
N/A
3
6
1
2"
2
3
1
1
N/A
2
2
1
3
Lead
Date Concentration
Component
EWO
EWY
ARD
FLR
WCH
WST
EWI
BDY
BDY
EWO
EWY
FDN
N/A
FLR
FLR
FLR
FLR
N/A
ARD
EWI
FLR
FLR
N/A
WCH
WST
BDY
FDN
BDY
EWO
EWY
FDN
N/A
ARD
EWI
FLR
WCH
WST
FDN
ARD
FLR
FLR
N/A
WCH
WST
BDY
FDN
Sample
Dust •
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Dust •
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust •
Soil
Soil
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Hediun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuum
Vacuun
Vacuun
Vacuun
Vacuun
Wipe
Wipe
Wipe
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuun
Vacuum
Vacuum
Vacuun
Vacuun
Vacuum
Vacuun
Vacuun
Vacuum
Vacuum
Vacuun
Vacuun
Vacuun
Sample Type
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Field Blank
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Field Blank
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Collected
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
04/02/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
03/26/92
(ug/g)
1188.48
151.36
505.95
134.09
88.36
140.02
494.04
259!?3
304.16
454.65
184.92
436.67
.
126.87
195.40
m
.
,
252! 38
125 !03
609.36
769.94
797.77
438.77
132.31
75.99
698.17
26.72
495.94
19.70
117.86
m
437.87
118.51
36.97
2107.94
216.19
42.30
323.63
38.40
37.57
268.91
346.36
98.21
21.35
45.39
Lead
Loading
(ug/ft2)
1420.95
p
24! 00
14.15
65.89
36.89
1162.13
.
698.53
,
_
169!33
224.95
8.45
25.78
m
179J96
4!33
458.85
684.55
B
2092 !o4
158.48
.
3908!20
_
,
1122.84
23.48
5.24
2491.59
29.92
m
89.05
18.22
9.52
5807!92
14.90
.
Dust
Loading Below
(ug/ft2) Detection
1195.60
4?!44
105.50
745.69
263.49
2352.30
1536! 40
<
1334 !60
1151.20
m
u
<
7is!o5
34!60
753.00
889.10
4769! 32 <
1197.79 <
.
7880 ! 40
! <
2564! 35
198.10
141.60
1182.00
138.41
.
275.15
474.40
253.30
1 6768 ! 23
151.68
m
-------
o
M
I
(D
•U
01
«Q
(D
Component Sample Medium Sample Type
Lead Lead
Date Concentration Loading
Collected (ug/g) (ug/ft2)
Dust
Loading Below
(ug/ft2) Detection
32
29
30
31
19
21
13
15
17
11
07
OS
12
01
02
24
25
26
27
06
28
03
04
22
20
U
16
18
BAT
BD1
BSM
DIN
DIN
DIN
DIN
EXT
FRO
FRO
FRO
FRO
FRO
HAL
HAL
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LDY
LFT
LVG
RGT
RGT
RGT
RGT
3
1
2
2
2
2
6
6
4
6
1
3
5
1
1
4
4
4
4
N/A
N/A
2
2
N/A
1
5
5
3
ARD
FLR
UCH
UST
BOY
BDY
EWO
EWY
FDN
EUI
FLR
ARD
EUI
FLR
FLR
FLR
FLR
FLR
'FLR
N/A
N/A
UCH
UST
N/A
BDY
EWO
EUY
FDN
Dust
Dust
Dust
Dust
Soil
Soil
Dust
Soil
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Soil
Dust
Soil
Soil
- Vacuum
- Vacuum
- Vacuum
- Vacuum
• Vacuum
- Vacuum
• Vacuum
- Vacuum
• Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
• Uipe
- Uipe
- Vacuum
- Uipe
- Vacuum
- Vacuum
- Vacuum
Regular
Regular
Regular
Regular
Regular
Field Sfde-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Vacuun Uipe Comparison
Vacuum Uipe Comparison
Vacuun Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Field Blank
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
•
•
03/16/92
03/16/92
03/16/92
03/16/92
•
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
•
03/16/92
•
03/16/92
03/16/92
03/16/92
03/16/92
•
•
•
247.05
117.11
459.77
408.98
•
58.96
74.12
1460.46
427.58
348.80
287.46
220.85
885.97
142.02
230.55
388.52
134.89
154.17
•
.
-334.57
•
189.85
205.62
•
•
•
235.58
426.07
188.22
292.00
•
•
397! 09
274.60
3254.40
63.24
m
.
.
6316.50
•
•
1304.50
337.93
478.00
25.45
193.53
58.98
680.95
448.34
75.75
63.52
•
.
501.73
209.31
9
•
•
•
1222.89
•
•
»
1607)31
2344.80
7078.25
154.63
m
m
•
4325.00
•
m
4538.00
1530.10
539.52
179.20
839.40
151.80
5048.20
2908.00
•
•
•
•
2642.78
1017.94
m
m
•
•
2870.16
.
•
-------
o
n>
•o
0)
vQ
(D
Unit Sample
ID ID
93 19
14
16
17
21
13
15
10
12
07
24
25
26
27
28
08
09
05
11
01
02
06
03
04
20
18
22
Room or
Yard
Location
BAG
BAG
BAG
BAG
BAG
EXT
FRO
FRO
HAL
HAL
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LVG
LVG
LVG
LVG
LVG
LVG
LVG
RGT
RGT
RGT
General
Sample
Location
2
5
5
4
4
6
6
3
5
1
4
4
4
4
N/A
2
2
3
6
1
1
N/A
2
2
1
3
N/A
Lead
Date Concentration
Component
BDY
EUO
EUY
FDN
FDN
EUO
EUY
ARD
EUI
FLR
FLR
FLR
FLR
FLR
N/A
UCH
UST
ARD
EUI
FLR
FLR
N/A
UCH
UST
BDY
FDN
N/A
Sample
Soil
Dust -
Soil
Soil
Soil
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Regular
Regular
Regular
Regular
Field Blank
Collected
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
04/01/92
(ug/g)
98.88
1303.25
59.31
67.18
79.97
135 !94
126.25
86.53
149.74
74.71
202.26
92.71
m
m
m
4067.58
2218.35
1456.63
107.97
64.76
74.74
548.52
8785.71
248.07
114.64
223.86
.
Lead
Loading
(ug/ft2)
3381 '. 40
!
a
18s!42
31J71
611.54
12.90
3.84
9.61
5.24
3.00
.
39410.37
1315.08
28.93
177.62
16.59
11.17
m
18883! 56
19.40
.
t
m
DUSt
Loading Below
(ug/ft2) Detection
2594 !&0
|
.
1386.00
366 ! 48
4083 !°0
172.70
19.00
103.70
m
<
<
9688.89
592.82
19.86
1645.00
256.20
149.40
<
214?!35
78.20
m
<
09
-------
cf
•0
D>
w
10
Unit Sample
ID ID
94 14
16
32
29
30
31
19
13
15
17
22
12
05
01
02
24
25
26
27
28
03
04
10
11
07
06
08
09
20
18
21
Room or
Yard
Location
BAC
BAC
BAT
BD1
BD1
BD1
BD1
BD2
EXT
FRO
FRO
FRO
FRO
FRO
HAL
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
KIT
LVG
LVG
LVG
LVG
LVG
LVG
RGT
RGT
RGT
General
Sample
Location
5
5
3
1
2
2
2
6
6
4
N/A
5
3
1
1
4
4
4
4
N/A
2
2
3
6
1
N/A
2
2
1
3
3
Lead
Date Concentration
Component
EWO
EUY
ARD
FLR
UCH
UST
BOY
EWO
EUY
FDN
N/A
EUI
ARD
FLR
FLR
FLR
FLR
FLR
FLR
N/A
UCH
UST
ARD
EUI
FLR
N/A
UCH
UST
BOY
FDN
FDN
Sample
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Oust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Medium
Vacuum
Vacuum
Vacuun
Vacuun
Vacuun
Vacuun
Vacuum
Vacuum
Vacuum
Vacuun
Vacuun
Vacuum
Uipe
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Field Side-by-Side
Vacuun Uipe Comparison
Vacuun Uipe Comparison
Vacuun Uipe Comparison
Vacuun Uipe Comparison
Field Blank
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Field Side-by-Side
Collected
03/18/92
03/18/92
m
03/18/92
03/18/92
03/18/92
03/18/92
m
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
03/18/92
(ug/g)
509.91
216.69
m
317.64
308.08
344.51
35.79
m
lisiffi
2348.06
188.12
224.53
m
21.67
327.26
24.12
96.74
31.53
144.27
.
.
.
3984.00
1781.19
84.97
282.24
182.08
563.87
1301.04
416.25
107.11
339.90
307.04
Lead
Loading
(ug/ft2)
7408.40
.
m
40863.60
85.62
34.90
4.10
B
m
14021.00
m
m
m
19.91
31953.60
5.31
35.64
38.61
18.96
28.61
25.33
m
19.12
318.50
3471.30
836.45
49.67
m
2158.85
22.21
B
.
.
Dust
Loading Below
(ug/ft2) Detection
14528.90
.
>
128646.00
277.90
101.31
114.67 <
m
t
5971 .30
t
^
• *
918.70
97639.20
220.20
368.40
1224.50
131.40
.
.
m
4.80
178.81
40852.80
2963.60
272.80
• *
1659.32
53.35
w
e
.
-------
(D
Unit Sample
ID ID
95 19
21
14
16
17
13
15
05
12
01
02
03
04
20
18
22
10
11
07
06
08
09
Room or
Yard
Location
BAC
BAC
BAC
BAC
BAC
FRO
FRO
KIT
KIT
KIT
KIT
KIT
KIT
LFT
LFT
LFT
LVG
LVG
LVG
LVG
LVG
LVG
General
Sample
Location
2
2
5
5
4
6
6
3
5
1
1
2
2
1
3
N/A
3
6
1
N/A
2
2
Lead
Date Concentration
Component
BDY
BDY
EWO
EUY
FDN
EWO
EUY
ARD
EUI
FLR
FLR
UCH
UST
BDY
FDN
N/A
ARD
EUI
FLR
N/A
UCH
UST
Sample
Soil
Soil
Dust -
Soil
Soil
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Medium
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Field Blank
Regular
Regular
Collected
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
03/12/92
(ug/g)
77.29
95.12
79.81
47.37
43.50
145.64
83.05
201.88
26.17
854.14
70.58
591.53
342.78
41.66
46.17
m
263!90
132.56
35.82
435.72
139.15
3114.95
Lead
Loading
(ug/ft2)
.
22.13
83.44
m
2)08
5.33
5.21
0.92
544.67
8.77
m
f
s!si
314.95
8.16
m
84il6
231.23
Dust
Loading Below
(ug/ft2) Detection
|
277!SO
572.90
13.27
203i60
6.10
13.10 <
920.78
25.57
<
2o! 87
2375 !90
227.90
604 '.BO
74.23
I
•0
P>
ua
(D
f*
o
-------
(D
(D
>
4k
Room or
Unit Sample Yard
ID ID
96 05
24
25
26
27
04
10
07
33
08
09
13
15
01
02
06
28
21
19
17
22
11
32
12
29
30
31
20
14
16
18
Location
BAT
BAT
BAT
BAT
BAT
BAT
BD1
BD2
BD2
8D2
BD2
BD2
BSN
EXT
FRO
FRO
HAL
HAL
HAL
HAL
KIT
LFT
LFT
LFT
LFT
LVG
P02
P02
P02
P02
P02
POR
RGT
RGT
RGT
RGT
General
Sample
Location
3
4
4
4
4
2
3
1
2
2
6
6
1
1
N/A
N/A
2
4
N/A
6
3
5
1
2
2
1
5
5
3
Lead
Date Concentration
Component
ARD
FLR
FLR
FLR
FLR
UST
ARD
FLR
UCH
UCH
UST
EUO
EUY
FLR
FLR
N/A
N/A
BDY
BDY
FDN
N/A
EUI
ARD
EUI
FLR
UCH
UST
BDY
EUO
EUY
FDN
Sample
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Dust -
Dust -
Dust -
Soil
Soil
Soil
Soil
Dust -
Dust -
Dust -
Dust -
Dust -
Dust -
Soil
Dust -
Soil
Soil
Medium
Vacuum
Vacuum
Vacuum
Uipe
Uipe
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Uipe
Vacuum
Vacuum
Vacuirn
Vacuum
Vacuum
Vacuum
Vacuum
Sample Type
Regular
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Vacuum Uipe Comparison
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Field Blank
Field Blank
Field Side-by-Side
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Collected
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
a
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
e
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
03/13/92
m
03/13/92
03/13/92
03/13/92
03/13/92
(ug/g)
331.57
167.13
1387.82
.
.
m
434J03
385.32
31595.17
28337.07
2923.75
m
151 .es
54.23
6217.62
1724.32
-122.15
.
m
76.66
159.24
91.00
.
214.43
217.21
124.48
92.58
828.73
162.17
m
79.09
35.65
190.28
82.44
Lead
Loading
(ug/ft2)
57.37
3.71
408.58
7.22
15.02
a
190.57
290.23
46843.00
17061.75
636.87
m
155^27
.
11641.25
2365.43
^
.
e
^
m
m
m
3522.60
8.90
265.34
48.54
11449.67
16.16
m
\
23.40
m
.
Oust
Loading Below
(ug/ft2) Detection
173.02
22.20
294.41
,
.
m
439! 07
753.20
1482.60
602.10
217.83
m
1022!50
.
1872.30
1371.80
<
<
|
a
m
<
16428!lO
40.96
2131.50
524.30
13815.96
99.68
|
656i30
e
.
97
EXT
-------
Unit
ID
99
O
M
c
3
(D
Q>
«Q
(D
>
*k
IS)
Room or General
Sample Yard Sample
ID Location Location Component
Sample Hediun Sample Type
Lead Lead
Date Concentration Loading
Collected (ug/g) (ug/ft2)
Dust
Loading Below
(ug/ftZ) Detection
H
16
10
07
08
09
OS
01
02
24
25
26
27
06
28
03
04
19
13
15
17
22
11
12
20
18
21
BAC
BAC
BAT
BD1
BD2
BD2
BD2
8D2
BD3
BSH
DIN
DIN
DIN
DIN
DIN
DIN
DIN
DIN
DIN
DIN
DIN
EXT
FRO
FRO
FRO
FRO
FRO
HAL
KIT
LDY
LVG
RGT
RGT
RGT
5
5
3
1
2
2
3
1
1
4
4
4
4
N/A
N/A
2
2
2
6
6
4
N/A
6'
5
1
3
3
EUO
EWY
ARD
FLR
WCH
WST
ARD
FLR
FLR
FLR
FLR
FLR
FLR
N/A
N/A
WCH
UST
BDY
EUO
EUY
FDN
N/A
EUI
EUI
BDY
FDN
FDN
Dust
Soil
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Dust
Soil
Dust
Soil
Soil
Soil
Dust
Dust
Soil
Soil
Soil
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Wipe
- Wipe
- Vacuum
- Wipe
- Vacuum
- Vacuum
- Vacuum
- Vacuum
- Vacuum
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Regular
Field Side-by-Side
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Vacuum Wipe Comparison
Field Blank
Field Blank
Regular
Regular
Regular
Regular
Regular
Regular
Field Blank
Regular
Regular
Regular
Regular
Field Side-by-Side
03/16/92
03/16/92
m
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
f
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
03/16/92
m
03/16/92
m
03/16/92
03/16/92
03/16/92
163.59
148.14
2988isi
140.87
422.53
.
1645J10
347.13
627.17
54.08
94.50
m
•
-207.89
m
3061.29
5917.30
521 !38
197.55
212.56
197.97
.
310.23
e
194!44
m
240 ! 05
210.37
229.23
203.13
B
412!s7
32.25
510.51
.
ZuizS
204.53
1071.78
485.09
63.79
31.76
19.22
f
u
26435.50
16710.45
848 ! 52
m
.
_
326o!60
379! 04
j
.
.
1241.70
•
i3a!is
228.90
1208.23
•
13o!25
589.20
1708.90
8969.80
675.00
•
a
•
8635 ! 40
2824.00
4295 ! 20
|
m
1051o!30
1949! 40
•
•
•
-------
APPENDIX B
ADDITIONAL EXPLANATORY VARIABLES CONSIDERED
FOR INCLUSION IN THE STATISTICAL MODELS
-------
APPENDIX B
Factors considered for inclusion in the models.
Abatement
Proportion of interior abated by encapsulation, as
opposed to enclosure
Proportion of exterior abated by encapsulation, as
opposed to enclosure
Proportion of room abated by encapsulation, as opposed
to enclosure
Proportion of interior abated by each of four specific
removal methods
Proportion of exterior abated by each of four specific
removal methods
Proportion of room abated by each of four specific
removal methods
Phase of abatement
Abatement contractor
Mean of Log XRF measures obtained at location during HUD
Demonstration
Remodeling
During six months prior to sampling
During sampling
Number of residents over 18 years
Number of residents between 7 and 17 years of age
Number of residents under age 7
Number of pets
Number of months at residence
Indicator of ownership
Volume II - Page B-l
-------
Occupation
Paint removal Auto body repair
Building demolition Salvage
Home remodeling Chemical plant
Welding Glass work
Plumbing Lead smelter
Sand blasting oil refinery
Indicator of work clothes worn home
Indicator of work clothes washed at home
Activities (at home)
Car or bicycle painting
Paint removal
Soldering pipes or electronic parts
Soldering jewelry
Use of artists' paint
Car maintenance
Cleanliness
Frequency of vacuuming carpets
Frequency of vacuuming uncarpeted floors
Frequency of sweeping uncarpeted floors
Frequency of wet mopping uncarpeted floors
Frequency of furniture dusting
Frequency of window stool dusting
Frequency of window stool dusting
Frequency of pets scratching the carpet or furniture
Frequency of pets chewing the carpet or furniture
Sampling Deviations
Small nozzle used instead of large nozzle
Sample collected from outside of air duct instead of
inside
Substrate type
Substrate condition
Proximity of house to lead smelter
Age of house
Volume II - Page B-2
-------
APPENDIX C
CONCISE LISTING OF STATISTICAL MODELING RESULTS
BY COMPONENT AND MEASUREMENT TYPE
-------
Mixed Model Results by Sample Type
Sample
Type
ARD
ARD
ARD
Denominator for
Degrees of Multiplicative Std
Response Parameter Freedom Effect Estimate Error
CONC INTERCEPT
Observations
STUDVID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
Condition Effect
Condition: Damaged
Condition: Good
Condition: Peeling
Contractor Effect
Contractor: A
Contractor: B
Contractor: C
Contractor: D
DUSTS ILL
EXTAIRDT
UNOPUNCP
DUST INTERCEPT
Observations
STUDVID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
Condition Effect
Condition: Damaged
Condition: Good
Condition: Peeling
Contractor Effect
Contractor: A
Contractor: B
Contractor: C
Contractor: D
DUSTSILL
EXTAIRDT
UNOPUNCP
LOAD INTERCEPT
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
Condition Effect
Condition: Damaged
Condition: Good
Condition: Peeling
Contractor Effect
Contractor: A
Contractor: B
Contractor: C
Contractor: D
DUSTSILL
EXTAIRDT
UNOPUNCP
35.00
86.00
•
.
35.00
35.00
35.00
35.00
35.00
•
35.00
35.00
35.00
35.00
35.00
35.00
35.00
35.00
35.00
35.00
86.00
•
B
35.00
35.00
35.00
35.00
35.00
m
35.00
35.00
35.00
35.00
35.00
35.00
35.00
35.00
35.00
35.00
86.00
.
•
35.00
35.00
35.00
35.00
35.00
.
35.00
35.00
35.00
35.00
35.00
35.00
35.00
35.00
35.00
331.91
•
•
2.21
1.59
2.01
0.79
•
1.52
1.00
6.71
•
2.34
0.77
0.91
1.81
1.03
0.78
0.98
202.46
•
m
2.73
3.11
1.80
0.91
•
28.44
1.00
2.54
•
0.24
1.36
0.83
1.87
0.96
0.26
0.98
75.81
•
•
3.24
4.70
3.99
0.73
•
40.56
1.00
27.77
B
0.55
1.01
0.78
3.35
0.99
0.18
0.97
5.80
•
•
0.79
0.46
0.70
-0.24
•
0.42
0.00
1.90
•
0.85
-0.26
-0.10
0.59
0.03
-0.24
-0.02
5.31
•
1.45
1.00
1.14
0.59
-0.10
•
3.35
0.00
0.93
•
-1.44
0.31
-0.19
0.63
-0.04
-1.35
-0.02
4.33
.
1.52
1.18
1.55
1.38
-0.31
•
3.70
0.00
3.32
.
-0.61
0.01
-0.25
1.21
-0.01
•1.74
-0.03
0.19
•
•
m
0.23
0.24
0.23
•
0.83
•
0.50
•
0.35
0.13
0.15
0.28
0.01
0.19
0.01
0.48
•
0.79
•
0.57
0.63
0.34
•
1.34
.
1.07
m
0.96
0.36
0.40
0.79
0.04
0.33
0.02
0.52
.
0.86
•
0.61
0.68
0.39
•
1.55
•
1.21
•
1.03
0.38
0.43
0.84
0.04
0.38
0.02
Observed
Significance
Level
.0000
•
m
m
.0488
.0057
.3012
.0022
.6139
m
.0005
.0155
.0220
.0537
.5024
.0407
.0602
.1992
.0394
.0000
m
.0008
•
.0530
.3557
.7768
.0460
.0176
m
.3911
.3559
.1411
.3961
.6350
.4324
.3625
.0003
.4566
.0000
•
.0019
.
.0165
.0490
.4318
.0047
.0224
!0092
.5036
.5609
.9736
.5689
.1617
.7761
.0001
.1869
Volume II - Page C-l
-------
Mixed Model Results by Sample Type (Contd)
Sample
Type
BOY
EUI
EUI
EUI
Denominator for
Degrees of Multiplicative std
Response Parameter Freedom Effect Estimate Error
CONC INTERCEPT
LOCATION(STUDYID)
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
POD
HUDYRBLT
PNTREMJB
SLDRPIPE
CONC INTERCEPT
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
NUN7T017
REHPAINT
Substrate Effect
Substrate: Carpet
Substrate: Linoleum
Substrate: Plastic
Substrate: Tile
Substrate: Uood
VACMUNCP
DUST INTERCEPT
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
NUM7T017
REMPAINT
Substrate Effect
Substrate: Carpet
Substrate: Linoleum
Substrate: Plastic
Substrate: Tile
Substrate: Uood
VACMUNCP
LOAD INTERCEPT
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
NUM7T017
RENPAINT
Substrate Effect
Substrate: Carpet
Substrate: Linoleum
Substrate: Plastic
Substrate: Tile
Substrate: Uood
VACMUNCP
20.00
B
120.00
B
m
20.00
20.00
20.00
20.00
20.00
34.00
90.00
.
.
34.00
34.00
34.00
34.00
34.00
34.00
.
34.00
34.00
34.00
34.00
34.00
34.00
90.00
•
•
34.00
34.00
34.00
34.00
34.00
34.00
•
34.00
34.00
34.00
34.00
34.00
34.00
90.00
m
m
34.00
34.00
34.00
34.00
34.00
34.00
m
34.00
34.00
34.00
34.00
34.00
108.62
B
m
m
1.23
1.18
1.27
0.84
0.45
1.46
181.71
.
m
2.31
0.85
0.95
1.28
0.81
1.10
.
1.00
0.94
0.76
1.08
1.34
1.06
3052.85
B
.
2.88
1.19
1.24
1.31
0.78
0.97
.
1.00
0.15
0.01
0.03
0.27
0.99
531.48
B
.
4.38
1.05
1.15
1.63
0.64
1.06
•
1.00
0.15
0.01
0.03
0.35
1.06
4.69
0.44
m
0.38
0.21
0.16
0.24
-0.18
-0.81
0.38
5.20
m
0.49
0.84
-0.16
-0.05
0.25
-0.21
0.10
.
0.00
-0.06
-0.27
0.07
0.29
0.06
8.02
a
•
1.06
0.17
0.22
0.27
-0.24
-0.03
m
0.00
-1.90
-5.05
-3.65
-1.33
-0.01
6.28
•
•
1.48
0.05
0.14
0.49
-0.45
0.06
m
0.00
-1.87
-5.18
-3.53
-1.06
0.06
0.14
0.22
m
0.25
m
0.17
0.18
0.04
0.36
0.15
0.24
m
0.41
m
0.27
0.31
0.26
0.11
0.05
m
m
0.23
0.69
0.41
0.37
0.02
0.25
.
.
0.28
0.31
0.29
0.11
0.05
m
m
0.27
0.77
0.45
0.42
0.02
0.35
.
B
0.38
0.44
0.41
0.15
0.07
•
m
0.38
1.08
0.62
0.58
0.02
Observed
Significance
Level
.0000
.0001
f
.0146
m
.3469
.2118
.0002
.0376
.0199
.0000
m
.1537
.
.5612
.8764
.3410
.0602
.0675
.9001
t
.7838
.6927
.8579
.4377
.0013
.0000
_
.
!S385
.4921
.3642
.0304
.5224
.0000
m
.0000
.0000
.0000
.0030
.7614
.0000
•
m
.9021
.7537
.2439
.0052
.3840
.0000
.
.0000
.0000
.0000
.0766
.0230
Volume II - Page C-2
-------
Mixed Model Results by Sample Type (Contd)
Sample
Type
EUO
EUO
EWO
EUO
EUY
FDN
FLR
Denominator for
Degrees of Multiplicative Std
Response Parameter Freedom Effect Estimate Error
CONC INTERCEPT
Observations
STUDY ID
Standard Deviation Estimate
Avg Abatement Effect
POD
VACNUNCP
DUST INTERCEPT
Observations
STUDY ID
Standard Deviation Estimate
Avg Abatement Effect
POD
VACNUNCP
LOAD INTERCEPT
Observations
STUDYID
Standard Deviation Estimate
LOAD Avg Abatement Effect
POD
VACMUNCP
CONC INTERCEPT
LOCATION* STUDYID)
Observat i ons
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
POD
PROXIHTY
CONC INTERCEPT
LOCATION(STUDYID)
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
POO
SO
NUDYRBLT
LOG XRF
NOUNER
NUMMONTH
REHPAINT
UELDJB
CONC INTERCEPT
LOCATION(STUDYID)
Observat i ons
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
SI
Substrate Effect
Substrate: Carpet
Substrate: Concrete
Substrate: Linoleum
Substrate: Tile
Substrate: Uood
VACNUNCP
WELDJB
46.00
97.00
.
,
46.00
46.00
46.00
46.00
97.00
•
•
46.00
46.00
46.00
46.00
97.00
•
•
46.00
46.00
46.00
12.00
.
109.00
m
t
12.00
12.00
12.00
14.00
m
88.00
,
.
14.00
14.00
14.00
14.00
14.00
14.00
14.00
14.00
14.00
105.00
•
233.00
.
>
105.00
105.00
105.00
105.00
105.00
,
105.00
105.00
105.00
105.00
105.00
105.00
183.84
m
.
2.44
1.19
1.00
1.05
1151.37
m
m
3.30
1.95
1.05
0.96
219.99
m
m
4.35
2.24
1.06
1.00
110.66
•
•
m
1.49
1.18
1.18
1.37
194.49
^
•
B
1.31
1.17
0.72
0.65
0.73
1.28
0.34
0.94
0.84
2.09
99.15
.
.
m
2.16
1.07
1.20
0.84
1.14
m
1.00
3.56
1.06
1.15
2.70
1.03
3.94
5.21
.
0.52
0.89
0.17
0.00
0.04
7.05
B
0.40
1.19
0.67
0.05
-0.04
5.39
m
0.91
1.47
0.81
0.06
0.00
4.71
0.71
.
0.37
0.40
0.17
0.16
0.31
5.27
0.44
.
0.21
0.27
0.16
-0.33
-0.43
-0.31
0.24
-1.09
-0.07
-0.18
0.74
4.60
0.73
_
oioo
0.77
0.07
0.18
-0.17
0.13
m
0.00
1.27
0.06
0.14
0.99
0.03
1.37
0.22
m
0.41
.
0.26
0.29
0.02
0.25
m
0.50
m
0.30
0.33
0.02
0.37
m
0.69
•
0.44
0.50
0.03
0.18
0.38
.
0.35
•
0.22
0.24
0.11
0.15
0.26
.
0.24
m
0.21
0.33
0.12
0.05
0.09
0.18
0.02
0.03
0.36
0.19
0.35
_
•
_
0.22
0.23
0.22
0.06
m
m
0.69
0.20
0.35
0.27
0.01
0.38
Observed
Significance
Level
.0000
•
.0969
.
.5090
.9943
.0142
.0000
a
.5230
a
.0291
.8760
.0317
.0000
.
.0762
•
.0710
.9072
.9967
.0000
.0005
.
.2666
.
.4578
.5175
.0169
.0000
.0034
•
.4375
t
.4584
.3350
.0029
.0000
.0146
.0000
.0014
.0002
.0619
.0000
.0000
.
•
.7614
.4218
.4375
.0311
.0024
•
.0687
.7818
.6898
.0003
.0070
.0005
Volume II - Page C-3
-------
Mixed Model Results by Sample Type (Contd)
Sample
Type
FLR
FLR
FLR
FLU
UCH
Denominator for
Degrees of Multiplicative Std
Response Parameter Freedom Effect Estimate Error
DUST INTERCEPT
LOCATION(STUDTID)
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
P1D
PR
SI
Substrate Effect
Substrate: Carpet
Substrate: Concrete
Substrate: Linoleum
Substrate: Tile
Substrate: Wood
VACHUNCP
UELDJB
LOAD INTERCEPT
LOCATION(STUDYID)
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
PID
LOAD PR
SI
Substrate Effect
Substrate: Carpet
Substrate: Concrete
Substrate: Linoleum
Substrate: Tile
Substrate: Wood
VACNUNCP
UELDJB
LOAD Observations
STUDYID
Standard Deviation Estimate
PID
POD
Contractor Effect
Contractor: A
Contractor: B
Contractor: C
Contractor: D
Intercept
N OWNER
CONC INTERCEPT
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
SI
SO
Condition Effect
105.00
•
233.00
•
•
105.00
105.00
105.00
105.00
105.00
.
105.00
105.00
105.00
105.00
105.00
105.00
105.00
.
233.00
m
,
105.00
105.00
105.00
105.00
105.00
.
105.00
105.00
105.00
105.00
105.00
105.00
65.00
.
.
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
32.00
29.00
83.00
,
m
29.00
29.00
29.00
29.00
29.00
29.00
348.05
•
•
.
2.36
1.69
1.52
0.67
0.91
m
1.00
4.24
0.14
0.11
0.57
0.99
2.86
35.25
•
m
.
2.56
1.83
1.87
0.57
1.03
m
1.00
13.76
0.14
0.12
1.45
1.02
10.92
m
1.76
1.48
0.44
.
4.70
0.68
1.58
0.12
18.21
0.38
1571.83
m
m
3.07
0.81
1.86
0.66
1.31
0.59
.
5.85
0.84
a
0.49
0.86
0.52
0.42
•0.40
-0.10
•
0.00
1.45
•1.98
-2.24
-0.55
-0.01
1.05
3.56
1.27
•
•
0.94
0.60
0.63
-0.56
0.03
•
0.00
2.62
-1.93
-2.10
0.37
0.02
2.39
o!73
0.56
0.39
-0.82
•
1.55
•0.38
0.46
-2.11
2.90
-0.97
7.36
.
0.84
1.12
-0.21
0.62
-0.42
0.27
-0.53
.
0.26
0.45
•
0.44
•
0.29
0.32
0.25
0.08
^
.
0.90
0.24
0.43
0.34
0.02
0.53
0.30
0.53
.
t
0.34
0.36
0.33
0.09
•
•
1.08
0.32
0.55
0.41
0.02
0.58
o!45
.
0.38
0.38
•
0.51
0.19
0.19
0.45
0.21
0.34
0.52
_
o!62
m
o!55
0.63
0.36
0.15
0.18
.
Observed
Significance
Level
.0000
.0005
•
.2224
.
.0786
.1891
.1169
.2325
.0000
.
.1110
.0000
.0000
.1107
.5604
.0507
.0000
.0000
.
.
.0826
.0811
.0971
.7426
.0000
.
.0169
.0000
.0002
.3694
.2495
.0001
!0080
13083
.0381
.0012
.0048
.0546
.0250
.0001
.0000
.0074
.0000
!o648
17065
.3383
.2508
.0778
.0076
.0865
Volume II - Page C-4
-------
Mixed Model Results by Sample Type (Contd)
Sample
Type Response
UCH CONG
Parameter
Condition: Chalking
Condition: Good
Condition: Peeling
Int Removal Effect
Int Removal: Chan Strip
Int Removal: Heat Gun
Int Removal: Rem/Repl
NOZZLDEV
Substrate Effect
Substrate: Concrete
Substrate: Metal
Substrate: Plastic
Substrate: Wood
Denominator for
Degrees of
Freedom
29.00
m
29.00
29.00
29.00
29.00
29.00
29.00
29.00
29.00
29.00
29.00
•
Multiplicative
Effect
4.07
1.00
2.35
•
1.05
1.33
1.03
1.18
•
0.58
0.41
0.23
1.00
Estimate
1.40
0.00
0.86
m
0.05
0.28
0.03
0.17
•
-0.55
•0.90
-1.49
0.00
Std
Error
0.99
•
0.39
•
0.13
0.17
0.16
0.45
•
1.42
0.49
0.81
•
Observed
Significance
Level
.1674
•
.0364
.0763
.7082
.0967
.8735
.7082
.1729
.7038
.0737
.0779
•
UCH DUST INTERCEPT 29.00 2450.92
Observations 83.00
STUDYID
Standard Deviation Estimate . 3.25
Avg Abatement Effect 29.00 0.90
PID 29.00 0.33
PR 29.00 0.80
SI 29.00 1.02
SO 29.00 0.83
Condition Effect 29.00
Condition: Chalking 29.00 0.71
Condition: Good . 1.00
Condition: Peeling 29.00 0.95
Int Removal Effect 29.00
Int Removal: Chem Strip 29.00 0.85
Int Removal: Heat Gun 29.00 0.79
Int Removal: Rem/Repl 29.00 0.97
NOZZLDEV 29.00 2.26
Substrate Effect 29.00
Substrate: Concrete 29.00 0.28
Substrate: Metal 29.00 0.94
Substrate: Plastic 29.00 0.13
Substrate: Uood . 1.00
UCH LOAD INTERCEPT 29.00 3908.71
Observations 83.00
STUDYID
Standard Deviation Estimate . 5.28
Avg Abatement Effect 29.00 0.73
PID 29.00 0.60
PR 29.00 0.53
SI 29.00 1.34
SO 29.00 0.49
Condition Effect 29.00
Condition: Chalking 29.00 2.84
Condition: Good . 1.00
Condition: Peeling 29.00 2.25
Int Removal Effect 29.00
Int Removal: Chem Strip 29.00 0.90
Int Removal: Heat Gun 29.00 1.06
Int Removal: Rem/Repl 29.00 1.00
NOZZLDEV 29.00 2.61
Substrate Effect 29.00
Substrate: Concrete 29.00 0.15
Substrate: Metal 29.00 0.37
Substrate: Plastic 29.00 0.03
Substrate: Uood . 1.00
7.80
8.27
0.51
0.73
o'.ST
.0000
0.72
1.18
•0.11
•1.12
•0.22
0.02
0.18
m
0.34
0.00
0.06
•
0.16
0.23
0.03
0.81
•
1.28
0.06
2.03
0.00
0.69
.
0.54
0.61
0.37
0.14
0.18
•
1.01
•
0.39
•
0.12
0.16
0.15
0.45
.
1.46
0.48
0.80
•
.2789
.
!8463
.0796
.5576
.8921
.3120
.9436
.7372
m
.8877
.1733
.2092
.1612
.8335
.0831
.0802
.3889
.8927
.0163
•
.0000
!l200
1.66
-0.32
-0.51
-0.64
0.29
-0.71
•
1.04
0.00
0.81
.
-0.11
0.05
0.00
0.96
.
-1.87
-0.98
-3.54
0.00
•
0.78
0.89
0.53
0.21
0.26
•
1.44
a
0.56
•
0.18
0.23
0.22
0.6S
.
2.08
0.69
1.15
.
•
.6822
.5706
.2313
.1669
.0097
.3460
.4741
•
.1554
.5073
.5517
.8147
.9825
.1496
.0294
.3748
.1620
.0044
•
Volume II - Page C-5
-------
Mixed Model Results by Sample Type (Contd)
Denominator for
Sample Degrees of Multiplicative Std
Type Response Parameter Freedom Effect Estimate Error
UST CONC INTERCEPT
Observations
STUDY ID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
NUHONE
PHASE 1
PHASE 2
PHASE 3
PHASE Effect
SINCE.CL
UST DUST INTERCEPT
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
NUHOHE
PHASE 1
PHASE 2
PHASE 3
PHASE Effect
SINCE.CL
UST LOAD INTERCEPT
Observations
STUDYID
Standard Deviation Estimate
Avg Abatement Effect
PID
PR
NUHONE
PHASE 1
PHASE 2
PHASE 3
PHASE Effect
SINCE CL
59.00
111.00
•
•
59.00
59.00
59.00
59.00
59.00
59.00
59.00
59.00
59.00
59.00
111.00
•
•
59.00
59.00
59.00
59.00
59.00
59.00
59.00
59.00
59.00
59.00
111.00
•
•
59.00
59.00
59.00
59.00
59.00
59.00
59.00
59.00
59.00
412.54
,
•
3.63
1.70
2.71
0.65
1.53
6.73
1.09
0.44
.
0.67
92.29
•
•
2.94
1.16
1.72
0.86
1.96
1.84
0.97
0.77
.
0.93
37.93
•
,
5.82
1.95
4.64
0.56
3.04
12.43
1.04
0.33
•
0.62
6.02
•
0.74
1.29
0.53
1.00
-0.44
0.43
1.91
0.08
-0.82
•
-0.40
4.52
•
0.43
1.08
0.15
0.54
-0.15
0.67
0.61
-0.03
-0.26
B
-0.08
3.64
.
0.90
1.76
0.67
1.53
-0.58
1.11
2.52
0.04
-1.09
•
-0.48
0.29
•
0.56
•
0.37
0.48
0.31
0.38
0.59
0.50
0.62
m
0.18
0.22
•
0.43
•
0.28
0.36
0.26
0.28
0.44
0.38
0.46
a
0.13
0.38
•
0.74
.
0.49
0.62
0.42
0.50
0.77
0.66
0.81
.
0.23
Observed
Significance
Level
.0000
m
.0828
.
.1599
.0415
.1633
.2676
.0020
.8716
.1952
.0132
.0256
.0000
!3117
_
.6005
.1350
.5547
.0210
.1701
.9360
.5708
.4198
.5682
.0000
•
.1456
•
.1750
.0170
.1759
.0300
.0018
.9536
.1840
.0115
.0411
Volume II - Page C-6
-------
APPENDIX D
LABORATORY CONTROL CHARTS FOR
QUALITY CONTROL SAMPLES
-------
I
(D
-------
o
t->
i
n>
o>
vQ
(D
D
I
PERCENT
150
140
130
120
1 1 O
1 U
100
on
80
70
60
50
RECOVERY
—
—
0
D ^ ©
8
i i i
CONTROL LIMIT
WARNING LIMIT
WARNING LIMIT
CONTROL LIMIT
BATCH CODE
Figure D-2. Performance - control chart of individual spiked sample recovery data,
wipe samples.
-------
<
o
c
M
M
Q>
n>
o
i
I *
ww
PERCENT RECOVERY
150
140
130
120
1 10
100
90
80
70
60
50
—
0
Y-""'y "" ----- .. r>< -
o . o °
O "~n""
-ft ° 08 P o
o 0 o o
0
Q
0
1 1 1 1 1 II
01 23456789101112
CONTROL LIMIT
WARNING LIMIT
WARNING LIMIT
CONTROL LIMIT
BATCH CODE
Figure D-3. Performance - control chart of individual spiked sample recovery data,
soil samples.
-------
PERCENT RECOVERY
c
3
(0
H
M
1
V
PI
fD
a
i
1 OU
160
140
120
100
80
60
40
20
O
1 00
O
X, O
f) A O TN
— A ^ ^ O O A O
l_ (^) v_/ ^- ' A /"\ ^p •— '
O O 0
- n*®^Q* |t(**n^ ltmDi*y •**
-PI Dr-]Q^[-i^™g|D Qnn ^ n
a
0
— i — ' — i— J — i — i — i — i — i — i- iii i - i i m ?
2 3 4 5 6 7 8 9 10111213141516171819202122232425
BATCH CODE
O SRM 1646
SRM 2704
D SRM 1646 DF=1
CONTROL LIMIT
WARNING LIMIT
Pigure D-4. Performance - control chart of individual blind reference
material recovery data, vacuum cassette samples.
-------
PERCENT RECOVERY
o
c
3
(D
H
H
1
Ol
^o
fl)
i
Ul
1 OU
160
140
1 20
100
ao
60
40
-
—
^^
* * •
S
~
—
1 i i
O
D
\
SRM 1646
SRM 2704
SRM 1646 DF= 1
CONTROL LIMIT
WARNING LIMIT
BATCH CODE
Figure D-5. Performance - control chart of individual blind reference
material recovery data, wipe samples.
-------
PERCENT RECOVERY
o
C
(D
M
H
1
T3
01
vQ
(D
D
CM
1 OU
160
140
1 90
100
on
60
40
C
-
O
—
j. O
— O ..../TV .
u D OS *
@ — A i — i ' — ' A
LJ
—
—
i i i i i i i i i i i
) 123 456 78 9 10 111
C
<
r
""!
2
O SRM 1646
SRM 2704
D SRM 1646 DF=1
CONTROL LIMIT
WARNING LIMIT
BATCH CODE
Figure D-6. Performance - control chart of individual blind reference
material recovery data, soil samples.
-------
(D
H
Q)
(D
D
RANGE PERCENT
50
40
30
10
CDi
CONTROL LIMIT
WARNING LIMIT
0123456789 10111213141516171819202122232425
BATCH CODE
Figure D-7. Performance - control chart of individual spiked and spiked duplicat*
recovery data, vacuum cassette samples.
-------
c
3
ID
I
T3
0)
D
I
00
RANGE F
50
40
30
20
10
n
C
3ERCENT
—
—
O
o
o 0
1 (T) ft\
—*&— (y "
' 1 2 3 A
BATCH CODE
[
CONTROL LIMIT
WARNING LIMIT
Figure D-8. Performance - control chart of individual spiked and spiked duplicate
recovery data, wipe samples.
-------
RANGE PERCENT
vJU
40
|
I
H
H
on
/:U
1
O
vo 10
n
O
0 o
O
0 0
0 °
o 0
- 0 ~ 0 0
o o
1 Q — i , m (T> i i rn Q i m
01 23456789 10 11 12
BATCH CODE
CONTROL LIMIT
WARNING LIMIT
Figure D-9. Performance - control chart of individual spiked and spiked duplicate
recovery data, soil samples.
------- |